Potential and Limitations of Palladium-Cinchona Catalyst for the Enantioselective Hydrogenation of a Hydroxymethylpyrone

Article abstract of DOI:10.1006/jcat.2000.2890

The palladium-catalyzed enantioselective hydrogenation of 4-hydroxy-6-methyl-2-pyrone afforded up to 85% excess to the (S)-enantiomer of the corresponding 5,6-dihydropyrone, under very mild conditions (1 bar, room temperature). This is the highest enantioselectivity achieved so far with chirally modified Pd, demonstrating the potential of this catalyst in the enantioselective hydrogenation of unsaturated compounds. A complicating feature of the reaction is the limited stability of cinchonidine under reaction conditions, which results in a decline of the initial enantiomeric excess (ee) with reaction time. Continuous feeding of a minute amount of cinchonidine during reaction allows maintenance of the high initial ee with an overall substrate/modifier molar ratio of ca. 20.

Full text of DOI:10.1006/jcat.2000.2890

Journal of Catalysis 193, 1–4 (2000)
doi:10.1006/jcat.2000.2890, available online at http://www.idealibrary.com on
PRIORITY COMMUNICATION
Potential and Limitations of Palladium–Cinchona Catalyst for the
Enantioselective Hydrogenation of a Hydroxymethylpyrone
W.-R. Huck, T. Mallat, and A. Baiker1
Laboratory of Technical Chemistry, Swiss Federal Institute of Technology, ETH-Zentrum, CH-8092 Zu¨rich, Switzerland
E-mail: baiker@tech.chem.ethz.ch
Received March 17, 2000; revised April 3, 2000; accepted April 3, 2000
tetrahydro-derivative 3 with only 17% ee. Among the het-
erogeneous metal hydrogenation catalysts only Pd affords
the semihydrogenation of 1 to the dihydro-derivative 2 (6).
Ni forms predominantly the cis 3, while the saturated lac-
tone 4 is the major product on Pt due to its high activity in
hydrogenolysis of the C–O(H) bond.
Intrigued by the pharmaceutical importance of chiral 2-
pyrone derivatives and the shortcomings of homogeneous
catalysts in the enantioselective hydrogenation of 1 to 2,
we have studied this reaction over Pd, Pt, and Ni in the
presence of cinchonidine (CD) as chiral modifier.
The palladium-catalyzed enantioselective hydrogenation of 4-
hydroxy-6-methyl-2-pyrone afforded up to 85% excess to the (S)-
enantiomer of the corresponding 5,6-dihydropyrone, under very
mild conditions (1 bar, room temperature). This is the highest enan-
tioselectivity achieved so far with chirally modified Pd, demon-
strating the potential of this catalyst in the enantioselective hy-
drogenation of unsaturated compounds. A complicating feature
of the reaction is the limited stability of cinchonidine under re-
action conditions, which results in a decline of the initial enan-
tiomeric excess (ee) with reaction time. Continuous feeding of a
minute amount ofcinchonidine duringreaction allowsmaintenance
of the high initial ee with an overall substrate/modifier molar ratio of
c
ca. 20.
2000 Academic Press
2. EXPERIMENTAL
Key Words: enantioselective; regioselective; asymmetric; hy-
drogenation; cinchonidine; palladium; platinum; nickel; 2-pyrone
derivative; 4-hydroxy-6-methyl-2-pyrone.
The 5-wt% Pd/alumina (Engelhard 40692, metal disper-
sion 0.21, determined by TEM), 5-wt% Pt/alumina (Engel-
hard 4759, metal dispersion 0.27, determined by TEM), and
Raney-Ni (Fluka, 83440) catalysts were used as received. A
5-wt% Pd/titania (metal dispersion 0.18, determined by H2
chemisorption) was prepared by neutralizing an aqueous
H2PdCl4 solution with NaCO3 at room temperature, in the
presence of TiO2 (P25, Degussa, 55 m2/g). The 4-hydroxy-
6-methyl-2-pyrone (1, Fluka 98% ) was purified before use
by column chromatography (silica gel, dry hexane/ethyl ac-
etate 1 : 1) and recrystallization. All solvents were distilled
before use. CD (Fluka) was used as received.
The reactions at 1 bar were carried out in a magnetically
stirred 100-ml glass reactor. According to a standard pro-
cedure, 40 mg catalyst in 20 ml solvent was pretreated with
H2 for 5 min, at 25–30 C. Then the appropriate amount of
modifier and 300 mg 1 were added, and the stirring was
started. A 100-ml stainless steel autoclave equipped with a
50-ml glass liner and a PTFE cap and stirrer was used for
high-pressure experiments.
1. INTRODUCTION
The structure of 2-pyrone and its hydrogenation products
is present in many natural compounds. The partial hydro-
genation products can be used as chiral intermediates, for
example, in the synthesis of tetrahydrolipstatin, a potent an-
tiobesity drug, and that of an HIV protease inhibitor (1–3).
For this reason the enantioselective hydrogenation of sub-
stituted 2-pyrones with homogeneous transition metal com-
plexes has recently become a subject of great interest. A
chiral Ru catalyst afforded excellent ee (up to 98% ) and
good chemoselectivities to 5,6-dihydropyrones, but only a
mixture of trans and cis tetrahydropyrones formed when
the substrate was not alkylated in C-3 position (4, 5).
There is only one report that mentions (without de-
tails) the heterogeneous enantioselective hydrogenation of
a substituted 2-pyrone (3). Hydrogenation of 1 (Scheme 1)
with the Raney Ni–tartaric acid–NaBr system yielded the
Conversion, yield, and selectivity were determined by
an HP 6890 gas chromatograph (HP-5 column). The enan-
tiomers of 2 were separated after methylation with a cy-
closil B column (J&W). Derivatization was carried out in
1 To whom correspondence should be addressed.
1
0021-9517/00 $35.00
c
Copyright
2000 by Academic Press
All rights of reproduction in any form reserved.

2
HUCK, MALLAT, AND BAIKER
SCHEME 1
3 ml methanol with 100 mg trimethylorthoformate in the
presence of acidic ion exchange resin (Diaion RCP1 60H)
and ca. 0.1 mmol hydrogenation product (50 C, 15 h). The
products 2, 3, and 4 (Scheme 1), after isolation by flash
chromatography (silica gel, hexane/ethyl acetate 1 : 1) were
identified by NMR and GC-MS analysis and by optical
rotation.
FIG. 1. Conversion dependence of enantiomeric excess (ee) in the
hydrogenation of 1 in acetonitrile (standard conditions, 5 wt% Pd/titania,
3 mg CD, reaction time 1–24 h).
3. RESULTS
influence of catalyst support (titania or alumina) on the rate
and chemoselectivity was small. In the following part, the
study of the enantioselective transformation of 1 to 2 was
carried out with supported Pd catalysts.
Hydrogenation of 1 with Pd/alumina, Raney Ni, and
Pt/alumina was complete within 1 h under ambient con-
ditions, leading predominantly to 2, 3, and 4, respectively
(Scheme 1). The activity and selectivity of Raney Ni was
barely influenced by the presence of CD, but Pt was com-
pletely poisoned by the addition of only 3 mg alkaloid
(Table 1).
Pd preserved its good chemo- and regioselectivity to the
dihydro-derivative 2 in the presence of CD, though the re-
action rate decreased remarkably. For example, addition of
3 mg CD to 40 mg Pd/alumina increased the reaction time
by a factor of 25. Over 95% selectivity to 2 was achieved
at around 80% conversion of 1 in various solvents, except
acetonitrile (Table 1). The chemoselectivity successively
decreased by 10–25% at higher conversion and especially
under pressure (5–50 bar). The major by-product was the
tetrahydroderivative 3, except in acetic acid, which solvent
promoted the hydrogenolysis of the C–O(H) bond (7). The
Pd/alumina or Pd/titania with CD as chiral modifier af-
forded 20–50% ee to (S)-2 in various solvents, such as
acetic acid, 2-propanol, 3-pentanone, dimethylformamide,
and water. Replacing CD with the diastereomer cincho-
nine yielded the opposite enantiomer (R) in excess, as ex-
pected. In most solvents the ee decreased slowly with con-
version when only 1–3 mg CD was applied, corresponding
to a modifier/Pds molar ratio of ca. 1–3. In acetonitrile
the initial ee was higher, but the loss of enantioselectiv-
ity with increasing conversion was strikingly fast (Fig. 1).
The differential curve shows that the actual (incremental)
ee dropped to zero above 5% conversion of 1 (4 h reaction
time). When the reaction was repeated under similar con-
ditions but Pd/titania was prehydrogenated for 4 h in the
presence of CD, no ee was observed. NMR experiments
TABLE 1
Influence of Solvent and Pressure on the Chemioselectivity of Pd, Pt, and Ni Catalysts
(40 mg Catalyst, 20 ml Solvent, and 300 mg 1, rt)
Selectivity to
CD
(mg)
Pressure
(bar)
Time
(h)
Conv.
(% )
2
(% )
3
(% )
4
(% )
Catalyst
Solvent
Raney Ni
Pt/Al2O3
Pt/Al2O3
Pd/Al2O3
Pd/Al2O3
Pd/TiO2
Pd/TiO2
isopropanol
isopropanol
isopropanol
acetic acid
acetic acid
isopropanol
acetonitrile
10
0
3
10
10
3
1
1
1
1
30
1
1
1
10
24
7
90
100
0
78
80
75
77
0
2
100
0
—
2
8
3
0
98
—
2
7
0
—
96
83
97
75
24
24
3
1
25
0

ENANTIOSELECTIVE HYDROGENATION OF HYDROXYMETHYLPYRONE
3
10 mg) but an increased amount of dosed CD above 0.5
mg/h.
4. DISCUSSION
The chemical behaviour of 2-pyrones can be described
by the reactivity of an endocyclic ester possessing a con-
jugated double bond system, or it may be attributed to a
pseudo-aromatic character. Indications to the former ap-
proach are the facile hydrogenation of 2-pyrones under
ambient conditions, the ring opening by the attack of hy-
drides, and the missing NMR evidence for aromaticity (9).
On the other hand, the indications for partial aromaticity
are that hydrogenation leads to tetrahydro-pyrones except
SCHEME 2
indicated that Pd catalysed saturation of the vinyl group
and the hetereoaromatic ring of CD, while hydrogenation
of the unfunctionalized aromatic ring of CD at room tem-
perature and atmospheric pressure was minor even after
72 h (Scheme 2).
When higher amounts (10–20 mg) of modifier were used,
the loss of ee with increasing conversion ceased or slowed
down considerably in all solvents investigated, but hydro-
genation of 1 was also sluggish. For example, in an experi-
ment, 20 mg CD and 40 mg Pd/titania were used under stan-
dard conditions in acetonitrile containing 0.25 vol% water
(substrate/modifier molar ratio 35). After 6 h the conver-
sion was only 4% , though the ee to the (S)-product reached
77% . The highest enantioselectivity, of 85% , was achieved
under similar conditions, but the reaction was carried out
at 32 C with 22 mg CD-hydrochloride as chiral modifier. In
this case 4 h was necessary to get 2% conversion.
To overcome these difficulties, the reaction was started
with small amounts of modifier and some CD was added
during reaction (Fig. 2). This strategy has been applied suc-
cessfully for optimizing the concentration of CD on the
Pt surface and thus the stereochemical outcome of ethyl
pyruvate hydrogenation (8). As illustrated in Fig. 2, dosing
of at least 0.5 mg/h alkaloid in 1 ml solvent stabilised the
initial high ee in acetonitrile. On the other hand, no en-
hancement in ee could be achieved when the experiments
were repeated with the same initial amount of CD (2.5–
==
when the C C double bond in the intermediate is stabilised
by two alkyl groups at the C-3 and C-4 positions, or a hy-
droxy or alkoxy group at the 4 position (10, 11). Also the-
oretical calculations showed a small amount of aromatic-
ity (9), and methylation of the carbonyl group to form the
aromatic pyryllium derivative cannot be explained with-
out assuming a partially aromatic character (12). Accord-
ingly, in many publications this structure is described as
pseudo-aromatic.
Independent of considering 1 as an endocyclic enol es-
ter or a pseudo-aromatic compound, the 85% ee achieved
with CD-hydrochloride is the highest enantioselectivity re-
ported so far for chirally modified Pd. Limitations arise
from the low reaction rate and moderate stability of the
modifier under reaction conditions. A possible solution to
maintain the initial enantioselectivity up to high conver-
sions is the continuous feeding of minute amounts of mod-
ifier. This strategy is demonstrated in Fig. 2. Further im-
provement by optimization of reaction conditions aiming at
increasing reaction rate and optical yield (conversion ee)
is possible. As an example, increasing the amount of cata-
lyst from 20 to 200 mg (at constant Pd/modifier ratio) af-
forded a conversion of 22% in 4 h, and an ee of 82% . In the
enantioselective hydrogenation of -functionalized olefins,
Pd modified by cinchona- or vinca-type alkaloids afforded
52–72% ee (13–16). Note that higher ee has never been
reported even at lower conversions. Efforts aiming at the
enantioselective hydrogenation of (hetero)-aromatic com-
pounds with chirally modified Ni and Rh catalysts were
barely efficient (17, 18).
The results presented in Figs. 1 and 2 demonstrate the
application limit of cinchona alkaloids as chiral modifiers.
Palladium is more active in the (partial) hydrogenation of
the aromatic ring system of CD than other platinum met-
als (19). This transformation leads to a weaker adsorption
and a loss of enantio-differentiating ability of the modifier.
This side reaction is likely accelerated by the protonation of
the quinoline N-atom by the acidic substrate. It has been re-
ported that the pKa of 1 (4.73) is similar to that of acetic acid
(4.76) (4). When the target reaction is slow, consumption
FIG. 2. Influence of feeding CD during hydrogenation of 1 in acetoni-
trile (standard conditions, 5 wt% Pd/titania).

4
HUCK, MALLAT, AND BAIKER
of the modifier results in a significant drop in enantiose-
lectivity during reaction. The feeding of additional CD to
the batch reactor can compensate for the detrimental effect
of modifier consumption. The smallest feed of CD, which
provided constant ee, was 0.5 mg/h. Taking this value as the
approximate rate of CD consumption during hydrogena-
tion of 1 under ambient conditions, we calculated the rate
of modifier consumption related to that of hydrogenation
of 1. These calculations indicated that the hydrogenation of
20 molecules of 1 is accompanied by the consumption of
1 CD molecule.
Despite the partial hydrogenation of the aromatic rings
of CD during reaction, the overall substrate/modifier ratio
of ca. 20 is not unusually low for chirally modified Pd.
For comparison, achieving 72% ee in the hydrogenation
of phenyl-cinnamic acid required a substrate/CD ratio that
was lower by sixfold (16), and an equimolar amount of
ephedrine was necessary to obtain the best ee in the hy-
drogenation of pyruric acid oxime (20). Apparently, the
low modifier/substrate ratio is, so far, a general feature of
Pd-catalysed enantioselective hydrogenation.
REFERENCES
1. Spino, C., Mayes, N., and Desfosse´s, H., Tetrahedron Lett. 37, 6503
(1996).
2. Romines, K. R., Morris, J. K., Howe, W. J., Tomich, P. K., Horng,
M.-M., Chong, K.-T., Hinshaw, R. R., Anderson, D. J., Strohbach,
J. W., Turner, S. R., and Mizsak, S. A., J. Med. Chem. 39, 4125
(1996).
3. Schmid, R., Broger, E. A., Cereghetti, M., Crameri, Y., Foricher, J.,
Lalonde, M., Mu¨ller, R. K., Scalone, M., Schoettel, G., and Zutter, U.,
Pure Appl. Chem. 68, 131 (1996).
4. Fehr, M. J., Consiglio, G., Scalone, M., and Schmid, R., New J. Chem.,
1499 (1998).
5. Fehr, M. J., Consiglio, G., Scalone, M., and Schmid, R., J. Org. Chem.
64, 5768 (1999).
6. Bacardi, R., and Moreno-Manas, M., Tetrahedron Lett. 21, 551
(1980).
7. Rylander, P. N., in “Catalysis in Organic Synthesis” (W. H. Jones, Ed.),
p. 155. Academic Press, New York, 1980.
8. LeBlond, C., Wang, J., Liu, J., Andrews, A. T., and Sun, Y.-K., J. Am.
Chem. Soc. 121, 4920 (1999).
9. Brogden, P. J., Gabbutt, C. D., and Hepworth, M., in “Comprehensive
Heterocyclic Chemistry” (A. R. Katritzky, C. W. Rees, A. J. Boulton,
and A. McKillop, Eds.), Vol. 3, p. 634. Pergamon, Oxford, 1984.
10. Wiley, R. H., and Hart, A. J., J. Am. Chem. Soc. 77, 2340 (1955).
11. Smyrniotis, P. Z., Miles, H. T., and Stadtman, E. R., J. Am. Chem. Soc.
80, 2541 (1958).
5. CONCLUSIONS
12. Sib, S., Tetrahedron 31, 2229 (1975).
13. Borszeky, K., Mallat, T., and Baiker, A., Catal. Lett. 41, 199 (1996).
14. Tungler, A., Nitta, Y., Fodor, K., Farkas, G., and Mathe, T., J. Mol.
Catal. A 149, 135 (1999).
15. Borszeky, K., Mallat, T., and Baiker, A., Tetrahydron Asym. 10, 4781
(1999).
16. Nitta, Y., and Kobiro, K., Chem. Lett., 897 (1996).
17. Nasar, K., Fache, F., Lemaire, M., Gallezot, P., Besson, M., and Be´ziat,
J. C., J. Mol. Catal. 87, 107 (1994).
18. Laine, R. M., Hum, G., Wood, B. J., and Dawson, M., Stud. Surf. Sci.
Catal. 7, 1478 (1999).
The Pd-catalysed hydrogenation of 4-hydroxy-6-methyl-
2-pyrone afforded 77–85% excess to the (S)-enantiomer of
4-hydroxy-6-methyl-5,6-dihydro-2-pyrone, in the presence
of CD or CD-hydrochloride as chiral modifiers. Presently
this is the only catalytic enantioselective hydrogenation
method available for this transformation. Limitations arise
from the limited stability of the modifiers under reaction
conditions. Although continuous dosing of a minute quan-
tity of modifier can alleviate this limitation, the develop-
ment of similarly functioning but more stable modifiers is
desirable.
19. Bond, G., and Wells, P. B., J. Catal. 150, 329 (1994).
20. Borszeky, K., Mallat, T., Aeschiman, R., Schweizer, W. B., and Baiker,
A., J. Catal. 161, 451 (1996).