|
The efficiency of photovoltaic energy conversion for single junction silicon solar-cells is limited to 30%. This limit can be shifted up to 60% by using multi-gap solar-cell stacks. Usually we need different solids or composites to get a gap width variation. An alternative way to realize this may be the exploitation of quantum size effects which leads to a change of the fundamental optical absorption edge. Candidates are Si/SiO2 superlattices with different thickness sequences of the Si quantum layers consisting of homogeneous crystalline Si layers with abrupt Si/SiO2 interfaces. Both should have low densities of states. Sample preparation is performed by thermal deposition of amorphous silicon layers (<10 nm) onto quartz glass (SiO2) substrates under ultrahigh vacuum conditions. Subsequent oxidation with atomic oxygen at substrate temperatures of 600°C leads to the formation of 1-2 nm thick SiO2 layers with nearly no sub-oxide species at the interfaces. Annealing of the a-Si/SiO2 structure is necessary for Si crystallization and the improvement of the interface quality. Crystallization of sandwiched Si quantum layers requires higher temperatures (1000°C) than bulk a-Si and results in a nano-crystalline structure with some amorphous fraction, detected by Raman spectra. Photoconduction measurements (PC) were performed at single and multiple Si/SiO2 quantum wells where the Si layers were provided with two evaporated Mg contacts. This allows to measure the spectral dependence of internal quantum efficiency Yint of photoconductivity which is proportional to the product of absorption coefficient, majority carrier mobility and their lifetime. Yield values down to 10-6 were measured. The spectral dependence correlates well with the structural change of the a-Si layer after annealing and the resulting nano-crystalline structure with some amorphous fraction. The onset of Yint shifts from 2eV (a-Si phase) to 3.6 eV (c-Si phase). To which extent the low yield values can be increased by reduction of interface recombination and/or enhancement of mobility by reducing the scattering centre concentration becomes the next interesting question. The answer will be decisive for the true impact of such structures on photovoltaics.
|