Communication: Resolving the three-body contribution to the lattice energy of crystalline benzene: Benchmark results from coupled-cluster theory

Abstract

Coupled-cluster theory including single, double, and perturbative triple excitations [CCSD(T)] has been applied to trimers that appear in crystalline benzene in order to resolve discrepancies in the literature about the magnitude of non-additive three-body contributions to the lattice energy. The present results indicate a non-additive three-body contribution of 0.89 kcal mol −1 , or 7.2% of the revised lattice energy of −12.3 kcal mol −1 . For the trimers for which we were able to compute CCSD(T) energies, we obtain a sizeable difference of 0.63 kcal mol −1 between the CCSD(T) and MP2 three-body contributions to the lattice energy, confirming that three-body dispersion dominates over three-body induction. Taking this difference as an estimate of three-body dispersion for the closer trimers, and adding an Axilrod-Teller-Muto estimate of 0.13 kcal mol −1 for long-range contributions yields an overall value of 0.76 kcal mol −1 for three-body dispersion, a significantly smaller value than in several recent studies.

Density-functional approach to the three-body dispersion interaction based on the exchange dipole moment

“…Table I contains the C 6 and C 9 estimates of a set of well-separated homonuclear triplets of H, N, O, F, Si, P, S, Cl, Ne, He, Ar, Kr using aug-cc-PVQZ, G3LARGE 26 (in round brackets) and 6-311+G(3df,3p) (in square brackets). There is a small to moderate dependence of the results on the basis set (percentage deviations in single digits) with the exception of H and He triplets where the basis set dependence is more noticeable, especially for C 9 .…”

“…The atoms in those triplets are not close in space and thus are not suitable for the optimization of a damping formula. Very recently, Sherrill and Co. 9 have reported accurate coupled-cluster theory values of the threebody interaction in crystalline benzene. They proposed that the non-additive three-body dispersion component can be deduced by subtracting the Møller-Plesset perturbation theory (MP2) energy from the energy of the CCST(T).…”

“…The significance of the non-additive three-body dispersion term is still not quite clear and is being debated. [7][8][9] It was estimated to be as large as 10% of the stabilization energy in some molecular crystals 10 and about 46% of the energy difference between folded and elongated polyalanine decamers. 7 The three-body dispersion contribution was also shown to be crucial to the a) Author to whom correspondence should be addressed.…”

Intra‐ and Intermolecular Dispersion Interactions in [n]Cycloparaphenylenes: Do They Influence Their Structural and Electronic Properties?

“…(1). Note that a satisfactory agreement has been recently found between highly accurate CCSD(T) (and thus many-body) calculations and this correction for trimers of crystalline benzene [50]. c) Another way to incorporate part of the missing non-covalent interactions is to rely on a nonlocal (-NL) correction to the electronic energy in the form:…”

Noncovalent Interactions in Organic Electronic Materials

“…We note that the intermolecular angles found in the oligoacene crystal structures (approximately 50° to 80° depending on the oligoacene) lie near the least stable intermolecular interaction energies. This points to a limitation in the current analysis, suggesting that important terms are not fully accounted for in a two-body SAPT0 approach; [129][130][131][132] thus, models aiming to predict crystal packing need to go beyond such approaches. 133 The analysis of the non-covalent interaction energies of oligoacene dimers reveals the complexity of molecular crystal engineering, with the small energy differences among rather different configurations providing a clear picture of why polymorphism is such a common phenomenon in these systems.…”