Highly-uniform, hollow glass spheres, which are used for inertial-confinement fusion targets, are formed from metalorganic gel powder feedstock in a drop-tower furnace. The modelling of this gel-to-sphere transformation has consisted of three phases: gel thermochemistry, furnace-to-gel heat transfer, and gravity-driven degradation of the concentricity of the molten shell.

The heat transfer from the furnace to the free-falling gel particle was modelled with forced convection. The gel mass, dimensions, and specific heat as well as furnace temperature profile and furnace gas conductivity, were controlled variables. This model has been experimentally verified. In the third phase, a mathematical model was developed to describe the gravity-driven degradation of concentricity in molten glass shells.

Crystals grown from the melt exhibit compositional and structural defects which limit the exploitation of their full potential in solid state electronics. The origin of these defects is related primarily to gravity-induced convective currents in the melt. In semiconductor compounds additional problems are introduced from variations in stoichiometry. Progress has been made recently in relating qualitatively, and in some instances quantitatively, the growth parameters to the materials properties of the crystals, and in turn to their electronic properties. Overcoming the presence of gravitational forces in space eliminates or minimizes convective interference and, thus, the quantitative assessment of the key growth parameters controlling the chemical and structural perfection of single crystals becomes possible. Results (obtained on earth) will be presented on the growth-property relationships of elemental and compound semiconductors. The advantages, as well as the limitations, that zero gravity conditions present for crystal growth will be discussed.

In the Czochralski growth of single crystals from large melts, fluid flow phenomena have a major effect on interface shape, growth striations, defect density and the length of crystals which can be grown from a melt of given volume and thermal geometry. Because of the technical difficulties encountered in making direct measurements in molten oxides, simulation experiments have been extensively utilized to gain insight into melt behavior.

Both temperature profile and flow geometry results from simulation experiments are discussed. This data is supported by direct melt observations and results from the characterization of grown crystals. When reviewed together, this information offers new insights into the complex behavior of Czochralski growth processes, including the role of thermal gradients, crystal rotation, and surface tension driven (Marangoni) convection.

To maintain a constant growth rate in the absence of convection it is necessary to program down the temperature. Trial and error numerical computations were performed to find the first approximation for the required isothermal dissolution period, linear ramp; and polynomial temperature variation period. The linear ramp rate was limited by the specified maximum temperature gradient. The isothermal dissolution turned out to be unnecessary. The linear ramp had to be stopped before the polynomial period was begun to avoid overshooting the specified maximum growth rate. After a few hours the temperature profile approached steady state behavior.

The growth of crystals from solution is greatly influenced by buoyancy driven convection. In a low-g environment, convection is greatly suppressed and diffusion becomes the predominant mechanism for thermal and mass transport. An experiment to grow TGS crystals by solution technique during the orbital Spacelab III mission has been designed. Crystals are grown by a new and unique technique of extracting heat from the crystal through a sting. The cooling at the sting tip is responsible for the desired supersaturation near the growing crystal. Calculations indicate that the cooled sting technique for solution crystal growth is necessary in low-g to maintain a maximum growth rate of 1 mm/day. Results of groundbased work in support of the flight experiment are discussed.

A series of Hg1−xCdxTe alloy crystals was grown by high-temperature-gradient directional solidification at furnace translation rates ranging from 0.068 to 1.12 μm/s. For several ingots, the measured longitudinal compositional profiles were fitted to theoretical profiles to estimate the magnitude of D, the liquid HgTe-CdTe interdiffusion coefficient. The best-fit value of D was about 5.5 × 105 cm2/s. The majority of the ingots showed significant radial compositional variations along the growth axis. These variations are attributed, at least in part, to fluid flows ahead of the growth interface. The results are discussed in terms of the heat transfer characteristics of the alloy/ampule/ furnace system and the effects of these characteristics on the shape and stability of the growth interface in a 1-g environment.

Theoretical models are described for calculations of charge-carrier concentrations, Fermi energy, and conduction-electron mobility as functions of x, temperature, and ionized- and neutral-defect concentrations in Hg1−xCdxTe alloys. Measurements are reported of electron concentration and electron mobility from 5–300 K for alloys with 0.17 < ˗ < 0.30. The electrical data are in reasonable agreement with theory and were analyzed to obtain estimates of donor and acceptor state concentrations. The electron inobilities are calculated in terms of a microscopic theory of electrical conduction derived from the solution of the Boltzmann equation for the perturbed steady-state electron distribution function and show that longitudinal opticalphonon and charged and neutral defect scattering are the dominant mobility-limiting mechanisms.

The chemical vapor transport and crystal growth properties of the Hgl−xCdxTe iodine system (x=0.2, 0.5 and 0.7) are investigated as a function of transport agent pressure and orientation of the density gradient with respect to the gravity vector to understand the effects of convection on mass transfer and crystal morphology. A thermodynamic analysis of the solid-gas phase system yielded the composition of the vapor and the partial pressures of the gaseous species as a function of temperature and total pressure to assist in the elucidation of the dominant transport mode. Optical and scanning electron microscopy, electron microprobe, chemical etching, X-ray diffraction and energy dispersive spectroscopy techniques were employed to examine the surface and bulk morphology and compositional homogeneity of single crystals obtained under different conditions.

Using a law of mass action approach in which lattice defects are treated as chemical species, the variation of the defect concentrations as a function of .the physicochemical conditions of preparation has been established via measurements of electrical characteristics which are directly related to the defect concentrations.

Continued development of a system designed to grow crystals by physical vapor transport in the environment of Spacelab III will be described, with special emphasis on simulation of expected space conditions, adjustment of crystal growth parameters, and on board observation and control of the experiment by crew members and ground personnel.

A critical factor in the use of mercuric iodide for semiconductor detectors of x-rays and gamma-rays is the crystalline quality of the material. The twofold purpose of the Spacelab III experiment is therefore to grow single crystals with superior electronic properties as an indirect result of the greatly reduced gravity field during the growth, and to obtain data which will lead to improved understanding of the vapor transport mechanism

The experiments planned to evaluate the space crystals, including gamma-ray diffractometry and measurements of stoichiometry, lattice dimensions, mechanical strength, luminescense, and detector performance will be discussed.

Gravity-driven convection can alter the diffusive-advective mass transport behavior in the growth of crystals by physical vapor transport. Specially designed and constructed transparent furnaces are being used in our laboratory to study the effects of gravity in crystal growth of the compound semiconductors PbTe and CdTe.

Extensive experimentation has been carried out in which the kinetics and morphology of dendritic growth were measured as a function of thermal supercooling, solute concentration, and spatial orientation of the dendritic growth axis. The crystal growth system studied is succinonitrile [NC(CH2)2CN] with additions of argon (up to 0.1 mole %). This system is especially useful as a model for alloy studies because kinetic data are available for high purity (7–9's) succinonitrile. The addition of argon provides a simple, controllable dilute solute that now permits the first comparably detailed dendritic growth studies on binary alloys.

One dramatic influence of the solute, at fixed thermal supercooling, is to increase the growth velocity (to a maximum) and correspondingly decrease intrinsic crystal dimensions (tip radius). Morphological measurements will be described in detail relating tip size, perturbation wavelength, supercooling, and solute concentration. The analysis of these effects based on morphological stability theory will also be discussed. Finally, experiments permitting the separation of convective and diffusive heat transport during crystal growth of succinonitrile will be described briefly. These studies clearly underscore the importance of gravitationally-induced buoyancy effects on crystal growth kinetics and morphology.

Although slow convective flow is difficult to detect in solidifying metals, it can readily be observed in transparent materials by observing the motion of small neutrally buoyant particles. An excellent material for such studies is succinonitrile, which solidifies with an unfaceted solid/liquid interface and which has well characterized physical properties. For studies of solute-induced convection, ethanol is a useful addition to succinonitrile because it has a lower density and a somewhat similar molecular structure.

Samples of high purity and ethanol-doped succinonitrile are solidified unidirectionally in a vertical temperature gradient. Latex microspheres, 2μm in diameter, are suspended in the liquid to delineate convective flow. Convective and morphological stability are observed as a function of solute concentration and growth velocity. These measurements are compared to theoretical calculations which predict the transition from stability to instability as a function of solidification conditions. The predicted transitions occur at low concentrations and solidification velocities, so that extreme care is required to eliminate the effects of impurities or thermally-induced convection.

During the processing of monotectic alloys large compositional segregation of the liquid can precede final solidification. Under normal treatment a coarse scale phase distribution is obtained as a result of convection and sedimentation effects which may be minimized by microgravity processing. In undercooled droplets (5–20μ) of Bi-Ga alloys with compositions near the monotectic point, a finescale segregation has been observed and is similar to that formed during a microgravity treatment of bulk samples. Near the critical point of the miscibility gap, Bi-rich alloys exhibit an undercooling prior to phase separation, while in Ga-rich alloys the onset of phase separation is at the miscibility gap boundary. Similarly, in Cu-Pb alloys an undercooling below the miscibility gap prior to phase separation is observed for Cu-rich alloys, but not for Pb-rich alloys near the critical point. These observations are consistent with the operation of a critical point wetting behavior which can modify the liquid segregation pattern during microgravity treatment.

The surface composition of liquid gallium-tin alloys has been studied in an Auger electron spectrometer (AES) as a function of bulk composition and temperature. The sessile drop samples were cleaned by argon ion bombardment sputtering of the liquid. This technique produced surfaces that were entirely free of impurities within the sensitivity of AES and remained so for many days. Tin was found to be strongly adsorbed at the liquid-vacuum interface. Surface concentrations based on Auger measurements are in reasonably good agreement with values calculated from surface tension measurements interpreted in terms of a monolayer depth distribution model for the adsorbed tin.

This research is an effort to determine whether surface tension driven convection in molten tin can be eliminated by the formation of one to several monolayers of oxide on the molten tin surface. Initial work is presented describing Auger and electron loss spectroscopies used to detect SnD and SnO2 on molten tin surfaces in a UHV system; Progress on efforts to produce and monitor controlled oxide layers of 8 to 20 Å upon floating zones in a disk geometry is presented.

The role of gravity on Bridgman-Stockharger directional solidification of eutectic Bi/MnBi has been studied in reduced gravity aboard NASA sounding rocket SPAR flight experiments and contrasted with normal gravity investigations. The directional solidification of eutectic Ri/MnBi results in a low volume fraction, faceted/nonfaceted aligned rod eutectic whose MnRi rod size, interrod spacing, thermal and magnetic properties are sensitive functions of solidification processing conditions. The morphology of the low-gravity samples showed striking differences compared with identically processed,normal gravity samples grown in the same apparatus. The MnBi rod diameter and interrod spacing distributions were significantly smaller, approximately 50%, for the lowgravity samples compared with identically processed one gravity samples. Accompanying the smaller MnBi rod diameters observed in the flight samples, was an increase in permanent magnet properties which reached greater than 97% of the theoretical maximum.

Gravitationally induced thermal instabilities in one-gravity which result in irregular interface movement and associated difficulty of the faceted MnBi phase to branch are suggested to explain the morphological differences between one and low gravity solidification.

Previously we reported on a theoretical treatment of the influence on freezing rate of sudden changes in translation rate in the Bridgman-Stockbarger technique [11]. This has now been extended to consideration of a linear ramped translation rate and an oscillatory freezing rate. Oscillations above a few hertz are found to be highly damped in smalldiameter apparatus.

An experimental test was made of the theoretical predictions for a sudden change of translation rate. MnBi-Bi eutectic was solidified with current induced interface demarcation.The experimental results correspond reasonably well with theory if the silica ampoule wall is assumed either (1) to contribute only a resistance to heat exchange of sample with the furnace wall, or (2) to transmit heat effectively in the axial direction by radiation.

In an attempt to explain the fact that a finer microstructure is obtained in space, MnBi-Bi microstructure is being determined when the freezing rate is rapidly increased or decreased. Preliminary results indicate that fiber branching does not occur as readily as does fiber termination.

Four samples of the Al-In system having monotectic and hypermonotectic compositions were solidified under microgravity during the NASA-SPAR IX flight of January 20th,1981. The experimental thermal and physico-chemical conditions actually achieved have been analysed. Radiographic and metallographic observations of the samples show a non-regular dispersed primary phase inside the monotectic matrix. These observations are commented on the basis of capillarity and solidification aspects.