Mechanistic Studies on the Catalytic Oxidative Amination of Alkenes by Rhodium(I) Complexes with Hemilabile Phosphines

Article abstract of DOI:10.1002/cctc.201200510

Cationic rhodium(I) complexes with P,O-functionalised arylphosphine ligands are efficient catalysts for the regioselective anti-Markovnikov oxidative amination of styrene with piperidine. The mechanism of the catalytic reaction has been investigated by spectroscopic means under stoichiometric and catalytic conditions. In the presence of piperidine, the catalyst precursor [Rh{κ²-P,O-Ph₂P(CH₂)₃OEt}₂]⁺ (5) gave the piperidine complex [Rh{κ¹-P-Ph₂P(CH₂)₃OEt}₂(HNC₅H₁₀)₂]⁺ (8) that was transformed into the neutral amido-piperidine species [Rh{κ¹-P-Ph₂P(CH₂)₃OEt}₂(NC₅H₁₀)(HNC₅H₁₀)] (9) under thermal conditions. NMR studies performed in the presence of styrene under catalytic conditions showed that 9 is a key species in the catalytic oxidative amination of styrene. Related cyclooctadiene-containing catalyst precursors [Rh(cod){κ¹-P-Ph₂P(CH₂)₃OEt}_n]⁺ (n=1, 2) also gave 9 under the same conditions. The proposed catalytic cycle has been established by a series of DFT calculations including the transition states of the key steps that have been identified and characterised. These studies have shown that, after elimination of the enamine, regeneration of catalytic active species takes place by direct transfer of the proton of a piperidine ligand to the alkyl group resulting from the insertion of styrene into the Rh-H bond and formation of ethylbenzene. Against the expectations, the formation of a dihydride intermediate by NH oxidative addition is a highly energy-demanding process. Catalyst 5 has also been applied for the oxidative amination of substituted vinylarenes with several secondary cyclic and acyclic amines.

Full text of DOI:10.1002/cctc.201200510

CHEMCATCHEM
FULL PAPERS
DOI: 10.1002/cctc.201200510
Mechanistic Studies on the Catalytic Oxidative Amination
of Alkenes by Rhodium(I) Complexes with Hemilabile
Phosphines
M. Victoria Jimꢀnez,* M. Isabel Bartolomꢀ, Jesffls J. Pꢀrez-Torrente,* Daniel Gómez,
F. Javier Modrego, and Luis A. Oro[a]
Cationic rhodium(I) complexes with P,O-functionalised aryl-
phosphine ligands are efficient catalysts for the regioselective
anti-Markovnikov oxidative amination of styrene with piperi-
dine. The mechanism of the catalytic reaction has been investi-
gated by spectroscopic means under stoichiometric and cata-
lytic conditions. In the presence of piperidine, the catalyst pre-
cursor [Rh{k2-P,O-Ph2P(CH2)3OEt}2]+ (5) gave the piperidine
complex [Rh{k1-P-Ph2P(CH2)3OEt}2(HNC5H10)2]+ (8) that was
transformed into the neutral amido–piperidine species [Rh{k1-
P-Ph2P(CH2)3OEt}2(NC5H10)(HNC5H10)] (9) under thermal condi-
tions. NMR studies performed in the presence of styrene under
catalytic conditions showed that 9 is a key species in the cata-
lytic oxidative amination of styrene. Related cyclooctadiene-
containing catalyst precursors [Rh(cod){k1-P-Ph2P(CH2)3OEt}n]+
(n=1, 2) also gave 9 under the same conditions. The proposed
catalytic cycle has been established by a series of DFT calcula-
tions including the transition states of the key steps that have
been identified and characterised. These studies have shown
that, after elimination of the enamine, regeneration of catalytic
active species takes place by direct transfer of the proton of
a piperidine ligand to the alkyl group resulting from the inser-
tion of styrene into the RhÀH bond and formation of ethylben-
zene. Against the expectations, the formation of a dihydride in-
termediate by NH oxidative addition is a highly energy-de-
manding process. Catalyst 5 has also been applied for the oxi-
dative amination of substituted vinylarenes with several secon-
dary cyclic and acyclic amines.
Introduction
The direct addition of ammonia, primary or secondary amines
to alkenes affords amines, products of the formal NÀH bond
addition to the unsaturated C=C bond, or enamines as a result
of an oxidative amination.[1] Both product classes are highly
valuable building blocks in fine-chemical synthesis, and partic-
ularly amines are also of great relevance for bulk chemicals
synthesis. The intermolecular hydroamination is an efficient al-
ternative to the classic organic synthesis procedures for the
preparation of nitrogen-containing compounds. However, the
uncatalysed reactions have very high energy barriers thereby
requiring the use of a catalyst. In contrast to the rapid devel-
opment of the catalysed intramolecular amination methods to
produce cyclic aminoalkanes or aminoalkenes, the intermolecu-
lar hydroamination–oxidative amination reactions have experi-
enced less progress, and, in particular, the anti-Markovnikov
functionalisation of unactivated olefins with amines still re-
mains a challenge in catalysis. In this context, mechanistic
studies are of key importance for the development of efficient
catalysts to achieve this important goal.
The general mechanisms commonly accepted for lantha-
nide[1,2] or group 4 metal hydroamination catalysts[3] are the
migratory insertion of the alkene into a metal-amide species
followed by proton transfer,[4] and the [2+2] addition of al-
kenes and metal-imido intermediates.[1b,5] On the other hand,
two comprehensive mechanisms have been described for late-
transition-metal-based hydroamination catalysts (Scheme 1):
the nucleophilic attack at the coordinated alkene (olefin activa-
tion) and the oxidative addition of the amine on the metal
centre, followed by olefin insertion into the metalÀamide bond
(amine activation).
The key step of the first mechanistic model is the activation
of the alkene by coordination to a Lewis acidic metal centre,
which renders the p-system susceptible to nucleophilic attack
by the amine. Subsequent protolytic cleavage of the metalÀ
carbon bond in the zwitterionic 2-ammoniumalkyl complex af-
fords the amine product after decoordination from the metal
centre. Interestingly, several experimental findings on late-tran-
sition-metal-catalysed hydroamination reactions supporting
this mechanism have been recently reported, for both intermo-
lecular[6] and intramolecular versions.[7]
[a] M. V. Jimꢀnez, M. I. Bartolomꢀ, Prof. Dr. J. J. Pꢀrez-Torrente, D. Gꢁmez,
Dr. F. J. Modrego, Prof. Dr. L. A. Oro
Departamento de Quꢂmica Inorgꢃnica
Instituto de Sꢂntesis Quꢂmica y Catꢃlisis Homogꢀnea
Universidad de Zaragoza-C.S.I.C.
50009-Zaragoza (Spain)
E-mail: vjimenez@unizar.es
On the other hand, activation of the amine by oxidative ad-
dition to a coordinatively unsaturated late transition metal in
low oxidation states has been proposed frequently as a poten-
tial pathway for hydroamination catalysed by ruthenium(0),
perez@unizar.es
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201200510.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 263

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
dine.[14,15] Interestingly, the catalyst containing P,O-functional-
ised arylphosphine ligands with a 3-alkoxypropyl hemilabile
moiety, Ar2P(CH2)3OR (R=Me, Et, nBu), exhibited a higher cata-
lytic activity than catalysts containing P,N-functionalised phos-
phine ligands. An important feature of these flexible ligands is
their potential for a weak interaction with the rhodium centre
through the oxygen atom forming a six-membered metallocy-
cle ring. Although each single reaction step of the generally
accepted mechanism appears feasible,[1] the reaction pathway,
both for the catalytic hydroamination and oxidative amination,
still remains ambiguous with regard to the initial oxidative ad-
dition of the amine.[16] In the present work, we have been par-
ticularly interested in determining the reaction mechanism to
provide a basis for the design of improved catalysts. Thus,
mechanistic studies under both stoichiometric and catalytic
conditions have been performed. In addition, a plausible cata-
lytic cycle and a potential role of the ether-functionalised
phosphine ligands have been established by DFT calculations.
Results and Discussion
Scheme 1. Mechanisms for late-transition-metal-based hydroamination
catalysts.
Rhodium complexes with hemilabile phosphines as efficient
catalysts for oxidative amination
rhodium(I), iridium(I), copper(I), palladium(0) or platinum(0)
metal complexes. The oxidative addition of the amine results
in the formation of a hydrido–amido intermediate. Subse-
quently, insertion of the alkene into the MÀN or MÀH bond fol-
lowed by reductive elimination affords the amine product and
regenerates the active metal species (Scheme 1). Although the
NÀH bond activation has been postulated as the first step in
the mechanism, there are only a few examples of direct forma-
tion of amido complexes from amines, most of them in iridium
complexes.[8,9] Migratory insertion of alkenes into the MÀNR2
bond is a favourable process in late-transition-metal com-
plexes. Theoretical calculations for the migratory insertion of
the alkene in complexes [(PMe3)2Rh(h2-alkene)(NH2)] evidenced
low-energy barriers with energies even lower than those of the
migratory insertion into a RhÀCH3 bond.[10] Similar results have
been reported for group 10 hydrido–amido metal complexes
[MH(NR2)(PEt3)2] (M=Pd, Pt), in which the insertion of alkenes
into the MÀNR2 bond is preferred over the insertion into the
MÀH bond.[11] Finally, the last step of the catalytic cycle leading
to the amine product is reductive elimination.[12] However, re-
ductive elimination competes with b-hydride elimination re-
sulting in the formation of the enamine product (oxidative
amination), which is necessary in this case as a hydrogen ac-
ceptor to regenerate the active species. The oxidative amina-
tion products were first observed by Brunet et al. in the hydro-
amination of styrene with aniline catalysed by the system
[RhCl(PEt3)2]2/LiPhNH that mainly affords the Markovnikov hy-
droamination compound, N-(1-phenylethyl)aniline, in a regiose-
lective way.[13]
Cationic rhodium(I) complexes containing P,O-functionalised ar-
ylphosphine ligands exhibited an outstanding catalytic activity
for the regioselective anti-Markovnikov oxidative amination of
styrene with piperidine, a benchmark reaction for the intermo-
lecular alkene hydroamination.[14,15] In particular, complexes
with arylphosphine ligands with a 3-ethoxypropyl hemilabile
moiety exhibited a superior catalytic activity.[17]
We have observed that the catalyst structure has a strong in-
fluence on the catalytic activity (Table 1). In particular, com-
plexes featuring two k1-P-coordinated hemilabile ligands, [Rh-
(cod){Ar2P(CH2)3OEt}2]+ (Ar=Ph, 3; Ar=4-tolyl, 4), exhibited
higher catalytic activity than the complexes [Rh(cod){Ar2P-
(CH2)3OEt}]+ (Ar=Ph, 1; Ar=4-tolyl, 2) with only one hemila-
bile ligand k2-P,O coordinated (entries 1–4). Interestingly, bi-
sphosphine complexes [Rh{Ar2P(CH2)3OEt}2]+, with two k2-P,O
coordinated ligands with a cis disposition (Ar=Ph, 5; Ar=4-
tolyl, 6), showed an outstandingly high catalytic activity (en-
tries 5 and 6). The best catalytic performance was observed
with catalyst [Rh{(4-tolyl)2P(CH2)3OEt}2][PF6] (6) giving turnover
frequencies up to 85 hÀ1 with excellent enamine selectivity
(96%).[18]
Activation of the precatalyst
The cationic complexes [Rh{Ph2P(CH2)3OEt}2]+ (5) and [Rh{(4-
tolyl)2P(CH2)3OEt}2]+ (6) are reactive species. Both compounds
quickly reacted with piperidine to give piperidine complexes
as result of the decoordination of the hemilabile fragment of
the functionalised phosphine ligands. Addition of one equiva-
lent of piperidine to a solution of 5 in [D8]THF at room temper-
ature gave the compound [Rh{Ph2P(CH2)3OEt}2(HNC5H10)]+ (7)
with k1-P and k2-P,O coordinated ligands (Scheme 2). Com-
pound 7 is fluxional as it was evidenced in the 31P{1H} NMR
We have recently found that rhodium(I) complexes contain-
ing hemilabile phosphine ligands of the type Ar2P(CH2)nZ (Z=
OR, NR2, SR; n=2, 3) are active catalysts for the regioselective
anti-Markovnikov oxidative amination of styrene with piperi-
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 264

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Table 1. Oxidative amination of styrene with piperidine catalysed by
complexes [Rh(cod)n{Ar2P(CH2)3OEt}m]BF4 (n=0, 1; m=1, 2).[a]
Conversion[b] Selectivity[c]
Figure 1. 31P{1H} NMR ([D8]THF, 253 K) spectra for the reaction of [Rh{Ph2P-
(CH2)3OEt}2][PF6] (5) with a) 0.8, b) 1, c) 2, and d) 40 equiv. piperidine.
Catalyst
t
[h] [%]
1
2
3
4
5
6
[Rh(cod){Ph2P(CH2)3OEt}]+, 1
[Rh(cod){(4-tolyl)2P(CH2)3OEt}]+, 2
[Rh(cod){Ph2P(CH2)3OEt}2]+, 3
[Rh(cod){(4-tolyl)2P(CH2)3OEt}2]+, 4
[Rh{Ph2P(CH2)3OEt}2]+, 5
20 82
20 88
97:3
96:4
97:3
97:3
96:4
96:4
terised in solution at low temperature. Both species are in
equilibrium as it was shown by a 31P{1H} NMR study at variable
temperature on a [D8]THF solution of 5 and two equivalents of
piperidine.
8
8
1
1
96
98
93
97
[Rh{(4-tolyl)2P(CH2)3OEt}2]+, 6
[a] Reaction conditions: styrene (4 mmol), piperidine (1 mmol), solvent
(2.5 mL), [Rh]=0.01m in THF (2.5 mol%), 808C. [b] Standard conditions of
analysis. Conversion relative to piperidine (%) determined by GC analysis
using tetradadecane as internal standard. [c] Selectivity: (E)-1-styrylpiperi-
dine/1-phenethylpiperidine ratio.
As can be seen in Figure 2, this equilibrium is shifted to 8 at
temperatures down to 178 K. Increasing the temperature re-
sults in the formation of 7 as a consequence of the equilibrium
displacement. The calculated equilibrium constant at 273 K is
spectrum at 298 K that showed two broad resonances at d
ꢀ51 and 33 ppm. However, the 31P{1H} NMR spectrum at 253 K
consisted of two doublets of doublets at d 50.24 (JPÀRh =215.0)
and 30.23 ppm (JPÀRh =161.0) with a JPÀP coupling constant of
54.4 Hz (Figure 1), in full agreement with the proposed struc-
ture. In contrast, reaction of 5 with an excess of piperidine
(1:40) in [D8]THF gave [Rh{Ph2P(CH2)3OEt}2(HNC5H10)2]+ (8).
Complex 8 is also fluxional, but at 253 K the 31P{1H} NMR spec-
trum (Figure 1) showed a sharp doublet at d 38.29 ppm
(JPÀRh =170.2 Hz), which is in accordance with a symmetric mol-
ecule with two piperidine ligands and two ether-phosphine li-
gands k1-P coordinated with a cis disposition. Compounds 7
and 8 could not be isolated in the solid state and were charac-
Figure 2. Variable-temperature 31P{1H} NMR ([D8]THF) study for the reaction
of 5 with 2 equiv. of Hpip (piperidine).
K=2.62. The thermodynamic parameters estimated from the
van’t Hoff representations (lnK vs. 1/T) in the temperature
range 193–288 K were DH8=(6.9Æ0.4) kcalmolÀ1 and DS8=
(29.0Æ1.7) calKÀ1 molÀ1 (see the Supporting Information). DFT
calculations on these species have shown that both complexes
7 and 8 are more stable than 5 by À7.2 and À19.6 kcalmolÀ1,
respectively, which is in full agreement with the thermodynam-
ically calculated data. The optimised structure of 8 revealed
a long-range interaction (ꢀ4 Š) between the oxygen atom of
an ethoxy group and the NH group of one of the piperidine li-
gands. However, several conformers of both complexes with
a hydrogen bond (ꢀ2.9 Š) are only slightly higher in energy
(less than 4 kcalmolÀ1 difference), which could be in part re-
sponsible for the enhanced stability of 8 over 7.
If a mixture of the piperidine compounds 7 plus 8, obtained
by reaction of 5 with a slight excess of piperidine (1:3) in
[D8]THF, was heated at 343 K for 2 h, a clean formation of the
new species 9 was observed. Its 31P{1H} NMR spectrum at room
Scheme 2. Sequential formation of piperidine complexes from 5.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 265

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
temperature consisted of two doublets of doublets at d 39.60
and 36.31 ppm with JPÀRh coupling constants of 165.0 and
173.1 Hz, and JPÀP of 51.3 Hz (Figure 3). On the other hand, the
1H NMR spectrum evidenced the presence of the piperidonium
by an ethoxy group of one of the hemilabile ligands, a solvent
molecule (THF) or piperidine. A possible dimerisation of this
unsaturated species could be ruled out because a symmetrical
amido-bridged dinuclear complex would be formed which is
not compatible with the observed spectroscopic data.[10b,20]
DFT calculations have shown that the amido species [Rh{Ph2P-
(CH2)3OEt}2(NC5H10)(THF)] (9b) and [Rh{Ph2P(CH2)3OEt}2(NC5H10)-
(HNC5H10)] (9c-exo) are À12.0 and À7.3 kcalmolÀ1 more stable,
respectively, than [Rh{Ph2P(CH2)3OEt}2(NC5H10)] (9a) with one of
the hemilabile phosphine ligands k2-P,O coordinated (Figure 4).
Figure 3. 31P{1H} NMR spectrum ([D8]THF, 298 K) of 9.
cation [H2NC5H10]+ which showed a broad signal at d 4.38 for
the NH2 group and two resonances at d 2.85 and 1.57 ppm for
the methylene protons.[19] The simultaneous formation of 9
and [H2NC5H10]+ strongly supports the formulation of 9 as an
amido-rhodium(I) complex, for which the structure, supported
by DFT calculations, is shown in Scheme 3. Unfortunately, we
Figure 4. Possible structures for the neutral amido species 9.
However, optimisation of the structure of 9 with the NH
proton in a relative endo disposition to the amido ligand (9c-
endo) resulted in a net stabilization energy of À12.11 kcalmolÀ1
relative to 9a. In this optimised structure, the hydrogen bond
between the NH of the coordinated piperidine and the nitro-
gen atom of the amido ligand (NH···N distance of 1.856 Š) pro-
duces a narrow NÀRhÀN angle of 73.918 (Figure 5). This inter-
action probably blocks the two coordination sites thereby pre-
venting the piperidine exchange and, in contrast with com-
pound 8, 9 is static.
Thus, DFT calculations allow us to formulate 9 as a neutral
amido–piperidine square-planar rhodium(I) complex [Rh{Ph2P-
(CH2)3OEt}2(NC5H10)(HNC5H10)] stabilised by a NH···N hydrogen
bond (Scheme 3).
Scheme 3. Formation of the amido–piperidine rhodium(I) 9 from 5.
The reactivity of the cyclooctadiene complexes [Rh(cod){k2-
P,O-Ph2P(CH2)3OEt}]+ (1) and [Rh(cod){k1-P-Ph2P(CH2)3OEt}2]+ (3)
with piperidine was also monitored by NMR spectroscopy. The
addition of one equivalent of piperidine to a suspension of
complex [Rh(cod){Ph2P(CH2)3OEt}]+ (1) in [D8]THF gave an equi-
molecular mixture of complexes [Rh(cod){Ph2P(CH2)3OEt}2]+ (3)
and [Rh(cod)(HNC5H10)2]+ (10) as a result of a ligand redistribu-
tion reaction (Scheme 4). Compound 10 has been identified in
were unable to observe the amido ligand by 15N NMR spectros-
copy; however, it has been possible to identify the protons
and carbons of the amido ligand, all of them inequivalent,
through H,1H COSY and H,13C HSQC correlation experiments
(see the Supporting Information). In sharp contrast, if a solution
of complex [Rh{Ph2P(CH2)3OEt}2(HNC5H10)]+ (7) in [D8]THF was
heated for 20 h, formation of 9 was never observed. This result
supports the active role of piperidine in the formation of this
key amido–rhodium species from 8.
Most probably, 9 resulted from the deprotonation of one of
the coordinated piperidine ligands in 8 by the second piperi-
dine ligand (intramolecular) or, alternatively, by an external pi-
peridine molecule (intermolecular). Elimination of [H2NC5H10]+
produced a coordination vacant site which could be occupied
1
1
1
the H NMR spectrum by comparison with a sample independ-
ently prepared by reaction of [Rh(cod)2]+ with two equivalents
of piperidine in [D8]THF.
Addition of 40 equiv. of piperidine to a suspension of com-
plex [Rh(cod){Ph2P(CH2)3OEt}2]+ (3) in [D8]THF and heating the
mixture for approximately 20 min at 333 K, or 24 h at room
temperature, gave a dark red solution containing a new com-
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 266

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
The spectroscopic data for 11 suggest a symmetrical structure
with equivalent phosphine ligands and are compatible with
a square-pyramidal structure with the amido ligand in the
apical position. More importantly, if a solution of 11, prepared
by reaction of 3 with an excess of piperidine, was heated at
333 K for 4 h, the only species observed in the 31P{1H} NMR
spectrum was the amido complex 9 (Scheme 5). The transfor-
mation of 11 into 9 requires the elimination of cyclooctadiene.
Notably, investigation of some of the catalytic solutions by
GC–MS led to the detection of a peak with an m/z ratio of 191
corresponding to 1-(cycloocta-1,5-dien-1-yl)piperidine.[14] Thus,
the oxidative amination of the coordinated 1,5-cyclooctadiene
ligand is necessary for the generation of active catalytic species
if using complexes 1 and 3 as catalyst precursors.
The catalytic active species at work
The precedent reactivity studies have shown that the three dif-
ferent types of catalyst precursors react with piperidine under
thermal conditions to give the same species. Spectroscopic
data and DFT calculations allow us to propose the neutral
amido–piperidine
complex
[Rh{Ph2P(CH2)3OEt}2(NC5H10)-
(HNC5H10)] (9) as a key species in the oxidative amination of
styrene with piperidine. To confirm this hypothesis, further
NMR studies were performed in the presence of styrene under
catalytic conditions.
Thus, styrene and piperidine were added to a solution of 5
in [D8]THF with a molar ratio of 160:40:1, heated in a thermo-
1
statised bath at 338 K and then monitored by H NMR spec-
troscopy at the same temperature (Figure 6). Under these cata-
lytic conditions, the formation of (E)-1-styrylpiperidine and eth-
ylbenzene was detected after 20 min. The two resonances at d
6.49 and 5.26 ppm correspond to the C=H protons of (E)-1-
styrylpiperidine. Ethylbenzene was identified by the
characteristic triplet and quartet resonances at d 1.21
and 2.58 ppm of the ethyl substituent. As expected,
an increase in the intensity of these resonances and
a concomitant decrease of the resonances corre-
sponding to styrene was observed during the reac-
tion time.
Figure 5. DFT geometry-optimized structure of a) 9c-exo and b) 9c-endo.
In a parallel experiment, a [D8]THF solution con-
taining catalyst 5, piperidine and styrene at
a 1:40:160 molar ratio was heated in a thermostatised
bath at 338 K and monitored regularly by
31P{1H} NMR spectra at 273 K. After 20 min, a mixture
of the cationic piperidine complex 8 and the neutral
species 9 was observed. Interestingly, 9 was the only
species detected after heating the mixture at 338 K
for 1 h (Figure 7).
Scheme 4. Ligand redistribution reaction involving 1 and piperidine.
The results above suggest that the piperidine–
amido complex 9 could be a competent catalyst for the cata-
lytic hydroamination of styrene. Formation of 9 from the cata-
lyst precursor 5 requires that the amine is activated through
the piperidine intermediate species 8. However, to rule out the
possible involvement of the olefin in the generation of an
active species, we have studied the reactivity of 5 with styrene.
Addition of a large excess of styrene to a solution of 5 in
pound 11. The 31P{1H} NMR spectrum showed a doublet at d
30.80 ppm (JPÀRh =200.2 Hz), and the 1H NMR spectrum evi-
denced the presence of coordinated cyclooctadiene. Interest-
ingly, compound 11 has also been observed in the reaction of
[Rh(cod){Ph2P(CH2)3OEt}2]+ (3) with LiNC5H10 in [D8]THF or C6D6.
Thus, this compound has been tentatively identified as the
neutral amido complex [Rh(cod){Ph2P(CH2)3OEt}2(NC5H10)] (11).
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 267

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
styrene, and, more importantly, in the steady forma-
tion of the amido–piperidine 9, even at room tem-
perature.
The spectroscopic information obtained from the
reactivity studies confirms that the mechanism of oxi-
dative amination of styrene proceeds through the ac-
tivation of the amine, regardless of the catalyst pre-
cursor, and involve the same key catalytic species
that is most probably the neutral amido–piperidine
[Rh{Ph2P(CH2)3OEt}2(NC5H10)(HNC5H10)] (9). Conse-
quently, the catalytic activity of the different catalyst
precursors must be related to the concentration of
this species. This interpretation is supported by the
similar regioselectivity, which was approximately 96%
in (E)-1-styrylpiperidine, observed for the three types
of catalytic precursors (Table 1).
Scheme 5. Formation of the amido–piperidine rhodium(I) 9 from 3.
Figure 7. 31P{1H} NMR spectra at 273 K of the oxidative amination of styrene
with piperidine catalysed by complex 5 at different reaction times. Styrene/
piperidine/catalyst ratio 160:40:1; [Rh]=0.01m in 0.5 mL [D8]THF. Conversion
to enamine: t=20 min, 13%; t=1 h, 35%; t=9 h, 98%. OP=phosphine
oxide.
Mechanism of the catalytic oxidative amination
The regioselective intermolecular anti-Markovnikov hydroami-
nation and oxidative amination of aromatic olefins with secon-
dary amines catalysed by the system [Rh(cod)2]BF4/2PPh3 was
studied by Beller et al.[21] The reaction proceeds through inde-
pendent catalytic amination cycles and kinetic investigations
have shown that in both cases the reaction rates are depen-
dent both on styrene and catalyst concentrations, and inde-
pendent of the amine concentration. Mechanistic studies al-
lowed the conclusion that rhodium hydride complexes were
participating in the catalytic cycle. The proposed mechanism,
which is shown in Scheme 6, follows an amine activation path-
way, in which the key hydrido–amido intermediate is formed
by oxidative addition of the amine.
Figure 6. 1H NMR spectra at 338 K of the oxidative amination of styrene with
piperidine catalysed by complex 5 at different reaction times. Styrene/piperi-
dine/catalyst ratio 160:40:1; [Rh]=0.01m in 0.5 mL [D8]THF.
The formation of the amido complex 9 from [Rh{Ph2P-
(CH2)3OEt}2(HNC5H10)2]+ (8) in the presence of piperidine
(Scheme 3) probably takes place through the intermolecular
deprotonation of a coordinated piperidine ligand. An alterna-
tive pathway for the formation of 9 could be the oxidative ad-
dition of piperidine followed by deprotonation of the resulting
hydride complex; however, we failed to observe any hydride
intermediate in the formation of 9 by 1H NMR spectroscopy,
both under stoichiometric and catalytic conditions. On the
[D8]THF (160:1) resulted in the formation of a symmetric spe-
cies 12, which was observed at d 33.93 ppm (JPÀRh =204.1 Hz)
in the 31P{1H} NMR spectrum at 273 K. The spectroscopic data
for 12 suggest the formation of an olefin adduct [Rh{Ph2P-
(CH2)3OEt}2(styrene)n]+, which has never been observed by
31P{1H} NMR spectroscopy under catalytic conditions before. In
agreement with this, addition of 40 equiv. of piperidine result-
ed in the formation of 8 by replacement of the coordinated
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 268

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Scheme 6. Mechanism proposed by Beller for the oxidative amination of ar-
omatic olefins with secondary amines.[21, 26] Ar=aryl.
other hand, the intermolecular transfer of amido ligands from
organometallic amides to unactivated olefins has been recently
reported.[22,23] On the basis of these results, we propose that
the oxidative amination catalysed by rhodium complexes con-
taining P,O-functionalised phosphine ligands proceeds through
the catalytic cycle shown in Scheme 7
The first step in the proposed mechanism is the replacement
of the piperidine ligand by styrene in 9 to give a key styrene–
amido intermediate. Then, insertion of styrene into the rhodi-
umÀamide bond should result in the formation of an alkyla-
mine intermediate, giving the enamine product by a b elimina-
tion process. Finally, regeneration of the catalytic active species
requires the hydrogenation of a second styrene molecule as it
is discussed below.
The influence of the styrene concentration on the activity
and selectivity was also investigated (Table 2). Under the stan-
dard catalytic conditions a molar ratio of catalyst/amine/olefin
of 1:40:160 was used (olefin/amine ratio of 4). In general, the
catalytic activity increased with increasing the olefin concentra-
tion. For example, for catalyst 5 an increase in the olefin/amine
ratio to 8 resulted in an upsurge of the conversion from 66%
to 94% in only half an hour (entry 2, Table 2). The positive in-
fluence of the olefin concentration on the catalytic activity can
be rationalised by a shift of the equilibria involving the amino–
amido species
9 towards the olefin–amido intermediate
Scheme 7. Proposed mechanism for the oxidative amination of styrene with
piperidine catalysed by rhodium complexes containing P,O-functionalised
phosphine ligands. Ph=phenyl; TS=transition state.
a (Scheme 7), or from e towards d intermediates. In addition,
Table 2. Oxidative amination of styrene with piperidine catalysed by rho-
dium bisphosphine complexes at different olefin/amine ratios.[a]
this positive effect is also in agreement with the hydrogen ac-
ceptor role of the olefin in the generally accepted oxidative
amination mechanism[1,24] and related borylation and silylation
processes.[25]
Catalyst
t [h]
Styrene/piperidine ratio
4
2
8
X[b] [%]
S[c]
X[b] [%]
S[c]
X[b] [%]
S[c]
On the other hand, the observed selectivity is practically the
same for olefin/amine ratios higher than 4 and only a slightly
decrease of selectivity was observed with an olefin/amine ratio
of 2 (Table 2). These results are parallel to those obtained by
Beller et al.[21,26] and contrast with those observed by Hartwig
et al. for the anti-Markovnikov hydroamination of vinylarenes
with secondary amines using a rhodium complex of DPEphos
3
5
6
4
0.5
0.5
50
42
46
94:6
93:7
93:7
92
66
75
96:4
95:5
96:4
100
94
96
96:4
95:5
95:5
[a] Reaction conditions: piperidine/catalyst ratio 40:1, [Rh]=0.01m in
2.5 mL THF, 808C. [b] Conversion under standard analysis conditions rela-
tive to the piperidine determined by GC analysis with tetradadecane as
internal standard. [c] (E)-1-styrylpiperidine/1-phenethylpiperidine ratio.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 269

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
(bis(2-diphenylphosphinophenyl)ether).[27] The higher enamine/
amine ratios at higher concentrations of vinylarene was inter-
preted as the participation of two vinylarenes in the transition
cycle is shown in Figure 8. The energies reported are gas-
phase values with ZPE corrections. The substitution processes
were assumed to require low activation energy values.[29–31]
The neutral square-planar rhodium(I) amino–amido complex
9 was the precursor for the catalytic active species: the olefin–
amido intermediate [Rh{Ph2P(CH2)3OEt}2(NC5H10)(styrene)] (a,
Scheme 7). The resulting complex a presents a distorted
square-planar geometry, in which the amido N-atom forms an
angle of 95.088 with the CÀC centroid of the olefin group that
lies roughly perpendicular to the coordination plane. DFT cal-
culations showed that the styrene complex is less stable than
the piperidine complex 9 by +7.64 kcalmolÀ1, and conse-
quently the replacement of the piperidine ligand in 9 by sty-
rene is an endothermic process.
state that controls the selectivity in
intermediate.
a hydrido–rhodium
DFT calculations on the mechanism of the catalytic oxidative
amination
As it is apparent from the proposed catalytic cycle, the hemila-
bile ligand can play an important role in the mechanism be-
cause of its potential for the reversible protection of one or
more coordination sites by an alternating “on/off” mechanism.
To understand the outstanding catalytic activity of the com-
plexes with hemilabile P,O-functionalised phosphine ligands,
we studied the key steps of the proposed mechanism by DFT
calculations.
The styrene coordination was followed by the insertion into
the RhÀamido bond to produce an aminoalkyl intermediate
b (Scheme 7). The formation of a metallacyclic a-amino-b-ary-
lalkyl group stabilizes the system by À16.17 kcalmolÀ1. The
process requires an activation energy of DE= +14.27 kcal
molÀ1 through the transition state TSa–b (Figure 8), which has
been confirmed by frequency calculations. In the transition
state the CÀN distance has reduced from the initial value of
3.027 Š (NÀRhÀC angle of 88.778) to 2.188 Š (NÀRhÀC angle of
The catalytic cycle was studied by a series of DFT calcula-
tions on the successive steps (Scheme 7) by using the B3LYP
functional as implemented in Gaussian 09,[28] and with
a Lanl2dz basis and ECP for rhodium and 6-31G** basis for the
other elements. The transition states of the key steps were
found and characterized. The energy profile for the catalytic
Figure 8. Calculated energy profile of the catalytic cycle for the catalytic oxidative amination of styrene in Scheme 7.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 270

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
54.148); see Figure 9a. Inspection of the imaginary mode gives
the connection to the starting and final species of the process.
The structural features of the metallacyclic aminoalkyl insertion
product deserve further comments. The insertion process af-
fords a coordinatively unsaturated species that could be stabi-
lized by coordination of a pending ethoxy group on a hemila-
bile phosphine ligand. However, the weak coordination ability
of this group is pointed out by the fact that optimization in va-
cuo starting from a chelating ether-phosphine ligand leads to
an h3-benzyl coordination of the aminoalkyl ligand. Neverthe-
less, this species is less stable than b with the aminoalkyl
ligand k2-C,N coordinated. The two protons of the methylene
group in the four-membered metallacycle have equatorial and
axial dispositions, respectively. The rhodium–axial-proton dis-
tance, the shortest one, was driven to shorter values searching
for a transition state towards the elimination process leading
to the (E)-1-styrylpiperidine reaction product.
The formation of the p-styrylpiperidine complex c by b-hy-
dride elimination requires +4.94 kcalmolÀ1 through a transition
state, TSb–c,with an activation energy of +7.55 kcalmolÀ1
(Figure 8). The hydrogen distances in TSb–c are 1.623 Š to the
carbon atom and 1.625 Š to the Rh centre (Figure 9b). The
rhodium coordination environment is almost planar with both
P and C atoms and the eliminated hydrogen atom in roughly
the same plane. On the other hand, the coordinated styrylpi-
peridine ligand in intermediate c can be replaced by any of
the following ligands: a molecule of styrene, a molecule of pi-
peridine or the ethoxy group of the free arm of the phosphine.
Whereas the last process is slightly endothermic (+4.93 kcal
molÀ1), the replacement of the olefin by styrene to give d
(À9.39 kcalmolÀ1) or by piperidine to give e (À12.72 kcalmolÀ1
)
are exothermic and close in energy which suggest that both
species could be in equilibrium (Figure 8).
The catalytic reaction yielded equimolar amounts of aminat-
ed and hydrogenated styrene products, which suggests that
the hydrogenation is concomitant with the amination process.
From here on, two paths leading to styrene hydrogenation
could be devised. In one way, the coordination and oxidative
addition of piperidine in e would generate an amido–dihydrido
complex that could react with styrene leading to hydrogena-
tion. In this pathway the resulting dihydrido complex is
26.3 kcalmolÀ1 above the hydrido-piperidine reactant, and the
NÀH oxidative addition process requires an activation energy
of 32.09 kcalmolÀ1 through the TSNH transition state (dashed
line pathway, Figure 8).
The second pathway, which is the less energetically demand-
ing, occurs through substitution of p-styrylpiperidine in com-
plex c by styrene (solid-line pathway, Figure 8). This complex
can generate an alkyl intermediate f by a facile olefin insertion
into the RhÀH bond, which is slightly endothermic by only
0.5 kcalmolÀ1, and requires only 4.32 kcalmolÀ1 of activation
energy (TSd–f, Figure 10a). Coordination of piperidine in the re-
sulting vacant coordination site stabilizes the product by
À6.56 kcalmolÀ1 (g, Figure 8). From this point, a direct transfer
of the amine proton from the piperidine ligand to the alkyl
group requires just an activation energy of 22.38 kcalmolÀ1
(TSg–a, Figure 10b) to give the hydrogenated product ethylben-
zene (+7.55 kcalmolÀ1) completing the catalytic cycle. The dif-
ference in energy between both alternatives suggests that this
second pathway is the operating one in the catalytic cycle.
The results described above suggest that the formation of
the amine product, 1-phenethylpiperidine (Table 1) could take
place from the metallacyclic aminoalkyl intermediate b,
through the coordination of piperidine and proton transfer of
the piperidine proton to the aminoalkyl ligand resulting in the
regeneration of the active catalyst species a. On the other
hand, the oxidative addition of piperidine to b seems to be un-
likely in the light of the precedent results. Thus, the high selec-
tivity for the enamine product is a consequence of the stability
of the metallacyclic aminoalkyl intermediate b with a conforma-
Figure 9. Structures for transition states a) TSa–b and b) TSb–c
.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 271

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
bile arms of the phosphine ligands on the same side of the
molecular plane towards the metal centre, facilitating the coor-
dination of both ethoxypropyl fragments and, in turn, the pro-
tection of the two accessible coordination sites necessaries for
the catalysis (Figure 11). Thus, the observed hemilabile effect
Figure 11. cis Arrangement of the O-functionalised phosphine ligands in an
intermediate model showing a p–p intramolecular interaction.
could be related with the stabilisation of polar species in solu-
tion, as the key intermediate [Rh{Ph2P(CH2)3OEt}2(NC5H10)-
(HNC5H10)] (9), or unsaturated catalytic intermediates by transi-
ent O-coordination as for example, those resulting after the de-
coordination of the p-styrylpiperidine ligand in c or the styrene
insertion in d.
Scope and limitations of the anti-Markonikov oxidative ami-
nation of aromatic olefins
The catalyst [Rh{Ph2P(CH2)3OEt}2][PF6] (5) has been applied for
the oxidative amination of substituted vinylarenes with several
amines (Table 3). The reactions either of electron-poor or elec-
tron-rich vinylarenes with piperidine in the presence of
2.5 mol% of 5 gave the corresponding (E)-1-styrylpiperidine
Table 3. Anti-Markovnikov oxidative amination of vinylarenes with secon-
dary amines catalysed by 5.[a]
Figure 10. Structures for transition states a) TSe–f and b) TSg–a
.
tion that facilitates the b-hydride elimination to the p-styrylpi-
peridine intermediate c.
Amine
R’’
t
[h]
Conversion[b]
[%]
Selectivity[c]
The computational study on the full catalytic cycle for the
oxidative amination of styrene with piperidine has failed to
identify the role of the hemilabile fragment of the functional-
ised phosphine ligands. The 3-ethoxypropyl moieties of both
phosphine ligands are far away from the rhodium centre in the
optimised structures of the intermediates. In addition, these
groups do not seem to have any influence on the transition
states involved in the keys steps of the catalytic cycle.
1
2
3
4
5
6
7
8
9
piperidine
piperidine
piperidine
piperidine
piperidine
piperidine
morpholine
pyrrolidine
Et2NH
p-Me
m-Me
o-Me
p-tBu
p-CF3
p-OMe
H
H
H
H
H
2
3
7
94
95
72
58
64
79
84
5
95:5
94:6
93:7
91:9
89:11
94:6
6
11
8
1
2
2
8
95:5
100:0
100:0
100:0
100:0
However, the relative cis arrangement of the functionalised
phosphine ligands in the intermediates along the catalytic
cycle could play an important role. Determination of the mo-
lecular structures of complexes 5 and 6 allowed the identifica-
tion of a p–p intramolecular interaction between two phenyl
groups of the two phosphines.[14,15] This interaction provides
a rigid scaffold in the intermediates that directs both hemila-
11
87
85
Et2NH
nBu2NH
10
6
[a] Reaction conditions: styrene (4 mmol), piperidine (1 mmol), solvent
(2.5 mL), [Rh]=0.01m in THF, 808C. [b] Standard analysis conditions; con-
version relative to piperidine (%) determined by GC analysis with tetrada-
decane as internal standard. [c] Enamine/amine ratio.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 272

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
derivatives with excellent selectivities (entries 1–6). In general,
the catalytic activity was negatively affected by the presence
of a substituent in the styrene. In particular, ortho substituents
largely influenced the catalytic activity although the steric in-
fluence was also observed for substrates with substituents in
para position (entries 1–4). The introduction of electron density
in the styrene aromatic ring in the para position resulted in
a slight decrease in the catalytic activity, which was significant-
ly higher for the more reactive electron-poor styrenes (entry 5).
These results suggest a complex interplay between the elec-
tronic and steric effects exerted by the substituent. In addition
to the decrease of activity, the substitution on the vinylarene
resulted in a moderate decrease of the enamine selectivity,
typically from 96% to 90%, especially at long reaction times.
Other secondary amines are also suitable for the oxidative
amination of styrene (Table 3, entries 7–10). Morpholine,
a cyclic amine less basic than piperidine (piperidine, pKb =2.88;
morpholine, pKb =5.67) showed a similar activity and complete
selectivity to the enamine product. However, pyrrolidine with
a similar basicity (pKb =2.74)[32] was much less active. Unex-
pectedly, the catalytic system with N-methylpiperazine showed
no activity, probably because of a catalyst deactivation. In con-
trast, the reaction of [Rh{Ph2P(CH2)3OEt}2]+ (5) with 1 equiv. of
N-methylpiperazine afforded the compound [Rh{Ph2P-
(CH2)3OEt}2(N2C5H12)]+ (13), which was isolated as a yellow
solid in 78% yield (Scheme 8). The 31P{1H} NMR spectrum of 13
featured two doublet of doublets at d=41.61 (JPÀRh =228.1 Hz)
Conclusions
The mechanism of the anti-Markovnikov oxidative amination
of styrene with piperidine catalysed by cationic rhodium(I)
complexes containing P,O-functionalised arylphosphine ligands
has been investigated by spectroscopic means under stoichio-
metric and catalytic conditions. Reactivity studies on the cata-
lyst precursor [Rh{k2-P,O-Ph2P(CH2)3OEt}2]+ (5) indicated that
the reaction proceeds by an amine activation mechanism
through the piperidine intermediate species [Rh{k1-P-Ph2P-
(CH2)3OEt}2(HNC5H10)]2[PF6] (8). which in the presence of piperi-
dine and under thermal conditions is transformed into a neutral
amido species with formation of the piperidonium cation
[H2NC5H10]+. The identity of this neutral amido–piperidine spe-
cies as [Rh{k1-P-Ph2P(CH2)3OEt}2(NC5H10)(HNC5H10)] (9) has been
established by DFT calculations. Optimization of the structure
of 9, with an endo disposition of the NH proton relative to the
amido ligand resulted in a net stabilization of À12.11 kcal
molÀ1 relative to the energy of the structure of [Rh{Ph2P-
(CH2)3OEt}2(NC5H10)] with one of the hemilabile phosphine li-
gands k2-P,O-coordinated and the other k1-P-coordinated, as
a consequence of the NH···N hydrogen bond. NMR studies per-
formed in the presence of styrene under catalytic conditions
showed that 9 is a key species in the catalytic oxidative amina-
tion of styrene. Related cyclooctadiene-containing catalyst pre-
cursors [Rh(cod){k1-P-Ph2P(CH2)3OEt}n]+ (n=1, 2) also gave 9
under the same conditions, suggesting that the oxidative ami-
nation of the diene ligand in these precursors could be a re-
quirement for the generation of the active catalytic species,
and different catalytic activity is probably related to the con-
centration of this species.
The catalytic cycle has been thoroughly studied by a series
of DFT calculations disclosing the most plausible intermediates.
In addition, the transition states of the key steps have been
found and characterized. The first step of the proposed mecha-
nism is the replacement of the piperidine ligand in 9 by sty-
rene to give the catalytic active species [Rh{k1-P-Ph2P-
(CH2)3OEt}2(NC5H10)(styrene)]. The catalytic cycle takes place at
two stages. The first stage leads to the oxidative amination of
styrene and is followed a concomitant styrene hydrogenation
stage, which closes the cycle. The most energetically demand-
ing step of the styrene amination stage is the amido insertion
into the coordinated styrene ligand. This step is followed by
a facile b elimination process, which gives the styrylpyperidine
reaction product after coordination of piperidine or styrene to
the resulting hydrido complex. Thus, two alternate paths for
the second stage of the catalytic cycle are possible. The coordi-
nation of piperidine should be followed by an oxidative addi-
tion step of the NH group to the rhodium atom. However, a rel-
evant aspect of the whole catalytic cycle is that this NH oxida-
tive addition step is more energetically demanding than the al-
ternate, operating, mechanism which regenerates the catalyti-
cally active amido species through a direct proton transfer
from a coordinated piperidine ligand to an alkyl group that is
formed after styrene coordination to the hydrido complex.
The optimised structures of the key intermediates in the cat-
alytic cycle revealed that the relative cis arrangement of the
Scheme 8. Reaction of 5 with N-methylpiperazine to afford stable com-
pound 13.
and 40.12 ppm (JPÀRh =220.6 Hz), and a JPÀP of 54.3 Hz. The
chelating coordination of the N-methylpiperazine ligand,
which drives a change of the coordination mode of the phos-
phine ligands to k1-P,O, is responsible for the inequivalence of
the phosphine ligands. Therefore, the high stability of 13 pre-
cludes the formation of any active amido complexes and con-
sequently the compound is inactive.
The catalytic systems based on acyclic secondary amines
were much less active in the oxidative amination of styrene.
However, conversions higher than 85% could be typically
reached in 6–8 h for Et2NH and nBu2NH (entries 9 and 10,
Table 3). Interestingly, a complete selectivity to the enamine
product was observed in spite of the long reaction times re-
quired to get acceptable conversions. The catalytic system
based on iPr2NH showed no activity probably as a consequence
of the steric effects.[33]
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 273

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
In situ preparation of [Rh{k1-P-Ph2P(CH2)3OEt}2(NC5H10)(HNC5H10)]
(9): [Rh{k2-PO-Ph2P(CH2)3OEt}2][PF6] (5) (0.020 mmol) reacted with
piperidine (0.060 mmol) in [D8]THF (0.5 mL, in an NMR tube). The
solution that contained a mixture of 7 and 8 was heated at 343 K
for 2 h. 1H NMR evidenced the formation of 9 and [H2NC5H10]+.
Data for 9: 1H NMR (298 K, [D8]THF): d=7.53–7.35 (m, 20H, Ph),
functionalised phosphine ligands allows for a p–p intramolecu-
lar interaction between two phenyl groups. This interaction
provides a rigid scaffold that directs both hemilabile arms to-
wards the same side of the molecular plane, facilitating the
protection of the accessible coordination sites at the rhodium
centre. As a consequence, the observed hemilabile effect could
be related with the stabilization of polar species in solution or
unsaturated catalytic intermediates by transient O-
coordination.
3.87 (brs, 1H, CH2 amido), 3.59 (brs, 2H, CH2O), 3.42 (q, 2H, JHÀH
=
7.2 Hz, CH2CH3), 3.36 (q, 2H, JHÀH =7.0 Hz, CH2CH3), 3.25 (brs, 2H,
CH2O), 3.09 (brs, 1H, CH2 amido), 3.01 (brs, 1H, CH2 amido), 2.85
(s, CH2), 2.44, 2.43, 2.28, 1.98, 1.97 (brs, 1H each, CH2 amido), 1.81
(brs, 2H, CH2P), 1.78 (brs, 2H, CH2), 1.57 (s, CH2), 1.55 (brs, 1H, CH2
amido), 1.40 (brs, 1H, CH2 amido), 1.34 (brs, 2H, CH2P), 1.27 (brs,
2H, CH2), 1.12 (t, 3H, JHÀH =7.0 Hz, CH2CH3), 1.10 ppm (t, 3H, JHÀH
7.5 Hz, CH2CH3). 31P{1H} NMR (298 K, [D8]THF): d=39.60 (dd, JPÀRh
165.0 Hz, JPÀP =51.3 Hz), 36.31 ppm (dd, JPÀRh =173.1 Hz, JPÀP
=
=
=
Experimental Section
General
51.3 Hz). 13C{1H} NMR (298 K, [D8]THF): d=131.28–127.89 (m, Ph),
70.52 (d, JCÀP =13.4 Hz, CH2O), 70.38 (d, JCÀP =13.6 Hz, CH2O), 69.94
(d, CH2CH3), 65.61 (CH2CH3), 53.28, 49.40 (CH2 amido), 29.53 (d,
Reactions were performed under exclusion of air by using standard
Schlenk techniques. Solvents were dried by known procedures and
distilled under argon just prior to use, or obtained from a solvent
purification system (Innovative Technologies). CDCl3, CD2Cl2,
[D8]THF (Euriso-top) were dried using activated molecular sieves.
The functionalised hemilabile phosphine, Ph2P(CH2)3OEt,[17] and cat-
alyst precursors 1–6[15] were prepared following published meth-
ods (see the Supporting Information).
J
CÀP =25.9 Hz, CH2P), 29.40 (CH2 amido), 27.42 (d, JCÀP =20.9 Hz,
CH2P), 26.11 (CH2), 25.92 (CH2 amido), 24.49 (CH2), 23.86 (CH2
amido), 22.22 (d, JCÀP =14.5 Hz, CH2), 22.20 (d, JCÀP =17.3 Hz, CH2),
1
14.65 (CH2CH3), 14.59 ppm (CH2CH3). Data for [H2NC5H10]+: H NMR
(298 K, [D8]THF): 4.38 (br, 2H, NH2+), 2.85 (s, CH2), 1.57 ppm (s,
CH2). 13C{1H} NMR (298 K, [D8]THF): d=46.61, 26.11, 24.49 ppm
(CH2).
Reaction of [Rh{k2-P,O-Ph2P(CH2)3OEt}2][PF6] (5) with
piperidine
Reaction of [Rh(cod){k2-P,O-Ph2P(CH2)3OEt}][BF4] (1) with
piperidine
Formation
of
[Rh{k1-P-Ph2P(CH2)3OEt}{k2-P,O-Ph2P(CH2)3OEt}-
Piperidine (2 mL, 0.02 mmol) was added at RT to a solution of
1 (11.40 mg, 0.02 mmol) in [D8]THF (0.5 mL, NMR tube). Spectro-
scopic NMR analysis showed the formation of an equimolar mix-
ture of [Rh(cod){k1-P-Ph2P(CH2)3OEt}2][BF4] (3) and [Rh(cod)-
(HNC5H10)][PF6] (7): To a solution of [Rh{k2-P,O-Ph2P(CH2)3OEt}2][PF6]
(5, 16.0 mg, 0.02 mmol) in [D8]THF (0.5 mL, in an NMR tube) at RT,
piperidine (2.00 mL, 0.02 mmol) was added to give an orange solu-
1
1
tion of 7. H NMR (253 K, [D8]THF): d=8.35–6.59 (m, 20H, Ph), 4.42,
(HNC5H10)2][BF4] (10). Data for 10: H NMR (298 K, [D8]THF): d=4.03
4.07 (brs, 2H, CH2O), 3.46 (q, 4H, JHÀH =6.9 Hz, CH2CH3), 3.43 (q,
4H, JHÀH =6.8 Hz, CH2CH3), 3.31 (brs, 1H, CH2O), 3.31–1.99 (m, 8H,
4H CH2 pip+4H CH2), 1.76–1.17 (m, 4H, CH2 pip, 4H CH2), 1.19 (t,
6H, JHÀH =7.2 Hz, CH2CH3), 1.14 (t, 6H, JHÀH =6.9 Hz, CH2CH3), 0.74
(brs, 1H, CH2 pip), 0.38 ppm (br, 1H, CH2 pip). 31P{1H} NMR (253 K,
[D8]THF): d=50.24 (dd, JPÀRh =215.0 Hz, JPÀP =54.4 Hz), 30.23 ppm
(dd, JPÀRh =161.0 Hz, JPÀP =54.5 Hz). 13C{1H} NMR (253 K, [D8]THF):
d=137.34 (d, JCÀP =47.0 Hz, Cipso), 135.33 (d, JCÀP =46.9 Hz, Cipso),
132.23–128.28 (Co, Cp, Cm), 70.72 (CH2CH3), 69.83 (CH2O), 68.61
(CH2O), 65.94 (CH2CH3), 48.87, 48.56 (CH2 pip), 29.15 (CH2), 27.61
(CH2 pip), 26.90 (CH2), 24.04, 24.19 (CH2 pip), 22.26, 20.28 (CH2),
14.90, 12.69 ppm (CH2CH3). MS/ESI+ (MeOH, m/z): 731 ([M]+).
(br, 4H, =CH cod), 2.80 (brs, 4H, CH2 pip), 2.41–2.33 (m, 8H, CH2
cod), 1.79–1.56 ppm (br m, 6H, CH2 pip). 13C{1H} NMR (298 K,
[D8]THF): d=81.19 (br d, JCÀRh =13.2 Hz, =CH cod), 48.99, 29.86
(CH2 pip), 27.86 (CH2 cod), 25.84 ppm (CH2 pip).
Reaction of [Rh(cod){k1-P-Ph2P(CH2)3OEt}2][BF4] (3) with
piperidine
In situ preparation of [Rh(cod){NC5H10}{Ph2P(CH2)3OEt}2][BF4] (11): Pi-
peridine (70 mL, 0.712 mmol) was added to a solution of 3 (15 mg,
0.018 mmol) in [D8]THF (0.5 mL, NMR tube). The solution was
heated at 333 K for 20 min., or alternatively left at 298 K for
24 h.1H NMR (298 K, [D8]THF): d=7.33–7.05 (m, 20H, Ph), 4.93 (br,
4H, =CH cod), 3.18 (q, 4H, JHÀH =7.0 Hz, CH2CH3), 3.08 (t, 4H,
Reaction of 5 with piperidine in excess; formation of [Rh{k1-P-Ph2P-
(CH2)3OEt}2(HNC5H10)2][PF6] (8): To a solution of [Rh{k2-PO-Ph2P-
(CH2)3OEt}2][PF6] (5), 16.0 mg, 0.02 mmol) in [D8]THF (0.5 mL, in
NMR tube) at RT, an excess of piperidine (8.00 mL, 0.08 mmol) was
added. The colour of the solution immediately changed to orange
J
HÀH =6.3 Hz, CH2O), 2.85 (brs, 2H, CH2), 2.77 (m, 4H, CH2 cod), 1.89
(brs, 4H, CH2 cod), 1.57 (brs, 2H, CH2), 0.97 ppm (t, 6H, JHÀH
7.0 Hz, CH2CH3). 31P{1H} NMR (298 K, [D8]THF): d=30.62 ppm (d,
PÀRh =199.5 Hz). 13C{1H} NMR (298 K, [D8]THF): d=130.52 (t, J=
=
1
because of the formation of 8. H NMR (193 K, [D8]THF): d=7.89–
J
7.16 (m, 20H, Ph), 3.37 (q, 4H, JHÀH =6.9 Hz, CH2CH3), 3.34 (brs, 4H,
CH2O), 3.14 (m, 2H, CH2 pip), 2.91 (t, 2H, JHÀH =11.0 Hz, CH2 pip),
2.84 (m, 2H, CH2 pip), 2.49 (t, 2H, JHÀH =11.0 Hz, CH2 pip), 2.04 (m,
2H, CH2 pip), 1.92 (brs, 4H, CH2), 1.69 (brs, 2H, CH2 pip), 1.50–1.32
5.8 Hz, Co), 125.91 (Cp), 125.30 (t, J=4.2 Hz, Cm), Cipso (not observed),
100.20 (m, =CH cod), 69.32 (t, J=7.6 Hz, CH2O), 63.72 (CH2CH3),
48.01 (CH2 pip), 30.94 (CH2 cod), 28.00 (CH2), 27.71 (CH2 pip), 24.22
(CH2 pip), 26.81 (t, J=13.0 Hz, CH2P), 12.85 ppm (CH2CH3).
(m, 8H, 4H CH2 pip+2H CH2 pip+4H CH2), 1.16 (6H, t, JHÀH
=
7.0 Hz, CH2CH3), 0.42 ppm (CH2 pip). 31P{1H} NMR (193 K, [D8]THF):
d=38.29 ppm (d, JPÀRh =170.2 Hz). 13C{1H} NMR (193 K, [D8]THF):
d=136.52 (d, JCÀP =44.3 Hz, Cipso), 132.69–129.00 (m, Ph), 70.62
(CH2O), 66.04 (CH2CH3), 50.26 (CH2 pip), 46.71 (CH2 pip), 27.95 (CH2
pip), 27.49 (CH2P), 26.30 (CH2 pip), 25.01 (CH2 pip), 23.76 (CH2),
15.13 ppm (CH2CH3).
Synthesis of [Rh{k1-P-Ph2P(CH2)3OEt}2(N2C5H12)][PF6] (13)
N-methylpiperazine (8.40 mL, 0.088 mmol) was added to a solution
of [Rh{k2-PO-Ph2P(CH2)3OEt}2][PF6] (5) (60.0 mg, 0.088 mmol) in THF
(4 mL) at RT and the solution stirred for 1 h. The volatiles were re-
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 274

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
moved under reduced pressure and the residue disaggregated by
washing with cold n-hexane (3 times 4 mL) at 08C to give a yellow
solid. Yield: 78%; elemental analysis of C39H54N2F6O2P3Rh calc. (%):
C, 52.47; H, 6.10; N, 3.14; found: C, 52.33; H, 6.25; N, 3.10. Fast
gratefully acknowledged. M.I.B. and D.G. thank the Spanish
MICINN and the IUCH (Instituto Universitario de Catꢃlisis Homo-
gꢀnea) for a predoctoral fellowship.
1
atom bombardment (FAB)+–MS (m/z,%): 892 ([M]+, 100). H NMR
Keywords: density functional calculations
catalysis · amination · reaction mechanisms · rhodium
· homogeneous
(253 K, [D8]THF): d=7.84–7.31 (m, 20H, Ph), 3.55 (brs, 1H, CH2À
amine), 3.52 (brs, 1H, CH2Àamine), 3.38 (m, 9H, 4H CH2O+4H
CH2CH3 +1H CH2Àamine), 3.23 (brs, 1H, NH), 2.71 (brs, 1H, CH2À
amine), 2.42 (brs, 2H, CH2P), 2.36 (brs, 1H, CH2Àamine), 2.16 (brs,
2H, CH2P), 2.06 (brs, 1H, CH2Àamine), 2.00 (brs, 1H, CH2Àamine),
1.93 (brs, 2H, CH2), 1.81 (brs, 2H, CH2), 1.73 (brs, 1H, CH2Àamine),
1.70 (s, 3H, CH3), 1.12 ppm (m, 6H, CH2CH3); 31P{1H} NMR (253 K,
[D8]THF): d=41.61 (dd, JPÀRh =228.1 Hz, JPÀP =54.3 Hz), 40.12 ppm
(dd, JPÀRh =220.6 Hz, JPÀP =54.3 Hz); 13C{1H} NMR (253 K, [D8]THF):
d=138.25 (d, JCÀP =40.6 Hz, Cipso), 136.98 (d, JCÀP =40.6 Hz, Cipso),
133.67–129.30 (Ph), 71.65 (dd, JCÀP =16.3 Hz, JCÀRh =6.3 Hz, CH2O),
71.22 (d, JCÀP =15.1, CH2O), 66.96, 66.63 (CH2CH3), 57.00 (NCH2),
48.61 (NCH3), 46.37, 33.15 (NCH2), 29.09 (d, JCÀP =4.7 Hz, CH2), 27.91
(NCH2), 27.35 (d, JCÀP =34.7 Hz, CH2P), 26.13 (d, JCÀP =29.9 Hz,
CH2P), 25.97, 24.08 (CH2CH3), 23.23 ppm (d, JCÀP =3.4 Hz, CH2).
[1] a) J.-S. Ryu, G. Y. Li, T. J. Marks, J. Am. Chem. Soc. 2003, 125, 12584–
12605; b) T. E. Müller, K. C. Hultzsch, M. Yus, F. Foubelo, M. Tada, Chem.
Rev. 2008, 108, 3795–3892; c) R. Taube in Applied Homogeneous Cataly-
sis with Organometallic Compounds (Eds.: B. Cornils, W. A. Herrmann),
Wiley-VCH, Weinheim, 1996, pp. 513–524.
[2] a) A. L. Reznichenko, H. N. Nguyen, K. C. Hultszch, Angew. Chem. 2010,
122, 9168–9171; Angew. Chem. Int. Ed. 2010, 49, 8984–8987; b) K. C.
Hultzsch, Adv. Synth. Catal. 2005, 347, 367–391; c) J. Y. Kim, T. Living-
house, Org. Lett. 2005, 7, 1737–1739; d) S. Hong, T. J. Marks, Acc. Chem.
Res. 2004, 37, 673–686; e) G. A. Molander, J. A. C. Romero, Chem. Rev.
2002, 102, 2161–2185.
[3] a) N. Hazari, P. Mountford, Acc. Chem. Res. 2005, 38, 839–849; b) F.
Pohlki, S. Doye, Chem. Soc. Rev. 2003, 32, 104–114.
[4] a) A. Motta, I. L. Fragalµ, T. J. Marks, Organometallics 2006, 25, 5533–
5539; b) S. Tobisch, Chem. Eur. J. 2006, 12, 2520–2531; c) M. R. Gagnꢀ,
T. J. Marks, J. Am. Chem. Soc. 1989, 111, 4108–4110.
[5] a) J. A. Bexrud, J. D. Bear, D. C. Leitch, L. L. Schafer, Org. Lett. 2005, 7,
1959–1962; b) B. D. Ward, A. Maisse-FranÅois, L. H. Gade, P. Mountford,
Chem. Commun. 2004, 704–705; c) Z. K. Sweeney, J. L. Salsman, R. A.
Andersen, R. G. Bergman, Angew. Chem. 2000, 112, 2429–2433; Angew.
Chem. Int. Ed. 2000, 39, 2339–2343.
[6] a) J. Zhang, C. G. Yang, C. He, J. Am. Chem. Soc. 2006, 128, 1798–1799;
b) D. Karshtedt, A. T. Bell, T. D. Tilley, J. Am. Chem. Soc. 2005, 127,
12640–12646; c) J. M. Seul, S. Park, J. Chem. Soc. Dalton Trans. 2002,
1153–1158; d) M. Beller, C. Breindl, M. Eichberger, C. G. Hartung, J.
Seayad, O. R. Thiel, A. Tillack, H. Trauthwein, Synlett 2002, 1579–1594.
[7] a) Z. Liu, H. Yamamichi, S. T. Madrahimov, J. F. Hartwig, J. Am. Chem. Soc.
2011, 133, 2772–2782; b) K. D. Hesp, S. Tobisch, M. Stradiotto, J. Am.
Chem. Soc. 2010, 132, 413–426; c) X. Li, A. R. Chianese, T. Vogel, R. H.
Crabtree, Org. Lett. 2005, 7, 5437–5440; d) T. E. Müller, M. Berger, M.
Grosche, E. Herdtweck, F. P. Schmidtchen, Organometallics 2001, 20,
4384–4393.
[8] I. Mena, M. A. Casado, P. García-OrduÇa, V. Polo, F. J. Lahoz, A. Fazal,
L. A. Oro, Angew. Chem. 2011, 123, 11939–11942; Angew. Chem. Int. Ed.
2011, 50, 11735–11738.
[9] a) E. Morgan, D. F. MacLean, R. McDonald, L. Turculet, J. Am. Chem. Soc.
2009, 131, 14234–14236; b) J. Zhao, A. S. Goldman, J. F. Hartwig, Sci-
ence 2005, 307, 1080–1082; c) M. S. Driver, J. F. Hartwig, J. Am. Chem.
Soc. 1996, 118, 4206–4207; d) M. D. Fryzuk, C. D. Montgomery, Coord.
Chem. Rev. 1989, 95, 1–40; e) A. L. Casalnuovo, J. C. Calabrese, D. Mil-
stein, Inorg. Chem. 1987, 26, 971–973.
Catalytic hydroamination experiments
General procedure: The catalytic hydroamination reactions were
performed under an argon atmosphere in a thick glass reaction
tube fitted with a greaseless high-vacuum stopcock. In a typical ex-
periment, the reactor was charged with a solution of the catalyst
(0.020 mmol) in THF (2 cm3), 2 mg of molecular sieves in powder
(4 Š) and the reactants in the following order: piperidine
(0.800 mmol, 79 mL), tetradecane as internal standard (0.350 mmol,
91 mL) and styrene (3.24 mmol, 371 mL). The mixture was stirred at
RT until the catalyst was completely dissolved, and then placed in
a thermostatised oil bath at the required temperature.
The yield and selectivity were determined by using GC (HP 6890N
with an ionization detector fitted to a HP Ultra-1 column (25 m”
0.32 mm”internal diameter 0.17 mm) under the following condi-
tions: Initial T=508C for 4 min, ramp 158minÀ1, and final T=2508C
for 10 min. Calibration was made with the internal standard tetra-
decane. Organic compounds were identified by GC–MS recorded
in the mass range 1–1000 m/z on an Agilent 6890 GC-Agilent 5973
MS, equipped with a polar capillary column HP-5 MS (30 m”
0.25 mm”internal diameter 0.25 mm).
Theoretical calculations
[10] a) J. W. Tye, J. F. Hartwig, J. Am. Chem. Soc. 2009, 131, 14703–14712;
b) P. Zhao, C. Krug, J. F. Hartwig, J. Am. Chem. Soc. 2005, 127, 12066–
12073.
[11] R. L. Cowan, W. C. Trogler, J. Am. Chem. Soc. 1989, 111, 4750–4761.
[12] a) D. S. Glueck, L. J. Newman-Winslow, R. G. Bergman, Organometallics
1991, 10, 1462–1479; b) M. S. Driver, J. F. Hartwig, J. Am. Chem. Soc.
1997, 119, 8232–8245.
[13] J. J. Brunet, D. Neibecker, K. Philippot, Tetrahedron Lett. 1993, 34, 3877–
3880.
[14] M. V. Jimꢀnez, J. J. Pꢀrez-Torrente, M. I. Bartolomꢀ, F. J. Lahoz, L. A. Oro,
Chem. Commun. 2010, 46, 5322–5324.
All computations were performed by using the Gaussian 09 (Revi-
sion A.02) package.[28] The structures were fully optimised without
geometrical constraints and the stationary points (minima and TS)
were confirmed by frequency calculations. Connections between
the transition states and the minima were checked by visual in-
spection of the negative frequency calculation and IRC in some
cases. Calculations were performed by using the B3LYP functional
and the basis sets used were: LANL2DZ and its associated ECP for
rhodium and 6-31G** for the rest of atoms. The structures have
been depicted using CYLview.[34]
[15] M. V. Jimꢀnez, M. I. Bartolomꢀ, J. J. Pꢀrez-Torrente, F. J. Lahoz, L. A. Oro,
ChemCatChem 2012, 4, 1298–1310.
[16] A. Uhe, M. Hçlscher, W. Leitner, Chem. Eur. J. 2010, 16, 9203–9214.
[17] M. V. Jimꢀnez, J. J. Pꢀrez-Torrente, M. I. Bartolomꢀ, L. A. Oro, Synthesis
2009, 1916–1922.
[18] TOF50 =Turnover frequency estimated at 50% of piperidine conversion.
[19] D. A. Buckingham, C. R. Clark, M. G. Weller, G. J. Gainsford, Chem.
Commun. 1982, 779–781.
Acknowledgements
Financial support from the Ministerio de Economꢂa y Competitivi-
dad (MINECO/FEDER), projects CTQ2010-15221 and MULTICAT
(CSD2009-00050), and Gobierno de Aragꢁn/FSE (Group E07) is
[20] J. J. Brunet, C. Commenges, D. Neibecker, K. Philippot, L. Rosenberg,
Inorg. Chem. 1994, 33, 6373–6379.
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 275

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
[21] M. Beller, M. Eichberger, H. Trauthwein, Angew. Chem. 1997, 109, 2306–
2308; Angew. Chem. Int. Ed. Engl. 1997, 36, 2225–2227.
[22] a) P. S. Hanley, J. F. Hartwig, J. Am. Chem. Soc. 2011, 133, 15661–15673;
b) P. S. Hanley, D. Markovic, J. F. Hartwig, J. Am. Chem. Soc. 2010, 132,
6302–6303.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[23] P. B. White, S. S. Stahl, J. Am. Chem. Soc. 2011, 133, 18594–18597.
[24] K. D. Hesp, M. Stradiotto, ChemCatChem 2010, 2, 1192–1207.
[25] a) J. M. Brown, G. C. Lloyd-Jones, J. Am. Chem. Soc. 1994, 116, 866–878;
b) J. M. Brown, G. C. Lloyd-Jones, J. Chem. Soc. Chem. Commun. 1992,
710–712.
[29] X. Yuan, S. Bi, Y. Ding, L. Liu, M. Sun, J. Organomet. Chem. 2010, 695,
1576–1582.
[30] K. H. Hopmann, J. Conradie, Organometallics 2009, 28, 3710–3715.
[31] K. H. Hopmann, N. F. Stuurman, A. Muller, J. Conradie, Organometallics
2010, 29, 2446–2458.
[26] M. Beller, H. Trauthwein, M. Eichberger, C. Breindl, J. Herwig, T. E. Müller,
O. R. Thiel, Chem. Eur. J. 1999, 5, 1306–1319.
[32] S. P. Anthony, S. Varughese, S. M. Draper, Chem. Commun. 2009, 7500–
7502.
[27] M. Utsunomiya, R. Kuwano, M. Kawatsura, J. F. Hartwig, J. Am. Chem.
Soc. 2003, 125, 5608–5609.
[33] nBu2NH, pKa (benzene)=10.66; iPr2NH, pKa (benzene)=11.13; iEt2NH,
pKa (benzene)=10.97; a) J. A. Tayh, R. M. Scott, J. Mol. Struct. 1990, 237,
297–305; b) M. Abduljaber, R. M. Scott, J. Mol. Struct. 1990, 237, 285–
295; c) V. Frenna, N. Vivona, G. Consiglio, D. Spinelli, J. Chem. Soc. Perkin
Trans. 2 1985, 1865–1868.
[28] Gaussian 09, Revision A.02, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
[34] C. Y. Legault, Universitꢀ de Sherbrooke, 2009 CYLview, 2009.
Received: July 27, 2012
Published online on October 30, 2012
ꢁ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 2013, 5, 263 – 276 276