As the record single-junction efficiencies of perovskite solar cells now rival those of CIGS, CdTe, and multicrystalline silicon, they are becoming increasingly attractive for use in tandem solar cells, due to their wide, tunable bandgap and solution processability. Previously, perovskite/silicon tandems were limited by significant parasitic absorption and poor environmental stability. Here, we improve the efficiency of monolithic, two-terminal, 1 cm 2 perovskite/silicon tandems to 23.6% by combining an infrared-tuned silicon heterojunction bottom cell with the recently developed cesium formamidinium lead halide perovskite. This more stable perovskite tolerates deposition of a tin oxide buffer layer via atomic layer deposition that prevents shunts, has negligible parasitic absorption, and allows for the sputter deposition of a transparent top electrode. Furthermore, the window layer doubles as a diffusion barrier, increasing the thermal and environmental stability to enable perovskite devices that withstand a 1000-hour damp heat test at 85 °C and 85% relative humidity.

Semi-transparent perovskite solar cells with silver nanowires are stacked on CIGS and Si to achieve solid-state polycrystalline tandems with efficiency improvement.

This work presents Neural Equivariant Interatomic Potentials (NequIP), an E(3)-equivariant neural network approach for learning interatomic potentials from ab-initio calculations for molecular dynamics simulations. While most contemporary symmetry-aware models use invariant convolutions and only act on scalars, NequIP employs E(3)-equivariant convolutions for interactions of geometric tensors, resulting in a more information-rich and faithful representation of atomic environments. The method achieves state-of-the-art accuracy on a challenging and diverse set of molecules and materials while exhibiting remarkable data efficiency. NequIP outperforms existing models with up to three orders of magnitude fewer training data, challenging the widely held belief that deep neural networks require massive training sets. The high data efficiency of the method allows for the construction of accurate potentials using high-order quantum chemical level of theory as reference and enables high-fidelity molecular dynamics simulations over long time scales.

With the advent of efficient high-bandgap metal-halide perovskite photovoltaics, an opportunity exists to make perovskite/silicon tandem solar cells. We fabricate a monolithic tandem by developing a silicon-based interband tunnel junction that facilitates majority-carrier charge recombination between the perovskite and silicon sub-cells. We demonstrate a 1 cm2 2-terminal monolithic perovskite/silicon multijunction solar cell with a VOC as high as 1.65 V. We achieve a stable 13.7% power conversion efficiency with the perovskite as the current-limiting sub-cell, and identify key challenges for this device architecture to reach efficiencies over 25%.