'Diopium', a chiral phosphoniophosphine derived from Kagan's diop. Rhodium complexes and reducing catalytic properties

Article abstract of DOI:10.1016/S0022-328X(02)01433-X

Cationic chiral monophosphines are devised as a novel type of ligands for catalytic transformations of polar substrates with polar reagents. A phosphonio-phosphine derived from (R,R)-diop ('methyldiopium') proved to act as a chiral version of Baird's ω-phosphonio-phosphine ligands ('phophos') in Rh(I) complexes in both 1:1 and 2:1 P:Rh stoichiometry. Their versatile structure in CDCl₃ solution has been studied by NMR spectroscopy in the presence of various anions (Cl^-, I^-, BF₄^-, PF₆^-). The in situ 2.5:1 methyldiopium-Rh catalytic system slowly hydrogenated (Z)-α-acetamidocinnamic acid in 63% conversion and 5% ee. This system was, however, specifically active in a novel hydrogen transfer procedure by a 1:1 HCOOH:NEt₃ mixture under mild conditions (40 C, in the absence of DMSO), for which the corresponding diop-Rh system was not active. Itaconic acid was thus quantitatively reduced to racemic methylsuccinic acid. (Z)-α-acetamidocinnamic acid was reduced in up to 85% conversion to N-acetylphenylalanine in various solvents. Though still moderate, the ee's depend on the solvent: 10% in (R) product in THF, 14% in (S) product in acetonitrile.

Full text of DOI:10.1016/S0022-328X(02)01433-X

Journal of Organometallic Chemistry 654 (2002) 180ꢁ186
/
www.elsevier.com/locate/jorganchem
‘Diopium’, a chiral phosphoniophosphine derived from Kagan’s
diop. Rhodium complexes and reducing catalytic properties
Lydie Viau, Remi Chauvin *
Laboratoire de Chimie de Coordination du CNRS, 205 Route de Narbonne, F-31077 Toulouse Cedex 4, France
Received 29 October 2001; accepted 3 April 2002
Abstract
Cationic chiral monophosphines are devised as a novel type of ligands for catalytic transformations of polar substrates with polar
reagents. A phosphonio-phosphine derived from (R,R)-diop (‘methyldiopium’) proved to act as a chiral version of Baird’s v-
phosphonio-phosphine ligands (‘phophos’) in Rh(I) complexes in both 1:1 and 2:1 P:Rh stoichiometry. Their versatile structure in
CDCl3 solution has been studied by NMR spectroscopy in the presence of various anions (Clꢀ, Iꢀ, BF4ꢀ, PF6ꢀ). The in situ 2.5:1
methyldiopiumꢁ
was, however, specifically active in a novel hydrogen transfer procedure by a 1:1 HCOOH:NEt3 mixture under mild conditions
(40 8C, in the absence of DMSO), for which the corresponding diopꢁRh system was not active. Itaconic acid was thus
/
Rh catalytic system slowly hydrogenated (Z)-a-acetamidocinnamic acid in 63% conversion and 5% ee. This system
/
quantitatively reduced to racemic methylsuccinic acid. (Z)-a-acetamidocinnamic acid was reduced in up to 85% conversion to N-
acetylphenylalanine in various solvents. Though still moderate, the ee’s depend on the solvent: 10% in (R) product in THF, 14% in
(S) product in acetonitrile. # 2002 Published by Elsevier Science B.V.
Keywords: Diop; Phosphoniophosphines; Rhodium complexes; Hydrogen transfer; Asymmetric catalysis
1. Introduction
performances consisted in the introduction of functional
groups enabling secondary ligandꢁmetal interactions
[7,3d,3e]. In particular, chelating dativeꢁdative-electro-
/
Thirty years after its discovery by Kagan [1], diop
stands as a historical benchmark amongst chiral dipho-
sphines [2], while the chemical simplicity of its chiral
skeleton is still inspiring the search for new ligands of
asymmetric hydrogenation catalysts [3]. As a further
example, ‘methyldiopium’ 1 [4], a monophosphonium
salt of diop (Scheme 1), is proposed here as a tentative
prototype to examine the scope of hybrid dative-
electrostatic interactions in the rhodium-catalyzed re-
duction of acrylic acid derivatives.
/
static interactions have been invoked to account for high
ee’s obtained in rhodium-catalyzed hydrogenation of
acrylic acid derivatives in the presence of chiral amino-
diphosphines ligands [8]. Keeping in mind recent
observations that certain monophosph(on)ite-rhodium
catalysts also exhibit high enantioselectivity in catalytic
hydrogenation [9], we may devise new ligands suscep-
tible to be engaged in a simple dative-electrostatic
interaction with a metal center. A single P0
/Rh dative
bond might be sufficient to ensure a chiral organization,
which would simply result from a formal cationic charge
on a monophosphine backbone. In most of the chiral
quaternary ammonium-diphosphine ligands reported so
far, the positive charges are rigidly repelled at the
periphery of the complex in order to ensure water-
solubilty [10]. By contrast, the present prospect requires
the possibility of back folding of the cationic charge
onto an electron-rich metal center or to its polar ligands
[11]. The metal would thus be embedded in a enzyme-
like chiral electrostatic pocket where a ‘chiral zwitter-
After the early discoveries of chiral monophosphineꢁ
/
rhodium catalysts [5], a rigid chelation of the transition
metal center by diphosphine ligands was recognized as a
desirable feature for the design of highly enantioselective
hydrogenation catalysts [6]. A further conceptual ad-
vance aimed at improving the diphosphine catalysts’
* Corresponding author. Tel.:
553003.
ꢂ/33-5-61-333113; fax: 33-5-61-
E-mail address: chauvin@lcc-toulouse.fr (R. Chauvin).
0022-328X/02/$ - see front matter # 2002 Published by Elsevier Science B.V.
PII: S 0 0 2 2 - 3 2 8 X ( 0 2 ) 0 1 4 3 3 - X

L. Viau, R. Chauvin / Journal of Organometallic Chemistry 654 (2002) 180ꢁ
/186
181
[NH4][PF6] (0.150 g, 1.6 mmol) in dichloromethane (5
ml) and water (5 ml) for 6 h. The organic phase was then
separated, washed with water (3ꢄ10 ml), dried over
/
MgSO4, filtered and dried under vacuum, affording 1×
/
PF6 as a white powder (0.056 g, 72%). 31P{1H}-NMR
1
(CDCl3, 81 MHz) d: ꢀ
1P; PF6ꢀ); ꢀ
PꢂMe).
/
143.95 (sept, JPF
ꢃ713.3 Hz,
/
/
24.87 (s, 1P; CH2PPh2); 23.21 (s, 1P;
Scheme 1. (R,R)-methyldiopium salts from (R,R)-diop.
ionic control’ of the reaction could take place [12]. A
related strategy has been envisioned by Brunner with
2.2. Characterization of complex 2×
/
BF4
chiral pyridiniumꢁ
the hydrosilylation of ketones [13].
Baird et al. have designed achiral phosphoniumꢁ
phosphine ligands PPh2(CH2)nPꢂMe3 (nꢃ
2, 3, 6, 10)
for which they coined the generic term ‘n-phophos’ [14].
Phophosꢁrhodium complexes (2:1) were shown to be
/
oxazolidineꢁ/rhodium catalysts for
A solution of [NH4][BF4] (0.022 g, 0.21 mmol) in
water (1 ml) was added to diopium iodide 1×I (0.067 g,
0.10 mmol) in dichloromethane (5 ml). After stirring for
/
/
/
2 h, the organic phase was separated, washed with water
(2ꢄ
dissolved in methanol (3 ml), and added to a solution of
[Rh(cod)Cl]2 (0.026 g, 0,032 mmol) in a 1:1 methanolꢁ
/
5 ml), and concentrated. The product (1×
/
BF4) was
/
catalytic precursors for the hydrogenation of alkenes in
biphasic media [15]. A chiral version of IV-phophos
/
dichloromethane mixture (1 ml). The mixture was
stirred for 9 h at room temperature (r.t.). The solvents
(nꢃ
1.
/4) is here investigated through the diopium ligand
were evaporated to dryness, giving the single complex 2×
/
BF4 as a light brown solid which could be characterized
without further purification. 1H-NMR (200 MHz) d:
2. Experimental
1.35, 1.41 (2 s, 6H; C(CH3)2); 1.91ꢁ
/
2.56 (m, 8H;
14 Hz, 3H; CH3Pꢂ); 2.80ꢁ
3.02 (m, 2H; CH2P); 3.13 (broad, 2H; CH ÄCH(cod)
trans to Cl); 3.20ꢁ
3.61 (m, 3H; PꢂCH2CHO); 4.20 (m,
1H; RhPCH2CHO); 5.15ꢁ
5.55 (broad, 2H; CH Ä
2
CH2(cod)); 2.67 (d, JPH
ꢃ
/
/
All reactions were carried out under a nitrogen
atmosphere using Schlenk tube and vacuum line tech-
nics. THF and diethylether were distilled over Naꢁ
benzophenone. Dichloromethane and acetonitrile were
distilled over P2O5. Triethylamine was distilled over
KOH. Commercial synthesis grade methanol (SDS) was
degassed by bubbling argon. (ꢀ)-(R,R)-diop, (Z)-a-
acetamidocinnamic acid and methyl iodide were pur-
chased from Fluka. Formic acid was purchased from
Riedel-de-Haen. Ammonium tetrafluoroborate was pur-
¨
chased from Aldrich. [RhCl(cod)]2 was prepared from
cyclooctadiene and RhCl3×
according to a modified described procedure (no car-
bonate was added) [16]. NMR spectra were recorded in
CDCl3 solution, on Brucker AC 200, AM250 and AMX
400 spectrometers. Positive chemical shifts at low field
are expressed in ppm by internal reference to TMS for
1H and 13C, and by external reference to 85% H3PO4 in
D2O for 31P. Optical rotations were measured in a 1 dm
/
/
/
/
/
CH(cod) cis to Cl [19a,19]); 7.2ꢁ8.0 (m, 20H; aromatic
/
CH). 31P{1H}-NMR (81 MHz) d: 23.16 (s, 1P;
1
PꢂCH3); 29.41 (d, JPRh
ꢃ149.4 Hz, 1P; PRh).
/
/
13C{1H}-NMR (62.9 MHz) d: 8.41 (d, JPC
ꢃ
/
55.7 Hz;
2
PꢂCH3); 26.16, 26.41 (2 s; C(CH3)2); 25.30ꢁ
/
27.03
(broad m; CH2P, CH2Pꢂ and CH2(cod) [19a]); 75.32
(m; CHO); 110.02 (s; CMe2); 118.77, 120.15 (2 d,
1JPC
ꢃ
/
86.6, 87.3 Hz; (ipso-C)2Pꢂ); 128.68ꢁ
134.76 (m;
/
/
3H2O (JohnsonꢁMattey)
/
aromatic C). As reported for [RhCl(nbd)(II-phophos)],
the olefinic carbons of the diene ligand did not give
observable 13C-NMR signals [15a].
2.3. VTP 31P-NMR analysis of complexes 2×
/
I and 3×
/
Cl
Methyldiopium iodide 1×I (0.026 g, 0.04 mmol) and
/
cell with a Perkinꢁ
2.1. (R,R)-Methyldiopium
1×I was prepared from (R,R)-diop and methyl iodide
/
Elmer 241 polarimeter.
[Rh(cod)Cl]2 (0.010 g, 0.02 mmol) were dissolved in
CDCl3 (0.5 ml) in an NMR tube. Heating of the sample
for 10 min at 50 8C was required to observe the
disappearance of the free 31PIII signal of 1. NMR
spectra were recorded at 162 MHz, at 293 K and 233
/
according to a described procedure in 97% yield
according to stoichiometry [4]. Complementary analy-
K. 31P{1H}-NMR (293 K) d: 25.78 (s; PꢂCH3); 28ꢁ
32
/
(residual coalescence signal; PRh). 31P{1H}-NMR (233
1
tical data: melting point (m.p.): 106ꢁ
/
109 8C; FAB (B
/
K) d: 25.87 (broad s, 1P; CH3Pꢂ); 34.27 (d, JPRh
ꢃ
/
1
0) m/z: 127 ([I]ꢀ); FAB (ꢀ0) m/z: 513 ([1]ꢂ). Tetra-
fluoroborate and hexafluorophosphate of 1 were ob-
tained by anion metathesis. For instance: diopium
/
144.1 Hz, 0.7P; PRh of 3×
Hz, 0.3P; PRh of 2×
also observed at d: 26.04 (CH3Pꢂ), 32.57 (P Ä
assigned by comparison with an authentic sample
/
Cl); 35.47 (d, JPRh
ꢃ
/149.3
/
I). Diopium oxide impurities were
/O), and
iodide 1×
/
I (0.076 g, 0.2 mmol) was treated with

182
L. Viau, R. Chauvin / Journal of Organometallic Chemistry 654 (2002) 180ꢁ186
/
prepared by controled oxidation of 1×
/
I with H2O2 in
protons of the PhPh?PRh units by 1HÃ
/
1H selective
1
acetone.
decoupling). 13C-NMR (62.9 MHz) d: 7.78 (dq, JPC
ꢃ
/
1
56.7 Hz, JCH
1
ꢃ
/
135.8 Hz; CH3Pꢂ); 26.42 (dm, JPC
ꢃ
/
2.4. Characterization of complex 4
55.3 Hz, CH2Pꢂ); 25.86, 26.39 (2 s; C(CH3)2); 28.28
(broad m, CH2PRh); 30.03, 31.32 (2 m; CH2 cod); 74.61
1
(broad d, JCH
[RhCl(cod)]2 (0.016 g, 0.032 mmol) and methyldio-
pium iodide 1×I (0.084 g, 0.13 mmol) were dissolved in
ꢃ
/
148.4 Hz; CHÃ
/
O); 78.80 (broad d,
1
/
1JCH
CH cod); 99.30 (d; JCH
ꢃ
/
150.9 Hz; CHÃ
/
O); 97.51 (d; JCH
157.7 Hz; ÄCH cod); 109.81
86.2 Hz; (ipso-C)Pꢂ);
88.2 Hz; (ipso-C)Pꢂ); 127.41ꢁ
136.88
)FAB m/e: 1527.4
ꢃ
/
151.1 Hz; Ä
/
1
CH2Cl2 (3 ml). A solution of [NH4][BF4] (0.010 g, 0.096
mmol) in water (3 ml) was added, and the mixture was
stirred for 2 h at r.t., and the organic phase was
ꢃ
/
/
1
(s; C(CH3)2); 119.54 (d, JCP
ꢃ
/
1
119.82 (d, JCP
ꢃ
/
/
separated, then concentrated. Crude complex 4×
/Y2
(m; aromatic C). (ꢂ
/
)ESMS and (ꢂ
/
(YꢃBF4, I) was obtained as an orange oil and
/
([Rh(cod)(1)2(PF6)2]ꢂ) minor peak with consistent iso-
topic pattern); 513.2 ([1]ꢂ).
characterized as such, exactly as described for
[Rh(nbd)(n-phophos)2]ꢂ complexes [15a]. 1H-NMR
(250 MHz) d: 1.34, 1.44 (2 s, 12H; C(CH3)2); 2.00ꢁ
/
2.6. Procedure for the catalytic reduction of (Z)-a-
acetamidocinnamic acid with 1:1 HCOOH:NEt3
2
2.50 (m, 8H; CH2(cod)); 2.67 (d, JPH
CH3Pꢂ); 2.80ꢁ
3.22 (m, 4H; CH2P); 2.35ꢁ
CH2Pꢂ); 3.59, 4.20 (2 m, 2ꢄ
2H; (CHO)2); 5.60ꢁ
(m, 4H; CH ÄCH); 7.20ꢁ8.00 (m, 40H; aromatic CH).
Secondary signals at dꢃ5.37 (m) and 1.5ꢁ2.5 (m) might
ꢃ
/
14 Hz, 6H;
3.65 (m, 4H;
5.70
/
/
/
/
[RhCl(cod)]2 (0.005 g, 0.1 mmol), methyldiopium
/
/
iodide 1×I (0.032 g, 50 mmol) and acetamidocinnamic
/
/
/
acid 5 (0.148 g, 0.72 mmol) were dissolved in THF (5
ml). The mixture was stirred for 15 min. at 20 8C.
Triethylamine (0.5 ml, 3.6 mmol) and formic acid (0.129
ml, 3.6 mmol) were then added, and the medium was
stirred for 9 h at 40 8C. After cooling to r.t., 2 M
aqueous (aq.) NaOH was added, and the aqueous phase
be assigned to a side-product of general formula
[Rh(cod)2]X. 31P{1H}-NMR (81 MHz) d: 23.11 (s, 2P;
1
CH3Pꢂ); 31.87 (d, JPRh
ꢃ144.9 Hz, 2P; PRh).
/
13C{1H}-NMR (62.9 MHz) d: 8.26 (d, JPC
ꢃ56 Hz;
/
1
CH3Pꢂ); 25.96 (d, JPC
ꢃ
55 Hz, CH2Pꢂ); 26.90, 27.23
/
1
(2 s; C(CH3)2); 29.38, 30.10, 32.29, 32.89 (4 s;
CH2(cod)); 32.40 (d, JPC 27.5 Hz; CH2PRh); 74.07ꢁ
75.48, 81.05 (2 m; (CHO)2); 103.35 (very broad; Ä
CH(cod)); 110.44 (s; C(CH3)2); 120.22 and 120.40 (2
was washed with ether (3ꢄ
N HCl, and filtered through celite. The filtrate was
extracted with ether (3ꢄ20 ml), dried over MgSO4 and
/
10 ml), then acified with 10
ꢃ
/
/
/
/
concentrated. The resulting grey solid was analyzed by
1H-NMR in DMSO-d6 solution showing the signals of
the starting material 5 and those of the N-acetylpheny-
lalanine 6 only: the calculated conversion was 54%. The
enantiomeric excess in (R)-6 (10%) was estimated from
the optical rotation of a diluted solution of the product:
1
d, JCP
other
ꢃ
/
86.2 Hz; (ipso-C)2Pꢂ); 127.83ꢁ
aromatic C). )ESMS m/e:
/
135.68 (m;
851.1
(ꢂ
/
([RhI(cod)(1)]ꢂ) with consistent isotopic pattern).
2.5. Characterization of complex 5
[a]2D5ꢃ
/
ꢀ
/
5.58 (cꢃ0.297, absolute EtOH).
/
RhCl(cod)]2 (0.010 g, 0.021 mmol) and methyldio-
pium hexafluorophosphate 1×PF6 (0.054 g, 0.082 mmol)
/
were dissolved in CH2Cl2 (3 ml). A solution of
[NH4][PF6] (0.037 g, 0.227 mmol) in water (3 ml) was
added. The mixture was stirred for 2 h at r.t. while the
color turned from yellow to orange. The organic phase
3. Results and discussion
3.1. Solution structure of diopiumꢁ
/rhodium complexes
was separated, washed with water (3ꢄ6 ml) and then
/
In continuation of previous reports on the coordina-
tion chemistry of phosphonium derivatives of diop
(isodiop in a chlororhenium cluster [17], methyldiopium
1 in carbonyliron complexes [4]), the solution structure
concentrated. Complex 5×
/
(PF6)3 was obtained as a light
orange solid (0.050 g, 73%). 31P{1H}-NMR (81 MHz) d:
1
22.05 (s, 2P; PꢂMe);16.45 (d, JRhP
ꢃ144.0 Hz, 2P;
/
1
143.87 (sept, JPFꢃ
713.9 Hz, 3 P; PF6ꢀ). H-
NMR (200 MHz) d: 0.63, 1.10 (2 s, 12H; C(CH3)2);
1
PRh); ꢀ
/
/
of 1ꢁ/Rh(I) complexes in 1:1 and 2:1 stoichiometry has
2
been investigated.
1.90ꢁ
6H; CH3Pꢂ); 2.25ꢁ
3.62ꢁ3.82 (m, 4H; OCH Ã
cod); 4.92 (m, 2H; CHꢃcod); 6.87 (broad t, JHH
Hz, 2H; p-CH PhPRh) 7.18ꢁ7.31 (m, 10H; aromatic
CH Ph?PRh); 7.46 (broad t, 3JHH
7.4 Hz, 4 H; m-CH
PhPRh); 7.64ꢁ
7.78 (m, 20H; aromatic CH Ph2Pꢂ); 8.25
(broad td, 3JHH
7.3 Hz, J(RhP)H 10.2 Hz, 4 H; o-CH
PhPRh) (assignment of the diastereotopic Ph and Ph?
/
2.15 (m, 8H; CH2(cod)); 2.44 (d, JPH
3.15 (m, 8H; CH2P and CH2Pꢂ);
CHO) 4.46 (m, 2H; ÄCH
7.3
ꢃ/13.7 Hz,
/
/
/
/
3.1.1. Diopium:rhodium stoichiometryꢃ
Reaction of 1×BF4 with [Rh(cod)Cl]2 lead to the
formation of the complex 2×BF4 (Scheme 2). It is a
/
1:1
3
/
ꢃ
/
/
/
/
:
/
chiral homologue of Baird’s squareꢁplanar complex
/
/
[RhCl(nbd)(II-phophos)]ꢂ, with similar spectroscopic
1
:
/
:
/
characteristics: d
23.16 (s) [15a].
/
31P ꢃ
/
29.41 (d, JPRh
ꢃ149.4 Hz) and
/

L. Viau, R. Chauvin / Journal of Organometallic Chemistry 654 (2002) 180ꢁ
/186
183
Scheme 2. 1:1 Diopiumꢁrhodium complexes in the presence of various anions.
/
The iodide analog 1×
/
I exhibits a different behavior.
I
After heating the CDCl3 solution of 1×
/
and
[Rh(cod)Cl]2 for 10 min at 50 8C, the 31P-NMR
spectrum at 293 K and 162 MHz displays a sharp signal
for the phosphonium center at dꢃ
broad signal at dꢃ28ꢁ32. Decoalescence starts at 273
K, and finally gives rise to two sharp doublets at 233 K:
/
25.78, and a very
/
/
1
dꢃ
1JRhP
those of isomers 2×
1JRhP coupling constant of the cis-X Ã
squareꢁplanar RhXP2L complexes with similar d and
1JPRh values (ca. 35 ppm and 140 Hz, respectively), were
found to be a few Hz greater for XꢃCl than for Xꢃ
[18]. The integration is thus also in accordance with a
greater stability of the RhÃCl bond with respect to the
RhÃI bond. The interconversion of these species by a
fast chlorideꢁiodide exchange at r.t., could occur via an
associationꢁdissociation mechanism, involving a 18-
/
34.27 (70%, JRhP
144.1 Hz). These PRh units can be assigned to
I and 3×Cl, respectively. Indeed, the
RhÃP unit in
ꢃ
/
149.3 Hz) and 35.47 (30%,
ꢃ
/
/
/
/
/
/
31P
/
/
I
/
/
/
/
electron zwitterionic rhodate intermediate A [12]. Simi-
lar elusive [RhCl2(cod)P2]ꢀ species have been previously
mentionned [19a]. Within this hypothesis, the free
Scheme 3. 2:1 Diopiumꢁrhodium complexes in the presence of iodide,
tetrafluoroborate and hexafluorophosphate anions.
/
enthalpy of activation DG" for the 2×
/
I?
/
3×/Cl inter-
conversion can be estimated from the Eyring equation
[19b]:
characteristics was formed even from a large excess
[NH4][PF6]: integration of the 31P-NMR spectrum
showed that the PF6ꢀ anions compensate the charges
ꢀ
ꢁ
Tc
Kc
pDn
pffiffi
DG" ꢃRTc 10:32ꢂlog
cal: molꢀ1
;
kc ꢃ
;
of the phosphonium groups but not that of the Rh(I)
1
23.08 (s, 2P); 32.35 (d, JPRh
2
center: d
143.99 (sept, 1JPF
iodide association is thus rather strong and is revealed in
/
31P ꢃ
/
ꢃ145.2 Hz,
/
Rꢃ1:987 cal Kꢀ1 molꢀ1
where Tc (:293 K) is the coalescence temperature, and
Dn (ꢃ194 Hz at 233 K) is the separation in Hertz of the
31PRh signals at the slow exchange limit. Thus, DG"
14 kcal molꢀ1. This value is greater than the average
activation barrier (109
2 kcal molꢀ1) for Berry pseu-
2P); ꢀ
/
ꢃ
/
713.3 Hz, 2P). The rhodiumꢁ
/
/
the positive electrospray mass spectrum by the main
/
monocationic fragment [RhI(cod)(1)]ꢂ at m/zꢃ
/
851.1.
The 1JPRh coupling constant has the order of magnitude
(14595 Hz) reported for equivalent P atom(s) in related
squareꢁ
:
/
/
/
/
planar complexes: RhX(diene)P [15a], RhXP?P2
dorotation in related five-coordinated d8 rhodium com-
plexes [19c]. The putative intermediate A, in equilibrium
[18], or [Rh(diene)P2]ꢂ [3d,3e,6b,15a,20]. This suggests
with 2×
therefore lie ca. 4 kcal molꢀ1 higher in energy than 2×
and 3×Cl.
/
I or 3×
/
Cl via a Berry transition state, should
that the geometry of the Rh(cod)P2 unit in 4 is close to
/
I
squareꢁplanar as well. However, the unusual deshield-
/
ing of the 31P and olefinic 1H chemical shifts for a
(RCH2Ph2P)2RhI(cod) unit beyond 30 and 5.5 ppm,
respectively, could also be indicative of a non disso-
ciated Rhꢀ ꢀ ꢀI center [3d,3e,6b,20], with a halide cis to
the cod ligand and trans to the PPh2 termini [6b]. NMR
and MS analyses thus support a flattened square-
pyramidal structure for 4 in a solvent of low polarity
such as CDCl3, with a bulky iodide anion associated in
an apical position.
/
3.1.2. Diopium:rhodium stoichiometryꢃ
/2:1
In biphasic CH2Cl2ꢁH2O mixture in the presence of a
/
slight excess tetrafluoroborate, a single diopium com-
plex was formed from [RhCl(cod)]2 and four equivalent
of 1×
/
I. It can be formulated as 4 on the basis of 31P-
1
23.11 (s); 31.87 (d, JPRh
(Scheme 3). A complex with similar spectroscopic
NMR: dꢃ
/
ꢃ144.9 Hz)
/

184
L. Viau, R. Chauvin / Journal of Organometallic Chemistry 654 (2002) 180ꢁ186
/
This interpretation was confirmed by the preparation
of the analogous complex in the absence of iodide ion.
Diopium hexafluoroposphate 1×PF6 was prepared by
I with an excess [NH4][PF6], and then
reacted with [RhCl(cod)]2 in the presence of [NH4][PF6].
A novel cationic rhodium complex 5 was formed which
metric reduction of ketones by chromium hydrides [28].
The 1:1 acidꢁbase reducing mixture was also selected in
order to lower the acidity of the medium, and thus
prevent an eventual acid-catalyzed ring opening of the
dioxolane ring of the diopium ligand 1.
/
/
metathesis of 1×
/
The in situ 5:1 1×/I:[Rh(cod)Cl]2 system in 3% catalytic
exhibited spectroscopic characteristics different to those
5 ppm) and 31P chemical shifts
ratio [27], and five equivalents of 1:1 HCOOH:NEt3
mixture in THF at 40 8C for 9 h, allowed (Z)-a-
acetamidocinnamic acid 6a to be reduced to (R)-7a in
54% conversion and 10% ee (Scheme 4). A kinetic
monitoring showed that in this case, the reaction was
completed in 5 h. The use of ten equivalents of the
reducing agent did not improve much the conversion
(ca. 60%), whatever was overall concentration of the
substrate: a competing deactivation of the rhodium
catalyst by the formiate mixture with a kinetic order
similar to that of the reduction process [29], would
account for the leveling of the kinetic curve below
complete conversion.
1
of 4: the olefinic H (B
/
of 5 (B
cationic (RCH2Ph2P)2RhI(cod) unit [3d,3e,6b,20]: d1H
4.46 (2 H), 4.92 (2 H). d31P ꢃ22.05 (s);16.45 (d,
1JRhP
144.0 Hz). NMR data show a perfect C2
/
20 ppm) now lye in the classical range for a
/
ꢃ
/
/
/
ꢃ
/
symmetry of the complex, with characteristic well
1
resolved H-NMR pattern for one of the two diaster-
eotopic types of phenyl rings of the Ph?(Ph)PÃ
/
Rh units
(Section 2). This pattern is not observed in 4 and
suggests here a frozen orientation of the corresponding
RhÃ
/
PÃ
/
Ph? sequence.
3.2. Catalytic properties
Under the same conditions, itaconic acid 6b was
successfully reduced to methylsuccinic acid 7b in 100%
conversion, but without significant ee By contrast, the a-
keto function of pyruvic and phenylglyoxylic acids was
not reduced at all.
The effect of the diluting solvent on the reduction of
6a was studied (Table 1). While low conversion occurred
in methanol and DMSO, a better conversion (85%) and
a better and reversed enantioselectivity (14% in (S)-7a)
were obtained in acetonitrile.
Both the preformed complex 5 and the in situ 5:1 1×
I:[Rh(cod)Cl]2 system were tested in catalytic hydro-
genation of (Z)-a-acetamidocinnamic acid 6a under 1
bar H2. Complex 5 was not active, and this is reminis-
cent of the peculiar inactivity in CÄC hydrogenation of
the closely related structural analog [Rh(cod)(P-
/
/
MePh2)2][PF6] [21]. The in situ catalyst, however, was
more active, and smoothly afforded (ꢀ)-N-acetyl-(R)-
/
phenylalanine 7a in 63% conversion and 5% ee after 22
h. This rather low ee is similar to those given by many
monodentate chiral ligands, but the cationic character of
1 could exhibit specific effects in the presence of polar
reducing agents used in transfer hydrogenation.
Surprisingly, the reducing system in THF is specific of
the monophosphine ligand: no conversion was observed
with diop as ligand. This observation is consistent with
Rocha Gonsalves’ report that the activity of the diopꢁ
/
rhodium catalyst requires the use of DMSO as solvent
[27]. It also suggests that the active species for this
Since early exploratory works on the use of
rutheniumꢁdiop catalysts for asymmetric hydrogen
/
reducing system requires a trans PÃ
/
RhÃ
/
P arrangement.
transfer from alcohols to itaconic acid [22] and acet-
ophenone [23], ruthenium-catalyzed asymmetric hydro-
gen transfer to acrylic acid derivatives [24] and ketones
[25] has been made highly efficient through the use of a
Although no mechanistic study has been tackled (con-
trary to the simplicity of the medium in direct hydro-
genation procedures, the complexity medium in the
hydrogen transfer procedure with excess HCOOH and
NEt3, prevents in situ 1H-NMR studies of the rhodiumꢁ
/
5:2 formic acidꢁtriethylamine azeotropic mixture. Few
/
applications of this system were also reported with
rhodium and iridium catalysts [24a,26]. Rocha Gon-
salves and coworkers showed that the 5:2 HCOOH:-
NEt3 mixture can be used for the reduction of N-acetyl-
6 interaction), the mechanism is therefore surely quite
different to those proposed for chelating ligands-cod-
rhodium catalysts [30]. Despite the remote position of
the chirality center of diopium with respect to the
catalytic center, the low but significant ee’s suggest
dehydroaminoacids with chiral diphosphineꢁrhodium
/
catalysts in DMSO as a specific solvent [27]. For
example, with (R,R)-diop as ligand, (Z)-a-acetamido-
cinnamic acid 6a was converted to (R)-N-acetyl-pheny-
lalanine 7a in 100% conversion (18 h, 3% catalytic ratio)
and 50% ee The latter substrates and polar reducing
agent are a priori suited for favoring electrostatic
interactions around the catalytic center. In order to
avoid the use of DMSO, an unconvenient and oxidizing
solvent [27], we considered a 1:1 HCOOH:NEt3 mixture
in THF at 40 8C, recently described for the stoichio-
Scheme 4. Hydrogen transfer catalysis by the in situ 2.5:1 diopiumꢁ
rhodium association.
/

L. Viau, R. Chauvin / Journal of Organometallic Chemistry 654 (2002) 180ꢁ
/186
185
[2] (a) H.B. Kagan, in: E.N. Jacobsen, A. Pfaltz, H. Yamamoto
(Eds.), Comprehensive Asymmetric Catalysis, vol. 1 (Chapter 2),
Springer, Heidelberg, 1999, p. 9.J;
Table 1
Catalytic reduction of (Z)-a-acetamidocinnamic acid 6a and itaconic
acid 6b (0.7ꢁ
/
1 mmol) with the in situ 1:5 [RhCl(cod)]2 ꢁ
/
1×I system
(b) M. Brown, in: E.N. Jacobsen, A. Pfaltz, H. Yamamoto (Eds.),
Comprehensive Asymmetric Catalysis, vol.
1 (Chapter 5),
Substrate Reducing agent/sub- Solvent Conversion
strate
Percent ee
of 7
Springer, Heidelberg, 1999, p. 121.
[3] (a) F. Robert, D. Sinou, J. Organomet. Chem. 604 (2000) 99;
(b) A. Borner, Eur. J. Inorg. Chem. (2001) 327;
(%)
a
6a
1 bar H2
MeOH 63
No
5 (S)
3 (R)
¨
(c) J. Holz, R. Kadyrov, S. Borns, D. Heller, A. Borner, J.
Organomet. Chem. 603 (2000) 61;
b
b
b
b
5:5 HCO2H:NEt3
5:5 HCO2H:NEt3
5:5 HCO2H:NEt3
5:5 HCO2H:NEt3
10:10 HCO2H:-
41
MeOH 15
MeCN 85
¨
ꢁ/
14 (S)
10 (R)
10 (R)
(d) A. Borner, J. Waed, K. Kortus, H.B. Kagan, Tetrahedron
¨
Asymmetry 4 (1993) 2219;
THF
THF
54
60
(e) L.B. Fields, E.N. Jacobsen, Tetrahedron Asymmetry 4 (1993)
b
NEt3
10:10 HCO2H:-
2229.
[4] P. Leglaye, B. Donnadieu, J.-J. Brunet, R. Chauvin, Tetrahedron
Lett. 39 (1998) 9179.
6b
THF
100
0
b
NEt3
[5] (a) W.S. Knowles, M.J. Sabacky, J. Chem. Soc. Chem. Commun.
(1968) 1445;
a
Hydrogenation. [Rh]2/1×I/6aꢃ1/5/100, concentration of 6a: 0.13
M in methanol, 208C, 22 h. Optical rotation of 7a: [a]Dꢃꢂ2.28
(cꢃ1.0, EtOH).
(b) L. Horner, H. Siegel, H. Buthe, Angew. Chem. Int. Ed. Engl. 7
¨
(1968) 942.
b
Hydrogen transfer. [Rh]2/1×I/6ꢃ1/5/66, concentration of 6: 0.13
[6] (a) H.B. Kagan, T.P. Dang, J. Am. Chem. Soc. 94 (1972) 6429;
(b) D. Sinou, H.B. Kagan, J. Organomet. Chem. 114 (1976) 325.
[7] (a) M. Sawamura, Y. Ito, Chem. Rev. 92 (1992) 857;
(b) C.R. Landis, R. Sawyer, E. Somsook, Organometallics 19
(2000) 994;
M, 48C, 9 h.
that the Ph2PꢂMe terminus of the monophosphine 1
could play a role just as does the R1R2PÄ
/O terminus of
(c) G.J. Rowlands, Tetrahedron 57 (2001) 1865.
[8] (a) I. Yamada, M. Ohkouchi, M. Yamaguchi, T. Yamagishi, J.
Chem. Soc. Perkin Trans. 1 (1997) 1869;
chiral b-aminophosphine oxide ligands in transfer
hydrogenation of ketones by ruthenium catalysts [31].
The efficiency of the putative electrostatic interaction
should, however, be enhanced in the more rigid envir-
onment of the homologous methylbinapium ligand [4], a
Hayashi-type phosphonium-MOP ligands [32]. In pre-
liminary experiments, however, the conditions used for
(b) I. Yamada, K. Fukui, Y. Aoki, S. Ikeda, M. Yamaguchi, T.
Yamagishi, J. Organomet. Chem. 539 (1997) 115;
(c) I. Yamada, M. Yamaguchi, T. Yamagishi, Tetrahedron
Asymmetry 7 (1996) 3339.
[9] (a) C. Claver, E. Fernandez, A. Gillon, K. Heslop, D.J. Hyett, A.
Martorell, A.G. Orpen, P.G. Pringle, Chem. Commun. (2000)
961;
the complexation of 1ꢁ4 were not suitable for the
/
bulkier binapium ligand, and an ylide strategy can be
devised [33].
(b) M.T. Reetz, G. Mehler, Angew. Chem. Int. Ed. Engl. 39
(2000) 3889.
[10] (a) I.T. Horvath (Ed.), J. Mol. Catal. A 116 (1997);
(b) P. Kalck, F. Monteil, in: F.G.A. Stone, R. West (Eds.), Adv.
Organomet. Chem., vol. 34, Academic Press, San Diego, 1992, p.
219;
4. Conclusion
(c) W. Beck, U. Nagel, Z. Anorg. Allg. Chem. 458 (1979) 22;
(d) I. Toth, B.E. Hanson, Tetrahedron Asymmetry 1 (1990) 895.
[11] R. Chauvin, Eur. J. Inorg. Chem. (2000) 577.
[12] J.J. Brunet, R. Chauvin, G. Commenges, B. Donnadieu, P.
Leglaye, Organometallics 15 (1996) 1752.
It has been shown that for the cationic chiral
phosphine 1, the use of a suitable polar reducing reagent
lead to improved activity and enantioselectivity with
respect to the use of molecular hydrogen. These
preliminary results pave the way for further explorations
of the catalytic properties of chiral cationic monopho-
sphine ligands. Alternatively, instead of rather soft
phosphonium group, harder cationic ammonium group
could also could also be considered by resorting to chiral
ammoniophosphinite ligands (ephosiums, valphosiums)
[12,34], or to quaternary ammonium derivatives of
Sinou’s diop-related amino-monophosphines [3a]. The
[13] H. Brunner, R. Storiko, Eur. J. Inorg. Chem. (1998) 783.
[14] I. Kovacs, M.C. Baird, J. Organomet. Chem. 502 (1995) 87 (and
references therein).
[15] (a) E. Renaud, R.B. Russel, S. Fortier, S.J. Brown, M.C. Baird, J.
Organomet. Chem. 419 (1991) 403;
(b) E. Renaud, M.C. Baird, J. Chem. Soc. Dalton Trans. (1992)
2905.
[16] G. Giordano, R.H. Crabtree, R.M. Heintz, D. Forster, D.E.
Morris, Inorganic Synthesis, vol. 28, Wiley, New York, 1991, p.
88.
[17] J.-D. Chen, F.A. Cotton, J. Am. Chem. Soc. 113 (1991) 2509.
[18] (a) T.H. Brown, P.J. Green, J. Am. Chem. Soc. 92 (1970) 2359;
(b) C.J. Elsevier, B. Kowall, H. Kragten, Inorg. Chem. 34 (1995)
4836.
latters would be chiral versions of Baird’s (IV)ꢁamphos
/
ligands [35].
[19] (a) K. Vrieze, H.C. Volger, A.P. Praat, J. Organomet. Chem. 14
(1968) 185;
(b) H. Friebolin, Basic One- and Two Dimensional NMR
References
Spectroscopy, VCH, Weinheim, 1991, pp. 269ꢁ271;
/
(c) P. Meakin, J.P. Jesson, J. Am. Chem. Soc. 95 (1973) 7272.
[20] (a) U. Wertz, H.-A. Brune, J. Organomet. Chem. 363 (1989) 377;
(b) R.R. Schrock, J.A. Osborn, J. Am. Chem. Soc. 93 (1971) 2397.
[1] T.P. Dang, H.B. Kagan, J. Chem. Soc. Chem. Commun. (1971)
481.

186
L. Viau, R. Chauvin / Journal of Organometallic Chemistry 654 (2002) 180ꢁ186
/
[21] (a) S. Sato, I. Matsuda, M. Shibata, J. Organomet. Chem. 377
(1989) 347;
[28] J.-J. Brunet, R. Chauvin, P. Leglaye, Eur. J. Inorg. Chem. (1999)
713.
(b) T. Muraoka, I. Matsuda, K. Itoh, J. Am. Chem. Soc. 122
(2000) 9552.
[29] S.H. Strauss, K.H. Whitemire, D.F. Shriver, J. Organomet.
Chem. 174 (1979) C59.
[22] K. Yoshinaga, T. Kito, K. Ohkubo, Bull. Chem. Soc. Jpn. 56
(1983) 1786.
[30] M. Bernard, V. Guiral, F. Delbecq, F. Fache, P. Sautet, M.
Lemaire, J. Am. Chem. Soc. 120 (1998) 1441.
[23] R. Chauvin, J. Mol. Catal. 62 (1990) 147.
[24] (a) H. Brunner, E. Graf, W. litner, K. Wutz, Synthesis (1989) 743.;
(b) J.M. Brown, H. Brunner, W. Leitner, M. Rose, Tetrahedron
Asymmetry 2 (1991) 331.
[31] A.M. Maj, K.M. Pietrusiewicz, I. Suisse, F. Abgossou, A.
Mortreux, Tetrahedron Asymmetry 10 (1999) 831.
[32] T. Hayashi, Acc. Chem. Res. 3 (2000) 354.
[33] L. Viau, C. Lepetit, G. Commenges, R. Chauvin, Organometallics
20 (2001) 808.
[25] A. Fujii, S. Haschiguchi, N. Uematsu, T. Ikariya, R. Noyori, J.
Am. Chem. Soc. 118 (1996) 2521.
[34] J.-J. Brunet, R. Chauvin, J. Chiffre, S. Huguet, P. Leglaye, J.
Organomet. Chem. 566 (1998) 117.
[26] (a) W. Leitner, J.M. Brown, H. Brunner, J. Am. Chem. Soc. 115
(1993) 152;
[35] (a) R.T. Smith, M.C. Baird, Organometallics 2 (1983) 1138;
(b) R.T. Smith, R.K. Ungar, L.J. Sanderson, M.C. Baird,
Organometallics 2 (1983) 1138;
(b) D.G.I. Petra, P.C.J. Kamer, A.L. Spek, H.E. Schoemaker,
P.W.N.M. Van Leeuwen, J. Org. Chem. 65 (2000) 3013.
[27] A.M.d’A. Rocha Gonsalves, J.C. Bayon, M.M. Pereira, M.E.S.
Serra, J.P.R. Pereira, J. Organomet. Chem. 553 (1998) 199.
(c) D.J. Brauer, J. Fischer, S. Kucken, K.P. Langhans, O. Stelzer,
N. Weferling, Z. Naturforsch. B Chem. Soc. 49 (1994) 1511.