Smooth C(alkyl)-H bond activation in rhodium complexes comprising abnormal carbene ligands

Article abstract of DOI:10.1039/c1dt11116g

Rhodation of trimethylene-bridged diimidazolium salts induces the intramolecular activation of an alkane-type C-H bond and yields mono- and dimetallic complexes containing a formally monoanionic C,C,C-tridentate dicarbene ligand bound to each rhodium centre. Mechanistic investigation of the C_alkyl-H bond activation revealed a significant rate enhancement when the carbene ligands are bound to the rhodium centre via C4 (instantaneous activation) as compared to C2-bound carbene homologues (activation incomplete after 2 days). The slow C-H activation in normal C2-bound carbene complexes allowed intermediates to be isolated and suggests a critical role of acetate in mediating the bond activation process. Computational modelling supported by spectroscopic analyses indicate that halide dissociation as well as formation of the agostic intermediate is substantially favoured with C4-bound carbenes. It is these processes that discriminate the C4- and C2-bound systems rather than the subsequent C-H bond activation, where the computed barriers are very similar in each case. The tridentate dicarbene ligand undergoes selective H/D exchange at the C5 position of the C4-bound carbene exclusively. A mechanism has been proposed for this process, which is based on the electronic separation of the abnormal carbene ligand into a cationic N-C-N amidinium unit and a metalla-allyl type M-C-C fragment.

Full text of DOI:10.1039/c1dt11116g

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Dalton
Transactions
Cite this: Dalton Trans., 2011, 40, 9911
www.rsc.org/dalton
PAPER
Smooth C(alkyl)–H bond activation in rhodium complexes comprising
abnormal carbene ligands†
Anneke Kru¨ger,a L. Jonas L. Ha¨ller,b Helge Mu¨ller-Bunz,a Olha Serada,c Antonia Neels,c
Stuart A. Macgregor*b and Martin Albrecht*a
Received 14th June 2011, Accepted 27th July 2011
DOI: 10.1039/c1dt11116g
Rhodation of trimethylene-bridged diimidazolium salts induces the intramolecular activation of an
alkane-type C–H bond and yields mono- and dimetallic complexes containing a formally monoanionic
C,C,C-tridentate dicarbene ligand bound to each rhodium centre. Mechanistic investigation of the
Calkyl–H bond activation revealed a significant rate enhancement when the carbene ligands are bound to
the rhodium centre via C4 (instantaneous activation) as compared to C2-bound carbene homologues
(activation incomplete after 2 days). The slow C–H activation in normal C2-bound carbene complexes
allowed intermediates to be isolated and suggests a critical role of acetate in mediating the bond
activation process. Computational modelling supported by spectroscopic analyses indicate that halide
dissociation as well as formation of the agostic intermediate is substantially favoured with C4-bound
carbenes. It is these processes that discriminate the C4- and C2-bound systems rather than the
subsequent C–H bond activation, where the computed barriers are very similar in each case. The
tridentate dicarbene ligand undergoes selective H/D exchange at the C5 position of the C4-bound
carbene exclusively. A mechanism has been proposed for this process, which is based on the electronic
separation of the abnormal carbene ligand into a cationic N–C–N amidinium unit and a metalla-allyl
type M–C–C fragment.
Two major pathways have been put forward for Calkyl–H
Introduction
bond activation,4f,4i,6 and they involve either (complex-assisted)
s-bond metathesis7 or oxidative additions.8 The former is perhaps
most prominently illustrated with Hartwig’s rhodium(III) catalyst
The selective activation of inert bonds, in particular C–H bonds
in alkanes and related compounds, has significant implications
in many different areas of chemistry.1 Foremost, mild routes
to Calkyl–H bond activation provide access to unprecedented
retrosynthetic opportunities in organic synthesis2 and yield highly
desirable processes for a more efficient fossil fuel utilisation.3
Although promising advances have been made in homogeneous
transition metal-mediated activation of unfunctionalised C–H
bonds,4 drawbacks persist such as the limited substrate scope,
or the harsh reaction conditions required to achieve substantial
turnover frequencies, and industrial processes are relatively rare.5
for mild and selective alkane borylation.9 Most other alkane
activation catalysts operate, however, along the oxidative addition
route and include the insertion of a metal centre into the C–H
bond.4 Recent calculations have suggested that the metal centre
for oxidative addition may be either nucleophilic, electrophilic,
or amphiphilic.10 A further variation is ambiphilic metal–ligand
assisted (AMLA) Calkyl–H bond activation involving chelating
bases, and this is thought to be widely operative in synthesis.11
A conceivable approach to develop new systems for (catalytic)
Calkyl–H bond activation thus consists of using electron-rich tran-
sition metals that are bound by strong electron donor ligands, thus
providing a high level of nucleophilicity to promote C–H oxidative
addition. Ligands derived from N-heterocyclic carbenes (NHCs)
may be particularly useful spectator ligands for such application,
as NHCs entail high s-basicity and low p-acidity paired with
a rather covalent bonding to the metal centre.12 Late transition
metals comprising NHC ligands have indeed been reported to acti-
vate unreactive bonds.13 Further enhancement of the donor ability
imparted by the NHC ligand may be achieved upon changing the
ligand coordination. Substantially stronger donor properties were
noted for NHCs bound via C4 rather than C2,14 often referred
aSchool of Chemistry and Chemical Biology, University College Dublin,
Belfield, Dublin 4, Ireland. E-mail: martin.albrecht@ucd.ie; Fax: +353-
17162501; Tel: +353-17162504
bSchool of Engineering and Physical Sciences, William Perkin Build-
ing, Heriot-Watt University, Edinburgh, U.K., EH14 4AS. E-mail:
s.a.macgregor@hw.ac.uk
cXRD Application Lab, CSEM, Rue Jaquet-Droz 1, CH-2002 Neuchaˆtel,
Switzerland
† Electronic supplementary information (ESI) available: Crystallographic
and computational details and computed structures and energies of
all species. CCDC reference numbers 817392–817395. For ESI and
crystallographic data in CIF or other electronic format see DOI:
10.1039/c1dt11116g
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 9911–9920 | 9911
©

View Article Online
to as abnormal bonding.15 In this coordination mode, the metal-
bound carbon experiences a lower s-withdrawing effect due to the
presence of only one nitrogen atom in the a position. Indeed,
abnormally bound carbene complexes have shown intriguing
catalytic activity in bond activation processes, surpassing in many
instances that of normal carbene analogues.16 Within this context,
we recently communicated the smooth intramolecular activation
of an alkane-type C–H bond in a rhodium complex featuring
abnormally bound NHC ligands.17 Upon expanding on these
studies we have identified key parameters that govern the C–H
bond activation step. Bond activation in these abnormal carbene
complexes presumably is a chelate-assisted process, thus disclosing
an additional feature of this particular ligand.
Formation of either the monometallic complexes 2 or the
dimetallic complexes 3 was dependent on the coordination
ability of the solvent. Purification of the complexes by column
chromatography using CH2Cl2 and acetone as eluents gave the
dimetallic species 3, whereas addition of MeCN afforded the
monometallic complexes 2 exclusively. All monomeric complexes
displayed two sets of 1H NMR signals in solution (CD3CN), which
have been attributed to two isomers featuring the iodide ligand
trans to the alkyl ligand and trans to one of the carbene ligands,
respectively. Generally, the signals appear in approximate 2 : 1
ratio, which indicates a statistical distribution of the iodide ligand
over all three available positions, and thus a similar trans influence
of the abnormal carbene and alkyl ligands. In all complexes but
2b, the major isomer of 2 revealed significant broadening of the
carbene resonances both in the 1H and 13C NMR spectra, while in
the minor species, the trimethylene resonances were broad. This
behaviour was assigned to fluxional coordination of the iodide
ligand, either trans to one of the carbene ligands (major isomer)
or trans to the alkyl ligand (minor isomer). In complex 2b the ratio
is inverted (1 : 1.2) with the major species featuring a well-resolved
carbene resonance at dC 146 (1JRhC = 46 Hz), while the minor
species, attributed to the isomer with a mutual cis arrangement
of iodide and alkyl ligands, displays broad carbene signals and
a sharp doublet for the rhodium-bound alkyl carbon (dC 34.0,
1JRhC = 28.5 Hz). For the dimeric complexes 3, two major sets
of signals in approximately 2 : 1 ratio were identified by NMR
spectroscopy. These sets were tentatively assigned to syn and anti
eclipsed conformations of the two tridentate ligands. The protons
of the trimethylene linker and likewise the NCH2 protons of the
butyl wingtip groups in 3a and 3d are diastereotopic, in line with
a hindered rotation about the Rh–C bonds.19
Results and discussion
C–H bond activation in C2-substituted diimidazolium salts
Reaction of the trimethylene-linked diimidazolium salts 1 with
hydrated RhCl3 in the presence of NaOAc yielded the tridentate
dicarbene complexes 2, and after column chromatography the
dimetallic complexes 3 (Scheme 1). Formation of 2 implies a triple
C–H bond activation process including activation of the alkyl
C–H bond in the linker in addition to the expected18 heterocyclic
C–H bond activation. The activation of the linker occurred under
relatively mild conditions, i.e. in moist MeCN at 82 ◦C and in
an aerobic atmosphere. Variation of the substituents within the
heterocycle (aryl vs. alkyl group at C2 and N1, bulky iPr vs.
linear alkyl groups at N1) did not have a significant effect on
the bond activation process and the ligand consistently adopted
a tridentate bonding mode. No intermediate was detected when
stopping the reaction at low conversion. Hence the aliphatic C–H
bond activation is a general process, provided the rhodium centre is
bound by abnormal carbene ligands. An abnormal bonding mode
is enforced in 1a–d by protection of the imidazolium C2 position.
The dimeric structure of 3b was unambiguously confirmed by
an X-ray diffraction analysis of a single crystal that was grown
after exchange of the non-coordinating anion from I- to PF6 .
-
The molecular structure is almost C2v-symmetric (Fig. 1). The
Fig. 1 ORTEP representation of the cationic portion of 3b (30%
probability). Hydrogen atoms and co-crystallised solvent atoms have
˚
been omitted for clarity. Selected bond lengths (A): Rh1–C1 1.974(17),
Rh1–C7 1.939(18), Rh1–C5 2.075(15), Rh1–I1 2.8052(16), Rh1–I2
2.7808(17), Rh1–I3 2.7934(19). Selected bond angles (deg): C1–Rh1–C7
85.6(7), C1–Rh1–C5 83.4(6), C5–Rh1–C7 84.1(7), C1–Rh1–I2 174.1(5),
C5–Rh1–I1 178.0(5), C7–Rh1–I3 175.4(5); the environment around the
Rh2 centre is similar.
Scheme 1 Metallation of a trimethylene-linked diimidazolium salt: (i)
RhCl3, NaOAc, MeCN (reflux), then KI; (ii) CH2Cl2, RT; (iii) MeCN, RT.
9912 | Dalton Trans., 2011, 40, 9911–9920
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Ccarbene–Rh–Ccarbene bite angles of the two crystallographically
independent sites are slightly different (85.6(7)◦ and 89.7(8)◦ re-
spectively), which suggests some steric flexibility of the chelate. As
reported previously for 3c,17 the Rh–Ccarbene bonds are substantially
˚
shorter than the Rh–Calkyl bonds (average 1.94(3) vs. 2.074(1) A),
˚
yet the Rh–I bonds are all similar (average Rh–I 2.80(2) A). These
data suggest similar trans influences of alkyl and carbene ligands,
in line with the statistical mixture of isomers found in solution
(see above).
Normal and mixed normal–abnormal dicarbene complexes
In order to evaluate effects that are critical for inducing the smooth
Calkyl–H bond activation observed during the formation of 2 and
3, rhodation of ligand precursor 1e was investigated in more
detail (Scheme 2). In this precursor, the imidazolium C2-positions
are not protected and hence available for metal coordination via
the normal carbene coordination mode. Upon exposure of 1e to
identical conditions as the diimidazolium salts 1a–d, i.e. hydrated
RhCl3 and NaOAc in refluxing MeCN, a mixture of products was
obtained after 16 h (Scheme 2). Complex 5 was isolated from this
mixture in 21% yield after column chromatography. This complex
features a bidentate coordinated ligand with both carbenes in the
normal bonding mode; a close analogue was previously reported
by Peris and coworkers.19a
Careful analysis of the reaction mixture before chromatographic
separation revealed, in addition to 5, also the presence of the
tridentate biscarbene complexes 4 and 6,20 both resulting from
Calkyl–H bond activation. In complex 4 the tridentate ligand
features an abnormally and a normally bound carbene, whereas in
complex 6 both carbenes are in the normal bonding mode. All three
complexes were isolated by fractional crystallisation and were fully
characterised by X-ray crystallography and NMR spectroscopy.
Complexes 4, 5b, and 6 all crystallised as monomers with the
rhodium centre in a slightly distorted octahedral environment
(Fig. 2). Like in complex 3, the Rh–Ccarbene bonds in complexes
4 and 6 are significantly shorter than the Rh–Calkyl bonds. The
normal carbene ligand in 4 is trans to a MeCN ligand and is closer
Scheme 2 Metallation of 1e. Reagents and conditions: RhCl3, NaOAc,
MeCN (reflux), then KI (for 4,5a, and 6) or KBr (for 5b).
Table 1 Selected bond lengths and angles for complexes 4, and 5b,
and 6a
4
5b
6
Rh1–C1
Rh1–C7
Rh1–C5
C1–Rh1–C7
C1–Rh1–C5
C5–Rh1–C7
1.995(3)
1.962(3)
2.071(3)
85.07(11)
82.81(11)
81.59(11)
1.978(10)
1.970(8)
—
98.9(4)
—
1.995(8)
1.988(8)
2.063(8)
91.5(3)
80.6(3)
81.7(3)
—
˚
to the rhodium centre (Rh–C7 1.962(3) A) than the abnormal
a
˚
Bond lengths in A, bond angles in deg.
˚
carbene ligand (Rh–C1 1.995(3) A) (Table 1). Longer Rh–Ccarbene
bonds were also noted for the all-normal carbene complex 6 (Rh–
˚
Ccarbene 1.995(8) and 1.988(8) A), where the carbenes are positioned
The dicarbene bite angle in complex 5b is unusually large,
trans to iodide, which has a stronger trans influence than MeCN.21
C1–Rh1–C7 is 98.9(4)◦,19,21,22 and the dihedral angle between
Fig. 2 ORTEP representation of the cationic complex 4 (a; 40% probability), neutral dicarbene complex 5b (b; 50% probability), and the neutral
tridentate complex 6 (c; 50% probability). Hydrogen atoms and non-coordinating iodide in 4 omitted for clarity.
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 9911–9920 | 9913
©

View Article Online
the imidazolylidene rings and the metal xy coordination plane
(defined by C1, Rh1, and C7) is much smaller (20.8–29.0◦) than
in the tridentate complex 4, where the heterocycles are positioned
nearly perpendicularly (dihedral angle between 80.9◦ and 84.6◦).
The conformational flexibility of the ligand imparted by the
flexible trimethylene linker is further illustrated by the synrotatory
twisted arrangement of the imidazolylidene rings, thus avoiding
steric congestion and yielding a propeller-like arrangement rather
than the generally observed V-shaped conformation.19 Steric
considerations may also account for the mutual trans arrangement
of the iodide and one carbene ligand in complexes 4 and 6.
Even though complexes 4 and 6 have a related tridentate ligand,
only complex 6 is overall neutral while complex 4 is monocationic.
The stabilisation of overall cationic rhodium centre in 4 may be
a direct consequence of the larger mesoionic character of the
C4-bound imidazolylidene ligand as compared to the normal
analogue in 6,23 thus illustrating the stronger donor power of the
abnormal carbene.14,24 Bond length analysis in the heterocycles
suggests a vinyl-type bonding of the C4-bound carbene ligand in
4, including remote charge stabilisation by the cationic amidinium
fragment.
as four doublets at dH 7.23, 7.08, 7.07 and 6.97 (3JHH = 1.9 Hz).
In line with the observed behaviour of complexes 2 and 4, iodide
coordination at lower temperature is surmised to be rigid on the
NMR time scale, thus producing an asymmetric cationic complex
with the iodide presumably cis to the alkyl ligand.25
Mechanistic studies
1
Based on the characteristic H NMR signals, it was possible to
identify and quantify complexes 4, 5, and 6 also in composite
mixtures. Since kinetic studies through in situ measurements
were hampered by the inhomogeneity of the reaction mixture,
1
the progress of the reaction was monitored through H NMR
measurement of samples taken at regular intervals. A typical
plot of the relative concentrations of 4, 5, and 6 during the
first 12 h of the reaction is shown in Fig. 3. In addition to
these three complexes, the reaction mixture was composed of
the diimidazolium salt 1e and presumably some monocarbene
species,26 which could not be confidently identified and quantified
in the product mixture.
In solution, complex 4 displays a complicated 1H NMR signal
pattern that simplifies upon heating to 60 ◦C. At this elevated
temperature, the inequivalence of the carbene bonding modes
is reflected by the presence of four singlet resonances for the
imidazolium protons. Most characteristic is the low-field reso-
nance at dH 8.31 due to the C2-bound hydrogen of the abnormal
carbene residue. The two butyl groups resonate as two disparate
sets of signals both in the 1H and 13C NMR spectra. Similarly, the
NCH2 protons of the trimethylene linker appear as two distinct
sets of AB doublets, one at dH 4.38 and 3.83 (2JHH = 13.0 Hz)
and the other at dH 4.32 and 3.67 (2JHH = 12.0 Hz). The proton
attached to the rhodium-bound carbon is very broad due to the
coupling to four inequivalent protons and to the rhodium centre.
At room temperature, three different species were identified from
the split of the lowest field singlet into three resonances between
8.4 and 8.2 ppm in approximate 1 : 4 : 6 ratio. The presence of three
species may be rationalised with rigid iodide coordination at room
temperature, either trans to the alkyl group or to either of the two
different carbene ligands and may thus reflect the different trans
influence of normally and abnormally bound NHC ligands and
the alkyl ligand, respectively.
Fig. 3 Time-dependent evolution of 4, 5 and 6 from the reaction of 1e
with RhCl3 (determined by 1H NMR spectroscopy).
Analysis of the reaction mixture composition allowed key
factors to be identified that govern the formation of the different
cyclometallated products. Thus, initial production of complex 4
is relatively fast and essentially complete after 2 h. No mono-
or bidentate intermediate comprising a C4-bound carbene was
observed and the concentration of the tridentate complex did not
change any further, even though the reaction mixture contained
residual monocarbene species (see below). Only after heating
for several days, was a slow decrease of the concentration of 4
observed. This decrease was attributed to complex decomposition,
hence revealing a limited long-term stability of this complex at
elevated temperatures.
Formation of the bidentate dicarbene species 5 occurred at
slightly lower rates than the tridentate complex 4 and preceded
the formation of the tridentate complex 6. Only after about 2 h
and significant build-up of 5, was Calkyl–H activation induced
and complex 6 started to appear. The increase of 6 was more
pronounced than the apparent consumption of 5, presumably
because of continuous formation of this complex from 1e or the
corresponding monocarbene species. At extended reaction times
(12–24 h), the rates for the formation of 6 and the consumption of 5
became essentially equal, yet low. After 60 h, the 5 : 6 ratio is about
The NMR spectrum of complex 5b displays a single set of
3
signals for the imidazolylidene ligands (dH 7.61 and 7.41, JHH
=
2.5 Hz), indicating Cs-type symmetry in solution. All N-bound
CH2 protons are diastereotopic and resonate as four sets of
multiplets due to the rigid conformation of the metallacycle on
the NMR time scale even at elevated temperatures (up to 60 ◦C).19
Obviously, the crystallographically deduced C1 symmetry is not
preserved in solution and swinging about the Ccarbene–Rh bond is
fluxional.
A similar Cs symmetric conformation was deduced for complex
6 at 70 ◦C. The aromatic signals of the tridentate ligand are shifted
to higher field as compared to the bidentate ligand in 5b and
appear at dH 7.15 and 7.03. The Cs symmetry supposedly arises
from dissociation of one iodide ligand and fluxional coordination
of the second iodide in the formally cationic system (see complex
2 above). Upon lowering the temperature, the structure becomes
◦
unsymmetrical and at -10 C the imidazolium protons resonate
9914 | Dalton Trans., 2011, 40, 9911–9920
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
˚
˚
1 : 2, indicating that Calkyl–H bond activation from 5 is very slow
and not complete even after several days, thus sharply contrasting
the instantaneous bond activation in the formation of 4.
slightly shorter (average 1.98 A cf . 2.01 A in 5a¢). The remaining
Rh–ligand bond distances are very similar in the two structures.
One possible pathway for Calkyl–H activation in 5a¢ and 5a¢ab
is shown in Fig. 4 and involves initial iodide dissociation and
rearrangement of the bis-NHC ligand to give intermediates A
and Aab. Both these species feature an upright orientation of
the carbene rings and greatly reduced bidentate chelate angles
(A: 92◦; Aab: 86.4◦). This arrangement allows the C5–H1 bond
to engage in agostic bonding with the Rh centre. The strength
of these interactions appears to be comparable in A and Aab,
as judged by the similar Rh ◊ ◊ ◊ H1 and C5–H1 distances (ca.
The data further suggest that formation of 4 and 6 are mutually
independent processes and do not result from interconversion
of an abnormal carbene bonding mode to a normal one or
vice versa.27 However, complexes 5 and 6 seem to be intercon-
nected. Indeed, separate experiments starting from pure 5 revealed
a slow Calkyl–H bond activation in the presence of NaOAc in
refluxing MeCN, and conversion to 6 occurred over several days.
Hence, product selectivity should be biased at an early stage of the
reaction, most likely at the initial heterocyclic C–H bond activation
process. This model suggests that initial activation of the C4–H in
1e is energetically not much different from C2–H bond activation.
The substantially lower acidity of the C4-bound proton28 may thus
be compensated by steric (de)shielding, suggesting the product
selectivity to be sensitive to wingtip group modifications.29
˚
˚
1.91 A and 1.14 A respectively). The proximity of the agostic C–H
bond cis to the k -OAc ligand in A and Aab allows these systems
2
to effect C–H activation via an ambiphilic metal–ligand assisted
(AMLA) mechanism.11,30 This proceeds via TS(A–B)/TS(A–B)ab
and involves dissociation of one Rh–O bond such that the free
acetate arm becomes available as an intramolecular base that can
deprotonate the agostic C–H bond. TS(A–B)ab displays a slightly
later geometry with shorter Rh ◊ ◊ ◊ C5 and Rh ◊ ◊ ◊ H1 contacts,
somewhat greater C5 ◊ ◊ ◊ H1 elongation and a shorter O1 ◊ ◊ ◊ H1
distance. These C–H activation transition states lead to the 5-
coordinated metalated intermediates B (E = -0.6 kcal mol-1) and
Bab (E = +1.7 kcal mol-1). The structures of these species closely
resemble their preceding transition states, with the full formation
of the Rh–C5 bond and proton transfer forming a bound acetic
acid molecule (see the ESI for full details†).
Fig. 4 also shows the relative free energies of the various
species implicated in Calkyl–H activation, where a correction for
the MeCN solvent has also been included via the PCM approach.
The overall barrier for C–H activation from 5a¢ab is 24.9 kcal mol-1,
4.4 kcal mol-1 lower than that for C–H activation from 5a¢. This
difference is consistent with the greater reactivity of the abnormal
NHC systems towards Calkyl–H activation seen experimentally for
Theoretical analyses
Density functional theory (DFT) calculations employing the BP86
functional have been performed to obtain further insight into
how the NHC binding mode affects the Calkyl–H bond activation
process. Calculations considered the mechanism of Calkyl–H bond
activation from the key intermediate 5a¢ (the prime indicating use
of the N–Me analogue of complex 5 used experimentally) and
its abnormal NHC analogue 5a¢ab, i.e. the presumed precursor
of complex 2 (Fig. 4). The computed structure of 5a¢ compares
well with the experimental structure of 5b; in particular the
large bidentate angle (C1–Rh–C7 = 99.8◦) and the propeller
arrangement of the ligand are correctly reproduced. These features
are also apparent in the computed structure of 5a¢ab in which the
C1–Rh–C7 angle is slightly larger (102.1◦) and the Rh–C distances
˚
Fig. 4 Computed geometries with selected distances in A for C–H activation from 5a¢ and 5a¢ab (H atoms, excepting those of the C5 methylene group,
are omitted for clarity). Free energies (kcal mol-1) include a correction for MeCN solvation (PCM method).
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 9911–9920 | 9915
©

View Article Online
complexes 2 and 4. Perhaps surprisingly, however, the origin of this
difference does not lie in the Calkyl–H activation step itself, where
a slightly lower barrier is actually computed from intermediate
A (12.4 kcal mol-1 cf . 13.1 kcal mol-1 from Aab). Instead it
is the greater ease of iodide dissociation and rearrangement to
form the agostic intermediate, lower by 5.1 kcal mol-1, which is
responsible for the reduced overall barrier in the C4-bound NHC
system. To probe the reasons for this difference we computed the
alternative forms of A and Aab where the NHC ligands maintain
the conformations seen in 5a¢ and 5a¢ab. The energies of these
species, +13.3 kcal mol-1 and +10.2 kcal mol-1 respectively, show
iodide dissociation is ca. 3 kcal mol-1 easier in the C4-bound NHC
system. The subsequent rearrangement is also slightly more facile
in this case and contributes to the 5.1 kcal mol-1 overall difference
in the formation of A and Aab. This latter component may reflect
a greater donation from C4-bound NHC and hence the more
effective stabilisation of the agostic intermediate. While this is not
particularly apparent in 5a¢ and 5a¢ab, where the Rh–I distances
and even the Rh–O distances (trans to the NHC ligands) are very
˚
similar, the Rh–O distances in Aab are ca. 0.02 A longer than in
A. Trends in the computed natural charges also suggest greater
donation from the NHC upon iodide loss. For 5a¢ the computed
charge on Rh is +0.367 and is actually less positive than the value
of +0.384 in 5a¢ab. In contrast, for intermediate A the charge on
Rh increases to +0.389, whereas in Aab there is almost no increase
in charge on Rh (+0.385). Although these effects are rather subtle
they do point to a greater ability of the C4-bound NHC to stabilise
the agostic intermediate and that this, together with the enhanced
propensity of iodide dissociation, is responsible for promoting the
C–H activation reaction in the C4-bound NHC system.31
Scheme 3 Reactivity of complexes 2c, 4, and 7 towards D3PO4.
Mechanistically, the deuterium incorporation may be strongly
related to the isotope exchange observed in free normal carbenes,32
and may thus involve transient sp2-to-sp3 rehybridisation at C5.
Rehybridisation of heterocyclic carbon atoms in NHC complexes
has precedents33 and may be promoted by the abnormal bonding
mode of the carbene, in which contributions from mesoionic
resonance structures are more pronounced. In such a limiting
structure, the ligand may be separated into a vinylic metal-
bound C–C fragment and a N–C–N amidinium cation as charge-
compensating unit (Scheme 4).15b When bound to the metal,
formally a metalla-allyl anion is formed, which may be reversibly
protonated either at the metal centre (reinforcing a vinyl-type
ligand bonding as in A) or at the C5 position (producing a C Rh
carbene complex B; Scheme 4). Alternatively, a 1,3-sigmatropic
shift of the metal-bound hydrogen to C5 may allow for accessing
intermediate B from A, which upon unselective deprotonation
would rationalise the observed H/D exchange at C5. In an attempt
to verify this hypothesis, complex 2c was treated with MeOTf
(5 equiv.), which should lead to irreversible C–C bond formation
at the imidazolylidene C5 position. When following the reaction
Reactivity and stability of the complexes
Both the bidentate and tridentate complexes 2–6 are remarkably
stable towards air, moisture and elevated temperatures. Slow
decomposition was only noted when heated to 80 ◦C for extended
periods of time (>3 days). The complexes were also seemingly
robust when treated with moderately strong acids such as HOAc
and H3PO4. However, when treated with deuterated acids such
as D3PO4 at RT, complexes 2 and 4 were deuterated at the C5
position to give 2–D2 and 4–D, respectively (Scheme 3). The
1
deuteration was monitored by H NMR spectroscopy through
the disappearance of the C5-bound heterocyclic proton signal
1
(dH 6.78 for 2c) and was confirmed by the appearance of a
at room temperature by H NMR using 4,4¢-dimethylbiphenyl
2
resonance at the same frequency in the H NMR spectrum. No
such reaction was observed when the carbene was normally bound
as in complexes 5 and 6. This reactivity pattern is also illustrated
by the exclusive monodeuteration of the mixed abnormal/normal
dicarbene complex 4. In this complex, only the C5-bound proton
was exchanged in the presence of D3PO4 while the more acidic
C2-bound proton was stable and remained unaffected. In line
with this observation, the methylene-bridged dicarbene complex
7 underwent H/D exchange under identical reaction conditions
and gave 7–D2 (Scheme 3).
Heterocyclic H/D exchange occurred much slower in complexes
4 and 7 than in 2. After 1 h at room temperature, 85% deuteration
was observed in 2 as compared to 33% in 4 (7% in 7). Moderate
warming to 60 ◦C accelerated the exchange considerably and
reached 72% in 4 after the same reaction time (83% in 7).
Scheme 4 Proposed mechanism for H/D exchange in abnormal carbene
metal complexes.
9916 | Dalton Trans., 2011, 40, 9911–9920
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
as internal standard, significant decomposition of the complex
was observed after 30 min. Presumably, complex degradation was
induced by the formation of triflic acid, which may originate
All other reagents are commercially available and were used as
received. Unless otherwise stated NMR spectra were recorded
at 25 ◦C on Bruker and Varian spectrometers operating at
400, 500 or 600 MHz (1H NMR) and 100, 125 or 150 MHz
from H+/CH3 substitution at C5.34 Independent measurements
+
1
revealed that the carbene complex 2c is indeed instable towards
such a strong acid. Performing the reaction in the presence
of 1,8-bis(dimethylamino)naphthalene as proton-sponge did not
completely prevent decomposition, yet a selective decrease of the
signal due to the C5-bound proton to 32% with respect to the
other ligand signals was noted, indicating a substantial degree of
deprotonation. Although the aliphatic region of the spectrum was
rather complicated, a new resonance appearing at dH 2.60 and
integrating for three protons was tentatively attributed to the C5-
methlyated analogue of 2c. These results point to a cooperative
mode of action of the complex, involving the metal centre and the
abnormal carbene for bond activation.35 Such cooperativity may
be key for the smooth C–H activation in complexes containing
abnormal carbenes as opposed to analogues with normally bound
carbenes only.
(13C{ H} NMR), respectively. Chemical shifts (d in ppm, coupling
constants J in Hz) were referenced to residual solvent resonances.
Assignments are based on homo- and heteronuclear shift cor-
relation spectroscopy. Elemental analyses were performed by the
Microanalytical Laboratory at the Federal Institute of Technology
in Zurich, Switzerland and at the University College Dublin,
Ireland.
Synthesis of 1a
A solution of 2-phenyl N-n-butyl imidazole (3.47 g, 17.3 mmol)
and 1,3-dibromopropane (1.75 g, 8.7 mmol) was stirred in
refluxing toluene (20 mL) for 16 h. The formed precipitate was
isolated by centrifugation, washed with toluene, and recrystallised
from CH2Cl2/Et2O to yield 1a as an off-white solid (3.12 g,
3
60%). 1H NMR (CD3CN, 500 MHz): d 7.94 (d, 2H, JHH
=
=
2.5 Hz, Himi), 7.75 (t, 2H, 3JHH = 7.5 Hz, Haryl), 7.66 (t, 4H, 3JHH
7.5 Hz, Haryl), 7.55 (m, 3H, Haryl, Himi), 3.97 (t, 4H, 3JHH = 7.5 Hz,
NCH2CH2CH2N), 3.89 (t, 4H 3JHH = 7.5 Hz, NCH2CH2CH2CH3),
Conclusions
The intramolecular Calkyl–H bond activation mediated by rhodium
carbene complexes has been investigated by using both C2-
alkylated and C2-protonated diimidazolium salt precursors. These
studies revealed that the activation of both the second heterocyclic
C–H bond and in particular the alkyl C–H bond require less energy
and proceed at significantly faster rates when the rhodium centre
is ligated by an abnormally bound, strongly electron-donating
carbene ligand. Specifically C2-protected diimidazolium salts,
where the abnormal binding mode is enforced, undergo smooth
Calkyl–H bond activation and yield tridentate complexes at ambient
temperatures. Theoretical analyses suggest that the ease of forma-
tion of the agostic precursors to C–H activation discriminates
the two systems. This results from a combination of more facile
iodide dissociation in the abnormally bound carbene complexes,
presumably due to the pronounced mesoionic character of C4-
bound imidazolylidenes, and the ensuing stabilisation of agostic
intermediates. Subsequent chelate-assisted C–H bond activation
entails similar barriers of ca. 13 kcal mol-1 in each case.
In the presence of mild acids, selective H/D exchange at
the abnormally bound carbene heterocycle was observed. This
reactivity pattern may be general for C4-bound imidazolylidenes
and related ligand systems such as C4-bound 1,2,3-triazolylidenes
and may constitute a new pathway for cooperating metal-centred
reactivity. Accordingly, these abnormal carbene ligands may act
as transient proton acceptors through their mesoionic resonance
form, thus transforming the metal-bound carbanion to a formally
neutral carbene donor site comprising a remote cationic residue.
Preliminary work in our laboratories suggests that this type
of metal–ligand cooperativity is general and applicable for the
activation of different E–H bonds.
3
2.16 (quintet, 2H, JHH = 7.5 Hz, NCH2CH2CH2N), 1.65
3
(quintet, 4H, JHH
= 7.5 Hz, NCH2CH2CH2CH3), 1.19
3
(sextet, 4H, JHH = 7.5 Hz, NCH2CH2CH2CH3), 0.78 (t,
3
1
6H, JHH
=
7.5 Hz, NCH2CH2CH2CH3). 13C{ H} NMR
(CD3CN, 125 MHz): d 145.5 (NCN), 133.7 (CarylH), 131.5
(CarylH), 130.8 (CarylH), 123.3 (Cimi), 123.0 (Cimi), 122.1
(Caryl), 49.4 (NCH2CH2CH2CH3), 46.3 (NCH2CH2CH2N),
32.3 (NCH2CH2CH2CH3), 30.9 (NCH2CH2CH2N), 20.0
(NCH2CH2CH2CH3), 13.5 (NCH2CH2CH2CH3). HR-MS
(ESI+): Calc. for C29H38BrN4 ([M - Br]+): 521.2280. Found:
521.2280. Anal. calc. for C29H38Br2N4 (602.45) ¥ H2O: C, 56.14;
H, 6.50; N, 9.03. Found: C, 56.12; H, 6.47; N, 9.00%.
Synthesis of 1b
According to the procedure described for 1a, compound 1b was
obtained from 2-methyl N-mesityl imidazole (1.20 g, 6.0 mmol)
and 1,3-dibromopropane (0.61 g, 3.0 mmol) as an off-white solid
(1.52 g, 73%). 1H NMR (CD3CN, 500 MHz): d 8.18 (d, 2H,
3JHH = 2.0 Hz, Himi), 7.36 (d, 2H, 3JHH = 2.0 Hz, Himi), 7.14 (s, 4H,
Hmes), 4.55 (t, 4H, 3JHH = 8.0 Hz, NCH2CH2CH2N), 2.66 (m, 2H,
NCH2CH2CH2N), 2.51 (s, 6H, CimiCH3), 2.36 (s, 6H, CmesCH3),
1
1.99 (s, 12H, CmesCH3). 13C{ H} NMR (CD3CN, 125 MHz) d
146.2 (NCN), 142.3 (CmesCH3), 136.1 (CmesCH3), 131.3 (CmesN),
130.6 (CmesH), 123.6 (Cimi), 123.3 (Cimi), 46.6 (NCH2CH2CH2N),
30.9 (NCH2CH2CH2N), 21.1 (CmesCH3), 17.6 (CmesCH3), 10.8
(CimiCH3). Anal. calc. for C29H38Br2N4 (602.45) ¥ 2 CH3CN: C,
57.90; H, 6.48; N, 12.28. Found: C, 57.92; H, 6.40; N, 12.28%.
Spectroscopic data for 2a
The monometallic species 2a was obtained by dissolving complex
3a in MeCN. Two species are distinguishable in the NMR spectra
in a 2 : 1 ratio. Major species: 1H NMR (CD3CN, 500 MHz):
d 7.60 (m, 6H, Haryl), 7.42 (d, 4H, JHH = 8.0 Hz, Haryl), 7.08
(br, 2H, Himi), 4.22 (m, 2H, NCHHCHCHHN), 3.83 (m, 5H,
Experimental section
General comments
3
The 1,2-disubstituted imidazoles36 and compounds 1e, 2c–d, 5, and
7 were prepared according to previously reported procedures.17–19
RhCH + NCH2CH2CH2CH3), 3.49 (m, 2H, NCHHCHCHHN),
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 9911–9920 | 9917
©

View Article Online
3
1.61 (quintet, 4H, JHH = 7.5 Hz, NCH2CH2CH2CH3), 1.14
(1JRhC = 29.6 Hz, RhCH), 32.9, 32.7 (NCH2CH2CH2CH3), 20.3,
20.2(2 ¥ NCH2CH2CH2CH3), 14.0, 13.8(2 ¥ NCH2CH2CH2CH3).
3
(sextet, 4H, JHH = 7.5 Hz, NCH2CH2CH2CH3), 0.76 (t, 6H,
3JHH = 7.5 Hz, NCH2CH2CH2CH3). Minor species: 1H NMR
(CD3CN, 500 MHz): d 7.60 (m, 6H, Haryl), 7.46 (dd, 4H,
The carbene carbon signal is not well resolved in the 13C{ H}
1
spectrum. Anal. calc. for C58H70I4N8Rh2 (1592.66) ¥ CH2Cl2:
C, 40.89; H, 4.23; N, 6.36. Found: C, 40.91; H, 4.30; N,
6.57%.
3JHH = 8.0 Hz, JHH = 1.5 Hz, Haryl), 6.93 (s, 2H, Himi), 4.22
4
(m, 2H, NCHHCHCHHN), 3.83 (m, 4H, NCH2CH2CH2CH3),
3.72 (m, 1H, RhCH), 3.49 (m, 2H, NCHHCHCHHN), 1.66
3
(quintet, 4H, JHH = 7.5 Hz, NCH2CH2CH2CH3), 1.14 (sextet,
Synthesis of 3b
3
3
4H, JHH = 7.5 Hz, NCH2CH2CH2CH3), 0.79 (t, 6H, JHH
=
1
7.5 Hz, NCH2CH2CH2CH3). 13C{ H} NMR (CD3CN, 125 MHz):
d 132.0 (CarylH), 130.5 (CarylH), 130.2 (CarylH), 121.8 (Cimi),
106.2 (Caryl), 59.4 (NCH2CHCH2N), 48.0 (NCH2CH2CH2CH3),
33.0 (NCH2CH2CH2CH3), 20.1 (NCH2CH2CH2CH3), 13.6
(NCH2CH2CH2CH3). The carbene carbon, NCN and RhCH
According to the procedure described for 3a, complex 3b was
produced from 1b (277 mg, 0.4 mmol), RhCl3·H2O (105 mg,
0.4 mmol), NaOAc (263 mg, 3.2 mmol) and KI (266 mg, 1.6 mmol)
as a brown solid (189 mg, 74%). Two major sets of signals are
distinguishable in the NMR spectra in an approximate 2 : 1 ratio.
1H NMR (CD2Cl2, 500 MHz): d 7.05–6.91 (s, 4H, Hmes), 6.65, 6.63,
6.63 (s, 2H, Himi), 4.67–4.58 (m, 1H, RhCH), 4.58–4.41 (m, 2H,
NCHHCHCHHN), 3.94–3.81 (m, 2H, NCHHCHCHHN), 2.33,
2.32 (s, 6H, CimiCH3), 2.18 (s, 6H, CmesCH3), 1.99 (s, 6H, CmesCH3),
1
signals are not resolved in the 13C{ H} spectrum. The signals of
the major and minor species differ by less than 0.3 ppm.
Spectroscopic data for 2b
1.75, 1.72 (s, 6H, CmesCH3). 13C{ H} NMR (CD2Cl2, 125 MHz):
1
The monometallic species 2b was obtained by dissolving complex
d 140.6 (NCN), 135.5 (br, CmesCH3), 132.1 (br, CmesCH3), 129.9,
129.8 (2 ¥ Cmes), 121.7 (br, Cimi), 60.52, 60.17 (2 ¥ NCH2CHCH2N),
1
3b in MeCN. Two species are distinguishable in the H NMR
spectrum in a 1.2 : 1 ratio. Major species: 1H NMR (CD3CN,
360 MHz): d 7.09, 7.04 (2 ¥ s, 2H, Hmes), 6.55 (s, 2H, Himi), 4.43
(br, 2H, NCHHCHCHHN), 4.04 (br, 1H, RhCH), 3.90 (m, 2H,
NCHHCHCHHN) 2.33 (s, 6H, CimiCH3), 2.19 (s, 6H, CmesCH3),
37.7 (d, JRhC = 32.0, RhCH), 21.4 (CimiCH3), 17.5 (CmesCH3),
1
10.6 (CmesCH3). The carbene carbon signal and CmesN signals
are not well resolved in the 13C{ H} spectrum. Anal. calc. for
1
C58H70I4N8Rh2 (1592.67) ¥ CH2Cl2: C, 42.24; H, 4.33; N, 6.68.
1
1.98 (s, 6H, CmesCH3), 1.75 (s, 6H, CmesCH3). Minor species: H
Found: C, 42.16; H, 4.73; N, 6.61%.
NMR (CD3CN, 360 MHz): d 7.07, 7.02 (2 ¥ s, 2H, Hmes), 6.69,
6.40 (2 ¥ br, 2H, Himi), 4.35 (m, 2H, NCHHCH2CHHN), 3.90
(m, 3H, RhCH + NCHHCHCHHN), 2.32 (s, 6H, CimiCH3), 2.14
(br, 6H, CmesCH3), 1.96 (br, 6H, CmesCH3), 1.73 (br, 6H, CmesCH3).
Synthesis of 4
According to the general procedure described for 2a–b, starting
from 1e (353 mg, 0.74 mmol), RhCl3·H2O (194 mg, 0.74 mmol),
NaOAc (484 mg, 5.9 mmol) and KI (490 mg, 3.0 mmol). The
crude product was purified by column chromatography (SiO2).
Complex 5 was eluted selectively with CH2Cl2/acetone (5 : 2),
and a mixture of complexes 4, 5, 6 and some monocarbene
species were obtained as a brown solid by increasing the polarity
of the mobile phase (CH2Cl2/acetone 5 : 3). Gradient column
chromatography followed by fractional crystallisation induced
by slow diffusion of Et2O into a solution of the crude mixture
of 4 and 6 in CH3CN and acetone (1 : 1) gave two different
types of crystals that were manually separated. Complex 4
crystallised as red rods, complex 6 as yellowish needles. 1H NMR
1
13C{ H} NMR (CD3CN, 125 MHz): d 146.0 (1JRhC = 46.2 Hz,
Cimi), 141.2 (s, NCN), 136.1 (CmesCH3), 132.3 (CmesCH3), 130.1
(Cmes), 120.9 (d, 2JRhC = 2.9 Hz, Cimi), 59.0 (NCH2CHCH2N), 34.0
1
(d, JRhC = 28.5 Hz, RhCH), 21.0 (CimiCH3) 17.4, 17.3, 11.0 (3 ¥
1
CmesCH3). The CmesN signal is not well resolved in the 13C{ H}
spectrum. The signals of the major and minor species differ by
less than 0.3 ppm.
Synthesis of 3a
Compound 1a (301 mg, 0.5 mmol), RhCl3·H2O (132 mg, 0.5 mmol)
and NaOAc (328 mg, 4.0 mmol) were stirred in refluxing MeCN
(30 mL) for 16 h. Then KI (332 mg, 2.0 mmol) was added
and stirring at reflux was continued for a further hour. The
reaction mixture was cooled to room temperature and all volatiles
were removed under reduced pressure. The residue was purified
by column chromatography (SiO2; CH2Cl2/acetone, 5 : 2), thus
affording 3a as a brown solid (0.24 g, 60%). Two major sets of
signals are distinguishable in the NMR spectra in an approximate
(CD3CN, 600 MHz, 333 K): d 8.31, 7.07, 6.91, 6.86 (4 ¥ s,
2
1H, Himi), 4.38 (br, 1H, NCHHCHCH2N), 4.32 (dd, 1H, JHH
=
3
12.0 Hz, JHH = 7.7 Hz, NCH2CHCHHN), 4.21 (br, 3H, RhCH
+ NCH2CH2CH2CH3), 3.97 (t, 2H, NCH2CH2CH2CH3), 3.83
2
2
(d, 1H, JHH = 13.0 Hz, NCHHCHCH2N), 3.67 (d, 1H, JHH
=
12.0 Hz, NCH2CHCHHN), 1.80 (m, 2H, NCH2CH2CH2CH3),
1.75 (quintet, 2H, JHH = 7.3 Hz, NCH2CH2CH2CH3), 1.38
(sextet, 2H, JHH = 6.0 Hz, NCH2CH2CH2CH3), 1.26 (m,
3
1
3
2 : 1 ratio. H NMR (CD2Cl2, 500 MHz): d 7.61–7.53 (m, 6H,
3
Haryl), 7.36–7.30 (m, 4H, Haryl), 7.02, 6.93, 6.92 (s, 2H, Himi),
4.60–4.14 (m, 3H, RhCH + NCHHCHCHHN), 3.89–3.68 (m,
4H, NCH2CH2CH2CH3), 3.67–3.46 (m, 2H, NCHHCHCHHN),
1.77–1.66 (m, 4H, NCH2CH2CH2CH3), 1.25–1.18 (m, 4H,
2H, NCH2CH2CH2CH3), 0.98, 0.90 (2
¥
t, 3H, JHH
=
7.4 Hz, NCH2CH2CH2CH3). 13C{ H} NMR (CD3CN, 150 MHz):
1
d
132.2, 122.6, 121.6, 119.5 (4 ¥ Cimi), 60.7, 60.4 (2 ¥
NCH2CHCH2N), 49.8, 49.4 (2 ¥ NCH2CH2CH2CH3), 35.5
(RhCH), 33.7, 33.0 (2 ¥ NCH2CH2CH2CH3), 20.6, 20.1 (2 ¥
NCH2CH2CH2CH3), 14.2, 13.7 (2 ¥ NCH2CH2CH2CH3). The
3
NCH2CH2CH2CH3), 0.84, 0.83, 0.82 (t, JHH = 7.0 Hz, 6H,
1
NCH2CH2CH2CH3). 13C{ H} NMR (CD2Cl2, 125 MHz): 140.0
1
(NCN), 131.7, 131.6 (2 ¥ CarylH), 130.1, 130.0 (2 ¥ CarylH),
129.75 (CarylH), 125.2 (Caryl), 122.1, 121.3 (2 ¥ Cimi), 61.6
(NCH2CHCH2N), 48.7, 48.1 (2 ¥ NCH2CH2CH2CH3), 38.5
carbene carbon signals are not well resolved in the 13C{ H}
spectrum. Anal calc. for C21H33I2N6Rh (726.24): C, 34.73; H, 4.58;
N, 11.57. Found: C, 34.76; H, 4.53; N, 11.73%.
9918 | Dalton Trans., 2011, 40, 9911–9920
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Synthesis of 6
Werner Foundation, and Sasol Ltd for financial support of this
work.
Method A: According to the procedure described for the synthesis
of complex 4.
Notes and references
Method B: Complex 5 (80 mg, 0.16 mmol) and NaOAc
(107 mg, 1.3 mmol) were stirred together in refluxing acetonitrile
(20 mL) for 6 days. Solid KI (224 mg, 1.6 mmol) was added and
the reaction mixture refluxed for 2 more days. The product was
purified by column chromatography (SiO2, CH2Cl2/acetone first
10 : 1, then 10 : 3) and 6 was obtained as a yellow solid (26 mg,
1 (a) R. Jazzar, J. Hitce, A. Renaudat, J. Sofack-Kreutzer and O.
Baudoin, Chem.–Eur. J., 2010, 16, 2654; (b) R. Giri, B.-F. Shi, K.
M. Engle, N. Maugel and J.-Q. Yu, Chem. Soc. Rev., 2009, 38,
3242.
2 M. S. Chen and M. C. White, Science, 2007, 318, 783.
3 (a) V. Vidal, A. Theolier, J. Thivolle-Cazat and J.-M. Basset, Science,
1997, 276, 99; (b) A. S. Goldman, A. H. Roy, Z. Huang, R. Ahuja, W.
Schinski and M. Brookhart, Science, 2006, 312, 257.
1
23% yield) after solvent removal. H NMR (CD3CN, 500 MHz,
343 K): d 7.15, 7.03 (2 ¥ s, 1H, Himi), 4.56 (br, 1H, RhCH),
4.35 (m, 4H, NCHHCHCHHN + NCHHCH2CH2CH3), 4.06
(m, 2H, NCHHCH2CH2CH3), 3.67 (m, 2H, NCHHCHCHHN),
4 (a) R. H. Crabtree, Chem. Rev., 1985, 85, 245; (b) A. E. Shilov and G.
B. Shul’pin, Russ. Chem. Rev., 1987, 56, 442; (c) B. A. Arndtsen and
R. G. Bergman, Science, 1995, 270, 1970; (d) J. Corker, F. Lefebvre,
C. Lecuyer, V. Dufaud, F. Quignard, A. Choplin, J. Evans and J.-M.
Basset, Science, 1996, 271, 966; (e) R. A. Periana, O. Mironov, D.
Taube, G. Bhalla and C. J. Jones, Science, 1998, 280, 560; (f) J. A.
Labinger and J. E. Bercaw, Nature, 2002, 417, 507; (g) A. S. Goldman
and K. I. Goldberg (ed.), Activation and Functionalization of C–H
Bonds, ACS Symposium Series 885, Wiley, Washington, D.C., 2004;
(h) R. H. Crabtree, J. Organomet. Chem., 2004, 689, 4083; (i) R. G.
Bergman, Nature, 2007, 446, 391; (j) J. F. Hartwig, Nature, 2008, 455,
314; (k) C. Coperet, Chem. Rev., 2010, 110, 656–680; (l) J. Choi, A.
MacArthur, M. Brookhart and A. S. Goldman, Chem. Rev., 2011, 111,
1761; For intramolecular processes, see e.g. ref. 1a and (m) D. Lapointe
and K. Fagnou, Chem. Lett., 2010, 39, 1118.
1.87, 1.79 (2 ¥ m, 2H, NCH2CH2CH2CH3), 1.40 (sextet, 4H,
3JHH = 7.4 Hz, NCH2CH2CH2CH3), 0.98 (t, 6H, JHH
=
3
7.3 Hz, NCH2CH2CH2CH3). 13C{ H} NMR (CD3CN, 125 MHz):
1
d
123.0, 122.9, 120.4, 120.2 (4 ¥ Cimi), 60.4, 60.1 (2 ¥
NCH2CHCH2N), 50.1, 49.8 (2 ¥ NCH2CH2CH2CH3), 36.0
(RhCH), 33.9, 33.3 (2 ¥ NCH2CH2CH2CH3), 20.6, 20.4 (2 ¥
NCH2CH2CH2CH3), 14.1 (NCH2CH2CH2CH3). Carbene carbon
signal not resolved. Anal. Calc. for C19H30I2N5Rh (685.19): C,
33.31; H, 4.41; N, 10.22. Found: C, 33.32; H, 4.14; N, 10.61%.
5 H. Arakawa, M. Aresta, J. N. Armor, M. A. Barteau, E. J. Beckman,
A. T. Bell, J. E. Bercaw, C. Creutz, E. Dinjus, D. A. Dixon, K. Domen,
D. L. DuBois, J. Eckert, E. Fujita, D. H. Gibson, W. A. Goddard, D.
W. Goodman, J. Keller, G. J. Kubas, H. H. Kung, J. E. Lyons, L. E.
Manzer, T. J. Marks, K. Morokuma, K. M. Nicholas, R. Periana, L.
Que, J. Rostrup-Nielson, W. M. H. Sachtler, L. D. Schmidt, A. Sen, G.
A. Somorjai, P. C. Stair, B. R. Stults and W. Tumas, Chem. Rev., 2001,
101, 953.
Structure determination and refinement of 3b, 4, 5b and 6
Suitable single crystals were mounted on Stoe Imaging Plate
Diffractometer Systems37 with a two-circle goniometer (3b and
4) or a one-circle j goniometer (5b), or on a Bruker SMART
APEX CCD diffractometer (6). Graphite-monochromators (Mo-
6 Z. Lin, Coord. Chem. Rev., 2007, 251, 2280.
7 R. N. Perutz and S. Sabo-Etienne, Angew. Chem., Int. Ed., 2007, 46,
2578.
˚
Ka radiation, l = 0.71073 A). For 3b and 4, a semi-empirical
absorption correction was applied using MULscanABS as imple-
mented in PLATON.38 An empirical absorption correction was
applied for 6 using the SADABS routine,39 and no absorption
correction was applied for 5b. The structures were solved by direct
methods using the program SHELXS-9740 and refined by full
matrix least squares on F2 with SHELXL-97. The hydrogen atoms
were included in calculated positions and treated as riding atoms
using SHELXL-97 default parameters. All non-hydrogen atoms
were refined anisotropically.
Crystals of complex 3b contained one half-occupied CH2Cl2 per
complex molecule. Disorder was found in one mesityl substituent
and in the PF6- anion, and more pronounced in the co-crystallised
solvent molecules. The SQUEEZE option in PLATON was
used to calculate the potential solvent accessible volume (1848
8 (a) A. Dedieu, Chem. Rev., 2000, 100, 543; (b) S. Sakaki, Top.
Organomet. Chem., 2005, 12, 31; (c) M. Lersch and M. Tilset, Chem.
Rev., 2005, 105, 2471.
9 (a) H. Chen, S. Schlecht, T. C. Semple and J. F. Hartwig, Science, 2000,
287, 1995; (b) I. A. I. Mkhalid, J. H. Barnard, T. B. Marder, J. M.
Murphy and J. F. Hartwig, Chem. Rev., 2010, 110, 890.
10 In this context, electrophilic metals imply a larger charge transfer from
the C–H bond to the metal, whereas nucleophilic centres engage in
larger backbonding, i.e. charge transfer from the metal to the C–
H s* orbital. In amphiphilic centres bonding and backbonding are
nearly balanced. See: D. H. Ess, W. A. Goddard and R. A. Periana,
Organometallics, 2010, 29, 6459.
11 (a) Y. Boutadla, D. L. Davies, S. A. Macgregor and A. I. Poblador-
Bahamonde, Dalton Trans., 2009, 5820; (b) C. E. Kefalidis, O. Baudoin
and E. Clot, Dalton Trans., 2010, 39, 10528; (c) D. Balcells, E. Clot and
O. Eisenstein, Chem. Rev., 2010, 110, 749.
12 (a) D. Bourissou, O. Guerret, F. P. Gabbai and G. Bertrand, Chem. Rev.,
2000, 100, 39; (b) W. A. Herrmann, Angew. Chem., Int. Ed., 2002, 41,
1290; (c) F. E. Hahn and M. C. Jahnke, Angew. Chem., Int. Ed., 2008,
47, 3122; (d) S. Diez-Gonzalez, N. Marion and S. P. Nolan, Chem. Rev.,
2009, 109, 3612; (e) L. Mercs and M. Albrecht, Chem. Soc. Rev., 2010,
39, 1903.
13 Intermolecular C–H bond activation: (a) M. Muehlhofer, T. Strassner
and W. A. Herrmann, Angew. Chem., Int. Ed., 2002, 41, 1745; For
examples of intramolecular C–H bond activation, see: (b) J. Huang,
E. D. Stevens and S. P. Nolan, Organometallics, 2000, 19, 1194; (c) R.
Dorta, E. D. Stevens and S. P. Nolan, J. Am. Chem. Soc., 2004, 126,
5054; (d) S. Burling, B. M. Paine, D. Nama, V. S. Brown, M. F. Mahon,
T. J. Prior, P. S. Pregosin, M. K. Whittlesey and J. M. J. Williams, J.
Am. Chem. Soc., 2007, 129, 1987; (e) V. Lavallo and R. H. Grubbs,
Science, 2009, 326, 559; C–C bond activation: (f) R. F. R. Jazar, S. A.
Macgregor, M. F. Mahon, S. P. Richards and M. K. Whittlesey, J. Am.
Chem. Soc., 2002, 124, 4944; C–F bond activation: (g) V. P. W. Bo¨hm, C.
W. K. Gsto¨ttmayr, T. Weskamp and W. A. Herrmann, Angew. Chem.,
Int. Ed., 2001, 40, 3387.
˚
A, containing about 385 electrons). Eight hexane molecules
(8 ¥ 50 electrons) per unit cell were included in all further
calculations. Due to the disorder, the corresponding parts of the
ligand were refined as having the same thermal values. Compound
5b crystallised as a racemic twin, therefore the TWIN refinement
was applied in SHELXL. For 6, the absolute structure was
determined, the final Flack parameter is -0.03(3). Further details
on data collection and refinement are summarised in Table S1.†
Acknowledgements
We thank F. Fehr, Y. Ortin, and J. Muldoon, for analytical
assistance and the Swiss National Science Foundation, the Alfred
This journal is
The Royal Society of Chemistry 2011
Dalton Trans., 2011, 40, 9911–9920 | 9919
©

View Article Online
14 (a) A. R. Chianese, A. Kovacevic, B. M. Zeglis, J. W. Faller and R. H.
Crabtree, Organometallics, 2004, 23, 2461; (b) M. Heckenroth, A. Neels,
M. G. Garnier, P. Aebi, A. W. Ehlers and M. Albrecht, Chem.–Eur. J.,
2009, 15, 9375.
15 (a) P. L. Arnold and J. Pearson, Coord. Chem. Rev., 2007, 251, 596;
(b) M. Albrecht, Chem. Commun., 2008, 3601; (c) O. Schuster, L.
Yang, H. G. Raubenheimer and M. Albrecht, Chem. Rev., 2009, 109,
3445.
26 (a) W. A. Herrmann, J. Schwarz and M. G. Gardiner, Organometallics,
1999, 18, 4082; (b) M. Heckenroth, E. Kluser, A. Neels and M. Albrecht,
Dalton Trans., 2008, 6242.
27 This conclusion is further supported by separate experiments involving
the exposure of pure complex 4 or 6 to the reaction conditions used
for their preparation (KI, NaOAc, MeCN, reflux for several hours). In
either case, the complexes were stable and did not show any detectable
interconversion to the other complex.
16 (a) M. Heckenroth, E. Kluser, A. Neels and M. Albrecht, Angew.
Chem., Int. Ed., 2007, 46, 6293; (b) R. Lalrempuia, N. D. McDaniel,
H. Mueller-Bunz, S. Bernhard and M. Albrecht, Angew. Chem.,
Int. Ed., 2010, 49, 9765; (c) A. Prades, E. Peris and M. Albrecht,
Organometallics, 2011, 30, 1162; (d) A. Kru¨ger and M. Albrecht, in N-
heterocyclic carbenes: From Laboratory Curiosities to Efficient Synthetic
Tools, ed. Silvia Diez-Gonzalez (Ed.), RSC, Cambridge, UK, 2011, p.
134.
17 A. Kru¨ger, A. Neels and M. Albrecht, Chem. Commun., 2010, 46, 315.
18 L. Yang, A. Kru¨ger, A. Neels and M. Albrecht, Organometallics, 2008,
27, 3161.
19 (a) J. A. Mata, A. R. Chianese, J. R. Miecznikowski, M. Poyatos, E.
Peris, J. W. Faller and R. H. Crabtree, Organometallics, 2004, 23, 1253;
(b) C. H. Leung, C. D. Incarvito and R. H. Crabtree, Organometallics,
2006, 25, 6099.
28 (a) A. M. Magill and B. F. Yates, Aust. J. Chem., 2004, 51, 1205; (b) A.
M. Magill, K. J. Cavell and B. F. Yates, J. Am. Chem. Soc., 2004, 126,
8717.
29 C. E. Ellul, M. F. Mahon, O. Saker and M. K. Whittlesey, Angew.
Chem., Int. Ed., 2007, 46, 6343.
30 (a) D. L. Davies, S. M. A. Donald and S. A. Macgregor, J. Am. Chem.
Soc., 2005, 127, 13754; (b) D. Garcia- Cuadrado, A. A. C. Braga, F.
Maseras and A. M. Echavarren, J. Am. Chem. Soc., 2006, 128, 1066;
(c) D. L. Davies, S. M. A. Donald, O. Al-Duaij, J. Fawcett, C. Little
and S. A. Macgregor, Organometallics, 2006, 25, 5976; (d) D. Garcia-
Cuadrado, P. de Mendoza, A. A. C. Braga, F. Maseras and A. M.
Echavarren, J. Am. Chem. Soc., 2007, 129, 6880; (e) S. I. Gorelsky, D.
Lapointe and K. Fagnou, J. Am. Chem. Soc., 2008, 130, 10848; (f) D.
R. Stuart, M. Bertrand-Laperle, K. M. N. Burgess and K. Fagnou, J.
Am. Chem. Soc., 2008, 130, 16474; (g) M. Albrecht, Chem. Rev., 2010,
110, 576.
20 In earlier work (ref. 17), only complex 5 was isolated after column
chromatography as this was the expected product (cf. ref. 19a) and
hence purification was not further pursued. Diligent analysis of the
reaction mixture in further experiments revealed the formation of two
more products, viz. 4 and 6, as reported here.
31 Eclipsed interactions may also contribute to the observed reactivity
differences. These would be expected to be more pronounced in C2-
bound A rather than C4-bound Aab because of the different position of
N-substituents relative to the axial iodide ligand.
21 (a) A. Pidcock, R. E. Richards and L. M. Venanzi, J. Chem. Soc. A,
1966, 1707; (b) R. Mason, R. McWeeny and A. D. C. Towl, Discuss.
Faraday Soc., 1969, 47, 20; (c) R. G. Pearson, Inorg. Chem., 1973, 12,
712; (d) J. N. Harvey, K. M. Heslop, A. G. Orpen and P. G. Pringle,
Chem. Commun., 2003, 278; (e) C. R. Landis, T. Cleveland and T. K.
Firman, J. Am. Chem. Soc., 1998, 120, 2641.
22 (a) M. Albrecht, R. H. Crabtree, J. Mata and E. Peris, Chem. Commun.,
2002, 32; (b) M. Poyatos, M. Sanau and E. Peris, Inorg. Chem., 2003,
42, 2572.
23 IUPAC. Compendium, of Chemical Terminology, 2nd ed. (the “Gold
Book”) compiled by A. D. McNaught and A. Wilkinson, Blackwell
Scientific Publications, Oxford, 1997.
24 E. Aldeco-Perez, A. J. Rosenthal, B. Donnadieu, P. Parameswaran, G.
Frenking and G. Bertrand, Science, 2009, 326, 556.
32 M. K. Denk and J. M. Rodezno, J. Organomet. Chem., 2000, 608, 122.
33 S. Gru¨ndemann, A. Kovacevic, M. Albrecht, J. W. Faller and R. H.
Crabtree, J. Am. Chem. Soc., 2002, 124, 10473.
34 Despite that experiments were conducted under strictly anhydrous
conditions, water as a proton source cannot be fully excluded.
35 For related ligand cooperativity, see: (a) D. B. Grotjahn, Chem.–Eur.
J., 2005, 11, 7146; (b) C. Gunanathan, Y. Ben-David and D. Milstein,
Science, 2007, 317, 790; (c) H. Gru¨tzmacher, Angew. Chem., Int. Ed.,
2008, 47, 1814; (d) T. Zweifel, J.-V. Naubron and H. Gru¨tzmacher,
Angew. Chem., Int. Ed., 2009, 48, 559; (e) A. Kru¨ger and M. Albrecht,
Aust. J. Chem., 2011, 64, 1113.
36 A. A. Gridnev and I. M. Mihaltseva, Synth. Commun., 1994, 24, 1547.
37 Stoe & Cie (2000). IPDS-I Bedienungshandbuch. Stoe & Cie GmbH,
Darmstadt, Germany.
38 A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7.
39 Bruker (2001). SADABS, Bruker AXS Inc., Madison, Wisconsin,
USA.
25 Exclusive arrangement of the iodide ligand cis to the alkyl ligand
is in agreement with the a significantly different trans influence of
the normal NHC and the alkyl ligand. Notably, this difference is
considerably smaller between abnormal NHC and the alkyl group,
cf . NMR spectroscopic data of 2 and 4.
40 G. M. Sheldrick, Acta Crystallogr., Sect. A: Found. Crystallogr., 2008,
A64, 112.
9920 | Dalton Trans., 2011, 40, 9911–9920
This journal is
The Royal Society of Chemistry 2011
©