α-amino acid induced rate acceleration in aqueous biphasic Lewis acid catalyzed Michael addition reactions

Article abstract of DOI:10.1002/anie.200700560

It's as easy as ABC: α-Amino acids may become very important tools in the development of economical and environmentally friendly aqueous biphasic catalysis. Alanine accelerated an Yb(OTf)₃-catalyzed Michael addition by a factor of 138. The solubilizing and stabilizing properties of alanine allowed the metal catalyst to be "heterogenized" in the aqueous phase and recycled multiple times without appreciable loss of activity.

Full text of DOI:10.1002/anie.200700560

Angewandte
Chemie
DOI: 10.1002/anie.200700560
Catalysis in Water
a-Amino Acid Induced Rate Acceleration in Aqueous Biphasic Lewis
Acid Catalyzed Michael Addition Reactions**
Karolina Aplander, Rui Ding, U. Marcus Lindström,* Johan Wennerberg,* and Sara Schultz
Water is of increasing importance as a cheap, safe, and
environmentally benign reaction medium.[1] Indeed, aqueous
biphasic catalysis is one of the most promising strategies
towards more economical and greener processes.[2] However,
progress in the development of efficient catalysts that are also
highly water-soluble has been slow.[1e] With regard to the
natural abundance of a-amino acids, it is somewhat surprising
that their use as ligands in Lewis acid catalysis remains largely
unexplored. A likely explanation is their low solubility in the
organic solvents used ubiquitously for metal-catalyzed reac-
tions. In water, however, the solubility of a-amino acids is not
limiting, and their potential as ligands in aqueous Lewis acid
catalysis deserves exploration.[3]
(MVK) to afford adduct 1 [Eq. (1)]. The result was a quite
extraordinary effect on the initial (for the first 120 min)
second-order rate constant (Figure 1). The use of Yb(OTf)3
(10%) alone led to
10À4 mÀ1 minÀ1). However, the inclusion of alanine (12%)
resulted in remarkable rate acceleration (k = 1.79 ”
a very slow reaction (k = 1.3 ”
a
À
The Michael addition is one of the most important C C
bond-forming reactions. Although transition-metal-catalyzed
Michael addition reactions in organic solvents have been
developed with a great deal of success,[4] catalytic efficiency
and substrate scope in aqueous media have so far been
limited.[5] Recent significant advances have been the use of
solid-phase Lewis acid catalysis[6] and our own discovery that
small 1,2-dibasic ligands, N,N,N’,N’-tetramethylethylenedi-
amine (tmeda) in particular, can have a remarkable effect
on aqueous Michael addition reactions catalyzed by
ytterbium triflate.[7] Herein we report our first results on the
use of a-amino acids as rate-accelerating ligands.
First, we set out to measure the effect of alanine on the
catalytic activity of ytterbium triflate in the aqueous biphasic
reaction between ethyl acetoacetate and methyl vinyl ketone
[*] Dr. U. M. Lindström
Department of Chemistry
McGill University
MontrØal, Quebec, H3A 2K6 (Canada)
Fax: (+1)514-398-3797
Figure 1. Initial second-order rate constants, k, of the aqueous bipha-
sic Michael addition [Eq. (1)] in the presence and absence of d-
alanine.[8] [A]: concentration of ethyl acetoacetate.
E-mail: marcus.lindstrom@mcgill.ca
Homepage: http://web.mac.com/marcus71
10À2 mÀ1 minÀ1), which corresponds to a relative second-
order rate of 138 for the reaction in the presence of alanine.[8]
This positive ligand effect is to our knowledge one of the
strongest disclosed for any metal-based Lewis acid catalyzed
reaction.[9–11] The large difference in initial rate constants
offers a considerable synthetic advantage. After 16 h, the
ligand-accelerated reaction afforded 1 in 96% yield, whereas
a yield of only 12% was observed for the reaction without
alanine. The ligand-free reaction was reported previously to
take 5 days to reach completion.[5d]
Dr. J. Wennerberg, S. Schultz
DuPont Chemoswed
P. O. Box 839, 20180 Malmö (Sweden)
Fax: (+46)40-186-395
E-mail: johan.wennerberg@swe.dupont.com
K. Aplander,[+] Dr. R. Ding[+]
Division of Organic Chemistry
Lund University
P. O. Box 124, 22100 Lund (Sweden)
[+] These two authors contributed equally to this article.
[**] This research was partially supported by the Swedish Research
Council and the Crafoord Foundation. R.D. acknowledges a post-
doctoral fellowship from the Carl Trygger Foundation.
We then applied the Yb(OTf)3/alanine combination in
reactions between various donors and acceptors (Table 1).
The products were isolated in high yields (88–97%) after
relatively short reaction times (2–3 h) when nearly equivalent
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2007, 46, 4543 –4546
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4543

Communications
Table 1: Michael addition reactions in aqueous suspension catalyzed by
Yb(OTf)3/d-alanine.
in which the Michael addition was performed under “organo-
catalytic” conditions (see the Supporting Information) gave
the products with very low yields and ee values. These results
gave us confidence that an enamine mechanism is not a
significant contributor to the acceleration or to the enantio-
selectivities we observe (see below).[15]
Entry
Donor
Acceptor
Cond.[a]
t [h]
Yield [%][b]
1
A
3
96
We propose that the a-amino acid accelerates the reaction
by acting as a ligand. The deprotonated amino acid first
displaces one of the labile triflate ligands of Yb(OTf)3 to give
A (Scheme 1); a considerable amount of experimental data
2
A
3
97
3
4
A
A
2
2
88
90
5
6
B
B
4
78
94
24
7
8
B
B
24
30
90
41
[a] Reaction conditions: A) donor (1 mmol), acceptor (1.1 mmol), Yb-
(OTf)3 (0.1 mmol), d-alanine (0.12 mmol), NaOH (0.12 mmol), water
(1 mL), 608C; B) donor (1 mmol), acceptor (3 mmol), Yb(OTf)3
(0.2 mmol), d-alanine (0.24 mmol), NaOH (0.24 mmol), water (2 mL),
608C. [b] Yield of the isolated product.
Scheme 1. Formation of the postulated “amino acid activated” catalytic
complex B that leads to the addition product.
exists to support the formation of 1:1 complexes between
lanthanides and a-amino acids in water.[14] Upon the addition
of the reactants, another triflate ligand is displaced by a
bidentate b-ketoenolate to give complex B. Simultaneous
coordination of the enone brings the reactants into proximity
for the addition step. A clue to the origin of the large ligand
effect may then be derived from a recent study by Christoffers
and co-workers, who showed that the coordination of
negatively charged ligands can enhance significantly the
reactivity of a b-ketoenolate towards an electrophile, pre-
amounts of the donor and MVK were used (Table 1,
entries 1–4). Although the use of internal olefins as acceptors
is normally much more challenging than the use of terminal
olefins, the desired product was isolated in 78% yield from
the reaction between ethyl acetoacetate and 2-cyclohexen-1-
one (3 equiv) after only 4 h (Table 1, entry 5). High yields
(90–94%) were observed for the addition of acetylacetone to
cyclic enones if the reaction times were extended to 24 h
(Table 1, entries 6 and 7). Even the unreactive acceptor
benzalacetone underwent a slow but clean Michael addition
with acetylacetone to provide the adduct in 41% yield after
30 h, along with recovered starting material (Table 1, entry 8).
A ligand acceleration effect of the magnitude observed in
this study is counter-intuitive. In fact, a negative ligand effect,
that is, rate deceleration, is the most commonly observed
outcome of the addition of a basic ligand to a Lewis acid
catalyzed reaction.[12] Although a significant amount of
research has gone into understanding transition-metal-cata-
lyzed addition reactions of 1,3-dicarbonyl compounds to
enones,[13] as well as a-amino acid–lanthanide complex-
ation,[14] any discussion about ligand effects must start by
recognizing and disproving other potential catalytic pathways.
Is the rate acceleration really due to ligand–metal binding and
not to some alternative mechanism? First, we eliminated the
possibility that small amounts of acid, potentially formed
upon hydrolysis of the metal salt, had a catalytic effect; no
addition product 1 was observed after 16 h at 258C in the
presence of triflic acid (5%). Second, a series of experiments
À
sumably by weakening the covalent character of its M O
bond.[13b] The authors showed that the coordination of
common counterion ligands, such as ClÀ, leads to fast
protonation of the enolate by the acidic reaction mixture
and a retarded reaction rate. In our system, however, the
dibasic character of the a-amino acid ligand may give a
kinetic advantage to the reaction between the b-ketoenolate
and a coordinated electrophilic Michael acceptor.
We then screened a number of a-amino acids for their
ability to induce asymmetry in the reaction between acetyl-
acetone and 2-cyclohexen-1-one. The observed ee values of
adduct 2 were modest (16–53% ee; Scheme 2). Nevertheless,
our results compare favorably to the best previously reported
ee value of 36% for aqueous Lewis acid catalysis with native
a-amino acid ligands.[3a] Higher enantioselectivities have been
attained by using synthetically modified a-amino acids
(76% ee)[3a] or by using a cosolvent (56% ee).[16] Unexpect-
edly, the sterically more demanding a-amino acids valine and
isoleucine gave similar results (38% ee) to that with alanine
4544
www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4543 –4546

Angewandte
Chemie
biphasic Lewis acid catalysis. A ligand acceleration factor of
138 was measured for alanine in an ytterbium triflate
catalyzed Michael addition. High-yielding reactions were
observed with various donors and a wider range of acceptors
than could be used previously under conditions of aqueous,
non-solid-phase Lewis acid catalysis. The enantioselectivities
of the reactions were the highest observed with native a-
amino acids as ligands for Lewis acids in water alone. Still in
its infancy, aqueous biphasic Lewis acid catalysis has tremen-
dous potential in the development of more economical and
more environmentally friendly processes. We will pursue the
opportunities revealed herein in terms of the large ligand
effect and the efficient and extremely simple recycling of the
catalyst. However, we first require deeper insight into the
mechanism behind the rate acceleration and the atypical
relationship between reaction temperature and selectivity.
Scheme 2. Yb(OTf)3/a-amino acid catalyzed asymmetric Michael addi-
tion between acetylacetone and 2-cyclohexen-1-one (amino acid,
ee value of 2: Asn, 16%; Gln, 18%; Ser, 22%; Leu, 25%; Phe, 32%;
Pro, 33%; Trp, 36%; Val, 38%; Ile, 38%; Ala, 39%; Ala (608C), 53%).
(39% ee), the sterically least demanding a-amino acid.[17] To
our further surprise, we observed higher enantioselectivity in
the reaction with alanine at 608C (53% ee) than at 408C
(39% ee). This atypical relationship between temperature
and selectivity was also observed in the addition of acetyl-
acetone to benzalacetone (15% ee at 608C versus 9% ee at
408C).[18] An explanation for this observation cannot be
derived from our current knowledge of the mechanism, but Experimental Section
Typical procedure: d-Alanine (6.3 mg, 0.07 mmol) and ytterbium
the phenomenon may stem from variations in solubility with
temperature and/or a mechanism dominated by entropy
effects.[19] The lowest ee values were observed when a-
amino acids with polar side chains were used (Asn, Gln,
Ser; 16–22% ee).
triflate (36.5 mg, 0.06 mmol) were stirred in aqueous NaOH (0.06m,
1.2 mL) in a round-bottomed flask for 15 min at room temperature.
Ethyl 2-oxocyclohexanecarboxylate (94 mL, 0.59 mmol) and methyl
vinyl ketone (53 mL, 0.65 mmol) were then added. The reaction flask
was sealed, and the mixture was stirred vigorously at 608C for 3 h. The
reaction mixture was then diluted with water and extracted with ethyl
acetate (6 mL). The organic phase was dried over MgSO4, filtered,
and concentrated. Flash column chromatography of the residue on
silica gel (heptane/ethyl acetate 2:1) afforded 3 (136 mg, 96%).
Finally, we performed a recycling experiment under
aqueous biphasic conditions (Figure 2). Ethyl 2-oxocyclohex-
anecarboxylate (6.3 mmol) and MVK (6.9 mmol) were added
to a solution of the catalyst (10% Yb, 12% d-Ala), and the
resulting suspension was stirred at 608C for 8.5 h. When the
stirring was stopped, the mixture separated slowly into two
distinct phases. Following the facile separation of the mixture
in a separating funnel, the aqueous catalytic phase was
returned to the reaction flask, and another batch of reactants
was added. Five such cycles provided consistently complete
conversion of the reactants into the desired adduct 3, without
the use of an organic solvent. Remarkably, no product was
detected in the absence of the catalyst.
Received: February 7, 2007
Published online: May 4, 2007
Keywords: amino acids · aqueous media · Lewis acid catalysis ·
.
ligand effects · Michael addition
[1] For some recent reviews, see: a) Organic Reactions in Water
(Ed.: U. M. Lindström), Blackwell Publishing, Oxford, 2007;
b) S. Kobayashi, C. Ogawa, Chem. Eur. J. 2006, 12, 5955; c) M. C.
Pirrung, Chem. Eur. J. 2006, 12, 1312; d) C. J. Li, Chem. Rev.
2005, 105, 3095; e) Aqueous-Phase Organo-
metallic Catalysis, 2nd ed. (Eds.: B. Cornils,
W. A. Herrmann), Wiley-VCH, Weinheim,
2004; f) K. Manabe, S. Kobayashi, Chem. Eur.
J. 2002, 8, 4094.
In conclusion, we have shown that natural a-amino acids
have significant potential as chiral ligands for aqueous
[2] For a comprehensive treatise on the concept
of aqueous biphasic catalysis, see Ref. [1e];
see also recent research on the closely related
“on water” concept: S. Narayan, V. V. Fokin,
K. B. Sharpless in ref. [1a], pp. 350–365; S.
Narayan, J. Muldoon, M. G. Finn, V. V. Fokin,
H. C. Kolb, K. B. Sharpless, Angew. Chem.
2005, 117, 3339; Angew. Chem. Int. Ed. 2005,
44, 3275.
[3] For the most prominent precedent, see: a) S.
Otto, J. B. F. N. Engberts, J. Am. Chem. Soc.
1999, 121, 6798; see also: b) J. Gyarmati, C.
Hajdu, Z. Dinya, K. Micskei, C. Zucchi, G.
Pµlyi, J. Organomet. Chem. 1999, 586, 106.
[4] a) H. Li, Y. Wang, L. Tang, F. Wu, X. Liu, C.
Guo, B. M. Foxman, L. Deng, Angew. Chem.
Figure 2. The high water solubility of the Yb(OTf)3/alanine catalyst allows simple isolation
of the Michael adduct and repeated recycling of the catalytic aqueous phase.
2005, 117, 107; Angew. Chem. Int. Ed. 2005,
44, 105; b) J. Christoffers, A. Baro, Angew.
Angew. Chem. Int. Ed. 2007, 46, 4543 –4546
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
4545

Communications
Chem. 2003, 115, 1726; Angew. Chem. Int. Ed. 2003, 42, 1688;
[12] D. J. Berrisford, C. Bolm, K. B. Sharpless, Angew. Chem. 1995,
107, 1159; Angew. Chem. Int. Ed. 1995, 34, 1059.
c) N. Krause, A. Hoffmann-Röder, Synthesis 2001, 171.
[5] a) S. Shirakawa, S. Kobayashi, Synlett 2006, 1410; b) Y. Mori, K.
Kakumoto, K. Manabe, S. Kobayashi, Tetrahedron Lett. 2000, 41,
3107; c) E. Keller, B. L. Feringa, Synlett 1997, 842; d) E. Keller,
B. L. Feringa, Tetrahedron Lett. 1996, 37, 1879.
[6] T. Kawabata, M. Kato, T. Mizugaki, K. Ebitani, K. Kaneda,
Chem. Eur. J. 2005, 11, 288.
[7] R. Ding, K. Katebzadeh, L. Roman, K.-E. Bergquist, U. M.
Lindström, J. Org. Chem. 2006, 71, 352.
[13] a) J. Comelles, M. Moreno-Maꢀas, E. Pꢁrez, A. Roglans, R. M.
Sebastiµn, A. Vallribera, J. Org. Chem. 2004, 69, 6834; b) S.
Pelzer, T. Kauf, C. van Wꢂllen, J. Christoffers, J. Organomet.
Chem. 2003, 684, 308; c) J. Christoffers, Eur. J. Org. Chem. 1998,
1259.
[14] For a review, see: C. Kremer, J. Torres, S. Domꢃnguez, A.
Mederos, Coord. Chem. Rev. 2005, 249, 567.
[15] Aqueous enamine reactions in the presence of catalytic amounts
of amino acids and peptides have been reported recently: a) N.
Mase, K. Watanabe, H. Yoda, K. Takabe, F. Tanaka, C. F.
Barbas III, J. Am. Chem. Soc. 2006, 128, 4966; b) P. Dziedzic, W.
Zou, J. Hafrꢁn, A. Cꢄrdova, Org. Biomol. Chem. 2006, 4, 38.
[16] a) T. Darbre, M. Machuqueiro, Chem. Commun. 2003, 1090;
b) R. Fernandez-Lopez, J. Kofoed, M. Machuqueiro, T. Darbre,
Eur. J. Org. Chem. 2005, 5268.
[17] This small variation in the ee values observed with amino acids
with large and small side chains further supports our claim that
an enamine mechanism is not involved significantly.
[18] A similar nonlinear relationship between temperature and
enantioselectivity was observed previously for a Lewis acid
catalyzed Michael addition: A. Corma, S. Iborra, I. Rodriguez,
M. Iglesias, F. Sµnchez, Catal. Lett. 2002, 82, 237.
[8] For details of the kinetic experiments that led to Figure 1,
including the derivation of rate constants, see the Supporting
Information.
[9] Many reports on ligand effects in Lewis acid catalysis are based
on qualitative observations rather than kinetic data. It is
therefore difficult to make any definitive statement on how
this ligand effect compares with those observed previously. It
certainly matches the strongest Lewis acid ligand effect that we
have been able to find in recent literature: W. G. Lewis, F. G.
Magalion, V. V. Fokin, M. G. Finn, J. Am. Chem. Soc. 2004, 126,
9152.
[10] For examples of modest ligand effects in aqueous media, see:
a) S. Kobayashi, N. Aoyama, K. Manabe, Chirality 2003, 15, 124;
b) S. Kobayashi, N. Aoyama, K. Manabe, Synlett 2002, 483;
c) ref. [3a].
[19] We thank one of the reviewers for pointing out this possibility.
[11] For another example of ligand-accelerated Yb(OTf)3 catalysis,
see: V. K. Aggarwal, A. Mereu, G. J. Tarver, R. McCague, J. Org.
Chem. 1998, 63, 7183.
4546 www.angewandte.org
ꢀ 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2007, 46, 4543 –4546