An AI-discovered all-in-one ink yields cost-effective perovskite solar modules exceeding 22% efficiency

Read the full article See related articles

Discuss this preprint

Start a discussion What are Sciety discussions?

Listed in

This article is not in any list yet, why not save it to one of your lists.
Log in to save this article

Abstract

Screening optimal formulations from billions of candidates remains a key bottleneck for developing industrial perovskite photovoltaic inks, as the vast combinatorial parameter space renders trial-and-error approaches impractical. Here, we discover an all-in-one ink that, through a one-step, quench-free deposition, self-stratifies into three distinct functional layers. The discovery is accelerated by a self-constructed AI-driven framework named PerovSeek that integrates pre-trained prediction models with Bayesian optimization and high-throughput experimentation. This ink eliminates the need for separate deposition of charge-transport and passivation layers, reducing fabrication complexity by over 60%. When deployed in a slot-die coater, it yields perovskite solar modules with a certified power conversion efficiency of 22.09% (an aperture area of 695 cm2), representing one of the highest efficiencies reported for solution-processed large-area modules. By combining the all-in-one ink with quench-free deposition, this work bridges the gap between laboratory efficiency and industrial viability, establishing a direct pathway to low-cost, high-throughput manufacturing of perovskite solar modules.

Article activity feed