Heavily doped Si:P emitters of crystalline Si solar cells: recombination due to phosphorus precipitation

Abstract

1 Introduction The emitter of crystalline silicon solar cells is usually formed in mass production by flowing POCl 3 through a furnace, which creates a phosphorussilicate glass (PSG) layer at the surfaces of the p-type wafers, from where phosphorus diffuses into the silicon. Only very recently has the phosphorus concentration in the PSG layer been measured [1]. It is firstly substantially higher than the solubility of P in Si, and secondly it changes only slightly with POCl 3 flow or other process parameters such as temperature. In this paper, a consequence of these recent findings is identified: the high flow of P into Si causes a far larger amount of (extrinsic) Shockley-Read-Hall (SRH) recombination than (intrinsic) Auger recombination. This implies that todays emitters are not Auger limited, but are instead SRH limited, and explains the large discrepancies between measured and simulated saturation current densities, J 0e , observed persistently over the last two decades in industrially fabricated Si solar cells. This clarification correlates strongly with recent cell efficiency improvements by reducing inactive phosphorus

Recombination processes in passivated boron-implanted black silicon emitters

“…23,36,43,[46][47][48][49][50] The performance of textured emitters can be simulated in two dimensions, which thus allows using two-dimensional bSi doping profiles generated by a separate process simulator. 48 However, we show that, implying some approximations in the bSi regions, useful insight on bSi emitter recombination mechanisms can be obtained using the PC1D freeware.…”

“Zero-charge” SiO2/Al2O3 stacks for the simultaneous passivation of n+ and p+ doped silicon surfaces by atomic layer deposition

“…Simulations using the free-ware program EDNA [41], using Fermi-Dirac statistics, the Auger parameterization from Richter et al [42] and the band-gap narrowing model of Schenk [43], indicate that the Auger limit of the n þ region is J 0n þ , Auger ¼16 fA/cm 2 . This significant difference ΔJ 0 of 34 fA/cm 2 can indicate the presence of other recombination processes in the n þ doped region, potentially due to SRH recombination via inactive phosphorus precipitates [44,45].…”

PC1Dmod 6.1 – state-of-the-art models in a well-known interface for improved simulation of Si solar cells

“…Min et al [29] recently suggested that SRH recombination close to the surface actually is an important contributing factor for the total 0 , and can therefore be used to explain the observed differences in simulated and experimental 0 data for the cases when only adapting 0 is not sufficient. Resolving the discussion of which BGN model that should be used for device simulation and the potential role of SRH recombination in solar cell emitters is beyond the scope of this paper.…”

“…We however want to provide a new tool for further investigation in this topic, and have therefore also implemented an option for adding position-dependent SRH recombination, which is defined by importing a text file specifying the SRH time constants 0 and 0 at different positions. In order to verify the implementation and to demonstrate a potential use of this feature we have used PC1Dmod 6.1 simulations to reproduce the original data published in the work of Min et al [29], which simulated the 0 values of a series of emitters fabricated on planar substrates. The additional SRH recombination in the emitter was assumed to be proportional to the concentration of inactive phosphorus at each point, taken as the difference between secondary ion mass spectrometry (SIMS) and electrochemical capacitance-voltage (ECV) measurements, which measures the concentration of total and active dopants, respectively.…”

“…As in Ref. [29] the SRV (via the 0 parameter) were adjusted to best match the experimental 0 values for each of four different emitters produced with increasing POCl 3 flow during phosphorus diffusion. As seen in Fig.…”