Another Angle on Benchmarking Noncovalent Interactions

For noncovalent interactions, the CCSD­(T)-coupled cluster method is widely regarded as the “gold standard”. With localized orbital approximations, benchmarks for ever larger complexes are being published, yet FN-DMC (fixed-node quantum Monte Carlo) intermolecular interaction energies diverge to a progressively larger degree from CCSD­(T) as the system size grows, particularly when π-stacking is involved. Unfortunately, post-CCSD­(T) methods like CCSDT­(Q) are cost-prohibitive, which requires us to consider alternative means of estimating post-CCSD­(T) contributions. In this work, we take a step back by considering the evolution of the correlation energy with respect to the number of subunits for such π-stacked sequences as acene dimers and alkadiene dimers. We show it to be almost perfectly linear and propose the slope of the line as a probe for the behavior of a given electron correlation method. By going further into the coupled cluster expansion and comparing with CCSDT­(Q) results for benzene and naphthalene dimers, we show that CCSD­(T) does slightly overbind but not as strongly as suggested by the FN-DMC results.