Despite the successful examples commented above, and others that can be found
in the literature, we are still far from a regular inclusion of additives in the chemical
models of organometallic systems.
4.2 Conformational Complexity
Even when there is no speciation problem and the reactive species are fully identified
another important issue for the proper chemical modeling of the system is the
recognition that most species have a large variety of conformations and that it is
essential to consider the appropriate ones in the study. This was not a problem in the
early days of computational mechanistic studies, when simple models were used
such as PH 3 or PMe 3 for any experimentally used phosphine. But nowadays, when
the actual full ligands are customarily included in the computed systems, one needs
to have confidence that the relevant (lowest energy) conformer(s) are considered.
Conformational diversity is often neglected in computational studies of transition
metal complexes, even when relatively large systems are present. The importance of
conformational search in mechanistic investigations was illustrated through the
analysis of the errors that could be caused by a wrong choice of conformers in the
computational study of the Suzuki–Miyaura cross-coupling between CH 2 ¼ CHBr
and CH 2 ¼ CHB(OH) 2 catalyzed by Pd(PPh 3 ) 2] or Pd(P
i Pr 3 ) 2 [57]. Figure 9 displays
the Gibbs energy profile for the oxidative addition step to the Pd(P
i
Pr 3 ) 2 complex,
highlighting the energy differences between the least and the most stable conformer
of each of the species present in the reaction pathway. In all cases the error bar is
about 10 kcal mol
À1 , leading to wide oscillations of the computed barrier for
randomly chosen conformations.
A recent study has emphasized the importance and necessity of conformational
analysis for appropriate choice of rotamer prior to any further mechanistic study
[58]. A careful conformational search shows that the energy profile for the CO 2
insertion into a nickel hydride bond of POCOP
i
Pr nickel hydride complex can be
either exergonic or endergonic depending on the rotamer choice. The POCOP
i Pr
Ni-formato complex product of the CO 2 insertion reaction has an energy difference
between the lowest and highest energy rotamers as high as 16.8 kcal mol
À1
[58]. Moreover, not only do conformer effects modify particular reaction barriers,
but often the lowest barrier reaction pathway proceeds from a conformer that is not
the lowest energy conformer, as evinced in the DFT study of C–C bond forming
elementary step (reductive elimination) at Ni bisphosphine catalysts with varying
phosphine side chains [59]. These studies show that errors resulting from random
selection of conformers can be of the same order or magnitude, or even larger, that
the ones coming from a poor choice of DFT functional or the use of a small basis set.
Conformational sampling implies generating a large number of conformations
whose structure needs to be optimized for allowing energy ranking. When the
system becomes large the conformational space grows, and this can be a troublesome
and time-consuming work. Thus, different strategies have been devised for
What Makes a Good (Computed) Energy Profile?
15
in the literature, we are still far from a regular inclusion of additives in the chemical
models of organometallic systems.
4.2 Conformational Complexity
Even when there is no speciation problem and the reactive species are fully identified
another important issue for the proper chemical modeling of the system is the
recognition that most species have a large variety of conformations and that it is
essential to consider the appropriate ones in the study. This was not a problem in the
early days of computational mechanistic studies, when simple models were used
such as PH 3 or PMe 3 for any experimentally used phosphine. But nowadays, when
the actual full ligands are customarily included in the computed systems, one needs
to have confidence that the relevant (lowest energy) conformer(s) are considered.
Conformational diversity is often neglected in computational studies of transition
metal complexes, even when relatively large systems are present. The importance of
conformational search in mechanistic investigations was illustrated through the
analysis of the errors that could be caused by a wrong choice of conformers in the
computational study of the Suzuki–Miyaura cross-coupling between CH 2 ¼ CHBr
and CH 2 ¼ CHB(OH) 2 catalyzed by Pd(PPh 3 ) 2] or Pd(P
i Pr 3 ) 2 [57]. Figure 9 displays
the Gibbs energy profile for the oxidative addition step to the Pd(P
i
Pr 3 ) 2 complex,
highlighting the energy differences between the least and the most stable conformer
of each of the species present in the reaction pathway. In all cases the error bar is
about 10 kcal mol
À1 , leading to wide oscillations of the computed barrier for
randomly chosen conformations.
A recent study has emphasized the importance and necessity of conformational
analysis for appropriate choice of rotamer prior to any further mechanistic study
[58]. A careful conformational search shows that the energy profile for the CO 2
insertion into a nickel hydride bond of POCOP
i
Pr nickel hydride complex can be
either exergonic or endergonic depending on the rotamer choice. The POCOP
i Pr
Ni-formato complex product of the CO 2 insertion reaction has an energy difference
between the lowest and highest energy rotamers as high as 16.8 kcal mol
À1
[58]. Moreover, not only do conformer effects modify particular reaction barriers,
but often the lowest barrier reaction pathway proceeds from a conformer that is not
the lowest energy conformer, as evinced in the DFT study of C–C bond forming
elementary step (reductive elimination) at Ni bisphosphine catalysts with varying
phosphine side chains [59]. These studies show that errors resulting from random
selection of conformers can be of the same order or magnitude, or even larger, that
the ones coming from a poor choice of DFT functional or the use of a small basis set.
Conformational sampling implies generating a large number of conformations
whose structure needs to be optimized for allowing energy ranking. When the
system becomes large the conformational space grows, and this can be a troublesome
and time-consuming work. Thus, different strategies have been devised for
What Makes a Good (Computed) Energy Profile?
15