Proline-catalyzed ketone-aldehyde aldol reactions are accelerated by water

Article abstract of DOI:10.1055/s-2004-831296

Proline-catalyzed aldol reactions between acetone or 4-thianone and different aldehydes are accelerated by addition of 1-10 equivalents of water to the reaction medium, allowing stoichiometric aldol reactions to proceed at acceptable rates.

Full text of DOI:10.1055/s-2004-831296

LETTER
1891
Proline-Catalyzed Ketone–Aldehyde Aldol Reactions are Accelerated
by Water
W
ater-Accelerated
n
P
roline-Cata
n
lyzedAldol
i
Reacti
k
ons a I. Nyberg, Annina Usano, Petri M. Pihko*
Laboratory of Organic Chemistry, Helsinki University of Technology, P.O.Box. 6100, FI-02015 HUT, Finland
Fax +358(9)4512538; E-mail: Petri.Pihko@hut.fi
Received 8 April 2004
the List–Barbas procedure also allows the use of less vol-
atile ketones, such as 4-thianone,6 as the donor compo-
nent.
Abstract: Proline-catalyzed aldol reactions between acetone or 4-
thianone and different aldehydes are accelerated by addition of 1–
10 equivalents of water to the reaction medium, allowing stoichio-
metric aldol reactions to proceed at acceptable rates.
The substrate range for the proline-catalyzed aldol reac-
Key words: aldehydes, aldol reactions, ketones, proline catalysis, tions includes simple acyclic ketones such as acetone and
cyclic ketones such as cyclohexanone.1 We initially stud-
water
ied the reaction of acetone with iso-butyraldehyde, benz-
aldehyde and p-nitrobenzaldehyde (Table 1 and
Table 2).7,8 In the reactions with DMSO as the solvent
The direct enantioselective aldol reaction catalyzed by
(Table 1, entries 1 and 2), both aldol product 1 and aldol
condensation product 2 were formed. Switching over to
DMF as the solvent (Table 1, entries 3–16), the product 1
was formed smoothly in a clean reaction and only a mini-
mal amount of the enone 2 was observed. With these re-
actants, the highest conversions were obtained in the
reactions with 100 mol% H2O (Table 1, entries 5 and 11).
However, with p-nitrobenzaldehyde, up to 1000 mol%
H2O was required for the highest conversion (Table 2, en-
try 7). Further addition of more water (up to 100 vol%,
corresponding to 5500 mol%) had a deleterious effect on
the rates of the reactions, and typically no product forma-
tion could be observed. Also, the enantioselectivity suf-
fered, in agreement with the results obtained by Barbas
and co-workers.1a,e
proline or its derivatives is one of the most promising cat-
alytic approaches to asymmetric synthesis of new C-C
bonds. These reactions are of tremendous importance
since they allow the direct construction of biologically
and medicinally important targets in high enantiomeric
excess using metal-free catalysts and very benign reaction
conditions.1 Proline and its derivatives operate by bifunc-
tional catalysis: the amine functionality activates the aldol
donor molecules by converting them into enamines,
whereas the carboxylic acid moiety provides a hydrogen
bond to the acceptor.2 The enamine attacks the carbonyl
group of the aldehyde acceptor with high enantiofacial se-
lectivity. This mechanism is identical to that used by class
I aldolases.3
The optimized conditions reported for the proline-cata-
lyzed aldol reaction between ketone donors and aldehyde
acceptors typically require the use of 30–40 mol% of pro-
line and an large excess of ketone in anhydrous DMSO (or
DMF).1 With more expensive ketones as donors, this is
neither economical nor practical because it complicates
the purification of the product, particularly if a substantial
excess of the ketone is used and the ketone is not volatile.
This is especially true in cases where the product needs to
be purified by chromatography since the starting ketones
and the aldol products sometimes have similar Rf values.
A similar water-accelerating effect was also observed in
reactions performed using an excess of ketone (28:1, ace-
tone–iso-butyraldehyde in DMSO) according to the orig-
inal List–Barbas procedure.1,9 Using DMF as the co-
solvent and water to accelerate the reaction, we could ob-
tain the aldol product 3 from acetone and p-nitrobenzalde-
hyde in 90% yield and 76% ee in only 24 hours reaction
time!10
Compared to acetone, cyclic ketones react more slowly in
proline-catalyzed aldol reactions.11 To our surprise, under
the standard List–Barbas conditions, very little progress
could be observed in the reaction with 4-tert-butylcyclo-
hexanone and benzaldehyde (Scheme 1). While addition
of water (300 mol%) did increase the reaction rate using
the List–Barbas conditions, it was only the stoichiometric
conditions with H2O that gave a synthetically viable reac-
tion rate and yield.12 After eight days, the aldol products
5a and 5b were obtained in 45% yield and 74% ee (for
5a).
Under stoichiometric conditions, the direct aldol reaction
can be much more difficult to accomplish since the equi-
librium constants for many ketone–aldehyde aldol reac-
tions are just barely on the side of the products.4 In
addition, proline can undergo a variety of reactions with
the aldehyde acceptor.5
In this Letter, we report that the stoichiometric aldol reac-
tion becomes feasible if water is added to the reaction
mixtures to accelerate the reaction. Our modification of
To further illustrate the advantage of the water-accelerat-
ed reactions, we turned to 4-thianone, a cyclic, nonvolatile
ketone that could serve as a potential polypropionate
building block (Scheme 2, see below).6 The effect of
SYNLETT 2004, No. 11, pp 1891–1896
Advanced online publication: 06.08.2004
DOI: 10.1055/s-2004-831296; Art ID: G11204ST
© Georg Thieme Verlag Stuttgart · New York
0
6
.0
9
.2
0
0
4

1892
A. I. Nyberg et al.
LETTER
Table 1 Effect of Water on Proline-Catalyzed Stoichiometric Aldol Reactions Using Acetone as the Ketone Component
O
O
L-proline (10 mol%)
O
O
OH
+
H
R
+
H2O additive
r.t.
R
R
1 mmol
1 mmol
1
2
Entry
1
R
Conditions
H2O (mol%)a
1 (%)b
60
ee (%)c
N.d.e
N.d.
90
2 (%)
28
22
<1
<1
4
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
i-Pr
Ph
DMSO, 5dd
DMSO, 5 dd
DMF, 9 d
DMF, 9 d
DMF, 9 d
DMF, 9 d
DMF, 9 d
DMF, 9 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 7 d
DMF, 10 d
0
100
0
2
60
3
31
4
50
39 (32)f
91
5
100
200
300
500
0
42
36
29
29
30
37
45 (44)f
35
28
21
2
92
6
91
3
7
89
3
8
90
9
9
71
2
10
11
12
13
14
15
16
Ph
50
69
4
Ph
100
200
300
500
63
3
Ph
66
2
Ph
67
2
Ph
73
3
Ph
50 vol%
100 vol%
N.d.
N.d.
N.d.
N.d.
Ph
0
a Amount of added H2O (mol%) except in entries 15 and 16; in these entries the reactions were 1 M in aldehyde and ketone.
b Conversion as determined by 1H NMR.
c The ee was determined by chiral HPLC analysis (Chiralcel AD).
d 30 mol% L-proline.
e Not determined.
f Isolated yield.
a) or b)
water was very pronounced in aldol reactions with this
ketone as the donor (Table 3). Only trace amounts of the
desired product 6 were obtained in the absence of water
(Table 3, entries 1 and 7). With iso-butyraldehyde as the
acceptor, addition of 300 mol% water gave the highest
yield and diastereoselectivity (entry 4), whereas the enan-
tioselectivities were unaffected.13
O
a) i) 0 or ii) 300 mol% H2O
L-proline (30 mol%), DMSO
O
+
H
Ph
b) i) 0 or ii) 500 mol% H2O
L-proline (10 mol%), DMF, r.t.
t-Bu
1 mmol
O
a) 25 mmol
b) 1 mmol
Prolonging the reaction time or using an excess of alde-
hyde increased the conversion but led to somewhat de-
creased enantioselectivities (entries 3 and 5). In reactions
with benzaldehyde, the highest yield was obtained with
500 mol% of water (entry 10). In these reactions, the ad-
dition of water also significantly improved the enantiose-
lectivities, and ee values of >90% were consistently
achieved provided that at least 300 mol% of water was
added (entries 9–12).
Conversion after 5d:
O
OH
OH
Ph
Conditions
a) i) 0 mol% H2O: <10%
a) ii) 300 mol% H2O: 15%
2:1 5a: 5b
b) i) 0 mol% H2O: 21%
b) ii) 500 mol% H2O: 42%
2:1 5a:5b
Ph
+
t-Bu
t-Bu
5b
5a
With benzaldehyde, using the standard List–Barbas
conditions with an excess of 4-thianone in DMSO and
20 mol% proline, only 7% conversion was reached after
Scheme 1 Effect of water on reactions of 4-tert-butylcyclohexa-
none and benzaldehyde. The conversions were determined by 1H NMR
from the crude reaction mixture.
Synlett 2004, No. 11, 1891–1896 © Thieme Stuttgart · New York

LETTER
Water-Accelerated Proline-Catalyzed Aldol Reactions
1893
Table 2 Effect of Water on Proline-Catalyzed Stoichiometric Aldol
Reactions with Acetone and p-Nitrobenzaldehyde
O
O
L-proline (10 mol%)
+
H
H2O additive
r.t.
NO2
1 mmol
O
1 mmol
O
OH
+
NO2
ee (%)c 4 (%)
NO2
H2O
4
3
Entry Conditions
3 (%)b
(mol%)a
1
2
DMF, 6 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 3 d
DMF, 15 d
DMF, 15 d
0
10
58
65
69
68
69
66
49
47
–
<1
<1
<1
<1
<1
<1
<1
<1
<1
<1
50
20
Figure 1 Effect of water on the rate of the reaction of 4-thianone
and iso-butyraldehyde (1 mmol 4-thianone, 2 mmol iso-butyralde-
hyde, 10 mol% L-proline, 1 mL DMF, r.t.).
3
100
24
4
200
27
Here, too, the stoichiometric method (Table 3, entry 10) is
the method of choice.
5
300
32
6
500
41 (41)d
Entries 13–15 in Table 3 further illustrate the utility of the
optimized conditions for stoichiometric aldol reactions
with 4-thianone. Very high enantio- and diastereoselectiv-
ities were obtained with other aldehydes as well and the
products were readily isolated. The products derived from
aromatic aldehydes are all highly crystalline solids.
7
1000
1500
78 (82)d
8
52
0
9
50 vol%
100 vol%
10
0
–
We have also demonstrated that the sulfur bridge in the 4-
thianone aldol product can be cleaved to afford potential
polypropionate building blocks. After TBS protection of
the alcohol to prevent unwanted retro-aldolization reac-
tions, desulfurization was readily achieved by Raney-Ni
reduction in 50% unoptimized yield (Scheme 2).14
a Amount of added H2O (mol%) except in entries 9 and 10; in these
entries the reactions were 1 M in aldehyde and ketone.
b Conversion as determined by 1H NMR.
c The ee values were determined by chiral HPLC analysis (Chiralcel
AD).
d Isolated yield.
The reactions between 4-thianone, iso-butyraldehyde and
different amounts of H2O were monitored by 1H NMR. As
shown in Figure 1, the effect of added water was even
more dramatic with 200 mol% of iso-butyraldehyde. The
1H NMR spectra revealed that part of the proline and the
aldehyde under anhydrous conditions condensed to form
a bicyclic oxazolidinone,5a,1b thereby tying up the catalyst.
The oxazolidinone is, however, readily hydrolyzed upon
addition of water.5a,15 Accordingly, the amount of oxazo-
seven days under dry conditions (>90% ee, 2.5:1
anti:syn). In comparison, addition of 500 mol% water re-
stored the catalytic activity (55% conversion after seven
days, 97% ee, 15:1 anti:syn). However, the aldol product
was very difficult to separate from the large excess of
starting ketone, it was still present in the crude product.
R = i-Pr:
O
O
S
OH
L-proline
(10 mol%)
O
OTBS
1) TBSCl,
O
imid. (77%)
R
+
H
R
DMF
H2O
2) H2, Raney-
Ni (50%)
S
6a
1 equiv
1 equiv
8
R = i-Pr: 95% ee, yield 50%
8:1 anti:syn
R = Ph: 98% ee, yield 60%
Ref. 6
other polypropionate
building blocks
Scheme 2 Proline-catalyzed, water-assisted anti-aldol reactions allow access to polypropionate building blocks such as 8.
Synlett 2004, No. 11, 1891–1896 © Thieme Stuttgart · New York

1894
A. I. Nyberg et al.
LETTER
Table 3 Proline-Catalyzed Stoichiometric Aldol Reactions with 4-Thianone and Different Aldehydes with Added Watera
O
S
OH
O
S
O
S
OH
O
H2O
O
L-proline (10 mol%)
R
R
R
+
+
+
H
R
DMF, r.t.
S
1 mmol
1 mmol
6b
6a
7
Entry
1
R
H2O (mol%)b
0
6a+6b (%)c
ee (%)d
96
6a:6be
2:1
7 (%)c
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
< 1
i-Pr
11
2
i-Pr
100
36
96
10:1
3
i-Prf
100
49e
50g
65
87
12:1
4
i-Pr
300
95
15:1
9:1
5
i-Prh
300
83
6
i-Pr
500
42
96
10:1
7
Ph
0
3
65
>20:1
>20:1
>20:1
>20:1
>20:1
>20:1
20:1
8:1
8
Ph
100
22
86
9
Ph
300
37g
60g
41
93
10
11
12
13
14
15
Ph
500
98
Ph
1000
1500
500
92
Ph
32
95
p-NO2C6H4
p-CF3C6H4
m-BrC6H4
45g
76g
69g
>99
>98
>98
500
500
16:1
a Reaction times: 9 d (R = i-Pr) or 5 d (R = aromatic).
b Amount of added H2O.
c Conversion as determined by 1H NMR.
d The ee of the anti isomer, determined by chiral HPLC analysis (Chiralcel OD).
e Determined by 1H NMR.
f Reaction time: 13 d.
g Isolated yield.
h 200 mol% iso-butyraldehyde.
lidinone detected was directly related to the amount of vol%, corresponding to ca 200 mol%).1a,f However, high-
water added. After 6 days, the reaction mixtures contained er concentrations of water resulted in erosion of enantio-
8 mol% (25 mol% H2O), 5 mol% (50 mol% H2O), 5 mol% selectivity.1e Recently, the presence of water was found to
(100 mol% H2O) and 3 mol% (300 mol% H2O) of the ox- improve the enantioselectivity of the proline-catalyzed
azolidinone. As the reactions proceeded, the amount of Mannich reaction between 9-tosyl-3,4-dihydro-b-carbo-
the oxazolidinone decreased to almost undetectable levels line with ketones.17 In enantioselective organocatalytic
in all reactions examined.
Mukaiyama–Michael reactions, the addition of two
equivalents of water led to optimal reaction efficiency and
stereoselectivities.18
In principle, adding more proline should also increase the
rate of the reaction. Unfortunately, L-proline is only
slightly soluble in DMF and the conversion was not sig- Very recently, Yamamoto and co-workers19 discovered a
nificantly improved by doubling the amount of catalyst. proline-derived tetrazole that functions as an effective al-
With 10 mol% of proline, a conversion of 42% (de 14:1, dol catalyst in the presence of at least 100 mol% water. In
ee 92%) was obtained, whereas 20 mol% of proline gave their case, only hydrated aldehydes were reactive, and the
a conversion of 51% (de 10:1, ee 90%).16
role of water could readily be explained by this require-
ment.
Previously, proline-catalyzed aldol reactions have been
reported to tolerate a small amount of water in DMSO (<4
Synlett 2004, No. 11, 1891–1896 © Thieme Stuttgart · New York

LETTER
Water-Accelerated Proline-Catalyzed Aldol Reactions
1895
(5) (a) Seebach, D.; Boes, M.; Naef, R.; Schweizer, B. J. Am.
Chem. Soc. 1983, 105, 5390. (b) Orsini, F.; Pelizzoni, F.;
Forte, M.; Sisti, M.; Bombieri, G.; Benetollo, F. J.
In our stoichiometric reactions, no other by-products be-
sides a small amount of the aldol condensation product
were formed. In all reactions studied, at least 100 mol% of
water is necessary for optimal reaction rates and/or enan-
tioselectivities. Moreover, because the reactions can be
performed in concentrated solutions (1 M) with only 10
mol% proline, the reactions can readily be scaled up.
Heterocycl. Chem. 1989, 26, 837. (c) Rizzi, G. P. J. Org.
Chem. 1970, 35, 2069. (d) Orsini, F.; Pelizzoni, F.; Forte,
M.; Destro, R.; Gariboldi, P. Tetrahedron 1988, 44, 519.
(6) (a) Hayashi, T. Tetrahedron Lett. 1991, 32, 5369. (b)Ward,
D. E.; Sales, M.; Sasmal, P. K. Org. Lett. 2001, 3, 3671.
(c) Ward, D. E.; Sales, M.; Man, C. C.; Shen, J.; Sasmal, P.
K.; Guo, C. J. Org. Chem. 2002, 67, 1618. (d) Karisalmi,
K.; Rissanen, K.; Koskinen, A. M. P. Org. Biomol. Chem.
2003, 1, 3193.
(7) General Experimental Procedure. To ensure that all
reactions were performed under otherwise identical
conditions, all reactions were conducted in flame-dried
glassware under an argon atmosphere. DMF and DMSO
were dried by distillation over 4 Å molecular sieves, and
acetone was dried over anhyd CaSO4. To a mixture of 1 mL
anhyd DMF or DMSO, ketone donor (1 mmol) and L-proline
(10 mol% or 30 mol%) was added aldehyde (1 mmol) and
H2O (0–1500 mol%). The flask was capped and the reaction
mixture was stirred at r.t. under argon for 3–13 d. The
reaction was then quenched with H2O (10 mL, Table 3) or
with a sat. NH4Cl solution (10 mL, Table 1 and Table 2) and
then extracted with Et2O (3 × 10 mL). The organic layer was
dried (Na2SO4), filtered and concentrated. The pure aldol
products were obtained by flash column chromatography
[silica gel, pentane–Et2O (Table 1), hexane–EtOAc
(Table 2) or hexane–MTBE (Table 3)].
In conclusion, we have demonstrated that water has an ac-
celerating effect on proline-catalyzed ketone–aldehyde al-
dol reactions. This allows the use of stoichiometric
amounts of both ketone donor and the aldehyde acceptor,
thereby improving the overall economy of the process. In
addition, aldol reactions with an excess of ketone are also
improved by the addition of water. The excellent enantio-
and diastereoselectivities obtained in the 4-thianone aldol
reactions described herein allow us to rapidly construct
polypropionate building blocks6,20 by water-accelerated
proline catalysis.
Further studies to expand the scope of these water-assist-
ed, proline-catalyzed reactions are in progress and the re-
sults will be communicated in due course.
Acknowledgment
Financial support from the Academy of Finland (Grant No.
104397), Ministry of Education and the National Technology
Agency is gratefully acknowledged. The PMP group is the recipient
of the Outstanding Junior Research Group Award of Helsinki Uni-
versity of Technology. We thank Dr. Melanie Clarke for helpful
comments on the manuscript and Drs. Jari Koivisto and Juho Helaja
for NMR assistance.
(8) The racemic samples for the reactions with acetone, 4-tert-
butylcyclohexanone and 4-thianone (entries 13–15, Table 3)
as the donor were prepared by using racemic proline as
catalyst. For other reactions with 4-thianone, the racemic
samples were obtained using the procedure of Hayashi.6a
(9) Reaction conditions: 14.0 mmol acetone, 0.5 mmol
isobutyraldehyde, 30 mol% L-proline, 1 mL DMSO, r.t.,
50 h; 0 mol% H2O gave a conversion of 59% (ee 95%), 300
mol% H2O a conversion of 68% (ee 95%).
References
(10) Reaction conditions: 27 mmol acetone, 1 mmol p-nitro-
benzaldehyde, 20 mol% L-proline, 1 mL DMF, r.t., 24 h, 555
mol% (3 vol%) H2O; yield 90% (100% conversion), ee 76%.
Previously, List and Barbas had obtained this aldol product
in 68% yield and 76% ee in dry DMSO (see ref.1a,b).
(11) As an example, under the List–Barbas conditions, the
reactions with cyclohexanone and isobutyraldehyde require
five days for reasonable yields (ref.2d). It should be noted that
the stoichiometric reactions with cyclohexanone and
isobutyraldehyde were also accelerated (0% water: <30%
conversion after 4 d; 300 mol% water: ca. 60% conversion).
(12) (a) The assignment of the stereochemistry of 5a and 5b is
based on the 1H NMR coupling constants of the
(1) (a) Sakthivel, K.; Notz, W.; Bui, T.; Barbas, C. F. III J. Am.
Chem. Soc. 2001, 123, 5260. (b) List, B.; Lerner, R. A.;
Barbas, C. F. III J. Am. Chem. Soc. 2000, 122, 2395.
(c) List, B.; Pojarliev, P.; Castello, C. Org. Lett. 2001, 3,
573. (d) Notz, W.; List, B. J. Am. Chem. Soc. 2000, 122,
7386. (e) Córdova, A.; Notz, W.; Barbas, C. F. III Chem.
Commun. 2002, 3024. (f) For a review of proline-catalyzed
asymmetric reactions, see: List, B. Tetrahedron 2002, 58,
5573.
(2) Only one proline molecule is involved in the transition state:
(a) Hoang, L.; Bahmanyar, S.; Houk, K. N.; List, B. J. Am.
Chem. Soc. 2003, 125, 16. (b) For density functional studies
of the mechanism and quantum mechanical predictions of
the stereoselectivities, see: Arnó, M.; Domingo, L. R. Theor.
Chem. Acc. 2002, 108, 232. (c) Rankin, K. N.; Gauld, J. W.;
Boyd, R. J. J. Phys. Chem. 2002, 106, 5155.
(d) Bahmanyar, S.; Houk, K. N.; Martin, H. J.; List, B. J. Am.
Chem. Soc. 2003, 125, 2479.
(3) See for example: Zhong, G.; Lerner, R. A.; Barbas, C. F. III
Angew. Chem. Int. Ed. 1999, 38, 3738; and references
therein.
(4) (a) Guthrie, J. P.; Cossar, J.; Taylor, K. F. Can. J. Chem.
1984, 62, 1958. (b) Guthrie, J. P.; Wang, X.-P. Can. J.
Chem. 1992, 70, 1055.
COCHCH(OH)Ph proton of the product (400 MHz, CDCl3),
characteristic shifts of 5a: d = 4.90 [d, J = 9.9 Hz, 1 H,
COCHCH(OH)Ph], 2.61 [br ddd, J = 9.9, 6.0, 4.9 Hz, 1 H,
COCHCH(OH)Ph]; shifts of 5b: d = 5.19 [d, J = 4.9 Hz, 1
H, COCHCH(OH)Ph], 2.68 [br q, J = 5 Hz, 1 H,
COCHCH(OH)Ph)], indicating that the cyclohexanone ring
does not adopt the chair conformation, as would be expected
for the equatorial products. The major products also do not
match the data reported for the corresponding equatorial
product, see: Busch-Petersen, J.; Corey, E. J. Tetrahedron
Lett. 2000, 41, 6941. (b) In addition, the stereochemical
model presented in ref.2d clearly predicts an axial attack of
the carbonyl electrophile to the enamine intermediate. Only
minor amounts (<5% each) of the corresponding equatorial
isomers were formed.
Synlett 2004, No. 11, 1891–1896 © Thieme Stuttgart · New York

1896
A. I. Nyberg et al.
LETTER
oxazolidinones may simply reflect the fact that the
equilibrium constants for oxazolidinone formation are
higher with aldehydes than with ketones. The observation by
List et al. that cyclohexanone forms the oxazolidinone with
an equilibrium constant ca. five times that of acetone (0.68
vs. 0.12) fits very nicely with our observations that the effect
of added water is more dramatic with cyclic ketones such as
4-thianone. (b) List, B.; Hoang, L.; Martin, H. J. Proc. Natl.
Acad. Sci. U.S.A. 2004, 101, 5839.
(13) All new compounds gave satisfactory analytical and spectral
data. Selected characterization data: (3S,1¢S)-3-[(1¢-
hydroxy-2¢-methyl)propyl]-tetrahydrothiopyran-4-one
(Table 3, Entry 4): The ee was determined by HPLC
(Chiralpak OD column, hexane–i-PrOH, 98:2, flow rate 0.7
mL/min; tminor = 26.1 min; tmajor = 20.7 min, l = 254 nm). Rf
(1:1, hexane–MTBE) = 0.27. [a]D23 –72 (c 1.0, MeOH, 95%
ee). 1H NMR: d = 3.50 (t, J = 5.1 Hz, 1 H), 2.93–2.77 (m, 5
H), 2.70–2.65 (m, 2 H), 1.83–1.70 (m, 1 H), 0.91 (d, J = 6.8
Hz, 3 H), 0.87 (d, J = 6.8 Hz, 3 H). 13C NMR: d = 212.4,
76.1, 55.7, 44.9, 33.5, 30.9, 29.9, 19.9, 16.0. IR (film): 3488,
2960, 2874, 1704, 1426, 1274, 990 cm–1. HRMS (ESI): m/z
calcd for C9H16O2NaS: 211.0769. Found: 211.0762.
(3S,1¢R)-3-[(1¢-hydroxy-1¢-phenyl)methyl]-
(16) Reaction conditions: 1 mmol 4-thianone, 1 mmol iso-
butyraldehyde, 1 mL DMF, r.t., 10 d.
(17) (a) Itoh, T.; Yokoya, M.; Miyauchi, K.; Nagata, K.; Ohsawa,
A. Org. Lett. 2003, 5, 4301. (b) In contrast with these
results, related proline-catalyzed Mannich reactions with
glyoxylate-derived imine electrophiles were not improved
by the addition of water. See: Notz, W.; Tanaka, F.;
Watanabe, S.-i.; Chowdari, N. S.; Turner, J. M.;
Thayumanavan, R.; Barbas, C. F. III J. Org. Chem. 2003, 68,
9624.
(18) Brown, S. P.; Goodwin, N. C.; MacMillan, D. W. C. J. Am.
Chem. Soc. 2003, 125, 1192.
tetrahydrothiopyran-4-one (Table 3, Entry 10): The ee
was determined by HPLC (Chiralpak OD column, hexane–i-
PrOH, 90:10, flow rate 0.5 mL/min; tminor = 40.8 min;
t
major = 31.0 min, l = 254 nm). Rf (1:1, hexane–
MTBE) = 0.19. Mp 126–127 °C. [a]D23 –73 (c 1.1, MeOH,
98% ee). 1H NMR: d = 7.32–7.22 (m, 5 H), 4.90 (dd, J = 8.9,
3.0 Hz, 1 H), 3.34 (s, 1 H), 2.96–2.86 (m, 3 H), 2.78 (m, 1
H), 2.70 (m, 1 H), 2.49 (dd, J = 13.8, 9.9 Hz, 1 H), 2.44 (ddd,
J = 13.8, 5.2, 1.8 Hz, 1 H). 13C NMR: d = 211.8, 140.2,
128.7, 128.3, 126.9, 73.8, 59.7, 44.5, 32.9, 30.9. IR (KBr
pellet): n = 3421, 2914, 1690, 1426, 1275, 1051, 706 cm–1.
Anal. calcd for C12H14O2S: C, 64.8%; H, 6.4%. Found: C,
64.4%; H, 6.3%. HRMS (ESI): m/z calcd for C12H14O2NaS:
245.0612. Found: 245.0605.
(19) Torii, H.; Nakadai, M.; Ishihara, K.; Saito, S.; Yamamoto, H.
Angew. Chem. Int. Ed. 2004, 43, 1983.
(20) 4-Thianone is an excellent surrogate for 3-pentanone, which
is unreactive under proline catalysis (see ref.1a). For an
earlier example of the 4-thianone strategy in polypropionate
synthesis, see: Woodward, R. B.; Logusch, E.; Nambiar, K.
P.; Sakan, K.; Ward, D. E.; Au-Yeung, B.-W.; Balaram, P.;
Browne, L. J.; Card, P. J.; Chen, C. H.; Chênevert, R. B.;
Fliri, A.; Frobel, K.; Gais, H.-J.; Garrat, D. G.; Hayakawa,
K.; Heggie, W.; Hesson, D. P.; Hoppe, D.; Hoppe, I.; Hyatt,
J. A.; Ikeda, D.; Jacobi, P. A.; Kim, K. S.; Kobuke, Y.;
Kojima, K.; Krowicki, K.; Lee, V. J.; Leutert, T.;
Malchenko, S.; Martens, J.; Matthews, R. S.; Ong, B. S.;
Press, J. B.; Rajan Babu, T. V.; Rousseau, G.; Sauter, H. M.;
Suzuki, M.; Tatsuta, K.; Tolbert, L. M.; Truesdale, E. A.;
Uchida, I.; Ueda, Y.; Uyehara, T.; Vasella, A. T.;
Vladuchick, W. C.; Wade, P. A.; Williams, R. M.; Wong, H.
N.-C. J. Am. Chem. Soc. 1981, 103, 3210.
(14) (a) A slight erosion in the diastereopurity of the product
(12:1 to 8:1 anti:syn) was observed during the purification of
the product by preparative TLC. The spectroscopic data of 8
match those reported earlier by Paterson and co-workers.
See: Paterson, I.; Wallace, D. J.; Cowden, C. J. Synthesis
1998, 639. (b) [a]D23 +29 (c 1.0, CHCl3, 96% ee, 8:1
anti:syn), lit.: [a]D23 +12.3 (c 0.6, CHCl3).
(15) (a) L-Proline is not racemized in this process (see ref.5a).
Very recently, List and co-workers15b have reported that
ketones may also condense with proline to form the
oxazolidinones. Our inability to detect these ketone-derived
Synlett 2004, No. 11, 1891–1896 © Thieme Stuttgart · New York