Researchers from Soochow University, The Hong Kong Polytechnic University, Suzhou Maxwell Technologies, the University of Oxford, Wuxi EliTe Solar, Hangzhou Zhongneng Photoelectricity Technology, Shenzhen Polytechnic University and Suzhou National Laboratory have developed an indium-free recombination layer for monolithic perovskite/silicon tandem solar cells, based on titanium oxynitride (TiOxNy), reporting power conversion efficiencies of 33.3% on 1.0 cm² devices and 30.6% on industrial-size, 207.87 cm² tandems.
Monolithic perovskite/silicon tandems are widely seen as a route past the efficiency ceiling of single-junction photovoltaics, but turning lab-scale records into durable, manufacturable modules depends heavily on one thin layer: the recombination interconnect that joins the two subcells. That layer has to do three things at once - support fast charge recombination, stay optically transparent so light reaches both subcells, and form a chemically robust interface that holds up over time. Today's leading options each fall short somewhere: indium-containing transparent conductive oxides (TCOs) raise cost and supply concerns, while silicon-based tunnel junctions introduce parasitic optical losses.
The team's answer is to move away from both approaches and use conductive titanium oxynitride as a single, multifunctional recombination layer. TiOxNy is designed to support efficient vertical charge recombination between the perovskite and silicon subcells while suppressing lateral current leakage - a distinction that matters for long-term stability, since incomplete or inefficient recombination allows carriers to accumulate, degrading interfaces and accelerating performance loss during operation.
The material's role isn't purely electrical. TiOxNy also provides anchoring sites for self-assembled monolayers (SAMs), attaching through a tridentate binding configuration. Those SAMs help shape the energy landscape at the interface, improving contact selectivity and cutting down on recombination pathways that would otherwise waste photogenerated carriers. On the optical side, the layer is engineered to stay sufficiently transparent that both subcells can still harvest light efficiently - avoiding the usual trade-off between electrical performance and manufacturability.
Alongside the 33.3% and 30.6% PCE results, the devices showed improved operational durability, supporting the idea that TiOxNy stabilizes interfacial chemistry against the stresses that typically degrade tandem stacks, and not just their initial performance. Reaching 30.6% at 207.87 cm² is the more significant number for industrial translation: it suggests the recombination layer concept holds up under the structural and processing complexity of large-area fabrication, rather than being a lab-only effect confined to small test cells.
Being indium-free, with what the researchers describe as minimal optical penalty, TiOxNy is positioned as a scalable interconnection strategy for next-generation tandem photovoltaics. If the approach holds up across full manufacturing lines, it could help move perovskite-silicon tandems further from record-chasing lab devices and closer to high-yield, widely deployable production.
