CPCM model
To test the performance of CPCM solvation model in such copolymerization
systems, single-point solvation calculations have been performed at the
level of PBE0/6-31+G(d), which was shown to be suitable for CPCM
model.34 As shown
in Table 3, in the case of ethylene insertion, all the tested
functionals underestimated the reaction energy barrier, among which the
hybrid M06-HF performs the worst with an error of 2.3 kcal/mol. For MA
insertion, all the functionals shown in Table 3 perform well. In the
case of VB insertion, like the case of SMD model, GGA andmeta -GGA functionals underestimated the reaction energy barrier
with errors of larger than 1.0 kcal/mol. The hybrid functional LC-ωPBE
with D3 dispersion correction overestimated the barrier (17.9 vs15.9±0.1 kcal/mol). This can be ascribed to that the functional LC-ωPBE
itself considered the long-range interaction, and further D3 dispersion
correction may lead to overcorrection. The hybrid functional BHandH also
overestimated the energy barrier (17.3 kcal/mol). When larger basis set
6-31+G(d,p) was used for the non-metal atoms, similar results was
obtained (Table S1). Therefore, with the CPCM solvent model, the lower
theoretical level of 6-31+G(d) for single-point calculation is
sufficient to obtain accurate energy barriers in such copolymerization
system. It is noteworthy that, although some functionals produced errors
of > 1.0 kcal/mol, the errors are not so large (<
3 kcal/mol), suggesting a wide range of functionals suitable for dealing
with such (α-diimine)Pd-catalyzed copolymerizations.