Investigating the multifunctional nature of bimetallic FeMoP catalysts using dehydration and hydrogenolysis reactions

Article abstract of DOI:10.1016/j.apcata.2016.06.011

Herein, we investigate the effects of altering the multifunctional nature of a series of FeMoP catalysts for biomass deoxygenation reactions. Unsupported FeMoP catalysts were synthesized at reduction temperatures between 650?°C and 850?°C. The reduction temperature used during the synthesis of FeMoP altered the surface properties of these materials (i.e., acidity and CO-titrated metal sites) while leaving the bulk crystal structure of FeMoP unaffected. This alteration in the surface properties significantly affected the catalytic performance of FeMoP materials in the deoxygenation reactions. The catalytic performance of the FeMoP catalysts was first investigated using the acid catalyzed, non-aqueous, dehydration of cyclohexanol. Aside from showing high selectivities to cyclohexene (>99%) for all catalysts, the experiments demonstrated that FeMoP-650 showed the highest initial rate due to the largest number of surface acid sites among these catalysts. In addition, these FeMoP catalysts were examined using the hydrodeoxygenation (HDO) of phenol. While all FeMoP catalysts were highly selective to benzene (～90%) at conversions less than 15%, studies at higher conversions of phenol provided evidence that catalysts with a higher amount of acid sites maintained enhanced selectivity to benzene. Furthermore, the stability of FeMoP catalysts for phenol HDO was tested with a 48?h time-on-stream (TOS) study. FeMoP exhibited excellent stability, as evidenced by the retention of the initial reaction rate, selectivity to benzene, and oxidation state of surface species throughout the TOS run.

Full text of DOI:10.1016/j.apcata.2016.06.011

Applied Catalysis A: General 524 (2016) 85–93
Contents lists available at ScienceDirect
Applied Catalysis A: General
journal homepage: www.elsevier.com/locate/apcata
Investigating the multifunctional nature of bimetallic FeMoP catalysts
using dehydration and hydrogenolysis reactions
Dallas J. Rensel, Jongsik Kim, Yolanda Bonita, Jason C. Hicks∗
University of Notre Dame, Department of Chemical and Biomolecular Engineering, 182 Fitzpatrick Hall, Notre Dame, IN 46556, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Herein, we investigate the effects of altering the multifunctional nature of a series of FeMoP catalysts for
biomass deoxygenation reactions. Unsupported FeMoP catalysts were synthesized at reduction temper-
atures between 650 ◦C and 850 ◦C. The reduction temperature used during the synthesis of FeMoP altered
the surface properties of these materials (i.e., acidity and CO-titrated metal sites) while leaving the bulk
crystal structure of FeMoP unaffected. This alteration in the surface properties significantly affected the
catalytic performance of FeMoP materials in the deoxygenation reactions. The catalytic performance of
the FeMoP catalysts was first investigated using the acid catalyzed, non-aqueous, dehydration of cyclo-
hexanol. Aside from showing high selectivities to cyclohexene (>99%) for all catalysts, the experiments
demonstrated that FeMoP-650 showed the highest initial rate due to the largest number of surface acid
sites among these catalysts. In addition, these FeMoP catalysts were examined using the hydrodeoxygena-
tion (HDO) of phenol. While all FeMoP catalysts were highly selective to benzene (∼90%) at conversions
less than 15%, studies at higher conversions of phenol provided evidence that catalysts with a higher
amount of acid sites maintained enhanced selectivity to benzene. Furthermore, the stability of FeMoP
catalysts for phenol HDO was tested with a 48 h time-on-stream (TOS) study. FeMoP exhibited excellent
stability, as evidenced by the retention of the initial reaction rate, selectivity to benzene, and oxidation
state of surface species throughout the TOS run.
Received 26 February 2016
Received in revised form 18 May 2016
Accepted 7 June 2016
Available online 9 June 2016
Keywords:
Hydrodeoxygenation
Dehydration
Bimetallic
Phosphides
Catalyst
Biofuels
© 2016 Elsevier B.V. All rights reserved.
1. Introduction
erential hydrogenation of CAROMATIC-H over CAROMATIC-O and (2)
chemical susceptibility to coking and oxidation resulting from the
Lignocellulosic biomass is an abundant and renewable resource
from which to derive value-added chemicals and hydrocarbons
[1–5]. In particular, the lignin portion of biomass, primarily consist-
ing of phenolic ethers, is the largest source of aromatic compounds
in nature [6]. It has been demonstrated that phenolic ethers can be
catalytically transformed into a wide range of aromatic feedstocks
primarily via deoxygenation reactions (e.g. hydrodeoxygenation
(HDO)) [6]. Many transition metal catalysts, such as sulfides [7,8],
noble metals [9–12], and carbides [13,14], have been studied for
lignin and lignin model compound deoxygenation reactions at
elevated temperatures (e.g., >200 ◦C) [15]. While these catalysts
have shown high activities for these reactions, they suffered from
either low selectivities to aromatic products (i.e., noble metals
[16–18]) and/or low stability during the reaction (i.e., sulfides
[19,20], oxides [21], and carbides [22]) due in part to (1) pref-
oxygenated products.
To obtain a good balance of activity, selectivity, and stability
for catalytic deoxygenation of lignin and its derivatives, the cata-
lysts should provide the following surface properties: (1) acid sites
to aid in the removal of OH functional groups innate to lignin
derivatives [23,24], (2) metal sites to adsorb lignin derivatives and
catalyze HDO [13,25], and (3) coking and oxidation resistance under
the reaction conditions employed [19–21]. Mono-metallic transi-
tion metal phosphides represent an interesting and amenable class
of catalysts that can provide a unique combination of surface acidic
and metallic species active for the deoxygenation of lignin [25–27].
Monometallic phosphide catalysts (e.g., Ni2P, Co2P, MoP, and FeP
[28]), however, have been reported with low yields or low selectivi-
ties toward aromatic products in these reactions and exhibited low
stability due potentially to surface oxidation during the reactions
[28–30]. A highly promising method to modify the surface proper-
ties of phosphide catalysis is to incorporate a second metal into the
mono-metallic structure to form a bimetallic structure. Bimetallic
phosphides have been reported as catalysts with increased selec-
∗
Corresponding author.
E-mail address: jhicks3@nd.edu (J.C. Hicks).
http://dx.doi.org/10.1016/j.apcata.2016.06.011
0926-860X/© 2016 Elsevier B.V. All rights reserved.

86
D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
2. Experimental methods
2.1. Chemicals
All materials were used as-received with no further purifica-
tion: citric acid (Alfa Aesar, 99%), FeNO3·9H2O (Alfa Aesar, 99%),
(NH4)6Mo7O24•4H2O (Alfa Aesar, 99%), (NH4)2HPO4 (Amresco,
98%), phenol (Sigma-Aldrich, 99%), cyclohexanol (Alfa Aesar, 99%),
benzene (Alfa Aesar, 99%), cyclohexene (Alfa Aesar, 99%), cyclohex-
ane (Acros Organics, 99%) decane (Alfa Aesar, 99%), and Davisil®
silica gel (Sigma Aldrich, Grade 635, 60–100 mesh). All gas cylin-
ders (purity of ≥99.995%) were purchased from Airgas: Ar, He, N2,
H2, 1% O2/He, 2% NH3/He, and 30% CO/He.
2.2. Synthesis of FeMoP catalysts
FeMoP catalysts were synthesized according to a slightly
modified procedure we previously reported [27]. FeNO3·9H2O,
(NH4)6Mo7O24·4H2O, and (NH4)2HPO4 were added sequentially to
an aqueous citric acid solution (0.4 M in deionized water) while
achieving a molar ratio of Fe:Mo:P of 1:1:1. The solution was par-
tially dried using a rotary evaporator to form a viscous liquid prior
to being transferred to a drying dish. The viscous solution was fur-
ther dried and calcined under an air atmosphere with a ramp rate
of 1.5 ◦C min−1 to 200 ◦C and held at 200 ◦C for an hour. Subse-
quently, resulting intermediate was ground into a powder and then
subjected to the calcination under the air atmosphere at 550 ◦C
with a ramp rate of 1 ◦C min−1 and held at 550 ◦C for six hours.
The resulting material was a mixture of FePO4 and MoO3. After-
wards, the resulting powder was reduced using a Lindberg Blue M
programmable tube furnace under the following condition: drying
at 100 ◦C for an hour under an Ar atmosphere with the flow rate
of 60 mL min−1 followed by two-step reduction under a H2 atmo-
sphere with the flow rate of 160 mL min−1 (1st step: holding at
260 ◦C for an hour with the ramp rate of 5 ◦C min−1 (between 25 ◦C
to 260 ◦C); 2nd step: holding at target temperatures (i.e., 650, 750,
800, and 850 ◦C) for two hours with the ramp rate of 5 ◦C min−1
(between 260 ◦C to target temperature)). After the reduction, the
resulting FeMoP catalysts were cooled to room temperature under
the H2 atmosphere, passivated using a 1% O2/He with the flow rate
of 60 mL min−1 for an hour, and then transferred to a glovebox
purged with a N2 atmosphere (>99%). The resulting FeMoP cata-
lysts were denoted as FeMoP-XXX, where XXX indicates the target
reduction temperature employed under the 2nd reduction step.
Fig. 1. Schematic representation of the multifunctional properties of FeMoP cat-
alytically deoxygenating two lignin model compounds.
tivity to aromatic products compared to mono-metallic phosphides
in a variety of deoxygenation reactions [29,31–33]. In addition to
retaining surface acidity and surface metal sites, the incorpora-
tion of the second transition metal provides another method to
enhance the chemical, electronic, and surface properties of the
phosphide catalysts [25]. Moreover, the catalytic properties of
bimetallic phosphides can be tuned for improved performance by
altering phosphide synthesis conditions, expanding the application
of these materials to a variety of reactions [27,33]. However, there
has not been a systematic study performed to elucidate the impact
of synthetic conditions on surface properties and HDO catalytic
performance.
In spite of the aforementioned potential of these materials, there
have been few reports detailing the effects of bimetallic phosphide
synthesis conditions on their catalytic surface properties [33]. In
a previous study, we employed bimetallic FeMoP as a catalyst to
cleave CAROMATIC-O bonds of anisole, phenol, and phenethoxyben-
zene, which resulted in the production of deoxygenated aromatic
compounds (e.g. benzene, toluene, ethylbenzene) with high selec-
tivities (∼90%) over the production of saturated compounds (e.g.
cyclohexane) [27]. The FeMoP catalyst also showed good stability
for the deoxygenation of phenol, as it was able to catalyze this trans-
formation for five consecutive reaction cycles with almost no loss
in selectivity [27].
Herein, we varied the FeMoP reduction temperature to inves-
tigate the effects of synthesis conditions on catalytic surface
properties (i.e., the number of surface acid and metal sites) and how
these surface properties subsequently affected catalytic perfor-
mance in deoxygenation reactions. Specifically, the multifunctional
surface properties of the FeMoP catalysts were investigated using
the catalytic deoxygenation of two lignin model compounds, cyclo-
hexanol (with CALIPHATIC-O) and phenol (with CAROMATIC-O)], as
probe reactions. The effect of acid sites present on the FeMoP
surface was probed via the non-aqueous, dehydration of cyclo-
hexanol, whereas the effect of metal sites (i.e. CO-titrated metal
sites) on the FeMoP surface was studied via the HDO of phenol
(Fig. 1). In addition, mechanistic studies were performed in order
to understand how the combination of acid sites and metal sites
affected the product selectivity for the HDO of phenol. In addition
to detailed catalytic studies using these two reactions, the stability
of the FeMoP catalysts was also evaluated via a 48h time-on-stream
(TOS) experiment for the HDO of phenol. From these detailed exper-
iments, it is evident that FeMoP catalysts can have a significant
impact on deoxygenation reactions pertinent to biomass utilization
and biofuels/bio-chemicals production.
2.3. Characterization
X-ray diffraction (XRD) patterns were collected using a Bruker
DaVinci Advanced D8 X-ray diffractometer equipped with a Cu K␣
radiation source (␭ = 1.5418 Å) under a 2␪ range of 20–60◦, a scan
speed of 3.5 s./step, and a size step of 0.02◦/step. Crystallite size was
evaluated using the Scherrer equation shown in Eq. (1), where K is
the shape factor (0.9, based on prior work on phosphides [34]), ␭
is the wavelength of the Cu K␣ source, ␤ is the full width half max
(FWHM) value of the peak, and ␪ is the Bragg angle.
Kꢀ
crystallite size =
(1)
ˇcosꢁ
In situ transmission electron microscopy (TEM) images were
acquired on a FEI Titan 80–300 microscope at 300 keV, equipped
with in situ sample holder along with heating chip (DENS Solu-
tions Corporation). The TEM samples were prepared by sonicating
the catalyst in iso-propyl alcohol for 5 min, placing one drop of the
suspension on the in situ sample holder, and then drying under
vacuum (10−2 torr) at room temperature for 2 h prior to analysis.
NH3 temperature programmed desorption (NH3-TPD) experiments

D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
87
were conducted on a Micromeritics Chemisorb 2750 equipped with
3. Results and discussion
a thermal conductivity detector (TCD) and TPx temperature con-
troller. The catalyst was re-reduced under a H2 atmosphere by
ramping at a rate of 10 ◦C min−1 to 400 ◦C for an hour. Chemisorp-
tion of NH3 was performed at 100 ◦C for 3 h. Subsequently, NH3
desorption was performed in flowing He by ramping at 10 ◦C min−1
to 230 ◦C, holding for 42 min, then ramping at 10 ◦C min−1 to 400 ◦C,
and holding for 42 min. The desorbed area was quantified based on
NH3 calibrations determined using the 2% NH3/He. CO chemisorp-
tion experiments were conducted on the Micromeritics Chemisorb
2750 instrument following the same pre-treatment steps as the
NH3-TPD experiments. Pulse chemisorption of CO was performed
at 35 ◦C using 0.1 mL injections of a 30% CO/He. The CO uptake
on each material was obtained by summing the area of each peak
and subsequently quantified based on CO calibrations determined
using 30% CO/He. Brunauer-Emmett-Teller surface area (SBET) of
the catalyst was analyzed using a Quantachrome Nova 2200e. A
PHI VersaProbe II X-ray Photoelectron Spectrometer was used to
obtain X-ray photoelectron (XP) spectra of the catalysts to analyze
the surface species. The Fe 2p3/2, Mo 3d5/2, and P 2p3/2 peaks were
used to compare binding energies based on a C 1s reference peak
located at 284.5 eV.
3.1. FeMoP catalysts syntheses and characterizations
The FeMoP catalysts were synthesized following a procedure we
previously reported [27] except with varying reduction tempera-
tures ranging from 400 ◦C to 850 ◦C (Fig. S1). This led to a series of
FeMoP catalysts denoted as FeMoP-XXX (where XXX indicates the
reduction temperature). The bulk crystal structures of the resulting
FeMoP catalysts were investigated using X-ray diffraction (XRD).
The purpose of this XRD analysis was to determine the range of
reduction temperatures suitable for the production of the crys-
talline FeMoP structure. Aside from showing a peak assigned to
a FeMoP diffraction (Fig. S2), the catalysts synthesized at 600 ◦C or
below exhibited multiple diffractions attributed to bulk phases of
Mo, MoO2, MoO3, FePO4, and Fe2PO7, and thus were not studied
as part of this work. In contrast, as shown in Fig. S3, the catalysts
synthesized at 650 ◦C or greater showed only diffractions assigned
to the crystalline FeMoP bulk phase, and thus were utilized in this
study. Notably, the increase in reduction temperatures used during
the synthesis of FeMoP catalysts resulted in the continuous growth
of FeMoP crystallites, as evidenced by the increase in average crys-
tallites size from 22 nm (650 ◦C) to 44 nm (850 ◦C) estimated using
the Scherrer equation (Table 1). This crystallite growth was also
consistent to transmission electron microscopy (TEM) images of
FeMoP-550 heated in situ from 650 ◦C to 850 ◦C (Fig. S4). In this
experiment, a single region containing the catalyst was imaged at
different temperatures to show the change in morphology. These
images revealed a significant morphological change at 800 ◦C, likely
resulting from the aggregation of FeMoP crystallites. In agreement
with the trend observed in both XRD and in situ TEM experiments,
the increase in reduction temperatures also led to the decrease in
BET surface areas (SBET) of the resulting FeMoP catalysts as deter-
mined by N2 physisorption experiments (Table 1).
The surface properties were investigated using x-ray photo-
electron spectroscopy (XPS). XPS analysis provided evidence the
surface of the FeMoP catalysts was composed of Fe, Mo, and P
species that were close to their zero valence states or at a higher
oxidation state (Fig. S5). The oxidation state of the species close
to their zero valence state was not clearly assigned due to inter-
atomic effects in phosphide materials that induced a Madelung
potential (electrostatic interaction), as reported for several MP and
M2P materials (M = metal) [35,36]. These XP spectra also showed
the presence of oxidized species such as Fe3+, Mo6+, and P5+ located
at binding energies of 711.1 eV, 232.1 eV, and 133.5 eV, respec-
tively [27,37,38]. The oxidized surface species were likely formed
by either oxidation of the catalyst surface during passivation under
1% O2 in He during synthesis or by the incomplete reduction to the
FeMoP structure, both of which could provide Lewis and Brønsted
surface acid sites (M␦+ as Lewis acid sites and M-OH and/or P-
OH as Brønsted acid sites) on the catalysts. Additionally, several
prior studies that synthesized monometallic phosphide catalysts
reported the evolution of phosphide gas during reduction of the
precursors [26,34,39]. To test the potential loss of P on the FeMoP
catalysts, XPS was used to determine the ratio of metal atoms to P
atoms. The surface composition of P on FeMoP reduced at 650, 750,
800, and 850 ◦C was 34, 32, 32, and 31 at.%, respectively. While it is
possible that some P is removed at higher reduction temperatures,
the surface concentration of P was fairly unaffected.
2.4. Dehydration of cyclohexanol and HDO of phenol
All reactions were performed in a custom-built continuous flow
reactor. For the kinetic studies, the reaction was operated in the
forward reaction-dominant regime with the conversion of 15%
or lower. In a typical experiment, 30 mg of the catalyst particles
were diluted using 100 mg of Davisil® grade 635 and subsequently
packed in a 316 stainless steel reactor equipped with a stainless
mesh steel screen (10 ␮m, McMaster-Carr) and quartz wool (VWR
Scientific) to immobilize the catalysts particles in the reactor. The
reactor was operated in an up-flow configuration. Flow rates of
reactant solutions (0.13 M of cyclohexanol dissolved in decane;
0.13 M of phenol dissolved in decane) were controlled using a high
pressure liquid chromatography (HPLC) pump (Hitachi L-6000),
whereas flow rates of gases (100 mL min−1 of N2 for cyclohexanol;
100 mL min−1 of H2 for phenol) were regulated using a mass flow
controller (Aalborg GFC17). Reaction temperature was set using
PID controllers and maintained at 180–220 ◦C (for cyclohexanol)
and 350–400 ◦C (for phenol) using a heat tape (Omega Engineering)
equipped on the reactor. HDO reactions were operated in an excess
hydrogen flow with the feed molar ratio of H2 to phenol as ∼34.
Catalytic results (i.e., conversion and selectivity) were quantified
using an Agilent 7890A gas chromatograph (GC) directly coupled
with an Agilent 5975C mass spectrometer (MS) based on external
calibration curves of both reactants and products. Throughout all
reaction runs, the carbon balance was 96% or greater. Conversion
and selectivity were calculated based on Eqs. (2) and (3).
moles of reactant converted
conversion =
selectivity =
× 100
(2)
(3)
moles of reactant fed
moles of specific product
moles of all products
× 100
NH3-temperature programmed desorption (TPD) was used to
quantify the surface acid concentration (␳ACID, ␮mol m−2, in
Table 1) present on the FeMoP catalysts (Fig. S6) [29]. Prior reports
investigating NiMoP, MoP, and Ni2P have shown that two types
of acid sites, Brønsted and Lewis, were present on the surface of
phosphide materials [29,30,40]. One report suggested that NH3
desorbed from these two acid sites at different temperatures,
All reported reaction rates are initial rates calculated by using
the rate expression discussed in the text and extrapolating to the
initial solution concentration. The fitted rate law was validated
based on the reactions run at 400 ◦C with varying concentrations
of phenol and varying concentrations of H2 using the FeMoP-650
catalyst.

88
D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
Table 1
Physical properties of the FeMoP catalysts.
c
d
e
catalyst
average crystallite sizea,b (nm)
SBET (m2 g−1
)
␳
(␮mol m−2
)
NCO (␮mol g−1
)
ACID
FeMoP-650
FeMoP-750
FeMoP-800
FeMoP-850
22.0
26.4
39.2
44.2
6.8
2.7
1.8
1.5
3.9 0.4
3.2 0.3
2.2 0.2
1.0 0.1
28.5
2
9.6 0.7
5.3 0.4
2.0 0.2
a
Via Scherrer equation.
Average of values evaluated at (011), (200), and (211).
Via N2 physisorption.
Areal acid concentration (␳ACID) via NH3-TPD and SBET
Via CO-pulsed chemisorption.
b
c
d
e
.
∼210 ◦C for Brønsted acids and ∼320 ◦C for Lewis acid sites [29].
The NH3 TPD plots in Fig. S6 shows two separate peaks, one at low
temperature (∼210 ◦C) and at high temperature (∼310 ◦C), which
suggested that two different types of acid sites were present on
the FeMoP materials. As shown in Table 1, the increase in reduc-
tion temperatures resulted in a decrease in ␳ACID, which suggested
that thermochemical reduction under a H2 atmosphere eliminated
surface acid sites on FeMoP catalysts at elevated temperatures. In
addition, the loss of surface area from sintering observed in in situ
TEM images could also lead to the decrease in ␳ACID. The reduction
in Brønsted acid sites was likely caused by H2 that thermochem-
ically cleaved the OH groups on the surface of MP or M1M2P,
liberating H2O [41–43]. This was suggested based on the phase
transition of metal oxide and phosphate materials, which showed
the gradual elimination of surface Brønsted acid sites at reduction
temperatures of 500 ◦C or greater [41–43]. The Lewis acid sites also
decreased with an increase in reduction temperature. One possi-
ble reason for this decrease could be the heterolytic cleavage of H2
to form M␦+-H␦− surface species at higher reduction temperatures
[44,45]. Other reports on supported metal phosphide catalysts (e.g.,
Ni2P [40]) have clearly shown Brønsted and Lewis acid sites from
Fourier Transform Infrared spectroscopy (FT-IR) analysis. Attempts
to collect spectroscopic data of these acid sites on the unsupported
FeMoP catalysts were unsuccessful because these samples were
dark in color and had a poor spectroscopic reflectivity, which in
turn showed only weak peak intensities in the FT-IR spectra. In
addition to the investigation of ␳ACID innate to the FeMoP catalysts,
the quantity of CO-titrated surface metal sites (i.e. Lewis acid metal
sites and metallic sites [46,47], NCO) in the FeMoP catalysts were
also determined via CO pulsed chemisorption experiments. Sim-
ilar as the trend of ␳ACID, the increase in reduction temperatures
also reduced the quantity of NCO (Table 1). The loss of CO-titrated
surface sites in the FeMoP catalysts at elevated reduction tempera-
tures was likely due to the loss in the surface area from the growth
of FeMoP crystallites (Fig. S4).
reaction-limited regime (XCYCLOHEXANOL ≤15%) and utilized an iden-
tical amount of catalyst for each reaction (30 mg).
A 1st-order reaction assumption was used for the calculation
of reaction rate based on previous reports with solid acid cata-
lysts [16,51]. The initial rate of cyclohexanol consumption for each
catalyst was calculated from the rate and concentration measured
during reaction. The measured rate was used to determine the rate
constant, which was then multiplied by the initial concentration of
cyclohexanol to obtain the initial reaction rate. It should be noted
that neither mass transport limitations nor reaction heat effects
played significant roles during the kinetic evaluation. This was evi-
denced by three control reaction runs with FeMoP-650 at 220 ◦C
(Fig. S7) which showed consistent values of -r0,COH/MASS under the
employed reaction conditions (particle size = 60–100 mesh, flow
rate = 1.0 mL min−1 solution, diluent ratio = 4 parts silica gel to 1
part catalyst), regardless of varying the size of the catalytic par-
ticles, flow rate of cyclohexanol, and amount of the diluent. The
FeMoP-800 catalyst was excluded in this study due to its similar
surface properties (e.g., ␳ACID and SBET in Table 1) to those of FeMoP-
850. As expected, these reactions evolved cyclohexene with high
selectivities (>99%) at all temperatures with all catalysts (FeMoP-
650, FeMoP-750, and FeMoP-850). Values of -r0,COH/MASS for the
FeMoP catalysts showed that the -r0,COH/MASS values decreased
with an increase in the reduction temperatures used to synthesize
the FeMoP catalysts. Furthermore, as stated above, an increase in
reduction temperature also led to lower values of ␳
(Fig. 2a).
ACID
Finally, the initial forward reaction rates when normalized on the
number of acid sites per gram (-r0, COH/ACID) for FeMoP catalysts
reduced at 650, 750, and 850 ◦C were nearly identical at any given
reaction temperature (Fig. 2b). Therefore, surface acid sites were
the active sites for the dehydration of cyclohexanol.
Kinetic data for the dehydration of cyclohexanol at five temper-
atures (180, 190, 200, 210, and 220 ◦C) were evaluated in order to
obtain activation energies (EA) for the FeMoP catalysts. The FeMoP
catalysts not only followed Arrhenius behavior under the employed
reaction temperatures (Fig. 2b) but all materials also exhibited sim-
ilar values of EA (∼130 kJ mol−1) in this reaction. Previous studies
reported that similar values of EA suggested a similar pathway for
this reaction when using acid catalysts (e.g. HZSM-5 [17], ␥-Al2O3
[52], and H3PO4 [17]). This suggests that all FeMoP materials cat-
alyzed the reaction following a similar mechanistic pathway.
3.2. Deoxygenation on the FeMoP catalysts
3.2.1. Dehydration of cyclohexanol
The effects of altering surface properties of the FeMoP on cat-
alytic performance were then investigated using two lignin model
compounds. Previously, surface acid sites (i.e. Brønsted and Lewis
acid sites) in various catalysts have been shown to catalyze the
dehydration of aliphatic alcohols with high selectivities to aliphatic
alkenes [48,49]. Therefore, to determine if the acid sites on the
FeMoP catalysts were accessible and catalytically active, the dehy-
dration of cyclohexanol (a CALIPHATIC-OH compound, denoted as
COH), a model compound that can be generated from the hydro-
genation of lignin derivatives, was studied as a probe reaction
[16,50]. The catalytic performances of the FeMoP catalysts for this
reaction were evaluated using a flow reactor under a N2 atmo-
sphere (5.2 MPa) at 220 ◦C. All reaction runs were operated in the
3.2.2. HDO of phenol
The catalytic role of CO-titrated metal sites (NCO) present on
the surface of the FeMoP catalysts was investigated using another
lignin model compound, phenol (CAROMATIC-O, denoted as PHEN),
as a probe molecule. These CO-titrated metal sites are likely a com-
bination of near zero-valent metal sites [53,54] and Lewis acid sites
(or coordinatively unsaturated sites (CUSs) [30]), both of which
could be active sites for the HDO of biomass model compounds
[30,55,56]. In this study, phenol dissolved in decane was fed to
the reactor packed with the FeMoP catalysts at 400 ◦C and reacted
under a H2 atmosphere (5.2 MPa). Similar to the dehydration of

D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
89
Fig. 2. Initial reaction rate of cyclohexanol dehydration normalized on (a) mass (-r0,COH/MASS) (b) and number of acid sites (-r0,COH/ACID) at several reaction temperatures. Inset
in Fig. 2b shows the fitted Arrhenius plots based on the data plotted in Fig. 2b. The inset in Fig. 2b is a plot of ln(kapp) vs 1000/TReaction (K−1).
cyclohexanol, identical amounts (30 mg) of FeMoP catalysts were
used in each reaction run to compare the initial reaction rate nor-
malized on the mass of each catalysts (-r0,PHEN/MASS) resulting from
operation at XPHENOL ≤∼15%. Again, to ensure that the reactions
were operated in the forward reaction-dominant regime, three con-
trol experiments were performed which demonstrated negligible
mass and heat transport effects during the reaction runs (Fig. S8).
The product selectivity in the forward reaction-dominant regime
favored the production of benzene (∼90% selectivity) for all FeMoP
catalysts at 400 ◦C (Fig. S9).
Phenol HDO was then performed at varying reaction tem-
peratures, and the results are depicted in Fig. 3. A decrease in
-r0,PHEN/MASS was observed on FeMoP catalysts synthesized at
higher reduction temperatures (Fig. 3a). Additionally, as noted
above (Table 1), the NCO sites decreased on the FeMoP catalysts
synthesized at higher reduction temperatures. When these rates
were normalized on NCO (-r0,PHEN/CO, Fig. 3b), all FeMoP catalysts
had similar values of -r0,PHEN/CO, which suggested that the CO-
titrated metal sites were the primary active sites for the HDO of
phenol. The normalization on CO-titrated metal sites was in agree-
ment with previous studies in which surface metals sites (NiMoP
[29] and FeNiP [57]) of bimetallic phosphides were suggested as
active sites for the HDO of model compounds (e.g. anisole and 2-
methyltetrahydrofuran). The -r0,PHEN/CO values for FeMoP-650 were
slightly lower than those for FeMoP-750 and FeMoP-850. This could
be due to the different accessibility of the reactant to active sites of
the much smaller crystallites observed in FeMoP-650 compared to
FeMoP-750 and FeMoP-850 (Table 1 and Fig. S4).
Multiple prior studies have investigated two site mechanisms
for HDO reactions of various CAROMATIC-O model compounds. On
Mo2C surfaces, one distinct metal site has been proposed to coor-
dinate with the aromatic ring in the reactant while another metal
site participates in the CAROMATIC-O bond cleavage [13,58–60]. Sev-
eral other studies have proposed that the O functionality in the
reactant binds to a hydroxyl or acid site while a reduced metal
site facilitates H2 dissociation used to cleave the CAROMATIC-O bond
[24,61–63]. Yet another type of mechanism that has been proposed
for HDO reactions is a Mars-van Krevelen or reverse Mars-van
Krevelen mechanism, which have been reported on oxide catalysts
where surface vacancies form to facilitate the reaction [21,64]. For
the FeMoP catalysts reported here, it is possible that a surface acid
site binds to the OH in phenol to destabilize the Caromatic-O, while
reduced metal surface sites dissociate H2 onto the surface. Using the
above kinetic data in conjunction with reported mechanisms in lit-
erature [61,63], a two site Langmuir–Hinshelwood mechanism was
proposed (Eqs. (4)–(9)), where R, S1, and S2 indicate phenyl (C6H5)
and two surface sites, respectively. H2 first dissociatively adsorbs
on two S1 sites (Eq. (4)), while phenol adsorbs on S2 (Eq. (5)). In
the following step, a surface bound hydrogen species protonates
the absorbed phenol (Eq. (6)), followed by the addition of another
surface H species to form benzene (Eq. (7)). Two assumptions were
made to link this two-site mechanism to values of -r0,PHEN-MASS on
the FeMoP catalysts. Firstly, Eq. (7) was assumed irreversible and
the rate-determining step (RDS) [63]. Eq. (6) could also be the RDS,
but the derived rate expression in Eq. (10) would have the same
form. Secondly, all elementary steps, except Eq. (5), were assumed
to be pseudo-steady state, which provided a site-balance involv-
ing all chemical species adsorbed on both S1 and S2 during the
HDO of phenol. The resulting −rPHEN/MASS based on the aforemen-
tioned assumptions can be expressed in Eq. (10), where k, K1, and
K2 denote reaction rate constant, adsorption coefficient for phenol,
and adsorption coefficient for H2, respectively.
H2 + 2S1 ꢀ 2H − S1
(4)
(5)
(6)
(7)
(8)
(9)
R − OH + S2 ꢀ R − OH − S2
R − OH − S2 + H − S1 ꢀ R − OH2 − S2 + S1
R − OH2 − S2 → OH − S2 + R − H
OH − S2 + H − S1 ꢀ H2O − S2 + S1
H2O − S2 ꢀ H2O + S2
⎡
⎤
ꢀ
ꢃ
1
2
ꢁ
ꢂ
K2C0,H
K1C0,PHEN
⎢
⎣
⎥
⎦
2
ꢁ
ꢂ
−r0,PHEN/MASS = k
(10)
ꢄ
1
2
ꢁ
ꢂ
1 + K1C0,PHEN
1 + K2C0,H
2
Eq. (10) could be further simplified based on an additional study
where -r0,PHEN/MASS was monitored as a function of C0,H2 and C0,PHEN
(Fig. 4).
The initial phenol HDO rate (-r0,PHEN/MASS) showed 0th-order
dependence on the initial hydrogen concentration C0,H2 leading
to the simplified rate expression in Eq. (11). In this simplification,
because the initial hydrogen concentration (C0,H2) used during the
reaction was far greater than the C0,PHEN, the contribution of C0,H2
to -r0,PHEN/MASS was assumed constant and therefore incorporated
into the value of kꢀ (Eq. (11)).
ꢀ
ꢃ
K1C0,PHEN
−r0,PHEN/MASS = kꢀ
(11)
ꢁ
ꢂ
1 + K1C0,PHEN
Having proposed
a
mechanism and derived
a
rate law,
data for -r0,PHEN/MASS of FeMoP-650 (Fig. 4) were fitted to a

90
D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
Fig. 3. Initial reaction rates for phenol HDO normalized on (a) mass (-r0,PHEN/MASS) and (b) CO-titrated metal sites (-r0,PHEN/CO) at various reaction temperatures. Inset in Fig. 3b
shows the fitted Arrhenius plots based on the data plotted in Fig. 3b. The inset in Fig. 2b is a plot of ln(kꢀ) vs 1000/TReaction (K−1).
Fig. 4. Plot of -r0,PHEN/MASS at various initial concentrations C0,H2 (black circles) and
0,PHEN (blue hollow squares) using FeMoP-650 for the HDO of phenol at 400 ◦C. (For
Fig. 5. Initial rate of phenol HDO on FeMoP-650 normalized on mass (-r0,PHEN/MASS
fitted to Langmuir-Hinshelwood kinetic model.
)
C
interpretation of the references to color in this figure legend, the reader is referred
to the web version of this article.)
and, furthermore, the surface metal sites on the FeMoP catalysts
were likely the active sites for this reaction.
Langmuir–Hinshelwood kinetic model with an excellent agree-
ment (R2 ∼ 0.97, Fig. 5).
3.2.3. Mechanistic evaluation
Furthermore, the value of K1 was extracted from this data,
which was determined as 2.0 M−1 at 400 ◦C (Eq. (11)). Consider-
ing that the values of C0,PHEN ranged from 0.05 M to 0.13 M, the
values of resulting K1C0,PHEN were close to unity and could not be
neglected. Based on the fitted data in Fig. 5, the value of kꢀ was
calculated to be 1160 min−1. Based on the rate expression in Eq.
(10), the apparent activation energies (EA) of the FeMoP catalysts
were determined by performing reactions at three additional reac-
tion temperatures (350 ◦C, 380 ◦C, and 390 ◦C). The values of EA
(∼ 100 kJ mol−1, Fig. 3b) were similar for all FeMoP catalysts, which
again implied a similar reaction pathway on these materials. Fur-
thermore, in addition to consistently showing high selectivity to
benzene (SBEN) at all reaction temperatures (Fig. S9), the FeMoP
catalysts also exhibited values of initial forward rates of phenol
normalized on NCO (-r0,PHEN/CO) that were almost invariant at each
reaction temperature (Fig. 3b). The similar EA values and -r0,PHEN/CO
for all FeMoP catalysts suggested that the FeMoP materials cat-
alyzed the HDO of phenol along a similar reaction pathway [17,52]
The reaction pathway on the active sites of the FeMoP catalysts
(CO-titrated metal sites) was investigated by observing the product
selectivities from phenol HDO at different conversions. This was
motivated by our previous batch reactions, using anisole as the
model CAROMATIC-O compound, operated at high values of XANISOLE
(i.e. >25–99%), which showed the FeMoP catalysts synthesized via
the reduction at 750 ◦C or lower led to an increase in aromatic prod-
ucts (e.g. benzene and toluene) compared to the FeMoP analogues
synthesized at higher reduction temperatures [27]. As shown in
Fig. 6, the FeMoP catalysts showed very different values of SBEN at
higher values of XPHEN (e.g., ∼70%).
The SBEN on both FeMoP-650 and FeMoP-750 consistently
reached values of ∼90% throughout the high conversion study.
FeMoP-850, in contrast, showed a decrease in values of SBEN from
∼90% to ∼75%, when increasing the values of XPHEN up to ∼70%.
The change in product selectivity at different conversions of phe-
nol was likely due to surface coverage of the reactant and product
molecules. At lower values of XPHEN the surface is covered primar-

D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
91
previous reports have demonstrated that the key CUSs in molyb-
denum sulfide catalysts, which are similar in surface properties to
phosphide catalysts (i.e. Brønsted, Lewis, and metallic sites), pre-
dominantly formed on edges and rims of planes [68–70]. Thus, a
reduction of edge and rim atoms compared to surface atoms via sin-
tering could also decrease the number of CUSs relative to metallic
surface atoms [68,69]. Therefore, the significant crystallite growth
seen in the Scherrer analysis of XRD (Table 1) and morphology
change observed in the in situ TEM analysis (Fig. S4) could lead
to a reduction in the ratio of edge and rim atoms relative to metal-
lic surface atoms, which would ultimately lead to a decrease in
aromatic selectivity. Of note, it is possible that the water formed
during the HDO of phenol could potentially interact with the sur-
face to influence the acidity and thus the reactivity of the catalysts.
This effect was tested by pulsing pure water for intervals of ∼15 min
during a time-on-stream study while operating at close to 20% phe-
nol conversion using FeMoP-850. From the results shown in Fig.
S10, the addition of water did not significantly affect -r0,PHEN/CO or
SBEN. Therefore, from these studies, the water produced from the
reaction is unlikely to affect the reactivity of the catalysts under
these operating conditions.
The HDO and hydrogenation pathways on the FeMoP catalysts
were evaluated by co-feeding a reactant mixture of phenol and
benzene at 400 ◦C. For these control experiments, FeMoP-650 and
FeMoP-850 were used as the catalysts due to their different SBEN
profiles (Fig. 6). In these reactions, phenol (0.13 M) mixed with
benzene (30 mol% with respect to the concentration of phenol, or
0.039 M) was used as the feed stream. The 0.039 M concentration
of benzene was added in order to mimic the onset of hydrogenated
products that were shown in Fig. 6 (SBEN started to decrease at
XPHEN of ≥20%). Instead of investigating SBEN, the production rate of
cyclohexene and cyclohexane (rALIPHATIC, ␮mol min−1) was calcu-
lated at different residence times (Fig. S11). For the FeMoP-650 and
FeMoP-850 catalysts, the addition of benzene to the feed stream
decreased XPHEN. It is likely that benzene competitively adsorbed
on the surface, which would result in a decreased XPHEN. The addi-
tion of benzene to the feed stream also slightly increased the values
of rALIPHATIC for FeMoP-650 (0.12 ␮mol min−1 to 0.34 ␮mol min−1
at ∼70% conversion) (Fig. S11). Therefore, benzene could be hydro-
genated on the FeMoP-650 catalyst under these conditions. This
result also suggested that phenol was first deoxygenated to ben-
zene and then subsequently hydrogenated to aliphatics through
tandem reactions. The addition of benzene enhanced the values of
rALIPHATIC for FeMoP-850 (0.76 ␮mol min−1 to 2.54 ␮mol min−1 at
∼70% conversion). This large increase in rALIPHATIC on FeMoP-850
suggested that the decrease in SBEN observed in Fig. 6 was a result
of subsequent hydrogenation reactions. Therefore, it is possible that
either the hydrogenation sites on FeMoP-850 were more active or
present in a higher density (per mass basis) than the hydrogenation
sites on FeMoP-650. To verify that FeMoP-850 was more active for
benzene hydrogenation than FeMoP-650, reactions were run with
only benzene to determine the rate of consumption of benzene (-
r0,BEN,CO) (Fig. S12). FeMoP-850 was ∼ 2.5 times more active than
FeMoP-650 when normalized on CO-titrated sites. The results of
these residence time studies were in line with two suggested reac-
tion pathways, shown in Fig. 7, where the acid sites facilitate HDO
on phenol followed by benzene hydrogenation on metallic sites
(Fig. 8).
Fig. 6. Selectivity to benzene (SBEN) at various conversions of phenol (XPHEN) on the
FeMoP catalysts at 400 ◦C.
Fig. 7. (a) HDO pathway for CUSs (green circle) and (b) hydrogenation pathway for
metallic sites (red square) on the FeMoP surface. (For interpretation of the references
to color in this figure legend, the reader is referred to the web version of this article.)
ily with phenol, and thus HDO reactions dominate. At higher values
of XPHEN, the increased formation of benzene leads to higher sur-
face coverages of benzene, which would allow for its subsequent
hydrogenation on the FeMoP catalysts. One possible reason for the
different values of SBEN among the FeMoP catalysts at higher XPHEN
are due to the loss of CUSs (i.e. Lewis acid sites) on the FeMoP
catalysts synthesized at higher reduction temperatures. Further-
more, the surface characterization in Table 1 shows that FeMoP-650
and FeMoP-750 contained a similar amount of acid and CO-titrated
sites, while FeMoP-800 and FeMoP-850 had a higher amount of
CO-titrated sites compared to acid sites. Therefore at least two
types of metal sites likely existed on the FeMoP catalysts, metal-
lic sites and CUSs (i.e. Lewis acid sites) [30,40]. The CUSs were
of particular interest because prior reports with sulfide and phos-
phide catalysts suggested the CUSs to be the active sites for HDO
of model compounds [30,55,56,65]. Additionally, it has been pre-
viously proposed in multiple independent studies that metallic
sites are capable of hydrogenating aromatic rings to form aliphatic
products (i.e. cyclohexane and cyclohexene, denoted as aliphatic)
[30,66,67]. Thus it is proposed for these FeMoP materials (and sum-
marized schematically in Fig. 7) that the surface likely contained
CUSs that catalyzed the direct HDO of CAROMATIC-O, likely leading
to the large SBEN for FeMoP-650 [30,55,56], and metallic sites that
further hydrogenated the adsorbed CAROMATIC, likely resulting in
the higher rate of formation of aliphatic products for FeMoP-850
[24,30,50]. Higher reduction temperatures may facilitate the loss
of CUSs (i.e. Lewis acid sites) by removing them via heterolytic
cleavage of H2 [44,45], as shown by the decrease in values of NACID
with increasing reduction temperatures (Table 1). Furthermore,
3.2.4. Stability in the HDO of phenol
The catalytic stability of FeMoP-650 (catalyst with -r0,PHEN/MASS
)
was examined via a time-on-stream (TOS) study over 48 h under
5.2 MPa of H2 at 400 ◦C. For 48 h, FeMoP-650 was able to maintain
activity (-r0,PHEN/CO ∼9.5 ␮molPHEN ␮molCO−1 min−1) and selec-
tivity (SBEN ∼90%), which demonstrated the catalytic stability of
FeMoP-650.

92
D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
materials acted as non-aqueous phase acid catalysts for the dehy-
dration of cyclohexanol at 180 and 220 ◦C, and FeMoP-650 showed
the highest mass normalized rate due to the largest amount of
surface acid species. All FeMoP catalysts yielded similar activation
energy values for the dehydration of cyclohexanol (∼130 kJ mol−1),
which suggested that all FeMoP catalysts followed a similar reac-
tion pathway. In addition, the CO-titrated surface metal sites were
studied via hydrodeoxygenation (HDO) of phenol at 350–400 ◦C.
At conversions <15%, all FeMoP catalysts were ∼90% selective to
benzene and displayed similar activation energies (∼100 kJ mol−1),
suggesting a similar HDO reaction pathway on these FeMoP cat-
alysts. However, at conversions >20% the selectivity to benzene
decreased on FeMoP-850. When benzene was added into the feed
stream, FeMoP-650 favored HDO of phenol with minimal hydro-
genation of benzene while FeMoP-850 was significantly more
active for the hydrogenation of benzene to aliphatic compounds.
The high selectivity to benzene on FeMoP-650 was potentially due
to the increased number of coordinatively unsaturated sites, which
could be altered based on reduction temperature. Furthermore, the
FeMoP catalysts provided long-term stability due in part to in situ
reduction of surface species in the presence of H2 gas, as evidenced
by a 48 h time-on-stream study and XP spectra after the reaction.
Therefore, based on the high selectivities to the aromatic prod-
uct and high catalytic stability, FeMoP is highly applicable to other
lignocellulosic biomass relevant deoxygenation reactions.
Fig. 8. TOS study of the FeMoP-650 catalyst for the HDO of phenol.
The catalytic stability of FeMoP-650 after the TOS run was
further investigated using XPS in order to study the change in oxi-
dation state of the surface species. The resulting XP spectrum was
compared with that prior to the TOS run in order to determine if the
surface of FeMoP-650 was altered due to the reaction conditions
(Fig. S13). The C 1 s region in the XP spectrum after the TOS run
was first observed to investigate the potential formation of surface
carbidic species that have been reported as active species for this
reaction [71–73]. The XP spectrum did not show a peak associated
with surface carbidic species (located at binding energies between
280 and 283 eV). Therefore, the possible role of surface carbidic
species to catalyze this reaction was excluded. In addition, the other
Acknowledgements
We thank NSF for support through the CAREER program (CBET-
1351609). We also thank the ND Energy Materials Characterization
Facility for the use of the XRD and XPS, the Center for Environmental
Science and Technology (CEST) at the University of Notre Dame, and
Notre Dame Integrated Imaging Facilities (NDIIF) for partial sup-
port in utilizing their instrumentation. We thank Dr. Rouvimov and
DENSsolutions for their time and expertise in acquiring the images
from the in situ TEM work. We also thank Dr. Paul McGinn for valu-
able conversations and use of his crystallographic software. We
acknowledge VESTA 3 and Mercury CSD for access to their three-
dimensional visualization of crystal, volumetric, and morphology
data.
surface elements observed in the regions of Fe 2p 3/2, Mo 3d
,
5/2
and P 2p 3/2, were composed of two surface species (i.e. near zero-
valent species and oxidized species [53]). For the reduced surface
species, their oxidation states were retained throughout the TOS
run, as evidenced by unchanged binding energies in the XP spectra
(Fe␦+ at ∼706.7 eV; Mo␦+ at ∼227.6 eV; P␦− at ∼129.2 eV) [53,54,74].
In contrast, oxidized surface species were reduced during the TOS
run, as demonstrated by the decrease in binding energies of these
oxidized species. During the TOS run, surface P species assigned
to phosphate (at 133.5 eV) were also reduced to generate surface
species assigned to pyrophosphate species (132.7 eV) [27]. In par-
ticular, surface species of Fe3+ (at 711.1 eV) and Mo6+ (at 232.1 eV)
inherent to the FeMoP-650 were reduced to Fe2+ (710.2 eV) and
Mo5+ (230.7 eV). These results suggested that the oxidized surface
species underwent an in situ reduction during the reaction due to
the presence of elevated pressure H2. Therefore FeMoP-650 show-
cased enhanced catalytic stability due to in situ reduction of the
surface, which prevented complete surface oxidation and deacti-
vation.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.apcata.2016.06.
011.
References
[1] D.M. Alonso, S.G. Wettstein, J.A. Dumesic, Chem. Soc. Rev. 41 (2012)
8075–8098.
[2] E. Furimsky, Catal. Today (2013).
[3] J.C. Hicks, J. Phys. Chem. Lett. 2 (2011) 2280–2287.
[4] H. Li, S. Pattathil, M.B. Foston, S.-Y. Ding, R. Kumar, X. Gao, A. Mittal, J.M.
Yarbrough, M.E. Himmel, A.J. Ragauskas, M.G. Hahn, C.E. Wyman, Biotechnol.
Biofuels 7 (2014) 1–11.
[5] K. Barta, P.C. Ford, Acc. Chem. Res. 47 (2014) 1503–1512.
[6] P.C.A. Bruijnincx, B.M. Weckhuysen, Nat. Chem. 6 (2014) 1035–1036.
[7] Y. Romero, F. Richard, S. Brunet, Appl. Catal. B Environ. 98 (2010) 213–223.
[8] V.N. Bui, D. Laurenti, P. Afanasiev, C. Geantet, Appl. Catal. B Environ. 101
(2011) 239–245.
4. Conclusions
[9] C. Zhao, Y. Kou, A.A. Lemonidou, X. Li, J.A. Lercher, Angew. Chem. Int. Ed. 48
(2009) 3987–3990.
[10] N. Yan, Y. Yuan, R. Dykeman, Y. Kou, P.J. Dyson, Angew. Chem. Int. Ed. 49
(2010) 5549–5553.
[11] W. Zhang, J. Chen, R. Liu, S. Wang, L. Chen, K. Li, ACS Sustainable Chem. Eng. 2
(2014) 683–691.
[12] T. Omotoso, S. Boonyasuwat, S.P. Crossley, Green Chem. 16 (2014) 645–652.
[13] W.-S. Lee, Z. Wang, R.J. Wu, A. Bhan, J. Catal. 319 (2014) 44–53.
[14] J. Kim, G.T. Neumann, N.D. McNamara, J.C. Hicks, J. Mater. Chem. A 2 (2014)
14014–14027.
In summary, we have studied the multifunctional surface char-
acteristics of FeMoP catalysts as they pertain to the deoxygenation
of biomass model compounds. The surface properties were altered
by changing the reduction temperature used during the catalyst
syntheses (650–850 ◦C). It was determined that the materials with
the highest acidity (␳ACID) and CO-titrated metal sites (NCO) were
attainable at the lowest reduction temperatures. These FeMoP

D.J. Rensel et al. / Applied Catalysis A: General 524 (2016) 85–93
93
[15] H. Wang, J. Male, Y. Wang, ACS Catal. 3 (2013) 1047–1070.
[48] B. Katryniok, S. Paul, F. Dumeignil, ACS Catal. 3 (2013) 1819–1834.
[49] S. Sato, F. Sato, H. Gotoh, Y. Yamada, ACS Catal. 3 (2013) 721–734.
[50] C. Zhao, J. He, A.A. Lemonidou, X. Li, J.A. Lercher, J. Catal. 280 (2011) 8–16.
[51] A. Vjunov, M.Y. Hu, J. Feng, D.M. Camaioni, D. Mei, J.Z. Hu, C. Zhao, J.A. Lercher,
Angew. Chem. Int. Ed. 53 (2014) 479–482.
[52] H. Knözinger, H. Bühl, K. Kochloefl, J. Catal. 24 (1972) 57–68.
[53] A.P. Grosvenor, S.D. Wik, R.G. Cavell, A. Mar, Inorg. Chem. 44 (2005)
8988–8998.
[54] A.P. Grosvenor, R.G. Cavell, A. Mar, J. Solid State Chem. 180 (2007) 2702–2712.
[55] M. Badawi, J.-F. Paul, E. Payen, Y. Romero, F. Richard, S. Brunet, A. Popov, E.
Kondratieva, J.-P. Gilson, L. Mariey, A. Travert, F. Maugé, Oil Gas Sci. Technol.
68 (2013) 829–840.
[56] A.Y. Bunch, X. Wang, U.S. Ozkan, Appl. Catal. A 346 (2008) 96–103.
[57] A. Cho, J. Shin, A. Takagaki, R. Kikuchi, S. Oyama, Top Catal. 55 (2012) 969–980.
[58] X. Liu, A. Tkalych, B. Zhou, A.M. Köster, D.R. Salahub, J. Phys. Chem. C 117
(2013) 7069–7080.
[59] A.S. Rocha, A.B. Rocha, V.T. da Silva, Appl. Catal. A 379 (2010) 54–60.
[60] B. Zhou, X. Liu, J. Cuervo, D.R. Salahub, Struct. Chem. 23 (2012) 1459–1466.
[61] X. Zhu, L. Nie, L.L. Lobban, R.G. Mallinson, D.E. Resasco, Energy Fuels 28 (2014)
4104–4111.
[62] Y. Hong, H. Zhang, J. Sun, K.M. Ayman, A.J.R. Hensley, M. Gu, M.H. Engelhard,
J.-S. McEwen, Y. Wang, ACS Catal. 4 (2014) 3335–3345.
[63] R.C. Nelson, B. Baek, P. Ruiz, B. Goundie, A. Brooks, M.C. Wheeler, B.G.
Frederick, L.C. Grabow, R.N. Austin, ACS Catal. 5 (2015) 6509–6523.
[64] D.R. Moberg, T.J. Thibodeau, F.G. Amar, B.G. Frederick, J. Phys. Chem. C 114
(2010) 13782–13795.
[16] C. Zhao, W. Song, J.A. Lercher, ACS Catal. 2 (2012) 2714–2723.
[17] J. He, C. Zhao, J.A. Lercher, J. Catal. 309 (2014) 362–375.
[18] C. Zhao, S. Kasakov, J. He, J.A. Lercher, J. Catal. 296 (2012) 12–23.
[19] 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. El Fallah, J.P.
Gilson, L. Mariey, A. Travert, F. Maugé, J. Catal. 282 (2011) 155–164.
[20] B.M. Vogelaar, N. Kagami, T.F. van der Zijden, A.D. van Langeveld, S. Eijsbouts,
J.A. Moulijn, J. Mol. Cat. A Chem. 309 (2009) 79–88.
[21] T. Prasomsri, T. Nimmanwudipong, Y. Roman-Leshkov, Energy Environ. Sci. 6
(2013) 1732–1738.
[22] P.M. Mortensen, H.W.P. de Carvalho, J.-D. Grunwaldt, P.A. Jensen, A.D. Jensen,
J. Catal. 328 (2015) 208–215.
[23] R. Rousseau, D.A. Dixon, B.D. Kay, Z. Dohnalek, Chem. Soc. Rev. 43 (2014)
7664–7680.
[24] X. Zhu, L.L. Lobban, R.G. Mallinson, D.E. Resasco, J. Catal. 281 (2011) 21–29.
[25] S.T. Oyama, T. Gott, H. Zhao, Y.-K. Lee, Catal. Today 143 (2009) 94–107.
[26] R. Prins, M. Bussell, Catal. Lett. (2013) 1–24.
[27] D.J. Rensel, S. Rouvimov, M.E. Gin, J.C. Hicks, J. Catal. 305 (2013) 256–263.
[28] H.Y. Zhao, D. Li, P. Bui, S.T. Oyama, Appl. Catal. A 391 (2011) 305–310.
[29] K. Li, R. Wang, J. Chen, Energy Fuels 25 (2011) 854–863.
[30] V.M.L. Whiffen, K.J. Smith, Energy Fuels 24 (2010) 4728–4737.
[31] Y. Yang, A. Gilbert, C. Xu, Appl. Catal. A 360 (2009) 242–249.
[32] S. Ted Oyama, H. Zhao, H.J. Freund, K. Asakura, R. Włodarczyk, M. Sierka, J.
Catal. 285 (2012) 1–5.
[33] R. Wang, K. Smith, Catal. Lett. 144 (2014) 1594–1601.
[34] P. Bui, J.A. Cecilia, S.T. Oyama, A. Takagaki, A. Infantes-Molina, H. Zhao, D. Li, E.
Rodríguez-Castellón, A. Jiménez López, J. Catal. 294 (2012) 184–198.
[35] A.P. Grosvenor, R.G. Cavell, A. Mar, J. Solid State Chem. 181 (2008) 2549–2558.
[36] S. Diplas, O.M. Løvvik, J. Phys. Condens. Matter. 21 (2009) 245503.
[37] B.J. Tan, K.J. Klabunde, P.M.A. Sherwood, Chem. Mater. 2 (1990) 186–191.
[38] M. Shetty, K. Murugappan, T. Prasomsri, W.H. Green, Y. Román-Leshkov, J.
Catal. 331 (2015) 86–97.
[65] A.Y. Bunch, U.S. Ozkan, J. Catal. 206 (2002) 177–187.
[66] J.A. Widegren, M.A. Bennett, R.G. Finke, J. Am. Chem. Soc. 125 (2003)
10301–10310.
[67] A.C.A. Monteiro-Gezork, R. Natividad, J.M. Winterbottom, Catal. Today 130
(2008) 471–485.
[68] M. Daage, R.R. Chianelli, J. Catal. 149 (1994) 414–427.
[69] S. Kasztelan, H. Toulhoat, J. Grimblot, J.P. Bonnelle, Appl. Catal. 13 (1984)
127–159.
[70] E.J.M. Hensen, P.J. Kooyman, Y. van der Meer, A.M. van der Kraan, V.H.J. de
Beer, J.A.R. van Veen, R.A. van Santen, J. Catal. 199 (2001) 224–235.
[71] I.N. Shabanova, V.A. Trapeznikov, J. Electron. Spectros. Relat. Phenom. 6
(1975) 297–307.
[39] X. Wang, P. Clark, S.T. Oyama, J. Catal. 208 (2002) 321–331.
[40] Y.-K. Lee, S.T. Oyama, J. Catal. 239 (2006) 376–389.
[41] F. Al-Kharafi, H. Al-Kandari, A. Katrib, Catal. Lett. 123 (2008) 269–275.
[42] B. Hou, H. Zhang, H. Li, Q. Zhu, Chin. J. Chem. Eng. 20 (2012) 10–17.
[43] E. Muneyama, A. Kunishige, K. Ohdan, M. Ai, J. Catal. 158 (1996) 378–384.
[44] P. Hirunsit, K.-i. Shimizu, R. Fukuda, S. Namuangruk, Y. Morikawa, M. Ehara, J.
Phys. Chem. C 118 (2014) 7996–8006.
[72] L. Ramqvist, K. Hamrin, G. Johansson, A. Fahlman, C. Nordling, J. Phys. Chem.
Solids 30 (1969) 1835–1847.
[45] S. Klacar, H. Gronbeck, Catal. Sci. Technol. 3 (2013) 183–190.
[46] A. Seidel, A. Gutsze, B. Boddenberg, Chem. Phys. Lett. 275 (1997) 113–118.
[47] V.A. Nasluzov, V.V. Rivanenkov, A.M. Shor, K.M. Neyman, U. Birkenheuer, N.
Rösch, Int. J. Quantum Chem. 90 (2002) 386–402.
[73] C.S. Kuivila, J.B. Butt, P.C. Stair, Appl. Surf. Sci. 32 (1988) 99–121.
[74] P.E.R. Blanchard, A.P. Grosvenor, R.G. Cavell, A. Mar, J. Mater. Chem. 19 (2009)
6015–6022.