Novel α-fluorinated cyclic phosphite and phosphinite ligands and their Rh-complexes as suitable catalysts in hydroformylation

Article abstract of DOI:10.1016/j.jorganchem.2005.04.038

Asymmetrical cyclic phosphite and phosphinite ligands of a novel type bearing either trifluoromethyl or pentafluorophenyl group were synthesized using >PCl or >PN< species and racemic fluorinated alcohols. The P-ligands were converted to complexes of Rh^III(L)(Cp^*)Cl ₂ type (where L = phosphite or phosphinite) and, in two instances, their stereostructures were evaluated by X-ray analysis. These complexes along with in situ systems, formed from Rh(CO)₂(acac) precursor and the corresponding ligand, were tested in the hydroformylation of styrene. Both systems provided excellent hydroformylation activities at 100 °C. Using the Rh^I in situ systems, moderate and high regioselectivities towards the branched aldehyde (2-phenyl-propanal) were obtained at 100 and 40 °C, respectively.

Full text of DOI:10.1016/j.jorganchem.2005.04.038

Journal of Organometallic Chemistry 690 (2005) 3456–3464
www.elsevier.com/locate/jorganchem
Novel a-fluorinated cyclic phosphite and phosphinite ligands
and their Rh-complexes as suitable catalysts in hydroformylation
a,
Irina Odinets , Tamas Kegl , Elena Sharova , Oleg Artyushin , Evgenii Goryunov ,
b
a
a
a
*
´
Galina Molchanova , Konstantin Lyssenko , Tatyana Mastryukova ,
´
a
a
a
c
Gerd-Volker Ro¨schenthaler , Gyo¨rgy Keglevich , Laszlo Kollar
d
e
´
´
´
a
A.N. Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences, Thiophosporous Laboratory, Vavilova Street,
28, Moscow 119991, Russian Federation
b
´
Research Group for Petrochemistry of the Hungarian Academy of Sciences, H-8200 Veszprem, Hungary
c
Institut fuˆr Anorganische & Physikalische Chemie, Universita¨t Bremen, Germany
Department of Organic Chemical Technology, Budapest University of Technology and Economics, 1521 Budapest, Hungary
´
Department of Inorganic Chemistry, University of Pecs and Research Group for Chemical Sensors of the Hungarian Academy of Sciences,
d
e
´
H-7624 Pecs, Hungary
Received 7 February 2005; revised 22 April 2005; accepted 26 April 2005
Available online 15 June 2005
Abstract
Asymmetrical cyclic phosphite and phosphinite ligands of a novel type bearing either trifluoromethyl or pentafluorophenyl group
were synthesized using >PCl or >PN< species and racemic fluorinated alcohols. The P-ligands were converted to complexes of
RhIII(L)(Cp*)Cl2 type (where L = phosphite or phosphinite) and, in two instances, their stereostructures were evaluated by X-ray
analysis. These complexes along with in situ systems, formed from Rh(CO)2(acac) precursor and the corresponding ligand, were
tested in the hydroformylation of styrene. Both systems provided excellent hydroformylation activities at 100 ꢀC. Using the RhI
in situ systems, moderate and high regioselectivities towards the branched aldehyde (2-phenyl-propanal) were obtained at 100
and 40 ꢀC, respectively.
ꢁ 2005 Elsevier B.V. All rights reserved.
Keywords: Cyclic phosphite- and phosphinite ligands; Rhodium complexes; Stereostructure; Homogeneous catalysis; Hydroformylation
1. Introduction
formylation (e.g., that of vinylaromatics) hold an enor-
mous potential in synthesis [1]. To find an appropriate
balance of chemo-, regio and enantioselectivities, as
well as a satisfactory catalytic activity, the search for
efficient hydroformylation catalysts still remains a
challenge.
Naturally, the most important variable in a homoge-
neous catalytic system is the ligand itself. Transition me-
tal catalysts applying ligands with phosphorus donors
are used almost exclusively [2]. Although mostly mono-
and ditertiary phosphines are used, phospholes have
also been thoroughly investigated during the systematic
structural variation of the ligands [3]. Many examples
Hydroformylation is probably one of the most de-
tailed investigated homogeneous catalytic reaction.
Since the discovery of Roelen in 1938, hundreds of co-
balt, rhodium and platinum complexes have been
tested as catalysts in the hydroformylation of a great
variety of olefins. Both the hydroformylation of simple
substrates like propene and the asymmetric hydro-
*
Corresponding author. Tel.: +7 095 1135326; fax: +7 095 1355085.
E-mail address: odinets@ineos.ac.ru (I. Odinets).
0022-328X/$ - see front matter ꢁ 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.04.038

I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
3457
are known where insignificant changes in the structure
of the ligand resulted in profound change of the catalytic
activity.
cost and to simplify the introduction procedure, mostly
short ꢁpony-tailsꢁ are used nowadays. Till now, the most
studied and applied strategy has been focused on the
development of monodentate fluorinated phosphines.
They were used in hydroformylation [19], hydroboration
[20], cross-coupling reactions [21] and in other processes.
Regarding fluorine-containing phosphite ligands, a ser-
ies of symmetrical tris(polyfluoroalkyl)phosphites was
synthesized and their Ni complexes were tested in the
cyclodimerization of isoprene [22]. Only the Ni complex
with P[OCH(CF3)2]3 having a-CF3 groups on the sec-
ondary carbon atom was found to demonstrate a sur-
prisingly high activity and selectivity in contrast to
complexes with ligands based on primary polyfluoroal-
kanoles [22,23].
As a part of our continuing efforts regarding the sys-
tematic variation of ligands in rhodium-catalysed reac-
tions, our studies were extended to asymmetrical cyclic
phosphites with benzoanellation possessing a trifluo-
romethyl or pentafluorophenyl moiety in a-position. In
the present paper, the synthesis of novel P-ligands, their
use in the preparation of rhodium complexes, as well as
the test of the complexes in hydroformylation reaction
are discussed.
In recent years, the interest was shifted from r-donor
phosphines to p-acceptor phosphites and amidophosph-
ites. This is connected with the ease of their synthesis,
their high stability, absence of oxidation-sensitive P–C
bonds in the molecule and higher stability of the corre-
sponding metal complexes. Although the number of
such examples is less than those of phosphines, several
applications have been described [4]. Most of the papers
published up till now on rhodium–phosphite catalysed
hydroformylation reactions involve triphenylphosphite
[5], trialkylphosphite (trimethylphosphite [6] and tricy-
clohexylphosphite [7]) ligands. There are several cata-
lytic systems containing sterically bulky phosphites like
2-t-butyl-phenyl derivatives [8] or other 2-substituted
aryl derivatives [9]. In this respect, most of the recent
publications are devoted to the application of Rh–phos-
phite catalytic systems [10].
The application of chiral heterobidentate phosphine–
phosphite derivatives (e.g., BINAPHOS [11] and its ana-
logues [12]) with various rhodium precursors proved to
be a real breakthrough in enantioselective hydroformy-
lation of styrene providing enantiomeric excesses (e.e-
s) of practical importance. Diphosphite analogues of
the corresponding chiral diphosphines (BINAP [13],
CHIRAPHOS [14], BDPP [15]) were also tested in enan-
tioselective hydroformylations. Chiral diphosphites of
novel type possessing 2,2-dimethyl-dioxolane ring were
also used in the enantioselective hydroformylation of
styrene [16]. A review devoted to the advances in Rh-
catalyzed asymmetric hydroformylation using phosphite
ligands of different types has been published recently
[17].
2. Results and discussion
2.1. Synthesis of the ligands and RhIII complexes thereof
In order to obtain the ligands, phosphorylation of the
racemic fluorinated alcohols either by P-chlorides (ꢀchlo-
rideꢁ method), or by amides (ꢀamideꢁ method) of the corre-
sponding phosphorus(III) acids may be applied in
principle. We found that the ꢀchlorideꢁ methodology
could be successfully applied only for ligands 1a–e with
the 1,3,2-benzodioxaphospholene moiety that are not
sensitive towards oxidation (Scheme 1). In contrast, phos-
phinites are very sensitive towards oxidation, especially in
On the other hand, fluorinated ligands also find
application in the field of catalysis and are of high inter-
est in fluorous biphasic catalysis (FBC) [18]. For this
purpose, the fluorine containing ꢀpony-tailsꢁ are intro-
duced in a variety of known ligands and to reduce the
(+/-)
R2PCl +
-HOCH(CX3)Ar/Py
O
R2=
R=Ph
O
O
_
_
P
OCH(CX )Ar
3
(+/-) Ph2POCH(CX3)Ar
(+/-)
O
2b
1a-e
.
cat. Py HCl
1/2[Cp*RhCl2]2
_
(+/-) Ph2P(O)OCH(CX3)Ar
3b
LRhCp*Cl2
4a-e
X=F, Ar=C 6H5 (a), 4-ClC6H4 (b), 3-ClC6H4 (c), 3-CF3C6H4 (d);
X=H, Ar=C 6F5 (e)
Scheme 1.

3458
I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
1/2[Cp*RhCl2]2
(+/-)
-HOCH(CF3)Ar
1
(+/-)
-L RhCp*Cl2
(+/-)
-Et2POCH(CF3)Ar
Et2PNEt2
6a,b
5a,b
spontaneous
resolution for 6a
Ar=C6H5 (a), 4-ClC6H4 (b)
(-)-(5a)RhCp*Cl2
(+)-(5a)RhCp*Cl2
Scheme 2.
solutions. Thus, using the ꢀchlorideꢁ methodology, phos-
phorylation of the diphenyl- or diethylchlorophosphine
is accompanied by partial oxidation even under pro-
tective atmosphere of argon apparently catalyzed by
the pyridine hydrochloride formed. Moreover, the
typical isolation procedure in the reaction of diphenyl-
chlorophosphine with 4-ClC6H4CH(CF3)OH afforded
the corresponding phosphinate 3b instead of phosphinite
2b.
To overcome this problem, the substitution of dieth-
ylamino moiety of the phosphorus atom for the corre-
sponding residue of the racemic alcohol may be used.
The interaction proceeds at an elevated temperature in
the absence of any solvent according the procedure de-
scribed by us earlier for 5a [24] (Scheme 2). The other
diethylphosphinite ligand 5b was obtained similarly by
the ꢀamideꢁ procedure in nearly quantitative yield (see
Tables 1,2).
Interaction of the ligands 1a–e and 5a,b with dimeric
(pentamethylcyclopentadienyl)rhodium dichloride in
CH2Cl2 solution resulted in the corresponding com-
plexes 4a–e and 6a,b, respectively. They have been
obtained as dark red crystalline compounds incorporat-
ing chlorinated hydrocarbons. In some cases, the sol-
vates were disrupted by heating the complexes in
vacuo to afford the solvent-free forms as orange-red
powders. In the 31P NMR spectra, the cyclic phosphite
(4a–e) and the diethylphosphinite complexes (6a,b) were
characterized by dP chemical shifts appearing at ca. 135
and 155 ppm, respectively. Sharp doublets with charac-
teristic coupling constant (1JPRh) ranging between 241
and 246 Hz for 4a–e and 155.5–156.5 Hz for 6a,b were
observed. It should be noted that the values of chemical
shifts for the complexes are rather close to those of the
corresponding P-ligands (1a–c and 5a,b) (the Dd is ca.
10 and about 2 ppm, respectively) that may be in accord
with rather weak coordination bonds.
Single crystal X-ray analysis was carried out for rep-
resentative RhIII complexes bearing either a 1,3,2-benzo-
dioxaphospholene moiety, or two ethyl substituents on
the phosphorus atom, namely for 4d and 6a, respec-
tively. The X-ray analysis carried out for complex 4d re-
vealed that it crystallizes as a solvate with one molecule
of CH2Cl2 (Fig. 1). The complex 4d crystallizes as race-
mate (space group Pca21). The phosphorus atom in 4d is
Ligands 1a–e and 5a,b appeared as sharp singlets in
the 31P NMR spectra at ca. 126 or 154 ppm, respec-
tively. The trifluoromethyl group resonates at 0.0–
4
0.9 ppm (either as a singlet or as a doublet with a JPF
coupling constants up to 6.2 Hz) in the 19F NMR spec-
tra. In the 1H and 13C NMR spectra of the compounds,
the signals of the protons and skeletal carbon atoms can
be found in the typical regions in agreement with the
suggested structure.
Table 1
Yields and elemental analyses of 1a–e 3b, 5b and RhIII complexes 4a–e, 6b
Compound
Yield (%)a
m.p. (ꢀC)
Found (calcd.) (%)
Formula
C
H
1a
1b
1c
1d
1e
3b
4a
4b
4c
4d
4e
5b
6b
a
100(88)
93b
Oil
Oil
53.28(53.52)
–
3.20(3.21)
–
C
–
14H10F3O3P
97(80)
96(61)
100(78)
90(68)
73(58)
87(62)
92(63)
100(53)
100(67)
100(96)
100(84)
49–50(hexane)
42–43(hexane)
57–59(hexane)
60
48.56(48.23)
47.22(47.14)
48.01(48.02)
58.59(58.48)
44.02(44.20)
42.68(42.90)
43.76(43.83)
40.21(40.23)
43.61(43.73)
43.54(43.58)
43.18(43.12)
2.74(2.60)
2.30(2.37)
2.31(2.30)
3.71(3.68)
3.90(3.94)
3.74(3.64)
3.81(3.68)
3.22(3.38)
3.54(3.52)
4.68(4.57)
5.76(4.96)
C14H9ClF3O3P
C15H9F6O3P
C
C20H15ClF3O2P
14H8F5O3P
96–97
108–109
77–78
C24H25Cl2F3O3PRh Æ 0.5CH2Cl2
C24H24Cl3F3O3PRh Æ 0.5CH2Cl2
C24H24Cl3F3O3PRh Æ
C25H24Cl2F6O3PRh Æ CH2Cl2
C24H23Cl2F5O3PRh
C12H15ClF3OPSc
105–107
162–163
100–102/0.1
145–146
C22H35Cl3F3OPRh Æ
The yield according to NMR data, in brackets the yield of isolated analytically pure compound.
It was used for complexation without further purification.
Analytical data for thione derivative Et2P(S)OCH(CF3)C6H4-Cl-4 purified by column chromatography (silica gel, hexane:acetone 100:3).
b
c

Table 2
Selected 31P, 19F, 1H and 13C NMR parameters for compounds obtained
Compound
NMR spectra: ppm, J Hz, CDCl3
31P
19F
1H
13C
1
dp (1JP–Rh
127.09
)
dCF3 d (4JP–F) or s
0.29(1.2)
dOCH dq (3JP–H/3JH–F
4.94(8.4/6.4)
)
Cꢀpdð JC ꢁp
–
Þ
dOCH dq (2JP–C/2JC–F
73.77(2.1/33.6)
)
d
CF3dq (3JP–C/1JC–F
)
d
POCHC(Ar)c
dPOC(Ar) d (2JP–C
)
ꢀ
p
1a
122.76(3.8/281.1)
131.78
145.00(7.5)
144.43(7.6)
–
1b
1c
1d
126.54
126.13
125.57
–2.27
0.34
0.19(6.1)a
4.85–4.91
4.84–4.89 m
4.96(7.6/6.4)
–
–
–
–
–
–
–
–
122.35(3.8/281.4)
–
132.90
–
73.02 q
2JC–F = 34.4
63.60 s
144.82(7.4)
144.17(7.6)
145.20 (7.7)
145.05 (7.9)
b
1e
125.52
5.27mc
(3JP–H = 6.8)
–
22.27 s
115.20 dt
(2JP–C = 3.1;
3JC–F = 13.6)
2b
3b
4a
125.07
36.42
136.56 d (246.2)
0.55 (6.18)
1.25
1.25
–
–
–
–
–
–
5.63–5.70 m
5.71(8.8/6.0)
–
1.78(6.4)
–
76.42 q
2JC–F = 35.6
–
–
135.83
112.43
(6.6/47.8)
–
145.39(8.1)
145.62(7.7)
–
4b
4c
4d
4e
135.92 d (244.5)
135.51d(243.6)
135.93 d (244.0)
134.94 d (241.3)
1.07
1.29
5.94(8.8/6.4)
5.91–5.97 m
6.18(7.6/6.4)
5.99
1.74(6.4)
1.74(6.6)
1.74(6.8)
1.76(6.7)
–
–
1.21a
–
–
–
–
–
–
–
–
b
(3JP–H = 7.1) apparent qv
4.92–5.01 m
4.88(8.9/6.9)
5a
5b
6a
6b
a
157.94
159.04
0.83(5.25)
0.39(5.1)
2.60
–
–
153.50(156.5)
155.05(155.5)
5.72–5.78 m
5.75–5.82 m
1.62(3.6)
1.63(3.6)
–
d (3-CF3-Ar): 1d 14,97 s; 4d 15,13 s.
b
c
d (C6F5) 1e: ꢁ80,48 to ꢁ80,70 (m, p-F), ꢁ68,64 to ꢁ68,80 (m, o-F), ꢁ88,01 to ꢁ88,26 (m, m-F); 4e: ꢁ83,92 to ꢁ84,11 (m, p-F), ꢁ76,08 to ꢁ76,25 (m, o-F), ꢁ63,64 to ꢁ63,79 (m, m-F).
3
d (CH3-CH) 1,55 d JH–H = 6,8 Hz.

3460
I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
Fig. 1. Perspective view of 4d with the representation of atoms by the
thermal ellipsoids at 50% probability level. The selected bond lengths
˚
(A): Rh(1)–P(1) 2.259(1), Rh(1)–Cl(1) 2.440(1), Rh(1)–Cl(2) 2.439(1),
Fig. 2. Perspective view for the one of two independent molecules in
complex 6a with the representation of atoms by the thermal ellipsoids
at 50% probability level.
Rh(1)–Cpcent 1.855(6), P(1)–O(1) 1.627(3), P(1)–O(2) 1.655(4), P(1)–
O(3) 1.654(4), O(1)–C(1) 1.442(5), O(2)–C(10) 1.424(7), O(3)–C(11)
1.410(7); Bond angles (ꢀ): P(1)–Rh(1)–Cl(2) 89.48(6), P(1)–Rh(1)–Cl(1)
87.44(6), Cl(2)–Rh(1)–Cl(1) 94.82(5), O(1)–P(1)–O(3) 104.1(2), O(1)–
P(1)–O(2) 98.2(2), O(3)–P(1)–O(2) 96.07(15), O(1)–P(1)–Rh(1)
122.15(12), O(3)–P(1)–Rh(1) 115.35(15), O(2)–P(1)–Rh(1) 116.69(16),
C(1)–O(1)–P(1) 126.3(3).
drawing 1,3,2-benzdioxaphospholene moiety, as it is in
4d, resulted in the shortening of the Rh–P bond
˚
(2.284(1) versus 2.259(1) A) and the P–O(1) bond
˚
(1.641(3) versus 1.627(3) A), as well as the elongation
˚
of the Rh–Cl bonds (2.404(1) versus 2.440(1) A).
In a similar manner, the interaction of phosphite 1a
with Rh(CO)2(acac) resulted in the formation of
Rh(L)(CO)(acac) complex (L = 1a) that was character-
characterized by a distorted P-pyramid with angles for
O(1)P(1)O(2) and O(1)P(1)O(3) reduced up to 96.1(1)–
98.2(2)ꢀ and with an increased angle for Rh(1)P(1)O(1)
(122.1(1)ꢀ). The dioxaphospholene ring possesses an
envelope conformation with a deviation of the P(1)
ized by its 31P NMR spectra (dP 126.1, JP–Rh(I)
=
˚
atom by 0.25 A. The benzodioxaphospholene ring is al-
241.0 Hz). Comparing with the RhIII complex contain-
ing the same ligand (4a), the smaller dP shift may suggest
a weaker coordination bond with RhI. This complex is
stable in solution under argon, but decomposed during
the isolation attempted.
most coplanar with the Cp* ligand with an angle of 8ꢀ.
The analysis of intramolecular contacts have revealed
that the rather acidic hydrogen atom at C(1) participates
in the bifurcated RhCl2ꢂ ꢂ ꢂH contact with Clꢂ ꢂ ꢂH dis-
˚
tances equal to 2.69 and 2.62 A (for Cl(1) and Cl(2)
atom, respectively). It should be noted that from geo-
metrical point of view these contacts are comparable
2.2. Hydroformylation in the presence of rhodium–
phosphite catalysts
˚
with the intermolecular Hꢂ ꢂ ꢂCl ones (2.55–2.68 A)
formed with a solvate CH2Cl2 molecule. Therefore, it
is suggested that the mutual disposition of RhCl2 and
C(CF3)(C6H4CF3) fragments is controlled not only by
anomeric interactions of the O(1) electron lone pairs
with Rh(1)–P(1) and P(1)–O(2)/P(1)–O(3) antibonding
orbitals, but by intramolecular Clꢂ ꢂ ꢂH contacts as well.
The X-ray analysis of 6a reported earlier in our pre-
liminary communication [24] has revealed that it crystal-
Styrene as the model substrate was reacted with CO/H2
(1/1) at 40 or 100 ꢀC at 100 bar in the presence of
Rh(L)(Cp*)Cl2 4a,c–e, 6a (where L represents 1a, c–e
and 5a, respectively) ꢀpreformedꢁ or rhodium–L in situ
catalysts (Fig. 3) The phosphorus ligands applied differ
significantly both in electronic and in steric properties.
It is worth noting, that the Taft constants (rI) for CF3
and C6F5 groups are 0.32–0.41 [25] and 0.31 [26], respec-
tively, so ligands 1a and 1e are rather similar both from
electronic and steric point of view.
In addition to the formation of the two formyl regioi-
somers, 2-phenyl-propanal (A) and 3-phenyl-propanal
(B), that of ethylbenzene (C) arising from hydrogenation
was also expected (Scheme 3).
All ꢀpreformedꢁ catalysts (4a, 4d, 4e and 6a) were ac-
tive at 100 ꢀC under the given conditions. Practically
lizes in the chiral space group (P212121) giving the
D
enantiopure crystals (½aꢃ ꢄ 900ꢅ (c 1, CHCl3)) (Fig. 2).
20
Therefore, the crystallization of the 6a leads to the
spontaneous resolution of enantiomers and using fur-
ther the latter as seeds may allow obtaining the chiral
catalyst from a racemic solution. In this study of cata-
lytic activity, the racemic compound was, however,
used. It should be noted that the exchange of two elec-
tron-releasing ethyl groups in 6a for the electron-with-

I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
3461
O
O
P
O
P
P
O
O
O
F3C
F3C
Cl
F
1a
1c
O
O
F
F
O
O
P
O
F
O
F3C
H3C
CF3
1e
1d
F
Et
Et
P
O
F3C
5a
Fig. 3. Ligands used for catalytic experiments either in ꢀpreformedꢁ or in in situ systems.
CO/H2
PhCH CH2
PhCH(CHO)CH3 + PhCH2CH2CHO + PhCH2CH3
A
B
C
Scheme 3.
complete conversions have been obtained in most cases
in up to 3 h (i.e., TOF values up to 2000) except for the
diethylphosphinite derived catalyst 6a. Decreased activ-
ity was observed with the complex of phosphinite ligand
having donor ethyl groups at the phosphorus atom (6a).
The reaction was highly chemoselective towards the
formation of aldehydes. (The chemoselectivity of the hyd-
ro- formylation was higher than 96% in all cases.) The
regioselectivities towards A fall in the range of 44–54%,
i.e., nearly a 1:1 mixture of branched (A) and linear (B)
aldehydes has been obtained.
Contrary to the results obtained with the ꢀpreformedꢁ
catalysts, the in situ formed rhodium catalysts proved to
be active even at 40 ꢀC. At elevated reaction times (up to
72 h) the hydroformylation resulted in practically com-
plete conversions both with 1a and 1d (Table 3).
Excellent chemoselectivities (above 98%) have been
obtained in all cases both at 40 and 100 ꢀC except for
the 3-chlorophenyl ligand 1c at 40 ꢀC (79%). The che-
moselectivities are higher than those obtained with pre-
formed catalysts and varied from 98% to 100%.
The regioselectivities show strong temperature depen-
dence. At 40 ꢀC, the branched aldehyde A predominated
that was formed in excellent regioselectivities (95% in all
three cases). The regioselectivities observed at 100 ꢀC
show a good agreement with those detected with pre-
formed systems involving 1a. Slightly higher branched
selectivity (60%) has been obtained with 1d.
According to a generally accepted mechanism, the
low activity of the ꢀpreformedꢁ catalysts at 40 ꢀC can
be interpreted as follows: (i) The reduction of the RhIII
precursor to RhI species, needed for the complexation of
Table 3
Hydroformylation of styrene in the presence of in situ catalysts formed from Rh(CO)2(acac) and 2L (L = 1a, 1c and 1d)a
Conv.b (%)
TOF (hꢁ1
)
c
RC (%)
d
Rbr (%)
Run
Ligand
Temperature (ꢀC)
R. time (h)
1
2
3
4
5
6
a
1a
1a
1d
1d
1c
1c
40
72
2
100
100
100
100
47.5
100
100
95.8
56.9
95.4
60.3
95.1
57.3
>55
>2000
>55
100
40
97.9
100
72
2
100
40
98.5
79.4
98.1
>2000
>26
>2000
72
2
100
Reaction conditions: p = 100 bar (CO:H2 = 1:1); [Rh]/[styrene] = 1/4000; solvent: toluene.
Determined by GC.
b
c
Chemoselectivity (A + B)/(A + B + C) · 100.
Regioselectivity towards branched aldehyde A/(A + B) · 100.
d

3462
I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
olefin and carbon monoxide, as well as oxidative addi-
tion of hydrogen (which is probably the slowest step in
the whole process [14]), is difficult at low temperature.
(ii) Although the Rh(L)(Cp*)HCl complex formed in
hydrogenolysis should release hydrogen chloride in
reductive elimination in order to form a coordinatively
unsaturated complex containing soft RhI centre. In the
absence of a base it is not a facile reaction. The
Rh(L)*(Cp)HnCl2ꢁn (n = 1,2) type RhIII complexes are
not ready to activate the olefin even if a vacant coordi-
nation site is available.
The increased activity of the in situ system formed
from Rh(CO)2(acac) precursor and the corresponding
phosphite ligand is due to the facile formation of RhI
catalytic intermediates. The Rh(L)2(CO)(H) key-inter-
mediate is easily available in the in situ system and could
be formed even at low temperature [27]. Furthermore,
an alternative catalytic route involving the direct forma-
tion of catalytically active ꢀP-ligand-freeꢁ hydrido-
carbonyl-rhodium species cannot be excluded.
Considering the ligands used in in situ systems, there
is a sharp difference between the 3-chlorophenyl-con-
taining (1c) and the other two phosphites (1a and 1d)
at 40 ꢀC. Namely, the chemoselectivity towards hydro-
formylation is definitely lower in the presence of 1c,
accompanied by the decrease in catalytic activity. There-
fore, different species have to be operative in hydro-
formylation and further intermediates have to be
involved in the catalytic cycle. The activation of the
aryl-chloro bond by RhI species (even in small extent)
might be resulted in the formation of rhodium-chloro
species. Substantially, there are two possibilities for the
formation of catalytically active species: (i) hydrogenol-
ysis of the Rh–Cl bond yielded monohydrido-rhodium
species, (ii) activation of dihydrogen in oxidative addi-
tion results in RhIII species containing Rh(H)2Cl moiety
which should undergo reductive elimination by loosing
HCl (facile elimination in the presence of an amine but
not in the absence of that).
standard drying agents. Toluene was distilled under ar-
gon from sodium in the presence of benzophenone. Sty-
rene was freshly distilled before use. The NMR spectra
were recorded on a ꢀBruker AMX-4000 spectrometer in
CDCl3 solutions using residual proton signal or the
characteristic 13C chemical shift of the deutero solvent
as an internal standard (1H or 13C, respectively) and
85% H3PO4 (31P) as an external standard. The 19F chem-
ical shifts were determined with trifluoroacetic acid (d
19F 76.53 ppm) as an external standard. GLC analyses
were carried out with a HP-5890/II gas chromatograph
using a 15 m HP-5 column (temperature program: initial
temperature: 200 ꢀC (2 min), rate: 10 ꢀC/min, final tem-
perature: 300 ꢀC).
The starting 1-(2,3,4,5,6-pentafluorophenyl)-1-etha-
nol was purchased from Aldrich and used in synthesis
without purification. a-Trifluoromethylbenzyl alcohols
were synthesised by reduction of the correspond-
ing ketones according the procedure developed by us
earlier [28]. 2-chloro-1,3,2-benzodioxaphospholene was
obtained by known procedure [29]. For the synthesis
of ligand 5a and its RhIII complex 6a, see [24].
3.2. Synthesis of the P-ligands
3.2.1. 2-(2,2,2-Trifluoro-1-phenylethoxy)-1,3,2-
benzodioxaphospholene 1a
To a cooled (ꢁ5ꢀ to 0ꢀ) solution of the equimolar
mixture of pyridine (0.65 g, 8.2 mmol) and ( )-2,2,2-tri-
fluoro-1-phenyl-1-ethanol (1.44 g, 8.2 mmol) in a mix-
ture solvent (hexane:ether; 4:1; 100 ml), under argon
and with stirring, 2-chloro-1,3,2-benzodioxaphosphole
(1.43 g, 8.2 mmol) was added dropwise at the same tem-
perature over 0.5 h. The mixture was allowed to warm
to room temperature and stirred for additional 3 h.
The precipitate was filtered off under argon, washed
with Et2O (2 · 30 ml) and discarded. The filtrate was
evaporated in vacuo yielding crude ligand 1a as an oil
1
(2.56 g), in a purity of ca. 98% according 31P and H
In conclusion, we obtained a series of novel fluori-
nated phosphite P-ligands and their RhIII complexes.
Representative examples of the prepared complexes or
in situ systems formed from Rh(CO)2(acac) precursor
and the corresponding P-ligand have been tested in the
hydroformylation of styrene demonstrating high cata-
lytic activity.
NMR. For purification, hexane (10 ml) was added to
the crude oil and the mixture was chilled to ꢁ20 ꢀC.
After decanting the hexane phase and concentration in
vacuo, 1a was isolated in a pure form as colourless oil
(2.27 g, 88%).
2-[1-(4-Chlorophenyl)-2,2,2-trifluoroethoxy]-1,3,2-ben-
zodioxaphospholene 1b, 2-[1-(3-chlorophenyl)-2,2,2-tri-
fluoroethoxy]-1,3,2-benzodioxaphospholene 1c, 2-[2,2,
2-trifluoro-1-[3-(trifluoromethyl)phenyl]ethoxy]-1,3,2-
benzodioxaphospholene 1d, and 2-[1-(2,3,4,5,6-penta-
fluorophenyl)ethoxy]-1,3,2-benzodioxaphospholene 1e
were obtained similarly. All of the ligands were purified
as above. Table 1 shows the yields, melting points and
analytical data the ligands 1a–e obtained, NMR data
are summarized in Table 2. In principle, the crude prod-
ucts may be used in the complexation reactions without
further purification.
3. Experimental
3.1. General
All reactions were conducted under an inert atmo-
sphere of dry argon by using Schlenk glassware and vac-
uum line techniques. Solvents were freshly distilled from

I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
3463
3.2.2. 2-[2,2,2-Trifluoro-1-(4-chlorophenylethyl)]
diethylphosphinite 5b
3.5. Hydroformylation experiments
Equimolar mixture of diethyl-(N,N-diethylamino)-
phosphinite (1.10 g, 6.84 mmol) and ( )-2,2,2-trifluoro-
1-(4-chlorophenyl)ethanol (1.44 g, 6.84 mmol) was feed
into the distillation flask that was heated up to 155 ꢀC
and maintained under these conditions over 45 min.
During that time, the diethylamine formed distilled off.
The residue was distilled in vacuo yielding 1.75 g
(86%) of 5b; b.p. 100–102 ꢀC/0.5 torr. Analytical data
were obtained for the corresponding sulphide derivative
obtained by the addition of sulphur to the compound in
benzene solution after purification by column chroma-
tography (see Tables 1 and 2).
In a typical experiment a solution of 0.00625 mmol of
preformed catalyst 4a,d,e, or 6a in 7.5 ml toluene con-
taining 2.6 ml (25 mmol) of styrene was transferred un-
der argon into a 20 ml stainless steel autoclave. (In
case of the in situ catalysts 1.6 mg (0.00625 mmol)
Rh(CO)2(acac) precursor and 0.0125 mmol of 1a (or
1c or 1d) were dissolved in 7.5 ml toluene, 25 mmol sty-
rene was added and the homogeneous solution was
transferred under argon into a stainless steel autoclave.)
The reaction vessel was pressurized to 100 bar total pres-
sure (CO/H2 = 1/1) and placed in an oil bath and the
mixture was stirred with a magnetic stirrer for the
appropriate reaction time. The pressure was monitored
throughout the reaction. After cooling and venting of
the autoclave, the pale yellow solution was removed
and immediately analysed by GC.
3.3. Synthesis of the (p-pentamethylcyclopentadienyl)
(phosphite)-or(phosphinite)rhodium dichlorides 4a–e, 6b
(typical procedure)
The solution of [RhCp*Cl2]2 (128 mg, 0.207 mmol) in
3 ml of CH2Cl2 was added to a solution of either phos-
phite 1a–e or phosphinite 5b (0.407 mmol) in 7 ml of
CH2Cl2. After stirring the immediately formed red solu-
tion for 1 h at room temperature, the solvent was evap-
orated in vacuo up to the volume of ca. 0.5 ml and then
about 4 ml of pentane was added. The crystal precipi-
tated (dark red prisms) was filtered off and dried in
vacuo to afford the desired complexes (Tables 1 and 2).
Acknowledgements
This work was partially supported by Russian Basic
Research Foundation (Grant No. 05-03-32692), the
Federal Program on Support of Leading Scientific
Schools (Grant No. SS-1100.2003.3) and German Sci-
ence Foundation 436 RUS 113/766/1-1 as well as the
Hungarian Research Foundation (OTKA TS 44800
and T 042479).
3.4. X-ray crystallography
References
Crystals of 4d suitable for X-ray diffraction
were grown by slow evaporation of CH2Cl2 from the
solution of the complex. At 120 K crystals of 4d
(C26H26Cl4F6O3PRh) are orthorhombic, space group
[1] H. Brunner, W. Zettlmeier, Handbook of enantioselective catal-
ysis with transition metal compounds, VCH, Weinheim, 1993.
[2] C.D. Frohning, C.W. Kohlpainter, in: B. Cornils, W.A. Herr-
mann (Eds.), Applied homogeneous catalysis with organometallic
compounds, vol. 1, VCH, Weinheim, 1996, p. 29.
˚
Pca21, a = 23.347(8), b = 10.913(4), c = 12.383(4) A,
,
3
V = 3155(2) A , Z = 4, M = 776.15, dcalc = 1.634 g cmꢁ3
˚
l(Mo Ka) = 9.93 mꢁ1, F(000) = 1552. Intensities of
33209 reflections were measured with a Smart 1000
´
´
[3] G. Keglevich, T. Kegl, T. Chuluunbaatar, B. Dajka, P. Matyus,
´
B. Balogh, L. Kollar, J. Mol. Catal. A: Chemical 200 (2003) 131
(and references cited therein).
˚
CCD diffractometer at 120 K (k(Mo Ka) = 0.71072 A,
[4] (a) I. Ojima, C-Y. Tsai, M. Tzamarioudaki, D. Bonafoux, in:
L.E. Overman et al. (Eds.), The hydroformylation reaction in
organic reactions, vol. 56, Wiley, New York, 2000, pp. 1–354
(Chapter 1);
2h < 58ꢀ), and 8319 independent reflections (Rint
=
0.0550) were used in the further refinement. The struc-
ture was solved by direct method and refined by the
full-matrix least-squares technique against F2 in the
anisotropic–isotropic approximation. The refinement
converged to wR2 = 0.0935 and GOF = 1.062 for all
independent reflections (R1 = 0.0468 was calculated
against F for 6330 observed reflections with I > 2r
(I)). All calculations were performed using SHELXTL
PLUS 5.0 [30]. Crystallographic data (excluding struc-
ture factors) for the structure of 4d have been deposited
to the Cambridge Crystallographic Data Centre; No.
CCDC-262146. Copies of the data can be obtained free
of charge on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ UK (Fax: (internat.) +44-1223/
336-033; deposit@ccdc.cam.ac.uk).
(b) A. Alexakis, C. Benhaim, Eur. J. Org. Chem. (2002) 3221.
[5] (a) W. Chen, Y. Xu, S. Liao, J. Mol. Catal. 88 (1994) 277;
´
(b) A.M. Trzeciak, J.J. Ziolkowski, R. Choukroun, J. Organomet.
Chem. 525 (1996) 145;
(c) A. Cerotti, L. Garlaschelli, G. Longoni, M.C. Malatesta, D.
Strumolo, A. Fumagalli, S. Martinengo, J. Mol. Catal. 24 (1984)
309;
(d) B. Breit, S.K. Zahn, Tetrahedron Lett. 39 (1998) 1901;
(e) B. Breit, Angew. Chem. Int. Ed. Engl. (1996) 2835.
[6] (a) C. Claver, Ph. Kalck, M. Ridmy, A. Thorez, L.A. Oro, M.T.
Pinillos, M.C. Apreda, F.H. Cano, C. Foces-Foces, J. Chem. Soc.
Dalton Trans. (1988) 1523;
(b) S.N. Poelsma, P.M. Maitilis, J. Organomet. Chem. C15 (1993)
451;
(c) C.C. Cumbo, K.K. Bhatia, US Patent 4,024,197, 1977;
C.C. Cumbo, K.K. Bhatia, Chem. Abstr. 87 (1977) 52753.

3464
I. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 3456–3464
[7] D.J. Bergmann, E.M. Campi, W.E. Jackson, Q.J. McCubbin,
A.F. Patti, Tetrahedron Lett. 53 (1997) 43158.
[15] (a) J.E. Babin, G.T. Whiteker, US Patent, US 5,360,938, 1994;
J.E. Babin, G.T. Whiteker, Chem Abstr. 122 (1994) 186609;
(b) G.J.H. Buisman, L.A. van der Veen, A. Klootwijk, W.G.J.
Delang, P.C.J. Kamer, P.W.N.M. vanLeeuwen, D. Vogt, Organo-
metallics 16 (1997) 2929;
´
´
[8] (a) A. Polo, C. Claver, S. Castillon, A. Rulz, J.C. Bayon, J. Real,
C. Mealli, D. Masi, Organometallics 11 (1992) 3525;
(b) B. Breit, R. Winde, K. Harms, J. Chem. Soc. Perkin Trans. 1
(1997) 2681;
´ ´ ´ ´
ˆ
(c) S. Cserepi-Szucs, I. Toth, L. Parkanyi, J. Bakos, Tetrahedron:
´ ´
(c) E. Fernandez, A. Ruiz, C. Claver, S. Castillon, Organomet-
allics 17 (1998) 2857.
Asymmetry 9 (1998) 3135.
[16] G.J.H. Buisman, P.C.J. Kamer, P.W.N.M. van Leeuwen, Tetra-
hedron: Asymmetry 4 (1993) 1625.
[9] (a) A. van Rooy, J.N.H. de Bruijn, K.F. Roobeek, P.C.J. Kamer,
P.W.N.M. Vanleeuwen, J. Organomet. Chem. 507 (1996) 69;
(b) T. Omatsu, T. Onishi, Y. Tokitoh, Eur. Pat. Appl., EP
747,373, 1996;
[17] M. Dieguez, O. Pamies, C. Claver, Tetrahedron: Asymmetry 15
(2004) 2113.
[18] A.P. Dobbs, M.R. Kimberley, J. Fluorine Chem. 118 (2002) 3.
´ ´
[19] I.T. Horvath, J. Rabai, Science 266 (1994) 72.
´
[20] J.J.J. Juliette, I.T. Horvath, J.A. Gladysz, Angew. Chem. Int. Ed.
(c) T. Omatsu, T. Onishi, Y. Tokitoh, Chem. Abstr. 126 (1997)
74739;
(d) J.R. Briggs, D.L. Packett, D.R. Bryant, A.G. Phillips, D.J.
Schreck, K.D. Olson, E.B. Tjaden, A.S. Guram, T.C. Eisensch-
midt, E.S. Bragaham, PCT Int. Appl. WO 97,40,001, 1996;
(e) J.R. Briggs, D.L. Packett, D.R. Bryant, A.G. Phillips, D.J.
Schreck, K.D. Olson, E.B. Tjaden, A.S. Guram, T.C. Eisensch-
midt, E.S. Bragaham, Chem. Abstr. 128 (1998) 14353.
[10] P.W.N.M. van Leeuwen, C. Claver (Eds.), Rhodium catalyzed
hydroformylation, Kluwer, Dordrecht, 2000.
Engl. 36 (1997) 1610;
J.J.J. Juliette, D. Rutherford, I.T. Horvath, J.A. Gladysz, J. Am.
´
Chem. Soc. 121 (1999) 2696.
[21] B. Betzemeier, P. Knochel, Angew. Chem. Int. Ed. Engl. 36 (1997)
2623.
[22] P.W.N.M. van Leeuwen, C.F. Roobeek, Tetrahedron 37 (1981)
1973.
[23] D. Dakternieks, R. Schmutzler, G.-V. Ro¨schenthaler, J. Fluorine
Chem. 11 (1978) 387.
[11] (a) K. Nozaki, N. Sakai, T. Nanno, T. Higashijima, S. Mano, T.
Horiuchi, H. Takaya, J. Am. Chem. Soc. 119 (1997) 4413;
(b) T. Nanno, N. Sakai, K. Nozaki, H. Takaya, Tetrahedron:
Asymmetry 6 (1995) 2583;
[24] I.L. Odinets, O.I. Artyushin, E.V. Sharova, E.I. Goryunov, D.G.
Golovanov, K.A. Lyssenko, P.V. Petrovskii, T.A. Mastryukova,
Mendeleev Comm. 13 (2003) 102.
(c) T. Horiuchi, T. Ohta, E. Shirakawa, K. Nozaki, H. Takaya,
Tetrahedron 53 (1997) 7795.
[25] R.W. Taft, J. Phys. Chem. 64 (1960) 1805.
[26] M. Charton, Progr. Phys. Org. Chem. 13 (1981) 119.
[27] C.K. Brown, G. Wilkinson, J. Chem. Soc. A (1970) 2753.
[28] L.S. Zacharov, E.I. Goryunov, S.T. Ioffe, L.L. Morozov, T.M.
Scherbina, M.I. Kabachnik, Izv AN USSR Ser. Khim. (1976)
1834.
[12] K. Nozaki, Y. Itoi, F. Shibahara, E. Shirakawa, T. Ohta, H.
Takaya, T. Hiyama, J. Am. Chem. Soc. 120 (1998) 4051.
[13] (a) N. Sakai, K. Nozaki, K. Mashima, H. Takaya, Tetrahedron:
Asymmetry 3 (1992) 583;
´
ˆ
(b) S. Cserepi-Szucs, G. Huttner, L. Zsolnai, J. Bakos, J.
[29] P.C. Crofts, J.H.H. Markes, N.D. Rygon, J. Chem. Soc. (1958)
4250.
Organomet. Chem. 586 (1999) 70.
[14] G.J.H. Buisman, E.J. Vos, P.C.J. Kamer, P.W.N.M. van Leeu-
wen, J. Chem.Soc.Dalton Trans (1995) 409.
[30] G.M. Sheldrick, SHELXTL Version 5, Software Reference Manual,
Siemens Industrial Automation, Madison, WI, 1994.