Role of phosphorus in the formation of selective palladium catalysts for hydrogenation of alkylanthraquinones

Article abstract of DOI:10.1016/j.apcata.2019.117293

Effective Pd-P catalysts for the hydrogenation of 2-ethyl-9,10-anthraquinone in a toluene/1-octanol medium were obtained. Phosphorus modification increases the selectivity of the active quinones from 69% to 93%–97%, and the deposition of Pd-P catalyst particles on a carbon support increases their activity fivefold due to increased dispersion without a decrease in selectivity to the target product. Using energy dispersive X-ray spectroscopy, X-ray powder diffraction, high-resolution transmission electron microscopy and X-ray photoelectron spectroscopy, the composition, size and state of the surface layers of Pd-P catalysts were determined. Surface enrichment by the electron-deficient palladium and structural disorder of Pd-P catalyst particles were shown to account for the increase in selectivity to active quinones during hydrogen peroxide production by the anthraquinone method.

Full text of DOI:10.1016/j.apcata.2019.117293

AppliedCatalysisA,General589(2020)117293
Contents lists available at ScienceDirect
Applied Catalysis A, General
journal homepage: www.elsevier.com/locate/apcata
Role of phosphorus in the formation of selective palladium catalysts for
hydrogenation of alkylanthraquinones
⁎
Lyudmila B. Belykha, , Nikita I. Skripova, Tatyana P. Sterenchuka, Vladlen V. Akimovb,
Vladimir L. Tausonb, Tatyana A. Savanovicha, Fedor K. Schmidta
a Department of Chemistry, Irkutsk State University, 1 K. Marx str., 664003 Irkutsk, Russia
b A.P. Vinogradov Institute of Geochemistry SB RAS, 1a Favorsky str., 664033 Irkutsk, Russia
A R T I C L E I N F O
A B S T R A C T
Keywords:
Effective Pd-P catalysts for the hydrogenation of 2-ethyl-9,10-anthraquinone in a toluene/1-octanol medium
were obtained. Phosphorus modification increases the selectivity of the active quinones from 69% to 93%–97%,
and the deposition of Pd-P catalyst particles on a carbon support increases their activity fivefold due to increased
dispersion without a decrease in selectivity to the target product. Using energy dispersive X-ray spectroscopy, X-
ray powder diffraction, high-resolution transmission electron microscopy and X-ray photoelectron spectroscopy,
the composition, size and state of the surface layers of Pd-P catalysts were determined. Surface enrichment by
the electron-deficient palladium and structural disorder of Pd-P catalyst particles were shown to account for the
increase in selectivity to active quinones during hydrogen peroxide production by the anthraquinone method.
2-ethyl-9,10-anthraquinone
Hydrogenation
Palladium phosphorus catalysts
XPS
1. Introduction
Among the four methods of H2O2 production (electrolysis of sul-
phuric acid to peroxodisulphuric acid and its hydrolysis to H2O2 and
The theoretical and practical problems faced during the production
of hydrogen peroxide (H2O2) by the anthraquinone method have
prompted continuous research interest in developing new and im-
proving presently known catalysts for this method [1]. H2O2 is an im-
portant chemical compound widely used in textile (cotton and wood
fabric bleaching), chemical (reagents), pulp and paper (wood pulp
bleaching) and food (sterilised packaging of milk and fruit juice and
antiparasitic treatment of salmon) industries, in wastewater treatment
and in silicon wafer preparation for the manufacture of printed circuit
sulphuric acid; synthesis of H2O2 from hydrogen and oxygen; liquid-
phase oxidation of 2-propanol with oxygen from air; sequential hy-
drogenation and oxidation of alkylanthraquinones), the anthraquinone
method is prevalent [2,3]. The global industrial H2O2 production by
this method reaches 99%.
The anthraquinone method involves catalytic hydrogenation of al-
kylanthraquinones (2-ethyl-, 2-tert-butyl- and 2-pentylanthraquinones)
to alkylanthrahydroquinones and their subsequent mild oxidation with
oxygen in the air.
(1)
boards [2]. The application of H2O2 in the large-scale production of
propylene oxide led to the second significant expansion of the market
for H2O2 consumption (2008) [2].
The selective catalytic hydrogenation of alkylanthraquinones to al-
kylanthrahydroquinones is key stage of H2O2 production by this
method. However, the hydrogenation of alkylanthraquinones is a
⁎ Corresponding Author.
E-mail address: belykh@chem.isu.ru (L.B. Belykh).
https://doi.org/10.1016/j.apcata.2019.117293
Received 3 July 2019; Received in revised form 6 September 2019; Accepted 7 October 2019
Availableonline12October2019
0926-860X/©2019ElsevierB.V.Allrightsreserved.

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
complex series–parallel reaction, in which, along with reduction of the
carbonyl group in 2-ethyl-9,10-anthraquinone (eAQ), side reactions
occur [4]. These include, initially, saturation of aromatic rings of 2-
ethyl-9,10-anthrahydroquinone (eAQH2) to 2-ethyl-5,6,7,8-tetrahydro-
9,10-anthrahydroquinone (H4eAQH2), 2-ethyl-1,2,3,4-tetrahydro-9,10-
anthrahydroquinone (iso-H4eAQH2), and 2-ethyl-1,2,3,4,5,6,7,8-octa-
hydro-9,10-anthrahydroquinone (H8eAQH2) (eAQH2 → H4eAQH2 (iso-
H4eAQH2) → H8eAQH2) and the hydrogenolysis of the CeOH bond in
eAQH2 along with anthrone formation [5]. On an industrial scale, to
reduce the contribution of side processes, catalytic hydrogenation of
alkylanthraquinone is not performed until its complete transformation.
Alkylanthraquinone conversion does not usually exceed 60% [1].
Therefore, developing targeted approaches to synthesising highly se-
lective hydrogenation catalysts that reduce only the carbonyl group of
alkylanthraquinones without accelerating hydrogenolysis of the CeO
bond in alkylanthrahydroquinones or affecting the aromatic ring re-
mains an urgent problem [1]. Overcoming this is also essential for other
practically important processes: synthesis of drugs (vitamins A, B, E and
K), biologically active additives (dietary supplements) and inter-
mediates for the perfume industry, including the semihydrogenation
stages of acetylene alcohols, alpha-beta unsaturated carbonyl com-
pounds (aldehydes and ketones) and quinones [6,7].
2.2. General hydrogenation procedure
2.2.1. Hydrogenation of eAQ and product analysis
Hydrogenation of eAQ was performed in a thermostated ‘duck’ type
glass vessel at 50 °C and an initial hydrogen pressure of 2 atm in the
presence of in situ formed palladium catalysts. A solution of Pd(acac)2
(0.0152 g, 5 × 10−5 mol) in a mixture of aromatic hydrocarbon/oc-
tanol (ratio 7:10) solvents was reduced by hydrogen at 90 °C at 2 atm in
the presence of white phosphorus at different P:Pd ratios (P:Pd = 0, 0.3,
0.7 and 1.0) for 30–45 min until the quantitative transformation of Pd
(acac)2. The resulting black-brown suspension or ‘solution’ was cooled
to 50 °C under hydrogen pressure of 2 atm; subsequently, 3 ml of eAQ
solution in toluene (2.116 mmol, 0.4993 g) was injected using a syr-
inge. Hydrogenation was performed with intensive mixing, excluding
the reaction in the diffusion region. After absorption of 1.0, 1.2 or
1.4 mol H2⋅(eAQ mol)−1, samples were collected for analysis. To this
end, an aliquot (5 mL) of the reaction mixture was filtered from the
catalyst in an argon atmosphere. The aliquot of filtrate (5 mL) was
taken and oxidised with oxygen in the air for 15–20 min. Subsequently,
30 ml of water was added and H2O2 was extracted from the organic
layer (5 mL) into water (30 mL) by stirring. The amount of H2O2 ex-
tracted was determined by titrating an aliquot (10 mL) of the pre-
acidified water layer with KMnO4 (0.1 N). The KMnO4 concentration
was controlled using a primary standard of oxalic acid or ammonium
oxalate.
Most studies in this area focus on palladium catalysts. Palladium is
characterised by its relatively low reactivity in the hydrogenation of
aromatic rings and hydrogenolysis of the CeOH bond. Depending on
the carrier nature and texture, the modifiers and the palladium particle
size, the selectivity of the deposited palladium catalysts varies from
50% to 93%–94% [8–10]. An increase in catalyst selectivity is usually
accompanied by a decrease in their activity [11].
Hydrogenation of eAQ is performed in two stages: during the first
stage, at a high speed, hydrogenation of eAQ to eAQH2 is predominant,
while during the second stage, products of the conversion of eAQH2 are
formed: H4eAQH2, H8eAQH2 and 2-ethyl-9(10)-anthrones at lower
speed. Insignificant amounts of transformation products of eAQH2 are
formed during the first stage of hydrogenation. Therefore, the reaction
rate r (molH2 min−1) at each stage was determined from the kinetic
curve (the dependence of the amount of hydrogen absorbed on the
reaction time). The rate of hydrogenation of eAQ to eAQH2 (r1) was
calculated based on the tangent of the angle of inclination of the
straight sections of the kinetic curves in the absorption interval 0.1–0.5
eq. of hydrogen. The rate of conversion of eAQH2 (r2) was calculated
using the slope ratio of straight-line sections of the kinetic curves in the
interval of absorption 1.1–1.2 eq. of hydrogen. Since Pd-P catalysts are
not individual substances, the catalytic activity (A) was used to com-
pare their properties in eAQ hydrogenation. The activity of Pd-P cata-
lysts in the first and second sections of the kinetic curve was calculated
as the ratio of the reaction rate (r1 or r2) to the initial concentration of
Pd(acac)2.
The traditional way to regulate the catalyst properties is by con-
trolling surface properties (acidic/basic) and the structure of the carrier
pores [12], and by modification [13–16]. The addition of ammonium
formate or ammonium halogenides inhibits hydrogenation of the aro-
matic ring of eAQH2; the introduction of small amounts of amines ac-
celerates the oxidation of H4eAQH2; the addition of activated alumi-
nium oxide accelerates the dehydrogenation of H4eAQH2 [1].
For the first time, we have discovered that elemental phosphorus
(P4) promotes the selectivity of palladium catalysts in the hydrogena-
tion of eAQ [17]. The H2O2 yield increased from 69% (for the palla-
dium–carbon (Pd/C) catalyst) to 96%–98% (for the palladium–pho-
sphorus (Pd-P) catalyst, P:Pd = 0.3). However, phosphorus
modification at the reduction stage of Pd(acac)2 by hydrogen reduced
the catalytic activity of Pd-P containing particles by an order of mag-
nitude in comparison with that of Pd-black [17]. To develop active and
selective hydrogenation catalysts and determine the causes of phos-
phorus modification effect, the nature and properties of colloidal so-
lutions of Pd-P particles and those of Pd-P particles deposited on a
carbon support were studied.
The compositions of the transformation products of eAQ were
analysed by GLC on a chromatograph (Chromatec-Crystal 5000.2; ca-
pillary column length: 30 m; phase: poly(5% diphenyl
/ 95%
dimethylpolysiloxane)–(BPX-5); flame ionisation detector) under the
following temperature programming conditions: 160 °C (3 min), 270 °C
(20 min) and heating rate of 40 °C min−1. In parallel, intermediates and
2. Experimental
reaction products were identified using
a GCMS-QP2010 Ultra
2.1. General procedures
Shimadzu gas chromatography-mass spectrometer (capillary column
GsBP⋅5MS, length: 30 m, phase: poly(5% diphenyl / 95% dimethyl
polysilphenylenesiloxane)).
The solvents (benzene, toluene, xylene, 1-octanol and 2-octanol)
were purified by standard methods [18]. Pd(acac)2 was synthesised
according to the reported procedure [19]. White phosphorus was me-
chanically purified from surface oxidation products immediately before
use and washed in anhydrous benzene. A solution of white phosphorus
in benzene was prepared and stored under an inert atmosphere in a
2.2.2. Catalyst recycling
The stability of the Pd-P/C catalyst in eAQ hydrogenation was stu-
died in a 170 ml batch thermostated glass reactor (internal diameter
40 mm) at 50 °C, 2 bar hydrogen pressure and vigorous stirring using a
magnetic stirrer (1000 rpm). The Pd-P/C catalyst (0.603 g, 2.87 × 10−5
mol Pd) was loaded into the reactor in an argon flow, purged with
Schlenk flask. 31P nuclear magnetic resonance (NMR) data:
=−522 ppm (s).
δ
2

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
hydrogen for 15 min, then 32 ml of 1-octanol and 32 ml of eAQ toluene
solution (2 g, 8.474 mmol eAQ) were added. The reactor was sealed and
a hydrogen pressure of 2 bar was created. After uptake of 1.2 mol H2
(mol Pd)−1, the hydrogenation of eAQ was stopped. A clear emerald-
colored solution was carefully decanted into a Schlenk flask in argon,
and an aliquot of the solution (10 ml) was used to determine the yield of
hydrogen peroxide. The catalyst in the reactor was washed with a
mixture of toluene/1-octanol, dried in vacuum (50 °C/2 Torr) and
tested in the hydrogenation of the next portion of eAQ without adding
fresh catalyst. After five cycles of eAQ hydrogenation, the catalyst was
washed twice with a mixture of toluene/1-octanol, then with toluene
and dried in vacuum (50 °C/2 Torr). Data from ICP analysis after hy-
drogenation: Pd – 0.54 wt. %, P – 0.057 wt. %.
occurred. When the reaction was complete, the resulting catalytic
system was cooled to room temperature and transferred in an inert
atmosphere to a Schlenk flask. The liquid above the sediment was
decanted, and the precipitate was sequentially washed with benzene
and diethyl ether in an argon atmosphere and dried under vacuum
(50 °C/1 torr). The yield was 0.2420 g. Data from ICP analysis: Pd –
0.46 wt. %, P – 0.66 wt. %.
2.3. Characterization techniques
UV spectra were recorded on an SF-2000 spectrophotometer
(Russia) in quartz cuvettes having an absorbing layer thickness of
0.1 cm. The conversion of Pd(acac)2 was controlled using the absorp-
tion band at 330 nm (ε330 = 10,630 L⋅ cm−1⋅ mol−1).
2.2.3. Preparation of the catalysts
Additionally, the products of the interaction of Pd(acac)2 with ele-
mental phosphorus were analysed on a gas chromatography-mass
spectrometer (GCMS-QP2010 Ultra Shimadzu, capillary column
GsBP⋅5MS, length: 30 m, phase: poly(5% dephenyl/95% dimethyl
polysilphenylenesiloxane). Ionisation was initiated by electron impact,
with an ionisation energy of 70 eV. The obtained mass spectra were
compared with the reported data (comparison libraries, Wiley, NIST,
NIST05).
Preparation of the Pd-P catalyst for the TEM. The Pd-P catalyst for the
transmission electron microscopy (TEM) study was prepared in situ as
done when studying its properties in the hydrogenation of eAQ. A 1 ml
of phosphorus solution in benzene (1.5 × 10–5 mol relative to atomic
phosphorus) was added dropwise to a solution of Pd(acac)2 (0.0152 g,
5 × 10–5 mol) in 7 ml of toluene, placed in a thermostated reaction
vessel under hydrogen flow and then stirred at room temperature for
5 min. Subsequently, 1-octanol (10 mL) was added, the temperature
was increased to 90 °C and the catalyst was formed with intensive
stirring of the reaction mixture in hydrogen for 30–45 min until quan-
titative transformation of Pd(acac)2 occurred. The transformation of Pd
(acac)2 was controlled by UV spectroscopy based on the absorption
band at 330 nm (ε330 = 10,630 L⋅ cm−1⋅ mol−1). The resulting black-
brown colloidal solution was cooled to 30 °C. A drop of Pd-P catalyst
solution was applied to a carbonised copper grid (200 mesh) and dried
at room temperature in a box under an inert atmosphere.
The Pd-P catalyst was analysed by XRD using a XRD-7000 S dif-
fractometer (Shimadzu Co., Japan) (CuKα radiation, Ni filter,
=1.5418 Å).
λ
TEM images were obtained using an electron microscope (Tecnai
G2, FEI, USA) with an accelerating voltage of 200 kV. The images were
recorded using a CCD camera (Soft Imaging System, Germany). Local
elemental analysis was performed using energy dispersive X-ray spec-
troscopy (EDX, Phoenix) using a Si (Li) detector. The particle structure
parameters in the images parameters were measured using iTEM 5.0
and DigitalMicrographs 1.94.1613 software. The analysis of structure
periodicity and image filtration were analysed using Fourier methods,
i.e. fast Fourier transformation and inverse fast Fourier transformation.
The average size was determined by processing an area containing at
least 100 particles.
Preparation of the Pd-P catalyst for the XPS and XRD. Preparation of
Pd-P catalyst sample for X-ray photoelectron spectroscopy (XPS) and X-
ray powder diffraction (XRD) studies: To a solution of Pd(acac)2
(0.4566 g, 1.55 × 10−3 mol) in 68 ml of toluene in a thermostated re-
action vessel under hydrogen, 3.1 ml of phosphorus solution
(4.5 × 10−4 mol, calculated relative to atomic phosphorus) in benzene
was added dropwise and stirred for 5 min at room temperature.
Subsequently, 1-octanol (44.5 mL) was added, the temperature was
increased to 90 °C and the catalyst was formed by intensive stirring of
the reaction mixture in hydrogen for 30–45 min until quantitative
transformation of Pd(acac)2 occurred. When the reaction was complete,
the resulting catalytic system was cooled to room temperature and
transferred in an inert atmosphere to a Schlenk flask. The solvents were
then distilled under vacuum (2/3 of the volume), and diethyl ether was
added to form a black precipitate. This was washed sequentially with
benzene and diethyl ether in an argon atmosphere and dried under
vacuum (50 °C/1 torr). The yield was 0.1090 g. Data from ICP analysis:
Pd–90.32 wt. %, P – 8.27 wt. %.
X-ray photoelectron spectra were obtained using a photoelectron
spectrometer SPECS (SPECS, Germany) equipped with a PHOIBOS 150
MCD
9 energy analyser, using monochromatised AlKα radiation
(1,486.74 eV). A survey spectrum was recorded with a step of 1 eV at an
analyser transmission energy of 20 eV, and high-resolution spectra
(narrow scans) in steps of 0.1 eV and a transmission energy of 10 eV.
The C1s carbon line (285.0 eV) was recorded for spectrum calibration.
Non-uniform charging was compensated by flood the sample with low-
energy electrons. The samples were etched for 2 min by Ar+ ions using
a PU-IQE112/38 scanning ion gun (‘SPECS’) at an accelerating voltage
of 2.5 kV and an ion current of 20 μA, which provided an etching rate of
∼1 nm min−1. The experimental data were processed using the
CasaXPS program. The spin–orbit splitting of doublet lines Pd 3d5/2–3/2
and P 2p3/2–1/2 was approximated by two Lorentz–Gaussian curves with
doublet separations of 5.26 and 0.84 eV, respectively, at an area ratio of
3/2 for Pd 3d5/2–3/2 and 2/1 for P 2p3/2–1/2. The full width at half
maximum (FWHM) of spectral lines was determined according to the
reported data for similar substances and conditions for obtaining
spectra ≤1.5 eV (up to 1.2 eV for phosphorus). It was presumed that if
the FWHM exceeds 1.5 eV (1.2 eV for phosphorus), these lines should be
approximated by several curves corresponding to different chemical
forms of the element. Since the oxygen 1s spectral line heavily overlaps
with the 3p3/2 palladium line, the number of chemical forms of palla-
dium and their line positions were coordinated with the second doublet
line Pd 3p1/2 in the survey spectrum.
Preparation of the Pd-P/C catalyst. Preparation of Pd-P/C catalyst
sample for XPS and XRD studies: Toluene (10 mL) and 1-octanol
(10 mL) were added to a mixture of Pd(acac)2 (0.0152 g, 5.0 × 10−5
mol) and coal (0.2607 g, Sibunit type, fraction 0.2–0.4 mm, S = 450
m2 g−1) in a thermostated reaction vessel, in a hydrogen flow. The
reaction mixture was stirred at room temperature for 60 min.
Subsequently, 0.37 ml of phosphorus (1.5 × 10−5 mol, calculated re-
lative to atomic phosphorus) solution in benzene under hydrogen was
added dropwise to the reaction vessel and stirred for 10 min at room
temperature. The temperature was then increased to 90 °C, and the
catalyst was formed by intensive stirring of the reaction mixture in
hydrogen for 45 min until quantitative transformation of Pd(acac)2
3

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
ICP analysis. The catalysts samples were decomposed in acidic so-
Table 2
Effect of the catalytic system composition on the palladium catalyst properties
in the hydrogenation of 2-ethyl-9,10-anthraquinone: P:Pd = 0.3, CPd
lutions under heating in accordance with the procedure described in
[20]. The solutions were analyzed on the Agilent 7500ce unit manu-
factured by Agilent Technologies with quadrupol mass analyzer (Agi-
lent Tech., Santa Clara, CA, USA). Measurements were performed under
the following optimal conditions: plasma power 1550 W, reflected
power 1 W, the carrying gas flow rate 0.8 L/min, auxiliary gas flow
0.13 L min−1, integration time 0.1 s, points on mass – 3, oxides (CeO/
Ce, 156/140) = 0.3%. The device was tuned by Tuning solution with
the concentration of 10 μg L−1 (10 ppb) Li, Co, Y, Ce, Tl. The calcula-
tion of elements concentrations was carried out by external calibration
method using multi-element certified solutions 68A-A, 68A-C (High-
purity Standards) containing all of the elements from Li to U. The ca-
libration solutions were diluted with 2% HNO3 down to the con-
centration 50–200 μg L−1 (50, 100 and 200 ppb) of each element. The
drift tracking of the device was carried out by internal standard In
(10 ppb). Calculation of the concentrations was based on the most
abundant isotopes, giving signals, free of overlapping or subject to
minimum isobar and polyatomic disturbances. Elements detection
limits are at the level of 1 ppt (Pd) and 50 ppb (P), determination errors
are 5 and 20%, accordingly.
=2.5 mmol L–1
;
νeAQ =2.116 mmol; solvent
–
toluene/1-octanol
(10 mL:10 mL), T =50 °C, P(H2) = 2 atm.
№
Catalyst
ν(Pd)×105, mol Pd, wt.
Yield (H2O2),
%
Aa, min–1
%
1
Pd(acac)2–H2
5.0
–
69c
40д
96c
70b
65c
77b
50c
61b
64c
79b
87c
87b
93c
78b
94c
12.6
2
3
Pd(acac)2–0.3P–H2 5.0
–
1.2
8.0
Pd/C
5.0
2
4
5
6
7
8
Pd/C
1.25
0.62
5.0
0.5
8.6
1.6
4.0
7.0
11.0
Pd/C
0.25
2.0
Pd-P/C
Pd-P/C
Pd-P/C
1.25
0.62
0.5
0.25
а A – activity in the initial reaction stage (mol H2 (mol eAQ min)–1
;
b,c,d - yield of
H2O2 was determined after absorption of 1 (b), 1.2 (c), 1.4 (d) mol H2(mol
eAQ)–1, respectively.
3. Results and discussion
As noted above, modification of palladium hydrogenation catalysts
with elemental phosphorus increases the H2O2 yield obtained by the
anthraquinone method from 69% to 96%–98% [17]. However, colloidal
solutions of Pd-P catalyst are almost an order of magnitude less active
in eAQ hydrogenation than Pd black formed during reduction of Pd
(acac)2 by hydrogen in the absence of white phosphorus. To optimise
the conditions of eAQ hydrogenation catalysed by colloidal solutions of
Pd-P particles, various reaction parameters were varied. The nature of
the aromatic hydrocarbon (benzene, toluene and xylene) used as a
solvent and higher alcohols (1-octanol and 2-octanol) did not affect the
H2O2 yield (Table 1). As the temperature increased from 30 °C to 80 °C,
the release of H2O2 in the presence of Pd-P catalyst (P:Pd = 0.3) de-
creased from 97% to 86% (Table 1). An active and selective catalyst for
eAQ hydrogenation was obtained by Pd-P catalyst formation in the
presence of a carbon support. The catalytic properties of Pd-P/C cata-
lyst are presented in Table 2.
catalysts (P:Pd = 0.3) depended on the palladium content on the
carbon carrier. As the palladium content on the carbon varied from 2 to
0.5 wt. %, the activity of the Pd-P/C catalysts (P:Pd = 0.3) doubled
(from 4 min−1 for the 2 wt. % Pd-P/C catalyst to 7–11 min−1 for the
0.25–0.5 wt. % Pd-P/C catalyst) without reducing the H2O2 yield
(Table 2). The Pd-P/C catalyst (0.25–0.5 wt. %) formed in hydrogen is
comparable in activity with the 2 wt. % Pd/C catalyst and is almost
comparable in activity with heterogeneous catalysts described in the
literature (for Pd/Al2O3, Pd/SiO2 and Pd/C, A = 12–16 min−1 at T
=62 °C and PH2 = 1 atm [21]). Hydrogenation of eAQ in the presence
of a 0.5 or 1 wt. % Pd-P/C catalyst (P:Pd = 0.3) ensures a H2O2 yield
reaching 93%–94% (Table 2).
To establish the stability of the Pd-P/C catalyst, five portions of the
substrate were sequentially hydrogenated in a batch reactor. The
The activity of the Pd-P/C catalyst (P:Pd = 0.3) in eAQ hydro-
genation increases fivefold on average in comparison with colloidal
solutions of Pd-P particles (Table 2). The initial activity of the Pd-P/C
Table 1
Effect of reaction conditions on the properties of the Pd-P catalyst (P: Pd = 0.3)
in the hydrogenation of 2-ethyl-9,10-anthraquinone: CPd =2.5 mmol L–1
;
[eAQ]/[Pd] = 42.3; solvent – toluene/1-octanol (10 mL:10 mL), P(H2) = 2 atm.
а
b
Solvent
T, °C
Yield H2O2
A, min–1
A1:A2
%
g H2O2⋅(g Pd)–1
A1
A2
benzene– 1-octanol
xylene – 1-octanol
toluene – 2-octanol
toluene – 1-octanol
toluene – 1-octanol
toluene – 1-octanol
30
30
30
30
50
80
94
92
98
97
96
86
88.3
86.5
92.1
91.2
90.2
80.8
0.93
1.28
0.63
0.40
1.17
1.80
0.25
0.20
0.08
0.28
0.11
0.17
3.7
6.4
7.8
1.4
10.6
10.6
a
– after the absorption of 1.2 mol H2 (mol eAQ)–1
;
b – A1, A2 – activities in the
Fig. 1. Yield of H2O2 as a result of eAQ hydrogenation and oxidation of active
initial reaction stage and after absorption of 1 mol H2 (mol eAQ)–1, respectively.
quinones: Recycling of the Pd-P/C catalyst.
4

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
experiments were carried out at 50 °C and a hydrogen pressure of 2 bar,
without adding fresh catalyst to the reaction system. As Fig. 1 illus-
trates, the Pd-P/C catalyst is relatively stable after five eAQ hydro-
genation cycles: the yield of hydrogen peroxide has remained almost
constant.
are observed in the electron diffraction pattern (Fig. 2a, inset), as well
as significant broadening of diffraction maxima in the XRD curve
(Fig. 2c). The CDS calculated using the Selyakov–Scherrer formula is
1.6 nm. The broadening of the diffraction peak may be due to both the
small size of the CDS and crystal lattice defects. A large half-width of
the diffraction peak at 71° (d/n =1.3241 Å) in the diffraction curve of
the Pd-P catalyst in comparison with the half-width of the peak at
39.59° (d/n =2.2765 Å) indicates distortion of the palladium crystal
structure. For palladium crystallites, d/n =2.2458 Å (111) and 1.37537
(220) Å (# 00-046-1043).
To determine the causes of the modification effect of phosphorus on
the activity and selectivity of palladium catalysts in eAQ hydrogena-
tion, their physical characteristics were studied.
Based on TEM, colloidal solutions of Pd-P particles and Pd-P/C
catalysts containing different palladium mass contents have differ cat-
alyst particle size. In the Pd(acac)2-0.3P-H2 system, high-contrast par-
Similar results were obtained for the Pd-P catalyst formed in the Pd
(acac)2-nP-H2 system at the ratio P:Pd = 1.0 (solvent: toluene/1-oc-
tanol). Pd-P catalyst particles (P:Pd = 1.0) having a diameter of
25–70 nm are composed of smaller particles having a diameter of
ticles with an average diameter of 34.4
13.7 nm were formed
(Fig. 2a). The mean coherent domain size (CDS) of the Pd-P catalyst
particles does not exceed 1.5 nm (Fig. 2b). The structural disorder of the
Pd-P catalyst particles (P:Pd = 0.3) is verified by the electron diffrac-
tion data (Fig. 2a, inset) and XRD (Fig. 2c, inset). Diffuse Debye rings
7–10 nm (Fig. 3a). The average particle diameter is 36.4
21.6 nm.
On the diffraction curve of the Pd-P catalyst (P:Pd = 1.0), two
Fig. 2. TEM (a), HRTEM (b) images, EDX spectrum (c) and XRD pattern (c, inset) of the Pd-P catalyst (P:Pd = 0.3) (the dotted lines indicate the reflexes for
crystalline palladium).
5

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
Fig. 3. TEM (a), HRTEM (b) images, EDX spectrum (c) and XRD pattern (c, inset) of the Pd-P catalyst (P:Pd = 1.0) (the dotted lines indicate the reflexes for
crystalline palladium).
broadened maxima with d/n =2.2533 Å and 1.3547 Å, are observed;
the half-width of the peak at 69° (d/n =1.3547 Å) also predominates
over the half-width of the peak at 40.01° (d/n =2.2533 Å). The CSD
calculated by the Selyakov–Scherrer formula is 1.8 nm (Fig. 3с, inset).
Crystallinity disturbance in the palladium catalyst formed in the
presence of white phosphorus is due to incorporation of phosphorus
(metalloid) into the particles. EDX analysis of selected areas of a Pd-P
catalyst sample (P:Pd = 0.3) indicate that it has a homogeneous com-
position; the P:Pd ratio for different sections is 0.23:1 (Fig. 2c). For the
Pd-P catalyst (P:Pd = 1.0), the P:Pd ratio for different sections varies
from 0.32:1 to 0.2:1 (Fig. 3c).
of elemental phosphorus, distortion of the palladium crystal structure
and the formation of structurally disordered Pd-P solid solutions. As the
P:Pd ratio increases, the size of the base particles decreases and the
degree of polydispersity of the Pd-P catalyst increases due to diffuse
aggregation of the catalyst particles (Figs. 2 and 3).
For the Pd-P/C catalyst (0.5 wt. %, P:Pd = 0.3), the average particle
size is 3.6
0.9 nm (Fig. S1, SI2). For the Pd-P/C catalyst (2.0 wt. %,
P:Pd = 0.3), the average particle size (6.5
smaller than that of colloidal solutions of Pd-P particles
1.3 nm) is five times
(34.4
13.7 nm) (Figs. 2a, 4a ). Particle size comparison confirms that
the main reason for the increased activity of the Pd-P/C catalyst in eAQ
hydrogenation in comparison with colloidal solutions of Pd-P particles
is the increased dispersion. Based on the high-resolution transmission
Thus, the common features of Pd-P catalyst particles formed in to-
luene/1-octanol solution at different P:Pd ratios are the incorporation
6

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
Fig. 4. TEM (a, c) and HRTEM (b, d) images of the 2 wt. % Pd-P/C catalyst (P:Pd = 0.3) (a, b) and 0.5 wt. % Pd/C catalyst (c, d).
electron microscopy (HRTEM) data, Pd-P/C catalyst particles are also
structurally disordered. The CDS of the Pd-P catalyst particles does not
exceed 0.6–1.6 nm (Fig. 4b).
(Pdδ+) in palladium phosphides, solid palladium solutions with phos-
phorus (Eb(Pd 3d5/2) =336.2 eV) [25,26] and small palladium clusters
on the carbon support (Eb(Pd 3d5/2) =336.1 eV) [27].
Due to the low palladium content in the Pd-P/C catalyst (0.5 wt. %;
P:Pd = 0.3) and structural disorder of particles, XRD proved to be un-
informative (Fig. S2, SI2). The diffraction curve of the 0.5 wt. % Pd-P/C
catalyst (P:Pd = 0.3) showed only the reflexes of the carbon support (d/
n = 3.5447, 2.0907, 1.7217 and 1.2137 Å) with relative intensities (I/
I0) 100, 13, 2 and 2, respectively.
The P 2p X-ray spectrum contains a peak with a binding energy of
130.5 eV (Fig. 5c). The position of the P 2р3/2 level at 130.5 eV is
characteristic of the phosphorus state in metal phosphides (Eb(P 2р3/2
)
=
128.5–130 eV) [28], elemental phosphorus (Eb(P 2р3/2) =
129.9–130.4 eV) [29,30] and phosphorus in metal alloys (Eb(P 2р3/2) =
130.1 eV). Considering the electron microscopy data (Fig. 2b, c) and P
2p XPS spectrum (Fig. 5c), the Pd 3d5/2 binding energy of 336.0 eV is
related to palladium in structurally disordered solid solutions. The
molecular formula of the surface compound (Pd19P) does not match the
composition of any known phosphides (Pd15P2, Pd6P, Pd4.8P, Pd3P,
Pd5P2 or PdP2 [31]). The composition of the surface compound (Pd19P)
also indicates the formation of a solid solution between palladium and
phosphorus during Pd-P/C catalyst formation by the low-temperature
method due to phosphorus incorporation in the palladium crystal
structure.
To determine the state of the surface layer, the 0.5 wt. % Pd-P/C
catalyst sample (P:Pd = 0.3) was studied by XPS. The surface compo-
sition of the Pd-P/C catalyst is represented by four elements: Pd, P, C
and O (Fig. 5). The ratios of the contents of the chemical forms of
elements are given in Table 3.
The X-ray photoelectron spectrum of Pd 3d is represented by three
spin-orbit doublets, 3d5/2–3/2, characterising the three chemical states
of palladium (Fig. 5a). The binding energy of the photoelectrons of the
high-intensity component Pd 3d5/2, with a maximum of 335.6 eV, is
close to that of palladium in the reduced state, Pd(0) [22]. The binding
energy of Pd 3d5/2, 338.0 eV, is characteristic of Pd2+ in Pd(acac)2
[23,24]. The Pd(acac)2 precursor content, among three chemical forms
of palladium, does not exceed 15%.
The electron-deficient palladium form (Pdδ+) prevails in the surface
layer of the Pd-P/C catalyst. The surface concentration of the Pdδ+ form
(indicated as PdxP in Table 3) exceeds the Pd(0) concentration by 3.25
times (Table 3). Surface enrichment by the electron-deficient form of
palladium is also characteristic of Pd-P catalyst particles obtained
without a carbon support (Table 4).
The high-intensity component with the binding energy Eb(Pd 3d5/2
)
=336.0 eV corresponds to an electron-deficient form of palladium
7

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
Fig. 5. Pd 3d (a, b) and P 2р (c, d) XPS lines of the Pd-P catalyst before (a, c) and after (b, d) argon etching.
Table 3
Binding energies and concentrations of elements in different states in the surface layer of the Pd-P/C catalyst before argon etching.
Spectral line
Binding energy, eV
FWHM, eV
Element concentration, at %
Most likely chemical species*
Pd 3d5/2
335.6
336.0
338.0
130.5
133.6
135.0
532.1
533.3
534.5
536.1
538.4
284.7
284.9
285.8
287.3
289.1
291.0
0.9
1.1
1.5
0.7
1.2
0.8
1.9
1.7
1.5
2.0
3.0
0.7
1.1
1.5
1.5
1.5
1.5
0.5
0.1
3.4
0.1
0.3
0.1
0.02
0.05
0.03
–
Pd0 (0.20)
PdxP (0.65)
Pd(acac)2 (0.15)
P 2p3/2
O 1s
PdxP (0.17)
2−
R–O–PO3
(0.53)
3−
PO4
(0.30)
Pd 3p3/2 (Pd0, PdxP)
2.2
–
O–C + O–P + C–O–C + O = C (0.65); (AC, R–O–PO32−, R–OH, Pd(acac)2)
Pd 3p3/2 (Pd(acac)2)
0.5
0.7
27.9
38.4
19.2
6.7
3.8
–
C = O (0.15) (AC**)
O2/C (0.20) (AC**)
C 1s
96.0
C = C (0.29) (AC: graphite)
C–C + CH2 + CH3 (0.40) (AC, R–O–PO32−, R–OH, Pd(acac)2)
C–O–P; C–O–C (0.2) (R–O–PO32−, R–OH)
C = O + CO2 (0.07) (АC**, Pd(acac)2)
O-C = O (0.04), (AC**: COOH, COOR)
sat π-π* (AC**)
* The number in parentheses is the atomic fraction of the species.
** AC – Activated carbon.
8

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
Table 4
(Table 3, Fig. 6a). The carbon binding energy of the CeOPe fragment
coincides with that in alcohols and common ethers. The oxygen binding
energy of the O 1s level of the OeP bond (Eb(O 1s) = 533.3 eV) co-
incides with that in ethers (Table 3, Fig. 6c) [34]. Therefore, the for-
mation of octyl esters of phosphoric acid in the reaction system Pd
(acac)2-P in toluene/1-octanol medium was confirmed by additional
gas chromatography-mass spectrometry.
Binding energies and concentrations of elements in different states in the sur-
face layer of the Pd-P catalyst (P:Pd = 0.3).
Binding energy, eV
Pd 3d5/2
P 2p
O 1s
C 1s
а
335.5 (0.23)
335.8 (0.72)
337.4 (0.05)
130.0 (0.35)
532.4
285.0 (0.83)
286.8 (0.12)
289.1 (0.05)
131.1 (0.13)
533.1 (1.0)
b
The presence of phosphorus on the Pd-P/C catalyst surface in the
reduced and oxidised states is due to the following chemical reactions
at the catalyst formation stage. Phosphorus introduced into the system
as a modifier does not only react with palladium(0) atoms produced
during the reaction (2) to form solid solutions of PdxP under mild
conditions.
133.5 (0.18) 134.6 (0.34)
534.5
16.5
Element concentration, at %
13.6
9.2
60.7
[a]- Pd 3p3/2 (Pd0; PdxP); [b] - Pd 3p3/2 (Pd(acac)2).
It should be noted that after Pd-P/C catalyst argon etching (2 min),
the same four elements (Pd, P, C and O) again represent the surface
layer composition. The ratios of the elemental concentrations (at. %)
coincide (Table 5); however, in the near-surface layer, there is no Pd(0)
(Table 5). The composition of the PdxP compound in the near-surface
layer (Pd23.5P) is close to that of the solid solution in the surface layer.
In the surface layer of the Pd-P/C catalyst, phosphorus is pre-
Pd(acac)2 + H2
Pd + 2Hacac
(2)
xPd + P PdxP
(3)
With respect to Pd(acac)2, both phosphorus and hydrogen act as
reducing agents [35] (EHo PO /P = -0.502 V; EPod
= +0.99 V):
2+
o
3
3
/Pd
3Pd(acac)2 + 2 P+ 6H2 O 3Pd + 6Hacac + 2H3PO3
(4)
dominantly present in the oxidised state in the form of pentavalent
Phosphorous acid formed and its salts also possess reducing prop-
3−
tetracoordinated phosphorus: PO4
(Eb(P 2р3/2) = 135.0 eV) and
erties with respect to Pd(acac)2 [36]
(
EHo PO /H PO
4
=
−0.276 V;
3
3
3
2−
R–O–PO3
(Eb(P 2р3/2) = 133.6 eV) [32] (Fig. 5c, Table 3). The
EPod
o= +0.99 V).
2+
presence of the oxidised phosphorus form (PO43−) is primarily asso-
ciated with the oxidation of surface phosphides of transition metals or
Pd-P solid solutions with oxygen at the sample preparation stage [33].
According to literature, phosphorus in alkyl phosphate attributes to one
/Pd
Pd(acac)2 + H3PO3 + H2 O Pd + 2Hacac + H3PO4
(5)
Due to the neutralisation of phosphoric acids with 1-octanol (sol-
vent), phosphoric acid octyl esters are formed.
of the high-intensity components of phosphorus (Eb(P 2р3/2
) =
All the results of the kinetic experiments and physical and chemical
analysis data for Pd-P catalyst particles and Pd-P/C catalyst allow us to
determine the main causes of the effect of modification of phosphorus
on activity and selectivity in eAQ hydrogenation.
133.6 eV). The presence of octyl phosphate on the surface of the Pd-P/C
catalyst cannot be unambiguously confirmed based only on the XPS
data (P 2p3/2, C 1s, O 1s) for several reasons, including the low surface
concentration of phosphorus and overlapping of the photoelectron
spectra of the R–O–P fragments with that of the carbon support.
Particularly, the XPS C 1s spectrum is decomposed into six com-
ponents (Fig. 6a) and the XPS O 1s spectrum is decomposed into five
components (Fig. 6c). The components of the C 1s line correlate with
graphite carbon (Eb(C 1s) = 284.7 eV), aliphatic carbon (Eb(C 1s) =
284.9 eV), carbon in alcohols, esters or CeOPe fragments (Eb(C 1s) =
285.8 eV), carbon in carbonyl groups (Eb(C 1s) = 285.8 eV) and carbon
in carboxyl or complex ester groups (Eb(C 1s) = 289.1 eV) [34]
The formation of palladium catalysts in hydrogen in the presence of
elemental phosphorus under mild conditions leads to the incorporation
of phosphorus into palladium particles, the formation of structurally
disordered solid solutions and the appearance of a partial positive
charge on palladium. The hydrogenation reaction requires activation of
molecular hydrogen and alkylanthraquinone. Pd catalysts having high
electron density on palladium are more active in anthraquinone hy-
drogenation [37,38]. The appearance of a partial positive charge on
palladium in Pd-P catalyst particles weakens the activation of molecular
Table 5
Binding energies and concentrations of elements in different states in the surface layer of the Pd-P/C catalyst after argon etching.
Spectral line
Binding energy, eV
FWHM, eV
Element concentration, at %
Most likely chemical species*
Pd 3d5/2
336.0
337.8
130.0
131.4
134.2
135.8
532.2
533.5
534.2
535.7
538.5
284.7
284.9
286.1
287.5
289.3
291.0
1.109
1.5
0.5
0.1
0.4
0.1
0.01
0.01
0.07
0.01
–
PdxP (0.83)
Pd(acac)2 (0.17)
PdxP (0.09)
P 2p3/2
0.5
0.5
P (0.08)
2−
1.197
0.596
2
R–O–PO3
(0.68)
3−
PO4
(0.15)
O 1s
C 1s
3.4
Pd 3p3/2 (Pd0, PdxP)
2
2.0
–
O–C + O–P + C–O–C + O = C (0.60) (AC**, R–O–PO32−, R–OH, Pd(acac)2)
Pd 3p3/2 (Pd(acac)2)
1.5
2.0
0.6
0.8
38.4
38.4
11.5
4.8
2.9
–
C = O (0.17) (AC**)
3.0
O2/C (0.23) (AC)
0.634
0.989
1.494
1.5
96.0
C = C (0.40) (AC: graphite)
C–C + CH2 + CH3 (0.40) (AC, R–O–PO32−, R–OH, Pd(acac)2)
C–O–P; C–O–C (0.12) (R–O–PO32−; R–OH)
C = O + CO2 (0.05) (АC**, Pd(acac)2)
O-C = O (0.03), (AC: COOH, COOR)
sat π-π* (AC**)
1.5
1.5
* The number in parentheses is the atomic fraction of the species.
** AC – Activated carbon.
9

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
Fig. 6. C 1s (a, b) and O 1s (c, d) XPS lines of the Pd-P/C catalyst before (a, c) and after (b, d) argon etching.
hydrogen and the HeH bond break required for catalytic hydrogenation
[39,40].
content. Second, palladium solid solutions with phosphorus contain
more strongly bound hydrogen [43], which is disadvantageous for
hydrogenation of the aromatic ring [44]. Third, the formation of
structurally disordered solutions is less favourable for adsorption of
alkylanthraquinones by the plane of the aromatic ring for its subsequent
hydrogenation [45]. Based on DFT calculations [45], palladium crys-
tallites facetted by the (111) planes are most favourable for saturation
of alkylanthraquinone aromatic rings.
Thus, the predominance of electron-deficient palladium on the
surface leads to the low activity of colloidal solutions of Pd-P particles
during the hydrogenation reaction. Increased dispersion of the Pd-P/C
catalyst (d = 6.1
1.5 nm; P:Pd = 0.3) in comparison with colloidal
solutions of Pd-P particles (d = 34.4
13.7 nm) (Figs. 2a, 3 a) and the
Pd/C catalyst (d = 14.1 6.1 nm (TEM); 13.5 nm (CDS)) (Fig. S3, SI2)
leads to an increase in its activity in eAQ hydrogenation and partially
compensates for the negative effect of phosphorus on catalytic activity.
Colloidal solutions of Pd-P particles and Pd-P/C catalyst are char-
acterised by their high selectivity for active quinones. At least three
factors define the promoting effect of phosphorus on selectivity. First,
the incorporation of phosphorus into palladium catalyst particles re-
duces the concentration of so called ‘non-selective’ hydrogen in the
near-surface layer [41] favourable for side processes. The solubility of
hydrogen in palladium phosphides or solid solutions of palladium with
phosphorus is lower than that in palladium crystallites [42]. The so-
lubility of hydrogen is inversely correlated with the phosphorus
4. Conclusion
The properties of colloidal solutions of Pd-P particles and Pd-P/C
catalysts in the hydrogenation of eAQ in toluene/1-octanol medium
were studied. The results of catalytic experiments showed that phos-
phorus modification increases the selectivity to active quinones from
69% to 93%–97%. The deposition of Pd-P particles on the carbon
support increases their activity fivefold with no drop in selectivity for
the target product as a result of its increased dispersion. With its high
selectivity, the 0.5 wt. % Pd-P/C catalyst activity is comparable with
10

L.B. Belykh, et al.
AppliedCatalysisA,General589(2020)117293
those of the heterogeneous catalysts described in the literature: Pd/
Al2O3, Pd/SiO2 and Pd/C.
[13] S. Wang, K. Gao, W. Li, J. Zhang, Appl. Catal. A Gen. 531 (2017) 89–95.
[14] D. Ding, X. Xu, P. Tian, X. Liu, J. Xu, Y.-F. Han, Chinese J. Catal. 39 (2018)
673–681.
Using EDX analysis, XRD, HRTEM and XPS, the composition, size
and state of the surface layers of Pd-P catalysts were determined.
During Pd-P catalyst particle formation by the low-temperature method
in toluene/1-octanol medium, structurally disordered solid solutions
are formed due to phosphorus incorporation into the palladium crystal
structure. The surface of Pd-P catalyst particles is enriched with elec-
tron-deficient palladium. Surface enrichment by the electron-deficient
palladium and structural disorder of Pd-P catalyst particles are re-
sponsible for the increase in selectivity for active quinones in H2O2
production by the anthraquinone method.
[15] R. Bi, Q. Wang, C. Miao, J. Feng, D. Li, Catal. Lett. 149 (2019) 1286–1296.
[16] J. Zhang, K. Gao, S. Wang, W. Li, Y. Han, RSC Adv. 7 (2017) 6447–6456.
[17] T.P. Sterenchuk, L.B. Belykh, N.I. Skripov, S.B. Sanzhieva, K.L. Gvozdovskaya,
F.K. Schmidt, Kinet. Catal. 59 (2018) 585–592.
[18] A.J. Gordon, R.A. Ford, Handbook of Practical Data, Techniques, and References,
Wiley, New York, 1972.
[19] J.C. Mollens and L.L. Wood, USA, Patent 3,474,464, 1969.
[20] T.Iu. Zheleznova, I.V. Vlasova, A.V. Shilova, Analitika I control 19 (2015) 363–372,
https://doi.org/10.15826/analitika.2015.19.4.004.
[21] A. Drelinkiewicz, A. Waksmundzka-Gora, J. Mol. Catal. A Chem. 246 (2006) 1–49.
[22] P. Miquel, Y. Yamin, K. Lombaert, C. Dujardin, M. Trentesaux, L. Gengembre,
P. Granger, Surf. Interface Anal. 45 (2010) 585–592.
[23] L.N. Mazalov, S.V. Trubina, N.A. Kryuchkov, O.A. Tarasenko, S.V. Trubin,
G.I. Zharkova, J. Struct. Chem. 48 (2) (2007) 259–268.
[24] D.S. Selishchev, N.S. Kolobov, A.V. Bukhtiyarov, E.Y. Gerasimov, A.I. Gubanov,
D.V. Kozlov, Appl. Catal. B: Environ. 45 (2018) 585–592.
[25] T. Inui (Ed.), Successful Design of Catalysts, Elsevier Science Publishers B. V.,
Amsterdam - Printed in the Netherlands, 1988.
Acknowledgements
This work was supported by the Russian Science Foundation (pro-
ject no. 17-73-10158). XPS spectra of the samples were obtained in the
Krasnoyarsk regional Centre for Collective Use SB RAS. Electronic
images of the samples were obtained in the Baikal center of nano-
technologies. ICP-MS analyses were done at the Common Use Center of
Ultramicroanalysis in Limnological Institute SB RAS with assistance of
A.P. Chebykin.
[26] M. Zhao, Chem. Asian J. 11 (2016) 461–464.
[27] L. Nikoshvili, A. Bykov, T. Khudyakova, T. LaGrange, F. Héroguel, J.S. Luterbacher,
V.G. Matveeva, E.M. Sulman, P.J. Dyson, L. Kiwi-Minsker, Ind. Eng. Chem. Res. 45
(2017) 13219–13227.
[28] A.P. Grosvenor, R.G. Cavell, A. Mar, J. Solid State Chem. 181 (2008) 2549–2555.
[29] P.E.R. Blanchard, A.P. Grosvenor, R.G. Cavell, A. Mar, Chem. Mater. 20 (2008)
7081–7088.
[30] R. Rego, A.M. Ferraria, A.M. Botelho do Rego, M. Cristina Oliveira, Electrochim.
Acta 87 (2013) 73–81.
[31] S. Carenco, Y. Hu, I. Florea, O. Ersen, C. Boissière, C. Sanchez, N. Mézailles, Dalton
Trans. 42 (2013) 12667–12674.
Appendix A. Supplementary data
[32] J. Bedia, J.M. Rosas, J. Rodrıguez-Mirasol, T. Cordero, Appl. Catal. B: Environ. 94
(2010) 8–18 2002.
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.apcata.2019.117293.
[33] H. Sun, J. Xu, G. Fu, X. Mao, L. Zhang, Y. Chen, Y. Zhou, T. Lu, Y. Tang,
Electrochim. Acta 59 (2012) 279–283.
[34] M.A. Puziy, O.I. Poddubnaya, A. Ziatdinov, Appl. Surface Sci. 252 (2006)
8036–8038.
References
[35] L.B. Belykh, N.I. Skripov, T.P. Sterenchuk, S.S. Kolesnikov, F.K. Schmidt, J. Nanosci.
Nanotechnol. 17 (2017) 696–704.
[1] J.M. Campos-Martin, G. Blanco-Brieva, J.L.G. Fierro, Angew. Chem. Int. Ed. 45
(2006) 6962–6984.
[36] R. Prins, M.E. Bussell, Catal Lett. 142 (2012) 1413–1436.
[37] E. Yuan, C. Wu, X. Hou, M. Dou, G. Liu, G. Li, L. Wang, J. Catal. 347 (2017) 1–49.
[38] R. Kosydar, A. Drelinkiewicz, E. Lalik, J. Gurgul, Appl. Catal. A Gen. 402 (2011)
1–49.
[2] R. Ciriminna, L. Albanese, F. Meneguzzo, M. Pagliaro, ChemSusChem. 9 (2016)
3374–3381.
[3] S. Ranganathan, V. Sieber, Catalysts 8 (2018) 1–22.
[4] Y. Han, Z. He, S. Wang, W. Li, J. Zhang, Catal. Sci. Technol. 5 (2015) 2630–2639.
[5] A. Drelinkiewicz, A. Waksmundzka-Gora, J. Mol. Catal. A Chem. 246 (2006)
167–172.
[39] M.J. Maccarrone, C.R. Lederhos, G. Torres, C. Betti, F. Coloma-Pascual,
M.E. Quiroga, J.C. Yori, Appl. Catal. A Gen. 441–442 (2012) 90–98.
[40] R.M. Mironenko, O.B. Belskaya, A.V. Lavrenov, V.A. Likholobov, Kinet. Catal. 59
(2018) 347–354.
[6] M. Crespo-Quesada, A. Yarulin, M. Jin, Y. Xia, L. Kiwi-Minsker, J. Am. Chem. Soc.
45 (2011) 12787–12794.
[41] G. Vilе, D. Albani, N. Almora-Barrios, N. Lоpez, J. Pеrez-Ramirez, ChemCatChem. 8
(2016) 21–33.
[7] D.Y. Murzin, Appl. Catal. A Gen. 292 (2005) 1–49.
[8] A. Biffis, R. Ricoveri, S. Campestrini, M. Kralik, K. Jerabek, B. Corain, Chem. Eur. J.
8 (2002) 2962–2967.
[42] M. Stojewski, J. Kowalska, R. Jurczakowski, J. Phys. Chem. C 113 (2009)
3707–3712.
[43] B.T.B. Flanagan, G.E. Biehl, J.D. Clewley, S. Kundqvist, Y. Anderson, J.C.S. Faraday
I. 76 (1980) 196–208.
[9] A. Drelinkiewicz, A. Waksmundzka-Gora, J.W. Sobczak, J. Stejskal, Appl. Catal. A
Gen. 45 (2007) 219–228.
[44] B. Liu, M. Qiao, J. Wang, K. Fan, Chem. Commun. (2002) 1236–1239.
[45] Y. Han, Z. He, S. Wang, W. Li, J. Zhang, Catal. Sci. Technol. 5 (2015) 2630–2639.
[10] X. Li, H. Su, G. Ren, S. Wang, J. Braz. Chem. Soc. 27 (2016) 1060–1066.
[11] A. Drelinkiewicz, M. Hasik, J. Mol. Catal. A Chem. 177 (2001) 149–164.
[12] R. Bi, Q. Wang, C. Miao, J. Feng, D. Li, Catal. Lett. 149 (2019) (2000) 1286–1296.
11