Metal Oxide Catalysts in Relation to Environmental Protection and Energy Conversion

Abstract

Metal oxide catalysts have played important roles in the removal of harmful gases and the effective utilization of methane. For the CO/NO removal using metal oxide catalysts, the lattice oxygens and oxygen defects in the metal oxide often act as active sites. Therefore, the development of highly active metal oxide catalysts requires rational designs of the packing structure of O²⁻ ions and the redox ability of constituent metals. As part of such efforts, oxygen storage materials that rapidly store and release oxygen have been studied extensively, and such materials are applied as catalyst materials. In addition, for the formation of C₂ hydrocarbons by oxidative coupling of methane, although many supported catalysts have been continually investigated, metal oxide catalysts containing multiple metals with unique coordination structures are being recently developed.

10.1 General Background for Environmental Protection and Energy Conversion

Global warming is becoming a serious problem, leading to an urgent need for efficient processes for the removal of harmful gases, as well as the development of renewable energy resources. The development of highly functional catalyst materials is therefore highly needed [1, 2]. Furthermore, from the viewpoint of resources and economy of platinum group metals (PGMs) such as Rh and Pd, it is also desired to develop catalyst materials that do not contain PGMs or reduce the use of PGMs minimized as much as possible [3,4,5,6]. Nowadays, base metal catalysts have drawn attention with researchers focusing on the development of metal oxides that exhibit excellent redox properties. This chapter introduces metal oxide catalysts that contribute to environmental protection and the effective utilization of methane. In particular, the chemical characteristics of CeO₂-based oxides with a fluorite-type structure and perovskite-type oxides are outlined from the perspective of the packing structure of O²⁻ ions. In addition, oxygen storage materials, which are applicable to various fields involving catalytic reactions, are introduced from the viewpoint of the structural changes associated with the insertion and extraction of O²⁻ ions. Finally, oxidative coupling on metal oxides, which contributes to the effective utilization of methane, is introduced.

10.2 CO and VOC Oxidations

10.2.1 Overview

Catalytic combustion is effective for removing CO and volatile organic compounds (VOCs). This process uses a catalyst material that purifies harmful gases to CO₂ and H₂O at temperatures typically below 300 °C. Such a procedure does not induce flame combustion and is much safer than simple combustion processes that ignite with a flame. The process is also advantageous for environmental protection because of the suppressed evolution of incomplete combustion products and thermal NO_x, which are formed by exposure to N₂ and O₂ in air at high temperatures. The use of PGMs is indispensable for improving catalytic activity in the low-temperature range [7], whereas many studies have been conducted on metal oxide catalysts without PGM species for catalytic combustion. Highly active metal oxide catalysts and their crystal structures have been the subject of many publications [8], including CeO₂-based oxides, perovskite-type oxides, and spinel-type oxides. In general, all catalyst materials utilize metal oxides containing elements with excellent redox properties.

This section begins by presenting the relationship between the redox ability of metal oxides and the reaction mechanisms, followed by the structural characteristics of each metal oxide for typical catalyst materials.

10.2.2 Reaction Mechanism (Mars-Van Krevelen Mechanism)

In the catalytic combustion on a PGM catalyst, adsorption of reaction substrates and activation of O₂ molecules are known to occur on the surface of PGM. That is, the catalytic reaction proceeds according to a reaction based on the Langmuir–Hinshelwood mechanism (or Eley–Rideal mechanism) (Fig. 10.1) [9]. Therefore, when designing supported PGM catalysts, the high dispersion of PGM particles and the control of the electronic state by forming PGM alloy nanoparticles should be studied [10, 11]. In contrast, regarding metal oxide catalysts, the combustion mechanism involving lattice oxygen based on the Mars-van Krevelen (MvK) mechanism has become common [12,13,14]. Here, CO oxidation (Eqs. 10.1 and 10.2) is presented as an example to explain this reaction mechanism. First, CO reacts with lattice oxygen (O_L) at the oxide surface to generate CO₂ followed by an O₂ molecule filling the generated oxygen vacancies (V_O).

$$ {\text{CO}} + {\text{O}}_{{\text{L}}} \to {\text{CO}}_{2} + {\text{V}}_{{\text{O}}} $$

(10.1)

$$ {\text{O}}_{{2}} + {\text{V}}_{{\text{O}}} \to {\text{2O}}_{{\text{L}}} $$

(10.2)

Fig. 10.1

CO oxidation based on a Langmuir–Hinshelwood and b Mars-van Krevelen mechanisms

The transfer of lattice oxygen in the metal oxide is a key step in this reaction mechanism. It is therefore preferable to optimize the redox properties of the constituent metal species when designing metal oxide catalysts.

10.2.3 CeO₂-Based Materials

Ce⁴⁺ ions contained in CeO₂ are easily reduced to Ce³⁺ and this redox plays a crucial role in the use of CeO₂ as a catalyst [15]. CeO₂ crystallizes in a fluorite-type structure in which Ce⁴⁺ ions form a face-centered cubic lattice, with O²⁻ ions occupying the tetrahedral voids (Fig. 10.2). In other words, the O²⁻ ions compose a primitive cubic structure, with Ce⁴⁺ ions located at the body-centered sites. In the case of transition-metal oxides such as NaCl-type MnO and spinel-type Co₃O₄, the ionic radius of transition metals such as Fe³⁺ (0.55 Å) and Co³⁺ (0.545 Å) is much smaller than that of the O²⁻ ion (1.40 Å). This difference in radius results in a close-packed structure composed of the O²⁻ ions (Fig. 10.3) where the transition metal ion occupies the tetrahedral or octahedral voids formed by the O²⁻ ions. For CeO₂, on the other hand, the size of the Ce⁴⁺ ions (0.97 Å) is larger than that of the transition metal ions, indicating that the O²⁻ ions do not form a close-packed structure and prefer a primitive cubic structure instead. This characteristic packing structure of the O²⁻ ions facilitates their fast diffusion in the CeO₂ crystal lattice, even though their diffusion depends on the defect chemistry as well [16]. Thus, the redox property of cerium and the transfer ability of O²⁻ ions in CeO₂ are promising features for the above-mentioned MvK mechanism.

Fig. 10.2

Crystal structure of CeO₂ with a fluorite structure. a Positions of Ce ions and b O²⁻ ions in the unit cell. c Structure constructed by Ce and O²⁻ ions. d Crystal structure providing a clear view of the coordination environment of the Ce ions. Panels (a–c) are shown based on actual ionic radii: yellow balls, Ce⁴⁺ ions; red balls, O²⁻ ions. The schematic crystal structures in this chapter are drawn by VESTA program (K. Momma and F. Izumi. J. Appl. crystallogr., 2011, 44, 1272)

Fig. 10.3

Crystal structure of Co₃O₄ with a spinel structure. a Positions of Co ions and b O²⁻ ions in the unit cell. c Structure constructed by Co and O²⁻ ions. d Crystal structure providing a clear view of the coordination environment of the Co ions. Panels (a–c) are shown based on actual ionic radii: blue balls, Co²⁺ ions; green balls, Co³⁺ ions; red balls, O²⁻ ions

Many studies on combustion catalysts using CeO₂ exist, analyzing the correlation between the oxygen release property and combustion activity and reporting that the reactivity of lattice oxygen in CeO₂ was drastically improved by PGM loading [17,18,19]. CeO₂ is often combined with CuO or Co₃O₄ as a PGM-free catalyst [20,21,22,23]; the formed composites exhibit high mobility of lattice oxygen, resulting in high catalytic activity for VOCs combustion. Most of these reaction behaviors are explained by the MvK mechanism. Meanwhile, the reactivity of lattice oxygen in CeO₂ is closely related to the exposed crystal plane. Theoretical studies indicate that the (110) and (100) planes have high catalytic activity, emphasizing the importance of morphology control of the CeO₂ nanoparticles [24, 25]. In fact, rod-shaped CeO₂ particles, in which (110) planes are preferentially exposed, exhibit high VOC combustion activity. Furthermore, it has been demonstrated that the catalytic oxidation of CO with rod-shaped CeO₂ proceeds at lower temperatures than with cubic or octahedrally shaped CeO₂ grains.

10.2.4 Perovskite Oxides

Since 3d transition metals can adopt a wide range of valence states, transition metal-based oxide catalysts have been extensively studied for combustion reactions from the perspective of the redox ability of their metal species. Many reports exist on perovskite oxides, which are generally represented by the chemical formula ABO₃ [6, 26,27,28]. The A site of the perovskite structure is often composed of alkali metals, alkaline-earth metals, or rare-earth metals, whereas the B site contains transition metals. Unlike the close-packed structure of the O²⁻ ions in NaCl-type and spinel-type transition metal oxides described in the previous section, the A-site cation and O²⁻ ion form a close-packed structure. This is due to the ionic radius of the A-site cation in the perovskite structure being similar to that of the O²⁻ ion (Fig. 10.4). The B-site metal occupies the octahedral interstice consisting of O²⁻ ions. The ideal perovskite structure belongs to the cubic system, whereas many perovskite oxides are known to exhibit tetragonal or orthorhombic distortion according to the ionic radii of the metals occupying the A and B sites.

Fig. 10.4

Crystal structure of ABO₃ with a perovskite structure. a Positions of B ions and b A ions, and O²⁻ ions in the unit cell. c Crystal structure with a clear view of the coordination environment of the B ion. d Relationship between the A and B ions. Panels (a, b, and d) are shown based on actual ionic radii: Purple balls, A ions (alkali or alkali-earth elements); brown balls, B ions (transition metal elements); red balls, O²⁻ ions

An important advantage of perovskite oxides from the viewpoint of catalyst material design is that the constituent elements of the A and B sites can be chemically substituted. For example, partial substitutions of the A site can adjust the valence state of the B-site element or promote the formation of oxygen vacancies according to the charge balance. The substitution of Sr²⁺ for La³⁺ in La³⁺Co³⁺O₃ is expected to induce oxygen defects and control the Co valence state. In addition, partial substitutions of the B site can modulate the redox capacity of the B-site cation and the mobility of the O²⁻ ions. The possibility of adjusting these properties through the choice of dopants is a critical factor for efficient catalytic reactions based on the MvK mechanism.

Among the La-based transition metal oxides, LaM³⁺O₃, LaMnO₃, and LaCoO₃ are known to be active in CO oxidation [27]. Furthermore, the Sr-for-La-substituted material, La_1-xSr_xMnO₃, exhibits excellent catalytic performance. There are many other reports on the development of combustion catalysts based on perovskite structures, and the reader is referred to various review articles for further detail [6, 28, 29].

10.2.5 Other Metal Oxides

One of the most representative materials as a metal oxide catalyst for CO oxidation is Co₃O₄ with a spinel-type structure [30]. The Co₃O₄ catalyst having a rod-like morphology shows excellent catalytic ability even at room temperature. In general, the spinel structure presents a composition of AB₂O₄, where the A cation occupies the tetrahedral interstice formed in the close-packed structure of the O²⁻ ions, and the B cation occupies the octahedral interstice. The A and B cations are typically divalent and trivalent, respectively. For example, when clarifying the valence of Co species of Co₃O₄, it can be expressed as Co²⁺[Co³⁺]₂O₄. Since the 3d transition metal can be distributed at the A and B sites in the spinel structure, binary and ternary mixed oxides can be easily obtained by substituting various elements. Furthermore, it is possible to adjust the redox ability by selecting the metal species and multicomponent spinel oxides exhibit excellent catalytic ability. For example, spinel-type CoMn₂O₄ shows a higher ability to release lattice oxygen than a mixture of Co₃O₄ and MnO_x, a feature that is useful for toluene oxidation [31,32,33].

Other types of metal oxide catalysts are hexagonal rare-earth-iron mixed oxides (REFeO₃). REFeO₃ crystallizes in an orthorhombic perovskite structure with Pbnm being a stable form for all the rare-earth elements, while hexagonal REFeO₃ (h-REFeO₃) with P6₃cm exists as a metastable phase [34]. In the unit cell of h-REFeO₃, layers of iron ions and RE ions are stacked alternately, with the iron ions having a unique structure with trigonal bipyramidal 5-coordination (Fig. 10.5). When h-YbFeO₃ is synthesized by a solvothermal reaction, h-YbFeO₃ nanocrystals with thin hexagonal plate morphology are obtained, in which the crystal growth in the c-axis direction is suppressed. h-YbFeO₃ can act as a high-performance catalyst for combustion reactions of hydrocarbons [35, 36], and when modified with Mn species, it exhibits higher CO oxidation activity than PGM catalysts [37, 38].

Fig. 10.5

Crystal structure of a orthorhombic perovskite REFeO₃ with Pbnm and b hexagonal REFeO₃ with P6₃cm. Crystal structure is shown as a clear view of the coordination environment of the Fe ion. Purple balls, RE³⁺ ions; green balls, Fe³⁺ ions; red balls; O²⁻ ions

10.3 NO Oxidation and Reduction

10.3.1 Overview

NO_x abatement is becoming increasingly important due to the growing awareness of environmental issues, such as acid rain caused by NO_x emission from thermal power plants or automobiles [39,40,41]. Although the direct decomposition of NO is thermodynamically favorable, it is kinetically very difficult and often requires temperatures above 600 °C. Therefore, NO reduction to N₂ using hydrocarbons and NH₃ as reductants is widely applied. In contrast, since NO reduction is difficult in an oxygen-rich atmosphere, NO_x-trap materials are sometimes used. In such materials, as NO₂ is more easily adsorbed on the materials’ surface than NO, a material with NO oxidation ability in addition to NO_x adsorption ability would be ideal. A typical NO_x-trap catalyst, Pt/Ba/Al₂O₃, oxidizes NO to NO₂ at the Pt site and traps NO₂ at the BaO site [39, 42,43,44]. The NO_x species adsorbed on the catalyst are instantaneously reduced to N₂ by injecting a reductant such as hydrocarbons. Recently, PGM-free metal oxide catalysts that have both NO oxidation and NO_x storage capacity have been reported [45, 46].

In this section, NO_x storage materials involving NO oxidation mechanisms are firstly described, followed by catalyst materials for effective NO reduction based on the MvK-type reaction mechanism. Since most metal oxide catalysts following these reaction mechanisms are related to combustion catalysts [47], the characteristics of their crystal structures are included in Sect. 10.2.

10.3.2 Catalyst Materials

The reaction mechanism for NO oxidation on metal oxides is similar to the previously described MvK mechanism [47]. Many perovskite materials have been reported as catalyst materials for NO oxidation or NO_x storage, such as La_1-xSr_xCoO₃ which are known to exhibit a NO oxidation activity comparable to Pt/Al₂O₃ [48, 49]. Ruddlesden-Popper type perovskite Sr₃Fe₂O_7-${_{\delta}}$, which has a layered structure consisting of a perovskite layer (SrFeO_3-${_{\delta}}$ layer) and a rock salt layer (SrO layer), shows excellent NO_x storage properties [45]. For these materials, the NO oxidation process involves MvK-type NO oxidation with lattice oxygen, as shown in Eq. (10.3) [50].

$$ {\text{NO}} + {\text{O}}_{{\text{L}}} \to {\text{NO}}_{2} $$

(10.3)

The chemical formula of Sr₃Fe₂O_7−δ can be rewritten as SrO · 2SrFeO_3−δ, where the SrO layer is incorporated in the structure. In fact, monodentate nitrate (NO₃⁻) and ionic nitrate species were observed on the surface of Sr₃Fe₂O_7−δ after NO_x adsorption, suggesting that the SrO layer contributes to the adsorption of NO_x as shown in Eq. (10.4).

$$ {\text{SrO}} + 2{\text{NO}}_{2} + O^{*} \to {\text{Sr}}({\text{NO}}_{3} )_{2} $$

(10.4)

The oxygen source (O*) in Eq. (10.4) is either lattice oxygen species or oxygen species produced by the disproportionation of NO₂. Apart from the NO oxidation properties derived from the SrFeO_3−δ layer, Sr₃Fe₂O_7−δ also presents NO_x adsorption properties derived from the SrO layer, resulting in a high NO_x storage capacity (Fig. 10.6).

Fig. 10.6

Crystal structure of Sr₃Fe₂O₇₋${_{\delta}}$. Crystal structure is shown as a clear view of the coordination environment of the Fe ions. Purple balls, Sr²⁺ ions; green balls, Fe³⁺ or Fe⁴⁺ ions; red balls, O²⁻ ions

These materials require reaction temperatures above 250 °C since NO oxidation triggers NO_x adsorption. In contrast, passive NO_x adsorber (PNA) materials, in which Pd species are highly dispersed in zeolite materials such as ZSM-5 and SSZ, are known to effectively trap NO_x at lower temperatures [51,52,53]. The NO molecule is adsorbed as it is on the Pd species without any NO oxidation process. However, since the NO-to-Pd ratio is 1, a large amount of Pd loading is essential for high NO_x storage capacity. It has been reported recently that perovskite-type SrTi_1−xMn_xO₃ catalysts can efficiently adsorb NO at temperatures below 200 °C even without PGM loading [46].

Regarding NO reduction on metal oxides, the reduction mechanism has been recently proposed according to Eq. (10.5) (Fig. 10.7) [54, 55]:

$$ {\text{NO}} + {\text{V}}_{{\text{O}}} \to 1/2{\text{N}}_{2} + {\text{O}}_{{\text{L}}} $$

(10.5)

Fig. 10.7

Reaction mechanism of NO–CO reaction based on the MvK mechanism. a CO oxidation using lattice oxygen (O_L) and b NO reduction using oxygen vacancy (V_o)

In this reaction, oxygen vacancies are regarded as active sites instead of lattice oxygen that was the case for the MvK mechanism described in Eq. (10.1). During this reduction mechanism, the continuous generation of oxygen vacancy sites is difficult. Therefore, a high temperature or a reducing agent is required for reducing the NO to N₂. For NO reduction processes using CO as a reducing agent, the overall reaction can be written as follows:

$$ {\text{CO}} + {\text{O}}_{{\text{L}}} \to {\text{CO}}_{2} + {\text{V}}_{{\text{O}}} $$

(10.6)

$$ {\text{NO}} + {\text{V}}_{{\text{O}}} \to {\text{N}} + {\text{O}}_{{\text{L}}} $$

(10.7)

$$ {\text{N}} + {\text{N}} \to {\text{N}}_{2} $$

(10.8)

A supported PGM catalyst is often used for NO reduction. The reasons are that oxygen vacancy sites can be induced by the PGM loading, and that PGM sites can be used for NO/CO adsorption [38, 55]. For example, in procedures involving automotive exhaust purification, when a very small amount of Pd is supported on the Mn-modified hexagonal YbFeO₃, the NO reduction and CO oxidation are drastically enhanced at low temperatures. In this material, the arrangement of Mn cations on the MnO (111) plane is very close to that of Yb cations on the hexagonal YbFeO₃ (001); that is, the MnO (111) and YbFeO₃ (001) planes are bonded epitaxially. The role of Pd species was theoretically studied by the density functional theory (DFT) using a model in which Pd is supported by the MnO (111). It has been proven that the Pd loading promotes adsorption and dissociation of NO molecules. PGM-free materials have also been developed with Cu and Cr-modified CeO₂ catalysts exhibiting high catalytic activity following the MvK-type NO reduction mechanism [56]. It has also been found that spinel-type NiFe₂O₄ [57, 58] or multicomponent spinel-type catalysts containing four metallic elements (Cu, Ni, Al, Cr) are also effective for NO reduction [59].

10.4 Metal Oxides with Oxygen Storage Performance

10.4.1 Overview

Oxygen storage materials (OSMs) are oxides that can reversibly store and release lattice oxygens in response to variations in temperature and/or oxygen partial pressure. Therefore, this class of materials can be used in various oxygen-related technologies such as exhaust purification in automobiles, chemical looping reactions, and oxygen separation membranes/oxygen sensors [60,61,62,63,64]. OSMs show an interesting crystallographic feature which could be advantageous to rapid and reversible reactions; upon oxygen storage and release, slight structural changes take place. For many transition metal oxides such as spinel-type structures, the crystal structure is composed of close packing O²⁻ ions such that oxygen desorption may significantly deform the lattice framework. However, as described in Sect. 10.2, the packed structures of O²⁻ ions of fluorite-type CeO₂ and perovskite-type oxides are different from those of typical metal oxides. Materials with such characteristic O²⁻ ion-filled structures often store and release oxygen reversibly while maintaining the cationic arrangement. Chemical reactions in which the lattice framework is retained is called “topotactic” [65, 66]. The term “topotactic” is often used for Li intercalation/deintercalation of positive electrode materials in Li-ion secondary batteries. This section introduces the insertion/extraction processes of lattice oxygens based on topotactic reactions.

10.4.2 CeO₂–ZrO₂ Solid Solution

CeO₂–ZrO₂ solid solution (CZ) with fluorite structure is the most well-known oxygen storage material for automotive catalysts [67]. This material reversibly stores and releases lattice oxygens depending on the oxygen concentration in the exhaust gas, contributing to oxygen-pressure control around PGM species that act as active sites. It is known that CZ crystallizes in a fluorite-type structure when the material absorbs oxygen, and the oxygen storage capacity can be adjusted by the mixing ratio of Ce and Zr. For this material, oxygen storage/release proceeds via redox between Ce⁴⁺ and Ce³⁺. It can therefore be expected that the oxygen storage capacity depends on the Ce content. However, the oxygen storage capacity of CZ with about 30 mol% Zr is reported to be much higher than that of Zr-free CeO₂. During redox processes, the ionic radii of Ce and Zr are responsible for this phenomenon; when oxygen is released from the crystal lattice involving a reduction of Ce⁴⁺ (0.97 Å) to Ce³⁺ (1.14 Å), the lattice volume expands, resulting in local lattice distortions. The presence of smaller Zr⁴⁺ ions (0.84 Å) at the cation site adjacent to Ce³⁺ mitigates the lattice distortions to form a stable crystal structure containing Ce³⁺ and oxygen deficiencies [68].

During the preparation of CZ exhibiting improved oxygen storage performance, the formation of pyrochlore-type Ce₂Zr₂O₇ consisting of an ordered arrangement of Ce and Zr should be formed under a reducing atmosphere at temperatures above 1000 °C [67]. Upon oxygen storage, a phase transition from pyrochlore to fluorite-type structure occurs. Although these two crystal structures are distinct from each other, structural analyses proved that the cationic arrangement is maintained during the phase transition [69]. In other words, topotactic reactions proceed upon oxygen storage and release.

10.4.3 Perovskite and Layered Perovskite Oxides

The topotactic oxygen storage/release also occurs in mixed oxides with perovskite-related structures [70,71,72,73]. For this class of materials, since a close-packed structure is formed by A-site cations and O²⁻ ions, the release of oxygen proceeds with minor structural changes. For example, when perovskite-type SrFeO_3−δ is deoxygenated under a hydrogen-containing atmosphere, Sr₂Fe₂O₅ with an oxygen-deficient perovskite structure named “brownmillerite-type” is formed without rearrangements at the cation sites. Details of the brownmillerite-type structure are discussed below. However, severe reductive treatments of SrFeO_3−δ at high temperatures lead to structural collapse and the formation of metallic iron. To suppress this structural collapse, a partial substitution of Fe with Ti was reported to be effective by enhancing the structural stability [72].

On the other hand, Sr₃Fe₂O_7−δ with a layered perovskite structure cannot decompose into metallic iron even by severe reductive treatments [66]. The improved phase stability of Sr₃Fe₂O_7−δ is suggested to originate from SrO double layers within the crystal lattice, which would stabilize the oxygen-deficient structure. During oxygen storage/release, the crystal structure remains essentially unchanged with the I4mmm tetragonal space group (Fig. 10.8). The Fe species in Sr₃Fe₂O_7−δ are Fe³⁺/Fe⁴⁺ mixed valent, and the redox between Fe³⁺ and Fe⁴⁺ is responsible for the oxygen storage and release. The oxygen storage performances are higher than those of Pt/CZ (Fig. 10.9) and the partial replacement of Fe with Ni can improve the oxygen storage capacity of Sr₃Fe₂O_7−δ. Structural analysis of deoxygenated Sr₃(Fe_0.8Ni_0.2)₂O_7−δ showed that the reduction of Ni³⁺ to Ni²⁺ proceeds in addition to the reduction of Fe⁴⁺ to Fe³⁺ [74]. This phenomenon indicates that the Ni³⁺ substitution in Sr₃Fe₂O_7−δ induces the redox activity of the Fe³⁺ site. Indeed, the Pd/Sr₃(Fe_0.8Ni_0.2)₂O_7−δ catalyst shows superior catalytic activity to Pd/Sr₃Fe₂O_7−δ and Pd/CZ for the purification of automotive exhaust gases under fluctuating oxygen concentration, most likely owing to the enhanced oxygen storage capacity of Sr₃(Fe_0.8Ni_0.2)₂O_7−δ.

Fig. 10.8

Topotactic oxygen release and storage between Sr₃Fe₂O_7−δ and Sr₃Fe₂O₆. Crystal structure is shown as a clear view of the coordination environment of the Fe ions. Purple balls, Sr²⁺ ions; green balls, Fe³⁺ or Fe⁴⁺ ions; red balls, O²⁻ ions

Fig. 10.9

Oxygen storage performances of Pt/CeO₂-ZrO₂ and Sr₃Fe₂O_7−δ. These values were obtained using thermogravimetric analysis under H₂–O₂ cycles at 500 °C

10.4.4 Brownmillerite Oxides

As described in Sect. 10.4.3, perovskite oxides are potential OSMs owing to their topotactic oxygen storage/release capability. In the perovskite-type lattice, oxygen deficiencies often form long-range orders, leading to specific crystal phases that enhance structural stability. The brownmillerite (BM)-type structure is a member of the perovskite family with an oxygen deficiency-ordered structure [75]. The BM structure consists of a layered arrangement of tetrahedral BO₄ and octahedral BO₆ units, as illustrated in Fig. 10.10a, in contrast to the perovskite structure that only contains BO₆ units. The general formula for BM oxides is A₂B₂O₅ or ABO_2.5. This structural type is noteworthy owing to the high density of oxygen deficiencies among stable perovskite-derived lattices, that is, 1/6 with respect to the stoichiometric perovskite composition ABO₃. The high density of oxygen deficiencies in BM oxides is beneficial for enhancing their oxygen storage capacity.

Fig. 10.10

Crystal structure of a Ca₂AlMnO₅ with a brownmillerite structure and b YBaCo₄O₇. Crystal structure is shown as a clear view of the coordination environments of the Mn, Al, and Co ions. Purple balls, Ca²⁺ or Ba²⁺ ions; blue balls, Y³⁺ ions; green balls, Mn³⁺ or Co²⁺ ions; orange balls, Al³⁺ ions; red balls, O²⁻ ions

Several BM-type OSMs have been reported including (Sr,Ca)FeO_2.5 and Sr(Fe,Co)O_2.5 [76,77,78,79,80]. These oxides have attracted attention as potential materials for oxygen absorption that can selectively separate oxygen gas from the atmosphere. This technology is promising as a novel method of oxygen gas production, being more efficient than cryogenic O₂/N₂ distillation and conventional pressure-swing-adsorption using zeolites as N₂ sorbents. Ca₂AlMnO₅ is another BM-type OSM presenting alternating stacking of tetrahedral AlO₄ and octahedral MnO₆ [81]. This oxide stores excess oxygen topotactically and is then transformed into an oxygen-rich form, Ca₂AlMnO_5.5. The excess oxygen can then be released in a highly reversible manner [82]. The resultant oxygen storage capacity exceeds 3.0 wt%, which is even larger than that of CZ. Ca₂AlMnO_5+δ, with excess oxygen amounts δ ranging from 0 to 0.5, is promising in terms of elemental abundance, but its high operating temperatures (~550 °C) are an obstacle to future practical applications. Efforts have been made to lower the operating temperatures, such as Ga-for-Al or Sr-for-Ca substitutions and atomic-defect engineering [83,84,85,86].

10.4.5 Other Nonstoichiometric Oxides

Besides the aforementioned CeO₂-based materials and perovskite-related oxides, novel OSMs of other structural types have been extensively studied. Such materials are noteworthy because their unique atomic arrangements could induce characteristic redox behaviors and hence open up the possibility of new areas of applications. In this section, two OSMs, YBaCo₄O_7+δ and YMnO_3+δ, are briefly introduced. Noticeably, these oxides have the following common features, which could be crucial factors for high-performance OSMs. (1) Both the oxides contain cationic sites with smaller coordination numbers (CN) than six. Such “coordinatively unsaturated (CN < 6)” sites act as reaction centers for oxygen storage and release. (2) These oxides incorporate excess oxygen topotactically to transform into distinct oxygen-rich forms. Remarkable oxygen storage/release triggered by structural transformations arises even with slight variations in temperature and/or oxygen partial pressure.

YBaCo₄O_7+δ (denoted as “YBCO”) has a characteristic structure consisting of a 3D network of corner-sharing CoO₄ tetrahedra, as illustrated in Fig. 10.10b [87]. The corner-sharing framework favors excess-oxygen incorporation. Upon oxygen intake, a part of the CoO₄ tetrahedra transforms into CoO₆ octahedra, leading to large oxygen nonstoichiometry ranging 0 ≤ δ ≤ 1.5 [88, 89]. The oxygen storage/release processes are highly reversible, being controlled by both temperature and oxygen partial pressure. The maximum oxygen storage capacity reaches 4.2 wt%, corresponding to Δδ = 1.5 [90]. This value is much larger than that for CZ, making this oxide a promising candidate for high-performance OSM. YBCO works at relatively low temperatures of 300–400 °C in air, and such low operating temperatures are crucial for this oxide to be applied to oxygen-gas production [91]. It should also be pointed out that YBCO shows compositional flexibility, and its oxygen storage characteristics can be widely controlled by chemical substitutions [92].

Oxygen-stoichiometric YMnO₃ crystallizes in a hexagonal structure consisting of an alternate stacking of corner-shared MnO₅ trigonal-bipyramids and layers of edge-shared YO₇ polyhedra. This structure is essentially identical to that of h-YbFeO₃ shown in Fig. 10.5b. This oxide can incorporate excess oxygen into interstitial positions in the MnO₅ layer, resulting in the oxygen-rich form [93]. Upon oxygen intake, various oxygen-rich phases appear in the YMnO_3+δ system (denoted as “YMO”) depending on the amount of excess oxygen (δ). These phases are labeled as Hex1 (space group: R3c) with δ ≈ 0.28, Hex2 (Pca2₁) with δ ≈ 0.41, and Hex3 (P6₃mc) with δ ≈ 0.45 [94, 95]. YMO exhibits remarkable oxygen storage/release even at 250 °C or lower, with the maximum oxygen storage capacity reaching 2.5 wt%. The large oxygen storage capacity and low operating temperature ensure that YMO can be a promising OSM. Although YMO is effective as a material for oxygen-gas production, its sluggish oxygen storage/release kinetics are an obstacle to practical applications. To address this issue, various material tailoring protocols were implemented, and partial substitutions of lanthanoids at the Y site have proven effective. Materials with larger lanthanoids, such as Sm_0.25Y_0.75MnO_3+δ and Ce_0.15Tb_0.15Y_0.70MnO_3+δ, were found to show enhanced oxygen storage kinetics at low temperatures [96, 97].

10.5 CH₄ Conversion: Oxidative Coupling of Methane

10.5.1 Overview

Concerns regarding the depletion of natural carbon resources have increased interest in developing ways to effectively utilize methane. Methane is the primary component of natural gas and is the most promising alternative to petroleum. However, methane is a chemically stable molecule with no functional groups, limiting its use to simple combustion for mobility and energy generation. The oxidative coupling of methane (OCM) to C₂ hydrocarbons, such as C₂H₆ and C₂H₄, is a possible route for facile conversion into commodity chemicals [98, 99]:

$$ {\text{CH}}_{4} + 1/4{\text{O}}_{2} \to 1/2{\text{C}}_{2} {\text{H}}_{6} + 1/2{\text{H}}_{2} {\text{O}}\quad \Delta {\text{H}} = - 89\;{\text{kJ}}\;{\text{mol}}^{ - 1} $$

(10.9)

$$ {\text{CH}}_{4} + 1/2{\text{O}}_{2} \to 1/2{\text{C}}_{2} {\text{H}}_{4} + {\text{H}}_{2} {\text{O}}\quad \Delta {\text{H}} = - 141\;{\text{kJ}}\;{\text{mol}}^{ - 1} $$

(10.10)

OCM is a fascinating reaction that is direct, exothermic, and unaffected by thermodynamic constraints. However, to date, no economically feasible process has been established. The reason for this is the absence of suitable catalysts with high C₂ yields (>30%) and long-term durability to meet the practical requirements. The deep oxidation of methane to CO and CO₂ hinders OCM, causing the deterioration of the C₂ selectivity.

10.5.2 Existing Potential Catalysts and OCM Reaction Mechanism

Since the pioneering work [100] of Keller and Bhasin in the 1980s, approximately 2300 reports on OCM have been published. Statistical analysis of the literature data on OCM catalysts suggests several key factors that positively affect the OCM catalytic activity [101]. Catalysts containing alkali, alkali-earth, and rare-earth metals tend to show high catalytic activity. In fact, Li-doped MgO (Li/MgO) [102] and a composite consisting of manganese oxides and sodium tungstates on silica support (Mn-Na₂WO₄/SiO₂) [103, 104] are known to be OCM-active. Unfortunately, both catalysts face technical drawbacks: in long-term operations, they suffer from deactivation due to the vaporization of alkali metals. Regarding the practical application of OCM technology, alternative concepts for catalyst development may help improve the catalytic performance, as reported recently.

The catalytic mechanism of OCM has been extensively investigated. Previous studies have indicated that the formation of C₂ hydrocarbons predominantly involves the gas-phase association of methyl radicals (⋅CH₃) generated at oxygen sites on the catalyst surface [105, 106]. This finding implies that methane activation, that is, the hydrogen abstraction reaction by the catalyst, is one of the dominating factors affecting OCM activity. Several research groups have reported the importance of the electronic state of oxygen species on the catalyst surface. However, there is no concrete view of the role of such oxygen species, and the activation factors of highly active OCM catalysts are not completely understood.

10.5.3 Recent Progress in the Development of Complex Metal Oxide Catalysts

The aforementioned Mn-Na₂WO₄/SiO₂ catalyst is promising because of its relatively high catalytic activity; typically, C₂ yields are larger than 25% [103, 104]. Nevertheless, the primary activation factor for OCM is still debated owing to the chemical/structural complexity, causing difficulties in further improving its catalytic performance. To overcome this obstacle, single-phase oxides that contain multiple elements seem noteworthy. Single-phase catalysts provide well-defined active sites that are advantageous for mechanistic studies. On the lookout for novel OCM catalysts, few studies have been conducted based on such a strategy, and most of the previous works have focused on perovskite-type oxides [99]. Wang et al. characterized three model La₂B₂O₇ compounds with Ti⁴⁺, Zr⁴⁺, and Ce⁴⁺ at the B site to prove the phase structure-OCM activity relationship [107]. La₂Ti₂O₇, La₂Zr₂O₇, and La₂Ce₂O₇ crystallize in perovskite-related, pyrochlore, and defective fluorite structures, respectively. The OCM activity followed the order: La₂Ce₂O₇ > La₂Zr₂O₇ > La₂Ti₂O₇. The authors suggested that superoxide ${{\rm{O}}_2^ - }$ is the active oxygen species, and that the structural type influences the concentration of ${{\rm{O}}_2^ - }$ anions, emphasizing the importance of the crystallographic viewpoint.

Matsumoto et al. recently discovered the high OCM activity of crystalline Li₂CaSiO₄ (Fig. 10.11) [108]. The C₂ selectivity and CH₄ conversion reached 77.5% and 28.3%, respectively, at 750 °C with a feed gas ratio of CH₄/O₂ = 4/1. In addition, Li₂CaSiO₄ showed stable OCM performance in a durability test over 50 h without phase decomposition and compositional changes. The significant OCM activity of this catalyst is likely to originate from a combination of multiple cations in the crystal lattice; the local coordination around oxygen would give a dual effect of strong basicity and lattice stability, leading to simultaneous enhancements of the CH₄ conversion and C₂ selectivity. Notably, Li₄SiO₄ and Li₄GeO₄ were also found to be OCM-active, with performances comparable to Li₂CaSiO₄ [109]. The high OCM performance of these catalysts is believed to originate from the crystallographic similarities to Li₂CaSiO₄, where every oxygen site is surrounded by one Si/Ge atom and multiple Li atoms, which fulfill the OCM activation factors.

Fig. 10.11

Crystal structure of Li₂CaSiO₄. Crystal structure is shown as a clear view of the coordination environments of the Li and Si ions. Purple balls, Ca²⁺ ions; brown balls, Li⁺ ions; blue balls, Si⁴⁺ ions; red balls, O²⁻ ions