The role of alkali modifiers (Li, Na, K, Cs) in activity of 2%Pd/Al 2O3 catalysts for 2-ethyl-9,10-anthraquione hydrogenation

Article abstract of DOI:10.1016/j.apcata.2011.05.036

Present research concentrates on the role of alkali modifiers (Li, Na, K, Cs) in activity of 2%Pd/Al₂O₃ catalyst for 2-ethyl-9,10-anthraquinone (eAQ) hydrogenation. The catalysts with various content of alkali modifier (Me/Pd molar ratio ranges from 0.5 up to 4, Me-alkali metal) were prepared by impregnation of pre-reduced 2%Pd/Al₂O ₃ catalyst with appropriate alkali metal carbonates. The XPS, EDS and TEM measurements show that alkali promoters are introduced into alumina matrix. The microcalorimertic experiments of CO adsorption prove the interaction of CO with catalysts leading to stronger bonding of carbon monoxide by alkali doped catalysts. The presence of alkali promoters in Pd/Al₂O₃ catalyst plays an essential role in the whole eAQ hydrogenation process. The nature of alkali promoter and its content (Me/Pd atomic ratio) in catalyst are of importance. As the alkalinity of promoter increases going from Li to Cs all the effects caused by their presence become stronger. In the presence of alkali doped catalysts the content of 2-ethyloxoanthrone (OXO, isomer of 2-ethyl-9,10-anthrahydroquinone) formed is higher than that on un-doped 2%Pd/Al₂O₃. On the other hand, reactions in the "deep hydrogenation" stage comprising the formation of 2-ethyl-5,6,7,8- tetrahydro-9,10-anthraquinone (H₄eAQ) and the transformation of OXO to 2-ethylanthrone and other degradation products are remarkably inhibited. In particular, the formation of 2-ethylanthrone via hydrogenolysis of OXO isomer is strongly suppressed. The Cs-doped catalyst exhibits the highest activity to OXO among all the catalysts tested whereas the ability of Cs-doped catalysts to the formation of anthrone is most effectively inhibited. The role of alkali modifiers is considered to be associated with stronger interactions between the catalyst and quinone reagents, and in particular OXO isomer. Moreover, in the reagent adsorption the centres of support nearby the palladium particles may also participate by affecting the mode of reagents adsorption.

Full text of DOI:10.1016/j.apcata.2011.05.036

Applied Catalysis A: General 402 (2011) 121–131
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
The role of alkali modifiers (Li, Na, K, Cs) in activity of 2%Pd/Al2O3 catalysts for
2-ethyl-9,10-anthraquione hydrogenation
R. Kosydar, A. Drelinkiewicz∗, E. Lalik, J. Gurgul
Institute of Catalysis and Surface Chemistry, Polish Academy of Sciences, 30-238 Kraków, Niezapominajek 8, Poland
a r t i c l e i n f o
a b s t r a c t
Article history:
Present research concentrates on the role of alkali modifiers (Li, Na, K, Cs) in activity of 2%Pd/Al2O3 cat-
alyst for 2-ethyl-9,10-anthraquinone (eAQ) hydrogenation. The catalysts with various content of alkali
modifier (Me/Pd molar ratio ranges from 0.5 up to 4, Me-alkali metal) were prepared by impregnation
of pre-reduced 2%Pd/Al2O3 catalyst with appropriate alkali metal carbonates. The XPS, EDS and TEM
measurements show that alkali promoters are introduced into alumina matrix. The microcalorimertic
experiments of CO adsorption prove the interaction of CO with catalysts leading to stronger bonding of
carbon monoxide by alkali doped catalysts. The presence of alkali promoters in Pd/Al2O3 catalyst plays
an essential role in the whole eAQ hydrogenation process. The nature of alkali promoter and its content
(Me/Pd atomic ratio) in catalyst are of importance. As the alkalinity of promoter increases going from Li
to Cs all the effects caused by their presence become stronger. In the presence of alkali doped catalysts
the content of 2-ethyloxoanthrone (OXO, isomer of 2-ethyl-9,10-anthrahydroquinone) formed is higher
than that on un-doped 2%Pd/Al2O3. On the other hand, reactions in the “deep hydrogenation” stage com-
prising the formation of 2-ethyl-5,6,7,8-tetrahydro-9,10-anthraquinone (H4eAQ) and the transformation
of OXO to 2-ethylanthrone and other degradation products are remarkably inhibited. In particular, the
formation of 2-ethylanthrone via hydrogenolysis of OXO isomer is strongly suppressed. The Cs-doped
catalyst exhibits the highest activity to OXO among all the catalysts tested whereas the ability of Cs-
doped catalysts to the formation of anthrone is most effectively inhibited. The role of alkali modifiers is
considered to be associated with stronger interactions between the catalyst and quinone reagents, and in
particular OXO isomer. Moreover, in the reagent adsorption the centres of support nearby the palladium
particles may also participate by affecting the mode of reagents adsorption.
Received 22 February 2011
Received in revised form 25 May 2011
Accepted 26 May 2011
Available online 2 June 2011
Keywords:
2-Ethylanthraquinone
Palladium catalysts
Hydrogenation
Alkali modifier
© 2011 Published by Elsevier B.V.
1. Introduction
products”. They are not oxidized to form H2O2 and thus represent
a loss of starting eAQ [3–7]. Reduction of the content of degra-
The catalytic hydrogenation of 2-ethyl-9,10-anthraquinone
(eAQ) into 2-ethyl-9,10-anthrahydroquinone (eAQH2) is one of the
key steps in the industrial synthesis of hydrogen peroxide [1,2].
In this method eAQ is hydrogenated in the presence of supported
palladium catalysts to yield 2-ethyl-9,10-anthrahydroquinone
(eAQH2). Oxidation of eAQH2 results in hydrogen peroxide with the
regeneration of starting eAQ. However, even on the most selective
catalyst, eAQ slowly disappears due to its consumption in various
side reactions yielding number of products termed as “degradation
dation products formed in the conventional route of hydrogen
peroxide synthesis is an important aspect of process improvement,
since the formation of these by-products substantially reduces the
amount of active quinones, eAQ and 2-ethyl-5,6,7,8-tetrahydro-
9,10-anthraquinone (H4eAQ) [6]. It has been demonstrated by Chen
[6] that in industrial practice the selectivity of the desired reaction,
i.e. hydrogenation of quinone to hydroquinone should be above
99%. However, in cyclic hydrogenation/oxidation operations, the
degradation products accumulate in the solution and consequently
a regeneration step has to be involved. This clearly shows that
the contribution of side reactions consuming active quinones must
be strongly limited. Thus, better understanding the course of side
reactions yielding degradation products become one of the priority
issues in developing industrial anthraquinone process [6].
Abbreviations: eAQ, 2-ethyl-9,10-anthraquinone; eAQH2, 2-ethyl-9,10-anthra-
hydroquinone; OXO, 2-ethyl-10-hydroxy-9-anthrone (2-ethyloxanthrone); H4eAQ,
2-ethyl-5,6,7,8-tetrahydro-9,10-anthraquinone; H4eAQH2, 2-ethyl-5,6,7,8-tetra-
hydro-9,10-anthrahydroquinone; H8eAQ, 2-ethyl-1,2,3,4,5,6,7,8-octahydro-9,10-
anthraquinone; eAN, 2-ethylanthrone (2-ethyl-10-anthrone and 2-ethyl-9-
anthrone); H4eAN, 2-ethyl-5,6,7,8-tetrahydroanthrone.
The degradation products are formed in reactions involving car-
bonyl group, such as the hydrogenolytic cleavage of C–O bonds
producing mainly 2-ethylanthrone (eAN). Upon further hydrogena-
tion, eAN is slowly converted to products containing hydrogenated
∗
Corresponding author.
E-mail address: drelinki@chemia.uj.edu.pl (A. Drelinkiewicz).
0926-860X/$ – see front matter © 2011 Published by Elsevier B.V.
doi:10.1016/j.apcata.2011.05.036

122
R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
Table 1
Physicochemical characteristic of catalysts.
Catalyst
BET surface area (m2/g)
Porosity (cm3/g)
Pore size (nm)
Surface concentration (at%) obtained by EDS, average data
As-prepared catalysts Recovered catalysts
Al
Pd
Me
Me/Pd
Al
Pd
Me
Me/Pd
Pd/Al
165
162
0.27
0.27
4.3
4.3
97.4
2.6
Li–Pd/Al
0.5Na–Pd/Al
Na–Pd/Al
K–Pd/Al
Cs–Pd/Al
4Cs–Pd/Al
95.6
95.3
95.5
95.6
88.5
3.3
2.5
2.7
2.6
2.3
1.1
0.33
0.74
0.67
0.69
4
97.1
95.2
96.2
96
2.3
2.7
2.3
2.1
2.4
0.9
2.1
1.5
1.8
9
0.39
0.77
0.65
0.85
3.8
149
0.25
4.3
1.85
1.80
1.80
9.2
154
158
0.25
0.20
3.8
3.8
88.6
aromatic ring (2-ethyl-5,6,7,8-tetrahydroanthrone H4eAN) as well
as to 2-ethylanthracene (eANT) [8–10]. However, apart from the
degradation reactions, a successive saturation of phenyl ring in
eAQH2 to give H4eAQ (active quinone) and 2-ethyl-1,2,3,4,5,6,7,8-
octahydro-9,10-anthraquinone (H8eAQ) takes place. The literature
data showed that the hydrogenolytic C–O reactions are preceded by
the tautomerization of hydroquinone form to 2-ethyl-10-hydroxy-
9-anthrone (2-ethyloxoanthrone) (OXO) (Scheme 1) [1,2,11–13].
In case of anthraquinone molecule, the equilibrium concentra-
tions of enol–keto forms in solutions were reported to depend on
medium properties and, accordingly, acidic conditions favoured
the formation of keto-form [12,13]. Hence, during the consump-
tion of a desired hydroquinone-form a number of consecutive and
parallel reactions can be observed which are known as “deep hydro-
genation” stage. In our continuing work concentrating on variables
determining the course of quinone degradation we focused our
attention on studying the role of alkaline promoters Li, Na, K and
Cs in catalytic performance of Pd/Al2O3. The present studies have
been undertaken because our previous results showed that highly
alkaline medium formed due to high content of Na2CO3 in Pd/SiO2
catalyst together with humidity in the eAQ solution enhanced the
transformation of eAQ via hydrogenolytic reactions which pro-
duced degradation products [14]. On the other hand, a beneficial
effect of modifiers as alkali and alkali earth metals on the behaviour
of Pd-catalysts in terms of catalytic activity, selectivity and resis-
tance to deactivation was observed in various reactions, among
them hydrogenation of ␣, ␤-unsaturated aldehydes (cinnamyl and
crotyl) [15–17], aromatic ketones (acetophenone) [18], phenols
[19–22] as well as alkynes such as 2-butyne-1,4-diol [23]. An elec-
tronic effect, induced by alkali over the metal particles has been
stressed by several authors to explain these activity and selectiv-
ity changes [15]. Accordingly, the alkali promoter tends to donate
electron to the metal which leads to higher charge density of the
palladium particles. The promoting roles of sodium in Pd/C catalyst
observed in hydrogenation of biphenol [22] as well as strontium
in supported Pd-catalysts (Al2O3, SiO2, MgO, hydrotalcite, zeolite)
during hydrogenation of phenol [20] were explained by electronic
effects. The increase in basicity of Pt/CaCO3 catalysts due to doping
with Li, Na, K and Cs-carbonates was reported to be responsible for
the increase in electron density of Pt which resulted in more selec-
tive hydrogenation of C C in 2-butyne-1,4-diol to form olefinic
diol, 2-butene-1,4-diol [23]. An enhanced activity and selectivity
of alkali (Li, Na, K) doped Pt/graphite for hydrogenation of ␣, ␤-
unsaturated aldehydes to alcohols was related to the presence of
electropositive metal which after donation of electrons to platinum
becomes a highly active centre for the activation of C O group thus
facilitating its adsorption [15–17]. Cinnamaldehyde was adsorbed
on the modifier surface via donation of a lone pair of electrons from
the oxygen atom [17].
The role of alkali promoters (Ca, K, Cs) in hydrogenation of phe-
nol to cyclohexanone on Pd/Al2O3 catalyst was explained by the
modification of acid–base properties of alumina support due to par-
tial neutralization of acid sites and generation of basic sites [19,21].
The authors concluded that phenol was adsorbed on the surface of
the support through the oxygen of the O–H group as a phenolate
species and high basicity of the support stabilized the phenolate
forms. Reaction occurred mainly between such phenolate species
and hydrogen activated on the Pd-sites [19,21].
In the present work 2%Pd/Al2O3 catalysts modified by alkali Li,
Na, K and Cs-promoters were prepared by impregnation technique.
The effect of the nature of alkali promoter as well as its content in
catalysts was studied. Particular attention was focused on the role
of alkali promoters in the hydrogenation of quinone reduced forms,
namely eAQH2 and its tautomer OXO.
2. Experimental
2.1. Preparation of catalysts
2%Pd/Al2O3 catalyst (abbreviated as Pd/Al) was prepared using
alumina support ␥-Al2O3, Procatalyse, SPH 1515 batch, grain
diameter 50–150 ␮m, the content of Na as Na2O was given by
Manufacturer to be 670 ppm. The support was impregnated with
PdCl2 solution of very low acidity (pH ∼4) prepared using 0.06 M
HCl (PdCl2 concentration 2.3 × 10−3 mol/dm3). The obtained sam-
ple was calcined in air at 450 ◦C for 5 h and reduced with hydrogen
at 300 ◦C for 3 h. The catalyst was stored in a desiccator until use.
It should be stressed that prior to the catalytic experiments, the
catalysts were activated with hydrogen “in situ” in the reactor.
Alkali doped catalysts were prepared by impregnation of
the pre-reduced 2%Pd/Al2O3 catalyst with an aqueous solu-
tion of appropriate alkali metal carbonates (concentration
2.83 × 10−2 mol/dm3), namely Li2CO3, Na2CO3, K2CO3 and Cs2CO3.
The sample of 2%Pd/Al2O3 catalyst (dried overnight at 120 ◦C) was
Scheme 1. Tautomerization equilibrium of hydroquinone–oxoanthrone forms.

R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
123
added into solution of alkali metal carbonates and impregnation
2.4. Hydrogenation experiments
was carried out at temperature of 80–90 ◦C up to complete evap-
oration of liquid. The obtained alkali doped catalysts were dried
overnight at 120 ◦C. In typical procedure the catalysts of Me/Pd
molar ratios of ca. 0.7 were obtained (Table 1). By changing the
volume of alkali carbonate solutions, the catalysts having their
respective ratios of Na/Pd and Cs/Pd other than 0.7 were also pre-
pared. The alkali doped catalysts of typical Me/Pd molar ratio of ca.
0.7 are abbreviated as Li–Pd/Al, Na–Pd/Al, K–Pd/Al and Cs–Pd/Al.
For catalysts having a different content of alkali promoter the
ratio of Me/Pd is given before the name (4Cs–Pd/Al corresponds
to Cs/Pd ∼ 4).
Hydrogenation of eAQ was performed in agitated glass reac-
tor at atmospheric pressure of hydrogen and temperature of 55 ◦C
following a previously described procedure [9,14]. A mixture of
p-xylene and 2,6-dimethylheptane-4-ol (diisobutyl carbinol) with
the volume ratio of 8:2 was used as the solvent system. Before the
hydrogenation experiment the catalyst was wetted with reagent
solution and was activated “in situ” inside the reactor by passing
through the reactor at first nitrogen gas (15 min) and then hydro-
gen gas (20 min at room temperature and then for next 15 min at
55 ◦C). Following this procedure, the reaction was started by con-
tacting the catalyst with the hydrogenated solution. The progress
of hydrogenation was followed by measuring hydrogen uptake as
a function of reaction time. In a typical hydrogenation experiment
0.45 g of catalyst in 30 cm3 of solution (initial eAQ concentration
20 g/dm3) was used and the reaction was carried out up to reaching
a consumption of ca. 2–3 mol of H2 per 1 mol of anthraquinone ini-
tially present in the reactor. Reproducibility of the hydrogenation
experiments was ca. 5% over 2–3 catalysts batches.
2.2. Characterization of catalysts
The specific surface areas of samples were calculated from the
nitrogen adsorption–desorption isotherms at 77 K in an Autosorb-
1, Quantachrome equipment. Prior to the measurements, the
samples were heated at 120 ◦C for 16 h.
The X-ray photoelectron spectroscopy (XPS) measurements
were carried out with a hemispherical analyzer (SES R4000, Gam-
madata Scienta). The unmonochromatized Al K␣ (1486.6 eV) X-ray
source with the anode operating at 11 kV and 17 mA current emis-
sion was applied to generate core excitation. All binding energy
values were charge-corrected to the carbon C 1s excitation which
was set at 285.0 eV. The samples were pressed into indium foil
and mounted on a special holder. All spectra were collected at
pass energy of 100 eV except survey scans which were collected
at pass energy of 200 eV. Intensities were estimated by calculat-
ing the integral of each peak, after subtraction of the Shirley-type
background, and fitting the experimental curve with a combina-
tion of Gaussian and Lorentzian curves of variable proportions
(70:30).
In the course of reaction the samples of solution were taken
at appropriate time intervals and the composition of solution was
analysed by HPLC following the analytical methods described in our
previous papers [9,14]. The concentration of reagents eAQ, H4eAQ,
eAN (2-ethylanthrone) and OXO at reaction time t (number of moles
at time t: {nt(eAQ)}, {nt(H4eAQ)}, {nt(eAN)} and {nt(OXO)}) was
analysed by HPLC in a re-oxidized solution.
The conversion of eAQ (K, %), selectivity to H4eAQ (S-Tetra, %)
and selectivity to anthrone (S-eAN, %) were calculated according to
the equations
nt(eAQ)
K =
× 100%
(1)
no(eAQ) − nt(eAQ)
where no(eAQ), initial concentration of eAQ
Field emission scanning electron microscope JEOL JSM – 7500 F
equipped with the X-ray energy dispersive (EDS) system was used
for the X-ray (EDS) measurements of surface concentration (at%) of
Al, Pd and alkali promoters Na, K, Cs in catalysts.
nt(H4eAQ)
S-Tetra =
S-eAN =
× 100%
× 100%
(2)
(3)
no(eAQ) − nt(eAQ)
nt(eAN)
no(eAQ) − nt(eAQ)
TEM and STEM studies were performed on FEI Tecnai G2 trans-
mission electron microscope operating at 200 kV equipped with
EDAX EDX and HAADF/STEM detectors. Samples for analysis were
prepared by placing a drop of the suspension of sample in ethanol
or THF onto a carbon-coated copper grid, followed by evaporating
the solvent.
Compounds termed “degradation products” were analysed also
by GC–MS method [9]. Reproducibility of the chromatographic
analysis was ca. 5%.
3. Results and discussion
2.3. Microcalorimetry of CO adsorption
The textural properties of studied catalysts are collected in
Table 1. The specific surface area and porosity of initial un-doped
2%Pd/Al2O3 catalyst only slightly decrease after modification by
alkali promoters. Likewise, the average pore diameter only slightly
decreases from 4.3 nm in the un-doped Pd/Al to 3.8 nm in the
Cs–Pd/Al catalyst. It indicates that the original pore structure of
2%Pd/Al2O3 catalyst remains largely unaffected by doping with typ-
ical amount of alkaline promoters (Me/Pd ∼ 0.7). The X-ray energy
dispersive measurements (EDS) showed the presence of Al, Pd and
alkaline promoters whereas no Cl was observed. The surface con-
centrations (average data, in at%) of Al, Pd and alkaline promoters,
Na, K, Cs determined by EDS are collected in Table 1. The measure-
ments were performed for fresh catalysts and for selected samples
removed from the reactor after the hydrogenation experiment. No
sodium was detected by the EDS in the un-doped Pd/Al catalyst,
which is in agreement with the composition of alumina support
given by the manufacturer (see Section 2). The lithium content
could not be determined by the EDS technique in the Li–Pd/Al cat-
alyst because of too low sensitivity of the method for this element.
The Me/Pd atomic ratios of “typical” catalysts, namely Na–Pd/Al,
The experiments were performed using the Microscale gas flow-
through microcalorimeter. A mixture of 3% CO in diluent nitrogen
was used for all experiments. The amount of catalyst (ca. 0.16 g,
of catalyst, corresponds to ca. 30 ␮mol of Pd) was such to fill the
microcalorimetric cell of ca. 0.15 cm3. The flow-through measure-
ments of CO adsorption were carried out at RT and atmospheric
pressure, with the flow rate of gas mixture 3 cm3/min. The cata-
lyst was first heated at 105 ◦C (16 h) in the flow of nitrogen and
then subjected to an adsorption/desorption cycle of hydrogen at
the same temperature in order to remove the surface oxygen. Fol-
lowing this pre-treatment, the sample was cooled down in the flow
of nitrogen and the system was equilibrated thermally at RT. The
adsorption measurement was initiated by replacing the nitrogen
gas by the mixture of N2 with 3% CO. The calorimetric curve was
recorded and concurrently the uptake of CO was measured using a
down-stream thermoconductivity detector (TCD). Calibration was
made “in situ” for each measurement using a Teflon-encapsulated
electric coil located axially and surrounded by the sample within
the microcalorimetric cell.

124
R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
Fig. 2. STEM micrograph of un-doped 2%Pd/Al2O3 catalyst.
the surface concentrations of Cs at different areas on the catalyst
surface. This effect was even stronger at 4-times higher loading of
Cs (catalyst 4Cs–Pd/Al), in which case a few analysed areas showed
the concentration of Cs markedly higher than a majority of other
analysed places (Fig. 1 and Table 2). Thus, at the high loading of Cs
(Cs/Pd ∼ 3–4), an accumulation of some of Cs-species occurs and,
as a result, a number of areas enriched with cesium appear on the
catalyst surface.
The EDS data for the recovered catalysts (Table 1) show that
the composition of catalysts in general and in particular the Me/Pd
atomic ratio is preserved after the hydrogenation experiments. This
was observed within the whole range of Me/Pd ratios in the cata-
lysts.
The micrographs of the un-doped Pd/Al and the cesium
doped 4Cs–Pd/Al catalysts obtained by the HAADF STEM registra-
tion mode of transmission electron microscope are displayed in
Figs. 2 and 3, respectively. This registration mode is sensitive to Z-
number of elements and provides the micrographs in which the Pd
particles are bright spots whereas the support is black field. More-
over, electron diffraction as well as EDS analyses can be performed
using this registration mode.
Fig. 1. SEM micrographs of Cs–Pd/Al and 4Cs–Pd/Al catalysts with marked areas
where EDS analysis was performed.
K–Pd/Al and Cs–Pd/Al are practically the same, within the range
0.67–0.74. At this loading of alkaline promoter, its distribution
throughout the catalyst grains was practically homogenous in all
studied catalysts. As an example the data obtained for Cs–Pd/Al cat-
alyst are provided showing almost constant surface concentration
of Cs in all areas analysed by the EDS technique (Fig. 1 and Table 2).
Doping of initial Pd/Al catalyst with a content of Cs2CO3 3-times
higher than that in “typical” procedure resulted in certain non-
homogeneity of Cs distribution evidenced by visible differences in
In the micrographs of both catalysts a number of Pd-particles
of size within wide range, 2–10 nm is observed. The individual Pd-
particles as well as their aggregates appear in both catalysts and
no essential difference in the Pd-particles state can be observed. In
order to better recognize the location of alkali promoter the EDS
analysis was performed along the line marked in Fig. 3. It can be
observed that the bright spots represent the areas enriched with Pd
(Pd particles) whereas cesium is distributed throughout the whole
catalyst. It indicates that similarly to what was observed for alu-
mina supported catalysts [19,21], cesium appears to be introduced
into the alumina support. This conclusion is also supported by the
results of XPS studies (Figs. 4–6). The XPS spectra of Pd 3d, Al 2p
and C 1s were registered for un-doped Pd/Al and two alkali doped
Li–Pd/Al and Cs–Pd/Al catalysts. In the spectrum of un-doped Pd/Al
catalyst only one state of Al at binding energy of 74.5 eV appears
which is in agreement with the energy of 74.7 eV reported in the
literature for Pd/␥-Al2O3 sample [24]. On the other hand, the spec-
tra of Li and Cs-doped catalysts both show two states of Al-species.
Apart from the Al 2p energy of 74.5 eV, a small contribution of new
Al state of higher binding energy can be observed (Fig. 4). It should
be stressed that binding energy may change depending on the near-
Table 2
Surface concentration of Al, Cs and Pd (in at%) determined by EDS technique for
Cs–Pd/Al and 4Cs–Pd/Al catalysts.
Spectrum
Cs–Pd/Al
4Cs–Pd/Al
Al
Cs
Pd
Cs/Pd
Al
Cs
Pd
Cs/Pd
1
2
3
4
5
6
7
8
9
95.9
95.7
97.3
96.2
96.1
96.5
95.9
96.5
95.9
96.3
1.4
1.5
1.4
1.6
1.8
1.5
1.9
1.7
2.0
1.6
2.7
2.8
1.3
2.1
2.1
2.0
2.2
1.8
2.1
2.1
0.52
0.54
1.1
0.76
0.86
0.75
0.86
0.94
0.95
0.76
83.7
86.5
92.6
91.0
93.1
85.5
92.9
91.0
89.7
15.7
12.1
6.3
7.2
5.1
14.1
4.4
8.5
0.6
1.3
1.2
1.9
1.8
0.4
2.7
0.5
0.8
26.2
9.30
5.25
3.79
2.83
35.2
1.63
17.0
11.9
9.5
10

R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
125
74.7
Cs-Pd/Al
Al 2p
76.6
78.8
Li-Pd/Al
75.1
74.3
71.7
74.5
Pd/Al
82
80
78
76
74
72
70
68
Binding Energy (eV)
Fig. 4. XPS Al 2p spectra of un-doped Pd/Al, Li–Pd/Al and Cs-Pd/Al catalysts.
Cs-Pd/Al
C 1s
285.0
286.7
289.9
Fig. 3. STEM micrograph of cesium doped 4Cs–Pd/Al catalyst and the distribution
of Pd and Cs along the marked line determined by EDS analysis.
est environment of the atom because this energy is determined by
the electron density. In fact, higher values of Al 2p binding energy
were reported for number of compounds, for example MgAl2O4
76.3 eV, Al(OH)3 75.7–75.9 eV, AlO(OH) 76.7 eV [24]. Thus, the for-
mation of a new Al state in the present catalysts can be explained
as a result of interaction between lithium or cesium-species and
alumina support. This observation is consistent with the results
reported by Qi et al. [25] who showed that the addition of alkali
modifier affected the acid–base properties of the alumina support.
This was result of strong interaction between the modifier and the
vacant site of alumina support [25].
Li-Pd/Al
285.0
285.0
289.4
286.5
Pd/Al
In the XPS C 1s spectra of all three un-doped and alkali-doped
catalysts a strong peak at binding energy of 285 eV dominates. This
carbon energy is commonly detected in samples subjected to the
XPS technique and is ascribed to the presence of carbon contam-
inations. However, in the spectra of alkali-doped catalysts (Fig. 5)
a new carbon state at energy of 289.2 eV is observed. This energy
288.4
2−
is commonly ascribed to the carbon in carbonate CO3 ions [24].
This shows incorporation of lithium/cesium carbonates into alu-
mina support. However, the presence of other Li, Cs-species, like
hydroxide cannot be excluded because of using aqueous solutions
of alkali carbonates. On the other hand, no remarkable changes in
the Pd state are observed in the Cs, Li-doped catalysts (Fig. 6). In
292
290
288
286
284
282
Binding Energy (eV)
Fig. 5. XPS C 1s spectra of un-doped Pd/Al, Li–Pd/Al and Cs–Pd/Al catalysts.

126
R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
lysts since the CO uptake of the respective samples is comparable
Cs-Pd/Al
Li-Pd/Al
Pd/Al
Pd 3d
335.1
to that of the parent, un-doped Pd/Al sample. The same is also true
for the material doped with Cs in lower proportion, 0.5Cs–Pd/Al. In
contrast, however, both the Cs-doped materials show remarkably
higher molar heats of CO adsorption, 196 kJ/mol and 202 kJ/mol.
This seems to suggest that no less than two different effects of Cs
may be in operationin the so modifiedmaterials, separatelyrespon-
sible for increasing of the CO uptake, on one hand, and for increasing
of the heat of adsorption, on the other. The higher molar heats of
adsorption may indicate a stronger bonding of CO with palladium
in the alkali doped catalysts, and especially strong in case of the
most alkaline Cs promoter. This shows that the Cs may affect the
electronic state of palladium stronger than the other alkaline mod-
ifiers can do. It also shows that the effect of increasing the molar
heat is independent of the Cs/Pd ratio in the catalyst, suggesting
that here the effect of promoter is somehow delocalized, i.e. the
cesium is likely to change the electron density of Pd-particles using
the alumina support as a mediator rather than by involving any
kind of direct Pd–Cs interactions. Such direct interactions, on the
other hand, seem more likely to be responsible for the effect of
increasing of the uptake of CO, since it is more strongly dependent
on the amount of the Cs promoter added. Hence, our results are in
agreement with literature data presenting the interaction of alkali
promoters with both, alumina support and Pd-particles [19,25,26].
336.2
336.6
335.3
335.0
336.1
3.1. Hydrogenation experiments
In typical analytical procedure commonly used in eAQ hydro-
genation experiments, samples of the reaction mixture taken from
the reactor were oxidized by air and extracted with water before
chromatographic (GC, HPLC) analysis. The hydroquinone form (QH)
is thermodynamically unstable and when contacted with air it
is spontaneously and quickly oxidized forming hydrogen perox-
ide and restoring quinone (eAQ). Analogous reactivity is exhibited
by tetra-form, although a longer time (ca. 15–30 min) is required
for the quinone form to be restored completely. Thus, in the re-
oxidized sample of reaction mixture used for chromatographic
analysis the hydroquinone forms were absent and only the cor-
responding quinone forms were present. The OXO-isomer which is
present in the reaction mixture (under hydrogen atmosphere) upon
contacting with air isomerizes slowly to the hydroquinone form
(QH) [10,11,27]. Our previous results showed that the isomeriza-
tion of OXO to QH is relatively slow reaction and the concentration
of OXO practically did not change during the time period of ca. 3 h
from the moment of sample contacting with air [27]. Longer time
of contacting with air resulted in slow decrease of OXO content
whereas increased that of eAQ. This was associated with the iso-
merization of OXO to hydroquinone followed by its oxidation to
the quinone form. Therefore, in order to determine the content of
OXO (before it isomerizes to QH), all samples of reaction mixtures
were analysed immediately after withdrawing from the reactor.
The literature data and our previous results revealed crucial
role of OXO in the formation of degradation products [1,27]. The
OXO isomer can be considered as the precursor for these prod-
ucts because the conversion of OXO by the hydrogenolysis reaction
produces anthrone (eAN) which can be further transformed via
aromatic ring hydrogenation to give other degradation products
like (H4eAN). Moreover, the hydrogenolysis reaction of eAN yields
the next degradation products such as 2-ethylanthracene and its
derivatives [1,8,9].
346 344 342 340 338 336 334 332 330
Binding Energy (eV)
Fig. 6. XPS Pd 3d spectra of un-doped Pd/Al, Li–Pd/Al and Cs–Pd/Al catalysts.
the spectra of all three, un-doped and Li-, Cs-doped catalysts the
Pd 3d5/2 peak component at binding energy of 335 eV characteris-
tic of Pd-metal is the dominating [24]. The second palladium state
observed in the spectra of all three catalysts is the one represented
by small and highly energetic peak at ca. 336.1 eV. This energy is
commonly ascribed to the Pd␦+ state which may be formed due to
partial (surface) oxidation of Pd-metal particles.
Table 3 presents the microcalorimetric experiments of CO
adsorption that provided data concerning the uptake of CO (␮mol/g
catalyst) by the catalyst samples, making it possible to determine
the molar heats of adsorption of CO (kJ/mol CO). All the experiments
used a comparable sample mass of ca. 0.16 g to fill the calorimetric
cell, so that each sample contained ca. 30 ␮mol of Pd. Moreover,
for all the catalysts studied, the adsorption of CO turned out to be
totally irreversible at RT, as subsequent exposure of the samples to
pure N2 carrier yielded only a negligible desorption effect in each
case.
Table 3 shows a relatively higher uptake of CO for two
samples, K–Pd/Al and Cs–Pd/Al, respectively, 76.4 ␮mol/g and
81.1 ␮mol/g. For other samples the uptake ranges from 63.4 ␮mol/g
to 66.3 ␮mol/g. This indicates that doping with the Li and Na cations
does not change the CO sorption capability of such modified cata-
Table 3
Microcalorimetric results of adsorption of CO.
Catalyst
CO uptake
Molar heat of CO
(␮mol/g catalyst)
uptake (kJ/mol CO)
The hydrogen uptake curves obtained in the whole hydrogena-
tion test performed at high concentration of catalysts (15 g/dm3)
are displayed in Fig. 7a, whereas the curves restricted to the
initial stage of experiments are shown in Fig. 7b. In the initial
stage of reaction featuring the reduction of eAQ to hydroquinone
(quinone–hydroquinone stage) the rate of hydrogen uptake is very
Pd/Al
64.9
66.3
63.9
76.4
63.4
81.1
155
176
157
181
196
202
Li–Pd/Al
Na–Pd/Al
K–Pd/Al
0.5Cs–Pd/Al
Cs–Pd/Al

R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
127
600
450
300
150
200
(a)
eAQ
150
Pd/Al
Na-Pd/Al
K-Pd/Al
100
Pd/Al
Li-Pd/Al
K-Pd/Al
Na-Pd/Al
OXO
50
0
Na-Pd/Al
K-Pd/Al
Pd/Al
Cs-Pd/Al
0
0
100
200
Time [min]
300
400
0
50
100
150
Time [min]
200
250
300
350
300
Fig. 8. The change of eAQ and OXO content in hydrogenation experiment performed
(b)
at conditions: catalyst concentration 1 g/dm3, eAQ concentration 20 g/dm3.
The results obtained by XPS technique (Figs. 4–6) as well as
CO chemisorption experiments (Table 3) reveal the interaction of
alkali-promoters with alumina matrix. Moreover, the promoters
affected to some extent the electronic state of palladium. These
effects may lead to some changes in the adsorption ability of
Pd-centres towards reagents, among them carbonyl groups, like
200
100
0
C
O, C–OH. Therefore, it may be expected that enhanced reactivity
towards quinone on alkali-doped catalysts would result in the for-
mation of higher amount of H2O2. In order to verify this effect, the
hydrogenation experiments were carried out under the same reac-
tion conditions (catalyst concentration 1 g/dm3) and the content of
H2O2 formed after 50 min of reaction, i.e. the time corresponding
to ca. 70% of quinone reduction was determined in oxidized solu-
tion (by titration method). The experiments were performed using
un-doped Pd/Al and two cesium doped Cs–Pd/Al and 4Cs–Pd/Al cat-
alysts. In the presence of all studied catalysts practically the same
amount of H2O2 was determined, although the rate of hydrogen
uptake was higher in the presence of alkali-doped catalysts. Thus,
the observed effect may be associated with hydrogen spillover on
the supports which is facilitated in the presence of alkali promoters.
According to the literature observations among numerous variables
affecting the effect of hydrogen spillover, electronic properties of
the support exhibit an important role [28,29]. The effect of hydro-
gen spillover in the hydrogenation of eAQ seems to be interesting
and further work is in progress in order to find a role of other
experimental parameters.
Pd/Al
Na-Pd/Al
K-Pd/Al
Cs-Pd/Al
0
10
20
30
Time [min]
Fig. 7. Hydrogen uptake curves obtained in hydrogenation experiment performed
at conditions: catalyst concentration 15 g/dm3, eAQ concentration 20 g/dm3.
high and after the consumption of ca. first equivalent of hydrogen
a break on the hydrogen uptake curves appears thus evidencing
almost complete hydrogenation of quinone to hydroquinone. Fur-
ther hydrogenation occurs at a relatively low rate and is termed as
“deep hydrogenation”.
Two series of hydrogenation experiments were performed. In
the series I experiments, the main reaction stage, e.g. hydrogena-
tion of eAQ to hydroquinone was examined. The catalytic tests were
carried out using very low catalysts concentration 1 g/dm3. In the
series II experiments, the hydrogenation tests were carried out at
high concentration of catalysts (15 g/dm3) and the course of “deep
hydrogenation” stage was examined.
As shown in Fig. 7b on all alkali-doped catalysts the rate
of hydrogen uptake in the first quinone–hydroquinone stage is
higher compared to that on the un-doped Pd/Al catalyst. This may
be considered to be the result of enhanced reactivity of alkali-
doped catalysts for the C O hydrogenation, similarly to what
was observed in hydrogenation of various carbonyl compounds
[15–17]. The presence of electropositive metal (Li, Na, K) was
reported to facilitate the activation of the carbonyl C O group via
interaction of lone pair of C O electrons with alkali modifiers.
However, in the quinone–hydroquinone stage, apart from the
desired hydroquinone (QH) product, its isomer, OXO, the pre-
cursor of degradation products is formed. The nature of alkali
promoter and its content (Me/Pd ratio) both affect the for-
mation of OXO-isomer. From the very beginning of the first,
quinone–hydroquinone stage, the content of eAQ continuously and
slowly decreases whereas the content of OXO isomer grows. This
effect is observed at both, low (Fig. 8) as well as at high (Fig. 9)
catalyst concentrations. It can be observed that as the reaction
progresses the content of OXO attains the maximum, which quite
well correlates with a break on the hydrogen uptake curves. The
decrease of eAQ content (number of moles) reflects quantitatively
the increase of OXO content. The traces of H4eAQ were observed
only when the content of OXO reached the maximum i.e. when the
rate of hydrogen uptake started to slow down. This implies that in
“quinone–hydroquinone” stage all the reacted quinone was in the

128
R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
form of OXO-isomer and even if other products were formed their
content was negligible.
200
150
100
50
Pd/Al
Li-Pd/Al
K-Pd/Al
0.5Cs-Pd/Al
Cs-Pd/Al
4-Cs-Pd/Al
As shown in Fig. 9 in the presence of alkali-doped catalysts
the maximum content of OXO is higher than that formed on un-
doped Pd/Al2O3 catalyst. Moreover, the maximum content of OXO
depends on the nature, Cs, K, Na, of alkali-promoters (Figs. 8 and 9).
The maximum content of OXO on K-doped catalyst is higher com-
pared to that produced on Na-doped sample and both are higher
compared to the content of OXO formed on un-doped Pd/Al catalyst
(Fig. 8).
This effect cannot be explained on the basis of keto–enol
tautomerization equilibrium for hydroquinone–OXO which was
observed in the solution of these isomers (Scheme 1) [11–13]. In
view of this scheme, under alkaline conditions the QH–OXO iso-
merization was shifted to the hydroquinone form, whereas acid
conditions favoured the keto-form. An opposite effect is observed
in the present work. It may be speculated that the tautomerization
of hydroquinone molecules adsorbed on the centres of alkali-doped
catalysts may be easier compared to the isomerization in solu-
tion, i.e. under conditions when the isomerization is accomplished
due to the interactions with solvent and polarity of solvents is
parameter determining its extent. However, it cannot be excluded
that alkali promoters enhanced reactivity of catalysts towards the
hydrogenation of quinone directly to the OXO isomer. This reactiv-
ity may be associated with facilitated effect of hydrogen spillover
observed in the presence of alkali promoters. This effect is consis-
tent with the recent results of Gao et al. [30] who stated by “in situ
spectroscopic techniques” that on Pt/Al2O3 catalyst the hydrogena-
tion of C O group in acetophenone to C–OH was accomplished due
to the presence of spilt-over hydrogen.
From the series II experiments the role of alkali promoters in the
“deep hydrogenation” of eAQ can be clearly observed. The results
reported by Petr et al. [8] as well as our previous studies [9] showed
that in the course of “deep hydrogenation” stage the amount of
H4eAQH2 formed was high whereas the amount of other products,
i.e. eAN and in particular the products of the eAN hydrogenation
(H4eAN), was remarkably lower. Analogous effect can be observed
from the present results.
The obtained reagent distribution profiles are displayed in Fig. 9.
It can be seen that doping of catalyst with alkaline promoter results
in strong inhibition of “deep hydrogenation” stage, i.e. the reac-
tions transforming both reduced forms, hydroquinone QH and OXO.
The nature and the content of alkali promoter in the catalyst are
observed to determine the extent of reaction inhibition. This effect
is evidenced by the data in Fig. 9 showing slow decrease of eAQ
content vs. the time of hydrogenation experiment.
In the presence of all catalysts, un-doped Pd/Al2O3 and alkali
doped samples the content of eAQ strongly decreases during
the first few minutes of reaction whereas its further decrease is
definitively slower. The initial strong decrease of eAQ content cor-
responds to the formation of OXO. As observed before (Fig. 8)
the OXO isomer is formed during the first quinone–hydroquinone
stage, reaching the maximum content at the breaking point on
hydrogen consumption curve which corresponds to almost com-
plete reduction of quinone (eAQ) form.
In this figure, the content of eAQ is given, because for the
HPLC analysis re-oxidized solutions were subjected. However, as
described before, the content of oxidized quinone form, eAQ, cor-
responds to that of hydroquinone (QH) form which occurred under
hydrogen atmosphere. Therefore, the slow decrease of eAQ content
in Fig. 9 may be considered as the decrease of hydroquinone (QH)
content during the “deep hydrogenation” stage.
0
0
0
0
100
200
300
400
Time [min]
150
Pd/Al
100
50
Li-Pd/Al
K-Pd/Al
Cs-Pd/Al
4Cs-Pd/Al
0
100
200
300
400
Time [min]
60
40
20
0
Pd/Al
Li-Pd/Al
K-Pd/Al
0.5Cs-Pd/Al
Cs-Pd/Al
4Cs-Pd/Al
100
200
Time [min]
300
400
A complicated shape of eAQ decrease curves on all studied cat-
alysts (Fig. 9) makes difficult the calculation of the conversion
rates. Thus, in order to compare the activity of catalysts, the time
required for 50% conversion during the “deep hydrogenation” stage
Fig. 9. Reagents distribution profile in “deep hydrogenation” experiments. Reaction
conditions: catalyst concentration 15 g/dm3, eAQ concentration 20 g/dm3.

R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
129
Table 4
“Deep hydrogenation” stage of eAQ.
−1
−1
)
Catalyst
Timea (min)
R (H4eAQ) (mol min−1 gcat
)
R (eAN) (mol min−1 gcat
Selectivity at 50% eAQ conversion
S-Tetra (%)
S-eAN (%)
Pd/Al
Li–Pd/Al
76
100
135
146
143
86
1.29
0.78
0.37
0.59
0.54
0.59
0.44
0.20
0.29
0.30
44.6
39.2
42.4
33.7
38
26.9
26
23.5
8.9
16.8
20
17.9
13.6
14.6
8.3
0.5Na–Pd/Al
Na–Pd/Al
K–Pd/Al
0.5Cs–Pd/Al
Cs–Pd/Al
4Cs–Pd/Al
0.25
0.18
160
180
0.10
0.05
11.2
a
Reaction time required for 50% eAQ conversion.
is assumed (Table 4). This time is the shortest for un-doped Pd/Al
catalyst but it remarkably grows in the presence of all alkali-doped
catalysts. This effect clearly demonstrates strong inhibition of QH
conversion in “deep hydrogenation” stage in the presence of alkali-
doped catalysts and the inhibition depends on the nature and the
content of alkali promoter in the catalysts. At typical Me/Pd molar
ratio of ∼0.7, the smallest suppression is observed on Li-doped cata-
lyst (Table 4). The inhibition is much more effective in the presence
of Na, K and Cs-promoters and their content in catalyst plays a role.
As the content of Cs-promoter increases, the activity of catalyst
for the QH conversion decreases. The catalysts doped with various
amounts of Na follow the same trend.
hand, it is generally accepted that the hydrogenolysis reactions
of C–OH to form C–H are catalysed by acid centres. Reducing of
catalysts acidity by doping with alkali promoter (Na, K) resulted
in strong suppression of C–OH hydrogenolysis during hydrogena-
tion of acetophenone [18]. In the present work, as the alkalinity
of the promoter increases going from Li to Cs, the formation of
anthrone is more effectively suppressed. This may be ascribed to
improved alkalinity of studied catalysts due to the alkali promoters.
150
Pd/Al
According to the literature data H4eAQH2 is formed via hydro-
genation of phenyl ring in the hydroquinone QH and the reaction
proceeds according to a dual site mechanism [4,5]. Therefore, the
observed inhibition of QH conversion leads to decrease in the rate
of H4eAQH2 formation and on all alkali-doped catalysts this rate is
lower compared to that on Pd/Al catalyst (Table 4).
Li-Pd/Al
K-Pd/Al
100
0.5Cs-Pd/Al
Cs-Pd/Al
The higher alkalinity of the promoter going from Li to Cs, the rate
of H4eAQH2 formation is lower (Table 4). Furthermore, when the
content of Cs-promoter grows going from 0.5Cs–Pd/Al to 4Cs–Pd/Al
catalysts, the rate of H4eAQH2 formation decreases even more.
Thus, the ability of alkali-doped catalysts towards the hydrogena-
tion of aromatic ring in the hydroquinone form is generally lower
compared to that of the un-doped Pd/Al catalyst, and an extent
of this effect appears to be determined by the nature of alkali-
promoter.
The effect of alkali promoters in the subsequent transforma-
tion of OXO via hydrogenolysis reaction to anthrone product, eAN
can be recognized from Fig. 10. Similarly to what observed in the
series I experiments (Fig. 9), the maximum content of OXO iso-
mer is higher on alkali-doped catalysts compared to the un-doped
Pd/Al catalyst and the OXO content is especially high in the pres-
ence of Cs-doped catalysts. As the “deep hydrogenation” progresses
the content of OXO decreases whereas the anthrone product, eAN,
is formed. It should be noted, that no other products were observed
by chromatographic analysis and the mass balance calculated for
eAQ, OXO, H4eAQ and eAN confirmed this observation.
However, the transformation of OXO to the anthrone product
is slowed down as the result of alkali doping. Except from the Li-
doped catalyst, on all other alkali modified catalysts, the rate of
anthrone formation is lower compared to that on un-doped Pd/Al
catalyst (Table 4). The nature of alkali promoter is of importance
and the formation of anthrone is most effectively inhibited in the
presence of Cs-doped catalysts. The inhibition effect of Cs-promoter
is also a function of the Cs content as it clearly getting more pro-
nounced going from 0.5Cs–Pd/Al to 4Cs/Pd–Al catalysts. Previous
studies revealed that the transformation of OXO to anthrone is
described by a first order reaction kinetics [4,5,27]. Hence, one can
expect that the higher content of OXO on alkali-doped catalysts
should result in higher rate of anthrone formation and conse-
quently the content of anthrone should be higher. On the other
4Cs-Pd/Al
50
0
0
20
40
60
80
100
K [%]
40
30
20
10
0
Pd/Al
Li-Pd/Al
Na-Pd/Al
K-Pd/Al
Cs-Pd/Al
0
20
40
60
80
100
K [%]
Fig. 10. The content of H4eAQ and 2-ethylanthrone vs the conversion of eAQ. Reac-
tion conditions: catalyst concentration 15 g/dm3, eAQ concentration 20 g/dm3.

130
R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
This conclusion is supported by the observation that although on
catalyst with more alkaline Cs promoter the content of OXO is the
highest further conversion of OXO to anthrone is more effectively
suppressed.
the facilitated adsorption of quinone in the carbonyl group-bonded
configuration.
Thus, the enhanced alkalinity of the catalyst due to higher
content of alkali promoter or more alkaline promoter facilitates
the formation of OXO isomer whereas the reactions in “deep
hydrogenation” stage are more and more inhibited. The Cs-doped
catalysts exhibited the highest activity to OXO among all the cat-
alysts tested and their activity in “deep hydrogenation” stage was
most effectively inhibited. The molar heat of CO chemisorption was
determined to be the highest for Cs-doped catalysts thus evidencing
the strongest modification of catalyst properties by the Cs-species.
This modification results in facilitated interaction of OXO with
the catalyst and suppression of reactions in “deep hydrogenation”
stage.
Let us compare the role of alkali promoter in the formation
of H4eAQH2 and anthrone. The data in Table 4 show that rela-
tive to un-doped Pd/Al catalyst, in the presence of alkali-doped
catalysts the formation of anthrone is less inhibited than the for-
mation of H4eAQH2. This is evidenced by stronger decrease of
the rate of H4eAQH2 formation than that of anthrone due to the
presence of alkali promoters. As the result, on alkali-doped cat-
alysts, the selectivity towards H4eAQH2 is lower than on Pd/Al
catalyst. An opposite effect is observed relative to the selectivity to
anthrone, which increases on alkali-doped catalysts. These selec-
tivity results are supported by the relationships in Fig. 10 showing
the content of H4eAQH2 and eAN moles against the conversion
of eAQ, which was calculated according to Eq. (1). This conver-
sion presents the whole amount of reacted eAQ, towards H4eAQ,
OXO and anthrone. At any given conversion of eAQ the content of
H4eAQ is the highest in the presence of un-doped Pd/Al catalyst and
decreases on all alkali-doped catalysts. In particular, the low con-
tent of H4eAQH2 is formed in the presence of Cs-doped catalysts.
For the entire range of eAQ conversions, the Li- and Na-doped cata-
lysts both produced the content of anthrone considerably higher
compared to that on the un-doped Pd/Al catalyst, whereas the
K-doping has very little effect in this respect. In contrast, signifi-
cantly lower content of anthrone is formed on Cs-doped catalyst
(Fig. 10).
In summary, alkali promoters influenced reactivity of Pd/Al2O3
catalyst for the whole eAQ hydrogenation process. This reveals that
acid–base properties of catalyst play an important role in hydro-
genation of eAQ, similarly to what was observed in hydrogenation
of number of carbonyl reactants. The effect of alkalinity has already
been observed in our previous studies showing that high content
of Na2CO3 in 0.5%Pd/SiO2 catalyst enhanced the transformation of
eAQ via hydrogenolytic reactions to form degradation products [4].
However, these studies did not provide data enabling evaluation
the role of alkali-promoters in the course of particular reactions
and especially in the formation and further transformation of OXO-
isomer. Recently, the role of acid–base properties of Pd catalysts
supported on SiO2–Al2O3 with various content of SiO2 in hydro-
genation of eAQ was also studied by Feng et al. [31]. According
to the authors, the acid sites on the surface of catalyst partici-
pated in the hydrogenation of eAQ because they acted as adsorption
sites for eAQ molecules. The adsorbed eAQ molecules were subse-
quently activated and hydrogenated by spilt-over hydrogen species
formed on metal surface. Similar effect involving the participation
of acid–base sites of the supports for adsorption and activation
of phenol molecules was postulated by other researches [19,21].
The acid–base properties of supports exhibited a primary role
because they determined the geometry of phenol molecule adsorp-
tion whereas the reaction occurred between phenol chemisorbed
on the support (phenolate species) nearby the metal particles and
hydrogen chemisorbed on the Pd-sites [19]. In the presence of
Pd catalysts supported on basic MgO, the phenol molecule was
predominantly anchored to the surface through the oxygen atom
forming phenolate form. On more acidic supports like alumina “co-
planar” adsorption of phenol molecule involving O–H and aromatic
ring was involved because of the interaction between benzene ring
and the acid sites of the support. An analogous model was suggested
for the hydrogenation of naphthalene [32]. The acidic sites of the
support acted as adsorption sites for the aromatic compound, and
in the vicinity of Pd particles these adsorbed species reacted with
hydrogen spilt-over from the metal. The enhanced activity of Pd
catalysts with high content of Na2CO3 for the formation of degrada-
tion product observed in our previous work [4] was also related to
4. Conclusions
The presence of alkali (Li, Na, K, Cs) promoters plays an
important role in the performance of Pd/Al2O3 catalysts for eAQ
hydrogenation process. The XPS, EDS and TEM measurements show
that alkali promoters are introduced into alumina matrix. The
microcalorimertic experiments of CO adsorption prove the inter-
action of CO with catalysts leading to stronger bonding of carbon
monoxide by alkali doped catalysts. The alkaline modifiers exhibit
multiple effects on the whole eAQ hydrogenation process. The rate
of the first quinone–hydroquinone stage increases whereas the rate
of reactions in “deep hydrogenation” stage remarkably decreases.
The nature of alkali promoter and its content (Me/Pd atomic ratio)
in catalyst are of importance. As the alkalinity of promoter increases
going from Li to Cs all the effects caused by their presence become
stronger. Enhanced alkalinity of catalysts favours the formation of
OXO, the isomer of hydroquinone. As the consequence, the con-
tent of OXO formed on alkali-doped catalysts is higher compared
to that on un-doped Pd/Al2O3 sample. These effects are ascribed to
stronger interactions between the catalyst and quinone reagents
and may concern especially the OXO molecule. However, in the
reagent adsorption the centres of support nearby the palladium
particles may also participate by affecting the mode of reagents
adsorption. The facilitated interaction of OXO with the catalyst
results in the suppression of “deep hydrogenation” stage. On the
other hand, the conversion of OXO to anthrone is strongly inhib-
ited on alkali doped catalysts. This is ascribed to reduced acidity of
catalysts due to alkaline promoters. Among all catalysts tested, the
Cs-doped catalysts exhibit the highest activity for the formation of
OXO whereas their ability to the formation of anthrone is strongly
inhibited.
References
[1] Ulmann’s Encyclopedia of Industrial Chemistry, vol. 13A, VCH, Weinheim, 1908,
p. 443.
[2] W.M. Weigert, H. Delle, G. Kabish, Chem. Ztg. 99 (1975) 101.
[3] E. Santacesaria, M. Di Serio, A. Russo, U. Leone, R. Velotti, Chem. Eng. Sci. 54
(1999) 2799.
[4] E. Santacesaria, M. Di Serio, R. Velotti, U. Leone, J. Mol. Catal. 94 (1994) 37.
[5] E. Santacesaria, M. Di Serio, R. Velotti, U. Leone, J. Mol. Catal. 99 (1995) 151.
[6] Q. Chen, Chem. Eng. Process. 47 (2008) 787.
[7] R. Halder, A. Lawal, Catal. Today 125 (2007) 48.
ˇ
[8] J. Petr, L. Kurz, Z. Be˘lohlav, L. Cerveny´, Chem. Eng. Process. 43 (2004) 887.
[9] A. Drelinkiewicz, A. Waksmundzka-Gora, J. Mol. Catal. A 246 (2006) 167–175.
[10] S.W. Sjawcillo, W.I. Sawuszkina, E.M. Zhernowskaja, Zh. Obsh. Khim. 28 (1959)
1752.
[11] Houben-Weyl, Methoden der Organishen Chemie, Vol. 7, teil 3c, Georg Thieme
Verlag, Stuttgart, 1979, p. 270.
[12] J. Houben, Das Antracen and die Antrachinone, Georg Thieme Verlag, Leizieg,
1929.
[13] K.H. Meyer, Annalen 379 (1911) 37.
[14] A. Drelinkiewicz, A. Waksmundzka-Góra, J. Mol. Catal. A 258 (2006) 1–9.
[15] P. Maki-Arvela, J. Hajek, T. Salmi, D.Yu. Murzin, Appl. Catal. A 292 (2005) 1–49.
[16] P. Gallezot, D. Richard, Catal. Rev. Sci. Eng. 40 (1998) 81–126.

R. Kosydar et al. / Applied Catalysis A: General 402 (2011) 121–131
131
[17] W. Koo-Amornpattana, J.M. Winterbottom, Catal. Today 66 (2001) 277–287,
Pt/C.
[24] NIST Standard Reference Database 20, Version 3.2 (Web Version), National
Institute of Standards and Technology, Gaithersburg, 2001.
[18] R.V. Malyala, C.V. Rode, M. Arai, S.G. Hedge, R.V. Chaudhari, Appl. Catal. A 193
(2000) 71–86.
[19] G. Neri, A.M. Visco, A. Donato, C. Milone, M. Malentacchi, G. Gubitosa, Appl.
Catal. A 110 (1994) 49–59.
[20] U.R. Pillai, E. Sahle-Demessie, Appl. Catal. A 281 (2005) 31–38.
[21] S. Scire, C. Crisafulli, R. Maggiore, S. Minico, S. Galvagno, Appl. Surf. Sci. 93 (1996)
309–316.
[22] H.-B. Cho, J.-Ch. Lee, Y.-H. Park, Catal. Today 111 (2006) 417–422.
[23] M.M. Telkar, C.V. Rode, V.H. Rane, R.V. Chauhari, Catal. Commun. 6 (2005)
725–730.
[25] C. Qi, T. Bai, A. Lidun, React. Kinet. Catal. Lett. 54 (1995) 131–137.
[26] X.N. Mahata, K.V. Raghavan, V. Vishwanathan, Appl. Catal.
183–187.
A 182 (1999)
[27] R. Kosydar, A. Drelinkiewicz, J.-P. Ghany, Catal. Lett. 139 (2010) 105–113.
[28] B.K. Hodnett, B. Delmon, Stud. Surf. Sci. Catal. 27 (1986) 53–78.
[29] R.T. Braunschweig, F. Roessner, J. Mol. Catal. A 127 (1997) 61–84.
[30] F. Gao, A.D. Allian, H. Zhang, S. Cheng, M. Garland, J. Catal. 241 (2006) 189–199.
[31] J.-T. Feng, H.-Y. Wang, D.G. Evans, X. Duan, D.-Q. Li, Appl. Catal. A 382 (2010)
240–245.
[32] J. Zheng, M. Guo, Ch. Song, Fuel Process. Technol. 89 (2008) 467–474.