A C-H activation-CO2-carboxylation reaction sequence mediated by an 'Iridium(dppm)' species. Formation of the anionic ligand (Ph 2P)2C-COOH

Article abstract of DOI:10.1039/b809331h

The reaction of [Ir₂(μ-Cl)₂(coe)₄] with 1,1-bisdiphenylphosphinomethane under CO₂ atmosphere affords the complex [IrCl(dppm)(H){(Ph₂P)₂C-COOH}]₂ by initial CH activation followed by formal insertion of CO₂ into the C-H bond of the formed diphosphanylmethanide ligand. The Royal Society of Chemistry.

Full text of DOI:10.1039/b809331h

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A C–H activation–CO2-carboxylation reaction sequence mediated
by an ‘Iridium(dppm)’ species. Formation of the anionic ligand
(Ph2P)2C–COOHw
Jens Langer,*z Marı
a Jose Fabra, Pilar Garcıa-Orduna, Fernando J. Lahoz and
´ ´
´
Luis A. Oro*
Received (in Cambridge, UK) 3rd June 2008, Accepted 9th July 2008
First published as an Advance Article on the web 21st August 2008
DOI: 10.1039/b809331h
The reaction of [Ir2(l-Cl)2(coe)4] with 1,1-bisdiphenylphosphino-
methane under CO2 atmosphere affords the complex
[IrCl(dppm)(H){(Ph2P)2C–COOH}]2 by initial CH activation
followed by formal insertion of CO2 into the C–H bond of the
formed diphosphanylmethanide ligand.
The activation of small molecules like CO2 and the modifica-
tion of alkanes by C–H activation are among the goals of
modern organometallic and coordination chemistry. Although
in both fields significant progress has been reported during the
last decades,1,2 there are still only a few examples of transition
metal complexes, which are able to selectively functionalize
alkanes by a C–H activation–CO2 carboxylation sequence.1
Initial investigations by Herskovitz showed that iridium com-
pounds can be used to carboxylate acetonitrile,3 but the
products were not isolated. Later Behr et al. demonstrated,
that the use of the more acidic malonodinitrile (pKa = 12)
made possible the isolation of the products and their
complete characterization. They were able to determine the
X-ray structures of [IrCl(depe)2(H)]+[HOOCC(CN)2]ꢀ and
[Ir(H)2(PMe3)4]+[HOOCC(CN)2]ꢀ.4
We herein report the iridium mediated transformation of
the less acidic dppm (1,1-bisdiphenylphosphinomethane,
pKa = 29.95) into an unpredicted anionic ligand via a C–H
activation–carboxylation sequence.
The system dppm–[Ir2(m-Cl)2(coe)4] (coe = cyclooctene)
was chosen as a promising candidate for CO2 activation due
to the fluxional behavior of the resulting product ‘‘dppm2IrCl’’
in solution as reported by Tejel et al.6 The products and
intermediates 1a–1d (see Scheme 1) are species with potential
reaction sites for CO2. The iridium(I) complex [Ir(dppm)2]Cl is
similar to complexes of the composition [IrL4]Cl (L2 = dmpe
Scheme 1 Proposed mechanism for the formation of 2.
or L = PMe3) which react with CO2 either to form metalla-
carboxylates7 or to oxidatively couple two CO2 molecules.8
Interestingly, the reaction of [Ir2(m-Cl)2(coe)4] with dppm
(1 : 4 ratio) under a CO2 atmosphere in various solvents (for
instance toluene, thf, acetone or acetonitrile) leads to the
formation of a white product 2 within minutes. IR measure-
ments show strong bands at 1580 and 1270 cmꢀ1 which were
not observed in absence of CO2. Although similar to those
reported for the metallacarboxylate [IrCl(CO2)(dmpe)2] (1550
and 1230 cmꢀ1), the obtained product shows different reacti-
vity.7 For instance, no reaction was observed with methyl
triflate in toluene suspension.
´
´
Departamento de Quımica Inorganica, Instituto Universitario de
Catalisis Homogenea, Instituto de Ciencia de Materiales de Aragon,
´
´
´
Universidad de Zaragoza-Consejo Superior de Investigaciones
´
Cientıficas, 50009-Zaragoza, Spain. E-mail: oro@unizar.es;
j.langer@uni-jena.de; Fax: +34 976 761143; Tel: +34 976 761143
w Electronic supplementary information (ESI) available: Full experi-
mental details and analytical data for 2 and 3. NMR data for
[Li{OOCCH(Ph2P)2}(tmeda)], [Ir(dppm)2]PF6 and [Ir(dcypm)2]Cl.
CCDC reference numbers 689715 (1d) and 689716 (2). For ESI and
crystallographic data in CIF or other electronic format see DOI:
10.1039/b809331h
An X-ray diffraction studyy of crystals of 2, obtained from
acetone, shows that the unexpected carboxylation of a dppm
z Current address: Institut fur Anorganische und Analytische Chemie,
FSU Jena, August-Bebel-Straße 2, 07743 Jena, Germany.
ꢁc
This journal is The Royal Society of Chemistry 2008
4822 | Chem. Commun., 2008, 4822–4824

View Article Online
a slightly distorted octahedral environment (Fig. 1). Com-
pared to 1d (see Fig. 2) the coordination sphere of the iridium
atom has not changed too much and similar distances and
angles around the metal were observed. In the formed anionic
(Ph2P)2C–COOH ligand the carbon atoms, C(1) and C(2),
clearly exhibit sp2 hybridizations (bond angles around C(1)
and C(2) summing up 359.27 and 360.01) and the P–C(1)
bonds (average 1.764(3) A) are significantly shorter than in the
standard coordinated dppm (av. 1.847(3) A), consistent with a
P–C partially multiple bond character and with the delocaliza-
tion of the negative charge in the P(1)–C(1)–P(2) fragment.
The carboxylic acid group is nearly coplanar with the
P(1)–C(1)–P(2) fragment forming a dihedral angle of 8.14(10)1,
very similar to that observed in the neutral carbodiphosphorane
CO2 adduct (Ph3P)2C–CO2 (10.01).9 Like other carboxylic acids,
2 forms dimers in the solid state through a double hydrogen bond
(O(1)–O(2)0 2.600(4) A), which makes both C(2)–O bond
distances identical. In contrast, the corresponding lithium salt
was described as a [bis(diphenylphosphino)acetate],10 but neither
infrared nor NMR data were reported. Therefore, the related
base adduct with tmeda (N,N,N0N0-tetramethylethylendiamine)
was prepared and NMR measurements support the description
of the ligand as a [bis(diphenylphosphino)acetate] (see supporting
information).
Fig. 1 Molecular structure of 2. (Most of H atoms have been omitted
for clarity). Selected distances (A) and angles (1) are: Ir–P(1) 2.3463(9),
Ir–P(2) 2.3340(9), Ir–P(3) 2.3206(9), Ir–P(4) 2.3405(9), Ir–Cl 2.4806(9),
P(1)–C(1) 1.763(4), P(2)–C(1) 1.764(4), P(1)/P(2)–C(Ph) 1.822(2),
P(3)–C(3) 1.840(4), P(4)–C(3) 1.853(4), P(3)/P(4)–C(Ph) 1.819(2),
C(1)–C(2) 1.421(5), C(2)–O(1) 1.296(4), C(2)–O(2) 1.295(4);
P(1)–Ir–P(3) 172.25(3), P(2)–Ir–P(4) 177.22(3), Cl–Ir–H(0) 175.5(15),
P(1)–C(1)–P(2) 100.17(18), P(1)–C(1)–C(2) 129.1(3), P(2)–C(1)–C(2)
130.0(3), P(3)–C(3)–P(4) 95.65(17).
We propose the mechanism displayed in Scheme 1 to
explain the formation of complex 2. Initially, a C–H bond
activation in the proposed square-planar complex 1c afforded
1d. Afterwards, a new C–C bond is formed by nucleophilic
attack of the diphosphanylmethanide carbon to the carbon of
CO2, followed by proton migration. A similar insertion of
alkynes into the C–H bond of a diphosphanylmethanide
ligand in fac-[Mn(CNtBu)(CO)3{(PPh2)2C–H}] was reported
by Ruiz et al.11 The structure of 1d, determined by X-ray
diffraction is shown in Fig. 2.
ligand has taken place. Compound 2 (Scheme 1) crystallizes in
ꢀ
the space group P1 and it shows the iridium centre surrounded
by four phosphorus atoms, a hydride and a chloride ligand in
Like in related diphosphanylmethanide complexes,12 the
formed four membered ring Ir,P(1),C(1),P(2) is planar. The
distances Ir–P(1) (2.3534(8) A) and Ir–P(2) (2.3390(8) A) are sligh-
tly longer than reported for [Ir(dppm-H)(H)(tp)] (tp = trispyr-
azolylborate) (2.276 and 2.269 A),13 but very similar to those
observed in 2 (see Fig. 1). The major difference between 1d and 2
concerns the P–C(1) bond distances in the methanide moieties,
being remarkably longer in 2 (1.764(3) A) than those observed in
1d (1.725(3) A), reflecting the different electronic properties of C(1)
due to the presence of the carboxylic acid function in 2.
The formation of 1d as intermediate can be verified by blocking
the C–H activation reaction. This has been done by exchanging
the anion in the starting material, since a coordinating anion is
necessary for the C–H activation to proceed via 1c or by
increasing the steric crowding around the iridium centre to
prevent the attack of the chloride anion to form such an inter-
mediate. Using a similar synthetic pathway to that used for the
preparation of 1a, complexes [Ir(dppm)2]PF6 and [Ir(dcypm)2]Cl
(dcypm = 1,1-bisdicyclohexylphosphinomethane) were generated
in situ and found to be stable in solution. C–H activation was not
observed. Consequently, these compounds did not react with CO2
under the applied conditions.
Fig. 2 Molecular structure of 1d. Selected distances (A) and angles (1)
are: Ir–P(1) 2.3534(8), Ir–P(2) 2.3390(8), Ir–P(3) 2.3258(8), Ir–P(4)
2.3277(8), Ir–Cl 2.4945(8), P(1)–C(1) 1.726(3), P(2)–C(1) 1.723(3),
P(1)/P(2)–C(Ph) 1.832(2), P(3)–C(3) 1.843(3), P(4)–C(3) 1.847(3),
P(3)/P(4)–C(Ph) 1.819(2); P(1)–Ir–P(3) 173.62(3), P(2)–Ir–P(4)
176.24(3), Cl–Ir–H(0) 176.1(17), P(1)–C(1)–P(2) 100.47(18).
It is interesting to note, that the newly formed ligand does
not act as a P,P,O scorpionate ligand Instead the remaining
hydrogen of the former CH2 group of dppm is transferred to
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4822–4824 | 4823

View Article Online
Notes and references
y X-Ray diffraction data for 1d and 2 were recorded at 100(2) K on a
Bruker-AXS Smart CCD diffractometer, using a o scan technique
with Mo-Ka radiation (l = 0.71073 A). Both structures were solved
by direct methods (SHELXS-97) and refined on F2 by full-matrix least-
squares techniques using the SHELXL-97 program. Non-hydrogen
atoms were refined with anisotropic displacement parameters, and
most hydrogen atoms were refined from observed positions. The
hydride ligands were included from electrostatic potential calculations
(HYDEX program).
Scheme 2
Crystallographic data for complex 1d: C50H44ClIrP4ꢂ1.5C7H8,
ꢀ
M = 1134.58, triclinic, space group P1, a = 10.6704(7), b = 11.7099(8),
c = 20.3601(14) A, a = 85.152(1), b = 76.129(1), g = 88.756(1)1,
V = 2461.0(3) A3, Z = 2, m = 2.938 mmꢀ1, Dcalc = 1.531 Mg mꢀ3
F(000) = 1146, 30 715 reflections collected, 11 798 unique (Rint =
,
0.0309), R1 = 0.0328 (F2 4 2s(F2)), wR2 = 0.0788 (all data) and
GOF = 1.105. CCDC 689715.
Complex 2: C51H44ClIrO2P4ꢂ3C3H6O, M = 1214.63, triclinic, space
ꢀ
group P1, a = 12.3699(12), b = 13.1164(13), c = 17.4537(17) A, a =
83.760(2), b = 79.883(2), g = 79.480(2)1, V = 2732.3(5) A3, Z = 2, m =
2.658 mmꢀ1, Dcalc = 1.476 Mg mꢀ3, F(000) = 1232, 32 782 reflections
collected, 11 929 unique (Rint = 0.0437), R1 = 0.0348 (F2 4 2s(F2)), wR2
= 0.0744 (all data) and GOF = 1.036. CCDC 689716.
Scheme 3 Acid induced decarboxylation of 2.
the carboxylate function and the anionic PCP fragment is
reformed. The similarity of the resulting compound to carbo-
diphosphorane CO2 adducts of the type [(R3P)2C–CO2]9 is
remarkable (see Scheme 2). 2 should be treated as their
metalorganic analogue and is potentially useful as a carbene
precursor like other carbene CO2 adducts.14
1 For reviews on CO2 activation see: M. Aresta and A. Dibenedetto,
Dalton Trans., 2007, 2975–2992; T. Sakakura, J.-C. Choi and H.
Yasuda, Chem. Rev., 2007, 107, 2365–2387; X. Yin and J. R. Moss,
Coord. Chem. Rev., 1999, 181, 27; D. H Gibson, Chem. Rev., 1996,
96, 2063–2095; W. Leitner, Coord. Chem. Rev., 1996, 153, 257; P.
Braunstein, D. Matt and D. Nobel, Chem. Rev., 1988, 88, 747; A.
Behr, Angew. Chem., 1988, 100, 681; D. Walther, Coord. Chem.
Rev., 1987, 79, 135.
2 For reviews on C–H activation see: K. R. Campos, Chem. Soc.
Rev., 2007, 36, 1069–1084; J. A. Labinger and J. E. Bercaw,
Nature, 2002, 417, 507–514; C. Slugovc, I. Padilla-Martinez, S.
Sirol and E. Carmona, Coord. Chem. Rev., 2001, 213, 129–157; G.
Dyker, Angew. Chem., Int. Ed., 1999, 38, 1699–1712; A. E. Shilov
and G. B. Shul’pin, Chem. Rev., 1997, 97, 2879–2932.
3 A. D. English and T. Herskovitz, J. Am. Chem. Soc., 1977, 99,
1648–1649.
4 A. Behr, E. Herdtweck, W. A. Herrmann, W. Keim and W.
Kipshagen, J. Chem. Soc., Chem. Commun., 1986, 1262–1263; A.
Behr, E. Herdtweck, W. A. Herrmann, W. Keim and W. Kipshagen,
Organometallics, 1987, 6, 2307–2313.
5 F. G. Bordwell, W. S. Matthews and N. R. Vanier, J. Am. Chem.
Soc., 1975, 97, 442–443.
6 C. Tejel, M. A. Ciriano, S. Jimenez, L. A. Oro, C. Graiff and A.
Tiripicchio, Organometallics, 2005, 24, 1105–1111.
7 T. Herskovitz, J. Am. Chem. Soc., 1977, 99, 2392–2393; R. L.
Harlow, J. B. Kinney and T. Herskovitz, J. Chem. Soc., Chem.
Commun., 1980, 813–814.
The low solubility of 2 in common organic solvents prevents
its further investigation in solution. Chlorinated solvents
slowly dissolve and react with 2 within days under release of
CO2. While an untraceable mixture of products was formed in
chloroform, the cation [IrCl(dppm)2(H)]+ was the only phos-
phorous containing species observed in dichloromethane. This
species probably results from the attack of in situ formed HCl.
Therefore, the reactivity of 2 towards other acids and bases
was also briefly examined. Reaction with NH4PF6 leads to
[IrCl(dppm)2(H)]PF6 3 (see Scheme 3) with complete loss of
CO2, underlining its similarity to carbodiphosphorane CO2
adducts, which react in a comparable way.15
Attempts to remove the proton of the carboxylic acid
function with neutral bases like triethylamine were unsuccess-
ful even when the base was used in high excess. The use of an
excess of anionic bases like KOtBu was found to be more
promising. Indeed, deprotonation of the acid function was
achieved, as judged by infrared measurement, but it was
complicated by the competing deprotonation of the second
dppm ligand.
8 T. Herskovitz and L. J. Guggenberger, J. Am. Chem. Soc., 1976,
98, 1615–1616.
9 W. Petz, C. Kutschera, M. Heitbaum, G. Frenking, R. Tonner and
B. Neumuller, Inorg. Chem., 2005, 44, 1263–1274.
Other lithium and potassium bases are under current in-
vestigation to provide a clean access to the deprotonated form
of 2, which should be a valuable precursor for heterobimetallic
compounds. Further studies to use 2 as CO2 transfer reagent
to C–H acid compounds like reported for N-heterocyclic
carbene CO2 adducts,16 are also in progress.
10 K. Issleib and H. P. Abicht, J. Prakt. Chem., 1970, 312, 456–465.
11 J. Ruiz, R. Quesada, V. Riera, E. Castellano and O. Piro,
Organometallics, 2004, 23, 175–177.
12 D. E. Chebi, P. E. Fanwick and I. P. Rothwell, Organometallics,
1990, 9, 2948–2952; J. R. Torkelson, O. Oke, J. Muritu, R.
McDonald and M. Cowie, Organometallics, 2000, 19, 854–864.
13 J. S. Wiley and D. M. Heinekey, Inorg. Chem., 2002, 41, 4961–4966.
14 A. M. Voutchkova, L. N. Appelhans, A. R. Chianese and R. H.
Crabtree, J. Am. Chem. Soc., 2005, 127(50), 17624–17625.
15 C. N. Matthews, J. S. Driscoll and G. H. Birum, Chem. Commun.,
1966, 736–737.
This work was funded by MEC (Grants CTQ2006-01629
and CONSOLIDER INGENIO-2010 CSD2006-0015) and the
Deutsche Forschungsgemeinschaft (DFG) with a grant to J.L.
(LA 2474/1-1).
16 I. Tommasi and F. Sorrentino, Tetrahedron Lett., 2005, 46, 2141–2145.
ꢁc
This journal is The Royal Society of Chemistry 2008
4824 | Chem. Commun., 2008, 4822–4824