Sample Preparation

Phase Diagram Sample Preparation

By Robert D. Shull

National Bureau of Standards

The procedures by which samples are prepared for phase diagram studies are examined and

critically evaluated. The three key elements that require attention (aHoy purity, homogeneity,

and equilibrium) are separately addressed, and several examples of bad procedure are

presented with information on their past and future consequences. The origin of commonly

confronted problems are described and special procedures are suggested for their circum-

vention. Additionally, new methods for the early detection of some sample problems are

presented, and the usefulness of rapidly solidified materials (as specimens) in phase diagram

studies is illustrated.

Of the many problems that befall phase diagram studies,

one of the most damaging and yet difficult to detect is that

of poor specimen characterization. In most instances that

this problem arises, the cause is found to be due to inatten-

tion to this important aspect of the program because of its

"triviality". Certainly, the most interesting and rewarding

aspect of any investigation is the interpretation of the

measurements performed on the samples. However, it

should not be construed that the preparation of the sam-

ples is by any means an inconsequential prerequisite to

this end. Obviously, the results of a study will depend on

the state of the alloys on which the measurements were

made. ]f carelessness is exercised on the development of

the specimens, especially such that their constitution and

thermomechanical history is not well-characterized, then

the conclusions drawn from their measurement will most

likely be in error.

A clear example of such a situation would be the gross

contamination of an alloy with tungsten by the inadver-

tent touching of the tungsten electrode in an arc furnace

with the alloy during the melting operation. The alloy, if

initially comprised of two elements, becomes a ternary

alloy and (according to the phase rule) allows an extra

degree of freedom for the subsequent establishment of

phase equilibria. The binary phase diagram determined

from such ternary alloys (if for some unknown reason they

were unrecognized as such and still used) is likely to be

quite different from the true diagram. Unfortunately, how-

ever, most instances of improper sample preparation are

much more subtle than this example. It is also unfortunate

that, usually in these more subtle instances, the problem

is found after a great deal of time and effort has already

been expended on using that sample.

This paper is concerned with sample preparation in gen-

eral, especially as it impacts alloy phase diagram studies

specifically. Alloy contamination is only one of three major

areas in this regard that must be addressed. The other two

problem areas are alloy homogeneity and alloy equili-

brium. Homogeneity ensures that all parts of the sample

are equivalent in composition. Consequently, the phase

equilibria attained in one part of the sample will be the

same as those attained in a different part of the sample.

Equilibrium should be ascertained so that the reported

phase diagram could properly be labeled an equilibrium

diagram. Specimens used in its determination should, con-

sequently, be appropriately stabilized. Additionally, if a

sample is truly at equilibrium in a given state, then cer-

tainly it does not matter how that state was achieved. If,

however, the sample is not at equilibrium--a condition

which is more generally the circumstance-- then the state

of the alloy (and, therefore, the measurement value deter-

mining it) will depend on the path by which that state was

reached. Such problems-- those that are directly traceable

to the preparation procedures of the specimens--are not

easily detected. This is doubly true for the reader of the

investigation results, where complete information is un-

available and vague terms such as "standard procedures"

frequently appear. This paper is designed to aid this latter

group of people (in formulating their evaluations of a

particular study) as well as to help the future investigator

(in designing his experimental program) by indicating

the places where problems in the above-listed three areas

are likely to occur and of their possible consequences. In

what follows, each problem area will be dealt with in the

natural order in which it is confronted during the prepara-

tion of the alloy for measurement: i.e., alloy purity, ho-

mogeneity, and equilibrium. Methods for their detection

and suggestions for best circumventing these problems are

then presented.

Alloy Purity

By necessity, alloy purity questions obviously must start

with purity questions of the constituent elements. A phase

diagram study should always begin with an analysis ofth e

beginning materials used in the investigation, if for no

other reason than to ascertain that the or.der of the system

(binary, ternary, quaternary, etc.) being studied matches

that of the alloys being used. At this stage, optical micros-

copy of the starting elements has been found very useful,

by being capable of the quick detection of insoluble con-

taminants. On the other hand, the qualitative detection of

soluble contaminants (including interstitial elements) is

possible by the relatively fast differential thermal anal-

ysis (DTA) method. This very powerful and widely used

phase diagram tool relies on the detection of the heat ab-

sorbed or released during a phase transformation in a

specimen as heat is supplied at a very constant rate to its

adiabatic enclosure. However, its usefulness in validating

elemental purities (especially in regard to the difficult-to-

detect interstitial contents) has not been generally recog-

nized. Figures 1 and 2 are shown to illustrate this novel

capability. In these and later figures herein containing

DTA data, the vertical axis depicts the differential tem-

perature between the sample and an A1203 reference cruci-

ble (the latter having a total heat content close to that of

the specimen and containing no phase transitions in the

temperature range of interest). The horizontal axis (implic-

itly, a time scale) indicates the programmed furnace tem-

perature. Large deviations from a horizontal line indicate

heat released (upward deviations) or absorbed (downward

deviations) during a phase transformation. For a congru-

ent transformation (i.e., the phenomenon of melting or

Bulletin of Alloy Phase Diagrams Vol. 4 No. 1 1983 5