Palladium-catalyzed asymmetric hydrogenation of furan carboxylic acids

Article abstract of DOI:10.1016/S0021-9517(03)00184-2

Enantioselective hydrogenation of aromatic and heteroaromatic compounds is the field where chirally modified metal hydrogenation catalysts have the biggest potential compared to homogeneous chiral transition metal complexes. Here we report the hydrogenation of furan and benzofuran carboxylic acids over a cinchonidine-modified 5 wt% Pd/Al₂O₃ catalyst. (S)-Tetrahydrofuran-2-carboxylic acid was synthesized in 4 h at rt and 30 bar with 95% yield and 32% ee. The ee was lower in the hydrogenation of methylfuran carboxylic acids but up to 100% de was achieved. In the slow hydrogenation of benzofuran-2-carboxylic acid, the ee went up to 50% at 29% yield. The potential application of the method is limited by the competing hydrogenation of the quinoline rings of cinchonidine in the latter reaction, necessitating the feeding of small amounts of cinchonidine during reaction. Still, this simple method using an easily available chiral modifier and catalyst affords the highest rate and ee reported so far in the catalytic asymmetric hydrogenation of furan and benzofuran carboxylic acids, and it may be an attractive route in combination with optical resolution. We assume that the reaction mechanism is analogous to that described for α,β-unsaturated carboxylic acids over the same catalyst, involving a 1:2-type interaction between the cinchonidine and the acid dimer.

Full text of DOI:10.1016/S0021-9517(03)00184-2

Journal of Catalysis 219 (2003) 52–58
www.elsevier.com/locate/jcat
Palladium-catalyzed asymmetric hydrogenation of furan carboxylic acids
∗
Mihaela Maris, Wolf-Rüdiger Huck, Tamas Mallat, and Alfons Baiker
Institute for Chemical and Bioengineering, Swiss Federal Institute of Technology, ETH-Höngerberg HCI, CH-8093 Zürich, Switzerland
Received 11 February 2003; revised 2 April 2003; accepted 4 April 2003
Abstract
Enantioselective hydrogenation of aromatic and heteroaromatic compounds is the field where chirally modified metal hydrogenation
catalysts have the biggest potential compared to homogeneous chiral transition metal complexes. Here we report the hydrogenation of
furan and benzofuran carboxylic acids over a cinchonidine-modified 5 wt% Pd/Al O catalyst. (S)-Tetrahydrofuran-2-carboxylic acid was
2
3
synthesized in 4 h at rt and 30 bar with 95% yield and 32% ee. The ee was lower in the hydrogenation of methylfuran carboxylic acids but
up to 100% de was achieved. In the slow hydrogenation of benzofuran-2-carboxylic acid, the ee went up to 50% at 29% yield. The potential
application of the method is limited by the competing hydrogenation of the quinoline rings of cinchonidine in the latter reaction, necessitating
the feeding of small amounts of cinchonidine during reaction. Still, this simple method using an easily available chiral modifier and catalyst
affords the highest rate and ee reported so far in the catalytic asymmetric hydrogenation of furan and benzofuran carboxylic acids, and it
may be an attractive route in combination with optical resolution. We assume that the reaction mechanism is analogous to that described for
α,β-unsaturated carboxylic acids over the same catalyst, involving a 1:2-type interaction between the cinchonidine and the acid dimer.

2003 Elsevier Inc. All rights reserved.
Keywords: Asymmetric hydrogenation; Pd/Al O ; Cinchonidine; Furan carboxylic acids; Benzofuran carboxylic acid
2
3
1
. Introduction
Hydrogenation of aromatic and heteroaromatic com-
pounds is a challenging topic in enantioselective catalysis
as most of the reactions are characterized by very poor
ees [1–5]. Recently, the hydrogenation of furan-2-carboxylic
acid was attempted with homogeneous Rh diphosphine cat-
alysts [6]. Up to 24% ee was obtained at a substrate/catalyst
molar ratio of 20 but the reaction was very slow requiring
Scheme 1. Hydrogenation of furan and benzofuran carboxylic acids over
wt% Pd/Al O modified by cinchonidine (CD).
5
2
0 h at 100 bar to achieve 3% yield. In the absence of a
2 3
useful catalytic method, the optically active tetrahydrofuran-
carboxylic acids are prepared by conventional methods such
as optical resolution [7].
Intrigued by the importance of chiral tetrahydrofuran-
carboxylic acids as useful intermediates [8], we studied the
enantioselective hydrogenation of some furan carboxylic
acids over heterogeneous catalysts (Scheme 1). In an early
paper, Mitsui et al. [9] reported the chemoselective satura-
tion of the aromatic ring of 1 to form tetrahydrofuran-2-car-
boxylic acid without hydrogenolysis-typeside reactions over
Pd/C and Raney Ni catalysts. It is also known that supported
and unsupported Pd modified by naturally occurring chiral
compounds, such as cinchona and vinca alkaloids, are ef-
fective in the enantioselective hydrogenation of C=C bonds
in endo- and exocyclic alkenones (up to 55% ee) [10–14],
α, β-unsaturated carboxylic acids (up to 72% ee) [15–21],
and 2-pyrone derivatives (up to 94% ee) [22–24]. In the hy-
*
Corresponding author.
E-mail address: baiker@tech.chem.ethz.ch (A. Baiker).
0021-9517/$ – see front matter  2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0021-9517(03)00184-2

M. Maris et al. / Journal of Catalysis 219 (2003) 52–58
53
◦ −1
3
drogenation of other functionalized olefins chirally modified
Pd is far less efficient [25–27].
split injection (250 C, 20:1), He carrier gas (42 cm s ),
◦
◦
FID detector (275 C), 80–180 C column temperature.
Derivatization was carried out by heating ca. 5 mg car-
boxylic acid and 0.1 mL of a 1 M solution of trimethylchloro-
silane in MeOH for 2 h. After evaporation of the solvent the
residue was dissolved in 1 mL toluene or ethyl acetate. Hy-
drogenation of 7a was followed by NMR analysis.
Here we report that a cinchona-modified Pd/Al2O3 is the
best catalyst among the Pt-group metal hydrogenation cata-
lysts for the synthesis of chiral tetrahydrofuran-carboxylic
acids, and a comparison to the only known homogeneous
Rh catalyst [6] is also in favor of Pd.
In the hydrogenation of 1a, the (S)-(−)-1b enantiomer
was formed in excess over CD-modified Pd, as confirmed
by using commercially available (R)-(+)-tetrahydrofuran-2-
carboxylic acid (ee 98%, Lancaster). On the basis of analo-
gous separation of the products we assume that also in the
hydrogenation of 3a and 5a the (S)-enantiomer formed in
excess. Two diastereomers (cis and trans) may be formed in
the hydrogenation of 2a and 4a. All four enantiomers could
be detected by GC analysis but determination of ee was re-
liable only for the two cis-isomers. In these reactions the
major enantiomer was not identified.
2
. Experimental
Furan-2-carboxylic acid (1a, Fluka, 98%), 3-methylfu-
ran-2-carboxylic acid (2a, Lancaster, 98%), furan-3-carbo-
xylic acid (3a, Aldrich, 99%), 2-methylfuran-3-carboxylic
acid (4a, Lancaster, 97%), benzofuran-2-carboxylic acid
(
9
5a, Fluka, 98%), and pyrrole-2-carboxylic acid (7a, Fluka,
8%) were purified by sublimation followed by recrystal-
lization from hexane (1a, 2a, 3a, 4a) or toluene (5a). Methyl-
-furoate (6a, Fluka, 97%) (Scheme 2) was used after distil-
The incremental ee was calculated as follows: ꢀee =
2
(
ee1 · yield − ee2 · yield )/(yield − yield ).
lation in vacuum. Cinchonidine (CD, Fluka, 98% alkaloid)
and all other chemicals were used as received.
1
2
1
2
CD consumption during reaction was followed by UV–
vis spectroscopy using a Cary 400 spectrophotometer. Be-
fore analysis the samples were diluted with 2-propanol by a
factor of 5. NMR measurements were carried out on a DPX
A 100-mL autoclave equipped with a 50-mL glass liner,
a PTFE cover, and a magnetic stirrer (500 rpm) was used
for hydrogenations. Total pressure and H2 uptake were con-
trolled by a computerized constant volume constant pressure
equipment (Büchi BPC 9901). Under standard conditions
the catalyst pretreatment and the hydrogenation reactions
were carried out at room temperature and 30 bar. At first,
3
00 spectrometer.
3
. Results and discussion
4
0 mg 5 wt% Pd/Al2O3 catalyst (Engelhard 40692, Pd dis-
3
.1. Choice of catalyst and reaction conditions
At first, various Pt, Rh, Ru, and Pd catalysts supported
persion: 0.21 as determined by TEM) was prereduced with
hydrogen for 5 min in 10 mL solvent. Then 0.45 mmol sub-
strate and 68 µmol CD modifier were added (S/M = 6.5)
and the reaction was started. No external mass transport lim-
itation (no effect of stirring rate) was observed in the rather
slow hydrogenation of the furan ring. Intraparticle diffusion
effects cannot be ruled out completely, but are unlikely due
to the relatively low rate compared to the corresponding
racemic (unmodified) hydrogenation.
on Al2O3, active C, TiO2, or CaCO3 were tested in the hy-
drogenation of furan-2-carboxylic acid (1a). Cinchonidine
was applied as chiral modifier in toluene, at room tempera-
ture and 30 bar (standard conditions). A 5 wt% Pd/Al2O3
provided 83% conversion in only 3 h and 31% ee to (S)-
1
b. Substituting CD with its diastereomer cinchonine led to
somewhat lower conversion (72%) and to inversion of con-
figuration (27% (R)-1b). The chemoselectivity of cinchona-
modified Pd/Al2O3 was also good, at least 95%, depending
on the conditions (solvent, pressure, and temperature). CD-
modified Rh/Al2O3 afforded only 4% ee at 5% conversion
and supported Ru and Pt were inactive.
The role of catalyst pretreatment was investigated in the
hydrogenation of benzofuran-2-carboxylic acid (5a). Prere-
duction of the catalyst stored in air before addition of the
substrate was necessary to reduce the oxidized metal surface
and thus avoid metal leaching by the acidic substrate. Al-
ready 5 min stirring in H2 at room temperature in the solvent
was sufficient to reduce the surface Pd oxides. In the hydro-
genation of 5a the best results were obtained when CD and
the substrate were added together to the catalyst slurry af-
ter prereduction. When the modifier was added first to the
prereduced catalyst and stirred for 2 min before addition
of the substrate, the ee was not affected but the hydrogena-
Conversion and enantioselectivity were determined after
derivatization using an HP 6890 gas chromatograph and a
Chirasil-DEX CB (Chrompack 7502, 25 m × 0.25 mm ×
2
1
3
50 nm) capillary column for the hydrogenation products
b, 2b, 4b, 5b, 6b and a CYCLOSILB (J&W 7419117,
0.0 m × 0.25 mm × 0.25 µm) column for 3b. Conditions:
Scheme 2. Hydrogenation of methyl-2-furoate 6a and pyrrole-2-carboxylic
acid 7a. Standard conditions, 10 mL 2-propanol, S/M = 52 and S/M = 6.5,
respectively.

54
M. Maris et al. / Journal of Catalysis 219 (2003) 52–58
Table 1
The role of solvent polarity in the hydrogenation of 1a and 5a under standard conditions (20 mL solvent)
N
T
Substrate
Solvent
E
Time (h)
Conversion (%)
ee (%)
H O
AcOH
1.00
0.64
0.54
0.27
0.21
0.10
1
3
3
3
3
3
100
88
38
19
12
83
10
22
35
39
42
31
2
2-Propanol
1a
3-Pentanone
THF
Toluene
AcOH
0.64
0.54
0.27
0.21
4
4
4
4
8
6
3
3
23
45
34
19
2-Propanol
3-Pentanone
5a
THF
Fig. 2. Effect of substrate/modifier molar ratio (S/M) on the conversion
dependence of ee in the hydrogenation of 5a. Standard conditions, 10 mL
Fig. 1. Variation of ee with conversion in the hydrogenation of 1a. Standard
conditions, 20 mL toluene, S/M = 6.5.
2-propanol, S/M = 6.5, 26, and 52 corresponding to 20, 5, and 2.5 mg CD,
respectively.
tion rate dropped to less than one-half. In the reverse case,
when adding first the substrate and stirring it with the pre-
reduced catalyst for 2 min before adding the modifier, the
ee decreased by 10–25%, compared to the reaction with si-
multaneous addition of the reaction components. The likely
explanation for these effects is the strong, almost irreversible
adsorption of CD and benzofuran-2-carboxylic acid on the
Pd surface.
were carried out at room temperature and 30 bar (standard
conditions).
The effect of solvent depended on the structure of the
substrate as illustrated in Table 1 on two examples. In the
hydrogenation of furan-2-carboxylic acid (1a), in most sol-
vents except toluene, the ee increased with decreasing sol-
vent polarity characterized by the empirical solvent para-
meter E TN [29], while the rate of conversion decreased. In
the reduction of benzofuran-2-carboxylic acid (5a) the ef-
fect of solvent polarity on the reaction rate was similar but
no clear correlation could be established between solvent po-
larity and ee. Note that toluene and water could not be used
in the latter reaction due to the low solubility of the reac-
tant. In the solvents tested, hydrogenation of 5a was 5–15
times slower than that of 1a. The difference is attributed to
the stronger adsorption of 5a on Pd, and to the larger number
The reaction temperature and pressure had only a small
influence on the catalyst efficiency. For example, in the hy-
drogenation of furan-2-carboxylic acid in toluene, the reac-
◦
tion rate increased with increasing temperature (10–40 C),
as expected, and the ee varied in a narrow range 31–36%
◦
with a maximum at 25 C. Similarly, the hydrogenation
rate increased at higher pressure in the range 5–30 bar and
the ee barely changed. Higher pressures, however, were
detrimental to the enantioselectivity: e.g., hydrogenation of
benzofuran-2-carboxylicacid was attempted to accelerate by
working at 85 bar but the ee dropped to about one-half. We
assume that at this high pressure the quinoline ring of the
modifier is rapidly hydrogenated, leading to a weaker ad-
sorption of the modifier and a loss in ee [22,28]. This point
will be discussed below. Accordingly, further experiments
0
of Pd sites covered by one molecule.
3.2. Competing hydrogenation of cinchonidine
The above parameter study was obscured by the some-
times strong variation of ee with conversion. For exam-

M. Maris et al. / Journal of Catalysis 219 (2003) 52–58
55
Fig. 4. Conversion of cinchonidine (CD) followed by UV–vis analysis and
variation of ee with time in the hydrogenation of 5a. Standard conditions,
Fig. 3. Hydrogenation of CD over Pd/Al O in the absence of substrate
followed by UV–vis spectroscopy. Standard conditions, 10 mL 2-propanol.
2
3
10 mL 2-propanol, S/M = 52.
ple, during hydrogenation of furan-2-carboxylic acid (1a) in
toluene the ee slightly increased with conversion (Fig. 1),
while it was constant in acetic acid (not shown). The conver-
sion dependence of enantioselectivity was affected also by
the substrate/modifier ratio as shown in Fig. 2 for the hy-
drogenation of benzofuran-2-carboxylic acid (5a).
Variation of ee with time and substrate/modifier ratio in
solution is attributed to changes in the actual surface concen-
tration of the modifier. The observation that ee increases with
increasing modifier concentration is a common feature of all
hydrogenation reactions over chirally modified metal cata-
lysts [30]. In some cases this correlation is valid up to very
high modifier concentrations [31], while in other reactions a
maximum is observed [32]. A plausible explanation for the
decay of ee at higher CD concentrations is the change of the
adsorption mode of alkaloid on the metal surface which di-
minishes the ee.
Scheme 3. Hydrogenation of cinchonidine (CD) over 5 wt% Pd/Al2O3 un-
der standard conditions, in 2-propanol (S/M = 52).
The relative amount of modifier remaining in solution after
filtering off the catalyst was determined by subtracting the
reference spectrum of CD (initial solution, corresponding to
0 min) from the measured spectrum (Fig. 3). Since the band
of partially hydrogenatedCD (280–335 nm) is much broader
than the characteristic band of CD (315 nm), quantification
of CD concentration is not hindered.
Over Pd catalysts, hydrogenation of the quinoline ring
system of CD is an important side reaction that weakens its
adsorption on Pd and thus reduces the actual surface CD con-
centration with time [22,28]. As a consequence, the ee varies
also with time (conversion). This effect is clearly indicated in
Fig. 2 by the rapid loss of ee at the highest substrate/modifier
molar ratio of 52, corresponding to the lowest modifier con-
centration. At higher modifier concentrations there is suf-
ficient unreacted modifier in solution or on the support to
replace the hydrogenated modifier on the metal surface, and
thus the ee does not drop with increasing conversion, but
may even increase by approaching the optimum surface con-
centration (S/M = 6.5). Note that hydrogenationof the vinyl
group of CD is fast but this transformation has no influence
on the enantioselection [33,34].
Fig. 4 illustrates the consumption of CD on the 5 wt%
Pd/Al2O3 catalyst in the presence of hydrogen. The exper-
iment corresponding to a substrate/modifier molar ratio of
52 in 2-propanol, shown in Fig. 2, was repeated at 1 and
30 bar, but in the absence of benzofuran-2-carboxylic acid
(5a). After 1 min preadsorption, in both cases 86% of the
initial amount of CD remained in solution, the rest being
adsorbed on the catalyst (starting point of the two CD hy-
drogenation curves in Fig. 4). At 30 bar no more alkaloid
molecule remained in solution after 45 min, as illustrated in
Figs. 3 and 4. However, it is very likely that at this point
some unhydrogenated CD was present on the surface of Pd
and Al2O3 support. As expected, the saturation of the quino-
line ring of CD was much slower at 1 bar: 52% of the initial
amount of CD was still present in solution after 45 min, as
determined by UV–vis analysis (Fig. 4).
The rate and extent of the hydrogenation of CD were
followed by UV–vis measurements, which method offers a
sensitive detection of the (partial) saturation of the quinoline
ring system that is the chromophore moiety of the alkaloid.
Details of the method have been reported earlier [35,36].
To confirm the transformation of CD, the experiment at
30 bar was repeated and after 45 min the solution was ana-
lyzed by NMR. In agreement with the UV–vis analysis, con-
version of the alkaloid was complete and the only detectable

56
M. Maris et al. / Journal of Catalysis 219 (2003) 52–58
anchoring moiety of the modifier providing fixed adsorption
on the Pd surface [32,37–39]. The hydrogenated modifier
is expected to adsorb weaker and thus afford lower ee. The
calculated incremental ee was constant after 120 min at the
level of 13–14%, confirming the assumption.
A practical solution to this difficulty is the feeding of
small amounts of modifier to the reaction mixture [22,32].
For example, when starting the experiment with 2.5 mg CD,
the ee dropped rapidly with the conversion of benzofuran-
2
-carboxylic acid (5a) (Fig. 5). Repeating the experiment
with a stepwise feeding of 0.5 mg/h CD was sufficient to
maintain the ee above the initial value even when feeding
was ceased after 7 h. Under these conditions the overall
substrate/modifier ratio was 22.
3
.3. Effect of substrate structure
Fig. 5. Influence of feeding CD (0.5 mg/h) during hydrogenation of 5a.
Standard conditions, 10 mL 2-propanol, initial S/M = 52.
The efficiency of CD-modified Pd/Al2O3 in the hydro-
genation of some furan and benzofuran carboxylic acids
varied in a broad range (Table 2). Changing the carboxyl
group from 2 to 3 position in the furan ring (3a) lowered the
enantioselectivity but increased the reaction rate. This shift
may be attributed to electronic effects as indicated by the
different acidity of 3a and 1a (pKa = 3.16 and 4.03, respec-
tively) [40]. Assuming an acid-base type interaction between
the substrate and the 1,2-aminoalcohol-type modifier (see
below), this interaction is likely influenced by the pKa value
of the carboxyl group. Besides, the adsorption strength of
the substrate on Pd may also depend on the acidity of the
carboxyl function.
product was formed by saturation of the C=C double bond
and the heteroaromatic ring (Scheme 3).
When both 5a and CD were present in the reaction mix-
ture, hydrogenation of CD was somewhat slower, due to
competing adsorption of substrate and modifier on the Pd
surface. For example, after 45 min there was still about 10%
of the initial amount of CD present in solution. Note that in
this series the analysis of CD consumption was less accurate
as the signals overlapped and 5a and 5b had to be separated
from solution before analysis.
Using these data, interpretation of the decay of ee with
time is straightforward. For convenience, the curve in Fig. 2
corresponding to a substrate/modifier molar ratio of 52 is
replotted in Fig. 4. Clearly, the drop in ee is induced by the
saturation of the quinoline ring system of CD, which is the
Insertion of a methyl group (2a and 4a) slowed down the
formation of tetrahydrofuran-carboxylic acids and lowered
the ee. The low reactivity of trisubstituted C=C bonds on Pd
and other hydrogenation catalysts is well known [41]. A fur-
ther explanation for the loss in activity and selectivity is the
Table 2
a
Structural effects in the enantioselective hydrogenation of furan carboxylic acids
b
Substrate
S/M molar ratio
Time (h)
Yield (%)
ee (%)
de (%)
Degree of rate deceleration
6
6
2
.5
.5
2
4
2
22
95
47
36
32
18
–
–
–
4.5
–
2.1
1a
c
5
6
6
.5
.5
3
20
6
56
15
10
100
92
1.9
–
2
a
a
d
3
6.5
3
100
23
–
5.6
4a
52
3
2
22
100
1.2
6
22
.5
4
23
1
6
29
2
45
50
47
–
–
–
8.6
–
3.2
e
5a
52
a
0
.446 mmol substrate, 10 mL 2-propanol, 40 mg 5% Pd/Al O , standard conditions.
2 3
b
c
d
e
Initial rate of conversion in the absence of CD, related to the rate observed with CD.
In 20 mL toluene.
At 40 C and 3 bar.
◦
CD is fed stepwise in the first 7 h.

M. Maris et al. / Journal of Catalysis 219 (2003) 52–58
57
steric hindrance by the methyl group against the interaction
with the modifier. Two diastereomers (cis and trans) may be
formed in the hydrogenation of 2a and 4a but under standard
conditions only the two enantiomers of one diastereoisomer
could be detected by GC analysis. In the hydrogenation of
C=C double bonds the very high diastereoselectivity to the
cis isomer over supported Pd, including cinchona-modified
Pd/Al2O3, is not unusual [42–44]. To prove that really the
thermodynamically less stable cis isomers were produced,
the hydrogenation of 2a was repeated under conditions that
favor isomerization [45]. Increasing the temperature from 20
quinuclidine N of CD. A second H-bond involving the OH
function of CD stabilizes the structure of the complex on the
metal surface. We proposed a 1:2-type interaction [19] that
is supported by catalytic data, spectroscopic analysis, and
ab initio calculations [43,49,50]. In this structure the unsat-
urated carboxylic acid, present as dimers in apolar medium,
interacts via two H-bonds with the alkaloid modifier, as de-
picted in Fig. 6. Theoretical calculations revealed that the
1:1 model is less feasible due to steric hindrance against ad-
sorption on the Pd surface. We assume that hydrogenation of
furan carboxylic acids obeys an analogous mechanism that
may be described by replacing the R (alkenyl) group in the
substrate by a furyl group (Fig. 6). The validity of this model
is currently under investigation in our laboratory.
◦
to 40 C and lowering the surface hydrogen concentration
by reducing the pressure from 30 to 3 bar decreased the di-
astereomeric excess (de) to 92% (Table 2).
The presence of the carboxyl function in the molecule
was inevitable for achieving significant enantioselectivity,
as indicated by the hydrogenation of the ester derivative
4
. Conclusions
6
a (Scheme 2). Apparently, an acid-base interaction be-
tween the substrate and the basic quinuclidine N of CD is
crucial in the enantio-differentiating step. A further limi-
tation to the application range of cinchona-modified Pd is
the failed hydrogenation of the analogous pyrrole derivative
Cinchona-modified Pd/Al2O3 is a moderately effective
heterogeneous catalyst for the enantioselective hydrogena-
tion of furan and benzofuran carboxylic acids. At room tem-
perature and 30 bar, this catalyst system afforded 32% ee at
full conversion of furan-2-carboxylic acid (1a) to its tetrahy-
dro derivative 1b (3 h, S/M = 6.5), or 50% ee at 29% con-
version of benzofuran-2-carboxylic acid (5a) to its dihydro
derivative 5b (23 h, S/M = 22). In the latter reaction the
slow hydrogenation of the quinoline rings of CD was com-
pensated and the loss of ee prevented by the stepwise feeding
of small amounts of CD. These ee values are the highest
obtained so far in the asymmetric hydrogenation of furan
derivatives [6]. Under standard conditions, at room temper-
ature and 30 bar, hydrogenation of methylfuran carboxylic
acids 2a and 4a afforded 100% de to the cis isomers. At
higher temperature and lower hydrogen concentration (pres-
sure) the diastereoselectivity decreased likely due to isomer-
ization of a reaction intermediate.
7
a (Scheme 2). It is known that pyrrols are more resistant
against hydrogenation than the corresponding benzene and
furan derivatives [41].
Finally, in all reactions addition of CD resulted in a rate
deceleration by a factor of 1.2 to 9; some examples are
shown in Table 2. The lower reaction rate of C=C bond
hydrogenation over Pd modified by cinchona or vinca alka-
loids is apparently a general feature of these catalyst sys-
tems [31,46,47]. In fact, poisoning of Pd by strongly ad-
sorbed N-heteroaromatic compounds, including quinoline,
is well known [41].
3
.4. Substrate–modifier interaction
As noted above, the complete loss of ee in the hydrogena-
A major limitation of cinchona-modified Pd is the com-
peting hydrogenation of the alkaloid modifier, necessitating
low substrate/modifier molar ratios (6.5–22). For compari-
son, in the hydrogenation of ketopantolactone only 4 ppm
CD (S/M = 237, 000) was sufficient to induce over 90% ee
on Pt/Al2O3 [51].
We assume that the nature of substrate–modifier interac-
tion on the Pd surface resembles that of alkenoic acid hydro-
genation over cinchona-modified Pd. An important conse-
quence of this analogy is that the method might be extended
to the enantioselective hydrogenation of other aromatic car-
boxylic acids. Enantioselective hydrogenation of aromatic
compounds is an unsolved, challenging problem.
tion of 6a indicates that the carboxyl function of furan car-
boxylic acids is involved in the substrate–modifier interac-
tion. From this respect the hydrogenation of furan carboxylic
acids resembles that of α, β-unsaturated carboxylic acids
over cinchona-modified Pd. Two different models have been
described for the latter reaction, involving 1:1 or 1:2-type
interactions (Fig. 6). According to Nitta’s model [48], the
deprotonated carboxyl group interacts with the protonated
Acknowledgment
Fig. 6. Schematic representation of cinchonidine (CD)–alkenoic acid inter-
actions involving 1:1 and 1:2 (acid dimer) structures.
Financial support by the Swiss National Science Founda-
tion is gratefully acknowledged.

58
M. Maris et al. / Journal of Catalysis 219 (2003) 52–58
References
[25] T. Yoshida, K. Harada, Bull. Chem. Soc. Jpn. 44 (1971) 1062.
[26] H.U. Blaser, H. Hönig, M. Studer, C. Wedemeyer-Exl, J. Mol. Catal.
A 139 (1999) 253.
[
[
[
[
[
1] R.M. Laine, G. Hum, B.J. Wood, M. Dawson, Stud. Surf. Sci. Ca-
tal. 7B (1981) 1478.
2] C. Exl, E. Ferstl, H. Hönig, R. Rogi-Kohlenprath, Chirality 7 (1995)
[
27] G.V. Smith, J. Cheng, R. Song, in: R.E. Malz (Ed.), Catalysis of Or-
ganic Reactions, Dekker, New York, 1996, p. 479.
28] G. Bond, P.B. Wells, J. Catal. 150 (1994) 329.
29] C. Reichardt, Solvents and Solvent Effects in Organic Reactions,
VCH, Weinheim, 1988.
[
[
211.
3] K. Nasar, F. Fache, M. Lemaire, J.C. Beziat, M. Besson, P. Gallezot,
J. Mol. Catal. 87 (1994) 107.
4] S. Nakagawa, N. Haruna, D.E. Acosta, T. Endo, T. Sugimura, A. Tai,
Chem. Lett. (1999) 1055.
5] S. Coman, E. Angelescu, A. Petride, M. Banciu, V.I. Parvulescu, in:
M.E. Ford (Ed.), Catalysis of Organic Reactions, Dekker, New York,
[
30] A. Baiker, H.U. Blaser, in: G. Ertl, H. Knözinger, J. Weitkamp (Eds.),
Handbook of Heterogeneous Catalysis, Vol. 5, VCH, Weinheim, 1997,
p. 2422.
[
[
31] K. Borszeky, T. Mallat, A. Baiker, Catal. Lett. 41 (1996) 199.
32] C. Leblond, J. Wang, J. Liu, A.T. Andrews, Y.-K. Sun, J. Am. Chem.
Soc. 121 (1999) 4920.
2000, p. 483.
[
[
[
[
6] M. Studer, C. Wedemeyer-Exl, F. Spindler, H.U. Blaser, Monatsh.
Chem. 131 (2000) 1335.
7] M. Morimoto, H. Sato, JP 2002080405, 2002, to Toray Industries,
CAN 136:232193.
8] K. Nishimoto, Y. Hirayama, Y. Fujima, M. Ikenaka, JP 2002171994,
[33] W.-R. Huck, T. Mallat, A. Baiker, Catal. Lett. 69 (2000) 129.
[34] H.U. Blaser, H.P. Jalett, D.M. Monti, A. Baiker, J.T. Wehrli, Stud.
Surf. Sci. Catal. 67 (1991) 147.
[35] W.-R. Huck, T. Mallat, A. Baiker, Adv. Synth. Catal. 345 (2003) 255.
[36] W.-R. Huck, T. Bürgi, T. Mallat, A. Baiker, J. Catal. 216 (2003) 276.
[37] D. Ferri, T. Burgi, A. Baiker, J. Catal. 210 (2002) 160.
[38] K. Borszeky, T. Bürgi, Z. Zhaohui, T. Mallat, A. Baiker, J. Catal. 187
(1999) 160.
[39] W.-R. Huck, T. Bürgi, T. Mallat, A. Baiker, J. Catal. 200 (2001) 171.
[40] C. Cativiela, J.L. Dejardin, J. Elguero, J.I. Garcia, E. Gonzalez, J.A.
Mayoral, Collect. Czech. Chem. Commun. 55 (1990) 72.
[41] S. Nishimura, Handbook of Heterogeneous Catalytic Hydrogenation
for Organic Synthesis, Wiley, New York, 2001, p. 497.
[42] M. Bartók, À. Molnàr, in: S. Patai (Ed.), The Chemistry of Double-
Bonded Functional Groups, Suppl. A3, Wiley, 1997, p. 843.
[43] K. Borszeky, T. Mallat, A. Baiker, Tetrahedron: Asym. 10 (1999)
4781.
2
002, to Nagase, CAN 137:32152.
9] S. Mitsui, Y. Ishikawa, Y. Takeuchi, Nippon Kagaku Zasshi 81 (1960)
86, CAN 56:2289.
2
[10] T. Tarnai, A. Tungler, T. Màthé, J. Petró, R.A. Sheldon, G. Toth, J. Mol.
Catal. A 102 (1995) 41.
[
[
11] C. Thorey, F. Henin, J. Muzart, Tetrahedron: Asym. 7 (1996) 975.
12] A. Tungler, Y. Nitta, K. Fodor, G. Farkas, T. Màthé, J. Mol. Catal.
A 149 (1999) 135.
[13] C. Thorey, S. Bouquillon, A. Helimi, F. Henin, J. Muzart, Eur. J. Org.
Chem. 13 (2002) 2151.
[
[
14] G. Fogassy, A. Tungler, A. Lévai, J. Mol. Catal. A 192 (2003) 189.
15] J.R.G. Perez, J. Malthete, J. Jacques, C. R. Acad. Sci. Paris 300 (1985)
169.
[
[
16] Y. Nitta, Y. Ueda, T. Imanaka, Chem. Lett. (1994) 1095.
17] T.J. Hall, P. Johnston, W.A.H. Vermeer, S.R. Watson, P.B. Wells, Stud.
Surf. Sci. Catal. 101 (1996) 221.
[44] K. Borszeky, T. Mallat, A. Baiker, J. Catal. 188 (1999) 413.
[45] K. Borszeky, T. Mallat, A. Baiker, Catal. Lett. 59 (1999) 95.
[46] Y. Nitta, K. Kobiro, Chem. Lett. (1995) 165.
[
[
[
[
18] G.V. Smith, J. Cheng, R. Song, Catal. Lett. 45 (1997) 73.
19] K. Borszeky, T. Mallat, A. Baiker, Tetrahedron: Asym. 8 (1997) 3745.
20] Y. Nitta, Top. Catal. 13 (2000) 179.
[47] T. Tarnai, A. Tungler, T. Màthé, J. Petró, R.A. Sheldon, G. Tóth,
J. Mol. Catal. A 102 (1995) 41.
[48] Y. Nitta, A. Shibata, Chem. Lett. (1998) 161.
[49] D. Ferri, T. Bürgi, A. Baiker, J. Chem. Soc., Perkin Trans. 2 (1999)
1305.
21] I. Kun, B. Török, K. Felföldi, M. Bartók, Appl. Catal. A 203 (2000)
71.
[22] W.-R. Huck, A. Mallat, A. Baiker, J. Catal. 193 (2000) 1.
[23] W.R. Huck, T. Mallat, A. Baiker, New J. Chem. 26 (2002) 6.
[24] W.-R. Huck, T. Mallat, A. Baiker, Catal. Lett. 80 (2002) 87.
[50] D. Ferri, T. Bürgi, A. Baiker, J. Chem. Soc., Perkin Trans. 2 (2002)
437.
[51] M. Schürch, N. Künzle, T. Mallat, A. Baiker, J. Catal. 176 (1998) 569.