Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Much has been written over the last few decades about the potential use of CO2 as a feedstock in the chemical industry, and particular attention has been given to its possible use in the conversion of methane and other hydrocarbons using so-called dry reforming to give syngas, a mixture of CO and hydrogen. The CO2 or dry reforming of methane can be represented by the equation

$$ {\mathrm{CH}}_4+{\mathrm{CO}}_2\to 2\mathrm{CO}+2{\mathrm{H}}_2 $$
(6.1)

In any practical application, in order to achieve acceptable conversions, this reaction would have to be carried out at very high temperature (up to as much as 1,000 °C) when the conversion would be close to chemical equilibrium. The reaction is always accompanied by the reverse water–gas shift reaction, generally fully at equilibrium:

$$ {\mathrm{CO}}_2+{\mathrm{H}}_2\rightleftarrows \mathrm{CO}+{\mathrm{H}}_2\mathrm{O} $$
(6.2)

A consequence of the combination of these two reactions is that the CO:H2 ratio obtained is generally rather lower than the value of 1:1 to be expected if only the reaction of Eq. 6.1 occurred. The reaction system is further complicated by the fact that the conditions used generally favour carbon deposition. Carbon deposition can be thought of as occurring by one of three interdependent reactions:

$$ 2\mathrm{CO}\rightleftarrows \mathrm{C}+{\mathrm{CO}}_2 $$
(6.3)
$$ {\mathrm{CH}}_4\rightleftarrows \mathrm{C}+2{\mathrm{H}}_2 $$
(6.4)

and

$$ \mathrm{C}\mathrm{O}+{\mathrm{H}}_2\rightleftarrows \mathrm{C}+{\mathrm{H}}_2\mathrm{O} $$
(6.5)

In most practical cases, whether or not carbon deposition is likely to occur is determined by thermodynamic considerations; further, as Boudouard carbon (Eq. 6.3) is generally expected at low temperatures, carbon deposition under dry reforming conditions is most likely to occur by either methane decomposition (Eq. 6.4) or CO reduction (Eq. 6.5). The problem of carbon deposition is a key feature of the process and we will return to it many times below.

A search of Web of Science using the search terms “methane, CO2 and reforming” shows that more than 1,900 papers with these three words in the titles and/or abstracts had been published between 1987 and mid-2013. Table 6.1 shows the approximate number of papers that include these search terms published up to 1994 and in each 5-year period since then, together with the approximate total number of citations to these papers given in each period. It is clear that there is little sign of the interest in the subject showing any decline; indeed, the yearly number of publications reached an all-time high of more than 200 in 2012, this being significantly higher than the previous maximum of 160 in 2011. As will be discussed further below, it is highly questionable if this level of activity is justified by the potential applicability of the process. The only justification for such a level of activity is probably that the work reported provides new insights into particular types of catalysts, CO2 reforming being used as a model reaction.

Table 6.1 The numbers of papers on the topic of dry reforming of methane published during various time periods since 1987 together with the numbers of citations listed in those time periods
Full size table

Other combinations of search terms to those used in the search of the Web of Science discussed above show up significant additional material.Footnote 1 This review is based largely on the material gleaned from the search related to Table 6.1. However, some references are also included that did not show up in this search, in most cases because the keywords of the original articles were different; there is also no doubt that some important papers will have been omitted for the same reason.

Interestingly, the first paper on CO2 reforming of methane listed in the search outlined in Table 6.1 is a relatively under-cited one by Fish and Hawn on the use of CO2 reforming in a thermochemical cycle [1], a topic to which we will return briefly below. The second paper on the list, by Gadalla and Bowers [2], is concerned with the use of a range of commercial steam reforming catalysts, showing that the most promising results were obtained with alumina supports containing magnesium or calcium oxide. (A further paper by Gadalla and Somer [3], published in 1989, gives more details on this subject.) As will be discussed in a later section, Ni catalysts based on Mg-containing supports still appear to be among the most effective catalysts for the process.

Many of the papers that have been published over the last 20 years or so justify the work carried out by claiming that a process involving the consumption of both methane and CO2 would help to provide a solution to the emission of these two greenhouse gases. It needs to be emphasised from the outset that such claims are largely unjustified since the quantity of CO2 that might be used in syngas production by CO2 reforming would be negligible compared with the very large emissions of CO2 currently causing concern, even if it was possible to collect the CO2 in a sufficiently pure form for the CO2 reforming reaction. Further, dry reforming produces syngas with a composition usable only in limited applications; as shown above (see Eq. 6.1), using methane as co-reactant, the syngas ratio, CO/H2, is approximately 1.0. Much more suitable ratios can be attained by the addition of either steam or oxygen to the reactant stream, such addition also having the benefit that the potential for carbon deposition on the catalysts used in the process is significantly reduced.Footnote 2 An alternative to the dry reforming process using CO2 in the primary feed is to operate the steam reforming reaction:

$$ {\mathrm{CH}}_4+{\mathrm{H}}_2\mathrm{O}\to \mathrm{CO}+3{\mathrm{H}}_2 $$
(6.6)

and then to carry out the reverse water–gas shift reaction (reverse of reaction (6.2)) in order to achieve the desired CO/H2 ratio:

$$ {\mathrm{CO}}_2+{\mathrm{H}}_2\to \mathrm{CO}+{\mathrm{H}}_2\mathrm{O} $$
(6.7)

There are several advantages to be achieved using this approach, these including the following: (i) the problem of C deposition can largely be avoided since the catalysts and process conditions for the steam reforming reaction under carbon-free conditions are well established commercially; and (ii) the CO2 required for the process can be extracted from the effluent of the reformer furnace (burning natural gas) used to heat the tubular steam reforming reactor system. Hence, the only additional processing step required is the separation unit to remove CO2 from the combustion gases, a step also required if dry reforming is to be carried out. Further, although the cost of a steam reforming plant is very high and a major contributor to the all-over cost of any process requiring a syngas feed, it is unlikely that there will be much reduction in the total operation cost if dry reforming is used.

Accepting, for the reasons given above, that dry reforming is unlikely to solve the greenhouse gas problem, other justifications have to be found for the very extensive work on the subject that has been carried out over recent years. As will be discussed in more detail in the following sections, much of the most recent work has been devoted to trying to find catalysts that are resistant to carbon deposition while having suitable activities and stabilities under reaction conditions.Footnote 3 Although much of the work published recently undoubtedly adds incrementally to the literature of the subject, there appear to have been no major breakthroughs: most, if not all, of the catalysts that have been examined under suitable conditions unfortunately show gradual deactivation due to carbon deposition. Further, many of the catalyst systems studied bear remarkable similarity to ones that have been studied previously for other closely related reactions, for example, the steam reforming of methane. In many cases, such similarities do not appear to have been recognised by the authors of these papers.

This review traces briefly the development of processes used for the production of syngas from methane (and higher hydrocarbons) and discusses the interest in the use of CO2 as a co-reactant, paying particular attention to the conversion of methane. It then considers the thermodynamics of such conversion processes, highlighting the problem of carbon deposition and the need for operation at high temperature. Recognising the similarity between dry reforming and steam reforming and the fact that the latter is a well-established industrial process, mention is made of the catalysts used commercially for the steam reforming of methane and of higher hydrocarbons, with a short digression on the catalysts used for methanation. The review then considers various groups of papers that have been devoted to studies of different catalyst types, particular attention being given to the use of noble metals (particularly Pt) and other transition metals such as Ni, Co and Mo, on a variety of different supports. The review concentrates on the literature on methane reforming, but it should be recognised that many of the catalysts developed will be equally applicable to the reforming of higher hydrocarbons, the only complication being that methane is also a potential product when the methanation reaction is thermodynamically feasible (at lower temperatures, see Sect. 6.2.2).

2 Historical Background

2.1 Syngas Production from Hydrocarbons

Before embarking on a more detailed discussion of the dry reforming reaction and the catalysts that have been used for that reaction, it is sensible first to give a brief outline of the use and importance of the steam reforming of methane as a source of syngas. The subject of the steam reforming of methane (Eq. 6.6) has been reviewed by various different authors, from the point of view of both the processes involved [4] and of the catalysts used [5]. Although, as pointed out by Rostrup-Nielsen [4], some papers and patents on the steam reforming of hydrocarbons had been published as early as 1868, the first industrial steam reformer was commissioned in Baton Rouge in 1930 by the Standard Oil of New Jersey, and this was followed by another reformer commissioned by ICI in Billingham in 1936; in both cases, the main product required was hydrogen, and so the CO formed was removed using the water–gas shift reaction. While steam reforming in the United States was usually carried out with methane as the feed hydrocarbon, the raw material of choice in Europe until the 1960s was naphtha; with the advent of European natural gas supplies in the second half of that decade, methane also became the feedstock of choice.Footnote 4 It is interesting to note that although there has been some limited commercial interest in the use of CO2 as a feed instead of water, the only plant using a methane/CO2 feedstock has been a pilot plant operated by Haldor Topsoe in Texas in which a sulphur-modified steam reforming catalyst is used [6]. It is also worth noting that in the so-called Midrex process for the direct reduction of iron ore, a mixed feed of methane with CO2 and steam is used in a reformer; the CO2 and steam for the reforming step are produced during the reduction of the iron ore, and the syngas produced in the reformer is fed directly to the reduction reactor [6, 7].

Although other active materials, notably the noble metals, have been examined for use as the active components in steam reforming catalysts, the only ones that have been used commercially are those including Ni as the active component. This is not only a matter of cost: Ni is the only non-noble transition metal that is maintained in its metallic state under steam reforming conditions since the equilibrium

$$ \mathrm{Ni}\mathrm{O}+{\mathrm{H}}_2\rightleftarrows \mathrm{Ni}+{\mathrm{H}}_2\mathrm{O} $$
(6.8)

lies to the right-hand side as long as the ratio H2/H2O is greater than about 0.01; in other words, the nickel will be in its reduced (and active) form as long as there is approximately 1 % hydrogen in the reactant gas [5]. (With the exception of the noble metals, much higher proportions are needed for most of the other transition metals.) A variety of different Ni-containing formulations have been commercialised. However, among the most commonly used materials are the ICI (now Johnson Matthey) 46–1 catalyst (predominant components in the reduced form are Ni, alumina and K2O with CaO, SiO2 and MgO added as a cement binder) and the Haldor Topsoe catalyst (containing Ni, MgO and alumina). Other catalyst manufacturers supply similar materials. The potassium of the ICI formulation is added to help minimise C deposition, and the MgO of the Haldor Topsoe material serves a similar function. As mentioned above, the catalyst used in the Haldor Topsoe pilot plant for dry reforming (see above) is partially sulphided [4, 6, 8]. This was based on work on the subject of S passivation for application in the steam reforming process [9].

2.2 Methanation

From a historical point of view but also in relation to work on the formulation of catalysts, it is interesting to note that there occurred in the late 1970s and early 1980s a period of very significant academic and industrial research activity on the topics of methanation and methanation catalysts. This interest arose because natural gas supplies in the United States could not keep up with demand. As a result, it was desirable to find means of converting syngas (produced from coal) into so-called synthetic natural gas (SNG):

$$ \mathrm{CO}+3{\mathrm{H}}_2\to {\mathrm{CH}}_4+{\mathrm{H}}_2\mathrm{O} $$
(6.9)

This reaction is the reverse of the steam reforming reaction (Eq. 6.6) and is highly exothermic. Hence, much of the work from that period concentrated on finding catalysts with high stability at higher temperatures that also resisted carbon deposition. Many of the catalysts examined and reported in many papers of that period had great similarity to those that have been examined more recently for dry reforming. This author reviewed the literature on methanation and steam reforming in 1985 [10].

2.3 Methane Coupling

Following on from the great interest from the catalysis community in the topic of methanation, the subject of the oxidative coupling of methane to give C2H6 and C2H4 became predominant during the second half of the 1980s as a result of a paper by Keller and Bhasin published in 1982 [11]. These authors showed that it was possible, using a cyclic process in which methane and oxygen were fed in turn to a catalyst bed, to obtain reasonable yields of C2 hydrocarbons, the catalysts being operated at temperatures around 800 °C. Keller and Bhasin used a range of α-alumina-supported materials (including the oxides of Cd, Sn, Sb, Tl, Bi and Pb) as catalyst and postulated that an oxidation–reduction mechanism was involved. Hinsen and Baerns were the first to carry out the reaction with methane and oxygen fed simultaneously, and they obtained good yields using a lead/γ-alumina catalyst [12]. A paper by Ito, Wang, Lin and Lunsford published in 1985 introduced the use of Li/MgO materials [13]. Many more similar catalysts were introduced over the next decade, and a number of very extensive reviews have been written on the subject [14–17].

It was soon recognised that the reaction occurs by a combination of heterogeneous reactions and gas-phase radical reactions and that the products included not only ethane and ethylene but also a mixture of higher hydrocarbons, water, CO, CO2 and hydrogen. The participation of gas-phase reactions as well as limitations introduced by the explosive limits of methane/oxygen mixtures leads to a limit in the potential yields of C2 hydrocarbons produced in the reaction, the best values being just above 20 %, too low to enable the reaction to have commercial application with current natural gas prices. Although papers on the subject were published in a variety of different journals, a very good overview of the work that was carried out is to be found by examining the proceedings of the earlier “Natural Gas Conversion” meetings [18–21], from the first of the series held in 1987 in Auckland, NZ, to the fourth meeting held in 1998 in Messina, Italy; by 2001, when the 6th meeting of the series was held in Alaska, the topic of methane coupling had disappeared. What had been gained from all this work was a much greater understanding in the academic community of the operation of catalytic reactors at elevated temperatures and the problems of catalyst stability. (It is of interest to note that renewed interest in methane coupling has recently been evident, this being spurred by the current low prices of natural gas.)

2.4 Dry Reforming and Partial Oxidation of Methane

When at the beginning of the 1990s an interest developed in the production of syngas from natural gas using either partial oxidation or “dry reforming”, it was not surprising that the research community that had been involved in the work on methane coupling transferred its interest to studying one or other of these two reactions. This transfer is clearly seen in the proceedings of the “Natural Gas Conversion” meetings referred to above. The majority of the papers under the title “Methane Conversion” in the second meeting held in Oslo in 1990 [18] were related to oxidative coupling, there being none on partial oxidation or CO2 reforming. This bias was continued at the third meeting held in Sydney in 1993 [19] with more than 20 papers on methane coupling and only four on dry reforming; two of the latter were reviews of the subject by J. R Rostrup-Nielsen and by J. H. Edwards and A. M. Maitra (discussed further in the following section). However, by the time of the fourth meeting held in the Kruger National Park in South Africa in 1995 [20], there were 16 papers on oxidative coupling, 13 papers on partial oxidation and 8 papers on CO2 reforming. By the fifth meeting held in Taormina, Italy, in 1998 [21], the latter two topics had become the predominant subjects of interest related to methane conversion. The relevant papers from these meetings will be outlined in Sect. 6.4.1.

As discussed above (see Table 6.1), there has since 1996 been an explosive development of interest in CO2 reforming and partial oxidation, so much so that it is virtually impossible to write a comprehensive review covering every aspect of the work that has been carried out on either topic. The following sections will therefore concentrate on some of the more important types of catalyst that have been investigated, particular attention being paid to papers describing those types of catalyst that have reasonable stabilities under “dry reforming” conditions and to papers giving new insights into the mechanism of the reaction or advancing other approaches to the conversion process. No great detail on any of the papers is given; instead, summaries are given in tabular form of their main conclusions. Further, no attempt is made to discuss any of the parallel literature on partial oxidation, exceptions being when it has been shown that the addition of oxygen to the “dry reforming” mixture has beneficial effects or when the results on the partial oxidation reaction on a particular catalyst have some relevance to the dry reforming reaction on identical or similar catalysts. In selecting the literature for inclusion, use has been made of citation indexing, the searches being based on the use of Web of Science and/or Scopus. As a starting point, some of the papers appearing in the proceedings of the earlier Natural Gas Conversion meetings referred to above have been used to identify some of the major themes of research. Scopus and Web of Science searches have then been used to identify some of the (in most cases) more recent key papers related to each of these themes, and some of the main conclusions of these are tabulated. Finally, the main objectives and conclusions of some very recently published papers on the subject that have not yet achieved a significant citation record are summarised.

Before embarking on our summary of the literature, a brief outline of the thermodynamic limitations associated with the dry reforming reaction and associated side reactions is given since these limitations have some importance in discussing the papers from the literature.

3 Thermodynamics of the CH4 + CO2 Reaction

The dry reforming of methane (Eq. 6.1) is an endothermic process, and so the maximum conversion calculated thermodynamically increases with increasing temperature. This is illustrated schematically in Fig. 6.1 that shows the thermodynamic conversions for various feed compositions as a function of temperature of reaction [22]. It can be seen that a temperature in excess of about 850 °C is necessary to obtain adequate conversions using a stoichiometric CH4/CO2 feed (1:1).

Fig. 6.1
figure 1

Thermodynamically calculated conversions of methane as a function of temperature for a series of different feed ratios. Calculations carried out using HSC chemistry (version 1.10, Outokumpu Research, Finland) (Reprinted from Ref. [22]. Copyright 2005, with permission from Elsevier)

Full size image

It is to be expected that the water–gas shift reaction (Eq. 6.2) will be in equilibrium under CO2 reforming conditions and so the CO/H2 ratio of the product gas can be calculated easily from a knowledge of the feed compositions and the measured methane conversions. It should be noted that the so-called selectivity of the reaction is very frequently given in publications reporting the behaviour of novel catalysts; unless it is shown very clearly that the water–gas reaction is not in equilibrium for that catalyst system, such values have no real significance as they are thermodynamically rather than kinetically controlled.

Figure 6.2 shows the yield of carbon to be expected if carbon deposition is possible for two possible cases: whether or not the water–gas shift reaction is considered to be in equilibrium [22]. This diagram was constructed assuming that there is a closed system and that the amount of carbon formed relates to the amount fed as methane and CO2; the results should therefore be seen only as a guide as to when carbon would form in a continuous flow reactor. It is clear that carbon formation is possible over the whole range of temperatures in both cases considered (with or without the water–gas shift reaction being equilibrated), and that the amount of carbon formed will be lower at high temperatures. When the water–gas shift reaction does not take place, the amount formed at the highest temperatures is much lower; however, in most cases reported in the literature, the water–gas shift reaction is at equilibrium at the higher temperatures and so this situation does not apply. Rostrup-Nielsen uses a different approach, favouring the use of so-called carbon limit diagram, developed first for the steam reforming reaction [6, 23]. He shows that in the absence of the addition of steam, the CO2 reforming reaction operates under conditions when carbon can form regardless of the CH4/CO2 ratio. He concludes that the noble metals, most of which do not form carbides (necessary intermediates in carbon formation) have a greater potential for use as CO2 reforming than do catalyst formulations using nickel. The topic of carbon deposition will be handled further after a discussion of the types of catalyst studied.

Fig. 6.2
figure 2

Thermodynamically calculated proportions of carbon formed under CO2 reforming conditions with CH4/H2O = 1.0; (■), with the reverse water–gas shift reaction in equilibrium; (□), without the reverse water–gas shift reaction (Reprinted from Ref. [22]. Copyright 2005, with permission from Elsevier)

Full size image

4 Catalysts for the Dry Reforming of Methane

4.1 Papers on CO2 Reforming Published in the Proceedings of the Natural Gas Conversion Symposia up to 1998

Tables 6.2, 6.3, and 6.4 list the relevant papers presented at the Natural Gas Conversion Symposia between 1993 and 1998: the third (held in 1993, published in 1994 [19]), the fourth (held in 1995, published in 1997 [20]) and the fifth (held and published in 1998 [21]). The following paragraphs first summarise two review articles from the third meeting and then outline some of the important aspects of the other papers, this being in preparation for a more general discussion of the literature on the subject since about 1997.

Table 6.2 Papers from NGC 3
Full size table
Table 6.3 Papers from NGC4
Full size table
Table 6.4 Papers from NGC V, SSSC 119 (1998) (Italy)
Full size table

In his review delivered at the third symposium [6], Rostrup-Nielsen of Haldor Topsoe A/S (Lyngby, Denmark) gave a very clear exposition of the then-available information on the CO2 reforming of methane, paying particular attention to earlier work on the subject. He noted that the reaction was examined by Fischer and Tropsch as early as 1928 [49]; these scientists studied a series of base metal catalysts and found that nickel and cobalt gave the best results, the product gas compositions being close to thermodynamic equilibrium.

Rostrup-Nielsen also summarises the work of Bodrov and Apel’baum who showed in 1967 [50] that the kinetic expression for the CO2 reforming reaction over a nickel film was similar to what they had found previously for the steam reforming reaction [51], this indicating that the mechanism was very similar for both reactions.

Rostrup-Nielsen discussed industrial processes that involve CO2 reforming, these including the SPARG process making use of a sulphur-passivated Ni catalyst. The original research relating to that process is detailed in a research paper by Dibbern et al. referred to above [8]. The technology used, adding traces of H2S to the reactor feed, is similar to that previously introduced by Haldor Topsoe to enable steam reforming to be carried out at low steam/methane ratios [9]. Rostrup-Nielsen also presented in his review some data on the CO2 reforming of methane using the noble metals and Ni supported on MgO. He concluded that the noble metals are much less susceptible to carbon poisoning due to the fact that they do not dissolve C in their bulk. He then described briefly some work in which Rh and Ru catalysts, the most promising of the metals studied, were used for carbon-free operation in a pilot plant under conditions in which carbon is favoured thermodynamically. He concluded, however, that because of their scarcity neither of these metals is likely to be used in other than niche applications. In conclusion, Rostrup-Nielsen discussed the limitations of possible processes for the use of CO-rich gases from CO2 reforming. He pointed out that the amounts of CO2 used in such processes will be virtually insignificant in comparison with the total worldwide emissions of CO2. For example, the CO2 that would be required for the production of 5 million tons of acetic acid per year (the current global rate of production) by the reaction:

$$ {\mathrm{CH}}_4+{\mathrm{CO}}_2\to {\mathrm{CH}}_3\mathrm{COOH} $$
(6.10)

would correspond to the emission from only one 500 MW coal-based electricity power station. Yearly worldwide production of methanol is four times higher, and so supplying syngas via CO2 reforming of methane would require the output from four such power plants. More importantly, the current technology for CO2 extraction makes the use of CO2 from flue gases uneconomical and so other sources of CO2 would have to be considered. The article by Rostrup-Nielsen contains many other details (and warnings) related to the potential of a process for CO2 reforming. Anyone working in the area should therefore read it with care.

A second review article was delivered at the same symposium; this was reported by Edwards and Maitra of CSIRO (North Ryde, Australia) [24]. Having also discussed the various reactions involved and their thermodynamics, these authors gave a very comprehensive review of work on CO2 reforming published up until 1993. They showed that most of the noble metals and also nickel had been studied and that oxides such as alumina and magnesia had been used as supports for the active metals. They also pointed out that carbon deposition rates seem to depend on the support used and possibly on catalyst morphology, two factors that emerge repeatedly in more recent publications. They mentioned specifically the paper by Gadalla and Somer referred to above [3] in which the use of a Ni/MgO catalyst for up to 125 h was reported; at the end of the experiment, these authors observed only a minute trace of carbon deposition.

Edwards and Maitra also discussed the results of two papers where the activities of a series of metals are reported. In the first, by Takayasu et al. [52], Ni, Ru, Rh, Pt and Pd catalysts supported on “ultrafine single crystal MgO” were compared, the activities being in that order. In the second, by Ashcroft et al. [53], catalysts comprising of Ni, Ir, Rh, Pd and Ru supported on alumina were examined, the order of activity being in the order given. While Takayasu et al. had found that Ru was one of the most active metals, Ashcroft et al. reported that it is one of the least active. (Rostrup-Nielsen has also reported it as being very active [6].) However, both groups found that Ni and Rh are among the most active metals. Listing other papers on the topic, Edwards and Maitra also pointed out that Solymosi et al. [54] and Masai et al. [55] had previously reported that Rh supported on alumina is one of the most active catalysts for the reaction.

Edwards and Maitra closed their review [24] with a discussion of the potential application of thermochemical heat pipe applications for the recovery, storage and transmission of solar energy, showing that the CO2 reforming reaction can be coupled with the methanation reaction to provide a method of energy transmission (a topic also mentioned briefly by Rostrup-Nielsen in his more extensive review [6]). We return briefly to this topic later (see Table 6.13).

The research papers on CO2 reforming of methane from the meetings of the Natural Gas Conversion series listed in Tables 6.2, 6.3, and 6.4 cover the use of most of the metal/support combinations that have since been examined in more detail. The vast majority of these papers reported work involving nickel, this being supported on a variety of different oxides, in particular, alumina [6, 25, 27–29, 33, 35, 37, 41, 42], silica [31, 38, 40, 46], magnesia [6], lanthana [30, 38] and zirconia [25]. Some attention was also given to the use of catalysts derived from nickel-containing compounds or from solid solutions of NiO in MgO [48]. Most of the noble metals are also featured, particular attention being focused on Pt [2, 25, 32], Rh [26, 27, 34], Ru [34] and Ir [31], again on a variety of supports. It is interesting to note that there was a wide range in the temperatures of operation, from 400 to 900 °C. As was pointed out above, operation at high temperature is necessary to give acceptably high conversions and so the results at lower temperatures, although giving information on relative activities, have less practical relevance. It should also be noted that the paper by York et al. [42] reports the use of carbides, while the papers by Kikuchi and Chen [33] and Ponelis and Van Zyl [34] both describe the use of membrane systems.

4.2 Highly Cited Publications from the General Literature

In the following sections, we give a listing of the more significant papers from the recent literature under headings suggested by the brief outline given in the last paragraph. Owing to the very large literature on the subject of CO2 reforming of methane, the papers selected for mention under each heading are chosen in relation to their citation records, recognising that those papers most frequently referred by others are likely to be the most important in that field. In the tables, the number of citations to each of the publications given by the Web of Science in May 2013 is shown for comparison purposes; clearly this is only a snapshot of the situation, and it should also be recognised that younger publications will have correspondingly lower numbers of citations than older ones.Footnote 5 A reader with a particular interest in one or another type of catalyst would gain an up-to-date picture of the literature on that topic by doing his/her own citation study based on the appropriate key papers listed in the tables.

Table 6.5 lists some of the most cited reviews on the subject of CO2 reforming. Some of these are very general, covering a whole range of topics from basic aspects to pilot plant results [58], while others are much more focused on one type of catalyst [60, 63]. The review by Armor [64] is of particular interest as it not only covers both steam reforming and CO2 reforming but also deals with matters such as problems associated with purification of CO2. Kodama [66] reviews the subject of fuel production using sunlight.

Table 6.5 Some of the most cited reviews on CO2 reforming listed in the Web of Science
Full size table

Nickel, although very active for the conversion of methane by either steam reforming or CO2 reforming, suffers from the problem that it rapidly becomes deactivated by the formation of carbon, this carbon generally being in the form of nanotubes with a nickel crystallite at the tip. Much of the effort has therefore been in trying to find suitable additives for the catalysts that help to minimise this form of carbon deposition, often by modifying the Ni particle size or the support surface in the region of the nickel crystallites. It is now reasonably well established that small nickel crystallites are much less susceptible to C growth, but anchoring the active metal to the support is also known to hinder filament growth. Table 6.6 lists some of the papers reporting the use of catalysts consisting of Ni supported on alumina or modified aluminas. A number of additives are seen to improve the behaviour of the catalysts in relation to the problem of carbon deposition, particularly the alkali [67, 75] and alkaline earth metals [67, 74], lanthana [71] and ceria [68, 73]; ceria is particularly effective.

Table 6.6 Some of the most cited papers dealing with Ni supported on Al2O3 or modified Al2O3
Full size table

The effect of reducing the particle size is seen particularly with catalysts based on Ni on MgO; see Table 6.7. This is because the precursor to the active catalyst often involves the formation of a solid solution of NiO in MgO; see, for example, references [77, 80, 83]. Catalysts based on supports such as MgAl2O4 are also very effective [81–83].

Table 6.7 Some papers dealing with Ni supported on MgO
Full size table

Another support that has received some attention is zirconia, sometimes alone but also often promoted by other species such as ceria [84, 85] or magnesia [83]; see Table 6.8. The effect of Ni particle size on stability has been shown for a zirconia support by Lercher et al. [86]. Other supports studied have included titania [88], lanthana [89], silica [90] and a La–Sr–Ni–Co perovskite [91]; see Table 6.9.

Table 6.8 Some papers dealing with Ni supported on zirconia
Full size table
Table 6.9 Some papers dealing with nickel on other supports
Full size table

Noble metals have the advantage over Ni in that they do not dissolve carbon, and so the problem of the growth of carbon filaments does not occur with them. As a result, there has been a significant amount of work using catalysts containing noble metals on a variety of supports. Table 6.10 lists some highly cited papers describing the use of supported Pt catalysts on various different supports. Work from this laboratory [60, 95] and also from Bitter et al. [96] has been concerned with the use of Pt on zirconia, this usually being promoted by a small quantity of alumina. Other promoters such as La2O3 or CeO2 are also possible [93].

Table 6.10 Some papers dealing with Pt catalysts on various supports
Full size table

The other noble metals have also received significant attention, particularly Rh [97–102], Ir [99, 101] and Pd [101, 103]; see Table 6.11. As discussed by Rostrup-Nielsen, a series of successful pilot plant tests have been carried out using a catalyst based on Rh. However, the cost of Rh is prohibitively high so that this work is unlikely to be applied.

Table 6.11 Some papers dealing with catalysts containing Rh, Ru, Ir or Pd
Full size table

Because of the high cost of the noble metals and the resultant problems of operating large scale using them, a significant amount of attention has been paid to the possibility of using MoC2 or WC as the active component of a catalyst, it being well established that these carbides exhibit properties similar to those of the noble metals. Claridge et al. ([105], see Table 6.12) showed that both of the carbides were active for the reaction, but they reported that there was some evidence of deactivation caused by oxidation of the carbides. Work from this laboratory [106] showed that this was probably due to the fact that the entrance to the CO2 reforming bed operates under slightly oxidising conditions, this causing the carbide to oxidise, and that the exit part of the bed was stable. The activity of a Mo2C material supported on ZrO2 could be improved significantly by promotion with Bi species.

Table 6.12 Some papers dealing with carbides as catalysts
Full size table

Finally, Table 6.13 lists some references describing work in various laboratories on energy transport systems involving the CO2 reforming reaction as the endothermic part of the system. It is interesting to note that despite the comments above, the metal of choice for this work appears to be Rh, probably because of the reliable performance far outweighs the disadvantage of cost.

Table 6.13 Some papers discussing the use of dry reforming as part of an energy transportation system
Full size table

4.3 Some Recent Publications on CO2 Reforming of Methane

In conclusion, we list in Table 6.14 a number of recent publications on the subject of CO2 reforming chosen rather at random from the large number that have been published recently. Many of the themes that we have highlighted above return frequently. While some of the papers report significant improvements in previously described catalyst systems (e.g. [110, 111, 123]), others use the reaction purely as a test reaction (generally at rather lower temperatures) to provide a method of characterising a catalyst system (e.g. [116, 121]). One example of an improvement is to be found in work from this laboratory ([123]; see also [87]) that shows that a very stable catalyst with good activity can be achieved with the Ni–Zr–Mg–O system as long as K+ ions are included in the structure.

Table 6.14 Some recent papers (2013) on methane dry reforming
Full size table

5 Conclusions

The dry reforming of methane has provided a fertile area of research for more than two decades, and some significant advances have been made in the design of catalysts suitable for the reaction. However, it is at present unlikely that the reaction will be used in practice unless it is combined with either partial oxidation or steam reforming in order to avoid the problems associated with carbon formation. This review has concentrated almost exclusively on the CO2 reforming of methane, paying little attention to the reactions of other hydrocarbons. However, many of the constraints encountered equally to the reforming of other molecules such as naphtha with the added complication that methane may be formed as a product.

Dedication

The author wishes to dedicate this review to the memory of the late Laszlo Guczi with whom he collaborated in a joint project under the auspices of an ERA-Chemistry project (coordinated by Alain Keinemann) until shortly before his death in December 2012.