Direct asymmetric aldol reactions of glycine Schiff base with aldehydes catalyzed by chiral quaternary ammonium salts

Article abstract of DOI:10.1002/1521-3773(20021202)41:23<4542::AID-ANIE4542>3.0.CO;2-3

A practical and environmentally benign chemical process for the synthesis of optically active β-hydroxy-α-amino acids involves the reaction of an aldehyde, glycine Schiff base 2, in the presence of catalytic N-spiro chiral quaternary ammonium bromide 1, T

Full text of DOI:10.1002/1521-3773(20021202)41:23<4542::AID-ANIE4542>3.0.CO;2-3

COMMUNICATIONS
[27] The catalytic rate (R) of the oxidation of benzyl alcohol was expressed
by R ¼ k2[C]/{1 þ k2/(k1[A])}, where [C] and [A] are the amount of Ru/
Al2O3 and the concentration of benzyl alcohol, respectively, and k1
and k2 are the rate constants for step 1 and step 2, respectively, and at
356 K these were 38 hꢀ1 gꢀ1 and 42m hꢀ1 gꢀ1, respectively.
[28] J. W. Mellor, A Comprehensive Treatise on Inorganic and Theoretical
Chemistry, Vol. XV, Longmans & Green, London, 1936, pp. 498 544.
[29] B.-Z Wan, J. H Lunsford, Inorg. Chim. Acta 1982, 65, 29; M. G.
Cattania, A. Gervasini, F. Morazzoni, P. Scotti, D. Strumolo, J. Chem.
Soc. Faraday Trans. 1 1987, 83, 3619.
cally pure, C2-symmetric chiral quaternary ammonium salt
1[14] as a phase-transfer catalyst (Scheme 1). This approach
provides a practical and environmentally benign chemical
process for the synthesis of optically active b-hydroxy-a-
amino acids.
Initially, we examined the direct asymmetric aldol reaction
of prochiral glycine Schiff base 2 and 3-phenylpropanal as a
representative acceptor under phase-transfer conditions.
Thorough optimization of the catalyst structure and reaction
conditions revealed that treatment of 2 with 3-phenylpropanal
(2 equiv) in toluene/aqueous NaOH (1%) (v/v 1.25:1; 2 equiv
of base for 2) in the presence of chiral quaternary ammonium
Direct Asymmetric Aldol
Reactions of Glycine Schiff Base
with Aldehydes Catalyzed by
Chiral Quaternary Ammonium
Salts**
Takashi Ooi, Mika Taniguchi,
Minoru Kameda, and Keiji Maruoka*
As naturally occurring a-amino acids as
well as components of many complex bio-
logically active cyclic peptides and enzyme
inhibitors, optically active b-hydroxy-a-amino
acids are extremely important chiral units,
especially from the pharmaceutical view-
point.[1] Furthermore, they are useful chiral
building blocks in organic synthesis[2] as
exemplified by their transformation into b-
lactams,[3] b-halo-a-amino acids,[4] and aziri-
dines.[5] Accordingly, numerous methods for
the asymmetric synthesis of b-hydroxy-a-
Scheme 1. Direct catalytic asymmetric aldol reaction of glycine Schiff base 2 with aldehydes in
the presence of 1 under phase-transfer conditions. a) (R,R)-1 (2 mol%), toluene/aqueous
NaOH (1%), 08C, 2 h; b) HCl (1n)/THF.
amino acids have been elaborated, most of
which unfortunately involve multistep proce-
dures and/or the inevitable use of a stoichio-
8]
metric amount of chiral auxiliaries.[6 In this
regard, the construction of their primary structure with the
correct stereochemistry by the direct catalytic asymmetric
aldol reaction of a glycine donor with aldehyde acceptors has
been considered to be an ideal protocol.[9] However, there
have been few successful examples to date,[10,11] except for the
chemoenzymatic process with glycine-dependent aldola-
ses.[9b,12] We report herein an efficient and direct asymmetric
aldol reaction of glycine Schiff base 2[13] with aldehydes under
organic/aqueous biphasic conditions by using enantiomeri-
salt (R,R)-1a (2 mol%)[14d] at 08C for 2 h and subsequent
hydrolysis with HCl (1n) in THF resulted in the formation of
the corresponding b-hydroxy-a-amino ester
3
(R ¼
PhCH2CH2) in 76% yield with the anti/syn ratio of 3.3:1.
The enantiomeric excess of the major anti isomer was
determined to be 91% by chiral HPLC analysis (Table 1,
entry 1). Significantly, use of (R,R)-1b (which contains a 3,5-
bis(3,5-bis(trifluoromethyl)phenyl)phenyl substituent) as a
catalyst enhanced both diastereo- and enantioselectivities in
this system (anti/syn 12:1; 96% ee for anti isomer) (Table 1,
entry 2).[15]
[*] Prof. K. Maruoka, Dr. T. Ooi, M. Taniguchi, M. Kameda
Department of Chemistry
A variety of aldehydes were examined for this direct
asymmetric aldol reaction with (R,R)-1 and the results are
listed in Table 1. Generally, the reaction proceeded smoothly
at 08C for 2 h to afford the anti isomer predominantly, with
excellent enantioselectivity. Heptanal, an aliphatic aldehyde
with a long hydrocarbon chain, was found to be a good
candidate (Table 1, entry 3), thus indicating the feasibility of
direct asymmetric synthesis of a variety of lipo b-hydroxy-a-
amino acids, a useful component for the preparation of
lipophilic peptides and glycopeptides with characteristic high
Graduate School of Science, Kyoto University
Sakyo, Kyoto, 606-8502 (Japan)
Fax : (þ 81)75-753-4041
E-mail: maruoka@kuchem.kyoto-u.ac.jp
[**] This work was partially supported by a Grant-in-Aid for Scientific
Research from the Ministry of Education, Culture, Sports, Science,
and Technology, Japan. M.K. is grateful to the Japan Society for the
Promotion of Science for Young Scientists for a Research Fellowship.
Supporting information for this article is available on the WWW under
http://www.angewandte.org or from the author.
4542
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4542 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 23

COMMUNICATIONS
Table 1. Direct asymmetric aldol reactions of 2 with aldehydes by chiral
phase-transfer catalysis of 1.[a]
Experimental Section
Aqueous NaOH (1%; 2.4 mL) was added at 08C under an argon
atmosphere to a solution of tert-butyl glycinate benzophenone Schiff base
2 (88.6 mg, 0.3 mmol) and (R,R)-1b (9.9 mg, 2 mol%) in toluene (3.0 mL),
and 3-phenylpropanal (79.0 mL, 0.6 mmol) was introduced dropwise. The
whole mixture was stirred for 2 h at 08C, and water and diethyl ether were
then added. The ether phase was separated and washed with brine. The
organic phase was dried over Na2SO4 and concentrated. The crude product
was dissolved in THF (8.0 mL) and treated with HCl (1.0n, 1.0 mL) at 08C
for 1 h. After removal of THF in vacuo, the aqueous solution was washed
with diethyl ether three times and neutralized with NaHCO3. The mixture
was then extracted with CH2Cl2 three times. The combined extracts were
dried over MgSO4 and concentrated. Purification of the residue by column
chromatography on silica gel with MeOH/CH2Cl2 (1:15) as eluent afforded
the corresponding b-hydroxy-a-amino ester 3 (R ¼ PhCH2CH2) (56.8 mg,
0.214 mmol, 71%, anti/syn 12:1). anti-3 (R ¼ PhCH2CH2): 1H NMR
(400 MHz, CDCl3): d ¼ 7.30 7.26 (m, 2H; Ph), 7.22 7.16 (m, 3H; Ph), 3.77
(ddd, J ¼ 3.2, 4.4, 10.0 Hz, 1H; CHOH), 3.47 (d, J ¼ 4.4 Hz, 1H; CHNH2),
2.91 2.84 (ddd, J ¼ 4.8, 9.2, 13.6 Hz, 1H; PhCH), 2.73 2.65 (ddd, J ¼ 8.0,
8.0, 13.6 Hz, 1H; PhCH), 1.85 (br s, 3H; OH and NH2), 1.75 1.65 (m, 1H;
PhCH2CH), 1.62 1.53 (m, 1H; PhCH2CH), 1.41 ppm (s, 9H; tBu). syn-3
(R ¼ PhCH2CH2): d ¼ 7.29 7.25 (m, 2H; Ph), 7.22 7.16 (m, 3H; Ph), 3.70
(ddd, J ¼ 4.8, 5.2, 7.6 Hz, 1H; CHOH), 3.24 (d, J ¼ 5.2 Hz, 1H; CHNH2),
2.90 2.82 (ddd, J ¼ 6.2, 9.0, 13.6 Hz, 1H; PhCH), 2.74 2.67 (ddd, J ¼ 7.2,
8.8, 13.6 Hz, 1H; PhCH), 2.17 (br s, 3H; OH and NH2), 1.85 1.78 (m, 2H;
PhCH2CH2), 1.46 ppm (s, 9H; tBu). The enantiomeric excess of the major
anti isomer was determined to be 96% by HPLC analysis of its N-benzoate:
chiral column, DAICEL Chiralcel OD-H, hexane/2-propanol (20:1), flow
Entry
R
Catalyst Yield [%][b] anti/syn[c] ee [%][d]
1
2
3
4
5
6
7
8
9
PhCH2CH2
PhCH2CH2
CH3(CH2)4CH2
iPr3SiOCH2
¼
1a
1b
1b
1b
76
71
65
72
62
71
58
40
78
3.3:1
12:1
10:1
20:1
6.3:1
2.4:1
2.3:1
2.8:1
1.2:1
91
96
91
98
80
90
92[e]
95
93[f]
CH2 CHCH2CH2 1b
CH2 CHCH2CH2 1a
¼
CH3
c-C6H11
c-C6H11
1a
1a
1a
[a] Unless otherwise specified, the reaction was carried out with 2 equivof
aldehyde in the presence of (R,R)-1 (2 mol%) in toluene/aqueous NaOH
(1%) at 08C for 2 h. [b] Yield of isolated product. [c] Determined by
1H NMR analysis. [d] Enantiomeric excess of anti-3, which was determined
by HPLC analysis of its N-benzoate by using a chiral column (DAICEL
Chiralcel OD-H) with hexane/2-propanol or hexane/ethanol as solvent.
[e] The reaction with (S,S)-1a as catalyst displayed similar reactivity and
selectivity. [f] Use of dibutyl ether as solvent.
enzymatic stability and enhanced drug transport activity.[16]
The reaction with a-triisopropylsiloxyacetaldehyde cleanly
produced the desired b-hydroxy-a-amino ester 3 (R ¼ iPr3-
SiOCH2) in 72% yield with virtually complete stereochemical
control (anti/syn 20:1; 98% ee) (Table 1, entry 4), which
parallels the l-threonine aldolase catalyzed aldol reaction
used for the synthesis of the monobactam antibiotic caruno-
man and its analogues.[12b] Moreover, a key building block for
the synthesis of the carbacephem antibiotic loracarbef,
previously prepared by a chemoenzymatic process with serine
hydroxymethyltransferase (SHMT),[12f] was readily assembled
with 4-pentenal as acceptor; (R,R)-1a resulted in a higher
enantioselectivity than did (R,R)-1b (Table 1, entries 5 and 6).
We also found that l-allo-threonine tert-butyl ester can be
obtained by the reaction of 2 with acetaldehyde in the
presence of (R,R)-1a, which confirmed that the absolute
configuration of the newly created a stereocenter is S
(Table 1, entry 7).[17] This method allows a facile preparation
of non-natural d-allo-threonine because of the ready avail-
ability of the enantiomerically pure catalyst (S,S)-1a. Inter-
estingly, (R,R)-1a was also a suitable catalyst for the reaction
with a-substituted aldehydes such as cyclohexanecarbalde-
hyde as acceptor; to attain sufficient reactivity, dibutyl ether
was used as solvent (Table 1, entries 8 and 9).
In summary, the direct asymmetric aldol reaction of glycine
Schiff base 2 with aldehyde acceptors proceeds under mild
organic/aqueous biphasic conditions with excellent stereo-
chemical control by using chiral quaternary ammonium salts
as catalysts. This reaction offers a powerful chemical method
for the synthesis of optically active b-hydroxy-a-amino acids,
and complements the aldolase-based chemoenzymatic proc-
esses. Therefore, the present system can be regarded as an
artificial glycine-dependent aldolase; its operational simplic-
ity, environmentally friendly conditions, and suitability for
large-scale reaction represent distinct advantages for practical
industrial applications.
rate ¼ 0.5 mLminꢀ1
,
retention times 23.06 min (minor) and 39.79 min
(major).
Received: July 16, 2002 [Z19746]
[1] a) Glycopeptide Antibiotics (Ed.: R. Nagarajan), Marcel-Dekker,
New York, 1994; b) Amino Acids, Peptides and Proteins; Special
Periodical Reports; Vols. 1 28, Chemistry Society, London, 1968
1995; c) M. A. Blaskovich, G. Evindar, N. G. W. Rose, S. Wilkinson, Y.
Luo, G. A. Lajoie, J. Org. Chem. 1998, 63, 3631, and references
therein.
[2] a) G. M. Coppola, H. F. Schuster, Asymmetric Synthesis: Construction
of Chiral Molecules Using Amino Acids, Wiley, Toronto, 1987; b) A.
Goleciowski, J. Jurczak, Synlett 1993, 241; c) J.-P. Genet, Pure Appl.
Chem. 1996, 68, 593.
[3] M. J. Miller, Acc. Chem. Res. 1986, 19, 49.
[4] a) S. V. Pansare, J. C. Vederas, J. Org. Chem. 1987, 52, 4804; b) R.
Badorrey, C. Cativiela, M. D. Diaz-de-Villegas, J. A. Galvez, Tetrahe-
dron: Asymmetry 2000, 11, 1015.
[5] D. Tanner, Angew. Chem. 1994, 106, 625; Angew. Chem. Int. Ed. Engl.
1994, 33, 599.
[6] For recent examples, see: a) A. M. P. Koskinen, H. Hasilla, V. T.
Myllymaki, K. Rissanen, Tetrahedron Lett. 1995, 36, 5619; b) H. Shao,
M. Goodman, J. Org. Chem. 1996, 61, 2582; c) T. Nagamitsu, T.
Sunazuka, H. Tanaka, S. Omura, P. A. Sprengeler, A. B. Smith III. , J.
Am. Chem. Soc. 1996, 118, 3584; d) C. Cativiela, M. D. Diaz-de-
Villegas, J. A. Galvez, J. I. Garcia, Tetrahedron 1996, 52, 9563; e) M.
Horikawa, Y. Shigeri, N. Yumoto, S. Yoshikawa, T. Nakajima, Y.
Ohfune, Bioorg. Med. Chem. Lett. 1998, 8, 2027; f) C. A. Hutton, Org.
Lett. 1999, 1, 295; g) P. D. Felice, G. Porzi, S. Sandri, Tetrahedron:
Asymmetry 1999, 10, 2191.
[7] With enolates of glycine derivatives: a) S. Kanemasa, T. Mori, A.
Tatsukawa, Tetrahedron Lett. 1993, 34, 8293; b) Y. N. Belokon, K. A.
Kochetkov, N. S. Ikonnikov, T. V. Strelkova, S. R. Harutyunyan, A. S.
Saghiyan, Tetrahedron: Asymmetry 2001, 12, 481; c) J. B. MacMillan,
T. F. Molinski, Org. Lett. 2002, 4, 1883.
[8] For a few examples in which a catalytic amount of chiral sources is
employed, see: a) Y. Ito, M. Sawamura, E. Shirakawa, K. Hayashizaki,
T. Hayashi, Tetrahedron 1988, 44, 5253; b) H. Suga, K. Ikai, T. Ibata,
Tetrahedron Lett. 1998, 39, 869; c) D. A. Evans, J. M. Janey, N.
Magomedov, J. S. Tedrow, Angew. Chem. 2001, 113, 1936; Angew.
Chem. Int. Ed. 2001, 40, 1884.
[9] For a recent review on the direct catalytic asymmetric aldol reaction,
see: a) B. Alcaide, P. Almendros, Eur. J. Org. Chem. 2002, 1595; see
Angew. Chem. Int. Ed. 2002, 41, No. 23
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4543 $ 20.00+.50/0
4543

COMMUNICATIONS
also: b) R. Mahrwald, Chem. Rev. 1999, 1095; c) T. D. Machajewski,
C.-H. Wong, Angew. Chem. 2000, 112, 1406; Angew. Chem. Int. Ed.
2000, 39, 1352.
ofluoromethane would have positive optical rotation, assum-
ing that the polarizabilities decrease in the order Br> Cl >
H > F.[2] Hargreaves and Modarai were able to resolve the
enantiomers of 1-bromo-1-chloro-1-fluoroacetone and con-
vert them into the corresponding bromochlorofluoromethane
enantiomers.[3] They reported the specific rotations to be
þ 0.20 and ꢀ0.13. In 1973 Applequist predicted, using an
atom dipole interaction model and assuming the polarizabil-
ities to be in the order, Br> Cl > F > H, that (S)-bromochlor-
ofluoromethane would have positive optical rotation.[4] How-
ever, the different polarizability orders chosen by Brewster
and Applequist would yield opposite conclusions.[5] Collet and
co-workers achieved optical resolution of bromochlorofluoro-
methane by enantioselective inclusion in cryptophan C.[6]
Using the NMR resonance signals of (þ) and (ꢀ) enantiomers
encapsulated in cryptophan, they estimated the enantiomeric
excess and maximum rotation values at five different wave-
lengths. In a different approach, Wilen et al. also resolved the
enantiomers of bromochlorofluoromethane using brucine[5]
and addressed the ambiguity remaining in the absolute
configuration of bromochlorofluoromethane; Brewster and
Applequist had arrived at the same assignment but using
mutually conflicting polarizability trends. Wilen et al. con-
cluded that ™experimental determination of the absolute
configuration of bromochlorofluoromethane is a challenge∫.[5]
Using a quantum-mechanical static method, the specific
rotation of (R)-bromochlorofluoromethane at the sodium D
line was predicted (without the Lorentz factor) to be ꢀ6.[7,8]
Comparison of this value with the experimental value of ꢀ1.78
reported by Collet et al. for enantiopure (ꢀ)-bromochloro-
fluoromethane, and based on the comparison of experimental
and ab initio predicted Raman optical-activity (ROA) spectra,
it was concluded that the absolute configuration of bromo-
chlorofluoromethane is (S)-(þ) and (R)-(ꢀ).[7]
[10] For a synthesis that leads to products with low ee values, see: C. M.
Gasparski, M. J. Miller, Tetrahedron 1991, 47, 5367.
[11] For a Mukaiyama-type reaction with ketene silyl acetal of glycine
Schiff base, see: M. Horikawa, J. Busch-Petersen, E. J. Corey,
Tetrahedron Lett. 1999, 40, 3843.
[12] a) A. Saeed, D. W. Young, Tetrahedron 1992, 48, 2507; b) V. P.
Vassilev, T. Uchiyama, T. Kajimoto, C.-H. Wong, Tetrahedron Lett.
1995, 36, 5063; c) T. Kimura, V. P. Vassilev, G.-J. Shen, C.-H. Wong, J.
Am. Chem. Soc. 1997, 119, 11734; d) T. Miura, M. Fujii, K. Shingu, I.
Koshimizu, J. Naganoma, T. Kajimoto, Y. Ida, Tetrahedron Lett. 1998,
39, 7313; e) N. Wymer, E. J. Toone, Curr. Opin. Chem. Biol. 2000, 4,
110; f) B. G. Jackson, S. W. Pedersen, J. W. Fisher, J. W. Misner, J. P.
Gardner, M. A. Staszak, C. Doecke, J. Rizzo, J. Aikins, E. Farkas, K. L.
Trinkle, J. Vicenzi, M. Reinhard, E. P. Kroeff, C. A. Higginbotham,
R. J. Gazak, T. Y. Zhang, Tetrahedron 2000, 56, 5667.
[13] For reviews on the use of glycine Schiff base for the preparation of
enantiomerically pure a-amino acids, see: a) T. Abellan, R. Chinchil-
la, N. Galindo, G. Guillena, C. Najera, J. M. Sansano, Eur. J. Org.
Chem. 2000, 2689; b) M. J. O©Donnell, Aldrichimica Acta 2001, 34, 3.
[14] a) T. Ooi, M. Kameda, K. Maruoka, J. Am. Chem. Soc. 1999, 121,
6519; b) T. Ooi, M. Takeuchi, M. Kameda, K. Maruoka, J. Am. Chem.
Soc. 2000, 122, 5228; c) T. Ooi, M. Kameda, H. Tannai, K. Maruoka,
Tetrahedron Lett. 2000, 41, 8339; d) T. Ooi, K. Doda, K. Maruoka,
Org. Lett. 2001, 3, 1273; e) T. Ooi, M. Takeuchi, K. Maruoka, Synthesis
2001, 1716.
[15] Attempted reaction of 2 with 3-phenylpropanal using O-allyl-N-(9-
anthracenylmethyl)cinchonidinium bromide as catalyst under other-
wise similar conditions gave rise to 3 (R ¼ PhCH2CH2) in 60% yield
with an anti/syn ratio of 1:5.3; the enantiomeric excess of the major syn
isomer was 25%. Similar syn selectivity was observed in the
Mukaiyama-type reaction catalyzed by the cinchonidine-derived
bifluoride salt.[11]
[16] M. M. Palian, R. Polt, J. Org. Chem. 2001, 66, 7178.
[17] Treatment of the aldol product anti-3 (R ¼ CH3) with HCl (6n) in
MeOH gave rise to the corresponding b-hydroxy-a-amino acid
hydrochloride; its absolute configuration was determined to be
2S,3S by comparison of the optical rotation value with that of the
hydrochloride salt of commercially available l-allo-threonine.
Previous conclusions[7] on the absolute configuration were
based on quantum-mechanical calculations of both ROA and
specific rotation, carried out at the Hartree Fock (HF) level
using smaller basis sets. Since then, evidence has been
collected[9] to indicate that the HF level calculations, because
of lack of electron correlation, are not quantitatively accurate.
DFT,[10] which includes electron correlation using density
functionals, has now been widely accepted as the preferred
approach for predicting molecular properties accurately. For
reliable quantum mechanical predictions of specific rotations,
those predicted at the HF level using small basis sets are not
considered to be adequate and predictions with DFT and
larger basis sets are necessary. Electron correlation and larger
basis sets are also important for molecular polarizability
derivatives, and hence Raman properties. Because the abso-
lute configuration of bromochlorofluoromethane was sug-
gested[7] based on HF calculations, the previously predicted
specific rotation and ROA parameters of bromochlorofluoro-
methane must be verified using DFTand larger basis sets. The
DFT method has been implemented recently for the pre-
diction of specific rotation[11a] and of ROA[11b] in the quantum-
mechanical program DALTON.[12] Thus it is important to
reinvestigate the specific rotation and ROA of bromochloro-
fluoromethane. Recently Grimme[13] has reported DFT pre-
dictions of [a]D for bromochlorofluoromethane. Because
The Absolute Configuration of
Bromochlorofluoromethane**
Prasad L. Polavarapu*
Bromochlorofluoromethane is the simplest chiral molecule
that is used as an example to illustrate chirality or asymmetric
carbon atoms, yet its absolute configuration remains uncer-
tain. The synthesis of bromochlorofluoromethane in a pure
form was achieved by Berry and Sturtevant in 1942.[1] They
determined its boiling point, melting point, and refractive
dispersion, but the individual enantiomers were not resolved.
Despite the unavailability of enantiomers and their optical
rotation values, Brewster hypothesized that (S)-bromochlor-
[*] Prof. Dr. P. L. Polavarapu
Department of Chemistry
Vanderbilt University
Nashville, TN 37235 (USA)
Fax : (þ 1)615 322 4936
E-mail: prasad.l.polavarapu@vanderbilt.edu
[**] I thank NSF (CHE0092922) for research grants and Dr. K. Ruud for
the developmental version of DALTON program which was used for
the calculations reported here.
4544
¹ 2002 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
0044-8249/02/4123-4544 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2002, 41, No. 23