Search Documents

By R. F. Mehl, C. E. Birchenall

R. E. Hoffman and D. Turnbull—The authors have presented evidence which they have interpreted as indicating that the rate of self diffusion is not intrinsically more rapid at grain boundaries than within the grain. Grain-size effects which apparently exist are attributed rather to impurities concentrated at the grain boundaries. In view of our own experiments and the existing evidence, we believe that the support for this hypothesis is not convincing. We have in progress an investigation in which the rate of self diffusion of silver is being measured over an extended temperature range in both single-crystal and polycrystalline specimens. The results of the single-crystal experiments and some preliminary data on fine-grained polycrystalline specimens have already been reported:' and it is anticipated that a complete report will be published in the near future. The self-diffusion coefficient of large-grained polycrystalline silver (1 grain per sq mm) has previously been measured by Johnson" between 730" and 940°C. The diffusion coefficients which we have measured in single crystals (210 plane normal to diffusion direction) agree within experimental error with values calculated from an extrapolation of Johnson's curve down to temperatures as low as 500 °C. However, it has been demonstrated that the overall self-diffusion rate in fine-grained polycrystalline specimens (initial grain size of 0.003 cm) becomes measurably larger than the overall rate in a single crystal at a temperature of 600°C, and the discrepancy between the two rates becomes greater as the temperature is further decreased. In fact, it has been possible to obtain satisfactory penetration curves for polycrystalline specimens using the sectioning technique at temperatures as low as 400°C. At this temperature, the penetration is 50 to 100 times greater in the polycrystalline specimens than in a single crystal. Fisher" has developed an analysis whereby the ratio of the rate of the unit diffusion process at the grain boundary to the corresponding rate within the grain can be calculated from the penetration curves and an assumption as to the width of a grain boundary. This analysis applied to our data indicates that the unit process at the grain boundary is faster by a factor of 10' at 475°C when the grain boundary width is taken to be 5. The silver used in most of these experiments was obtained from the Handy and Harmon Co. and listed as 99.97 pct pure. Preliminary experiments on 99.999 pct silver from the Jarrel-Asch Co. indicate a grain-size effect of the same order of magnitude as in the less pure silver. Nominally, these purities are as good, at least, as that of the carbonyl iron used by the authors, but of course if an impurity effect does exist its magnitude might be very dependent upon the nature of both major and minor constituents. The authors have cited the work of other investigators who have found no grain-size effects. Neither Steigman, Shockley and Nix" nor Maier and Nelson8 were able to correlate self-diffusion coefficients of copper with grain size. However, all their measurements were performed at or above 750°C; and on the basis of our work with silver, no grain-size effect would be expected at temperatures above about 0.7 of the absolute melting temperature unless the grain size were exceedingly small. Likewise, in the investigation of the self diffusion of lead by Seith and Keil,14 the lowest temperature at which the diffusion coefficient was measured in polycrystalline specimens was 207°C, which is still sufficiently high so that the lack of a grain-size effect is not surprising. Finally, in those experiments on iron from which they concluded that there was no grain size effect, Drs. Birchenall and Mehl seem to have no information as to the actual grain sizes immediately prior to and following the diffusion anneal. Without this information, we believe that their own experiments offer little support for their hypothesis. F. S. Buffington, I. D. Bakalar, and M. Cohen—The results given in this paper agree in order of magnitude with those tentatively reported by us.27 However, significant differences exist in the two sets of data, and it may be well to make an explicit comparison. The diffusion studies at M.I.T. were conducted on somewhat higher purity iron (99.98 pct Fe) than the grades used by the authors, but this is undoubtedly not the answer. Fig. 4 shows the diffusion results of both laboratories for the gamma phase, omitting the authors' data on the commercial steels, while fig. 5 presents a similar comparison for the alpha phase. The divergence is much more marked in the latter case than in the former. In connection with the M.I.T. determinations, all of the runs in the gamma range and those above 800 °C in the alpha range were conducted with specimens having a relatively thick (0.002 cm) electrodeposit of radioactive iron. This practice minimizes any possible error due to extraneous diffusion that may occur during the heating to and cooling from the operating temperature. An exact solution of Fick's law for these boundary conditions was used in calculating the diffusion coefficients. At a later time, three runs were made below 800°C, using very thin electrodeposits similar to those of the authors, and the points fell considerably below the values expected from the extrapolation of the results based on the specimens with the thick deposits (compare dash-dot line in fig. 5). However, in the runs with the thin deposits, deviations of 100 pct were found between the individual specimens, whereas the maximum deviation with the thick deposits was less than 25 pct. Accordingly, it is not known at the moment whether the M.I.T. points below 800 °C should be given as much weight as those above 800°C. If this were done, the frequency factor would be of the order of 400 cm2 per sec, which is quite high. In other metals, the frequency factor for self diffusion lies between about 0.1 and 10 cm2 per sec. As the points below

Jan 1, 1951

By F. E. Jaumot, R. L. Smith

Self-diffusion in zinc at temperatures below 200°C has been studied using both single crystal and polycrystal samples. Anomalous results were obtained for single crystal samples, the data indicating that in some cases grain boundary diffusion predominated. These anomalous results are presumed to be due to low angle lineage boundaries in the single crystals. When volume diffusion occurred in either single or polycrystal samples, the values for the diffusion coefficient were as much as several orders of magnitude larger than would be expected from high temperature data. ONSIDERABLE work has been done on self-^-*diffusion at temperatures in the neighborhood of two thirds of the melting temperature (in OK) and higher, but very little has been done at lower temperatures. This is particularly true of self-diffusion in single crystals. This paper reports work on self-diffusion in zinc at temperatures below 200 °C where both polycrystalline and single crystal samples have been used. The measurements were made using both the sectioning and absorption techniques. The experiments on diffusion at low temperatures were initiated primarily for three reasons: to obtain data from single crystal diffusion covering a wide range of temperatures, to obtain values of the activation energy for grain boundary diffusion, and to obtain, if possible, a lower limit to the temperatures at which the absorption technique may be used. It is possible to determine whether volume or grain boundary diffusion predominates at low temperatures by analyzing the manner in which the concentration of the diffusing atoms varies with the depth of penetration. The analysis is possible only for the sectioning technique. For volume diffusion and for the boundary conditions of the present experiment, the curve of the logarithm of the concentration, C, vs the square of the penetration depth, x, is a straight line with the slope proportional to — 1/D. If grain boundary diffusion predominates, Fisher' has shown that a curve of In C vs the first power of the penetration depth should be a straight line. This linear dependence of concentration on depth has been observed experimentally by several investigators. (See, for example, refs. 3 and 4.) Following Fisher's analysis, the grain boundary diffusion coefficient can be written as where 8 is the grain boundary width, D, is the volume diffusion coefficient at the temperature involved, t is the time of diffusion, C is the concentration of diffusing material, and x is the depth of penetration. A more general analysis employing the same model as Fisher's but with fewer assumptions has been given by Whipple.' In contrast to Fisher's analysis the more general analysis of Whipple does not always yield a straight line relation between In C and x. Recently Turnbull and Hoffman have discussed both analyses and have applied the Fisher-Whipple model to a more refined dislocation model of the grain boundary. They have also discussed the possibility of a large effect of grain boundary direction on diffusion. In the case of the absorption technique, for which results are also given, one simply gets a value for the diffusion coefficient with no indication of the predominating mechanism. Experimental Techniques Single crystals of 99.99+ pct Zn were grown using a modified Bridgman method. Polycrystal samples were prepared in three grain sizes (small, about 4 mm2; medium, about 9 mm'; and large, about 25 mmz) by quenching, air cooling, and furnace cooling. The grains of the polycrystalline samples tended to be columnar, with some preferred texture such that the c-axis was inclined at about 20" to 25" from the normal to the diffusion face, particularly in the medium and large grain size samples. The samples were cut from the rod, polished, and heavily etched, and faces cut accurately parallel using a jeweler's lathe. They were again etched to remove the lathe smear and annealed for 3 hr at 250" to 300°C. After annealing, they were etched, examined, and the orientation determined by Laue back reflection photographs. They were then electroplated with Zn. The plated surfaces were examined for uniformity; nonuniform coatings were removed by etching and the samples replated without additional preparation. For the diffusion anneals, the samples were placed in pairs, with active faces together, in evacuated Pyrex tubes. Constant temperature baths were used, and the variation in temperature did not exceed ±l°C, Diffusion runs were made at l00°,

Jan 1, 1957

By L. Himmel, R. F. Mehl, C. E. Birchenall

The rates of self-diffusion of iron in artifically prepared wustites of various compositions have been determined using the decrease in surface activity technique. Similar measurements are reported for artificial magnetites of nearly stoichiometric composition and for natural hematite single crystals. These data, together with appropriate thermodynamic data derived from the present study or from the literature, are used to calculate the rates of oxidation of iron and its oxides, taking as a basis the theoretical rate equations developed by Wagner. The calculations are compared with experimental rate constant data for these same reactions and are shown to provide essential confirmation for the Wagner theory. For the most part, the results are also consistent with the previously proposed transport mechanisms in the oxides of iron. WHEN a metal is exposed to an oxidizing atmosphere at elevated temperatures, a superficial oxide layer is formed which then thickens according to some characteristic rate law. Provided that the oxide layer is dense and continuous and does not present any paths for gaseous diffusion, continued growth must inevitably involve the migration of the reactants—either the metal or oxygen or both—through the lattice of the solid oxide. This is true whether the reaction rate is controlled by the diffusion process itself or by a slow step in a reaction at one of the interfaces. If the overall rate is limited by diffusion, and if the concentrations or thermodynamic activities of the reactants at the boundaries of the oxide layer are independent of time, the familiar parabolic rate law is observed. Under these conditions, Wagner1 has shown that the rate constant may be expressed in terms of the mobilities of the reacting components and the appropriate concentration or activity gradients which are established across the growing oxide layer. In presenting the phenomenological basis for the theory, Wagner has emphasized the role played by lattice defects in diffusion processes in oxides. The essential validity of this theory has been confirmed by experiment for a number of simple systems including the growth of Cu2O on copper.'-' One of the objectives of the present study has been to extend the Wagner approach to the more complex process of the oxidation of iron and its oxides, and particularly to wustite, which is normally the major constituent of the oxide scales formed on iron at temperatures above approximately 570°C. For this purpose the rates of self-diffusion of iron in w?stite, magnetite, and hematite have been measured by tracer techniques and the necessary thermodynamic data have been obtained from equilibrium studies involving the oxides and their atmospheres. These data have subsequently been used to test the underlying assumptions of the Wagner theory as well as the detailed mechanism of the scaling of iron previously proposed by Davies, Simnad, and Birchenall.5 Lengthy treatments of diffusion and conduction phenomena in ionic crystals have been published." * The latest, by Jost,8 also summarizes much of the existing experimental data. In general, however, the oxides of iron have not been considered specifically and hence a brief review of the available

Jan 1, 1954

By C. E. Birchenall, R. H. Condit, M. J. Brabers

In the oxidation of pure iron above 700°C the overall rate is determined mainly by the rapid growth of wiistite, through which iron ions can diffuse rapidly.' Nickel added to the iron progressively decreases the stable range of composition of wustite.' When enough nickel is added to eliminate wiistite as a stable phase, the spinel is the fast growing phase. This investigation was undertaken to test the hypothesis that nickel substituting for iron in magnetite might decrease the rate of spinel growth by reducing the iron mobility. Nickel ferrite boules grown by the flame fusion process were purchased from the Linde Co. Chemical analysis gave an atomic ratio of iron to nickel of 2.88 and showed no other cations. They were cut into discs and ground to give flat, parallel faces. The experimental procedure employed was like that of Eisen and Birchenall in which the distribution of radioactivity diffused in from a vapor deposited surface layer was determined by counting the residual activity through a series of ground faces. The diffusivities were calculated in a manner consistent with the requirements discussed by Condit and BirchenallP The details will be given in connection with a more extensive spinel diffusion study now in progress. The results of the measurements are shown in the figure in comparison with earlier data on magnetite1 and zinc ferrite. The uncertainties arising from scatter in the activity vs distance relationship and uncertainties in the temperatures are shown by flags. Despite the scatter it is clear that substitution of a considerable part of the iron in magnetite by nickel, in this case about one third decreases the iron mobility substantially. sachs' determination of the distribution of iron and nickel in scales on iron-nickel alloys appears to show that nickel is less mobile than iron in the spinel layer. It is evident, therefore, that nickel plays several roles in reducing the oxidation rate of iron-nickel alloys. 1) It reduces progressively the stability range of wiistite. ' 2) It reduces progressively the stability range of the iron-nickel spinel phase in terms of the variation in cation to anion ratio.' 3) As shown in this investigation, it reduces the cation mobility in the spinel phase which is the faster growing phase in the absence of wiistite. The fe was obtained from the Oak Ridge National Laboratory. The research was supported by the Air Force Office of Scientific Research (ARDC) under contract number AF 18(600)-967.

Jan 1, 1961

By H. I. Aaronson, H. A. Domian

The self-diffusivity of Ag10 has been measured as a function of temperature and composition in AgMg. a CsCl-type intermetallic compound with a substitutional defect structure on both sides of the stoichiometric (50 at. pct Mg) composition. These measurements show that In D increases linearly, and both In Do and Q decrease linearly with deviation from stoichiometry. The slopes of these lines are significantly steeper in magnesium-rich alloys. Equations for the variation of Q with composition have been derived by application of the theory of random walk processes on the assumption that diffusion in compounds of this type takes place by the six-jump cycle mechanism of Huntington, Elcock, and McCombie. The favorable agreement obtained with experiment furnishes fresh support for the correctness of this assumption. A principal result of this treatment is that. although the concentration of vacancies is considerably higher in the silver sublattice than in the magnesium sublattice, the acfivation energies for movement of the magnesium sublattice vacancies are so much lower that these vacancies are responsible for effectively all diffusion of silver in this compound. A tendency for vacancies of both types to be kinetically bound to wrong atoms is also noted. THIS study was undertaken to examine the influence of substitutional defects upon self-diffusion in an intermetallic compound, AgMg, with a CsCl-type crystal structure. The CsCl structure is exception- ally simple. At the equiatomic or stoichiometric composition in a binary alloy, it consists of two interpenetrating simple cubic sublattices, each of which is exclusively populated by the atoms of one species. When atomic size and other factors are favorable, deviations from stoichiometry can be entirely accommodated by the substitution of atoms of the species present in excess of 50 at. pct on the sublattice of the other species on a one-for-one basis. Under this condition, the only vacancies present are those formed by thermal activation. The excess atoms thus incorporated into the sublattice of the minority species may be considered as substitutional defects in this sublattice. Several X-ray studies have shown that the defect structure of AgMg is entirely of the substitutional type on both sides of the stoichiometric composition.1"3 The thermodynamic investigation of Kachi indicates that, aside from a low level of thermal disorder and these constitutional defects, order is maintained to the solidus temperature. This compound is therefore a particularly useful vehicle for a study of the influence of substitutional defects upon diffusion kinetics. Other advantages of AgMg for this purpose include: the broad composition region, as much as approximately 10 at. pct on either side of the stoichiometric (50 at. pct Mg), in which the compound is stable; the comparatively low temperature range of the solidus (820°C and below),5 making diffusion anneals near this range experimentally convenient; the reasonably good machinability of the compound, permitting utilization of the standard sectioning method of radiotracer analysis; and the availability of a good tracer, A for one of the constituent species. While this investigation was in progress, Hagel and westbrooks reported data on the self-diffusion

Jan 1, 1964

By D. L. Feucht, R. L. Longini

The semiconductor heterojunction is considered in terms of simple models which may lead to an understanding of move complex heterojunctions. Metallurgical and electrical properties of hetero-junctions aye discussed including the interface structure, energy -band diagram, and carrier transbovt across the interface. It is found that in a heterojunction all mechanisms such as injection, tunneling, and junction recombination found in simple junctions play modified voles. INTERFACES between materials (grain boundaries, the electrical junction between two differently doped materials in a single crystal, the oxide-metal interface, or metal-metal junctions) are of considerable importance in many situations. These various interfaces all have one very fundamental thing in common. Quantum mechanically speaking, the wave functions of the electrons in one material may penetrate the other material but, in general, only to the extent of angstroms. From an electrical point of view the conduction mechanism changes as a current passes through such junctions. In some cases the change is tremendous, in others almost negligible. The interface, then, is the locus of a change of conduction mechanisms. Some of these, particularly in semiconductors, are well-understood. The ordinary p-n junction in a single crystal can be the locus of an injection mechanism or a tunneling process, depending on conditions. The mechanisms are probably best understood in semiconductors because of the possible simplified view of particlelike conduction. The bands are either nearly filled or nearly empty and band overlap is seldom involved. The same fundamentals are probably important in other situations too but they are very difficult to look at naively. Although the simple look at the semiconductor case only gives us a relatively rough picture which must then be refined, the other systems, which involve a more complex situation, immediately are in many ways too difficult. There are too many initial choices of complex systems and therefore it is not possible to be even reasonably certain of any one model. Because of the relative simplicity of semiconductors, their good and controllable structure, and because of the ability to make many measurements on them not normally available to either metals or insulators! they are probably the best understood materials. It is therefore desirable to use them as a tool to further the understanding of interfaces in general. Semiconductor-heterojunction concepts were first proposed by kroemer1 in 1957. This was followed several years later by reports on the fabrication and experimental characteristics of heterojunction structures by Anderson2 and Diedrich and jotten.3 I) THE HETEROJUNCTION STRUCTURE To get down to hardware, when we refer to a semiconductor heterojunction we imply that there exists an intimate contact between different semiconductor materials. We could put two pieces of material together, complete with oxide layers, we could remove the oxides, or we could even melt the interface and hopefully get wetting and a good "bond" on solidifying. In fact we could by some means grow a crystal of one material using the other as a seed. Essentially we are interested only in the last two because they are the simplest to look at analytically. The degree of perfection of fit varies greatly and is reflected somewhat in the arc welder's joint strength. The lattice match of the two materials, their orientation, and so forth. is obviously necessary for a good bond but so is the continuity of any polar bonds which are involved such as in the III-V semiconductors. The mechanical misfit between two similar lattices can be described in terms of edge dislocations. The edge-type dislocations must be very close together for the usual misfit and there must be dislocations for each of several different Burger's vectors in order to produce a lattice match. The .'dangling bonds'' resulting will be involved in producing interface charge. Order of magnitude estimates of the charge density extrapolated from low densities of dislocations in homogeneous materials give 5 x 1013 cm-2 Ge-Si and 1 X 1012 cm-2 Ge-GaAs electronic charges. Edge dislocations also act as very active recombination centers between holes and electrons. One lattice "matching" difficulty usually exists even if two structures have essentially the same lattice constants as they will have different coefficients of therma1 expansion. Thus, on cooling from the usually high temperature of fabrication to room temperature, dislocations are produced, a good fit not existing at both temperatures. In brittle materials this shrinkage may even result in cracking. For the Ge-Si interface the mismatch is about 2 x 10 -6 per degree whereas it is less than 10"7 per degree between germanium and GaAs. The exact effect of the misfit is dependent on the thickness of the materials involved. For a very

Jan 1, 1965

By A. J. Shaler

The subject of the mechanism of sintering has received much attention in the past few years, particularly since the beginning of the series of AIME seminars in powder metallurgy of which this paper introduces the fourth. In the first of these, F. N. Rhines1 brought together and discussed the available experimental data on the sintering of pure metallic powder, and succeeded in bringing to a sharp focus the attention of workers in this field on the established observations which a satisfactory theory must explain. Several other authors3,5,6 have, in the last few years, studied the phenomena that occur when cold metallic powders, loose or in the form of compacts, are first brought to elevated temperatures. Some workers' in the field of friction have recently studied the adhesion of solid metal surfaces when they are brought into close contact. These researches have indicated that several separate mechanisms operate simultaneously, at least during the first part of the sintering process. Some of them have been called transient mechanisms4 because they are in general not absolutely necessary to sintering. Powders may be so prepared and so treated that these transient phenomena do not take place during subsequent sintering. This does not mean, of course, that their industrial and scientific importance is any less than that of the steady-state phenomena. The latter are changes that go on during sintering no matter how the powders are made or treated; they cannot be divorced from sintering. One way to analyze the process of sintering into its component parts is perhaps to distinguish between these transient and steady-state phenomena. Some of the transient phenomena have been studied in the past few years. Huttig3 has shown that, when the temperature of metallic powder is slowly raised, the following events generally occur in order: (1) physically adsorbed gases are desorbed; (2) there is an atomic rearrangement of the surface, a sort of two-dimensional "surface-reciystallization"; (3) there is a breakdown of chemically adsorbed surface compounds; (4) there is a recrystalliza-tion in the volume of the metal. All these changes are shown by Huttig and his coworkers to be completed fairly rapidly at lower temperatures than those generally used in sintering and are therefore not a part of the mechanism whereby the density of a mass of powder continues to change after long heating at an elevated temperature. But the first and third of these changes release gases in quantities which may or may not help to control the steady-state mechanisms, depending on when the voids become isolated from the outside of the compact. Among the phenomena studied by Steinberg and Wulff,8 there is the effect on sintering of residual stresses arising from the pressing operation. They found that the lateral surfaces of a green compact of iron are under a longitudinal residual tension-stress of the order of magnitude of half the yield-point for solid iron. If the outside surface is in tension, the core must be under longitudinal compression. When the compact is heated, the surface residual stress is thermally relieved first, and the compact therefore initially expands in the direction of its axis. This is a transient phenomenon, if for no other reason than the possibility of sintering unpressed powders, as demonstrated by Delisle,9 Libsch, Volterra and Wulff10 and others.1 The subject of recrystallization is dealt with further in a separate section, in view of its prominent place in sintering literature. It, too. is one of these transient phenomena. Among the steady-state parts there may be distinguished the attraction between particles and its consequences, the spheroidization of voids in the compacts, and the densification or swelling of the compact. There is considerable evidence4,7 showing that cold metallic surfaces, when brought to within a few interatomic distances of one another, are attracted to each other by forces of the order of many thousands of pounds per square inch. A calculation, discussed in greater detail in another section, shows that this force changes but slightly when the temperature of the surfaces approaches the melting point. Actual measurements of forces of adhesion of this magnitude have been made by Bradley12 on some nonmetals, but none has yet been made on cold or hot metals. This force is of sufficient magnitude to cause some plastic deformation in powder compacts, as will be shown below. A second force of steady-state nature is due to the surface tension, which probably has the same origin as the force of attraction between surfaces.164 A paper by Udin, Shaler, and Wulff1,3 gives the results of precise direct measurements of its value for solid copper. The demonstration of the tendency for the surface tension to shrink a pore was long ago given by Gibbs.17 He showed that its effect on a curved surface between two phases is equivalent to a pressure perpendicular to that

Jan 1, 1950

By George R. Cowan

A number of studies hare been reported of the effects produced in metals subjected to deformation by shock waves with maximum pressures ranging from tens to hundreds of kilobars. On the basis of the equations for the flow of mass, momentum, and energy through a stationary shock front, the macroscopic stress-strain curve for the resulting shock deformation can be calculated within narrow limits from the experimentally determined Hugoniol curve. In relatively weak shocks which are preceded by an elastic wave, the stress rises above the clastic limit only as plastic deformation proceeds cold thus the shock has a long toe. In strong shocks that override the elastic wave a high stress is applied without prior plastic deformation. A more important effect of increasing the shock pressure is the generation of shear stresses, called supercrilical shear stresses, that exceed the strength of the perfect lattice. A change in the mechanism of deformation is expected to result from the onset of supercritical shear. The shock disordering of ordered Cu3Au in strong shocks appears to be an example of such a change. It is suggested that the formation of fine twins in copper and nickel and the formation of structures which enable visible twins to be formed in the rarefaction ware, observed in copper and presumably in disordered Cu3 Au, are related to the occurrence of supercritical shear in shock dcformation. In recent years several studies1,2 have been made of the changes in structural and mechanical properties of metals produced by the passage through the metals of strong shock-compression waves ranging from about 50 to 800 kbar pressure. Recent work involving dynamic measurements of the shock compression "Hugoniot" curves 3-8 of many metals has developed techniques and provided data required to obtain the shock pressure and the (transient! plastic deformation produced in the shock-conlpression experirnents.9 Shock deformation has been found to be much more effective than slow deformation in changing the mechanical properties of metals, when the two are compared on the basis of equal plasti strain, Holtzman and Cowan9 made quantitative estimates of the shearing stress occurring in a shock front in a metal by assuming that the shearing stress is similar to that occurring in a shock front in a viscous, heat-conducting fluid, with the addition of a yield stress. Taylor's solution9 for a weak shock was used to estimate pairs of values of shearing stress and thickness of the shock front obtained by assumed choices of the ratio of effective kinetic viscosity to thermal diffusivity. It was noted from these values that. unless the shock front is extremely thin. heat conduction has slight effect, and the shearing stress is nearly independent of the mechanism of deformation. This mechanism does, however, determine the thickness of the shock front and the rate of strain. Furthermore, since the maximum possible shearing stress occurring in shocks of moderate strength does not greatly exceed the shear stress occurring in conventional slow deformation, the mechanism of deformation is not expected to be qualitatively different. The greater effectiveness of shock deformation in changing the mechanical properties of metals can be attributed partly to the fact that dislocations, when driven by near-conventional stresses, cannot keep up with the shock front, thus necessitating a higher dislocation density than required for an equivalent slow strain. The fast uni-axial strain occurring in the thin shock front would also be expected to cause a larger number of dislocation intersections to occur. In the upper range of shock pressures that have been studied the estimated values of the shearing stress exceeded the estimated shear strength of a perfect crystal. Under these circumstances it is reasonable to expect that the mechanism of deformation might be considerably different from that involved in slow deformation. Except for the observation by smith1 of twins in shocked copper, the effects of shock waves on metals did not show any obvious or large changes in properties that would indicate the onset of a change in the mechanism of deformation. The recent investigation of the effect of shock waves on ordered and disordered specimens of Cu3Au by Beardmore, Holtzman, and ever" showed a spectacular decrease in the amount of long-range order retained by initially ordered Cu3Au when the shock pressure was raised from 290 to 370 kbar. Since Dr. Holtzman and I suspected that this behavior probably was due to the onset of a shearing stress in the shock front in Cu3Au which exceeded the limiting shear strength of the perfect crystal. it was considered appropriate to examine directly the shock-front equations for a solid. and to obtain a sound estimate of the shearing stress occurring in the front using equation of state data obtained from shock studies. In this paper an estimate is made of the

Jan 1, 1965

By W. V. Green

The creep-rupture behavior of commercial powder-metallurgy tungsten rod is reported for temperatures of 2250°, 2500°, 2700°, and 2800°C, stresses up to 7000 psi, and times up to 4 hr. The temperature dependence and the stress dependence of creep rates are evaluated. Rupture time is fitted to the Larson-Miller parameter and compared to lower-temperature and higher stress results reported by pugh.4 THE tensile strength of single-crystal wire1 and of polycrystalline wire2 have been measured by Goucher. More recently Bechtold3 and pugh4 have studied the mechanical properties of tungsten rod between room temperature and 1200°C. Since tungsten has the highest melting point of all metals, it should retain useful strength to very high temperatures, and should find many applications in missiles, rockets, and so forth. Therefore, the mechanical usefulness of polycrystalline tungsten rod was investigated in the 2250" to 2800°C temperature range by short-time creep-rupture tests. MATERIALS TESTED Creep specimens were machined from 1/2-in. round commercial tungsten rod which was manufactured from powder by cold pressing, sintering, and hot swaging. Purity reported by the vendor was 99.95 pct W. Spectrochemical analyses showed "traces" (less than 0.01 pct) of silicon, iron, and titanium, and "faint traces" (less than 0.001 pct) of calcium and manganese. Carbon content was 30 ppm and nitrogen content was 70 to 110 ppm. No oxygen analysis was obtained. Room-temperature stress-strain tests on the as-received tungsten gave badly scattered results. Threaded end specimens broke at the threads. When these specimens were retested using shoulder grips, the specimens broke at the shoulders. No fractures occurred in the gage lengths. The fracture stresses were: 127,000; 137,000; and 154,000 psi for three specimens which fractured at the shoulders. The microstructure of the as-received tungsten rod is shown in Fig. 1. SPECIMENS USED Fig. 2 shows the two types of creep specimens tested. The threaded-end type specimen, which was used in most tests, had a 3/4-in. gage length and 1/4-in. gage diam. The cone-end specimen, which was used primarily to investigate the effects of a smaller gage diameter and the practicability of the modified grips and specimen, had a 3/4-in. gage length and a 1/8-in. gage diam. In both cases, the gage length terminated at 0.005-in. high shoulders which served as fiducial marks for optical strain measurements. Surface finish on the gage length was 16 µ-in. rms. Creep rates and rupture times were not significantly different for the two types of specimens and were not affected by electropolishing the specimens before testing. TESTING PROCEDURE The constant-load creep-testing machine used has been described by Smith, Olson, and Brown.5 In it the specimen is held vertically by grips of similar material and on the axis of a cylindrical tungsten or tantalum heater-tube. The specimen is loaded by means of the specimen grips, a pull rod (which is sealed to the chamber lid by a vacuum-tight bellows), a mechanical lever system, and hanging dead weights. Both specimen and grips are heated by radiation from the heater-tube, which is heated by its own electrical resistance. An optical system views the incandescent specimen through slits in the radiation shielding and heater-tube. The optical system projects an enlarged image of the specimen gage length on a ground-glass screen. Periodic gage-length measurements are made on this image with two cathetometers. Thorium oxide coatings on the specimen threads or conical end surfaces completely prevented diffusion-welding of specimens to grips, and were used

Jan 1, 1960

By T. L. Chu, G. A. Gruber

Silica films hare been rleposited 011 silicon substmtes at 400° to 600°C by a chemical-transport technique using hydrogen fluoride as the transport agent ill a closed system. This transport takes place from a source materia1 1071: temperature to substrates at higher temperatures, as indicated by the thermochemistry of the transport reaction. The experimental variables of- the transport process, such as the substrate temperature, the pressure pi the transport agent, and so forth, have been studied. The rate -determining step of the transport process appears to he the ),ale of chemical reaction in the source region. The transported films are similar to thermally grown silica films in physical proper-ties with the exception of 'some what higher dissolrrtion rates. SILICA films deposited on suitable substrates serve many purposes in electronic devices. They are used for the fabrication of tunneling devices, the surface passivation of devices, and the shielding of devices from nuclear radiation: and as selective masks against the diffusion of specific impurities into semiconductors. Doped silica films can also be used as sources for the diffusion of impurities into semiconductors. Several oxidation and deposition techniques for the preparation of silica films have been developed to meet the requirements of these applications. The therma1 oxidation of silicon by oxygen or steam at temperatures above 900 C is commonly used in silicon technology. The deposition techniques are perhaps more advantageous since they usually require lower temperatures and are not limited to silicon substrates. Silica films have been deposited on silicon and other substrates by reactive sputtering and chemical reactions. The sputtering of silicon in an oxygen atmosphere is capable of depositing good-quality silica films on silicon' and gallium arenide. Many chemical reactions are known to yield silica at room temperature or higher. These reactions may involve intermediate steps. However, the final step yielding silica should take place predominately on the substrate surface in order to produce adherent films. When silica is formed in the gas phase by volume reactions, no adherent deposit can be obtained. Generally, the experimental conditions of a reaction can be varied so that the surface reaction predominates over the volume reaction. The chemical reactions which have been used successfully for the deposition of silica films are briefly as follows. The pyrolysis of alkoxysilanes in an inert atmosphere or under reduced pressure has been employed to deposit silica films on germanium3 and silicon4 at 650" to 750°C in a flow system. The deposition of silica films from alkoxysilanes has also been achieved at nearly room temperature by a low-pressure plasma. Device quality silica films have been deposited on germanium and gallium arsenide by the deposition of an amorphous thin silicon film followed by oxidation at 600" to 700" . Silica films for high-temperature capacitors have been produced by the hydrolysis of silicon tetrabromide at 950°C in argon and hydrogen atmospheres.7 We have developed a chemical-transport technique for the deposition of silica films on semiconductor substrates at relatively low temperatures. The thermochemistry of the transport reaction, the experimental variables of the transport process, and the properties of the transported silica films are described in this paper. THERMOCHEMICAL CONSDERATIONS The transport of solid substances by chemical reactions in the presence of a temperature gradient has been used for the preparation of films and crystals of many electronic materials. In this technique, a gaseous reagent is chosen so that it reacts reversibly with the solid substance under consideration to form volatile products. Since the equilibrium constants of most reactions are temperature-dependent, the transport of these products to regions of suitable temperature in the reaction system would cause the reverse reaction to take place. depositing the original solid. When the equilibrium is shifted toward the formation of the solid as the temperature is decreased, the solid is transported from a high-temperature zone to a lower-temperature region, and vice versa. This chemical-transport technique can be carried out in a closed or gas-flow system. In a closed system, chemical equilibrium is presumably established in the different temperature regions of the system, and the transport agent regenerated in the deposition region repeats the transport process in a cyclic manner. The local chemical equilibrium may not be approached in a flow system: however, this system offers a greater degree of flexibility. Silica reacts reversibly with hydrogen fluoride and this reaction was chosen for the transport process. The over-all reaction between silica and hydrogen fluoride may be written as: SiO2(s) + 4HF(g-) = SiF4Ur) + 2H2O(^)

Jan 1, 1965

By J. D. Lubahn

The influence of precipitation from solid solution on the subsequent deformation resistance of alloys is well known. However, the influence of precipitation or aging that occurs simultaneously with deformation may be entirely different and of considerable importance. It is thought1 that the presence or absence of a yield point jog in impure iron at room temperature and the presence or absence of discontinuous flow at higher temperatures may be directly related to aging phenomena, particularly since the yield point jog is absent when the carbon and nitrogen are sufficiently removed.2 As a preliminary attack on the problem of simultaneous aging and deformation, three kinds of tests—constant strain rate tensile tests, constant load creep tests, and variable strain rate tensile tests—were carried out on an age hardenahle aluminum alloy. Constant Strain Rate Tensile Tests Stress-strain curves at a strain rale of about 0.001 min-1 were obtained for specimens of 61S that had been quenched into liquid nitrogen after a one-hour solution treatment at 521°C, then aged at room temperature 10 min., 20 min., 4 hr, and 26.5 hr before testing. Strain-time curves were obtained on the same chart with the stress-strain curve by means of an auxiliary pen, driven parallel to the axis of the chart drum at constant speed by a synchronous motor and gear chain. The angle of rotation of the chart drum was proportional to strain. The strain rates reported are the measured slopes of the strain time curves. Constant strain rate was approximated (within a factor of 2) by manual adjustment of the oil inlet valve of the hydraulic testing machine in such a way that the course of the recorder pen was parallel to pre-drawn strain-time lines of the desired slope. The specimens were machined from 9/16 in. rod. The cylindrical portion was 0.357 in. in diam by 13/4 in. long and

Jan 1, 1950

By J. H. Brophy

Experimental evidence in recent years shows that nickel coated hydrogen reduced tungsten powder can be sintered to 98 pct of theoretical density at 1100°C. New data indicate that the sintering rate is the same for nickel contents ranging from 0.125 wt pct (mon-atomic layer) to 5.0 wt pct Ni coated on 0.561. tungsten powder. Nickel tungsten made by coreduction from oxides sinters in a kinetically similar way, but the rate tends to increase with higher nickel contents. The activation of sintering can be accomplished with a minimum amount of nickel if coated powder is used. Transverse rmpture strength was found to increase in specimens containing 0.25 pct Ni as density increased during the initial stage of the carrier-phase sintering process. Upon the onset of grain growth in the final stage, strength was found to decrease. With increasing nickel contents up to 4 pct, strength increased. Maximum strengths in three-point loading were observed at 73,000 psi for 0.25 pct Ni and 93,000 psi for 5 pct Ni. These were comparable to that of massive tungsten recrystal-lized in the presence of nickel and tested in the same way. The results were sensitive to the mode of powder treatment and ductility was negligible in all cases. The study and application of sintered bodies of nickel and tungsten is not new; however, the analysis of the mechanism by which small additions of nickel increase the sintering rate of tungsten promises to Contribute new information regarding sintering theory in general. Initial experimental evidence was presented by the authors showing that nickel-coated hydrogen-reduced tungsten powder may be sintered to 98 pct of theoretical density at 1100°C in 16 hr.' Simultaneously, a kinetic analysis of the process showed that it took place in two distinct stages. In the first stage, nickel evidently served as a carrier phase through which tungsten atoms could move. This stage was kinetically similar to the 'solution-precipitationo step postulated when the carrier phase is liquid3 although in the nickel tungsten case, there is no evidence of the existence of a liquid phase either in equilibrium phase relationships or in mi-crostructures at the temperatures under consideration. In the present work, new data for nickel-coated tungsten powder are offered in support of the proposed mechanism, and a more explicit study of the second stage of densification is presented. Concurrently, the sintering kinetics of these coated powders are compared to Ni-W powder made by coreduction of oxides. In this way the present work can be compared to earlier results in nickel activated sintering of tungsten in which the coreduction process was emplyed. The ultimate utility of such an activated sintering process will be determined to some extent by the mechanical and physical properties of the final prod-

Jan 1, 1962

By R. L. Fleischer

IN 1925 a criterion for localized deformation in a single crystal was derived by Taylor and Elam.' They noted that for a single active slip plane the slip plane area is constant, but the direction of the stress with respect to the slip direction changes during deformation so that, when the increase in stress due to this rotation is large enough relative to the work hardening, instability occurs. (By instability is meant continued slip on the active slip plane.) The instability criterion may be written t>dt/d?/m tan2? [1] where t and ? are resolved shear stress and strain, m the Schmid factor for resolving shear stress, and ? is the angle between the slip direction and the axis of tension. The formula points out that regardless of the atomic mechanism involved, for large flow stresses and low work hardening rates, slip on a single plane—once started—should continue. The case of alloying provides a direct possibility of adjusting the parameters in Eq. [I]: As an alloying element is added to a pure metal, the flow stress is raised and the work hardening in easy glide may be either decreased2 or unaltered.3 Since flow stresses in easy glide may rise to the kg/mm2 range for alloy crystals while easy glide slopes may be less than % kg/mm2, 2 it is possible for the inequality to be satisfied. Indeed Garstone and Honeycombe3 report that slip becomes coarser with increasing stress in the easy glide range, where the flow stress would be increasing and the slope would be essentially constant. Another case where high flow stresses and low hardening rates accompany coarse slip is that of quench hardening.4 Maddin and Cottrell report that quenched aluminum crystals display slip markings resembling those of a brass. A similar effect has been observed after irradiation.'

Jan 1, 1960

By S. K. Nowak

The lithium solubility limit in solid aluminum was determined by the use of micro-graphic techniques. The solubility limit thus established was shown to be a true equilibrium by checking the reversibility of AlLi phase precipitation and dissolution around the equilibrium temperature. The experimental results fit a straight line by plotting log N(LI) against 1/T. The metallurgical reasons explaining the lack of agreement among previous investigations are discussed. THE solubility limit of lithium in solid aluminum was determined by Shamray and Saldau1 by micrography and by Grube et al.' and also Voss-Kiihler- by measurement of electrical resistivity. Since the addition of lithium induces extremely small changes in the aluminum lattice, X-ray techniques could not be employed for this purpose; in fact, the addition of lithium does not cause an expansion of the lattice as might be expected from a consideration of atomic radii, but causes a small contraction from 4.040 to 4.037A3, 4 when the solvent aluminum is saturated with this element. An analysis of the results from previous investiga-tions'" shows that with improvement of investigation techniques the solubility limit tends toward a lower lithium content. Again, some of the author's own experimental results on the action of atomic hydrogen on A1-Li alloys seemed to indicate that the solubility of lithium should be much lower, i.e., about 0.3 pet at 300°C: instead of 1.15 pet, the lowest published value.4 For the foregoing reasons it was decided to investigate this problem. The aluminum and the lithium used for this purpose were of the highest purity. The lithium solubility limit was determined by metallographic techniques, in particular by the use of a reagent especially developed for the purpose of revealing the AlLi phase. The solubility limit thus determined was shown to be a true equilibrium by checking the reversibility of AlLi phase dissolution and precipitation around the equilibrium temperature. All of the experimental work was carried out in 1951 at the Centre d'Etudes de Chimie Metallur-gique at Vitry, France, under the direction of Georges Chaudron.

Jan 1, 1957

By R. R. Joseph, K. A. Gschneidner

The solid solubility of magnesium in the close-packed modifications of lanthanum, cerium, praseodymium, neodymium, gadolinium, dysprosium, and lutetium was determined from approximately 250°C to the eutectoid temperature by the X-ray parametric technique. The maximum solubility at the eutectoid temperature was found to be 9.4 at. pet in La, 8.2 in Ce, 9.2 in Pr, 8.2 in Nd, 14.1 in Gd, 13.5 in Dy, and 11.5 in Lu. The solubility when compared at a homologous temperature was found to increase in a smooth manner with decreasing size factor from lanthanum to lutetium. It was found that the apparent atomic radius of magnesium when dissolved in the lanthanide solvent was constant and could be accounted for by consideration of several models dealing with departures from Vegard's law. The eutectoid temperatures for the lanthanide-rich La-Mg, Ce-Mg, and Nd-Mg systems were found to be 544", 505°, and 551°c, respectively. A study of the solid solubility of magnesium in the lanthanide metals was initiated because the valency and electron egativity difference are essentially constant and favorable for solid-solution formation, and thus one could examine the effect of size on terminal solid solutions. The size factors based on metal radii range from 14.6 pet for lanthanum to 7.5 pet for lutetium. Since the electronegativity difference is small, 0.06 unit or less, application of the Hume-~othery' and Darken and Gurry2 criteria for solid-solution formation indicates that the solubility of magnesium would be expected to be small for lanthanum (between 5 and 10 at. pet) and to increase as we proceed along the lanthanide series. A search of the literature revealed no data on the solubility of magnesium in the lanthanide metals except for the solubility of magnesium in cerium at 450oCA typical phase diagram for the lanthanide-rich end of the lanthanide-magnesium phase diagram is shown in Fig. 1. EXPERIMENTAL PROCEDURES Preparation of Alloys. The lanthanide metals used in the preparation of alloys for this investigation were prepared from the corresponding fluoride by metallothermic reduction techniques previously described by Daane.4 The chemical analyses of the metals used are listed in Table I.

Jan 1, 1965

By A. U. Seybolt

The solubility limit of oxygen in columbium has been determined in the range between 775' and 1100°C by means of lattice parameter measurements and microscopic examination. The solubility is a function of temperature and varies, in the range given above, from 0.25 to 1.0 pct O, respectively. BECAUSE of the marked deleterious effect of oxygen upon the mechanical properties of some of the transition metals, it is desirable to know something about the solubility of oxygen in these metals. The brittleness caused by oxygen in solution is particularly marked in the case of the group VA elements, vanadium, columbium, and tantalum. The solubility of oxygen in vanadium has already been reported in an earlier paper,' and Wasilewski2 has given a value (0.9 wt pct) for the solid solubility of oxygen in tantalum at 1050°C. Brauer3 in 1941 investigated the Cb-0 system up to Cb2O5, but made no real effort to investigate the extent of oxygen solubility in the metal. He made the observation, however, that this solubility must be less than 4.76 atom pct (0.86 wt pct) oxygen. This estimate was made from X-ray diffraction results on the alloys CbO, CbO, and CbO; all alloys consisted of the terminal (Cb) solid solution plus CbO, but the last alloy containing 4.76 atom pct 0 showed only three very weak CbO lines. It is surprising that Brauer, by examining only three alloys, arrived at an estimate of the solubility which agrees very well with the results to be reported herein. Experimental Procedure A columbium strip obtained from Fansteel Metallurgical Products was cut into strips, 0.020x1/2x2 in. Two holes, about 3/16 in. in diameter, were made near the ends of the strips in order to hold them against a flat steel block for mounting in a General Electric X-ray spectrometer for lattice parameter measurements. The same holes were used to hang the specimens inside a fused silica vacuum furnace tube which was part of a Sieverts' gas absorption apparatus. The apparatus and method of adding oxygen gas has been previously described.1 According to the supplier, the columbium obtained had the analysis given in Table I. After degreasing the samples, approximately 0.001 in. was etched from each side of the samples in order to remove possible surface impurities from the last rolling operation. For this purpose the following cold acid pickle was found satisfactory: 8 parts HNO3, 2 parts H2O2 and 1 part HF. Various Cb-O compositions were obtained up to 0.75 wt pct O by the gas absorption and diffusion technique. After the sample had absorbed all the oxygen gas added at 1000°C, an additional 24 hr was allowed for homogenization. This treatment appeared to be adequate, as shown by the linearity of the lattice parameter-composition plot. More concentrated alloys were prepared by arc melting mixtures of Cb and Cb2O5 since it was very time-consuming to make Cb-0 alloys in the neighborhood of 1 pct O, or over, by the diffusion method. When the flat strip specimens were used, they were ready for the X-ray spectrometer after cooling from the Sieverts' apparatus. The cooling rate obtained by merely allowing the hot fused silica furnace tube to radiate to the atmosphere (when the furnace was lowered) was sufficiently fast to keep the dissolved oxygen in solution. Arc-melted alloys were reduced to —200 mesh powder in a diamond mortar, wrapped in tantalum foil, sealed off in evacuated fused silica tubes, and then heat treated as indicated in Table 11. The fused silica tubes were quickly immersed in cold water without breaking the tubes after the heat treatments. The tantalum foil prevented reaction between the fused silica and the sample; there was no reaction between the powdered samples and the foil at 1000°C, but some trouble was experienced at 1100°C. At this temperature level a reaction between the sample and the foil was sometimes observed, which resulted in erroneous parameter values. Experimental Results Hardness Tests: Since most of the X-ray samples were in the form of flat strip, it was convenient to obtain Vickers hardness numbers as a function of oxygen content. Compared to the V-O case,' oxygen hardens columbium much more slowly, presumably because of the larger octahedral volume in colum-bium (about 12.0 compared to 9.3Å3 in vanadium), hence, requiring less lattice strain for solution. The plot of VHN vs wt pct O is shown in Fig. 1.

Jan 1, 1955

By B. L. Dunic, Terkel Rosenqvist

rr has long been suspected that sulphur has a small but finite solid solubility in iron, but up to the present more accurate data have been lacking. The survey given by Hansen' illustrates the divergence of data and opinions. Solubilities between 0.01 and 1.5 pct have been suggested, and opinions vary regarding the effects of temperature and phase transformations in iron on the sulphur solubility. The extent of solid solution and its variation with temperature might be expected to be an important factor in the occurrence of hot shortness in steel. Therefore, the object of the present investigation was to study this problem more systematically. It was furthermore hoped that solubility data for sulphur might contribute to an understanding of the factors which determine solid solubilities in general. Experimental Procedure The present work was based on the measurement of the chemical activity of sulphur in the solid solution by means of equilibrium with a mixture of hydrogen and hydrogen sulphide. When the solid solubility limit is reached, the sulphur activity in the iron becomes equal to that in the coexisting sulphide phase. A gas mixture of a controlled H,S/H, ratio was passed over pure metallic iron heated to a temperature in the range 900" to 1500". The amount of sulphur absorbed by the iron was determined by analysis, and from a plot of percentage of sulphur vs. gas composition the sulphur activity and the solubility limit were determined. As the rate of diffusion of sulphur in solid iron is very small, often several days were needed to be sure that equilibrium conditions had been established. To produce a gas flow with less than 1 pct H,S of constant composition for several days, was in itself a problem. The most satisfactory arrangement was obtained by a continuous synthesis of hydrogen sulphide from the elements. This was done in the part of the apparatus shown to the left in Fig. 1. The hydrogen used was first purified in tank batches by passage through activated charcoal cooled in liquid nitrogen under a pressure of 150

Jan 1, 1953

By F. M. Smith, R. H. Moore, J. R. Morrey

Electrodiffusion in y cerium reported by Henrie has been confirmed and a Preliminary estimate made of the relative rates of electrodiffusion of iron, cobalt, and nickel. These diffuse to the anode at rates decreasing in that order. In addition, copper and manganese exhibit slow, but detectable, diffusion to the anode and molybdenum exhibits detectable diffusion to the cathode. The electrodiffusion of carbon, .zirconium, antimony, magnesium , and silicon in y cerium could not be detected. Iron and cobalt diffuse in y cerium at rates proportional to the current density and with no apparent dependence on temperature. Decreasing polarization of iron and cobalt with increasing temperature, which cancels the expected rate increase, would account for this behavior. The electrodiffusion rate of iron in y uranium and in E plutonium has been measured. Diffusion of iron is anode-directed. Tin was found to diffuse to the cathode, in y uranium, at an appreciable rate. In all of these solvent metals, negative ions diffuse to the anode and positive ions to the cathode. The potential field effect appears to account satisfactorily for these results. FROM early experimental work summarized by Jost1 and Seith,2 the driving force for electrodiffusion was attributed to the potential field acting on ions in a metal. More recently, Heuman,3 Wever,4 and Huntington5 have shown that momentum interchange between conduction electrons and mobile entities in the metal contributes to electrodiffusion. Electron momentum interchange is anodically directed and the direction of diffusion resulting from the field force is dependent upon the charge on the diffusing entity. These two effects may either reinforce or oppose each other. Glinchuk6 has pointed out that momentum transfer in defect conductors should be cathode-directed and this appears to be the case as demonstrated by wever's4 work on iron. Barnett's7 work, on the other hand, indicates that, even in defect conductors, electrons show a negative E/m ratio when accelerated with respect to the lattice and should lead to anode-directed momentum transfer. In discussing this problem, Wever and seith8 suggest that defect electrons interact preferentially with activated ions so as to allow a net movement toward the cathode while still maintaining an electron momentum transfer in the anode direction. Williams and Huffine9 and Henriel0 have demonstrated that electrodiffusion may be useful for purification of yttrium and cerium. In yttrium, Williams and Huffine note that movement of several metallic impurities toward the anode is in keeping with observations in most other metallic systems and indicates that yttrium remains a normal electronic conductor at least to 1230°C. Close inspection of their data shows, however, that oxygen, nitrogen, and the transition elements diffused toward the anode, while nontransition elements diffused toward the cathode. This suggests that potential field effects may have been appreciable. The present work was concerned with the applicability of electrodiffusion as a technique for purification of plutonium, but, because of the obvious hazard inherent in work with this metal, experiments to develop the technique were carried out using cerium and uranium. The results of electrodiffusion measurements on these metals and on plutonium are reported here. EXPERIMENTAL The metal specimens prepared for this work were 6 in. long, 1/4 to 1/2 in. wide, and 1/16 to 3/32 in. thick. The uranium specimens were machined from a bar which analyzed 310 ppm Fe and the electrodiffusion of iron was followed by spectrographic and by chemical analysis. Cerium and plutonium specimens were cut from sheet rolled from ingots obtained from molten salt-metal equilibrations during which radioactive tracers were introduced. The electrodiffusion of the tracers was subsequently determined by counting methods. The specimens were electrolyzed between nickel electrodes containing resistance heaters used to equalize the specimen and electrode temperatures, thereby reducing thermal gradients. The temperature of the electrodes adjacent to the ends of the specimen was measured with chromel-alumel thermocouples which were connected to the heater controls. The surface temperature of the specimen at a point midway between the electrodes was measured with a sapphire rod pyrometer, the output of which controlled the dc power supply. This assembly was enclosed within an evacuable chamber containing a quartz viewing window. The temperature of the specimen over its entire length could be scanned with a portable pyrometer through

Jan 1, 1965

By E. A. Gulbransen, K. F. Andrew

Thermodynamic information on the solubility of hydrogen in exothermic metals is limited. Thus, the overall solubility decreased as the temperature rose, which suggests the heat of solution of hydrogen in the metal is negative; yet the terminal solubility in the metal phase increased, which suggests an endothermic reaction. A thermodynamic analysis, therefore, was made of the several equilibria involved when hydrogen dissolved in the metal. Equations were derived expressing the solubility, terminal solubility, and decomposition pressures of hydrogen in terms of the heat, entropy, and free energy of solution of hydrogen in the given phase. The relations between the thermodynamic functions for the a-zirconium and 8-hydride phase were developed. The thermodynamic quantities were determined experimentally by the measurement of the decomposition pressures over a broad composition and temperature range. Two new methods of analyzing the experimental data were developed for determination of the terminal solubility. STUDIES', ' of the impact strength of zirconium at room temperature indicated that small quantities of hydrogen, of the order of 10 parts per million, embrittles the metal. Micrographic' and X-ray diffraction' studies showed this embrittlement to result from the precipitation of a second phase upon slow cooling. These studies suggested that the terminal solubility of hydrogen in the metal phase increased with increase of temperature. This behavior appeared to contradict the effect of temperature on exothermic hydrogen occluders such as zirconium. For these metals, the solubility of hydrogen should decrease with increase of temperature. Before these facts could be understood, a thorough analysis of the several equilibria was essential. It was the purpose of this paper to consider the several equilibria involved when hydrogen was reacted with zirconium from both the theoretical and experimental points of view. Since the overall solubility of hydrogen in the metal has been considered by a number of workers, it was proposed to study the composition range 0 to 6 atomic pct H in detail. Literature Schwartz and Mallett' determined the terminal solubilities of hydrogen in the metallic phase from diffusion studies. They extrapolated concentration-penetration curves to the boundary between the solid solution phase and the hydride phase at the surface. Their results showed values of 2.6 atomic pct at 400°C and 5.2 atomic pct at 500°C. Extrapolations of the data to room temperature suggested very low values for the terminal solubilities. This study confirmed the hydrogen embrittlement results and indicated that the terminal solubilities increased as the temperature was raised. The solution of hydrogen in zirconium over the complete concentration range was studied by Hall, Martin, and Rees3 and by others. The results showed that hydrogen was absorbed up to a maximum composition of 66 atomic pct or ZrH This composition represented the t-hydride phase. The maximum amount of hydrogen absorbed increased with pressure and decreased with temperature. Due to this temperature behavior, zirconium was classed as an exothermic occluder." In 1930 Hägg studied the crystal structures of the Zr-H alloys over a broad composition range. Four hydride phases were found. However, the range of homogeneity of the several phases was not studied. Gulbransen and Andrew" recently studied the phase diagram of this system by X-ray diffraction methods and by a thermodynamic method based on decomposition pressure measurements. The results showed that only two hydride phases existed for low temperatures, the 6 and the .-phases of Hagg.' Both phases had broad regions of homogeneity. The lower limit of the two-phase region of a and d-phases suggested a terminal solubility of 4 atomic pct at 500°C. This value was in agreement with the results of Schwartz and Mallett.' To avoid hydrogen embrittlement at low temperatures, it was necessary to heat the metal in high vacuum at elevated temperatures. To choose the conditions of temperature and high vacuum, it was

Jan 1, 1956

By S. H. Moll, R. E. Ogilvie

The investigation of solid-state diffusion phenomena may lead to much information concerning binary alloys. In particular, a study of the concentration gradients present in multiphase diffusion couples can lead to the determination of the solubility limits of the single-phase fields in the phase diagram. The specific concentration gradient which results from the heat treatment of a diffusion couple depends not only upon the time and temperature of diffusion, but also upon the number of phases existing in equilibrium at the diffusion temperature. Such concentration gradients possess sharp discontinuities whose terminal points correspond to the solubilities existing at the limits of the two-phase field in the phase diagram at the diffusion temperature. In general, there will be a discontinuity in the concentration curve for each two-phase field which exists in the phase diagram between the concentration limits of the couple. A number of investigators have analyzed the gradients present in multiphase diffusion couples. 1-8 The iron-titanium phase diagram exhibits a y loop in the temperature range from 900" to 1400°C, and the limit of the a + y field lies at some composition which is less than 1 wt pct Ti for any temperature. It was felt that a diffusion study could yield values not only for the diffusion coefficients involved, but also for the chemical solubilities in the phase diagram. The concentration gradients were analyzed by an X-ray absorption technique developed by Ogilvie.6 The absorption of a monochromatic X-ray beam is measured in steps parallel to the diffusion direction and the resulting intensity gradients are transformed into concentration gradients by applying the laws of X-ray absorption in a binary system. This technique has been applied by ogilvie6 to the systems Ni-Au, Cu-Au, Fe-Cr, and Ni-Cr; by Gelles7 to the systems Be-Fe, and Be-Ni and by Hilliarde to the system Al-Zn. Linear X-Ray Absorption Analysis—The use of X-ray absorption analysis as a tool for determining concentration gradients arises from the fact that the beam is absorbed according to the quantity and species of elements present in the sample and not according to their state of aggregation. This method is more advantageous than the chemical analysis of machined layers since it is less time consuming and less expensive and can analyze a very small area. IF a homogeneous binary alloy (A + B) is analyzed with two different monochromatic X-ray beams I0(2) and lo(2) the following relation between the transmitted intensities, the intrinsic absorption coefficients of A and B for the two different radiations and the weight fraction of each element may be derived.' In (I/Io) (a+b) xa(u/p)a(2) + xb(u/p)8(2) Eq. [1] is independent of specimen thickness and density, but requires that either element A or B possess an absorption edge between the wavelength limits of the two monochromatic beams. I, is evaluated for each radiation by measuring the transmitted intensity through a varying number of foils of a suitable absorber. A plot of the natural logarithm of the transmitted intensity as a function of the number of foils, when extrapolated to zero foils, yields the value of I,. If a plot of the ratio in Eq. [ I] is calculated as a function of xA, then a master curve results. From

Jan 1, 1960