Synthesis of enantiomerically enriched Baylis-Hillman alcohols from their acetates: Combination of kinetic resolution during the salt formation with (DHQD)2PHAL and following asymmetric induction during hydrolysis with NaHCO3 as a water surrogate

Article abstract of DOI:10.1016/S0040-4039(02)02274-8

Enantiomerically enriched Baylis-Hillman alcohols 2a-d (S) were prepared in 25-42% yields with optical purities of 54-92% ee by using the combined concept of kinetic resolution during the salt formation of racemic Baylis-Hillman acetates 1a-d with (DHQD)₂PHAL and the asymmetric induction during the S_N2′ type reaction with sodium bicarbonate as a water surrogate.

Full text of DOI:10.1016/S0040-4039(02)02274-8

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 9141–9146
Synthesis of enantiomerically enriched Baylis–Hillman alcohols
from their acetates: combination of kinetic resolution during the
salt formation with (DHQD)2PHAL and following asymmetric
induction during hydrolysis with NaHCO3 as a water surrogate
Jae Nyoung Kim,* Hong Jung Lee and Ji Hyeon Gong
Department of Chemistry and Institute of Basic Science, Chonnam National University, Kwangju 500-757, South Korea
Received 10 September 2002; accepted 4 October 2002
Abstract—Enantiomerically enriched Baylis–Hillman alcohols 2a–d (S) were prepared in 25–42% yields with optical purities of
54–92% ee by using the combined concept of kinetic resolution during the salt formation of racemic Baylis–Hillman acetates 1a–d
with (DHQD)2PHAL and the asymmetric induction during the SN2% type reaction with sodium bicarbonate as a water surrogate.
© 2002 Elsevier Science Ltd. All rights reserved.
The Baylis–Hillman reaction allows the direct prepara-
tion of a-methylene-b-hydroxycarbonyl compounds by
base-catalyzed reaction of a,b-unsaturated carbonyl
compounds with aldehydes.1 Much effort has been
devoted to the synthesis of the Baylis–Hillman adducts
in an enantioselective or diastereoselective manner.2–6
By using the modified acrylic acid derivatives with
chiral alcohols or chiral amines, limited success has
for p-methoxyphenol and is given as an example in this
paper. More recently, Basavaiah and co-workers have
published the asymmetric synthesis of the propargyl
ether of the Baylis–Hillman adduct via the deracemiza-
tion process.7 In situ generation of the quinidinium salt
followed by the nucleophilic substitution reaction with
propargyl alcohol gave the product in 32–47% isolated
yield with 25–40% enantiomeric purity.
been achieved.2
A remarkable improvement was
reported by Leahy and co-workers by using Oppolzer’s
chiral sultam-derived acrylamide in the Baylis–Hillman
reaction.2a The use of chiral catalyst, tertiary amines3,4
or tertiary phosphines,5 has also been studied with low
% ee. Recently, Hatakeyama and co-workers have
reported an unprecedented level of asymmetric induc-
tion using the cinchonidine-derived chiral amine cata-
lyst, (3R,8R,9S)-10,11-dihydro-3,9-epoxy-6%-hydroxy-
cinchonane.4 They obtained Baylis–Hillman adducts in
31–58% yields with 91–99% ee by using this catalyst.
Baylis–Hillman acetates or bromides can form the
quarternary ammonium salt, which can be used as an
electrophile toward many nucleophiles.7,8 The forma-
tion of ammonium salt is quantitative for DABCO
(1,4-diazabicyclo[2.2.2]octane) in a short time.8,9 The
reaction of the ammonium salt and nucleophile gave
the allylic substitution product in good yield.8 We
intended to study a deracemization type process with
oxygen and nitrogen nucleophiles in the presence of a
chiral cinchona alkaloid in this context in order to
obtain better results than in Basavaiah’s paper.7
Trost et al. have reported on a deracemization process
of the Baylis–Hillman carbonate with phenols in the
presence of a palladium catalyst and a chiral phosphine
ligand.6 The best result was up to 92% ee (yield=69%)
We initially examined the reaction of p-methoxyphenol
and the Baylis–Hillman acetate 1b in the presence of
cinchonidine, (DHQD)2PHAL and (DHQ)2PHAL in
aqueous THF (see Table 1). To our disappointment,
however, the chemical yield of the product 3a was low
(11–48%). Moreover, the % ee of the product did not
exceed 38% (entries 1–3 in Table 1). Thus, we next
examined other nucleophiles such as p-toluenesulfon-
amide and methanesulfonamide under similar reaction
Keywords: Baylis–Hillman alcohol; kinetic resolution; cinchona alka-
loid; water surrogate; asymmetric induction.
* Corresponding author. Fax: +82-62-5303389; e-mail: kimjn@
chonnam.chonnam.ac.kr
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02274-8

9142
J. N. Kim et al. / Tetrahedron Letters 43 (2002) 9141–9146
Table 1. Effect of chiral amine and nucleophile on the deracemization of 1b
formation in the cases of cinchona alkaloids,9 the use of
conditions (entries 4–7 in Table 1). We found that the
% ee reached up to 70% when we used sulfonamide as
the nucleophile. However, the chemical yields of the
products were quite low. The low yield might be due to
the incomplete and slow salt formation9 and low reac-
tivity of sulfonamide in the reaction medium. Recently,
we have reported the facile synthesis of Baylis–Hillman
adducts of N-tosylimines from the Baylis–Hillman
acetates of benzaldehydes via the achiral version using
DABCO.8c The salt formation was complete when we
used DABCO as previously mentioned and the follow-
ing substitution reaction with p-toluenesulfonamide
occurred in good yield. Due to incomplete and slow salt
larger amounts of chiral amine did not improve the
yield. The use of aqueous DMF in order to quicken the
reaction and improve the yields failed due to the fast
formation of the rearranged acetate. The use of triethyl-
amine improves the chemical yields; however, as shown
in entry 6, the % ee was dramatically decreased.7 Low
% ee (4% yield, 8% ee, not shown in Table 1) with
N-methyl p-toluenesulfonamide as the nucleophile
implies that a certain type of interaction such as hydro-
gen bonding (between the incoming nucleophile and the
ammonium salt) might be involved for the asymmetric
induction.

J. N. Kim et al. / Tetrahedron Letters 43 (2002) 9141–9146
9143
The unfavorable data for the deracemization type pro-
cess made us try another approach for the synthesis of
AQN and (DHQD)2PYR as the chiral amine catalyst.
As a water surrogate, NaHCO3, KHCO3, Na2CO3 and
Cs2CO3 were examined.
a
Baylis–Hillman adduct in an enantiomerically
enriched form. Kinetic resolution of the Baylis–Hillman
adduct with acetic anhydride in the presence of a chiral
amine catalyst was examined.10,11 The reaction of the
Baylis–Hillman adduct 2b (racemic) in the presence of
acetic anhydride, triethylamine and (DHQD)2PHAL
was quenched at around half conversion. However,
HPLC analysis of the remaining alcohol (41%) or the
generated acetate (45%) showed that they are com-
pletely racemic mixtures. Without triethylamine, acetyl-
ation reaction was quite slow.
It is difficult to say from the results whether the first
salt-formation step or the second SN2%-type stage is the
reason for the asymmetric induction. There might be a
combined interaction. Thus, we examined the salt for-
mation reaction between the Baylis–Hillman acetate 1b
(racemic) and (DHQD)2PHAL or (DHQ)2PHAL. The
remaining Baylis–Hillman acetate was separated and its
% ee was examined after the reaction at room tempera-
ture for 2 days with chiral cinchona alkaloid (0.2
equiv.). The Baylis–Hillman acetates 1b were isolated in
62 and 72%, respectively. Their % ee was 27 and 6% ee,
respectively. The results mean that the difference in
reactivity for the salt formation (kinetic resolution) is
present between the two enantiomeric acetates (R) and
(S). Moreover, (DHQD)2PHAL is more effective than
(DHQ)2PHAL.
In another experiment, we examined the deracemization
type process with triphenylsilanol or sodium bicarbon-
ate as a water surrogate in the presence of commercially
available chiral amine catalyst (Scheme 1). As Trost et
al. have reported in a similar experiment, triphenylsi-
lanol was ineffective as a water surrogate.12 When we
used sodium bicarbonate interesting results were
observed. The reaction of the Baylis–Hillman acetate 1a
in aqueous THF in the presence of sodium bicarbonate
and (DHQD)2PHAL (0.2 equiv.) was examined (entry 1
in Table 2). At the point of half-conversion (around 3.5
days), the separated alcohol (42%) showed 84% ee,
while the remaining acetate (24%) showed 53% ee.13
Appreciable amounts of rearranged alcohol and unre-
acted ammonium salt were observed as the side prod-
ucts. As shown in Table 2, the use of other chiral amine
catalysts and water surrogates showed similar results.
We used (DHQD)2PHAL, (DHQ)2PHAL, (DHQD)2-
The structure of the generated ammonium salt from
(DHQD)2PHAL and 1b would be E-form as previously
reported by us and Basavaiah et al.7,8 As mentioned in
the above paragraph, maximum % ee of the recovered
acetate would not exceed 27% if there were only kinetic
resolution factor. Thus, the observed data (vide supra,
entry 1 in Table 2, 84% ee for the Baylis–Hillman
alcohol and 53% ee for the recovered acetate) must
result from the combined effects of kinetic resolution
during salt formation and the asymmetric induction
process during the hydrolysis of the ammonium salt
Scheme 1.
Table 2. Effects of chiral catalyst and water surrogate on the deracemization of 1a
Entry
Catalyst
Water surrogate
Time/temp. (°C)
2a
1a
1
2
3
4
5
6
7
8
9
(DHQD)2PHAL (0.2 equiv.)
(DHQD)2PHAL (0.2 equiv.)
(DHQD)2PHAL (0.2 equiv.)
(DHQD)2PHAL (0.2 equiv.)
(DHQD)2PHAL (0.2 equiv.)
(DHQD)2AQN (0.2 equiv.)
(DHQD)2PYR (0.2 equiv.)
(DHQD)2PHAL (0.2 equiv.)
(DHQ)2PHAL (0.2 equiv.)
NaHCO3 (3.0 equiv.)
NaHCO3 (3.0 equiv.)
KHCO3 (2.0 equiv.)
Na2CO3 (3.0 equiv.)
Cs2CO3 (3.0 equiv.)
NaHCO3 (3.0 equiv.)
NaHCO3 (3.0 equiv.)
NaHCO3 (2.0 equiv.)
NaHCO3 (2.0 equiv.)
3.5 days/40–50
16 days/rt
42%a, 84% ee (S)
60%, 81% ee (S)
27%, 79% ee (S)
33%, 75% ee (S)
25%, 71% ee (S)
39%, 72% ee (S)
15%, 76% ee (S)
31%, 79% ee (S)
8%b, 53% ee (R)
24%, 53% ee (R)
Not isolated
3.5 days/40–50
3.5 days/40–50
3.5 days/40–50
3.5 days/40–50
3.5 days/40–50
40 h/40–50
17%, 56% ee (R)
35%, 52% ee (R)
43%, 66% ee (R)
24%, 53% ee (R)
41%, 39% ee (R)
52%, 22% ee (R)
51%, 10% ee (S)
40 h/40–50
a [h]2D2=+85.5° (c 1.11, MeOH).
b [h]2D2=−56.9° (c 1.11, MeOH).

9144
J. N. Kim et al. / Tetrahedron Letters 43 (2002) 9141–9146
with NaHCO3 by the chiral ammonium moiety of the
salt. Asymmetric induction in the nucleophilic substitu-
tion stage with sodium bicarbonate could be another
reason for the 84% ee. In fact, when the alcohol was
isolated over 50%, the % ee was not reduced signifi-
cantly (entry 2 in Table 2, 60% yield, 81% ee). Thus,
without affecting the % ee, the desired Baylis–Hillman
alcohol could be theoretically obtained in over 50%
yield.
gave the improved % ee (up to 92% ee). However, low
yield (25%) and long reaction time diminished the
value. Other Baylis–Hillman acetates 1b–d showed sim-
ilar results. When we used the aliphatic Baylis–Hillman
acetate derived from hexanal, elimination of acetic acid
followed by subsequent Diels–Alder reaction was the
principle pathway.8c
The absolute configuration of the obtained Baylis–Hill-
man alcohol 2a was deduced from the optical rotation
by comparison with that reported.15 The optical rota-
tion of the (R)-form of the Baylis–Hillman adduct 2a
was reported as −111.1°. In our case, we obtained
+85.5° for the isolated alcohol (entry 1 in Table 3, 42%,
84% ee). The optical rotation value can be estimated as
77% ee based on the reported data, which was in
agreement with the HPLC data (S-form is major, 84%
ee). When we used (DHQ)2PHAL (entry 9 in Table 2)
the optical rotation of the alcohol was −56.9°, indicat-
ing the major isomer is the (R)-form as expected.
The reaction of the Baylis–Hillman acetate and
(DHQD)2PHAL produced a quarternary ammonium
salt with E-configuration as previously reported.7,8 Salt
formation of the (S) isomer occurs faster than the (R)
form (vide infra). Sodium bicarbonate then attacks the
electrophilic b-carbon toward the si face and subse-
quently the (S) form of the Baylis–Hillman alcohol was
obtained as a major product (Fig. 1). Hydrogen bond-
ing between the hydrogen atom of the bicarbonate and
the oxygen atom of the (DHQD)2PHAL moiety would
cause the approach of the bicarbonate ion toward the si
face. To obtain greater insight on the hydrogen bond-
ing effect, we examined the reaction with sodium car-
bonate, potassium bicarbonate and cesium carbonate.
However, we could not find any significant changes in
chemical yields and in % ees (Table 2). The reaction in
THF, instead of using aqueous THF in order to maxi-
mize the hydrogen bonding, failed due to the incom-
plete salt formation in dry THF and subsequent low
yield of the alcohol product.
As stated above, we presumed the reaction mechanism
as follows. Salt formation between the acetate and
chiral amine, although incomplete, must be the first
step. Nucleophilic attack of sodium bicarbonate at the
benzylic position occurred in a stereoselective manner.
Direct hydrolysis of the acetate by sodium bicarbonate
could not be excluded completely, which could diminish
the optical purity of the products. However, the possi-
bility of direct hydrolysis with sodium bicarbonate was
excluded experimentally. As a blank test, direct hydroly-
sis of the acetate group into alcohol in the absence of
(DHQD)2PHAL was never observed (NaHCO3, aq.
THF, 40–50°C, 3 days). Thus, our conclusion is as
follows: The (S) form of acetate reacts faster than the
(R) form with (DHQD)2PHAL. The generated ammo-
nium salt reacts with sodium bicarbonate preferentially
at the si face to give the (S) form of Baylis–Hillman
alcohol as the major product. Most of the (R) form
acetate remained intact.
The distance between the oxygen atom of the methoxy
group at the quinoline moiety and the carbon atom at
the benzylic position of the ammonium salt is estimated
,
to be about 7.38 A in the energy-minimized conforma-
tion (MM2 calculation). Considering the size of the
,
bicarbonate anion (about 3.24 A by MM2) and normal
H-bonding distance (about 1.79 A), the distance
14
,
between the oxygen atom of bicarbonate and the reac-
tion center (benzylic carbon) is calculated as 2.35 A. In
,
other words, the bicarbonate ion fits well into the
roomy pocket of the ammonium salt as shown in Fig.
1. The counter cation, sodium, potassium or cesium,
did not affect the results significantly.12 The effect of
the anion, bicarbonate or carbonate, was also not
observed.12
In summary, we could obtain the enantiomerically
enriched Baylis–Hillman alcohols in 25–42% yield with
optical purity of 54–92% ee by using the combined
concept of kinetic resolution during salt formation and
the asymmetric induction during the SN2% type reaction
with sodium bicarbonate as a water surrogate. Such a
combined concept should inspire new avenues for other
asymmetric synthesis.
We then examined the reaction of some Baylis–Hillman
acetate 1a–d as shown in Table 3. As compared in
entries 1 and 2, the reaction of 1a at room temperature
Figure 1. Proposed TS model for the (S) selectivity.

J. N. Kim et al. / Tetrahedron Letters 43 (2002) 9141–9146
9145
Table 3. Deracemization of the Baylis–Hillman acetates
Acknowledgements
5. (a) Li, W.; Zhang, Z.; Xiao, D.; Zhang, X. J. Org. Chem.
2000, 65, 3489; (b) Hayase, H.; Shibata, T.; Soai, K.;
Wakatsuki, Y. Chem. Commun. 1998, 00, 1271.
6. Trost, B. M.; Tsui, H.-C.; Toste, F. D. J. Am. Chem. Soc.
2000, 122, 3534.
This work was supported by a Korea Research Foun-
dation Grant (KRF-2002-015-CP0215).
7. Basavaiah, D.; Kumaragurubaran, N.; Sharada, D. S.;
Reddy, R. M. Tetrahedron 2001, 57, 8167.
References
8. (a) Chung, Y. M.; Gong, J. H.; Kim, T. H.; Kim, J. N.
Tetrahedron Lett. 2001, 42, 9023; (b) Basavaiah, D.;
Kumaragurubaran, N. Tetrahedron Lett. 2001, 42, 477;
(c) Kim, J. N.; Lee, H. J.; Lee, K. Y.; Gong, J. H. Synlett
2002, 173 and references cited therein.
1. (a) Basavaiah, D.; Rao, P. D.; Hyma, R. S. Tetrahedron
1996, 52, 8001; (b) Drewes, S. E.; Roos, G. H. P.
Tetrahedron 1988, 44, 4653; (c) Ciganek, E. Organic
Reactions; John Wiley & Sons: New York, 1997; Vol. 51,
pp. 201–350; (d) Langer, P. Angew. Chem., Int. Ed. 2000,
39, 3049; (e) Kim, J. N.; Lee, K. Y. Curr. Org. Chem.
2002, 6, 627.
2. (a) Brzezinski, L. J.; Rafel, S.; Leahy, J. W. J. Am. Chem.
Soc. 1997, 119, 4317; (b) Drewes, S. E.; Emslie, N. D.;
Khan, A. A. Synth. Commun. 1993, 23, 1215; (c) Isaacs,
N. S.; Gilbert, A.; Heritage, T. W. Tetrahedron: Asymme-
try 1991, 2, 969; (d) Yang, K.-S.; Chen, K. Org. Lett.
2000, 2, 729; (e) Bauer, T.; Tarasiuk, J. Tetrahedron:
Asymmetry 2001, 12, 1741.
3. (a) Oishi, T.; Oguri, H.; Hirama, M. Tetrahedron: Asym-
metry 1995, 6, 1241; (b) Barrett, A. G. M.; Cook, A. S.;
Kamimura, A. Chem. Commun. 1998, 00, 2533; (c)
Marko, I. E.; Giles, P. R.; Hindley, N. J. Tetrahedron
1997, 53, 1015.
4. (a) Iwabuchi, Y.; Nakatani, M.; Yokoyama, N.;
Hatakeyama, S. J. Am. Chem. Soc. 1999, 121, 10219; (b)
Iwabuchi, Y.; Furukawa, M.; Esumi, T.; Hatakeyama, S.
Chem. Commun. 2001, 2030.
9. Salt formation with cinchona alkaloids was incomplete in
all cases.
10. For the enantioselective acylation with kinetic resolution
strategy, see: (a) Vedejs, E.; Chen, X. J. Am. Chem. Soc.
1996, 118, 1809; (b) Vedejs, E.; Chen, X. J. Am. Chem.
Soc. 1997, 119, 2584; (c) Jarvo, E. R.; Copeland, G. T.;
Papaioannou, N.; Bonitatebus, P. J., Jr.; Miller, S. J. J.
Am. Chem. Soc. 1999, 121, 11638; (d) Vedejs, E.; Dau-
gulis, O.; Diver, S. T. J. Org. Chem. 1996, 61, 430; (e)
Taylor, S. J.; Morken, J. P. Science 1998, 280, 267; (f)
Ruble, J. C.; Fu, G. C. J. Org. Chem. 1996, 61, 7230; (g)
Copeland, G. T.; Jarvo, E. R.; Miller, S. J. J. Org. Chem.
1998, 63, 6784; (h) Kawabata, T.; Nagato, M.; Takasu,
K.; Fuji, K. J. Am. Chem. Soc. 1997, 119, 3169; (i)
Spivey, A. C.; Fekner, T.; Spey, S. E.; Adams, H. J. Org.
Chem. 1999, 64, 9430.
11. For the symmetric acetylation of the Baylis–Hillman
adducts with vinyl acetate catalyzed by Pseudomonas
AK,11a selective hydrolysis of the Baylis–Hillman acetates

9146
J. N. Kim et al. / Tetrahedron Letters 43 (2002) 9141–9146
by pig-liver esterase,11b and other kinetic resolution tech-
by HPLC analysis (DAICEL CHIRALCEL OD, hexane/
EtOH=19:1, 1.0 mL/min). The retention time (tR) of the
alcohol was 9.31 min for the (R) enantiomer and 8.66
min for the (S) isomer (vide infra). Enantiomeric excess
of 2a was calculated as 84% from the two peak areas. The
optical rotation of the Baylis–Hillman alcohol 2a was
determined as [h]2D2=+85.5° (c 1.11, MeOH). We pre-
sumed the absolute configuration of the major isomer
was (S) from the sign of specific rotation based on the
reported data.15 The value of optical rotation corre-
sponds to 77% ee based on the reported data of Drewes
([h]2D6=−111.1° (c 1.11, MeOH) for the (R) isomer),15
which was in agreement with our HPLC data (84% ee).
14. Smith, M. B.; March, J. March’s Advance Organic Chem-
istry; John Wiley & Sons, 2001; p. 99.
nique,11c see: (a) Burgess, K.; Jennings, L. D. J. Org.
Chem. 1990, 55, 1138; (b) Basavaiah, D.; Rao, P. D.
Synth. Commun. 1994, 24, 917; (c) Takagi, M.;
Yamamoto, K. Tetrahedron 1991, 47, 8869.
12. Trost, B. M.; McEachern, E. J. J. Am. Chem. Soc. 1999,
121, 8469.
13. Typical procedure: To a stirred solution of the Baylis–
Hillman acetate 1a (234 mg, 1.0 mmol) and
(DHQD)2PHAL (156 mg, 0.2 mmol) in aqueous THF (10
mL, H2O/THF=1:2) was added sodium bicarbonate (252
mg, 3.0 mmol). The reaction mixture was stirred at
40–50°C for 3.5 days. After usual workup and column
chromatographic purification (hexane/ether/dichloro-
methane=80:10:10) the Baylis–Hillman alcohol (2a, 81
mg, 42%) and the remaining acetate (1a, 56 mg, 24%)
were isolated. Enantiomeric excess (ee) was determined
15. Drewes, S. E.; Emslie, N. D.; Field, J. S.; Khan, A. A.;
Ramesar, N. Tetrahedron: Asymmetry 1992, 3, 255.