CONCULUSION
The benchmark of DFT for α-diimine palladium catalyzed insertions of
ethylene, methyl acrylate, and vinyl bromide has been studied, during
which sixty-seven functional methods have been screened on the basis of
experimental kinetic data.
For the ethylene/methyl acrylate copolymerization featuring the
insertions of the both monomers, when the single-point calculation is
performed at the level of M05-2X(SMD)/6-31G(d)∩SDD, 11 hybrid
functionals (M06, BHandH, mPW1PW91, HSEh1PBE, mPW3PBE, LC-ωPBE, PBE0,
M06-HF, M06-2X, mPW1PBE, M05-2X), 5 functionals with D3 dispersion
correction (BPBE-D3, M062X-D3,
PBE0-D3,
PBEPBE-D3, LC-ωPBE-D3), and 3 functionals with D3BJ dispersion
correction (B3PW91-D3BJ, PBE0-D3BJ, PBEPBE-D3BJ) used for geometrical
optimization could give better results compared with other tested
functionals. However, GGA and some meta -GGA functionals showed
poor performance in the geometry optimization with respect to calculated
free energy barrier in solution (SMD model). The hybrid functional B1B95
produced the worst result. When the CPCM model is applied at the level
of PBE0(CPCM)/6-31+G(d)∩SDD, the aforementioned 11 hybrid functionals
except M06-HF also showed good performance. At this single-point
calculation level, the M06, M062X-D3, LC-ωPBE-D3, and PBE0-D3BJ
geometries can also produce accurate free energy barrier in solution.
For the copolymerization ethylene and vinyl bromide, when the
single-point calculation is carried out at the level of
PBE0(CPCM)/6-31+G(d)∩SDD, the tested hybrid functionals except BHandH,
TPSSh, and M06-2X perform well. However, the GGA and meta -GGA
functionals have poor performance, like the case of ethylene/MA
copolymerization. And, LC-ωPBE-D3 overestimated the energy barrier of
vinyl bromide insertion, possibly due to the over-correction of
dispersion force. At the level of M05-2X(SMD)/6-31G(d)∩SDD for
single-point calculations, the GGA and meta -GGA still have poor
performance. The dispersion correction augmented PBEPBE functional
underestimated the barrier and showed worst performance among the tested
functionals. The hybrid functionals except M06, TPSSh, and M06-HF
perform well at this level.
It is noted that the performance of the common B3LYP functional with or
without dispersion is not ideal in these systems. Compared with SMD
solvation model, CPCM model slightly underestimated the energy barrier
of ethylene insertion, but it is still within the acceptable error
range.
ASSOCIATED CONTENT
Supporting Information
Additional table and figures showing the effect of dispersion
corrections on the error of ethylene insertion energy barrier and
monomer insertion energy barriers calculated at various levels together
with SMD and CPCM solvation models as well as Cartesian coordinates of
optimized stationary points. The Supporting Information is available
free of charge on the ACS Publications website.
AUTHOR INFORMATION
Corresponding Author
* Y.L. (luoyi@dlut.edu.cn)
Notes The authors declare no competing financial interest.