SEARCH · Engineering Papers
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Major silicon-gate CMOS/SOS processes are described. Sapphire substrate preparation is also discussed, as well as the following process variations: (1) the double epi process; and (2) ion implantation.
Growth configurations were developed which produced crystals having low residual stress levels. The properties of a 106 mm diameter round crucible were evaluated and it was found that this design had greatly enhanced temperature fluctuations arising from convection in the melt. Thermal modeling efforts were directed to developing finite element models of the 106 mm round crucible and an elongated susceptor/crucible configuration. Also, the thermal model for the heat loss modes from the dendritic web was examined for guidance in reducing the thermal stress in the web. An economic analysis was prepared to evaluate the silicon web process in relation to price goals.
A multi-step silicon etching process has been developed to fabricate silicon-based terahertz (THz) waveguide components. This technique provides precise dimensional control across multiple etch depths with batch processing capabilities. Nonlinear and passive components such as mixers and multipliers waveguides, hybrids, OMTs and twists have been fabricated and integrated into a small silicon package. This fabrication technique enables a wafer-stacking architecture to provide ultra-compact multi-pixel receiver front-ends in the THz range.
Chemical engineering analyses involving the preliminary process design of a plant (1,000 metric tons/year capacity) to produce silicon via the technology under consideration were accomplished. Major activities in the chemical engineering analyses included base case conditions, reaction chemistry, process flowsheet, material balance, energy balance, property data, equipment design, major equipment list, production labor and forward for economic analysis. The process design package provided detailed data for raw materials, utilities, major process equipment and production labor requirements necessary for polysilicon production in each process.
Two computer codes were developed for describing flow reactors in which high purity, solar grade silicon is produced via reduction of gaseous silicon halides. The first is the CHEMPART code, an axisymmetric, marching code which treats two phase flows with models describing detailed gas-phase chemical kinetics, particle formation, and particle growth. It can be used to described flow reactors in which reactants, mix, react, and form a particulate phase. Detailed radial gas-phase composition, temperature, velocity, and particle size distribution profiles are computed. Also, deposition of heat, momentum, and mass (either particulate or vapor) on reactor walls is described. The second code is a modified version of the GENMIX boundary layer code which is used to compute rates of heat, momentum, and mass transfer to the reactor walls. This code lacks the detailed chemical kinetics and particle handling features of the CHEMPART code but has the virtue of running much more rapidly than CHEMPART, while treating the phenomena occurring in the boundary layer in more detail.
Mathematical models and computer codes based on these models, which allow prediction of the product distribution in chemical reactors for converting gaseous silicon compounds to condensed-phase silicon were developed. The following tasks were accomplished: (1) formulation of a model for silicon vapor separation/collection from the developing turbulent flow stream within reactors of the Westinghouse (2) modification of an available general parabolic code to achieve solutions to the governing partial differential equations (boundary layer type) which describe migration of the vapor to the reactor walls, (3) a parametric study using the boundary layer code to optimize the performance characteristics of the Westinghouse reactor, (4) calculations relating to the collection efficiency of the new AeroChem reactor, and (5) final testing of the modified LAPP code for use as a method of predicting Si(1) droplet sizes in these reactors.
Methods to improve the radiation tolerance of silicon cells for spacecraft use are described. The major emphasis of the program was to reduce the process-induced carbon and oxygen impurities in the junction and base regions of the solar cell, and to measure the effect of reduced impurity levels on the radiation tolerance of cells. Substrates of 0.1, 1.0 and 10.0 ohm-cm float-zone material were used as starting material in the process sequence. High-dose, low-energy ion implantation was used to form the junction in n+p structures. Implant annealing was performed by conventional furnace techniques and by pulsed laser and pulsed electron beam annealing. Cells were tested for radiation tolerance at Spire and NASA-LeRC. After irradiation by 1 MeV electrons to a fluence of 10 to the 16th power per sq cm, the cells tested at Spire showed no significant process induced variations in radiation tolerance. However, for cells tested at Lewis to a fluence of 10 to the 15th power per sq cm, ion-implanted cells annealed in vacuum by pulsed electron beam consistently showed the best radiation tolerance for all cell resistivities.
A procedure was perfected for the production of components used in engineering applications of silicon nitride. Particles of complex geometry that combine remarkable mechanical properties with a high density are obtained. The process developed, in contrast to the "hot pressing" method, does not use external pressures, and in contrast to the reaction bonding method, final densities close to the theoretical value are obtained.
The program aims at developing mathematical models and computer codes based on these models, which allow prediction of the product distribution in chemical reactors for converting gaseous silicon compounds to condensed-phase silicon. The major interest is in collecting silicon as a liquid on the reactor walls and other collection surfaces. Two reactor systems are of major interest, a SiCl4/Na reactor in which Si(l) is collected on the flow tube reactor walls and a reactor in which Si(l) droplets formed by the SiCl4/Na reaction are collected by a jet impingement method. During this quarter the following tasks were accomplished: (1) particle deposition routines were added to the boundary layer code; and (2) Si droplet sizes in SiCl4/Na reactors at temperatures below the dew point of Si are being calculated.
Silicon-dominant anodes are of great interest because of their potential to boost the cell-level energy of state-of-the-art Li-ion batteries. While silicon materials have been extensively studied, understanding interactions at the electrode level has recieved little attention, especially the coating process of Si particles, which plays an equally important role in unlocking the full potential of silicon anodes. Herein, the electrode processing of a Si-dominated anode (52.8 wt%, 3.5–4.5 mAh cm −2 ) is being investigated to understand the relationship of processing on the morphology and properties of Si anodes at the electrode level. It has been found that almost-undetectable Si agglomerates easily form during electrode processing, which grow into largeprotrusions after lithiation and trigger potential internal shorting and self-discharge problems. A facile slurry filtration step is proposed to homogenize the particle distribution within Si-dominant electrodes which improves the electrochemical performance and storage stability of Si-based Li ion batteries.
Process of growing expitaxial silicon carbide (SiC) layers on SiC substrates so that epitaxial growth is perpendicular to c-axis by chemical vapor deposition process at temperatures of 1590 to 1660 K minimizes variations in stacking sequence and problems associated with high temperatures.
Thirty-five (35) furnace runs were carried out during this quarter, of which 25 produced a total of 120 web crystals. The two main thermal models for the dendritic growth process were completed and are being used to assist the design of the thermal geometry of the web growth apparatus. The first model, a finite element representation of the susceptor and crucible, was refined to give greater precision and resolution in the critical central region of the melt. The second thermal model, which describes the dissipation of the latent heat to generate thickness-velocity data, was completed. Dendritic web samples were fabricated into solar cells using a standard configuration and a standard process for a N(+) -P-P(+) configuration. The detailed engineering design was completed for a new dendritic web growth facility of greater width capability than previous facilities.
Mechanisms for the SiCl4/Na and SiF4/Na reaction systems were examined. Reaction schemes which include 25 elementary reactions were formulated for each system and run to test the sensitivity of the computed concentration and temperature profiles to the values given estimated rate coefficients. It was found that, for SiCl4/Na, the rate of production of free Si is largely mixing-limited for reasonable rate coefficient estimates. For the SiF4/Na system the results indicate that the endothermicities of many of the reactions involved in producing Si from SiF4/Na cause this system to be chemistry-limited rather than mixing-limited.
Two major topics are treated: methods to measure and evaluate impurity effects in silicon and comprehensive tabulations of data derived during the study. Discussions of deep level spectroscopy, detailed dark I-V measurements, recombination lifetime determination, scanned laser photo-response, conventional solar cell I-V techniques, and descriptions of silicon chemical analysis are presented and discussed. The tabulated data include lists of impurity segregation coefficients, ingot impurity analyses and estimated concentrations, typical deep level impurity spectra, photoconductive and open circuit decay lifetimes for individual metal-doped ingots, and a complete tabulation of the cell I-V characteristics of nearly 200 ingots.
A barrier crucible design which consistently maintains melt stability over long periods of time was successfully tested and used in long growth runs. The pellet feeder for melt replenishment was operated continuously for growth runs of up to 17 hours. The liquid level sensor comprising a laser/sensor system was operated, performed well, and meets the requirements for maintaining liquid level height during growth and melt replenishment. An automated feedback loop connecting the feed mechanism and the liquid level sensing system was designed and constructed and operated successfully for 3.5 hours demonstrating the feasibility of semi-automated dendritic web growth. The sensitivity of the cost of sheet, to variations in capital equipment cost and recycling dendrites was calculated and it was shown that these factors have relatively little impact on sheet cost. Dendrites from web which had gone all the way through the solar cell fabrication process, when melted and grown into web, produce crystals which show no degradation in cell efficiency. Material quality remains high and cells made from web grown at the start, during, and the end of a run from a replenished melt show comparable efficiencies.
Results of a study to utilize an Nd:glass laser for production line annealing of ion implantation induced damage in solar cells are reported. Czochralski-grown and sawn Si wafers 7.6 cm in diam, 0.35 mm thick, were implanted with phosphorus junctions and boron BSFs. Annealing with electron beam, laser, and firing of an Al paste to form the BSFs in different cells was compared. The laser was employed at 1.06 and 0.53 micron and in combination of both, with a 20-50 nsec pulsewidth, and energy densities of 1.2, 1.5, 1.9, and 2.1 J/sq cm. Best optical coupling was observed with the combined wavelengths and a 20 nsec pulse, using energy densities less than 1.5 J/sq cm. Although the Al sintered cells displayed the best characteristics, laser annealing is concluded to offer electrically active, defect-free, shallow junction Si substrates for high efficiency cells.
The effects of oxidation of Si which established that vacancies (V) and Si self interstitials (I) coexist in Si at high temperatures under thermal equilibrium and oxidizing conditions are discussed. Some essential points associated with Au diffusion in Si are then discussed. Analysis of Au diffusion results allowed a determination of the I component and an estimate of the V component of the Si self diffusion coefficient. A discussion of theories on high concentration P diffusion into Si is then presented. Although presently there still is no theory that is completely satisfactory, significant progresses are recently made in treating some essential aspects of this subject.