Catalytic deoxygenation of octanoic acid over supported palladium: Effects of particle size and alloying with gold

Article abstract of DOI:10.1021/cs501865n

Catalytic deoxygenation of octanoic acid (OA) to n-heptane was investigated over silica-supported Pd and PdAu catalysts at 260 °C and 1 atm in a fixed-bed microreactor. Pd/SiO₂ catalysts were prepared by incipient wetness (IW) and ion exchange (IE). Bimetallic catalysts were prepared using an IE procedure that is known to produce supported PdAu nanoparticles. The Pd nanoparticles (7.5 nm average size) in the Pd/SiO₂ (IW) catalyst exhibit well-defined (100) and (111) facets, as evidenced by high-resolution electron microscopy (HREM) and diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) of adsorbed CO. As expected, the smaller nanoparticles (1.5 nm average size) in the Pd/SiO₂ (IE) catalyst display strong linear and bridging CO DRIFTS bands. The PdAu/SiO₂ (1/1 atomic ratio) catalyst contains 5 nm alloy nanoparticles with Pd-rich surfaces, as evidenced by HREM with energy-dispersive X-ray (EDX) analysis and in situ EXAFS spectroscopy. DRIFTS thermal desorption experiments demonstrated that alloying with Au reduced the CO adsorption energy on surface Pd sites. The Pd/SiO₂ (IE) catalyst initially exhibited OA decarboxylation and decarbonylation activity but lost decarboxylation activity rapidly with time on stream (TOS). In contrast, the Pd/SiO₂ (IW) catalyst had only decarbonylation activity, deactivated less rapidly with TOS, and could be regenerated by heating in H₂ to remove OA residues. Alloying Pd with Au was found to improve catalyst stability without significantly affecting decarbonylation activity, as evidenced by the equivalent OA turnover frequencies of the Pd/SiO₂ (IW) and PdAu/SiO₂ (2/3) catalysts. The geometric and electronic effects of alloying reduce the CO adsorption energy and mitigate self-poisoning by OA and related species.

Full text of DOI:10.1021/cs501865n

Subscriber access provided by UNIVERSITY OF MICHIGAN LIBRARY
Article
Catalytic Deoxygenation of Octanoic Acid over Supported
Palladium: Effects of Particle Size and Alloying with Gold
KEYI SUN, Adria R. Wilson, Simon T. Thompson, and H. Henry Lamb
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/cs501865n • Publication Date (Web): 10 Feb 2015
Downloaded from http://pubs.acs.org on February 18, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
ACS Catalysis is published by the American Chemical Society. 1155 Sixteenth Street
N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 39
ACS Catalysis
1
2
3
4
5
6
7
8
9
Catalytic Deoxygenation of Octanoic Acid over Supported Palladium:
Effects of Particle Size and Alloying with Gold
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Keyi Sun1, Adria R. Wilson2, Simon Thompson1, and H. Henry Lamb1*
1Department of Chemical and Biomolecular Engineering, North Carolina State
University, Raleigh, NC 27695ꢀ7905, USA
2Department of Chemistry, Duke University, Box 90354, Durham, NC 27708, USA
*Corresponding author. Eꢀmail: lamb@ncsu.edu
Keywords: fatty acid, decarboxylation, decarbonylation, deactivation, selfꢀpoisoning,
bimetallic catalyst
ACS Paragon Plus Environment

ACS Catalysis
Page 2 of 39
1
2
3
4
Abstract
5
6
7
8
Catalytic deoxygenation of octanoic acid (OA) to nꢀheptane was investigated over
silicaꢀsupported Pd and PdAu catalysts at 260ºC and 1 atm in a fixedꢀbed microꢀreactor.
Pd/SiO2 catalysts were prepared by incipient wetness (IW) and ion exchange (IE).
Bimetallic catalysts were prepared using an IE procedure that is known to produce
supported PdAu nanoparticles. The Pd nanoparticles (7.5ꢀnm average size) in the Pd/SiO2
(IW) catalyst exhibit wellꢀdefined (100) and (111) facets, as evidenced by highꢀresolution
electron microscopy (HREM) and diffuse reflectance infrared Fourier transform
spectroscopy (DRIFTS) of adsorbed CO. As expected, the smaller nanoparticles (1.5ꢀnm
average size) in the Pd/SiO2 (IE) catalyst display strong linear and bridging CO DRIFTS
bands. The PdAu/SiO2 (1:1 atomic ratio) catalyst contains 5ꢀnm alloy nanoparticles with
Pdꢀrich surfaces, as evidenced by HREM with energyꢀdispersive xꢀray (EDX) analysis
and in situ EXAFS spectroscopy. DRIFTS thermal desorption experiments demonstrated
that alloying with Au reduced the CO adsorption energy on surface Pd sites. The Pd/SiO2
(IE) catalyst initially exhibited OA decarboxylation and decarbonylation activity but lost
decarboxylation activity rapidly with time on stream (TOS). In contrast, the Pd/SiO2 (IW)
catalyst had only decarbonylation activity, deactivated less rapidly with TOS, and could
be regenerated by heating in H2 to remove OA residues. Alloying Pd with Au was found
to improve catalyst stability without significantly affecting decarbonylation activity, as
evidenced by the equivalent OA turnover frequencies of the Pd/SiO2 (IW) and PdAu/SiO2
(2:3) catalysts. The geometric and electronic effects of alloying reduce the CO adsorption
energy and mitigate selfꢀpoisoning by OA and related species.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Page 3 of 39
ACS Catalysis
1
2
3
4
1. Introduction
5
6
7
8
Bioethanol and biodiesel are widely employed today as substitutes for gasoline and
diesel, respectively; however, because of their lower energy densities, poor oxidative
stabilities, and incompatibility with engines and transportation fuel infrastructure,1ꢀ2 these
firstꢀgeneration biofuels typically are employed in blends with fossil fuels. Secondꢀ
generation biofuels that overcome these deficiencies are being developed to provide
“dropꢀin” replacements for conventional petroleumꢀderived hydrocarbon fuels, including
jet fuel. With advances in algae cultivation technology, large areal yields of trigylcerides
and fatty acids (FAs) can be produced autotrophically using solar energy.3 Catalytic
deoxygenation can be used to convert FAs to hydrocarbons for transportation fuels via
decarboxylation and decarbonylation using supported Pd catalysts.4ꢀ6 These reactions are
illustrated below for octanoic acid (OA), the smallest FA found in typical bioꢀderived
feedstocks. OA decarboxylation produces nꢀheptane and CO2:
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
C7H15COOH → nꢀC7H16 + CO2
In contrast, OA decarbonylation produces heptene(s), CO and H2O.
C7H15COOH → CO + C7H14 + H2O
(1)
(2)
Typically, the decarboxylation pathway is preferred for biofuel production, because it
eliminates the H2 required for alkene hydrogenation and avoids the generation of CO, a
noble metal catalyst poison.
Murzin and coworkers have reported extensively on liquidꢀphase deoxygenation of
FAs for biofuels production.4ꢀ5, 7ꢀ10 Various supported transition metals, including Ni, Ru,
Pd, Ir, Os and Rh on silica, alumina and activated carbon, were evaluated for semiꢀbatch
liquidꢀphase stearic acid (SA) deoxygenation in dodecane.5 Pd on activated carbon (Pd/C)
was identified as highly active and selective for SA decarboxylation to nꢀheptadecane at
3
ACS Paragon Plus Environment

ACS Catalysis
Page 4 of 39
1
2
3
4
5
6
7
8
9
300°C and 17.5 atm under flowing He. Immer and Lamb found that 5 wt.% Pd/C
deactivated very rapidly under He in fedꢀbatch operation; however, quasiꢀsteady state SA
decarboxylation activity was achieved under 5% H2 with low net H2 consumption.11 Ford,
et al. investigated the influence of the support and FA alkyl chain length on initial
deoxygenation kinetics using supported Pd catalysts at 300°C and 15 atm under 5% H2.12
When compared to 5 wt.% Pd/C, 5 wt.% Pd/SiO2 catalysts were less active and exhibited
much lower CO2 selectivity. A wellꢀdispersed 5 wt.% Pd/Al2O3 catalyst had high initial
SA decarboxylation activity but deactivated rapidly. For evenꢀnumbered C10ꢀC18 FAs, as
carbon number decreased, reaction completion time increased and CO2 selectivity and
initial decarboxylation rate decreased over 5% Pd/C. Simakova, et al.9 reported that a 1
wt.% Pd/C catalyst containing 1.7ꢀnm Pd particles had the highest turnover frequency
(TOF) (1.1 × 10ꢀ1 sꢀ1) for deoxygenation of SAꢀpalmitic acid mixtures in the liquid phase
(dodecane solvent) at 300°C and 17.5 atm under 5% H2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Gasꢀphase deoxygenation of aliphatic acids (with the exception of acetic acid) over
Pd catalysts has been studied less extensively. Early work by Maier, et al.13 showed that
Pd/SiO2 was active for gasꢀphase deoxygenation of OA at 330°C and 1 atm in flowing
H2. OA deoxygenation was assumed to occur via decarboxylation; however, no reaction
was observed when an inert N2 carrier gas was employed. Boda, et al. studied OA
deoxygenation over Pd/C at 360°C under elevated H2 pressure (10 or 17 atm) in a
continuous flow reactor. OA decarbonylation was proposed to occur via a formic acid
intermediate that rapidly decomposes to CO and H2O.14 LugoꢀJosé, et al. reported TOFs
of 5ꢀ10 × 10ꢀ3 sꢀ1 for gasꢀphase deoxygenation of propanoic acid to ethane over Pd/SiO2
4
ACS Paragon Plus Environment

Page 5 of 39
ACS Catalysis
1
2
catalysts at 200°C and 1 atm under 20% H2.15 Alotaibi, et al. reported a similar TOF for
propanoic acid decarbonylation to ethene over Pd/SiO2 at 250°C under H2 at 1 atm.16
Pdꢀcontaining bimetallic catalysts have attracted much attention because of their
enhanced performance when compared to monometallic Pd catalysts.17 Among the
various bimetallics, PdꢀAu catalysts have been extensively investigated and employed for
many important applications including vinyl acetate synthesis,18 lowꢀtemperature CO
oxidation,19 and formic acid decomposition.20 Bimetallic enhancement of catalyst
activity, selectivity and stability may be attributed to geometric (ensemble) and electronic
(ligand) effects. Both types of effects have been used to describe the catalytic
modification of Pd by Au.21 Herein, we use PdAu alloys to probe the structure sensitivity
of alkyl carboxylic acid deoxygenation over Pd. Previous computational work by
Neurock indicated that the activation energy for acetic acid deoxygenation is increased
substantially by substituting inert Au atoms for Pd atoms in the active surface ensemble.22
Our group previously reported xꢀray absorption spectroscopy of PdAu/SiO2 catalysts
prepared by the method of Lam and Boudart,23 and consequently, it was a convenient
starting point for this work.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In this work, gasꢀphase deoxygenation of OA to nꢀheptane was investigated over
Pd/SiO2 and PdAu/SiO2 catalysts at atmospheric pressure in a fixedꢀbed microꢀreactor.
Pd/SiO2 catalysts prepared by incipient wetness impregnation and ion exchange were
used to elucidate particleꢀsize effects on reaction pathways, kinetics and catalyst
deactivation. The monometallic catalysts were characterized by CO chemisorption,
diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS), transmission
electron microscopy (TEM), and temperatureꢀprogrammed hydride decomposition
5
ACS Paragon Plus Environment

ACS Catalysis
Page 6 of 39
1
2
3
4
5
6
7
8
9
(TPHD). PdAu/SiO2 catalysts were prepared using an ion exchange procedure known to
produce supported PdAu alloy nanoparticles.23 In addition to the above techniques, the
bimetallic catalysts were characterized by analytical TEM with energy dispersive xꢀray
(EDX) analysis and in situ extended xꢀray absorption fine structure (EXAFS)
spectroscopy. Alloying Pd with Au is shown to improve catalyst stability without
sacrificing OA deoxygenation activity.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2. Experimental Methods
2.1 Materials
OA (98+%) was purchased from SigmaꢀAldrich and used as received. Ultraꢀhighꢀ
purity (99.999%) He and H2 were purchased from National Welders. Pd(NO3)2·xH2O and
[Pd(NH3)4][NO3]2 were purchased from Strem Chemicals. [Au(en)2]Cl3 (en = ethyleneꢀ
diamine) was prepared from HAuCl4 (SigmaꢀAldrich), as described in the literature.24
The silica support was Aerosil 300 from Evonik.
2.2 Catalyst preparation
Pd/SiO2 catalysts were prepared by incipient wetness (IW) impregnation and ion
exchange (IE). The Pd/SiO2 (IW) catalyst were prepared by impregnation of silica with
an aqueous Pd(NO3)2 solution of appropriate volume to form a thick paste. Specifically,
Pd(NO3)2·xH2O was dissolved in 19 mL of deionized (DI) water and added slowly to 4 g
of silica. The resultant paste was dried at 40ºC overnight and then calcined at 300 ºC in
flowing O2 for 2 h. The Pd/SiO2 (IE) catalyst were prepared by contacting the silica
support with a dilute aqueous solution of [Pd(NH3)4][NO3]2, as described previously.25
After 24 h, the solid was recovered by filtration and washed 3 times with DI water. The
resultant paste was dried at 40 ºC overnight and calcined at 250 ºC in flowing O2 for 2 h.
6
ACS Paragon Plus Environment

Page 7 of 39
ACS Catalysis
1
2
3
4
5
6
7
8
9
The bimetallic catalyst PdAu/SiO2 was prepared by ion exchange of [Pd(NH3)4][NO3]2
and [Au(en)2]Cl3 with silica from a pH 9 slurry.26 After filtration and washing with DI
water, the resultant paste was dried at 100 ºC in air overnight, resulting in a purplishꢀ
brown powder. The Pd and Au loadings were determined by Galbraith Laboratories using
inductively coupled plasmaꢀoptical emission spectrometry (ICPꢀOES).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2.3 Catalyst characterization
Pd dispersion was measured by volumetric CO chemisorption using a Micromeritics
ASAP 2020c instrument. Prior to chemisorption measurements, the samples were
reduced in H2 at 300ºC for 2 h followed by evacuation at 300°C for 4 h. Two CO
adsorption isotherms were recorded at 35ºC: an initial isotherm and another after
subsequent evacuation of the sample at 35ºC for 30 min. The Pd dispersion was
calculated based on the difference between the two isotherms (representing strongly
adsorbed CO). A CO to surface Pd atom stoichiometry of 0.6 was used in dispersion
calculations.27 Equivalent particle diameter (d) was calculated from the fractional
dispersion (D) using the equation below: 28
1.12
(3)
d =
D
TPHD measurements were performed using a Micromeritics AutoChem II 2920
instrument. The catalyst was heated at 5°C/min to 300°C in 5% H2/Ar and held for 1 h.
After cooling to ~40°C in flowing 5% H2/Ar, H2 evolution during a second temperature
ramp to 300°C was monitored using a thermal conductivity detector (TCD).
DRIFT spectra of adsorbed CO were measured using a Bruker Vertex 70 FTIR
spectrometer equipped with a Harrick Praying Mantis diffuseꢀreflectance accessory and a
highꢀtemperature in situ DRIFTS cell. Samples were reduced in the cell under flowing H2
7
ACS Paragon Plus Environment

ACS Catalysis
Page 8 of 39
1
2
3
4
5
6
7
8
9
at 300 ºC for 30 min, purged with He, and then cooled to room temperature under
flowing He. For each sample, 1ꢀmL pulses of 5% CO/He were used to saturate the
catalyst surface until no significant change in the DRIFT spectrum was observed. All
spectra consist of 128 scans collected at 4 cmꢀ1 resolution using a liquid N2ꢀcooled MCT
detector. Thermal desorption experiments were conducted in the DRIFTS cell after CO
chemisorption at 25°C. The catalyst temperature was increased incrementally from 25 to
260ºC under flowing He. After 15 min at each desorption temperature, to avoid thermal
artifacts the sample was cooled to 25°C, and the DRIFT spectrum was recorded.
Conventional transmission electron microscopy (TEM) and highꢀangle annular darkꢀ
field scanning TEM (HAADFꢀSTEM) were used to determine particle size distributions
and examine the atomic scale structure of the supported catalysts. Brightꢀfield TEM
images were acquired on a FEI Tecnai G² Twin instrument with 0.3ꢀnm resolution using
a 200 kV acceleration voltage (Duke University). HAADFꢀSTEM imaging of the Pd/SiO2
(IW) catalyst was done using a JEOL 2010F microscope operating at 200 kV. An FEI
Titan STEM with EDX capability was used to characterize the elemental composition of
the PdAu nanoparticles.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Pd K and Au L3 edge xꢀray absorption spectroscopy (XAS) of the PdAu/SiO2 (1:1)
catalyst was conducted on beam line Xꢀ11A of the National Synchrotron Light Source at
Brookhaven National Laboratory. The storage ring operated at 2.58 GeV with a current
of 150ꢀ275 mA. The Si(311) monochromator was calibrated using Au and Pd foils. The
monochromator was detuned by 30% at the Au L3 edge to minimize higher harmonics.
Measurements were made in transmission mode: For Au L3 edge XAS, the I0 ionization
chamber contained flowing N2 and the I chamber contained flowing Ar. For Pd edge
8
ACS Paragon Plus Environment

Page 9 of 39
ACS Catalysis
1
2
3
4
5
6
7
8
9
XAS, the I0 and I chambers contained flowing Ar. A sample (~250 mg) was pressed into
a rectangular slot in a stainless steel sample holder that was then mounted in an in situ
XAS cell.29 The sample was reduced at 300oC in flowing 5% H2/He for 1 h, purged
briefly with He, and cooled in flowing He to 25°C. Au L3 (3 scans) and Pd K (5 scans)
edge spectra were recorded with the sample under He.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
XAS data were processed to isolate EXAFS spectra from the background using
Athena software.30 Replicate scans were averaged to increase the signalꢀtoꢀnoise ratios.
The resulting χ spectra (k space) were Fourier transformed (Pd K: ꢁk = 2.4–12.2 Åꢀ1, Au
L3: ꢁk = 2.5–12.7 Åꢀ1) and fitted in R space (Pd K: ꢁR = 1.4–3.3 Å, Au L3: ꢁR = 1.4–3.4
Å) using Artemis software.31 PdꢀPd, PdꢀAu, AuꢀAu, and AuꢀPd singleꢀscattering paths
were generated using FEFF.32 Amplitude reduction factors, S02 , for the PdꢀPd and AuꢀAu
paths, 0.76 and 0.80, were determined by fitting the EXAFS of Pd and Au foils,
respectively. The reported EXAFS parameters: N (coordination number), R (interatomic
distance), σ2 (DebyeꢀWaller factor), and ꢀE0 (inner potential shift) are based on
simultaneous fits of knꢀweighted spectra (n = 1, 2, and 3).
2.4 OA deoxygenation experiments
OA deoxygenation experiments were conducted in a 0.5ꢀin. stainless steel tubular
reactor using 0.5 g catalyst powder suspended between quartz wool plugs. Catalyst
temperature was measured using a Type K thermocouple in direct contact with the bed.
Gas flow rate and composition to the reactor were set by mass flow controllers (Brooks),
and OA was fed via highꢀpressure syringe pump (ISCO Serial D) at 4 ꢂL/min (typical
flow rate). The standard in situ catalyst pretreatment procedure was heating at 300 ºC for
9
ACS Paragon Plus Environment

ACS Catalysis
Page 10 of 39
1
2
3
4
5
6
7
8
9
1 h in flowing H2. Typically, the catalyst was brought on stream at 260ºC in flowing 10%
H2/He or H2 (100 sccm), and the reactions conditions were maintained for 8ꢀ10 h.
The reactor effluent was analyzed using dual onꢀline gas chromatographs (GCs). A
Shimadzu GCꢀ2010 equipped with a flame ionization detector (FID) was used to measure
the reactant and product concentrations in the reactor effluent. Periodic sampling was
achieved by using a heated Valco 6ꢀport injection valve with a 100ꢀꢂL sample loop. An
Econocap ECꢀ1 capillary column was used, and the GC oven temperature was
programmed as follows: 10ºC /min ramp from 70 to 100ºC, then an increase to 240ºC at
40ºC/min. FID responses were calibrated using a series of different concentrations of OA,
nꢀheptane, 1ꢀheptene, octanal, 1ꢀoctanol, heptanal, and diheptylketone in suitable
solvents. An SRI GC equipped with a TCD and a Restek ShinCarbon ST 100 column was
used for monitoring CO and CO2 concentrations. The GCꢀTCD was operated
isothermally at 100°C. A standard calibration mixture containing 5% CO, 5% CO2, 5%
H2 and 1% CH4 standard calibration gas (National Welders) was used to calibrate the
GCꢀTCD for CO, CO2 and CH4 response factors.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OA deoxygenation rates at 260°C in 10% H2/He were determined from nꢀheptane
yields measured under differential conversion conditions. Differential conversion (≤10%)
of OA was achieved by increasing the OA injection rate and the feed flow rate of 10%
H2/He proportionally to increase the weight hourly space velocity (WHSV) at constant
OA concentration. In some instances, it was necessary to reduce the catalyst mass by
diluting the bed with inert quartz chips. Small concentrations (<1% each) of impurities
and/or thermal reaction products, diheptylketone, heptanal, and octanal, were detected by
GCꢀFID in blank (empty tube) experiments with OA (4 ꢂL/min) in 10% H2 at 260°C.
10
ACS Paragon Plus Environment

Page 11 of 39
ACS Catalysis
1
2
3
4
5
6
The absence of external and internal diffusional limitations was confirmed using the
Mears and WeiszꢀPrater criteria, respectively.33
7
8
9
3. Results and Discussion
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.1. Characterization of fresh catalysts
ICPꢀOES and CO chemisorption results for the Pd/SiO2 and PdAu/SiO2 catalysts are
given in Table 1. As expected, the Pd/SiO2 (IE) catalyst has a much higher Pd dispersion
than the Pd/SiO2 (IW) catalyst. The calculated dispersion of the Pd/SiO2 (IE) catalyst
exceeds 100% indicating that the CO/Pd stoichiometry for these particles is greater than
0.6. Applying a CO/Pd stoichiometry of ~0.8, as suggested by Fagherazzi, et al.34 based
on smallꢀangle xꢀray scattering measurements on 2.0ꢀnm silicaꢀsupported Pd particles,
would result in a calculated Pd dispersion of 100%. The average particle sizes estimated
from CO chemisorption are ~1 and ~7 nm, for the Pd/SiO2 (IE) and Pd/SiO2 (IW)
catalysts, respectively. The Pd dispersions of the PdAu bimetallic catalysts are equivalent
and lie between those of the Pd monometallic catalysts. The Pd/SiO2 (IE) and PdAu/SiO2
catalysts have comparable Pd loadings but very different surface Pd atom densities. The
estimated size of the alloy particles in the PdAu/SiO2 catalysts is ~3.5 nm; however, this
estimate assumes negligible irreversible CO adsorption on Au and a uniform Pd
distribution in the particles, i.e., no Pd surface enrichment.
TEM images of the Pd/SiO2 (IW), Pd/SiO2 (IE) and PdAu/SiO2 (1:1) catalysts are
shown in Figure 1. The Pd/SiO2 (IW) catalyst (Fig. 1a) has an average particle size of 7.5
nm with a broad size distribution (Fig. 2a) that is skewed toward larger sizes. A darkꢀ
field HAADFꢀSTEM image (Fig 1b) illustrates that individual nanoparticles in this
catalyst exhibit wellꢀdefined lowꢀindex facets. The Pd/SiO2 (IE) catalyst contains 1ꢀ2 nm
11
ACS Paragon Plus Environment

ACS Catalysis
Page 12 of 39
1
2
3
4
5
6
7
8
9
Pd particles, seen as bright spots in the darkꢀfield STEM image (Fig. 1c). The average
particle size is 1.5 nm with a narrow size distribution (Fig 2b). The average particle sizes
from TEM agree well with those determined by CO chemisorption for the monometallic
catalysts. The nanoparticles in the PdAu/SiO2 (1:1) catalyst (Fig. 1d) have a 5ꢀnm
average diameter with a relatively broad size distribution (Fig. 2c). The average particle
size from TEM is significantly larger than from CO chemisorption, consistent with Pd
surface enrichment (at least in the presence of adsorbed CO).27 The alloy surface
composition estimated from the chemisorption data and the TEM average particle size is
~70% Pd. The EDX maps (Fig. 3) of an individual nanoparticle from the PdAu/SiO2 (1:1)
catalyst provide direct evidence of alloying; however, there are more Pd atoms than Au
atoms near the nanoparticle surface (based on the larger diameter in the Pd pixel map).
Because the EDX measurements were made ex situ, we could not ascertain if the
observed Pd surface enrichment was induced by air exposure.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The PdAu/SiO2 (1:1) catalyst was analyzed by XAS at the Pd K and Au L3 edges.
The EXAFS spectra were fit in R space using 3 adjustable parameters for each scattering
path: PdꢀAu, PdꢀPd, AuꢀAu and AuꢀPd (where the absorber atom is listed first and the
backscatterer second). The inner potential corrections (ꢀE0 ) for each path for a given
absorber were fixed at the values in Table 2. When the Pd K and Au L3 EXAFS spectra
were fit independently, the results (Table 2) are in reasonable agreement with those
reported by Reifsnyder and Lamb, who analyzed the EXAFS spectra of similarly
prepared PdAu/SiO2 catalysts using a combination of empirical and theoretical
references.26 Moreover, the results conform to the consistency criteria originally proposed
by Via, et al.35 Strong interference between the EXAFS signals from homometallic and
12
ACS Paragon Plus Environment

Page 13 of 39
ACS Catalysis
1
2
3
4
5
6
7
8
9
heterometallic scattering paths is apparent in each spectrum (Fig. 4). The highꢀquality fits
of the EXAFS spectra illustrated in Fig. 4 are confirmed by the small Rꢀfactors in Table
2; an Rꢀfactor less than 0.05 indicates a good fit. The heterometallic (PdꢀAu and AuꢀPd)
distances are equivalent within experimental uncertainty but shorter than the Vegard’s
Law prediction for a 1:1 PdAu alloy (2.815 Å). The PdꢀPd distance is equivalent to the
PdꢀAu distance; however, the AuꢀAu distance is 0.03 Å longer than the AuꢀPd distance.
The total coordination number for Au (11.1) is greater than for Pd (8.8) indicating that Pd
atoms preferentially occupy surface sites (and conversely, Au atoms are in a more bulkꢀ
like environment). The larger AuꢀAu coordination number (relative to AuꢀPd) and longer
AuꢀAu distance evidence that the supported alloy nanoparticles have Auꢀrich cores. Thus,
the Pd surface enrichment indicated by CO chemisorption and EDX analysis is confirmed
by in situ EXAFS spectroscopy.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Bulk hydride formation and decomposition have been used to assess the degree of
alloying in Pdꢀcontaining bimetallic catalysts.36ꢀ37 For relatively large supported Pd
particles (~10 nm) previously saturated with hydrogen, a βꢀPd hydride decomposition
peak is observed at 70ꢀ90°C in flowing 10% H2.37ꢀ38 For smaller Pd particles, the higher
surfaceꢀtoꢀvolume results in a lower H/Pd ratio39 and a lower hydride decomposition
temperature.40ꢀ41 For PdAu alloys containing 50 at.% Au or more, the βꢀPd hydride peak
is greatly diminished (or completely absent) due to the effects of alloying on bulk
electronic structure.36, 42 As shown in Figure 5, the Pd/SiO2 (IW) catalyst exhibits an
intense wellꢀdefined TPHD peak at 55°C in flowing 5% H2/Ar. The measured H/Pd ratio
(0.65) agrees well with the H/Pd ratio for bulk βꢀPd hydride.39, 42 The Pd/SiO2 (IE)
catalyst displays a much smaller TPHD peak at 50°C that is broad and skewed toward
13
ACS Paragon Plus Environment

ACS Catalysis
Page 14 of 39
1
2
3
4
5
6
7
8
9
higher temperatures. The measured H/Pd ratio (0.16) is consistent with highly dispersed
supported Pd particles.39, 41, 43 The TPHD spectrum of PdAu/SiO2 (1:1) does not contain
any βꢀPd hydride decomposition features indicating the absence of segregated Pd
particles in this catalyst. Apparently, the combination of relatively small particle size (5
nm) and alloying with Au is sufficient to completely suppress bulk hydride formation.
CO DRIFTS was used to probe the adsorption sites of the Pd/SiO2 and PdAu/SiO2
catalysts. The spectra of the Pd/SiO2 catalysts (Fig. 6) exhibit COꢀstretching bands in the
2000ꢀ2100 and 1800ꢀ2000 cmꢀ1 regions corresponding to linear (atop) and bridging
(multiꢀcoordinated) CO species, respectively 44. The DRIFT spectrum of the Pd/SiO2
(IW) catalyst (Fig. 6a) contains a strong twoꢀfold bridging CO peak at 1990 cmꢀ1 and a
weak linear CO peak at ~2090 cmꢀ1. When considered with the HAADFꢀSTEM image in
Fig. 1b, this spectrum establishes that the Pd/SiO2 (IW) catalyst contains supported Pd
nanoparticles with wellꢀdefined (100) and (111) facets.45ꢀ46 The unusually sharp and
narrow band at 1990 cmꢀ1 in Fig. 6(a) may be assigned to CO adsorbed on (100) facets
and CO bridge bonded to particle steps and edges, with significant “intensity borrowing”
from bridging CO on (111) facets, as discussed in Lear, et al. and references therein.46
The CO DRIFT spectrum of our Pd/SiO2 (IW) catalyst is remarkably similar to that of a
7.3 wt.% Pd/Al2O3 catalyst prepared from tetraamminepalladium (II) tetraazide (Fig. 4 of
Ref. 46). In contrast, the Pd/SiO2 (IE) catalyst exhibits strong linear and twoꢀfold
bridging CO peaks at 2100 and 1930 cmꢀ1, respectively (Fig. 6b). Previous research has
shown that small supported Pd particles having a higher fraction of edge and corner sites
that accommodate atop CO species.47 Moreover, the twoꢀfold bridging CO peak shifts to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
14
ACS Paragon Plus Environment

Page 15 of 39
ACS Catalysis
1
2
3
4
5
6
lower frequency because the ratio of (111) to (100) facets increases with decreasing
particle size.48ꢀ49
7
8
9
DRIFT spectra of the PdAu/SiO2 catalysts (Fig. 7) exhibit two strong peaks at 2070
and 1960 cmꢀ1 that are assigned to linear and twoꢀfold bridging CO species; however,
these features are much less intense than those of the Pd/SiO2 catalysts because of the
lower surface Pd atom densities. A shoulder at 2105 cmꢀ1 in Fig. 7a is assigned to linearly
adsorbed CO atop Au atoms,50 and its intensity increases with Au content (Fig. 7b). In
addition, the linearꢀtoꢀbridging intensity ratio increases,23, 51 and the CO peaks narrow
with increasing Au content. The νCO of atop CO on Pd sites at saturation coverage is
shifted downward by ~30 cmꢀ1 relative to atop CO on 1ꢀnm Pd particles. Goodman and
coworkers assigned peaks at 2060ꢀ2085 cmꢀ1 to atop CO on isolated Pd sites of PdAu
alloy films and AuPd (100) single crystals.52 The νCO of the twoꢀfold bridging species lies
between those assigned to bridging CO on Pd(100) and Pd(111) surfaces at saturation
coverage and indicates the presence of vicinal Pd atoms on the alloy surface.27 Goodman
and coworkers assigned a peak at 1958 cmꢀ1 to CO adsorbed on contiguous Pd surface
sites of an ionꢀsputtered PdAu (100) single crystal. A band at ~1970 cmꢀ1 was assigned to
CO on contiguous Pd sites created by COꢀinduced segregation of Pd atoms to the
surface.52
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To assess CO adsorption energies, stepwise thermal desorption experiments were
conducted in flowing He beginning with catalysts saturated with CO at 25°C. As
illustrated by the CO DRIFT spectrum of Pd/SiO2 (IE), linear CO desorbs at lower
temperatures and more completely than bridging CO indicating a lower heat of
adsorption (adsorption energy). Concomitant with CO desorption (reducing surface
15
ACS Paragon Plus Environment

ACS Catalysis
Page 16 of 39
1
2
3
4
5
6
7
8
9
coverage), the linear and bridging peaks shift to lower wavenumbers consistent with
reduction in dipolar coupling between adjacent adsorbed CO molecules.53 Qualitatively
similar thermal desorption behavior is shown for the Pd/SiO2 (IW) catalyst in Figure 6a.
For both monometallic catalysts, a significant fraction of the adsorbed CO remains in
bridging sites after thermal desorption at 260°C. Moreover, the position and broad shape
of the band after 260°C desorption indicates a mixture of CO in twoꢀfold and threeꢀfold
bridging sites. The thermal desorption behavior of the PdAu/SiO2 catalysts (Fig. 7) is
different in several aspects. The atop and twoꢀfold bridging CO peak positions do not
shift appreciably during thermal desorption. In addition, the fractional surface coverage
by bridging CO after thermal desorption at 260°C is less than for the monometallic
samples, especially for the PdAu/SiO2 (2:3) catalyst. These observations are consistent
with the expected geometric and electronic effects of alloying Pd with Au. Alloying
creates Pd surface ensembles that are diluted by Au and, consequently, less affected by
dipolar coupling. Moreover, alloying with Au weakens the interaction between CO and
surface Pd atoms, consistent with electron transfer from Au to the Pd d band.21, 54ꢀ55
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.2. OA deoxygenation catalysis
The supported Pd catalysts were evaluated for OA deoxygenation under 10% and
100% H2 at 260°C and 1 atm. Under 10% H2, the Pd/SiO2 (IW) catalyst gave an initial
OA conversion of ~55% and deactivated to a stable conversion of ~40% after 4ꢀ5 h time
on stream (TOS) (Fig. 8). Selectivity to nꢀheptane was >93%; minor products included
octanal, 1ꢀoctanol, and heptanal. CO yield and OA conversion with TOS agree closely
indicating that decarbonylation is the dominant reaction pathway. The nꢀheptane product
may result either from rapid hydrogenation of heptene(s) formed via Eq. 2 or from
16
ACS Paragon Plus Environment

Page 17 of 39
ACS Catalysis
1
2
decarbonylation of octanal resulting from OA hydrogenation.56 The water gas shift
(WGS) reaction, in principle, can disguise the deoxygenation pathway; however, the
equilibrium constant for WGS is 71.4 at 260°C.57 Under 10% H2, the equilibrium CO2
selectivity in our OA deoxygenation experiments would be approximately 50%; however,
CO2 was not detected as a product indicating negligible WGS activity for this catalyst.
Under 100% H2, initial OA conversion was ~100% (Fig. 8); however, the catalyst
deactivation rate was significantly higher, and stable activity was not achieved even after
>8 h TOS. In this case, CO yield with TOS was ~3% lower than OA conversion;
selectivity to nꢀheptane was >95%. Higher H2 partial pressure should favor OA
hydrogenation to octanal that subsequently undergoes decarbonylation to nꢀheptane.56
The rapid deactivation of the Pd/SiO2 (IW) catalyst under 100% H2 may be related to its
increased decarbonylation activity. Endogenous CO will block active sites if it is not
desorbed completely (irreversibly) under reaction conditions.11 As evidenced by
DRIFTS, CO does not desorb completely from this catalyst at 260°C in flowing He, and
CO poisoning may be responsible for the observed catalyst deactivation.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
OA deoxygenation results for the Pd/SiO2 (IE) catalyst are shown in Figure 9. Initial
deoxygenation activity was lower than for the Pd/SiO2 (IW) catalyst despite the more
than twoꢀfold greater surface Pd atom density. Under 10% H2, Pd/SiO2 (IE) gave an
initial OA conversion of 35ꢀ40% and exhibited both decarbonylation and decarboxylation
activity; however, the initial decarboxylation activity (as measured by CO2 yield)
decayed to essentially zero after 5 h TOS. Good agreement is observed between the
nꢀheptane yield and the CO + CO2 yield (Fig. 9) demonstrating closure of the material
balance. After 5 h TOS stream, the catalyst approached a stable conversion of ~15% with
17
ACS Paragon Plus Environment

ACS Catalysis
Page 18 of 39
1
2
3
4
5
6
7
8
9
90% nꢀheptane selectivity. Octanal, heptanal, and diheptylketone were minor products.
Diheptylketone was detected only at low conversions (<20%), and it represented <1% of
the reaction products. The Pd/SiO2 (IE) catalyst exhibited similar initial activity under
100% H2 (Fig. 9). Unfortunately, temporal GCꢀTCD data are not available for this
experiment; however, it seems safe to assume that CO was the main C1 product under
these conditions. Selectivity to nꢀheptane was 90ꢀ95%. In contrast to the Pd/SiO2 (IW)
catalyst, the Pd/SiO2 (IE) catalyst deactivated more slowly with TOS under 100% H2. In
this case, the higher H2 partial pressure may be more effective at preventing selfꢀ
poisoning by strongly adsorbed (i.e., fully or partially dehydrogenated) intermediates on
the small 1ꢀ2 nm Pd particles or inhibiting polymerization of heptenes. Theoretical
investigations of the deoxygenation of propanoic and butanoic acids on Pd(111) have
revealed the importance of dehydrogenated surface intermediates in the decarboxylation
and decarbonylation pathways.58ꢀ59
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Under 10% H2, the PdAu/SiO2 (2:3) catalyst achieved an initial OA conversion of
~45% despite its lower Pd surface atom density, and it did not deactivate significantly
with TOS (Fig. 10). Stable OA conversion was sustained for more than 8 h with >97%
selectivity to nꢀheptane. CO was the only C1 product. Performance of the PdAu/SiO2
(1:1) catalyst was closely similar under these reaction conditions. Under 100% H2, the
PdAu/SiO2 (2:3) catalyst had modestly lower activity and selectivity to nꢀheptane.
Several minor products, including octanal, 1ꢀoctanol, and diheptylketone, were favored
under 100% H2. A 1.9 wt.% Au/SiO2 catalyst was prepared by depositing 4.3ꢀnm
colloidal Au particles onto Aerosil 300, as described by Wilson, et al.60 After calcination
at 300°C in air and in situ reduction at 300°C in H2, the Au/SiO2 catalyst had no
18
ACS Paragon Plus Environment

Page 19 of 39
ACS Catalysis
1
2
3
4
5
6
7
8
9
appreciable activity for OA deoxygenation at 260°C (4 ꢂL/min under 10% H2). OA
conversion was less than 5%, and the major products are consistent with either silicaꢀ
catalyzed reactions (diheptylketone) or thermal decomposition. As the PdAu/SiO2
catalysts have much lower Pd loadings (and Pd surface atom densities) than the Pd/SiO2
catalysts, clearly, alloying Pd with Au improves catalyst performance. The improved
stability may result from a decrease in CO adsorption energy. The order of residual CO
coverage at 260ºC (determined from DRIFTS) is as follows: Pd/SiO2 (IE) ≈ PdSiO2 (IW)
> PdAu/SiO2 (1:1) > PdAu/SiO2 (2:3), and this order correlates well with catalyst
deactivation behavior. Under 100% H2, the PdAu/SiO2 (2:3) catalyst (Fig. 10) had
modestly lower activity (40% OA conversion) and nꢀheptane selectivity (>93%);
however, activity declined only very slowly with TOS.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.3. Characterization of used catalysts
The dispersion of the used Pd/SiO2 (IW) catalyst after postꢀreaction treatment in H2
at 300°C is equivalent to that of the fresh catalyst (Table 1). In contrast, chemisorption
measurements on the used Pd/SiO2 (IE) catalyst (after an equivalent postꢀreaction
treatment) indicate a ~70ꢀ80% decrease in dispersion. TEM imaging of the used Pd/SiO2
(IE) catalyst did not evidence particle sintering under reaction conditions, leading to the
inference that irreversibly adsorbed OA (or other organic residues) is responsible for loss
of Pd active sites.61
DRIFT spectra of the used catalysts (Fig. 11) contain new features that are assigned
to adsorbed OA species. The spectra of the used catalysts display peaks near 2900 cmꢀ1
and 1700 cmꢀ1 that may be assigned to alkyl CꢀH and C=O stretching frequencies,
respectively. The observed peaks are consistent with IR spectrum of OA;62 however, the
19
ACS Paragon Plus Environment

ACS Catalysis
Page 20 of 39
1
2
C=O stretching frequency is red shifted by ~65 cmꢀ1 relative to gasꢀphase OA, consistent
with either liquidꢀphase or adsorbed OA. Subsequently, the used catalysts were treated in
flowing H2 at 300 ºC for 1 h and then purged with He. As shown in Fig. 11, OA was
removed from the Pd/SiO2 (IW) and PdAu/SiO2 (2:3) catalysts, but remained on the
Pd/SiO2 (IE) catalyst. Thus, DRIFTS analysis indicates that OA is adsorbed more
strongly on the Pd/SiO2 (IE) catalyst that contains smaller (1ꢀ2 nm) Pd nanoparticles.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3.4. TOF measurements
Specific rates of OA decarbonylation over the Pd/SiO2 (IW) and PdAu/SiO2 (2:3)
catalysts were measured at differential conversion by increasing WHSV at constant OA
concentration. Steadyꢀstate TOFs were calculated using the measured rates and the
surface Pd atom densities of the fresh catalysts from CO chemisorption. The TOFs for
gasꢀphase OA deoxygenation over Pd/SiO2 (IW) and PdAu/SiO2 (2:3) under 10% H2 at
260°C and 1 atm are 4.1 × 10ꢀ2 and 4.3 × 10ꢀ2 sꢀ1, respectively. The equivalence of these
TOFs suggests that OA decarbonylation requires only a relatively small ensemble of Pd
atoms. These TOFs are compared in Table 3 to previously reported values for
deoxygenation of aliphatic carboxylic acids in the gas phase over Pd/SiO2 catalysts.
Meaningful comparisons are difficult owing to the range of reaction conditions employed.
The reported TOFs for gasꢀphase deoxygenation of propanoic acid to ethane and
propionaldehyde over Pd/SiO2 catalysts are nearly an order of magnitude lower than our
OA results; however, the reaction conditions are not directly comparable (lower
temperature and higher H2 partial pressure).15ꢀ16 A similarly low TOF was reported for
propanoic acid decarbonylation over a 0.5 wt.% Pd/SiO2 at 250°C in H2; however, the
20
ACS Paragon Plus Environment

Page 21 of 39
ACS Catalysis
1
2
3
4
5
6
product was ethene not ethane. The higher TOFs for OA (when compared to propanoic
acid) may be explained (at least in part) by its longer alkyl chain.12
7
8
9
4. Conclusions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Pd/SiO2 catalysts are active and highly selective for OA decarbonylation to
nꢀheptane at 260°C under H2; however, they deactivate significantly with TOS, and the
deactivation behavior is related to Pd particle size. A highly dispersed Pd/SiO2 catalyst
containing 1ꢀ2 nm Pd particles deactivates rapidly and irreversibly under 10% H2
apparently due to selfꢀpoisoning by OA and strongly adsorbed (i.e., fully or partially
dehydrogenated) intermediates. TEM images of the used catalyst show negligible change
in Pd particle size, ruling out sintering as a key reason for deactivation. Moreover, this
catalyst deactivates much more slowly under 100% H2, suggesting that hydrogen may
inhibit the accumulation of unsaturated species on the catalyst surface. A Pd/SiO2 catalyst
containing much larger Pd nanoparticles deactivates less rapidly with TOS in 10% H2 and
can be regenerated by heating in H2 to remove OA residues. Higher H2 partial pressure
substantially increases the initial decarbonylation rate, but this catalyst deactivates more
rapidly under 100% H2, apparently due to CO poisoning. Alloying Pd with Au was found
to improve catalyst stability without significantly affecting decarbonylation activity.
Thermal desorption experiments demonstrate a reduction in the CO adsorption energy on
Pd surface sites of the alloy nanoparticles. The equivalence of OA TOFs over Pd/SiO2
(IW) and PdAu/SiO2 (2:3) suggests that decarbonylation requires only a relatively small
ensemble of Pd atoms.
21
ACS Paragon Plus Environment

ACS Catalysis
Page 22 of 39
1
2
3
4
Acknowledgments
5
6
7
8
This work was supported by the National Science Foundation under an EFRI HyBi grant
(EFRIꢀ0937721). Xꢀray absorption measurements were conducted using beam line Xꢀ
11A of the National Synchrotron Light Source at Brookhaven National Laboratory. The
authors gratefully acknowledge the assistance of Dr. Xianqin Wang (New Jersey Institute
of Technology) for her assistance.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
References
(1)
(2)
(3)
(4)
Smith, B.; Greenwell, H. C.; Whiting, A. Energy Environ Sci 2009, 2, 262ꢀ271.
Knothe, G. Fuel Process Technol 2005, 86, 1059ꢀ1070.
Chisti, Y. Biotechnol Adv 2007, 25, 294ꢀ306.
Kubičková, I.; Snåre, M.; Eränen, K.; MäkiꢀArvela, P.; Murzin, D. Y. Catal
Today 2005, 106, 197ꢀ200.
(5)
Snåre, M.; Kubičková, I.; MäkiꢀArvela, P.; Eränen, K.; Murzin, D. Y. Ind. Eng.
Chem. Res. 2006, 45, 5708ꢀ5715.
(6)
(7)
Immer, J. G.; Kelly, M. J.; Lamb, H. H. Appl Catal, A 2010, 375, 134ꢀ139.
MäkiꢀArvela, P.; Kubickova, I.; Snåre, M.; Eränen, K.; Murzin, D. Y. Energy
Fuels 2007, 21, 30ꢀ41.
(8)
(9)
Snåre, M.; Kubičková, I.; MäkiꢀArvela, P.; Chichova, D.; Eränen, K.; Murzin, D.
Y. Fuel 2008, 87, 933ꢀ945.
Simakova, I.; Simakova, O.; MäkiꢀArvela, P.; Simakov, A.; Estrada, M.; Murzin,
D. Y. Appl Catal, A 2009, 355, 100ꢀ108.
(10) Lestari, S.; MäkiꢀArvela, P.; Bernas, H.; Simakova, O.; Sjöholm, R.; Beltramini,
J.; Lu, G. M.; Myllyoja, J.; Simakova, I.; Murzin, D. Y. Energy Fuels 2009, 23,
3842ꢀ3845.
(11) Immer, J. G.; Lamb, H. H. Energy Fuels 2010, 24, 5291ꢀ5299.
(12) Ford, J. P.; Immer, J. G.; Lamb, H. H. Top. Catal. 2012, 55, 175ꢀ184.
(13) Maier, W. F.; Roth, W.; Thies, I.; Schleyer, P. V. R. Chem. Ber. 1982, 115, 808ꢀ
812.
(14) Boda, L.; Onyestyák, G.; Solt, H.; Lónyi, F.; Valyon, J.; Thernesz, A. Appl.
Catal., A 2010, 374, 158ꢀ169.
(15) LugoꢀJosé, Y. K.; Monnier, J. R.; Williams, C. T. Appl. Catal., A 2014, 469, 410ꢀ
418.
(16) Alotaibi, M. A.; Kozhevnikova, E. F.; Kozhevnikov, I. V. Appl. Catal., A 2012,
447, 32ꢀ40
(17) Coq, B.; Figueras, F. J. Mol. Catal. A 2001, 173, 117ꢀ134.
(18) Han, Y.F.; Wang, J. H.; Kumar, D.; Yan, Z.; Goodman, D. J. Catal. 2005, 232,
467ꢀ475.
(19) Xu, J.; White, T.; Li, P.; He, C.; Yu, J.; Yuan, W.; Han, Y. F. J. Am. Chem. Soc.
2010, 132, 10398ꢀ10406.
22
ACS Paragon Plus Environment

Page 23 of 39
ACS Catalysis
1
2
3
4
5
(20) Huang, Y.; Zhou, X.; Yin, M.; Liu, C.; Xing, W. Chem. Mater. 2010, 22, 5122ꢀ
5128.
(21) Gao, F.; Goodman, D. W. Chem. Soc. Rev. 2012, 41, 8009ꢀ8020.
(22) Neurock, M. Forment, G. F.; Waugh, K. Dynamics of Surfaces and Reaction
Kinetics in Heterogeneous Catalysis; Elsevier: 1997; p 23ꢀ24
(23) Lam, Y.; Boudart, M. J. Catal. 1977, 50, 530ꢀ540.
(24) Block, B.; Bailar Jr, J. C. J. Am. Chem. Soc. 1951, 73, 4722ꢀ4725.
(25) Reifsnyder, S. N.; Otten, M. M.; Sayers, D. E.; Lamb, H. H. J. Phys. Chem. B
1997, 101, 4972ꢀ4977.
(26) Reifsnyder, S. N.; Lamb, H. H. J. Phys. Chem. B 1999, 103, 321ꢀ329.
(27) Kunz, S.; Iglesia, E. J. Phys. Chem. C 2014, 118 , 7468ꢀ7479.
(28) Xu, X.; Szanyi, J.; Xu, Q.; Goodman, D. W. Catal. Today 1994, 21, 57ꢀ69.
(29) Kampers, F.; Maas, T.; Van Grondelle, J.; Brinkgreve, P.; Koningsberger, D. Rev.
Sci. Instrum. 1989, 60, 2635ꢀ2638.
(30) Ravel, B.; Newville, M. J. Synchrotron Radiat. 2005, 12, 537ꢀ541.
(31) Newville, M. J. Synchrotron Radiat. 2001, 8, 322ꢀ324.
(32) Rehr, J. J.; Albers, R. Rev. Mod. Phys. 2000, 72, 621.
(33) Fogler, H. S. Elements of Chemical Reaction Engineering, 3rd Ed,; Prentice Hall,
Upper Saddle River, New Jersey, 1999; p 761.
(34) Fagherazzi, G.; Benedetti, A.; Polizzi, S.; Di Mario, A.; Pinna, F.; Signoretto, M.;
Pernicone, N. Catal. Lett. 1995, 32, 293ꢀ303.
(35) Via, G.; Drake Jr, K.; Meitzner, G.; Lytle, F.; Sinfelt, J. Catal. Lett. 1990, 5, 25ꢀ
33.
(36) Ziemecki, S.; Michel, J.; Jones, G. React. Solid. 1986, 2, 187ꢀ202.
(37) Bonarowska, M.; Pielaszek, J.; Juszczyk, W.; Karpiński, Z. J. Catal. 2000, 195,
304ꢀ315.
(38) Chang, T. C.; Chen, J. J.; Yeh, C. T. J. Catal. 1985, 96, 51ꢀ57.
(39) Boudart, M.; Hwang, H. J. Catal. 1975, 39, 44ꢀ52.
(40) Pinna, F.; Signoretto, M.; Strukul, G.; Polizzi, S.; Pernicone, N. React. Kinet.
Catal. Lett. 1997, 60, 9ꢀ13.
(41) Nag, N. K. J. Phys. Chem. B 2001, 105, 5945ꢀ5949.
(42) Luo, S.; Wang, D.; Flanagan, T. B. J. Phys. Chem. B 2010, 114, 6117ꢀ6125.
(43) Bonivardi, A. L.; Baltanás, M. A. J. Catal. 1992, 138, 500ꢀ517.
(44) Bradshaw, A.; Hoffmann, F. Surf. Sci. 1978, 72, 513ꢀ535.
(45) Kelly, M. J.; Kim, J.; Roberts, G. W.; Lamb, H. H. Top. Catal. 2008, 49, 178ꢀ186.
(46) Lear, T.; Marshall, R.; LopezꢀSanchez, J. A.; Jackson, S. D.; Klapötke, T. M.;
Bäumer, M.; Rupprechter, G.; Freund, H. J.; Lennon, D. J. Chem. Phys. 2005,
123, 174706.
(47) Giorgi, J. B.; Schroeder, T.; Bäumer, M.; Freund, H. J. Surf. Sci. 2002, 498, 71ꢀ
77.
(48) Szanyi, J.; Kuhn, W. K.; Goodman, D. W. J. Vac. Sci. Technol. A 1993, 11, 1969ꢀ
1974.
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(49) Rieck, J. S.; Bell, A. T. J. Catal. 1987, 103, 46ꢀ54.
(50) Yi, C. W.; Luo, K.; Wei, T.; Goodman, D. J. Phys. Chem. B 2005, 109, 18535ꢀ
18540.
(51) Kugler, E.; Boudart, M. J Catal 1979, 59, 201ꢀ210.
23
ACS Paragon Plus Environment

ACS Catalysis
Page 24 of 39
1
2
3
4
5
6
7
8
9
(52) Gao, F.; Wang, Y.; Goodman, D. W. J. Phys. Chem. C 2009, 113, 14993ꢀ15000.
(53) Matolin, V.; Gillet, E. Surf. Sci. 1990, 238, 75ꢀ82.
(54) Ruff, M.; Takehiro, N.; Liu, P.; Nørskov, J. K.; Behm, R. J. ChemPhysChem
2007, 8, 2068ꢀ2071.
(55) Liu, P.; Nørskov, J. K. Phys. Chem. Chem. Phys. 2001, 3, 3814ꢀ3818.
(56) Rozmysłowicz, B.; MäkiꢀArvela, P.; Tokarev, A.; Leino, A. R.; Eränen, K.;
Murzin, D. Y. Ind. Eng. Chem. Res.2012, 51, 8922ꢀ8927.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(57) Bartholomew, C. H.; Farrauto, R. J. Fundamentals of industrial catalytic
processes. John Wiley & Sons: New York; 2011
.
(58) Lamb, H. H.; Sremaniak, L.; Whitten, J. L. Surf. Sci. 2012.607,130ꢀ137
(59) Lu, J.; Behtash, S.; Heyden, A. The Journal of Physical Chemistry C 2012, 116,
14328ꢀ14341.
(60) Wilson, A. R.; Sun, K.; Chi, M.; White, R. M.; LeBeau, J. M.; Lamb, H. H.;
Wiley, B. J. J. Phys. Chem. C 2013, 117, 17557ꢀ17566.
(61) Ping, E. W.; Pierson, J.; Wallace, R.; Miller, J. T.; Fuller, T. F.; Jones, C. W.
Appl. Catal., A 2011, 396, 85ꢀ90.
(62) Mallard, W.; Linstrom, P. NIST Standard Reference Database 2000, 69.
24
ACS Paragon Plus Environment

Page 25 of 39
ACS Catalysis
1
2
3
4
Table 1: Catalyst composition and CO chemisorption results.
5
6
7
8
9
Catalyst
Composition
Strongly
Pd
Particle size Particle size
adsorbed CO Dispersiona
(ꢀmol/g)
(nm)
(CO)
(nm)
(TEM)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1.78% Pd/SiO2 (IW)
0.63% Pd/SiO2 (IE)
15.7 ± 0.6
15.7%
(15.0%)
130%
(22.5%)
33.0%
7.1
1.1
3.4
3.4
7.5
1.5
5.0
−
(15.0 ± 0.5)b
45.9 ± 0.4
(8.0 ± 1.2)b
8.2 ± 0.4
0.44% Pd
0.91% Au/SiO2 (1:1)
0.42% Pd
1.2% Au/SiO2 (2:3)
7.8 ± 0.9
33.0%
aComputed using 0.6 CO per surface Pd atom and assuming negligible strongly adsorbed
CO on Au.
bUsed catalyst following ex situ reduction at 300°C in flowing H2.
25
ACS Paragon Plus Environment

ACS Catalysis
Page 26 of 39
1
2
3
4
Table 2: EXAFS spectroscopy results for PdAu/SiO2 (1:1) catalyst.
5
6
7
8
9
absorber backꢁ
scatterer
N
R
(Å)
σ2
ꢂE0
(eV) factor
Rꢁ
(10ꢁ3Å2)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Pd
Pd
4.2 ± 0.6 2.79 ± 0.01 7.7 ± 1.2
4.6 ± 0.6 2.79 ± 0.01 8.0 ± 1.0
7.0 ± 1.2 2.83 ± 0.01 8.1 ± 1.3
4.1 ± 0.9 2.80 ± 0.01 8.7 ± 1.5
ꢀ6
ꢀ6
3
0.02
0.03
Au
Au
Pd
Au
3
26
ACS Paragon Plus Environment

Page 27 of 39
ACS Catalysis
1
2
3
4
Table 3. TOF values for deoxygenation of alkyl carboxylic acids over Pd/SiO2.
5
6
7
Catalyst
Acid
T (°C) % H2
P
(atm)
TOF x 103 Reference
(sꢁ1)
1.1 wt.% Pd/SiO2
4.0 wt.% Pd/SiO2
0.5 wt.% Pd/SiO2
1.78 wt.% Pd/SiO2
0.42 wt.% Pd
propanoic
propanoic
propanoic
octanoic
200
200
250
260
20
20
100
10
1
1
1
1
11
5.3
9.4
[15]
[15]
[16]
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
41 ± 2.3
This work
1.2 wt.% Au/SiO2
octanoic
260
10
1
43 ± 3.7
This work
27
ACS Paragon Plus Environment

ACS Catalysis
Page 28 of 39
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
(d)
(c)
Figure 1. TEM images of (a, b) Pd/SiO2 (IW), (c) Pd/SiO2 (IE), and (d) PdAu/SiO2 (1:1)
catalysts.
28
ACS Paragon Plus Environment

Page 29 of 39
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(a)
(b)
(c)
Figure 2. TEM particle size distributions of (a) Pd/SiO2( IW) and (b) Pd/SiO2 (IE) and (c)
PdAu/SiO2 (1:1) catalysts.
29
ACS Paragon Plus Environment