Phosphonic acid modifiers for enhancing selective hydrodeoxygenation over Pt catalysts: The role of the catalyst support

Article abstract of DOI:10.1016/j.jcat.2019.03.011

Bifunctional catalysts comprised of both noble metal and Br?nsted acid sites are of growing interest for many reactions, such as the hydrodeoxygenation (HDO) of oxygenates produced by the deconstruction of lignocellulosic biomass. One method of preparing a bifunctional metal-acid catalyst is to modify the supporting material of a metal catalyst with acid-containing ligands, such as phosphonic acids (PAs), which provide tunable Br?nsted acid sites at the metal-support interface. Here, we explore the potential for PA modification to improve HDO rates on Al₂O₃, TiO₂, CeO₂, and SiO₂-Al₂O₃ supports. PAs containing either alkyl or carboxylic acid (CA) tails were used to modify Pt catalysts on each support. PA modification improved HDO rates for the model compound benzyl alcohol when applied to Pt supported by Al₂O₃, TiO₂, and CeO₂, but had a negative effect on the HDO rate of Pt/SiO₂-Al₂O₃. Measurements of the relative strength of Br?nsted acid sites present on the modified and unmodified catalysts suggested that PAs improved HDO when they provided new or stronger Br?nsted acid sites. Additionally, the strength of the Br?nsted acid sites provided by PA modifiers were tunable by altering the tail functionality, which affected the rates of HDO. These results suggest that PAs can be used to modify a variety of supports to prepare bifunctional acid-metal catalysts.

Full text of DOI:10.1016/j.jcat.2019.03.011

Journal of Catalysis 372 (2019) 311–320
Contents lists available at ScienceDirect
Journal of Catalysis
journal homepage: www.elsevier.com/locate/jcat
Phosphonic acid modifiers for enhancing selective hydrodeoxygenation
over Pt catalysts: The role of the catalyst support
Patrick D. Coan a, Michael B. Griffin b, Peter N. Ciesielski c, J. Will Medlin a,
⇑
a Department of Chemical and Biological Engineering, University of Colorado – Boulder, Boulder, CO 80309, United States
b National Bioenergy Center, National Renewable Energy Laboratory, Golden, CO 80401, United States
c Biosciences Center, National Renewable Energy Laboratory, Golden, CO 80401, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Bifunctional catalysts comprised of both noble metal and Brønsted acid sites are of growing interest for
many reactions, such as the hydrodeoxygenation (HDO) of oxygenates produced by the deconstruction of
lignocellulosic biomass. One method of preparing a bifunctional metal-acid catalyst is to modify the sup-
porting material of a metal catalyst with acid-containing ligands, such as phosphonic acids (PAs), which
provide tunable Brønsted acid sites at the metal-support interface. Here, we explore the potential for PA
modification to improve HDO rates on Al2O3, TiO2, CeO2, and SiO2-Al2O3 supports. PAs containing either
alkyl or carboxylic acid (CA) tails were used to modify Pt catalysts on each support. PA modification
improved HDO rates for the model compound benzyl alcohol when applied to Pt supported by Al2O3,
TiO2, and CeO2, but had a negative effect on the HDO rate of Pt/SiO2-Al2O3. Measurements of the relative
strength of Brønsted acid sites present on the modified and unmodified catalysts suggested that PAs
improved HDO when they provided new or stronger Brønsted acid sites. Additionally, the strength of
the Brønsted acid sites provided by PA modifiers were tunable by altering the tail functionality, which
affected the rates of HDO. These results suggest that PAs can be used to modify a variety of supports
to prepare bifunctional acid-metal catalysts.
Received 23 December 2018
Revised 22 February 2019
Accepted 10 March 2019
Keywords:
Platinum
Hydrodeoxygenation
Phosphonic acids
Self-assembled monolayers
Ó 2019 Elsevier Inc. All rights reserved.
1. Introduction
tion of aromatic alcohols over Pt catalysts is highly competitive
with dehydrogenation/decarbonylation steps and can be coverage
The development and understanding of multifunctional cata-
lysts, which contain two or more distinct reactive sites, is an
increasingly important area of heterogeneous catalysis research
[1–4]. By incorporating different functionalities onto a single cata-
lyst, complex reactions can be performed. These multifunctional
catalysts can be used to accomplish cascade reactions over a single
substrate or to improve the rate of a desirable reaction via cooper-
ative interactions of the individual active sites [5–8]. One specific
application for bifunctional heterogeneous catalysts is the upgrad-
ing of pyrolysis oil, produced from lignocellulosic biomass [9].
Pyrolysis oil is made up of thousands of components, the majority
of which have a high oxygen content relative to traditional crude
oil. Thus, a hydrodeoxygenation (HDO) step is required to upgrade
pyrolysis oil to a mixture that is chemically similar to petroleum
and compatible with existing refining technology [10]. Benzyl alco-
hol has been used as a probe compound for the lignin portion of
pyrolysis oil [11,12]. Previous work has shown that the deoxygena-
dependent [13]. These competitive pathways for benzyl alcohol
are shown in Fig. 1, where HDO of benzyl alcohol results in the for-
mation of the desirable product toluene, while dehydrogenation
and subsequent decarbonylation steps lead to the formation of
benzaldehyde and benzene, respectively. Although oxygen can be
removed during decarbonylation steps, direct CAO bond scission
is preferred because it maximizes the carbon yields of bio-oil
feedstocks.
Several studies have shown that efficient HDO of biomass-
derived oxygenates occurs through a combination of hydrogena-
tion and dehydration reactions, requiring a bifunctional catalyst
that contains active sites for each step [10,14–16]. Noble metals
(e.g. Pt, Pd) are known to be efficient hydrogenation catalysts as
they readily activate H2, while Brønsted acid sites are known to
perform dehydration reactions [17,18]. Thus, an effective HDO cat-
alyst may be prepared by combining noble metal and Brønsted acid
sites. One method for preparing such a bifunctional catalyst is to
modify the support with acid-containing ligands [19]. Previous
work has shown that self-assembled monolayers (SAMs) deposited
from phosphonic acid (PA) precursors can be applied to both
⇑
Corresponding author.
E-mail address: medlin@colorado.edu (J.W. Medlin).
https://doi.org/10.1016/j.jcat.2019.03.011
0021-9517/Ó 2019 Elsevier Inc. All rights reserved.

312
P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
behavior on diverse supported systems. Our previous work showed
that the strength of Brønsted acid sites incorporated by PA
modifiers is dependent on the tail functionality, with short-chain
CA tails providing stronger acid sites than alkyl tails [20].
2-Phosphonoethanoic acid (2PEA) was chosen as the phosphono-
carboxylic acid (PCA) because it was found to provide the strongest
Brønsted acid sites and the greatest enhancements in DO rates in
our previous study [20]. Propylphosphonic acid (PPA) was chosen
for comparison because it has a simple alkyl tail that is similar in
size to 2PEA. The four supports (Al2O3, TiO2, CeO2, SiO2-Al2O3) were
chosen so as to allow us to examine the effects of using reducible
versus irreducible supports and the role of support acidity on the
potential for PAs to improve HDO rates.
2. Experimental methods
2.1. Materials
The
c-Al2O3 (3 mm powder, 99.997% metals basis) and CeO2
(15–30 nm APS powder, >99.5%) supports were purchased from
Alfa Aesar. The TiO2 support (P25, >99.5%) was purchased from
Acros Organics. The SiO2-Al2O3 support (GRACE DavicatÒ SIAL
3115, ꢀ13 wt% Al2O3, ꢀSi:Al = 5.7) was provided by W.R. Grace.
2-Phosphonoethanoic acid (98%), propylphosphonic acid (95%),
benzyl alcohol (99.8%), and pyridine (ꢁ99%) were purchased from
Sigma Aldrich. HPLC-grade tetrahydrofuran (>99.9%) was pur-
chased from OmniSolv. Chloroplatinic acid hexahydrate (H2PtCl6-
Á6H2O, 99.9%) was purchased from Strem Chemicals. All gases
(Ultra-high purity H2, He, and Ar, 30% N2/He, and 10% CO/He) were
obtained from Airgas.
Fig. 1. (a) Reaction mechanism of benzyl alcohol over Pt catalysts, showing
competitive hydrodeoxygenation and decarbonylation pathways (adapted from Ref.
[21]) (b) Schematic representation of PA modifiers bound to supported metal
catalysts, and structures of the modifiers propylphosphonic acid (PPA) and 2-
phosphonopropionic acid (2PEA).
2.2. Catalyst preparation and characterization
metal-oxide catalysts and the metal-oxide support of supported
metal catalysts to influence reactivity [20–23]. PAs are known to
bind to a variety of metal-oxides, forming up to three direct P-O-
M (M = metal) bonds, and can provide Brønsted acid sites (P-OH)
at the surface [24–30]. For supported noble metals, PAs have been
used to modify Al2O3-supported Pt and Pd catalysts and improve
HDO in both the gas and liquid phases by incorporating Brønsted
acidity at the metal-support interface [20–22]. It has been shown
that the structure of the PA precursor can be adjusted to optimize
the acid strength and HDO performance through inductive effects
and intramolecular cooperativity [20,21]. As an additional benefit,
PAs have also been found to improve the stability of metal-oxide
supports in the harsh and acidic environments often encountered
in pyrolysis oil upgrading [31]. Previous work has also analyzed
the thermal stability of PA modifiers on a variety of metal-oxide
surfaces, and found that the modifiers remain intact to tempera-
tures at or above 400 °C in inert environments [20,21,31,32].
It is valuable to understand how changing the support can affect
the enhancements observed by PA modification. Because PAs may
interact differently with different supports [24–26], it is not known
whether PAs will generally retain Brønsted acidity. As shown in
Fig. 1, monodentate and possibly bidentate binding modes provide
Brønsted acid sites on the catalyst support, while tridentate bind-
ing does not provide Brønsted acidity, but may provide surface
crowding at the metal-support interface. It is also unclear whether
the deposition of acid-containing PAs will improve HDO activity
and selectivity relative to unmodified catalysts regardless of the
support. In this study, we prepared four different supported Pt cat-
alysts and deposited PAs containing either an alkyl or carboxylic
acid (CA) tail to each (Fig. 1). These various supported catalysts
exhibited differences in surface area, acid site strength, and
metal dispersion; therefore, our main objective was to understand
how application of PAs influenced properties and catalytic
Supported Pt catalysts were prepared by incipient wetness
impregnation of Pt onto each support. Prior to catalyst synthesis,
each support was dried in air at 120 °C for at least 12 h. Next, a
solution of chloroplatinic acid hexahydrate (H2PtCl6Á6H2O) in dis-
tilled water was prepared and mixed with the desired mass of sup-
port in order to achieve the desired Pt weight loading of
approximately 3%. The resulting mixture was dried at 120 °C for
12 h to remove water, and the dried solid was then calcined in
air at 400 °C for 4 h. Lastly, the calcined solid was reduced in 10%
H2/He at 250 °C for 2 h, resulting in the formation of metallic Pt
nanoparticles.
BET surface area of each catalyst support was measured using a
Micrometrics (Norcross, GA) Chemisorb 2720. Samples were pre-
treated at 220 °C for 1 h in helium (20 sccm), then cooled to room
temperature prior to nitrogen saturation with 30% nitrogen in
helium. Adsorption and desorption of nitrogen were measured,
and the desorption measurements were used to determine the
powder surface area. Triplicate measurements were performed
on each substrate, and the associated errors were propagated with
errors from ICP-OES analysis to determine SAM surface density
(Equations S7–S16). Samples were also analyzed with an ARL
3410+ inductively coupled optical emission spectrometer (ICP-
OES) to measure elemental composition. A blank and three stan-
dards were used for calibration. Standards were made by diluting
certified standards. The samples were dissolved using a modified
technique developed by Farrell, Matthes and Mackie (1980) [33].
Briefly, this procedure starts by adding 5 ml of a 7:3 mixture of
hydrochloric acid and hydrofluoric acid and then 2 ml of nitric acid
to the digestion tubes, which are heated to 95 °C in a digestion
block (HotBlock by Environmental Express) for approximately
2 h. Samples were then cooled and brought up to 50 ml with a
1.5 wt% boric acid solution. The samples were then reheated to

P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
313
95 °C for approximately 15 min before being cooled to room tem-
perature for analysis. Phosphonate coverages and the Pt weight
loading were calculated from the data collected by this procedure,
using the equations given in the Supporting Information (Equa-
tions S7–S16 for phosphonate coverages and Equations S17–S20
for Pt weight loadings).
and CA acid tails, respectively. High temperature measurements
were performed under 100 ml minÀ1 Ar, by heating the sample to
incremental temperatures and holding for 15 min, then cooling
to 50 °C prior to collection of spectra. Background spectra of
uncoated samples at 50 °C were subtracted to produce the
reported spectra. Infrared spectroscopy analysis following pyridine
adsorption was performed under a flow of 100 ml minÀ1. All spec-
Transmission electron microscopy (TEM) was performed on
each of the uncoated catalysts to evaluate the particle size of the
resulting Pt nanoparticles. Catalyst particles were suspended in
ethanol at ꢀ1% wt/vol. A volume of 7 ml of the suspension was
drop-cast onto 200 mesh, carbon-coated copper grids (SPI Supplies,
West Chester, PA). Samples were air dried prior to imaging. Imag-
ing was performed with a FEI Tecnai G2 20 Twin 200 kV LaB6 TEM
(FEI, Hilsboro, OR) at an accelerating voltage of 200 kV. Images
were captured with a Gatan UltraScan 1000 camera (Gatan,
Pleasanton, CA). At least 20 images were obtained from at least 5
different particle clusters for each sample. Particle size analysis
was performed manually using ImageJ software on a sample size
of at least 100 particles. Representative images of each uncoated
catalyst are provided in Fig. S1. The average Pt particle sizes on
Pt/Al2O3, Pt/TiO2, Pt/SiO2-Al2O3, and Pt/CeO2 were found to be
6.7 1.2 nm, 1.9 0.2 nm, 19.2 2.5 nm, and 4.0 1.9 nm, respec-
tively, as reported in Table S1. As a complementary measure, the
metal dispersion and Pt particle sizes of the uncoated catalysts
were also measured by CO chemisorption on a Micrometrics
(Norcross, GA) Chemisorb 2720. Samples were reduced in H2 at
200 °C for 2 h, then cooled to room temperature prior to perform-
ing injections of 10% CO/He. For Pt/Al2O3, Pt/TiO2, Pt/SiO2-Al2O3,
and Pt/CeO2, the metal dispersion was found to be 58.0% 1.7%,
55.1% 3.2%, 2.4% 0.2%, and 44.8% 1.6%, respectively. These
values and the corresponding estimated particle sizes are reported
in Table S1.
Phosphonate SAMs were prepared on the synthesized catalysts
(Pt/Al2O3, Pt/TiO2, Pt/SiO2-Al2O3, Pt/CeO2) by a liquid deposition
technique [23]. First, the mass of PA required for a full monolayer
(assuming monodentate binding) on a desired mass of catalyst was
determined by Equation S1, using reported values for the density of
surface hydroxyl groups on each support [34–37]. Next, a 10 mM
solution of PA in THF was prepared, containing a three-fold excess
of PA relative to the number of moles required for a full monolayer.
The desired mass of prepared catalyst powder was then added to
the PA solution, and the resulting suspension was stirred for 12–
16 h at ambient conditions. The solid was then separated by cen-
trifugation in an Eppendorf 5804 centrifuge at 8000 rpm for
8 min and annealed at 120 °C for 6 h. The resulting solid was
washed and centrifuged three times (at 8000 rpm for 8 min), using
the same amount of THF used to prepare the initial solution, in
order to remove any physisorbed PAs. Finally, the SAM coated
powder was dried under vacuum at room temperature.
tra were collected using 100 scans, and at a resolution of 4 cmÀ1
.
Samples were first loaded into a closed cell attachment (Harrick)
for DRIFTS and pretreated in He at 200 °C for at least 1 h. Samples
were then cooled to 50 °C and a background spectra was collected.
Pyridine was then dosed into the cell by flowing He through a pyr-
idine bubbler, held at 0 °C, for 15 min. Next, the system was
flushed in He for 1 h while holding at 50 °C prior to the collection
of spectra. For higher temperature measurements, samples were
heated to a desired temperature and held for 10 min, then cooled
back to 50 °C for collection of spectra.
2.3. Reactor studies
Hydrogenation reactions were performed in a Pyrex tube,
packed bed, continuous-flow reactor at atmospheric pressure.
Helium was bubbled through a liquid reactant (benzyl alcohol)
held in a heated water bath at 60 °C and mixed with H2 and
make-up He upstream of the reactor. The resulting feed stream
had gas-phase mole fractions of
YH2 = 15% and Ybenzyl
alco-
hol = 0.075%, and a total gas flow rate of 83 sccm. Reactions were
run at a temperature of 180 °C. The mass of each catalyst used
was varied as needed to achieve a desired conversion. The feed
stream and reactor effluent were analyzed using an SRI Instru-
ments 8610C gas chromatograph equipped with a Restek MXTÒ-
5 capillary column and a flame ionization detector. Each experi-
ment was run to steady state, which was defined as running for
at least 200 min and having a conversion within 0.5% over 4 con-
secutive measurements, taken over a period of 1 h. Average con-
version and selectivity were determined from these steady state
measurements, and used with the number of active metal sites
to calculate the reported turnover frequencies by Equations S2–
S6. The number of metal active sites were determined by CO
chemisorption of the unmodified catalysts. Because previous work
has shown that PAs do not bind to the surface of supported or
unsupported Pt nanoparticles, these values are not expected to
change after surface modification, and were thus used to calculate
turnover frequencies of the modified catalysts as well [21].
3. Results and discussion
3.1. Catalyst synthesis and characterization
Powder X-ray diffraction (XRD) data were collected using a
Rigaku Ultima IV diffractometer with a Cu K
a
source (40 kV,
Pt/Al2O3, Pt/TiO2, Pt/SiO2-Al2O3, and Pt/CeO2 catalysts were
modified with PPA and 2PEA. Thus, including the unmodified cat-
alysts, a total of twelve catalysts were tested and compared in this
study. Table 1 presents the total surface area of the unmodified cat-
alysts and the Pt weight loading. Note that the BET surface area is
reported for the unmodified catalyst support; previous work using
44 mA). Diffraction patterns were collected in the 2q range of
20–80° at a scan rate of 4° minÀ1. Samples (10–20 mg) were sup-
ported on a glass sample holder with a 0.2 mm recessed sample
area and were pressed into the recession with a glass slide to
obtain a uniform z-axis height. Data were compared to reference
card files from the International Center for Diffraction Data (Pt:
00-001-1190, Rutile TiO2: 00-001-1292, Anatase TiO2: 00-001-
0562, CeO2: 00-002-1306, Al2O3: 00-004-0858) to confirm the
identity and phase of the sample.
Infrared spectroscopy experiments were performed with 100
scans at 4 cmÀ1 resolution using a Thermo Scientific Nicolet 6700
FT-IR. A closed cell attachment (Harrick) was used for diffuse
reflectance infrared Fourier transform spectroscopy (DRIFTS).
Spectra were collected in both the hydrocarbon stretching region
and in the carbonyl stretching region, to verify the identity of alkyl
c-Al2O3 found that modification with PAs of similar size led to a
decrease in BET surface area of approximately 25% [31]. We carried
out DRIFTS experiments to confirm that catalysts were modified
with the intended alkyl and carboxylic acid functionalities. DRIFT
spectra of the modified catalysts were collected in the region of
2700–3000 cmÀ1 to characterize the CH3 and/or CH2 stretching
modes, shown in Fig. 2a. For catalysts modified with PPA, intense
peaks were observed at approximately 2965 cmÀ1 (asymmetric
methyl stretching), 2910 cmÀ1 (symmetric methyl stretching),
2934 cmÀ1 (asymmetric methylene stretching) and 2880 cmÀ1

314
P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
Table 1
Catalyst support BET surface areas, Pt loadings, and densities of surface modifiers. Errors associated with BET surface area represent the standard deviation of three separate
measurements on the starting support material prior to deposition of Pt. Pt loadings and densities of surface modifiers were determined by ICP-OES analysis, and the errors
represent the standard deviation of three separate measurements.
Catalyst support
-Al2O3
TiO2 (P25)
SiO2-Al2O3
CeO2
BET surface area (m2 gÀ1
)
Pt wt%
PPA surface density (Phosphonates nmÀ2
)
2PEA surface density (Phosphonates nmÀ2
)
c
56
48
3
5
3.1 0.2
2.7 0.1
3.0 0.1
2.8 0.1
4.9 0.8
3.4 0.6
2.2 0.1
3.1 0.4
5.0 0.5
3.5 0.6
4.1 0.2
3.3 0.4
354 12
42
4
Fig. 2. FT-IR spectra of phosphonate coated Pt catalysts in (a) the hydrocarbon stretching region and (b) the carboxyl stretching region. Unmodified Pt catalysts for each
support were used as the background for measurements of the coated catalysts. Spectra were scaled to improve visibility of relevant peaks.
(symmetric methylene stretching). The relatively high frequency of
the asymmetric CH2 stretching peak indicates considerable disor-
der in the monolayer, which is common for short-chain alkyl
ligands [38]. Modification of catalysts using 2PEA resulted in a car-
boxylic acid-terminated monolayer, characterized by strong peaks
in the region of 1700–1720 cmÀ1 (C@O stretching) shown in
Fig. 2b, indicating the presence of unbound CA tails. Asymmetric
methylene stretching near 2960 cmÀ1 was also observed in 2PEA-
modified catalysts. Overall, both modifiers produced a somewhat
disordered monolayer, as indicated by the higher frequencies of
asymmetric CH2 stretching, consistent with previous observations
of monolayers having short chains and/or polar tails [28,38]. Fur-
thermore, greater disorder has been shown to be consistent with
under-coordinated binding of PAs, suggesting the presence of
Brønsted acidity at the PA head group [28]. Spectra of the PPA-
modified catalysts were also collected at high temperatures to ver-
ify the stability of these modifiers on each of the supports. Fig. S2
shows the spectra of each PPA-modified catalyst in the hydrocar-
bon stretching region at 50 °C and 400 °C. Comparison of these
spectra suggests that the PA modifiers remain intact up to
400 °C, consistent with previous work [20,21,31,32].
ICP-OES analysis was used to determine the Pt wt% of the
uncoated catalysts, as well as the density of PA modifiers on each
coated catalyst, which are reported in Table 1. The measured Pt
loadings exhibited good agreement with the nominal loading of
3 wt%. By detecting the relative amounts of phosphorus in each
of the coated samples and using the measured BET surface areas
of each support, the average density (in phosphonates nmÀ2) of
2PEA and PPA were determined on each catalyst. The highest cov-
erages of approximately 5 phosphonates nmÀ2 were observed on
Pt/Al2O3. On Pt/TiO2 and Pt/CeO2, coverages of around 3 phospho-
nates nmÀ2 were measured. Pt/SiO2-Al2O3 was the only catalyst
found to have significantly different coverages of the two modi-
fiers; PPA had a coverage of only 2.2 phosphonates nmÀ2, while
2PEA had a higher coverage of 4.1 phosphonates nmÀ2. One possi-
ble explanation for the higher coverage of 2PEA on this support is
that 2PEA contains stronger acid sites than PPA, which may lead to
greater interaction with the strong Brønsted acid sites present on
the SiO2-Al2O3 support during deposition. However, further work
is needed to identify the cause of this result.
3.2. Pyridine DRIFTS studies
XRD was performed to analyze the supporting materials both
before and after PA modification. The results, shown in Fig. S3, con-
firmed the crystallinity of Al2O3 and CeO2. The TiO2 was deter-
mined to be mixed phase, as evidenced by a combination of
diffractions from rutile and anatase crystallites. These results also
show that deposition of PAs had no detectable effect on the bulk
structure of the supporting materials. Additionally, previous work
has suggested that PAs do not bind to the surface of Pt nanopar-
ticdles [21], and thus, the only apparent change to the prepared
catalysts after PA deposition was the presence of PAs on the surface
of the catalyst support.
Pyridine DRIFTS was used to characterize the acid sites on each
of the twelve catalysts used in this study, shown in Fig. 3. Pyridine,
when dosed onto acidic solid surfaces, exhibits characteristic
stretching modes associated with Lewis and Brønsted acid sites
[39,40]. Specifically, pyridine coordinately bound or hydrogen-
bonded to Lewis acid sites exhibits stretching at ꢀ1450 cmÀ1
(1440 cmÀ1–1460 cmÀ1), whereas pyridinium ions adsorbed to
Brønsted acid sites generally exhibit stretching at ꢀ1540 cmÀ1
,
but have been observed in the range of 1540–1550 cmÀ1
[39,41–43]. Additionally, peaks near 1490 cmÀ1 are observed, and

P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
315
Fig. 3. FT-IR spectra of modified and unmodified (a) Pt/Al2O3 catalysts, (b) Pt/TiO2 catalysts, (c) Pt/SiO2-Al2O3 catalysts, and (d) Pt/CeO2 catalysts, following pyridine dosing at
50 °C. Uncoated catalysts are shown in black (bottom), PPA coated catalysts are shown in red (middle), and 2PEA coated catalysts are shown in blue (top). Spectra prior to
pyridine dosing were used as the backgrounds for all experiments. Spectra have been scaled as needed to improve visibility of relevant peaks.
represent a combination of both pyridine binding to Lewis acid
sites (1488 cmÀ1–1503 cmÀ1) and pyridinium ion adsorption to
Brønsted acid sites (1485 cmÀ1–1500 cmÀ1), although it has been
suggested that this peak can be used to indicate Brønsted acidity
[39]. Lewis acid sites on metal-oxides are associated with under-
coordinated surface metal ions, and are expected to be present
on all supports used here [44,45]. Brønsted acid sites are associated
with surface hydroxyl groups that are partially positively charged,
and thus able to protonate an adsorbed base [44,46].
On the Pt/Al2O3 catalysts, shown in Fig. 3a, both unmodified and
modified catalysts exhibited Lewis acidity, as evidenced by the
peaks near 1450 cmÀ1 and 1490 cmÀ1. On uncoated Pt/Al2O3, no
Brønsted acidity was detected, as indicated by the lack of a clear
peak at 1540 cmÀ1. After PA modification, however, distinct peaks
appeared around 1540 cmÀ1, suggesting that both the PPA and
2PEA modifiers provided Brønsted acid sites at the catalyst surface.
After modification with 2PEA, the appearance of the Brønsted acid
peak (1540 cmÀ1) was accompanied by a significant reduction in
the Lewis acid peak (1450 cmÀ1), and an increase in the peak at
1490 cmÀ1, which represents binding to either Lewis or Brønsted
acid sites [39,40]. Thus, modification of Pt/Al2O3 with 2PEA
appeared to simultaneously increase Brønsted acidity and reduce
small peak at 1546 cmÀ1 was observed after dosing pyridine onto
the uncoated catalyst, suggesting a small amount of Brønsted acid-
ity [41,42]. After modification with PAs, this peak shifted to
1540 cmÀ1, indicating a change in the nature of the Brønsted acid
sites when PAs were present. Again, modification with 2PEA
resulted in a significant reduction of the peak at 1450 cmÀ1. On
the Pt/SiO2-Al2O3 catalysts, shown in Fig. 3c, Brønsted acidity
was detected on the unmodified and PA-modified catalysts. This
catalyst exhibited both Brønsted and Lewis acidity, as indicated
by the peaks at 1540 cmÀ1, 1490 cmÀ1, and 1450 cmÀ1. Modifica-
tion with PAs did not alter the appearance of these three peaks.
Additionally, modification with 2PEA again led to a reduction in
the strength of the Lewis acid peak relative to the Brønsted acid
peak.
Unexpectedly, there were no Brønsted acid peaks resolved
around 1540 cmÀ1 on any of the Pt/CeO2 catalysts, as shown in
Fig. 3d. Although the absence of Brønsted acidity on unmodified
CeO2 was expected, PA modification of the supports was expected
to introduce Brønsted acid sites, similar to the Al2O3 support. The
absence of a clearly resolvable peak near 1540 cmÀ1 could indicate
that PAs adsorb in a tridentate configuration in which acid sites are
not introduced, or that any acid sites are simply weaker. One pos-
sible indication in support of the latter hypothesis is that on the
PA-modified Pt/CeO2 catalysts a significant increase in the peak
the availability of Lewis acid sites on native
c-Al2O3. The Pt/TiO2
catalysts, shown in Fig. 3b, exhibited a similar result, although a

316
P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
at 1490 cmÀ1 was observed. Because the peak at 1490 cmÀ1 is rep-
resentative of pyridine binding to either Brønsted or Lewis acid
sites, the increased intensity of this peak may suggest the presence
of Brønsted acid sites after PA modification. The reactivity results
presented below offer further support to this conjecture. Overall,
the results of these experiments indicate the PA modifiers provide
or increase the number of Brønsted acid sites to Pt/Al2O3 and Pt/
TiO2 catalysts, and possibly on Pt/CeO2. Conversely, PA modifica-
tion appears to have little to no effect on the presence of Brønsted
acid sites on Pt/SiO2-Al2O3.
Previous work has suggested that the strength of Brønsted acid
sites present in bifunctional acid-metal catalysts can be correlated
to the DO activity, with stronger acid sites providing the highest
DO rates [20,21]. The relative strength of the Brønsted acid sites
was therefore measured on uncoated and coated catalysts by
observing the pyridinium desorption temperature with DRIFTS.
In our previous study, this approach was used for Pd/Al2O3 cata-
lysts modified with a variety of PAs, and it was found that short-
chain PCAs, such as 2PEA, provided stronger acid sites compared
to alkyl PAs, such as PPA [20]. A similar result would be expected
2PEA with the surface, resulting in the weakening of nearby Lewis
acid sites.
3.3. Hydrodeoxygenation of benzyl alcohol
Each catalyst was evaluated for the HDO of benzyl alcohol, with
the goal being to determine how PA modification of each support
affects the HDO performance relative to the uncoated catalysts.
Catalysts were assessed at 5–15% steady state conversion of benzyl
alcohol, except for the Pt/SiO2-Al2O3 catalysts, which were
assessed at less than 1% conversion. Due to the highly acidic nature
of the SiO2-Al2O3 support, higher conversions resulted in the for-
mation of high molecular weight products, likely formed by con-
densation and coupling reactions, that quickly deactivated the
catalysts at higher conversions. The overall rates, product forma-
tion rates, and product selectivities of each catalyst are summa-
rized in Table S2. Control experiments were also performed using
the unmodified catalyst supports, as well as each support modified
with 2PEA. In these experiments, the only reactivity observed was
low rates of dehydrogenation to form benzaldehyde (Fig. 1a). The
rates of dehydrogenation (per mass of catalyst) accounted for less
than 10% of the dehydrogenation rates observed when Pt was pre-
sent, except on SiO2-Al2O3, where the support was found to
account for nearly 80% of the dehydrogenation activity.
When Al2O3 and TiO2 supports were used, the effect of PA mod-
ification led to comparable changes in HDO performance. For both
Pt/Al2O3 and Pt/TiO2, modification with PPA led to a small decrease
in the rate of production of toluene, the HDO product, as shown in
Fig. 5. However, benzaldehyde production was found to decrease
as well, so that the selectivity of toluene increased from 71% to
83% after PPA modification of Pt/Al2O3, and from 57% to 62% after
modification of Pt/TiO2, as shown in Fig. S4. When these catalysts
were modified with 2PEA, which has been shown to provide stron-
ger Brønsted acid sites than alkyl PAs [20], the rate of toluene pro-
duction over Pt/Al2O3 increased by 1.5x, and by 1.9x over Pt/TiO2.
Additionally, the 2PEA-modified catalysts further increased the
toluene selectivity to 92% over 2PEA/Pt/Al2O3 and 83% over 2PEA/
Pt/TiO2. The highest HDO rate of all catalysts studied here was
measured over 2PEA/Pt/TiO2, which was nearly double the rate
measured over uncoated Pt/TiO2. Consistent with previous obser-
vations, the catalysts containing stronger acid sites (Fig. 4) there-
fore yielded the higher rates of HDO.
for TiO2 supported catalysts, because
c-Al2O3 and TiO2 contain
Lewis acid sites of similar strength [18]. Indeed, on the Pt/TiO2 cat-
alysts, it was found that the Brønsted acid strength followed the
trend Pt/TiO2 < PPA/Pt/TiO2 < 2PEA/Pt/TiO2. As shown in Fig. 4a,
pyridine desorbed from Brønsted acid sites on Pt/TiO2 at 240 °C,
as indicated by the disappearance of the peak at 1544 cmÀ1. On
PPA/Pt/TiO2, this peak persisted up to 260 °C, and up to 280 °C on
2PEA/Pt/TiO2. Conversely, on the Pt/SiO2-Al2O3 catalysts, the
Brønsted acid sites present after PA modification were found to
be weaker than those present on the unmodified catalyst. Instead,
the Brønsted acid sites inherent to the SiO2-Al2O3 support were
found to be strongest, so that the acid strength trended as PPA/
Pt/SiO2-Al2O3 < 2PEA/Pt/SiO2-Al2O3 < Pt/SiO2-Al2O3. As shown in
Fig. 4b, pyridine remained adsorbed to Brønsted acid sites on Pt/
SiO2-Al2O3 up to at least 500 °C, evidenced by the persistence of
the peak at 1546 cmÀ1. On 2PEA/Pt/SiO2-Al2O3, this peak only per-
sisted up to 300 °C, and to 200 °C on PPA/Pt/SiO2-Al2O3. It is noted
that pyridine desorption occurred below 400 °C on all PA-modified
catalysts tested. Because it has been shown that PA modifiers
remain intact on the surface up to temperature of at least 400 °C,
it can be concluded that desorption of pyridine is dependent on
the strength of adsorption to the acid sites provided by PA modi-
fiers, and not due to the removal or decomposition of these sites
[20,21,31,32].
On both the TiO2 and SiO2-Al2O3 supported catalysts, it was also
found that PA modification weakened the strength of Lewis acid
sites. This can be seen by observing the presence and relative size
of the peak at or around 1450 cmÀ1 at increasing temperature. PA
modifiers likely blocked Lewis acid sites on the support, lowering
the signal even at low temperatures. Additionally, the peak around
1450 cmÀ1 was found to weaken significantly or even disappear at
higher temperatures when PA modifiers were used. In general, by
comparing the spectra of 2PEA- and PPA-modified catalysts, it
appeared that 2PEA weakened the Lewis acid sites more so than
PPA. One possible explanation for this difference is that the two
modifiers have a different impact on the electronic structure of
the support via through-surface inductive effects. However, previ-
ous computational work has indicated that changing the pendant
organic ligands at positions remote from the P atom does not influ-
ence the local charges present on the support, even when strongly
electron-withdrawing (e.g., AF) or -donating (ANH2) groups are
employed [23]. Another possible explanation is that the 2PEA mod-
ifier provides more steric hindrance or blocks Lewis acid sites
through a direct carboxylic acid-surface interaction, reducing the
availability of Lewis acid sites on the support. Finally, this differ-
ence may be due to a direct interaction between the CA tail of
In comparing the coated and uncoated catalysts, modification of
Pt/CeO2 yielded somewhat similar results to those found for the Pt/
Al2O3 and Pt/TiO2 catalysts, with the major difference being the
effect of PPA. While this PA had little effect on the HDO perfor-
mance of Pt/Al2O3 and Pt/TiO2, it increased the rate of toluene pro-
duction by
a factor of 1.7 when deposited onto Pt/CeO2.
Modification with 2PEA further increased the rate of HDO, as
expected, to nearly 2.3 times that of the uncoated catalyst. How-
ever, as opposed to uncoated Pt/Al2O3 and Pt/TiO2, the rate of
HDO measured over uncoated Pt/CeO2 was quite low, and the
toluene selectivity was only 6%, making it an undesirable catalyst
for this reaction even after PA modification. Although the HDO
rates increased after modification with PPA and 2PEA, the selectiv-
ity of toluene was only increased to 9% and 12%, respectively, as
shown in Fig. S4. From the results of the TiO2- and CeO2-
supported catalysts, which both showed increased rates of HDO
after PA modification, it can be concluded that high reducibility
of the support does not hinder the promotional effects of PAs.
Although some previous reports have suggested that an oxygen-
vacancy mechanism may be responsible for deoxygenation over
reducible supports, the promotional effects observed here indicate
that the PAs are still able to promote activity via a dehydration/
hydrogenation mechanism [47,48].

P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
317
Fig. 4. Temperature-dependent pyridine DRIFTS spectra of coated and uncoated (a) Pt/TiO2 catalysts and (b) Pt/SiO2-Al2O3 catalysts. The Brønsted acid strengths of each
catalyst were found to trend with the HDO activity.
Unlike the other three supports discussed above, PA modifica-
tion of Pt/SiO2-Al2O3 had a negative effect on HDO performance,
regardless of the PA used. Uncoated Pt/SiO2-Al2O3 was found to
have a similar toluene production rate to that of uncoated Pt/
Al2O3, and a toluene selectivity of 60%. After modification with
PPA, the toluene production rate decreased by over 75%, and the
toluene selectivity decreased to 48%. When 2PEA was used as the
modifier, the toluene production rate was decreased by 68% com-
pared to the uncoated catalyst, while the toluene selectivity
improved to 67%. Although 2PEA/Pt/SiO2-Al2O3 performed better
overall than PPA/Pt/SiO2-Al2O3, both catalysts performed signifi-
cantly worse for HDO than the unmodified catalyst. Moreover,
modification of Pt/SiO2-Al2O3 resulted in a substantial decrease
in the overall activity of the catalyst, regardless of the modifier
used, shown in Fig. S4.
Catalyst rates were also analyzed on the basis of interfacial area,
assuming hemispherical particles. The results of this analysis are
reported in Tables S3 and S4 for particle sizes measured by CO
chemisorption and by TEM analysis, respectively. While this anal-
ysis does not affect the product selectivities or relative rates mea-
sured on a given catalyst before and after PA modification, it does
affect the relative rates between the different supported catalysts,

318
P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
Fig. 5. Turnover frequencies (TOF) of products formed from benzyl alcohol under hydrogenation conditions. T = 177 °C, yBA = 0.075% , yH2 = 15% , total flow of 83 sccm. The
mass of catalyst was varied as needed to achieve a steady state conversion of 10 5% for the Al2O3, TiO2, and CeO2 supported catalysts, and a steady state conversion of less
than 1% for the SiO2-Al2O3 supported catalysts. Error bars represent the standard deviation of three separate experiments.
due to the different particle sizes on each support. Turnover fre-
quencies calculated on this basis caused the Pt/SiO2-Al2O3 catalysts
to increase relative to the other catalysts used in this study, due to
the much larger Pt particle sizes measured on the SiO2-Al2O3 sup-
port, particularly when particle sizes determined by CO chemisorp-
tion were used in the calculation. The use of interfacial area for rate
normalization may be more appropriate for comparisons between
the different supports used here, and are in better accordance with
the expectation that SiO2-Al2O3 supports should provide acid sites
that are highly active for the deoxygenation reaction.
Overall, the results of these experiments suggest that PAs can
improve HDO by one of two mechanisms. The first is to provide
Brønsted acid sites at the metal-support interface of catalysts that
do not have inherent Brønsted acidity, such as Al2O3 and possibly
CeO2. The second is to provide stronger and more tunable acid sites
to supports that may already have weak Brønsted acidity, such as
TiO2. However, when a support with strong Brønsted acidity is
used, such as SiO2-Al2O3, PA modification only serves to replace
or block these sites with potentially weaker sites, resulting in a loss
of HDO activity. Thus, one way that different metal-oxide supports
can result in different changes in HDO activity is that the Brønsted
acid sites on the support can be different in character, and so the
blocking or replacing of these sites by PAs has different effects
on the HDO activity (either increasing or decreasing).
surface oxygen atom to be associated with a more positively
charged phosphorus atom, and consequently a stronger Brønsted
acidity. We previously observed that pyridine desorbed from
2PEA-modified Al2O3 at 250 °C, whereas in this study pyridine
remained adsorbed to 2PEA/Pt/TiO2 up to temperatures of 280 °C
(Fig. 4), indicating that PAs have stronger Brønsted acidity after
deposition onto TiO2 than Al2O3. This is consistent with DFT calcu-
lations suggesting that the surface oxygen atoms present in TiO2
carry a more negative charge those present in Al2O3 [49,50]. Fur-
thermore, these results are consistent with the greater extent of
increase in HDO activity observed after 2PEA-modification of Pt/
TiO2 compared to Pt/Al2O3 (Fig. 5). Thus, a second way that differ-
ent metal-oxide supports can lead to different changes in HDO
activity is that the electronic structure of the support may influ-
ence the Brønsted acid strength of the modifier, which in turn
effects the degree of improvement in HDO.
While the electronic structure of the support can also influence
the Lewis acidity, it is proposed that Lewis acid sites were not a
major participant in HDO over the catalysts used in this study, as
PA modification was found to weaken the Lewis acid sites, but gen-
erally improved HDO rates. Finally, the effect of PA modification to
a mostly inert support, such as SiO2, may be of interest, though PAs
were found to not deposit onto SiO2 using the deposition proce-
dure described herein.
A further question that may arise is how the electronic structure
of the metal-oxide support influences the strength of Brønsted acid
sites in PA modifiers, and thus the extent of potential improvement
on HDO activity due to modification. Because PAs bind directly to
oxygen in the support, these atoms essentially serve as a ligand
to the PA, and so the electronic structure of these atoms may influ-
ence the PA strength. One would expect a more negatively charged
4. Conclusion
PAs were used to modify the support surface of Pt/Al2O3,
Pt/TiO2, Pt/SiO2-Al2O3, and Pt/CeO2 catalysts, and were tested in
the HDO of benzyl alcohol. Over Al2O3, TiO2, and CeO2-supported

P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
319
[9] P.N. Ciesielski, M.B. Pecha, V.S. Bharadwaj, C. Mukarakate, G.J. Leong, B. Kappes,
M.F. Crowley, S. Kim, T.D. Foust, M.R. Nimlos, Advancing catalytic fast pyrolysis
through integrated multiscale modeling and experimentation: challenges,
progress, and perspectives, WIREs Energy Environ. 7 (4) (2018) 1–19.
[10] D.A. Ruddy, J.A. Schaidle, J.R. Ferrell III, J. Wang, L. Moens, J.E. Hensley, Recent
advances in heterogeneous catalysts for bio-oil upgrading via ‘‘ex Situ
Catalytic Fast Pyrolysis”: catalyst development through the study of model
compounds, Green Chem. 16 (2014) 454–490.
[11] K.L. Deutsch, B.H. Shanks, Hydrodeoxygenation of Lignin Model Compounds
over a Copper Chromite Catalyst, Appl. Catal. A 447–448 (2012) 144–150.
[12] A.M. Robinson, J.E. Hensley, J.W. Medlin, Bifunctional catalysts for upgrading
of biomass-derived oxygenates: a review, ACS Catal. 6 (8) (2016) 5026–5043.
[13] L.O. Mark, A.H. Jenkins, H. Heinz, J.W. Medlin, Furfuryl alcohol deoxygenation,
decarbonylation, and ring-opening on Pt (111), Surf. Sci. 677 (2018) 333–340.
[14] R.C. Nelson, B. Baek, P. Ruiz, B. Goundie, A. Brooks, M.C. Wheeler, B.G.
Frederick, L.C. Grabow, R.N. Austin, Experimental and theoretical insights into
the hydrogen-efficient direct hydrodeoxygenation mechanism of phenol over
Ru/TiO2, ACS Catal. 5 (11) (2015) 6509–6523.
[15] D.D. Laskar, B. Yang, H. Wang, J. Lee, Pathways for biomass-derived lignin to
hydrocarbon fuels, Biofuels Bioprod. Bioref. 7 (2013) 602–626.
[16] M. Besson, P. Gallezot, C. Pinel, Conversion of biomass into chemicals over
metal catalysts, Chem. Rev. 114 (2014) 1827–1870.
[17] G.W. Huber, R.D. Cortright, J.A. Dumesic, Renewable alkanes by aqueous-phase
reforming of biomass-derived oxygenates, Angew. Chem. Int. Ed. 43 (12)
(2004) 1549–1551.
[18] A. Corma, Inorganic solid acids and their use in acid-catalyzed hydrocarbon
reactions, Chem. Rev. 95 (3) (1995) 559–614.
catalysts, deposition of 2PEA was found to increase the rate of
HDO, while deposition of PPA only improved the HDO rate when
applied to Pt/CeO2. When PAs were used to modify the support
of Pt/SiO2-Al2O3, which natively contains strong Brønsted acidity,
the HDO performance was found to decrease by as much as 76%.
Pyridine DRIFTS measurements suggested that improved rates of
HDO were observed when modification with PAs led to incorpora-
tion of Brønsted acid sites or provided stronger Brønsted acid sites
than those present prior to modification. Additionally, all catalysts
modified with 2PEA, which is known to provide stronger acid sites
than PPA, performed better than the equivalent catalyst modified
with PPA. Overall, the results reported here show that PAs can be
deposited onto a variety of catalyst supports and can provide sub-
stantial enhancements in HDO rates and selectivity, with the
exception of a supporting material that already contains an abun-
dance of strong acid sites. More broadly, these results suggest that
PAs can be used to modify a variety of metal oxide supports, and
are capable of strongly influencing catalyst reactivity, opening
the door to a wide variety of materials and chemistries that can
be examined using PA modification.
[19] N.C. Nelson, Z. Wang, P. Naik, J.S. Manzano, M. Pruski, I.I. Slowing, Phosphate
modified ceria as a brønsted acidic/redox multifunctional catalyst, J. Mater.
Chem. A 5 (2017) 4455–4466.
Acknowledgements
[20] P.D. Coan, L.D. Ellis, M.B. Griffin, D.K. Schwartz, J.W. Medlin, Enhancing
This work was funded by the US Department of Energy, Office of
Science, Basic Energy Sciences Program, Chemical Sciences, Geo-
sciences and the Biosciences Division, under Grant No. DE-
SC0005239. Additional support for sample characterization was
provided by the U.S. Department of Energy, Office of Energy Effi-
ciency and Renewable Energy (EERE), Bioenergy Technologies
Office (BETO), under Contract DE-AC36-08-GO28308 at the
National Renewable Energy Laboratory (NREL). The research was
performed in collaboration with the Chemical Catalysis for Bioen-
ergy Consortium (ChemCatBio), a member of the Energy Materials
Network (EMN). The views expressed in this article do not neces-
sarily represent the views of the U. S. Department of Energy or
the United States Government. P.D.C. acknowledges partial support
from the Department of Education Graduate Assistantship in Areas
of National Need (GAANN). The authors gratefully acknowledge Dr.
Fred Luiszer for performing ICP-OES analysis. P.D.C. also wishes to
thank Dr. Jing Zhang and Dr. Pengxiao Hao for useful discussions.
cooperativity in bifunctional acid
À
Pd catalysts with carboxylic acid-
functionalized organic monolayers, J. Phys. Chem. C 122 (2018) 6637–6647.
[21] J. Zhang, L.D. Ellis, B. Wang, M.J. Dzara, C. Sievers, S. Pylypenko, E. Nikolla, J.W.
Medlin, Control of interfacial acid-metal catalysis with organic monolayers,
Nat. Catal. 1 (2018) 148–155.
[22] P. Hao, D.K. Schwartz, J.W. Medlin, Phosphonic acid promotion of supported Pd
catalysts for low temperature vanillin hydrodeoxygenation in ethanol, Appl.
Catal. A 561 (2018) 1–6.
[23] L.D. Ellis, R.M. Trottier, C.B. Musgrave, D.K. Schwartz, J.W. Medlin, Controlling
the surface reactivity of titania via electronic tuning of self-assembled
monolayers, ACS Catal. 7 (2017) 8351–8357.
[24] S. Pawsey, K. Yach, L. Reven, Self-assembly of carboxyalkylphosphonic acids on
metal oxide powders, Langmuir 18 (13) (2002) 5205–5212.
[25] T. Bauer, T. Schmaltz, T. Lenz, M. Halik, B. Meyer, T. Clark, Phosphonate- and
carboxylate-based self-assembled monolayers for organic devices:
a
theoretical study of surface binding on aluminum oxide with experimental
support, ACS Appl. Mater. Interfaces 5 (13) (2013) 6073–6080.
[26] R. Luschtinetz, G. Seifert, E. Jaehne, H.J.P. Adler, Infrared spectra of
alkylphosphonic acid bound to aluminium surfaces, Macromol. Symp. 254
(2007) 248–253.
[27] R. Luschtinetz, A.F. Oliveira, J. Frenzel, J.O. Joswig, G. Seifert, H.A. Duarte,
Adsorption of phosphonic and ethylphosphonic acid on aluminum oxide
surfaces, Surf. Sci. 602 (7) (2008) 1347–1359.
[28] R. Luschtinetz, A.F. Oliveira, H.A. Duarte, G. Seifert, Self-assembled monolayers
of alkylphosphonic acids on aluminum oxide surfaces–a theoretical study, Z.
Anorg. Allg. Chem. 636 (8) (2010) 1506–1512.
[29] H. Dietrich, T. Schmaltz, M. Halik, D. Zahn, Molecular dynamics simulations of
phosphonic acid-aluminum oxide self-organization and their evolution into
ordered monolayers, Phys. Chem. Chem. Phys. 19 (2017) 5137–5144.
[30] M. Nilsing, S. Lunell, P. Persson, L. Ojamäe, Phosphonic acid adsorption at the
TiO2 anatase (101) surface investigated by periodic hybrid HF-DFT
computations, Surf. Sci. 582 (1–3) (2005) 49–60.
Appendix A. Supplementary material
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.jcat.2019.03.011.
References
[31] T.V. Cleve, D. Underhill, M.V. Rodrigues, C. Sievers, J.W. Medlin, Enhanced
[1] G.W. Huber, J.N. Chheda, C.J. Barrett, J.A. Dumesic, Production of liquid alkanes
by aqueous-phase processing of biomass-derived carbohydrates, Science 308
(5727) (2005) 1446–1450.
[2] M.B. Griffin, G.A. Ferguson, D.A. Ruddy, M.J. Biddy, G.T. Beckham, J.A. Schaidle,
Role of the support and reaction conditions on the vapor-phase deoxygenation
of m-cresol over Pt/C and Pt/TiO2 catalysts, ACS Catal. 6 (4) (2016) 2715–2727.
[3] J.C. Serrano-Ruiz, J.A. Dumesic, Catalytic routes for the conversion of biomass
into liquid hydrocarbon transportation fuels, Energy Environ. Sci. 4 (1) (2011)
83–99.
[4] J.N. Chheda, G.W. Huber, J.A. Dumesic, Liquid-phase catalytic processing of
biomass-derived oxygenated hydrocarbons to fuels and chemicals, Angew.
Chem. Int. Ed. 46 (2007) 7164–7183.
[5] R.K. Zeidan, S.J. Hwang, M.E. Davis, Multifunctional heterogeneous catalysts:
SBA-15-containing primary amines and sulfonic acids, Angew. Chem. Int. Ed.
45 (2006) 6332–6335.
hydrothermal stability of
c-Al2O3 Catalyst supports with alkyl phosphonate
coatings, Langmuir 34 (2018) 3619–3625.
[32] L.D. Ellis, J. Ballesteros-Soberanas, D.K. Schwartz, J.W. Medlin, Effects of metal
oxide surface doping with phosphonic acid monolayers on alcohol
dehydration activity and selectivity, Appl. Catal. A (2018).
[33] R.F. Farrell, S.A. Matthes, A.J. Mackie, Simple low-cost method for the
dissolution of metal and mineral samples in plastic pressure vessels, Rep.
Investig. – Bur. Mines (1980) 8480.
[34] M. Digne, P. Sautet, P. Raybaud, P. Euzen, H. Toulhoat, Use of DFT to achieve a
rational understanding of acid–basic properties of
226 (2004) 54–68.
c-alumina surfaces, J. Catal.
[35] R. Mueller, H.K. Kammler, K. Wegner, S.E. Pratsinis, OH surface density
of SiO2 and TiO2 by thermogravimetric analysis, Langmuir 19 (1) (2003) 160–
165.
[36] H. Yuan, D. Guo, X. Li, L. Yuan, W. Zhu, L. Chen, X. Qiu, The effect of CeO2 on Pt/
CeO2/CNT catalyst for CO electrooxidation, Fuel Cells 9 (2) (2009) 121–127.
[37] J.G. Larson, W.K. Hall, Studies of the hydrogen held by solids. VII. The exchange
of the hydroxyl groups of alumina and silica-alumina catalysts with
deuterated methane, J. Phys. Chem. 69 (9) (1965) 3080–3089.
[38] A. Ulman, Formation and structure of self-assembled monolayers, Chem. Rev.
96 (1996) 1533–1554.
[6] R.K. Zeidan, M.E. Davis, The effect of acid-base pairing on catalysis: an efficient
acid-base functionalized catalyst for aldol condensation, J. Catal. 247 (2007)
379–382.
[7] E.L. Margelefsky, R.K. Zeidan, M.E. Davis, Cooperative catalysis by silica-
supported organic functional groups, Chem. Soc. Rev. 37 (2008) 1118–1126.
[8] J.M. Notestein, A. Katz, Enhancing heterogeneous catalysis through cooperative
hybrid organic-inorganic interfaces, Chem. Eur. J. 12 (2006) 3954–3965.

320
P.D. Coan et al. / Journal of Catalysis 372 (2019) 311–320
[39] E.P. Parry, An infrared study of pyridine adsorbed on acidic solids.
Characterization of surface acidity, J. Catal. 2 (5) (1963) 371–379.
[40] C.A. Emeis, Determination of integrated molar extinction coefficients for
infrared absorption bands of pyridine adsorbed on solid acid catalysts, J. Catal.
141 (2) (1993) 347–354.
[41] C.R. Reddy, Y.S. Bhat, G. Nagendrappa, B.S.J. Prakash, Brønsted and Lewis
acidity of modified montmorillonite clay catalysts determined by FT-IR
spectroscopy, Catal. Today 141 (2009) 157–160.
[42] B. Chakraborty, B. Viswanathan, Surface acidity of MCM-41 by in situ IR
studies of pyridine adsorption, Catal. Today 49 (1999) 253–260.
[43] A. Platon, W.J. Thomson, Quantitative Lewis/Brønsted ratios using DRIFTS, Ind.
Eng. Chem. Res. 42 (2003) 5988–5992.
[44] M.I. Zaki, M.A. Hasan, L. Pasupulety, Surface reactions of acetone on Al2O3,
TiO2, ZrO2, and CeO2: IR spectroscopic assessment of impacts of the surface
acid-base properties, Langmuir 17 (2001) 768–774.
[46] M.I. Zaki, H. Knözinger, Carbon monoxide – a low temperature infrared probe
for the characterization of hydroxyl group properties on metal oxide surfaces,
Mater. Chem. Phys. 17 (1987) 201–215.
[47] P.M. de Souza, R.C. Rabelo-Neto, L.E.P. Borges, G. Jacobs, B.H. Davis, D.E.
Resasco, F.B. Noronha, Hydrodeoxygenation of phenol over Pd catalysts. Effect
of support on reaction mechanism and catalyst deactivation, ACS Catal. 7 (3)
(2017) 2058–2073.
[48] S.M. Schimming, O.D. LaMont, M. König, A.K. Rogers, A.D. D’Amico, M.M. Yung,
C. Sievers, Hydrodeoxygenation of guaiacol over Ceria – Zirconia catalysts,
ChemSusChem 8 (2015) 2073–2083.
[49] F. Labat, P. Baranek, C. Adamo, Structural and electronic properties of selected
rutile and anatase TiO2 surfaces: an Ab initio investigation, J. Chem. Theory
Comput. 4 (2008) 341–352.
[50] H.A. Dabbagh, M. Zamani, B.H. Davis, Nanoscale surface study and reactions
mechanism of 2-butanol over the
c-alumina (100) surface and nanochannel: a
[45] A.L. Farragher, Surface vacancies in close packed crystal structures, Adv.
Colloid Interface Sci. 11 (1979) 3–41.
DFT study, J. Mol. Catal. A Chem. 333 (2010) 54–68.