Single-chamber process development of microcrystalline sicicon solar cells and high-rate deposited intrinsic layers
Thèse de doctorat : Université de Neuchâtel, 2005 ; 1835.
The "Micromorph" tandem solar cell concept consisting of an amorphous and a microcrystalline silicon solar cell is considered to be one of the most promising concepts for the next solar cell generation. To translate this concept into action, efficient industrial processes have to be developed. Thereby important issues have to be considered, such as e.g. the development of an economical... MoreAdd to personal list
- The "Micromorph" tandem solar cell concept consisting of an amorphous and a microcrystalline silicon solar cell is considered to be one of the most promising concepts for the next solar cell generation. To translate this concept into action, efficient industrial processes have to be developed. Thereby important issues have to be considered, such as e.g. the development of an economical single-chamber process and the achievement of high deposition rates for intrinsic microcrystalline silicon ( c Si:H). The present thesis focuses on the single-chamber process. The development was done under three aspects: the aspect of layer optimisation with respect to their structural and electrical characteristics, the chamber-design optimisation and the solar-cell tuning. For the present single-chamber PECVD deposition system microcrystalline p doped c Si:H layers could be obtained using trimethylboron (TMB) doping gas. At low source-gas concentration values, a plasma excitation frequency as high as 110.0 MHz was found to be necessary to attain a sufficiently high electrical conductivity and good crystallinity for these p-doped c-Si:H layers. The search for alternative methods led to the discovery of a crucial pre-deposition plasma treatment with H2 as well as with CO2 on the LPCVD ZnO layer; this treatment has a beneficial effect on crystallisation during the subsequent growth of the overlying p-doped c-Si:H layers. In order to obtain high deposition rates for intrinsic c-Si:H layers, pressures up to 8 mbar were used. When working in the high-pressure regime, deposition uniformity and powder formation is critical. The use of a novel cylindrical design of the plasma confinement box - the electrode is designed as a box that defines the space of the bulk plasma- led to a homogeneous deposition. Thereby, the deposition rate on ZnO could be increased up to about 28 Å/s. At the same time working at an inter-electrode distance dgap of 9.5 mm made polysilicon (powder) formation even disappear. In our single-chamber PECVD deposition system the use of a standard deposition process (p-doped c-Si:H/ intrinsic c-Si:H/ n-doped c-Si:H ) led to solar cells with poor characteristics (VOC << 500 mV & FF< 0.5) accompanied by an annealing initiated metastable phenomenon "JSC-degradation ? JSC-regeneration". Only the introduction of a "SF6/O2-cleaning & a-Si:H covering layer" as chamber treatment after the deposition of the p-doped layer allowed us to overcome the boron cross-contamination between p layer and subsequent i-layer - this was proved by SIMS. We suspect that boron contamination is the cause of the metastability mentioned above, via Fei+ centres. Hereby, Fei+ irons are activated by the annealing process and act then as recombination centres within the intrinsic layer. The best c-Si:H solar cell with a typical microcrystalline VOC of 516 mV, a FF of 0.685 and a JSC of 18.5 mA/cm2 leading to an efficiency of 6.5 %. SIMS-Analysis allowed us to quantify the boron contamination in the intrinsic layer for the various cases studied. An equilibrium reaction is suspected to changing by annealing an "inactive" iron atom within a FeB complex into a dissociated, very strongly active Fei+ ion. Finally, a novel structuring method by ZnO lift-off led to a substantial increase in the uniformity of the solar cell characteristics (VOC and FF) compared to the previous standard structuring method.