Computer Based Optimization of Amorphous Silicon Materials for Solar Cells

Grant # 93-01-02
Principal Investigator: Rana Biswas
Co-principal investigator: C. M. Soukoulis
Organizations: Microelectronics Research Center, ASC-I, Iowa State University
Participants: Q. Li
Current period: 7/1/93 – 5/31/97
Technical Area: Renewable Energy

Background
Solar energy offers a clean and effectively inexhaustible source of energy. A technically and economically-feasible material for solar cells is thin film hydrogenated amorphous silicon (a-Si:H). Solar cells fabricated with amorphous silicon have the advantage of low cost and well- established processing know-how. The main problem is that these solar cells degrade their performance in sunlight with their efficiency degrading up to 20% under prolonged exposure.

The causes for this degradation effect is unclear. Intensive research in the U.S. focuses on new ways to reduce the degradation effect in solar cells and the amorphous silicon films. The goals being investigated model, and establish the causes of light-induced degradation in amorphous silicon films and thus help optimize the quality of these films.

Our project is part of the metastability and middle-band gap team organized by NREL (National Renewable Energy Laboratory) and EPRI (Electric Power Research Institute). Team members are from NREL, UCLA, University of Oregon, ISU, NIST(Boulder, CO), University of North Carolina, and Penn State University together with industrial members from United Solar Systems, Solarex and Energy Conversion Devices. The goal of these teams is to achieve a stable amorphous silicon module with 15% efficiency. EPRI jointly funds our research and our program is a high-priority part of a long-range research thrust. We are actively collaborating with NREL scientists on ongoing research and work with other team members.

Work to Date
We have investigated the atomic mechanisms underlying the degradation effect in amorphous silicon using the technique of molecular dynamics which simulates the behavior of a collection of Si and H atoms at any temperature. Computer-generated models of a-Si:H are being employed. During the first phase of our project, we developed new tight-binding models for describing the behavior of silicon and hydrogen atoms that incorporate the electronic properties of this system. We have fitted the parameters of the model to data on crystalline silicon and the known properties of hydrogen in crystalline silicon.

We applied these molecular dynamics techniques to model defect creation in hydrogenated amorphous silicon models. It is recognized that a large concentration of hydrogen (5% or more) is needed for growing device-quality material. The new finding of our research is that the hydrogen also plays a key role in causing the light-induced degradation effect. This is caused by the relocation of hydrogen over atomic length scales and motion of hydrogen in the film, creating metastable defects.

The specific mechanism we found for hydrogen rebonding is that a Si-H bond is broken and the light-induced degradation effects in solar cells, limiting the efficiency. The density of such defects is controlled by the distribution of silicon-silicon bond lengths, a new variable that characterizes the quality of the material. We are also studying the motion of H through such an amorphous network.

The basic finding is that the degradation effect in amorphous silicon is an intrinsic effect intimately connected to the underlying structure. It is not connected to the presence of trace impurities or contaminants, so improving the purity of the material will not have a significant effect. The degradation effect is directly caused by hydrogen. Our research suggests that improving solar cells by reducing the degradation is possible by controlling the hydrogen content and the way in which hydrogen is bonded in the network. Presently, solar cell manufacturers are manipulating the way hydrogen is incorporated into the network. In addition, the degradation effect will be reduced by increasing the ordering of the network and making it closer to crystalline rather than amorphous. Replacing hydrogen by deuterium may also generate significant improvement.

Our research results have been published in the Physical Review, Applied Physics Letters and proceedings of the Materials Research Society. Presentations to industry and University groups have been made at the EPRI/NREL review meetings, Materials Research Society and American Physical Society meetings.