Industrial Solar Cells Featuring Carrier Selective Front Contacts

We present an industrially applicable approach to integrate boron doped carrier-selective passivating contacts to the front side of silicon solar cells. A single BBR3 diffusion simultaneously forms the highly doped p region within the Poly-Si as well as the lowly doped buried junction below the AlOX/SiNX stack. A self-limiting wet etching process allows the removal of the Poly-Si at the noncontacted region between the fingers, whilst stopping at the wafer/SiOX interface. The final structure combines the high-level passivation and transparency of the AlOX/SiNX together with the good passivation and transport properties of the carrier selective contact. In Quokka 3 [1] simulations, the main solar cell parameters are analysed as a function of the saturation current density for the contacted region versus the non-contacted. The simulations show that implementing the Poly-Si below the metallization on the front side can boost the efficiency from 23.1% to 23.5%, so an effective gain of 0.4%. In recent years, carrier-selective passivating contacts based on a thin SiOX layer covered by a highly doped Poly-Si have demonstrated their potential in world record devices [2], [3]. To date they have been applied mainly to the solar cell rear side. Thanks to their excellent contact properties, the front side often then becomes the limiting efficiency factor, namely the recombination losses at the contact openings [4]. These losses are targeted by using a Poly-Si contact on the front. However, using a full area PolySi is unfavourable, when compared with the low parasitic absorption of an Al2O3/SiNX stack. Therefore a patterned front Poly-Si below the metallization becomes the focus of our research. We propose a process sequence forming the shallow-doped region (200-400 Ω/sq) below the Al2O3/SiNX stack in a single diffusion step together with the doping of the Poly-Si with a surface concentration above 10 cm , providing low resistance with the metallization. In this contribution intrinsic Poly-Si layers are prepared by Jinko solar upon a thin thermal SiOX (<2nm), grown on 180 μm thick double side textured n-type wafers, having a bulk resistivity of 2 Ωcm. Symmetrical test samples are fabricated in a tube furnace either in a POCl3 or a BBr3 atmosphere and lead to implied open circuit voltages (iVOC) above 730 mV and 700mV, respectively. A self-limiting wet chemical etching process allows the removal of the Poly-Si while keeping the shallow diffusion at the wafer surface, shown in Figure 1b, together with the sheet resistance RSH. After etchingoff the Poly-Si, buried junctions are re-passivated by Al2O3/SiNX. The achieved iVOC values of ~690mV 0 50 100 150 200 250 10 10 10 689 mV 332 W/sq 686 mV 248 W/sq B o ro n c o n c . [c m -1 ] depth [nm] 600°C therm. SiOX: 50nm Poly-Si 80nm Poly-Si 700°C therm. SiOX: 50nm Poly-Si 691 mV 701 W/sq wafer Figure 1 a) Sketch of the measured symmetrical test structure after the poly-Si is etched off and the surface is repassivated by Al2O3/SiNX. b) Boron doping profiles measured by ECV for different SiOX temperatures and Poly-Si thicknesses. The sheet resistance as well as the iVOC values after re-passivation with an Al2O3/SiNX stack are indicated next to each condition. a) b) (J0 ~ 30mA/cm) are in agreement with literature values for Al2O3/SiNX passivation of such doping profiles [5]. This proves that the wafer surface can be re-passivated between the fingers with high quality after etching off the poly-Si layer. The potential and limitations of this approach is investigated using Quokka 3 [1] simulations. A metallization covering 5% of the illuminated area is used representing an industrial solar cell (Figure 2ad). The key photovoltaic parameters (fill factor (FF), open-circuit voltage (VOC) and efficiency) are plotted as a function of the saturation current density J0 for the contacted versus non-contacted region. An optimum for the FF is observed for an RSH between 200-500Ω/sq, which enables 10fA/cm < J0,non contacted < 30fA/cm. The VOC is mainly driven by J0, non-contacted due to the larger surface area. The contribution of J0,contact becomes more apparent in case of low J0, non-contacted. The white star on the graphs indicates a state-of-the-art contact without Poly-Si below the metallization (J0,contact = 500fA/cm; J0, non-contacted = 20fA/cm). The black star indicates a contact using the proposed approach with Poly-Si below the metallization using values which have been achieved experimentally in our lab (J0,contact = 15fA/cm; J0, non-contacted = 30fA/cm). The simulations show that, using this approach, the VOC improves by 10mV (from 698mV to 708mV) leading to an improvement in efficiency from 23.1% to 23.5%. The effective gain of 0.4% evidences the strong potential of this approach to boost the efficiency of the solar cell while keeping the fabrication process simple.