Angewandte
Communications
Chemie
carbon atoms, in either m-h2:h2 or m-h1:h1 bonding modes,[27,28]
highlighting the differences between these isosteres.[29]
Surprisingly, the amino-borane in 4 is quite strongly
bound. It is only slowly displaced by excess acetonitrile (7%
in 50 min) to give a mixture of species, one of which is
[Rh(LiPr)(NCMe)2][BArF ].[22] No reaction occurs with tolu-
4
ene,
which
might
be
expected
to
form
a
[Rh(LiPr)(h6-C6H5Me)]+ complex if a monomeric {Rh(LiPr)}+
fragment were accessible.[30] Addition of cyclohexene, shown
to be a probe for free H2B NH2, gave no reaction. In
[2]
=
contrast,
H2
rapidly
reacts
with
4
to
form
[Rh2(LiPr)2(H)2(m-H)3][BArF ].[22]
4
There are two limiting forms for the structure of 4 (and
quasi-isostructural 3): 1) a bridging amino-borane at two RhI
centers, or 2) a bridging borylene dihydride (RhIII), Scheme 3.
The observed d(11B) chemical shift of 51 ppm is more
consistent with the former as amino-boranes bound to one
metal center show chemical shifts around 40–50 ppm,[12,13,24,31]
while bridging borylenes[32] are generally observed between
90 and 100 ppm.[23,33]
To probe the bonding of the amino-borane ligand in 4,
DFT calculations were used as the basis for a Quantum
Theory of Atoms in Molecules (QTAIM) analysis of the total
electron density. The results are presented in Figure 2A,
along with selected bond critical point (BCP) metrics. Fig-
ure 2B provides comparative BCP data for the bridging
borylene complex C, the hydridoborate complex D, and
Figure 2. A) Contour plot of the electron density of the central part of
4 presented in the {Rh1B1Rh2} plane with projected stationary points,
bond paths, bond critical points (BCP; green), and ring critical points
(RCP; red); the associated table shows selected BCP metrics (a.u.;
average data for indicated bonds) and computed d(11B) chemical
shifts. B) Calculated BCP metrics (a.u.; average data for indicated
[(PPh3)2Rh(H)(m-H)(m-Cl)2Rh(H)(PPh3)2]+, E,
a
well-
defined RhIII dimer with both terminal and bridging
hydrides.[34] Average data are presented for all complexes
where appropriate, although the discussion will focus on the
bonding around a single rhodium center (Rh1).
bonds) for comparator complexes C (including the computed 11
B
chemical shift), D and E (1(r)=electron density, 51(r)=Laplacian of
electron density, e=bond ellipticity, H(r)=local energy density). All
geometries are based on the crystallographically determined heavy
atom positions with hydrogen atoms optimized with the BP86 func-
tional. For a full summary of parameters see Figures S24–27 and
associated Tables in the Supporting Information.
In 4, the {Rh1/B1/H1} moiety displays bond paths
between all three centers, and these enclose a ring critical
À
À
point (RCP). Thus, 4 has direct Rh1 B1 and Rh1 H1
À
bonding interactions, while the B1 H1 bond is also intact.
À
Comparison with the Rh1 B1 interaction in C provides
similar 1(r) and H(r) values, but highlights a much reduced
bond ellipticity (e) of 0.08; this low value indicates dominant
s-bond character, whereas the value of 0.47 in 4 reflects the
asymmetry introduced by the B1-H1 unit. In D, the absence of
Rh-B BCPs confirms a lack of any direct Rh-B interaction,
ligand in 4 interacts with the rhodium centers through
À
stretched B H bonds that engage in strong Rh-H and Rh-B
interactions. Further support for this assertion comes from the
computed d(11B) chemical shifts (Figure 2) and the
Pipek–Mezey localized orbitals, where a strong bonding
interaction spanning all three Rh1, B1, and H1 centers was
identified (see Figure 3).
À
and this also reduces the average ellipticity of the Rh1 H1
À
and B1 H1 bonds. Also noticeable are the higher values of
p(r) and H(r) for the terminal B1-H4 bond in D compared to
the bridging B-H bonds in both that structure and, in
particular, 4, all of which is consistent with a weakening of
the latter. For E, the Rh1-H1 BCP has larger values for 1(r)
and H(r) than the Rh1-H1 BCP in 4, as well as a minimal
The mechanism of the room temperature fluxional
process observed for 4 was also probed with DFT calculations
and a single transition state was found to account for this
process (Scheme 4). This is accessed by cleavage of one (blue)
B-H bond to give a transition state structure featuring two
Rh-H-Rh bridging hydrides; movement of the original (red)
Rh-H-Rh hydride into a Rh-H-B bridging position then
completes the exchange (4’). Repeating this process from 4’
exchanges a second B-H hydrogen (black) into the Rh-H-Rh
bridging position (4’’). The computed free energy of activa-
tion is 55.2 kJmolÀ1, somewhat higher than the experimental
value (39.2 Æ 1.6 kJmolÀ1) but still consistent with facile room
temperature exchange.
À
e value. These data indicate a terminal Rh H s-bond and
stress the differences in bridging character of H1 and H2 in 4.
À
À
BCP data for the Rh1 H3 Rh2 bonds in 4, D, and E are very
similar, suggesting that this moiety varies little across these
three systems.
Taken together, the QTAIM analyses suggest that 4 is best
described as a m-amino-borane RhI species; a m-borylene
hydride RhIII formulism can certainly be ruled out in light of
À
À
À
the intact B1 H1/B1 H2 bonds and the lack of Rh1 H1/
Understanding how bimetallic species such as 3 and 4 are
formed, and subsequently react, is important for delineating
À
Rh2 H2 terminal hydride character. The m-amino-borane
Angew. Chem. Int. Ed. 2016, 55, 6651 –6656
ꢀ 2016 The Authors. Published by Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
6653