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.
The various steps involved in obtaining quantitative information of structural defects in crystalline silicon samples are described. Procedures discussed include: (1) chemical polishing; (2) chemical etching; and (3) automated image analysis of samples on the QTM 720 System.
Solar cells fabricated from HEM cast silicon yielded up to 15% conversion efficiencies. This was achieved in spite of using unpurified graphite parts in the HEM furnace and without optimization of material or cell processing parameters. Molybdenum retainers prevented SiC formation and reduced carbon content by 50%. The oxygen content of vacuum cast HEM silicon is lower than typical Czochralski grown silicon. Impregnation of 45 micrometers diamonds into 7.5 micrometers copper sheath showed distortion of the copper layer. However, 12.5 micrometers and 15 micrometers copper sheath can be impregnated with 45 micrometers diamonds to a high concentration. Electroless nickel plating of wires impregnated only in the cutting edge showed nickel concentration around the diamonds. This has the possibility of reducing kerf. The high speed slicer fabricated can achieve higher speed and longer stroke with vibration isolation.
A new calculation of the effects of thermal stresses during growth on silicon ribbon quality is reported. Thermal stress distributions are computed for ribbon growth under a variety of temperature profiles. A growth rate of 55 cu cm/min with a single ribbon was achieved. The growth of RTR ribbon with a fairly uniform parallel dendritic structure was demonstrated. Results with two approaches were obtained for reducing the Mo impurity level in polycrystalline feedstock. Coating the Mo substrate with Si3N4 does not effect thermal shear separation of the polyribbon; this process shows promise of improving cell efficiency and also increasing the useful life of the molybdenum substrate. A number of solar cells were fabricated on RTR silicon grown from CVD feedstock.
The primary objective of this contract is to develop equipment and methods for the economic production of single crystal ingot material by the continuous Czochralski (CZ) process. Continuous CZ is defined for the purpose of this work as the growth of at least 100 kilograms of ingot from only one melt container. During the reporting period (October, 1977 - September, 1978), a modified grower was made fully functional and several recharge runs were performed. The largest run lasted 44 hours and over 42 kg of ingot was produced. Little, if any, degradation in efficiency was observed as a result of pulling multiple crystals from one crucible. Solar efficiencies observed were between 9.3 and 10.4% AMO (13.0 and 14.6% AMI) compared to 10.5% (14.7% AMI) for optimum CZ material control samples. Using the SAMICS/IPEG format, economic analysis of continuous CZ suggests that 1986 DoE cost goals can only be met by the growth of large diameter, large mass crystals.
The JPL Continuous Czochralski Growth Facility is now essentially complete and is functional. In this reporting period, a silicon lump recharging device was designed, and a prototype was built which performed well in simulated recharge tests. The large chamber (designed to accommodate a 14-inch hot zone) was put into operation initially with the standard hot zone in order to compare performance with previous experience. Some modifications were made to the hot zones in the larger chamber due to considerably larger radiative heat losses. Several short (one-ingot) runs were performed and four continuous runs were attempted. The largest continuous run lasted 64 hours and produced 57 kilograms of ingot.
Significant progress is reported in fabricating a 4 sq cm cell having a 10.1 percent conversion efficiency and a 10 sq cm cell having a 9.2 percent conversion efficiency. The continuous (SCIM) coater succeeded in producing a 16 sq cm coating exhibiting unidirectional solidification and large grain size. A layer was grown at 0.2 cm/sec in the experimental coater which was partially dendritic but also contained a large smooth area approximately 100 micron m thick. The dark characteristic measurements of a typical SCC solar cell yield shunt resistance values of 10K ohms and series resistance values and 0.4 ohm. The production dip-coater is operating at over 50 percent yield in terms of good cell quality material. The most recent run yielded 13 good substrates out of 15.
A technique for growing limited-length ribbons continually was demonstrated. This Rigid Edge technique can be used to recrystallize about 95% of the polyribbon feedstock. A major advantage of this method is that only a single, constant length silicon ribbon is handled throughout the entire process sequence; this may be accomplished using cassettes similar to those presently in use for processing Czochralski waters. Thus a transition from Cz to ribbon technology can be smoothly affected. The maximum size being considered, 3 inches x 24 inches, is half a square foot, and will generate 6 watts for 12% efficiency at 1 sun. Silicon dioxide has been demonstrated as an effective, practical diffusion barrier for use during the polyribbon formation.
The various steps involved in the chemical polishing and etching of silicon samples are described and the data on twins, grain boundaries and dislocation pits from fifty-three (53) samples are discussed.
The various steps involved in the chemical polishing and etching of silicon samples are described. Data on twins, dislocation pits, and grain boundaries from thirty-one (31) silicon sample are also discussed. A brief review of the changes made to upgrade the image analysis system is included.
During the reporting period, a successful 100 kilogram run was performed. Six ingots of 13 cm diameter were grown, ranging in size from 15.5 kg to 17.7 kg. Melt replenishment methods included both poly rod and lump feed material. Samples from each ingot were prepared for solar cell fabrication and analyses, impurity analysis, and structural studies. The furnace was converted to the 14-inch hot zone and preliminary heat runs were performed. Two sucessful runs were demonstrated, by growing 25 kg ingots from 30 kg melts. Also, a 100 kg run was attempted, utilizing the 14 inch crucible hot zone, but was prematurely terminated due to excessive monoxide which accumulated on the viewports and a seed failure.
Several 20 cm diameter silicon ingots, up to 6.3 kg. were cast with good crystallinity. The graphite heat zone can be purified by heating it to high temperatures in vacuum. This is important in reducing costs and purification of large parts. Electroplated wires with 45 um synthetic diamonds and 30 um natural diamonds showed good cutting efficiency and lifetime. During slicing of a 10 cm x 10 cm workpiece, jerky motion occurred in the feed and rocking mechanisms. This problem is corrected and modifications were made to reduce the weight of the bladeheat by 50%.
The crystallinity of large HEM silicon ingots as a function of heat flow conditions is investigated. A balanced heat flow at the bottom of the ingot restricts spurious nucleation to the edge of the melted-back seed in contact with the crucible. Homogeneous resistivity distribution over all the ingot has been achieved. The positioning of diamonds electroplated on wirepacks used to slice silicon crystals is considered. The electroplating of diamonds on only the cutting edge is described and the improved slicing performance of these wires evaluated. An economic analysis of value added costs of HEM ingot casting and band saw sectioning indicates the projected add on cost of HEM is well below the 1986 allocation.
The growth of 100 kg of silicon single crystal material, ten cm in diameter or greater, and 150 kg of silicon single crystal material 15 cm or greater utilizing one common silicon container material (one crucible) is investigated. A crystal grower that is recharged with a new supply of polysilicon material while still under vacuum and at temperatures above the melting point of silicon is developed. It accepts large polysilicon charges up to 30 kg, grows large crystal ingots (to 15 cm diameter and 25 kg in weight), and holds polysilicon material for recharging (rod or lump) while, at the same time, growing crystal ingots. Special equipment is designed to recharge polysilicon rods, recharge polysilicon lumps, and handle and store large, hot silicon crystal ingots. Many continuous crystal growth runs were performed lasting as long as 109 hours and producing as many as ten crystal ingots, 15 cm with weights progressing to 27 kg.
The results of a program to demonstrate the feasibility of a low angle silicon ribbon growth process are described. Twenty-six experimental runs were performed. Ribbons were grown at pull rates from 5 to 68 cm/min. Ribbon lengths up to 74 cm were grown while widths varied from 5 to 25 mm. Thicknesses varied from 0.6 to 2.5 mm, with typical values of about 1 mm.
Statistically significant quantitative structural imperfection measurements were made on samples from ubiquitous crystalline process (UCP) Ingot 5848 - 13C. Important correlation was obtained between defect densities, cell efficiency, and diffusion length. Grain boundary substructure displayed a strong influence on the conversion efficiency of solar cells from Semix material. Quantitative microscopy measurements gave statistically significant information compared to other microanalytical techniques. A surface preparation technique to obtain proper contrast of structural defects suitable for quantimet quantitative image analyzer (QTM) analysis was perfected and is used routinely. The relationships between hole mobility and grain boundary density was determined. Mobility was measured using the van der Pauw technique, and grain boundary density was measured using quantitative microscopy technique. Mobility was found to decrease with increasing grain boundary density.
Data from the spherical double probe electric-field experiment on ISEE-1 were used to study plasmasheet/lobe boundary crossings during substorms, identified by plasma measurements and by using the electric field probes as a reference for measurements of the spacecraft potential. There are strong electric fields, with a dominant dawn-to-dusk component, throughout the boundary layer outside the plasmasheet for contracting and expanding motions of the plasmasheet and for different magnetic field directions. Characteristic amplitudes and durations are 5 to 10 mV/m and 5 to 15 min. The corresponding E x B vectors are always towards the plasmasheet.
Noting that the stiffness of the girder increases very rapidly as Beta increases, the result can be summed up as follows: When the cross stress preponderates in one direction and when the web plate is to be given the dimensions commensurate to its stresses, it is advisable (regardless of any ensuing structural difficulties) to set the uprights at about Beta = 120 degrees, thereby lowering the weight of the plate wall 15 percent (in contrast to Beta = 90 degrees), and raising the stiffness 55 percent. But, when the cross stresses alternate and are approximately of the same intensity in both directions, or, if the web plate thickness is determined by other structural reasons, then Beta = 90 degrees should be chosen.
Incompresible, thin sheet flows have been of research interest for many years. Those studies were mainly concerned with the stability of the flow in a surrounding gas. Squire was the first to carry out a linear, invicid stability analysis of sheet flow in air and compare the results with experiment. Dombrowski and Fraser did an experimental study of the disintegration of sheet flows using several viscous liquids. They also detected the formulation of holes in their sheet flows. Hagerty and Shea carried out an inviscid stability analysis and calculated growth rates with experimental values. They compared their calculated growth rates with experimental values. Taylor studied extensively the stability of thin liquid sheets both theoretically and experimentally. He showed that thin sheets in a vacuum are stable. Brown experimentally investigated thin liquid sheet flows as a method of application of thin films. Clark and Dumbrowski carried out second-order stability analysis for invicid sheet flows. Lin introduced viscosity into the linear stability analysis of thin sheet flows in a vacuum. Mansour and Chigier conducted an experimental study of the breakup of a sheet flow surrounded by high-speed air. Lin et al. did a linear stability analysis that included viscosity and a surrounding gas. Rangel and Sirignano carried out both a linear and nonlinear invisid stability analysis that applies for any density ratio between the sheet liquid and the surrounding gas. Now there is renewed interest in sheet flows because of their possible application as low mass radiating surfaces. The objective of this study is to investigate the fluid dynamics of sheet flows that are of interest for a space radiator system. Analytical expressions that govern the sheet geometry are compared with experimental results. Since a space radiator will operate in a vacuum, the analysis does not include any drag force on the sheet flow.