路径(计算)
位阻效应
动能
计算机科学
点(几何)
化学
基本反应
统计物理学
反作用坐标
生物系统
数学优化
迭代法
化学反应
代理(统计)
反应速率
热力学
物理
材料科学
拓扑(电路)
单点
补语(音乐)
作者
Nicholas Casetti,Connor W. Coley
标识
DOI:10.26434/chemrxiv.15011670/v1
摘要
Automated reaction exploration can generate more candidate elementary steps than can be practically characterized with transition-state searches. Thermodynamic screening provides an inexpensive first filter, but endpoint energies contain no information about the path connecting reactants and products. Here, we introduce SPARTA (Steric Path Alignment for Reactive Transformation Assessment), a geometric kinetic proxy that generates reaction-biased endpoint conformers, refines them through iterative image-dependent pair potential path optimization, and evaluates electronic single-point energies along the resulting path. Across two large reaction-barrier datasets, the SPARTA barrier discriminates highand low-barrier reactions better than reaction energy. Across curated reaction classes, SPARTA reproduces experimentally established steric trends when kinetic differences are governed by accessibility to the reactive configuration, while identifying clear limitations for stereoelectronically or orbital-symmetrycontrolled transformations. Applied to glucose pyrolysis and formose reaction networks, a single SPARTA threshold retains more than 98% of reactions with reference barriers below 50 kcal mol−1. In the glucose network, adding SPARTA after thermodynamic screening reduces the number of candidate transitionstate searches from 13,436 to 4,760 while retaining all 241 reactions below 20 kcal mol−1. These results establish the geometric accessibility of the reaction path as a useful complement to thermodynamic screening for reducing the cost of automated reaction-network exploration.
科研通智能强力驱动
Strongly Powered by AbleSci AI