So far, intrinsic hydrogenated amorphous silicon (a-Si:H(i)) has been commonly used below molybdenum oxide (MoO_x) to form a good contact. An a-Si:H(i)/MoO_x stack gives good surface passivation, but often results in poor carrier selectivity after exposure to slightly elevated temperatures >130 °C (Geissbühler et al., 2015) [1]. For this reason, we have investigated an alternative interface layer, a very thin Al₂O₃ tunneling layer (<2 nm), deposited by atomic layer deposition (ALD), that can provide surface passivation, higher transparency and thermal stability without affecting the hole transport across the contact. To demonstrate this new passivating contact a 6” moly-poly cell, with an Al₂O₃/MoO_x stack at the front side and n-type doped polysilicon at the rear side, was made using a high- throughput spatial ALD tool, and E-beam PVD, for the Al₂O₃ and MoO_x layers, respectively. This resulted in an efficiency of 18.2% with a V_oc of 651 mV, a FF of 75.6% and a J_sc of 36.9 mA/cm². A post-deposition anneal (PDA) of the thin Al₂O₃ interlayer has significant effect on the Al₂O₃ thickness, layer stoichiometry, contact selectivity, and sputtering-induced damage. Annealing at higher T_PDA (350–600 °C) results in ineffective hole carrier transport and makes the stack more sensitive to ITO damage. The best performing device was, therefore, made using an Al₂O₃ layer without a PDA treatment. Moreover, we have found that this solar cell structure is thermally stable up to at least 210 °C, and even slightly improves under annealing which makes this device industrially appealing.

@en

Electron beam (E-beam) deposited molybdenum oxide (MoO_x) has been investigated for its potential to replace p-type hydrogenated amorphous silicon (a-Si:H) in Si heterojunction (SHJ) solar cells. Excellent passivation was achieved for our best MoO_x/c-Si junction based device, reaching an average implied V_oc (iV_oc) of 734 mV on textured, commercially available 6-inch Cz wafers. This confirms the compatibility of MoO_x as a hole selective layer with industrial SHJ cell processing. A hole barrier was, however, observed for our MoO_x-based solar cells due to inefficient hole extraction. The formation of this hole barrier can be related to annealing of MoO_x and the presence of a native oxide grown on the intrinsic a-Si:H interface layer below. Pre-annealing, followed by an HF treatment on the a-Si:H(i) layer prior to MoO_x deposition, proved to be useful to mitigate the formed barrier, while making it more stable under standard SHJ annealing conditions.

@en

We present large-area "moly-poly" cells, with a front side MoO_x/a-Si:H(i) passivating contact and a rear-side poly-Si/SiO_x stack, and we have demonstrated that MoO_x based c-Si solar cell technology can be scaled to industrial wafer size. Excellent surface passivation was achieved using MoO_x and poly-Si, leading to implied V_oc values above 700 mV, and a final cell V_oc of 687 mV. However, some care needs to be taken to avoid parasitic optical losses in the infra-red (IR) spectral range due to free-carrier absorption (FCA). These losses were investigated by comparing poly-Si layers of different thicknesses, deposited by low-pressure or plasma-enhanced chemical vapor deposition (LPCVD or PECVD), at the rear side of moly-poly cells. We found that ultra-thin PECVD layers are most suitable for solar cell applications due to a very good trade-off between surface passivation and reduced FCA. Based on this result, a 18.1% efficient 9.2 × 9.2 cm² moly-poly cell was made, which is the highest reported efficiency so far for moly-poly cells. Finally, we present a preliminary study of the parasitic IR losses in the MoO_x layer itself, when deposited on either a-Si:H or SiO_x passivation layers.

@en

High quality passivating contacts can be realized by using the combination of a thin interfacial oxide (SiOx) and doped polysilicon (polySi). Recombination losses are minimized by providing very good passivation between the thin hydrogenated oxide and the cSi, a high field effect by the highly doped polySi [1-2], combined with the low level penetration of dopants in the wafer [2-3]. To realize this low level in-diffusion of dopants, several interacting options are evaluated in this work: the quality of the thin oxide layer (growth method), combined with a diffusion blocking method (nitridation), doping concentration levels in the polySi and temperature of diffusion. It is shown that for Phosphorus (P)-doped polySi, in-diffusion can be reduced by adding an i-layer in between the oxide and the highly doped polySi, lowering the overall doping level in the system slightly. For Boron (B)-doped polySi, in-diffusion can be blocked by nitridation of the SiO2 layer.@en

An Ar‑H₂ plasma treatment was applied on an ultrathin RCA oxide to create well-passivated silicon wafers with symmetric c‑Si/SiO_x:H/a‑Si:H passivation layer stacks. The effective lifetime of these samples increased from 10 μs to 4 ms after annealing at 200 °C through Ar‑H₂ plasma treatment of the oxide. The results indicate that the plasma treatment can modify the RCA oxide and this enables atomic hydrogen diffusion at low annealing temperature, leading to a well passivated c‑Si/SiO_x:H interface. This might provide new possibilities to use wet chemical oxides in c-Si solar cells, for example as tunnel contacts.

@en