Fixed-node diffusion Monte Carlo shows promise for modeling reaction thermochemistry of hydrocarbon-based radicals.

This article evaluates the use of fixed-node diffusion Monte Carlo (FN-DMC), a quantum Monte Carlo method, for accurately predicting activation energies and reaction enthalpies in free radical polymerization (FRP) reactions. Using methyl radical addition to ethylene as a model, the study systematically analyzes FN-DMC parameters—including trial wavefunctions, basis set sizes, and effective core potentials (ECPs)—and benchmarks results against high-level ab initio methods and experimental data. FN-DMC demonstrates chemical accuracy with mean absolute deviations of 3.5(7) kJ/mol for radical addition and 1.4(8) kJ/mol for hydrogen abstraction enthalpies relative to the Weizmann-1 (W1) composite method, outperforming the widely used G3X(MP2)-RAD method. Given its favorable computational scaling and parallelizability, FN-DMC with single Slater–Jastrow trial wavefunctions and ECPs is recommended for further testing on larger, more complex polymerization systems where traditional benchmarking methods are computationally prohibitive.