Computed Boys and Bernadi CP correction for the S66×8 benchmark set with HF, B3LYP, and BLYP with the basis sets (a) MINIS and (b) 6-31G*. The set consists of 66 systems (abscissa) with eight varying distances. The CP correction for each distance (ordinate) results in 528 data points.
Deviations between the Boys and Bernadi CP correction (ΔBB-CP) and the gCP correction (ΔgCP) for the S66×8 dimers in kcal/mol for B3LYP/6-31G* and HF/MINIS. The set consists of 66 systems (abscissa) with eight varying distances. The deviation (ΔgCP − ΔBB-CP) for each distance (ordinate) results in 528 data points. Positive values indicate an overestimation.
Potential energy curve of of a peptide-ethene complex (taken from the S66 set, depicted in the top right corner). The dimers are displaced along the vectors connecting their center-of-mass points by the multipliers 0.8, 0.9, 0.95, 1.00, 1.05, 1.10, 1.25, 1.50, 1.75, 2.00. The dispersion energy E disp from SAPT and DFT-D3 is plotted along with the Boys and Bernadi counterpoise (BB-CP) and gCP correction.
View of the folded (top) and extended (bottom) PM6-DH+ optimized (gas-phase) conformers of crambin.
Relaxed rotational energy profile of the FGF tripeptide in kcal/mol. The dihedral angle δ between the atoms indicated in black is varied in steps of 20° and held fixed in subsequent optimizations.
B97-D3 and HF-D3 interaction energies for 24 protein-ligand complexes with and without gCP correction. The LPNO-CEPA/1 and B97-D3/def2-TZVP reference values are taken from Ref. 132.
Comparison of BB-CP- and gCP-optimized structures with uncorrected ones for the S22 benchmark set. Deviation from the center-of-mass distances of the MP2 reference structures is plotted in Å. (a) HF-D3/SV, (b) B3LYP-D3/6-31G*, and (c) HF-D3/MINIS.
HF-D3/def2-QZVP optimized structure of 9-helicene. Two intramolecular distances of raw and gCP-corrected HF-D3/SV structures are compared with the shown reference structure.
TPSS-D3/def2-QZVP structure of the (RhCl(PH3))2 dimer.
Parameters for the gCP correction and fit quality (RMS deviation in kcal/mol) of the fit against Boys and Bernadi CP values for the S66×8 dimer geometries. The arithmetic mean of the BB-CP correction is also given (in kcal/mol).
Mean absolute deviations (MADs) for S66 benchmark set in kcal/mol. The column “+ gCP” indicates that the gCP correction is added to the DFT-D3 results. The methods are ordered according to increasing basis set quality.
Mean absolute deviations (MADs) for DFT-NL for the S66 benchmark set in kcal/mol. The column “+ gCP” indicates that the gCP correction is added to the DFT-NL results. The DFT-NL calculations have been performed non-self-consistently, except those indicated by (SC).
Mean absolute deviations (MADs) for S22 benchmark set in kcal/mol. The column “+ gCP” indicates that the gCP correction is added to the DFT-D3 or DFT-NL results. The DFT-NL calculations have been performed non-self-consistently.
Energy decomposition analysis of the interaction energy of the benzene-water complex (system 54) of the S66 benchmark set in kcal/mol. The “def2” label is omitted for clarity. The individual terms are electrostatic interaction ES, Pauli, or exchange repulsion EX, induction energy IND, charge-transfer energy CT, and the high-order-coupling term MIX. The gCP correction and the Boys and Bernadi counterpoise correction BB-CP are given. The BB-CP-correction only affects the EX,CT and MIX terms.
Selected systems from the WATER27 set as a test case for a cluster-type CP correction. ssf-CP and gCP values in kcal/mol are compared for the HF/MINIS and DFT(=B3LYP)/def2-SVP level of theory.
Relative energies (kcal/mol) of two crambin conformers (extended and folded) calculated with BLYP-D3 and various basis sets. Structures are optimized at the PM6-DH+ level of theory.
Intramolecular BSSE calculated for the FGF-tripeptide in kcal/mol. Comparison with Jensen's ACP approach.
Article metrics loading...
Full text loading...