An Overview of Trustworthy and Sustainable AI Development for Catalyst Screening: The Role of Explainable AI (XAI)

The significance of predicting the trustworthiness of Explainable Artificial Intelligence (XAI) in the context of sustainable AI-based catalyst screening lies in ensuring the reliability and interpretability of AI-driven decisions in catalysis research. Trustworthy XAI can enhance the transparency and accountability of AI models, crucial for gaining the confidence of scientists and stakeholders in the findings. By accurately predicting the performance and reliability of catalysts, XAI aids in identifying the most promising candidates efficiently, reducing the need for extensive experimental trials, thus saving time and resources. Moreover, trustworthy XAI facilitates better understanding of the underlying mechanisms driving catalyst performance, leading to more informed and targeted research directions. In sustainable AI-based catalyst screening, where environmental and economic considerations are paramount, the integration of XAI ensures that AI models are not only accurate but also aligned with sustainable development goals. This fosters the development of greener and more efficient catalytic processes, ultimately contributing to the advancement of sustainable technologies. The ability to predict and ensure the trustworthiness of XAI in catalyst screening thus represents a critical step towards leveraging AI for sustainable and impactful scientific advancements. This review discusses various techniques commonly used in sustainable AI for catalyst screening, highlighting different XAI methods with their benefits and drawbacks in the catalyst screening process. It suggests integrating multiple XAI tools to overcome individual limitations and presents significant findings in catalyst screening. The review also addresses challenges and explores future directions in this field.