Passivation of thin film silicon solar cells

Abstract

Passivation of electronic defects is a necessary step in production of solar cells based on polycrystalline silicon thin film on glass. Standard passivation method is plasma hydrogenation, which however, represents the second most expensive production step. We used solid phase crystallized (SPC) silicon samples to explore an alternative cheaper passivation approach using annealing in water vapour. Open-circuit voltage VOC measured by Suns-VOC method was the key parameter to determine success of the treatment. Annealing of SPC Si thin film solar cells in water vapour was explored in the temperature range from 145°C to 650°C under steam pressure from atmospheric pressure to 1.0 MPa at the exposure times of 5-225 minutes. For this purpose a special passivation chamber enabling an independent control of a sample temperature and steam pressure was designed and built. We achieved the best VOC of 360 mV (from the starting 220 mV) demonstrating the annealing in water vapour as a possible low cost alternative passivation method. Some SPC poly-Si solar cells were annealed in hydrogen gas, a mixture of steam and hydrogen gas, or a mixture of steam and oxygen. These experiments uncovered that neither hydrogen gas nor the mixtures are able to passivate silicon as effectively as water vapour. While the plasma hydrogenation represents a saturation of silicon dangling bonds by hydrogen radicals, annealing in water vapour is an oxidation of silicon and hydrogen acts just as a catalyst. On the basis of the realized experiments and a review of scientific literature, principles of the water vapour passivation were described, explained and presented as a model of steam passivation. Still, the plasma hydrogenation can achieve better VOC for the same samples. We achieved the best VOC of 497 mV for the same SPC samples. This value resulted from optimization of the plasma hydrogenation parameters at the Helmholtz-Zentrum Berlin during which we suggested to use 1) the higher hydrogen pressure of 300-1,000 Pa in comparison with a commonly used 100 Pa, 2) the longer exposure time of 15-20 minutes, and mainly 3) to keep the usually omitted bias voltage Vbias constant during the whole passivation process up to the plasma termination. Since all experiments in the hydrogen plasma were realized as a closed system without a hydrogen flux with very satisfying results, the generally accepted necessity to run the plasma hydrogenation process with a continuous hydrogen flux was called into question. The development of the crystalline silicon on glass solar cells led to replacement of SPC by liquid phase crystallized (LPC) for which VOC over 600 mV can be achieved. We tested the effect of passivation for LPC poly-Si samples crystallized either by a laser or an electron beam (with the SiOx diffusion barrier deposited either by plasma enhanced chemical vapour deposition or by physical vapour deposition). In these experiments hydrogen plasma increased the VOC from the typical value of 535 mV to 570 mV for most of the parameter values. Some SPC Si samples treated in the hydrogen plasma were analyzed by both Suns-VOC method and also optical pump transient terahertz probe spectroscopy which represents optical method for measurement of photogenerated carrier transport at ultrafast time scales. While each of these methods characterizes the solar cell in a different state, a clear correlation between VOC and the lifetime of charge carriers was observed. Terahertz spectroscopy analyzes the sample before the photogenerated charge carriers can be redistributed by a space charge region and therefore VOC is not built up yet. In contrary to this, Suns-VOC method characterizes the cell at a quasi-steady state, when VOC is already built up.