The reductive amination of benzaldehyde over Pd/C catalysts: Mechanism and effect of carbon modifications on the selectivity

Article abstract of DOI:10.1002/1099-0690(200007)2000:13<2501::AID-EJOC2501>3.0.CO;2-S

The effects of the acidity of the carbon surface on the reactivity and selectivity of carbon supported noble metal catalysts in the amination of benzaldehyde with ammonia was studied. The equilibiria involved in the homogeneous solution were studied with ₁H NMR. At the start of the reductive amination, dibenzylimine is rapidly formed, which is subsequently hydrogenated to give dibenzylamine. Benzylamine is formed by disproportionation of dibenzylimine to benzylamine and benzylimine. An increase in the number of acidic sites on the carbon support results in higher reaction rates. This may be ascribed to an acid catalysis in the establishment of relevant homogeneous equilibria and/or reduction of the concentration of the inhibiting gem-diamine intermediate at the catalyst surface. A fully selective synthesis of dibenzylamine is achieved by hydrogenation of the intermediate dibenzylimine.

Full text of DOI:10.1002/1099-0690(200007)2000:13<2501::AID-EJOC2501>3.0.CO;2-S

FULL PAPER
The Reductive Amination of Benzaldehyde Over Pd/C Catalysts: Mechanism
and Effect of Carbon Modifications on the Selectivity
Annemieke W. Heinen,[a] Joop A. Peters,*[a] and Herman van Bekkum[a]
Keywords: Aminations / Aldehydes / Hydrogenations / Amides / Activated carbon
The effects of the acidity of the carbon surface on the reactiv- benzylamine and benzylimine. An increase in the number of
ity and selectivity of carbon supported noble metal catalysts
in the reductive amination of benzaldehyde with ammonia
was studied. The equilibria involved in the homogeneous so-
acidic sites on the carbon support results in higher reaction
rates. This may be ascribed to an acid catalysis in the estab-
lishment of relevant homogeneous equilibria and/or reduc-
lution were studied with 1H NMR. At the start of the reduc- tion of the concentration of the inhibiting gem-diamine inter-
tive amination, dibenzylimine is rapidly formed, which is
mediate at the catalyst surface. A fully selective synthesis of
subsequently hydrogenated to give dibenzylamine. Benzyl- dibenzylamine is achieved by hydrogenation of the interme-
amine is formed by disproportionation of dibenzylimine to
diate dibenzylimine.
Introduction
benzylamine that is formed, however, can react with benzal-
dehyde in a similar reaction to dibenzylamine, which in turn
can react with benzaldehyde yielding tribenzylamine. It is
obvious, therefore, that the selectivity of this process is a
major problem. In addition, waste products such as benzyl
alcohol and toluene can be formed. Several review papers
concerning the reductive amination of carbonyl compounds
have been published.[1Ϫ4]
Dibenzylamine is an industrially interesting compound
that is used in rubber and tire compounding, as corrosion
inhibitor, and as an intermediate in the pharmaceutical in-
dustry. It is produced by a reductive amination of benzal-
dehyde with ammonia, via benzylamine (Scheme 1). The
The most frequently used catalytic system for this reac-
tion is Pd/C. The advantages of activated carbon as a sup-
port are its cheapness, its stability in both strongly acidic
and strongly alkaline environments, and, due to its high
porosity, its large surface area. Another advantage, espe-
cially in the case of noble metal catalysts, is the easy recov-
ery of the precious metal by burning off the carbon. How-
ever, since carbon originates from natural products, like
peat or wood, every batch of activated carbon can have dif-
ferent properties, which may be reflected in the performance
of the catalyst.
In this study, the various equilibria and hydrogenation
reactions involved in the reductive amination of benzal-
dehyde are investigated. To investigate the effect of the car-
bon properties on the reductive amination, we modified a
batch of carbon by oxidative and reductive treatments, to
obtain supports with different amounts of oxygen surface
groups. This changes both the acidity and hydrophilicity of
the support.[5]
Results and Discussion
Catalyst Characterization
Palladium was supported on activated carbon, pretreated
with (NH4)2S2O8. The effect of different oxidation treat-
ments on the acidity of this carbon has been described pre-
viously.[5] Upon treatment of carbon with (NH4)2S2O8,
mainly carboxylic acid sites are formed, which can be con-
cluded from neutralization experiments and IR spectra. The
Scheme 1. Overall steps in the reductive amination of benzaldehyde
[a]
Laboratory of Applied Organic Chemistry and Catalysis, Delft
University of Technology,
Julianalaan 136 2628 BL Delft, The Netherlands
Fax (internat.): ϩ 31 15/278 4289
E-mail: J.A. Peters@tnw.tudelft.nl
Eur. J. Org. Chem. 2000, 2501Ϫ2506
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1434Ϫ193X/00/0713Ϫ2501 $ 17.50ϩ.50/0
2501

A. W. Heinen, J. A. Peters, H. van Bekkum
FULL PAPER
Table 1. Relative intensity of 1700 cmϪ1 band (I1700/1580 cmϪ1) and
average Pd particle size (dPd) of the Pd/C catalysts ion exchanged
with Naϩ and reduced at different temperatures (Tred) in 10%
H2/N2
Tred [°C]
cation
I1700/1580 cmϪ1
dPd [nm]
150
Naϩ
Hϩ
Hϩ
Hϩ
0.15
0.30
0.23
0.18
3.0
3.0
3.2
4.9
300
400
neutralization experiments with NaOH revealed 1.4 meq/g
acidic sites to be present on the oxidized carbon. Another
measure of acidity is the intensity, in the IR spectra, of the
1700 cmϪ1 band (carboxylic acid groups[6,7]) relative to that
of the 1580 cmϪ1 band (C-skeleton).[5,8] Upon oxidation,
this ratio increased from 0.0 to 0.45.
Scheme 2. Equilibria leading to benzylimine and dibenzylimine
After oxidation, the carbon was impregnated with an
aqueous solution of Pd(NH4)Cl2 and reduced (10% H2/N2)
at different temperatures during 2 h. As Table 1 indicates,
reduction at temperatures ranging from 150 to 400 °C re-
sulted in a decrease of the amount of carboxylic acid groups
present on the carbon. As shown previously[5], the hydro-
philicity decreases with the loss of oxygen surface sites, and
increases upon sodium exchange. The Pd particle size distri-
bution of the catalysts is displayed in Figure 1. After reduc-
tion at 150 °C, most particles measure 3 nm; when applying
400 °C the average particle size increases to about 5 nm.
Scheme 3. Hydrobenzamide (C) in equilibrium with NH3, leading
to compound D, and possibly E
The formation of hydrobenzamide was studied before by
Dobler[12] and Ogata et al.[11] The latter authors monitored
the reaction with UV spectroscopy. In the present study we
used 1H NMR to follow this reaction, in an attempt to
better identify the products and intermediates formed. A
1
typical H NMR spectrum is shown in Figure 2. We mon-
itored the disappearance of benzaldehyde and the formation
of hydrobenzamide. Hydrobenzamide was present in two of
the three possible isomers (ZZ), (EE), and (ZE). It crystal-
lizes from more concentrated mixtures of benzaldehyde and
ammonia in methanol as white crystals with a sharp melt-
ing point of 101 °C, which suggests that these crystals con-
sist of just one isomer.
Figure 1. Pd particle size distribution of Pd/SX1GNS versus cata-
lyst reduction temperature
Hydrobenzamide Formation
The homogeneous steps in the reductive amination, the
equilibria leading to the imines, are shown in Scheme 2. In
the reaction between benzaldehyde and ammonia, the
benzylimine formed is highly reactive and readily forms hy-
drobenzamide (compound C, Scheme 3). The formation of
this compound was already reported in 1837 by Laurent.[9]
Two mechanisms have been proposed for its formation:
Figure 2. 1H NMR spectrum of a mixture of benzaldehyde and
NH3 in [D4]MeOH. C1 and C2 indicate resonances of the different
isomers of hydrobenzamide (C, Scheme 3); D and E mark the res-
onances of compound D and E (Scheme 3), respectively
At molar ratios NH3/benzaldehyde greater than 1, and
(i) trimerization of benzylimine with loss of ammonia,[10] upon addition of NH3 to hydrobenzamide, additional peaks
1
and (ii) condensation of benzaldehyde with either two at δ ϭ 8.4, 5.3, 4.9, and 1.2 appear in the H NMR spec-
benzylimine or α-amino benzyl alcohol molecules, followed trum, which indicate the formation of compound D
by dehydration to give hydrobenzamide.[11]
2502
(Scheme 3). A peak at δ ϭ 5.1 appeared at the same time,
Eur. J. Org. Chem. 2000, 2501Ϫ2506

Reductive Amination of Benzaldehyde Over Pd/C Catalysts
FULL PAPER
Table 2. Second order rate constants for conversion of benzal-
dehyde at different ammonia to benzaldehyde ratios (MeOH solu-
tion)
[NH3] [benzaldehyde] additive
T
k2
()
()
[°C] [10Ϫ4 L molϪ1sϪ1
]
0.33
0.39
0.33
0.67
0.67
3.33
0.33
0.37
0.33
0.33
0.33
0.33
Ϫ
25 0.8
0.31  NH4NO3 25 16.3
Ϫ
Ϫ
0.53  D2O
Ϫ
40 2.7
25 0.9
25 1.0
25 1.2
librium, the concentration of α-amino benzyl alcohol de-
pends on the starting concentrations of NH3 and benzal-
dehyde, which explains the observed second order kinetics.
Figure 3b shows the effect of temperature on this reac-
tion. Higher temperatures result in a faster reaction. This
explains the slightly higher rate constants observed by
Ogata et al.,[11] that were measured at 30 °C. The equilibria,
however, shift back to benzaldehyde and ammonia[12] at
higher temperatures. This was clearly observed at 90 °C by
the formation of benzaldehyde starting from a solution con-
taining hydrobenzamide and NH3.
Dibenzylimine Formation
Upon reaction of benzaldehyde with benzylamine, forma-
tion of dibenzylimine is observed. The formation of this
relatively stable imine (See Scheme 2) is faster than the reac-
tion between benzaldehyde and ammonia at the same tem-
perature (25 °C, Figure 4); a complete conversion is reached
within 1.5 h. In contrast to the formation of hydrobenzam-
ide, a strong effect of the D2O concentration is observed.
Apparently, D2O shifts the equilibrium to benzaldehyde
Figure 3. Relative intensity of benzaldehyde peak (δ ϭ 9.9) vs. time
during the reaction with ammonia to hydrobenzamide in
[D4]MeOH at (a) 25 °C and different benzaldehyde/NH3/D2O ra-
tios: (᭜) 1:1:0; (᭿) 1:2:0; (ᮀ) 1:2:1.6; (•) 1:10:0; and (b) benzal-
dehyde/NH3 ϭ 1:1 and different temperatures: (᭿) 25 °C and
(•) 40 °C
1
possibly due to the formation of the benzylimine cyclo-
trimer (compound E, Scheme 3).
and benzylamine. From the H NMR intensities, equilib-
rium constants K1, K2, and K were calculated to be 1.5, 27,
and 40, respectively. K, K1, and K2 are defined as:
The formation of hydrobenzamide was followed, over
1
time, with H NMR. In Figure 3a the relative intensity of
K1 ϭ [PhCH(OD)(NH2)]/[PhCHO][PhCH2NH2]
K2 ϭ [C14H13N][D2O]/[PhCH(OD)(NH2)]
K ϭ K1 · K2ϭ [C14H13N][D2O]/[PhCHO][PhCH2NH2]
the aldehyde peak (δ ϭ 9.9) is shown, during reactions with
different ratios of benzaldehyde to ammonia. The only in-
termediate observed is possibly α-amino benzyl alcohol, as
suggested by a small peak at δ ϭ 5.5 (Ͻ 15% of the initial
benzaldehyde intensity). The system was checked for the
formation of benzaldehyde dimethyl acetal by using
CH3OD as solvent. No peaks corresponding to the acetal-
CH3 shifts were observed.
Only a small amount of aminocarbinol (δ ϭ 5.5) was
observed. Its intensity decreased at the same rate as that for
As expected from earlier reports,[11,12] the reaction of
benzaldehyde with NH3 shows second order kinetics
(Table 2). The second order rate constants k2, displayed in
Table 2, were determined by fitting the data with the follow-
ing equation:
v ϭ k2[NH3][PhCHO]
with [NH3] ϭ [NH3]tϭ0 Ϫ /3([PhCHO]tϭ0 Ϫ [PhCHO])
2
Addition of water had no effect on the reaction rate (see
Figure 3a). Acid, in the form of NH4NO3, accelerates the
hydrobenzamide formation considerably (Table 2), which
suggests that the acid catalyzed dehydration[13] of the
Figure 4. Relative intensity of the benzaldehyde peak (δ ϭ 9.9)
vs. time during the reaction with benzylamine to dibenzylimine in
[D4]MeOH (25 °C) at different benzaldehyde/benzylamine/D2O
aminocarbinol is the rate-determining step. Due to the equi- ratios: (ο) 1.1:1:0; (᭜) 1:2.1:0; (᭿) 1:1.1:2.6
Eur. J. Org. Chem. 2000, 2501Ϫ2506
2503

A. W. Heinen, J. A. Peters, H. van Bekkum
FULL PAPER
benzaldehyde. The molar ratio benzaldehyde/aminocarbinol
remained at the constant value of 12.
Two possible mechanisms may explain the results ob-
tained: either benzylamine is formed by hydrogenation of
The reaction of benzylamine with benzylimine (see benzylimine (or compound A), formed from dibenzylimine
Scheme 2) could not be studied because benzylimine cannot via the gem-diamine intermediate (compound B, Scheme 2),
be isolated. Starting from dibenzylimine and ammonia, no or it is formed by direct hydrogenolysis of compound B.
reaction occurred. Even at 90 °C, dibenzylimine is stable.
In both cases, the ammonia concentration will affect the
selectivity. The absence of benzylamine in the reaction mix-
ture during the first 20 min (Figure 5) is due to its con-
sumption by reaction with benzaldehyde or benzylimine
with the formation of dibenzylimine. The stability of di-
benzylimine under these reaction conditions was shown by
the NMR experiments.
From Figure 6 we can also conclude that water does not
significantly affect the reaction rate and selectivity. Sodium
exchange, however, decreased the activity of the catalyst, as
is shown by the larger amount of unconverted dibenzylim-
Reductive Amination
The reductive amination was started after 1 h of equilib-
ration at 90 °C under nitrogen, by applying the H2 pressure.
Under the conditions applied, neither tribenzylamine nor
by-products like benzyl alcohol and benzaldehyde dimethyl
acetal were observed. Only a minor amount of toluene
(Ͻ 1%) was formed. The NMR experiments showed that
hydrobenzamide is not stable under these conditions, so ini-
tially mainly benzaldehyde and ammonia are present.
ine. It may be noted that the H-form of the catalyst will
ϩ
mainly occur as carbonϪCOOϪNH4
.
Hydrogenation of Dibenzylimine
Since dibenzylamine and benzylamine are formed start-
ing from dibenzylimine, the hydrogenation of this com-
pound in the absence of NH3 was studied separately. Reac-
tion conditions similar to those used in the reductive amin-
ation were applied: 40 bar H2, 90 °C, 0.9 g dibenzylimine,
and 0.017 g catalyst. In this reaction, dibenzylamine was
formed with 100% selectivity.
Both the linear relationship between the initial reaction
rate and the initial dibenzylimine concentration and the
straight line obtained when ln(cdibenzylimine) was plotted
Figure 5. Concentrations of benzylamine (᭿), dibenzylamine (ο),
and dibenzylimine (᭜) during the reductive amination of benzal-
dehyde; 80 mL of 2  NH3/MeOH, 17 mg Pd/SX1GNS150, 1 g of against time, suggest first order kinetics in dibenzylimine.
benzaldehyde, 90 °C and 40 bar H2. *At this point, the mass bal-
The rate constant was not affected by the impeller speed,
but increased linearly with the amount of catalyst. External
transport limitation can therefore be ruled out.
ance was Ͻ 100%, most probably due to the formation of hydro-
benzamide when the sample was cooled down. This compound was
not analyzed, but could be precipitated on the catalyst
The catalyst reduction temperature does not affect the
hydrogenation rate (Table 3). However, the sodium ex-
changed catalyst showed a higher activity. The hydrogena-
tion of dibenzylimine in these experiments is significantly
faster than during the reductive amination on the same
catalysts. This indicates inhibition effects during the latter
process.
Addition of benzylamine does not retard the hydrogena-
tion of dibenzylimine (see Table 3). Apparently, dibenzylim-
ine adsorbs much more strongly on the Pd surface. Upon
addition of NH3, benzylamine is formed (Figure 7). During
the reductive amination, a similar effect of ammonia on the
selectivity was observed (see Figure 6). This confirms our
Figure 5 shows the concentrations of benzylamine, di-
benzylimine, and dibenzylamine as a function of time. This
figure displays a rapid formation of dibenzylimine. It is only
after the complete conversion into dibenzylimine that
benzylamine and dibenzylamine are observed. Additionally,
Figure 6 shows that the ammonia concentration has a
strong effect on the selectivity to the two products. Higher
NH3 concentrations increase the benzylamine production.
Table 3. Initial rate of dibenzylimine hydrogenation with different
catalysts and in the presence of benzylamine; 0.9 g of dibenzylim-
ine, 17 mg of catalyst, 80 mL of MeOH, 90 °C, 40 bar H2
catalyst
additive
r0 [mol LϪ1hϪ1
]
Pd/SX1GNS150
Ϫ
0.46
0.46
0.47
0.43
0.74
1.5 g benzylamine
Figure 6. Composition of the reaction mixture after 2 h in the re-
ductive amination of benzaldehyde with ammonia over 5% Pd/
SX1GNS150. Effect of the cationic form of the catalyst, of addition
of water and of NH3 concentration; 80 mL solution, 17 mg of
Pd/SX1GNS150, 1 g of benzaldehyde, 90 °C and 40 bar H2
Pd/SX1GNS300
Pd/SX1GNS400
NaPd/SX1GNS150
Ϫ
Ϫ
Ϫ
2504
Eur. J. Org. Chem. 2000, 2501Ϫ2506

Reductive Amination of Benzaldehyde Over Pd/C Catalysts
FULL PAPER
rapidly, and benzylamine does not inhibit the reaction, we
suggest that the adsorption of compounds A and/or B
(Scheme 2) on the catalytic surface is stronger than the ad-
sorption of dibenzylimine, which would explain the inhibi-
tion effects observed during hydrogenation of dibenzylimine
in the presence of NH3.
Catalyst Acidity and Selectivity
Figure 8 shows the concentrations of dibenzylimine,
benzylamine, and dibenzylamine obtained with the different
Pd/C catalysts after 2 h of reductive amination. Compar-
ison of Figure 5 and 8 suggests that the difference in prod-
uct distributions have to be ascribed to a difference in activ-
ity, rather than to a difference in selectivity.
Figure 8. Composition of the reaction mixture after 2 h in the re-
ductive amination of benzaldehyde with ammonia over 5% Pd/C
catalysts, reduced at different temperatures; 80 mL of 2  NH3/
MeOH, 17 mg of catalyst, 1 g of benzaldehyde, 90 °C and 40 bar H2
First, it must be noted that the available Pd surface de-
creases with increasing reduction temperature. However, the
activity of the differently reduced catalysts in the hydro-
genation of dibenzylimine was approximately the same
(Table 3).
The number of acidic groups on the carbon surface de-
creases with increasing reduction temperature during the
catalyst preparation. This effect of the acidic surface sites is
shown by comparing the H- and Na-form, and is remark-
able, considering the small amount of Hϩ (eq NH4ϩ) pres-
ent in the system. Hϩ is only present on the catalytic sur-
face, leading to locally high Hϩ concentrations and a large
effect on the catalytic performance in the reductive amin-
ation.
As shown by NMR (see above), the establishment of the
homogeneous equilibria is acid catalyzed. Probably, this
also plays an important role in the reductive amination.
Furthermore, the local concentrations of A and B near the
acidic sites at the carbon surface may be relatively low and,
Figure 7. Hydrogenation of dibenzylimine in the presence of (a) 0 
NH3, (b) 1  NH3, and (c) 4  NH3; Benzylamine (᭿), dibenzylam-
ine (ο), and dibenzylimine (᭜); 80 mL of NH3/MeOH, 17 mg of
Pd/SX1GNS150, 0.9 g of dibenzylimine, 90 °C and 40 bar H2
suggestion that dibenzylimine forms benzylamine in the consequently, the inhibiting effect of these species might de-
presence of ammonia during the reductive amination. crease upon increasing the amount of acidic sites at the
With increasing NH3 concentration, the overall dibenzyl- catalyst surface.
imine conversion slows down, whereas the initial rate re-
The performance of the Naϩ exchanged catalyst, which
mains the same, indicating that inhibition effects play a role. is the most hydrophilic, but least acidic one, is in between
From these observations, we must conclude that the hy- the two most acidic catalysts. The hydrogenating activity of
drogenation of dibenzylimine to dibenzylamine competes NaPd/SX1GNS150 is higher than the activity of Pd/
with the formation of benzylamine, either via hydro- SX1GNS150 (Table 3), but any inhibition will be more
genolysis of compound B, or hydrogenation of benzylimine severe, due to the lower acidity. The importance of the
(see Scheme 2). Because benzylimine is hydrogenated very local acidity on the catalyst is clear from the fact that
Eur. J. Org. Chem. 2000, 2501Ϫ2506
2505

A. W. Heinen, J. A. Peters, H. van Bekkum
FULL PAPER
A sodium exchanged catalyst was prepared from Pd/SX1GNS150
by exchange of the carboxylic acid protons with 0.05  NaOH.
This catalyst is coded NaPd/SX1GNS150.
addition of 50 mg SX1GNS to a reaction with 17 mg
NaSX1GNS150 did not influence the activity or selectivity.
The metal loadings and impurities in and on the support were
measured using X-ray fluorescence spectroscopy, XRF (Philips
PW1480). The particle diameter and the dispersion of Pd were
studied with transmission electron spectroscopy (TEM) using a
Philips CM 30 T electron microscope, combined with Energy Dis-
persive analysis of X-rays (EDX).
Conclusions
In the reductive amination of benzaldehyde on Pd/C cata-
lysts, the two main products both result from dibenzylimine.
Benzylamine is formed by hydrogenation of benzylimine,
which is formed by transimination of dibenzylimine with
ammonia. The gem-diamine intermediate adsorbs strongly
on the catalyst, inhibiting the hydrogenation of dibenzylim-
ine. Acidic supports catalyze the establishment of the equi-
libria involved, and possibly reduce the concentration of the
inhibiting intermediate, which results in higher hydrogena-
tion rates.
Reductive Amination: Methanol was saturated with NH3. Of the
resulting 7  solution 23 mL was introduced into a Parr 4842 auto-
clave, made of Hastelloy C276. Methanol (57 mL), the Pd/C cata-
lyst (0.017 g), and benzaldehyde (1 g) were added. The reaction
mixture was heated to the desired temperature under a N2 atmos-
phere and, after 1 h, the H2 pressure was applied. Unless otherwise
stated, these experiments were performed at 90 °C and 40 bar H2.
The hydrogenation of dibenzylimine was performed in the same
autoclave and under similar conditions. Dibenzylimine (0.9 g) was
prepared by mixing of equimolar amounts of benzylamine and
benzaldehyde. Full conversion into dibenzylimine was obtained.
Samples taken during the reaction were analyzed with a Varian
Star 3400 gas chromatograph (CP Sil-5 CB column), applying a
temperature gradient from 50 to 300 °C.
Experimental Section
Materials: Benzaldehyde was obtained from Fluka Chemie AG
(Buchs,
Switzerland).
Benzylamine,
dibenzylamine,
and
Pd(NH3)4Cl2 · H2O were purchased from Aldrich Chemical Comp.,
Inc. (Milwaukee, USA). D2O and deuterated methanol (CD3OD),
were obtained from the Cambridge Isotope Laboratories (Andover,
USA) and CH3OD and (NH4)S2O8 from Merck KGaA (Darm-
stadt, Germany). The steam activated, peat based activated carbon
SX1G was a gift from Norit NV (Amersfoort, The Netherlands).
Acknowledgments
Dr. P. J. Kooyman of the National Centre of High Resolution Elec-
tron Microscopy, Delft University of Technology is acknowledged
for performing the transmission electron microscopy investigations.
Prof. Dr. L. Lefferts, Dr. H. Pinxt, and Dr. H. Oevering of DSM
Research BV are acknowledged for the helpful discussions. Thanks
are also due to Norit NV, for providing the activated carbon
samples, and E. Wurtz for assisting the autoclave experiments. The
Innovation Oriented Program on Catalysis (IOP-katalyse) is
acknowledged for the financial support.
1H- and 13C NMR Experiments: To study the equilibria, H NMR
1
experiments were performed on a Varian Unity-Inova 300 spectro-
meter. [D4]Methanol was used as solvent. Kinetic experiments were
1
performed by following the H resonance of benzaldehyde at δ ϭ
9.9 as a function of time. 1H NMR spectra were recorded at regular
time intervals, during 12 h.
Carbon Preparation and Characterization: SX1G (10 g) was added
to 175 mL of a saturated solution of (NH4)2S2O8 in 1  aqueous
H2SO4. After 22 h of stirring at room temperature, the carbon was
filtered off, washed until the filtrate was free of sulfate (as tested
with BaCl2), and dried as described above. The carbon treated in
this way is denoted as SX1GNS.
[1]
W. S. Emerson, Organic Reactions 1948, 4, 174Ϫ255.
P. A. Rylander, Catalytic Hydrogenation over Platinum Metals,
Academic Press, New York, 1967, pp. 291Ϫ303.
M. V. Klyuev, M. L. Khidekel, Russ. Chem. Rev. 1980, 49,
14Ϫ27.
[2]
[3]
The amount of acid sites on the carbon surface was determined by
selective neutralization with NaOH, according to the method of
Boehm.[14] To 100 mg of carbon, 10 mL of 0.05  NaOH was ad-
ded. After shaking the suspension for four days, the carbon was
centrifuged, washed, and filtered over a 0.45 µm filter (Chromofil)
and then titrated with 0.05  HCl.
[4]
T. Mallat, A. Baiker, in: Handbook of Heterogeneous Catalysis,
Vol. 5 (Eds.: G. Ertl, H. Knözinger, J. Weitkamp), VCH,
Weinheim, 1997, pp. 2334Ϫ2348.
[5]
A. W Heinen, J. A. Peters, H. van Bekkum, Appl. Catal., A
2000, 194Ϫ195, 193Ϫ202.
V. A. Garten, D. E. Weiss, J. B. Willis, Aust. J. Chem. 1957,
10, 295Ϫ308.
W. M. Prest, R. A. Mosher, in: Colloids and Surfaces in Repro-
graphic Technology (Eds.: M. Hair, M. D. Croucher), Wash-
ington, 1982, p.225Ϫ247.
P. Vinke, M. van der Eijk, M. Verbree, A. F. Voskamp, H. van
Bekkum, Carbon 1994, 32, 675Ϫ686.
M. A. Laurent, Ann. 1937, 21, 1130Ϫ1134.
T. I. Crowell, R. K. McLeod, J. Org. Chem. 1967, 32,
4030Ϫ4033.
Y. Ogata, A. Kawaski, N. Okumura, J. Org. Chem. 1964, 29,
1985Ϫ1988.
F. Dobler, Z. Physik. Chem. 1922, 101, 1Ϫ33.
[6]
[7]
An indication of the amount of carboxylic acid sites was obtained
by infrared spectroscopy. Infrared spectra were recorded with a
PerkinϪElmer spectrum 1000 FT-IR spectrometer. KBr pellets
were used, containing 2 mg of carbon in 250 mg of KBr. The spec-
tra were obtained by co-adding 20 spectra with a resolution of 4
cmϪ1. The original spectra were corrected for a curved baseline.
[8]
[9]
[10]
[11]
Catalyst Preparation and Characterization: 5 wt.-% Pd/C catalysts
were prepared by incipient wetness impregnation of 0.95 g
SX1GNS with 127 mg of Pd(NH3)4Cl2 in 2 mL of water. After dry-
ing at 120 °C, the catalyst was reduced in a 10% H2/N2 flow (2 h)
at different temperatures, and passivated at room temperature with
a slowly increasing O2 concentration. A Pd catalyst based on
SX1GNS and reduced at 150 °C is denoted as Pd/SX1GNS150.
[12]
[13]
E. H. Cordes, W. P. Jencks, J. Am. Chem. Soc. 1962, 84,
832Ϫ837.
[14]
H. P. Boehm, Adv. Catal. 1966, 16, 179Ϫ274.
Received February 28, 2000
[O00082]
2506
Eur. J. Org. Chem. 2000, 2501Ϫ2506