Surface passivation and characterisation of crystalline silicon by wet chemical treatments

Abstract

This thesis concerns the development of low temperature surface passivation films by wet chemical treatments. The low temperatue processing minimises the energy requirements, and the excellent surface passivation is adequate for achieving high solar cell efficiencies, both of which can contribute to a reduction in the cost of photovoltaics and its drive towards grid parity. Firstly we describe two refinements to the analysis of photoconductance (PC) and capacitance-voltage (CV) measurements. The first refinement concerns accurate measurements of the surface recombination velocity S by accounting for non-uniform carrier profiles during a PC measurement. Contrary to other methods, our technique permits an accurate determination of S (<5% uncertainty), provided the bulk lifetime is known. The second refinement concerns accurate measurements of the interface defect density, oxide charge and the oxide thickness from a CV measurement by accounting for leakage current through the dielectric film, inductance and series resistance. We develop a procedure to temporarily attain a very high level of surface passivation for silicon wafers immersed in hydrofluoric acid. When applied during a photoconductance measurement, the procedure permits an accurate assessment of the bulk lifetime, even for high-lifetime samples ({u03C4}bulk> 5 ms) that are otherwise very difficult to measure. Here we show that the HF passivation is greatly enhanced by illuminating the wafers just prior to a measurement, and that the HF passivation depends critically on the surface preparation, where the best passivation is attained after etching the wafers in tetramethylammonium hydroxide. We assess the level of passivation for a range of HF concentrations and wafer resistivities, and we demonstrate that S < 5 cm/s can be attained on 0.8-1000 ohms-cm n- and p-type silicon wafers. We next investigate the passivation attained by growing thin (<3 nm) silicon dioxide films in nitric acid at low temperature (100 degrees). With no anneal, the films offer very little passivation, however by subjecting the nitric silicon dioxide to a 1050 degree nitrogen anneal, S of <30 cm/s is attained. We then investigate a two-step nitric acid oxidation method, previously demonstrated to grow thicker silicon dioxide layers (>3 nm). X-ray photoelectron spectroscopy and CV measurements revealed that the nitric silicon dioxide do not increase with longer immersion times, but rather decrease from 2.4 nm to 2.0 nm. The passivation achieved by this alternative method does reduce S <30 cm/s, however high temperature annealing is still required. Post annealing of the nitric silicon dioxide films, we show that a correlation exists between S and the bulk lifetime, where a lower bulk lifetime correlates to a higher S. Finally, we show that low S of <20 cm/s can be achieved by a DC/AC anodic oxidation procedure (using nitric acid) when the silicon dioxide is subject to a low temperature (400 degree) anneal in oxygen and FG. We demonstrate that the nitric acid purity has a substantial impact on the repeatability of S. When the nitric acid impurity level is in the ppm range, S of 10-1000 cm/s results. However when the impurity level is reduced to ppb, S ranges between 15-30 cm/s. We show that these anodic silicon dioxide films degrade when exposed to the ambient air, however from our investigation, we find that water moisture is the main source of degradation observed in our anodic silicon dioxide films.