Search

Indium iodide (InI) is a promising wide energy band gap nuclear detector material. It is ideal for space experiments because it is non-toxic and has a relatively low melting point of only 351 degrees Centigrade. However, it has been established that melt-grown crystals contain a large amount of second phase inclusions/precipitates. The typical size of inclusions are 1 to 27 microns in diameter, while the volume fraction of all sizes is 300 to 600 parts per million. The SEM-EDS (Scanning Electron Microscopy / Energy Dispersive X-Ray Spectroscopy) analysis of the inclusions has revealed that they all contain oxygen and some contain carbon. At present, under sponsorship of NASA and CASIS (Center for the Advancement of Science in Space), we are conducting ground-based experiments with InI in preparation for the flight experiments to be conducted in the SUBSA (Solidification Using a Baffle in Sealed Ampoules) furnace in the Microgravity Science Glovebox at the International Space Station, planned for the summer/fall of 2017. Earth-based experiments include melt and vapor growth conducted in the SUBSA ground unit, measurements of the volumetric expansion coefficient of the melt, and measurements of the wetting angle of molten InI. Finite element modeling has been conducted to optimize the design of the flight ampoules. Alloying with Tl and Ga has given promising results.

During the past 40 years, dozens of semiconductor crystal growth experiments have been conducted in space laboratories. The subsequent analysis of the space-grown crystals revealed (i) that weak convection existed in virtually all melt-growth experiments, (ii) de-wetting significantly reduced the level of stress-induced defects, and (iii) particularly encouraging results were obtained in vapor-growth experiments. In 2002, following a decade of ground based research in growing doped Ge and GaSb crystals, a series of crystal growth experiments was performed at the ISS, within the SUBSA (Solidification Using a Baffle in Sealed Ampoules) investigation. Te- and Zn-doped InSb crystals were grown from the melt. The specially designed furnace provided a side-view of the melt and precise seeding measurement of the growth rate. At present, under sponsorship of CASIS (Center for the Advancement of Science in Space, www.iss-casis.org), we are conducting ground-based experiments with indium mono-iodide (InI) in preparation for the "SUBSA II" ISS investigation, planned for 2017. The experiments include: i) Horizontal Bridgman (HB) growth and ii) Vapor Transport (VT) growth. Finite element modeling will also be conducted, to optimize the design of the flight ampoules, for vapor and melt growth.

Three Zn-doped InSb crystals were directionally solidified under microgravity conditions at the International Space Station (ISS) Alpha. The distribution of the Zn was measured using SIMS. A short diffusion-controlled transient, typical for systems with k greater than 1 was demonstrated. Static pressure of approximately 4000 N/m2 was imposed on the melt, to prevent bubble formation and dewetting. Still, partial de-wetting has occurred in one experiment, and apparently has disturbed the diffusive transport of Zn in the melt.

In directional solidification, dewetting is characterized by the lack of contact between the crystal and the crucible walls, due to the existence of a liquid meniscus at the level of the solid-liquid interface. This creates a gap of a few tens of micrometers between the crystal and the crucible. One of the immediate consequences of this phenomenon is the dramatic improvement of the quality of the crystal. This improvement is partly due to the modification of the solid-liquid interface curvature and partly to the absence of sticking and spurious nucleation at the crystal-crucible interface. Dewetting has been, commonly observed during the growth of semiconductors in crucibles under microgravity conditions where it appears to be very stable: the gap between the crystal and the crucible remains constant along several centimetres of growth. The physical models of the phenomenon are well established and they predict that dewetting should not occur in microgravity, if sufficient static pressure is imposed on the melt, pushing it towards the crucible. We present the results of InSb(Zn) solidification experiments conducted at the International Space Station (ISS) where, in spite of a spring exerting a pressure on the liquid, partial dewetting did occur. This surprising result is discussed in terms of force exerted .by the spring on the liquid and of possibility that the spring did not work properly. Furthermore, it appears that the segregation of the Zn was not affected by the occurrence of the dewetting. The data suggest that there was no significant interference of convection with segregation of Zn in InSb.

In 2002, within the SUBSA (Solidification Using a Baffle in Sealed Ampoules) investigation, seven doped InSb crystals were grown in microgravity at the International Space Station. The key goals of the SUBSA investigation are: (a) to clarify the origin of the melt convection in space laboratories; (b) to reduce melt convection to the level which allows reproducible diffusion-controlled segregation; (e) to explore the submerged baffle process and liquid encapsulation in microgravity. 30 crystal growth experiments were conducted in the ground unit, to optimize the design of flight ampoules and to test the transparent SUBSA furnace developed by TecMasters Inc. The specially designed furnace, allowed observation of the crystal growth process (melting, seeding, motion of the solid-liquid interface, etc.). In the summer of 2002, eight crystal growth experiments were conducted in the Microgravity Science Glovebox (MSG) facility at the ISS. Four Te-doped (k = 0.5) and three Zn-doped (k2.9) crystals were grown on undoped seeds. In one experiment, we were not able to seed and grow. The seven grown crystals were sectioned and analyzed using SIMS. The design of the SUBSA ampoules, the segregation data and the video images obtained during the SUBSA flight experiments will be presented and discussed.

Solidification Using a Baffle in Sealed Ampoules (SUBSA) will be the first materials science experiment conducted in the Microgravity Science Glovebox (MSG) Facility at the International Space Station (ISS) Alpha. The launch is schedule for May 31, 2002. Using the specially developed furnace, 10 Te and Zn-doped single crystals of InSb will be directionally solidified in microgravity. A key goal of the SUBSA investigation is to (i) clarify the origin of the melt motion in space laboratories and (ii) to reduce the magnitude of the melt motion to the point that it does not interfere with the transport phenomena. These goals will be accomplished through a special ampoule and furnace design. A disk-shaped baffle, positioned close to the freezing front, is used to reduce melt motion. Furthermore, the solidification will be visualized by using a transparent furnace, with a video camera, continuously sending images to the earth. This allows detection of bubbles and melt de-wetting that could cause surface tension driven convection. In preparation for the space experiments, 30 ground-based experiments were conducted. The results of ground based tests and numerical modeling will be presented. Based on numerical modeling, 12 mm 1D silica ampoules were selected. The small diameter ampoule favors closer placement of the baffle to the interface, without excessive radial segregation caused by forced convection while providing more damping of natural convection. The parts in the silica ampoule include 2 carbon springs made by Energy Science Laboratories, Inc., a pyrocarbon-coated graphite cylinder, pyrocarbon-coated graphite a baffle with the shaft and the InSb charge with the seed crystal grown by W.A. Bonner of Crystallod Inc.

The composition of semiconductor crystals produced in space by conventional melt-growth processes (directional solidification and zone melting) is affected by minute levels of residual micro-acceleration, which causes natural convection. The residual acceleration has random magnitude, direction and frequency. Therefore, the velocity field in the melt is apriori unpredictable. As a result, the composition of the crystals grown in space can not be predicted and reproduced. The method for directional solidification with a submerged heater or a baffle was developed under NASA sponsorship. The disk-shaped baffle acts as a partition, creating a small melt zone at the solid-liquid interface. As a result, in ground based experiment the level of buoyancy-driven convection at the interface is significantly reduced. In several experiments with Te-doped GaSb, nearly diffusion controlled segregation was achieved.

The character of natural buoyant convection in rigidly contained inhomogeneous fluids can be drastically altered by vibrating the container. Vibrations are expected to play a crucial influence on heat and mass transfer onboard the International Space Station (ISS). It is becoming evident that substantial vibrations will exist on the ISS in the wide frequency spectrum. In general, vibrational flows are very complex and governed by many parameters. In many terrestrial crystal growth situations, convective transport of heat and constituent components is dominated by buoyancy driven convection arising from compositional and thermal gradients. Thus, it may be concluded that vibro-convective flow can potentially be used to influence and even control transport in some crystal growth situations.

One of the most promising approaches for the reduction of convection during the crystal growth of conductive melts (semiconductor crystals) is the application of magnetic fields. Current technology allows the experimentation with very intense static fields (up to 80 KGauss) for which nearly convection free results are expected from simple scaling analysis in stabilized systems (vertical Bridgman method with axial magnetic field). However, controversial experimental results were obtained. The computational methods are, therefore, a fundamental tool in the understanding of the phenomena accounting during the solidification of semiconductor materials. Moreover, effects like the bending of the isomagnetic lines, different aspect ratios and misalignments between the direction of the gravity and magnetic field vectors can not be analyzed with analytical methods. The earliest numerical results showed controversial conclusions and are not able to explain the experimental results. Although the generated flows are extremely low, the computational task is a complicated because of the thin boundary layers. That is one of the reasons for the discrepancy in the results that numerical studies reported. Modeling of these magnetically damped crystal growth experiments requires advanced numerical methods. We used, for comparison, three different approaches to obtain the solution of the problem of thermal convection flows: (1) Spectral method in spectral superelement implementation, (2) Finite element method with regularization for boundary layers, (3) Multiquadric method, a novel method with global radial basis functions, that is proven to have exponential convergence. The results obtained by these three methods are presented for a wide region of Rayleigh and Hartman numbers. Comparison and discussion of accuracy, efficiency, reliability and agreement with experimental results will be presented as well.