Oxidative dehydrogenation of a biomass derived lignan - Hydroxymatairesinol over heterogeneous gold catalysts

Article abstract of DOI:10.1016/j.jcat.2011.05.025

Synthesis of the lignan oxomatairesinol via oxidative dehydrogenation of the naturally occurring lignan hydroxymatairesinol was studied over gold catalysts supported on C, TiO₂, SiO₂, Al₂O ₃, and MgO. In order to investigate the reaction performance over the gold catalyst, synthesis of lignan oxomatairesinol was carried out in different organic solvents/water mixtures under synthetic air and nitrogen atmosphere at 373 K, and using also isolated hydroxymatairesinol isomers as a starting material. The results were compared with those obtained over palladium catalysts. Synthesized supported gold catalysts as well as the corresponding supports were characterized by TEM, XRD, ICP-OES, CO₂-TPD, FTIR (using pyridine as a probe molecule), and XPS. Gold catalysts were shown to display superior performance compared with palladium ones: the activity was 4 times higher, with selectivity toward oxomatairesinol being 100%, while 60-85% were obtained over palladium catalysts. In contrast to palladium, the activity of gold catalysts is high in aerobic conditions and water-propan-2-ol mixture. However, activity and selectivity of gold catalysts were shown to be dependent on the electronic state of the metal and, similar to palladium catalysts, on the support acidity.

Full text of DOI:10.1016/j.jcat.2011.05.025

Journal of Catalysis 282 (2011) 54–64
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Oxidative dehydrogenation of a biomass derived lignan – Hydroxymatairesinol
over heterogeneous gold catalysts
Olga A. Simakova a,d, Elena V. Murzina a, Päivi Mäki-Arvela a, Anne-Riikka Leino b, Betiana C. Campo a,
Krisztián Kordás b, Stefan M. Willför c, Tapio Salmi a, Dmitry Yu. Murzin a,
⇑
a Laboratory of Industrial Chemistry and Reaction Engineering, Process Chemistry Centre, Åbo Akademi University, FI-20500 Åbo/Turku, Finland
b Microelectronics and Materials Physics Laboratories, EMPART Research Group of Infotech Oulu, University of Oulu, FI-90014 Oulu, Finland
c Laboratory of Wood and Paper Chemistry, Process Chemistry Centre, Åbo Akademi University, FI-20500 Åbo/Turku, Finland
d Graduate School of Materials Research, Åbo Akademi University, Porthansgatan 3-5, FI-20500 Åbo/Turku, Finland
a r t i c l e i n f o
a b s t r a c t
Article history:
Synthesis of the lignan oxomatairesinol via oxidative dehydrogenation of the naturally occurring lignan
hydroxymatairesinol was studied over gold catalysts supported on C, TiO2, SiO2, Al2O3, and MgO. In order
to investigate the reaction performance over the gold catalyst, synthesis of lignan oxomatairesinol was
carried out in different organic solvents/water mixtures under synthetic air and nitrogen atmosphere
at 373 K, and using also isolated hydroxymatairesinol isomers as a starting material. The results were
compared with those obtained over palladium catalysts. Synthesized supported gold catalysts as well
as the corresponding supports were characterized by TEM, XRD, ICP-OES, CO2-TPD, FTIR (using pyridine
as a probe molecule), and XPS. Gold catalysts were shown to display superior performance compared
with palladium ones: the activity was 4 times higher, with selectivity toward oxomatairesinol being
100%, while 60–85% were obtained over palladium catalysts. In contrast to palladium, the activity of gold
catalysts is high in aerobic conditions and water–propan-2-ol mixture. However, activity and selectivity
of gold catalysts were shown to be dependent on the electronic state of the metal and, similar to palla-
dium catalysts, on the support acidity.
Received 18 March 2011
Revised 26 May 2011
Accepted 26 May 2011
Available online 7 July 2011
Keywords:
Gold catalysts
Palladium catalysts
Selective oxidation
Lignans
Oxomatairesinol
Hydroxymatairesinol
Ó 2011 Elsevier Inc. All rights reserved.
1. Introduction
and cardiovascular diseases [6]. Besides the anticarcinogenic
effects, lignans are proven to have antioxidative effects [7–9]. Stud-
Lignans are natural phenolic compounds found in different parts
of plants, such as wooden parts, roots, leaves, stem, seeds, flowers,
and fruits. However, plants mainly containlignans in small amounts,
moreover, usually as glycosidic conjugates that are associated with
fibercomponents, makingtheisolationprocesscomplicated[1]. Flax
seed, sesame seed, and rye grains are among the richest known food
sources of lignans. Coniferous trees contain a large number of differ-
ent lignans in unconjugated form. The type and amount of lignan de-
pend on the species and part of the tree. Thus, Norway spruce (Picea
ies of the lignans matairesinol (MAT) and oxidized MAT (oxoMAT)
for radical and superoxide scavenging activities have shown that
MAT has the highest radical scavenging activity while oxoMAT
exhibits the highest superoxide scavenging activity.
Lignans or lignan derivatives, due to their antioxidative activity
[10] and UV-protection properties [11], can be applied for cosmetic
and pharmaceutical uses (skin- and hair-care products), as well as
color-keeping agents for textile industry, wherein the active agent
is selected from the group consisting of hydroxymatairesinol
(HMR), lariciresinol, secoisolariciresinol, isolariciresinol, oxoma-
abies) knots contain large amounts of lignans:
todolactol A derivatives, secoisolariciresinol, lariciresinol, pinoresi-
nol, matairesinol (MAT), isolariciresinol, -conidendric acid (ConiA),
a-conidendrin (Coni),
tairesinol (oxoMAT),
a-conidendrin (Coni), liovil, picearesinol,
a
syringaresinol, and nortrachelogenin. Among this group of lignans,
only HMR can be extracted in considerable amounts from the
Norway spruce (Picea abies) knots, while oxoMAT cannot be ob-
tained in large quantities from natural resources.
Several valuable lignans have been synthesized using HMR as a
starting material [12,13]. For instance, MAT can be produced
through hydrogenolysis of HMR (K-acetate adduct of HMR) under
hydrogen pressure over Pd/C catalyst in dichloroethane and in eth-
anol by using Raney-Nickel or Pd/C catalysts [14,15]. Another lig-
nan that is of interest is oxoMAT, which can be obtained through
lignan A, isohydroxymatairesinol, and hydroxymatairesinol (HMR)
[2–4]. The latter is the most abundant one, constituting 65–
85 wt.% of the total lignans and occurs in unconjugated free form [5].
Lignans are known as human health-promoting agents. A lig-
nan-rich diet, for example, decreases the risk for various cancers
⇑
Corresponding author. Fax: +358 2 215 4479.
E-mail address: dmurzin@abo.fi (D.Yu. Murzin).
0021-9517/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2011.05.025

O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
55
oxidative dehydrogenation of HMR. The formation of oxoMAT from
HMR by light-irradiation was reported in [16], while in [17], 2,3-di-
chloro-5,6-dicyano-1,4-benzoquinone (DDQ) was used as an oxi-
dizing agent. An attempt has been made over Pd/C to perform
the reaction via heterogeneous catalysis, providing a more com-
mercially attractive synthesis of oxoMAT. Unfortunately, oxoMAT
was not the only product in this reaction [18]. There is no addi-
tional information available about this synthesis carried out over
heterogeneous catalysts. Hence, the need for industrially applica-
ble catalytic methods for the selective synthesis of lignan oxoMAT
(oxomatairesinol) exists.
In this paper, a new approach for oxoMAT synthesis utilizing
heterogeneous Au catalysts is reported. The obtained data were
compared with the results over heterogeneous Pd/C catalysts,
which are also active in this reaction. OxoMAT was obtained via
oxidative dehydrogenation of HMR, and the reaction network is
presented in Scheme 1.
The lignan HMR exists in nature as a mixture of two diastereo-
mers of HMR: (7R,8R,8’R)-(-)-7-allo-hydroxymatairesinol (HMR 1)
and (7S,8R,8’R)-(-)-7-allo-hydroxymatairesinol (HMR 2), which
can be isomerized into each other. The ratio between HMR 2/
HMR 1 is typically equal to 2. Both HMR 1 and HMR 2 can be trans-
formed by oxidative dehydrogenation to oxoMAT.
In this study, the activity of gold catalysts, the influence of cat-
alyst support, solvent, and reaction atmosphere were investigated
in oxidative dehydrogenation of hydroxymatairesinol.
porous carbon material Sibunit (SBET = 450 m2/g, micropore area
37.7 m2/g, average pore diameter 9 nm) was used as a support.
Palladium on activated carbon catalyst (Degussa) was dried
overnight at 353 K and then reduced under flowing hydrogen at
423 K.
2.1.2. Gold catalysts
Several methods of catalyst preparation have been applied: di-
rect ion exchange (DIE) of the gold species with the hydroxyl
groups of the support surface as described in [20] and deposi-
tion–precipitation with urea (DPU) [21]. Carbon supported gold
catalysts were prepared via the immobilization of PVA stabilized
gold sols by the carbon support (Sibunit) [22]. The detailed proce-
dures are given below.
2.1.2.1. Au/Al2O3 preparation by direct ion exchange (DIE). Hydrogen
tetrachloroaurate (HAuCl4 99.9% ABCR, Darmstadt) was diluted
with deionized water up to the concentration of 5ꢁ10ꢀ4 M in the
amount corresponding to the final Au loading of 2 wt.%. The solu-
tion was heated up to 343 K and powdered Al2O3 (UOP, A-201,
SBET = 200 m2/g) was added. The slurry was mixed for 1 h, washed
with ammonium hydroxide (4 M) for 1 h, dried overnight at 353 K,
and calcined in air at 573 K for 4 h. The obtained catalyst was
marked as Au/Al2O3-DIE.
2.1.2.2. Au/Al2O3, TiO2, SiO2, MgO preparation by deposition–precip-
itation with urea (DPU). An aqueous solution of HAuCl4 with the
concentration of 5 ꢂ 10ꢀ4 M in the amount corresponding to the fi-
nal Au loading of 2 wt.% was mixed with urea (Sigma–Aldrich,
99.5%). The amount of urea was calculated to achieve the concen-
tration of 0.21 M. The solution was heated up to 354 K and then the
powdered supports such as Al2O3 (UOP, A-201, SBET = 200 m2/g),
TiO2 (Degussa AG, Aerolyst 7708, anatase >70%, SBET = 45 m2/g),
SiO2 (Merck, SBET = 480–540 m2/g), and MgO (Sigma–Aldrich,
99%) were added. The slurry was mixed for 24 h at 354 K and then
filtered, washed with water, dried overnight at 353 K, and calcined
at 573 K for 4 h. The catalyst supported on alumina was denoted as
Au/Al2O3-DPU to distinguish it from a catalyst prepared by direct
ion exchange.
2. Experimental section
2.1. Catalysts
2.1.1. Palladium catalysts
Two powdered Pd catalysts were utilized in HMR oxidation:
Pd/Sibunit prepared by the authors (denoted as Pd/C) and a com-
mercial one 5 wt.% Pd/C (Degussa).
Carbon supported palladium catalysts were prepared by using
deposition of Pd(II)-hydroxochlorocomplexes in the amounts cor-
responding to the final metal loading (4 wt.%). Support was
impregnated with Na2CO3 aqueous solution at pH = 8; thereafter,
palladium precursor H2PdCl4 (10ꢀ3 M) aqueous solution was added
dropwise in the ratio corresponding to Na:Pd = 21 mol/mol as de-
scribed in [19], followed by drying of the catalyst at 353 K in air
overnight and reduction under flowing hydrogen at 423 K. Meso-
2.1.2.3. Au/C preparation by gold sols immobilization by car-
bon. HAuCl4 (99.9% ABCR, Darmstadt) was dissolved in deionized
water up to a concentration of 60 lg/ml. While maintaining this
O
H3CO
O
HO
HMR 2
HO
O
H3CO
OCH3
OH
O
HO
O
O
H3CO
OCH3
O
OH
HO
HO
HMR 1
oxoMAT
OCH3
OH
Scheme 1. Reaction pathway of lignan hydroxymatairesinol (HMR) transformations over heterogeneous gold catalysts.

56
O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
solution under vigorous stirring, 2 wt.% solution of PVA was added
at Au/PVA weight ratio equal to 1:5. Freshly prepared 0.1 M solu-
tion of NaBH4 was added dropwise at Au/NaBH4 weight ratio equal
to 1:100. Within a few minutes of sol generation, the sol was
immobilized by adding mesoporous carbon material Sibunit
(SBET = 450 m2/g, micropore area 37.7 m2/g, average pore diameter
9 nm). The nominal metal loading was 2.5 wt.%. After 2 h, the slur-
ry was filtered and the catalyst was washed thoroughly with
deionized water, followed by drying overnight at 333 K in air.
2.2.5. Support acidity measurements
The acidity of the supports was determined by Fourier trans-
formed infrared spectroscopy (FTIR) using pyridine as a probe mol-
ecule. Thin self-supporting wafers of the adsorbents were prepared
by pressing and placed in the cell. The cell was evacuated to
ꢃ10ꢀ6 Torr at 723 K for 1 h. Thereafter, the cell was cooled to
373 K, and background (Bkg) spectra were taken as an average of
accumulated 100 scans over the frequency range (4000–
400 cmꢀ1). Pyridine was than adsorbed onto the sample at 373 K
for 30 min. The subsequent desorption was performed at 523 K,
623 K, and 723 K. The spectra were taken between each temperature
ramp at 373 K. Adsorption bands around 1610 and 1450 cmꢀ1 were
characterizing Lewis acid sites, while the adsorption signals at 1545
and 1630 cmꢀ1 were corresponding to the Brønsted acidity of the
adsorbent [23].
2.2. Catalyst characterization
2.2.1. Metal particle size determination
X-ray diffraction experiments were performed by using CuK
a
radiation (Siemens D5000 diffractometer equipped with a graphite
monochromator to suppress fluorescent and Cu-Kb radiation). The
average crystallite size of the gold catalyst particles was estimated
using Scherrer’s equation from the peak half-widths of Au(1 1 1)
2.3. Hydroxymatairesinol (HMR) oxidation to oxomatairesinol
(oxoMAT)
reflection measured at 2
H
ꢃ 38.3°. Instrumental broadening and
2.3.1. Starting material
unresolved CuKa1 and a2 radiation were neglected (error less than
2%).
Hydroximatairesinol was isolated from Norway spruce knots as
described in [24]. Lignan was extracted from ground knots by ace-
tone–water mixture. The extract was concentrated in a rotary
evaporator and then purified by flash chromatography. Two diaste-
reomers of HMR: (7R, 8R, 8’R)-(-)-7-allo-hydroxymatairesinol
(HMR 1) and (7S, 8R, 8’R)-(-)-7-allo-hydroxymatairesinol (HMR 2)
Electron microphotographs of the samples were taken by a LEO
912 OMEGA energy-filtered transmission electron microscope by
using 120 kV acceleration voltage. Histograms of the particle size
distribution were obtained by counting at least 100 particles on
the micrographs for each sample.
were obtained as
a mixture with the ratio between them
HMR 2:HMR 1 = 2:1 mol/mol. The purity of HMR was determined
by GC to be 95%. The major contaminant was the lignan coniden-
drin and conidendric acid (see Fig. 1).
Both isomers HMR 1 and HMR 2 could be oxidized, and the isom-
erization between them occurs. For the experiments, the obtained
mixture containing isomers at the ratio of HMR 2:HMR 1 = 2 mol/
mol was utilized as a reactant. However, to study each isomer activ-
ity in the oxoMAT synthesis, HMR 1 (92% of purity) and HMR 2 (85%
of purity) were used as starting materials. The separation of the iso-
mer from the mixture was performed by flash chromatography.
2.2.2. Gold metal loading analysis
The metal loading was determined by inductively coupled plas-
ma-optical emission spectroscopy (ICP-OES) in a PerkinElmer, Op-
tima 5300 DV spectrometer. For that purpose, 60 mg of the sample
was treated with 4 ml of aqua regia, digested in a microwave oven,
diluted to 100 ml, and analyzed by spectrometer.
2.2.3. XPS study of the metal oxidation state
The XPS measurements were carried out in a PHI Quantum
2000 Scanning ESCA Microprobe spectrometer with Al anode. Gold
supported on TiO2, Al2O3, and SiO2 was analyzed as powder
mounted in a double sided adhesive tape. The survey and multi
spectra were collected at 187.75 eV and 11.75 eV pass energy,
respectively. During the analyses, electron and ion beams were
applied over the catalyst to compensate the charge effect. Spectra
deconvolution was performed with the free software XPS Peak 4.1.
A combination of 30% Lorentzian and 70% Gaussian curves was fit-
ted after subtraction of a Shirley background. For quantitative anal-
ysis, the peaks areas were affected for the correspondent
sensitivity factors.
2.3.2. Oxidation/dehydrogenation of hydroxymatairesinol
The reaction was performed under atmospheric pressure in a
stirred 200 ml glass reactor, equipped with a heating jacket (using
silicon oil as the heat transfer medium), a reflux condenser (cooling
medium set 253 K), oil lock, pitched-blade turbine, and stirring
baffles. In a typical experiment, 250 mg of catalyst was put into
reactor in case of gold catalysts and 100 mg in case of palladium.
To study the support activity, 245 mg of alumina was charged into
the reactor. The catalyst grain size was 45–63 lm to suppress the
internal mass transfer limitations. Furthermore, the reaction mix-
ture was stirred at 1000 rpm to avoid external mass transfer limi-
tations, and the experiments were performed in the kinetic regime.
The catalysts were preactivated in situ by heating under hydrogen
(AGA, 99.999%) flow (100 ml/min) until 393 K, and thereafter, the
reactor was cooled down to the reaction temperature 343 K under
2.2.4. Support basicity measurements
Magnesia, titania, and alumina basicity was measured by tem-
perature-programmed desorption of CO2 (CO2-TPD) by using
Micromeritics AutoChem 290. Measurements were performed in
a conventional flow-through reactor with CO2 as the probe mole-
cule. Two hundred and fifty milligram of metal oxide was flushed
with a stream of helium (100 ml/min) to clean the sample at
823 K for 30 min before cooling to 323 K. A stream of CO2 was al-
lowed to saturate the solid particles for 30 min, and weakly bound
CO2 and other physisorbed molecules were removed by a stream of
pure He for 1 h. Chemisorbed CO2 was desorbed by heating to
1173 K (ramp of 10 C/min), while He was used as a carrier gas.
The effluent from the reactor was analyzed by a TC detector. CO2
calibration was performed by pulse injection of 1 ml CO2 to deter-
mine the actual amount of CO2 desorbed from the magnesia.
(a)
(b)
O
O
H3CO
H3CO
OH
OH
O
HO
HO
Coni
ConiA
OCH3
OCH3
OH
Fig. 1. Structure of lignan: (a)
OH
a
-conidendrin (Coni); (b) a-conidendric acid (ConiA).

O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
57
Table 2
nitrogen (AGA, 99.999%) flow (100 ml/min). To perform the reac-
tion over palladium catalysts, propan-2-ol (Sigma–Aldrich, 99.8%)
was used as a solvent. Gold catalysts were studied in different sol-
vents and their mixtures with deionized water (Table 2). The fol-
lowing organic solvents: ethanol (Altia Corporation, 99.5%),
propan-2-ol (Sigma–Aldrich, 99.8%), butan-1-ol (Acros, 99%), pen-
tan-2-ol (The British Drug Houses Ltd., 95%), cyclohexanol (J.T. Ba-
ker, 99%), 1,4-Dioxane (Labscan, 99.8%), and tetrahydrofuran (THF)
(Labscan, 99.8%), mixed with water were applied as solvents as
well. The reactant solution (100 ml) with an HMR concentration
of 1 mg/ml was poured into the reactor. The gas flow was changed
to synthetic air, except the case of dehydrogenation per se over pal-
ladium and gold catalysts, where the reaction was performed un-
der nitrogen flow. The stirring was started at reaction time set to
zero and the first sample was withdrawn.
The effect of solvent on the activity and selectivity in HMR oxidative dehydrogena-
tion. Experiments were performed over 250 mg 2 wt.% Au/Al2O3 (DIE), HMR
concentration was 1 mg/ml, the total volume of solution was 100 ml.
Solvent
Selectivity
toward
OxoMAT, %
Conversion Activity,ꢂ 105 ꢂ mol/ pH of the
after 4 h, % l ꢂ sec ꢂ gcat
HMR
solution
Alcohols
Ethanol
Propan-
2-ol
Butan-
2-ol
–
–
0
0
n.a.a
n.a.
5
5
–
–
0
0
n.a.
n.a.
5
5
Pentan-
2-ol
Etanol in water, vol.%
2
0
30
n.a.
5
Propan-2-ol in water, vol.%
2
5
10
20
50
80
100
100
100
100
100
100
70
70
53
52
52
30
6
6
3
3
2
2
5
5
5
5
5
5
2.3.3. Product analysis
Samples were withdrawn from the reactor at different time
intervals and analyzed by gas chromatography (GC) using a HP-1
column (length 25 m, inner diameter 0.20 mm, film thickness
0.11
l
m) and a flame ionization detector (FID) operating at
l (concentration of 1 mg/ml) of the
573 K. For the analysis, 100
l
Butan-2-ol, vol.%
sample was taken and 2 ml of the internal standard for the GC con-
taining mainly betulinol (0.02 mg/ml) and C21:0 fatty acid
(0.02 mg/ml) dissolved in methyl tert-butyl (MTBE, C5H12O) was
added. Samples were placed in a water bath at 313 K, where the
MTBE and solvent were evaporated in a stream of nitrogen. There-
after, the samples were dried in an oven at 313 K under vacuum for
20 min. The prepared samples were silylated prior to the analysis
as follows: dried samples were dissolved in the silylation mixture
2
0
30
65
24
15
n.a.
9
5
5
5
5
Pentan-2-ol, vol.%
35b
2
Cyclohexanol, vol.%
75c
2
0.2
0.1
Acetone, vol.%
83c
2
1,4- Dioxane, vol.%
of 80
Fluka), 25
and 25 l pyridine (99.0%, J.T. Baker). The samples were kept at
l
l N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA, 98%,
2
80
100
100
100
-
44
15
0
1.3
0.6
n. a.
5
6
6
l
l trimethylchlorsilane (TMCS, 98%, Acros Organics),
l
Tetrahydrofuran (THF), vol.%
2
80
100
343 K for 40 min and transferred to vials and then analyzed by
GC. One microliter of the silylated sample was injected with an
autosampler. The injection temperature was 533 K and the split ra-
tio was 1:20. Hydrogen served as a carrier gas. The initial temper-
ature of the column was 393 K (for 1 min), and the temperature
was increased at a rate 6 K/min to 573 K (for 10 min). The peaks
were identified by analysis with a gas chromatograph–mass spec-
trometer operating at the same GC conditions.
55
–
–
21
0
0
0.5
n. a.
n. a.
5
5.5
6
a
b
c
n.a. = Not active in oxoMAT was formation. The main by-product was lignan.
Conidendric acid (ConiA).
Hydroxymatairesinol (HMR 1).
tion histogram, while magnesia has a bimodal one. The reason
for such phenomena is the alkaline nature of magnesia. During
the deposition, the pH increases due to the urea decomposition,
which changes the charge of support surface. The rate of the gold
precursor adsorption is linearly proportional to the difference be-
tween the pH value of media and the Point Zero Charge of the sup-
port (PZC), which is about 12 in case of magnesia [25]. Due to urea
decomposition, the pH of media is increasing, which results in slow
adsorption of gold complexes. At the same time, gold(III) chlorohy-
droxocomplexes tend to form gold(III) polynuclearhydroxocom-
plexes [26], which form agglomerates of gold particles in the
solution and on the support surface. In case of slow adsorption,
3. Results and discussion
3.1. Catalysts properties
3.1.1. Au and Pd particle size and metal loading
Characterization data confirmed that gold particles were suc-
cessfully deposited as nano-particles in case of all utilized sup-
ports. The obtained TEM images and XRD patterns are shown in
Fig. 2, where d is an average metal particle size.
Gold particles supported on titania, silica, alumina, and meso-
porous carbon Sibunit have maxima in the particle size distribu-
Table 1
Dehydrogenation of HMR over Pd/C catalysts at 343 K under the nitrogen flow. Conditions: catalyst mass is corresponding to 5 mg of Pd mass, 100 mg of HMR, 100 ml of propan-
2-ol.
Catalyst
Activitya, ꢂ 104 moloxoMAT/l ꢂ s ꢂ gPd
Conversion after 4 h, %
Selectivityb, %
OxoMAT
Yield after 4 h, %
OxoMAT
MAT
MAT
5 wt.% Pd/C (Degussa)
4 wt.% Pd/C (Sibunit)
7.2
5.3
92
87
62
87
38
13
64.4
73.1
27.6
13.9
a
Activity of catalyst in the oxoMAT synthesis was calculated according to Eq. (3).
Selectivity was calculated at 4% of HMR conversion.
b

58
O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
(a) 2 wt.% Au/TiO2
d = 1.9 1.0 nm
50
40
30
20
10
0
300
250
200
150
100
50
0
2
4
6
8 101214161820
20 25 30 35 40 45 50
Particle size (nm)
2θ (degree)
(b) 2 wt. % Au/SiO
d = 1.6 0.8 nm
2
1200
1000
800
50
40
30
20
10
0
600
400
200
20 25 30 35 40 45 50 55 60 65
0 2 4 6 8 101214161820
Particle size (nm)
2θ (degree)
(c) 2 wt. % Au/MgO
d = 2-3 nm and 6-8 nm
50
40
30
20
10
0
120
100
80
60
40
20
0
2 4 6 8 10 12 14 16 18 20
25 30 35 40 45 50 55
Particle size (nm)
2θ (degree)
(d) 2.5 wt. % Au/C
d= 3.3 1.2 nm
50
40
30
20
10
0
100
80
60
40
34 36 38 40 42 44 46 48
0
2 4 6 8 10 12 14 16 18 20
Particle size (nm)
2θ (degree)
Fig. 2. TEM images and XRD patterns of the catalysts: (a) 2 wt.% Au/TiO2; (b) 2 wt.% Au/SiO2; (c) 2 wt.% Au/MgO; (d) 2.5 wt.% Au/C; (e) 2 wt.% Au/Al2O3-DIE; (f) 2 wt.% Au/
Al2O3-DPU.
more gold(I) polynuclearhydroxocomplexes are formed in the
solution followed by deposition of them onto magnesia surface
as bigger particles.
Gold loadings analyzed by ICP-OES are summarized in Table 3.
The amount of supported gold was found to be in the range 1.4–
2.4 wt.% depending on the support type.
The XRD results showed that in the most cases, Au reflections
are missing indicating the presence of a very small gold particle
size (less than 3 nm). Although an Au(1 1 1) reflection was ob-
served in case of carbon and silica supports, the contribution of
big particles is too small to be visible in the particle size
distribution.
Pd catalyst supported on Sibunit has an average particle size
4.6 2.2 nm. The catalyst Pd/C supplied by Degussa has a maxi-
mum in the particle size distribution at about 3 nm. XRD study
has confirmed the results on the particle size obtained by TEM.
3.1.2. Gold oxidation state
The X-ray photoelectron (XPS) spectra of the Au 4f core level of
Au/Al2O3 catalysts show only one band corresponding to Au0.
However, in case of Au/TiO2 and Au/SiO2, the contribution of two
metal states was detected: Au0 and Audꢀ. These observations could
be explained by two phenomena: electron transfer from the sup-
port toward the gold particles [27,28] or a dominant effect of the
atoms at low coordinated sites as proposed by Radnik and et al.
[29]. The first one could be applied only to the titania supported

O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
59
(e) 2 wt. % Au/Al2O3-DIE
d = 2.2 1.1 nm
50
40
30
20
10
0
XRD is not reliable
0
2 4 6 8 10 1214 16 18 20
Particle size (nm)
(f) 2 wt. % Au/Al2O3-DPU
d = 2.3 0.6 nm
50
40
30
20
10
0
XRD is not reliable
0
2 4 6 8 10 12 14 16 18 20
Particle size (nm)
Fig. 2 (continued)
Table 3
Kinetic results on HMR oxidative dehydrogenation at 343 K under the synthetic air flow using 2 vol.% propan-2-ol in water as a solvent. Data was calculated according to Eqs. (1)–
(3).
Entry Catalyst
Metal loading,
wt.%
Average Au particle size,
nm
Activity, ꢂ104 moloxoMAT
l ꢂ s ꢂ gAu
/
Conversion of HMR after
240 min, %
Selectivitya to
oxoMAT, %
1
2
3
4
5
Au/MgO
Au/C
Au/SiO2
Au/TiO2
Au/Al2O3-
DPU
1.4 0.10
1.6 0.60
2.4 0.20
1.9 0.20
2.0 0.14
2.0 1.0
3.3 1.2
1.6 0.8
1.9 1.0
2.3 0.6
n.a.
0.06
5
13
23
100
6
13
21
50
0
100
100
100
100
b
6
Au/Al2O3-
DIE
Al2O3
2.1 0.15
2.2 1.0
29
70
100
7
8
0
0
–
–
0
0
33
100
0c
0c
MgO
n.a. = Not active.
a
Selectivity to oxoMAT was measured at 10% of HMR conversion.
Selectivity was calculated at 6% of conversion.
The product of the reaction over Al2O3 was HMR 1 and over MgO was ConiA.
b
c
catalyst, since silica demonstrates weak interactions with noble
metal clusters [30]. The effect of low coordinated atoms could be
detected in case of both carriers, silica and titania, since gold par-
ticle size is lower than 2 nm.
In order to study the alumina acidity and basicity changes dur-
ing the catalyst preparation gold free, alumina support was treated
in the same way as during the catalyst preparation by DIE (conse-
quent washing with ammonia and water followed by drying and
calcination). The obtained sample denoted as Al2O3-DIE was ana-
lyzed by pyridine FTIR and CO2-TPD techniques.
Measurements on the acidity and basicity are collected in Table
4. Lewis acidity is increasing in the order: MgO = SiO2 ꢄ Al2O3-
DIE < Al2O3 < TiO2. The amount of alumina Lewis acid sites de-
creased during the treatment according to the catalyst preparation
procedure. At the same time, the basicity of alumina before and
after treatment was the same. Among the analyzed samples, mag-
nesia possesses the highest basicity.
3.1.3. Support acidity and basicity
The nature of supports acid sites was studied by FTIR using pyr-
idine as a probe molecule. Since Selli and Forni [31] have found a
good correlation between the acidities obtained by FTIR measure-
ments and those obtained by TPD for a number of different acidic
catalysts, the pyridine FTIR technique can be applied for quantita-
tive determination of acid sites. FTIR spectral interpretation was
performed according to the commonly accepted procedure which
attributes adsorption at about 1610 and 1450 cmꢀ1 to Lewis acid
sites and peaks at 1545 and 1630 cmꢀ1 to pyridine adsorbed on
Brønsted acid sites. Brønsted acidity was not observed in any of
the samples, while the number of Lewis acid sites was different.
3.2. Catalytic results
Oxidative dehydrogenation of HMR was studied over different
types of Pd and Au supported catalysts. Effects of different solvents

60
O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
Table 4
to compounds, which are poisoning gold catalysts, as for example,
Acidity and basicity of the supports.
acetone via propan-2-ol oxidation [34] or other aldehydes and car-
boxylic acids generated from corresponding alcohol. The formation
of acetone from propan-2-ol was detected in the present work by
using high-performance liquid chromatography (HPLC), analysis
procedure is described in [35]. Thus, when an alcohol–water mix-
ture was applied, gold catalysts became active in the oxoMAT syn-
thesis due to a minor effect of alcohols oxidation. Moreover, the
catalytic activity is increasing with the increasing water content
in the reaction mixture. The highest activity and selectivity were
reached applying 2 and 5 vol.% propan-2-ol in water as a solvent.
Utilization of cyclohexanol–water mixture as a solvent was an
attempt to diminish the contribution of solvent oxidation via steric
hindrance [36]. Application of this solvent resulted in lower cata-
lytic activity and selectivity to oxidation/dehydrogenation reaction
compared with propan-2-ol–water mixture.
Support
Acid sites (
l
mol/g)
Basic sites (
l
mol/g)
Brønsted
Lewis
Brønsted
Lewis
Al2O3
Al2O3-DIE
SiO2
TiO2
MgO
0
0
0
0
0
50
10
0
110
110
n.d.
84
150
0
250
n.d. = Not determined.
and reaction atmospheres were studied over gold catalysts and
compared with the performances over palladium catalysts.
Turn-over frequency, yield, and conversion were calculated as
follows:
In order to investigate the effect of acetone, which could be
formed during the reaction as a by-product due to oxidation of pro-
pan-2-ol from the solvent, the reaction was performed in acetone–
water solution. Although the content of acetone was only 2 vol.%,
the activity of the catalyst dramatically decreased. The selectivity
toward desired oxoMAT was 83%, while the less active isomer
HMR 1 was also produced.
To exclude the competitive oxidation of alcohol from the sol-
vent and the reactant HMR, solvents such as 1,4-dioxane and
THF were utilized. Application of these solvents showed that the
catalyst was becoming more active with increasing of water con-
tent in the solution. No HMR oxidation/dehydrogenation to oxoM-
AT in pure 1,4-dioxane and THF was detected. This result could be
explained by further transformation of these solvents on the cata-
lyst active sites in reactions such as ring opening, consequent oxi-
dation and polymerization [37–39].
20min
OxoMAT
mol
TOF ¼
ð1Þ
molsurfaceAu ꢂ 1200 s
TOF was calculated by dividing the initial reaction rate, calculated
from the slope of the concentration–time plot, to the amount of ex-
posed catalytic sites. The amount of surface Au was calculated
according to the size of gold particles observed by TEM and catalyst
metal loading.
C0HMR2 ꢀ CHMR2
Conversion ¼
ꢂ 100%
ð2Þ
ð3Þ
C0HMR2
Activity ¼ molOxoMAT=volume ꢂ time0 ꢂ mcat
3.2.1. Solvent effects
3.2.1.1. Palladium catalysts. HMR dehydrogenation over palladium
catalysts was carried out at the same reaction conditions as de-
scribed previously in [18]. Pd/Sibunit was tested using as a solvent
propan-2-ol per se and 2 vol.% propan-2-ol in water. It was shown
that the presence of water in the solvent significantly changes the
catalyst activity and selectivity as well (Table 1). Palladium catalyst
has higher activity applying propan-2ol per se as a solvent.
The selective aerobic oxidation of HMR was studied using a
range of organic solvents and their mixtures with water. It was ob-
served that increasing water content in the reaction mixture re-
sulted in increasing conversion of HMR to the desired product—
oxoMAT. The same tendency was discovered in [40], where water
had shown to have a promotion effect on the activity of gold cata-
lyst. In case of HMR oxidation, the role of water could be related to
two phenomena: first, suppressing of described above possible
deactivation due to organic solvent interaction with catalyst sur-
face and second, activation of adsorbed oxygen by adsorbed water,
due to formation of hydroperoxyl-like intermediate species (OOH)
followed by their decomposition and releasing of Oꢅ and H2O [37].
3.2.1.2. Gold catalysts. The oxidative dehydrogenation of HMR was
performed in different solvents by using 2 wt.% Au/Al2O3 as a cat-
alyst. Since HMR is poorly soluble in water and hydrocarbons,
but well soluble in alcohols, the reaction was performed in differ-
ent alcohols and their mixtures with water. In order to avoid the
competitive oxidation of HMR and alcohols over the gold catalyst,
the reaction was also performed using 1,4-dioxane and tetrahydro-
furan (THF) and their mixtures with water as a solvent. Gold cata-
lysts were not active in this reaction, when ethanol, propan-2-ol,
butan-1-ol, pentan-2-ol were used as solvents. However, applying
mixtures of alcohol and water as a solvent results in the formation
of the desired product—oxoMAT. The formation of oxoMAT was
also observed in solvents different from alcohols, e.g., 1,4-dioxane
and tetrahydrofuran (THF) and their mixtures with water. Since
the most probable by-products, Coni and ConiA, can be formed
with increasing of the solution pH, the pH value of each reaction
mixture was measured and found to be similar in all cases. Thus,
the product distribution is not related to the difference in the solu-
tion pH. The results are presented in the Table 2.
3.2.2. Effect of catalysts: active metal, support and preparation method
3.2.2.1. Palladium catalysts. The tested Pd/C catalysts have shown
different catalytic behavior (Table 1) with the highest activity
achieved over 5 wt.% Pd/C (Degussa), which could be associated
with larger support acidity [18]. Conversion of HMR over Pd/C (De-
gussa) after 4 h was similar to the case of Pd on Sibunit supported
catalysts.
Over palladium catalysts HMR undergoes two types of transfor-
mation: dehydrogenation to oxoMAT and hydrogenolysis to MAT
[18]. The studied Pd catalysts have shown different selectivity to-
ward these products (Table 1). In particular, selectivity and yield
of oxoMAT were the highest in case of Pd/C (Sibunit) catalyst.
The reaction kinetics over this catalyst is shown in Fig. 3. The
increasing of HMR 1 concentration in the beginning of the reaction
was related to the isomerization of HMR 2 to HMR 1.
Gold catalysts were inactive for the oxidative dehydrogenation
of HMR, when pure alcohols were applied as solvents. However,
application of alcohol with water mixtures results in oxoMAT for-
mation due to HMR oxidative dehydrogenation. These changes
could be related to the activity of gold catalysts in the alcohols oxi-
dation [32,33]; thus, the catalyst interacts with the alcohol in the
solvent rather than HMR. Alcohol can be oxidized over gold leading
3.2.2.2. Gold catalysts.
3.2.2.2.1. Catalytic activity. The tested gold catalysts have shown
different activities toward oxidation/dehydrogenation of HMR (Ta-
ble 3). Among the catalysts active in selective HMR oxidation/
dehydrogenation, the most active catalyst was gold supported on

O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
61
0.020
0.016
0.012
0.008
0.004
0.000
The conversion of HMR was the highest with Au/MgO, but this
catalyst was not selective to oxoMAT (Table 3). The conversion of
HMR over other Au supported catalysts increases with increasing
of supports acidity (Table 4).
Furthermore, the Au/Al2O3-DIE catalyst was more active than
Au/Al2O3-DPU giving after 240 min 70% conversion, whereas only
50% conversion was achieved over the latter catalyst.
Moreover, both supports, alumina and magnesia, without sup-
ported gold particles were active in the HMR isomerization (Table
3, entries 7 and 8). However, no oxoMAT was formed in the ab-
sence of gold, thus, indicating that the presence of gold particles
is needed for the synthesis of the desired product—oxoMAT.
OxoMAT
MAT
HMR 1
HMR 2
0
50
100 150 200 250
Time, min
Fig. 3. Transformation of HMR to oxoMAT and MAT over Pd/C (Sibunit) catalyst
under nitrogen flow at 343 K in propan-2-ol.
3.2.2.2.3. Yield and selectivity. Except magnesia supported gold cat-
alyst, all the catalysts displayed 100% selectivity toward oxoMAT,
while Pd catalysts tested in [18] and described in this paper have
shown lower selectivity toward oxoMAT.
alumina (Table 3, entry 6). The reaction over titania and silica sup-
ported catalysts resulted in relatively low yields of HMR (entries 3
and 4, Table 3). The lowest activity was obtained over the carbon
supported gold catalyst. As was already studied in [18], in case of
palladium, catalysts with a higher support acidity had higher activ-
ity in dehydration. The same tendency was detected in case of the
gold catalysts. The acidity of the carbon supported catalyst cannot
be measured by the same technique as applied for metal oxides,
e.g., pyridine desorption studied by FTIR. However, the acidity of
carbons is typically lower than those of metal oxides. The obtained
data on carbon supported gold catalyst showed the lowest activity
in the HMR transformation. The gold catalysts supported on differ-
ent metal oxides exhibited the following trend in TOF for the HMR
transformation: MgO ꢄ SiO2 < Al2O3 (Table 3), which follows the
same order of increasing Lewis acidities (Table 4). It is worth to
note that although titania possesses higher acidity than applied
alumina, the activity of Au/TiO2 is lower. This difference could be
attributed to the different state of the supported gold (Audꢀ), de-
tected by XPS, and related to the strong metal-support interaction.
The HMR oxidation reaction is in fact selective oxidation of sec-
ondary alcohols. TOF of oxoMAT synthesis over gold catalysts was
changing in the range 1 ꢂ 10ꢀ3 sꢀ1 to 20 ꢂ 10ꢀ3 sꢀ1, which is much
lower than TOF values calculated for other secondary alcohols oxi-
dation reaction over supported gold [41]. This difference could be
related to the steric hindrance of OH-group in HMR presented in
different conformations [42].
The alumina supported gold catalysts displayed metal particle
size distributions, measured by TEM and XRD techniques that were
independent of the preparation method; however, their activity was
different. Since the methods of catalysts preparation, namely DPU
and DIE, have different mechanisms of gold particles deposition;
consequently, the gold particles size distribution could be different.
The DPU method allows formation of [Au(Cl)4ꢀx(OH)x]ꢀ complexes
[43] and their growth onthe support surface similarly as in the water
solution, forming polymeric gold (III) chlorohydroxocomplexes [26]
during the deposition procedure. The other method used for gold
deposition on the alumina surface, DIE, implies the adsorption of
hydrolyzed gold precursor as [Au(Cl)4ꢀx(OH)x]ꢀ followed by ammo-
nia washing, which efficiently removes the remaining chloride ions
[20]. Thus, it was supposed that DIE method results in formation of
smaller particles not visible either by TEM or detected by XRD tech-
niques. This assumption could explain the difference in the activity
of the tested catalysts supported on alumina, while more active cat-
alyst is prepared by DIE method resulting in supposedly higher dis-
persion of gold catalysts.
In the case of Au/MgO catalysts, HMR was transformed into an-
other lignans: conidendrin (Coni) and conidendric acid (ConiA)
(see Fig. 1) by HMR dehydration followed by cyclization. This
behavior is related to the basicity of magnesia, as ConiA is normally
formed with increasing pH value of the solution.
The isomers of HMR, HMR 1 and HMR 2, can be transformed
into each other. At the same time, there was no isomerization ob-
served over gold catalysts active in HMR oxidation. As was already
mentioned, alumina per se was active in the isomerisation of
HMR 2 to HMR 1. However, under the same reaction conditions,
alumina supported gold particles were able to change the reaction
selectivity toward oxoMAT formation.
The dependence of the oxoMAT yield on time for all the tested
catalysts is shown in Fig. 4. The yield and conversion were calcu-
lated according to Eqs. (2) and (3).
The yield of oxoMAT was found to be dependent on the catalyst
support. Since oxoMAT was the only product of HMR oxidative
dehydrogenation, the yield of oxoMAT is increasing in the same or-
der as catalysts activity: MgO ꢄ SiO2 < TiO2 < Al2O3 (Table 3).
According to the shape of the curves presented on the Fig. 3, during
Table 5
Calculated reaction rate constants.
-1
Catalyst
k ꢂ 102, sꢀ1 ꢂ gAu
2 wt.% Au/Al2O3-DIE
2 wt.% Au/Al2O3-DPU
2 wt.% Au/TiO2
2 wt.% Au/SiO2
2 wt.% Au/C
6.4 0.8
5.4 0.4
5.0 0.2
4.0 0.4
0.4 0.04
100
80
60
40
20
0
a
b
c
d
e
f
0
50
100 150
Time, min
200 250
3.2.2.2.2. Activity. Gold catalysts have shown different activity in
the reaction of oxoMAT synthesis. The rate constants for the
HMR transformation reaction were calculated assuming the kinetic
curve described by the first order kinetics. The results are pre-
sented in the Table 5.
Fig. 4. OxoMAT yield as a function of time at 343 K using 2 vol.% propan-2-ol in
water as solvent in air flow over 250 mg catalysts: (a) Au/Al2O3-DIE
(d = 2.2 1.0 nm); (b) Au/Al2O3-DPU (d = 2.3 0.6 nm); (c) Au/TiO2
(d = 1.9 1.0 nm); (d) Au/SiO2(d = 1.6 0.8 nm); (e) Au/C (d = 3.3 1.2 nm); (f) Au/
MgO (d = 2.0 1.0 nm and 7.0 1.0 nm).
a

62
O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
0.020
(a)
(b)
0.020
0.016
0.012
0.008
0.004
0.000
0.016
OxoMAT
0.012
0.008
HMR1
0.004
0.000
HMR2
0
50
100 150 200 250
Time, min
0
50
100 150 200 250
Time, min
Fig. 5. Dehydrogenation of HMR in 2 vol.% propan-2-ol in water over 250 mg of 2 wt.% Au/Al2O3-DIE catalyst at 343 K under the flow of: (a) synthetic air; (b) nitrogen.
(I)
(II) 2.0
2.0
1.6
1.2
0.8
0.4
0.0
a
c
b
1.6
1.2
0.8
0.4
0.0
e
f
g
d
0
50
100 150 200 250
0
50
100
150
200
250
Fig. 6. Ratio HMR 2/HMR 1 vs. the reaction time of the HMR oxidative dehydrogenation reaction in 2 vol.% propan-2-ol in water in air flow at 70 °C over 250 mg of: I—a)Au/C;
(b) Au/TiO2; (c) Au/SiO2; (d) Au/MgO; II—(e) Au/Al2O3-DPU; (f) Au/Al2O3-DIE; (g) Al2O3.
the reaction, the deactivation is taking place most probably due to
stronger adsorption of oxidation products than reactants. It should
be explicitly mentioned that Au/MgO displayed zero activity to-
ward oxoMAT formation, although Fig. 4 might give a wrong
impression due to the properties of graphical software.
3.2.4. Reactivity of HMR 1 and HMR 2 over gold
In case of the applications of the gold catalysts described in this
paper, no isomerization between HMR 1 and HMR 2 was observed.
However, the molar ratio between these two isomers changes dif-
ferently during the reaction, dependent on the catalyst. The results
were also compared with the activity of the alumina support with-
out gold particles, since alumina catalyzes the isomerization of
HMR 2 to HMR 1 as presented on Fig. 6.
Two isomers, HMR 1 and HMR 2, have different reactivity in
oxidative dehydrogenation. As was already mentioned, HMR 2 is
the only isomer transformed in the most cases, while in case of
the highly active alumina supported gold catalyst prepared by
DIE, HMR 1 was also oxidized to oxoMAT (Fig. 7a). Comparison of
the catalytic behavior of the alumina support and Au/Al2O3 toward
the HMR transformation (Fig. 5a and 8) showed explicitly the
activity of alumina toward the isomerization of HMR 2 to HMR 1.
The conversion of HMR 2 after 4 h was 33%, giving exclusively
HMR 1 as a product. Thus, HMR 1 consumption during the oxida-
tion could be either from isomerization of HMR 2 or oxidative
dehydrogenation to oxoMAT. In order to investigate the routes of
HMR isomers transformation, the reaction was performed using
the isolated isomers HMR 1 and HMR 2 separately as starting
materials under the same conditions as for their mixture over
Au/Al2O3-DIE catalyst.
3.2.3. Effect of the reaction atmosphere
3.2.3.1. Palladium catalysts. Applied Pd/C catalysts were not active
in the presence of oxygen (under the air flow). Thus, all the further
described results were obtained under the flowing nitrogen in line
with previous workers [18]. The reason for such behavior is that Pd
catalysts could be deactivated in the presence of oxygen due to Pd
oxide formation, which is catalytically not active.
3.2.3.2. Gold catalysts. Since oxoMAT formation was observed over
Pd catalysts under the nitrogen flow, activities of gold catalysts in
the oxoMAT synthesis were studied under the same reaction con-
ditions in the absence of oxygen. The kinetic curves shown in the
Fig. 5 clearly demonstrate that the presence of oxygen significant
increases the catalyst activity.
The HMR conversion over Au/Al2O3-DIE was 17% after 4 h under
nitrogen atmosphere (Fig. 5b), whereas under oxygen, the conver-
sion was much higher being 68% (Fig. 5a). The selectivity toward
oxoMAT was 100% in both cases: in the absence and presence of
oxygen. These results confirmed that the HMR transformation over
gold includes the dehydrogenation route, not requiring presence of
oxygen.
Procedure of gold catalyst preparation could suppress the activ-
ity of alumina toward isomerization HMR 2 to HMR 1 due to sup-
port acidity changing (Table 4). The adsorption of alcohols on
alumina surface takes place due to interactions with Lewis acid site
(Al) and Brønsted base site as studied in [45]. HMR hydroxyl-group
interactions with alumina surface can be described in the same
way as secondary alcohols adsorption on alumina. Thus, the fol-
lowing explanation could be proposed: due to the decreasing
It was already reported that gold catalysts have catalytic activ-
ity in the reaction of alcohols oxidation in the absence of oxygen.
However, the presence of oxygen leads to removing of adsorbed
hydrogen atoms by co-adsorbed oxygen [44], thus, increasing the
reaction rate.

O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
63
(a)
(b)
0.024
0.020
0.016
0.012
0.008
0.004
0.000
0.024
HMR 1
0.020
0.016
0.012
HMR 2
OxoMAT
HMR 1
0.008
OxoMAT
0.004
HMR 2
0.000
0
50
100 150 200 250
Time, min
0
50
100
150 200 250
Time, min
Fig. 7. Kinetic results on the oxidation/dehydrogenation of: (a) HMR 1; (b) HMR 2 over Au/Al2O3-DIE catalyst at 343 K in 2 vol.% propan-2-ol in water under synthetic air flow.
Activity of gold catalysts is high in the solvent containing mainly
water when oxygen is used as an oxidizing agent.
0.020
HMR 1
HMR 2
Palladium catalysts exhibited high catalyst activity only in pro-
pan-2-ol and under a nitrogen atmosphere. Contrary to palladium,
gold catalysts were not active either in propan-2-ol or other stud-
ied organic solvents and exhibited lower activity in the absence of
oxygen than in aerobic conditions. In addition, application of gold
catalysts afforded 100% selectivity toward the desired product.
Hydroxymatairesinol (HMR) was applied as a mixture of two
diastereomers of HMR: (7R,8R,8’R)-(-)-7-allo-hydroxymatairesinol
(HMR 1) and (7S,8R,8’R)-(-)-7-allo-hydroxymatairesinol (HMR 2),
which can be isomerized into each other. HMR isomers have differ-
ent activity toward oxomatairesinol (oxoMAT) synthesis, which
can be explained by different adsorption strength of diastereomers.
Such assumption requires further theoretical investigation by
quantum chemical methods, which is now in progress.
0.016
0.012
0.008
0.004
0.000
0
50
100 150
Time, min
200 250
Fig. 8. Concentration vs. time dependency of HMR transformation in 2 vol.%
propan-2-ol in water under synthetic air flow at 70 °C over 245 mg of Al2O3.
amount of Lewis acid sites the isomerization reaction is sup-
pressed, while oxidation of HMR is taking place via the adsorption
on the gold surface followed by alkoxy-intermediate formation
[41].
Acknowledgments
This work is part of the activities at the Åbo Akademi Process
Chemistry Centre within the Finnish Centre of Excellence Pro-
gramme (2000–2011) by the Academy of Finland. The authors ex-
press their gratitude to Dr. Lari Vähäsalo for HMR preparation and
Mr. Sten Lindholm for the ICP-OES analysis.
The obtained kinetic curves (Fig. 7) demonstrate that both iso-
mers oxidized to oxoMAT and no isomerization to each other
was observed. However, isomer HMR 1 is less active, than
HMR 2, reaching only 17% of conversion after 4 h, while during
the same time, HMR 2 is converted to 52% extent. These results
support the assumption that conversion of HMR 1 in the mixture
with HMR 2 is related to the oxidation of this isomer to oxoMAT.
Dehydrogenation of the isolated isomers over Pd/C catalysts
was studied previously. The isomer HMR 2 also reacted with higher
reaction rate. Quantum chemical calculations had showed that
HMR 1 formed more stable intermediate than HMR 2, therefore,
rendering it less active [18]. The same explanations could probably
be used in case of the reaction over gold catalysts.
References
[1] N.M. Saarinen, A. Smeds, S.I. Makela, C. Eckerman, M. Reunanen, M. Ahotupa,
S.M. Salmi, A.A. Franke, L. Kangas, R. Santti, Nutr Cancer 36 (2000) 207.
[2] R. Ekman, Holzforschung 30 (1976) 79.
[3] S. Willför, P. Eklund, R. Sjöholm, M. Reunanen, R. Sillanpää, S. von Schoultz, J.
Hemming, L. Nisula, B. Holmbom, Holzforschung 59 (4) (2005) 413.
[4] P.C. Eklund, S. Willför, A. Smeds, B.R. Holmbom, R.E. Sjöholm, J. Nat. Prod. 67
(6) (2004) 927.
[5] S. Willför, J. Hemming, M. Reunanen, C. Eckerman, B. Holmbom, Holzforschung
57 (2003) 27.
[6] H. Adlercreutz, Phytoestrogens and human health, in: K. Korach (Ed.),
Reproductive and Developmental Toxicology, Marcel & Dekker, NY, 1998, pp.
299–371.
4. Conclusions
[7] S.M. Willför, M.O. Ahotupa, J.E. Hemming, M.H.T. Reunanen, P.C. Eklund, R.E.
Sjöholm, C.S.E. Eckerman, S.P. Pohjamo, B.R. Holmbom, J. Agr. Food Chem. 51
(2003) 7600.
[8] S. Yamaguchi, T. Sugahara, Y. Nakashima, A. Okada, K. Akiyama, T. Kishida, M.
Maruyama, T. Masuda, Biosci. Biotechnol. Biochem. 70 (2006) 1934.
[9] P.C. Eklund, O.K. Långvik, J.P. Wärnå, T.O. Salmi, S.M. Willför, R.E. Sjöholm, Org.
Biomol. Chem. 3 (2005) 3336.
[10] H. Korte, M. Unkila, M. Hiilovaara-Teijo, V.-M. Lehtola, M. Ahotupa, Patent WO/
2004/000304 A1, 2003.
[11] T. Ahlnäs, Patent WO 2008/020112 A1, 2008.
[12] P. Eklund, R. Sillanpää, R. Sjöholm, J. Chem. Soc. Perkin Trans. 1 (2002) 1906.
[13] P.C. Eklund, R.E. Sjöholm, Tetrahedron 59 (2003) 4515.
[14] P. Eklund, A. Lindholm, J.-P. Mikkola, A. Smeds, R. Lehtilal, R. Sjöholm, Org. Lett.
5 (2003) 491.
[15] R. Sjöholm, P. Eklund, J.-P. Mikkola, R. Lehtila, M. Sodervale, A. Kalapudas, US
Patent 0080299, 2005.
Gold catalysts were for the first time studied in oxidative dehy-
drogenation of a lignan, hydroxymatairesinol (HMR) to another lig-
nan, oxomatairesinol (oxoMAT), which is a promising antioxidant
and anticarcinogenic compound for pharmaceutical and cosmetic
products, as well as a color-keeping agent in textile industry. De-
spite a complicated chemical structure of the substrate, the reac-
tion in essence is a transformation of secondary alcohol to a
corresponding ketone. Activity and selectivity of gold catalysts
were compared with those of palladium catalysts which exhibited
selectivity to oxomatairesinol (oxoMAT) of 70%. Investigation of
the reaction showed that the activity of utilized gold catalysts
strongly depends on several parameters, such as the catalyst
support, gold oxidation state, solvent, and gaseous atmosphere.
[16] F. Kawamura, M. Miyachi, S. Kaway, J. Wood Sci. 44 (1998) 47.

64
O.A. Simakova et al. / Journal of Catalysis 282 (2011) 54–64
[17] P. Eklund, A. Lindholm, J.-P. Mikkola, A. Smeds, R. Lehtilä, R. Sjöholm, Org. Lett.
5 (4) (2003) 491.
[18] H. Markus, A.J. Plomp, T. Sandberg, V. Nieminen, J.H. Bitter, D.Yu. Murzin, J.
Mol. Catal. A. 274 (2007) 42.
[29] J. Radnik, C. Mohr, P. Claus, Phys. Chem. Chem. Phys. 5 (2003) 172.
[30] G.C. Bond, in: B. Imelik, C. Naccache, G. Courier, H. Praliaud, P. Meriaudeau, P.
Gallezot, G.A. Martin, J.C. Vedrine (Eds.), Metal-Support and Metal Additive
Effects in Catalysis, Elsevier, Amsterdam, 1982 (Chapter 1).
[19] O.A. Simakova, P.A. Simonov, A.V. Romanenko, I.L. Simakova, React. Kinet.
Catal. Lett. 95 (1) (2008) 3.
[31] E. Selli, L. Forni, Micropor. Mesopor. Mater. 31 (1999) 129.
[32] L. Prati, F. Porta, Appl. Catal. A 291 (1–2) (2005) 199.
[20] S. Ivanova, V. Pitchon, Y. Zimmermann, C. Petit, Appl. Catal. A 298 (2006) 57.
[21] E.V. Murzina, A.V. Tokarev, K. Kordás, H. Karhu, J.-P. Mikkola, D.Yu. Murzin,
Catal. Today 131 (2008) 385.
[22] F. Porta, L. Prati, M. Rossi, S. Coluccia, G. Martra, Catal. Today 61 (2000) 165.
[23] M. Zaki, M. Hasan, F.A. Al-Sagheer, L. Pasupulety, Colloids Surf. A 190 (2001)
261.
[33] S. Demirel, P. Kern, M. Lucas, P. Claus, Catal. Today 122 (3–4) (2007) 292.
[34] J. Gong, D.W. Flaherty, T. Yan, C.B. Mullins, ChemPhysChem 9 (17) (2008) 2461.
[35] M. Käldström, N. Kumar, T. Heikkilä, M. Titta, T. Salmi, D.Yu. Murzin,
ChemCatChem 2 (5) (2010) 539.
[36] J. Clayden, N. Greeves, S. Warren, P. Wothers, Organic Chemistry, Oxford
University Press, Oxford, 2000. p. 1508.
[24] H. Markus, P. Mäki-Arvela, N. Kumar, N.V. Kul’kova, P. Eklund, R. Sjöhölm, B.
Holmbom, T. Salmi, D.Yu. Murzin, Catal. Lett. 103 (2005) 125.
[25] S. Ardizzone, C.L. Bianchi, B. Vercelli, Colloids Surf. A 144 (1–3) (1998) 9.
[26] P.A. Simonov, V.A. Likholobov, in: A. Wieckowski, E.R. Savinova, C.G. Vayenas
(Eds.), Catalysis and Electrocatalysis at Nanoparticle Surfaces, Marcel Dekker,
New York, 2003, p. 409.
[27] P. Claus, A. Bruückner, C. Mohr, H. Hofmeister, J. Am. Chem. Soc. 122 (11)
(2000) 430.
[28] A. Sanchez, S. Abbet, U. Heiz, W.-D. Schneider, H. Häkkinen, R.N. Barnett, U.
Landman, J. Phys. Chem. A 103 (1999) 9573.
[37] H. Liang, H. Gao, Q. Kong, Z. Chen, J. Chem. Eng. Data 51 (2006) 119.
[38] M. Sommovigo, H. Alper, J. Mol. Catal. 88 (1994) 151.
[39] A. Aouissi, S.S. Al-Deyab, H. Al-Shahri, Molecules 15 (2010) 1398.
[40] X. Yang, X. Wang, C. Liang, W. Su, C. Wang, Z. Feng, C. Li, J. Qiu, Catal. Commun.
9 (2008) 2278.
[41] A. Abad, C. Almela, A. Corma, H. Garcia, Tetrahedron 62 (2006) 6666.
[42] A. Taskinen, P. Eklund, R. Sjöholm, M. Hotokka, J. Mol. Catal. A 333 (2010) 136.
[43] F. Moreau, G.C. Bond, A.O. Taylor, J. Catal. 231 (2005) 105.
[44] Y. Guan, E.J.M. Hensen, Appl. Catal. A 361 (2009) 49.
[45] S. Cai, K. Sohlberg, J. Mol. Catal. A 193 (2003) 157.