ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR DEVELOPMENT OF PROCEDURES FOR THE T-DENTFICATION OF MINOR PHASES IN HEAT-RESISTANT ALLOYS BY ELECTRON DIFFRACTION by L. 0. BROCKWAY Professof-of Chemistry and W. C., BIGELOW Research Assocdiate PROGRESS REPORT NO. 9 for the period 15 Januakiy to 15 April 1954 PROJECT NO. 2020 CONTRACT NO. AF-33(616)-23 EXPENDITURE ORDER NO. R465 Br-1 To AERONAUTICAL RESEARCH LABORATOR (WCRRL) RESEARCH DIVISION WRIGET AIR DEVELOPMT CENTER WRIGE-PATtERSON AIR FORCE BASE, OHIO

I Im

ACKNOWLEDG1EMNTS The heat-treated alloy specimens used in this work were furnislhed by Professor J, W. Freeman and Dr. C. L, Corey of the Department of Chemical and Metallurgical Engineering. The electron microscope studies were carried.out-on tpe electron microseope of the School of Public Health through the courtesy of Professor Thomas Francis and Dr. Rs E, Hartman., Much of the experimental work on the N-155 and. Inconel-X Alloys was done by Mrs J, A. Amy and Miss Rosemary Jacobson. The interest and cooperation of the above persons have contributed much to the work of this project and are sincerely appreciated. ii

SUMMARY Electron diffraction and x-ray diffraction studies of specimens of Inconel-X alloy aged for periods up to 1000 hours at 12000, 14000, and 1600~F have been made. Columbium carbonitride, titanium nitride and a complex M23C6-type carbide have been identified in the specimens aged at 1200~ and 1400QF while only the columbium carbonitride and the titanium nitride were found in the specimens aged at 1600~F. Prelimary studies have also indentified an intermetallic T' phase in the specimen aged 1000 hours at 14000F. Recent results have also provided additional evidence for the occurrence of an Fe2W phase in N-155 alloy.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN DEVELOPMENT OF PROCEDURES FOR THE IDENTIFICATION OF MINOR PHASES IN HEAT-RESISTANT ALLOYS BY ELECTRON DIFFRACTION INTRODUCTION This report covers the work carried out during the period of 15 January to 15 April 1954 on Project No. 2020 of the Engineering Research Institute under contract AF-33(616)-23 for the Aeronautical Research Laboratory of the Wright Air Development Center. This project has as its objective the development of procedures for adapting the electron diffraction method of Heidenreich, Sturkey, and Woods' to the inderntification of minor phases of heat-resistant alloys, and the application of this method to the study of the influence of controlled high-temperature aging. on the development of minor phases in several typical alloys. The basic procedures of the electron diffraction method of minor phase identification have been presented in the literature 1 2,3, and in previous reports on this project and therefore will be described only briefly here. Specimens are prepared for electron diffraction examination in three steps: (1) they are polished to produce flat surfaces suitable for microscopic and diffraction examinations, (2) they are etched using reagents and conditions which selectively attack the matrix phase and leave the minor phase particles protruding in relief from the surfaces, and (3) they are thoroughly rinsed with appropriate solvents to remove the products and reagents of the polishing and etching treatments and to produce clean, dry surfaces. Electron diffraction patterns are obtained from the protruding minor phase particles by directing an electron beam across the surfaces at a grazing angle and recording the diffracted electrons photographically. The patterns thus obtained are analocous to the familiar x-ray powder diffraction patterns and are employed similarly to identify or characterize the minor phases..The surfaces used in the electron diffraction studies are also examined by optical and electron microscopy to correlate the microstructures and the diffraction results. Previous work on this project has included (1) the investigation of various polishing, etching, and r insing procedures for preparation of alloy specimens for electron diffraction examination, (2) a study of electron microscope replica techniques for the evaluation of the surface conditions of the specimens used in the electron diffraction studies, (3) an

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN extensive study of the influence of high-temperature aging on the minor phases of 16-25-6 alloy, and (4) preliminary studies of the minor phases of heattreated S816, Inconel-X, and low-carbon N-155 alloys. This work has been described in previous progress reports. In addition, a paper describing the studies of etching agents has been submitted for publication in a technical journal and a second paper describing the work on the 16-25-6 alloy is in preparation. The work carried out during the period of this report has been primarily concerned with the investigation -of some special topics which have arisen in connection with the studies of the Inconel-X and N-155 alloys. RESULTS AND CONCLJSIONS Inconel-X Alloy Previous studies of this alloy were limited to specimens aged at 1400~ and 1600'F and were concerned primarily with development of etching procedures for preparing these specimens for the electron diffraction and electron microscopic examinations. Recently, additional studies on the etching of this alloy have been made, and the studies of the minor phases bave been extended to include aseries of specimens aged at 1200F. In the etching of this alloy, considerable difficulty has been encountered due to a pronounced tendency for the surfaces of the specimens to become contaminated with adherent, insoluble products formed dring-'the etching treatments.4,5 The formation of these products is extremely undesirable since they may prevent the electron beam from striking the minorphase particles or they may produce diffraction patterns which confuse the identification of the minor phases. This difficulty has been largely overcome through the use of a new etch consisting of 85 ml ethyl alcohol (95%), 5 ml of hydrofluoric acid (48%), and 10 ml of glycerol. This etch is used electrolytically at current densities of the order of 0.2 ampere per square inch. Because of its low conductivity, a potential of the order of 200 volts is required to produce this current. In some cases a visible black product is produced on the surfaces of the specime1ns; however, this can be removed by inmersing the specimens for several minutes in a solution consisting of 3 parts water and 1 part hydrochloric at id (355%). This etch is essentially a modification of the etch used by Taylor and Floyd* in studies of the mraicrostructures of various nickel-base alloys.6 * 85 ml water, 5 ml hydrofluoric acid, 10 ml glycerol.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN a. Taylor and Floyd'.s Reagent. b. Modified Reagent. Figure 1. Surfaces Obtained by Etching Inconel-X Specimen, AgedlOOO hours at 1400~F, With Hydrofluoric Acid Reagents. X10, 0000 Figure 2. Microstructures of Heat -Treated Inconel-X Specimens.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN It has a considerable advantage over the reagent of Taylor and Floyd for the present purposes, however, since the latter reagent frequently removes the minor-phase particles from the surfaces of the Inconel-X specimens. The etching results obtained with the two reagents on a specimen aged 1000 hours at 1400'F are compared in Figure 1, and the results obtained from two additional specimens with the new reagent are shown in Figure 2. In all of these cases the new etch produces surfaces in which the minor phase particles are clearly exposed and which are well suited for study by electron diffraction and electron microscopy. Use has also been made of aqua regia and of a solution consisting of 10 ml hydrochloric acid (35%) and 90 ml glycerol in preparing the Inconel-X specimens for the electron diffraction studies. The aqua. regia is used as an immersion etch whTle the HC1-glycerol reagent is used electrolytically. Both of these etches attack the matrix grains in preferential directions depending on crystallographic orientations of the grains in the surface. The resulting surfaces are thus somewhat rough and are not well suited for studing the microstructures by electron microscopy; however, the minor-phase particles are not dislodged and satisfactory electron diffraction results can be obtained from them. The results obtained from a specimen aged 1000 hours at 14000F using the HC1-glycerol reagent are shown in Figure 3. a. Electron Micrograph b. Electron Diffraction X10, 000. Pattern. Figure 35 Results Obtained by Etching Inconel-X Alloy With HCl-Glycerol Reagent. Specimen Aged 1000 hours at ll-00~F.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Using these different etching procedures, electron diffraction patterns of minor phases have been obtained from specimens aged for periods of 1, 10, 100 and 1000 hours at temperatmr es of 1200~, 1400o, and 1600oF. The minor phases identified from each specien are summarized in Table I. X-ray diffraction studies have also been made on carbides separated from a few selected specimens by the method of Mahla and Nielsen7, using a 10% solution of bromine in anhydrous methanol to digest the matrix phase. The diffraction patterns were-taken with Cu-Di radiation using a powder camera of 143.6 mm diameter to obtain accurate, high-resolution data. The minor phases identified in this way are also given in Table I'and are enclosed in square brackets to distinguish them from the electron diffraction results. As indicated in Table I, the only minor phase identified by electron diffraction in the specimerns aged at 16000F is the columnbium carbonitride. This is also true of the specimen aged for 1 hour at 12000F. In the other specimens aged at 12007F and in those aged at 1400'F, an M23C8 carbide, is found to be present in addition to the columbium carbonitride. Table I. Minor Phases Identified in Specimens of Inconel-X Alloy Aged at Elgh Temperatures* Tims of Temperature of Aging Aging (hrs ) 1200"F 1400 oF 1600-~F 1 Cb(C,N) Cb(C,N) + M23C. Cb(C,N) [Cb(C,N) + TiN] 10 Cb(C,N) + M2sC6 Cb(C,N) + M23C6 Cb(C,N) 100 Cb(C,N) + M23C6 Cb(C,N) +M2306 Cb(C) + N) [Cbh(C,N) + TiN] l000 Cb(C,N) + M2306 Cb(C,N) + M23C6 Cb(C,N) [Cb(C,N) + M23C6 + TiN + r ] Results enclosed in square brackets, [1, were obtained by x-ray diffraction, the others by electron diffraction.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN The x-ray diffraction results from the extracted carbides confirm these identifications but also provide evidence for the occurrence of small amounts of titanium nitride in each of the specimens studied. Reflections of the titanium-nitride phase were obtained in only a few of the electron diffraction patterns and then werem so weak as to be difficult to measure reliably. The lattice parameters of the columbium carbonitride, titanium nitride and the M23C6 carbide as dstermined from the x-ray data are respectively 4.38 A, 4.22 A, and 10.63 A. The values calculated from the electron diffraction patterns agree with these values within the linits of accuracy of the electron diffraction method (+ 1 percent). From these results it is evident that the columbium carbonitride occurs as a mincor phase in this alloy over the entire range of heat treatments considered here. The same appears to be true of the titanium nitride, although the data on this phase are somewhat limited at present. On the other hand, the M23C6 carbide is strongly influenced by the agingtreatment, for although it occurs in virtually all of the specimens aged at 12000 and 1400*F, it has not been detected in any of the specimens aged at 16000F. The fact that this phase has not been found in the specimen aged 1 hour at 12000F is not considered particularly significant since only preliminaxry studies have been made on this specimen to date. In addition to these studies of the carbide phases, consideration has been given to the occurrence in this alloy of an intermetallic 7' phase based on Ni3Al. This phase has been observed by several investigators in other nickel-base8alloys containing titanium and aluminuxm, and has been shown by Nordheim to be of importance in the precipitation hardening of such alloys. This phase usually appears as globular particles, although Taylor and Floyd6 have reported instances in which it developed as cubeshaped particles aligned along definite directions in the matrix grains. It has a superlattice structure based on a cubic unit cell whose dimensions differ but slightly, if at all, from those of the matrix phase. Due to the limited accuracy of the electron-diffraction method, reflections from this phase would be indistinguishable from matrix reflections in the electron diffraction patterns; therefore, high-resolution x-ray methods- are required to detect this phase. The procedure generally used is to scan the region of the (1ll)-matrix reflectin from a bulk specimen with a geiger-counter diffractometer using Cr-Ka radiation to obtain maximum sensitivity and resolution. The presence of the 7' phase may be indicated by the presence of a weak (111) reflection from this phase adjacent to the strong (111) matrix reflection. The absence of such a reflection is not definite evidence that the 7' phase is not. present, however, since its reflection may coincide with that of the matrix phase. In some cases the very weak (100) and (110) superlattice reflections of the r' phase can be detected and will resolve this amibiguity.

T- I;'i` -'',i,''....... i r ~- 4-4. ~_ 3 1.: / o X - ~t,:!, i ~ I 7 i o3 tL':L0 1X+ t 012t @ Pn0 i S. L tt | 0i |; 4 jD 0"f 7-.... V 1 t k 41 7;:-:-= ~L -t —Li-444-L t | -] o a |0 ~ S TH i E-4 L i0 II/i~ -,~,,~-,,o,~,=..... ~'-. —i —' i..._.. 1 K.. r.' 3 -I-) cL —, ~~- u'r r~~~~FT

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN X-ray diffraction studies of this kind have been made on specimens of Inconel-X alloy aged 100 and 1000 hours at 14000~F 1000 hours at 1600~F and 1000 hours at 12000F. The specimens were prepared for these studies by mechanical polishing followed by a rather heavy etching treatment with the modified etch described above. This procedure resembles closely the one used by Nordheim.I Definite evidence for the presence of the 7' phase in this alloy has been obtairned, but only for the case of the specimen aged 1000 hours at 14000F. The traces of the (111) reflections from this specimen and from the specimens aged 1000 hours at 1200( and 16000F are reproduced in Figure 4. The weak (111) reflection of the 7' phase is evident in the trace from the 1400~F specimen but no in the other traces. In no case were the (100) or (110) superlattice reflections detected. These results are considered to be only preliminary since the sensitivity of the method is strongly dependent on the condition of the surfaces of the specimens. It is possible that the etching treatments used were not adequate to remove all of the worked metal formed in the mechanical polishing operation. This would not only decrease the intensities of the reflections but would tend to reduce the resolution of the 7 and y' reflections. Therefore it appears advisable to carry out some further studies using different etching techniques and particularly using electrolytically polished specimens. Nevertheless, these present results show an interesting correlation with the microstructures of the specimens as revealed by eleqtron microscopy. The micrograph from the specimen aged 1000 hours at 1400~F show the presence of numerous small cubic particles which are arranged along definite rows in the matrix grains (see Figure 1). These particles do not have the form usually exhibited by the carbide and nitride phases indentified;. in this specimb but closely resemble the cubic y' particles observed by Taylor and Floyd.8 Furthermore, they do not show in the micrographs from the specimins aged at 1200~ and 1600~F (see Figure 2). On this basis it appears that these are the particles of the 7' phase. N-155 Al Recent work on this alloy has been concerned primarily with the question of the occurrence of an intermetallic Fe2W phase as a result of certain conditions of solution-treatment or high-temperature aging. Previously,4 electron diffraction patterns agreeing closely with the pattern reported for this phase were obtained from specimens solution-treated for 10 hours at 22000F and aged for 1000 hours at temperatures of 14000 and 1600~F. While this identification was considered to be only tentative,.the tter was of inteest since similar patterns were not obtained from specimens solution-treated,- for 2 hours and aged for periods up to 100 hours at these tenmperatures. \ 8 1

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Extensive electron diffraction examinations have been made of the specimens of the previous investigations using a variety of polishing and etching procedures. In all cases the specimens solution-treated for the 2-hour period at 22000F gave only patterns of the columbium carbonitride, M23C6, and matrix phases. A typical pattern is reproduced in Figure 5a. In several instances, however, both specimens solution-treated for 10 hours gave patterns corresponding to the Fe2W pha-se. The best patterns of this type were obtained from mechanically polished surfaces after etching with HF-ethyl alcohol-glycerol reagent described above, or with a solution consisting of 300 ml of sulfuric acid (96%), 150 ml of phosphoric acid (85%) and 50 ml of water. One of these patterns is reproduced in Figure 5b, and a. Cb(C,N) and M23C6. b. "Fe2W. Figure 5. Electron Diffraction Patterns Obtained From N-155 Alloy. the interplanar spacings and intensities of the diffraction rings of this pattern are compared with those reported for Fe2W in Table II. The etching treatments required to obtain this pattern are rather severe (3 min at 0.2 ampere per sq. in.) and roughen the surfaces to such an extent that replicas for electron microscopic examination cannot be obtained. With lighter etching treatments the microstrctures of the specimens are clearly revealed (Figure 6) but the patterns corresponding to Fe2W are not obtained. X-ray diffraction studies have been made of the minor phases separated by the method of Mahla and Nielsen7 from the specimens which produce the "Fe2W" pattern and from several other specimens. Patterns of the columbium carbonitride and the M23C6 phases were obtained from all specirnens, but in no case was there an indication of the Fe2W phase.

Table II. Interplanar Spacings and Relative Intensities* for the Electron Diffraction Patterns Obtained From N-155 Alloy Specimens Solution-Treated for 10 hours at 2200~F. E. D. Pattern Fe2W Pattern** dhkl I dhkl 3.61 vw 3.60 vw 2.80 m 2.80 m 2.44 vw - 2. 34 s 2.36 s 2.17 s 2.17 s 2.14 - - 2,04 m 2.,05 m 2.01 $ 2.0'2 s 1.99 s 1.98 s 1.92;w 1.94 w.1.81 vw 1.82 1.74 w 1.74 w 1.61 w 1.61 vvw _ - 11.55 in 1.51 w 153 m 1.49 w 1.50 vw 1.44, vw 1.44 w 13.9 VW 1.39 w 1.35 vw 1.37 m 1.33 m 1.33 5 1.29' m 1.29 s 1.23 m 1,24 s w -Weak, m = Medium, s = Strong, v = Very. ** ASTM Card 3-0920. 10

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN a. Aged 1000 hrs at 14000F:, b. Aged 1000hrs at 16000F. Figure 6. Electron Micrographs Showing Microstructures of N-155 Specimens Which Give the "Fe2W" Patterns, X10,000. The electron diffraction results presented here support the previous identification of an Fe2W phase in the N-155 alloy. The fact that indications of this phase have been obtained from specimens solution-treated for the 10-hour period but not from those solution-treated for the 2-hour period suggests that the solution treatment may be the principal factor in the formation of this phase. The fact that the x-ray studies give no indications of the Fe2W is not considered important since an intermetallic phase of this type would probably dissolve in the bromine reagent usedto separate the minor phases from the matrix. It is possible, however, that the electron diffraction patterns are produced by one or more products formed on the specimens by the etching treatments. This possibility cannot at present be eliminated in view of the fact that very heavy etching treatments are required to produce the patterns and particularly because then surfaces could not be examined by electron microscopy for conta-mination. 11

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN DISCUSSION The results on the minor phases of Inconel-X alloy which have been described here are more extensive and more complete than any previously reported. Four different minor phases have been identified in this alloy: titanium nitride, columbium carbonitride, the complex M2sC8 carbide, and the intermetallic T' phase. Of these, only the first two have been reported in previous studies of this alloy. ~ In addition, the present work has shown that the development of these phases is strongly influenced by the temperature of the heat treatment. Considering the good agreement between the electron diffraction, electron microscope, and x-ray diffraction results, it appears that the studies of the present series of specimens of this alloy are nearing completioni Some further x-ray diffraction and electron microscope studies will be carried out to determine more closely the range of conditions under which the r' phase occurs and to provide a more complete correlation with the microstructure of the alloy. In addition, it appears advisable to make careful x-ray diffraction studies of the carbides extracted from a few more specimens to give a more complete corrlation of the x-ray and electron diffraction methods. In the case of the N-155 alloy, the present results lend scme support to the occurrence of the Fe2W phase in this alloy; however, additiomal work is proposed on this question. In th meantime the studies of the carbide phase of this alloy are being continued.

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN BIBLIOGRAPTI 1. Ieidenreich, R. D., Sturkey, L., and Wood, H. W.., J. Applied Phys, rI, 127 (1946) 2. Heidenreich, R, D., "Electron Microscopy" in Physical Methods in Chemical Analn sis, Vol. I, edited by W G. Berl, Academic Press, Inc.* New York, 1950, p. 535. 3. Heidenreich, R. Do, "Electron Diffraction and Electron Microscopy" in dern esearch TechnAe s in Physical Metallxrgy, American Socitty for Metals, Cleveland, 1953. 4. Brockway, L. 0. and Bigelow, W. C., Progress Report No. 7, Project 2020, Eng. Res,. Inst. Unit. of Mich., to Aeronautical Research Laboratory (WCRRL), WADC, 15 October 1953. 5. Brock,way, L. O. and Bigelow, W. C., Progress Report No. 3, Project 2020, Eng* Res- nst.,, Univ. of Michb to Aeronautical Research Laboratory (WCRRL),'WADC, 15 April 1953. 6. Taylor, A. and Floyd, R. W., J. Inst. Metals, 81, 1 451 (1953). 7. Mahla, E. M. and Nielsen, N. A., Transactions Ame, Sc. Metals, 43, 290, (1951)* 8. Nordhieim, R., Doctoral Thesis, Departrent of Metallurgy, Massaehusetts Institute of Technaoigr, 1953, 9. Frey, D. N,, Freeman, J. W,, and White, A. E.,National Advisory Committee for Aeronautics, Technical Note 2385, 1951. 10. Rosenbaum, B. M., National Advisory Comminttee for Aeronautics, Technical Note 1580, 1948. 11* Brockway, L. 0. and Bigelow, W. C., Annual Summary Report for 1952 to 19553t Project 2020, Eng. Res. Inst., Univ. of Mich., to Aeronautical Research Laboratory (WCRRL), WADC, 15 January 19535 13

UNIVERSITY OF MICHIGAN 11111111111111111111111111111111111151 13 Ie II 02526 1382 x