Phosphine-olefin ligands: A facile dehydrogenative route to catalytically active rhodium complexes

Article abstract of DOI:10.1039/b608129k

Facile, metal-mediated, (acceptorless) dehydrogenation of tricyclopentyl phosphine directly affords rhodium chelating phosphine-olefin complexes, some of which are catalytically active for 1,4-additions. The Royal Society of Chemistry 2006.

Full text of DOI:10.1039/b608129k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Phosphine–olefin ligands: a facile dehydrogenative route to catalytically
active rhodium complexes{
Thomas M. Douglas, Je´roˆme Le Noˆtre, Simon K. Brayshaw, Christopher G. Frost* and Andrew S. Weller*
Received (in Cambridge, UK) 8th June 2006, Accepted 4th July 2006
First published as an Advance Article on the web 21st July 2006
DOI: 10.1039/b608129k
in dehydrogenation of one of the cyclopentyl rings and formation
of the phosphine–olefin complex [Rh(dppe){PCyp2(g2-
C5H7)}][BArF4] (1) (Scheme 2). Remarkably, this dehydrogenation
reaction occurs rapidly (minutes) at room temperature without a
hydrogen acceptor and is quantitative (by NMR spectroscopy, see
ESI). The reaction is also selective: only the 3,4-alkene is formed
and reaction occurs on only one ring. The removal of the halide to
generate a vacant site, subsequent C–H activation of the
cyclopentyl ring via an agostic intermediate, b-elimination and
finally H2 loss is the most likely mechanism.2
Facile, metal-mediated, (acceptorless) dehydrogenation of
tricyclopentyl phosphine directly affords rhodium chelating
phosphine–olefin complexes, some of which are catalytically
active for 1,4-additions.
The selective activation and functionalisation of alkanes is one of
the major goals in the field of organometallic chemistry.1
A
particularly useful case of alkane functionalisation is dehydrogena-
tion to form alkenes. The dehydrogenation of alkanes usually
requires either high temperatures or the use of a hydrogen
acceptor.2 In cases where the process is chelate assisted
(intramolecular), mild conditions can be used for dehydrogenation.
For example, Sabo-Etienne et al. have reported that treatment of
the bis-dihydrogen complex Ru(PCy3)2H2(g2-H2)2 with ethene
affords the cyclohexenyl complex Ru(PCy3){PCy2(g3-
C6H8)}H(g2-H2CLCH2) A,3 while Gru¨tzmacher et al. have
reported that ligands based upon dibenzocycloheptyl phosphine
undergo dehydrogenation on complexation with a metal centre, to
give phosphine–olefin complexes such as B.4
The assignment of (1) as a rhodium(I) complex containing a
phosphine–olefin was made using NMR spectroscopy and single
1
crystal X-ray crystallography.{ The H NMR spectrum shows a
signal attributable to a coordinated alkene at d 4.89, which shows
one-bond correlation to a signal at d 96.2 [dd, J 10 Hz, 9 Hz] in the
13C NMR spectrum, fully consistent with a coordinated alkene.
The 31P NMR spectrum shows the expected three signals (see ESI),
two with characteristically large trans P–P coupling [J(PP) 283 Hz].
In the solid-state (Fig. 1) the presence of a coordinated CLC
double bond was confirmed by a short C–C distance [C3–C4
The synthesis of mixed phosphine–olefin ligands is of significant
interest.4–7 Such ligands have been used by Hayashi et al.
(C Scheme 1) in the catalysis of 1,4-addition reactions6 and by
Gru¨tzmacher in hydrogenation reactions.7 However, the synthesis
of such ligands is multistep and can be non-trivial. Here we report
the straightforward and facile synthesis of a rhodium complex
containing a phosphine–olefin ligand (D) by the dehydrogenation
of a cyclopentyl group in the commercially available tricyclopen-
tylphosphine (PCyp3) ligand.8 We also show that such complexes
are effective catalysts for 1,4-addition reactions. That the active
catalyst species is available in excellent yields and purity from
readily available starting materials in two simple steps makes this
methodology particularly attractive.
˚
1.372(3) A] together with M–C distances typical of M–olefin
˚
coordination [viz. 2.241(2), 2.242(2) A]. Additionally (given the
usual caveats regarding the location of hydrogens by X-ray
diffraction), only one hydrogen for C3 and C4 respectively was
located in the final difference map. The newly formed phosphine–
olefin ligand has a bite angle (/ P–Rh–centroid of C3/C4) of 84u,
which is the same as that of dppe.
The intramolecular dehydrogenation of cyclopentane was also
observed in other rhodium complexes. Treatment of [Rh(nbd)Cl]2
with PCyp3 gives RhCl(nbd)(PCyp3), which when treated with
Na[BArF4] in fluorobenzene results in dehydrogenation and
formation
of
the
fluorobenzene
complex
[Rh(g6-
C6H5F){PCyp2(g2-C5H7)}][BArF4] (2) (Scheme 3). Use of benzene
as co-solvent gives the analogous benzene complex (3). In the
formation of (2) and (3), the coordinated norbornadiene acts as a
hydrogen acceptor, and norbornene is observed in the reaction
mixture (by GC or 1H NMR). The 1H and 13C{1H} NMR spectra
Treatment of RhCl(dppe)(PCyp3) with sodium tetrakis(1,3-
bis(trifluoromethyl)phenyl)borate {Na[BArF4]} in CH2Cl2 results
Scheme 1
Department of Chemistry, University of Bath, Bath, UK BA2 7AY.
E-mail: a.s.weller@bath.ac.uk; c.g.frost@bath.ac.uk
{ Electronic supplementary information (ESI) available: Experimental
details, spectroscopic and crystallographic data for all the new
compounds. See DOI: 10.1039/b608129k
Scheme 2
3408 | Chem. Commun., 2006, 3408–3410
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Complex (2) acts as a convenient source of the synthetically
useful {PCyp2(g2-C5H7)Rh}+ fragment, with other ligands displa-
cing the weakly bound fluorobenzene. Reaction with benzene gives
(3), while the reaction with two electron donors such as THF or
CH3CN gives the 16-electron, Rh(I), complexes [Rh{PCyp2(g2-
C5H7)}L2][BArF4] [L = THF (4), MeCN (5), Scheme 4], which
1
have been characterised by H, 13C and 31P NMR spectroscopy
(see ESI).
In previous studies, Hayashi et al. have established that hybrid
phosphine–olefin ligands can offer distinct advantages over
bidentate phosphine or diene ligands in the rhodium-catalysed
1,4-addition reaction.6 The convenient preparation of (2) in only
two steps from commercially available starting materials, coupled
with the subsequent ease of fluorobenzene displacement makes it
an attractive complex for use in catalytic studies. Thus, the 1,4-
addition of phenylboronic acid to 2-cyclohexenone
I was
performed using 1 mol% of complex (2) as shown in Scheme 5.9
The substrate was efficiently converted to the desired product II,
which was isolated in 93% yield after column chromatography.
Having established that (2) is an effective catalyst for the 1,4-
addition process, the chemoselectivity was investigated in a
competitive 1,2- versus 1,4-addition reaction. The addition of
phenylboronic acid to cinnamaldehyde III was examined at room
temperature using 1 mol% of complex (2) as depicted in Scheme 6.
Interestingly, in the presence of (2) the 1,4-addition product IV
was obtained exclusively. Although the conversion was modest at
room temperature the 1,2-addition product V was not detected by
1H NMR of the crude reaction mixture. The excellent chemo-
selectivity of (2) towards 1,4-addition is consistent with previous
observations involving rhodium complexes with olefin ligands.10
Miyaura et al. have noted the chemoselectivity switch to 1,2-
addition in the presence of one equivalent of t-Bu3P but the
selectivity is sensitive to the choice of reaction conditions, with
Fig. 1 ORTEP plot (50% ellipsoids) of the cationic portion of (1).
˚
Hydrogen atoms omitted for clarity. Selected bond lengths (A) and angles
(u): Rh–P1 2.3284(5), Rh–P2 2.2818(5), Rh–P3 2.2709(5), Rh–C3 2.242(2),
Rh–C4 2.241(2), C3–C4 1.372(3), P3–Rh–P2 83.823(19), P3–Rh–P1
174.42(2), P2–Rh–P1 100.068(19). Ct(3/4)–Rh–P3 93.01(6), Ct(3/4)–Rh–
P2 172.15(6), Ct(3/4)–Rh–P1 83.63(6). Ct(3/4) is the centroid of the C3/C4
bond.
Scheme 3
of (3) show signals arising from the coordinated arene, together
with a signal characteristic of a coordinated alkene [1H: d 4.3,
13C: d 62.72 d, J(RhC) 16 Hz, HCLCH]. Similar signals are
observed in the spectra of (2). (2) and (3) have similar 31P{1H}
NMR spectra, with (2) showing additional coupling due to the
fluorine [J(FP) 3.8 Hz]. The solid-state structures (Fig. 2) confirm
the presence of a coordinated CLC double bond by short C–C
Scheme 4
˚
distances [(2) C13–C14 1.401(4); (3) C13–C14 1.414(5) A] together
with M–C distances typical of M–olefin coordination.
Scheme 5
Fig. 2 ORTEP plot (50% ellipsoids) of the cationic portion of (2) (left)
and (3) (right). Hydrogen atoms omitted for clarity. Selected bond lengths
˚
(A) (2): Rh–P 2.2412(6), Rh–C13 2.125(2), Rh–C14 2.128(2), C13–C14
1.401(4); (3): Rh–P 2.219(3), Rh–C13 2.133(6), Rh–C14 2.114(4), C13–C14
1.414(5).
Scheme 6
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3408–3410 | 3409

View Article Online
that PCyp3 may not always act as a completely innocent ligand
when complexed with a late-transition metal.
Table 1 Addition of phenylzinc chloride to activated alkenes
catalysed by complex (2)a
Notes and references
{ Crystallographic data. Intensity data were collected at 150 K on a Nonius
Kappa CCD, using graphite monochromated MoKa radiation (l =
˚
0.71073 A). 1: C79H66BF25P3Rh, M = 1696.95, monoclinic, space group
˚
˚
˚
P21/c (Z = 4), a = 23.8390(2) A, b = 13.11800(10) A, c = 26.0540(2) A, b =
113.06u. V = 7496.40(10) A , m = 0.398 mm21, 2hmax = 60u. 127183
3
(2)
(mol%) time (h) (Yield %)c
Reaction Conversionb
˚
reflections collected, 21803 unique [R(int) = 0.0666]. Final wR2 = 0.1152 (all
Entry R1
R2 R3 R4
data), R1 = 0.0416 [I . 2s(I)]. 2: C53H42BF25PRh, M = 1298.56,
˚
monoclinic, space group C2/c (Z = 8), a = 19.55500(10) A, b =
3
1
2
3
4
5
a
–(CH2)3–
–(CH2)3–
H
H
H
H
H
H
1.0
0.1
2
16
2
16
16
100 (89)
100
100 (94)
100
100 (85)
˚
˚
˚
16.35700(10) A, c = 34.5800(2) A, b = 106.38u, V = 10611.70(10) A , m =
0.477 mm21, 2hmax = 58.2u, 94762 reflections collected, 14840 unique
[R(int) = 0.0633]. Final wR2 = 0.0938 (all data), R1 = 0.0428 [I . 2s(I)]. 3:
C53H43BF24PRh, M = 1280.56, monoclinic, space group C2/c (Z = 8), a =
OMe
OMe
Ph
H
H
H
CH2CO2Me 1.0
CH2CO2Me 0.1
Ph H
1.0
˚
˚
˚
19.53700(10) A, b = 16.43900(10) A, c = 34.2820(2) A, b = 106.53u, V =
All reactions are performed on 0.5 mmol of substrate.
Isolated yield after column
10555.15(10) A , m = 0.476 mm21, 2hmax = 63u, 147742 reflections collected,
3
˚
b
c
Determined by 1H NMR.
chromatography.
17437 unique [R(int) = 0.0475]. Final wR2 = 0.0941 (all data), R1 = 0.0365
[I . 2s(I)]. CCDC 610374–610376. For crystallographic data in CIF or
other electronic format see DOI: 10.1039/b608129k
cationic rhodium complexes predominantly affording the 1,4-
addition product.11
1 B. A. Arndtsen, R. G. Bergman, T. A. Mobley and T. H. Peterson, Acc.
Chem. Res., 1995, 28, 154–162; R. H. Crabtree, J. Chem. Soc., Dalton
Trans., 2001, 2437–2450.
2 C. M. Jensen, Chem. Commun., 1999, 2443–2449.
3 A. F. Borowski, S. Sabo-Etienne, M. L. Christ, B. Donnadieu and
B. Chaudret, Organometallics, 1996, 15, 1427–1434.
4 S. Deblon, L. Liesum, J. Harmer, H. Scho¨nberg, A. Schweiger and
H. Gru¨tzmacher, Chem.–Eur. J., 2002, 8, 601–611.
5 J. Thomaier, S. Boulmaaˆz, H. Scho¨nberg, H. Ru¨egger, A. Currao,
H. Gru¨tzmacher, H. Hillebrecht and H. Pritzkow, New J. Chem., 1998,
947–958.
6 R. Shintani, W.-L. Duan, T. Nagano, A. Okada and T. Hayashi,
Angew. Chem., Int. Ed., 2005, 44, 4611–4614.
7 P. Maire, S. Deblon, F. Breher, J. Geier, C. Bo¨hler, H. Ru¨egger,
H. Scho¨nberg and H. Gru¨tzmacher, Chem.–Eur. J., 2004, 10,
4198–4205.
8 Dehydrogenation of the PCyp3 ligand also occurs with a hydrogen
acceptor in certain ruthenium systems related to A. S. Sabo-Etienne,
personal communication. See also: M. Grellier, L. Vendier, B. Chaudret,
A. Albinati, S. Rizzato, S. Mason and S. Sabo-Etienne, J. Am. Chem.
Soc., 2005, 127, 17592–17593 for the related complex Ru(PCyp3)2H2(g2-
H2)2.
9 T. Hayashi and K. Yamasaki, Chem. Rev., 2003, 103, 2829–2844;
K. Fagnou and M. Lautens, Chem. Rev., 2003, 103, 169–196; N. Krause
and A. Hoffmann-Ro¨der, Synthesis, 2001, 171–196; M. P. Sibi and
S. Manyem, Tetrahedron, 2000, 56, 8033–8205; S. Darses and
J.-P. Genet, Eur. J. Org. Chem., 2003, 4313–4327.
The utility of arylzinc nucleophiles in the rhodium-catalysed 1,4-
addition reaction is known but to date there are relatively few
examples.12 To further explore the scope of (2) in 1,4-additions, the
addition of phenylzinc chloride to 2-cyclohexenone I, dimethyl
itaconate and chalcone was investigated and the results are shown
in Table 1.
It is useful to note that water is not required for catalyst
turnover when arylzinc reagents are employed in rhodium-
catalysed 1,4-addition reactions.12 The inert conditions are clearly
beneficial to the catalytic activity of (2). In the addition to
2-cyclohexenone I, the reaction reached completion after 2 hours
at room temperature with 1 mol% of catalyst and within 16 hours
with just 0.1 mol% of catalyst (Table 1, entries 1–2). In a repeat of
the addition with 1 mol% of catalyst, NMR analysis of the crude
reaction mixture revealed the presence of the intact {PCyp2(g2-
C5H7)Rh}+ fragment in conjunction with complete conversion of I
to the intermediate silyl enol ether. Furthermore, the catalyst
species was still active. Upon the addition of a further 0.5 mmol of
substrate and appropriate reactants, an overall 90% conversion
was observed after 2 more hours.
An efficient 1,4-addition was also realised with the less reactive
1,19-disubstituted alkene dimethyl itaconate.13 As before, the
addition of phenylzinc chloride to dimethyl itaconate was complete
after 2 hours at room temperature with 1 mol% of (2) and within
16 hours in the presence of 0.1 mol% of (2) (Table 1, entries 3–4).
Remarkably, the more challenging acyclic a,b-unsaturated ketone
chalcone also undergoes complete conversion to the 1,4-addition
product at room temperature with 1 mol% of (2) (Table 1,
entry 5).14
10 J.-F. Paquin, C. Defieber, C. R. J. Stephenson and E. M. Carreira,
J. Am. Chem. Soc., 2005, 127, 10850–10851; T. Hayashi, N. Tokunaga,
K. Okamoto and R. Shintani, Chem. Lett., 2005, 34, 1480–1481;
N. Tokunaga and T. Hayashi, Tetrahedron: Asymmetry, 2006, 17,
607–613.
11 M. Ueda and N. Miyaura, J. Org. Chem., 2000, 65, 4450–4452.
12 R. Shintani, T. Yamagami, T. Kimura and T. Hayashi, Org. Lett., 2005,
7, 5317–5319; R. Shintani, N. Tokunaga, H. Doi and T. Hayashi, J. Am.
Chem. Soc., 2004, 126, 6240–6241.
13 For examples of 1,4-addition of organoboron derivatives to dimethyl
itaconate, see: R. J. Moss, K. J. Wadsworth, C. J. Chapman and
C. G. Frost, Chem. Commun., 2004, 1984–1985; K. J. Wadsworth,
F. K. Wood, C. J. Chapman and C. G. Frost, Synlett, 2004, 2022–2024.
14 T. Koike, X. Du, A. Mori and K. Oskada, Synlett, 2002, 301–303; S. Oi,
Y. Honma and Y. Inoue, Org. Lett., 2002, 4, 667–669; S. Oi, M. Moro,
H. Ito, Y. Honma, S. Miyano and Y. Inoue, Tetrahedron, 2002, 58,
91–97.
In summary, a concise and practical synthesis of a rhodium
complex containing a hybrid phosphine–olefin ligand is presented.
That this occurs by a facile transfer dehydrogenation of a
cyclopentyl group in PCyp3 not only suggests future strategies
for the synthesis of this important class of ligand, but also shows
3410 | Chem. Commun., 2006, 3408–3410
This journal is ß The Royal Society of Chemistry 2006