1

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

2
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.

3
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....................................... 8
Refractory Metals: Ti, Zr, Hf, Cr, Mo, Nb,
Method Development........................................................ 8
Re, Ta, V and W..................................................40
Sectioning............................................................................... 9 Titanium.............................................................40
Abrasive Wheel Cutting............................................ 9 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................................................................48
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..............................................23
Ceramics.......................................................................57
Electrolytic Polishing...............................................24
Composites.................................................................59
Manual “Hand” Polishing........................................24
Printed Circuit Boards.............................................61
Automatic Polishing................................................25
Eletronic Materials....................................................62
Polishing Cloths.........................................................25
Polymers......................................................................65
Polishing Abrasives..................................................27
Etching...................................................................................67
Examples of Preparation Procedure............................29
Etching Procedures..................................................67
The Traditional Method..........................................29
Selective Etching.......................................................68
Contemporary Methods.........................................29
Electrolytic Etching and Anodizing....................71
Procedures for Specific Materials.................................32
Heat Tinting................................................................71
Periodic Table of Elements....................................33
Interference Layer Method...................................72
Light Metals: Al, Mg and Be...................................34
Commonly Used Etchants for
Aluminum..........................................................34
Metals and Alloys......................................................73

4
 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................................................ 133
Binary Operations.....................................................98

Common Applications......................................... 100

Written by George F. Vander Voort, Director, Research & Technology, with contributions from the Buehler laboratory staff, past and present.

5
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-µm, or even a 3-µm diamond finish
ent of interest is revealed selectively and with
may yield a surface suitable for routine examina-
adequate contrast, then the actual tasks of
tion, although finer abrasives are usually needed
examination, interpretation, documentation and
for 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 240 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

6
 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 E45, E1122 and E1245 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
processes produce damage; some methods When developing a sampling
(such as flame cutting or dry abrasive cutting) plan for an unfamiliar part or
produce extreme amounts of damage. Tradi- component, determine the orientation of the
tional laboratory sectioning procedures, using piece relative to the original wrought or cast
abrasive cut-off saws, introduce a minor amount starting material and prepare sections on the
of damage that varies with the material being longitudinal, transverse and planar surfaces,
cut and its thermal and mechanical history. It is or radial and transverse surfaces, to reveal the
generally unwise to use the original face cut in desired information. Remember that the micro-
the shop as the starting point for metallographic structure may look much more homogeneous
preparation as the depth of damage at this than it is on a transverse plane compared to a
location may be quite extensive. This damage longitudinal or planar surface. If the starting
must be removed if the true structure is to be material was cold worked, or not fully recrys-
revealed. However, because abrasive grinding and tallized after hot working, grain structures
polishing steps also produce damage, where will appear to be equiaxed and smaller on a
the depth of damage decreases with decreasing transverse plane, but non-equiaxed and larger
abrasive size, the preparation sequence must on a longitudinal or planar surface.
be carefully planned and performed; otherwise,
preparation-induced artifacts will be interpreted
as structural elements.

7
Goals of Specimen Preparation

GOALS OF SPECIMEN PREPARATION If these characteristics are met, then the true
structure will be revealed and can be interpreted,
The preparation procedure and the prepared measured and recorded. The preparation method
specimen should have the following characteris- should be as simple as possible, should yield con-
tics to reveal the true microsctructure*: sistent high quality results in a minimum of time
and cost and must be reproducible.
• Deformation induced by sectioning, grinding
and polishing must be removed or be shallow Preparation of metallographic specimens [1-
enough to be removed by the etchant. 3] generally requires five major operations: (a)
sectioning, (b) mounting (optional), (c) grinding,
• Coarse grinding scratches must be removed;
(d) polishing and (e) etching (optional).
very fine polishing scratches may be tolerable
for routine production work.
Method Development
• Pullout, pitting, cracking of hard particles, The methods presented in this book use times
smear, and other preparation artifacts, must be that are conservative so that the vast majority of
avoided. compositions of alloys or materials of a specific
type can be prepared to yield the true microstruc-
• Relief (i.e., excessive surface height variations ture without artifacts or observable scratches.
between structural features of different hard- They were developed using an 8-inch (200 mm)
ness) must be minimized; otherwise portions of diameter platen system with six 1.25-inch (30 mm)
the image will be out of focus at high magnifica- diameter mounted specimens. For each category,
tions. except beryllium, a wide variety of specimens with
different processing histories were prepared using
• The surface must be flat, particularly at the stated methods.
edges (if they are of interest) or they cannot
be imaged. For any metal or material category, there are
compositions or processing conditions that
• Coated or plated surfaces must be kept flat if make those specimens easier, or more difficult,
they are to be examined, measured or photo- to prepare compared to the “average” specimen.
graphed. In general, even the most difficult specimens of
a given type can be prepared perfectly using
• Specimens must be cleaned adequately
these methods with no more than one etch and
between preparation steps, after preparation,
re-polish (repeat last step) cycle. For perfect color
and after etching.
etching results, it may be necessary to follow the
cycle with a brief vibratory polish, depending upon
• The etchant chosen must be either general or
the material.
selective in its action (reveal only the phase
or constituent of interest, or at least produce
When working with easier to prepare specimens,
strong contrast or color differences between
or for routine work where the same degree of
two or more phases present), depending upon
perfection is not required, the user can reduce the
the purpose of the investigation, and must
times, or eliminate one or two steps, or the final
produce crisp, clear phase or grain boundaries,
step, and get satisfactory results. Modifications
and strong contrast.
to the recommended procedures are left to the
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.

8
 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 sectioning of
polish. In the case of relatively brittle materials, an annealed CP titanium specimen (mod. Weck’s reagent, 100X,
sectioning may be accomplished by fracturing 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.
Figure 1. AbrasiMatic 300 Abrasive Cutter This may, or may not, be a problem. But, thicker

9
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 but for larger parts, it may not be possible to orient
than a thinner wheel, everything else being held the specimen properly for optimal cutting. When
constant. Consequently, for cases where the kerf a round wheel is used to cut a round bar in chop
loss or heat generation must be minimized, select mode, the contact area is initially very small. As the
the thinnest available wheel of the proper bond cut progresses, the cut widens until the maximum
strength and abrasive. Buehler’s AcuThin Cut-off diameter is reached. After this, the contact area
Wheels offer the thinnest possible wheels for decreases until sectioning is completed. Under
delicate abrasive sectioning. Buehler also has application of a constant load, the pressure on
diamond cut-off blades with resin bonded or the abrasive particles in the cut decreases as the
Rimlock1  metal bonded blade forms in sizes from contact area increases. If the pressure applied to
8-inch (203 mm) to 12-inch (305 mm) diameters. the grains is inadequate for cutting, then heat is
Resin-bonded diamond blades are ideal for cutting generated which may not be dissipated by the
very hard cemented carbide specimens; Rimlock coolant, which causes deformation damage, phase
or continuous rim blades are recommended for changes and possibly burning or melting.
cutting rocks and petrographic specimens.
A chop cutter may be set up to pulse as the
Historically, the most common cutter design has load is applied, that is, the wheel is fed into the
been the so-called “chop” cutter. Basically, the specimen, and then feeding is halted momen-
blade is attached to a motor and the operator tarily, Figure 3b. Pulsing causes the wheel to be
pulls a handle to drive the blade downward into stress shocked which removes both dull and
the work piece. Because of the design, the blade sharp abrasive from the wheel as new abrasive
moves through an arc as it is pulled downward, is exposed to the cut. While a better cut may
Figure 3a. For efficient cutting, the piece must be be obtained, cutting time and wheel wear are
oriented to minimize the contact area with the increased.
wheel. For small parts, this is generally easy to do,

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

Optimal cutting, with the least damage to the Consequently, cut times are longer. The wheel tra-
specimen, is obtained by keeping the pressure on verse length must be greater than the specimen
each abrasive particle in the wheel constant and width.This machine must be programmed for each
as low as possible by keeping the contact area size part to obtain cutting.
between wheel and specimen constant and low.
This idea is called the minimum area of contact Orbital cutting, Figure 3d, combines the best
cutting (MACC) method. features of chop, oscillation and traverse-and-
increment cutters, while minimizing cutting time.
Another option for MACC is the “traverse and incre- As the wheel is fed into the specimen, the arbor
ment” cutter design, Figure 3c. A strip of material is also rotated in a small ellipse planar with the
is removed by moving either the wheel or the wheel permitting the wheel to cut a small strip
work piece perpendicular to the cut direction. of material from the specimen. But no time is
After this, the cutter must move back to the side lost, as with the traverse and increment cutter,
where the cut began. The wheel is dropped down as the wheel is ready to cut on the original side.
a fixed increment and another strip is removed. In With the elliptical orbit and a controlled feed, the
this approach, the wheel must move to the end of conditions for minimum area of contact cutting
the stroke, return to the starting point, and then are obtained, irrespective of work piece size.
be moved an increment downwards in each cycle. Special programming is not needed. The orbital

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 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.

3d 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

cutting concept is available in the Delta cutter

from Buehler. These also feature Smartcut, which
will sense motor overloading during a cut, if it
occurs, and reduce the feed rate automatically.
When it is safe to return to the original feed rate,
Smartcut will do so.

Precision Saws
Precision saws, Figure 4, are commonly used in
metallographic preparation to section materi-
als that are small, delicate, friable, extremely
hard, or where the cut must be made as close as
possible to a feature of interest, or where the Figure 4. IsoMet 5000 Linear Precision Saw
cut width and material loss must be minimal. As
the name implies, this type of saw is designed to
make very precise cuts. They are smaller in size Helpful Hints For
than the usual laboratory abrasive cut-off saw Sectioning
and use much smaller blades, typically from 3-
When cutting a difficult speci-
to 8-inch (76 to 203 mm) in diameter.These blades
men with the recommended
can be of the non-consumable type, made of
consumable abrasive wheel, if the cutting ac-
copper-based alloys or copper plated steel with
tion has become very slow, pulse the applied
diamond or cubic boron nitride abrasive bonded
force. This will help break down the abrasive
to the periphery of the blade; or, they can be con-
bonding, exposing fresh, sharp abrasives to
sumable blades using alumina or silicon carbide
enhance the cutting action. However, if you are
abrasives with a rubber-based bond. Blades for the
using a resin- bonded diamond blade to cut
precision saws are much thinner than the abrasive
cemented carbides, or other very hard materi-
wheels used in an abrasive cutter and the load
als, do not pulse the applied force, as this will
applied during cutting is much less. Consequently,
shorten wheel life.
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
an abrasive cutter can be cut with a precision saw,
the cutting time will be appreciably greater but the
depth of damage will be much less. Precision saws
are widely used for sectioning sintered carbides,
ceramic materials, thermally sprayed coatings,
printed circuit boards, electronic components,
bone, teeth, etc. Table 2 lists selection criteria for
precision saw blades. Buehler’s blades for precision
saws are available with different abrasive sizes
and bonds to provide optimum cutting for a wide
variety of applications.

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, titanium
alloys, bone, teeth, and tungsten carbide
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, barium titanate,
rare earth oxides, and computer chips
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
Low Concentration Cubic Boron Nitride X X X X X X
(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
High Concentration Cubic Boron Nitride X X X X X X
(CBN) for use with Iron and Cobalt Base
Alloys, Nickel Base Super Alloys, and Lead
Based 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 42CrMo4 alloy steel
specimen mounted in EpoMet resin (1000X, 2% nital).
Never rapidly cool a thermosetting resin mount
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

Most castable mounting materials require
neither pressure nor external heat and are recom-
mended 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 8620 alloy
used castable resins due to their low cost and
steel was degraded by a shrinkage gap between the specimen and short curing time. However, shrinkage can be a
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 (and some epoxies) do generate consider- a compatible metal on the surface to be protected
able heat during curing and this can be strongly (electroless nickel, deposited using the EdgeMet
influenced by the molding technique used. Nelson Kit has been widely used) is generally considered
[4] measured the exotherm produced by polymer- to be the most effective procedure. However, im-
izing an acrylic resin using two procedures: a glass age contrast at an interface between a specimen
mold on a glass plate (insulative) and an alumi- and the electroless nickel may be inadequate for
num mold on an aluminum plate (conductive). certain evaluations. Figure 9 shows the surface of
Polymerization produced a maximum exotherm a specimen of 1215 free-machining steel that was
of 132°C using the insulative approach but only salt bath nitrided. One specimen was plated with
42°C using the conductive approach. Note that electroless nickel; both were mounted in EpoMet
132°C is not much less than the 150°C temperature G. It is hard to tell where the nitrided layer stops
used in hot compression mounting! Nelson also for the plated specimen, Figure 9a, which exhibits
measured 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 visibility due to a
lack of contrast between the protective nickel plating and the salt
during polymerization. The mold material also bath nitrided surface (arrow) of 1215 free-machining carbon steel;
can influence curing time and temperature. and, (b, bottom) good contrast and visibility between EpoMet resin
and nitrided surface (arrow) plus excellent edge retention (1000X,
For example, EpoxiCure cures fastest in 2% nital).
SamplKup plastic molds, slower in phenolic ring
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 EPDM mounting cups. temperature under pressure produce much tighter
All of these factors must be considered if the exo- mounts. Gaps that form between specimen and
therm 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 a
edge 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-woven, napless surfaces for polishing with
diamond abrasives (rather than softer cloths such
as canvas, billiard and felt) maintains flatness. Fi-
nal polishing with low nap cloths for short times
introduces very little rounding compared to use
b
of higher 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
b
less expensive phenolic compounds. The best
results are obtained with EpoMet G or EpoMet Figure 12. Excellent edge retention for (a, top) complex coated
(arrow) sintered carbide insert (1000X, Murakami’s reagent) and
F an epoxy based thermosetting resin that for (b, bottom) iron nitrided (arrow points to a white iron nitride
contains a filler material. For comparison, Figure layer; an embedded shot blasting particle is to the left of the
arrow) H13 hot work die steel specimen (1000X, 2% nital). EpoMet
11 shows micrographs of a nitrided 1215 speci- was used in both cases.
men mounted in a phenolic resin (Figure 11a),
and in methyl methacrylate (Figure 11b), at 1000X.
These specimens were prepared in the same speci- Very fine aluminum oxide spheres have been
men holder as those shown in Figure 9, but neither added to epoxy mounts to improve edge reten-
displays acceptable edge retention at 1000X. Fig- tion, but this is really not a satisfactory solution
ure 12 shows examples of perfect edge retention, as the particles are extremely hard (~2000 HV)
as also demonstrated in Figure 9. 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.
• Apex DGD and UltraPrep metal-bonded or
resin-bonded diamond grinding discs produce
excellent flat surfaces for a wide variety of ma-
Figure 13. Flat Edge Filler shot was added to EpoxiCure resin to terials.
improve the edge retention of this annealed H13 hot work die
steel specimen (500X, 4% picral). • Use “hard” napless surfaces for rough polishing
(until the final polishing step), such as TexMet,
UltraPol or UltraPad cloths, and fine polishing,
• Properly mounted specimens yield better edge such as a TriDent cloth. Use a napless, or a low- to
retention than unmounted specimens, as round- medium-nap cloth, depending upon the mate-
ing is difficult, if not impossible, to prevent at a rial being prepared, for the final step and keep
free edge. Hot compression mounts yield better the polishing time brief.
edge preservation than castable resins.
• Rigid grinding discs, such as the ApexHercules
• Electrolytic or electroless Ni plating (e.g., with H and S discs, produce excellent flatness and
the EdgeMet Kit) of the surface of interest edge retention and should be used whenever
provides excellent edge retention. If the possible.
compression mount is cooled too quickly after
polymerization, the plating may be pulled
away from the specimen leaving a gap. When Helpful Hints For
this happens, the plating is ineffective for edge Mounting
retention.
Epoxy is the only resin that will
• Thermoplastic compression mounting materials physically adhere to a specimen.
are less effective than thermosetting resins. The If its viscosity is low, epoxy can be drawn into
best thermosetting resin for edge retention is pores and cracks by vacuum impregnation.
EpoMet G, an epoxy-based resin containing a Acrylics are too viscous and boil to easily for
hard filler material. vacuum impregnation.

• Never hot eject a thermosetting resin after Castable resins are sensitive to shelf life,
polymerization and cool it quickly to ambient which can be extended by keeping them in a
(e.g., by cooling it in water) as a gap will form refrigerator. It is a good practice to date your
between specimen and mount due to the containers when you get them.
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 depth is high and embedding of alumina abrasive

in specimens can be a problem.
Grinding should commence with the finest grit
size that will establish an initially flat surface and Other materials have also been used both for
remove the effects of sectioning within a few the planar grinding stage or, afterwards, to
minutes. An abrasive grit size of 180 to 240 (P180 replace SiC paper. For very hard materials
to P280) is coarse enough to use on specimen such as ceramics and sintered carbides, one, or
surfaces sectioned by an abrasive cut-off wheel. more, metal-bonded or resin-bonded diamond
Hack-sawed, bandsawed, or other rough surfaces disks (the traditional type) with grit sizes from
usually require abrasive grit sizes in the range of about 240- to 9-µm can be used. The traditional
120- to 180-grit. The abrasive used for each suc- metal- or resin-bonded diamond disc has diamond
ceeding grinding operation should be one or two spread uniformly over its entire surface. Apex
grit sizes smaller than that used in the preced- DGD and Apex DGD Color are also available as
ing step. A satisfactory fine grinding sequence resin bonded diamond grinding discs covering a
might involve SiC papers with grit sizes of 220- or variety of material applications. An alternate type
240-, 320-, 400-, and 600-grit (P240 or P280, P400, of disc, the UltraPrep disc, has diamond particles
P600 and P1200). This sequence is used in the applied in small spots to the disk surface, so that
“traditional” approach. surface tension is lessened. UltraPrep metal-
bonded discs are available in six diamond sizes
As with abrasive-wheel sectioning, all grinding from 125- to 6-µm while UltraPrep resin bonded
steps should be performed wet provided that discs are available in three diamond sizes from 30-
water has no adverse effects on any constituents to 3-µm. Another approach uses a stainless steel
of the microstructure. Wet grinding minimizes woven mesh UltraPlan cloth on a platen charged
specimen heating, and prevents the abrasive from with coarse diamond, usually in slurry form, for
becoming loaded with metal removed from the planar grinding. Once planar surfaces have been
specimen being prepared. obtained, there are several single-step procedures
available for avoiding the finer SiC papers. These
Each grinding step, while producing damage
include the use of platens, thick woven polyester
itself, must remove the damage from the previ-
cloths, silk, or rigid grinding disks. With each of
ous step. The depth of damage decreases with
these, an intermediate diamond size, generally
the abrasive size but so does the metal removal
9- to 3-µm, is used.
rate. For a given abrasive size, the depth of
damage introduced is greater for soft materials
than for hard materials. Grinding Media
The grinding abrasives commonly used in the
For automated preparation using a multiple- preparation of metallographic specimens are
specimen holder, the intital step is called planar silicon carbide (SiC), aluminum oxide (Al 2O 3),
grinding. This step must remove the damage from emery (Al2O3 - Fe3O4), composite ceramics and
sectioning while establishing a common plane diamond. Emery paper is rarely used today in
for all of the specimens in the holder, so that each metallography due to its low cutting efficiency.
specimen is affected equally in subsequent steps. SiC is more readily available as waterproof
Silicon carbide and alumina abrasive papers are paper than aluminum oxide. Alumina papers, such
commonly used for the planar grinding step and as PlanarMet Al 120-grit paper, do have a better
are very effective. Besides these papers, there cutting rate than SiC for some metals [3]. These
are a number of other options available. One abrasives are bonded to paper, polymeric or cloth
option is to planar grind the specimens with a backing materials of various weights in the form
conventional alumina grinding stone on a ma- of sheets, discs and belts of various sizes. Limited
chine such as the PlanarMet (see Figure 17b). This use is made of standard grinding wheels with
requires a special purpose machine, as the stone abrasives embedded in a bonding material. The
must rotate at a high speed, ≥1500 rpm, to cut abrasives may be used also in powder form by
effectively. The stone must be dressed regularly charging the grinding surfaces with the abrasive
with a diamond tool to maintain flatness, damage in a premixed slurry or suspension. SiC particles,
particularly with the finer size papers, embed
20
 Grinding

Figure 14. SiC grit particle (arrow) embedded in a 6061-T6 Figure 15. 6µm diamond particles (arrows) embedded in lead
aluminum weldment (500X, aqueous 0.5% HF). (1000X).

readily when grinding soft metals, such as Pb, Sn, standards use the same methods for sizing the
Cd and Bi (see Figure 14). Embedding of diamond abrasives and the same standards to calibrate these
abrasive is also a problem with these soft metals devices (sieving for the coarsest grits, sedimen-
and with aluminum, but mainly with slurries when tation grading for intermediate grits (240-600
napless cloths are used, see Figure 15. or P280-P1200), and the electrical resistance
method for very fine grit sizes). The grit size
CarbiMet 2 SiC paper is made according to the ANSI/ numbering systems differ above 180- (P180) grit,
CAMI standard (B74. 18-1996) while SiC papers but equivalent sizes can be determined using
manufactured in Europe are made according to Table 5. As with many standards, they are not
the FEPA standard (43-GB-1984, R 1993). Both mandatory and manufacturers can, and do, make
some of their papers to different mean particle
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

sizes than defined in these specifications. There is a

philosophical difference in the two systems. ANSI/
CAMI papers use a wider particle size distribution
(centered on the mean size) than FEPA papers. A
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 Figure 17a. SuperMet grinder
surface finish after a particular grinding step, as it
sectioning damage. Manual grinding work is per-
is the residual damage in the specimen that may
formed on rotating “wheels”; that is, motor-driven
prevent us from seeing the true microstructure at
platens, Figure 17a and Figure 18, upon which the
the end of the preparation sequence.
grinding paper is attached.

Grinding Equipment Alternately, in the cases where large sections need

Stationary grinding papers, often used by students, to be ground to planarity or a high degree of stock
but uncommon in industrial use, are supplied in removal is necessary, then a planar grinder that
strips or rolls, such as for use with the HandiMet uses a grinding stone should be considered (figure
2 roll grinder. The specimen is rubbed against the 17b). Such devices can quickly bring even quite
paper from top to bottom. Grinding in one direc-
tion is usually better for maintaining flatness than
grinding in both directions. This procedure can be
done dry for certain delicate materials, but water is
usually added to keep the specimen surface cool
and to carry away the grinding debris.

Belt grinders, such as the SurfMet I (Figure 16), are

usually present in most laboratories. An alterna-
tive approach is to use a high speed disc grinder,
such as the SuperMet grinder (Figure 17a). These
types of devices use coarse abrasive papers from
60- to 240-grit, 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
Figure 17b. PlanarMet stone grinder

Figure 16. SurfMet I belt grinder

Figure 18. EcoMet 250 grinder/polisher

22
 Polishing

large sections of un-mounted material to a planar

condition with a good surface finish relatively
quickly and so are ideal for macro work

Lapping is an abrasive technique in which the

abrasive particles roll freely on the surface of a
carrier disc. During the lapping process, the disc
is charged with small amounts of a hard abrasive
such as diamond or silicon carbide. Lapping disks
can be made of many different materials; cast iron Figure 21. “Comet tails” at large nitrides in an annealed H13 hot
and plastic are used most commonly. Lapping work die steel specimen (200X, DIC).

produces a flatter specimen surface than grinding,

but it does not remove metal in the same man- (height differences between different constituents,
ner as grinding. Some platens, referred to as laps, or between holes and constituents (Figure 23
are charged with diamond slurries. Initially the and 24). Polishing usually is conducted in several
diamond particles roll over the lap surface (just stages. Traditionally, coarse polishing generally
as with other grinding surfaces), but they soon was conducted with 6- or 3-µm diamond abrasives
become embedded and cut the surface produc- charged onto napless or low-nap cloths. For hard
ing microchips. materials, such as through hardened steels, ceram-
ics and cemented carbides, an additional coarse
POLISHING polishing step may be required. The initial coarse
polishing step may be followed by polishing with
Polishing is the final step, or steps, in producing a
1-µm diamond on a napless, low nap, or medium
deformation-free surface that is flat, scratch free,
nap cloth. A compatible lubricant should be used
and mirror-like in appearance. Such a surface is
sparingly to prevent overheating or deformation
necessary to observe the true microstructure
of the surface. Intermediate polishing should be
for subsequent metallographic interpretation,
performed thoroughly so that final polishing
both qualitative and quantitative. The polishing
may be of minimal duration. Manual, or “hand”
technique used should not introduce extraneous
polishing, is usually conducted using a rotating
structures such as disturbed metal (Figure 19), pit-
wheel where the operator rotates the specimen
ting (Figure 20), dragging out of inclusions,“comet
in a circular path counter to the wheel rotation
tailing” (Figure 21), staining (Figure 22) or relief
direction.

Figure 22. Staining (arrow) on the surface of a of Ti-6% Al-2% Sn-

4% Zr-2% Mo prepared specimen (200X).

Mechanical Polishing
The term “mechanical polishing” is frequently
used to describe the various polishing procedures
involving the use of fine abrasives on cloth. The
cloth may be attached to a rotating wheel or a
vibratory polisher bowl. Historically, cloths have
been either stretched over the wheel and held
Figure 19. (top) Preparation damage (arrows) in annealed CP
titanium (500X, DIC, Kroll’s reagent). Figure 20. (bottom) Pitting in place with an adjustable clamp on the platen
(arrow) on the surface of a cold-drawn brass (Cu-20% Zn) specimen periphery, or held in place with a pressure sensi-
(100X)
23
Polishing

Electrolytic Polishing
Electrolytic polishing can be used to prepare
specimens with deformation free surfaces. The
technique offers reproducibility and speed. In
most cases, the published instructions for elec-
trolytes tell the user to grind the surface to a
600-grit finish and then electropolish for about
1- to 2-minutes. However, the depth of damage
a after a 600 (P1200) grit finish may be several
micrometres but most electropolishing solutions
remove only about 1-µm per minute. In this case,
the deformation will not be completely removed.
In general, electropolished surfaces tend to
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
b alloys, one phase will polish at a different rate than
Figure 23. Examples of freedom from relief (a, top) and minor another, leading to excessive relief. In some cases,
relief (b, bottom) at the edges of large primary hypereutectic
silicon particles in an as-cast Al-19.85% Si specimen (200X,
one phase may be attacked preferentially and
aqueous 0.5% HF). inclusions are usually attacked. Consequently, elec-
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.
a

Manual “Hand” Polishing

Aside from the use of improved polishing cloths
and abrasives, hand-polishing techniques still
follow the basic practice established many
years ago:
1. Specimen Movement. The specimen is held
with one or both hands, depending on the op-
erator’s preference, and is rotated in a direction
b
counter to the rotation of the polishing wheel. In
Figure 24. Differences in relief control in a brazed specimen
containing shrinkage cavities: a) (top) poor control; and, b) addition, the specimen is continually moved back
(bottom) good control (100X, glyceregia). and forth between the center and the edge of the
tive adhesive (PSA) bonded to the back of the wheel, thereby ensuring even distribution of the
cloth. If a stretched cloth moves under the applied abrasive and uniform wear of the polishing cloth.
pressure during polishing, cutting will be less ef- (Some metallographers use a small wrist rotation
fective. If an automated polishing head is used, while moving the specimen from the center to the
stretched cloths are more likely to rip, especially edge of one side of the wheel.) After each step, the
if unmounted specimens are being prepared. In specimen is rotated 45 to 90° so that the abrasion
mechanical polishing, the specimens are held by is not unidirectional.
hand, held mechanically in a fixture, or merely 2. Polishing Pressure. The correct amount of ap-
confined within the polishing area, as with the plied pressure must be determined by experience.
VibroMet 2 polisher. In general, a firm hand pressure is applied to the
24 specimen.
 Polishing

3. Washing and Drying. The specimen is washed more at a time and can be used for all grinding and
by swabbing with a liquid detergent solution, polishing steps.These devices enable the operator
rinsed in warm running water, then with ethanol, to prepare a large number of specimens per day
and dried in a stream of warm air. Alcohol usually with a higher degree of quality than hand polish-
can be used for washing when the abrasive carrier ing and at reduced consumable costs. Automatic
is not soluble in water or if the specimen cannot polishing devices produce the best surface flat-
tolerate water. Ultrasonic cleaning may be needed ness and edge retention.There are two approaches
if the specimens are porous or cracked. for handling specimens. Central force utilizes a
specimen holder with each specimen held in place
4. Cleanness. The precautions for cleanness, as
rigidly. The holder is pressed downward against
previously mentioned, must be strictly observed
the preparation surface with the force applied to
to avoid contamination problems. This involves
the entire holder. Central force yields the best edge
the specimen, the metallographer’s hands, and
retention and specimen flatness. If the results after
the equipment.
etching are inadequate, the specimens must be
placed back in the holder and the entire prepara-
tion sequence must be repeated. Instead of doing
this, most metallographers will repeat the final
step manually and then re-etch the specimen.

The second method utilizes a specimen holder

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-
Figure 25. EcoMet 300 grinder/polisher and AutoMet 300 power men can be replaced in the holder to repeat the
head
last step, as planarity is achieved individually rather
Automated Polishing than collectively. The drawback to this method is
that slight rocking of the specimen may occur, es-
Mechanical polishing can be automated to a high
pecially if the specimen height is too great, which
degree using a wide variety of devices ranging
degrades edge retention and flatness.
from relatively simple systems, Figure 25, to rather
sophisticated, minicomputer, or microprocessor
controlled devices, Figure 26. Units also vary in Polishing Cloths
capacity from a single specimen to a half dozen or The requirements of a good polishing cloth
include the ability to hold the abrasive media,
long life, absence of any foreign material that
may cause scratches, and absence of any pro-
cessing chemical (such as dye or sizing) that
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.
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 15 - 3 Diamond, Hard Non-woven Used for Harder Hard Ferrous
Al2O3 Pad used for Specimens and Materials
Durability Increased Flatness and Ceramics
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
* Interchangable in many of the following methods

26
 Polishing

Polishing Abrasives diamond was introduced, first of the monocrys-

Polishing usually involves the use of one or more talline form, similar in morphology to natural
of the following abrasives: diamond, aluminum diamond, and then in polycrystalline form. MetaDi
oxide (Al 2O 3), and amorphous silicon dioxide II diamond pastes and MetaDi suspensions use
(SiO2) in collidal suspension. For certain materi- synthetic monocrystalline diamond while MetaDi
als, cerium oxide, chromium oxide, magnesium Supreme suspensions and paste use synthetic
oxide or iron oxide may be used, although these polycrystalline diamonds. Figure 27 shows the
are used infrequently. With the exception of dia- shape differences between monocrystalline and
mond, these abrasives are normally suspended in polycrystalline diamonds. Studies have shown that
distilled-water, but if the metal to be polished is cutting rates are higher for many materials using
not compatible with water, other suspensions such polycrystalline diamond compared to monocrys-
as ethylene glycol, alcohol, kerosene or glycerol, talline diamond.
may be required. The diamond abrasive should
be extended only with the carrier recommended Colloidal silica was first used for polishing wafers
by the manufacturer. Besides the water-based of single crystal silicon where all of the damage
diamond suspension, Buehler also offers oil-based on the wafer surface must be eliminated before
suspensions that are needed to prepare materials a device can be grown on it. The silica is amor-
sensitive to water. phous and the solution has a basic pH of about
9.5. The silica particles are actually nearly spherical
Diamond abrasives were introduced to metallog- in shape, Figure 28, the polishing action is slow,
raphers commercially in the late 1940s. However,
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 Figure 28. Amorphous silica particles in colloidal silica (TEM,
natural diamond was used initially, and is still 300,000X).

available as MetaDi diamond paste. Later, synthetic

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
surface with running water. Then, cleaning is sim-
b pler. Amorphous silica will crystallize if allowed to
Figure 27. Comparison of monocrystalline (a, top) and evaporate. Crystalline silica will scratch specimens,
polycrystalline (b, bottom) synthetic diamond grain shapes (SEM,
450X). so this must be avoided. When opening a bottle,
27
Polishing

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

28
 Examples of Preparation Procedures

EXAMPLES OF PREPARATION varied, as well, to suit the needs of the work,

PROCEDURES or the material being prepared. This is the “art”
of metallography.
The “Traditional” Method
Over the past forty years, a general procedure has Contemporary Methods
been developed that is quite successful for prepar- New concepts and new preparation materials have
ing most metals and alloys. This method is based been introduced that enable metallographers to
on grinding with silicon carbide waterproof papers shorten the process while producing better, more
through a series of grits, then rough polishing with consistent results. Much of this effort has centered
one or more diamond abrasive sizes, followed by upon reducing or eliminating the use of silicon
fine polishing with one or more alumina suspen- carbide paper in the grinding steps. In all cases,
sions of different particle size. This procedure will an initial grinding step must be used, but there
be called the “traditional” method, and is described are a wide range of materials that can be chosen
in Table 7. instead of SiC paper. There is nothing wrong with
the use of SiC for the first step, except that it has a
This procedure is used for manual or automated short life. If an automated device is used that holds
preparation, although manual control of the a number of specimens rigidly (central force), then
force applied to a specimen would not be as the first step must remove the sectioning damage
consistent. Complementary motion means that on each specimen and bring all of the specimens
the specimen holder is rotated in the same in the holder to a common plane. The first step is
direction as the platen, and does not apply to often called “Planar grinding”. SiC paper is most
manual preparation. Some machines can be often used for this step although one more than
set so that the specimen holder rotates in the sheet may be needed according to the surface area
direction opposite to that of the platen, called to be planed down and how much material needs
“contra.” This provides a more aggressive action to be ground back. If it is such that more than a few
but was not part of the “traditional” approach sheets of SiC are needed then the use of a more
when automated. The traditional method is not aggressive / more durable surface such as ZirMet
rigid, as other polishing cloths may be substituted paper or diamond grinding discs (Apex DGD, Apex
and one or more of the polishing steps might DGD Color, UltraPrep diamond discs or UltraPlan
be omitted. Times and pressures could be stainless steel mesh cloth (charged with diamond)
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 2 120- (P120) grit 6 (27) 240-300 Until
abrasive discs SiC water Comp. Plane
(waterproof paper) cooled
CarbiMet 2 220- or 240- (P240 or P280) 6 (27) 240-300 1:00
abrasive discs grit SiC water cooled Comp.
CarbiMet 2 320- (P400) grit 6 (27) 240-300 1:00
abrasive discs SiC water cooled Comp.
CarbiMet 2 400- (P600) grit 6 (27) 240-300 1:00
abrasive discs SiC water cooled Comp.
CarbiMet 2 600- (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

may be used. In the case that a lot of material has materials, such as titanium and zirconium alloys,
to be removed then a stone grinding device such an attack polishing solution is added to the slurry
as the PlanarMet should be considered. With this to enhance deformation and scratch removal
huge array of products to choose from, how can and improve polarized light response. Contra
the metallographer decide what to use? Each of rotation (head moves in the direction opposite
these products has advantages and disadvantages, to the platen) is preferred as the slurry stays on
and this is only the first step. the cloth better, although this will not work if the
head rotates at a high rpm. Examples of generic
One or more steps using diamond abrasives on contemporary preparation practices follow in
napless surfaces usually follow planar grinding. Tables 8 to 10. Specific procedures for a wide range
PSA-backed silk, nylon or polyester cloths are of materials are given in the next section.
widely used. These give good cutting rates,
maintain flatness and minimize relief. Silk cloths, The starting SiC abrasive size is chosen based upon
such as the UltraPol cloth, provide the best flatness the degree of surface roughness and depth of
and excellent surface finishes relative to the cutting damage and the hardness of the material.
diamond size used. UltraPad cloth, a thicker, hard, Never start with a coarser abrasive than necessary
woven cloth, is more aggressive, gives nearly as to remove the cutting damage and achieve planar
good a surface finish, similar excellent flatness, conditions in a reasonable time.
and longer life than an UltraPol cloth. Synthetic
chemotextile pads, such as TexMet, give excellent A similar scheme can be developed using
flatness and are more aggressive than silk. They rigid grinding discs, such as the ApexHercules
are excellent for retaining second phase particles H Disc. These discs are generally restricted to
and inclusions. Diamond suspensions are most materials above a certain hardness, such as
popular with automated polishers as they can be 175 HV, although some softer materials can be
added easily during polishing; although it is still prepared using them. This disc can also be used
best to charge the cloth initially with diamond for the planar grinding step. An example of such a
paste of the same size to get polishing started practice, applicable to nearly all steels (results are
quickly. Final polishing could be performed with marginal for solution annealed austenitic stainless
a very fine diamond size, such as 0.1-µm diamond, steels) is given in Table 9.
depending upon the material, your needs and
The planar grinding step could also be
personal preferences. Otherwise, final polishing is
performed using a 45-µm metal bonded, or a
performed with MasterMet colloidal silica or with
30-µm resin bonded, UltraPrep disc or with the
MicroPolish or MasterPrep alumina slurries using
ApexHercules H rigid grinding disc and 15- or
napless, or low to medium nap cloths. For some
30-µm diamond, depending upon the material.

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 2 120- 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 Pad 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

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.
its wear rate will be greater than the H disc when
used to prepare very hard materials. A generic
four step practice is given below for soft metals
and alloys.

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 2 120- 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 Pad 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 2 220- 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. Different
is further modified, especially in the case of iron mount sizes per se should not be used as the basis
based alloys, as required due to the wide range of for calculations as the area of the specimens within
properties that can be experienced. the mount is more important than the mount
area itself. For example, the same specimen cross
The periodic table of the elements categorizes sectional area could be present in 1-, 1.25- or 1.5-
metallic, nonmetallic and metalloid elements inch (25, 32 or 38 mm) diameter polymer mounts.
according to similarities in atomic structure and Further, the number of specimens in the holder
is a good starting point, aided by similarities in can vary substantially. It is necessary to put at
physical properties and behavior, for grouping 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 (32 mm) or three 1.5-inch (38 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 (32 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 (38 mm) diameter specimens, the pressures
with an 8-inch (200 mm) diameter platen with six and times could be reduced to about 70% of the
1.25-inch (32 mm) diameter specimens in a 5-inch standard values.
(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

are presented below.While MgO was the preferred
LIGHT METALS: Al, Mg and Be
final polishing abrasive for aluminum and its alloys,
it is a difficult abrasive to use and is not available
Aluminum
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 2 220 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
Dendritic segregation in as cast 206 aluminum revealed by color
a perfect finish using standard methods. A
etching with Weck’s reagent (polarized light, 200X).
vibratory polisher can be used quite effectively
For many aluminum alloys, excellent results can with MasterMet colloidal silica to produce
be obtained using a four step procedure, such as deformation-free, scratch-free surfaces for
shown below in Table 12. This procedure retains all anodizing or color etching and examination
of the intermetallic precipitates observed in alumi- with polarized light or Nomarski DIC.
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 2 220- 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 2 220- to 320- (P240 5 (22) 200-250 Until
abrasive discs to P400) grit SiC* Comp. Plane
(waterproof paper) water cooled
TexMet 9-µm MetaDi oil-based 5 (22) 120-150 6:00
pad diamond slurry Contra
TexMet 3-µm MetaDi oil-based 5 (22) 120-150 5:00
pad diamond slurry Contra
TexMet 1-µm MetaDi oil-based 5 (22) 120-150 4:00
pad diamond slurry Contra
ChemoMet ~0.05-µm 6 (26) 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, cross- embedding. Another technique is to use both
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 2 220- 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
(top) Cadmium dendrites and a Cd-Bi eutectic in an as-cast Cd
to prepare; however, lead alloys are considerably – 20% Bi alloy (unetched, crossed polarized light plus sensitive
easier.Tin, which is allotropic with a body-centered tint, 50X). (bottom) As-cast microstructure of Zn – 0.1% Ti – 0.2%
Cu alloy exhibiting mechanical twins and a three-phase eutectic
tetragonal crystal structure at room temperature, (alpha-Zn, Cu-Zn and Zn15Ti) in a cellular pattern (Palmerton
is soft and malleable and less sensitive to twin- reagent, crossed polarized light plus sensitive tint, 200X).

ning. Pure tin, like pure lead, is very difficult to

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 eutectic mixture mized. Grinding of these metals is always difficult,
of antimony and Cu2Sb, in an as-cast Sb – 30% Cu specimen
(unetched, polarized light, 200X). as SiC particles tend to embed heavily.

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
For best results, follow this with a vibratory polish the paper with wax before grinding. Johnson
using MasterMet colloidal silica on MicroCloth (Wear, Vol. 16, 1970, p. 351-358) showed that
pad for times up to 1-2 hours. This will improve candle wax is far more effective than paraffin.
polarized light response for the hcp Cd and Zn However, if the layer is applied too thickly, the
and the rhombohedral Bi. The 1-µm alumina metal-removal rate in grinding will be reduced
step will remove any embedded silicon carbide drastically.
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 2 320- (P400) grit SiC 4 (18) 150-250 Until
abrasive discs water cooled Comp. Plane
(waterproof paper) (wax coated)*
CarbiMet 2 400- (P600) grit SiC 4 (18) 150-250 0:30
abrasive water cooled Comp.
discs (wax coated)*
CarbiMet 2 600- (P1200) grit SiC 4 (18) 150-250 0:30
abrasive water cooled Comp.
discs (wax coated)*
CarbiMet 2 MicroCut 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.
cloths deagglomerated
alumina suspension
MicroCloth or MasterPrep 4 (18) 80-150 4:00
VelTex 0.05-µm alumina Contra
cloths 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

5022 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 quench) Ti
authors have stated that titanium alloys should – 3% Cr alloy after tint etching with Beraha’s reagent (polarized
not be mounted in phenolic resins as the alloys 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 final polishing abrasive to obtain the best results,

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 2 320- (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
cloths 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

especially for commercially pure titanium, a rather effective). Polarized light response of CP titanium
difficult metal to prepare free of deformation can be improved by following this procedure with
for color etching, heat tinting and/or polarized a brief vibratory polish using colloidal silica.
light examination of the grain structure. Attack
polishing solutions added to the abrasive slurry
Zirconium and Hafnium
or suspension must be treated with great care to
avoid burns. Use good, safe laboratory practices Pure zirconium and pure hafnium are soft, ductile
and it is advisable to wear protective gloves. This hexagonal close-packed metals that can deform
three step practice could be modified to four steps by mechanical twinning if handled aggressively in
by adding a 3- or 1-µm diamond step, but this is sectioning and grinding. As with most refractory
usually unnecessary, see Table 16. metals, grinding and polishing removal rates are
low and eliminating all polishing scratches and
A number of attack polishing agents have been deformation can be difficult. It may even be pos-
used. The simplest is a mixture of 10 mL hydrogen sible to form mechanical twins in compression
peroxide (30% concentration – avoid skin contact) mounting. Both can contain very hard phases
and 50 mL colloidal silica. Some metallographers that make relief control more difficult. To improve
add either a small amount of Kroll’s reagent to polarized light response, it is common practice
this mixture, or a few mL of nitric and hydrofluoric to chemically polish specimens after mechanical
acids (avoid contact). These latter additions may polishing. Alternatively, attack polishing additions
cause the suspension to gel. In general, these acid can be made to the final polishing abrasive slurry,
additions do little to improve the action of the hy- or vibratory polishing may be employed. Table 17
drogen peroxide (the safer 3% concentration is not is a four-step procedure that can be followed by
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
Basket-weave alpha-beta structure of as-cast Ti – 6% Al – 4% V also been used. All of these attack polishing addi-
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 2 320- (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: Cr, reported to vary with crystallographic orientation.

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 revealing approaches are highly recommended, espe-
tungsten grains surrounded by a nickel-iron matrix (unetched, cially if attack polishing is performed. Following
Nomarski DIC, 200X).
is a generic four step practice suitable for these
metals and their alloys, see Table 18.
Pure chromium is soft and brittle; but, when en-
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 2 320- (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 pad or 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 (top) Wrought tungsten - 10 atomic % Ti containing a small
(middle) bright field and (bottom) Nomarski DIC which reveals the amount of alpha-Ti, beta-Ti-W eutectic and grains of beta-Ti,W of
cold work much better (Murakami’s reagent, 500X). varying composition and crystal 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
deformation-free grain structures is difficult. Sheet
steels present the same problem, which can be
complicated by protective coatings of zinc, alu-
minum 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 porosity (Klemm’s I
premium, double vacuum melt alloy. A wide range 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)
Apex DGD or 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)
Apex DGD or 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 2 120- 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 vibra-
the hardness of the alloy. Planar grinding can tory polish using colloidal silica on MicroCloth or
also be performed using 45-µm metal-bonded 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 white) stainless
grit SiC paper only if it is a very hard martensitic steel held at 816º C for 48 hours which forms sigma phase (orange
particles) revealed by electrolytic etching with aqueous 20%
stainless steel, such as type 440C. For the mar- NaOH at 3 V dc, 9 s (500X).
tensitic grades, planar grinding can be performed
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 2 120- 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 phosphorous-
deoxidized arsenical bronze that was annealed and lightly cold
usually are not necessary using the contempo- 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 2 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 cloth or 3-µm MetaDi Supreme 5-6 120-150 3:00
TexMet 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
cloths 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 2 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 or alumina suspension or Contra
ChemoMet 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 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 cloth or 3-µm MetaDi Supreme 6 (27) 100-150 3:00
TexMet 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 cloths 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 – 0.05% Be) after hot rolling and
Cu or Fe. Preparation of cobalt and its alloys is annealing revealing annealing twins (Beraha’s reagent, crossed
somewhat similar to that of refractory met- 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 2 220- 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 3-µm MetaDi Supreme 6 (27) 100-150 5:00
pad diamond suspension* Contra
TriDent cloth or 1-µm MetaDi Supreme 6 (27) 100-150 3:00
TexMet 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
cloths 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, Os, Pd, for metallographers. Damaged surface layers are
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 and the silver phase is uncolored
pure form, even its alloys are infrequent subjects (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 2 abrasive 220- to 240- (P240 to P280) moderate 400 Until
discs (waterproof paper) grit SiC water cooled Plane
CarbiMet 2 320- (P400) grit SiC moderate 400 1:00
abrasive discs water cooled
CarbiMet 2 400- (P600) grit SiC moderate 400 1:00
abrasive discs water cooled
CarbiMet 2 600- (P1200) grit SiC moderate 400 1:00
abrasive discs water cooled
TriDent 6-µm MetaDi Supreme moderate 400 2:00
cloth 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 for the diamond steps, as it will hold the abrasive
and is relatively easy to prepare. Rh is sensitive in its surface well, which minimizes embedding.
to surface damage in sectioning and grinding. Only diamond paste is used, as slurries will be more
Ruthenium is a hard, brittle metal that is not too prone to embedding, as we observed in high gold
difficult to prepare. alloys. Use only a small amount of distilled water
as the lubricant. Do not get the cloth excessively
Stewart (Tech-Notes, Vol. 2, Issue 5) has described wet. Final polishing is with a ChemoMet I cloth
a method for preparing jewelry alloys. Invariably, and MasterPrep alumina. Due to their excellent
these are small pieces, due to their cost, and corrosion resistance, colloidal silica is not effective
must be mounted. Stewart uses EpoMet G resin as an abrasive for precious metals. The ChemoMet
for most specimens. If transparency is need, he pad has many fine pores to hold the abrasive. The
uses TransOptic resin. Fragile beads and balls are 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 or with 50 mL of MasterPrep suspension. This will
MicroCloth pads to further enhance the quality thicken, so add about 20-30 mL water to make
of the preparation; but, a 1-µm diamond finish is it thinner. A 3 to 6 minute attack- polish step will
adequate for most work. Attack-polishing has been remove the fine polishing scratches. Wear protec-
used to prepare gold and its alloys and chemical tive gloves as the chromium trioxide solution is a
polishing has been performed after mechanical strong oxidizer.
polishing of silver, but neither practice is com-
monly 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 18-karat gold
embedding of abrasives. In this method, only one (Neyoro 28A: 75% Au – 22% Ag – 3% Ni) revealing annealing twins
silicon carbide step is used. TexMet pads are used (equal parts 10% NaCN and 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 2 220- to 320- (P240 to 3 (13) 150-240 Until
abrasive discs P400) grit water cooled Comp. Plane
TexMet 9-µm MetaDi II 3 (13) 150-240 5:00
pad diamond paste* Comp.
TexMet 3-µm MetaDi II 3 (13) 150-240 3:00
pad diamond paste* Comp.
TexMet 1-µm MetaDi II 3 (13) 150-240 2:00
pad diamond paste* Comp.
ChemoMet MasterPrep 6 (26) 100-150 2:00
cloth 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 with fluorescent illumination, the epoxy-filled

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

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. Alternate 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 ApexHercules 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 revealing a small Microstructure of a steel substrate covered by two thermally-
amount of porosity (black spots), linear detachments (elongated sprayed layers, a NiAl bond coat and yittria-zirconia top
black lines), and inclusions (gray particles) (unetched, 100X). coat (unetched, 100X). The bond coat 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)
Apex DGD or 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 pad or 3-µm 6 (27) 120-150 3:00
TriDent cloth MetaDi Comp.
Supreme
diamond
suspension*
MicroCloth, MasterPrep 8 (36) 120-150 2:30
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

54
 Sintered Carbides

SINTERED CARBIDES metal-bonded UltraPrep discs or the ApexHercules

H rigid grinding discs. A further option would be
Sintered carbides are very hard materials made the use of the traditional metal-bonded diamond
by the powder metallurgy process and may discs.
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
Listed below are two alternate procedures and 0.4% Nb (form complex carbides, the dark regions in the
for preparing sintered carbides using Apex DGD matrix) and coated with alumina (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)
Apex DGD or UltraPrep 45-µm diamond 6 (27) 240-300 5:00**
metal-bonded disc water cooled Contra
Apex DGD or UltraPrep 9-µm diamond 6 (27) 240-300 4:00
metal-bonded disc water cooled Contra
TriDent cloth, UltraPol cloth, 3-µm MetaDi 6 (27) 120-150 3:00
TexMet pad or Supreme diamond Contra
Nylon cloth 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 cloth, UltraPol cloth, 3-µm MetaDi 6 (27) 120-150 3:00
TexMet pad or Supreme diamond Comp.
Nylon cloth 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 etched with
aqueous 1% chromium trioxide at 2 V dc, 30 s to reveal the
Microstructure at the surface of a multilayered, CVD-coated WC - martensitic matrix (1000X).
8% Co cutting tool. The arrows point 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 matrix (dark spots at
bottom) for wear resistance (Murakami’s reagent, 1000X). Helpful Hints For
Sintered Carbides
The final step in these two procedures, Tables The ability to see the boundar-
32 and 33, employs either MasterMet colloidal ies between the WC particles
silica or MasterPrep alumina suspensions as they and cobalt binder in the as-polished condition
will produce the best definition of the structure, depends on the surface of the polishing cloth
particularly if the surfaces have a complex series used in the last step. To see no boundaries,
of coatings for improved wear resistance, as used which makes it easier to see graphite or eta
in some coated carbide inserts. However, if such a phase, use a nap-less surface, such as TriDent,
coating is not present, the matrix grain structure TexMet or UltraPol pads. To see the phase
can be observed quite clearly after the 3-µm dia- boundaries between WC particles and the
mond step. For routine examination there is little cobalt binder, use a medium nap surface, such
need to go beyond the third step. It is also possible as MicroCloth, MasterTex pads.
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 unknown)
But, it is possible to break out grains or produce (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)
Apex DGD or 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, 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)
Apex DGD or 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 cloth or 6-µm MetaDi 6 (27) 120-150 4:00
TexMet pad Supreme diamond Comp.
suspension*
TexMet or 3-µm MetaDi 8 (36) 120-150 4:00
Nylon or Supreme diamond Contra
UltraPol suspension and
cloths MasterMet
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 spheres added to
5 LC and 10 LC Series. Thinner blades produce 7075 aluminum (100X original magnification, magnification bar is
100 µm long). Etched with aqueous 0.5% HF. (bottom) Carbon-
less damage. Section as close as possible to the reinforced polymer composite (original at 100X; magnification
plane of interest. bar is 100 µm long). Specimen is as-polished and viewed with
Nomarski DIC illumination.
Composites can be damaged by compression
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 2 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 pad, 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.
cloths 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 cloth, UltraPad cloth, 15-µm MetaDi 6 (27) 120-150 4:00
or TexMet pad Supreme diamond Contra
suspension*
UltraPol cloth, TriDent cloth, 6-µm MetaDi 6 (27) 120-150 3:00
or TexMet pad Supreme diamond Contra
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 ‘populated’ 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 informa-
posed of layers of woven glass fiber cloth in a tion on these cases. A generalized method for non
polymeric matrix. Flex circuits, which are becoming populated PCB’s of both the rigid and flex varieties
quite common, do not typically contain glass fiber, 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 im-
mediately prior to mounting in acrylic. This will
“wick” the acrylic into the through-holes and
minimize the presence of problematic bubbles
Micrograph showing a plated-through hole in a printed circuit in these critical locations.
board (resin and fibers) with copper foil, electroless-plated copper
and Pb-Sn solder (equal parts ammonium hydroxide and 3%
hydrogen peroxide, 200X).

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

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

Surface Size Specimen (rpm)/Direction (min:sec)
CarbiMet 2 abrasive 180- (P180) grit SiC 5 (22) 120-150 Until
discs (waterproof paper) water cooled Contra Plane
CarbiMet 2 320- (P400) grit SiC 5 (22) 120-150 1:30
abrasive discs water cooled Contra
CarbiMet 2 600- (P1200) grit SiC 5 (22) 120-150 1:30
abrasive discs water cooled Contra
TexMet pad or 3-µm MetaDi II 5 (22) 150-240 1:30
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)
CarbiMet 2 abrasive 600 (P1200) grit Manual 100-150 Until
abrasive disc SiC water cooled Plane
(waterproof paper)
TexMet 3-µm diamond MetaDi II 7 (31) 100-150 2:00
pad diamond paste/ Contra
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)
CarbiMet 2 600 (P1200) 7 (31) 100-150 Until
abrasive disc grit SiC Comp. Plane
(waterproof paper) water cooled
TexMet pad 3-µm MetaDi II 7 (31) 100-150 Until
diamond paste/ Contra 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, 100X).
resins are preferred as the heat from a mounting
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 2 320- (P400) grit SiC 4 (18) 200-250 Until
abrasive discs water cooled Contra Plane
(waterproof paper)
CarbiMet 2 400- (P600) grit SiC 4 (18) 200-250 0:30
abrasive discs water cooled Contra
CarbiMet 2 600- (P1200) grit SiC 4 (18) 200-250 0:30
abrasive discs water cooled Contra
CarbiMet 2 MicroCut 800- (P1500) grit SiC 4 (18) 200-250 0:30
abrasive discs water cooled Contra
TexMet pad or 3-µm MetaDi II 5 (22) 100-120 4:00
TriDent diamond paste Comp.
cloth 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 worked and cold size. Grain boundary etchants are given in [1-3, 9].
drawn high-purity zirconium viewed with polarized light (200X). Problems associated with grain boundary etch-
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; and (c, bottom) Beraha’s
ing pearlite (a higher magnification is required to
reagent (100 mL water, 10 g Na2S2O3 and 3 g K2S2O5) at 100X. see the very finely spaced cementite that is more
lightly colored). Etching with boiling Murakami’s
reagent, Figure 32c, colors the iron phosphide
darkly and will lightly color the cementite after
68
 Etching

prolonged etching. The ferrite could be colored reveal the microstructure of a dual phase stainless
preferentially if Klemm’s I was used. steel in the hot rolled and annealed 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. Glyceregia, a

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 and d
measurement of delta ferrite, ferrite in dual phase
grades, and sigma phase. Figure 33 shows exam-
ples of the use of a number of popular etchants to

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

69
Etching

etchant for stainless steels, was not suitable for Selective etching techniques are not limited to
this grade, Figure 33b, as it appears to be rather iron-based alloys, although these have more thor-
orientation sensitive. Many electrolytic etchants oughly developed than for any other alloy system.
have been used for etching stainless steels, but Selective etching of beta phase in alpha-beta
only a few have selective characteristics. Of the copper alloys has been a popular subject. Figure
four shown in Figures 33c to f, only aqueous 60% 34 illustrates coloring of beta phase in Naval Brass
nitric acid produced any gray level discrimination (UNS 46400) using Klemm’s I reagent. Selective
between the phases, and that was weak. All re- etching has a long historical record for identifica-
vealed 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 ferrite
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 stainless steel by
ishing has brought out the outlines of the cobalt
electrolytic etching with (g) aqueous 20% NaOH (4 V dc, 10 s) and binder phase, but this image is not particularly
with (h) aqueous 10 N KOH, (3 V dc, 4 s); and by immersion color
etching (i) with Beraha’s reagent (100 mL water, 10 mL HCl and 1 g useful for image analysis. Etching in a solution of
K2S2O5) at 200X. hydrochloric acid saturated with ferric chloride,
Figure 36c, attacks the cobalt and provides good
uniform contrast for measurement of the cobalt
binder phase. Follow ing this with Murakami’s
70
 Etching

reagent at room temperature reveals the edges of Electrolytic Etching and Anodizing
the tungsten carbide grains, useful for evaluation The procedure for electrolytic etching is basi-
of the WC grain size, Figure 36d. cally the same as for electropolishing except that
voltage and current densities are considerably
lower. The specimen is made the anode, and
some relatively insoluble but conductive mate-
rial 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
a boundaries, or for coloring ferrite (as illustrated)
in Figure 33, delta ferrite, sigma or chi phases. An-
odizing is a term applied to electrolytic etchants
that develop grain coloration when viewed with
crossed-polarized light, as in the case of aluminum,
tantalum, titanium, tungsten, uranium, vanadium
and zirconium [2]. Figure 37 shows the grain struc-
ture of 5754 aluminum revealed by anodizing with
Barker’s reagent, viewed with crossed-polarized
light. Again, color image analysis makes this
b
image useful now for grain size measurements.

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
d unmounted polished specimen is placed face up
Figure 36. Microstructure of WC-Co cutting tool: a) as-polished in an air-fired furnace at a set temperature and
revealing graphite particles; b) relief polished revealing the
cobalt binder phase; c) after immersion in Chaporova’s etch (HCl held there as an oxide film grows on the surface.
saturated with FeCl3) to attack and “darken” the cobalt; and, d) Interference effects, as in tint etching, create col-
after (c) plus Murakami’s reagent to outline the WC grains (1000X).
oration for film thicknesses within a certain range,
about 30-500 nm. The observed color is a function
of the function of the film thickness. Naturally, the

71
Etching

thermal exposure cannot alter the microstructure. method. Suitable materials for the production
The correct temperature must be determined by of evaporation layers have been summarized in
the trial-and-error approach, but the procedure [22,23]. The technique is universally applicable,
is reproducible and reliable. Figure 38 shows the but does require a vacuum evaporator. Its main
grain structure of CP titanium revealed by heat weakness is difficulty in obtaining a uniformly
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 air, such as Al, Ni, Cr, stainless
Figure 38. Grain structure of annealed CP titanium revealed by steels, Nb (Cb),Ti and Zr. However, if the etchant
heat tinting on a hot plate (100X, polarized light plus sensitive forms a film, as in tint etchants, then immersion
tint).
must be used as swabbing will keep the film
from forming. Keller’s reagent reveals the grain
Interference Layer Method
size of certain aluminum alloys by forming a
The interference layer method [2], introduced film. This will not occur if the etch is used by
by Pepperhoff in 1960, is another procedure for 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 primary
aberration and coma is placed in front of the light 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 the to obtain standard magnifications, some sys-
primary image produced by the objective; the tems require other magnifications, such as 6.3X.
eye can then use the full resolution capability of Higher power eyepieces, such as 12, 15, 20, or
the objective. The microscope produces a virtual 25X, are also useful in certain situations. The
image of the specimen at the point of most dis- overall magnification is found by multiplying
tinct vision, generally 250 mm (10-inch) from the the objective magnification, Mo, by the eyepiece
eye. The eyepiece magnifies this image, permit- magnification, Me. If a zoom system or bellows
ting achievement of useful magnifications. The is also used, the magnification should be altered
standard eyepiece has a 24 mm diameter field of accordingly.
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 Using the upright microscope, the specimen
the eyepieces, but in the tube lens. Eyepieces, is placed on a slide on the stage. Because the
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 an analyzer (another Polaroid filter) before the
is used. For this reason, there is a practical limit eyepiece, as illustrated in Figure 44. The polarizer
to the amount of contrast that can be achieved. produces plane polarized light that strikes the
Illumination also becomes uneven as the degree surface and is reflected through the analyzer to
of “obliqueness” increases. Since differential in- the eyepieces. If an anisotropic metal is examined
terference contrast systems have been available, with the analyzer set 90° to the polarizer, the grain
oblique illumination is rarely offered as an option structure will be visible. However, viewing of an
on new microscopes. isotropic metal (cubic metals) under such condi-
tions will produce a dark,“extinguished” condition
Dark Field. Another method that can be used (complete darkness is not possible using Polaroid
to distinguish features not in the plane of the filters). Polarized light is particularly useful in
polished and etched surface of a metallographic 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 co-
herent interference in the image plane. This leads
to two slightly displaced (laterally) images differ-
ing in phase (λ/2) that produces height contrast.
The image produced reveals topographic detail
Figure 43. Schematic diagram showing the light path in dark field somewhat similar to that produced by oblique illu-
illumination.
mination but without the loss of resolution. Images
can be viewed with natural colors similar to those
Polarized Light. Because many metals and metallic observed in bright field, or artificial coloring can be
and nonmetallic phases are optically anisotropic, introduced by adding a sensitive tint plate.
polarized light is particularly useful in metal-
lography. Polarized light is obtained by placing a As an example of the use of these different imaging
polarizer (usually a Polaroid filter) in front of the modes, Figure 46 shows the micro- this happens
condenser lens of the microscope and placing because only a structure of an aluminum bronze

85
Light Optical Microscopy

However, as the martensite is noncubic in crystal

structure, it responds well to crossed- polarized
light, Figure 46b. Of course, not every material can
be examined using all four illumination modes, but
there are many cases where two or more modes
can be effectively utilized to reveal more informa-
tion than given by bright field alone.

Figure 45. Schematic diagram showing the light path for c

differential interference contrast (DIC) illumination.

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, Figure 46c, and differential inter-
d
ference contrast, Figure 46d, make use of this relief Figure 46. Martensitic microstructure of heat treated, eutectoid
to reveal much more detail than in bright field. 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
which objective you are using
a 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.

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 HARDNESS decision as it made the Vickers test very easy

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
covers the broadest hardness range, but it is not
In 1925, Smith and Sandland of the UK devel-
commonly used.
oped a new indentation test for metals that were
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 indentation
in hardness units expressed not in the standard, (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 and

Figure 50. Schematic of the Knoop indenter and the shape of an Bias
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 gf paration-induced damage could be removed thus
load produced a Knoop hardness of 375. (The kgf/ permitting bias-free low-load testing. However, the
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.
confidence level) if the difference between the If the spacing is too small, the new indent will be
mean values of the tests at the two loads is statisti-
cally significant or not.

90
 Microindentation Hardness Testing

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-
ing measurement precision. They state that this
Figure 51. OmniMet MHT
causes indents to be undersized by a constant
amount. However, as undersizing would in- An automated system can be programmed to
crease the measured hardness, not decrease it, make any number of indents in either a defined
resolution limits per se cannot explain this pattern (x number of equally spaced indents,
problem. For the Knoop indent, the chief or x number between two chosen points) or a
problem is image contrast at the indent tips random pattern (locations selected at random
that results in undersized indents. This problem, with a mouse). Curved patterns may be used, not
plus the variable indent shape, both result in just straight-line patterns. The system then makes
increasing HK with decreasing test load. For the the indents at the requested load and locations,
Vickers test, one would assume that undersizing or measures each indent, calculates the hardness
oversizing would be equally likely; but, experience (and desired conversions to other scales), and
suggests that oversizing is much more commonly prints/plots the results. Statistical analysis can also
encountered for low loads and small indents. be preformed.

In general, Vickers indents are easier to mea-

Automation sure by image analysis than Knoop indents due
Microindentation hardness testing is tedious, to the lower contrast at the tips of the Knoop
so anything that can be done to simplify test- indents (leads to undersizing and higher HK
ing is valuable, especially for laboratories that values). Metal flow (plastic deformation) around
do substantial testing. Many adjuncts to indent the indent edges can interfere with measurement
measurement have been tried and a variety of precision. Vickers indents, like the ones shown in
such systems are available. There has been con- Figure 48, exhibit excellent contrast and shape
siderable interest in applying image analyzers to and are easily measured. If the magnification
the indent measurement task. Further, with stage is too high (and this may be influenced by the
automation, it is possible to automate the indent- numerical aperture of the objective) for a given
ing process itself, and with the same equipment. indent size, image contrast suffers and correct
Figure 51 shows the Buehler OmniMet MHT auto- detection of the indent will be very difficult. On the
mated microindentation hardness testing system. other hand, if the indent is very small on the screen,
This system can be used in the fully automatic, it will be hard for the system to detect it automati-
semiautomatic or manual mode, depending upon cally. In this case, use a higher magnification, or use
the nature of the testing. OmniMet MHT features the semiautomatic measurement mode if this is
a tracing function which allows the operator to not possible. The operator fits a box around the
outline the exterior of a specimen,. rapidly set indent with a mouse to measure the indent.
multiple indent paths, document the placement
of hardness profiles, and automatically generate Microindentation hardness testing is a very
reports with a profile map, creating a visual record valuable tool for the materials engineer but it
of the entire measurement. must be used with care and a full understanding of
the potential problems that can occur. Always try

91
Microindentation Hardness Testing

to use the highest possible load depending upon

the indent spacing or closeness to edges. Try to Helpful Hints For
keep indents greater than 20-µm in diameter, as Microindentation
recommended in ASTM standard E384. If you have Hardness Testing
not read E384, or it has been some time since you
read it (all ASTM standards must be reviewed and If you suspect that the test
revised, if necessary, every five years), get the latest results obtained on a specimen are question-
copy and go over it. The Precision and Bias section able, verify the tester using a certified test
(and appendix X1 of E384) contains a wealth of block. When testing thin coatings, use the
practical advice. Automation of indentation and Knoop indenter with the long axis parallel
measurement is possible and, for laboratories that to the surface. If you have a specimen where
do substantial testing, greatly reduces operator the hardness changes rapidly, Vickers indents
fatigue while reducing testing times. 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 analysis ments. The OmniMet Capture Advanced system
expands upon the Capture Basic capabilities by
The progress in computer and video technology including additional interactive measurements
has created a movement toward electronic image (radius, angle, and counting), adding a results
acquisition. These images can be used in software window, and allowing the exporting of data to
applications such as word processing or desktop Microsoft® Excel. The OmniMet Express and Om-
publishing programs allowing for fast report niMet Enterprise products include automated
generation and electronic distribution. In addition, measurement capabilities in addition to the man-
associated data can be stored with the images in a ual measurements found in the OmniMet Capture
database allowing for easy retrieval of information products. OmniMet Express can incorporate a vari-
through searches. ety of application specific metallographic analysis
modules. OmniMet Enterprise is a full featured
When adding imaging capabilities to the labo-
system including a routine builder which enables
ratory, it is important to consider your goals.
the user to generate custom analysis solutions to
Oftentimes, images are just another step in the
meet their specific needs.
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
condition. Additional functionality of an imag-
ing system might include a scale marker overlay
and point-to-point or other operator interactive
measurements. Fully automated image analysis
includes detection of the features of interest
based on grey level or color differences as well
as morphological characteristics such as size and
shape. Automated imaging applications are based
Figure 52. OmniMet Digital Imaging System
on a fundamental series of steps shown in Table
45. Depending on your goals, some or all of these
steps may be utilized. Image Capture
Image capture is a term used to describe image
Aligned with these imaging goals, Buehler offers a acquisition by means of a camera and frame
series of upgradeable OmniMet imaging products grabber or a digital camera. Because of the many
(Figure 52). The OmniMet Capture Basic System choices of camera types, a video microscopy
provides digital or analog image capture, manual system must be flexible. Analog CCD cameras,
interactive measurements (length, parallel width, black and white or color, or digital CCD or CMOS
area, curve (string lenth), text annotation and scale are most frequently used. Component video (Y/C
bars), basic report generation in Microsoft® Word, or S-Video) and composite video signals and a
and databasing of saved images and measure- number of color video standards such as NTSC,

Table 45. Primary Imaging Steps

Imaging Steps Description

Capture 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

PAL and SECAM are generally supported. Images An image on a monitor consists of hundreds of
acquired in the materials laboratory are optimized thousands of pixels, arranged in such a manner
in real time by adjusting brightness, contrast, and that they appear to be connected. Figure 53
color saturation. The analog output camera signal displays a cast iron image with a small segment
is then digitized utilizing an analog input frame zoomed such that the pixels are observed.
grabber board. Table 46. Bit Depth

Bits
Alternatively, digital cameras may be integrated
1 21 or 2 tones, black and white
with an IEE 1394 firewire or USB connection. Since
8 28 or 256 grey scale shades
the late 1990’s digital video cameras have become
the preferred choice for scientific imaging applica- 24 224 or 16.7 million colors

tions. This is largely accounted for by two factors: 32 232 or 4.29 billion colors

1) higher pixel densities in the acquired images

potentially provide higher spatial resolution and Each pixel in a bitmap is stored in one or more
2) the cost range of digital cameras has merged data bits. If an image is monochromatic (black and
with the combined cost of an analog camera and white), then one bit is enough to store information
specialized capture board. Capturing an image for one pixel. If an image is colored or uses various
with a digital camera is often a two-stage process, shades of grey, then additional bits are required to
first a preview image is viewed for focusing and store color and shading information. Table 46 lists
selecting the field of interest and then a snapshot some of the more common bit depths referred to
of the image is taken. The preview image window in imaging software and camera literature.
displays the image at a lower pixel density and
the refresh rate of the camera will determine When working in grayscale, 256 values are used
the amount of lag time between the operator to represent each shade of grey, starting with
adjusting the microscope and the new image black at 0 and transitioning to white at 255. For
being displayed. In order to be comparable to an example, the information stored for the image in
analog camera, the refresh rate should approach Figure 53 would be a combination of the x and y
25 frames per second. positions and a number from 0 to 255. For color
images there are two common models used to
The resulting image files are based on two primary numerically represent color, RGB and HLS. Each
formats, bit-map and vector. The majority of scien- requires storing three values, in addition to the x
tific imaging programs are based on bit-mapped and y position of the pixel, in order to recreate the
images. Bit-mapped graphics are essentially grids, image. In general, you will notice the file sizes for
consisting of rows and columns of pixels. When color images are larger than grey scale images of
the image is zoomed to observe details, the indi- the same pixel density.
vidual pixels will become visible. Pixels, or picture
elements, are the smallest part of an image that a The RGB model is based on three primary colors
computer, printer or monitor display can control. – red, green and blue – that can be combined in
various proportions to produce different colors.
This scheme is considered additive because, when
combined in equal portions, the result is white.
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
Figure 53. A small segment of an image zoomed to display the body diagonal of the cube contains luminance, in
individual pixels..

94
 Image Capture and Analysis

other words, colorless information about how light true. The problem with large array images is that
or dark the signal is. they consume a lot of storage space. As a result,
these images are not recommended for posting on
The HLS model uses the concept of Hue, Lumi- web sites or email use. It is important to take into
nance, and Saturation. Hue is the color tone. For account what you intend to do with your images
example, an etchant might highlight the different when selecting the appropriate camera.
phases as unique shades of brown or blue. Satu-
ration is complementary to the hue. It describes In order to perform measurements, the image or
how brilliant or pure – as opposed to washed the image source must first be calibrated. Calibra-
out or grayish a particular color is. Luminance is tion is achieved by assigning a known distance to
correlated with the light intensity and viewed as a pixel count. When using a light microscope this
comparable to the greyscale values where zero is is accomplished with a stage micrometer. Each of
black and 255 is white. The HLS color space can be the objectives will have its own unique calibration
viewed as a double cone, in which the z-axis of the factor. If the aspect ratio of the camera pixels is un-
cone is the grayscale progression from black to known, it can be determined by calibrating in both
white, distance from the central axis is the satura- the x and y direction. If the aspect ratio is known,
tion, and the direction or angle is the hue. then calibration only needs to be completed in the

The illustration in Figure 54 represents pixels of

nine different color values. These values are shown
in Table 47 for the greyscale, RGB, and HLS models.
When the pixels are not true grayscale tones, the
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 5120 x 4096. 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 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

x direction. This calibration technique is repeated In addition to measurements, other annota-

for each objective or working magnification. It is tions are often placed on electronic images.
also possible to calibrate images that have been Annotations can be as simple as a label that is
imported from another source. The only require- referenced in the image caption or a complete text
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 coating thickness
(Figure 55). An operator will typically make several 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
is graphed. When working in grayscale with an
Figure 55. Operator interactive tools founds in imaging software. 8-color image, the same 256 values correspond

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

to luminance. Figure 59 demonstrates a yellow Binary Operations

bitplane being assigned to the coating (Figure 58) After the thresholding process has been accom-
based on the HLS values of the pixels as discussed plished, the different phases and features in an
previously. image are represented by different bitplane color
overlays. It is possible that more than one phase
or feature of interest is detected by the same bit-
plane color because of a similar gray level range.
A binary operation is the process that separates
and classifies features within the same bitplane,
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.
Figure 58. A TSC color image captured at 1200 x 1600 pixels.
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.
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-

Figure 60. The effect of the Binary commands, Close, Fill, and
Grid on the detected coating.

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.

To define the individual chords, where the grid

Figure 59. Thresholding the coating with a yellow bitplane overlaps the coating, a Boolean command is
overlay using the HLS color model.

98
 Image Capture and Analysis

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
segments of the red grid bitplane that overlap
the yellow coating bitplane. Figure 61 shows the

Figure 62. The histogram of results representing the chords

shown in Figure 61.

as a histogram offering the minimum, maximum,

mean and standard deviation of the measure-
ments across the coating (Figure 62).

Most measurements fall into one of two categories:

Figure 61. The chords that result from the overlapping coating
and grid lines.
field or feature-specific measurements. The grid
lines across the coating are typical of feature mea-
red chords which result from the AND command. surements. Each individual feature is measured and
OR combines two source bitplanes together to then plotted based on frequency in the histogram
form a single destination bitplane. For example, if and contributes to the overall statistics. In the case
two different inclusion types, oxides and sulfides, of grain or particle size measurements, features on
are detected initially, they can be combined to the edge of the image are eliminated because the
measure the total inclusion content. complete size of the feature is unknown.

One of the most powerful binary functions is the Field measurements are performed on the entire
ability to categorize individual features based field of view or a framed portion of the image,
on shape or size. It can be a single specification providing the sum of the individual measurements
(>5-µm) or multiple specifications (>2 and <5) in the selected area. Statistical information for field
separated by an AND or an OR. If a maximum allow- measurements is only generated if multiple fields
able pore size is specified, all pores that exceed that are analyzed. This can be employed to observe
limit can be flagged. The limit itself can be based microstructural variations within different fields of
on average diameter, maximum diameter, area, a specimen. A fairly common field measurement
aspect ratio, etc. Additionally, the shape of the pore is area percentage. An area percentage measure-
may have some significance. For example, in cast ment results from dividing the number of pixels
materials, a gas bubble tends to be circular while a in the bitplane of interest by the total number of
shrinkage cavity will have a more complex shape. pixels in the image. In the case of the coating, there
Common shape factors examine the relationship is a need to measure the area percent porosity rela-
between the perimeter and area of the feature. tive to the coating area and not the entire image
For this coating, the pores were separated from area. This is calculated by dividing the number of
the elongated oxides to insure that the porosity pixels in the bitplane of interest, i.e. the pores in the
level was not overestimated. coating, by the total number of pixels representing
the coating.
Returning to Figure 61, where grid lines are placed
across the coating, the length of all the individual
grid lines can be measured.The data can be shown

99
Image Capture and Analysis

Common Applications ity” is another ability of image analysis.These same

Throughout the past 20 years of Buehler’s involve- techniques are applicable to any constituent that
ment in image analysis, a number of applications can be detected.
stand out and may be of interest to most image
analysis users. Below is a listing of these applica- Phase Percentage. The area percent of various
tions: phases in a microstructure influences the proper-
ties. The tensile strength of grey iron, for example,
Grain Size. Image analysis provides a rapid and is directly related to the percentage of pearlite in
accurate means for determining grain size ac- its microstructure. In addition, in a single image,
cording to ASTM E112. Even if etching is unable multiple phases can be detected, measured and
to produce complete grain boundaries, or if there presented in one graph. For example, when evalu-
are twins that could skew the data, binary modifi- ating the inclusion content of steels, it would be
cations can be employed to make corrections. If a useful to examine the overall inclusion content as
specification cites maximum grain size limitations, well as isolating particular inclusion types, such as
the excessively large grains may be transferred to oxides and sulfides.
a different bitplane color to provide visual and
numerical feedback.
Helpful Hints For Image
Porosity. Porosity is detrimental to the physical Capture and Analysis
properties of most engineering materials. Image
Determine the required pixel
analysis is able to characterize the pores according
density of a camera system
to the total number of pores, number per unit area,
based on the size of the features of interest
maximum size, average size and the size distribu-
and acceptable file size
tion in the form of a histogram.
When distributing an image, always add a
Linear Measurements. Simple point-to-point scale bar
measurements are widely used for making occa-
sional measurements; however, in cases where a When saving images for analysis at a later
high quantity of measurements and more statistics time, verify that the image is stored as cap-
are required, automated image analysis is time tured, i.e., without any filters applied.
saving. A coating or layer is detected based on
the pixel values and then, after binary isolation
of the coating, grid lines are superimposed. Us-
ing Boolean logic to evaluate the common pixels
between the layer and grid lines, the result is many
chords representing the thickness at given points
in the coating.

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 “nodular-

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 result in dangerous
ASTM E 2014 (Standard Guide on Metallographic chemical spills and a clean-up nightmare.
Laboratory Safety) and relevant Material Safety
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 those
alcohol; then dissolve the specified salts. A mag-
reagents that are used regularly. Glass-stoppered
netic stirring device is of great value, as shown in
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 add the liquid ingredients water, acetone, ethyl ether, ethylene glycol, glycerol
to the solvent by dripping them down a glass stirring rod. If the
solution is more dangerous than this one, wear protective gloves
(glycerin), kerosene, petroleum ether, trichloroeth-
and use a face shield. If mixing generates substantial heat, it is a ylene, butyl cellosolve, and alcohols, such as amyl
good practice to place a cooling jacket around the beaker.
alcohol, ethanol, methanol, and isopropyl alcohol.
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. The acid ingredient a rancid odor that is used in electropolishing so-
was poured into an empty acid bottle and the solvents were added
without stirring or cooling. The solution may erupt in the metallo-
lutions. It is combustible and may form explosive
grapher’s face at any moment. It is recommended to keep the peroxides. It is toxic in contact with the skin, can
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 perchloric
Hydrofluoric acid (HF) is a clear, colorless, fuming acid solutions as explosions can result. The mixture
liquid or gas with a sharp, penetrating odor. It is of perchloric acid and acetic anhydride, which
very dangerous to the eyes, skin, and upper respi- was developed by Jacquet, is difficult to prepare
ratory tract. The substance can be absorbed into and highly explosive. Jacquet has reviewed the
the body by inhalation, through the skin and by accidents involving perchloric acid and has de-
ingestion. A harmful concentration of the gas in scribed safety procedures (30). The worst accident
air can be reached quickly, making it very danger- occurred on February 20, 1947, in an electroplating
ous to handle. Exposure by ingestion, inhalation factory in Los Angeles. In this accident 17 people
or contact, can be fatal. Undissociated HF poses were killed and 150 were injured (29). Medard,
a unique threat in that it can destroy soft tissues Jacquet, and Sartorius have prepared a ternary
and result in decalcification of the bone. Moreover, diagram showing safe com- positions of perchlo-
the effects may be delayed. Laboratories where ric acid, acetic anhydride, and water, Figure 66.
HF is used should stock an antidote kit to be used Anderson, however, states that accidents have still
in case of exposure. Although it is a relatively occurred with solutions in the “safe” region of this
weak mineral acid, HF will attack glass or silicon diagram (27). Electropolishing solutions composed
compounds, and should be measured, mixed, of perchloric acid and acetic anhydride are not
and stored in polyethylene vessels. HF reacts with recommended. Many companies forbid the use of
many compounds, including metals, and will liber- such mixtures, and some cities have banned their
ate explosive hydrogen gas. use. Electropolishing solutions of perchloric acid
and alcohol, with or without organic additions, and
Orthophosphoric acid (H3PO4), a colorless thick liq- mixtures of perchloric acid and glacial acetic acid
uid or hygroscopic crystal, is a medium strong acid. are safe to use. Nevertheless, in using these “safe”
It is corrosive to the skin, eyes and respiratory tract. mixtures, one should follow the formula instruc-
Phosphoric acid decomposes on contact with al- tions carefully, mix only small quantities, keep the
cohols, aldehydes, cyanides, sulfides, ketones, and temperature under control, and avoid evaporation.
can react with halogenated organic compounds 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 color-
Sax, however, states that mixtures of nitric and less, odorless crystals that are water-soluble. It is
acetic acids are dangerous (32). an irritant to the skin, eyes and respiratory tract;

109
Laboratory Safety

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 liq-
uid that should only be handled by well-qualified
Bases personnel. Its vapors are extremely irritating to the
Bases, such as ammonium hydroxide (NH4OH), eyes, skin, and mucous membranes. Skin contact
potassium hydroxide (KOH), and sodium produces deep, penetrating burns that are slow to
hydroxide (NaOH), are commonly used in metallo- heal. Contact with organic matter can cause fires.
graphy, chiefly in etchants.
Chromic acid (H2CrO4) is formed when chromium
Ammonium hydroxide is a colorless liquid with a trioxide (CrO3) is dissolved in water. CrO3 is used in
strong, obnoxious odor. Solutions are extremely electropolishing solutions (see previous comment
corrosive and irritating to the skin, eyes, and and Table 49 about explosive nature of mixtures
mucous membranes. It reacts exothermically with acetic anhydride). Dilute aqueous solutions
with sulfuric acid and other strong mineral acids, are widely used for attack polishing. It is a powerful
producing boiling solutions. oxidant; always wear gloves when using it for at-
tack polishing or use automatic devices and avoid
Sodium and potassium hydroxides are strong bas- contact potential. Chronic or long-term inhalation
es, available as white deliquescent pellets that are exposure may produce asthma-like reactions.
soluble in water. They can rapidly absorb carbon
dioxide and water from the air. Solutions stored in

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-
graphers. Since they are powerful oxidizers, they Purchase relatively small quanti-
pose a dangerous fire hazard, especially when in ties of chemicals that are used
contact with organic materials. They may evolve routinely, especially those that have a short
toxic fumes, such as oxides of nitrogen and lead, shelf life, or are dangerous.
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
In general, the metallographic laboratory is a fireproof cabinet.
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 14. G. F. Vander Voort, “Phase Identification

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

113
References

25. G. F. Vander Voort, “Operator Errors in the Norman V. Steere (ed.), Handbook of Laboratory
Measurement of Microindentation Hardness,” Safety, 2nd ed., CRC Press, Boca Raton, FL., 1971.
Accreditation Practices for Inspections, Tests, and A. Keith Furr (ed.), Handbook of Laboratory Safety,
Laboratories, ASTM STP 1057, ASTM, Philadelphia, 3rd ed., CRC Press, Boca Raton, FL., 1989.
1989, pp. 47-77.
L. J. Diberardinis et al., Guidelines for Laboratory
26. G. F. Vander Voort and Gabriel M. Lucas, “ Design: Health & Safety Considerations, J. Wiley &
Microindentation Hardness Testing,” Advanced Sons, N.Y., 1987.
Materials & Processes, Vol. 154, No. 3, September
1998, pp. 21-25. S. R. Rayburn, The Foundations of Laboratory Safety,
Brock-Springer Series in Contemporary Bioscience,
27. R. L. Anderson, “Safety in Metallographic Springer-Verlag, N.Y., 1989.
Laboratory,” Westinghouse Research Laboratory
Science Paper, No. 65 – 1P30 – METLL – P2, March C. A. Kelsey and A. F. Gardner (eds.), Radiation Safety
29, 1965. for Laboratory Technicians, Warren H. Green, Inc.,
St. Louis, MO., 1983.
28. G. F. Vander Voort, Metallography: Principles
and Practice, ASM International, Materials Park, OH., N. I. Sax, Dangerous Properties of Industrial
1999, pp. 148-159. Materials, 5th ed., Van Nostrand Reinhold Co., N.Y.,
1979.
29. R. C. Nester and G. F. Vander Voort, “Safety in the
Metallographic Laboratory,” Standardization News, Prudent Practices for Handling Hazardous
Vol. 20, May 1992, Chemicals in Laboratories, National Academy Press,
pp. 34-39. Washington, D.C., 1981.

30. P. A. Jacquet, “The Safe Use of Perchloric-Acetic Prudent Practices for Disposal of Chemicals from
Electropolishing Baths,” Met. Finish, Vol. 47, 1949, Laboratories, National Academy Press, Washington,
pp. 62-69. D.C., 1983.

31. S. M. Comas, R. Gonzalez Palacin and D. Vassallo, N. Proctor and J. Hughes, Chemical Hazards in the
“Hazards Associated with Perchloric Workplace, J. B. Lippincott Co., Philadelphia, 1978.
Acid-Butylcellosolve Polishing Solutions,” F. A. Patty (ed.), Industrial Hygiene and Toxicology,
Metallography, Vol. 7, 1974, pp. 47-57. Volume II – Toxicology, 3rd ed., Wiley-Interscience,
32. N. I. Sax, Dangerous Properties of Industrial N.Y., 1980.
Materials, 5th ed., Van Nostrand Reinhold Co., Inc., L. Bretherick, Handbook of Reactive Chemical
N.Y., 1979. Hazards, 2nd ed., Butterworths, London, 1979.
33. A.E. Dawkins, “Chromic Acid-Acetic Anhydride R. J. Lewis, Sr. (ed.), Hawley’s Condensed Chemical
‘Explosion,’” J. Iron and Steel Institute, Vol. 182, 1956, Dictionary, 12th ed., Van Nostrand Reinhold, New
p. 388. York, 1993.

Selected Bibliography of Books on Useful Web sites on laboratory safety:

Laboratory Safety http://ull.chemistry.uakron.edu
ASTM E-2014, “Standard Guide on Metallographic http://www.hhmi.org/research/labsafe/lcss/lcss.
Laboratory Safety”. html
N. Van Houten, Laboratory Safety Standards for http://www.cdc.gov/niosh/ipcsneng/nengsyn.html
Industry, Research, Academe, Sci. Tech. Pubs., Lake
Isabella, CA. 1990. http://www.state.nj.us/health/eoh/rtkweb/ rtkhsfs.
htm
E. Gershey, A. Wilkerson and E. Party,
Laboratory Safety in Practice, Van Nostrand
Reinhold, N.Y., 1991.
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 bond 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 resin EpoxiCure cast epoxy resin with Conductive SamplKwick cast acrylic resin with
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 Incomplete mixing; 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
melting 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 E562: Determining Volume Fraction by Systematic

STANDARDS Manual Point Count

A247: Visual Classification of Graphite in the E766: Calibrating the Magnification of a Scanning
Microstructure of Cast Iron Electron Microscope

A892: Defining and Rating the Microstructure of E768: Preparing and Evaluating Specimens for
High Carbon Bearing Steels Automatic Inclusion Assessment of Steel

B390: Evaluating Apparent Grain Size and E807: Metallographic Laboratory Evaluation
Distribution of Cemented Tungsten Carbides E883: Reflected-Light Photomicrography
B588: Measurement of the Thickness of Transparent E930: Estimating the Largest Grain Observed in a
or Opaque Coatings by Double-Beam Interference Metallographic Section (ALA Grain Size)
Microscope Technique
E975: X-Ray Determination of Retained Austenite in
B657: Metallographic Determination of Steel with Near Random Crystallographic
Microstructure in Cemented Tungsten Carbide Orientation
B681: Measurement of Thickness of Anodic E986: Scanning Electron Microscope Beam Size
Coatings on Aluminum and of Other Transparent Characterization
Coatings on Opaque Surfaces Using the
E1077: Estimating the Depth of Decarburization of
Light-Section Microscope
Steel Specimens
B748: Measurement of the Thickness of
E1122: Obtaining JK Inclusion Ratings Using
Metallic Coatings by Measurement of Cross
Automatic Image Analysis
Section with a Scanning Electron Microscope
E1180: Preparing Sulfur Prints for Macrostructural
B795: Determining the Percentage of Alloyed or
Examination
Unalloyed Iron Contamination Present in Powder
Forged Steel Parts E1181: Characterizing Duplex Grain Sizes
B796: Nonmetallic Inclusion Level of Powder E1245: Determining the Inclusion or Second-Phase
Forged Steel Parts Constituent Content of Metals by Automatic Image
Analysis
B847: Measurement of Metal and Oxide Coating
Thickness by Microscopical Examination of a Cross E1268: Assessing the Degree of Banding or
section Orientation of Microstructures
C664: Thickness of Diffusion Coating E1351: Production and Evaluation of Field
Metallographic Replicas
E3: Preparation of Metallographic Specimens
E1382: Determining Average Grain Size Using
E7: Standard Terminology Relating to
Semiautomatic and Automatic Image Analysis
Metallography
E1508: Quantitative Analysis by Energy-Dispersive
E45 Determining the Inclusion Content
Spectroscopy
of Steel
E1558: Electrolytic Polishing of Metallographic
E82: Determining the Orientation of a
Specimens
Metal Crystal
E1920: Metallographic Preparation of Thermal
E112: Determining Average Grain Size
Spray Coatings
E340: Macroetching Metals and Alloys
E1951: Calibrating Reticles and Light Microscope
E381: Macroetch Testing Steel Bars, Billets, Blooms, Magnifications
and Forgings
E2014: Metallographic Laboratory Safety
E384: Microindentation Hardness of Materials
E2015: Preparation of Plastics and Polymeric
E407: Microetching Metals and Alloys Specimens for Microstructural Examination
F1854: Stereological Evaluation of Porous Coatings
on Medical Implants
123
ASTM/ISO Standards

ASTM HARDNESS TESTING ISO3887: Steel, Non-Alloy and Low Alloy –

STANDARDS Determination of Depth of Decarburization
ISO4499: Hardmetals: Metallographic
B578: Microhardness of Electroplated Coatings
Determination of Microstructure
B721: Microhardness and Case Depth of Powder
ISO4524-1: Metallic Coatings – Test Methods for
Metallurgy Parts
Electrodeposited Gold and Gold Alloy Coatings –
C730: Knoop Indentation Hardness of Glass Part 1: Determination of Coating Thickness
C849: Knoop Indentation Hardness of Ceramic ISO4964: Steel – Hardness Conversions
Whitewares
ISO4967: Steel – Determination of Content of
C1326: Knoop Indentation Hardness of Advanced Non-Metallic Inclusions – Micrographic Method
Ceramics Using Standard Diagrams
C1327: Vickers Indentation Hardness of Advanced ISO4968: Steel – Macrographic Examination by
Ceramics Sulphur Print (Baumann Method)
E10: Brinell Hardness of Metallic Materials ISO4970: Steel – Determination of Total or Effective
E18: Rockwell Hardness and Rockwell Superficial Thickness of Thin Surface-Hardened Layers
Hardness of Metallic Materials ISO5949: Tool Steels and Bearing Steels –
E92: Vickers Hardness of Metallic Materials Micrographic Method for Assessing the
Distribution of Carbides Using Reference
E140: Standard Hardness Conversion Tables for Photomicrographs
Metals
ISO9042: Steels – Manual Point Counting Method
E448: Scleroscope Hardness Testing of Metallic for Statistically Estimating the Volume Fraction of a
Materials Constituent with a Point Grid
E1842: Macro-Rockwell Hardness Testing of Metallic ISO9220: Metallic Coatings – Measurement of
Materials Coating Thickness – Scanning Electron Microscope
Method
ISO METALLOGRAPHY STANDARDS
ISO14250: Steel – Metallographic Characterization
ISO643: Steels – Micrographic Determination of the of Duplex Grain Size and Distribution
Ferritic or Austenitic Grain Size
ISO945: Cast Iron: Designation of Microstructure of
Graphite
ISO1083: Spheroidal Graphite Cast Iron –
Classification
ISO1463: Metallic and Oxide Coatings –
Measurement of Coating Thickness – Microscopical
Method
ISO2624: Copper and Copper Alloys – Estimation of
Average Grain Size
ISO2639: Steel – Determination and Verification of
the Effective Depth of Carburized and Hardened
Cases
ISO3754: Steel – Determination of Effective Depth
of Hardening After Flame or Induction Hardening
ISO3763: Wrought Steels – Macroscopic Methods
for Assessing the Content of Non-Metallic
Inclusions

124
 ISO/Other Nattional Standards

ISO HARDNESS STANDARDS Other National Standards

ISO1355: Metallic Materials – Hardness Test – France
Calibration of Standardized Blocks to be Used for NFA04-10: Determination of the Ferritic and
Rockwell Superficial Hardness Testing Machines Austenitic Grain Size of Steels
(Scales 15N, 30N, 45N, 15T, 30T and 45T)
Germany
ISO4516: Metallic and Related Coatings – Vickers
and Knoop Microhardness Tests DIN50150: Testing of Steel and Cast Steel;
Conversion Table for Vickers Hardness, Brinell
ISO4545: Metallic Materials – Hardness Test – Hardness, Rockwell Hardness and Tensile Strength
Knoop Test
DIN50192: Determination of the Depth of
ISO4546: Metallic Materials – Hardness Test – Decarburization
Verification of Knoop Hardness Testing Machines
DIN50600: Testing of Metallic Materials;
ISO4547: Metallic Materials – Hardness Test – Metallographic Micrographs; Picture Scales
Calibration of Standardized Blocks to be Used for and Formats
Knoop Hardness Testing Machines
DIN50601: Metallographic Examination;
ISO6441-1: Paints and Varnishes – Determination of Determination of the Ferritic or Austenitic
Microindentation Harness – Part 1: Knoop Hardness Grain Size
by Measurement of Indentation Length
DIN50602: Metallographic Examination;
ISO6441-1: Paints and Varnishes – Determination of Microscopic Examination of Special Steels Using
Microindentation Harness – Part 2: Knoop Hardness Standard Diagrams to Assess the Content of
by Measurement of Indentation Depth Under Load Non-Metallic Inclusions
ISO6507-1: Metallic Materials – Hardness Test – SEP1510: Microscopic Test of Steels for Grain Size
Vickers Test – Part I: HV 5 to HV 100 by Comparison with Standard Charts
ISO6507-2: Metallic Materials – Hardness Test – SEP1570: Microscopical Examination of Special
Vickers Test – Part 2: HV 0, 2 to Less Than HV 5 Steels for Non-Metallic Inclusions Using Standard
ISO6508: Metallic Materials – Hardness Test – Micrograph Charts
Rockwell Test (Scales A, B, C, D, E, F, G, H, K) SEP1572: Microscopic Testing of Free-Cutting Steels
ISO9385: Glass and Glass-Ceramics – Knoop for Non-Metallic Sulphide Inclusions by Means of a
Hardness Test Series of Pictures

ISO10250: Metallic Materials – Hardness Testing –

Tables of Knoop Hardness Values for Use in Tests Italy
Made on Flat Surfaces UNI3137: Extraction and Preparation of Samples
ISO14271: Vickers Hardness Testing of Resistance UNI3138: Macrographic Analysis
Spot, Projection and Seam Welds (low load and
UNI3245: Microscopic Examination of Ferrous
microhardness)
Materials - Determination of Austenitic or Ferritic
Grain Size of Plain Carbon and Low-Alloy Steels
UNI4227: Determination of Metallographic
Structures
UNI4389: Nonferrous Metals and Their Alloys:
Determination of the Dimension of Their Crystal
Grains

125
Other National Standards

Japan PN-66/H-04505: Microstructure of Steel Products.

JISB7724: Brinell Hardness – Verification of Testing Templates and Evaluation.
Machine PN-75/H-04512: Nonferrous Metals. Reagents for
JISB7725: Vickers Hardness – Verification of Testing Revealing Microstructure.
Machines PN-75/H-04661: Gray, Spheroidal Graphite and
JISB7730: Rockwell Hardness Test – Calibration of Malleable Cast Iron. Metallographic Examination.
Standardized Blocks Evaluation of Microstructure

JISB7734: Knoop Hardness Test – Verification of PN-76/H-04660: Cast Steel and Iron. Microscopic
Testing Machines Examination. Sampling and Preparation of Test
Pieces.
JISB7735: Vickers Hardness Test – Calibration of the
Reference Blocks PN-84/H-04507/01: Metals. Metallographic Testing
of Grain Size. Microscopic Methods for
JISB7736: Brinell Hardness Test – Calibration of Determination of Grain Size.
Standardized Blocks
PN-84/H-04507/02: Metals. Metallographic Testing
JISG0551: Methods of Austenite Grain Size Test for of Grain Size. Methods of Revealing the
Steel Prior-Austenitic Grains in Non-Austenitic Steels.
JISG0552: Method of Ferrite Grain Size Test PN-84/H-04507/03: Metals. Metallographic Testing
for Steel of Grain Size. Macroscopic Method of Revealing the
JISG0553: Macrostructure Detecting Method for Prior-Austenitic Grain Size by the Fracture Method.
Steel, Edition 1 PN-84/H-04507/04: Metals. Metallographic Testing
JISH0501: Methods for Estimating Average Grain of Grain Size. A Method of Testing for Overheating
Size of Wrought Copper and Copper Alloys of Steel.
JISR1610: Testing Method for Vickers Hardness of PN-87/H-04514: Steel, Cast Steel, Cast Iron.
High Performance Ceramics Macrostructure Examination. Baumann’s Test.
JISR1623: Test Method for Vickers Hardness of Fine
Ceramics at Elevated Temperatures Russia
JISZ2243: Brinell Hardness Test – Test Method GOST801: Standard for Ball and Roller Bearing Steel

JISZ2244: Vickers Hardness Test – Test Method GOST1778: Metallographic Methods of

Determination of Nonmetallic Inclusions
JISZ2245: Rockwell Hardness Test – Test Method
GOST5639: Grain Size Determination
JISZ2251: Knoop Hardness Test – Test Method
JISZ2252: Test Methods for Vickers Hardness at
Sweden
Elevated Temperatures
SIS111101: Estimating the Average Grain Size of
Metals
Poland
SIS111102: Estimating the Austenitic Grain Size of
PN-57/H-04501: Macroscopic Examination of Steel. Ferritic and Martensitic Steels
The Deep Etching Test.
SIS111111: Methods for Assessing the Slag
PN-61/H-04502: Reagents for Macrostructure Tests Inclusion Content in Steel: Microscopic Methods
of Iron Alloys
SIS111114: Determination of Slag Inclusions –
PN-61/H-04503: Reagents for Microstructure Tests Microscopic Methods – Manual Representation
of Iron Alloys
PN-63/H-04504: Microstructure Testing of Steel
Products. Iron Carbide. Ghost Structure.
Widmanstätten’s Structure.

126
 Other National Standards

Sweden (continued) United Kingdom

SIS111116: Steel – Method for Estimation of the BS860: Tables for Comparison of Hardness Scales.
Content of Non-Metallic Inclusions – Microscopic
BS4490: Methods for Micrographic Determination
Methods – Jernkontoret’s Inclusion Chart II for the
of the Grain Size of Steel
Assessment of Non-Metallic Inclusions
BS5710: Macroscopic Assessment of the
SIS110340: Hardness Test – Vickers Test HV 0, 2 to
Non-Metallic Inclusion Content of Wrought Steels
HV 100 – Direct Verification of Testing Machines
BS6285: Macroscopic Assessment of Steel by
SIS1103 41: Hardness Test – Vickers Test HV 0,2 to
Sulphur Print
HV 100 – Indirect Verification of Testing Machines
Using Standardized Blocks BS6286: Measurement of Total or Effective
Thickness of Thin Surface-Hardened Layers in Steel
SIS1103 42: Hardness Test – Vickers Test HV 0,2 to
HV 100 – Direct Verification of Standardizing BS6479: Determination and Verification of Effective
Machine for Calibration of Standardized Blocks Depth of Carburized and Hardened Cases in Steel

SIS110343: Hardness Test – Vickers Test HV 0,2 to BS6481: Determination of Effective Depth of
HV 100 – Calibration of Standardized Blocks Hardening of Steel after Flame or Induction
Hardening
SIS112516: Metallic Materials – Hardness Test –
Vickers Test HV 5 to HV 100 BS6533: Macroscopic Examination of Steel by
Etching with Strong Mineral Acids
SIS112517: Metallic Materials – Hardness Test –
Vickers Test HV 0,2 to Less Than HV 5 BS6617: Determination of Decarburisation in Steel

SIS117020: Determination of the Depth of

Decarburization in Steel

127
Buehler Trademarks

registered trademarks of unregistered trademarks of

Buehler Ltd.: Buehler Ltd.:
Apex AcuThin
ApexHercules Delta
AutoMet EpoColor
Buehler PhenoCure
CarbiMet 2 Phoenix Beta
ChemoMet Phoenix Vector
DuoMet TransOptic
EcoMet TriDent
EdgeMet Kit UltraPad
EpoHeat UltraPlan
EpoKwick UltraPol
EpoMet UltraPrep
EpoThin WhiteFelt
EpoxiCure
FluorMet
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.) Iron and steel:

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

I filters, 80
graticules, 81
Image analysis
illumination modes, 79, 83-86
binary operations, 98-99
Köhler illumination, 83-84
feature-specific measurements, 99
light sources, 77, 79
field measurements, 99
magnification, 81, 82
measurements, 96, 99-100
numerical aperture, 80, 82, 83
thresholding, 96, 99-100
objective lens, 77, 79, 80
Inclusion measurement, 7
ray diagram, 78
Individual force, 19
resolution, 82, 83
reticles, 81
130
 Index

M (cont.) Polymers:
Microscope, reflected light (cont): etchants, 77
stage, 78, 81-89 preparation
stand, 82 (see also: castable resins)
working distance, 80, 81 Precision saws, 12-13
Mounting: Precious metals:
castable resins, 14, 15-17 etchants, 76-77
clamp mounting, 14 Preparation, 51-52
compression mounting, 14-15, 17-19, 117 Preparation method development:
conductive resins, 16, 117 6, 8, 29, 31, 32

edge retention (preservation), 14, 15, 17-19, Preparation problems:

117, 121-122 “comet tails,” 23, 121-122
electroless nickel plating, 17, 19 disturbed metal, 23
low-melting point metals, 38 edge retention (preservation), 14, 15, 17-19, 117,
phenolic resins, 14 121-122

presses, 17, 19 embedded abrasive, 20, 21, 35, 39, 64,

121-122
purposes, 14
excessive relief, 23, 24, 121-122
recommendations, 14, 16
pitting, 23
shrinkage gaps, 14, 15, 17, 18, 19
pull outs, 9, 121-122
thermoplastic resins, 14-15, 18, 117
scratches, 121-122
thermosetting resins, 14-15, 17-19, 117
smear, 63, 64, 121-122
troubleshooting guides, 117
troubleshooting guide, 121-122
Mount size, 21
water sensitivity, 36, 37
N Preparation procedures
Nickel and alloys: contemporary methods, 29-31
etchants, 76 Al and alloys, 34-35
preparation, 49 Be, 37
Nomarski DIC, 19, 43, 44, 59, 65, 85, 86 ceramics, 57-58
Co and alloys, 50
O-P
composites, 59-60
Oblique illumination, 84-85
Cu and alloys, 48
Other National Standards, 125-127
electronic devices, 62-64
Periodic table of the elements, 33
Fe and alloys, 45-47
Polarized light, 9, 35, 36, 37, 38, 40, 41, 42, 45, 48, 50,
59, 67, 71, 85, 86 Hf and alloys, 41-42

Polishing: low-melting point alloys, 38-38, 62-64

abrasives, 27-28, 120 Mg and alloys, 36

automated, 25 Ni and alloys, 49

cloths, 19, 24, 25-26, 28 polymers, 65-66

electrolytic, 24 precious metals, 51-52

head position, 19, 32 printed circuit boards, 61

lubricants, 27 Refractory metals, 40-44

manual, 24 Sintered carbides, 55-56

recommendations, 120 Ti and alloys, 40-41

vibratory, 27, 35-36, 37, 39, 41 TSC and TBC coated metals, 53-54
Zr and alloys, 41-42
131
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 Fax: (86) (28) 8619.9419

41 Waukegan Road, P. O. Box 1 Buehler - GuangZhou
Lake Bluff, Illinois 60044-1699 Tel: (86) (20) 3839.5933
Tel: (847) 295-6500 • Fax: (847) 295-7979 Fax: (86) (20) 8365.2222
Sales: 1-800-BUEHLER (800-283-4537) Buehler - Shanghai
www.buehler.com Tel: (86) (21) 6468.2199
Fax: (86) (21) 6333.0809
INTERNATIONAL Buehler - Shen Yang
Argentina - Buenos Aires Tel: (86) (24) 2325.5656
ARO S.A. Fax: (86) (24) 2396.1702
Tel: (54) (11) 4331.5766
Buehler - Nanjing
Fax: (54) (11) 4331.3572
Tel: (86) (25) 5186.0487
Australia - Victoria Fax: (86) (25) 8375.1010
Biolab Davidsons Buehler - Wuhan
Tel: (61) (3) 9580.4366 Tel: (86) (27) 8360.2980
Fax: (61) (3) 9580.6499 Fax: (86) (27) 8362.9193
Austria - Wolfsgraben Buehler - Xian
Andreas Grimas Tel: (86) (29) 8836.4931
Tel: (43) 2233.78610 Fax: (86) (29) 8837.7035
Fax: (43) 2233.78619 Colombia - Bogota
Bahrain- Saudi Arabia Arotec Colombiana S.A.
Abdulla Fouad Tel: (57) (1) 288.7799
Tel: (966) 2639.4455 Fax: (57) (1) 285.3604
Fax: (966) 2639.6725 Costa Rica - San Jose
Bangladesh - Dhaka Capris S.A.
Sensor International Tel: (506) 232.9111
Tel: (88) 028.629.621 Fax: (506) 232.8525
Fax: (88) 028.629.946 Croatia - Zagreb
Brazil - Sao Paulo Arting Zagreb d.o.o NDT
Instrumental Instrumentos de Medicao LTDA Tel: (385) (1) 346.0065
Tel: (55) (11) 5011.0901 Fax: (385) (1) 346.0067
Fax: (55) (11) 5012.4650 Czech Republic - Praha
Bulgaria - Sofia Hanyko Praha S.R.O.
EVIT Bulgaria Tel: (420) (251) 550.017
Tel: (359) (2) 983.3338 Fax: (420) (257) 187.611
Fax:(359) (2) 983.3338 Ecuador - Quito
Canada - Ontario Sistec (Sistemas Tecnologicos)
Buehler Canada Tel: (593) (2) 253.1118
Tel: (905) 430.4684 Fax: (593) (2) 253.1118
Fax: (905) 430.4647 Egypt - Cairo
Chile - Vina del Mar GSC2000 (Egypt)
G. Busch & Cia. Ltda. Tel: (20) 2781.0165
Tel: (56) (32) 2692064 Fax: (20) 2585.5326
Fax: (56) (32) 2692064 Estonia - Tallinn
China Deltalab,c/o Tallinn Technical University
Hong Kong - China Tel: (372) 539.94294
Buehler Asia/Pacific Fax: (372) 6203.153
Tel: (852) 2307.0909 Finland - Helsinki
Fax: (852) 2307.0233 Knorring OY A.B.
Buehler - Beijing Tel: (358) (9) 56041
Tel: (86) (10) 8844.9599 Fax: (358) (9) 565.2463
Fax: (86) (10) 8561.5320
Buehler - ChengDu
Tel: (86) (28) 8619.9419
133
Worldwide Sales Offices

France - Dardilly Kuwait - Safat

Buehler France Al-Sedan Trading & Contracting Co.
Tel: (33) 800 897.371 Tel: (965) 483.4510 ext. 242
Fax: (33) 800 880.527 Fax: (965) 484.0629
Germany - Düsseldorf Lithuania and Baltics- Vilnius
Buehler GmbH AMTEST
Tel: (49) (211) 974100 Tel: (370) (5) 238.8272
Fax: (49) (211) 974.1079 Fax: (370) (5) 238.8271
Greece - Koropi Malaysia - Seri Kambangan
B. Caravitis S.A. CLMO Technology SDN. BHD.
Tel: (30) (210) 602.8317 Tel: (60) (3) 8942.1238
Fax: (30) (210) 664.4886 Fax: (60) (3) 8948.1661
Hungary - Budapest Malta - Marsa
Grimas HmbH Attard & Co., Ltd.
Tel: (36) (1) 420.5883 Tel: (356) (21) 237555
Fax: (36) (1) 276.0557 Fax: (356) (21) 220186
India - Agra Mexico - Mexico D.F.
Censico International Private, Ltd. Microanalisis S.A. de C.V.
Tel: (91) (562) 2524777 Tel: (52) (55) 5660.9828
Fax: (91) (562) 28520222 Fax: (52) (55) 5660.9831
Ireland - Dublin New Zealand - Auckland
Labquip Ltd. Biolab Ltd.
Tel: (353) (1) 6434586 Tel: (64) (9) 980.6700
Fax: (353) (1) 6434611 Fax: (64) (9) 980.6788
Israel - Givataim Norway - Moss
Microtech Micronova AS
Tel: (972) (3) 573.1122 Tel: (47) 69241230
Fax: (972) (3) 573.1123 Fax: (47) 69241229
Italy – Firenze Pakistan – Karachi
Nikon Instruments, S.p.a. FACO Trading
Tel: (39) 055.300920 Tel: (92) (21) 452.3096
Fax: (39) 005.300993 Fax: (92) (21) 452.6898
Japan - Tokyo Peru - Lima
Sankei Co., Ltd. Cimatec S.A.
Tel: (81) (3) 5805.0514 Tel: (51) (1) 336.5151
Fax: (81) (3) 5805.0524 Fax: (51) (1) 336.5279
Osaka Philippines - Rizal
Tel: (81) (6) 6327.3850 Mr. Edzel Estaris
Fax: (81) (6) 6327.1993 Tel: (63) (917) 501.6347
Shizuoka Pref. Poland - Katowice
Tel: (81) (54) 287.6722 Testing
Fax: (81) (54) 287.6907 Tel: (48) (32) 757.4597
Tsukuba Fax: (48) (32) 757.4815
Tel: (81) (298) 52.3061
Portugal - Porto
Fax: (81) (298) 52.3063
Gravimeta, Lda.
Yokohama Tel: (351) (226) 184.232
Tel: (81) (467) 41.1221 Fax: (351) (226) 184.619
Fax: (81) (467) 48.3610
Qatar - Doha
Jordan - Amman AL-Obeidly Trading Technology
SMS Scientific & Medical Supplies Co. Tel: (974) 632.8509
Tel: (962) (6) 462.4907 Fax: (974) 436.0323
Fax: (962) (6) 462.8258

134
 Worldwide Sales Offices

Romania - Bucarest Taiwan - Taipei

Total Control S.R.L. Sanpany Instruments Co. Ltd.
Tel: (40) (21) 619.3054 Tel: (886) (2) 2392.3433
Fax: (40) (21) 212.3151 Fax: (886) (2) 2392.3463
Russian - Moscow Thailand - Bangkok
S-Instruments Jebsen & Jessen Marketing (T) Ltd.
Tel: (7) (095) 203.1125 Tel: (66) (2) 787.8118
Fax: (7) (095) 203.1067 Fax: (66) (2) 787.8109
Russian - St. Petersburg Turkey - Istanbul
RVS Company Ltd. Mega
Tel: (7) (812) 786.9516 Tel: (90) (216) 326.4535
Fax: (7) (812) 252.0136 Fax: (90) (216) 326.3286
Russian - St. Petersburg Ukraine - Kiev
Nienschanz Innovation Technology
Tel: (7) (812) 326.1090 ext. 2184 Tel: (38) (093) 154.8832
Fax: (7) (812) 325.5864 Fax: (38) (044) 468.9161
Serbia & Montenegro - Belgrade United Arab Emirates - Dubai
Trimid d.o.o GSC2000 (UAE)
Tel: (381) (11) 444.4149 Tel: (971) (4) 881.5270
Fax: (381) (11) 3440416 Fax: (971) (4) 881.6778
Slovakia - Bratislava United Kingdom - Coventry
MITAR s.r.o Buehler UK
Tel: (421) (2) 6428.8474 Tel: (44) 800 707.6273
Fax: (421) (2) 6428.8484 Fax: (44) 2476.582159
Slovenia - Kamnik Uruguay - Montevideo
Team Trade d.o.o Analitica SRL
Tel: (386) (1) 8310.700 Tel: (598) (2) 600.3468
Fax: (386) (1) 8313.496 Fax: (598) (2) 600.3776
South Africa - Johannesburg Venezuela - Caracas
Apollo Scientific cc Dequimem C.A.
Tel: (27) (11) 466.7666 Tel: (58) (212) 242.7512
Fax: (27) (11) 466.7672 Fax: (58) (212) 241.6354
South Korea - Kyung-Ki Vietnam - Hochiminh City
Key One Engineering Co., Ltd. Ms. Do Thi Thuy
Tel: (82) (31) 783.2199 Tel: (84) 8822.7160
Fax: (82) (31) 783.2499 Fax: (84) 9413.43696
Spain - Llanera
Biometa Tecnologia y Systemas, S.A.L.
Tel: (34) (902) 244343
Fax: (34) (985) 269169
Sultanate of Oman
Business International Group LLC
Tel: (968) 771.4752
Fax: (968) 771.3924
Sweden - Sundbyberg
Mikron AB
Tel: (46) (8) 564.82140
Fax: (46) (8) 564.82141
Switzerland - Schlieren
Prufmaschinen AG Prufag
Tel: (41) 44 746.4030
Fax: (46) 44 746.4039

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

material scientists and

metallographers need to find

effective solutions to preparation

issues quickly. For over 67 years

Buehler has been providing

technical solutions to the

materials industry on specimen

preparation 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