Q-Chem also supports the non-perturbative 1155 (variational) energy and charge decomposition analysis in the second generation ALMO-EDA framework. The advantage of this method over the perturbative CT analysis method is that the energy and charge decompositions are both exact and there are no higher-order terms left. Currently, this method is implemented for restricted SCF calculations only.
Set GEN_SCFMAN_EDA2 = 1 and EDA_VCT_A = 1 to use the non-perturbative CT analysis method (note: this feature is currently not compatible with jobs with EDA2 0 and needs to be performed as a separate calculation) The result will print out the total energy and charge change in the charge transfer process, as well as the pairwise energy and charge changes between fragments, with fragment labels starting from 0. Set EDA_COVP_THRESH = to print out COVPs that accounts for more than % of the energy or charge change between each fragment pair, whose default value is set to 5. Set EDA_PCT_A = 1 to do the perturbative CT analysis as well. To visualize the COVPs, set EDA_SAVE_COVP = 1 and GUI = 2. The saved COVPs will be written to an .fchk file, with the occupied orbitals called “Localized Alpha MO Coefficients (Boys)" and the corresponding virtual orbitals called “Localized Beta MO Coefficients (Boys)" (note: this is a workaround and visualization of COVPs in .fchk files will be enabled in future releases). One caveat is that the current implementation does not support the OCC_RI_K algorithm, and it will give wrong results if this REM is set true.
$molecule 1 1 -- 0 1 H -0.73946 0.94887 0.78379 O -1.16910 0.63297 -0.02844 H -2.12156 0.70793 0.14730 -- 1 1 Na -0.17266 -0.04338 -1.86190 $end $rem method = b3lyp sym_ignore = true symmetry = false basis = 6-31g* scf_algorithm = diis thresh = 12 incfock = 0 mem_total = 16000 scf_convergence = 9 child_mp = 1 child_mp_orders = 232 ! nDQ gen_scfman_eda2 = 1 eda_pct_a = 1 eda_vct_a = 1 eda_covp_thresh = 5 eda_save_covp = 1 gui = 2 $end