The gettering effectiveness of various backside gettered polysilicon, silicon nitride, or poly + nitride film structures on n-type (100) Czochralski silicon wafers from a single ingot during simulated complementary-metal-oxide-semiconductor process cycles has been investigated by synchrotron section topography, breakdown voltage of 25-nm gate oxide, and minority carrier lifetime measurements. Interfacial structure was studied by cross-section transmission electron microscopy. Our studies show that for successful implementation of thin film backside gettering, good control of interactions between intrinsic gettering and thin film backside gettering during device processing cycles is critical. Thin film deposition generally increases oxygen precipitation. No extended defects propagate toward the bulk silicon. The structural and electrical quality of a denuded zone is found important. The best device yield (90%) is observed from the middle section wafers (with a 2 μm thick polysilicon backsurface coverage) which develop no recognizable bulk precipitates and stacking faults after the complete thermal cycles.

Previous workers have shown that germane does not decompose below 700 K on silica substrates. However, if the germane is irradiated with an ArF or KrF laser, germanium films will grow at temperatures down to 610 K. In the work reported here, the role of the laser in the growth process was explored. First, the species impinging on the substrate during ArF laser-assisted chemical vapor deposition (LCVD) were analyzed with a mass spectrometer. The analysis showed that when germane is irradiated with an ArF laser under growth conditions, digermane is formed in the gas phase. The digermane then impinges on the substrate and pyrolyses. Many other species have been reported previously to be formed transiently during the laser pulse. However, we were not able to detect any other species with a high enough flux onto the substrate to condense at an appreciable rate. Next, temperature programmed decomposition was used to see if photochemical decomposition of chemisorbed germane played an important role in film growth. It was found that a laser dose comparable to that needed for film growth did not produce an appreciable change in the decomposition spectra. Lastly, film growth studies were used to examine the role of the digermane formed by the laser in the film growth process. It was found that a mixture of germane and digermane will decompose on a silica substrate down to 580 K to produce a germanium film. The growth characteristics of the germane/digermane mixtures were virtually the same as those seen in LCVD. Thus, it appears that under the conditions used previously for LCVD, the main role of the ArF laser is to convert some of the germane source gas to digermane. It also has been found that under other conditions, one can grow germanium films down to at least 300 K via digermane photolysis.

The reactions of SiCl4 and GeCl4 with oxygen and oxides of nitrogen have been studied in a continuous flow system. SiCl4 is found to react with nitrous oxide at a considerably lower temperature (700 °C vs 1050 °C) than with oxygen. In both cases the formation of silicon oxychlorides (hexachlorodisiloxane and octachlorocyclo-tetrasiloxane) was observed for the first 300 °C after reaction initiates. Activation energies of 115 kcal/mole and 40 kcal/mole were determined for the reactions with oxygen and nitrous oxide, respectively (under conditions of 100-fold excess oxidizer). GeCl4 is also found to react with nitrous oxide at a lower temperature than with oxygen; however, in this case no oxychloride species were observed. Activation energies of 53 kcal/mole and 48.6 kcal/mole were determined for the reactions with oxygen and nitrous oxide, respectively (under conditions of 70-fold excess oxidizer). Reactions of GeCl4 with nitric oxide and nitrogen dioxide are also reported. The use of N2O as an oxidant in optical fiber fabrication processes is discussed.

Microwave plasma assisted CVD diamond films have been studied by transmission electron microscopy. Cross-section TEM as well as plan-view TEM methods were used to investigate various structural defects formed in the films. It was found that diamond films contain a large number of stacking faults and twins which lie on the {111} planes in diamond. With an increase in methane concentration during the deposition process, the density of these defects increases and their dimensions become smaller. The tendency for forming these structural defects is of concern in developing tailored structures and properties of diamond films.

Diamond synthesis from the vapor phase has been studied using hot filament assisted CVD, microwave plasma assisted CVD in an open system, and chemical transport process in a closed system using a gas mixture of hydrocarbon diluted with hydrogen gas. The deposited materials were identified as a cubic diamond with x-ray diffraction and Raman scattering measurement. The deposition process of diamond was considered on the basis of these experimental results and a review of the relevant field. The results suggest that the diamond synthesis proceeded by the deposition process, with the etching process operating simultaneously, and that atomic hydrogen played a very important role as the etching agent for non-diamond carbon. Also, the deposition process would be considered as the method that utilizes the bonding energy difference on each crystal surface of the diamond. The complicated problems on diamond synthesis originate from the fact that a third allotropic form exists in carbon, that is, the carbyne group.

C-cut sapphire was implanted with 400 keV Co+ ions to doses of 3 × 1017/cm2 and 5 × 1017/cm2 and then annealed in air sequentially at 1273, 1373, and 1473 K. Before and after the annealing, the implanted surfaces were examined by RBS and XRD methods. A nearly homogeneous layer of 350 nm thickness was formed on the surface when sapphire was implanted to a dose of 5 × 1017/cm2 and then annealed at 1473 K. For sapphire implanted to a dose of 3 × 1017/cm2 and then annealed at 1473 K, the implanted layer had a complex structure: the top surface layer of 70 nm thickness was α-alumina with Co3+ ions at the substitutional sites and the near-surface layer (70 nm–300 nm) was a mixed phase of CoAl2O4 and α-alumina.

The influence of coherency strains on phase equilibria in a two-phase microstructure is examined for a binary or pseudobinary alloy system possessing a consolute critical point (chemical miscibility gap). The qualitative features of phase equilibria, including the limits of metastability (chemical spinodal), are shown to depend critically on the mechanical loading conditions and the geometric arrangement of the phases in the microstructure. If the elastic state of a phase in a two-phase coherent system is independent of the presence of the other phase, then the equilibrium characteristics usually associated with fluid systems should be observed, even though the system is nonhydrostatically stressed. If the elastic state of a phase depends upon the presence of the other phase, then the equilibrium characteristics that have come to be associated with coherent systems should be observed; tie lines and field lines do not coincide, the common tangent construction is invalid, and Gibbs phase rule is not applicable.

The Young's modulus of silica aerogels is measured, using the three point flexural technique. Various parameters have been investigated, such as the concentration of silicon compounds as well as the catalyting conditions used to develop the initial alcogel and the heat treatment performed to densify the material. The elastic behavior of aerogels depends on the conditions of gel preparation, and the Young's modulus is shown to obey the power dependence E ≃ ρ3.7. This scaling exponent is compared to the percolation model predictions. On the other hand, the elastic behavior of different sets of aerogels is related to the spatial arrangement of the particles pointed out by fractal features.

The fluoride anion has been shown to be one of the most effective catalysts in accelerating the polycondensation of alkoxide-derived silica gels. A detailed pore analysis study has been employed to investigate the effect of NaF on the pore structure of the resulting xerogels and its evolution during thermal heat treatment up to 800 °C. Addition of NaF to tetramethylorthosilicate-sols leads to an increase in average pore size, and the pore size distribution becomes narrower. By changing the fluoride concentration and the heating temperature, the surface properties of the xerogels can be tuned over a wide range. The possible application of the F− anion catalyzed sol gel process to prepare porous host materials for the studies of fluids in restricted geometries is also discussed.

Alkali metal ion exchange isotherms at a total ionic strength of 0.001 M were determined at 25 °C on a 1.13 nm anomalous [Al3+ + Na+]-substituted tobermorite with the formula, Ca5Na0.75Al0.9Si5.1O16(OH)2 · 6.03H2O. The Kielland plots of Na+/K+, Na+/Rb+, and Na+/Cs+ exchange reactions showed straight lines. The slopes were steeper for Cs+ and Rb+ exchange reactions, as compared to K+ exchange. The cation-exchange selectivity for alkali metals was found to increase as follows: Cs+>Rb+>K+>Na+. A new evaluation method of the separation factor (αMN/KMd/KNd,Kd: distribution coefficient) for a combination of two cations at infinite dilution was proposed in terms of the selectivity coefficient (KMNa) which can be easily determined from the Kielland plot. An extremely large separation factor for Cs (αCsNa = 112) was found at infinite dilution. These basic studies of cation exchange selectivity are of relevance in cation separation and purification and nuclear waste disposal.

The Panel was charged with assessing the present scientific understanding of the size-dependent physical and chemical properties of clusters, the methods of synthesis of macroscopic amounts of size-selected clusters with desired properties, and most importantly, the possibility of their controlled assembly into new materials with novel properties. The Panel was composed of both academic and industrial scientists from the physics, chemistry, and materials science communities, and met in January 1988.

In materials (insulators, semiconductors, and metals) with strong chemical bonding, there is extensive spatial delocalization of valence electrons, and therefore the bulk physical properties which depend upon these electrons develop only gradually with cluster size. Recent research using supersonic-jet, gas-aggregation, colloidal, and chemical-synthetic methods indeed clearly establishes that intermediate size clusters have novel and hybrid properties, between the molecular and bulk solid-state limits. A scientific understanding of these transitions in properties has only been partially achieved, and the Panel believes that this interdisciplinary area of science is at the very heart of the basic nature of materials. In Sec. V (Future Challenges and Opportunities), a series of basic questions for future research are detailed. Each question has an obvious impact on our potential ability to create new materials.

Present methods for the synthesis of useful amounts of size-selected clusters, with surface chemical properties purposefully controlled and/or modified, are almost nonexistent, and these fundamentally limit our ability to explore the assembly of clusters into potentially novel materials. While elegant spectroscopic and chemisorption studies of size-selected clusters have been carried out using molecular-beam technologies, there are no demonstrated methods for recovery and accumulation of such samples. Within the past year, the first reports of the chemical synthesis of clusters with surfaces chemically modified have been reported for limited classes of materials. Apparatus for the accumulation and consolidation of nanophase materials have been developed, and the first promising studies of their physical properties are appearing. In both the chemical and nanophase synthesis areas, clusters with a distribution of sizes and shapes are being studied. Progress on macroscopic synthetic methods for size-selected clusters of controlled surface properties is the most important immediate goal recognized by the Panel. Simultaneous improvement in physical characterization will be necessary to guide synthesis research.

Assuming such progress will occur, the Panel suggests that self-assembly of clusters into new elemental polymorphs and new types of nanoscale heterogeneous materials offers an area of intriguing technological promise. The electrical and optical properties of such heterogeneous materials could be tailored in very specific ways. Such ideas are quite speculative at this time; their implementation critically depends upon controlled modification of cluster surfaces, and upon development of characterization and theoretical tools to guide experiments.

The Panel concluded that a number of genuinely novel ideas had been enunciated, and that in its opinion some would surely lead to exciting new science and important new materials.