Gallium Doped Silicon for PERC Solar Cells—Carrier Lifetime Potential and Instability

Mainstream solar cell production now preferentially uses gallium doped silicon wafers. Gallium doped substrates offer inherently better carrier lifetime stability than boron doped ones, without a need for post-cell production stabilisation processes. The transition to gallium occurred very rapidly, as no changes to cell lines were required. Although the benefits are clear, the properties of the material are relatively poorly understood, and a set of specific research issues specific to gallium doped silicon solar cells has started to emerge.<br/>The first issue concerns potential performance limitations specific to gallium doped silicon. In terms of carrier lifetime – the most important property for solar cells – the highest reported historical values were typically ~1 ms. We have evaluated the performance limit of modern gallium doped silicon materials with state-of-the-art surface passivation [1]. We find this to result in considerably higher lifetimes than usually reported, with lifetimes &gt; 9 ms achieved in some cases. The lifetimes we measure in ungettered gallium doped silicon are similar to those for boron doped silicon which had undergone both gettering and boron-oxygen stabilisation processes. As such, equivalent high performance can be achieved with gallium doped silicon substrates without the need for stabilisation processing.<br/>A second issue is the stability of gallium doped solar cells, particularly those of the market-dominating passivated emitter and rear cell (PERC) variety. We use a photoluminescence imaging proxy method to study the performance of commercially produced cells under illumination for very long times (&gt; 3000 h in some cases). Our published reports have shown gallium doped PERC cells undergo a degradation in performance, which then recovers, reminiscent of light and elevated temperature induced degradation (LeTID) [1, 2]. The magnitude of the degradation with gallium doping is less than in equivalent boron doped PERC cells. Stabilisation processes of the kind used to mitigate boron-oxygen light induced degradation do not appear to influence behaviour in our gallium doped cells. A dark pre-anneal (up to 300 °C) can influence the LeTID-like behaviour, although the behaviour has been found to be inconsistent between batches, highlighting the need for further controlled studies. Ongoing work aims to address the dependence of degradation and recovery on cell substrate resistivity, with factors such as unintentional boron doping also being taken into account.<br/>The final issue we address is carrier lifetime changes which occur in illuminated gallium doped silicon wafers. Historically, lifetimes in gallium doped silicon wafers are assumed to be stable, however it is now clear that this is not always the case. Kwapil <i>et al.</i> have recently reported LeTID behaviour in gallium doped wafers [3] and interestingly the degradation has a larger magnitude when the samples are exposed to relatively low levels of illumination (0.1 Suns). We will report new data which support these findings, using samples with different levels of gallium.<br/>In summary, our work generally confirms the positive effects of a transition from B to Ga doping for solar cells. However, given the recent switch, understanding and mitigating these newly found degradation processes is one of the most pressing issues in silicon photovoltaic research at present.<br/>[1] N. E. Grant, P. P. Altermatt, T. Niewelt, R. Post, W. Kwapil, M. C. Schubert, J. D. Murphy, Solar RRL, <b>5</b> 2000754 (2021), doi: 10.1002/solr.202000754.<br/>[2] N. E. Grant, J. R. Scowcroft, A. I. Pointon, M. Al-Amin, P. P. Altermatt, J. D. Murphy, Solar Energy Materials & Solar Cells, <b>206 </b>110299 (2020), doi: 10.1016/j.solmat.2019.110299.<br/>[3] W. Kwapil, J. Dalke, R. Post, T. Niewelt, Solar RRL, <b>5</b> 2100147 (2021), doi: 10.1002/solr.202100147.