Device physics of hydrogenated amorphous silicon solar cells

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




Eric Allan Schiff


Hydrogenated, Amorphous silicon, Solar cells

Subject Categories



This dissertation reports measurements on and modeling of hydrogenated amorphous silicon (a-Si:H) nip solar cells. Cells with thicknesses from 200-900 nm were prepared at United Solar Ovonic LLC. The current density-voltage ( J-V ) relations were measured under laser illumination (685 nm wavelength, up to 200 mW/cm2) over the temperature range 240 K--350 K. The changes in the cells' open-circuit voltage during extended laser illumination (light-soaking) were measured, as were the cell properties in several light-soaked states. The J-V properties of cells in their as-deposited and light-soaked states converge at low-temperatures. Electromodulation spectra for the cells were also measured over the range 240 K--350 K to determine the temperature-dependent bandgap.

These experimental results were compared to computer calculations of J-V relations using the AMPS (©Pennsylvania State University) computer code. Bandtail parameters (for electron and hole mobility and recombination) were consistent with published drift-mobility and transient photocurrent measurements on a-Si:H. The open-circuit voltage and power density measurements on as-deposited cells, as a function of temperature and thickness, were predicted well. The calculations support a general "hole mobility limited" approach to analyzing a-Si:H solar cells, and indicate that the doped electrode layers, the as-deposited density of dangling bonds, and the electron mobility are of secondary importance to as-deposited cells.

For light-soaked a-Si:H solar cells, incorporation of a density of dangling bonds in the computer calculations accounted satisfactorily for the power and open-circuit voltage measurements, including the low-temperature convergence effect. The calculations indicate that, in the light-soaked state at room-temperature, electron recombination is split nearly evenly between holes trapped in the valence bandtail and holes trapped on dangling bonds. The result supports Stutzmann, Jackson, and Tsai's 1985 conjecture that dangling bond creation results only from bandtail recombination events.

We compared the predictions of the hydrogen-collision model proposed by Branz with the kinetics of the open-circuit voltage as light-soaking progressed. We obtained satisfactory agreement for the initial phases of light-soaking with the conjecture that only bandtail recombination leads to dangling bond creation, and the computer calculations for this recombination channel's diminishment in the cell as the dangling bond density grows.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.