The S66 noncovalent interactions benchmark reconsidered using explicitly correlated methods near the basis set limit

摘要

The S66 benchmark for noncovalent interactions has been re-evaluated usingexplicitly correlated methods with basis sets near the one-particle basis setlimit. It is found that post-MP2 "high-level corrections" are treatedadequately well using a combination of CCSD(F12*) with (aug-)cc-pVTZ-F12 basissets on the one hand, and (T) extrapolated from conventionalCCSD(T)/heavy-aug-cc-pV{D,T}Z on the other hand. Implications for earlierbenchmarks on the larger S66x8 problem set in particular, and for accuratecalculations on noncovalent interactions in general, are discussed. At a slightcost in accuracy, (T) can be considerably accelerated by using sano-V{D,T}Z+basis sets, while half-counterpoise CCSD(F12*)(T)/cc-pVDZ-F12 offers the bestcompromise between accuracy and computational cost.