Tandem solar cells (SCs) based on perovskite and silicon represent an exciting possibility for a breakthrough in photovoltaics, enhancing SC power conversion efficiency (PCE) beyond the single-junction limit while keeping the production cost low. A critical aspect to push the tandem PCE close to its theoretical limit is the development of high-performing semitransparent perovskite top cells, which also allow suitable near-infrared transmission. Here, we have developed highly efficient semitransparent perovskite SCs (PSCs) based on both mesoporous and planar architectures, employing Cs0.05(MA0.17FA0.83)0.95Pb­(I0.83Br0.17)3 and FA0.87Cs0.13PbI2Br perovskites with band gaps of 1.58 and 1.72 eV, respectively, which achieved PCEs well above 17 and 14% by detailed control of the deposition methods, thickness, and optical transparency of the interlayers and the semitransparent electrode. By combining our champion 1.58 eV PSCs (PCE of 17.7%) with an industrial-relevant low-cost n-type Si SCs, a four-terminal (4T) tandem efficiency of 25.5% has been achieved. Moreover, for the first time, 4T tandem SCs’ performances have been measured in the low light intensity regime, achieving a PCE of 26.6%, corresponding to revealing a relative improvement above 9% compared to the standard 1 sun illumination condition. These results are very promising for their implementation under field-operating conditions.

Perovskite solar cells (PSCs) with transparent electrodes can be integrated with existing solar panels in tandem configurations to increase the power conversion efficiency. A critical layer in semi-transparent PSCs is the inorganic buffer layer, which protects the PSC against damage when the transparent electrode is sputtered on top. The development of n-i-p structured semi-transparent PSCs has been hampered by the lack of suitable p-type buffer layers. In this work we develop a ptype CuO x buffer layer, which can be grown uniformly over the perovskite device without damaging the perovskite or organic hole transport layers. The CuO x layer has high hole mobility (4.3 ± 2 cm 2 V -1 s -1 ), high transmittance (>95%), and a suitable ionization potential for hole extraction (5.3 ± 0.2 eV). Semi-transparent PSCs with efficiencies up to 16.7% are achieved using the CuO x buffer layer. Our work demonstrates a new approach to integrate n-i-p structured PSCs into tandem configurations, as well as enable the development of other devices that need high quality, protective p-type layers.

Process optimization of photovoltaic devices is a time-intensive, trial-and-error endeavor, without full transparency of the underlying physics, and with user-imposed constraints that may or may not lead to a global optimum. Herein, we demonstrate that embedding physics domain knowledge into a Bayesian network enables an optimization approach that identifies the root cause(s) of underperformance with layer-by-layer resolution and reveals alternative optimal process windows beyond global black-box 2 optimization. Our Bayesian-network approach links process conditions to materials descriptors (bulk and interface properties, e.g., bulk lifetime, doping, and surface recombination) and device-performance parameters (e.g., cell efficiency), using a Bayesian inference framework with an autoencoder-based surrogate device-physics model that is 100x faster than numerical solvers. With the trained surrogate model, our approach is robust and reduces significantly the time-consuming experimentalist intervention, even with small numbers of fabricated samples. To demonstrate our method, we perform layerby-layer optimization of GaAs solar cells. In a single cycle of learning, we find an improved growth temperature for the GaAs solar cells without any secondary measurements, and demonstrate a 6.5% relative AM1.5G efficiency improvement above baseline and traditional black-box optimization methods.

No abstract