In heterogeneous catalysis, unravelling the distinct structural and compositional nature of active sites provides a good platform in modulating important catalytic properties toward improved activity and selectivity. Obviously, despite the direct propane dehydrogenation (PDH) being a commercial process, oxidative dehydrogenation of propane (ODHP) remains a highly prospective technology for propene production due to its leveraged thermodynamic advantage and less susceptibility to catalyst deactivation. The present work has systematically and comprehensively reviewed recent advances on key aspects related to ODHP catalysts such as nature of the active sites, active sites isolation, accessibility of active sites, active sites in layered-double hydroxide (LDH) derived mixed oxide catalysts, intercalated LDH precursors, co-precipitated LDH precursors, oxides-support interaction, oxygen vacancy, surface oxide reducibility, and acid–base property. Furthermore, discussion on accelerated discovery of novel catalysts via high-throughput experimentations (HTE) and high-throughput quantum computations (HTC) has been provided. Integrated frameworks via machine learning and Python language are hugely augmenting in this pursuit. It is hoped that this perspective provides insights on the design and tuning of heterogeneous catalysts, and in particular the development of efficient catalysts that could prompt industrialization and commercialization of ODHP process.

在非均相催化中，揭示活性位点的独特结构和组成性质可为调节重要的催化性能以提高活性和选择性提供一个良好的平台。显然，尽管直接丙烷脱氢（PDH）是一种商业方法，但是丙烷的氧化脱氢（ODHP）由于其杠杆作用的热力学优势和对催化剂失活的敏感性较小，仍然是丙烯生产的高前景技术。本工作系统地和全面地回顾了与ODHP催化剂相关的关键方面的最新进展，例如活性部位的性质，活性部位的分离，活性部位的可及性，层状双氢氧化物（LDH）衍生的混合氧化物催化剂中的活性部位，插层LDH前体，共沉淀LDH前体，氧化物-载体相互作用，氧空位，表面氧化物还原性和酸碱性质。此外，提供了关于通过高通量实验（HTE）和高通量量子计算（HTC）加速发现新型催化剂的讨论。通过机器学习和Python语言的集成框架在这种追求中得到了极大的增强。希望这种观点能为多相催化剂的设计和调整提供见解，尤其是开发有效的催化剂，可以促进ODHP工艺的工业化和商业化。通过高通量实验（HTE）和高通量量子计算（HTC）加速了新型催化剂的发现。通过机器学习和Python语言的集成框架在这种追求中得到了极大的增强。希望这种观点能为多相催化剂的设计和调整提供见解，尤其是开发有效的催化剂，可以促进ODHP工艺的工业化和商业化。通过高通量实验（HTE）和高通量量子计算（HTC）加速了新型催化剂的发现。通过机器学习和Python语言的集成框架在这种追求中得到了极大的增强。希望这种观点能为多相催化剂的设计和调整提供见解，尤其是开发有效的催化剂，可以促进ODHP工艺的工业化和商业化。