Impact of dopant compensation on the electrical properties of silicon for solar cells applications

Abstract

This thesis aims at understanding the mechanisms limiting the efficiency of compensated silicon solar cells (containing boron and phosphorus in the bulk). Such dopant compensation is common in solar grade materials, especially in silicon from the metallurgical route, and can potentially lead to a degradation of the materials electronic properties. We experimentally show that a thermal oxidation can create an n-type layer at the surface of compensated p-type silicon. This n-type layer is further shown to interfere with device performance and material characterization. We investigate the impact of compensation on the minority carrier lifetime, in particular for recombination through defects. Metastable defects such as chromium-boron pairs and the boron-oxygen defect are shown to degrade the lifetime of compensated n-type silicon. The boron-oxygen defect in compensated n-type silicon is then experimentally investigated. It is shown that if not mitigated, the boron-oxygen defect leads to a strong reduction in implied VOC. The defect is also shown to be fundamentally different in compensated n-type silicon compared to p-type silicon. Its concentration does not depend on the net doping and its recombination activity is dominated by a shallow defect rather than a deep defect. Through a theoretical investigation, we show that the carrier mobility is also affected by compensation. Both theory and experiments confirm that the mobility is reduced by the combined presence of acceptors and donors. Compensation not only increases the amount of ionized impurities and decreases the amount of free carriers, it also affects the scattering cross section of ionized impurities and free carriers. Theoretical calculations show a relatively weak influence of the compensating impurities on the mobility. However experimental results suggest a stronger influence of compensating impurities. This results in mobilities slightly lower than predicted by advanced model such as Klaassen's model. A new method to measure the sum of the majority and minority carrier mobility in silicon is introduced. Measurement of the influence of dopant density, injected carriers and temperature on the mobility sum are made and compared to data available in the literature.