Aerobic oxidation of secondary pyridine-derivative alcohols in the presence of carbon-supported noble metal catalysts

Article abstract of DOI:10.1016/j.cattod.2012.01.012

Pt/C and PtBi/C catalysts prepared on a synthetic mesoporous carbon were evaluated in the oxidation of secondary aromatic alcohols (1-phenylethanol, α-methyl or phenyl pyridinemethanol) with air in different dioxane/water mixtures at 100 °C under 10 bar air. The observed activity for all substrates over both types of catalysts was strongly improved with the addition of water to the dioxane solvent as a result of different interactions of the alcohols with the metallic surface in the apolar dioxane solvent or polar aqueous medium. A promoting effect of bismuth was observed for all substrates. However, the reaction rate was also dramatically influenced by the nature of the aromatic moiety, the nature of the α-group, and the position of the substituent on the pyridine moiety. As a general rule the reactivity was meta < para < ortho and the pyridine derivatives with a phenyl group were more reactive. α-Phenyl-2-pyridinemethanol was totally converted to 2-benzoylpyridine with a selectivity of 96% in dioxane/water 50/50 vol% in the presence of a 2.7% Pt-0.9% Bi/C. Nevertheless, platinum leaching was detected, which could be limited with the promotion by bismuth.

Full text of DOI:10.1016/j.cattod.2012.01.012

Catalysis Today 203 (2013) 133–138
Contents lists available at SciVerse ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Aerobic oxidation of secondary pyridine-derivative alcohols in the presence of
carbon-supported noble metal catalysts
Antonio Frassoldati, Catherine Pinel, Michèle Besson∗
Institut de recherches sur la catalyse et l’environnement de Lyon, IRCELYON, UMR5256, CNRS-Université Lyon 1, 2 Avenue Albert Einstein, 69626 Villeurbanne Cedex, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Pt/C and PtBi/C catalysts prepared on a synthetic mesoporous carbon were evaluated in the oxidation
of secondary aromatic alcohols (1-phenylethanol, ␣-methyl or phenyl pyridinemethanol) with air in
different dioxane/water mixtures at 100 ◦C under 10 bar air. The observed activity for all substrates over
both types of catalysts was strongly improved with the addition of water to the dioxane solvent as a
result of different interactions of the alcohols with the metallic surface in the apolar dioxane solvent
or polar aqueous medium. A promoting effect of bismuth was observed for all substrates. However, the
reaction rate was also dramatically influenced by the nature of the aromatic moiety, the nature of the
␣-group, and the position of the substituent on the pyridine moiety. As a general rule the reactivity was
meta < para ꢀ ortho and the pyridine derivatives with a phenyl group were more reactive. ␣-Phenyl-2-
pyridinemethanol was totally converted to 2-benzoylpyridine with a selectivity of 96% in dioxane/water
50/50 vol% in the presence of a 2.7% Pt–0.9% Bi/C. Nevertheless, platinum leaching was detected, which
could be limited with the promotion by bismuth.
Received 21 November 2011
Received in revised form 19 January 2012
Accepted 19 January 2012
Available online 23 February 2012
Keywords:
Secondary alcohol
(Hetero) aromatic
Oxidation
Platinum catalyst
Solvent effect
© 2012 Elsevier B.V. All rights reserved.
1. Introduction
ketone were obtained respectively, after 48 h in the oxidation
of ␣-methyl-2-pyridinemethanol in water/trifluorobenzene under
The development of clean, selective and atom-efficient reac-
tions remains an area of intense interest in organic synthesis.
This is particular the case for oxidation of primary and secondary
alcohols to carbonyl compounds; the challenge is the oxidation
with molecular oxygen in the presence of heterogeneous cata-
lysts. A substantial effort has been devoted to this type of reaction
and noble metal-catalyzed oxidation with air or pure oxygen is
a clean alternative. Platinum, palladium, ruthenium, and more
recently gold are proposed as efficient catalysts under mild condi-
tions [1–4]. Whilst many reports concerned the selective oxidation
of benzyl alcohol derivatives to either the corresponding alde-
hydes or acids [2,5–7], the oxidation of secondary aliphatic or
aromatic alcohols to the corresponding ketones has been less
studied [8,9]. The heterocyclic carbonyl derivatives are interesting
synthons in pharmaceutical industry [10–16]. For instance, hetero-
cyclic thiosemicarbazone derivatives represent an important series
of compounds with potentially beneficial pharmacological proper-
ties [11]. As examples, different metal complexes with 2-formyl
and 2-acetylpyridine thiosemicarbazones showed antiviral and
antitumor activity [13,14]. Using a polymer incarcerated gold or
gold–platinum bimetallic catalyst, 58% and 80% yields to the desired
ambient conditions [15,16]. 1-Phenylethanol was completely con-
verted to acetophenone under the same conditions within 3 h or
5 h, respectively. Consequently, the aerobic oxidation of pyridine
derivatives is still challenging.
In this work, we investigated the oxidation with air of differ-
ent ␣-methyl or ␣-phenyl pyridinemethanols in the presence of
Pt/C catalyst. Preliminary experiments were performed with 1-
phenylethanol to examine the impact of the nature of the aromatic
ring. Previously, we reported on the positive influence of water in
the aerobic oxidation of 2-octanol; with increasing water contents
in 1,4-dioxane/water solvent, a dramatic enhancement of the cat-
alytic activity was observed [8]. In line with our previous work, this
parameter, as well as the nature and the position of the substituents
on the heteroaromatic moiety were examined.
2. Experimental
2.1. Catalysts preparation and characterization
The supported catalysts were prepared on a synthetic carbon
from MAST Carbon Technology Ltd. The carbon was prepared by
carbonisation at 800 ◦C of a porous polymeric resin prepared from
a phenolic resin and hexamethylenetetramine in ethyleneglycol,
washing, and activation by treatment with CO2 at 850 ◦C with 30%
burn-off. Different batches of platinum catalyst were prepared by
∗
Corresponding author. Tel.: +33 0472445358; fax: +33 0472445399.
E-mail address: michele.besson@ircelyon.univ-lyon1.fr (M. Besson).
0920-5861/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
doi:10.1016/j.cattod.2012.01.012

134
A. Frassoldati et al. / Catalysis Today 203 (2013) 133–138
a
700
600
500
Scheme 1. ␣-Substituted pyridinemethanol derivatives.
400
300
200
100
0
impregnation of the carbon with an aqueous solution of H2PtCl6
to obtain a metal loading of 3–5 wt.%, followed by liquid phase
reduction with formaldehyde in alkaline medium, as described in
[8]. The Pt catalyst was also used as the precursor for preparing
a bimetallic Pt–Bi/C catalyst via a redox surface reaction from an
aqueous BiONO3 solution, using glucose as the reductant, according
to previously reported procedure [17]. By this way, Bi is directly and
preferentially deposited on the noble metal. A molar Bi/Pt ratio of
0.35 was chosen, since this composition has been shown to be often
the optimum in the oxidation of a series of alcohols [1,2] (Scheme 1).
The metal content of the samples were determined by using ICP-
OES (inductively coupled plasma-optical emission spectroscopy).
Surface area and pore volume of the carbon sample were measured
by nitrogen adsorption experiments at 77 K, using a Micromeritics
ASAP 2020 automated system. The sample was degassed for 3 h at
623 K before the measurement. Transmission electron microscopy
(TEM) was performed using a JEOL 200EX microscope operating at
200 kV. The samples were first suspended in ethanol and agitated in
an ultrasonic bath. A drop of the suspended catalyst was applied to
a copper mesh grid with a carbon film, and the ethanol was allowed
to evaporate.
0
0.2
0.4
0.6
0.8
1
relative pressure P/P0
0.6
0.5
0.4
0.3
0.2
0.1
0
0.06
0.05
0.04
0.03
0.02
0.01
0
b
1
10
100
pore diameter (nm)
2.2. Catalytic tests
Fig. 1. Characterization of the carbon support: (a) N2 sorption isotherm (+ adsorp-
tion, desorption), (b) pore size distribution (ꢀ cumulative pore volume,
desorption dV/dD volume).
ꢁ
ꢁ
The reaction tests were performed under a set of realistic con-
ditions in a stirred autoclave reactor of 300 mL made of Hastelloy
in batch mode equipped with a magnetically driven stirrer set. In a
typical oxidation reaction, the reactor was loaded with a solution of
alcohol (15 mmol) in solvent (150 mL), and the supported catalyst,
to get a 1 mol% Pt with respect to the alcohol. This ratio corresponds
to ca. 27 mg Pt introduced in the reactor. After sealing and purging
with argon, the reactor was heated to the desired reaction temper-
ature of 373 K, and then air was introduced up to the pressure of
10 bar. Efficient stirring was started and this time was considered
to be zero time for the reaction.
Liquid samples were periodically removed from the reactor via
a sample diptube. The samples from the reaction were filtered
and then analyzed via gas chromatography (GC) equipped with a
flame ionisation detector (FID) and a HP5 column (30 m × 0.25 mm,
0.25 ␮m film thickness). The external calibration method was used
to quantify reactant and products.
ı (ppm) = 1.49 (3H, d, J = 6 Hz); 4.94 (1H, q, J = 6 Hz); 7.21 (1H, m);
7.28 (1H, m); 7.73 (1H, m); 8.48 (1H, m). Purity = 99% (NMR).
˛-Methyl-3-pyridinemethanol. Using similar procedure, 92%
of colourless oil was obtained. 1H NMR (250 MHz, CDCl3): ı
(ppm) = 1.47 (3H, d, J = 7 Hz); 3.78 (1H, s); 4.92 (1H, q, J = 7 Hz);
7.23 (1H, m); 7.71 (1H, m); 8.36 (1H, m); 8.44 (1H, m). Purity = 99%
(NMR).
˛-Phenyl-2-pyridinemethanol. Using similar procedure, 86% of
white solid was obtained. 1H NMR (250 MHz, CDCl3): ı (ppm) = 5.76
(1H, s); 7.17–7.37 (7H, m); 7.62 (1H, m). m.p. = 74 ◦C. Purity = 99%
(NMR).
˛-Phenyl-3-pyridinemethanol. Using similar procedure, 93% of
white solid was obtained. 1H NMR (250 MHz, CDCl3): ı (ppm) = 5.92
(1H, s); 7.17–7.42 (6H, m); 7.70 (1H, m); 8.30 (1H, m); 8.50 (1H, m).
m.p. = 54 ◦C. Purity = 99% (NMR).
˛-Phenyl-4-pyridinemethanol. Using similar procedure, 86% of
white solid was obtained. 1H NMR (250 MHz, CDCl3): ı (ppm) = 5.76
(1H, s); 7.29–7.34 (7H, m); 8.46 (2H, m). m.p. = 116 ◦C. Purity = 98%
(GC).
The metal analysis in samples was carried out using a ICP-OES
Horiba Jobin Yvon (Activa) instrument.
2.3. Synthesis of the heteroaromatic alcohols
Except ␣-methyl-4-pyridinemethanol which was commercial
(Sigma–Aldrich), the other ␣-heteroaromatic alcohols were syn-
thesized from the commercial corresponding ketones, according to
the procedure described in [18].
3. Results and discussion
Synthesis of ˛-methyl-2-pyridinemethanol. General procedure.
20 g (0.17 mol) of 2-acetyl-pyridine were dissolved in 250 mL of
EtOH. 12.5 g (0.3 mol) of NaBH4 were added slowly. After 1 h
at room temperature, the conversion of the substrate was com-
plete. 250 mL of water were added and the solution was heated
at 90 ◦C for 15 min. After cooling, the product was extracted with
AcOEt. The combined organic solutions were dried over MgSO4,
the solvent was evaporated to give 13 g (70%) of ␣-methyl-2-
pyridine methanol as a pale green oil. 1H NMR (250 MHz, CDCl3):
3.1. Catalyst characterization
Fig. 1a shows the dinitrogen adsorption–desorption isotherm
of the support with a hysteresis loop at a relative pressure
0.7 < P/P0 < 1.0. Correspondingly, the resulting carbon was shown
to be mesoporous with the main pore sizes centered at 11 nm in
diameter (Fig. 1b). Moreover, a BET surface area of 1265 m2 g−1 was
calculated. Elemental analysis revealed that the carbon was poorly
functionalized, and that it was rather hydrophobic.

A. Frassoldati et al. / Catalysis Today 203 (2013) 133–138
135
0.10
0.10
0.08
0.06
0.04
0.02
0.00
0.08
0.06
0.04
0.02
0.00
0
10
20
30
40
50
time (h)
0
1
2
3
4
5
6
time (h)
Fig. 3. Oxidation of ␣-mmethyl-2-pyridinemethanol (Py-2-(CH(OH)-Me) to 2-
acetylpyridine (2-AcPy) in dioxane/water 50/50 vol.%. Reaction conditions: 150 mL
of 0.1 mol L−1 alcohol, 373 K, 10 bar total air pressure, 4.8% Pt/C, alcohol/Pt molar
ratio 100. (ꢁ) 2-MePyMeOH), (ꢀ) 2-AcPy, and (+) molar balance.
Fig. 2. Oxidation of 1-phenylethanol in different dioxane/water solvent. Reaction
conditions: 150 mL of 0.1 mol L−1 alcohol, 373 K, 10 bar total air pressure, 2.8% Pt/C
(plain lines) or 2.7% Pt–0.9% Bi/C (dotted line), (alcohol/Pt molar ratio 100) in (ꢁ)
dioxane, (᭹) dioxane/water 75/25 vol.%, or (ꢂ, ꢂ) dioxane/water 50/50 vol.%.
by using aqueoussolventinsteadof toluene. Furthermore, the activ-
ity in the oxidation of benzylalcohol (hydrophobic alcohol) and of
glycerol (hydrophilic alcohol) in the presence of Pd catalysts sup-
ported on carbon nanotubes, the surface of which was modified
to change the hydrophobic/hydrophilic properties, was different in
polar (water) or apolar (cyclohexane) solvent. The promotion by
water was also revealed in the aerobic oxidation of benzyl alcohol
over Pd nanoparticles confined in mesoporous SBA-16 [23]. Using
acetonitrile, dioxane, DMF and toluene as solvents, the conversions
were low, whilst using water led to high conversion. This was par-
tially explained by the polarity of water which forced the benzyl
alcohol on the catalyst surface with a moderate polarity. It was
also suggested that water helps the abstraction of the ␣-hydrogen
from the alcohol under hydrophobic surroundings [24]. Promoting
effect of water was also observed over different supported metallic
catalysts such as Pt or Ru [25,26].
As reported earlier [7], the XRD diffractogramm of the supported
platinum catalysts was very similar to the one of the raw support,
with no diffraction lines attributed to Pt, suggesting a good disper-
sion of the metallic phase. According to TEM investigations, the Pt/C
catalyst contained platinum particles homogeneously distributed
over the support with a mean size of ca. 2 nm.
3.2. Oxidation of 1-phenylethanol
We first investigated the oxidation of 1-phenylethanol to ace-
tophenone as a model for aromatic secondary alcohol, in the
solvent with different compositions. Fig. 2 gives the results of 1-
phenylethanol consumption as a function of time when the reaction
was performed with Pt/C in dioxane, dioxane/water 75/25 vol.%,
and dioxane/water 50/50 vol.%.
3.3. Oxidation of ˛-substituted pyridinemethanol derivatives
When dioxane was used as the solvent, only a very low conver-
sion of 20% after 2 h was observed. The catalytic activity followed
the same trend as for previously studied oxidation of 2-octanol
[8], i.e. the activity increased by increasing the amount of water
in the solvent. Conversions after 2 h were improved to 66% and 90%
in dioxane/water solvent of 75/25 and 50/50 composition, respec-
tively. The use of a PtBi/C catalyst still increased the conversion
with a total conversion of the alcohol after 3–4 h in dioxane/water
50/50 vol.%.
The scope for oxidation reaction was extended to the oxidation
of different ␣-methyl or ␣-phenyl pyridinemethanol derivatives to
the corresponding acetyl or benzoylpyridine, with the substituent
located in 2-, 3-, or 4-position respective to the nitrogen.
3.3.1. ˛-Methyl derivatives
Oxidation of ␣-methyl-2-pyridinemethanol (Py-2-CH(OH)-Me)
was very low in dioxane. The conversion measured was less than
10% after 6 h. Consequently, all other reactions were performed in
the presence of a mixture of dioxane and water. To illustrate, Fig. 3
shows the reaction concentration profiles for the oxidation of Py-
2-CH(OH)-Me in dioxane/water 50/50 vol.%, using a Pt/C catalyst.
The initial reaction rate of oxidation of Py-2-CH(OH)-Me was
appreciable. However, as the reaction progressed, it declined
remarkably. Conversion increased from 38% after 6 h to only 55%
after 48 h. Compared to 1-phenylethanol oxidation in the same sol-
vent composition (Fig. 2), this substrate behaved differently, due
to the presence of the pyridine moiety. The strong adsorption of
N-containing compounds on the metal surface, possibly leading to
deactivation or even poisoning, has been described in the literature
[27]. Also, these compounds are well known to coordinate strongly
to the metal. Therefore, it is not surprising that differences in the
reaction rate are likely. The reaction was very selective to the ketone
and the molar balance was nearly total.
In terms of selectivity, all three conditions produced excellent
selectivity to acetophenone. Also, no leaching of Pt was detected in
the liquid phase.
As in the case of 2-octanol, a limited reaction rate was observed
in dioxane, and we can again conclude to a beneficial effect of
water addition during the oxidation of an alcohol. In addition, the
measured activities are significant and compare favourably with
those in 2-octanol oxidation [8]. One may note that both substrates
are rather hydrophobic, as well as the metallic surface since very
low oxygen functionalities are present. Dioxane which is apolar
(ETN = 0.164) [19] may adsorb strongly on the catalyst surface and
limit the accessibility for the alcohol. On contrary, addition of water
(ETN = 1) to dioxane makes the solvent more polar. It may then be
argued that the catalytic activity could be related to the different
affinity of the catalyst surface, the substrates and the oxidation
products in the different solvents, resulting in a higher concen-
tration of the substrate close to the Pt particles in dioxane/water
medium.
The influence on the activity of the Pt and PtBi catalysts of
the solvent composition was further studied for all ␣-methyl
pyridinemethanol derivatives. Since these substrates and their cor-
responding ketones are soluble in water, pure water which is the
This phenomenon may be complex, as observed by Prati et al.
[20–22]. Activity of bimetallic Au–Pd and Au–Pt catalysts during
oxidation of n-octyl, cinnamyl or benzylic alcohols was enhanced

136
A. Frassoldati et al. / Catalysis Today 203 (2013) 133–138
a
0.10
0.10
0.08
0.06
0.04
0.02
0.00
0.08
0.06
0.04
0.02
0.00
0
2
4
6
0
4
8
12
16
20
24
time (h)
time (h)
Fig. 5. Oxidation of ␣-phenyl-2-pyridinemethanol to 2-benzoylpyridine in diox-
ane/water 50/50 vol.% in the presence of 2.7% Pt/C (ꢁ, ♦) or 2.7% Pt–0.9% Bi/C (ꢂ, ꢂ).
Reaction conditions: 0.1 mol L−1 alcohol, alcohol/Pt molar ratio = 100, 373 K, 10 bar
air.
0.10
0.08
0.06
0.04
0.02
0.00
b
Py-4-CH(OH)-Me oxidation was performed at a lower temperature
of 333 K over a Pt and PtBi catalyst. This observation was confirmed;
no other peak than the alcohol or the corresponding acetyl deriva-
tives were detected by GC. NMR analysis of the crude mixture was
carried out at the end of the reaction. Only the signals correspond-
ing to the alcohol and the corresponding ketone were observed by
NMR analysis. We thus assumed that either insoluble by-products
of unknown nature were formed, or that some important adsorp-
tion of the reaction alcohol and ketone occurred on the catalyst
surface.
0
4
8
12
16
20
24
time (h)
Additionally, comparison of the results in the presence of Pt or
PtBi catalysts clearly showed that the addition of bismuth drasti-
cally promoted the catalytic activity in both solvents. The beneficial
effect of Bi was particularly important in the case of the ortho
derivative.
A last experiment was performed on Py-3-CH(OH)-Me to exam-
ine whether the rapid decrease of the reaction rate observed could
be attributed to some “over-oxidation” of platinum. In most cases,
over-oxidation is reversible, and initial activity can be restored by
substituting oxygen with an inert gas [28–30]. The reaction was
started in dioxane/water solvent over Pt–Bi/C as usually, and after
5 h of reaction (conversion was 42%), the reaction mixture was
cooled down, the reactor was purged with argon, heated to 373 K
and the oxidation was re-started. The described procedure had no
significant effect on conversion, and the reaction proceeded as in
the previous experiment. Thus, reversible over-oxidation may be
excluded as the cause for deactivation.
c
0.10
0.08
0.06
0.04
0.02
0.00
0
4
8
12
16
20
24
time (h)
Fig. 4. Oxidation of ␣-methyl-pyridinemethanol derivatives, (a) ortho, (b) meta, and
(c) para in the presence of Pt (plain lines) or PtBi (dotted lines). Reaction conditions:
150 mL of 0.1 mol L−1 alcohol, substrate/Pt molar ratio = 100, 373 K, 10 bar air in (ꢀ)
dioxane, (ꢁ, ꢂ) dioxane/water 50/50 vol.%, or (♦, ꢂ) water.
3.3.2. ˛-Phenyl derivatives
The pyridine methanol derivatives with a phenyl substituent
are not soluble in water. Therefore, they were tested in dioxane
and dioxane/water 50/50 vol.%. Fig. 5 shows the evolution of the
concentration of ␣-phenyl-2-pyridinemethanol (Py-2-CH(OH)-Ph)
in both solvents in the presence of 2.7% Pt/C. The reaction was also
tested in the presence of 2.7% Pt–0.9% Bi/C.
The reaction of Py-2-CH(OH)-Ph was still very low in pure diox-
ane in the presence of the monometallic catalyst (8% conversion
after 6 h, not shown). Interestingly, the addition of water sig-
nificantly increased the reaction rate, and total conversion was
obtained after 6 h reaction. The use of PtBi catalyst still drasti-
cally enhanced the reaction rate and the conversion was complete
after 4 h. The molar balance was very satisfying in all runs, and the
selectivity to the ketone was 96% at total conversion.
solvent of choice from an environmental and safety point of view
was also tested. The results of these reactions are summarized in
Fig. 4a–c.
Whatever the operating conditions, the reaction rates were
moderate in the presence of Pt catalyst. Though the performances
were improved with water addition, none of the reaction could
attain complete conversion after 24 h.
The molar balance was systematically calculated all over the
reactions and was found dependent on the solvent and on the
position of the substituent on the pyridine moiety. Very satis-
fying results were obtained in dioxane/water 50/50 vol.% in the
presence of Pt or PtBi catalysts, except for Py-4-CH(OH)-Me oxi-
dation catalyzed by the platinum catalyst. The oxidation took place
smoothly in water, but a deficit was observed very rapidly from
the beginning of the reaction. This deficit was in the range 15–20%
in the oxidation of Py-2 or 3-CH(OH)-Me, but it attained 40% in
the oxidation of Py-4-CH(OH)-Me. In another set of experiences,
The noticeable feature again is that, when the substituent was
in the meta or para position, the reactions in dioxane/water were
characterized by a much lower efficiency. In a similar manner,
Py-3 or 4-CH(OH)-Ph can be converted to the desired ketone.

A. Frassoldati et al. / Catalysis Today 203 (2013) 133–138
137
Table 1
Oxidation of the ␣-substituted pyridinemethanol derivatives after 6 h reaction in dioxane/water 50/50 vol.%.
Substrate
Catalyst
Conversion% (selectivity%)
Pt (Bi) leaching (%)
Ortho
Meta
Para
Ortho
Meta
Para
OH
Me
N
N
Pt/C
PtBi/C
38 (90)
76 (97)
17 (82)
42 (83)
22 (50)
31 (77)
5
21
8 (< 1)
49
18 (4)
< 1 (15)
OH
Pt/C
PtBi/C
100 (96)
100 (92)
19 (70)
47 (95)
30 (75)
53 (95)
8
< 1
< 1 (< 1)
< 1
<1 (<1)
3 (n.m.)
However, conversion after 6 h in the presence of the Pt catalyst was
only 19 and 30%, respectively; it was increased to 47 and 53% over
the bimetallic catalyst, respectively. The selectivity was ca. 70–75%
over Pt and was improved to 96% over PtBi.
by reinforcing the adsorption of the ortho substrate through ␲–␲
interactions of the phenyl ring, leading to a higher rate (Fig. 6b).
Another issue from Table 1 that deserves comment is the impor-
tant leaching into solution of platinum and bismuth. In the case of
the methyl group, the leaching of platinum was the highest for the
meta and para derivatives with the lower reactivity. Since the leach-
ing might be associated with a high complexation of the metal with
the nitrogen atom, this would confirm, as suggested above, that the
better performances of the ortho substrate are related to a lower
interaction of the pyridine nitrogen with platinum, as a result of
steric hindrance of this atom (Fig. 6a).
The leaching of platinum was significantly decreased by pro-
motion of the platinum catalysts by bismuth. Since promotion by
bismuth increases the reaction rates, this suggests again that an
increased reaction rate is related to a lowering of the interaction
of the nitrogen doublet and the metal. Indeed, bismuth which is
selectively deposited on platinum may play the role of a diluent,
which may disfavour this negative adsorption. It was also noted
that the leaching of bismuth was strongly affected by the nature
of the substrate. In the oxidation of the hindered phenyl deriva-
tives, leaching was negligible or not measurable within the limits
of the analysis method. In contrast, during the oxidation of pyridine
methanol substrates, the leaching depended on the position of the
substituent and significant amount of bismuth was dissolved as far
as the ortho-substituent was concerned. In all cases, the presence
of bismuth significantly reduced the leaching of platinum, and this
corresponded to higher reactivity.
3.3.3. Discussion on the nature and the position of the substituent
Table 1 summarizes the conversion and selectivity to ketone
measured after 6 h for the oxidation of the different ␣-substituted
pyridine derivatives over both types of catalysts.
Some of the samples taken out from the reactor showed yellow
or even red colours. One important issue is to determine whether
or not leaching of the active phase occurs during the course of the
reaction. To address this point, the final solutions were analysed for
Pt and Bi by ICP-OES. Leaching of the metals is reported in Table 1 as
the metal depletion on the fresh catalyst. It should be kept in mind
that no leaching was detected during 1-phenylethanol oxidation.
The results indicate that the reaction is highly sensitive to the
position and the nature of the substituent group on the pyridine
moiety. The reasons for these differences can be ascribed to the dif-
ferent adsorption strengths of the pyridine nitrogen and hydroxyl
group onto the metallic surface.
Whether the substituent was methyl or phenyl, the ortho
pyridinemethanol showed the highest conversion after 6 h. In
this case, one may note that the nitrogen atom in pyridine is
very close to the hydroxyl group compared to the other deriva-
tives and the substituent could then strongly disfavour the strong
nitrogen–metal interaction and may improve the accessibility of
the hydroxyl group to the catalyst surface, facilitating oxida-
tive dehydrogenation. Moreover, an internal hydrogen bond could
occur and limit the availability of the nitrogen doublet to the surface
(Fig. 6a). In the other derivatives (meta or para) the strong adsorp-
tion of the pyridine via the nitrogen electron doublet would move
the alcohol function far away from the surface. In this respect also,
the replacement of the methyl group by a phenyl group is important
Ruthenium is also known to be an efficient metal for the
selective oxidation of alcohols [24]. For comparison, an oxidation
experiment of ␣-methyl-4-pyridinemethanol was performed on a
2% Ru/C catalyst prepared on the same support from RuCl3 using the
same preparation procedure as described for Pt. In dioxane/water
50/50 vol.%, conversion after 6 h was 37% with 75% selectivity to the
ketone, compared to 22% and 50%, respectively, in the presence of
Fig. 6. Adsorption mode of ␣-substituted pyridinemethanol derivatives on the metallic surface.

138
A. Frassoldati et al. / Catalysis Today 203 (2013) 133–138
the platinum catalyst. Interestingly, no leaching was detected. This
can be correlated to the higher stability of ruthenium catalysts com-
pared to platinum observed previously during the total oxidation of
nitrogen-containing substrates at higher operating conditions [31].
Further studies are needed to confirm the potential of Ru catalysts.
[4] G.J. Hutchings, Chem. Commun. (2008) 1148–1164.
[5] C. Donze, P. Korovchenko, P. Gallezot, M. Besson, Appl. Catal. B: Environ. 70
(2007) 621–629.
[6] P. Korovchenko, C. Donze, P. Gallezot, M. Besson, Catal. Today 121 (2007) 13–21.
[7] Z. Opre, J.-D. Grunwaldt, T. Mallat, A. Baiker, J. Mol. Catal. A: Chem. 242 (2005)
224–232.
[8] A. Frassoldati, C. Pinel, M. Besson, Catal. Today 173 (2011) 81–88.
[9] B.N. Aggarwal, P. Das, Tetrahedron Lett. 52 (2011) 4954–4956.
[10] J. Boudreau, M. Doucette, A.N. Ajjou, Tetrahedron Lett. 47 (2006) 1695–1698.
[11] R.A. Finch, M.-C. Liu, S.P. Grill, W.C. Rose, R. Loomis, K.M. Vasquez, Y.-C. Cheng,
A.C. Sartorelli, Biochem. Pharm. 59 (2000) 983–991 (references therein).
[12] S. Shimizu, N. Watanabe, T. Kataoka, T. Shoji, N. Abe, S. Morishita, H. Ichimura,
Pyridine and Pyridine Derivatives Ullmann’s Encyclopedia of Industrial Chem-
istry, Wiley VCH Verlag, GmbH & Co KGaA, 2000.
[13] D. Kovala-Demertzi, A. Papageorgiou, L. Papathanasis, A. Alexandratos, P.
Dalezis, J.R. Miller, M.A. Demertzis, Eur. J. Med. Chem. 44 (2009) 1296–1302.
[14] D. Kovala-Demertzi, A. Boccarelli, M.A. Demertzis, M. Coluccia, Chemotherapy
53 (2007) 148–152.
[15] C. Lucchesi, T. Inasaki, H. Miyamura, R. Matsubara, S. Kobayashi, Adv. Synth.
Catal. 350 (2008) 1996–2000.
[16] H. Miyamura, R. Matsubara, S. Kobayashi, Chem. Commun. (2008) 2031–2033.
[17] A.B. Crozon, M. Besson, P. Gallezot, N. J. Chem. (1998) 269–273.
[18] T.D. Lash, K. Pokharel, J.M. Serling, V.R. Yant, G.M. Ferrence, Org. Lett. 9 (2007)
2863–2866.
[19] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, Wiley-VCH
Verlag, GmbH&Co KGaA, 2002.
[20] A. Villa, D. Wang, N. Dimitratos, D. Su, V. Trevisan, L. Prati, Catal. Today 150
(2010) 8–15.
4. Conclusions
The oxidation of 1-phenylethanol and of pyridinemethanol
derivatives substituted by an alpha-methyl or phenyl group were
oxidized to the corresponding carbonyl compound using Pt/C
or Pt–Bi/C catalysts at 373 K under 10 bar air. These substrates
were poorly reactive in pure dioxane. However, addition of water
to dioxane solvent very significantly improved the activity of
the Pt catalyst. Promotion by bismuth still increased the cat-
alytic performances. Compared to 1-phenylethanol, the presence
of the pyridine moiety dramatically decreased the reaction rate,
except for ␣-phenyl-2-pyridinemethanol. Total conversion of this
␣-phenyl substrate was obtained in the presence of Pt/C or Pt–Bi/C
after 4 h or 6 h, respectively, with a selectivity to 2-benzoylpyridine
of 96%. Leaching of platinum was more or less important depend-
ing on the derivative. It became very low or negligible using the Bi
promoted catalyst.
[21] N. Dimitratos, A. Villa, D. Wang, F. Porta, D. Su, L. Prati, J. Catal. 244 (2006)
113–121.
[22] A. Villa, M. Plebani, M. Schiavoni, C. Milone, E. Piperopoulos, S. Galvagno, L.
Prati, Catal. Today (2011), doi:10.1016/j.cattod.2011.09.041.
[23] Z. Ma, H. Yang, Y. Qin, Y. Hao, G. Li, J. Mol. Catal. A: Chem. 331 (2010) 78–85.
[24] R. Anderson, K. Griffin, P. Johnston, P.L. Alsters, Adv. Synth. Catal. 345 (2003)
517–523.
[25] X. Yang, X. Wang, C. Liang, W. Su, C. Wang, Z. Feng, C. Li, J. Qiu, Catal. Commun.
9 (2008) 2278–2281.
[26] X. Yang, X. Wang, J. Qiu, Appl. Catal. A: Gen. 382 (2010) 131–137.
[27] M. Besson, P. Gallezot, Catal. Today 81 (2003) 547–559.
[28] P.C.C. Smits, B.F.M. Kuster, K. van der Wiele, H.S. van der Baan, Appl. Catal. 33
(1987) 83–96.
Acknowledgements
The authors gratefully acknowledge Région Rhône-Alpes for
awarding a doctoral fellowship to A. Frassoldati via the Cluster
Chimie. They are also grateful to Activation for also providing finan-
cial assistance to this study.
References
[29] H.E. van Dam, A.P.G. Kieboom, H. Van Bekkum, Appl. Catal. 33 (1987)
361–372.
[30] P.J.M. Dijkgraaf, H.A.M. Duisters, B.F.M. Kuster, K. van der Wiele, J. Catal. 112
(1988) 337–344.
[31] N. Grosjean, C. Descorme, M. Besson, Appl. Catal. B: Environ. 97 (2010) 276–328.
[1] M. Besson, P. Gallezot, Catal. Today 57 (2000) 127–141.
[2] T. Mallat, A. Baiker, Chem. Rev. 104 (2004) 3037–3058.
[3] L. Prati, M. Rossi, J. Catal. 176 (1998) 552–560.