Unraveling the impact of nuclear quantum effects on proton affinity using nuclear electronic orbital-density functional theory: A Comprehensive Benchmark Study

This work presents a benchmarking of the nuclear electronic orbital-method used for multicomponent density functional theory calculations (NEO-DFT). We check the influence of the different parameters necessary in the method at the example of predicting proton affinity values. Here, NEO-DFT accounts for nuclear quantum effects of the proton being transferred. We used a test set of 72 molecules for this benchmark and took experimental proton affinities as reference. Our investigation of exchange correlation functionals reveals that CAM-B3LYP demonstrates the best performance. In the framework of NEO-DFT, the interaction with the quantum nuclei is treated via electron-proton correlation functionals. Here, the epc17-2 gave a similar performance to the computationally more demanding epc19. Compared to DFT (treating nuclei classically), NEO DFT provides a better value for the proton affinities when electron-proton correlation functionals are used. With respect to the electronic basis sets, the def2-QZVP basis set provides the most accurate values compared to def2-TZVP or def2-SVP, albeit with increased computational demands. Additionally, the nuclear basis sets show negligible effects on proton affinity accuracy and lack any trend. Overall, this study underlines the performance of NEO-DFT as an efficient approach to treat nuclear quantum effects using the example of calculating proton affinity values. The findings give a guideline towards the choice of optimal parameter sets in NEO-DFT calculations in the future.