Top Curr Chem (Z) (2018) 376:24
1 3
Møller–Plesset perturbation theory (MP2) levels. As shown in Fig. 7c, indeed,
the choice of geometry optimization method may have a significant effect on the
simulations of 2D electronic spectra, and it should be considered as another factor that could affect comparisons with experimental data.
Fig. 7 Benchmark computations of the adenine excited-state manifold (a), showing the 2DES maps in
the UV-pump/UV-probe window (2DUV–UV) as computed (b) at the CAS(12,10), RAS(12,10|2,4),
RAS(12,10|2,8) and RAS(12,10|2,12) levels of theory and using the ANO-L(432,21)-aug basis set, or
(c) at the RAS(12,10|2,12) level with the ANO-L(432,21)-aug basis set while using the GS geometry
optimized at the CAS(8,8)/6-31G*, MP2/ANO-L(432,21)-aug and DFT(B3LYP)/6-31G* levels. Note
that only the (variable) RAS3 spaces are indicated in the picture, for simplicity. Reproduced from data
reported in Ref. [57]
82
Reprinted from the journal
1 3
Møller–Plesset perturbation theory (MP2) levels. As shown in Fig. 7c, indeed,
the choice of geometry optimization method may have a significant effect on the
simulations of 2D electronic spectra, and it should be considered as another factor that could affect comparisons with experimental data.
Fig. 7 Benchmark computations of the adenine excited-state manifold (a), showing the 2DES maps in
the UV-pump/UV-probe window (2DUV–UV) as computed (b) at the CAS(12,10), RAS(12,10|2,4),
RAS(12,10|2,8) and RAS(12,10|2,12) levels of theory and using the ANO-L(432,21)-aug basis set, or
(c) at the RAS(12,10|2,12) level with the ANO-L(432,21)-aug basis set while using the GS geometry
optimized at the CAS(8,8)/6-31G*, MP2/ANO-L(432,21)-aug and DFT(B3LYP)/6-31G* levels. Note
that only the (variable) RAS3 spaces are indicated in the picture, for simplicity. Reproduced from data
reported in Ref. [57]
82
Reprinted from the journal