FACE(ノースフェイス) 【送料無料】バッグ・小物 ラップトップバッグtargus

Special Issue Article
Partial Oxidation of Methane to Syngas over Pt/Rh/MgO
Catalyst Supported on FeCralloy Woven Fibre†
Z. Ma, P. Ouzilleau, C. Trevisanut, C. Neagoe, S. Lotfi, D. C. Boffito, G. S. Patience*
Department of Chemical Engineering, Polytechnique Montreal, C.P. 6079, Succ. CV Montreal,
H3C 3A7, QC, Canada
* Corresponding Author: G. S. Patience
email: [email protected]
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process, which
may lead to differences between this version and the Version of Record. Please cite this
article as doi: [10.1002/cjce.22428]
Received 23 June 2015; Revised 21 August 2015; Accepted 15 September 2015
The Canadian Journal of Chemical Engineering
This article is protected by copyright. All rights reserved
DOI 10.1002/cjce.22428
This article is protected by copyright. All rights reserved
Integrating a high pressure syngas step with Fischer-Tropsch (FT) in a single vessel reduces investment and operating costs to synthesize GtL liquids.
Methane catalytic partial oxidation (CPOX) to produce syngas for FT is an
economic opportunity for micro-refineries. Many metals and metal oxides selectively convert natural gas to CO and H2 but they also form coke, which must
be removed intermittently otherwise it deactivates the catalyst and can foul
the reactor and process lines. Here, we prepared a mass fraction of 1 % Pt/Rh
(Pt/Rh = 9) catalyst supported on MgO over FeCralloy woven fibre via Solution
Combustion Synthesis. At 900 ◦C, from 0.1 MPa to 2 MPa and with a 2:1 feed
composition of CH4 to O2 , we consumed all the oxygen and obtained a H2 /CO
ratio of 2 (ideal for FT). The catalyst at low pressure and a 0.1 s residence time
converted 90 % of the methane at 90 % CO selectivity. At 2 MPa, we obtained
a CO yield of 50 % (<88 % conversion and 57 % selectivity). Thermodynamics predict that less than 5 % coke forms below 900 ◦C. At high pressure and
short residence time (0.1 s), the coke yield (presumed to be coke crystallites)
was 24 %. Increasing the residence time to 0.3 s reduced the amount of coke by
33 % because it is metastable. This article is protected by copyright. All rights reserved
Keywords: catalytic partial oxidation, methane, syngas, FeCralloy, Pt, Rh,
MgO, coke, thermodynamics
EIA (U.S. Energy Information Administration) projects a 56 % increase in
the world’s energy consumption by 2040, with fossil fuels accounting for nearly
80 % of it [1]. In North America, natural gas, including tight gas, shale gas
and coal bed methane, is the fastest-growing carbon source. It competes with
coal as the world’s second largest energy source (Figure 1). The world relies on
liquid fuels as the predominant energy vector, but environmental, geopolitcal
and macroeconomic events greatly affect their price [2]. This motivated many
nations to develop alternative solutions to produce liquid fuels [3, 4, 5], such as
This article is protected by copyright. All rights reserved
gas-to-liquids GtL products [2]. Natural gas is as abundant as oil and coal but
most of it is trapped in hydrates underneath the ocean or in remote regions of
the world.
Associated natural gas that is dissolved in petroleum is often flared or, worse,
vented. Instead of wasting this resource, converting it to a liquid fuel not only
reduces its negative environmental impact [6], but also generates an additional
revenue stream. The first step to diesel (FT fuels) is to partially oxidize methane
to syngas with air (POX), steam (steam methane reforming, SMR) or a combination of the two (auto thermal reforming, ATR). POX reaction temperature
runs as high as 1500 ◦C, but metals and metal oxides catalyze the reation below
1000 ◦C [7]. The reaction is mildly exothermic (∆H298
= −36 kJ mol−1 ) and
the product H2 /CO ratio is close to 2, which is optimum for FT liquid fuels [8].
This process requires a relatively pure hydrocarbon stream: sulphur compounds
and metals must be removed upstream of the syngas reactor. Therefore, besides
reducing anthropogenic CO2 and CH4 emissions, the process eliminates toxic
The choice of catalyst is a crucial component of the CPOX reaction. Structured catalysts like gauze and woven fibres combine mechanical elasticity and
resistance to thermal stress at extremely high space velocities (12 000 000 h−1 )
[4, 9]. Catalysts are mainly noble metals, such as Pt, Pd, Rh and Ir, because of
their high reactivity [10]. Another advantage of a structured catalyst is that the
final design of the reactor allows greater throughput. Reactors are compact and
easy to handle for conducting experimental studies and are readily designed for
an industrial scale [7, 10].
Following the initial discovery of Schmidt and Hickman [4, 11], researchers
have studied various metallic gauze catalysts. Methane conversion and CO
selectivity are low on Pt gauze operating between 200 ◦C to 900 ◦C [7]. Adding
Rh to the Pt (Pt/Rh = 9) increased the conversion to 30 % [12]. Surface
oxides of Pd, Pt and Pt-Ir form and degrade catalytic activity more than Pt/Rh
[7, 12, 13, 14, 15]. CO selectivity was 95 % at 1000 ◦C and 0.1 MPa in a 15 mm
quartz reactor operating with a CH4 /O2 ratio of 2 [13]. Pd and Ni metals on
stainless steel gauzes and FeCralloy woven metal fibres convert almost 100 % of
the methane at 90 % CO selectivity with little coke ( 0.1 %) [16, 17].
Washcoating is the most common method to deposit metals and binders onto
structured supports. Other methods also includes eletrochemical deposition
[18, 19, 20, 21] and in situ spray pyrolysis [22, 23], etc. Solution Combustion
Synthesis (SCS) is a quick process to prepare powders but can be adapted to
coat nanocrystalline metal oxides, such as Al2 O3 , and zero-valent metal onto
monoliths, honeycombs and gauzes. Prepared by SCS, Ru/Al2 O3 loaded on
monolith catalyst for oxy-steam reforming of methane [24], and palladium on
zirconiastabilised lanthanum manganese oxide perovskite on FeCralloy gauzes
for natural gas combustion [25] are quite active.
Coke deactivates catalyst but also promotes methane combustion to CO2
rather than CO. MgO supports resist carbon formation on Ni and Pt based
catalysts [7, 26, 27, 28, 29]. Rather than encapsulating active sites, the carbon
on MgO is amorphous and in the form of non-deactivating whiskers due to the
This article is protected by copyright. All rights reserved
formation of ductile cast iron [30].
In this paper, we prepared a new structured catalyst by depositing Pt/Rh/MgO
on FeCralloy woven fibres by SCS, and evaluated the catalyst’s performance in
a CPOX reaction. We also discuss coke formation with respect to equilibrium
thermodynamics. We focused on the impact of pressure: conditions ranged
from atmospheric pressure to the high required pressures of a Fischer-Tropsch
reaction (typically 2 MPa).
Catalyst preparation
The support was a FeCralloy woven fibre, with a mass fraction of 20 %
chromium, 5 % aluminum, yttrium > 0.1 %, 0.3 % silicon, 0.08 % manganese,
0.03 % copper, 0.03 % carbon and the balance is iron. We cut the woven fibres
into 10 mm×10 mm squares. They were then washed in a water/acetone solution
(1:1) in an ultrasonic bath for 30 min. A muffle furnace (a Neytech model 3-550,
Vulcan Multi-stage programmable furnace) dried the samples at 120 ◦C for 1 h
and calcined at 1000 ◦C for 4 h.
We deposited MgO by dipping the 1.8 g FeCralloy squares in a 10 mL aqueous solution containing magnesium nitrate (Mg(NO3 )2 ·6H2 O, Sigma Aldrich)
as a precursor and urea (CO(NH2 )2 , Sigma Aldrich) as the organic fuel. We
calculated the masses of the reagents according to Vita et al. [24]. The stoichiometric ratio between oxidizing (O: total valences of oxidizers, i.e. nitrates)
and reducing species (F: total valences of fuel, i.e. urea) was 1. The concentration of the solutions were 1.7 mol L−1 and 3.6 mol L−1 for Mg(NO3 )2 ·6H2 O
and CO(NH2 )2 respectively. We placed the solution in the furnace at 600 ◦C for
2 min. This procedure was repeated 3 times to reach a mass fraction of 36.3 %
of MgO on the bare woven fibres. To test the mechanical stability, we put MgO
coated fecralloy fibre in a 40 % intensity ultrasonic bath for 30 min. The mass
was constant after 1 h drying at 120 ◦C.
Before depositing the active components (Pt and Rh), the samples calcined
at 1000 ◦C for 3 h at a heating and cooling rate of 5 ◦C min−1 . To minimize the
catalyst cost, we must minimize the mass of the precious metals while simultaneously maximizing the activity. An autothermal methane catalytic partial
oxidation reaction was rather insensitive to noble metal loading from 0.05 % to
10 % [31]. A middle loading of 5 % is common for the initial study [32]. We compare the catalyst activity of a pure Pt/Rh (Pt/Rh=9) metal gauzes, with 1 %
Pt and Rh which is a much lower metal loading for starting. We deposited Pt
and Rh with the same procedure adopted for MgO. We repeated the dipping
3 times with 2 mL aqueous solution containing tetraammineplatinum(II) nitrate (Pt(NH3 )4 (NO3 )2 , Sigma Aldrich), Rhodium(III) chloride (RhCl3 , Sigma
Aldrich) and urea at a concentration of 0.057 mol L−1 , 0.012 mol L−1 and 0.125 mol L−1
respectively. The final mass fraction of Pt and Rh was 1 % with a Pt/Rh weight
ratio of 9:1. Finally, the catalyst calcined at 1000 ◦C in static air for 4 h with a
heating and cooling rate of 5 ◦C min−1 . To confirm that the total mass fraction
This article is protected by copyright. All rights reserved
of Pt and Rh on the catalyst was 1 %, we weighed the catalyst before and after
deposition and analyzed the catalyst by EDS. We inspected the morphological
change of the FeCralloy during the preparation steps with a 44302-B Deluxe
Handheld digital Microscope (Fig. 2).
Catalyst characterization
A Philips X-Pert MPD diffractometer equipped with a Cu Kα radiation
at 50 kv and 40 mA measured the diffraction of the crystalline phases of the
1 % Pt/Rh on MgO over FeCralloy catalyst. The scanning rate was 0.02 ◦ s−1
with 2θ varying from 20◦ to 90◦ . We attributed the peaks according to the
PCPFWIN database and calculated the crystallite size of MgO and Pt with the
Scherrer equation (particle shape factor = 0.9). We performed scanning electron microscopy (JSM-7600A) with an EDS (Energy Dispersive Spectroscopy)
detector to assess the surface composition [33].
Experimental setup
The reactor for methane partial oxidation was a 30 cm long quartz reactor
with an ID of 8 mm (Fig. 3). We cut three circular disks of the fecralloy woven
metal so that they fit snugly inside the reactor. The total weight of the three
disks was 0.27 g and we placed them in the centre of the reactor without reduction pretreatment [34]. An electrical furnace heated the reactor to 900 ◦C. A
thermocouple within the catalyst monitored the temperature. A back pressure
regulator maintained a constant reactor pressure. The gases entered from the
bottom of the reactor. An Agilent gas-chromatograph 7890 B measured the gas
composition following the ASTM D1945 standard. A mass-spectrometer (Hiden, HPR-20/QIC) monitored the gas phase composition on-line at the exit of
the reactor for the duration of the experiments.
The conversion of methane (XCH4 ) and selectivity towards CO (SCO ), H2
(SH2 ), CO2 (SCO2 ) and the coke based on the inlet CH4 (SC ) were calculated
as follows:
XCH4 (%) =
SCO (%) =
SH2 (%) =
− qCH
× 100
− qCH
SCO2 (%) =
× 100
out ) × 2
− qCH
× 100
− qCH
This article is protected by copyright. All rights reserved
× 100
ml min−1
ml min−1
ml min−1
gCH4 g−1
cat h
Table 1: Design of experiments. Reaction conditions: T = 900 ◦C, CH4 /O2 =
2, mass of catalyst = 0.27 g (mass of catalyst* = 0.2 g). The value of gas flow
Q is under 25 ◦C and atmosphere pressure.
SC (%) =
− qCH
− qCO
− −qCO
× 100
Where: qCH
and qCH
are the molar flow rates of methane at the reactor
entrance and exit, respectively; qCO
, qH
and qCO
are the molar flow rates of
CO, H2 and CO2 at the reactor exit.
We changed pressure, residence time for each 60 min catalytic test. We fed
O2 /Ar (O2 % = 30 %) to remove the coke and regenerate the catalyst between
Catalyst characterization
The XRD analysis of the 1 % Pt/Rh over MgO/FeCralloy detected the phases
of Pt (JCPDS no. 04-0802), MgO (JCPDS no. 45-0946) and FeCr (JCPDS no.
34-0396) (Fig. 4 ). It did not detect any diffraction peak attributable to Rh,
due to its low concentration. The sharp Pt peaks at 2θ equal to 40.0◦ , 46.5◦
and 67.7◦ correspond to large crystallites on the MgO surface. We calculated
the particle size of the main crystallites with the Scherrer equation, which was
54 nm and 47 nm for Pt and MgO, respectively.
The SCS deposited a uniform MgO layer on the bare FeCralloy woven fibres
(Fig. 2). The SEM analysis also detected a thick homogeneous layer of MgO.
(Fig. 5 (a)). The MgO on the FeCralloy surface sintered to give a thicker and
more uniform layer after 40 h of operation (Fig. 5 (b)). This indicates that
MgO effectively covered the bare FeCralloy, and thereby decrease any coking
that would be due to Fe.
On the catalyst surface (Fig. 6 (a)), we confirmed the deposition of Mg,
O, Pt and Rh on the catalyst surface by EDS analysis (Fig. 6 (b)-(d)). The
This article is protected by copyright. All rights reserved
dispersion of Pt and Rh are uniform on the catalyst surface. We did not detect
Fe on the surface (Fig. 7). Comparing with the sum of Mg and O (95.4 % in
EDS), the total mass fraction of Pt and Rh is close to 1 % (2.5 % in EDS). We
did not find obvious coke on the surface of used catalyst. The regeneration with
O2 /Ar is effective to remove the coke, and also maintain the catalytic activity
of the catalyst.
Catalytic tests
We tested the catalyst activity under CPOX conditions (Table 1 ). In the
first tests we kept the feed flow rate constant at a weight hourly space velocity
(WHSV) equal to 15 gCH4 g−1
at elevated pressure (Fig. 8 (a) to (d)). At
cat h
atmospheric pressure the methane conversion decreased to 45 % and stabilized
after 30 min (Fig. 8 (a)). Methane conversion increased to 80 % when the
pressure increased to 2 MPa. At 0.1 MPa, the selectivity of CO rapidly increased
in the first 30 min to 45 % (Fig. 8 (b)).
The CO selectivity improved as the pressure increased and reached a maximum at 0.5 MPa (above 55 %). H2 selectivity followed the same trend (Fig. 8
(c)). The CPOX reaction at 2 MPa gave the highest H2 selectivity (60 % after
30 min) among all the tested pressures. The CO2 selectivity is lower at elevated pressures (Fig. 8 (d)). It decreased from around 40 % to 23 % when going
from atmospheric pressure to 2 MPa. Based on the carbon mass balance, higher
pressure produces more coke. More importantly, the average H2 /CO ratio at
2 MPa is 2 (1.9 at 0.5 MPa and 1 MPa, 1.7 at 0.1 MPa), which is ideal for the
Fischer-Tropsch reaction.
The gas residence time across the FeCralloy was 0.015 s at a WHSV of 15
gCH4 g−1
at 0.1 MPa. With such short contact time, methane conversion is
cat h
only 45 %. At higher pressures while keeping WHSV constant, the residence
time increases (0.075 s at 0.5 MPa, 0.15 s at 1 MPa and 0.3 s at 2 MPa), which is
the main contribution to the increase of the methane conversion. Therefore, to
study the effect of pressure, we then test the CPOX reaction with the constant
residence time of 0.1 s under pressures up to 2 MPa with the same methane to
oxygen ratio of 2 and temperature of 900 ◦C.
Methane conversion was high at atmospheric pressure and a residence time
of 0.1 s (around 90 %, Fig. 9 (a)). At higher pressure, the conversion of methane
drops to around 75 %; however, it changes little from 0.5 MPa to 2 MPa. Pressure affects both CO and H2 selectivity (Fig. 9 (b) and (c)). At 0.1 MPa and
0.5 MPa both selectivities increase quickly in the first 30 min and then SCO
remains constant around 90 % while SH2 increases up to 99 %. The average
selectivities (after 30 min operation) are 9 % and 53 % at 1 MPa and 44 % and
34 % at 2 MPa, respectively. The CO2 selectivity (Fig. 9 (d)) is around 10 %
at 0.1 MPa and 0.5 MPa, around 20 % at 1 MPa and 2 MPa, which also indicates an increasing coke formation with the rising pressure. The average H2 /CO
ratio lies between 1.8 to 2.1 from 0.1 MPa to 1 MPa, however dropping to 1.6
at 2 MPa. To maintain constant residence time, we increased the flow rates of
methane and oxygen/argon proportionally to the operating pressure, but this
disrupted the reactor operation.
This article is protected by copyright. All rights reserved
Coke formation analysis
Coke formation is an inevitable problem when catalytically converting methane
to syngas: it deactivates the catalyst and disrupts the reactor and process lines.
Previous work has either neglected the impact of carbon formation on the thermodynamics [7, 35] or ignored the thermodynamic difference between solid carbon in its graphite form and in its coke form [36, 37].
Considering the importance of carbon deposition [38], we recalculated the
thermodynamic equilibrium of coke with a thermodynamic model (FactSage⃝
Thermochemical Software) [39] [40].
The thermodynamic calculations considers the thermodynamic equilibrium
of all possible reactions including the following chemical species: H2 , CH4 , N2 ,
O2 , H2 O, CO, CO2 . Any possible reactions combining these chemical species
like partial oxidation and Boudouard reactions are included as well as the following:
H2 + O2 ⇌ H2 O
CO + O2 ⇌ CO2
CH4 ⇌ C + 2H2
CO + H2 ⇌ C + H2 O
CH4 + CO2 ⇌ 2CO + 2H2
CH4 + H2 O ⇌ CO + 3H2
CO + H2 O ⇌ CO2 + H2
The Factsage Gibbs energy minimization procedure finds the minimum of
Gibbs energy of the system at constant T and total P by distributing the moles
of C, O and H between the solid phase coke and different gas species. This
results in the equilibrium phase assemblage (with the internal composition of
the phases given as species composition). The resulting partial pressures of
the gaseous species then fully respect all equilibrium constants resulting from
This article is protected by copyright. All rights reserved
all potential reaction between the chosen species. These techniques of Gibbs
energy minimization using Lagrange multipliers were developed around 1960
[41, 42] and are very standard today in computing phase equilibria.
We selected the 7 chemical species to compare with the thermodynamic
calculations of Enger [7]. However, extending this selection of species to 472 did
not affect the thermodynamic calculations presented in the present paper (For
example, methane conversion was modified by less than 0.07 %). The selected
species are the most stable at 900 ◦C and 0.1 MPa to 2 MPa.
The key advantage of the FactSage software is the large number of solution
databases to calculate the solution thermodynamics based on pure substance
properties, interactions parameters and order/disorder modelling. For example,
the database for coke is unique to FactSage and relies on established principles such as the assumed turbostratic structure of coke crystallites and the
kinetic/thermodynamic aspects of graphitization of pyrocarbons (which is directly related to methane coking, as pyrocarbons are formed from the pyrolysis
of methane at high temperatures).
Solid carbon (i.e. coke or graphite) formation regime (Fig. 10) varies with
crystallite size La (the diameter of crystallites composing the coke). The larger
the La (nearing graphite) results in a more stable coke due to high surface
energies. Generally, the formation of large coke crystallites proceeds through
the growth of smaller ones, which requires an equivalent activation energy. It
requires a higher activation energy at higher temperatures to form large crystallites of stable coke. The minimal size for stable solid carbon can be 4 times
larger at an operating temperature of 900 ◦C than 750 ◦C. The presence of active
metals on the surface of the support lowers this activation energy, kinetically
hindering coke formation at high temperatures.
We compared equilibrium coking of CH4 calculated by FactSage⃝
with experimental SC based on a mass balance (Fig. 11). According to equilibrium
thermodynamics, pressure has little impact on the equilibrium conditions for
temperatures close to 1200 K (900 ◦C) or higher [7]. Our kinetic data indicate
that the CPOX reaction has not reached equilibrium, the amount of coke is
above the equilibrium thermodynamic value. The experimental results (Fig. 8
and 9) suggest that high coking rates are associated with a high selectivity of
CO2 . This indicates that CPOX reaction does not reach the Boudouard equilibrium, which is expressed by
C + CO2 ⇌ 2CO
Higher pressure promotes the reverse Boudouard reaction and thus more coke
and CO2 form. At lower pressures (0.1 and 0.5 MPa), CPOX produces coke
much closer to the equilibrium value.
The global kinetics of the CPOX reaction are assumed to be fast because
of the short reactor residence time (less than 0.3 s). Thermal decomposition of
methane to coke and diatomic hydrogen (8), as the first reaction step, is faster
than the other reaction steps of CPOX, because it is not strongly limited by
This article is protected by copyright. All rights reserved
oxygen diffusion between reaction sites [7]. Therefore the generation rate of
metastable coke through thermal decomposition of methane will be higher than
the consumption rate of metastable coke to produce CO or CO2 . Increasing
pressure will lower the generation of metastable coke as it will restrain the
decomposition rate of CH4 to form 2 moles of H2 . This kinetic effect is more
important than the thermodynamic impact of increasing pressure on catalyst
Residence time also affects coking: as we increase the residence time, the
amount of metastable coke is reduced and both selectivity and conversion approach equilibrium. The coke selectivity (SC ) decreased from 24 % (τ = 0.1 s) to
18 % (τ = 0.3 s) at 2 MPa. This is possibly due to a higher methane conversion
and CO selectivity at a relatively longer residence time of 0.3 s. Lower coke also
corresponds to a higher H2 seletivity at 2 MPa. Based on the thermodyanmic
calculations, the coke formed is metastable.
Combining CPOX with FT to make synthetic fuel has advantages over standard technologies with respect to operation and investment. The challenge of
integrating CPOX and FT is converting methane to suitable syngas at high pressure. Methane cokes CPOX catalyst at high pressure and consequently both CO
and H2 selectivity are low High operating temperatures may partially alleviate
the negative effects of coke. However, the catalyst and the FeCralloy thermal
stability can become limiting factors. The FeCralloy woven fibre is thermally
stable only up to 1050 ◦C.
Coating the FeCralloy support with a layer of MgO layer may minimize coke.
We synthesized a 1 % mass fraction of Pt/Rh catalyst supported on MgO over
FeCralloy woven fibre. At 2 MPa and a residence time of 0.3 s, syngas yields
were 50 % CO with a H2 /CO ratio of 2. All the oxygen was consumed for all
experiments, which is critical for the FT step that requires reducing conditions.
Comparison with thermodynamics clearly demonstrate that the measured
coke formation is metastable. Coke conversion to CO through the Boudouard
reaction is assumed to be the limiting factor in metastable coke elimination.
Increasing residence time should thus reduce metastable coke formation and
the associated negative effects on conversion and H2 selectivity. At 2 MPa, a
residence time of 0.3 s reduced coke formation by 33 % and increase CO and
H2 yield than a residence time of 0.1 s . The present study did not explore the
effect of temperature. Higher temperatures should further reduce the amount
of metastable coke.
[1] E.I.A.,
“International Energy Outlook 2014”,
ergy Information Admistration, accessed on 12
http://www.eia.gov/forecasts/ieo/ 2014.
This article is protected by copyright. All rights reserved
U. S. EnJanuary 2015,
[2] E.I.A., “EIA projects world energy consumption will increase 56% by 2040”,
U. S. Energy Information Admistration, accessed on 12 January 2015,
http://www.eia.gov/todayinenergy/detail.cfm?id=12251 2013.
[3] US 6746658 B2 (2004), H. Dindi, W. H. Manogue, J. J. Barnes.
[4] US 5648582 A (1997), L. D. Schmidt, D. A. Hickman.
[5] US 20090026420 A1 (2009), B. Xu.
[6] S. Gibb, Chemical Engineering News 2015, 93, 21.
[7] B. C. Enger, R. Lodeng, A. Holmen, Applied Catalysis A: General 2008,
346, 1.
[8] G. P. V. D. Laan, A. A. C. M. Beenackers, Catalysis Reviews 1999, 41,
[9] T. K. Das, G. Jacobs, P. M. Patterson, W. A. Conner, J. Li, B. H. Davis,
Fuel 2003, 82, 805.
[10] A. Basile, Handbook of Membrane Reactors: Reactor Types and Industrial
Applications, Woodhead Publishing, Sawston, Cambridge 2013.
[11] D. A. Hickman, L. D. Schmidt, Journal Catalysis 1992, 138, 267.
[12] K. H. Hofstad, T. S., O. A. Rokstad, A. Holmen, Catalysis Letters 1997,
45, 97.
[13] M. Fathi, K. H. Hofstad, T. S., O. Rokstad, A. Holmen, Catalysis Today
1998, 42, 205.
[14] J. M. Australia, “Catalyst gauze” Johnson Matthey Australia, accessed
on 15 January 2015, http://www.matthey.com.au/products/catalyst-gauze
[15] US 0126628 (2015), G. S. Patience, D. C. Boffito.
[16] A. L. Tarasov, L. M. Kustov, Catalysis Industry 2013, 5, 14.
[17] RU 2491118 C1 (2012), A. L. Tarasov, L. M. Kustov.
[18] F. Basile, P. Benito, G. Fornasari, V. Rosetti, E. Scavetta, D. Tonelli,
A. Vaccari, Applied Catalysis B: Enviromental 2009, 91, 563.
[19] P. Benito, M. Monti, W. D. Nolf, G. Nuyts, G. Janssen, G. Fornasari,
E. Scavetta, F. Basile, K. Janssens, F. Ospitali, D. Tonelli, A. Vaccari,
Catalysis Today 2015, 246, 154.
[20] P. Benito, M. Monti, I. Bersani, F. Basile, G. Fornasari, E. Scavetta,
D. Tonelli, A. Vaccari, Catalysis Today 2012, 197, 162.
This article is protected by copyright. All rights reserved
[21] E. Verlato, S. Barison, S. Cimino, F. Dergal, L. Lisi, G. Mancino, M. Musiani, L. Vazquez-Gomez, International Journal Hydrogen Energy 2015,
39, 11473.
[22] C. Neagoe, D. C. Boffito, Z. Ma, C. Trevisanut, G. S. Patience, Catalysis
Today 2015, accepted.
[23] C. Trevisanut, S. M. Jazayeri, S. Bonkane, C. Neagoe, A. Mohamadalizadeh, D. C. Boffito, C. L. Bianchi, C. Pirola, C. G. Visconti, L. Lietti,
N. Abatzoglou, L. Frost, J. Lerou, W. Green, G. S. Patience, Canadian
Journal Chemical Engineering 2015, accepted.
[24] A. Vita, G. Cristiano, C. Italiano, S. Specchia, F. Cipiti, V. Specchia,
International Journal Hydrogen Energy 2014, 39, 18592.
[25] S. Specchia, C. Galletti, V. Specchia, Studies Surface Science Catalysis
2010, 175, 59.
[26] S. Tang, J. Lin, K. L. Tan, Catalysis Letters 1998, 51, 169.
[27] M. Soick, O. Buyevskaya, M. Hohenberger, D. Wolf, Catalysis Today 1995,
32, 163.
[28] I. H. Son, S. J. Lee, I. Y. Song, W. S. Jeon, I. Jung, D. J. Yun, D. W.
Jeong, J. O. Shim, W. J. Jang, H. S. Roh, Fuel 2014, 136, 194.
[29] J. B. Claridge, M. L. H. Green, S. C. Tsang, A. P. E. York, A. T. Ashcroft,
P. D. Battle, Catalysis Letters 1993, 22, 299.
[30] B. I. Imasogie, A. A. Afonja, J. A. Ali, Materials Science Technology 2000,
16, 194.
[31] N. J. Degenstein, R. Subramanian, L. D. Schmidt, Applied Catalysis A:
General 2006, 305, 146.
[32] R. Horn, K. A. Williams, N. J. Degenstein, A. Bitsch-Larsen, D. D. Nogare,
S. A. Tupy, L. D. Schmidt, Journal Catalysis 2007, 249, 380.
[33] S. M. Ibrahim, A. A. Badawy, G. A. El-Shobaky, H. A. Mohamed, Canadian
Journal Chemical Engineering 2014, 92, 676âĂŞ684.
[34] R. Abbasi, G. Huang, G. M. Istratescu, L. Wu, R. E. Hayes, Canadian
Journal Chemical Engineering 2015, 93, 1474.
[35] P. Rofouie, M. Moshkelani, M. Perrier, J. Paris, Canadian Journal Chemical Engineering 2014, 92, 593.
[36] S. Ayabe, H. Omoto, T. Utaka, R. Kikuchi, K. Sasaki, Y. Teraoka,
K. Eguchi, Applied Catalysis A: General 2003, 241, 261.
[37] E. Newson, T. B. Truong, International Journal Hydrogen Energy 2003,
28, 1379.
This article is protected by copyright. All rights reserved
[38] D. A. Hickman, L. D. Schmidt, AIChE Journal 1993, 39, 1164.
[39] C. W. Bale, P. Chartrand, S. A. Degterov, G. Eriksson, K. Hack, R. B.
Mahfoud, J. Melancon, A. D. Pelton, S. Petersen, Calphad 2002, 26, 189.
[40] P. Ouzilleau, A. E. Gheribi, G. Eriksson, D. K. Lindberg, P. Chartrand,
Carbon 2015, 85, 99.
[41] R. C. Oliver, S. E. Stephanou, R. W. Baier, Chem. Eng. 1962, 69, 121.
[42] G. Eriksson, E. Rosen, Chemica Scripta 1973, 4, 193.
Figure 1: Until 1960, coal was the leading energy source in the world. Natural gas rivals
coal, while nuclear and hydroelectric energy represent less than 10 % of the total. In 2013,
renewable energy contributed 1.5 % of the total energy supply.
This article is protected by copyright. All rights reserved
Figure 2: (a) uncoated FeCralloy woven fibre; (b) coated MgO over FeCralloy; (c) 1 % Pt/Rh
on MgO over FeCralloy
30 % O2
Electric heater
ooo oo
oo o
Figure 3: The reactor was an 8 mm ID quartz tube filled to 300 mm with sand to preheat
and distribute the gases to the catalyst. Four mass flow controllers (MKS) metered the gases
to the reactor and the analytical instruments to calibrate the feed composition as well as the
This article is protected by copyright. All rights reserved
Figure 4: XRD pattern of 1 % Pt/Rh on MgO/FeCralloy woven metal fibre catalyst
Figure 5: SEM micrographs of 1 % Pt/Rh on MgO/FeCralloy (a) x1000 before reaction; (b)
x1000 after reaction.
This article is protected by copyright. All rights reserved
Figure 6: (a) x10 000 SEM micrographs of the one typical surface of Pt/Rh/MgO FeCralloy
(b) Mg dispersion; (c) Pt dispersion; (d) Rh dispersion.
Figure 7: The composition of the surface of Pt/Rh/MgO fecralloy via SEM-EDS analysis.
This article is protected by copyright. All rights reserved
Figure 8: (a) Methane conversion; (b) CO selectivity; (c) H2 selectivity; (d) CO2 selectivity
vs. time under elevated pressures with CH4 /O2 of 2, 900 ◦C, WHSV = 15 gCH g−1
cat h
This article is protected by copyright. All rights reserved
Figure 9: (a) Methane conversion; (b) CO selectivity; (c) H2 selectivity; (d) CO2 selectivity
vs. time under elevated pressures with CH4 /O2 of 2, 900 ◦C, residence time of 0.1 s.
This article is protected by copyright. All rights reserved
Figure 10: Coke formation regime for CPOX calculated at various crystallite sizes (i.e. activation energies) and pressures with a CH4 /O2 ratio of 2
Figure 11: Experimental and thermodynamically predicted coking of CH4 at 900 ◦C for various
pressures and residence times (RT) with a CH4 /O2 ratio of 2
This article is protected by copyright. All rights reserved