Applied Catalysis A, General 623 (2021) 118286

Contents lists available at ScienceDirect

Applied Catalysis A, General

journal homepage: www.elsevier.com/locate/apcata

Review on the catalytic tri-reforming of methane - Part II:

Catalyst development
Xuan-Huynh Pham a, U.P.M. Ashik b, Jun-Ichiro Hayashi b, c, Alejandro Pérez Alonso a, d,
Daniel Pla d, Montserrat Gómez d, Doan Pham Minh a, *
a
Université de Toulouse, IMT Mines Albi, UMR CNRS 5302, Centre RAPSODEE, Campus Jarlard, F-81013 Albi Cedex 09, France
b
Institute for Materials Chemistry and Engineering, Kyushu University, 6-1, Kasuga Koen, Kasuga 816-8580, Japan
c
Transdisciplinary Research and Education Center of Green Technology, Kyushu University, Kasuga 816-8580, Japan
d
Laboratoire Hétérochimie Fondamentale et Appliquée, UMR CNRS 5069, Université Toulouse 3 – Paul Sabatier, 118 route de Narbonne, 31062, Toulouse Cedex 9,
France

A R T I C L E I N F O A B S T R A C T

Keywords: Methane reforming allows the production of synthesis gas (syngas) which is an important gas mixture feedstock
Tri-reforming of methane for the production of chemicals and energy carriers. Steam reforming of methane (SRM) and partial oxidation of
Active phase methane (POM) have been deployed at large industrial scale, while dry reforming of methane (DRM) and more
Support
recently tri-reforming of methane (TRM) are intensively studied. TRM simultaneously combines SRM, POM and
Catalyst promoter
Syngas
DRM in a unique process and allows overcoming several weaknesses of each individual methane reforming
Catalyst design process: e.g. regulation of the molar ratio of H2/CO by controlling feed composition or adaptation to the variation
in biogas composition as renewable resource. TRM process strongly requires a solid catalyst. To date, the design
of efficient TRM catalysts remains a challenge. This work reviews recent achievements on the development of
catalysts for TRM, and provides a guideline for future work related to TRM catalysts.

1. Introduction related to the catalyst deactivation by coke deposition and/or by ther­

mal sintering [3–8,9]. Thus far, SRM and POM are already deployed at
Methane reforming makes possible the production of synthesis gas the industrial scale for syngas production from natural gas [10–12],
(syngas), which is an important gas mixture feedstock for various in­ while DRM is still at a research and development stage. However, SRM is
dustrial sectors. Its annual production has continuously increased during an energy-intensive process because of its high temperatures (ca. 900 ◦ C)
the last decades [1,2]. Mostly, water, oxygen and carbon dioxide are to favor methane conversion, and high steam-to-methane ratio (S/C =
used to reform methane according to the following equations: 3–4) to limit the catalyst deactivation by coke deposition [10,13,14].
Steam reforming of methane (SRM): Consequently, energy is lost during syngas cooling to the temperature of
downstream processes (i.e. Fisher-Tropsch synthesis (FTS): 220–250 ◦ C;
CH4 + H2 O→CO + 3H2 ΔH298 = + 206 kJ/mol (1)
water-gas shift (WGS): 200− 450 ◦ C, etc.) [15,16]. For POM, several
Partial oxidation of methane (POM): drawbacks can be involved (i) catalyst bed hot-spots; (ii) high cost of
syngas purification to remove residual O2; (iii) catalyst deactivation by
CH4 + 1/2O2 →CO + 2H2 ΔH298 = -36 kJ/mol (2)
re-oxidation of metallic nanoparticles, and coke deposition; iv) fire
Dry reforming of methane (DRM): hazard of O2 and oxygen-enriched mixtures; etc. [17]. For DRM, despite
a consequent effort devoted to this reaction during the last decades, the
CH4 + CO2 →2CO + 2H2 ΔH298 = + 274 kJ/mol (3)
development of an economically viable solid catalyst, which resists to
SRM and DRM are strongly endothermic reactions, while POM is a carbon formation and thermal sintering, is still a crucial challenge for a
slightly exothermic one. Generally, these reactions are favorable at high future commercialization [18].
temperatures, and require a solid catalyst [3–8]. A common feature is Tri-reforming of methane (TRM) combines SRM, POM and DRM in a
unique process (Eq. 4) [19]. TRM exhibits several advantages compared

* Corresponding author.
E-mail address: doan.phamminh@mines-albi.fr (D. Pham Minh).

https://doi.org/10.1016/j.apcata.2021.118286
Received 6 May 2021; Received in revised form 25 June 2021; Accepted 9 July 2021
Available online 13 July 2021
0926-860X/© 2021 Elsevier B.V. All rights reserved.
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

to each individual SRM, POM or DRM process, including: limitation of 2. Active catalytic phase
coke formation by steam; heat consumption reduction by POM
exothermic reaction; and CO2 valorization into syngas. Especially, TRM As stated above, TRM allows producing syngas with a desired H2/CO
not only offers the possibility of controlling the molar ratio of H2/CO by ratio, but its study has only started since the early 2000s [34–36]. The
adjusting the feed composition (Eq. 4), which is not the case of the in­ TRM process needs a solid catalyst to get significant rates and yields at
dividual SRM, POM or DRM, but also it is particularly adapted for syngas reasonable temperatures. However, the development of an active, se­
production from renewable resources such as biogas and/or landfill gas, lective, and stable catalyst is still a major challenge for TRM. The main
which contain mixtures of CO2, H2O and O2 together with CH4 [20]. This criterion in the catalyst selection is its ability to catalyze the three
means that biogas and/or landfill gas reforming does not need a specific reforming reactions (Eq. 1, Eq. 2, and Eq. 3). Likewise, coke deposition,
separation step to perform syngas production via TRM. Moreover, flue and thermal catalyst sintering are main catalyst deactivation factors,
gas from waste incinerators, which mainly contains CO2, H2O and O2, along with the risk of catalyst re-oxidation by oxygen in the feed mixture
can also be used to reform methane via TRM [21]. In addition, according [29,30]. Accordingly, catalysts used for SRM, POM and DRM are not
to Świrk et al. [22,23], TRM can be applied not only to reduce green­ suitable for TRM [19].
house gas emissions, but also to store energy under the form of chemical Nickel is generally considered as the active phase, immobilized on a
energy vectors. As for SRM, POM and DRM reactions, a solid catalyst is solid support, e.g. Al2O3, ZrO2, CeO2, etc. Given that the structure and
required to enable TRM at high temperatures (> 700 ◦ C) [3,4,24]. Under composition of the resulting catalytic material determine its perfor­
such severe conditions, the design of highly-efficient TRM catalysts mance [35], this section focuses on the correlation between catalyst
represents the crucial challenge in the frame of environmentally sus­ preparation methods, both the nature of the active metal species as well
tainable chemical industries [25–30]. as catalyst loadings, and the efficiency of the as-prepared catalytic ma­
terials towards TRM, taking into account their physicochemical char­
aCH4 + bH2O + cO2 + dCO2 → eCO + fH2 (4) acteristics. New trends in the engineering of the active phase in TRM will
Effort has been devoted to the development of TRM catalysts as also be discussed (Scheme 1).
evidenced by the growing number of articles published on TRM, which
has quickly increased in recent years (Fig. 1). Thus, after the foremost 2.1. Catalyst preparation methods
work published by Song in 2001 [21], Moon [31] reviewed the
reforming of gaseous hydrocarbons featuring a section on TRM (2008), To date, various approaches have been applied to synthesize cata­
while Soloviev et al. [4] reviewed the oxidative reforming of methane, lysts for TRM, including hydrothermal process, combustion synthesis,
including a discussion on the thermodynamics of TRM (2018). In 2019, microemulsion, co-precipitation, and impregnation.
Arab Aboosadi and Yadecoury [3] reviewed the impact of feed compo­
sition and reaction temperature in TRM. More recently, in 2020, Zhao 2.1.1. Impregnation
et al. [32] reviewed biogas reforming, with a sub-section on TRM. It is Impregnation is commonly used in heterogeneous catalysis thanks to
also worth to mention that no work has been reported on TRM at large its practical ease, low cost, and low waste formation [37]. The dried
scale yet. impregnated materials are usually activated by calcination and/or
The present contribution belongs to a series of two review papers on reduction to obtain the required catalyst. Song and Pan [27] described
TRM. The first one, entitled “Review on the catalytic tri-reforming of the impregnation of Ni(NO3)2 on various supports such as CeO2, ZrO2,
methane - Part I: impact of operating conditions, catalyst deactivation and MgO, and the mixed oxide of Ce and Zr. Support was suspended in an
regeneration” [33], has discussed the aspects of thermodynamic, mech­ aqueous solution of Ni(II) precursor under agitation for 1 h, followed by
anism, operating condition impact, and catalyst deactivation and drying, grounding into powder, calcination under air, and reduction
regeneration. The second one (this work) provides a comprehensive under hydrogen. In a TRM reaction using a fixed-bed reactor (reaction
overview on the rational design of TRM catalysts, including the conditions: 100 mg of catalyst, 1 bar, 700− 850 ◦ C, feed composition of
following main sections: (i) active catalytic phases; (ii) catalyst supports; CH4:CO2:H2O:O2 = 1:0.48:0.54:0.1 with CH4 flow rate =25 mL/min),
(iii) and catalyst promoters. To the best of our knowledge, this is the first the following performance trends were found for CO2 conversion:
review devoted to the rational design and development of TRM cata­ Ni/MgO > Ni/MgO/CeZrOx > Ni/CeO2 ~ Ni/ZrO2 > Ni/CeZrOx. The
lysts. Current challenges, outcomes and critical recommendations are highest catalytic efficiency of Ni/MgO could be tentatively explained by
also addressed. the following reasons. First, MgO support had the highest specific sur­
face area and the highest basicity (MgO ~ CeZrOx > ZrO2 > CeO2). High
specific surface area generally favors metal dispersion. Second, Ni/MgO
showed enhanced CO2 adsorption in comparison to other catalysts
prepared by the same impregnation method. Thus, CO2, which is more
acidic than O2 and H2O, can preferentially be adsorbed on the support
surface and enhance CH4 reforming to produce syngas while promoting
coke gasification [38,39]. This result evidences a strong effect of the
support in the catalytic efficiency of the catalytic material [27]. For
further details, the characteristics of various catalyst supports for TRM
are presented in the section 3 of this review.
The efficiency of the catalyst is directly related to the size of active
metals, so it is important to optimize the parameters of the synthesis of
the targeted catalytic materials. Lino et al. [40] prepared MgAl2O4
support (denoted as MA) by co-precipitation method. Then, MA was
modified by addition of ZrO2 and with X-ZrO2 (X = Ce, La, Y or Sm with
the molar ratio of X to Zr = 0.25, denoted as XZr phase) using incipient
wetness impregnation (IWI). Finally, Ni/XZr/MA (10 wt.% Ni) catalysts
were prepared by IWI method, followed by air calcination at 750 ◦ C. The
addition of Ce, La, Sm, and Y as promoters together with ZrO2 (XZr
Fig. 1. Number of articles found on the Web of Science using “tri-reforming” phase) facilitates the reduction of NiO to Ni(0) and shifts the tempera­
and “methane” as keywords (data on June 21, 2021). ture programmed reduction (TPR) peak to lower temperature. Among

2
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Scheme 1. Summary of the main parameters in consideration for the development of active phases for TRM catalysts.

the different promoters investigated, only CeZr led to a decrease of Ni 2.1.2. Precipitation
particle size (14 nm) in comparison with the referent Ni/MA catalyst (16 Precipitation allows a good control of the particle size and catalyst
nm). Obviously, in TRM reaction (reaction conditions: 750 ◦ C, 85 mg of structure. The surface characteristics, crystalline nature, metal-support
catalyst, CH4:CO2:H2O:O2:N2 = 3:1:1.4:0.5:2 molar ratio and CH4 flow interaction, and thereby catalytic performance can be tuned by chang­
rate = 51.5 N mL min− 1), Ni/XZr/MA catalyst provided the highest H2 ing the parameters of the synthesis. This was demonstrated by varying
(68 %) and CO (63 %) yields. Also, this catalyst showed the lowest rate calcination temperatures, pH values, and reflux times in the synthesis of
of solid carbon formation, indicating the important role of the Ni particle Ni-CaO-ZrO2 catalysts [41]. Nitrate salts of Ni(II), Ca(II) and Zr(IV) were
size in methane reforming for syngas production and for coke resistance. co-precipitated at 8-14 pH range using a NaOH solution, and refluxed in
Thus, by selecting a good promoter, the authors increased the Ni deionized water for 1− 10 h. Different calcination temperatures of 500,
dispersion, improving consequently the catalyst efficiency. 600 and 700 ◦ C were applied. The authors found that the metal-support
In the section 2.5, the decrease of the Ni particle size to the level of interaction (MSI) was enhanced at high calcination temperatures and
isolated metal atoms is discussed together with its impact in the DRM high pH values, which provided stable catalysts for TRM. The reflux at
reaction, one of the main reactions in TRM process. To the best of our relative short time was also crucial to increase the Ni dispersion within
knowledge, no work on isolated metal atoms in TRM has been reported the CaO-ZrO2 support. The optimal conditions for the catalyst prepara­
in the open literature thus far. tion encompassed a 10–12 pH treatment, 24 h of reflux, and further
Despite the technical simplicity of the impregnation process, this calcination at 700 ◦ C. The as-prepared catalytic materials enabled TRM
methodology provides limited control on the particle size of the active (reaction conditions: 700 ◦ C, molar ratio of feed CH4:CO2:O2:H2O =
phase. This major drawback frequently renders the catalysts prepared by 1:0.5:0.375:0.25, Gas Hourly Space Velocity (GHSV) = 34,000 mL h− 1
impregnation less efficient than those prepared by other methods, e.g. g−cat1) with up to 70 % methane conversion [41].
hydrothermal or precipitation methods [34]. Majewski and Wood [42] adopted the Stöber cum precipitation

Fig. 2. a) Conversion of CH4 and CO2 (average value from 4 h reaction) over Ni@SiO2 catalyst at different temperatures. Reaction conditions: molar ratio of the feed
CH4:CO2:H2O:O2:He = 1:0.5:0.5:0.1:0.4, CH4 gas flow rate 25 mL min− 1, 0.2 g of catalyst. b) and c) Scanning Electron Microscopy (SEM) micrographs of the Ni@SiO2
core@shell catalyst after 4 h reaction at b) 550 ◦ C, and c) 750 ◦ C. Reprinted with permission from [42], Copyright (2019) Elsevier.

3
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

method to obtain a core@shell structured Ni/SiO2 catalyst, in which reduction degree (78 % with Ni-Mg-Al_me@900), and thus increases
silica spheres were covered by a Ni-film. The as-prepared 11 wt.% both CH4 and CO2 conversion rates. Thus, both doped catalysts exhibi­
Ni@SiO2 catalyst exhibited enhanced catalytic properties (Fig. 2a). The ted higher catalytic activity and stability in comparison with non-doped
catalyst produced whisker or carbon nanotubes at 550 ◦ C in TRM, while catalysts. In the case of Ni-Cu-Mg-Al catalyst, Ni-Cu alloy was formed
such carbon accumulation was precluded at 750 ◦ C. However, particle which reduced the size of Ni crystallites (6.5 nm by XRD analysis, which
cluster formation was slightly high at high temperature, as shown in is the smallest value among the five as-prepared catalysts) and occupied
Fig. 2 b and c. kink and edge Ni sites, explaining its high catalytic activity, due to the
Kumar and Pant [29] developed different Ni-based catalysts sup­ small Ni particles which favors reforming reactions. However, the
ported on hydrotalcite-derived Mg-Al mixed oxide, with and without Cu characterization of the used Ni-Cu-Mg-Al catalyst revealed a
or Zn acting as promoters, via the co-precipitation method. First, a re-oxidation of Ni-Cu alloy, explaining a slight catalytic deactivation
Ni-Mg-Al catalyst was prepared by adding Na2CO3 to an aqueous solu­ (Fig. 3). In the case of Ni-Zn-Mg-Al, Ni-Zn alloy was also formed but with
tion of nitrates salts of Ni(II), Mg(II) and Al(III) at ambient conditions. a higher electron density on Ni, as observed by XPS in comparison with
Then, Cu doped Ni-Mg-Al (Ni-Cu-Mg-Al) and Zn-doped Ni-Mg-AL Ni-Cu-Mg-Al catalyst. This fact explained the highest catalytic activity of
(Ni-Zn-Mg-Al) catalysts were prepared by the same method using the Ni-Zn-Mg-Al catalyst. In addition, zinc-doped catalyst was not
appropriate nitrate salt Cu(NO3)2 and Zn(NO3)2. Additionally, two other re-oxidized during TRM, in agreement with its good catalytic stability
Ni-Mg-Al catalysts were prepared by the “memory method”. The latter is [29].
a sequential approach, which consisted of the co-precipitation of Mg-Al
support, followed by air calcination at 500 or 900 ◦ C. Then, Ni deposi­ 2.1.3. Hydrothermal/solvothermal synthesis
tion on these supports was carried out to obtain Ni-Mg-Al_me@500 and The hydrothermal treatment is an advanced catalyst synthesis
Ni-Mg-Al_me@900 catalysts. As showed in Fig. 3, despite its highest methodology, which is generally carried out under controlled temper­
specific surface area (132 m2 g− 1) and its highest basic site density (107 ature and pressure conditions [43]. The solvothermal approach is
μmol g− 1), the non-doped Ni-Mg-Al catalyst performed the lowest cat­ similar to the hydrothermal one, but involving organic solvents instead
alytic performance, due to its lowest degree reduction of Ni (only 47 %). of water [44]. A stainless-steel sealed autoclave is needed to withstand
Changing the catalyst preparation to the memory method improves Ni high pressure upon heating the solvent above its boiling point. This
single-step high temperature-pressure procedure generally results in
highly-crystalline materials. Key factors of this process include precursor
concentration, solvent, stabilizing agent, reaction time, and tempera­
ture. Singha et al. [34] prepared Ni/ZrO2 catalysts using cetyl­
trimethylammonium bromide, Zr(IV) propoxide, and Ni(II) nitrate
hexahydrate as metal precursors under hydrothermal conditions (180 ◦ C
for 24 h in a sealed autoclave, pH 12), followed by filtration, drying at
100 ◦ C for 12 h and air calcination at 600 ◦ C for a further 6 h. The
resulting catalyst exhibited a uniform morphology with 54 m2g− 1 sur­
face area, 10− 40 nm Ni particle size. This catalyst had also a better
Ni-dispersion, and better MSI in comparison with a catalyst prepared by
impregnation. Consequently, this as-prepared catalyst revealed superior
performance in terms of conversion and H2/CO ratio at relative low
reaction temperatures and gave improved energy efficiency, compared
with other catalysts obtained via the impregnation method. Moreover,
the impregnated catalyst rapidly deactivated, which was attributed to a
surface area decline by a rapid particle agglomeration at high temper­
ature. The authors claimed that both the poorly-dispersed Ni and the
particle agglomeration in the impregnated catalyst favor reverse WGS
reaction, thus reducing the H2/CO ratio [34].
Singha and co-workers [45] extended their investigation to Ni-M­
gO-CeO2-ZrO2 composite nanoporous catalysts. They developed a new
nanoporous CeO2-ZrO2 support by a straightforward solvothermal
method. Ethanol solutions of zirconium isopropoxide, cerium nitrate
and magnesium nitrate were mixed at 28 ◦ C under stirring for 8 h, fol­
lowed by drying at 60 ◦ C and calcination at 400 ◦ C for 5 h. The formed Ni
nanoclusters were deposited on the surface of these supports by the
deposition-precipitation method using urea as precipitating agent. After
studying a series of catalysts with different compositions and supports,
they concluded that the observed catalytic performance can be
explained by a combined contribution of multiple factors such as: spe­
cific surface area, particle size, metal dispersion, oxygen storage ca­
pacity, metal support interaction, and acidic/basic properties of the
catalysts. In TRM reaction (reaction conditions: pressure (1 atm), 800
◦
C, GHSV = 20,000 mL g− 1 h− 1, feed molar ratio O2:CO2:H2O:CH4:He =
1:1:2.1:5:18), they observed the highest performance by the 5Ni-Mg­
CeZr catalyst, encompassing 95 % CH4 conversion with a H2:CO molar
ratio near to 2, and enhanced catalytic stability during 100 h of reaction.
Fig. 3. Conversion rates of CH4 (A) and CO2 (B) in TRM over (a) Ni-Mg-Al; (b)
Ni-Mg-Al_me@500, (c) Ni-Mg-Al_me@900, (d) Ni-Cu-Mg-Al and (e) Ni-Zn-Mg- This catalytic stability is basically attributed to the extended MSI, small
Al. Reaction conditions: molar ratio of CH4:CO2:H2O:O2:N2 = Ni particle size (undetectable by XRD even at high Ni loading of 15 wt
1:0.23:0.46:0.07:0.28, GHSV = 49,200 mL h− 1 g− 1, 800 ◦ C, 1 bar. Reprinted %), and the promoter effect of Mg to prevent coke formation. It is
from [29], Copyright (2020), with permission from Elsevier. noteworthy to highlight that the solvothermal method employed

4
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

provided the means to tune the characteristics of catalytic materials by density of oxygen vacancy sites [53]. In TRM at 800 ◦ C and 1 bar, these
adjusting the synthesis parameters [45]. catalysts showed high methane conversion (up to ca. 96 % [53]) and a
In DRM, Das et al. [46] prepared surface-modified mesoporous good stability during a long time-on-stream of ca. 150 h [54].
alumina support by solvothermal approach using a mixture of ethanol To-date, the works of Pino’s group discussed above seemed to be the
and HNO3 acid as solvent mixture, aluminum isopropoxide as Al pre­ first ones devoted to TRM. Taking into account the interesting results
cursor, and cerium and magnesium nitrate salts as promoter precursors. obtained, further investigation would be conducted on this synthesis
A copolymer, poly(ethylene glycol)–block–poly(propylene glycol)– way to explore its potential in catalyst development for TRM.
block–poly(ethylene glycol), was also used as template to build the
mesoporous structure of the support. The synthesis was carried out 2.1.6. Polymerized complex route
under soft conditions (28 ◦ C, stirring overnight), followed by an air This method was applied to the preparation of nickel-based perov­
calcination at 700 ◦ C for 6 h. The resulting mesoporous support favored skite-type NdM0.25Ni0.75O3 (M = Cr, Fe) catalysts for TRM [55]. First, a
the formation of well-dispersed and small Ni particles (4− 5 nm) by mixture of nitrate salts of metals is prepared in deionized water at
nickel nitrate preparation using urea in an aqueous medium. After a desired molar ratio. Then, citric acid is added to this mixture at the
reduction under H2 at 800 ◦ C, the catalysts showed a high catalytic molar ratio of citric acid / metals = 1.1 / 1 for complexation of metal
activity (> 90 % of CH4 and CO2 conversion) in DRM at 800 ◦ C, feed cations at 60 ◦ C for 30 min. After that, ethylene glycol is added and the
composition CH4:CO2:He = 1:1:8, 34 mL min− 1 gas flow rate; 0.2 g temperature of the mixture is adjusted to 90 ◦ C to perform the
catalyst. The highest turnover frequency (TOF) reached ca. 1.57 and poly-esterification reaction, leading to the formation of a polymeric
1.40 s− 1 for respectively CH4 and CO2 conversion. The prepared cata­ resin. After drying at 110 ◦ C for 16 h, the resin is finally calcined at 500
lysts showed also a good stability at 700 ◦ C during ca. 100 h of ◦
C for 30 min, then 800 ◦ C for 4 h. This synthesis allowed obtaining
time-on-stream, thanks to a strong MSI, as revealed by TPR with catalysts having perovskite-like structure with the presence of nickel
reduction peaks above 450 ◦ C. oxide particles, which can be reduced into metallic nickel under
To sum up, hydrothermal/solvothermal methodology seems to be an hydrogen. In TRM at 850 ◦ C and 1 bar, the prepared catalysts showed a
adequate approach to prepare Ni supported catalysts with high Ni relatively good catalytic performance with methane conversion in the
dispersion. However, to date, this methodology is scarcely applied for range of ca. 55–70 % without notable catalyst deactivation during 14 h
TRM process. of time-on-stream. However, it is noticed that this catalyst structure
requires much higher Ni content in comparison with Ni-supported
2.1.4. Microemulsion catalyst, without outstanding catalytic properties, which might be a
Oil-in-water and water-in-oil represent the two of the most widely weakness of this preparation method.
applied microemulsion-based synthetic routes [47]. Water, oil, surfac­
tant, and a co-surfactant are the main components of a microemulsion 2.1.7. Effect of catalyst preparation methods
system. The structural features of a catalyst can efficiently be controlled Catalyst synthesis method along with other preparation parameters
with this methodology, which renders more homogeneous metal parti­ (e.g. nickel precursor, calcination temperature, etc.) can impact param­
cles size and distribution of the catalytic materials than the ones ob­ eters such as metal loading, metal particle size, metal dispersion,
tained by conventional approaches. However, this approach is seldom strength of MSI etc., and thus, the performance of a catalyst in TRM.
the one subsequently applied to TRM. Kim et al. [48] synthesized a Singha et al. [34] examined the effect of catalyst synthesis method
multi-yolk-shell nanotube structured NiCe@SiO2 catalyst via a reverse toward TRM activity (reaction conditions: 60 mg of catalyst, molar ratio
microemulsion method. The Ni-Ce-yolk of the prepared catalyst was of O2:CO2:H2O:CH4:He = 1:1:2.1:5:18, 500–800 ◦ C, reaction time =6 h,
confined inside a SiO2 shell, leading to a remarkable metal-support GHSV = 20,000–400,000 mL g− 1 h− 1). Using a hydrothermal approach,
interaction. TPR analyses confirmed that the multi–yolk–shell nano­ they prepared ZrO2 support (denoted as ZrOHT 2 ) and different Ni/ZrO2
tube catalyst showed ca. 20 times higher hydrogen consumption. Cata­ catalysts (denoted as xNi/ZrOHT
2 , where x means Ni loading equal to 2.3;
lysts with small yolk showed enhanced resistance to carbon deposition 4.8 and 9.5 wt%). By impregnation method, two catalysts containing 5
by instantaneous solid carbon oxidation. Furthermore, yolks larger than Imp
wt% Ni were prepared using ZrOHT
2 support (denoted as 5NiZrO2 ) and a
30 nm exhibited stable activity at high oxidizer-methane feed ratio.
Overall, the major advantage of the microemulsion method over con­ commercial ZrO2 support (denoted as 5Ni/ZrOImpCom 2 ). These catalysts
ventional synthesis approaches is the convenient control of catalyst were investigated in TRM reaction at 500− 800 ◦ C. As highlighted in
morphology and appropriate MSI [48]. Fig. 4 for three catalysts containing the similar Ni loading of 4.8− 5 wt%,
Several works related to the synthesis of catalysts prepared by the catalyst prepared by the hydrothermal method was much more
microemulsion for DRM reaction have been reported [49,50]. However, active than those prepared by the impregnation method in the temper­
the work of Kim et al. is thus far the only one featuring the synthesis of ature range investigated [34]. The impregnated catalysts were inactive
TRM catalysts by microemulsion methods. Thus, more research is below 600 ◦ C, and despite their initial activity above 600 ◦ C, they
needed to better explore the potential of this synthesis methodology. rapidly deactivated due to the thermal sintering of Ni-containing phases.
On the other hand, the catalyst prepared by a hydrothermal method
2.1.5. Combustion synthesis exhibited high and stable catalytic performance with good selectivity
This method was mostly applied by Pino’s group for the preparation towards syngas. Authors claimed that catalyst synthesis methods (and
of different nickel containing catalysts towards TRM, such as Ni/CeO2 also the metal loading) can tune the surface and metal-support inter­
[51,52], and Ni/CeO2 doped with different amounts of La [53,54]. In action features, hence directly influencing catalytic behavior. In fact,
this method, a mixture of (NH4)2Ce(NO3)6, Ni(NO3)2⋅6H2O and oxa­ when hydrothermal method was adopted, the tetragonal ZrO2 with
lyldihydrazide (fuel) was used. A precursor of promoter such as La (111) plane was the major exposed surface for feed gases which may
(NO3)2⋅6H2O can also be added to the mixture. Then, the mixture was explain their improved catalytic performance compared to catalysts
homogenized by ball milling with a minimum amount of water, before prepared by impregnation, where this phase was absent. Besides, the
being introduced to a muffle furnace at 350 ◦ C to give a flaming com­ impregnation method resulted in large NiO crystallites, which was
bustion. After the complete elimination of water, the mixture was confirmed by highly intense and sharp NiO peak as evidenced by XRD,
ignited to burn with a flame leaving a solid sponge form. The latter is pointing to the presence of large particles and consequently a decrease
crushed prior to TRM experiments, yielding small Ni particles (ca. 9 nm of the specific surface area of the catalyst, together with a decrease in Ni
[53]) on the surface of the support. The addition of La as promoter can dispersion [34]. In addition, compared to impregnation materials, the
not only decrease Ni particle size to ca. 3 nm, but also increase the catalysts prepared by hydrothermal method had higher specific surface

5
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Fig. 4. Effect of catalyst preparation method on (a) CH4 conversion, (b) CO2 conversion, (c) H2O conversions and (d) H2/CO ratios. Reaction conditions: reaction
time = 6 h, temperature = 500–800 ◦ C, GHSV = 80,000 mL g− 1 h− 1, molar ratio of O2:CO2:H2O:CH4:He = 1:1:2.1:5:18 [34].

areas, and showed a better reducibility of NiO species, which contrib­ species offered by the reverse microemulsion method. Combustion
uted to their higher global catalytic efficiency in TRM reaction. Also, as synthesis method also led to high performing TRM catalyst (Table 1)
shown in Table 1, Anchieta et al. [59] proved that a 5 wt%Ni/ZrO2 [53].
catalyst, prepared by impregnation method, exhibited lower catalytic Walker et al. [35] investigated the effect of the synthesis method (wet
performance (at the same reaction temperature and with a smaller impregnation and deposition precipitation) on the catalytic perfor­
GHSV) in comparison with 4.8Ni/ZrOHT 2 catalyst, prepared by hydro­
mance of Ni–MgO–(Ce,Zr)O2 catalysts in TRM. CeO2-ZrO2 supports
thermal method [34]. Similarly, NiCe/SiO2 prepared by reverse micro­ (denoted as (Ce,Zr)O2) were first prepared by co-precipitation. Then, Ni
emulsion exhibited higher catalytic performance than the counterpart and Mg loading on these supports were performed by wet impregnation
prepared by impregnation, despite the higher Ni dispersion of the latter and deposition-precipitation. The wet impregnation method enhanced
(Table 1) [48]. This is explained by a controlled morphology (multi­ the Mg loading compared to the deposition-precipitation approach.
–yolk–shell nanotube) and synergetic interactions of Ni–Ce and Ni–Si Consequently, the catalyst prepared by wet impregnation had higher Mg

Table 1
Catalytic TRM Ni–ZrO2 catalysts prepared by different methods; − : data not available; SSA: specific surface area.
Catalyst Preparation method SSA Ni Ni mean Reaction T GHSV CH4 CO2 H2O H2/CO Ref.
(m2 dispersion particle size (◦ C) (mL g− 1
conv. conv. conv. molar
g− 1) (%) (nm) h− 1) (%) (%) (%) ratio

2.3 wt.%Ni/ZrOHT
2
Hydrothermal 45.5 18.8 5.87 800 80,000 98.3 93.1 94.5 1.99

9.5 wt.%Ni/ZrOHT
2 Hydrothermal 52.3 14.6 8.89 800 80,000 98.4 97.7 97.6 1.98
4.8 wt.%Ni/ZrOHT
2 Hydrothermal 54.0 18.2 6.23 800 80,000 98.5 98.1 98.6 1.99 [34]
5 wt.%Ni/ZrOImp
2 Impregnation 10.2 3.7 26.45 800 80,000 84.4 89.3 39.9 1.62
5 wt.%Ni/ZrOImpCom
2 Impregnation 7.9 2.7 33.24 800 80,000 45.2 36.3 21.9 1.53
Precipitation cum
5 wt.%Ni/ZrO2 17.0 − 14.5 800 45,000 78.0 40.0 − 1.20 [59]
wet impregnation
NiCe/SiO2 Impregnation 486.0 1.71 59.1 750 60,000 78 72 − 1.7 [48]
(Ni: 7.2− 8.2 wt%)
(Ce: 4.6− 5.3 wt%)
NiCe/SiO2-multi- Reverse
400.3 0.31 327.3 750 60,000 79 75 − 1.7 [48]
yolk-shell microemulsion
(Ni: 7.2− 8.2 wt%)
(Ce: 4.6− 5.3 wt%)
Combustion
5 wt.%Ni/La–CeO2 − − − 800 30,000 96.0 86.5 − 1.7 [53]
synthesis
Combustion
5 wt.%Ni/CeO2 − − − 800 30,000 93.0 83.0 − 1.7 [53]
synthesis

6
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

loading, which in turn improved the basicity of the catalyst, increased 2.2. Effects of the nature of the active phase
CO2 adsorption on the surface catalyst, and favored CO2 conversion.
In addition to the synthesis methodology, tuning parameters (e.g. The commonly used catalyst for TRM reaction is that constituted of
nickel precursor, calcination temperature, etc.) can improve catalyst zero-valent nickel particles supported on different types of supports. The
performance. García-Vargas et al. [56] compared Ni/CeO2 and Ni/SiC catalytic characteristics of d- and f-block elements in SRM, POM and
catalysts prepared from different nickel salts (nitrate, acetate, chloride DRM are well-known [5,58]. Hence, researchers are very watchful when
and citrate) by impregnation method. For both supports used, nickel deciding the active phase for TRM purposes as the risk of re-oxidation of
chloride and nickel citrate precursors resulted in catalysts with larger Ni low valent catalysts by oxygen in the feed stream is a major concern
particles (92− 116 nm on CeO2, 70− 71 nm on SiO2), in comparison with [30]. To the best of our knowledge, a comparative assessment of the
the catalyst prepared from nickel nitrate and nickel acetate (57− 87 nm catalytic performance of different metal-based catalysts has not yet been
on CeO2, 50− 52 nm on SiO2), which in turn led to a lower catalytic reported. Table 2 compares the activity, selectivity, and stability of the
behavior in TRM. For a given nickel precursor, the catalysts prepared active phase studied in the literature for TRM. A full compilation of
with SiC support were more active and more stable than the counter­ results on Ni-based catalysts in TRM is provided in the Supplementary
parts prepared with CeO2 support. This fact is explained by the nature of Information (SI. 1).
SiC support, resulting in smaller Ni particles and stronger MSI in com­ Kozonoe et al. [30] studied Fe-based catalyst for TRM using
parison with the catalysts prepared with CeO2 support. multi-walled carbon nanotubes (MWCNT) as catalyst support. Fe
MSI is also a key parameter to control catalyst performance, and can deposition (5 wt %) inside the pores of MWCNT support was performed
be tuned by selecting an appropriate support and thermal pretreatment by impregnating this support with a solution of iron nitrate in etha­
conditions during the catalyst preparation [41,57]. Kumar et al. [57] nol/water, followed by drying at 50 ◦ C for 16 h, air calcination at 350 ◦ C
studied Ni catalysts supported on different supports (TiO2, SBA-15, for 2 h and H2 reduction at 400 ◦ C for 2 h. Then, Co and/or Cu deposition
MgO, and Al2O3) applied in TRM. These catalysts were prepared by (5 wt%) outside of the above Fe@MWCNT material by the similar
wet impregnation technique. After the impregnation step, the residual impregnation approach in order to obtain Fe@MWCNT/Co,
water was removed under vacuum and the resulting solid was dried at Fe@MWCNT/Cu, and Fe@MWCNT/CoCu catalysts. The
110 ◦ C for 15 h, and further calcined at 400− 950 ◦ C for 5 h (10 ◦ C/min). Fe@MWCNT/Cu catalyst exhibited the highest methane conversion (62
The calcination temperature induced a remarkable impact on MSI: %, Table 2), explained by the highest metal reduction degree of this
relative low calcination temperature led to weak MSI, while high tem­ catalyst (98 %) in comparison with that of Fe@MWCNT/Co (53 %) and
perature gave stronger MSI. For instant, Ni/MgO@400 (calcined at 400 Fe@MWCNT/CoCu (50 %). Furthermore, the Fe@MWCNT/Cu catalyst
◦
C) had weaker MSI than Ni/MgO@850 (calcined at 850 ◦ C), as revealed had a very good catalytic stability, while the Fe@MWCNT/CoCu cata­
by TPR analyses. Consequently, most of Ni in Ni/MgO@850 was not lyst showed an initial catalytic deactivation. A remarkable difference in
reduced and thus inactive in TRM. On the other hand, most of Ni in the H2/CO molar ratio was observed with the activity of
Ni/MgO@400 could be reduced and was active in TRM. However, under Fe@MWCNT/Cu compared to that with Fe@MWCNT/CoCu (Table 2).
the same TRM conditions (CH4:CO2:H2O:O2:N2 = 1:0.23:0.46:0.07:0.28, The significant variation, up to a factor of 2, was attributed to the in­
space velocity =17,220 mL h− 1 g− 1, 800 ◦ C, 1 bar), Ni/MgO@400 fluence of metal nature and feed composition. In all cases, the H2/CO
exhibited higher deactivation rate than Ni/MgO@850, despite its higher molar ratio below 1 indicated the occurrence of water involved in
initial activity. This is partially due to Ni sintering. In the case of side-reactions such as RWGS. It is worth to note that Fe@MWCNT/CuCo
Ni/Al2O3, TPR analyses showed the presence of several reduction peaks favors C–O bond cleavage to produce dehydration products and hence
at ca. 270− 900 ◦ C for Ni/Al2O3@400, corresponding to both weak and resulting in low CO production.
strong MSI; only one broad reduction peak around 750− 900 ◦ C for After TRM experiments, the structure of the catalyst was completely
Ni/Al2O3@950, corresponding to strong MSI. In both cases, high extent modified. MWCNT was destroyed and only less than 15 wt% of carbon
of Ni reduction was reached (89 and 82 % for the catalysts calcined remained in the used catalyst, while new metal oxides were formed (e.g.
respectively at 400 and 950 ◦ C). In TRM reaction (CH4:CO2:H2O:O2:N2 α-Fe2O3, γ-Fe2O3, CuO, Cu2O and Co3O4). The Fe and Cu particles
= 1:0.23:0.46:0.07:0.28, space velocity =17,220 mL h− 1 g− 1, 800 ◦ C, 1 remained small in the used Cu or Co doped catalysts (Fig. 5). However,
bar), Ni/Al2O3@950 was consequently more stable than the morphology of the bimetallic Fe@MWCNT/CoCu catalyst changed
Ni/Al2O3@400, due to strong MSI in Ni/Al2O3@950, which limited and exhibited a heterogeneous surface with the presence of large
thermal sintering. Surprisingly, Ni/Al2O3@950 was also much more particles.
active than Ni/Al2O3@400. This result is explained by the formation of Considering the data shown in Table 2, even though the reaction
monodispersed Ni atom at high calcination temperature. Therefore, it conditions are not identical, the performance of Ni-based catalysts is far
seems that strong MSI is required to limit or prevent thermal sintering superior to that observed for Fe-based catalysts. For example, while
but this parameter must be controlled together with other properties almost all Fe-based catalysts performed a CH4 conversion in 40–60 %
such as Ni reducibility and dispersion, which depend on the nature of range, the Ni-based catalysts led to higher CH4 conversion (up to 98 %).
each support (for a detailed discussion on the nature of supports, see Especially, under TRM reaction conditions (fixed-bed down flow
Section 3). reactor: 60 mg of catalyst, 1 atm, 800 ◦ C, feed composition of O2:CO2:
H2O:CH4:He = 1:1:2.1:5:18 with GHSV = 80,000 mL g− 1 h− 1), the Ni/
2.1.8. Conclusions ZrO2 catalyst prepared by hydrothermal treatment did not exhibit
To sum up, this section summarizes in a critical way the different deactivation for 100 h, obtaining nearly full CH4 conversion [34]. This
methodologies for the synthesis of catalysts applied in TRM process outstanding catalyst performance is explained by a high Ni dispersion
reported to date. The impregnation method is the simplest way and can (18.2 %, with mean Ni particle size of 6.2 nm) and a strong MSI (TPR
be used for a given adequate support (see Section 3). Hydrothermal peaks around 340 and 420 ◦ C) of this catalyst. In another work, the
treatment and precipitation methods both provide the means to effec­ influence of the ionic liquid on the morphology, acido-basicity, and Ni
tively tailor the catalyst surface properties to enhance TRM efficiencies. dispersion was investigated on the catalytic performance of Ni/ZrO2
It is worth to note that the hydrothermal treatment can provide highly- [59]. The use of an ionic liquid during the catalyst preparation by
crystalline materials, despite its lower yields in comparison to precipi­ impregnation reduced the number of strong acid sites of the support,
tation approaches [43]. Hitherto, the microemulsion method and com­ which improved the catalyst efficiency in TRM.
bustion synthesis are rarely implemented in TRM research, despite the As previously mentioned in [61], it is important to highlight that to
promising initial results reported in the literature, and thus merit to be date, only few works have been devoted to noble metal catalysts for TRM
explored in future work. (see Supplementary Information (SI. 1)).

7
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Table 2
Comparison of the activity, selectivity, and stability of the catalytically-active phase studied in TRM.
Initial conversion
Specific surface Degree of GHSV (mL T Time on H2/CO
Catalysts XCH4 XCO2 XH2O Ref.
area (m2 g− 1) reduction (%) g− 1 h− 1) (◦ C) stream (h) molar ratio
(%) (%) (%)

Fe@MWCNT/Cu 139.3 98.0 60,000 800 27 62.0 38.8 69.0 0.3

Metal loadings: 5 wt% Fe, 5 wt% Cu
Fe@MWCNT/Co 23.5 53.0 63,000 800 27 46.0 36.0 35.0 0.5 [30]
Metal loadings: 5 wt% Fe, 5 wt% Co
Fe@MWCNT/CuCo 23.5 50.0 60,000 800 27 44.2 35.7 65.6 0.7
Metal loadings: 5 wt% Fe, 5 wt% Cu
and Co
4.8 wt.%Ni–ZrO2 HT 54.0 18.2 80,000 800 100 98.5 98.1 98.6 1.99 [34]
Metal loading: 4.8 wt% Ni
Ni–ZrO2–ionic liquid 11.0 84.3 48,000 800 5 39.0 20.5 − 2.4 [59]
Metal loading (5.8 wt% Ni)
Ni/Ce-Zr-Al2O3 151.0 (calcined) 7.3 161.5 h− 1(*) 800 8 98.0 41.0 − 2 [60]
Fresh/calcined catalyst (13.0/10.6
131.0 (reduced)
(Ni) 3.0/2.7 (Ce) 4.0/3.6 (Zr)
Ni/CeZr/MgAl2O4 91.0 100.0 − 750 6 74.0 34.5 − 1.1 [40]
Metal loading (by EDS chemical
analysis): 8 wt% Ni
Ni@SiO2 26.1 − − 750 4 73.0 91.1 − 2 [42]
Metal loading: 11 wt% Ni
WHSV
Ni/Zeolite L (cylindrical 30− 60 nm) 95.32 − 1 800 1.5 82.0 24.0 − 1.9
=161.5 h−
Metal loading: 11.92 wt% Ni [61]
Rh-Ni/zeolite L (cylindrical 30− 60 WHSV
64.12 − 1 800 1.5 97.0 34.0 − 1.6
nm) =161.5 h−
Metal loadings: 1.09 wt% Rh and
13.17 wt% Ni

(− : not mentioned; HT: hydrothermal method; MWCNT: multi-walled carbon nanotubes; *: weight hourly space velocity).

2.3. Effects of metal loading superior performance over other compared catalysts, and the reason was
attributed to the promotion effect of Zn via electron transfer to enrich Ni
Even though the active phase is required in TRM catalysis, its excess (0) [29]. Various promoters were investigated in TRM for improving the
may result in the agglomeration to form giant particles that in combi­ stability and activity of the active phase (see Section 4).
nation with poor dispersion can even promote the reverse WGS reaction,
thus lowering the H2/CO ratio. The number of accessible active sites,
2.4. New trend in the engineering of the active phase in TRM
their size and reducibility are the major factors to control the activity
and selectivity [62,63]. Generally, small metal particles favor methane
Recently, advances in material science open a new frontier in het­
reforming [63,64]. Hence, the optimization of the active phase is rele­
erogeneous catalysis, related to single atom catalysis [66,67]. In single
vant. Singha et al. [34] synthesized Ni/ZrO2 catalysts containing 2.3, 4.8
atom catalysts (SAC), isolated metal atoms, singly-dispersed on a sup­
and 9.5 wt% Ni by hydrothermal method using cetyl­
port, constitute the active phase. This specific structural feature allows
trimethylammonium bromide as a cationic surfactant affording catalysts
maximizing the efficiency of metal atom utilization, along with
which exhibit specific surface areas of 45.5, 54.0 and 52.3 m2 g− 1, and
providing uniform active sites [68]. First, in a SAC, all metal atoms can
average Ni particle sizes of 5.87, 6.23 and 8.89 nm for respectively 2.3,
be exposed to reactants instead of being buried in nanoparticles or bulk
4.8 and 9.5 wt% Ni loading. Thus, the increase of Ni loading led to the
metal. Second, this specific structure provides a low-coordination
increase of the average Ni particle size and subsequently a decreased H2
environment of metal centers, which promotes adsorption ability and
selectivity in TRM reaction (Fig. 6), which is attributed to the favored
enhances catalytic performance. Third, quantum size effects can be
RWGS over large Ni-particles. Among the prepared catalysts, the one
fostered since the confinement of electrons leads to a discrete energy
having 4.8 wt% Ni loading exhibited the highest catalyst performance in
level distribution and a distinctive HOMO-LUMO gap. Furthermore,
TRM reaction, explained by a high Ni dispersion and a strong MSI (see
chemical bonding effects and charge transfer between metal and sup­
Section 2.2. for the effects of the nature of the active phase on TRM
ports and associated interfaces can be enhanced, which allows regu­
process).
lating MSI [66].
Singha et al. [45] also investigated MgCeZr-supported Ni catalysts
In methane reforming reactions, small metal particles play a crucial
containing 2.5, 5 and 10 wt% of Ni. The catalyst 5 wt%Ni/MgCeZr
role [69], as reported by Tang et al. [70] pointing to 10 nm of mean
showed the best catalytic performance given its enhanced reducibility
diameter as a threshold of Ni nanoparticle size for the intrinsic carbon
(in comparison with 10 wt%Ni/MgCeZr) and its large availability of
deposit limitation. Recently, Ni-based SACs have been applied in DRM
active sites (in comparison with 2 wt%Ni/MgCeZr) [45]. Pino et al. [65]
[71–74]. Even being considered as thermodynamically-unstable
studied Ni-La-CeO2 catalysts containing 1.76, 3.66, 7.88 and 10.25 wt%
because of high surface free energy [66] (Fig. 7), SAC demonstrated
Ni in TRM (500 mg catalyst, 800 ◦ C, 1 bar, GHSV = 30,000 h− 1, and
high catalytic activities and stability under DRM conditions [71,72,74].
molar ratio of CH4:CO2:H2O:O2 = 1:0.46:0.46:0.1). The catalysts having
Thus, Akri et al. [71] successfully prepared and tested
1.76 and 3.66 wt% Ni showed similar CH4 (ca. 98 %) and CO2 (ca. 85–87
atomically-dispersed Ni single atoms supported on Ce-doped hydroxy­
%) conversions, which were higher than those obtained with the cata­
apatite catalysts (containing 0.5, 1 and 2 wt.% of Ni) in DRM. SACs
lysts containing 7.88 and 10.25 wt% Ni. The authors explained this
showed higher activity, stability and coke-resistance compared to con­
result by the low Ni dispersion at high Ni loading.
ventional nanoparticle-based catalyst. Notably, SACs only activate the
In addition to the role of the active phase, metals were also exten­
first C–H bond in CH4 and do not promote methane cracking, explaining
sively studied as catalyst promoters in TRM. Ni/Zn-Mg-Al exhibited
their excellent coke-resistance. Interesting results in DRM were also

8
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Their performance is illustrated in Scheme 2 on the basis of different

physico-chemical and thermal properties, and will be discussed
hereafter.

3.1. Alumina

Alumina possesses several properties required for a convenient TRM

catalyst support, such as high specific surface area, high thermal sta­
bility and low cost among others. In TRM, Ni/Al2O3 catalysts have
intensively been examined [19,28,76–79]. Solov’ev et al. [19] per­
formed TRM using Ni/Al2O3 catalysts prepared from structured cordi­
erite with and without additives (CeO2 and La2O3). Catalysts without
additives showed low CH4 conversion (ca. 15–20 %) at 700− 800 ◦ C
under various TRM conditions (molar ratio of CH4:CO2:O2:H2:Ar =
1:0.75:0.35:0.75:15; 1:0.65:0.3:0.75:16; 1:0.7:0.2:0.8:17;
1:0.9:0.25:0.75:15; 1:0.9:0.3:0.7:14; 1:0.55:0.2:1.0:16; WHSV = 12,000
h− 1) due to the re-oxidation of Ni(0) to NiO. The addition of promoters
overcame this problem, reaching up to 100 % of methane conversion in
the presence of La2O3 at ca. 650 ◦ C, and nearly 95 % of methane con­
version using CeO2 as additive at ca. 700 ◦ C.
By controlling the synthesis conditions of Al2O3 supports, Yoo et al.
[80] evidenced the beneficial effect of high specific surface area and
large pore volume in TRM reaction. Ni/Al2O3, containing 20 wt% Ni
were prepared by sol-gel method, with or without a supercritical carbon
dioxide drying step resulting respectively in mesoporous nickel–alumina
aerogel catalyst (NAA) and mesoporous nickel–alumina xerogel catalyst
(NAX). CO2 drying allowed increasing the specific surface area (370 and
320 m2 g− 1 for NAA and NAX, respectively) and average pore size (12.7
and 7.2 nm for NAA and NAX, respectively). Consequently, NAA
exhibited higher CH4 conversion and higher H2 yield than NAX, which
correlated with the higher nickel dispersion of the former.
In the Ni/Al2O3 catalyst described by Jiang et al. [81], three different
Ni species can exist as evidenced by TRP analysis (Fig. 9): free NiO (peak
at ca. 400 ◦ C); NiO interacting strongly with the support (peak at ca. 600
◦
C); and spinel NiA12O4 (peak at ca. 820 ◦ C), as evidenced by TPR. The
reduction of the first two Ni species are generally possible under H2
resulting in the formation of zero-valent Ni which is active in methane
Fig. 5. SEM images of catalysts after TRM. Top: Fe@MWCNT/Cu - a) 10000x,
b) 40000x, c) EDS mapping; Bottom: Fe@MWCNT/CoCu - a) 15000x, b) reforming, albeit spinel NiAl2O4, usually formed at high calcination
30000x, c) EDS mapping. The monometallic catalyst maintained the small temperatures, is inactive in methane reforming [57,77,82], and its
particle size, while the bimetallic catalyst led to agglomerates. Reprinted from content can reach up to 90 % of the nickel species [83]. Thus, even if the
[30], Copyright (2019), with permission from Elsevier. reduction of NiAl2O4 is difficult and requires harsh conditions (tem­
perature ca. 800− 900 ◦ C) [57,81,85], it is strongly recommended to
obtained by Zuo et al. [72] using Ni/MgO SACs, and by Tang et al. [74] appropriately reduce it to Ni(0) in order to optimize the Ni/Al2O3
using Ni/CeO2, Ru/CeO2 and Ni-Ru/CeO2 SACs catalysts. High catalytic catalyst performance [28,57,83–85]. Indeed, NiAl2O4 reduction results
activity and stability of Ni-Ru/CeO2 SACs catalysts were attributed to in well-dispersed Ni particles, which enhance the catalytic activity and
the formation of isolated single atoms of Ni(0) and Ru(0). Nanoparticles limit carbon deposits [57,84,86,87].
were absent, even after DRM reaction at 600 ◦ C (Fig. 8). The develop­
ment of SACs will probably open new prospects for the design of 3.2. Silica
highly-efficient heterogeneous catalysts applied in TRM. The combina­
tion of isolated atoms with small clusters (< 2 nm), which can lead to Silica is generally a catalyst support of low cost, high thermal sta­
“cooperative effects”, as evidenced for other transformations (recently bility, and high specific surface area [42]. Different silica structures have
reviewed by Serp [75]) represents a strategy to be explored towards the been investigated to prepare ordered mesoporous silica supported nickel
development of TRM catalysts. (Ni/SBA-15) [57,84,88], core-shell Ni/SiO2 [42] and multi-yolk-shell
Ni/SiO2 (Fig. 10) [48]. Unlike Ni/Al2O3 catalysts, no metal-support
3. Catalyst supports compound is formed and the catalytic performance mostly depends on
silica morphology [57,84].
Supports play an important role in heterogeneous catalysis. In TRM, NiO species can diffuse and remain confined inside silica pores, and
the following criteria are commonly required for an efficient catalyst their reduction needs high temperatures (ca. 850 ◦ C), leading to small Ni
support: (i) high thermal stability, (ii) high specific surface area for (0) particles, which in turn, enhance catalytic activity. Also, NiO
active phase dispersion, (iii) adequate MSI for metal reduction capacity nanocrystallites located inside micropores need higher reduction tem­
and thermal sintering resistance, (iv) high basicity for adsorption of CO2 peratures than those occluded in mesopores [57,84]. Moreover, the
and coke limitation, (v) frequently presence of oxygen vacancies and confinement of Ni particles in the pores limits their thermal sintering in
high oxygen storage capacity for coke elimination, (vi) and affordable TRM [89]. Furthermore, these pores can serve as channels for reactants
cost. As shown in SI. 1, various supports have been investigated in TRM, to diffuse through, while silica shell-structure can suppress carbon
including: metal oxides, mixed oxides, zeolites, and silicon carbide. filament growth [48]. Notably, the direct contact between Ni clusters
and unreduced or partially reduced Ni ions within the SiO2 matrix on the

9
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Fig. 6. Effect of Ni loading during TRM over Ni–ZrO2 on (a) CH4 conversion, (b) CO2 conversion, (c) H2O conversion, and (d) H2/CO ratio. Reaction conditions:
reaction time = 6 h, temperature = 500–800 ◦ C, GHSV = 80,000 mL g− 1 h− 1, molar ratio of O2:CO2:H2O:CH4:He = 1:1:2.1:5:18. Reprinted from [34], Copyright
(2016), with permission from Elsevier.

3.3. Magnesia

Magnesia displays the advantage of having high thermal stability and

high basicity [42,57,84,93]. However, controversial results have been
reported for this support. Despite Song et al. [27] and Tomishige [94]
stated that Ni/MgO catalysts were highly active for CH4 and CO2 con­
version, other studies differed [57,84,93]. This could be due to the
formation of a NiO-MgO solid solution, requiring particular reduction
conditions. At high temperatures, NiO progressively diffuses into MgO
lattice to form NiO-MgO solid solution at any NiO/MgO ratio, which is
difficult to reduce [27,57,84]. This sensitive reduction can take place at
high temperatures, resulting in highly-dispersed and active Ni(0) par­
ticles [93]. Therefore, Fedorova et al. [93] used porous nickel ribbons
covered with a MgO underlayer as support to deposit nickel particles by
impregnation. After reduction at 900 ◦ C under H2, highly-dispersed
nickel particles (3–5 nm), epitaxially bound to MgO underlayer were
observed. These catalysts exhibited a high activity and stability in TRM
Fig. 7. Surface free energy and specific activity as a function of metal sizes. (75 % CH4 conversion at 750 ◦ C, 1 bar, CH4:CO2:H2O:O2:N2 =
Reprinted with permission from [66]. Copyright (2013) American Chemi­ 24:18:18:3.5:36.5, flow rate of gas mixture = 25 L h− 1, 0.4 g of the
cal Society. catalyst).

wall surface provides an anchoring effect and thus helps to prevent 3.4. Ceria
thermal sintering, as previously evidenced by Quek et al. [90].
Furthermore, the stabilization of Ni nanoparticles inside SiO2 nanotubes Ceria has also been investigated in TRM reaction given its high
has also been recently reported for DRM reaction by Li et al. [91]. On the thermal stability, high density of basic sites, and high oxygen-storage
other hand, when Ni nanoparticles are distributed on the external sur­ capacity and mobility [27,41,51,51,53,56,85].
face of SiO2 support, thermal sintering takes place, causing a Ni particle Ohtake et al. [95] demonstrated the high thermal stability of CeO2
size increase from ca. 2 nm for the fresh catalyst to ca. 14 nm after its use synthesized by hydrothermal method. The specific surface area of this
for 1400 min at 700 ◦ C under DRM conditions [92]. In addition, support was respectively 139, 131, 73 and 50 m2 g− 1 when it was
silica-supported Ni catalysts could also partially loss their specific sur­ calcined at respectively 400, 600, 800 and 900 ◦ C. Thus, even at high
face area, (from 68.1 m2 g− 1 to 21.4− 59.6 m2 g− 1), which affects the temperatures (800− 900 ◦ C), usual temperature for methane reforming,
catalyst stability during the TRM reaction [42]. this support could maintain a relatively high specific surface area (> 50

10
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Fig. 8. Aberration-corrected STEM images of Ni-Ru/CeO2 SAC catalyst recovered after DRM reaction at 600 ◦ C, illustrating the absence of nanoparticles on the
surface of the support. Reprinted with permission from [74]. Copyright (2019) American Chemical Society.

Scheme 2. Illustration of the impact of supports in the design of TRM catalysts.

solution, and thus oxygen vacancies, is favored by increasing nickel

content (range studied: 1.8–31.0 wt%) [51].
Ni can also form stronger O–Ni–O–Ce– bonds in comparison with
Ni− O bond (in NiO crystal) by the insertion of Ni(II) ions in the fluorite-
type structure of CeO2 [96]. For that purpose, CeO2 support was pre­
pared by refluxing an aqueous solution of (NH4)2Ce(NO3)6 at pH 9 using
(NH4)2CO3 under 500 rpm stirring. The resulting precipitate was finally
recovered by centrifugation and dried at 90 ◦ C for 12 h. Ni deposition
(2.5− 10 wt%) was performed by wet impregnation. The as-prepared
CeO2 was added to an aqueous solution of nickel nitrate at 80 ◦ C and
500 rpm. The solvent was eliminated under continuous stirring and the
resulting solid was calcined at 500 ◦ C for 6 h. In the range of Ni loading
studied (2.5− 10 wt%), the catalyst containing more than 2.5 wt% Ni led
to surface defects through the formation of the –O–Ni–O–Ce super
structure over CeO2 crystal. The high activity and stability of these
catalysts in partial methane oxidation into syngas can be explained by
the presence of low-coordinated O atoms [96]. Moreover, the enhanced
MSI in Ni/CeO2 catalysts resulted in small Ni particles in the reduced
Fig. 9. TPR profile of the Ni/Al2O3 catalyst. Reprinted from [81], Copyright catalyst, and thus inducing a high catalytic performance in TRM [51]. In
(2007), with permission from Elsevier. addition, the high density of basic sites in Ni/CeO2 (as determined by
CO2-TPD analysis) enhances CO2 adsorption, which in turn favors coke
m2 g− 1). Also, CeO2 is well known by its high redox capacity, promoted gasification as highlighted by Pino et al. [53].
by Ni sites, creating oxygen vacancies in CeO2 lattice, which in turn can
stabilize the Ni sites [74]. Moreover, they provide new adsorption sites
3.5. Zirconia
for O2, H2O and CO2, thus, enhancing coke gasification [27,51,51]. As
examples, Lo Faro et al. [51] investigated TRM over 1.75 wt.%Ni/CeO2
Zirconia exhibits the required properties for a support such as: high
catalyst. XRD analyses evidenced the insertion of Ni(II) ions in CeO2
thermal stability, high ionic conductivity, high mechanical strength,
lattice, shifting CeO2 diffraction peaks to higher Bragg angles with
high fracture toughness and hardness [34]. In addition, zirconia is often
respect to pure CeO2. NiO-CeO2 solid solution was formed, giving oxy­
selected as catalyst support or promoter thanks to the presence of defects
gen vacancies from its reduction [51]. The formation of NiO-CeO2 solid
on its crystal surface, where oxygen vacancies are easily created.

11
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Fig. 10. Illustration of the core–shell, yolk–shell, and multi–yolk–shell nanotube supported nickel structures. Green: Ni; white: void; orange: SiO2. Reprinted from
[48], Copyright (2019), with permission from Elsevier (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of
this article.).

Moreover, ZrO2 provides low concentration of Lewis acid sites and a [84,98]. In ethanol reforming, TiOx species were able to suppress carbon
moderate MSI with Ni species [34,59]. ZrO2 and ZrO2-based materials deposition and to enhance MSI by the formation of new active sites of
(e.g. yttria-stabilized zirconia - YSZ) have been intensively studied in Ni-O-Ti3+ at the boundary between metallic Ni particles and TiO2 sup­
TRM [34,41,59,56,84,97]. port [99]. For Ni/TiO2 catalysts applied in the TRM reaction (see Section
Singha et al. [34] investigated TRM over Ni/ZrO2 catalysts prepared 2.3), low calcination temperature (400 ◦ C) is preferred, leading to
by hydrothermal method. Average Ni particle size increased with the moderate MSI and suitable conditions for the reduction of these catalysts
increase of Ni loading (optimal value being 4.8 wt%). High nickel [57]. At higher temperature (850 ◦ C), most of Ni was non-reducible and
dispersion and strong MSI explained the excellent activity and stability inactive in TRM [57]. The addition of promoters such as MgO could also
of these catalysts (> 95 % CH4, CO2 and H2O conversions at 800 ◦ C cause a decrease of Ni reduction degree as revealed by Jiang et al. for
without deactivation for more than 100 h) [34]. Anchieta et al. [59] Ni/MgxTi1-xO catalysts (with x = 0 to 1) [98]. Without MgO, Ni could
prepared a ZrO2 support by precipitation method with the use of an ionic completely be reduced under H2 below 650 ◦ C. In the presence of MgO,
liquid (1-hexadecil-3-metilimidazolium bromide). Then, Ni deposition TPR peak temperatures increased while TPR peak surfaces decreased
(5 wt%) was done by IWI method. In TRM reaction (conditions: 850 ◦ C, signifying the increase of non-reducible Ni [98]. Thus, for TiO2-based
molar ratio of CH4:CO2:H2O:O2 = 1.0:0.5:0.5:0.1, total gas flow =185 nickel supported catalysts, it is strongly recommended to determine the
mL min− 1, GHSV = 45,000 mL g− 1 h− 1), this catalyst, having a specific preparation conditions to optimize Ni reduction by H2 reduction.
surface area of 11 m2 g− 1, showed promising catalytic performance with
ca. 85 % CH4 conversion while precluding coke deposition. The use of an
ionic liquid during support preparation tunes acid-basic properties of 3.7. Mixed oxides
the support, and consequently improves the catalytic performance. In
fact, under the same conditions of TRM, the catalyst prepared in the Mixed oxides allow improving physico-chemical properties required
absence of the ionic liquid, showing a specific surface area of 17 m2 g− 1, for TRM catalyst supports. MgO and CaO with high basicity, CeO2 and
gave only ca. 55 % of CH4 conversion. The catalytic performance of ZrO2 with high oxygen storage capacity, and Y2O3 and La2O3 with high
Ni/ZrO2 could be highly improved by the addition of Mg playing the role oxygen lability are particularly targeted in TRM.
of promoter [97]. In this work, Sun et al. [97] synthesized Ni/ZrO2 and MgAl2O4, a common support used in SRM and DRM, was also
Ni/MgO-ZrO2 by co-precipitation method using nitrate precursors of Ni investigated in TRM reaction [40,76,79] thanks to its high specific
(II), Mg(II) and Zr(IV), and a 5 wt% KOH solution. The precipitate was surface area, high thermal sintering resistance and high basicity.
filtered, dried at 110 ◦ C for 12 h and calcined under air at 800 ◦ C for 4 h. Well-dispersed Ni particles formed on MgAl2O4 surface could be stabi­
The specific surface areas obtained were 28 and 39 m2 g− 1 for respec­ lized by SMSI. The presence of Mg in the spinel structure results in a high
tively Ni/ZrO2 and Ni/MgO-ZrO2. Under the same conditions in TRM basicity, that favors CO2 adsorption and coke gasification [40]. Jiang
process (reaction conditions: 0.2 g catalyst, GHSV = 30,000 mL g–1 h–1, et al. [98] combined TiO2 with MgO to obtain MgxTi1− xO (x = 0.25; 0.5;
molar ratio of CH4:CO2:H2O:O2:N2 = 1.0:0.45:0.45:0.1:0.4, 800 ◦ C, 1 0.75) composites. Ni/MgxTi1− xO catalysts exhibited similar Ni disper­
bar), Ni/MgO-ZrO2 led to ca. 98 % CH4 conversion (exhibiting a good sion but higher catalytic performance than Ni/TiO2 and Ni/MgO. The
catalyst stability during 58 h of time-on-stream), while Ni/ZrO2 showed moderate interaction of Ni with MgxTi1− xO support favors both Ni
less than 80 % CH4 conversion. This different catalytic behavior is reducibility and catalyst stability [98]. Walker et al. [35] evaluated the
explained by the presence of MgO, which improved the basicity of the influence of the synthesis methodology on the behavior of Ni–MgO–(Ce,
support and stabilized the tetragonal ZrO2 structure. Zr)O2 catalysts under various TRM reaction conditions. The molar ratios
of Ce/Zr and Ni/Mg, Ni deposition technique, and Ni content were found
as key factors impacting the catalyst performance [35].
3.6. Titania Lee et al. investigated Ni/CeO2-ZrO2-Al2O3 catalysts in methane
reforming processes [100]. The supports CeO2-ZrO2-Al2O3 were pre­
Titania is a support largely used in heterogeneous catalysis. It has pared by IWI method using a commercial θ-Al2O3 support (167 m2 g− 1)
also been applied in TRM reaction by several authors [57,84,98]. The and solutions of Ce(IV) and Zr(IV) precursors. Ni deposition (3 wt%) was
reactivity and stability of Ni/TiO2 depend on the formation and reduc­ also performed by IWI method using an aqueous solution of nickel ni­
tion of NiTiO3 solid solution [57]. The formation of NiTiO3 is favored at trate salt, followed by drying at 100 ◦ C and air calcination at 550 ◦ C for 6
high calcination temperature (850 ◦ C), exhibiting a high stability and h. The resulting catalyst showed very good catalytic performance in
consequently a reluctance towards reduction [57,84,98]. Moreover, TRM at 800 ◦ C and 3 bar affording 96 and 82 % of respectively CH4 and
TiOx species (x < 2) can be formed, that can migrate onto metallic Ni CO2 conversion.
particles, partially covering them and thus decreasing the Ni dispersion Si et al. [41] investigated the impact of co-precipitation conditions on

12
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

the performance of Ni-CaO-ZrO2 catalysts. As previously mentioned (see Ni/CeZrO. In another study working at 800 ◦ C, the following trend was
Section 2.1.2), working under optimal conditions (10-12 pH, 24 h reflux observed: Ni/Al2O3 ~ Ni/CeO2 > Ni/SiC > Ni/YSZ [57]. More recently,
time, and calcination at 700 ◦ C), led to catalysts with specific surface Kumar et al. [57,84] found the following tendency: Ni/Al2O3 >
areas of ca.177− 225 m2 g− 1. The corresponding catalysts showed high Ni/SBA-15 > Ni/ZrO2 > Ni/CeO2-ZrO2 > Ni/TiO2 > Ni/MgO. Song and
methane conversion (up to 70 % at 700 ◦ C and 1 bar; other conditions: Pan [27] showed that Ni/MgO was more active than Ni/Al2O3, which is
0.3 g of catalyst; molar ratio of CH4:CO2:O2:H2O = 1:0.5:0.375:0.25; in contrast to the results found by Kumar et al. [57,84]. These apparent
GHSV = 34,000 mL h− 1 g−cat1) and good stability (practically no loss of inconsistencies between different reports must serve to override
activity during 10 h of reaction). simplistic comparisons and highlight that catalyst efficiency depends on
La-Ce-O mixed oxide (Ce1− 3xLa2xNixO2, x = 0.10; 0.20; and 0.25) many parameters such as synthesis methodology, pretreatment condi­
supported Ni catalysts were studied by Pino et al. [54]. Nickel existed tions, metal particle size, textural and structural properties of supports,
under both zero-valent Ni nanoparticles and cationic Ni(II) inserted in etc.
the cubic fluorite structure of CeO2. Despite the low specific surface area Thus far, most of the conventional supports in heterogeneous catal­
of the synthesized catalysts (< 4 m2 g− 1), high cationic character of Ni ysis have been investigated in TRM and each of them responds to a
(II) along with the combination of Ni(0) in close contact with La-Ce-O greater or lesser extent to the criteria required for TRM catalyst sup­
matrix resulted in high catalytic activities (ca. 1.8 molCH4 s− 1 g−Ni1) and ports. At the current stage of research and development on TRM catalyst,
high catalytic stability while precluding carbon deposition during 6 h at it is difficult to quantitatively compare the different supports studied in
800 ◦ C and atmospheric pressure in TRM process. this reaction. Nevertheless, we have tried to propose a tentative rating
Dong et al. [78] tested various Ni-based catalysts supported on table for the main criteria required for TRM catalyst supports (Table 3).
complex mixtures of ZrO2 doped with Y2O3 and another oxide among Accordingly, the range of potential impact of each criterion for the
CeO2, MgO, SiO2, TiO2, CaO to distort the crystal lattice of ZrO2. These design of TRM catalyst is assigned and rated from 1 to 3 (least preferred
catalysts were able to overcome coke deposition by facilitating oxygen to most preferred). The total score of the rating serves to compare TRM
transfer and increasing oxygen storage and supply. Similar work was catalyst supports.
conducted by Kang et al. [101] who investigated NiO–YSZ–CeO2 cata­ On the basis of the works reported in the literature and the rating in
lytic systems, reaching full conversion of CO2 and CH4 above 800 ◦ C. Table 3, Fig. 11 shows the relative comparison of the supports studied in
TRM. Among the simple oxides, ZrO2 and CeO2 seem to be the best
3.8. Other supports choices. Mixed oxides offer different possibilities to improve physico-
chemical and thermal properties. Thus, for future works on TRM, it is
Carbon-based supports and zeolites are largely studied and applied in recommended to focus on mixed oxides, CeO2 or even ZrO2 to optimize
heterogeneous catalysis, but much less in TRM. To date, Kozonoe et al. the design of performing TRM catalysts.
[30,102] and Izquierdo et al. [61] seemed to be the first teams who
respectively investigated a carbon-based support and a zeolite support 4. Catalyst promoters
(zeolite L) in TRM. MWCNT-supported nickel catalyst, with cerium as
promoter, showed very promising catalytic performance (up to 96.8 % Catalyst performance can be improved by using a promoter. In TRM,
of CH4 conversion at 750 ◦ C, and a good stability for 44 h of reaction) the promoters can be classified into two groups: i) s-block elements
under TRM conditions using a gas mixture of CH4:CO2:H2O:O2:N2 = (alkali and alkaline earth metals) such as Na(I), K(I), Ca(II), and Mg(II);
1:0.34:0.23:0.5:2.1 and space velocity of 1250 mL g− 1 min− 1 [102]. In ii) d- and f-block elements such as Zr(IV), La(III), Ce(IV), Pt(0), and Rh
the case of the zeolite L, this support was initially stabilized by calci­ (0).
nation at 800 ◦ C before Ni or Ni-Rh deposition. The resulting catalysts
showed higher methane conversions and hydrogen yields compared to 4.1. s-block elements
those obtained with alumina-based catalysts. This catalytic behavior
was explained by the high metal dispersions and strong MSI achieved Alkali (Na(I), K(I)) and alkaline earth (Mg(II), Ca(II)) elements have
with the zeolite L [61]. been investigated as promoters of TRM catalysts by different teams [27,
Silicon carbide (β-SiC) has been largely applied in heterogeneous 35,97,106,119]. The addition of Mg(II) in Ni-based catalysts resulted in
catalysis during the last decades [103,104]. In TRM, this support was a high catalytic activity, stability and could attenuate the coke deposi­
particularly studied by García-Vargas et al. [56,85,105–108]. Ni/β-SiC tion [27,120]. As previously reported in section 3, the reduction of
catalysts showed relatively high catalytic activity and stability in TRM, NiO-MgO solid solution enhances nickel dispersion and MSI, increases
explained by a good Ni dispersion and adequate MSI [61,105–108].
However, SiC can be irreversibly oxidized into SiO2 by O2 at high tem­
Table 3
peratures [109]. Tentative rating of TRM catalyst supports.
Hydroxyapatite (Ca10(PO4)6(OH)2), considered as a relatively-new
Rating
support in heterogeneous catalysis, has been found as a good catalyst Criteria Indicator
support in DRM [110–117]. A recent work Phan [118] is the first study 1 2 3
of hydroxyapatite-based catalysts in TRM with promising results. High Thermal stability Threshold temperature 300− 700 700− 900 > 900
methane conversions (up to 90 %) and particularly high catalytic sta­ for phase/structure
bility were achieved with 5 wt/%Ni/HA at 800 ◦ C and 1.6 bar total changing (◦ C)
Specific surface BET (m2∙g− 1) 2− 10 10− 100 > 100
pressure for 300 h of reaction (conditions in a fixed bed reactor: 340 mg
area
of catalyst, at 800 ◦ C, molar ratio of CH4:CO2:O2 = 1:0.67:0.1 and S/C = MSI Reduction temperature 300− 500 500− 700 > 700
0.9, CH4 flow rate =45 mL min− 1). (◦ C)
Basicity Density of basic sites (*) Low High Very
high
3.9. Support comparison and conclusions
Oxygen vacancies Density of oxygen Low High Very
and oxygen vacancies and high
Several studies have compared different supports under the same storage capacity extractable oxygen from
operating conditions in TRM [27,57,84,85]. Song and Pan [27] per­ support (*)
formed TRM over Ni supported on different supports at 850 ◦ C. The Affordable cost Cost (*) Expensive Medium Cheap

catalytic activity carries on the following trend for CH4 conversion: (*): To date, data from the literature in TRM reaction are not enough to better
Ni/MgO > Ni/MgO-CeZrO > Ni/CeO2 ~ Ni/ZrO2 ~ Ni/Al2O3 > rate this criterion.

13
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

Fig. 11. Relative comparison of TRM catalyst supports on the basis of the nominated criteria, taking into consideration the operation conditions of this process.

basicity, and suppresses carbon deposition, due to the roles of Mg(II) vacancies of CeO2 improved Ni dispersion, and induced the formation of
acting as catalyst promoter [27,120], as reported for different supports Ce(III) sites, which in turn enhanced catalytic activity. Strong basic sites,
in TRM: Al2O3 [119], Al2O3-based support [119], ZrO2 [27,97], created by La2O3, promoted CO2 adsorption which favored coke elimi­
CeO2-ZrO2 [27,35], β-SiC [106], NiMo carbides [121]. Fig. 12 shows an nation. Orlyk et al. [82] and Solov’ev et al. [19] reported that CeO2 and
example of the effect of Mg(II) addition to Ni-CeZrO catalyst [27]. La2O3 addition avoided catalytic deactivation by providing high O-va­
For other alkali and alkaline earth metals, no clear common trend cancies for coke gasification and promoting Ni reduction capacity. Zou
was observed. Orlyk et al. [82] indicated that the addition of K2O and et al. [121] stated that the addition of La(III) to NiMo carbide prevented
Na2O to Ni/Al2O3 improved Ni dispersion and thus enhanced catalyst Ni particles from sintering, and suppressed carbon deposition, which
activity. In contrast, García-Vargas et al. [106] pointed out that Na(I), K explains the high catalytic performance of these catalysts. Moreover,
(I) and Ca(II) were not useful for Ni/β-SiC because they enhanced the CeO2 is reported to be active for catalyst regeneration after sulfur
oxidation rate of this support into SiO2 during calcination, thus modi­ deactivation in TRM [60].
fying the catalyst structure. Zou et al. [121] indicated that the addition Lino et al. [40] showed the promoter effect of Ce–ZrO2 addition to
of K(I) transformed active α-Al2O3 to less active θ-Al2O3. Moreover, for Ni/MgAl2O4 catalyst in TRM reaction. The conversions of CH4 and CO2
NiMo carbides, the addition of K(I) suppressed the redox ability of this were kept stable at respectively ca. 75 and 40 % over doped catalyst
support, thus decreasing the catalytic activity [121]. (reaction conditions: 750 ◦ C, 1 bar, 85 mg of catalyst, molar ratio of CH4:
CO2:H2O:O2:N2 = 3:1:1.4:0.5:2, CH4 inlet flow rate = 51.5 NmL min− 1),
4.2. d- and f-block elements while non-doped counterpart linearly deactivated during the reaction.
This behavior is explained by the adequate basic properties of Ce–ZrO2,
The addition of La2O3 or CeO2 to TRM catalysts enhances their cat­ which limits coke formation, and by the favorable effect of Ce–ZrO2 on
alytic activity by increasing MSI, increasing basicity for CO2 adsorption, NiO reduction. Especially, when the addition of Ce–ZrO2 is synergized
and limiting catalytic deactivation by re-oxidation of Ni(0) [19,82,53, with Mg in the Ni-Mg/Ce–ZrO2/Al2O3 catalysts, a superior catalytic
121]. Pino et al. [53] studied the influence of La(III) loading on the activity and a high coke resistance were observed by creating new
performance of Ni/CeO2 catalyst. By adding 10 wt.% La(III), CO2 and weakly acidic sites, basic sites and redox ability [79,122].
CH4 conversions increased from 93 to 96 % and 83 to 86.5%, respec­ The addition of a small amount (< 0.5 wt.%) of noble metals such as
tively, and no carbon deposition was observed (reaction conditions: 800 Pt(0) or Rh(0) to Ni-based catalysts limits the catalyst deactivation by Ni
◦
C, 1 bar, WHSV = 30,000 h− 1, molar ratio of CH4:CO2:H2O:O2 = re-oxidation during TRM reaction [123]. In addition, these metals can
1:0.46:0.46:0.1). A further increase in La(III) loading decreased CO2 and favor the reduction and the dispersion of the active phase via spill-over
CH4 conversions, while the molar H2/CO ratio remained stable at about effect, and promote surface carbon gasification by providing high sur­
1.62–1.65. The strong interaction of Ni with La2O3 and surface oxygen face oxygen species [83,124]. Jiang et al. [123] investigated Pt-modified
Ni/MgO catalysts. The Ni-Pt alloy formed in these catalysts was active
for both methane reforming and methane partial oxidation reactions,
leading to different reaction zones inside the catalyst bed (such as
auto-thermal zone followed by an oxygen absence zone). The formation
of Ni-Pt alloy prevented the re-oxidation of Ni(0) during TRM reaction.
Izquierdo et al. [83] studied various Ni- and Rh-Ni-based catalysts in
TRM. The addition of 1 wt.% Rh to Ni/Ce-Al2O3 catalyst lowered the
reduction temperature, which is explained by spillover effect of Rh, and
limited the formation of NiAl2O4, which is reluctant to undergo reduc­
tion. In addition, the catalyst showed a high specific surface area of
156.8 m2 g− 1 and a strong MSI (evidenced by the TPR peaks at 751 ◦ C
and 813 ◦ C). Consequently, Rh-Ni/Ce-Al2O3 catalyst exhibited high
catalytic performance (99.5 % CH4 conversion and 62.8 % H2 yield) in
TRM at 800 ◦ C and 1 atm (reaction conditions: 340 mg of catalyst; molar
ratio of CO2/CH4 = 0.67, O2/CH4 = 0.25, S/C = 1.0, WHSV =161 h− 1,
reaction time of 90 min). Similar results on the promoter effect of Rh
Fig. 12. Example of the effect of Mg(II) addition to the Ni-CeZrO catalyst in
were also reported for Ni/zeolite L [61]. However, the main drawback of
TRM at different reaction temperatures. Reaction conditions: 100 mg catalyst, 1
atm, molar ratio of CH4:CO2:H2O:O2 = 1:0.48:0.54:0.1 (CH4 flow rate =25 mL/ noble metals is their high cost.
min). Reprinted from [27], Copyright (2004), with permission from Elsevier. On the basis of the rating criteria in Table 3 (except the specific

14
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

surface area which could not be assessed for all promoters), a relative
comparison between the two groups of promoters is presented in Fig. 13.
Globally, d- and f-block elements, when they are used at low content (e.
g. < 0.5 wt%), have been found to be more advantageous than s-block
elements. Thus, depending on the support used during the catalyst
preparation, the presence of a promoter can be envisaged in order to
correct drawbacks exhibited by the support (e.g. acidity, metal reduc­
ibility etc.).

5. Conclusions and outlook

Catalytic TRM is an alternative to SRM, POM and DRM to produce

syngas from CH4 using simultaneously H2O, CO2, and O2 as oxidants.
This review is specifically focused on the development of efficient TRM
catalysts, topic intensively investigated during the last two decades. Fig. 13. Relative comparison of TRM catalyst promoters on the basis of the
Through this review, the following conclusions and analyses can be nominated criteria, taking into consideration the operation conditions of
inferred: this process.

- Nickel is largely studied as the active metal in TRM. Small Ni preparation. U.P.M. Ashik: Data curation, Writing- Original draft
nanoparticles, together with a compromise between strong MSI and preparation. Jun-Ichiro Hayashi: Validation, Funding acquisition.
metal reducibility are recommended. SAC investigation in TRM is Alejandro Pérez Alonso: Data curation, Visualization. Daniel Pla:
also highly suggested for the future work. Supervision, Reviewing, Validation, Funding acquisition. Montserrat
- Oxides and mixed oxides are also largely studied as catalyst supports Gómez: Supervision, Reviewing, Validation, Funding acquisition. Doan
for Ni dispersion in TRM. Particular attention should be paid on the Pham Minh: Conceptualization, Writing- Original draft preparation,
dispersion of the active phase (Ni) and its adequate interaction with Editing, Supervision, Reviewing, Validation, Funding acquisition.
the support to tune its reducibility, but also its stability during TRM
reaction. Mixed oxides doped with MgO, CeO2, ZrO2, and La2O3 are Declaration of Competing Interest
strongly suggested for future works.
- The use of an appropriate promoter is also highly recommended for The authors report no declarations of interest.
improving Ni reducibility, suppressing coke deposition, and
enhancing catalyst stability.
Acknowledgments

Taking into account the results reported in the literature, the

Part of this work was financially supported by ANR (France) via
following recommendations can be considered for the future works on
CARNOT M.I.N.E.S in the context of the project HyTREND, Occitanie
catalyst development for TRM:
region, and IMT Mines Albi. A. P. A., D. P., and M. G. thank the Uni­
versité Toulouse 3 – Paul Sabatier and the Centre National de la
- The utilization of Ni(0) as the main active phase: Ni appears as the
Recherche Scientifique (CNRS) for their financial support. A. P. A.
best choice to catalyze TRM reaction because of its high efficiency,
thanks the Université Fédérale de Toulouse and IMT Mines Albi for his
and its large availability with relative low cost in comparison with
doctoral fellowship. The authors also thank Prof. D.V.N. Vo for technical
classical noble metal catalysts. Small Ni clusters and isolated metal
help.
atoms as well as their cooperative effect should be explored in TRM.
For that, the development of specific synthesis methods is needed,
Appendix A. Supplementary data
and the characterization of these species must be carefully
conducted.
Supplementary material related to this article can be found, in the
- The utilization of a noble metal such as Pt at low content (< 0.5 wt.
online version, at doi:https://doi.org/10.1016/j.apcata.2021.118286.
%) as promoter: this favors Ni reduction and solid carbon suppres­
sion, and thus enhances catalyst stability.
References
- The utilization of a spinel such as MgAl2O4 doped with CeO2 or ZrO2
as catalyst support: alumina-based supports are well known by their [1] K. Liu, C. Song, V. Subramani, Hydrogen and Syngas Production and Purification
high specific surface area and thermal stability in heterogeneous Technologies, John Wiley & Sons, Hoboken, New Jersey, 2009.
catalysis. The addition of magnesium oxide to form spinel structure [2] https://www.marketsandmarkets.com/ (accessed on May 05, 2021).
[3] Z. Arab Aboosadi, M. Farhadi Yadecoury, Int. J. Chem. React. Eng. 17 (2019),
enhances the basicity of the alumina support, which is favorable for https://doi.org/10.1515/ijcre-2019-0108, 20190108.
TRM reaction implying CO2 conversion, while the addition of CeO2 [4] S.O. Soloviev, I.V. Gubareni, S.M. Orlyk, Theoretic. Experiment. Chem. 54 (2018)
or ZrO2 improves the MSI and provides surface oxygen vacancies 293–315, https://doi.org/10.1007/s11237-018-9575-5.
[5] A.H. Elbadawi, L. Ge, Z. Li, S. Liu, S. Wang, Z. Zhu, Catal. Rev. (2020), https://
which reduces coke formation. doi.org/10.1080/01614940.2020.1743420.
- Scale-up the process: the proof-of-concept of TRM has been demon­ [6] A.I. Osman, Chem. Eng. Technol. 43 (2020) 641–648, https://doi.org/10.1002/
strated at laboratory scale but fundamental research towards reac­ ceat.201900339.
[7] S. Arora, R. Prasad, RSC Adv. 6 (2016) 108668–108688, https://doi.org/
tion optimization and validation of the best TRM catalyst hits for
10.1039/C6RA20450C.
each series is currently ongoing. Robust TRM catalyst systems able to [8] N.A.K. Aramouni, J.G. Touma, B.A. Tarboush, J. Zeaiter, M.N. Ahmad, Ren. Sust.
operate under optimal conditions are required to enable further Ener. Rev. 82 (2018) 2570–2585, https://doi.org/10.1016/j.rser.2017.09.076.
studies at large scale to realize and accelerate the deployment of this [9] W.J. Jang, J.O. Shim, H.M. Kim, S.Y. Yoo, H.S. Roh, Catal. Today 324 (2019)
15–26, https://doi.org/10.1016/j.cattod.2018.07.032.
technological advancement of societal relevance. [10] B.V.R. Kuncharam, A.G. Dixon, Fuel Process. Technol. 200 (2020), https://doi.
org/10.1016/j.fuproc.2019.106314, 106314.
CRediT authorship contribution statement [11] R. Ma, B. Xu, X. Zhang, Catal. Today 338 (2019) 18–30, https://doi.org/10.1016/
j.cattod.2019.06.025.
[12] P. Arku, B. Regmi, A. Dutta, Chem. Eng. Res. Des. 136 (2018) 385–402, https://
Huynh Pham Xuan: Data curation, Writing- Original draft doi.org/10.1016/j.cherd.2018.05.044.

15
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

[13] G. Kolios, B. Glöckler, A. Gritsch, A. Morillo, G. Eigenberger, Fuel Cells Weinh. [50] M. Usman, W.M.A. Wan Daud, RSC Adv. 6 (2016) 38277–38289, https://doi.org/
(Weinh) 5 (2005) 52–65, https://doi.org/10.1002/fuce.200400065. 10.1039/C6RA01652A.
[14] U. Hamid, A. Rauf, U. Ahmed, M.S.A.S. Shah, N. Ahmad, Fuel 266 (2020), [51] A. Vita, L. Pino, F. Cipitì, M. Laganà, V. Recupero, Fuel Proc. Technol. 127 (2014)
https://doi.org/10.1016/j.fuel.2020.117111, 117111. 47–58, https://doi.org/10.1016/j.fuproc.2014.06.014.
[15] A.Y. Krylova, Solid Fuel Chem. 48 (2014) 22–35, https://doi.org/10.3103/ [52] M. Lo Faro, A. Vita, L. Pino, A.S. Aricò, Fuel Proc. Technol. 115 (2013) 238–245,
S0361521914010030. https://doi.org/10.1016/j.fuproc.2013.06.008.
[16] D.B. Pal, R. Chand, S.N. Upadhyay, P.K. Mishra, Renew. Sustain. Ener. Rev. 93 [53] L. Pino, A. Vita, F. Cipitì, M. Laganà, V. Recupero, Appl. Catal. B: Env. 104 (2011)
(2018) 549–565, https://doi.org/10.1016/j.rser.2018.05.003. 64–73, https://doi.org/10.1016/j.apcatb.2011.02.027.
[17] A.P. York, T. Xiao, M.L. Green, Topics Catal. 22 (3–4) (2003) 345–358, https:// [54] L. Pino, A. Vita, M. Laganà, V. Recupero, Appl. Catal. B Env. 148–149 (2014)
doi.org/10.1023/A:1023552709642. 91–105, https://doi.org/10.1016/j.apcatb.2013.10.043.
[18] A.M. Ranjekar, G.D. Yadav, J. Ind. Chem. Soc. 98 (2021), https://doi.org/ [55] K.T. de C. Roseno, R.A. Antunes, R.M.B. Alves, M. Schmal, Catal. Lett. (2021),
10.1016/j.jics.2021.100002, 100002. https://doi.org/10.1007/s10562-021-03600-0.
[19] S.A. Solov’ev, Y.V. Gubareni, Y.P. Kurilets, S.N. Orlik, Theor. Exp. Chem. 48 [56] J.M. García-Vargas, J.L. Valverde, A. de Lucas-Consuegra, B. Gómez-Monedero,
(2012) 199–205, https://doi.org/10.1007/s11237-012-9262-x. P. Sánchez, F. Dorado, Appl. Catal. A Gen. 431–432 (2012) 49–56, https://doi.
[20] S. Jain, D. Newman, A. Nzihou, H. Dekker, P. Le Feuvre, H. Richter, F. Gobe, org/10.1016/j.apcata.2012.04.016.
C. Morton, R. Thompson, Global Potential of Biogas, World biogas association, [57] R. Kumar, K. Kumar, K.K. Pant, N.V. Choudary, Int. J. Hydr. Ener. 45 (2019)
2019. 1911–1929, https://doi.org/10.1016/j.ijhydene.2019.11.111.
[21] C.S. Song, Chem. Innov. 31 (2001) 21–26. [58] D. Pham Minh, T.J. Siang, D.V.N. Vo, T.S. Phan, C. Ridart, A. Nzihou, D. Grouset,
[22] K. Świrk, T. Grzybek, M. Motak, Tri-reforming as a process of CO2 utilization and Chapter 4 - Hydrogen production from biogas reforming: An overview of steam
a novel concept of energy storage in chemical products, E3S Web of Conferences reforming, dry reforming, dual reforming, and tri-reforming of methane.
14 (2017), https://doi.org/10.1051/e3sconf/20171402038, 02038. “Hydrogen Supply Chains - Design, Deployment and Operation”, Academic Press,
[23] K. Świrk, J. Grams, M. Motak, P. Da Costa, T. Grzybek, J. CO2 Utiliz. 42 (2020), 2018, pp. 111–166, https://doi.org/10.1016/B978-0-12-811197-0.00004-X.
https://doi.org/10.1016/j.jcou.2020.101317, 101317. [59] C.G. Anchieta, E.M. Assaf, J.M. Assaf, Int. J. Hydr. Ener. 44 (2019) 9316–9327,
[24] Z. Li, Z. Wang, S. Kawi, ChemCatChem 11 (2019) 202–224, https://doi.org/ https://doi.org/10.1016/j.ijhydene.2019.02.122.
10.1002/cctc.201801266. [60] U. Izquierdo, I. García-García, A.M. Gutierrez, J.R. Arraibi, V.L. Barrio, J.
[25] Q. Wei, X. Gao, L. Wang, Q. Ma, Fuel 27 (2020), https://doi.org/10.1016/j. F. Cambra, P.L. Arias, Catalysts 8 (2018) 12, https://doi.org/10.3390/
fuel.2020.117631, 117631. catal8010012.
[26] N. Gao, M. Cheng, C. Quan, Y. Zheng, Fuel 27 (2020), https://doi.org/10.1016/j. [61] U. Izquierdo, V.L. Barrio, K. Bizkarra, A.M. Gutierrez, J.R. Arraibi, L. Gartzia,
fuel.2020.117702, 117702. J. Bañuelos, I. Lopez-Arbeloa, J.F. Cambra, Chem. Eng. J. 238 (2014) 178–188,
[27] C. Song, W. Pan, Catal. Today 98 (2004) 463–484, https://doi.org/10.1016/j. https://doi.org/10.1016/j.cej.2013.08.093.
cattod.2004.09.054. [62] R.K. Singha, A. Yadav, A. Agrawal, A. Shukla, S. Adak, T. Sasaki, R. Bal, Appl.
[28] L.J.L. Maciel, E.A.M. de Souza, V.O. Cavalcanti-Filho, A. Knoechelmann, C.A. Catal. B: Env. 191 (2016) 165–178, https://doi.org/10.1016/j.
M. de Abreu, React. Kinet. Mech. Catal. 101 (2010) 407–416, https://doi.org/ apcatb.2016.03.029.
10.1007/s11144-010-0232-9. [63] S. Zhang, S. Muratsugu, N. Ishiguro, M. Tada, ACS Catal. 3 (2013) 1855–1864,
[29] R. Kumar, K.K. Pant, Appl. Surf. Sci. 515 (2020), https://doi.org/10.1016/j. https://doi.org/10.1021/cs400159w.
apsusc.2020.146010, 146010. [64] V.A. Kondratenko, C. Berger-Karin, E.V. Kondratenko, ACS Catal. 4 (2014)
[30] C.E. Kozonoe, R. de Paiva Floro Bonfim, R.M.B. Alves, M. Schmal, Fuel 256 3136–3144, https://doi.org/10.1021/cs5002465.
(2019), https://doi.org/10.1016/j.fuel.2019.115917, 115917. [65] L. Pino, A. Vita, C. Italiano, C. Fabiano, M. Lagana, V. Recupero, Adv. Sci.
[31] D.J. Moon, Catal. Surv. Asia 12 (2008) 188–202, https://doi.org/10.1007/ Technol. 93 (2014) 19–24, https://doi.org/10.4028/www.scientific.net/
s10563-008-9051-7. AST.93.19.
[32] X. Zhao, B. Joseph, J. Kuhn, S. Ozcan, iScience 23 (2020), https://doi.org/ [66] X. Yang, A. Wang, B. Qiao, J. Li, T. Zhang, Acc. Chem. Res. 46 (2013) 1740–1748,
10.1016/j.isci.2020.101082, 101082. https://doi.org/10.1021/ar300361m.
[33] D. Pham Minh, H. Pham Xuan, T.J. Siang, D.V.N. Vo, Appl. Catal. A Gen. 621 [67] F. Chen, X. Jiang, L. Zhang, R. Lang, B. Qiao, Chinese J. Catal. 39 (2018)
(2021), https://doi.org/10.1016/j.apcata.2021.118202, 118202. 893–898, https://doi.org/10.1016/S1872-2067(18)63047-5.
[34] R.K. Singha, A. Shukla, A. Yadav, S. Adak, Z. Iqbal, N. Siddiqui, R. Bal, Appl. [68] N. Cheng, L. Zhang, K. Doyle-Davis, X. Sun, Electrochem. Ener. Rev. 2 (2019)
Energy 178 (2016) 110–125, https://doi.org/10.1016/j.apenergy.2016.06.043. 539–573, https://doi.org/10.1007/s41918-019-00050-6.
[35] D.M. Walker, S.L. Pettit, J.T. Wolan, J.N. Kuhn, Appl. Catal. A Gen. 445–446 [69] Y.H. Hu, Catal. Today 148 (2009) 206–211, https://doi.org/10.1016/j.
(2012) 61–68, https://doi.org/10.1016/j.apcata.2012.08.015. cattod.2009.07.076.
[36] U.P.M. Ashik, S. Asano, S. Kudo, D. Pham Minh, S. Appari, E. Hisahiro, J. [70] S. Tang, L. Ji, J. Lin, H.C. Zeng, K.L. Tan, K. Li, J. Catal. 194 (2000) 424–430,
I. Hayashi, Catalysts 10 (2020) 21, https://doi.org/10.3390/catal10010021. https://doi.org/10.1006/jcat.2000.2957.
[37] J.R.A. Sietsma, A. Jos van Dillen, P.E. de Jongh, K.P. de Jong, Application of [71] M. Akri, S. Zhao, X. Li, K. Zang, A.F. Lee, M.A. Isaacs, W. Xi, Y. Gangarajula,
ordered mesoporous materials as model supports to study catalyst preparation by J. Luo, Y. Ren, Y.T. Cui, L. Li, Y. Su, X. Pan, W. Wen, Y. Pan, K. Wilson, L. Li,
impregnation and drying, in: E.M. Gaigneaux, M. Devillers, D.E. de Vos, B. Qiao, H. Ishii, Y.F. Liao, A. Wang, X. Wang, T. Zhang, Nat. Commun. 10 (2019)
S. Hermans, P.A. Jacobs, J.A. Martens, P. Ruiz (Eds.), Scientific Bases for the 1–10, https://doi.org/10.1038/s41467-019-12843-w.
Preparation of Heterogeneous Catalysts, 162, Elsevier, Amsterdam, 2006, [72] Z. Zuo, S. Liu, Z. Wang, C. Liu, W. Huang, J. Huang, P. Liu, ACS Catal. 8 (2018)
pp. 95–102. 9821–9835, https://doi.org/10.1021/acscatal.8b02277.
[38] R. Raudaskoski, E. Turpeinen, R. Lenkkeri, E. Pongrácz, R.L. Keiski, Catal. Today [73] M. Akri, A. El Kasmi, C. Batiot-Dupeyrat, B. Qiao, Catalysts 10 (2020) 630,
144 (2009) 318–323, https://doi.org/10.1016/j.cattod.2008.11.026. https://doi.org/10.3390/catal10060630.
[39] T. Horiuchi, K. Sakuma, T. Fukui, Y. Kubo, T. Osaki, T. Mori, Appl. Catal. A Gen. [74] Y. Tang, Y. Wei, Z. Wang, S. Zhang, Y. Li, L. Nguyen, Y. Li, Y. Zhou, W. Shen, F.
144 (1996) 111–120, https://doi.org/10.1016/0926-860X(96)00100-7. F. Tao, P. Hu, J. Am. Chem. Soc. 141 (2019) 7283–7293, https://doi.org/
[40] A.V.P. Lino, E.M. Assaf, J.M. Assaf, J. CO2 Util. 33 (2019) 273–283, https://doi. 10.1021/jacs.8b10910.
org/10.1016/j.jcou.2019.06.016. [75] P. Serp, Nanoscale 13 (2021) 5985–6004, https://doi.org/10.1039/
[41] L.J. Si, C.Z. Wang, N.N. Sun, X. Wen, N. Zhao, F.K. Xiao, W. Wei, Y.H. Sun, J. Fuel D1NR00465D.
Chem. Technol. 40 (2012) 210–215, https://doi.org/10.1016/s1872-5813(12) [76] A.V.P. Lino, Y.N. Colmenares Calderon, V.R. Mastelaro, E.M. Assaf, J.M. Assaf,
60011-5. Appl. Surf. Sci. 481 (2019) 747–760, https://doi.org/10.1016/j.
[42] A.J. Majewski, J. Wood, Int. J. Hydr. Ener. 39 (2014) 12578–12585, https://doi. apsusc.2019.03.140.
org/10.1016/j.ijhydene.2014.06.071. [77] I. Sebai, A. Boulahaouache, M. Trari, N. Salhi, Int. J. Hydr. Ener. 44 (2019)
[43] G. Huang, C.H. Lu, H.H. Yang, in: X. Wang, X. Chen (Eds.), Novel Nanomaterials 9949–9958, https://doi.org/10.1016/j.ijhydene.2018.12.050.
for Biomedical, Environmental and Energy Applications, Elsevier, Amsterdam, [78] D.J. Moon, J.S. Kang, W. S. Nho, D.H. Kim, S. D. Lee, B. G. Lee, Ni-based catalyst
2019, pp. 89–109. for Tri-reforming of methane and its catalysis application for the production of
[44] H. Zhong, T. Mirkovic, G.D. Scholes, in: D.L. Andrews, G.D. Scholes, G. syngas, US patent: US20080260628A1 (2008).
P. Wiederrecht (Eds.), Comprehensive Nanoscience and Technology, Academic [79] X. Zhao, D.M. Walker, D. Maiti, A.D. Petrov, M. Kastelic, B. Joseph, J.N. Kuhn,
Press, Amsterdam, 2011, pp. 153–201. Ind. Eng. Chem. Res. 57 (2018) 845–855, https://doi.org/10.1021/acs.
[45] R.K. Singha, S. Das, M. Pandey, S. Kumar, R. Bal, A. Bordoloi, Catal. Sci. Technol. iecr.7b03669.
6 (2016) 7122–7136, https://doi.org/10.1039/C5CY01323B. [80] J. Yoo, Y. Bang, S.J. Han, S. Park, J.H. Song, I.H. Song, J. Mol. Catal. A Chem. 410
[46] S. Das, S. Thakur, A. Bag, M.S. Gupta, P. Mondal, A. Bordoloi, J. Catal. 330 (2015) (2015) 74–80, https://doi.org/10.1016/j.molcata.2015.09.008.
46–60, https://doi.org/10.1016/j.jcat.2015.06.010. [81] H.T. Jiang, H.Q. Li, Y. Zhang, J. Fuel Chem. Technol. 35 (2007) 72–78, https://
[47] E. McEvoy, S. Donegan, J. Power, K. Altria, A. Marsh, in: I.D. Wilson (Ed.), doi.org/10.1016/S1872-5813(07)60012-7.
Encyclopedia of Separation Science, Academic Press, Oxford, 2007, pp. 1–10. [82] S.N. Orlyk, M.R. Kantserova, T.K. Shashkova, E.V. Gubareni, V.I. Chedryk, S.
[48] S. Kim, B.S. Crandall, M.J. Lance, N. Cordonnier, J. Lauterbach, E. Sasmaz, Appl. A. Soloviev, Theor. Exp. Chem. 49 (2013) 22–34, https://doi.org/10.1007/
Catal. B: Env. 259 (2019), https://doi.org/10.1016/j.apcatb.2019.118037, s11237-013-9290-1.
118037. [83] U. Izquierdo, V.L. Barrio, J. Requies, J.F. Cambra, M.B. Güemez, P.L. Arias, Int. J.
[49] B. Safavinia, Y. Wang, C. Jiang, C. Roman, P. Darapaneni, J. Larriviere, D. Hydr. Ener. 38 (2013) 7623–7631, https://doi.org/10.1016/j.
A. Cullen, K.M. Dooley, J.A. Dorman, ACS Catal. 10 (2020) 4070–4079, https:// ijhydene.2012.09.107.
doi.org/10.1021/acscatal.0c00203. [84] R. Kumar, K. Kumar, N.V. Choudary, K.K. Pant, Fuel Proc. Technol. 186 (2019)
40–52, https://doi.org/10.1016/j.fuproc.2018.12.018.

16
X.-H. Pham et al. Applied Catalysis A, General 623 (2021) 118286

[85] J.M. García-Vargas, J.L. Valverde, F. Dorado, P. Sánchez, J. Mol. Catal. A Chem. [107] J.M. García-Vargas, J.L. Valverde, A. de Lucas-Consuegra, B. Gómez-Monedero,
395 (2014) 108–116, https://doi.org/10.1016/j.molcata.2014.08.019. F. Dorado, P. Sánchez, Int. J. Hydr. Ener. 38 (2013) 4524–4532, https://doi.org/
[86] J. Du, J. Gao, F. Gu, J. Zhuang, B. Lu, L. Jia, G. Xu, Q. Liu, F. Su, Int. J. Hydr. Ener. 10.1016/j.ijhydene.2013.02.001.
43 (2018) 20661–20670, https://doi.org/10.1016/j.ijhydene.2018.09.128. [108] J.M. García-Vargas, J.L. Valverde, J. Díez, P. Sánchez, F. Dorado, Appl. Catal. B:
[87] J.R.H. Ross, M.C.F. Steel, A. Zeini-Isfahani, J. Catal. 52 (1978) 280–290, https:// Env. 164 (2015) 316–323, https://doi.org/10.1016/j.apcatb.2014.09.044.
doi.org/10.1016/0021-9517(78)90142-2. [109] A.R. Sane, P.M. Nigay, D. Pham Minh, C. Toussaint, A. Germeau, N. Semlal,
[88] D. Sun, X. Li, S. Ji, L. Cao, J. Nat. Gas Chem. 19 (2010) 369–374, https://doi.org/ R. Boulif, A. Nzihou, J. Therm. Anal. Calorim. 140 (2020) 2087–2096, https://
10.1016/S1003-9953(09)60096-7. doi.org/10.1007/s10973-019-08964-5.
[89] W. Cai, L. Ye, L. Zhang, Y. Ren, B. Yue, X. Chen, H. He, Materials 7 (2014) [110] Z. Boukha, M. Kacimi, M.F.R. Pereira, J.L. Faria, J.L. Figueiredo, M. Ziyad, Appl.
2340–2355, https://doi.org/10.3390/ma7032340. Catal. A Gen. 317 (2007) 299–309, https://doi.org/10.1016/j.
[90] X.Y. Quek, D. Liu, W.N.E. Cheo, H. Wang, Y. Chen, Y. Yang, Appl. Catal. B: Env. apcata.2006.10.029.
95 (2010) 374–382, https://doi.org/10.1016/j.apcatb.2010.01.016. [111] T.S. Phan, A.R. Sane, B. Rêgo de Vasconcelos, A. Nzihou, P. Sharrock, D. Grouset,
[91] Z. Li, Z. Wang, B. Jianga, S. Kawi, Catal. Sci. Technol. 8 (2018) 3363–3371, D. Pham Minh, Appl. Catal. B: Env. 224 (2018) 310–321, https://doi.org/
https://doi.org/10.1039/C8CY00767E. 10.1016/j.apcatb.2017.10.063.
[92] V.G. de la Cruz-Flores, A. Martinez-Hernandez, M.A. Gracia-Pinilla, Appl. Catal. A [112] B. Rêgo de Vasconcelos, D. Pham Minh, N. Lyczko, T.S. Phan, P. Sharrock,
Gen. 594 (2020), https://doi.org/10.1016/j.apcata.2020.117455, 117455. A. Nzihou, Fuel 226 (2018) 195–203, https://doi.org/10.1016/j.
[93] Z.A. Fedorova, M.M. Danilova, V.I. Zaikovskii, T.A. Krieger, Mater. Lett. 263 fuel.2018.04.017.
(2020), https://doi.org/10.1016/j.matlet.2019.127156, 127156. [113] B. Rêgo de Vasconcelos, D. Pham Minh, P. Sharrock, A. Nzihou, Catal. Today 310
[94] K. Tomishige, Catal. Today 89 (2004) 405–418, https://doi.org/10.1016/j. (2018) 107–115, https://doi.org/10.1016/j.cattod.2017.05.092.
cattod.2004.01.003. [114] B. Rêgo de Vasconcelos, D. Pham Minh, E. Martins, A. Germeau, P. Sharrock,
[95] N. Ohtake, M. Katoh, S. Sugiyama, J. Ceram. Soc. Japan 125 (2017) 57–61, A. Nzihou, Int. J. Hydr. Ener. 45 (2020) 18502–18518, https://doi.org/10.1016/
https://doi.org/10.2109/jcersj2.16255. j.ijhydene.2019.08.068.
[96] P. Pal, R.K. Singha, A. Saha, R. Bal, A.B. Panda, J. Phys. Chem. C 119 (2015) [115] B. Rêgo de Vasconcelos, D. Pham Minh, E. Martins, A. Germeau, P. Sharrock,
13610–13680, https://doi.org/10.1021/acs.jpcc.5b01724. A. Nzihou, Chem. Eng. Technol. 43 (4) (2020) 698–704, https://doi.org/
[97] L.Z. Sun, Y.S. Tan, Q.D. Zhang, H.J. Xie, Y.Z. Han, J. Fuel Chem. Technol. 40 10.1002/ceat.201900461.
(2012) 831–837, https://doi.org/10.1016/s1872-5813(12)60032-2. [116] T.Q. Tran, D. Pham Minh, T.S. Phan, Q.N. Pham, H. Nguyen Xuan, Chem. Eng.
[98] H. Jiang, H. Li, H. Xu, Y. Zhang, Fuel Proc. Technol. 88 (2007) 988–995, https:// Sci. (2020), https://doi.org/10.1016/j.ces.2020.115975.
doi.org/10.1016/j.fuproc.2007.05.007. [117] Z. Boukha, M.P. Yeste, M.A. Cauqui, J.R. González-Velasco, Appl. Catal. A Gen.
[99] J.L. Ye, Y.Q. Wang, Y. Liu, H. Wang, Int. J. Hydr. Ener. 33 (2008) 6602–6611, 580 (2019) 34–45, https://doi.org/10.1016/j.apcata.2019.04.034.
https://doi.org/10.1016/j.ijhydene.2008.08.036. [118] T.S. Phan, Élaboration, caractérisation et mise en oeuvre d’un catalyseur dans le
[100] S.H. Lee, W. Cho, W.S. Ju, B.H. Cho, Y.C. Lee, Y.S. Baek, Catal. Today 87 (2003) reformage du biogaz en vue de la production de l’hydrogène vert, PhD thesis, IMT
133–137, https://doi.org/10.1016/j.cattod.2003.10.005. Mines, Albi, 2020. http://theses.fr/s164217.
[101] J.S. Kang, D.H. Kim, S.D. Lee, S.I. Hong, D.J. Moon, Appl. Catal. A Gen. 332 [119] X. Zhao, H.T. Ngo, D.M. Walker, D. Weber, D. Maiti, U. Cimenler, A.D. Petrov,
(2007) 153–158, https://doi.org/10.1016/j.apcata.2007.08.017. B. Joseph, J.N. Kuhn, Chem. Eng. Commun. 205 (2018) 1129–1142, https://doi.
[102] C.E. Kozonoe, R.M.B. Alvesa, M. Schmal, Fuel 281 (2020), https://doi.org/ org/10.1080/00986445.2018.1434162.
10.1016/j.fuel.2020.118749, 118749. [120] G. Nahar, D. Mote, V. Dupon, Renew. Sust. Ener. Rev. 76 (2017) 1032–1052,
[103] M.J. Ledoux, C. Pham-Huu, CATTECH 5 (2011) 226–246, https://doi.org/ https://doi.org/10.1016/j.rser.2017.02.031.
10.1023/A:1014092930183. [121] H. Zou, S. Chen, J. Huang, Z. Zhao, Int. J. Hydr. Ener. 41 (2016) 16842–16850,
[104] A. Majid, N. Rani, M.F. Malik, N. Ahmad, N.A. Hassan, F. Hussain, A. Shakoor, https://doi.org/10.1016/j.ijhydene.2016.07.108.
Ceram. Int. 45 (2019) 8069–8080, https://doi.org/10.1016/j. [122] W. Cho, T. Song, A. Mitsos, J.T. McKinnon, G.H. Ko, J.E. Tolsma, D. Denholm,
ceramint.2019.01.167. T. Park, Catal. Today 139 (2009) 261–267, https://doi.org/10.1016/j.
[105] J.M. García-Vargas, J.L. Valverde, J. Díez, F. Dorado, P. Sánchez, Int. J. Hydr. cattod.2008.04.051.
Ener. 40 (2015) 8677–8687, https://doi.org/10.1016/j.ijhydene.2015.05.032. [123] H. Jiang, H. Li, H. Fan, Appl. Mech. Mater. 252 (2013) 255–258, https://doi.org/
[106] J.M. García-Vargas, J.L. Valverde, J. Díez, P. Sánchez, F. Dorado, Appl. Catal. B: 10.4028/www.scientific.net/AMM.252.255.
Env. 148–149 (2014) 322–329, https://doi.org/10.1016/j.apcatb.2013.11.013. [124] D. Pham Minh, T.J. Siang, N.D.V. Vo, T.S. Phan, A. Nzihou, D. Grouset, in:
C. Azzaro-Pantel (Ed.), Hydrogen Supply Chains Design, Deployment and
Operation, Academic Press, London, 2019, pp. 111–166.

17