Artificial intelligence in adsorption process optimization: A 2014–2024 bibliometric analysis of research trends and developments

Abstract

• First bibliometric analysis of AI applications in adsorption (2014–2024). • Machine learning dominates research (ANNs, SVMs), but hybrid models remain rare. • Key challenges identified: data scarcity, model interpretability, material-specific limitations. • Practical roadmap proposed: standardized databases, SHAP integration, experimental validation. • Future directions: Focus on hybrid AI models and industrial-scale validation. This study offers a bibliometric analysis of research trends and developments in the application of Artificial Intelligence (AI) to adsorption process optimization from 2014 to 2024, with an annual growth rate of 26.54 % in the number of publications indexed in the Scopus database. Utilizing the Scopus database, we analyzed key trends, influential authors, leading institutions, and international collaborations. The analysis reveals a growing interest in AI-driven adsorption, with Machine Learning (ML) techniques increasingly employed for modeling, prediction, and material design. Key findings highlight the significant contributions of researchers from Iran, China, and India and the prominent role of affiliations like the Islamic Azad University. Keyword co-occurrence analysis identifies ”artificial neural networks,” ”machine learning,” ”adsorption,” and ”optimization” as central themes. AI demonstrates significant potential to improve adsorption efficiency and sustainability, although challenges remain concerning data availability, model interpretability, and generalizability across materials. Future research should focus on developing standardized adsorption databases, improving model transparency, and validating AI-driven predictions through experimental studies. Addressing these obstacles, especially the need for reliable data and interpretable models, will be critical for integrating AI tools into real-world adsorption systems. By grounding future AI developments in experimental validation and practical challenges, researchers can accelerate meaningful progress in environmental remediation and sustainable resource recovery.