SMALL-ANGLE

SCATTERING OF X-RAYS

ANDRE GUINIER
Professor, Universite de Paris (France)

GERARD FOURNET
Lecturer, Ecol•~ Superieure de
Physique et Chimie, Paris

Translation by
CHRISTOPHER B. WALKER
Institute for the Study of Metals
University of Chicago

New York JOHN WILEY S SONS, Inc.

London CHAPMAN S HALL Ltd.
1955
STRUCTURE OF MATTER SERIES
MARIA GO EPPERT MAYER
Aduisor!I Editor

SMALL-ANGLE
SCATTER! NG OF X-RAVS
COPYRIGHT @ 1955
BY
JOHN WILEY & SONS, INC.

All Rights Reserved

Thia book ur any part thereof muat not
be reproduced in any form without
the written permiss.:On of the publiaher.

· Library of Congress Catalog Card Number: 55-9772

PRINTED IN THE UNITED STATES OF AMERICA
PREFACE

X-ray diffraction was first utilized in establishing the atomic structure

of crystals. Later the technique of X-ray diffraction found other appli-
cations, however, and branched off from pure crystallography, extending
to studies of imperfections in crystals, sizes of crystallites, and even to
studies of the atomic structure of amorphous bodies. These fields of
application of X-rays were made possible by further developments in
the theory of the diffraction of X-rays by matter and also by improve-
ments in experimental methods.
The small-angle scattering of X-rays is one of these fields that has
been rather recently opened. Although the first observations were made
in 1930 (295] particular attention has been given to this field only since
the late l 930's. At the present time a large, ever-increasing number of
laboratories are interested in small-angle scattering, as is shown by the
number of references compiled in the bibliography of this book.
For these reasons it seemed worth while to us to devote a monograph
to this specific branch of X-ray diffraction. In fact, the theories that
are used in this field are generally not discussed in textbooks on X-rays.
They are quite distinct from the concepts that are customarily associated
with X-ray diffraction; almost no use of Bragg's law will be made in this
book, except to point out that the habit, so natural to crystallographers,
of interpreting every detail in a diffraction pattern in terms of lattice
distances sl1ould be discarded. The experimental aspect also is different;
small-angle scattering in general cannot be studied with the usual
apparatus of a crystallography laboratory; special cameras and some-
times special tubes are required.
Since the late 1930's many theoretical works have appeared in this
field; starting from different points of view, these have occasionally
arrived at different, but non-contradictory, results. In a parallel man-
ner, apparatus based on quite varied principles have been used in ex-
perimental methods. We believed that it was now time to collect and
evaluate the results that have been obtained from the different ap-
proaches. Our object has been to make the new research in this field
more rapid and more efficient. Finally, we have also tried to evaluate
the different attempts at applications in order to specify those which
are the most fruitful.
v
vi PREFACE

The plan of this book is as follows: in a first, short chapter we present

the phenomenon of small-angle scattering and investigate its physical
significance.
The second chapter is devoted to a discussion of the progress realized
in the theoretical study of small-angle scattering. We have tried to
show the problems that have actually been solved and the limitations
that now appear to us as difficult to overcome.
In a third chapter we discuss the experimental methods that have been
employed, trying not to treat all the details but giving the general prin-
ciples that should be satisfied in a small-angle scattering system. Evi-
dently these techniques will be similar whether the objective is the study
of continuous scattering or the study of crystalline diffraction patterns.
Thus it will be seen that problems are mentioned in this section which
are not considered from a theoretical point of view in the second chapter.
The fourth chapter is devoted to the problem of the interpretation of
the experimental results and includes several examples which demonstrate
the validity of the theoretical results.
In a fifth chapter we compare the results of small-angle X-ray scat-
tering with the results of other physical methods for measuring particle
sizes, such as interpretations of Debye-Scherrer line widths and measure-
ments with the electron microscope.
The sixth and last chapter is devoted to a discussion of the applications
of small-angle X-ray scattering. These are found in a number of diverse
fields, such as chemistry, biology, and metallurgy. Some applications
are of technical interest, as, for example, the study and testing of cata-
lysts. Others are of interest to theoretical physics, as, for example, the
structure of liquid helium below the X-point.
Although the object of the first chapters of this book is to present all
the theoretical and experimental data necessary to the specialist in X-ray
diffraction, the last chapter has been written without use of mathematics
and without details of X-ray techniques so that it can be read without
difficulty by a non-specialist. Our object has been to present the differ-
ent types of problems that can be studied by small-angle scattering and
the results that have actually been obtained up to the present. Thus
a chemist, biologist, or metallurgist should be able to decide from this
whether or not any given problem can be approached effectively by
means of X-rays.
In this monograph we have tried more to give a logical, ordered pres-
entation of this subject than to give a complete compilation of all the
published papers. Any gaps can be filled by the reader by referring to
the bibliography. Let us point out that several general articles on small-
angle scattering have now appeared: the article by Hosemann [84] and
PREFACE vii

another by Porod [137] are particularly noted. As a result we have

been able to shorten our discussion on several points, since the reader
can find the complete development of these ideas in the works cited.
When reference is made in the text to a formula in the same chapter,
the formula is denoted by a single number, as, for example, 36. When
the formula has been developed in a different chapter, it is denoted by a
double number, such as 2.36 (equation 36 of Chapter 2).
If a bibliographic reference appears as numbers within brackets, [ ],
the reference will be found in the general bibliography at the end of the
book. References appearing as "Author (year)" are tabulated in a
special bibliography at the end of each chapter.
Our sincere thanks are extended to Dr. R. S. Bear, Dr. W.W. Beeman,
Dr. J. W. M. DuMond, Dr. A. N. J. Heyn, Dr. R. A. Van Nordstrand,
and Dr. C. B. Walker for having made available to us papers which are
as yet unpublished and drawings or original photographs which they have
authorized us to reproduce here. Permission has been given to repro-
duce a number of illustrations from technical journals, for which we wish
to thank both the authors and the publishers.
We are particularly grateful to Professor P. P. Ewald, who encouraged
us to publish this book, and to Professor W.W. Beeman, whose criticism
and advice were very helpful in the final editing of our manuscript.
Finally we want to thank Dr. C. B. Walker for the careful translation
which has made the original manuscript more accessible to many readers.
A. GurnrER
G. FOURNET
Paris, France
August, 1955
CONTENTS
1. ORIGIN AND CHARACTERISTICS OF SMALL-ANGLE X-RAY
SCATTERING

2. GENERAL THEORY. 5
2.1. SCATTERING PRODUCED BY A SINGLE PARTICLE 5
2.1.1. Fixed Particle . 5
2.1.1.1. Centrosymmetric particle 6
2.1.2. Moving Particle 7
2.1.2.l. Centrosymmetric particle 8
2.1.2.2. Spherically symmetric particle 10
2.1.2.3. Calculation of the average intensity 10
2.1.2.4. The characteristic function of the particle -yo(r) 12
2.1.2.5. General properties of F 2 (h) . 16
2.1.2.6. A tabulation of the average intensity distributions for particles of
different shapes 19
2.1.2.7. Particle with preferred orientations 23
2.1.3. Concept of a Radius of Gyration of a Particle 24
2.1.3.1. Moving particle 24
2.1.4. Spherically Symmetric Particle 28
2.1.5. The Distribution of Scattering from a Fixed Particle 28
2.2. SCATTERING PRODUCED BY A GROUP OF IDENTICAL PARTICLES 30
2.2.1. General Results for Fluids 30
2.2.l.1. Basic hypotheses . 30
2.2.1.2. Consequences of hypothesis H1 31
2.2.1.3. Consequences of hypothesis H2 32
2.2.1.4. General expression for the scattered intensity 33
2.2.2. Widely Separated Particles 35
2.2.2.1. Expression for the scattered intensity 36
2.2.2.2. Remarks on the Babinet principle of reciprocity 38
2.2.3. Influence of the Closer Packing of Particles 40
2.2.3.1. General considerations . 40
2.2.3.2. Scattered intensity and the equation of state 42
2.2.3.3. Limiting value for the intensity scattered at very small angles 46
2.2.3.4. Thermodynamic expression for the intensity 48
2.2.3.5. Fluids and crystals 52
2.2.3.6. Secondary maxima 54
2.2.3.7. Remarks on Fourier transformations . 57
2.2.4. Particles Unrestricted by Hypotheses H1 and H2 60
2.3. SCATTERING BY GROUPS COMPOSED OF SEVERAJ, TYPES OF PARTICLES 65
2.3.1. General Theory 65
2.3.2. Widely Separated Particles 66
2.3.3. Influence of the Closer Packing of Particles 67
2.4. GENERAL CASE 70
2.4.1. Limiting Value of the Scattered Intensity at Very Small Angles 71
ix
x CONTENTS

2.4.2. Calculation of the Scattered Intensity as a Function of p(X) 75

2.4.3. Matter of Uniform Density and Random Distribution 78

3. EXPERIMENT AL EQUIPMENT 83
3.1. GENERAL CONSIDERATIONS . 83
3.1.l. Operational Principles 83
3.1.2. Influence of the Monochromatization of the Primary lladiation 85
3.2. SYSTEMS EQUIPPED WITH COLLIMATORS 86
3.2.1. Collimator Formed by Two Slits . 86
3.2.1.1. Calculation of the optimum collimator 89
3.2.2. Collimator with Circular Openings 91
3.2.3. Collimator with Slits of Finite Height for the Study of Circularly
Symmetrical Diffraction Patterns . 91
3.3. SYSTEMS USING MONOCHROMATIC RADIATION 94
3.3.1. Source of Monochromatic lladiation 94
3.3.2. Balanced Filters 95
3.3.3. Monochromatization by Crystalline Diffraction . 96
3.3.3.1. Plane monochromator . 96
3.3.3.2. Bent crystal monochromator 100
3.3.3.3. Combination of two bent crystal monochromators . 102
3.3.3.4. Monochromator with a point focus 104
3.3.3.5. Double monochromator with plane crystals 109
3.3.4. Measurement of the Total Scattered Intensity 110
3.4. METHODS OF CORRECTION OF EXPERIMENTAL SCATTERING CURVES 111
3.4.1. Correction for the Effect of Beam Width 112
3.4.2. Correction for the Effect of Beam Height 114
3.4.2.1. Slit correction for infinite height 116
3.4.2.2. Case of a beam of arbitrary height 118
3.5. CONSTRUCTION OF LOW-ANGLE SCATTERING SYSTEMS 120
3.5.1. Slit Construction 120
3.5.2. Stopping the Direct Beam 121
3.5.3. Absolute Measurements 121
3.5.4. Vacuum Apparatus . 123

4. METHODS OF INTERPRETATION OF EXPERIMENTAL RESULTS 126

4.1. IDENTICAL PARTICLES . 126
4.1.1. Widely Separated, Identical Particles . 126
4.1.1.1. Equal probability of all orientations . 126
4.1.1.2. Identical particles with a definite orientation 134
4.1.2. Dense Groups of Identical Particles 135
4.1.2.1. Analysis of the scattering curve . 135
4.1.2.2. Interpretation of a maximum in a scattering curve 140
4.1.2.2.1. Interpretation in terms of an average distance . 141
4.1.2.2.2. Interpretation in terms of an average t•olume 145
4.1.2.2.3. Interpretation by means of an interparticle interference function 146
4.1.2.2.4. Predictions of the correct theory . 146
4.1.2.3. Conclusions 147
4.2. GROUPS OF NON-IDENTICAL PARTICLES 148
4.2.1. Determination of the Average Radius of Gyration for the Group of
Particles 149
CONTENTS xi

4.2.2. Attempts at Determining the Statistical Distribution of the Particles 151

4.2.3. Determination of the Specific Surface 156
5. COMPARISOK OF THE RESULTS FROM SMALL-ANGLE SCAT-
TERI.'.'iG WITH THE RESULTS OF OTHER METHODS OF MEAS-
UREMENT OF SMALL PARTICLES 161
5.1. COMPARISON WITH THE EI,ECTRON MICROSCOPE 161
5.2. COMPARISON WITH THE METHOD OF DEBYE-SCHERRER LINE WIDTHS 163
6. THE APPLICATIONS OF THE SMALL-ANGLE SCATTERING OF
X-RAYS 167
6.1. LARGE MOLECULES 167
6.1.1. Dilute Solutions 167
6.1.2. Concentrated Solutions 171
6.2. HIGH POLYMERS 176
6.2.1. Study of Solutions 176
6.2.2. Study of Fibers . 177
6.2.3. Ordered Arrangements of Micelles 183
6.3. FINELY DISPERSED SOLIDS. CATALYSTS 187
6.3.1. Carbons 188
6.3.1.1. Practical study of carbon blacks 189
6.3.1.2. Structure of different varieties of carbon 190
6.3.2. Catalysts 192
6.3.3. Colloidal Solutions 194
6.4. SUBMICROSCOPIC HETEROGENEITIES IN SOLIDS. APPLICATIONS TO PHYS-
ICAL METALLURGY 195
6.4.1. Heterogeneities in Pure Metals 195
6.4.2. Heterogeneities in Solid Solutions . 197
6.4.2.1. Equilibrium solid solutions 197
6.4.2.2. Supersaturated solid solutions: Age-hardening . 199
6.4.2.3. Structural characteristics directly related to the small-angle scat-
tering 200
6.4.3. Examples of Small-Angle Scattering by Age-Hardening Alloys 203
6.4.3.1. Aluminum-silver alloy: First stage of hardening 203
6.4.3.2. Aluminum-silver alloy: Second stage of hardening . 208
6.4.3.3. Aluminum-copper alloy 211
6.5. ABSOLl:TE MEASL'REMENTS OF THE !:'!TENSITY OF SCATT~JRil\G AT ZERO
ANGLK MEASUREMENTS OF THE COMPRESSIBILITY OF A FLUID 213
BIBLIOGRAPHY 217
AUTHOR IXDEX TO BIBLIOGHAPHY . 261
AUTHOR INDEX TO n;xT 265
SUBJECT INDEX 267
1. ORIGIN AND CHARACTERISTICS OF
SMALL-ANGLE X-RAY SCATTERING
The fundamental relation describing the diffraction of X-rays by
crystalline matter, ;. = 2d sin(), shows that the angle of diffraction, (},
varies inversely with the separation of the diffracting lattice planes. In
ordinary crystals, particularly those of inorganic matter, the majority of
the observed lattice spacings are of the same order of magnitude as the
X-ray wavelengths generally e~ployed, so that the angles () are usually
rather large. This advantageous condition has had important con-
sequences, both in the discovery of the phenomenon of X-ray diffraction
and in its employment in studies of crystal i;tructures.
The study of small-angle X-ray diffraction was introduced when it
became desirable to detect large lattice spacings, of the order of tens or
hundreds of interatomic distances. These spacings are found in some
particular minerals and in certain complex molecules, such as the high
polymers or proteins. In studies of the structures of macromolecular
crystals the X-ray diffraction patterns must be extended to include very
small angles. For example, with Cu Krx. radiation and a spacing of 100 A
the diffraction angle (} is equal to 0.45°, and, with a period of 1000 A,
() equals 0.045° or 2'. This illustrates the importance of small-angle
scattering techniques in such fields as biochemistry, for example.
One might consider using longer-wavelength X-rays to obtain larger
diffraction angles for a given lattice spacing. This is not generally
feasible, however, since the long-wavelength X-rays are absorbed to a
very great extent in matter, which not only complicates the necessary
diffraction apparatus and the means of detection of the X-rays but also
considerably diminishes the intensity of the diffracted beam. For these
practical reasons we must recognize a gap in the spectrum of useful
electromagnetic radiation extending from wavelengths of the order of 2 A
up to those of the remote ultraviolet.
In studying crystals with large periodicities only the operational
technique is different, since the interpretation of the patterns is based on
the same principles as the usual structure determinations. The difficulties
encountered are greater, however, as a result of the complexity of the
unit cell and the imperfection of the crystals. One can intuitively picture
"perfect" crystals as being formed only by the grouping of small numbers
l
2 SMALL-ANGLE SCATTERING OF X-RAYS

of atoms bound by strong forces. In molecular and macromolecular

crystals the degree of perfection is much less; only rarely is the theory of
diffraction by perfect crystals a good approximation in small-angle
diffraction phenomena. In this domain the theory of diffraction by
imperfect crystals assumes particular importance, as is illustrated by the
correlation of small-angle diffraction and the diffraction by imperfect
crystals in an X-ray study of high polymers by Hosemann [84). Since
diffraction by imperfect crystals is a theoretical problem not confined to
small-angle scattering and one that has been well discussed elsewhere,
we shall not examine it further in this monograph.
If a sample has a non-periodic structure or if its lattice has been
sufficiently perturbed, the diffraction patterns are not limited to spots or
lines hut contain more or less extended regions of scattering. Let us
examine schematically the origin of this scattering at small angles.
It is well known that the diffraction pattern of a sample can he simply
described in terms of a reciprocal, or Fourier, space. If we designate by
p(X) the electronic density of the diffracting body at a point defined by the
vector X, then A (h), the transform of p(x) at the point defined by the vector
h in reciprocal space, is given by
A(h) = J
p(x)e-•h•J: dx (l)

The theory of X-ray· diffraction is based on the fact that A (h) represents
the amplitude of the diffracted radiation when h is defined as
h = (2rr/A)(S - S0 )
where ). is the wavelength of the radiation and s0 and s are unit vectors in
the direction of the incident and diffracted rays, respectively. The
magnitude of h is then equal to (4rr sin 0)/)., where 20 is the scattering
angle (the angle between the incident and scattered rays). Thus scattering
at very small angles corresponds to small values of h.
Equation l can be interpreted as follows: the scattered intensity
observed for conditions corresponding to a certain value of h is equal to
the square of the value of A (h), where A (h) is the component corresponding
to h in the development of p(X) in a Fourier series. For small values of h,
that is, at very small angles, the terms in p(x) that primarily control the
magnitude of A(h) are those that show a periodicity of x = 2rr/h, a
periodicity large with respect to the X-ray wavelength. These general
considerations show again that diffraction at very small angles (less than a
few degrees) gives information concerning the structure of matter on a
scale that is large compared to the X-ray wavelength.
It has been experimentally observed that certain samples cause an
intense, continuous scattering below angles of the order of 2° without
GENERAL CHARACTERISTICS 3

producing the usual type of diffraction effects found on ordinary X-ray

patterns. This was first ob::;erved by Krishnamurti [295] and Warren
[171] for certain varieties of finely divided carbons, carbon blacks, and
various other substances, all having in common the characteristic of being
present as fine particles of submicroscopic size. Actually it was later
recognized that the continuous scattering in the neighborhood of the direct
beam is related to the existence of matter in the form of small particles,
or, more generally, to the existence of heterogeneities in the matter, these
heterogeneities having dimensions from several tens to several hundred
times the X-ray wavelength. This offers another example of the general
relation previously cited.
It is relatively easy to describe qualitatively the central scattering due
to the presence of small particles. This is analogous to the well-known
phenomenon of optical diffraction, where a halo is produced by the passage
of a light ray in a powder whose grain dimensions are of the order of a
hundred times the wavelength of the light.
Let us consider a particle bathed in a beam of X-rays; all the electrons
are then sources of scattered waves. When the scattering direction is
the same as that of the incident ray, these scattered rays are all in phase,
and, as the scattering angle increases, the difference in phase between the
various scattered waves also increases. The amplitude of the resultant
scattered wave then decreases with increasing angle because of increasing
destructive interference; it becomes zero when there are as many waves
with phases between 0 and TT as there are between 7T and 277. This will
occur for a scattering angle of the order of 20 = A(D, D being the
"average dimension" of the particle, demonstrating how the study of the
continuous central scattering offers a method for obtaining particle
dimensions.
This method is applicable only for particles whose sizes lie within
certain limits. If D is too large the scattering is limited to angles so
small as to be inaccessible to experiment, and if D is too small, of the
order of several wavelengths, the scattering is widely spread but too weak
to be observable.
These rough qualitative conHiderations can be made more precise. To
show exactly on which factors the small-angle scattering depends, let us
consider a small particle that has been cut from a section of matter of
electronic density p(x). Let us define a '"form factor" of this particle,
s(x) (Ewald (1940) ), that has the value I whm the vector x lies within the
particle and the Yalue 0 when x lies outside the particle. The amplitude
of radiation scattered by this particle, as found from equation 1, is then

A 1 (h) = Jp(x) s(x)e -ih·z dx (2)

SMALL-ANGLE SCATTERING OF X-RAYS

There is a general theorem related to the operation of "folding" in the

theory of Fourier transformations stating that, if A(h) and S(h) are
respectively the Fourier transforms of p(x) and s(x), then

A 1 (h) = f
A(y)S(h - y) dy (3)

where y is a variable of integration.

Given the dimensions of the region in which s(x) is different from zero,
its transform, S(h), is fully determined, and, if the particle has dimensions
of several tens to several hundreds of atomic diameters, S(h) will be
different from zero only for very small values of h.
Let us consider now the function A(h). If we first assume that the
sample is of constant electronic density, p(x) = k, the transform A(h)
acts as a Dirac delta-function,1 being zero everywhere except at h = 0,
where it is infinite. For the more general case of a homogeneous body
whose electronic density shows periodicities only on an atomic or molecular
scale, the transform A(h) shows a large number of peaks. However, all
these peaks except the one for h = 0 are produced for values of h well
outside the domain in which S(h) has a non-zero value.
Then, since A(y) is essentially a Dirac delta-function about y = 0, it
may be predicted that around the origin of the reciprocal space the
amplitude A 1 (h) is simply proportional to S(h), the function p(x) not
intervening. The scattering around the center is thus practically independent
of the "short-range order" of the ato~, depending only on the exterior form
and dimensions of the particle.
Small-angle scattering thus appears as a means of studying the dimen-
sions of colloidal particles, and it is in this direction that the technique has
been generally exploited. It was quickly realized, however, that the
assumptions adopted in the first theoretical approaches (widely separated,
identical particles) were not being satisfied in the constitution of real
samples. Interpretation of the scattering then demanded that the
theory be generalized to take into account the diversity of particles sizes
and the effect of the closer packing of the particles. Also, without
speaking of particles, the possibility should be considered of obtaining an
expression for the intensity scattered near the center in terms of the
electronic density at all points of the sample. The theoretical approaches
to these and other problems are discussed in the following chapter.

REFERENCE FOR CHAPTER

Ewald, P. P. (1940), Proc. Phys. Soc. (London), 52, 167.

1 The Dime delta-function 6(x) is zero for x # 0, infinite for x = 0, and

J6(x) dx = l.
2. GENERAL TIIEORY
In this study we shall consider only coherent scattering, neglecting
Compton scattering which is always small at small angles. We shall
discuss only the single-scattering process, disregarding the phenomenon of
multiple scattering [31), [33).

Incident rays

Fig. 1. Diffraction by a single particle.

We shall assume always that the transverse dimensions of the X-ray

beam are large enough so that a large number of particles are irradiated,
yet sufficiently small compared to the sample-receiver distances so that
the beam can be likened to a single ray in the macrogeometry of the
experimental apparatus.

2.1. SCATIERING PRODUCED BY A SINGLE PARTICLE

2.1.1. FIXED PARTICLE
The classical formula in the theory of X-ray diffraction gives the ampli-
tude of radiation scattered by the point Mk (Fig. I) (of scattering factor
fk) in the direction defined by the unit vector s as

(1)
where A. designates the amplitude scattered by one electron for the same
conditions; 0, an arbitrary origin serving to describe the path differences
between different rays; and s0 , the unit vector defining the direction of
the incident radiation. Let us designate by h the vector (2TT/A)(s - 80 ).
If 20 represents the angle of scattering, .zS_ ss 0 , the magnitude of h is
h = (47r sin O)/A.
5
6 SMALL-ANGLE SCATTERING OF X-RAYS

The total amplitude of radiation scattered by a particle is then

A(h) = _LAk = A.(h)_Ljke-ih·OMk (2)
k k
and the scattered intensity, the product of the amplitude A and its
complex conjugate A*, is
J(h) = A. 2 (h)_L_LJd1 cos (h · M,)'tl1) (3)
k j
The intensity scattered by one electron
J8 (h) = A 0 2 (h) = 7.90 x I0- 26 J 0p- 2 X
1 + cos 2 20 (4)
2
is a function only of(), that is, of the magnitude of h; 1 0 represents the
intensity of the incident beam, and p is the distance between the particle
and receiver, expressed in centimeters.

2.1.1.1. Centrosymmetric Particle

If the particle possesses a center of symmetry, the expression for the
diffracted amplitude can be simplified, for, if the origin is taken at the
center of symmetry, then to each vector OMk there corresponds another
vector -OMk. Therefore
A(h) = _LAk = A.(h)L:f,. cos (h · OMk)
k k
We shall define the structure factor of the particle as the ratio of the
total scattered amplitude to the amplitude of radiation scattered by one
electron under the same conditions:
2 Ak(h)
F(h)--k_ _ (5)
- A,(h)
The scattered intensity is then
J(h) = I.(h)[Lfk cos (h · 0Mk)] 2 = I.(h)F 2 (h) (6)
k
The term "point Mk" has been used to refer to and define the structure
of a particle. In considering a large particle the basic element in its
description is the atom; the point Mk then refers to the "center of the
kth atom," and the scattering factor fk is the scattering factor of this kth
atom. As f k varies with the scattering angle, it should be denoted by
fk(h). However, in the angular range where the structure factor of a
large particle is different from zero, fk(h) can be effectively considered as a
constant, equal to fk(O). For example, the structure factor of a molecule
of human hemoglobin is effectively zero for all angles such that
h > 0.15, and in this range the variation of the scattering factor of a
carbon atom is less than 0.4 per cent.
GENERAL THEORY 7

When a small particle (an atom, for example) is being considered, the
point Mk will refer to a volume element, small even on the angstrom
scale, surrounding the point Mk. The scattering factor A then equals
Pk dvk, where Pk is the electronic density of the particle in the neighborhood
of the point Mk, and dvk is the volume element considered.
In general we will find it convenient to describe the structure of a
particle in terms of elements which are small enough so that the scattering
factors of these elements can be considered as constants, independent of
the angle of scattering, over the range in which the structure factor of the
particle under consideration is different from zero.

2.1.2. MOVING PARTICLE

In the majority of low-angle scattering investigations, such as exami-
nations of solutions, suspensions, and emulsions, the particles are capable
of motion. This motion can always be described as the sum of a trans-
lation and a rotation. A translation, defined by a vector V, introduces
the multiplicative factor e-ih·V in the expression for the scattered
amplitude, but this has no effect on the scattered intensity. Only
rotations intervene in the calculation of an average intensity.
When the probabilities of different orientations are defined, we can
obtain from equations 3 or 6 the expression for the observed average
intensity

this relation defining the average of the square of the structure factor.
-- --2
There would be a temptation to describe F 2(h) as equal to F(h) , the
square of the average of the structure factor. However, in order that the
average of a product, ah, be equal to the product of the averages of a and
b, it is necessary that the variables be completely independent, that is,
that knowledge of the value of a in no way modifies the probabilities of the
different values of b. This limitation is not met by the structure factors,
since a = b = F(h). The only general case in which F 2 and F2 are
equal is that pertaining to spherically symmetric particles, for then a
rotation of the particle around its center does not modify the distribution
of scattering centers and consequently leaves F(h) unchanged. For this
case one finds

In this section, the discussion is restricted almost entirely to considering

all particle orientations as equally probable; a treatment of the more
general case will be found at the end of the chapter. When this assump-
tion is made, the only mathematical problem is one of calculating the
8 SMALL-ANGLE SCATTERING OF X-RAYS

average of the function, cos (h • r), as the vector r, of magnitude r, takes

all orientations with equal probability. To calculate this average, let us
define the angle between the vectors h and r as the angle rp, a variable
with limits of 0 and Tr radians. The probability that this angle is con-
tained between the values cp and cp +
dcp is equal to t sin cp dcp. The
average of the phase function, cos (h. r), is then

l"
0
sin cp
cos (hr cos rp) - - drp
2

= Jof"'2 cos (hr cos cp) sin cp dcp

1 f"'2
= - hr Jo cos (hr cos cp) d(hr cos rp)

= - -1 Lo cosudu
hr hr

leading to the classic result

sin hr
cos(h·r)=~ (7)

The result depends only on the magnitude of h; the distribution of

scattered intensity thus contains an axis of revolution coinciding with
the incident beam.
Equation 3 then resolves into the expression for F2(h) 1 expressed by
Debye (1915),
_ sin(h!MV I>
F 2 (h) = 2.2.fds J<-i (8)
k; hlM.tM1 1
2.1.2.l. Centrosymmetrlc Particle
When a center of symmetry exists, application of equation 7 to equation
5 results in a simple expression for the average of the structure factor:
- I
_ "" / sin (h OM:k ll (9)
F(h)- fJk hiOMkl
Generalization of this equation to include particles with a continuous
distribution of scattering points leads to the following expression:
- f I
sin (h OMk !l
F(h) = J/(Mk) I
h OMk! dvk
1 The notation! (h) will be used when the function depends only on the magnitude

of h; the notationf(h) will demonstrate dependence of the function on both magnitude

and direction of h.
GENERAL THEORY 9

The coefficient of the term sin hr/hr in this integral, obtained by con-
sidering the ensemble of points defined by OMk \ = r, is I
i,
r+dr
p(Mk) dvk = ,O(r)411r 2 dr

this defining the function ,O(r). The generalization of equation 9 then

takes the form
-- f
F(h) =
o
00 sin hr
,O(r) --47rr2 dr
hr
(10)

We see thus that the average of the structure factor is uniquely deter-
mined by the distribution of scattering centers as a function of their
distance from the center of the particle.
Equation 3 shows that the parameters possessing physical significance
in the expression for the intensity are the distances \ M~1 between I
each of the pairs of scattering centers. Nevertheless, for convenience of
calculation one might on occasion prefer an expression for the intensity in
I I I
which the distances \ OMk and OM; are the essential parameters,
where 0 designates the center of symmetry of the particle. Fournet [48)
has shown this to be

F 2(h)=

' ' { f, f ) 11(2 + l)

00 J2 1 (h\OM
"+ 11
l)J 2 7>+
1
1•
(hi OM
;
I> P (cos <I> )
}
p~O :p
/c
f; t k i hv' I OMk 11 OM; I 2" kJ

(11)

where Pm represents the Legendre polynomial of order m, and <l>ki• the

angle "i:,_MkOM 1• [The Legendre polynomial of order m, P m(x), can be
described as the coefficient of the term ym in the expansion of the function
(1 - 2yx +
y 2 )-li2.] In certain cases this equation can be employed
more simply than equation 8 (Fournet [48]).
Fournet has employed equation 11 to illustrate the difference between
-- --2
F 2 (h) and F(h) . If we ~aluate the sum of terms for p = 0,

then, on transforming the Bessel functions into sine functions with the
relation J 112 (x) = v' (2/11x) sin x, we find that the sum of these terms is
equal to
IO SMALL-ANGLE SCATTERING OF X-RAYS

which is the square of the average of the structure factor. Thus we can
write

F 2(h) = F(h) 2

+ LL {!kf; f 7T(2p + !) J 2P+l/2(h IOMk ll J 2v+I/2(h IOM; ll p 2v (cos tl>k;)}

k j p=l hV IOMk I IOM, I
(12)
2.1.2.2. Spherically Symmetric Particle
A particularly important case to be considered-is that of the spherically
symmetric particle. The electronic density at any point depends only on
the distance r of this point from the center of the particle and can thus
be denoted by p(r).
The structure factor is then obtained from equation IO, replacing
p(r) by p(r):

F(h) =
o
oo
i sin hr
p(r) - - 47Tr 2 dr
hr
(13)

For this particular case, rotation of the particle does not modify the
amplitude of scattered radiation, leading to the relation

(14)

2.1.2.3. Calculation of the Average Intensity

The calculation of the average intensity can be made by several methods.
(a) Analytical Method: The intensity scattered by the particle in an
arbitrary position is calculated (see equation 3). Then the expression
is averaged, taking into account the different orientations, in a manner
similar to that employed by Guinier ([65], p. 195) and Fournet ([48],
p. 45). This method is particularly simple when applied to a spherically
symmetric particle; equations rn and 14 can then be used directly.
(b) Geometrical 1V!ethod: Kratky and Porod [108]. Equation 8 can be
generalized intuitively to allow the consideration of a particle of volume
V, defined by an electronic density p(Mk); the resulting expression is

- 2
F (h) =
i i . .·
v v
p(.Mk)p(.M;)
h Mr)I;
II
sin (h M1'J;
I ll dvk dv; (15)

Let us consider the coefficient of sin hr/hr in the integral, assuming for
the moment that p is a constant. This coefficient is obtained by con-
I
sidering the ensemble of terms where MkM; = r. The number of I
electrons at distances between r and r +
dr from a volume element dvk
of the particle is simply p{Vk(r +
dr) - Vk(r)}, in which Vk(r) designates
GENERAL THEORY 11

the part of the volume of the particle situated at a distance smaller than
or equal to r from dvk. When we now consider all possible positions of
dvk, we can introduce a function, p(r), defined by the relation

( p{Vk(r + dr) - Vk(r)}p dvk = p 2p(r) dr (16)

Jv
The average of the square of the structure factor can then be expressed as
-- = f
F 2 (h) p2
o
00 sin hr
p(r) - - dr
hr
(17)

In order to determine the physical significance of p(r), let us describe

the volume element dv 1 of equation 15 in a system of spherical coordinates
centered on the point Mk, for which dv1 = r 2 dw dr. Equation 15 then
becomes

(18)

The point M 1 in the integral with respect to dw dr is any point in the

particle situated at a distance r from the point 1lfk, where
OM; - OMk = r with Ir I = r
and the integral extends only over the volume V of the particle. This
integral can be extended over all space by '\\Tit.ing

F 2 (h) = r {p(OMk) Jor Jor4" p(OMk + r)--,;;:--

Jv
00
sin hr
r 2
}
dw dr dvk (19)

on condition that p(OMk +

r) is taken equal to the density of the particle
p if the point OMk +
r is inside the particle, and to zero if the point is
outside.
\Ve can now write that the partial integral

fvp(OMk)p(OMk + r) dvk
is equal to the product of p 2 times the volume V(r) of the solid common
to the particle and to the "ghost" of the particle translated by the vector
r (Wilson (1949)) (Fig. 2). V(r) is evidently a function of the direction of
the vector r. If we introduce the average value, as defined by the
relation

L4" V(r) dw = 41T V(r)

equation 19 becomes
-- 1
F 2 (h) = p2
o
00
- sin hr
V(r)-- 41Tr 2 dr
hr
12 SMALL-ANGLE SCATTERING OF X-RAYS

Let us now introduce a function y 0 (r), defined as

V(r) V(r)
Yo(r) = V(o) =V (20)

Our last equation then becomes

--
F2(h) = V pa !"°o
y 0(r) sin
--
hr
hr 4m'll dr (21)

Fig. 2. A representation of the function V(r).

A comparison of equations 17 and 21 shows that the functions p(r) and

y 0 (r) are related by the following expression:

(22)

2.1.2.4. 'The Characteristic Function of the Particle yo(r)

The characteristic function y 0 (r) was introduced by Porod [137]. It
has no intuitive connection with the form of the particle.
y 0 (r) represents the probability that a point at a distance r in an arbi-
trary direction from a given point in the particle will itself also be in the
particle.
Let us consider an arbitrary line in the particle, terminating on its
boundaries to form a segment of length M, and let us further consider
an arbitrary point on this segment. The probability that a second point
on the line at a distance r from the first is also inside the segment M is:
y M(r) = 1 - (r/M) if r < M and is zero if r > M (Fig. 3). If g(M) 1 is
1 A precise definition of g(M) is as follows: Through a point r in the particle

there will pass an infinite set of randomly oriented lines. If gr(M) is the distri-
bution function for the lengths M of these lines, then g(M) is the average of this
function as the point r takes all positions in the particle, i.e.,
GENERAL THEORY 13

the distribution function for the group of such lines in the particle, then

y 0 (r) = ("'
JM~r
(1 - !.) g(M) dM
M
(23)

It can be shown from equation 23 that

g(M) = M (d2~0)
dr r-M

r
Fig. 3 The function ')'.v(r) for a single segment of length M.

The function y 0 (r) possesses the following general properties:

l. At r = 0, y 0 (r) has the value unity; as r increases, y 0 (r) decreases,
always staying positive, and becomes zero beyond the value r = R 1
corresponding to the line of maximum length through the particle.
2. An integration from zero to infinity of the two sides of equation 16
gives

p2 L00
p(r)dr= J/Vpdvk=p 2 V 2

which, when combined with equation 22, leads to the relation

L"' 47T1' 2 y 0 (r) dr = V

3. The initial swpe of y 0 (r) is a function of the external surface of the
particle, S. Let us trace around the particle the shell of thickness r
(Fig. 4), where r is small with respect to the dimensions of the particle.
14 SMALL-ANGLE SCATTERING OF X-RAYS

We can now calculate y 0 (r) by means of equation 16, neglecting the terms
smaller than r 3 •
fl[Vk(r + dr) - Vk(r)]p dvk = 411p 2Vr2y 0(r) dr

Fig. 4. Calculation of the initial slope of the characteristic

function y 0 (r).

For a point Mk' in the inner volume V' = V - Sr

Vk(r + dr) - Vk(r) = 47Tr 2 dr
and therefore
{ p[ Vk(r + dr) - Vk(r)]p dvk = 411p 2r 2( V - Sr) dr
Jv·
For a point llfk in the shell at a depth x from the surface (Fig. 4),
Vk(r + dr) - V 1k) = 27TT(r + x) dr
and therefore
{ p[Vk(r + dr) - Vk(r)]p dvk = i"'~211p 2 r(r + x)S dx dr = 311r 3 p2S dr
Jahell~Sr x-o
GENERAL THEORY 15

Thus as a first approximation

4rrp 2r 2Vy 0(r) dr = 47Tp2r 2 ( V - ~) dr

or
y 0 (r) = 1 - (S/4 V)r + ··· (24)

l'o(r)

~
'\
~

""
O'--~~~~~~-,L-~~"-_,_~~===-~L-~-+
""
R 4R 2R r
3

Fig. 5. The functions p(r) and y 0 (r) for the sphere of radius R.

As an example, let us consider a spherical particle of radius R. The

volume V(r) = V(r) common to two spheres ofradius R whose centers are
separated by the distance r is given by a simple geometrical calculation as
V(r) = (7r/12)(2R - r) 2 (4R + r)
Consequently,
Yo(r) = 1 - :~ + 116 (~)3
Equation 24 gives a similar result when V is replaced by (4/3)7T R 3 and S
by 4rrR2 (Fig. 5).
16 SMALL-ANGLE SCATTERING OF X-RAYS

Figure 3 shows that, for the line in the particle of length M, JyM(r) dr
= M /2. The integral of the characteristic function of the particle is thus

L0 oo y 0 (r) dr = LOOM
0 2
-
-g(M) dM = !
2
(25)

The integral of the characteristic function is therefore equal to one-half

of an average length of all the lines contained in the particles.
Thus for a spherical particle

It can be verified that (3/2)R is the average length of the lines passing
through all the points in a sphere in all directions and terminating on its
boundaries.
We see therefore that this function shows properties analogous to
those of the Fourier transform of the profiles of Debye-Scherrer lines
broadened by the effect of the small size of a crystal (Bertaut (1950)).

2.1.2.5. General Properties of ¢(h)

From these general properties of the function y 0 (r) we can deduce the
following consequences for the function }' 2(h):
1. The value of F 2 (h) at h = 0, F 2(0), is

F 2(0) = Vp 2 L 00
47Tr 2y 0(r) dr = V 2p 2

This is the square of the total number of electrons in the particle. All
the scattered waves are in phase and the amplitudes are added.
2. The value of F 2 (h) at small values of h is found from equation 21
by making the expansion
sin hr h 2r 2 h4r4
--,;:;:-- = 1 - 6 + 120 + ...
Then, by introducing the factor F 2 (0), this equation becomes

= h2 1
F2(0) { 1 - - -
loo
F2(h)
v 0 47Tr4y 0(r) dr
6

+ -120 -v 0 477r 8y 0(r) dr + ...}

h4 1 ioo
(21a)
GENERAL THEORY 17

Thus, as h increases from zero, F 2 (h) decreases following a parabolic

curve. The curvature of this curve is determined by an integral in which
the values of y 0(r) for larger play a predominant part because of the factor
r 4 • In §2.1.3.1 we shall see a simple and much more important expression
for the curvature of F 2 (h) at small angles.
3. A useful representation of the value of F 2(h) for large values of h
can also be obtained from the function y 0 (r). This comes from the fact
that, since hF2(h) and ry 0 (r) are related by a Fourier transform, the high-
angle part of the curve of F 2 (h) corresponds to the part of the curve of
y 0 (r) at small values of r, and an approximate expression for this part of
y 0 (r) is known.
y 0 (r) can be expressed as a polynomial in r, of which the first two
terms are known:
y 0 (r) = 1 - (S/4 V)r + ···
We also know that y 0 (r) becomes zero beyond r = R 1 . Therefore, by
making the substitutions hr = y and hR1 = u, equation 21 becomes

-
F2(h) = - -
47T v p2
h3 o
l" ( y - - s y2
4 Vh
+ -r1.y3
h2
+ · · ·) sin y dy

By integrating by parts the following formulas can be established:

L" y sin y dy = -u cos u + sin u

L" y sin y dy
2 = -u 2 cos u + 2u sin u + 2 cos u - 2

i" yn sin y dy = -un cos u + nun-l sin u - L" n(n - l)yn-2 sin y dy

Therefore
-- 2Trp 2S A f 1 (u,h) cos u f 2 (u,h) sin u
F 2 (h) = --y;;t + h6 + ... + ha + h3

At large values of h the principal term in F 2 (h) is 2Trp 2S/h 4 , to which

are added damped oscillations of pseudoperiod hR1 /27r. The average
curve of the continuous decrease of F 2 (h) is therefore given as

(26)

This depends ·uniquely on the external surface of the particle.

18 SMALL-ANGLE SCATTERING OF X-RAYS

4. A Fourier inversion of equation 21 gives

ry 0 (r) = -2
7T
i"" v
0
hF 2
- -2(h)
47Tp
- sin hr dh

or
y 0 (r) =- 1
- -
2 2
2TTpVo
i"" -- sin hr
h2F2(h)-h-dh
r
(27)

Evaluated at r = 0, this becomes

L'' h 1 F2(h) dh = 27T2p2V (28)

The integral of h2 F 2(h} depends only on the volume of the particle a.nd
not on its form. This is a. particular illustration of a general theorem
regarding the integral in reciprocal space of the intensity scattered by an
arbitrary object, which relates this integral to the total number of scatter-
ing electrons in the object.
5. Let us calculate an average value l of the length of all the lines
contained in a particle by evaluating the integral L"" y (r) dr.
0 By making
the substitution y =hr, the integral of equation 27 becomes

l = 2 ["' y 0 (r) dr = ~ ["" ["' hF2 (h) sin hr dh dr

Jo 7T p V Jo Jo r

= -1- l"'siny
- - dy
TT 2 p2 V o y
i"' -
o
hF2(h) dh
or
l
27Tp V o
= -1-
2
f"'
hF 2 (h) dh (29)

This integral can be expressed in terms of the total energy E scattered in

all the low-angle scattering region. On a film placed at a distance p from
the sample, the area that receives the rays scattered through the small
angles contained between 20 and 20 + d(20) can be written to a first
approximation as
da !:::'. 2TTp228 d(28)
or
da !:::'. (J.2/277) p 2h dh
Equations 4 and 6 then give

E = J.(h)I F (h) da2 = ;: 1 0 x 7.90 x 10-2sf F2(h)hdh

= 7.90 X 10-2s;.21op2vz (30)
GENERAL THEORY 19

All the results of the preceding discussion are still valid if the electronic
density of the particle is not a constant but shows fluctuations around an
average value p, if these fluctuations are such that statistically the sur-
roundings of all the atoms in the particle are the same.
If, on the other hand, p varies from one part to another of the particle
(for example, a hollow particle, etc.), equation 16 can be generalized by
introducing the function nk(r), which represents the number of electrons
situated at distances smaller than or equal to r from the volume element
dvk enclosing the point Mk. In order to modify our notation as little as
possible, we redefine p(r) by the relation

f [nk(r + dr) - nk(r)]p(Mk) dvk = p2p(r) dr

where pis the average electronic density. F 2 (h) can now be obtained by
replacing p by p in equation 17, but it is necessary to note carefully
that p(r) is no longer uniquely :determined by the geometry of the particle.

2.1.2.6. A Tabulation of the Average Intensity Distributions for Particles of

Different Shapes
We list below the average intensity distributions for particles of different
shapes which take all orientationR with equal probability. The intensity
distribution function tabulated is i(h), rather than F 2 (h), which is defined
by the relation

where n = V pis the total number of electrons in the particle; i(O) is then
always equal to unity.
(a) Sphere of radius R (Rayleigh (1914)) (Fig. 6),

• _
i(h) -
2
<l> (hR) -
_ [
3
sin hR - hR cos hR] 2 _ 9n
h3 R3 - 2
[J(hR)312
312(hR)] 2
(31)

(b) Ellipsoid of revolution, axes 2a, 2a, 2va (Guinier [65]) (Fig. 7),

i(h) =Jo("'2 <1> 2 (ha\/ cos 2 a + v2 sin2 0) cos ad(} (32)

Another equation has been developed for this case by Schull and Roess
[155], employing hypergeometric functions.
(c) Cylinders ofrevolution of diameter 2R and height 2H (Fournet [48])

. 1"'
i(h) =
o
2 sin 2 (hHcos0)
h2H2 cos 2 (j
X
4J1 2 (hRsin0). (}d£J
h 2 R 2 sin 2 (}
sm u (33)
20 SMALL-ANGLE SCATTERING OF X-RAYS

~Mean asymptotic curve

I
I

0.5 I
I
I
I

\~
\V Exponential approximation

',~
'
10 hR

,1
\I

i.
5 I 5

0 5 10 15 hR
Fig. 6. Scattered intensity from a sphere of radius R, ((> 1 (hR). The
curve is drawn with different scales for the various ranges of hR
( x 1000 for 4 < hR < IO; x 10,000 for hR > IO).
h1 R2
Expcnential approximation: e - -5- (equation 39); mean asymptotic
curve: ~ (h~)• (equation 26).

(d) Rod of infinitesimal transverse dimensions and length 2H (Neuge-

bauer (1943)) (Fig. 8a)
. Si(2hH) sin 2 (hH) (a4)
i(h) = hH - h2H'1.

where Si(x) = l"' 0

sin t
-dt
t
GENERAL THEORY 21

(e) Flat disc of infinitesimal thickness and diameter 2R (Kratky and

Porod [108]) (Fig. Sb)

. 2 [
i(h) = h 2 R 2 1 -
1
hRJ1 (2hR)
] (35)

These various functions i(h) behave according to the predictions of the

general study: at h = 0, i(h) is unity and the tangent to the curve is

l.O

0.5 Exponential approximation

ha-v¥-
Fig. 7. Scattered intensity from ellipsoids of revolution of axes
2a, 2a, 2va. The abscissae have been chosen so that the radius of gyration
of each ellipsoid corresponds to the same length (§2.l.3.l, p. 26).
a'
h1 2+••
Exponential approximation: e - 5 . -3- .

horizontal, and, as h increases, i(h) decreases parabo1ically, tending

finally towards zero along a curve which oscillates somewhat about a
curve varying as h- 4 • For narrow cylinders or thin discs whose small
dimension is €, this asymptotic law is valid only if h ~ (1/€). If in these
cases his large with respect to l/H or l/R but small with respect to 1/€,
equations 34 and 35 show that the curves decrease respectively as h-1
(cylinder) and h- 2 (disc).
An examination of Figs. 6, 7, and 8 shows that particles of very different
forms can have nearly the same scattering curves.
Tables 1-3 will facilitate numerical calculations of equations 31
through 35.
22 SMALL-ANGLE SCATTERING OF X-RAYS

Table 1
sin x - x cos x
<I>(.r) = 3 xa and cIJ2(x)

x !IJ(x) cIJ2(x) x !IJ(x) cIJ2(x)

0.000 1.000 1.000 2.100 0.622 0.388
0.100 0.999 0.998 2.250 0.575 0.330
0.200 0.996 0.992 2.500 0.499 0.248
0.300 0.991 0.982 3.000 0.346 0.119
0.400 0.983 0.968 3.200 0.288 0.083
0.500 0.975 0.951 3.500 0.205 0.042
0.600 0.964 0.930 4.000 0.0875 0.0076
0.700 0.952 0.906 4.493 0.0000 0.0000
0.800 0.937 0.879 5.000 -0.057 0 0.003 25
0.900 0.921 0.849 5.600 -0.085 0 0.007 22
1.000 0.903 0.816 5.760 -0.086 3 0.007 45
1.200 0.863 0.745 5.800 -0.086 0 0.007 40
1.400 0.816 0.668 6.000 -0.084 3 0.007 10
1.600 0.766 0.587 7.300 -0.023 0 0.000 53
1.800 0.702 0.502 7.720 -0.000 0.000 00
2.000 0.654 0.427
Table 2
'( Si(2x) sin2 x
ix)=-- - - -
x x2
x i(x) x i(x)
0.0 1.000 1.6 0.768
0.2 0.996 1.8 0.719
0.4 0.984 2.0 0.673
0.6 0.961 2.2 0.627
0.8 0.931 2.4 0.583
1.0 0.898 3.0 0.473
1.2 0.858 3.5 0.406
1.4 0.813 4.0 0.357
Table 3
x Ji(x) x Jl(x)
0.0 0.0000 2.6 0.4708
0.2 0.0995 2.8 0.4097
0.4 0.1960 3.0 0.3391
0.6 0.2867 3.2 0.2613
0.8 0.3688 3.4 0.1792
1.0 0.4401 3.6 0.0955
1.2 0.4983 3.8 0.0128
1.4 0.5419 4.0 -0.0660
1.6 0.5699 4.2 -0.1386
1.8 0.5815 4.4 -0.2028
2.0 0.5767 4.6 -0.2566
2.2 0.5560 4.8 -0.2984
2.4 0.5202 5.0 -0.3276
GENERAL THEORY 23

0
0 5 10 hR
(a)

1.0

~Asymptotic curve

\
~
0.5

(b)
Fig. 8. (a) Scattered intensity from rods of length 2H. Asymptotic
curve: 7r/(2hH) (<lquation 34). (b) Scattered intensity from fiat circular
discs of radius R. Asymptotic curve: 2/(h 2 R 2 ) (equation 35).

2.1.2.7. Particle with Preferred Orientations

For simplicity we shall consider only particles with a center of sym-
metry; the structure factor is then given by equation 5. Let us fix the
particle in a system of three mutually perpendicular axes, Ox, Oy, Oz. The
space in which the particle is found is described by a second set of three
mutually perpendicular axes, OX, 0 Y, OZ. The centers of these two
systems of axes can be made to coincide without loss of generality, since
only relative orientations arc of interest. Euler's angles, x, 0, and <p,
24 SMALL-ANGLE SCATTERING OF X-RAYS

will be employed to mark tlie orientation of the particle system with

respect to the spatial system. (In this paragraph () will designate only
the Euler angle, the scattering angle appearing only in terms of h.) We
arbitrarily orient the spatial axes so that h is directed along the axis
OZ. The amplitude scattered by a particle of orientation, (), cp, can be
denoted as: A.(h)F(h, (), cp). Then, by designating by P 1 (h, ()) and
P 2 (h, rp) the probability density functions of() and rp (where the notation
P 1 (h, 0) is to recall the particular choice of OZ), the average intensity is
found as

This approach will be useful whenever exterior physical conditions

impose a preferred orientation as, for example, when molecules are
oriented by the flow of a solution.

2.1.3. CONCEPT OF A RADIUS OF GYRATION OF A PARTICLE

2.1.3.1. Moving Particle
We shall consider primarily particles for which all orientations are
equally probable.
Let us rewrite equation 8, expanding the trigonometric function in a
power series:

(37)

The first term of the expansion is equal to CL:fk) 2 , that is, F 2 (0).
k
To describe the second term, let us consider a point, 0, chosen so that
2,fkOMk = 0. The point 0 then defines the electronic center of mass of
k
the particle. Employing this point as an origin, we can ~Tite

IM~; 1 =IOMk1 +IOM;1

2 2 2 - 21OMk11 om:, Icos <I>ki

In the second term of our expansion in equation 37 the contribution of

the factor I OMk I
2 is

h2 h2
- 6 't'f.!d; I OMk 1 =
2 - 6 f.hl.fk I OMk 12
GENERAL THEORY 25

The contribution of the factor I OM; 12 is the same. The contribution of

the angularly dependent term is zero, since

22.,2,fd; IOMk 11 OM; I cos <l>ki = 22,f; I OM; I{2.fk IOMk I cos <l>k;}
k j j k

and the sum over k can be recognized as being the projection of 2.,fkOMk
on the vector OM;· Thus equation 37 reduces to the following:

or

We can now introduce the parameter R 0 , defined by the relation

(38)

By analogy with classical mechanics, R 0 can be considered as the electronic

radius of gyration of the particle about its electronic center of mass. We
can thus hope to have a good approximation by writing (Guinier [65])

h 2 R 0'
-3-
(39)

where n = 2,fk, the total number of electrons in the particle. This

k
relation, the law of Guinier, coincides with the exact expression for terms
up to the 4th power of h and, like the exact. expression, vanishes with
increasing h.
Equation, 39 can be derived from the general relation, equation 12, in
which F 2 (h) and F(h) 2 were expressed in an explicit manner. For
simplicity we shall limit the derivation to the case of particles possessing
a center of symmetry.
The principal part of the double summation over indices k and j
involves terms in h4 or still higher powers. 'l'h1ts any constant term and
the term in h 2 in the expression for the average intensity must arise from the
square of the average amplitude and consequently must depend only on the
distribution of scattering centers as a function of their distance ~
center of the particle (see equation 9). The ratio of the coeffic#f.at~~
26 SMALL-ANGLE SCATTERING OF X-RAYS

term involving h 2 to the constant term should then be a universal function

of this distribution. A calculation

shows that this function is simply

1 2.,/,.r.,.2 1
- _,._ _ = -Ro2
3 2.,/,. 3
"'
confirming the results of the preceding paragraph.
The curves of the scattering by two particles having the same radius
of gyration then coincide at very small angles, and it is only the tails of
the curves, due to terms in h4 , h6 , etc., which show the influence of the
forms of the particles. When the scattering curves of particles of two
different forms are to be compared, it is thus essential to choose two
particles having the same radius of gyration (see Fig. 7). For example,
to a sphere of radius R, one should compare an ellipsoid of revolution of
axes

J+
2 --R
2
3
v2
J+
2 --R
2
'3
v2
2vJ 2+v3
2 R

v being the ratio of the unequal axes. Certain authors (Kratky and
Porod (108)) have compared reduced curves (curves of the function
i(h), chosen so that the abscissae for both are the same for the ordinate
i(h) = 1/2), but this arbitrary choice does not permit the separation of the
influences of the dimensions and of the forms of the particles.
In order to eliminate any possible confusion, we want now to emphasize
separately two points:
1. The validity of the concept of a radius of gyration. We must re-
emphasize that the concept of a radius of gyration is sound, whatever
the form of the particle (viz., Guinier (65], p. 191, or the discussion of
equation 12), since it seems that some authors have recognized this
concept only for spherical particles. The magnitude of the slope of the
curve of log J(h) vs. h 2 at the origin is always equal to one-third of the
square of the radius of gyration. The influence of the form of the particle
manifests itself particularly at larger values of h in the form of deviations
of the curve oflog J(h) from the extrapolation of its tangent at the origin.
2. The validity of the approximate law of Guinier. To illustrate this,
let us consider a family of ellipsoids of revolution of the same radius of
GENERAL THEORY 27

gyration, R 0 , and of half axes R,R, and vR; v and Rare thus variables.
The first terms in the power series describing the intensity are given by

F 2 (h) = F(0) 2 [ 1 - -T
R
h2 2
+ ··· ]

In this family of ellipsoids there are two for which the expression for the
scattered intensity coincides with the law of Guinier (equation 39) up to
terms of the 6th power of h; these are the flat ellipsoid defined by v = 0.24
and the elongated ellipsoid, where v = 1.88.
Again, in a family of cylinders of revolution of diameter 2R and height
2H there is a cylinder for which the same precision of agreement is found;
this is the cylinder for which H/R = V30/11 = 1.65.
These examples show that the particles that obey the approximate
law of Guinier closely are those that are nearly isodiametric (see Fig. 6
for the case of a sphere).
On the other hand, rather wide differences will be found for the curves
of very elongated ellipsoids, thin discs, and narrow cylinders.
Finally, let us call attention to the fact that experiments have verified
that the exponential law, equation 39, is a very good approximation for a
large number of scattering curves, a surprisingly large number in view of
the approximations involved in the derivation.
The precision with which a radius of gyration can be measured in an
experiment depends to a certain extent on the form of the particles. The
determination of the radius of gyration to the same degree of precision
for a series of particles having the same radius of gyration but different
forms requires information on the scattering at smaller and smaller angles,
the further the particle departs from a spherical form.
Kratky and Porod [108] have given an approximate formula for particles
in the form of narrow cylinders or thin discs which is valid for values of h
large with respect to the reciprocal of the large particle dimension but
small with respect to the reciprocal of the small dimension. In its
dependence on the small dimension of the particle this scattering function
behaves as though the low-angle region of the curve (the exponential
approximation) were involved, whereas in its dependence on the large
dimension the scattering behaves as though the tail of the curve (the
curves in h- 1 and h- 2 , respectively, for cylinders and discs) were involved.
An accurate calculation based on equation 33 gives as the relation for
narrow cylinders of diameter 2R and length 2H, containing n electrons
each,

(40)
28 SMALL-ANGLE SCATTERING OF X-RAYS

and for thin discs of diameter 2R and thickness 2H

- 2 sin 2(hH) 2n2 _ h2 II'
F 2h ,_, n 2 - - • ,_, - - e 3 (41)
- h 2R 2 (hH) 2 - h2 R 2
Let us also recall the equation relating the radius of gyration of a
particle and its characteristic function y 0{r). By comparing the expansion
of F 2 (h), equation 2la, p. 16, with equation 39, this is found to be

Ro2 -
1
= 2V
i"' 41Tr4yo(r) dr =
f"' io
r4yo(r) dr
(42)
o "'
2 r 2 y 0 (r) dr
0

We note that the radius of gyration can then be defined either by

equation 39, where, following the method of Guinier, it is determined
from a trace of the curve of log I vs. h 2 , or by equation 42, in which case
it is convenient to determine the function y 0 (r) with the aid of equation
27 and then R 0 2 with equation 42.

2.1.4. SPHERICALLY SYMMETRIC PARTICLE

The calculation of the average intensity is considerably simplified when
the particle possesses spherical symmetry. The electronic density
function p(r) is then sufficient to determine A(h) and, consequently,
F 2 (h). Conversely, a Fourier transformation of equation 13 gives

p(r) = - 12
27T r o
i"'
hF(h) sin hr dh (43)

and the radius of gyration, from its definition, is determined by the

relation

i"'r 4 p(r)dr
R 2_~0_ _ __

o - Loo r2p(r) dr
2.1.5. THE DISTRIBUTION OF SCATTERING FROM A FIXED PARTICLE
Figure 9 shows the geometrical relations between the film, the orienta-
tion of the particle, and the distribution of scattered intensity on the film.
We are interested primarily in the distribution of scattered radiation
along the line AB. The corresponding scattering vectors h are all con-
tained in the plane fixed by B0 and AB. The limiting direction of h
as the scattering angle 2(J tends to zero is marked by the unit vector 10
perpendicular to Bo (since 4 hs 0 = (7r/2) - (J; see Fig. 1).
GENERAL THEORY 29

Let us consider how the concept of a radius of gyration appears under

these circumstances by studying the series expansion of equation 3. An
analogous case has already been considered in §2.1.3, giving immediately
the result

where the point 0 is defined by the relation

Fig. 9. The relation between the orientation of a particle and the

distribution of scattered intensity in the plane of observation.

Then
F2(b) ffk (h • OMk)2
F2(0) = 1- 2.Jk + ...
k

As a· first approximation, the scalar product b · OMk for small values of

20 is equal to the product of the magnitude h and the scalar product
10 • OMk. This scalar product, 10 • OMk, is equal in magnitude to the
distance dk(l 0 ) of the point Mk from the plane Il (10 ) through 0 perpen-
dicular to 10 • The notation dk(l 0 ) serves to recall the dependence on the
orientation of 10 • We can now write the expansion of F 2(h) as

F 2 (h)
- - = l - h2
f fk dk (10) + ...
2

F2(0) 2.A
k

The coefficient of the term in h2 can be designated as the square of an

30 SMALL-ANGLE SCATTERING OF X-RAYS

average inertial distance, D(l 0 ), of the particle with respect to the par-
ticular plane I1(10 ). Then, as in the law of Guinier, this expression can
be written to a good approximation as
(44)
Equation 44 explains the distribution of scattered radiation found on
the film of Fig. 9. The particle is presenting its largest dimension in the
direction parallel to AB. The average inertial distance, the factor of
primary importance in determining the scattered intensity, is thus also
a maximum for this particular direction. Equation 44 then shows that
it is along this line AB that the decrease in intensity with increasing h
will be the most rapid.

2.2. SCATTERING PRODUCED BY A GROUP OF IDENTICAL PARTICLES

2.2.1. GENERAL RESULTS FOR FLUIDS'
2.2.1.1. Basic Hypotheses
Scattering experiments are rarely cond~-cted with a single particle as
the scatterer; thus it is of more practical importance to calculate the
intensity scattered by a group of particles. In this section we shall
consider the simplest such case, that of a group of identical particles.
We shall also restrict our study by requiring that the structure of the
ensemble of particles satisfy two hypotheses. To simplify later references,
these hypotheses will be denoted as H 1 and H 2 •
Hypothesis H 1 • We shall assume that:
1. All particles, each possessing a center of symmetry, can, with equal
probability take all possible orientations.
2. The knowledge of the relative positions of two particles in no way
modifies the probabilities of their different orientations.
The second part of hypothesis H 1 is always realized for spherically
symmetric particles. For the more general case it would seem that, if
the particles are not too densely packed and if their shapes are not too
anisotropic, this hypothesis should be good at least as a first approximation.
Hypothesis H 2 • We shall assume that the group of particles is iso-
tropic and without order at long distances.
In order to clarify hypothesis H 2 , let us define Rk as the vector from
an arbitrary origin to the center of the kth particle. The vector joining
the centers of the kth and jth particles is then (Rk - R,). Hypothesis
H 2 then requires that all vectors (Rk - R 1) of the same magnitude have
an equal probability of orientation in all directions (isotropic) and that
for long distances (viz., 1000 A) the probability offinding vectors (Rk - R;)
1 We include in the term "fluids" gases, liquids, solutions, and suspensions. Thus
the word fluid refers to all matter satisfying hypotheses H 1 and H 2 •
GENERAL THEORY 31

of a given magnitude, r, is a continuous function of rand is nearly constant.

This hypothesis is well justified for fluids.

2.2.1.2. Consequences of Hypothesis H 1

Let us consider a group of particles, each possessing a center of
symmetry. We shall designate by rk 1 the vector extending from the
center of the kth particle to a point l of scattering factor fki in the same
particle.
The amplitude of radiation scattered by this group of particles is then
given as (cf. equation 2)
A(h) = A.(h) L {2:fk 1e -ih· (R>+r.,l}
k l

since the position of the scattering point, l, is defined by the vector sum
Rk + rkz·
By virtue of the center of symmetry of each particle, this can be written
as
A(h) = A.(h),2:e-ih·R•2:.fk1 cos (h · rk 1) (45)
k l

The sum 2fkz cos (h • rk 1) corresponds to the structure factor Fk(h) of

l
the kth particle (see equation 5); since the part.ides are identical, the
index k serves only to mark the kth particle.
The mathematical formulations of equation 45 and equation 2 are
identical, with the sum over the index l playing the role of the quantity
fk, so the scattered intensity can be found easily by analogy with equatio113:
/(h) = I.(h),2:2:[2:fki cos (h · rk 1 )][Lf;m cos (h · r,ml1 cos [h · (Rk - R,)]
k j l "' (46)

In equation 46 indices k and j refer to particles, arnl indiees l and m

refer to the different seattering points i11 partides k and j, respectively.
The intensity caleulated above is that furnishe(l by one certain con-
figuration of the ensemble of particles. In the comse of time this
configuration changes, the partieles shifting awl changing orientation.
Thus we can observe only average intensities. lVe shall discuss here only
the general method of eakulating the average intensity; for details, sec•
Fournet [48].
First we separate the terms where k = j in equation 46. These terms
represent the intensity seattered by one particle multiplied by the average
number of particles being examinecl (Hee §2.2.1.3). For the calculation
of the other terms, use is made of the second part of hypothesiH H 1 ,
which t>nablt>s us to separate the calculation of the averaging of the
32 SMALL-ANGLE SCATTERING OF X-RAYS

orientations of the particles from the averaging of the positions of the

particles. The final result is:

/(h) = I,(h){N F 2(h) + F(hJ2:2:kN,k

2 cos [h · (Rk - R;)]} (47)

where F 2 (h) and F(h) 2 are the quantities defined earlier (p. 7). The
problem that remains is the calculation of the average of the double sum
of terms.

2.2.~.3. Consequences of Hypothesis H,

Let us consider two very small volume elements, ~vk and ~v;, located
by vectors Rk and R1, which are contained in the volume V irradiated by
the X-rays. The contribution of this elementary pair, ~vk, ~v1 , to the
desired average is nk; cos [h · (Rk - R 1)], where nk; is the number of pairs
of particles found in these volume elements, one particle being in ~vk and
the other in ~v1 . Now, letting the elementary volumes ~vk and ~v 1
approach the volume elements dvk and dv 1, small even on an atomic scale,
we introduce a probability function Pk;• in terms of which the probability
of finding a particle in dvk and at the same time a different particle in
dv 1 is: p 2 =Pk; dvk dv 1• It is essential to consider that the particle in
dv 1 is different from that in dvk, since we have excluded the case of k = j
in the double sum above. The sum of the cosine terms is now found as
the integral (see equation 7)

rr
JvJv
I
sin (h Rk - R;
hjRk-R;I
ll
Pk;dvkdv;

The volume element dvk being infinitesimal, we shall describe a particle

as being in dvk if the center of the particle is found there.
Before calculating Pk;• let us carefully define the experimental con-
ditions. We shall designate by V 0 the total volume offered to the particles.
This volume contains a well-defined number of particles, N 0 • The
volume V 0 must be distinguished from the volume V actually irradiated
by the X-rays.
1. If V 0 is smaller than or equal to V, then it follows that N = N 0 ,
where N is the number of particles in V.
2. If V 0 is larger than V, the case generally met experimentally, then
N cannot be known exactly, as previous authors have implicitly assumed.
Only statistical information, such as an average value of N, can be known.
We shall assume that V 0 is large as compared with V, as is true of most
experiments, and in any event a situation which is easily realizable. Let
us now evaluate the probability Pk; dvk dv;. We know that the probability
of occurrence of an ensemble of two events is equal to the probability of
GENERAL THEORY 33

the first multiplied by the probability of the second as modified by the

knowledge that the first event exists. The probability of occurrence of the
first event, finding a particle in dvk, is (N 0 / V 0) dvk. When this is realized:
1. The remaining N 0 - 1 particles are distributed through a volume
V0 - dvk.
2. If the centers of the two volumes dvk and dv;, each containing a
particle, are separated by a distance r, the centers of the particles are
separated by a distance approximately equal tor. It is obvious that the
different distances (and uniquely the distances, if the matter being
examined is composed of only one phase) cannot all be equally probable;
if the particles are spheres of radius R, the distance between particles
cannot be. less than 2R. This behavior will be described by a function
P(r) such that
N0 N0 - l
Pt; dvk dv; =- dvk P(IRk - R;j) dv; (48)
V0 V0 - dvk
The manner in which P(r) has been introduced shows that this function
will tend towards unity as r increases, for then the condition discussed
above disappears or, rather, does not play a part. We can neglect dvk
with respect to V 0 , and usually 1 is negligible compared to N 0 (see §2.4).
By introducing the average volume offered to each particle, Vi = (V 0/N 0 ),
a quantity characteristic of the ensemble of particles, equation 48
reduces to

(49)

and the desired average of the double sum becomes

2 2 cos [h • (Rk -
kj*k
R;)] = ii V V
sinhrk.
--'
hrki
dvkdv1
P(rt;) - -
Vi Vi
(50)

2.2.1.4. General Expression for the Scattered Intensity

Since P(r) tends toward unity as r increases, the structure of the
ensemble of particles might be better characterized by the function
(1 - P(r)), this function being different from zero only for small values of
r. By making the substitution P(r) = 1 - (1 - P(r)) equation 50 takes
on the following form:
34 SMALL-ANGLE SCATTERING OF X-RAYS

J,et us examine separately each of these terms.

FirBt term: The contribution of this term to the total scattered intensity
is
-
I 1 ( h ) =I.(h )F (h)
2 ii
v V
sin hrk; dvkdv;
----
hrki V1 Vl
which (cf. equation 15) can be interpreted as representing the intensity
scattered by a "particle" of volume V with a uniform electronic density,
p = F(h)/v1 • Given the size of this particle, / 1 (h) is effectively zero for
all obBervable angleB. (Fournet [48] discusses the reasoning of Compton
and Allison (1935) on this subject.)
It is important to note the simple, general interpretation of the term
/ 1 (h). A further discussion of this factor is given by James (1948),
where the scattering of a spherical volume is considered. To find simply
and schematically the limit h0 , beyond which / 1 (h) is negligible, it might
be considered that for h = h0 the largest phase difference between rays
scattered by two points in the particle will be of the order of 211 radians.
Then, if the average dimension of the volume irradiated is designated by
D 0 , h 0 is determined by h0 = 211/D 0 •
Second term: Let us consider first the integration with respect to dv11::

f sin hrki [l _ Ph;)] dvk

Jv hrk; v1

Since the function [l - P(rk;)] approaches zero rapidly as r increases, by

neglecting boundary effects this term can be written as

i O
00 sin hr 4m2
- - [l - P(r)]- dr
hr V1
(51)

independent of the index j. The further integration with respect to j

dv.
then results simply in multiplying equation 51 by a factor -2, which is
-
J V1
equal to N, the average number of particles in the irradiated volume V.
The final relation for the scattered intensity is thus (Fournet [44])

/(h) = I 1 (h) + I.(h)N {F2(h) - F(h)

2
f"' sin hr [l - P(r)]4m2 dr} (52)
v1 Jo hr
/ 1 (h) is completely unobservable, and so for h > h 0 this becomes

-
/(h) - {
= I.(h)N - - -
F2(h) F(h)
-
2

V1
i"'
0
sin
--
hr
hr [l - P(r)]411r 2 dr} (53)

If the particles considered are spherically symmetric, we have seen

GENERAL THEORY 35

( §2.1.2) that there is no distinction between the average of the square of

the structure factor F 2 (h) and the square of the average of the structure
factor F(h) 2 • For this case equation 53 can be simplified to the following:

l(h) = I 0(h)N F 2 (h) {1 - !._ ( 00 [l - P(r)] sin hr 417r2 dr} (54)

vi Jo hr
This expression was derived by Zernicke and Prins [309] and by Debye
and Mencke [269]. Equation 53 thus appears as a generalization of this
last expression.
The integral figuring in equations 53 and 54 has the dimensions of a.
volume. We shall define this as the volume of perturbation, v2 (h):

v2 (h) = i "'
o
sin hr
[l - P(r)] - - 47TT 2 dr
hr
We shall see in §2.2.3.1 that this function, which has also been called the
"characteristic volume" (Porod [137]), is actually a function of two
variables: v2 = v 2 (h, vi)·
To summarize, we have shown that the scattered intensity can be
expressed as a sum of two terms:

l(h) = I.(h)F(h) 2 ( ( sin hrt1 dvkdv 1

Jv Jv hrk; vi vi

the first of these terms being negligible with respect to the second for
h > h 0 • The value of h 0 is defined by the relation h 0 D 0 = 277, D 0 being
the average dimension of the irradiated volume. In the remainder of this
section we shall designate by "intensity I(h)" only the second term of
equation 52, and when a misunderstanding is possible we shall employ
the expression "observable intensity" to denote this second term.

2.2.2. WIDELY SEPARATED PARTICLES

The general expression which we have just established shows the
influence of interparticle interferences, through the intermediary of
v2 (h), on the scattered intensity. Let us now consider in detail the case of
rather widely separated particles, for which the expression for the
intensity takes on a particularly simple form. We shall later indicate the
criterion which defines "widely separated particles," but for the moment
we can indicate that a good example of such a system is a gas under low
pressure.
36 SMALL-ANGLE SCATTERING OF X-RAYS

2.2.2.1. Expression for the Scattered Intensity

If the particles are widely separated, the average volume Vi offered to
each will be large. Any irregularities of the function [l - P(r)] demon-
strate that certain interparticle distances are favored while others are less
probable. These irregularities are the more marked, the more closely the
particles are packed, for, in order to contain more matter in a given volume,
the degree of organization of this matter must be increased.
We see thus for two reasons that the ratio of v2 (h) to Vi is very small
for widely separated particles. More rigorously, the part of v2 (h)
inyolving the integration from r = 0 to r equal to the smallest particle
diameter will not vary as the particles become more separated, but, since
v1 increases, the corresponding part of the ratio of v 2 (h) to Vi decreases.
In the limit of large separations, we find the classical result

I(h) = I.(h)NF2(h) (55)

where we recall that this expression, derived from equation 53, is valid
only for h > h 0 (h 0 has been defined on p. 34).
The intensity of radiation scattered by an ensemble of widely separated
particle8 is thus identical on a relative scale to the mean intensity scattered by
one isolated partick; in obtaining the intensity relative to an ensemble of
particles it is necessary only to multiply the intensity scattered by one
particle by the average number of particles, N.
Realizing the practical importance of this simple result, it is opportune
to underline its significance and to recognize its limits of validity. Later
(§2.2.3.2) we shall demonstrate the connection existing between this
expression and the equation of state for ideal gases, pvi = kT.
Let us compare equation 55 with a well-known problem in optics. It
is often indicated in the literature that the intensity of scattering by
identical elements distributed at random is formed by the addition of the
intensities scattered by each element. This is not always correct, for
- ---
if h = 0, we are led to the result /(0) = N F 2(0), whereas the exact result
is known to be /(0) = N 2 F 2(0). The usual reasoning behind the above
statement consists of describing the double sum
L L cos (h. (Rk - Ri))
.I: j #i:

as containing as many positive terms as negative terms, so that con-

sequently the sum is zero. But if h = 0 all the cosine terms are equal to
unity and this reasoning is no longer true. As a criterion for applicability
we can say that this reasoning is correct when the largest phase difference
between particles reaches27Tradians; that is, if D 0 is the average dimension
of the volume offer"<l to the particles, the reasoning is correct when
GENERAL THEORY 37

h > (2TT/D 0 ) = h 0 . This is simply a restatement of the result that was

established in §2.2.1.3.
The classical reasoning is a little too simplified, if not incorrect, and it
should be replaced with that due to Lord Rayleigh (1919), who tries first
to calculate the probability that the intensity will be between I and I dl +
and then afterwards calculates the average intensity. By correcting one
error and slightly modifying the reasoning of Lord Rayleigh to render it
applicable to the problem of X-ray scattering, equation 55 can be obtained
as a first approximation when h > h 0 • The corrective terms that appear
are negligible when N is very large (Fournet [48]).

Remarks
For the simple case of widely separated, spherical particles, the scattered intensity
(see equation 47) is given by

l(h) = 1.(h)F 2 (h){N + :E :E cos (h • (R. - R,))}

kjiok

Let us compare the mathematical structure of the square of the structure factor of
the particle, F 2 (h). with the bracketed term. The intensity scattered by a spherically
symmetric particle, whose p scattering centers each have the same scattering factor
f, can be described as (cf. equation 3)

J,(h)F'(h) = 1,(h)/ 2 :E :E cos (h • M,Mml

lm
= l.(h){pf' + / 2 :E :E cos (h • (rm - r,))} (56)
l miol

The previous discussion shows that the second term is negligible wi~h respect to
the first when h is greater than h 1 = 211"/rl, rl being the average dimension of the
particle. For angles where h is the order of h 1/IO, however, each of the terms of
equation 56 has approximately the same value, so that the term pf' is negligible
compared to the double sum.
The total scattered intensity is thus

J(h) = I,(h)j 2 {p + :E :E cos (h •(rm - r,))}

I miol
x {N + :E :E cos (h · (R. - R;)l} (57)
k jiok

The symmetry of this relation is evident; mathematically the description is the

same, whether for an ensemble of points in a particle or for an ensemble of particles,
but the physical results are very different.
For 0 < h < h 0 , the values of the bracket concerning the ensemble of particles
are not interesting to the physicist, since they cannot be reached experimentally.
For h 0 < h < h 1 , the values of the bracket concerning the ensemble of points in a
particle are essential ancl permit the determination of the radius of gyration. For
an angle of the order of h 1 /10 and for quasi-homogeneous particles, the second term
of the first bracket and the first term of the second bracket are the terms that are
important in this expression. The two parts of this equation are shown separately
in Fig. 10, in which curve a refers to the first factor and curve b to the second.
38 SMALL-ANGLE SCATTERING OF X-RAYS

2.2.2.2. Remarks on the Babinet Principle of Reciprocity

This principle will be considered at this point not because it is applicable
uniquely to ensembles of widely separated particles but rather because our
recent concern with just such ensembles permits us to treat it very quickly.
First term,equation 57

h
Curve a

Second term, equation 57

ho
Curve b
Fig. IO. A graphical representation of the two factors of equation 57.
For clarity h 0 has been greatly exaggerated.

Let us first recall the simplest enunciation of Babinet's theorem:

complementary objects produce the same diffraction effects. The con-
cept of complementary objects will be more precisely defined in the follow-
ing paragraph, but we can give a simple illustration of a pair of such
objects: a screen pierced with circular holes, and an ensemble of circular
discs, each disc corresponding in size and position to a particular hole.
The general expression for the amplitude scattered by matter con-
tained in a volume V is
A 1(h) = A,(h) f/ 1(r)e-ih·r dr
GENERAL THEORY 39

where p1(r) is the electronic density in the volume element dr whose

position is fixed by the vector r. Let us now consider a complementary
space, whose electronic density p 2(r) is defined as: p 2(r) = p0 - p1(r),
where p0 is a constant. The amplitude scattered by this complementary
space is then
A2(h) = A.(h) fv[Po - P1(r)]e-ih·r dr

The problem now is to compare

11(h) = A1(h) A1*(h)
and
1 2(h) = A 2 (h) A 2 *(h)
If we write
A 0 (h) = A 1 (h) + A 2 (h) = A.(h) f
•v
p0e-ih·r dr
then the function,

is a maximum for h = 0
[1 0 (0) = I.(O)V 2 p0 2]
decreases rapidly with increasing h, and effectively becomes zero beyond
an angle defined by h0 = 27T/ D 0 , D 0 being the average dimension of V.
We can now calculate the following:
1 2(h) = [A 0 (h) - A 1(h)][A 0 *(h) - A 1 *(h)]
= 11(h) + 1 0(h) - A 1(h)A 0*(h) - A 0(h)A 1*(h) (58)
If h > h 0 , J 0 (h) is effectively zero, and consequently A 0 (h) and A 0 *(h)
are also zero. Then J 1 (h) =I 2 (h), the usual statement of the reciprocity
principle. However, if h < h0 , equation 58 shows that this principle is
no longer true.
Thus the principle of reciprocity can be applied only to calculate the
intensity scattered at angles such that the corresponding values of h are
greater than the limit 27T/ D 0 • The dimension D 0 refers to the average
dimension of the volume in which a "complementary" electronic density
is defined.
In the experimental systems generally used for the study of low-angle
scattering, the principle of reciprocity can be applied to an ensemble of
particles (that is, an ensemble of cavities in a homogeneous medium can be
considered equally well in place of an ensemble of particles), since the
intervening dimension is defined by the transverse dimension of the beam,
generally of the order of 1 mm. wide, but it cannot be applied to each single
particle (as, for example, replacing a spherical particle of 100 A diameter
containing a concentric spherical cavity of 20 A diameter by a simple
40 SMALL-ANGLE SCATTERING OF X-RAYS

sphere of 20 A diameter). This latter method has been employed by

some authors, leading to incorrect results; we cite for example the article
of Kratky and Porod ([108], p. 45.ff), in which these authors tried to
evaluate the intensity scattered by a "finite packing of lamellae" of
submicroscopic dimensions by replacing the lamellae with the interstices
contained between them. In a later article, however, Porod is in agree-
ment with the ideas expressed here.
Another application of these ideas is the calculation of the scattering
from particles which, instead of being in a vacuum, are im'TflR,rsed in a
homogeneous medium of electronic density Po (for example, the solvent of a
colloidal suspension). The scattering body can be considered as the
superposition of a continuous medium of density Po and particles of
density p - Po· The scattered amplitude is the sum of the amplitudes
scattered by the continuous medium and by the fictitious particles. The
first of these is zero throughout the region accessible to experiment. The
observed scattering is therefore simply that due to the particles of density
p - Po· All the equations which have been established are thus valid on
condition that p is considered as the difference between the electronic density
of the particle and that of the surrounding medium. The small-angle
scattering becomes zero if the particles have a density equivalent to that
of the surrounding medium, even if they have a quite different atomic
structure.

2.2.3. INFLUENCE OF THE CLOSER PACKING OF PARTICLES

2.2.3.1. General Considerations

We have just treated the simple case of widely separated particles, in

which the total scattered intensity, proportional to the intensity relative
to a single particle, generally decreases continuously with increasing
scattering angle. It is only for the very particular case of particles with
a strongly marked internal structure (for example, CC1 4 molecules) that
intensity maxima at non-zero angles can be observed for widely separated
particles.
It is well known that numerous liquids whose elementary particles
possess simple structures give rise to maxima of intensity at non-zero
angles. Zernicke and Prins [309] established their well-known formula
(equation 54) in order to explain these results. We should point out that
it is difficult to study the effect of the closer packing of particles on the
scattering distribution from this equation; when the concentration of
matter increases, v1 decreases, but there is al,so an unknown change in the
function P(r). Writing the equation for the intensity in the form given by
Zernicke and Prins could lead to the assumption that the concentration
of matter has no influence on the function P(r). To eliminate the
GENERAL THEORY 41

possibility of this assumption, we propose to modify slightly the expression

of the Zernicke-Prins formula, writing:

l(h) = I.(h)N F 2(h) {1 - ~ roo [l - P(r, V1)] sin hr 47Tr2 dr} (59)
v1 Jo hr
in which the functional dependence of P(r, v1 ) appears explicitly.
Numerous authors have studied the influence of the mutual approach of
particles on the distribution of scattered radiation, assuming a priori a
function P(r) more or less well chosen but independent of the concen-
tration of the matter. Among the latest attempts we may cite that of
Yudowitch [186]. We believe that it is difficult to determine the validity
of the results thus obtained, for in these studies of the influence of con-
centration on the intensity l(h), one of the most important functions
determining this intensity has been assumed a priori to be independent of
concentration.
The real problem then in any such study is to obtain the function
P(r, v1 ). This problem is difficult. We feel certain that its solution
will require a profound analysis of the thermodynamics of ensembles of
particles. \Ve can hope that by such an approach, equations 53 and 54,
which may be called "Zernicke-Prins type equations," concerning only
the geometry of the ensembles ?f particles, can be transformed into
"thermodynamic equations" by the introduction of certain intrinsic
characteristics of the particles.
It is first necessary to find variables that can intervene in a definition of
P(r). We have already noted one such variable, v1 . The function P(r)
is connected to the probability of seeing a certain configuration of two
particles realized, so that the calculation of probabilities introduces itself
naturally into the problem. If Boltzmann statistics are employed we
need to introduce both the temperature and the potential energy ll>(r) of
a pair of particles whose centers are separated by a distance r.
The problem of calculating the function P(r) from v1 , T, and ll>(r) is the
central problem of the latest kinetic theories of fluids (Yvon (1935),
Kirkwood (1935), Born and Green [259]). Our own problem is not limited
simply to the case of fluids; we are interested to a large extent in solutions
of large molecules, suspensions, etc. In each of these cases the functions
P(r) and IJ>(r) can always be defined. We shall introduce later the
variable, pressure, and the equation of state of the matter being con-
sidered. This presents no difficulty in problems concerning gases or
liquids. Moreover, it is well known that in solutions the osmotic pressure
plays a role analogous to that of pressure in fluids and that the equation of
state of ideal solutions can be expressed in the form pv 1 = kT. If
there is difficulty in extending these concepts to the study of emulsions,
42 SMALL-ANGLE SCATTERING OF X-RAYS

we can always formally define the pressure by evaluating the change of

free energy of the system with respect to the total volume offered.
In the different kinetic theories of fluids cited above, the integral
equation determining P(r) is of the form,

P(r) = F (P(r) , <l>(r)) (60)

Vi kT

where F designates an integration of the functions P(r)/v1 and <l>(r)/k'J'.

In establishing this equation it has been necessary to make a physical
approximation known as the "principle of superposition." We shall
not give the details of this principle, for these can be found in the article
of Kirkwood and Boggs (1942) in which this principle is described and
employed for the first time. We may describe equation 60 as having been
established in a very general manner from considerations of classical
Boltzmann statistics, the only assumption being that of the Yalidity of
the principle of superposition.

2.2.3.2. Scattered Intensity and the Equation of State

We have just seen that the kinetic theories of fluids furnish a relation
between P(r) and <l>(r); this shows the manner in which the relation
between the function l(h) and <l>(r) can be established. At this point it
seems advantageous to introduce the equation of state, a more familiar
quantity than the corresponding mutual potential energy function.
This can be written in the following form (Green [273], Yvon (1949)):

p = -kT -
v1
- 12
6v1
i"'
0
P(r)<I>' (r) 47Tr 3 dr (61)

If the integral is neglected, this reduces simply to the ideal gas law,

kT NAkT RT
p=-=--=-
V1 N Avl V
where NA is Avogadro's number.
Let us first consider the case for which the function <l>(r) is identically
equal to zero. The equation determining P(r) then states that P(r) - 1.
Two important relations resulting from this particular function P(r) can
now be derived:
1. The expression for the scattered intensity becomes (see equation 53)

I(h) = I.(h)NF 2(h) (55)

the result for widely separated particles.

GENERAL THEORY 43

2. The equation of state, obtained from the general expression of

equation 61, becomes
(62)
the ideal gas law.
This result shows that the domains of validity of equation 55 and
equation 62 are the same. These two expressions are rigorously true only
for point particles, small even on the angstrom scale, as is required when
the potential energy is defined as <l>(r) = 0, independent of r. In
§2.2.3.4 we shall discuss the conditions under which equation 55 is a good
approximation.
Let us remark also that if an attempt is made to find a function P(r)
which is independent of v1 and which is to be defined by equation 60, the
only possible solution will be found to be P(r) = 1.
Now let us leave the case of widely separated particles and try to treat
the general case. By developing P(r, v1 ) in a series expansion with v1 as
the variable, we find
P(r, v1 ) = P 0 (:r) + v1 Pi(r) + · · ·
where the term P 0 (r) is identical to e-IT>(r)/kT.
Thus a first approximation of equation 60 is
P(r, V1) = e-lll(r)/kT (63)
For this approximation the equation of state becomes
(21T)s12 p(O)]
pv1 = kT [ 1 - - - -
2 v1
and the scattered intensity distribution is (cf. equation 54)

l(h) = I.(h)N F 2 (h) [ 1 + (217)312

~ p(h)
]
(64)

where the function p(h) is defined by the relation

hp(h) = 2
--:=
v21T o
i"" roc(r) sin hr dr (65)

with
oc(r) = e-IT>(r)JltT - 1

An approximate solution for hard spheres of radius R and volume v 0 ,

with no interactions other than impenetrability, has been considered by
Debye [265]. With the probability function defined (cf. equation 63) as
P(r) = 0 0 < r< 2R
P(r) = 1 r>2R
44 SMALL-ANGLE SCATTERIN"G OF X-RAYS

equations 31 and 64 lead to the following expression for the scattered

intensity:

(66)

where the function cf>(x) is described by equation 31. We shall consider

the validity of this expression in §2.2.3.3.
The solution P(r, v1 ) = e-'11(r)/kT has been proposed as a general
solution by Raman (1924), who believed that the solution was exact.
His oversight was the following: if an ensemble of only two particles is
considered, the probability that these particles are at a distance r from
one another is truly e-<ll(r)/kT, but, if an ensemble of a large number of
particles is considered, the probability that any two particles are at a
distance r cannot be the same, since interactions with the other particles
must be taken into account.
A second approximation of equation 60 based on the more complete
theories of Born and Green [259) leads to the following result (Fournet
[44 ], [45), [49]):

I(h) = I,(h)N{F 2 (h) + F(h) 2 L'1(27T) _"fi~h)- e{J(h)

} (67)

where e designates a constant approximately equal to unity.

For the simple case of spherically symmetric particles, equation 67
becomes

Let us emphasize that equations 67 and 68 have been derived from

certain results of the theory of Born and Green with no additional
physical hypotheses or mathematical simplifications.
A simple outline of the derivation of equation 67 is as follows:
The modification added by Rodriguez (1949) to the simple calculation of Green
consisted in writing
(69)

In obtainingf(r) from equation 60 the value of the function tX{r) = e-<r>(r)/kT - 1 is

assumed to be different from zero only for small values of r, and in certain cases the
product tX(r) f{r) is assumed to be replaceable by the terms tX(r) (E - 1), thus defining
(e - 1) as a mean value of f(r) for small values of r. From this, the function f(r)
can be described by the Fourier transform

rf(r) =
l
v'2;
r 00 e 2h{3 2 (h)
J_°" v,( 217 )_
.
312 _ e{i(h) sm hr dh
(70)

neglecting terms inf2(r),f"(r), etc.

GENERAL THEORY 45

The calculation of v,(h) requires a knowledge of the function [P(r) - I]. From
equation 69,
P(r) - I = [o:(r) + I] [I + f(r) + · · · ] - I
or
P(r) - 1 "" £o:(r) f(r) +
Thus we can write

-
v (h) I (2,,.) 3 "
-'- = ,,,-- .
Vi Vi
2
1-
V 27T
l"'
0
r[Eo:(r) + f(r)] sin hr dh
which, by means of equations 65 and 70, becomes
_ v,(h) = ~ (2,,.)31 2 [£h/3(h) -+- h £ 2 /3'(h) ] (71)
Vi h v1 ' v1 (2,,.)- 3 12 - £{J(h)
Equation 67 is easily found from this last equation.
Let us try to analyze quite generally the different relations which we
have established by employing a criterion often used in the statistical
theory of ensembles (theory of fluids, the order-disorder transformations
in alloys, magnetism): the inclusion in the calculation of particles in units,
pairs, triplets, etc. We must point out that a perfect theory should take
account of all such possible groups.
Equation 55, correct for widely separated particles, takes account of
particles only in units; that is, all interactions between particles have been
neglected.
The discussion we have given of the work of Raman with respect to
equation 64 shows that this expression takes into account. particles
considered in units and in pairs; that is, interactions between particles
have been limited to a sum of interactions between pairs of particles
isolated in space.
In considering the validity of equations 67 and 68 we must first point
out that the Kirkwood-Boggs principle of superposition used in establishing
equation 60 is presented in the form of a relation, good to a first approxi-
mation, between the properties of particles considered in pairs and the
properties of particles considered in triplets. Now in the form that
Rodriguez has given to the theory of Born and Green, one determines the
second approximation to the functionf(r), defined by equation 60 and the
relation
P(r) = e-«>(r)/kTef(r)

The first appmximation, that of f(r) - 0, furnished the solution of

Raman, which takes into account only the influence of doublet terms.
Thus we can affirm that equations 67 and 68, established from the theory
of Born and Green, take account of triplet terms at least to a first approxi-
mation, contrary to the opinion of Oster and Riley [128].
To summarize these results, we have listed the expressions of certain
equations of state and the corresponding expressions for the scattered
intensity.
46 SMALL-ANGLE SCATTERING OF X-RAYS

Summary
EQUATION OF STATE SCATTERED INTENSITY
pV= RT l(h) = 1,(h)NF2(h)

pV =
(2.,,.)3f2(J(O)]
RT [ 1 - - - - -
2v1
l(h) = l,(h)N F 2(h) [ 1 + < 2 :~312 {J(h)]
--- 1
Born and Green l(h) = l,(h)NF 2(h) (2 .,,.) 312
1 - --E{J(h)
vl

2.2.3.3. Limiting Value for the Intensity Scattered at Very Small Angles
We have just examined the relation between the equation of state and
the scattered intensity, 1(h). Let us now restrict the problem to deter-
mining which thermodynamic variables are related to 1(0). From a
previously established relation, equation 53, we find

1(0) = 1,(0)N {n2 - n2 v~~)} = 1.(0)Nn2{1 - v 2~~)}

n being the number of electrons contained in a particle.
The function v 2 (0) can be evaluated as follows: we have seen (§2.2.1.3)
that Pk; dvk dv; represents the probability that there is at the same time
a particle in dvk and a different particle in dv1• The double integral of
Pk; dvk dv1 extended over the domain V should then give the average
number of pairs of particles existing in the volume V. Designating by N
the number of particles in V at a certain instant, we find that

rr
Jv Jv
Pk, dvk dv; = rr
Jv Jv
P(rk;) dvk dv; = N(N - 1) = N 2
v1 v1
- N
since the average of a sum always equals the sum of the averages of each
term.
Now by introducing the function (1 - P(r)) in terms of these quantities,
we have
- -= J~f
N2 -- -~ - f -~I
N
V1 V1 V1
[l - P(rk;)] -~
V1

By making use of the calculations employed in §2.2.1.4, and neglecting

only very small terms, we obtain the following expression:

N2- N = N2 - N f"'[l - P(r)] 411r 2 dr

.lo V1
from which we find

-
v1
1 i o
00
V2(0)
[1- P(r)]411r 2 dr=--
Vi
= - N2 +
N
N2 + N = 1- - - -
N2 -
N
N2
GENERAL THEORY 47

With this value of v2 (0), we find that

J(O) = J.(O)n2(N 2 - N2 ) = J.(O)n2(N - N)2 (72)
This result was first established by the work of Einstein and Smoluchow-
ski; other papers developing this topic which might be cited are those
of Zernicke and Prins [309), Bhatia and Krishnan (1948), Yvon (1947),
and Fournet and Guinier [53).
Equation 72, which is a result of the hypotheses H 1 and H 2 imposed on
the structure of ensembles of particles, shows that the observable scattered
intensity (seep. 35) at very small angles is a consequence of the existence
of fluctuations in the sample. A thermodynamic description of /(0) can
be found by recalling a classical result of the kinetic theory of gases, 1

kT
yfl+···=
(N-N)
~
2
=
(N-N) 2
N2 +·" (73)

where fl is the isothermal coefficient of compressibility,

fl=_~
V
(oV)
Op T

V being the total volume offered to the gas. If N is large and the matter
is not near its critical point, the terms neglected in equation 73 are small
when one writes
-kT
/(0) = J.(O)n2N - fJ (74)
V1
thus establishing a simple relation between /(0) and the isothermal com-
pressibility.
Values for /(0) have been predicted by several different theories. The
expression given by Debye for a model of hard spheres is (cf. equation 66)

/(0) = I.(O)n2N ( 1 - 8~0)

where v 0 is the volume of each sphere. (The coefficient, 8, has been

omitted in several references in the literature.) The maximum value of
v0/v1 , that for close-packed hexagonal or cubic systems, is 0.74, leading to
the prediction of negative intensities for a large domain (v 0 /v1 > 0.125).
The linear model of Kratky and Porod [108), with assumptions similar to
those of Debye, leads to a similar factor: (l -2l 0/l 1 ), [2 3 = 8]; since the
maximum possible value of l 0 /l 1 is unity, again negative intensities are
possible.
1 See, for example, R. C. Tolman, The Principles of Statistical Mechanics, Oxford,

1946, p. 647.
48 SMALL-ANGLE SCATTERING OF X-RAYS

Equations 67 and 68 give negative results if Vi is smaller than (27T) 3' 2

E/3(0). This is not a valid criticism, however, since Fournet [49] has
shown that the passage of the Zernl.cke-Prins type equation to equations
67 and 68 is possible only if there are no roots to the following equation:
Vi - (27T) 3f2 E{J(h) = 0
When roots to this equation exist, as happens for liquids, equation 67 is
no longer correct and must be replaced by another which does not predict
negative intensities (for details, see the article of Fournet cited above).

2.2.3.4. Thermodynamic Expression for the Intensity

Let us now consider the general possibilities of employing equation 68,

I(h) = I,(h)NF 2(h) \ 12 (68)

V1 - (27T) E{J(h)

in a study of spherical· particles.i Fournet [45], [49] has employed this

equation together with the potential energy function <l>(r) determined by
Lennard-Jones (1937) to predict correctly the scattering by gaseous and
liquid argon at 150° K.; this is shown in Fig. 11. When considering a
fluid of hard spheres, for which {J(h) can be calculated, equation 68
becomes
I
J(h) = I,(h)N<l>2(hR) Bv
(75)
l +- 0 E<l>(2hR)
Vi

The corresponding curves are given in Fig. 12 for various values of

(8v 0 E/vi)- Let us recall that, in practice, E can be taken as equal to I.
·we can now establish a precise criterion for defining the term "widely
separated particles." Equation 68 can be written in the form

I - 2 I
(h) - I,(h)N F (h) (271 )312 E{J(h)
1-----

which allows us to say that the equation relative to widely separated

particles is accurate to within n per cent when the ratio (27T) 312 E{J(h)/vi
has the value of n per cent.
A certain characteristic behavior of the scattered intensity can be
predicted from the mathematical structure of equation 68. If the value
1 We shall consider only spherically symmetrical particles in this paragraph in

order to simplify the discussion. For the general case, as expressed by equation 67,
the results are analogous.
GENERAL THEORY 49

l(h)

2 3 4 h(A.-1)
(a)

l(h) l(h)

(F 2(h) u = 0.200 A.- 3

.... , .... u = 0.143 A.- 3
'y,,..F 2 (h) ' '\
\
\
\ u = O.lOOA.- 3
\ \
\ \
\

2 3 4 2 3
(b) (c)

Fig. 11. (a) Experimental curves of the scattering by liquid and

gaseous argon at pressures near the condensation preBBUre at 149.3° K.
(Eisenstein and Gingrich [40]). The dotted curve represents the
square of the structure factor, F 2 (h). (b) Theoretical curve (equation
68) of the scattering from gaseous argon at 149.3° K. and at condensation
pressure. (c) Theoretical curves of the scattering from liquid argon of
several densities at 149.3° K. (u = (211/'•t:(v,)- 1 [see equation 68]).
.50 SMALL-ANGLE SCATTERING OF X-RAYS

of {:J(h) is zero when h equals some value h1 , the reduced intensity,

l(h)/NI.(h), evaluated at h1 is a constant, F 2(h 1 ), regardless of the con-
centration of the matter. This fact, predicted from equation 68, is
verified quite well by the results found for argon by Eisenstein and
Gingrich [40], when their results are considered on the basis of one

hR
Fig. 12. Scattering curve for non-interacting hard spheres (equation 75)_

temperature (see Fig. 7 of the reference cited); t.he condition of a common

temperature for the curves is necessary, since {:J(h) depends on the
temperature, as is shown in equation 65_
If the coefficient of F 2 (h) in equation 68 is considered, it is seen that
the maxima of this function always occur at the same angles (those such
that {:J(h) is a maximum), regardless of the concentration of scattering
matter; the only effect of a change in concentration is to accentuate
the maxima to a greater or lesser degree. This same result is found if
the expression derived from only the first approximation is used. (Per-
sonal communication from G. W. Brindley.)
A detailed study of equation 68 by Fournet [48) has shown that for the
general case, in which F 2 (h) decreases in the observable region with
increasing h, the intensity maxima are produced at larger and larger
angles, the greater the concentration of scattering matter. This is
illustrated in Fig. 13, in which we have plotted representative curves of
GENERAL THEORY 51

a(h)
IT\
/ I \
/ I \

/
I
/
.....-r--,\\ ',,.
I
I
I
I
Ihm h

F 2(h)
I
I
I
I
I
I
I
I
I

l(h)

Fig. 13. Scattering from a fluid for three different concentrations of

particles. The solid curves correspond to a very low concentration,
the dashed curves to an average concentration, and the dotted
curves to a high concentration. Note that the maxima of a(h),
the interpa.rticle interference function, and I(h), the observed
intensity, occur at different scattering angles.

the terms involved as a function of h for three values of v1 ; the solid

curves refer to the case of v1 approaching infinity (i.e., widely separated
particles), the dashed curves refer to a smaller value of v1 , and the dotted
cmves to still smaller values of v1 (i.e., still higher concentrations). The
first function, the interparticle interference function
52 SMALL-ANGLE SCATTERING OF X-RAYS

is equivalent to J(h) 2 ; the second is the function F 2 (h); and the

I.(h) NF (h)
third, the product F 2 (h)a(h), is proportional to the intensity. It can be
seen that the product of the continually decreasing function F 2 (h) and
the function a(h), with its only slightly accentuated maximum, results in a
function J(h) having a still more diffuse maximum situated at a smaller
value of h than that for the original function, a(h), for each of these cases.
The position of the observed maximum thus depends markedly on the
function F 2 (h), that is, on the structure of the pa.rticle. The position
of the intensity maximum depends in a very complex way on the
structure of the arrangement of the particles and on the particular structure
of each particle.

2.2.3.5. Fluids and Crystals

In an examination of a crystalline substance by means of an experi-
mental method such as Debye-Scherrer photography, we find that the
function for a perfect crystal which plays the role of a(h), that is, the
function I h 1_t'~ 2 (h) , is identically equal to zero except for certain
.( )
specific values of h, at which points it takes on very large values. The
product of this function with the function F 2 (h) then gives a function
I(h) which shows the same structure as a(h}, in that it also is identically
zero except for certain specific values of h. The positions of the intensity
maxima (the Debye-Scherrer lines in this example) are identical to those
of the function a(h) and thus can immediately furnish information on the
structural arrangement of the particles (see Fig. 14). This illustrates
one of the essential differences between the classical problems of X-ray
crystallography and the problems that are treated here: the degree of
order in a crystal is in general such that the function J(h)/F 2 (h) presents
sharp maxima. As a result the maxima of the function J(h) occur at
the same values of h as the maxima of the function J(h)/F 2(h). This
result does not apply to fluids, for they are much less ordered than any
crystal.
We should like now to offer a physical explanation of the fact that the
maximum of the function J(h)/F 2 (h) is produced at a constant angle for a
fluid, independent of its concentration. For this very qualitative
explanation let us make the approximation that P(r) = e-/!}(r)fkT. The
factor of physical importance in this problem is the arrangement of par-
ticles around any one particle. At very low concentrations the probability
of finding a particle in a volume element dv 1 is dv 1/v1 • If it is known that
this element, dv1, is at a distance r from another particle, the probability
GENERAL THEORY

a(h)

l(h)

h
Fig. 14. Scattering from perfect crystals. The maxima of a{h) and
l(h) occur at the same scattering angles.

becomes (dv1/v 1)e-lfl(rl/kT, and the probability density, d(r), from its
definition, is (l/v1 ) e-ll>(r)/kT. We have plotted this probability density
in Fig. 15 for two concentrations, v1 ' and v1 H. Whereas the mean
probability density increases when the concentration increases, the
ratio of probability densities for arbitrarily chosen r 1 and r 2 remains a
constant. Since the position of the maximum of l(h)/F2(h) is related to
this ratio of probability densities, it can thus be understood that the
position of the maximum will remain inva.riant.
54 SMALL-ANGLE SCATTERING OF X-RAYS

To summarize, we have shown that there is a great deal of difference

between the behavior of the function directly connected to the arrange-
ment of particles, /(h)/F 2 (h), and the function I(h) which is observed
experimentally. In §4.1.2.2 we shall return to this point to discuss its
important consequences in the interpretation of experiments.

d(r)

r
Fig. 15. A scheme.tic representation of the probability density d(r)
for two different concentrations of pe.rticles.

2.2.3.6. Secondary Maxima

In the course of experiments on the scattering of X-rays by suspensions
of latex, Yudowitch [186] and Danielson, Shenfil, and DuMond [25] found
several maxima in the scattering curve (Fig. 16). A study of this same
suspension by electron microscopy showed that the latex globules were
spherical and very regular in size, variations in diameter being less than
10 per cent. These globules were relatively close-packed, so that an
explanation based on the theoretical development of §2.2.3.4 can readily
explain the principal maximum, the maximum at the smallest angle.
The other maxima, which we shall refer to as secondary maxima, appear
to be more difficult to interpret.
Yudowitch advanced the hypothesis that these secondary maxima were
due to the particular form of the square of the particle structure factor,
F 2 (h). Let us re-examine the function F 2(h), considering the latex
globules as analogous to hard spheres of constant electronic density, this
being the same model that we have heretofore designated as "Debye's
model."
GENERAL THEORY 55

The intensity scattered by one such sphere is given by the relation

F 2 (h) = 11> 2 hR = [ sin hR - hR cos hR] 2

( ) 3 haRa (31)

The positions of the maxima and minima of this function are then given
by the solutions of the equation

21l>(u)ll>'(u) = 21l>(u) (u
2 - 3) sin u + 3u cos u = 0
u'

Fig. 16. A microphotometer curve of the diffraction pattern of latex

particles showing the secondary diffraction rings. Intervals at the
bottom correspond to a scattering angle of 3 minutes of arc (Danielson,
Shenfil, and DuMond [25]).

where u = hR. The solutions corresponding to the minima are those for
which ll>(u) = 0. The positions of the maxima are given as solutions
of the equation

(76)

At these positions we note that

• 2 9
$2 u - sm um -
( ml - u
m
2 - u
m
4 + 3u 2 + 9
m
(77)

A first approximation to the solutions of equation 76 is found by

placing um= kTr, where the first maximum is that fork= 0, the second
for k = 2, the third, k = 3, etc. A better, second-order approximation
(except fork= 0) is obtained in writing
(78)
56 SMALL-ANGLE SCATTERING OF X-RAYS

the corresponding intensities being given by

9 81
l(um) = k'Tr' + k67r6 (79)

Thus we can establish the following table:

Index of Maximum 1 2 3 4 5 6
k 0 2 3 4 5 6
"m• equation 78 5.80 9.10 12.33 15.52 18.69
"m• exact 0.000 5. 76
103 x Im, equation 79 7.37 1.28 0.39 0.15 0.07
103 x Im, exact 1000 7.45
It should be noted that the second-order approximation gives co1Tect
values even for k = 2.
Evidently the ratio of I(um) at its first maximum, at zero angle, to the
value of l(um) at its second maximum is very large, actually a value of
1.3 X 102 , but, more important, the ratio of the intensities of any two
other successive maxima is small; the ratio of the second to the third is
only about 5. It is thus possible to observe experimentally several of the
maxima that ·theory predicts for the case of widely Beparated particles.
Next the effect of interparticle interference should be considered. We
shall employ equation 75, which has been established for the case of
particles with no mutual interaction other than impenetrability; though
this represents an approximation, it should be sufficient for the larger
angles. Neglecting constant factors (see §2.2.3.4), the scattered intensity
is
<f>2(u)
l 2(u) = - - - - -
8v0
1 + -V1
e<l>(2u)

Expanding this function, l 2 (u), around the point u = kTr x, we find +

that to a second approximation this function is a maximum for
u = kTr - (3/k7r} + ...
the same result as that found for the maxima of <1> 2 (u}. The values of the
intensity maxima are (cf. equation 79):
9 81 27c
l2(um} = k'n' + k67T6 + 4k67r6
where c designates the ratio (8v0 e/v1 }, which has a maximum value of
about 6.
We see thus that for this model the poBitionB and magnitudeB of the
Becondary maxima are only Blightly modified in paBsing from a very dilute
BYBtem to a denBe ByBtem. This is not true for the principal maximum;
the principal maximum occurs at zero angle for dilute solutions, and as the
GENERAL THEORY 157

concentration increases it is displaced towards larger angles, occurring at

values of u between 0 and 2.5 for systems of average concentration.
The explanation of the secondary maxima observed by Yudowitch is
thus given by a complete calculation of interparticle interferences. These
maxima are present in the representative curves of F 2(h) for a single
particle, and interparticle interferences, instead of removing these maxima,
actually reinforce them slightly, as is shown in Fig. 17.

l(h)

10·10- 3

5 hR
Fig. 17. Scattering from hard spheres. The solid curve represents
the intensity scattered by a single sphere of radius R at large value of
hR (see Fig. 6). The dotted curve represents the scattered intensity
per sphere for a dense group of spheres ((8t10E)/v1 = 15).

We must point out that this explanation of the secondary maxima is

based on the choice of a model of the particles, but the hard-sphere model
seems particularly valid for suspensions of latex globules.
Let us now consider as a second example the secondary maxima that
can be observed in the scattering curves of liquid or gaseous argon as
determined by Eisenstein and Gingrich (Fig. 11). A comparison of
these curves with the square of the structure factor shows immediately
that these secondary maxima are due uniquely to interparticle inter-
ferences.

2.2.3.7. Remarks on Fourier Transformations

When considering two mutually reciprocal spaces that are connected
by a Fourier transformation, as, for example, the real space containing the
particles and the reciprocal space of the variable h, in which densities a.re
58 SMALL-ANGLE SCATTERING OF X-RAYS

related to scattered intensities, it is known that an unevenness in density

in one space corresponds to a periodic variation of density in the other
space, with the period in the second space being related to the position
of the unevenness in the first. We know thus that a Debye-Scherrer line,
which is a discontinuity in h space, is determined by the periodic dis-
tribution of certain crystallographic planes in real space, and the position

Real space Reciprocal space

p(r)

Hard
sphere
Poi------~

R r h

p(r)

Argon
atom

r
Fig. 18. Schematic curves of p(r) and F•(h) for a hard sphere and
for an argon atom. The unevenness in p(r) for the hard sphere causes
a certain periodicity in its F"{h).

of the line is determined by the period, or distance between consecutive

planes, in real space (Bragg's law). These remarks can now be applied in
a discussion of the intensity scattered by a hard homogeneous sphere and
by an atom of argon. Both particles are characterized in the real space
by the electronic density p(r) at a distance r from the center of a particle
and in the h space by the function F(h), related to p(r) by a Fourier
integral (see §2.1).
Since the density p(r) is more uneven, in a general sense, for hard
spheres than for argon atoms, the graph of F 2(h) for hard spheres will
GENERAL THEORY 59

demonstrate a certain periodicity not found in the corresponding curve

for argon, as is seen in Fig. 18.
Let us now consider an ensemble of hard spheres and an ensemble of
argon atoms. The effects of interparticle interferences, which must be
taken into account, will be determined in real space in terms of the
potential <l>(r) associated with the forces acting between two particles

if?(r) fj(h)

Hard
sphere

r h

if?(r) fj(h)

Argon
atom

Fig. 19. Schematic curves of tll(r) and {J(h) for hard spheres and for
argon atoms. The argon interaction potential is the more uneven, so
its function {J(h) has the more marked periodic character.

separated by a distance r. We have previously shown that the influence

of <l>(r) is felt through the intermediary of a function {J(h), which is defined
as the Fourier transform of or;(r), where or;(r) = e-IJJ(rl/kT - 1. Thus we
can say immediately that, since the curve of ¢l(r) with respect to argon is
the most uneven, the function fj(h) of argon will have the more marked
periodic character. This is shown in Fig. 19.
The scattered intensity is found by combining the functions F 2 (h) and
p(h) (see §2.2.3.2). The above discussion shows why the secondary
60 SMALL-ANGLE SCATTERING OF X-RAYS

maxima from argon are due almost entirely to interparticle effects,

whereas form and internal structure are the predominant factors for the
case of hard spheres.

2.2.4. PARTICLES UNRESTRICTED BY HYPOTHESES H 1 AND H 1

We have assumed from the beginning of §2.2 that the scattering
matter being examined satisfies hypotheses H 1 and Hz. We should like
now to remove these restrictions. If the details of the calculations which
led to equation 47 are considered, it is seen that the second part of our
hypothesis-that which excludes all possible relations between relative
positions and probabilities of orientations of particles (i.e., all possible
relations between positions and structure factors)-has enabled us to
calculate the averages by describing the average of the factor

as the product of the averages of F1:(h), F 1(h), and cos (h • (R1: - R 1)).
The first part of hypothesis H 1 gave knowledge of the average of F(h),
which was convenient for our considerations.
We should like now to try to consider the general problem in order to
determine the characteristic magnitudes that are involved in this
question.
The general relation, equation 46, shows that only information relative
to pairs of partides is necessary. Thus it is sufficient to define the
statistical correlations existing between two particles. We shall describe
these by means of the development offered by Fournet [48]:
p 1(F, h) designates the a priori probability density function of the
scattering factor F of a particle for a scattering angle corresponding to h.
We shall assume this function to be identical for all particles. If we have
no information concerning the surroundings of a particle, the probability
that its scattering factor for a given value of h is contained between the
values F 0 and F 0 + dF 0 is equal to p 1(F0 , h) dF 0 •
p 2(F 1, Fk, r, h) designates the probability density function of the
scattering factor F; of a jth particle for a scattering angle corresponding
to h when it is known that the scattering factor for the same angle of a
kth particle has a value F1c, where Rk - R1 = r. As a consequence of the
indistinguishability of particles, p 2 must be the same for the same vector
r, regardless of the position of the center of the kth particle.
There is no contradiction in stating that the probability density p 1 is
the same for all particles, while defining p 2 as has been done above, if it
is realized that the function p 2 concerns ensembles of factors Fk and F 1,
GENERAL THEORY 61

whereas p 1 governs the factor Fk or F 1, considered by itself. The relation

between p 1 and p 2 can be expressed as

£ 00
P2(F;, Fk, r, h)p1 (Fk, h) dFk - p 1 (F;, h) (80)

Now let us consider the intensity of radiation scattered by an ensemble

of identical, arbitrarily shaped particles,

/(h) = l,(h)2:2:Fk(h)F,(h) cos (h · (Rk - R;)), (81)

k j

We can group separately the terms for which. k = j. An analogous

problem has been treated in §2.2.1.2, in which tlie result for the sum of
these terms was shown to be

I,(h)N F 2 (h) = I,(h)N L 00

F 2p1(F, h) dF

In the general term of equation 81 for which k =I= j there are found
three types of variables:
1. The scattering factor Fk (F; is related to Fk by means of the functions
P2l·
2. The angle Z)_h(Rk - R1).
3. The distance Rk - R1 I I·
In order to evaluate the group of terms for which k =I= j we shall perform
successive integrations over each of these variables.

Average of the Variable F k

Let us consider first only the functions F1; as variables. The group of
terms of equation 61 for which k =I= j can now be written as

2: 2: {cos (h · (Rk - R;))

k Nk Jof '° Fkp1(Fk, h)· Jof"' F 1p 2(F;, Fk, r, h) dF1 dFk}
(82)
We shall assume that correlation between values of the scattering factors
does not occur for large distances of separation; the knowledge of Fk gives
I I
no information about F 1 if Rk - R1 is large compared to nearest
neighbor distances.
It now follows that p 2 (F 1 , Fk, r, h) tends toward the function p 1 (F 1 , h)
for large r. Thus it is useful to express the quantity p 2 as
p 2 (F;, Fk, r, h) = p 1(F 1, h) - {p 1(F1, h) - p 2 (F;, Fk, r, h)} (83)
On replacing p 2 by this expression, equation 82 breaks down into two
terms:
62 SMALL-ANGLE SCATTERING OF X-RAYS

First term: This term describes the scattering when the structure
factors Fk and F; follow the same law of probability, independent of one
another. We have already solved an analogous problem (§2.2.1.2),
which gave as a result

F(h) 2 {L L cos (h. (Rk - R;))}

kN•k

where F(h) = i 00
Fp1 (F, h) dF

Second term: This term is of the form

- L L {b(r, h) cos (h • (Rk - R;))}
kh'lc
where

b(r, h) = f" {'" Fkp 1 (Fk, h) {p1(F;, h)- p 2 (F;, Fk, r, h)}F1 dF;dFk (84)

The bracketed term in the integrand approaches zero when becomes Ir I

large, and consequently the function b(r, h) behaves in the same manner.
Thus in evaluating this second term we can neglect boundary effects and
treat the summation over the index j as independent of k:
- L L b(r, h) cos (h • r)
kr;•O

where the sum extends over all vectors r which exist in the sample (except
r = 0). The sum over k then results simply in multiplying this result by
the average number of particles. The final expression for the scattered
intensity is then

J(h) = J,(h)N{F 2 (h) - L b(r, h) cos (h • r)}

r;<O

+ J.(h)F(h)2 {L L cos (h. (Rk - R;))} (85)

k j;<k

or
J(h} = J,(h)N{[F 2 (h) - F(h) 2 ] - L b(r, h) cos (h • r)}
r;<O

+ J.(h)F(h) 2 LL cos (h. (Rk - R;)) (86)

k j

Average with Respect to the Angle .zS_ h (Rk - R1)

Keeping the magnitudes of the distances fixed, let us assume that all
orientations of the vector r are equally probable. The function p 2 then
depends only on the distance r between centers k and j. The terms in
cos (h · r} on averaging will then be replaced by terms in sin hr/hr (a
GENERAL THEORY 63

calculation illustrated several times, viz., equation 7). The resulting

expression for the intensity is

/(h) = I.(h)N{[F 2 (h) - F(h)2] - ,L b(r, h) sin hr}

r7"0 hr

(87)

Average with Respect to Distance r

For this final averaging we proceed as in §2.2.1.3, defining a function
P(r) which is related to the probability of finding the centers of two
different particles at a distance r. When no external field is applied, ·an
orientations of a vectorial distance r are equally probable, so that we can
begin with equation 87, modified only so that the term -NI.(h)F(h) 2 is
included in the double sum over k and j. Analogous problems have
already been treated in §2.2.1.4.
The summations of equation 87 must be replaced by integrations, taking
into account the probability of realization of the different distances, r.
By replacing P(r) with 1 - (1 - P(r)] and considering only angles for
which h > h 0 ( §2.2.1.4), only the term in (1 - P(r)] need be considered
in the third term of equation 87. This reasoning cannot be applied to the
second term of equation 87, since b(r, h) tends to zero as r becomes large.
The final expression is thus (Fournet (48])

-/(h) = - {-- F(h)2

I.(h)N F 2 (h)- - -
V1
i""
0
sin hr
[ 1-P(r)]-h- 4?Tr2dr
r

- -l
V1
i""
0
P(r)b(r, h) sinhr
- - 4?Tr2 dr}
hr
(88)

The first two terms of this expression are identical to those derived on
the assumption of complete independence between the orientations and
the positions of the particles. The term in b(r, h) thus appears as a
corrective term, necessary for the description of the general case.
The complete expression, equation 88, must be used, for example, in
calculating the intensity scattered by a dense ensemble of identical
ellipsoids of axes 2R, 2R, and 2vR. The scattering factor of an ellipsoid
for a given scattering angle, 20, depends on the orientation of the ellipsoid.
If it is known that the distance between the centers of two ellipsoids is
2R, the axes of these ellipsoids cannot be oriented in a completely arbi-
trary manner, and, consequently, their scattering factors must be related.
We can see by this example how a relation between scattering factors
SMALL-ANGLE SCATTERING OF X-RAYS

Fk and F; and the distance between their centers can be introduced. It

is this dependence which introduces the supplementary term in equation 88.
We have just seen that in the general case, in which the orientations
and the positions of particles are related, the expression for the scattered
intensity contains two characteristic functions, P(r) and b(r, h). This
makes it impossible to determine separately these characteristic functions
from the experimental curve of J(h). Equation 88 can serve only to predict
intensities for certain models.
T.o proceed further it would probably be necessary to introduce thermo-
dynamic considerations, defining a potential energy of interparticle forces
not as a function <l>(r) but rather as <l>(r, qi), where at least one angular
variable is necessary, and then relating the functions P(r) and b(r, h) to
this potential. This problem has not yet been undertaken.
Remarks
In establishing equation 86 we have allowed only the structure factors to undergo
variations, so that a simple interpretation can be given to this equation by applying
it to a study of crystals and neglecting therme.1 effects.
The third term of equation 86 represents the intensity sc11.ttered by the s11.mple
under inspection, if it is supposed th11.t e.11 the scattering centers are identical, with
scattering factors F(h). The first term varies only slowly with h, while the second
term can present intensity maxima that will probably be less sharp than those
created by the third term. From this we can see the essentie.1 role of the function
b(r, h), a function that is analogous to the Patterson distribution function for crystal
structure analyses. If we assume that each of the three intensity terms can be
separated, all the information obt.ainable from experiments will be contained in the
function b(r, h).
In the study of crystals it is often assumed that the structure factors of different
atoms all follow the same law, F(h). the magnitudes depending on a coefficient eque.l
to the atomic number Z of the atom considered. If we assume this to be true, e.11
the formulas we have developed can be considerably simplified, for the functions
p 1 . and p 1 can be treated as being functions uniquely of the Zk.
The function b(r, h) becomes
b(r, h) = b(r, h) = F 1 (h)I:I:Z.tP1(Z.) [p1(Z1) - p,(Z1, z •. r)]Z,
kj
= F 1 (h)c(r)
The final result is then (Fournet [48])
J(h) = I,(h)N F 1(h){[Z' - Z1J - I:c(r) cos (h • r)}
¥;"0

+ J,(h)F 1 (h)Z1 I:I: cos (h • (R. - R1)) (89)

k j
The function c(r) can e.lways be calculated from any given model. We have
employed this technique in the recalculation of the scattering by a linear model of
an e.lloy with partie.l short-range order, a model first studied by Guinier and Griffoul
(1948). We quickly obtained the same results as those given by these authors.
Equation 86 can be easily used to obtain the intensities scattered by more compli·
cated models. This equation, a particular case of which has been given by
GENERAL THEORY 65

Mac Gillavry and Strijk (1946), can be viewed as a generalization of the expression
obtained by von Laue (1941) in considering completely disordered metallic solid
solutions, in which all scattering centers were assumed to follow independently the
same law. When bound by this assumption, p 2 is equal to p 1 and the function
b(r, h) becomes identically zero. The second term of equation 86 disappears and,
on noting that the two types t>f atoms, A and B, are present in proportions p,,_ and
PB the Laue equation is obtained:

F2 - F2 = P.AFA'l. + PBFB 2 - (pAF A + PBFBF"

= PA-PB(F A - FB) 2

2.3. SCATTERING BY GROUPS COMPOSED OF SEVERAL TYPES OF

PARTICLES
This problem is quite complicated, and the few equations that can be
derived have not yet been employed in experimental studies.

2.3.1. GENERAL THEORY

Simple calculations, in every way analogous to those which have been
developed in §2.2.1.4, give the following relation for the scattered intensity
(Fournet [48]):

I(h) = I.(h)N { fpkFk 2 (h)

- -- 1
+ Lk jLPkP;Fk(h)F;(h) -
V10
i 00
[Pk1(r) -
sin hr
l] - - 4rrr 2 dr
hr
}
(90)

where Fk(h) designates the structure factor of the particle of type k; pk,
the probability that one of the N particles is of the type k; v1 , the average
volume offered to each particle, regardless of its type; and Pk;(r), a
probability function analogous to the function P(r) defined in §2.2.1.3,
which applies to a pair of particles of type k and j. Evidently
Pk1(r) = P1k(r)
This general expression can best be appreciated by comparing it with
equation 53, the expression applicable to an ensemble of identical particles.
From the first terms we obtain

F2(h) = LPkFk 2(h)

k

while from the second terms

and
66 SMALL-ANGLE SCATTERING OF X-RAYS

By a simple substitution we now find

LLPkP;Fk(h)F;(h)Pk;(r)
P(r) = _k~3-·- - - - - - - -
LLPkP;Fk(h )F;(h)
k j

which demonstrates that in the general case a function equivalent to

P(r) but dependent on the single variable r does not exist.
It is equally impossible to obtain information by means of a Fourier
transformation. Equation 90 can be written in the form

The right side of the equation depends on the variable h not only in the
term sin hr/hr but also in Fk(h) and F;(h), and this prevents the effective
application of a Fourier transf~rmation. Results can be obtained by this
technique only if the assumption is made that Fk(h) = ad(h), that is,
that the functions F;(h) differ only by a constant factor. On making
this assumption, we find

L LPkP;akaiPk;(r)
with P(r) = _k_i_ _ _ __ (92)
LLPkP;aka;
k j

from which the function P(r) can easily be obtained by means of a Fourier
transformation.
With the exception of this case, which itself would be hard to interpret,
it is difficult to use equation 90 without supplying some model.

2.3.2. WIDELY SEPARATED PARTICLES

'-'Then the concentration of scattering matter becomes small equation 90
reduces to the following form:

l(h) = I.(h)N LPkFk 2 (h) (93)

k

which describes the total intensity as being given simply by the addition
of the intensities scattered by each of the different types of particles, each
GENERAL THEORY 67

weighted by its respective probability. On expressing each Fk 2 (h) by the

approximate law of Guinier (equation 39), equation 93 becomes

(94)

This relation shows that the total curve can still be represented by an
exponential function if the Guinier approximation is valid for all of the
individual particles, particularly those that are the largest.
These conditions are rarely satisfied in practice. Thus, it is more
interesting to consider the tails of the scattering curves. The principal
part of the curve of Fk 2 (h) at large values of h for a homogeneous particle
of density Pk and external surface Skis given by the function (2TTpk 2Sk)/h 4
(equation 26). The extra terms acting as damped oscillations contain
the functions cos hRlk and sin hR1k, where Rlk is the maximum dimension
of the kth particle. When the curves for a large number of particles of
different dimensions are added together, it is probable that the sum of the
oscillating terms will be zero. In fact, such oscillations have never been
observed on a scattering curve unless the particles of the sample were
extremely uniform in size (Fig. 16). The asymptotic behavior of the
observed curve is then

(95)

This shows that, for a given angle within the domain of validity of
equation 26, the scattered intensity is proportional to the total surface of
the group of diffracting particles if all particles have the same electronic
density.
In this discussion we have assumed that the particles were separated
sufficiently so that there were no interparticle interferences. In §2.4.3 we
shall see that equation 95 is also valid for the case of packed powders.

2.3.3. INFLUENCE OF THE CLOSER PACKING OF PARTICLES

Evidently one can try to effect the same sort of generalization of
equation 90 as was done with equation 53, the equation of the scattering
relative to a single species of particle. The first step is to generalize the
theory of Born and Green (Fournet [51], and Rushbrooke and Scoins
68 SMALL-ANGLE SCATTERING OF X-RAYS

(1951)) so as to allow the consideration of this case. The general

expression can be obtained in placing
- cJ>k1(rl +hi«>
Pk;(r) = e kT
Next the functions g(h) and /J(h) are defined as

hgk;(h) = 1
---= Joorfk1(r) sin hr dr
v'27T - 00

h/Jk;(h) =-= 1
00

v'27T - 00
fr(e-cJ>k1(r)/kT -1) sinhrdr

These two functions are connected by the set of relations

where the E;k designates a mean value of (f;k(r) + 1) near the origin,
r=O.
With the introduction of these functions, equation 90 becomes

-{ - (277)3/2 -- }
I(h) = I,(h)N 2._pkFk 2 (h) +- - 2..2..PJcP;Fk(h)F;(h)[gk1(h) + €kA 1(h)]
k Vl k j
(96)
The scattered intensity then depends only on the functions Fk(h) and
/Jk;(h), since the gki(h) are expressible in terms of the /J 1m(h). We have
given the complete expression for the intensity scattered by ensembles of
two types of spherically symmetric particles in another article (Fournet
[48]). This complicated equation was applied to mixtures of homogeneous
spheres of the same matter but of radii R and 2R; the curves representing
the variation of scattered intensity as a function of angle are given in
Fig. 20. Two parameters were included in the calculation: k, the ratio
of the volume occupied by the particles to the total volume offered them;
and x, the ratio of the mass of smaller particles to the total mass of the
particles. Short-dashed curves correspond to k = 0.5, long-dashed
curves to k = 0.125, and full-line curves to the case of infinitely separated
particles.
For each concentration, x per cent, the curves have been normalized
so that the ordinate at h = 0 is equal to unity. The essential feature of
the curves is that for constant k, the intensity curves are more sharply
varying, the more homogeneous the mixture.
We believe that this statement is generally true; it is difficult to
conceive of a not too compact heterogeneous mixture manifesting a
GENERAL THEORY 69

I I

hR hR

I I

4 hR 4 hR

I I

x=O%

4 hR 4 hR
Fig. 20. Scattered intensity from mixtures of spheres of radii R and
2R. The parameter x denotes the ratio of the mass of the small spheres
to the total mass of the mixture, and k represents the ratio of the
effective volume of the spheres to the volume occupied by the
mixture. - - - , k = O; - - - , k = 0.125; -----, k = 0.500.
70 SMALL-ANGLE SCATTERING OF X-RAYS

reasonable degree of order. In our model, the curve fork= 0.5 possesses
a maximum only if the mass of smaller particles represents more than 95
per cent or less than 25 per cent of the total mass.
We must remark that this model was based on spherical particles, and
that consequently there was a favorable opportunity for observing intensity
maxima. In generalizing the conclusion drawn from a study of this
example, it thus would be a temptation to state: in a mixture of non-
identical particles of arbitrary forms and with random distribution (no
long-range order) it is improbable that the packing of particles will lead to
large changes in the scattering curves and that thus the laws for widely
separated particles can furnish the orders of magnitude of the scattering
phenomena.
This idea is in agreement with the calculations of Hosemann [81],
[84], who showed that, for arbitrary particles and conveniently chosen
functions PM(r), the packing of the particles caused little change in the
distribution of scattered intensity. We refer the reader to these works
for the details of the calculations and results.
Conclusions contrary to these have been given by Kratky and Porod
[108], [137], in considering the influence of packing on an ensemble of
heterogeneous particles. However, they considered uniquely a linear
model composed of a series of parallel plates of different thicknesses
situated at variable distances from one another. In the limiting case the
space is completely occupied by the plates, causing the central scattering
at observable angles to disappear entirely. The packing of particles thus
creates notable changes in the distribution of scattered intensity. We
believe that this is a rnsult which depends on the linear character of
the model.

2.4. GENERAL CASE

There are often substances which give rise to strong small-angle scatter.
ing that cannot be described as a group of well-defined particles arranged
in a more or less close-packed fashion. This is true, for example, of
matter which displays submicroscopic porosity, such as activated carbon.
The physical characteristic that can accurately define these substances is
the electronic density p(x) found at the point defined by the vector x.
The intensity scattered by such a substance is then given by the general
relation
(97)

V being the volume irradiated by the X-rays. It is well known that it is

not possible to determine p(x) from the experimental data. Indeed,
GENERAL THEORY 71

low-angle scattering experiments made with different samples taken from

a given material give the same experimental curve. Thus it is obvious
that the central scattering depends on some statistical property of p(x)
that defines the state of heterogeneity or porosity of the substance. It is
particularly interesting to try to determine the general characteristics
that are necessary for the production of an observable scattering at small
angles. When considered in this general form, the problem presents
special difficulties; we shall first discuss these difficulties and then present
the results of the attempts that have been made in this field.

2.4.1. LIMITING VALUE OF THE SCATTERED INTENSITY AT VERY

SMALL ANGLES
The property we shall try to calculate is the "experimental" limit of
the scattered intensity as the scattering angle tends to zero, that is, the

l(h) l(h) l(h)

l(O)

M N
I(h-0) - - - - - - - - - - - - • : - - -

ho h ho

Fig. 21. (a) Schematic representation of a real scattered intensity

distribution. (b) The observable intensity and its extrapolated value
at h = 0. (cl A possible result for a calculated intensity distribution.

intensity that can be obtained by extrapolation of the results of measure-

ments which, as will be seen later, cannot be extended to angles smaller
than several minutes of arc, even with the most perfect experimental
system.
Let us recall some results of our discussion of an ensemble of particles.
The curve of the scattered intensity shows a very important singularity
in the neighborhood of h = 0, since for extremely small values of h the
amplitudes of the waves scattered by all the electrons in the scatterer add
together and the scattered intensity approaches the value,

/(0) = I,(O)n 2N 2 (98)

The width of this central peak, defined by the parameter h 0 (see Fig.
2la), depends on the dimensions of the volume V explored by X-rays and
is smaller by several orders of magnitude than the minimum observable
angle.
72 SMALL-ANGLE SCATTERING OF X-RAYS

A correct calculation of the desired "limiting intensity" should give a

formula which reproduces the real curves down to a value of h equal to h2 ,
the experimental limit, and then, eliminating the central peak, remains
practically independent of h (Fig. 2lb) in the range between h2 and 0. This,
for example, is the result obtained in the calculations of §2.2.3.3, which
pertained to the particular case of an ensemble of particles contained in a
volume V 0 large with respect to the volume V of the scatterer explored
by the X-rays. The result of this calculation gave
(72)
Frequently it can happen that, in trying to eliminate the central peak,
a term will be discarded in the course of the calculation which will bring
about a marked change in the curve, such as that depicted in Fig. 2lc.
Such a formula does not give correctly the limiting value of the intensity
when h is made equal to zero. It can be quite correct for h > h 0 , but it
will not be useful, since the intensity cannot be easily calculated if h is
non-zero.
In the case of an ensemble of particles, the intensity 1(0) for h = 0 is
of the order of N 2 , while the limiting value should be of the order of
N 2 - N2 , that is, of the order of N. (This limit is exact for the case of
ideally separated particles, viz., equation 55.) Thus we see that in the
region of very small angles the limiting intensity, which will be expressed
as the difference between the exact expression for the intensity and another
term which must be made more definite, is of the order of l/N times the
actual intensity. This order of magnitude indicates immediately tlie
rigor and exactness that must be maintained in all phases of tlie calculations.
As an illustration of the effects of a slight inexactness, let us reconsider
the reasoning that we have developed in §2.2.1.3 and §2.2.1.4. We have
assumed that the volume V 0 is large with respect to the volume V. If
we vary V 0 (and thus N 0 , the total number of particles), keeping V and
v1 constant, equation 72 predicts a limiting intensity of zero when V 0 = V;
this is then a good example of the type of curve represented in Fig. 2lc.
In order to obtain a relation which will more accurately describe this
case, the manner in which the function P(r) was introduced must be
reviewed, noting specifically that the probability PkJ dvk dv; is defined
by the relation
N0 N0 - 1
Pk; dvk dv; = - dvk P(rk;) dv;
V0 V 0 - dvk
The infinitesimal volume element, dvk, is always negligible as compared
with V 0, but this cannot always be said for the comparison of unity with
respect to N 0 , particularly when V and V 0 are approximately equal (and,
GENERAL THEORY 73

thus, N and N 0 ). On reconsidering the complete calculation, beginning

with the relation

we obtain as an expression for the observable intensity (cf. equation 53)

l(h) = l 0 (h)N {F 2(h)- F(h) 2 !_ No - 1 f'°[l - P(r)]sinhhr 471'1'2dr} (99)

1v N0 Jo r

In the limit as h - 0, we then find

(100)

which gives as an expression for the limiting value of the observable

intensity (Fournet [48])

l(h - 0) = I.(O)n 2 [ (N2 - F) + ~ J-2

= I.(O)n2 [ N2 - N2 + N ;J (101)
- -2 -
The factor (N 2 - N) is of the order of N, so that, when V/V 0 is small
with respect to unity, equation 101 is to a first approximation equivalent
to our earlier result, equation 72. The advantage of equation 101 is that
a reasonable result is permitted in the case for which V = V0 ,
lv=v 0 (h---+ 0) = I.(O)n 2N
We can see by this example how approximations which at first sight
are quite logical can completely upset the results of this type of calculation,
We have defined the limiting intensity by an extrapolation of a part of
the curve of Fig. 2lb, but we have not yet determined whether this
quantity has any physical meaning. We know, for example, that the
actual curve has a shape which depends on the form of the sample,
whereas the limiting intensity should depend only on the statistical
properties of the distribution of scattering centers in the sample. It is
not obvious that the operation of extrapolation as described should lead
to such a result. In any event, we have not yet clearly defined a criterion
for the determination of the extrapolated part of the curve (part MN of
Fig. 2lb). Questions such as these form the obstacles encountered when
an attempt is made to resolve the problem of low-angle scattering as
calculated from the electronic density function.
As a first approach, let us offer without real proof a qualitative
74 SMALL-ANGLE SCATTERING OF X-RAYS

treatment; then we shall present the solution given by Debye and Bueche
[322] and that of Porod [137] for a more particular case.
We want to calculate the scattered intensity for a value h2 near the
minimum observable angle (for example, 2(} = l'). Let us divide the
irradiated volume V into a series of 1W equal volumes, v1 , v2 , • • • Vw

where each volume is of the order of (~:) 3 . (In the example chosen,
V; 1:::; 1µ 3 for Cu K-x radiation; this is still small with respect to the
volume V which is of the order of 1 mm 3 . in ordinary experiments.)
The amplitude of the radiation scattered by the volume element V; is then
A;(h). We shall assume that there will be negligible interference effects
among the waves scattered by the different volume elements when
h > h 2 ; this hypothesis is reasonable, for the phase difference between
waves will always be much larger than 277, since the centers of the volumes
are separated by distances greater than 277/h2 • Consequently a very
probable value for the total intensity is simply the sum of the elementary
intensities, I= 21
i
A; 12 . In other words, the observed intensity is
effectively M times the average intensity scattered by an elemental
volume. This result is analogous to that for the problem of a variable
number of particles, N, in a volume V ( §2.2.3.3), except that the observed
intensity is determined by an average over space instead of an average
over time as N fluctuates in a constant volume V.
Thus by analogy we are led to the following description of the limiting
observed scattered intensity. It is proportional to the mean square
fluctuation in the number of electrons in the volume v;, and will be zero
if the number of electrons in the different volumes v, is fixed. If Pi is
the electronic density of the volume V; and p is the average density as
determined by all the volume elements, the observed scattered intensity
will be proportional to
y2 ____
M M2 (p; - p)2 = Vv;(p, - p)2 (102)

The volumes v, = V/1lt do not really intervene as such, since the mean
square fluctuation is inversely proportional to the volume v;, but their
consideration is essential in determining the magnitude of the volume to
be used in the calculation of P;·
Equation 102 shows that the Umiting low-angle scattering is a consequence
of the heterogeneities in the scattering medium, but it also allows for the
stipulation that this heterogeneity must exist on a scale of several tens
to several thousand angstroms if the scatte:iing is to be observable. Let
us give several examples of the application of this simple rule. The fact
GENERAL THEORY 75

that matter is formed of atoms instead of being continuous is not a cause

of low-angle scattering since the corresponding density fluctuations are
averaged out in a volume of the order of 1µ 3 • A lattice with periodic
perturbations of density (for example, a period of the order of 50 A) is
heterogeneous, but it produces no small-angle scattering since the volume
V; contains a large number of periods, and hence the average density in
this volume is approximately constant.
A typical case is offered by the Al-Ag alloy that will be studied further
in §6.4.3.1. The silver atoms assemble themselves into spherical clusters
while remaining on the sites of the solid solution lattice. Around these
clusters is left a spherical shell lacking in silver. The shell diameter
is of the order of 50 A, and a large number of these clusters are randomly
distributed throughout the solid solution.
This alloy gives a pattern of central scattering containing a diffuse ring
whose radius corresponds to a Bragg distance of the order of 25 A, but
the 8cattered inten8ity decrea<Jes toward the center, and careful measurements
show that it approaches zero. If the small silver clusters exist in the
average solid solution without being accompanied by the shell-like
regions lacking in silver, only a normal central scattering is observed.
Equation 102 furnishes the explanation of these facts. In the first case
the cluster contains a number of excess atoms which is just the number of
atoms lacking in the shell-like region. The volumes V; in which the
densities p; should be examined are large with respect to the cluster
dimensions; thus they contain the same number of silver atoms as they
would if the solution were homogeneous. On this scale there are no
electronic density fluctuations and the intensity scattered at the center
should be zero. If there is no spherical shell around these clusters, these
clusters will play the role of particles with an electronic density different
from that of the surrounding medium; if the clusters are distributed at
random in the solid solution and are not too closely packed, they give
rise to a central scattering which is characteristic of their size.

2.4.2. CALCULATION OF THE SCATTERED INTENSITY AS A FUNCTION

OF p(x)
We shall now briefly present the solution of Debye and Bueche [322].
These authors define a function ri(x) by means of the relation

p(X) = Po + 'YJ(X)
where Po designates the average density of the substance. We can then
write

J ri(x) dx = O
76 SMALL-ANGLE SCATTERING OF X-RAYS

By means of equation 97 we can immediately write the expression for the

scattered intensity as (cf. §2.1.1)

l(h) = l,(h) fv Iv [Po+ '17(J:k)] [po+ 11(:1:,)]e-•'h•Czt-:11> dxk dJ:i

Recognizing the difficulties discussed previously, we shall take account

of more details in our calculation than was done by the original authors.
We will divide the intensity expression into the four terms corresponding
respectively to terms in p 0 2 , p 017(:1:k), p 0 17(:1:1), and '17(J:k)17(:1:1) and consider
ea.ch term separately.
First term:
11(h) = l,(h)p@'1· fv fv e-•h·Czt-:11> dJ:k dx,

We have already studied analogous terms (p. 34) and have seen that
they correspond to intensities that are practically zero for all observable
angles, since Vis the order of 1 mm 3 •
Second and third terms: Since the second and third terms are complex
conjugates, we can write
12 (h) + 1 (h) =
3 2Re {l.(h)p 0 fv 11(:1:11:) d:1:11: Iv e-ih·Czt-z1l dx1}
or
1 2(h) + 1 3(h) = 21,(h) Re {fv 11(:1:k)e-ih•zt dJ:k f
v p 0eih· "'1 dz 1}
Therefore
12 (h) + l 3 (h) < 21.(h) IA1 *(h) I ·IL ri(x}e-ih·z dz I
I
The first factor A 1 *(h) I is the modulus of the amplitude from which
11 (h) is derived. As we shall see below, the second factor is just the
modulus of A 4 (h), the amplitude corresponding to the fourth term.
Therefore 1 2 and13 as well as 11 are negligible as compared with 1 4 •
Fourth term:

l 4(h} = l,(h) fvfv 11(zk}17(:1:1)e-ih·C:1t-•1> dxk dx1

Let us place :1:1 = J:k + r. The above equation then transforms to

I 4 (h} = I.(h) fvfv '17(J:k)ri(zk + r)eib·r dJ:k dr

Let us consider first the integral with respect to :1:11:,

Iv 11(:1:)17(:1: + r} dx
GENERAL THEORY 77

This integral depends Ul).iquely on r. For r = 0, the value of the

integral is proportional to ;ji, the mean value of the square of the density
fluctuations, since
Iv 17(x)17(x) dx = f v 17 2 (x) dx = 17 2 V

Conforming to the notation of Debye and Bueche, let us define a

function y(r) by the relation

fv 17(x)17(x + r) dx = ;j°iVy(r) (103)

The equation for I 4 (h) then takes the form

I 4(h) = I.(h)17 2 V fv y(r)eih·r dr

This integral can undergo two modifications:
1. For large r there is no relation between the fluctuations of 17(x) and
17(x+ r); the function y(r) thus tends to zero as r increases. The
integration over the domain V can then be replaced by an integration over
an infinite region.
2. It can often be assumed that, in the region of vectors r of small
magnitude, the only domain in which y(r) is different from zero, the
properties of y(r) depend only on the magnitude r.
The final expression for the intensity is then
- ("' sinhr
I(h) = I 4 (h) = l 6 (h)172V Joy(r)--,;- 4'1T1'2 dr (104)

This is the expression developed by Debye and Bueche to describe the

intensity scattered by the matter under examination.
Let us consider the limiting value of I(h) ash - 0:

1(0) = J.(O) fvfv 17(xk)1J(X 1) dxt dx1

= I.(O) Iv 17(xk) dxk L 17(x1) ~xi (104:a)

The integral J17(x) dx is zero in a large volume on the average, but the
irregularity of the distribution of matter in the volume V irradiated by
X-rays gives it a value which fluctuates around zero. The two integrals
of equation 104:a are taken over the same volume; the two factors are
thus not independent, and the average of the product is not the product
of the average of each factor. Therefore, although the average value of
each integral is zero, the average value of J(O) is not zero. A calculation
shows that the limiting value for the intensity scattered at zero-angle is
given by equation 102.
78 SMALL-ANGLE SCATTERING OF X-RAYS

2.4.3. MATIER OF UNIFORM DENSITY AND RANDOM

DISTRIBUTION
Porod [137] studied a more particular system, that of a sample made
up of a random distribution of matter of constant density. The density
p(x) in the volume V of the sample can take only two values, p and 0. If
c is the fraction of the volume occupied by matter, the average density of
the sample is pc. In the parts occupied by matter r; = p(l - c), and in
the empty regions r; = - pc. Therefore
r;2 = p2c(l - c) (105)
Porod defines the distribution of matter by a function, Z(r), which
represents the probability that a point in the volume at a distance r from
a point occupied by matter is itself also occupied. It is assumed that this
probability is a function only of the distance r (an isotropic sample) and
that there is no long-range order, so that Z(r) tends toward c as r approaches
infinity. We can therefore put
Z(r) = c + (l - c)y(r) (106)
where the function y(r), called the characteristic function of the sample,
has the value unity when r is zero and approaches zero as r becomes very
large. This is identical to the function Debye and Bueche introduced
by the definition
fr;(x)r;(x + r) dx = ;Ji Vy(r) (103)

In order to show this, let us first point out that Z'(r) = (1 - c) cy(r) +
represents the probability that a point in the volume at a distance r from
a point in empty space (unoccupied) is itself also in empty space. This
can be verified easily by equating the two relations each of which describes
the probability of one point being occupied and the other unoccupied,
c(l - Z(r)) = (1 - c) (l - Z'(r))
Now in order to calculate the integral in equation 103 we must first set
up a table of probabilities for the different situations at points x and
x + r.
Nature of Points Value of
Probability of Occurrence
x and x + r 1](X)1](X +
r)

Occupied Occupied p 2 (1 - c) 2 cZ(r) = c2 + c(l - c)y(r)

Occupied Empty - p 2 c(l - c) c(l - Z(r)) = c(l - c) (1 - y(r))

Empty Occupied -p 2 c(l - c) (1 - c) (1 - Z'(r)) = c(l - c)(l - y(r))

Empty Empty p•c• (1 - c)Z'(r) = (1 - c)' + c(l - c)y(r)

GENERAL THEORY 79

Now by carrying out the integration

~ J17(x)17(x + r) dx = p 2c(l - c)y(r)

where equation 105 has been used, it can be seen that equations 103 and
106 define the same function.
Thus the scattered intensity can be determined immediately from
equation 104,

J(h) = f 6 (h) V p2c(l - c) l'°

o
sin hr
y(r) - - 4irr2 dr
hr
(107)

This expression can be linked to the equation relative to a single particle,

equation 21. Let us consider a very dilute system of identical particles
of arbitrary orientations. The probability Z(r) is then approximately
equal to the function we have called the characteristic function of a
particle, y 0 (r) (see p. 12), since by virtue of the dilution of the system
there is only a negligible chance of finding an occupied point outside of
the particular particle in which the origin point is chosen. Then since
I - c ~ 1 and c is negligible, equation 106 gives y(r) = y 0(r). In
addition Ve is the total volume of particles, Nv 0 • Therefore equation
107 is equivalent to equations 21 and 55.
Mathematically, J(h) is determined entirely by a knowledge of y(r).
However, the calculation of y(r), very complex for an isolated particle,
is rarely possible for systems of particles of a given arrangement. Con-
versely, the characteristic function of the sample can be determined from
the experimental measurement of J(h). This function y(r) contains all
the information that can be obtained from the small-angle scattering
experiments, but unfortunately this function does not give a direct image
of the structure and is quite far from defining it. The effects of both the
form of the particles and their mutual arrangement are intermixed in
the single function y(r). Theories are discussed in §2.2.3 which have as
their object the separation of these two effects.
Nevertheless, several parameters having simple and precise inter-
pretations can be obtained from y(r), as was done for the function of a
single particle.
y(r) is equal to unity when r is zero and tends asymptotically to zero
as r becomes large. It can take on negative values, greater than -c/(1 -c),
while the function for an isolated particle is always positive.
1. The slope of the curve at the origin is
80 SMALL-ANGLE SCATTERING OF X-RAYS

The slope of Z(r) at r = 0 can be calculated for the complex system in the
same way as for the isolated particle. If S is the total surface area of the
matter contained in the volume V, the real volume of matter being only
Vc, then from equation 24

(dZ)
dr r-o = -4 Ve
1 S
Therefore

( dy) - - 1 ~- - 1 8
dr r-o - 4c(l - c) V - 4c(l - c) •:i>

where S •:i>' the specific surface, is the surface area per unit volume of the
sample. From this, following the reasoning leading to equation 26, the
asymptotic behavior of the intensity curve is found as
2TTp 1S
I(h) R:i IB(h) h' (108)

The absolute value of the scattered intensity in the tail of the curve
depends only on the total surface area of the matter in the sample. If
the object is made up of n identical particles of volume v and surface a,
the total free surface S is always approximately equal to na, whatever
the degree of aggregation of the particles, provided that these are of some
arbitrary form and that they will not become distorted. The intensity
can be written as
l(h) = I.(h)np2(2TTB/h4)
It is equal to n times the average intensity scattered by one particle.
This shows that at large angles interparticle interferences are negligible,
even for particles of uniform size. It is therefore valid to apply equation
95 to dense systems. This does not mean that the intensity curve at
small angles will not be modified considerably by the action of particles
drawing closer together. Let us point out also that the above argument
will not be valid for particles in the form of broad platelets parallel to
one another, since the packing together of such particles can make the
interfaces disappear, decreasing the total surface and thus the scattered
intensity by a large amount. We have already mentioned (p. 70) that the
effect of interferences for one-dimensional systems is much larger than for
a powder of irregular grains.
2. The area of the curve f 00
y(r) dr can be calculated either from the
f"' Jo
integral Jo hl(h) dh (see equation 29) or from the total energy E scattered
GENERAL THEORY 81

in the low-angle region for a given incident beam intensity (see equation
30). Equation 107 gives

l.!!.= i<Xl r dr=

2 o y( )
1 i"'
hl(h)
--dh
47r V p 2c(l - c) o I.(h)
(109)

and
f cc 1 1 E
(110)
Jo y(r) dr = 2V p 2c(l - c) J.2p 2 I.(h)

where p is the distance from the sample to the film. l 0 is a parameter that
has been called the distance of heterogeneity by Kratky and Porod. In an
isolated particle l (equation 25) represents a mean value of the diameters
passing through every point of the particle in all directions, but what is the
geometrical significance of l 0 for an arbitrary system1 Let us draw a
straight line in an arbitrary direction from a point in matter. This line
will be divided into segments which are alternately occupied and
unoccupied by matter. The ratio of the total length of occupied segments
to that of unoccupied segments is equal to the ratio of the occupied and
unoccupied volumes, that is, e/(1 - c). The probable occupied length in a
line of length L drawn from the point chosen as origin is found from the
definition of Z(r) (equation 106) to be

LLZ(r) dr =Le + (1 - e) LLy(r) dr

,..., Le + (1 - e) L
cxi y(r) dr

This length is larger than the mean occupied length Le of an arbitrary

section of length L, because by imposing the condition that the section L
starts from a point occupied by matter the chances of finding an occupied
section are increased. This excess length is simply (l - c) (l0 /2).
3. In the general case there is a normalization relation analogous to
equation 28,

L"'h 2l(h) dh = 211 2I.(h)p2Vc(I - e) (111)

4. The roles played by e and (1 - e) in equation 107 are symmetric.

In addition the characteristic function y(r) is the same for an object and
for its complementary object (Fig. 66, p. 192). Therefore the same
scattered intensity is found for two complementary objects in the angular
region in which equation 107 is valid ( §2.2.2.2).
82 SMALL-ANGLE SCATTERING OF X-RAYS

REFERENCES FOR CHAPTER 2

Bertaut, E. F. (1950), Acta Cryst., 3, 14.
Bhatia, A. B., and Krishnan, K. S. (1948), Proc. Roy. Soc. (London), A192, 181.
Compton, A. H., and Allison, S. K. (1935), X-Rays in Theory anr/ Experiment,
Macmillan, New York.
Debye, P. (1915), Ann. Physik, 46, 809.
Guinier, A., and Griffon!, R. (1948), Acta Cryst., 1, 188.
James, R. W. (1948), The Optical Principles of the Diffraction of X-Rays, Bell,
London.
Kirkwood, J. G. (1935), J. Chem. Phys., 3, 300.
Kirkwood, J. G., and Boggs, E. M. (1942), J. Chem. Phys., 10, 394.
von Laue, M. (1941), Rontgenstrahlinterferenzen, Springer, Leipzig, p. 175.
Lennard-Jones, J. E. (1937), Physica, 4, 941.
Mac Gillavry, C.H., and Strijk, B. (1946), Physica, 11, 369.
Neugebauer, T. (1943), Ann. Physik, 42, 509.
Raman, C. V. (1924), Phil. Mag., 47, 671.
Rayleigh, Lord (1914), Proc. Roy. Soc. (London), A90, 219.
Rayleigh, Lord (1919), Phil. Mag., 6, 37, 321.
Rodriguez, A. E. (1949), Proc. Roy. Soc. (London), Ai96, 73.
Rushbrooke, G. S., and Scoins, H. I. (1951), Phil. Mag., 42, 582.
Wilson, A. J.C. (1949), X-Ray Optt'cs, Methuen, London; Wiley, 'N'ew York.
Yvon, J. (1935). Actualites sci. et ind., Hermann, Nos. 202, 542, 543.
Yvon, J. (1947), J. phys. radium, VII, 7, 201.
Yvon, J. (1949), Nuovo cimento, Suppl., 2, 187.
3. EXPERIMENTAL EQUIPMENT
3.1. GENERAL CONSIDERATIONS
The object of an X-ray scattering experiment, whether in the small-
angle region or in the usual domain of investigation, is the determination
of the variation of the intensity scattered by a sample as a function of the
scattering direction, this direction in general being defined by two para-
meters. In the important particular case in which the scattering is
circularly symmetric about an axis coincident with the incident beam,
only one parameter, the scattering angle, is involved, and the object
of the experiment is simply the determination of the relative value of
J(h), with h = (47T sin B)/}•.
Experiments can also furnish a second quantity, less frequently
employed, which is the absolute value of the scattering coefficient, a.
This is defined by means of the relation, I = I 0 a dm d.Q, where I is the
power scattered by the particle of mass dm in the solid angle d.Q, and I 0
is the intensity of the incident beam striking the sample. The sample is
assumed to be small enough to be non-absorbing.

3.1.1. OPERATIONAL PRINCIPLES

The method employed to realize the objectives discussed above is not
different in principle from that used in all experiments in X-ray crystal-
lography. Special difficulties are encountered, however, in investigating
the scattering at very small angles.
1. Geometrical Definition of the Incident Beam. Following the notation
of Fig. 22, let MM' be the portion of the sample which is irradiated. Each
point of the sample will receive a beam of rays whose divergence depends
on the constitution of the incident beam. Rays will converge at the point
of observation P which have been scattered by each of the points of the
sample through angles varying in an interval 2d() about a mean value,
20. The interval df) is practically independent of B, so that good definition
of the scattering angle in relative value is more difficult to obtain, the
closer a scattering angle of zero is approached.
Furthermore, there will always be an angular region inaccessible to
experiment; this is the region between N and N' of Fig. 22, in which
the scattered radiation received at any point is completely overshadowed
by the much greater intensity of the direct beam at this point. Thus, to
83
84 SMALL-ANGLE SCATTERING OF X-RAYS

investigate scattering at small angles, it is necessary to reduce both the

cross section and the divergence of the primary beam, the restriction being
greater, the smaller the limiting angle of observation that is desired. As
a result, the beams employed are much less intense than those used in
ordinary techniques, so that a determination of the best geometrical
conditions is essential.

x-~ra~y~~~~~~~~;::T'~~;;:;:;~~~~;;;;~~~~~R
source
N'

Observation D'
plane
Fig. 22. Slit system for a small-angle aoattering apparatus.

2. Parasitic Scattering. The measurement of the intensity received at

the point of observation is a correct measure of the intensity scattered by
the sample only if there is no parasitic scattering. The term parasitic
scattering refers to the radiation received at the point of observation
when the sample is withdrawn from the beam. If a Geiger counter or
ionization chamber is employed as a detector, the parasitic scattering
can easily be subtracted from the observed scattering to give the corrected
value, but this procedure is acceptable only if the correction is small. If
photographi~ detection is employed, it is very difficult to make the
correction by the above procedure. The reduction of the parasitic scattering
is thus· the second important requirement, and here again the suppression
is more difficult, the smaller the angles at which scattering is to be
observed.
Thus we can say that the quality of a small-angle scattering apparatus
is characterized by the power of the beam for a given fineness of dimensions
and by the angle beyond which all parasitic scattering is eliminated.
EXPERIMENTAL EQUIPMENT 85

A third critical property of such an apparatus is the spectral purity of

the primary radiation. We shall consider successively equipment without
and with monochromatization, showing the different domains of
application of each.

3.1.2. INFLUENCE OF THE MONOCHROMATIZATION OF THE

PRIMARY RADIATION
Use may be made of either filtered or crystal monochromated radiation,
depending on the nature of the sample to be studied.
1. The total radiation from the anode, with the usual filtering to remove
the K{3, may be used in a study of low-angle crystalline diffraction effects
that are analogous to the usual high-angle phenomena, differing only in
that the effective lattice spacings are very large. These patterns contain
lines, spots, or rings at well-defined angles, and the corresponding inten-
sities are considerably larger than those at intermediate angles on the
patterns. Thus, as with ordinary diffraction patterns, the diffraction
effects due to the characteristic radiation emerge from the continuous
background of diffraction and scattering caused by the continuous
spectrum. Often the primary objective of such a study is to determine
the position of the lines or spots, and in these circumstances even a rather
strong parasitic scattering may be tolerated.
2. The opposite case is the study of continuous scattering of the type
that has been described in the first part of this book. This continuous
scattering is often extremely weak and is superposed on the scattering
of various other origins, such as the inactive parts of the sample (the
solvent, for example, when the scattering of particles in solution is studied).
Given the actual state of the theory, it is essential in this type of problem
to have a precise evaluation of the function J(h).
It is easy to see that in certain cases the influence of the continuous
spectrum may be considerable, since, in addition to the intensity J(h) due
to the principal radiation of wavelength A. 0 , one will also observe a scattering
of the form

fI (~ h) f(A.) dA.

where j(A.) is the distribution function of the energy in the continuous

spectrum. The effect of all the continuous spectrum can thus be large
with respect to that of the characteristic radiation. Experiments by
several authors have proved that an investigation cannot be made free
from all objection without the use of monochromatized radiation.
When Geiger-counter detection is used, the elimination of the continuous
spectrum by the double filter method of Ross (Kirkpatrick (1939)) is often
86 SMALL-ANGLE SCATTERING OF X-RAYS

sufficient, but this method is not easily applied with photographic

techniques. The most practical and the most general method of pro-
ducing monochromatic radiation involves the use of a cryBtal mono-
chromator. Since the use of a monochromator profoundly modifies the
geometry of the equipment, we shall study separately the system with
collimation, designed for st~dies of crystalline diffraction, and the system
wit:ii monochromatization, especially adapted to the study of the continuous
scattering.

3.2. SYSTEMS EQUIPPED WITH COLLIMATORS

The beam (see Fig. 22) is defined by two apertures, 0 1 and 0 2 , separated
by a distance v, which are placed before a source of radiation large enough
to illuminate the entire opening. As the edges of 0 2 are touched by the
direct beam, they are sources of scattered and diffracted radiation. Thus
it is necessary to protect the plane of observation by introducing a third
aperture, 0 3 , at a distance w from 0 2 , whoBe edgeB clOBely approach but do
not touch the incident beam. Apertures 0 2 and 0 3 then define the region
DD' which is not exempt from parasitic scattering, The sample is placed
after 0 3 , as close to it as the supports will permit. The sample and the
opening 0 3 are thus at approximately the same distance from the plane
of observation; this distance is denoted by B.
We want to determine the form and dimensions to give to the various
apertures in order to obtain the "best results" for our measurements.
We must first specify the known quantities of the problem and the criterion
of quality to be required. This cannot be done until the nature of the
pattern given by the sample has been specified.

3.2.1. COLLIMATOR FORMED BY TWO SLITS

Let us consider the simple problem of the study of the equatorial line
of a crystalline "fiber pattern." Our attention is thus devoted to
diffraction effects in only one dimension. The collimator openings can
then be infinitely long slits parallel to the "fiber axis" (perpendicular to
the plane of Fig. 22), and the diffraction spots will appear as lines. We
shall impose the following conditions:

I. Let us define the angular uncertainty of the pattern as the variation

in scattering angle of the rays arriving at a point Pin the plane of observa-
tion. Such rays scattered from an incident ray M 1 R are scattered through
an angle equal as a first approximation to P R/B, whatever the position
of the diffracting point M 1 on the sample. Thus the maximum variation
of the scattering angle, 2d(J, for the group of rays converging at P (the
EXPERIMENTAL EQUIPMENT 87

angular uncertainty of the pattern), is measured by the quantity,

A = a/s, where a is the width of the primary beam in the plane of observa-
tion, and s is the sample-to-film distance. The largest lattice spacing
which can give rise to a line distinct from the direct beam will then be

dmax =A/A
2. The scattering angle inside of which parasitic radiation is found is
B/2 = b/2s, where b is the width of the part DD' of the film receiving
scattered radiation in the absence of a sample. The upper limit of the
lattice spacings which will register outside of all parasitic scattering is
then d'max = 21/B. Obviously B is larger than A, and usually it is
larger than 2A, so that d'max < dmax·
3. Either Geiger-Muller counters or photographic plates may be
employed as detectors, but the conditions that apply to each are different.
(i) If a Geiger-Muller counter is used, it must be equipped with an
entrance slit so that the divergence of the rays scattered by a point on
the sample that enter the counter is fixed and clearly less than A (A/10,
for example). If this condition is satisfied, it is possible, at least
theoretically, to correct the observed pattern for the effect of the width
of the direct beam (see §3.4).
For a given angular uncertainty, A, the counter slit width is proportional
to s. Since, for constant A, scan be arbitrarily chosen without affecting
the measured power, it is then advantageous to employ a large value of s
so that the counter slit can be more easily constructed. The only
restriction is that s must be less than a limiting value, s 0 , determined by
the mechanical conditions and obstructions.
(ii) The limitations of the photographic method arise from the grain
size of the film, which is always rather large for emulsions sensitive to
X-rays. X-ray patterns cannot be usefully enlarged by a factor of 10.
The resolving power of these films is of the order of a hundredth of a
millimeter; thus the exploring slit of the microphotometer should have
a width of this order of magnitude, £. Consequently the sample-to-film
distance s has a lower limit, sP, such that £/sP is clearly smaller than A
(for example, A/10, as suggested in the preceding case).
4. The study of very small angles necessitates the use of beams which
are very narrow an_d, consequently, of low power. It is essential for the
success of the experiment that the system be found which allows the most
powerful beam, while satisfying the geometrical conditions previously
enumerated. Specifically, it is necessary to try to maximize the number
of photons received by a counter placed at the center of a diffraction line
or to maximize the blackening of this line on a film. The most desirable
88 SMALL-ANGLE SCATTERING OF X-RAYS

form for a sample is that of a small plate intercepting the entire beam,
the thickness being chosen in accordance with its absorption coefficient.
(It is well known that the optimum thickness is that for which the ratio
of the transmitted to incident intensity is the factor, l/e ~ 1/3.) Now,
for simplicity let us assume that the point of observation is in a region of
the pattern in which the variation of intensity with angle is very small,
as, for example, in the center of a rather wide diffraction line. Then, for
a given slit width, the intensity of the radiation entering the counter will
be proportional to the total power of the primary beam incident on the
sample per unit collimator slit height, 11 . When photographic detection
is employed, the blackening at the same point will be proportional to
11/s (we are considering one-dimensional diffraction effects, so the factor
Ifs rather than l/s 2 intervenes). These are the factors that must be maxi-
mized respectively in the construction of the collimator when the detector
is a Geiger counter or a photographic film.
Bolduan and Bear [205], in an analogous calculation, chose a criterion
which seems to us to be on a less general level; they maximized not the
total power of the incident beam but rather the intensity of the radiation
striking the plane of observation at the center of the direct beam. Their
conclusions are clearly different from those we shall draw. This shows
that, if in a given experimental problem some of our assumptions are not
satisfied, it will be necessary to discard our conclusions and to make an
analogous calculation with appropriately modified factors.
5. The source of X-rays is assumed t.o be an X-ray tube with a
rectangular focal spot of large length and of width l, so oriented that the
long dimensions of the focal spot and the slits are parallel. The emerging
rays make an angle, cc, with the plane of the target. If the power per unit
area delivered to the target by the incident electrons is&, the intensity
of the emitted X-rays will be proportional to &/cc, if cc is larger than a
limiting value, cc 0 , of the order of 1° or 2° (Bolduan and Bear [205]). We
shall fix cc at this optimum value, cc 0 , and we shall place the first slit of the
collimator close to the focal spot; the width, p, of the slit 0 1 is then
determined as the projection of the focal spot, lcc 0 .
Let r be the width of the second slit, 0 2 , which is placed at a distance
v from 0 1 . The power of the beam defined by the collimator 0 1 0 2 will
then be proportional to

& r &lr
-p-=-
IXo V V

The third slit, 0 3 , of width q, is placed at a distance w from 0 2 • Let

us recall that the ratio a/s has been denoted by A and the ratio b/s by B.
EXPERIMENTAL EQUIPMENT 89

The following relations, deduced from simple geometrical considerations,

exist between these different quantities (Bolduan and Bear [205]):
(p + r)/v <A (1)
w = 2r/(B-A) (2)

+ v(B2~ A)]
q = r [l (3)

r 2(p + r) + v(B - A)
8 = --- x -------- (4)
B- A Av - (p + r)

3.2.1.1. Calculation of the Optimum Collimator

(a) Geiger-Muller Counter Detector. The collimator should be chosen
so as to maximize pr/v and thus r/v, since p is fixed. For a given value of
v the choice of maximum r by means of equation l leads to the relation
rmu.x =Av - p. This value of r corresponds to an infinite value of s,
however, as is seen from equation 4. But, as rand s vary in the same
way, ifs is made equal to its maximum value, s0 , there will be an optimum
value of r, r 0 (v), for each value of v. Thus, by letting v vary, a maximum
for the ratio r.(v) can be found, which determines the value to be selected
v
for v. The rest of the geometry of the system is then determined by
means of equations 2, 3, and 4.
(b) Photographic Detection. For this system the collimator should be
chosen so as to maximize the function 11 /s = 9lr/vs. This requires that
s be a minimum, that is, thats have the values,, determined by the grain
size. A calculation similar to that outlined for the previous case but
with s 0 replaced by s 21 then leads to the determination of the optimum
value of v and the values of the other unknowns of the system.
EXAMPLE. The focal spot of the tube has a width of l mm. For an
angle of emergence of l 0 the source corresponds to a first slit of width
p = 20 µ. The parameter describing the angular uncertainty of the
pattern, A, is chosen as A = 10- 3 • This means that for Cu Ka. radiation
(A = l.54 A) the diffraction line for a lattice spacing of 154;0 A will just
be separated from the trace of the direct beam. In addition we shall
require that the parasitic scattering be stopped beyond an angle corre-
sponding to a lattice spacing of 1200 A; this determines B as
B = 2.5 x 10-3 •
(a) Counter Detection. Let the largest possible sample-to-counter
distance be s0 = 500 mm. The efficiency of the system is given by
curve a of Fig. 23 in which the ratio r/v is plotted as a function of the slit
90 SMALL-ANGLE SCATTERING OF X-RAYS

separation v. The maximum of the curve determines the following

parameters: -
v= lOOmm. w= 77mm.
r0 = 58 µ q = 120 µ
(b) Photographic Detection. Let the minimum sample-to-film distance
be: s'P = 100 mm. If the film is investigated with a microphotometer

riv

0.6
o.s

(a)

r/v

0.5

0.3

O.O .___5_._0_ _ _ _ _l_._00-----1~50----v~

(b)
Fig. 23. Calculated collimator efficiencies.
(a) Counter detection. (b) Photographic detection.

with a 20µ slit, this slit will then be 5 times as narrow as the trace of the
direct beam on the film. The efficiency of this system is given by curve b
of Fig. 23, in which again the ratio r/v is plotted as a function of the slit
separation v.
The optimum conditions as determined from this curve are:
v = 70mm. w =31 mm.

rjl = 24µ q =47 µ

EXPERIMENTAL EQUIPMENT 91

The maxima in the efficiency plots of both these systems are very fiat.
The slit separation v could vary between 60 and 190 mm. in the first
system or between 50 and 100 mm. in the second without causing a
10 per cent loss. Consequently there is considerable latitude for accom-
modation of particular supplementary conditions.
The preceding calculations are given as an example, but, as we have
already emphasized, they are valid only if all the conditions imposed by
the hypotheses are satisfied. If, for example, there is available a demount-
able tube whose focal spot dimensions can be varied, it is advantageous to
diminish the size of the focal spot because the brilliance of the focal spot
increases, the smaller its dimensions, at least until the diameter approaches
a limiting value between 50 and 100 µ. Such a change of the dimensions
of the focal spot has an important influence on the values taken for r
and v.

3.2.2. COLLIMATOR WITH CIRCULAR OPENINGS

Let us now consider a sample which gives rise to a scattering in various
directions that does not show rotational symmetry, such as, for example,
fixed crystals without special orientations. For this study it seems
logical to employ cylindrical collimation. We shall preserve the same
notation as used in the preceding section, except that p, r, w, a, and b will
refer to diameters instead of widths of the apertures and the beam
cross sections. The power of the primary beam is then proportional
!}J p2r2
t o - - . The quantities to be maximized for Geiger-Muller counter and
oco v2 &' p2r2 &' p2r2
photographic plate detection are then respectively - - 2- and - 22 .
cc 0 v <7.o t' s
These are the same conditions as were found for the slit collimators.
Actually this system is not often employed, because it is more difficult
to obtain adjustable holes than adjustable slits. In practice two slits
oriented at right angles are generally used, so that the beam is defined by
square or rectangular openings. This general problem is too complex,
and we shall consider it only for the case in which the sample gives a
circularly symmetrical pattern, that is, when the diffracted intensity is a
function only of the diffraction angle 20.

3.2.3. COLLIMATOR WITH SLITS OF FINITE HEIGHT FOR THE STUDY

OF CIRCULARLY SYMMETRICAL DIFFRACTION PATTERNS
Our study will be extended only to the intensity diffracted in the plane
perpendicular to the Large slit dimension. Experience with ordinary
powder patterns has shown that intensity considerations are best served
by the employment of beams defined by slits. Let us consider how this
concept may be made more specific for a study of small-angle scattering.
92 SMALL-ANGLE SCATTERING OF X-RAYS

Following the notation of Fig. 24, the primary beam, defined by

rectangular openings of dimensions Pv p 2 , and r1 , r 2 , cuts the plane of
observation (II) in a rectangle of sides ai and a 2 , these being widths and
heights respectively. Let P be the point of observation in the horizontal
plane; the radiation scattered to P from the central ray of the primary
beam is scattered through an angle 2(} = OP/s. Similarly, radiation
scattered to P from any ray whatsoever of the primary beam which

Fig. 24. Collimator with slits of finite height and width.

intercepts the plane (II} at the point M is scattered through an angle

which, to a first approximation, is 2(} = 1ll P/s. The maximum un-
certainty of the scattering angle is then given by

A= PMmax - PMrnin = PQ- PR

8 8

The use of this expression for A to characterize the accuracy of the

pattern imposes too severe a condition, since the edges of the beam will
intervene much less than the central part (primarily because of a pen-
umbra] effect). A more accurate calculation, however, not only would be
more complex but also would lack generality.
It can easily be seen that if u = 4 OPQ

R'Q a2
tanu= - - = - - - -
R' P 4s(} a1 +
EXPERIMENTAL EQUIPMENT 93

and
A= -PQ -
8
-PR = ('2(}
8
+ -2s
a1 ) - 1 -
cos u
( 2(} - -a 1 )
2s
Ifni and a2 are small with respect to 2s(}, theo.
al a22
A = -; + 1682(} (6)

For large scattering angles the angular uncertainty A is thus primarily

due to the width of the slit, but as small angles are approached the
influence of the slit height becomes preponderant. For example, if our
system is-to be designed so that the two causes of error, the width and the
height of the slit, are to contribute equal parts to A, then if we choose A
as 10- 3 , as in a previous example, the choice must be a1 /s = a 22/16s 2(}
= 0.5 X 10- 3 ,
For 2(} = 0.1 radian, this requires a slit height 40 times the width; on
the other hand, for 2(} = 0.001 radian, the ratio of height to width is only
a factor of 4.
Our primary problem is to determine the best arrangement of two rec-
tangular slits for a collimator to be used for studies of small-angle dif-
fraction. It is impossible to solve this problem for the completely general
case. The angular uncertainty of the general pattern is unlike that of the
one-dimensional pattern; it cannot be a constant but must increase with
decreasing angle. Thus only particular cases can be treated, of which we
offer two examples:
(a) Let us consider the system described on p. 89, which employs a
counter as a detector, and let us replace the infinite slits by slits of the
same widths placed in the same positions but of heights p 2 and r 2 ,
respectively. The sample gives a circularly symmetrical diffraction
pattern. An accuracy such that A = 10- 2 is wanted at an angle
2() = 0.01 radian. Since a1 /s = 10- 3 , equation 6 gives a2 /s = 0.027;
thus, since 8 = 500 mm., it follows that a 2 = 13.5 mm. Now the total
power of the beam for a given value of the slit separation v is proportional
to the product of the areas of the two openings. The beam power will
thus be a maximum when the product p 2r 2 is maximized, since p 1 and r 1
are fixed. There is in addition a geometrical relation between p 2 and
r 2 (see Fig. 24)

v 8

From these relations the optimum parameter values are found to be

p2 = 1.35 mm. and r 2 = l.14 mm.
SMALL-ANGLE SCATTERING OF X-RAYS

(b) As a second problem let us cite one that has been treated by
Yudowitch [241]. If the collimator is formed by two identical apertures
separated by a distance v, what form should be given them so that the
angular uncertainty of the pattern for a given beam intensity should be a
minimum at a given angle? The first hypothesis implies a fixed area of
the collimator openings, Pi.P2 = r 1r 2 = S. Now the width and height of
the beam in the plane of observation at-a given distance, s, are propor-
tional, respectively, to p 1 and p 2 , and thus they are related by the
expression a 1a 2 = kS.
Equation 6 then shows that A is a minimum when
a1/s = 2A/3 (7)
~ 2/16s 2 0
= A/3
The form of the openings is determined by these relations.
The principal conclusion of this discussion is that the variation of the
scattering angle among the various rays reaching the plane of observation
at one point increases as the average scattering angle decreases. It is
therefore necessary to reduce more and more the collimator openings and,
consequently, the power of the primary beam, the greater the accuracy
desired. This is the major difficulty in experiments concerning small.
angle diffraction.
This difficulty can be avoided by either eliminating the cause of the
error or by calculating a correction to the experimental results. In the
following paragraphs we shall discuss the efforts that have been made in
these two directions.

3.3. SYSTEMS USING MONOCHROMATIC RADIATION

A mortochromatic primary beam is advantageous, at least theoretically,
for every type of investigation, but it is indispensible for the correct
interpretation of continuous scattering.

3.3.1. SOURCE OF MONOCHROMATIC RADIATION

A first approach is to employ the same system as previously described
but to replace the ordinary source by one giving monochromatic radiation.
It is well known that the continuous spectrum from a tube is weaker, the
smaller the atomic number of the target material and the lower the applied
potential. The characteristic radiation of light elements can be excited
by very low voltages; a tube with an aluminum target operated at 4 kv.
emits practically pure Al Ka. radiation (Hosemann [80), Yudowitch [186)).
This technique, however, allows the use of only very long wavelengths
(Al Ka. : ), = 8.34 A); although this offers the advantage of increased dis-
persion of the pattern, proportional to I., unfortunately absorption is so
EXPERIMENTAL EQUIPMENT 95

great that the very small sample thicknesses that are required can be
prepared only with great difficulty.
In another method employing long wavelengths, Henke and DuMond
[76""] isolated the Cu L radiation (13.3 A) by total reflection. The
incident radiation struck the surface of a polished glass mirror at an angle
such that only X-rays oflong wavelengths were totally reflected (Compton
and Allison (1935)). The form of the mirrors was ellipsoidal, so that the
radiation coming from a point source was focused to a point image.

3.3.2. BALANCED FILTERS

The filtering technique usually employed as a means of decreasing the
relative percentage of the K{J radiation cannot be applied here, since it
does not appreciably affect a large part of the continuous spectrum.

Fig. 25a. A schematic diagram of the Beeman apparatus with slit

collimation. Tis the X-re.y source; F 1 and F 1 , the collimator slits;
S, the sample; F 8 and F,, the counter slits; C, the counter.

Ross and later Kirkpatrick (1939) developed a balanced filter technique

that uses two patterns, each made with a different filter; the two filters,
whose K edges bracket the wavelength to be employed, are of such
thicknesses that the difference in the scattered intensity of the two
patterns is due almost exclusively to the selected wavelength. For
example, if Cu Ka. radiation is to be used, one filter would be of nickel
6.84 x 10-3 mm. thick, and the other would be of cobalt 7.60 x 10-3 mm.
thick. Kratky [112] has used balanced filters in conjunction with
photographic detection, but this is a difficult technique, since the
differences in intensity of two patterns or of the two halves of one pattern
cannot be determined with great accuracy. However, the balanced
filter method becomes one of the best techniques when it is used in com-
bination with a Geiger counter.
Figure 25a shows a schematic drawing of the apparatus used in the
laboratory of Beeman [376]. The source of X-rays is a rotating anode
tube whose high power has been shown to be necessary for studies of
96 SMALL-ANGLE SCATTERING OF X-RAYS

weakly scattering substances. The beam is defined by a collimator

composed of two slits separated by a distance of 30 to 50 cm. The
specimen is placed from 15 to 25 cm. beyond the second slit. A second
pair of slits with the same separation as the first pair are placed on an
arm which rotates about an axis through the specimen up to an angle of 5°
with respect to the incident beam. The third slit, F 3 , acts as the aperture
0 3 in the general diagram of Fig. 22, serving to stop the parasitic scattering
and not defining the geometry of the useful beam. The background
intensity may be less than _I0-7 of that of the direct beam. The double
filter is placed immediately in front of the counter. The useful intensity
is measured by the difference between the number of counts obtained when
first one filter and then the other is employed.
i:I:'he balanced filter is ineffective if the sample, under the action of the
continuous spectrum, emits fluorescence radiation of the same wavelength
as that of the principal primary radiation, since the intensity to be
measured is found as the difference between very large numbers.
Beeman and his collaborators later used a proportional counter as a
detector. When this is employed with the appropriate electronic cir-
cuitry, the sensitivity of the apparatus is restricted to wavelengths lying
in a narrow band around that of the chosen radiation (Arndt, Coates, and
Riley (1953)).
In a new apparatus shown in Fig. 25b, Beeman and his coworkers
replaced the collimator slits by pinholes in order to eliminate the difficult
corrections for the effects of slit height ( §3.4.2). The intensity of the
incident beam is reduced considerably by this modification. In order to
make possible reliable measurements (for samples that give a circularly
symmetric pattern), the counter slit is made in the form of a ring-like
aperture centered exactly on the axis of the direct beam. The scattering
angle is then varied by displacing the counter and sHt along the axis of the
direct beam, either towards or away from the sample.

3.3.3. MONOCHROMATIZATION BY CRYSTALLINE DIFFRACTION

Crystalline diffraction offers the most precise method for mono-
chromatization of radiation and the best method for use with photo-
graphic measurements. We shall thus consider at length the different
techniques that have been proposed.

3.3.3.1. Plane Monochromator

The simplest system is that employing a flat single crystal. The usual
crystals can vary in type from the perfect crystal, which gives a sharply
defined, low-intensity beam, to the very imperfect crystal, which gives an
EXPERIMENTAL EQUIPMENT 97

intense but poorly defined beam whose divergence is equal to the range
of orientations of the reflecting crystalline planes.
The characteristics of the different crystals that are most easily obtained
have been compared by Lipson, Nelson, and Riley (1945), who recommend
for small-angle scattering studies the use of the (100) planes of calcite.

Fig. 25b. A sketch of the Beeman apparatus with pinhole collimation.

A is the rotating target tube; B, the collimating pinholes; 0, the
vacuum connecting tube; D, a vacuum extension tube; E, the guard
ring and specimen; F, an extensible plastic tube filled with helium;
G, the beam stop for the direct beam; H, the annular slit before the
proportional counter J and preamplifier K; L, a sliding base for the
annular slit and counter.

These planes, though of weak reflecting power, give an extremely sharp

beam in which the Ka. 1 and Ka. 2 reflections can be separated. Quartz
(lOll) planes give a still weaker beam. Pentaerythritol, on the other
hand, gives a very intense reflection, but not only is the beam divergence
very large, but also the crystal is unstable.
To concentrate the power of the diffracted beam Fankuchen (1937)
proposed to cut the surface of the crystal in such a manner as to diminish
the cross section of the reflected beam, as is shown in Fig. 26. If the
98 SMALL-ANGLE SCATTERING OF X-RAYS

angle between the surface and the reflecting crystalline planes is d.. and
the Bragg angle is(}, the angles made by the primary and diffracted beams
with the surface are, respectively, (} + oc and (} - oc. This gives the ratio
sin((} - oc)
of cross sections as . Evans, Hirsch, and Kellar (1948)
sin((}+ oc)
studied in detail the practical possibilities of this process and reached the
following conclusions: there is a maximum theoretical intensity gain of 2,

Fig. 26. The plane monochromator proposed by Fankuchen.

but this factor cannot be attained in practice. There exists an optimum

angle of cut, slightly smaller than (}, The gain obtained depends to a
large extent on the state of the crystal surface; the best results were
obtained with calcite crystals which were cut, polished, and then very
lightly rubbed with either very fine emery paper or the point of a pencil.
A perfect crystal also collimates the diffracted radiation, functioning in
this respect as a slit at an infinite distance viewed through a negligible
angle. Figure 27 shows the arrangement of the diffraction apparatus
employing a plane, perfect monochromator. Radiation from a source of
width p is diffracted into a beam of constant width (neglecting the
spectral width of the characteristic radiation); thus the width of the
direct beam at the plane of observation is p. The sample is at a distances
from the plane of observation and at a distance w from the monochro-
mating crystal. It is necessary to place a slit. just before the sample to
eliminate the strong parasitic scattering arising from that part of the
surface of the monochromator touched by the direct, polychromatic
beam. The angular uncertainty of the scattering pattern is given by the
quantity A = p/s, and the angular region within which there is parasitic
scattering is B/2, where

B = b/s = (2p/w) + (p/s)

Thus for given values of A and B these relations determine the values
of w and s, the source width p being fixed.
EXPERIMENTAL EQUIPMENT 99

In determining the angular region obscured by parasitic radiation we

have assumed that only the useful region on the monochromating crystal
receives radiation. This requires that the equipment include a second
slit, F, of width p, placed as close as possible to the monochromating
crystal. This slit plays no role in the collimation of the incident beam
but acts only to reduce the parasitic scattering. After the mono-
chromator crystal has been adjusted so as to produce the reflected beam,

X-ray source

Fig. 27. A small-angle scattering apparatus employing a plane

monochromator.

the independently adjustable edges of this slit should be closed to the

precise width at which the reflected beam begins to be narrowed: the
extent of the region containing parasitic scattering will then be a minimum.
If the monochromator crystal is not perfect, there will be a divergence
of the reflected beam equal to the disorientation of the crystalline planes.
This disorientation is the minimum value of A that can be realized.
Another cause of divergence of the reflected beam is the oc1 -cx2 doublet
nature of the characteristic K radiation. For example, the divergence of
Cu Kcx radiation reflected from the cleavage planes of calcite is 0.65 X 10- 3
radian. This divergence is larger, the greater the diffraction angle of
the monochromator. Thus Lipson, Nelson, and Riley (1945) have sug-
gested the use of a crystal of gypsum, which has a very large spacing for
its cleavage planes (7.85 A). This would reduce the Cu Kcx divergence to
0.25 x Io-a radian, making possible theoretically the resolution from the
direct beam of reflections due to spacings as great as 6000 A.
100 SMALL-ANGLE SCATTERING OF X-RAYS

The primary disadvantage in the use of plane crystals as mono-

chromators is the low power of the reflected beam, which seriously limits the
possibilities for practical application of this method. Two other methods
will now be discussed that offer several important improvements: (1) an
increase in intensity of the monochromatic beam: (2) a decrease in the
angular uncertainty of the pattern; (3) an elimination of the parasitic
scattering beyond quite small angles. These methods depend on the use
of the bent crystal monochromator and the double monochromator.

(a) (b)

Fig. 28. Johann bent crystal monochromators:

(a) symmetrical; (b) asymmetrical.

3.3.3.2. Bent Crystal Monochromator

Bent crystals were originally introduced in X-ray spectroscopy, serving
as analyzers capable of producing intense spectra; they can serve equally
well as monochromators. When they are used as monochromators, a
system employing reflection geometry is particularly advantageous. The
simplest of these is that described by Johann (1931): A thin plate is bent
elastically (or plastically) to a radius of curvature R; then, with the
arrangement shown in Fig. 28a, rays from the source S, diffracted at an
angle (} by the bent crystal, are focused approximately at a point F, with
the source, the crystal, and the focus lying on a circle C of diameter R.
This focusing is exact if the surface of the crystal is coincident with the
arc of the circle, the radius of curvature of the lattice planes remaining
the same, i.e., R. This is accomplished by the technique of Johansson
(1933): The plate is first cut cylindrically to a radius Rand then curved
onto the circle of radius R/2; the radius of curvature of the diffracting
EXPERIMENTAL EQUIPMENT 101

planes is then R, as before. The Johansson technique is of interest only

if beams of large divergence (greater than 1°) are to be used, which, as we
shall see, is not true for small-angle scattering studies.
If the crystal slab is cut so that the reflecting planes make an angle <X
with the surface of the slab, an asymmetric monochromator is obtained
(Guinier (1946)) for either the Johann or Johansson systems (Fig. 28b).
The ratio of the distances of the source and the focus from the crystal is
sin (0 - <X)
. (O • This arrangement has the advantage that the distance of
sm + <X)
the focus from the crystal can be made large without having to place the
source farther away from the crystal.
For small-angle scattering studies, the bent crystal not only serves as
a monochromator but it also has the advanf,age of producing a beam which
is geometrically very well defined; the diffracted beam converges to a fine
line even when the beam divergence, and hence power, is quite large.
Thus, if the plane of observation passes through the focus F, normal to
the beam, the angular uncertainty of the pattern is not dependent on the
beam power, as was true for the system with a simple collimator. The
width of the trace of the primary beam is determined primarily by the
separation of the Ka. doublet, the effects arising from the geometry of the
system being much smaller. For example, ifthe monochromator consists
of a thin slab of quartz cut at an angle of a.= 3° to the reflecting (lOll)
planes and curved to a radius of 500 mm., the distance from the crystal
to the point of focus for Cu K<X radiation is 140 mm., and the separation
of the <Xi-IX:! doublet corresponds to an angle of 2'; the observed beam
width in the plane of observation is then 0.15 mm. If the sample is
placed in the beam at a distances= 100 mm. from the plane of observa-
tion, the corr~sponding uncertainty of the pattern will be A = 1.5 X 10- 3 •
In this as in the other systems, particularly those employed for the
study of continuous scattering, the minimizing of the angle beyond which
parasitic scattering is eliminated is essential. This is done by placing
just in front of the sample a slit whose edges approach the beam as closely
as possible, as is shown in Fig. 29. If dis the focal distance, CF, and w is
the divergence of the beam, the angle obscured by the parasitic scattering
will be
B b dw
-z,._FMD=-=-=--
2 2s d-s
It is therefore necessary to restrict the distance s to the smallest value
sJ>for which the pattern will be legible (cf. p. 87). If, in the example
treated above (d = 140 mm.), we choose sJ> = 80 mm., the obscured angle
will be B/2 = 2.2w. We are thus led to the following conclusion: although
102 SMALL-ANGLE SCATTERING OF X-RAYS

the convergent property of the beam makes the angular uncertainty of

the pattern independent of the beam divergence, nevertheless this diver-
ge.nee must be strongly limited in order to diminish the range of angles
which are obscured, since experiments have shown that the parasitic
intensity is so large that no experiments are possible in this region.

F
b

D'
-------j"t

X-ray source

Fig. 29. A small-angle scattering system employing a

bent crystal monochromator.

The system with a bent crystal monochromator then appears to be

hampered by the same restriction as the systems with a plane mono-
chromator or a collimator; less and less power can be employed, the
smaller the diffraction angle at which a measurement is desired. There
remains one important advantage for this system over the previous
systems employing monochromatization, however; as a result of the
precise convergence of the direct beam, there is a smaller angular
uncertainty in the pattern.

3.3.3.3. Combination of Two Bent Crystal Monocbromators

The principal cause of the troublesome parasitic scattering lies in the
fact that the surface of the monochromator is irradiated by the very
intense direct beam, while use is made of the radiation of a chosen wave-
length, which forms only a small fraction of the total power of the beam.
This fact led to the idea of employing a double monochromator, in which
the focus, F 1 , of the first bent crystal served as the source for a second, so
oriented as to reflect the entire incident beam into a beam converging to
a new focus, F 2 (see Fig. 30). Fournet [48] has studied the geometry of
this system in detail and has shown that the most desirable arrangement
is that in which the two crystals are oriented in the antiparallel position,
since the parallel position allows a narrow spectral band to be reflected
succest1ively by the two crystals at their optimum mutual orientation.
The point of interest of this system is that, even without any protective
EXPERIMENTAL EQUIPMENT 103

slits, the parasitic scattering is almost completely eliminated beyond an

angle of about ten minutes from the direct beam. The sample can then be
placed quite close to the second monochromator, so that with very
asymmetric monochromators of large radius of curvature, sample-to-film
F1

Sample

Film

X-ray
source
Fig. 30. A double bent crystal monochromator.
F 1, intermediate focus; F 1 , final focus.

distances of 300 mm. can be obtained. But, although it has been verified
that films which are certainly more easily readable are obtained with such
large distances (at the price of increased exposure times), it has proved
to be practically impossible to reduce further the angular limit imposed
by the remaining parasitic scattering. Fournet has made a detailed
theoretical and experimental study of this scattering; it seems to be
caused principally by irregularities in the curvature of the lattice planes
of the two crystals. These irregularities are the results of defects in the
cutting of the crystal and the crystal holders. The mechanical difficulties
in making perfect cylinders are well known. Thus from this point of
view it would seem advantageous to employ uncut slabs of quartz (the
Johann technique). The best method for cutting the holders is that
described by DuMond (1947).
In addition to this effect it seems probable that simple elastic curving
creates a crystalline imperfection that should cause diffuse scattering.
As yet no experimental or theoretical analysis of this effect has been made.
In any event, even for a perfect crystal, there will still be diffuse
scattering of thermal origin which will be particularly intense in the
neighborhood of the direction of a Bragg reflection. This scattering,
however, has never been troublesome in the apparatus we have used.
The following are the details of construction and the characteristics of
the double monochromator system we have developed.
First Monochrmnator. Quartz slab, cut parallel to (1011) planes;
Johansson technique; elastically bent to 500 mm. radius of curvature.
Focusing distance of 115 mm. for Cu Koc radiation.
Second Monochromator. Quartz slab, cut at 3° to (lOll) planes;
Johansson technique; elastically bent to 500 mm. radius of curvature.
Focusing distance of 140 mm. The sample is generally placed 100 mm.
from the film.
104 SMALL-ANGLE SCATTERING OF X-RAYS

Numerous trials have been made with quartz slabs of different curva-
tures constructed according to the Johann method. We have verified
that it is possible, with suitable though delicate adjustments, to eliminate
one of the components of the KrJ. doublet in the twice reflected beam.
The parasitic scattering at a point in the observation plane correspond-
ing to a scattering angle of 15' (the angle for a lattice spacing of 300 A) is
1/100,000th the intensity of the primary beam. If the scattering to be
studied is not too weak, it can be registered beyond an angle corresponding
to a 400 A spacing, with the necessary correct.ion for scattering from
sources other than the sample remaining reasonable.
At first glance it might be thought that the double monochromatization
and the large distance from X-ray tube to film would require a prohibitive
increase in the exposure time. Actually the loss of power of this system
as compared to that with a single monochromator is not too large. This
arises from the fact that with a single monochromator it is necessary to
use beams of very small divergence in order to limit the parasitic scatter-
ing, whereas with the double monochromator the final beam has more than
a 1° divergence.
Nevertheless, for ease of adjustment of the second monochromator
with a fluorescent screen, it is almost indispensable to have a very power-
ful source of X-rays. Our system employs a tube with a rotating target.
With this tube operating at 45 kv. and 45 ma. the exposure times are only
one-third as long as the exposures necessary with our previous system, which
employed a single monochromator coupled to an ordinary X-ray tube
operated at 30 kv. and 7.5 ma. For example, good patterns of a 19 per
cent solution of hemoglobin (Fig. 31) were obtained with one hour
exposures.
In summary, the system of two bent crystal monochromators is a
substantial improvement over the simple system of a single mono-
.chromator. It does not require the difficult adjustment of a slit in front
of the sample (see Fig. 29), and the patterns are much purer. Its range
of application, however, is somewhat limited; parasitic scattering prevents
approaching the direct beam to angles smaller than those corresponding
to 300 or 400 A for Cu KrJ. radiation (Fig. 31). We do not think that it
will be possible to better this limit with such a system. To register
larger spacings it is necessary to return to a system of successive slits, and
with that arrangement the double crystal monochromator.is no longer of
interest and a simple crystal monochromator must be used.

3.3.3.4. Monochromator with a Point Focus

Both single and double bent crystal monochromators have the advantage
of diminishing the width of the trace of the direct beam and, consequently,
EXPERIMENTAL EQUIPMENT 105

I
u
L..I
(a) (e)

I
u L..J
(b) (f)

L.......J L.....J
(c) (g)

L......l L......l
(d) (h)
:Fig. 31. Some examples of patterns obtained with the double
monochromator. The scale indicates 2° in 20. Patterns a, b, c, and d
were made with incident beams of small height and patterns e, f, g,
and h with beams of large height. X-ray tube input: 45 kv. and 40 ma.
(a) Carbon black. Exposure time: T = I hr.
(b) Solution of human hemoglobin (19%). T = I hr.
(c) Human red cells. T = 9 hr.
(d) Solution of albumin serum (23%). T = 1 hr.
(e) Chrysotile. T = I hr.
(j) Hemocyanine (snail blood). T = I hr.
(g) Egg yolk. T = 2 hr.
(h) Same sample as ind. T = 3 hr.
106 SMALL-ANGLE SCATTERING OF X-RAYS

increasing the accuracy of the pattern. The diffracted beam, though of

small width, is large in height, and this causes an angular uncertainty in
the pattern which, as we have seen (p. 93), becomes very important when
approaching very small angles. As in the system with collimators, this
height can be reduced only with the aid of slits, and these also reduce
the usable beam power.
The bent crystal focuses the rays in one dimension. The ideal system
would be one which concentrates the rays to a point instead of a line.
This is the objective of attempts that have been made to develop mono-
chromators with a point focus.
The first technique consists of curving the lattice planes onto a toroidal
surface. In order to describe this, let us consider the cylindrical mono-
chromator giving an image, F, of the point source, S, shown in Fig. 28.
Now let us imagine this system rotated about the axis, SF. The cylinder
generates a toroid, and, if the lattice planes were curved along this
surface, the radiation from S would converge to a point at F.
The toroid, however, is not a developable surface for crystals. A
crystalline slab cannot be shaped onto a toroid without breaking it or
introducing such perturbations in the lattice that all hope for good images
is lost. A solution of this problem consists in deforming a crystal
plastically rather than elastically (Wilsdorf (1948); DuMond [210];
Shacklett [149"]) and in following the plastic deformation with a "poly-
gonization" treatment (Guinier (1952)) by means of a suitable anneal.
This treatment transf~rms the distorted crystal into a mosaic of small
crystallites whose reflecting planes are very closely tangential to the
toroidal surface. Cauchois, Tiedema, and Burgers (1950) have shown
that excellent images can be obtained with an aluminum crystal cylin-
drically deformed and polygonized, which offers hope that a good toroidal
crystal can also be made. Hagg and Karlsson [219'] succeeded in
preparing a toroidal crystal of aluminum with which they obtained a
beam 50 times as intense as that from a cylindrically bent quartz crystal.
However, it has not yet been established that this kind of monochromator
is perfect enough to be usable in small-angle scattering experiments.
Another technique, based on the successive employment of two crystals,
is due to DuMond and his collaborators [229]. The two crystals are bent
slabs which have been cut so that the reflecting planes make a small angle
with the surface of the slab. Their arrangement is shown in Fig. 32.
The rays coming from the point source S are diffracted by the first crystal,
A, and converge towards the vertical line focus, F, parallel to the generator
of the surface of A. This converging beam appears to come from a
virtual line source, F', lying in the horizontal plane and situated on the
circle of center M and of radius SM, where M is the midpoint of the arc
EXPERIMENTAL EQUIPMENT 107

SA F. Let 0 be the midpoint of the line FF'. If we now pivot the rigid
system of source S and crystal A by a 90° rotation about the axis FF'
followed by a 180° rotation about a vertical axis through 0, the crystal
A is transformed into a crystal at B, and the source comes to the point P.
The beam coming from P and diffracted by B again is characterized by

F ~---------PYr-!'t-+r
I
I Projection
I upon
I vertical
I
I plane
I
I
I
I
I
I
I
I
r: _______________ , I _________ 1I
I
I I
I I
I
I
I
I Projection
I
I upon
I
I
horizontal
I plane
I
I
I
I
I
_______ l'...1.'-1--~-U-
F

Fig. 32. Horizontal and vertical projections of a point-focusing

double monochromator according to DuMond et a!. [229].

the same two foci, F and F'. Thus, by virtue of ordinary optical prin-
ciples, the beam reflected by the first crystal A and then by the second
crystal B will converge to the point P, giving a point image of the point
source S. Shenfil, Danielson, and DuMond [229] have studied in detail
the geometry of this system for the case of an extended source of X-rays.
They showed that for a given position of the crystals there was only a
small part of the area of the source from which the Koc radiation could
be twice reflected. The experimental results have confirmed the geo-
metrical predictions, and the authors have found a focus of dimensions
0.19 mm. X 1.41 mm. for a focusing distance of the order of 65 cm.
The optimum settings for the two crystals can be found only by means of
a systematic application of numerous adjustments; without such an
approach, no results can be obtained.
An important factor for such a system is the beam intensity remaining
after this double diffraction. A primary cause of loss of intensity is the
108 SMALL-ANGLE SCATTERING OF X-RAYS

polarization of the beam upon diffraction. If the Bragg angle for the
crystals were 45°, there would be complete polarization of the beam after
the first diffraction, and the intensity diffracted by the second crystal
would be zero. It is thus necessary to use planes affording small Bragg
angles. The first tests of this system were made with quartz crystals,

~ I ~_:-=~-----.J
+ .
Gas- f1 lled
~:~~d
cathode
\:_~nnular
stop
--- Photocell
X-ray tube monitor
Fig. 33. Sketch of a high-intensity small-angle diffraction unit for long
wavelength X-rays. A large solid angle of X-ray beam is totally reflected
and focused to a point by an ellipsoidal mirror of ground and polished
Pyrex. Reflection angles are very nearly equal to the critical angle
of reflection for the desired line radiation, so that all harder background
radiation is essentially cut off. (Henke and DuMond [76"''].)

using the (310) planes with a Bragg angle of 40°. The final beam intensity
was so weak that only samples that scattered strongly could be studied.
The authors have suggested the use of topaz rather than quartz crystals,
predicting an intensity increase of a factor of 4.8. A consideration of the
long path length of the beam brought out the advantage of placing the
apparatus in a vacuum or a helium atmosphere; the gain from this
modification would be a factor of 3.5. Also the use of an X-ray source
with a smaller angle of emergence from the target surface would permit
a further increase in intensity. When all these factors are considered, the
authors believe that a total intensity gain of the order of a factor of 40
over the first system can be obtained. Such an arrangement, associated
with either a powerful X-ray tube or one with a fine focus, would then be
capable of producing adequate scattered intensities.
The use of the two monochromators obviously gives this system the
same advantages (accuracy of beam definition, absence of parasitic
scattering) as those of the system in which the two crystals are arranged
with parallel axes of bending.
Another point-focusing system has been developed by Furnas [437),
in which transmission by a mica crystal is used after the first reflecting
quartz crystal; the principle of the technique is the same as that developed
by DuMond.
Let us also mention again the system of Henke and DuMond [76"")
which uses total reflection from a mirror having the shape of an ellipsoid
of revolution (Fig. 33). A ring-shaped screen placed after the mirror
EXPERIMENTAL EQUIPMENT 109

stops the direct radiation and allows only the reflected rays to reach the
point P.
Even if we could succeed in producing a beam that converged strictly
to a point, there would still remain one source of error, resulting from the
examination of the films with a microphotometer. Ge,ierally the

X-ray source

I
I
{Limiting slitsy /

First crystal

Fig. 34. A double plane crystal spectrometer for small-angle

diffre.ctiQil measurements.

exploring slit of this apparatus has a height of the order of 1 mm., and
this height will introduce an error in the measurements at very small
angles.

3.3.3.5. Double Monochromator with Plane Crystals

This system, proposed first by Fankuchen and Jellinek [212] and used
by Kaesberg, Beeman, and Ritland [221], is employed for quite different
purposes than those of the system with two bent crystals. Interest is
centered not on suppressing the parasitic scattering as much as possible but
rather on recording the crystalline reflections due to very large lattice
spacings.
The apparatus, shown in Fig. 34, may be described as a double crystal
spectrometer, with the crystals set in the parallel position.1 The two
crystals are very good specimens of cleaved calcite, and the goniometers
are constructed with sufficient precision to permit. the determination of
orientations to within seconds of arc. The detector, a Geiger-Muller
counter, is placed after the second crystal. The first crystal is held fixed,
and the intensity is measured as a function of the angle of rotation of the
second crystal. In the absence of a sample, a plot of this intensity
distribution gives a curve known as a "rocking curve," whose width is a
measure of the perfection of the crystal. Neither the spectral width of
the radiation nor the width of the slits affects the shape of this curve.
1 The theory of the double crystal spectrometer can be found treated in detail in

X·Rays in Theory and Experiment, by Compton and Allison (1935).

110 SMALL-ANGLE SCATTERING OF X-RAYS

The scattering by a sample is determined by performing the same

operation with the sample placed between the two crystals. The intensity
measured in excess of the normal rocking curve when the second crystal
makes an angle, e, with the first then represents the intensity scattered by
the sample through this angle. Since the width of the rocking curve of
good calcite crystals does not exceed 10", diffraction lines corresponding to
lattice spacings of 20,000 A can be resolved from the direct beam. The
angular uncertainty of the pattern, in the sense in which this has pre-
viously been defined (p. 87), can thus be 20 times as small as that of other
systems. It must be pointed out, however, that this double crystal
spectrometer functions schematically as a system of two slits of
infinite height. There is therefore a very important inaccuracy due to
the slit height at very small angles just as in the other systems employing
slits, except for samples so oriented as to give a diffraction pattern con-
sisting of parallel lines. Also there is no arrangement for suppressing the
parasitic scattering arising principally from the second crystal. This is the
source of the very strong small-angle parasitic scattering shown in the
curves published by Kaesberg and his collaborators, and others. This
system is not adapted to the study of weak, continuous scattering, but it is the
best for the study of crystalline diffraction arising from very large lattice
spacings.
Let us mention also a somewhat similar system devised by Banerjee
and Maitra [2], in which a single plane crystal is used, the first crystal of
Fig. 34 being replaced by a simple slit collimator whose width determines
the accuracy that can be attained. For a given setting of the CTystal two
lines are found in the plane of observation: one corresponds to the
characteristic Koc radiation scattered by the sample through an angle E
and then reflected by the crystal, and the other corresponds to a radiation
of unknown wavelength from the continuous spectrum which has
traversed the sample without deviation and has then been reflected by the
crystal (Laue reflection). The position of this second line determines the
orientation of the crystal and thus the angle of scattering, e, of the first
line. Since the intensity of a wavelength of the characteristic spectrum is
not very great, the intensity of the Laue line will be of the same order of
magnitude as that of the characteristic radiation scattered by the
sample.

3.3.4. MEASUREMENT OF mE TOTAL SCATTERED INTENSITY

Another use of the double crystal spectrometer has been suggested by
Warren [174]. The total energy scattered at small angles by a sample is
determined by comparing the energies received by the counter when the
sample is placed between the two crystals (position 1) and when it is
EXPERIMENTAL EQUIPMENT 111

placed just in front of the counter (position 2). In this second position
the scattered rays enter the counter, whereas in the first position they are
eliminated by the second crystal. Exploitation of this unique measure-
ment is possible by means of supplementary hypotheses concerning the
form of the particles in the sample, by neglecting all interparticle inter-
ferences, and by assuming that when the sample is in position 1 the
fraction of scattered energy superposed upon the transmitted beam and
reflected by the second crystal is small. Warren has shown that, if the
sample is assumed to be constituted of spheres of radius r, the difference
between the mass absorption coefficients of the sample in positions 1
and 2, µ 8 .A., is proportional tor. The numerical relation was shown to be

where A. and r are expressed in angstroms and p is the density of the sample.
The measurements cited by Warren for a carbon black give a value of r
in good accord with that deduced from an analysis of the low-angle
continuous scattering curve.
If the assumption of spherical particles is not made, the measurement
of the total scattered energy can still be used by means of equations 2.30
and 2.110.

3.4. METHODS OF CORRECTION OF EXPERIMENTAL

SCATTERING CURVES
We have already emphasized that the inaccuracy produced by the
dimensions of the direct beam increases as the center of the pattern is
approached. This is particularly serious in a study of continuous
scattering, where there is need for the entire curve of the intensity and
not simply the positions of the maxima, as is the situation in studies of
diffraction lines or spots. The errors are due to the width and the height
of the direct beam. We have seen that the bent crystal monochromator
with its convergent beam allows a reduction of the error due to the beam
width without diminishing the beam intensity. It is not actually possible,
in the absence of point monochromators, to avoid the error due to the
beam height. Instead of trying to reduce this error at the price of a very
great loss in intensity, methods of correction to determine the real curve
from the measured curve have been sought. Such methods exist for the
correction of errors due both to the width and to the height of the beam,
but, because of the persistence of the slit height error noted above, it is
particularly the latter correction that is important.
The cross section of the direct beam in the plane of observation is
shown in Fig. 35a. Let the intensity of the direct beam at the point
112 SMALL-ANGLE SCATTERING OF X-RAYS

P(x, y) be denoted by i(x, y) dx dy. In order to simplify the notation

for the following calculations, we have adopted as coordinates of P

x = (2TT0Pz)/s).

y= (21TOP~)fs).

where s is the sample-to-film distance.

y y

dx
p
Py ~ H

M
0 Px x x

(b)

(a)
Fig. 35. Calculation of the effect of beam dimensions.

Radiation scattered from this part of the direct beam to the point of
observation, M, will be scattered by the sample through an angle equal,
as a first approximation, to 20(x, y) = PM/s, so that the parameter h
has the value h(x, y) = (2TTPM)/s}..
We shall assume that the scattered intensity depends only on the
scattering angle, this dependence being expressed by the function l(h).
The point of observation Mis defined by the parameter h.u = (21TOM)/s).,
and the observed intensity at this point is

J(h) = ff i(x, y)I(h(x, y)) dx dy

The problem is to deduce l(h) from the measured valuesJ(h).

3.4.1. CORRECTION FOR THE EFFECT OF BEAM WIDTH

We shall consider a beam of rectangular cross section in which the
distribution of intensity along the vertical direction is similar for all
values of the horizontal coordinate, that is, i(x, y) can be written
EXPERIMENTAL EQUIPMENT 113

LetJ1 (h) be the scattered intensity that would be obtained at a point

a distance (hls)/27T along the x-axis from an infinitely narrow incident
beam containing the same power as the real direct beam, the intensity
distribution of this incident beam in the vertical direction being pro-
portional to i 2 (y). Let i1.0(x) dx be the fraction of the total power of the
real direct beam striking the plane of observation between the abscissae,
x and x + dx; that is,
. i 1 (x)
J
ii.o(x) = i1 (x) dx

The contribution of this elementary direct beam to the observed scattered

intens!ty at Mis i1.0(x)J1 (h - x) dx. Therefore

J(h) = f i1.0(x)J1(h - x) dx (8)

An integral equation of this form can be solved by means of Fourier

transforms. We shall assume that J(h), J 1 (h), and i1.0(x) are even
functions. Their transforms are then defined by the following equations:

J*(u) = f 0

00
J(h) cos 27ThU dh

J l *(u) = f_"""" Jl(h) COS 27ThU dh

i* 1.0 (u) = f_: i1.0(x) cos 211xu dx

One can show that equation 8 can be transformed to the following

equation:
(9)

The correction is found by calculating the transforms of the two known

functions,J(h) and i1.0(x), and in forming their quotient. The transform
of this quotient will be the desired function, J 1 (h).
Let us point out thatJ(h) and i 1 .0 (x) are measured by means of a counter
or microphotometer and are not known precisely because of the finite
width of the exploring slits. But, if the same slit is used to explore the
direct beam and the scattered intensity, equation 9 is still correct, since
an application of the same process shows that the effects due to the
"slit transmission" function disappear.
Theoretically one can thus correct exactly for the effect of the width
by this process, regardless of the magnitude of this width. Actually we
find in practice that the mathematical operations for calculating the
transforms give precise results only on condition that the function i 1•0 (x)
114 SMALL-ANGLE SCATTERING OF X-RAYS

is zero outside of an interval which is small with respect to that in which

f(h) undergoes notable variations. If, for example, the width of a
continually decreasing central scattering were not much larger than that
of the direct beam, the transform of i1.0(x) would have zeroes in the
region in whichf1 (h) is to be calculated, which would lead to a very poor
determination of f 1 (h). For the same reasons it is necessary that the
exploring slit used to determine i1.0(x) be small with respect to the width
of the direct beam (less than 10 or 20 per cent).

3.4.2. CORRECTION FOR THE EFFECT OF BEAM HEIGHT

After the correction for the beam width has been made, we must
correct for the effect due to a beam which is infinitely narrow but has a
cl•rtain height along the y-axis. Let M again be the point of observation
(Fig. 35b), and let the position of the point R in the direct beam be
designated by the coordinate y = (2TTOR)/s).. The fraction of the total
beam power contained in the segment of dimension dy, at the ordinate y,
will be denoted by i2,0(Y) dy, where
. i2(y)
i2.o(Y) = Ji2(Y) dy

If /(h) is the true scattering distribution (i.e., that obtained with a

direct beam of infinitely small radius passing through 0), the observed
intensity at the point M is

f1(hl = Ji2.o(y)J(Vh 2 + y2 J dy (10)

where
h = (2TTOM)/s).
A very important fact must be pointed out here: when the true
scattering distribution is represented by a Gaussian curve, the observed
curve is proportionaltotherealcurve. This is easily seen, forif /(h) = Ae-.t"l•'

f 1 (h) = [f 2 . 0 (y)e-k'u' dy] Ae-k'h' =constant X I(h) (11)

If the beam height is large and the intensity of the beam is a constant
c, independent of y, equation 11 becomes

(lla)

This demonstrates that the proportionality factor between the measured

and true intensity distributions depends on the coefficient in the exponent.
This fact, first pointed out by Hosemann [79], is of considerable interest,
EXPERIMENTAL EQUIPMENT llll

since an exponential distribution is often a good approximation of the

curves of low-angle scattering.
Shull and Roess [155] have pointed out a possible extension of this
remark. If the experimental curve is not sufficiently well represented
by an exponential function, one can try to separate it into a series of
exponentials. Thus, if the measured intensity can be written as
$ 1 (h) = 2,T;e-klh1
i

the transform of l(h) will be the sum of the transforms of the component
exponentials. The expression for the true intensity distribution will
then be
l(h) = J. T/e-k/h•
i
where
I ki
T• = y1; Ti
This method is not generally usable, since not only is the separation
of an arbitrary function into a series of exponentials sometimes impossible, 1
but also, even when it is possible, a general method for performing the
separation is lacking.
When the equation of the scattering curve for an incident beam of
point-like cross section is known, the scattering curve obsenred for a beam
of infinite height can always be calculated by means of equation 10.
This was illustrated for an exponential scattering curve in equation 11.
Several other such results are found useful:
(i) If the asymptotic shape of the intensity curve is of the form of
Kh- 4 (equations 2.26 and 2.108), the observed curve varies asymptotically
as h- 3 . This is seen from the following:

J1(h) = K I-co
C()

(h2
1
+ y2)2dy
By making the substitution y = h tan ex

..?i(h) =
2K
h3 Jo
f"' 2 -(I_+_t_a_n_dcx_cx_)_-co_s_
2 2 2 _:i:

2K f"' 2 7TK
= h3 Jo cos 2 a da = 2h 3 (llb)

or

1 The group of functions e -k/h' does not form an orthogonal group.

116 SMALL-ANGLE SCATTERING OF X-RAYS

(ii) In equations 2.28 and 2.111 the integral i 00

h 2/(h) dh appears. In
terms of the observed intensity, this should be replaced by the integral

Ii"'
-
2 0
h.F1(h) dh. This can be shown as follows:

["' h.F1(h) dh =
Jo
["'
Jo
f 00
-oo
hl(vh 2 + y 2) dh dy
By making the change of variables y = z sin ct and h = z cos ct so
that
y2 +
h,2 = z2 and dh dy = z dz dct

Jo
["' h.F1(h) dh = f
Jo
f"'
00
2
-,,,/2
I(z)z 2 cos ct dz drr..

= 2 L 00
z2I(z) dz (Ile)

In the same way it can be seen that

(lld)

These results are valid only if the beam can be considered as infinitely
high even for the most distant parts of the curve.
(iii) Schmidt, Kaesberg, and Beeman (1954) have calculated numerically
the scattering curve observed for spherical particles in a dilute system
when the incident beam is of infinite height. (The scattering for an
incident beam of point-like cross section is given by equation 2.31.)

3.4.2.1. Slit Correction for Infinite Height

A rigorous method of correction for this case has been described by
DuMond [209] and by Guinier and Fournet [216), [217). It is rigorous
only for the particular case in which the beam is of uniform intensity and
infinite height. However, if the low-angle scattering decreases con-
tinuously from the center, this method can be applied if the beam intensity
i 2 (y), is constant up to a value of y such that .F1 (y) is negligible. If the
pattern consists of one ring of scattering, the beam must have a height
at least equal to the diameter of the ring for this method to be applicable.
The true intensity distribution is determined by the following equation:

I(h) = - -1 i"" .F
1TC O
1 '(Vh 2 + u 2 ) du
yh,2 + u2
(12)
EXPERIMENTAL EQUIPMENT 117

where c designates the constant value of i 2.0 (y},

.f1'(vh2 + u2) = d.f1(V~)

d(vh2 + u2)
and u represents a variable of integration of no physical significance.
The derivation of this equation may be outlined as follows:
Equation 10 can be written 118

J.fh) = 2ci J(Vh•

00
+ y•) dy
Differentiation with respect to h gives the following:

.f1'(h) = 2ci 0
oo
J'( V h'
+ y•
+ y 1) • I
v h'
h
dy

Dividing by h and then changing the variable h' into h' + u• gives

.f1 '(Vh 1 + u')

--~~~=2c
Vh'+u•
i""
l'(Vh' + u• + y•)
o
dy
Vh'+u•+y•
Integrating with respect to u

"" .f1 '(Vh' + u') iooioo l'(Vh' + u' + y')

i --===~ du = 2c du dy
o v'h• + u' o o Vh' + u• + y•
A second change of variables, u = r cos 0, y = r sin 0, then gives

i o
+ u") du
oo .f1 '(Vh'
---====--
v'h• + u•
= 2c
i"' i""
o
2

o
l'(Vh' + r')
Vh' + r•
I r I dr dO

= C7r[J(Vh 2+ r')]F-+OO = -CWJ(h)

r=O
on condition that l(h) approaches zero as h approaches infinity, which is true for
scattering experiments at very small angles.

The use of equation 12 requires the determination of the derivative of

the experimentally determined function 5 1(h). Then, for each value of
h , a curve of ..1'i'(Vh + u ) is
2 2 • t race d and a graphical m
. t egrat'10n is
.
vh + u
2 2

performed. The process is tedious, but it can be precise, as has been

verified by Guinier and Fournet [217].
When the function.f1(h) differs only slightly from a Gaussian function,
the calculation may be shortened considerably by writing
(13)

where T, k, and f(h) are determined by means of an auxiliary graph of

ll8 SMALL-ANGLE SCATTERING OF X-RAYS

log.f1 (h) as a function of h 2 • Application of equation 12 to equation 13

then gives

I (h) Tk -k'h'
= ---= e - -
1f"'f'(Vh2 + u2) du
cV7T 7TC o Vh 2 +u 2

This approach leads to a. much greater accuracy in the determination of

the function l(h).

3.4.2.2. Case of a Beam of Arbitrary Height

If i 2 (y) depends on y and, in particular, if the beam is of finite height
(for example, i 2 (y) is a constant between -y 0 and Yo and is zero outside of
this interval), the property discussed above of curves of the form e _.,..,,. is
still valid. For this type of curve the measured distribution can be used
directly, whatever the intensity distribution of the direct beam, if only
the shape of the curve is important. If absolute measurements of
intensity must be made, equa1rion 11 must be employed to determine the
proportionality factor.
The general problem of the arbitrary curve that cannot be resolved
into a sum of exponentials remains to be considered. The solution we shall
outline here was pointed out by Kratky, Porod, and Kahovec [111].
The equation to be solved is

(10)

in which i2.0(Y) is some arbitrary function of y. Porod has shown that

the solution for this general equation, a relation analogous to equation 12
but containing another function, g(u), is

/(h) = - ~ f "'.f1~Yh2 + u2) g(u) du (12a)

1TC Jo Jt2 + u2
The function g(u) is determined by the following condition: on making
the change of variables u = r cos() and y = r sin 0, the integral
("'2
Jo i2.0(r sin O)g(r cos 0) d() should be a constant, i.e., independent of r.
If i 2 •0 is a constant, g(u) is a constant, and equation l2a reduces correctly
to equation 12.
The authors did not indicate a method, even numerical, for determining
g(u) in the general case.
When the beam is of uniform intensity and limited to the interval
between -y 0 and y 0 , these authors indicated a solution which, though
not rigorous, should certainly be a good approximation. An exact
EXPERIMENTAL EQUIPMENT 119

solution of the function g(u) was determined for values of u less than
Yo v2. For large values of u they found that the function tended towards
the value g(u) = u/y 0 , oscillating about this quantity. For their applica-
tion they adopted the following values:

g(u) = 1 u <Yo
g(u) = 2 Yo :S: u < YoV2
u
g(u) = -
Yo
u > y0 v2
In practice, the method of application of this correction is the following:
the function

F(u) = - ----
+ u2)
.f1'(vh2
yh2 + u2
is constructed as in the previous case for an infinite slit. The integration
is more complicated because of the presence of the function g(u). First
the integration of the function F(u) is performed over value's of u ranging
from 0 to y 0 • To this is added twice the value of the integration over
values of u ranging from Yo to Yo v2. The final contribution is that given
by the integral of the product (u/y 0 )F(u) over values of u ranging from
y0v2 to infinity. Kratky and Porod have shown that this last contri-
bution is given by the ordinate of the experimental curve at an abscissa
of Vh 2 + 2y 0 2 . Tests by the authors of this method have shown that
the accuracy obtained is of the same order as that found in the correction
for beams of infinite height by means of equation 12.
It is not yet possible to say that definite, complete solutions exist
for the problem of correction for the height of the incident beam. Thus
in practice one should either employ very high beams of uniform intensity
and make an important, but calculable, correction, or else diminish the
height of the beam as much as possible to diminish the error. We believe
that this last solution is to be recommended, provided that the study is
not extended to very small angles.
Actually, if the correction is small, an approximate method pointed out
both by Fournet and Guinier [216] and by Franklin [214] can be used.
The beam is assumed to be of uniform intensity, c, and limited to values
of y between -y0 and y 0 . The function I(h) is expanded in terms of
.f1 (h) and its derivatives, the first terms being

1
(y
2cy l(h)=.rd: (h)--l - 0 )
2 d11 (h)
h ---- 1
- 0) (y 4 [
hd.f1 (h) d 2Y 1 (h)]
- - - h2 - -
0 6 h dh 360 h dh dh 2
120 SMALL-ANGLE SCATTERING OF X-RAYS

This method still requires the determination of the slopes of the experi-
mental curve, but these enter only in the corrective terms.
The above relation can be \I-Tit.ten, to a slightly poorer approximation, as

2cy 0 l(h) =f1 (J h2 - ~~:)

3.5. CONSTRUCTION OF LOW-ANGLE SCATTERING SYSTEMS
Systems employing collimators or monochromators have been built
following the principles given in the preceding sections, and detailed
descriptions of the different pieces of equipment used can be found in the
references given in the general bibliography. Our discussion here will be
limited to indicating the difficulties generally encountered and the
solutions that have been proposed.

3.5.1. SLIT CONSTRUCTION

In systems employing either collimators or monochromators, slits play
an essential :r:ole in defining the beam and in eliminating the parasitic
scattering. These must be very carefully made, and, in the majority of
cases, their edges should be independently adjustable to a precision of
approximately 0.01 mm. Finally,. they should be designed so as to
produce the least possible parasitic scattering when they are touched by
the direct beam.
Kratky et al. [111] have pointed out the existence of a particular small-
angle scattering effect arising from the slits which may be described as
follows: when the volume of matter bathed by the X-rays is rather small,
as the result of a high absorption coefficient or of the form of the slit edge,
a small-angle scattering is created, with the irradiated volume playing the
role of the scattering particle.
In order to limit the extent of this scattering, the volume effectively
bathed by the direct beam must be increased. This can be done by
causing the slit to have sharply angled edges made of a substance of low
atomic weight. However, if the slit edges are very thin and not very
absorbing, the beam is not sharply defined. The best method is to define
the beam precisely with slits formed of steel cylinders with diameters of
2 to 3 mm., for example, and to place a third slit after the first two whose
function is to intercept the parasitic scattering. The usable region in the
plane of observation is then the part protected by the third slit. This
third slit in particular should be very carefully made so that its edges can
be independently positioned about 0.01 mm. away from the main beam.
The only important slit in a system that employs a monochromator is
the one that stops the parasitic scattering but does not touch the main
EXPERIMENTAL EQUIPMENT 121

beam. The nature of the edges of this slit is not important, but their
construction should be very precise.
The ll.djustment of the slits cannot be made visually by means of a.
fluorescent screen, since in general the luminosity of the pattern is
insufficient to allow the necessary accuracy to be obtained. The beam
form may be recorded photographically, successive exposures determining
whether the edges of the slit to be adjusted are touching the beam. If a
Geiger-Muller counter is to be used, the power of the direct beam may be
measured and each of the edges advanced in turn until the measured
power begins to diminish.

3.5.2. STOPPING THE DIRECT BEAM

It is essential to prevent the direct beam from striking the photo-
graphic film. It is necessary also that the device used to stop the direct
beam does not itself produce scattering. A final requirement of the beam
stop is that it should not mask the film in its useful region, that is, outside
the zone fogged by parasitic scattering. The construction of a beam stop
that will satisfy these conditions is made difficult by the fineness of the
direct beam. The most common device is a fl.at band of metal placed
several millimeters from the film. The width of the band is made slightly
larger than the beam width, yet slightly smaller than the width of the
fogged zone (the quantities a and b, respectively, of §3.2.1), as is shown in
Fig. 36a.
It is advantageous to use a rather thin band, so that in the course of the
normal exposure the trace of the direct beam will appear on the film with
a blackening comparable to that of the scattered radiation. By this
means a positioning mark is always available. If the direct beam is mono-
chromatized there is a much more important point of interest to the trace
of the direct beam, which is that its intensity is proportional to the
intensity of the incident beam. This fact allows the measurements of
the scattered intensity to be made in absolute units (see §3.5.3).
With a Geiger counter detector, the slit before the counter plays the
role of the beam stop. It should therefore be thick enough to stop the
radiation completely and constructed in such a way as to prevent any rays
scattered by the slit from entering the counter (Fig. 36b).

3.5.3. ABSOLUTE MEASUREMENTS

The difficulty in absolute measurements arises from the great difference
in magnitude between the intensity of the direct beam and that of the
scattered radiation. Such measurements obviously are worth while only
when the direct beam is strictly monochromatic.
If the primary intensity is reduced by an absorber, one essential
122 SMALL-ANGLE SCATTERING OF X-RAYS

precaution should be observed. In general a monochromator crystal

reflects not only the useful wavelength but also the harmonies of wave-
lengths J./2, J./3, etc. If the absorbing band has a much smaller coefficient
of absorption for the wavelength A/2 than for A, the trace of the direct
beam contains the radiation A/2 in a considerable proportion and thus

~-- . I

- )__.Sam-pie
(b
-----1 E Slit

Sample

(c)

Fig. 36. Beam stops; (a) usual device with photographic detection;
(b) arrangement for counter measurements; (c) arrangement proposed
by Kratky.

does not give correct information on the intensity of the primary radiation
of wavelength A. This source of error can be avoided by choosing for the
absorbing band a substance having a coefficient of absorption for J./2 at
least as large as that for ). as a result of its K absorption discontinuity.
The proportion of J./2 in the trace is then equal to or less than its pro-
portion in the direct beam. For example, copper or nickel serve well for
Cu Kl'J. radiation and silver serves for Mo Kl'J.. The beam stops in our
present system are made from copper sheets of 0.20- to 0.25-mm. thickness.
This reduces the Cu Kr1.. direct beam in a ratio of 1.2 X 10-4 to 1.2 X 10- 5•
The error due to the J./3 radiation is not suppressed, but in general this is
unimportant since the proportion of A/3 in the primary beam is very small.
An inaccuracy in the process of normalization to absolute units that
requires the measurement of the total power of the direct beam arises from
EXPERIMENTAL EQUIPMENT 123

the fact that the coefficient of absorption is never known to a high enough
accuracy to permit a sufficiently sure determination of the weakening of
the direct beam.
Kratky and coworkers [111] have developed a very ingenious method
for reducing the intensity of the direct beam to a quantity measurable by
photographic techniques. They use an X-ray tube powered by unrectified
high voltage. The image of the direct beam is received on a photographic
plate which is moved in the plane of observation at such a speed that the
traces due to successive emissions are resolved. Then by microphoto-
metering one such trace the incident energy can be evaluated for a.n
exposure time of 1/50 second (for 50-cycle current); this is 1/180,000 of
the normal exposure time of 1 hour generally necessary for a scattering
pattern. In order that the measurement furnish the correct factor, the
X-ray tube power must be kept constant. It is not necessary that the
displacement of the plate be at constant velocity, so this motion can be
provided simply by hand.
Geiger counters cannot measure intensity ratios in excess of 1000
because of counter saturation at high intensities. Therefore, if a counter
is to be used for absolute measurements, we must either reduce the
intensity of the direct beam by absorption or else measure the scattering
from a sample whose scattering power is known (an ideal gas [53] or some
light material whose high-angle scattering is essentially free from inter-
atomic interferences).
Ionization chambers were generally discarded when Geiger counters
were developed because electrometer measurements were much more
difficult than the counting of impulses. These chambers, however,
have a big advantage in that precise and direct absolute measurements can
be made, for if a sufficiently high collection voltage (800 volts) is used there
is no recombination of ions in the chamber, and the charge collected by
the central electrode per unit time is proportional to the X-ray flux, even
for very intense beams. Laval (1939) and his coworkers (Olmer (1948);
Curien (1952)) measured the ionization current by compensating the
collected charges with charges induced on the chamber itself (capacity y)
for low beam intensities and with tJharges induced across a condenser of
capacity C for high intensities (Fig. 37). The sensitivity of the chamber is
reduced by the ratio of the capacities y/C in the latter case, and this
ratio can be measured very accurately.

3.5.4. VACUUM APPARATUS

It is essential to eliminate air scattering in small-angle scattering
experiments, since this scattering can very often be stronger than the
intensity scattered by the sample. The best method for eliminating it
124 SMALL-ANGLE SCATTERING OF X-RAYS

is to include all the apparatus between the tube and the detector either in
an evacuated chamber [48] or, if this is not possible, in an enclosure filled
with hydrogen or helium (Fig. 25b) [376], which scatters only weakly. If
such a large vacuum chamber cannot be used, a system must be employed
in which the air is eliminated in that part of the beam path between the
sample and the film or counter.

Ionization
chamber

2v.
Fig. 37. Ionization chamber circuit for absolute measurements of
the scattered intensity. For low intensities switches 1 and 2 are closed
and 3 and 4 are open. For high intensities switches 1 and 2 are open
and 3 and 4 are closed. The electrometer E is maintained at its zero
by adjusting the potentiometer P. If an adjustment of n divisions is
required in time t, the ionization current is given by (n/t)y or (n/t)O for
the low- and high-intensity settings, respectively (Curien (1952)).

Kratky [lll] proposed the following system to simplify the apparatus.

The beam stop is moved to a distance s' from the film, as is shown in
Fig. 36c, thus suppressing the path of the direct beam in air in the regions
closest to' the plane of observation. These regions are the most harmful,
since the scattered intensity measured at a point varies inversely as the
square of the distance from the scatterer.
To find the order of magnitude of the improvement, let us consider a
slit collimator for which the fogged region has a width b twice that of the
direct beam, a. Here the beam stop can be placed at a point approxi-
mately one-third of the distance from the film to the sample without
screening a part of the useful region. With the beam stop against the
film, the scattering from the air traversed by the direct beam between the
sample and the film that is received at a point a distance l from the direct
beam trace will be proportional to the integral
r· dx
Jo x2 + l 2
EXPERIMENTAL EQUIPMENT 125

With the arrangement proposed by Kratky, the scattering will be pro-

I.•
portional to
dx
•/3 x2 + z2
The gain by this arrangement, proportional to s/l, is approximately a
factor of 20 at an angle 20 = 1°. The method is thus simple and efficient
for very small angles.

REFERENCES FOR CHAPTER 3

Arndt, U. W., Coates, W. A., and Riley, D. P. (1953), Proc. Phys. Soc. (London),
66, 1009.
Cauchois, Y., Tiedema, T. J., and Burgers, W. G. (1950), Acta Cryst., 3, 372.
Compton, A. H., and Allison, S. K. (1935), X-Rays in Theory and Experiment,
Macmillan, New York.
Curien, H. (1952), Bull. soc.jraw;. mineral. et crist., '!5, 197.
DuMond, J. W. M. (1947), Rev. Sci. Instr., 18, 617.
Evans, R. C., Hirsch, P. B., and Kellar, J. N. (1948), Acta 01'yst., 1, 124.
Fankuchen, I. (1937), Nature, 139, 193.
Guinier, A. (1946), Cornpt. nnd., 223, 161.
Guinier, A. (1952) (Ed., W. Shockley), Imperfections in Nearly Pef'ject Crystals,
Wiley, New York, p. 402.
Johann, H. H. (1931), Z. Physik, 69, 185.
Johansson, T. (1933), Z. Physik, 83, 507.
Kirkpatrick, P. (1939), Rev. Sci. Instr., 10, 186.
Laval, J. (1939), Bull. soc. /ran~. mineral. et crist., 62, 137.
Lipson, H., Nelson, J.B., and Riley, D. P. (1945), J. Sci. Instr., 22, 184.
Olmer, P. (1948), Bull. soc.jran~. mineral. et crist., 71, 144.
Schmidt, P., Kaesberg, P., and Beeman, W. W. (1954) Biochim. et Biophys. Acta,
U, l.
Wilsdorf, H. (1948), Naturwiss., 35, 313.
4. METHODS OF INTERPRETATION
OF EXPERIMENTAL RESULTS
In Chapter 2 the small-angle scattering effects were calculated for
scattering bodies of given structure. Then in Chapter 3 we reviewed the
experimental methods and equipment that are actually employed. Now
we can approach the real problem, which is the determination of the
structure of the scatterer from the patterns of the small-angle scattering by
means of the theories developed in Chapter 2. We shall examine
successively the different cases that were considered in that chapter:
widely separated identical particles, closer-packed identical particles,
groups of particles of different sizes, etc. Then, from what is known of
any scattering body, we can determine in advance of an investigation the
particular case out of those enumerated into which it must fall. We
shall also give the criteria of validity of the hypotheses which are adopted.

4.1. IDENTICAL PARTICLES

Let us begin with the simple case of widely separated particles.

4.1.1. WIDELY SEPARATED, IDENTICAL PARTICLES

4.1.t.l. Equal Probability of All Orientations
We have seen (equation 2.55) that the scattered intensity is given by
F 2 (h), apart from a multiplicative constant. Consequently, we can
write
log /(h) =log F 2 (h) + constant (1)
The approximate law of Guinier (equation 2.39)
111 Re1
F 2 (h) = F 2(0)e - - 3- (2)
can be written as
-- h2Ro2
log10 F 2 (h) = - - 3- log10 e + constant (3)

Therefore
- h2R2
log10 /(h) = - ~log 10e +constant (4)

The method of analysis is thus simple; the curve of log /(h) vs. h 2
(hereafter denoted by log J(h 2 )) is plotted, and from its slope the radius
126
INTERPRETATION OF RESULTS 127

of gyration is determined. In the experimental systems generally used

the parameter which is most easily obtained directly is the tangent of the
scattering angle 20 (tan 20 = x/s, where x represents the distance from
the trace of the direct beam to the point on the film being considered,
and s is the sample-to-film distance). In the small-angle region we are
considering, an angle, its sine, and its tangent can be interchanged
without error. Therefore
sin 0 27T
h= 4rr-y ~ 'T tan 20

and
- 411'2
log10 /(h) =-(tan 20) 2 322 R 0 2 log10 e + constant (5)

The slope, -p, of the representative curve of log /(h) as a function of

(tan 20) 2 is thus equal to -(47T 2/3A. 2)R 0 2 log10 e. Therefore

I
Ro=-
27T
J--A.Vp
3
log10 e
(6)

For example, with Cu Koc radiation (A. = l.540 A)

R 0 = 0.645Vp (Angstroms) (7)
If, instead of using the approximate law of Guinier, we use the exact
formula for the structure factor,
-
F2(h) = -2
[
F (0) I - T
h2R 2
+ och 4 + · · ·] (8)

we find
- h2R 2
log10 I(h) = - - -0 log e
3
10 + fJh"' + · · · + constant (9)

From this we can see that, rigorously speaking, the slope p that should
be used in equation 6 is not the average slope of the curve of log I(h 2 ) but
rather the value of the slope at the origin, that is, the slope of the curve
at h = 0.
We know (cf. Chapter 3) that measurements of the scattered intensity
cannot be accurately made below an angle O' which is very small but not
zero (this corresponds to the value h' for the parameter h). Therefore an
experiment cannot give directly the slope that is required. However,
the slope at the origin is the limiting value of the slopes for different
values of h 2 as h 2 tends to zero, so that, if measurements are made of the
slope -p(h 2 ) of the curve log J(h 2) at several points, the limiting value,
-p(O), can be determined by means of an auxiliary graph. This extra-
polation is very precise if the experimental curve shows only a small
curvature over a relatively large part of the small-angle region.
128 SMALL-ANGLE SCATTERING OF X-RAYS

For a particle of given form and radius of gyration R 0 there is a limiting

value, h" R 0 , of the product hR 0 beyond which it is evident that the approxi-
mate law of Guinier is no longer valid. This is illustrated by the graphs
of log J(h 2 R 02 ) for a sphere (Fig. 38) and for a homogeneous, one-
dimensional rod (Fig. 39), in which the clotted lines represent the extra-
polations of the slope at the origin. For the spherical particle the
log I

-1

-2

1.69 2 3 4 5 (hR0)2

Fig. 38. Log I vs. (hR 0 ) 2 for spherical particles.

Guinier formula is a very good approximation at small angles, remaining

valid up to the limit, h" R 0 = 1.3. For the rod the Guinier formula is not
as good an approximation. It is precisely for particles of this rod-like
shape that the difference between the approximate law and the exact
expression is greatest; here h" R 0 is only of the order of 0.7. If we assume
that the slope of the curve at h' will correctly determine the radius of
gyration if h' :S:: h" /2, we can conclude that an experimental system
characterized by a lower angular limit, h', will be capable of correctly
measuring the radius of gyration of spherical particles up to an upper
limiting radius of 0.65/h' A, whereas for rod-like particles the limiting
radius will be only 0.35/h' A. Actually it is possible to extend this limit
somewhat. Instead of simply tracing the curve of log J(h 2 ), we can
plot the curve of the slope, -p(h 2 ), and extrapolate to the origin, h 2 = 0.
This method of extrapolation is illustrated in Fig. 40 for the unfavorable
ease of a homogeneous one-dimensional rocl in a system such that h' R 0
= 1.04. The value of -p(O) obtained by extrapolation, 0.0492, is very
INTERPRETATION OF RESULTS 129
log I

-1

-2

0.49 2 3 4 5
Fig. 39. Log I vs. (hR 0 ) 1 for homogeneous one-dimensional rods.

0.050

1Limit of the experiments

0.040

0.030

0.020

0.010

2 3 (hRof
Fig. 40. Graph of the slope of log
-p(h1 ) as a function of (hB0 ) 1 •
J(h1 )
The slope -p(O) is found by extrapolation.
130 SMALL-ANGLE SCATTERING OF X-RAYS

nearly the correct value, 0.0483, even though the value of h' R 0 is con-
siderably larger tha.n h" R 0 = 0. 7.
Within the limits just defined, it is always easy to determine the radius
of gyration of the particles being examined. The radius of gyration is the
only precise parameter which can be determined by small-angle sr,attering
experiments without invoking supplementary hypotheses. It is evident that
this one para.meter does not entirely define the particle. The curves of
log J(h 2) for particles having the same radius of gyration but different
forms will have the same slope at the origin and will coincide so long as
h < h", but beyond this point they will diverge. The use of the entire
curve ii). the determination of the form of the particle is very difficult in
practice, since rather large changes in the form of the particles produce
only small effects on the shape of the tail of the scattering curves.
Theoretically, we could trace a priori a family of curves of log I as a
function of log h for different forms of the particles by means of the
equations established in §2.1, and then, neglecting simple translations,
determine which of the different curves most closely corresponds to the
experimental curve as plotted in the same coordinates.
Let us, however, cite a rather remarkable exception. The experi-
mental curve for very homogeneous spherical particles can be made to
coincide with the theoretical curve <l> 2 (hR) (see equation 2.31) over a very
large region; in fact, the successive maxima predicted by theory have been
experimentally demonstrated ([186], [25], [76'"']) (Fig. 41). This last
result proved tha.t the form of the particles was spherical a.nd allowed the
determination of the radius of the particle. This radius agreed well with
that calculated from only the very low-angle part of the curve by means of
equation 6.
In the general case, once the radius of gyration is determined, further
details about the form of the particle can be obtained either by employing
auxiliary information or else by making certain hypotheses. For ex-
ample, if the volume and the general form of the particle are known, its
dimensions can be calculated.
As a particular example, let us consider the study of horse hemoglobin
given by Fournet [48). The particles, which can be assumed to be mole-
cules, were found to have a radius of gyration equal to 23 A. The
molecular weight of this substance is 66, 700, and the density of its
solution in water is 1.33 (Perutz (1946)). This auxiliary information
determined the volume of each molecule as 83,400 A3 .
If we assume that the molecule has the form of a right cylinder of
radius R and height 2H, two relations between R and H can be deter-
mined: the volume of the molecule is V = 2TTR 2H = 83,400 A 3 , and the
square of the radius of gyration is R 2 = (R 2/2) + (H2/3) = (23)2.
INTERPRETATION OF RESULTS 131

1.0mm.

Fig. 41. Scattering pattern of uniform spherical particles (Dow

Latex) obtained with a point-focusing monochromator (Fig. 32).
The first visible ring is the fifth maximum (Fig. 6); on the original film,
the rings are visible up to the seventeenth maximum. Sample-to-film
distance: 66 cm_ Cu Koc radiation. Exposure time: 129 hours.
(Danielson, Shenfil, and DuMond [251.)
132 SMALL-ANGLE SCATTERING OF X-RAYS

These relations can be expressed as two curves in the R-H plane (Fig. 42)
whose intersections define the dimensions of the only possible cylinders
which can fit all our information. Figure 42 shows that there are two
possibilities: (a) 2R = 42 A, 2H = 120 A; (b) 2R = 60 A, 2H = 30 A.
This last size is very close to that proposed by Perutz [407], which was

Hin A

50
v = 83,400A3

40 ,,,
II ' \
I \
I \
I \
30 \ \
\ \
\ I
\ \
\
'' \
I
20 ' I
'\
1',
--Cylinder
I
I ''
10 I
- - - - Ellipsoid of revolution
'
I
I
I
0
0 10 20 30 40 R in A
Fig. 42. A graphical approach in the study of the form of particles.
The full line curves correspond to cylinders of radius R and height 2H
and the dotted curves to ellipsoids of axes 2R, 2R, 2H. This figure
shows the information that can be determined from a knowledge of the
radius of gyration and the volume of a particle.

2R = 57 A, 2H = 34 A. Actually these dimensions will not be sharply

defined but will rather be described as zones of possibilities with widths
dependent upon the errors in the molecular weight and the radius of
gyration.
Similarly, if we assume that the molecule has the form of an ellipsoid of
revolution, two relations can be defined between the half axes, Rand H.
The curves of these relations will again lead to two zones of possibilities
centered on the intersections of the curves (see Fig. 42). This graphical
method of solution is of interest because by its use we can see directly the
consequences of a modification of the data on the form and dimensions of
the particle.
We have just seen that it is advantageous to know the volume of the
particle in order to obtain more detailed information from its radius of
INTERPRETATION OF RESULTS 133

gyration. It is possible to determine this volume ifthe intensity measure-

ments are made in absolute units. We know that the zero-angle
scattered intensity, J(O), obtained by an extrapolation of the curve of
log J(h 2 ), is given as
1(0) = J 0 (0)Nn2 (10)
This expression signifies that /(0), expressed in electron units, is equal to
the product of the average number of particles N and the square of the
number of electrons in each particle, n 2 • The total number of electrons
in the scattering body, Nn, can be determined from a knowledge of the
mass of the sample that is active in scattering and the chemical composition
of the scatterer. Therefore the measurement of J(O) in electron units
enables us to determine the number of electrons in each particle, n
= Nn 2/Nn, and from this we can deduce not only the mass of ea.ch
particle but also its volume, if its density is known. The application of
this method requires several precautions:
(a) We have assumed that there is negligible absorption of the X-rays
in the sample, and, if this condition is not satisfied, the necessary cor-
rection must be made.
(b) In the case of solutions of large particles, the number of electrons
per particle, n, to be considered is, to a. first approximation, the difference
between the number of electrons effectively contained in a. particle and
the number of electrons contained in the same volume of the solvent.
Results obtained with the methods described in this section (§4.1.1.1)
will be correct only if the hypotheses we have imposed are satisfied.
The relative interparticle distances can in general be checked easily if the
density of the sample and the density of matter in the particle are known.
It is more difficult to be sure of the uniformity of the particles. A
definite criterion cannot be found in the X-ray data alone. Nevertheless
it can be said that, if the curve of log J(h 2 ) maintains its linearity over a.
large pa.rt of the small-angle region, it is probable that there is not a very
great variation in the sizes of the particles. More precisely, such a.
linearity means that the sample does not contain a notable proportion of
particles having a radius of gyration ktrger than that determined from the
linear part of the curve, for the presence of such particles produces a
curvature in the curve at very small angles; on the other hand, very small
particles can be present. A bending of the curve of log J(h 2 ) with
increasing angle cannot be interpreted unambiguously, for this may be
due either to the particular shape of supposedly identical particles or
to the presence of smaller particles mixed in with those larger ones which
primarily determine the initial linear part of the curve (see, for example,
p. 151).
134 SMALL-ANGLE SCATTERING OF X-RAYS

4.1.1.2. Identical Particles with a Definite Orientation

Any particle of this system can be derived from any other one by
means of a translation. The characteristic of the resulting pattern is
that there is no longer circular symmetry around the incident beam. In
the general case it is therefore necessary to use a beam of very small
height, since the methods of correction for beam height effects described in
§3.4.2 are valid only if the intensity of scattered radiation depends solely
on the scattering angle, 20.
Let us assume, then, that the beam cross section is as nearly like a point
as possible. The curves of scattered intensity as a function of angle are
determined for different azimuths about s0 (Fig. 3, p. 13). From §2.1.5
(equation 2.44), the approximate exponential function describing the
intensity scattered for an orientation defined by the vector 10 is
J(h) = n2Ne-1i'D'<lo> (11)

Measurements for each orientation of 10 then lead to the determination

of an average inertial distance of the particle with respect to the plane
11(10 ) that is perpendicular to 10 and passes through the center of gravity
of the particle. Let us point out that this equation differs from the
approximate law of Guinier in that there is no coefficient, 1/3, in the
exponent.
The determination of inertial distances for different orientations gives
a rather precise idea of the form of the particle. This is the principal
advantage of the study of oriented particles as compared to the study of
particles without definite orientation.
It is obviously advantageous to direct the incident beam along one of
the principal axes of the particle. Let us take as an example an ellipsoid
of revolution of axes a, a, and va. We shall direct the incident X-ray
beam normal to the unequal axis. The small-angle scattering spot will
then have the form of an ellipse. The variation of intensity with angle
is (equation 2.44) e-aY.11•15 in the direction parallel to the unequal axis,
and e-a•h•/5 in the perpendicular direction. Thus two experimental
measurements have determined the parameters a and v which define the
ellipsoid.
It can be said quite generally, by means of equation 2.44, that the
small-angle scattering spot will be elongated in the direction of the
minimum dimension of the particle. An ellipsoid or an elongated rod
gives a spot elongated in the direction perpendicular to the long axis. A
plate-like particle gives a spot elongated in the direction perpendicular to
the plane of the platelet. Thus the general aspect alone of the pattern
furnishes qualitative information about the general f~rm and orientations
of the particles.
INTERPRETATION OF RESULTS 135

If the particles are so elongated that one of their dimensions exceeds

the upper limit that can be measured by the equipment, the scattering
spot appears as a narrow streak whose width is a function only of the
dimensions of the incident beam, and no quantitative information about
this long particle dimension can be obtained from the pattern. In this
case, in studying the distribution of intensity along the streak an incident
beam of large height may equally well be used.

4.1.2. DENSE GROUPS OF IDENTICAL PARTICLES

4.1.2.1. Analysis of the Scattering Curve
It has been recognized since the first studies of small-angle scattering
that quite often the particles cannot be considered as being widely
separated. This is true when the graph of the intensity has such a form
that it is evident that the exponential law is no longer a valid approxi-
mation. If the interparticle interferences cannot be neglected, the
applicability of the results of the methods we have previously described
must certainly be doubted.
This is the reason why it is very important to study the influence of the
drawing together of particles on the scattering. In §2.2 we discussed the
solutions that had been given for this difficult problem, but no solution,
even approximate, has been advanced which has not required several
restrictive hypotheses. These hypotheses were that the particles are
identical and either that they are spherically symmetric or that their
orientations are in no way dependent on the separation of their centers.
Equation 2.68, the fundamental relation for the scattering by spherically
symmetrical particles (see §2.2.3.2), gives the observed intensity /(h) as

(2.68)

where v1 , the volume offered to each particle, is equal to V 0 /N 0 • The

constant € can be taken as unity (see p. 44). The interparticle inter-
ferences which modify the curve of F 2 (h) are taken into account in the
function /J(h). The first task in the interpretation of the experimental
curves is to obtain the function F 2 (h) from the measured function l(h),
which takes us back to the problem we have just discussed. If the
concentration of the particles is not too high, the function {J(h) only
slightly modifies the curve of F 2 (h). We can again trace the usual curve
of log /(h 2 ) and verify that it still has a linear portion at small angles.
By applying equation 6 to this linear part we can determine an apparent
radius of gyration, R 0 A; this is not the true radius of gyration, R 0 , which
would be found from the curve of log F 2(h) as a function of h 2• A simple
136 SMALL-ANGLE SCATTERING OF X-RAYS

mathematical manipulation of equation 2.68 leads to the following

relation (Fournet [48]):
R 2 - R 2 - ~ {J"(O) (12)
OA - o 2v1(211)-3/2_{J(0)

R 0 A obviously approaches R 0 as the average volume v1 increases, that is,

as the concentration decreases. Furthermore, other experimental results
(the decrease of the intensity J(O) with increasing concentration),

log I

h2
Fig. 43. Log I vs. hi for spherical particles packed together
in increasing concentrations.

together with certain theoretical considerations (the theory of lyophobic

colloids given by Verwey and Overbeck (1945)), have shown that, if
equations 2.67 and 2.68, deduced from the theory of Born and Green, are
to be used, it is necessary to adopt a function, {J(h), which is negative for
I I
small values of h. Now, when {J(h) is a maximum at h = 0, fJ"(O) is of
the opposite sign to that of fJ(O), so that in our case {J"(O) must be positive.
Therefore R 0 A is less than R 0 and becomes even smaller, the greater the
concentration of particles.
As an illustration of this, let us consider the model studied earlier
INTERPRETA1'ION OF RESULTS 137

(§2.2.3.4, equation 2.75), which is composed of a group of hard spheres

showing no interactions other than impenetrability. We have reproduced
in Fig. 43 the curves oflog J(h 2 ) for three values of the ratio (8v 0 t:/v1 ) : 0.00,
0.25, and 0.50. These curves verify not only that the slope at the origin
varies with concentration in the manner we have indicated but also that
the curves depart more and more from linearity as the concentration
increases.
log I

Fig. 44. Log I vs. h2 for a 9.8 per cent solution of horse hemoglobin.

This is just the aspect Fournet has pointed out on some real experi-
mental curves. The curve of log /(h 2 ) (Fig. 44) is a straight line for very
small concentrations (5 and 10 per cent), whereas for higher concentrations
the curves, after an initial straight-line portion at small values of h, drop
below the extrapolated straight line.
Summarizing the foregoing discussion, the method to employ to be
sure of avoiding all error is the following: After having determined a
radius of gyration from particles in a solution, repeat the experiment
with a more dilute solution and compare the two results. If, within the
limits of experimental error, the results are the same, the measurement is
correct. If they are different, then by making several measurements on
samples of decreasing concentration, the radius of gyration can be deter.
mined by an extrapolation to zero concentration.
If the concentration is high, the curve of F 2(h) is so deformed that it is
138 SMALL-ANGLE SCATTERING OF X-RAYS

no longer useful to try to employ the log J(h 2 ) representation. Let us

now show how it is possible to obtain the curve corresponding to an
infinitely dilute solution, F 2 (h), by means of the curves of I(h) for several
different concentrations, even though these concentrations are all rather
high. In addition, the function {J(h) will also be determined.
Let x be the weight percentage of particles in the solution; d(x), the
density of the solution at x per cent concentration; and m, the mass of one
particle. The volume per particle, v1(x), is then

1 lOOm
v1(x)= =--
X d(x) xd(x) (13)
100 m

Equation 2.68 shows that there will be certain points, defined by {J(h) = 0,
which will be common to all curves of the function

(14)

It is difficult in general to determine accurately the absolute value of

I(h), so that the functions f(x, h) will probably be known only to within
some multiplicative constant. In such a circumstance we can select
a priO'l'i some point to be the common point and then try to verify whether
the normalized curves reproduce the different characteristics of the
function
(15)

Let us give as an example the scattering curves for human hemoglobin

(Fournet [48]) (Fig. 45) which have been normalized to a common point
at a scattering angle of 2.4 X 10- 2 radian. The behavior of the functions
is in good agreement with the predictions that can be made from equation
2.68: as the concentration increas{ls, the intensity curves become more
uneven; at a certain concentration (in this case, 32 per cent) a maximum
appears, and this maximum becomes more accentuated and shifts to
larger angles as the concentration further increases.
Let us emphasize that, if the experiments are not capable of accurately
evaluating the ratio of the intensities scattered by solutions of different
concentrations, it is still possible in general to determine whether the
experimental facts can be interpreted by means of equation 2.68. A
common point, chosen a priori, is imposed on the several curves. An
examination of these normalized curves then shows whether or not the
curve corresponding to a concentration of x per cent is always contained
INTERPRETATION OF RESULTS 139

between the curves for x + dx per cent and x - dx per <.>ent as is pre-
dicted by the theory. By successive approximations we can see whether
it is possible to find a common point, defined by an angle k,_, for which
these conditions will be satisfied by all the experimental curves.
I

x=10%

1• Angle
2)(10- 2 3 >< 10- 2 radian
Fig. 45. Experimental scattering curves of Bolutions of human
hemoglobin of different concentrations.
When this common point can be determined, each normalized curve
represents a curve of the function/(x, h), apart from an arbitrary constant
the same for each curve; this function is
F 2(h)
f(x, h) = 1 - (211)312 £,8(h)[v1 (x)]-l (l5a)

Then by use of equation 13, which describes the dependence of the mean
volume, v1 (x), on the concentration, x, we find
1 1 - (211) 312 £{J(h)x d(x)J0-2m-1 1 - kx d(x),B(h)
(16)
f(x, h) F2(h) F 2 (h)
For each value of h, the curve of/-1 (x, h) as a function of x starts from
a value of l/F 2 (h) at x = 0 and increases in an approximately linear
manner. The slope of this curve as a function of x is
-k/3(h)F- 2 (h){d(x) + xd'(x)} (17)
and, since d(x) is in general a function which increases with x, the slope
then increases in absolute value as x increases.
140 SMALL-ANGLE SCATTERING OF X-RAYS

If the curves of this function are taken from the curves of Fig. 45, in
which h 1 ~2.4 X 10- 2 radian, the behavior wehavepredicted will be found.
If the common point of the curves represented in Fig. 45 had been taken
as an angle smaller than 2.4 X 10- 2 radian (2.2 x 10- 2 radian, for
example), the general appearance of the curves of f(x, h) would still be
satisfying, though less so, but the same cannot be said for the curves of
J- 1 (x, h). For a value of h equal to 1.5 X 10- 2 , the extrapolation of the
curve of J-1(x, h) gives a negative value for this function at the origin,
x = 0, which is physically impossible. This example shows that the
general appearance of the family of curves ofj(x, h) as a function of his
less sensitive than that of the curves of J-1 (x, h) as a function of x in
determining whether the common point has been well chosen.
The representation of J-1 (x, h) as a function of xis of further interest,
in that the Junction F 2 (h) is determined by the reciprocal of the ordinate at
x = 0. We must admit that experimentally the plots of these curves
will not be very meaningful unless the experimental results are known to a
high degree of precision. This requires that a very careful study be made
of a large number of solutions. Naturally, accurate measurements, such
as are possible with Geiger-Miiller counter detectors, are indispensable.
The function {J(h) can also be determined from the same curves by use of
equation 16. Now {J(h) is related to <l>(r) by the equation

h{J(h)
1
=. 1 _
Jco r{e
- kT
<l>(r)
- I} sin hr dr
v 27T - co

Therefore theoretically we can obtain the potential energy, <l>(r), of the

interaction between particles. Small-angle scattering experiments are thus
capable of giving not only the size of the particles but also the law of their
interaction. This method, which is complex, to be sure, but legitimate,
as is shown by the calculations of Chapter 2, has not yet been exploited as
thoroughly as it should have been.
So far we have considered only spherically symmetrical particles. It
is not impossible to treat the case of particles of arbitrary shape if we
assume that their relative orientations are independent of their spatial
distribution. The formulas are very complex, and the indispensable
hypothesis is quite hard to justify in numerous cases. In addition, there
are only a very few applications of such a treatment. Therefore it will not
be developed here.

4.1.2.2. Interpretation of a Maximum in a Scattering Curve

Disregarding these complex but accurate methods, many authors have
been content to use a more simple interpretation of the results of experi-
ments made on dense systems of particles, taking account only of the
INTERPRETATION OF RESULTS 141

position of the maximum or bump on the scattering curve. This pro-

cedure should be closely examined, since evidently it is not established on
a rigorous basis. It is only a rough approximation from which erroneous
conclusions can be drawn if the procedure is given more worth than it
actually deserves.
In ·order to simplify th~ notation in this section we shall consider only
spherically symmetrical particles, but it is easy to extend the discussion
to all cases for which there is complete independence between the
orientations of the arbitrarily shaped particles and the distances between
their centers. For spherically symmetrical particles F2(h) and F(h) 2 are
identical, and the scattered intensity can be written in the form

I(h) = I.(h)N F 2 (h)a(h) (18)

The product Na(h) is equal to the intensity in electron units which would
be scattered by N electrons, each positioned at the center of one of the N
irradiated particles. The function a(h) depends markedly on the con-
centration of the particles, so we shall denote the function by a(h, c) or
a(h, v1 ). For example, the function a(h, v1 ) for the system described by
equation 2.68 is
a(h V) - Vi (19)
' I - ,v1 - (21T) 312 e{J(h)

The angular positions of the maxima of the function a(h) will be denoted
by h1 , h2, • • • h1 (or 01 , 02 , • • • 81), and those for the function J(h) by
hM, hN, ... (or ()M, ()N, .. ·).
We shall now give the different possible interpretations.

4.1.2.2.1. Interpretation in Terms of an Average Distance

The presence of a maximum or even a singularity such as a bump in the
scattering curve has often been interpreted as showing the existence of an
"average distance frequently realized between neighboring particles."
First let us observe that it is impossible to define the concept of neigh-
boring particles in a substance which has no long-range order, except in the
very particular case of a linear model. This concept of an "average
distance," which seems simple and natural and which has so often been
employed, is in reality never defined in a precise manner.
Let us now look at the arguments which have been advanced by those
who employ this idea (for example, Mattoon, Stearns, and Harkins [366]).
The formula that is most often cited is that due to Ehrenfest (1915):

(20)
142 SMALL-ANGLE SCATTERING OF X-RAYS

Actually this formula describes the scattering angle 20 111 , for the maximum
in the intensity scattered by a group of widely separated particles, when
each particle, having the form of a dumbbell composed of two points
separated by a distance d, takes all possible orientations with equal
probability. 1 The Ehrenfest rel,ation therefore cannot be used to interpret

0 5 10 hd

Fig. 46. A graph of~ (I + sinhd hd) cll 1 ( ~·) for varioug ratios of d /d.
0

The maximum and minimum of the curve, shown joined by the dotted
line, disappear for d 0 /d > 0.52.

the structure of any fluid. In order to give every possible opportunity to

those who would employ this challenged formula, let us 'iUake the
assumption-which cannot be even a rough first approximation-that
the intensity scattered by a fluid composed of spheres of diameter d0 is
proportional to the intensity scattered by a pair of these spheres separated
by the distance J, and then calculate this intensity. This is found to be

/(h) = constant ( I + sinhdhd) <I>2 (hd

</) (21)

In Fig. 46 we have given the normalized curves of this function for

1 The intensity scattered by such a system is proportional to ( 1 + si~hd) (Fig.

46), the position of the first maximum being given by equation 20.
INTERPRETATION OF RESULTS 143

different values of the ratio d 0 /d, d being kept constant. The curve for
d 0 = 0 corresponds to the case described by the Ehrenfest relation. As
the ratio d 0 /d increases, there is a limiting value beyond which the intensity
I(k) becomes a function that decreases continuously with increasing
angle. Now experiments have shown that the observation of a maximum
is more probable, the higher the concentration of matter, which for our
model means a greater ratio of d0 /d. But it is precisely for these con-
ditions that equation 21 that supposedly relates the position of a mazimum
to an average distance will not allow a maximum to appear. This
internal contradiction thus gives a second reason for abandoning this
usage of the Ehrenfest equation.
It is still possible to introduce an average distance into the inter-
pretation of the presence of a maximum by considering a model in which
the centers of the particles are arranged on a more or less distorted face.
centered cubic lattice. In such a face-centered cubic paracrystal, as it
has been called by Hosemann, each molecule is surrounded by 12 neighbors
which are clearly closer than all other molecules, but the distance from the
origin molecule to these 12 neighbors varies from one to the next,
fluctuating about a mean distance d. In this model it is possible to
define a mean distance of approach which naturally will be larger than the
diameter of a molecule. The cell edge of a cubic pa.racrystal is equal to
dV2'. The scattering angle 281 'for the first maximum of the function
a(k), that due to (111) planes, is then determined by Bragg's law to be

2dv2
;. = 2<411 sin 81 = v3 sin 81
or

-
2d sin 81 = J32 ). = 1.22). (22a)

Hosemann has shown that the function a(k) shows maxima not only at
the angle 01 but also at angles 02 , 03 , etc., corresponding to planes in the
average lattice of higher indices. The angles for these maxima are
similarly defined:
2d sin 82 = v2;. = l.4U (22b)

2d sin 03 = 2). = 2.00.A. (22c)

etc. However, these maxima are broader and weaker, the higher the
corresponding index and the more distorted the latticet so that it is
poBBible that only the first two or three can be observed.
144 SMALL-ANGLE SCATTERING OF X-RAYS

This model imposes a certain relation between sin 81 and the volume
concentration, c (the ratio of the volume of the particles to the total
volume of the system). The maximum value of this volume concentration
for spherical particles is 0.74. We can thus write that
c/0.74 = (d 0 /d) 3 (23)

where d0 is the diameter of the sphere. This relation can be transformed to

the following:

Equation 22a then can take the form

1.22.I. 1.35.I.
- - - c1i3(0.74)- 113 = - - - c113 = d0 =constant (24)
2 sin 81 2 sin 81

We see thus that the sin 8; are proportional to c1 i 3 •

This paracrystalline model leads to the prediction of several marked
maxima in the function a(h). The value of sin 8; for each maximum
varies in proportion to the one-third power of the volume concentration.
Patterns that agree with all these results have been only rarely
observed (a good application of this model is that of Bernal and Fankuchen
[402]), but nevertheless a method of interpretation based on this model
has often been employed where the experiment did not show definitely
that the paracrystalline model was a good approximation. For example,
this method was applied by Riley and Oster [371] to the results for
solutions of human hemoglobin obtained by Riley and Herbert [486].
These scattering curves showed only one slightly marked maximum (two
maxima were observed at the highest concentration only). Since the
maxima of l(h) were very broad, a similar breadth must exist for the
maxima of the function a(h). We have already explained in §2.2.3.4
that in a circumstance like this one the intervention of the factor F 2 (h)
will make the positions 8M of the maxima of l(h) very different from the
positions 8; of the maxinia of a(h) (cf. Fig. 13). Therefore conclusions
that are valid for the latter maxima, as drawn from equations 22 and 24,
cannot be applied to the intensity maxima. The c113 dependence has
almost never been validly observed (Philippoff [369]). Riley and Oster
have observed that the position of the first maximum obeys a relation of
the form
sin OM= kMc113 (25)

If one is to satisfy both equation 25 and the relation

sin 81 = k1c113 (26)

INTERPRETATION OF RESULTS 145

very special conditions must be imposed on the functions Fll(h) and

a(h), so we can say that in general equation 26 cannot be satisfied if the
experimental results obey equation 25. An entirely different inter-
pretation of equation 26 will be given in §4.1.2.2.4.
Finally, let us point out that the proposed model requires that, when
the concentration changes, the same type of short-range order continues
to exist while the interparticle distance varies. This would probably
require that the interaction potential change with dilution, a mechanism
which does not seem very probable.
We have just discussed the attempts at interpretation of real solutions,
isotropic in three dimensions, in terms of an average distance. We could
give an analogous discussion for cylindrically symmetrical systems. The
values of sin (}i would then vary as c112 if a quasi-crystalline model were
adopted. This is the experimental relation that was observed by Bernal
and Fankuchen [402].

4.1.l.l.l. Interpretation in Terms of an Average Volume

In order to escape the criticisms based on the impossibility of defining
an average distance, we can try to interpret the presence of a maximum
in the scattering curve in terms of the existence of a certain average
volume v1 defined as the ratio of the total volume offered to the particles,
V 0 , to the number of particles, N 0 • Equations 2.67 and 2.68 show that
this is not possible; f3(h) plays just as essential a role as does v1 , for it is
the ratio of f3(h) to v1 which appears. For example, let us consider two
identical enclosures of volume V 0 , one containing N 0 argon atoms and
the other containing N 0 non-interacting hard spheres, whose volumes
can be chosen equal to those of the argon atoms. These two groups of
particles characterized by the same average volume v1 give very different
scattering patterns, as is shown by a comparison of their functions {J(h)
given in Fig. 19. By generalizing this remark we can say:
1. Although the average volume per particle may be the same in several
groups of particles, nevertheless the statistical arrangement of the par-
ticles may be quite different in the different groups, since this depends
on the function <l>(r).
2. The scattering of X-rays, through the intermediary of the function
P(r), is very sensitive to these differences.
An example of this can be found by a comparison of the behavior of a.
solution of tobacco virus, which demonstrates the presence of order for
very small values of the ratio v0 /v1 (Bernal and Fankuchen [402]), with
the behavior of an ordinary solution of a protein (hemoglobin, for example).
Thus the interpretations that depend on the intervention of the volume
v1 cannot be justified.
146 SMALL-ANGLE SCATTERING OF X-RAYS

4.1.2.2.3. Interpretation by Means of an Interparticle Interference

Function
Oster and Riley [128] have proposed to get information from the
scattering curves of solutions of proteins by means of a function, a(h),
obtained from the experimental study of liquid mercury. Equations
2.67 and 2.68 show that a(h) depends only on the ratio {J(h)/v 1 , where the
function {J(h) is determined by means of a Fourier transformation of a
term involving <l>(r) (see equation 2.65). Each type of particle has its
own potential <l>(r) and, consequently, an individual function {J(h). Thus
a method of interpretation employing a given a(h)-type function cannot
be generally correct. This method could be used nevertheless as a first
approximation when the particles being examined are somewhat analogous
to the particles from which the a(h) function is determined. However,
it is not at all obvious that the data from mercury are valid for proteins.
A further criticism of this particular comparison follows from the use of
liquid mercury, since it is known that certain properties of solutions of
large molecules (for example, osmotic pressure) can be likened to the
properties of gases but not to those of liquids.

4.1.2.2.4. Predictions of the Correct Theory

We would like now to summarize the facts that can be predicted about
the angular position of a possible intensity maximum from the two types
of equations we have discussed in §2.2.3.
The "geometrical" equation for spherically symmetrical particles,

l(h) = - { li"'
I,(h)N F 2(h) 1 - -
Vl 0
[l - sinhr
P(r)] - - 47rr2 dr}
hr

shows that the position of a possible maximum of l(h) depends primarily

on two factors:
1. The function P(r). The value of the integral for a certain value,
h 0 , of the parameter h depends on all the values of the function P(r)
and not simply on the value P(r 0 ) corresponding to a certain distance,
r 0 • Thus there cannot be a law relating a value r 1 , corresponding to the
first maximum of the function P(r), to the value h 1 at which the integral
presents its minimum value. The function P(r) must be known over its
entire range in order to calculate hl'
2. The structure factor F(h) and thus the internal structure of the
particle. This factor, though often neglected, can be important. In
most cases F 2 (h) decreases with increasing angle, so that the maximum
intensity occurs at a value of h smaller than h1 . Moreover, all other
characteristics being equal, the maxima will appear at smaller angles and
INTERPRETATION OF RESULTS 147

to a lesser degree (even disappearing), the less compact the distribution

of scattering centers in the interior of a particle.
We see therefore that in the general case it is difficult to define the relation
between the existence of an intensity maximum and the structure of the group
of particles being examined, so that it is especially necessary to abandon the
idea that the intensity of radiation scattered through a certain angle contaiM
specific information about a certain interparticle distance (as, for example,
the probability of the existence of this distance).
The "thermodynamical" relation for spherically symmetrical particles
(Fournet [48])
- 1
/(h) = I.(h)N F2(h) (27r)S/2E{J(h)
1-----
"1
shows that the maximum of a(h, v1 ) is always produced at the same
angle h1 , regardless of the concentration of particles in the sample. The
influence of the concentration is felt uniquely in the steepness of the
variations of a(h) as a function of h (Fig. 13). The intensity, /(h),
determined by the product of a(h) and F 2 (h), begins to show a maximum
when the concentration reaches a rather high value. This maximum is
situated at an angle smaller than h1. As the concentration increases,
the curve of a(h) becomes sharper, and the position of the maximum
moves towards hi· The maximum in l(h) shifts towards larger anglea as
the concentration increases. This behavior is illustrated by the curves of
Fig. 13. These were constructed with a function F 2(h) which followed
the law of Guinier and a function {J(h) chosen so that the observed maxi-
mum obeyed the law found by Riley and Oster (equation 25). The
behavior of this {J(h) is not at all improbable.
Equation 2.67 is an approximate expression, but nevertheless we can
still state rigorously that the displacement of /(h) is explained in large
part by the mechanism we have indicated in §2.2.3.4 and is not due to a
variation in the angular position of the maximum of a(h). Equation
2.67 also shows that the primary factor which determines a(h) is the
function {J(h). We have used this fact in our discussion in §4.1.2.2.2 and
"1
§4.1.2.2.3 (Fournet (1955)).

4.1.2.3. Conclusions
We have just demonstrated the complicated factors regulating the
positions of the maxima of the scattered intensity. Let us now summarize
the information of this section by considering the two following cases:
1. The intensity maximum is not very sharp. This means that the
148 SMALL-ANGLE SCATTERING OF X-RAYS

corresponding maximum of the function a(h) is also not very sharp, and
consequently the constitution of the matter is quite different from that
of a crystal. The theoretical calculation of the position h1 of the maximum
of a(h) is then often difficult. The structure factor of the particle, F(h),
will contribute strongly to the form of the curve of the intensity, I(h).
The maximum of I (h) will be found at a position, h M• which is often quite
different from h1 . No simple significance is attached to this one para-
meter, hM. Two procedures can be followed:
If precise information is wanted, a complete study must be made of
the family of curves of I(h) as a function of concentration and then an
attempt must be made to compare this group of curves with theoretical
results obtained from some suitable model.
If only approximate information is wanted, we can simply apply Bragg's
law to the diffuse peak and determine an order of magnitude (to within 30
to 50 per cent) of some geometrical dimension characteristic of the matter
being studied. It is impossible to determine the meaning of this geo-
metrical dimension by this one measurement (an example is shown in
§6.4.3.l for Al-Ag alloys).
2. The observed intensity maximum is sharp. Although it is not
rigorous, we are probably justified in considering the sample being studied
as a paracrystalline substance, and there will probably be a simple
interpretation of the maxima of a(h). Since these maxima are very sharp,
the factor F 2 (h) will have very little influence on the positions of the
intensity maxima (h M is almost exactly h1 ) and will affect only the
magnitude of the intensity (see Fig. 14).
This section has been deliberately developed at length in order to show
clearly how easy it is to make rather serious errors in interpretation.
The introduction of X-ray techniques into the studies of large molecules
has not always succeeded in clarifying the ideas. This is due to the fact
that the non-specialists in the scattering of X-rays have often been
deceived by the dangerously precise results that have been published
(certain authors have not hesitated to correct Bragg's law with a co-
efficient with three significant figures) and have thus by inference been
able to obtain information (forms of particles, number of layers of water
with which each is surrounded, etc.) which really has no serious justifica-
tion and which can be completely erroneous. It should always be
remembered that Bragg's law, so familiar in crystallography, applies only
to good crystals.

4.2. GROUPS OF NON-IDENTICAL PARTICLES

We shall now consider the case where it is not possible to assume that
all the particles are identical. The theoretical calculations of §2.3 show
INTERPRETATION OF RESULTS H9

that there is only one case for which a useful result can be obtained-
particles separated far enough so that interparticle interferences a.re
negligible. Let us state immediately that we cannot give a rigorous
method of interpretation when the particles are of several types and are
closely packed. Section 6.3 gives the uses that can be made of the
experimental results by adoptihg an empirical point of view.

4.2.1. DETERMINATION OF THE AVERAGE RADIUS OF GYRATION FOR

THE GROUP OF PARTICLES
The curve of log /(h 2) is generally concave, with a positive second
derivative. If it is found that at small angles the curvature diminishes
enough so that the tangent at the origin can be determined with reasonable
accuracy, by applying equation 6 to the slope of this tangent a mean
radius of gyration R 0 M can be obtained. This parameter represents a.
mean value defined by the following relation:
LP.tR2okFA:2(0) LP.tn/·R2ok
RI - I: I: (27)
OM - LP.tFA:2(0) 2.P.tnk2
k k

where nk represents the number of electrons contained in the particle

of type k and Pk represents the proportion of this type.
Equation 27 can be simplified if it can be assumed that all the particles
are geometrically similar, the number of electrons n11: being proportional
to R 3w With this assumption we find
LJikR8 ok
R2oM = ;pkRSok (28)
k

It is easy to generalize these formulas for the case of a continuous

distribution of particle sizes by first introducing a probability density
function, p(R0 ), where p(R0 ) dR 0 designates the probability that a
particle has a radius of gyration contained between R 0 and R 0 + dR 0 ,
and then replacing the summation by an integration.
We can empley equally well a mass distribution function, m(R0 ),
where m(R0 ) dR 0 gives the total mass of particles whose radius of gyration
is contained between R 0 and R 0 + dR 0 • With this approach we find
R2 - f m(Ro)Ro5 dRo (29)
oM - fm(Ro)Ro3 dRo
These equations show that l.arge radii of gyration are favored in the
e:eperimentally determined average. This fact can be illustrated rather
easily by considering the very simple model composed of two different
150 SMALL-ANGLE SCATTERING OF X-RAYS

l(h) l(h) l(h)

0 2 3 4 hRo 0 2 3 4 hRo 0 2 3 4 hRo

(a) (b) (c)

Fig. 47. Scattered intensity for dilute systems of particles:

(a} Identical particles of radius of gyration R 01 ; (b) identical particles
of radius of gyration R 01 = 1/ 1R 01 ; (c) a mixture of equal proportions
of the particles of a and b.

log l(h)

Slope corresponding to R 01

Slope corresponding to Ro2

0 IO

Fig. 48. Log I vs. h 2 R 01 for a mixture of particles of radii of gyration

R 01 and R 01 , where R 01 = 1 / 1 RD1' Same sample as for Fig. 47c.
INTERPRETATION OF RESULTS 151

types of particles (1 and 2) with R 01 greater than R 02 • Figure 47 shows

three scattering curves: curve a is that for particles 1 only, curve b is
that for particles 2 only, and curve c is that for a mixture of 50 per cent
of particles 1 and 50 per cent of particles 2 (by weight). At very small
angles, the variation of the intensity for the mixture is almost entirely
due to curve a, since curve b shows only a very slight variation in this
region; thus large radii are favored in the experimentally measured
radius of gyration. The influence of the small particles is only perceptible
at larger angles, where the intensity scattered by the large particles is.
practically reduced to zero.
This example will also permit us to demonstrate another important
aspect characteristic of mixtures. Let us assume that the structure
factors of the different particles follow the law of Guinier exactly. At
very small angles we have seen that the apparent radius of gyration is
effectively R 01 ; at larger angles (h1 ) the curve c is due almost entirely to
curve b, since the contribution from curve a is effectively zero, and the
apparent radius of gyration, as defined by the slope of log Ic(h 2 ) at h1 2,
is practically equal to R 02 • The curve of log Jc(h2 ) is thus characterized
by a decided concavity, as is shown in Fig. 48. We may generalize this
remark to say that the curve oflog J(h 2) for a mixture will usually present
such a concave appearance. Let us make quite clear that this is only a
general observation and is not a proof of a one-to-one correspondence;
for example, the curve of log I(h 2 ) for a group of identical homogeneous
rods demonstrates the same concavity.

4.2.2. ATIEMPTS AT DETERMINING THE STATISTICAL DISTRIBUTION

OF THE PARTICLES
It is easy to calculate the intensity curve by means of equation 2.93,
J(h) = I.(h)NJ..pkFk 2 (h)
k
if the distribution of particles is given, that is, if the form, the size, and the
relative fraction Pk of the particles of each type are given.
Conversely, several authors (Hosemann [79], [81]; Shull and Roess
[155]) have tried to deduce the distribution of the particles from the
experimental curve. But this problem does not have a unique solution.
Hypotheses must be made, and the results depend on these hypotheses.
Several of the methods that have been proposed will be discussed in this
section.
Hypotheses are not made on the form of the particles, but it is assumed
that they are all geometrically similar, so that the number of electrons nk
in a particle of type k, defined by a radius of gyration R 0k, is proportional
to Raok·
152 SMALL-ANGLE SCATTERING OF X-RAYS

It is assumed that the particles of type k give a scattering curve

represented by the exponential law, e-<h'R'o.>1 3 • We have already
discussed the validity of this (p. 26). This approximation must be valid
for the largest particles of the group for all values of the parameter h at
which the scattered intensity is important; if small particles are present
this scattering will extend to large angles, and in this case the complete
calculations are correct only if the shapes of the particles are such that the
exponential relation is a very good approximation up to rather large
values of hR 0 •
Next it is assumed that the distribution of the radii of gyration of the
particles is represented by a Maxwellian distribution of the form (Whit-
taker anil Watson (1927))

(30)

where m(R 0 )dR0 represents the proportion of the particles whose radius
of gyration is between R 0 and R 0 + dR 0 • By means of the two para-
meters, r 0 and n, we can satisfactorily describe a wide range of distributions.
The arithmetic mean of the radii of gyration is given by

r (~ + 1)
RoJtI = ro (n 1)
r 2- + 2-

and the fractional standard deviation of this value, V t1R 0 2/R 0, is

l/V2(n + lJ.
The scattered intensity is given by equation 2.93 as proportional to
the integral
I m(R )
- a Ok
R
-
Ok
2
nk e
_ h2 R'ot
-3- dR
Ok

As a result of the form chosen for m(R 0 k) the integration can be carried
out, giving
constant
l(h) = __[_1_+_(h-~0-)--2]_,..(n_,+-..,.4)"""/2 (31)

The problem is now to determine r 0 and n in such a manner that this

equation will suitably represent the experimental curve.
INTERPRETATION OF RESULTS 153

Several methods of calculation have been given. The method due to

Hosemann is based on the use of a curve of h 2I(h) as a function of h. This
curve always presents a maximum at some position hM greater than zero.
A second position, hT, is defined by the intersection with the h-axis of the
tangent to the curve at the inflection point Won the high-angle side of the
curve (Fig. 49).
h 2 l(h)

h
Fig. 49. The Hosemann method for analyzing the intensity
distribution from a mixture of particles rs41.

Hosemann derived the following relation:

hp 1
2-h-M - 1 ~ V-;=2(=n=+=l)

From this equation the parameter n can be determined. Then the

arithmetic mean of the radii of gyration can be found by the relation

+ 2)
RoM = h M
1 J + 2r (
-
n
6
r( n-
2
n ; l)
The mathematical proof of these results, as drawn from equations 30 and
31, will be found in an article by Hosemann [84].1
In the method of Shull and Roess, equation 31 is first "\\Titten in the
form
n+4
log I = constant - -
2
- log ( h 2 + r2 3)
0
(32)

1 Hosemann considers systems of spheres instead of general particles defined by

their radius of gyration. His formulas have been changed accordingly.
154 SMALL-ANGLE SCATTERING OF X-RAYS

Then the experimental intensity curve is constructed with the coordinate

system log I, log (h 2 + oc), and a value of°' is sought for which the curve
of log J(h 2 + oc) is a straight line. Then the parameters, r 0 and n, are
determined from this constant (r 0 2 = 3/°') and from the slope of the line
(slope = (n + 4)/2).
One can also construct the family of curves of log I as a function of
log (hr 0 ) 2 for different values of n and then try to superpose the curve of
log J(h 2 ) on one of these. When this is done, the translation of the origin
along the h-axis gives r0 •
It can be stated that, although it may be easy to determine the average
dimension of the particles, it is difficult to select the form of the distri-
bution law, because the difference between theoretical curves for groups
with the same average dimensions but with different distribution laws are
very small (see Fig. 50).
Roess and Shull [148] have made calculations for spherical particles
and particles in the form of ellipsoids of revolution with a variable axial
ratio, v, in which they employed the true structure factor instead of the
exponential law. They obtained exact results by expressing them in
terms of generalized hypergeometric functions. 'They were thus able to
obtain a family of curves (Fig. 50) plotted with the coordinates log I and
log a 2h 2 , in which each curve was defined by two parameters, n and v.
The parameter v described the form of the particles; n, the statistical
distribution; and a, the dimensions of the average particle. The
important conclusions of this distinguished work are that there is little
variation in the form of the curves and also that the same forms are found
for different pairs of the parameters n and v. If then we consider the
experimental uncertainty of measurements, particularly at high angles, it
appears that it is difficult to choose from the family of curves that par-
ticular one which gives the best coincidence with the experimental curve.
Therefore the parameter n and, consequently, the statistics of the distri-
bution are poorly determined. Again, very different values can be found,
depending on the choice of v, that is, the form of the particle. This
demonstrates the important limitations of low-angle scattering methods
when the particles are not homogeneous.
Let us now briefly mention other contributions to this subject. Shull
and Roess have repeated their calculations using a "rectangular"
distribution (uniform distribution between a minimum and a maximum
size). It has also been shown (Riseman [145]) that, for a system of
spherical particles, it is mathematically possible to determine the
statistical distribution of sizes from the experimental curve without
imposing a priori some form on this intensity curve. However, there
seems to be little physical interest in such calculations.
INTERPRETATION OF RESULTS 155

1.0

0.6
0.4

0.2
I

0.1

0.06
0.04

0.02

0.01
0.1 0.2 0.4 0.6 1.0 2 4 6 10 40 60 100
Ro2h2
(a)

1.0
0.7

0.4

I
0.11-----+--t---t---t-----t---t---";c't"~:-'l<~'t---t---r-----1

O.o? l----+--t---t---+----c---t--~-"<i'.--"<c-'<-t--'~-t---r-----1

(b)

Fig. 50. Calculated curves of I vs. (hR 0 ) 1 • (a} Maxwellian distri-

butions of spheres characterized by the pa.re.meter n; (b) Maxwellian
distributions of oblate ellipsoids (axial ratio: v = 0.25). (Roess
and Shull [148].)
156 SMALL-ANGLE SCATTERING OF X-RAYS

4.2.3. DETERMINATION OF THE SPECIFIC SURFACE

In the preceding sections we have tried to use the low-angle part of the
scattering curve. But in systems composed of particles of various sizes,
that low-angle part which can be represented by an exponential approxi-
mation is often inaccessible to experiment. Also, if the particles are
packed together the low-angle part is perturbed by interference effects.
It is therefore more advantageous under these conditions to exploit the
other possible approximation, that for the tail of the scattering curves. It
would undoubtedly be possible to determine the characteristic function
y{r) {§2.4.3) from the experimental curve without making any hypotheses,
for, mathematicaJly, ry{r) is the Fourier transform of hl(h). But
in practice the function y{r) can be used to obtain one important
quantity, the value of the specific surface of the specimen, and this is deter-
mined by the tail of the scattering curve.
The fundamental equation, equation 2.108, is valid both for systems
composed of arbitrary particles distributed in an arbitrary manner
and for an arbitrary distribution of matter of constant density p.
This. . h h I. l aJ, h4/(h)
equation s ows t at, Jor arge v ues of h, the quantity 217 - -p-21-.-(h-)
approaches the value of S, the total, free surface of the scattering body. If
the scattering body has unit volume or unit mass, then S represents the
specific surface per cubic centimeter or per gram of the sample. The
method requires that the measurements of the scattered intensity be
extended up to rather large angles, where the intensities are rather weak.
That the product h 4l(h) becomes constant must be verified. Finally, the
intensity of the direct beam should be measured in order to obtain an
absolute value for S. However, if samples of the same matter but of
different structures are compared, then, by examining samples all of which
have the same mass, the ratio of the products h 4I(h) gives the ratio of the
specific surfaces of the specimens.
The assumption is made in these equations that the electronic density
in the interior of the particles is rigorously constant, and this, by virtue of
the atomic structure of matter, is never the case. This atomic structure
gives rise to a high-angle scattering pattern such as that for an amorphous
body, for example. The intensity to be expected from this effect in the
low-angle scattering region is weak but not zero. It is negligible in the
small-angle part of the region of scattering arising from the heterogeneity
of the sample, but it can become perceptible towards the outer part of the
curve, the part in which we are interested at this point. Other sources of
extraneous radiation must also be added: fluorescence radiation, Compton
scattering, thermal scattering by crysta.ls, various parasitic scatterings,
INTERPRETATION OF RESULTS 157

etc. The product h 4/(h) thus will begin to increase when h becomes too
large. The asymptotic law is also not valid for too small values of h.
There is therefore only a certain range of validity, whose extent depends
on the particular sample, in which this relation can be used in making
measurements of specific surfaces. These measurements lmve significance
only if it is verified tlmt the product h 4I(h) remains constant over a reasonable
interval.
If the cross section of the incident beam is not pointlike but rather has a
large height, then it is the product h3J1(h) which should be constant at
large values of h (equation 3.llb).
We can avoid the measurement of the intensity of the direct beam in
determining the absolute value of the specific surface by making use of the
normalization relation, equation 2.111. If Vis the total volume of the
specimen and c is the fraction of this volume occupied by matter (of
electronic density p),

L'' h 2J(h) dh = 211 21.(h)pSVc(I - c) (33)

Ifs.pis the specific surface per unit volume, then 8 = vs,p, and

S _ lim [h41(k)] _ lim [h'l(k)]

•P - • 2 - 7TC(l - c) fm
V211p l.(k) Jo h•I(k) dk (34)

When the scattering experiment is made with an incident beam of large

height andJ1(h) is the measured intensity, equations 3.llb and 3.llc give

lim [h3J1(k)] = i lim [h I(h)]

4

and

i0
m h2J(k) dh = -1
2
i"°
0
h.F1(h) dk
Therefore
s.p = 4c(l - c) lim [h3f1(h)]
f"' k.F1(h) dh <35>

The difficulty in evaluating the integrals Lm h 1(h) dk or L"" hJ1{h) dh

2

is that J(h) or J 1(h) is known only beyond a certain minimum value of h.

The contribution of the unknown small-angle part of the curve is certainly
relatively small, particularly in the integral L""h 2J(h) dk, since here the
158 SMALL-ANGLE SCATTERING OF X-RAYS

measured intensity is multiplied by h 2 • Nevertheless, there is still some

uncertainty which causes this method to be less precise than the method
employing the intensity of the direct beam.
In interpreting certain results, Porod introduced the length

l 0 =4Vc/S

which he designated as a "range of inhomogeneity." This length repre-

sents an average1 of the diameters of the parts of the specimen occupied
by matter; for instance, for a collection of spheres of radius R, 10 = !R.
Another parameter derived from the characteristic function of the
specimen which can be rather easily determined experimentally is the
distance of heterogeneity or extent of coherence, l 0 (Porod [137]), given by
the integral 2 f "'y(r) dr. Equations 2.109 and 2.110 determine l0 respec-
Jo f"'
tively in terms of the integral Jo hl(h) dh and the total scattered energy E,
the quantity obtained by measuring all the small-angle scattering outside
of the direct beam in a counter or ionization chamber. These relations
are the following:

l=
• 277 V p 2c(I1- c)l8 (h) o
hlhdh
( ) i"" (36)

and

l = 1 ~ (37)
• V p2c(l - c)I.(h) A_2p2

where p is the distance from the sample to the receiver.

By making use of the normalization relation, equation 33, these
become:
i"'hl(h) dh
l =71--'=-0- - - - (38)
• L""h 2I(h) dh

and
271 2 E
z. = A_2p2 f" h2J(h) dh

1 There are several ways of calculating the "average value" of the diameters of a

particle. The average value considered here is not the same as the average used
in §2.1.2.4 (IR for the sphere).
INTERPRETATION OF RESULTS 159

For incident beams of large height, equations 38, 3.llc, and 3.lld give

f'°) ..F1(h) dh
l = 2
c L'' h..F1(h) dh
(39)

Direct measurements of the total scattered energy have not yet been
employed, but, if the sample scatters rather strongly, the method devised
by Warren (§3.3.7) could be used. The evaluation of the integrals

3
\
\
\
\
2 \
\
\
\.
\.
\.
\.
\.
\.
o~~~"--~-'........~~~~~~~~__,~
0 0.5 logx

Fig. 51. Log f 1 vs. log x for a sa.mple of naphthalene. black. f 1 is

the scattered intensity for an incident beam of infinite height, and x iB
the distance on the film. The dashed line indicates a line of slope, -3.
(L. Kahovek, G. Porod, H. Ruck [97'].)

fhf 1 (h) dh and particularly ff1 (h) dh cannot be made very accurately
because of the unknown small-angle part of the curve. If the small-
angle scattering increases rapidly with decreasing angle (Fig. 51) the
extrapolation of the curve to zero-angle is very uncertain.
As an illustration of the application of these methods let us discuss the
results obtained by Kahovek, Porod, and Ruck (97'] for a sample of
naphthalene-black in which c = 0.12. The experiments were made with
monochromatized Cu Koc radiation. The vertical dimensions of the slits
defining the primary beam were large, so the beam could be considered
as infinitely high. The intensity measurements were made with a Geiger
counter. Figure 51 shows the curve of the variation of log./1 as a function
of log x, where x is the distance between the point of observation and the
direct beam. The curve verified the relation lim x3f1(x) = A, a constant.
160 SMALL-ANGLE SCATTERING OF X-RAYS

In addition the following quantities were determined:

f .F1(X) dx = E

f x.F1(x) dx =Q
From these it was found that for this sample,
S h A
V = 4 ; c (1 - c) Q= 7.5 m. 2/cm. 3

4Vc
l0 =s =650A

xE
l =2--=600A
• hQ

An electron-microscope study of this specimen showed spherical grains

with a diameter of approximately 1000 A.

REFERENCES FOR CHAPTER 4

Debye, P., and Menke, H. (1931), Physik. Z., 31, 797.
Ehrenfest, P. (1915), Proc. Amsterdam Acad., 17, 1132 and 1184.
Fournet, G. (1955), J. phys. radium, 16, 395.
Perutz, M. F. (1946), Trans. Faraday Soc., 42, 187.
Verwey, E. J. W., and Overbeck, J. T. G. (1945), Theory of the Stability of Lyophobic
Colloids, Elsevier, Amsterdam.
Whittaker, E. T., and Watson, G. L. (1927), A Course of Modem Analysis,
Cambridge, London.
5. COMPARISON OF THE RESULTS FROM
SMALL-ANGLE SCATTERING WITH THE
RESULTS OF OTHER METHODS OF
MEASUREMENT OF SMALL PARTICLES

In verifying these theories and the experimental methods which arise

from them, it is very important to compare the results of small-angle
scattering analyses with the data from other methods, particularly those
methods which yield results more directly. Furthermore, in ocder to
judge the possible applications of X-ray methods, it is necessary to
c.ompare their advantages and disadvantages with those of other physical
methods.
With this intent, Shull and Roess [155] and Van Nordstrand and Hach
[126'] studied the size of particles of a very fine powder (alumina) by
X-rays and then determined the total surface of the same sample by
absorption methods. The results were in reasonable accord, particularly
in the experiments of Van Nordstrand and Hach. The method of these
authors ( §4.2.3) is independent of the rather arbitrary hypotheses
demanded by the first method ( §4.2.2). Nevertheless these results do
not give the desired direct verification.
The real verifications have been obtained by working with particles
that are visible to the electron microscope or by working with crystals
whose dimensioUB can be determined from the width of their Debye-
Seherrer diffraction lines.

5.1. COMPARISON WITH THE ELECTRON MICROSCOPE

A good, well-handled electron microscope has a resolving power of the

order of 20 to 30 A. It is therefore capable of giving reasonably accurate
images of particles having diameters of several hundreds of angstroms.
This is precisely the size of particles that can be studied by small-angle
scattering. Therefore such images are the best means of checking the
X-ray results, provided that a sample has been chosen for which a rather
accurate calculation of the scattering can be made.
The discussion of Chapter 4 has shown that it is necessary to choose a
suspension containing identical particles if the calculatioUB are to be
161
162 SMALL-ANGLE SCATTERING OF X-RAYS

accurate. Now very regular preparations of latex spheres1 exist which

are actually used for calibration of electron microscopes. The mean
diameter of these particles is 2780 A. Yudowitch [192] studied the
small-angle scattering of these particles and, from measurements on five
successive rings of the observed secondary maxima (§2.3.6) (Fig. 41),
determined a diameter of 2740 A. Measurements of these particles were
repeated in three different laboratories, with the following results:
D = 2750 A, 2732 A, 2692 A. These numbers give an indication of the
accuracy that can be obtained with the X-ray methoc;l.
Testing the validity of the determination of the radius of gyration from
the slope of the curve of log J(h 2) is much more important because of the
greater generality of this method. This also has been verified several
times. As an example, let us cite the result obtained by Turkevitch,
Hubbell, and Hillier [162] for particles of colloidal gold which were pre-
pared so that the variation in size was a minimum, about 10 per cent.
The X-ray results gave a diameter of 824 A, and the electron microscope,
700 A. Fournet [48] studied much smaller particles of colloidal silver
and found a diameter of 130 A with X-rays and 120 A with the electron
microscope. In this last case there was a somewhat larger variation in
particle sizes than in the first.
Thus it can be concluded that the validity of the formulas used in
Chapter 4 has been proved experimentally. Let us also point out that
periodicities in images of fibers (collagen) have been found with the electron
microscope that had already been determined by small-angle X-ray
diffraction. Direct images could be compared in a straightforward way
with the models that had been determined by calculations from the
diffraction patterns. This is an interesting result and one which cannot
be achieved in the ordinary domain of X-ray diffraction, since the
structure models that are determined in the ordinary domain are on a
scale of the order of angstroms, completely outside the range of even
the most powerful imaging apparatus.
If the electron microscopy method is compared to that of the small-
angle scattering, the advantages of the first are quite evident. Both
the form and the size of the visible particles can be determined, regardless
of the heterogeneity or of the compactness of the specimen. The accuracy
of the results of the electron microscope is thus much greater than that
of the X-ray results when a good sample preparation can be made.
A shortcoming of the microscope is that the specimen must be dried
before it can be placed in the vacuum. This drying may cause either the
destruction of or a serious, unpredictable modification in the sample.
The X-ray technique, however, can be applied directly to suspensions
1 Dow Latex 580 G.
COMPARISON WITH OTHER TECHNIQUES 163

or colloidal solutions. Certain other difficulties can often occur during

the preparation of samples for the microscope; quite often the particles
will agglomerate, so that the elementary particle can be isolated only
with great difficulty. The image observed on the screen is the shadow of
the apparent contour of the grain, and this size is a function not
only of the nature of the sample but also of the condition of its
preparation.
In addition, if we consider a grain that contains numerous cavities,
these cavities will produce a scattering of X-rays although they are
invisible to the electron microscope. An example of this type of specimen
is carbon after an activation treatment; the grain has an unchanged
exterior form, but the activation has resulted in the emptying of internal
cavities, considerably modifying properties that are related to the total
surface of the carbon.
On the whole, if there is disagreement between the X-ray results and
those of the microscope, it can be predicted that the numbers obtained
with the X-ray technique will probably be the smaller.

5.2. COMPARISON WITH THE METHOD OF DEBYE-SCHERRER LINE

WIDTHS
If the sample is composed of very small crystallites, it is known that
the diffraction lines are broadened and, that, from a study of their widths,
or, more exactly, of the form of the lines, the dimensions of the crystallite
can be determined (Wilson (1949), Bertaut (1950)). The line broadening
is accompanied by the appearance of scattering at small angles, which
can be considered as arising from the broadening of the (000) diffraction
point. There is a very close relation between these two phenomena. For
instance, let us consider the diffraction by a crystallite having a lattice
that is assumed to be perfect. We can apply the general formulas (equations
1.2, 1.3) to determine the scattered amplitude. Now in this case p(x)
is periodic, so its transform, A(h), has a non-zero value only when h
coincides with one of the vectors of the reciprocal lattice. The inter-
pretation of equation 1.3 is thus the following: each point of the reciprocal
lattice is surrounded by the same region of scattering as is found around
the (000) point (the origin of reciprocal space), each of these regions being
determined by the "form factor" of the crystal, that is, by the transform
of s(x) (Ewald (1940)). In this case there is no difference between study-
ing the small-angle scattering and studying the broadening of the high-
angle reflections.
When the sample is composed of many crystallites arranged at random,
Bertaut (1950) has shown that the width of a D'ebye-Scherrer line can be
used to determine the root mean square thickness of the crystallite in a
164 SMALL-ANGLE SCATTERING OF X-RAYS

direction perpendicular to the corresponding lattice planes. The low-

angle scattering gives the radius of gyration of the particle and, thus, a
mean square value of atomic distances from the center of the particle.
There is thus a significant analogy between the results of the two methods,
but the averaging is made in a particular direction when the reflection
from a well-defined set of crystallographic planes is used, whereas there
is no such privileged direction operating in low-angle scattering.
There are several important differences between the two methods.
First, the complete profile of a diffraction line can be determined, since
nothing prevents measurements at the center of the line, whereas in
low-angle scattering the undeviated direct beam prevents measurements
from being made down to zero angle. Therefore the methods of analysis
of the patterns are different. When studies are made of larger particles,
which give only slightly broadened reflections, it can be predicted that
the same difficulties will not be encountered with the line broadening
measurements as with the low-angle scattering. Actually, the line
broadening technique can detect crystallites whose size is greater than
500 A with ordinary, small-diameter cameras, while such a particle
dimension requires systems with large sample-to-film distances for an
effective study of the small-angle scattering.
The small-angle scattering is much more intense than the scattering
around the lines, because all the crystallites of the powder contribute to
this, regardless of their orientations, whereas only the small number of
particles which are correctly oriented will contribute to the scattering
at a point on the diffraction line; also, the atomic scattering factor is a
maximum at low angles.
Comparisons have been made between low-angle scattering results and
those from line broadening, wherein an "average" dimension of the
crystallite was determined by means of approximate formulas such as
that due to Scherrer (Wilson (1949)), and the agreement between them is
satisfactory. At present much more accurate methods, based on a
Fourier analysis of the profile of the line, are known, but to our knowledge
they have not been used for a comparison with the low-angle scattering
analysis.
Such a study could reveal differences arising from two sources:
1. It is possible that the grains of the material would be made up of
more than one crystal (Fig. 52). Now the width of a line depends on the
size of the individual crystallite; if this is joined to crystallites of other
orientations, the line width is not affected. It can thus be predicted
that the small-angle scattering would give a larger crystallite size tba.n
that found from the line widths. It is possible that the sample would
not even be made up of separate grains; an example would be a deformed
COMPARISON WITH OTHER TECHNIQUES 165

metal in which there were no appreciable density fluctuations. The

small-angle scattering would then disappear completely, and the diffraction
lines would still be broadened.
2. If the crystallite is not perfect, the lattice perturbations would
produce a broadening of the lines. The separation of particle-size effects
from those of lattice distortions is a difficult problem and has been widely
discussed. Let us recall, for example, the long controversy over the
origin of the broadening of the diffraction lines of a cold-worked metal.

Fig. 52. Sketch of a polycrystalline grain. The solid line outlines the
grain as revealed by small-angle scattering, and the inner boundaries
are those revealed by Debye-Scherrer line widths.

In any event, the relative influence of the two factors varies with the
order of the reflection, while the influence of disorder is negligible in the
low-angle scattering pattern, as we have shown in the introduction. If,
therefore, the line broadening is attributed only to the influence of the
crystallite size, the crystallite dimension that is so determined will be
smaller than that found from the low-angle scattering pattern. For
example, certain authors have calculated the particle dimensions in several
varieties of carbon from the widths of the lines. Their results are not
correct, since there is considerable distortion in the crystallites. The
correct method, used by Franklin (1951), requires the use of the entire
curve of diffracted intensity, including the low-angle scattering.
We shall see in Chapter 6 an example of the separation of the effects
due to the size and to the imperfections of the crystallite that is based
on a comparison of the scattering at small angles with that around a
Bragg reflection (Al-Ag alloy).
When the degree of disorder increases, the material passes from the
crystalline state to an amorphous state, and then the various diffraction
rings are completely independent of the size of the particles. However,
the particle size can still be determined from the low-angle scattering.
166 SMALL-ANGLE SCATTERING OF X-RAYS

For example, a high polymer, such as Perspex, gives the same diffraction
ring when it is in a solid state as when it is dissolved in acetone. The
solution, however, gives in addition a small-angle scattering pattern from
which the size of the micelles of the colloidal solution can be determined.
Consequently the method employing the widths of lines can furnish the
same information as that found from small-angle scattering only if the
particles are practically perfect crystallites. If this is the case, it can be
more easily and more accurately employed when ~he crystallites have a
rather large volume.

REFERENCES FOR CHAPTER 5

Bertaut, E. F. (1950), Acta Cryst., 3, 14.
Ewald, P. P. (1940), Proc. Phys. Soc. (London), 52, 167.
Franklin, R. E. (1950), Acta Cryst., 3, 158.
Franklin, R. E. (1951), Acta Cryst., 4, 253.
Wilson, A. J. C. (1949), X-Rays Optics, Methuen, London.
6. THE APPLICATIONS OF THE
SMALL-ANGLE SCATIERING OF X-RAYS
We shall review in this chapter the principal problems to which the
methods described in the preceding chapters have been applied.
We would like to show how these methods have been employed in
relation to the known facts of a problem and the objectives of the investi-
gation. Without summarizing all the different investigations, a list of
which is given in the bibliography, we shall try to give a general picture
of the results that have been obtained and of the possibilities of the
methods a.t their present state of development, both for fundamental and
applied research.
We shall distinguish two classes of problems.
1. The first will concern the study of well-defined particles, either
large molecules or the grains of a finely dispersed substance.
2. The other series of problems will concern the study of a solid that
contains heterogeneities on a submicroscopic scale; these can be so
complex that a particle model cannot be employed.

6.1. LARGE MOLECULES

A conclusion of the theoretical study of small-angle scattering was that
such scattering can be most favorably exploited when the particles are
identical and widely separated from one another. If the particle is a
molecule that is chemically well defined, we can be sure that the first
hypothesis is verified, and it is always possible to disperse the molecules
in a solution or suspension dilute enough to satisfy the second.

6.1.1. DILUTE SOLUTIONS

The experimental conditions are good if the particles have a radius of
gyration1 of the order of 10 to 50 A, or a volume of approximately 8 X 10 3
to 2 X 10 5 A3 • The corresponding weight of the particle is from I to
50 X 10- 20 g. if the density is between I and 2. The gram-molecular
weight of such molecules is between 5,000 and 250,000. This is the
1 The introduction of this po.re.meter, the radius of gyre.tion, is found on p. 24,

where it was defined as the root mee.n aqua.re of the distances of a.toms from the
center of gravity of the pe.rticle, ee.ch diste.nce being modified by a coefficient eque.l
to the atomic nwnber of the e.tom.
167
168 SMALL-ANGLE SCATTERING OF X-RAYS

order of magnitude of the molecular weights of many important biological

compounds, as, for example, the proteins.
These remarks explain why the study of proteins offers one of the best
applications of this method. As a matter of fact a large number of
investigations have already been carried out on proteins, as is shown in
the bibliography at the end of the text. We believe that it is in this field
that small-angle scattering can give the most valuable and most important
results from a general point of view.
The experimental work begins with the preparation of a dilute solution
in which the molecule to be studied is found in a pure state (within
several per cent).
It is essential that the electronic density of the solvent be as different as
possible from that of the molecule, because the smaller this difference, the
weaker will be the low-angle scattering.
Frequently aqueous solutions produce a rather intense scattering, so
that solutions down to a concentration of only a few per cent can be
studied.
The object of the experiment is to determine the curve of the scattered
intensity as a function of scattering angle. A photographic determination
of the scattering pattern must therefore be followed by a microphoto-
metering of the film. It is certain that Geiger-Miiller counter measurements
can be made to a higher degree of accuracy than photographic measurements,
and this accuracy is very desirable, particularly for solutions of molecules,
since the theoretical interpretation of this case is on a rather sound basis.
The measurements of the radius of gyration of the molecules can be
made by the method of §4.1.1.1. If the shape of the molecules is not too
far from spherical, the description of the experimental curve as an
exponential in -kh 2 is a sufficiently accurate approximation over a large
angular region. Here an incident beam of large height can be employed
without having to make corrections before interpreting the curves of
log /(h 2) (seep. 114). This is important from the point of view of rapidity
of measurement. As we have already indicated, the radius of gyration
does not completely define the form and the size of the molecules, but
(and here is a point of view on which we must insist) it is a well-defined
geometrical parameter which can be considered as a characteristic of a
molecule of any shape.
The experimental techniques, as illustrated, for example, in the counter-
equipped apparatus of Beeman (Fig. 25, p. 95), have reached such a
degree of simplicity and accuracy that one can envisage a broad program
of systematic measurements of radii of gyration of different proteins or
other large molecules in biochemical laboratories.
There are many services which these measurements could render. We
APPLICATIONS 169
have already seen how a knowledge of the radius of gyration, together
with other data, can give indications of the shape and the real dimensions
of the molecule. 1 Figure 53a gives the curves of log J(h 2 ) for a series of
proteins, and it can be seen that the straight lines are defined well enough
to permit an accurate determination of the radius of gyration. In Fig.
53b the curves of log I vs. log h are plotted for two ellipsoids of the same
radius of gyration with axial ratios of 2 and 3. The experimental points
for ovalbumin fall just between these two curves, indicating that the

log I log I
2

Lysocym

OL---'-~~-'-~,-1-::----,:-'::-~
0 1·10- 2 2·10- 2 3·10- 2 h 2 (A- 2 ) 0.02 0.05 0.1 0.2 h(A- 1 )
(a) (b)
Fig. 53. (a) Log I vs. h' for solutions of different proteins; (b) log I
vs. log h for ellipsoids of axial ratios 2 and 3 (theoretical). The small
circles O show the experimental points obtained for a solution of
ovalbumin. (Ritland, Kaesberg, and Beeman [492].)

ovalbumin molecule has an axial ratio of about 2.5. With the latest
improvements in experimental techniques (use of a powerful X-ray tube
for studies of dilute solutions in which interparticle interferences are
negligible, correction of the curves for the effect of beam height), the error
in a determination of the radius of gyration is of the order of 2 to 3 per
cent. This parameter could also be used as a means of identification, if
no simpler method were possible. Finally, the single aspect of the
linearity of the curve of log J(h 2 ) is an indication of the uniformity of size
of the particles and thus constitutes a means of inspection of the purity
of preparations. For example, the curves of log J(h2) for euglobulin and
pseudoglobulin extracts from the serum of horse blood are strongly
1 In addition to the other data previously mentioned (p. 130),the sedimentation

and diffusion constants can be used toward this end (Ritland, Kaesberg, and Beeman
(492]).
170 SMALL-ANGLE SCATTERING OF X-RAYS

convex (Ft<urnet [48]), and other techniques have shown that the globulins
are complex mixtures.
Dervichian, Fournet, and Guinier [419] have studied the denaturing of
hemoglobin and albumin serums with urea. When the small-angle
scattering characteristic of the molecules disappears, the molecules have
been broken up into small fragments. If all the molecules are not

100.---,~~-,-~~-,-"T'"--rl~-.--~--.

80 5% T.B.S.V.
60 0.5 x 10-mm. slits

40

~ 20
=
~
~
~ 10
~ 8
"' 6
c:
~
4

1
60-=---40::'="-~--:27
o~~~l0:--'-8~6'---4..l..-~--'2
Angle, radians x 10-3
Fig. 54. Scattered intensity for a l'i per cent solution of tomato bushy
stunt virus. The dotted curve is the theoretical curve for spheres of
310-A radius. (Leonard, Anderegg, Shulman, Kaesberg, and Beeman
[478].)

destroyed, the patterns decrease in intensity without changing their

appearance. A change of appearance of the patterns occurs when new
particles are formed.
Beeman and coworkers have obtained especially accurate results for
virus macromolecules [376]. The experimental curve showed several
peaks which the authors were able to make coincide with the successive
peaks in a scattering pattern of homogeneous spheres. They thus demon-
strated that these viruses were spherical and found their diameters (Fig.
54); for example, the diameter of the tomato bushy stunt virus was
310 A. Then, by a comparison of this diameter with other molecular
APPLICATIONS 171

constants, they were able to determine the degree of internal hydration

of the molecules.
It is quite evident that the information drawn from the low-angle
scattering concerns only the exterior form and not the structure of the
molecule, while much more information on the structure is contained in
the diffraction patterns of crystallized proteins.
By comparing the work of Fournet on hemoglobin in solution [48]
with that of Perutz and coworkers on crystallized hemoglobin [407], it
can be seen that these two. methods are undertakings of a completely
different order of difficulty. The crystallographic method requires a
considerable amount of work (and sometimes only for an uncertain result)
both in the preparation of a usable crystal and in the interpretation of
the patterns, whereas a low-angle scattering investigation of a solution
can now be considered a routine operation.

6.1.2. CONCENTRATED SOLUTIONS

Another field of investigation in the study of large molecules such as
proteins has been opened by the progress of more recent theoretical
approaches. This is the study of the apatial distribution of molecules in
solutions at high concentrations.
The organization of the molecules in a solution can be successfully
determined. The method to be applied for this purpose has been described
in §4.1.2.1. Without repeating the detail'I of the technique, let us recall
only that, if a solution is very concentrated and if the particles show a
high degree of organization, the uniformly decree.sing scattering curve of
the dilute solution is deformed, showing a bump or even a maximum.
A practice.I consequence of this shape of the curve is that for a quanti-
tative study either the correction for the slit height must be made or else
beams used that are as small as possible. Obviously this complicates the
determination of the curves. Here again only Geiger-Muller counter
measurements can convey the accuracy required for the theoretical
calculations. However, photographic patterns can often give immediately
qualitative information which will help to decide whether a precise study
is worth being undertaken.
The theoretical study showed that very carefully performed experi-
ments made on a series of solutions of various concentrations can lead to
the determination of the mutual potential energy of a pair of molecules as
a function of the separation of their centers, if it is assumed that the
molecules are spherically symmetric.
As an example of studies of this type, we cite those carried out by
Beeman and his collaborators for various solutions of spherical viruses.
Figure 55 gives the low-angle part of the scattering curves for solutions
172 SMALL-ANGLE SCATTERING OF X-RAYS

of turnip yellow mosaic virus of various concentrations (Schmidt,

Kaesberg, and Beeman (1954)). A maximum is observed on the curves
for concentrations greater than 5 per cent. Now, theoretical calculations

50 •A
oB
() c
45 eD

40

35

~30
·;;;
<:
2<:
- 25

20

15

10

Scattering angle, radians x 10- 3

Fig. 55. Interparticle interference effects in the low-angle scattering
from concentrated turnip yellow mosaic virus solutions. The ordinates
are adjusted so that the curves would coincide if scattering were pro-
portional to concentration at all angles. Curves A, B, 0, and D refer
to concentrations of 1.6, 3.2, 6.3, and 12.8 per cent, respectively.
(Schmidt, Kaesberg, and Beeman (1954).)

based on the hard sphere model (§2.2.3.4 and Fig. 12) show that this
maximum is perceptible for 8v 0E/v1 > 2, where E is a constant approxi-
mately equal to unity, v0 is the volume of the molecule, and v1 is the
reciprocal of the number of molecules per unit volume. By applying
APPLICATIONS 173

this theoretical result to the observations on the virus solutions, a volume

v 0 is found which is 4 times as great as that determined from the experi-
mental scattering curves of dilute solutions. From this it is concluded
that the molecules do not behave as hard spheres which can come in

I I
Arbitrary
units

(b)

I I

(c)

(e)
Fig. 56. Scattering from red corpuscles in various preparations.
(a) Artificia.l hemoglobin solution; (b) normal red cells prepared with
heparin; (c) sickle cells prepared with heparin; (d) fetal red cells
prepared with heparin; (e) normal red cells prepared with sodium
citrate.

contact with one another. Instead, there must be repulsive forces which
give the molecule an effective collision diameter of about l.6 times their
actual diameter.
Differences in the form of the scattering curves have been observed
between hemoglobin in normal human red corpuscles and hemoglobin in
abnormal <iorpuscles (fetus blood, blood affected by the sickle-cell disease)
174 SMALL-ANGLE SCATTERING OF X-RAYS

[420]. Hemoglobin extracted from these various cells and put in dilute
solutions gives the same scattering curve, independent of its source. All
these hemoglobin molecules therefore have the same radius of gyration,
but the red corpuscles themselves give different curves in which the bump
is more or less accentuated, depending on the type of corpuscle (Fig. 56).
From this qualitative observation we can deduce that the organization
of the molecules is less for the sickle cells than for the normal corpuscles
but is greater for the fetus corpuscles.
Let us point out also the work of Riley and Oster [371] on hemoglobin
and on ova.lbumin serum in concentrated solutions that was based on the
6Qr-~~---,.--~~---.~~~-.

ol.-~--1~.:'.:=±~==~
0 2 3 ·10- 2 radian
9
Fig. 57. Bcatt.ering from a dilufe solution of hemoeyanine. The
int.ensity in arbitrary units is plotted against 6 = Akf4.r (Cu Kcx
radiation). (Kratky, Sekore., and Friedrieh-Freska (474).)

method of interpretation discussed in §4.1.2.2.1. The same method of

interpretation was used by Bateman, Hsu, Knudsen, and Yudowitch
[382] in a. series of experiments on red cells swollen to different degrees in
solutions of variable tonicity. Their results confirm those of Riley and
Oster but are not in accord with the works previously cited. A basic
assumption whose correctness is not proved in these works is that the
molecules form a sort of para.crystalline structure, which expands without
appreciably changing the nature of the molecular arrangement when the
concentration decreases.
Hemocyanine molecules [416], [474] are peculiar in that their scatter-
ing curve shows a maximum even for very dilute solutions (2 per cent),
the position of the maximum being independent of the concentration
of the solution (Fig. 57). It is therefore improbable that this maximum
can be explained in terms of interparticle interferences. Isolated particles
can give such patterns if it is admitted that the particles a.re complex,
ea.ch being made up, for example, of a.n agglomeration of a certain number
APPLICATIONS 175

of molecules joined together. This hypothesis agrees with the description

suggested by observations with an electron microscope (Fournet [48]).
As a last example, a considerable number of investigations have been
devoted to aqueous soap solutions in the presence of electrolytes (see
Bibliography). These solutions give a small-angle pattern containing
rather sharp rings, often designated as I and M bands. In the majority
of the investigations the two rings have been interpreted in terms of
"distances" having simple physical meanings. For example, the M band
was related to the spacing between two layers of molecules and the I
band to an intermicelle distance by means of an equation analogous to
4.24.
It is certainly not permissible to separate the two peaks in this way or
to apply to these micelles relations that are valid for crystalline arrange-
ments. The complexity of the patterns of the soaps arises from the fact
that the micelle has an internal structure, so that the curve of its structure
factor shows peaks. In addition, the micelles are rather ordered, so that
there are strong interferences that deform the scattering curves of the
individual micelles. An a priori analysis of the data is practically
impossible. We can only assume different models on which calculations
can be based (Corrin [359]), but as Hughes has pointed out [362], [363],
there are many widely different models that can explain certain experi-
mental facts.
It seems, therefore, that it is necessary at the present time to reconsider
the majority of the "quantitative" results for soap solutions that have
been arrived at by overly simplified reasoning. It is necessary to retain
only those results that have been obtained by a restrained application of
a correct theory to the patterns, leading to a modest but sure
interpretation.
There is no doubt that often the numbers that are given are not even
approximately correcc, and the so-called "conclusions" from the X-ray
patterns can only bring confusion into questions that have not yet been cleared
up if these conclusions are adopted by non-specialists.
In several cases the tendency to order among the large molecules
becomes so strong that they tend towards a regular arrangement, and it
becomes legitimate to adopt a crystalline model. The dimensions of the
basic units are such that these lattices have very large cells, so that
diffraction lines appear at very small angles. Bernal and Carlisle [9]
found that the molecules of wet turnip yellow virus form a cubic, diamond-
type lattice. The distance between molecules is 304 A and the layer of
water between neighboring molecules is 70 A thick. In this case the inter-
molecular forces succeed in bringing about good regularity over long
distances.
176 SMALL-ANGLE SCATTERING OF X-RAYS

Another example is furnished by the red corpuscles of a rat. After

being subjected to a special treatment (Ponder (1945)), these give a
small-angle pattern containing a series of extremely sharp lines that coin-
cide with those from hemoglobin crystals prepared from pure solutions.
The organization of the molecules has here resulted in crystallization.
At this stage the interpretation of the pattern becomes simple, because the
classical procedures of crystallography can again be applied. However,
we wish to repeat that it is justifiable to apply these procedures only if
the experimental results permit it; the indispensable condition is that the
pattern contain many well-marked lines.
Despite the difficulties we have not hesitated to stress, it is certain that
the results already obtained show the chemist and biochemist interested in
large molecules that small-angle scattering patterns should no longer be
neglected, as has generally been done. It is obviously in this region of a
diffraction pattern in which the forms of scattering are found that show the
organization of matter on the scale of large molecules. Even if a complete
interpretation is not possible, the pattern can serve to characterize a
product; therefore it can be a tool for following and checking preparations
or fractionations. For example, lipides extracted from egg yolk give rise
to two rather sharp diffraction lines. When these lipides are dissolved
in alcohol a separation into two phases is observed; the upper phase
always gives rise to two lines, while nothing is observed for the lower phase.

6~.HIGH POLYMERS
6.2.1. STUDY OF SOLUTIONS
The large molecules formed by polymerization offer a natural field of
application for small-angle scattering. But here a difficulty is found that
did not exist for proteins; in general, the particles are not all identical,
since the number of elementary monomeric units can vary from one
polymer to another. It should therefore be expected ( §4.2) that quite
often the X-ray results cannot be very precise. If there is only a small
variation in the sizes of the polymers in the mixture, an average radius of
gyration can be determined, as defined in §4.2.1. When a large range of
sizes is present and when the mixture contains some very large particles,
this measurement is difficult. It must not be forgotten that, from its
definition, the "average" radius that is found is near that of the largest
particles. Consequently this radius can be determined only if an appar-
atus is used that allows measurements at very small angles. If the shape
of the particles is known and if it is assumed that they are all similar, it is
possible to derive the distribution function for the sizes of the particles
from the experimental measurements (§4.2.2). In order that the experi-
ments be accurate, the concentration of particles must be small, and it is
APPLICATIONS 177

essential that they be contained in a medium of very different electronic

density so that there will be sufficient scattered intensity.
Kratky and Wurster [476], in a study of cellulose fibers submitted to a
series of swelling agents, verified that the intensity of scattering at a given
angle was proportional to the square of the difference of the electronic
density between the crystallized cellulose and the liquid in which the
micelles were immersed. An unknown electronic density in particles can
thus be determined by finding the solvent that cancels the small-angle
scattering.
The result of such small-angle scattering experiments could be the
determination of molecular weights of high polymers. Actually, to the
best of our knowledge, there has not yet been a systematic employment of
X-rays towards this end, either because of difficulties in the method for
particles that are too large or of too heterogeneous a distribution, or
because other physical methods are simpler and more certain.
Kratky and Porod [469] have studied the scattering from coiled chain
molecules in solution. Calculations predicted a small-angle scattering
curve of the form e-kh", but this part is not visible since the radius
of gyration of the molecule is too large. They further predicted that for
increasing angles the intensity would vary first as h- 2 followed by a curve
in h- 1 • Thus, if a plot of the measured intensity multiplied by h 2 were
made, we would obtain a horizontal straight-line section followed by
another straight line of negative slope. The transition point between the
two parts of the curve would be related to the mode of coiling of the
molecule, or, more exactly, to the "distance of persistence" in which on
the average the molecule does not change its orientation appreciably.
An experiment was made on polyvinylbromide in methylnaphthalene.
The theoretical results were not verified very clearly, but nevertheless the
measured distance of persistence was of the expected order of magnitude.

6.2.2. STUDY OF FIBERS

X-rays have been used to a great extent in the study of polymers in the
form of fibers (see Bibliography). In fibers the particles have a common
average orientation, and this, by simplifying the interpretation of the
scattering patterns, increases the interest in them considerably. Thus it
was noticed a long time ago that ramie fibers gave a low-angle scattering
spot that was elongated in the direction perpendicular to the fiber axis.
In the direction parallel to the fiber axis (the fiber is then placed perpen-
dicular to the height of the beam if this is defined by slits), the scattering is
not perceptible if measurements cannot be made at extremely small
angles. This is immediate proof, in advance of any quantitative
measurements, of the existence of elementary particles that are very
178 SMALL-ANGLE SCATTERING OF X-RAYS

B (a)
-*-
a b c

Fig. 58. (a) Sme.11-e.ngle scattering from cylindrical e.nd le.melle.r

micelles on e. helix a.bout the fiber a.xis; (b) ame.ll-e.ngle patterns of
vegetable fibers: a, coir; b, Tula ixtle; c, J e.ume.ve ixtle. (Heyn [446].)
APPLICATIONS 179

elongated in the direction of the fiber axis. Hosemann [80] has succeeded
in estimating the length of these cellulose micelles as being greater than
3000A.
The first result to be obtained from the patterns is therefore the demon-
stration of the form and orientation of the elementary particles in a very
simple, qualitative manner. Heyn [442] has studied vegetable fibers, in
which the micelles are arranged in spirals around the fiber axis. The
pattern is composed of a cross, the angle between the branches being
twice the angle that the axis of the micelles makes with the fiber axis.
This pattern can be explained by assuming that the micelles are not
cylindrical but are rather in the form of lamella. To show this, let us
consider a cylindrical particle making the angle a. with the fiber axis,
which is normal to the X-ray beam (see Fig. 58). This causes a streak of
scattering elongated in the direction OX, the line of intersection with the
film of the plane normal to the axis of the micelle. When the particle
rotates around the fiber axis, OX sweeps out ;i AOB = 2ot. But the
variation of the intensity along OX will maintain itself more or less
constant throughout the interior of this angle because the average
inertial distance of the cylindrical particle in the direction of OX is
independent of the position of the particle. On the other hand, if the
particle is a thin plate, the direction OX is parallel to the direction of
minimum thickness only for the limiting directions, OA and OB. There-
fore a cross rather than a fan-shaped pattern should be observed if the
particles have the form of lamella, and this is the result given by experi-
ment.
Heyn [446] has also used small-angle scattering to show the orientation
phenomenon in fibers of viscose drawn under different conditions.
The fiber that has been studied the most with X-rays is cellulose, for
which the existence of a low-angle scattering in the form of a streak
perpendicular to the fiber axis was long ago observed (Mark (1932);
Kratky [466]; Hosemann [452]; P. H. Hermans [440]; etc.). But it
cannot yet be said that a clear picture of the micelle structure of cellulose
has been achieved. This structure is certainly very complex, and the
different investigators have tried to interpret the scattering patterns by
means of predetermined models which were necessarily oversimplified.
This has led to quite varied results which do not agree among themselves.
A method of quantitative interpretation of such a pattern, based on the
assumption of no interparticle interferences, was given in §4.1.1.2. An
"average inertial distance" with respect to the fiber axis is obtained
(equation 4.11) that plays the role of a radius of gyration for disoriented
particles and that is a characteristic of the breadth of the particle per-
pendicular to its axis. Heyn [447], in a study of different fibers in a wet,
180 SMALL-ANGLE SCATTERING OF X-RAYS

swollen state, found quite linear curves of log 1(h 2 ) and determined
inertial distances of the order of 8 A for jute and 12 A for ramie. If the
particles are cylindrical, these correspond to diameters of 22 and 35 A,
respectively.
Many fibers give curves of log J(h 2 ) having a curvature too accentuated
to allow a valid determination of the radius of gyration. Attempts have
been made to determine the combination of particles of various sizes that
would give the observed pattern. Hosemann [80], by applying the
methods of §4.2.2, has determined the statistical distribution of the
diameters of particles of acetylcellulose.
Nevertheless, the question arises whether it is legitimate to apply
methods to cellulose fibers that are valid theoretically only for a system
of particles of low concentration, since the known, rather high density
of this fiber leads to the prediction that the micelles should be very
tightly packed (see Bibliography for numerous papers by Kratky and
Heyn). Actually this approximation is perhaps not as bad as could be
feared, since the individual particles are not of uniform size. It has
been shown ( §2.3.3) that irregularity of sizes considerably diminishes the
influence of interparticle interferences for random-shaped particles. It is
just as true, however, that there are certainly cases in which the appear-
ance of the scattering curve proves that interferences are playing a part
(see §6.2.3).
A supplementary item of information, the absolute value of the scattered
intensity, can also be used, as was done in comparing dry and wet cellulose
(Fournet and Antzenberger [434']). This is not useful in finding a priori
the structure of the cellulose, but it enables a decision to be made on
whether a model of the structure is acceptable. As a result of this work,
cellulose can no longer be considered as made up of particles isolated by
empty spaces; rather we must assume the existence of dense particles
having an ordered internal structure, each surrounded by amorphous,
cellulosic matter of slightly lower density. The volume occupied by the
"particles" is of the order of 25 per cent of the total volume, so that it is
legitimate, for dry cellulose, to use the approximation of a dilute system.
When the cellulose is swollen by water, the density of the intermediate
regions becomes smaller and the contrast of the "particles" increases.
This causes the very definite increase in the intensity scattered by the
system that is observed experimentally. At the same time the particles
become more organized (§6.2.3).
Kratky and his collaborators take as a base for their interpretation a
model drawn from general knowledge about the structure of fibers. The
micelles are taken to be in the form of thin platelets having large surfaces.
These platelets are stacked on one another in an approximately parallel
APPLICATIONS 181

manner like pages in a book, but these "pages" are of irregular thicknesses
and are separated from their neighbors by intervals that are also more or
less irregular. The complete fiber is formed of groups of micelles having
in common a general direction of the fiber but with slight misorientations.
We have seen in a previous section that micellar plates of thickness d and
random orientation in a system of low concentration give a scattered
intensity in the region accessible to experiment which is proportional to
; 1 d (8in~7j2 )r (equation 2.41). 1 Furthermore a group of Ia.mellar
micelles will give an intensity proportional to Ji.-1<1>, where «II is a one.
dimensional function which depends on the distribution of empty a.nd
occupied segments on the normal to the plane of the lamellas. The
method of Kratky and Porod [137] is thus to muUiply the observed intenaity
by h. 2 and to interpret the resulting function «II as the scattering pattern
of a linear structure. This pattern depends not only on the distribution
of thicknesses of the micelles but also on their arrangement. We have
already pointed out (p. 70) that the importance of interferences is par·
ticularly great for linear structures. The controversies that have arisen
between the partisans of the theory of isolated particles (which neglects
interparticle interferences) and those who give these interferences &
dominant place in the explanation of small-angle scattering phenomena
have been obscured by the fact that the two groups are working from
different models.
Zernicke and Prins [309], J. J. Hermans (1944), Porod [137], and
Hosemann [276), [278) have calculated the scattering from different
models of linear structures for given statistical distributions of segments
and the intervals which separate them. One of these models is the
following: Segments of equal length or with small fluctuations a.round an
average length a.re separated by intervals such that the degree of expansion,
the.t is, the ratio of the total length of the structure to the total length of
the segments, is fixed. The total length of the intervals between segments
is given, but the division of this length into individual intervals is randomly
made. For an infinite degree of expansion we observe the scattering
from isolated particles (Fig. 59), and as the density of the system increases
the curve becomes deformed, just as was found for the closer packing of
spherical particles (Fig. 12). The difference between these cases is that
the intensity scattered by dense linear systems decreases in absolute
value until it becomes zero when there is no interval between segments,
whereas spheres in a compact structure still give a crystalline diffraction
1 This expression describes equation 2.41 in a slightly modified form, where l/R 1
has been replaced by d; this follows from the fact that Rid is proportional to n,
the total number of electrons in the particle.
182 SMALL-ANGLE SCATTERING OF X-RAYS

pattern. Janeschitz-Kriegl [457] has studied the small-angle scattering

pattern oft hreads of viscose in various degrees of distention. By compar-
ing the experimentally determined curve of h 2J(h) with the group of
curves of Fig. 59, he concluded that the micelles had an average thickness
of 50 A with a fluctuation of 50 per cent, but the determination was not
very accurate.
Another model is one of segments whose distribution of lengths follows
a Gaussian curve, these segments being separated by small, equal intervals.

1.0 . . - : : : - - - - - - - - - - - - - - - - - - ,

q=r:JJ

0.5
3

1.5

1.2 1.1
0
0 0.5 1.0
ah
211"
Fig. 59. Scattered intensity from a linear structure composed of
segments of uniform length, a, separated by intervals which follow the
distribution law for perfect disorder, H(y) = (l/b)e<-11/•>, where b is
the average interval. The degree of expansion, q, is equal to (a + b)Ja.

This is therefore a very dense system. The scattering curve is similar

to that for a liquid or, rather, a paracrystal. The intensity is very weak
at very low angles, and there are large maxima in the neighborhood of
the angles at which would be found the maxima from a linear lattice of
period equal to the average micelle thickness. lVhen the fluctuations of
thickness are large, only the first two such maxima are observable. The
experimental curve of h 2 I(h) for non-distended fibers of viscose [457] is of
this type and gives a mean micelle thickness of. 40 A. However, the
agreement between theoretical and experimental results is rather
qualitative.
APPLICATIONS 183

6.2.3. ORDERED ARRANGEMENTS OF MICELLES

Certain high polymers give a completely different type of small-angle
scattering pattern. Instead of a scatt{lring that decreases regularly from
the center, bumps or maxima are observed that are more or less pro-
nounced. These patterns are similar to those already pointed out in

0.3 h = 4X8 (A-1)

Fig. 60. Absolute measurements of the intensity per C 6 H 100 6 group

scattered by ramie fiber. Curve a: dry fiber; curve b: wet fiber.

connection with concentrated solutions, but with the added complication

that the pattern no longer has circular symmetry, showing lines or diffuse
spots instead of rings.
This effect has been noted on patterns of substances such as moist
cellulose (Figs. 60 and 61), the polyamides (nylon), and the polyethylenes

(a)- (b)
••
Fig. 61. Small-angle scattering pattern of "Fortisan" rayon.
(a) Fiber in the dry state; (b) fiber swollen in water. Natural size of
photograph; sample-to-film distance 20 cm.; Cu Ka. radiation.
(Heyn (1953).)

[442], [441'], [441"], [480]. Although complete use has not yet been
successfully made of the data, these observations apparently are the
origin of an important application of small-angle scattering methods.
184 SMALL-ANGLE SCATTERING OF X-RAYS

Figure 63 gives an idea of the effects that are observed. It should be noted
that the use of the photographic method is almost indispensable in this kind
of study. The general appearance of the pattern that will be found is
not known a priori, and a correct interpretation of a Geiger-Muller
counter measurement without previous photography would be practically
impossible.
The explanation of these phenomena is based on the existence of an
order among the particles, or, in a more general manner, on the existence
of certain large-scale regularities in the density of matter.

·jt:·
c

~-
·'i

;, ./
Fig. 62. Small-angle diffraction pattern from a plain, dry fiber of
kwigaroo tail tendon. Fundamental period of sample: b 0 = 627 A.
(Bear and Bolduwi (397].)

As we have already pointed out, the problem is simple when the sample
can be considered in terms of a well-defined, true periodicity. The most
typical example is provided by the fibers of collagen (kangaroo tail)
(Bear and Bolduan [397]) (Fig. 62) or fibers of myosin, in which the
periods along the fiber axis are respectively 700 and ] 46 A and in which
the periodicity is so perfect that the pattern contains several tens of lines,
up to very high orders (40th). In other substances periodicity exists in
the plane perpendicular to the fiber axis. This structure is manifested
by quite sharp lines, small in number {often only two), on the equator of
the pattern. This is exemplified by chrysotile (Fankuchen and Schneider
[434)) and by the tobacco mosaic virus. These can be explained by
assuming that the molecules show a crystalline type of arrangement
{hexagonal). The different molecules of tobacco mosaic virus are
APPLICATIONS 185

separated by a fixed distance; Bernal and Fankuchen [402] found this

to be equal to 152 A for the dry state and 450 A when the virus is in
solution. The arrangement is rather imperfect, resulting in a broadening
of the reflections.
These viruses are exceptional cases. Generally the degree of regularity
is much smaller, which seems natural, given the weak forces that can be
exerted between the large basic units. The difficulties in interpreting
the observed facts are consequently greater, and a complete explanation
of the facts is still far from being achieved. Two methods of approach

I ~---1-_J~!-- ----i--- ----------~

(a) (b) (c)
Fig. 63. Small-angle pattern of synthetic fibers. (a) Copolamide,
stretched, T = 20° C; (b) same material, stretched, T = 200° C;
(c) same material drawn from the melt, unstretched, T = 20° C. Fiber
. vertica
. I. Sca1e given
. . terms o f A 2" A . (H ess and
ruus m 2 sin 0 = h
Kiessig [44.l'].)

are generally adopted: the use of either a model of a perturbed crystal

in which there is no long-range order (a paracrystal) or a model of a
gaseous arrangement in which the disposition of neighboring molecules
is somewhat ordered. At present we can only repeat the conclusions
offered in the discussion of ordering between molecules of proteins or
soaps. It is dangerous to apply results valid for crystalline arrangements
to these cases without caution. The application of Bragg's law should
be distrusted, even when it is believed to be "corrected" by the use of a
factor somewhat larger than unity (1.23, 1.22, 1.16, 1.05, etc.).
It is certainly regrettable that these negative remarks cannot be
followed by a correct general interpretation, but this has not yet been
developed.
As an example of the general case, let us consider the patterns of
synthetic polyamide fibers (Hess and Kiessig [441 '], [441 "]). In addition
to a central spot elongated perpendicular to the fiber axis, the pattern
also contains two other regions of scattering whose aspect and position
vary according to the treatment given to the fiber (Fig. 63). For example,
a copolamide gives rise to two rather diffuse streaks perpendicular to
the fiber axis. The centers of these streaks are found by Bragg's law to
correspond to a distance of the order of 80 A. Hess and Kiessig proposed
186 SMALL-ANGLE SCATTERING OF X-RAYS

as a model of this structure a regular arrangement of micelles with a

periodicity of 80 A along the fiber axis. The smearing out of the spots
was interpreted as being the result of the small number of micelles
coherently grouped together. When the temperature of the fiber is
raised, the distance of the spots from the center of the pattern diminishes
a little, which, according to the above model, would mean that the
periodicity increases by a lengthening of the elementary micelle. But, in
trying to take account of all the observations, this model leads to contra-
dictions, as Hosemann [84] has pointed out. If it were true that the
width of the spot, interpreted as a crystalline reflection, is due to the
limited size of the crystallite, a diffuse spot of the same size should be
found around the center of the pattern, the (000) reflection, as was pointed
out on p. 163. This is not the case; although different treatments of the
fiber can modify the two other spots, the central spot does not vary.
Moreover, no spots corresponding to higher orders of reflection have been
detected. Thus the "crystalline" model is not a useful approximation.
The starting point for a correct explanation is to suppose that the fiber
contains micelles between which there is a lower density of matter.
These micelles are distributed with a certain degree of order, particularly
in the direction of the fiber axis. This organization of the micelles among
themselves will cause the large scattering maxima that are observed. It
may be possible to define this organization quantitatively but only on
two conditions: (1) We must begin with a scattering curve that has been
well determined experimentally. This means that a counter should be
used to determine precisely the shape of the scattering distribution that
the photographic technique has revealed. (2) We must consider the
complete curve of the scattering. The simple determination of the position
of the maximum has no significance. This maximum, in fact, can be
displaced either by a change in the organization of the micelles, as, for
example, when the organization becomes more regular (seep.140),or by the
agglomeration of a larger number of micelles without the micelles themselves
changing dimensions. These effects are opposite to those foreseen by
Hess and Kiessig. This shows, for example, that the theory of the effect
of temperature can be completely changed.
A detailed quantitative study of the large-scale organization of polymers
in fibers has not yet actually been approached. Not only are the experi-
mental data merely qualitative (the observations of Heikins, Hermans,
and Weidinger [439] on the fiber G, moist "Fortisan," and those of Kratky,
Schauenstein, and Sekora [470] on natural silk), but also the theoretical
difficulties have not yet been fully resolved. The progress already
accomplished in the study of imperfect crystalline structures leads to the
hope that this new aspect of the structure of polymers, the importance
APPLICATIONS 187

of which cannot be questioned, will soon be taken up. In any event the
experimental techniques are already sufficient to undertake this kind
of work.

6.3. FINELY DISPERSED SOLIDS. CATALYSTS

\\re shall now consider the case in which the particle is no longer a
molecule but is instead a grain of matter of very small dimensions, as for
example those found in colloidal suspensions or in extremely fine powders
such as certain catalysts. The problem is to carry out a granulometric
measurement with X-rays when the grains are too small to be studied by
other methods. The particle dimensions most favorable to an X-ray
study are from several tens to a thousand angstroms. This upper limit
can certainly be exceeded with X-rays, but the experimental difficulties
involved in a measurement of a dimension of, let us say, 5000 A (0.50 µ)
are not justified in practice, since other methods are much simpler for
particles of this size.
The major difficulty limiting the effectiveness of X-ray measurements
is that in general the particles are not uniform in either size or form. Also,
if the sample is in the form of a powder, the particles are generally packed
closely against one another. This represents therefore the least favorable
case for the interpretation of small-angle scattering. Realizing the
complexity of the data, there is no hope of drawing from the single
scattering curve I(h) a complete description of the distribution of forms
and sizes of the particles being studied. Calculations such as those made
for molecules in concentrated solutions are not at all applicable.
Nevertheless, there is an important favorable element in this problem,
which is that the non-uniformity in the sizes of the particles diminishes
the influence of interferences even between nearest neighboring particles
(p. 70). It is therefore not unreasonable to assume that the observed
intensity is the sum of the intensities diffracted by the individual particles.
The first studies of fine powders based on theories for the low-angle
part of the scattering curve have been able to give only qualitative results.
There arc almost always large particles in the mixture for which these
theories are not valid, so that it is only rarely that an average radius of
gyration can be obtained which has real significance (§4.2.1). The
attempts at determining the distribution of grain sizes depend so greatly
on the initial hypotheses that the results should be applied only with
many precautions.
The methods of analysis based on theories for the tails of the scattering
curve (§4.2.3) seem capable of giving more interesting results. In no
case can a complete picture of the structure be given by small-angle
scattering experiments, but a parameter whose definition is quite precise,
188 SMALL-ANGLE SCATTERING OF X-RAYS

the specific surface of the specimen, can be successfully obtained without

resort to arbitrary hypotheses. Likewise, the radius of gyration can be
determined accurately and without arbitrary hypotheses from experi-
ments on dilute systems of particles of uniform size. Since the specific
surface is a very important property of fine powders, the methods of
analysis of §4.2.3 should give rise to very interesting applications. These
methods, however, a.re too new for a. sure judgment of their practical
value to be given.

log!

Fig. M. Log J(h1 ) for three samples of carbon black: 1. Standard

grade 6, channel black. 2. Sample P33; large particles formed by
thermal decomposition. 3. Carbolac l; channel black of extremely
fine particle size. (Biscoe and Warren (10].}

6.3.1. CARBONS
Severa.I varieties of carbon, both artificial and natural, are found in
very fine grain sizes. As a matter of fact, it was on these substances that
small-angle scattering was first observed and immediately attributed to
the state of division of the matter. The pattern of small-angle scattering
is very dependent on the type of carbon. Practically invisible for
anthracite, it is intense and contained within small angles for lampblack;
it becomes more or less enlarged for the different preparations of carbon
black, and, finally, it is very intense and extends over a large angular
region for certain active carbons. A typical example is given by the
comparison of three carbon blacks used as a charge in the rubber industry.
Figure 64 reproduces the curves obtained by Biscoe and Warren [10],
represented as curves of log /(h 2 ). The characteristic of each of these,
the general rule in studies of powders, is that the curves of log J(h 2 ) are not
linear but are curved in such a way that the slope increases with decreas-
ing scattering angle. The intensity incr(ll8.Ses so rapidly as the small
angles are approached that it is not possible to obtain the complete curve
APPLICATIONS 189

by photometry of a single film; a series of films of different exposures

must be made and their results combined. Two situations can occur.
As in curve 1, the variation in slope may become small enough near the
low-angle limit so that an extrapolation of the curve to zero-angle is
possible. This means that the largest particles in the sample do not have
a radius of gyration too large to escape the measurement (p. 149). The
slope at the origin then gives the average radius of gyration as defined in
§4.2.1 (R 0 = 210 A). In the other case (curves 2 and 3), the curvature
of the curve at small angles is such that an accure.te extrapolation cannot
be made. The average re.dius of gyration cannot be measured without
extending the measurements to smaller scattering angles. All that can
be said is that there is a lower limit to the average re.dius. The inter-
pretation of these curves can only be qualitative. The genera.I aspect of
the curves shows that sample 2 contains many large particles and only a.
few small ones. In sample 3 there is apparently a. mixture of large
particles with a large proportion of very small particles, the large particles
giving rise to the very steep initial part of the curve and the small particles
causing the only slightly inclined tail of the curve.
If the statistical methods described in §4.2.2 are applied, it is necessary
to make use of a very precise experimental curve extended down to very
small angles. It is necessary, furthermore, to apply corrections to this
curve for the effects of slit height ( §3.4.2), e.nd these corrections are not
negligible for curves of log I(h 2 ) that are very non-linear.

6.3.1.1. Practical Study of Carbon Blacks

Although it is difficult to obtain a rigorous description of the grains of a.
sample, the X-ray data can nevertheless have useful technical applications.
The scattering curve by itself characterizes the state of division of the
matter and often can be used to distinguish the various qualities of carbon
blacks and even to classify them qualitatively as to degree of fineness.
Such information is simple to obtain. The intensity scattered by carbons
is very high; it is not necessary to monochromatize the radiation, and the
Geiger-Muller counter measurements, which are preferable to photography,
are very rapid.
It has been reported [126'] that satisfactory results can be obtained
with a commercial spectrometer to which only very simple modifications
and adjustments have been made (it is necessary to use small slits and to
place the sample holder so that the powder sample, in the form of a plate
of appropriate thickness, is normal to the incident beam). The results of
the measurements are much more interesting if care has been taken to
insure that the scattering mass has been held constant for a series of
samples. To illustrate this, let us consider the following schematic
190 SMALL-ANGLE SCATTERING OF X-RAYS

example: one carbon black is made up entirely of fine particles, and a

second is made up of 50 per cent of the same particles and 50 per cent of
grains which are so large that they produce no scattering in the accessible
angular region. The two scattering curves will be superposable, but, if
the scattering masses are the same, the intensity from the second sample
will be only one-half the absolute value of that given by the first. There-
fore, if only fine particles are active particles, the quality of the sample
is directly characterized by the absolute value of the scattered intensity.
The proper execution of such an experiment requires that the X-ray
tube input and the geometry of the scattering apparatus be held constant.
The constancy of the scattering mass, also necessary, can be easily main-
tained, as, for example, by the following technique: a fixed mass of
powder is placed in a small cylindrically shaped cup with a mica bottom.
The powder is spread into a uniform layer and then compressed by a
cover that also has a mica window. In this manner the mass per unit
surface will be the same for all the samples, and, since the cross section of
the beam is invariant, the scattering mass is then constant from one
experiment to the next.

6.3.1.2. Structure of Different Varieties of Carbon

Brusset [13], [18] has used small-angle scattering to compare natural
carbons of various origins anrl has shown that they are differentiable
from the point of view of their granular structure. Riley [141] found an
anthracite from \Vales that gave a very singular pattern containing a
ring of scattering. IWey compared this pattern to those of liquids and
suggested that in the carbon there were particles in contact with one
another. Let us point out that analogous patterns have Leen founcl for
age-hardening alloys (p. 204) and that a different interpretation has been
given them. This interpretation involved particles which were isolated
but showed internal density variations. It would Le interesting to see
whether an analogous explanation could be valid for these carbons.
R. E. Franklin [56'], [56"J has studied the structure of carbon blacks
obtained by carbonization of organic substances. She obtained an atomic
distribution function such as is obtained for liquids by a :Fourier inversion
of the complete intensity curve, including Loth the small-angle scattering
and the high-angle diffraction maxima. In this way she determined at the
same time both the structure of the carbon and the dimensions of the
particles. These particles were for the most part isolated layers of the
graphite ;;tructurc, the particle produce(l by pyrolysi;; of polyvinylidene
chloride at 1000° C. having a diameter of only 16 A. Tlw size of the
particles increased with the temperature of carbonization. Franklin also
used a Fourier transformation to interpret only the curve of small-angle
APPLICATIONS 191

scattering. However, this transform, analogous to the characteristic

function of Porod [137] (§2.4.3), does not have a simple physical signifi-
cance because of the non-uniformity of the particles and because of their
interactions.
Another interesting example is that of activated charcoal. Cocoanut
charcoal, before activation treatment, gives only a rather limited and
weak, low-angle scattering. However, after activation the scattering is
extremely intense and extends out to angles of several degrees. Actually,
this is the substance that gives the strongest low-angle scattering observed.
log!

Fig. 65. Log J(h') for activated charcoals: (a) activated cocoanut
charcoal; (b) and (c) charcoal activated by two different treatments.
(Brusset [13].)

The curve oflog /(h 2 ) for this sample shows a very pronounced curvature
followed by a long linear part of small slope (Fig. 65). It might then be
said that the sample is made up of a mixture of particles of very different
sizes, the smallest particles having a radius of gyration not larger than
10 A. Yet, given the method of activation of the charcoal, this model
is not satisfactory, for it is known that the activation does not have the
effect of producing a fine powder but rather that it produces cavities in
the interior of a grain by dissociation of the tarry substances that fill it.
The active surface is thus increased, which gives large absorbing properties
to the charcoal.
The model of isolated material particles is, therefore, completely
inadequate in this problem. We could more validly adopt the opposite
point of view, the small holes in the interior of grains of large dimensions
playing the role of small scattering particles. The important factor in
192 SMALL-ANGLE SCATTERING OF X-RAYS

the scattering is the difference in density between the particle and the
surrounding medium; two complementary objects (Fig. 66) in which the
holes of one correspond to the solid parts of the other give exactly the
same scattering pattern in the region accessible to experiment ( §2.2.2.2).
The 10 A radius of gyration would accordingly correspond to the smallest
pores of the active charcoal.
It would certainly be quite interesting to undertake measurements of
the specific surface of carbons by the method of §4.2.3. This would be
a precise and probably convenient method for characterizing the porosity
of a carbon.

Fig. 66. A sketch of the structure of two complemente.ry objects.

6.3.2. CATALYSTS
This is a domain in which the applications of small-angle s~attering are
the most immediate since the granulometry of a catalyst is an important
factor in the value of the product.
Pulverized catalytic nickel is obtained by various processes. It is
known that Raney nickel is more active than Sabatier nickel, and small-
angle scattering shows immediately that the first is made up of much
smaller grains than the second [65].
The difficulties of interpretation already pointed out for the carbons are
found again for the catalysts. Nevertheless, determinations of the
statistical distribution of grain sizes have been successfully made. Elkin,
Shull, and Roess [41] applied the method of §4.2.2 to desiccated silica and
alumina gels as well as to oxides of nickel and iron. Then, from the
measured distribution of grain sizes, they calculated the specific surface
of the catalyst and obtained good agreement with the result given by
classical gas adsorption methods. This very interesting result gives some
confidence in the hypotheses used in the calculations and shows that at
least in certain cases a precise analysis of the patterns deserves to be made.
As has been done for the carbon blacks, certain laboratories are
beginning to use small-angle scattering to characterize a catalyst from a
technical standpoint. The following example makes the most of the
advantages of this method.
. APPLICATIONS 193

Van Nordstrand and Hach [126'] showed that the form of the scattering
curve was progressively modified as an alumina catalyst was subjected to
heating at higher and higher temperatures, the small particles disappear-
ing by partial fusion. If a catalyst has partially lost its activity after
3000

A
A. Harshaw Uvergel
B. Alcoa-1000°F.
1000 C. Alorco >.<" Balls
D. Alorco R2200

100

~
c"'
.$
c:

10

lLL.LJ.~'--~~~~--'-~~~~~~~~-'-~~~~~~~-,-J
100 10 1 0.1
Square of angle, radians 2 x 10 4
Fig. 67. Log I vs. log h' for different samples of alumina catalysts.
The dotted line shows the theoretical 1i-• dependence. (Van Nordstrand
and Hach [126'].)

use, this can arise from two causes: either a part of the product has been
rendered completely inactive, the rest still being good, or else the whole
catalyst has been slightly modified. The remedies to be considered are
different in the two cases. Small-angle scattering will allow a sure
diagnosis: in the first, the form of the curve will not have varied, hut its
absolute intensity will have decreased; in the second, the form of the
194 SMALL.ANGLE SCATTERING OF X-RAYS

curve is modified. The same authors found empirically that the intensity
Sl'attered at rather large angles varied as h- 4 (Fig. 67) and that in this
region the intenHity at a given angle for a given mass of catalyst was
proportional to the specific surface of the catalyst as determined by the
nitrogen adsorption method (Fig. 68). This is an excellent verification
of Porod's theory (§2.4.3), though it was apparently unknown to these
authors. After a preliminary calibration a measurement of the specific
surface could be made with X-rays in five minutes with a very simple

>- 400
~
I
x
~
"'e
<( 200

200 400 600 800

Area by N2, m.2/gm.
Fig. 68. Specific surface areas as measured by X-rays and by nitrogen
adsorption. (Van Nordstrand and Hach [126'].)

apparatus such as a commercial Geiger counter spectrometer (p. 95).

They found also that the measurement of small-angle scattering curves
could sometimes advantageously replace the determination of gaseous
adsorption-desorption isotherms.

6.3.3. COLLOIDAL SOLUTIONS

Colloidal micelles generally are of a size favorable to study by X-rays,
and when colloids of high atomic weight, such as certain metals, are
considered, the very high intensity of the small-angle scattering makes
the measurement easy to carry out.
Fournet [48] studied a solution of colloidal silver and obtained a curve
of log /(h 2 ) which had a long linear part; from this the radius of gyration
could be easily determined, the result being 50 A. The same colloid after
flocculation gave a curve of log /(h 2 ), which could be superposed almost
exactly with the curve of the first solution except at very small angles.
This shows that, although flocculation has caused a drawing together of
APPLICATIONS 195

the particles, only a small percentage of them at the most have clustered
into larger units. This experiment shows also that it is legitimate to apply
the theory of widely separated particles to systems of a rather high density
if the particles of the system are only slightly non-uniform.
The results of this experiment were confirmed by a measurement with
the electron microscope. The micelles appeared as approximately
spherical particles with a rather uniform diameter of about 120 A. The
sphere of radius of gyration of 50 A has a diameter of 130 A.
Porod [137] found that colloidal molybdic acid in the dry state gave a
different pattern from that of the solution, the pattern showing a ring
of scattering r1tther than the pattern of isolated particles. The difference
in the behavior of this colloid as compared to silver may be associated
with the fact that, whereas flocculation of silver is irreversible, it is
reversible for molybdic acid.

6.4. SUBMICROSCOPIC HETEROGENEITIES IN SOLIDS. APPLICATIONS

TO PHYSICAL METALLURGY
The existence of small-angle scattering is in a very general way charac-
teristic of the existence of submicroscopic heterogeneities in the scatterer.
The example in which the scatterer is composed of small, separate particles
is only a particular case of which we have seen numerous applications
(solutions, suspensions, and powders). A continuous solid can also give
rise to low-angle scattering if its electronic density shows fluctuations, the
particular conditions having been analyzed in §2.4.l. We have already
mentioned several examples in which the sample was not composed of
distinct particles (certain fibers of high polymers). We shall now con-
sider certain metals and alloys that offer interesting examples of hetero-
geneities. Often these heterogeneities have been disclosed only by the
phenomenon of small-angle scattering.

6.4.1. HETEROGENEITIES IN PURE METALS

We shall not consider here either metals in the form of fine powders or
metals in the form of colloidal grains, since these problems have already
been discussed.
Several experiments (Blin and Guinier [11']; Blin (1954); Hayes and
Smoluchowski [74'"]) have shown that a piece of metal that to the eye or
even to the microscope is pure, sound, and homogeneous can give rise to
small-angle scattering. Since this is very weak in intensity, its study
requires a well-designed apparatus from which all parasitic scattering has
been eliminated. A Geiger-Muller counter must be used as the detector,
since photographic techniques are not of high enough sensitivity. The
196 SMALL-ANGLE SCATTERING OF X-RAYS

sample should be in the form of a thin plate whose thickness is approxi-

mately the optimum thickness for transmission experiments (p. 88), that
is, of the order of some hundredths of a millimeter.
This scattering appears when the metal has been severely cold-worked,
particularly by tensile elongation. It has been observed on nickel,
copper, aluminum, and zinc. The scattering is of the continuous type,
the intensity decreasing with increasing distance from the center of the
pattern. The curve traced with the usual coordinates (log J(h 2 )) shows a.
rather long linear part that corresponds to a radius of gyration of the
order of 6 A. When the cold-worked metal is annealed, the scattering at
small angles is not modified until recovery has been completed. Then
the scattered intensity becomes stronger at very small angles and decreases
much more rapidly with increasing scattering angle, so that the curve of
log /(h.2) corresponds to a much larger radius of gyration.
Thus far we have not tried to interpret these observations in terms of
the structure of the sample. The values of the radii of gyration serve
only to summarize the facts in a rather convenient form that allows us
to contrive probable models of the structure. Lattice defects, such as
dislocations, introduced by deformation of the metal, produce at the most
only very small variations in density, and these a.re certainly so localized
that they cannot be invoked to explain this scattering phenomenon.
Moreover, such defects produce a broadening of the Debye-Scherrer lines,
and actually there is no positive correlation between this broadening and
the observed scattering at small angles. Vacancies or missing atoms in
the lattice also correspond to a radius of gyration that is too small. We
are therefore led to the assumption that the heterogeneities found in a
deformed metal are submicroscopic cavities formed by the coalescence of
vacancies, these cavities, if they are spherical, having a diameter of the
order of 15 A. It is natural to assume that the form of the cavities would
depend on the directions of the stresses experienced by the metal. Their
shapes, therefore, would not be spherical, but it is difficult to describe them
more precisely.
These cavities are not sensitive to anneals at low temperatures at which
the atomic mobility is small, since the lattice could then become more
perfect only by local rearrangement of the atoms. At high temperatures,
at the point at which the metal is transformed by the appearance of new
recrystallized crystals, these cavities either disappear or are eventually
fused together to give a smaller number of much larger cavities retained
in the recrystallized metal.
The hypothesis of the existence of cavities in the deformed metal is
confirmed by the slightly smaller density of the metal after cold-working.
However, such measurements of density are very critical and are not
APPLICATIONS 197

sufficiently precise, so that apparently this new aspect of imperfections

in metals can be more easily studied with X-rays.

6.4.2. HETEROGENEITIFS IN SOLID SOLUTIONS

The schematic structure of a solid solution is the following: the atoms
of the various constituent metals are distributed on the lattice points of a
single lattice. Such a solution is certainly homogeneous on a microscopic
scale but is not so on an atomic scale. The heterogeneities can become
important if the distribution of atoms is not perfectly random, as, for
example, if the atoms of one type tend to agglomerate into clusters. In
such circumstances, moreover, the predominance of one type of atom in a
cluster can lead to a deformation of its lattice, the atoms being displaced
more or less from the lattice points of the average lattice of the solid
solution. These irregularities or defects in periodicity, arising both
from the nature of the atoms and from their positions, modify the
crystalline reflections and cause abnormal diffuse scattering outside the
positions of the Bragg reflections. In particular, small-angle scattering
can appear.
The study of the real structure of a solid solution requires a knowledge
of the complete pattern of scattering and diffraction and not solely a
knowledge of the small-angle region of the pattern. Nevertheless, in the
following paragraphs we shall consider only the consequences of small-
angle scattering in order to show how simply this furnishes conditions
that have to be satisfied by the structure models contrived to explain the
total pattern. It is therefore essential that the small-angle part of the
X-ray pattern be not neglected in any such study. It is impossible to
explore the very-small-angle region with standard large-angle diffraction
apparatus, and consequently investigators who have limited themselves
to such equipment have often been led to propose structure models that
are in contradiction with the results drawn from the simple appearance
of the small-angle scattering patterns.

6.4.2.1. Equilibrium Solid Solutions

When an alloy is at a temperature such that the solid solution is the
only equilibrium phase and at which there is no ordered state, as, Jor
example, all dilute solid solutions, it was thought for a long time that the
equilibrium state of the solid solution was one of perfect disorder. More
precisely stated, this means that the nature of an atom has no influence
at all on the nature of the atoms that occupy neighboring sites in the
lattice.
The number of atoms of one type in a volume containing a given number
of lattice sites will undergo statistical fluctuations about the mean value
198 SMALL-ANGLE SCATTERING OF X-RAYS

required by the composition of the alloy. There is a certain probability,

therefore, that a group of atoms of one type can be accidentally produced
on a group of neighboring sites. This probability can be easily calculated,
and it of course rapidly becomes very small when the number of sites in
the group increases. These fluctuations of composition correspond to
local fluctuations of electronic density and thus cause a scattering at small
angles.
A simple calculation shows that this scattering is independent of the
scattering angle. 1 If the sample is a binary alloy, AB, containing p
I

2· 211
Fig. 69. Scattered intensity from an equilibrium solid solution of
Al-Ag (20 per cent Ag by weight) at two temperatures. (Walker,
Blin, and Guinier [169].)

atomic per cent of atoms of type B, the scattered intensity per atom of the
solid solution is: I = I.p( 1 - p) (f A - f nl 2 , where f A and f B designate
the scattering factors of the two types of atoms, equal to the atomic
numbers at zero angle, and where I. is the intensity scattered by one
electron. This formula, first deduced by Laue, is a particular application
of the general relation, equation 2.72.
The intensity of this scattering is very weak. Even in the favorable
circumstance of an equiatomic mixture of atoms of very different atomic
numbers, it is just at the limit of the intensity measurable with a very
sensitive Geiger-Miiller counter-equipped apparatus.
Walker, Blin, and Guinier [169] measured the scattering from an Al-Ag
alloy containing 20 weight per cent Ag at a temperature at which the
solid solution is in equilibrium (T > 450° C.). The scattered intensity
shows a maximum at zero-angle and is considerably stronger than the
intensity predicted by Laue's equation. The maximum in the scatter-
ing at zero-angle becomes less pronounced at higher temperatures (Fig.
69). These experiments show that the atoms are not distributed com-
pletely at random. The atoms of silver have a tendency to bunch
1 Aside from the slow decrease with angle due to the variation of the atomic

scattering factors.
APPLICATIONS 199

together and form very small nuclei rich in silver. As we have already
stated, such nuclei can be produced in a perfectly disordered crystal by
statistical fluctuations, but the experiment proves that there are more
such nuclei than would be predicted by the statistics of an ideally dis-
ordered state. These results are in agreement with the theoretical
discussion of binary solid solutions given by Fournet (1953a). The very
slow decrease of the scattered intensity with increasing angle allows the
inference that the nuclei are very small.
The value of a radius of gyration would not have much meaning, since
these nuclei are certainly non-uniform both in size and shape. There is a
special method for interpreting the experimental data in this case which
was first used to determine the degree of short-range order in solid solutions.
By means of a Fourier transformation of the experimental curve we can
determine the probability of finding two atoms of silver at distances
equal to those of first neighbors, second neighbors, etc., in the lattice.
This is related to the function y(r) introduced in §2.4.3. This method,
however, requires a knowledge not simply of the scattering curve at small
angles but of the total scattering curve. Another method consists of
calculating the function I(h) theoretically and comparing this with the
experimental curve (Fournet (1953b)). The complete solution of the
problem thus extends beyond our particular subject. Nevertheless,
from the small-angle scattering we can obtain qualitative proof of the
existence of clusters of atoms, and it can be immediately seen that these
clusters can only be of small size.

6.4.2.2. Supersaturated Solid Solutions: Age-Hardening

The heterogeneities that we have discussed so far have been rather
small and, consequently, have produced a small-angle scattering that is
very weak in intensity and difficult to study. Intense scattering
phenomena have been observed, however, for certain supersaturated
solid solutions, that is, solutions that are out of the region of single-phase
equilibrium.
Let us consider a binary solid solution, AB, for which the limit of
solubility of B in A increases with temperature. The solid solution at a
given composition is in equilibrium above a temperature T 0 (Fig. 70);
it is in this region of temperature that the experiments described in the
preceding section have been made. When this solid solution is brought
to room temperature by quenching, it is no longer in the equilibrium
state; indeed, the room-temperature equilibrium state consists of two
phases, a solid solution r:t. of normal concentration and a {J phase in which
the excess dissolved B atoms are assembled.
By thermal treatments at fairly low temperatures (T < T 0 ) the solid
200 SMALL-ANGLE SCATTERING OF X-RAYS

solution can be made to evolve towards the stable state. Generally the
most stable state, corresponding to the precipitation of the {J phase, is
not produced directly. The precipitation is preceded by intermediate
stages in which the excess atoms are clustered together but do not yet
form the true final precipitate. These structural transformations are
shown by a betterment of the mechanical properties of the alloy called
age-hardening.
The heterogeneities of the hardened alloy are too small in size to be
visible to the microscope, but they can cause observable small-angle

T
Homogeneous
solid solution

a phase a+ {3 phases

Temperatures
} of annealing

o.__~~~~~-'-~~~~~~~__..
Pure %B
A
Fig. 70. Schematic phase diagram of an age-hardening alloy.

scattering. This scattering plays a fundamental role in the study of

age-hardening, a phenomenon that is not only of scientific interest in the
development of the physics of metals but also of considerable technical
interest, in that it affects many applications of light alloys. As we have
already pointed out, it is somewhat artificial to isolate the study of
scattering at small angles from the rest of the pattern. However, we
shall stress particularly the direct information that the . small-a.ngle
scattering can provide, especially when the simplicity of the phenomena
makes the method useful for technical applications.

6.4.2.3. Structural Characteristics Directly Related to the Small-Angle

Scattering
The difficulty in studies of the structure of hardened alloys a.rises from
the simultaneous presence of lattice deformations and local changes in
composition produced by the clustering of the dissolved atoms. It is for
APPLICATIONS 201

precisely such a condition that' an interpretation is facilitated by the

consideration of small-angle phenomena. This follows from two prop-
erties of small-angle scattering which, in view of their importance in
this matter, we shall recall at this time.
J . 8mall-angle scattering depends primarily on heterogeneities of com-
position and little, if at all, on deformations of the lattice. A schematic
example will demonstrate this.
Let us suppose that in the lattic.e of a solid solution the dissolved B
atoms assemble in irregularly arranged small clusters or zones without

III
H--\~
11'frl
.l

ll 1-t-1 Iili
I
(a) (b)

Fig. 71. (a) Structure of a zone without lattice deformation. Foreign

atoms have clustered on the matrix lattice sites. (b) Inclusion in the
matrix lattice of a grain with a different lattice but with the same
composition.

producing any deformation of the general lattice (Fig. 71). The local
heterogeneities, in which the electronic density is different from that of
the surrounding medium, will cause small-angle scattering similar to that
produced by sm~ll, isolated particles of the same exterior form. But, in
addition, the small particles in this case are small crystals; they will
therefore produce high-angle diffraction spots also, these being somewhat
enlarged as a result of the small particle size. It can be shown ( §5.1) that
each point in the reciprocal lattice, including the origin or (000) point, is
replaced by a small diffracting region centered on each point. This is the
explanation of the broadening of Bragg diffraction spots. The essential
conclusion to be drawn from this calculation is that the regions around
each of the points are identical, any one being derivable from any other
by a simple translation. Thus, if the alloy shows heterogeneity without
deformation, the study of the scattering in the neighborhood of the center
gives all the information that can be furnished by X-ray diffraction.
Now let us consider as a second example the opposite case, deformation
in a lattice of uniform composition. Let us suppose that the lattice of a
homogeneous solid solution includes regions in which the lattice has
202 SMALL-ANGLE SCATTERING OF X-RAYS

slightly different parameters from those of the matrix. The fluctuations

of the parameters are necessarily small, since metals always have a com-
pact structure, so that the density of the medium can undergo only very
small variations. Therefore, since the whole metal has a quasi-constant
density, there is no scattering at the center of the pattern. On the other
hand, the Bragg angle of a given high-angle reflection will vary for the
regions with different lattices, the variations in parameters producing an
effect which is greater, the larger the Bragg angle of the reflection (the
higher the indices of the reflection). The primary reciprocal lattice
points will be surrounded by a region of diffraction whose extent will
increase in proportion to the distance of the point from the center. The
center itself is not surrounded by any scattering region.
Another type of deformation that is often found in face-centered cubic
metallic structures is the slippage of (111) planes over one another. This
causes the formation of more or less extended regions with a hexagonal
close-packed structure which are called stacking faults (Barrett (1952),
p. 259). These stacking faults produce regions of scattering around
certain reciprocal lattice points that are elongated along the [111] lines of
the reciprocal lattice normal to the (111) planes showing the faults.
There is no scattering around the center, since, regardless of the mode of
stacking of the (111) planes, the density remains rigorously constant at all
points of the metal lattice.
2. The two simple examples discussed above form extreme cases
between which are those structures generally found in practice, where a
variation in composition is connected with a deformation of the lattice,
either in that the B atoms have a different diameter from the A, or in that
there is a tendency of the cluster to form a crystalline structure different
from that of the original lattice. The distribution of the regions of
scattering in reciprocal space can then become very complex, but to a
first approximation the scattering in the neighborhood of the center.depends
only on the exterior surface of the heterogeneous regions and not on their
internal structure. Hence, if in the example of Fig. 7lb we suppose that
the atoms of the zones are B atoms or at least a mixture containing a large
proportion of B, the small-angle scattering phenomena will be exaetly the
same as for the case of Fig. 7la. The variations of the parameters
will have no influence on this scattering.
In order that the small-angle scattering be intense enough to he
detectable, it is necessary that the variations in electronic density be large
and, consequently, that there be a large difference between the atomic
numbers of the alloying elements. When these elements are adjacpnt or
near neighbors in the periodic table, the method is not applicable. Thi;; is
true for alloys sueh as Al-Mg or Al-Si, and also for Cu-Zn ancl Ni-Fe.
APPLICATIONS 203

6.4.3. EXAMPLES OF SMALL-ANGLE SCATIERING BY

AGE-HARDENING ALLOYS
It is found that, in the different systems of age-hardening alloys that
have been studied, the small-angle scattering patterns present quite
varied appearances, showing that the process of precipitation varies
according to the nature of the atoms. We shall give examples correspond-
ing to two quite distinct types, stressing only the manner in which the
X-ray technique is utilized, and not the results pertinent to the field of
physical metallurgy.

6.4.3.1. Aluminum-Silver Alloy: First Stage of Hardening

The metal is an alloy containing 20 weight per cent of silver. After

an anneal at a temperature greater than 450° C. (the temperature at
which the alloy becomes homogeneous) the sample is brought to room
temperature by a rapid quench and then studied after anneals of various
lengths at temperatures of the order of 50 to 200° C.
The sample is prepared in the form of a thin foil of suitable thickness
(0.04 mm. for Cu Koc radiation). It is preferable, as we have already
pointed out, to begin the investigation with the use of photographic
techniques, since it is not possible to predict a priori the appearance of the
low-angle scattering in these cases.
Two facts can be drawn from the first observations ([l 70], Walker and
Guinier (1953), Belbeoch and Guinier (1954)):
1. The patterns are the same regardless of the state of crystallization
of the metal, whether it be in a microcrystallized state (cold-worked) or
whether it be a single crystal of arbitrary orientation.
2. The scattering does not decrease continuously from a maximum at
zero-angle (Fig. 72). Instead, when the height of the beam has been
reduced sufficiently, the pattern has the form of a rather sharply defined
ring whose diameter varies according to the thermal treatment given the
sample. The mean scattering angle of the ring corresponds by Bragg's
law to a distance of the order of magnitude of 50 A. We have already
called attention to other low-angle scattering patterns that contained a
ring of scattering ( §4.1.2.2), but the ring is more sharply defined in this
case, and, more important, the intensity of the scattering inside the ring
is very weak near the center.
It is obvious that one must try to interpret these observations in terms
of the formation of clusters of silver atoms. Since the orient~tion of the
crystal has no effect on the appearance of the pattern, it can be concluded
that these clusters are isotropic and have a symmetry which at least on the
average is rigorously spherieal. But, in order that the scattering not be a
204 SMALL-ANGLE SCATTERING OF X-RAYS

maximum at zero-angle, it would be necessary the.t the particles show

a very strong interaction between themselves, analogous to the inter-
action between the molecules of a liquid. This we.a the interpretation
advanced by Riley rt41 l for the ring observed on the small-angle patterns

(a)

s 21r
(b)
Fig. 72. (a) Small-angle scattering from Al-Ag (20 per cent Ag by
weight) water-quenched from 520° C. (b) A graph of intensity vs.
scattering angle for this sample.

of certain carbons. However, this model of a structure does not seem

very probable for an alloy. A ring of such sharpness could only come
from close-packed particles of a very uniform size. How can this
regularity be justified?. In addition, when the alloy is annealed, the ring
remains sharply defined and contracts. This would mean that the
clusters become uniformly larger and change into another close-packed
arrangement in which the interparticle distances are somewhat greater.
Such a transformation is difficult to imagine.
Another model, which is more acceptable, has been proposed by Walker
and Guinier (1953). It is assumed that there are independent particles
that do not give rise to interparticle interferences, and that the single
APPLICATIONS 205

particle has such an internal structure as to give a ring-type pattern

instead of the classical continuous scattering curves. We have already
called attention to an analogous interpretation for hemocyanine (p. 17 4).
Actually, if the particles we considered in Chapter 2 have continually
decreasing scattering curves, it is because their structure can be assumed
to be of effectively uniform depsity. The scattering curve can become
much more complex, however, if there are regions of different electronic
density in the interior of each particle.
·when the silver atoms cluster around a particular point in the solid
solution, they migrate by diffusion, but as the annealing temperature is
not very high the atoms are not very mobile. It is therefore easy to
imagine. that the silver atoms clustered into a nucleus leave a shell-like
region emptied of silver atoms. Thus the schematic particle would be
made up of a spherical nucleus of high electronic density p1 surrounded by
a spherical shell of a density p 2 that is less than the average density Po of
the medium in which the particle is immersed. This concept requires
the supplementary condition that f p dv = Po V, since all the atomic
movements have taken place at the interior of the total volume V of
the particle.
Let us consider the following simple model: the two parts of the particle
are defined by two concentric spheres of radii R 1 and R 2 . The condition
relating Pv p 2 , and p 0 is:

or

In a calculation of the scattering, the particle can be replaced by a set of

two superposed concentric spheres, one of radius R 1 and density (p1 - p 2),
and the other of radius R 2 and negative density, -(p 0 - p2 ). The
scattered intensity will be the square of the algebraic sum of the amplitudes
scattered by these two spheres. The amplitudes at zero-angle are equal
and opposite, as is seen from equation 2.10, but the scattering due to the
small sphere R1 decreases more slowly with angle than that of the large
sphere R 2 • The curve of the intensity, therefore, starts from zero at
zero-angle, and it obviously again becomes zero at large angles, so that it
shows a maximum quite similar to the one found experimentally (Fig. 73).
This model is obviously arbitrarily chosen, but a better model of a
spherically symmetric, heterogeneous particle can be obtained by
quantitatively determining the electronic density distribution p(r) that
would produce the observed scattering curve.
206 SMALL-ANGLE SCATTERING OF X-RAYS

Equation 2.13 gives the scattPrcd intensity as

'° [rsin hr
l(h) = 1,(h) Jo p(r)-,;;-- 47Tr2 dr
] 2

A Amplitude

h
(a)

h
(b)

Fig. 73. (a) The amplitude of scattering for two spheres: A,, a.
sphere of radiuR R 1 and density (p 1 - p 2 ); A,. a sphere of radius R 0
and density ( p 0 - p 2 ). (b) Scattered intensity for the concentric spheres
model: I= (A 1 - A 2 ) 2•

where h = (47T sil1 ())/A. The density p(r) can be obtained from the
experimental curve l(h) by a Fourier transformation:

27T r o
00
p(r) = - I 2- 1
h Jl(h\
- - ' sin hr dh
le(h)
A second method of approach, which does not require the assumption
of spherically symmetric particles, is to use equation 2.21 and the Fourier
transform of hl(h),
ry 0 (r) oc l 00
hl(h) sin hr dh
APPI.ICATIONS 207

If J(h) is normalized so that y 0 (0) = 1, then y 0 (r) is related to the prob-

ability of finding an atom of silver a distance r away from another atom
of silver (Fig. 74).
In reality the distribution of matter in the interior of the particles is not
continuorn~. The aluminum and silver atoms are situated on sites of the

1.0

Fig. 74. The probability of finding an Ag atom at a distance r from

another Ag atom in the quenched AI-Ag alloy.

known lattice, and the problem is to find their statistical distribution.

The solution to this problem given by Cowley (1950) leads to similar
equations.
Theoretically, the same results about the structure of the solid solution
could be obtained by using the phenomena of the scattering around any
Bragg reflection of the crystal, but it is much easier experimentally to use
the low-angle part of the pattern. Tne principal advantage of this is that
the metal can be used directly regardless of its state of crystallization.
The method can thus be applied to cold-worked metals or to metals in
208 SMALL-ANGLE SCATTERING OF X-RAYS

the form of very fine crystals such as are used in metallurgical techniques.
A determination of the structure of a hardening alloy of this type can be
made even on samples usually employed for other physical or mechanical
tests.
6.4.3.2. Aluminum-Silver Alloy: Second Stage of Hardening
When the same Al-Ag alloy is given anneals at higher and higher
temperatures, the ring decreases in size until only an intense blur at very
small angles is observed; then finally a completely different phenomenon

Fig. 75. Small-angle scattering from Al-Ag, water-quenched from

520° C. and annealed 10 days at 140° C.

appears. When the sample is polycrystalline, with not too fine a crystal
size, the pattern contains several streaks of scattering, the direction and
number depending on the orientation of the crystals (Fig. 75). A dis-
placement or rotation of the sample will modify the number and orienta-
tion of the observed streaks. A study of this effect should be made with
a monocrystal whose axes can be oriented at will with respect to the
incident beam.
The object of the experiment is to find the form and extent of the
scattering regions around the center of reciprocal space and to determine
their orientations with respect to the axes of the reciprocal lattice of the
solid solution (this is a face-centered cubic crystal with parameters
practically equal to those of aluminum). The result of the study is shown
in Fig. 76. The regions ofscattering are directed along [Ill] axes. They
are very narrow, and their length depends on the thermal treatment of
the alloy. More exactly, the intensity decreases with increasing angle,
the scattering for Cu Ka. being perceptible up to scattering angles of the
order of 2° or 3° for the longest streaks.
The observations on the polycrystalline samples mentioned above are
then explained in the following manner: The reciprocal space of the
sample contains a number of scattering regions in the form of straight
lines centered on the origin, the number being equal to 4 times the number
of crystals irradiated by the incident beam. It is known that the pattern
APPLICATIONS 209

for a given position of the sample can be depicted by the cut through
reciprocal space of the Ewald sphere of reflection. Near the center of
the pattern this spherical surface can be replaced by the plane normal to
the direction of the incident beam. If there are a sufficient number of
crystals and if these are randomly oriented, there are always a certain
number of scattering regions that are very near this plane normal to the
beam. In order that a streak may appear, it is necessary theoretically
that the scattering regions be in this plane. In practice, as a result of the
divergence of the rays of the primary beam, it is sufficient that these

(001]
[111]

(010]

/
/ [l l l]
/
/
/

[100] [ 11 l]
Fig. 76. The representation of the observed streaks in the reciprocal
lattice of the matrix Al-Ag crystal.

domains be close to the plane. A certain number of streaks will be

observed whose directions cannot be predicted a priori, since the orienta-
tions of the crystals are unknown. If the number of irradiated crystals is
too large, as in a cold-worked material, the streaks will be very numerous
and very weak, since the individual crystals will be quite small. As a
result the streaks will merge to form a continuous scattering that is
spread out but too weak to be detectable by photography.
Let us return to the consideration of a single crystal. What are the
particles that can explain the small-angle scattering in reciprocal space
shown in Fig. 76. The anisotropy of the pattern proves that, unlike the
particles in the first stage, these are anisotropic and are oriented with
210 SMALL-ANGLE SCATTERING OF X-RAYS

respect to the matrix crystal. In §2.1.3.4 and §4.1.1.2, it was shown that
a wide, thin platelet gives a pattern of scattering that is elongated along
the normal to the platelet. The data of Fig. 76 are then explained by
supposing that the matrix crystal contains four series of particles in the
form of platelets parallel to the four sets of (Ill) planes. This is a par-
ticula:dy acceptable supposition, since the precipitates visible to the
microscope at a later stage also have the form of platelets parallel to the
(111) planes (Widmanstatten structure) (Barrett (1952), p. 542).
The diameter of the pl,atelet determines the thickness of the scattering
region, or the width of the streak on the film. This measured width
appears to be determined uniquely by the experimental conditions, such
as crystal dimensions and beam dimensions. Consequently all that can
be said is that the diameter of the platelet is at least of the order of a
thousand angstroms. The thickness of the platelet determines the length
of the streak. If the curve of the variation of intensity along the streak
were known, theoretically we could apply equation 2.44 to determine a
mean inertial distance with respect to the plane of the platelet. It is
more simple and satisfactory to evaluate the length of the streak, e, in
angular units and then to obtain the thickness of the platelet by means
of the following reasoning: The scattering by a thin, wide platelet of
thickness d will become zero at an angle equal to e = Afd. Therefore the
order of magnitude of the thickness of the platelet is d = )./e. In the
case described above, the thickn'ess is determined in this way to be
.-30 to 50 A. The platelets become thicker as the annealing continues.
Here again the consideration of the small-angle scattering has given
no information about the internal structure of these platelets. The
large-angle patterns show that these are grains of a precipitate with a
hexagonal structure in which the lattice is derived from that of the solid
solution by a slippage of (lll) planes over one another. The diffraction
spots of these precipitates are surrounded by regions of scattering that
are also directed along the [Ill] axes. These regions are much longer
than the region around the center of the pattern. In order to explain
these high-angle scattering phenomena we must suppose that the platelets
of precipitate parallel to (111) planes do not have a thickness of more than
10 A, a figure that is in contradiction with the small-angle scattering data.
The correct interpretation, which explains all the observations, is that
the platelets have the dimensions determined from the small-angle
scattering but that their crystalline lattice is not perfect. The principal
cause of the high-angle streaks of scattering along the (Ill) directions
is the presence of "stacking faults" in the precipitates (Barrett (1952)).
Although the complete study required tha~ the investigation be carried
out on a single crystal, let us observe that, once the facts just established
APPLICATIONS 211

are known, a simple small-angle scattering experiment on a metal without

special preparation allows us to determine the actual state of the alloy, to
distinguish between the two stages of hardening, and to obtain in a simple
manner quantitative results on the dimensions of the clusters or sub-
microscopic grains of precipitate.

6.4.3.3. Aluminum-Copper Alloy

Several other systems give phenomena analogous to those of the
aluminum-silver system, but this is not a general precipitation process.
Frequently the low-angle scattering patterns are very different in appear-
ance, and often they are less easy to measure and interpret than in the
example we have just discussed.
An aluminum-copper age-hardening alloy in not too fine a crystalline
form gives a pattern of streaks coming from the center even in the first
stage of hardening [64 ], [274]. These streaks are directed along the
[100] axes of the crystal of the solid solution; they can be very long
(from 10° to 15° with Cu Krt. radiation) and they are ofrather low intensity.
In order to study them in practice, it is necessary to work with a single
crystal. 'Ve can deduce from the form and orientation of these streaks
that the heterogeneous regions are very thin platelets parallel to the three sets
of (JOO) planes. The simplest region that could be imagined would be
formed of clusters of copper atoms on one or two (100) planes of the solid
solution lattice in zones having a diameter of several hundred angstroms.
Figure 77 reproduces the low-angle part of the pattern corresponding to
an alloy annealed at 100° C. for 100 hours and oriented so that the [100]
axis is parallel to the incident beam; the two other fourfold axes are
horizontal and vertical, respectively. The two branches of the cross
correspond to two of the three series of zones; the third, oriented normal
to the beam, should theoretically give a circularly symmetrical scattering,
but it is too limited in diameter to be visible. It is easy to calculate the
distribution of scattering from a thin platelet parallel to the direct beam. 1
We can thus obtain the dimensions of the platelet that gives a low-angle
scattering identical to that observed. It is by this type of calculation,
which we shall not reproduce here, that the zone dimensions reported in the
first papers [64] were obtained. But this model of zones of clustered
copper atoms docs not allow an explanation of the other abnormal and
complex scattering phenomena observed at high angles. In order to
explain these it is necessary to assume that the lattice of the zones is
deformed. Since the copper atom is much smaller than the aluminum
1 It is curious to discover that in this very particular case we cannot make the

approximation usually rnado in Chapter 2, which was to neglect the variation of the
direction of the vector h with the suattering angle.
212 SMALL-ANGLE SCATTERING OF X-RAYS

atom, it seems reasonable that a clustering of copper would cause such a

contraction in the spacing of the crystalline planes.
The low-angle scattering of this alloy extends to rather large angles,
actually joining the other regions of scattering. Thus the single "cross"
of Fig. 77 cannot be isolated from the other regions of scattering that were
not represented there. The procedure for separation of the effect of the
heterogeneity from the effect of the lattice deformation is not applicable
here. An attempt at a more complete explanation is no longer within

i.,.h,,.·
,_
-,f

Fig. 77. Small-angle pattern from a single crystal of an Al-Cu alloy

(4 per cent Cu by weight). The [100) axis is parallel to the incident
beam and normal to the film, and the [010] axis is vertical. Mo Ka.
radiation; sample-to-film distance, 4 cm. Enlarged 2 X.

the compass of a discussion of small-angle scattering but rather is in that

of crystalline imperfections in general.
It is no less true, however, that, qualitatively and even semiquanti-
tatively speaking, an examination of the small-angle scattering gives the
orientation, the form, and the order of magnitude of the dimensions of the
heterogeneous regions.
The examples treated above show the capabilities of small-angle
scattering in a study of a solid state reaction. We could also find applica-
tions to non-metallic substances. When the presence of heterogeneities
of a dimension less than several hundred angstroms is suspected, the
X-ray method should be attempted. Let us point out that the zones of
clustering in hardening alloys have been demonstrated only by X-ray
APPLICATIONS 213

methods; they are not visible to the electron microscope, even though
they are of a size greater than the resolving power of this instrument.
This is due to the fact that the zones are not sensitive to the methods of
etching that have so far been employed.

6.5. ABSOLUTE MEASUREMENTS OF THE INTENSITY OF SCATTERING

AT ZERO ANGLE. MEASUREMENTS OF THE COMPRESSIBILITY OF
A FLUID
In the majority of the applications cited up to this point only a know.
ledge of the form of the scattering curve is required. In several cases we
have seen that a comparison between two scattering curves had to be
made (p. 190), but the only requirement was that the incident intensity be
constant, the absolute magnitude not being needed.
It is possible to determine experimentally the ratio of the scattered
intensity to the incident intensity. These absolute measurements
( §3.5.3) are difficult because of the disproportion between the two quanti-
ties to be measured. We have shown, nonetheless, how these measure-
ments can be successfully made with a reasonable accuracy, of the order
of 5 per cent.
The formulas employing the absolute value of the limiting intensity were
established in Chapter 2 and the possible applications follow from these.
In the case of identical, widely separated particles, the equation for
the limiting intensity is
l(O) = I.Nn2
I,, the intensity scattered by one electron, can be calculated as a function
of the incident intensity by means of Thomson's equation (p. 6) (I.(O)
= 7.9 X l0- 26 / 0p-2 ); N is the total number of particles, and ni8 the number
of electrons per particle. l(O) can certainly not be determined directly by
experiment, but if the curve of log /(h 2 ) is linear at low angles it can
easily be obtained by extrapolation. The total mass of the scattering
particles gives the value of the product, Nn. The small-angle scattering
experiment will therefore give the value of n, which, let us recall, can be
used to describe the form of the particle more precisely ( §4.1.1.1).
If, on the other hand, the form of the particle is known, the volume
of the particle, and thus n, can be determined from the radius of gyration.
In such a case the X-ray measurement gives the value of N. In this way
the number of particles or the concentration of particles in a sample can
be determined. There are certain cases in which this method, complex
though it is, is the only possible method. Let us cite as an example the
case of microcavities in cold-worked metals, in which Blin (1954) deter-
mined the total volume of the cavities by X-ray measurements. The
214 SMALL-ANGLE SCATTERING OF X-RAYS

number of clusters of atoms in a hardening alloy can also be determined

in this way.
Another expression for the absolute value of the scattered intensity was
given in §2.2.3.3 for the example in which the particles were molecules of a
fluid (this expression is not restricted to macromolecules only). The
relation (equation 2.74, p. 47) is
-kT
/(0) = I.(O)n 2N -{1
1.'1

N is the average number of molecules irradiated; Tis the temperature of

the fluid; v1 , the average volume offered to each molecule (v1 = V/N);
and {1 is the isothermal compressibility,

{1 = _ ~
V oP
(oV) T

An interesting application of equation 2.74 is the experimental measure-

ment of the coefficient of compressibility of a fluid by X-rayB. This measure-
ment has recently been made for liquid helium at temperatures above and
below the .A.-point (Tweet [162']). The intensity scattered by helium is
very weak. However, since the intensity varies very little with scattering
angle, the Geiger-Muller counter detector can be adjusted to receive very
divergent beams, and this, together with the use of a powerful source (a
rotating anode tube), makes the measurement possible. The container
for the scatterer was filled successively with helium and with a gas that
scattered as a perfect gas. The value of {1 could be derived from the ratio
of the measured intensities and from the ratio of the densities.
At 4° K. the scattered intensity showed a rather marked maximum at
zero-angles (Fig. 78), just as is observed for all fluids in the neighborhood
of their critical point (see Fig. 11 for argon). The value of {1 determined
from these X-ray measurements agreed well with the value of the com-
pressibility as measured by completely different methods.
At 2° K., below the .A.-point, the intensity does not vary perceptibly
with the scattering angle, and is slightly higher than the value of (kT/v1 ){1
by an amount greater than the experimental uncertainty. These experi-
ments show that the statistics of the helium atoms at this temperature
are different from those of the normal gas, and they will serve IU! a test of
theoretical predictions.

REFERENCES FOR CHAPTER 6

Barrett, C. S. (1952), Structure of Metals, McGraw-Hill, New York.
Belbeoch, B., and Guinier, A. (1954), Gompt. rend., 238, 1003.
Blin, J. (1954), Th/>se Ing. Doct., Univ. Paris.
APPLICATIONS 215

l(h)
I.(h)n 2 N

0 8= :; radian
(a)

I(h)
le(h)n2fii

T= 2.0"K.

--------""""!'l'~~..._~-::-o~~~~"'O"'"~--~~'O'-~---

5·10- 3 O= i~ radian
(b)

Fig. 78. The scattered intensity in absolute units for liquid helium
above and below the A-point, at 4.16° K. and 2.0° K., respectively.
The point EB indicates the theoretical value of the scattered intensity at
zero-angle derived from equation 2. 74. (Tweet [162'].)
216 SMALL-ANGLE SCATTERING OF X-RAYS

Cowley, J. (1950), J. Appl. Phys., 21, 24.

Fournet, G. (1953a), Acta Met., 1, 38'3.
Fournet, G. (1953b), Compt. rend., 237, 75.
Hermans, J. J. (1944), Rec. trav. chim., 63, 5.
Heyn, A. N. J. (1953), Textile Research J., 23, 782.
Mark, H. (1932), Phyaik und Chemie der Cellulose, Springer, Berlin.
Ponder, E. (1945), J. Gen. Physiol., 29, 89.
Schmidt, P., Ke.esberg, P., and Beeman, W.W. (1954), Biochim. et Biophya. Ar,ta,
14, I.
Walker, C. B., and Guinier, A. (1953), Acta Met., 1, 568.
BIBLIOGRAPHY
This compilation of 569 titles on small-angle X-ray scattering is a
revision of the 398-title bibliography issued in March 1952 by the American
Crystallographic Association. It is divided for convenience into seven
sections as follows:
I. General l-197 Isotropic and non-periodic small-angle X-ray scattering.
II. Geometry 198-244 Camera geometry and geometric errors.
III. Neutrons 245-258 Small-angle neutron scattering.
IV. Basic Theory 259-309 Selection of theoretical work basic to development.
v. Light 310-356 Analogous and associated light scattering.
VI. Solutions 357-373 Selection of work on isotropic emulsions.
VII. Fibers and
Proteins 374-506 Selection from literature-mostly periodic and non-
isotropic.

The division into these sections is arbitrary; many of the references

might well be listed in more than one section. However, each reference is
listed only once in the section considered most appropriate. Sections
I-III were made as complete and up to date as possible. Sections IV-
VII are incomplete, as the definition of what is properly associated with
small-angle scattering is here somewhat indefinite.
Within each section the arrangement is alphabetical by first author,
and chronological under each author. The numbering of references is
continuous through all seven sections. The author index is alphabetical
and notes all references to each author by number. The type of reference
is coded as follows:
A. Published article or letter.
P. Presented paper.
a. Published abstract.
B. Book.
T. Thesis.

For articles or letters the information includes the authors, journal,

volume, inclusive pages, year, and title. If only one page number is
given, the reference is complete on that page. For papers the information
includes the authors, society, paper number, place and date of meeting,
and title. Papers abstracted in a journal have the same information as
articles. For books the information includes the authors, publisher's name
and location, the year of publication, the title, and usually the pages of
217
218 SMALL-ANGLE SC1\TTERING OF X-RAYS

interest. '1-'hesis information includes the author, institution, degree,

year, and title.
For a few items the above information is not complete. Often a brief
annotation is appended to the reference. The language of the title
indicates the language of the paper. It was necessary to translate a
few titles back to the approximate original from English; and the Russian
names were transliterated and titles translated.
The compound and transliterated names have been alphabetized as
follows:
Boyes-Watson B von N ardroff N
DeVore D Van Nordstrand N
Danilov D Nowacki N
DuMond D Phillipovich p
Frey-Wyssling F Porai-Koshits p
Friedrich-Freska F Serikov s
J aneschitz-Kriegl J Shishakov s
Kachkurzov K Sokolov s
Kalinina K Tades T
LaMer L Toraldo Di Francia T
Lennard-Jones L Van de Hulst v
MacArthur M Van Rijssel v
McBain M van der Wyk w
Mc Reynolds M Zubko z
McRoberts M

To avoid tedious renumbering and reindexing, 66 late additions to the

bibliography were incorporated into appropriate positions and numbered
with primes.

J. GENERAL
Bale, H. T
Univ. N. Dakota, M.S. (1953). "Acetic Acid near the Critical Point,"
See No. 180.
l' Banerjee, K. P
Intern. Union Cryst., No. B-8, Cambridge, ::lllass. (July 1948). "::lllethod for
Studying Extremely Low-Angle Scattering of Monochromatic X-Rays."
Not presented.
2 Banerjee, K., and Maitra, J.C. A
Indian J. Phys., 25, 141-144 (1951). "On a Method of Studying Small-Angle
Scattering of Monochromatic X-Rays." Single crystal spectrometer.
3 Barton, H. M., Jr. T
Polytechnic Inst., Brooklyn, M.S. (1950). "A Study of Electron Density by
Low-Angle X-Ray Scattering." Electron density of carbon blacks from total
power in small-angle scattering.
4 Barton, H. M., Jr. Pa
Am. Phys. Soc., No. H-6, Washington (April 1950). Phys. Rev., 79, 211 (1950).
"Average Electron Density ::vt:easurements by Low-Angle X-Ray Scattering."
BIBLIOGRAPHY 219

5 Barton, H. M., Jr., and Brill, R. A

J, Appl. Phys., 21, 783-785 (1950). "Average Electron Density Measurements
by Low-Angle X-Ray Scattering." See No. 174.
6 Bauer, S. H. A
J. Chem. Phys., 13, 450-451 (1945). "Inversion of Low-Angle Scattering Data
for Particle Size Distribution." See Nos. 145, 147, 155.
7 Baule, B., and Kratky, 0. A
Z. physik. Chem., B-52, 142-152 (1942). "Die Dehnung der amorphfln Bereiche
bei affiner Verzerrung eines micellaren Netzee."
8 Beeman, W. W., and Kaesberg, P. A
Phys. Rev., 72, 512 (1947). "X-Ray Scattering at Very Small Angles."
Use of double crystal spectrometer. Comparieon of multiple refraction and
scattering. See Nos. 173, 206, 208, 209, 212, 226, 230, 238.
8' Beeman, W. W., and Neynaber, R. H. Pa
Am. Phys. Soc., No. 17, Washington (April 1954). Phya. Rw., 95, 617 (1954),
"Equipment for the Measurement of Very Small X-Ray Scattering Cross
Sections."
9 Bernal, J. D., and Carlisle, C. H. PA
Discussions Faraday Soc., 11, 227-229 (1951 ). "The Place of X-Ray Diffraction
Methods in the Estimation of the Size and Mutual Arrangements of Colloidal
Particles."
10 Biscoe, J., and Warren, B. E. A
J. Appl. Phys., 13, 364-371 (1942). "An X-Ray Study of Carbon Black."
Particle size of carbon black from small-angle scattering compared with
crystallite size from line broadening.
11 Blin, J., and Guinier, A. A
Gompf. rend., 233, 1288-1290 (1951). "Diffusion des rayons X aux tree petits
angles par lea metaux ecrouis." Cavities in Ni and Cu of about 18 A.
11' Blin, J., and Guinier, A. A
Compt. rend., 236, 2150-2152 (1953). "Etude experimentale des cavites
submicroscopiques au sein des metaux ecrouis."
11 N Bowman, J. C., Hayes, S., and Smoluchowski, R. Pa
Am. Phys. Soc., No. Z-1, Chapel Hill, N. Carolina (March 1953). Phys. Rev.,
91, 244 (1953). "X-Ray Study of Graphitization."
12 Brentano, J. C. M. P
ASXRED, No. 17, Philadelphia (December 1949). "Small-Angle Diffraction
from Compact Powders." Theory of scattering from irregular discontinuities
related to structure breaks.
13 Brusset, H. T
Univ. Paris, D .S.P. (194 7). "Contribution a l'etude du carbone et des charbons
mineraux." Comprehensive study of structure of solids by small-angle X-ray
scattering, with special application to carbons.
14 Brusset, H. A
Compt. rend., 225, 1002-1003 (1947). "Diffusion centrale des rayons X et
microscopie electronique." Six samples of C black examined by small-angle
scattering and electron microscopy. Sizes by electron microscopy several
times those from X-ray measurements. It is concluded that in electron
microscopy the particles usually observed are aggregates.
15 Brusset, H. A
Compt. rend., 228, 323-327 (1949). "Etude par la diffusion centrale des rayons
X de !'activation du carbone."
220 SMALL-ANGLE SCATTERING OF X-RAYS

16 Brusset, H. A
Bull. soc. chim. France, 16, 319-323 (1949). "Etude de l'etat de division
fine des gels de silice." ·
17 Brusset, H., Devaux, J., and Guinier, A. A
Oompt. rend., 216, 152-154 (1943). "Etude de quelques charbons par la
diffusion des rayons X sous tres petits angles." Carbons alike in wide angle
patterns distinguished by their small-angle scattering.
18 Brusset, H., and Kikindai, T. A
Oompt rend., 231, 858-860 (1950). "La diffusion centrale des rayons X
produite par quelques colloides mineraux." Particle sizes of several minerals.
18' Buttrey, J. W. T
Univ. Mo., Ph.D. (1953). "Small-Angle Scattering from Liquid Paraffins near
Their Critical Points."
19 Carroll, B., and Fankuchen, I. A
J. Chem. Phys., 16, 153-154 (1948). "Small-Angle X-Ray Scattering from
Metal Deposits Made by Evaporation." Black deposits of evaporated Al and
Cu show small-angle scattering characteristic of colloidal particles.
19' Charlesby, A. A
J. Polymer Sci., 10, 201-211 (1953). "Investigation of Halo Patterns of
Amorphous Polymers." Debye formula for several molecular models to
explain halos.
20 Christman, A. C., Jr. T
Penn. State College, M.S. (1950). "X-Ray Determination of Particle Size of
Mercuric Oxide."
21 Clark, G. L. B
McGraw-Hill, New York (1947). Applied X-Rays. Third edition, p. 504.
Brief discussion of Guinier's method. Fuller treatment in fourth edition (Jan.
1955).
22 Clark, G. L., Eckert, A. C., Jr., and Burton, R. L. A
Ind. Eng. Chem., 41, 201-208 (1949). "Commercial and Experimental Carbon
Blacks. An X-Ray Comparison." No two carbons give identical X-ray
patterns. The scattering at small angles increases with temperature of heat
treatment at low temperatures, indicating a growth of clusters.
23 Daams, H., and Arlman, J. J. A
Appl. Sci. Research, B-2, 217-226 (1951). "Design of and Measurements with
a Simple Double Crystal Spectrometer for Small-Angle Scattering." Latex
peaks. MgO correlated with electron microscopy. See Nos. 25, ll5, 192.
24 Danielson, W. E. T
Calif. Inst. Tech., Ph.D. (1952). "Application of the Point Focusing X-Ray
Monochromator to the Study of Low-Angle Diffraction."
25 Danielson, W. E., Shenfil, L., and DuMond, J. W. M. A
J. Appl. Phys., 23, 860-865 (1952). "Latex Particle Size Determination
Using Diffraction Peaks Obtained with the Point Focusing X-Ray Mono-
chromator." See Nos. 23, II5, 192.
26 Danilov, V. I., and Serikov, A. S. A
Doklady Akad. Nauk, SSSR, 83, 71-74 (1952). "The Scattering of X-Rays at
Small Angles and the Pore Structure of Activated Carbon." (In Russian).
27 Danilov, V. I., and Zubko, A. M. A
Doklady Akad. Nauk, SSSR, 82, 385-388 (1952). "On the Fine Structure of
Activated Carbons." (In Russian.)
BIBLIOGRAPHY 221

28 Davis, B. A
J. Franklin Inst., 204, 29-39 (1927). "Refraction of X-Rays." Experimental
central scattering attributed to refraction.
29 Davis, B., and Slack, C. M. Pa
Am. Phys. Soc., No. 8, Washington (April 1926). Phy.<. Rev., 27, 796-797
(1926). "Refraction of X-Rays in Prisms." Broadening of beam by granular
materials.
30 Dexter, D. L. T
Univ. Wisc., Ph.D. (1950). "Studies in Small-Angle Scattering of X-Rays and
in Solid State Physics."
30' Dexter, D. L. Pa
Am. Phys. Soc., No. FA-10, Washington (May 1953). Phys. Rev., 91, 448
(1953). "Small-Angle Scattering of X-Rays from Cold-Worked Crystals."
30• Dexter, D. L. A
Phys. Rev., 90, 1007-1012 (1953). "The Small-Angle Scattering of X-Rays
from Cold-Worked Solids." Scattering from either cavities or edge-type
dislocations.
31 Dexter, D. L., and Beeman, W. W. A
Phys. Rev., 76, 1782-1786 (1949). "Multiple Diffuse Small-Angle Scattering
of X-Rays." Multiple scattering from large spheres. Particles sizes are
determined from the variation of the width of the scattering curve with sample
thickness. Same as Part III of No. 30.
32 Dexter, D. L., and Beeman, W.W. P
Con/. Small-Angle X-Ray Scattering, No. 7, Columbia, Mo. (October 1949).
"Multiple Diffuse Small-Angle Scattering of X-Rays."
33 Dexter, D. L., and Beeman, W. W. Pa
Am. Phys. Soc., No. V-3, Chicago (November 1949). Phys. Rev., 77, 761
(1950). "Multiple Diffuse Small-Angle Scattering of X-Rays.
34 Donnay, .J. D. H., and Shull, C. G.
ASXRED Bibliography (1946). "X-Ray Small-Angle Scattering Annotated
Bibliography," Sixty-nine references arranged by year to 1945. Indexed as
non-Laue reflections, small-angle scattering, large Bragg spacings, and non-
Bragg reflections.
34' Dorgelo, H. B., and Daams, H. A
Ned. Tijdschr. Natuurk., 17, 209-233 (1951). "De Bespaling van Grootte,
Vorm en Grootte-Distributie van Submicroscopische Deeltjes."
35 Dragsdorf, R. D. T
Mass. Inst. Tech., Ph.D. (1948). "Small-Angle X-Ray Scattering." Refrac-
tion versus scattering discussed. Interparticle interference peaks noted.
See Nos. 87, 168, 254, 256, 355.
36 Dragsdorf, R. D. p
ASXRED, No. 31, Columbus (December 1948). "Small-Angle X-Ray
Scattering.''
37 Dragsdorf, R. D. P
Con/. Small-Angle X-Ray Scattering, No. 12, Columbia, Mo. (October 1949).
"Small-Angle X-Ray Scattering." Refraction versus scattering. Inter-
particle interference.
38 Eisenstein, A. S. T
Univ. Mo., Ph.D. (1942). "The Diffraction of X-Rays by Argon in the Liquid,
Vapor, and Critical Regions."
222 SMALL-ANGLE SCATTERING OF X-RAYS

39 Eisenstein, A. S., and Gingrich, N. S. A

Phys. Rev., 58, 307-309 (1940). "The Diffraction of X-Rays by Liquid Argon."
40 Eisenstein, A. S., and Gingrich, N.S. A
Phys. Rev., 62, 261-270 (1942). "The Diffraction of X-Rays by Argon in the
Liquid, Vapor, and Critical Regions." Small-angle scattering from argon near
saturated vapor curve. See Nos. 49, 59, 180.
41 Elkin, P. B., Shull, C. G., and Roese, L. C. A
Ind. Eng. Chem., 37, 327-331 (1945). "Specific Surface and Particle Size
Distribution in Silica Alumina Gels." Correlations given between specific
surface and particle size distribution obtained from small-angle scattering.
42 Fankuchen, L., and Jellinek, M. H. B
Academic Press, New York (1948). Advances in Catalysis. Chapter on X-ray
diffraction.
43 Firth, F. G. A
Rubber Age and Synthetics, London, 57, 561-564 (1945). Use of Low-Angle
X-Ray Scattering in the Study of Catalysts, Viruses, and Other Materials."
Changes indicating giowth of Al 10 3 particles. Lines of diffuse maxima along
meridian with nylon at angles corresponding to Bragg period of 100 A. Review
and discussion of developments.
44 Fournet, G. A
Compt. rend., 228, 1421-1423 (1949). "Diffusion des rayons X aux petits
angles par des systemes denses de particules identiques." Small-angle scatter-
ing equation including interparticle interference derived for gases, based
on Born and Green's partition function. See Nos. 56, 120, 126, 128, 137, 186,
268, 284.
45 Fournet, G. A
Compt. rend., 228, 1801-1803 (1949). "Application de la theorie de Born et
Green au calcul de l'intensite diffuse par les ge.z et !es liquides." Extension of
equation including interparticle interference to apply to liquids as well as gases.
See Nos. 259, 260, 261, 273.
46 Fournet, G. A
Compt. rend., 229, 1071-1073 (1949). "Generalisation de la theorie des fluides
de Born et Green aux melanges du plusieurs especes de molecules. Calcul de la
diffusion par !es rayons X." Extension of equation including interparticle
interference to fluids of two sizes of spheres.
47 Fournet, G. TB
Paris, Doctorat d'Etat-Sciences (1950). Masson et c, Paris (1951). "Etude
theorique et experimentale de la diffusion des rayons X par Jes ensembles
denses de particules."
48 Fournet, G. A
Bull. soc./rant;. mineral. et crist., 74, 39-ll3 (1951). Same as No. 47.
49 Fournet, G. A
Acta Cryst., 4, 293-301 (1951). "Diffusion des rayons X par les fluides."
Intensity from Born and Green-argon. See Nos. 40, 59, 180.
50 Fournet, G PA
Discussions Faraday Soc., 11, 121-125 (1951). "Influence of the Size and
Shape of Particles on the Interpretation of the X-Ray Diffuse Diagrams."
Critical discussion of validity of relating interparticle intensity maximum to a
"mean" distance.
BIBLIOGRAPHY 223

51 Fournet, G. A
J. phys. radium, 12, 592-595 (1951). "Generalisation de la tMorie cinetique
des fluides de Born et Green aux ensembles de pe.rticules de plusieurs especes
differentes." Effects due to proximity of particles and interpe.rticle forces are
separated. Particles of several kinds.
52 Fournet, G. PA
Intern. Union Chem., Stre.sbourg (June 1952). J. Polymer Sci., 9, 539-545
(1952). "Interpretation des die.grammes de diffusion des rayons X par des
systemes denses de pe.rticules."
53 Fournet, G., and Guinier, A. A
Compt. rend., 228, 66-68 (1949}. "Interpretation de la valeUl' limitee de la
diffusion des rayons X aux tres fe.ibles angles." Scattered intensity at zero
angle attributed to density fluctuations which may be treated thermodynami-
cally.
54 Fournet, G., and Guinier, A. A
J. phys. radium, 11, 516-520 (1950). "L'ete.t actual de la tMorie de la diffusion
des rayons X aux petits angles." See No. 74.
55 Fran9ois, J. A
Compt. rend., 230, 1282-1284 (1950). "Influence de la temperature sur la
dimension des pe.rticules d'oxyde de nickel." Usual slope analysis correlated
with electron microscopy.
56 Frank, R. M., and Yudowitch, K. L. A
Phys. Rev., 88, 759-760 (1952). "Small-Angle X-Ray Scattering from Com-
pact Identical Particles." Novel treatment using an average radial electron
density. See Nos. 44, 120, 126, 128, 137, 186, 268, 284.
56' Franklin, R. E. A
Acta Cryst., 3, 107-121 (1950). "The Interpretation of Diffuse Die.grams of
Carbon."
56" Franklin, R. E. A
Proc. Roy. Soc. (London), A-209, 196-218 (1951). "Crystallite Growth in
Gre.phitizing and Non-Graphitizing Carbons."
57 Gingrich, N. S. P
Conj. Small-Angle X-Ray Scattering, No. 1, Columbia, Mo. (October 1949).
"Introductory Remarks." Brief review.
58 Glocker, R. A
FIAT Rev. Ger. Sci., 1939-1946, I, 1-14 (1948). "Physik der fasten Karper."
Review of German work to 1946. Crystal monochrome.tors. Fibers.
59 Graham, W., and Lund, L. H. A
J. Chem. Phys., 19, 1380--1382 (1951). "Small-Angle X·Ray Scattering from
Argon." See Nos. 40, 49, 180.
60 Gray, J. A. A
Oan. J. Research, 12, 408-409 (1935). "Note on the Scattering of X-Rays at
Small Angles."
61 Gray, J. A., and Zinn, W. H. A
Oan. J. Research, 2, 291-293 (1930). "New Phenomena in X-Ray Scattering."
Early observation of small-angle scattering by charcoal, related to particle size.
See No. 130.
62 Guinier, A. A
Compt. rend., 206, 1374-1376 (1938). "La diffusion des rayons X sous !es tree
faibles angles e.ppliques a l'etude de fines particles et de suspensions colloide.les."
224 SMALL-ANGLE SCATTERING OF X-RAYS

63 Guinier, A. A
Compt. rend., 206, 1641-1643 (1938). "Un nouveau type de diagrammes de
rayons X." Particle size determination from small-angle scattering. Central
streaks.
64 Guinier, A. T
Univ. Paris, Doctorat (1939). "La diffraction des rayons X aux tres petits
angles; application a l'etude de phenomlmes ultramicroscopiques." Basic
discussion of X-ray scattering at small angles by small discrete particles.
Considers effect of particle shape but not of size distribution.
65 Guinier, A. A
Ann. phys., 12, 161-237 (1939). "La diffraction des rayons X aux tres petits
angles; application a l'etude de phenomenes ultramicroscopiques." Same as
No. 64.
65' Guinier, A. A
Compt. rend., 208, 894-896 (1939). "La diffusion des rayons X par Jes
cristaux et les corps microcristallises."
66 Guinier, A. A
J. phys. radium, 2, 124-136 (1942). "Le mecanisme de la precipitation dans
un cristal de solution solide metallique. Cas des systemes aluminum-cuivre et
aluminum-argent."
67 Guinier, A. A
J. chim. phys., 40, 133-150 (1943). "Determination de la taille des particules
submicroscopiques par les rayons X." Summary of work by Guinier, Hose-
mann, and Kratky.
67' Guinier, A. A
Metaux, corrosion, usure, 18, 209-213 (1943). "Etude du mecanisme de la
precipitation d'une solution solide sursaturee aluminium-zinc." Clusters in
Al-Zn age-hardened alloy.
68 Guinier, A. B
Dunod, Paris (1945). Radiocrystallcgraphie. Chapter 12 and Appendix 8.
See No. 72.
69 Guinier, A. A
Proc. Phys. Soc. (London), 57, 310-324 (1945). "Imperfections of Crystal
Lattices as Investigated by the Study of X-Ray Diffuse Scattering." Small-
angle method mentioned.
70 Guinier, A. PA
Colloq. tres grosses molecules, Liege (April 1948). Bull. soc. chim. Belges, 57,
286-297 (1948). "Determination de la taille et de !'arrangement mutuel des
grosses molecules par la diffusion des rayons X aux petits angles." Two
curved crystals. Hemoglobin, hemocyanine in solution.
71 Guinier, A. PA
Discusswns Faraday Soc., 8, 344-347 (1950). "Study of Catalysts by Scatter-
ing of X-Rays at Small Angles." Use of small-angle scattering supplementing
other data to determine size, shape, and dispersion of catalysts.
72 Guinier, A. B
Hilger and Watts, London (1952). X-Ray Crystallographic Technology.
Translation of No. 68 by T. L. Tippell.
72' Guinier, A A
Z. Metallkunde, 43, 217-223 (1952). "Les donnees de la diffusion des rayons
X aux petits angles sur la structure des alliages durcissants. Cas de l'alliage
aluminium-argent." Clusters in Al-Ag.
BIBLIOGRAPHY 225
73 Guinier, A. PA
Intern. Union Chem., Strasbourg (June 1952). J. Polymer Sci., 9, 557-563
(1952). "Les po~sibilites de la diffusion des rayons X aux petits angles dans
l 'etude des grosses molecules.''
~~~A ~
Am. Phys. Soc., No. V-1, Chapel Hill, N. Carolina (March 1953). Phys. Rw.,
91, 239 (1953). "Possibilities of X-Ray Small-Angle Scattering for Study of
Crystal Imperfections."
74 Guinier, A., and Fournet, G. P
Oonf. Small-Angle X-Ray Scattering, No. 6, Columbia, Mo. (October 1949).
"L'etat actual de la theorie de la diffusion des rayons X aux petits angles."
Review of scattering by systems of dilute and compact, uniform and non-
uniform particles. Read in English translation. Same as No. 54.
74' Guinier, A., and Jacquet, P. A
Rev. met., 41, 1-16 (1944). "Etude du durcissement des alliages cuivre-
glucinium." Clusters in age-hardened alloys of Cu-Be.
74" Gupta, N. N. A
Proc. Nall. Inst. Sci. India, A-18, 379-388 (1952). "The Optical Principles of
the Low-Angle Scattering ofX-Rays."
74'8 Hayes, S., and Smoluchowski, R. Pa
Am. Phys. Soc., No. Z-2, Chapel Hill, N. Carolina (March 1953). Phys. Rev.,
91, 244 (1953). "Small-Angle X-Ray Scattering in Proton Irradiated Dia-
mond."
74•• Hayes, S., and Smoluchowski, R.
Appl. Sci. Research, B-4, 10-24 (1954). "Small-Angle X-Ray Scattering
Study of Imperfections in Copper."
75 Hayes, S., Smoluchowski, R., and Turner, R. W. Pa
Am. Phys. Soc., No. H-4, Cambridge (January 1953). Phys. Rev., 90, 350
(1953). "Small-Angle X-Ray Scattering in Deformed Crystals."
76 Hendricks, S. B. A
Z. Krist., A-83, 503-504 (1932). "Die Glanzkohle, eine andere Kohlenstoff.
form." Diffuse scattering approximately uniform in density between central
image and interference maxima.
76' Henke, B. T
Calif. Inst. Tech., Ph.D. (1953). Total reflection focusing of copper L lines.
Latex application.
76" Henke, B.
Calif. Inst. Tech., Special Tech. Rep. 24, O.N.R., No. 3, A.E.C. (June 1953).
"Diffraction of Long Wavelength X-Rays."
76'• Henke, B. Pa
Am. Phys. Soc., No. B-1, Albuquerque, N.M. (September 1953). Phys. Rev.,
92, 1079 (1953). "The Size Determination of Submicroscopic Particles by
Long Wavelength X-Ray Diffraction."
70•• Henke, B., and DuMond, J. W. M. A
Phys. Rev., 89, 1300 (1953). "Low-Angle X-Ray Diffraction with Long Wave-
lengths."
77 Henry, N. F. M., Lipson, H., and Wooster, W. A. B
Van Nostrand, New York (1951). The Interpretation of X-Ray Diffraction
Photographs. (P. 217.)
78 Heywood, H. A
Nature, 159, 717-718 (1947). "Particle Size Analysis." Report on February
1947 Conf. Inst. Chem. Engrs. and Soc. Chem. Ind.
226 SMALL-ANGLE SCATTERING OF X-RAYS

79 Hosemann, R. A
z. Physik, 113, 751-768 (1939). "Theorie der Rontgenstrahlenstreuung und
Partikelhaufen. Aufstellung des Aggregationsdiagrammes." Theory developed
for small-angle scattering from single oriented rotational ellipsoids.
80 Hosemann, R. A
Z. Physik, 114, 133-169 (1939). "Neues Rontgenographisches Verfahren zur
Bestimmung des submikroskopischen Feinbaues eines Stoffes. Diskussion des
Zellulosemodells." Description of apparatus and cellulose data. Application
of theory of No. 79.
80' Hosemann, R. A
Z. Elektrochem., 53, 331-334 (1949). "Theorie der Rontgenstreuung an
statistisch Ungeordneten linearen Gittern."
81 Hosemann, R. A
Kolloid-Z., 117, 13-41 (1950). "Rontgeninterferenzen an kolloiden Systemen.
Zusammenhang zwischen Packungsdickte, Polydispersitat, und reiner Partikel-
streuung." General theory for small-angle scattering for randomly and
regularly packed as well as dilute systems.
82 Hosemann, R. A
Kolloid-Z., 119, 129-145. (1950). "Grossenstatistik der parakristallinen
Bereiche in festen hochrnolekularen Stoffen bei kontinuierlicher Kleinwinkel-
streuung."
83 Hosemann, R. A
Z. Physik, 128, 465-492 (1950). "Der ideale Parakristall und die von ihrn
gestreute kohiirente Rontgenstrahlen."
84 Hosemann, R. A
Ergeb. exakt. Naturw., 24, 142-221 (1951). "Die Erforschung der Struktur
hochmolekularer und kolloider Stoffe mittels Kleinwinkelstreuung." Com-
prehensive, 188 references.
85 Hosemann, R. PA
Deut. Mineral. Gesell., Sektion fiir Kristallkunde (1951). Zur Struktur und
Materie der Festkorper, 127-222 (1951). "Der statistische Charakter der
Feinstruktur hochmolekularer und kolloider Stoffe." Lengthy review with
48 references.
86 Hosemann, R., and Bagchi, S. N. A
Acta Cryst., 5, 749-762 (1952). "Existenzbeweis fiir eindeutige Rontgen-
strukturanalyse durch Entfaltung. I. Entfaltung zentrosymmetrischer
endlicher Massenverteilungen."
87 Hubbell, H. H., Jr. T
Princeton, Ph.D. (1947). "X-Ray Scattering at Small Angles." Wavelength
dependence checked. Secondary intensity plateaus noted. See No. 161.
88 Hubbell, H. H., Jr. p
Conj. Small-Angle X-Ray Scattering, No. 2, Columbia, Mo. (October 1949).
"Observed and Expected Intensity Distribution in the Small-Angle Scattering
from Carbon and Other Substances."
89 Imelik, B., and Carteret, Y. A
Compt. rend., 231, 280--282 (1950). "Etude de gel de silice par la diffusion
centrale des rayons X."
90 Imelik, B., and Carteret, Y. A
Bull. soc. chim. France, 18, 864-867 (1951). "Etude de gels de silice."
BIBLIOGRAPHY 227

91 Imelik, B., Teichner, S., and Carteret, Y. A

J. chim. phys., 48, 438-446 (1951). "Etude comparative de la texture des gels
de silice par le diffraction des rayons X aux petits angles et !'adsorption
d'azote a basse temperature."
92 Jellinek, M. H. T
Polytechnic Inst., Brooklyn, Ph.D. (1946). "The X-Ray Examination of
Gamma Alumina." Line breadth and small-angle studies of heat treatment
of gamma alumina. Attempted size distribution from small-angle scattering.
93 Jellinek, M. H., and Fankuchen, I. A
Ind. Eng. Chem., 37, 158-164 (1945). "X-Ray Diffraction Examination of
Gamma Alumina." Crystal and particle size. Filtered radiation and long
collimating system used in small-angle scattering study on heat-treated
gamma alumina.
94 Jellinek, M. H., and Fankuchen, I. Pa
Am. Phys. Soc., No. D-4, New York (January 1945). Phys. Rev., 67, 201
(1945). "Low-Angle Scattering of X-Rays." Use of Geiger counter instead of
film for continuous recording over 1000 : l intensity range.
95 Jellinek, M. H., and Fankuchen, I. A
Ind. Eng.- Chem., 41, 2259-2265 (1949). "X-Ray Examination of Pure
Alumina Gel." Both wide-angle and small-angle measurements indicate an
increase in crystallite and particle size with increase in severity of heat treat-
ment, negating structures proposed earlier in ~o. 93.
96 Jellinek, M. H., Solomon, E., and Fankuchen, I. A
Ind. Eng. Chem., Anal. Ed., 18, 172-175 (1946). "Measurement and Analysis
of Small-Angle X-Ray Scattering." Geiger recording of small-angle X-ray
scattering. Graphical analysis of size distribution given for alumina and
carbon black.
97 Kaesberg, P. T
Univ. Wisc., Ph.D. (1949). "The Use of the Double Crystal Spectrometer in
the Analysis of Small-Angle Scattering."
97' Kahovek, L., Porod, G., and Ruck, H.
Kolloid Z., 133, 16 (1953). "Rontgenkleinwinkeluntersuchungen an dichtge-
packten kolloiden Systemen."
98 Kaufman, H. S., and Fankuchen, I A
Anal. Chem., 24, 20-22 (1952). "X-Ray Diffraction." General review.
99 Kaufman, H. S., Saches, A., Alfrey, T., and Fankuchen, I. A
J. Am. Chem. Soc., 70, 3147 (1948). "Side-Chain Crystallization in Alkyl
Polycrylates."
100 Kikindai, T. A
Compt. rend., 230, 1772-1774 (1950). "Taille de particules d'oxyde de cuivre
colloidal." Slope analysis of sintered copper.
101 Klug, H. P., and Alexander, L. B
Wiley, New York (1954). X-Ray Diffraction Procedures for Polycrystalline and
Amorphous Materials, Chapter 12. Apparatus and techniques. See Nos.
193, 205.
102 Kratky, 0. A
Naturwissenschaften, 30, 542-543 (1942). "Die Berechnung von Dimensionen
kolloider Teilchen ans der unter kleinster Winkeln abgebeugten Rontgen-
strahlen." C<illulose measurements considering close-packing.
228 SMALL-ANGLE SCATTERING OF X-RAYS

103 Kratky, 0. A
Beih. Z. Ver. deut. Ohem., 47, 79-80 (1943). "Untersuchung iiber die Klein-
winkelstreuung." Size, shape, density, and swelling interpretation.
104 Kratky, 0. Aa
Monatsh. Ohem., 76, 311-318 (1946-1947). Anz. Akad. Wiss. Wien, Math.-
naturw. Kl., 83, 29-30 (1946). "Die Struktur amorpher Festkorper." Packing
in quasi-crystals.
105 Kratky, 0. A
Monatsh. Ohem., 76, 325-349 (1947). "Die Abhangigkeit der Rontgen-
kleinwinkelstreuung von Grosse und Form der kolloiden Teilchen in verdiinnten
Systemen." Scattering from non·spherical particles as a combination of
spheres. Hemocyanine.
106 Kratky, 0. A
Acta Phys. Austriaca, 4, 502 (1951). "Teilchengrossen-bestimmungen an
Metallsolen mittels der Rontgenkleinwinkelmethode."
106' Kratky, 0. A
Osterr. Ohem-Z., 54, 149 (1953). "Grosse und Gestalt von Kolloidteilchen und
Makromolekulen nach der Rontgenkleinwinkelmethode."
107 Kratky, 0., and Porod, G. A
Acta Phys. Austriaca, 2, 133-147 (1948). "Die Abhangigkeit der Rontgen-
kleinwinkelstreuung von Form und Grosse der kolloiden Teilchen in verdiinnten
Systemen." Approximations of Guinier (gyration radius), Kratky (sphere-
clusters), and Porod to determine particle form.
108 Kratky, 0., and Porod, G. A
J. Oolwid Sci., 4, 35-70 (1949). "Diffuse Small-Angle Scattering of X-Rays in
Colloid Systems." Theory for dense systems, particularly layer-packed and
sphere-clusters. Cellulose applications.
109 Kratky, 0., and Porod, G. P
12th Intern. Congr. Pure Appl. Chem., Section 9, No. 27, New York (September
1951). "Problems of Small-Angle Scattering of Densely Packed Systems."
llO Kratky, 0., Porod, G., and Kahovek, L. P
Pittsburgh Oonf. X-Ray Electron Diffr., No. 4, Pittsburgh (November 1950).
"Recent Advances in the Technique and Evaluation of X-Ray Small-Angle
Scattering Investigations." Collimation technique and corrections. Particle
volume from absolute intensity.
111 Kratky, 0., Porod, G., and Kahovek, L. A
Z. Elektrochem., 55, 53-59 (1951). "Einige Neuerungen in der Technik und
Auswertung von Rontgenkleinwinkelmessungen.''
111' Kratky, 0., and Schossberger, F. A
Z. Elektrochem., 43, 666-667 (1937). "-Ober den micellaren Aufbau des
Kautschuks."
112 Kratky, 0., and Sekora, A. A
Naturwissenschajten, 31, 46-47 (1943). "Bestimmung von Form und Grosse
geloster Teilchen aus den unter kleinsten Winkeln diffus abgebeugten Rontgen-
strahlen." Approximation for non-spherical particles by model of agglomer-
ated spheres.
ll3 Kratky, 0., Sekora, A., and Traer, R. A
Z. Elektrochem., 48, 587-601 (1942). "tl'ber die an kolloiden Systemen unter
kleinen Winkeln diffus abgebeugten Rontgenstrahlung." Lamellar micelles.
Camera. Solutions.
BIBLIOGRAPHY 229
114 Kratky, 0., Treiber, E., and Schurz, J. A
Z. Elektrochem., 56, 143-146 (1952). "Photometrierung durch automatische
Kornzahlung." Counting developed grains; usable to low densities.
114' Kratky, 0., and Worthmann, W. A
Monatsh. Chem., 76, 263-281 (1947). "tl'ber die Bestimmbarkeit der Kon-
figuration geloster organischer Molekiile durch interferometrische Vermessung
mit Rontgenstrahlen."
115 Leonard, B. R., Jr., Anderegg, J. W., Kaesberg, P., and Beeman, W.W. A
J. Appl. Phys., 23, 152 (1952). "The Size of Latex Particles by X-Ray
Scattering." See Nos. 23, 25, 192.
115' Li, Y. Y., and Smoluchowski, R. P
Am. Phys. Soc., No. E-2, New York (January 1954). "Diffraction of X-Rays
by Lattice Imperfections."
115" Li, Y. Y., and Smoluchowski, R. P
Am. Phys. Soc., No. W-10, Detroit (March 1954). "The Fourier Analysis of
X-Ray Small-Angle Scattering."
116 Lonsdale, K. B
Bell, London (1948). Crystals and X-Rays. Pp. 145-147. Brief description
of source and application of small-angle scattering.
117 Lund, L. H. P
ASXRED, Columbus, Ohio (1948). "Interparticle Interference Effects in
Small-Angle X-Ray Scattering." Interference modified intensity distribution.
118 Lund, L. H. T
Univ. Mo., Ph.D. (1949). "Radial Density Functions and Interference of
X-Rays Scattered at Small Angles." Particle-to-particle interferences worked
out for various aggregates.
119 Lund, L. H. P
Conj. Small-Angle X-Ray Scattering, No. 8, Columbia, Mo. (October 1949).
"Interparticle Interference Effects in Small-Angle X-Ray Scattering."
Intensity distribution shown for aggregates of 2 to 12 particles.
120 Lund, L. H., and Vineyard, G. H. A
J. Appl. Phys., 20, 593-597 (1949). "Interparticle Interference Effects in the
Small-Angle X-Ray Scattering from Fine Powder." Interparticle interference
significant for one-third or more close packing. Intensity distribution from
several aggregates computed. See Nos. 44, 56, 126, 128, 137, 186, 268, 284.
121 :Milligan, W. 0. P
Am. Chem. Soc., Anal. Div., No. 34, Atlantic City (September 1952). "X-Ray
and Electron Diffraction Examination of Colloidal Materials."
122 Miiller, F. H., and Erbring, H. A
FIAT Rev. Ger. Sci., 1939-1946. "Physik der Fliissigkeiten und Gase,"
Review and bibliography.
123 von Nardroff, R. Pa
Am. Phys. Soc., No. 9, 'Washington (April 1926). Phys. Rev., 27, 797 (1926).
"The Refraction of X-Rays Applied to the Determination of the Diameters of
Small Particles." Multiple refraction theory.
124 von Nardroff, R. A
Phys. Rev., 28, 240-246 (1926). "Refraction of X-Rays by Small Particles."
Theory of multiple refraction !ff X-rays at particle surfaces.
125 Noakes, C. E. T
Univ. Toronto, Ph.D. (1951). "An Investigation of the Small-Angle Scatter-
ing of X-Rays by Small Particles."
230 SMALL-ANGLE SCATTERING OF X-RAYS

126 Noakes, G. E., and Allin, E. J. A

Can. J. Phys., 31, 40-48 (1953). "Interference Effects in the Small-Angle
Scattering of X-Rays by 8mall Particles." Aluminum radiation. Colloidal
gold. Seo ~os. 44, 56, 120, 128, 137, 186, 268, 284.
126' Van Nordstrand, R. A., and Hach, K. M. P
Catalysis Club, Chicago (May 1953). "Small-Angle Scattering of Alumina
and Silica Gels." Empirical relation of scattering to adsorption measured
surface.
126" Van Nordstrand, R. A., and Johnson, M. F. L. P
Pittsburgh Conj. X-Ray Electron Diffr. and ACA, No. 23, Pittsburgh (November
1954). "Small-Angle X-Ray Scattering of Cracking Catalysts-Comparison
with Adsorption Isotherms." Slope of linear log-log plot of scattered intensity
vs. angle correlated with surface measurements for several types of catalysts.
127 Nowacki, W. A
Schweiz. Chem.-Z. u. Tech. Ind., 29, 235-238, 267-270 (1946). "Kleinwinkel·
rontgeninterferenzen und submikroskopische Struktur." Review with 42
references. See Nos: 127', 130'.
127' Nowacki, W. A
Uspekhi Khim., 16, 319-326 (1947). "The Scattering of X-Rays at Small
Angles by Submicroscopic Structure. II. The Application of X-Ray Scatter-
ing at Small Angles to the Determination of Submicroscopic Structure."
(In Russian.) See No. 130'.
128 Ostor, G., and Riloy, D. P. A
Acta Cryst., 5, 1-6 (1952). "Scattering from Isotropic Colloidal and Macro-
molecular Systems." See Nos. 44, 56, 120, 126, 137, 186, 268, 284.
129 Oster, G., and Riley, D. P. A
Acta Cryst., 5, 272-276 (1952). "Scattering from Cylindrically Symmetric
Systems."
130 Penley, H. H., and Gray, J. A. A
Can. J. Research, A-15, 45-47 (1937). "Scattering of X-Rays at Very Small
Angles." Scattering from charcoal and powdered graphite in terms of
refraction and gas-type scattering. See No. 61.
130' Porai-Koshits, Ye. A. A
Uspekhi Khim., 16, 315-319 (1947). "The Scattering of X-Rays at Small
Angles by Submicroscopic Structure. I. Concerning the Scattering of X-Rays
at Small Anglos." (In Russian.) See Nos. 127, 127'.
131 Porai-Koshits, Ye. A. A
Izvest. Akad. Nauk S.S.S.R., Ser. Fiz., 15, 195-201 (1951). "The Study of
Matter in the Dispersed State with the Aid of Small-Angle X-Ray Scattering."
(In Russian.) Uses crystal monochromatization and vacuum camera.
131' Porai-Koshits, Ye. A., Kalinina, A. M., and Phillipovich, V. N. A
Doklady Akad. Nauk, S.S.S.R., 86, 985-988 (1952). "A Study of the Structure
of Several Silica Gels by the Method of X-Ray Scattering at Small Angles."
(In Russian.)
132 Porai-Koshits, Ye. A., and 8okolov, L'. G. A
Doklady Akad. Nauk, S.S.S.R., 72, 477•480 (1950). "Diffuse Scattering of
X-Rays at Small Angles." (In Russian.)
133 Porod, G. T
Univ. Graz, doctoral (1947).
BIBLIOGRAPHY 231

134 Porod, G. A
Acta Phy8. Austriaca, 2, 255-292 (1948). "Die Abhiingigkeit der Rontgen-
kleinwinkelstreuung von Form und Grosse der kolloiden Teilchen in ver-
diinnten Systernen, IV." Calculation of scattering from dilute monodispersed
rods, needles, plates, ellipsoids, cylinders, and spheres_
135 Porod, G. A
Acta Phys. Austriaca, 3, 66-81 (1949). "Abhiingigkeit der Rontgenklein-
winkelstreuung von Form und Anordmmg der kolloiden Teilchen in dicht-
gepackten Systernen." Scattering from proposed platelet structure model for
cellulose fibers.
136 Porod, G. A
Z. Naturforsch., A-4, 401-414 (1949). "Theorie der diffusen Rontgenklein-
winkelstreuung an kolloidcn Systemen." Scattering by (I) dilute systems of
globules, rods, and plates, (2) oriented and unoriented compact systems, and
(3) boundaries and edges.
137 Porod, G. A
Kolloid-Z., 124, 83-ll4 (1951); 125, 51-57, 109-122 (1952). "Die Ri:intgen-
kleinwinkelstreuung von dichtgepackten kolloiden Systemen I, II." See
Nos. 44, 56, 120, 126, 128, 186, 268, 284.
138 Posner, A. P
Am. Dental Assoc., St. Louis (September 1952). "Discussion of the Inorganic
Portion of Human Enamel."
139 Posner, A. S. A
Natl. Bur. Standards Tech. News Bull., 37, 85-86 (1953). "X-Ray Scattering
Study of Tooth Structure.''
139' Posner, A. S. T
Univ. Liege, Ph.D. (1954). "A Study of the Constitution of Certain Mineral-
ogical, Biological, and Synthetic Calcium Phosphates."
139" Posner, A. 8., and Stephenson, S. R. A
J. Am. Dental Assoc., 46, 257-264 (1953). "Isomorphous Substitution in
Enamel Apatite."
140 Raether, H. A
FIAT Rev. Ger. Sci'., 1939-1946, I, 102-108 (1948). "Struktur der festen
Korper, VI. Struktur der diinnen Filmen."
141 Riley, D. P. PA
Brit. Coal Ul'ilization Research Assoc. Conj., London (1944). "The Ultra-Fine
Structure of Coals and Cokes." (Pp. 232-239.) ''The Low-Angle Scattering
of X-Rays by Various Coals." Coals investigated give either liquid-type or
gas-type scattering-a diffuse ring or central scattering.
142 Riley, D. P. PA
Brit. Coal Utilization Research Assoc. Conj., London (1944). "The Ultra-Fine
Structure of Coals and Cokes." (Pp. 246-257.) "The Crystallinity of Certain
High Ash-Content Coals." Both small- and wide-angle diffuse X-ray rings.
143 Riley, D. P. PA
Inst. of Physics, X-Ray Analysis Group, Royal Inst. (April 1950). Nature,
165, 870-872 (1950). "Production and Application of High-Intensity X-Ray
Beams." Conference report includes a few small-angle applications.
144 Riley, D. P. P
Conj. du Palais de la Decouverle, Univ. Paris (1948). "La structure des
charbons."
232 SMALL-ANGLE SCATTERING OF X-RAYS

145 Riseman, J. A
Acta Cryst., 5, 193-196 (1952). "Particle Size Distribution from Small-Angle
Scattering." Theoretical distribution for spheres. See Nos. 6, 147, 155.
145' Ritland, W. N., and Beeman, W.W. P
Pittsburgh Conj. X -Ray Electron Diffr. and ACA, No. 21, Pittsburgh (November
1954). "The Small-Angle Scattering of X-Rays by Several Glasses."
146 Ritter, H. L., and Erich, L. C. A
Anal. Chem., 20, 665-670 (1948). "Pore Size Distribution in Porous Materials.
Interpretation of Small-Angle X-Ray Scattering Patterns." Theory of
small-angle scattering applied to porous aggregates, and results reduced to
continuous distributions of pore size. A semi-empirical method of correcting
for collimation is developed and applied.
146' Robinson, W. H., Li, Y. Y., and Smoluchowski, R. P
Am. Phys. Soc., No. W-9, Detroit (March 1954). "Small-Angle Scattering of
X-Rays by Irradiated Diamond."
146" Robinson, W. H., Li, Y. Y., and Smoluchowski, R. P
Pittsburgh Conj. X-Ray Electron Diffr. and ACA, No. 22, Pittsburgh (November
1954). "Small-Angle Scattering of X-Rays by Diamond." No vacancy
clusters found, but 1500 A surface imperfections.
147 Roess, L. C. A
J. Chem. Phys., 14, 695-697 (1946). "A Simple Method of Obtaining a Particle
Mass Distribution by Inverting the X-Ray Intensity Scattered at Small
Angles." See Nos. 6, 145, 155.
148 Roess, L. C., and Shull, C. G. A
J. Appl. Phys., 18, 308-313 (1947). "X-Ray Scattering at Small Anglos by
Finely Divided Solids. II. Exact Theory for Random Distribution of
Spheroidal Particles." Families of scattering curves for spheroids. Fit
between experimental and calculated scattering curves for mass distribution.
149 Ruess, G. A
Monatsh. Chem., 76, 253-262 (1946-1947). "tlber die Struktur des Glaz-
kohlenstoffs." Includes small-angle scattering.
149' Schmidt, P. T
Univ. Wisc., Ph.D. (1953). Small-angle scattering from turnip yellow mosaic
virus.
149' Shacklett, R. T
Calif. Inst. Tech., Ph.D. (1953). Point-focusing monochromator applications.
150 Sharrah, P. C. P
Conj. Small-Angle X-Ray Scatt,ring, No. 11, Columbia, Mo. (October 1949).
"Surface Coating on Magnesium." Examination of surface film using small-
angle scattering.
151 Shenfil, L., Danielson, W. E., and DuMond, J. W. M.
Tech. Report 15, O.N.R. Contract N60NR-244, T.O. IV, NR017-602. "A
Point-Focusing Monochromator for the Study of Low-Angle X-Ray Diffraction;
Determinations of the Size Distributions of Carbon Blacks and a Precision
Determination of the Size of Dow Latex Particles." See Nos. 25, 229.
152 Shishakov, N. A. A
,Zhur. Eksptl. i. Teoret. Fiz. (J. Exptl. Theoret. Phys. U.S.S.R.), 19, 431-437
(1949). "Displacement of the Interference Maxima with Variation the of Size
and Shape of Very Small Crystals." (In Russian.)
BIBLIOGRAPHY 233

153 Shull, C. G. p
ASXRED and AAAS, Gibson Island, Md. (August 1944). "Some Experi-
mental Studies of Small-Angle Scattering."
154 Shull, C. G., Elkin, P. B., and Roess, L. C. A
J. Am. Chem. Soc., 70, 1410-1414 (1948). "Physical Studies of Gel Micro-
structure." Pore size by adsorption and small-angle scattering.
155 Shull, C. G., and Roess, L. C. A
J. Appl. Phys., 18, 295-307 (1947). "X-Ray Scattering at Small Angles by
Finely Divided Solids. I. General Approximate Theory and Applications."
Review of theory for independent particles. Procedure for obtaining mass
distribution and correcting for slit geometry. Examples. See Nos. 6, 145,
147, 148.
156 Smith, A. E. Pa
Am. Phys. Soc., No. B-2, Los Angeles (January 1947). Phys. Rev., 71, 277
(1947). "Determination of Particle Size by Low-Angle Scattering of X-Rays."
Effect on small-angle scattering of particle shape. Check with line broadening.
Note: item 157 is omitted
158 Stone, L. L., and Yudowitch, K. L. P
Pittsburgh Conj. X-Ray Electron Diffr., No. 11, Pittsburgh (November 1950).
"Applications of a High Resolution Small-Angle Scattering Camera."
Extension by fine geometry to 1/4 micron particles. Application to bacterial
nuclei and latex. See No. 192.
159 Turkevich, J., and Hubbell, H. H., Jr. Pa
Am. Phys. Soc., No. G-12, New York (January 1948). Phys. Rev., 73, 1250
(1948). "Experimental Check on Intensity Distribution in Small-Angle
X-Ray Scattering." Guinier formula checked on wavelength dependence.
Secondary maximum detected.
160 Turkevich, J., and Hubbell, H. H., Jr. P
Am. Chem. Soc., Phys. Inorg. Div., No. 117, Atlantic City (SeptAmber 1949).
"Small-Angle Scattering of X-Rays by Colloidal Particles."
161 Turkevich, J., and Hubbell, H. H., Jr. A
J. Am. Chem. Soc., 73, 1-7 (1951). "Low-Angle X-Ray Diffraction of Colloidal
Gold and Carbon Black."
162 Turkevich, J., Hubbell, H. H., Jr., and Hillier, J. PA
Discussions Faraday Soc., 8, 348-352 (1950). "Electron Microscopy and
Small-Angle X-Ray Scattering." Correlation on catalysts.
162' Tweet, A. G. Pa
Am. Phys. Soc., No. W-2, Washington (May 1953). Phys. Rev., 91, 488-489
(1953). "Small-Angle Scattering of X-Rays from Helium I and Helium II."
163 Vineyard, G. H. Pa
Am. Phys. Soc., No. C-7, Washington (April 1948). Phys. Rev., 74, 1209
(1948). "The Interpretation of Small-Angle X-Ray Scattering Observed
from Argon." Scattering from argon near critical point re-examined for
pretransition phenomena.
164 Vineyard, G. H. A
Phys. Rev., 74, 1076-1083 (1948). "Small-Angle X-Ray Scattering from
Amorphous Materials." Derivation of scattering from inhomogeneities.
165 Vineyard, G. H. P
ASXRED, No. 29, Columbus (December 1948). "Small-Angle Scattering
from Amorphous Materials."
234 SMALL-ANGLE SCATTERING OF X-RAYS

166 Vineyard, G. H. P
Conj. Small-Angle X-Ray Scattering, No. 13, Columbia, Mo. (October 1949).
"Ray Optics for Larger Particles?" Particle size, refractive index, and wave-
length determine whether refraction or diffraction theory is valid.
167 Vineyard, G. H. A
PhyB. Rev., 82, 453 (1951). "Concerning Certain Anomalous Small-Angle
Diffraction Effects."
168 Vineyard, G. H. A
PhyB. Rev., 85, 633-636 (1952). "Geometrical Optics and the Theory of
Multiple Small-Angle Scattering." See Nos. 35, 254, 256, 355.
168' Vineyard, G. H. P
Am. Phys.Soc., No. E-1,NewYork(January 1954). "Small-Angle Scattering."
169 Walker, C. B., Blin, J., and Guinier, A. A
Compt. rend., 235, 254--255 (1952). "Mise en evidence des htlterogeneites
d'une solution solide en equilibre."
170 Walker, C. B., and Guinier, A. A
Compt. rend., 234, 2379-2381 (1952). "Les phenomenas de pre-precipitation
dans l'alliage aluminium-argent." Silver nuclei.
171 Warren, B. E. A
J. Chem. Phys., 2, 551-555 (1934). "X-Ray Diffraction Study of Carbon
Black." Small-angle scattering. Interpretation without any BBBumption as
to crystallinity-but packing density variations.
172 Warren, B. E. Pa
Am. Phys. Soo., No. 132, Washington (April 1936). PhyB. Rev., 49, 885 (1936).
"Small-Angle X-Ray Scattering." C-black and SiO,. Interpreted as scatter-
ing and interference between particles. All atoms of a particle thought in
phase. Particle size measurement discUBBed.
173 Warren, B. E. P
ASXRED, No. 40, Ste. Marguerite Sta., Quebec (June 1947). "Double
Crystal Spectrometer Measurements of Total Energy of Small-Angle Scatter-
i:ng." Specimen moved from between crystals to between second crystal and
counter to obtain total small-angle scattering and hence an average particle
size.
174 Warren, B. E. A
J. Appl. Phys., 20, 96-97 (1949). "A Method for Measuring the Total Power
of Small-Angle X-Ray Scattering." Total power in small-angle scattering
from the difference between two settings of double crystal spectrometer for an
average particle size. See No. 5.
175 West, W. J. P
Soc. Rheology, No. 25, Chicago (October 1951). "Size Determinations of Clay
Particles in Water Suspensions by Use of Low-Angle X-Ray Diffraction."
176 West, W. J. A
J. Colloid Sci., 7, 295-305 (1952). "Size Determinations of Clay Particles in
Water Suspensions by Use of Low-Angle X-Ray Diffraction." Two crystal
spectrometer. Largest particles exhibit most swelling.
177 Whitman, J. T
Fla. State Univ., M.S. (1950). "Accurate Determination of Particle Size by
the Line Broadening Method Using Soft X-Rays; Critical Examination of
Corrections of Such Methods."
BIBLIOGRAPHY 235
178 Whitman, J., and Yudowitch, K. L. PA
Florida Acad. Sci., No. 84 (December 1950). Quart. J. Florida Acad. Sci., 14,
43-46 (1951). "X-Ray Line Breadth Corrections." Crystallite size compared
with particle size.
179 Wild, R. L. T
Univ. Mo., Ph.D. (1950). "The Small-Angle Scattering of X-Rays by
Nitrogen."
180 Wild, R. L. A
J. Chem. Phys., 18, 1627-1632 (1950). "The Small-Angle Scattering of X-Rays
by Nitrogen." Compressibility and inhomogeneity size for nitrogen near
critical point. Compressibility values check; inhomogeneous regions found
to be only 4-8 A. See Nos. 40, 49, 59.
181 Wild, R. L. P
Pittsburgh Conj. X-Ray Electron Diffr., No. I, Pittsburgh (November 1950).
"The Small-Angle Scattering of X-Rays by Nitrogen." Evidence of inhomo-
geneities near critical point.
182 Wood, L. A. T
Stanford Univ., Ph.D. (1939). "The Absolute Amount of Adsorption in the
Surface of Solutions."
183 Yudowitch, K. L. T
Univ. Mo., Ph.D. (1948). "Particle Size Determination by the Scattering of
Soft X-Rays." Secondary maxima observed. Interparticle interference
peaks observed and theory suggested. Optimum collimation scheme outlined.
Examples of colloidal gold with vacuum-soft aluminum X-rays.
184 Yudowitch, K. L. Pa
Electron }\[icr. Soc. Amer., No. 41, Toronto (November 1948). Anal. Chem.,
20, 993 ( 1948). "Particle Size Correlation with X-Ray Methods." Soft X-rays
extend small-angle technique to overlap electron microscopy range. See
No. 186.
185 Yudowitch, K. L. P
Pittsburgh Conj. X-Ray Electron Diffr., No. 18, Pittsburgh (November 1948).
"Particle Size Determination by Soft X-Ray Scattering." See No. 186.
186 Yudowitch, K. L. A
J. Appl. Phys., 20, 174-182 (1949). "Particle Size Determination by Soft
X-Ray Scattering." See Nos. 44, 56, 120, 126, 128, 137, 268, 284.
187 Yudowitch, K. L. A
Phys. Today, 2, 33 (1949). "Measuring Colloidal Piirticles."
188 Yudowitch, K. L. PA
Florida Acad. Sci., No. 65 (December 1949). Quart. J. Florida Acad. Sci., 12,
246-250 (1949). "Particle Size by X-Ray Scattering." Line broadening and
small-angle scattering.
189 Yudowitch, K. L. P
Electron Mier. Soc. Amer., No. 12, Detroit (September 1950). "Latex Size by
X-Rays." See No. 192.
190 Yudowitch, K. L.
Final Report, Chem. Corps, Camp Detrick, Md., Contract W-18-064-CM-229
(1950). "An Investigation of Particle Structure, Packing, and Size
Distribution."
191 Yudowitch, K. L.
Final Report, Chem. Corps, Camp Detrick, Md., Contract DA-18-064-CML-464
(1951). "Investigation of Particle Structure and Size Distribution."
236 SMALL-ANGLE SCATTERING OF X-RAYS

192 Yudowitch, K. L. A
J. Appl. Phys., 22, 214-216 (1951). "Latex Particle Size from X-Ray Diffrac-
tion Peaks." Particle size of spheres from series of secondary maxima.. See
Nos. 23, 115, 151.
193 Yudowitch, K. L. A
Rev.Sci. Instr., 23, 83-90 (1952). "Small-Angle X-Ray Scattering Technique."
Comprehensive review of design for peak and slope analysis.
194 Yudowitch, K. L. P
AOA, Te.miment, Pe.. (June 1952). "Interparticle Interference Effects."
See No. 56.
195 Yudowitch, K. L. P
Am. Chem. Soc., Anal. Div., No. 35, Atlantic City (September 1952). "Small-
Angle Scattering Technique, Interpretation, and Applications." See No. 196.
196 Yudowitch, K. L. A
Anal. Chem., 25, 721-724 (1953). "Low-Angle X-Re.y Scattering Technique,
Interpretation, and Application." Elementary review.
197 Yvon, M. J. A
J. phys. radium, 7, 201-202 (1946). "Le coefficient de compressibilite des
fluides deduit du die.gramme diffusion des re.yona X." Liquid compressibility
from small-angle data.

Il. GEOMETRY
198 Alexander, L. P
Pittsburgh Conj. X-Ray Electron Diffr., Pittsburgh (November 1947).
Geometry.
199 Alexander, L. A
J. Appl. Phys., 19, 106S--1071 (1948). "Geometrical Factors Affecting the
Contours of X-Re.y Spectrometer Maxima. I. Factors Causing Asymmetry."
3rd dimension error.
200 Alexander, L. A
J. Appl. Phys., 21, 126--136 (1950). "Geometrical Factors Affecting the
Contours ofX-Ray Spectrometer Maxima. II. Factors Ce.using Broadening."
The pure diffraction contour is broadened by X-ray source width, flat rather
than curved sample surface, vertical divergence of the X-ray beam, penetration
of the sample by the beam, receiving slit width.
200' Alexander, L. A
Brit. J. Appl. Phys., 4, 92-93 (1953). "The Effect of Vertical Divergence on
X-Re.y Powder Diffraction Lines."
201 Alexander, L., and Klug, H. P. P
ASXRED, Ithaca, N.Y. (June 1949). Geometry.
202 Alexander, L., and Klug, H.P. A
J. Appl. Phys., 21, 137-142 (1950). "Determination of Crystallite Size with
the X-Re.y Spectrometer." Convolution analysis for effect of geometrical
factors in broadening pure diffraction contour.
203 Bergman, M. E., and Fe.nkuchen, I. A
Rev. Sci. Instr., 20, 696 (1949). "Modification of X-Re.y Diffraction Micro-
Camera to Permit Study of Long Spacings."
204 Bertaut, F. A
Oompt. rend., 228, 1597-1599 (1949). "Methode de correction de le. forme des
re.ies Debye-Scherrer." Correction of line profile.
BIBLIOGRAPHY 237
205 Bolduan, 0. E. A., and Bear, R. S. A
J. Appl. Phys., 20, 983-992 (1949). "Effective Use of Collimating Apertures
in Small-Angle X-Ray Diffraction Cameras." Optimum conditions for pinhole
and slit aperture systems are developed. Construction and performance of
some cameras for periodic structures.
206 Broussard, L. A
Rev. Sci. Instr., 21, 399-400 (1950). "Improved Double Crystal Spectrometer
for Small-Angle Scattering Measuring." Second crystal and counter rotate
rigidly about specimen. See Nos. 8, 209, 212, 226, 238.
207 Clark, G. L., and Corrigan, K. E. A
Ind. Eng. Chem., 23, 815-820 (1931). "Industrial and Chemical Research
with X-Rays of High Intensity and with Soft X-Raya." Use of longer wave-
lengths to increase diffraction angles: Mg target.
208 Davis, B., and Slack, C. M. A
Phys. Rev., 27, 18-22 (1926). "Measurement of the Refraction of X-Rays in a
Prism by Means of the Double X-Ray Spectrometer." Double crystal spectra·
meter used to measure small angular deflection. See No. 230.
208' Despujols, J. A
Compt. rend., 235, 716-718 (1952). "Monochromateur pour rayons X B. foyer
ponctuel."
209 DuMond, J. W. M. A
Phys. Rev., 72, 83-84 (1947). "Method of Correcting Low-Angle X-Ray
Diffraction Curves for the Study of Small Particle Sizes." Two crystal
techniques described, correction for slit height. See Nos. 8, 173, 206, 208,
212, 226, 230, 238.
210 DuMond, J. W. M. A
Rev. Sci. Instr., 21, 188-189 (1950). "Point Focus Monochromators for Low-
Angle Diffraction." Two types of point-focusing crystal arrangements
-described.
211 DuMond, J. W. M., Shenfil, L., Danielson, W., and Henke, B. P
Pittsburgh Can/. X-Ray Electron Diffr., No. 10, Pittsburgh (1950). "Anastig-
matic Point-Focusing X-Ray Monochromator for Cu Ka. Radiation for Use in
Low-Angle X-Ray Diffraction Studies."
212 Fankuchen, I., and Jellinek, M. H. Pa
Am. Phys. Soc., No. D-5, New York (January 1945). Phys. Rev., 67, 201
(1945). "Low-Angle X-Ray Scattering." Describes a double crystal method
for obtaining scattering. Geiger counter. Scattering at extremely small
angles can be studied. Specimen between crystals in parallel and antiparallel
arrangements. See Nos. 8, 173, 206, 208, 209, 226, 230, 238.
212' Finean, J. B. A
J. Sci. Instr., 30, 60-61 (1953). "A Versatile Camera for Low-Angle Diffraction
Studies."
213 Fournier, F. A
Compt. rend., 227, 833-834 (1948). "Sur la realisation et !'utilisation de
faisceaux de rayons X tres fins, de l'ordre de quelques microns." Glass capillaries.
214 Franklin, R. E. A
Acta Cryst., 3, 158-159 (1950). "A Rapid Approximate Method for Correcting
Low-Angle Scattering Measurements for the Influence of the Finite Height of
the Beam."
238 SMALL-ANGLE SCATTERING OF X-RAYS

215 Gokhale, B. G. A
Compt. rend., 230, 636-638 (1950). "Sur l'effet de la largeur de fente du
microphotometre sur la largeur enregistree des raies dans !es spectres des
rayons X."
216 Guinier, A., and Fournet, G. A
J. phys. radium, 8, 345-351 (1947). "Facteurs de correction dans los mesures
de la diffusion des rayons X aux faibles angles." Consideration of error in
small-angle pattern due to finite cross section of incident beam.
217 Guinier, A., and Fournet, G. A
Nature, 160, 501 (1947). "Correction of Measurements of Low-Angle X-Ray
Scattering." The correction for the finite height of the monochromator beam
is worked out for spherical and ellipsoidal particles, and shown graphically.
218 Guinier, A., and Fournet, G. P
CSA-ASXRED, New Haven, Conn. (1948). "A Two-Curved Crystal Mono-
chromator for the Study of Low-Angle Scattering."
219 Guinier, A., and Fournet, G. A
Compt. rend.,· 226, 656-659 (1948). "Emploi d'un monochromateur double
pour l'etude de la diffusion des rayons X aux tres faibles angles." A slit is
placed at the focus of the first curved crystal monochromator, and the rays
focused on the film by the second monochromator. Background is greatly
reduced, at the expense of increase in exposure time. Spacings to 250 A.
219' Hagg, G., and Karlsson, N. A
Acta Cryst., 5, 728-730 (1952). "Aluminum Monochromator with Double
Curvature for High-Intensity Powder Photographs."
220 Hubbell, H. H., Jr. Pa
Am. Phys. Soc., No. G-13, Chicago (December 1947). Phys. Rev., 73, 1251
(1948). "The Effect of Apparatus Geometry on Small-Angle X-Ray Scatter-
ing Curves." Smearing of secondary maxima by slits.
220' Huxley, H. E. A
Acta Cryst., 6, 457-465 (1953). "Optimum Geometric Conditions in the
Design and Use of X-Ray Diffraction Tubes and Cameras." See Nos. 125,
205, 241.
221 Kaesberg, P., Beeman, W. W., and Ritland, H. N. Pa
Am. Phys. Soc., No. OB-4, New York (February 1950). Phys. Rev., 78, 336
(1950). "Double Crystal and Slit Methods in Small-Angle X-Ray Scattering."
Resolution of about lO seconds obtained using a calcite double crystal spectra·
meter. Wing background at larger angles makes detection of weak scattering
difficult. Method inferior to a slit system for low resolution requirement.
See No. 226.
222 Kahovek, L., and Treiber, E. A
Z. Elektrochem., 55, 437-439 (1951). "tlber eine Rontgenkiivette fiir Klein-
winkelaufnahmen.''
222' Kiessig, H. A
Kolloid-Z., 98, 213-221 (1942). "Rontgenuntersuchung grosser Netzebenen-
abstande und Untersuchung stromender Losungen." Describes apparatus.
223 Lely, J. A., and Van Rijssel, T. W. A
Acta Cryst., 2, 337-338 (1949). "X-Ray Collimator Producing a Beam of Very
Small Divergence and Large Intensity." Slits constructed of polished glass
blocks with a critical angle Jess than 15' produce strong beams of small
divergence.
BIBLIOGRAPHY 239

224 Lipson, H., Nelson, J.B., and Riley, D. P. A

J. Sci. Instr., 22, 184-187 (1945). "Monochromatic X-Radiation."
225 Ritland, H. N., Kaesberg, P., and Beeman, W. W. P
Conj. Small-Angle X-Ray Scattering, No. 3, Columbia, Mo. (October 1949).
"Double Crystal and Slit Methods in Small-Angle X-Ray Scattering."
Intensity wings with double crystal spectrometer examined; reduced by third
crystal.
226 Ritland, H. N., Kaesberg, P., and Beeman, W. W. A
J. Appl. Phys., 21, 838-841 (1950). "Double Crystal and Slit Methods in
Small-Angle X-Ray Scattering." Experimental comparison of slit vs. crystal
collimation. Intensity wings from crystal examined. See Nos. 8, 209, 212.
227 Shenfil, L. T
Calif. Inst. Tech., Ph.D. (1952). "Point-Focusing X-Ray Monochromator for
Low-Angle Scattering Studies."
228 Shenfil, L., Danielson, W. E., and DuMond, J. W. M. Pa
Am. Phys. Soc., No. B-5, Houston (November 1951). Phys. Rev., 85, 739
(1952). "A Point-Focusing Monochromator for Low-Angle X-Ray Diffraction."
229 Shenfil, L., Danielson, W. E., and DuMond, J. W. M. A
J. Appl. Phys., 23, 854-859 ( 1952). "A Point-Focusing X-Ray Monochromator
for the Study of Low-Angle Diffraction." See Nos. 25, 151, 438.
230 Slack, C. :M. A
Phys. Rev., 27, 691-695 (1926). "Refraction of X-Rays in Prisms of Various
Materials." Double crystal spectrometer for beam contour. Granular
material. See No. 208.
231 Solomon, E., and Jellinek, M. H. A
Rev. Sci. Instr., 15, 348-349 (1944). "An Inexpensive Densitometer for the
Evaluation of Low-Angle X-Ray Scattering Films."
232 Spencer, R. C. A
Phys. Rev., 38, 618-629 (1931). "Additional Theory of the Double X-Ray
Spectrometer.''
233 Spencer, R. C. A
Phys. Rev., 48, 473 (1935). "The Effect of the Spectrometer on the Width of
Spectral Lines."
234 Spencer, R. C. A
Phys. Rev., 52, 761 (1937). "A Theorem on the Effect of Vertical Divergence."
235 Spencer, R. C. Pa
Am. Phys. Soc., No. 31, Chicago (November 1938). Phys. Rev., 55, 239 (1939).
"The Correction of Experimental Curves for the Resolving Power of the
Apparatue. ''
236 Spencer, R. C. Pa
Am. Phys. Soc., No. 33, Providence (June 1941). Phys. Rev., 60, 172 (1941).
"Optimum Design of Physical Apparatus."
237 Spencer, R. C. A
J. Appl. Phys., 20, 413-414 (1949). "Discussion of Geometrical Factors
Affecting X-Ray Spectrometer Maxima."
238 Wild, R. A
Rev. Sci. Instr., 22, 537-538 (1951). "Automatic Focusing Condition for
Double Crystal Spectrometer." See Nos. 8, 206, 209, 212, 226.
239 Yudowitch, K. L. Pa
Am. Phys. Soc., No. B-1, Cambridge, Mass. (June 1949). Phys. Rev., 76, 455
(1949). "Collimation Error in Small-Angle X-Ray Scattering."
240 SMALL-ANGLE SCATTERING OF X-RAYS

240 Yudowitch, K. L. p
Conj. Small-Angle X-Ray Scattering, No. 4, Columbia, Mo. (October 1949).
"Proper Design for Minimizing Collimation Error in Small-Angle Scattering."
241 Yudowitch, K. L. A
J. Appl. Phya., 20, 1232-1236 (1949). "Collimation Error in Small-Angle
X-Ray Scattering." Minimization of range of angles for rays reaching a single
point on film leads to optimum slit shape and spacing.
242 Yudowitch, K. L. P
ACA, No. 1, Chicago (October 1951). "Powder Pattern Geometry."
243 Y udowitch, K. L. A
J. Opt. Soc. Amer., 42, 437-440 (1952). "Diffraction Geometry."
244 Yudowitch, K. L. A
J. Opt. Soc. Amer., 43, 50-52 (1953). "Recording Radially Symmetric
Scattering."

m. NEUTRONS
245 Halpern, 0. Pa
Am. Phya. Soc., No. H-3, Houston (November 1947). Phya. Rev., 73, 653
(1948). "Small-Angle Diffraction in Lattices." Theoretical paper on small-
angle scattering of light and material waves.
246 Halpern, 0. P
Pittsburgh Conj. X-Ray Electron Diffr., No. 2, Pittsburgh (November 1950).
"Small-Angle Diffraction of Neutrons." Theoretical consideration of scatter-
ing cross sections for neutrons,
247 Halpern, 0., and Gerjouy, E. Pa
Am. Phys. Soc., No. H-5, Pasadena (June 1948). Phys. Rev., 74, 1562 (1948).
"On Diffraction under Small Angles by Three-Dimensional Lattices." General
theory for small-angle scattering cross sections for light, X-rays, and neutrons.
248 Halpern, 0., and Gerjouy, E. A
Phys. Rev., 76, 1117-1129 (1949). "Small-Angle Diffraction of Neutrons and
Similar Wave Phenomena." Diffraction cross-section dependence on nature
and orientation of scatterer, and on scattering angle.
249 Heller, R. B. P
Conj. Small-Angle X-Ray Scattering, No. 9, Columbia, Mo. (October 1949).
"Small-Angle Scattering of Neutrons at Domain Boundaries in Magnetized
Iron." Multiple magnetic refraction of neutrons in iron.
250 Hughes, D. J., Burgy, M. T., Heller, R. B., and Wallace, J. W. A
Phya. Rev., 75, 565-569 (1949). "Magnetic Refraction of Neutrons at Domain
Boundaries." Refraction of neutrons at magnetic domain boundaries.
251 Krueger, H. H. A., Meneghetti, D., Ringo, G. R., and Wensberg, L. A
Phys. Rev., 80, 507-510 (1950). "Small-Angle Scattering of Thermal
Neutrons." Experimental examination of (1) Halpern and Gerjouy's, (2) Van
de Hulst's, and (3) von Nardroff's theories.
252 McReynolds, A. W., and Weiss, R. J. Pa
Am. Phya. Soc., No. V-6, Chicago (November 1950). Phya. Rev., 81, 326
(1951). "Small-Angle Scattering of Thermal Neutrons (Experimental)."
253 Meneghetti, D. Pa
Am. Phys. Soc., No. X-7, Washington (April 1949). Phys. Rev., 76, 188-189
(1949). "Small-Angl1> Scattering of Neutrons." Scattering in agreement with
X-ray refraction theory.
BIBLIOGRAPHY 241
254 Meneghetti, D. p
Pittsburgh Conj. X-Ray Electron Diffr., No. 5, Pittsburgh (November 1950).
"Small-Angle Scattering of Thermal Neutrons." Multiple scattering of
neutrons agrees in range of study with both refraction and diffraction theories.
See Nos. 35, 168, 2.56, 355.
254' Shull, C. G., and Wilkinson, M. K. p
Am. Phys. Soc., No. Y-2, Detroit (March 1954). "Small-Angle Neutron
Scattering by Iron at High Temperatures."
255 Weiss, R. J. p
ACA, No. H-2, New Hampton, N.H. (August 1950). "Neutron Small-Angle
Scattering.''
256 Weiss, R. J. T
N.Y.U., Ph.D. (1950). "Small-Angle Scattering of Neutrons." Verification
of diffraction theory for small phase change in traversing particle, and of
refraction theory for large phase change in traversing particle. See Nos. 35,
168, 254. 355.
257 Weiss, R. J. A
Phys. Rev., 83, 379-389 (1951). "Small-Angle Scattering of Neutrons."
258 Weiss, R. J., and Pasternack, S. Pa
Am. Phys. Soc., No. V-7, Chicago (November 1950). Phys. Rev., 81, 326
(1951). "Small-Angle Scattering of Thermal Neutrons (Theoretical)." Phase
change in neutron in traversing particle to determine validity of refraction or
diffraction theory.

IV. BASIC THEORY

259 Born, M., and Green, H. S. A
Proc. Roy. Soc. (London), A-188, 10-18 (1946). "A General Kinetic Theory
of Liquids. I. The Molecular Distribution Functions."
260 Born, M., and Green, H. S. A
Proc. Roy. Soc. (London), A-190, 455-473 (1947). "A General Kinetic Theory
of Liquids. III. Dynamical Properties."
261 Born, M., and Green, H. S. A
Proc. Roy. Soc. (London), A-191, 168-181 (1947). "A General Kinetic Theory
of Liquids. IV. Quantum Mechanics of Fluids."
262 Corner, J., and Lennard-Jones, J.E. A
Proc. Roy. Soc. (London), A-178, 401-414 (1941). "The Neighbor Distribution
Function in Monatomic Liquids and Dense Gases."
263 Debye, P. A
Ann. Physik, 43, 49-95 (1914). "Interferenz von Rontgenstrahlen und
Warmbewegung."
264 Debye, P. A
Ann. Physik, 46, 809-823 (1915). "Zerstreuung von Rontgenstrahlen."
Scattering from non-crystalline substances.
265 Debye, P. A
Physik. Z., 28, 135-141 (1927). "t:'ber die Zerstreuung von Rontgenstrahlen
an amorphen Korpern."
266 Debye, P. A
Physik. Z., 31, 348-350 (1930). "Rontgenzerstreuung an Fliissigkeiten und
Gasen." Diffraction by microscopic drop of liquid; large angles due to mole-
cular arrangement, immerumrably small angles due to boundary of drop.
242 SMALL-ANGLE SCATTERING OF X-RAYS

267 Debye, P. A
Physik. Z., 31, 419-428 (1930). "Ri:intgeninterferenzen und Atomgri:isse."
Distribution of scattered X-ray intensity from polyatomic gases makes it
necessary to treat atoms as extended scattering domains.
268 Debye, P. A
J. Math. and Phys., 4, 133-147 (1925). "Note on the Scattering of X-Rays."
Liquid and gas scattering theory.
269 Debye, P., and Menke, H. A
Physik. Z., 31, 797-798 (1930). "Bestimmung der inneren Struktur von
Fliissigkeiten mit Rontgenstrahlen." Quasi-crystalline structure for mercury.
270 Debye, P., and Menke, H. B
Akademische Verlagsgesellschaft, Leipzig (1931). Ergebnis der technische
Rontgenkunde. II. Method of Zernicke and Prins extended to liquid scattering.
271 Gingrich, N. S. A
Revs. Mod. Phys., 15, 90-110 (1943). "The Diffraction of X-Rays by Liquid
Elements." Inversion of liquid scattering data. Some patterns show
low-angle intensity.
272 Gingrich, N. S., and Warren, B. E. A
Phys. Rev., 46, 248-251 (1934). "The Interpretation of X-Ray Diffraction
Patterns of a Fluid at Various De'nsities." The method of Zernicke and Prins
applied to obtain scattering from a fluid at various densities for assumed radial
densities.
273 Green, H. S. A
Proc. Roy. Soc. (London), A-189, 103-117 (1947). "A General Kinetic Theory
of Liquids. II. Equilibrium Properties." See Nos. 259, 260, 261.
274 Guinier, A. A
Acta Cryst., 5, 121-130 (1952). "Interpretation de la diffusion anormale des
rayons x par !es alliages a durcissement structural."
275 Harvey, G. G. A
Phys. Rev., 46, 441-445 (1934). "The Effect of Pressure on the Intensity of
X-Rays Scattered by Nitrogen at Small Angles." Scattering curves at various
pressures resemble those of a liquid or solid. Concluded due to the closeness
of scattering centers rather than preferred orientation.
276 Hosemann, R. A
Z. Physik, 127, 16-39 (1950). "Rontgeninterferenzen an Strichgittem mit
Fliissigkeitsunordnung.''
277 Hosemann, R. A
Z. Physik, 128, 1-35 ( 1950). "Rontgeninterferenzen an Stoffen mit fliissigkeits-
statistischen Gittersti:irungen.''
278 Hosemann, R. A
Z. Elektrochem., 54, 23-26 (1950). "Kristallitgrossenbestimmung in Hoch·
molekularen. Grenzen der Laueschen Methode."
279 Hosemann, R., and Bagchi, S. N. A
Acta Cryst., 5, 612-614 (1952). "The Interference Theory of Ideal Para.
crystals."
280 Kachkuruzov, G. A., and Todes, 0. M. A
Doklady Akad. Nauk, S.S.S.R., 66, 77-79 (1949). "Principles of X-Ray
Diffraction Analysis of Polycrystalline Materials." (In Russian.) Review of
methods for particle sizes in polydisperse systems.
BIBLIOGRAPHY 243

281 Kirkwood, J. G. Pa
Am. Chem. Soc., New York (April 1935). J. Chem. Phys., 3, 300-313 (1935).
"Statistical Mechanics of Fluid Mixtures."
282 Kirkwood, J. G. A
J. Chem. Phys., 7, 919-927 (1939). "Molecular Distribution in Liquids."
283 Kirkwood, J. G., and Boggs, E. M. A
J. Chem. Phys., 10, 394-402 (1942). "The Radial Distribution Function in
Liguids."
284 Kirkwood, J. G., Maun, E. K., and Adler, B. J. A
J. Chem. Phys., 18, 1040-1047 (1950). "Radial Distribution Functions and the
Equation of State of a Fluid Composed of Rigid Spherical Molecules."
285 Kirkwood, J. G., and Mazur, J. PA
lntern. Union Chem., Strasbourg (June 1952). J. Polymer Sci., 9, 519-524
(1952). "The Radial Distribution Functions of Electrically Charged Macro-
molecules in Solution."
286 Kottler, F. A
J. Franklin lnst., 250, 339-356, 419-442 (1950). "The Distribution of Particle
Sizes I, II." Distribution law from crystal growth law. Use of weighted
squares.
287 Krishnamurti, P. A
Indian J. Phys., 2, 491-500 (1927-1928). "X.Ray Diffraction and Its
Bearing on the Molecular Complexity in the Liquid State."
288 Krishnamurti, P. A
Indian J. Phys., 2, 501-507 (1927-1928). "X-Ray Diffraction in Solutions and
Liquid Mixtures, Part I."
289 Krishnamurti, P. A
Indian J. Phys., 3, 209-223 (1928-1929). "The Diffraction of X-Rays by
Aqueous Solutions of Canesugar, Levulose, and Glucose."
290 Krishnamurti, P. A
Indian J. Phys., 3, 225-240 (1928-1929). "X-Ray Diffraction of Crystal
Powders and Liquids in Relation to Their Constitution."
291 Krishnamurti, P. A
Indian J. Phys., 3, 307-329 (1928-1929). "On the Nature of Dextrin, Gelatin,
and Sodium Oleate Solutions as Revealed by X-Ray Diffraction."
292 Krishnamurti, P. A
Indian J. Phys., 3, 331-355 (1928-1929). "X.Ray Diffraction in Liquid
Mixtures." Rings from binary mixtures.
293 Krishnamurti, P. A
Indian J, Phys., 3, 507-522 (1928-1929). "X.Ray Diffraction in Liquids and
Solutions, and the Molecular Structure Factor."
294 Krishnamurti, P. A
Indian J. Phys., 4, 99-108 (1929-1930). "X-Ray Diffraction by Amorphous
Solids." Diffuse rings shift as amorphous resins change physical state.
295 Krishnamurti, P. A
Indian J. Phys., 5, 473-500 (1930). "Studies in X-Ray Diffraction. Part I:
The Structure of Amorphous Scattering. Part II: Colioidal Solutions and
Liquid Mixtures." Particle size and molecular weights related to extent of
small-angle scattering.
296 LaMer, V. K., and Sinclair, D. A
0.S.R.D. Report 1857. "VerificationofMieTheory."
244 SMALL-ANGLE SCATTERING OF X-RAYS

297 Mathematical Tables Project (National Bureau of Standards). Columbia

Univ. Press, New York (1947). Tables of Spherical Bessel Functions. (Vol. 1:
pp. 25-47.) 3/2 order Bessels for argument 0 to 25.
298 Oster, G. A
Rec. trav. chim., 68, 1123-1136 (1949). "Visible Light and X-Ray Scattering
by Concentrated Solutions of Macromolecules." Interparticle interference for
spheres and other shapes.
299 Patterson, A. L. A
Z. Krist., 90, 517-542 (1935). "A Direct Method for the Determination of the
Components of Interatomic Distances in Crystals."
300 Patterson, A. L. Pa
Am. Phys. Soc., No. 128, Washington (April 1936). Phys. Rev., 49, 884 (1936).
"The Determination of the Size and Shape of Crystal Particles by X-Rays.''
Interference functions for several shapes.
301 Patterson, A. L. A
Phys. Rev., 56, 972-977 (1939). "The Diffraction of X-rays by Small Crystal-
line Particles.'' Ca.lculation of the shape functions for spheroids and various
polyhedra.
301' Prins, J. A. A
Physica, 6, 315-325 (1926). "Diffractie van Riintgenstre.len door Kwik.''
302 Prins, J. A. A
Naturwissenschaften, 19, 435-442 (1931). "Die Molekillanordnung in Fltissig-
keiten und die damit zuse.mmenhiingenden Beugungserscheinungen." Re.dial
distributions and scattering to fairly small angles.
303 Prins, J. A. A
Z. Naturforsch., 6a, 276-277 (1951). "Der kolloidale Stand." Close-packed
colloidal crystallites.
304 Re.man, C. V., and Krishne.murti, P. A
Nature, 124, 53-54 (1929). "A New X-Ray Effect.'' Centre.I diffuse scattering
observed on graphite. Interpreted as due to free electrons in crystal structure.
305 Raman, C. V., and Ramanathan, K. R. A
Proc. Indian Assoc. Cultivation Sci., 8, 127-162 (1932). "Diffre.ctionofX-Re.ys
in Liquids, Fluid Crystals, and Amorphous Solids." A fluid cannot be treated
as a continuum for X-ray scattering except at very small angles. An equation
is derived which takes account only of the average structure of the fluid.
305' Randall, J. T. B
Chapman and Hall, London (1934). The Diffraction of X-Rays aml Electrons
by Amorphous Solids, Liquids, and Gases.
306 Stewart, G. W. A
Revs. Mod. Phys., 2, 116-122 (1930). "X-Ray Diffraction in Liquids." A
review of experimental results.
307 Warren, B. E. A
J. Appl. Phys., 8, 645-654 (1937). "X-Ray Determination of the Structure of
Liquids and Glass.'' Effect of physical discontinuities in structure on sme.ll-
angle scattering discussed. Si0 1 example qualitatively discussed.
308 Warren, B. E., and Biscoe, J. A
J. Am. Oeram. Soc., 21, 49-54 (1938). "The Structure of Silica Glass by
X-Ray Diffraction Studies."
309 Zernicke, F., and Prins, J. A. A
Z. Physik, 41, 184-194 (1927). "Die Beugung von Riintgenstrahlen an
Fltissigkeiten a.ls Effekt der Molekillanordnung.''
BIBLIOGRAPHY 245

V. LIGHT
310 Bethe, H. A., and Rohrlich, F. A
Phys. Rev., 86, 10--16 (1951). "Small-Angle Scattering of Light by a Coulomb
Field."
311 Blumer, H. A
Z. Physik, 32, 119-134 (1925). "Strahlungsdiagramme kleiner dielektrischer
Kugeln. I."
312 Blumer, H. A
Z. Physik, 38, 304-328 (1926). "Strahlungsdiagramme kleiner dielektrischer
Kugeln. II."
313 Blumer, H. A
Z. Physik, 38, 920--947 (1926). "Die Zerstreuung des Lichtes an kleinen
Kugeln."
314 Blumer, H. A
Z. Physik, 39, 195-214 (1926). "Die Farbenzerstreuung an kleinen Kugeln."
315 Bommel, H. A
Helv. Phys. Acta, 21, 289-298 (1948). "Lichtstreuung von den durch Utraschall
orientierten Kolloidalteilchen."
316 Brillouin, L. A
J. Appl. Phys., 20, 1110-ll25 (1949). "The Scattering Cross Section of
Spheres for Electromagnetic Waves."
317 Bromwich, T. J. A
Philos. Trans., A-220, 175-206 (1920). "Scattering of Plane Electric Waves
by Spheres."
318 Bueche, A. M. A
J. Am. Chem. Soc., 74, 2373 (1952). "Size Distribution of Spheres by Light
Scattering."
319 Bueche, A. M., and Debye, P. Pa
Am. Phys. Soc., No. N-5, New York (January 1949). PhyB. Re11., 75, 1308
(1949). "Scattering by Inhomogeneous Solids." See No. 322.
320 Bueche, A. M., and Debye, P. Pa
Am. Phys. Soc., No. G-4, Chicago (November 1950). Phys. Rev., 81, 3030
(1951). "A Study of Crystalline Sizes in Polymers by a Light Scattering
Method."
321 Debye, P. A
Ann. Physik, 335, 57-136 (1909). "Der Lichtdruck aufKugeln von behebigem
Material." From dissertation at Munich.
322 Debye, P., and Bueche, A. M. A
J. Appl. Phys., 20, 518-525 (1949). "Scattering by an Inhomogeneous
Solid." Scattering measurements used to obtain the average square of the
fluctuation in refractive index or electron density and a correlation function
which measures the degree of correlation between two fluctuations as a function
of this distance of separation. See No. 319.
323 Debye, P., and Bueche, A. M. B
Reinhold, New York (1950). Colloid Chemistry, Theoretical and Applied
(J. Alexander). "Scattering by Inhomogeneous Materials." (Vol. 7, pp.
33-46.)
324 De Vore, J. R., and Pfund, A.H. A
J. Opt. Soc. Amer., 37, 826-832 (1947). "Optical Scattering by Dielectric
Powders of Uniform Particle Size." Experimental verification of Mie minima
in the scattering from uniform particles.
246 SMALL-ANGLE SCATTERING OF X-RAYS

325 Engelhard, V. H., and Friess, H. A

Kolloid-Z., 81, 129-143 (1937). "Uber die Brauchbarkeit des Mieeffectes zur
Teilchengrosseitbestinunung weisser Aerosole, insbesondere grossdisperser
weisser Aerosolc. ''
326 Forrester, A. T., and Mittcnthal, L. A
Phys. Rev., 81, 268-271 (1951). "Small-Anglo Diffraction of Light by a Glass
Cube." Results incompatible with Halpern and Gerjouy theory. Born
approximation apparently unjustified.
327 Gans, R. A
Ann. Physik, 76, 29-38 (1925). "Strahlungsdiagramme ultramikroskopi-
scher Teilchen."
328 Hart, R. "\V., and Montroll, E. "\V. A
J. Appl. Phys., 22, 376-386 (1951). "On the Scattering of Plane Waves by
Soft Obstacles. I. Spherical Obstacles."
329 Humphreys, W. J. B
Lippincott, Philadelphia (1940). Physics of the Afr. (Pp. 547-554.) Optical
corona discussed; analogous to X-ray secondary maxima.
330 Jobst, V. G. A
Ann. Physik, 76, 863-888 (1925). "Zur Farbentheorie kolloidaler Metall-
suspensionen.''
331 Johnson, I., and La}for, V. K. A
J. Am. Chem. Soc., 69, 1184-1193 (1947). "The Determination of the Particle
Size of Monodisperscrl iiysiems Ly the Scattering of Light."
332 Kenyon, A. ii., and La~Ier, V. K. Pa
Am. Chem. Soc., New York (September 1947). J. Colloid Sci., 4, 163-184 (1949).
"Light Scattering Properties of }Ionodispersed Sulphur Sols. I. Mono-
chromatic l"ltraYiolct Angular Scattering. II. Effect of the Complex Index
of Refraction upon Transmittance." See No. 350.
333 La:\for, V. K., and Barnes, M. D. A
J. Colloid Sci., 1, 71-77 (1946). "Monorlispersed Hydrophobic Colloidal
Dispersions and Light Scattering Properties. I. Preparation and Light
Scattering Properties of Monodispersed Colloidal Sulphur."
334 LaMer, V. K., and Barnes, M. D. A
J. Colloid Sci., l, 79-91 (1946). "Monodispersed Hydrophobic Colloidal
Dispersions anrl Light Kcattering Properties. II. Total Scattering from
Transmittance as a Basis for tho Calculation of Particle Size and Concentra-
tion."
335 Ljunggren, T. A
Arkiv 11fot_ Astron. Fysik, A-.36, No. 14, 1-36 (Hl48). "Contributions to the
Theor,v of Diffraction of Elect.romagnetic ·waves by Spherical Particles."
336 LoYe, A. E. H. A
Proc. London .J1ath. Soc., .JO, 308-321 (1899). "The Scattering of Electric
"\VaYes by a Dielectric Sphere."
337 Mie, G. A
Ann. Physik, 2.5, 3i7-445 (1908). "Beitrage zur Optik triiber Medien,
speziell kolloidaler Metallosungen."
338 Mitra, M. CIT. A
Indian J. Phys., 3, 175-180 (1928-1929). "Photographs of Coronas in Mono-
chromatic Light." "\Vater droplets-multiple rings.
BIBLIOGRAPHY 247

339 Natl. Bur. Standards, Appl. Math. Series, No. 4 (1949). U.S. Govt. Printing
Office, Washington, D.C. "Tables of Scattering Functions for Spherical
Particles."
340 Pijper, A. A
Med. J. S. Africa, 14, 211-218 (1918). "Diffraction Phenomena in Cultures of
Microorganisms.''
341 Preston, T. B
Macmillan, New York (1924). Theory of Light. (Pp. 228-236.) Theory of
corona rings in light scattering, similar to X-ray secondary maxima.
342 Rayleigh, Lord. B
Cambridge Univ. Press, London (1899). Scientific Papers. (Vol. I, pp. 92-93,
Vol. IV, p. 400.)
343 Rayleigh, Lord. A
Proc. Roy. Soc. (London), A-84, 25-46 (1911). "The Incidence of Light upon
a Transparent Sphere of Dimensions Comparable with the Wavelength."
344 Ruedy, R. A
Can. J. Research, A-19, 117-125 (1941). "Absorption of Light and Heat
Radiation by Small Spherical Particles. I. Absorption of Light by Carbon
Particles."
345 Ruedy, R. A
Can. J. Research, A-21, 79-88 (1943). "Absorption of Light by Small Drops of
Water."
346 Ruedy, R.
Can. J. Research, A-21, 99-109 (1943). "Scattering of Light by Small Drops
of Water."
347 Ruedy, R. A
Can. J. Research, A-22, 53-66 (1944). "Transmission of Light by \Vater Drops
1 to 5 Microns in Diameter."
348 Ryde, J. N., and Cooper, B. S. A
Proc. Roy. Soc. (London), A-131, 451-475 (1931). "Scattering of Light by
Turbid Media." Particle size determination in opal glasses by means of light
diffusion.
349 Sinclair, D. A
J. Opt. Soc. Amer., 37, 475-480 (1947). "Light Scattering by Spherical
Particles." Test of Mie and Rayleigh equations.
350 Sinclair, D., and LaMer, V. A
Chem. Revs., 44, 245-267 (1949). "Light Scattering as a Measure of Particle
Size in Aerosols. The Production of )'[onodisperscd Aerosols." Light
scattering by aerosols and determinations of particle size by methods based on
measurement of (l) transmission, (2) intensity, (3) color, (4) polarization.
Technique for the production of monodispersed aerosols. See No. 332.
350' Sivarajan, S. R. A
Proc. Indian Acad. Sci., A-37, 418-423 (1953). "Light Scattering in Gold
Sols."
35 l Storruste, A. A
K. Norske Videnskab. Selskabs Forh., 21, 88-91 (1948). "Scattering of Waves
by Circular Discs."
352 Stratton, J. A. B
McGraw-Hill, New York (1941). Electromagnetic Theory. (Pp. 563-573.)
248 SMALL-ANGLE SCATTERING OF X-RAYS

353 Stratton, J. A., and Houghton, H. G. A

Phys. Rev., 38, 159-165 (1931). "A Theoretical Investigation of the Trans-
mission of Light through Fog."
354 Toraldo Di Francia, G. A
Atti accad. naz. Lincei, (Rend., Classe sci. fis. mat. e nat.), 4, 730--735 (1948).
"Sulla luce diffusee da un vasto numero di aperture a saso su di uno schermo
opaco."
355 Van de Hulst, H. C. B
Duwaer and Zonen, Amsterdam (1946). Recherches Astronomiques de
l'Observatoire d' Utrecht XI, Part I. "Optics of Spherical Particles." Pains-
taking analysis of scattering by spheres. Cases for different relative values of
parameters discussed separately to derive range of validity of simplifying
approximations.
3"56 van der Wyk, A. J. A. A
Arch. sci., (Geneva), 3 (No. 6), 419-444 (1950). "Diffusion de la lumiere par Jes
colloldes.''

VI. SOLUTIONS
357 Bergman, P. G. Pa.
Am. Phys. Soc., No. C-6, Washington (April 1948). Phys. Rev., 74, 1209
(1948). "Low-Angle Scattering of Colloidal Solutions." Distribution effect
in concentrated monodispersed hydrosols.
358 Brady, G. W. A
J. Chem. Phys., 19, 1547-1550 (1951). "Fourier Analysis of the X-Ray
Scattering from Soap Solutions." Spherical micelle model. Nearest neighbor
distance.
359 Corrin, :\'.[. L. A
J. Chem. Phys., 16, 844-845 (1948). "Interpretation of X-Ray Scattering
from Solutions of Long-Chain Electrolytes on the Basis of a Spherical Micelle."
Small-angle patterns from long-chain molecules as readily explained by
distribution of spheres as by periodic layers.
360 Davies, J. T., and Philippoff, W. A
Nature, 164, 1087 (1949). "A Test of the Interpretation of X-Ray Patterns
of Micelles."
361 Harkins, W. D. A
J. Chem.. Phys., 16, 156-157 (1948). "A Cylindrical Model for the Small Soap
Micelle."
361' Harkins, W. D. B
Reinhold, New York (1952). The Physical Chem.:stry of Surface Films.
Soap micelles.
362 Hughes, E. W. A
Nature, 165, 1017 (1950). "A Test of the Interpretation of X-Ray Patterns
of Micelles." Soap solutions. Criticism of No. 360. Different models fit
data-none unique.
363 Hughes, E. W. P
A.CA, No. A-5, New Hampton, N.H. (August 1950). "The X-Ray Diffraction
Patterns from Soap Solutions." It is concluded that it is not possible with
best X-ray data to positively establish any particular micelle model.
364 Mattoon, R. W., and Mathews, M. B. A
J. Chem. Phys., 17, 496-497 (1949). "Micelles in Non-Aqueous Media."
Small-angle scattering from aerosols.
BIBLIOGRAPHY 249

365 Mattoon, R. W., Stearns, R. S., and Harkins, W. D. A

J. Chem. Phys., 15, 209-211 (1947). "Structure of Soap Micelles as Indicated
by a Previously Unrecognized X-Ray Diffraction Band."
366 Mattoon, R. W., Stearns, R. S., and Harkins, W. D. A
J. Chem. Phys., 16, 644-658 (1948). "A New Long-Spacing X-Re.y Diffraction
Band, and the Relations of Other Bands." Band position independent of
concentration; attributed to micelle thickness.
367 McBain, J. W., and Marsden, S. S., Jr. A
J. Chem. Phys., 15, 211-212 (1947). "Non-Ionic Detergents as Association
Colloids Giving Long X-Re.y Spacings in Aqueous Solution."
368 Philippoff, W. A
J. Colloid Sci., 4, 169-191 (1950). "Micelles and X-Re.ys." 53 references.
369 Philippoff, W. PA
Discussions Faraday Soc., 11, 96-107 (1951). "The Micelle and Swollen
Micelle on Soap Micelles."
369' Porod, G. PA
Intern. Union Chem., Strasbourg (June 1952). J. Polymer Sci., 10, 157-166
(1953). "X-Re.y and Light Scattering by Che.in Molecules in Solution."
Debye formula to deduce low-angle scatter.
370 Riley, D. P. A
Brit. Sci. News, 3, 7-10 (1949). "X-Ray Investigations of the Structure of
Liquid Disperse Systems." Protein solutions and oil-water emulsions.
371 Riley, D. P., and Oster, G. PA
Discussions Faraday Soc., 11, 107-116 (1951). "Some Theoretical and Experi-
mental Studies of X-Re.y and Light Scattering by Colloidal and Macromolecular
Systems."
372 Schulman, J. H., Me.talon, R., and Cohen, M. PA
Discussions Faraday Soc., 11, 117-121 (1951). "X-Re.y and Optical Prop-
erties of Spherical and Cylindrical Aggregates in Long Chain Hydrocarbon
Polyethylene Oxide Systems."
372' Schulman, J. H., McRoberts, T. S., and Riley, D. P. A
J. Physiol. (London), 107, 49P-50P (1948). "X-Ray Investigation of Trans-
parent Oil-Water Systems."
373 Schulman, J. H., and Riley, D. P. A
J. Colloid Sci., 3, 383-405 (1948). "X-Ray Investigation of the Structure of
Transparent Oil-Water Disperse Systems. I." Structure suggested water
spheres in oil and oil spheres in water.

Vll. FIBERS AND PROTEINS

374 Adams, J. A
Nature, 167, 78 (1951). "Low-Angle X-Re.y Scattering from Ramie and Jute."
375 Anderegg, J. W. T
Univ. Wisc., Ph.D. (1952). "Small-Angle Scattering from Serum Albumin
Solutions." Slit corrections. Concentration and pH effects.
376 Anderegg, J. W., Beeman, W. W., and Shulman, S. Pa
Am. Phys. Soc., No. C-7, Washington (May 1952). Phys. Rev., 87, 186 (1952).
"Small-Angle X-Ray Scattering from Serum Albumin Solutions."
377 Anderegg, J. W., Beeman, W. W., Shulman, S., and Ke.esberg, P. P
ACA, Tamiment, Pa. (June 1952). "Small-Angle X-Re.y Scattering from
Serum Albumin Solutions."
250 SMALL-ANGLE SCATTERING OF X-RAYS

378 Arnett, L. M., Meibohm, E. P.H., and Smith, A. F. A

J. Polymer Sci., ii, 737-738 (I 950).
"New Observations on the Small-Angle
X-Ray Diffraction of Synthetic Fibers."
379 Astbury, W. T. Aa
J. Intern. Soc. Leather Trades' Chemists, 24, 69-92 (1940). Nature, 14/i,
421-422 (1940). "The Molecular Structure of the Fibers of the Collagen
Group."
380 Astbury, W. T. A
Proc. Roy. Soc. (London), B-134, 303-328 (1947). "Structure of Biological
Fibers and the Problem of Muscle." Resume of X-ray studies on structure of
proteins of the keratin and collagen groups. 66 references.
381 Astbury, W. T., and Weibull, C. A
Nature, 163, 280-282 (1949). "X-Ray Diffraction Study of the Structure of
Bacterial Flagella." Flagella behave like long protein macromolecules,
studied in dried films. Give patterns indicating relation to alpha-keratin in
hair or muscle.
382 Bateman, J.B., Hsu, S. S., Knudsen, J. P., and Yudowitch, K. L. A
Arch. Biochem. and Biophys., 4/i, 411-422 (1953). "Hemoglobin Spacing in
Erythrocytes." Intermolecular interference ring varied by swelling and
shrinking blood cells. See Nos. 415, 486.
383 Bateman, J.B., Yudowitch, K. L., Hsu, S. 8., and Knudsen, J.P. P
Soc. Exptl. Biol. Med. (Baltimore Section), Edgewood, Md. (October 1952).
"Hemoglobin Spacing in Erythrocytes."
384 Baule, B., Kratky, 0., and Treer, R. A
Z. physik. Chem., B-50, 255-297 (1941). "Die iibermolekulare Aufbau der
Hydratcellulose die Einfiihrung der Blattchenmicelle in die Theorie der
Deformationsvorgange.''
385 Bawden, F. C., Pirie, N. \V., Bernal, J. D., and Fankuchen, I. A
Nature, 138, 1051-1052 (1936). "Liquid Crystalline Substances from Virus-
Infected Plants." Intramolecular spacings of infected sap from tomato and
tobacco plants. Preparation of crystalline protein. Camera 40 cm. long.
3' angular width beam gives spacings up to 1,200 A.
386 Bear, R. S. A
J. Am. Chem. Soc., 64, 727 (1942). "Long X-Ray Diffraction Spacings of
Collagen." Fine slits, long collimating tubes. Ni filtered Cu; specimen-to-film
distance: 26 cm. Twenty and more orders of spacing about 640 A observed
in dried beef tendon.
387 Bear, R. S. A
J. Am. Chem. Soc., 65, 1784-1785 (1943). "Long X-Ray Diffraction Spacings
of the Keratins." Chart of meridional and near meridional reflections. Low
values are listed, since meridional and near meridional arcs had to be considered
together, since large-spacing layer linv components could not be resolved
satisfactorily.
388 Bear, R. S. A
J. Am. Chem. Soc., 66, 1297-1305 (1944). "X-Ray Diffraction Studies on
Protein Fibers. I. The Large Fiber-Axis Period of Collagen." Wet and
dried samples differ. 640-680 A period. 30 orders from dried samples.
389 Bear, R. S. A
J. Am. Chem. Soc., 66, 2043-2050 (1944). "X-Ray Diffraction Studies on
Protein Fibers. II. Feather Rachis, Porcupine Quill Tip, and Clam Muscle."
Classification of protein fibers, relations between wide- and low-angle
diffractions.
BIBLIOGRAPHY 251

390 Bear, R. S. Pa
ASXRED and AAAS, No. A-1, Gibson Island, Md. (August 1944). Phys.
Rev., 67, 196 (1945). "Small-Angle Diffraction Studies on Protein Fibers."
Fiber axis periods from 20-40 layer lines.
391 Bear, R. S. A
J. Am. Chem. Soc., 67, 1625 (1945). "Small-Angle X-Ray Diffraction Studies
on Muscle." Calculated spacings change with cosine of angle, indicating
diffraction elements as two-dimensional.
392 Bear, R. s. p
Intern. Union Cryst., Harvard (1948). "The Large Unit Cells of Protein
Fibers."
393 Bear, R. S. p
ACA, Tamiment, Pa. (June 1952). "Tho Structure of Collagen."
394 Bear, R. S., and Bolduan, 0. E. A. P
CSA and ASX RED, New Haven (April 1948). "The Nature of the Order of
Large Size Exhibited by Collagen Fibrils." Dry tendon model of cylindrical
fibrils.
395 Bear, R. S., and Bolduan, 0. E. A. A
Acta Cryst., 3, 230-235 (1950). "Deficiencies in Order of Large Size in Fibrous
Systems.''
396 Bear, R. S., and Bolduan, 0. E. A. A
Acta Cryst., 3, 236-241 (1950). "Diffraction by Cylindrical Bodies with
Periodic Axial Structure."
397 Bear, R. S., and Bolduan, 0. E. A. A
J. Appl. Phys., 22, 191-198 (1951). "Periodic Statistical Distortion of
Unidirectionally Ordored Diffractors, with Application to Collagen."
398 Beeman, W. W. P
Pitt8burgh Conj. X-Ray Electron Diffr., No. 6, Pittsburgh (November 1950).
"Some Problems of Interpretation in the Small-Angle Scattering from Proteins.''
Not delivered.
399 Beeman, W. W. P
Am. Chem. Soc., No. J-17, Boston (April 1951). "The Small-Angle Scattering
of X-Rays by Protein Solutions."
400 Bergman, M. E., and Fankuchen, I. A
Science, 113, 415 (1951). "X-Ray Diffraction Studies of Inclusion Bodies
Found in Plants Infected with Tobacco Mosaic Virus."
401 Bernal, J. D., and Fankuchen, I. A
Nature, 139, 923-924 (1937). "Structure Types of Protein Crystals from Virus
Infected Plants." Oriented protein gels. None of the reflections in directions
other than at right angles to the molecular length is due to the packing of the
molecules, but rather to regularities within them, as shown by the fact that
such rE-flections persist unchanged when the molecules are in dilute solution.
402 Bernal, J. D., and Fankuchen, I. A
J. Gen. Physiol., 25, 111-165 (1941). "X-Ray and Crystallographic Studies
of Plant Virus Preparations. I. Introduction and Preparation of Specimen.
II. Modes of Aggregation of the Virus Particles. III. Structure of the
Particles. Biological Implications." See No. 385.
403 Bernal, J. D., Fankuchen, I., and Riley, D. P. A
Nature, 142, 1075 (1938). "Struct1J.re of Crystals of Tomato Bushy Stunt
Virus Preparations." Two lines were observed of spacings 279 A and 160 A,
respectively, corresponding to 340 A diameter spheres.
252 SMALL-ANGLE SCATTERING OF X-RAYS

404 Bolduan, 0. E. A. p
Electron Mier. Soc. Amer., Pittsburgh (1946). "A Long Spacing X-Ray
Diffraction Camera." Camera design for resolution of multi-order reflections
at small angles.
405 Bolduan, 0. E. A., and Bear, R. S. A
J. Polymer Sci., 5, 159-168 (1950). "Unidirectional Nature of the Large
Periodic Structure of Collagen Fibrils."
406 Bolduan, 0. E. A., apd Bear, R. S. A
J. Polymer Sci., 6, 271-284 (1951). "General Nature of the Extension of
Structure Transverse to Collagen Fibril Axes."
407 Boyes-Watson, J., Davidson, E., and Perutz, M. F. A
Proc. Roy. Soc. (London), A-191, 83-129 (1947). "An X-Ray Study of Horse
Methemoglobin." Cell contains variable liquid of crystallization. From
data, horse hemoglobin molecules are concluded to be cylinders 57 A in
diameter and 34 A high.
408 Cannan, C. M. M., and Bear, R. S. P
Am. Chem. Soc., No. C-72, Boston (April 1951). "Small-Angle X-Ray Diffraction
Analysis of the Ultrastructure of the Myofilement."
409 Clark, G. L., and Corrigan, K. E. A
Phys. Rev., 40, 639 (1932). "The Crystal Structure of Insulin." Large
spacings (80-130 A) reported for insulin, using Mg and Al radiation.
410 Clark, G. L., and Parker, E. A. A
Science, 85, 203-204 (1937). "Diffraction of X-Rays at Very Small Angles by
Celluloses and Rayons."
411 Clark, G. L., Parker, E. A., Schaad, J. A., and Warren, W. J, A
J. Am. Chem. Soc., 57, 1509 (1935). "New Measurements of Previously
Unknown Large Interplanar Spacings in Natural Materials." Equatorial
maxima in cellulose: Nerve, collagen, keratin, rubber, cellulose.
412 Corey, R. B., and Wyckoff, R. W. G. A
J. Biol. Chem., 114, 407-414 (1936). "Long Spacings in Macromolecular
Solids." X-ray diffraction photographs registering large spacings. General
structures discussed.
413 Cowan, P., and Hodgkin, D. A
Acta Gryst., 4, 160-161 (1951). "A Comparison of X-Ray Measurements on
Air-Dried Tobacco Necrosis Protein Crystals with Electron Microscope Data."
414 Crowfoot, D., and Fankuchen, I. A
Nature, 141, 522-523 (1938). "Molecular Weight of Tobacco Seed Globulin."
Three lines recorded on X-ray photographs which correspond to the first three
reflections from a cubic face centered lattice of dimension a = 123 A.
415 Dervichian, D. G., Fournet, G., and Guinier, A. A
Gompt. rend., 224, 1848-1850 (1947). "Mise en evidence d'uno structure
submicroscopique dans Jes globules rouges par la diffusion des rayons X aux
petits angles." Shoulder in pattern attributed to spacing between hemo·
globins. Cell order between crystalline and solution.
416 Dervichian, D. G., Fournet, G., and Guinier, A. PB
Barcroft Memorial Haemoglobin Conf., Cambridge, England (June 1948).
Interscience, New York (1949). "An X-Ray Investigation of Haemoglobin
and Haemocyanin in Aqueous Solution," Haemoglobin, pp. 131-134. Hemo-
globin radius of gyration and hemocyanine spacing determined.
BIBLIOGRAPHY 253

417 Dervichian, D. G., Fournet, G., and Guinier, A. P

Sixth Intern. Congr. Exptl. Cytology, Stockholm. "Evidence for the Sub-
microscopic Structure of Red Cells by the Scattering of X-Rays at Small
Angles." Shoulder in pattern attributed to spacing between hf'moglobins.
Cell order between crystalline and solution.
418 Dervichian, D. G., Fournet, G., and Guinier, A. p
Pittsburgh Conj. X-Ray Electron Difjr., No. 3, Pittsburgh (November 1950).
"Small-Angle Scattering Studies on Egg Yolk." Persistent ring at Bragg
spacing of 260 A.
419 Dervichian, D. G., Fournet, G., and Guinier, A. A
Biochem. et Biophys. Acta, 8, 145-149 (1952). "Etude par la diffusion des
rayons X des modifications subies par certaines proteines."
420 Dervichian, D. G., Fournet, G., Guinier, A., and Ponder, E. A
Compt. rend., 235, 324-326 (1952). "Organisation submicroscopique des
hematies contenantes des hemoglobines anormales."
421 Ehrenberg, W., and Franks, A. A
Nature, 170, 1076-1077 (1952). "Small-Angle X-Ray Scattering." Total
reflection focusing for resolution of large periods.
421' Engstrom, A., and Finean, J.B. A
Nature, 171, 564 (1953). "Low-Angle X-Ray Diffraction of Bone."
422 Fankuchen, I. A
Phys. Rev., 53, 909 (1938). "On the Structure of 'Built-Up' Films on Metals."
423 Fankuchen, I. A
J. Gen. Physiol., 24, 315-316 (194i). "An X-Ray and Crystallographic Study
of Ribonuclease."
424 Fankuchen, I. A
Ann. N.Y. Acarl. Sci., 41, 157-168 (1941). "Evidence from X-Rays regarding
the Structure of Protein Molecules." Molecular weights determined by X-ray
methods in good agreement with those obtained by other methods.
425 Fankuchen, I. PA
Cold Spring Harbor Symposia Quant. Biol., 9, 198-203 (1941). "X-Ray
Diffraction Studies of Protein Preparations." Continuous and discontinuous
scattering.
426 Fankuchen, I. A
J. Am. Chem. Soc., 64, 2504-2505 (1942). "X-Ray and Optic Measurements
of fl-Lactoglobin." Spacings of 60, 63, llO A.
427 Fankuchen, I. A
J. Riol. Chem., 150, 57-59 (1943). "X-Ray Diffraction Data on Ferritin and
Apoferritin." Large spacings.
428 Fankuehen, I. A
Ann. Rev. Biochem., 14, 207-224 (1945). "X-Ray Studies on Compounds of
Biochemical Interest." 85 references.
429 Fankuchen, I. A
Nature, 168, 684-685 (1951). "The Lotmar-Picken X-Ray Diffraction
Diagram of Muscle."
430 Fankuchen, I. B
Academic Press, New York (1945). Advances in Protein Chemistry, edited by
M. L. Anson and J. T. Edsall, Vol. 2, pp. 387-405, "X-Ray Diffraction and
Protein Structure."
254 SMALL-ANGLE SCATTERING OF X-RAYS

431 Fankuchen, I., and Mark, H. A

Re-cord Chem. Progr. (Kresge-Hooker Sci. Lib.), 4, 54-57 (1943). "Improved
X-Ray Technique for the Study of Natural and Synthetic Fibers."
432 Fankuchen, I., and Mark, H. A
J. Appl. Phys., 15, 364-370 (1944). "X-Ray Studies of Chain Polymers."
Small-angle scattering for larger polymer periods. Apparatus and technique
described.
433 Fankuchen, I., and Mark, H. B
Interscience, New York (1946). Currents in Biochemical Research, edited by
D. E. Green, pp. 439-452, "X-Ray Diffraction and the Study of Fibrous
Proteins."
434 Fankuchen, I., and Schneider, M. A
J. Am. Chem. Soc., 66, 500-501 (1944). "Low-Angle X-Ray Scattering from
Chrysotiles." Dry gels of tobacco mosaic virus, 30 cm. specimen-to-film
distance, spacings up to 500 A.
434' Fournet, G., and Antzenberger, P. A
Compt. rend., 236, 394-396 (1953). Etude et interpretation de la diffusion des
rayons X aux petits angles par la cellulose." Disordered regions swell when
wet.
435 Frey-Wyssling, A. B
Elsevier, New York (1948). Submicroscopic Morphology of Protoplasm and
Its Derivatives. (Pp. 67-81.)
436 Friedrich-Freska, H., Kratky, 0., and Sekora, A. A
Naturwissenschaften, 33, 78 (1944). "Auftreten von neuen Rontgeninter-
ferenzen bei Einlagerung von Jod in Seidenfibrin von Bombyn-mori Typ."
Meridian reflections indicating a 70 A period.
437 Furnas, T. C., Jr. T
Mass. Inst. Tech., Ph.D. (1952). "Point Focus X-Re.y Monochromator for
Diffraction Studies of Protein Fibers." See No. 229.
438 Furnas, T. C., Jr. P
ACA, Tamiment, Pa. (June 1952). "Diffraction Diagrams of Protein Fibers
Taken with a New Point Focus Monochromator."
439 Heikens, D., Hermans, P.H., and Weidinger, A. A
Nature, 170, 369-370 (1952). "Occurrence of Small-Angle X-Ray Diffraction
Maxima in Some Artificial Cellulose Fibres."
440 Hermans, P.H. B
Elsevier, New York (1946). Contribution to the Physics of Cellulose Fibres.
(Pp. 158-187.)
441 Hermans, P. H. B
Elsevier, New York (1949). Physics and Chemistry of Cellulose FibTes. (Pp.
244-265.)
441' Hess, K., and Kiessig, H. A
Z. physik. Chem., 193, 196-217 (1944). "Uber Langperiodeninterferenzen und
micellaren Faserfeinbau bei vollsynthetischen Fasern (Polyamide und
Polyester)."
441" Hess, K., and Kiessig, H. A
Kolloid-Z., 130, 10-19 (1953). "Ober Langperiodeninterferenzen bei syn-
thetischen Fasern."
442 Heyn, A. N. J. A
J. Am. Chem. Soc., 70, 3138-3139 (1948). "Small-Angle X-Ray Scattering of
Various Cellulose Fibers." Measure of micelle.r orientation from small-angle
patterns.
BIBLIOGRAPHY 255

443 Heyn, A. N. J. PA
Fiber Soc., Boston (1948). Textile' Research J., 19, 163-172 (1949). "Small-
Angle X-Ray Scattering in Various Cellulose Fibers and Its Relation to the
Micellar Structure." Scattering related to micellar size, distance, and
orientation.
444 Heyn, A. N. J. P
Conj. Small-Angle X-Ray Scattering, No. 5, Columbia, Mo. (October 1949).
"Small-Angle Scattering by Cellulose Fibers." Equatorial scattering modified
by swelling fibers and by torsion.
445 Heyn, A. N. J. A
J. Am. Chem. Soc., 71, 1873-1875 (1949). "Small-Angle Scattering of X-Rays
and the Micellar Structure in Cellulose Fibers." Distribution of the micelles,
rather than their size, determines the scattering.
446 Heyn, A. N. J. A
J. Am. Chem. Soc., 72, 2284-2285 (1950). "Small-Angle X-Ray Scattering by
Cellulose Fibers: Experimental Study of the Orientation Factor in Model
Filaments and Rayons." Effects of controlled orientation changes.
447 Heyn, A. N. J. A
J. Arn. Chem. Soc., 72, 5768-5769 (1950). "A Quantitative Evaluation of
Small-Angle X-Ray Scattering by Various Cellulose Fibers for the Deter-
mination of Crystallite Size with Special Reference to the Problem of Inter-
particle Interference." Effect of dense-packing of micelles overcome by
swelling fibers.
448 Heyn, A. N. J. A
Rayon and Synthetic Textiles, 31, No. 9: 29-43, No. 10: 42-43 (1950). "How
to Identify Synthetic Fibers by Their X-Ray Diagram." Systematic examin-
ation of small-angle patterns from a number of fibers.
449 Heyn, A. N. J. B
Textile Book, Now York (1951). "Identification of Synthetic Fibers by
X-Ray Diffraction," Chap. 19 of American Handbook of Synthetic Fibers.
450 Heyn, A. N. J. B
Wiley, New York (1952). "X-Ray Diagrams of ~atural and Synthetic
Fibers," Chap. XXI of Textile Fibers, edited by l\Iatthews. Sixth edition.
450' Heyn, A. N. J. P
ACA, Cambridge, Mass. (April 1954). "Small-Angle X-Ray Scattering by
Fibers."
451 Hosemann, R. A
Z. physik. Chem., A-179, 356-360 (1937). "Quantitative Bestimmung der
Orientierung der Kristallite in Faserstiiffen."
452 Hosemann, R. A
Z. Elektrochem., 46, 535-555 (1940). "Riintgenographische Verteilungszu-
standes in einem Faserstoff. Diskussion des Cellulosemodells." Micelle size
distribution in cellulose and paraffin.
453 Hosemann, R. A
Acta Cry.•t., 4, 520-530 (1951). "Die parakristalline Feinstruktur natiirlicher
und synthetischer Eiweisse. Visuelles Naherungsverfahren zur Bestimmung
der Schwankungstensoren von Gitterzellen."
454 Hosemann, R. PA
Kolloid Ges. (September Hl51). Kolloid-Z., 125, 149-156 (1952). "Bestim-
mung der statistischen Strukturparameter der Mizellgitter hochmolekularer
Faserstoffe.
256 SMALL-ANGLE SCATTERING OF X-RAYS

455 Hsu, S. S. T
Fla. State Univ., M.S. (1950). "An X-Ray Investigation of Red Blood
Corpuscles." Hemoglobin separation in swollen and shrunken red cells. See
No. 382.
456 Hsu, S. S., and Yudowitch, K. L. P
Pittsburgh Conj. X-Ray Electron Diffr., No. 8, Pittsburgh (November 1950).
"Hemoglobin Spacing in Erythrocytes." Small-angle intensity shoulder due
to hemoglobin spacing varied by swelling erythrocytes. See No. 382.
456' Janeschitz-Kriegl, H., and Kratky, 0. A
Z. Elektroche:m., 57, 42-50 (1953). "Die Abhengigkeit der Rontgenklein-
winkelstreuung von Quellungsmittel bei regenerierter Cellulose."
457 Janeschitz-Kriegl, H., Kratky, 0., and Porod, G. A
Z. Elektrochem., 56, 146-154 (1952). "Rontgenkleinwinkelmessungen an
porosen Hydratzellulosefaden."
458 Kaesberg, P. P
ACA, Tamiment, Pa. (June 1952). "The Size, Shape, and Internal Hydration
of Several Viruses."
459 Kaesberg, P., Anderegg, J. W., Leonard, B. R., and Beeman, W. W. Pa
Am. Phys. Soc., No. C-6, Washington (May 1952). Phys. Rev., 87, 186 (1952).
"On the Size, Shape, and Hydration of Tomato Bushy Stunt Virus."
460 Ke.esberg, P., Ritland, H. N., and Beeman, W. W. A
Phys. Rev., 74, 71-73 (1948). "The Use of the Double Crystal Spectrometer
in the Analysis of Bragg Reflections at Very Small Angles." Two-crystal
spectrometer used for nine orders of scattering from large periodicity in wet
collagen.
461 Ke.esberg, P., Ritland, H. N., and Beeman, W.W. Pa
Am. Phys. Soc., No. A-5, Madison, Wisc. (June 1948). Phys. Rev., 74, 1255-
1256 (1948). "Small-Angle Reflections from Collagen." Double-crystal
spectrometer for eight orders of Bragg reflections from collagen.
462 Knudsen, J. P. T
Fla. State Univ., M.S. (1951). "X-Ray Investigations of Hemoglobin in
Human Erythrocytes and Particulate Inclusions in Bacteria and Fungi
Spores." Continuation of Nos. 455, 499. See No. 382.

Note: Item 463 1°s omitted.

464 Kratky, 0. A
Kolloid-Z., 64, 213-222 (1933). "Zurn Deformationsmechanismus der
Feserstoffe. I."
465 Kratky, 0. A
Kolloid-Z., 68, 347-350 (1934). "Zurn Deformationsmechanismus der
Faserstoffe. II. Die Ordnung der Mizellen von Filmen in kleinsten Berei·
chen."
466 Kratky, 0. A
Naturwissenschaften, 26, 94 (1938). "Die Berechnung der Mizelldimensionen
von Faserstoffen aus den unter kleinsten Winkeln abgebeugten Interferenzen."
Micelle size of ramie. Closely packed colloids.
467 Kratky, 0. A
Monatsh. Chem., 77, 224-250 (1947). "Neuere Ergebnisse aus der Rontgeno-
graphie der Eiweissstoffe." Review on kangaroo tail with regard to amino
acid connections. Egg whites.
BIBLIOGRAPHY 257
468 Kratky, 0. A
J. Polymer Sci., 3, 195-215 (1948). "Low-Angle Scattering in Polymers."
Review including insulin solutions.
469 Kratky, 0., and Porod, G. A
Rec. trav. chim., 68, 1106--1122 (1949). "Riintgenuntersuchung geloster
Fadenmolektile." Internal spacings in coiled chain molecules.
470 Kratky, 0., Schauenstein, E., and Sekora, A. A
Nature, 165, 527-528 (1950). "A New Type of Lattice with Large Periods in
Silk." New equatorial reflections indicating a network distance of 45 A.
471 Kratky, 0., Schauenstein, E., and Sekora, A. A
Nature, 170, 796--797 (1952). "Relations between Small-Angle Interferences
of Silk and the Wax Content."
472 Kratky, 0., and Schossberger, F. A
Z. physik. Chem., B-39, 145--154 (1938). "Ein Beitrag zur micellar Struktur
der Cellulose." Scattered intensity is increased by introduction of finely
divided gold into ramie. Diffuse scattering is ascribed to an enlargement of
small-angle Bragg reflections due to some approximate orderliness of a large-
scale arrangement.
473 Kratky, 0., and Sekora, A. A
J. Makromol. Chem., 1, 113-121 (1943). "Die Auffindung von grossen Netze-
benenabstiinden bei Kiinguruh-Schwanzsehne. Ein Beitrag zur molekularen
Struktur der Faserproteine." Long period in kangaroo tail related to inter-
ferences with separate amino acids with assumed period structure.
474 Kratky, 0., Sekora, A., and Friedrich-Freska, H. a
Anz. Akad. Wiss. Wien, Math. Naturw. Kl., 83, 30--35 (1946). "Dber Grosse
und Form von gelosten Hiimocyaninmolektilen aus Messungen der Riintgen-
kleinwinkelstreuung." Aggregates of 260 A spheres. Compared with electron
microscopy and centrifugation_ Maxima noted for neighboring spheres.
475 Kratky, 0., Sekora, A., and Weber, H. H. A
Naturwissenschaften, 31, 91 (1943). "Neue Kleinwinkelinterferenzen bei
Myosin." Myosin fibers showed 3 new weak reflections of 33, 42, and 66 A,
representing a lateral regularity, the 66 A dimension being the fiber thickness.
476 Kratky, 0., and Wurster, A. A
Z. Elektrochem., 50, 249-255 (1944). "Die Abhiingigkeit der Kleinwinkel-
streuung Lier Riintgenstrahlung von Quellungsmittel die Hydratcellulose."
Scattered intensity found dependent on third power of difference between
electron densities of cellulose and swelling centers.
477 Leonard, B. R., Jr. T
Univ. Wisc., Ph.D. (1951). "Small-Angle Scattering from Certain Spherical
Viruses (Southern Bean Mosaic, Tobacco Necrosis, and Tomato Bushy Stunt)."
478 Leonard, B. R., Jr., Anderegg, J. W., Kaesberg, P.,. Shulman, S., and
Beeman, W.W. A
J. Chem. Phys., 19, 793-794 (1951). "On the Size, Shape, and Hydration of
Southern Bean Mosaic Virus and Tobacco Necrosis Virus in Solution."
Secondary maxima show viruses are spherical and internally hydrated.
478' MacArthur, I. A
Nature, 152, 38-41 (1943). "Structure of Alpha-Keratin."
4 79 Mark, H. B
Springer, Berlin (1932). Technologie der Textiljasern. "I. Physik und
Chemie der Cellulose."
258 SMALL-ANGLE SCATTERING OF X-RAYS

480 Meibohm, E. P.H., and Smith, A. F. PA

Pittsburgh Conj. X-Ray Electron Dijjr., No. 7, Pittsburgh (November 1950).
J. Polymer Sci., 7, 449-456 (1951). "Observations on Small-Angle Inter-
ference Maxima in Synthetic Organic Polymers." Non-meridional maxima
observed in several orders.
481 Miller, E., Fankuchen, I., and Mark, H. Pa
Am. Phys. Soc., No. PC-2, New York (January 1949). Phys. Rev., 75, 1314
(1949). "Polymerization in the Solid State."
482 Miller, E., Fankuchen, I., and Mark, H. A
J. ,Appl. Phys., 20, 531-533 (1949). "Polymerization in the Solid State."
Spacings of 5 to 90 A.
483 Perron, R. A., and Wright, B. A. A
Nature, 166, 863-864 (1950). "Alteration of Collagen Structure by Irradiation
with Electrons."
484 Perutz, M. F. A
Nature, 161, 204-205 (1948). "Submicroscopic Structure of the Red Cell."
Mean distance between neighboring hemoglobin molecules in red cell.
485 Posner, A. P
ACA, No. 67, Ann Arbor, Mich. (June 1953). "Particle Size and Particle Size
Distribution in Apatites."
486 Riley, D. P., and Herbert, D. A
Biochim. et Biophys. Acta, 4, 374-384 (19i'i0). "Molecular Size, Shape, and
Aggregation in Concentrated Protein Solutions as Revealed by X-Ray
Scattering-Hemoglobin and Egg-Albumin." Diffuse rings related to sizes
and shapes of packed molecules.
487 Riley, D. P., and Oster, G. A
Trans. Faraday Soc., 46, 791 (1950). "X-Ray Diffraction Studies of Solutions
of Desoxyribonucleic Acid (Na Salt)."
488 Riley, D. P., and Oster, G. A
J. chim. phys., 47, 715 (1950). "Etude aux rayons X de solutions d'acide
thymonucleique a diverses concentrations."

Note: Item 489 is omitted.

490 Ritland, H. N. 'I
Univ. Wisc., Ph.D. (1949). "Small-Angle X-Ray Scattering from Protein
Solutions."
491 Ritland, H. N., Kaesberg, P., and Beeman, \V. \V. P
Conj. Small-Angle X-Ray Scattering, Ko. 10, Columbia, Mo. (October 1949).
"Information on the Shapes and Hydrations of Protein Molecules in Solution
from Small-Angle X-Ray Scattering.'' Gyration radii of six proteins measured.
Axial ratio deduced and used to find extent of hydration.
492 Ritland, H. N., Kaesberg, P., and Beeman, W. W. A
J. Chem. Phys., 18, 1237-1242 (1950). Tech. Report, O.::>r.R. Contract N7
ONR-285, T-0 III, NR 017-605 (1949). "An X-Ray Investigation of the
Shapes and Hydrations of Several Protein Molecules in Solution." Gyration
radii of five proteins from small-angle scattering used with molecular weights,
densities, and fractional ratios to get axial ratios and hydrations.
493 Schmitt, F. 0. A
J. Am. Leather Chemists' Assoc., 46, 538-547 (1951). "Structural and
Chemical Studies on Collagen."
BIBLIOGRAPHY 259

494 Sen, M. K., and Mukherjee, R. R. A

J. Textile Inst. Proc. and Abstr., 43, 114-121 (1952). "The Structure of Jute.
I. The X-Ray Diffraction Pattern."
495 Shulman, S., Anderegg, J., Leonard, B. R., Jr., and Kaesberg, P. P
Am. Chem. Soc., No. C-74, Boston (April 1951). "An X-Ray Examination of
Tobacco Necrosis Virus in Solution."
496 Shurman, M. M., and Kaesberg, P. P
Pitt<Jburgh Conj. X-Ray Electron Diffr., No. 9, Pittsburgh (November 1950).
"X-Ray Evidence concerning the Structure of Collagen." Fourier synthesis
of pattern from beef tendon in agreement with electron micrographs.
497 Stone, L. L. T
Fla. State Univ., M.S. (1950). "An X-Ray Examination of Selected Bac-
teria." Small-angle evidence for bacterial nuclei.
498 Vineyard, G. H. A
Acta Cryst., 4, 281 (1951). "Patterson Transforms and the Interpretation of
X-Ray Scattering from Fibers."
499 Weibull, C. A
Nature, 165, 482-483 (1950). "X-Ray Diffraction Pattern Given by Bacterial
Flagella." Periodicities detected in flagella.
500 Wright, B. A. A
Nature, 162, 23 (1948). "Low-Angle X-Ray Diffraction Pattern of Collagen."
27 orders. Found 630-672 A for 20-100 per cent humidity.
501 Wright, B. A. P
CSA and ASX RED, Kew Haven (1948). "Information concerning the Water
Uptake of Collagen as Evidenced by Its Low-Angle X-Ray Diffraction Pat-
tern.''
502 Wright, B. A., and Cole, P. A. A
Rev. Sci. Instr., 20, 355-356 (1949). "An Improved Low-Angle X-Ray
Diffraction Camera." Collagen pattern with high resolution and facility.
503 Wright, B. A., and Wiederhorn, N. M. P
ASXRED, Ithaca, N. Y. (June 1949). "An X-Ray Diffraction Study in
Shrinkage of Collagen."
504 Wright, B. A., and Wiederhorn, N. M. A
J. Polymer Sc·i., 7, 105-120 (1951). "Studies Concerned with the Structure of
Collagen. I. An X-Ray Investigation of the Denaturization of Collagen."
505 Wrinch, D. A
J. Chem. Phys., 20, 1332-1333 (1952). "Evidence for Globulite Molecules in
Horse Hemoglobin."
505' Wyckoff, R. W. G., and Corey, R. B. A
Science, 81, 365-366 (1935). "X-Ray Diffractions from Hemoglobin and
Other Crystalline Proteins."
505" Wyckoff, R. \V. G., Corey, R. B., and Biscoe, J, A
Science, 82, 175-176 (1935). "X-Ray Reflections of Long Spacings from
Tendon."
506 Zahn, H., and Winter, U. A
Kolloid-Z., 128, 142-153 (1952). "t'.rber die Langperiodenreflexes im Rontgeno-
gramm von Polyethanfaden."
AUTHOR INDEX TO BIBLIOGRAPHY

Adams, J., 374 Bowman, J.C., 11•

Adler, B. J., 284 Boyes-Watson, J., 407
Alexander, L., 101, 198, 199, 200, 200', Brady, Q_ W., 358
201, 202 Brentano, J_ C. M., 12
Alfrey, T., 99 Brill, R., 5
Allin, E. J., 126 Brillouin, L., 316
Anderegg, J. W., 115, 375, 376, 377, 459, Bromwich, T. J., 317
478, 495 Broussard, L., 206
Antzenberger, P., 434' Brusset, H., 13, 14, 15, 16, 17, 18
Arlman, J. J., 23 Bueche, A. M., 318, 319, 320, 322, 323
Arnett, L. M., 378 Burgy, M. T., 250
Astbury, W. T., 379, 380, 381 Burton, R. L., 22
Buttrey, J. W-, 18'
Bagchi, S. N., 86, 279
Bale, H., l Cannan, C. M. M., 408
Banerjee, K., l ', 2 · Ca.rlisle, C. H., 9
Ba.mes, M. D., 333, 334 Carroll, B., 19
Barton, H. M., Jr., 3, 4, 5 Carteret, Y., 89, 90, 91
Bateman, J.B., 382, 383 Cha.rlesby, A., 19'
Ba.uer, S. H., 6 Christman, A- C., Jr., 20
Baule, B., 7, 384 Clark, G. L_, 21, 22, 207, 409, 410, 411
Bawden, F. C., 385 Cohen, M., 372
Bear, R. S., 205, 386, 387, 388, 389, 390, Cole, P. A., 502
391, 392, 393, 394, 395, 396, 397, 405, Cooper, B. S., 348
406, 408 Corey, R. B., 412, 505', 505N
Beeman, W. W., 8, 8', 31, 32, 33, 115, Corner, J., 262
145', 221, 225, 226, 376, 377, 398, 399, Corrigan, K. E., 207, 409
459, 460, 461, 478, 491, 492 Corrin, M. L., 359
Bergman, M. E., 203, 400 Cowa.n, P., 413
Bergman, P. G., 357 Crowfoot, D. (Hodgkin), 414
Berna.I, J. D., 9, 385, 401, 402, 403
Bertaut, F., 204 De.ams, H., 23, 34'
Bethe, H. A., 310 Danielson, W. E., 24,25, 151,211,228,229
Biscoe, J., 10, 308, 505
6 Da.nilov, V. I., 26, 27
Blin, J., 11, 11', 169 Davidson, E., 407
Blumer, H., 311, 312, 313, 314 Davies, J. T., 360
Boggs, E. M., 283 Davis, B., 28, 29, 208
Bolduan, 0. E. A., 205, 394, 395, 396, Debye, P., 263, 264, 265, 266, 267,268,
397. 404, 405, 406 269, 270, 319, 320, 321, 322, 323
Bommel, H., 315 Dervichia.n,D. G., 415, 416, 417, 418, 419,
Born, M., 259, 260, 261 420
Note: Numbers refer to numbered items in the Bibliography, not to pa.gee.
261
262 SMALL-ANGLE SCATTERING OF X-RAYS

Despujols, J., 208' Guinier, A., 72', 73, 73', 74, 74', 169, 170,
Devaux, J., 17 216, 217, 218, 219, 274, 415, 416, 417,
DeVore, J. R., 324 418, 419, 420
Dexter, D. L., 30, 30', 30n, 31, 32, 33 Gupta, N. N., 74"
Donnay, J. D. H., 34
Dorgelo, H. B., 34' Hach, K. M., 126'
Dragsdorf, R. D., 35, 36, 37 Hagg, G., 219'
DuMond, J. W. M., 25, 76"", 151, 209, Halpern, 0., 245, 246, 247, 248
210, 211, 228, 229 Harkins, W. D., 361, 361', 365, 366
Hart, R. W., 328
Eckert, A. C., Jr., 22
Harvey, G. G., 275
Ehrenberg, W., 421
Hayes, S., 11•, 74"', 74"', 75
Eisenstein, A. S., 38, 39, 40
Heikens, D., 439
Elkin, P. B., 41, 154
Heller, R. B., 249, 250
Engelhard, V. H., 325
Hendricks, S. B., 76
Engstrom, A., 421'
Henke, B., 76', 76', 76'", 76"'', 211
Erbring, H., 122
Henry, N. F. M., 77
Erich, L. C., 146
Herbert, D., 486
Fankuchen, I., 19, 42, 93, 94, 95, 96, 98, Hermans, P. H., 439, 440, 441
99, 203, 212, 385, 400, 401, 402, 403, Hess, K., 441', 441'
414, 422, 423, 424, 425, 426, 427, 428, Heyn, A. N. J., 442, 443, 444, 445, 446,
429, 430, 431, 432, 433, 434, 481, 482 447, 448, 449, 450, 450'
Finean, J. B., 212', 421' Heywood, H., 78
Firth, F. G., 43 Hillier, J., 162
Fournier, F., 213 Hodgkin, D. (Crowfoot), 413
Forrester, A. T., 326 Hosemann, R., 79, 80, 80', 81, 82, 83, 84,
Fournet, G., 44, 45, 46, 47, 48, 49, 50, 51, 85, 86, 276, 277, 278, 279, 451, 452,
52, 53, 54, 74, 216, 217, 218, 219, 415, 453, 454
416, 417, 418, 419, 420, 434' Houghton, H. G., 353
Frarn;ois, J., 55 Hsu, S. S., 382, 383, 455, 456
Frank, R. M., 56 Hubbell, H. H., Jr., 87, 88, 159, 160, 161,
162, 220
Franklin, R. E., 56', 56', 214
Hughes, D. J., 250
Franks, A., 421
Hughes, E. W., 362, 363
Frey-Wyssling, A., 435
Humphreys, W. J., 329
Friedrich-Freska, H., 436, 474
Huxley, H. E., 220'
Friess, H., 325
Furnas, T. C., Jr., 437, 438
Imelik, B., 89, 90, 91
Gans, R., 327
Gerjouy, E., 247, 248 Jacquet, P., 74'
Gingrich, N. S., 39, 40, 57, 271, 272 Janeschitz-Kriegl, H., 456', 457
Glocker, R., 58 Jellinek, M. H., 42, 92, 93, 94, 95, 96,
Gokhale, B. G., 215 212, 231
Graham, W., 59 Jobst, V. G., 330
Gray, J. A., 60, 61, 130 Johnson, I., 331
Green, H. S., 259, 260, 261, 273 Johnson, M. F. L., 126"
Guinier, A., 11, 11', 17, 53, 54, 62, 63, 64,
65, 65', 66, 67, 67', 68, 69, 70, 71, 72, Kachkuruzov, G. A., 280

Note: Numbers refer to numbere~ items in the Bibliography, not to pages.

AUTHOR INDEX TO BIBLIOGRAPHY 263

Kaesberg, P., 8, 97, 115, 221, 225, 226, Meibohm, E. P. H., 378, 480
377, 458, 459, 460, 461, 478, 491, 492, Meneghetti, D., 251, 253, 254
495, 496 Menke, H., 269, 270
Kahovek, L., 97', llO, ll l, 222 Mie, G., 337
Kalinina, A. M., 131' Miller, E., 481, 482
Karlsson, N., 219' Milligan, W. 0., 121
Kaufman, H. S., 98, 99 Mitra, M. N., 338
Kenyon, A. S., 332 Mittenthal, L., 326
Kiessig, H., 222', 441', 441" Montroll, E. W., 328
Kikindai, T., 18, 100 Mukherjee, R. R., 494
Kirkwood, J. G., 281, 282, 283, 284, 285 Muller, F. H., 122
Klug, H. P., 101, 201, 202
Knudsen, J. P., 382, 383, 462 von Nardroff, R., 123, 124
Kottler, F., 286 Natl. Bur. Standards, 339
Kratky, 0., 7, 102, 103, 104, 105, 106, Nelson, J. B., 224
106', 107, 108, 109, llO, ll l, 111', ll2, Neynaber, R. H., 126"
113, ll4, ll4', 384, 436, 456', 457,; Noakes, G. E., 125, 126
464, 465, 466, 467, 468, 469, 470, 471, Van Nordstrand, R. A., 126', 126"
472, 473, 474, 475, 476 Nowacki, W., 127, 127'
Krishnamurti, P., 287, 288, 289, 290, 291, Oster, G., 128, 129, 298, 371, 487, 488
292, 293, 294, 295, 304
Krueger, H. H. A., 251 Parker, E. A., 410, 411
Pasternack, S., 258
LaMer, V. K., 296, 331, 332, 333, 334, Patterson, A. L., 299, 300, 301
350 Penley, H. H., 130
Lely, J. A., 223 Perron, R. A., 483
Lennard-Jones, J. E., 262 Perutz, M. F., 407, 484
Leonard, B. R., Jr., 115, 459, 477, 478, Pfund, A. H., 324
495 Philippoff, W., 360, 368, 369
Li, Y. Y., 115', 115", 146', 146" Phillipovich, V. N., 131'
Lipson, H., 77, 224 Pijper, A., 340
Ljunggren, T., 335 Pirie, N. W., 385
Lonsdale, K., 116 Ponder, E., 420
Love, A. E. H., 336 Porai-Koshits, Ye. A., 130', 131, 131',
Lund, L. H., 59, 117, 118, 119, 120 132
Porod, G., 97', 107, 108, 109, 110, 111,
MacArthur, I., 478'
133, 134, 135, 13~ 137, 369', 457, 469
Maitra, J. C., 2
Posner, A., 138, 139, 139', 139", 485
Mark, H., 431, 432, 433, 479, 481, 482
Preston, T., 341
Marsden, S. S., Jr., 367
Prins, J. A., 301', 302, 303, 309
Matalon, H., 372
Mathematical Tables Pr., 297 Raether, H., 140
Mathews, :M. B., 364 Raman, C. V., 304, 305
Mattoon, R W., 364, 365, 366 Ramanathan, K. R., 305
Maun, E. K., 284 Randall, J. T., 305'
Mazur, J., 285 Rayleigh, Lord, 342, 343
McBain, J. W., 367 Riley, D. P., 128, 129, 141, 142, 143, 144,
McReynolds, A. W., 252 224, 370, 371, 372', 373, 403, 486, 487'
McRoberts, T. S., 372' 488

Note: Numbers refer to numbered items in the Bibliography, not to pages.

264 SMALL-ANGLE SCATTERING OF X-RAYS

Riseman,-J"., 145 Teichner, S., 91

Ritland, H. N., 145', 221, 225, 226, 460, Todes, 0. M., 280
461, 490, 491, 492 Toraldo Di Francia, G., 354
Ritter, H. L., 146 Treer, R., 113, 384
Robinson, W. H., 146', 146h Treiber, E., 114, 222
Roess, L. C., 41, 147, 148, 154, 155 Turkevich, J., 159, 160, 161, 162
Rohrlich, F., 310 Turner, R. W., 75
Ruck, H., 97' Tweet, A. G., 162'
Ruedy, R., 344, 345, 346, 347
Ruess, G., 149 Van de Hulst, H. C., 355
Ryde, J. N., 348 Van Rijssel, T. W., 223
Vineyard, G. H., 120, 163, 164, 165, 166,
Saches, A., 99 167, 168, 168', 498
Schaad, J. A., 411
Sche.uenstein, E., 470, 471 We.Iker, C. B., 169, 170
Schmidt, P., 149' Wallace, J. W., 250
Schmitt, F. 0., 493 Warren, B. E., 10, 171, 172, 173, 174,
Schneider, M., 434 272, 307' 308
Schossberger, F., 111', 472 Weber, H. H., 475
Schulman, J. H., 372, 372', 373 Weibull, C., 381, 499
Schurz, J., 114 Weidinger, A., 439
Sekore., A., 112, 113, 436, 470, 471, 473, Weiss, R. J., 252, 255, 256, 257, 258
474, 475 West, W. J., 175, 176
Sen, M. K., 494 Whitman, J ., 177, 178
Serikov, A. S., 26 Wiederhom, N. M., 503, 504
Shacklett, R., 149' Wild, R. L., 179, 180, 181, 238
She.rrah, P. C., 150 Wilkinson, M. K., 254'
Shenfil, L., 25, 151, 211, 227, 228, 229 Winter, U., 506
Shishakov, N. A., 152 Wood, L.A., 182
Shull, C. G., 34, 41, 148, 153, 154, 155, Wooster, W. A., 77
254' Worthmann, W., 114'
Shulman, S., 376, 377, 478, 495 Wright, B. A., 483, 500, 501, 502, 503,
Shurman, M. M., 496 504
Sinclair, D., 349, 350 Wrinch, D., 505
Sivarajan, S. R., 350' Wurster, A., 476
Slack, C. M., 29, 208, 230 Wyckoff, R. W. G., 412, 505', 505"
Smith, A. E., 156
Smith, A. F., 378, 480 van der Wyk, A. J. A., 356
Smoluchowski, R., llh, 74'", 74.. , 75,
115', 115"' 146', 146' Yudowitch, K. L., 56, 158, 178, 183, 184,
Sokolov, U. G., 132 185, 186, 187, 188, 189, 190, 191, 192,
Solomon, E., 96, 231 193, 194, 195, 196, 239, 240, 241, 242,
Spencer,R.C., 232, 233, 234, 235, 236, 237 243, 244, 382, 383, 456
Stearns, R. S., 365, 366 Yvon, M. J., 197
Stephenson, S. R., 139"
Stewart, G. W., 306 Zahn, H., 506
Stone, L. L., 158, 497 Zernicke, F., 309
Storruste, A., 351 Zinn, W. H., 61
Stratton, J. A., 352, 353 Zubko, A. M., 27

Note: Numbers refer to numbered items in the Bibliography, not to pages.

AUTHOR INDEX TO TEXT
Allison, 8. K., 34, 95, 109 Evans, R. C., 98
Anderegg, J. W., 170 Ewald, P. P., 3, 163, 166, 209
Antzenberger, P., 180
Arndt, U. W., 96 Fankuchen, 1.,97, 109, 144, 145, 184, 185
Fournet, G., 34, 44, 47, 48, 50, 60, 63, 65,
Banerjee, K., 110 67, 102, 116, 117, 119, 130, 136, 138,
Barrett, C. 8., 202, 210, 214 147, 160, 162, 170, 171, 175, 180, 194,
Bateman, J. B., 174 199, 216
Bear, R. 8., 88, 89, 184 Franklin, R. E., 119, 165, 166, 190
Beeman, W.W., 95, 96, 97, 109, 116, 168, Friedrich-Freska, H., 174
169, 170, 171. 172 Furnas, T. C., Jr., 108
BelMoch, B., 203, 214
Bernal, J. D., 144, 145, 175, 185 Gingrich, N. 8., 49, 50, 57
Bertaut, F., 16, 163, 166 Green, H. 8., 41, 42, 44, 46, 67, 136
Bhatia, A. B., 47 Griffoul, R., 64
Biscoe, J., 188 Guinier, A., 25, 47, 64, 67, 101, 106, 116,
Blin, J., 195, 198, 213, 214 117, 119, 147, 151, 170, 195, 198, 203,
Boggs, E. M., 42 204
Bolduan, 0. E. A., 88, 89, 184
Born, M., 41, 44, 46, 67, 136 Hach, K. M., 161, 193, 194
Bragg, W. H., 58 Hagg, G., 106
Brindley, G. W., 50 Harkins, W. D., 141
Brusset, H., 190, 191 Hayes, S., 195
Bueche, A. M., 74, 75, 77 Heikens, D., 186
Burgers, W. G., 106 Henke, B., 95, 108
Hermans, P.H., 179, 181, 186, 216
Carlisle, C.H., 175 Hillier, J., 162
Cauchois, Y., 106 Hirsch, P. B., 98
Coates, W. A., 96 Hosemann, R., 2, 70, 94, 114, 143, 151,
Compton, A. H., 34, 95, 109 153, 179, 180, 181, 186
Cowley, J., 207, 216 Hsu, 8. 8., 174
Curien, H., 123, 124 Hughes, E. W., 175

Danielson, W. E., 54, 55, 107, 131 James, R. W., 34

Debye, P., 8, 35, 43, 47, 74, 75, 77, 160 Janeschitz-Kriegl, H., 182
Dervichian, D. G., 170 Johann, H. H., 100, 101, 103
DuMond, J. W. M., 54, 55, 95, 103, 106, Johansson, T., 100, 101, 103
107, 108, 116, 131
Kaesberg, P., 109, 110, 116, 169, 170, 172
Ehrenfest, P., 141, 142, 143, 160 Kahovek, L., 118, 159
Einstein, A., 47 Karlsson, N., 106
Eisenstein, A. 8., 49, 50, 57 Kellar, J. N., 98
Elkin, P. B., 192 Kiessig, H., 185, 186
265
266 AUTHOR INDEX TO TEXT

Kirkpatrick, P., 85, 95 Riseman, J., 154

Kirkwood, J. G., 41, 42 Ritland, H. N., 109, 169
Knudsen, J.P., 174 Rodriguez, A. E., 44
Kratky, 0., 21, 26, 27, 40, 47, 70, 81, 95, Roess, L. C., 19, 115, 151, 153, 154, 155,
118, 119, 120, 122, 123, 124, 174, 177, 161, 192
179, 180, 181, 186 Ross, P. A., 85, 95
Krishnamurti, P., 3 Ruck, H., 159
Krishnan, K. S., 47 Rushbrooke, G. S., 67

von Laue, M., 65, 198 Schauenstein, E., 186

Lave.I, J., 123 Scherrer, P., 164
Lennard-Jones, J. E., 48 Schmidt, P., 116, 172, 216
Leonard, B. R., 170 Schneider, M., 184
Lipson, H., 97, 99 Scoins, H. I., 67
Sekora, A., 174, 186
MacGillavry, C. H., 65 Sh;i.cklett, R., 106
Maitra, J. C., 110 Shenfil, L., 54, 55, 107, 131
Mark, H., 179, 216 Shull, C. G., 19, 115, 151, 153, 154, 155,
Mattoon, R. W., 141 161, 192
Menke, H., 35, 160 Shulman, S., 170
Smoluchowski, R., 195
Nelson, J. B., 97, 99 Stearns, R. S., 141
Neugebauer, T., 20 Strijk, B., 65
Van Nordstrand, R. A., 161, 193, 194
Tiedema, T. J., 106
Turkevich, J., 162
Olmer, P., 123 Tweet, A. G., 214, 215
Oster, G., 45, 144, 146, 147, 174
Overbeck, J. T. G., 136, 160 Verwey, E. J. W., 136, 160

Patterson, A. L., 64 Walker, C. B., 198, 203, 204, 216

Perutz, M. F., 130, 132, 160, 171 Warren, B. E., 3, 110, 159, 188
Philippoff, W., 144 Watson, G. L., 152, 160
Ponder, E., 176, 216 Weidinger, A., 186
Porod, G., 12, 21, 26, 27, 35, 40, 47, 70, Whittaker, E. T., 152, 160
74, 78, 81, 118, 119, 158, 159, 177, 181, Wilsdorf, H., 106
194, 195 Wilson, A. J.C., 11, 163, 164, 166
Prins, J. A., 35, 40, 47, 181 Wurster, A., 177

Raman, C. V., 44 Yudowitch, K. L., 41, 54, 57, 94, 162, 174
Rayleigh, Lord, 19, 37 Yvon, M. J., 41, 42, 47
Riley, D. P., 45, 96, 97, 99, 144, 146, 147,
174, 190, 204 Zernicke, F., 35, 40, 47, 181
SUBJECT INDEX
Absolute measurements, 121 Crysta.ls, 52
Adsorption methods, 161 Cylinder, homogeneous, 19, 27
Age-hardening, 199
Albumin, 105, 169 Debye-Scherrer lines, 163, 196
Alloys, Al-Ag, 75, 165, 198, 203, 208 Debye's model (see Spheres, he.rd)
Al-Cu, 211 Disc, homogeneous, 21, 23
Alumina., 193 Distance of heterogeneity, 81, 158
Angular uncertainty, 87
Egg yolk, 105, 176
Approximation for ta.ii of low-angle scat-
Electron microscope, 161
tering curve, 67, 80, 115
Ellipsoid, homogeneous, 19, 26, 169
Argon, 48, 58, 59
Ellipsoids, distribution of sizes, 154
Equation of state, 42, 46
Be.binet's theorem, 38
Exponential approximation, 25, 27, 30,
Beam stops, 121, 124
114, 128
for absolute measurements, 121
Bragg's law, use of, 148 Fibers, 177, 185
Filters, be.le.need (Ross), 85, 95
Carbon hie.ck, 105, 188 Flocculation, 194
Catalysts, 192 Fluids (see Particles)
Cellulose, 177, 179, 180, 181, 183 Fluids, theories of, 41, 45
Characteristic function, isolated particle, Form factor, 3
12 Fourier transformations, 4, 16, 18, 57,
system of particles, 77, 78 206
Che.rcoe.l, activated, 191
Chrysotile, 105, 184 Globulin, 169
Coiled cha.in molecules, 177 Gold, colloidal, 162
Cold-worked mete.ls, 195 Guinier's law (see Exponential approxi-
Colle.gen, 184 mation)
Collime.tors, 86
optimum, 89 Helium, liquid, 214
with circular apertures, 91 Hemocye.nine, 105, 174
with rectangular apertures, 91 Hemoglobin, 104, 105, 130, 137, 144, 170,
with slits of infinite height, 86 171, 173, 176
Colloidal solutions, 194 Heterogeneous matter, 70
Complementary objects, 38, 81 High polymers, 176
Compressibility, 47, 213 Hypotheses Hi and H2, 30, 60
Compton scattering, 5
Correction, beam height, beam of finite Inertial distance, 30
height, 118 Inhomogeneity, range of, 158
beam of infinite height, 114, 116 Ionization chamber, 123
beam width, 112
Counter, Geiger-MUiler, 85, 87, 89, 93, 121 >.-point, 214
proportional, 96 Latex, 131, 162
267
268 SUBJECT INDEX

Laue scattering, 65 Photographic detection, 87, 89

Limiting value, zero-angle intensity, 46, Polyamides, 185
71 Polyethylene, 183
Polymers, high, 176
Macromolecules, solutions of, 167 Potential energy, interparticle, 41
Maximum in diffraction patterns of fluids, Probability function P(r), 40, 146
50, 140, 147, 171 Proteins, 167
Metals, cold-worked, 195
Microcavities in metals, 196, 213 Radius of gyration, 24, 26, 127, 130, 149,
Molecules, 167 167, 169, 189
coiled chain, 177 Rayon, 183
Monochromatic source, Al Ka, 94 Reciprocity principle, 38
Monochromatization by total reflection, Red cells, 105
95, 108 Rod, homogeneous, 20, 23
Monochromators, 96
bent crystal, 100 Scattering, Thomson, 6
plane crystal, 96 Secondary maxima, 54, 130
Fankuchen cut, 97 Silk, 186
point focusing, 104 Silver, colloidal, 162
two crystal, bent, 102 Slits, construction, 120
plane, 109 scattering from, 120
Multiple scattering, 5 Soap, 175
Solid solutions, equilibrium, 197
Naphthalene black, 159 supersaturated, 199
Nickel, catalytic, 192 Specific surface, 156, 188, 194
Nylon, 183 Sphere, homogeneous, 19, 20
Spheres, distribution of sizes, 154
Paracrystal, 143 hard, 43, 54, 58, 59, 137
Parasitic scattering, 84, 87 mixtures of different sizes, 68
Particle, centrosymmetric, 6, 8 Stacking faults, 202, 210
fixed, 5, 28 Statistical distribution of particles, 151
moving, 7, 24 Superposition principle, 42, 45
spherical, 10, 28 Surface area, internal, 80
Particles, identical and densely packed, particle, 13, 67
experiment, 135
theory, 33, 40 Total reflection, 95, 108
identical and widely separated, exper- Total scattered energy, 18, 81, 110, 158
iment, 126
theory, 35 Vacuum chambers, 123
in a homogeneous medium, 40 Virus molecules, 170, 172, 175, 184
mixtures of several types, 65, 149 Viscose, 179, 182
oriented, experiment, 134
theory, 23, 60 Zones, Guinier-Preston, 201, 204, 211