BUEHLER® SUM-MET ™
The Science Behind Materials Preparation
A Guide to Materials Preparation & Analysis

BUEHLER® SUM-MET™ - The Science Behind Materials Preparation
Copyright © 2007 Buehler Ltd.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in
any form or by an means, electronic, mechanical, photocopying, recording or otherwise, without permission of
the copyright holder.

ISBN Number: 0-9752898-0-2

FN NUMBER: FN01255

Printed in the United States of America

First Printing: 1M0107

This book was produced digitally by Buehler Ltd.

Trademarks
All terms mentioned in this book that are known to be trademarks or service marks have been appropriately
capitalized.

Warning and Disclaimer

Every effort has been made to make this book as complete and as accurate as possible, but no warranty or fit-
ness is implied. The information provided is on an “as is” basis. The authors and the publisher shall have neither
liability nor responsibility to any person or entity with respect to any loss or damages arising from the information
contained in this book.

Table of Contents

Table of Contents............................................................ 4 Magnesium.......................................................36

Introduction........................................................................ 6 Beryllium............................................................37

Sampling............................................................................... 7 Low Melting Point Metals:

Sb, Bi, Cd, Pb, Sn and Zn.................................38

Goals of Specimen Preparation............................. 9
Refractory Metals: Ti, Zr, Hf, Cr, Mo,
Method Development.................................................... 9
Nb, Re, Ta, V and W...........................................40
Sectioning..........................................................................10
Titanium.............................................................40
Abrasive Wheel Cutting..........................................10
Zirconium and Hafnium...............................41
Precision Saws............................................................12
Other Refractory Metals: Cr, Mo, Nb,
Mounting of Specimens............................................14
Re, Ta, V and W..................................................43
Clamp Mounting.......................................................14
Ferrous Metals............................................................45
Compression Mounting.........................................14
Copper, Nickel & Cobalt..........................................48
Castable Resins for Mounting..............................15
Copper................................................................45
Edge Preservation....................................................17
Nickel...................................................................49
Grinding..............................................................................20
Cobalt..................................................................50
Grinding Media..........................................................20
Precious Metals..........................................................51
Grinding Equipment................................................22
Thermally-Spray Coated Specimens..................53
Polishing.............................................................................23
Sintered Carbides.....................................................55
Mechanical Polishing..............................................24
Ceramics.......................................................................57
Electrolytic Polishing...............................................24

Manual “Hand” Polishing........................................24 Composites.................................................................59

Automatic Polishing................................................25 Printed Circuit Boards.............................................61

Polishing Clothes......................................................25 Eletronic Materials....................................................62

Polishing Abrasives..................................................27 Polymers......................................................................65

Examples of Preparation Procedure.................29 Etching.................................................................................67

The Traditional Method..........................................29 Etching Procedures..................................................67
Contemporary Methods.........................................29 Selective Etching.......................................................68

Procedures for Specific Materials......................32 Electrolytic Etching and Anodizing....................71

Periodic Table of Elements....................................33 Heat Tinting................................................................71

Light Metals: Al, Mg and Be...................................34 Interference Layer Method...................................72

Aluminum..........................................................34 Commonly Used Etchants for

Metals and Alloys......................................................73

Table of Contents

Light Optical Microscopy.........................................78 Laboratory Safety...................................................... 101

The Light Microscope..............................................79 Laboratory Equipment........................................ 102

Microscope Components......................................79 Personal Protective Equipment (PPE)............. 102

Resolution...................................................................82 Chemicals, Storage and Handling.................... 103

Depth of Field............................................................83 Etchants..................................................................... 103

Imaging Modes.........................................................83 Solvents..................................................................... 104

Microindentation Hardness Testing..................87 Acids........................................................................... 106

The Vickers Test.........................................................87 Bases........................................................................... 110

The Knoop Test..........................................................88 Other Chemicals..................................................... 110

Factors Affecting Accuracy, Precision Summary......................................................................... 112

and Bias........................................................................89 References..................................................................... 113

Automation.................................................................91 Appendices.................................................................... 115

Image Capture & Analysis.......................................93 ASTM Metallography Standards....................... 123

Acquisition..................................................................93
ISO Standards.............................................................. 124
Clarification.................................................................96
Other National Standards..................................... 125
Operator Interactive Measurements.................96
Buehler Trademarks.....................................128
Automated Measurements...................................96
Index................................................................................... 129
Thresholding..............................................................96
Worldwide Sales Offices....................................... 132
Binary Operations.....................................................98

Common Applications............................................99

Written by George F. Vander Voort, Director, Research & Technology, with contributions from Wase Ahmed, George Blann, Paula Didier, Don Fisher,
Scott Holt, Janice Klansky and Gabe Lucas.

Introduction

INTRODUCTION used in polishing exhibit a range of resilience

depending on whether they are non-woven poly-
Proper specimen preparation is essential if the
meric or chemo-textiles, woven napless materials,
true microstructure is to be observed, identi-
or napped materials. Resilience influences edge
fied, documented and measured. Indeed, when
retention, relief control and cutting rate (“aggres-
metallographers have difficulty seeing or mea-
siveness”). Polishing must remove the damage
suring microstructures, these problems can
remaining after grinding, with the final step reduc-
almost always be traced to improper specimen
ing the damage to where it is insignificant, or can
preparation procedures, or failure to prop-
be removed by the etchant. In this way, the true
erly execute the preparation steps. When a
microstructure is revealed. For most metals and
specimen is prepared properly, and the structure
alloys, and many non-metallic materials, polishing
is revealed properly, or a phase or constitu-
down to a 1-, or even a 3-µm diamond finish may
ent of interest is revealed selectively and with
yield a surface suitable for routine examination,
adequate contrast, then the actual tasks of
although finer abrasives are usually needed for
examination, interpretation, documentation and
more critical work, such as failure analysis.
measurement are generally quite simple. Experi-
ence has demonstrated that getting the required Historically, polishing was performed with abra-
image quality to the microscope is by far the sive slurries, most commonly using alumina, but
greatest challenge. Despite this, the specimen with limited use of MgO, Cr2O3 or CeO for specific
preparation process is often treated as a trivial materials. Diamond abrasives were introduced in
exercise. However, specimen preparation qual- the late 1940s. Cloths such as canvas, billiard, felt,
ity is the determining factor in the value of the and cotton are not used as much today as they
examination. This is in agreement with the classic tend to promote relief and edge rounding, or
computer adage, “garbage in = garbage out.” other artifacts. Modern preparation methods have
focused on reducing the number of SiC grinding
Historically, people tend to divide specimen prep-
steps and replacing SiC paper with other abrasives
aration (after mounting) into two steps, grinding
and surfaces, although there are some materials
and polishing, further sub-divided, somewhat
that respond better to grinding with a series of
arbitrarily, into coarse and fine stages. Years ago,
finer and finer SiC papers. But, most materials can
these divisions were quite clear. Both processes
be prepared with one grinding step, often called
involve the same mechanism, chip formation by
planar grinding when an automated system is
micro-machining. Grinding is abrasive material re-
used, and two or more polishing steps. The second
moval from the specimen surface using randomly
step uses relatively hard, non-resilient surfaces
oriented, fixed abrasives on a substrate such as
with a relatively coarse diamond size (15- or 9-µm,
paper or cloth. The abrasive size range has gener-
for example) and the abrasive is added as a slurry
ally been in the range of about 150 to 5 µm, with
or spray. Initially, the particles move between the
SiC paper used exclusively until recently.
specimen surface and the substrate. Technically,
Historically, polishing has been defined in terms this is lapping. But, after a short time, the diamond
of industrial polishing rather than metallographic particles become embedded in the surface and
polishing, simply stating that it is a smoothing of produce cutting action. We tend to call this pol-
the surface to a high luster. A polished surface was ishing, although the historic definitions are not as
defined as “a surface that reflected a large propor- clear here. If the substrate is a rigid grinding disk,
tion of incident light in a specular manner.”These the action may be more correctly called grinding
definitions are true as far as they go; but they are than polishing, as the surface finish will be dull.
inadequate for metallography. Polishing also pro- The same abrasive when used on a hard, woven
duces micro- machining action, but scratch depths surface such as an UltraPol silk cloth, will yield a
are shallower in polishing, as the abrasives are much more lustrous appearance, but with less
smaller. They are not fixed initially to a substrate, stock removal and shallower scratches. Napped
but are applied as a paste, slurry or spray. After a cloths, if they are used, are generally restricted to
short time, the abrasive particles become fixed the final step with the finest abrasives.
in the cloth producing cutting action. The cloths

Sampling

Sampling Many metallographic studies require more than

one specimen and sectioning is nearly always
The specimens selected for preparation must be required to extract the specimens. A classic
representative of the material to be examined. example of multiple specimen selection is
Random sampling, as advocated by statisticians, the evaluation of the inclusion content of steels.
can rarely be performed by metallographers. A One specimen is not representative of the whole
good exception is the testing of fasteners where lot of steel, so sampling becomes important. ASTM
a production lot can be sampled randomly. But, standards E 45, E 1122 and E 1245 give advice on
a large forging or casting cannot be sampled at sampling procedures for inclusion studies.To study
random as the component would be rendered grain size, it is common to use a single specimen
commercially useless. Instead, test locations are from a lot. This may or may not be adequate,
chosen systematically based on ease of sampling. depending upon the nature of the lot. Good engi-
Many material specifications dictate the sampling neering judgment should guide sampling, in such
procedure. In failure analysis studies, specimens cases. In many cases, a product specification may
are usually removed to study the origin of the define the procedure rigorously. Grain structures
failure, to examine highly stressed areas, to exam- are not always equiaxed and it may be misleading
ine secondary cracks, and so forth. This, of course, to select only a plane oriented perpendicular to
is not random sampling. It is rare to encounter the deformation axis, a “transverse” plane, for such
excessive sampling as testing costs are usually a study. If the grains are elongated due to process-
closely controlled. Inadequate sampling is more ing, which does happen, the transverse plane will
likely to occur. usually show that the grains are equiaxed in shape
and smaller in diameter than the true grain size.To
Normally, a specimen must be removed from
study the effect of deformation on the grain shape
a larger mass and then prepared for examina-
of wrought metals, a minimum of two sections are
tion. This requires application of one or more
needed — one perpendicular to, and the other
sectioning methods. For example, in a manu-
parallel to, the direction of deformation.
facturing facility, a piece may be cut from a bar
of incoming metal with a power hacksaw, or an
abrasive cutter used dry, i.e., without a coolant.
The piece is then forwarded to the laboratory Helpful Hints For
where it is cut smaller to obtain a size more Sampling
convenient for preparation. All sectioning
When developing a sampling
processes produce damage; some methods
plan for an unfamiliar part or
(such as flame cutting or dry abrasive cutting)
produce extreme amounts of damage. Tradi- component, determine the orientation of the
piece relative to the original wrought or cast
tional laboratory sectioning procedures, using
starting material and prepare sections on the
abrasive cut-off saws, introduce a minor amount
longitudinal, transverse and planar surfaces,
of damage that varies with the material being
or radial and transverse surfaces, to reveal the
cut and its thermal and mechanical history. It is
desired information. Remember that the micro-
generally unwise to use the original face cut in
structure may look much more homogeneous
the shop as the starting point for metallographic
than it is on a transverse plane compared to a
preparation as the depth of damage at this
location may be quite extensive. This damage longitudinal or planar surface. If the starting
material was cold worked, or not fully recrys-
must be removed if the true structure is to be
revealed. However, because abrasive grinding and tallized after hot working, grain structures
will appear to be equiaxed and smaller on a
polishing steps also produce damage, where
transverse plane, but non-equiaxed and larger
the depth of damage decreases with decreasing
on a longitudinal or planar surface.
abrasive size, the preparation sequence must
be carefully planned and performed; otherwise,
preparation-induced artifacts will be interpreted
as structural elements.

Goals of Specimen Preparation

GOALS OF SPECIMEN If these characteristics are met, then the true

PREPARATION structure will be revealed and can be interpreted,
measured and recorded. The preparation method
The preparation procedure and the prepared should be as simple as possible, should yield con-
specimen should have the following characteris- sistent high quality results in a minimum of time
tics to reveal the true microsctructure*: and cost and must be reproducible.

• Deformation induced by sectioning, grinding Preparation of metallographic specimens [1-

and polishing must be removed or be shallow 3] generally requires five major operations: (a)
enough to be removed by the etchant. sectioning, (b) mounting (optional), (c) grinding,
(d) polishing and (e) etching (optional).
• Coarse grinding scratches must be removed;
very fine polishing scratches may be tolerable
for routine production work. Method Development
The methods presented in this book use times
• Pullout, pitting, cracking of hard particles, that are conservative so that the vast majority of
smear, and other preparation artifacts, must be compositions of alloys or materials of a specific
avoided. type can be prepared to yield the true microstruc-
ture without artifacts or observable scratches.
• Relief (i.e., excessive surface height variations They were developed using an 8-inch (200 mm)
between structural features of different hard- diameter platen system with six 1.25-inch (30 mm)
ness) must be minimized; otherwise portions of diameter mounted specimens. For each category,
the image will be out of focus at high magnifica- except beryllium, a wide variety of specimens with
tions. different processing histories were prepared using
the stated methods.
• The surface must be flat, particularly at
edges (if they are of interest) or they cannot For any metal or material category, there are
be imaged. compositions or processing conditions that
make those specimens easier, or more difficult,
• Coated or plated surfaces must be kept flat if to prepare compared to the “average” specimen.
they are to be examined, measured or photo- In general, even the most difficult specimens of
graphed. a given type can be prepared perfectly using
these methods with no more than one etch and
• Specimens must be cleaned adequately re-polish (repeat last step) cycle. For perfect color
between preparation steps, after preparation, etching results, it may be necessary to follow the
and after etching. cycle with a brief vibratory polish, depending upon
the material.
• The etchant chosen must be either general or
selective in its action (reveal only the phase When working with easier to prepare specimens,
or constituent of interest, or at least produce or for routine work where the same degree of
strong contrast or color differences between perfection is not required, the user can reduce the
two or more phases present), depending upon times, or eliminate one or two steps, or the final
the purpose of the investigation, and must step, and get satisfactory results. Modifications
produce crisp, clear phase or grain boundaries, to the recommended procedures are left to the
and strong contrast. user based upon their knowledge of the degree
of difficulty in preparing their specimens and the
desired quality of preparation.

* Not all of these apply to every specimen.

Sectioning

SECTIONING
Bulk samples for subsequent laboratory sectioning
may be removed from larger pieces using methods
such as core drilling, band or hack-sawing, flame
cutting, or similar methods. However, when these
techniques are used, precautions must be taken to
avoid alteration of the microstructure in the area
of interest. Laboratory abrasive-wheel cutting is
recommended to establish the desired plane of
Figure 2. Cutting damage (top) and a “burr” after
polish. In the case of relatively brittle materials, sectioning of an annealed CP titanium specimen
sectioning may be accomplished by fracturing (mod. Weck’s reagent, 100X, polarized light plus
sensitive tint).
the specimen at the desired location.
Wheels consist of abrasive particles, chiefly
alumina or silicon carbide, and filler in a binder
Abrasive-Wheel Cutting material that may be a resin, rubber, or a mixture
The most commonly used sectioning device in of resin and rubber. Alumina (aluminum oxide) is
the metallographic laboratory is the abrasive the preferred abrasive for ferrous alloys and silicon
cut-off machine, Figure 1. All abrasive-wheel carbide is the preferred abrasive for nonferrous
sectioning should be performed wet. An ample metals and minerals. Wheels have different bond
flow of coolant, with an additive for corrosion strengths and are recommended based on the
protection and lubrication, should be directed suitability of their bond strength and abrasive
into the cut. Wet cutting will produce a smooth type for the material to be sectioned. In general,
surface finish and, most importantly, will guard as the hardness of a material increases, abrasives
against excessive surface damage caused become dull more quickly, and the binder must
by overheating. Abrasive wheels should be break-down and release the abrasives when
selected according to the manufacturer’s they become dull so that fresh abrasive particles
recommendations. Table 1 summarizes Buehler’s are available to maintain cutting speed and ef-
recommendations for our machines. Specimens ficiency. Consequently, these wheels are called
must be fixtured securely during cutting, and “consumable” wheels because they wear away
cutting pressure should be applied carefully to with usage. If they do not wear at the proper rate,
prevent wheel breakage. Some materials, such as dull abrasives will rub against the region being
CP (commercial purity) titanium, Figure 2, are quite cut generating heat and altering the existing true
prone to sectioning damage. microstructure. If this heat becomes excessive, it
can lead to grain or particle coarsening, softening
or phase transformations, and in extreme case, to
burning or melting. Different materials have dif-
ferent sensitivities to this problem. But, the need
to balance the wheel break-down rate with the
hardness of the piece being sectioned, produces
the various recommendations listed for cutting
different materials and metals with different hard-
nesses, such as steels.

The size of the cut-off machine also affects wheel

parameters. As the diameter of the wheel is
increased, to permit sectioning of larger speci-
mens, the wheel thickness is generally increased.
Even so, for a given diameter wheel, there may be
a range of thicknesses available. Thicker wheels
are stronger but remove more material in the cut.
This may, or may not, be a problem. But, thicker
Figure 1. Delta Automatic Orbital Cutter

Sectioning

Table 1. Buehler’s Abrasive Cutting Recommendations

Available Diameters
Recommended Use Bond Abrasive (Inches) (mm)
General Usage Blades
Tools Steels 60 HRC and Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,
Above Carburized Steels 14, 16, 18 350, 400, 455
Hard Steel 50 HRC Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,
14, 16, 18 350, 400, 455
Medium Hard Steel Rubber Resin Al2O3 9, 10, 12, 230, 250, 300,
35-50 HRC 14, 16, 18 350, 400, 455
Soft or Annealed Steel Rubber Al2O3 9, 10, 12, 230, 250, 300,
15-35 HRC 46-90 HRB 14, 16, 18 350, 400, 455
Medium Hard Nonferrous Materials, Rubber SiC 9, 10, 12, 230, 250, 300,
Uranium, Titanium, Zirconium 14, 16, 18 350, 400, 455
Soft Nonferrous Materials Rubber SiC 9, 10, 12, 230, 250, 300,
Aluminum, Brass, etc. 14, 16, 18 350, 400, 455
Superalloys Rubber Al2O3 10, 12, 250, 300,
14, 16, 18 350, 400, 450
Thin Blades to Minimize Kerf Loss and Cutting Deformation
Tool, Hard Steel, Rubber Al2O3 5, 9, 10*, 12 130, 230,
≤ 45 RC 250*, 300
Medium Hard, Soft Steel Rubber Al2O3 5, 7, 9, 130, 180, 230,
≥ 45 HRC 10, 12, 14 250*, 300, 350
Hard or Soft Nonferrous Materials Rubber SiC 7 180
*Rubber Resin Bond
Refer to Buehler’s Consumables Buyers Guide for ordering information, and exact dimensions of arbor sizes, outer diameter, and thickness of Buehler Metabrase,
Delta, and Acu-Thin Abrasive Cut-off Wheels

wheels do generate more heat during cutting oriented to minimize the contact area with the
than a thinner wheel, everything else being held wheel. For small parts, this is generally easy to do,
constant. Consequently, for cases where the kerf but for larger parts, it may not be possible to orient
loss or heat loss must be minimized, select the thin- the specimen properly for optimal cutting. When
nest available wheel of the proper bond strength a round wheel is used to cut a round bar in chop
and abrasive. Buehler’s AcuThin Cut-off Wheels mode, the contact area is initially very small. As the
offer the thinnest possible wheels for delicate cut progresses, the cut widens until the maximum
abrasive sectioning. Buehler also has diamond diameter is reached. After this, the contact area
cut-off blades with resin bonded or Rimlock1  metal decreases until sectioning is completed. Under
bonded blade forms in sizes from 8-inch (203 mm) application of a constant load, the pressure on
to 12-inch (305 mm) diameters. Resin-bonded the abrasive particles in the cut decreases as the
diamond blades are ideal for cutting very hard contact area increases. If the pressure applied to
cemented carbide specimens; Rimlock or continu- the grains is inadequate for cutting, then heat is
ous rim blades are recommended for cutting rocks generated which may not be dissipated by the
and petrographic specimens. coolant, which causes deformation damage, phase
changes and possibly burning or melting.
Historically, the most common cutter design has
been the so-called “chop” cutter. Basically, the A chop cutter may be set up to pulse as the
blade is attached to a motor and the operator load is applied, that is, the wheel is fed into the
pulls a handle to drive the blade downward into specimen, and then feeding is halted momen-
the work piece. Because of the design, the blade tarily, Figure 3b. Pulsing causes the wheel to be
moves through an arc as it is pulled downward, stress shocked which removes both dull and
Figure 3a. For efficient cutting, the piece must be sharp abrasive from the wheel as new abrasive
is exposed to the cut. While a better cut may

1
Rimlock is a registered trademark of Felker Operations.
10
Sectioning

be obtained, cutting time and wheel wear are of permitting larger size pieces to be cut properly
increased. compared to a chop cutter with the same wheel
diameter.
Optimal cutting, with the least damage to the
specimen, is obtained by keeping the pressure on Another option for MACC is the “traverse and incre-
each abrasive particle in the wheel constant and ment” cutter design, Figure 3d. A strip of material
as low as possible by keeping the contact area is removed by moving either the wheel or the
between wheel and specimen constant and low. work piece perpendicular to the cut direction.
This idea is called the minimum area of contact After this, the cutter must move back to the side
cutting (MACC) method. Oscillation style cutters, where the cut began. The wheel is dropped down
such as the OscilliMet, minimize the contact area a fixed increment and another strip is removed. In
by oscillating perpendicular to the cut contact this approach, the wheel must move to the end of
area, Figure 3c. This approach has the advantage the stroke, return to the starting point, and then

Figure 3. Illustration of cutter types and the wheel path during sectioning

CUTTING STYLE AND WHEEL PATH

Wheel Path

Sample
3a Chop Cutting
The traditional form of machine
operation. Wheel contact arc is
governed by specimen size. Generally
a struggle with large/difficult parts.

3b Chop Cutting with Pulsing

Wheel contact still governed by specimen size.
The pulsing action gives a shock to the wheel
causing abrasives to break away. Wheel wear is
generally high with this method.

3c Oscillation
The wheel is reciprocated over a
small distance. This minimizes the contact
arc. Available only on manual labor-intensive
cutters.

3d Traverse and Increment

The wheel contact arc can be precisely
controlled via depth increment. The traverse
stroke must always exceed the part length
to avoid a change in wheel contact arc area.
Machine needs to be set for each part. Action
is slow.

3e Orbital
Similar in action to traverse and increment
but on a curved path. Simpler and quicker in
operation. Part size is irrelevant as the orbital
action produces a minimum contact arc area
during cutting.

11
Sectioning

be moved an increment downwards in each cycle.

Consequently, cut times are longer. The wheel tra-
verse length must be greater than the specimen
width.This machine must be programmed for each
size part to obtain cutting.

Orbital cutting, Figure 3e, combines the best

features of chop, oscillation and traverse-and-
increment cutters, while minimizing cutting time.
As the wheel is fed into the specimen, the arbor
is also rotated in a small ellipse planar with the
wheel permitting the wheel to cut a small strip
of material from the specimen. But no time is Figure 4. IsoMet 5000 Linear Precision Saw
lost, as with the traverse and increment cutter,
as the wheel is ready to cut on the original side. an abrasive cutter can be cut with a precision saw,
With the elliptical orbit and a controlled feed, the the cutting time will be appreciably greater but the
conditions for minimum area of contact cutting depth of damage will be much less. Precision saws
are obtained, irrespective of work piece size. are widely used for sectioning sintered carbides,
Special programming is not needed. The orbital ceramic materials, thermally sprayed coatings,
cutting concept is available in the Delta cutter printed circuit boards, electronic components,
from Buehler. These also feature Smartcut, which bone, teeth, etc. Table 2 lists selection criteria for
will sense motor overloading during a cut, if it precision saw blades. Buehler’s blades for precision
occurs, and reduce the feed rate automatically. saws are available with different abrasive sizes
When it is safe to return to the original feed rate, and bonds to provide optimum cutting for a wide
Smartcut will do so. variety of applications.

Precision Saws Helpful Hints For

Sectioning
Precision saws, Figure 4, are commonly used in
metallographic preparation to section materi- When cutting a difficult speci-
als that are small, delicate, friable, extremely men with the recommended
hard, or where the cut must be made as close as consumable abrasive wheel, if the cutting ac-
possible to a feature of interest, or where the tion has become very slow, pulse the applied
cut width and material loss must be minimal. As force. This will help break down the abrasive
the name implies, this type of saw is designed to bonding, exposing fresh, sharp abrasives to
make very precise cuts. They are smaller in size enhance the cutting action. However, if you are
than the usual laboratory abrasive cut-off saw using a resin- bonded diamond blade to cut
and use much smaller blades, typically from 3- cemented carbides, or other very hard materi-
to 8-inch (7.6 to 20.3 mm) in diameter.These blades als, do not pulse the applied force, as this will
can be of the non-consumable type, made of shorten wheel life.
copper-based alloys and copper plated steel with
diamond or cubic boron nitride abrasive bonded
to the periphery of the blade; or, they can be con-
sumable blades using alumina or silicon carbide
abrasives with a rubber-based bond. Blades for the
precision saws are much thinner than the abrasive
wheels used in an abrasive cutter and the load
applied during cutting is much less. Consequently,
less heat is generated during cutting and damage
depths are reduced. While pieces with a small sec-
tion size, that would normally be sectioned with

12
Sectioning

Table 2. Selection Criteria for Precision Saw Blades

3″ x 0.006″ 4″ x 0.012″ 5″ x 0.015″ 6″ x 0.020″ 7″ x 0.025″ 8″ x 0.035″

(75mm x (100mm x (125mm x (150mm x (180mm x (200mm x
Blade Series 0.2mm) 0.3mm) 0.4mm) 0.5mm) 0.6mm) 0.9mm)
Diamond Wafering Blades
Series 30HC Diamond, for use X** X** X
with plastics, polymers, and rubber
Series 20HC Diamond, for aggressive X* X X
general sectioning of ferrous and
nonferrous materials including
titanium alloys
Series 15HC Diamond, for general X X X X X X
use with ferrous and nonferrous
alloys, copper, aluminum, metal
matrix, composites, PC boards,
thermal spray coatings, and
titanium alloys
Series 20LC Diamond, for use with X* X
hard/tough materials structural
ceramics, boron carbide, boron
nitride, and silicon carbide
Series 15LC Diamond, for use with X X X X X X
hard/brittle materials structural
ceramics, boron carbide, boron
nitride, and silicon carbide
Series 10LC Diamond, for use with X X X* X
medium to soft ceramics, electronic
packages, unmounted integrated
circuits, GaAs, AIN and glass fiber
reinforced composites
Series 5LC Diamond, for use with X X
soft friable ceramics, electronic
packages, unmounted integrated
circuits, composites with fine
reinforcing media, CaF2, MgF2,
and carbon composites
IsoCut Wafering Blades
Cubic Boron Nitride (CBN) abrasive
blades work well for many tough
materials giving significantly shorter
cut times
For iron, carbon steels, high alloy X X X X X X
steels, cobalt alloys, nickel super-
alloys, and lead alloys
General Usage Abrasive Cut-off Bond/
Wheels 0.03″ (0.8mm) thick Abrasive
For ferrous materials, stainless R/Al2O3 X
steels, cast irons, and
thermal spray coatings
For tough nonferrous metals, R/SiC X
aluminum, copper, titanium,
uranium, zirconium
AcuThin Abrasive Cut-off Bond/
Wheels 0.019″ (0.5mm) thick Abrasive
For sectioning small, delicate
specimens or where minimal
deformation and kerf loss is the
primary concern
Tool, hard steel, ≥ 45 HRC R/Al2O3 X
Medium hard, soft steel ≤ 45 HRC R/Al2O3 X
* Alternate blade thickness of 0.020″ (0.5 mm) ** Alternate blade thickness of 0.030″ (0.8 mm)
For a complete listing of Buehler consumable supplies for use with the Isomet Precision Saws, please refer to Buehler’s Consumables Buyers Guide.

13
Mounting of Specimens

MOUNTING OF SPECIMENS Clamp Mounting

Clamps have been used for many years to
The primary purpose of mounting metallo-
mount cross sections of thin sheet specimens.
graphic specimens is for convenience in handling
Several specimens can be clamped conveniently
specimens of difficult shapes or sizes during the
in sandwich form making this a quick, convenient
subsequent steps of metallographic prepara-
method for mounting thin sheet specimens.
tion and examination. A secondary purpose is to
When done properly, edge retention is excellent,
protect and preserve extreme edges or surface
and seepage of fluids from crevices between
defects during metallographic preparation. The
specimens does not occur. The outer clamp
method of mounting should in no way be injurious
edges should be beveled to minimize damage
to the microstructure of the specimen. Pressure
to polishing cloths. If clamps are improperly used
and heat are the most likely sources of injurious
so that gaps exist between specimens, fluids and
effects.
abrasives can become entrapped and will seep out
Phenolic plastics were introduced to metal- obscuring edges and can cause cross contamina-
lography in 1928 for encapsulating specimens tion. This problem can be minimized by proper
by hot compression mounting. Prior to that tightening of clamps, by use of plastic spacers
time, specimens were prepared unmounted, or between specimens, or by coating specimen sur-
mounted in flours of sulfur, wax or low-melting faces with epoxy before tightening.
point alloys, such as Wood’s metal. These “mount-
ing compounds” were not without their problems. Compression Mounting
Introduction of polymers was a big improvement The most common mounting method uses
over these methods. Subsequently, many polymers pressure and heat to encapsulate the specimen
were evaluated for use as mounting compounds, with a thermosetting or thermoplastic mounting
as they were introduced to the market. Develop- material. Common thermosetting resins include
ment of castable resins in the 1950’s added new phenolic (PhenoCure resin), diallyl phthalate and
resins to the metallographer’s tool chest. The sim- epoxy (EpoMet resin) while methyl methacrylate
plicity of mounting without using a press, and their (TransOptic resin) is the most commonly used
low curing temperatures, made castable resins an thermoplastic mounting resin. Table 3 lists the
attractive alternative. characteristics of the hot compression mounting

Table 3. Characteristics of Buehler’s Compression (Hot) Mounting Resins

Best Edge Retention; Near Zero

Very Low Shrinkage; Best Edge electrical
Least Fine Particle Size; Retention; Very Resistance;
Materials Expensive Fills Smallest Crevices Low Shrinkage SEM - EDS/WDS Clear
Ceramics PhenoCure EpoMet F EpoMet G ProbeMet TransOptic
Steels PhenoCure EpoMet F EpoMet G ProbeMet TransOptic
Plated PhenoCure EpoMet F EpoMet G ProbeMet TransOptic
Layers
Aluminum PhenoCure ProbeMet TransOptic
Copper/ PhenoCure ProbeMet TransOptic
Brass
Black, Red, Black Black Bronze Transparent
Color or Green
Temperature 300°F/150°C 300°F/150°C 300°F/150°C 300°F/150°C 350°F/177°C
4200 psi/ 4200 psi/ 4200 psi/ 4200 psi/ 2100 psi/
Pressure 290 bar 290 bar 290 bar 290 bar 145 bar

14
Mounting of Specimens

resins. Both thermosetting and thermoplastic

materials require heat and pressure during the
molding cycle; but, after curing, mounts made
of thermoplastic resins must be cooled under
pressure to at least 70 °C while mounts made of
thermosetting materials may be ejected from
the mold at the maximum molding temperature.
However, cooling thermosetting resins under pres-
sure to near ambient temperature before ejection
will significantly reduce shrinkage gap formation. Figure 6. Excellent edge retention of a borided
Never rapidly cool a thermosetting resin mount 42CrMo4 alloy steel specimen mounted in EpoMet
resin (1000X, 2% nital).
with water after hot ejection from the molding
temperature. This causes the metal to pull away Also, placing a number of mounted specimens in
from the resin producing shrinkage gaps that a holder for semi- or fully-automated grinding and
promote poor edge retention, see Figure 5, be- polishing is easier with standard mounts than for
cause of the different rates of thermal contraction. unmounted specimens. Mounted specimens are
EpoMet resin, a thermosetting epoxy, provides the easier on the grinding/polishing surfaces than
best edge retention, Figure 6, of these resins and unmounted specimens.
is virtually unaffected by hot or boiling etchants
while phenolic resins are badly damaged.

Figure 7. SimpliMet 3000 Automatic Mounting Press

Castable Resins for Mounting

Cold mounting materials require neither pre-
ssure nor external heat and are recommended
for mounting specimens that are sensitive to
heat and/or pressure. Acrylic resins, such as Vari-
Dur and SamplKwick resins, are the most widely
Figure 5. Edge retention of this improperly carbur-ized
used castable resins due to their low cost and
8620 alloy steel was degraded by a shrinkage gap short curing time. However, shrinkage can be a
between the specimen and the phenolic mount: a)
top, 500X; b) bottom, 1000X (2% nital).
problem with acrylics. Epoxy resins, although
more expensive than acrylics, are commonly used
Mounting presses vary from simple laboratory because epoxy will physically adhere to specimens,
jacks with a heater and mold assembly to full auto- have low shrinkage, and can be drawn into cracks
mated devices, as shown in Figure 7. An advantage and pores, particularly if a vacuum impregnation
of compression mounting is production of a chamber, Figure 8, is employed and a low viscos-
mount of a predicable, convenient size and shape. ity epoxy, such as EpoThin resin, is used. Epoxies
Further, considerable information can be engraved are very suitable for mounting fragile or friable
on the backside – this is always more difficult with specimens and corrosion or oxidation specimens.
unmounted specimens. Manual “hand” polishing Dyes or fluorescent agents may be added to ep-
is simplified, as the specimens are easy to hold. oxies for the study of porous specimens such as

15
Mounting of Specimens

Cast epoxy resins provide better edge reten-

tion than cast acrylic resins, mainly due to the
better adhesion of epoxy to the specimen and
their lower shrinkage. Acrylics usually do not
bond to the specimen and, because they shrink
more during curing, a gap is inevitability formed
between specimen and mount. When a shrink-
age gap is present, edge retention is usually
poor. To improve edge retention with castable
mounts, the specimen can be plated, for example,
with electroless nickel using the EdgeMet kit, or
Figure 8. Vacuum Impregnation Equipment Flat Edge Filler particles can be added to the resin.
To obtain electrical conductivity, Conductive Filler
thermal spray coated specimens. Most epoxies
particles can be added to the castable resin.
are cured at room temperature, and curing times
can vary from 2 to 8 hours. Some can be cured at
When preparing castable resin mounts, particu-
slightly elevated temperatures in less time, as long
larly epoxy mounts by manual (“hand”) methods,
as the higher temperature does not adversely
the metallographer will observe that the surface
affect the specimen.Table 4 lists the characteristics
tension between the mount and the working
of Buehler’s castable resins.
Table 4. Characteristics of Buehler’s Castable Resins

Peak Shore D Cure Recommended

Name Type Temperature Hardness* Time Mold Comments
EpoThin Epoxy 80 °F 78 Best Edge 9 Any Low viscosity, low shrinkage,
(27 °C) Retention hours transparent, best for vacuum
impregnation
EpoxiCure Epoxy 82 °F 82 Best Edge 6 Any Moderate hardness, transparent,
(28 °C) Retention hours low shrinkage
EpoxiCure NC Epoxy 82 °F 82 Excellent Edge 6 Any General purpose epoxy; viscosity
(28 °C) Retention hours 400-600 cps at 77 °F, (25 °C) good
adherence to specimen; can be used
to vacuum impregnate voids
(viscosity can be reduced by
warming to 122 ° F (50 °C)); good
for heat-sensitive specimens (very
low exotherm during polymerization)
EpoKwick Epoxy 185 °F 82 Good 90 Sampl-Kup Faster epoxy, some shrinkage,
(85 °C) Edge Retention minutes transparent
EpoColor Epoxy 175 °F 82 Good Edge 90 Sampl-Kup Dye-enhanced epoxy, displays red
(79 °C) Retention minutes under darkfield and polarized light
EpoHeat Epoxy 295 °F 82 Best Edge 90 Any Fast, heat cure epoxy system with
(Transparent) (79 °C) Retention minutes low shrinkage and low viscosity (32
cps at 176 °F (80 °C)). Excellent for
impregnation. Requires oven cure.
Golden color.
VariDur 3000 Acrylic 252 °F 90 15-30 Any Minimal shrinkage, good edge
(122 °C) minutes retention
VariDur Acrylic 170 °F 85 Fair Edge 10 Any Fast cure, harder acrylic opaque,
(77 °C) Retention minutes abrasive resistant
SamplKwick Acrylic 175 °F 80 Poor 5-8 Any Very fast cure, translucent,
(79 °C) Edge Retention minutes some shrinkage
VariKleer Acrylic 212 °F 84 10-15 Sampl-Kup Designed for use with a pressure
(Transparent) (100 °C) minutes vessel, this general use acrylic
hardens crystal clear with minimal
shrinkage. To achieve crystal clear
samples, reusable silicon or EPDM
molds should not be used.
* Hardness differences appear negligible but abrasion resistance has a significant effect on edge rounding

16
Mounting of Specimens

surface is much higher than with a compression Edge Preservation

mount. This can make holding the mount more Edge preservation is a classic metallographic
challenging. If automated devices are used, the problem and many “tricks” have been promoted
metallographer may hear “chatter” (noise) during (most pertaining to mounting, but some to grind-
rough grinding due to the greater surface ten- ing and polishing) to enhance edge flatness. These
sion. The chatter can be reduced or stopped by methods [2] include the use of backup material
changing to contra mode (head and platen rotate in the mount, the application of coatings to the
in opposite directions.) surfaces before mounting or the addition of a
filler material to the mounting resin. Plating [2] of
Acrylics do generate considerable heat during a compatible metal on the surface to be protected
curing and this can be strongly influenced by the (electroless nickel, deposited using the EdgeMet
molding technique used. Nelson [4] measured the Kit has been widely used) is generally considered
exotherm produced by polymerizing an acrylic to be the most effective procedure. However, im-
resin using two procedures: a glass mold on a age contrast at an interface between a specimen
glass plate (insulative) and an aluminum mold on and the electroless nickel may be inadequate for
an aluminum plate (conductive). Polymerization certain evaluations. Figure 9 shows the surface of
produced a maximum exotherm of 132°C using a specimen of 1215 free-machining steel that was
the insulative approach but only 42°C using the salt bath nitrided. One specimen was plated with
conductive approach. Note that 132°C is not electroless nickel; both were mounted in EpoMet
much less than the 150°C temperature used in G. It is hard to tell where the nitrided layer stops
hot compression mounting! Nelson also mea- for the plated specimen, Figure 9a, which exhibits
sured the exotherm produced when an epoxy poor image contrast between the nickel and the
resin was cured in a phenolic ring form placed on nitrided surface. This is not a problem for the non-
a pasteboard base. Although this was an insula- plated specimen, Figure 9b.
tive approach, the maximum temperature during
polymerization was only 7 °C, a vast improvement
over the acrylics.

Nelson’s work applies to specific acrylic and

epoxy resins molded upon specific conditions.
While the epoxy that he used exhibited a low
exotherm, this does not imply that all epoxy
resins will exhibit such low exotherms in polym-
a
erization. Epoxy resins that cure in short time
periods develop much higher exotherms, that
can exceed that of acrylic resins. In addition to
the speed of curing of the epoxy resin, other fac-
tors do influence the magnitude of the exotherm
during polymerization. The larger the mass of
epoxy in the mount, the faster it will set and the
greater the exotherm. Indeed, very large mounts
can generate enough heat to crack extensively.
Heating the resin makes it less viscous and b
speeds up curing, also generating more heat Figure 9. Example of (a, top) poor coating edge
during polymerization. The mold material also visibility due to a lack of contrast between the
protective nickel plating and the salt bath nitrided
can influence curing time and temperature. surface (arrow) of 1215 free-machining carbon steel;
and, (b, bottom) good contrast and visibility between
For example, EpoxiCure cures fastest in EpoMet resin and nitrided surface (arrow) plus
SamplKup plastic molds, slower in phenolic ring excellent edge retention (1000X, 2% nital).
forms, and still slower in the reuseable rubber Introduction of new technology has greatly
mounting cups. Consequently, the exotherm will reduced edge preservation problems [5,6]. Mount-
be greater when using the SamplKup type mold ing presses that cool the specimen to near ambient
and lowest when using the rubber mounting temperature under pressure produce much tighter
cups. All of these factors must be considered if the mounts. Gaps that form between specimen and
exotherm must be minimized. 17
Mounting of Specimens

resin are a major contributor to edge rounding, as

shown in Figure 5. Staining at shrinkage gaps may
also be a problem, as demonstrated in Figure 10.
Use of semi-automatic and automatic grinding/
polishing equipment, rather than manual (hand)
preparation, increases surface flatness and edge
retention. To achieve the best results, however, the
position of the specimen holder, relative to the
platen, must be adjusted so that the outer edge a
of the specimen holder rotates out over the edge
of the surface on the platen during grinding and
polishing, particularly for 8-inch diameter (200
mm) platens. The use of harder, woven or non-wo-
ven, napless surfaces for polishing with diamond
abrasives (rather than softer cloths such as canvas,
billiard and felt) maintains flatness. Final polishing
with low nap cloths for short times introduces
very little rounding compared to use of higher
b
nap, softer cloths.
Figure 11. Poor edge retention (arrows) at the salt
bath nitrided surface of 1215 free-machining carbon
steel mounted in (a, top) phenolic resin and in (b,
bottom) methyl methacrylate and polished in the
same holder as those shown in Figure 9 (1000X, 2%
nital).

Figure 10. Staining (thin arrows) due to etchant bleed

out from a shrinkage gap (wide arrows) between
the phenolic mount and the M2 high speed steel
specimen (500X, Vilella’s reagent).
a
These procedures will produce better edge
retention with all thermosetting and thermoplas-
tic mounting materials. Nevertheless, there are still
differences among the polymeric materials used
for mounting. Thermosetting resins provide better
edge retention than thermoplastic resins. Of the
thermosetting resins, diallyl phthalate provides
little improvement over the much less expensive
phenolic compounds.The best results are obtained b
with EpoMet G an epoxy based thermosetting
Figure 12. Excellent edge retention for (a, top)
resin that contains a filler material. For comparison, complex coated (arrow) sintered carbide insert
Figure 11 shows micrographs of the nitrided 1215 (1000X, Murakami’s reagent) and for (b, bottom) iron
nitrided (arrow points to a white iron nitride layer; an
specimen mounted in a phenolic resin (Figure 11a), embedded shot blasting particle is to the left of the
and in methyl methacrylate (Figure 11b), at 1000X. arrow) H13 hot work die steel specimen (1000X, 2%
nital). EpoMet was used in both cases.
These specimens were prepared in the same speci-
men holder as those shown in Figure 9, but neither Very fine aluminum oxide spheres have been
displays acceptable edge retention at 1000X. Fig- added to epoxy mounts to improve edge reten-
ure 12 shows examples of perfect edge retention, tion, but this is really not a satisfactory solution
as also demonstrated in Figure 9. as the particles are extremely hard (~2000 HV)
and their grinding/polishing characteristics are
incompatible with softer metals placed inside
18
Mounting of Specimens

the mount. Recently, a soft ceramic shot (~775 • Automated grinding/polishing equipment
HV) has been introduced that has grinding/pol- produces flatter specimens than manual “hand”
ishing characteristics compatible with metallic preparation.
specimens placed in the mount. Figure 13 shows
• Use the central force mode (defined later in the
an example of edge retention with the Flat Edge
text) with an automated grinder/polisher as this
Filler soft ceramic shot in an epoxy mount.
method provides better flatness than individual
pressure mode (defined later in the text).
Following are general guidelines for obtaining the
best possible edge retention. All of these factors • Orient the position of the smaller diameter
contribute to the overall success, although some specimen holder so that, as it rotates, its periph-
are more critical than others. ery slightly overlaps the periphery of the larger
diameter platen.
• Use PSA-backed SiC grinding paper (if SiC is
used), rather than water on the platen and a
peripheral hold-down ring, and PSA-backed
polishing cloths rather than stretched cloths.
• UltraPrep metal-bonded or resin-bonded
diamond grinding discs produce excellent flat
surfaces for a wide variety of materials.
Figure 13. Flat Edge Filler shot was added to • Use “hard” napless surfaces for rough polishing
Epoxicure resin to improve the edge retention of this
annealed H13 hot work die steel specimen (500X, 4% (until the final polishing step), such as TexMet
picral). 1500, UltraPol or UltraPad cloths, and fine pol-
ishing, such as a TriDent cloth. Use a napless, or
• Properly mounted specimens yield better edge a low- to medium-nap cloth, depending upon
retention than unmounted specimens, as round- the material being prepared, for the final step
ing is difficult, if not impossible, to prevent at a and keep the polishing time brief.
free edge. Hot compression mounts yield better
edge preservation than castable resins. • Rigid grinding discs, such as the ApexHercules
H and S discs, produce excellent flatness and
• Electrolytic or electroless Ni plating (e.g., with edge retention and should be used whenever
the EdgeMet Kit) of the surface of interest possible.
provides excellent edge retention. If the
compression mount is cooled too quickly after
Helpful Hints For
polymerization, the plating may be pulled
Mounting
away from the specimen leaving a gap. When
this happens, the plating is ineffective for edge Epoxy is the only resin that will
retention. physically adhere to a specimen.
If its viscosity is low, epoxy can be drawn into
• Thermoplastic compression mounting materials
pores and cracks by vacuum impregnation.
are less effective than thermosetting resins. The
Acrylics are too viscous for vacuum impreg-
best thermosetting resin for edge retention is
nation.
EpoMet G, an epoxy-based resin containing a
hard filler material. Castable resins are sensitive to shelf life,
• Never hot eject a thermosetting resin after which can be extended by keeping them in a
polymerization and cool it quickly to ambient refrigerator. It is a good practice to date your
(e.g., by cooling it in water) as a gap will form containers when you get them.
between specimen and mount due to the
differences in thermal contraction rates. Fully
automated mounting presses cool the mounted
specimen to near ambient temperature under
pressure and this greatly minimizes gap forma-
tion due to shrinkage.
19
Grinding

GRINDING Other materials have also been used both for

the planar grinding stage or, afterwards, to
Grinding should commence with the finest grit replace SiC paper. For very hard materials
size that will establish an initially flat surface and such as ceramics and sintered carbides, one, or
remove the effects of sectioning within a few more, metal-bonded or resin-bonded diamond
minutes. An abrasive grit size of 180 to 240 (P180 disks (the traditional type) with grit sizes from
to P280) is coarse enough to use on specimen about 70- to 9-µm can be used. The traditional
surfaces sectioned by an abrasive cut-off wheel. metal- or resin-bonded diamond disc has dia-
Hack-sawed, bandsawed, or other rough surfaces mond spread uniformly over its entire surface.
usually require abrasive grit sizes in the range of An alternate type of disc, the UltraPrep disc, has
120- to 180-grit. The abrasive used for each suc- diamond particles applied in small spots to the
ceeding grinding operation should be one or two disk surface, so that surface tension is lessened.
grit sizes smaller than that used in the preced- UltraPrep metal-bonded discs are available in six
ing step. A satisfactory fine grinding sequence diamond sizes from 125- to 6-µm while UltraPrep
might involve SiC papers with grit sizes of 220- or resin bonded discs are available in three diamond
240-, 320-, 400-, and 600-grit (P240 or P280, P400, sizes from 30- to 3-µm. Another approach uses
P600 and P1200). This sequence is used in the a stainless steel woven mesh UltraPlan cloth on
“traditional” approach. a platen charged with coarse diamond, usually
in slurry form, for planar grinding. Once planar
As with abrasive-wheel sectioning, all grinding
surfaces have been obtained, there are several
steps should be performed wet provided that
single-step procedures available for avoiding the
water has no adverse effects on any constituents
finer SiC papers. These include the use of platens,
of the microstructure. Wet grinding minimizes
thick woven polyester cloths, silk, or rigid grinding
specimen heating, and prevents the abrasive from
disks. With each of these, an intermediate diamond
becoming loaded with metal removed from the
size, generally 9- to 3-µm, is used.
specimen being prepared.

Each grinding step, while producing damage Grinding Media

itself, must remove the damage from the previ- The grinding abrasives commonly used in the
ous step. The depth of damage decreases with preparation of metallographic specimens are
the abrasive size but so does the metal removal silicon carbide (SiC), aluminum oxide (Al 2O 3),
rate. For a given abrasive size, the depth of emery (Al2O3 - Fe3O4), composite ceramics and
damage introduced is greater for soft materials diamond. Emery paper is rarely used today in
than for hard materials. metallography due to its low cutting efficiency.
SiC is more readily available as waterproof
For automated preparation using a multiple- paper than aluminum oxide. Alumina papers, such
specimen holder, the intital step is called planar as PlanarMet Al 120-grit paper, do have a better
grinding. This step must remove the damage cutting rate than SiC for some metals [3]. These
from sectioning while establishing a common abrasives are bonded to paper, polymeric or cloth
plane for all of the specimens in the holder, so backing materials of various weights in the form
that each specimen is affected equally in subse- of sheets, discs and belts of various sizes. Limited
quent steps. Silicon carbide and alumina abrasive use is made of standard grinding wheels with
papers are commonly used for the planar grinding abrasives embedded in a bonding material. The
step and are very effective. Besides these papers, abrasives may be used also in powder form by
there are a number of other options available. charging the grinding surfaces with the abrasive
One option is to planar grind the specimens in a premixed slurry or suspension. SiC particles,
with a conventional alumina grinding stone. This particularly with the finer size papers, embed
requires a special purpose machine, as the stone readily when grinding soft metals, such as Pb, Sn,
must rotate at a high speed, ≥1500 rpm, to cut Cd and Bi (see Figure 14). Embedding of diamond
effectively. The stone must be dressed regularly abrasive is also a problem with these soft metals
with a diamond tool to maintain flatness, damage and with aluminum, but mainly with slurries when
depth is high and embedding of alumina abrasive napless cloths are used, see Figure 15.
in specimens can be a problem.
20
Grinding

Figure 14. SiC grit particle (arrow) embedded in a Figure 15. 6µm diamond particles (arrows) embedded
6061-T6 aluminum weldment (500X, aqueous 0.5% in lead (1000X).
HF).
CarbiMet SiC paper is made in the USA method for very fine grit sizes). The grit size
according to the ANSI/CAMI standard (B74. numbering systems differ above 180- (P180) grit,
18-1996) while SiC papers manufactured in but equivalent sizes can be determined using
Europe are made according to the FEPA Table 5. As with many standards, they are not
standard (43-GB-1984, R 1993). Both standards mandatory and manufacturers can, and do, make
use the same methods for sizing the abrasives some of their papers to different mean particle
and the same standards to calibrate these sizes than defined in these specifications. There is a
devices (sieving for the coarsest grits, sedimen- philosophical difference in the two systems. ANSI/
tation grading for intermediate grits (240-600 CAMI papers use a wider particle size distribution
or P280-P1200), and the electrical resistance (centered on the mean size) than FEPA papers. A

Table 5. European/USA Equivalency Grit Guide

European/USA Equivalency Grit Guide

FEPA (Europe) ANSI/CAMI (USA)
Grit Number Size(µm) Grit Number Size(µm) Emery Grit
P60 269.0 60 268.0
P80 201.0 80 188.0
P100 162.0 100 148.0
P120 127.0 120 116.0
P180 78.0 180 78.0 3
P240 58.5 220 66.0 2
P280 52.2 240 51.8
P320 46.2
P360 40.5 280 42.3 1
P400 35.0 320 34.3 0
P500 30.2 360 27.3
P600 25.8 400 22.1 00
P800 21.8
P1000 18.3 500 18.2 000
P1200 15.3 600 14.5
P1500 12.6 800 12.2 0000
P2000 10.3 1000 9.2
P2500 8.4 1200 6.5
P4000* 5.0*
The chart shows the midpoints for the size ranges for ANSI/CAMI graded paper according to ANSI standard B74.18-1996 and for FEPA graded paper according to
FEPA standard 43-GB-1984 (R1993). The ANSI/CAMI standard lists SiC particles sizes ranges up to 600 grit paper. For finer grit ANSI/CAMI papers, the particles sizes
come from the CAMI booklet, Coated Abrasive (1996).
*FEPA grades finer than P2500 are not standardized and are graded at the discretion of the manufacturer. In practice, the above standard values are only
guidelines and individual manufactures may work to a different size range and mean value.

21
Grinding

broader size range allows cutting to begin faster

at lower pressures than with a narrower size range,
so less heat is generated and less damage results.
However, the broader size range does produce a
wider range of scratch depths; but, these should
be removed by the next step in the preparation
sequence. Generation of less damage to the struc-
ture is considered to be more important than the
surface finish after a particular grinding step, as it
is the residual damage in the specimen that may
prevent us from seeing the true microstructure at
the end of the preparation sequence.

Figure 17. SuperMet grinder

Grinding Equipment
platens, Figure 18, upon which the SiC paper is
Stationary grinding papers, often used by students, attached.
but uncommon in industrial use, are supplied in
strips or rolls, such as for use with the HandiMet Lapping is an abrasive technique in which the
2 roll grinder. The specimen is rubbed against the abrasive particles roll freely on the surface of a
paper from top to bottom. Grinding in one direc- carrier disc. During the lapping process, the disc
tion is usually better for maintaining flatness than is charged with small amounts of a hard abrasive
grinding in both directions. This procedure can be such as diamond or silicon carbide. Lapping disks
done dry for certain delicate materials, but water is can be made of many different materials; cast iron
usually added to keep the specimen surface cool and plastic are used most commonly. Lapping
and to carry away the grinding debris. produces a flatter specimen surface than grinding,
but it does not remove metal in the same man-
Belt grinders, such as the SurfMet I or the Du- ner as grinding. Some platens, referred to as laps,
oMet II belt grinders, Figure 16, are usually present are charged with diamond slurries. Initially the
in most laboratories. An alternative approach is diamond particles roll over the lap surface (just
to use a high speed disc grinder, such as the Su- as with other grinding surfaces), but they soon
perMet grinder, Figure 17. These devices are used become embedded and cut the surface produc-
with coarse abrasive papers from 60- to 240-grit, ing microchips.
and are mainly used for removing burrs from
sectioning, for rounding edges that need not
be preserved for examination, for flattening cut
surfaces to be macro-etched, or for removing
sectioning damage. Manual grinding work is per-
formed on rotating “wheels”; that is, motor-driven

Figure 18. Beta grinder/polisher

Figure 16. DuoMet II belt grinder

22
Polishing

POLISHING
Polishing is the final step, or steps, in producing a
deformation-free surface that is flat, scratch free,
and mirror-like in appearance. Such a surface is
necessary to observe the true microstructure
for subsequent metallographic interpretation,
both qualitative and quantitative. The polishing
technique used should not introduce extraneous
structures such as disturbed metal (Figure 19), pit-
Figure 22. Staining (arrow) on the surface of a of Ti-
ting (Figure 20), dragging out of inclusions,“comet 6% Al-2% Sn-4% Zr-2% Mo prepared specimen (200X).

materials, such as through hardened steels, ceram-

ics and cemented carbides, an additional coarse
polishing step may be required. The initial coarse
polishing step may be followed by polishing with
1-µm diamond on a napless, low nap, or medium
nap cloth. A compatible lubricant should be used
sparingly to prevent overheating or deformation
of the surface. Intermediate polishing should be
performed thoroughly so that final polishing
may be of minimal duration. Manual, or “hand”
polishing, is usually conducted using a rotating
wheel where the operator rotates the specimen
in a circular path counter to the wheel rotation
direction.

Mechanical Polishing
The term “mechanical polishing” is frequently
Figure 19. (top) Preparation damage (arrows) in used to describe the various polishing procedures
annealed CP titanium (500X, DIC, Kroll’s reagent). involving the use of fine abrasives on cloth. The
Figure 20. (bottom) Pitting (arrow) on the surface of a
cold-drawn brass (Cu-20% Zn) specimen (100X)

tailing” (Figure 21), staining (Figure 22) or relief

(height differences between different constituents,
or between holes and constituents (Figure 23
and 24). Polishing usually is conducted in several
stages. Traditionally, coarse polishing generally
was conducted with 6- or 3-µm diamond abrasives
charged onto napless or low-nap cloths. For hard a

b
Figure 23. Examples of freedom from relief (a, top)
Figure 21. “Comet tails” at large nitrides in an
and minor relief (b, bottom) at the edges of large
annealed H13 hot work die steel specimen (200X,
primary hypereutectic silicon particles in an as-cast
DIC).
Al-19.85% Si specimen (200X, aqueous 0.5% HF).

23
Polishing

be wavy rather than flat and focusing may be

difficult at high magnifications. Further, elec-
tropolishing tends to round edges associated with
external surfaces, cracks or pores. In two-phase
alloys, one phase will polish at a different rate than
another, leading to excessive relief. In some cases,
one phase may be attacked preferentially and
inclusions are usually attacked. Consequently, elec-
a trolytic polishing is not recommended for failure
analysis or image analysis work, except possibly as
a very brief step at the end of a mechanical polish-
ing cycle to remove whatever minor damage may
persist. Electropolishing has been most successful
with soft single-phase metals and alloys, particu-
larly where polarized light response needs to be
maximized.

b Manual “Hand” Polishing

Figure 24. Differences in relief control in a brazed Aside from the use of improved polishing cloths
specimen containing shrinkage cavities: a) (top)
poor control; and, b) (bottom) good control (100X, and abrasives, hand-polishing techniques still
glyceregia). follow the basic practice established many
years ago:
cloth may be attached to a rotating wheel or a
vibratory polisher bowl. Historically, cloths have 1. Specimen Movement. The specimen is
been either stretched over the wheel and held held with one or both hands, depending on the
in place with an adjustable clamp on the platen operator’s preference, and is rotated in a direction
periphery, or held in place with a pressure sensi- counter to the rotation of the polishing wheel. In
tive adhesive (PSA) bonded to the back of the addition, the specimen is continually moved back
cloth. If a stretched cloth moves under the applied and forth between the center and the edge of the
pressure during polishing, cutting will be less ef- wheel, thereby ensuring even distribution of the
fective. If an automated polishing head is used, abrasive and uniform wear of the polishing cloth.
stretched cloths are more likely to rip, especially (Some metallographers use a small wrist rotation
if unmounted specimens are being prepared. In while moving the specimen from the center to the
mechanical polishing, the specimens are held by edge of one side of the wheel.) After each step, the
hand, held mechanically in a fixture, or merely specimen is rotated 45 to 90° so that the abrasion
confined within the polishing area, as with the is not unidirectional.
VibroMet 2 polisher.
2. Polishing Pressure. The correct amount of
applied pressure must be determined by experi-
Electrolytic Polishing ence. In general, a firm hand pressure is applied
Electrolytic polishing can be used to prepare to the specimen.
specimens with deformation free surfaces. The
3. Washing and Drying. The specimen is
technique offers reproducibility and speed. In
washed by swabbing with a liquid detergent
most cases, the published instructions for elec-
solution, rinsed in warm running water, then with
trolytes tell the user to grind the surface to a
ethanol, and dried in a stream of warm air. Alcohol
600-grit finish and then electropolish for about
usually can be used for washing when the abrasive
1- to 2-minutes. However, the depth of damage
carrier is not soluble in water or if the specimen
after a 600- (P1200) grit finish may be several
cannot tolerate water. Ultrasonic cleaning may be
micrometres but most electropolishing solutions
needed if the specimens are porous or cracked.
remove only about 1-µm per minute. In this case,
the deformation will not be completely removed.
In general, electropolished surfaces tend to

24
Polishing

4. Cleanness. The precautions for cleanness, as ness and edge retention.There are two approaches
previously mentioned, must be strictly observed for handling specimens. Central force utilizes a
to avoid contamination problems. This involves specimen holder with each specimen held in place
the specimen, the metallographer’s hands, and rigidly. The holder is pressed downward against
the equipment. the preparation surface with the force applied to
the entire holder. Central force yields the best edge
retention and specimen flatness. If the results after
Automated Polishing
etching are inadequate, the specimens must be
Mechanical polishing can be automated to a high placed back in the holder and the entire prepara-
degree using a wide variety of devices ranging tion sequence must be repeated. Instead of doing
from relatively simple systems, Figure 25, to rather this, most metallographers will repeat the final
sophisticated, minicomputer, or microprocessor step manually and then re-etch the specimen.
controlled devices, Figure 26. Units also vary in
capacity from a single specimen to a half dozen or The second method utilizes a specimen holder
more at a time and can be used for all grinding and where the specimens are held in place loosely.
Force is applied to each specimen by a piston,
hence the term “individual force” for this approach.
This method provides convenience in examining
individual specimens during the preparation cycle,
without the problem of regaining planarity for all
specimens in the holder on the next step. Also, if
the etch results are deemed inadequate, the speci-
men can be replaced in the holder to repeat the
last step, as planarity is achieved individually rather
than collectively. The drawback to this method is
that slight rocking of the specimen may occur, es-
pecially if the specimen height is too great, which
degrades edge retention and flatness.

Figure 25. Beta & Vector grinder/polisher and power

head
Polishing Cloths
The requirements of a good polishing cloth
polishing steps.These devices enable the operator include the ability to hold the abrasive media,
to prepare a large number of specimens per day long life, absence of any foreign material that
with a higher degree of quality than hand polish- may cause scratches, and absence of any pro-
ing and at reduced consumable costs. Automatic cessing chemical (such as dye or sizing) that
polishing devices produce the best surface flat- may react with the specimen. Many cloths of
different fabrics, weaves, or naps are available
for metallographic polishing. Napless or low nap
cloths are recommended for coarse polishing
with diamond abrasive compounds. Napless,
low, medium, and occasionally high nap cloths
are used for final polishing. This step should be
brief to minimize relief. Table 6 lists Buehler’s
line of polishing cloths, their characteristics and
applications.

Polishing Abrasives
Polishing usually involves the use of one or more
of the following abrasives: diamond, aluminum
Figure 26. BuehlerVanguard 2000 fully automatic
specimen preparation system

25
Polishing

Table 6. Cloth Selection Guide

Recommended
Premium µm Size Abrasive Cloth Usage
Cloths Range Types Characteristics Guide Applications
UltraPad 15 - 3 Diamond Hard Woven, Used to Replace Ferrous
No Nap with High Multiple SiC Materials and
Material Removal Grinding Steps Thermal
Spray Coatings
UltraPol 15 - 1 Diamond Hard Woven, Excellent Surface Minerals, Coal,
No Nap Finish Used to Retain and Ceramics,
Flatness in Medium Inclusion Retention
to Hard Specimens in Steels, and
Refractory Metals
Nylon 15 - 1 Diamond Medium-Hard Used to Retain Ferrous Materials,
Woven, No Nap Flatness and Sintered Carbides
Hard Phases and Cast Irons
TexMet P 15 - 3 Diamond Hardest Perforated Used for Material Ceramics, Carbides,
Non-woven Pad Removal and Petrographic, Hard
for High Material Flatness of Metals, Glass, Metal
Removal Hard Specimens Matrix Composites
TexMet 2500 15 - 3 Diamond, Hard Non-woven Used for Harder Hard Ferrous
Al2O3 Pad used for Specimens and Materials
Durability Increased Flatness and Ceramics
TexMet 1500 9 - 0.02 Diamond, Soft Most Popular Cloth Ferrous and
Al2O3, SiO2 Non-woven for Intermediate Nonferrous Metals,
Pad for Steps, Good for Ceramics, Electronic
Surface Finish Most Materials Packages, PCB’s,
Thermal Spray
Coatings, Cast Irons,
Cermets, Minerals,
Composites, Plastics
TriDent 9 - 1 Diamond Softer, Used to Maximize Ferrous and
Durable, Flatness and Nonferrous Metals,
Woven Synthetic, Retain Phases while Microelectronics,
No Nap Providing Excellent Coatings
Surface Finish
VerduTex 15 - 1 Diamond Durable, Medium Used for Rough Ferrous and Nonferrous
Hard Synthetic and Final Polishing Composite Material
Silk and Plastic
(CFK, GFK)
VelTex 9 - 1 Diamond Short-napped, Ideal for Steel. Nonferrous and
Synthetic Final Polishing Soft Metals
Velvet
WhiteFelt 6 - 0.02 Diamond, Soft and General Usage Ferrous and
Al2O3, SiO2 Durable Matted for Intermediate Nonferrous Materials
Wool Cloth to Fine Steps
MicroCloth 5 - 0.02 Diamond, Synthetic Cloth Most Popular Ferrous and
Al2O3, SiO2 with Medium Nap General Nonferrous
Usage Final Metals, Ceramics,
Polishing Cloth Composites, PCB’s,
Cast Irons, Cermets,
Plastics, Electronics
MasterTex 1 - 0.05 Al2O3, SiO2 Soft Synthetic Softer Final Soft Nonferrous and
Velvet with Polishing Cloth Microelectronic
Low Nap Packages
ChemoMet 1 - 0.02 Al2O3, SiO2 Soft Synthetic General Usage Titanium, Stainless
Pad, Micro-nap Pad that Removes Steels, Lead/Tin
Smear Metal Solders, Electronic
from Tough Packages, Soft
Materials during Nonferrous
Chemomechanical Metals, Plastics
Polishing

26
Polishing

oxide (Al 2O 3), and amorphous silicon dioxide first of the monocrystalline form, similar in
(SiO2) in collidal suspension. For certain materi- morphology to natural diamond, and then in
als, cerium oxide, chromium oxide, magnesium polycrystalline form. MetaDi II diamond pastes use
oxide or iron oxide may be used, although these synthetic monocrystalline diamond while MetaDi
are used infrequently. With the exception of dia- Supreme suspensions and paste use synthetic
mond, these abrasives are normally suspended in polycrystalline diamonds. Figure 27 shows the
distilled-water, but if the metal to be polished is shape differences between monocrystalline and
not compatible with water, other suspensions such polycrystalline diamonds. Studies have shown that
as ethylene glycol, alcohol, kerosene or glycerol, cutting rates are higher for many materials using
may be required. The diamond abrasive should polycrystalline diamond compared to monocrys-
be extended only with the carrier recommended talline diamond.
by the manufacturer. Besides the water-based
diamond suspension, Buehler also offers oil-based Colloidal silica was first used for polishing wafers
suspensions that are needed to prepare materials of single crystal silicon where all of the damage
sensitive to water. on the wafer surface must be eliminated before
a device can be grown on it. The silica is amor-
Diamond abrasives were introduced to metallog- phous and the solution has a basic pH of about
raphers commercially in the late 1940s. However, 9.5. The silica particles are actually nearly spherical
their use in metallography goes back to at least
the late 1920s, as Hoyt [7] mentions a visit to
the Carboloy plant in West Lynn, Massachusetts,
where he saw sapphire bearings being polished
with diamond dust in an oil carrier. He used some
of this material to prepare sintered carbides and
published this work in 1930. Diamond abrasives
were first introduced in a carrier paste but later
aerosol and slurry forms were introduced. Virgin
natural diamond was used initially, and is still Figure 28. Amorphous silica particles in colloidal
available in MetaDi diamond paste and suspen- silica (TEM, 300,000X).
sions. Later, synthetic diamond was introduced,
in shape, Figure 28, the polishing action is slow,
and is due to both chemical and mechanical
action. Damage-free surfaces can be produced
more easily when using colloidal silica than with
other abrasives (for final polishing). Etchants can
respond differently to surfaces polished with col-
loidal silica. For example, an etchant that produces
a grain contrast etch when polished with alumina
may instead reveal the grain and twin boundaries
a
with a “flat” etch when polished with colloidal silica.
Color etchants frequently respond better when
colloidal silica is used producing a more pleasing
range of colors and a crisper image. But, clean-
ing of the specimen is more difficult. For manual
work, use a tuft of cotton soaked in a detergent
solution. For automated systems, stop adding
abrasive about 10-15 seconds before the cycle
ends and. For the last 10 seconds, flush the cloth
b surface with running water. Then, cleaning is sim-
Figure 27. Comparison of monocrystalline (a, top) and pler. Amorphous silica will crystallize if allowed to
polycrystalline (b, bottom) synthetic diamond grain evaporate. Crystalline silica will scratch specimens,
shapes (SEM, 450X).

27
Polishing

so this must be avoided. When opening a bottle,

clean off any crystallized particles than may have
formed around the opening. The safest approach
is to filter the suspension before use. Additives are
used to minimize crystallization, as in MasterMet 2
Colloidal Silica, greatly retarding crystallization.

For routine examinations, a fine diamond

abrasive, such as 1-µm, may be adequate as the
last preparation step. Traditionally, aqueous fine
alumina powders and slurries, such as the Mi-
croPolish II deagglomerated alumina powders and
suspensions, have been used for final polishing Figure 29. VibroMet 2 vibratory polisher
with medium nap cloths. Alpha alumina (0.3-µm
size) and gamma alumina (0.05-µm size) slurries (or
suspensions) are popular for final polishing, either Helpful Hints For
in sequence or singularly. MasterPrep alumina Grinding/Polishing
suspension utilizes alumina made by the sol-gel Specimens that contain cracks
process, and it produces better surface finishes or pores that are not filled with
than alumina abrasives made by the traditional epoxy may require ultrasonic cleaning to re-
calcination process. Calcined alumina abrasives move abrasive and debris from the openings
always exhibit some degree of agglomeration, to avoid contaminating the next step.
regardless of the efforts to deagglomerate them,
while sol-gel alumina is free of this problem. Excessive ultrasonic cleaning vibrations can
MasterMet colloidal silica suspensions (basic pH damage the structure of certain soft metals
around 9.5) are newer final polishing abrasives and alloys, particularly precious metals.
that produce a combination of mechanical and
To remove a worn out tightly adhering cloth,
chemical action which is particularly beneficial
soak the platen under hot water for a few min-
for difficult to prepare materials. Vibratory polish-
utes.This loosens the glue bond and allows you
ers, Figure 29, are often used for final polishing,
to pull it off easier.
particularly with more difficult to prepare materi-
als, for image analysis studies, or for publication
quality work.

28
Examples of Preparation Procedures

EXAMPLES OF PREPARATION be omitted. Times and pressures could be

PROCEDURES varied, as well, to suit the needs of the work,
or the material being prepared. This is the “art”
The “Traditional” Method of metallography.
Over the past forty years, a general procedure has
been developed that is quite successful for prepar- Contemporary Methods
ing most metals and alloys. This method is based New concepts and new preparation materials have
on grinding with silicon carbide waterproof papers been introduced that enable metallographers to
through a series of grits, then rough polishing with shorten the process while producing better, more
one or more diamond abrasive sizes, followed by consistent results. Much of this effort has centered
fine polishing with one or more alumina suspen- upon reducing or eliminating the use of silicon
sions of different particle size. This procedure will carbide paper in the grinding steps. In all cases, an
be called the “traditional” method, and is described initial grinding step must be used, but there are a
in Table 7. wide range of materials that can be chosen instead
of SiC paper.There is nothing wrong with the use of
This procedure is used for manual or automated SiC for the first step, except that it has a short life. If
preparation, although manual control of the an automated device is used that holds a number
force applied to a specimen would not be as of specimens rigidly (central force), then the first
consistent. Complementary motion means that step must remove the sectioning damage on each
the specimen holder is rotated in the same specimen and bring all of the specimens in the
direction as the platen, and does not apply to holder to a common plane. This first step is often
manual preparation. Some machines can be called “planar grinding.” SiC paper can be used for
set so that the specimen holder rotates in the this step, although more than one sheet may be
direction opposite to that of the platen, called needed. Alternatively, the metallographer could
“contra.” This provides a more aggressive action use PlanarMet Al alumina paper, conventional
but was not part of the “traditional” approach metal or resin bonded diamond discs or the
when automated. The traditional method is not UltraPrep diamond discs, the UltraPlan stainless
rigid, as other polishing cloths may be substituted steel mesh cloth (diamond is applied during use),
and one or more of the polishing steps might the ApexHercules H and S rigid grinding discs

Table 7. The Traditional Procedure for Preparing Most Metals and Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 120- (P120) grit 6 (27) 240-300 Until
abrasive discs SiC water Comp. Plane
(waterproof paper) cooled
CarbiMet 220- or 240- (P240 or P280) 6 (27) 240-300 1:00
abrasive discs grit SiC water cooled Comp.
CarbiMet 320- (P400) grit 6 (27) 240-300 1:00
abrasive discs SiC water cooled Comp.
CarbiMet 400- (P600) grit 6 (27) 240-300 1:00
abrasive discs SiC water cooled Comp.
CarbiMet 600- (P1200) grit 6 (27) 240-300 1:00
abrasive discs SiC water cooled Comp.
Canvas 6-µm MetaDi 6 (27) 120-150 2:00
cloth diamond paste with Comp.
MetaDi fluid extender
Billiard or 1-µm diamond paste with 6 (27) 120-150 2:00
Felt cloths MetaDi fluid extender Comp.
MicroCloth Aqueous 0.3-µm 6 (27) 120-150 2:00
cloth α − alumina Comp.
MicroPolish slurry
MicroCloth Aqueous 0.05-µm 6 (27) 120-150 2:00
cloth γ - alumina Comp.
MicroPolish slurry
Comp. = Complementary (platen and specimen holder both rotate in the same direction)

29
Examples of Preparation Procedures

(diamond is applied during use), or lapping platens MasterPrep alumina slurries using napless, or low
of several types (diamond is applied and becomes to medium nap cloths. For some materials, such as
embedded in the surface during use). With this titanium and zirconium alloys, an attack polishing
huge array of products to choose from, how can solution is added to the slurry to enhance
the metallographer decide what to use? Each of deformation and scratch removal and improve
these products has advantages and disadvantages, polarized light response. Contra rotation (head
and this is only the first step. moves in the direction opposite to the platen) is
preferred as the slurry stays on the cloth better,
One or more steps using diamond abrasives on although this will not work if the head rotates at
napless surfaces usually follow planar grinding. a high rpm. Examples of generic contemporary
PSA-backed silk, nylon or polyester cloths are preparation practices follow in Tables 8 to 10.
widely used. These give good cutting rates, Specific procedures for a wide range of materials
maintain flatness and minimize relief. Silk cloths, are given in the next section.
such as the UltraPol cloth, provide the best flatness
and excellent surface finishes relative to the The starting SiC abrasive size is chosen based upon
diamond size used. UltraPad cloth, a thicker, hard, the degree of surface roughness and depth of
woven cloth, is more aggressive, gives nearly as cutting damage and the hardness of the material.
good a surface finish, similar excellent flatness, Never start with a coarser abrasive than necessary
and longer life than an UltraPol cloth. Synthetic to remove the cutting damage and achieve planar
chemotextile pads, such as the TexMet 1500 conditions in a reasonable time.
and 2500 cloths, give excellent flatness and are
more aggressive than silk. They are excellent for A similar scheme can be developed using
retaining second phase particles and inclusions. rigid grinding discs, such as the ApexHercules
Diamond suspensions are most popular with H Disc. These discs are generally restricted to
automated polishers as they can be added easily materials above a certain hardness, such as
during polishing; although it is still best to charge 175 HV, although some softer materials can be
the cloth initially with diamond paste of the same prepared using them. This disc can also be used
size to get polishing started quickly. Final polishing for the planar grinding step. An example of such a
could be performed with a very fine diamond size, practice, applicable to nearly all steels (results are
such as 0.1-µm diamond, depending upon the marginal for solution annealed austenitic stainless
material, your needs and personal preferences. steels) is given in Table 9.
Otherwise, final polishing is performed with
MasterMet colloidal silica or with MicroPolish or

Table 8. Generic Four-Step Contemporary Procedure for Many Metals and Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 120- to 240- 6 (27) 240-300 Until
abrasive discs (P120 to P280) Comp. Plane***
(waterproof paper) grit SiC water cooled
UltraPol 9-µm MetaDi 6 (27) 120-150 5:00
cloth Supreme diamond Comp.
suspension*
TexMet 1500 or 3-µm MetaDi 6 (27) 120-150 4:00
TriDent cloth Supreme diamond Comp.
suspension*
MicroCloth, ~0.05-µm 6 (27) 120-150 2:00
VelTex or MasterMet Contra**
ChemoMet colloidal silica or
cloths MasterPrep
alumina suspensions
* Plus MetaDi fluid extender as desired
** With heads that rotate at >100 rpm, contra may spray the abrasive above the splash guard
*** Until Plane = All specimens in the holder are ground to a common plane and the damage from sectioning has been removed
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = platen and specimen holder rotate in opposite directions

30
Examples of Preparation Procedures

The planar grinding step could also be its wear rate will be greater than the H disc when
performed using a 45-µm metal bonded, or a used to prepare very hard materials. A generic
30-µm resin bonded, UltraPrep disc or with the four step practice is given below for soft metals
ApexHercules H rigid grinding disc and 15- or and alloys.
30-µm diamond, depending upon the material.
Rigid grinding discs contain no abrasive; they The planar grinding step can be performed with
must be charged during use and suspensions are the 30-µm resin bonded diamond disc, or with
the easiest way to do this. Polycrystalline diamond a second ApexHercules S disc and 15- or
suspensions are favored over monocrystalline 30-µm diamond, depending upon the metal or
synthetic diamond suspensions for most metals alloy. For some very difficult metals and alloys,
and alloys due to their higher cutting rate. a 1-µm diamond step on a TriDent cloth (similar to
step 3, but for 3 minutes) could be added, and/or
The ApexHercules S rigid grinding disc, designed a brief vibratory polish (use the same cloths and
for soft metals and alloys, is used in a similar abrasives as for step 4) may be needed to produce
manner. This disc is quite versatile and can be perfect publication quality images. Four steps may
used to prepare harder materials as well, although suffice for routine work.

Table 9. Four-Step Contemporary Procedure for Steels Using a Rigid Grinding Disc

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 120- to 240- 6 (27) 240-300 Until
abrasive discs (P120 to P280) Comp. Plane
waterproof paper grit SiC, water cooled
ApexHercules H 9-µm MetaDi 6 (27) 120-150 5:00
rigid grinding disc Supreme diamond Comp.
suspension*
TexMet 1500 or 3-µm MetaDi 6 (27) 120-150 4:00
TriDent cloth Supreme diamond Comp.
suspension*
MicroCloth, ~0.05-µm 6 (27) 120-150 2:00
VelTex or MasterMet Contra
ChemoMet colloidal silica
cloths or MasterPrep
alumina suspensions
*Plus MetaDi fluid extender as desired
Comp.= Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 10. Four-Step Contemporary Procedure for Nonferrous Metals Using a Rigid Grinding Disc

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/ Direction (min:sec)
CarbiMet 220- to 320- 5 (22) 240-300 Until
abrasive discs (P280 or P400) grit Comp. Plane
waterproof paper SiC water cooled
ApexHercules S 6-µm MetaDi 5 (22) 120-150 5:00
rigid grinding disc Supreme diamond Comp.
suspension*
TriDent 3-µm MetaDi 5 (22) 120-150 4:00
cloth Supreme diamond Comp.
suspension*
MicroCloth, ~0.05-µm 5 (22) 120-150 2:00
VelTex or MasterMet Contra
ChemoMet colloidal silica
cloths or MasterPrep
alumina suspensions
*MetaDi fluid extender as desired
Comp.= Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

31
Procedures for Specific Materials

PROCEDURES FOR SPECIFIC specimen holder is used with the 12-inch (300
MATERIALS mm) diameter platen, the head position should
be adjusted so that the specimens rotate out over
Following are our recommendations for preparing the platen’s edge.
a wide range of materials arranged, in the case of
metals, according to common characteristics, basi- In general, it is difficult to make accurate predic-
cally in accordance with their classification in the tions about variations in time and load when using
periodic table of the elements. This break down different format platens and mount sizes. Differ-
is further modified, especially in the case of iron ent mount sizes per se should not be used as the
based alloys, as required due to the wide range of basis for calculations as the area of the specimens
properties that can be experienced. within the mount is more important than the
mount area itself. For example, the same specimen
The periodic table of the elements categorizes cross sectional area could be present in 1-, 1.25-
metallic, nonmetallic and metalloid elements or 1.5-inch (25, 30 or 40 mm) diameter polymer
according to similarities in atomic structure and mounts. Further, the number of specimens in the
is a good starting point, aided by similarities in holder can vary substantially. It is necessary to put
physical properties and behavior, for grouping at least three mounts in a central-force holder to
elements and their alloys that have reasonably obtain a proper weight balance. But, two speci-
similar preparation procedures. The periodic table mens could be dummy mounts (mounts without
(pg. 29) has the various groups color coded for specimens). In general, as the area of the specimen
convenience. in the mount increases, as the diameter and the
number of specimens increases, the time and pres-
Of course, metallographers prepare materials
sure must increase to obtain the same degree of
other than metals. These have been grouped
preparation perfection.
and sub-divided according to the nature of these
materials, as follows: If the area of the mounts is used, rather than the
• Sintered Carbides area of the embedded specimens, and we as-
sume that we are preparing six 1-inch (25 mm),
• Ceramics
six 1.25-inch (30 mm) or three 1.5-inch (40 mm)
• Composites: Metal Matrix, Polymer Matrix and diameter specimens, then we can adjust the pres-
Ceramic Matrix sures relative to the standard values in the tables
(for six 1.25-inch (30 mm) diameter specimens)
• Printed Circuit Boards
using an 8 inch (200 mm) diameter platen format
• Microelectronic Devices machine. A comparison of the cross-sectional
• Plastics and Polymers areas would suggest that for the six 1-inch (25
mm) specimens, the pressures could be reduced
The procedures discussed in this book were de- to 65% of the standard values while for three 1.5-
veloped and tested extensively using machines inch (40 mm) diameter specimens, the pressures
with an 8-inch (200 mm) diameter platen with and times could be reduced to about 70% of the
six 1.25-inch (30 mm) diameter specimens in a 5- standard values.
inch (125 mm) diameter holder. These procedures
Changing the platen diameter increases the dis-
produced the same results when the specimens
tance that the specimens travel in a given time.
were prepared using a 10-inch (250 mm) or a 12- For example, for the same rpm values, specimens
inch (300 mm) diameter platen. When using the travel 1.25 and 1.5 times as far using a 10- or 12-
8-inch (200 mm) platen and a 5-inch (125 mm) inch (250 or 300 mm) diameter platen, respectively,
diameter specimen holder, the machine’s head compared to an 8-inch (200 mm) platen. This sug-
position was adjusted so that the specimens rotate gests that the times could be reduced to 80 or
out over the outer periphery of the surface of the 67% of that used for the 8-inch (200 mm) diameter
platen. This procedure makes maximum use of the platen when using 10- or 12-inch (250 or 300 mm)
working surface and improves edge retention.This diameter platens. While preparation times may be
alignment practice is less critical with the larger shorter using a larger format machine, consumable
diameter platens when using the same diameter cost is greater and larger machines are needed
holder. However, if the 7-inch (175 mm) diameter only when the specimen throughput needs are
very high or with large specimens.
32
Procedures for Specific Materials

PERIODIC TABLE OF ELEMENTS

33
Light Metals

LIGHT METALS: Al, Mg and Be

4
Be
13
12 Al
Mg

Five and four-step practices for aluminum alloys

LIGHT METALS: Al, Mg and Be are presented below.While MgO was the preferred
final polishing abrasive for aluminum and its alloys,
Aluminum it is a difficult abrasive to use and is not available
in very fine sizes. Colloidal silica has replaced
Aluminum is a soft, ductile metal. Deformation
magnesia as the preferred abrasive for the final
induced damage is a common preparation prob-
step and is finer in size. For color etching work, and
lem in the purer compositions. After preparation,
for the most difficult grades of aluminum, a brief
the surface will form a tight protective oxide layer
vibratory polish may be needed to completely
that makes etching difficult. Commercial grades
remove any trace of damage or scratches. The
contain many discrete intermetallic particles with
five-step practice is recommended for super pure
a variety of compositions. These intermetallic par-
(SP) and commercially pure (CP) aluminum and for
ticles are usually attacked by etchants before the
wrought alloys that are difficult to prepare.
matrix. Although the response to specific etchants
has been used for many years to identify these
240- or 320-grit (P280 and P400) SiC waterproof
phases, this procedure requires careful control.
paper may also be used for the planar grinding
Today, energy-dispersive analysis is commonly
step. An UltraPol cloth produces better surface
performed for phase identification due to its
greater reliability.

Table 11. Five-Step Procedure for Aluminum Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 220 to 320
abrasive discs (P240 to P400) 5 (22) 240-300 Until
(waterproof paper) grit SiC water cooled Comp. Plane
UltraPol or 9-µm MetaDi 5 (22) 120-150 5:00
UltraPad cloths Supreme diamond* Comp.
TriDent cloth 3-µm MetaDi 5 (22) 120-150 4:00
Supreme suspension* Comp.
TriDent cloth 1-µm Metadi 5 (22) 120-150 2:00
paste* Comp.
MicroCloth, ~0.05-µm 5 (22) 120-150 1:30
VelTex or Mastermet Contra
ChemoMet cloth colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = platen and specimen holder rotate in opposite directions

34
Light Metals

finishes than an UltraPad cloth, but the UltraPad napless cloths may also be used for the final step
cloth has a longer useful life. A ChemoMet cloth with colloidal silica and they will introduce less
is recommended when edge retention is critical. relief than a low or medium nap cloth, but may
Pure aluminum and some alloys are susceptible not remove fine polishing scratches as well. For
to embedment of fine diamond abrasive particles, very pure aluminum alloys, this procedure could
especially when suspensions are used. If this oc- be followed by vibratory polishing to improve the
curs, switch to diamond in paste form, which is surface finish, as these are quite difficult to prepare
much less likely to cause embedding. SP and CP totally free of fine polishing scratches.
aluminum can be given a brief vibratory polish
(same products as last step) to improve scratch
control, although this is generally not required.
MasterPrep alumina suspension has been found to
be highly effective as a final polishing abrasive for
aluminum alloys, however, the standard alumina
abrasives made by the calcination process are
unsuitable for aluminum.

Deformed, elongated grain structure of extruded

6061-F aluminum after shearing revealed by anodizing
with Barker’s reagent (polarized light, 100X).

Helpful Hints For

Aluminum
Many low-alloy aluminum speci-
mens are difficult to polish to a
Dendritic segregation in as cast 206 aluminum perfect finish using standard methods. A vibra-
revealed by color etching with Weck’s reagent
(polarized light, 200X). tory polisher can be used quite effectively with
MasterMet colloidal silica to produce deforma-
For many aluminum alloys, excellent results can
tion-free, scratch-free surfaces for anodizing or
be obtained using a four step procedure, such as
color etching and examination with polarized
shown below in Table 12. This procedure retains all
light or Nomarski DIC.
of the intermetallic precipitates observed in alumi-
num and its alloys and minimizes relief. Synthetic

Table 12. Four-Step Procedure for Aluminum Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/ Direction (min:sec)
CarbiMet 220- to 320- (P240 5 (22) 240-300 Until
abrasive discs to P400) grit SiC Comp. Plane
(waterproof paper) water cooled
UltraPol 6-µm MetaDi 6 (27) 120-150 6:00
cloth Supreme Comp.
diamond
suspension*
TriDent cloth 3-µm MetaDi 6 (27) 120-150 4:00
Supreme diamond Comp.
suspension*
MicroCloth ~0.05-µm 6 (27) 120-150 2:00
VelTex or MasterMet Contra
ChemoMet colloidal silica
cloth or MasterPrep
alumina suspensions
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = platen and specimen holder rotate in opposite directions

35
Light Metals

Magnesium MasterPolish is nearly water-free and yields excel-

Preparation of magnesium and its alloys is rather lent results as the final abrasive. After the last step,
difficult due to the low matrix hardness and the wash the specimens with ethanol. Cleaning after
higher hardness of precipitate phases that lead the last step, without using water, is difficult. Hold-
to relief problems, and from the reactivity of the ing the specimen under running water for about
metal. Mechanical twinning may result during a second eased the cleaning problem and did
cutting, grinding, or handling if pressures are ex- not appear to harm the microstructure. Cosmetic
cessive. Final polishing and cleaning operations cotton puffs can scratch the surface when swab
should avoid or minimize the use of water and etching. For best results, etch-polish-etch cycle
a variety of solutions have been proposed. Pure may be needed. Magnesium has a hexagonal
magnesium is attacked slowly by water while Mg close-packed crystal structure and will respond
alloys may exhibit much higher attack rates. Some to polarized light. To enhance the response, use
authors state that water should not be used in any a brief vibratory polish with the materials used in
step and they use a 1 to 3 mixture of glycerol to the last step.
ethanol as the coolant even in the grinding steps.
Always grind with a coolant, as fine Mg dust is a
fire hazard. Because of the presence of hard inter-
metallic phases, relief may be difficult to control,
especially if napped cloths are used. Following
is a five-step procedure for magnesium and its
alloys, see Table 13.

As-cast microstructure of a magnesium alloy with

2.5% of rare earth elements, 2.11% Zn and 0.64% Zr
showing a film of the rare earth elements in the grain
boundaries, alloy segregation within grains (“coring”
revealed by color variations within grains) and a few
mechanical twins in the grains (acetic-picral etch,
crossed polarized light plus sensitive tint, 100X).

Mechanical twinning in deformed high-purity

magnesium (99.8% Mg) (acetic-picral etch, crossed
polarized light plus sensitive tint, 50X).

Table 13. Four-Step Procedure for Magnesium Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 220- to 320- (P240 5 (22) 200-250 Until
abrasive discs to P400) grit SiC* Comp. Plane
(waterproof paper) water cooled
TexMet 1500 9-µm MetaDi oil-based 5 (22) 120-150 6:00
pad diamond slurry Contra
TexMet 1500 3-µm MetaDi oil-based 5 (22) 120-150 5:00
pad diamond slurry Contra
TexMet 1500 1-µm MetaDi oil-based 5 (22) 120-150 4:00
pad diamond slurry Contra
ChemoMet ~0.05-µm 3(13) 120-150 1:30-3:00
cloth MasterPolish or Contra
MasterPrep alumina
*SiC surfaces were coated with wax to minimize embedment
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = platen and specimen holder rotate in opposite directions

36
Light Metals

Beryllium For the final step, mix hydrogen peroxide (30%

Beryllium is also a difficult metal to prepare and concentration – avoid physical contact!) with the
presents a health risk to the metallographer. Only MasterMet colloidal silica in a ratio of one part
those familiar with the toxicology of Be, and are hydrogen peroxide to five parts colloidal silica.
properly equipped to deal with these issues, should Oxalic acid solutions (5% concentration) have also
work with the metal.The grinding dust is extremely been used with alumina for attack polishing. For
toxic. Wet cutting prevents air contamination but optimal polarized light response, follow this with
the grit must be disposed of properly. As with Mg, vibratory polishing using a one-to-ten ratio of
Be is easily damaged in cutting and grinding form- hydrogen peroxide to colloidal silica.
ing mechanical twins. Light pressures are required.
Although some authors claim that water cannot
be used, even when grinding Be, others report no Helpful Hints For
difficulties using water. Attack-polishing agents Magnesium
are frequently used when preparing Be and many For polishing, apply diamond
are recommended [2]. Table 14 shows a four-step paste and use lapping oil, or use
practice for beryllium. Buehler’s oil-based MetaDi diamond suspen-
sions, available down to a 0.10-µm size, with a
medium nap cloth.

Helpful Hints For

Light Metals
Small diamond abrasive sizes
are prone to embedment when
applied as a suspension or by spraying. Charge
the cloth using diamond paste to eliminate
Grain structure of wrought, P/M beryllium (unetched,
embedding. Another technique is to use both
cross-polarized light, 100X). a fine diamond paste and MasterPrep alumina
suspension – simultaneously – to prevent em-
bedding.

Table 14. Four-Step Procedure for Beryllium

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 220- to 320- (P240 4 (18) 240-300 Until
abrasive discs to P400) grit SiC Comp. Plane
(waterproof paper) water cooled
UltraPol 6-µm MetaDi 4 (18) 120-150
cloth Supreme Comp. 5:00
diamond
suspension*
TriDent 3-µm MetaDi 4 (18) 120-150 4:00
cloth Supreme Comp.
diamond
suspension*
MicroCloth ~0.05-µm 3 (13) 80-120 2:00
VelTex or MasterMet Contra
ChemoMet colloidal silica
cloth
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

37
Low Melting Point Metals

LOW-MELTING POINT METALS:

Sb, Bi, Cd, Pb, Sn and Zn

50 51
Sn Sb
30
Zn 82 83
Pb Bi
48
Cd

LOW-MELTING POINT METALS:

Sb, Bi, Cd, Pb, Sn and Zn
As pure metals, antimony, bismuth, cadmium,
lead, tin, and zinc are all very soft and difficult to
prepare. Pure antimony is quite brittle, but alloys
containing Sb are more common. Bismuth is a
soft metal, but brittle, and not difficult to prepare.
However, retaining bismuth particles in free
machining steels is difficult. Cadmium and
zinc, both with hexagonal close-packed crystal
structures, are quite prone to mechanical twin
formation if sectioning or grinding is performed
too aggressively. Zinc is harder than tin or lead and
tends to be brittle. Zinc is widely used to coat sheet
steel (galvanized steel) for corrosion protection,
and is a common metallographic subject. Pure zinc
is very difficult to prepare. Lead is very soft and
ductile and pure specimens are extremely difficult
to prepare; however, lead alloys are considerably (top) Cadmium dendrites and a Cd-Bi eutectic in
an as-cast Cd – 20% Bi alloy (unetched, crossed
easier.Tin, which is allotropic with a body-centered polarized light plus sensitive tint, 50X). (bottom)
tetragonal crystal structure at room temperature, As-cast microstructure of Zn – 0.1% Ti – 0.2% Cu
alloy exhibiting mechanical twins and a three-phase
is soft and malleable and less sensitive to twin- eutectic (alpha-Zn, Cu-Zn and Zn15Ti) in a cellular
pattern (Palmerton reagent, crossed polarized light
ning. Pure tin, like pure lead, is very difficult to plus sensitive tint, 200X).

prepare. Due to their low melting points, and low

recrystallization temperatures, cold setting resins
are usually recommended as recrystallization may
occur during hot compression mounting. Some of
these metals in the pure, or nearly pure form, will
deform under the pressures used in compression
mounting. Alloys of these metals are easier to
prepare, as they are usually higher in hardness.
Heating of surfaces during grinding must be mini-
Proeutectic dendrites of Cu2Sb surrounded by a mized. Grinding of these metals is always difficult,
eutectic mixture of antimony and Cu2Sb, in an as-cast
Sb – 30% Cu specimen (unetched, polarized light, as SiC particles tend to embed heavily.
200X).

38
Low Melting Point Metals

Many authors have recommended coating the

SiC paper surface with bees wax, but this does not Helpful Hints For
solve the problem. Paraffin (candle wax) is much Low-Melting Point
better for reducing embedding. Embedding is Metals
most common with the finer grit size papers. Dia- Embedment of fine abrasive
mond is not a very effective abrasive with these particles is a common problem when preparing
metals. Alumina works quite well. Following is a soft, low-melting point specimens. To reduce
procedure for these alloys, see Table 15. embedment of SiC abrasive particles, coat
the paper with wax before grinding. Johnson
For best results, follow this with a vibratory polish (Wear, Vol. 16, 1970, p. 351-358) showed that
using MasterMet colloidal silica on MicroCloth or candle wax is far more effective than paraffin.
NanoCloth pads for times up to 1-2 hours. This However, if the layer applied is too thick, the
will improve polarized light response for the hcp metal-removal rate in grinding will be reduced
Cd and Zn and the rhombohedral Bi. The 1-µm drastically.
alumina step will remove any embedded silicon
carbide particles much more effectively than a
diamond abrasive step.

Table 15. Six-Step Procedure for Low-Melting Point Metals

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 320- (P400) grit SiC 4 (18) 150-250 Until
abrasive discs water cooled Comp. Plane
(waterproof paper) (wax coated)*
CarbiMet 400- (P600) grit SiC 4 (18) 150-250 0:30
abrasive water cooled Comp.
discs (wax coated)*
CarbiMet 600- (P1200) grit SiC 4 (18) 150-250 0:30
abrasive water cooled Comp.
discs (wax coated)*
CarbiMet 800- (P1500) grit SiC 4 (18) 150-250 0:30
abrasive water cooled Comp.
discs (wax coated)*
MicroCloth or 1-µm 5 (22) 120-150 4:00-5:00
VelTex MicroPolish II Comp.
pads deagglomerated
alumina suspension
MicroCloth or MasterPrep 4 (18) 80-150 4:00
VelTex 0.05-µm alumina Contra
pads suspension**
*Rub candle wax lightly across revolving disc prior to grinding
**See vibratory polish recommendation in text
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

39
Refractory Metals

REFRACTORY METALS:
Ti, Zr, Hf, Cr, Mo, Nb, Re, Ta, V, and W

22
50 23 24
SnTi V Cr

40 41 42
Zr Nb Mo

72 73 74 75
Hf Ta W Re

REFRACTORY METALS: Ti, Zr,

Hf, Cr, Mo, Nb, Re, Ta, V and W

Titanium
Pure titanium is soft and ductile, but is very easily
damaged by twinning in sectioning and grind-
ing. Preparation of commercially pure titanium,
which is a popular grade, is very difficult, while
preparation of the alloys is somewhat easier. Some
Alpha at the surface of heat treated (1038 ºC, water
authors have stated that titanium alloys should quench) Ti – 3% Cr alloy after tint etching with
not be mounted in phenolic resins as the alloys Beraha’s reagent (polarized light, 500X).

can absorb hydrogen from the resin. Further, it a castable resin with a very low exotherm (long
is possible that the heat from mounting could curing times favor lower heat generation, and vice
cause hydrides to go into solution. This is also versa). Titanium is very difficult to section and has
possible with castable resins if the exothermic low grinding and polishing rates. The following
reaction of polymerization generates excessive practice for titanium and its alloys demonstrates
heat. If the hydride phase content is a subject of the use of an attack-polishing agent added to the
interest, then the specimens must be mounted in

Table 16. Three-Step Procedure for Titanium and Titanium Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 320- (P400) grit 6 (27) 240-300 Until
abrasive discs SiC, water cooled Comp. Plane
(waterproof paper)
UltraPol 9-µm MetaDi 6 (27) 120-150 10:00
cloth Supreme diamond Contra
suspension*
MicroCloth or ~0.05-µm 6 (27) 120-150 10:00
VelTex MasterMet Contra
pads colloidal silica plus
attack polish agent**
*Plus Metadi fluid extender as desired **See text for agent
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

40
Refractory Metals

to obtain the best results, especially for com- effective). Polarized light response of CP titanium
mercially pure titanium, a rather difficult metal can be improved by following this procedure with
to prepare free of deformation for color etching, a brief vibratory polish using colloidal silica.
heat tinting and/or polarized light examination
of the grain structure. Attack polishing solutions
Zirconium and Hafnium
added to the abrasive slurry or suspension must
be treated with great care to avoid burns. Use Pure zirconium and pure hafnium are soft, ductile
good, safe laboratory practices and it is advisable hexagonal close-packed metals that can deform
to wear protective gloves. This three step practice by mechanical twinning if handled aggressively in
could be modified to four steps by adding a 3- or sectioning and grinding. As with most refractory
1-µm diamond step, but this is usually unneces- metals, grinding and polishing removal rates are
sary, see Table 16. low and eliminating all polishing scratches and
deformation can be difficult. It may even be pos-
A number of attack polishing agents have been sible to form mechanical twins in compression
used. The simplest is a mixture of 10 mL hydrogen mounting. Both can contain very hard phases
peroxide (30% concentration – avoid skin contact) that make relief control more difficult. To improve
and 50 mL colloidal silica. Some metallographers polarized light response, it is common practice
add either a small amount of Kroll’s reagent to to chemically polish specimens after mechanical
this mixture, or a few mL of nitric and hydrofluoric polishing. Alternatively, attack polishing additions
acids (avoid contact). These latter additions may can be made to the final polishing abrasive slurry,
cause the suspension to gel. In general, these acid or vibratory polishing may be employed. Table 17
additions do little to improve the action of the hy- is a four-step procedure that can be followed by
drogen peroxide (the safer 3% concentration is not either chemical polishing or vibratory polishing.

Several attack polishing agents have been used

for Zr and Hf. One is a mixture of 1-2 parts hy-
drogen peroxide (30% concentration – avoid
all skin contact) to 8 or 9 parts colloidal silica.
Another is 5 mL of a chromium trioxide solution
(20 g CrO3 to 100 mL water) added to 95 mL
colloidal silica or MasterpPrep alumina slurry.
Additions of oxalic, hydrofluoric or nitric acids have
also been used. All of these attack polishing addi-
Basket-weave alpha-beta structure of as-cast Ti – 6%
Al – 4% V revealed by heat tinting (polarized light,
100X).

Table 17. Four-Step Procedure for Zirconium and Hafnium

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 320- (P400) 5 (22) 200-250 Until
abrasive discs grit SiC Comp. Plane
(waterproof paper) water cooled
UltraPol or 9-µm MetaDi Supreme 5 (22) 150-200 5:00
UltraPad cloths diamond suspension* Contra
TriDent cloth 3-µm MetaDi Supreme 5 (22) 150-200 3:00
diamond suspension* Contra
MicroCloth, ~0.05-µm 6 (27) 120-150 7:00
VelTex or MasterMet Contra
ChemoMet colloidal silica plus
cloths attack polish agent**
*Plus MetaDi fluid extender as desired **See text for agent
Comp. = Complementary (platen and specimen holder both rotate in the same direction
Contra = Platen and specimen holder rotate in opposite directions

41
Refractory Metals

tions must be handled with care as they are strong

oxidizers. Skin contact must be avoided. Chemical
polishing solutions are reviewed in [2]. Cain’s has
been popular. Use under a hood and avoid skin
contact. Ann Kelly developed an excellent chemi-
cal polish for refractory metals, such as Zr, Hf, and
Ta. It consists of 25 mL lactic acid, 15 mL nitric acid
and 5 mL hydrofluoric acid. Swab vigorously for up
to 2 minutes. To prepare ultra-pure Zr and Hf, the
above method is unsatisfactory and a procedure,
such as in Table 15 (add a 5- or 3-µm alumina
step) is needed, as diamond is ineffective. Use the
chromium trioxide attack polish with the alumina
and conclude with Kelly’s chemical polish.

(top) Cold worked microstructure of wrought

zirconium alloy XL (Zr - 3% Sn - 0.8% Nb - 0.8% Mo)
in the as-polished condition at 200X with polarized
light plus sensitive tint. (bottom) Microstructure (α-Zr
+ AlZr3) of wrought Zr - 8.6% Al in the as-polished
condition at 200X with polarized light.

Helpful Hints For

Refractory Metals
Due to their very low rate of
grinding and polishing, the
more aggressive contra rotation gives suitable
surfaces in less time.

Refractory metal preparation is aided by

using attack-polishing additives. Polarized
light response may be improved by following
preparation by swabbing with a chemical
polishing solution.

(top) Equiaxed grain structure of wrought zirconium

(unetched, crossed polarized light, 100X). (middle)
Equiaxed grain structure of wrought hafnium
(unetched, crossed polarized light plus sensitive
tint, 100X). (bottom) Fine grain recrystallized
microstructure of wrought Zr - 1.14% Cr strip in the
as-polished condition at 200X using polarized light
plus sensitive tint.

42
Other Refractory Metals

OTHER REFRACTORY METALS: reported to vary with crystallographic orientation.

Cr, Mo, Nb, Re, Ta, V and W Rhenium is very sensitive to cold work and will
form mechanical twins. Tantalum is softer than
These refractory metals have body-centered
niobium and more difficult to prepare as it easily
cubic crystal structures (except for rhenium
forms damaged layers in sectioning and grinding.
which is hexagonal close packed) and are soft
Tantalum may contain hard phases that promote
and ductile when pure, but some are brittle in
relief control problems. Vanadium is a soft, ductile
commercial form. They can be cold worked easily,
metal but may be embrittled by hydrogen; other-
although they do not work harden appreciably, so
wise it can be prepared much like a stainless steel.
it may be difficult to get completely deformation-
Tungsten is not too difficult to prepare, although
free microstructures.
grinding and polishing rates are low. Hard carbides
and oxides may be present in these metals that
introduce relief control problems.

Mechanical polishing often incorporates an

attack-polishing agent in the final step or is
followed by vibratory polishing or chemical pol-
ishing. Manual preparation of these metals and
their alloys tends to be very tedious due to their
low grinding and polishing rates. Automated
Microstructure of powder-made W – 5.6% Ni – 2.4% Fe
approaches are highly recommended, espe-
revealing tungsten grains surrounded by a nickel-iron cially if attack polishing is performed. Following
matrix (unetched, Nomarski DIC, 200X).
is a generic four step practice suitable for these
Pure chromium is soft and brittle; but, when en- metals and their alloys, see Table 18.
countered commercially, for example, as a plated
Many attack-polishing additives [2] have been
layer, it is hard and brittle. Chromium alloys are
suggested for these metals and their alloys.
relatively easy to prepare, although difficult to etch.
A good general purpose attack polish consists
Molybdenum may be tough or brittle depending
of a mixture of 5 mL chromium trioxide solution
upon composition. It is susceptible to deformation
(20 g CrO3 in 100 mL water) to 95 mL MasterMet
damage in sectioning and grinding. Pure niobium
colloidal silica. Avoid skin contact as this is a
(columbium) is soft and ductile and difficult to
strong oxidizing solution. A number of chemical
prepare while its alloys are harder and simpler to
prepare. Grinding and polishing rates have been

Table 18. Four-Step Procedure for Refractory Metals

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 320- (P400) 6 (27) 150-250 Until
abrasive discs grit SiC Comp. Plane
(waterproof paper) water cooled
UltraPol or 9-µm MetaDi 6 (27) 150-200 10:00
UltraPad Supreme diamond Contra
cloths suspension*
TexMet 1500 3-µm MetaDi 6 (27) 150-200 8:00
TriDent cloth Supreme diamond Contra
suspension*
MicroCloth, ~0.05-µm 6 (27) 120-150 5:00
VelTex or MasterMet Contra
ChemoMet colloidal silica plus
cloths attack polish agent**
*Plus Metadi fluid extender as desired **See text for agent
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

43
Other Refractory Metals

polishing solutions have been suggested [2]. For

Nb, V and Ta, use a solution consisting of 30 mL
water, 30 mL nitric acid, 30 mL hydrochloric acid
and 15 mL hydrofluoric acid. Swab or immerse at
room temperature. An alternative chemical pol-
ish for Nb, V and Ta consists of 120 mL water, 6 g
ferric chloride, 30 mL hydrochloric acid and 16 mL
hydrofluoric acid. Eary and Johnson recommend
immersing the specimen in this solution 1 minute
for V, 2 minutes for Nb and 3 minutes for Ta. Vibra-
tory polishing is also very helpful for these metals
and their alloys.

(top) Non-recrystallized grain structure of wrought

pure molybdenum, longitudinal plane (500X, polarized
light, etch of water, hydrogen peroxide (30% conc)
and sulfuric acid in 7:2:1 ratio).
(bottom) Equiaxed alpha grain structure in wrought
pure vanadium (200X, etch: glycerol-nitric acid-
hydrofluoric acid, 1:1:1ratio).

(top) Alpha grains in Mo – 47.5% Re. The small spots

are sigma phase (Murakami’s reagent, 200X).
Deformed alpha grains in W – 25% Re containing
sigma phase (black spots) revealed using (middle) (top) Wrought tungsten - 10 atomic % Ti containing
bright field and (bottom) Nomarski DIC which reveals a small amount of alpha-Ti, beta-Ti-W eutectic and
the cold work much better (Murakami’s reagent, grains of beta-Ti,W of varying composition and crystal
500X). orientation (500X, Kroll’s reagent/Murakami’s reagent
at room temperature). (bottom) Fine grain boundary
precipitates (not identified) in wrought, cold worked
Fan Steel 85-03 alloy (Nb - 28% Ta - 10.5% W - 0.9%
Zr), longitudinal plane (500X, etchant: lactic acid-
nitric acid-hydrofluoric acid, 30:10:5 ratio).

44
Ferrous Metals

FERROUS METALS: Fe

26
Fe

FERROUS METALS identified in steels. Addition of 12 or more percent

chromium dramatically reduces the corrosion
Iron-based alloys account for a large portion of rate of steels, producing a wide range of stain-
all metals production. The range of compositions less steel alloys. Tool steels cover a wide range of
and microstructures of iron-based alloys is far compo sitions and can attain very high hardnesses.
wider than any other system. Pure iron is soft and
ductile. Development of scratch-free and deforma-
tion-free grain structures is difficult. Sheet steels
present the same problem, which can be compli-
cated by protective coatings of zinc, aluminum
or Zn-Al mixtures. In general, harder alloys are
much easier to prepare. Cast irons may contain
graphite, which must be retained in preparation.
Inclusions are frequently evaluated and quantified.
Volume fractions can vary from nearly 2% in a free
machining grade to levels barely detectable in a Lath martensitic microstructure of an unconsolidated
powder metallurgy steel gear containing substantial
premium, double vacuum melt alloy. A wide range porosity (Klemm’s I reagent, polarized light, 200X).
of inclusion, carbide and nitride phases has been

Table 19. Four-Step Procedure for Hard Steels using the BuehlerHercules H disc

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 45-µm 6 (27) 240-300 Until
metal-bonded diamond, Comp. Plane
disc water cooled
ApexHercules H 9-µm MetaDi Supreme 6 (27) 120-150 5:00
rigid grinding disc diamond suspension* Comp.
TriDent cloth 3-µm MetaDi Supreme 6 (27) 120-150 3:00
diamond suspension* Comp.
MicroCloth, MasterPrep 6 (27) 120-150 2:00
VelTex or 0.05-µm alumina Contra
ChemoMet suspension,
cloths or MasterMet
colloidal silica
*Plus Metadi fluid extender desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

45
Ferrous Metals

Preparation of ferrous metals and alloys is quite content and the potential for staining problems, it is
straightforward using the contemporary methods. best to use the MasterPrep alumina suspension for
Edge retention (see guidelines on pages 13-15) the final polishing step.
and inclusion retention are excellent, especially
if automated equipment is used. The following For softer steels, use the 30-µm resin bonded Ultra-
procedures are recommended and are adequate Prep disc, or 240- (P280) grit SiC paper, for step 1.
for the vast majority of ferrous metals and alloys. The ApexHercules S disc can be used in place of the
Table 9 presented a variation of the method given ApexHercules H disc but this is not usually necessary.
in Table 19. However, with either disc, use 6-µm MetaDi Supreme
for the second step instead of 9-µm, as shown in
This practice is also useful for cast iron specimens Table 20.
including graphitic cast irons. 240- (P280) grit SiC
paper can be substituted for the UltraPrep disc This practice is well suited for solution annealed
or another ApexHercules H disc can be used for austenitic stainless steels and for soft sheet steels.
planar grinding with 30-µm MetaDi Supreme UltraPol or UltraPad cloths could be substitute for
diamond suspension. Due to their high silicon the rigid grinding discs, if desired. For perfect pub-
lication quality images, or for color etch ing, follow

Table 20. Four-Step Procedure for Soft Steels using the BuehlerHercules S disc

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 30-µm 6 (27) 240-300 Until
resin-bonded diamond, Comp. Plane
disc water cooled
ApexHercules S 6-µm MetaDi 6 (27) 120-150 5:00
rigid grinding disc Supreme diamond Comp.
suspension*
TriDent cloth 3-µm MetaDi 6 (27) 120-150 3:00
Supreme diamond Comp.
suspension*
MicroCloth, MasterPrep 6 (27) 120-150 2:00
VelTex or 0.05-µm alumina Contra
ChemoMet suspension,
cloths or MasterMet
colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 21. Three-Step Procedure for Steels

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 120- to 320- 6 (27) 240-300 Until
Abrasive Discs (P120 to P400) Comp. Plane
(Waterproof Paper) grit SiC, water cooled
ApexHercules H or S 3-µm MetaDi 6 (27) 120-150 5:00
Rigid Grinding Disc Supreme Comp.
Diamond Suspension*
MicroCloth MasterPrep 0.05-µm 6 (27) 120-150 5:00
VelTex or Alumina Suspension, Contra
ChemoMet or MasterMet
Cloths Colloidal Silica
*Plus MetaDi Fluid extender as desired
Comp. = Complementary (platen and specimen holder rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

46
Ferrous Metals

this practice with a brief vibratory polish using the disc. For softer stainless steels, use the 30-µm
cloths and abrasives in the last step. resin-bonded diamond UltraPrep disc for planar
grinding. Another alternative is a second Apex
Many steels, particularly the harder steels, can be Hercules disc, either H or S, depending upon
prepared in three steps with excellent results. A the hardness of the grade, and a 30-µm MetaDi
recommended practice is given in Table 21. Supreme polycrystalline diamond suspension.
The solution annealed austenitic stainless steels
For soft alloys, use 240- or 320- (P280 or P400) and the fully ferritic stainless steels are the most
grit SiC paper; for harder alloys use 120 (P120), difficult to prepare. It may be helpful to add a 1-µm
180- (P180), or 240- (P280) grit SiC paper, diamond step on a TriDent cloth before the last
depending upon the starting surface finish and step, or to follow the last step with a brief vibratory
the hardness of the alloy. Planar grinding can polish using colloidal silica on MicroCloth, Nano-
also be performed using 45-µm metal-bonded or Cloth or a ChemoMet cloth, see Table 22.
30-µm resin-bonded UltraPrep diamond discs. For
softer steels, use the ApexHercules S disc for best
results. The UltraPol cloth can also be used for the
second step for steels of any hardness.

The following practice is recommended for

stainless steels and maraging steels. For solu-
tion annealed austenitic grades and for ferritic
stainless grades and for annealed maraging
grades, use the ApexHercules S disc or the Ultra-
Pol cloth for best results. Start with 120- (P120) Microstructure of a duplex (ferrite tan, austenite
grit SiC paper only if it is a very hard martensitic white) stainless steel held at 816º C for 48 hours
which forms sigma phase (orange particles) revealed
stainless steel, such as type 440C. For the mar- by electrolytic etching with aqueous 20% NaOH at 3
tensitic grades, planar grinding can be performed V dc, 9 s (500X).

using a 45-µm metal-bonded diamond UltraPrep

Table 22. Four-Step Contemporary Procedure for Stainless Steels and Maraging Steels

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 120- to 240- (P120 6 (27) 240-300 Until
abrasive discs to P280) grit SiC, Comp. Plane
(waterproof paper) water cooled
ApexHercules S 9-µm 6 (27) 120-150 5:00
rigid grinding disc or MetaDi Comp.
UltraPol cloth Supreme
Diamond
Suspension*
TriDent cloth 3-µm 6 (27) 120-150 5:00
MetaDi Comp.
Supreme
diamond
suspension*
MicroCloth, MasterPrep 6 (27) 120-150 2:30
VelTex or 0.05-µm alumina Contra
ChemoMet suspension,
cloths or MasterMet
colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

47
Copper, Nickel and Cobalt

Copper, Nickel and Cobalt

27 28 29
Co Ni Cu

COPPER, NICKEL and COBALT rary method followed by vibratory polishing, see
Table 23.
Copper
Planar grinding can be performed using the
Pure copper is extremely ductile and malleable.
45- or 15-µm metal-bonded or the 30-µm resin-
Copper and its alloys come in a wide range
bonded UltraPrep discs. Use the resin-bonded
of compositions, including several variants of
disc for the soft copper grades and copper
nearly pure copper for electrical applications
alloys.
to highly alloyed brasses and bronzes and to
precipitation hardened high strength alloys.
Copper and its alloys can be easily damaged by
rough sectioning and grinding practices and
the depth of damage can be substantial. Scratch
removal, particularly for pure copper and
brass alloys, can be very difficult. Following the
preparation cycle with a brief vibratory polish
using colloidal silica is very helpful for scratch
removal. Attack-polish additions have been
used in the past to improve scratch removal but Alpha grains containing annealing twins in
usually are not necessary using the contempo- phosphorous-deoxidized arsenical bronze that was
annealed and lightly cold drawn (Klemm’s I reagent,
polarized light, 50X).

Table 23. Five-Step Procedure for Cu and Cu Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet abrasive discs 220- to 320- (P240 to 5-6 240-300 Until
(waterproof paper) P400) grit SiC water cooled (22-27) Comp. Plane
UltraPol cloth or 6-µm MetaDi 5-6 120-150 5:00
ApexHercules S Supreme diamond (22-27) Comp.
rigid grinding disc suspension*
TriDent or 3-µm MetaDi Supreme 5-6 120-150 3:00
TexMet 1500 pads diamond suspension* (22-27) Comp.
TriDent 1-µm MetaDi Supreme 5-6 120-150 2:00
cloth Diamond Suspension* (22-27) Comp.
MicroCloth, ~0.05-µm MasterMet 5-6 120-150 1:30-2:00
VelTex or colloidal silica (22-27) Contra
ChemoMet or MasterPrep
pads alumina suspensions
*Plus MetaDi Fluid extender as desired
Comp.= Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

48
Copper, Nickel and Cobalt

Nickel If color etching is to be performed, follow the last

Nickel and its alloys have face-centered cubic step with a brief vibratory polish using the same
crystal structures and are prepared in basically materials as in the last step.This step is also helpful
the same way as austenitic stainless steels. for the most difficult to prepare solution annealed
Pure nickel is more difficult to prepare than the alloys. Alternatively, for the most difficult speci-
alloys. The Ni-Fe magnetic alloys are rather dif- mens, or when color etching is being performed,
ficult to prepare scratch free unless vibratory a 1-µm diamond step on a Trident cloth can be
polishing is used. The Monel (Ni-Cu) and the added before the final step.
highly corrosion resistant (Ni-Cr-Fe) alloys are more
difficult to prepare than the nickel-based super- For pure nickel, nickel-copper and nickel-iron
alloys. Solution annealed superalloys are always alloys, a five step practice is preferred, as given
more difficult to prepare than age hardened below. The planar grinding step can be performed
superalloys. Age hardened superalloys can be using either the 30-µm resin-bonded UltraPrep
prepared using the ApexHercules H disc; for all diamond disc or with 240- (P280) or 320- (P400)
other nickel alloys, use the ApexHercules S disc for grit SiC papers with equal success, see Table 25.
best results. The following practice works well for
nickel based superalloys (and Fe-Ni based super Attack-polishing agents are not often used with
alloys) and the highly corrosion resistant Ni-Cr-Fe these alloys to eliminate fine polishing scratches
alloys, see Table 24. or residual damage. If this is a problem, and some
of these grades are very difficult to get perfectly

Table 24. Four-Step Procedure for Nickel-Based Superalloys and Ni-Cr-Fe Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet abrasive discs 220- to 240- (P240 to P280) 6 (27) 240-300 Until
(waterproof paper) grit SiC water cooled Comp. Plane
ApexHercules H 9-µm MetaDi 6 (27) 120-150 5:00
or ApexHercules S Supreme diamond Comp.
rigid grinding disc suspension*
Trident cloth 3-µm MetaDi Supreme 6 (27) 120-150 5:00
diamond suspension* Comp.
MicroCloth, MasterPrep 0.05-µm 6 (27) 120-150 2:00-5:00
VelTex alumina suspension or Contra
ChemoMet pads MasterMet colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 25. Five-Step Procedure for Ni, Ni-Cu and Ni-Fe Alloys

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 30-µm diamond 5 (22) 200-300 Until
resin-bonded disc water cooled Comp. Plane
UltraPol cloth or 9-µm MetaDi 6 (27) 100-150 5:00
ApexHercules S Supreme diamond Comp.
rigid grinding disc suspension*
TriDent or 3-µm MetaDi Supreme 6 (27) 100-150 3:00
TexMet 1500 pads diamond suspension* Comp.
TriDent 1-µm MetaDi Supreme 6 (27) 100-150 2:00
cloth diamond suspension* Comp.
MicroCloth, ∼0.05-µm MasterMet 6 (27) 80-150 1:30-2:00
VelTex or colloidal silica or Contra
ChemoMet pads MasterPrep
alumina suspensions
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

49
Copper, Nickel and Cobalt

free of scratches and deformation damage, a brief Two steps of SiC paper may be needed to get the
vibratory polish, using the same materials as in the specimens co-planar. If the cut surface is of good
last step, will provide the needed improvement. quality, start with 320- (P400) grit paper. Cobalt
MasterPrep alumina may give better results than and its alloys are more difficult to cut than most
colloidal silica for the more pure nickel composi- steels, regardless of their hardness. Attack polish-
tions. ing has not been reported but chemical polishing
has been used after mechanical polishing. Morral
(2) has recommended two chemical polishing
solutions: equal parts of acetic and nitric acids
(immerse) or 40 mL lactic acid, 30 mL hydrochloric
acid and 5 mL nitric acid (immerse). A wide variety
of Co-based alloys have been prepared with the
above method without need for chemical polish-
ing. The 1-µm diamond step could be eliminated
for routine work.
Alpha grains containing annealing twins of solution
annealed and double aged Waspaloy nickel-based
superalloy (Beraha’s reagent, 100X).

Cobalt
Cobalt and its alloys are more difficult to
prepare than nickel and its alloys. Cobalt is a
tough metal with a hexagonal close-packed
crystal structure and is sensitive to deformation
damage by mechanical twinning. Grinding
and polishing rates are lower for Co than for Ni, Equiaxed grain structure of Elgiloy (Co – 20% Cr
– 15% Fe – 15% Ni – 7% Mo – 2% Mn – 0.15% C
Cu or Fe. Preparation of cobalt and its alloys is – 0.05% Be) after hot rolling and annealing revealing
somewhat similar to that of refractory met- annealing twins (Beraha’s reagent, crossed polarized
light plus sensitive tint, 100X).
als. Despite its hcp crystal structure, crossed
polarized light is not very useful for examining
cobalt alloys compared to other hcp metals and
alloys. Following is a practice for preparing Co and
its alloys, see Table 26.

Table 26. Five-Step Procedure for Cobalt

Abrasive/ Load Lb. (N)/ Base Speed Time
Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 220- to 320- 6 (27) 250-300 Until
abrasive discs (P240 to P400) grit Contra Plane
(Waterproof Paper) SiC water cooled
UltraPol cloth or 9-µm MetaDi 6 (27) 100-150 5:00
ApexHercules S Supreme diamond Contra
rigid grinding disc suspension*
TexMet 1500 3-µm MetaDi Supreme 6 (27) 100-150 5:00
pad diamond suspension* Contra
TriDent or 1-µm MetaDi Supreme 6 (27) 100-150 3:00
TexMet 1500 pads diamond suspension* Contra
MicroCloth, ~0.05-µm MasterMet 6 (27) 80-120 2:00-3:00
VelTex or colloidal silica or Contra
ChemoMet MasterPrep
pads alumina suspensions
*Plus Metadi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

50
Precious Metals

PRECIOUS METALS: Au, Ag, Ir,

Os, Pd, Pt, Rh, and Ru

44 45 46 47
Ru Rh Pd Ag

76 77 78 79
Os Ir Pt Au

PRECIOUS METALS: Au, Ag, Ir, for metallographers. Damaged surface layers are
Os, Pd, Pt, Rh and Ru easily produced and grinding and polishing rates
are low. It is quite difficult to prepare. Palladium is
Relatively few metallographers work with precious malleable and not as difficult to prepare as most
metals, other than those used in electronic devices. of the precious metals. Platinum is soft and mal-
Preparing precious metals within an integrated leable. Its alloys are more commonly encountered.
circuit is discussed later (see Microelectronic De- Abrasive embedment is a problem
vices).The precious metals are very soft and ductile,
deform and smear easily, and are quite challenging
to prepare. Pure gold is very soft and the most
malleable metal known. Alloys, which are more
commonly encountered, are harder and somewhat
easier to prepare. Gold is difficult to etch. Silver is
very soft and ductile and prone to surface dam-
age from deformation. Embedding of abrasives is
a common problem with both gold and silver and
their alloys. Iridium is much harder and more eas-
ily prepared. Osmium is rarely encountered in its Eutectic microstructure of Ag – 28% Cu where
Klemm’s reagent has colored the copper particles
pure form, even its alloys are infrequent subjects and the silver phase is uncolored (500X).

Table 27. Stewart’s Manual Procedure for Precious Metals

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet abrasive 220- to 240- (P240 to P280) moderate 400 Until
discs (waterproof paper) grit SiC water cooled Plane
CarbiMet 320- (P400) grit SiC moderate 400 1:00
abrasive discs water cooled
CarbiMet 400- (P600) grit SiC moderate 400 1:00
abrasive discs water cooled
CarbiMet 600- (P1200) grit SiC moderate 400 1:00
abrasive discs water cooled
MetCloth 6-µm MetaDi Supreme moderate 400 2:00
pad diamond suspension
MasterTex 1-µm MetaDi Supreme moderate 400 2:00
cloth diamond suspension
Comp. = Complementary (platen and specimen holder both rotate in the same direction)

51
Precious Metals

with Pt and its alloys. Rhodium is a hard metal are used for the diamond steps, as it will hold the
and is relatively easy to prepare. Rh is sensitive abrasive in its surface well, which minimizes em-
to surface damage in sectioning and grinding. bedding. Only diamond paste is used, as slurries
Ruthenium is a hard, brittle metal that is not too will be more prone to embedding, as we observed
difficult to prepare. in high gold alloys. Use only a small amount of
distilled water as the lubricant. Do not get the
Stewart (Tech-Notes, Vol. 2, Issue 5) has described cloth excessively wet. Final polishing is with a
a method for preparing jewelry alloys. Invariably, ChemoMet I cloth and MasterPrep alumina. Due to
these are small pieces, due to their cost, and their excellent corrosion resistance, colloidal silica
must be mounted. Stewart uses EpoMet G resin is not effective as an abrasive for precious metals.
for most specimens. If transparency is need, he The ChemoMet pad has many fine pores to hold
uses TransOptic resin. Fragile beads and balls are the abrasive. The cycle is given in Table 28.
mounted in castable resins. Grinding and polishing
was conducted at 400 rpm. His practice is shown For 18 karat gold and higher (≥75% Au), it is neces-
in Table 27. sary to use an attack polish agent in the final step.
An aqueous solution of 5 g CrO3 in 100 mL water
This can be followed by a brief vibratory polish works well. Mix 10 mL of the attack polish agent
using colloidal silica on MasterTex, ChemoMet, with 50 mL of MasterPrep suspension. This will
MicroCloth or NanoCloth pads to further enhance thicken, so add about 20-30 mL water to make
the quality of the preparation; but, a 1-µm diamond it thinner. A 3 to 6 minute attack- polish step will
finish is adequate for most work. Attack-polishing remove the fine polishing scratches. Wear protec-
has been used to prepare gold and its alloys and tive gloves as the chromium trioxide solution is a
chemical polishing has been performed after me- strong oxidizer.
chanical polishing of silver, but neither practice is
commonly performed. Alternate etch-polish cycles
may be needed to remove fine polishing scratches,
especially for annealed specimens.

A procedure for an automated system was

developed after experimentation with a num-
ber of precious metal specimens. Most of these
metals and alloys are quite soft, unless they have
been cold worked, and they are susceptible to Equiaxed grain structure of cold rolled and annealed
embedding of abrasives. In this method, only one 18-karat gold (Neyoro 28A: 75% Au – 22% Ag – 3% Ni)
revealing annealing twins (equal parts 10% NaCN and
silicon carbide step is used. TexMet 1000 pads 30% H2O2, 50X).

Table 28. Five-Step Automated Procedure for Precious Metals

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 220- to 320- (P240 to 3 (13) 150-240 Until
abrasive discs P400) grit water cooled Comp. Plane
TexMet 1500 9-µm MetaDi II 3 (13) 150-240 5:00
pad diamond paste* Comp.
TexMet 1500 3-µm MetaDi II 3 (13) 150-240 3:00
pad diamond paste* Comp.
TexMet 1500 1-µm MetaDi II 3 (13) 150-240 2:00
pad diamond paste* Comp.
ChemoMet MasterPrep 2 (9) 100-150 2:00
pad alumina suspension Comp.
* Use water as the lubricant, but keep the pad relatively dry
Comp. = Complementary (platen and specimen holder both rotate in the same direction)

52
Thermally-Spray Coated Specimens

THERMALLY-SPRAY COATED dye, may be added to the epoxy. When viewed

SPECIMENS with fluorescent illumination, the epoxy-filled
voids appear bright yellow-green. This makes it
Thermally sprayed coatings (TSC) and thermal easy to discriminate between dark holes and dark
barrier coatings (TBC) are widely used on many oxides, as would be seen with bright field illumi-
metal substrates. Invariably, these coatings are nation. Filling the pores with epoxy also makes
not 100% dense but contain several types of it easier to keep the pore walls flat to the edge
voids, such as porosity and linear detachments. during preparation. Aside from this mounting
Hot compression mounting is not recommended requirement, TSC and TBC specimens are prepared
as the molding pressure can collapse the voids. using all of the factors needed for good edge re-
Use a low-viscosity castable epoxy and use tention (see pages 13-15). A variety of procedures
vacuum infiltration to fill the connected voids can be used. The ApexHercules H rigid grinding
with epoxy. Fluorescent dyes, such as FluorMet disc produces exceptional edge flatness for these
Table 29. Four-Step Procedure for TSC and TBC Specimens with Metallic Coatings

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
ApexHercules H 30-µm MetaDi 5 (22) 240-300 Until
rigid grinding disc Supreme diamond Contra Plane
suspension*
ApexHercules H 6-µm MetaDi 5 (22) 150-200
rigid grinding disc Supreme diamond Contra 4:00
suspension*
TriDent 3-µm MetaDi 6 (27) 120-150 3:00
cloth Supreme diamond Comp.
suspension*
MicroCloth, MasterPrep 4 (18) 120-150 2:00
VelTex or 0.05-µm alumina Contra
ChemoMet suspension or
cloths MasterMet
colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 30. Four-Step Procedure for TSC and TBC Specimens with Metallic Coatings

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
ApexHercules H 30-µm MetaDi 5 (22) 240-300 Until
rigid grinding disc Supreme diamond Contra Plane
suspension*
UltraPol 9-µm MetaDi 5 (22) 150-200 4:00
cloth Supreme diamond Contra
suspension*
TriDent 3-µm MetaDi 6 (27) 120-150 3:00
cloth Supreme diamond Comp.
suspension*
MicroCloth, MasterPrep 4 (18) 120-150 2:00
VelTex or 0.05-µm alumina Contra
ChemoMet suspension or
cloths MasterMet
colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

53
Thermally-Spray Coated Specimens

specimens. Two four step procedures for TSC and The 30-µm resin bonded, or the 45-µm metal
TBC specimens using the Apex Hercules H rigid bonded UltraPrep diamond discs, can be substi-
grinding disc for specimens with metallic coatings tuted for the planar grinding step. (Table 29 and
and one procedure for specimens with ceramic 30). UltraPol or UltraPad cloths can be used in the
coatings are given, see Tables 29-31. second step. (Table 29 and 31).

NiCrAlY thermally-spray coated steel specimen Microstructure of a steel substrate covered by two
revealing a small amount of porosity (black spots), thermally-sprayed layers, a NiAl bond coat and yittria-
linear detachments (elongated black lines), and zirconia top coat (unetched, 100X). The bond coat
inclusions (gray particles) (unetched, 100X). contains pores, linear detachments and inclusions
while the top coat is quite porous.

Table 31. Four-Step Procedure for TSC and TBC Specimens with Ceramic Coatings

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 45-µm disc 5 (22) 240-300 Until
metal-bonded disc water cooled Contra Plane
ApexHercules H 9-µm 5 (22) 150-180 4:00
rigid grinding disc MetaDi Contra
Supreme
diamond
suspension*
TexMet 1500 or 2500 3-µm 6 (27) 120-150 3:00
or TriDent cloth MetaDi Comp.
Supreme
diamond
suspension*
MicroCloth, MasterPrep 8 (36) 120-150 2:30
ValTex or 0.05-µm alumina Contra
ChemoMet suspension or
cloths MasterMet
colloidal silica
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

54
Sintered Carbides

SINTERED CARBIDES ed UltraPrep discs or the ApexHercules H rigid

grinding discs. A further option would be the use
Sintered carbides are very hard materials made of the traditional metal-bonded diamond discs.
by the powder metallurgy process and may
by reinforced with several types of MC-type
carbides besides the usual tungsten carbide
( WC). The binder phase is normally cobalt
although minor use is made of nickel. Mod-
ern cutting tools are frequently coated with a
variety of very hard phases, such as alumina,
titanium carbide, titanium nitride and titanium
carbonitride. Sectioning is normally performed
with a precision saw, so surfaces are very good
and rough abrasives are not usually required. Surface of a WC – 11.5% Co cutting tool enriched with
1.9% Ta and 0.4% Nb (form complex carbides, the
Listed below are two alternate procedures dark regions in the matrix) and coated with alumina
for preparing sintered carbides using metal-bond- (arrows) for enhanced tool life (Murakami’s reagent,
1000X).

Table 32. Four-Step Procedure for Sintered Carbides using UltraPrep Discs

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 45-µm diamond 6 (27) 240-300 5:00**
metal-bonded disc water cooled Contra
UltraPrep 9-µm diamond 6 (27) 240-300 4:00
metal-bonded disc water cooled Contra
TriDent, UltraPol, 3-µm MetaDi 6 (27) 120-150 3:00
TexMet 1500 Supreme diamond Contra
or Nylon cloths suspension*
TriDent, ~0.05-µm MasterMet 6 (27) 120-150 2:00
UltraPol colloidal silica or Contra
or Nylon cloths MasterPrep alumina
suspensions
*Plus MetaDi fluid extender as desired
** Grind longer than “Until Plane”
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 33. Four-Step Procedure for Sintered Carbides using ApexHercules H Discs

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
ApexHercules H 30-µm MetaDi 6 (27) 240-300 5:00**
rigid grinding disc Supreme diamond Contra
suspension*
ApexHercules H 9-µm MetaDi 6 (27) 240-300 4:00
rigid grinding disc Supreme diamond Contra
suspension*
TriDent , UltraPol, 3-µm MetaDi 6 (27) 120-150 3:00
TexMet 1500 Supreme diamond Comp.
or Nylon cloths suspension*
MicroCloth, ~0.05-µm MasterMet 6 (27) 120-150 2:00
VelTex or colloidal silica or Contra
ChemoMet cloths MasterPrep alumina
suspensions
*Plus MetaDi fluid extender as desired
** Grind longer than “Until Plane”
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

55
Sintered Carbides

If a greater amount of material must be removed in

the planar grinding step, use either the 30-µm resin
bonded or the 45-µm metal-bonded UltraPrep disc
for step one, in Table 33.

b
Microstructure of cold isostatically pressed and age
hardened Ferro-Titanit Nikro 128 (Fe - 13.5% Cr - 9%
Co - 4% Ni - 30% TiC) cutting tool;
a) etched with Murakami’s reagent at room
temperature to darken the TiC; and b) electrolytic
Microstructure at the surface of a multilayered, CVD- etched with aqueous 1% chromium trioxide at 2 V dc,
coated WC - 8% Co cutting tool. The arrows point 30 s to reveal the martensitic matrix (1000X).
to the CVD layers of TiCN, TiN, TiC and alumina.
The region below the coatings is made higher in Co
to improve crack resistance and complex carbide
forming elements (Ta, Ti and Nb) are added to the Helpful Hints For
matrix (dark spots at bottom) for wear resistance
(Murakami’s reagent, 1000X).
Sintered Carbides
The ability to see the boundar-
The final step in these two procedures, Tables
ies between the WC particles
32 and 33, employs either MasterMet colloidal
and cobalt binder in the as-polished condition
silica or MasterPrep alumina suspensions as they
depends on the surface of the polishing cloth
will produce the best definition of the structure,
used in the last step. To see no boundaries,
particularly if the surfaces have a complex series
which makes it easier to see graphite or eta
of coatings for improved wear resistance, as used
phase, use a nap-less surface, such as TriDent,
in some coated carbide inserts. However, if such a
TexMet 1500 or UltraPol pads. To see the phase
coating is not present, the matrix grain structure
boundaries between WC particles and the co-
can be observed quite clearly after the 3-µm dia-
balt binder, use a medium nap surface, such as
mond step. For routine examination there is little
MicroCloth, MasterTex pads.
need to go beyond the third step. It is also possible
to use a 1-µm diamond step, similar to step three,
but for 2 minutes, as an alternative fourth step.

56
Ceramics

CERAMICS
Ceramic materials are extremely hard and brittle
and may contain pores. Sectioning must be per-
formed using diamond blades. If the specimen is to
be thermally etched, then it must be mounted in a
resin that permits easy demounting, and vacuum
infiltration of epoxy into the voids should not be
done. Deformation and smearing are not problems
with ceramics due to their inherent characteristics.
Microstructure of hot-pressed silicon nitride (binder
But, it is possible to break out grains or produce unknown) (unetched, 500X).
cracking during preparation.

Table 34. Four-Step Procedure Suitable for Most Ceramic Materials

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 45-µm diamond 8 (36) 120-150 Until
metal-bonded disc MetaDi fluid Comp. Flat
for coolant
ApexHercules H 9-µm MetaDi 6 (27) 120-150 4:00
rigid grinding disc Supreme diamond Comp.
suspension*
ApexHercules S 3-µm MetaDi 6 (27) 120-150 3:00
rigid grinding disc Supreme diamond Comp.
suspension*
TexMet 1500, 1-µm MetaDi 10 (44) 120-150 5:00
Nylon or Supreme diamond Contra
UltraPol suspension* or
cloths MasterPolish 2
suspension
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 35. Four-Step Procedure Suitable for Most Ceramic Materials

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 45-µm diamond 8 (36) 120-150 Until
metal-bonded MetaDi fluid Contra Flat
disc for coolant
UltraPad 15-µm MetaDi 6 (27) 120-150 6:00
cloth Supreme diamond Comp.
suspension*
UltraPad or 6-µm MetaDi 6 (27) 120-150 4:00
TexMet 1500 or Supreme diamond Comp.
TexMet 2500 cloths suspension*
TexMet 1500 or 3-µm MetaDi 8 (36) 120-150 4:00
TexMet 2500, Supreme diamond Contra
Nylon or suspension and
UltraPol MasterMet
cloths colloidal silica
(50:50 simultaneously)
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

57
Ceramics

Pullouts are a major problem to control as they

can be misinterpreted as porosity. Mechanical Helpful Hints For
preparation has been done successfully with laps, Ceramics
metal-bonded diamond discs, rigid grinding discs, Cut specimens as close as pos-
or hard cloths. SiC paper is rather ineffective with sible to the desired examination
most ceramics, as they are nearly as hard or as plane using a saw and blade that produces the
hard as the SiC abrasive. Consequently, diamond least possible amount of damage, producing
is commonly used for nearly all preparation steps. the best possible surface for subsequent prepa-
Automated preparation is highly recommended ration. Rough grinding is slow and tedious and
when preparing ceramic materials as very high does considerable damage to the structure and
forces arise between the specimen and the work- should be avoided as much as possible.
ing surface, often too high for manual preparation.
Listed are two generic procedures for preparing Coarse abrasives for grinding may be needed
ceramics using different approaches, see Tables at times. However, severe grain pull-outs can
34 and 35. result from use of coarse abrasives and it is
recommended to use the finest possible coarse
MasterPolish 2 suspension is specifically formu- abrasive for grinding.
lated for final polishing ceramic materials and will
yield a better surface finish than 1-µm diamond.

(top) Microstructure of an alumina (matrix), zirconia

(gray), silica (white) refractory material (unetched,
200X). (bottom) Silicon - SiC ceramic etched
electrolytically with 10% oxalic acid (500X).

58
Composites

COMPOSITES
Composites cover a wide range of composi-
tions, general grouped under the subcategories
metal-matrix composites (MMC), polymer-matrix
composites (PMC) and ceramic-matrix composites
(CMC). Consequently, preparation schemes are
quite difficult to generalize due to the extreme
range of materials employed, differences in hard-
ness and grinding/polishing characteristics; so,
relief control is a big issue. Pull out problems are
also very common, especially with PMCs. Section-
ing frequently produces considerable damage that
must be removed in the initial preparation step.
Mounting with castable epoxy resin along with
vacuum impregnation is frequently performed.

(top) Microstructure of a polymer-matrix composite,

graphite fabric-reinforced polysulfone (unetched,
polarized light, 100X). (bottom) Microstructure of
alumina fibers in an aluminum - lithium alloy matrix
(500X, as-polished, Nomarski DIC).

(top) Microstructure of a metal-matrix composite,

SiC in a CP titanium matrix (unetched, Nomarski DIC,
100X).

Helpful Hints For

Composites
Sectioning damage can propa-
gate down the fibers and can be
difficult to remove. Prevent damage by using
precision saw sectioning and diamond wafering
blades with a very fine abrasive, specifically like
(top) Synthetic foam made with hollow ceramic
5 LC and 10 LC Series. Thinner blades produce spheres added to 7075 aluminum (100X original
less damage. Section as close as possible to the magnification, magnification bar is 100 µm long).
Etched with aqueous 0.5% HF. (bottom) Carbon-
plane of interest. reinforced polymer composite (original at 100X;
magnification bar is 100 µm long). Specimen is as-
Composites can be damaged by compression polished and viewed with Nomarski DIC illumination.

mounting. Castable mounting resins with low

peak temperature are preferred.

Avoid aggressive grinding steps by using longer

times and finer abrasives.

59
Composites

Table 36. Four-Step Procedure for Metal-Matrix Composites

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 30-µm diamond 5 (22) 240-300 Until
resin-bonded disc water cooled Contra Plane
UltraPol, or 9-µm 5 (22) 150-180 4:00
UltraPad cloths MetaDi Supreme Contra
diamond suspension*
TriDent 3-µm 6 (27) 120-150 3:00
cloth MetaDi Supreme Comp.
diamond suspension*
ChemoMet ~0.05-µm MasterMet 6 (27) 120-150 2:00
cloth colloidal silica or Contra
MasterMet
alumina suspension
*Plus MetaDi fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

Table 37. Four-Step Procedure for Polymer-Matrix Composites

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet abrasive discs 320- (P400) grit 4 (18) 150-240 Until
(waterproof paper) SiC water cooled Comp. Plane
UltraPol, or 6-µm MetaDi 4 (18) 120-150 4:00
UltraPad Supreme diamond Comp.
cloths suspension*
TexMet 1500, 3-µm MetaDi 5 (22) 120-150 4:00
TriDent or Supreme diamond Comp.
Nylon cloths suspension*
ChemoMet or ~0.05-µm MasterMet 6 (27) 120-150 2:00-4:00
MicroCloth colloidal silica or Comp.
pads MasterPrep
alumina suspension
*Plus MetaDi Fluid extender as desired
Comp. = Complementary (platen and specimen holder both rotate in the same direction)

Table 38. Four-Step Procedure for Ceramic-Matrix Composites

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep 45-µm diamond MetaDi 6 (27) 240-300 Until
metal-bonded disc fluid for coolant Contra PLane
UltraPol, UltraPad, 15-µm MetaDi 6 (27) 120-150 4:00
or TexMet 1500 Supreme diamond Contra
or 2500 cloths suspension*
UltraPol, TriDent, 6-µm MetaDi 6 (27) 120-150 3:00
or TexMet 1500 Supreme diamond Contra
or 2500 cloths suspension*
UltraPol, 1-µm MetaDi 6 (27) 120-150 2:00
ChemoMet Supreme diamond Contra
or Nylon suspension and
cloths ~0.05-µm MasterMet
colloidal silica
(50:50 simultaneously)
*Plus MetaDi Fluid extender as desired
Contra = Platen and specimen holder rotate in opposite directions

60
Printed Circuit Boards

commonly found in the boards. This changes,

PRINTED CIRCUIT BOARDS however, when ‘packed’ boards with ceramic or
semiconductor components must be sectioned.
The vast majority of printed circuit boards (PCB’s) Please refer to methods of preparation for
are of a rigid variety, the bulk of which are com- ceramics or microelectronic materials for infor-
posed of layers of woven glass fiber cloth in a mation on these cases. A generalized method for
polymeric matrix. Flex circuits, which are becoming nonpacked PCB’s of both the rigid and flex variet-
quite common, do not typically contain glass fiber, ies is presented in Table 39.
but instead, their bulk is often composed of layers
of polyimide. The circuitry in both types of boards Quality control of manufactured PCB’s often
is composed of plated and/or foil metal. The metal demands statistical analysis based on plating
used is generally copper, while in a few cases, gold thickness measurements taken from the centers
and/or nickel plating may be present. Furthermore, of plated through-holes. However, to gener-
depending upon whether the boards have under- ate enough data for a representative sampling,
gone assembly or shock testing, solders of various numerous specimens must be prepared. This is
compositions might also be present. easiest when coupons are taken from PCB’s and
are ganged together in precise alignment. In this
Luckily for the metallographer, the variety of manner, the through holes in all of the coupons
materials present in PCB’s generally do not com- can be sectioned at one time, with one set of
plicate the preparation methods due to the fact consumables. The Buehler NelsonZimmer System
that extremely hard and brittle materials are not allows this type of automation with few adjust-
ments to the method listed below.

Helpful Hints For

Printed Circuit
Boards
To improve penetration of
acrylic mounting material into the through-
holes of a PCB specimen, dip the “pinned”
PCB coupons in liquid hardener immediately
prior to mounting in acrylic. This will “wick” the
Micrograph showing a plated-through hole in a acrylic into the through-holes and minimize
printed circuit board (resin and fibers) with copper the presence of problematic bubbles in these
foil, electroless-plated copper and Pb-Sn solder
(equal parts ammonium hydroxide and 3% hydrogen critical locations.
peroxide, 200X).

Table 39. Five-Step Procedure for Non-Packed Printed Circuit Boards

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet abrasive 180- (P180) grit SiC 5 (22) 120-150 Until
discs (waterproof paper) water cooled Contra Plane
CarbiMet 320- (P400) grit SiC 5 (22) 120-150 1:30
abrasive discs water cooled Contra
CarbiMet 600- (P1200) grit SiC 5 (22) 120-150 1:30
abrasive discs water cooled Contra
TexMet 1500, 3-µm MetaDi II 5 (22) 150-240 1:30
or TriDent cloth diamond paste* Contra
MicroCloth 0.05-µm MasterPrep 5 (22) 100-150 1:15
cloth alumina suspension Contra
*Plus MetaDi Fluid extender as desired
Contra = Platen and specimen holder rotate in opposite directions

61
Electronic Materials

ELECTRONIC MATERIALS phy of preparation in which we give our attention

specifically to the materials of interest.
The term,‘microelectronic materials’ encompasses
an extremely wide range of materials. This is due First and foremost in the class of ‘microelectronic
to the fact that most microelectronic devices materials’ is silicon. Silicon is a relatively hard, brittle
are composites, containing any number of in- material, which does not respond well to grinding
dividual components. For example, present day with large silicon carbide abrasives. Silicon carbide
microprocessor failure analysis might require the papers contain strongly bonded abrasive particles
metallographer to precisely cross section through which, when they collide with the leading edge of
a silicon chip plated with multiple thin-film layers a piece of silicon, create significant impact dam-
of oxides, polymers, ductile metals such as cop- age. In addition, they create tensile stresses on the
per or aluminum, and refractory metals such as trailing edge of silicon, which results in deep and
tungsten and/or titanium-tungsten. In addition, destructive cracking. Cutting close to the target
the packaging of such a device might contain area is preferable to grinding, but to accurately
materials of such varying mechanical properties approach the target area within a silicon device,
as toughened aluminum oxide and solder. The fine grinding is still necessary. This is the point at
solder materials may have compositions rang- which silicon preparation divides into two distinct
ing up to 97% lead. With such a vast number of classes. The first is preparation through traditional
materials incorporated into a single device, and metallographic techniques. The second is prepa-
with these materials having such highly disparate ration of unencapsulated silicon chips (or die),
mechanical properties, it is virtually impossible to using special fixturing and abrasives. This second
develop a general method for achieving perfect category is a specialized field of metallography,
metallographic results. Instead, we must focus on which is discussed in other texts [8] and is beyond
a few individual materials, and develop a philoso- the scope of this book. Therefore, we will focus on
standard metallographic techniques throughout
this discussion.

Standard preparation of epoxy encapsulated

silicon is similar to general metallographic
preparation, with the exception that only very
fine grades of silicon carbide paper are used.
Table 40 illustrates a typical process of prepara-
tion.

Aluminum circuitry on silicon adjacent to mineral

filled epoxy packaging, (1000x).

Table 40. Three-Step Procedure for Preparing Silicon in Microelectronic Devices

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
UltraPrep, Type A 15-µm diamond Manual 100-150 Until
diamond lapping film water cooled Plane
on TriDent cloth
TexMet 1500 3-µm diamond MetaDi II 7 (31) 240-300 2:00
pad diamond paste/ Comp
MasterPrep as lubricant
ChemoMet 0.02-µm 5 (22) 100-150 1:30
pad MasterMet 2 Contra
colloidal silica
Comp. = Complementary(platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions
Note: The first step is manual so that you can monitor the progress of grinding relative to the feature of interest. The remaining steps are accomplished using a
system that can apply individual force to the back of the spceimen. The first step can be automated if there is a significant amount of material to remove.

62
Electronic Materials

Using silicon carbide grinding paper with abrasive want to be sure that the nickel doesn’t smear, and
sizes coarser than 600- (P1200) grit may cause seri- that we are able to discern the true microstructure
ous damage to the silicon. Therefore, this method of the lead frame materials (refer to section listing
often requires precision cutting or cleaving near to preparation methods for copper and nickel).
the target area of the specimen prior to the initial
grinding step. Care must be taken to assure that When preparing a silicon device for the purpose
only precision saw blades designed especially for of inspecting the metalized, thin film circuitry,
microelectronic materials are used when cutting. the techniques would be the same as those listed
above. Again, the choice of final polishing agent
Silicon responds very well to the chemo-mechani- would be determined entirely by the features of
cal polishing effects of colloidal silica, but not as interest. For instance, aluminum circuitry responds
well to aluminum oxide. However, there are in- extremely well to the chemomechanical effects of
stances when an aluminum oxide product, such colloidal silica, but tungsten and titanium-tung-
as the MasterPrep suspension, should be used for sten are not as effectively polished with colloidal
final polishing. This relates back to the philosophy silica as are the materials by which they are sur-
of preparation mentioned earlier. An example of rounded. Therefore, colloidal silica often causes
the appropriate use of MasterPrep alumina would such refractory metals to stand in relief against
be: when a silicon die is attached to a lead frame a well polished background. This results in edge
material such as nickel-plated copper, and the lead rounding of the refractory metals and can make
frame and die attach materials are the materials interface analysis quite difficult. An alternative is to
of interest. In this case, we are not so concerned use an extremely fine diamond suspension for final
with the surface finish of the silicon, but instead, we polishing in order to reduce these effects.

When lead alloy solders are the materials of inter-

est in a microelectronic package, we can often
separate them into two general categories. The
first is the eutectic, or near-eutectic solders. These
tend to be ductile, but they respond well to the
same general preparation listed in Table 40. (Dur-
ing the 3-µm polishing step, the best results are
obtained by using diamond paste.

High temperature (96% Pb) solder under flip chip on

ceramic substrate, (200x).

Table 41. Three-Step Procedure for High-Temperature Solder on Ceramic Devices

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
ApexHercules S 9-µm MetaDi II 7 (31) 100-150 Until
rigid grinding disc diamond paste/ Comp. Plane
MasterPrep
as lubricant*
TexMet 3-µm MetaDi II 7 (31) 240-300 Until
1500 diamond paste/ Comp. all embedded
MasterPrep 9-µm diamond
as lubricant is removed
ChemoMet 0.02-µm 5 (22) 100-150 2:00
pad MasterMet 2 Contra
colloidal silica
*Apply sparingly
Comp. = Complementary(platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

63
Electronic Materials

Pastes contain a waxy base that reduces diamond Polishing of high-temperature solders attached to
embedding in ductile metals. Limit the use of ceramics often produces undesired edge round-
extenders in such cases. A few drops of water ing. However, this cannot be completely avoided.
lubrication does not break down the waxy base, The cushioning effects of polishing pads will tend
and therefore it retains the desired properties of to cause the abrasive to preferentially remove ma-
diamond paste.) terial from ductile materials faster than from hard,
brittle materials. By using stiff, flat polishing pads,
one can reduce this effect, but at the expense of
having additional material embedded in the high-
temperature solder. One effective preparation
technique is to perform a grind-polish-grind-pol-
ish type of procedure. In this technique, a polishing
step, which utilizes a napped polishing cloth, is
used in conjunction with a step-down in abrasive
size from the previous grinding step. This process
is used to polish out the embedded material. The
Silicon flip chip, solder balls, on PC board, (200x).
next step is a grinding process, utilizing another
The second category of lead solders is the high step-down in abrasive size, to flatten the specimen
temperature solders (in the range of 90% to 97% again. This continues to a sufficiently fine polish-
lead). These are difficult to prepare, and require ing stage at which final polishing will produce the
special consideration. This is especially true when desired result. Table 41 illustrates an example of
they are incorporated into a device that contains such a process.
a ceramic package. The mechanical properties
of these toughened ceramics require aggressive
grinding techniques, which cause ceramic debris Helpful Hints For
(and often, the abrasive used for the grinding op- Electronic Materials
eration) to became embedded in the solder. Silicon Consider what area of
carbide abrasives are a poor choice for grinding in the specimen is most
such cases, as silicon carbide is only slightly harder critical when choosing a final polishing
(and significantly more brittle) than the typical suspension. For example, the chemo- me-
packaging material they are intended to grind. As chanical nature of colloidal silicas (such as
the silicon carbide abrasive fractures, it produces MasterMet and MasterMet 2) make them
elongated shards, which embed deeply into high- ideal for polishing silicon, glass, oxides,
temperature solders. In addition, they do not grind aluminum, and to some degree copper. Howev-
the ceramic package effectively, and therefore er, polishing nickel platings and gold ball bonds
produce extreme edge rounding at the interface with these agents often results in smearing. For
with the solder. Diamond grinding produces a such non-reactive materials, use MasterPrep
more desired result since diamond abrasives aluminum oxide final polishing suspension to
are blockier in shape, and are therefore easier to produce a flat, scratch-free surface.
remove after they have embedded themselves. In
addition, they are capable of creating a flat surface
when ceramic packages are present.

High removal rates and good surface finishes have

been obtained on ceramic packages when using
diamond paste on the ApexHercules S discs.

64
Polymers

POLYMERS specimen. However, there may be a visibility

problem if a transparent, clear polymeric speci-
Plastics and polymers are normally quite soft. Many
men is mounted in a clear, transparent epoxy. In
different sectioning methods have been used. A
this case, EpoColor resin, with its deep red color,
sharp razor blade or scalpel, or even a pair of scis-
will produce excellent color contrast between
sors, can be used. Microtomes have been used, and
specimen and mount in darkfield or polarized
a surface cut after refrigeration in liquid nitrogen
light. Due to their light weight, a specimen may
requires very little subsequent preparation. A
float in the epoxy filled mold. To prevent this,
jeweler’s saw may be used, although the surface
put double sided adhesive tape on the bottom
is somewhat rough. The precision saw produces
of the mold cup and set the specimen on top of
excellent surfaces, while an abrasive cut-off saw
the tape. Always use practices that minimize the
yields a bit more damage and a rougher surface.
exotherm during polymerization.
Blades and wheels for sectioning polymers are
available from Buehler. Damage from sectioning Preparation of plastics and polymers for micro-
is quite easy to remove. structural examination follows the same basic
principles as for other materials. Automated
polishing devices are preferred to manual prepa-
ration. Rough grinding abrasives are unnecessary,
even for the planar grinding step. Pressures
should be lighter than used for most metals.

Micrograph showing multi-layered polymer meat

casing mounted in epoxy (dark field, 100X).

Surface quality can be degraded by abrasion

from the debris produced during grinding and
polishing. Mounted specimens are much easier
to prepare than nonmounted specimens. Castable Micrograph of high-density polyethylene containing
a filler (white particles) (unetched, Nomarski DIC,
resins are preferred as the heat from a mounting 100X).
press may damage or alter the structure of the

Table 42. Generic Procedure for Preparing Plastic and Polymeric Specimens

Abrasive/ Load Lb. (N)/ Base Speed Time

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 320- (P400) grit SiC 4 (18) 200-250 Until
abrasive discs water cooled Contra Plane
(waterproof paper)
CarbiMet 400- (P600) grit SiC 4 (18) 200-250 0:30
abrasive discs water cooled Contra
CarbiMet 600- (P1200) grit SiC 4 (18) 200-250 0:30
abrasive discs water cooled Contra
CarbiMet 800- (P1500) grit SiC 4 (18) 200-250 0:30
abrasive discs water cooled Contra
TexMet 1500 or 3-µm MetaDi II 5 (22) 100-120 4:00
TriDent diamond paste Comp.
cloths and MetaDi fluid
MasterTex 0.05-µm MasterPrep 3 (13) 100-120 4:00
cloth alumina suspension Contra
Comp. = Complementary (platen and specimen holder both rotate in the same direction)
Contra = Platen and specimen holder rotate in opposite directions

65
Polymers

Water is generally used as the coolant, although

some plastics and polymers may react with Helpful Hints For
water. In such cases, use a fluid that will not Polymers
react with the particular plastic or polymer. Plastics often have little con-
Embedding can be a problem with plastics and trast from the mounting media
polymers. ASTM E 2015 (Standard Guide for when viewed under a microscope. This makes
Preparation of Plastics and Polymeric Specimens thickness measurement and discerning edges
for Microstructural Examination) describes difficult. Mounting in EpoColor dye enhanced
procedures for preparing several types of plastics epoxy will resolve this situation by producing
and polymers. Table 42 contains a generic proce- excellent color contrast between the speci-
dure for preparing many plastic and polymeric mens and mount.
materials. Depending upon the material and the
surface roughness after sectioning, it may be pos- Polymers and plastics may react with fluids dur-
sible to start grinding with 400- (P600) or even ing sectioning and preparation. Always check
600- (P1200) grit SiC paper. your particular polymer and choose the proper
fluid, usually a water, or oil based solution to
If flatness is critical, the final step can be altered. avoid a reaction.
Use a TexMet pad with MasterPrep alumina abra-
sive slurry at 10 lbs. pressure (44N) per specimen,
120 rpm, contra rotation, for 3 minutes.

66
Etching

ETCHING screen magnifications that may make detection

of fine structures easier. However, resolution is
Metallographic etching encompasses all processes not improved beyond the limit of 0.2-0.3-µm for
used to reveal particular structural characteristics the light microscope. Microscopic examination
of a metal that are not evident in the as-polished of a properly prepared specimen will clearly re-
condition. Examination of a properly polished veal structural characteristics such as grain size,
specimen before etching may reveal structural segregation, and the shape, size, and distribution
aspects such as porosity, cracks, and nonmetallic of the phases and inclusions that are present.
inclusions. Indeed, certain constituents are best Examination of the microstructure will reveal
measured by image analysis without etching, prior mechanical and thermal treatments give
because etching will reveal additional, unwanted the metal. Many of these microstructural features
detail and make detection difficult or impossible. are measured either according to established im-
The classic examples are the measurement of age analysis procedures, e.g., ASTM standards, or
inclusions in steels and graphite in cast iron. Of internally developed methods.
course, inclusions are present in all metals, not just
steels. Many intermetallic precipitates and nitrides Etching is done by immersion or by swabbing (or
can be measured effectively in the as-polished electrolytically) with a suitable chemical solution
condition. that essentially produces selective corrosion.
Swabbing is preferred for those metals and
In certain nonferrous alloys that have non-cubic alloys that form a tenacious oxide surface layer
crystallographic structures (such as beryllium, with atmospheric exposure such as stainless steels,
hafnium, magnesium, titanium, uranium and aluminum, nickel, niobium, and titanium and their
zirconium), grain size can be revealed adequately alloys. It is best to use surgical grade cotton that
in the as polished condition using polarized light. will not scratch the polished surface. Etch time var-
Figure 30 shows the microstructure of cold-drawn ies with etch strength and can only be determined
zirconium viewed in cross-polarized light.This pro- by experience. In general, for high magnification
duces grain coloration, rather than a “flat etched” examination the etch depth should be shallow;
appearance where only the grain boundaries are while for low magnification examination a deeper
dark. etch yields better image contrast. Some etchants
produce selective results in that only one phase
will be attacked or colored. A vast number of
etchants have been developed; the reader is di-
rected to references 1-3, 9 and ASTM E 407. Table
43 lists some of the most commonly used etchants
for the materials described in this book.

Etchants that reveal grain boundaries are very im-

portant for successful determination of the grain
Figure 30. Mechanical twins at the surface of hot size. Grain boundary etchants are given in [1-3, 9].
worked and cold drawn high-purity zirconium viewed Problems associated with grain boundary etch-
with polarized light (200X).
ing, particularly prior austenite grain boundary
etching, are given in [2, 10 and 11]. Measurement
Etching Procedures of grain size in austenitic or face-centered cubic
Microscopic examination is usually limited to a metals that exhibit annealing twins is a commonly
maximum magnification of 1000X — the approxi- encountered problem. Etchants that will reveal
mate useful limit of the light microscope, unless grain boundaries, but not twin boundaries, are
oil immersion objectives are used. Many image reviewed in [2].
analysis systems use relay lenses that yield higher

67
Etching

Selective Etching color either the anodic or cathodic constituent in a

Image analysis work is facilitated if the etchant microstructure. Tint etchants are listed and illus-
chosen improves the contrast between the fea- trated in several publications [1-3, 16-21].
ture of interest and everything else. Thousands
of etchants have been developed over the years, A classic example of the different behavior of
but only a small number of these are selective in etchants is given in Figure 31 where low-carbon
nature. Although the selection of the best etchant, sheet steel has been etched with the standard
and its proper use, is a very critical phase of the nital and picral etchants, and also a color tint etch.
image analysis process, only a few publications Etching with 2% nital reveals the ferrite grain
have addressed this problem [12-14]. Selective boundaries and cementite. Note that many of the
etchants, that is, etchants that preferentially ferrite grain boundaries are missing or quite faint;
attack or color a specific phase, are listed in a problem that degrades the accuracy of grain size
[1-3, 9, 13 and 14] and illustrated in 13 and 14. ratings. Etching with 4% picral reveals the cement-
Stansbury [15] has described how potentiostatic ite aggregates (one could not call this pearlite as it
etching works and has listed many preferential is too nonlamellar in appearance and some of the
potentiostatic etching methods. The potentiostat cementite exists as simple grain boundary films)
offers the ultimate in control over the etching pro- but no ferrite grain boundaries. If one is interested
cess and is an outstanding tool for this purpose. in the amount and nature of the cementite (which
Many tint etchants act selectively in that they can influence formability), then the picral etch is
far superior to the nital etch as picral revealed only
the cementite. Tint etching with Beraha’s solution
(Klemm’s I could also be used) colored the grains
according to their crystallographic orientation.
With the development of color image analyzers,
this image can now be used quite effectively to
provide accurate grain size measurements since
all of the grains are colored.

a
Figure 32 shows a somewhat more complex
example of selective etching. The micrograph
shows the ferrite-cementite-iron phosphide ter-
nary eutectic in gray iron. Etching sequentially
with picral and nital revealed the eutectic, Figure
32a, surrounded by pearlite. Etching with boiling
alkaline sodium picrate, Figure 32b, colored the

Figure 32a. The ternary eutectic (α-Fe3C-Fe3P) in gray

cast iron revealed by etching (a, top) in picral and
nital to “outline” the phases

c
cementite phase only, including in the surround-
Figure 31. Microstructure of low-carbon sheet steel
etched with (a, top) 2% nital, (b, middle) 4% picral;
ing pearlite (a higher magnification is required to
and (c, bottom) Beraha’s reagent (100 mL water, 10 g see the very finely spaced cementite that is more
Na2S2O3 and 3 g K2S2O5) at 100X.
lightly colored). Etching with boiling Murakami’s
reagent, Figure 32c, colors the iron phosphide

68
Etching

lightly color the cementite after prolonged etch- etchants to reveal the microstructure of a dual
ing. The ferrite could be colored preferentially if phase stainless steel in the hot rolled and annealed
Klemm’s I was used. condition. Figure 33a shows a well delineated
structure when the specimen was immersed in
ethanolic 15% HCl for 30 minutes. All of the phase
boundaries are clearly revealed, but there is no
discrimination between ferrite and austenite.
Twin boundaries in the austenite are not revealed.

c
Figure 32b. (top) boiling alkaline sodium picrate to
color the cementite. Figure 32c. (bottom) boiling
Murakami’s reagent to color the phosphide (200X).

Selective etching has been commonly applied

to stainless steels for detection, identification d
and measurement of delta ferrite, ferrite in dual
phase grades, and sigma phase. Figure 33 shows
examples of the use of a number of popular

f
Figure 33. Microstructure of a duplex stainless steel
revealed by electrolytic etching with (c) aqueous 60%
b HNO3 (1 V dc, 20 s); with (d) aqueous 10% oxalic acid
Figure 33. Microstructure of a duplex stainless (6 V dc, 75 s); with (e) aqueous 10% CrO3 (6 V dc, 30
steel revealed using (a, top) alcoholic 15% HCl by s); and, with (f) aqueous 2% H2SO4 (5 V dc, 30 s) at
immersion (30 min.); and, with (b, bottom) glyceregia 200X.
by swabbing (2 min.) at 200X.

69
Etching

Glyceregia, a etchant for stainless steels, was not Selective etching techniques are not limited to
suitable for this grade, Figure 33b, as it appears to iron-based alloys, although these have more thor-
be rather orientation sensitive. Many electrolytic oughly developed than for any other alloy system.
etchants have been used for etching stainless Selective etching of beta phase in alpha-beta
steels, but only a few have selective characteristics. copper alloys has been a popular subject. Figure
Of the four shown in Figures 33c to f, only aqueous 34 illustrates coloring of beta phase in Naval Brass
60% nitric acid produced any gray level discrimina- (UNS 46400) using Klemm’s I reagent. Selective
tion between the phases, and that was weak. All etching has a long historical record for identifica-
revealed the phase boundaries nicely, however.
Two electrolytic reagents, shown in Figures 33g
and h, are commonly used to color ferrite in dual
phase grades and delta ferrite in martensitic
grades. Of these, aqueous 20% sodium hydroxide,
Figure 33g, usually gives more uniform coloring of
the ferrite. Murakami’s and Groesbeck’s reagents
have also been used for this purpose. Tint etchants
have been developed by Beraha that color the fer-
rite phase nicely, as demonstrated in Figure 33i.
Figure 34. Beta phase colored in Naval Brass (Cu-
39.7% Zn-0.8% Sn) by immersion color etching with
Klemm’s I reagent (200X).

tion of intermetallic phases in aluminum alloys.

This method was used for many years before the
development of energy-dispersive spectroscopy.
Today, it is still useful for image analysis work. Fig-
ure 35 shows selective coloration of theta phase,
CuAl2, in the Al-33% Cu eutectic alloy. As a final
g example, Figure 36 illustrates the structure of a

h
Figure 35. Theta phase, CuAl2, colored in the α-Al/
CuAl2 eutectic in an as-cast Al-33% Cu specimen by
immersion color etching with the Lienard and Pacque
etch (200 mL water, 1 g ammonium molybdate, 6 g
ammonium chloride) at 1000X.

simple WC-Co sintered carbide, cutting tool. In

the as-polished condition, Figure 36a, the cobalt
binder can be seen faintly against the more gray-
ish tungsten carbide grains. A few particles of
i
graphite are visible. In Figure 36b, light relief pol-
Figure 33. Selective coloring of ferrite in a duplex
ishing has brought out the outlines of the cobalt
stainless steel by electrolytic etching with (g) binder phase, but this image is not particularly
aqueous 20% NaOH (4 V dc, 10 s) and with (h)
aqueous 10 N KOH, (3 V dc, 4 s); and by immersion useful for image analysis. Etching in a solution of
color etching (i) with Beraha’s reagent (100 mL water, hydrochloric acid saturated with ferric chloride,
10 mL HCl and 1 g K2S2O5) at 200X.
Figure 36c, attacks the cobalt and provides good
uniform contrast for measurement of the cobalt

70
Etching

ing this with Murakami’s reagent at room tem- Electrolytic Etching and Anodizing
perature reveals the edges of the tungsten carbide The procedure for electrolytic etching is basi-
grains, useful for evaluation of the WC grain size, cally the same as for electropolishing except that
Figure 36d. voltage and current densities are considerably
lower. The specimen is made the anode, and some
relatively insoluble but conductive material such
as stainless steel, graphite, or platinum is used for
the cathode. Direct current electrolysis is used
for most electrolytic etching. Electrolytic etching
is commonly used with stainless steels, either to
reveal grain boundaries without twin boundaries,
or for coloring ferrite (as illustrated) in Figure 33,
a
delta ferrite, sigma or chi phases. Anodizing is a
term applied to electrolytic etchants that develop
grain coloration when viewed with crossed-polar-
ized light, as in the case of aluminum, tantalum,
titanium, tungsten, uranium, vanadium and zir-
conium [2]. Figure 37 shows the grain structure
of 5754 aluminum revealed by anodizing with
Barker’s reagent, viewed with crossed-polarized
light. Again, color image analysis makes this
image useful now for grain size measurements.
b

c Figure 37. Grain structure of annealed 5754 aluminum

sheet revealed by anodizing with Barker’s reagent
(30 V dc, 2 min.) and viewing with polarized light plus
sensitive tint (100X).

Heat Tinting
Although not commonly utilized, heat tinting
[2] is an excellent method for obtaining color
contrast between constituents or grains. An
unmounted polished specimen is placed face up
d in an air-fired furnace at a set temperature and
Figure 36. Microstructure of WC-Co cutting tool: a) held there as an oxide film grows on the surface.
as-polished revealing graphite particles; b) relief
polished revealing the cobalt binder phase; c) after Interference effects, as in tint etching, create
immersion in Chaporova’s etch (HCl saturated with coloration for film thicknesses within a certain
FeCl3) to attack and “darken” the cobalt; and, d) after
(c) plus Murakami’s reagent to outline the WC grains range, about 30-500 nm. The observed color is a
(1000X).

71
Etching

function of the film thickness. Naturally, the ther- method. Suitable materials for the production
mal exposure cannot alter the microstructure. The of evaporation layers have been summarized in
correct temperature must be determined by the [22,23]. The technique is universally applicable,
trial-and-error approach, but the procedure is re- but does require a vacuum evaporator. Its main
producible and reliable. Figure 38 shows the grain weakness is difficulty in obtaining a uniformly
structure of CP titanium revealed by heat tinting. coated large surface area for image analysis mea-
surements.

Helpful Hints For

Etching
Many etchants can be used
by swabbing or by immersion.
Swabbing is preferred for those specimens that
form a tight protective oxide on the surface in
Figure 38. Grain structure of annealed CP titanium
air, such as Al, Ni, Cr, stainless steels, Nb (Cb), Ti
revealed by heat tinting on a hot plate (100X, and Zr. However, if the etchant forms a film, as
polarized light plus sensitive tint).
in tint etchants, then immersion must be used
as swabbing will keep the film from forming.
Interference Layer Method
Keller’s reagent reveals the grain size of certain
The interference layer method [2], introduced aluminum alloys by forming a film. This will not
by Pepperhoff in 1960, is another procedure for occur if the etch is used by swabbing.
obtaining a film over the microstructure that
generates color by interference effects. In this Many etchants, and their ingredients, do pres-
method, a suitable material is deposited on the ent potential health hazards to the user. ASTM E
polished specimen face by vapor deposition to 2014, Standard Guide on Metallography Labo-
produce a low-absorption, dielectric film with ratory Safety, describes many of the common
a high refractive index at a thickness within the problems and how to avoid them.
range for interference. Very small differences in
the natural reflectivity between constituents and
the matrix can be dramatically enhanced by this

72
Etching

Table 43. Commonly Used Etchants for Metals and Alloys3.

LIGHT METALS - Aluminum and Alloys

Composition Comments

1. 95 mL water Keller’s reagent, very popular general purpose reagent for Al and Al
2.5 mL HNO3 alloys, except high-Si alloys. Immerse sample 10-20 seconds, wash in
1.5 mL HCI warm water. Can follow with a dip in conc. HNO3. Outlines all common
1.0 mL HF constituents, reveals grain structure in certain alloys when used by
immersion.

2. 90-100 mL water General-purpose reagent. Attacks FeAl3, other constituents outlined. The
0.1-10 mL HF 0.5% concentration of HF is very popular.

3. 84 mL water Graff and Sargent’s etchant, for grain size of 2XXX, 3XXX, 6XXX, and
15.5 mL HNO3 7XXX wrought alloys. Immerse specimen 20-60 seconds with mild
0.5 mL HF agitation.
3 g CrO3

4. 1.8% fluoboric acid in water Barker’s anodizing method for grain structure. Use 0.5-1.5 A/in2, 30-45 V
dc. For most alloys and tempers, 20 seconds at 1 A/in2 and 30 V dc at 20 °C is sufficient. Stirring
not needed. Rinse in warm water, dry. Use polarized light; sensitive tint helpful.
Magnesium and Alloys
Composition Comments

5. 25 mL water Glycol etch, general purpose etch for pure Mg and alloys. Swab specimen
75 mL 3-5 ethylene glycol seconds for F and T6 temper alloys, 1-2 minutes for T4 and 0 temper
1 mL HNO3 alloys.

6. 19 mL water Acetic glycol etchant for pure Mg and alloys. Swab specimen 1-3 seconds
60 mL ethylene glycol for F and T6 temper alloys, 10 seconds for T4 and 0 temper alloys. Reveals
20 mL acetic acid grain boundaries in solution-treated castings and most wrought alloys.
1 mL HNO3
7. 100 mL ethanol For Mg and alloys. Use fresh. Immerse specimen for 15-30 seconds.
10 mL water Produces grain contrast.
5 g picric acid

LOW MELTING POINT METALS - Sb, Bi, Cd, Pb, Sn and Zn

Composition Comments

8. 100 mL water For Sb, Bi and alloys. Immerse specimen up to a few minutes.
30 mL HCI
2 g FeCI3
9. 100 mL water For Sb-Pb, Bi-Sn, Cd-Sn, Cd-Zn, and Bi-Cd alloys. Immerse specimen up
25 mL HCI to a few minutes.
8 g FeCI3
10. 95-99 mL ethanol For Cd, Cd alloys, Sn, Zn alloys, Pb and alloys, Bi-Sn eutectic alloy and Bi-
1-5 mL HNO3 Cd alloys. Can add a few drops of zephiran chloride. Immerse sample. For
Pb and alloys, if a stain forms, wash in 10% alcoholic HCI.

11. 100 mL water Pollack’s reagent for Pb and alloys. Immerse specimen 15-30 seconds.
10 g ammonium molybdate Other compositions used are: 100 mL: 9 g: 15 g and 100 mL: 10 g: 25 g.
10 g citric acid

12. 100 mL water For Sn-based babbitt metal. Immerse specimen up to 5 minutes.
2 mL HCI
10 g FeCI3
13. 200 mL water Palmerton reagent for pure Zn and alloys. Immerse specimen up to 3
40 g CrO3 minutes. Rinse in 20% aqueous CrO3.
3 g Na2SO4

14. 200 mL water Modified Palmerton reagent for Zn die-casting alloys. Immerse specimen
10 g CrO3 for several seconds, rinse in 20% aqueous CrO3.
1 g Na2SO4
3
When water is specified, always use distilled water. It is best to use reagent grade chemicals. Etchants can be quite dangerous and it is advisable to consult
with a chemist when dealing with new etchants that may be potentially dangerous. ASTM E 2014, Standard Guide on Metallographic Laboratory Safety, is
a valuable reference.

73
Etching

Table 43. Commonly Used Etchants for Metals and Alloys3.

REFRACTORY METALS: Ti, Zr, Hf, Cr, Mo, Re, Nb, Ta, W and V
Composition Comments
15. 100 mL water Kroll’s reagent for Ti alloys. Swab specimen 3-10 seconds or immerse
1-3 mL HF specimen 10-30 seconds.
2-6 mL HNO3
16. 200 mL water For Ti, Zr and alloys. Swab or immerse specimen. Higher concentrations
1 mL HF can be used but are prone to staining problems.
17. 30 mL lactic acid For Ti alloys. Swab specimen up to 30 seconds. Decomposes, do not store.
15 mL HNO3 Good for alpha-beta alloys.
30 mL HF
18. 30 mL HCI For Zr, Hf, and alloys. Swab specimen 3-10 seconds, or immerse specimen
15 mL HNO3 up to 120 seconds.
30 mL HF
19. 45 mL H2O (H2O2 or glycerol) Cain’s chemical polish and etch for Hf, Zr, and alloys. Can dilute aqueous
45 mL HNO3 solution with 3-5 parts water to stain the structure (swab specimen) after
8-10 mL HF chemical polishing. Chemically polish and etch specimen by swabbing 5-
20 seconds. Use polarized light.
20. 60 mL HCI Aqua regia. For Cr and alloys. Immerse or swab specimen up to 1 minute.
20 mL HNO3 Use under a hood with care, do not store.
21. 30 mL HCI Modified “Glyceregia”. For Cr and alloys. Immerse specimen up to a few
45 mL glycerol minutes.
15 mL HNO3
22. 100 mL water Murakami’s reagent. For Cr, Mo, Re, Ta-Mo, W, V and alloys. Use fresh, can
10 g KOH or NaOH immerse sample for up to 1 minute.
10 g K3Fe(CN)6
23. 70 mL water For Mo alloys. Immerse specimen 2 minutes. Wash with water and dry;
20 mL H2O2 (30%) immersion produces colors, swabbing produces grain-boundary etch.
10 mL H2SO4
24. 10-20 mL glycerol For Mo and Mo-Ti alloys. Immerse specimen for up to 5 minutes.
10 mL HNO3
10 mL HF
25. 100 mL water For Mo-Re alloys. Use at 20 °C by immersion.
5 g K3Fe(CN)6
2 g KOH
26. 50 mL acetic acid For Nb, Ta and alloys. Swab specimen 10-30 seconds.
20 mL HNO3
5 mL HF
27. 50 mL water DuPont Nb reagent. For Nb-Hf and Nb alloys.
14 mL H2SO4
5 mL HNO3
28. 50 mL water For Nb-Zr and Nb-Zr-Re alloys. Swab specimen.
50 mL HNO3
1 mL HF
29. 30 mL lactic acid For Re and W-Re alloys. Swab specimen.
10 mL HNO3
5 mL HF
30. 10 mL HF For V and alloys; grain-boundary etch for Ta alloys. Swab specimen. Equal
10 mL HNO3 parts used for Ta and high Ta alloys.
10-30 mL glycerol

IRON and STEEL

Composition Comments
31. 90-99 mL methanol or ethanol Nital. Most common etchant for Fe, carbon and alloy steels, cast iron.
1-10 mL HNO3 Reveals alpha grain boundaries and constituents. Excellent for martensitic
structures. The 2% solution is most common, 5-10% used for high alloy
steels (do not store). Use by immersion or swabbing of sample for up to
about 60 seconds.

74
Etching

Table 43. Commonly Used Etchants for Metals and Alloys3.

32. 100 mL ethanol Picral. Recommended for structures consisting of ferrite and carbide. Does
4 g picric acid not reveal ferrite grain boundaries. Addition of about 0.5-1% zephiran
chloride improves etch rate and uniformity.
33. 100 mL ethanol Vilella’s reagent. Good for ferrite-carbide structures. Produces grain
5 mL HCI contrast for estimating prior austenite grain size. Results best on
1 g picric acid martensite tempered at 572-932 °F (300-500 °C). Occasionally reveals
prior-austenite grain boundaries in high alloy steels. Outlines constituents
in stainless steels. Good for tool steels and martensitic stainless steels.
34. Saturated aqueous picric acid Bechet and Beaujard’s etch, most successful etchant for prior-austenite
solution grain plus small amount boundaries. Good for martensitic and bainitic steels. Many wetting agents
of a wetting agent have been used, sodium tridecylbenzene sulfonate is one of most
successful (the dodecyl version is easier to obtain and works as well). Use
at 20-100 °C. Swab or immerse sample for 2-60 minutes. Etch in ultrasonic cleaner (see ref.2, pg.
219-223). Additions of 0.5g CuCl2 per 100mL solution or about 1% HCI have been used for higher alloy
steels to produce etching. Room temperature etching most
common. Lightly back polish to remove surface smut.
35. 150 mL water Modified Fry’s reagent. Used for 18% Ni maraging steels, martensitic and
50 mL HCI PH stainless steels.
25 mL HNO3
1 g CuCl2
36. 100 mL water Alkaline sodium picrate. Best etch for McQuaid-Ehn carburized samples.
25 g NaOH Darkens cementite. Use boiling for 1-15 minutes or electrolytic at 6 V dc,
2 g picric acid 0.5 A/in2, 30-120 seconds. May reveal prior-austenite grain boundaries in
high carbon steels when no apparent grain boundary film is present.
37. 3 parts HCI “Glyceregia”. For austenitic stainless steels. Reveals grain structure,
2 parts glycerol outlines sigma and carbides. Mix fresh, do not store. Use by swabbing.
1 part HNO3
38. 100 mL ethanol Kalling’s no. 2 (“waterless” Kalling’s) etch for austenitic and duplex
100 mL HCI stainless steels. Ferrite attacked readily, carbides unattacked, austenite
5 g CuCl2 slightly attacked. Use at 20 °C by immersion or swabbing. Can be stored.
39. 15 mL HCI Acetic glyceregia. Mix fresh; do not store. Use for high alloy stainless
10 mL acetic acid steels.
5 mL HNO3
2 drops glycerol
40. 100 mL water Murakami’s reagent. Usually works better on ferritic stainless grades than
10 g K2Fe(CN)6 on austenitic grades. Use at 20 °C for 7-60 seconds: reveals carbides si
10 g KOH or NaOH sigma faintly attacked with etching up to 3 minutes. Use at 80°C (176°F) to
boiling for 2-60 minutes: carbides dark, sigma blue (not always attacked),
ferrite yellow to yellow-brown, austenite unattacked. Do not always get
uniform etching.
41. 100 mL water Use for stainless steels at 6 V dc. Carbides revealed by etching for
10 g oxalic acid 15-30 seconds, grain boundaries after 45-60 seconds, sigma
outlined after 6 seconds. 1-3 V also used. Dissolves carbides, sigma strongly attacked, austenite
moderately attacked, ferrite unattacked.
42. 100 mL water Used to color ferrite in martensitic, PH or dual-phase stainless steels.
20 g NaOH Use at 3-5 V dc, 20 °C, 5 seconds, stainless steel cathode. Ferrite outlined
and colored tan.
43. 40 mL water Electrolytic etch to reveal austenite boundaries but not twin boundaries in
60 mL HNO3 austenitic stainless steels (304, 316, etc.). Voltage is critical. Pt cathode
preferred to stainless steel. Use at 1.4 V dc, 2 minutes (see ref. 2, pgs. 235, 238 and 239).

COPPER, NICKEL and COBALT: Copper and Alloys

Composition Comments

44. 25 mL NH4OH General purpose grain contrast etch for Cu and alloys (produces a flat etch
25 mL water (optional) for some alloys). Use fresh, add peroxide last. Use under a hood. Swab
25-50 mL H2O2 (3%) specimen 5-45 seconds.
45. 100 mL water General purpose etch for Cu and alloys. Immerse or swab for 3-60
10 g ammonium persulfate seconds. Reveals grain boundaries but is sensitive to crystallographic
orientation.

75
Etching

Table 43. Commonly Used Etchants for Metals and Alloys3.

46. 100 mL water General purpose etch for Cu and alloys, particularly Cu-Be alloys.
3 g ammonium persulfate
1 mL NH4OH
47. 70 mL water Excellent general purpose etch, reveals grain boundaries well. Immerse
5 g Fe(NO3)3 specimen 10-30 seconds.
25 mL HCI

Nickel and Alloys

Composition Comments
48. 5 g FeCl3 Carapella’s etch for Ni and Ni-Cu (Monel) alloys. Use by immersion
2 mL HCI or swabbing.
99 mL ethanol
49. 40-80 mL ethanol Kalling’s no.2 etch (“waterless” Kalling’s) for Ni-Cu alloys and superalloys.
40 mL HCI Immerse or swab specimen up to a few minutes.
2 g CuCI2
50. 50 mL water Marble’s reagent for Ni, Ni-Cu, and Ni-Fe alloys and superalloys. Immerse
50 mL HCI or swab sample 5-60 seconds. Reveals grain structure of superalloys.
10 g CuSO4
51. 15 mL HCI “Glyceregia”, for superalloys and Ni-Cr alloys. Swab specimen for 5-60
10 mL glycerol seconds. Mix fresh. Do not store. Use under a hood.
5 mL HNO3
52. 60 mL glycerol Modified Glyceregia for superalloys. Reveals precipitates. Use under hood;
50 mL HCI do not store. Add HNO3 last. Discard when dark yellow. Immerse or swab
10 mL HNO3 specimen 10-60 seconds.

Cobalt and Alloys

Composition Comments
53. 60 mL HCI For Co and alloys. Mix fresh and age 1 hour before use. Immerse spe
15 mL water specimen for up to 30 seconds. Do not store.
15 mL acetic acid
15 mL HNO3
54. 200 mL ethanol General etch for Co and alloys. Immerse specimen 2-4 minutes.
7.5 mL HF
2.5 mL HNO3
55. 50 mL water Marble’s reagent, for Co high temperature alloys. Immerse or swab spe
50 mL HCI specimen for up to 1 minute.
10 g CuSO4
56. 80 mL lactic acid For Co alloys. Use by swabbing.
10 mL H2O2 (30%)
10 mL HNO3

PRECIOUS METALS: Au, Ag, Ir, Os, Pd, Pt, Rh and Ru

Composition Comments

57. 60 mL HCI For gold, silver, palladium and high noble metal alloys. Use under hood.
40 mL HNO3 Immerse specimen up to 60 seconds. Equal parts of each acid also used.

58. 60 mL HCI Aqua regia for pure gold, Pt and alloys, some Rh alloys. Use boiling for up
20 mL HNO3 to 30 minutes.

59. 1-5 g CrO3 For Au, Ag, Pd and alloys. Swab or immerse specimen up to 60 seconds.
100 mL HCI

60. 30 mL water For pure Pt. Use hot, immerse specimen up to 5 minutes.
25 mL HCI
5 mL HNO3

61. Conc. HCI For Rh and alloys. Use at 5 V ac, 1-2 minutes, graphite cathode,
Pt lead wires.

76
Etching

Table 43. Commonly Used Etchants for Metals and Alloys3.

62. Solution a For Ru. Mix 4 parts solution a to 1 part solution b, use 5-20 V ac, 1-2
100 mL water minutes, graphite cathode, Pt-lead wires.
40 g NaCI
Solution b
Conc. HCI
63. 50 mL NH4OH For pure Ag, Ag solders, and Ag-Pd alloys. Mix fresh. Swab specimen up to
20 mL H2O2 (3%) 60 seconds. Discard after etching. Use a 50:50 mix for sterling silver; for
fine silver, use 30% conc. hydrogen peroxide.
64. Solution a For gold alloys up to 18 karat. Mix equal amounts of a and b directly on
10 g NaCN the specimen using eye droppers. Swab and replenish the etchants
100 mL water until the desired etch level is obtained. If a brown stain forms, swab with
Solution b a to remove it.
H2O2 (30%)

SINTERED CARBIDES
Composition Comments
65. 100 mL water Murakami’s reagent, for WC-Co and complex sintered carbides. Immerse
10 g KOH or NaOH specimen seconds to minutes. Use 2-10 seconds to identify eta phase
10 g K3Fe(CN)6 (colored). Longer times attack eta. Reveals phase and grain boundaries.
Normally used at 20 °C.
66. 97 mL water For WC, MO2C, TiC, or Ni in sintered carbides. Use boiling for up to 60
3 mL H2O2 (30%) seconds. For coarse carbides or high Co content, use short etch time.
67. 15 mL water For WC, TiC, TaC, and Co in sintered carbides. Use at 20 °C for 5-30
30 mL HCI seconds.
15 mL HNO3
15 mL acetic acid
68. 100 mL H2O To darken Co (or Ni) binder phase. Mix fresh. Swab 10 seconds.
3 g FeCl3

CERAMICS and NITRIDES

Composition Comments
69. Phosphoric acid For alumina, Al2O3, and silicon nitride, Si3N4. Use by immersion at 250 °C
for up to a few minutes for Al2O3 or up to 15 minutes for Si3N4.
70. 100 mL water For magnesia, MgO. Use by immersion at 25-60 °C for several minutes.
15 mL HNO3
71. 100 mL water For titania, TiO2. Immerse up to a few minutes.
5 g NH4 FHF
4 mL HCI
72. 50 mL water For zirconia, ZrO2. Use by immersion in boiling solution for up to 5 minutes.
50 mL H2SO4
73. HCI For calcia, CaO, or MgO. Immerse for up to 6 minutes.
74. HF For Si3N4, BeO, BaO, MgO, ZrO2 and Zr2O3. Immerse for up to 6 minutes.

PLASTICS and POLYMERS

Composition Comments
75. 100 mL water For polypropylene (PP). Immerse for up to several hours at 70°C.
60 g CrO3
76. HNO3 For polyethylene (PE). Immerse for up to a few minutes.
77. Xylol Reveals spherolites in polyethene (PE). Immerse for up to a few days at
70 °C.
78. 70 mL water For polyoxymethylene (POM). Immerse up to 20 seconds.
30 mL HCI
79. Xylene For polyamid (PA) and polyethylene (PE). Use at 70 °C for 60 seconds.
For Nylon 6, use at 65-70 °C for 2-3 minutes. For Nylon 6, use at 75 °C for
3-4 minutes.

77
Light Optical Microscopy

LIGHT OPTICAL MICROSCOPy as nonmetallic inclusions, graphite, nitrides, cracks

or porosity, are best evaluated in the as‑polished
Metallurgical microscopes differ from biological condition as etching reveals other microstruc-
microscopes primarily in the manner by which tural features that may obscure such details. In
the specimen is illuminated due to the opacity of the as-polished condition, there is a minimum of
metals. Unlike biological microscopes, metallurgi- extra information for the observer to deal with
cal microscopes must use reflected light, rather which makes examination of these features most
than transmitted light. Figure 39 is a simplified efficient. Some materials do not require etching
light ray diagram of a metallurgical microscope. and, in a few cases, examination is less satisfactory
The prepared specimen is placed on the stage after etching.
with the surface perpendicular to the optical axis
of the microscope and is illuminated through the For most metals, etching must be used to
objective lens by light from a lamp or arc source. fully reveal the microstructure. A wide range of
This light is focused by the condenser lens into a processes are used, the most common being
beam that is made approximately parallel to the etching with either acidic or basic solutions. For
optical axis of the microscope by the half silvered most metals and alloys, there are a number of
mirror. The light then passes through the objec- general-purpose etchants that should be used
tive onto the specimen. It is then reflected from first to reveal the microstructure. Subsequent
the surface of the specimen, back through the examination may indicate that other more
objective, the half silvered mirror, and then to the specialized techniques for revealing the mi-
eyepiece to the observer’s eye, or to a camera port crostructure may be useful. For example, the
or a film plane. examination may reveal certain constituents
that the metallographer may wish to measure.
After the specimen has been properly sectioned Measurements, particularly those performed with
and polished, it is ready for examination. How- automatic image analyzers, will be simpler to per-
ever, in many cases, it is important to begin the form and more reliable if the desired constituent
examination with an as-polished specimen, can be revealed selectively, and numerous pro-
particularly for quality control or failure analysis cedures are available to produce selective phase
work. Certain microstructural constituents, such contrasting.

Figure 39. Schematic diagram showing the light path through an upright reflected light microscope operating with
bright field illumination.

78
Light Optical Microscopy

The Light Microscope by the late 1970s, these large metallographs had
When Henry Clifton Sorby made his first examina- become very expensive, too expensive for most
tion of the microstructure of iron on 28 July 1863, laboratories.
he was using a transmitted light petrographic mi-
croscope made by Smith, Beck and Beck of London In 1979, microscope manufacturers began to
that he had purchased in 1861. Although capable introduce very high quality, reasonably priced,
of magnifications up to 400X, most of his exami- compact metallographs. These microscopes
nations were conducted at 30, 60 or 100X and his can be obtained with a wide variety of objec-
first micrographs were produced at only 9X. The tive lens types and auxiliary accessories to meet
objective lenses of this microscope were equipped the metallographer’s needs. They are available
with Lieberkühns silvered concave reflectors for with universal vertical illuminators that permit
focusing light on opaque specimens. easy switching from one illumination mode to
another using the same set of objective lenses.
Sorby quickly realized that reflected light Furthermore, the manufacturers have introduced
produced with the Lieberkühns reflectors was new glass compositions and lens formulations,
inadequate and he developed two alternate generally by computer-aided design, for improved
methods for this purpose. Subsequently, others image contrast and brightness. Tungsten- hal-
developed vertical illuminators using prisms or ogen filament lamps have largely replaced
plane glass reflectors and Sorby’s systems were xenon arc lamps as the preferred light source.
not further developed. In 1886, Sorby reported on
the use of “very high power” (650X) for the study of Microscope Components
pearlite. This was accomplished using a 45º cover
Light Sources The amount of light lost during
glass vertical illuminator made for him by Beck.
passage from the source through a reflecting
type microscope is appreciable because of the
For many years, photomicroscopy was conducted
intricate path the light follows. For this reason, it
using specially built reflected microscopes known
is generally preferable that the intensity of the
as metallographs. These devices represented the
source be high, especially for photomicroscopy.
“top‑of‑the‑line” in metallurgical microscopes
Several types of light sources are used includ-
and were essential for quality work. In the late
ing tungsten-filament lamps, tungsten-halogen
1960’s and early 1970’s, manufacturers developed
lamps, quartz-halogen lamps, and xenon arc bulbs.
metallographs that were easier to use from the
Tungsten-filament lamps generally operate at low
eyepiece viewing position. The temperamental
voltage and high current. They are widely used for
carbon arc light sources were replaced by xenon
visual examination because of their low cost and
arc lamps. The unwieldy bellows systems for alter-
ease of operation.
ing magnification were replaced by zoom systems.
Vertical illuminators, previously using single posi-
Tungsten-halogen lamps are the most popular
tion objectives, were equipped with four to six
light source today due to their high light inten-
position rotatable nosepiece turrets to minimize
sity. They produce good color micrographs when
objective handling.The light path was deflected so
tungsten-corrected films are employed. Light
that the film plane was conveniently near at hand.
intensity can be varied easily to suit the view-
Universal type vertical illuminators and objective
ing conditions by adjusting a rheostat, a distinct
lenses were introduced so that the illumination
advantage over arc lamps.
mode could be readily switched from bright field
to darkfield, polarized light or differential interfer- Xenon arc lamps produce extremely high intensity
ence contrast. Such systems were very attractive in light, and their uniform spectra and daylight color
that separate vertical illuminators and objectives temperature makes them suitable for color pho-
were no longer needed for each illumination mode tomicrography. The first xenon lamps produced
and handling was eliminated. Exposure meters ozone, but modern units have overcome this
were also added at this time, chiefly as a result of problem. Light output is constant and can only be
the rising popularity of instant films where such reduced using neutral density filters.
devices are needed to minimize film wastage. But,

79
Light Optical Microscopy

Condenser. An adjustable lens free of spherical Objectives. The objective lens forms the pri-
aberration and coma is placed in front of the light mary image of the microstructure and is the most
source to focus the light at the desired point in the important component of the optical microscope.
optical path. A field diaphragm is placed in front The objective lens collects as much light as
of this lens to minimize internal glare and reflec- possible from the specimen and combines this
tions within the microscope. The field diaphragm is light to produce the image.The numerical aperture
stopped down to the edge of the field of view. (NA) of the objective, a measure of the light-col-
lecting ability of the lens, is defined as:
A second adjustable iris diaphragm, the aperture
diaphragm, is placed in the light path before
the vertical illuminator. Opening or closing this NA = n sin α
diaphragm alters the amount of light and the an-
gle of the cone of light entering the objective lens. where n is the minimum refraction index of the
The optimum setting for the aperture varies with material (air or oil) between the specimen and
each objective lens and is a compromise among the lens, and α is the half-angle of the most
image contrast, sharpness, and depth of field. As oblique light rays that enter the front lens of the
magnification increases, the aperture diaphragm objective. Light-collecting ability increases with
is stopped down. Opening this aperture increases α. The setting of the aperture diaphragm will
image sharpness, but reduces contrast; closing the alter the NA of the condenser and therefore the
aperture increases contrast, but impairs image NA of the system.
sharpness. The aperture diaphragm should not
be used for reducing light intensity. It should be The most commonly used objective is the
adjusted only for contrast and sharpness. achromat, which is corrected spherically for
one color (usually yellow green) and for lon-
Filters. Light filters are used to modify the light gitudinal chromatic aberration for two colors
for ease of observation, for improved photo- (usually red and green). Therefore, achromats are
microscopy, or to alter contrast. Neutral density fil- not suitable for color photomicroscopy, particular-
ters are used to reduce the light intensity uniformly ly at higher magnifications. Use of a yellow-green
across the visible spectrum.Various neutral density filter yields optimum results. However, achromats
filters, with a range of approximately 85 to 0.01% do provide a relatively long working distance, that
transmittance, are available. They are a necessity is, the distance from the front lens of the objective
when using an arc-lamp but virtually unnecessary to the specimen surface when in focus.The working
when using a filament lamp. distance decreases as the objective magnification
increases. Most manufacturers make long-working
Selective filters are used to balance the color distance objectives for special applications, for
temperature of the light source to that of the film. example, in hot stage microscopy. Achromats are
They may be needed for faithful reproduction of usually strain free, which is important for polarized
color images, depending on the light source used light examination. Because they contain fewer
and the film type. A green or yellow-green filter is lenses than other more highly corrected lenses,
widely used in black and white photography to internal reflection losses are minimized.
reduce the effect of lens defects on image quality.
Most objectives, particularly the lower cost achro- Semiapochromatic or fluorite objectives provide a
mats, require such filtering for best results. higher degree of correction of spherical and chro-
matic aberration. Therefore, they produce higher
Polarizing filters are used to produce plane quality color images than achromats. The apo-
polarized light (one filter) or crossed-polarized chromatic objectives have the highest degree of
light (two filters rotated 90° to each other to pro- correction, produce the best results, and are more
duce extinction) for examinations of anisotropic expensive. Plano objectives have extensive correc-
noncubic (crystallographic) materials. tion for flatness of field, which reduces eyestrain,
and are usually found on modern microscopes.

80
Light Optical Microscopy

With parfocal lens systems, each objective on the Eye clearance is the distance between the eye
nosepiece turret will be nearly in focus when the lens of the ocular and the eye. For most eyepieces,
turret is rotated, preventing the objective front the eye clearance is 10 mm or less – inadequate
lens from striking the specimen when lenses are if the microscopist wears glasses. Simple vision
switched. Many objectives also are spring loaded, problems, such as near sightedness, can be ac-
which helps prevent damage to the lens. This commodated using the fine focus adjustment.
is more of a problem with high magnification The microscope cannot correct vision problems
objectives, because the working distance can be such as astigmatism, and glasses must be worn.
very small. High eyepoint eyepieces are available to provide
an eye clearance of approximately 20 mm, neces-
Certain objectives are designed for use with oil sary for eyeglasses.
between the specimen and the front lens of the
objective. However, oil-immersion lenses are rarely Eyepieces are commonly equipped with various
used in metallography, because the specimen and reticles or graticules for locating, measuring,
lens must be cleaned after use. But, they do provide counting or comparing microstructures. The
higher resolution than can be achieved when air is eyepiece enlarges the reticle or graticule image
between the lens and specimen. In the latter case, and the primary image. Both images must be in
the maximum possible NA is 0.95; oil-immersion focus simultaneously. Special eyepieces are also
lenses produce a 1.3 to 1.45 NA, depending on produced to permit more accurate measure-
the lens and the oil-immersion used. Objective ments than can be made with a graticule scale.
magnifications from about 25 to 200X are avail- Examples are the filar-micrometer ocular or screw-
able, depending upon the manufacturer. Use of micrometer ocular. Such devices can be automated
oil also improves image contrast, which is valuable to produce a direct digital readout of the measure-
when examining low reflectivity specimens, such ment, which is accurate to approximately 1-µm.
as coal or ceramics.
A 10X magnification eyepiece is usually used;
Eyepieces. The eyepiece, or ocular, magnifies to obtain standard magnifications, some sys-
the primary image produced by the objective; tems require other magnifications, such as 6.3X.
the eye can then use the full resolution capabil- Higher power eyepieces, such as 12, 15, 20, or
ity of the objective. The microscope produces 25X, are also useful in certain situations. The
a virtual image of the specimen at the point of overall magnification is found by multiplying
most distinct vision, generally 250 mm (10-inch) the objective magnification, Mo, by the eyepiece
from the eye. The eyepiece magnifies this image, magnification, Me. If a zoom system or bellows
permitting achievement of useful magnifications. is also used, the magnification should be altered
The standard eyepiece has a 24 mm diameter field accordingly.
of view; wide field eyepieces for plano-type objec-
tives have a 30 mm diameter field of view, which Stage. A mechanical stage is provided for focus-
increases the usable area of the primary image. ing and moving the specimen, which is placed
Today, the wide field plano-objective is standard on the stage and secured using clips. The stage
on nearly all metallurgical microscopes. of an inverted microscope has replaceable
center stage plates with different size holes.
The simplest eyepiece is the Huygenian, which is The polished surface is placed over the hole for
satisfactory for use with low and medium power viewing. However, the entire surface cannot be
achromat objectives. Compensating eyepieces viewed, unless the specimen is smaller than the
are used with high NA achromats and the more hole and it is mounted. At high magnifications
highly corrected objectives. Because some lens it may not be possible to focus the objective
corrections are performed using these eyepieces, near the edge of the hole due to the restricted
the eyepiece must be matched with the type working distance.
of objective used. The newer, infinity-corrected
microscopes do not perform corrections in
the eyepieces, but in the tube lens. Eyepieces,
therefore, are simpler in infinity-corrected
microscopes.
81
Light Optical Microscopy

polished surface must be perpendicular to the Resolution

light beam, clay is placed between the specimen To see microstructural detail, the optical system
bottom and the slide. A piece of lens tissue is must produce adequate resolution, or resolving
placed over the polished surface, and the speci- power, and adequate image contrast. If resolu-
men is pressed into the clay using a leveling press. tion is acceptable but contrast is lacking, detail
However, pieces of tissue may adhere to the speci- cannot be observed. In general, the ability to
men surface. An alternative, particularly useful resolve two points or lines separated by dis-
with mounted specimens, is to use a ring instead tance d is a function of the wavelength, λ, of the
of tissue to flatten the specimen. Aluminum or incident light and the numerical aperture, NA, of
stainless steel ring forms of the same size as the that objective.
mounts (flattened slightly in a vise) will seat on the kλ
mount rather than the specimen. d=
NA
The upright microscope allows viewing of where k is 0.5 or 0.61. Figure 40 illustrates
the entire surface with any objective, and the this relationship for k = 0.61 and four light
operator can see which section of the specimen is wavelengths. Other formulas have also been
being viewed – a useful feature when examining reported.This equation does not include other fac-
specific areas on coated specimens, welds, and tors that influence resolution, such as the degree of
other specimens where specific areas are to be correction of the objectives and the visual acuity of
examined. For mounted specimens, an autolevel- the microscopist. It was based on the work of Abbe
ing stage holder can eliminate leveling specimens under conditions not present in metallography,
with clay. such as self-luminous points, perfect black-white
contrast, transmitted light examination, an ideal
The stage must be rigid to eliminate vibrations. point-light source, and absence of lens defects.
Stage movement, controlled by x and y microm-
eters, must be smooth and precise; rack and pinion Using the above equation, the limit of resolution
gearing is normally used. Many stages have scales for an objective with an NA of 1.3 is approximately
for measuring distances in the x and y directions. 0.2-µm. To see lines or points spaced 0.2-µm apart,
The focusing controls often contain rulings for the required magnification is determined by di-
estimating vertical movement. Some units have viding by the resolving power of the human eye,
motorized stages and focus controls. which is difficult to determine under observation
conditions. Abbe used a value of 0.3 mm at a dis-
A circular, rotatable stage plate may facilitate polar- tance of 250 mm – the distance from the eye for
ized light examination. Such stages, common for optimum vision. This gives 1500x. For light with a
mineralogical or petrographic studies, are gradu- mean wavelength of 0.55-µm, the required magni-
ated to permit measuring the angle of rotation. A fication is 1100 times the NA of the objective. This
rectilinear stage is generally placed on top of the is the origin of the 1000·NA rule for the maximum
circular stage. useful magnification. Any magnification above
1000·NA is termed “empty”, or useless.
Stand. Bench microscopes require a rigid stand,
particularly if photomicroscopy is performed on Strict adherence to the 1000·NA rule should be
the unit. The various pieces of the microscope questioned, considering the conditions under
are attached to the stand when assembled. which it was developed, certainly far different from
In some cases, the bench microscope is placed on those encountered in metallography. According to
a separate stand that also holds the photographic the Abbe analysis, for a microscopist with optimum
system. 20/20 vision and for optimum contrast conditions
and a mean light wavelength of 550 nm, the lowest
magnification that takes full advantage of the NA
of the objective is 500 times the NA.

82
Light Optical Microscopy

RESOLUTION CHART

Sodium (589 nm)

Green (546 nm)

Ultraviolet
(365 nm)

Blue (436 nm)

Figure 40. Influence of objective numerical aperture and light wavelength on the resolution of the light microscope.

This establishes a useful minimum magnification when longer wavelength light is used, as shown
to use with a given objective. It has been suggested in Figure 41.
that the upper limit of useful magnification for the
average microscopist is 2200·NA, not 1000·NA.
Imaging Modes
Most microscopical studies of metals are made
Depth of Field using bright field illumination. In addition to this
Depth of field is the distance along the optical type of illumination, several special techniques
axis over which image details are observed with (oblique illumination, dark field illumination,
acceptable clarity. Those factors that influence differential interference contrast microscopy
resolution also affect depth of field, but in the and polarized light microscopy) have particular
opposite direction. Therefore, a compromise must applications for metallographic studies.
be reached between these two parameters, which
is more difficult as magnification increases. This Nearly all microscopes using reflected or trans-
is one reason light etching is preferred for high- mitted light employ Köhler illumination because
magnification examination. The depth of field, df, it provides the most intense, most even illumina-
can be estimated from: tion possible with standard light sources. The
reflected light microscope has two adjustable
λ (n 2 −NA2)½ diaphragms: the aperture diaphragm and the field
df =
NA2 diaphragm, located between the lamp housing
and the objective. Both are adjusted to improve
where n is the refractive index of the medium illumination and the image. To obtain Köhler illu-
between the specimen and the objective (n ~ mination, the image of the field diaphragm must
1.0 for air), λ is the wavelength of light, and NA is be brought into focus on the specimen plane.
the numerical aperture. This equation shows that This normally occurs automatically when the
depth of field increases as the NA decreases and

83
Light Optical Microscopy

DEPTH OF FIELD CHART

Sodium (589 nm)

Green (546 nm)

Blue (436 nm)

Index of Refraction = 1.0

Figure 41. Influence of the objective numerical aperture and light wavelength on the depth of field of the light
microscope.

microstructural image is brought into focus. Oblique Illumination. The surface relief of a
The filament image must also be focused on metallographic specimen can be revealed using
the aperture diaphragm plane. This produces oblique illumination. This involves offsetting the
uniform illumination of the specimen imaged at condenser lens system or, as is more usually done,
the intermediate image plane and magnified by moving the condenser aperture to a position
the eyepiece. slightly off the optical axis. Although it should
be possible to continually increase the contrast
Bright Field. In bright field illumination, the achieved by oblique illumination by moving the
surface of the specimen is normal to the optical condenser farther and farther from the optical axis,
axis of the microscope, and white light is used. A the numerical aperture of a lens is reduced when
ray diagram for bright-field illumination is illus-
trated in Figure 42. Light that passes through the
objective and strikes a region of the speci-
men surface that is perpendicular to the beam
will be reflected back up the objective through
the eyepieces to the eyes where it will appear
to be bright or white. Light that strikes grain
boundaries, phase boundaries, and other
features not perpendicular to the optical axis will
be scattered at an angle and will not be collected
by the objective. These regions will appear to
be dark or black in the image. Bright field is
the most common mode of illumination used
by metallographers.
Figure 42. Schematic diagram showing the light path
in bright field illumination.

84
Light Optical Microscopy

this happens because only a portion of the lens and placing an analyzer (another Polaroid filter)
is used. For this reason, there is a practical limit before the eyepiece, as illustrated in Figure 44.
to the amount of contrast that can be achieved. The polarizer produces plane polarized light that
Illumination also becomes uneven as the degree strikes the surface and is reflected through the
of “obliqueness” increases. Since differential in- analyzer to the eyepieces. If an anisotropic metal
terference contrast systems have been available, is examined with the analyzer set 90° to the polar-
oblique illumination is rarely offered as an option izer, the grain structure will be visible. However,
on new microscopes. viewing of an isotropic metal (cubic metals) under
such conditions will produce a dark,“extinguished”
Dark Field. Another method that can be used condition (complete darkness is not possible using
to distinguish features not in the plane of the Polaroid filters). Polarized light is particularly useful
polished and etched surface of a metallographic in metallography for revealing grain structure and
specimen is dark field (also called dark ground) twinning in anisotropic metals and alloys (see the
illumination. This type of illumination (see appendix for crystal structure information) and for
Figure 43 for a ray diagram) gives contrast identifying anisotropic phases and inclusions.
completely reversed from that obtained with
bright-field illumination — the features that are
light in bright field will be dark in dark field, and
those that are dark in bright field will be light,
appearing to be self luminous in dark field. This
highlighting of angled surfaces (pits, cracks, or
etched grain boundaries) allows more positive
identification of their nature than can be derived
from a black image with bright field illumination.
Due to the high image contrast obtained and the
brightness associated with features at an angle to
the optical axis, it is often possible to see details
not observed with bright field illumination.

Figure 44. Schematic diagram showing the light

path in polarized light with an optional lambda plate
(sensitive tint plate).

Differential Interference Contrast (DIC).

When crossed polarized light is used along with
a double quartz prism (Wollaston prism) placed
between the objective and the vertical illuminator,
Figure 45, two light beams are produced which
exhibit coherent interference in the image plane.
This leads to two slightly displaced (laterally)
images differing in phase (λ/2) that produces
Figure 43. Schematic diagram showing the light path height contrast.The image produced reveals topo-
in dark field illumination.
graphic detail somewhat similar to that produced
by oblique illumination but without the loss of
Polarized Light. Because many metals and resolution. Images can be viewed with natural
metallic and nonmetallic phases are optically colors similar to those observed in bright field, or
anisotropic, polarized light is particularly useful artificial coloring can be introduced by adding a
in metallography. Polarized light is obtained by sensitive tint plate.
placing a polarizer (usually a Polaroid filter) in
front of the condenser lens of the microscope

85
Light Optical Microscopy

more detail than in bright field. However, as the

martensite is noncubic in crystal structure, it re-
sponds well to crossed- polarized light, Figure 46b.
Of course, not every material can be examined us-
ing all four illumination modes, but there are many
cases where two or more modes can be effectively
utilized to reveal more information than given by
bright field alone.

Figure 45. Schematic diagram showing the light c

path for differential interference contrast (DIC)
illumination.

structure of an aluminum bronze specimen that

was water quenched from the beta field forming
martensite. The specimen was prepared but not
etched. A minor amount of relief was introduced
in final polishing. Figure 46a shows the surface in
bright field illumination. Due to the relief present,
a faint image of the structure is visible. Dark field,
d
Figure 46c, and differential interference contrast,
Figure 46d, make use of this relief to reveal much Figure 46. Martensitic microstructure of heat
treated, eutectoid aluminum bronze (Cu–11.8% Al)
in the unetched condition viewed using: c) dark field
illumination; and, d) differential interference contrast
illumination (200X).

Helpful Hints
For Light Optical
Microscopy
It can be difficult to determine
a
which objective you are using with an inverted
microscope due to the limited view from
above, as the magnification value may not be
easily observed. However, each objective has a
color-coded ring on it that is usually visible.The
colors and corresponding magnifications are:
red – 5X, yellow – 10X, green – 20X, blue – 50X,
and white – 100X.

b
Figure 46. Martensitic microstructure of heat treated,
eutectoid aluminum bronze (Cu–11.8% Al) in the
unetched condition viewed using: a) bright field
illumination, b) polarized light

86
Microindentation Hardness Testing

MICROINDENTATION decision as it made the Vickers test very easy

HARDNESS TESTING to adopt. The ideal d/D ratio (d = impression
diameter, D = ball diameter) for a spherical
Microindentation hardness testing, more com- indenter is 0.375. If tangents are drawn to the
monly (but incorrectly) called microhardness ball at the impression edges for d/D = 0.375,
testing, is widely used to study fine scale changes they meet below the center of the impression
in hardness, either intentional or accidental. Heat at an angle of 136°, the angle chosen for the Vick-
treaters have utilized the technique for many years ers indenter.
to evaluate the success of surface hardening treat-
ments or to detect and assess decarburization.
Metallographers and failure analysts use the meth-
od for a host of purposes including evaluation of
136°
homogeneity, characterization of weldments, as an
aid to phase identification, or simply to determine 136
°
the hardness of specimens too small for traditional
bulk indentation tests.

Metallurgists and metallographers tend to

develop their own jargon, often as a matter of
linguistic simplicity, which is not always as rigor-
ously correct as it could be. Although the term
“microhardness” is generally understood by
its users, the word implies that the hardness is
extremely low, which is not the case. The ap-
plied load and the resulting indent size are small Figure 47. Schematic of the Vickers indenter and the
shape of an impression.
relative to bulk tests, but the same hardness
number is obtained. Consequently, ASTM Com- Use of diamond allowed the Vickers test to be used
mittee E-4 on Metallography recommends use to evaluate any material (except diamond) and,
of the term “microindentation hardness testing” furthermore, had the very important advantage
which could be given the acronym MHT. ASTM of placing the hardness of all materials on one
Standard E 384 fully describes the two most com- continuous scale. This is a major disadvantage
mon microindentation tests - the Vickers and the of Rockwell type tests where different scales (15
Knoop tests. standard and 15 superficial) were developed to
evaluate materials. Not one of these scales can
The Vickers Test cover the full hardness range. The HRA scale
In 1925, Smith and Sandland of the UK devel- covers the broadest hardness range, but it is not
oped a new indentation test for metals that were commonly used.
too hard to evaluate using the Brinell test. The
In the Vickers test, the load is applied smoothly,
hardened steel ball of the Brinell test limited
without impact, forcing the indenter into the
the test to steels with hardnesses below ~450
test piece. The indenter is held in place for 10 or
HBS (~48 HRC). (The harder tungsten carbide
15 seconds. The physical quality of the indenter
ball was not available in 1925. The WC indenter
and the accuracy of the applied load (defined
extends the Brinell test to metals up to 615
in E 384) must be controlled in order to get the
HBW (~58 HRC). The WC ball has now replaced
correct results. After the load is removed, the two
the steel ball for the Brinell test.) In designing
impression diagonals are measured, usually to
the new indenter, a square-based diamond
the nearest 0.1-µm with a filar micrometer, and
pyramid (see Figure 47), they chose a geometry
averaged. The Vickers hardness (HV) is calculated
that would produce hardness numbers nearly
using:
identical to Brinell numbers in the range where
both tests could be used. This was a very wise 1854.4L
HV =
d2

87
Microindentation Hardness Testing

where the load L is in gf and the average diagonal area of the indent rather than the projected area.
d is in µm (this produces hardness number units If the impression shape is distorted due to elastic
of gf/µm2 although the equivalent units kgf/mm2 recovery (very common in anisotropic materials),
are preferred; in practice the numbers are reported Figure 49, should the hardness be based on the
without indication of the units). The original Vick- average of the two diagonals? It is possible to cal-
ers testers were developed for test loads of 1 to culate the Vickers hardness based on the projected
120kgf that produce rather large indents. Recog- area of the impression, which can be measured
nizing the need for lower test loads, the National by image analysis. While rigorous studies of this
Physical Laboratory (UK) reported on use of lower problem are scant in the literature, the diagonal
test loads in 1932. Lips and Sack developed the first measurement is the preferred approach even for
low-load Vickers tester in 1936. distorted indents, at this time.

Because the shape of the Vickers indentation The Knoop Test

is geometrically similar at all test loads, the HV As an alternative to the Vickers test, particularly
value is constant, within statistical precision, over to test very thin layers, Frederick Knoop and his
a very wide test load range as long as the test associates at the former National Bureau of
specimen is reasonably homogeneous. Numerous Standards developed a low-load test using a rhom-
studies of microindentation hardness test results bohedral-shaped diamond indenter, Figure 50.
conducted over a wide range of test loads have
shown that test results are not constant at very
low loads. This problem, called the “indentation
size effect”, or ISE, has been attributed to funda-
mental characteristics of the material. However, the
same effect is observed at the low load test range
(1-10kgf ) of bulk Vickers testers [2] and an ASTM
interlaboratory “round robin” of indents made by
one laboratory but measured by twelve different
people, reported all three possible ISE responses
[24,25] for the same indents!

Since the 1960s, the standard symbol for Vickers

hardness per ASTM E 92 and E 384, has been HV.
This should be used in preference to the older,
obsolete symbols DPN or VPN. The hardness is
expressed in a standard format. For example, if
a 300 gf load is used and the test reveals a hard-
ness of 375 HV, the hardness is expressed as 375
HV300. Rigorous application of the SI system results Figure 48. (top) Example of a well-formed Vickers
in hardness units expressed not in the standard, indentation (400X). Figure 49. (bottom) Example of a
distorted Vickers indentation (400X).
understandable kgf/mm2 values but in GPa units
that are meaningless to most engineers and The long diagonal is seven times (7.114 actually)
technicians. ASTM recommends a ‘soft” metric as long as the short diagonal. With this indenter
approach in this case. shape, elastic recovery can be held to a minimum.
Some investigators claim there is no elastic re-
In the Vickers test, it is assumed that elastic
covery with the Knoop indent, but this cannot be
recovery does not occur once the load is
true as measurements of the ratio of long to short
removed. However, elastic recovery does occur,
diagonal often reveal results substantially different
and sometimes its influence is quite pronounced.
than the ideal 7.114 value.
Generally, the impression (Figure 48) appears to
be square, and the two diagonals have similar
lengths. As with the Brinell test, the Vickers hard-
ness number is calculated based on the surface

88
Microindentation Hardness Testing

unaffected (both halves of this diagonal of the

same approximate length).

The down side of the Knoop indent is that the

172° 30´ three dimensional indent shape will change with
test load and, consequently, HK varies with load.
0°
At high loads, this variation is not substantial.
13
Conversion of HK values to other test scales can
only be done reliably for HK values performed at
the standard load, generally 500gf, used to develop
the correlations. All hardness scale conversions
are based on empirical data. Conversions are not
precise but are estimates.

Factors Affecting Accuracy, Precision

Figure 50. Schematic of the Knoop indenter and the and Bias
shape of an impression
Many factors (see Table 44) can influence the
The Knoop test is conducted in the same man- quality of microindentation test results [26]. In the
ner, and with the same tester, as the Vickers test. early days of low-load (< 100gf ) hardness testing,
However, only the long diagonal is measured. This, it was quickly recognized that improper specimen
of course, saves some time. The Knoop hardness preparation may influence hardness test results.
is calculated from Most text books state that improper preparation
will yield higher test results because the surface
HK = 14229L contains excessive preparation induced deforma-
d2 tion. While this is certainly true, there are other
cases where improper preparation can create
where the load L is in gf and the long diagonal d excessive heat that will lower the hardness and
is in µm. Again, the symbol HK was adopted in the strength of many metals and alloys. So either prob-
early 1960’s while other terms; e.g., HKN or KHN, lem may be encountered due to faulty preparation.
are obsolete and should not be used. The Knoop For many years, it was considered necessary to
hardness is expressed in the same manner as the electrolytically polish specimens so that the pre-
Vickers hardness; i.e., 375 HK300 means that a 300 paration-induced damage could be removed thus
gf load produced a Knoop hardness of 375. (The permitting bias-free low-load testing. However, the
kgf/mm2 unit information is no longer reported). science behind mechanical specimen preparation,
chiefly due to the work of Len Samuels [3], has led
Aside from a minor savings of time, one chief merit to development of excellent mechanical specimen
of the Knoop test is the ability to test thin layers preparation procedures, and electropolishing is no
more easily. For surfaces with varying hardness, longer required.
such as case hardened parts, Knoop indents can
be spaced closer together than Vickers indents. There are several operational factors that must
Thus, a single Knoop traverse can define a hard- be controlled in order to obtain optimum test
ness gradient more simply than a series of two or results. First, it is a good practice to inspect the
three parallel Vickers traverses where each indent indenter periodically for damage, for example,
is made at different depths. Furthermore, if the cracking or chipping of the diamond. If you have
hardness varies strongly with the depth, the Vick- metrology equipment, you can measure the face
ers indent will be distorted by this change; that is, angles and the sharpness of the tip. Specifications
the diagonal parallel to the hardness change will for Vickers and Knoop indenter geometries are
be affected by the hardness gradient (i.e., there given in E 384.
is a substantial difference in the lengths of the
two halves of the diagonal), while the diagonal
perpendicular to the hardness gradient will be

89
Microindentation Hardness Testing

Table 44. Factors Affecting Precision and Bias in Microindentation Hardness Testing

Instrument Factors Measurement Factors Material Factors

Accuracy of the applied load Calibration of the measurement Heterogenity of the
system specimen
Inertia effects, Numerical aperture Strength of crystallographic
speed of loading of the objective texture, if present
Lateral movement of the Magnification Quality of specimen preparation
indenter or specimen
Indentation time Inadequate image quality Low reflectivity or transparency
Indenter shap deviations Uniformity of illumination Creep during indentation
Damage to the indenter Distortion in optics Fracture during indentation
Inadequate spacing between Operator’s visual acuity Oil, grease or dirt on indenter
indents or form edges indenter or specimen
Angle of indentation Focusing of the image

A prime source of error in the tests is the align- Because of the method of defining HV and HK
ment of the specimen surface relative to the (equations given above) where we divide by d2,
indenter. The indenter itself must be properly measurement errors become more critical as d get
aligned perpendicular (±1°) to the stage plate. smaller; that is, as L decreases and the material’s
Next, the specimen surface must be perpendicu- hardness increases (discussed later). So departure
lar to the indenter. Most testers provide holders from a constant hardness for the Vickers or Knoop
that align the polished face perpendicular to tests as a function of load will be a greater prob-
the indenter (parallel to the stage). If a specimen lem as the hardness increases. For the Knoop test,
is simply placed on the stage surface, its back HK increases as L decreases because the indent
surface must be parallel to its polished surface. Tilt- geometry changes with indent depth and width.
ing the surface more than 1° from perpendicular The degree of the change in HK also varies with
results in nonsymmetrical impressions and can test load being greater as L decreases.
produce lateral movement between specimen
and indenter. The occurrence of non- symmetri- The greatest source of error is in measuring the
cal indents is generally easily detected during indent as has been documented in an ASTM in-
measurement. terlaboratory test [24,25]. Place the indent in the
center of the measuring field, if it is not already
In most cases, errors in indenting with modern there, as lens image quality is best in the center.
testers are not the major source of error, although The light source should provide adequate, even
this can occur [26]. It is important to check the illumination to provide maximum contrast and
performance of your tester regularly using a resolution. The accuracy of the filar micrometer, or
certified test block. It is safest to use a test block other measuring device, should be verified using
manufactured for microindentation testing and a stage micrometer.
certified for the test (Vickers or Knoop) and the
load that you intend to use. Strictly speaking, Specimen preparation quality becomes more
a block certified for Vickers testing at 300 or important as the load decreases, and it must be
500 gf (commonly chosen loads) should yield es- at an acceptable level. Specimen thickness must
sentially the same hardness with loads from about be at least 2.5 times the Vickers diagonal length.
50 to 1000 gf. That is, if you take the average of Because the Knoop indent is shallower than
about five indents and compare the average at the Vickers at the same load, somewhat thinner
your load to the average at the calibrated load specimens can be tested. Spacing of indents is
(knowing the standard deviation of the test re- important because indenting produces plastic
sults), statistical tests can tell you (at any desired deformation and a strain field around the indent. If
confidence level) if the difference between the
mean values of the tests at the two loads is statisti-
cally significant or not.

90
Microindentation Hardness Testing

the spacing is too small, the new indent will be

affected by the strain field around the last indent.
ASTM recommends a minimum spacing (center
to edge of adjacent indent) of 2.5 times the Vick-
ers diagonal. For the Knoop test where the long
diagonals are parallel, the spacing is 2.5 times
the short diagonal. The minimum recommended
spacing between the edge of the specimen and
the center of the indent should be 2.5 times. Again,
Knoop indents can be placed closer to the surface
than Vickers indents.

Several studies have stated that the resolution

of the optical system is the major factor limit-
Figure 51. OmniMet MHT
ing measurement precision. They state that this
causes indents to be undersized by a constant with a mouse). Curved patterns may be used, not
amount. However, as undersizing would in- just straight-line patterns. The system then makes
crease the measured hardness, not decrease it, the indents at the requested load and locations,
resolution limits per se cannot explain this measures each indent, calculates the hardness
problem. For the Knoop indent, the chief (and desired conversions to other scales), and
problem is image contrast at the indent tips prints/plots the results. Statistical analysis can also
that results in undersized indents. This problem, be preformed.
plus the variable indent shape, both result in
increasing HK with decreasing test load. For the In general, Vickers indents are easier to mea-
Vickers test, one would assume that undersizing or sure by image analysis than Knoop indents due
oversizing would be equally likely; but, experience to the lower contrast at the tips of the Knoop
suggests that oversizing is much more commonly indents (leads to undersizing and higher HK
encountered for low loads and small indents. values). Metal flow (plastic deformation) around
the indent edges can interfere with measurement
precision. Vickers indents, like the ones shown in
Automation Figure 48, exhibit excellent contrast and shape
Microindentation hardness testing is tedious, so and are easily measured. If the magnification
anything that can be done to simplify testing is too high (and this may be influenced by the
is valuable, especially for laboratories that do numerical aperture of the objective) for a given
substantial testing. Many adjuncts to indent mea- indent size, image contrast suffers and correct
surement have been tried and a variety of such detection of the indent will be very difficult. On the
systems are available.There has been considerable other hand, if the indent is very small on the screen,
interest in applying image analyzers to the indent it will be hard for the system to detect it automati-
measurement task. Further, with stage automation, cally. In this case, use a higher magnification, or use
it is possible to automate the indenting process the semiautomatic measurement mode if this is
itself, and with the same equipment. Figure 51 not possible. The operator fits a box around the
shows the Buehler OmniMet MHT automated indent with a mouse to measure the indent.
microindentation hardness testing system. This
system can be used in the fully automatic, semi- Microindentation hardness testing is a very
automatic or manual mode, depending upon the valuable tool for the materials engineer but it
nature of the testing. must be used with care and a full understanding of
the potential problems that can occur. Always try
An automated system can be programmed to to use the highest possible load depending upon
make any number of indents in either a defined the indent spacing or closeness to edges. Try to
pattern (x number of equally spaced indents, keep indents greater than 20-µm in diameter, as
or x number between two chosen points) or a recommended in ASTM standard E 384. If you have
random pattern (locations selected at random

91
Microindentation Hardness Testing

since you read it (all ASTM standards must be re-

viewed and revised, if necessary, every five years), Helpful Hints For
get the latest copy and go over it. The Precision Microindentation
and Bias section (and appendix X1 of E 384) con- Hardness Testing
tains a wealth of practical advice. Automation of If you suspect that the test
indentation and measurement is possible and, for results obtained on a specimen are question-
laboratories that do substantial testing, greatly able, verify the tester using a certified test
reduces operator fatigue while reducing testing block. When testing thin coatings, use the
times. Knoop indenter with the long axis parallel
to the surface. If you have a specimen where
the hardness changes rapidly, Vickers indents
may be distorted badly in the direction of the
hardness gradient. Switch to a Knoop indenter
and place the long axis perpendicular to the
hardness gradient.

92
Image Capture and Analysis

Image capture and and multi-layer focusing as well as a powerful

analysis database for tracking projects and measurements.
The OmniMet Express and OmniMet Enterprise
The progress in computer and video technology products include automated measurements in
has created a movement toward electronic image addition to the operator interactive measurements
acquisition. These images can be used in software found in the other products. OmniMet Express has
applications such as word processing or desktop specific application modules while the OmniMet
publishing programs allowing for fast report Enterprise product enables the user to tackle any
generation and electronic distribution. In addition, application.
associated data can be stored with the images in a
database allowing for easy retrieval of information
through searches.

When adding imaging capabilities to the labo-

ratory, it is important to consider your goals.
Oftentimes, images are just another step in the
documentation process. For example, in a failure
investigation it is useful to capture the image of a
complete component before the sectioning pro-
cess. Or, an image of the microstructure might be
attached to a report as an indication of a pass/fail
Figure 52. OmniMet Digital Imaging System
condition. Additional functionality of an imag-
ing system might include a scale marker overlay
and point-to-point or other operator interactive Acquisition
measurements. Fully automated image analysis Image capture is a term used to describe image ac-
includes detection of the features of interest quisition by means of a camera and frame grabber
based on grey level or color differences as well or a digital camera. Because of the many choices of
as morphological characteristics such as size and camera types, a video microscopy system must be
shape. Automated imaging applications are based flexible. Analog CCD cameras, black and white or
on a fundamental series of steps shown in Table color, are most frequently used. Component video
45. Depending on your goals, some or all of these (Y/C or S-Video) and composite video signals and
steps may be utilized. a number of color video standards such as NTSC,
PAL and SECAM are generally supported. Images
Aligned with these imaging goals, Buehler offers acquired in the materials laboratory are optimized
a series of upgradeable OmniMet imaging prod- in real time by adjusting brightness, contrast, and
ucts (Figure 52). The OmniMet Image Capture and color saturation. The analog output camera signal
Report System focuses on the basics of image is then digitized utilizing an analog input frame
capture and generating simple image reports. The grabber board.
OmniMet Archive imaging product has additional
capturing features that enable image comparison

Table 45. Primary Imaging Steps

Imaging Steps Description

Acquisition Capture, load, or import an image
Clarification Develop the necessary contrast and clarity for detecting the features of interest
Thresholding Detect the features of interest
Binary Operations Clean-up any detection discrepancies, categorize features, and overlay grids
Measurements Conduct field or feature measurements
Data Analysis Evaluate statistics and relevance of the measurements
Archive Store images, annotations, and associated measurements in a database
Distribution Printing or electronic distribution of images and results

93
Image Capture and Analysis

Alternatively, digital cameras may be integrated Each pixel in a bitmap is stored in one or more
with a SCSI firewire or USB connection. Since the data bits. If an image is monochromatic (black and
late 1990’s digital video cameras have become white), then one bit is enough to store information
increasingly popular for scientific imaging applica- for one pixel. If an image is colored or uses various
tions. This is largely accounted for by two factors: shades of grey, then additional bits are required to
1) higher pixel densities in the acquired images store color and shading information. Table 46 lists
potentially provide higher spatial resolution and some of the more common bit depths referred to
2) the cost range of digital cameras has merged in imaging software and camera literature.
with the combined cost of an analog camera and Table 46. Bit Depth
specialized capture board. Capturing an image
Bits
with a digital camera is often a two-stage process,
1 21 or 2 tones, black and white
first a preview image is viewed for focusing and
8 28 or 256 grey scale shades
selecting the field of interest and then a snapshot
24 224 or 16.7 million colors
of the image is taken. The preview image window
32 232 or 4.29 billion colors
displays the image at a lower pixel density and
the refresh rate of the camera will determine
the amount of lag time between the operator When working in grayscale, 256 values are used
adjusting the microscope and the new image to represent each shade of grey, starting with
being displayed. In order to be comparable to an black at 0 and transitioning to white at 255. For
analog camera, the refresh rate should approach example, the information stored for the image in
25 frames per second. Figure 53 would be a combination of the x and y
positions and a number from 0 to 255. For color
The resulting image files are based on two primary images there are two common models used to
formats, bit-map and vector. The majority of scien- numerically represent color, RGB and HLS. Each
tific imaging programs are based on bit-mapped requires storing three values, in addition to the x
images. Bit-mapped graphics are essentially grids, and y position of the pixel, in order to recreate the
consisting of rows and columns of pixels. When image. In general, you will notice the file sizes for
the image is zoomed to observe details, the indi- color images are larger than grey scale images of
vidual pixels will become visible. Pixels, or picture the same pixel density.
elements, are the smallest part of an image that a
computer, printer or monitor display can control. The RGB model is based on three primary colors
An image on a monitor consists of hundreds of – red, green and blue – that can be combined in
thousands of pixels, arranged in such a manner various proportions to produce different colors.
that they appear to be connected. Figure 53 This scheme is considered additive because, when
displays a cast iron image with a small segment combined in equal portions, the result is white.
zoomed such that the pixels are observed. Likewise, an absence of all three is black. The three
primary colors are each measured as a value from
0-255. The colors produced by combining the
three primaries are a result of the relative values
of each primary. For example, pure red has a red
value of 255, a green value of 0, and a blue value
of 0. Yellow has a red value of 255, a green value of
255, and a blue value of 0. This system is modeled
as a cube with three independent directions: red,
green, and blue, spanning the available space. The
body diagonal of the cube contains luminance, in
other words, colorless information about how light
or dark the signal is.

The HLS model uses the concept of Hue, Lumi-

Figure 53. A small segment of an image zoomed to nance, and Saturation. Hue is the color tone. For
display the individual pixels..

94
Image Capture and Analysis

example, an etchant might highlight the different account what you intend to do with your images
phases as unique shades of brown or blue. Satu- when selecting the appropriate camera.
ration is complementary to the hue. It describes
how brilliant or pure – as opposed to washed In order to perform measurements, the image or
out or grayish a particular color is. Luminance is the image source must first be calibrated. Calibra-
correlated with the light intensity and viewed as tion is achieved by assigning a known distance to
comparable to the greyscale values where zero is a pixel count. When using a light microscope this
black and 255 is white. The HLS color space can be is accomplished with a stage micrometer. Each of
viewed as a double cone, in which the z-axis of the the objectives will have its own unique calibration
cone is the grayscale progression from black to factor. If the aspect ratio of the camera pixels is un-
white, distance from the central axis is the satura- known, it can be determined by calibrating in both
tion, and the direction or angle is the hue. the x and y direction. If the aspect ratio is known,
then calibration only needs to be completed in the
The illustration in Figure 54 represents pixels of x direction. This calibration technique is repeated
nine different color values. These values are shown for each objective or working magnification. It is
in Table 47 for the greyscale, RGB, and HLS models. also possible to calibrate images that have been
When the pixels are not true grayscale tones, the imported from another source. The only require-
greyscale number is not given.

Resolution is primarily the imaging system’s abil-

ity to reproduce object detail by resolving closely
spaced features. In digital microscopy, the resolu-
tion is examined in terms of both the microscope
limitations and the pixel array of the camera. The
clarity and definition of a digital image depends
largely on the total number of pixels used to create
it.Typical pixel array densities range from 640 x 480
to 3840 x 3072. Multiplying the two numbers gives
you the total number of pixels used to create the
image. Because a large number of pixels leads to
sharper images, you might conclude that it is al-
ways desirable to capture images with the largest
pixel array possible. That is not necessarily always
true. The problem with large array images is that
they consume a lot of storage space. As a result,
these images are not recommended for posting on
web sites or email use. It is important to take into
Figure 54. A representation of nine pixels with
different color values.

Table 47. Values based on color models

HLS Method RGB Method

Pixels Greyscale Hue Luminance Saturation Red Green Blue
1 192 0 192 0 192 192 192
2 150 0 150 0 150 150 150
3 128 0 128 0 128 128 128
4 0 127 255 255 0 0
5 85 127 255 0 255 0
6 170 127 255 0 0 255
7 60 98 100 255 255 0
8 310 36 79 148 31 129
9 25 100 176 169 113 31

95
Image Capture and Analysis

ment is to have a feature within the image that is description of a feature within the image. A recom-
of a known dimension. A scale marker is usually mended annotation is a scale marker in the lower
the best choice. corner of the image. This is particularly important
for distributing the image electronically because
the final size of the printout is at the discretion of
Clarification
the recipient.
Image clarification is also referred to as image
enhancement and is largely achieved through
the use of greyscale filters. Filters perform one Automated Measurements
of several standard functions: edge detection, When more detailed measurements are required
photographic enhancement, or grey level modi- or a large quantity of measurements is important,
fication. Image clarifications, which modify the it is useful to automate the process. For example,
values of the pixels across an entire image, are con- consider the coating that was measured in Figure
ducted in one of two ways: adjusting limits such as 56. It is possible to duplicate that measurement as
contrast, brightness and gamma or comparing the well as to evaluate any variation across the coating
values of neighboring pixels. The latter is referred and determine the percentage porosity within
to as a convolution or kernel operation. The neigh- the coating. A typical process for determining
borhood sizes are typically squares of pixels from coating thickness is outlined in Figure 57 and is
3x3, 5x5, 7x7, etc. A typical need is to increase the described below.
local contrast at the phase boundaries. Often times,
the use of a neighborhood transformation filter
results in a more narrow gray scale distribution of a
given phase, making the subsequent thresholding
or detection process easier.

Operator Interactive Measurements

Interactive measurements require the operator
to use a mouse to select the start and end points
for each measurement. The most common of
these are linear measurements (point-to-point),
parallel, radius or diameter, curvilinear, and angle
Figure 56. An example of an operator estimating
(Figure 55). An operator will typically make several coating thickness with the use of a parallel tool.
measurements over the area of interest. These
are either electronically transferred to a file or
Thresholding
transcribed by hand. This method is appropriate
for a limited number of measurements or where Thresholding is the method for representing
an average is of interest rather than variation. Fig- ranges of pixel grey or color values with different
ure 56 represents the most direct way to take an color bitplane overlays. The bitplanes are a binary
average measurement of thickness using a parallel layer placed over the image plane superimposing
line tool. One line is aligned with the interface and the phase(s) of interest. Most of the measurements
then the opposite line is aligned with the median are accomplished on this layer not on the actual
of the rough surface. image.

The first step of the thresholding process is the

development of a threshold histogram. Each
pixel is assigned a value and then the frequency
Figure 55. Operator interactive tools founds in is graphed. When working in grayscale with an
imaging software.
8-color image, the same 256 values correspond
In addition to measurements, other annota- to luminance. Figure 59 demonstrates a yellow
tions are often placed on electronic images. bitplane being assigned to the coating (Figure 58)
Annotations can be as simple as a label that is based on the HLS values of the pixels as discussed
referenced in the image caption or a complete text previously.

96
Image Capture and Analysis

Figure 57. A flowchart of the logic used when creating an automated measurement routine for coating
thickness.

97
Image Capture and Analysis

Binary Operations or feature of interest is detected by the same bit-

After the thresholding process has been accom- plane color because of a similar gray level range.
plished, the different phases and features in an A binary operation is the process that separates
image are represented by different bitplane color and classifies features within the same bitplane,
overlays. It is possible that more than one phase based on morphology or size. Additionally, not
all images are detected as desired. There may be
over-detected or under-detected regions that
need to modified. In general, binary operations
allow the operator to isolate the features that are
of interest.

Two common measurements on this type of coat-

ing would include the relative area fraction of
porosity as well as the average thickness. Before
either of these measurements can be made it is
necessary to define the total area of the coating.
Figure 58. A TSC color image captured at 1200 x
1600 pixels.
A combination of the binary commands Fill and
Close are used to eliminate the holes (undetected
areas) that are totally enclosed by the detected
bitplane. The thickness is determined by averag-
ing a series of chord measurements across the
coating, rather than a single measurement of
the yellow bitplane. An additional binary layer is
superimposed in the form of a grid using a Grid
command. Figure 60 demonstrates the effects of
these commands on the coating.

Figure 60. The effect of the Binary commands,

Close, Fill, and Grid on the detected coating.

To define the individual chords, where the grid

overlaps the coating, a Boolean command is
employed. Boolean operations perform logical
functions between bitplanes. The two most com-
mon are OR and AND. AND takes those portions
of two bitplanes that overlap and integrates
them into a single destination bitplane. In the
coating, for example, AND considers only those
Figure 59. Thresholding the coating with a yellow segments of the red grid bitplane that overlap
bitplane overlay using the HLS color model.

98
Image Capture and Analysis

the yellow coating bitplane. Figure 61 shows the

red chords which result from the AND command.
OR combines two source bitplanes together to
form a single destination bitplane. For example, if
two different inclusion types, oxides and sulfides,
are detected initially, they can be combined to
measure the total inclusion content.

Figure 62. The histogram of results representing

the chords shown in Figure 61.

then plotted based on frequency in the histogram

and contributes to the overall statistics. In the case
of grain or particle size measurements, features on
the edge of the image are eliminated because the
complete size of the feature is unknown.

Field measurements are performed on the entire

field of view or a framed portion of the image,
Figure 61. The chords that result from the providing the sum of the individual measurements
overlapping coating and grid lines.
in the selected area. Statistical information for field
measurements is only generated if multiple fields
One of the most powerful binary functions is the
are analyzed. This can be employed to observe
ability to categorize individual features based
microstructural variations within different fields of
on shape or size. It can be a single specification
a specimen. A fairly common field measurement
(>5-µm) or multiple specifications (>2 and <5)
is area percentage. An area percentage measure-
separated by an AND or an OR. If a maximum allow-
ment results from dividing the number of pixels
able pore size is specified, all pores that exceed that
in the bitplane of interest by the total number of
limit can be flagged. The limit itself can be based
pixels in the image. In the case of the coating, there
on average diameter, maximum diameter, area,
is a need to measure the area percent porosity rela-
aspect ratio, etc. Additionally, the shape of the pore
tive to the coating area and not the entire image
may have some significance. For example, in cast
area. This is calculated by dividing the number of
materials, a gas bubble tends to be circular while a
pixels in the bitplane of interest, i.e. the pores in the
shrinkage cavity will have a more complex shape.
coating, by the total number of pixels representing
Common shape factors examine the relationship
the coating.
between the perimeter and area of the feature.
For this coating, the pores were separated from
the elongated oxides to insure that the porosity Common Applications
level was not overestimated. Throughout the past 20 years of Buehler’s involve-
ment in image analysis, a number of applications
Returning to Figure 61, where grid lines are placed stand out and may be of interest to most image
across the coating, the length of all the individual analysis users. Below is a listing of these applica-
grid lines can be measured.The data can be shown tions:
as a histogram offering the minimum, maximum,
mean and standard deviation of the measure- Grain Size. Image analysis provides a rapid and
ments across the coating (Figure 62). accurate means for determining grain size ac-
cording to ASTM E112. Even if etching is unable
Most measurements fall into one of two categories: to produce complete grain boundaries, or if there
field or feature-specific measurements. The grid are twins that could skew the data, binary modifi-
lines across the coating are typical of feature mea-
surements. Each individual feature is measured and

99
Image Capture and Analysis

cations can be employed to make corrections. If a Phase Percentage. The area percent of various
specification cites maximum grain size limitations, phases in a microstructure influences the proper-
the excessively large grains may be transferred to ties. The tensile strength of grey iron, for example,
a different bitplane color to provide visual and is directly related to the percentage of pearlite in
numerical feedback. its microstructure. In addition, in a single image,
multiple phases can be detected, measured and
Porosity. Porosity is detrimental to the physical presented in one graph. For example, when evalu-
properties of most engineering materials. Image ating the inclusion content of steels, it would be
analysis is able to characterize the pores according useful to examine the overall inclusion content as
to the total number of pores, number per unit area, well as isolating particular inclusion types, such as
maximum size, average size and the size distribu- oxides and sulfides.
tion in the form of a histogram.

Linear Measurements. Simple point-to-point

Helpful Hints For
measurements are widely used for making occa-
Image Capture and
sional measurements; however, in cases where a
Analysis
high quantity of measurements and more statistics Determine the required pixel
are required, automated image analysis is time density of a camera system based on the
saving. A coating or layer is detected based on size of the features of interest and accept-
the pixel values and then, after binary isolation able file size
of the coating, grid lines are superimposed. Us-
ing Boolean logic to evaluate the common pixels When distributing an image, always add a
between the layer and grid lines, the result is many scale bar
chords representing the thickness at given points When saving images for analysis at a later
in the coating. time, verify that the image is stored as cap-
tured, i.e., without any filters applied.
Feature Shape and Size. The shape of the
graphite constituent in ductile irons is critical.
Ductile iron was developed such that the graphite
would occur in the form of spherical nodules with
the result of dramatically improved mechanical
properties. However, variations in chemistry and
other factors can cause the nodules to be irregular,
leading to some degradation of the properties.The
ability to monitor the graphite shape or determine
“nodularity” is another ability of image analysis.
These same techniques are applicable to any
constituent that can be detected.

100
Laboratory Safety

LABORATORY SAFETY
The metallographic laboratory is a relatively safe
working environment; however, there are dangers
inherent to the job. Included in these dangers is
exposure to heat, acids, bases, oxidizers and sol-
vents. Specimen preparation devices, such as drill
presses, shears and cutoff saws, also present haz-
ards. In general, these dangers can be minimized
if the metallographer consults documents such as
Figure 63. Carelessness in the laboratory can
ASTM E 2014 (Standard Guide on Metallographic result in dangerous chemical spills and a clean-up
Laboratory Safety) and relevant Material Safety nightmare.

Data Sheets (MSDS) before working with unfamil- Most reagents, chemical or electrolytic polishing
iar chemicals. Common sense, caution, training in electrolytes, and solvents should be used under a
basic laboratory skills, a laboratory safety program, ventilation hood designed for use with chemicals.
access to safety reference books – these are some Many of these chemicals used in metallo- graphy
of the ingredients of a recipe for laboratory safety. can cause serious damage on contact. It is best to
Table 48 lists the main requirements for a compre- assume that all chemicals are toxic and all vapors
hensive safety program. or fumes will be toxic if inhaled or will be damag-
ing to the eyes. A hood will prevent the working
Safe working habits begin with good housekeep- area from being contaminated with these fumes.
ing. A neat, orderly laboratory promotes safe However, you will not be protected when your
working habits, while a sloppy, messy work area head is inside the hood. In most cases, a protective
invites disaster. Good working habits include plastic or shatterproof glass shield can be drawn
such obvious, commonsense items as washing across the front of the hood for further protection
the hands after handling chemicals or before eat- from splattering or any unexpected reactions.
ing. Simple carelessness can cause accidents (see
Figure 63). For example, failure to clean glassware All laboratories should be equipped with a shower
after use can cause an accident for the next user. and eyewash for emergency use. This equipment
Another common problem is burns due to failure should be near the work area so that the injured
to properly clean acid spills or splatter. can reach it quickly and easily. Fire alarms and fire

Table 48. Elements of a Comprehensive Metallography Laboratory Safety Plan

1. Emergency Response 6. Equipment Use (JSAs)

a. First Responders a. Sectioning
b. Emergency Phone Numbers b. Mounting
c. First Aid c. Grinding and Polishing
d. Spills d. Electropolishing
2. Chemical Procurement, Distribution, and Storage e. Etching
a. MSDS Management 7. Facility
b. Labeling/Storage a. Facility and Equipment Maintenance
3. Laboratory SOPs b. Housecleaning
a. Handling Chemicals c. Fume Hood Air Flow
b. Mixing Chemicals d. Fire Extinguishers and Fire Protection
c. Rules e. Spill Response
d. Hygiene f. Showers and Eye Washes
g. Air Lines and Filters
4. Employee Health and Exposure
h. Plumbing
a. Personal Protective Equipment
i. Electricals
b. Monitoring of exposure levels
c. Medical Evaluations 8. Employee Training
d. Additional protections for employees working with a. Safety
particularly hazardous substances (carcinogens, toxins) b. Indications and Symptoms of Exposure
c. Job Functions
5. Disposal of Hazardous Materials
a. Etchants and Chemicals 9. Scheduled and Documented Safety Inspections and
b. Recirculating Tanks Meetings
c. Lubricants and Abrasives

101
Laboratory Safety

extinguishers (CO2 type) should be available and hood. Waste must be handled carefully and dis-
tested periodically. A good first aid kit and a chemi- posed of properly. Radioactive materials require
cal spill treatment kit should be readily available. special remote- handling facilities and elaborate
safety precautions.
Laboratory Equipment A drill press is frequently used in the laboratory.
Specimen preparation devices used in metal- Drilling holes in thin sections requires secure
lographic laboratories are generally quite safe to clamping; otherwise the sample can be grabbed
use. Information supplied by the manufacturer by the drill and spun around, inflicting serious
usually describes safe operating procedures. It is lacerations. Hair, ties, and shirt cuffs can become
good laboratory practice to prepare a job safety tangled in a drill, inflicting serious injuries. Safety
analysis (JSA) detailing potential hazards and glasses should always be worn when using drill
describing the safe operating procedure for each presses or when cutting or grinding. Mounting
piece of equipment. This information should be presses or laboratory heat-treatment furnaces
provided to all users and it must be revised and present potential burn hazards. It is a good
reviewed periodically. practice to place a “hot” sign in front of a labora-
tory furnace when it is in use. Gloves should be
Band saws or abrasive cutoff saws are commonly worn when working with these devices. Modern
used by metallographers.The cutting area of band mounting presses that cool the cured resin back
saws is exposed and potentially dangerous. Your to near room temperature dramatically reduce the
hands should never be used to guide the work potential for burns.
piece during cutting. A guiding device or block of
wood should always be used between the work It is occasionally necessary to heat solutions dur-
piece and your hands. After cutting is completed, ing their preparation or use. Although Bunsen
the saw should be turned off before pieces near burners are commonly employed for this purpose,
the blade are removed. Samples should be it is much safer to use a hot plate or water bath and
handled carefully, because considerable heat can thus avoid the use of an open flame. If a Bunsen
be generated. In addition, sharp burrs are often burner is used, the flame should never be applied
present, which should be carefully removed by directly to a flask, beaker, or dish. Plain- or asbes-
filing or grinding. Abrasive cutoff saws are safer to tos-centered wire gauze should always be placed
use because the cutting area is closed off during between the flame and the container.
use. The chief danger is from flying pieces from
a broken wheel. Fortunately, the closed cover
contains these pieces within the cutting chamber. Personal Protective Equipment (PPE)
Wheel breakage usually occurs when the part is Metallographer must take certain precautions
not firmly clamped in place or if excessive pressure to insure their personal safety. A laboratory coat
is applied, a bad practice from the standpoint of is useful for protecting the operator’s clothing,
specimen damage as well. and should be changed regularly and cleaned
professionally. When handling caustics, a rub-
Dust produced during grinding of metals is always berized or plastic coated apron provides better
dangerous. For certain metals, like beryllium, mag- protection. Gloves should be worn when handling
nesium, lead, manganese, and silver, the dusts are bulk samples, working with hot material, or using
extremely toxic. Wet grinding is preferred both for hazardous solutions. Lightweight surgeons’ gloves
dust control and for preventing thermal damage are very popular, because the operator retains
to the specimen. Bench grinders must be firmly the ability to “feel.” Many metallographers wear
mounted to prevent sudden movement. Care these to protect their skin when mounting with
must be exercised to avoid grinding one’s fingers epoxies, and polishing with oxide suspensions.
or striking the edge of a grinding belt, which will When using these gloves with chemicals, always
cause painful lacerations. With nearly all materi- inspect for holes, as they are easily punctured.
als, wet grinding is preferred and produces best Thick rubber gloves are often used for handling
results. For routine handling of dangerous metals, specimens during macroetching, chemical
grinding should be done wet under a ventilation polish ing, pickling, etc. The gloves should al-

102
Laboratory Safety

ways be checked first for small holes or cracks, These bottles are useful for holding solutions, such
because gloves can impart a false sense of security. as nital, that can build up gas pressure within a
The operator’s hands generally perspire when tightly stoppered bottle. A small hole can be drilled
using rubber gloves and it is sometimes difficult through the cap top to serve as a pressure relief
to tell if the moisture is due solely to perspiration vent. Tightly stoppered bottles of nital and some
or to leakage. Safety glasses should be worn dur- other solutions have exploded as the result of
ing processes that generate particulate matter. pressure buildup. Be certain that the reagent is safe
Goggles are appropriate for use with chemicals, to store and store only small quantities. All bottles
and a chemical face shield is recommended when should be clearly labeled. Polyethylene bottles are
handling large quantities of hazardous liquids. The required for etchants containing hydrofluoric acid,
appropriate PPE will be specified on the MSDS for which attacks glass.
most laboratory chemicals and products.
Most recipes for etchants or electrolytes list
the ingredients by weight if they are solids and
Chemicals, Storage and Handling by volume if they are liquids. In a few cases, all
Many of the chemicals used in metallography amounts are given in weight percentages. In most
are toxic, corrosive, flammable, or potentially cases, reagent compositions are not extremely
explosive. Whenever possible, purchase small critical. An ordinary laboratory balance provides
quantities that are likely to be used within a rea- adequate weighing accuracy, while graduated cyl-
sonably short time. Flammable solvents should be inders provide acceptable accuracy for volumetric
stored in fireproof steel cabinets. Acids and bases measurements. These devices should be cleaned
should be stored separately, again in fireproof steel after use to prevent accidents to the next user. For
cabinets. Strong oxidants must not be stored along weight measurements, a clean piece of filter paper,
with acids, bases or flammable solvents. or a cup, should be placed on the balance pan to
hold the chemical, to protect the pan surface, and
Reagent-grade chemicals or solvents of highest to facilitate transfer to the mixing beaker. A large
purity are recommended. Although more expen- graduated beaker is usually employed for mixing
sive, the amounts used are small and the gain in solutions.
safety and reliability compensates for the cost dif-
ference. Chemicals may deteriorate during storage. With many etchants, the mixing order is impor-
Exposure to light can accelerate deterioration of tant, especially when dangerous chemicals are
some chemicals. Hence, they should be stored in used. When water is specified, distilled water
a closed metal cabinet. should always be used, because most tap wa-
ter contains minerals or may be chlorinated or
Etchants fluorinated. Tap water can produce poor results or
unexpected problems. Cold water should always
Most laboratories mix commonly used reagents in
be used, never warm or hot water, which can
quantities of 250 to 1000 mL and then store them
cause a reaction to become violent. In mixing, one
as stock reagents. Many reagents can be safely
should start with the solvents, such as water and
handled in this manner. It is best to store only
alcohol; then dissolve the specified salts. A mag-
those reagents that are used regularly. Glass-stop-
netic stirring device is of great value, as shown in
pered bottles are commonly used as stock reagent
Figure 64. Then, the dangerous chemicals, such as
bottles. If these bottles are opened regularly, the
acids, should be added carefully and slowly while
stopper will not become “frozen”. However, if they
the solution is being stirred. Whenever sulfuric
are used infrequently, a frozen stopper often re-
acid (H2SO4) is specified, it should be added last.
sults. Holding the neck of the bottle under a stream
It should be added slowly, while stirring, and it
of hot water will usually loosen them. If thermal
should be cooled, if necessary, to minimize heat-
expansion does not free the stopper, the stopper
ing. Never just pour one liquid into another, as
can be gently tapped with a piece of wood. Glass
shown in Figure 65. If sulfuric acid is added to
bottles with plastic screw-on tops can be used so
long as the solution does not attack the plastic.

103
Laboratory Safety

be, and has been, done. Table 49 lists examples

from the literature (27-29) of chemical polishing
solutions, electrolytic polishing solutions and
etchants that have been involved in accidents.
Table 50 lists a number of commonly used chemi-
cals and incompatible chemicals.

Solvents
Numerous organic solvents are used for clean-
ing or are ingredients in chemical or electrolytic
polishing solutions or etchants, where they are
used to control ionization or the speed and mode
of attack. Commonly employed solvents include
Figure 64. When mixing etchants, use a magnetic
stirring plate with a magnetic bar for stirring. Slowly water, acetone, ethyl ether, ethylene glycol, glycerol
add the liquid ingredients to the solvent by dripping (glycerin), kerosene, petroleum ether, trichloroeth-
them down a glass stirring rod. If the solution is more
dangerous than this one, wear protective gloves and ylene, butyl cellosolve, and alcohols, such as amyl
use a face shield. If mixing generates substantial
heat, it is a good practice to place a cooling jacket
alcohol, ethanol, methanol, and isopropyl alcohol.
around the beaker. Most are flammable and their vapors can form
water without stirring, it can collect at the bottom explosive mixtures with air. They should be kept
of the beaker and enough local heating can occur closed when not in use and should be stored in a
to throw the contents out of the beaker. cool place away from heat and open flames.

Acetone (CH3COCH3) is a colorless liquid with a

fragrant mintlike odor. It is volatile and highly flam-
mable. It is an irritant to the eyes and the mucous
membranes. Its vapor is denser than air and can
travel along the ground and can be ignited at a
distance. Acetone can form explosive peroxides on
contact with strong oxidizers such as acetic acid,
nitric acid and hydrogen peroxide. It is an irritant
to the eyes and respiratory tract, and will cause the
skin to dry and crack.

Butyl cellosolve (HOCH2CH2OC4H9), or ethylene

glycol monobutyl ether, is a colorless liquid with
Figure 65. Illustration of a bad mixing practice. a rancid odor that is used in electropolishing so-
The acid ingredient was poured into an empty acid
bottle and the solvents were added without stirring
lutions. It is combustible and may form explosive
or cooling. The solution may erupt in the metallo- peroxides. It is toxic in contact with the skin, can
grapher’s face at any moment. It is recommended to
keep the hands clear of the area and remove rings.
be absorbed through the skin, can cause serious
damage to the eyes, irritation to the skin and
The literature contains references to a great respiratory tract.
many formulas for etchants, chemical polishes,
Carbitol (C2H5OCH2CH2OCH2CH2OH), or diethylene
and electrolytes that are potentially dangerous
glycol monoethyl ether, is a colorless, viscous sol-
or extremely dangerous. Few of these references
vent that is compatible with water and is used in
contain comments regarding safe handling
electropolishing solutions. It irritates the skin, eyes,
procedures or potential hazards. Fortunately, me-
mucous membranes, and upper respiratory tract,
tallographic applications involve small quantities
and is harmful if inhaled or swallowed.
of these solutions, and accidents do not usually
produce catastrophic results. However, even with
small solution volumes considerable damage can

104
Laboratory Safety

Table 49. Chemical and Electrolytic Polishing Solutions and Etchants Known to be Dangerous

Solution Use Problems

5 parts Lactic Acid Chemical polishing Lactic and nitric acids react autocatalytically.
5 parts HNO3 solution for Zr Explosion can occur if stored.
2 parts Water
1 part HF

50 parts Lactic Acid Chemical polishing Lactic and nitric acids react autocatalytically.
30 parts HNO3 solution for Ta, Nb Explosion can occur if stored.
2 parts HF and alloys

3 parts Perchloric Acid Electropolishing Mixture is unstable and can, and has, exploded, with
1 part Acetic Anhydride solution for Al heating, or in the presence of organic compounds
adding to the potential hazard.

60-90 parts Perchloric Acid Electropolishing Solution will explode at room temperature.
40-10 parts Butyl Cellosolve solution Solutions with ≤30% HCIO4 will be safe if T is <20°C.

100g CrO3 Electropolishing CrO3 was dissolved in water, cooled to about 20°C;
200mL Water solution the acetic anhydride was added very slowing with
700mL Acetic Anhydride stirring. The solution became warm to the touch.
About 20 seconds later it erupted from the beaker.

1 part HNO3 Electropolishing Mixture is unstable and cannot be stored

2 parts Methanol solution for Muntz
(Cu-40% Zn) metal

20mL HF Etchant for Nb, Ta, This etchant is unstable. At 20°C, it reacted after 18
10mL HNO3 Ti, V, Zr hours. At 30-35°C, it reacted after 8 hours with
30mL Glycerol violence. At 100°C, it will react after 1 minute.

20-30mL HCI Etchant for Ni and Two incidents occurred when a violent reaction
10mL HNO3 stainless steels resulted producing NO2 and a spray of acid
30mL Glycerol after the etch was left standing for 2-3 hours.

40mL Acetic Acid Etchant for Ni A closed bottled exploded about 4 hours after
40mL Acetone it was mixed. Authors state that solutions
40mL HNO3 without the acetic acid also cannot be stored.

10mL HNO3 Etchant The solution reacted spontaneously about 2 minutes

10mL Acetic Acid after mixing with evolution of heat and fumes (nitrous
20mL Acetone and nitric oxides). The mixed acids were poured into
the acetone. The beaker was externally cooled.

50mL Nitric Acid Etchant Mixtures have exploded violently after mixing
950mL Isopropyl Alcohol or about 20 minutes after mixing.

Table 50. Some Incompatible Chemicals

Chemical Use in Metallography Do Not Mix With the Following:

Acetic Acid Chemical polishing Chromic acid, glycol, hydroxol compounds,
electrolytic polishing nitric acid, peroxides, permanganates
Acetone Degreasing, cleaning, Concentrated solutions of nitric and
etchants sulfuric acid
Chromic Acid Electropolishing Acetic acid, flammable liquid, glycerol
Hydrogen Peroxide Chemical polishing, etchants Flammable liquid, organic materials
Nitric Acid (conc) Chemical polishing, etchants Acetic acid, chromic acid, flammable liquids,
isopropyl alcohol
Perchloric Acid Electropolishing Acetic anhydride, alcohol, some organics,
oil, grease
Sulfuric Acid Etchants Methyl alcohol, potassium chlorate,
potassium perchlorate and potassium
permanganate

105
Laboratory Safety

Ethylene glycol (HOCH2CH2OH) is a colorless, hy- Ethyl alcohol (CH3CH2OH), or ethanol, is a colorless,
groscopic liquid with a sweet taste (but do not inoffensive solvent commonly used in metallog-
swallow as it is poisonous) that reacts with strong raphy. Ethanol is miscible with water and rapidly
oxidants and strong bases. It is flammable. The absorbs up to 5% water from the air.The denatured
substance irritates the eyes, skin, and the respira- version is less expensive and contains 5% absolute
tory tract. methanol and is suitable for any recipe requiring
ethyl alcohol. It is a dangerous fire hazard, and
Glycerol (glycerin) (CH2OHCHOHCH2OH) is a color- its vapors are irritating to the eyes and upper
less or pale yellow, odorless, hygroscopic, syrupy respiratory tract. High concentrations of its vapor
liquid with a sweet, warm taste. It is relatively can produce intoxication. Because ethanol is com-
nontoxic and nonvolatile but can cause iritis (in- pletely burned in the body, it is not a cumulative
flammation of the iris). It is combustible and is a poison like methanol.
moderate fire hazard. Glycerol should never be
used in anhydrous solutions containing nitric Methyl alcohol (CH3OH) is an excellent, non-hy-
and sulfuric acids, because nitroglycerin can form. groscopic solvent, but it is a cumulative poison.
Glycerol should not be used with strong oxidizing Ingestion, inhalation or absorption through the
agents, such as chromium trioxide and potassium skin in toxic levels can damage the central nervous
permanganate, as an explosion may occur. Glycerol system, kidneys, liver, heart, and other organs.
is often added to aqua regia (glyceregia).This mix- Blindness has resulted from severe poisoning. It
ture decomposes readily and should be discarded is particularly dangerous because repeated low-
immediately after use. This etchant should not be level exposures can also cause acute poisoning as
allowed to stand for more than about 15 minutes a result of accumulation. Ethanol should be used
after mixing. whenever possible. When using methanol, always
work under a ventilation hood. Mixtures of metha-
Kerosene is occasionally employed in grinding nol and sulfuric acid can form dimethyl sulfate,
samples and with diamond paste as a lubricant. which is extremely toxic. Solutions of methanol
Only the deodorized form should be used. It is and nitric acid are more stable than mixtures of
flammable, but the vapors do not readily explode. nitric acid and higher alcohols.
Contact defattens the skin and can cause derma-
titis, irritation or infections. Isopropyl alcohol [CH3CH(OH)CH3], also known
as 2-propanol, is a clear, colorless liquid that,
Trichloroethylene (CHCl:CCl 2) is a stable, col- like ethanol, does not accumulate in the body,
orless liquid with a chloroform-like odor. although it does have a strong narcotic effect.
Effective laboratory ventilation is necessary. At It is a flammable liquid and is a dangerous fire
ambient temperatures it is nonflammable and hazard. Metallographers have used it as a substi-
nonexplosive,butbecomeshazardousathighertemp- tute for ethanol but isopropyl alcohol has quite
eratures. In the presence of strong alkalies, with different characteristics and should not be used.
which it can react, it can form explosive mixtures. Fatal injuries and explosions have been reported
In the presence of moisture, the substance can be due to its use.
decomposed by light to corrosive hydrochloric
acid. It is carcinogenic to humans, and toxic when
inhaled or ingested, which may cause acute poi-
Acids
soning. Inorganic and organic acids are common con-
stituents in chemical and electrolytic polishing
Amyl alcohol (CH3(CH2)4OH), or 1-Pentanol, is a solutions and in etchants. The inorganic, or mineral
colorless liquid with noxious odor. It is flammable acids, including the very familiar acids such as
and the vapors may form explosive mixtures at hydrochloric, nitric, perchloric, phosphoric, and
elevated temperatures. The substance reacts sulfuric, are highly corrosive and poisonous.They
violently with strong oxidants and attacks alka- should be stored in a cool, well-ventilated location
line metals. The fumes are irritating to the eyes, away from potential fire hazards and, of course,
upper respiratory tract, and skin. The substance is away from open flames. They should not be stored
toxic through ingestion, inhalation, or absorption in a location that receives direct sunlight.When the
through the skin. pure acids contact metals, most liberate hydrogen
106
Laboratory Safety

gas – a fire and explosion hazard.The organic acids Never add nitric acid to isopropyl alcohol. Ander-
are naturally occurring substances in sour milk, son (27) reported that a litre bottle of 5% nitric
fruits, and plants and include the following acids: acid in isopropyl alcohol was mixed and placed in a
acetic, lactic, citric, oxalic, and tartaric. cabinet. Although this had been done many times
in the past without problems, twenty minutes
Hydrochloric acid (HCl), commonly used in metal- later the bottle exploded destroying the cabinet,
lography, is a colorless gas or fuming liquid with other stored bottles, and throwing debris up to
a sharp, choking odor. It is very dangerous to the 20 feet away. Anderson (27) also reported that a
eyes and irritating to the nose and throat. It attacks metallographer was pouring a freshly mixed litre
the skin strongly causing severe burns. of 5% nitric acid in isopropyl alcohol into another
bottle when it exploded. The person died within
Nitric acid (HNO3), also commonly used in met- three hours without being able to tell anyone what
allography, is a colorless or yellowish fuming happened. As with the other explosion, this same
liquid, highly toxic and dangerous to the eyes. If procedure had been performed many times pre-
it contacts organic material or other easily oxi- viously without mishap. Anderson recommends
dizable materials, it can cause fires and possibly avoiding use of isopropyl alcohol completely.
explosions. When it reacts with other materials,
toxic oxides of nitrogen are produced. The oxides, Most metallographers consider nital to be very
which vary with the conditions, include nitrous safe to use, and indeed it is. However, even with
acid, nitrogen dioxide, nitric oxide, nitrous oxide, such an apparently safe solution, one can have
and hydroxylamine. A commonly encountered accidents. One such accident occurred when
problem involves pouring nitric acid into a gradu- an employee, not a skilled metallographer, was
ated cylinder that contains some methanol or replenishing a stock of 5% nitric acid in ethanol
ethanol from prior use. The brown fumes given using a procedure that he had claimed to have per-
off are quite harmful. Mixtures of nitric acid and formed many times previously (he was not taught
alcohols higher than ethanol should not be stored. the safe way to mix nital as it was a union chemist’s
Mixtures of concentrated nitric and sulfuric acids job to mix nital i.e., not his job). The worker began
are extremely dangerous, while strong mixtures of by adding the desired volume of concentrated
nitric acid and glycerin or glycols can be explosive. nitric acid into the container that contained a small
Aqua regia, a mixture of one part nitric acid and residual amount of stale 5% nital. To his surprise,
two to four parts hydrochloric acid, forms several the contents began “boiling” and spewing out of
products including nitrosyl chloride, an exception- the container along with dense, brown fumes. The
ally toxic gas. Aqua regia is a popular etchant but acid splashed the worker, resulting in burns on his
must be used with care under a hood. forehead, face and eyes. The small amount of aged
nitric acid solution (the concentration may have
Ethanol with additions of up to 3% nitric acid (nital) been increased due to evaporation of the alcohol),
can be safely mixed and stored in small quantities. present in the container when the fresh acid was
Higher concentrations result in pressure buildup in added, created a dangerous chemical reaction. An
tightly stoppered bottles. Explosions of 5% nitric experiment also showed that a similar reaction can
acid in ethanol have occurred as a result of failure occur when nitric acid is poured into a graduated
to relieve the pressure. If higher concentrations cylinder containing only remnants of ethanol or
are desired, they can be mixed daily, placed in an methanol.
open dish, and used safely. Discard the etchant
at the end of the day. Mixtures of methanol with Sulfuric acid (H2SO4) is a colorless, oily liquid that
up to 5% nitric acid are safe to use and store in is highly corrosive, a strong oxidizing agent, and
small quantities. Mixtures of methanol with more dangerously reactive. It reacts violently with bases
than 5% nitric acid if heated are subject to vio- and is corrosive to most metals forming flam-
lent decomposition. Mixtures of 33% nitric acid mable/explosive hydrogen gas. It reacts violently
in methanol have decomposed suddenly and with water and organic materials with evolution
violently. of heat. Upon heating, toxic sulfur oxides are

107
Laboratory Safety

formed. slowly to water with constant stirring. If material, can form highly unstable perchlorates,
added without stirring, it will produce a pocket of which can ignite and cause explosions. Regular
steam in the bottom of the vessel, throwing the use of perchloric acid requires that the ventila-
contents out of the vessel. Concentrated sulfuric tion system must be specifically designed and
acid can cause severe, deep burns on contact with maintained for perchloric acid. Special fume hoods
the skin, and permanent vision loss on contact with a waterfall-type fume washer will remove
with the eyes. Tissue is destroyed by the acid’s the perchlorate fumes before they can enter the
dehydrating action. Lungs may be affected by exhaust system.
long-term or chronic exposure to aerosol. Skin le-
sions, tooth erosion, and conjunctivitis are other Perchloric acid is very useful in electropolishing
long-term effects. solutions. Never electropolish samples mounted
in phenolic (Bakelite) or other plastics in perchlo-
Hydrofluoric acid (HF) is a clear, colorless, fuming ric acid solutions as explosions can result. The
liquid or gas with a sharp, penetrating odor. It is mixture of perchloric acid and acetic anhydride,
very dangerous to the eyes, skin, and upper respi- which was developed by Jacquet, is difficult to
ratory tract. The substance can be absorbed into prepare and highly explosive. Jacquet has re-
the body by inhalation, through the skin and by viewed the accidents involving perchloric acid
ingestion. A harmful concentration of the gas in and has described safety procedures (30). The
air can be reached quickly, making it very danger- worst accident occurred on February 20, 1947, in
ous to handle. Exposure by ingestion, inhalation an electroplating factory in Los Angeles. In this ac-
or contact, can be fatal. Undissociated HF poses cident 17 people were killed and 150 were injured
a unique threat in that it can destroy soft tissues (29). Medard, Jacquet, and Sartorius have prepared
and result in decalcification of the bone. Moreover, a ternary diagram showing safe com- positions of
the effects may be delayed. Laboratories where perchloric acid, acetic anhydride, and water, Figure
HF is used should stock an antidote kit to be used 66. Anderson, however, states that accidents have
in case of exposure. Although it is a relatively still occurred with solutions in the “safe” region
weak mineral acid, HF will attack glass or silicon of this diagram (27). Electropol-ishing solutions
compounds, and should be measured, mixed, composed of perchloric acid and acetic anhydride
and stored in polyethylene vessels. HF reacts with are not recommended. Many companies forbid the
many compounds, including metals, and will liber- use of such mixtures, and some cities have banned
ate explosive hydrogen gas. their use. Electropol-ishing solutions of perchloric
acid and alcohol, with or without organic additions,
Orthophosphoric acid (H3PO4), a colorless thick liq- and mixtures of perchloric acid and glacial acetic
uid or hygroscopic crystal, is a medium strong acid. acid are safe to use. Nevertheless, in using these
It is corrosive to the skin, eyes and respiratory tract. “safe” mixtures, one should follow the formula in-
Phosphoric acid decomposes on contact with al- structions carefully, mix only small quantities, keep
cohols, aldehydes, cyanides, sulfides, ketones, and the temperature under control, and avoid evapora-
can react with halogenated organic compounds tion. These solutions should not be stored.
forming organophosphorus nerve-gas type com-
pounds that are extremely toxic. It reacts violently Mixtures of acetic acid and 5-10% perchloric
with bases, and will generate hydrogen gas when acid have been commonly used to electropolish
it reacts with metals. iron-based alloys and are reasonably safe. Do
not use these solutions to electropolish bis-
Perchloric acid (HClO 4) is a colorless, fuming, muth, arsenic or tin, as explosions have occurred.
hygroscopic liquid. It is extremely unstable in Anderson suggests that arsenic, antimony, and
concentrated form and may explode by shock or tin may also be incompatible with perchloric
concussion when dry or drying, so commercially electrolytes (27). Do not store these electrolytes
available perchloric acids come in concentrations . Discard them when they become colored by
of 65 to 72%. In this form, contact with perchloric dissolved metallic ions (from electropolishing).
acid will cause irritation and burns, while its fumes Always keep these solutions cool; increasing the
are highly irritating to the mucous membranes. temperature increases the oxidizing power of
Contact with organic, or other easily oxidized perchloric acid.

108
Laboratory Safety

Figure 66. Diagram developed by Médard, Jacquet and Sartorius showing safe (A and 1 to 9) nonexplosive
electropolishing solutions and explosive (C, E and Los Angeles) compositions of perchloric acid, acetic anhydride
and water (brought in by the perchloric acid, but not corrected for the effect of acetic anhydride).

Comas et al. have studied the hazards associated Acetic anhydride [(CH3CO)2O], or acetic oxide, is a
with mixtures consisting of butyl cellosolve and colorless liquid with a very strong acetic odor. It
from 10 to 95% of 70% perchloric acid (31). Mix- can cause irritation and severe burns to the skin
tures with 60 to 90% acid were explosive at room and eyes. Acetic anhydride decomposes on heat-
temperature. Acid concentrations of 30% or less ing producing toxic fumes. It reacts violently with
were inflammable but were judged to be safe to boiling water, steam, strong oxidants (specifically
use as long as the operating temperature does sulfuric acid), alcohols, amines, strong bases, and
not exceed 20°C. others. It attacks metals, and is very corrosive,
especially in presence of water or moisture. It is
Acetic acid (CH3COOH) is a clear, colorless liquid extremely flammable and should be avoided. The
with a pungent odor. It is a weak acid that reacts electrolytic polishing mixtures of acetic anhydride
with strong oxidizers, bases and metals. It is flam- and perchloric acid (4-to-1 to 2-to-1 mixtures)
mable and is not easily ignited, although when developed by Jacquet, as mentioned above, are
heated, it releases vapors that can be ignited and exceptionally dangerous and should never be
can travel some distance to an ignition source. used. Dawkins (33) reported an accident involv-
Contact with the skin results in serious burns. ing a mixture of chromium trioxide and acetic
Inhalation of the fumes irritates the mucous mem- anhydride that had been used for electropolishing
branes. Anderson states that acetic acid is a good (Table 49).
solvent for nitric acid and that a 50% solution can
be prepared, but not stored, without danger (27). Citric acid [C3H4(OH)(COOH)3·H2O] comes as col-
Sax, however, states that mixtures of nitric and orless, odorless crystals that are water-soluble.
acetic acids are dangerous (32).

109
Laboratory Safety

It is an irritant to the skin, eyes and respiratory tract; dioxide and water from the air. Solutions stored in
no unusual problems are encountered except for flasks with ground glass stoppers may leak air and
occasional allergic reactions. freeze the stoppers, making re-opening difficult.
Dissolving NaOH or KOH in water will generate
Lactic acid (CH3CHOHCOOH) is a yellow or colorless considerable heat. Do not dissolve either in hot
thick liquid. It is damaging to the eyes. water. Never pour water onto these hydroxides;
always add the pellets slowly to the water. Alkali
Oxalic acid (COOHCOOH·2H 2 O) comes as metal hydroxides react violently with acid, and are
transparent, colorless crystals. It is poisonous if in- corrosive in moist air to metals like zinc, aluminum,
gested and irritating to the upper respiratory tin and lead forming flammable/explosive hydro-
tract and digestive system if inhaled. Skin contact gen gas. They are very corrosive to skin, eyes and
produces caustic action and will discolor and respiratory tract. Long-term exposure may lead
embrittle the fingernails. It is not compatible to dermatitis. Potassium hydroxide is somewhat
with nitric acid, as it reacts violently with strong more corrosive than sodium hydroxide.
oxidants. It can also form explosive compounds
due to reacts with silver.
Other Chemicals
Picric acid [(NO2)3C6H2OH], or 2,4,6-trinitrophenol, Hydrogen peroxide (H2O2) is available as a liquid in
comes as yellow crystals that are wet with 10 to concentrations of either 3 or 30%. The 3% solution
35% water. When picric acid is dry, it is a danger- is reasonably safe to use, while the 30% solution is
ous explosive. It is toxic and stains the skin and a very powerful oxidant whose effect on the skin
is incompatible with all oxidizable substances. is about as harmful as that produced by contact
Picrates, which are metal salts of picric acid, are with sulfuric acid. Hydrogen peroxide by itself is
explosive.When picrates are dry, they can detonate not combustible, but if brought in contact with
readily, possibly spontaneously. Purchase in small combustible materials, it can produce violent
quantities, keep it moist, and store it in a safe, cool combustion. Hydrogen peroxide is very damaging
place. If it starts to dry out, add a small amount of to the eyes. Because release of oxygen can cause
water to keep it moist. The maximum solubilities high pressures to develop within the container,
of picric acid in water and in ethanol are about the container caps are vented.
1.3 and 8 g per 100 mL, respectively. Picral can be
stored safely. During use, the solution should not Bromine (Br2), a fuming reddish brown liquid with
be allowed to dry out. The etching residue should a pungent, suffocating odor, is commonly used
be discarded at the end of the day to avoid poten- in deep-etching solutions. It is very corrosive,
tial explosions. reacting violently with easily oxidized substances,
including some metals. Bromine is a dangerous
liquid that should only be handled by well-quali-
Bases fied personnel. Its vapors are extremely irritating
Bases, such as ammonium hydroxide (NH4OH), to the eyes, skin, and mucous membranes. Skin
potassium hydroxide (KOH), and sodium contact produces deep, penetrating burns that
hydroxide (NaOH), are commonly used in metallo- are slow to heal. Contact with organic matter can
graphy, chiefly in etchants. cause fires.

Ammonium hydroxide is a colorless liquid with a Chromic acid (H2CrO4) is formed when chromium
strong, obnoxious odor. Solutions are extremely trioxide (CrO3) is dissolved in water. CrO3 is used in
corrosive and irritating to the skin, eyes, and electropolishing solutions (see previous comment
mucous membranes. It reacts exothermically and Table 49 about explosive nature of mixtures
with sulfuric acid and other strong mineral acids, with acetic anhydride). Dilute aqueous solutions
producing boiling solutions. are widely used for attack polishing. It is a powerful
oxidant; always wear gloves when using it for at-
Sodium and potassium hydroxides are strong bas- tack polishing or use automatic devices and avoid
es, available as white deliquescent pellets that are contact potential. Chronic or long-term inhalation
soluble in water. They can rapidly absorb carbon exposure may produce asthma-like reactions.

110
Laboratory Safety

Potassium permanganate (KMnO4), a black crystal- that can be encountered are summarized; others
line powder, is a powerful oxidant used in etchants. undoubtedly exist that are not covered.
It is a dangerous fire and explosion hazard, es-
pecially when in contact with organic materials. Most accidents can be prevented by simple com-
Ingestion produces serious damage. KMnO4 and monsense rules. It is best to assume that all metal
sulfuric acid should never be mixed together be- dust and all chemicals are hazardous. Inhalation
cause a violent explosion can result. of dust and fumes, ingestion, or bodily contact
should be avoided. Personal protective equip-
Potassium dichromate (K2Cr2O7), a bright orange ment is very useful, but it should not be used as a
crystalline powder, is another powerful oxidant substitute for good laboratory procedures. The use
that is used in etchants. Contact can cause ul- of such equipment does not guarantee freedom
ceration of the hands, severe damage to nasal from injury.
tissue, or asthma and allergies with long-term
exposure. The metallographic literature contains many ref-
erences to the use of dangerous materials, often
Cyanide compounds are occasionally used in without any mention of the dangers involved
metallographic applications. Potassium cyanide or safe handling procedures. This is unfortunate
(KCN) and sodium cyanide (NaCN) are extremely because the unwary may be injured. Many of us
dangerous and highly toxic. Exposure by eye or are tempted to experiment when a recommended
skin contact, or by ingestion, is fatal. NaCN and procedure does not work as claimed. The develop-
KCN vapors are intensely poisonous. They are par- ment of electrolytes, chemical polishing agents,
ticularly hazardous when brought in contact with or etchants should be left to those who are fully
acids or acid fumes because of liberation of hydro- versed in the potential dangers. Metallographic
gen cyanide, which is extremely toxic and highly laboratories should have some of the referenced
flammable. Potassium ferricyanide (K3Fe(CN)6), a safety publications (see bibliography) readily
ruby-red crystalline powder and an ingredient in available in the laboratory and these safety pub-
Murakami-type reagents, is poisonous but stable lications should be consulted when working with
and reasonably safe to use. new or infrequently used materials.

A number of nitrates, such as ferric nitrate Helpful Hints For

[Fe(NO 3) 3·6H 2O], lead nitrate [Pb(NO 3) 6], and Laboratory Safety
silver nitrate (AgNO3), are employed by metallo-
Purchase relatively small quanti-
graphers. Since they are powerful oxidizers, they
ties of chemicals that are used
pose a dangerous fire hazard, especially when in
routinely, especially those that have a short
contact with organic materials. They may evolve
shelf life, or are dangerous.
toxic fumes, such as oxides of nitrogen and lead,
and are poisonous and corrosive to the eyes, skin After pouring acids, carefully replace the cap,
and respiratory tract. wash the exterior of the bottle under running
water, and dry off the surface with a paper
Summary towel before replacing the acid bottle in its
fireproof cabinet.
In general, the metallographic laboratory is a
reasonably safe environment. However, depend- Schedule regular time periods for laboratory
ing on the materials being prepared, dangerous clean-ups. These should be weekly for high
situations can arise. Some hazards, such as the volume laboratories. A clean lab is an essential
preparation of radioactive or reactive metals, are ingredient in all laboratory safety programs.
quite obvious, while others are not. In the pre-
ceding discussion, some of the potential hazards

111
Summary

SUMMARY Removal rates also decrease as the abrasive size

decreases.Through an understanding of these fac-
Preparation of metallographic specimens is based tors, good, reproducible preparation procedures,
upon scientific principles that are easily under- such as presented in this guide, can be established
stood. Sectioning creates damage that must be for the materials being prepared. Automation in
removed by the grinding and polishing steps if specimen preparation offers much more than
the true structure is to be examined. Each section- reduced labor. Specimens properly prepared us-
ing process exhibits a certain amount of damage, ing automated devices consistently exhibit much
thermal and/or mechanical. Consequently, choose better flatness, edge retention, relief control and
a procedure that produces the least possible dam- freedom from artifacts such as scratches, pull
age and use the correct wheel or blade. Grinding out, smearing and comet tailing, than specimens
also causes damage, with the depth of damage prepared manually.
decreasing as the abrasive size decreases. Materi-
als respond differently to the same size abrasive,
so one cannot generalize on removal depths.

112
References

REFERENCES Van Nostrand Reinhold Co., NY, 1986,

pp. 1-19.
1. Metallography and Microstructures,
Vol. 9, Metals Handbook, 9th ed., American Society 15. E. E. Stansbury, “Potentiostatic Etching,” Applied
for Metals, Metals Park, OH, 1985. Metallography, Van Nostrand Reinhold Co., NY,
1986, pp. 21-39.
2. G. F. Vander Voort, Metallography:
Principles and Practice, ASM International, 16. E. Beraha and B. Shpigler, Color
Materials Park, OH, 1999. Metallography, American Society for Metals, Met-
als Park, OH, 1977.
3. L. E. Samuels, Metallographic Polishing
by Mechanical Methods, 3rd Ed., American 17. G. F. Vander Voort, “Tint Etching,”
Society for Metals, Metals Park, OH, 1982. Metal Progress, Vol. 127, March 1985,
pp. 31-33, 36-38, 41.
4. J. A. Nelson, “Heating of Metallographic Speci-
mens Mounted in ‘Cold Setting’ Resins”, Practical 18. E. Weck and E. Leistner, Metallographic In-
Metallography, Vol. 7, 1970, structions for Colour Etching by
pp. 188-191. Immersion, Part I: Klemm Colour Etching, Vol.
77, Deutscher Verlag für Schweisstechnik GmbH,
5. G. F. Vander Voort, “Trends in Specimen Prepara- Düsseldorf, 1982.
tion,” Advanced Materials &
Processes, Vol. 157, February 2000, 19. E. Weck and E. Leistner, Metallographic In-
pp. 45-49. structions for Colour Etching by
Immersion, Part II: Beraha Colour Etchants
6. G. F. Vander Voort, “Metallography for and Their Different Variants, Vol. 77/II,
Edge Retention in Surface Treated Sintered Car- Deutscher Verlag für Schweisstechnik GmbH, Düs-
bides and Steels,” Industrial Heating, seldorf, 1983.
Vol. 67, March 2000, pp. 85-90.
20. E. Weck and E. Leistner, Metallographic
7. S. L. Hoyt, Men of Metals, ASM, Metals Park, OH, Instructions for Colour Etching by
1979, pp. 99-100. Immersion, Part III: Nonferrous Metals, Ce-
8. J. Benedict, R. Anderson and S. J. Klepeis, “Recent mented Carbides and Ferrous Metals, Nickel
Developments in the Use of the Tripod Polisher for Base and Cobalt Base Alloys,
TEM Specimen Preparation,” Specimen Prepara- Vol. 77/III, Deutscher Verlag für Schweisstechnik
tion for Transmission Electron Microscopy GmbH, Düsseldorf, 1986.
of Materials – III, Materials Research Society, 21. P. Skocovsky, Colour Contrast in
Pittsburgh, PA, 1992, Vol. 254, pp. 121-140. Metallographic Microscopy, Slovmetal,
9. G. Petzow and V. Carle, Metallographic Etch- Ž ilina, 1993.
ing, 2nd ed., ASM International, Materials Park, OH, 22. H. E. Bühler and H. P. Hougardy, Atlas of
1999. Interference Layer Metallography, Deutsche
10. G. F. Vander Voort, “Grain Size Measurement,” Gesellschaft für Metallkunde, Oberursel, Germany,
Practical Applications of Quantitative Metal- 1980.
lography, STP 839, American Society for Testing 23. H. E. Bühler and I. Aydin, “Applications of the
and Materials, Philadelphia, 1984, pp. 85-131. Interference Layer Method,” Applied Metallog-
11. G. F. Vander Voort, “Wetting Agents in Metallog- raphy, Van Nostrand Reinhold Co., NY, 1986, pp.
raphy,” Materials Characterization, Vol. 35, No. 2, 41-51.
September 1995, pp. 135-137. 24. G. F. Vander Voort, “Results of an ASTM
12. A. Skidmore and L. Dillinger, “Etching E-4 Round-Robin on the Precision and Bias of
Techniques for Quantimet Evaluation,” Measurements of Microindentation Hardness
Microstructures, Vol. 2, Aug./Sept. 1971, Impressions,” Factors that Affect the
pp. 23-24. Precision of Mechanical Tests, ASTM STP
1025, ASTM, Philadelphia, 1989, pp. 3-39.
13. G. F. Vander Voort, “Etching Techniques
for Image Analysis,” Microstructural Science,
Vol. 9, Elsevier North-Holland, NY, 1981,
pp. 135-154.
14. G. F. Vander Voort, “Phase Identification
by Selective Etching,” Applied Metallography,
113
References

25. G. F. Vander Voort, “Operator Errors in the Mea- Norman V. Steere (ed.), Handbook of
surement of Microindentation Hardness,” Laboratory Safety, 2nd ed., CRC Press,
Accreditation Practices for Inspections, Boca Raton, FL., 1971.
Tests, and Laboratories, ASTM STP 1057, A. Keith Furr (ed.), Handbook of Laboratory
ASTM, Philadelphia, 1989, pp. 47-77. Safety, 3rd ed., CRC Press, Boca Raton, FL., 1989.
26. G. F. Vander Voort and Gabriel M. Lucas, “Microin- L. J. Diberardinis et al., Guidelines for Labora-
dentation Hardness Testing,” tory Design: Health & Safety
Advanced Materials & Processes, Vol. 154, No. Considerations, J. Wiley & Sons, N.Y., 1987.
3, September 1998, pp. 21-25.
S. R. Rayburn, The Foundations of
27. R. L. Anderson, “Safety in Metallographic Labora- Laboratory Safety, Brock-Springer Series in Con-
tory,” Westinghouse Research Laboratory temporary Bioscience, Springer-Verlag, N.Y., 1989.
Science Paper, No. 65 – 1P30 – METLL – P2,
March 29, 1965. C. A. Kelsey and A. F. Gardner (eds.),
Radiation Safety for Laboratory
28. G. F. Vander Voort, Metallography: Principles Technicians, Warren H. Green, Inc.,
and Practice, ASM International, Materials Park, St. Louis, MO., 1983.
OH., 1999, pp. 148-159.
N. I. Sax, Dangerous Properties of Industrial
29. R. C. Nester and G. F. Vander Voort, “Safety in the Materials, 5th ed., Van Nostrand Reinhold Co., N.Y.,
Metallographic Laboratory,” Standardization 1979.
News, Vol. 20, May 1992,
pp. 34-39. Prudent Practices for Handling Hazardous
Chemicals in Laboratories, National
30. P. A. Jacquet, “The Safe Use of Perchloric-Acetic Academy Press, Washington, D.C., 1981.
Electropolishing Baths,” Met. Finish, Vol. 47, 1949,
pp. 62-69. Prudent Practices for Disposal of
Chemicals from Laboratories, National Acad-
31. S. M. Comas, R. Gonzalez Palacin and D. Vassallo, emy Press, Washington, D.C., 1983.
“Hazards Associated with Perchloric Acid-Butylcel-
losolve Polishing Solutions,” Metallography, Vol. 7, N. Proctor and J. Hughes, Chemical Hazards in
1974, pp. 47-57. the Workplace, J. B. Lippincott Co., Philadelphia,
1978.
32. N. I. Sax, Dangerous Properties of Industrial
Materials, 5th ed., Van Nostrand Reinhold Co., Inc., F. A. Patty (ed.), Industrial Hygiene and
N.Y., 1979. Toxicology, Volume II – Toxicology, 3rd ed.,
Wiley-Interscience, N.Y., 1980.
33. A.E. Dawkins, “Chromic Acid-Acetic Anhydride
‘Explosion,’” J. Iron and Steel Institute, Vol. 182, L. Bretherick, Handbook of Reactive Chemical
1956, p. 388. Hazards, 2nd ed., Butterworths, London, 1979.
R. J. Lewis, Sr. (ed.), Hawley’s Condensed Chemi-
Selected Bibliography of Books on cal Dictionary, 12th ed., Van Nostrand Reinhold,
Laboratory Safety New York, 1993.
ASTM E-2014, “Standard Guide on Metallo-
graphic Laboratory Safety”. Useful Web sites on laboratory safety:
N. Van Houten, Laboratory Safety Standards http://ull.chemistry.uakron.edu
for Industry, Research, Academe, Sci. Tech. http://www.hhmi.org/research/labsafe/lcss/lcss.
Pubs., Lake Isabella, CA. 1990. html
E. Gershey, A. Wilkerson and E. Party, http://www.cdc.gov/niosh/ipcsneng/nengsyn.html
Laboratory Safety in Practice, Van Nostrand http://www.state.nj.us/health/eoh/rtkweb/ rtkhsfs.
Reinhold, N.Y., 1991. htm
A. A. Fuscaldo et al. (eds), Laboratory Safety:
Theory & Practice, Academic Press, San Diego,
CA. 1980.
S. B. Pal (ed.), Handbook of Laboratory Health
& Safety Measures, Kluwer Academic, Norwell,
MA., 1985.
114
Appendices

APPENDICES

Classification of Metals Using Common Characteristics and the Periodic Table

Atomic Atomic Density Melting Crystal

Group Element Symbol Number Weight g/cc@20°C Point,°C Structure
Alkaline Beryllium Be 4 9.015 1.848 1280 cph
Earth, IIA Magnesium Mg 12 24.312 1.74 650 cph
Coinage Copper Cu 29 63.545 8.94 1083 fcc
Metals, IB Silver Ag 47 107.87 10.49 961.9 fcc
Gold Au 79 197 19.32 1064.4 fcc
Group IIB Zinc Zn 30 65.38 7.1 419.57 cph
Cadmium Cd 48 112.4 8.65 320.9 cph
Group III Aluminum Al 13 26.98 2.7 660 fcc
Group IVA Tin Sn 50 118.69 7.29 232 bct
Lead Pb 82 207.19 11.35 327.5 fcc
Group IVB Titanium Ti 22 47.9 4.507 1670 cph
Zirconium Zr 40 91.22 6.44 1853 cph
Hafnium Hf 72 178.49 13.21 2225 cph
Group V Antimony Sb 51 121.75 6.79 630.5 rhombohedral
Bismuth Bi 83 208.98 9.78 271.3 rhombohedral
Group VB Vanadium V 23 50.94 6.1 1895 bcc
Niobium/ Nb/
Columbium Cb 41 92.91 8.66 2468 bcc
Tantalum Ta 73 180.95 16.63 2996 bcc
Group VIB Chromium Cr 24 52 7.14 1867 bcc
Molybdenum Mo 42 95.94 9.01 2615 bcc
Tungsten W 74 183.85 19.3 3410 bcc
Group VIIB Manganese Mn 25 54.94 7.3 1244 complex cubic
Rhenium Re 75 186.22 21.04 3175 cph
Group VIII Iron Fe 26 55.85 7.874 1536 bcc
Cobalt Co 27 58.93 8.832 1495 cph
Nickel Ni 28 58.71 8.8 1452 fcc
Platinum Ruthenium Ru 44 101.1 12.1 2310 cph
Metals, Rhodium Rh 45 102.91 12.44 1963 fcc
Group VIII Palladium Pd 46 106.4 12.16 1551 fcc
Osmium Os 76 190.2 22.48 3015-3075 cph
Iridium Ir 77 192.2 22.42 2410 fcc
Platinum Pt 78 195.09 21.37 1772 fcc

Transformation Temperatures for Allotropic Metals Listed Above

Element Transformation Temperature (C°/F°)
Beryllium (Be) α↔β 1256/2293
Cobalt (Co) ε↔α 427/801
Hafnium (Hf ) α↔β 1742/3168
Iron (Fe) Curie 770/1418
α↔γ 912/1674
γ↔δ 1394/2541
Manganese (Mn) α↔β 707/1305
β↔γ 1088/1990
γ↔δ 1139/2082
Nickel (Ni) Curie 358/676
Tin (Sn) α↔β 13.0/55.4
Titanium (Ti) α↔β 883/1621
Zirconium (Zr) α↔β 872/1602
Note: The Curie temperature is the temperature where a magnetic material losses its magnetism upon heating above and regains it upon
cooling below.

115
Appendices

Abrasive Cutting Troubleshooting Guides

Problem Possible Cause Suggested Remedy
Burning (bluish discoloration) Overheated specimen Increase coolant flow rate;
reduce cutting pressure; select a
wheel with softer bonding (faster
breakdown)

Rapid wheel wear Wheel bonding breaks down Select a wheel with harder bonding;
too rapidly reduce cutting pressure

Frequent wheel breakage Uneven coolant distribution; Adjust coolant flow to be even on
loose specimen fixturing; both sides of the wheel; clamp the
abrupt contact with specimen specimen more tightly; start cut
contact carefully

Resistance to cutting Slow wheel bond breakdown Select a wheel with softer bonding;
use a “pulse” cutting mode; use
cutter with oscillating motion or
with minimal area of contact
cutting ability

Stalled wheel Inadequate cutter capacity; Use cutter with greater horsepower;
pinched blade due to movement tighten the clamp on one side less
of specimen than on the other side; reduce
pressure or feed rate; use a cutter
with oscillating motion or with minimal
area of contact cutting ability

Abrasive Wheel Troubleshooting Guides

Pointed: Chisel: Glazed Surface:

Wheel too hard Nonuniform coolant flow Wrong wheel

Normal: Normal: Normal:

For solid materials For structural or For light tubing or
medium wall tubing other thin wall section

116
Appendices

Compression (Hot) Mounting Troubleshooting Guide

Thermosetting Resins (Epoxies, Diallyl Phthalates, and Phenolics)

Defect Probable Cause Suggested Remedy
Radial splits Specimen cross sectional area Use a larger mold size; decrease
or cracks too large; specimen with the specimen size; bevel sharp
sharp corners corners, if possible
Shrinkage gaps Specimen surfaces dirty; specimen Clean and dry specimens carefully;
cooled quickly after polymerization; after polymerization, cool under
wrong resin used pressure to near ambient; use
Epomet G or Epomet F resins
Circumferential cracks Resin contained moisture Keep resins dry during storage;
keep containers closed when not
using; dry resin by baking at
38-49°C (100-120°F)
Bulging or soft mount Inadequate curing Increase polymerization time
(polymerization) time and pressure

Mount appears Time at temperature too short; Increase polymerization time,
grainy and unfused temperature for polymerization temperature and pressure
too low; molding pressure too low

Thermoplastic Resins (Acrylics)

Defect Probable Cause Suggested Remedy
Cottonball Incomplete polymerization of resin; Use less resin; use longer
not enough time at temperaure heating and cooling periods;
use controlled linear cooling
Crazing Relief of internal stresses Cool mount to a lower temperature
upon ejection of mount before ejection; use controlled
linear cooling

Comparison of Six Edge Retention Compression Mounting Compounds

PhenoCure phenolic resin EpoMet thermosetting epoxy resin ProbeMet Cu-filled conductive resin

KonductoMet C-filled conductive EpoxiCure cast epoxy resin with SamplKwick cast acrylic resin with
resin Conductive Filler particles Conductive Filler particles

Micrographs showing the as-forged surface of a hardened modified 5130 alloy steel part mounted using a variety of
resins showing different degrees of edge retention. The specimens were polished simultaneously in the same holder
and were etched with 2% nital. The magnification bars are 20 µm long. Best results were obtained with EpoMet,
ProbeMet and EpoxiCure resin with the Conductive Filler particles.
117
Appendices

Castable Mounting Resin Selection Guide

Name Typical Cure Time Uses and Restrictions
SamplKwick 5-8 minutes Least expensive resin; used for mounting printed circuit boards
Acrylic Resin and polymers; good for x-ray diffraction work (no peak interferences);
high shrinkage (not recommended for edge retention)

VariKleer 10-15 minutes Designed for use with a pressure vessel, this general use acrylic
hardens crystal clear with minimal shrinkage. To achieve crystal clear
samples, reusable silicon or EPDM molds should bot be used.

VariDur 3000 15-30 minutes Minimal shrinkage, good edge retention

VariDur 10 minutes Inexpensive; (good edge rentention for acrylic); used with many metals
Acrylic Resin when the edge detail is not important

EpoKwick 90 minutes Fast curing epoxy; more shrinkage than other epoxies; viscosity 250-
Epoxy Resin 400 cps at 25 °C; not recommended for vacuum infiltration of voids due
to fast curing time; physically adheres to specimen; gives acceptable
edge retention; good for most specimens

EpoColor 90 minutes Provides color contrast between mount and specimen that can be
Epoxy Resin helpful when studying edges; viscosity 400-700 cps at 25 °C; good
for most specimens; but not recommended for vacuum impregnation

EpoxiCure 6 hours General purpose epoxy; viscosity 400-600 cps at 25 °C, good
Epoxy Resin adherence to specimen; can be used to vacuum impregnate
voids (viscosity can be reduced by warming to 50 °C); good for heat- sensitive specimens (very
low exotherm during polymerization)

EpoxiCure NC 5 hours General purpose epoxy - No hazard shipping fees.

EpoHeat 90 minutes Fast, heat cure epoxy system with low shrinkage and low viscosity
Low Viscosity Epoxy (32 cps at 176 °F (80 °C)). Excellent for impregnation. Requires oven
cure. Golden color.

EpoThin 9 hours Lowest viscosity (can be further reduced by warming resin to 50 °C)
Epoxy Resin and best penetration of voids during vacuum impregnation; good
adherence to specimen; good for heat-sensitive specimens
(lowest exotherm during polymerization)
Curing time is not a precise quantity as it can be influenced by the nature of the mold (how well, or how poorly, the exothermic heat of polymerization is extract-
ed), by the volume of the cast resin used, by the volume and nature of the specimen being encapsulated, by the temperature of the mix, by room temperature
and air circulation and by the age of the resin and hardener (i.e., curing time is longer for products beyond their normal shelf life). The values listed are typical for
metallic specimens cast in 1.25-inch (30-mm) diameter Sampl-Kup molds at 21 °C (70 °F) using fresh resin and hardener.

Greek Alphabet and its English Equivalents

Greek Letter Name English Equivalent Greek Letter Name English Equivalent
Α, α Alpha A, a Ν, ν Nu N, n
Β, β Beta B, b Ξ, ξ Xi X, x
Γ, γ Gamma G, g Ο, ο Omicron O, o
∆, δ Delta D, d Π, π Pi P, p
Ε, ε Epsilon E, e Ρ, ρ Rho R, r
Ζ, ζ Zeta Z, z Σ, σ Sigma S, s
Η,η Eta E, e Τ, τ Tau T, t
Θ, θ Theta Th Υ, υ Upsilon U, u
Ι, ι Iota I, i Φ, ϕ Phi Ph
Κ, κ Kappa K, k Χ, χ Chi Ch
Λ, λ Lambda L, l Ψ, ψ Psi Ps
Μ, µ Mu M, m Ω, ω Omega O, o

118
Appendices

Cast Resin (“Cold” Mounting) Troubleshooting Guide

Defect Probable Cause Suggested Remedy

Non-Curing Expired resin and/or hardener; Measure weights of resin and hardener;
Wrong resin – to – hardener ratio after prescribed time, and it has not
hardened, heat the mold, liquid and
specimen to 50 °C (120 °F) to force curing;
check age before using resin and hardener
Slow Curing Incomplete mixing; too little hardener Mix thoroughly; measure weights of resin
used; curing at too low a temperature and hardener; place in warm area*
Rapid Curing Too much hardener; curing at too high Measure weights of resin and hardener;
a temperature; mold size too large mix smaller quantities; increase air
(excessive resin volume) circulation; place in a cooler place
Gas Tunnels Excessive hardener used; mold size Measure weights of resin and
too large (excessive resin volume); hardener; increase air circulation;
use of a low thermal conductivity mold use higher thermal conductivity
molds
Mount Stuck in Mold Did not use mold release agent Apply Silicon Mold Release or Release Agent
to interior surfaces of mold
Solvent Softening Poor resistance to solvents; Heat the mold, liquid and specimen
mount may not be fully cured to 50 °C (120 °F) to force curing
Excessive Shrinkage Polymerization is too fast; too much Use an epoxy rather than an arcylic; use a
hardener used; wrong resin selected slow curing epoxy; add Flat Edge Filler
particles; use the correct resin – to –
hardener ratio
* Placing in an oven at 120°F (50°C) for up to two hours may accelerate the curing of epoxies

Guide to Fine Polishing Abrasives

Name Characteristics Applications
MicroPolish 1-, 0.3- and 0.05-µm alumina powder Inexpensive general-purpose abrasive
alumina powder for polishing most minerals and metals
MicroPolish II 1-, 0.3- and 0.05-µm deagglomerated Higher quality alumina suitable for
deagglomerated alumina powders and 5-, 1-, 0.3- and polishing most minerals and metals;
alumina powder 0.05-µm deagglomerated alumina produces a better surface finish than
and suspensions suspensions standard alumina abrasives of the same
size
MasterPrep 0.05-µm sol-gel alumina suspension Unique seeded-gel alumina produces
agglomerate-free better finishes than calcined alumina
seeded-gel alumina abrasives; excellent with minerals, metals,
suspension carbides, PCBs. Good for percious metals.
MasterMet 0.06-µm amorphous silica colloidal Chemo-mechanical action produces
colloidal silica suspension with a pH of ~10 producing a excellent surfaces for metals, minerals,
suspension chemo-mechanical polishing action polymers, PCBs, electronic devices,
etc., especially soft metals, but not for
precious metals
MasterMet 2 0.02-µm non-crystallizing amorphous Non-crystallizing version of colloidal silica
non-crystallizing silica colloidal suspension with a pH yields similar results as Mastermet but
colloidal silica of ~10.5 producing chemo-mechanical without the danger of crystallized silica
suspension polishing action particle contamination
MasterPolish Proprietary viscous blend of high Proprietary blend has little water and can
blended alumina purity 0.05-µm alumina and colloidal be used with materials that are sensitive
and colloidal silica with a slightly basic pH (can to water, such as Mg and most of its alloys;
silica suspension be diluted with water) good for most Fe-, Ni- or Co-based alloys
and MMCs
MasterPolish 2 Proprietary 0.06-µm abrasive suspension Proprietary blend designed for final
proprietary with a pH of ~10 for chemo-mechanical polishing of ceramic materials, nitrides,
chemo-mechanical polishing action carbides; and composites
suspension

119
Appendices

A Guide to Equivalent Loads for Different Mount Sizes

The two graphs plot the load applied per specimen, and the resulting pressure, as a function of mold diameter for
both Imperial and metric units. An example of the use of the graphs is shown. Assume that we have a practice
developed using 1.25-inch diameter mounts that calls for 5 lbs. load per specimen. We want to use this method with
mounts that are 1-inch in diameter. How much load should we use? Draw a line vertically at 5 lbs. (shown in green)
until it intersects the line for a 1.25-inch diameter mount. Then draw a horizontal line (shown in blue) from the
intersection point to the line for 1-inch diameter mounts. Then, draw a vertical line (shown in blue) down to the load
axis. This point is about 3 lbs. So, we would use 3 lbs per specimen for 1-inch mounts. In similar manner, if we want
to use 1.5-inch diameter specimens, the graph would suggest that we use about 7 lbs. per specimen (red lines).

120
Appendices

Preparation Troubleshooting Guide

Troubleshooting Definitions
Comet Tails: Dull, broad depressed lines emanating from a hard particle or from a hole in a specimen in a pattern that resembles the tail
of a comet. The hard particles may be nonmetallic inclusions or nitrides. This appears to be a material-specific phenomenon as repeated
efforts to produce this defect in randomly chosen specimens fail. Historically, with manually-prepared specimens, it was claimed to be
caused by unidirectional grinding; however, repeated efforts to create comet tails with only unidirectional grinding fail. Comet tails have
been observed on susceptible specimens prepared with automated machines using only complementary grinding and polishing with
specific combinations of head and platen speeds. In a specimen prone to comet tailing, using contra rotation in the final step will enhance
the problem compared to complementary rotation.

Edge Rounding: The edge or edges of a specimen are abraded at a faster rate than the interior leading to a change in flatness at the
edge such that the edge cannot be brought into focus along with the interior with the light microscope (this condition is more critical as
the magnification is raised due to the higher numerical aperture of the objective which reduces the depth of focus).

Embedding: Hard abrasive particles that become fixed in the surface of the softer specimen. This is a common problem with the low melt-
ing point alloys (e.g., Pb, Sn, Bi, Cd, Zn), aluminum (mainly fine diamond), and precious metals, but has been observed with refractory metals
such as titanium.

Pull Out: Removal of second phase particles (either softer or harder than the matrix) during preparation. Pull out may be enhanced if the
interface between the particle and matrix is weak or if the particle is particularly brittle.

Relief: Excessive height differences between second-phase particles and the matrix due to differences in grinding and polishing rates
between the matrix and the particles. Soft particles abrade faster than the matrix and are recessed below the matrix while the reverse is
observed for hard particles.

Scratches: A linear cut along the surface of a specimen due to contact with a correctly oriented abrasive particle; a groove caused by
the abrasion process. The width and depth of the cut is a function of the size of the abrasive particle, its angle to the specimen surface, the
applied pressure, and other factors. Scratches become a problem when they are excessively coarse for the abrasive size used, indicating
a problem with the preparation method. Very fine scratches may not be a problem in routine examination for grain size or inclusions but
could be a problem in failure analysis if critical fine detail is obscured.

Smear: Matrix flow over voids, cracks, or second phase particles that make detection of these features and measurement of their extent
difficult or impossible.

Stain: A contamination residue on the surface of a specimen, that may form around second-phase particles due to interactions between
the specimen and abrasives and/or lubricants, or may form on the matrix due to inadequate cleaning or drying or may form due to interac-
tions between the specimen and solvents after preparation or etching.

121
Appendices

Preparation Troubleshooting Guide Cont.

Defect Probable Cause Suggested Remedy

Comet Tails Poorly bonded very hard phase in softer Use hard, napless cloths; reduce applied pressure;
matrix; or, pores in matrix results in reduce step times; impregnate pores with epoxy or
unidirectional grooves emanating from wax; use complementary rotation. In manual
particles or hole. preparation, avoid unidirectional grinding.
Edge Rounding Shrinkage gaps between specimen and Avoid gaps by cooling thermosetting mount under
mount are main problem and, when pressure after polymerization. Use EpoMet resin
present, avoid napped cloths and long (least prone to shrinkage gaps); of castable resins,
polishing times. epoxy is best (adding Flat Edge Filler particles to
epoxy improves edge retention); use central force
rather than individual force; use ApexHercules rigid
grinding discs; use napless cloths.
Embedding Hard abrasive tends to embed in soft Coat SiC abrasive paper with paraffin wax
metals; finer particles are more likely (candlewax is best, soaps are also effective);
to embed than larger particles. beeswax is less effective; reduce applied pressure
rpm and times; avoid finer SiC abrasive papers;
avoid using diamond slurries with fine diamond sizes
(for M 3-mm diamond, paste embeds less than slurries).
SiC embedded in soft metals (e.g., Pb) can be removed
by polishing with alumina slurries.
Pull Outs Excessive grinding time on worn SiC Grind no more than 60 seconds on a sheet of SiC
paper; excessive polishing time on paper; use napless cloths; reduce applied pressure,
napped cloths; excessive pressure; use proper degree of lubrication.
and inadequate lubrication lead to pull
out of second-phase particles.
Relief Use of napped cloths, long polishing times, Use napless cloths, higher pressures and shorter
and low pressure create height differences times. If relief is observed using contra rotation in
between matrix and second-phase particles. last step, repeat the last step using complementary
rotation.
Scratches Contamination of grinding or polishing Maintain clean operating conditions, washing hands and
surfaces by coarser abrasives; pull out equipment between steps; for cracked or porous
of hard particles from matrix or broken specimens, use ultrasonic cleaning after each step
pieces of brittle materials; or inadequate (especially polishing); execute each step thoroughly
grinding or polishing times (avoid short cutting methods); when using the Apex
and pressures can leave a heavy magnetic disc
system, store the discs in a clean scratch pattern.
Drawer or cabinet and scrub surfaces if contamination
is believed to have occurred.
Smear Flow of softer metals may be caused by Use proper degree of lubrication, lower pressures
inadequate lubrication, excessive and rpms; use a medium napped cloth on final step;
pressures or rpms. lightly etch the specimen after final polishing and
repeat the last step; use the vibratory polisher to
remove smeared surface metal.
Stain Stains around second-phase particles can Some staining issues are unique to the specimen
be induced by interactions between but most result from using impure tap water (switch
abrasive and particle, perhaps affected to distilled water; sometimes de-ionized water is
by water quality; or they can occur due needed) or inadequate drying. When using colloidal
to inadequate cleaning, or improper silica, which can be hard to remove, direct the water
solvents used after the abrasive step. jet onto the cloth and wash the specimens (and
cloth) for the last 6-10 seconds of the polishing cycle.
This makes subsequent cleaning easy. Otherwise,
scrub the surface with water containing a detergent
solution using cotton. Then, rinse with alcohol and blow
dry with hot air, such as with the TorraMet dryer.
Compressed air systems can contain impurities,
such as oils, that can stain the surface.

122
ASTM Metallography Standards

ASTM METALLOGRAPHY E 562: Determining Volume Fraction by Systematic

STANDARDS Manual Point Count

A 247: Visual Classification of Graphite in the Micro- E 766: Calibrating the Magnification of a Scanning
structure of Cast Iron Electron Microscope

A 892: Defining and Rating the Microstructure of E 768: Preparing and Evaluating Specimens for
High Carbon Bearing Steels Automatic Inclusion Assessment of Steel

B 390: Evaluating Apparent Grain Size and Distribu- E 807: Metallographic Laboratory Evaluation
tion of Cemented Tungsten Carbides E 883: Reflected-Light Photomicrography
B 588: Measurement of the Thickness of Trans- E 930: Estimating the Largest Grain Observed in a
parent or Opaque Coatings by Double-Beam Metallographic Section (ALA Grain Size)
Interference Microscope Technique
E 975: X-Ray Determination of Retained Austenite
B 657: Metallographic Determination of Microstruc- in Steel with Near Random
ture in Cemented Tungsten Carbide Crystallographic Orientation
B 681: Measurement of Thickness of Anodic E 986: Scanning Electron Microscope Beam Size
Coatings on Aluminum and of Other Transparent Characterization
Coatings on Opaque Surfaces Using the Light-Sec-
tion Microscope E 1077: Estimating the Depth of
Decarburization of Steel Specimens
B 748: Measurement of the Thickness of
Metallic Coatings by Measurement of Cross E 1122: Obtaining JK Inclusion Ratings Using Auto-
Section with a Scanning Electron Microscope matic Image Analysis

B 795: Determining the Percentage of Alloyed or E 1180: Preparing Sulfur Prints for
Unalloyed Iron Contamination Present in Powder Macrostructural Examination
Forged Steel Parts E 1181: Characterizing Duplex Grain Sizes
B 796: Nonmetallic Inclusion Level of Powder E 1245: Determining the Inclusion or Second-Phase
Forged Steel Parts Constituent Content of Metals by Automatic Image
B 847: Measurement of Metal and Oxide Coating Analysis
Thickness by Microscopical E 1268: Assessing the Degree of Banding or Orien-
Examination of a Cross section tation of Microstructures
C 664: Thickness of Diffusion Coating E 1351: Production and Evaluation of Field Metal-
E 3: Preparation of Metallographic Specimens lographic Replicas

E 7: Standard Terminology Relating to E 1382: Determining Average Grain Size Using

Metallography Semiautomatic and Automatic Image Analysis

E 45 Determining the Inclusion Content E 1508: Quantitative Analysis by Energy-

of Steel Dispersive Spectroscopy

E 82: Determining the Orientation of a E 1558: Electrolytic Polishing of Metallographic

Metal Crystal Specimens

E 112: Determining Average Grain Size E 1920: Metallographic Preparation of Thermal

Spray Coatings
E 340: Macroetching Metals and Alloys
E 1951: Calibrating Reticles and Light
E 381: Macroetch Testing Steel Bars, Billets, Blooms, Microscope Magnifications
and Forgings
E 2014: Metallographic Laboratory Safety
E 384: Microindentation Hardness of Materials
E 2015: Preparation of Plastics and Polymeric Speci-
E 407: Microetching Metals and Alloys mens for Microstructural Examination
F 1854: Stereological Evaluation of Porous Coatings
on Medical Implants
123
ASTM/ISO Standards

ASTM HARDNESS TESTING ISO 2639: Steel – Determination and

STANDARDS Verification of the Effective Depth of
Carburized and Hardened Cases
B 578: Microhardness of Electroplated
Coatings ISO 3754: Steel – Determination of Effective Depth
of Hardening After Flame or Induction Hardening
B 721: Microhardness and Case Depth of Powder
Metallurgy Parts ISO 3763: Wrought Steels – Macroscopic Methods
for Assessing the Content of
C 730: Knoop Indentation Hardness of Glass Non-Metallic Inclusions
C 849: Knoop Indentation Hardness of ISO 3887: Steel, Non-Alloy and Low Alloy – Determi-
Ceramic Whitewares nation of Depth of Decarburization
C 1326: Knoop Indentation Hardness of Advanced ISO 4499: Hardmetals: Metallographic
Ceramics Determination of Microstructure
C 1327: Vickers Indentation Hardness of Advanced ISO 4524-1: Metallic Coatings – Test
Ceramics Methods for Electrodeposited Gold and Gold Alloy
E 10: Brinell Hardness of Metallic Materials Coatings – Part 1: Determination of Coating Thick-
ness
E 18: Rockwell Hardness and Rockwell Superficial
Hardness of Metallic Materials ISO 4964: Steel – Hardness Conversions

E 92: Vickers Hardness of Metallic Materials ISO 4967: Steel – Determination of Content of Non-
Metallic Inclusions – Micrographic Method Using
E 140: Standard Hardness Conversion Tables for
Standard Diagrams
Metals
ISO 4968: Steel – Macrographic Examination by
E 448: Scleroscope Hardness Testing of Metallic
Sulphur Print (Baumann Method)
Materials
ISO 4970: Steel – Determination of Total
E 1842: Macro-Rockwell Hardness Testing of Metal-
or Effective Thickness of Thin Surface-Hardened
lic Materials
Layers
ISO METALLOGRAPHY ISO 5949: Tool Steels and Bearing Steels – Micro-
STANDARDS graphic Method for Assessing the Distribution of
Carbides Using Reference Photomicrographs
ISO 643: Steels – Micrographic Determination of
the Ferritic or Austenitic Grain Size ISO 9042: Steels – Manual Point Counting Method
for Statistically Estimating the Volume Fraction of a
ISO 945: Cast Iron: Designation of
Constituent with a Point Grid
Microstructure of Graphite
ISO 9220: Metallic Coatings – Measurement of
ISO 1083: Spheroidal Graphite Cast
Coating Thickness – Scanning Electron Microscope
Iron – Classification
Method
ISO 1463: Metallic and Oxide Coatings
ISO 14250: Steel – Metallographic
– Measurement of Coating Thickness –
Characterization of Duplex Grain Size
Microscopical Method
and Distribution
ISO 2624: Copper and Copper Alloys – Estimation of
Average Grain Size

124
ISO/Other Nattional Standards

ISO HARDNESS STANDARDS Other National Standards

ISO 1355: Metallic Materials – Hardness Test France
– Calibration of Standardized Blocks to be Used for NF A04-10: Determination of the Ferritic and Aus-
Rockwell Superficial Hardness Testing Machines tenitic Grain Size of Steels
(Scales 15N, 30N, 45N, 15T, 30T and 45T)
Germany
ISO 4516: Metallic and Related Coatings – Vickers
and Knoop Microhardness Tests DIN 50150: Testing of Steel and Cast Steel; Conver-
sion Table for Vickers Hardness,
ISO 4545: Metallic Materials – Hardness Test Brinell Hardness, Rockwell Hardness and
– Knoop Test Tensile Strength
ISO 4546: Metallic Materials – Hardness DIN 50192: Determination of the Depth of Decar-
Test – Verification of Knoop Hardness burization
Testing Machines
DIN 50600: Testing of Metallic Materials;
ISO 4547: Metallic Materials – Hardness Test Metallographic Micrographs; Picture Scales
– Calibration of Standardized Blocks to be Used for and Formats
Knoop Hardness Testing Machines
DIN 50601: Metallographic Examination;
ISO 6441-1: Paints and Varnishes – Determination Determination of the Ferritic or Austenitic
of Microindentation Harness – Part 1: Knoop Hard- Grain Size
ness by Measurement of
Indentation Length DIN 50602: Metallographic Examination;
Microscopic Examination of Special Steels Using
ISO 6441-1: Paints and Varnishes – Determination Standard Diagrams to Assess the Content of Non-
of Microindentation Harness – Part 2: Knoop Hard- Metallic Inclusions
ness by Measurement of Indentation Depth Under
Load SEP 1510: Microscopic Test of Steels for Grain Size
by Comparison with Standard Charts
ISO 6507-1: Metallic Materials – Hardness Test
– Vickers Test – Part I: HV 5 to HV 100 SEP 1570: Microscopical Examination of Special
Steels for Non-Metallic Inclusions Using Standard
ISO 6507-2: Metallic Materials – Hardness Micrograph Charts
Test – Vickers Test – Part 2: HV 0, 2 to Less Than HV 5
SEP 1572: Microscopic Testing of Free-Cutting
ISO 6508: Metallic Materials – Hardness Test – Rock- Steels for Non-Metallic Sulphide Inclusions by
well Test (Scales A, B, C, D, E, F, G, H, K) Means of a Series of Pictures
ISO 9385: Glass and Glass-Ceramics – Knoop Hard-
ness Test Italy
ISO 10250: Metallic Materials – Hardness Testing UNI 3137: Extraction and Preparation of Samples
– Tables of Knoop Hardness Values for Use in Tests
UNI 3138: Macrographic Analysis
Made on Flat Surfaces
UNI 3245: Microscopic Examination of Ferrous
ISO 14271: Vickers Hardness Testing of Resistance
Materials - Determination of Austenitic or Ferritic
Spot, Projection and Seam Welds (low load and
Grain Size of Plain Carbon and Low-Alloy Steels
microhardness)
UNI 4227: Determination of Metallographic Struc-
tures
UNI 4389: Nonferrous Metals and Their Alloys:
Determination of the Dimension of Their Crystal
Grains

125
Other National Standards

Japan PN-63/H-04504: Microstructure Testing of Steel

JIS B 7724: Brinell Hardness – Verification of Testing Products. Iron Carbide. Ghost Structure. Widma-
Machine nstätten’s Structure.

JIS B 7725: Vickers Hardness – Verification of Testing PN-66/H-04505: Microstructure of Steel Products.
Machines Templates and Evaluation.

JIS B 7730: Rockwell Hardness Test – Calibration of PN-75/H-04512: Nonferrous Metals. Reagents for
Standardized Blocks Revealing Microstructure.

JIS B 7734: Knoop Hardness Test – PN-75/H-04661: Gray, Spheroidal Graphite and
Verification of Testing Machines Malleable Cast Iron. Metallographic Examination.
Evaluation of Microstructure
JIS B 7735: Vickers Hardness Test –
Calibration of the Reference Blocks PN-76/H-04660: Cast Steel and Iron. Microscopic Ex-
amination. Sampling and Preparation of Test Pieces.
JIS B 7736: Brinell Hardness Test – Calibration of
Standardized Blocks PN-84/H-04507/01: Metals. Metallographic Testing
of Grain Size. Microscopic Methods for Determina-
JIS G 0551: Methods of Austenite Grain Size Test for tion of Grain Size.
Steel
PN-84/H-04507/02: Metals. Metallographic Testing
JIS G 0552: Method of Ferrite Grain Size Test of Grain Size. Methods of Revealing the Prior-Aus-
for Steel tenitic Grains in Non-Austenitic Steels.
JIS G 0553: Macrostructure Detecting Method for PN-84/H-04507/03: Metals. Metallographic Testing
Steel, Edition 1 of Grain Size. Macroscopic Method of Revealing the
JIS H 0501: Methods for Estimating Prior-Austenitic Grain Size by the Fracture Method.
Average Grain Size of Wrought Copper and Copper PN-84/H-04507/04: Metals. Metallographic Testing
Alloys of Grain Size. A Method of Testing for Overheating
JIS R 1610: Testing Method for Vickers Hardness of of Steel.
High Performance Ceramics PN-87/H-04514: Steel, Cast Steel, Cast Iron. Macro-
JIS R 1623: Test Method for Vickers Hardness of Fine structure Examination. Baumann’s Test.
Ceramics at Elevated Temperatures
JIS Z 2243: Brinell Hardness Test – Test Method Russia
JIS Z 2244: Vickers Hardness Test – GOST 801: Standard for Ball and Roller
Test Method Bearing Steel

JIS Z 2245: Rockwell Hardness Test – GOST 1778: Metallographic Methods of

Test Method Determination of Nonmetallic Inclusions

JIS Z 2251: Knoop Hardness Test – GOST 5639: Grain Size Determination
Test Method
JIS Z 2252: Test Methods for Vickers Hardness at Sweden
Elevated Temperatures SIS 11 11 01: Estimating the Average Grain Size of
Metals
Poland SIS 11 11 02: Estimating the Austenitic Grain Size of
PN-57/H-04501: Macroscopic Examination of Steel. Ferritic and Martensitic Steels
The Deep Etching Test. SIS 11 11 11: Methods for Assessing the
PN-61/H-04502: Reagents for Macrostructure Tests Slag Inclusion Content in Steel:
of Iron Alloys Microscopic Methods

PN-61/H-04503: Reagents for Microstructure Tests

of Iron Alloys

126
Other National Standards

Sweden (continued) BS 6479: Determination and Verification of Effective

SIS 11 11 14: Determination of Slag Depth of Carburized and Hardened Cases in Steel
Inclusions – Microscopic Methods – BS 6481: Determination of Effective Depth of Hard-
Manual Representation ening of Steel after Flame or Induction Hardening
SIS 11 11 16: Steel – Method for Estimation of the BS 6533: Macroscopic Examination of Steel by Etch-
Content of Non-Metallic Inclusions – Microscopic ing with Strong Mineral Acids
Methods – Jernkontoret’s
BS 6617: Determination of Decarburisation in Steel
Inclusion Chart II for the Assessment of
Non-Metallic Inclusions
SIS 11 03 40: Hardness Test – Vickers Test
HV 0, 2 to HV 100 – Direct Verification of
Testing Machines
SIS 11 03 41: Hardness Test – Vickers Test HV 0,2 to
HV 100 – Indirect Verification of Testing Machines
Using Standardized Blocks
SIS 11 03 42: Hardness Test – Vickers Test
HV 0,2 to HV 100 – Direct Verification of
Standardizing Machine for Calibration of
Standardized Blocks
SIS 11 03 43: Hardness Test – Vickers
Test HV 0,2 to HV 100 – Calibration of
Standardized Blocks

SIS 11 25 16: Metallic Materials – Hardness Test

– Vickers Test HV 5 to HV 100

SIS 11 25 17: Metallic Materials – Hardness Test

– Vickers Test HV 0,2 to Less Than HV 5

SIS 11 70 20: Determination of the Depth of

Decarburization in Steel

United Kingdom
BS 860: Tables for Comparison of Hardness Scales.
BS 4490: Methods for Micrographic Determination
of the Grain Size of Steel
BS 5710: Macroscopic Assessment of the Non-Me-
tallic Inclusion Content of Wrought Steels
BS 6285: Macroscopic Assessment of Steel by
Sulphur Print
BS 6286: Measurement of Total or Effective Thick-
ness of Thin Surface-Hardened Layers in Steel

127
Buehler Trademarks

registered trademarks of unregistered trademarks

Buehler Ltd.: of Buehler Ltd.:
Apex AcuThin
ApexHercules Delta
Buehler EpoColor
CarbiMet PhenoCure
ChemoMet Phoenix Beta
DuoMet Phoenix Vector
EdgeMet Kit TransOptic
EpoHeat TriDent
EpoKwick UltraPad
EpoMet UltraPlan
EpoThin UltraPol
EpoxiCure UltraPrep
FluorMet WhiteFelt
HandiMet
IsoCut
IsoMet
KonductoMet
MasterMet
MasterPolish
MasterPrep
MasterTex
MetaDi
MetaDi Supreme
MicroCloth
MicroPolish
NanoCloth
NelsonZimmer
OmniMet
OscilliMet
PlanarMet
ProbeMet
SamplKup
SamplKwick
SimpliMet
SuperMet
SurfMet
TexMet
TorraMet
VariDur
VibroMet

128
Index

A Central force, 19, 29

Abrasive cut-off machines: Ceramics:
chop cutters, 10 etchants, 77
minimum area of contact cutting, 11 preparation, 57-60
orbital cutting, 11-12 Chemical polish, 41-42
oscillating cutting, 11 Clamp mounting, 14
pulse cutting mode, 10-11 Cobalt and alloys:
traverse-and-increment cutting, 11-12 etchants, 76
Abrasive cut-off wheels: preparation, 50
bond strength, 9 Cold mounting (see castable resins)
recommendations, 10 Colloidal silica, 27-28, 34, 120
thickness, 9 “Comet tails,” 23, 121-122
troubleshooting guide, 116 Composites:
Abrasive sectioning, 7, 9-12, 116 preparation, 59-60
Acids, 107-110 Compression mounting, 14-15, 17-19, 53, 117
Acrylic resins, 15-17, 117-119 Conductive mounts, 16, 117
Alumina, 20, 27, 28, 35, 40 Contra rotation, 29, 30-31
Alumina grinding paper, 20, 29 Coolant flow, 9
Aluminum and alloys: Copper and alloys:
etchants, 70, 71, 73 etchants, 70, 75-76
preparation, 34-35 preparation, 48
ASTM Standards, 123-124 Cutting damage, 7-9, 116
Attack polishing: D
Au, 52 Damage, sectioning, 7-9, 116
Be, 37 Dark field illumination, 65, 85, 86
Co, 50 Diamond, 27, 28
Hf, 41-42 Differential interference contrast, 23, 43, 44, 59, 65,
Nb, V and Ta, 43-44 85, 86
safety, 41 Digital imaging:
Ti, 41 acquisition, 93
Zr, 41-42 archiving, 93

B calibration, 95-96
cameras, 93-94
Bases, 110
clarification, 96
Beryllium
color, 94
preparation, 37
pixels, 94, 95, 96
Bright field illumination, 83,84,86
resolution, 95
Buehler trademarks, 130

C E
Edge preservation, 14, 15, 17-19, 121-122
Castable resins:
Electronic materials:
acrylic resins, 15-17, 118, 119
preparation, 62-64
Conductive Filler particles, 16
Embedded abrasive, 20, 21, 35, 39, 64, 121-122
epoxy resins, 15-17, 118, 119
Epoxy resin, 15-17, 117-119
exotherm, 16-17
Flat Edge Filler particles, 16,18,19
recommendations, 16, 148 129
troubleshooting guide, 119
Index

E (cont.) preparation, 31, 45-47

Etchants, 27, 73-77 ISO standards, 124-125
chemicals, 110-111 K-L
dangerous solutions, 105, 107, 108, 109 Knoop hardness, 88-89
mixing, 103-104 Light microscopy, 78-86
storage, 103 Low-melting point metals:
Etching, 67-77, 78 etchants, 73
anodizing, 71 preparation, 37-38, 62-64
electrolytic, 71
heat tinting, 71-72 M
interference layer method, 72 Magnesium and alloys:
procedure, 67 etchants, 73
selective etching, 68-71 preparation, 36
Magnesium oxide, 34
F Metals:
Ferrous metals: allotropic transformations, 116
etchants, 68-70, 74-75 periodic table, 33
preparation, 31, 45-47 properties, 116
Fluorescent dyes, 16, 53 Method Development, 7
G Microindentation hardness:
Grain size measurement, 7, 99-100 accuracy, precision, bias, 89-92
Grinding: automation, 91-92
alumina paper, 20 indentation size effect, 88
diamond, 19, 20 Knoop test, 88-89
embedding, 20,21 Vickers test, 84-88
emery paper, 20 Microscope, reflected light:
machines, 22 condenser, 80
planar, 20 depth of field, 83-84
SiC paper, 20 diaphragms, 80
“empty” magnification, 82-83
H eye clearance, 81
Hafnium and alloys: eye pieces, 81
etchants, 74 filters, 80
preparation, 41-42 graticules, 81
I illumination modes, 79, 83-86
Image analysis Köhler illumination, 83-84
binary operations, 98-99 light sources, 77, 79
feature-specific measurements, 99 magnification, 81, 82
field measurements, 99 numerical aperture, 80, 82, 83
measurements, 96, 99-100 objective lens, 77, 79, 80
thresholding, 96, 99-100 ray diagram, 78
Inclusion measurement, 7 resolution, 82, 83
Individual force, 19 reticles, 81
Iron and steel: stage, 78, 81-89
etchants, 68-70, 74-75 stand, 82

130
Index

M (cont.) Precision saws, 12-13

Microscope, reflected light (cont): Precious metals:
working distance, 80, 81 etchants, 76-77
Mounting: Preparation, 51-52
castable resins, 14, 15-17 Preparation method development:
clamp mounting, 14 6, 8, 29, 31, 32

compression mounting, 14-15, 17-19, 117 Preparation problems:

conductive resins, 16, 117 “comet tails,” 23, 121-122

edge retention (preservation), 14, 15, 17-19, disturbed metal, 23

117, 121-122 edge retention (preservation), 14, 15, 17-19, 117,
electroless nickel plating, 17, 19 121-122

low-melting point metals, 38 embedded abrasive, 20, 21, 35, 39, 64,
121-122
phenolic resins, 14
excessive relief, 23, 24, 121-122
presses, 17, 19
pitting, 23
purposes, 14
pull outs, 9, 121-122
recommendations, 14, 16
scratches, 121-122
shrinkage gaps, 14, 15, 17, 18, 19
smear, 63, 64, 121-122
thermoplastic resins, 14-15, 18, 117
troubleshooting guide, 121-122
thermosetting resins, 14-15, 17-19, 117
water sensitivity, 36, 37
troubleshooting guides, 117
Preparation procedures
Mount size, 21
contemporary methods, 29-31
N Al and alloys, 34-35
Nickel and alloys: Be, 37
etchants, 76 ceramics, 57-58
preparation, 49 Co and alloys, 50
Nomarski DIC, 19, 43, 44, 59, 65, 85, 86 composites, 59-60

O-P Cu and alloys, 48

Oblique illumination, 84-85 electronic devices, 62-64

Other National Standards, 125-127 Fe and alloys, 45-47

Periodic table of the elements, 33 Hf and alloys, 41-42

Polarized light, 9, 35, 36, 37, 38, 40, 41, 42, 45, 48, 50, low-melting point alloys, 38-38, 62-64
59, 67, 71, 85, 86 Mg and alloys, 36
Polishing: Ni and alloys, 49
abrasives, 27-28, 120 polymers, 65-66
automated, 25 precious metals, 51-52
cloths, 19, 24, 25-26, 28 printed circuit boards, 61
electrolytic, 24 Refractory metals, 40-44
head position, 19, 32 Sintered carbides, 55-56
lubricants, 27 Ti and alloys, 40-41
manual, 24 TSC and TBC coated metals, 53-54
recommendations, 120 Zr and alloys, 41-42
vibratory, 27, 35-36, 37, 39, 41
Polymers:
etchants, 77
preparation
131
(see also: castable resins)
Index

O-P (cont.) Silicon dioxide, 27

specific procedures, 32-66 Sintered carbides:
development, 32 etchants, 70-71, 77
load, 32 preparation, 55-56
platen size, 32 Smartcut, 12
time, 32 Solvents, 104, 106
“traditional” method, 29 Sorby, Sir Henry Clifton, 81
Printed circuit boards: Specimen preparation:
etchant, 61 goals, 6, 10
preparation, 61 Methods
Pull outs, 7, 121-122 Contemporary, 29-31
Specific materials, 32-66
R
traditional, 29
Refractory metals:
(see: preparation procedures)
etchants, 74
Problems, 7, 121-122
preparation, 40-44
(see: preparation problems)
Relief, 8, 23, 24
Stains, staining, 17, 18, 23, 121-122
Rigid grinding discs, 19, 30-31, 45-47, 49-50, 53-54,
Standards, metallography and hardness,
55, 57, 63
123-127
S Steels (see: Fe and alloys or Irons and steels)
Safety: T
acids, 107
Thermally spray-coated metals
bases, 110
preparation, 53-54
chemical storage, 103
Thermoplastic resins, 14-15, 18
dust, 102
Thermosetting resins, 14-15, 17-19
equipment operation, 102
Titanium and alloys:
fume hoods, 101
etchants, 74
MSDS, 101, 103
heat tinting, 72
personal protective equipment, 102-103
preparation, 40-41
publications, 114
Trademarks, Buehler Ltd., 128
safety plan, 101
solvents, 104, 106 U-Z
web sites, 114 Ultrasonic cleaning, 28
Sampling, 7 Vacuum impregnation, 15-16, 53, 57, 59
Sectioning, 9-13 Vickers hardness, 87-88
Sectioning damage, 7, 8, 9 Water-sensitive materials, 27, 36, 37, 120
Sectioning planes, 7 Worldwide Sales Offices, 132-135
Shrinkage gaps, 14, 15, 17, 18, 19, 122 Zirconium and alloys:
Silicon etchants, 74
preparation, 62-63 preparation, 41-42
Silicon carbide paper, 20-22
ANSI/CAMI grit size, 21-22
FEPA grit size, 21-22

132
Worldwide Sales Offices

BUEHLER HEADQUARTERS Buehler - ChengDu

Tel: (86) (28) 662.8833
41 Waukegan Road, P. O. Box 1
Lake Bluff, Illinois 60044-1699 Fax: (86) (28) 662.5844
Tel: (847) 295-6500 • Fax: (847) 295-7979 Buehler - Shanghai
Sales: 1-800-BUEHLER (800-283-4537) Tel: (86) (21) 6391.2588
Fax: (86) (21) 6391.2277
INTERNATIONAL Buehler - Wuhan
Argentina - Buenos Aires Tel: (86) (27) 8360.2980
ARO S.A. Fax: (86) (27) 8360.2779
Tel: (54) (11) 4331.5766
Buehler - Xian
Fax: (54) (11) 4331.3572
Tel: (86) (29) 766.0608
Australia - Victoria Fax: (86) (29) 766.0603
Biolab Industrial Technologies Colombia - SantaFe de Bogota
Tel: (61) (3) 9580.4366 Arotec Colombiana S.A.
Fax: (61) (3) 9580.6499 Tel: (57) (1) 288.7799
Austria - Wolfsgraben Fax: (57) (1) 285.3604
Andreas Grimas Costa Rica - San Jose
Tel: (43) 2233.78610 Capris S.A.
Fax: (43) 2233.78619 Tel: (506) 232.9111
Bahrain-Manama Fax: (506) 232.8525
Yousuf Mahmood Husain, WLL Croatia - Zagreb
Tel: (973) 1727.6176 Industrijska Defektoskopija (IDEF)
Fax: (973) 1722.3830 Tel: (385) (1) 364.6637
Bangladesh - Dhaka Fax: (385) (1) 364.6635
A.Q. Chowdhury Science & Synergy Czech Republic - Praha
Tel: (880) (2) 966.9288 Hanyko Praha S.R.O.
Fax: (880) (2) 885.4428 Tel: (420) (2) 5718.7612
Belgium - Namur Fax: (420) (2) 5718.7610
Analis S.A. Ecuador - Quito
Tel: (32) (81) 22.50.50 Sistec (Sistemas Tecnologicos)
Fax: (32) (81) 23.07.79 Tel: (595) (2) 574.9198
Brazil - Sao Paulo Fax: (595) (2) 253.1118
Instrumental Instrumentos de Medicao LTDA Egypt - Cairo
Tel: (55) (11) 5011.0901 Scientific Equipment Trading Institution (SETI)
Fax: (55) (11) 5012.4650 Tel: (20) (2) 574.9198
Bulgaria - Sofia Fax: (20) (2) 575.5662
ET KBR-Ivan Merchandjiiski Estonia - Tallinn
Tel: (359) (2) 979.1061 Tallinn Technical University
Fax:(359) (2) 979.0933 Tel: (372) (5) 238.8272
Canada - Ontario Fax: (372) (5) 238.8271
Buehler Canada Finland - Helsinki
Tel: (905) 201.4686 Knorring OY A.B.
Fax: (905) 201.4683 Tel: (358) (9) 56041
Chile - Vina del Mar Fax: (358) (9) 565.2463
G. Busch & Cia. Ltda. France - Dardilly
Tel: (56) (32) 692.064 Buehler S.A.R.L.
Fax: (56) (32) 692.064 Tel: (33) (4) 37 59 81 20
China Fax: (33) (4) 37 59 81 29
Hong Kong - China Germany - Düsseldorf
Buehler Asia/Pacific Buehler GmbH
Tel: (852) 2307.0909 Tel: (49) (211) 974100
Fax: (852) 2307.0233 Fax: (49) (211) 974.1079
Buehler - Beijing
Tel: (86) (10) 8563.1122
Fax: (86) (10) 8561.5320 133
Worldwide Sales Offices

Greece - Koropi Lithuania and Baltics- Vilnius

B. Caravitis S.A. AMTEST
Tel: (30) (210) 602.8317 Tel: (370) (5) 238.8272
Fax: (30) (210) 664.4886 Fax: (370) (5) 238.8271
Hungary - Budapest Malaysia - Selangor Darul Ehsan
Grimas HmbH CLMO Technology SDN. BHD.
Tel: (36) (1) 420.5883 Tel: (60) (3) 8942.1238
Fax: (36) (1) 276.0557 Fax: (60) (3) 8948.1661
India - Agra Malta - Marsa
Censico International Private, Ltd. Attard & Co., Ltd.
Tel: (91) (562) 2524777 Tel: (356) (21) 237555
Fax: (91) (562) 28520222 Fax: (356) (21) 220186
Iran - Tehran Mexico - Mexico D.F.
A B Sangari & Sons Microanalisis S.A. de C.V.
Tel: (98) (21) 8964958 Tel: (52) (55) 5660.9828
Fax: (98) (21) 8959396 Fax: (52) (55) 5660.9831
Ireland - Lucan Morocco - Casablanca
Labquip Ltd. Establissements Boyer
Tel: (353) (1) 6434586 Tel: (212) 227.1571
Fax: (353) (1) 6434611 Fax: (212) 226.1149
Israel - Givataim Netherlands - Zoeterwoude
Microtech Paes Nederland BV
Tel: (972) (3) 573.1122 Tel: (31) (71) 545.0850
Fax: (972) (3) 573.1123 Fax: (31) (71) 545.0802
Italy – Firenze New Zealand - Auckland
Nikon Instruments, S.p.a. Biolab Ltd.
Tel: (39) 055.300920 Tel: (64) (9) 980.6700
Fax: (39) 005.300993 Fax: (64) (9) 980.6788
Japan - Tokyo Norway - Oslo
Sankei Co., Ltd. Micronova AS
Tel: (81) (3) 5805.0514 Tel: (47) 69241230
Fax: (81) (3) 5805.0524 Fax: (47) 69241229
Osaka Pakistan – Karachi
Tel: (81) (6) 6327.3850 Faco Trading
Fax: (81) (6) 6327.1993 Tel: (92) (21) 452.3096
Shizuoka Pref. Fax: (92) (21) 452.6898
Tel: (81) (54) 287.6722 Peru - Lima
Fax: (81) (54) 287.6907 Cimatec S.A.
Tsukuba Tel: (51) (1) 336.5151
Tel: (81) (298) 52.3061 Fax: (51) (1) 336.5279
Fax: (81) (298) 52.3063
Philippines - Rizal
Yokohama Mr. Edzel Estaris
Tel: (81) (467) 41.1221 Tel: (63) (2) 6582668
Fax: (81) (467) 48.3610 Fax: (63) (2) 6582668
Jordan - Amman Poland - Katowice
SMS Scientific & Medical Supplies Co. Testing
Tel: (962) (6) 462.4907 Tel: (48) (32) 757.4597
Fax: (962) (6) 462.8258 Fax: (48) (32) 757.4815
Kuwait - Safat Portugal - Linda-a-Velha
Al-Sedan Trading & Contracting Co. LaboControle
Tel: (965) 242.0469 Tel: (351) (21) 419.7945
Fax: (965) 240.0111 Fax: (351) (21) 415.1430

134
Worldwide Sales Offices

Qatar - Doha Sultanate of Oman

AL-Obeidly Trading Technology Business International Group LLC
Tel: (974) 469.4111 Tel: (968) 771.4752
Fax: (974) 469.4449 Fax: (968) 771.3924
Romania - Bucarest Sweden - Sundbyberg
Total Control S.R.L. Mikron AB
Tel: (40) (21) 619.3054 Tel: (46) (8) 564.82140
Fax: (40) (21) 212.3151 Fax: (46) (8) 564.82141
Russian - Moscow Switzerland - Dietikon
ZAO S-Instruments Prufmaschinen AG Prufag
Tel: (7) (812) 786.9516 Tel: (41) 1746.4030
Fax: (7) (812) 252.0136 Fax: (46) 1746.4039
Russian - St. Petersburg Taiwan - Taipei
RVS Company Ltd. Sanpany Instruments Co. Ltd.
Tel: (7) (495) 203 03 08 Tel: (886) (2) 2392.3433
Fax: (7) (495) 203 10 67 Fax: (886) (2) 2392.3463
Russian - St. Petersburg Thailand - Bangkok
Nienschanz Jebsen & Jessen Marketing (T) Ltd.
Tel: (7) (812) 326 59 25 Tel: (66) (2) 714.3939
Fax: (7) (812) 623 58 74 Fax: (66) (2) 714.3900
Saudi Arabia - Jeddah Turkey - Istanbul
K.I. Abdulkadir & Partners Mega
Tel: (966) (2) 640.5846 Tel: (90) (216) 326.4535
Fax: (966) (2) 640.4710 Fax: (90) (216) 326.3286
Serbia & Montenegro - Belgrade Ukraine - Kyiv
Trimid d.o.o ICC-Ukraine
Tel: (381) (11) 444-4149 Tel: (38) (44) 235-6565
Fax: (381) (11) 3440416 Fax: (38) (44) 235-0195
Republic of Singapore - Singapore United Kingdom - Coventry
Microactive Scientific Pvt. Ltd. Buehler UK
Tel: (65) (6) 786.3386 Tel: (44) (0) 2476.582158
Fax: (65) (6) 786.3183 Fax: (44) (0) 2476.582159
Slovakia - Bratislava Uruguay - Montevideo
MITAR Gustavo E. Arias
Tel: (421) (7) 6428.8474 Tel: (598) (2) 600.3703
Fax: (421) (2) 6428.8484 Fax: (598) (2) 600.3776
Slovenia - Kamnik Venezuela - Caracas
Team Trade d.o.o Dequimem C.A.
Tel: (386) (1) 8310-700 Tel: (58) (212) 242.7512
Fax: (386) (1) 8313-496 Fax: (58) (212) 241.6354
South Africa - Johannesburg Yugoslavia & Mecedonia - Belgrade
Apollo Scientific cc Trimid PO
Tel: (27) (11) 466.7666 Tel: (381) (11) 444.4149
Fax: (27) (11) 466.7672 Fax: (381) (11) 456870
South Korea - Kyung-Ki Vietnam - Hochiminh City
Key One Engineering Co., Ltd. Ms. Thuy Do Thi
Tel: (82) (31) 706.1128 Tel: (848) 825-0245
Fax: (82) (31) 706.1150 Fax: (848) 825-0244
Spain - Oviedo
Biometa Tecnologia y Systemas, S.A.L.
Tel: (34) (902) 244343
Fax: (34) (985) 269169

135
Buehler is the Science Behind Materials Preparation & Analysis™
In today’s competitive world,

material scientists and metal-

lographers need to find effective

solutions to preparation issues

quickly. For over 67 years

Buehler has been providing tech-

nical solutions to the materials

industry on specimen prepara-

tion theory, best lab practices,

and material applications.

In response to this need

and to further our quest for

excellence in customer service, Buehler’s staff of professional material scientists have written BUEHLER®

SUM-MET™ The Science Behind Materials

Preparation - A Guide to Materials Preparation.

BUEHLER® SUM-MET™ is a summation of Buehler’s knowledge base in metallurgical and materials applications and

includes step-by-step information on sectioning, mounting, and grinding/polishing. BUEHLER® SUM-MET™ gives

extensive theory and information on preparing specific materials, etching, microscopy, microhardness testing,

imaging, lab safety, consumables usage and troubleshooting.

This book on CD is supplemented with the free Buehler e-Club on-line materials database at www.buehler.com.

There you can learn more information on your material by alloy, see the recommended preparation procedure and

view an interpretation of the microstructure. More information and materials are continuously being added. The

e-Club also provides free downloadable technical publications such as the Buehler Tech-Notes and Laboratory

Posters to help your lab with its daily work.

Buehler is the world’s premier manufacturer of scientific equipment and supplies for use in materials analysis.

Buehler’s products are used throughout the world in manufacturing facilities, quality laboratories, and universities

to enable material characterization, ensure quality, and perform materials research.

USA $120.00
(CAN $160.00)

ISBN 0-9752898-0-2 FN01255

136