Chiral rhodium complexes derived from electron-rich phosphine-phosphites as asymmetric hydrogenation catalysts

Article abstract of DOI:10.1021/om200933b

Two new chiral cationic rhodium(I) complexes derived from electron-rich dicyclohexylphosphine-phosphite ligands were prepared from enantiopure Sharpless epoxy ethers. The best-performing catalyst system, which bears a less bulky methyl ether moiety, exhibited remarkably high enantioselectivity (up to 99% ee) and reactivity (up to >2500 TON) in asymmetric hydrogenation reactions of various functionalized alkenes (α-(acylamino)acrylates, itaconic acid derivatives, α-substituted enol esters and α-arylenamides). Our synthetic methodology has been successfully applied to the enantioselective synthesis of the antiepileptic drug (R)-lacosamide (Vimpat).

Full text of DOI:10.1021/om200933b

Article
pubs.acs.org/Organometallics
Chiral Rhodium Complexes Derived From Electron-Rich
Phosphine-Phosphites as Asymmetric Hydrogenation Catalysts
Pablo Etayo,† Jose L. Nunez-Rico,† and Anton Vidal-Ferran*,†,‡
́
́
̃
†
̈
Institute of Chemical Research of Catalonia (ICIQ), Avgda. Paısos Catalans 16, 43007 Tarragona, Spain
‡Catalan Institution for Research and Advanced Studies (ICREA), Passeig Lluís Companys 23, 08010 Barcelona, Spain
S
* Supporting Information
ABSTRACT: Two new chiral cationic rhodium(I) complexes derived from electron-rich dicyclohexylphosphine-phosphite
ligands were prepared from enantiopure Sharpless epoxy ethers. The best-performing catalyst system, which bears a less bulky
methyl ether moiety, exhibited remarkably high enantioselectivity (up to 99% ee) and reactivity (up to >2500 TON) in
asymmetric hydrogenation reactions of various functionalized alkenes (α-(acylamino)acrylates, itaconic acid derivatives,
α-substituted enol esters and α-arylenamides). Our synthetic methodology has been successfully applied to the enantioselective
synthesis of the antiepileptic drug (R)-lacosamide (Vimpat).
diastereomeric transition states to improve the performance
of the catalytic systems. In this way, an array of enantiopure 1,2-
P-OP derivatives (mostly phosphine-phosphite ligands) have
been developed in an efficient manner for Pd-mediated
asymmetric allylic substitution reactions,5 Ir-mediated asym-
metric hydrogenation of heteroaromatic compounds,6 and Rh-
mediated asymmetric hydrogenation of functionalized alkenes.7
Regarding the latter asymmetric transformation, our group
recently reported the highly enantioselective hydrogenation of a
structurally diverse range of substrates catalyzed by cationic Rh
complexes of diphenylphosphine-phosphite ligand 1 (Figure 1).
INTRODUCTION
■
The design and development of novel, efficient, and modular
chiral phosphorus ligands and their application in several
transition-metal-mediated asymmetric transformations of inter-
est constitute an expanding research area within academia and
industry.1 From a practical perspective, enantioselective
hydrogenation offers several advantages for the asymmetric
organic synthesis of enantiomerically pure compounds (optimal
atom economy, broad substrate scope, high reactivity and
selectivity, and operational simplicity). As a consequence, the
asymmetric hydrogenation of prochiral substrates (alkenes,
imines, ketones, heteroaromatic compounds), mostly catalyzed
by chiral Ir, Rh, or Ru complexes, is certainly among the most
efficient and reliable methodologies in asymmetric catalysis.2
Accordingly, many industrial methods for the production of
optically active pharmaceuticals, agrochemicals, fragrances, fine
chemicals, and natural chemicals rely on catalytic asymmetric
hydrogenation reactions.3
Despite the remarkably advanced state of the field, numerous
research groups are still actively pursuing new catalytic systems
that show higher activity and/or improved enantioselectivity for
challenging substrates or for recently discovered pharmacolog-
ically active compounds. In this context, we described the
preparation of a library of enantiomerically pure P-OP ligands4
(phosphine-phosphinites and phosphine-phosphites) whose
catalytic properties were assessed in different enantioselective
transformations. Our research group has employed a highly
modular ligand design in conjunction with a fine-tuning
methodology guided by computational analysis of the
Figure 1. Previously developed “lead” P-OP ligand (left) and general
structure of the new Rh−(P-OP) complexes (right).
β-Substituted α-(acylamino)acrylates, itaconic acid derivatives
and analogues, and α-arylenamides have been hydrogenated
Received: October 5, 2011
Published: November 22, 2011
© 2011 American Chemical Society
6718
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
with high enantioselectivities (up to 99% ee) using [Rh(1)]+
complexes as catalysts.7c,d We envisioned that the highly
modular design of our phosphine-phosphite ligands should
enable further optimization of the catalytic systems by
modifying the steric and electronic properties of the phosphino
group (replacement of the previously featured PPh2 moiety by a
more bulky and more electron-rich PCy2 group). Furthermore,
the catalytic performance of Rh complexes derived from such
electron-rich phosphine-phosphite ligands (general structure
depicted in Figure 1) can be tuned by varying the steric bulk of
the substituent attached to the pendant ether fragment (methyl
(2a) and triphenylmethyl (2b) ether groups). We report here
the synthesis of two newly designed chiral cationic Rh(I)
complexes derived from electron-rich dicyclohexylphosphine-
phosphites and the evaluation of these compounds as catalyst
precursors in asymmetric hydrogenation reactions of function-
alized alkenes.
protected minor regioisomer 5b could not be isolated in a
chemically pure form.
The preparation of the target Rh complexes 8a,b was
successfully accomplished in three subsequent chromatogra-
phy-free synthetic steps starting from the borane-protected
phosphino alcohols 4a,b (Scheme 2). First, the free
Scheme 2. Synthesis of Rhodium Complexes Derived from
Electron-Rich Dicyclohexylphosphine-Phosphite Ligands
RESULTS AND DISCUSSION
■
Synthesis of [Rh(nbd)(Cy2P-OP)]BF4 (8a) and [Rh(nbd)-
(Cy2P-OP)]BF4 (8b). The stepwise synthetic route toward the
target Rh−(P-OP) complexes started with the ring opening of
enantiopure Sharpless methyl- and triphenylmethyl-substituted
epoxy ethers 3a,b by using lithium dicyclohexylphosphide
(generated in situ from HPCy2 and n-BuLi) as the phosphorus
nucleophile (Scheme 1). The ring opening of these epoxy
dicyclohexylphosphino alcohols 6a,b were quantitatively
obtained under optimized conditions by deprotection of the
borane adducts 4a,b using neat diethylamine with heating over
reflux at 90 °C for 4 h.8 Phosphino alcohols 6a,b were
subsequently derivatized to the phosphine-phosphite ligands
7a,b by treatment with (Sa)-BINOL-derived chlorophosphite
in the presence of triethylamine as an auxiliary base. The
O-phosphorylation reactions were monitored by 31P{1H} NMR
spectroscopy (using sealed capillaries filled with DMSO-d6 for
deuterium shimming and locking) in order to optimize the
reaction time required for the reactions to reach completion.
31P{1H} NMR analysis of the reaction mixture leading to the
phosphine-phosphite 7a showed the complete disappearance of
the phosphorus signal belonging to the starting phosphino
alcohol 6a (singlet at 12.4 ppm) after 3 h at room temperature.
During this time two doublets (4JP−P = 18.4 Hz) appeared, and
these correspond to the phosphino and phosphite moieties in
ligand 7a (13.3 and 158.4 ppm, respectively). In contrast,
31P{1H} NMR analysis of the reaction mixture with the bulkier
triphenylmethyl ether derivative after 3 h at room temperature
indicated incomplete conversion. Complete conversion was
achieved on stirring the reaction mixture overnight at room
temperature, and the resulting spectrum revealed two
doublets at 12.2 and 158.1 ppm with identical long-range
31P−31P coupling constants (4JP−P = 18.5 Hz). These signals
are consistent with the expected phosphino and phosphite
groups in ligand 7b.
Scheme 1. Ring Opening of Sharpless Epoxy Ethers with
Lithium Dicyclohexylphosphide
ethers proceeded smoothly at −30 °C to room temperature,
and the intermediate dicyclohexylphosphino alcohols, which
proved to be rather prone to oxidation, were subsequently
protected in situ as the corresponding borane adducts in order
to make handling and storage more convenient. The ring-
opening reactions of epoxides 3a,b with lithium dicyclohex-
ylphosphide were stereospecific―as previously observed for
other phosphorus nucleophiles7c―and took place in a highly
regioselective manner (marked preference for nucleophilic attack
1
at the more reactive benzylic position). H and 31P{1H} NMR
analyses of the crude mixtures indicated regioisomer ratios (rr)
of 94/6 and 96/4 for the methyl and triphenylmethyl ether
derivatives, respectively. Purification of the mixtures by column
chromatography afforded the desired major regioisomers 4a,b on
the gram scale in 87% and 71% yields, respectively. Further-
more, a very small amount of the methyl ether protected minor
regioisomer 5a was also isolated (3% yield) and this compound
was also fully characterized. However, the triphenylmethyl ether
The capability of the bidentate dicyclohexylphosphine-
phosphites 7a,b to form stable, well-defined chiral Rh chelates
was clearly demonstrated after converting them into the
cationic Rh(I) complexes 8a,b by using [Rh(nbd)2]BF4 as
the metal precursor.7c Complexes 8a,b were prepared in a
straightforward manner by reacting stoichiometric amounts of
the crude ligands 7a,b9 with the aforementioned Rh precursor
in DCM at room temperature for 2 h (Scheme 2). In this way,
6719
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
the target Rh−(P-OP) complexes 8a,b were obtained from 4a,b
in three steps (namely deprotection, O-phosphorylation, and
complexation), and they were isolated in 60% and 52% overall
yields, respectively. The 31P{1H} NMR spectra of 8a,b showed
two sharp doublet of doublets centered around 40 ppm
(phosphino moiety) and 140 ppm (phosphite moiety) for each
Rh complex. The multiplicity of these phosphorus signals is due
to a direct 31P−103Rh (1J ≈ 270 Hz) coupling along with a
geminal 31P−31P coupling (2J ≈ 60 Hz).10 The NMR data and
HRMS measurements for 8a,b unequivocally confirm the
formation of the expected 1/1 Rh/ligand chelates.
Comparison of the Catalytic Performance of Rh
Complexes 8a,b in Asymmetric Hydrogenation of
Functionalized Alkenes. Once the envisioned Rh(I)
complexes 8a,b had been prepared, their catalytic activity
and selectivity were evaluated in the enantioselective
hydrogenation of various prochiral olefins. In the first set
of experiments, several model substrates from different
representative families of functionalized alkenes were chosen
in order to compare the efficiency provided by the sterically
different Rh complexes 8a,b. For this purpose, one
β-unsubstituted (9a), one β-aryl-substituted (9c), and one
β-alkyl-substituted (9k) α-(acylamino)acrylate were selected
as test substrates along with dimethyl itaconate 10a, its
analogue 10c, and the enol ester phosphonate 11 (see
Table 1 for the substrate structures). The hydrogenation
reactions were carried out under standard screening con-
ditions7c,d (1.0 mol % of [Rh(nbd)(P-OP)]BF4 as the catalyst
precursor, 20 atm of H2, THF or DCM as solvent, room
temperature, overnight). The results of the comparative study,
along with those previously obtained with the Rh complex of
diphenylphosphine-phosphite 1 (ligand structure depicted in
Figure 1), are summarized in Table 1. In all cases complete
conversion was achieved for the functionalized alkenes under
investigation and the catalytic performance of the catalyst
precursors 8a,b, with sterically different substituents at the
R-oxy position, did not show differences in terms of the
catalytic activity. Nonetheless, a significant negative effect on
the reaction enantioselectivity can be observed on comparing
the less hindered methyl ether protected Rh complex 8a with
the bulkier triphenylmethyl ether counterpart 8b, which gave
rise to dramatic decreases in the ee values for α-(acylamino)-
acrylates 9a,c,k, dimethyl itaconate 10a, and the related alkene
10c. These results clearly indicate that the steric hindrance
arising from the different size of the R-oxy substituent in the
chiral backbone of the ligand has a significant influence on
the enantioselectivity of the reaction.11 One exception to this
general catalytic behavior was found for the enol ester phos-
phonate 11, which was hydrogenated with excellent enantio-
selectivity regardless of the Rh complex used as the catalyst
precursor (Table 1, entries 17 and 18).
Enantioselective Hydrogenation of Functionalized
Alkenes Catalyzed by Rh Complex 8a. Having obtained
very promising results from the preliminary catalyst screening
studies, we proceeded to investigate the efficiency of the best-
performing Rh complex 8a in more detail in the enantiose-
lective hydrogenation of α-(acylamino)acrylates (see Table 2
for the substrate structures). Regarding the optimal catalyst
loading of the Rh complex 8a, a catalyst loading as low as 0.04
mol % (2500/1 substrate/catalyst ratio, >2500 TON) still gave
quantitative conversion in the hydrogenation of model
substrate 9a, without the enantioselectivity being affected
(99% ee; compare entry 1 in Table 2 and entry 2 in Table 1)
and without requiring an increase in either the H2 pressure or
the reaction time. The N-Cbz-protected analogue 9b was also
efficiently hydrogenated under standard conditions (Table 2,
entry 2). In addition to the β-aryl-substituted α-(acylamino)-
acrylic acid derivative mentioned above (substrate 9c; Table 1,
entry 5), other substrates of this kind (9d−i) were also
successfully reduced with full conversions and remarkably high
enantioselectivities (91−98% ee; Table 2, entries 3−8).
Interestingly, two analogous substrates of the model alkene
9c, one in which the methyl ester functionality is replaced by a
free carboxylic acid group (9d) and the other a Weinreb amide
(9e), were also efficiently hydrogenated with high enantiose-
lectivity (Table 2, entries 3 and 4). The enantioselective
hydrogenation of the free acid 9d proceeded quantitatively
under standard conditions, whereas the more difficult substrate
9e required an increased catalyst loading (up to 2.0 mol %) in
order to reach full conversion. Despite the synthetic versatility
of the functional group known as the Weinreb amide (N-methoxy-
N-methylcarbamoyl group) as an antecedent of a number of
functional groups,12 the results described herein constitute, to
the best of our knowledge, the first reported example of the
enantioselective hydrogenation of the Weinreb amide derivative
9e. The presence of carbamate-type amino protecting groups as
well as electron-donating and electron-withdrawing substituents
at different positions of the aromatic ring were all well-tolerated
by the catalyst precursor 8a (Table 2, entries 5−8).
Table 1. Asymmetric Hydrogenation of Functionalized
Alkenes Catalyzed by Rh Complexes 8a,b
a
a
All reactions were run at room temperature with 1.0 mol %
catalyst loading for 18 h at 20 atm of H2. Full conversion was
b
1
achieved in all cases, as determined by H NMR. Determined by
c
GC or HPLC analysis using chiral stationary phases. Absolute
configuration was assigned by comparison of the specific rotation with
d
e
reported data. 1 refers to [Rh(nbd)(1)]BF4. Results from ref 7c.
f
Results from ref 7d.
6720
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
itaconic acid derivatives 10a−c gave high enantioselectivities
(Table 3, entries 1−3) of up to 97% ee for dimethyl itaconate 10a.
Table 2. Asymmetric Hydrogenation of α-
(Acylamino)acrylates Catalyzed by Rh Complex 8a
a
Table 3. Asymmetric Hydrogenation in DCM of other
a
Functionalized Alkenes Catalyzed by Rh Complex 8a
a
See Table 1 for details on the asymmetric hydrogenations.
The hydrogenations of the α-arylenol acetates 12a,b proved
to be less enantioselective (Table 3, entries 4 and 5), with
the p-nitrophenyl-substituted derivative 12b affording the
highest ee value. Finally, α-arylenamides 13a,b were both
hydrogenated with moderate enantioselectivity (Table 3,
entries 6 and 7).
a
b
See Table 1 for details on the asymmetric hydrogenations. The
absolute configuration was tentatively assigned by analogy, on the basis
of the stereochemical outcome for analogous substrates.
Enantioselective Synthesis of the Antiepileptic Drug
(R)-Lacosamide. To demonstrate the synthetic utility of our
methodology, we studied the catalytic performance of Rh complex
8a in the asymmetric hydrogenation of β-alkoxy-substituted
α-(acylamino)acrylate derivatives 14a−c (Scheme 3). It is worth
The excellent results achieved in the enantioselective hydro-
genation reactions of substrates 9j−m (Table 2, entries 9−12)
clearly indicate the good catalytic performance of the Rh
complex 8a for β-alkyl-substituted α-(acylamino)acrylates. At
the same time, the results obtained for these differently
N-protected β-methyl-substituted alkenes (9j−m) strongly
evidence the tolerance of the catalytic system to a broad
variety of amino protecting groups. It is also worth mentioning
that the industrially relevant alkene 9m, which is the precursor
of the antiseizure agent levetiracetam,13 has proven to be a
challenging substrate in Rh-mediated asymmetric hydro-
genation, as only moderate enantioselectivities have been
achieved even under optimized conditions.14 Fortunately,
the catalyst precursor 8a showed quite efficient behavior for
substrate 9m and gave the desired product with 94% ee
(Table 2, entry 12).
Scheme 3. Enantioselective Synthesis of the Antiepileptic
Drug (R)-Lacosamide
We further extended the substrate scope of the enantiose-
lective hydrogenation catalyzed by the Rh complex 8a to other
representative families of functionalized alkenes: namely, ita-
conic acid derivatives, α-substituted enol esters, and α-aryle-
namides. The results obtained in these reactions are
summarized in Table 3. The asymmetric hydrogenation of
mentioning that precedents for the enantioselective hydrogenation
of α-amino acid precursors bearing β-alkoxy substituents have not
previously been reported in the literature. Nevertheless, we
6721
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
successfully applied our strategy to this unexplored field and
developed an enantioselective approach for the synthesis of
(R)-lacosamide15 (15c, Scheme 3), the active pharmaceutical
ingredient of the antiepileptic agent Vimpat.16 The results
obtained in the hydrogenation reactions of lacosamide
precursors 14a−c are summarized in Table 4.
evolves to the corresponding elimination products 16 through
hydrogenation.
In comparison to the Rh complex derived from P-OP ligand
1, the use of complex 8a led to a marked enhancement in the
selectivity (i.e., 98/2 15a/16a ratio) and a slightly reduced
enantioselectivity (i.e., 94% ee) under the same hydrogenation
conditions, albeit with only moderate conversion (Table 4,
entry 2). The reaction was taken to completion by increasing
the catalyst loading to 2.0 mol % (Table 4, entry 3). The
formation of the side product 16a was almost suppressed by
switching the reaction solvent from THF to DCM. Further-
more, this reaction medium allowed the desired product 15a to
be obtained with very high enantioselectivity (96% ee; Table 4,
entry 4). Most remarkably, the asymmetric hydrogenation of
carboxylic acid derivative 14b in DCM, a solvent in which
substrate 14b was only partially soluble, provided lacosamide
precursor 15b as the exclusive reaction product with excellent
enantioselectivity (i.e., 99% ee; Table 4, entry 5). The two
R-configured hydrogenation products 15a,b constitute direct
precursors of the chiral drug (R)-lacosamide (15c). Indeed,
this chiral pharmaceutical target can be readily prepared in
one or two steps from 15b and 15a, respectively, according to
literature procedures.15c
Finally, the enantioselective total synthesis of (R)-lacosamide
15c was also attempted by exploring the Rh-mediated
asymmetric hydrogenation of the tailored substrate 14c (see
Scheme 3). The hydrogenation reaction of alkene 14c catalyzed
by Rh complex 8a proceeded with 96% selectivity toward the
desired product 15c, which was obtained in 96% ee when THF
was used as the solvent (Table 4, entry 6). These results were
significantly improved on using DCM as the reaction solvent.
In this way, the antiepileptic drug (R)-lacosamide (15c) was
efficiently prepared with total selectivity and excellent
enantioselectivity (i.e., 99% ee; Table 4, entry 7). The measured
specific rotation for 15c is in excellent agreement with the
reported value,15c as indicated in Scheme 3, unambiguously
confirming the absolute configuration of the hydrogenated
product.
Table 4. Asymmetric Hydrogenation of 14 Catalyzed by Rh
a
Complex 8a
amt of
cat.
15/16
(molar
ratio)
ee of
15 (%)
(config)
conversn
(%)
entry substrate solvent (mol %)
b
1
2
3
4
5
6
7
14a
14a
14a
14a
14b
14c
14c
THF
THF
THF
DCM
DCM
THF
DCM
1.0
1.0
2.0
2.0
2.0
2.0
2.0
>99
44
85/15
98/2
97 (R)
94 (R)
94 (R)
96 (R)
99 (R)
96 (R)
>99
>99
>99
>99
>99
97/3
99/1
>99/1
96/4
>99/1
99 (R)
a
b
See Table 1 for details on the asymmetric hydrogenations. Results
from ref 17 obtained on using in situ generated [Rh(nbd)(1)]BF4.
The enantioselective hydrogenation of the methyl ester
derivative 14a has very recently been performed using the Rh
complex of P-OP ligand 1 (see Figure 1 for the ligand structure),
and this furnished very good results (Table 4, entry 1) under
standard reaction conditions (1.0 mol % catalyst loading,
20 atm of H2, THF, room temperature, 18 h).17 However,
although the conversion of the starting alkene 14a reached
completion, a selectivity of only 85% (85/15 15a/16a ratio)
toward the desired product 15a was obtained when
diphenylphosphine-phosphite ligand 1 was employed. Accord-
ing to the literature,18 the formation of the MeOH-elimination
products 16 may be rationalized as a side reaction of the
hydride intermediates 19 involved in the hydrogenation
catalytic cycle of alkenes 14 mediated by Rh complexes of
P-OP ligands7c (see Scheme 4). An elimination process19 from
hydride intermediates 19 releases MeOH along with the Rh
complex of 2-acetamidoacrylic acid derivatives 21, which further
Scheme 4. Formation of the MeOH-Elimination Products 16 under Hydrogenation Conditions
6722
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
of the solvent under vacuum the resulting crude phosphine-phosphite
ligand 7a, obtained as a spongy white solid, was straightforwardly used
in the next step. By operating inside a glovebox, the previous crude
compound 7a (0.80 theoretical mmol, 74% chemical purity by
31P{1H} NMR) was dissolved in anhydrous DCM (8.0 mL) and then
[Rh(nbd)2]BF4 (216 mg, 0.58 mmol) was added to the reaction
mixture, which was stirred at room temperature for 2 h. After this
period the solvent was evaporated off under vacuum until its volume
was reduced approximately to one-fourth. Then, anhydrous Et2O
(30.0 mL) was slowly syringed over the remaining solution. In this
manner a great deal of orange solid was immediately formed inside the
orangish solution. The solvent was filtered off under an Ar atmosphere
by employing a filter cannula, and the resulting solid was washed with
additional volumes of anhydrous Et2O (2 × 10.0 mL). The solvent was
again filtered off by employing the same filter cannula. Drying of the
resulting solid under vacuum for several hours afforded 463 mg
(60% overall yield) of pure Rh(I) complex 8a as an orange powder.
Norbornadienerhodium(I) tetrafluoroborate complex of (11bSa)-4-
((1R,2S)-1-dicyclohexylphosphino-3-methoxy-1-phenylpropan-2-
yloxy)dinaphtho[2,1-d:1′,2′-f ][1,3,2]dioxaphosphepine 7a (8a):
[α]D25 = +114.9° (c 1.02, CH2Cl2); 1H NMR (500 MHz, CD2Cl2) δ 0.75−
2.15 (m, 22H, 18CH2(PCy2), 2CH(PCy2), and CH2(nbd)), 2.24−2.37
(m, 2H, CH2(PCy2)), 3.35 (s, 3H, OMe), 3.39 (ddd, J = 9.9, 5.5, and
2.2 Hz, 1H, CHHOMe), 3.44−3.50 (m, 1H, CHHOMe), 3.59 (dd, J =
13.2 and 2.2 Hz, 1H, CHPCy2), 4.05−4.11 (m, 2H, CHhead of bridge(nbd)
and CHalkene(nbd)), 4.17−4.21 (m, 1H, CHhead of bridge(nbd)), 5.00−
5.08 (m, 1H, CHOP), 5.62−5.68 (m, 1H, CHalkene(nbd)), 6.15−6.21
(m, 1H, CHalkene(nbd)), 6.47−6.53 (m, 1H, CHalkene(nbd)), 7.10−7.46
(m, 8H, Harom), 7.53 (d, J = 9.0 Hz, 1H, Harom), 7.55−7.69 (m, 3H,
CONCLUSION
■
In summary, we have developed a convenient synthetic route
for the preparation of new chiral cationic rhodium complexes
derived from tunable electron-rich dicyclohexylphosphine-
phosphite ligands with substituents of varying size in the
R-oxy position of the ligand. These rhodium complexes have
been efficiently applied as catalyst precursors in enantioselective
hydrogenation reactions of various functionalized alkenes such
as α-(acylamino)acrylates, itaconic acid derivatives, α-substi-
tuted enol esters, and α-arylenamides. The best-performing
catalyst system (8a) gave excellent enantioselectivities (up to
99% ee) and catalytic activities (>2500 TON) for a wide array
of substrates (24 examples). The present methodology has
been successfully employed for the preparation of the chiral
antiepileptic drug (R)-lacosamide. Further applications of these
rhodium complexes in other asymmetric transformations are
currently under investigation, and the results will be reported in
due course.
EXPERIMENTAL SECTION
■
General Considerations. All syntheses were carried out on using
chemicals as purchased from commercial sources unless otherwise
stated. All manipulations and reactions were run under an inert
atmosphere using anhydrous solvents, either in a glovebox or with
standard Schlenk-type techniques. Glassware was dried under vacuum
and heated with a hot air gun before use. All solvents were dried by
using a Solvent Purification System (SPS). Silica gel 60 (230−
400 mesh) was used for column chromatography. NMR spectra were
recorded in CDCl3 unless otherwise cited, using a 400 or 500 MHz
H
arom), 7.75 (d, J = 8.8 Hz, 1H, Harom), 8.04 (d, J = 8.4 Hz, 1H, Harom),
8.09 (d, J = 9.0 Hz, 1H, Harom), 8.15 (d, J = 8.4 Hz, 1H, Harom), 8.33
(d, J = 8.8 Hz, 1H, Harom); 13C{1H31P} NMR (125 MHz, CD2Cl2) δ
25.9 (CH2(PCy2)), 26.1 (CH2(PCy2)), 26.6 (CH2(PCy2)), 26.9 (CH2-
(PCy2)), 27.1 (CH2(PCy2)), 27.2 (CH2(PCy2)), 28.2 (CH2-
(PCy2)), 28.3 (CH2(PCy2)), 30.2 (CH2(PCy2)), 35.1 (CH2(PCy2)),
35.4 (CH(PCy2)), 36.9 (CH(PCy2)), 38.0 (CHPCy2), 55.5
(CHhead of bridge(nbd)], 55.6 (CHhead of bridge(nbd)), 59.0 (OMe), 72.6
(CH2(nbd)), 72.9 (CH2OMe), 80.8 (CH−OP), 81.4 (CHalkene(nbd)),
89.8 (CHalkene(nbd)), 101.1 (CHalkene(nbd)), 104.2 (CHalkene(nbd)),
120.5 (CHarom), 121.2 (CHarom), 122.0 (Cq arom−(O)Cq arom), 123.0
(Cq arom−(O)Cq arom), 126.0 (CHarom), 126.3 (CHarom), 126.5 (CHarom),
126.9 (CHarom), 127.1 (CHarom), 127.3 (CHarom), 128.2 (CHarom), 128.6
(CHarom), 128.8 (CHarom), 128.9 (2CHarom), 130.7 (CHarom), 131.5
(CHarom), 131.8 (Cq arom), 131.9 (Cq arom), 132.2 (Cq arom), 132.4
(Cq arom), 132.5 (Cq arom), 146.5 (Cq arom−O), 146.8 (Cq arom−O);
1
spectrometer. H NMR and 13C NMR chemical shifts are quoted in
ppm relative to residual solvent peaks, whereas 31P{1H} NMR
chemical shifts are quoted in ppm relative to 85% phosphoric acid in
water and 19F{1H} NMR chemical shifts are quoted in ppm relative to
BF3·OEt2 in CDCl3. High-resolution mass spectra (HRMS) were
recorded by using an ESI ionization method in positive mode. Melting
points were measured in open capillaries and are uncorrected.
Enantiomeric excesses were determined by GC or HPLC on using
chiral stationary phases.
Synthesis of [Rh(nbd)(7a)]BF4 (8a). The reaction mixture was
prepared into an Ace pressure tube loaded with a stirring bar by
working under an N2 atmosphere inside a glovebox. First, the borane
adduct 4a (300 mg, 0.80 mmol) was charged into the pressure tube
and then HNEt2 (16.4 mL, 160 mmol) was syringed inside. Once the
pressure tube had been sealed under an inert atmosphere, it was taken
out of the glovebox and the reaction mixture was further heated to
90 °C and stirred for 4 h. After the mixture was cooled to room
temperature, the pressure tube was introduced into a glovebox, where
the reaction mixture was filtered through a short deoxygenated silica
gel pad (ca. 3.0 cm) on employing anhydrous toluene (2 × 10.0 mL)
as the solvent for the elution. The filtrate was collected in a Schlenk
flask under an inert atmosphere, and once outside the glovebox, the
solvent was evaporated off under vacuum. The resulting phosphino
alcohol 6a, obtained as a white solid, was subsequently introduced into
a glovebox in order to carry out the next step therein. The freshly
prepared phosphino alcohol 6a (0.80 mmol theoretical) was dissolved
in anhydrous toluene (20.0 mL). On the other hand, commercially
available (Sa)-BINOL-derived chlorophosphite (317 mg, 0.88 mmol)
was loaded into a vial and then anhydrous toluene (10.0 mL) and
distilled Et3N (0.22 mL, 1.59 mmol) were successively syringed into
the vial. The previous solution was syringed drop by drop over the
solution containing the starting phosphino alcohol. After completion
of the addition, the reaction mixture was stirred for 3 h at room
temperature under an N2 atmosphere inside the glovebox. The
reaction mixture was treated by addition of anhydrous Et2O (2 ×
10.0 mL). The resulting whitish suspension was filtered through a
short deoxygenated silica gel pad (ca. 3.0 cm), and the filtrate was
collected in a Schlenk flask under inert atmosphere. After evaporation
1
31P{1H} NMR (202 MHz, CD2Cl2) δ 38.5 (dd, JP−Rh = 141.4 Hz,
1
2
2JP−P = 59.8 Hz, 1P, P−C), 141.1 (dd, JP−Rh = 272.9 Hz, JP−P
=
59.8 Hz, 1P, P−O); 11B{1H} NMR (128 MHz, CD2Cl2) δ −1.16
(s, BF4−); 19F{1H} NMR (376 MHz, CD2Cl2) δ −153.2 (s, BF4);
HRMS (ESI+): m/z [M − BF4]+ calcd for C49H54O4P2Rh 871.2552,
found 871.2565.
Synthesis of [Rh(nbd)(7b)]BF4 (8b). The reaction mixture was
prepared into an Ace pressure tube loaded with a stirring bar by
working under an N2 atmosphere inside a glovebox. First, the borane
adduct 4b (482 mg, 0.80 mmol) was charged into the pressure tube,
and then HNEt2 (16.4 mL, 160 mmol) was syringed inside. Once the
pressure tube had been sealed under an inert atmosphere, it was taken
out of the glovebox and the reaction mixture was further heated at
90 °C and stirred for 4 h. After the mixture was cooled to room
temperature, the pressure tube was introduced into a glovebox, where
the reaction mixture was filtered through a short deoxygenated silica
gel pad (ca. 3.0 cm) on employing anhydrous toluene (2 × 10.0 mL)
as the solvent for the elution. The filtrate was collected in a Schlenk
flask under an inert atmosphere, and once outside the glovebox, the
solvent was evaporated off under vacuum. The resulting phosphino
alcohol 6b, obtained as a white solid, was subsequently introduced into
a glovebox in order to carry out the next step therein. The freshly
prepared phosphino alcohol 6b (0.80 mmol theoretical) was dissolved
in anhydrous toluene (20.0 mL). On the other hand, commercially
available (Sa)-BINOL-derived chlorophosphite (317 mg, 0.88 mmol)
6723
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
was loaded into a vial and then anhydrous toluene (10.0 mL) and
distilled Et3N (0.22 mL, 1.59 mmol) were successively syringed into
the vial. The previous solution was syringed drop by drop over the
solution containing the starting phosphino alcohol. After completion
of the addition, the reaction mixture was stirred for 18 h at room
temperature under an N2 atmosphere inside the glovebox. The
reaction mixture was treated by addition of anhydrous Et2O (2 ×
10.0 mL). The resulting whitish suspension was filtered through a
short deoxygenated silica gel pad (ca. 3.0 cm), and the filtrate was
collected in a Schlenk flask under an inert atmosphere. After
evaporation of the solvent under vacuum the crude phosphine-
phosphite ligand 7b, obtained as a spongy white solid, was
straightforwardly used in the next step. By operating inside a glovebox,
the crude compound 7b (0.80 mmol theoretical, 65% chemical purity
by 31P{1H} NMR) was dissolved in anhydrous DCM (8.0 mL) and
then [Rh(nbd)2]BF4 (190 mg, 0.51 mmol) was added to the reaction
mixture, which was stirred at room temperature for 2 h. After this
period the solvent was evaporated off under vacuum until its volume
was reduced approximately to one-fourth. Then, anhydrous Et2O
(30.0 mL) was slowly syringed over the remaining solution. In this
manner a great deal of orange solid was immediately formed inside the
orangish solution. The solvent was filtered off under an Ar atmosphere
by employing a filter cannula, and the resulting solid was washed with
additional volumes of anhydrous Et2O (2 × 10.0 mL). The solvent was
again filtered off by employing the same filter cannula. Drying of the
resulting solid under vacuum for several hours afforded 496 mg (52%
overall yield) of pure Rh(I) complex 8b as an orange powder.
Norbornadienerhodium(I) tetrafluoroborate complex of (11bSa)-4-
((1R,2S)-1-dicyclohexylphosphino-1-phenyl-3-triphenylmethoxypro-
pan-2-yloxy)dinaphtho[2,1-d:1′,2′-f ][1,3,2]dioxaphosphepine 7b (8b):
autoclave reactor (HEL Cat-24 parallel pressure multireactor). The
autoclave was purged three times with H2 gas at 10 atm, and finally,
the autoclave was pressurized under 20 atm of H2 gas. The reaction
mixture was stirred at room temperature for 18 h (overnight reaction).
The autoclave was subsequently depressurized slowly, and further the
reaction mixture was filtered through a short pad of SiO2 and eluted
with EtOAc (1.0 mL). The resulting solution was concentrated under
vacuum, and thus the conversion was determined by 1H NMR analysis
and the enantiomeric excess was determined by GC or HPLC analysis
on using chiral stationary phases.
ASSOCIATED CONTENT
* Supporting Information
■
S
Text and figures giving experimental procedures, analytical and
spectral characterization data for ligand precursors, substrates,
and products, analysis data for the enantioselectivities of
hydrogenation products, and NMR spectra of new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: avidal@iciq.es.
■
ACKNOWLEDGMENTS
■
We thank the MICINN (Grant CTQ2008-00950/BQU),
DURSI (Grant 2009GR623), Consolider Ingenio 2010
(Grant CSD2006-0003), and the ICIQ Foundation for financial
support. J.L.N.-R. thanks the ICIQ Foundation for a
predoctoral fellowship. We thank Prof. Dr. P. W. N. M. van
Leeuwen for helpful discussions.
26
1
[α]D = +34.4° (c 0.85, CH2Cl2); H NMR (400 MHz, CD2Cl2) δ
0.75−2.19 (m, 23H, 9.5CH2(PCy2), 2CH(PCy2), and CH2(nbd)),
2.39−2.50 (m, 1H, 0.5CH2(PCy2)), 3.21 (dd, J = 9.1 and 9.1 Hz, 1H,
CHHOMe), 3.29−3.37 (m, 1H, CHHOMe), 3.93 (bd, J = 13.8 Hz,
1H, CHPCy2), 4.05−4.15 [m, 2H, CHhead
bridge(nbd) and
of
CHalkene(nbd)], 4.19−4.26 [m, 1H, CHhead of bridge(nbd)], 4.82−4.95
(m, 1H, CHOP), 5.59−5.68 (m, 1H, CHalkene(nbd)), 6.12−6.19 (m,
1H, CHalkene(nbd)), 6.45−6.54 (m, 1H, CHalkene(nbd)), 6.90 (d, J =
8.8 Hz, 1H, Harom), 7.06−7.63 (m, 26H, Harom), 7.70 (d, J = 8.8 Hz,
1H, Harom), 7.79 (d, J = 8.8 Hz, 1H, Harom), 7.97 (d, J = 8.3 Hz, 1H,
REFERENCES
■
(1) For recent reviews, see for example: (a) Phosphorus Ligands in
Asymmetric Catalysis, 1st ed.; Borner, A., Ed.; Wiley-VCH: Weinheim,
Germany, 2008; Vols. I−III. (b) Teichert, J. F.; Feringa, B. L. Angew.
Chem., Int. Ed. 2010, 49, 2486. (c) Gladiali, S.; Alberico, E.; Junge, K.;
Beller, M. Chem. Soc. Rev. 2011, 40, 3744. (d) Luhr, S.; Holz, J.;
Borner, A. ChemCatChem 2011, 3, 1708.
(2) For recent reviews, see for example: (a) Church, T. L.; Andersson,
P. G. Coord. Chem. Rev. 2008, 252, 513. (b) Fleury-Breg
̈
H
arom), 8.12 (d, J = 8.3 Hz, 1H, Harom), 8.30 (d, J = 8.8 Hz, 1H, Harom);
13C{1H31P} NMR (125 MHz, CD2Cl2) δ 25.8 (CH2(PCy2)), 26.4
(CH2(PCy2)), 26.6 (CH2(PCy2)), 26.9 (2CH2(PCy2)), 27.2
(CH2(PCy2)), 28.3 (CH2(PCy2)), 28.6 (CH2(PCy2)), 30.6
(CH2(PCy2)), 34.7 (CH2(PCy2)), 35.5 (CH(PCy2)), 36.6 (CH(PCy2)),
37.5 (CHPCy2), 55.6 (CHheadofbridge(nbd)), 55.7 (CHheadofbridge(nbd)),
62.3 (CH2OCPh3), 73.0 (CH2(nbd)), 80.8 (CHalkene(nbd)), 81.1
(CHOP), 87.6 (OCPh3), 90.1 (CHalkene(nbd)), 101.5 (CHalkene(nbd)),
104.5 (CHalkene(nbd)), 120.3 (CHarom), 120.5 (CHarom), 122.0
(Cq arom−(O)Cq arom), 122.8 (Cq arom−(O)Cq arom), 126.0 (CHarom),
126.3 (CHarom), 126.5 (CHarom), 126.9 (CHarom), 127.0 (CHarom),
127.2 (CHarom), 127.4 (CHarom), 127.8 (CHarom), 128.0 (CHarom),
128.2 (CHarom), 128.4 (CHarom), 128.5 (CHarom), 128.6 (CHarom),
128.9 (CHarom), 131.0 (CHarom), 131.6 (CHarom), 131.9 (Cq arom), 132.0
(Cq arom), 132.1 (Cq arom), 132.2 (Cq arom), 132.5 (Cq arom), 143.1
(Cq arom), 146.4 (Cq arom−O), 146.4 (Cq arom−O); 31P{1H} NMR
̈
̈
́
eot, N.; de la
Fuente, V.; Castillon, S.; Claver, C. ChemCatChem 2010, 2, 1346.
́
(c) Xie, J.-H.; Zhu, S.-F.; Zhou, Q.-L. Chem. Rev. 2011, 111, 1713.
(d) Woodmansee, D. H.; Pfaltz, A. Chem. Commun. 2011, 47, 7912.
(3) For leading references, see: (a) Blaser, H.-U.; Pugin, B.; Spindler,
F.; Thommen, M. Acc. Chem. Res. 2007, 40, 1240. (b) Nugent, T. C.;
El-Shazly, M. Adv. Synth. Catal. 2010, 352, 753. (c) Green Chemistry in
the Pharmaceutical Industry, 1st ed.; Dunn, P. J.; Wells, A. S.; Williams,
M. T., Eds.; Wiley-VCH: Weinheim, Germany, 2010. (d) Busacca,
C. A.; Fandrick, D. R.; Song, J. J.; Senanayake, C. H. Adv. Synth. Catal.
2011, 353, 1825.
(4) For a comprehensive review describing the synthesis of this kind
of phosphorus ligand and their applications in asymmetric catalysis,
1
2
(162 MHz, CD2Cl2) δ 36.9 (dd, JP−Rh = 140.7 Hz, JP−P = 60.4 Hz,
1
2
see: Fernan
Chem. Rev. 2011, 111, 2119.
(5) Panossian, A.; Fernandez-Per
Tetrahedron: Asymmetry 2010, 21, 2281.
(6) Nunez-Rico, J. L.; Fernandez-Per
Vidal-Ferran, A. Organometallics 2010, 29, 6627.
(7) (a) Fernandez-Perez, H.; Pericas, M. A.; Vidal-Ferran, A. Adv. Synth.
Catal. 2008, 350, 1984. (b) Donald, S. M. A.; Vidal-Ferran, A.; Maseras,
F. Can. J. Chem. 2009, 87, 1273. (c) Fernandez-Perez, H.; Donald, S. M.
A.; Munslow, I. J.; Benet-Buchholz, J.; Maseras, F.; Vidal-Ferran, A. Chem.
Eur. J. 2010, 16, 6495. (d) Fernandez-Perez, H.; Etayo, P.; Nunez-Rico,
J. L.; Vidal-Ferran, A. Chim. Oggi 2010, 28, XXVI.
́ ́
dez-Perez, H.; Etayo, P.; Panossian, A.; Vidal-Ferran, A.
1P, P−C), 141.0 (dd, JP−Rh = 273.6 Hz, JP−P = 60.4 Hz, 1P, P−O);
11B{1H} NMR (128 MHz, CD2Cl2) δ −1.16 (s, BF4 ); 19F{1H}
−
NMR (376 MHz, CD2Cl2) δ −152.8 (s, BF4); HRMS (ESI+): m/z
[M − BF4]+ calcd for C66H66O4P2Rh 1099.3491, found 1099.3555.
General Procedure for the Rh-Mediated Asymmetric Hydro-
genations. A solution of the required amount of [Rh(nbd)-
(P-OP)]BF4 (8a,b; 0.04−2.0 mol %) and the corresponding
functionalized alkene among substrates 9−14 (0.10 mmol) in
anhydrous and degassed THF or DCM (0.50 mL) was prepared
inside a glass vessel under an N2 atmosphere in a glovebox. In all cases
the molar concentration of a given substrate in the reaction medium
was adjusted to 0.20 M. Once the reaction mixture had been loaded,
the glass vessel was then placed into one of the holes of a steel
́
́
ez, H.; Popa, D.; Vidal-Ferran, A.
́
́
́
ez, H.; Benet-Buchholz, J.;
̃
́
́
̀
́
́
́
́
́
̃
(8) Reaction under pressure carried out safely inside an Ace pressure tube.
6724
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725

Organometallics
Article
(9) Dicyclohexylphosphine-phosphites 7a,b proved to be highly
sensitive to oxidation by air and, as a consequence, once formed and
after a simple workup, 7a,b were immediately used as crude
compounds without purification by column chromatography. For-
tunately, the corresponding rhodium complexes were cleanly formed
from those crude compounds and these were stable in air, which
provided great operational convenience for handling.
(10) See the Experimental Section for details.
(11) The influence of the steric environment at this position was
previously observed for other chiral ligands derived from Sharpless
epoxy alcohols when applied in several asymmetric transformations.
See, for example: (a) Puigjaner, C.; Vidal-Ferran, A.; Moyano, A.;
Pericas
M. A.; Puigjaner, C.; Riera, A.; Vidal-Ferran, A.; Gom
F.; Muller, G.; Rocamora, M. Chem. Eur. J. 2002, 8, 4164. (c) Jimeno,
C.; Vidal-Ferran, A.; Pericas, M. A. Org. Lett. 2006, 8, 3895. (d) Popa,
D.; Puigjaner, C.; Gomez, M.; Benet-Buchholz, J.; Vidal-Ferran, A.;
Pericas, M. A. Adv. Synth. Catal. 2007, 349, 2265. (e) Popa, D.;
Marcos, R.; Sayalero, S.; Vidal-Ferran, A.; Pericas, M. A. Adv. Synth.
Catal. 2009, 351, 1539.
̀
, M. A.; Riera, A. J. Org. Chem. 1999, 64, 7902. (b) Pericas
̀
,
́
ez, M.; Jimenez,
́
̀
́
̀
̀
(12) See: Balasubramaniam, S.; Aidhen, I. S. Synthesis 2008, 3707
and references therein.
(13) (a) Ates, C.; Surtees, J.; Burteau, A.-C.; Marmon, V.; Cavoy, E.
International Patent WO 014080, 2003. (b) Kenda, B. M.; Matagne,
A. C.; Talaga, P. E.; Pasau, P. M.; Differding, E.; Lallemand, B. I.;
Frycia, A. M.; Moureau, F. G.; Klitgaard, H. V.; Gillard, M. R.; Fuks,
B.; Michel, P. J. Med. Chem. 2004, 47, 530.
(14) Khiri, N.; Bertrand, E.; Ondel-Eymin, M.-J.; Rousselin, Y.;
́
Bayardon, J.; Harvey, P. D.; Juge, S. Organometallics 2010, 29, 3622.
(15) For total syntheses of (R)-lacosamide, see: (a) Choi, D.;
Stables, J. P.; Kohn, H. J. Med. Chem. 1996, 39, 1907. (b) Andurkar,
S. V.; Stables, J. P.; Kohn, H. Tetrahedron: Asymmetry 1998, 9, 3841.
(c) Morieux, P.; Stables, J. P.; Kohn, H. Bioorg. Med. Chem. 2008,
16, 8968. (d) Muthukrishnan, M.; Mujahid, M.; Sasikumar, M.;
Mujumdar, P. Tetrahedron: Asymmetry 2011, 22, 1353.
(16) See Perucca, E.; Yasothan, U.; Clincke, G.; Kirkpatrick, P. Nat.
Rev. Drug Discovery 2008, 7, 973 and references therein.
(17) Etayo, P.; Nunez-Rico, J. L.; Fernan
A. Chem. Eur. J. DOI: 10.1002/chem.201103014.
́
dez-Per
́
ez, H.; Vidal-Ferran,
́
(18) For a closely related precedent of such a side reaction, see
Ramer, S. E.; Moore, R. N.; Vederas, J. C. Can. J. Chem. 1986, 64, 706.
(19) For related transformations, see: Matsuda, T.; Shiose, S.; Suda,
Y. Adv. Synth. Catal. 2011, 353, 1923 and references therein.
6725
dx.doi.org/10.1021/om200933b | Organometallics 2011, 30, 6718−6725