Strategies for the Direct Catalytic Valorization of ... - ACS Publications

3 downloads 113 Views 850KB Size Report
Mar 31, 2016 - (8) Parkyns, N. D.; Warburton, C. I.; Wilson, J. D. Catal. Today. 1993, 18, 385−442. (9) Rostrup-Nielsen, J. R. In Nato Sci. Ser. II Math; Derouane ...
Perspective pubs.acs.org/acscatalysis

Strategies for the Direct Catalytic Valorization of Methane Using Heterogeneous Catalysis: Challenges and Opportunities Alma I. Olivos-Suarez,† À gnes Szécsényi,†,‡ Emiel J. M. Hensen,‡ Javier Ruiz-Martinez,*,§,∥ Evgeny A. Pidko,*,‡,⊥ and Jorge Gascon*,† †

Catalysis Engineering, Chemical Engineering Department Delft University of Technology, Julianalaan 136, 2628 BL Delft, The Netherlands ‡ Inorganic Materials Chemistry group, Schuit Institute of Catalysis, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands § AkzoNobel - Supply Chain, Research & Development, Process Technology SRG, 7418 AJ Deventer, The Netherlands ∥ Inorganic Chemistry and Catalysis, Debye Institute for Nanomaterials Science, Utrecht University, Universteitsweg 99, 3584 CG Utrecht, The Netherlands ⊥ Institute for Complex Molecular Systems, Eindhoven University of Technology, P.O. Box 513, 5600 MB Eindhoven, The Netherlands ABSTRACT: In this Perspective, we highlight the main challenges to be addressed in the development of heterogeneous catalysts for the direct functionalization of methane. Along with our personal view on current developments in this field, we outline the main mechanistic, engineering, and catalyst design issues that have hampered implementation of new technologies and highlight possible paths to overcome these problems.

KEYWORDS: methane activation, methane valorization, heterogeneous catalysis, reaction mechanism, radical rebound mechanism, C−H bond activation, alkane functionalization

1. INTRODUCTION Fuels and chemical feedstocks are important commodities with a great direct impact on the development of society that currently relies on fossil fuels.1 Transportation (air, ground, and sea) and manufacture of goods, from petrochemical feedstocks to plastics and rubber industries, depend heavily on oil as a raw material.2 With the rapidly growing world population, the need to increase standards of living, and the dwindling world oil reserves per capita,3 a sustainable plan to couple an economically viable energy model with an environmentally friendly solution is the topic of many debates. Although more and more attention is given to research into renewables, the interest in improving current oil technologies is still pragmatically valid because most of the technologies applying renewable energies are in a very early stage of development, and it will be rather difficult to implement them within a realistic time frame.4 An essential prerequisite to implement greener technologies is the efficient use of nonrenewable sources of hydrocarbons. Naturally, this highlights the necessity to further develop efficient technologies for the valorization of methane. Indeed, © 2016 American Chemical Society

the enormous gas reserves found (208 trillion cubic feet proven),5 environmental sustainability, and lower overall costs point to natural gas as the primary source for energy and chemicals in the near future and to methane hydrates as the most important source of hydrocarbons in the long term. Hence, it is not surprising that methane valorization has been a hot topic over the last few decades, as highlighted in several excellent reviews.6−10 Natural gas is a mixture of gaseous hydrocarbons with varying quantities of nonhydrocarbons, which normally are considered impurities. Methane is the main component of natural gas, followed by a range of hydrocarbons such as propane and butane.11 It is also a byproduct from oil refining and chemical processing. It has potential value as a cleaner source of fossil energy and as raw material provided that it can be brought economically to the point of use.7 In this respect, it would be highly desirable to convert methane to a product Received: February 11, 2016 Revised: March 30, 2016 Published: March 31, 2016 2965

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis

products),9 the direct alternatives have proven difficult to control because of low yields, selectivity, and productivity.16 In addition to this, the engineering of syngas production is a highly developed and optimized technology, especially at large scales. Table 1 summarizes the major processes used for syngas production, that is, steam methane reforming10 (SMR), partial oxidation or oxy-reforming (can be coal gasification or from methane), and autothermal reforming (formally a combination of SMR and partial oxidation).21 Methane reforming with CO2, known as dry reforming, is a potential technology that has not reached enough efficiency to be applied at industrial scale. However, it is very attractive because it utilizes CO2 as an oxidant, thus using two of the major green house gases.22 Indirect routes via syngas have progressed substantially in terms of commercial development. One of the best-known synthesis gas processes is the methanol-to-gasoline (MTG) process developed by Mobil.28 Synthesis gas produced from natural gas is converted to methanol, followed by the selective conversion of methanol to yield a mixture of aromatic products that boil in the gasoline range. The methanol-to-hydrocarbons (MTH) process has gained industrial attention, and intense development efforts are ongoing. Several alternative processes are commercially available, and these include the Mobil Oil MTG process, the Tøpsoe integrated gasoline synthesis (TIGAS) process, and the Lurgi methanol to propylene (MTP) process.29 In the Norsk Hydro/UOP MTO process, primarily ethylene and propylene are produced as polymerization feedstocks.30 In a similar spirit, Shell has developed the Shell middle distillate synthesis (SMDS).5 In this process, syngas, produced from natural gas, undergoes two consecutive catalytic steps: (1) a high-molecular-weight wax is produced via Fischer−Tropsch synthesis, (2) up to 85% of the wax is hydrocracked to a middle distillate (gasoline and kerosene). The first commercial SMDS plants have started up during the past decade making use of remote natural gas fields as methane source for the reforming step (i.e., Sasol’s Oryx GTL project in Qatar, the SasolChevron Escravos GTL project in Nigeria and the Shell’s Pearl GTL project in Qatar). Analysis of the economics of these processes reveals that a majority of the capital investment is associated with synthesis gas generation.31 This hampers syngas production from remote and small sources of natural gas where, in most cases, it is just flared and is the main motivation for the search of new processes in which methane is initially activated and preferably converted to a valuable chemical in a single step. This challenge has prompted intense research into the development of homo- and heterogeneous catalysts for this reaction, avoiding in this way the generation of syngas and eventually allowing the use of such technologies at much smaller scales. The recent developments in the field of heterogeneous catalysis in this direction are the main topic of this perspective article. Although excellent reviews on methane activation have been published in the past few years,6,31−35 in this Perspective, we strongly focus on mechanistic and practical aspects that need to be considered in

(chemical or fuel) that could be easily transported. Although compressed natural gas is a feasible transportation fuel for truck and bus fleets, there are still doubts about the feasibility of running smaller vehicles on this technology.12 Moreover, bearing in mind that methane is, without any doubt, the main potential source of carbon for the synthesis of chemical commodities, its transformation into more useful products is of the utmost importance,8,9 and from the chemocatalytic point of view, the direct activation of methane is one of the remaining grand challenges.13,14 The challenge of methane activation is related to the high stability of this compound. Methane is a very stable and symmetrical molecule that does not possess any dipolar moment or functionality that would allow for directing chemical reactions. The activation of the methane C−H bond in the gas phase usually requires high temperature and leads mostly to radical reactions with intrinsic low selectivity.15,16 Because of these reasons, nowadays the industrial transformation of methane into useful chemicals and liquid fuels is only feasible via synthesis gas,10 a mixture of molecular hydrogen and carbon monoxide, that can be further transformed to methanol17 or to hydrocarbons under moderate reaction conditions (423−623 K and 10−100 bar) via the Fischer−Tropsch synthesis (Figure 1).18 Other important

Figure 1. Overview of the different routes for the valorization of methane.

processes based on syngas are the Haber−Bosch process for ammonia production19 and the oxo-process20 for higher chain aldehydes/alcohols production. Although indirect routes are chemically inelegant ways of converting methane (CH4 first has to be oxidized to CO and further reduced to the final desired

Table 1. Overview of the Principal Reactions in Syngas Production process

main reaction

ΔHro (kJ mol−1)

H2/CO ratio

temp (K) [P (atm)]

steam reforming

CH4 + H 2O ⇌ CO + 3H 2

206

3:1

1050−1250 (20−30 atm)

dry reforming

CH4 + CO2 ⇌ 2CO + 2H 2

247

1:1

>1000

oxy-reforming

CH4 + 0.5O2 ⇌ CO + 2H 2

−36

2:1

>1000 (1 atm)

2966

refs 10,23 22,24,25 26,27

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis

2.1. Activation of the C−H Bond. Most of the literature dedicated to methane activation focuses on understanding the mechanism of C−H bond activation, namely, the cleavage of the strongest and the least-polar carbon−hydrogen bond.42 Excellent examples by Shilov43 and by Copéret44 offer representative overviews of strategies in homogeneous and heterogeneous catalysis. For metal-based catalytic systems, Shilov classifies C−H activation into three main mechanisms, as described in the following sections. True Activation. This is one of the most widely encountered mechanisms in TM catalysis: the C−H bond cleavage is promoted through the direct interaction with the metal at the catalytic site, resulting in the formation of a σ M−C bond. This type of reactivity can be further classified according to the electronic nature of the interaction between the C−H bond and the catalyst (see Scheme 2).45,46 With respect to their relevance

catalyst design. We present a thorough analysis of mechanistic aspects of direct methane activation with emphasis on both C− H bond cleavage and oxidant activation (crucial and largely overlooked in the case of oxidative routes). We focus here mostly on mechanistic knowledge derived from homogeneous catalysis and enzyme chemistry. After this, we highlight the most striking developments in the field of direct activation of methane using heterogeneous catalysts. At the end of the Perspective, we share our opinion on future research directions for catalyst development.

2. MECHANISTIC CONSIDERATIONS FOR THE DEVELOPMENT OF DIRECT CH4 VALORIZATION ROUTES Methane functionalization under mild conditions has been long considered difficult, if not impossible, because of the nonpolar character of its C−H bonds and its low-lying highest occupied molecular orbital (HOMO) and high-lying lowest unoccupied molecular orbital (LUMO). These fundamental properties make methane rather inert toward most common organic chemistry strategies for functionalization. Activation through either nucleophilic or electrophilic attack has proven difficult to achieve and control. Alternative approaches involving partial oxidation have been intensively investigated since the 1960s.36−41 In principle, the ability to activate this rather unreactive C−H bond should enable numerous reactions for methane functionalization. However, the C−H bond activation, although a prerequisite, is not the only challenge in methane functionalization (see Scheme 1). In

Scheme 2. Schematic Representation of the “True” C−H Bond Activation Mechanisma

Scheme 1. Main Components of the Catalytic Cycle for Methane Functionalization

a

Descripton of schematic. (a) Substitution: the catalyst can be represented by Lewis acid−base pair or formally σ-bond metathesis; (b) Insertion: based on the electronic character of the transition state can be further classified as oxidative addition or electrophilic activation.

for catalytic methane functionalization, we have encountered three classes: oxidative addition, electrophilic activation,47 and Lewis acid/base.44 Conceptually, the latter has been used extensively to describe mostly heterogeneous systems. However, it is related to electrophilic activation and mechanistically is a σ-bond metathesis.48 Purely based on electronics, analyses of the M−C bonding schemes at the transition states of several complexes have shown that C−H bond activation can be classified as nucleophilic, ambiphilic, and electrophilic.49 For both, the insertion and the substitution mechanism, the type of reactivity is determined by the total charge transfer from the metal (occupied dπ orbital) to the C−H bond (empty σ* orbital) as well as the back-donation from C−H bond (occupied σ orbital) to the metal (empty dσ orbital). Thus, oxidative addition can be described as a purely nucleophilic activation (formally only donation from metal to C−H occurs).46 For active centers that are composed of electron-deficient metals, the active site can be described as a Lewis acid/base pair. The Lewis acidic metal site (M) polarizes the C−H bond to promote its heterolytic dissociation, yielding a σ-bonded M− CH3 species. The base site (X) accepts a proton and provides

fact, subsequent reactions, where the functionalization occurs, have received far less attention, in spite of the growing evidence that the activation of the oxidant represents a key challenge in closing the catalytic cycle. Thus, to achieve selective methane oxidation, the cleavage of the C−H bond has to be achieved, but equally important, the activation of the oxidant to form and regenerate the active site has to be compatible with the C−H bond activation step. Indeed, selectivity is an issue because the low reactivity of the C−H bond requires the use of either harsh reaction conditions or highly reactive reagents42−44 and this is the main paradigm in methane activation. 2967

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis

by the interaction with the surrounding basic ligands. Typically, oxygen atoms at the coordination sphere (or at the surface for heterogeneous catalysts) are responsible for the hydrogen atom abstraction (HAA)59,60 from methane, hence activating the C− H bond. This “fake” C−H bond activation mechanism results in the formation of “free” methyl radicals (Scheme 3).

substantial stabilization to the reaction products (Scheme 2a). In heterogeneous catalysis, such M−X acid−base pairs are commonly formed on oxide surfaces by undercoordinated Lewis acidic surface metal sites and neighboring basic oxygen centers. One of the most important examples of such mechanisms is the methane activation in oxidative coupling (OCM) by Li-doped MgO catalyst (Figure 2). Recent

Scheme 3. Schematic Representation of the “Fake” Activation Mechanism

Common oxidation reactions using bulk metal oxides usually react via nucleophilic oxygen leading to oxidative dehydrogenation of alkanes or deeper oxidation products via the Mars−van Krevelen mechanism (e.g., methane to formaldehyde).61 It has to be noted that extensive literature can be found about this reaction for production of propene, but this is beyond the scope of this perspective. Most of the examples on partial methane oxidation, including homogeneous, heterogeneous, and enzymatic catalytic systems, involve a fake activation mechanism with the aid of electrophilic oxygen.62 The metal at the active site can have variable oxidation state, and (although it can be argued) 3d transition metals are the most favorable for oxidation reactions via electrophilic oxygen active centers (vide infra). Due to the natural abundance of Fe-based enzymes for hydroxylation reactions, most of the research for characterization and toward understanding of the active species in catalysis has focused on FeIV-oxo cores. However, there is no fundamental reason that this reactivity cannot be encountered for other 3d late transition metals (i.e., Cu or Co). Based on molecular orbital theory, there are two possibilities in which an electrophile can approach methane and effectively carry out the HAA: the σ-channel and π-channel mechanism.63−66 In the first, the electrophilic orbital that participates in the H-abstraction (from a σCH) is an empty σ*(Fe-3dz2 O2pz) of the metal-oxo complex, while in the second case, an empty π*(Fe-3dxz/yz O-2px/y) promotes the reaction. Figure 3 shows the simplified frontier molecular orbital diagram and schematic representation of the electron-accepting orbitals of the high spin (HS) [FeO(H2O)5]2+ and the low spin (LS) [FeO(NH3)4(H2O)]2+ species, model compounds investigated by Baerends and Kazaryan.67 The LUMO of the HS compound (Figure 3a) is the dxy(β) orbital; however, this orbital cannot overlap with the HOMO of methane, because it is exclusively located on the iron center. The electron acceptor LUMO+1 is the σ*(α), orbital able to accept an α electron from methane, coupled antiferromagnetically to a methyl radical in the transition state. It was observed that the spin state is affected by ligand effects (more electron-donating ligands lead to low spin configurations), and both the spin state and the ligand field significantly influence the reactivity of the FeO2+ species and therefore the C−H bond activation mechanism. In the more active HS [FeO(H2O)5]2+ species, the σ-channel mechanism is energetically favored, whereas for the analogue low spin (LS) [FeO(NH3)4(H2O)]2+ species, a π-channel mechanism is preferred, where the π*(Fe-3dx/y O-2px/y) of the FeO2+ participates. The presence of the stronger-field NH3 ligands in the [FeO(NH3)4(H2O)]2+ species results in a low-spin ground state and effectively destabilizes or “pushes up” the

Figure 2. Reaction energy diagram for methane activation by a cluster model of MgO surface edge sites (black: Mg, red: oxygen). Reproduced with permission from ref 50. Copyright 2014 Wiley.

experimental and theoretical studies in this reaction revealed that the C−H bond activation takes place over the acid−base pairs involving Mg2+ and O2− ions on the edges and steps of the MgO surface.50 Additional examples on the Lewis acid−base pair mechanism have been demonstrated for a wide variety of heterogeneous catalysts based on main group elements, ranging from metal oxides (i.e., γ-Al2O351) to zeolite-based catalysts51 (Zn- and Gaexchanged high-silica zeolites52−54). Theory indicates that the formation of σ-type adsorption complexes55,56 between the Lewis acid site and CH4 facilitates the heterolytic C−H cleavage. The strength of the proton-accepting base part of the Lewis pair plays a key role in the heterolytic C−H bond activation process. As already anticipated above, oxidative addition (Scheme 2b) involves electron-rich transition-metal centers in low oxidation states (usually by late-transition-metal complexes) and C−H bond activation proceeds over a single transition-metal site. This reaction starts with the formation of a σ-complex with the coordinatively unsaturated metal center. The back-donation of two electrons from the metal d-orbitals to the σ* C−H orbital is the driving force for the C−H bond cleavage, yielding a hydride and CH3 anion fragment bound to the 2e− oxidized transition-metal species.33,57 In the third mechanism, the metal center has to be in a high oxidation state, electron-deficient, and a coordinatively unsaturated species, thus facilitating electrophilic activation.47 The organometallic M−C species is encountered only as a transient intermediate, and due to the high oxidation state of the metal center, these catalysts can withstand polar media such as water or strong acids. This stability enables the use of strong oxidants, contrary to species that undergo oxidative addition where only weak or no oxidants are applicable. Pioneering in this field, Shilov chemistry58 using PtIV salts is perhaps one of the most important breakthroughs in C−H bond activation. Importantly, a methylplatinium(IV) was observed supporting the Pt mediated C−H bond activation via the organometallic intermediate. Fake Activation. This type of activation occurs when the metal site is not accessible and methane activation is dominated 2968

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis

Figure 4. Schematic representation of orbital interactions involved in C−H bond activation by the reactive O−* radical in Cu-oxo clusters.

initio calculations to study alkane activation by an FeIV-oxobased MOF. It was found that the reactive sites in the ground state are dominated by the [FeIVO]2+ configuration in a quintet state (64%). This electronic configuration is however not reactive. The elongation of the Fe−O bond stabilizes a septet electronic configuration that corresponds to an antiferromagnetically coupled [FeIII−O*−]2+ radical species able to activate methane via the σ-channel mechanism.71,72 An interesting and puzzling observation was that such MOF materials promote the C−H bond cleavage in ethane and methane with barriers of 58 and 69 kJ mol−1, respectively, calculated by DFT. Such a small difference in activation energies would suggest that both reactions can potentially be carried out at near ambient conditions. However, while the cleavage of the C−H bond in ethane could be observed experimentally, the exposure of the iron-containing MOF catalyst to methane did not result in any reaction. Fenton-Type Mechanisms. Finally, when neither the metal nor its ligands are directly involved in the C−H bond activation, the reaction proceeds through a Fenton-type mechanism. The C−H bond in this case is activated by free radicals (see Scheme 4). A representative example can be found

Figure 3. Frontier Kohn−Sham orbitals of (a) HS [FeO(H2O)5]2+ S = 2 and (b) [FeO(NH3)4(H2O)]2+ S = 1 and the schematic representation of the electron-accepting molecular orbital. The alfa spin orbitals are to the left and the beta spin orbitals to the right.

LUMO dx2‑y2 as well as the higher-lying σ*(Fe-3dz2 O-pz), resulting in a π*(Fe-3dxz/yz O-2px/y) energetically more favorable for the electrophilic attack. The σ*(α) orbital is parallel to the Fe−O bond, and thus, in the transition state for maximal overlap, the Fe, O, and H atoms are aligned. In contrast, the electron-accepting orbital π*x/y(β) (LUMO in the LS [FeO(NH3)4(H2O)]2+) is perpendicular to the Fe−O bond, and therefore, the forming O−H bond is not aligned with the Fe−O bond in the transition state. Not surprising, the activation via the σ-channel mechanism for both HS complexes, [FeO(H2O)5]2+ and [FeO(NH3)4(H2O)]2+, proceeds with much lower barriers than via the π-channel. Nevertheless, for [FeO(NH3)4(H2O)]2+, the ground state is the LS, and the preferred mechanism becomes the π-channel. The groups of He, Neese, and Solomon have put forward an alternative proposal, where radical-anion character was postulated to stem from the reactive oxygen ligand and to be key for the efficient hydrogen atom abstraction (HAA) for methane (Scheme 3).65,68−70 This mechanistic concept has been supported by DFT studies on C−H activation by various metal-oxo clusters, including a promising heterogeneous methane oxofunctionalization system based on Cu-containing high-silica zeolite catalytic system (Figure 4).70 In a theoretical study, He et al. revealed a correlation between the spin density on the reactive oxygen center of small metal-oxo clusters and the barrier of C− H bond activation.69 Despite these leads, the role of the putative oxygen radical anion in the C−H bond activation mechanism is still under debate.67 Additional advanced theoretical analysis of the orbital interaction mechanism is necessary for the FeO2+ core as well as for other metalmediated HAA systems.16 Metal organic frameworks (MOFs) are providing excellent tools for the development of ideal “fake activation” catalysts and with it model systems to study the C−H bond activation mechanism. Gagliardi et al. applied high-level multireference ab

Scheme 4. Simplified Equations of the Fenton-Type Reaction Mechanism

in the works of Shulpin et al.,73−75 who applied transition-metal complexes as catalysts for the oxidation of methane as well as other alkanes. In these reactions, the oxidant source is H2O2 or O2, and the metal facilitates the radical decomposition of H2O2 to form reactive species capable of hydrogen abstraction from alkanes. The radical chain reaction is depicted on Scheme 3. In such a mechanism, the activation of methane by the OH radical is highly exothermic (ΔH ≅ −60 kJ mol−1) and proceeds with a barrier of only 15 kJ mol−1 to form a methyl radical and water.76 2.2. Activation of the Oxidant as a Key Step in Oxidative Direct Functionalization Routes. After the C−H cleavage step, the activation of the oxidant to form and regenerate an appropriate reactive site is crucial. This issue has been recently discussed in an excellent and comprehensive perspective by Conte and co-workers77 and in a review by Roduner, Laschat, and co-workers.78 An interested reader is 2969

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis invited to study those works for a deeper insight into O2 activation routes. Here we will only summarize and briefly discuss the main reaction pathways relevant to methane oxidation. The main difficulty of applying O2 directly for the selective oxidation of hydrocarbons is associated with the triplet ground state of the oxygen molecule. Wiegner’s selection rule states that the multiplicity of the system has to be preserved in the course of a chemical reaction (eq 1). Sr,1 ± Sr,2 ± ··· ± Sr, n = Sp,1 ± Sp,2 ± ··· ± Sp, k

Scheme 5. Simplified Reaction Scheme for the Autoxidation of Hydrocarbons in the Presence of a Radical Initiator

(1)

where Sr and Sp are the spin states of reactants and products, and n and k are the number of reactants and products, respectively. This implies that the reaction between a substrate with a singlet (CH4, S = 0) and a substrate with a triplet (O2, S = 1) electron configuration to form products in the singlet state (CH3OH + H2O, S = 0) is not allowed. Such a spin-forbidden process can only take place when a hydrocarbon reacts with O2 in one of its singlet-excited states (Figure 5) lying 157 (1Σg+) and 94 kJ mol−1 (1Δg) above the

of reactivity. The radical initiator creates the alkyl radical species (R·) by abstracting a hydrogen atom from the alkane (R-H). The subsequent reaction of R· with O2, practically barrierless, proceeds with preservation of the spin state and oxygen diffusion is the only limitation to the reaction rate.77 Alkane chain propagation is also a quite favorable reaction and shows a barrier of around 60 kJ mol−1.81 The radical chain can be easily terminated by radical recombination toward the diamagnetic oxidation products. A transition-metal center can facilitate the spin inversion of triplet oxygen. There are two possible mechanisms for spin inversion, namely, single electron transfer or via intersystem crossing by spin orbit coupling.82 Thus, in catalysis, a metal site in an appropriate spin state can react with the ground state triplet O2, resulting in a favorable active species for the selective oxo-functionalization of a singlet alkane. This reaction can result in different intermediate species that can be further classified depending on the total number of electrons transferred from the metal center to the oxygen molecule (Figure 6). The complete reduction requires four electrons (eq 2) and is highly influenced by the proton availability, and thus, the redox potential of these intermediate species depends heavily upon pH or effective acidity.83 One electron reduction of oxygen is often difficult because it possesses the highest barrier. This reduction is favored in protic media, and thus, the H atom abstraction is favored over the one-electron reduction.

Figure 5. Molecular orbital diagram of molecular oxygen in (a) triplet ground state (3Σg−), (b) singlet excited state with electrons on different orbitals (1Σg+), and (c) singlet excited state with electrons on the same orbitals (1Δg). Reproduced with permission from ref 77. Copyright 2015 Royal Society of Chemistry (RSC).

ground state. However, the respective excited-state reaction pathways are usually quite unselective, especially when taking place at elevated temperatures.77 There are different possibilities to overcome the spin forbidden limitation for oxidation reactions such as excitation to the singlet state (1Δg), autoxidation or, the most important concerning catalysis, the activation of (3Σg) O2 by the aid of a metal center. Excitation of the triplet to the singlet state via energy transfer from photoexited sensitizers79 can be used for photodegradation of organic compounds,80 but in general, singletstate reactivity is difficult to control because of the high reactivity of the singlet O2. Autoxidation reactions use the aid of an initiator to start a radical chain reaction (Scheme 5). The decomposition of (trace amount of) peroxides, either thermally (above 403 K) or catalyzed by metal species present in the reaction mixture can lead to uncontrolled and nonselective type

Figure 6. Reactions of transition metal with molecular O2. 2970

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis O2 + 4e̅ → 2O2 − (2) Based on metal-mediated O2 activation, there are two possible mechanisms for O2 activation: (1) heterolytic reduction (Figure 6c), where the four-electron-transfer path results in oxo-species Mn+2O and (2) homolytic reduction (Figure 6a,b), when the number of transferred electrons from the metal to oxygen is less than four, affording intermediate oxygen oxidation states (e.g., peroxo and superoxo anions) and hence radical intermediates. In principle, all the peroxo intermediate species are able to abstract a hydrogen atom from a C−H bond and homolytically cleaving it.62 These radical species are often pointed as undesirable because it can lead to uncontrolled nonselective radical reactions or even to autoxidation. The formation of Mn+2O species as oxidants, via the heterolytic reduction mechanism, is the preferred path. In this case, the stoichiometric problem84 of how to generate the MO species to favor the oxygen atom transfer reaction is important (see Scheme 6), and two different scenarios have to

Scheme 6. Concerted vs Stepwise Reactivity of both O-Atom Equivalents Figure 7. Catalytic cycle of O2 activation and CH4 hydroxylation in sMMO. The reaction pathway from MMOHred to MMOox was oversimplified to make emphasis on the mechanism for the regeneration of the MMOHox into the MMOHred. For more details, see refs 91−94.

the cooperation of the metal active center with functional groups at the protein matrix and/or reactions with cofactors such as NAD(P)H and FADH2. There are alternatives in enzymatic systems to conserve energy via flavin-based electron bifurcation. In systems where the electron flow is not thermodynamically favored, the flavin-based electron bifurcation mechanisms is often used by acetogens as means of coupling non- favored (endergonic) reaction with other exergonic processes.89,90 In particular, the catalytic cycle for the sMMO enzyme involves NADH (coupled to FAD as transfer agent) as source of protons and electrons. The three main parts of the sMMO are the MMOH where catalysis takes place with the substrate and O2,92 and two other important components, namely, the FAD-based reductase MMOR and the mediator MMOB.91 Although the MMOB plays the intermediary role for MMOR and MMOH to interact, MMOR is important to balance the 2H+ and 2e− and to manage to couple the NADH with the reduction of MMOHox.94 This insight into the oxidation processes in biological systems has inspired chemists to construct synthetic schemes following the enzymatic approach but using alternative lowercost stoichiometric reducing agents to replace NAD(P)H.93 The generation of electrons and protons within the catalytic cycle can be accomplished by using a combination of iron or zinc powder with carboxylic acids in the well-studied Gif systems.95 Ultimately, the use of molecular H2 as the stoichiometric reducing agent is desired. Such a selective oxidation one-pot system has been reported by Otsuka and coworkers.38 An alternative catalyst design involves the spatial separation of the substrate oxidation and the active-site regeneration steps. The selective oxidation reaction in such a scheme requires the use of preformed reactive oxygen species, which after the

be considered: (1) producing 1 equiv of MO per O2 molecule (having to scarify the other oxygen as, i.e., H2O) or (2) generating two neighboring MO species. Naturally, the second option is more attractive; however, the steric control for partial oxidation plays a key role. In dinuclear homogeneous complexes, as well as in surface reactions, the reactivity of the “first” MO species is intrinsically affected by the “second” provided that the oxygen atom transfer proceeds stepwise. More desirable is to have a complete dissociation of the two MO equivalents (formally active site isolation); however, this is the most unlikely scenario in terms of catalyst design. In spite of the usually claimed exquisite chemistry of natural systems, enzymes produce stoichiometric amounts of water during O2 activation, demonstrating that scenario 1 is probably the most realistic approach. The O2 activation mechanism represents a common strategy employed in nature and synthetic chemistry in selective oxidation reactions. In the soluble methane monooxygenase enzymes (sMMO), the reduced metals at the active site (MMOHred) donates four electrons to the adsorbed O2 molecule, resulting in its complete reduction to O2− species (intermediate Q, Figure 7).85 In the particulate methane monooxygenase (pMMO) enzymes, the accurate determination of the active site and reaction mechanism is still awaited; however, DFT calculations predict complete reduction of O2,86−88 which in turn acts as proton-accepting site upon C−H bond activation. However, to establish a true catalytic cycle (i.e., to promote the formation of the oxidation products and the regeneration of the initial active site), the assistance from other components of the enzymatic system is required. These include 2971

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis completion of the oxidation step has to be separately regenerated. Most commonly, H2O2, O3, N2O, and tert-butyl hydroperoxide are used as reactive oxygen donors in these schemes. However, other substances like HNO3, H2SO4, NaClO, and NaClO2 have also been applied.77 A very important and well-known example of this class of selective oxidation systems is the homogeneous Periana “Catalytica” system. On the basis of the studies of electrophilic reactions with alkenes from Shilov,58 Bercaw, and Labinger,96 Periana first employed HgII as a soft powerful electrophile able to oxidize methane to methyl bisulfate in concentrated sulfuric acid.97 Although it was a huge step forward, using Hg has the disadvantage that this metal cannot be easily modified by ligands. Thus, the effort was focused on stabilizing Pt species through different ligands in concentrated sulfuric acid. Few ligands withstand the oxidizing and acidic conditions,98,99 and remarkably, 2,2′-dipyrimidine binds Pt so effectively that it not only survives the reaction media but also stabilizes PtII and avoids oxidation to PtIV.100 In both cases, the selectivity is remarkably only toward methane because the oxidized product is protected by the bisulfate. It was found that weakly basic counteranions facilitate the reaction and that electron-deficient C−H bonds are less likely to react. In this system, H2SO4 is used as the oxidant but is also the key to stabilize the metal center. Upon reaction with methane, H2SO4 is reduced to SO2 and has to be separately reoxidized to regenerate sulfuric acid (Figures 8 and 9).101

Figure 9. Pt-based Catalytica system with sulfuric acid to help to close the catalytic cycle.

of its environmental attractiveness, the industrial applicability of H2O2 for methane oxo-functionalization is still limited mainly because of the relatively high price of the oxidant compared to that of the target product. Nevertheless, from the academic perspective, the use of H2O2 as well as other alternative oxidants110−112 can provide crucial mechanistic insights into the different steps of the methane oxidation reaction. In particular, previous studies identified two main H2O2 activation paths relevant to methane oxidation: (a) Fe2+-Fe3+/H2O2 systems, where the formation of hydroxyl and hydroperoxy radicals and thus, the homolytic reduction of oxygen is observed; and (b) two-electron Fe3+-Fe5+ mechanism, where formally the O−O bond in the peroxide is broken, and thus, a heterolytic mechanism is operational.113 It is important to note that the discrimination between the different mechanisms of O2 activation represents a great challenge in view of the high complexity of the catalytic systems and the difficulties associated with obtaining relevant experimental information regarding the state of active sites formed under the catalytic conditions. Modern computational modeling techniques can provide crucial mechanistic insights and reveal the fundamental factors controlling the catalyst performance at each of the steps of the selective methane oxofunctionalization process.50,54,67,69,71,114,115 To summarize the above mechanistic considerations, it is clear that whereas a wide variety of systems can promote the C−H bond activation, the formation of the active sites and their regeneration within a single catalytic process using molecular O2 as the oxidant represent the key challenge toward a practical direct route, specially when mild reaction conditions are desired. All the successful catalyst systems employ in one way or another O2 shuttle strategies, in which the C−H bond and O2 activation steps are mechanistically separated. Interestingly, nature does not provide exception to this rule. It also does not use directly O2 as the actual oxidant for methane activation, but it employs more complicated multistep schemes involving the use of stoichiometric reductants to form and regenerate the active sites within the catalytic cycles.

Figure 8. Mercury catalysis to activate methane and produce methyl bisulfate in sulfuric acid. The reaction is carried taking advantage of the soft character of HgII (i.e., good oxidizing properties) and by an electrophilic displacement that manages to activate the C−H bond.47,97

A conceptually similar reaction scheme creates a basis for heterogeneous catalyst systems employing Cu-modified zeolites. In these systems, catalytic methane oxidation is established in a three-step process involving sequential calcination, methane activation, and methanol product desorption steps.102−104 The exchangeable Cu sites inside the zeolite micropores play the role of O2 shuttle, similar to that of H2SO4 in the Periana system. The use of H2O2 as the oxidant conceptually bridges the above-mentioned approaches. On one hand, the possibility to directly synthesize hydrogen peroxide through a low-temperature catalytic oxidation of H2105,106 renders the latter as the O2 shuttle that can be used to establish a truly catalytic system for the selective oxidation of methane to methanol.107−109 In spite 2972

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis

3. PROGRESS IN THE DEVELOPMENT OF HETEROGENEOUS CATALYSTS FOR THE DIRECT ACTIVATION OF METHANE 3.1. High-Temperature Routes (>473 K). Oxidative Routes. Oxidative Coupling of Methane (OCM). The oxidative coupling of methane (OCM) (discovered in the early 80s but not industrially applied yet) consists of the direct coupling of two methyl radicals in the gas phase.40 The main products of this reaction (in addition to CO2 and CO) are ethane and ethene (eqs 3 and 4).116 Several metal oxides have been up to now tested as catalysts and although the pioneering work of Keller and Bhasin117 (MnOx/Al2O3 as catalyst) and Hinsen and Baerns118 provided the basis for this reaction, the work of Ito and Lunsford119 marks an important step in the OCM field. They tested a Li/MgO catalysts that afforded up to 27% methane conversion with 50% C2 selectivity at temperatures ∼973 K.

the formation and coupling of two methyl radicals will happen only in the presence of O2, which is required to compensate the endothermic process of the methyl radical desorption.50,124Among the important factors involved in selectivity are the acidity of the metal oxide support and promoters for the C−H activation such as chlorine. In fact, it was found that acid supports are essential to achieve high selectivity toward ethylene.125 On these lines, most of the catalysts developed for this reaction are based on metal oxides. Labinger and co-workers,96 based on the argument that radical abstraction is insensitive to the nature of the hydrocarbon molecule, clearly demonstrated that, in OCM, selectivity will always decrease upon increasing conversion. Indeed, experimental results show that maximum productivity can only be obtained at 30−40% conversion with selectivities below 80%.121 In spite of these clear issues, recent developments by Siluria (http://siluria.com/) seem to bring this technology closer to application. Direct Conversion to Oxygenates. High- and mediumtemperature catalytic selective oxidations constitute a technological challenging class of processes. The high exothermicity of the reaction and complex kinetics scheme leads to byproducts and relatively low selectivities.126,127 Clearly, a direct conversion of methane to methanol is energetically the most efficient manner to convert methane to a valuable liquid that can be very easily integrated in existing chemocatalytic routes for the synthesis of higher alkenes. The kinetics and mechanism of methane oxidation have been studied for over a century, and (full) combustion products have been identified since the beginning of last century. Methanol formation is promoted at high pressures, whereas H2, CO, CO2, formaldehyde, and formic acid are observed when the reaction is carried out at atmospheric pressure.37 The direct conversion of methane to C1 oxygenates has been reported under catalytic and noncatalytic conditions, using methane-rich mixtures in order to limit the oxidation reaction and to avoid explosive conditions. Usually only good selectivities are achieved at low conversions. Noncatalytic gas-phase reactions are able to achieve selectivity to methanol up to 83% at conversion of methane of 12% at 50 atm (5% air) and 603 K.128 Although these results might seem promising, the required high residence time hampers industrial implementation. This reaction is also catalyzed by MoO3- and V2O5-type oxides at temperatures ranging from 623 to 773 K.129 Considerable efforts have been devoted to developing active and selective catalysts; however, the mechanism has proven difficult to clarify, and the low methanol yields seem to correlate with the poor selectivity of the O-insertion into the C−H bond after the first Habstraction.61 V oxide catalysts supported on silica were identified as one of the most active and selective catalysts. One of the key factors for the remarkable performance is the dispersion of V on the silica support to promote isolated oxide species.130,131 Several spectroscopic studies have been devoted to elucidate the molecular bond directly involved in the catalytic oxidation reaction. In situ Raman spectroscopy studies under reaction conditions demonstrated the inactivity of VO species for the activation of methane and suggest that V+5 species on the silica surface (V−O−Si) are key to activate and selectively convert methane to formaldehyde132 The role of this V−O−Si species can be indirectly assessed by changing the support composition. In this way, the support was proven to be critical for the catalytic performance, which suggests that the oxygen in the V−O-support is highly important. The nature of

2CH4 + 1/2O2 ⇌ C2H6 + H 2O ΔH1073 K = − 1465 kJ mol −1 (3)

2CH4 + O2 ⇌ C2H4 + 2H 2O ΔH1073 K = − 1163 kJ mol −1 (4)

Due to the high temperature and the radical nature of this process, achieving yields higher than 20% (with selectivities of ca. 80%) in single pass operation has proved challenging.120,121 On the other hand, because of the exothermic nature of the reaction, formation of hot spots are of great concern for actual implementation. The formation of methyl radicals in the gas phase with Li/ MgO catalysts has been very recently identified by Synchrotron VUV photoionization mass spectroscopy, and it was confirmed that reaction was initiated by those radicals.122 As briefly touched upon above, C−H bond activation takes place over acid−base pairs involving Mg2+ and O2− ions on the edges and steps of the MgO surface in an heterolytic fashion (see scheme s-bond).50 More recent studies have demonstrated that Li promotes the formation of low-coordinated O2− ions at the edges, corners, and kinks of MgO.50,123 Thus, the activity of the catalyst has been attributed to morphological features of the catalyst.50 Indeed, a combination of catalytic performance and TEM revealed multiple (100) surface steps in a very active fresh MgO catalyst, contrary to the poorly active MgO catalyst showing microsteps along the (111) facets after more than 200 h on stream (see Figure 10).50 In agreement with mechanistic studies on Li/MgO and Sr/MgO,124 and the identification of the methyl radicals in the gas phase,122 calculations show that

Figure 10. TEM images of pure MgO after 6 h (left) and after reaching steady state (230 h on stream).Reproduced with permission from ref 50. Copyright 2014 Wiley. 2973

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis CH3Cl → 1/nCnH 2n + HCl

this V−O-support moiety was further elucidated by X-ray absorption measurements (XAS). The sensitivity of X-ray absoption fine structure (XAFS) spectroscopy to the electronic density of vanadium demonstrated that the electron density of V−O-support species directly depends on the ionicity of the support.133 Recently, IR and in situ Raman spectroscopy brought more chemical insight into the nature of the active species. This study was based on the effect of water on selective oxidation of methane to formaldehyde and suggested that the most active species are hydroxylated monomeric V species anchored to the support with two Si−O−V bonds.134 For Mo-based catalysts, mononuclear molybdates were determined as the species with the highest specific activity.135 In situ XANES and Raman spectroscopy shows that these Mo species are pentacoordinated MoVIOx with a terminal MoO group..136,137 In these materials, the catalytic cycle would involve the formation of reduced MoIV species that react with O2 forming a peroxide specie, which unfortunately was not spectroscopically observed.136 Halogenation and Oxyhalogenation. Halogenation of alkanes (eqs 5 and 6) is a well-known reaction that, in the absence of a catalyst, yields a mixture of mono- and multihalogenated products. In principle, this reaction can be carried out by F2, Cl2, Br2, and I2. The reaction with F2 is extremely undesirable due to the corrosive nature of fluorine (the side product is HF). At the same time, I2, because of its high polarizability, produces a rather unstable CH3I product that can easily decompose back to I2. Thus, of major interest has been the reaction with Cl2 and with Br2.

Nonoxidative Routes. Reductive Coupling to Acetylene and Other Higher Hydrocarbons. Noncatalytic methane pyrolysis at high temperature affords acetylene, ethylene, benzene and H2. The chemistry of this reaction is wellknown and strongly affected by equilibrium limitations. This together with the fact that, at such high temperatures, process engineering is challenging, have been the main troubles found by the noncatalytic reductive coupling. The thermal pyrolysis of methane at 1173−1273 K) produces a mixture of C2 and aromatics. Short reaction times, low partial pressures and dilution of the gas feed by recirculating the H2 produced have been strategies that attracted attention for a possible industrial implementation.7 Thus, efforts to improve this technology have stimulated research not only to improve the engineering but also to seek a catalytic process. Dissociative chemisorption of methane on metallic surfaces is a well-known mechanism that many heterogeneous reactions share. Activation of C−H on metal surfaces in fact has been well documented and, since the pioneering studies of the groups of Amariglio141 and van Santen,142,143 it was shown that the homologation144 of methane (formation of C−C bonds) over a metallic surfaces occurs via chemisorbed CHx species that are able to dimmerize (polymerize or chain growth, eq 9) through a mechanism that may be shared with other reactions such as FTS.142 CH4 + M → CHx ··· M +

x → alkanes + H 2 + M 2H 2 (9)

CH4 + Cl 2 ⇌ CH3Cl + HCl ΔH298 K = −417 kJ mol −1

Due to the nature of the chemisorbed species, fast formation of coke is the main issue. It is thought that from the carbonaceous deposits, Cα(carbidic) are responsible for the homologation reaction, whereas the Cγ(graphitic) and Cβ(amorphous) carbon lead to deactivation by coke formation. C−H activation can be achieved by metals on different supports (zeolites among the most important) to form adsorbed CHx. The metals that have been investigated are first-row early transition metals such as Mo, Fe, V, Cr, although van Santen also demonstrate the activity of late transition metals (Ru, Co, and Rh over silica) as catalyst for the homologation of methane with olefins. In this reaction, a two-step homologation was proposed:145 (1) constant removal of H2 as a strategy to achieve full coverage of CHx species onto the surface and (2) consequently hydrogenation at a somehow lower temperature, avoiding the hydrogenolysis. Despite the vast knowledge of the reaction, the intrinsic low efficiency represents the major problem that avoids the commercialization of this technology. Methane Aromatization. Thermodynamically, it is more favorable to produce aromatics from a nonoxidative reaction than to produce olefins;146 however, the formation of coke is even more favored under these conditions.147 Nevertheless, it has been demonstrated that, using an appropriate catalyst, methane can be converted to benzene. The reaction coproduces large amounts of H2. A variety of catalysts have been reported; however, by far, the best-reported continues to be Mo on ZSM-5.148−150 The success of this catalyst resides in its bifunctional nature. Mo sites are responsible for the activation of C−H via the formation of carbides and the acidic sites of the zeolite are believed to promote the formation of aromatics.151 Although this mechanism has not been fully elucidated, it is thought that Mo-carbide active sites produce the C2 species, which can further react at the Brönsted acid

(5) CH4 + Br2 ⇌ CH3Br + HBr ΔH298 K = −117.0 kJ mol −1 (6)

These reactions were found to be controlled toward the monohalogenation of methane in the presence of various acidic metal oxides in different supports by Olah et al.138 At temperatures as low as 453−523 K, selectivities to the monohalogenated product above 90% at conversions up to 58% have been achieved. A major drawback of the halogenation strategy is the regeneration of the gas (Cl2, Br2), which increases the operational costs. In principle, this is not a reaction for the direct conversion of methane to oxygenates; however, it opens the possibility to integrate into one step the halogenation with the hydrolysis of the intermediates to afford directly the oxygenated products. Last decade, a multifunctional catalyst (MoO2Br2(H2O)2 supported in a zinc-modified MCM48 zeolite was reported to integrate all three steps in a one single vessel.139 The main highlight of this example is the ability of the system to use the Br from the catalyst as oxidant carrier, and at the same time being able to regenerate the Br into the catalyst via a metathesis mechanism of the halogenated methane with the metal oxide pair. Thus, although this is a reaction that employs halogenation intermediates, addition of Br2 gas in not necessary, but instead the oxidant source is O2. Oxyhalogenation is an alternative for the halogenation of methane without the need to use the halogen gas (thus avoiding regeneration), and it has been shown that this principle is also applicable to produce propene in a two-step reaction (eqs 7 and 8).140 CH4 + HCl + 0.5O2 → CH3Cl + H 2O

(8)

(7) 2974

DOI: 10.1021/acscatal.6b00428 ACS Catal. 2016, 6, 2965−2981

Perspective

ACS Catalysis sites. Despite several attempts to find more active catalysts than the Mo/ZSM-5, only deactivation has been slightly mitigated by using zeolites with bigger pore sizes and somehow similar acidic properties as ZSM-5 (MCM-22, MCM-49).152−154 This was however at the cost of a worse Mo dispersion, thus leading to comparable activity. Other strategies to overcome carbon deposition involve the use of additional metals as promoters such as Cu,155 Ru,156 or K.157 Co-feeding of mild oxidants to the feed has also been employed to mitigate coke formation.146 A great deal of effort has been devoted to understanding the synthesis of Mo/ZSM-5 catalysts and speciation and role of Mo on the catalytic performance by a battery of analytical tools. Ma et al. investigated the preparation of Mo/ZSM-5 catalyst by an impregnation158 by means of NH3-TPD, 1H MAS NMR, and ESR spectroscopy. The results of this particular study suggest that during catalyst calcination, MoO3 clusters on the zeolite surface are able to migrate into the structure micropores and disperse into mononuclear Mo species, replacing the zeolite Brønsted acid sites. They proposed that Mo atoms both on the external surface of the zeolite and inside the micropores in close contact to framework Al sites are key for the aromatization of methane. Later, Zheng et al. used ultrahigh filed 95 Mo NMR spectroscopy and reported direct spectroscopic evidence of the formation of these Mo oxide species in close contact to framework Al sites.159 A further combination of solid-state NMR experiments and catalytic performance data lead to the conclusion that these Mo species are the active species for CH4 aromatization. Under reaction conditions, Mo chemistry is very dynamic and those changes are key to understanding the performance of the material. First, there is an induction period where Mo oxide species are transformed into an active phase. Lunsford et al. and Solymosi et al. performed XPS studies and found that during the induction period, CH4 transforms progressively from Mo6+ oxide species to Mo2C.148,149,151 They tentatively proposed Mo2C as active species for the formation of C2H4, which can subsequently be converted into C6H6 on the zeolite Brønsted acid sites. In situ Mo K-edge XAFS studies by Liu et al. confirmed the formation of Mo2C during the reaction with methane.160 However, the exact nature of the Mo active species was not yet identified. Very recently, Gao et al. reported a multipronged approach using density functional theory (DFT), in situ UV−vis diffuse reflectance spectroscopy, in situ IR spectroscopy, and operando Raman spectroscopy to determine the nature and catalytic activity of the different Mo species.161 They found that Mo oxide monomers anchored to double or single Al atoms in the zeolite framework were giving more active Mo-carbide structures than those anchored to Si on the external surface of the of the zeolite. The authors give a new vista for rational design of catalyst and optimization of reaction and regeneration conditions by limiting the anchoring of Mo oxide species to framework Al sites. Nonoxidative Coupling. In a recent example, with a Fe based system, Bao and co-workers have shown the possibility to activate methane under nonoxidative conditions, with high conversions (up to 48% at 1363 K) and selectivity toward ethylene of 52%.162 The outstanding performance of this system is thought to be a consequence of the isolation of monatomic iron sites, leading to high methane conversion without further formation of CO2 and/or coke. The formation of methyl radicals under reaction conditions is proposed, and these species are then dissociated to the gas phase to further

react and form the C2+ and aromatic products. EXAFS experiments and DFT studies suggest the active site to be a Fe species coordinated to a Si atom and two carbons (Figure 11).

Figure 11. Structural features of 0.5%Fe@SiO2. STEM-HAADF image of the catalyst after reaction, with the inset showing the computational model of the single iron atom bonded to two C atoms and one Si atom within silica matrix. Reproduced with permission from ref 162. Copyright 2014 American Association for the Advancement of Science (AAAS).

This active site is confined in silica lattices and is responsible for the formation of the methyl radical species. Although application of this catalytic system requires high temperatures, the high stability under this thermal conditions on stream (up to 60 h) might enable an interesting area of research. High-Temperature Activation of Methane Using Nonconventional Energy Sources. Among the proposed strategies to activate methane, dielectric barrier discharges have demonstrated to convert green house gases into alkanes, alkenes, and oxygenates.163 In these systems, it is claimed that high methane conversions can be achieved. However, more fundamental studies are necessary in order to elucidate the mechanism of activation and to evaluate the applicability of this technology. More recent examples include the use of deep UV light,164,165 plasma,163 and microwave irradiation.166 Although quite interesting results have been reported, especially in the case of deep UV activation, understanding of these technologies is still in its infancy. 3.2. Low-Temperature Routes (