Mo2C Modification by CO2, H2O, and O2: Effects of Oxygen Content and Oxygen Source on Rates and Selectivity of m-Cresol Hydrodeoxygenation

Article abstract of DOI:10.1021/acscatal.6b02762

Vapor-phase m-cresol hydrodeoxygenation rates on oxygenate-modified Mo₂C catalysts prepared by pretreating fresh Mo₂C catalysts in 1 kPa of O₂, H₂O, and CO₂ at 333 K showed that (i) molecular oxygen has a higher propensity to deposit oxygen (O/Mo_bulk before HDO = 0.23 ± 0.02) on fresh Mo₂C, compared to CO₂ and H₂O (O/Mo_bulk before HDO a‰ 0.036), as assessed from temperature-programmed surface reaction with H₂, and (ii) oxygen adsorbed in amounts exceeding ～0.06 ± 0.01 of O/Mo_bulk poisons the metal-like sites for toluene synthesis as inferred from a 10-fold decrease in toluene synthesis rate per gram on the O₂-1 kPa (333 K)-Mo₂C compared to that on fresh Mo₂C, H₂O-1 kPa (333 K)-Mo₂C, and CO₂-1 kPa (333 K)-Mo₂C catalysts. Invariant turnover frequencies of toluene synthesis measured from in situ CO titration among the O₂-, H₂O-, and CO₂-modified samples demonstrate that the effect of adsorbed oxygen is independent of the oxygen source.

Full text of DOI:10.1021/acscatal.6b02762

Research Article
pubs.acs.org/acscatalysis
Mo2C Modification by CO2, H2O, and O2: Effects of Oxygen Content
and Oxygen Source on Rates and Selectivity of m‑Cresol
Hydrodeoxygenation
Cha-Jung Chen and Aditya Bhan*
Department of Chemical Engineering and Materials Science, University of Minnesota−Twin Cities, 421 Washington Avenue SE,
Minneapolis, Minnesota 55455, United States
S
* Supporting Information
ABSTRACT: Vapor-phase m-cresol hydrodeoxygenation
rates on oxygenate-modified Mo2C catalysts prepared by
pretreating fresh Mo2C catalysts in 1 kPa of O2, H2O, and CO2
at 333 K showed that (i) molecular oxygen has a higher
propensity to deposit oxygen (O/Mobulk before HDO = 0.23
0.02) on fresh Mo2C, compared to CO2 and H2O (O/
Mobulk before HDO ≈ 0.036), as assessed from temperature-
programmed surface reaction with H2, and (ii) oxygen
adsorbed in amounts exceeding ∼0.06
0.01 of O/Mobulk
poisons the metal-like sites for toluene synthesis as inferred
from a 10-fold decrease in toluene synthesis rate per gram on
the O2−1 kPa (333 K)−Mo2C compared to that on fresh
Mo2C, H2O−1 kPa (333 K)−Mo2C, and CO2−1 kPa (333 K)−Mo2C catalysts. Invariant turnover frequencies of toluene
synthesis measured from in situ CO titration among the O2-, H2O-, and CO2-modified samples demonstrate that the effect of
adsorbed oxygen is independent of the oxygen source.
KEYWORDS: m-cresol, hydrodeoxygenation, lignin, biofuels, oxygenate-modification
Oxygen deposition occurred during this transient; concurrently,
cyclohexane rates dropped to almost zero and benzene was
observed, suggesting that hydrogenation of benzene to
cyclohexane was inhibited by oxygen deposition (see Figure
4a in ref 25). A similar effect of oxygen deposition was reported
in our recent studies, noting that the hydrogenation rates of
benzene and toluene on Mo2C were irreversibly inhibited when
water or methanol was co-fed (see Figure 5 in ref 10). These
results demonstrate that oxygen deposition under reaction
environments alters catalyst composition and function; the
effect of the source of oxygen on HDO catalysis is yet unclear.
Metal-like sites were found to be involved in selective
1. INTRODUCTION
Transition-metal carbides have been shown to catalyze
hydrodeoxygenation (HDO) reactions to convert biomass-
derived compounds to fuels and chemicals.1−8 These catalysts
are selective in cleaving C−O and CO bonds at ambient H2
pressure and low temperatures (423−623 K),1,2,7,9,10 such that
the aromaticity of furan and benzene rings can be retained
during HDO.10,11 Molybdenum carbides exhibit stable surfaces
that vary with surface coverage of O and C, as a result of
reaction environments or synthesis protocols,12 and display
heterogeneity in both bulk13 and surface14,15 compositions, and
in surface structures, including distinct binding sites16−19 and
nonstoichiometric surface compositions.13,20 This complexity in
bulk and surface structure of molybdenum carbides is
accentuated in the presence of oxygenates during HDO of
bio-oil or pyrolysis vapor on the oxophilic transition-metal
carbides,21−24 which can deposit oxygen on the catalyst. Surface
oxygen species can poison the metal-like sites required for
HDO25 and/or generate Brønsted acid sites.26
+
benzene synthesis (>90% C6 selectivity) from vapor-phase
hydrodeoxygenation of anisole on Mo2C and W2C catalysts, as
co-feeds of carbon monoxide were noted to inhibit HDO rates
under reaction conditions and turnover frequencies of benzene
−1
synthesis at 423 K (∼10−3 mol molCO s−1 for Mo2C and
∼10−4 mol molCO s−1 for W2C) were noted to be invariant
−1
with different CO co-feed pressures.25,27 The turnover
Oxygen (∼4.7 × 10−4 mol gcat−1) and carbon (equivalent of
frequency of benzene synthesis from anisole HDO on a fresh
Mo2C formulation (∼1.7 × 10−3 mol molCO s−1) was found
−1
∼1.4 × 10−3 mol gcat C6 species) deposition was observed
−1
to be almost the same as that on an O2-modified Mo2C catalyst
during an initial transient observed during anisole HDO on
bulk β-Mo2C that had been passivated in 1% O2/He and
pretreated in H2 for 1 h at 723 K prior to reaction at 423 K.25
Cyclohexane was observed in the effluent during this initial
transient, as observed from online mass spectrometric analysis.
Received: September 27, 2016
Revised: December 17, 2016
© XXXX American Chemical Society
1113
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
(∼1.5 × 10−3 mol molCO s−1; O/Mobulk (molar ratio) =
0.075), while the benzene synthesis rate normalized per gram
catalyst was three times higher on the fresh Mo2C, indicating
that the number of sites responsible for anisole HDO on Mo2C
decreases with O2 treatment.25
chemical titration and m-cresol HDO studies on fresh and
oxygenate-modified Mo2C catalysts showed that adsorbed
oxygen poisons the metal-like sites responsible for m-cresol
HDO and that the effect of adsorbed oxygen is independent of
the source of oxygen.
−1
Oxygen-modified tungsten carbide materials prepared by
exposing freshly synthesized tungsten carbides to O2 at various
temperatures have been reported to exhibit bifunctionality.28−30
Alkane (n-hexane and n-heptane) hydrogenolysis rates on
bifunctional tungsten carbides decreased concurrently with an
increase of alkane isomerization rates as the extent of oxygen
modification, quantified by temperature-programmed reduction
(TPR) and ex situ CO chemisorption, increased.28,29 Sullivan et
al.26 reported that a ∼30-fold increase in propylene synthesis
rates per gram Mo2C catalyst from isopropyl alcohol
dehydration at 415 K was observed when O2 co-feed was
varied from 0 to 13.5 kPa. Brønsted acid sites were found to be
responsible for isopropyl alcohol dehydration to propylene on
Mo2C formulations with oxygen co-feed (0−13.5 kPa), as
evidenced by the invariance of propylene rates normalized to
acid sites titrated in situ by 2,6-di-tert-butylpyridine.26 The O 1s
peak at 532.5 eV assigned to surface hydroxyl species observed
from X-ray photoelectron spectroscopy (XPS) on a Mo2C
catalyst after H2 pretreatment was correlated with the surface
acidity of Mo2C measured from temperature-programmed
desorption of ammonia (NH3-TPD).1 These surface hydroxyl
groups (O 1s peak at 532.5 eV) were also observed on spent
Mo2C catalysts from the hydrodeoxygenation of acetic acid at
623 K without exposure to air.1 These results show that a
change in site densities of metal and acid functions on tungsten
carbide formulations can result from changing O2 pretreatment
conditions and the functionality of Mo2C can be adjusted to
primarily acidic by varying the O2 co-feed pressure during the
reaction.
The propensity for O2, CO2, and H2O as oxidants to oxidize
pure β-Mo2C surface was evaluated using density functional
theory (DFT) calculations on β-Mo2C (011) and (101)
surfaces with mixed Mo/C terminations.31 The reaction
energies for surface oxidation by O2, CO2, and H2O to deposit
0.125 monolayer (ML) to 1 ML of oxygen on these surfaces
showed that O2−Mo2C reactions have a much larger enthalpy
(−13 eV to −2 eV) than those of H2O−Mo2C and
CO2−Mo2C (approximately −2 eV to 0 eV or >0 eV)
reactions at 0 K.31 However, it is still unclear (i) how the
differences in the oxidation propensity of O2, CO2, and H2O
translate to the amount of oxygen adsorbed on bulk Mo2C,
which may contain residual oxygen in the structure after
catalyst synthesis,32 and (ii) what are the resulting con-
sequences of oxygen loadings from different oxygen sources on
HDO rates and selectivity catalyzed by Mo2C.
2. MATERIALS AND METHODS
2.1. Kinetics and In Situ Chemical Titration Studies for
m-Cresol HDO. 2.1.1. Catalyst Synthesis and Character-
ization. Molybdenum carbide catalysts were synthesized based
on protocols reported in a prior report.10 Ammonium
molybdate tetrahydrate (∼1.2 g, Sigma, 99.98%, trace metal
basis; sieved, 177−400 μm, (NH4)6Mo7O24·4H2O) was loaded
to a tubular quartz reactor (inner diameter (ID) of 10 mm)
placed in a tube furnace (Applied Test System, Series 3210)
controlled by a Watlow Temperature Controller (96 series)
and was treated in a gas mixture (total flow rate of ∼2.75 cm3
s−1) of 15 vol %/85 vol % of CH4 (Matheson, 99.97%) and H2
(Minneapolis Oxygen, 99.999%) at ∼623 K for 5 h at a
ramping rate of ∼0.06 K s−1 from room temperature (RT),
then treated at ∼863 K for 3 h at a ramping rate of ∼0.047 K
s−1. Passivation of the resulting material was done at RT using
1% O2/He (Matheson, Certified Standard Purity; total flow ≈
1.67 cm3 s−1) for ∼2 h.
N2 adsorption/desorption (Micromeritics ASAP 2020)
measurements at ∼77 K were used to determine the
Brunauer−Emmett−Teller (BET) surface area of Mo2C
catalysts; the sample was degassed (<10 μm Hg) at 523 K
for at least 4 h before N2 adsorption. X-ray diffraction (XRD)
(using a Bruker, Model D8 Discover, 2D X-ray diffractometer
with a two-dimensional VÅNTEC-500 detector) was used to
determine the bulk structure of molybdenum carbide after m-
cresol HDO. Two-dimensional images were collected using Cu
Kα X-ray radiation with a graphite monochromator and a 0.8
mm point collimator measured in three measurement frames at
2θ = 25°, 55°, and 85° with a 900 s frame/dwell, which were
then converted to one-dimensional (1D) intensity vs 2θ for
analysis. Ex situ CO chemisorption (Matheson, 99.5%) uptake
at 323 K for Mo2C samples was measured using a Micro-
meritics ASAP 2020 instrument, in which ∼0.14 g passivated
Mo2C was treated in H2 at 723 K for 1 h, followed by degassing
(∼2 μm Hg) at 723 K for 2 h, after being evacuated at 383 K
(∼2 μm Hg) for 0.5 h. The uptake of irreversibly adsorbed CO
was then obtained from the difference between two adsorption
isotherms (from 100 mm Hg to 450 mm Hg at 323 K)
extrapolated to zero pressure; the second isotherm was taken
after the cell was degassed at ∼2 μm Hg to remove weakly
adsorbed species. Ex situ CO chemisorption measurements
were conducted within a seven-day period prior to the kinetic
studies for vapor-phase m-cresol HDO.
We report the effects of oxygenate modification on Mo2C for
m-cresol HDO on Mo2C catalysts treated in 1 kPa of O2, CO2,
or H2O at 333 K. Independent kinetic and in situ chemical
titration studies showed that two distinct sitesone of which
has metal-like characteristicsare involved in m-cresol HDO.
Temperature-programmed surface reaction with H2 (TPSR
with H2) was used to quantify the amount of oxygen
incorporated in the catalyst from oxygenate treatment (O/
Mobulk before HDO) and/or from m-cresol HDO (O/Mobulk
after HDO). Molecular oxygen has a higher propensity to
deposit oxygen on Mo2C (O/Mobulk before HDO = 0.23
0.02), compared to CO2 and H2O (O/Mobulk before HDO ≈
0.036), as inferred from TPSR with H2 before HDO. In situ
X-ray photoelectron spectroscopy (XPS) measurements were
performed using an SSX-100 spectrometer (Surface Science
Laboratories, Inc.) with an Al Kα X-ray source operating at 200
W on 1 mm × 1 mm area samples. High-resolution spectra
were collected using 50 eV pass energy and 0.1 eV step−1. The
atomic percentages were calculated from the survey spectrum
(150 eV pass energy and 1 eV step−1) using the ESCA Hawk
software. The lowest energy Mo 3d5/2 peak was used as the
reference (228.0 eV for Mo2C).33,34 A combination of
Gaussian/Lorentzian functions with the Gaussian percentages
being at 80% or higher was used for curve fitting.
2.1.2. Kinetic Studies for Vapor-Phase Hydrodeoxygena-
tion (HDO) of m-Cresol. Steady-state vapor-phase HDO
1114
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
Scheme 1. Experimental Procedures Used in This Study to Investigate the Effect of Oxygenate Pretreatment on Mo2C for
Hydrodeoxygenation (HDO)
reactions of m-cresol on Mo2C catalyst were carried out in a
tubular quartz reactor (ID = 10 mm) placed in a tube furnace
(Applied Test System, Series 3210), controlled by a Watlow
Temperature Controller (96 series). The reaction temperature
was monitored by a thermocouple inserted in the outer
thermowell of the reactor. The reactant gas mixture was
comprised of m-cresol (0.03%)/H2 (84%)/He (balance)
(mol %) at a total flow rate of ∼3.33 cm3 s−1 and a total
pressure of ∼112 kPa. m-Cresol (Sigma, FG, ≥98%) was added
to the flow line, using a syringe pump (KD Scientific, Model
100). Reactor effluents were analyzed with a flame ionization
detection (FID) device, using an online gas chromatography
(GC) system (Agilent, Model 7890) with a methyl-siloxane
capillary column (HP-1, 50 m × 320 μm × 0.52 μm). All
passivated molybdenum carbide samples were treated in H2
(Minneapolis Oxygen, 99.999%; ∼1.67 cm3 s−1) at 773 K (∼0.1
K s−1) for 1 h prior to reaction. All flow lines were heated to at
least 398 K via resistive heating to prevent condensation of the
synthesized following the procedures described in Section
2.1.1, and was subsequently cooled to the oxygenate treatment
temperature (333, 363, or 423 K) in the same CH4/H2 flow.
The freshly synthesized Mo2C was then exposed to a gas
stream containing the oxygenate (0.05−6 kPa)O2, CO2, or
H2O and an inert (Ar or He) at a total gas flow rate of 1.67
cm3 s−1 at ambient pressure for 2 h; the corresponding
oxygenate treatment pressures are summarized in Scheme 1.
Water was added to the flow lines by a syringe pump (KD
Scientific, Model 100). The catalyst was then heated to 423 K
in 0.5 h and held at 423 K for 1.5 h in a flow of helium (10%)/
H2 (balance) (mol %) at ambient pressure and at a total flow
rate of ∼1.83 cm3 s−1 to remove any loosely bound species
from the catalyst. The nomenclature for the oxygenate-
modified Mo2C catalysts reported in this work is of the form,
oxygenate−treatment pressure (temperature)−Mo2C. A freshly
synthesized molybdenum carbide catalyst that was treated with
1 kPa O2 at 333 K, for example, is denoted as O2−1 kPa (333
K)−Mo2C. All oxygenate-modified Mo2C catalysts studied in
this work are listed in Scheme 1.
+
compounds. m-Cresol conversion and C6 product selectivity
were calculated as follows:
2.2.2. m-Cresol HDO on Fresh/Oxygenate-Treated Mo2C.
Vapor-phase HDO was used as a probe reaction to study the
effect of oxygenate modification on Mo2C catalysts. The reactor
setup is the same as that described in section 2.1.2. The
reactant flow in all oxygenate treatment studies was comprised
of m-cresol (1%)/He (10%)/H2 (balance) (mol %), the total
flow rate was 1.83 cm3 s−1 at a total pressure of ∼107 kPa and a
temperature of 423 K.
m‐cresol conversion (%)
(sum of moles of C in products)out
=
× 100
(moles of C in m‐cresol)in
(1)
moles of C+6 product i
moles of C+6 products
C+6 product selectivity (%) =
× 100
(2)
2.2.3. Temperature-Programmed Surface Reaction (TPSR)
with H2. Temperature-programmed surface reaction (TPSR)
with H2 was performed to quantify the amount of oxygen
deposited on a fresh Mo2C catalyst from oxygenate
modification and on a fresh and an oxygenate-modified
Mo2C catalyst from m-cresol HDO at 423 K. The catalyst
was heated from 423 K to 773 K in 1 h and held at 773 K for
0.5 h in a flow of helium (10%)/H2 (balance) (mol %) at a
total flow rate of ∼1.83 cm3 s−1 at ambient pressure. The
oxygen uptake on the catalyst (i) after oxygenate treatment,
denoted as O/Mobulk before HDO, and (ii) after m-cresol
HDO, denoted as O/Mobulk after HDO, was estimated by the
H2O signal monitored by an online mass spectrometer (MKS,
Model Cirrus 200 Quadrupole mass spectrometer system)
during the TPSR process (Figure S1 in the Supporting
2.1.3. In Situ CO Titration for Vapor-Phase Hydro-
deoxygenation (HDO) of m-Cresol. In situ CO titration
studies were conducted to probe the identity and density of the
operational catalytic sites for m-cresol HDO on Mo2C
formulations. A co-feed of CO with Ar or He as an internal
tracer (0.0125, 0.025, or 0.0375 cm3 s−1 CO in 0.033 cm3 s−1
Ar) was introduced to the reactant mixture after steady-state m-
cresol synthesis rates were observed. The transient responses of
CO and Ar/He were monitored by an online mass
spectrometer (MKS, Model Cirrus 200 Quadrupole mass
spectrometer system). Toluene synthesis rates before and after
CO co-feed were quantified by online GC.
2.2. Oxygenate Treatment Studies. 2.2.1. Oxygenate
Treatment on Fresh Mo2C. Fresh Mo2C (∼1 gcat) was
1115
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
Figure 1. (a) Toluene selectivity and (b) m-cresol conversion for m-cresol hydrodeoxygenation (HDO) on Mo2C. Feed = m-cresol (0.04%)/H2
(balance) (mol %) at a total pressure of ∼108 kPa and at a temperature of 423 K; total flow rate ≈ 2.17 cm3 s−1; 0.14 gcat of Mo2C catalyst; ex situ
−1
CO uptake ≈ 94 μmol gcat
.
■
●
Figure 2. (a) Effect of ( ) H2 pressure and ( ) m-cresol pressure on turnover frequencies (TOFs) of toluene synthesis determined by ex situ CO
chemisorption (184−240 μmol gcat−1) on Mo2C catalysts at a total pressure of ∼110 kPa and a temperature of 423 K. H2 pressure was varied from 10
kPa to 110 kPa (balance being helium) at a m-cresol pressure of 0.03 kPa; the m-cresol pressure was varied from 0.03−1.5 at a H2 pressure of 50 kPa
(balance being helium); total flow rate ≈ 3.33 cm3 s−1; ∼0.02−0.06 gcat of Mo2C. (b) Temperature dependencies for TOF of toluene synthesis
determined by ex situ CO chemisorption (236 μmol gcat−1) from m-cresol HDO at a total pressure of ∼112 kPa and a temperature of 400−450 K. m-
Cresol pressure ≈ 0.028 kPa; total flow rate ≈ 3.33 cm3 s−1; ∼0.016 gcat of catalyst.
Information). Water was found to be the major oxygen-
containing compound eluted during TPSR, as also noted by
Lee et al.25 and Choi et al.32 for TPSR with H2 on Mo2C after
anisole HDO at 423 K and TPSR with H2 for the as-prepared
Mo2C catalysts, respectively. Scheme 1 illustrates the
experimental procedures for the oxygenate treatment studies.
An O/Mobulk value was obtained from eq 3.
carbide catalysts have been shown to selectively catalyze
vapor-phase anisole HDO at 423 K and ambient H2 pressure
(>90% C6 selectivity for benzene and cyclohexane).27,35
+
Similar to anisole HDO on Mo2C, high selectivity for toluene
+
(>90% C6 selectivity) was observed from m-cresol HDO on
Mo2C at ambient pressure and at temperatures between 423 K
and 483 K for ∼240 ks time-on-stream (TOS) (Figure 1a). A
decrease of <5% in m-cresol conversion at 423 K was observed
after two temperature cycles (423−483 K) were employed
during the ∼240 ks TOS (Figure 1b), demonstrating that the
catalyst was stable for m-cresol HDO. High selectivity (>75%)
to toluene from vapor-phase m-cresol HDO has also been
reported under ambient pressure on Pd/Fe2O3 catalysts at 573
K,36 Pt/H-BEA at 623 and 673 K,37,38 Pt/SiO2 at 553 K,39 and
10 wt % MoO3/ZrO2 (>99% toluene selectivity)40 at 593 K,
and at ∼5 atm on Pt/TiO2 at 623 K.41 m-Cresol conversion on
these catalysts, however, decreases with TOS.38−40
The turnover frequencies (TOFs) of toluene synthesis
determined by ex situ CO chemisorption on Mo2C are zero-
order-dependent on m-cresol pressure (0.03−1.5 kPa) and are
almost-half-order-dependent on H2 pressure (10−110 kPa)
(Figure 2a). The concurrent zero-order dependence on the m-
cresol pressure and the almost-half-order dependence on H2
pressure for toluene synthesis suggests that two distinct sites,
which catalyze (i) hydrogen dissociation and (ii) m-cresol
moles of oxygen uptake measured from TPSR
O/Mobulk
=
moles of Mo in Mo2C sample
(3)
Independent H2 TPSR experiments were conducted on the
fresh Mo2C, O2−1 kPa (333 K)−Mo2C, H2O−1 kPa (333 K)−
Mo2C, and CO2−1 kPa (333 K)−Mo2C catalysts, in which,
after the oxygenate treatment, the catalyst was heated to 773 K
in 1 h and held at 773 K for 0.5 h, then subsequently heated to
973 K in 0.5 h and held at 973 K for 2 h in a flow of helium
(10%)/H2 (balance) (mol %) at a total flow rate of ∼1.83 cm3
s−1 and at ambient pressure (Figure S1).
3. RESULTS AND DISCUSSION
3.1. Kinetics and In Situ Chemical Titration Studies for
m-Cresol HDO. 3.1.1. Kinetic Studies for Vapor-Phase
Hydrodeoxygenation (HDO) of m-Cresol. Molybdenum
1116
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
Figure 3. (a) Toluene synthesis rates versus time-on-stream (TOS) and (b) normalized transient mass spectrometer signals of toluene (m/z = 91),
Ar (m/z = 40), and CO (m/z = 28), as a function of time during the course of in situ CO titration for m-cresol HDO over ∼4 gcat of Mo2C catalyst
(75 μmol gcat−1 ex situ CO uptake) at a total pressure of ∼120 kPa and a temperature of 423 K. m-Cresol (0.3%)/H2 (balance) (mol %), total flow
rate ≈ 3.33 cm3 s−1, 28% conversion. Co-feed flow rates: 0.0125, 0.025, or 0.0375 cm3 s−1 CO in 0.033 cm3 s−1 Ar.
a
Table 1. O/Mobulk Ratio before and after m-Cresol HDO
b
O2−0.05 kPa (333 K)−Mo2C
O2−1 kPa (333 K)−Mo2C H2O−1 kPa (333 K)−Mo2C CO2−1 kPa (333 K)−Mo2C fresh Mo2C
O/Mobulk before HDO
O/Mobulk after HDO
conversion (%)
0.030 0.005
0.054 0.005
∼18
0.23 0.02
0.23 0.02
∼1.5
0.038 0.010
0.07 0.01
∼18
0.035 0.010
0.06 0.01
∼18
0
0.05 0.01
∼21
toluene synthesis rate
7.9
0.61
8.3
8.7
9.6
(× 10−8 mol gcat s−1
)
−1
toluene TOF
3.8 0.8
21
2.1 0.4
3
3.4 0.5
25
4.5 0.8
19
4.5 0.7
21
(× 10−3 mol molCO s−1
)
−1
metal-like site density
−1
(μmol gcat
)
BET surface area (m2 gcat
)
83
95
3
76
91
5
n/a
96
2
81
96
2
96
95
3
−1
toluene selectivity (%)
c
methylcyclohexanes
selectivity (%)
d
selectivity of others (%)
2
4
2
2
2
a
m-Cresol conversion, toluene synthesis rates, turnover frequencies of toluene synthesis, toluene selectivity, methylcyclohexane selectivity, and the
selectivity of others from m-cresol HDO on fresh Mo2C, O2−1 kPa (333 K)−Mo2C, H2O−1 kPa (333 K)−Mo2C, CO2−1 kPa (333 K)−Mo2C, and
O2−0.05 kPa (333 K)−Mo2C catalysts and their corresponding BET surface area after m-cresol HDO. Feed = m-cresol (1%)/He (10%)/H2
b
(balance) (mol %) at ∼107 kPa total pressure and at 423 K; total flow rate = 1.83 cm3 s−1; 1 gcat of catalyst. O2−0.05 kPa (333 K)−Mo2C treatment
c
d
conditions: 0.083 cm3 s−1 1% O2/He with 1.58 cm3 s−1 Ar. Methylcyclohexanes is the sum of methylcyclohexane and methylcyclohexenes. Others
is the sum of unidentified C6+ hydrocarbons having boiling points higher than 500 K, as inferred from the retention time of the species in the GC
chromatogram, which were quantified using m-cresol to give upper bounds of selectivity and rate.
activation, are required for m-cresol HDO on molybdenum
carbide catalysts in a Langmuir−Hinshelwood-type surface
reaction mechanism. These results are consistent with the
previously reported reaction mechanism for anisole HDO on
bulk Mo2C and W2C catalysts.27,35 The apparent activation
energy for toluene synthesis estimated from the Arrhenius plot
(Figure 2b) between 400 K and 450 K is ∼94 2 kJ mol−1 for
m-cresol HDO on Mo2C.
3.1.2. In Situ CO Titration for Vapor-Phase Hydro-
deoxygenation of m-Cresol. The number of operational sites
during vapor-phase m-cresol HDO was determined by in situ
CO titration and was subsequently used to assess the TOF of
toluene synthesis. A co-feed of pure CO (0.4−1.1 kPa) with Ar
as an internal tracer was introduced (the shaded areas in Figure
3a) after steady-state toluene synthesis rates were observed.
Toluene synthesis rates from m-cresol HDO on Mo2C catalyst
were inhibited in the presence of a CO co-feed and were
recovered when the co-feed was removed, suggesting that CO
is a reversible titrant for m-cresol HDO, similar to anisole HDO
on Mo2C and W2C catalysts.25,27 These results suggest that
metal-like sites42,43 are involved in Ar−OH bond cleavage.
The amount of CO adsorbed during the course of in situ
titration for m-cresol HDO was obtained by integrating the area
circumscribed by the Ar and CO signals in the mass
spectrometric measurement (τCO) (see Figure 3b) and
multiplying it by the corresponding CO flow rate. No CO2
signal was observed in the presence of a CO co-feed, suggesting
that the water−gas-shift reaction occurs at negligible rates
under the reaction conditions investigated. Less than 1% of the
adsorbed CO undergoes hydrogenation reaction to form CH4,
as estimated by multiplying τCO by the observed methane
synthesis rates (0.7−1.2 × 10−9 mol gcat−1 s−1) in the presence
of a CO co-feed. The TOF of toluene synthesis was estimated
by dividing the difference in toluene synthesis rates without and
with a CO co-feed by the amount of CO adsorbed. The TOF of
toluene synthesis from m-cresol HDO on Mo2C is ∼(3.6
0.7) × 10−3 mol molCO s−1 at 423 K, which is of the same
−1
order of magnitude as the TOF of benzene synthesis from
anisole HDO on Mo2C formulations at 423 K (∼1 × 10−3 mol
molCO−1 s−1),25,27 suggesting that the same catalytic sites cleave
Ar−OH and Ar-OCH3 bonds.
3.2. Oxygenate Treatment Studies. 3.2.1. Mo2C Mod-
ification by CO2, H2O, and O2. Edamoto et al.44 showed from
1117
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
surface science studies that O2 can adsorb dissociatively on a C-
or Mo-terminated Mo2C(0001) surface at room temperature,
as evidenced by XPS analysis in which the peaks on the C 1s
and Mo 3d spectra of an O2-treated Mo2C(0001) surface
broadened and shifted to higher binding energies, compared to
those measured on a Mo2C(0001) surface without exposure to
O2. DFT calculations from Shi et al.45 also showed that O2
adsorbs dissociatively and exothermically on low-index surfaces,
including pure or oxygen-covered β-Mo2C(100) and β-
Mo2C(011) surfaces with negative adsorption energies
(approximately −3 eV at 0 K). Lee et al.25 previously prepared
an O2-modified bulk β-Mo2C, using an approach similar to that
reported by Ribeiro et al.46,47 TPSR with H2 on the O2-
modified bulk β-Mo2C catalyst monitored by an online mass
spectrometer showed that oxygen was left behind on the
surface and was removed as H2O by H2.
Density functional theory (DFT) calculations in studying the
mechanism for water−gas-shift on Mo- or C-terminated β-
Mo2C(001) surfaces48 and on a β-Mo2C(001) slab49 suggest
that H2O and CO2 can dissociate on these surfaces. The
reaction energies for O* formation from H2O(g) (H2O(g) → O*
+ H2(g)) on Mo- and C-terminated β-Mo2C(001) surfaces with
0.2−1.2 monolayers of oxygen coverage were found to be −1.5
eV to −5.5 eV and −0.5 eV to −1.5 eV, respectively.48 Porosoff
et al.50 have probed and established CO2 dissociation on a
model Mo2C surface prepared by carburizing an Mo(110)
substrate, using ambient-pressure X-ray photoelectron spec-
troscopy (AP-XPS) and TPSR experiments monitored by mass
spectrometry. A C 1s peak at 283.6 eV assigned to an
oxycarbide (O−Mo−C) and an O 1s peak at 531.7 eV assigned
to an oxygen bonded to carbon were observed on the model
Mo2C surface both when exposed to 150 mTorr CO2 and
under CO2 hydrogenation reaction conditions (150 mTorr
CO2 and 553 mTorr H2 at 523 K),50 suggesting that CO2 is
dissociated to O and CO on a Mo2C/Mo(110) surface.
The amounts of oxygen deposited from oxygenate treatment
(O/Mobulk before HDO) on Mo2C samples at 333 K with 1
kPa of O2, H2O, and CO2 were quantified using TPSR with H2
at 773 K and are summarized in Table 1. The value of O/
Mobulk before HDO is ∼6 times higher on O2-modified Mo2C
(O/Mobulk = 0.23 0.02) than that on H2O (O/Mobulk = 0.038
in the Supporting Information) is observed on a freshly
synthesized Mo2C catalyst and it can only be removed at a
temperature higher than 823 K (Figure S1a), and (ii) the values
of O/Mobulk before HDO quantified from TPSR with H2 at 773
K are comparable to those quantified via TPSR with H2 at 973
K ( 10%) after subtracting the amount of residual oxygen (O/
Mobulk ≈ 0.07) in a freshly synthesized Mo2C catalyst (Table S1
in the Supporting Information). Therefore, TPSR with H2 at
773 K is used to quantify the oxygen loadings from oxygenate
modification or HDO reaction on Mo2C in this work.
3.2.2. m-Cresol HDO on CO2-, H2O-, or O2-Modified Mo2C.
Vapor-phase m-cresol HDO at 423 K was performed on fresh
Mo2C and on Mo2C samples treated with 1 kPa of O2, H2O,
and CO2 at 333 K to investigate the effects of oxygenate
modification on the metal-like properties of Mo2C. Toluene
synthesis rates from m-cresol HDO at 423 K reported in this
work (Table 1) are average values taken between 7.2−14.4 ks
TOS, since the decrease in toluene synthesis rates observed in
this regime is <5% on all the samples (Figure 4). Toluene
●
Figure 4. Toluene synthesis rates from m-cresol HDO over ( ) fresh
◇
△
Mo2C, ( ) H2O−1 kPa (333 K)−Mo2C, ( ) CO2−1 kPa (333 K)−
□
Mo2C, and ( ) O2−1 kPa (333 K)−Mo2C catalysts. Feed = m-cresol
(1%)/He (10%)/H2 (balance) (mol %) at a total pressure of ∼107
kPa and a temperature of 423 K; total flow rate = 1.83 cm3 s−1; 1 gcat of
catalyst.
0.010) and CO2-modified (O/Mobulk = 0.035
0.010)
synthesis rates from m-cresol HDO are similar on fresh Mo2C,
and H2O and CO2-modified Mo2C catalysts ((8.3−9.6) × 10−8
samples, demonstrating that O2 has a stronger propensity to
leave behind adsorbed oxygen on a fresh Mo2C. A temperature
exotherm (ΔT ≈ 20 K) was observed when 1 kPa of O2 was
introduced to the fresh Mo2C at 333 K; however, the
temperature exotherm observed when introducing 1 kPa of
H2O or CO2 to the fresh Mo2C at 333 K was only ∼2 K. These
experimental observations mirror the DFT studies reported by
Liu and Rodriguez,48 in which O2 adsorption energies on Mo-
terminated and C-terminated pure Mo2C(001) surfaces (−8.15
eV and −5.4 eV) were found to be larger than those for H2O
adsorption (−1.52 eV and −1.33 eV) and CO2 adsorption
(−0.88 eV and −0.27 eV) on the same surfaces. Thermody-
namic calculations also show that the oxidation propensity of
O2(g) is stronger than that of H2O(g) for the formation of
MoO2(s) (ΔGrxn, 600 K = −1301 kJ mol−1 vs ΔGrxn, 600 K = −17 kJ
mol−1) or MoO3(s) (ΔGrxn, 600 K = −1530 kJ mol−1 vs ΔGrxn, 600 K
= 183 kJ mol−1) from Mo2C(s).51,52
mol gcat s−1, Table 1), suggesting that the surface environ-
−1
ment of these catalysts is similar. However, the toluene
synthesis rate on O2-modified Mo2C is an order of magnitude
smaller (6.1 × 10−9 mol gcat−1 s−1) than rates measured on fresh
Mo2C, and H2O and CO2-modified Mo2C. The lower toluene
synthesis rate on O2-modified Mo2C, compared to that on the
fresh samples and H2O- and CO2-modified samples could
possibly result from (i) a decrease in the surface area due to the
large temperature exotherm (ΔT ≈ 20 K) during 1 kPa O2
treatment and/or (ii) a change in the number and/or the
identity of the active sites as discussed below and in section
3.2.3.
Brunauer−Emmett−Teller (BET) surface area measure-
ments were performed on the spent fresh Mo2C, O2−1 kPa
(333 K)−Mo2C, and CO2−1 kPa (333 K)−Mo2C catalysts
after m-cresol HDO, which were subsequently treated in 1%
O2/He at ∼5.8 cm3 s−1 for ∼1 ks at RT to passivate Mo2C,
which is pyrophoric. The BET surface area (Table 1) for the
spent O2−1 kPa (333 K)−Mo2C catalyst (76 m2 gcat−1) is only
∼20% lower than that for the spent fresh Mo2C catalyst (96 m2
We confirmed that the majority of the oxygen added to the
fresh Mo2C catalyst from the oxygenate treatment is accounted
for from TPSR with H2 at 773 K based on the observations that
(i) an amount of residual oxygen (O/Mobulk ≈ 0.07; Figure S1a
1118
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
Figure 5. Toluene synthesis rates vs time-on-stream (TOS) for m-cresol HDO on (a) fresh Mo2C, (b) O2−1 kPa (333 K)−Mo2C, (c) H2O−1 kPa
(333 K)−Mo2C, and (d) CO2−1 kPa (333 K)−Mo2C catalysts. Feed = m-cresol (1%)/Ar (10%)/H2 (balance) (mol %) at a total pressure of ∼107
kPa and a temperature of 423 K; total flow rate = 1.83 cm3 s−1; 1 gcat of catalyst. CO co-feed: CO (0.5%−0.93%)/m-cresol (1%)/He (10%)/H2
(balance) (mol %) at a total pressure of ∼107 kPa and a temperature of 423 K; total flow rate = 1.83 cm3 s−1 (CO pressure for panel (b) is 0.05−
0.07 kPa).
gcat−1), which cannot account for the observed 10-fold decrease
in toluene synthesis rate on O2−1 kPa (333 K)−Mo2C,
compared to that on the fresh Mo2C catalyst. Therefore, the
change in surface area is not the major cause for the decrease in
toluene synthesis rates on the O2−1 kPa (333 K)−Mo2C
catalyst.
Oxygen uptake on the catalyst after m-cresol HDO was
quantified subsequently by TPSR with H2, and the values of O/
Mobulk after HDO are reported in Table 1. Oxygen
HDO (Table 1, O/Mobulk ≈ 0.06 0.01)and (ii) toluene
synthesis rate is ∼10-fold lower on O2−1 kPa (333 K)−Mo2C
(Table 1), and the value of O/Mobulk after HDO (∼0.23
0.02) on this sample is four times higher than that on the other
three samples.
3.2.3. The Roles of Adsorbed Oxygen on CO2-, H2O-, or O2-
Modified Mo2C for m-Cresol HDO. In situ CO titration
experiments were conducted to probe the identity and the
density of active sites responsible for m-cresol HDO on fresh
Mo2C and on Mo2C samples treated with 1 kPa of O2, H2O,
and CO2 at 333 K. TOFs of toluene synthesis on fresh and
H2O and CO2-modified Mo2C catalysts are ∼(3.4−4.5) × 10−3
mol molCO−1 s−1 at 423 K (see Table 1, Figure 5, and Figure S3
in the Supporting Information) and are comparable to the TOF
of toluene synthesis measured on a Mo2C formulation that had
been passivated by 1% O2/He but pretreated in H2 prior to
accumulation was observed on the fresh Mo2C (O/Mobulk
≈
0.05 0.01) from m-cresol HDO at 423 K. A similar amount of
oxygen accumulation (O/Mobulk ≈ 0.045) from anisole HDO
on a passivated Mo2C that was treated in H2 at 723 K for 1 h
prior to the reaction study at ambient pressure and 423 K was
reported by Lee et al.,25 using TPSR with H2. The values of O/
Mobulk after HDO on H2O−1 kPa (333 K)−Mo2C and CO2−1
reaction studies (∼(3.6 0.7) × 10−3 mol molCO s−1 at 423
−1
kPa (333 K)−Mo2C catalysts are similar (Table 1, O/Mobulk
≈
0.06 0.01) but higher than the values of O/Mobulk before
K), as discussed in Section 3.1.2. The TOF of toluene synthesis
on the O2-modified Mo2C (2.1 × 10−3 mol molCO−1 s−1) is ∼2-
fold lower, compared to that on H2O- and CO2-modified
samples, which results from (i) the lower toluene synthesis rate
HDO on the same catalysts (Table 1, O/Mobulk ≈ 0.038 and
0.035), demonstrating that additional oxygen was deposited on
these catalysts from m-cresol HDO at 423 K. The values of O/
Mobulk after HDO and before HDO on O2−1 kPa (333 K)−
on O2−1 kPa (333 K)−Mo2C (Table 1, 6.1 × 10−9 mol gcat
−1
Mo2C were found to be similar (O/Mobulk ≈ 0.23
0.02),
s−1) and (ii) the narrow operating window for CO co-feed
pressure (0.05−0.07 kPa) required to concurrently achieve an
observable CO signal and a measurable CO residence time
during in situ CO titration on the mass spectrometer (Figure
S3b in the Supporting Information). Therefore, we conclude
that, within our ability to measure experimentally, TOFs of
toluene synthesis are similar across the fresh Mo2C, and O2-,
H2O-, and CO2-modified Mo2C catalysts, demonstrating that
the incorporation of oxygen in Mo2C does not alter the identity
of the metal-like sites for m-cresol HDO, similar to the
suggesting that the oxygen left behind from the treatment in 1
kPa of O2, which is in excess of that deposited with 1 kPa of
H2O or CO2 at 333 K, or from HDO conditions with 1 kPa of
m-cresol at 423 K, is not removed during HDO. These results
also show that the value of O/Mobulk after HDO correlates with
m-cresol HDO rates on oxygenate-modified Mo2C catalysts as
(i) toluene synthesis rates are the same on the fresh Mo2C,
H2O−1 kPa (333 K)−Mo2C, and CO2−1 kPa (333 K)−Mo2C
catalysts−samples with comparable values of O/Mobulk after
1119
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
Table 2. O/Mobulk Ratio before and after m-Cresol HDO and Toluene Synthesis Rate from m-Cresol HDO on O2-Modified
a
Mo2C Catalysts
b
c
d
O2−0.1 kPa (333 K)−Mo2C O2−0.25 kPa (333 K)−Mo2C O2−0.4 kPa (333 K)−Mo2C O2−1 kPa (333 K)−Mo2C
O/Mobulk before HDO
O/Mobulk after HDO
0.060 0.003
0.061 0.010
6.7
0.11 0.01
0.10 0.01
3.0
0.14 0.02
0.13 0.02
0.8
0.23 0.02
0.23 0.02
0.6
toluene synthesis r−a1te
(× 10−8 mol gcat s−1
)
a
m-Cresol HDO conditions: Feed = m-cresol (1%)/Ar (10%)/H2 (balance) (mol %) at a total pressure of ∼107 kPa and a temperature of 423 K;
b
c
total flow rate = 1.83 cm3 s−1; 1 gcat of catalyst. O2−0.1 kPa (333 K)−Mo2C: 0.167 cm3 s−1 1% O2/He in 1.5 cm3 s−1 Ar. O2−0.25 kPa (333 K)−
d
Mo2C: 0.416 cm3 s−1 1% O2/He in 1.25 cm3 s−1 Ar. O2−0.4 kPa (333 K)−Mo2C: 0.67 cm3 s−1 1% O2/He in 1 cm3 s−1 Ar.
observation reported for anisole HDO on an O2-treated Mo2C
catalyst.25
Ribeiro et al.,28 Iglesia et al.,29 and Boudart and co-workers30
showed that the bifunctionality of tungsten carbide formula-
tions can be adjusted by varying the oxygen loadings on O2-
modified catalysts prior to reaction studies, and Sullivan et al.26
reported that the catalytic function of Mo2C can be adjusted to
be primarily Brønsted acidic when O2 is co-fed to the system
during IPA dehydration in the absence of H2. Approximately
The density of the metal-like sites responsible for m-cresol
HDO was estimated by dividing the toluene synthesis rate by
the TOF of toluene synthesis (Table 1). The metal-like site
densities obtained from in situ CO titration for m-cresol HDO
are similar on fresh Mo2C, and H2O- and CO2-modified Mo2C
catalysts (∼19−25 μmol gcat−1), and are ∼7-fold lower on O2-
modified Mo2C (∼3 μmol gcat−1), which presumably results
from the metal-like sites being poisoned by the additional
+
2% selectivity for unidentified C6 hydrocarbons (denoted as
“others” in Table 1 and Table S3 in the Supporting
Information), which have boiling points of >500 K, as inferred
from the retention time of the species in the GC chromato-
gram, was observed on the fresh Mo2C, H2O−1 kPa (333 K)−
Mo2C, CO2−1 kPa (333 K)−Mo2C, and O2−0.05 kPa (333
amount of oxygen adsorbed (>0.06
0.01 O/Mobulk after
HDO) on the O2−1 kPa (333 K)−Mo2C catalyst (Table 1),
compared to that on fresh and CO2- and H2O-modified Mo2C.
Oxygen adsorbed in amounts exceeding O/Mobulk ≈ 0.06
0.01 on O2−1 kPa (333 K)−Mo2C was found to inhibit HDO
rate by poisoning the metal-like sites as discussed above. The
bulk structure of the spent and passivated fresh Mo2C and O2−
1 kPa (333 K)−Mo2C catalysts remained as β-Mo2C and no
MoO2 or MoO3 peaks were observed in the XRD patterns of
spent catalysts (Figure S2 in the Supporting Information),
indicating that the decrease in HDO rate did not result from
bulk oxidation of Mo2C. The percentage of surface Mo in Mo5+
and/or Mo6+ (231.5 eV) states measured from XPS on the
spent and passivated O2−1 kPa (333 K)−Mo2C (26.7%) was
found to be higher than that on the spent and passivated fresh
Mo2C (15.5%) (Table S2 in the Supporting Information),
suggesting that (i) the decrease in the number of metal-like
sites on O2−1 kPa (333 K)−Mo2C, compared to that on fresh
Mo2C (Table 1), could result from the surface Mo being
oxidized to higher oxidation states, and (ii) the higher amount
of oxygen deposited from 1 kPa of O2 correlates with a higher
fraction of surface Mo being in the Mo5+ and Mo6+ oxidation
states.
+
K)−Mo2C catalysts. The formation of unidentified C6
hydrocarbons during m-cresol HDO mirrors the observation
by Lee et al.11 for the formation of C10 products (∼30%
+
selectivity) in furfural HDO on Mo2C formulations at 423 K,
presumably because of the presence of acid sites on
molybdenum carbides.1 A higher selectivity for the unidentified
C6+ hydrocarbons (4%, Table 1) and a lower toluene selectivity
(∼91%, Table 1) are observed on the 1 kPa O2-modified
Mo2C, compared to the other four samples. However, the
+
synthesis rates per gram of catalyst of both the unidentified C6
hydrocarbons and toluene on O2−1 kPa (333 K)−Mo2C are
∼6-fold and ∼10-fold lower, compared to that on the other
four catalysts, respectively (Table S3 and Table 1). These
+
results suggest that the formation of the unidentified C6
hydrocarbons in m-cresol HDO cannot be selectively adjusted
by the oxygen loading on Mo2C and suggest that bifunction-
ality, metal and acid, is likely required for the formation of the
+
unidentified C6 hydrocarbons.
3.2.4. Changing O/Mobulk Ratios and HDO Rates by
Varying Oxygenate Treatment Conditions. Oxygen treatment
pressures varied between 0.1 and 1 kPa were used to investigate
the consequences of oxygen loadings between O/Mobulk = 0.06
and O/Mobulk = 0.23 on toluene synthesis rates from m-cresol
HDO on molybdenum carbides. A 2-fold increase in the value
of O/Mobulk after HDO (from ∼0.06 to 0.13) results in an 8-
An O2−0.05 kPa (333 K)−Mo2C catalyst (O/Mobulk before
HDO ≈ 0.03) was prepared to investigate the effect of oxygen
source on adsorbed oxygen as H2O−1 kPa (333 K)−Mo2C and
CO2−1 kPa (333 K)−Mo2C catalysts have similar values of O/
Mobulk before HDO (∼0.038 and ∼0.035, respectively; see
Table 1) and exhibit similar toluene synthesis rates, TOF of
toluene synthesis, and O/Mobulk after HDO (Table 1). Toluene
synthesis rates (7.9 × 10−8 mol gcat−1 s−1) and TOF of toluene
fold decrease in the toluene rate (from ∼6.7 × 10−8 mol gcat
−1
s−1 to 8.2 × 10−9 mol gcat−1 s−1) (Table 2). The drop in toluene
rate decreases as the value of O/Mobulk after HDO increased
from ∼0.13 to ∼0.23, in which the toluene synthesis rate on
O2−1 kPa (333 K)−Mo2C (O/Mobulk after HDO ≈ 0.23) is
∼70% of that on O2−0.4 kPa (333 K)−Mo2C (O/Mobulk after
HDO ≈ 0.13) (Table 2). These results suggest that (i) the
adsorbed oxygen does not poison the metal-like sites
selectivelysimilar to what has been reported by Shi et al.,31
using computational calculations that O2 adsorption can occur
preferentially on pure Mo2C surfaces, as inferred from the
reaction energy of surface oxidation by O2 at 0 K on
Mo2C(101) being ∼2 eV more negative than that on
synthesis (3.8 0.8 × 10−3 mol molCO s−1) from m-cresol
−1
HDO at 423 K and the value of O/Mobulk after HDO (0.054
0.005) on O2−0.05 kPa (333 K)−Mo2C catalyst were found to
be similar to that measured on H2O−1 kPa (333 K)−Mo2C,
CO2−1 kPa (333 K)−Mo2C, and fresh Mo2C catalysts (Table
1), clearly demonstrating that the effect of adsorbed oxygen on
Mo2C is independent of the oxygen source under the reaction
conditions investigated.53
Schaidle et al.1 reported the existence of surface acidity on a
Mo2C pretreated in H2 from XPS and NH3-TPD studies.
1120
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
Mo2C(011)and/or (ii) adsorbed oxygen from O2 can be
incorporated into the bulk structure of Mo2C at higher values
of the O/Mobulk ratio.13,54
oxygen, as inferred from in situ CO titration and m-cresol HDO
reactions on O2−0.05 kPa (333 K)−Mo2C, H2O−1 kPa (333
K)−Mo2C, and CO2−1 kPa (333 K)−Mo2C catalysts.
We showed that O2 can deposit ∼0.03 O/Mobulk (Table 1)
of oxygen on a fresh Mo2C at a treatment pressure of 0.05 kPa,
whereas 1 kPa of H2O, CO2, and m-cresol is required to achieve
the same value (Table 1). We increased H2O and CO2
treatment pressures (5−6 kPa) and temperatures (333, 363,
and 423 K) to investigate the amount of oxygen H2O and CO2
can deposit under ambient pressure. The values of O/Mobulk
before HDO on H2O−5 kPa (423 K)−Mo2C, CO2−5 kPa
(423 K)−Mo2C, CO2−6 kPa (363 K)−Mo2C, and CO2−6 kPa
(333 K)−Mo2C catalysts were found to be similar (Table 3, O/
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acscatal.6b02762.
■
S
Additional results for TPR, in situ CO titration, and m-
cresol HDO studies; XRD patterns and XPS data for
Mo2C formulations (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: abhan@umn.edu (A. Bhan).
ORCID
Aditya Bhan: 0000-0002-6069-7626
Notes
The authors declare no competing financial interest.
■
Table 3. O/Mobulk before m-Cresol HDO on H2O-Modified
and CO2-Modified Mo2C Catalysts
O/Mobulk before HDO
H2O−5 kPa (423 K)−Mo2C
CO2−5 kPa (423 K)−Mo2C
CO2−6 kPa (363 K)−Mo2C
CO2−6 kPa (333 K)−Mo2C
CO2−1 kPa (333 K)−Mo2C
0.050 0.004
0.046 0.005
0.050 0.002
0.053 0.003
0.035 0.010
ACKNOWLEDGMENTS
■
This research was supported by National Science Foundation
Catalysis and Biocatalysis Program (CBET Award No.
1510661). Parts of this work were carried out in the
Characterization Facility at the University of Minnesota,
which receives partial support from NSF through the
MRSEC program. We thank Ms. Seema Thakral for assistance
with the XRD analysis and Bing Luo for XPS measurements.
Mobulk ≈ 0.046−0.053) and comparable to the amount of
oxygen deposited from m-cresol HDO at 423 K (Table 1,
∼0.05
0.01 O/Mobulk). These results demonstrate that
oxygen adsorbed in amounts exceeding O/Mobulk ≈ 0.05
0.01 can be achieved by varying the O2 treatment pressure at a
treatment temperature of 333 K; however, oxygenates such as
CO2, H2O, and m-cresol can only deposit O/Mobulk ≈ 0.05
0.01, even at a higher treatment pressure and temperature (∼6
kPa and 423 K, respectively).
REFERENCES
■
(1) Schaidle, J. A.; Blackburn, J.; Farberow, C. A.; Nash, C.; Steirer,
K. X.; Clark, J.; Robichaud, D. J.; Ruddy, D. A. ACS Catal. 2016, 6,
1181−1197.
4. CONCLUSIONS
(2) Bej, S. K.; Thompson, L. T. Appl. Catal., A 2004, 264, 141−150.
(3) Xiong, K.; Yu, W.; Vlachos, D. G.; Chen, J. G. ChemCatChem
2015, 7, 1402−1421.
Kinetic and in situ chemical titration studies on vapor-phase m-
cresol hydrodeoxygenation (HDO) showed that two distinct
sitesone of them having metal-like site characteristicsare
involved in toluene synthesis. m-Cresol HDO was used as a
probe reaction to study the effect of oxygenate modification on
the metal-like function of Mo2C. The amount of adsorbed
oxygen (O/Mobulk) on O2−1 kPa (333 K)−Mo2C, H2O−1 kPa
(333 K)−Mo2C, and CO2−1 kPa (333 K)−Mo2C catalysts,
prepared by pretreating fresh Mo2C catalysts in 1 kPa of O2,
CO2, and H2O at 333 K, was quantified using temperature-
programmed surface reaction with H2 (TPSR). The value of O/
Mobulk before HDO on O2−1 kPa (333 K)−Mo2C (0.23
0.02) is ∼6 times higher than that on H2O−1 kPa (333 K)−
Mo2C and CO2−1 kPa (333 K)−Mo2C (O/Mobulk before
HDO ≈ 0.036), demonstrating that molecular oxygen has a
higher propensity to deposit oxygen on a fresh Mo2C sample.
The value of O/Mobulk after HDO was found to correlate with
toluene synthesis rates, suggesting that the relevant surface
environment for catalysis is that measured under reaction
conditions. A 10-fold decrease in toluene synthesis rates was
observed on O2−1 kPa (333 K)−Mo2C, compared to that on
fresh Mo2C, H2O−1 kPa (333 K)−Mo2C, and CO2−1 kPa
(333 K)−Mo2C catalysts; however, turnover frequencies
(TOFs) of toluene synthesis measured from in situ CO
titration on these samples are similar (∼(2.1−4.5) × 10−3 mol
(4) Han, J.; Duan, J.; Chen, P.; Lou, H.; Zheng, X.; Hong, H.
ChemSusChem 2012, 5, 727−733.
(5) Gosselink, R. W.; Stellwagen, D. R.; Bitter, J. H. Angew. Chem.,
Int. Ed. 2013, 52, 5089−5092.
(6) Jongerius, A. L.; Bruijnincx, P. C. A.; Weckhuysen, B. M. Green
Chem. 2013, 15, 3049−3056.
́
(7) Prasomsri, T.; Shetty, M.; Murugappan, K.; Roman-Leshkov, Y.
Energy Environ. Sci. 2014, 7, 2660−2669.
(8) Boullosa-Eiras, S.; Lødeng, R.; Bergem, H.; Stocker, M.;
̈
Hannevold, L.; Blekkan, E. A. Catal. Today 2014, 223, 44−53.
(9) Ren, H.; Yu, W.; Salciccioli, M.; Chen, Y.; Huang, Y.; Xiong, K.;
Vlachos, D. G.; Chen, J. G. ChemSusChem 2013, 6, 798−801.
(10) Chen, C.-J.; Lee, W.-S.; Bhan, A. Appl. Catal., A 2016, 510, 42−
48.
(11) Lee, W.-S.; Wang, Z.; Zheng, W.; Vlachos, D. G.; Bhan, A. Catal.
Sci. Technol. 2014, 4, 2340−2352.
(12) Wang, T.; Luo, Q.; Li, Y.-W. W.; Wang, J.; Beller, M.; Jiao, H.
Appl. Catal., A 2014, 478, 146−156.
(13) Ranhotra, G. S.; Haddix, G. W.; Bell, A. T.; Reimer, J. A. J. Catal.
1987, 108, 24−39.
(14) Ramírez-Caballero, G. E.; Burgos, J. C.; Balbuena, P. B. J. Phys.
Chem. C 2009, 113, 15658−15666.
(15) Gom
20, 215601.
́ ́
ez-Gualdron, D. A.; Balbuena, P. B. Nanotechnology 2009,
(16) Ko, E. I.; Madix, R. J. Surf. Sci. 1981, 109, 221−238.
(17) Stottlemyer, A. L.; Kelly, T. G.; Meng, Q.; Chen, J. G. Surf. Sci.
Rep. 2012, 67, 201−232.
(18) Wang, T.; Li, Y. W.; Wang, J. G.; Beller, M.; Jiao, H. J. J. Phys.
Chem. C 2014, 118, 3162−3171.
−1
molCO s−1), indicating that adsorbed oxygen poisons the
metal-like sites responsible for m-cresol. The effect of adsorbed
oxygen on toluene synthesis is independent of the source of
1121
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122

ACS Catalysis
Research Article
(19) Wang, T.; Li, Y.-W.; Wang, J.; Beller, M.; Jiao, H. J. Phys. Chem.
C 2014, 118, 8079−8089.
(20) Leary, K. J.; Michaels, J. N.; Stacy, M. J. Catal. 1986, 101, 301−
313.
(54) Sajkowski, D. J.; Oyama, S. T. Appl. Catal., A 1996, 134, 339−
349.
(21) Lee, J. S.; Oyama, S. T.; Boudart, M. J. Catal. 1987, 106, 125−
133.
(22) Oyama, S. T. The Chemistry of Transition Metal Carbides and
Nitrides; Blackie Academic: London, 1996.
(23) Ranhotra, G. S.; Bell, A. T.; Reimer, J. A. J. Catal. 1987, 108,
40−49.
(24) Schaidle, J. A.; Lausche, A. C.; Thompson, L. T. J. Catal. 2010,
272, 235−245.
(25) Lee, W.-S.; Kumar, A.; Wang, Z.; Bhan, A. ACS Catal. 2015, 5,
4104−4114.
(26) Sullivan, M. M.; Held, J. T.; Bhan, A. J. Catal. 2015, 326, 82−91.
(27) Lu, Q.; Chen, C.-J.; Luc, W. W.; Chen, J. G.; Bhan, A.; Jiao, F.
ACS Catal. 2016, 6, 3506−3514.
(28) Ribeiro, F. H.; Boudart, M.; Dalla Betta, R. A.; Iglesia, E. J. Catal.
1991, 130, 498−513.
(29) Iglesia, E.; Baumgartner, J. E.; Ribeiro, F. H.; Boudart, M. J.
Catal. 1991, 131, 523−544.
(30) Ribeiro, F. H.; Dalla Betta, R. A.; Guskey, G. J.; Boudart, M.
Chem. Mater. 1991, 3, 805−812.
(31) Shi, X.-R.; Wang, S.-G.; Hu, J.; Qin, Z.; Wang, J. Surf. Sci. 2012,
606, 1187−1194.
́
(32) Choi, J.-S.; Bugli, G.; Djega-Mariadassou, G. J. Catal. 2000, 193,
238−247.
(33) St. Clair, T. P.; Oyama, S. T.; Cox, D. F.; Otani, S.; Ishizawa, Y.;
Lo, R.-L.; Fukui, K.; Iwasawa, Y. Surf. Sci. 1999, 426, 187−198.
́
́
(34) Ovari, L.; Kiss, J.; Farkas, A. P.; Solymosi, F. J. Phys. Chem. B
2005, 109, 4638−4645.
(35) Lee, W.-S.; Wang, Z.; Wu, R. J.; Bhan, A. J. Catal. 2014, 319,
44−53.
(36) Hong, Y.; Zhang, H.; Sun, J.; Ayman, K. M.; Hensley, A. J. R.;
Gu, M.; Engelhard, M. H.; McEwen, J. S.; Wang, Y. ACS Catal. 2014,
4, 3335−3345.
(37) Sun, Q.; Chen, G.; Wang, H.; Liu, X.; Han, J.; Ge, Q.; Zhu, X.
ChemCatChem 2016, 8, 551−561.
(38) Zhu, X.; Nie, L.; Lobban, L. L.; Mallinson, R. G.; Resasco, D. E.
Energy Fuels 2014, 28, 4104−4111.
(39) Foster, A.; Do, P. M.; Lobo, R. Top. Catal. 2012, 55, 118−128.
(40) Shetty, M.; Murugappan, K.; Prasomsri, T.; Green, W. H.;
́
Roman-Leshkov, Y. J. Catal. 2015, 331, 86−97.
(41) Griffin, M. B.; Ferguson, G. A.; Ruddy, D. A.; Biddy, M. J.;
Beckham, G. T.; Schaidle, J. A. ACS Catal. 2016, 6, 2715−2727.
(42) Hammer, B.; Morikawa, Y.; Nørskov, J. Phys. Rev. Lett. 1996, 76,
2141−2144.
(43) Rodriguez, J. A.; Goodman, D. W. Science 1992, 257, 897−903.
(44) Edamoto, K.; Sugihara, M.; Ozawa, K.; Otani, S. Surf. Sci. 2004,
561, 101−109.
(45) Shi, X.-R.; Wang, S.-G.; Wang, J. J. Mol. Catal. A: Chem. 2016,
417, 53−63.
(46) Iglesia, E.; Ribeiro, F. H.; Boudart, M.; Baumgartner, J. E. Catal.
Today 1992, 15, 307−337.
(47) Ribeiro, F. H.; Dalla Betta, R. A.; Boudart, M.; Baumgartner, J.;
Iglesia, E. J. Catal. 1991, 130, 86−105.
(48) Liu, P.; Rodriguez, J. A. J. Phys. Chem. B 2006, 110, 19418−
19425.
(49) Tominaga, H.; Nagai, M. J. Phys. Chem. B 2005, 109, 20415−
20423.
(50) Porosoff, M. D.; Yang, X.; Boscoboinik, J. A.; Chen, J. G. Angew.
Chem., Int. Ed. 2014, 53, 6705−6709.
(51) Barin, I. Thermochemical Data of Pure Substances; VCH:
Weinheim, Germany, 1989.
(52) Barin, I.; Knacke, O. Thermochemical Properties of Inorganic
Substances; Springer−Verlag: New York, 1973.
(53) Sullivan, M. M.; Chen, C.-J.; Bhan, A. Catal. Sci. Technol. 2016,
6, 602−616.
1122
DOI: 10.1021/acscatal.6b02762
ACS Catal. 2017, 7, 1113−1122