Scholarlink

<p>This work introduces OpenAI Perovskite, an open-access platform for the systematic analysis of two-dimensional (2D) lead-halide perovskite solar cells. The platform integrates an arranged database and interpretable machine learning to analyse complex parameter interactions. It has been revealed through regression modelling and SHAP analysis that the device efficiency is heavily influenced by coupled composition–architecture ramifications. The interaction between the A-site cation and the electron transport layer (ETL) is recognised as the primary predictor of power conversion efficiency (PCE).&#xa0;OpenAI Perovskite provides a reproducible, interaction-centred benchmarking resource to guide hypothesis generation and accelerate experimental optimisation in 2D photovoltaics.</p> Graphic abstract <p></p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

OpenAI perovskite: A data-driven platform for exploration and performance analysis of two-dimensional perovskite solar cells

  • 期刊论文

  • 发表日期：

    2026年4月9日

查看全文
  • Abdul Hadi Ismail,
  • Nurul Izzah Hayati Mohd Noor,
  • Nur Aleya Nazifa Zulkifle,
  • Wan Nur Izzati Wan Abd Aziz,
  • Noorhafizah Muhammad,
  • Hasiah Salleh,
  • Norhayati Sabani,
  • Hasyiya Karimah Adli

摘要

This work introduces OpenAI Perovskite, an open-access platform for the systematic analysis of two-dimensional (2D) lead-halide perovskite solar cells. The platform integrates an arranged database and interpretable machine learning to analyse complex parameter interactions. It has been revealed through regression modelling and SHAP analysis that the device efficiency is heavily influenced by coupled composition–architecture ramifications. The interaction between the A-site cation and the electron transport layer (ETL) is recognised as the primary predictor of power conversion efficiency (PCE). OpenAI Perovskite provides a reproducible, interaction-centred benchmarking resource to guide hypothesis generation and accelerate experimental optimisation in 2D photovoltaics.

Graphic abstract