On the efficiency of phenol and cyclohexanone electrocatalytic hydrogenation - Effect of conditioning and working pH in acetic acid solution on palladium/fluorine-doped tin dioxide supported catalyst

Article abstract of DOI:10.1139/V10-020

The electrocatalytic hydrogenation (ECH) of phenol and cyclohexanone was performed on a conductive Pd/SnO₂:F catalyst. The catalyst was obtained by the impregnation method. We studied the influence of the pH of the supporting electrolyte, the conditioning pH, and the quantity of the conditioning charge passed before hydrogenation. Fourier transform infrared spectroscopy analysis showed that the functionalization of the catalyst surface by the acetic acid electrolyte depends on the pH. A direct correlation was observed between the efficiency of the hydrogenation, the pH of the electrolyte, and the electrode conditioning charge. Phenol hydrogenation was favored in acidic media, whereas cyclohexanone hydrogenation needed an acidic medium for conditioning and a basic medium for hydrogenation. The ECH rate appeared to depend on the functionalization of the catalyst surface, the adsorption of the target organic molecule on the catalyst, and its structural modification with the pH.

Full text of DOI:10.1139/V10-020

4
63
On the efficiency of phenol and cyclohexanone
electrocatalytic hydrogenation — Effect of
conditioning and working pH in acetic acid
solution on palladium/fluorine-doped tin dioxide
supported catalyst
Dihourahouni Tountian, Anne Brisach-Wittmeyer, Paul Nkeng, G e´ rard Poillerat,
and Hugues M e´ nard
Abstract: The electrocatalytic hydrogenation (ECH) of phenol and cyclohexanone was performed on a conductive
Pd/SnO2:F catalyst. The catalyst was obtained by the impregnation method. We studied the influence of the pH of the sup-
porting electrolyte, the conditioning pH, and the quantity of the conditioning charge passed before hydrogenation. Fourier
transform infrared spectroscopy analysis showed that the functionalization of the catalyst surface by the acetic acid electro-
lyte depends on the pH. A direct correlation was observed between the efficiency of the hydrogenation, the pH of the elec-
trolyte, and the electrode conditioning charge. Phenol hydrogenation was favored in acidic media, whereas cyclohexanone
hydrogenation needed an acidic medium for conditioning and a basic medium for hydrogenation. The ECH rate appeared
to depend on the functionalization of the catalyst surface, the adsorption of the target organic molecule on the catalyst,
and its structural modification with the pH.
Key words: electrocatalytic hydrogenation (ECH), functionalization, phenol, cyclohexanone, tin dioxide catalyst.
R e´ sum e´ : L’hydrog e´ nation e´ lectrocatalytique (HEC) du ph e´ nol et de la cyclohexanone a e´ t e´ r e´ alis e´ e avec un catalyseur de
Pd sur SnO2:F conducteur obtenu par impr e´ gnation. Les effets du pH de l’ e´ lectrolyte support, du pH de conditionnement
et de la quantit e´ de charge utilis e´ e pour le conditionnement ont e´ t e´ examin e´ s. La spectroscopie infra rouge par transform e´ e
de Fourier a montr e´ que la fonctionnalisation de la surface du catalyseur avec l’acide ac e´ tique d e´ pend du pH de l’ e´ lec-
trolyte. Une corr e´ lation directe entre l’efficacit e´ d’hydrog e´ nation, le pH de l’ e´ lectrolyte et la charge de conditionnement a
e´ t e´ d e´ montr e´ e. L’hydrog e´ nation du ph e´ nol est favoris e´ e en milieu acide tandis que celle de la cyclohexanone n e´ cessite un
conditionnement en milieu acide et une HEC en milieu basique. Le taux d’HEC d e´ pend ainsi de la fonctionnalisation de
surface du catalyseur, de l’adsorption de la mol e´ cule cible sur la surface fonctionnalis e´ e et de la structure de cette mol e´ -
cule en solution d e´ pendamment du pH.
Mots-cl e´ s : hydrog e´ nation e´ lectrocatalytique (HEC), fonctionnalisation, ph e´ nol, cyclohexanone, catalyseur de dioxyde
d’ e´ tain.
Introduction
ode and by electrolyzing water. This method is applicable to
hydrogenation of all kinds of unsaturated molecules, includ-
ing aromatic rings, carbonyl, nitro, and nitrile compounds.3
The efficiency of the ECH is dependent on numerous pa-
rameters, such as the unsaturated molecule to be hydrogen-
ated, the catalytic metal, the support of the metal, the
supporting electrolyte and its pH, the presence of organic
Electrocatalytic hydrogenation (ECH) of organic com-
pounds is industrially exploited to produce molecules that
have applications as chemical intermediates.1,2 This hydro-
genation method is gradually replacing the catalytic hydro-
genation (CH) process because of its easy implementation.
Indeed, the reactions are carried out at ambient temperatures
and pressures by producing chemisorbed hydrogen at a cath-
4
cosolvent or surfactant, and the current density. Each of
Received 2 December 2009. Accepted 23 January 2010. Published on the NRC Research Press Web site at canjchem.nrc.ca on 6 April
010.
2
´
D. Tountian. Laboratoire d’Electrochimie et de Chimie Physique du Corps Solide, Institut de Chimie- UMR 7177 CNRS- Universit e´ de
Strasbourg, 4 rue Blaise Pascal, 67000 Strasbourg, France; Laboratoire Sciences de Mat e´ riaux d’Electrodes, D e´ partement de Chimie-
Universit e´ de Sherbrooke, 2500 boul. de l’Universit e´ , Sherbrooke, QC J1K 2R1, Canada.
1
A. Brisach-Wittmeyer and H. M e´ nard. Laboratoire Sciences de Mat e´ riaux d’Electrodes, D e´ partement de Chimie- Universit e´ de
Sherbrooke, 2500 boul. de l’Universit e´ , Sherbrooke, QC J1K 2R1, Canada.
´
P. Nkeng and G. Poillerat. Laboratoire d’Electrochimie et de Chimie Physique du Corps Solide, Institut de Chimie- UMR 7177 CNRS-
Universit e´ de Strasbourg, 4 rue Blaise Pascal, 67000 Strasbourg, France.
1Corresponding author (e-mail: Anne.Wittmeyer@USherbrooke.ca).
Can. J. Chem. 88: 463–471 (2010)
doi:10.1139/V10-020
Published by NRC Research Press

4
64
Can. J. Chem. Vol. 88, 2010
Fig. 1. X-ray diffraction patterns of Pd/SnO2:F catalyst (a) before
and after ECH process at (b) pH 3 and (c) pH 12. (^) SnO2:F,
cassiterite structure; (^) tetragonal PdO peaks at 2q = 33.89,
production of the chemisorbed hydrogen (MHads) on the
metallic surface (M) by the electroreduction of water (acidic
or basic medium); the second (reactions 2–4) concerns the
hydrogenation reaction, which starts with the adsorption of
the unsaturated molecule (Y=Z) on the support (A), which
further reacts with the chemisorbed hydrogen; the last group
of reactions (reactions 5 and 6), which concerns the hydro-
gen evolution reaction, is a competition reaction with hydro-
genation. Depending on the parameters cited above, either
the hydrogenation process or the hydrogen evolution will be
predominant.
4
6.66, 54.76, and 71.48 in diffractogram (a) are superposed to the
SnO2:F peaks; (*) metallic Pd.
The behavior of a supported metal is different from the
bulk metal, and the synergy effect of a metal supported on
5
a matrix has been recently studied. The adsorption of the
unsaturated organic molecule is a key step. For the sup-
ported catalysts, this adsorption takes place on the support,6
thus the efficiency of the ECH is relative to the nature of the
catalyst matrix. To control both the adsorption of the unsatu-
rated compound and the desorption of the hydrogenated
product, recent studies have introduced new concepts: ex
7
8
situ and in situ functionalized materials. They consisted of
a modification of the catalyst matrix with organic functional
molecules before and during the ECH process, respectively.
These modifications controlled the adsorption of the target
molecule as in reverse chromatography. It was demonstrated
that the in situ modification of alumina surface depended on
the pH of the electrolyte when acetic acid was used.9
Scheme 1. Conversion of phenol into cyclohexanol via cyclohexa-
none.
Before the introduction of these new concepts, several
studies had focused on phenol and cyclohexanone. On RaNi
electrode, in alkaline media (pH > 13), ECH of cyclohexa-
none led to excellent chemical yields and current efficien-
cies, the best results being obtained with NaOH 0.14 mol/L.
However, when an acetate buffer was used, no hydrogena-
these parameters has an influence on the mechanism of the
ECH described below.
10
11
tion took place. The studies of Dube et al. in a phosphate
buffer (KH2PO4 + NaOH), pH 7, showed that Pd dispersed
on both alumina and activated carbon presented a bad elec-
trocatalytic activity compared with the other metals of the Pt
group. They also demonstrated that the yield of cyclohexa-
none electrohydrogenation was strongly dependent on the
nature of the non-conductive matrix used to disperse Ni
nanoparticles. Santana et al.12 found that the pH had no in-
fluence on the hydrogenation of 2-cyclohexen-1-one when a
nickel sacrificial anode was used. This study concluded that
the ECH of conjugated olefins or carbonyl groups is effi-
cient, whereas for non-conjugated substrates (the case of cy-
clohexanone), it is not.
´
þ
ꢁ
½
1aꢀ
1bꢀ
H3O þ e þ M ! MHads þ H2O Volmer reaction
H2O þ e þ M ! MHads þ OHꢁ
ꢁ
½
Kads
½
2ꢀ
3ꢀ
Y¼Z þ A Ð ðY¼ZÞads A Adsorption
Kdes
½
ðY¼ZÞads A þ 2MHads ! ðYHꢁZHÞads A þ 2M
Hydrogenation reaction
Fichter and al.13 reported a poor yield for the reduction of
phenol to cyclohexanol in 2 N H SO at a platinized Pt elec-
Kdes
½4ꢀ
ðYHꢁZHÞads A Ð YHꢁZH þ A Desorption
Kads
2
4
1
4
trode. On the contrary, Misra et al. showed that phenol
was reduced electrochemically in a divided cell at a Pt cath-
½5aꢀ
H2O þ MHads þ eꢁ ! M þ H2 þ OHꢁ
ode in 2 N HClO , in the presence of tetraethylammonium,
4
Heyrovsky reaction
with a 82.7% yield. More recently, the effect of pH on the
kinetics of phenol ECH was investigated with BaSO as the
4
þ
H3O þ MHads þ eꢁ ! M þ H2 þ H2O
adsorbent. The rate of reaction was considerably lower at
½
5bꢀ
6ꢀ
pH 14, and remained unchanged between acidic and neutral
Heyrovsky reaction
15
4
media. In acidic media (H SO ), Martel et al. found that
2
4
the phenol ECH depended on the catalyst material, whereas
in neutral media the results were linked to the supporting
electrolyte, i.e., the nature of ions in solution.
½
2MHads ! 2M þ H2 Tafel reaction
In this mechanism, three groups of reactions can be ob-
served. The first one (reactions 1a and 1b) is relative to the
Current density also influences the hydrogenation product
yield. This parameter acts as a balance between the two
Published by NRC Research Press

Tountian et al.
465
Fig. 2. ECH of phenol on Pd/SnO2:F in aqueous solution of acetic
acid (0.5 mol/L) as a function of the electrolyte pH and total
charge: (~) pH 12, (&) pH 6, (*) pH 5, and (^) pH 3; I = 20 mA.
Fig. 4. ECH of cyclohexanone on Pd/SnO2:F in aqueous solution of
acetic acid (0.5 mol/L) as a function of the total charge, at different
starting pH values without controlling the pH during the reaction:
(
&) pH 5, (^) pH 6, (*) pH 8, and (~) pH 12; I = 20 mA.
Fig. 3. FTIR spectra of SnO2:F powder subjected to aqueous solu-
tions of acetic acid at (a) pH 3, (b) pH 6, and (c) pH 12.
As shown by the ECH mechanism, the efficiency of hy-
drogenation is linked to the production of chemisorbed hy-
drogen (reactions 1a and 1b) by electroreduction of water.
Here, we present a study of the effect of quantity of charge
used for the conditioning of the catalyst, before the ECH
process, which determines the available quantity of chemi-
sorbed hydrogen. We show that the ECH efficiency depends
on the pH and the conditioning charge.
Experimental
Chemicals
Conductive fluorine-doped tin dioxide (SnO :F) of (r < 5
2
2
–1
Ucm, SBET = 6–10 m g ) was purchased from Keeling &
Walker Limited. Also, Pd(OOCCH ) as Pd precursor was
3
2
purchased from Aldrich.
Phenol, cyclohexanone, cyclohexanone-oxime, cyclohexa-
nol, and 3-methylcyclohexanol (Aldrich) were of +99%
grade of purity and were used as received. High-purity water
competitive reactions (3 and 5a, 5b). A high current density
(
from a Milli-Q unit) was used to prepare the solutions, and
favors the hydrogen evolution reaction, whereas a low cur-
rent density improves the hydrogenation efficiency.16
chloroform (Fisher, spectrophotometry-grade) was used for
the solubilization of Pd(OOCCH ) and the liquid–liquid ex-
This paper reports the results of a study on the effect of
the pH (3 to 12) on the efficiency of phenol and cyclohexa-
none ECH in acetic acid as electrolyte using Pd deposited
3 2
traction. Reticulated vitreous carbon (RVC, with 100 pores
per inch) was purchased from Electrolytica Inc. and used as
cathode. The electrolyte was prepared using glacial acetic
acid from Aldrich.
on conductive fluorine-doped tin dioxide (SnO :F) as cata-
2
lyst. Effectively, the ECH process implied a pH variation
during its proceeding as stated by the mechanism (reactions
1
and 5), and also the functionalization of the catalyst is a
Catalyst preparation
9
pH-depending phenomenon.
Pd/SnO :F catalyst was prepared by impregnating com-
2
Published by NRC Research Press

4
66
Can. J. Chem. Vol. 88, 2010
Table 1. Evolution of the pH as a function of the starting pH and
the quantity of charge passed.
Fig. 6. ECH of cyclohexanone on Pd/SnO2:F in aqueous solution of
acetic acid (0.5 mol/L) as a function of the total charge at two dif-
ferent conditioning pHs: (^) pH 5, and (~) pH 12; working pH = 12;
I = 20 mA.
Starting pH
Charge passed (C)
3
5
6
8
12
Conditioning
7
0
20
3.0
3.0
3.1
3.1
3.1
3.1
5.0
5.1
5.1
5.2
5.2
5.4
6.1
6.7
11.2
12.1
11.9
11.9
10.0
12.0
12.1
12.3
12.0
12.1
12.0
12.0
12.0
12.0
12.1
12.1
5
HEC process
5
0
1
1
00
50
Fig. 5. ECH of cyclohexanone on Pd/SnO2:F in aqueous solution of
acetic acid (0.5 mol/L) as a function of the total charge, at different
starting pHs with pH control during the reaction: (&) pH 5, (^)
pH 6, (*) pH 8, and (~) pH 12; I = 20 mA.
ders was recorded on a Bruker IFS 25 with a resolution of
–1
4
cm . The powders were immersed in the aqueous support-
ing electrolyte solutions containing acetic acid at pH 3, 6,
and 12. These pHs were adjusted with a NaOH (10 N) solu-
tion. The suspensions were filtered, and the powders were
washed with water. After drying at 50 8C for 15 h, the pow-
ders were probed by FTIR using powder not exposed to the
supporting electrolyte as a reference. The samples were dis-
persed in a KBr matrix and pressed in self-supporting disks.
Measurements of adsorption isotherms were carried out as
described below. Cylindrical vials (25 mL) were thoroughly
cleansed, rinsed with high purity water (from a Milli-Q
unit), and dried. SnO :F powder (1 g) was added to each
2
vial, except to the vials serving as blanks, 10 mL of acetic
acid solutions (0.5 mol/L) adjusted to pH 5 and 12 was
added to each vial. After 1 h of stirring in the presence of
SnO :F powders, some of the vials at pH 5 were adjusted to
2
pH 12. To each of these vials, cyclohexanone solution was
added to obtain the desired concentrations. The total volume
in each vial was 15 mL. The SnO2:F powders – cyclohexa-
none suspensions were then stirred with an orbital shaker at
320 trs/min and at ambient temperature (23 ± 2 8C). Adsorp-
tion isotherms were measured after 3 h (time required for
one ECH experiment). A 0.5 mL aliquot was withdrawn
from each vial, filtered, and analyzed by GC chromatogra-
phy to determine cyclohexanone concentration.
mercially available SnO :F. Pd(OOCCH ) was first solubi-
lized in chloroform and poured in a flask containing
2
3 2
SnO :F. After 1 h of stirring, the chloroform was evaporated
2
at low temperature. Water was added to the flask and it was
maintained at 80 8C until the complete evaporation of water.
The prepared catalyst with a nominal concentration of 10%
of Pd was dried overnight at 90 8C in a stove.
Catalyst characterization
Previous MEB characterization of the catalyst has been
Electrocatalytic hydrogenation procedure
1
7
described elsewhere. X-ray measurements were performed
using an X’pert Pro MRD of Panalytical diffractometer with
All electrocatalytic experiments were carried out in a two-
compartment dynamic cell18 under galvanostatic conditions
at ambient temperature. The cathode compartment was filled
with 29 mL of 0.5 mol/L solution of acetic acid in water at
the desired pH. The anodic compartment was filled with the
same solution. Prior to the ECH process, 200 mg of the cat-
˚
Cu Ka (wavelength = 1.5406 A) radiation. Diffractograms
of powders were recorded in 2q scan configuration, in the
10–80 8 2q range, at a fixed incident angle of 0.02 8. Fourier
transform infrared spectroscopy (FTIR) of the SnO :F pow-
2
Published by NRC Research Press

Tountian et al.
467
Scheme 2. Reaction scheme for the hydrogenation of cyclohexanone-oxime to cyclohexanone.
alyst was suspended in the catholic solution and dynamically
circulated (&1 L/min) through the reticulated vitrous carbon
cathode. The catalyst was then conditioned by passing a
known quantity of charge. This quantity of charge served to
reduce Pd(II) in the catalyst after preparation into metallic
Pd and to saturate the cathode material with chemisorbed
hydrogen. For our catalysts, containing theoretically 10% of
Pd, 31 8C was normally necessary to reduce the oxidized Pd,
assuming that Pd is totally in its oxidized form in the cata-
lyst before conditioning. The reduction of oxidized Pd is ki-
measure the quantity of cyclohexanone-oxime that disap-
peared from the solution in the cell.
Results and discussion
As we study the influence of the pH on the ECH effi-
ciency of phenol and cyclohexanone, we have to ensure that
the properties of our catalysts are not modified by changing
pH values. Indeed, XRD measurements carried out on the
Pd/SnO :F catalyst after the ECH process at pH 3 and
2
pH 12 (Fig. 1) showed no change in the catalyst support
(SnO2:F) structure, demonstrating that the catalyst support
remained stable in acidic and basic media. The only change
in the XRD peaks is linked to the electrochemical reduction
of Pd(II) to metallic Pd during the process. As a result,
netically faster than the formation of chemisorbed
hydrogen.19
For phenol electrohydrogenation, the quantity of condi-
tioning charge was fixed to 50 C for all the pH values,
whereas for cyclohexanone this quantity varied from 5 C to
SnO :F is a good matrix, which allows the study of the pH
2
7
5 C.
effect in a large range, contrary to alumina for example,
After conditioning, the coulometer was reset to zero; a
which is deteriorated in basic medium.15
volume of 1 mL of an aqueous unsaturated organic com-
pound solution (25.2 mg/mL of phenol or 26.3 mg/mL of
cyclohexanone) was then added to the cathodic compart-
ment, giving a total volume of 30 mL and an unsaturated
Conversion of phenol can lead to cyclohexanone forma-
tion and eventually to cyclohexanol (Scheme 1), depending
on the reaction conditions.
The ECH of phenol conducted on Pd/SnO :F in acetic
–
3
2
compound concentration of 8.94 ꢂ 10 mol/L. During the
ECH process, the constant current applied was 20 mA for
phenol hydrogenation, 20 and 5 mA for cyclohexanone hy-
drogenation.
acid solution mainly led to the formation of cyclohexanone
and presented different behaviors depending on the working
pH (Fig. 2). Two observations can be made. First, at pH 12,
the process was inefficient. Second, at pH 6 and below, the
more acidic the pH, the more efficient the ECH process be-
After passing different quantities of charge, 0.5 mL ali-
quots were withdrawn from the catholyte and added to
came. Considering the functionalization of the SnO :F sur-
2
1
mL of chloroform, saturated with NaCl, and shaken to-
gether for the extraction of the organic compounds. The or-
ganic phase was removed and dried on Na SO . Finally, an
face as a function of pH, the IR spectra were recorded. If
functionalization of the catalyst by the acetic acid had oc-
2
4
–1
curred, three bands at 1654, 1563, and 1466 cm would ap-
internal standard chloroform solution was added to 0.4 mL
of the organic phase, mixed, and 1mL of this solution was
injected in the gas chromatograph. The GC analyses were
carried out using an Agilent 6890 series chromatograph
equipped with an MS detector and 5HS capillary column
pear. Figure 3 shows that this was not the case at pH 12.
Therefore, it can be concluded that there was no acetic acid
on the catalyst surface. That absence of functionalization at
pH 12 is due to the competitive adsorption of the hydroxyl
anions,2
0,21
whereas at pH 3 and 6 (no excess of hydroxyl
(
30 m ꢂ 0.250 mm ꢂ 0.25 mm). Since hydroxide ions are
ions in solution), SnO :F surface is well-functionalized. As
2
produced (reaction 1) during the electrolysis, the pH can
change depending on the quantity of charge, so we followed
this pH in situ, or maintained its value using a concentrated
acetic acid solution.
8
shown by Cirtiu et al., the functionalization phenomenon
enhances the hydrogenation process by improving the ad-
sorption of the unsaturated molecule. That means that at
pH 12, the catalyst surface is unable to adsorb phenol mole-
culesbecause of the absence of functionalization. As shown
in the ECH mechanism, the adsorption is an important step
that will be the limiting reaction at this pH.
Chemisorbed hydrogen quenching procedure
The quantity of chemisorbed hydrogen produced during
the conditioning was estimated using cyclohexanone-oxime
as a quencher. Open-circuit hydrogenation (catalytic hydro-
However, one also must take into account that the struc-
ture of phenol is pH-dependent, as its pK is 9.96. Thus, at
a
3
genation) of cyclohexanone oxime was done to determine
pH 12, phenol exists as phenolate in solution. As ECH is to-
the quantity of active atomic hydrogen in solution or ad-
sorbed. After conditioning, the current was switched off,
and 1 mL of cyclohexanone-oxime solution (30 mg/mL in
tally inefficient at this pH, we suppose that this form (phe-
nolate) does not adsorb onto the non-functionalized SnO :F
2
at all. Such a strong decrease in hydrogenation rate at high
pH was pointed out by Wismeijer et al.22 in the catalytic hy-
drogenation (CH) process when Pd/C was used. They attrib-
uted this effect to the absence of adsorption of the phenolate
50/50 (v/v) water/methanol) was added to the cell. An ali-
quot was withdrawn from the cell at each hour, extracted
with diethyl ether, and injected to GC chromatograph to
Published by NRC Research Press

4
68
Can. J. Chem. Vol. 88, 2010
ꢃ
Table 2. Quantity of H available (mol) as a function of the charge passed for the conditioning.
ꢃ
Theoretical H
ꢃ
Experimental
conditions
With catalyst +
quantity formed Cyclohexanone-oxime
Cyclohexanone formed H quantity
pH
5
Entries (mol) N = Q/nF starting quantity (mol) after 3 h (mol)
available (mol)
4.40 ꢂ 10
–
4
–4
–4
–4
1
5.18 ꢂ 10
2.27 ꢂ 10
2.20 ꢂ 10
5
0 C charge
–
4
–4
–4
–4
1
2
5
2
5.18 ꢂ 10
2.27 ꢂ 10
1.50 ꢂ 10
3.10 ꢂ 10
–
–
4
4
–4
–4
Without cata-
lyst, without
charge
3
4
0
0
2.27 ꢂ 10
2.27 ꢂ 10
0.45 ꢂ 10
0.90 ꢂ 10
1
2
0
0
on the catalyst. Phenolate molecules are preferably found in
the aqueous phase because of the electrostatic repulsion with
the negatively charged carbon surface due to the OH ions.
A starting pH of 6 with final pH equal to 12 seemed to be
the best condition for ECH of cyclohexanone. Indeed, for a
starting pH of 6, it remained slightly acidic at the end of the
conditioning and went up to basic pH after a few quantity of
charge (20 C) was passed for the ECH. When starting with a
pH 8, only 7 C was sufficient to bring the pH to the basic
range during the conditioning (Table 1).
Then, we decided to control the pH of the solution to
know exactly the best working pH for cyclohexanone hydro-
genation. We realized ECH of cyclohexanone at these dif-
ferent pHs with in situ control of the pH (Fig. 5), and we
observed that the maximal efficiency was obtained for pH 8
and that it was near 50%. As a result, a starting pH of 6 was
not the only reason for the ECH efficiency without pH con-
trol (Fig. 4) because this process was kinetically limited
when the electrolyte was constantly maintained at pH 5 and
–
At pH < 10, we observed that the ECH process of phenol
becomes efficient. Moreover, we showed that the more
acidic the solution, the more efficient the hydrogenation. As
there is a significant difference between conversion of phe-
nol at pH 6 and at pH 3, we suggest the most acidic medium
is favorable for phenol electrohydrogenation. Indeed, it is
well-known that in a strong acid solution, the hydrogenation
of phenol is facilitated by some decrease in its resonance
2
3
stabilization energy. Also, the rate of enhancement of aro-
matics hydrogenation in superacidic solutions was attributed
to the removal of the aromatic resonance stabilization en-
ergy by Feiring24 and Mador.
25
Therefore, the ECH process of phenol on Pd/SnO :F in an
2
acetic solution will work, provided that the catalyst has been
previously functionalized at a pH £ 6, and it will be more
optimal in a more acidic medium. The pH values affected
the catalyst/solution interface that has a great influence on
the adsorption/desorption steps of the ECH process.
6
. Even if the catalyst surface was functionalized at these
pHs (Fig. 3), the process remained inefficient, which means
that the functionalization is not the only sufficient condition.
Comparing Fig. 4 and Fig. 5, we can suppose that the
starting pH must be acidic, whereas the working pH must
be basic to get the best efficiency. This is the reason why
the process is divided into two steps: the conditioning and
the working parts. We verified the impact of the condition-
ing pH by conducting ECH of cyclohexanone with two dif-
ferent procedures (Fig. 6): First, we conditioned the catalyst
at pH 5 during 50 C, and then adjusted the working pH to
Cyclohexanone is the intermediate product in the com-
plete ECH of phenol into cyclohexanol (Scheme 1), and it
is interesting to study its reactivity as a function of pH
when it is the starting product to see the influence of its ad-
sorption and reactivity on the ECH efficiency. If we con-
sider the ECH conducted at several pHs without controlling
the pH (Fig. 4), we observe that the maximal efficiency is
obtained for a starting pH equal to 6. Nevertheless, it is im-
portant to notice that, apart from the starting pH 5 and 12,
which remained stable during the process, the other starting
pH 6 and 8 change during the experiment for a final pH
equal to 12 (Table 1). Maintaining the pH is especially im-
portant for values higher than 5.5 due to the loss of the buf-
1
2 before adding cyclohexanone. Second, we used pH 12
both in conditioning and working conditions. Figure 6 per-
fectly shows that catalyst must be conditioned at pH 5 to be
able to convert cyclohexanone into cyclohexanol.
Then, it is important to know if there is a difference in the
quantity of chemisorbed hydrogen produced in acidic and
basic media. We know that the chemisorbed hydrogen’s pro-
fer capacity of acetic acid, which has a pK of 4.76. The
a
27
duction kinetics depends upon the metal and also the me-
dium (acidic or basic).19 To quantify the number of moles of
Hads produced by the conditioning in acidic versus basic me-
speed of pH change increases when the starting pH increases
(Table 1).
As observed for phenol hydrogenation, pH 12 is ineffi-
ꢃ
dium we used an H quencher, which completely reacts with
cient, owing to the non-functionalization of the catalyst sur-
face. Surprisingly, pH 5 is the less efficient medium,
although the surface is functionalized. In fact, Breitner et
ꢃ
all the H available. Before that, one must take into account
that during the conditioning, some electrons are used to re-
duce the oxidized Pd and some to produce chemisorbed hy-
drogen (reaction 1). To determine the quantity of
chemisorbed hydrogen produced during the conditioning,
we made an open-circuit hydrogenation of cyclohexanone-
oxime, which was added to the catholyte after switching off
the current at the end of the conditioning. After a total time
2
6
al. had made the same observation in the CH process for
rhodium and ruthenium catalysts, which hydrogenated car-
bonyl compounds rapidly in neutral and basic solutions and
slowly in acid solutions. They also noted that palladium was
ineffective to hydrogenate aliphatic ketones, but this be-
comes possible in our case by the ECH process.
Published by NRC Research Press

Tountian et al.
469
Fig. 7. Adsorption isotherm for cyclohexanone after 3 h in contact with F-doped SnO2 powders in aqueous solution of acetic acid (0.5 mol/
L) as a function of the initial pH: (~) pH 12, (^) pH 5, (&) 1 h of electrolyte–SnO2:F suspension stirring at pH 5 before adding cyclo-
hexanone at pH 12.
of 4 h, the cyclohexanone-oxime CH ended, i.e., its reaction
with active chemisorbed was total.
We calculated the theoretical quantity of H formed dur-
pH 12. Above that, its adsorption is maximal when the cata-
lyst is conditioned at pH 5 (functionalization) and when cy-
clohexanone is introduced at pH 12. Hence, the activated
cyclohexanone has a good adsorption onto the functional-
ized catalyst. This activation of cyclohexanone may be due
to the solvation, which differs in acidic and basic environ-
ments. The keto–enol equilibrium is also more favorable in
alkaline medium. Indeed, it was demonstrated that the al-
ꢃ
ing conditioning according to Scheme 2. Then, experiments
were conducted in the presence and absence of a catalyst,
and we determined the available quantities of active hydro-
gen (Table 2). This was done at a charge of 50 C during
conditioning. Table 2 shows that the maximum quantity of
ꢃ
–4
H produced is 4.4 ꢂ 10 mol (Table 2, entry 1) at pH 5
and 50 C conditioning of the electrocatalysis. We observed
that at pH 12 (Table 2, entry 2), the available quantity of
kene group is more reactive than the carbonyl group in the
ECH process.1
2,26
The increase in the rate of hydrogenation
of cyclohexanone in KOH solution in the CH process was
ꢃ
22
H was less important. However, it is interesting to remark
also observed by Wismeijer et al. It was assumed that the
that without either a catalyst or a charge, little reactivity of
cyclohexanone was observed at pH 5 (Table 2, entry 3),
whereas no reaction at all took place at pH 12 (Table 2, en-
promoted effect came from the reactive enolate anions or
ionized hydrated cyclohexanone species formed in the pres-
ence of OH . In our case, the FTIR spectrum (not presented)
of cyclohexanone at pH 12 shows an additional band com-
pared with that of cyclohexanone at pH 5, that band was at-
tributed to the enolate ion.
–
–4
try 4). Consequently, we assume that only 3.5 ꢂ 10 mol of
ꢃ
H (corresponding to 33.7 C) are really produced at pH 5 on
the catalyst over 50 C. In these conditions, the quantity of
active hydrogen available at pH 5 or pH 12 is in the same
range and the observed differences are in the experimental
error range. Then, experimentally, a maximum of 20 C is
used for the reduction of Pd(II) into Pd(0), which means
that all the Pd is not in Pd(II) form in the catalyst before
conditioning. We can assume that this quantity is always
the same whatever the quantity of charge passed for condi-
tioning.
From these observations, we can conclude that the differ-
ence in the efficiency of the ECH process due to the condi-
tioning pH is not linked to the available quantity of Hads
because this quantity is the same in acidic and basic media.
We can suppose that basic pH allows cyclohexanone to be
activated and this activated form of cyclohexanone may
have affinity with functionalized catalyst after conditioning
in acidic medium (no functionnalization in basic medium).
As pointed out by Dub e´ et al.,11 the efficiency also de-
pends on the kinetics of the transfer of a monoatomic hydro-
gen (Hads) from the metal site to the p-bond of the adsorbed
organic molecule. This transfer can differ from acidic to ba-
sic medium depending on the configuration of the substrate
in solution.
Because the quantity of the available Hads can influence
the hydrogenation, we were interested in studying its role in
the ECH of cyclohexanone on the same catalyst in acetic
acid electrolyte. For this study, the conditioning pH was 5
and the working pH was 12. To improve the efficiency of
the reaction, the constant current applied after conditioning
was 5 mA. As seen in the quenching of Hads, the experimen-
tal quantity of charge necessary to reduce all the Pd(II) to
Pd(0) was around 20 C. Then, we can estimate the approxi-
mate quantity of Hads as a function of conditioning charge
(Table 3). This table shows that if the quantity of charge
for the conditioning increases, the quantity of Hads available
Indeed, adsorption of cyclohexanone on SnO :F in presence
2
of acetic acid (Fig. 7) is more favored at pH 5 than at
Published by NRC Research Press

4
70
Table 3. Theoretical quantity of H formed and available (mol) as a function of the conditioning charge.
Can. J. Chem. Vol. 88, 2010
ꢃ
ꢃ
Experimental quantity of
charge QPd necessary to
reduce Pd(II) into Pd(0) (C)
Experimental quantity of charge
QHads for chemisorbed hydrogen
production with QHads = Q – QPd
Theoretical H quantity N formed and
available (mol) with N = QHads/(nF)
and n = 1
Charge Q passed for
conditioning (C)
5
20
20
20
20
0
5
30
55
None
–
–
–
5
4
4
25
50
75
5.18 ꢂ 10
3.11 ꢂ 10
5.70 ꢂ 10
Fig. 8. ECH of cyclohexanone on Pd/SnO2:F in aqueous solution of
acetic acid (0.5 mol/L) as a function of the total charge for different
quantities of conditioning charge: (&) 5 C, (~) 25 C, (*) 50 C,
and (^) 75 C; I = 5 mA; conditioning pH = 5 and working pH = 12.
Fig. 9. CH of cyclohexanone on Pd/SnO2:F in aqueous solution of
acetic acid (0.5 mol/L) as a function of the total charge for different
quantities of conditioning charge: (~) 25 C, (*) 50 C, and (^)
75 C; I = 5 mA; conditioning pH = 5 and working pH = 12.
theless, the catalytic hydrogenation of cyclohexanone, and
most of aliphatic ketones, was impossible on palladium cat-
alyst when hydrogen gas was used as source of chemisorbed
hydrogen.1
2,26
This is explained by the kinetic barrier (re-
lated to the low solubility and low dissociation of molecular
increases. Indeed, the more the charge passed before the be-
ginning of ECH, the faster the reaction (Fig. 8), which
shows that the kinetics of cyclohexanone hydrogenation is
affected. This observation can explain why a difference is
noticed in the ECH process according to the nature of the
metal used27 or the conductivity of the matrix supporting
28
hydrogen (H )) that is too high. So, it can be said that the
2
production of chemisorbed hydrogen is not systematic in CH
11
conditions. The ECH offers soft conditions of work in
which the production of chemisorbed hydrogen is obvious
as it was shown by the quenching. Additionally, in the ECH
29
process, the polarization of the catalyst and the cyclohexa-
1
7
the metallic nanoaggregates. Indeed, the quantity of Hads
generated is linked to these parameters.
none C=O bond takes place.9
To study only the influence of the conditioning charge di-
rectly on cyclohexanone hydrogenation, CH of cyclohexa-
none was conducted at the end of each conditioning
Conclusion
The in situ functionalization of the catalyst support affects
the electrocatalytic hydrogenation of phenol and cyclohexa-
none. This functionalization takes place in an acidic medium
(pH £ 6) with acetic acid as electrolyte. The effect of the pH
on the ECH process is explained by the catalyst functionali-
zation, which affects the catalyst/solution interface also by
the structural changes of the target molecule.
(
Fig. 9). Cyclohexanone was added to the catholyte after
the conditioning, i.e., in the presence of chemisorbed Hads
demonstrated by the quenching), and then we verified
(
whether or not it was hydrogenated after 100 min. Surpris-
ingly, Fig. 9 shows that cyclohexanone hydrogenation took
ꢃ
place and the quantity of H (Table 3) produced during con-
ditioning played an important role on the CH process. This
result demonstrated that when chemisorbed hydrogen is pro-
duced (by the electroreduction of water), cyclohexanone
could catalytically be hydrogenated on Pd catalyst. Never-
Hydrogenation of phenol to cyclohexanone increases as
the medium becomes more acidic. When the support surface
is functionalized, cyclohexanone is hydrogenated only in a
basic medium. For a complete hydrogenation of phenol
Published by NRC Research Press

Tountian et al.
471
with good kinetics, it would be interesting to conduct the
first step of hydrogenation, i.e., the formation of cyclohexa-
none in an acidic medium and then increase the pH to basic
values to facilitate the formation of cyclohexanol. The con-
ditioning charge influences the amount of chemisorbed hy-
drogen available for the hydrogenation of the organic
compound. This work has identified yet another advantage
of ECH versus CH, owing to the fact that the production of
chemisorbed hydrogen is systematic in the ECH process.
L.; M e´ nard, H. J. Appl. Electrochem. 2003, 33 (6), 541.
doi:10.1023/A:1024495505252.
(12) Santana, D. S.; Melo, G. O.; Lima, M. V. F.; Daniel, J. R.
R.; Areias, M. C. C.; Navarro, M. J. Electroanal. Chem.
2
004, 569 (1), 71. doi:10.1016/j.jelechem.2004.02.015.
13) Fichter, F.; Stocker, R. Ber. Bunsenges. Phys. Chem 1914,
7 (2), 2003. doi:10.1002/cber.19140470296.
14) Misra, R. A.; Sharma, B. L. Electrochim. Acta 1979, 24 (6),
27. doi:10.1016/0013-4686(79)87058-9.
15) Laplante, F.; Brossard, L.; M e´ nard, H. Can. J. Chem. 2003,
1 (3), 258. doi:10.1139/v03-027.
(
(
(
(
(
4
7
8
Acknowledgements
16) Mahdavi, B.; Chambrion, P.; Binette, J.; Martel, E.; Lessard,
J. Can. J. Chem. 1995, 73 (6), 846. doi:10.1139/v95-105.
17) Tountian, D.; Brisach-Wittmeyer, A.; Nkeng, P.; Poillerat,
G.; M e´ nard, H. J. Appl. Electrochem. 2009, 39 (3), 411.
doi:10.1007/s10800-008-9686-0.
We are thankful to ‘‘Coop e´ ration France–Burkina Faso’’,
to the French Government Funding, and to Universit e´ de
Sherbrooke for the financial and material support they pro-
vided.
(
18) Bannari, A.; Cirtiu, C.; Kerdouss, F.; Proulx, P.; M e´ nard, H.
References
Chem. Eng. Process. 2006, 45 (6), 471. doi:10.1016/j.cep.
2
005.11.007.
19) Weast, R. C., Ed. In Handbook of Chemistry and Physics,
6th ed.; 1975–1976; p. D-142.
20) Martinez, A. I.; Acosta, D. R.; Cedillo, G. Thin Solid Films
005, 490 (2), 118. doi:10.1016/j.tsf.2005.04.060.
(
(
(
1) Ruest, L.; M e´ nard, H.; Moreau, V.; Laplante, F. Can. J.
Chem. 2002, 80 (12), 1662. doi:10.1139/v02-182.
2) Tsyganok, A. I.; Otsuko, K. Appl. Cat. B Environ. 1999, 22
(
(
(
(
(
(
5
(
1), 15. doi:10.1016/S0926-3373(99)00028-4.
2
3) Robin, D.; Comtois, M.; Martel, A.; Lemieux, R.; Cheong,
21) Farrow, J. B.; Warren, L. J. Colloids Surf. 1988, 34 (3), 255.
A. K.; Belot, G.; Lessard, J. Can. J. Chem. 1990, 68 (7),
doi:10.1016/0166-6622(88)80104-5.
22) Wismeijer, A. A.; Kieboom, A. P. G.; Van Bekkum, H. Recl.
Trav. Chim. Pays Bas 1986, 105, 129.
1
218. doi:10.1139/v90-189.
(
4) Martel, A.; Mahdavi, B.; Lessard, J.; M e´ nard, H.; Brossard,
L. Can. J. Chem. 1997, 75 (12), 1862. doi:10.1139/v97-619.
5) Jaksic, M. M. Electrochim. Acta 2000, 19, 4085.
23) Koeberg-Telder, A.; Lambrechts, H. J. A.; Cerfontain, H.
(
(
6) Laplante, F.; Brossard, L.; M e´ nard, H. Can. J. Chem. 2003,
Recl. Trav. Chim. Pays Bas 1983, 102, 293.
24) Feiring, A. E. J. Org. Chem. 1977, 42 (20), 3255. doi:10.
1021/jo00440a012.
(25) Mador, I. L.; Rosan, A. M.; Crissey, R. K. J. Catal. 1984, 87
(1), 276. doi:10.1016/0021-9517(84)90187-8.
(26) Breitner, E.; Roginski, E.; Rylander, P. N. J. Org. Chem.
1959, 24 (12), 1855. doi:10.1021/jo01094a004.
(27) Wendt, H. In Electrochemical Hydrogen Technologies; Else-
vier, Amsterdam, 1990.
8
1 (3), 258. doi:10.1139/v03-027.
(
(
(
7) St-Pierre, G.; Chagnes, A.; Bouchard, N.-A.; Harvey, P. D.;
Brossard, L.; M e´ nard, H. Langmuir 2004, 20 (15), 6365.
doi:10.1021/la048977v. PMID:15248724.
8) Cirtiu, C. M.; Hassani, H. O.; Bouchard, N. A.; Rowntree, P.
A.; M e´ nard, H. Langmuir 2006, 22 (14), 6414. doi:10.1021/
la0519002. PMID:16800708.
9) Cirtiu, C. M.; Brisach-Wittmeyer, A.; M e´ nard, H. J. Catal.
2
007, 245 (1), 191. doi:10.1016/j.jcat.2006.10.010.
10) Cheong, A. K.; Bolduc, Y.; Lessard, J. Can. J. Chem. 1993,
1 (11), 1850. doi:10.1139/v93-232.
11) Dub e´ , P.; Kerdouss, F.; Laplante, F.; Proulx, P.; Brossard,
(28) Enyo, M. In Comprehensive Treatise of Elctrochemistry;
Conway, B. E.; Bockris, J. O’M., Yeager, E., and White, R.
E., Eds.; Plenum Press: New York, 1983.
(
(
7
(29) Kirilyus, I. V.; Zhuk, M. A. Sov. Electrochem. 1972, 8, 967.
Published by NRC Research Press