Benchmarking of CCSD(T)/CBS binding energies and density functional theory performance for small arsine clusters (AsH3)n (n = 2–6): unraveling the delicate balance between weak hydrogen bonding and dispersion interactions
摘要
Arsine (AsH3) clusters—heavy Group 15 hydride systems of critical importance in semiconductor manufacturing and atmospheric arsenic chemistry—lack rigorous high-accuracy ab initio quantification of their binding energetics, hindering the development of reliable computational methods for larger arsenic-containing systems. Here we present a systematic investigation of the structures and binding energies of arsine clusters ranging from dimers to hexamers, employing a validated composite scheme to establish benchmark energies at the CCSD(T)/CBS (coupled-cluster singles, doubles, and perturbative triples extrapolated to the complete basis set) limit. We further evaluate the performance of ten density functional theory functionals spanning hybrid, dispersion-corrected, and long-range corrected categories against our benchmark dataset. Our results reveal that dispersion interactions contribute nearly 40% of the total attractive energy in arsine clusters, nearly equal to electrostatic contributions from weak As-H⋯As hydrogen bonds—a distinct contrast to light Group 15/16 hydride clusters where electrostatic interactions dominate. Cooperativity analysis reveals a striking transition from anti-cooperative behavior in the trimer to cooperative strengthening in the tetramer and larger clusters, driven by the relief of angular strain and enhanced dispersion contributions. Among all tested functionals, M06-2X achieves the lowest mean absolute deviation (0.60 kcal/mol per monomer) among the four structurally robust functionals, while CAM-B3LYP exhibits the most consistent performance with the smallest maximum deviation (7.10 kcal/mol). Although APFD yields a lower total binding energy MAD (2.56 kcal/mol), its failure to locate all isomeric structures limits its general applicability. This work fills a critical gap in the benchmark dataset for heavy main-group hydride clusters, and the recommended M06-2X and CAM-B3LYP functionals enable cost-effective, accurate simulations of large arsine systems relevant to semiconductor deposition, atmospheric arsenic nucleation, and toxic gas detection.