Rhodium(III) and palladium(II) complexes of some P-heterocycles; synthesis and X-ray structures

Article abstract of DOI:10.1016/j.jorganchem.2004.08.051

Deoxygenation of the syn-3-phosphabicyclo[3.1.0]hexane 3-oxides bearing a 3-phenyl or a 3-(4-methylphenyl) substituent (1a,b) by trichlorosilane took place already at mild condition and resulted in the corresponding phosphines (2a,b) with retention of configuration at phosphorus, while in the case of 3-(2-methylphenyl)-3-phosphabicyclo[3.1.0]hexane (2c), the inversion of the phosphorus atom was observed in solution under ambient conditions that was evaluated by quantum chemical calculations. A further phosphine ligand (5) was obtained by the reduction of 4-dichloromethylene-1,4-dihydrophosphinine oxide (4). The phosphine ligands (2 and 5) were used in the preparation of Rh(III) complexes (3 and 6). A Pd(II) complex of type PdCl₂(5)₂ (7) was also prepared. The stereostructures of a series of Rh(III) complexes of 3-aryl-3-phosphabicyclo[3.1.0]hexanes (3b-syn, 3c-syn and 3c-anti) were elucidated by single crystal X-ray analysis confirming the relative position of the dichlorocyclopropane and the P-substituent.

Full text of DOI:10.1016/j.jorganchem.2004.08.051

Journal of Organometallic Chemistry 690 (2005) 704–712
www.elsevier.com/locate/jorganchem
Rhodium(III) and palladium(II) complexes of some
P-heterocycles; synthesis and X-ray structures
a,
a
a
Irina L. Odinets *, Natalya M. Vinogradova , Konstantin A. Lyssenko ,
a
a
a
Denis G. Golovanov , Pavel V. Petrovskii , Tatyana A. Mastryukova ,
b
Helga Szelke , Nora Balazsdi Szabo , Gyo¨rgy Keglevich
b
b
´
´
´
a
Thiophosphorus Laboratory, A.N.Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences,
Vavilova str., 28, 119991 Moscow, Russian Federation
b
Department of Organic Chemical Technology, Budapest University of Technology and Economics, H-1521 Budapest, Hungary
Received 6 April 2004; accepted 25 August 2004
Abstract
Deoxygenation of the syn-3-phosphabicyclo[3.1.0]hexane 3-oxides bearing a 3-phenyl or a 3-(4-methylphenyl) substituent (1a,b)
by trichlorosilane took place already at mild condition and resulted in the corresponding phosphines (2a,b) with retention of con-
figuration at phosphorus, while in the case of 3-(2-methylphenyl)-3-phosphabicyclo[3.1.0]hexane (2c), the inversion of the phospho-
rus atom was observed in solution under ambient conditions that was evaluated by quantum chemical calculations. A further
phosphine ligand (5) was obtained by the reduction of 4-dichloromethylene-1,4-dihydrophosphinine oxide (4). The phosphine lig-
ands (2 and 5) were used in the preparation of Rh(III) complexes (3 and 6). A Pd(II) complex of type PdCl2(5)2 (7) was also pre-
pared. The stereostructures of a series of Rh(III) complexes of 3-aryl-3-phosphabicyclo[3.1.0]hexanes (3b-syn, 3c-syn and 3c-anti)
were elucidated by single crystal X-ray analysis confirming the relative position of the dichlorocyclopropane and the P-substituent.
ꢀ 2004 Elsevier B.V. All rights reserved.
Keywords: Cyclic phosphine oxides; Deoxygenation; Cyclic phosphines; Rhodium and palladium complexes; Stereostructure
1. Introduction
available a valuable pool for novel P-ligands that have
never been prepared and studied [3,4].
Transition metal complexes incorporating phos-
phines and diphosphines as the ligands are frequently
used catalysts in hydroformylations and hydrogenations
[1]. Many catalysts involve phosphole or phosphole-
based ligands, the most pregnant example of which is
BIPNOR [2]. One of us developed a simple method
for the synthesis of a variety of six-membered P-hetero-
cycles, such as 3-phosphabicyclo[3.1.0]hexane 3-oxides
and 1,2- or 1,4-dihydrophosphinine oxides making thus
For the preparation of phosphines, the reduction of
the corresponding phosphine oxides by silane hydrides
[5] is the best choice due to the selectivity and efficiency.
This simple methodology was applied to obtain a series
of heterocyclic phosphine ligands referred also as k3-
phosphacyclanes in order to prepare their Rh(III) and
Pd(II) complexes and to introduce them for tests in cat-
alytic reactions.
In this paper, our interesting findings on the deoxy-
genation of the phosphine precursors and on the stereo-
structure of the P-ligands, as well as the transition metal
(Rh(III) and Pd(II)) complexes based on the cyclic phos-
phines are discussed. The results of the catalytic activity
*
Corresponding author. Tel.: +7 095 135 9356; fax: +7 095 135
5085.
E-mail address: odinets@ineos.ac.ru (I.L. Odinets).
0022-328X/$ - see front matter ꢀ 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.08.051

I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
705
of a variety of such complexes in hydroformylation and
hydroalkoxycarbonylation are described elsewhere [6].
from k5 to k3 phosphorus derivatives, the usual decrease
in JPC coupling costants could be observed.
The interaction of phosphines 2a,b with dimeric pen-
tamethylcyclopentadienyl rhodium dichloride resulted
in the typical formation of complexes of type
CP*RhCl2L (3a,b) in high yield. The structure of com-
plex 3b was independently confirmed by single crystal
X-ray analysis suggesting the cis disposition of the P-
p-tolyl substituent and the cyclopropane ring (3b-syn)
(Fig. 1). A comparison of the spectral features of com-
plexes 3a and 3b allowed us to conclude that complex
3a also has the syn-geometry. Thus the reduction of
phosphines 1a,b proceeds with the retention of
configuration.
2. Results and discussion
2.1. Synthesis of the 3-phosphabicyclo[3.1.0]hexane
ligands and their RhIII complexes
Earlier, one of us reported on the facile synthesis of
6,6-dichloro-1-methyl-3-phenyl-3-phosphabicyclo[3.1.0]
hexane 3-oxide (1a) by dichlorocyclopropanation of the
corresponding 2,5-dihydrophosphole oxide under phase
transfer catalytic conditions [3,7]. In contrast to the sim-
ilar P-alkoxy derivatives [3,8], dichlorocarbene adduct
1a was obtained as a single isomer with syn disposition
of the cyclopropane ring and the P-phenyl substituent
[9]. The aryl-analogues (1b,c) were obtained by a similar
procedure, but the o-tolyl derivative (1c) was formed as
a 6:4 mixture of syn and anti isomers, which were
separated by chromatography [10]. The structure of
the syn-isomer 1c was independently confirmed by X-
ray crystallography [10]. For further reduction proce-
dure to be discussed in this paper, the pure syn-isomers
1a–c were utilised. We wished to prepare the rho-
dium(III) and the palladium(II) complexes of the novel
ligands and to study their peculiarities.
It was shown that the deoxygenation of phosphine
oxides 1a,b with trichlorosilane proceeded very easily
even at 0 ꢁC in the 1:1 mixture of dichloromethane
and benzene to afford the desired phosphine 2a,b as
the only reaction product (Scheme 1). It should be noted
that at a higher temperature (even at ambient condi-
tions), the formation of a variety of side-products (up
to 45–50% in total) could be observed.
At the same time, a similar reduction procedure used
for 1c possessing o-tolyl substituent at the phosphorus
atom led to two closely-related forms of product (2c),
having dP signals in the phosphine region (in CDCl3 at
22.6 and 27.9 ppm) and the same pattern of signals in
the 13C NMR spectra. The ratio of the components de-
pended on the reaction time and was 95:5 after the
standard procedure (2 h at 0 ꢁC followed by 2 h at 20
ꢁC) and changed to 38:62 after a 6 daysꢀ standing. It
was assumed to have the syn and anti diastereomers of
2c in hands (Scheme 2, showing in the first approach a
tentative assignment).
We note that in the 13C NMR spectra of phosphines
2c-syn and 2c-anti, the signal pattern is also close to that
3
of their precursor 1c, but the high JPC value for the o-
methyl substituent (15.5 and 23.1 Hz in 2c-syn and 2c-
anti isomers, respectively) should also be mentioned, in
comparison with the 3.5 Hz detected on the o-methyl
substituent of 1c.
On addition of the Rh(III) precursor to the mixture
of phosphine isomers (2c-syn and 2c-anti) in different
times, the two different Rh(III) complexes (3c-syn and
3c-anti) were formed in the same ratio observed for
the phosphines 2c-syn and 2c-anti used in the complexa-
tion reaction. Both phosphine 2c-syn and complex 3c-
syn revealed upfield 31P NMR chemical shifts, as
compared with those for the other isomers (2c-anti and
3c-anti). The stereostructure of both isomers of complex
3c was confirmed by X-ray analysis justifying the
Phosphines 2a,b have upfield dP shifts of ca. 25 ppm
in CDCl3 as compared with those of the P-oxides 1a
and 1b (dP ca. 77 ppm). In the 13C NMR spectra of
2a,b, the signals of the skeletal carbon atoms can be
found in about the same regions as that for 1a,b. In
phosphines 2a,b, the signal of C(1) is shifted downfield
(Dd ꢀ4 ppm), while on the contrary, C(2) and C(4) are
shifted upfield, as compared with the oxides 1a,b. Going
Scheme 1.

706
I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
Fig. 2. The general view of 3c-syn in representation of atoms by
thermal motion ellipsoids at 50% probability level.
Fig. 1. The general view of 3b-syn in representation of atoms by
thermal motion ellipsoids at 50% probability level.
ppm). It is noted that the signals of the rhodium com-
plexes appeared as sharp doublets with characteristic
coupling constant (1JPRh) in the 31P NMR spectra. This
constant comprises a value of ca. 138 Hz for 138 Hz for
3(a–c)-syn and is a little bit higher (143.0 Hz) for 3c-anti.
In the 13C NMR spectra, the 1JPC coupling constants for
3a–c are approximately one and a half higher than for
2a–c, but twice as much lower than for 1a–c.
tentative assignment, the syn or the anti disposition of
the P-aryl substituent and the cyclopropane ring, both
in the complex (3c), and in the phosphine (2c) (Figs. 2
and 3). Hence, the assignment of the second signal at
dP 27.9 (or at dP 20.1 for a crude mixture) in the 31P
NMR spectrum of the mixture of phosphines to species
2c-anti was also confirmed.
The Rh(III) complexes 3a,b are characterized by dP
shifts that are close to the corresponding values of the
P-oxides 1a,b. For 3c, the complex with syn geometry
has an upfield chemical shift (dP 67.8 ppm), as compared
with that of the starting oxide 1c (dP 80.3 ppm), while
the signal for the 3c-anti is shifted downfield (dP 91.6
The geometry of the ligand in complexes 3b-syn and
3c-syn (Figs. 1 and 2) is practically identical having the
aryl ring and the dichlorocyclopropane moiety in cis
orientation (Table 1). The 5-membered hetero ring in
3b-syn and 3c-syn adopts an envelope conformation
˚
with a deviation of 0.205–0.344 A of the P(3) atom from
Scheme 2.

I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
707
tion of the phosphorus atom, the principal geometry
of complex 3c-anti (two independent molecules per unit
cell) is quite similar to the above mentioned syn ana-
logues (Table 1, Fig. 3). The main differences are re-
flected in the pronounced increase in the deviation of
the P(3) atom from the C(2)C(1)C(5)C(4) plane up to
˚
0.709 A, as well as in an increase in the
C(2)P(3)C(8)C(9) angle corresponding to the mutual
disposition of the aryl and the hetero rings.
It should be noted that complexes 3 form stable sol-
vates with chlorine containing solvents (CH2Cl2 or
CHCl3), which in some cases we failed to remove even
at prolonged drying in vacuo at an elevated tempera-
ture. Apparently, the solvates are stabilized by numer-
ous H. . .Cl interactions formed between the RhCl2
moiety and the dichloromethane or the chloroform mol-
ecules. Moreover, recrystallization of 3c-syn Æ 2CH2Cl2
from chloroform gave the corresponding pseudopoly-
morph modification with a chloroform solvate, which
structure was also confirmed by X-ray diffraction inves-
tigation (Table 1).
Fig. 3. The general view of one of independent molecules of 3c-anti in
representation of atoms by thermal motion ellipsoids at 50% proba-
bility level.
2.2. P-inversion in 6,6-dichloro-1-methyl-3-(2-methylphe-
nyl)-3-phosphabicyclo[3.1.0]hexane; quantum chemical
calculations
the C(2)C(1)C(5)C(4) plane towards the metal atom.
The dihedral angle between the ‘‘base’’ of envelope
and the cyclopropane moiety in both complexes is equal
to 70ꢁ. The angle between the Cp* and the aryl moiety is
slightly different and is equal to 23.6ꢁ and 18.9ꢁ for 3b-
syn and 3c-syn, respectively. The only difference ob-
served for 3b-syn and 3c-syn is the orientation of the
Cp* in respect to the RhCl2P moiety (see the
P(1)Rh(1)C(15)C(20) torsion angle).
It is known that trichlorosilane alone or more conven-
iently its pyridine complex reduces phosphine oxides
with retention of configuration. Only in the presence of
amines of base strength greater than about pKb 5, such
as triethylamine, the inversion was observed to take
place predominantly in silane reductions [11]. Thus,
one may assume that reduction of 3-phosphabicycloh-
exan-3-oxides 1a–c by trichlorosilane alone resulted in
the formation of the corresponding phosphine with
As it may be seen, the relative position of the P-aryl
substituent and the cyclopropane moiety in complex
3c-anti is trans. Despite the difference in the configura-
Table 1
˚
Selected bond lengths [A] and bond angles [ꢁ] of rhodium complexes 3b and 3c
3b-syn Æ CH2Cl2
3c-syn Æ 2CH2Cl2
3c-syn Æ 2CHCl3
3c-anti Æ 2CHCl3
Rh(1)–P(3)
P(3)–C(2)
P(3)–C(4)
P(3)–C(8)
2.283(1)
1.825(6)
1.834(5)
1.797(5)
1.802
2,3207(8)
1,848(3)
1.854(3)
1.828(3)
1.813
2.323(1)
1.844(4)
1.854(4)
1.808(5)
1.819
2.300(1)
1.834(4)
1.838(4)
1.826(4)
1.815
a
Rh–Cpcent
Rh–C(Cp)
Rh–Cl
2.143(5)–2.203(5)
2.393(1)–2.399(1)
130.4
2.155(3)–2.223(3)
2.402(1)–2.412(1)
124.1
2.152(4)–2.232(4)
2.408(1)–2.414(1)
132.6
2.146(4)–2.223(4)
2.383(1)–2.404(1)
131.7
P(3)Rh(1)Cp
Cl(1)Rh(1)Cl(2)
C(2)P(3)C(4)
90.90(5)
96.6(2)
0.205
88.75(3)
95.9(1)
0.266
88.93(4)
95.5(2)
0.344
87.29(4)
91.6(1)
0.79
b
dP(1)
c
dC(6)
P(3)Rh(1)C(15)C(20)
–1.19
–13.6
–1.22
8.3
–0.99
7.3
–1.18
7.4/–9.9
71.1
16.4
C(2)P(3)C(8)C(9)
Cp/Phd
56.9
23.6
61.3
18.9
59.5
22.0
a
Distance to centroid of the cyclopentadienyl ring.
b
c
Deviation of the P(3) atom from the C(2), C(3), C(4) and C(5) plane.
Deviation of the C(6) atom from the C(2), C(3), C(4) and C(5) plane.
Dihedral angle between the Ph and the cyclopentadienyl rings.
d

708
I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
preserved P-configuration that was followed by inversion
only in the case of compound 2c bearing an o-tolyl sub-
stituent. The ease of the inversion for 2c is a noteworthy
observation of us, as common phosphines are not known
to undergo P-inversion at 26 ꢁC. For example, the barrier
of inversion for a phospholane bearing a hydrogen atom
on the phosphorus atom is 39.4 kcal/mol [12].
Thus, our preliminary results based on simplified cal-
culations neglecting the specific and nonspecific solvata-
tions (that may have a considerable impact on the final
outcome) fit well the experimental observation that the
inversion proceeds more readily in the case of ortho-
methylphenyl substituted phosphine 2c than for that
with the P-phenyl phosphine (2a).
To consider such a possibility and to estimate the rea-
sons for the inversion observed only in the case of 2c,
quantum chemical calculations were carried out at the
at the B3PW91 DFT level of theory with the 6-31G* ba-
sis set for both the syn and the anti isomers of phos-
phines 2a and 2c. According to these data, the
corresponding anti isomers are thermodynamically more
stable in both cases. The difference in energy between the
two isomers is higher for the unsubstituted phosphine 2a
(1.79 kcal/mol) than that for 2c (0.83 kcal/mol). At the
same time, the inversion barrier is somewhat smaller
for 2c than that for 2a in the syn to anti direction
[DEa(2a) ꢁ Ea(2c) = 0.687 kcal/mol]. Although the dif-
ference is not significant, the tendency read from the en-
ergy values is in accord with the observation.
2.3. Preparation of a dihydrophosphinine ligand and its
complexation with Rh(III) and Pd(II)
The analogous series of reactions (reduction by tri-
chlorosilane followed by complexation) was applied to
4-dichloromethylene-1,4-dihydrophosphinine 1-oxide 4
[12]. In this case, the reduction and the subsequent com-
plexation with dimeric pentamethylcyclopentadienyl
rhodium dichloride furnished phosphine 5 and complex
6, respectively (Scheme 3).
The reaction of phosphine 5 with with trans-
dichloro(dibenzonitrile)palladium in benzene solution
at 0 ꢁC led to the formation of ML2 type complex 7.
At a higher temperature, the exothermic complexation
reaction proceeded to yield at least three different prod-
ucts which were not isolated and identified. Complex 7
precipitated from the dichloromethane solution as fine
yellow crystals after the addition of pentane. The struc-
ture of the complexes (6 and 7) were unambiguously
confirmed by NMR and MS data.
In the transition state for the inversion of 2a and 2c,
the planarization of the phosphorus atom leads to sig-
nificant shortening of the P–C01 bond lengths (where
C01 is the ipso carbon atom). The above mentioned bond
length is slightly different in transitions states of 2a and
˚
2c (1.772 and 1.782 A, respectively) that is the conse-
quence of some decrease of conjugation in 2c due to
the steric repulsion of the o-methyl substituent of the
Ph-ring and the methylene hydrogens of the a-carbon
atoms. Apparently such decrease of conjugation in 2c
intermediate planar state shifts the equilibrium to the
thermodynamically more preferred anti-isomer of
phosphine.
The 31P NMR signals for both the phosphine oxide 4
and the corresponding phosphine 5 are shifted upfield as
compared with those for the k5- and k3-phosphacyclanes
1 and 2, e.g., the dP for the latter one appears at –30.6, at
the same time, the dP for Rh(III) and Pd(II) complexes 6
and 7 can be found at ꢁ0.89 and ꢁ5.18, respectively. In
6, the 1JPRh coupling constant is approximately the same
Scheme 3.

I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
709
(142.1 Hz), as that in Rh(III) complexes 3. In the 13C
3. Experimental section
NMR spectra 5 and 6, the signal pattern is similar to
that of starting compound 4 with an expected decrease
in the JPC constants for the carbon atoms of heterocyclic
3.1. General
1
ring. The JPC coupling constant for 6 is approximately
six time higher than that is for 5, but is twice as much
lower than that is for 4 [13]. For phosphine 5, the in-
All reactions were conducted under an inert atmos-
phere of dry argon by using Schlenk glassware and vac-
uum line techniques. Solvents were freshly distilled from
standard drying agents. The NMR spectra were re-
corded on a ‘‘Bruker AMX-400’’ spectrometer in CDCl3
solutions using residual proton signals of deutero sol-
vent as an internal standard (1H, 13C) and 85% H3PO4
(31P) as an external standard. IR spectra were recorded
in KBr pellets on a Fourier-spectrometer ‘‘Magna-
IR750’’ (Nicolet) with a resolution of 2 cmꢁ1 after 128
scans. The mass-spectra were obtained on a Varian
MAT-311A spectrometer at 70 eV. The starting phos-
phine oxides used as precursors for the ligands were ob-
tained as described earlier [3,7,10,13].
2
3
crease in the JPC and in the JPC values for the o- and
the m-carbon atoms of the phenyl group that are even
1
higher than the JPC value for the ipso-carbon is also
noteworthy. This fact may be explained by the violation
of conjugation due to the difference in the mutual dispo-
sition of the unsaturated hetero ring and the phenyl
group in 5. For complex 6, the large 3JPC value detected
1
on C4 and on C7 atoms is to be mentioned. As the H
NMR spectra are concerned, the downfield shift for
the CH@ proton in complexes 6 and 7 along with the in-
crease in the JPH value seems to be interesting.
2
As regards the EI-mass spectra of all Rh(III) com-
plexes obtained (3a–c and 6), despite their stability un-
der usual conditions, they underwent immediate
rupture of the P–Rh bond resulting in two series of frag-
ments. Thus, the spectra revealed independent decom-
positions for the ligand part and for the metal
containing moiety. It should be mentioned that not only
the principle fragmentation patterns, but the intensities
for the rhodium containing moieties were quite similar
for compounds 3 and 6.
The complexes obtained have been tested as catalysts
in the hydroformylation and hydroalxoxycarbonylation
reaction of styrene. The chemoselectivity of hydro-
formylation was found to be excellent (higher than
99%) in all cases. A strong dependence of the regioselec-
tivity on the P-substituents was observed and the intro-
duction of a tolyl group instead of the phenyl one caused
an unexpectedly high increase of the catalytic activity
[6].
3.2. General procedure for the preparation of phosphines
2a–c and 5
To a cooled (0 ꢁC) solution of 0.3 mmol of precursor
1a–c or 4 in 10 ml of a 1:1 mixture of CH2Cl2–C6H6 was
added 0.15 ml (1.45 mmol) of HSiCl3 under argon
atmosphere and on stirring. The mixture was stirred at
the same temperature for 2 h and at ambient tempera-
ture for further 2 h. According to 31P NMR, the deoxy-
genation proceeded with high selectivity. Then the
volatile components were evaporated in vacuo to yield
the phosphines 2a–c and 5 quantitatively and in a pure
form according to 31P and 13C NMR (2c consisted of
a 95:5% mixture of the syn and the anti isomers). On
standing at room temperature for 6 days, the isomeric
mixture of 2c was converted to a 38:62% mixture of
the syn and the anti isomers.
The phosphines 2a–c and 5 were used in the subse-
quent complex formation without further purification.
Compound 2a-syn: 31P NMR (CDCl3) d 25.2, (dP
lit[14] 25.4); 13C NMR (CDCl3) d 21.2 (3JPC = 2.9, C1–
CH3), 27.0 (1JPC = 18.6, C4), 32.8 (1JPC = 17.6, C2),
40.5 (2JPC = 4.9, C5), 40.9 (C1), 72.6 (C6), 128.6
(2JPC = 7.8, C02), 129.5 (1JPC = 9.4 C01), 130.23
(3JPC = 13.5, C30 ), 132.3 (C04); FAB-MS, m/z 259
(M + H).
Compound 2b-syn: 31P NMR (CDCl3) d 24.5; 13C
NMR (CDCl3) d 21.2 (C1–CH3), 21.4 (Ar–CH3), 27.2
(1JPC = 11.8, C4), 33.1 (1JPC = 10.8, C2), 40.3 (C1), 40.8
(C5), 73.4 (C6), 126.02 (1JPC = 15.4, C10 ), 129.34 (2JPC
=7.2, C02), 131.22 (3JPC = 13.5, C30 ), 140.62 (C04); FAB-
MS, M+ found = 273.0342; C13H16Cl2P requires
273.0367 for the 35Cl isotopes.
Compound 2c-syn: 31P NMR (CDCl3) d 22.6 (95%);
13C NMR (CDCl3) d 21.6 (Ar–CH3), 21.6 (C1–CH3),
26.6 (1JPC = 3.7, C4), 37.7 (1JPC = 5.3, C2), 40.6 (C5),
40.6 (C1), 74.3 (3JPC = 7.2, C6), 125.64 (C05), 128.43
2.4. Concluding remarks
In summary, the reduction of 3-phosphabicy-
clo[3.1.0]hexane oxides and dihydrophosphinine oxides
by HSiCl3 was found to proceed easily at 0 ꢁC in con-
trast to the acyclic phosphine oxides that require reflux
at least in benzene solution [5]. In the case of 1-
methyl-3-(2-methylphenyl)-3-phosphabicyclo[3.1.0]hexane,
a novel inversion at the phosphorus atom was observed
at ambient conditions.
The new type of heterocyclic P-ligands were utilized
in the synthesis of Rh(III) and Pd(II) complexes. The
large Dd (>50 ppm) between 2 and 3 suggested a rather
strong coordination between the phosphorus and the
Rh(III) atoms in the above series. At the same time,
for rhodium(III) complex 6, the DdP of 29.7 suggested
only a weaker phosphorus–rhodium coordination, as
compared with that in complexes 3.

710
I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
(1JPC ꢀ7, C0 ), 128.49 (3JPC = 2.8, C0 ), 128.55 (C0 ),
272 ([Cp*RhCl ꢁ H]+, 10), 237 ([Cp*Rh ꢁ H]+, 18),
134 ([Cp* ꢁ H]+, 27). Anal. Calc. for C22H28Cl4RhP:
C, 46.51; H, 4.97; P, 5.45. Found: C, 46.16, 46.48; H,
5.21, 5.17; P, 5.80, 5.87%.
1
3
4
129.97 (2JPC = 4.0, C60 ), 139.71 (2JPC = 18.6, C02); FAB-
MS, M+ found = 273.0346, C13H16Cl2P requires
273.0367 for the 35Cl isotopes.
Compound 2c-anti: 31P NMR (CDCl3) d 27.9 (60%),
13C NMR (CDCl3) d 21.1 (C1–CH3), 21.3 (3JPC = 23.1,
Ar–CH3), 37.4 (1JPC = 16.9, C2), 39.3 (1JPC = 14.6, C4),
39.8 (2JPC = 5.7, C5), 45.2 (2JPC = 5.3, C1), C6 over-
lapped with the signals of CDCl3, 125.83 (C05), 128.43–
128.49 (C01 and C03 of both isomers overlapped), 128.75
(C04), 130.29 (2JPC = 4.2, C60 ), 135.48 (2JPC = 13.1, C02).
The total yield for the two isomers of 2c is 100%.
Compound 5: 31P NMR (C6D6) d –30.6; 13C NMR
(C6D6) d 23.1 (3JPC = 5.6, CH3), 128.7 (1JPC = 8.2, C2),
129.8 (C04), 131.8 (3JPC = 16.0, C03), 134.4 (1JPC = 16.2,
C10 ), 134.5 (2JPC = 21.5, C20 ), 134.7 (C3), 136.8
(3JPC = 11.4, C4), 144.4 (C7); FAB-MS, M+
found = 283.0191; C14H14Cl2P requires 283.0210 for
the 35Cl isotopes.
Compound 3b-syn: yield: 87%; m.p. 187–188 ꢁC; 31P
NMR (CDCl3) d 70.6 (JPRh = 136.6); 13C NMR
(CDCl3) d 8.6 (CP*–CH3), 20.2 (3JPC = 7.2, C1–CH3),
21.3 (Ar–CH3), 30.4 (1JPC = 31.7, C4), 36.4
(1JPC = 30.3, C2), 42.0 (C1), 43.2 (C5), 72.2 (C6), 98.0
(1JCRh = 4.1, Cp*), 128.6 (3JPC = 9.9, C03), 129.4
1
(1JPC = 33.7, C01) 130.8 (2JPC = 8.2, C20 ), 140.5 (C04); H
NMR (CDCl3) d 1.39 (d, 15H, CH3–Cp*, J = 3.6);
1.60 (s, 3H, CH3–C1), 1.99 (dd, 1H, C2–Hb,
2
2JHH = 8.0, JPH = 18.0), 2.34 (s, 3H, m-CH3–Ar), 2.64
2
2
(dd, 1H, cis-C2–Ha, JHH = 7.0, JPH = 15.2), 2.86 (part
of ABDX system, 1H, C4–Hb, J = 8.8, 8.8, 14.8), 2.96
(part of ABDX system, 1H, C4–HB, J = 7.2, 7.2, 14.8),
3.29 (m, 1H, C5–H), 5.27 (2H, CH2Cl2-solvate), 7.17
3
(d, 2H, o-C6H4CH3, JHH = 7.2), 7.68 (d, 1H, m-
3
C6H4CH3, JHH = 7.2), 7.71 (d, 1H, m-C6H4CH3,
3.3. Preparation of rhodium (III) complexes 3a–c and 6
3JHH = 7.2); Anal. Calc. for C23H30Cl4RhP Æ CH2Cl2:
C, 43.21; H, 4.83. Found: C, 43.18; H, 4.74%.
To the solution of 0.30 mmol of ligand 2a–c in 6 ml of
CH2Cl2 was added 89 mg (0.14 mmol) of [Cp*RhCl2]2 in
10 ml of CH2Cl2 at 25 ꢁC over a period of 10–15 min.
The red solution formed immediately was stirred for
0.5 h. According to 31P NMR, the formation of the
complex(es) was completed during this time. In the case
of 2c, the mixtures of isomers of the starting phosphines
were used in the ratio 2csyn:2c-anti = 95:5 and 38:62 to
obtain 3c-syn and 3c-anti, respectively. Then the solvent
was evaporated to dryness and 3 ml of CH2Cl2 was
added. The mixture was filtered and ꢀ1 ml of pentane
was added dropwise to the filtrate. The solution was al-
lowed to stand at 26 ꢁC for overnight. The red crystals
precipitated from the solution were filtered off and dried
in vacuo. For isolation of 3c-anti, the precipitated 35:75
mixture of the two 3c isomers was again recrystallized
using the same system of solvents (CH2Cl2-pentane).
Compound 3a-syn: yield: 87%; m.p. 191–195 ꢁC; 31P
NMR (CDCl3) d 70.3 (JPRh = 137.3), 13C NMR
(CDCl3) d 8.6 (Cp*–CH3), 20.1 (3JPC = 6.6, C1–CH3),
30.3 (1JPC = 31.5, C4), 36.3 (1JPC = 30.1, C2), 42.0 (C1),
43.1 (C5), 72.1 (C6), 98.0 (CP*), 127.9 (3JPC = 9.6, C30 ),
130.2 (4JPC = 2.5, C40 ), 130.7 (2JPC = 7.8, C20 ), 133.0
(1JPC = 31.0, C10 ); 1H NMR (CDCl3) d 1.42 (d, 15H,
CH3–Cp*, J = 3.6); 1.63 (s, 3H, CH3), 2.23 (dd, 1H,
Compound 3c-syn: yield: 67%; m.p. 169–170 ꢁC; 31P
NMR (CDCl3) d 67.8 (1JPRh = 139.9); 13C NMR
(CDCl3) d 8.5 (Cp*–CH3), 20.2 (C1–CH3), 22.8 (Ar–
CH3), 30.0 (C4), 34.5 (C2), 44.0 (C5), 73.9 (C6), 98.1
(Cp*), 124.75 (2JPC = 7.8, C60 ), 130.3 (C04), 130.5
(1JPC = 108.5, C01), 130.7 (3JPC = 4.5, C03), 131.5
1
(3JPC = 8.4, C05), 141.7 (2JPC = 10.2, C02–Me); H NMR
(CDCl3) d 1.35 (d, 15H, CH3–Cp*, J = 3.6), 1.59 (s,
3H, CH3–C1), 1.95–2.06 (br.m, 1H, C2–H), 2.50 (br.s,
4H, o-CH3–Ar overlapped with C2–H), 2.92–3.12
(br.m, 2H, C2–H2), 3.62–3.70 (br.m, 1H, C5–H), 5.27
3
(2H, CH2Cl2-solvate), 7.21 (d, 1H, C06-H, JHH = 7.4),
3
7.24 (d, 1H, C02–H, JHH = 7.2), 7.35 (dd, 1H, C05–H,
3
3JHH = 7.4, 7.6), 7.50 (t, 1H, C40 –H, JHH = 7.6). Anal.
Calc. for C23H30Cl4RhP: C, 47.45; H, 5.19. Found: C,
47.75, 47.60; H, 5.19, 5.27%.
Compound 3c-anti: yield: 32%; m.p.: 166–168 ꢁC; 31
P
NMR (CDCl3) d 91.6 (1JPRh = 143.4); 13C NMR
(CDCl3) 20.7 (C1–CH3), 22.6 (Ar–CH3), 29.5 (C4),
34.2 (1JPC = 27.4, C2), 40.4 (C1), 41.9 (C5), 75.1
(3JPC = 14.0, C6), 98.6 (d, Cp*, J = 2.6), 125.4
(2JPC = 6.8, C06), 130.9 (C40 ), 130.5 (1JPC = 88.5, C01),
131.4 (3JPC = 8.4, C50 ), 131.6 (3JPC = 8.6, C50 ), 132.3
(2JPC = 8.0, C20 –Me); Anal. Calc. for C23H30Cl4RhP: C,
47.45; H, 5.19. Found: C, 47.85; H, 5.08%.
2
2
C2–Hb, JHH = 7.6, JPH = 18.4), 2.70 (dd, 1H, cis-C2–
Complex 6 was obtained by the analogous procedure
using the appropriate amounts of [Cp*RhCl2]2 and
phosphine 5. Yield: 80%; m.p. >350 ꢁC; 31P NMR
(CDCl3) d –0.89 (1JRhP = 142.1); 13C NMR (CDCl3) d
8,7 (CH3–Cp*), 22.9 (3JPC =9.3, C8), 98.3 (Cp*), 121.7
(1JPC = 48.9, C2), 127.0 (1JPC = 86.0, C01-ipso), 127.8
(3JPC = 10.2, C30 ), 130.2 (C40 ), 131.1 (3JPC = 42.5, C4),
131.6 (3JPC = 9.2, C20 ), 140.4 (C3), and 152.4
2
2
Ha, JHH = 7.6, JPH = 15.2), 2.90 (part of ABDX sys-
tem, 1H, C4–Hb, J = 8.4, 8.4, 14.8), 3.01 (part of ABDX
system, 1H, C4–HB, J = 7.6, 7.6, 14.8), 3.32 (m, 1H, C5–
H, J = 8.4, 8.0, 14.8), 7.35–7.39 and 7.81–7.85 (2m,
3H + 2H,
Ph).
MS,
m/z
(rel.
int.)
258
([M ꢁ Cp*RhCl2]+, 6), 223 ([M ꢁ Cp*RhCl2 ꢁ Cl]+,
30), 187 ([M ꢁ Cp*RhCl2 ꢁ Cl ꢁ HCl]+, 9) and inde-
pendently 308 ([Cp*RhCl2]+, 8), 273 ([Cp*RhCl]+, 8),
1
(4JPC = 8.8, C7); H NMR (CDCl3) d 1.49 (d, CH3Cp*,

I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
711
2
J = 3.6, 15H), 2.29 (s, C8H3, 6H), 6.48 (d, JPH = 27.2,
C2H, 2H), 7.32–7.51 (m, m-, p-C6H5P, 3H), 7.78 (d,
2.32 (s, 12H, CH3), 6.4 (2JPH = 18.4, 4H, CH), 7.41–
7.47 (6H, m-, p-C6H5P), 7.70–7.74 (4H, o-C6H5P); IR
(KBr) m C@C 1605–1625 (br) cmꢁ1; Anal. Calc. for
C28H26Cl6P2Pd: C, 45.23; H, 3.52; P, 8.33. Found: C,
45.45, 45.55; H, 3.83, 3.37; P, 8.21, 8.16%.
3
3JHH = 7.6, o-C6H5P, 1H), 7.80 (d, JHH = 7.2,
o-C6H5P, 1H); IR (KBr) m C@C 1610–1625(br) cmꢁ1
.
MS, m/z (rel. int.) 282 ([M ꢁ Cp*RhCl2], 21), 247
([M ꢁ Cp*RhCl2 ꢁ Cl], 30), 211 ([M ꢁ Cp*RhCl2 ꢁ
Cl ꢁ HCl], 25), 196 ([M ꢁ Cp*RhCl2 ꢁ Cl ꢁ HCl ꢁ
CH3], 13), 181 ([M ꢁ Cp*RhCl2 ꢁ Cl ꢁ HCl ꢁ 2CH3],
15) and independently 308 ([Cp*RhCl2]+, 7), 273
([Cp*RhCl]+, 8), 237 ([Cp*Rh ꢁ H]+, 17), 236
([Cp*Rh ꢁ H]+, 23), 134 ([Cp* ꢁ H]+, 34). Anal. Calc.
for C24H28Cl4RhP: C, 48.68; H, 4.77; P, 5.23. Found:
C, 49.16, 49.48; H, 5.21, 5.17; P, 4.80, 4.87%.
3.5. X-ray Crystallography
Crystallographic data for the complexes are pre-
sented in Table 2. All structures were solved by direct
methods and refined by full-matrix least-squares against
F2 in the anisotropic (H-atoms isotropic) approxima-
tion, using the SHELXTL-97 package [15]. The absorp-
tion correction for all structures with the exception of
3c-syn Æ 2CH2Cl2 was applied semi-empirically from
equivalents using the SADABS v. 2.01 program [16].
The positions of hydrogen atoms in all structures were
calculated from geometrical point of view and included
in refinement in riding model approximation. Crystallo-
graphic data (excluding structure factors) for the struc-
tures reported in this paper have been deposited to the
Cambridge Crystallographic Data Centre as supplemen-
3.4. Preparation of palladium (II) complex 7
To a 10 ml C6H6 solution of crude ligand 5 obtained
from 4 (0.10 mmol) was added 32.0 mg (0.084 mmol) of
(PhCN)2PdCl2 in 5 ml of the same solvent at 0 ꢁC in a
period of 10 minutes and the solution was stirred for
0.5 h. According to 31P NMR, only one signal of the
corresponding complex was observed after this time.
After warming up to room temperature, the solvent
was evaporated and 3 ml of CH2Cl2 was added. The
mixture was filtered and ꢀ1 ml of pentane was added.
The solution was allowed to stand at 26 ꢁC for over-
night. The pale yellow precipitate was filtered off and
dried in vacuo to afford 57 mg (73%) of 7; m.p. >350
ꢁC; 31P NMR (CDCl3) d –5.2; 1H NMR (CDCl3) d
tary
No.
CCDC-225064
for
3b-syn Æ CH2Cl2,
CCDC-225065 for 3c-anti Æ 2CHCl3, CCDC-225066 for
3c-syn Æ 2CH2Cl2, CCDC-225067 for 3c-syn Æ 2CHCl3.
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.
E-mail: deposit@ ccdc.cam.ac.uk).
Table 2
Crystallographic data for complexes 3b,c
3b-syn Æ CH2Cl2
3c-anti Æ 2CHCl3
3c-syn Æ 2CHCl3
3c-syn Æ 2CH2Cl2
Formula
T (K)
C23H30Cl4PRh Æ CH2Cl
110
Orthorhombic, Pbca
C23H30Cl4PRh Æ 2CHCl3 C23H30Cl4PRh Æ 2CHCl3 C23H30Cl4PRh Æ 2CH2Cl2
2
120
120
Monoclinic, P21/n
153
Monoclinic, P21/n
ꢀ
Crystal system, space group
Unit cell dimensions
Triclinic, P1
˚
a (A)
24.514(4))
8.708(1)
26.115(4)
12.505(2)
16.558(2)
17.540(2)
102.814(2)
106.659(2)
90.395(2)
3383.1(7)
4(2)
19.548(2)
8.997(1)
20.190(2)
19.234(4)
8.835(2)
20.353(4)
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
110.164(2)
113.56(3)
3
˚
V (A )
5574.5(15)
8(1)
3333.3(7)
4(1)
3170.5(11)
4(1)
Z(Z0)
M
667.08
12.58
2704
820.89
13.58
820.89
13.78
1648
752.00
12.78
1520
l (cmꢁ1
F(000)
)
1648
1.612
qcalc (g cmꢁ3
2hmax (deg)
Diffractometer
)
1.590
56.00
1.636
58.00
1.680
58.00
58.00
Smart CCD
x
28,192 (0.0540)
16,887
10,800
Smart CCD
x
19,301 (0.0588)
6667
3756
Smart CCD
x
35,827(0.0888)
8760
6970
Syntex P21
h/2h
6415 (0.0299)
6227
4999
Scan mode
Number of reflections measured (Rint
Number of independent reflections
)
Number of reflections with I > 2r(I)
Number of parameters
R1
296
673
0.0579
341
323
0.0585
0.1209
1.083
0.0624
0.1467
1.085
0.0346
0.0939
1.003
wR2
GOF
Max/min peak (e A
0.1309
1.070
1.038/–0.992
ꢁ3
˚
)
1.28/–0.61
1.18/–0.300
0.70/–0.98

712
I.L. Odinets et al. / Journal of Organometallic Chemistry 690 (2005) 704–712
[3] Gy. Keglevich, Synthesis (1993) 931.
3.6. Quantum chemical calculations
[4] Gy. Keglevich, Rev. Heteroatom Chem. 14 (1996) 119.
[5] R. Engel, in: R. Engel (Ed.), Handbook of Organophosphorus
Chemistry, Marcel Dekker, Inc., New York, 1992 (Chapter 5).
For computational consistency no symmetry con-
strains were used for all calculated molecules. The calcu-
lations have been done with the B3PW91 DFT method
and 6-31G* basis set. Transition state calculations also
carried out within the same DFT method and basis set
via QST3 transition state search. Frequencies for all
molecules also were calculated to proof transition state
geometry and to consider zero point energy for inver-
sion barrier energy and activation energy for this proc-
ess. For each transition state geometry was found only
one negative frequency. All calculations were carried
out with GAUSSIAN 98 program package [17] in which
ultrafine grid (99590) and the (10ꢁ8 hartree) designation
were used for the SCF convergence. As a convergence
criteria threshold limit 2 · 10ꢁ6 and 4 · 10ꢁ6 a.u. were
used for maximum force and displacement, respectively.
´
[6] Gy. Keglevich, T. Kegl, I.L. Odinets, N.M. Vinogradova, L.
´
Kollar, Compt. Rend. Chim. 7 (2004) 779.
´
´
´
´
[7] Gy. Keglevich, I. Petnehazy, P. Miklos, A. Almasy, G. Toth, L.
To¨ke, L.D. Quin, J. Org. Chem. 52 (1987) 3983.
[8] Gy. Keglevich, J. Brlik, F. Janke, L. To¨ke, Heteroatom Chem. 1
(1990) 419.
´
[9] Gy. Keglevich, F. Janke, V. Fulo¨p, A. Kalman, G. Toth, L.
¨
´
´
To¨ke, Phosphorus, Sulfur, Silicon 54 (1990) 73.
´
´
´
[10] Gy. Keglevich, H. Szelke, A. Balint, T. Imre, K. Ludanyi, Z.
Nagy, M. Hanusz, K. Simon, V. Harmat, L. To¨ke, Heteroatom
Chem. 14 (2003) 443.
[11] L.D. Quin, A Guide to Organophosphorus Chemistry, Wiley,
New York, 2000 (Chapters 3 and 9).
´
[12] L. Nyulaszi, U. Bergstra¨sser, M. Regitz, P.v.R. Scleyer, New J.
Chem. (1998) 651.
´
[13] Gy. Keglevich, A. Szo¨llo¨sy, L. To¨ke, V. Fulo¨p, A. Kalman, J.
¨
´
´
Org. Chem. 55 (1990) 6361.
[14] Gy. Keglevich, L. To¨ke, Cs. Lovasz, K. Ujszaszy, G. Szalontai,
´
´
´
Heteroatom Chem. 5 (1994) 395.
[15] G.M. Sheldrick, SHELXTL Version 5, Software Reference Manual,
Siemens Industrial Automation, Madison, WI, 1994.
Acknowledgements
[16] G.M. Sheldrick, SADABS v. 2.01, Bruker/Siemens Area Detector
Absorption Correction Program, Bruker AXS, Madison, Wis-
consin, USA, 1998.
This work was partially supported by Russian Basic
Research Foundation (Grant No. 02-03-33073) and
the Federal Program on Support of Leading Scientific
Schools (Grant Nos. SS-1100.2003.3, 1060.2003.3 and
YC-1209.2003.03). Gy. K is grateful for the OTKA-sup-
port (No. T042479) of the project.
[17] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria,
M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Mont-
gomery, R.E. Stratmann, J.C. Burant, S. Dapprich, J.M.
Millam, A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas,
J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci,
C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A.
Petersson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick,
A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Clioslow-
ski, J.V. Ortiz, B.B. Stefanov, G. Liu, A. Liashenko, P.
Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox,
T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C.
Gonzalez, M. Challacombe, P.M.W. Gill, B. Johnson, W.
Chen, M.W. Wong, J.L. Andres, C. Gonzalez, M. Head-
Gordon, E.S. Replogle, J.A. Pople, GAUSSIAN 98, Gaussian,
Inc., Pittsburgh PA, 1998.
References
[1] (a) Iwao Ojima (Ed.), Catalytic Asymmetric Synthesis, second ed.,
Wiley-VCH, Wiley, 2000;
(b) B. Breit, W. Seiche, Synthesis (2001) 1.
[2] F. Mathey, F. Mercier, C.R. Acad. Sci. Paris, Se. IIb 324 (1997)
701.