《科学》 第392卷 第6799期 · 2026年5月14日 · 中文解读
AI引导设计在100°C下运行稳定的高效钙钛矿太阳能电池
AI-guided design of efficient perovskite solar cells operationally stable at 100°C · J. Guo et al.
约 41 分钟Research Article Summaries在小程序里点播,20 到 60 分钟做好
这篇讲什么
本文介绍了一种基于多智能体AI框架的设计方法,用于开发在100°C高温下运行1000小时后仍保持97%初始效率的钙钛矿太阳能电池。
原文开头
To realize global PSC optimization at the system level, we developed a new route using advanced multiagent AI techniques (13, 14). By analyzing the studies on high-performance PSCs recommended by large language models (LLMs) from a comprehensive literature database [dataset 1 in (15)], we explored various choices for perovskite components and interface materials using the insights from interconnected professional agents. Coupled with quantitative classical ML algorithms, our multiagent AI framework accurately predicted perovskite compositions and evaluated interface materials. For perovskite materials, AI prediction combined with thermodynamically driven single-crystal growth identified FA0.92Cs0.08PbI3 (FA, formamidinium) as the optimal composition, demonstrating the lowest trap density. …
摘自《科学》(Science)第392卷 第6799期 · 2026年5月14日,J. Guo et al.。仅引用开头一小段供了解文章,版权归原刊所有,全文请阅读原刊。

微信扫码,在小程序里听完整版
不用登录先听一篇 · 或在微信搜索小程序 晨间信号