Extended x-ray-absorption (EXAFS) and electron-energy-loss fine-structure (EXELFS) measurements have been performed on amorphous unhydrogenated silicon carbide, a-SiC, and amorphous hydrogenated silicon carbide, a-SiC:H. Two hydrogenated samples with hydrogen concentrations corresponding, respectively, to H flows of 4 sccm (20% of argon flow) and 8 sccm (40% of argon flow) during the reactive sputtering process, were analyzed (sccm denotes standard cubic centimeters per minute at STP). It is found that short-range order (SRO), consisting of the same tetrahedrally coordinated units present in cubic crystalline c-SiC (zinc-blende structure), where a Si atom is surrounded by nearly four C atoms and vice versa, does exist in all the amorphous samples.This SRO, however, is detected only at a level of the first C and Si coordination shells in a-SiC and a-SiC:H. The structural disorder of the first Si and C coordination shells in all forms of amorphous SiC is somewhat greater than c-SiC, and it decreases appreciably as hydrogen is added. The a-SiC sample exhibits large Si and C coordination numbers, almost identical to c-SiC, a low atomic density, and virtually the same Si -C bond length as c-SiC. These results indicate that a relatively small concentration of large voids exist in a highly disordered a-SiC matrix. The a-SiC:H samples, on the other hand, exhibit a decrease in the C coordination number relative to a-SiC, which is independent of H concentration, low Si and C atomic densities, comparable to a-SiC, and virtually the same Si coordination number as a-SiC. These EXAFS-EXELFS results are consistent with a model where part of the H is substituting for Si in the local tetrahedra surrounding C atoms, while the rest is located inside internal voids in the a-SiC:H samples. The surface of the voids is composed of C atoms which have at least one bond to H, and of Si atoms. Finally, a straightforward computational pro-0 cedure is applied to estimate the size of these voids. It is found that their average size 21 A, thus excluding the possibility that the voids might be point defects.

No abstract

Single crystals of TiCx (x=0.84) were exposed to a carbon source at temperatures near 2000°C. The penetration of carbon into the structure, reflected in a gradient in the value of x, was evaluated with the electron microprobe analyzer. The chemical diffusivity D̃ was then determined using the Boltzmann-Matano method of analysis. The experimental value of D̃ was found to increase with decreasing carbon vacancy concentration, 1−x, and to have the average value of 220 exp (−97 700/RT) cm2/sec. This value of D̃ is in good agreement with one generated from Sarian's 14C tracer measurements on TiC crystals and DePoorter's semitheoretical model for activity. Because of the abundance of vacancies in the carbon fcc sublattice of TiCx, the experimental activation energy represents an energy of motion only. Lower experimental values for the diffusion activation energy have been obtained by other workers with the phase-boundary advance method. The present work shows that the difference between the activation energy measured by the tracer method and that measured by the phase-boundary advance technique cannot be ascribed to the influence of a chemical gradient. Hence it is probably due to grain-boundary short circuiting.

The mechanisms which contribute to the plastic deformation of the intermetallic compound MoSiz have been determined for singlecrystal specimens of MoSi2, (Moo.~Nbo.ol)Siz and ( M O~.~~~T~~.~~~)~~ oriented along [13i] compressed at 1400°C and at a strain rate of 10-s-in vacuum. The results of compression testing of the three single crystals show significant differences between the yield stress values. For the binary MoSi2 single crystal, the yield stress was found to be 34MPa, and for the and (Moo.wzsT&.oo7s)Si2 crystals values of 79 and 38MPa respectively were recorded. An apparent violation of Schmid's law was found in deformation of samples of two of the crystals, namely the binary and Ti-containing version. Thus, while in the Nb-containing compound, the expected slip system, f (1 11){ 1 lo), was activated; this was not the case in the other two compounds. An alternative explanation involves the influence of climb, this being similar to the conclusions drawn in part I for polycrystalline MoSiz. 1. INTRODUCTIONIn part I (Evans, Scheltens, Woodhouse and Fraser 1993, it was shown that experimental results, including mechanical testing and microstructural observations, were consistent with the notion that the contribution to plastic strain from dislocation climb increased significantly with temperature and inverse strain rate at temperatures above 1200°C. Thus the brittle-to-ductile transition appears to be influenced more by climb than by increased dislocation mobilities in glide. Part of the evidence used to determine the contribution from climb involved the identification of active 'slip systems' in which not only are the slip planes physically unreasonable (i.e. not close packed) but also where their Schmid factors are so small that their activation would be considered to be unlikely. Such analysis based on Schrnid factors estimated in polycrystalline samples represents an approximation, and more precise conclusions may be derived from experiments involving single crystals. Hence, the 0 1997 US Government. research described in the following was performed in an attempt to determine unambiguously the factors influencing the brittle-to-ductile transition above 1200°C in MoSi2.The deformation behaviour of single crystals of MoSi2 has been studied by several workers (for example Umakoshi, Sakagami, Hirano and Yamane (1990b), Kimura, Nakamura and Hirano (1390), Boldt, Embury and Weatherley (1992), Maloy, Mitchell, Lewandowski and Heuer (1993a) and Ito, Inui, Shirai and Yamapchi (1995)). The latter paper reports the results of deformation of single crystals of MoSi2 as a function of crystal orientation over a wide range of temperatures, It is found that (111){110), (lOO]{Oll), (100){010), (100]{023) and (331]{013)are activated, depending on orientation. These determinations were the results of analyses of Burgers vectors using diffraction contrast in the transmission electron microscope and also surface slip markings following deformation (up to about 1200"C, since surface contamination precluded the observation of ...

No abstract