Higher emitter quality by reducing inactive phosphorus

In this work, we study the influence of the amount of active phosphorus atoms in the emitter region of silicon N+/P solar cells. After emitter formation by thermal diffusion, a part of phosphorus atoms in the vicinity of the surface is electrically inactive. This results in a high emitter saturation current density and in a poor quantum efficiency in the blue wavelength range. This emitter region is usually known as dead layer. We first introduce a simple method to determine the percentage of electrically active phosphorus atoms in emitters obtained from a low pressure POCl3 diffusion furnace. Different phosphorus diffusions are performed using different parameters (temperature, time, POCl3/O2 ratio, and preoxidation time). We then investigate the effect of the fraction of active phosphorus on the doping level and on the emitter passivation. We show that diffusion parameters can greatly influence the active phosphorus percentage, particularly the POCl3/O2 ratio or the addition of a preoxidation step. Concerning the emitter passivation, a linear variation of the emitter saturation current density as a function of the active phosphorus percentage is observed. Using appropriate POCl3/O2 ratio or preoxidation time, we obtain an emitter saturation current density around 100 fA/cm2.

► Inactive phosphorus amount in the emitter can be reduced by tailoring the diffusion parameters.
► A linear variation of the emitter saturation current density as a function of the active phosphorus percentage is observed.
► With the addition of the preoxidation step, emitter saturation current density below 100 fA/cm2 can be obtained.