Search Documents

By L. R. Weisberg, E. A. Miller

Single-crystalline films of silicon of good quality were vacuum-deposited on sapphire. The improved crystallinity was achieved by the strict exclusion of oxygen from the evaporation system, including the use of a "oxygen-free" silicon source. To minimize possible contamination of the source, the silicon was sublimed from the solid at about 1390° C. using a simple electron bombardment system to heat the silicon. This resulted in deposition rates of about 5 to 7 µ per hr and film thicknesses of 4 to 20 µ. The vacuum during deposition was ordinarily about 5x 10-8 Torr, DESPITE the success achieved in the chemical vapor deposition of silicon on sapphire using Silane,1-4 vacuum deposition has not provided comparable properties. For example, Chu et a1.5 vacuum-deposited silicon on single-crystalline substrates of A12O3, Sic, ThO2, and other compounds, and, although some epi- but the use of ultrahigh vacuum (about 1 to 3 x10-9 Torr) during deposition did not significantly improve the results in the oxygen-free system. The crystalline quality of films deposited on (0001) sapphire was found to be considerably superior to (1102) sapphire surfaces. Using (0001) sapphire, single-crystalline films were deposited at temperatures as low as 780° C and as high as 1075°C, with the best crystalline quality obtained at about 1000°C. At this temperature, hole mobilities over 250 sq cm per v-sec were achieved. taxy was indicated, single-crystalline films were not produced. More recently, Salama, Tucker, and young6 evaporated silicon on single-crystalline sapphire, but only poor crystallinity was observed, and low carrier mobilities were measured. In the present experiments, good-quality single-crystalline films of silicon on sapphire have been achieved by specifically excluding oxygen from the evaporation. This included not only the use of a "clean" vacuum system but, of special importance, the use of an oxygen-free silicon source. This has also been shown to provide a dramatic improvement in silicon homoepi-taxy.7 As a result, silicon-on-sapphire films with hole mobilities over 250 sq cm per v-set were achieved, and single-crystal films were grown as low as 780°C and as

Jan 1, 1969

By Richard O. Carlson

Ions of p31 and B 11 were implanted in natural insulating diamond macles. The thin (-0.4µ) layers showed sheet resistances of 107 to 1011 ohm per sq and activation energies of 0.17 to 0.34 ev above room temperature. However, no Hall effect could be measured, indicating that mobilities were less than I to 10 sq cm per (v-sec). Such low mobilities may be due to excessive scattering due to radiation damage in the bombarded layer or to very high concentrations of active compensating impurities. Annealing did not materially change the room -temperature properties, but the temperature was limited to -800°C by surface graphitization, the latter process probably being accelerated by lattice damage in our diamonds arising from the ion bombardment. A weak n-type thermoelectric power was detected after p31 irradiation, but it cannot be presumed that we have made the phosphorus ions active as donors in diamond. The natural diamonds have a high and uncontrolled concentration of impurities and, when coupled with the radiation damage and graphitization problems, would appear to seriously limit the quality of semiconductor that we can presently achieve by ion implantation in diamond. ALTHOUGH diamonds are usually thought of as insulating in terms of their electrical conductivity, it was found about 15 years ago that natural semiconducting diamonds do occur rarely, and these were designated as type IIb diamonds. These p-type semiconducting diamonds were found to be dominated by an impurity level 0.37 ev from the valence band. Evidence today based on correlation of the concentration of acceptor states from Hall effect measurements with the impurity concentrations determined by neutron activation analysis point to aluminum as the dominant acceptor impurity.' The compensating donor is believed to be nitrogen, which has a donor level 1.6 ev above the valence band.' However, only a small fraction of the total nitrogen content in the diamond is electrically active. An infrared absorption band at 7.8 p and ultraviolet absorption near 4 ev have been associated with nitrogen, the former providing a quantitative measure of nitrogen content.3 The nitrogen content is -l020 cm-3 in insulating or type I diamonds, but is less than 10'' in natural semiconducting diamonds. Much of this nitrogen is distributed in platelets oriented in (100) planes and not atomically dispersed in the diamond lattice.4 Semiconducting diamonds have been deliberately formed by incorporating impurities into the graphite charge in the high-pressure apparatus used to form diamonds.5 These manufactured semiconducting diamonds are always p-type, and while some show the 0.37 ev level due to aluminum, most such samples have been ascribed to impurity banding effects.6 The best natural or manufactured semiconducting diamonds have hole mobilities near 1500 sq cm per (v-sec) at room temperature, and those with the 0.37-ev level can be analyzed to reveal an acceptor concentration of 3 to 8 X 1016 cm-3 and a donor concentration 3 to 10 times lower.' NO bulk n-type diamonds have ever been reported, but the electron mobility has been measured as 1900 sq cm per (v-sec) by irradiating a diamond with ultraviolet to excite electrons out of the deep nitrogen donor level or other levels.7 New hope for the formation of n-type diamond has emerged from the ion-implantation method whereby desired impurities are introduced into a crystal lattice by bombardment with a high-energy beam of the impurity ion. It was found in the case of silicon that the usual donors such as phosphorus and arsenic and acceptors such as boron and gallium could be implanted into silicon. Wentorf and arrow' produced semiconducting layers on diamonds by an ionic bombardment in a glow discharge at potentials of about 2 kv. The observed typeness from thermoelectric probing seemed to depend on the atmosphere gas (nitrogen, argon, or hydrogen) rather than on the electrode material, but the nature of the conduction process in the thin damaged surface layer is completely unknown. A Russian group under Vavilov has attempted high-energy ion implantation in natural diamonds using boron and lithium ions9 and later also phosphorus, aluminum, and carbon ions.10 Their papers claim n-type layers from lithium, phosphorus, and carbon implant and a p-type layer from boron and aluminum implant, though the methods of type determination are not described in detail. Under government contract support, the Ion Physics Corp. has studied ion implantation in several semiconductors. While the bulk of their study1' was devoted to the irradiation of silicon, they did carry out a short study on boron and phosphorus implantation into natural diamonds. They did observe surface conducting layers but did not determine the typeness of the layers. For our experiments, phosphorus and boron were chosen as the dopant ions because of their respective donor and acceptor behavior in germanium and silicon. Moreover, they are the lightest mass dopants of the shallow level donors and acceptors from columns 111 and V of the Periodic Table (excluding nitrogen which is already present in diamond), and will have the largest depth penetration into the diamond lattice. EXPERIMENTAL Diamonds. As the target diamonds for our ion implantation study, we chose commercially available macles, which are flat, twinned diamond crystals. By scanning through the stock of a wholesale distributor

Jan 1, 1970

By Martin Rubenstein

CdS crystals have been grown from a number of metallic solvents such as bismuth, tin, lead, and cadmium. Etching studies have shown that plastic deformation occurs if the crystals are not removed from the solvent prior to the solidification of the solvent, on cooling. The deformed crystals show a umique exciton fluorescence as a function of edge dislocation density. If one grows the CdS in the eutectic alloy of the above four metals (commonly called Wood's metal) the crystals can be removed from the solvent with hot water and no plastic deformation occurs. In this paper, the liquidus solubility measurements of CdS, as a function of temperature, are presented. The data were obtained using a high -temperature filtration technique. CADMIUM-SULFIDE crystals have been grown from a number of metallic solvents1 such as cadmium, bismuth, tin, and lead. Liquidus solubilities of CdS in cadmium,2 bismuth,3 and tin4 have already been measured. Crystals of CdS, in all four metals, have been grown by solution growth: 1) by cooling a saturated solution and 2) by a solution transport method.1'"1 CdS crystals grown in these four solvents have a few characteristics in common: 1) 1.8°K photolumines-cent emission consisted mainly of the radiative recombination of the bound exciton commonly known as I,, 2) slip lines which could easily be seen by the naked eye, and 3) edge dislocation densities in the order of l05 per cu cm.1 It was decided that these slip lines and the high edge dislocation densities were caused by a plastic deformation of the CdS crystals. It was felt that this plastic deformation did not occur during the growth of the crystals nor during the cooling of the solution, but did occur when the solvent which was in contact with the crystals froze. If these assumptions were valid, the slip lines and the high number of dislocations could be reduced or eliminated by removing the crystals from the solvent before the solvent froze. Since crystals of CdS had already been grown separately in such solvents as bismuth, lead, tin, and cadmium, it was felt that crystals could be grown in a eutectic mixture of these four metals. In this work a eutectic (or near eutectic) mixture of bismuth, lead, tin and cadmium in the proportion 50, 26.5, 13.5, and 10 wt pct, respectively, was used to grow CdS crystals. Such a mixture has a melting point of about 70°C and is close in composition to the alloy commonly known as Wood's metals. If the crystals could be grown from this mixture of solvents, and if hot water (>75°C) could be used to separate the crystals of CdS from the metallic solvent, it was hoped that CdS crystals could be grown with little or no plastic deformation which had been ob- served when crystals were grown from these solvents uncombined. CdS crystals were grown from this low melting eutectic mixture of bismuth, lead, tin, and cadmium using the solvent transport method. CdS powder and the appropriate amount of metals were sealed in a quartz tube under a pressure of about 5 X 10-6 torr. This ampule was then placed in a vertical position in a furnace. The temperature was raised to about 900°C. The furnace was designed so that the top of the liquid column within the ampule was between 10° to 40°C higher than the bottom of the liquid column. These temperatures were measured on the outside of the quartz ampule. The ampule was maintained at temperature for 7 to 14 days (depending on the temperature at which transport was taking place) and then the furnace temperature was lowered until the temperature was about 125°C. The ampule was then removed from the furnace, placed in water maintained at about 90°C, and opened in this 90°C environment. The crystals could then be removed from this two-phase liquid (Wood's metal and water) by mechanically picking them out. Alternatively, the crystals could be quantitatively removed by adding an excess of mercury to the mixture of metals, crystals, and hot water. The hot solution of metals and the hot water could be evacuated using a small diameter tube connected to a vacuum. Small amounts of mercury and water could be removed by heating the crystals in vacuum. Crystals prepared using this technique showed no evidence of slip. However, some of these crystals did show edge dislocation densities as high as l04 per cu cm. Some few selected crystals showed no dislocations. Single crystals of CdS were grown as large as 5 by 5 by 0.5 mm. The ampules for the growth of these crystals were 13 mm O.D., 11 mm I.D., 150 mm! LIQUIDUS SOLUBILITY MEASUREMENTS The CdS starting materials was G.E. 118-8-2 powder which was fired in H2S at 1000°C, and then a vapor transport technique5 was applied to produce a "sound" mass of CdS. The Wood's metal was prepared by weighing out bismuth, lead, tin, and cadmium in the proportions of 50, 26.5, 13.5, and 10 wt pct, respectively. The bismuth, cadmium, and lead were from the American Smelting and Refining Co. (ASARCO) and all had purities of 99.999+ pct. The tin was 99.9999 pct spectroscopic grade from the Vulcan Materials Co. The appropriate mixture was placed in a quartz tube, evacuated to a pressure of 5 X 10-6 torr, melted to a liquid, cooled to room temperature under this same vacuum. This ingot was then placed in another quartz tube, evacuated to 5 x l0-6 torr, and sealed off under vacuum. The ampule was then horizontally placed in a furnace. The temperature was raised to 600°C, and over a period of several hours the ampule was vigorously shaken several times. The ampule was then removed from the furnace, and the metallic liquid was

Jan 1, 1970

By B. D. Lichter

In a previous paper, Oelsen, Schuermann, and Hey-nertl pointed out the possibility of obtaining complete thermodynamic functions for alloy systems from calorimetric measurements alone. Specifically, if the heat content of a heterogeneous, multicomponent system is known as a function of temperature and composition, then it is possible to derive the heats, entropies, and free energies of formation of liquid phases and of solid intermediate phases of fixed composition. The authors present equations and graphical techniques applicable to eutectic systems and systems with intermediate phases without solid solubility at lower temperatures. eijering has considered possible extension of the treatment to systems with complete solid-state miscibility at lower temperatures; however, experimental difficulties practically restrict the method to systems with very limited miscibility in the solid state. The purpose of this note is to present an alternative derivation of the heats and entropies of formation of liquid phases and of solid intermediate phases from heat-content measurements alone. Fig. 1 shows three types of phase diagram to which the analysis may be applied. If the heat content is determined as a function of composition x, at temperature T, we have quite generally and the heat of mixing of the liquid is obtained directly from heat-content measurements. If the temperature T lies in a two-phase region (solid-liquid) then

Jan 1, 1967

By N. A. Gjostein

From their study of the anisotropy of grain boundary mobility in aluminum, the authors conclude that tilt boundaries have a higher mobility than twist boundaries because the atomic misfit at the pure-twist boundary is relatively small compared to the misfit at a tilt boundary for a given amount of misorientation. If this is true, then one would expect a twist boundary to have a lower grain boundary energy than the corresponding tilt boundary. Fig. 6, taken from the work of Gjostein and Rhines,10 shows that this is indeed the case for [001]-cu tilt and twist boundaries in the high-angle region, thus lending support to the authors' conclusion.

Jan 1, 1968

By A. Hansson, G. E. Tardiff

WHILE it is generally known that cold-worked low-and medium-carbon steels exhibit substantial increases in tensile rupture ductility with increased deformation velocity172 (up to the von Karman limit), the observations for annealed or normalized steel do not permit simple generalization. Various investigators, for example, have reported both small increases1,3 and small decreases4,5' in rupture ductility with increased rates of deformation of normalized low-carbon steel. The purpose of the present note is to report on the dependence observed for strain-aged steel since preliminary work here indicates that strain aging results in the development of a consistent type of behavior that is understandable in the light of current knowledge of the lower yield dynamics for steels which deform inhomogeneously.6 This realization throws additional light on the influences responsible for the behavior of normalized steel. The results presented here were obtained for a 0.21 pct C, 0.012 pct N sheet (0.012 in. thick) steel which was received in a temper-rolled condition (approximately 3 pct reduction in thickness after process annealing); the as-received material had a spherodized structure with a ferritic grain size of ASTM 8. The material was selected for testing after 6 months room temperature aging, at which time the static lower yield and ultimate tensile strength measured 79 and 82 kpsi, respectively. ASTM 2-in. reduced section, 1/2-in.-wide specimens were used in smooth bar tests and 3/4-in.-wide specimens containing two 60-deg opposed edge notches (root radii = 0.002 in.) were used in notched bar tests. To facilitate strain measurement, the smooth bars were blued with ink and marked at 0.1-in. intervals along the gage length; the notched

Jan 1, 1968

By A. S. Argon, G. East

From their study of the anisotropy of grain boundary mobility in aluminum, the authors conclude that tilt boundaries have a higher mobility than twist boundaries because the atomic misfit at the pure-twist boundary is relatively small compared to the misfit at a tilt boundary for a given amount of misorientation. If this is true, then one would expect a twist boundary to have a lower grain boundary energy than the corresponding tilt boundary. Fig. 6, taken from the work of Gjostein and Rhines,10 shows that this is indeed the case for [001]-cu tilt and twist boundaries in the high-angle region, thus lending support to the authors' conclusion.

Jan 1, 1968

By R. J. Wasilewski

Elastic constants have been determined on single crystals of maximum-melting-temperature NiAl compound (50.6 at. pct Al) at 25°C. Temperature variations of Young&apos;s modulus in the three principal crystal directions between -150°C and +800°C exhibit no anomalies similar to those repovted for the isomor-phous compounds brass and AuCd. It is concluded that the ordered CsCl structure is stable throughout the temperuture range investigated NICKEL aluminide is a congruently melting inter-metallic compound of CsCl (B2) structure, stable over a rather wide composition range (40 to 54 pct Al). It exhibits a lattice-parameter maximum of a. = 2.883A at 50.6 pct 1,&apos; at which composition the excess aluminum atoms are randomly distributed on the vacant nickel sites;&apos; at higher nickel compositions the excess nickel atoms replace substitutionally in aluminum sites, while at higher aluminum compositions vacancies are present in the nickel sublattice.&apos; For compositions close to NiAl no significant variation in order has been observed at temperatures up to 900°C.3 At high nickel (65 pct) compositions the existence of a martensitic structural transformation has been reported.4 NiAl is isomorphous with the ordered B-brass structure and with the high-temperature structures of AuCd and TiNi alloys close to equiatomic composition. All of these exhibit anomalous elastic behavior as the temperature of their diffusionless transformation to lower-symmetry structures is approached.&apos;-&apos; This behavior has been explained as indicative of low stability of the CsCl structure in these compounds.8 The present paper reports the experimental determination of the elastic compliances of single-crystal NiA1. EXPERIMENTAL METHOD 1) Specimens. Polycrystalline NiAl was prepared by induction melting of carbonyl nickel and high-purity aluminum and casting into solid copper molds in gettered helium atmosphere. The castings thus produced were canned in mild steel and impact-extruded at 1100 to 1200°C to 38-in.diam rod. After the steel sheath was pickled off, the rods were converted to single crystals by repeated high-frequency satisfactorily produced the desired orientations in the form of crystals 38 in. diam by 8-12 in. long. These were subsequently center less-ground to 7 to 8 mm diam by 12 cm long, electropolished (in 80 pct methanol, 20 pct HzSO, at 25°C and 15 v), and their axial orientation determined to 1 deg by the Laue back-reflection technique. The rod dimensions were determined to 0.01 mm and the densities determined from geometrical dimensions and the weight of every specimen. The average density was 5.905 0.005 g per cu cm. 2) Elastic Data Measurements. Room-temperature elastic compliances Sii were obtained by determining to 0.1 cps the fundameGtal resonance frequencies in the longitudinal and torsional vibration modes. At these frequencies the following relations hold: E =p(2lfif =pvl= Sn- 2r[Sll- S12- -£„«] g -P(2iftf = pvf = (s<« - fr(sii - sw - ±s»y where vl and vt are the velocities, and f andft the resonance frequencies of the longitudinal and torsional waves, respectively; 1 is the length of the specimen, and r is the direction cosine factor—{cos2 a • cos2 0 + cos2 a . cos2 y + cos2 0 ¦ cos2 Y}. Given three crystals of orientations (loo), (110), and (111)) i.e., F = 0, 14, and 13, we can obtain six equations for the three unknown compliances, thus providing adequate internal check. Mode mixing can readily occur in the relatively long wavelength resonance utilized here. This could cause a significant error in the torsion wave velocities calculated from the apparent resonance frequenies. Therefore, throughout the present work, a method, described previosl,&apos; was used to ensure that no significant mode mixing took place at resonance. The apparatus used was originally designed for the Elastomat, Magnaflux Corp.. Chicago, 111. determination of flexural vibration resonance frequencies. At room temperature torsional vibration mode could be established by suitable positioning of the thin wires from the input and pick-up transducers, and longitudinal mode by cementing thin magnetic foils at the ends of the specimen, and inducing the vibration by means of electromagnetic transducers. Only flex-ural resonance frequencies could, however, be obtained reproducibly over the temperature range of -150 to 800°C. These data are considerably less accurate, the error in the modulus being estimated at -0.6 pct from the specimen dimensions and the room-temperature Poisson&apos;s ratio. The reproducibility of the resonance frequencies in repeat runs was, however, to better than l in 15,000, hence the relative elastic modulus error less than 0.02 pct. Therefore, the temperature variation of the principal Young&apos;s moduli obtained by flexural resonance, relative to room-temperature values, is believed to be satisfactorily accurate.

Jan 1, 1967

By Thomas L. Garrard, James A. Sprague, Rex B. McLellan, Samuel J. Horowitz

A simple rnodel for interstitial solid solutions has been devised in which each solute atom interacts with the solzlent lattice in such a way as to exclude an integral number of nearest-neighbor sites from being occupied by solute atoms. The vemaining interstitial sites ave not affected. In calculating the entropy of such a solution the increase in the number of interstitial sites available to solute atoms due to the ozlerlap of blocked sites is taken into account. The ther-modynamic functions and the solute activity fov the solution rnodel are calculated. It is shown that considevation of the overlap sues vise to a correction term of the order 8&apos; ( where 0 is the ratio of solute to solvent atoms) in the thevmodynamic functions of the solution and the solute activity. The theo,uetical equation for the solute activity was used to analyze the available data for the activity of carbon in austenite at three different temperatures. The use of a computer enabled many comparisons to be made between the experimental activity data and those predicted by both the complex blocking model and the simple, widely used blocking model in which the ozierlapping of blocked sites is not taken into consideration. Even though at the highest concentvations oj carbon the value of 0 is small it is shown that the best fit to the experimental data is found with the complex blocking model. RECENTLY it has been shown by analysis of gas-solid equilibriums that many interstitial solid solutions can be described by the quasi-regular solution model in which the partial energy of solution E, is independent of composition and, apart from a composition-independent term due to nonconfigurational factors, the partial entropy is that of an ideal solution. Dilute solutions of nitrogen in both bcc and fcc iron,&apos; dilute solutions of oxygen and silver,&apos; carbon in a Fe,&apos; and dilute solutions of carbon in ? Fe2 are all quasi-regular. In addition many hydrogen-metal solutions, both dilute and concentrated (with the notable exceptions of the H-Ni and H-Pd systems), can be described by this model.3 However, it has been shown that, at various temperatures, solutions of carbon in ? Fe begin to depart strongly from quasi-regular behavior at low concentrations. A plot of the activity of carbon in ? Fe relative to pure graphite vs atom fraction of carbon (Cc) calculated from Smith&apos;s data4 begins to depart from the (nonlinear) plot for a quasi-regular solution at about Cc = 0.02. Several statistical models have been proposed to account for the thermodynamic behavior of carbon in y Fe. Darken and smith4 set up a model in which a carbon atom in its octahedral site in the fcc lattice has either one or no neighboring carbon atoms in the twelve nearest-neighbor octahedral sites. It is assumed that the number of carbon atoms having more than one nearest neighbor is negligibly small and that a solute atom does not interact with another solute atom located at a greater distance than the first shell of interstitial sites. Thus this model contains two characteristic interaction energies. Darken and Smith showed that this model gives reasonable agreement with the measured variation of carbon activity ac with composition. It is noteworthy that Darken and Smith concluded from their analysis of the activity data using this model that there was a repulsive force between carbon atoms in austenite which reduces the concentration of C-C pairs below that corresponding to random mixing. Recently Aaronson, Domian, and pound5 have considered in detail the statistical model developed by Lachere and by Fowler and Guggenheim7 and have shown that it is compatible with the activity data for carbon in austenite. In Darken&apos;s model the energy of solution of carbon in austenite is dependent on temperature and concentration; the departure from Henry&apos;s law arises from the concentration dependence of the energy of solution. In Aaronson, Domian, and Pound&apos;s discussion of the Lacher, Fowler, Guggenheim model the interaction energy wy between adjacent carbon atoms in austenite is strongly temperature-dependent. Aaronson et al . suggested that this temperature dependence of wy and hence the energy of solution of carbon could be explained either by the formation of clusters of carbon atoms (despite the repulsive nature of wy) or by the dual occupancy of the octahedral and tetrahedral sites by the carbon atoms. However, a temperature-dependent heat of solution

Jan 1, 1968

By M. F. Burr, F. V. Lenel, G. S. Ansell

The mechanism of materials transport was studied in a model experiment in which a silver sphere was sintered to a silver plate under an applied external stress near 800° C. For this purpose the rate of growth of the neck between the sphere and the plate which is proportional to the strain rate of the creep process involved in a neck growth was determined by measuring the change in electrical resistance of the away. This rate of neck growth will depend upon the magnitude of the applied external stress. By measuring the rate just before and just after a change in the external load and therefore the applied stress, the relationship between strain rate and stress can be derived. For material transport by slip in a pure metal it is expected that the strain rate should be proportional to the 4.5th power of the stress, while for transport by diffusional flow the strain rate should be proportional to the first power of the stress. It was found that for stresses above 5 X 10&apos; dynes per sq cm (70 psi) the exponent in the stvain rate-stress relationship was near 4.5 while for a stress of 0.7 x 108 dynes per sq cm (10 psi) an exponent of 1.8 was determined indicating that under the conditions of the experiment the transport mechanism for creep in silver shifts from diffusional flow to slip in the region of 0.7 x 106 dynes per sq cm (10 psi). The importance of these findings for an understanding of the mechanism of material transport in conventional sintering (without an external load) is pointed out. In a paper by Early et al.,&apos; the mechanism of material transport during sintering was treated as a creep process. It was pointed out that two possible types of creep processes could operate during sintering. In one of these material transport takes place by vacancy diffusion due to a gradient in chemical potential. The vacancy sinks in this diffusion process may be either grain boundaries, as in Herring-Nabarro microcreep, or regions beneath a surface of small radius of curvature. The diffusive transport may be by volume, grain boundary, or surface diffusion. In the other mechanism, creep takes place by slip, i.e., dislocation motion with its rate generally controlled by dislocation climb. One method of distinguishing between these two modes of materials transport is to investigate the creep rate, i.e., the rate of material transport as a function of stress. For material transport by vacancy diffusion, the creep rate is proportional to the stress.&apos; For creep by slip, controlled by dislocation climb, the creep rate is proportional to a power of the stress. For pure metals this power is 4.5;374 for dispersion-strengthened material a fourth power relationship can be derived.5 This difference in the stress dependence of the strain rate is illustrated schematically in Fig. 1, in which the logarithm of strain rate is plotted vs the logarithm of the stress for material transport in a pure metal by vacancy flux and by dislocation climb. The stress range in which the two types of material transport overlap would be expected to be relatively small. In the experiments described by Early and coworkers compacts pressed from spherical copper powder were sintered isothermally under an applied axial external load in a dilatometer. During sintering the magnitude of the applied external load was changed and the rate of axial shrinkage of the compact before and after the load change determined. In order to find the stress dependence of the strain rate the magnitude of the stress acting on the neck area between the spherical particles before and after the load change had to be calculated. This stress is the resultant of the stresses due both to the externally applied load and to surface tension. The stress due to surface tension is a function of the geometry of the powder compacts. It can be determined, at least for simple geometrical arrays,&apos;76 but it is difficult to obtain exact values of the stresses for actual powder compacts when only average values of particle diameters and of neck diameters between particles are available. The calculation of stress and the strain rate-stress relationship is therefore only very approximate.

Jan 1, 1968

By Larry Kaufman, Joan B. Berkowitz-Mattuck, Edward V. Claugherty

The oxidation characteristics of hafnium and zirconiunr diboride were measured between 1200 and 2200&apos;K by a thermal- conductivity method which continuously ttzeasures the rate of reaction of oxygen with the diboride and by a metallographic air oridation method zuhich provides a measure of the total arr7ount of bovide conuerted to oxide for a given time interval. The oxidized specimens obtained from the tl~eritaal-coi~ductitrity method were also examined by quantitatire metal-lographic procedares. The significant results obtained in this investigation reveal that metal-rich compositions of lzafi~iutil diboride proride the most oxidation-resistatzt material up to 2000°K; hafnium diboride is tmre oxidation- resistant titan zirconium diboride at all tempevatures examined; the morphology of the oxide formed on H/B2 and Llie temperature coefficient qi. the oxidation rate constants change at the temperature of ttze monoclinic to tetragonal phase transition] in HfO2; the oxidation of neither HfB2 nor ZrB2, results iN catastrophic Jazlure at lorc. oxygen pressures; and pvefevetztial gvaLti boundary oxidation was not obsevued for either HfBi or ZYB, A comprehensive study of the high-temperature characteristics of refractory transition-metal di-borides is currently in progress. This program has included investigation of the physical, thermal, and thermodynamic properties of TiB2, ZrB2, HfB2, NbB2, and TaB2. In addition, aspects of the synthesis and fabrication of such materials have been studied. In view of the diverse nature of this research, a number of other laboratories have actively participated and contributed specific capabilities for analysis and characterization of these materials. As a consequence, an extensive description of the relevant properties of these compounds has emerged which is central in evaluating their high-temperature (1200" to 2500°K) performance. To date, information on thermodynamic stability, specific heat, and vaporization characteristics,1 hot hardness and electrical resistivity,1, 3 therma1 expansion:&apos;4 and thermal conductivity 1, 5 has been presented. This information has been generated on materials of the highest purity (98.5 to 99.9 wt pct Me + B) and density currently available. Samples fabricated by zone melting6 and high-pressure hot pressing"3&apos;7 techniques have been used to generate suitable specimens for all of the aforementioned studies. dation characteristics of the most oxidation-resistant of these materials, hafnium and zirconium diboride, is presented and a description of the synthesis and the experimental procedures used to prepare and characterize specimens is given. The high-temperature range under consideration (1200" to 2200°K) and the known dependence of oxidation characteristics on sample chemistry, density, and oxidation conditions required a close coupling of the synthesis, fabrication, and evaluation procedures.8 This was accomplished by continual surveillance of chemical composition of starting materials before and after specimen fabrication and by evaluation of density, phase constitution, and microstructural features prior to and after oxidation exposure. I) PROCUREMENT AND CHARACTERIZATION OF STARTING MATERIALS In view of the current state of the art in fabricating refractory boride materials, the methods used in preparing samples for the present study are given in detail as follows: starting materials were purchased in high-purity powder form and fabricated by high-pressure hot pressing into 0.40 by 1.00 in. bars from which oxidation specimens were obtained. The hafnium diboride used in this study was purchased from Wah Chang Corp.; the zirconium diboride from U.S. Borax and Chemical Co. These powders were routinely characterized by quantitative chemical analyses for metal, boron, carbon, oxygen, nitrogen, and iron, by qualitative emission-spectrographic analysis for trace impurities, by X-ray procedures for extraneous phase identification, and by powder densitometry for comparison with X-ray (theoretical) density. Hafnium and zirconium metal and elemental boron were also purchased as high-purity powders and characterized for impurities by emission-spectrographic analyses. The hafnium diboride was procured in three shipments which were designated as HfBl.g7(1), HfB1.88(2A), and HfB2.12(2). The indicated stoichiometry is based on the atomic ratio of total boron to total hafnium; the number in parentheses identifies the shipment number. Shipment 1 was 5 1b, shipment 2A, 1 1b, and shipment 2, 8 1b. The zirconium diboride was procured as a 20-1b shipment and designated as ZrB1.89(1). A small quantity of purified zirconium diboride was also supplied and designated ZrB1, 9(P). The averaged results for chemical analyses which were generally performed according to the procedures set forth in the compilation by KrieGe9 are presented in Table I. Qualitative spectrographic analyses indicated that Ca, Cr, Ti, Si, Zr (in H~B~), and A1 were present at levels between 0.01 and 0.10 wt pct. Other metallic elements were found to be less than 0.01 wt pct. Since it is virtually impossible to purchase these materials in the desired quantities (5 to 20 lb) as single-phase compounds it is necessary to obtain

Jan 1, 1968

By I. R. Kramer

The stress associated with the surface layer was deter-minedfor iron and molybdcnum. These measurements show that the surface layer plays a very important role in the plastic deformation of bcc metals. From the relationship between strain arui the surface layer stress and the back stress imposed by the internal obstacles, it is Proposed that the term of of the Hall-Petch equation is related only THE large number of studies on plastic flow of bcc metals have failed to delineate clearly the nature of the obstacles influencing dislocation movement. Since there have been numerous articles which have reviewed the mechanisms proposed to explain various aspects of plastic flow of bcc metals (for example, conrad1 and Keh and weissmann2), no attempt will be made here to discuss their conclusions. However, it should be to the surface stress while the term Kfd is the stress required to move dislocations through internal obstacles. Further, it was possible to derive the well-known empirical equation C = , and it is shown that the terprz a is comprised of two terms: a term Cs associated with the surface layer stress and Ci related to internal dislocation obstacles. pointed out that in all of these publications it has been tacitly assumed that plastic flow occurred uniformly throughout the cross section of the specimen (except, of course, for the Luders band region), It has generally been further assumed that the nature and distribution of the obstacles opposing the motion of dislocations are of the same type throughout. The role of the surface layer in bcc metals has not, as yet, been considered even though its role in fcc metals has been shown clearly to be of primary importance.3-5 Investigations on fcc metals have shown that at the onset of plastic deformation a surface layer acts to

Jan 1, 1968

By J. H. Bucher

The effect oj strengthening by the precipitation of CbC on the toughness offerrilic steel (0.11 pct C, 0.74 pct Mn, 0.02 pct Cb) was stlrdierl. A g-veater degree of pvecij~itatiotz strengthning is ohtaznable by aging the so11ution-treated steel rtt velaticely Low temperature was (1050" to 1l0O3F) than at higher temperatures (I150" to 1200°F). Tlzis ),eslrlts because a finer dispersion of CbC is obtained. However, the ductile-brittle transition te~ripertltlo.e is raised approximately tlzjeee tirues as IFLIIC~ perv a given yield stress increase) by aging in the lower tempareture range, as opposer1 to the Irigher temperaturee range. This is proposed to result pl-iir~avily because different precipitation distributions ar-e associated with the different aging -treaments. For tile steel aged at 1050" and 1100 F, heavy: precipitation on disloccctions occurs resulting in a high degree of rlislocution pinningh tl~us lowevirzg- the toilguiess. in tll c(.rise of the steel aged at 1150 &apos;a?zd 1300"F, Pvecipitatiol~ is irzove g-enerally distributed throughout the matrix rather limn preferentially on dislocations.; and Ike detlsitj, oj mobile dislocations is not sttyong-ly affected. THE most effective means of strengthening hot-rolled ferritic steels are grain refinement and precipitation hardening. It is well-known that grain refinement increases both the strength and the toughness. However, the upper limit of yield strength obtainable by grain refinement alone is about 60,000 psi,&apos;12 i.e., without the formation of acicular structures of lower toughness. The most economical means of obtaining higher strengths is the addition of small amounts of elements such as columbium, vanadium, and titanium which form carbonitrides. These precipitates are very effective in increasing the strength, but the ductile-brittle transition temperature is is Although it is agreed upon that precipitation decreases the toughness, there has been little effort to isolate the causative aspects of embrittlement. The purposes of this research are to assess quantitatively the embrittlement associated with different precipitate distributions and morphologies, and to correlate the degree of the embrittlement with the microstruc-ture. In this manner, a more fundamental insight into the mechanism of embrittlement may be gained. The steel selected for this purpose was a low-carbon, columbium-bearing steel which, after solution treatment, was aged in various time-temperature cycles while attempting to hold all other microstructural variables constant. MATERIALS AND PROCEDURE The composition of the steel used in this investigation is given in Table I. Starting with commercially hot-rolled material 0.074 and 0.280 in. thick, tensile blanks were cut from the thinner material and sub-size Charpy blanks from the thicker stock. The steel had been rolled using a low finishing temperature and rapid cooling after hot rolling to produce grain refinement. Also, chemical-extraction analysis and electron-microscopic examination revealed that the cooling was rapid enough to suppress any significant precipitation of CbC. Therefore, the as-rolled steel is essentially in the solution-treated condition in terms of the precipitation reaction, and it was used as the base material in this investigation. The room-temperature tensile properties of the as-rolled steel aged at 1100", 1150", and 1200°F for various time intervals were determined, see Table 11. The ductile-brittle transition behavior was studied for smooth and notched tensile specimens, and longitudinal half-size (0.200 in. thick) Charpy V notch specimens aged at 1100" and 1150°F for various times, also Table 11. In addition, tensile specimens aged for single times at 1050" and 1200°F were evaluated with respect to their ductile-brittle transition. Only one 1050°F treatment was used because of the extremely slow kinetics of precipitation at this temperature and because no significant difference between 1100°F aging was observed. Only one treatment at 1200°F was used because aging at this temperature alters the shape of the grain boundary cementite particles present in this steel and complicates analysis. The design of the tensile specimens is given in Fig.1.

Jan 1, 1968

By P. Chiotti, R. J. Hecht

Vapov pressure data of solid and liquid alloys along with phase diagram data WERE employed in calculating the thermodynamic properties for the Ca-Zn system. The relative excess partial molal free energy for zinc in the liquid alloys was fitted to the relation wnvre i is we tvmpvrutiirc in ucgrees neivtu ana in is the tr~ole fvaction. Relations for the standard free energy of fortnation for the eight cornpounds were developed. ThE phase diagram for the Ca-Zn system as determined by Messing et a1.&apos; is reproduced in Fig. 1. Earlier work on the phase diagram has been summarized by Hansen and Anderko2 and the very limited thermodynamic information for this system has been reviewed by Hultgren et a1.3 The heat of fusion for the compound CaZn2 has recently been determined to be 9260 cal per mole.4 Solubility data for calcium in zinc for the temperature range of 650" to 725°C were fitted by the relation and reported by Johnson and Dilloon5 In the present investigation zinc vapor pressure data were obtained by both the dewpoint and Knudsen methods in order to calculate the thermodynamic properties for liquid and solid alloys. The dewpoint method as employed in this work requires that only one of the components be significantly volatile, in this case zinc, and that the vapor pressure of pure zinc be known. For calculation purposes the vapor pressure data for pure zinc3 were fitted to the relations; for temperatures below 692.7K, the melting point of zinc. The corresponding free-energy relations for vaporization and sublimation of pure zinc are respectively. Some of the calculations also required knowledge of the free energy of fusion of zinc and calcium and the free energy for the 721°K allotropic transformation of calcium. The free energy of fusion of zinc was taken to be the difference between Eqs. [4] and [5], or The free energy of fusion and transition for calcium were based on the enthalpy values of 2040 and 222 cal per g-atom, respectively, as reported by Chiotti et al. 4 and the heat capacity data given by Hultgren et a1.3 and are represented by the relations, In the latter expression the difference in heat capacity between the a and 0 forms of calcium was neglected. MATERIALS AND EXPERIMENTAL PROCEDURES The metals used in preparing the alloys were Bunker Hill slab zinc, 99.99 pct pure, and Ames Laboratory calcium. The average impurities in the calcium in parts per million by weight, ppm, were N-17, H—77, Ba—216, Mg-15, Si—7, C—157, and Fe—9. The zinc was cleaned in dilute nitric acid and the calcium was used as received. The calcium was stored and handled in an argon-filled glove box and when removed from the dry box it was kept in an argon-filled desiccator. Tantalum was found to be a suitable alloy container. The alloy components were contained in a 1-in.-diam cylindri-

Jan 1, 1968

Jan 1, 1967

Jan 1, 1968

By Y. C. Liu, G. A. Alers

The anisotropy of Young&apos;s modules has been measured in a series of Cu-Zn alloys containing from 1 to 28 wt pct Zn and cold-rolled to 97.5 pct reduction in thickness. These modulus data show a transition in texture as a function of zinc content. In alloys containing 3 pct Zn or less, the texture is essentially that of copper. A transition is clearly seen to develop in 5 and 8 pct Zn alloys, and to have been completed in a 12 pct Zn alloy. Increasing the proportion of zinc to 28 pct does not produce any further changes in texture. The ease and clarity with which this texture transition can be seen in the modulus data is in great contrast with the inccertainty in the visual detertiination of the range of texture transition from a series of pole figures. It is also shown that the rolling texture of the high zinc content alloys cannot be a one-COIIIponenl texture, characterized by the (110)(112) ideal orientations, as usually assumed. A second, quite important component must also be present. Furthermore, the present data are inconsistent with matrix orientations on which the "oriented growth" hypothesis of the Formation of recrystallization textures is based. THE rolling textures of binary fcc alloy systems have generally been considered to lie between two terminal textures. One of these terminal textures is the copper type and the other is the 70:30 brass type. Alloys of intermediate solute content exhibit a texture which is said to be a transition between the two terminal textures. On the basis of efforts to correlate rolling and recrystallization textures, two concepts have been advanced in the literature on the nature of these transition textures. One school1 suggests that the addition of alloying elements, e.g., zinc, causes a gradual rotation about one of the (111) poles from the (358)( 523) ideal orientation of the copper-type rolling texture to the terminal {1 10)( 112) ideal orientation used to describe the rolling texture of 70:30 brass. This concept is depicted in Fig. 1 with the angle of rotation corresponding approximately to the zinc content and is designated here as the "rotational concept", since the ideal orientation of the transition textures varies continuously with alloy content. The other concept,&apos; the "mixture concept", suggests that the copper-type rolling texture is a two-component texture. {358) (523) and {110)(112). The addition of alloying elements affects only quantitatively the relative amounts of these two components—a decrease in the amount of material in the {385}(523) orientation and an increase of that in the {110)(112) orientation with increasing solute content. An experimental distinction between these two concepts is essential to a better understanding of both the mechanism of formation of rolling textures and the subsequent formation of recrystailization textures in fcc metals. Because of the lack of resolution inherent in pole figures, it has not been possible to differentiate between the two concepts on the basis of this data. Both concepts, in fact, make use of the same set of pole figures but interpret them in a manner acco-rding to the view held.&apos;" Experimental results obtained from inverse pole figures or selected-area electron diffraction3j4 simply -yield more refined pole figures but do not eliminate the ambiguity of interpretation. Quantitative measurement of an entirely different, orientation-dependent property must thus be considered

Jan 1, 1967

By C. M. Sellars, F. Maak

Electvomotie -force measurements hazle been made on ten Au-Ni alloys at temperatures 7754 825O, 900O, and 935°C using galvanic cells with solid electrolyte. Partial and ivtegral thermodynamic functions are derived from the determined activities. which show la9,ge positive deviations jvom Raoult&apos;s law. The enthalpies of mixing- are in close accord with calorinletric data and are of comparable accuracy. The results are used to determine the miscibility gap and the spinoidal curL1e, and are qlalitaticely interpreted in terms of the large size difference and electvonic interactions between gold and nickel. Tabulated nalues of the actielities and partial and inteqal free energies, entropies , and enthalpies of mixing are given in the appendix. The Au-Ni binary alloys are of particular interest thermodynamically as despite the large size difference between gold and nickel atoms they have a simple phase diagram.&apos; This shows complete solid solubility above a miscibility gap with a maximum at about 71 at. pct Ni and 810°C. Activity measurements (over a range of temperature) have been made previously on this system using an electrolytic-cell method with a liquid chloride eletrolyte. The accuracy of this data has, however, been queried3&apos;4 as it leads to values for the enthalpy of mixing which are too high when compared with calorimetric measurements.46 Also, the entropy of mixing is too high when compared with the values derived from specific-heat measrements. It was therefore considered worthwhile redetermining the activity data using a galvanic cell with a solid electrolyte (0.85 ZrO, 0.15 CaO). This method is based on a technique introduced by Kiukkola and aner&apos;&apos; and has been used to determine the thermodynamic data for Cu-Ni alloys.&apos; THE ELECTROLYTIC CELL The cell used to determine the activity of nickel in Au-Ni alloys was: The cell reaction, which is discussed for a similar cell by Rapp and Maak, gives a direct measurement of the activity of nickel in the alloys as where aNi(alloy) is the activity of nickel in the alloy, 3 is Faraday&apos;s constant, E is the measured electromotive force, R is the gas constant, and T is absolute temperature. For the valid application of this type of cell the arrangement must be ther modynamic ally stable, which means that: 1) The free energies of formation of the oxides fulfill the relation 2) The vapor pressures or dissociation pressures of each component must be so small that during the time of the experiment the amount of any material transported by the gas is negligible. 3) Side reactions which change the alloy composition must be negligible. These reactions are possible both with the surrounding gas atmosphere and with neighboring solid material. The first two conditions are fulfilled by the above cell.&apos; The third condition is approached by care in the experimental technique. The factors which control the attainment of equilibrium in this type of cell have been considered by Rapp and aak, who conclude that metallic diffusion in the alloy is the rate-control ling step. A lower temperature limit for the practical application of this cell is therefore set by the condition that equilibrium must be attained in reasonable times. EXPERIMENTAL MATERIALS AND TECHNIQUE The starting materials were mint gold (99.99 pct Au) and Mond nickel (99.9 pct Ni). Alloys, nominally 5 at. pct Ni and then each 10 at. pct across the system, were produced by vacuum-levitation melting compacts of mixed drillings of the pure metals. The nickel used for the reference tablets in the cells was melted in a similar manner. The alloy ingots, cast in boron nitride molds, were cold-forged and

Jan 1, 1967

By H. Morgolin, M. A. Greenfield, P. A. Farrar

In attempts to correlate mechanical properties with microstructure, it is desirable to determine the effect of ß grain size. Determination of 6 grain size is not a problem where a has precipitated at ß grain boundaries. However, in alloys which have been processed to produce an equiaxed a +ß structure the ß grain size is difficult to determine because ß grain boundaries do not etch. The ß grain size can sometimes be laboriously determined by searching for grains which have some a! outlining the 6 grains. This difficulty of ß grain size determination in equiaxed a + ß alloys can be overcome by heating below the ß transus temperature for a short time and water quenching. Fig. 1 shows the starting equiaxed a + ß structure of the Ti-5.25A1-5.5V-0.9Fe-0.5Cu alloy used in this investigation, and Fig. 2 shows the microstructure of the same sample water-quenched after heating 1 min at 1675"F, which is 25°F below the 0 transus. The heavily etched microstructure consists of some equiaxed a! in a structure showing considerable amounts of transformed ß. The ß grain

Jan 1, 1969

By H. I. Aaronson, W. L. Winterbottom, G. M. Pound

On the basis of a statistical thermodynamic treatment of the data of smithz2 on the activity of carbon in austenite, a Darken and smith23 deduced that the interaction energy, wy, between carbon atoms occupying nearest-neighboring interstitial sites is repulsive. Aaronson, Domian, and pound 24 (ADP) have applied two different statistical thermodynamic analyses to the same data. One analysis, a composite of the treatments of Darken and Smith,23 Speiser and Spret-nak," and Kaufman, Radcliffe, and Cohen,26 primarily takes account of the positional entropic effects of wy in dilute solutions. The other, due to Lacher 27 and Fowler and Guggenheim," is also applicable to more concentrated solid solutions. The equations which both analyses produce for a,, can be rearranged so that, when In a,, is plotted against the appropriate function of the carbon content of the austenite, the plot will have a slope of unity when the correct value of wy is employed. Since no other constant need be simultaneously evaluated in order to make this determination, ascertaining the most probable value of wy from Smith&apos;s data on the bases of these analyses was an elementary problem in least-squares statistics. ADP demonstrated that the wy&apos;s obtained from both analyses vary with temperature in a statistically significant manner. McLellan, Gerrard, Horowitz, and sprague 29 (MGHS), on the other hand, have recently made a new statistical thermodynamic study of a,, and have concluded from this study that wy is independent of temperature. The equation which they derived for a,, is: where 0 = ratio of the mole fraction of carbon atoms to that of iron atoms, z = number of sites from which occupancy is excluded by the presence of a carbon atom and the repulsive character of its interaction with nearest-neighboring carbon atoms (z = zb in Ref. 24), and x = a geometric parameter taking account of overlapping of the exclusion shells of nearby carbon atoms. z is 1 when wy = 0 and 13 when wy = Q. For a given value of z there is a definite value of X . In the absence of an equation connecting z and X, MGHS computed geometrically the value of x at several integral values of z; x increases from 0.47 to 8.67 as z is increased from 2 to 5. MGHS fitted Eq. [I] to the a,, data of Ellis, Davidson, and Bodsworth30 at 925° and 1050°C and to the data of smithz2 at 1000°C. Confining z to integral values in the range 4 to 6, and also somewhat restricting AG, the values of z, X, and AG were sought which would minimize the average root mean square percent deviation, D, of the calculated a,, values from those obtained experimentally. Finding that D was almost always smallest when z = 5, they concluded that z is independent of temperature. Since:23, 26 where R = gas constant and T = absolute temperature, and w,, was considered to be similarly invariant, this statement cannot be strictly true. The temperature range employed, however, was small and the interval between the values of z tested was sufficiently large so that the results could still be consistent with a constant The statistical treatment of the a,, data applied by MGHS is clearly quite approximate. Since more than three values of a,, are available at each temperature, z, and X, and AG have actually been over deter mined, rather than underdetermined as their treatment implies. In order to make more complete and exact use of the a,, data, we shall employ the Legendre Method of Least Squares, as described by Whittaker and Robinson. This technique permits conversion of the overdetermined system of equations produced by the substitution into Eq. [I] of the available pairs of a,, and 0 data at a given temperature into an even determined system. Solution of the latter system yields the most probable value of each of the three constants. This procedure will allow a statistically rigorous test to be made of the conclusion drawn by MGHS in respect of the temperature dependence of In order to obtain linear iiequations of conditions" from Eq. [I], the exact values of the three unknown constants are replaced by G/RT = =/RT + y, where z, i, and aG are estimated values of the constants. This allows the conditional equations to be written in the form:&apos;l where fi -f6, i, =/RT) = the right-hand side of Eq. [I] for a given pair of ayi and values. When applied to Eq. [I], these relationships become: The i equations of condition are reduced to three "normal equations" in the manner discussed by Whittaker and Robinson. The normal equations are then solved for a, ß, and y. Both of these operations are most conveniently performed by means of the computer technique of Winterbottom and Gjostein.32 This technique was used iteratively until the value of 0 (the most sensitive of the three correction factors) became 10.01. (Until they became small, a, 8, and y

Jan 1, 1969