Oxygen step-response experiments for methane oxidation over Pd/Al2O3: An in situ XAFS study

Article abstract of DOI:10.1016/j.catcom.2018.02.011

Methane oxidation over Pd/Al₂O₃ has been investigated by in situ XAFS characterization during oxygen step-response experiments. With a net-reducing feed gas, Pd is in a reduced state and the introduction of oxygen leads to oxidation of palladium as well as increased methane conversion. When the rate of Pd oxidation is slow, a transient surface oxidized state is observed with low activity for methane oxidation. The activity increases when palladium is further oxidized and the highest activity is observed for palladium oxide.

Full text of DOI:10.1016/j.catcom.2018.02.011

CatalysisCommunications109(2018)24–27
Contents lists available at ScienceDirect
Catalysis Communications
journal homepage: www.elsevier.com/locate/catcom
Short communication
Oxygen step-response experiments for methane oxidation over Pd/Al2O3: An
in situ XAFS study
Johan Nilssona,b, Per-Anders Carlssona,b, Natalia M. Martina,b, Peter Velina,b
Debora Motta Meirac, Henrik Grönbeckd,b, Magnus Skoglundha,b,*
,
a
Department of Chemistry and Chemical Engineering, Chalmers University of Technology, Göteborg, Sweden
Competence Centre for Catalysis, Chalmers University of Technology, Göteborg, Sweden
European Synchrotron Radiation Facility, Grenoble, France
b
c
d
Department of Physics, Chalmers University of Technology, Göteborg, Sweden
A R T I C L E I N F O
A B S T R A C T
Keywords:
Methane oxidation
Palladium
XAFS
In situ spectroscopy
Step-response
Methane oxidation over Pd/Al2O3 has been investigated by in situ XAFS characterization during oxygen step-
response experiments. With a net-reducing feed gas, Pd is in a reduced state and the introduction of oxygen leads
to oxidation of palladium as well as increased methane conversion. When the rate of Pd oxidation is slow, a
transient surface oxidized state is observed with low activity for methane oxidation. The activity increases when
palladium is further oxidized and the highest activity is observed for palladium oxide.
1. Introduction
reduced state during methane oxidation [10]. For Pd surfaces com-
pletely saturated with oxygen, a significant decrease in methane oxi-
Hydrocarbon emissions from natural gas vehicles consist mainly of
uncombusted methane. Methane is a strong greenhouse gas and it is
important to minimize the emissions of uncombusted methane. This can
be achieved using a catalytic converter that facilitates the total oxida-
tion of methane to carbon dioxide and water. As methane is the most
difficult hydrocarbon to oxidize catalytically, highly effective catalyst
designs are required to obtain sufficient conversion over a wide tem-
perature range.
Palladium is the most active metal for the catalytic total oxidation of
methane under lean (net-oxidizing) conditions [1, 2]. Under lean con-
ditions, the stable form of palladium at temperatures below ca. 790 °C is
palladium oxide. However, palladium can readily be reduced if the
temperature is raised or the feed gas is changed to a net-reducing
composition [3-5]. Palladium oxide has high activity for methane oxi-
dation and has been assigned as the active phase for methane oxidation
at low temperatures under lean conditions [3, 6, 7]. In particular, the
PdO(101) facet has been identified as highly active for methane oxi-
dation with low barrier for hydrogen abstraction on under-coordinated
surface sites [8, 9].
dation activity has been reported [11, 12]. It has been suggested that
the equilibrium between bulk Pd and surface PdO is important for the
activity at stoichiometric conditions [13].
In this report, we have investigated methane oxidation over a Pd/
Al2O3 catalyst using in situ X-ray absorption fine structure (XAFS) to
follow the oxidation state of Pd during the reaction. Oxygen step-re-
sponse experiments were performed where the feed gas composition
was switched rapidly from a rich composition containing only methane
to a net-oxidizing feed containing both methane and oxygen. The ef-
fluent stream from the reactor was monitored by mass spectrometry and
time-resolved XAFS in combination with linear combination fitting was
used to quantify the oxidation state of Pd as a result of the changing
feed gas composition. In this way, we were able to investigate how the
dynamics of Pd oxidation affect the catalytic activity for methane oxi-
dation.
2. Materials and methods
The 5 wt% Pd/Al2O3 catalyst was prepared through incipient wet-
ness impregnation by mixing 10 g of γ-alumina (Puralox SBa 200, Sasol)
with 10 g of aqueous Pd(NH3)2 solution (Alfa Aesar). The slurry was
stirred for 15 min, frozen with liquid nitrogen, and freeze-dried. The
freeze-dried powder sample was calcined in air at 500 °C for 1 h,
However, there are still uncertainties regarding the structure-ac-
tivity relationship for methane oxidation over Pd catalysts at inter-
mediate oxidation states. With a feed gas composition close to stoi-
chiometric, palladium has been reported to be predominately in a
*
Corresponding author at: Competence Centre for Catalysis, Chalmers University of Technology, Göteborg, Sweden.
E-mail address: skoglund@chalmers.se (M. Skoglundh).
https://doi.org/10.1016/j.catcom.2018.02.011
Received 21 December 2017; Received in revised form 10 February 2018; Accepted 15 February 2018
Availableonline20February2018
1566-7367/©2018ElsevierB.V.Allrightsreserved.

J. Nilsson et al.
CatalysisCommunications109(2018)24–27
starting from room temperature with a heating rate of 5 °C/min. The
calcined catalyst was characterized by X-ray diffraction using a Bruker
XRD D8 Advance instrument using monochromatic CuKα1 radiation.
The 2θ angle was varied from 20 to 65.9° and the step size and time was
0.029° and 1 s, respectively.
The energy-dispersive XAFS measurements were performed at the
ID24 beamline at the European Synchrotron Radiation Facility,
France [14]. X-ray absorption spectra were recorded in transmission
mode at the Pd K edge at 24.35 keV, with an energy range of ap-
proximately 23 500 to 26500 eV, and a time-resolution of 500 ms. A Pd
foil (Goodfellow, 99.99% purity) was used for energy calibration of the
XAFS spectra. A flow-through type catalytic cell, developed at ESRF was
used [14]. The powder sample was placed in a sample holder, with
diameter of 5 mm, made of Inconel. The sample holder was loaded with
approximately 40 mg of catalyst. The gas flows from top to bottom and
the sample is kept fixed at the bottom by a stainless steel gauze. The X-
rays penetrate the sample from side to side near the top of the catalyst
bed through carbon windows in the wall of the sample holder. Rapid
switching between feed gas compositions were facilitated by the small
dead volume of the cell (< 0.5 cm3) and performed using air-actuated
four-way gas valves (Valco, VICI).
Fig. 1. XRD pattern of the Pd/Al2O3 catalyst and the γ-alumina support. The reflections
from γ-Al2O3 are indicated by (*).
Helium was used as carrier gas and the total flow was kept constant
at 75 ml/min. Step-response experiments were performed at tempera-
tures of 320, 360, and 400 °C. The catalyst was reduced in a stream of
0.2% methane for 15 min before oxygen was added to the feed. For each
temperature, three different experiments were performed with an
oxygen concentration of 2.5, 1.2, and 0.6%, respectively, with net-
oxidizing conditions held for 15 min. The effluent stream from the re-
actor was continuously analyzed using mass spectrometry (Prisma,
Pfeiffer) following the m/z 4 (He), 15 (CH4), 32 (O2) and 44 (CO2).
XAFS data processing was performed using the Larch software [15].
Linear combination fitting was used to determine the fraction of Pd2+
in the catalyst. The spectrum of the fresh catalyst and a spectrum of the
catalyst reduced in H2 were used as references for Pd2+ and Pd0, re-
spectively. The reference spectra and the measured spectra for fitting
were normalized using the standard pre-edge function in Larch, by
subtracting the pre-edge and normalizing the spectra by the edge step.
The reference spectrum for Pd0 was assigned a linear coefficient x1 and
the Pd2+ reference spectrum a linear coefficient x2. Furthermore, x1
was set to be a free parameter while x2 was constrained to (1−x1).
Fitting was performed by minimizing the residual between the mea-
sured spectra and the linear combination of the reference spectra using
an energy range from 24000 to 24800 eV. The estimated standard error
in the fit of the linear coefficients is 0.02–0.03.
Fig. 2. XAFS spectra of the Pd/Al2O3 catalyst recorded in situ under a flow of 0.2% CH4
(solid black line) and 0.2% CH4 + 2.5% O2 (solid grey line, offset by 0.2) at 400 °C.
Included are also the reference spectra for Pd0 (dashed black line) and Pd2+ (dashed grey
line, offset by 0.2).
resolution in the energy-dispersive geometry. To quantify the oxidation
of Pd during the experiment, linear combination fitting was performed
by fitting the measured spectrum during the methane oxidation ex-
periment to reference spectra of Pd2+ and Pd0. From this procedure, the
fraction of oxidized Pd in the catalyst is determined.
Fig. 3 shows the outlet concentration of products and reactants as
well as the fraction of Pd2+ calculated from the linear combination
fitting of the XAFS spectra, for experiments performed at 320, 360, and
400 °C. During the net-reducing period, before oxygen is added to the
feed at 900 s, carbonaceous species accumulate on the surface of the
catalyst. The XAFS linear combination fitting gives a slight increase of
the fraction of Pd2+ during this period. This increase of the Pd2+
fraction is not a result of an oxygen induced oxidation of the catalyst
but rather due to the carbonaceous species that accumulate on the
catalyst. Changes in the XANES spectra similar to a slight oxidation
have been reported between Pd and PdC0.13 [16]. This behaviour is
more prominent for the experiments performed at high temperature
where the methane conversion is fairly high before the start of net-
oxidizing step. When oxygen is added to the feed at 900 s, the com-
bustion of the carbonaceous species results in an increased production
of CO2 at the start of the net-oxidizing period. The methane conversion
increases after oxygen has been added to the feed, and the increase of
the methane conversion is rapid for the experiments performed at
higher temperature and with higher oxygen concentration. A temporary
minimum in the methane conversion is observed in the beginning of the
net-oxidizing period in all experiments. However, the profile of the
3. Results and discussion
The XRD patterns for the Pd/Al2O3 catalyst and the pure γ-alumina
support are shown in Fig. 1. For alumina, reflections appear at
2θ = 37.6, 39.5, 45.9, and 60.9°. Compared to the support material, the
catalyst shows additional reflections at 2θ = 34.1, 42.1 and 55.0°.
Comparison with the JCPDS reference cards confirms that this set of
signals correspond to palladium oxide. Using the Scherrer equation and
the peak width at half maximum for the reflections at 2θ = 34.1 and
55.0°, the PdO crystallite size is estimated to be about 6 nm. However,
since XRD is not sensitive to very small nanoparticles (below 2–3 nm) it
should be noted that the average crystallite size can be overestimated
using this method.
Fig. 2 shows the XAFS spectra for the Pd/Al2O3 catalyst recorded in
situ before the oxygen step, in a flow of 0.2% methane, and after the
oxygen step, in a flow of 0.2% methane and 2.5% oxygen. The major
difference between the two spectra is found in the area just above the
absorption edge where the position of the first peak is shifted to lower
energies in the more oxidized Pd sample. Compared to a spectrum re-
corded at a conventional monochromator beamline, the increase in
intensity just after the absorption edge is lower due to the lower energy
25

J. Nilsson et al.
CatalysisCommunications109(2018)24–27
Fig. 3. Oxidation of 0.2% CH4 over the Pd/Al2O3 catalyst at 320, 360, and 400 °C. Oxygen with a concentration of 0.6, 1.2, or 2.5% is added to the feed at 900 s. The upper panels show
the methane conversion and the fraction of oxidized Pd determined by linear combination fitting, and the lower panels show the outlet concentrations of CO2 and O2.
methane conversion is different depending on the reaction conditions,
and the temperature dependent response for the methane conversion is
likely related to the oxidation rate of palladium.
i.e. more rapid oxidation at first followed by a slower oxidation.
The observed methane slip at the beginning of the net-oxidizing step
is similar to what has been observed previously in oxygen pulse-re-
sponse experiments [4, 17], however, by quantifying the rate of Pd
oxidation using linear combination fitting additional insight is gained
to the structure-activity relationship of methane oxidation over Pd.
When the rate of Pd oxidation is low, there is a clear temporary
minimum in the methane conversion, after the initial increase at the
start of the net-oxidizing step. When the Pd oxidation rate increases this
behaviour is much less pronounced. The temporary minimum in the
methane conversion is observed when the fraction of Pd2+ is close to
0.3. Consequently, in the experiments with a slower oxidation of Pd,
there is a longer period of time between the start of the net-oxidizing
period, and the observed temporary minimum in the methane conver-
sion. The most rapid Pd oxidation is observed for the experiments at
400 °C with 1.2 and 2.5% O2, here only a minor temporary decrease in
the methane conversion observed when the fraction of Pd2+ is about
0.4.
In the beginning of the net-oxidizing period, oxygen initially is
chemisorbed to the Pd surface. This leads to a surface oxidation re-
cognized as a rapid increase in the fraction of Pd2+ detected by the
XAFS linear combination fitting, from an initial value of 0.05–0.10 to an
intermediate oxidation state of 0.20–0.25. This fraction corresponds
well with the fraction of Pd surface atoms, which is 0.2 for 6 nm Pd
crystallites. Following the initial rapid increase, the Pd oxidation state
continues to increase with a slower rate, which corresponds to bulk
oxidation of Pd. The rate of the Pd bulk oxidation increases con-
siderably with higher oxygen concentration and temperature. For the
experiment at 320 °C with 0.6% O2 the fraction of Pd2+ is 0.4 at the end
of the 15 min net-oxidizing period, and for the experiment at 400 °C
with 2.5% O2 the corresponding fraction of Pd2+ is 0.9. The rate with
which the Pd nanoparticles are oxidized is likely related to the Pd
particle size. Small Pd nanoparticles, with a size on the order of one or
two nanometres, are expected to be bulk oxidized more rapidly com-
pared to particles that are considerably larger. In the present experi-
ments, this is observed as a bulk oxidation obeying a varying kinetics,
Several studies have reported that a surface oxide or an oxygen
saturated Pd surface has a low activity for methane oxidation [8, 11, 12,
18]. On the other hand, palladium oxide is highly active for methane
26

J. Nilsson et al.
CatalysisCommunications109(2018)24–27
oxidation [7, 8] and this is the reason for the different methane oxi-
dation behaviours depending on the Pd oxidation rate. In the experi-
ments where the oxidation of Pd is rapid, full methane conversion is
reached rapidly after the introduction of oxygen in the feed, with only a
minor temporary loss of conversion. Here, palladium oxide is formed
rapidly, and is the active phase for methane oxidation when oxygen is
present in the feed. When the rate of Pd oxidation is slow, additional
time is required to build a palladium oxide phase after the start of the
net-oxidizing step. Initially methane can be oxidized over reduced Pd,
but the surface is rapidly oxidized which leads to a temporary low ac-
tivity. The activity increases again when Pd has been oxidized to PdO.
over platinum and palladium, J. Catal. 122 (2) (1990) 280–294, http://dx.doi.org/
10.1016/0021-9517(90)90282-O.
[2] R. Burch, P.K. Loader, Investigation of Pt/Al2O3and Pd/Al2O3 catalysts for the
combustion of methane at low concentrations, Appl. Catal. B 5 (1–2) (1994)
149–164, http://dx.doi.org/10.1016/0926-3373(94)00037-9.
[3] S.K. Matam, M.H. Aguirre, A. Weidenkaff, D. Ferri, Revisiting the problem of active
sites for methane combustion on Pd/Al2O3 by operando XANES in a lab-scale fixed-
bed reactor, J. Phys. Chem. C 114 (20) (2010) 9439–9443, http://dx.doi.org/10.
1021/jp1019697.
[4] J. Nilsson, P.-A. Carlsson, S. Fouladvand, N.M. Martin, J. Gustafson, M.A. Newton,
E. Lundgren, H. Grönbeck, M. Skoglundh, Chemistry of supported palladium na-
noparticles during methane oxidation, ACS Catal. 5 (4) (2015) 2481–2489, http://
dx.doi.org/10.1021/cs502036d.
[5] J. Nilsson, P.-A. Carlsson, N.M. Martin, E.C. Adams, G. Agostini, H. Grönbeck,
M. Skoglundh, Methane oxidation over Pd/Al2O3 under rich/lean cycling followed
by operando XAFS and modulation excitation spectroscopy, J. Catal. 356 (2017)
237–245, http://dx.doi.org/10.1016/j.jcat.2017.10.018.
4. Conclusions
[6] S.K. Matam, A. Eyssler, P. Hug, N. van Vegten, A. Baiker, A. Weidenkaff, D. Ferri,
Elucidation of structure–activity relationships of model three way catalysts for the
combustion of methane, Appl. Catal. B 94 (1–2) (2010) 77–84, http://dx.doi.org/
10.1016/j.apcatb.2009.10.023.
Methane oxidation has been studied over a Pd/Al2O3 catalyst using
oxygen step-response experiments with in situ monitoring by X-ray
absorption fine structure. When flowing only methane over the catalyst,
the methane conversion is generally low. When oxygen is added to the
feed the methane conversion increases rapidly. The oxidation state of
palladium increases when oxygen is added to the feed. Upon oxygen
chemisorption the oxidized fraction of palladium increases rapidly due
to surface oxidation. Thereafter bulk oxidation proceeds. The rate of
palladium bulk oxidation increases with increasing oxygen concentra-
tion and temperature.
The rate of Pd oxidation influences also the methane conversion.
Initially when oxygen is introduced in the feed the methane conversion
increases rapidly which is assigned to oxidation over metallic Pd. When
the surface of the Pd nanoparticles is oxidized there is a temporary
decrease in the methane conversion. When a bulk oxide phase has been
formed, the conversion increases with palladium oxide as the active
phase.
[7] A. Hellman, A. Resta, N.M. Martin, J. Gustafson, A. Trinchero, P.-A. Carlsson,
O. Balmes, R. Felici, R. van Rijn, J.W.M. Frenken, J.N. Andersen, E. Lundgren,
H. Grönbeck, The active phase of palladium during methane oxidation, J. Phys.
Chem. Lett. 3 (6) (2012) 678–682, http://dx.doi.org/10.1021/jz300069s.
[8] N.M. Martin, M. Van den Bossche, A. Hellman, H. Grönbeck, C. Hakanoglu,
J. Gustafson, S. Blomberg, N. Johansson, Z. Liu, S. Axnanda, J.F. Weaver,
E. Lundgren, Intrinsic ligand effect governing the catalytic activity of Pd oxide thin
films, ACS Catal. 4 (10) (2014) 3330–3334, http://dx.doi.org/10.1021/cs5010163.
[9] M. Van den Bossche, H. Grönbeck, Methane oxidation over PdO(101) revealed by
first-principles kinetic modeling, J. Am. Chem. Soc. 137 (37) (2015) 12035–12044,
http://dx.doi.org/10.1021/jacs.5b06069.
[10] Y. Lu, M. Santhosh Kumar, G.L. Chiarello, P. Dimopoulos Eggenschwiler, C. Bach,
M. Weilenmann, A. Spiteri, A. Weidenkaff, D. Ferri, Operando XANES study of si-
mulated transient cycles on a Pd-only three-way catalyst, Catal. Commun. 39
(2013) 55–59, http://dx.doi.org/10.1016/j.catcom.2013.05.006.
[11] Y.-H.C. Chin, C. Buda, M. Neurock, E. Iglesia, Consequences of metal-oxide inter-
conversion for CeH bond activation during CH4 reactions on Pd catalysts, J. Am.
Chem. Soc. 135 (41) (2013) 15425–15442, http://dx.doi.org/10.1021/ja405004m.
[12] X. Weng, H. Ren, M. Chen, H. Wan, Effect of surface oxygen on the activation of
methane on palladium and platinum surfaces, ACS Catal. 4 (8) (2014) 2598–2604,
http://dx.doi.org/10.1021/cs500510x.
Acknowledgements
[13] Y. Lu, S. Keav, V. Marchionni, G.L. Chiarello, A. Pappacena, M. Di Michiel,
M.A. Newton, A. Weidenkaff, D. Ferri, Ageing induced improvement of methane
oxidation activity of Pd/YFeO3, Catal. Sci. Technol. 4 (9) (2014) 2919, http://dx.
doi.org/10.1039/C4CY00289J.
[14] S. Pascarelli, O. Mathon, T. Mairs, I. Kantor, G. Agostini, C. Strohm, S. Pasternak,
F. Perrin, G. Berruyer, P. Chappelet, C. Clavel, M.C. Dominguez, The time-resolved
and extreme-conditions XAS (TEXAS) facility at the European Synchrotron
Radiation Facility: the energy-dispersive X-ray absorption spectroscopy beamline
ID24, J. Synchrotron Radiat. 23 (1) (2016) 353–368, http://dx.doi.org/10.1107/
S160057751501783X.
[15] M. Newville, Larch: an analysis package for XAFS and related spectroscopies, J.
Phys. Conf. Ser. 430 (2013) 012007, http://dx.doi.org/10.1088/1742-6596/430/
1/012007.
[16] J.A. McCaulley, Temperature dependence of the Pd K-edge extended X-ray-ab-
sorption fine structure of PdCx (x ∼ 0.13), Phys. Rev. B 47 (9) (1993) 4873–4879,
http://dx.doi.org/10.1103/PhysRevB.47.4873.
The authors thank the European Synchrotron Radiation Facility
(ESRF) for providing beamtime. This work is financially supported by
the project “Unravelling catalytically active sites with X-ray absorption
spectroscopy”, Swedish Research Council (621-2011-5009); the project
“Time-resolved in situ methods for design of catalytic sites within
sustainable chemistry”, Swedish Research Council, Röntgen-Ångström
collaboration (No. 349-2013-567), the project “Novel two-dimensional
systems obtained on SiC as a template, for electronics, sensing and
catalysis”, Swedish Foundation for Strategic Research (RMA15-0024);
as well as the Competence Centre for Catalysis, which is financially
supported by the Swedish Energy Agency (22490-3), Chalmers
University of Technology and the member companies: AB Volvo, ECAPS
AB, Haldor Topsøe A/S, Volvo Car Corporation AB, Scania CV AB, and
Wärtsilä Finland Oy.
[17] N.M. Martin, J. Nilsson, M. Skoglundh, E.C. Adams, X. Wang, G. Smedler, A. Raj,
D. Thompsett, G. Agostini, S. Carlson, K. Norén, P.-A. Carlsson, Study of methane
oxidation over alumina supported Pd-Pt catalysts using operando DRIFTS/MS and
in situ XAS techniques, Catal. Struct. React. 3 (1–2) (2017) 24–32, http://dx.doi.
org/10.1080/2055074X.2017.1281717.
[18] P.-A. Carlsson, E. Fridell, M. Skoglundh, Methane oxidation over Pt/Al2O3 and Pd/
Al2O3 catalysts under transient conditions, Cat. Lett. 115 (1–2) (2007) 1–7, http://
dx.doi.org/10.1007/s10562-007-9057-1.
References
[1] R.F. Hicks, H. Qi, M.L. Young, R.G. Lee, Structure sensitivity of methane oxidation
27