Chemistry Reference
In-Depth Information
experiment nor that from a gas-phase ab initio calculation. A dominance of
delocalized bonding via the bridging ligand was thereby suggested [
77
].
The S% values comply with a delocalized picture of sources when the Mn-Mn
mp is taken as rp. The two Mn atoms are found to determine a small negative
contribution to the electron density at the Mn-Mn mp, with S% from each Mn atom
being about
8.5%. More than 50% of the density at the mp is instead determined
by the carbonyl groups. The B atom also serves as a sink (S%(B)
11.6%), as
likely the result of the polarization of its Laplacian density due to bonding to the
more electronegative C atom.
26
One anticipates, however, a large and positive S%
contribution from the global B
t
Bu ligand, analogously to that found for the bridging
carbonyls in the dinuclear carbonyl complexes, though an exact estimate of such a
cumulative contribution from the ligand cannot be exactly deduced from [
77
] (the
expected small S% contribution from the two Cp groups is not quoted).
A comparison between the LS profiles in Mn
2
(CO)
10
and [{Cp(CO)
2
Mn}
2
ΒΌ
(
-B
t
Bu)], along the Mn-Mn internuclear axis and with the Mn-Mn mp taken as a rp,
was also not reported in the paper, though it might provide interesting insights. One
could likely observe a comparatively larger drop of the negative LS around the mp
in the borylene complex. Indeed Table
2
of [
77
] reports a positive Laplacian value
at the Mn-Mn mp for the Mn
2
(CO)
10
compound which is halved with respect to the
one in the borylene complex, and the Laplacian is found to become even slightly
negative for Mn
2
(CO)
10
using the theoretical density. Hence, analogously to the LS
profiles comparison reported earlier for the unbridged and the double-bridged
Co
2
(CO)
8
compounds, a close examination of the mentioned Mn-Mn LS profiles
would allow to provide further evidence for indirect metal-metal bonding in the
borylene complex.
27
A parallel paper by Goetz et al. [
78
], with some common authors to [
77
], has
examined the theoretical electron densities in a number of gas-phase dinuclear
manganese complexes. These include, among other, two borylene-bridged com-
pounds, one of which coincides with that studied experimentally by Flierler et al.
[
77
] in the crystalline state. Both papers raise the important question of whether the
bridging boron ligand may be better described as a substituted borane or as a true
bridging borylene with an electron lobe from the B atom being directed at the
Mn-Mn mp. Goetz et al. [
78
] found that the preference between the two bonding
situations shows a large dependence on the adopted theoretical level. As a function
of the exact-exchange admixture in the DFT functional, the boron bridge may
involve either delocalized 3c bonding across the bridge or bonding through two
m
26
Indeed, in the dinuclear carbonyl complexes, the large S% contribution from the CO ligands to
the M-M mp or bcp is due to the dominant SF contribution from the indirectly bonded but more
electronegative O atom, rather than from the directly bonded, but relatively less electronegative, C
atom that exhibits negligible or sometimes even negative S% values (see for instance, Fig.
8
).
27
Note that these LS profiles comparisons are meaningful only because the bridged and unbridged
forms share similar M-M distances. If it were not so, comparison of the Laplacian values and of LS
profiles should take into account that the atomic Laplacian is smoothly decreasing in value with
increasing distance from the metal.
