CO2+CO and H2O+H2. Since the oxygen pressures are obtained through the equilibrium reactions:

(3.1)

and

(3.2)

the partial pressure of oxygen is given by:

(3.3)

where k1 and k2 are the equilibrium reaction constants. For constant ratios, the partial pressure of oxygen is independent of the total pressure. Thus these gas mixtures provide a means to obtain a range of oxygen pressures. Several techniques to mix these gases are discussed by Macchesney and Rosenberg [3.6].

In the study of corrosion in coal gasification atmospheres, gas mixtures such as CH4+H2 and H2S+H2 become important along with the ones listed above. As the gas mixture becomes more complex, the number of equations that must be solved to obtain the equilibrium gas composition at elevated temperatures and pressures also increases, making it convenient to use a program such as SOLGASMIX [3.7] for the calculations. One should not make the erroneous assumption that gas mixtures are the same at all temperatures since the equilibrium mixture is dependent upon the equilibrium constant, which is temperature-dependent.

3.5 CHARACTERIZATION METHODS

3.5.1 Microstructure and Phase Analysis

Visual Observation

The most obvious method of analysis is that of visual observation. The human eye is excellent at determining differences between a used and an unused ceramic. Such things

Copyright © 2004 by Marcel Dekker, Inc.

as variations in color, porosity, and texture should be noted. If no obvious changes have taken place, one should not assume that no alteration has occurred. Additional examination on a much finer scale is then required. Many times, visual observation can be misleading. For example, a sample may exhibit a banded variation in color, indicating a possible chemical variation. On closer examination, however, the color differences may be due only to porosity variation. An aid to visual observation is the dye penetration test. In this method, a sample is immersed into a solution such as methylene blue and then examined under a stereomicroscope.

Optical Microscopy

A compliment to visual observation is that of optical microscopy. Many people have devoted their entire lives to the study of ceramic microstructures through the examination of various sample sections and the use of some very sophisticated equipment. A preliminary examination should be conducted with a stereomicroscope and photographs taken. It is sometimes difficult to remember what a particular sample looked like after it has been cut into smaller pieces and/or ground to a fine powder for further analysis. A photographic record solves that problem.

The ceramics community has fallen into the habit of making only polished sections for observation by reflected light, when a tremendous amount of information can be obtained by observing thin sections with transmitted light. This trend has been brought about by the presence of many other pieces of equipment. Polished sections must be supplemented by x-ray diffractometry and also energy dispersive spectroscopy and/or scanning electron microscopy to obtain a full identification. A full identification can be made, however, with the use of thin sections. The only drawback is that an expert microscopist is required who understands the interaction of polarized and unpolarized light with the various features of the sample. It is true that the preparation of a thin section is more tedious than that of a polished section, but with today’s automatic equipment, there is not much difference. In addition, a thin

Copyright © 2004 by Marcel Dekker, Inc.

section does not require the fine polishing (generally down to submicron grit sizes) that a polished section does. The problem of pullouts does not interfere with the interpretation of the microstructure in transmitted light like it does in reflected light. The major drawback of a thin section, which should be on the order of 30 µm thick, is that it must be not greater than one crystal thick. With today’s advanced ceramics being produced from submicron-sized powders, many products do not lend themselves to thin section examination. In those cases, polished sections must suffice.

One major advantage of the light microscope over electron microscopes is the ability to observe dynamic processes. Timelapse video microscopy can be used to follow real-time corrosion processes. Obviously, room-temperature processes and those in aqueous media are the easiest to observe. Much of the latest work in the area of video microscopy has taken place in cell biology. Anyone interested in additional reading in this area should read the book by Cherry [3.8].

X-ray Diffractometry

Phase analysis is normally accomplished through the use of x- ray diffractometry (XRD), although optical microscopy can also be used. X-ray diffractometry is generally best done on powdered samples; however, solid flat surfaces can also be evaluated. Generally, a sample of about 1.5 g is necessary, but sample holder designs vary considerably and various sample sizes can be accommodated. Solid flat samples should be on the order of about 0.5 in. square. Powder camera techniques are available that can be used to identify very small quantities of powders. In multiphase materials, the minor components must be present in amounts greater than about 1–2 wt.% for identification. Once the mineralogy of the corroded ceramic is known, a comparison with the original uncorroded material can aid in the determination of the mechanism of corrosion.

Although quantitative XRD can be performed, the accuracy is dependent upon the sample preparation (crystal orientation plays a major role), the quality of the standards used, and the

Copyright © 2004 by Marcel Dekker, Inc.

care taken in reducing various systematic and random errors. Several articles have been published in the literature that the interested reader may want to consult before taking on the task of quantitative XRD [3.9–3.12]. The one by Brime [3.12] is especially good since it compares several techniques.

Although the author is unaware of the use of hightemperature XRD in the evaluation of corrosion, there is no technical reason why it could not be useful. The major problem with high-temperature XRD is the identification of multiple phases at temperatures where the peaks become sufficiently broadened to obscure one another*.

Scanning Electron Microscopy/Energy

Dispersive Spectroscopy

If an evaluation of the corroded surface is required and one does not want to destroy the sample totally, then an examination by scanning electron microscopy/energy dispersive spectroscopy (SEM/EDS) can yield valuable information. With most ceramics, however, the sample requires a conductive coating of carbon or gold before examination. If the same sample is to be used for both optical reflected light microscopy and SEM, the optical work should be done first. Quite often, the polished section prepared for optical examination is too large for the SEM, and the coating required for SEM may interfere with optical examination.

Chemical analysis by EDS can be quite useful in identifying phases observed in reflected light optical microscopy. Although the resolution of topographic features can be as good as several hundred angstroms in the SEM, the resolution of the EDS data is generally on the order of 1 µm. The EDS data also come from a small volume of sample and not just the surface. This may lead to the EDS signal originating from several overlapping

* High-temperature XRD also has many problems related to sample holders, sample temperature determination, and thermal expansion effects. Anyone considering this technique should have ample time to obtain results.

Copyright © 2004 by Marcel Dekker, Inc.