印度科学研究所的研究人员利用人工智能和机器学习技术，在二氧化碳转化过程中发现了大量隐藏的化学反应途径。1这项研究将已知参与CO2转化的化学反应数量从传统模型的152种扩展至近9389种，极大地拓展了对该过程的化学认知。1
新模型对二氧化碳转化量的预测能力提高了约40倍，显著增强了转化效率的预测精度。1研究团队发现了一种非传统的反应机制：氢可以作为完整的H2分子进行转移，而非仅作为单个原子，这一发现为开发更高效的甲醇生成催化剂提供了新的思路。1第一作者Anand Mohan Verma目前担任MNNIT Allahabad的助理教授，相关实验验证由Hindustan Petroleum Corporation Limited的绿色研发中心完成。1
Researchers at the Indian Institute of Science have leveraged artificial intelligence and machine learning to uncover a vastly expanded network of chemical reactions involved in converting carbon dioxide, potentially accelerating the development of more efficient fuel production methods. The AI-driven analysis identified approximately 9,389 chemical reactions participating in the CO2 conversion process, compared to just 152 reactions known through conventional modeling approaches.1
The expanded model dramatically improved predictions of CO2 conversion yields by roughly 40-fold, while revealing an unconventional reaction mechanism in which hydrogen transfers as a complete H2 molecule rather than as individual atoms.1 These insights are expected to inform the design of superior catalysts for methanol generation from CO2.1 The research, conducted in collaboration with the green research and development center at Hindustan Petroleum Corporation Limited, was published in Nature Communications in 2026.1 Anand Mohan Verma, the study's lead author and now an assistant professor at MNNIT Allahabad, spearheaded the investigation alongside collaborator Ananth Govind Rajan.1
评论
还没有评论，欢迎留下第一条。