Rhodium-catalyzed asymmetric 1,4-addition of aryl alanes to trisubstituted enones: Binap as an effective ligand in the formation of quaternary stereocenters

Article abstract of DOI:10.1002/anie.201003300

All for one and 1,4-all: Readily available aryl alanes are used in the rhodium-catalyzed asymmetric conjugate addition reaction with a variety of cyclic and acyclic enones. The enhanced reactivity of the system allows the use of the common binap ligand for the generation of quaternary benzylic stereocenters in excellent enantioselectivity (see scheme).

Full text of DOI:10.1002/anie.201003300

Angewandte
Chemie
DOI: 10.1002/anie.201003300
Asymmetric Catalysis
Rhodium-Catalyzed Asymmetric 1,4-Addition of Aryl Alanes to
Trisubstituted Enones: Binap as an Effective Ligand in the Formation
of Quaternary Stereocenters**
Christine Hawner, Daniel Mꢀller, Ludovic Gremaud, Abdellah Felouat, Simon Woodward,* and
Alexandre Alexakis*
The creation of quaternary all-carbon- or heteroatom-sub-
stituted stereogenic centers is an important transformation in
organic synthesis, and quaternary centers[1] are a common
motif in natural products. Several groups, ourselves among
them, have worked on this challenge and one way that has
been found to access these centers is through copper-
catalyzed conjugate addition reactions.[2] However, in such
reactions, the introduction of aryl groups presents an addi-
tional challenge owing to the steric hindrance of the sys-
tem.[2e–g,l] Special activation is required and aluminum organ-
yls are strong Lewis acids that can coordinate to the enone
and thus render the substrate more electrophilic; as such, they
have been especially successful in the copper-catalyzed aryl-
addition reaction.[3] Although rhodium in combination with
boron reagents is the metal of choice when introducing aryl
groups into a,b-unsaturated systems,[4] for a long time the
addition to trisubstituted substrates was limited to very few
activated substrates.[5] Only recently have a few articles
appeared that report the rhodium-catalyzed formation of
quaternary centers on enones,[6] of which one[6b] presents a
single example with poor yield. In the more general stud-
ies,[6a,c] several parameters in the original protocol[7] needed to
be changed:[7] boronic acids were replaced by tetraarylbor-
ates[6a] and boroxines,[6c] and chiral diene ligands were used
because commercially available binap did not give any
conversion. Furthermore, 2–4 equivalents of the nucleophile
(bearing four aryl groups per equivalent in the case of
tetraarylborates) were needed. These requirements left room
for the development of more atom-economical reactions for
the formation of benzylic quaternary centers that allow the
use of more easily accessible ligands.
Enones are interesting starting materials as they contain
several reactive sites that might be transformed regioselec-
tively. 1,4-Addition to enones has been studied more than 1,2-
addition,[8] but in 2007 an interesting article was published
reporting the rhodium-catalyzed addition of Me3Al and
PhAlMe2 to 2-cyclohexenone giving the 1,2-adduct as the
main product in very high ee.[9] Therefore, we were interested
in testing our previously reported methodology[3a] for the
generation of aryl alanes[10] by application in the rhodium-
catalyzed addition to 3-substituted 2-cyclohexenones, assum-
ing that an additional substituent in the b position would
result in a preference for 1,2-addition, and especially as
rhodium-catalyzed conjugate addition to trisubstituted
enones had not been reported at the time we started the
project. As we had obtained very high enantioselectivities in
copper-catalyzed 1,4-additions to the same substrates, we
were hoping to be able to develop a complementary method-
ology that would give us selective 1,2-addition with rhodium
and 1,4-addition with copper (Scheme 1).
Scheme 1. Enantioselective rhodium-catalyzed 1,4-addition to trisubsti-
tuted enones. Alk=alkyl, Ar=aryl.
In the first instance, we generated Me2AlPh[11] through a
transmetalation reaction from commercial PhLi and Me2AlCl
and, to our great surprise, we obtained the 1,4-adduct almost
exclusively and in very high ee using [{Rh(cod)Cl}2] (cod =
1,5-cyclooctadiene) and (R)-binap (L1; Scheme 2).
Intrigued by this inversion of the expected reactivity we
decided to pursue this reaction further. Furthermore, the
prospect of developing a method that would allow the use of
commercially available binap instead of dienes that require
multistep syntheses (although some dienes have been recently
commercialized) seemed highly interesting to us.
[*] Dr. C. Hawner, D. Mꢀller, L. Gremaud, A. Felouat, Prof. A. Alexakis
Department of Organic Chemistry, University of Geneva
30 quai Ernest Ansermet, 1211 Geneve 4 (Switzerland)
Fax: (+41)223-793215
E-mail: alexandre.alexakis@chiorg.unige.ch
Prof. S. Woodward
School of Chemistry, The University of Nottingham
Nottingham NG7 2RD (United Kingdom)
Fax: (+44)115-951-3564
E-mail: simon.woodward@nottingham.ac.uk
[**] We thank the European Community (Marie Curie Early Stage
Researcher Training, MEST-CT-2005-019780, INDAC-CHEM) for
financial support.
We first generated the alane from the corresponding aryl
iodide through an I/Li exchange with nBuLi, followed by
transmetalation with Me2AlCl, which would give us access to
a wide range of nucleophiles (Scheme 3). It should be noted
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201003300.
Angew. Chem. Int. Ed. 2010, 49, 7769 –7772
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
7769

Communications
ologies to access aryl alanes: In the first one (method A;
Scheme 4), the alane was obtained from the corresponding
aryl Grignard reagent through transmetalation with Me2AlCl,
and in the second (method B), a Br/Li exchange was
performed on an aryl bromide with nBuLi, followed by
transmetalation with Me2AlCl.
Scheme 2. Enantioselective rhodium-catalyzed 1,4-addition to trisubsti-
tuted enones.
Scheme 4. Methods for the generation of the alane nucleophile.
Scheme 3. Generation of aryl alanes from aryl iodides.
We used method A to optimize our reaction conditions.
Changing the aryl source from aryl iodides to Grignard
reagents did not improve the conversion, as a similar result
was obtained as for phenyl iodide (Table 2, entry 1). How-
ever, we observed an important improvement when the
addition of the magnesium salts was avoided (Table 2,
entry 2). It seems like the latter are also detrimental for the
reaction. Therefore, we tested dioxane as the solvent, as it is
known to precipitate magnesium salts, and it is a popular
solvent[4] for rhodium-catalyzed conjugate addition reactions.
The reaction temperature needed to be increased owing to
the melting point of dioxane. Under tetrahydrofuran-free
conditions, full conversion took place even at reduced catalyst
loading, but almost no enantiomeric excess was achieved
(Table 2, entry 3). It is thought that the active catalyst does
not form in dioxane, so that free [{Rh(cod)Cl}2] promoted this
reaction for the most part. Therefore, we had to find a way to
supply enough tetrahydrofuran to allow the formation of the
that one equivalent of nBuI is generated as a by-product in
this sequence. We then proceeded in testing the mixture in the
1,4-addition to 3-methyl-2-cyclohexenone (Table 1).
Table 1: Phenyl iodide as aryl source in the asymmetric conjugate
addition reaction.
Entry
Aryl source
Additive
Conv. [%][a]
ee [%]
1
2
3
4
PhLi
PhI
PhLi
PhLi
none
none
nBuI
98
29
1
99.6
98.5
n.d.
99.4
nBuBr
100
[a] Conversion was determined by 1H NMR analysis.
This table clearly shows that nBuI interferes in the
reaction, as 98% conversion was achieved starting from
commercial PhLi (Table 1, entry 1), but only 29% conversion
was observed when the aryl alane was derived from PhI
(Table 1, entry 2). To confirm this assumption, we added
1 equivalent of nBuI to commercial PhLi and generated the
alane from this mixture. Indeed, the conversion was severely
impaired, which means that nBuI is detrimental to this
reaction (Table 1, entry 3). It is possible that oxidative
addition of nBuI takes place at rhodium, leading to b-hydride
elimination on the butyl group and subsequent formation of a
catalytically ineffective Rh/hydride complex. However, this
phenomenon was not observed for the nBuBr, as full
conversion still took place in its presence (Table 1, entry 4).
Aryl iodides are thus not good aryl sources under these
conditions and we thus envisaged two alternative method-
Table 2: Optimization of reaction conditions using method A.
Entry T [8C] t [h] Equiv
Mg Rh cat./L1
salts [mol%]
Conv. ee [%]
alane
[%][a]
1
2
À20 to
3
3
3.0
3.0
yes 5:12.5
29
57
99.6
98.5
10
À20 to
no
5:12.5
10
5
3[b]
4[c]
o.n. 3.0
o.n. 1.2
no
no
3:9
3:9
96
99
7
98
5
[a] Conversion was determined by 1H NMR analysis. [b] Dioxane was
used as the solvent. [b] Dioxane containing 6% THF (v/v) was used as
the solvent. o.n.=overnight.
7770
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7769 –7772

Angewandte
Chemie
Table 4: Screening of different nucleophiles.
active catalyst, but no more than necessary, to minimize its
degradation. Therefore, we used dioxane as a solvent that
contained 6% (v/v) tetrahydrofuran. It seemed like this
amount of tetrahydrofuran was enough to promote the
coordination of binap to rhodium since 98% ee was obtained
with almost full conversion when only 1.2 equivalents of alane
was used with 3 mol% rhodium dimer (Table 2, entry 4).
With the optimized conditions in hand, we performed a
ligand screening (Table 3). (R)-Binap gave the best conver-
sion (Table 3, entry 1), but good results were also achieved
Entry Ar
Method Equiv Prod. Conv.
Yield
[%]
ee
[%]
alane
[%][a]
1
2
3
4
5
C6H5
C6H5
m-MeC6H4
p-MeC6H4
m-
A
B
A
A
A
1.2
1.5
1.2
1.2
1.0
2
2
3
4
5
99
99
99
93
99
71
n.d.
73
50
75
98
99
98
96
98
Table 3: Ligand screening using method A for alane generation.
OMeC6H4
p-
5
A
1.2
6
99
57
95
OMeC6H4
m-CF3C6H4
m-CF3C6H4
m-CF3C6H4
6
B
B
B
B
1.2
1.5
1.5
1.5
7
7
7
7
98
93
96
61
n.d.
n.d.
n.d.
99
99
>99
n.d.
7
8[b]
9[b,c] m-CF3C6H4
–
[d]
[a] Conversion was determined by 1H NMR analysis. [b] The whole
mixture was added, including the LiCl salts. [c] THF was used as solvent.
[d] A complex mixture was observed.
Entry
Ligand
Conv. [%][a]
ee [%]
of products, which underlines the importance of working in
dioxane (Table 4, entry 9).
1
2
L1
L2
L3
L3
99
96
86
95
98
99[b]
11
Finally, we investigated the substrate scope. Table 5 shows
that a wide range of linear and cyclic substrates can be used in
this transformation. Excellent enantioselectivities were
obtained in all cases, even for the linear substrates. The
olefin geometry of the enone influenced the absolute config-
uration of the product as E and Z substrates gave opposite
3
4[c]
17
[a] Conversion was determined by 1H NMR analysis. [b] The opposite
enantiomer was obtained. [c] The Rh/diene catalyst was preformed and
used as the isolated compound.
with (S)-OMe-biphep (L2), especially in terms of enantiose-
lectivity (Table 3, entry 2). The results with diene L3 were
disappointing. Good conversion was achieved when the active
catalyst was generated in situ, but the enantiomeric excess
was low (Table 3, entry 3). When the active catalyst was
preformed, higher conversion was achieved, but the enantio-
selectivity remained low (Table 3, entry 4).
Table 5: Screening of different substrates.
Entry
Product
Rh cat./L1
Conv.
[%]
Yield ee
[mol%]
[%]
[%]
Therefore, we decided to continue our investigations with
binap under these new conditions screening different nucle-
ophiles. It should be pointed out that for 3-ethyl cyclo-
hexenone, under the same reaction conditions, Me3Al gave
only the 1,4 adduct, albeit with 17% conversion and 45% ee.
We then tested different aryl nucleophiles (Table 4). Meth-
od B was also tested under the optimized conditions. Aryl
groups with different electron-donating and electron-with-
drawing substituents were added in good yields and nearly
perfect enantioselectivities. Method B was also successful, as
the same results were obtained for the addition of a phenyl
group (Table 4, entry 2). It was then also applied for the
addition of the electron-withdrawing trifluoromethyl-substi-
tuted aryl group. Our brief study showed that increasing the
number of equivalents of alane did not lead to improved
conversion (Table 4, entry 6 vs. entry 7) and that the whole
mixture could be used without detrimental effect (Table 4,
entry 7 vs. entry 8). This makes the use of method B very
convenient as no separation of salts and supernatant is
necessary. Furthermore, in the case of method B, the use
tetrahydrofuran as the sole solvent led to a complex mixture
1
2
3
4
8
3:9
96
71
66
99
99
9
5:12.5
3:9
35
46
56
10
11
100
86
>99
95
5:12.5
5[a]
6
12
13
5:12.5
5:12.5
91
55
52
93
100
>95
[a] The product was synthesized from (E)-4-methyl-6-phenylhex-3-en-2-
one. The corresponding Z substrate gave the opposite enantiomer in
88% ee.
Angew. Chem. Int. Ed. 2010, 49, 7769 –7772
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
7771

Communications
Chem. Int. Ed. 1998, 37, 388 – 401; c) L. Barriault, I. Denissova,
Tetrahedron 2003, 59, 10105 – 10146; d) L. E. Overman, C. J.
Douglas, Proc. Natl. Acad. Sci. USA 2004, 101, 5363 – 5367; e) A.
Baro, J. Christoffers, Adv. Synth. Catal. 2005, 347, 1473 – 1482;
f) C. Jiang, B. M. Trost, Synthesis 2006, 369 – 396.
enantiomers (Table 5, entry 5). Remarkably, even the chal-
lenging five-membered-ring substrate displayed excellent
reactivity, and near-perfect selectivity was achieved (Table 4,
entry 3). The reduced yield could be explained by the
sensitivity of the product towards column purification.
From mechanistic aspects, it is clear that the reaction
proceeds through the formation of an rhodium/aryl species,
and the enone is activated by the Lewis acidity of the
aluminum reagent.[6,12]
In summary, we have shown that aryl alanes are excellent
nucleophiles for the rhodium-catalyzed conjugate addition
reaction. Whilst elevated temperatures are usually required
with boron nucleophiles, only 1.2 to 1.5 equivalents of alane
afforded the same transformation overnight at 58C. Most
importantly, commercially available binap promoted the
reaction. An important electronic effect of the alkyl sub-
stituent in the 4-position of trisubstituted enones was
discovered which directs the nucleophilic attack to the 4-
position despite additional steric bulk. A range of electron-
donating and electron-withdrawing substituents were toler-
ated on the aromatic ring and the desired alanes were
generated from both the corresponding lithium or Grignard
precursors. Finally, a wide range of cyclic and acyclic
substrates led to excellent enantioselectivities in this trans-
formation.
[2] a) A. W. Hird, A. H. Hoveyda, J. Am. Chem. Soc. 2005, 127,
14988 – 14989; b) M. dꢁAugustin, L. Palais, A. Alexakis, Angew.
Chem. 2005, 117, 1400 – 1402; Angew. Chem. Int. Ed. 2005, 44,
1376 – 1378; c) N. Fuchs, M. dꢁAugustin, M. Humam, A. Alex-
akis, R. Taras, S. Gladiali, Tetrahedron: Asymmetry 2005, 16,
3143 – 3146; d) A. J. Wilsily, E. Fillion, J. Am. Chem. Soc. 2006,
128, 2774 – 2775; e) K. Lee, K. Brown, A. W. Hird, A. H.
Hoveyda, J. Am. Chem. Soc. 2006, 128, 7182 – 7184; f) D.
Martin, S. Kehrli, M. dꢁAugustin, H. Clavier, M. Mauduit, A.
Alexakis, J. Am. Chem. Soc. 2006, 128, 8416 – 8417; g) M. K.
Brown, T. L. May, C. A. Baxter, A. H. Hoveyda, Angew. Chem.
2007, 119, 1115 – 1118; Angew. Chem. Int. Ed. 2007, 46, 1097 –
1100; h) L. Palais, I. S. Mikhel, C. Bournaud, L. Micouin, C. A.
Falciola, M. Vuagnoux-dꢁAugustin, S. Rosset, G. Bernardinelli,
A. Alexakis, Angew. Chem. 2007, 119, 7606 – 7609; Angew.
Chem. Int. Ed. 2007, 46, 7462 – 7465; i) M. Vuagnoux-dꢁAugustin,
A. Alexakis, Chem. Eur. J. 2007, 13, 9647 – 9662; j) M. Vuag-
noux-dꢁAugustin, S. Kehrli, A. Alexakis, Synlett 2007, 2057 –
2060; k) H. Hꢂnon, M. Mauduit, A. Alexakis, Angew. Chem.
2008, 120, 9262 – 9264; Angew. Chem. Int. Ed. 2008, 47, 9122 –
9124; l) Y. Matsumoto, K. Yamada, K. Tomioka, J. Org. Chem.
2008, 73, 4578 – 4581; m) A. Wilsily, E. Fillion, Org. Lett. 2008,
10, 2801 – 2804; n) L. Palais, A. Alexakis, Chem. Eur. J. 2009, 15,
10473 – 10485; o) A. Wilsily, E. Fillion, J. Org. Chem. 2009, 74,
8583 – 8594.
[3] a) C. Hawner, K. Li, V. Cirriez, A. Alexakis, Angew. Chem. 2008,
120, 8334 – 8337; Angew. Chem. Int. Ed. 2008, 47, 8211 – 8214;
b) T. L. May, M. K. Brown, A. H. Hoveyda, Angew. Chem. 2008,
120, 7468 – 7472; Angew. Chem. Int. Ed. 2008, 47, 7358 – 7362.
[4] For reviews, see: a) K. Fagnou, M. Lautens, Chem. Rev. 2003,
103, 169 – 196; b) K. Yamasaki, T. Hayashi, Chem. Rev. 2003,
103, 2829 – 2844; c) G. Koripelly, A. Rosiak, M. Rꢃssle, J.
Christoffers, Synthesis 2007, 1279 – 1300; d) J. D. Hargrave,
J. C. Allen, C. G. Frost, Chem. Asian J. 2010, 5, 386 – 396.
[5] a) P. Mauleꢄn, J. C. Carretero, Chem. Commun. 2005, 4961 –
4963; b) R. Shintani, W.-L. Duan, T. Hayashi, J. Am. Chem.
Soc. 2006, 128, 5628 – 5629.
[6] a) R. Shintani, Y. Tsutsumi, M. Nagaosa, T. Nishimura, T.
Hayashi, J. Am. Chem. Soc. 2009, 131, 13588 – 13589; b) B. T.
Hahn, F. Tewes, R. Frꢃhlich, F. Glorius, Angew. Chem. 2010, 122,
1161 – 1164; Angew. Chem. Int. Ed. 2010, 49, 1143 – 1146; c) R.
Shintani, M. Takeda, T. Nishimura, T. Hayashi, Angew. Chem.
2010, 122, 4061 – 4063; Angew. Chem. Int. Ed. 2010, 49, 3969 –
3971.
Experimental Section
Method A: At 08C, Me2AlCl (1 equiv) was added to the aryl
magnesium bromide (1 equiv in THF) and reacted for 30 min at
08C. The salts were allowed to precipitate over 10–20 min at 08C and
the supernatant was used.
Method B: The aryl bromide (0.525 mmol) was dissolved in
anhydrous Et2O (0.5 mL) at RT. Then the flask was cooled to 08C and
nBuLi (0.33 mL, 0.525 mmol, 1.6m in heptane) was added. The
reaction mixture was stirred for 3 to 4 h at 08C before the addition of
Me2AlCl (0.525 mL, 0.525 mmol, 1m in hexanes) at 08C. A precipitate
of LiCl then appeared and the reaction mixture was stirred for 30 min
at 08C.
Conjugate addition reaction: THF (0.1 mL) was added to [{Rh-
(cod)Cl}2] (3 mol%) and ligand (9 mol%), and the mixture was
stirred for 2 min at RT. Then, dioxane (1.6 mL) was added and the
mixture was stirred for 30–50 min. The mixture was cooled to 58C and
the organoalane was added to the catalyst mixture, followed by the
substrate. The reaction was stirred overnight at 58C before quenching
with 10% HCl. The aqueous layer was extracted with EtOAc (2 ꢀ
20 mL) and the combined organic phases were washed with brine
before drying over MgSO4, filtration through silica, and evaporation
of the solvents. The crude mixture was purified by column chroma-
tography on silica gel using Et2O/pentane as an eluent.
[7] Y. Takaya, M. Ogasawara, T. Hayashi, J. Am. Chem. Soc. 1998,
120, 5579 – 5580.
[8] a) D. J. Ramꢄn, M. Yus, Angew. Chem. 2004, 116, 286 – 289;
Angew. Chem. Int. Ed. 2004, 43, 284 – 287; b) J. M. Betancort, C.
Garcia, P. J. Walsh, Synlett 2004, 749 – 760; c) H. Li, C. Garcia,
P. J. Walsh, Proc. Natl. Acad. Sci. USA 2004, 101, 5425 – 5427;
d) H. Li, P. J. Walsh, J. Am. Chem. Soc. 2005, 127, 8355 – 8361.
[9] J. Siewert, R. Sandmann, P. von Zezschwitz, Angew. Chem. 2007,
119, 7252 – 7254; Angew. Chem. Int. Ed. 2007, 46, 7122 – 7124.
[10] U. Imbery, A. T. Nguyen, K. Nickisch, J. Westermann, Eur. J.
Inorg. Chem. 1998, 295 – 298.
Some yields may be improved by scaling up the reactions to
2 mmol; for example, see Table 4, entry 1.
Received: May 31, 2010
Published online: September 6, 2010
[11] Et2AlPh did not give any conversion, probably owing to b-
hydride elimination.
Keywords: aryl alanes · asymmetric catalysis · Michael addition ·
quaternary stereocenters · rhodium
.
[12] Preliminary 31P NMR experiments show the characteristic
signals for the Ph-Rh-binap complex[4b] upon mixing PhMgBr
and the Rh-binap-Cl complex, but not with ArAlMe2; further
investigations are underway.
[1] a) K. Fuji, Chem. Rev. 1993, 93, 2037 – 2066; b) E. J. Corey, A.
Guzman-Perez, Angew. Chem. 1998, 110, 402 – 415; Angew.
7772
www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 7769 –7772