Advances in Sensing Compound Semiconductor Crystal Growth

"Monitoring compound semiconductor crystal growth using multifrequency eddy currents for sensing illustrates some of the extreme problems of measurement and interpretation that are not often seen in metallic systems. Using sensor arrays and scanning sensors have produced a qualitative and quantitative understanding of some of the high temperature electrical conductivity mechanisms in several compound systems. A detailed review of the CdTe/CdZnTe data is presented showing some of the complexity in the data as opposed to what is found in elemental semiconductors.HistoryThe most common goal in the use of eddy current measurement in crystal growth is to deter-mine the curvature of the liquid solid interface during growth to test if the heat transfer estimates of the thermal design were adequate for single crystal growth[l]. This aim could be met but represented only a small fraction of the data that was produced observing a growth cycle. A secondary consideration was what could be discovered in the recorded data about the growth and subsequent heat treatment as the boule cooled. This is usually a more difficult task because a crystal growth process is not a controlled experiment so much as a process that continually sees modification. Using a sensor arrays or a scanning loop has produced a qualitative and quantitative basis for understanding some of the high temperature electrical conductivity mechanisms. The implication of the these measurements of electrical conductivity variations are important in that they reflect the complex defect states that will occur on cooling and heat treatment following crystal growth. In contrast to the complex behavior of the conductivity of these compound semiconductors in the solid and liquid state we have the very well characterized system silicon to compare. Initially the interest on using eddy currents to monitor compound semiconductors came from G. Westphal at the Texas; Instruments Company in 1987 to monitor aspects of growth and compounding of GaAs. This was followed by work on solution growth of HgCdTe with D. Witter. The initial interest in the use of eddy cur-rents for monitoring CdTe crystal growth came from Prof. Wilcox at Clarkson University in the practical interest to determine and control the crystal interface curvature during vertical brigdeman growth. The first actual work was done by one of his former students. G. Neuge-bauer at the II-VI Corporation which was carried on by G. Rosen at Clarkson and R. Shetty at II-VI Corporation and with A. Socha at Johnson-Matthey Electronics. The work on CdTe extends over four years and yielded sufficient data to allow one to evaluate the utility of the information produced as compared to what could be obtained from monitoring an elemental semiconductor such as silicon."