Rhodium-NHC-complexes as potent catalysts in the hydroformylation of 1-octene

Article abstract of DOI:10.1016/j.jorganchem.2005.09.038

We present an overview of eight different rhodium-N-heterocyclic carbene (NHC) complexes 1-8 and their catalytic activity and selectivity in hydroformylation of 1-octene. It could be shown that activity can be increased by going from electron-rich NHC- to electron-poor NHC-ligands. However, no increase in the selectivity could be achieved by introducing bulky substituents to the NHCs.

Full text of DOI:10.1016/j.jorganchem.2005.09.038

Journal of Organometallic Chemistry 690 (2005) 6233–6237
www.elsevier.com/locate/jorganchem
Rhodium–NHC-complexes as potent catalysts
in the hydroformylation of 1-octene
a
b
b
a
Martin Bortenschlager , Jan Schutz , Denise von Preysing , Oskar Nuyken ,
¨
b
a,
*
Wolfgang A. Herrmann , Ralf Weberskirch
a
Department Chemie, Lehrstuhl fu¨r Makromolekulare Stoffe, TU Mu¨nchen, Lichtenbergstrasse 4, D-85747 Garching, Germany
Department Chemie, Lehrstuhl fu¨r Anorganische Chemie, TU Mu¨nchen, Lichtenbergstrasse 4, D-85747 Garching, Germany
b
Received 25 May 2005; received in revised form 15 September 2005; accepted 15 September 2005
Available online 27 October 2005
Abstract
We present an overview of eight different rhodium-N-heterocyclic carbene (NHC) complexes 1–8 and their catalytic activity and selec-
tivity in hydroformylation of 1-octene. It could be shown that activity can be increased by going from electron-rich NHC- to electron-
poor NHC-ligands. However, no increase in the selectivity could be achieved by introducing bulky substituents to the NHCs.
Ó 2005 Elsevier B.V. All rights reserved.
Keywords: Hydroformylation; NHC/rhodium catalysts; Steric/electronic effects
1. Introduction
phosphine ligands. However, phosphine ligands have some
disadvantages in common: in solution they are easily oxi-
dized by molecular oxygen. Furthermore, phosphines and
CO show similar binding constants to rhodium. Due to
the fact, that CO-pressures applied during hydroformyla-
tion are quite high, an excess of phosphine is required to
generate a sterically demanding environment around the
active rhodium-center, a prerequisite for high n/iso-ratios.
With the introduction of N-heterocyclic carbenes
(NHCs) [6–10] a novel class of ligands have been developed
that display similar binding properties than phosphines.
They act also as very strong r-donors and very poor p-
acceptors but form at the same time much more stable
metal–NHC-bonds in many cases [11]. The resulting com-
plexes are stable towards air and moisture. Recently, it
could even be shown that rhodium–NHC-bonds remain
stable under hydroformylation conditions (CO/H2-pres-
sure: up to 50 bar, temperature up to 100 °C) [12,13],
meaning that CO is not able to substitute NHC-ligands
in rhodium–NHC complexes. Therefore, rhodium–NHC-
complexes should be well suitable as catalysts for hydro-
formylation, even in aqueous systems, as has been
repeatedly proven [12–15]. More recently, we demonstrated
Hydroformylation of olefins is one of the most important
homogeneous catalyzed reactions in industry. Formally, it is
an addition of CO and H2 to a C–C-double bond yielding
the corresponding n- and iso-aldehydes [1]. In case of
aliphatic olefins of long chain double-bond-isomerization
may also occur during the reaction resulting in mostly inter-
nal olefins. Due to their superior properties n-aldehydes are
usually the preferred product for most applications. There-
fore, selectivity is very important in hydroformylation and
is expressed by the ratio of n-aldehyde to iso-aldehydes
(n/iso-ratio).
Nowadays most of the hydroformylation catalysts are
based on rhodium and phosphines (e.g., Wilkinson-catalyst
Rh(I)Cl(PPh3)3 [2]). The catalytic mechanisms of such sys-
tems, depending on the ligands employed, are well under-
stood [3–5]. Therefore, it is possible to optimize both
activity and selectivity by adjusting the structure of the
*
Corresponding author.
E-mail address: Ralf.Weberskirch@ch.tum.de (R. Weberskirch).
0022-328X/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.09.038

6234
M. Bortenschlager et al. / Journal of Organometallic Chemistry 690 (2005) 6233–6237
the first successful aqueous biphasic hydroformylation of 1-
octene with a NHC/rhodium catalyst when immobilized to
an amphiphilic polymer [16].
In this contribution, we give an overview of different
rhodium–NHC-complexes and their catalytic activity and
selectivity in hydroformylation of 1-octene. Sterical as well
as electronical variations of the NHC-ligand and their
influence on the catalytic results are discussed.
The catalysts (1–8) vary in both sterical demand and in
electron density around the rhodium-center. Both parame-
ters are determined by the attached NHC-ligand only. The
activity and selectivity of catalysts 1–4 was recently studied
in our group [13]. Catalysts 5–8 have been investigated un-
der the same reaction conditions and activities were calcu-
lated in all cases from the point of inflection in the slope of
aldehyde formation. As has been shown recently neither
the cyclooctadiene (COD) nor the halide has any influence
on the results of the hydroformylation of 1-octene (Table 1,
entries 1–4) [13]. In all cases, the catalytically active species
is supposed to be a Rh(NHC)-carbonyl-hydrido-complex
which is formed in situ under hydroformylation conditions
analogous to the Wilkinson-catalyst [19] (Fig. 2). First the
COD-ligand is substituted by CO. Second the rhodium-
hydrido-complex, the active species, is formed by an oxida-
tive addition of hydrogen and a consecutive reductive
2. Experimental part
All experiments were carried out in a 300 mL Parr high
pressure reactor. The reactor was evacuated, flushed with
nitrogen and filled with Rh-complex (10À5 mol), toluene
(100 mL), and 1-octene (5.6 g) leading to a substrate:cata-
lyst ratio of 5000. Undecane was added as internal stan-
dard. The mixture was pressurized twice to 30 bar
synthesis gas (CO:H2 = 1:1) to clean all supplies before
the pressure was adjusted to 50 bar with a back pressure
regulator. Then the autoclave was heated to 100 °C and
kept at this temperature. Samples were taken every 15–
30 min and the products were quantified via gas
chromatography.
Table 1
Catalytic hydroformylation activitya (TOFs) of complex (1–8)
Complex
TOF (hÀ1 c
)
Conversion (%)d
1b
2b
3b
4b
5
480
520
8.1
11.8
21.5
23.0
28.8
19.1
39.7
50.4
1340
1480
1785
1150
2410
3540
3. Results and discussion
A selection of imidazole- (5–6), xanthin- (7) and tetra-
zole-based rhodium–NHC-complexes (8) (Fig. 1) have been
characterized with regard to their catalytic activity and
selectivity of the hydroformylation of 1-octene. The results
have been compared to that of the recently published tetra-
hydropyrimidine-based NHC-complexes (1–4) [13]. The
synthesis and structural data of the rhodium–NHC-
complexes presented here have already been published else-
where by Buchmeiser and Herrmann, respectively [17,18].
6
7
8
a
Reaction conditions: T = 100 °C, p(H2/CO) = 50 bar, CO:H2 = 1:1,
substrate:catalyst ratio = 5000, c(Rh) = 10À4 mol/L, solvent:toluene.
b
Published by Buchmeiser et al. [13].
Based on the amount of aldehydes formed, olefin isomerization also
c
occurred.
d
Determined after 1 h reaction time (induction period is not taken into
account).
N
N
N
N
N
N
N
N
Mes
Mes
Mes
Mes
Rh
Rh
Rh
Rh
Cl
Br
Cl
Br
2
1
3
4
O
N
O
N
N N
N
N
N
N
Mes
Mes
Ad
Ad
N
N
N
N
Rh
Rh
Rh
Rh
Cl
Cl
I
I
6
5
7
8
Fig. 1. Rh–NHC-complexes (1–8) for the hydroformylation of 1-octene.

M. Bortenschlager et al. / Journal of Organometallic Chemistry 690 (2005) 6233–6237
6235
NHC
NHC
NHC
NHC
Rh
H2
CO
- HX
OC
OC
H
H
OC
OC
OC
OC
Rh
X
Rh
Rh
X
H
X
Fig. 2. Proposed mechanism for the formation of the active Rh(NHC)-carbonyl-hydrido-species.
elimination of HX. During the whole process, the rho-
dium–NHC-bond remains stable as successfully has been
shown by several groups [12,13,16]. Complex 7 has been
investigated by 13C NMR after 20 h hydroformylation
(100 °C, 50 bar CO/H2). The spectrum shows a signal for
the carbon atom of the Rh–NHC complex with a chemical
shift of 186.5 ppm and a coupling constant (Rh–C) of
J = 55.2 Hz. This is in good accordance with the literature
values found for compound 7 [18], where a chemical shift
of 189.6 ppm and a coupling constant of 52.2 is given con-
sidering the fact that compound 7 after 20 h hydroformyla-
tion should have lost the iodo as well as the COD ligand
(substituted by CO and H) according to the underlying
mechanism of hydroformylation.
For catalyst (5–8) time-conversion-rates were measured
and the initial activity (turnover frequency, TOF) was de-
rived from the maximum slope at the onset of hydroformy-
lation (Table 1). This method limits the uncertainties
coming from (i) different inductions periods for each cata-
lyst and (ii) the time for heating up the reaction vessel to
100 °C/50 bar pressure.
Selectivity was measured as a ratio between formed n-
aldehyde to formed iso-aldehydes. Due to the fact, that ole-
fin-isomerization occurred with all eight catalysts, the
n/iso-ratio dropped with increasing conversion in all cases.
Therefore, the selectivity of all catalysts is compared in a
n/iso-conversion-plot (Fig. 3).
The activities of the different catalysts (Table 1) range
from 480 hÀ1 for catalyst 1 up to 3500 hÀ1 for catalyst 8.
An exception is catalyst 6, which decomposes during
hydroformylation and precipitates as a brown solid that
cannot be further analyzed. With its extremely bulky ada-
mantyl-substituents the NHC-ligand seems no longer be
able to form a stable bond towards the rhodium in this case.
We assume that the increase in the catalytic activity can be
explained by a decrease of the electron-donating strength of
the NHC-ligands going from catalysts 1–8 as it is with
phosphine ligands [20,21]. While the tetrahydropyrimi-
dine-based NHC-complexes are very strong donor-ligands,
indicated by very low wavenumbers m(COI)/m(COII) =
2063/1982 cmÀ1 for the CO-complex of 1 and m(COI)/
m(COII) = 2062/1976 cmÀ1 for the CO-complex of 3 [17],
the purine- and tetrazole-based NHCs have poor electron-
donating properties for this class of ligand showing high
wavenumbers m(COI)/m(COII) = 2080/2009 cmÀ1 for the
CO-complex of 7 [18] and m(COI)/m(COII) = 2086/2015
cmÀ1 for the CO-complex of 8 [22]. The increase in activity
with a decrease of the electron-donating properties of the
NHC-ligand can be explained by the formation of the inter-
mediate rhodium-p-alkene complex (Fig. 4) which leads to
the rhodium-alkyl complex, the first step in the catalytic cy-
cle of hydroformylation. Although the low electron density
at the rhodium center make the formation of the p-complex
more difficult attack of the hydride to the double bond is
facilitated by the electron metal center.
Looking at the selectivities (Fig. 3) it becomes clear that
all catalysts presented here show similar n/iso-ratios with
increasing conversion. The n/iso-ratios are quite high at
the beginning of the reaction ranging from 1.5 to 2.5 but
are rapidly dropping with increasing conversion. In Fig. 5
are the different fractions occurring during hydroformyla-
tion depicted showing clearly the strong isomerization
behavior of the catalysts 5–8 which is the reason for low
n/iso ratioÕs after complete 1-octene conversion.
Due to olefin-isomerization, which is taking place as a side
reaction with all eight catalyst systems, iso-aldehydes can be
formed from the resulting internal olefins. This leads to a de-
crease of the n/iso-ratio down to 0.5 for all catalysts (except
for 6, which decomposes during the reaction). Olefin-isomer-
ization usually can be suppressed by providing a high sterical
demand around the rhodium-center of the catalyst. Regard-
ing phosphine-ligands this is achieved by either using an
NHC
Rh
R
NHC
NHC
Rh
OC
OC
OC
OC
OC
OC
Rh
H
R
R
H
Fig. 4. Formation of the rhodium-alkyl complex via an intermediate
Fig. 3. n/iso-selectivities for catalysts 1–8 as a function of conversion.
p-complex.

6236
M. Bortenschlager et al. / Journal of Organometallic Chemistry 690 (2005) 6233–6237
Fig. 5. Fraction of different compounds during the hydroformylation process of 1-octene in the presence of 5 (a), 6 (b), 7 (c) and 8 (d) as a function of
conversion (black = 1-octene, gray = internal octenes, white = n-aldehyde, crossed = iso-aldehydes).
excess of phosphine or by using bidentate phosphines [23–
26]. This way it is assured that at least two of the rhodium-
coordination centers are occupied by phosphine ligands. In
case of rhodium–mono-NHC-complexes the only way for
a high sterical demand around the metal center is by using
bulky carbene-ligands. The bulky character of carbene li-
gands is mostly determined by the substituents at the nitro-
gen atoms of the NHC. Going from methyl- (7, 8) over
iso-propyl- (1, 2) to mesityl-substituents (3, 4, 5) obviously
has no influence on the n/iso-ratio. The attempt to further en-
large the substituent by introducing adamantyl-substituents
to the NHC-ligand failed due to instability of the resulting
complex (6) towards hydroformylation conditions. It be-
comes clear that a suppression of olefin-isomerization and
therefore an improvement of the n/iso-ratio cannot be
achieved by a single bulky NHC-ligand.
investigations on rhodium–bis carbene-complexes and their
catalytic activity in hydroformylation have yet to be made.
Acknowledgment
Financial support provided by Freistaat Bayern was
gratefully acknowledged.
References
[1] K. Weissermel, H.J. Arpe, Industrial Organic Chemistry, VHC, New
York, 1993.
[2] D. Evans, J.A. Osborn, G.J. Wilkinson, J. Chem. Soc. A (1968) 3133.
[3] J.D. Unruh, J.R. Christenson, J. Mol. Cat. 14 (1) (1982) 19.
[4] J.J. Carbo, F. Maseras, C. Bo, P.W.N.M. van Leeuwen, J. Am.
Chem. Soc. 123 (31) (2001) 7630.
[5] L.A. van der Veen, P.C.J. Kamer, P.W.N.M. van Leeuwen,
CATTECH 6 (3) (2002) 116.
[6] H.-W. Wanzlick, E. Schikora, Chem. Ber. 94 (1961) 2389.
[7] H.J. Scho¨nherr, H.W. Wanzlick, Liebigs Ann. Chem. 731 (1970) 176.
[8] H.W. Wanzlick, H.J. Scho¨nherr, Angew. Chem. 80 (1968) 154.
[9] W.A. Herrmann, C. Ko¨cher, Angew. Chem. 109 (1997) 2256.
[10] W.A. Herrmann, Angew. Chem. 114 (2002) 1342.
[11] (a) For example, see: H. Schumann, J. Gottfriedsen, M. Glanz, S.
Dechert, J. Demtschuk, J. Organomet. Chem. 617–618 (2001) 588;
4. Conclusion
In summary, we have investigated the hydroformylation
activity and selectivity of four novel Rh–NHC-complexes
and compared the results to those Rh–NHC-complexes re-
cently published [13]. We could show that electron-poor
NHC-ligands lead to a higher hydroformylation activity
with TOFs up to 3500 hÀ1. However selectivity is still low
with all presented catalysts. Applying bulky substituents
to the NHC-ligand has no influence on the selectivity at
all. Introducing a second NHC-ligand to the rhodium-
complexes could be a promising solution to that. Further
¨
(b) K. Ofele, W.A. Herrmann, D. Mihalios, M. Elison, E. Herdtweck,
T. Priermeier, P. Kiprof, J. Organomet. Chem. 498 (1995) 1;
(c) M.H. Voges, C. Rommig, M. Tilset, Organometallics 18 (1999)
529;
M.V. Baker, B.W. Skelton, A.H. White, C.C. Williams, J. Chem.
Soc., Dalton Trans. (2001) 111;
(d) A.A.D. Tulloch, A.A. Danopoulos, S. Kleinhenz, M.E. Light,
M.B. Hursthouse, G. Eastham, Organometallics 20 (2001) 2027;

M. Bortenschlager et al. / Journal of Organometallic Chemistry 690 (2005) 6233–6237
6237
(e) J. Pytkowicz, S. Roland, P. Mangeney, J. Organomet. Chem. 631 [18] J. Schutz, W.A. Herrmann, J. Organomet. Chem. 689 (2004) 2995.
¨
(2001) 157.
[19] D. Cauzzi, M. Costa, L. Gonsalvi, M.A. Pellinghelli, G. Predieri, A.
Tiripicchio, R. Zanoni, J. Organomet. Chem. 541 (1997) 377.
[20] X. Wang, H.Y. Fu, X. Li, H. Chen, Catal. Commun. 5 (2004)
739.
[21] H. Klein, R. Jackstell, K.-D. Wiese, C. Borgmann, M. Beller, Angew.
Chem., Int. Ed. 48 (2001) 3408.
[12] M. Poyatos, P. Uriz, J.A. Mata, C. Claver, E. Fernandez, E. Peris,
Organometallics 22 (2003) 440.
[13] M. Bortenschlager, M. Mayr, O. Nuyken, M.R. Buchmeiser, J. Mol.
Catal. A: Chem. 233 (2005) 67.
[14] W.A. Herrmann, M. Elison, J. Fischer, C. Kocher, U.S. Patent
5,663,451, 1997.
[22] J. Schutz, Ph.D. Thesis, Technische Universita¨t Munchen, 2005.
¨
¨
[15] A.C. Chen, L. Ren, A. Decken, C.M. Crudden, Organometallics 19
(2000) 3459.
[23] L.A. van der Veen, P.C.J. Kramer, P.W.N.M. van Leeuwen, Angew.
Chem. 111 (1999) 349.
[16] M.T. Zarka, M. Bortenschlager, K. Wurst, O. Nuyken, R. Web-
erskirch, Organometallics 23 (2004) 4817.
[17] M. Mayr, K. Wurst, K.-H. Ongania, M.R. Buchmeiser, Chem. Eur.
J. 10 (2004) 1256.
[24] M. Matsumoto, M. Tamura, J. Mol. Catal. 16 (1982) 209.
[25] J.D. Unruh, J.R. Christenson, O.R. Richard, D.A. Young, Chem.
Ind. (Dekker) (1981) 309.
[26] B. Breit, W. Seiche, J. Am. Chem. Soc. 125 (2003) 6608.