n-Butane dehydrogenation to butene and butadiene has recently gained increasing attention owing to the exploitation and development of shale gas as well as the rapid growth in the demand for synthetic rubber worldwide. In this work, the full n-butane dehydrogenation reaction network involving 568 elementary steps on Pt is established by using a chemical informatics approach to loop over all of the atoms and chemical bonds in n-butane. By combining density functional theory (DFT) calculations, the Morgan molecular fingerprint method, and machine learning techniques, we have identified 208 elementary steps that contribute to the kinetically important reaction network, which presents some general guidelines for the formulation of mechanisms of great complexity. A detailed microkinetic analysis that ensures thermodynamic consistency is then performed, without and with the presence of H₂ cofeeding, to assess the n-butane catalytic activity and butene selectivity. It turns out that in the absence of H₂, the high coverages of the coke precursors give rise to a low catalytic activity due to the occupancy of a large number of active sites. The turnover frequencies for n-butane consumption and butene production rise rapidly as the H₂/n-C₄H₁₀ ratio goes up from 0 to 1.33. Meanwhile, the selectivity toward 1-butene increases as well, whereas the selectivities toward 2-butene and 1,3-butadiene are not sensitive to the H₂ partial pressure. The flux analysis reveals that the dominant reaction pathways for 1-butene and 2-butene follow the reverse Horiuti–Polanyi mechanism, and the byproducts are formed primarily by the C–C bond cleavage in CH₃CCHC*. The C–H bond activation in n-butane is identified by the sensitivity analysis as the rate-limiting step for the overall reaction while the selectivities toward butenes are found to be controlled dominantly by the ease with which n-butane can be activated and how readily butenes can be deeply dehydrogenated.

中文翻译：

---

通过机器学习和微动力学分析解读 Pt 催化正丁烷脱氢的动力学复杂性

近年来，随着页岩气的开采和开发以及全球合成橡胶需求的快速增长，正丁烷脱氢制丁烯和丁二烯越来越受到关注。在这项工作中，通过使用化学信息学方法循环遍历正丁烷中的所有原子和化学键，建立了涉及 Pt 上 568 个基本步骤的完整正丁烷脱氢反应网络。通过结合密度泛函理论 (DFT) 计算、摩根分子指纹法和机器学习技术，我们确定了有助于动力学重要反应网络的 208 个基本步骤，这为制定极其复杂的机制提供了一些通用指南。然后，在不存在和存在H ₂共进料的情况下进行确保热力学一致性的详细微动力学分析，以评估正丁烷催化活性和丁烯选择性。事实证明，在不存在H ₂的情况下，焦炭前体的高覆盖度由于占据大量活性位点而导致催化活性较低。随着H ₂ / n -C ₄ H ₁₀比率从0上升至1.33，正丁烷消耗和丁烯生产的周转频率迅速上升。同时，对1-丁烯的选择性也增加，而对2-丁烯和1,3-丁二烯的选择性对H ₂分压不敏感。通量分析表明，1-丁烯和2-丁烯的主要反应路径遵循反向Horiuti-Polanyi机制，副产物主要由CH ₃ CCHC*中的C-C键断裂形成。通过敏感性分析，正丁烷中的 C-H 键活化被确定为整个反应的限速步骤，而丁烯的选择性主要受正丁烷活化的难易程度以及活化方式的控制。丁烯容易深度脱氢。