• HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
Monday, November 16, 2026
BIOENGINEER.ORG
No Result
View All Result
  • Login
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
No Result
View All Result
Bioengineer.org
No Result
View All Result
Home NEWS Science News Technology

Transformer Model Predicts Lead-Free Perovskite Solar Cell Performance

by
November 10, 2026
in Technology
Reading Time: 2 mins read
0
Transformer Model Predicts Lead-Free Perovskite Solar Cell Performance

Transformer Model Predicts Lead-Free Perovskite Solar Cell Performance

Share on FacebookShare on TwitterShare on LinkedinShare on RedditShare on Telegram

The study, published in Discover Applied Sciences, introduces a Physics-Informed Feature Tokenizer Transformer, or FT-Transformer, to address the complexity of optimizing multi-dimensional parameters in these photovoltaic devices. Tin-based perovskites, specifically methylammonium tin triiodide (MASnI3) and formamidinium tin triiodide (FASnI3), are considered promising pathways for solar cells with improved efficiency, yet their optimization remains difficult due to the interplay of various physical factors.

The core innovation of this work lies in integrating physical constraints directly into the transformer architecture to maintain scientific integrity. Unlike standard machine learning models that may produce predictions inconsistent with fundamental physics, the FT-Transformer incorporates custom monostaticity constraints. These constraints enforce known semiconductor properties, such as the principle that higher defect densities lead to lower device efficiency. By embedding these physical laws into the model, the researchers aim to ensure that the predictions remain physically plausible, even when extrapolating to new parameter combinations.

To validate the model’s reliability, the team employed a comprehensive 5-fold cross-validation procedure using various random seeds. This approach was specifically designed to avoid partitioning bias, a common issue in machine learning where data splitting can artificially inflate performance metrics. The results demonstrated that the FT-Transformer possessed good architectural stability, producing consistent cross-validated R-squared values across different validation folds. For the MASnI3 material, the model achieved an R-squared value of 0.8024 with a standard deviation of 0.0135, while for FASnI3, it achieved a higher R-squared of 0.9574 with a standard deviation of 0.0024.

The study also compared the FT-Transformer against classical ensemble models, such as Random Forest. While the Random Forest model exhibited better empirical R-squared values in some instances, it failed to maintain physical consistency in its predictions. The FT-Transformer, by contrast, prioritized physical consistency, ensuring that its outputs adhered to the expected behavior of semiconductor materials.

Further analysis using gradient-based inverse design and SHAP explainability methods revealed distinct sensitivities between the two perovskite materials. The findings indicated that FASnI3 exhibits stronger sensitivity to defect densities at the interfaces compared to MASnI3.

The authors conclude that their cross-validated benchmarking framework offers a comprehensive and physics-based suite of tools to accelerate the development of perovskite technology. By combining the predictive power of transformer architectures with the rigor of physical constraints, the study provides a robust method for navigating the complex parameter space of lead-free perovskite solar cells.

Subject of Research: Machine learning modeling of lead-free perovskite solar cells

Article Title: Physics informed transformer modeling of lead free perovskite solar cells

Article References: Abdullah, S., Jamee, A. N., & Ifty, M. S. M. (2026). Physics informed transformer modeling of lead free perovskite solar cells. Discover Applied Sciences. https://doi.org/10.1007/s42452-026-09855-9

Image Credits: AI Generated

DOI: 10.1007/s42452-026-09855-9

Keywords: perovskite solar cells, machine learning, transformer model, lead-free materials, physics-informed, photovoltaics, cross-validation, defect density, SHAP analysis, inverse design, Physics, informed

News Source: Denise Maddox. (November 10, 2026). Transformer Model Predicts Lead-Free Perovskite Solar Cell Performance. Scienmag.

Tags: cross-validationdefect densityinformedinverse designlead-free materialsMachine Learningperovskite solar cellsphotovoltaicsphysicsphysics-informedSHAP analysisTransformer model
Share12Tweet7Share2ShareShareShare1

Related Posts

ShieldFL Framework Defends Non-IID Federated Learning Against Byzantine Attacks

ShieldFL Framework Defends Non-IID Federated Learning Against Byzantine Attacks

November 16, 2026
Quantum Inspired Model Directs Metaverse Server Resources in Simulated Cluster Tests

Quantum Inspired Model Directs Metaverse Server Resources in Simulated Cluster Tests

November 16, 2026

Event-Triggered Reinforcement Learning Cuts Control Overhead in Distributed IoT Sensor Clusters

November 16, 2026

Random Forest Outperforms CART in Mapping Land Cover Across Ajmer Region

November 16, 2026

POPULAR NEWS

  • Attention Graph Model Improves Dynamic Anomaly Detection Across Six Datasets

    Attention Graph Model Improves Dynamic Anomaly Detection Across Six Datasets

    29 shares
    Share 12 Tweet 7
  • Blockchain-Integrated Machine Learning Enhances Fraud Detection in Supply Chain Networks

    29 shares
    Share 12 Tweet 7
  • Dynamic Cognitive Models Improve External Dementia Transition Predictions But Face Major Limits

    29 shares
    Share 12 Tweet 7
  • Binary Hippopotamus Optimization Achieves High Accuracy Across Sixteen Cancer Datasets

    29 shares
    Share 12 Tweet 7

About

We bring you the latest biotechnology news from best research centers and universities around the world. Check our website.

Follow us

Recent News

ShieldFL Framework Defends Non-IID Federated Learning Against Byzantine Attacks

Surrounding Cereal and Non-Crop Lands Associate With Canola Weevil Numbers

Quantum Inspired Model Directs Metaverse Server Resources in Simulated Cluster Tests

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 87 other subscribers
  • Contact Us

Bioengineer.org © Copyright 2023 All Rights Reserved.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • Homepages
    • Home Page 1
    • Home Page 2
  • News
  • National
  • Business
  • Health
  • Lifestyle
  • Science

Bioengineer.org © Copyright 2023 All Rights Reserved.