Prof Henry Snaith FRS

Abstract:

Impressive power conversion efficiencies coupled with the relative ease of fabrication have made perovskite solar cells a front runner for next-generation photovoltaics. Although perovskite films and optoelectronic devices have been widely studied, relatively little is known about the chemistry of the precursor solutions. Here, we present a study on the hydrolysis of N,N-dimethylformamide, correlating how pH changes related to its degradation affect the crystallization of MAPbI3xClx perovskite films. By careful manipulation of the pH, and the resulting colloid distribution in precursor solutions, we fabricate perovskite films with greatly improved crystallinity, which when incorporated into photovoltaic devices reproducibly yield efficiencies of over 18%. Extending this method to the mixed cation, mixed halide perovskite FA0.83MA0.17Pb(I0.83Br0.17)3, we obtain power conversion efficiencies of up to 19.9% and open-circuit voltages of 1.21 V for a material with a bandgap of 1.57 eV, achieving the lowest yet reported loss in potential from bandgap to a VOC of only 360 mV.

Abstract:

Metal halide perovskite semiconductors offer rapid, low-cost deposition of solar cell active layers with a wide range of band gaps, making them ideal candidates for multijunction solar cells. Here, we combine optical and electrical models using experimental inputs to evaluate the feasible performances of all-perovskite double-junction (2PJ), triple-junction (3PJ), and perovskite-perovskite-silicon triple-junction (2PSJ) solar cells. Using parameters and design constraints from the current state-of-the-art generation of perovskite solar cells, we find that 2PJs can feasibly approach 32% power conversion efficiency, 3PJs can reach 33%, and 2PSJs can surpass 35%. We also outline pathways to improve light harvesting and demonstrate that it is possible to raise the performances to 34%, 37%, and 39% for the three architectures. Additionally, we discuss important future directions of research. Finally, we perform energy yield modeling to demonstrate that the multijunction solar cells should not suffer from reduced operational performances due to discrepancies between the AM1.5G and real-world spectrum over the course of a year.