Amplification of enantiomeric excess in a proline-mediated reaction

Article abstract of DOI:10.1002/anie.200453997

The evolution of homochirality from simple organic molecules can perhaps be modeled on the proline-mediated aminoxylation of aldehydes (see scheme). The observed accelerating reaction rate combined with an amplification of the enantiomeric excess of the product is attributed to an autoinductive reaction mediated by an adduct of proline and the product.

Full text of DOI:10.1002/anie.200453997

Angewandte
Chemie
Asymmetric Amplification
Amplification of Enantiomeric Excess in a
Proline-Mediated Reaction**
Suju P. Mathew, Hiroshi Iwamura, and
Donna G. Blackmond*
The origin of homochirality has intrigued scientists ever since
the biological importance of l-amino acids and d-sugars was
first recognized. Although a theoretical basis for the evolution
[*] Dr. S. P. Mathew, Prof. D. G. Blackmond
Department of Chemistry, Imperial College
London SW7 2AZ (United Kingdom)
Fax: (+44)207-594-8504
E-mail: d.blackmond@imperial.ac.uk
H. Iwamura
Mitsubishi Pharma Corporation
14 Sunayama, Hasaki-machi, Ibaraki 314-0255 (Japan)
[**] We are grateful to the Department of Chemistry, University of Hull,
where parts of the experimental work was carried out. Stimulating
discussions with Dr. B. G. Cox and Dr. D. L. Lathbury (AstraZeneca)
and Prof. Peter Beak (Illinois), technical assistance with NMR
experiments from Mrs. B. Worthington (Hull), and funding from
EPSRC/DTI/LINK and Mitsubishi Pharma are gratefully acknowl-
edged.
Supportinginformation for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2004, 43, 3317 –3321
DOI: 10.1002/anie.200453997
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3317

Communications
of high optical activity from a minute initial imbalance of
Intrigued by these reports, which imply a fundamental
mechanistic difference between the reaction shown in Equa-
tion (1) and other proline-catalyzed transformations, we
undertookcontinuous monitoring of reaction progress by
reaction calorimetry, a kinetic approach that we have
developed over the past several years as a mechanistic
probe of multistep reactions.[12,13] We observed, quite unex-
pectedly, that the rate of the reaction in Equation (1) rose
steadily throughout the course of the reaction (Figure 1). A
enantiomers was suggested more than half a century ago,[1]
experimental proof of such a concept eluded scientists until a
remarkable report by Soai and co-workers in 1995.[2] The Soai
reaction offered the first, and to date the only, example of an
asymmetric autocatalytic reaction employing a catalyst with a
very low enantiomeric excess and ultimately yielding the
catalyst with a very high enantiomeric excess catalyst as
product. While the Soai reaction serves as a mechanistic
model[3] for the evolution of homochirality, the dialkylzinc
chemistry involved in the reaction is unlikely to have been of
importance in an aqueous prebiotic environment. Therefore
speculation has continued concerning the types of trans-
formations that might have been directly responsible for the
development of high optical activity in biological systems. The
area of amino acid catalysis may hold significant clues to the
evolution of prebiotic chemistry. That prebiotic building
blocks such as sugars can be formed asymmetrically from such
reactions has recently led to speculation about the evolution
of biological homochirality through such routes.[4] We report
herein a proline-mediated reaction exhibiting an accelerating
reaction rate and an amplified, temporally increasing enan-
tiomeric excess of the product. Thus, catalysis with amino
acids is implicated in an autoinductive, selectivity-enhancing
process, providing the first general chemical strategy for the
evolution of biological homochirality from a purely organic
origin.
Figure 1. Reaction rate (filled black circles, correspondingto the left
axis) and percent conversion (shaded grey line, corresponding to the
right axis) versus time for a one-pot, two-consecutive-reaction
sequence of the reaction shown in Equation (1) (CHCl3 solvent,
278 K). Initial concentrations of aldehyde 1 and l-proline 4 were 2.07
and 0.07m, respectively. First reaction: 0.26m PhNO; second reaction:
0.24m PhNO.
The growing field of asymmetric aminocatalysis[5] makes
use of biomimetic strategies via enamine and iminium
intermediate species common to class I aldolase and ketoacid
decarboxylase enzymes. The proline-catalyzed intramolecular
Hajos–Parrish–Eder–Sauer–Wiechert reaction was the first
example employing a similar strategy in organic synthesis.[6,7]
More recently, this approach has been successfully expanded
by several groups, beginning with the first report by List,
Barbas, and Lerner[8] which demonstrated the proline-cata-
lyzed direct asymmetric intermolecular aldol reaction. A
recent addition to this rapidly growing list of aminocatalytic
transformations was offered by two independent and nearly
simultaneous reports of the proline-catalyzed a-aminoxyla-
tion of aldehydes shown in Equation (1).[9,10] Both studies
plot of conversion versus time accordingly showed the
hyperbolic shape indicative of an accelerating reaction rate.
This behavior suggests a process whereby the catalyst is
improving over time, as in autocatalytic or autoinductive
reactions, in which the reaction product either is itself a
catalyst or promotes the formation of a more effective
catalyst. Addition of the reaction product to fresh reagents
gave no discernible heat flow signal, indicating that the effect
is not due simply to an autocatalytic reaction. However, as
shown in Figure 1, when a second dose of reagents was added
to the crude reaction mixture containing the original proline
catalyst and the product of the first reaction, the initial rate at
the outset of this second reaction was as high as that at the end
of the first. These experiments indicate that the rising rate is
not due to substrate inhibition and suggest that the reaction is
mediated by a proline-product adduct that forms over the
course of the reaction and serves as an improved catalyst for
the reaction. Enantioselective autoinductive reactions have
been reported previously,[14,15] including the report by Danda
et al.[14] of an organocatalytic hydrocyanation of a substituted
benzaldehyde.
noted good yields and high ee values with 5–20 mol% catalyst
in reaction times measured on the order of tens of minutes.
This finding was all the more notable given the dramatically
lower activity of other proline-catalyzed reactions; for
example, completion of the aldol reaction between acetone
and isobutyraldehyde required 48 h at 30 mol% catalyst.[10]
Most, recently the scope of this reaction was widened to
include ketone substrates, which also exhibited rapid reaction
times.[11]
Amplification of the enantiomeric excess of the product is
a key feature of a chemical rationalization of the evolution of
biological homochirality. Considerable discussion has ensued
over the past several years about whether proline-catalyzed
reactions exhibit such nonlinear effects, since the Hajos–
Parrish reaction was cited in the context of Kaganꢀs mathe-
matical models for nonlinear effects.[16] The more recent
3318
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 3317 –3321

Angewandte
Chemie
observation by List and co-workers of a strictly linear
relationship between the catalyst and the ee value of the
product in both inter- and intramolecular aldol transforma-
tions,[17] however, has generally been taken to infer that amino
acid catalysis is unlikely to be involved in processes resulting
in amplification of enantiomeric excess.[18] We were intrigued
to find, therefore, that when the reaction shown in Equa-
tion (1) was carried out with non-enantiopure proline, the
enantiomeric excess of the product was higher than that
expected for a linear relationship (Figure 2) and that enan-
Kagan and co-workers provided the first explanations of
the phenomenon of nonlinear product enantioselectivity,[16]
and variations on these models have been proposed for both
catalytic[20,21] and autocatalytic reactions.[3] Scheme 1 outlines
Scheme 1. General mechanism for the product-induced kinetic amplifi-
cation of enantiomeric excess.
a general mechanism for a product-induced reaction in which
both rate and selectivity improve over time for the case in
which one enantiomer of the original catalyst 4 is present in
excess concentration relative to the other (noted as the
“major” and “minor” species, respectively). The first mole-
cules of reaction product 3 are formed through the less-
reactive pathways shown outside the dashed lines, after which
the reaction is dominated by the manifold shown inside the
dashed lines. Within this manifold, product 3 associates
reversibly with the original catalyst 4 to form the new
catalytic species 5, which exhibits higher catalytic activity than
4. As the concentration of 3 increases with increasing
turnover, a larger fraction of both hands of the original
catalyst 4 is driven toward further formation of 5, resulting in
increased reaction rate. If the mismatched catalyst-product
“cross-reaction” rates (3maj–4min and 4maj–3min) are suppressed
relative to those of the matched reaction, the enantiomeric
excess of the improved catalyst 5 will be higher than that of 4
and will increase over time. Accordingly, the enantio-
meric excess of the product 3 will increase over time and
will be amplified relative to that of the original catalyst 4.
The process described in Scheme 1 is kinetic. The
Figure 2. Final enantiomeric excess of the product as a function of the
fraction of l-proline in the reaction shown in Equation (1) carried out
usingmixtures of d- and l-proline. Initial concentrations of aldehyde 1,
nitrosobenzene 2, and total (l+d)-proline 4 were 2.07, 0.7, and
0.07m, respectively (CHCl3 solvent, 278 K). Conversions were greater
than 90%.
tiomeric excess rose over the course of the reaction
(Figure 3).[19] Rate acceleration and continuous improvement
of enantiomeric excess are requisite characteristics for
chemical models of the evolution of homochirality from
precursors of low optical activity.
enantiomeric excess of the product 3 will rise until the
formation of new catalyst 5 has reached equilibrium,
after which the ee value will slowly erode backto the
linear relationship. In agreement with this, we found that
reactions carried out with lower concentrations of pro-
line, which take longer to reach completion, exhibited
milder ultimate asymmetric amplification. However, it is
important to note that such erosion of enantiomeric
excess is predicted only for a closed system such as that
occurring in reaction vials in the laboratory. In an open
system, in which catalyst and product fluxes can exist
across the system boundaries,[22] the chemical propaga-
tion mechanism described in Scheme 1 would permit
enantiomeric excess to continue to rise. Kinetic ampli-
fication of enantiomeric excess as observed in the
present studies could be sustained, requiring only that
Figure 3. Reaction rate and enantiomeric excess of the product of the reac-
tion of Scheme 1 carried out with a mixture of l- and d-proline with 40% ee
(l). Reaction conditions: 278 K, total proline concentration (l+d): 0.071m,
nitrosobenzene: 0.70m, propionaldehyde: 1.95m.
Angew. Chem. Int. Ed. 2004, 43, 3317 –3321
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3319

Communications
the process of equilibration between the original catalyst 4
excess of species 5 relative to that of the original proline 4.
This case is more complex than in a simple first-order kinetic
resolution because the selectivity factor krel will be a function
of the temporally changing product concentration, as shown
in Equation (2):
and the product 3 to form the improved catalyst 5 is slower
than the 5-catalyzed formation of product 3 from substrates.
We may now relate the model shown in Scheme 1 to the
chemistry of the reaction in Equation (1) to suggest a
structure for the improved catalyst 5 and to rationalize its
higher activity. The key to the effectiveness of this system lies
in the fact that the reaction product 3 is multifunctional; it is
both an aldehyde and an amine. Scheme 2 suggests that
d½4LŠ
d½4DŠ
½4LŠ
½4DŠ
kfast½3RŠ þ kslow½3SŠ
kfast½3SŠ þ kslow½3RŠ
¼ krel
krel
¼
ð2Þ
We observed that extended room temperature
preequilibration of proline with excess propionalde-
hyde in CHCl3 produces a clear solution, as compared
to the cloudy suspension normally observed after
mixing for shorter times or at lower temperatures.
The reaction exhibits considerably lower activity fol-
lowing this preequilibration, resulting in less than 10%
conversion into 3 after 1 h[13] and giving linear behavior
in product enantioselectivity. This suggests that new
catalyst 5 cannot form as readily when the concen-
tration of free proline is low, and under such conditions
the initial reaction is directed toward the slower
proline-catalyzed pathway.
The product 3 may also be capable of condensing
with itself, and, in fact, it has been suggested that the
products of this reaction exist in oligomeric form.[10]
Preliminary 1H NMR studies of the crude reaction
product mixture reveal complex spectra.[13] In the case
of reactions carried out with non-enantiopure 4, the
formation of higher-order species could further skew
the enantiomeric excess of the improved catalyst
Scheme 2. Proposed mechanism for product induction in the reaction shown in
[Eq. (1)].
proline 4 may attackthe carbonyl group of the reaction
product 3 to form the new catalyst 5. This reaction is virtually
irreversible on the reaction timescale, since product racemi-
zation was not observed. This species 5 is a special amine
bearing an a-oxygen atom with lone pairs of electrons. The
“alpha effect”[23] describes the unexpectedly high activity of
such a nitrogen nucleophile, thought to be due in part to
stabilization of the transition state by the lone pair on the
oxygen a to the nucleophilic atom. Thus
species 5 relative to 4, resulting in either enhancement or
suppression of the nonlinear effect, depending on the relative
rates at which different species react in such condensation
reactions.
The nucleophilic nitrogen atom of product 3 might also
attackpropionaldehyde directly. Since the reaction was found
not to be autocatalytic, the enamine thus formed cannot be
5 may be a highly efficient competitor to
proline for nucleophilic attackon pro-
pionaldehyde, forming a new enamine,
6. This enamine may be competent to
attackPhNO, forming a transition state
such as 7 by interaction with the carbox-
ylic acid proton as a Brønsted acid
cocatalyst. This leads to the formation
of product 3 and regeneration of the
improved catalyst 5.
The observed asymmetric amplifica-
tion is rationalized as a kinetic resolu-
tion of the proline in the reaction with
product 3 to form 5, as shown in
Scheme 3. The “matched” reactions
4l–3R and 4d–3S dominate the enan-
tiopure cases. If we assume that the
E enamine is the stable product from
either the “matched” or “mismatched”
interaction, we can see that this com-
petitive process alters the enantiomeric
Scheme 3. Proposed kinetic resolution in the formation of 5 through the interaction of proline 4
with the reaction product 3.
3320
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2004, 43, 3317 –3321

Angewandte
Chemie
Strieter, D. G. Blackmond, S. L. Buchwald, J. Am. Chem. Soc.
competent to react with 2 to form a further molecule of
product 3; the lackof a Brønsted acid function as an
intramolecular cocatalyst may compromise its effective-
ness.[24] This argument leads naturally to speculation about
the potential for truly autocatalytic systems based on similar
chemistry. For example, reactions of aldehydic or keto acids
with a nitrogen-containing electrophile might yield an amine
nucleophile capable of reproducing itself. Investigations
along these lines are ongoing in our laboratories.
2003, 125, 13978; d) L. P. C. N. Nielsen, C. P. Stevenson, D. G.
Blackmond, E. N. Jacobsen, J. Am. Chem. Soc. 2004, 126, 1360.
[13] See Supporting Information for further details.
[14] H. Danda, H. Nishikawa, K. Otaka, J. Org. Chem. 1991, 56, 6740.
[15] A. H. Alberts, H. Wynberg, J. Am. Chem. Soc. 1989, 111, 7265.
[16] C. Puchot, O. Samuel, E. Duꢃach, S. Zhao, C. Agami, H. B.
Kagan, J. Am. Chem. Soc. 1986, 108, 2353.
[17] L. Hoang, S. Bahmanyar, K. N. Houk, B. List, J. Am. Chem. Soc.
2003, 125, 16.
[18] For example, no asymmetric amplification in product enantio-
selectivity was observed in the recent study cited in refer-
ence [4], which implicated amino acids as prebiotic catalysts.
[19] Asymmetric amplification distinguishes this reaction from the
organocatalytic example of autoinduction reported by Danda
et al.[14] in which a linear relationship with enantiomeric excess
was observed in cases in which the product was not added to the
reaction mixture.
[20] M. Kitamura, S. Suga, H. Oka, R. Noyori, J. Am. Chem. Soc.
1998, 120, 9800.
[21] D. G. Blackmond, Acc. Chem. Res. 2000, 33, 402.
[22] C. J. Welch, Chirality 2001 14, 425.
The experimental observation of an unexpectedly high,
accelerating reaction rate and an amplified, temporally
increasing enantiomeric excess of product in the proline-
mediated aminoxylation of aldehydes is consistent with a
mechanistic model for a selectivity-enhancing autoinductive
process as given in Schemes 1–3. This represents the first
example of a purely organic reaction exhibiting character-
istics that are key to a chemical rationalization of the
evolution of biological homochirality. A full kinetic model
of these experimental results will be published separately.
[23] J. O. Edwards, R. G. Pearson, J. Am. Chem. Soc. 1962, 84, 16.
[24] S. Bahmanyar, K. N. Houk, J. Am. Chem. Soc. 2001, 123, 11273.
Received: February 11, 2004 [Z53997]
Keywords: asymmetric amplification · asymmetric catalysis ·
.
autocatalysis · chirality · kinetics
[1] a) F. C. Frank, Biochim. Biophys. Acta 1953, 11, 459; b) M.
Calvin, Chemical Evolution, Oxford University Press, Oxford,
1969.
[2] For the first publications, see: a) K. Soai, T. Shibata, H. Morioka,
K. Choji, Nature 1995, 378, 767; b) T. Shibata, K. Choji, T.
Hayase, Y. Aizu, K. Soai, Chem. Commun. 1996, 1235; c) T.
Shibata, H. Morioka, T. Hayase, K. Choji, K. Soai, J. Am. Chem.
Soc. 1996, 118, 471; for a recent review, see: K. Soai, I. Sato,
Chirality 2002, 14, 548.
[3] a) D. G. Blackmond, C. R. McMillan, S. Ramdeehul, A. Schorm,
J. M. Brown, J. Am. Chem. Soc. 2001, 123, 10103; b) D. G.
Blackmond, Adv. Synth. Catal. 2002, 344, 156; c) F. G. Buono,
D. G. Blackmond, J. Am. Chem. Soc. 2003, 125, 8978; d) F. G.
Buono, H. Iwamura, D. G. Blackmond, Angew. Chem. 2004, 116,
2151; Angew. Chem. Int. Ed. 2004, 43, 2099; e) I. D. Gridnev,
J. M. Serafimov, H. Quiney, J. M. Brown, Org. Biomol. Chem.
2003, 1, 3811.
[4] S. Pizzarello, A. L. Weber, Science 2004, 303, 1151.
[5] a) B. List, Synlett 2001, 1675; b) B. List, Tetrahedron 2002, 58,
5573; c) E. R. Jarvo, S. J. Miller, Tetrahedron 2002, 58, 2481; d) S.
France, D. J. Geurin, S. J. Miller, T. Lectka, Chem. Rev. 2003, 103,
2985.
[6] U. Eder, G. Sauer, R. Wiechert, Angew. Chem. 1971, 83, 492;
Angew. Chem. Int. Ed. Engl. 1971, 10, 496.
[7] Z. G. Hajos, D. R. Parrish, J. Org. Chem. 1974, 39, 1615.
[8] B. List, R. A. Lerner, C. F. Barbas, J. Am. Chem. Soc. 2000, 122,
2395.
[9] G. Zhong, Angew. Chem. 2003, 115, 4379; Angew. Chem. Int. Ed.
2003, 42, 4247.
[10] S. Brown, M. Brochu, C. J. Sinz, D. W. C. MacMillan, J. Am.
Chem. Soc. 2003, 125, 10808.
[11] a) A. Bøgevig, H. Sundꢁn, A. Cꢂrdova, Angew. Chem. 2004, 116,
1129; Angew. Chem. Int. Ed. 2004, 43, 1109; b) Y. Hayashi, J.
Yamaguchi, T. Sumiya, M. Shoji, Angew. Chem. Int. Ed. 2004,
116, 1602; Angew. Chem. Int. Ed. 2004, 43, 1112.
[12] a) T. Rosner, A. Pfaltz, D. G. Blackmond, J. Am. Chem. Soc.
2001, 123, 4621; a) U. K. Singh, E. R. Strieter, D. G. Blackmond,
S. L. Buchwald, J. Am. Chem. Soc. 2002, 124, 14104; c) E. R.
Angew. Chem. Int. Ed. 2004, 43, 3317 –3321
www.angewandte.org
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3321