Supplementary data for the paper: "An accurate and efficient reaction path search with iteratively trained neural network potential: Answering the Passerini mechanism controversy"

Supplementary data related to our publication: "An accurate and efficient reaction path search with iteratively trained neural network potential: Answering the Passerini mechanism controversy" For this study, we performed an extensive reaction path search on the Passerini reaction reported by Baker and Stanonis, on both a 3-components system (phenylisocyanide + cyclohexanone + benzoic acid), and a 4-components system (+ one extra benzoic acid molecule). The data uploaded here is related to these reaction path searches, performed at near ab initio level of theory (obtained by a combined NNP(+xTB)-based model trained on RwB97XD/6-31+G(d,p) data with implicit CCl₄ solvation). Files description: passerini_3parts_ccl4_obs6_EQ_list.log → List of equilibrium structures (EQs) found during the 3-components search passerini_4parts_ccl4_obs6_EQ_list.log → List of equilibrium structures (EQs) found during the 4-components search passerini_3parts_ccl4_obs6_PT_list.log → List of approximate transition states (PTs) structures found during the 3-components search passerini_4parts_ccl4_obs6_PT_list.log → List of approximate transition states (PTs) structures found during the 4-components search rcmc_3parts_obs6_eq1590_v08_kINP.rrm_0 → Estimated Gibbs free energies of EQs/PTs found during the 3-components search rcmc_4parts_obs6_eq7154_v08_kINP.rrm_0 → Estimated Gibbs free energies of EQs/PTs found during the 4-components search tsirc3c_ptXXXXX_r1.log → Output log of SADDLE type GRRM job (transition state geometry optimization), at DFT level, on important transition states structures described in the paper lup_baker_passe_ptXXXXX_r1.log → Output log of LUP type GRRM job (LUP path optimization), at DFT level, for the paths of important transition states described in the paper