Nonlinear effects in Rh-catalyzed asymmetric olefin hydrogenation using mixtures of chiral monodentate P ligands

Article abstract of DOI:10.1002/anie.200503604

(Chemical Equation Presented) Well mixed: A 1:1 mixture of two different binaphthol-derived monophosphonite ligands (L) is employed in the Rh-catalyzed hydrogenation of prochiral olefins (see scheme; cod = cyclooctadiene). Variation of the enantiopurity o

Full text of DOI:10.1002/anie.200503604

Communications
are bonded to rhodium in the transition state ofhydro-
genation, which is in accord with NMR data, kinetic inves-
tigations, and observations ofconventional positive nonlinear
effects (NLEs).[5] Such detailed mechanistic investigations
have not yet been completed for the catalyst systems
involving the analogous phosphonites[2] or phosphoramidi-
tes,[3a] but preliminary results point to a similar mecha-
nism.[3b,6]
In 2002, we proposed a new tool in combinatorial
asymmetric transition-metal-catalyzed reactions:[7] The use
of mixtures of two different chiral monodentate ligands La
and Lb that led to the presence of three different catalysts in a
given reaction mixture, two homo-combinations MLaLa and
MLbLb and one hetero-combination MLaLb. Ifthe hetero-
combination is more reactive and more enantioselective than
either ofthe homo-combinations, enhanced enantioselectivity
will result. We demonstrated this effect in Rh-catalyzed olefin
hydrogenation.[7] Thus, the method allows for high catalyst
diversity with the possibility ofdiscovering superior catalysts,
but without the need to synthesize new ligands. Following our
initial report, Feringa and co-workers reported related results
using the analogous monodentate phosphoramidites 1.[8]
Herein, we address for the first time the question of
possible NLEs when using such mixtures. The situation is
considerably more complex than in the case ofclassical
systems that display conventional NLEs,[9] which means that
the usual mathematical treatment does not apply. When
Asymmetric Catalysis
DOI: 10.1002/anie.200503604
Nonlinear Effects in Rh-Catalyzed Asymmetric
Olefin Hydrogenation Using Mixtures of Chiral
Monodentate P Ligands**
a
varying the enantiomeric purity ofL from 100 to 0% ee
(racemate) while keeping the partner ligand pure, for
example, LbS, six different catalytic reactions need to be
considered [Eq. (1)]:
Manfred T. Reetz,* Yu Fu, and Andreas Meiswinkel
Several years ago three groups reported that certain binaph-
thol (BINOL)-derived monodentate phosphorus compounds
are excellent chiral ligands in Rh-catalyzed olefin hydro-
genation, namely, phosphites 1,[1] phosphonites 1 (R = alkyl,
aryl)[2] and phosphoramidites 1.[3] This came as a surprise
The stereochemical outcome ofa given hydrogenation
reaction depends upon the inherent enantioselectivity ofeach
ofthese processes, but also on the relative amounts of
catalysts present in the mixture and the six rate constants,
which can be expected to be different (except for k1 = k’1).
Because we are unable to separate the homo- from the
hetero-complexes and to use the latter without ligand
exchange, only k1 (or k’1) and k3 can be measured directly in
separate experiments. In the other cases, the rate ofproduct
formation results from the combined action of the relevant
rate processes. As a model reaction, we chose the Rh-
catalyzed hydrogenation ofitaconic acid dimethyl ester ( 2),
with compounds 1a and 1b serving as the two P ligands
[Eq. (2)]. In all cases, [Rh(cod)2]BF4 (cod = cyclooctadiene)
was treated with two equivalents of 1 to form [Rh(L)2(cod)]
(L = ligand) complexes as the precatalysts.
because it had been accepted for decades that chelating
bidentate diphosphines are necessary for high enantioselec-
tivity.[4] A recent mechanistic study by our group focused on
the phosphites and showed that two monodentate P ligands
[*] Prof. Dr. M. T. Reetz, Dr. Y. Fu, Dr. A. Meiswinkel
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1
45470 Mülheim/Ruhr (Germany)
Fax: (+49)208-306-2985
E-mail: reetz@mpi-muelheim.mpg.de
[**] We thank the Fonds der Chemischen Industrie for generous support
and Nils Theyssen for advice in measuring the hydrogenation rates.
1412
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 1412 –1415

Angewandte
Chemie
When utilizing the enantiopure homo-combinations [Rh]/
(S)-1a and [Rh]/(S)-1b in separate experiments, product 3
was shown to have the S configuration with 90 and 57% ee,
respectively.[7b] Upon using a 1:1 mixture of( S)-1a and (S)-1b,
the ee value increased to 96% (S).[7b] In the present study, we
first varied the enantiopurity of (S)-1a stepwise from 100 to
0% ee, while maintaining 100% ee of( S)-1b. The result of
these experiments show that the ee value of 3 decreases only
slightly from 96 to 92% ee (S enantiomer) as the racemic state
is approached (Figure 1, black line). Thus, one ofthe
components can be a racemate, and the result from a practical
point ofview is still acceptable.
Figure 2. NLE in the Rh-catalyzed hydrogenation of 2 using enantio-
pure (S)-1a and 1b of varying enantiopurity.
To gain some mechanistic insight, we first performed
NMR experiments. Upon treating [Rh(cod)2]BF4 with a 1:1
mixture of( S)-1a and (S)-1b, three species were identified by
NMR spectroscopy: [Rh{(S)-1a}2(cod)], [Rh{(S)-1b}2(cod)],
and [Rh{(S)-1a}{(S)-1b}(cod)] catalysts in a ratio of1:1:3,
which is not the statistical but rather the thermodynamic
value.[6,7b] Ofcourse, once the precatalysts convert into the
active species, the ratio may change (which we could not
measure). When performing the experiment with a 1:1
mixture ofpure ( S)-1a and rac-1b, a complex NMR spectrum
results which displays the same peaks as before in addition to
new signals that correspond to diastereomeric complexes.
Extensive work is required before a final NMR analysis can
be presented.
Kinetic studies ofasymmetric catalytic processes are
extremely useful in a mechanistic sense and therefore in
designing improved catalyst systems.[10] In the present case,
the relative rates ofhydrogenation of 2 were first measured
using the conventional homo-combinations [Rh]/(S)-1a and
[Rh]/(S)-1b. As shown in Figure 3, the bulky tert-butyl ligand
(S)-1b leads to the lowest catalyst activity followed by (S)-1a.
The synthetically useful combination that arises from a
mixture of( S)-1a and (S)-1b results in the highest rate
(Figure 3, red curve), which ofcourse is the result ofthree
reactions. As required by the mixture concept,[7] the two
homo-combinations are not only less enantioselective, they
are also less active, which is another way ofstating that the
hetero-combination displays enhanced enantioselectivity and
an increased rate. The pure mixed Rh complex of( S)-1a/(S)-
1b, ifit were possible to prepare as the sole species in the
reaction, would be even more enantioselective and reactive
than the mixture ofthe three species (the total amount ofRh
being constant).
Figure 1. NLE in the Rh-catalyzed hydrogenation of 2 using enantio-
pure (S)-1b and 1a of varying enantiopurity.
It was also ofinterest to scan the other region of
stereochemical space by once again maintaining 100%
S purity of 1b, but working in a regime in which the
R configuration of 1a dominates, namely, using (R)-1a at
different levels of enantiopurity. The experimental result is
also shown in Figure 1 (gray curve). In the extreme case of
enantiomerically pure (R)-1a and (S)-1b in a mixture, the
configuration of 3 remains as S, but the ee value decreases to
39%. This behavior indicates that the influence of the tert-
butyl component 1b dominates the stereochemical outcome.
We then performed the analogous series of experiments in
which the enantiopurity of( S)-1a was maintained at 100%
and the ee value of 1b was varied (Figure 2). The results point
to a similar behavior when varying the enantiopurity of( S)-1b
from 100 to 0% ee (Figure 2, black line). When working in the
regime in which more (R)- than (S)-1b is present (gray curve),
the quantitative results are somewhat different from those in
Figure 1. Nevertheless, the influence of the tert-butyl compo-
nent 1b dominates again. In this case, the effect finally leads
to inversion ofconigfuration, with ( R)-3 being formed
preferentially with 43% ee. This value does not quite
correspond to the expected 39% ee observed previously in
the case of( R)-1a/(S)-1b, but may be within experimental
error (Æ 2%).
In previous synthetic work regarding the use ofmix-
tures,[6,7] a trend was observed (with exceptions): Hetero-
combinations show enhanced enantioselectivity whenever
one component is “small” and the other is “large”, as in the
Angew. Chem. Int. Ed. 2006, 45, 1412 –1415
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1413

Communications
Figure 3. Relative rates of hydrogenation of 2 (Rh/2=1:5000, Rh/
1=1:2, 1a/1b=1:1, 1.3 bar H2, room temperature, CH2Cl2 as the
solvent).
Figure 4. NLE in the Rh-catalyzed hydrogenation of 2 using mixtures
of 1a and 1b, each with varying degrees of enantiopurity.
present case.[7] Rate enhancement because ofthis structural
feature is observed even when the absolute configurations of
the two components are opposite (mismatched), as in the case
of( S)-1a/(R)-1b (Figure 3, blue curve). Of course, this effect
is less pronounced relative to the use ofthe matched case ( S)-
1a/(S)-1b. The fact that the systems using the racemate of one
ofthe components are more reactive than ( S)-1a/(R)-1b is
now easily understood.
Although the rates ofthe six individual processes that
occur in a system comprising rac-1a/(S)-1b (or (S)-1a/rac-1b)
cannot be measured separately, the results ofthe kinetic study
are in line with the following conclusion: The combination
(S)-1a/(S)-1b constitutes the most reactive system, all others
such as diastereomeric (S)-1a/(R)-1b are less able to compete
effectively. If the amount of the diastereomeric ligand
combination (S)-1a/(R)-1b increases in the mixture by using
1b enriched with the R enantiomer, the reaction catalyzed by
the respective Rh complex begins to compete and in the
extreme case dominates, thus resulting in markedly lower or
even opposite enantioselectivity.
(R) in the extremes and that an enantioselectivity of
> 90% ee can be achieved even when both components
have enantiopurities ofonly 80% ee and even less (Figure 4,
red areas).
The basic phenomenon described above for itaconic acid
diester (2) was also observed in the case oftwo other
substrates 4a and 4b [Eq. (4)], although only a small portion
ofthe data points were collected in these cases (Table 1). For
example, in the case ofsubstrate 4b, racemic 1a can be used
with enantiopure 1b and the ee value is still a respectable
93% (Table 1, entry 6).
The present results are not only ofmechanistic value; they
also point to a practical feature, namely, the fact that
complete enantiopurity ofthe ligands is not essential.
Indeed, one ofthe components can even be racemic, with
the process still leading to > 92% ee. Although ofonly partial
practical value, we were interested from a theoretical view-
point to see how the system performs when varying the
enantiopurity ofboth components simultaneously. This
means that a total often reactions needs to be considered
[Eq. (3)].
Table 1: Asymmetric hydrogenation of olefins 4a,b using ligands 1a,b.
Entry
Ligand
ee of 5a [%][a]
ee of 5b [%][b]
(configuration)
(configuration)
1
2
3
4
5
6
(S)-1a
(S)-1b
(S)-1a/(S)-1b
(S)-1a/(R)-1b
(S)-1a/rac-1b
rac-1a/(S)-1b
76 (R)
13 (R)
95 (R)
49 (R)
83 (R)
70 (R)
92 (R)
93 (R)
98 (R)
82 (S)
62 (R)
93 (R)
[a] Conditions: Rh/4a=1:100, Rh/1=1:2, 1a/1b=1:1, 1.3 bar H2, room
temperature, CH2Cl2 as the solvent. [b] Conditions: Rh/4b=1:200, Rh/
1=1:2, 1a/1b=1:1, 1.3 bar H2, room temperature, CH2Cl2 as the
solvent.
The graphic representation ofthe data ofsuch a system-
atic search requires a three-dimensional format (Figure 4). It
can be seen that the ee values vary from 96% ee (S) to 96% ee
In summary, we have addressed for the first time the
question of nonlinear effects in Rh-catalyzed olefin hydro-
genation using a mixture of two different chiral monodentate
1414
www.angewandte.org
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2006, 45, 1412 –1415

Angewandte
Chemie
[11] a) This observation bears some resemblance to our previous
P ligands. Apart from the theoretical interest, the results are
ofpractical significance because the ligands do not need to be
100% enantiopure. Indeed, in some cases one ofthe ligand
components can be racemic[11] while still leading to > 92% ee.
It remains to be seen ifthe molecular basis ofthe observed
kinetic and stereochemical behavior can be uncovered by
theoretical studies.
work involving a diphosphite composed ofa chiral-backbone
diol and configurationally fluxional atropisomeric diphenol-
derived P heterocycles that lead to three diastereomeric ligand/
metal complexes with different kinetic profiles (D. G. Black-
mond, T. Rosner, T. Neugebauer, M. T. Reetz, Angew. Chem.
1999, 111, 2333 – 2335; Angew. Chem. Int. Ed. 1999, 38, 2196 –
2199; M. T. Reetz, T. Neugebauer, Angew. Chem. 1999, 111,
134 – 137; Angew. Chem. Int. Ed. 1999, 38, 179 – 181). Diphenol
itselfis achiral, but atropisomerism is created in the synthesis of
Received: October 12, 2005
Published online: January 20, 2006
the
P heterocycle. This basic phenomenon was reported
indepently by Noyori, Mikami, and co-workers in a study in
which they describe a Ru complex composed ofa chiral diamine
and an achiral diphenyl-derived diphosphine; upon complex-
ation, the latter turns into a configurationally fluxional atropi-
someric system that leads to two diastereomeric complexes with
different reaction rates and enantioselectivity (K. Mikami, T.
Korenaga, M. Terada, T. Ohkuma, T. Pham, R. Noyori, Angew.
Chem. 1999, 111, 517 – 519; Angew. Chem. Int. Ed. 1999, 38, 495 –
497). Since these papers appeared, many further examples have
been reported, including the extension by Balsells and Walsh
who showed that the enantioselectivity ofthe addition ofEt 2Zn
to aldehydes catalyzed by a chiral titanium alkoxide can be
increased by the addition of“achiral” meso-1,2-diamine, which is
really a pair offluxional enantiomers (J. B. Balsells, P. J. Walsh, J.
Am. Chem. Soc. 2000, 122, 1802 – 1803); see also: b) K. Mikami,
M. Yamanaka, Chem. Rev. 2003, 103, 3369 – 3400; c) J. W. Faller,
A. R. Lavoie, J. Parr, Chem. Rev. 2003, 103, 3345 – 3367; d) P. J.
Walsh, A. E. Lurain, J. Balsells, Chem. Rev. 2003, 103, 3297 –
3344.
Keywords: asymmetric catalysis · hydrogenation · kinetics ·
nonlinear effects · rhodium
.
[1] a) M. T. Reetz, G. Mehler, Angew. Chem. 2000, 112, 4047 – 4049;
Angew. Chem. Int. Ed. 2000, 39, 3889 – 3890; review: b) M. T.
Reetz, Chim. Oggi 2003, 21, 5 – 8; see also: c) I. V. Komarov, A.
Börner, Angew. Chem. 2001, 113, 1237 – 1240; Angew. Chem. Int.
Ed. 2001, 40, 1197 – 1200; d) T. Jerphagnon, J.-L. Renaud, P.
Demonchaux, A. Ferreira, C. Bruneau, Adv. Synth. Catal. 2004,
346, 33 – 36; e) J. Ansell, M. Wills, Chem. Soc. Rev. 2002, 31, 259 –
268; review ofmonodentate P ligands in general: f) F. Lagasse,
H. B. Kagan, Chem. Pharm. Bull. 2000, 48, 315 – 324.
[2] a) M. T. Reetz, T. Sell, Tetrahedron Lett. 2000, 41, 6333 – 6336;
b) C. Claver, E. Fernandez, A. Gillon, K. Heslop, D. J. Hyett, A.
Martorell, A. G. Orpen, P. G. Pringle, Chem. Commun. 2000,
961 – 962.
[3] a) M. van den Berg, A. J. Minnaard, E. P. Schudde, J. van Esch,
A. H. M. de Vries, J. G. de Vries, B. L. Feringa, J. Am. Chem.
Soc. 2000, 122, 11539 – 11540; b) M. van den Berg, A. J. Min-
naard, R. M. Haak, M. Leeman, E. P. Schudde, A. Meetsma,
B. L. Feringa, A. H. M. de Vries, C. E. P. Maljaars, C. E. Willans,
D. Hyett, J. A. F. Boogers, H. J. W. Henderickx, J. G. de Vries,
Adv. Synth. Catal. 2003, 345, 308 – 323; c) H. Bernsmann, M.
van den Berg, R. Hoen, A. J. Minnaard, G. Mehler, M. T. Reetz,
J. G. de Vries, B. L. Feringa, J. Org. Chem. 2005, 70, 943 – 951.
[4] a) W. S. Knowles, Angew. Chem. 2002, 114, 2096 – 2107; Angew.
Chem. Int. Ed. 2002, 41, 1998 – 2007; b) R. Noyori, Angew.
Chem. 2002, 114, 2108 – 2123; Angew. Chem. Int. Ed. 2002, 41,
2008 – 2022.
[5] M. T. Reetz, A. Meiswinkel, G. Mehler, K. Angermund, M. Graf,
W. Thiel, R. Mynott, D. G. Blackmond, J. Am. Chem. Soc. 2005,
127, 10305 – 10313.
[6] T. Sell, Dissertation Ruhr-Universität Bochum, 2002.
[7] a) M. T. Reetz, A. Meiswinkel, T. Sell, G. Mehler, patent
application DE-A 102476330, 2002; b) M. T. Reetz, T. Sell, A.
Meiswinkel, G. Mehler, Angew. Chem. 2003, 115, 814 – 817;
Angew. Chem. Int. Ed. 2003, 42, 790 – 793; c) M. T. Reetz, G.
Mehler, Tetrahedron Lett. 2003, 44, 4593 – 4596; d) M. T. Reetz,
X. Li, Tetrahedron 2004, 60, 9709 – 9714; e) M. T. Reetz, G.
Mehler, A. Meiswinkel, T. Sell, Tetrahedron: Asymmetry 2004,
15, 2165 – 2167.
[8] a) D. Peæa, A. J. Minnaard, J. A. F. Boogers, A. H. M. de Vries,
J. G. de Vries, B. L. Feringa, Org. Biomol. Chem. 2003, 1, 1087 –
1089; b) A. Duursma, R. Hoen, J. Schuppan, R. Hulst, A. J.
Minnaard, B. L. Feringa, Org. Lett. 2003, 5, 3111 – 3113.
[9] Reviews: a) C. Girard, H. B. Kagan, Angew. Chem. 1998, 110,
3088 – 3127; Angew. Chem. Int. Ed. 1998, 37, 2922 – 2959;
b) H. B. Kagan, Synlett 2001, 888 – 899; c) D. G. Blackmond,
Acc. Chem. Res. 2000, 33, 402 – 411; d) M. Avalos, R. Babiano, P.
Cintas, J. L. JimØnez, J. C. Palacios, Tetrahedron: Asymmetry
1997, 8, 2997 – 3017.
[10] D. G. Blackmond, Angew. Chem. 2005, 117, 4374 – 4393; Angew.
Chem. Int. Ed. 2005, 44, 4302 – 4320.
Angew. Chem. Int. Ed. 2006, 45, 1412 –1415
ꢀ 2006 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1415