Kinetic analysis and reaction mechanism for anisole conversion over zirconia-supported molybdenum oxide

Article abstract of DOI:10.1016/j.jcat.2019.06.046

Gas-phase catalytic conversion of anisole and its reaction intermediates was studied over a 10 wt% MoO₃/ZrO₂ catalyst at temperatures between 553 and 633 K and H₂ partial pressures (P_H2) ≤ 1 bar. Benzene, phenol, cresol and methyl anisole were identified as the primary products from the hydrodeoxygenation (HDO), hydrogenolysis, intra- and intermolecular alkylation of anisole, respectively. The anisole to benzene conversion featured a first-order dependence with respect to P_H2, while the conversion of phenol to benzene and m-cresol to toluene, showed P_H2 and P_oxygenate reaction orders of 1/2 and zero, respectively. A kinetic model showed that although the secondary pathway of phenol HDO to benzene has a rate constant ～3 times higher than that for the HDO of anisole to benzene, the anisole HDO pathway is dominant at low anisole conversions. Apparent orders of ～? with P_oxygenate for anisole hydrogenolysis and alkylation to form phenol, cresol, and methyl anisole implied the existence of different active sites than those responsible for HDO. Co-feed studies with H₂O, pyridine, and di-tert butyl pyridine (DTBP) indicated that the active-sites responsible for HDO have a Lewis acid character that is associated with oxygen vacancies and that is distinct from the nature of sites responsible for hydrogenolysis and alkylation. Accordingly, co-feeding CH₃OH resulted in increased phenol alkylation rates to form alkylated cresols along with inhibition of phenol to benzene HDO rates. A three-site model was proposed to unify the HDO, hydrogenolysis, and alkylation reactivity data obtained from the kinetic and co-feed studies.

Full text of DOI:10.1016/j.jcat.2019.06.046

Journal of Catalysis 376 (2019) 248–257
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Kinetic analysis and reaction mechanism for anisole conversion over
zirconia-supported molybdenum oxide
1
⇑
⇑
Manish Shetty , Eric M. Anderson, William H. Green , Yuriy Román-Leshkov
Department of Chemical Engineering, Massachusetts Institute of Technology, Cambridge, MA 02139, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Gas-phase catalytic conversion of anisole and its reaction intermediates was studied over a 10 wt% MoO3/
ZrO2 catalyst at temperatures between 553 and 633 K and H2 partial pressures (PH2) ꢀ 1 bar. Benzene,
phenol, cresol and methyl anisole were identified as the primary products from the hydrodeoxygenation
(HDO), hydrogenolysis, intra- and intermolecular alkylation of anisole, respectively. The anisole to ben-
zene conversion featured a first-order dependence with respect to PH2, while the conversion of phenol
to benzene and m-cresol to toluene, showed PH2 and Poxygenate reaction orders of 1/2 and zero, respec-
tively. A kinetic model showed that although the secondary pathway of phenol HDO to benzene has a rate
constant ꢁ3 times higher than that for the HDO of anisole to benzene, the anisole HDO pathway is dom-
inant at low anisole conversions. Apparent orders of ꢁ½ with Poxygenate for anisole hydrogenolysis and
alkylation to form phenol, cresol, and methyl anisole implied the existence of different active sites than
those responsible for HDO. Co-feed studies with H2O, pyridine, and di-tert butyl pyridine (DTBP) indi-
cated that the active-sites responsible for HDO have a Lewis acid character that is associated with oxygen
vacancies and that is distinct from the nature of sites responsible for hydrogenolysis and alkylation.
Accordingly, co-feeding CH3OH resulted in increased phenol alkylation rates to form alkylated cresols
along with inhibition of phenol to benzene HDO rates. A three-site model was proposed to unify the
HDO, hydrogenolysis, and alkylation reactivity data obtained from the kinetic and co-feed studies.
Ó 2019 Published by Elsevier Inc.
Received 18 December 2018
Revised 1 June 2019
Accepted 30 June 2019
Keywords:
Hydrodeoxygenation (HDO)
Alkylation
Kinetics
Anisole
Co-feed
Bifunctional catalysis
Molybdenum oxide
1. Introduction
lignin-derived phenolics into aromatic hydrocarbons at mild
conditions (593 K and H2 partial pressures ꢀ1 bar) without
The development of catalysts for the hydrodeoxygenation
(HDO) of lignin-derived phenolic compounds obtained from the
catalytic fast pyrolysis (CFP) and reductive catalytic fractionation
(RCF) of biomass is an important step in the valorization of biomass
into value-added chemicals and fuels [1–17]. These compounds,
including eugenol, syringol, and guaiacol, contain both
ring-saturation [26–28]. These MoO3-based catalysts were shown
to be a promising alternative to zeolites for the upgrading of bio-
oils obtained from the CFP of lignocellulosic biomass [29–32].
Shetty et al. demonstrated that anisole conversion on MoO3/ZrO2
catalysts formed HDO (e.g., benzene and toluene), hydrogenolysis
(e.g., phenol) and alkylation (e.g., methylated cresols and anisoles)
products [33]. While the product distributions and the impact of
MoO3 loading on the reactivity towards HDO and alkylation were
reported, detailed information on the reaction kinetics and HDO
mechanism were not elucidated for supported MoO3 catalysts.
Herein, we report a detailed investigation of the reaction
network, kinetics, and nature of active sites for the conversion of
anisole over a 10 wt% MoO3/ZrO2 catalyst. The reaction network
was determined by varying the catalyst contact times to determine
primary, secondary and higher-order products. Next, the kinetics
for anisole conversion to benzene, phenol, cresol and methyl ani-
sole, as well as the HDO of the intermediates phenol and m-
cresol to benzene and toluene, respectively, were studied on sup-
ported MoO3 catalysts over a temperature range of 553–633 K.
Caromatic-OH and Caromatic-OCH3 bonds that are cleaved during the
HDO process [11,18]. The fate of the reaction products hinges on
effective catalyst and reactor design, which in turn requires funda-
mental insights into the chemical kinetics and the nature of active
sites responsible for bond activation [19–25]. Recently,
Román-Leshkov and coworkers have demonstrated that bulk and
supported MoO3 catalysts are highly effective at the HDO of
⇑
Corresponding authors.
E-mail addresses: whgreen@mit.edu (W.H. Green), yroman@mit.edu
(Y. Román-Leshkov).
1
Present address: Institute of Integrated Catalysis (IIC), Pacific Northwest National
Laboratory (PNNL), 902 Battelle Boulevard, Richland, WA 99354, USA.
https://doi.org/10.1016/j.jcat.2019.06.046
0021-9517/Ó 2019 Published by Elsevier Inc.

M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
249
The nature of the active sites was probed with the aid of control
reactions, co-feed studies with pyridine, water (H2O), methanol
(CH3OH), and di-tert butyl pyridine (DTBP). Lastly, a macrokinetic
model involving oxygenate and H2 activation over three distinct
sites was proposed to described the observed reactivity data.
The kinetic experiments were performed at differential condi-
tions (<15% conversion) on a C-mol% basis (unless otherwise spec-
ified). The kinetic studies for determining reaction orders for the
HDO of anisole, phenol, and m-cresol were performed on 10 wt%
MoO3/ZrO2 and bulk MoO3 catalysts, while the kinetic studies for
the hydrogenolysis and alkylation reactions were performed on
the bare ZrO2 support and the 10 wt% MoO3/ZrO2 catalyst. The
reported reaction rates for 10 wt% MoO3/ZrO2 and ZrO2 were
obtained from steady-state data, while for bulk MoO3 it was
obtained from the maximum rate after the induction period. As
discussed in previous reports, bulk MoO3 undergoes an induction
period before reaching peak reactivity followed by gradual deacti-
vation [27,28]. The apparent kinetic orders with respect to anisole
and H2 were obtained by varying their partial pressures from
0.0049 to 0.0147 bar, and 0.2508–1.0032 bar, respectively. At
2. Materials and methods
2.1. Materials and reactivity measurements
The 10 wt% MoO3/ZrO2 catalyst and the ZrO2 support were syn-
thesized with procedures described in previous reports [28]. Bulk
MoO3 (99.97% trace metals basis, Sigma Aldrich) was used without
any purification or pretreatment. Reactivity studies were carried
out in the vapor-phase in a packed-bed, down-flow reactor. The
reactor consisted of a stainless-steel tube (0.95 cm OD, with wall
thickness of 0.089 cm) mounted in a single-zone furnace (Applied
Test Systems, Series 3210, 850 W/115 V). The temperature was
controlled by a temperature controller (Digi-Sense, model 68900-
10) connected to a K-type thermocouple (Omega, model TJ36-
CAXL-116u) mounted downstream in direct contact with the cata-
lyst bed. The catalyst was pelletized between 100 and 140 mesh,
PH2 < 1.0032 bar, the PTotal was 1.013 bar with N2 as the balance
gas. Solutions of 50 wt% in mesitylene and 40% in cyclohexane
were used for phenol and m-cresol, respectively (unless mentioned
otherwise). Mesitylene and cyclohexane did not react on contact
with the catalyst. The activation energy barriers were calculated
by varying the reaction temperatures between 573 and 613 K
(unless mentioned otherwise) for 10 wt% MoO3/ZrO2 and 553–
613 K for bulk MoO3.
mixed with inert
a-Al2O3 diluent (100–200 mesh, total 1 g), and
packed between two layers of
a
-Al2O3 (1.5 g each) in the middle
2.2. Co-feed experiments
of the furnace. Prior to the reaction, the reactor temperature was
ramped at a rate of ꢁ6 K minꢂ1 under N2 until reaching the reac-
tion temperature (553–613 K). Next, an oxygenated feed (anisole,
phenol solution in mesitylene or m-cresol solution in cyclohexane)
was delivered into the reactor via a capillary tube connected to a
syringe pump (Harvard Apparatus, model 703005) and mixed with
H2 gas at the inlet of the reactor. Catalyst contact times were chan-
ged by adjusting the flow rates of the oxygenate feed to vary the
weight hourly space velocity (WHSV), defined with respect to the
equivalent mass of MoO3 loaded [33]. Typically, flow rates of the
oxygenate feed ranged between 100 and 400 ml hꢂ1. The mass of
equivalent MoO3 was computed as the product of MoO3 loading
(10 wt%) and the total mass of catalyst loaded.
The co-feed experiments were carried out with water (H2O),
methanol (CH3OH), pyridine and ditertbutyl pyridine (DTBP).
Pyridine and DTBP were introduced in a solution with anisole from
a different syringe pump (Harvard apparatus, model 703310) while
maintaining the partial pressure of anisole (PAnisole) at a constant
value. The partial pressure of both pyridine and DTBP
was ꢁ0.0015 barꢄH2O or CH3OH were introduced as a saturated
feed by bubbling H2 through liquid H2O or CH3OH while maintain-
ing PAnisole at a constant value, with PH2O and PCH3OH values of 0.032
and 0.16 bar, respectively, as determined by vapor pressure data at
room temperature.
The reactor effluent lines were heated to 523 K to prevent con-
densation. The effluents were analyzed and quantified via an
online gas chromatograph (GC) fitted with a DB-5 column (Agilent,
30 m ꢃ 0.25 mm ID ꢃ 0.25 mm) and equipped with a mass selective
detector for identification (MSD, Agilent Technologies, model
5975C) and a flame ionization detector (FID, Agilent Technologies,
model 7890 A) for quantification. The GC parameters used for anal-
ysis were as follows: detector temperature 573 K, injector temper-
ature 548 K, split ratio 1:20. The initial and final oven temperatures
2.3. Fourier transform infrared (FTIR) spectroscopy
FTIR spectra of the 10 wt% MoO3/ZrO2 catalyst were acquired
from 4000 to 400 cmꢂ1 using a Bruker Vertex 70 spectrophotome-
ter by averaging 64 scans at a 2 cmꢂ1 resolution. Samples were
pressed into 7 mm diameter self-supporting pellets and placed in
a Harrick high temperature transmission cell equipped with KBr
windows. Samples were calcined in situ under flowing dry air
(50 ml minꢂ1), with a temperature ramp of 5 K minꢂ1 to 773 K,
and held at 773 K for 1 h. After the cell was cooled to room temper-
ature, dynamic vacuum of ꢁ0.1 Pa was established and a reference
spectrum of the bare material was acquired. Under a static vac-
uum, the cell was progressively dosed with pyridine and/or DTBP
vapor until physisorption peaks were observed. After overnight
evacuation at room temperature, spectra were collected as the
were 323 and 523 K, with a ramp of 10 K minꢂ1
.
The following equations were used to quantify experimental
data. The carbon balances were typically above 95%, (calculated
by comparing the total carbon moles of reaction products) as
shown in Fig. S1a. Our previous reports have shown that carbon
lost as coke is insignificant under the reaction conditions used in
this study [28]. Hence, the selectivity and yield were defined based
on the GC observable products.
temperature was increased in 50 K increments (rate 5 K minꢂ1
up to 623 K, and then cooled to room temperature. Similar acqui-
sition protocols were used after H2 treatment at 593 K.
)
Carbon moles of reactant consumed
Con
v
ersionðC ꢂ mol%Þ ¼
ꢃ 100
ð1Þ
Carbon moles of reactant fed
2.4. Ammonia-temperature programed desorption (NH3-TPD)
Carbon moles in product
NH3-TPD was performed in a quartz U-tube reactor setup
mounted in an insulated single-zone furnace (550 W/115 V, Carbo-
lite GTF 11/50/150B) connected to an online mass spectrometer
(MS, HIDEN Analytical HPR-20/QIC). 100 mg of catalyst (60–100
mesh) was loaded into a U-tube packed between quartz wool.
The catalyst bed was in contact with a K-type thermocouple
(Omega, model TJ36-CAXL-116u) that was connected to a temper-
Selecti
v
ityðC ꢂ mol%Þ ¼
ꢃ 100
Total carbon moles in products
ð2Þ
ð3Þ
Carbon moles in product
YieldðC ꢂ mol%Þ ¼
ꢃ 100
Carbon moles of reactant fed

250
M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
ature controller. The catalyst was calcined in O2 at 100 ml minꢂ1 at
773 K for 2 h at a ramp rate of 5 K minꢂ1 and cooled under 1% Ar in
He to 373 K (100 ml minꢂ1). The catalyst was exposed to several
pulses of NH3 totaling ꢁ10 ml of NH3 at STP. The catalyst was
purged with 1% Ar in He for 1 h at 373 K. The sample was then
heated to 823 K (10 K minꢂ1) and held for 30 min at 823 K. The
evolved NH3 was monitored with MS by tracking the m/z = 17 frag-
ment. The MS response was calibrated for each ammonia pulse
prior to the analysis. The acidity was estimated assuming a ratio
of acid-site/NH3 = 1.
2.6. Thermodynamic calculations
The free energies for the reaction network at 593 K were calcu-
lated using the Gaussian 03 software [34]. The geometry and the
energies were computed at the CBS-QB3 level of theory, using
the rigid rotor harmonic oscillator approximation.
3. Results
3.1. Physicochemical properties of 10 wt% MoO3/ZrO2
The physicochemical properties of the catalyst including BET
surface area, oxygen uptake, percentage of redox-active species
and acid site concentration are summarized in Table 1. The percent
of redox-active surface species was 31% as determined from oxy-
gen chemisorption data. The acid site concentration was ꢁ685 l-
mol/g from NH3-TPD data. At 10 wt% loading, the MoOx domains
on the ZrO2 support are amorphous and oligomeric in nature
[28]. The detailed physicochemical characterization of the catalysts
of the MoO3/ZrO2 catalyst including the powder X-ray diffraction
(PXRD) patterns and Raman spectra were reported in our previous
studies [28].
2.5. Reactor model
We assumed isothermal operation of a plug flow reactor with
negligible pressure drop across the catalyst bed. A packed bed reac-
tor (PBR) model was employed (see Eq. (4)):
dFi
dW
¼ Ri
ð4Þ
ð5Þ
with initial condition of Fi ¼ Fi;0 at W ¼ 0
In the above equations, W represents the catalyst mass, Fi and Ri
are the molar inlet flow rate and the net production rate of the
component i, respectively. The net rate of R of each component is
defined as the total sum of the formation and consumption of each
component.
3.2. Primary analysis of the products during anisole conversion
The primary and secondary products of anisole conversion were
determined by plotting product selectivities as a function of con-
X
version (0.6–44.0 C-mol %) at
a temperature of 593 K, a
Ri ¼
mijrj
ð6Þ
PAnisole = 0.0098 bar, and a PH2 = 1.0032 bar (Fig. 1). Extrapolating
j
product selectivity values of the major products to 0% conversion
showed that benzene, phenol, cresol and methyl anisole are the
primary aromatic products formed from anisole conversion (see
Scheme 1). We note that the higher alkylated aromatics and oxy-
genates were formed at high conversions (Scheme 2 and Fig. S2).
This product distribution suggests selective hydrodeoxygenation
(HDO) of the O-Caromatic bond to form benzene, hydrogenolysis of
the O-Caliphatic bond to form phenol, methyl migration from anisole
to form cresol, and the alkylation of anisole to form methyl anisole.
The latter product can be formed via intermolecular methyl migra-
tion, as suggested by Resasco and coworkers [35,36], intramolecu-
lar methyl migration, or secondary alkylation from surface
methoxy species. The negative free energies for the individual
reactions presented in Scheme 1 show that none of the primary
reactions are equilibrium limited at the reaction conditions inves-
tigated. Although benzene and CH3OH should be primary HDO
products, CH4 rather CH3OH than was detected, suggesting that
the HDO of methanol to methane is very fast at these conditions.
We note the bare ZrO2 support was active for the hydrogenolysis
and alkylation reactions to form phenol, cresols and methyl ani-
sole, albeit at rates ꢁ2–3 lower than MoO3/ZrO2 (see Table S5)
but was inactive for HDO to form benzene [33].
where mij is the stoichiometric coefficient of component i in
reaction j.
Since H2 is in stoichiometric excess with respect to anisole, the
change in the flow rate of H2 was not considered in the analysis.
The rate constants for the different reactions was determined by
fitting the output flow rate of all components from Fig. 1 to the
output of the system of ODEs in Eq. (4) in MATLABÒ using lsqcurve-
fit function (detailed analysis given in Supplementary Note 2).
3.3. Kinetic analysis of primary product formation
Steady-state kinetics for the formation of primary products
were obtained under differential conditions in the absence of
external and internal mass-transfer limitations as determined by
meeting the Mears and Weisz-Prater criteria, respectively (see
Supplementary Note 1). ZrO2 were used in addition to the 10 wt%
MoO3/ZrO2 catalyst in order to gain insights into the effects of
the support.
Fig. 1. . Selectivity of major products as a function of conversion of anisole on
10 wt% MoO3/ZrO2 catalyst. The selectivity is defined based on the carbon moles
included in aromatic carbons. Cresol is mainly o-cresol. Reaction conditions:
1.5–150 mg of 10 wt% MoO3/ZrO2. PTotal = 1.013 bar (0.0098 bar PAnisole, balance H2).
Temperature = 593 K. The solid trend lines serve as a guide to the eye. No other
products were detected at the lowest three conversion values.
3.3.1. Apparent reaction orders for benzene production
The benzene formation rate for the 10 wt% MoO3/ZrO2 catalyst
exhibited a near zero-order dependence (ꢂ0.03 0.08) with PAnisole

M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
251
Table 1
Textural properties, oxygen chemisorption values and acid site concentration from NH3-TPD for 10 wt% MoO3/ZrO2 catalyst.
Catalyst
BET surface area (m2/g)
112
Mo density (Mo/nm2)
3.7
Oxygen uptake (
l
mol/g)
% redox-active species
31
Acid-site concentration (
lmol/g)
10 wt% MoO3/ZrO2
107
685
Oxygen uptake and % redox-active species were reported from Shetty et al. [33]. The catalyst was reduced at 623 K for 2 h and oxygen uptake was measured at 303 K.
The rates of phenol, cresol and methyl anisole formation
increased ꢁ2–3 times on MoO3/ZrO2 as compared to ZrO2 support,
as shown in Table S5 when normalized to the mass of ZrO2. There-
fore, the sites on the ZrO2 support do not appear to play a domi-
nant role towards the formation of oxygenates. The ZrO2 support
featured phenol, cresol and methyl anisole reactions orders of
1.2 0.2, 1.2 0.2, and 1.0 0.2, respectively, with Panisole and near
half-order with PH2 (Fig. S6). These values suggest the contribution
of bimolecular pathways on the support for anisole conversion to
phenol and cresol, leading to reaction orders greater than 1. The
apparent activation energy for phenol formation from anisole on
ZrO2 was 85 12 kJ molꢂ1 in the temperature range 573–613 K
(Fig. S7), similar to MoO3/ZrO2. In contrast to bare ZrO2, the appar-
ent reaction orders with respect to anisole for MoO3/ZrO2 were all
<1 with PAnisole suggesting that the coverages of surface species
over the sites responsible for catalyzing these reactions in both
the catalyst and the bare support are different (Supplementary
Note 4).
3.4. Co-feed studies to determine the nature of active sites on 10 wt%
MoO3/ZrO2
The nature of the active sites for the HDO, hydrogenolysis and
alkylation reactions was interrogated with the aid of co-feed stud-
ies. Co-feeds that selectively inhibit product formation are useful
for developing insights into the site-requirements for catalysis that
Scheme 1. Free energy values for the primary reaction pathways and their
corresponding products from anisole conversion. All the free energy values are
reported in units of kJ molꢂ1 in the gas phase at 593 K.
are consistent with the kinetic studies. We note that the co-feeds
were largely reversible adsorbates and were not used for quantita-
tion. However, the reactions were not equilibrium limited as noted
earlier, hence the co-feeds are not expected to appreciably affect
the reaction equilibrium. Under our reaction conditions, H2 is pre-
sent in stoichiometric excess as compared to the oxygenate feed
(usually anisole). Introduction of a co-feed (H2O, pyridine and di-
tertbutyl pyridine) without change in the concentration of oxy-
genate leads to a negligible change in the PH2. Thus, any change
in reactivity can be attributed to the presence of the co-feed.
CH3OH was added in excess (PCH3OH = 0.16 bar, PCH3OH/Panisole ꢅ 16)
and alters the kinetics, effectively blocking many sites, and putting
more methyl groups on the surface than regular anisole
conversion.
(Fig. 2a). Benzene formation rates from phenol were also near zero-
order (0.02 0.08) with respect to PPhenol. These data are consistent
with previous reports of HDO rates of m-cresol and anisole on sup-
ported MoO3 and bulk Mo2C catalysts [28,37].
The rate of benzene production from anisole and phenol exhib-
ited apparent first-order (0.96 0.03) and half order (0.54 0.03)
dependencies, respectively, on PH2 (Fig. 2b). An apparent half-
order (0.56 0.05) with respect to PH2 was also observed for the
HDO of m-cresol to toluene (Fig. S3). The apparent activation
energy for anisole HDO was 116 2 kJ molꢂ1 in the temperature
range of 573–613 K, while phenol featured an apparent barrier of
98 3 kJ molꢂ1 in the same temperature range (Fig. 2c). The differ-
ence in the reaction order with PH2 and apparent activation ener-
gies for the two HDO reactions implicates differences in the
number of H additions prior to the rate determining step (RDS)
in the HDO mechanism for each substrate (vide infra).
3.4.1. Co-feed with H2O co-feed during anisole conversion
Fig. 4 shows the impact of H2O on the product distribution dur-
ing anisole conversion over 10 wt% MoO3/ZrO2. First, anisole
(0.0098 bar Panisole) mixed with H2 was introduced into the reactor.
After ꢁ1.2 h on stream, 0.032 bar PH2O (PH2O/PAnisole ꢅ 3) was co-
fed, leading to a decrease in HDO product (i.e., benzene and
toluene) yields from ꢁ8.8 and 3.6% to 1.6 and 0.7%, respectively,
which corresponds to an 80% reduction on a C-mol basis. Note that
at identical yields without the H2O co-feed, the PH2O/PAnisole ꢅ 0.15,
which indicates that inhibition by H2O was limited under normal
conditions. This inhibition effect was reversible, as shown by the
fast recovery of the original yields when the water co-feed was
stopped. A similar behavior was observed during the HDO of phe-
nol and m-cresol in the presence of a H2O co-feed (Figs. S9 and
S10). Although the HDO activity was decreased, the presence of
H2O during anisole HDO increased the yields of other oxygenate
products. For instance, phenol and cresol increased from 12.5
3.3.2. Apparent reaction orders for hydrogenolysis and alkylation on
10 wt% MoO3/ZrO2 catalyst and ZrO2
Fig. 3 illustrates the dependence of phenol, cresol and methyl
anisole formation rates on PAnisole and PH2 on the 10 wt% MoO3/
ZrO2 catalyst. Phenol, cresol and methyl anisole formation rates
showed orders with respect to PAnisole of 0.72 0.06, 0.55 0.09
and 0.43 0.03, respectively. With varying PH2, phenol, cresol,
and methyl anisole formation rates showed orders of 0.48 0.06,
0.71 0.03 and 0.43 0.03, respectively. The apparent activation
energy barriers for phenol, cresol and methyl anisole formation
from anisole were 93 16, 100 25 and 89 21 kJ molꢂ1, respec-
tively, in the temperature range 573–633 K (Fig. S5).

252
M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
Scheme 2. Reaction network for anisole conversion for hydrodeoxygenation (HDO), hydrogenolysis, intra- and intermolecular alkylation reactions.
(a)
6.0
(b)
(c)
7
Slope = 0.54 0.03
Slope = 0.96 0.03
Eact = 98 3 kJ/mol
Slope = 0.02 0.08
Slope = -0.03 0.08
6
5
4
3
2
5.5
5.0
4.5
4.0
3.5
3.0
6
5
4
3
Eact = 116 2 kJ/mol
0.00160
0.00165
0.00170
0.00175
-1.8 -1.5 -1.2 -0.9 -0.6 -0.3 0.0
ln (PH2
-5.4
-5.1
-4.8
-4.5
-4.2
1/T (1/K)
)
ln (Poxygenate
)
Fig. 2. Steady-state kinetic studies for anisole (open symbols) and phenol (closed symbols) HDO to benzene. The figure shows benzene production rates as a function (a)
anisole and phenol partial pressure, (b) hydrogen partial pressures (PH2), and (c) inverse temperature (1/T). Reaction conditions: (a) 25–30 mg MoO3/ZrO2. PTotal = 1.013 bar
(0.0049–0.0147 bar PAnisole for anisole HDO, 0.0053–0.0155 bar PPhenol and 0.0082–0.0244 Pmesitylene for phenol HDO, balance H2, temp = 593 K). (b) 30–40 mg MoO3/ZrO2.
0.0098 bar PAnisole and 0.2508–1.0032 bar PH2 for anisole HDO; 0.0106 bar PPhenol, 0.0164 bar Pmesitylene and 0.1972–0.986 bar PH2 for phenol HDO, balance N2, temp = 593 K.
(c) 30–40 mg MoO3/ZrO2. PTotal = 1.013 bar (0ꢂ.01098 bar PAnisole for anisole HDO, 0.0106 bar PPhenol and 0.0164 bar Pmesitylene for phenol HDO, balance H2), T = 573–613 K. The
reaction rate is represented in mmol C hꢂ1 gMoO3. The conversions were less than 15 C-mol% for all experiments.
and 9.9% to 24.5 and 15.5%, respectively. This effect was also
observed during anisole conversion on the bare ZrO2 support, as
shown in Fig. S11 for an experiment using ꢁ8 times more material.
cating competitive adsorption of CH3OH on the HDO sites. Further,
the introduction of the CH3OH co-feed resulted in the formation of
alkylated products including cresol and dimethyl phenol with
yields of ca. 10% and 5%, respectively, on a C-mol% basis of phenol
fed. Similar effects were observed when CH3OH was co-fed during
m-cresol HDO (Fig. S12). No alkylated products were observed for a
co-feed of CH3OH with benzene, suggesting the adsorption of phe-
nol through the O atom is critical for alkylation reactions to occur.
3.4.2. Alkylation with CH3OH co-feed during phenol HDO
In order to understand the role of methoxy species on oxy-
genate conversion, CH3OH was co-fed at a concentration of
PCH3OH = 0.16 bar during phenol HDO over 10 wt% MoO3/ZrO2
(Fig. 5). At a TOS = 4 h, a co-feed of CH3OH was introduced without
altering the value of PAnisole (resulting in a PCH3OH/PAnisole ꢅ 16). The
benzene yield decreased to almost zero, with a concomitant
increase in CH4 production (ꢁ5 mol% conversion of CH3OH), indi-
3.4.3. Co-feed with pyridine and DTBP
In order to further investigate the nature of active sites respon-
sible for anisole conversion, we performed co-feed experiments

M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
253
Fig. 4. Yield of benzene, toluene, phenol and cresol with and without co-feed of
H2O (0.032 bar PH2O). Reaction conditions: 150 mg 10 wt% MoO3/ZrO2.
Fig. 3. Steady-state kinetic studies for anisole conversion to phenol, cresol and
methyl anisole with respect to (a) anisole and (b) hydrogen concentrations.
Reaction conditions: 30 mg MoO3/ZrO2. PTotal = 1.013 bar (0.0049–0.0147 bar
P
Total = 1.013 bar (0.0098 bar PAnisole, balance H2 in the absence of H2O co-feed.
0.0098 bar PAnisole
,
0.032 bar PH2O
,
balance H2 with H2O co-feed).
Temperature = 593 K.
P
Anisole, balance H2, T = 593 K for a and 0.0098 bar PAnisole, 0.2508–1.0032 bar PH2
with balance N2 when PH2 < 1.0032 bar for (b). The reaction rate has units of mmol C
hꢂ1
g
.
-1
MoO3
with pyridine, a molecule that binds both to Lewis and Brønsted
acid sites, and DTBP, a molecule that binds only to Brønsted acid
sites [38]. As shown in Fig. 6, the pyridine co-fed
(Ppyridine = 0.0015 bar), decreased the yields of both HDO and oxy-
genate products, suggesting titration of sites responsible for HDO,
hydrogenolysis, and alkylation. Benzene and cresol selectivity val-
ues changed from 28 and 16%, to 39% and 11%, respectively
(Fig. S13). In contrast to the H2O co-feed experiments, the original
activity was not completely recovered after removing the pyridine
(ꢁ30% as compared to 45% prior to pyridine introduction), yet the
product distribution was not altered (Fig. S13). Pyridine co-feed
during phenol HDO showed similar results (Fig. S14).
The role of Brønsted acid sites for HDO was tested by introduc-
ing a co-feed of DTBP (PDTBP = ꢁ0.0015 bar) during anisole and phe-
nol HDO reactions. The bulky tert-butyl groups prevent DTBP from
binding to Lewis acid sites due to steric effects, but do not prevent
protonation by Brønsted acid sites [38]. As shown in Fig. 7, the
DTBP co-feed led to a negligible reduction in benzene and toluene
yields, which is in contrast to the results obtained with pyridine.
Minimal reduction in hydrogenolysis/alkylation activity was
observed in the presence of DTBP for anisole conversion. Specifi-
cally, the phenol, cresol, dimethyl phenol yield values decreased
from 6.0 to 4.8, 4.5 to 3.8, and 3.0 to 2.5%, respectively. Similar
results were obtained for phenol HDO (Fig. S15).
100
95
Benzene
0.16 bar PCH3OH
15
10
5
Cresol
Dimethyl
phenol
0
2
3
4
7
8
9
10
TOS (h)
Fig. 5. Yield of benzene, cresol and dimethyl phenol from phenol with and without
co-feed of methanol (ꢁ0.16 bar PCH3OH). The basis is the carbon moles of phenol fed.
Reaction conditions: 200 mg MoO3/ZrO2. PTotal = 1.013 bar (0.005 bar PPhenol
,
,
0.015 bar Pmesitylene, balance H2 in the absence of CH3OH co-feed. 0.005 bar PPhenol
0.16 bar PCH3OH, 0.015 bar Pmesitylene, balance H2 with CH3OH co-feed). Tempera-
ture = 593 K. 25 wt% phenol solution in mesitylene.

254
M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
Fig. 7. Yield of HDO products benzene, toluene, and oxygenate products phenol,
cresol, dimethyl phenol and methyl anisole with and without co-feed of DTBP
Fig. 6. Yield of HDO products benzene, toluene, and oxygenate products phenol,
cresol, dimethyl phenol and methyl anisole with and without co-feed of pyridine
(ꢁ0.0015 bar PDTBP). Reaction conditions: 100 mg 10 wt% MoO3/ZrO2. PTotal = 1.013 bar
(0.0098 bar PAnisole, balance H2 in the absence of DTBP co-feed. 0.0098 bar
(ꢁ0.0015 bar
Total = 1.013 bar (0.0098 bar PAnisole, balance H2 in the absence of pyridine co-feed.
0.0098 bar PAnisole 0.0015 bar Ppyridine balance H2 with pyridine co-feed).
Temperature = 593 K.
Ppyridine). Reaction conditions: 150 mg 10 wt% MoO3/ZrO2.
P
Anisole, 0.0015 bar PDTBP, balance H2 with DTBP co-feed). Temperature = 593 K.
P
,
,
methyl transfer reaction, and methyl anisole by intermolecular
alkylation. All the primary products can undergo sequential alkyla-
tion and/or HDO to form alkylated cresols, anisoles and benzenes,
with the exception of deoxygenated aromatics which cannot be
subsequently alkylated. A simplified kinetic model fitted for the
data in Fig. 1 (details in Supplementary Note 2) showed that
although the secondary pathway of phenol HDO to benzene has a
rate constant ꢁ 3 times higher than that for the HDO of anisole
to benzene, the anisole HDO pathway is dominant at low anisole
conversions (<15 C-mol%).
To analyze the effect of reaction temperature on the adsorption
of pyridine and DTBP on the catalyst surface, thermal treatment of
the catalyst at 623 K was carried out after pyridine and DTBP
adsorption and the FTIR spectra were analyzed. As shown in
Fig. S16c and S16d, the bands at 1530 and 1615 cmꢂ1 correspond-
ing to DTBP adsorption weaken post the thermal treatment. The
bands corresponding to both Lewis and Brønsted acid sites for pyr-
idine (1440, 1490, 1530, 1575, 1610 and 1640 cmꢂ1) reduced in
intensity post thermal treatment at 623 K. However, the reduction
in the intensity was not as significant as that seen for DTBP
(Figs. S16a and S16b).
Overall, our kinetic and co-feed studies suggested the involve-
ment of active sites with
a Lewis acid character for HDO
[27,28,33,48], consistent with our previously hypothesized path-
way driven by oxygen vacancies [26–28]. An apparent zero order
dependence on oxygenate concentration suggests that oxygenate
intermediates fully cover the oxygen vacancy sites at the reaction
conditions. We hypothesize that H2O and pyridine block these oxy-
gen vacancies and we posit that the drastic reduction of benzene
formation in the presence of CH3OH is due to the surface coverage
of HDO sites with methanol-derived intermediates.
The differences in reaction orders for oxygenate formation in
relation to HDO indicate that the activation of oxygenates toward
either HDO or hydrogenolysis/alkylation must occur on different
sites. The lower activation energy for phenol formation via
hydrogenolysis on both the 10 wt% MoO3/ZrO2 and ZrO2 as com-
pared to benzene formation via HDO further supports this hypoth-
esis. We reason that the hydrogenolysis and alkylation reactions
4. Discussion
Reaction network analyses have been used to obtain mechanis-
tic insights on reaction pathways, nature of active sites, and sup-
port effect on reactivity [18,35,36,39–44]. An analogous reaction
network analyses for anisole conversion demonstrate the bifunc-
tional nature of molybdenum supported on zirconia catalysts for
catalyzing HDO, hydrogenolysis and alkylation reactions
(Scheme 2). Indeed, previous studies on anisole deoxygenation
and alcohol dehydration have shown that MoO3/ZrO2 feature both
redox and acid sites [33,45–47]. Anisole is primarily converted to
benzene by selective HDO, phenol by hydrogenolysis, cresol by a

M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
255
are catalyzed by acid sites, as these sites are blocked by pyridine
but cannot be blocked to a significant extent neither by H2O,
CH3OH nor DTBP at the reaction conditions investigated and are
enhanced in the presence of H2O. Rather than promoting
expression can be obtained with the assumption of the H* addition
to Ph-S2 as the RDS.
pffiffiffiffiffiffiffiffiffiffiffiffiffi
rPhenol!Benzene ¼ k4K3 K1PH2
ð7Þ
Caromatic-O cleavage, these sites catalyze hydrolysis of the
The HDO of anisole to benzene is complicated. The HDO,
hydrogenolysis and alkylation reactions occur at comparable rates,
and their measurements are affected by secondary reactions
(Scheme 2). For example, CH3 groups needed for alkylation are pro-
vided by HDO and hydrogenolysis, which exhibit different reaction
orders. Our reaction kinetics and co-feed studies suggests that
there are 3 distinct active sites responsible for H2 dissociation (*),
HDO (S2) or hydrogenolysis/alkylation reactions (S3). The PH2
dependencies on conversion of anisole to benzene and phenol to
benzene are different, suggesting that either the RDS or H additions
prior to the RDS, are different. One possibility is that anisole is first
converted to phenol, followed by subsequent HDO of phenol to
benzene. However, our analysis of the primary product selectivities
with conversion (Fig. 1) suggests that this scenario is unlikely.
Another possibility is that significant differences between
-OCaliphaticH3 and –OH groups alter the barrier heights enough to
change the dominant mechanism. Here, we show one possible
mechanism consistent with our data, but we acknowledge other
possibilities as shown in Supplementary Notes 4–6. Further work
will be needed to fully elucidate the reaction mechanism.
In our model, molecular H2 dissociatively adsorbs on the cata-
lyst surface (*), while anisole (An) can adsorb on an oxygen
vacancy (S2), as suggested in our previous reports on bulk and sup-
ported MoO3 catalysts [27,28]. We posit that the cleavage of the
stronger Caromatic-O bond (R8) occurs after two H addition. In this
scheme, Reaction R8 is assumed to be the RDS with An-S2 to be
the MARI on S2. This is similar with assumptions made by Bhan
and coworkers on HDO reactions of oxygenates on Mo2C catalysts
[52]. We consider that both H* and CH3* are mobile on the catalyst
surface so that they can easily get to/from sites S2 and S3 to be
involved in HDO and oxygenate formation. Hence, we suggest H*
and CH3* as surface intermediates on the catalyst surface without
assigning to any particular site. This assumption was made to sim-
plify the kinetic analysis. Finally, the site S2 is regenerated with the
release of H2O.
O-Caliphatic bond, generating CH3OH in the presence of H2O
[35,36]. The hydrolysis increased the rates of phenol and cresol for-
mation by ꢁ 2 times (Fig. 4) when using a 20-fold increase in PH2O
(with 40% conversion of anisole as reference). Given that H2O is a
byproduct of anisole HDO, it may explain contribution towards a
small amount of phenol and cresol observed. Under differential
conditions (<15% conversion), the contribution of water-promoted
hydrolysis would be less than 2%. Therefore, the measured kinetics
under differential conditions are likely not significantly influenced
by the presence of H2O due to HDO of anisole. In turn, CH3OH read-
ily alkylates the aromatic oxygenates, but not the deoxygenated
arenes, suggesting the activation of the oxygenate through the O
atom is critical for alkylation, consistent with a report by Anderson
et al. on Mo-based polyoxometallates that feature both Lewis and
Brønsted acid sites [12]. We note that this pathway is different from
metal catalyzed hydrogenolysis of O-Caliphatic bonds [36,49].
We first consider a mechanism for phenol conversion to ben-
zene (Table 2). This is comparatively easier to interpret due to
the absence of alkylation and hydrogenolysis reactions. The exper-
imental data strongly indicate that H2 dissociation and HDO occur
on two distinct sites: (*) and (S2), respectively. On pre-reduced
MoO3/ZrO2 (70 ml minꢂ1 of 1.013 bar PH2 at 593 K for 1 h), only
trace amounts of cyclohexane was observed at both 523 and
593 K when ꢁ0.01 bar PBenzene was fed, suggesting very few metal
sites are present at reaction conditions. While we cannot fully rule
out the presence and influence of metallic sites for H2 dissociation,
it is also plausible that the sites responsible for H2 activation could
be undercoordinated Mo or Zr (for MoO3/ZrO2) moeities. Recent
experimental and computational studies on CeO2-ZrO2 and RuO2
have established that oxides can activate H2 with an energy barrier
less than 100 kJ molꢂ1 [50,51], which is within the range of
observed activation energies in our current study. We posit that
one H addition occurs on adsorbed phenol (Ph) prior to the rate-
determining step (RDS, R4). With the assumption that the most
abundant reactive intermediate (MARI) on S2 is Ph-S2, a reaction
rate expression with 0 and 0.5 reaction order with respect to Pphenol
and PH2, respectively, is obtained (Supplementary Note 4).
The formation of benzene (Bz) from anisole (An) can be found to
be represented by (Supplementary Note 4) Equation 10 as given
below
The formation of benzene (Bz) from phenol (Ph) is found to be
represented as given in Equation (7) below. A similar reaction rate
rAnisole!Benzene ¼ k8K1K6K7PH2
ð8Þ
Table 2
Elementary steps for hydrodeoxygenation (HDO) of phenol and anisole and formation of phenol, cresol and methyl anisole from anisole on MoO3/ZrO2.
Dissociative hydrogen adsorption
H2 þ 2ꢆ $ þ 2Hꢆ
Anisole adsorption on S3
An þ S3 $ An - S3
(R1)
(R15)
Phenol HDO
Phenol formation
(R2)
(R3)
(R16)
(R17)
Ph þ S2 $ Ph - S2
An - S3þHꢆ $ PhCH3 - S3 þꢆ
Ph - S2þ Hꢆ $ PhH - S2 þꢆ
RDS
PhCH3 - S3 þꢆ ! Ph - S3 þ CHꢆ3
RDS
(R4)
(R18)
Ph - S3 $ Ph þ S3
PhH - S2 ! Bz þ OH - S2
Anisole HDO
Methyl anisole formation
RDS
An þ S2 $ An - S2
(R5)
(R19)
An - S3 þ CHꢆ3 ! MAH - S3 þꢆ
An - S2þ Hꢆ $ PhCH3 - S2 þꢆ
MAH - S3 þꢆ $ MA - S3 þ Hꢆ
MA - S3 $ MA þ S3
(R6)
(R7)
(R20)
(R21)
PhCH3 - S2 þ Hꢆ $ BzOHCH3 - S2 þꢆ
RDS
(R8)
(R9)
Cresol formation
BzOHCH3 - S2 ! Bz þ CH3OH - S2
CH3OH - S2 þꢆ ! CH3ꢆ þ OH - S2
RDS
(R22)
Ph - S3 þ CHꢆ3 ! CrH - S3 þꢆ
CH3OH - S2 þꢆ $ Hꢆþ CH3O - S2
CrH - S3 þꢆ $ Cr - S3þ Hꢆ
Cr - S3 $ Cr þ S3
(R10)
(R11)
(R12)
(R13)
(R14)
(R23)
(R24)
CHꢆ3 þ Hꢆ ! CH4 þ 2ꢆ
CH3OH - S2 $ CH3OH þ S2
OH - S2þ Hꢆ $ H2O - S2 þꢆ
H2O - S2 $ H2O þ S2

256
M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
The formation of phenol on S3 and benzene on S2 from anisole
Appendix A. Supplementary material
can be hypothesized to be due to the difference in oxophilicity of
the sites. The oxophilicity of the sites plays a role in adsorption
of aromatic oxygenates by facilitating the adsorption through the
oxygen atom and repulsion of phenyl ring with the catalyst surface
[53]. Furthermore, it has been shown to reduce activation barriers
for direct Caromatic-O cleavage [49]. We hypothesize the oxophilic
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2019.06.046.
References
sites (S2) in MoO3/ZrO2 play
a similar role promoting the
[1] G.W. Huber, S. Iborra, A. Corma, Synthesis of transportation fuels from
biomass: chemistry, catalysts, and engineering, Chem. Rev. 106 (2006) 4044–
4098.
[2] G.W. Huber, A. Corma, Synergies between bio- and oil refineries for the
production of fuels from biomass, Angew. Chem. 46 (2007) 7184–7201.
[3] Q. Zhang, J. Chang, T. Wang, Y. Xu, Review of biomass pyrolysis oil properties
and upgrading research, Energy Convers. Manag. 48 (2007) 87–92.
[4] T.R. Carlson, G.A. Tompsett, W.C. Conner, G.W. Huber, Aromatic production
from catalytic fast pyrolysis of biomass-derived feedstocks, Top. Catal. 52
(2009) 241–252.
[5] E. Furimsky, Hydroprocessing challenges in biofuels production, Catal. Today
217 (2013) 13–56.
[6] A.J. Ragauskas, G.T. Beckham, M.J. Biddy, R. Chandra, F. Chen, M.F. Davis, B.H.
Davison, R.A. Dixon, P. Gilna, M. Keller, P. Langan, A.K. Naskar, J.N. Saddler, T.J.
Tschaplinski, G.A. Tuskan, C.E. Wyman, Lignin valorization: improving lignin
processing in the biorefinery, Science 344 (2014).
[7] K. Wang, P.A. Johnston, R.C. Brown, Comparison of in-situ and ex-situ catalytic
pyrolysis in a micro-reactor system, Bioresour. Technol. 173 (2014) 124–131.
[8] S. Van den Bosch, W. Schutyser, S.T. Renders, T. De Boe, S.F. Koelewijn, A.
Dewaele, T. Ennaert, O. Verkinderen, B. Goderis, C.M. Courtin, B.F. Sels,
Influence of bio-based solvents on the catalytic reductive fractionation of birch
wood, Green Chem. 17 (2015) 5035–5045.
[9] S. Van den Bosch, W. Schutyser, S.F. Koelewijn, T. Renders, C.M. Courtin, B.F.
Sels, Tuning the lignin oil OH-content with Ru and Pd catalysts during
lignin hydrogenolysis on birch wood, Chem. Commun. 51 (2015) 13158–
13161.
[10] S. Van den Bosch, W. Schutyser, R. Vanholme, T. Driessen, S.F. Koelewijn, T.
Renders, B. De Meester, W.J.J. Huijgen, W. Dehaen, C.M. Courtin, B. Lagrain, W.
Boerjan, B.F. Sels, Reductive lignocellulose fractionation into soluble lignin-
derived phenolic monomers and dimers and processable carbohydrate pulps,
Energy Environ. Sci 8 (2015) 1748–1763.
adsorption of anisole through the O atom and facilitates direct
Caromatic-O bond cleavage.
We assume the cleavage of the Caliphatic-O bond (R17) and CH3*
addition to An-S3 (R19) and Ph-S3 (R22) to be the RDS for phenol,
methyl anisole and cresol formation, respectively. Due to the cou-
pled nature of the reaction network, the exact reaction orders in
the kinetic expressions of phenol, methyl anisole and cresol forma-
tion rates on MoO3/ZrO2 could not be derived. However, we note
that these expressions correctly describe our kinetic data. The
detailed derivations can be found in Supplementary Note 4. We
further note that the kinetic rate expressions on ZrO2 with decou-
pled HDO and oxygenate formation pathways are consistent with
experimental data (Supplementary Note 4 through 6).
The presence of CH3OH or H2O leads to an increase in methoxy
or hydroxyl species on S2 (H2O-S2 and OH-S2, or CH3OH-S2 or
CH3O-S2), leading to a low coverage of intermediates from anisole
or phenol and low HDO rates. When CH3OH or H2O is introduced
with PH2O and PCH3OH significantly greater than PAnisole, HDO is
blocked and selectivities towards hydrogenolysis and alkylation
are also significantly altered. The alkylation of phenol when CH3OH
is fed (R19 and R22) suggests that CH3OH reacts through R9 lead-
ing to formation of CH3* species.
[11] E.M. Anderson, R. Katahira, M. Reed, M.G. Resch, E.M. Karp, G.T. Beckham, Y.
Román-Leshkov, Reductive catalytic fractionation of corn stover lignin, ACS
Sustain. Chem. Eng. 4 (2016) 6940–6950.
[12] E.M. Anderson, A. Crisci, K. Murugappan, Y. Roman-Leshkov, Bifunctional
molybdenum polyoxometalates for the combined hydrodeoxygenation and
alkylation of lignin-derived model phenolics, ChemSusChem (2017) 2226–
2234.
[13] E.M. Anderson, M.L. Stone, R. Katahira, M. Reed, G.T. Beckham, Y. Román-
Leshkov, Flowthrough reductive catalytic fractionation of biomass, Joule 1
(2017) 613–622.
[14] T. Iida, M. Shetty, K. Murugappan, Z. Wang, K. Ohara, T. Wakihara, Y. Roman-
Leshkov, Encapsulation of molybdenum carbide nanoclusters inside zeolite
micropores enables synergistic bifunctional catalysis for anisole
hydrodeoxygenation, ACS Catal. (2017) 8147–8151.
[15] Q. Lu, C.J. Chen, W. Luc, J.G.G. Chen, A. Bhan, F. Jiao, Ordered mesoporous metal
carbides with enhanced anisole hydrodeoxygenation selectivity, ACS Catal. 6
(2016) 3506–3514.
[16] E.M. Anderson, M.L. Stone, M.J. Hulsey, G.T. Beckham, Y. Roman-Leshkov,
Kinetic studies of lignin solvolysis and reduction by reductive catalytic
fractionation decoupled in flow-through reactors, ACS Sustain. Chem. Eng. 6
(2018) 7951–7959.
5. Conclusions
In summary, we carried out a comprehensive kinetic study for
the vapor-phase anisole HDO over a 10 wt% MoO3/ZrO2 catalyst.
Benzene, phenol, cresol and methyl anisole were confirmed as pri-
mary reaction products, while dimethyl phenol, toluene and
dimethyl anisole, and the higher alkylated phenols, benzenes and
anisoles were secondary and tertiary products, respectively. Co-
feed and kinetic analyses suggested the involvement of distinct
sites for the activation of anisole for HDO and for hydrogenoly-
sis/alkylation. The kinetic models suggest that anisole is activated
on an oxygen vacancy site with a Lewis acid character for HDO. The
formation of deoxygenated products on an oxygen vacancy may be
related to the oxophilicity of the undercoordinated surface. Our
study paves the way for a molecular understanding of catalysis
on bulk and supported molybdenum oxide catalysts for HDO.
[17] M. Shetty, D. Zanchet, W.H. Green, Y. Román-Leshkov, Cooperative Co0/CoII
sites stabilized by a perovskite matrix enable selective CꢂO and CꢂC bond
hydrogenolysis of oxygenated arenes, ChemSusChem 12 (2019) 2171–2175.
[18] L. Nie, D.E. Resasco, Kinetics and mechanism of m-cresol hydrodeoxygenation
on a Pt/SiO2 catalyst, J. Catal. 317 (2014) 22–29.
[19] K. Lee, G.H. Gu, C.A. Mullen, A.A. Boateng, D.G. Vlachos, Guaiacol
hydrodeoxygenation mechanism on Pt(111): insights from density
functional theory and linear free energy relations, ChemSusChem 8 (2015)
315–322.
Author contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
[20] M.M. Sullivan, C.J. Chen, A. Bhan, Catalytic deoxygenation on transition metal
carbide catalysts, Catal. Sci. Tech. 6 (2016) 602–616.
Declaration of Competing Interest
[21] M.J. Gilkey, C. Brady, D.G. Vlachos, B.J. Xuo, Characterization of oxidation states
in metal/metal oxide catalysts in liquid-phase hydrodeoxygenation reactions
with a trickle bed reactor, Ind. Eng. Chem. Res. 57 (2018) 5591–5598.
[22] C. Wang, A.V. Mironenko, A. Raizada, T. Chen, X. Mao, A. Padmanabhan, D.G.
Vlachos, R.J. Gorte, J.M. Vohs, Mechanistic study of the direct
hydrodeoxygenation of m-Cresol over WOx-Decorated Pt/C catalysts, ACS
Catal. 8 (2018) 7749–7759.
The authors declared that there is no conflict of interest.
Acknowledgment
[23] D.R. Moberg, T.J. Thibodeau, F.G. Amar, B.G. Frederick, Mechanism of
hydrodeoxygenation of acrolein on a cluster model of MoO3, J. Phys. Chem.
C 114 (2010) 13782–13795.
[24] D.H. Mei, A.M. Karim, Y. Wang, Density functional theory study of
acetaldehyde hydrodeoxygenation on MoO3, J. Phys. Chem. C 115 (2011)
8155–8164.
This research was funded by BP through the MIT Energy Initia-
tive Advanced Conversion Research Program. Eric Anderson
acknowledges support by the National Science Foundation, CBET
Award No 1454299.

M. Shetty et al. / Journal of Catalysis 376 (2019) 248–257
257
[25] K. Murugappan, E.M. Anderson, D. Teschner, T.E. Jones, K. Skorupska, Y.
Roman-Leshkov, Operando NAP-XPS unveils differences in MoO3 and Mo2C
during hydrodeoxygenation, Nat. Catal. 1 (2018) 960–967.
[26] T. Prasomsri, T. Nimmanwudipong, Y. Roman-Leshkov, Effective
hydrodeoxygenation of biomass-derived oxygenates into unsaturated
reaction network including hydrodeoxygenation reactions, Energy Fuels 25
(2011) 3417–3427.
[40] T. Nimmanwudipong, R.C. Runnebaum, D.E. Block, B.C. Gates, Catalytic
reactions of guaiacol: reaction network and evidence of oxygen removal in
reactions with hydrogen, Catal. Lett. 141 (2011) 779–783.
hydrocarbons by MoO3 using low H2 pressures, Energy Environ. Sci.
(2013) 1732–1738.
6
[41] R.C. Runnebaum, R.J. Lobo-Lapidus, T. Nimmanwudipong, D.E. Block, B.C.
Gates, Conversion of anisole catalyzed by platinum supported on alumina: the
reaction network, Energy Fuels 25 (2011) 4776–4785.
[42] T. Nimmanwudipong, C. Aydin, J. Lu, R.C. Runnebaum, K.C. Brodwater, N.D.
Browning, D.E. Block, B.C. Gates, Selective hydrodeoxygenation of guaiacol
catalyzed by platinum supported on magnesium oxide, Catal. Lett. 142 (2012)
1190–1196.
[27] T. Prasomsri, M. Shetty, K. Murugappan, Y. Román-Leshkov, Insights into the
catalytic activity and surface modification of MoO3 during the
hydrodeoxygenation of lignin-derived model compounds into aromatic
hydrocarbons under low hydrogen pressures, Energy Environ. Sci. 7 (2014)
2660–2669.
[28] M. Shetty, K. Murugappan, T. Prasomsri, W.H. Green, Y. Roman-Leshkov,
Reactivity and stability investigation of supported molybdenum oxide
catalysts for the hydrodeoxygenation (HDO) of m-cresol, J. Catal. 331 (2015)
86–97.
[43] R.C. Runnebaum, T. Nimmanwudipong, D.E. Block, B.C. Gates, Catalytic
conversion of compounds representative of lignin-derived bio-oils:
reaction network for guaiacol, anisole, 4-methylanisole, and cyclohexanone
conversion catalysed by Pt/ -Al2O3, Catal. Sci. Technol. 2 (2012) 113–118.
[44] R.C. Runnebaum, T. Nimmanwudipong, R.R. Limbo, D.E. Block, B.C. Gates,
Conversion of 4-methylanisole catalyzed by Pt/ -Al2O3 and by Pt/SiO2-Al2O3:
a
c
[29] S. Budhi, C. Mukarakate, K. Iisa, S. Pylypenko, P.N. Ciesielski, M.M. Yung, B.S.
Donohoe, R. Katahira, M.R. Nimlos, B.G. Trewyn, Molybdenum incorporated
mesoporous silica catalyst for production of biofuels and value-added
chemicals via catalytic fast pyrolysis, Green Chem. 17 (2015) 3035–3046.
[30] K. Murugappan, C. Mukarakate, S. Budhi, M. Shetty, M.R. Nimlos, Y. Román-
Leshkov, Supported molybdenum oxides as effective catalysts for the catalytic
fast pyrolysis of lignocellulosic biomass, Green Chem. 5548–5557 (2016).
[31] M.W. Nolte, J. Zhang, B.H. Shanks, Ex situ hydrodeoxygenation in biomass
pyrolysis using molybdenum oxide and low pressure hydrogen, Green Chem.
18 (2016) 134–138.
[32] G. Zhou, P.A. Jensen, D.M. Le, N.O. Knudsen, A.D. Jensen, Atmospheric
hydrodeoxygenation of biomass fast pyrolysis vapor by MoO3, ACS Sustain.
Chem. Eng. 4 (2016) 5432–5440.
[33] M. Shetty, K. Murugappan, W.H. Green, Y. Roman-Leshkov, Structural
properties and reactivity trends of molybdenum oxide catalysts supported
on zirconia for the hydrodeoxygenation of anisole, ACS Sustain. Chem. Eng. 5
(2017) 5293–5301.
c
reaction networks and evidence of oxygen removal, Catal. Lett. 142 (2012) 7–
15.
[45] J. Macht, C.D. Baertsch, M. May-Lozano, S.L. Soled, Y. Wang, E. Iglesia, Support
effects on Bronsted acid site densities and alcohol dehydration turnover rates
on tungsten oxide domains, J. Catal. 227 (2004) 479–491.
[46] J. Xu, A.M. Zheng, J. Yang, Y.C. Su, J.Q. Wang, D.L. Zeng, M.J. Zhang, C.H. Ye, F.
Deng, Acidity of mesoporous MoOx/ZrO2 and WOx/ZrO2 materials:
a
combined solid-state NMR and theoretical calculation study, J. Phys. Chem. B
110 (2006) 10662–10671.
[47] R.M. West, D.J. Braden, J.A. Dumesic, Dehydration of butanol to butene over
solid acid catalysts in high water environments, J. Catal. 262 (2009) 134–143.
[48] M. Shetty, B. Buesser, Y. Román-Leshkov, W.H. Green, Computational
Investigation on Hydrodeoxygenation (HDO) of acetone to propylene on
MoO3 (010) surface, J. Phys. Chem. C 121 (2017) 17848–17855.
a-
[49] Q.H. Tan, G.H. Wang, A. Long, A. Dinse, C. Buda, J. Shabaker, D.E. Resasco,
Mechanistic analysis of the role of metal oxophilicity in the
hydrodeoxygenation of anisole, J. Catal. 347 (2017) 102–115.
[50] S.M. Schimming, G.S. Foo, O.D. LaMont, A.K. Rogers, M.M. Yung, A.D. D’Amico,
C. Sievers, Kinetics of hydrogen activation on ceria-zirconia, J. Catal. 329
(2015) 335–347.
[51] A.V. Mironenko, D.G. Vlachos, Conjugation-driven ‘‘Reverse Mars–van
Krevelen”-type radical mechanism for low-temperature C-O bond activation,
J. Am. Chem. Soc. 138 (2016) 8104–8113.
[52] A. Kumar, S. Phadke, A. Bhan, Acetic acid hydrodeoxygenation on molybdenum
carbide catalysts, Catal. Sci. Technol 8 (2018) 2938–2953.
[53] P.M. de Souza, L. Nie, L.E.P. Borges, F.B. Noronha, D.E. Resasco, Role of oxophilic
supports in the selective hydrodeoxygenation of m-cresol on Pd catalysts,
Catal. Lett. 144 (2014) 2005–2011.
[34] M.J. Frisch, Gaussian 03, Gaussian, 2004.
[35] X.L. Zhu, R.G. Mallinson, D.E. Resasco, Role of transalkylation reactions in the
conversion of anisole over HZSM-5, Appl. Catal. A 379 (2010) 172–181.
[36] X.L. Zhu, L.L. Lobban, R.G. Mallinson, D.E. Resasco, Bifunctional transalkylation
and hydrodeoxygenation of anisole over a Pt/HBeta catalyst, J. Catal. 281
(2011) 21–29.
[37] W.S. Lee, Z.S. Wang, R.J. Wu, A. Bhan, Selective vapor-phase
hydrodeoxygenation of anisole to benzene on molybdenum carbide
catalysts, J. Catal. 319 (2014) 44–53.
[38] C.D. Baertsch, K.T. Komala, Y.H. Chua, E. Iglesia, Genesis of Bronsted acid sites
during dehydration of 2-butanol on tungsten oxide catalysts, J. Catal. 205
(2002) 44–57.
[39] T. Nimmanwudipong, R.C. Runnebaum, D.E. Block, B.C. Gates, Catalytic
conversion of guaiacol catalyzed by platinum supported on alumina: