SEARCH · Search NASA
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.
Web dendrite growth process pulls sheet of newly crystallized silicon from molten silicon. Jets of argon pull outside gas into melt cavity, preventing silicon oxide from passing through heat-shield hold and depositing on it. Generated by aspirators, reversed flow is used in web dendrite process, which produces sheets of single-crystal silicon for low-cost solar cells.
Technique produces reliable bonds of silicon crystal to molybdenum able to withstand continuous operation at 1,000 C in vacuum environment.
Wafer-like crystals suitable for making solar cells are formed by pouring molten Si containing suitable dopants into a mold of the desired shape and allowing it to solidify in a temperature gradient, whereby the large surface of the melt in contact with the mold is kept at less than 200 D and the free surface is kept at a temperature of 200-1000 D higher, but below the melting point of Si. The mold can also be made in the form of a slit, whereby the 2 sides of the mold are kept at different temperatures. A mold was milled in the surface of a cylindrical graphite block 200 mm in diameter. The granite block was induction heated and the bottom of the mold was cooled by means of a water-cooled Cu plate, so that the surface of the mold in contact with one of the largest surfaces of the melt was held at approximately 800 D. The free surface of the melt was subjected to thermal radiation from a graphite plate located 2 mm from the surface and heated to 1500 D. The Si crystal formed after slow cooling to room temperature had a columnar structure and was cut with a diamond saw into wafers approximately 500 mm thick. Solar cells prepared from these wafers had efficiencies of 10 to 11%.
The most developed process for silicon crystal growth is the Czochralski (CZ) method which was in production for over two decades. In an effort to reduce cost of single crystal silicon for photovoltaic applications, a directional solidification technique, Heat Exchanger Method (HEM), was adapted. Materials used in HEM and CZ furnaces are quite similar (heaters, crucibles, insulation, etc.). To eliminate the cost of high purity argon, it was intended to use vacuum operation in HEM. Two of the major problems encountered in vacuum processing of silicon are crucible decomposition and silicon carbide formation in the melt.
The power needed to zone silicon crystals by radio frequency heating was analyzed. The heat loss mechanisms are examined. Curves are presented for power as a function of crystal diameter for commercial silicon zoning.
Crack growth in single-crystal silicon at room temperature in air was evaluated by double torsion (DT) load-relaxation method and monitored by acoustic emission (AE) technique. Both DT and AE methods indicated lack of subcritical crack growth in silicon. At the critical stress intensity factor, the crack front was found to be jumping several times in a 'mirror' region and then followed by fast crack growth in a 'hackle' region. Hackle marks were found to be associated with plastic deformation at the tip of the fast moving crack. No dislocation etch pits were found in the 'mirror' region, in which crack growth may result from interatomic bonds broken at the crack tip under stress without any plastic deformation. Acoustic emission appears to be spontaneously generated from both interatomic bonds broken and dislocation generation at the moving crack tip during the crack growth in single-crystal silicon.
This innovation is the environmental qualification of a single-crystal silicon mirror for spaceflight use. The single-crystal silicon mirror technology is a previous innovation, but until now, a mirror of this type has not been qualified for spaceflight use. The qualification steps included mounting, gravity change measurements, vibration testing, vibration- induced change measurements, thermal cycling, and testing at the cold operational temperature of 225 K. Typical mirrors used for cold applications for spaceflight instruments include aluminum, beryllium, glasses, and glass-like ceramics. These materials show less than ideal behavior after cooldown. Single-crystal silicon has been demonstrated to have the smallest change due to temperature change, but has not been spaceflight-qualified for use. The advantage of using a silicon substrate is with temperature stability, since it is formed from a stress-free single crystal. This has been shown in previous testing. Mounting and environmental qualification have not been shown until this testing.
A compact low-cost cryocooling system has been designed, constructed, and tested at SLAC National Accelerator Laboratory. The cooling power is provided by natural convection and phase change of the liquid nitrogen. The initial application was to cool silicon crystal optics to the sub-100 K range. A silicon crystal of dimension (width × depth × height) 50 × 50 × 30 mm3 has been used with an electric heater on the top surface in this prototyping test. This system can effectively provide more than 80 W of cooling power to the optics with a consumption of liquid nitrogen less than 2.1 l/h. The vibration of the silicon crystal was monitored during the tests with added electric heater power on the crystal. We note the vibration of the silicon crystal due to liquid nitrogen boiling is negligible.
The present 1000 C and 1300 C oxidation tests on 111-oriented single-crystal Si and dense CVD Si3N4 notes the oxidation rates of the latter in wet O2, dry O2, wet inert gas, and steam atmosphere conditions to be several orders of magnitude lower than the rates for the former in identical atmospheric conditions. Although the parabolic rate constant for Si increased linearly as the water vapor pressure increased, the parabolic rate constant for Si3N4 exhibited a nonlinear dependency on water vapor pressure in the presence of O2. NO and NH3 formation at the reaction interface of Si3N4, and the counterpermeation of these reaction products, are noted to dominate reaction kinetics.
We have heated polished doped single-crystal silicon wafers in a single mode microwave cavity to temperatures where surface to surface bonding occurred. The absorption of microwaves and heating of the wafers is attributed to the inclusion of n-type or p-type impurities into these substrates. A cylindrical cavity TM (sub 010) standing wave mode was used to irradiate samples of various geometry's at positions of high magnetic field. This process was conducted in vacuum to exclude plasma effects. This initial study suggests that the inclusion of impurities in single crystal silicon significantly improved its microwave absorption (loss factor) to a point where heating silicon wafers directly can be accomplished in minimal time. Bonding of these substrates, however, occurs only at points of intimate surface to surface contact. The inclusion of a thin metallic layer on the surfaces enhances the bonding process.
The goals for the fabrication of single crystal silicon instrument mirrors include the following: 1) Develop a process for fabricating lightweight mirrors from single crystal silicon (SCS); 2) Modest lightweighting: 3X to 4X less than equivalent solid mirror; 3) High surface quality, better than lambda/40 RMS @ 633nm; 4) Significantly less expensive than current technology; and 5) Negligible distortion when cooled to cryogenic temperatures.
Suppression of fluid flow instabilities in the melt by the axial magnetic field in Czochralski silicon crystal growth (AMCZ) is investigated precisely by a high-sensitivity striation etching in conjunction with temperature measurements. The magnetic strength (B) was varied up to 4.0 kG, incremented in 0.5 kG/5 cm crystal length. The convection flow was substantially suppressed at B 1.0 kG. A low oxygen level of 2-3 ppma and a high resistivity of 400 ohm-cm is achieved in the AMCZ silicon crystals at B 1.0 kG. Details of the striation formation as a function of B are presented. Computer simulation of the magnetohydrodynamics of the AMCZ silicon crystal growth are discussed briefly with regard to the solute, especially oxygen segregation at B=0, 1.0, and 2.0 kG. Earlier studies in the inverted Bridgman growth of InSb and Ge, which have established the cause and effect relationship between the convection in the melt and the striation formation as well as the suppression of the convections in the melt by transverse magnetic field are reviewed.
Proton bombardment of high purity single crystal silicon
An apparatus for the Czochralski growth of silicon crystals is presented which is capable of producing multiple ingots from a single crucible. The growth chamber features a refillable crucible with a water-cooled, vacuum-tight isolation valve located between the pull chamber and the growth furnace tank which allows the melt crucible to always be at vacuum or low argon pressure when retrieving crystal or introducing recharge polysilicon feed stock. The grower can thus be recharged to obtain 100 kg of silicon crystal ingots from one crucible, and may accommodate crucibles up to 35 cm in diameter. Evaluation of the impurity contents and I-V characteristics of solar cells fabricated from seven ingots grown from two crucibles reveals a small but consistent decrease in cell efficiency from 10.4% to 9.6% from the first to the fourth ingot made in a single run, which is explained by impurity build-up in the residual melt. The crystal grower thus may offer economic benefits through the extension of crucible lifetime and the reduction of furnace downtime.
We report on the diffusive-ballistic thermal conductance of multi-moded single-crystal silicon beams measured below 1 K. It is shown that the phonon mean-free-path is a strong function of the surface roughness characteristics of the beams. This effect is enhanced in diffuse beams with lengths much larger than, even when the surface is fairly smooth, 510 nm rms, and the peak thermal wavelength is 0.6 microns. Resonant phonon scattering has been observed in beams with a pitted surface morphology and characteristic pit depth of 30 nm. Hence, if the surface roughness is not adequately controlled, the thermal conductance can vary significantly for diffuse beams fabricated across a wafer. In contrast, when the beam length is of order, the conductance is dominated by ballistic transport and is effectively set by the beam cross-sectional area. We have demonstrated a uniformity of +/-8% in fractional deviation for ballistic beams, and this deviation is largely set by the thermal conductance of diffuse beams that support the micro-electro-mechanical device and electrical leads. In addition, we have found no evidence for excess specific heat in single-crystal silicon membranes. This allows for the precise control of the device heat capacity with normal metal films. We discuss the results in the context of the design and fabrication of large-format arrays of far-infrared and millimeter wavelength cryogenic detectors.
Information is given on evaporation and segregation contributions to impurity profiles of floating zone crystals (FZ); high-purity silicon float zoning (FZ); minority-carrier lifetime measurement of heavily doped silicon crystals; the effect of some crystal growth parameters on minority-carrier lifetime; and defect investigations by X-ray topography in graphical and tabular form. It was concluded that evaporation contributes substantially to impurity reduction when FZ or cold-crucible growth is conducted in a vacuum; boron and gallium may be more favorable dopants than indium or aluminum for obtaining high minority-carrier lifetimes; minority-carrier lifetimes greater than 100 microseconds are feasible at a 2 times 10 to the 17th power cm-3 doping level; minority-carrier lifetime decreases with increasing crystal cooling rate and also with the presence of dislocations; the method used to clean silicon feed rods affects lifetime; and microdefect densities in dislocation-free FZ crystals appear to be lower with Ga doping than with B doping.
Plate shaped silicon crystals and their preparation by pouring a silicon melt into a suitable mold and then allowing it to solidify in a temperature gradient were investigated. The production of energy by direct conversion of solar energy into electrical energy by means of solar cells takes on increasing importance. While this type of energy production is already the prevailing form today in the realm of satellite technology, its terrestrial application has thus far encountered strict limitations owing to the high price of such solar cells. Of the greatest interest in this connection are silicon cells. A substantial reduction in the semiconductor material costs and the costs involved in the further processing to make solar cells are prerequisites for a rational market growth for solar energy.
Single crystal silicon carbide micro-sized tensile specimens were fabricated with deep reactive ion etching (DRIE) in order to investigate the effect of stress concentration on the room-temperature fracture strength. The fracture strength was defined as the level of stress at the highest stressed location in the structure at the instant of specimen rupture. Specimens with an elliptical hole, a circular hole, and without a hole (and hence with no stress concentration) were made. The average fracture strength of specimens with a higher stress concentration was larger than the average fracture strength of specimens with a lower stress concentration. Average strength of elliptical-hole, circular-hole, and without-hole specimens was 1.53, 1.26, and 0.66 GPa, respectively. Significant scatter in strength was observed with the Weibull modulus ranging between 2 and 6. No fractographic examination was performed but it was assumed that the strength controlling flaws originated from etching grooves along the specimen side-walls. The increase of observed fracture strength with increasing stress concentration was compared to predictions made with the Weibull stress-integral formulation by using the NASA CARES/Life code. In the analysis isotropic material and fracture behavior was assumed - hence it was not a completely rigorous analysis. However, even with these assumptions good correlation was achieved for the circular-hole specimen data when using the specimen data without stress concentration as a baseline. Strength was over predicted for the elliptical-hole specimen data. Significant specimen-to-specimen dimensional variation existed in the elliptical-hole specimens due to variations in the nickel mask used in the etching. To simulate the additional effect of the dimensional variability on the probabilistic strength response for the single crystal specimens the ANSYS Probabilistic Design System (PDS) was used with CARES/Life.