Natural gas is one of the main energy resources that occupies a pivotal role in the global energy paradigm. Natural gas reforming is among the most investigated reactions in heterogeneous catalysis due its direct connection with key industrial processes such as methanol or ammonia synthesis. This chapter will focus on the design of robust reforming catalysts able to overcome the traditional drawbacks of commercial catalysts such as sintering or coking.

The global consumption of natural gas has very much increased over recent decades.¹  This is mainly due to the industrialization of developing economies in regions with large gas resources (Africa and the Middle East) and to the progressive coal-to-gas switching,²  especially in the USA and China, which has been initiated to decarbonize energy and decrease energy-related CO₂ emissions.

Gas consumption is still supposed to grow by 50% until 2040, reaching a total volume of 5370 billion cubic metres (Bcm) (see Table 1.1). Although the direct use of gas as a transportation fuel is growing rapidly, this sector remains small, at less than 2% of total consumption in 2017, compared to industry (32%), buildings (21%) and power generation (39%). These four uses imply the combustion of natural gas to generate heat or power. In 2017, only 206 Bcm of natural gas (around 6% of total consumption) was not combusted but chemically converted.

The flow diagram in Figure 1.1 depicts the use or natural gas in 2017. Among the 206 Bcm that has been converted, around 90% has been converted by reforming reactions to hydrogen for ammonia synthesis or petrochemistry. Around 10%, after a step of reforming to produce synthesis gas (syngas), was further converted into methanol through the Fischer–Tropsch (FT) process to produce chemicals or fuels. Note that despite the strong industrial interest in direct routes of methane conversion to fuels and chemicals,³  this direct use, avoiding the energy-consuming reforming step, is almost equal to 0%. Only one oxidative coupling of methane (OCM) industrial unit producing ethylene has been identified to date.⁴ 

The steam reforming of methane (SRM) (eqn (1.1)) is a mature technology that is used industrially to produce hydrogen from natural gas.^5,6  The syngas thus obtained is compatible with methanol synthesis and the FT process.⁷  Syngas can also be generated by dry reforming of methane (DRM), using CO₂ as an oxidizing agent according to eqn (1.2). This process would be of particular interest in the context of the decarbonization of fuels, starting from biogas. The recent considerations around reducing greenhouse gas emissions clearly drove the research for efficient and stable DRM processes.⁸ 

The steam reforming and the dry reforming reactions are linked by the reverse water gas shift reaction (RWGS) (eqn (1.3)). Both reforming reactions are strongly endothermic and can be combined with the exothermic partial oxidation of methane (POX) (eqn (1.4)) to decrease energy demand⁹  and to allow for better thermal control of the catalytic reactor.¹⁰ 

The thermodynamics of the reforming reactions imply that they have to be operated at high temperature.¹¹  Depending on temperature, pressure and inlet composition, the formation of solid carbon, through methane decomposition (eqn (1.5)) or the Boudouard reaction (eqn (1.6)), may be favoured, which would be detrimental to the stability of the catalytic materials.¹² 

For methane steam reforming under a stoichiometric mixture, the maximal formation of carbon is expected at around 600 °C (Figure 1.2). The formation of carbon decreases at higher temperatures, in the range where high conversions are reached. At temperatures higher than 850 °C, almost total conversion can be achieved with close to 100% selectivity.¹¹  The use of steam to methane (S:C) ratios higher than 1 allows for the total suppression of the formation of solid carbon.¹³ 

The management of carbon formation is much trickier in the DRM reaction, which is one of the reasons why it has not been operated at an industrial scale. As is shown in Figure 1.3a, the formation of solid carbon is favoured at low temperatures and tends to decrease at higher temperatures. For a stoichiometric inlet mixture CH₄:CO₂, the formation of solid carbon is negligible at temperatures above 950 °C. These high temperatures are favourable for high conversions and total selectivities to syngas, but the energy required to operate the reactor under such conditions is prohibitive. Note that high temperatures would, in addition, promote catalyst deactivation by metal phase sintering. An excess of CO₂ in the mixture tends to decrease the temperature above which the solid carbon is no longer favoured (Figure 1.3b).

The zones of thermodynamic stability of solid carbon are illustrated in Figure 1.4.¹⁴ 

The two main problems for the efficient and stable activity of the catalytic materials in methane reforming reactions are the intensive carbon deposition and the active metal sintering, which are promoted by high reaction temperatures needed for thermodynamic and kinetic reasons.

Many researchers all over the world are trying to overcome these issues. Here, a general view on the currently developed strategies is presented combining the approaches for the improvement of catalytic materials, as well as the approaches for the modification of the reforming process itself.

The essential requirements of catalytic technologies for methane reforming reactions are high conversion rates to react the tremendous amounts of the feed gas, as well as high tolerance against deactivation due to active metal sintering and due to coke deposition.

The supported noble metal catalysts^15,16  are those that present the most attractive catalytic performances in methane reforming, with no strong tendency to deactivation by carbon deposition. Nevertheless, due to their high cost, they are not used at an industrial scale. Ni-based catalytic materials are generally preferred.

As discussed in the previous section, the catalysts for DRM are much more difficult to stabilize against the formation of carbon than are the catalysts for SRM, which can be efficiently operated under an excess of steam. In that latter case, they have to be resistant to sintering, accentuated by the steam and high temperatures.¹⁷ 

Based on the mechanism of coke formation, numerous attempts to reduce or to prevent coke deposition have been carried out by many researchers. Many of them compare the effect of using different kinds of catalyst components and supports. Taking into account the catalytic performance and the cost of catalysts, Ni-supported catalysts on alkaline-earth oxides and rare-earth oxides, especially magnesia, have been intensively studied. The intelligent design of Ni-based catalysts for methane reforming reactions should take into consideration the synthesis method, allowing different Ni structures to be obtained, as the geometric structure of the Ni catalysts plays an important role in the carbon deposition rate and carbon type and morphology. For example, simple thermal decomposition of precursor salts for the preparation of Ni reforming catalysts leads to the Ni (1 0 0) and Ni (1 1 0) orientation with many defects, thus leading to intensive carbon deposition during reforming reactions. The recently developed dielectric barrier discharge (DBD) plasma decomposition method mainly forms the Ni (1 1 1) structure with fewer defects, which suppressed carbon deposition.¹⁸  It was shown that carbon formed on the catalysts obtained by the plasma decomposition method was easy to react with O₂, H₂ and CO₂, indicating strong coke resistance.

The new synthetic methods, such as atomic layer deposition of active metal or the plasma decomposition method, still need further development for the design of coke-resistant catalysts. New characterization technologies should be developed first for convenient monitoring of Ni catalysts in real environments (e.g. in situ Transmission Electron Microscopy (TEM)). It was proposed¹⁸  to combine the DBD synthesis method with reduction by hydrogen for catalyst preparation as an alternative route.

The structure of the active metal is not the only factor that plays an important role. If the metal–support interface is expanded, the metallic Ni dispersion will be better and the metallic nanoparticle size will be smaller, ensuring superior and stable performance of the reforming catalysts. For Ni/La₂O₃ catalysts, the beneficial metallic Ni particles have a size of 4–6 nm with guarantied stability in DRM due to the suppression of coke deposition.

Figure 1.5 shows schematically coke precursor formation over larger metallic particles but no coke formation on smaller Ni particles.¹⁹  According to theoretical and experimental results, the DRM reaction mechanism on large and small Ni particles is proposed as follows: CH₄ is activated on the Ni surface with the formation of H₂ and activated forms of carbon; CO₂ adsorption forms bidentate carbonates at the interface between Ni and the support; the bidentate carbonates react with activated forms of carbon to form CO. Thus, only Ni in close contact with the support at the interface can avoid carbon deposition. If the Ni metal particles are too large, the activated forms of carbon will accumulate over the metallic Ni surface without close contact with the support and thus form coke (carbon fibres, graphite or other forms of coke). Another study confirmed that carbon deposition could be suppressed over the Ni metallic nanoparticles ≤8 nm in size.²⁰  However, small Ni particles are prone to aggregation at high methane reforming temperatures (≥700 °C) due to its thermodynamic instability. This low-stability issue can be addressed by increasing metal–support interaction or by immobilizing Ni nanoparticles inside porous supports. It would be beneficial to investigate the synergetic effect of small particle sizes coupled with the “confinement effect” using porous supports, which is a promising strategy for developing highly active and stable Ni catalysts with good economic viability.

The method of preparation/activation can strongly affect the stability of the Ni-based catalysts. For Ni/silica systems, it is reported²¹  that reduction–oxidation–reduction pre-treatment decreased the average Ni particle size and significantly enhanced catalytic performance in DRM. Accordingly, the initial reduction of Ni precursors after incipient wetness impregnation of silica was efficient at stabilizing catalytic performances, leading to negligible sintering and no carbon deposition.²² 

Stabilization of small Ni metallic nanoparticles could be also achieved by promoting the effect of noble metals, as well as through the modification of the support itself. A recent study²³  showed a comparison of the effects of different noble metals and also the effects of different supports for Ni reforming catalysts. Consequently, an economically promising catalyst for methane reforming reactions is expected to be based on an inexpensive non-noble metal (e.g. Ni) while adding a small yet influential amount of noble and/or non-noble metal, noble metals to enhance the catalytic properties and bimetallic materials showing better performance. Additionally, trimetallic catalysts can also produce similar outcomes. Researchers should investigate the alloying and synergetic effect of nickel and cobalt along with the excellent promoting effect of rhodium in Ni-Co-Rh trimetallic catalysts. From the number of different supports, perovskite-type catalysts could stand out from the crowd. Alternatively, a cheaper catalyst without Rh but with strong potential would confine Ni and Co nanoparticles in mesoporous structures, with catalytic properties enhanced by the “confinement effect”. The recommendations inspired by this study²³  for methane reforming catalyst design are presented schematically in Figure 1.6.

Regarding the choice of support for the active metal, it should be noted that anti-coking abilities can be enhanced by the presence of basic sites on the surface of the catalyst, promoting CO₂ dissociation for DRM of super-reforming reactions, for example. Metal oxide supports with higher basicity can alleviate the deactivation of the Ni surface due to coke formation by promoting CO₂ dissociation at basic sites. They can provide the oxygen atoms to neighbouring Ni sites covered with carbon species and thus facilitate gasification of the carbon species from the metal surface, thus regenerating the catalytically active site for reforming reaction.²⁰  Moreover, it has been reported that the promotion of alumina by alkali or alkaline earth was beneficial to protecting the Ni particles from sintering.²⁴ 

One more promising strategy for the support choice in the design of highly active and stable Ni-based catalysts is the use of mesoporous materials, as was mentioned above. A porous support represents a synergism of the “confinement effects” characteristics: control of metallic particles’ small size and suppression of coke due to the restricted pore volume of porous materials, and more precisely of mesoporous oxides,²⁵  SBA-15 or different core–shell or yolk–shell composite catalytic materials. Comparing catalysts where Ni is encapsulated in inner cavities of different diameters ranging from 5 to 50 nm, a larger inner cavity space is of vital importance to provide sufficient inner room for the survival of small, independent and highly active Ni core particles, as well as to the elimination of absorbed carbon intermediates over the Ni site under very high space velocity conditions. Most importantly, the unique confined structure of a shell with a small radius dimension (≤7 nm) can therefore act as a hard anti-coking framework. This important feature facilitates high coking resistance for the reforming processes with high space velocity. However, a cautionary point is that despite the advantages of inner space, an overly spacious interior cavity could reduce any anti-coking ability. In other studies, it was confirmed that the cavity space in the yolk–shell structures plays an important role in the elimination of coke precursors from the Ni metallic yolk (core), while at the same time preserving the metal yolk particles from sintering.²⁶  Various shell materials, including SiO₂, zeolite, CeO₂ and Al₂O₃, have been studied, and various synthetic methods were used for active metal incorporation, such as sol–gel, micro-emulsion, self-assembly, precipitation, atomic layer deposition, physical mixing, etc. In general, the core/yolk–shell catalysts exhibited better hydrocarbon (not only methane) reforming performance compared with the supported counterparts due to the “confinement effect”.

Even though the core–shell and yolk–shell catalysts show high selectivity to certain desired products, the conversions may be greatly affected.²⁷  A common and known problem of these structures is the high mass transfer barrier, especially when the shell has worm-like pores of small pore size. Therefore, the porosity and shell thickness should be tuned to achieve both high conversion and selectivity. In terms of the practical application of core/yolk–shell heterogeneous catalysts, they are unfortunately regarded as being less effective for the mass-scale production of catalysts.

The rapid deactivation and high coke formation rate associated with non-noble-based monometallic Ni catalysts, coupled with the expensive price of the more coke-resistant and more stable noble metal-containing catalysts, dictate that new and innovative approaches for catalyst design must be investigated to overcome these issues. In general, there is a need for a catalyst that is active with small particle size, well dispersed over a support with basic character and porous structure, has good metal–support interface but nonetheless good reducibility of active metal and has a certain tolerance to carbon formation.

The following is a summary of recommendations for further efficient catalyst development for methane reforming reactions:

• Innovative synthetic method preferentially allowing Ni (1 1 1) to be obtained.

• Control metallic Ni particle size to ≤8 nm.

• Ni-Co bimetallic and Ni-Co-Rh trimetallic catalysts seem to be most promising.

• Perovskite-type supports stand out from the crowd.

• Prefer supports with high basicity.

• Porous supports are desired for enhancing the “confinement effect” for Ni particles.

Methane transformation processes vary by their energetic character (exo- or endo-thermicity). Thus, the reaction could be directly affected by modification of such operating parameters as reaction temperature, gas hour space velocity (GHSV), reactor shape, packing loading and others. To obtain higher CH₄ conversions and increase the catalyst’s lifespan, it is mandatory to have an optimum set of the conditions. Temperature range and GHSV will not be discussed here; instead, this section will focus on recent developments in natural gas transformation technologies.

On the example of DRM as a promising technology for syngas production, a new reactor packing method was developed.²⁸  Different packing strategies allow the maintenance of the reaction temperature along the reactor within the specified range, as well as the conversion in an acceptable range for ±20% change in the feed flow rate. The proposed design can maintain the temperature in the region where coke is rarely generated.

Another efficient strategy for suppressing coke formation could be admixture of the feed with other different hydrocarbons. In thermo-catalytic decomposition of methane for the production of pure hydrogen, the main drawback limiting its industrial applications is catalyst deactivation due to carbon formation. Thus, co-feeding of CH₄ with other molecules such as C₆H₆, H₂S, CO₂, alkanes and alcohols can be utilized as an efficient method for significantly improving H₂ production yields and for switching the coke formed towards more active forms of carbon.²⁹  The co-feeding of methane with other feedstock molecules increases the production of more active carbon on the catalyst surface as compared to carbon eliminated from methane, and it is an effective solution to overcome rapid catalyst deactivation. The activity originating from different molecules is in the order C_benzene > C_acetylene > C_ethylene > C_propane and C_methane.³⁰  Limited work has been done on co-feeding of methane with other hydrocarbons to stabilize catalyst activity and deactivation time.

Sorption-enhanced reforming of methane is the combination of the reforming process and the CO₂ abstraction reaction to yield high-purity hydrogen. For implementation of this strategy, a bi-functional, highly active and stable material is always demanded, usually based on CaO in the mixture with Ca₃Al₂O₆ or other efficient CO₂ sorbents. The sorption–desorption cycling of the process in this case provides greater H₂ production together with little coke formation and almost completely suppressed active metal sintering.³¹ 

An emerging alternative to conventional technologies is CH₄ reforming based on chemical looping, which is a process designed mostly for syngas production. This is an option that is applicable in a diverse range of processes and could be divided into sub-reactions.³²  The link between the sub-reactions is a solid mediator, usually in the form of a metal oxide (e.g. Fe₃O₄, CeO₂, NiO, CuO), which is reduced and re-oxidized in a cyclic progression of the sub-reactions.³³  Figure 1.7 shows the scheme of the chemical looping for DRM, where the process globally corresponds to DRM, although in practice it consists of two independent gas–solid reactions:

The results of DRM chemical looping showed that carbon formed during exposure to methane can be fully removed during the re-oxidation by CO₂.³⁴  This technology allows for the avoidance of carbon accumulation and catalyst deactivation.

To equilibrate the energetic balance of methane reforming reactions for syngas or for hydrogen production, a combination of endothermic and exothermic reactions is used – so-called bi-reforming, co-reforming, tri-reforming or super-reforming.³⁵  Even a low concentration of O₂ in the feed gas dramatically suppresses carbon deposition. The simultaneous use of O₂ and H₂O may not only decrease the formation of carbon, but also gives an appropriate H₂:CO molar ratio to match the downstream use of the produced syngas. Additional combination with a membrane reactor has the advantages that the reaction can be operated at much lower temperatures and CO_x-free hydrogen can be obtained. This is favourable for use in fuel cells as an ideal small-scale energy system. The catalytic results using the membrane approach are still problematic because they are limited by the intrinsic performance of the catalysts and by diffusion across the membrane.³⁶ 

From an industrial point of view, the natural gas transformations in syngas and in pure hydrogen are characterized by severe economies of scale: plants must be very large. Thus, direct approaches for converting methane to valuable products that are economical on a small scale are of extreme interest. Such processes could rescue bio-methane or so-called stranded natural gas (flared, wasted or unused) that is produced in too small amounts and is too remote for economic transportation. Several commercialization attempts of emerging processes, pilot plant experiments or both for methane activation processes are collected in Table 1.2.³⁷ 

Oxidative coupling of methane to ethylene is currently being commercialized by Siluria. Ethylene is a valuable molecule of interest; as an alternative use, it could be oligomerized to fuel-type molecules. Commercial attempts reported by Catalytic, Inc., of Pt-catalysed oxidation of methane with SO₃ in concentrated sulfuric acid or oleum resulted in the highly selective formation of methylbisulfate.³⁸  This can be hydrolysed to methanol and under well-chosen conditions can be produced at rates that are in the industrial processes range.³⁹  Reaction of methane with SO₃ in oleum with substantially different chemistry is also the basis of the new methanesulfonic acid process; Methion has attempted to market the process.⁴⁰  Methanesulfonic acid could replace mineral acids in many special applications.

Currently, only the production of transportation fuels has a demand at a level corresponding to the industrial processes scale. However, other chemicals produced from methane with much higher added value compared to fuels and other chemicals considered as energy vectors could be attractive options.⁴¹ 

Even if active and stable catalytic performance can be achieved from the process of steam reforming of methane to produce syngas, the limitations of the development of dry reforming of methane are mainly due to rapid deactivation by carbon deposition.

Despite extensive work on the design and stabilization of DRM catalysts, coupled with engineering strategies to overcome deactivation, the industrial development of DRM seems hardly achievable.

Direct conversion of methane to chemicals and fuels remains the best strategy in the future to avoid the intermediate stage of reforming.