Sie sind auf Seite 1von 9

Journal of Catalysis 311 (2014) 144152

Contents lists available at ScienceDirect

Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat

Active sites of Ni2P/SiO2 catalyst for hydrodeoxygenation of guaiacol:


A joint XAFS and DFT study
Ji-Sue Moon a, Eung-Gun Kim b, Yong-Kul Lee a,
a
b

Laboratory of Advanced Catalysis for Energy and Environment, Department of Chemical Engineering, Dankook University, 126 Jukjeondong, Yongin 448-701, Republic of Korea
Department of Polymer Science and Engineering, Dankook University, 126 Jukjeondong, Yongin 448-701, Republic of Korea

a r t i c l e

i n f o

Article history:
Received 2 October 2013
Revised 19 November 2013
Accepted 22 November 2013
Available online 22 December 2013
Keywords:
Ni2P catalyst
Hydrodeoxygenation
Guaiacol
XAFS
DFT

a b s t r a c t
A Ni2P/SiO2 catalyst was prepared by temperature-programed reduction (TPR), and applied for the hydrodeoxygenation of guaiacol. The physical properties of the catalyst samples were characterized by N2
adsorption/desorption isotherms and CO uptake chemisorption. X-ray diffraction (XRD) and extended
X-ray absorption ne structure (XAFS) spectroscopy were used to obtain structural properties for the supported Ni2P catalysts. Hydrodeoxygenation (HDO) tests were performed in a continuous ow xed-bed
reactor at 523573 K, and 1 or 8 atm, and an LHSV of 2.0 h1. The Ni2P/SiO2 gave an HDO conversion over
90% with two different reaction pathways being identied; at 1 atm direct deoxygenation was dominant
to produce benzene, and at 8 atm prehydrogenation followed by deoxygenation was preferred to produce
cyclohexane. A combined X-ray absorption ne structure spectroscopy and density functional theory
analysis revealed that the active site of Ni2P catalysts is composed of threefold hollow Ni and P sites
which lead to adsorption of H or OH groups. These results suggest that relative populations of H or OH
groups on Ni or P sites of Ni2P surface have an impact on overall reaction pathways of the HDO.
2013 Elsevier Inc. All rights reserved.

1. Introduction
Bio-oils produced from the pyrolysis of biomass have a potential as an alternative fuel and petrochemical source and have several advantages over fossil fuels as a clean source of energy. Biooils are CO2-neutral and free of SOx emission upon combustion.
However, bio-oils cannot be directly used as fuel because of high
oxygen contents (4550%), which can cause the poor stability
and low volatility of fuel [1,2]. The oxygen can be catalytically removed via two pathways; catalytic cracking and hydrodeoxygenation (HDO). The catalytic cracking occurs at atmospheric
pressure condition and high temperature (>450 C) over acidic zeolite catalysts, but the cracking gives a lower yield of transport fuel
product due to the high amount of coke formation, reaching 50
60% [3]. Gayubo et al. studied catalytic cracking of bio-oil over
ZSM-5 zeolite and showed that phenol and guaiacol had low reactivity for hydrocarbons formation; deposition of a coke formed
thermally by condensation of 2-methoxyphenol was noticeable
[35]. The hydrodeoxygenation is generally conducted at high
pressure and temperature and results in high product yield with
high H/C ratios. As the reaction condition is similar to the traditional hydrotreating process, it can be grafted onto the existing
system [6].

Corresponding author.
E-mail address: yolee@dankook.ac.kr (Y.-K. Lee).
0021-9517/$ - see front matter 2013 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcat.2013.11.023

Traditionally, the hydrotreating catalysts are composed of metal


suldes like NiMoS/Al2O3 or CoMoS/Al2O3 [719]. However, these
catalysts can be poisoned in the lack of sulfur in the feed stream.
In other words, the active sites of the catalysts can be oxidized
and deactivated when the bio-oil without sulfur is applied as feedstock [13]. Therefore, alternative catalysts for HDO of bio-oil are
necessary to replace traditional metal sulde catalysts [2029].
Very recently, non-sulded metal phosphides such as Fe2P, Co2P,
Ni2P, MoP and WP have been tested for the HDO of guaiacol, and
Ni2P catalysts have shown the best activity in the HDO of guaiacol
among the metal phosphides [23,29].
In the present study, our attention is placed on investigating the
effect of reaction conditions on the catalytic activities and the
structural property of Ni2P by using X-ray absorption spectroscopy
(XAS). We also used density functional theory (DFT) calculations to
examine the possible structure and energetics of the phosphide
overlayers on the Ni2P (0 0 1) surface resulting from H2O dissociative adsorption.

2. Experimental
2.1. Synthesis of Ni2P catalysts
Supported Ni2P catalysts were prepared by incipient wetness
impregnation of aqueous metal phosphate precursors. The initial
Ni/P ratio in precursor was xed at 1/2. The amount of Ni loading

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

was maintained at 1.5 mmol/g of support using a commercial silica


(Cabot, Cab-O-Sil, M5, 200 m2 g1). The supported nickel phosphate precursor was synthesized by incipient wetness impregnation with a solution of nickel nitrate, Ni(NO3)26H2O (Alfa Aesar,
98%) and ammonium phosphate (NH4)2HPO4 (Samchun, 99%), followed by drying at 393 K for overnight and calcination at 673 K for
4 h. The resulting precursor phosphates were reduced to the corresponding phosphides at 873 K for 2 h using a hydrogen ow at
100 cm3 min1/g of sample. After reduction, the phosphides were
cooled to room temperature and typically were passivated under
a 0.2% O2/He ow for 4 h.
2.2. Catalyst characterization

145

reaction, the passivated catalysts were re-reduced at 723 K using


a hydrogen ow of 100 cm3 min1 for 2 h. The feed was prepared
by dissolving 1 wt% of guaiacol (Alfa Aesar, 98+%) in tridecane
(TCI, >98%) and was injected using a liquid pump at a liquid hourly
space velocity (LHSV) of 2.0 h1 along with 100 cm3 min1 of
hydrogen ow. The reactions were carried out at different temperatures of 523, 553 or 573 K, and pressures of 1 or 8 atm. Liquid
product compositions were analyzed with a HP-6890A gas chromatograph, equipped with a 60 m dimethylsiloxane column having 0.32 mm i.d. (Hewlett Packard, HP-1) and ame ionization
detector (FID), on samples collected at 34 h intervals. The products were identied with a GCMS (Agilent Technologies, 6890A,
HP-1 capillary column).

CO chemisorption uptake measurements were performed on


passivated, air-exposed samples re-reduced in hydrogen at 723 K
for 2 h to examine the dispersion of Ni2P particles on the support.
Pulses (100 ll) of CO at room temperature (300 K) were passed
over the sample to measure the total dynamic gas uptake. A
Micromeritics ASAP 2010 micropore size analyzer was used to
measure the N2 adsorption/desorption isotherms at 77 K, and the
specic surface area of the sample was calculated from the linear
portion of BrunauerEmmettTeller (BET) plots (P/Po = 0.01
0.10). The chemical composition of the samples was determined
by inductively coupled plasma-atomic emission spectroscopy
(ICP-AES) (PerkinElmer, Model Optima-4300 DV). X-ray diffraction
(XRD) patterns of the samples were analyzed using a diffractometer (Rigaku DMAX-2500) operated at 60 kV and 300 mA with Cu Ka
radiation (k = 0.15418 nm). Infrared spectra of phenol adsorbed on
the Ni2P/SiO2 were collected using a Fourier transform infrared
spectrometer (Nicolet 8700, Thermo Scientic) equipped with a
heatable in situ transmission cell with water-cooled KBr windows.
The samples were pressed into self-supporting wafers with a diameter of 20 mm and a weight of 13 mg. IR spectra were collected in
absorbance mode at a resolution of 2 cm1 and 64 scans in the region 4000400 cm1. In situ FTIR measurements on Ni2P/SiO2 were
made in a gas mixture of 5%H2/He. The Ni2P/SiO2 samples were
pretreated in the 5%H2/He ow at 723 K for 2 h. For the measurements of phenol adsorption, the samples were exposed to a gaseous phenol ow vaporized via a bubbler (1% phenol/hexane) at
353 K with the carrier at a ow rate of 150 cm3 min1 until saturation was achieved. The samples were outgassed under vacuum
(106 Torr) for 20 min at 353 K to remove gaseous and weakly adsorbed phenol. Measurements were made at 0, 5, 10, 15 and
20 min with outgassing (106 Torr) at 353 K.

DFT calculations were carried out to model the Ni2P surface.


First, the unit cell of bulk Ni2P was built based on crystallographic
data [30], the geometry of which was then optimized under threedimensional periodic boundary conditions to further rene the
structure. The model phosphide surface was constructed by using
the standard slab approach, in which a slab of nite thickness
was cut out of the Ni2P crystal at the (0 0 1) plane to expose an
atomic layer that has two types of Ni sites, denoted as Ni(I) and
Ni(II), as shown in Fig. 1. The rst type of Ni forms a trigonal prism
around each P atom on the corners of the unit cell, and the second
type of Ni forms another trigonal prism around the P atom located
inside the unit cell. Fig. 2 shows the supercell model used in the
present work. The slab consists of ve layers, each containing both
Ni and P atoms, and four surface unit cells, along with a 15 -thick
vacuum layer in the z-direction (perpendicular to the surface) to
minimize interactions between neighboring image slabs. Two bottommost layers were kept xed during geometry optimization to
simulate bulk constraints [3133].
All DFT calculations were performed using the DMol3 module
as part of the Material Studio package from Accelrys (version 5.5)
[35]. The double-numerical plus polarization (DNP) basis set with
a real space cutoff radius of 4.5 and PW91 exchangecorrelation
functional were used. Effective core potentials were used to treat
the core electrons of nickel.
Relative energies of the surface with different oxygen species
are calculated with similar methods previously used for sulde
catalysts [33,34]. For the adsorption of H2O, OH, and H on a clean
Ni2P surface, the energy changes were calculated using the following equations:

2.3. X-ray absorption ne structure (XAFS) studies

DEH2 O Ehydroxylated surface  Eclean surface

XAFS spectra at the Ni K-edge (8.333 keV) of reference and catalyst samples were recorded in the energy range 8.2339.283 keV
using synchrotron radiation at the beamline 8C, Pohang Light
Source (PLS). The X-ray ring at the PLS has a ux of 1  1010 photons s1 at 100 mA and 2.5 GeV. The X-ray beamline is equipped
with a Si (1 1 1) channel-cut monochromator and has an energy
range capability of 433 keV. The samples were prepared in a kapton sealed glass cell to avoid air-exposure. The X-ray absorption
spectra were recorded at ambient temperature in transmission
mode using ionization chambers for the detection of primary (I0,
100% N2) and transmitted (IT, 100% N2) beam intensities. The obtained XAFS data were analyzed by Winxas 3.1.
2.4. Activity test for HDO of guaiacol
Hydrodeoxygenation of guaiacol was measured in a packed bed
reactor at atmospheric pressure and 8 atm for 25 h. The supported
catalyst (0.326 g, Ni2P/SiO2) was packed in the reactor. Before the

2.5. Computational methods

 Efree H2 O
1
DEOH Ehydroxylated surface EFree H2
2
 Eclean surface  Efree H2 O

DEH Ehydroxylated surface  Eclean surface


1
 Efree H2
2

The thermodynamic equilibrium compositions were calculated


by the minimization of the Gibbs free energy, which was made
by RGibbs module in Aspen Plus 2006. The model compounds of
guaiacol, phenol, cyclohexane anisole, phenol, cyclohexanol, and
methoxy-cyclohexane were chosen as target compounds considered in the thermodynamic equilibrium calculation for the hydrodeoxygenation. The reaction temperature ranging from 300 to
1200 K, the hydrogen to oxygen ratio (H2/O ratio) ranging from

146

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

(A)

(B)

Fig. 1. Unit cell of Ni2P: (A) Ni2P (1 1 1), (B) top view.

Ni

(A)

(B)

Fig. 2. Supercell models of Ni2P surface: (A) Ni2P (0 0 1), (B) side view of 5-layered catalyst slab.

0.5 to 30, and the pressure ranging from 1 to 100 atm were applied
for the thermodynamic equilibrium calculation. The results are
presented in Supplementary materials.
3. Results and discussion
3.1. Physical properties of catalysts
The physical properties of the prepared catalyst samples are
summarized in Table 1. The supported Ni2P catalyst samples
underwent a loss in the BET surface area due to the pore occupation by the Ni2P loadings. The surface area remained unchanged
during the reaction.

Compared to the initial P/Ni molar ratio of 2.0/1.0 in the oxidic


precursor, elemental analysis showed that there was a decrease in
phosphorus content in the course of TPR for the fresh sample,
which gave the P/Ni molar ratio of 0.91/1.0. During the reaction,
a further loss of P was observed to give the P/Ni molar ratio of
0.64 and 0.62/1.0 for the spent Ni2P/SiO2 catalysts evaluated at 1
and 8 atm, respectively. The results indicate that the extra P employed in the oxidic precursor might have been removed via the
formation of PH3 during the course of TPR and hydrotreating condition, which corresponds to the Ni2P stoichiometry, as also suggested by previous reports [36,37]. The experimental CO uptake
for both spent samples was smaller than that for the fresh Ni2P
sample, and the decrease in CO uptake was found to be 32% and

Table 1
Physical properties and elemental analysis of the Ni2P catalyst samples.
Sample

CO uptake (lmol g1)

BET surface area (m2 g1)

Ni (wt%)

P (wt%)

Molar ratio P/Ni

Fresh
Spent (1 atm)
Spent (8 atm)

63.9
31.3
57.8

119.4
77.2
90.2

5.9
5.6
5.8

3.0
1.9
1.9

0.91/1
0.64/1
0.62/1

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

147

9% for the Ni2P samples tested at 1 and 8 atm, respectively. This


indicates that the surface active site of the catalyst is possibly
blocked by some species that prevent adsorption, such as carbon
or oxygen. The loss of active sites is more pronounced for the catalyst tested at 1 atm.
In order to conrm the phase of the fresh and spent Ni2P/SiO2
samples after HDO tests at atmospheric and pressured reaction
condition, XRD was measured as shown in Fig. 3. The XRD patterns
for the all Ni2P/SiO2 samples show three main peaks centered at
40.5, 44.8 and 47.5 corresponding to the Ni2P phase with little
changes in the peak positions after reaction. These results demonstrate that the Ni2P phase remains stable during the reaction.

0.150.30 nm, the shorter distance peak due to the NiP bond
and the longer distance peak to the NiNi bond [38]. The spent
samples show two distinct peaks located at almost the same positions as the fresh samples, indicating the maintenance of the Ni2P
phase during the reaction. For the sample tested under atmospheric reaction condition, the oscillation at 0.40.45 nm becomes
a little wider than the case of the spent sample tested under the
pressured reaction condition, which implies that the spent samples
tested at atmospheric pressure condition are slightly oxidized during the reaction.

3.2. XAFS studies

It has been suggested that the possible reaction mechanism of


guaiacol HDO involves a consecutive reaction pathway from guaiacol to catechol and then to phenol, which then can be transformed
to benzene or cyclohexene over conventional sulde catalysts [5
19]. More recently, the possibility of direct hydrogenolysis producing phenol from guaiacol by elimination of a methanol molecule
was also suggested in the presence of carbon-supported CoMoS
catalysts [17,18]. In a similar manner, Ni2P catalysts mainly produce phenol and benzene without formation of catechol in the
course of HDO of guaiacol [23].
Fig. 5 presents the catalytic activities in the HDO of guaiacol
over Ni2P/SiO2 at 573 K and at 1 or 8 atm. At 1 atm the HDO of
guaiacol gave benzene as the major product with a selectivity of
62% and anisole, cyclohexane, and phenol as minor products with
product selectivities of 30%, 7%, and <1%, respectively, with a guaiacol conversion of 83%. It is noteworthy that the formation of catechol, susceptible to coke formation, was not observed during the
reaction. These results are consistent with a previous work reported by Oyama et al. for the HDO over Ni2P/SiO2 catalyst [23]. Instead, at 8 atm the HDO of guaiacol gave rise to the formation of
cyclohexane (91%) as a major product and benzene (8%), anisole
(1%), phenol (<1%), and traces of cyclohexanol and methoxy-cyclohexane (<0.1%) as minor products, with a guaiacol conversion of
100%. Again, the formation of catechol was not observed. Additional activity tests at 4 and 15 atm also followed the pressuredependent behavior of the HDO as shown in Fig. S1 (Supplementary materials). The increase in pressure from 1 to 4 atm led to
the decrease in the formation of benzene from 62% to 17% with
the increase in the formation of cyclohexane from 7% to 72%. At
even higher pressure of 15 atm, these trends became more pronounced to yield cyclohexane of 99% (Fig. S1).
In order to conrm the pressure-dependent behavior of the
guaiacol HDO, thermodynamic equilibrium calculation was carried
out with varying reaction temperature and pressure, as presented
in Fig. S2 (Supplementary materials). Although the formation of
reaction products such as anisole, phenol, cyclohexanol, and methoxy-cyclohexane are possible, these species are found far less stable at the reaction condition than benzene or cyclohexane (Fig. S2).
At the lowest temperature of 300 K, cyclohexane appears dominant
and the equilibrium composition is kept constant until the temperature of 450 K at 1 atm, at which the formation of benzene is initiated and reached a maximum at 650 K at 1 atm. Upon the increase
in pressure from 1 to 100 atm, the thermodynamic equilibrium
composition of cyclohexane is found to shift to higher temperature
and remain dominant until 500 K at 20 atm and 600 K at 100 atm,
while that of benzene begins to increase exceeding 500 K at 20 atm
and 600 K at 100 atm (Fig. S2). These results indicate that at a lower temperature, the formation of cyclohexane is more thermodynamically favorable than benzene at 1 atm and the equilibrium
composition can shift toward higher temperature region with the
increase in pressure. The thermodynamic equilibrium calculations
are perfectly in line with the HDO tests with varying pressure conditions, demonstrating that the Ni2P catalyst is very active in the

Fig. 4 shows the Ni K-edge XAFS spectra of the fresh and spent
Ni2P/SiO2 catalyst samples collected after atmospheric or pressured reaction conditions. The X-ray absorption near edge structure (XANES) spectra of the fresh samples were measured after
being reduced at 673 K. The analysis of XANES region can provide
information on the symmetry and the oxidation state of the Ni absorber. The pre-edge intensity is related to the symmetry and the
occupancy of the 3d shell. The Ni K-edge XANES spectra show
two regions, which exhibit a pre-edge and main edge peak. The
pre-edge peak corresponds to the transition of a photoelectron
from the 1s to 3d orbital and the main edge peak corresponds to
the transition of 1s to 4p symmetry levels [36,38]. The XANES spectra of the spent samples retain a sharp absorption peak from the
1s4p transition, which is generally called a white line. Notably,
the spent sample collected after an atmospheric condition gives a
stronger white line than the case of pressured reaction condition,
which implies that the Ni2P/SiO2 underwent the surface oxidization particularly under atmospheric condition. These results are
in line with the lower CO uptake for the spent Ni2P catalyst tested
at 1 atm than the spent sample at 8 atm.
Shown also in Fig. 4 are the Ni K-edge EXAFS spectra and their
Fourier transforms for the fresh and spent Ni2P/SiO2 catalysts after
atmospheric and pressured reaction conditions. The EXAFS spectrum for bulk Ni2P as the reference comprises a little wider oscillation region in 30.080.0 nm1 due to the NiP contribution and a
narrower oscillation region in 80.0140.0 nm1 due to the NiNi
contribution, which give rise to two main peaks in the Fourier
transforms, centered at 0.175 nm and 0.240 nm, corresponding to
NiP and NiNi, respectively. Also for the fresh Ni2P/SiO2 sample,
the Fourier transform gives two distinct peaks in the region

Signal Intensity / a.u.

Spent Ni2P/SiO2 (8 atm)

Spent Ni2P/SiO2 (1 atm)


Fresh Ni2P/SiO2

Ni2P (JCPDS 3-953)


20

40

60

80

2 theta / degree
Fig. 3. XRD patterns for fresh and spent Ni2P/SiO2 catalyst samples.

3.3. HDO of guaiacol

148

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

Spent Ni2P/SiO2
(8 atm)

Spent Ni2P/SiO2
(1atm)

x(k)*k3

Spent Ni2P/SiO2
(1atm)

Fresh Ni2P/SiO2

Fresh Ni2P/SiO2

FT Magnitude / a.u.

Normalized absorbance / a.u.

Spent Ni2P/SiO2
(8 atm)

Bulk Ni2P

Ni2P(Bulk)
8.30

8.32

8.34

8.36

8.38

8.40

Photon Energy / keV

10

12

14

k /10 nm-1

Distance / 0.1 nm

Fig. 4. Ni K-edge XANES, EXAFS, Fourier transforms of fresh and spent Ni2P catalysts.

60

Benzene
Cyclohexane
Anisole
Phenol

40

60
40

20

20

0
5

10

15

20

25

30

Time on stream / h

Product distribution / %

80

80

(B) 8 atm

100

X HDO / %

Product distribution / %

100

100

80

Benzene
Cyclohexane
Anisole
Phenol

60

80
60

40

40

20

20

0
5

10

15

20

25

X HDO / %

(A) 1 atm

100

30

Time on stream / h

Fig. 5. Activity test for guaiacol HDO at 1 atm (left) and 8 atm (right) at 573 K.

HDO with satisfying the thermodynamic equilibrium composition


at the given reaction conditions.
Fig. 6 shows the conversion and product distribution at various
reaction temperatures, 523, 543, and 573 K, at an LHSV of 2.0 h1.
Again, at different pressure conditions, the products are quite different with a slight change in product selectivity with reaction
temperature. At 1 atm, the major product was benzene, followed
by phenol, anisole, and cyclohexane. As the temperature increased,
the conversion of guaiacol also rose from 78% to 87% and the product selectivity of benzene also increased from 54% to 62%. The
selectivity of cyclohexane then decreased from 24% to 8% and the
formation of anisole increased from 22% to 30%. The formation of
phenol appeared at the beginning of the reaction but disappeared
at the steady state. At 8 atm, in contrast, the major product for
the HDO of guaiacol was cyclohexane, followed by anisole and benzene. As the temperature increased, the conversion of guaiacol rose
to reach 100%. At 573 K, guaiacol was converted almost exclusively
to cyclohexane (90%), with the amount of anisole down from 11%
to 2%. These results clearly indicate that the Ni2P catalyst can provide parallel reaction pathways of direct hydrodeoxygenation

(DDO) and pre-hydrodeoxygenation (HYD), which give rise to the


formation of benzene and cyclohexane, respectively.
The catalytic stability was tested in the HDO with varying temperature steps of 573, 543, and 573 K in series at 1 or 8 atm, as presented in Figs. S3 and S4 (Supplementary materials). At 1 atm, the
decrease in reaction temperature lowered the HDO conversion
from 80% to 60%, but it was not fully recovered upon the temperature ramping, indicating that the catalyst underwent deactivation
probably due to the partial oxidation of the catalyst surface or coke
deposition. At 8 atm, the deactivation behavior was not observed
and the HDO activity was recovered upon the increase in temperature. These results suggest that the hydrogenation pathway is less
susceptible toward poisoning or coke deposition than the direct
deoxygenation pathway.
3.4. DFT calculations
Hydrodeoxygenation is generally conducted over traditional
hydrotreating catalysts which consist of metal suldes like NiMoS/Al2O3 or CoMoS/Al2O3 [519]. It is commonly suggested that

149

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

100

Benzene
Anisole

80

60

60

40

40

20

20

20

60

60

40

40

20

0
543

(B) 8 atm

80

80

573

Temperature /K

XHDO / %

Cyclohexane

80

523

100

100

(A) 1 atm

Product distribution / %

Benzene
Cyclohexane
Anisole

X HDO /%

Product distribution /%

100

0
523

543

573

Temperature / K

Fig. 6. Product distribution of guaiacol HDO at 1 atm (left) and 8 atm (right) at 523, 543, and 573 K.

the active sits of the catalysts are coordinatively unsaturated sites


at the edges of MoS2 slabs [39,40]. In addition, the surface conguration of the catalyst is known to strongly affect the reaction [39
42]. Delmon proposed that the active site for hydrogenation consists of threefold coordinatively unsaturated Mo atoms, and that
of hydrogenolysis is composed of coordinatively unsaturated Mo
atoms with neighboring SH or H groups which can be formed by
adsorption and dissociation of H2S [41]. Bunch and Ozkan also proposed the role of surface OH group in the HDO and revealed
hydrogenolysis function similarly to SH groups in sulde catalysts
[42]. In order to better understand the active site conguration of
Ni2P catalyst upon the production of H2O during the HDO, we
investigated three different congurations, namely, OH adsorption,
H2 dissociative adsorption, and H2O dissociative adsorption.
OH adsorption on the Ni2P (0 0 1) surface was calculated for four
possible bonding sites, Ni(I), 3Ni(I), P(II), and Ni(I)/P(II), as shown
in Fig. 7. The adsorption of an OH group on the Ni(I) site was stabilized with a bond length of 1.345 and an adsorption energy
of 0.238 eV, as shown in Fig. 7(A). The adsorption of the OH group
on three neighbor Ni(I) sites that form a threefold hollow (TFH), as
shown in Fig. 7(B), was calculated to have a NiO bond length of

1.776 and an adsorption energy of 0.501 eV, indicating a more


stable conguration than the adsorption on the single Ni(I) site.
The adsorption energies of the OH group on the P(II) site and Ni(II)
site were 0.163 eV (bond length: 1.676 ) and 0.303 eV, respectively, indicating less stable congurations than the others. In addition, the OH adsorption as a bridge between Ni(I) and P(II) sites
was not stable with bond lengths of 1.776 and 2.120 , respectively. These results demonstrate that the TFH-Ni(I) site acts most
favorably for adsorption of the OH group.
The calculated results for H2 dissociative adsorption on the Ni2P
(0 0 1) surface are displayed in Fig. 8. The hydrogen atom, placed
initially on either the Ni or P atom, has migrated to sit on the
TFH Ni(I) site with an adsorption energy of 0.637 eV, becoming
the most stable conguration for atomic hydrogen adsorption on
the Ni2P (0 0 1) surface. Our results show that both H and OH can
adsorb on the TFH-Ni(I) site, with the adsorption of atomic H being
more favorable than the OH group.
H2O dissociative adsorption on the Ni2P (0 0 1) surface was calculated for two possible congurations as shown in Fig. 9. It was
assumed that H2O dissociation results in the formation of adsorbed
H and OH, as with the H2S dissociation proposed by Nelson et al.

Fig. 7. Adsorption of OH on the surface of Ni2P (0 0 1) supercell.

150

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

(A)

(B)

Fig. 8. Adsorption of H on the surface of Ni2P (0 0 1) supercell.

Fig. 9. The structures and energetics for H2O dissociative adsorption.

[33]. One of the possible congurations can be of atomic H on the


TFH-Ni(I) site and OH on the nearest P(II) site which is surrounded
by Ni(II) atoms as shown in Fig. 9(A). The adsorption energy was
0.709 eV, and the bond length of POH was 1.718 and of Ni
H was 1.790 , indicating possible surface conguration of Ni2P
in the course of HDO reaction. In the reversed manner, the adsorp-

tion of the OH group on TFH-Ni(I) site and atomic H on the P(II) site
surrounded by Ni(II) atoms was also calculated. The adsorption energy was 0.638 eV, and the bond lengths of NiOH and PH were
2.025 and 1.595 , respectively, indicating less stable conguration than the case of H on TFH-Ni(I) and OH on P(II) sites. These results are in well accordance with a previous study on in situ FTIR

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

O H

Absorbance / A.U.

SiO2

3747

tion rises again, indicating that the phenol is only weakly


adsorbed on the SiO2 support. On the other hand, the peak assigned
to phenol adsorption on the POH groups on Ni2P persists even
after 20 min of outgassing (106 Torr) at 343 K, indicating that
the interaction of phenol with the POH groups on the Ni2P is
stronger than the SiOH groups on the support. These results thus
suggest that the POH groups on Ni2P catalyst well adsorb phenol
and the vicinity of POH and TFH Ni sites facilitates the HDO.

O H

Ni2P

2964

3399

151

3668
20 min
15 min

3.6. Active sites of Ni2P catalyst in HDO


10 min
5 min

0 min

3900

3600

3300

3000

2700

-1

Wavenumber / cm

Fig. 10. In situ FTIR spectra of phenol adsorption over Ni2P/SiO2 catalyst.

analysis of CO and pyridine adsorption on a hydrotreating Ni2P/


SiO2 catalyst [36], in which Ni centers and POH groups were also
identied for promoting hydrogenation and activation of S or N
compounds, respectively. In the similar manner, it can be thus suggested that the TFH-Ni(I) and neighboring P(II) sites of the Ni2P
surface could play a crucial role in HDO with dissociative adsorption of H2 or H2O during the HDO reaction.
3.5. In situ FT IR studies
In situ FTIR measurements of phenol adsorption as a function of
the outgassing time were carried out to characterize the catalyst
surface properties. For reduced Ni2P/SiO2, the adsorption of phenol
gave rise to IR bands at 3747, 3668, 3399, and 2964 cm1 at 343 K
as shown in Fig. 10. The broad band at 3399 cm1 corresponds to
CH vibrations in phenol, and the peak 2964 cm1 corresponds
to CH vibrations in hexane (solvent). The negative peaks at
3747 and 3668 cm1 correspond to hydroxyl vibrations of SiOH
and POH, respectively [36]. Since a background spectrum is subtracted from all spectra, the decrease in the negative peaks indicates recovery in the OH concentration with outgassing.
Initially, the OH bond on SiOH is strongly developed by its interaction with phenol, but as the phenol desorbs the OH concentra-

From these experimental and DFT calculation results, the reaction schematic over Ni2P catalyst in the HDO of guaiacol can be
proposed as shown in Scheme 1. Overall, it can be noted that direct
hydrodeoxygenation and pre-hydrodeoxygenation occur simultaneously over the Ni2P/SiO2 catalyst, in which the HDO of guaiacol
commonly produces anisole or phenol in the early stage of reaction
without the formation of catechol. At 1 atm, the direct hydrodeoxygenation pathway appears dominant to form benzene. Instead, at
a slightly increased pressure of 8 atm, the prehydrogenation pathway becomes pronounced to give cyclohexane. Based on the DFT
studies together with XAFS and in situ FTIR measurements, the active sites of Ni2P catalysts for the hydrodeoxygenation of guaiacol
can be suggested in terms of relative populations of H or OH groups
on Ni or P sites of Ni2P surface, inuencing overall reaction pathways. It can be thus proposed that the direct deoxygenation
(DDO) pathway is favored by the surface OH groups, while the prehydrogenation (HYD) pathway is preferred on the more reduced
surface of the Ni2P catalyst. Moreover, high dispersion of the active
sites will be of great importance in facilitating the HDO. Studies on
the correlation between the catalytic activity and the metal site
density titrated by the amount of CO uptake for the Ni2P catalysts
in hydrotreating [36,37] revealed that for the sample series with
different amounts of Ni2P loadings on the same surface area SiO2
support and the other series with same amounts of Ni and P precursors on the different surface area SiO2 supports, the more CO
uptake the Ni2P gave, the higher activity it showed in both cases.
4. Conclusion
The Ni2P/SiO2 catalysts demonstrated high and stable activity in
the hydrodeoxygenation (HDO) of guaiacol with a guaiacol conversion over 90% at 523573 K, and 1 or 8 atm, and a LHSV of 2.0 h1.
It was noteworthy that the conversion of guaiacol over the Ni2P
catalyst can be tunable to form benzene at 1 atm or cyclohexane
at 8 atm. The characterization of structural properties of fresh
and spent Ni2P catalysts made by XAFS and XRD analysis revealed

Scheme 1. Proposed reaction schematic of guaiacol HDO over Ni2P catalyst.

152

J.-S. Moon et al. / Journal of Catalysis 311 (2014) 144152

that the local structure of Ni2P phase was maintained at 8 atm but
slightly oxidized particularly at 1 atm although the bulk Ni2P phase
remained stable. DFT calculation for the catalyst surface conguration showed that threefold hollow (TFH) Ni site and neighboring P
site are responsible for adsorption of H and OH. In addition, dissociation of H2O was also possible on the TFH-Ni and P sites. All these
results suggest that the direct deoxygenation pathway is favored
over OH surface to form benzene, and the prehydrogenation pathway over the more reduced surface of Ni2P catalyst to generate
cyclohexane.
Acknowledgments
The authors are grateful for the funding supplied by KIST
(2E2280-11-212) and NRF (2012R1A1A2008651). The authors are
also thankful to Mr. Y.-T. Kwon at SK innovation for his valuable
help in FTIR measurement.
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.jcat.2013.11.023.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]

E. Furimsky, Appl. Catal. A 199 (2000) 147190.


A.V. Bridgwater, Catal. Today 29 (1996) 285295.
D.A. Bulushev, J.R.H. Ross, Catal. Today 171 (2011) 113.
A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Ind. Eng. Chem. Res.
43 (2004) 26102618.
A.G. Gayubo, A.T. Aguayo, A. Atutxa, R. Aguado, J. Bilbao, Ind. Eng. Chem. Res.
43 (2004) 26192626.
G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106 (2006) 40444098.
M.R. de Brimonta, C. Duponta, A. Daudina, C. Geantetb, P. Raybauda, J. Catal.
286 (2012) 153164.
V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B 101 (2011) 239
245.
V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B 101 (2011) 246
255.
B.S. Gevert, J.-E. Otterstedt, F.E. Massoth, Appl. Catal. 31 (1987) 119131.

[11] C. Dupont, R. Lemeur, A. Daudin, P. Raybaud, J. Catal. 279 (2011) 276286.


[12] A. Popov, E. Kondratieva, L. Mariey, J.M. Goupil, J.E. Fallah, J.-P. Gilson, A.
Travert, F. Maug, J. Catal. 297 (2013) 176186.
[13] O.I. Senol, E.-M. Ryymin, T.-R. Viljavam, A.O.I. Krause, J. Mol. Catal. A 277
(2007) 107112.
[14] Y. Romero, F. Richard, S. Brunet, Appl. Catal. B 98 (2010) 213223.
[15] Y. Romero, F. Richard, Y. Reneme, S. Brunet, Appl. Catal. A 353 (2009) 4653.
[16] O.I. Senol, E.M. Ryymin, T.R. Viljava, A.O.I. Krause, J. Mol. Catal. A: Chem. 277
(2007) 107112.
[17] M. Ferrari, R. Maggi, B. Delmon, P. Grange, J. Catal. 198 (2001) 4755.
[18] M. Ferrari, B. Delmon, P. Grange, Carbon 40 (2002) 497511.
[19] A.L. Jongerius, R. Jastrzebski, P.C.A. Bruijnincx, B.M. Weckhuysen, J. Catal. 285
(2012) 315323.
[20] A. Gutierrez, R.K. Kaila, M.L. Honkela, R. Siloor, A.O.I. Krause, Catal. Today 147
(2009) 239246.
[21] A.R. Ardiyanti, A. Gutierrez, M.L. Honkela, A.O.I. Karuse, H.J. Heeres, Appl. Catal.
A 407 (2011) 5666.
[22] W. Wang, Y. Yang, H. Luo, T. Hu, W. Liu, Catal. Commun. 12 (2011) 436440.
[23] H.Y. Zhao, D. Li, P. Bui, S.T. Oyama, Appl. Catal. A 391 (2011) 306310.
[24] J. Sun, A.M. Karim, H. Zhang, L. Kovarik, X.S. Li, A.J. Hensley, J.-S. McEwen, Y.
Wang, J. Catal. 306 (2013) 4757.
[25] C. Zhao, J. He, A.A. Lemonidou, X. Li, J.A. Lercher, J. Catal. 280 (2011) 816.
[26] C. Zhao, S. Kasakov, J. He, J.A. Lercher, J. Catal. 296 (2012) 1223.
[27] X. Zhu, L.L. Lobban, R.G. Mallinson, D.E. Resasco, J. Catal. 281 (2011) 2129.
[28] Y.T. Kim, J.A. Dumesic, G.W. Huber, J. Catal. 304 (2013) 7285.
[29] P. Bui, J.A. Cecilia, S.T. Oyama, A. Takagaki, A.I. Molina, H. Zhao, D. Li, E.R.
Castelln, A.J. Lpez, J. Catal. 294 (2012) 184198.
[30] R.W.G. Wyckoff, Crystal Structures, vol. 1, Interscience Publishers, New York,
1963.
[31] L.C. Grabow, A.A. Gokhalem, S.T. Evans, J.A. Dumesic, M. Mavrikakis, J. Phys.
Chem. C 112 (2008) 46084616.
[32] E.S. Bickford, S. Velu, C.S. Song, Catal. Today 99 (2005) 347357.
[33] A.E. Nelson, M. Sun, A.S.M. Junaid, J. Catal. 241 (2006) 180188.
[34] M. Badawi, J.F. Paul, S. Cristol, E. Payen, Y. Romero, F. Richard, S. Brunet, D.
Lambert, X. Portier, A. Popov, E. Kondratieva, J.M. Goupil, J.E. Fallah, J.P. Gilson,
L. Mariey, A. Travert, F. Maoug, J. Catal. 282 (2011) 155164.
[35] B. Delley, J. Chem. Phys. 113 (2000) 77567764.
[36] Y.-K. Lee, S.T. Oyama, J. Catal. 239 (2006) 376389.
[37] S.T. Oyama, X. Wang, Y.-K. Lee, W.-J. Chun, J. Catal. 221 (2004) 263273.
[38] H.-R. Seo, K.-S. Cho, Y.-K. Lee, Mater. Sci. Eng. B 176 (2011) 132140.
[39] R. Prins, Hydrotreating, in: G. Ertl, H. Knzinger, F. Schth, J. Weitkamp (Eds.),
Handbook of Heterogeneous Catalysis, vol. 6, second ed., Wiley, Darmstadt,
2008.
[40] H. Topse, B.S. Clausen, F.E. Massoth, Hydrotreating Catalysis, Science and
Technology, Springer, Berlin, 1996.
[41] B. Delmon, Catal. Lett. 22 (1993) 132.
[42] A.Y. Bunch, U.S. Ozkan, J. Catal. 206 (2002) 177187.

Das könnte Ihnen auch gefallen