supercomputing institute research bulletin online

Volume 15 Number 3

July 1999


Conductivity of Molten Semiconductors
Building of Protein Structures
Diamond Growth
Turbulent Flow and Heat Transfer in Propulsion Systems
Interns
Future Symposia
Colloquium Series
Special Seminars
Visitors
Research Reports

admium telluride (CdTe) and its alloys form an important class of technological materials used in a wide array of electro-optic devices that include high-performance infrared detectors and room temperature radiation detectors. However, the melt growth of large crystals needed for these applications has proven to be very difficult. To aid the understanding of inherent difficulties in growth, a team of researchers at the University of Minnesota has been investigating the properties of molten CdTe. These investigators include Fellows of the Supercomputing Institute, Professors Jeff Derby and Jim Chelikowsky of the Chemical Engineering and Materials Science Department and their graduate student researchers Vitaliy Godlevsky and Manish Jain.

Chelikowsky1.jpg
Figure 1: A snapshot of liquid Cadmium Telluride (CdTe) showing the atomic structure and distribution of electrons. The gray spheres localized around the Te atoms correspond to surfaces of constant electron density. This illustrates the ionic character of the liquid state‹most electrons are found near Te atoms and not Cd atoms. Only bonds between Te atoms are illustrated. The arrangement of Te atoms in the melt often resemble the chain structures found in elemental Te solids.
Simulating the properties of liquid semiconductors is a difficult problem since liquids possess no inherent symmetry or long range order. They can only be characterized in a statistical sense by sampling an ensemble of states over a long computational time period. University of Minnesota researchers have developed first-principles methods for predicting the electronic and structural properties of liquids. To model a liquid ensemble, a supercell geometry with periodic boundary conditions is defined, and the temperature of the liquid is controlled via a fictive heat bath. Unlike traditional molecular dynamics methods, the trajectories of particles within the liquid are calculated using fully quantum mechanical interatomic forces. By following trajectories over a long computational time period, statistical properties of the liquid can be extracted.

Chelikowsky2.jpg
Figure 2: The frequency dependent electrical conductivity of CdTe (dashed line) and GaAs (solid line) liquids. The dc-conductivity of CdTe in the liquid state is very low compared to GaAs.
In Figure 1, a snap-shot of liquid CdTe is illustrated. Owing to the quantum nature of the simulation, the electronic states are known at each time step; here, the distribution of the electrons around the nuclei is indicated. Most of the electronic charge remains localized on the Te atoms, corresponding to a very ionic configuration with significant electron transfer. Since the electrons are fairly localized, CdTe remains semiconductive in the liquid state. This behavior contrasts with what is exhibited for most semiconductors such as silicon, germanium, and gallium arsenide (GaAs), which are metallic in the melt. Figure 2 compares the computed electrical conductivity versus frequency for liquid CdTe and GaAs. The conductivity in the dc-limit for CdTe is approximately two orders of magnitude smaller than that for GaAs-a result consistent with experimental measurements.

Work by this group is continuing with an emphasis on diffusion properties within the liquid. For example, the diffusion of zinc in CdTe is an important quantity for modeling macroscopic growth, but it is difficult to measure. With the simulation techniques developed in this program, this becomes a straightforward calculation.

next article this issue

 
HOME | BULLETINS | CONTACT US| NEXT ARTICLE | THIS ISSUE

 

This information is available in alternative formats upon request by individuals with disabilities. Please send email to alt-format@msi.umn.edu or call 612-624-0528.
 


URL: http://
This page last modified on  
Website related questions or problems should be directed to webmaster@msi.umn.edu
The Supercomputing Institute does not collect personal information on visitors to our website. For the University of Minnesota policy, see www.privacy.umn.edu.