Volatile organic compounds (VOCs) are considered as a series of gaseous contaminants, which are not only carcinogenic to humans and animals, but also harmful to the environment [1−4]. Therefore, it is an urgent issue to minimize the emission of VOCs. Among the available technologies, the catalytic oxidation of VOCs is one of the most promising strategies owing to its lower operating temperature and less secondary pollution compared to those of direct thermal oxidation [5−7]. The catalysts, which play important roles in the catalytic oxidation, are classified into supported noble metal catalysts and transition metal oxide catalysts. Supported noble metal catalysts [8−10] exhibit high catalytic activity for VOC oxidation, while transition metal oxide catalysts, which are inexpensive and highly resistant to poisons, have attracted extensive attention. Among metal oxides, manganese oxides exhibit good catalytic activity in various catalytic oxidation reactions, such as HCHO oxidation [11], ethanol oxidation [12], and toluene oxidation [13], due to their naturally excellent physicochemical properties, including their controllable morphology [14], different crystal structures [15], and varied valences of Mn [16].

Nevertheless, the catalytic performance of individual metal oxides cannot satisfy the demands of practical applications; therefore, the formation of mixed metals oxides by introducing a hybrid metal is widely applied. Binary oxides, as the basic formation of mixed metal oxides, are common. Traditional methods, such as alkali-precipitation and impregnation, are applied to prepare the binary oxides [17, 18]. For example, Deng et al. [17] investigated the MnO₂@NiO core-shell hybrid nanostructure, which was prepared by uniformly decorating NiO nanosheets on the MnO₂ nanowire, and reported that it displayed much better activity toward benzene oxidation than individual MnO₂ nanowires because MnO₂@NiO exhibits better low-temperature reducibility and has a high number of active surface oxygen species. Putla et al. [18] reported that MnO_x/CeO₂ composites, formed by the deposition of MnO_x on CeO₂ nanocubes by a wet impregnation method, exhibited outstanding catalytic activities toward both diesel soot oxidation and benzylamine oxidation compared to pure CeO₂ nanocubes. The synergistic effect between the CeO₂ and MnO_x components is responsible for the excellent catalytic activity of the MnO_x/CeO₂ composites.

Recently, metal-organic frameworks (MOFs), considered as effective sacrificial templates for the synthesis of catalysts for VOC catalytic oxidation, have been attracting extensive attention [6, 19-21]. For instance, Jia et al. [19] revealed that a mesoporous CeO₂-MOF/350 catalyst, synthesized through the pyrolysis of Ce-MOF, exhibited good catalytic performance for toluene combustion. Zhang et al. [20] reported that Mn-100-Ar-O derived from Mn-MIL-100 exhibited superior catalytic activity. Wang et al. [6] prepared Mn–Co oxides using MOF as the precursor, and they reported that MOF-Mn1Co1 displayed excellent catalytic activity for toluene combustion. Tang et al. [21] prepared CeCoO_x catalysts by the pyrolysis of a MOF, and they reported that the 3D mesoporous rosette CeCoO_x catalyst possessed superior catalytic activity for toluene total oxidation. Evidently, the preparation of individual metal oxides and binary oxide catalysts by the direct thermal decomposition of MOF is conventional. Moreover, it has been reported that binary oxides with a constructed coupled interface can be obtained by the pyrolysis of MOFs, to be applied in lithium-ion batteries [22] and NO reduction [23]. Zhong et al. [22] synthesized yolk-shell MnO@ZnMn₂O₄/N-C nanorods derived from α-MnO₂/zeolitic imidazolate frameworks-8 (ZIF-8), which were applied as anode materials in lithium-ion batteries; they exhibited better electrochemical performances than pure MnO_x due to the synergistic effect between MnO and ZnMn₂O₄ in the unique yolk-shell nanorod structure. Wang et al. [23] revealed that the CeO₂-Co₃O₄ hybrid catalyst, obtained via the thermolysis of ZIF-67, exhibited better catalytic activity for NO reduction by CO than bare CeO₂ nanowires and Co₃O₄ catalysts, which was ascribed to the synergistic effect between the metal oxides. However, it is not common to synthesize binary oxides with a coupled interface constructed by the pyrolysis of a MOF, for application in catalytic toluene oxidation.

Inspired by the above-mentioned works, we report a bottom-down approach for designing high-performance MnO₂@Co₃O₄ catalysts, which exhibit remarkable catalytic activities for toluene oxidation in comparison to pure MnO₂ and Co₃O₄ catalysts. The catalysts were synthesized via ZIF-67, grown in situ on MnO₂ with various crystal structures, followed by direct thermal decomposition. The structure–activity relationship is confirmed using X-ray diffraction (XRD), Brunauer–Emmett–Teller (BET) studies, scanning electron microscopy (SEM), transmission electron microscopy (TEM), high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), hydrogen temperature-programmed reduction (H₂-TPR), X-ray photoelectron spectroscopy (XPS), and temperature-programmed desorption of oxygen (O₂-TPD). In situ diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) experiments were performed on pure α-MnO₂ and α-MnO₂@Co₃O₄ to further investigate the synergistic effect between α-MnO₂ and Co₃O₄ on the catalytic toluene oxidation. The enhanced catalytic activity can be attributed largely to the synergistic effect between α-MnO₂ and Co₃O₄.

Synthesis of α-MnO₂: KMnO₄ (40 mmol) was dissolved in 150 mL of deionized water. After stirring for 30 min, the above solution was added to another homogeneous solution containing 50 mmol MnSO₄ in 50 mL of deionized water. After stirring for another 30 min, 5 mL of concentrated nitric acid (68 wt%) was introduced with vigorous stirring. Subsequently, the obtained dark brown slurry was refluxed at 100 ℃ for 24 h and then cooled to room temperature.

The sample was collected by rinsing, centrifugation, drying, and annealing at 350 ℃ for 3 h at a heating rate of 1 ℃ min^–1 in static air.

Synthesis of α-MnO₂@ZIF-67: Seventy milligrams of the as-obtained α-MnO₂ and 350 mg of polyvinyl pyrrolidone (PVP) were added to 25 mL of methanol. After the α-MnO₂ sample was uniformly dispersed by ultrasonication for 30 min, the resultant solution was mixed with 291 mg of Co(NO₃)₂ 6H₂O. Thereafter, the slurry (A) was stirred for 30 min at room temperature. Subsequently, 493 mg of 2-methylimidazole was dissolved in 25 mL of methanol to form a solution (B), which was added to A under stirring for 30 min. The mixture was kept still for 12 h, after which the obtained precipitate was collected by centrifugation, rinsed thoroughly with methanol several times to remove the residual ions, and finally dried at 80 ℃ for 12 h. The sample was denoted as α-MnO₂@ZIF-67.

Synthesis of α-MnO₂@Co₃O₄: The α-MnO₂@ZIF-67 sample was annealing at 350 ℃ for 3 h at a heating rate of 1 ℃ min^–1 in static air.

Synthesis of β-MnO₂: 5 mmol MnSO₄ and an equal amount of ((NH₄)₂S₂O₈ were dissolved into 60 mL deionized water. After stirring for 30 min, the obtained homogeneous solution was transferred to a Teflon-lined autoclave with 100 mL capacity. Hydrothermal process was processed at 120 ℃ for 12 h, then the autoclave was cooled naturally to room temperature.

The sample was collected by rinsing, centrifuging, drying and annealing at 350 ℃ for 3 h with a heating rate of 1 ℃ min^–1 in static air.

Synthesis of β-MnO₂@ZIF-67: The synthesis of β-MnO₂ @ZIF-67 was the same as α-MnO₂@ ZIF-67 except for the usage of β-MnO₂ instead of α-MnO₂.

Synthesis of ZIF-67: The synthesis of ZIF-67 was the same as α-MnO₂@ZIF-67 except for without α-MnO₂.

The β-MnO₂@ZIF-67 and ZIF-67 samples annealed at 350 ℃ for 3 h with a heating rate of 1 ℃ min^–1 in static air were obtained β-MnO₂@Co₃O₄ and Co₃O₄-b, respectively.

The surface morphologies and microstructure of samples were investigated by field-emission scanning electron microscopy (SEM, ZEISS) and transmission electron microscopy (TEM, JEOL JEM-2010F), respectively. High-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) and energy-dispersive X-ray spectroscopy (EDS) element mapping were recorded on TEM instrument with a field emission gun operating at 200 kV. X-ray diffraction (XRD) were obtained on a Bruker D8 ADVANCE diffractometer with Cu K_α radiation (40 kV, 40 mA, scanning step = 0.02°) and further applied for determining the crystal structure of all samples. The chemical composition was determined by inductively coupled plasma-optical emission spectroscopy (ICP-OES, Perkin-Elmer plasma 8000). The Brunauer-Emmett-Teller (BET) testing was measured with a Micromeritics ASAP 2020 instrument at −196 ℃. Prior to the measurement, the sample was pretreated at 120 ℃ for 3 h under vacuum condition, and the Brunauer-Emmett-Teller (BET) model was used to calculate the specific surfaces area. The surface species was performed using XPS spectrum obtained by X-ray photoelectron spectroscopy (XPS) testing (Thermo ESCALAB 250Xi electron spectrometer), and every spectrum was referenced to the carbon 1s peak at 284.6 eV. Temperature-programmed reduction in H₂ (H₂-TPR) was conducted on a Micromeritics AutoChem II2920 instrument to get the reducibility information of the catalysts. Before each measurement, the catalyst (50 mg) in a U-shaped quartz reactor was heated 300 ℃ at a ramping rate of 10 ℃ min⁻¹ and maintained at the temperature for 1h in high purity Ar (99.999%). After the pretreatment, reactor was cooled down to room temperature. 10% H₂/Ar flow (30 mL min⁻¹) was employed and the samples were heated up to 700 ℃ at a ramping rate of 10 ℃ min⁻¹. For temperature-programmed desorption of O₂ (O₂-TPD), the catalyst (50 mg) in a U-shaped quartz reactor was heated 300 ℃ at a ramping rate of 10 ℃ min⁻¹ and maintained at the temperature for 1 h in high purity Ar (99.999%). After the pretreatment, reactor was cooled down to room temperature. 5 % O₂/He (30 mL min⁻¹) was switched into the reactor for 1h, following by purging with high purity He (99.999%) at the same temperature for 30 min to remove the unabsorbed molecules of O₂. The catalysts were heated from room temperature to 750 ℃ at a ramping rate of 10 ℃ min⁻¹ in high purity He (30 mL min⁻¹).

In situ diffuse reflection infrared Fourier transform spectroscopy (in situ DRIFTS) was recorded on a Nicolet iS50R FTIR spectrometer equipped with a liquid nitrogen-cooled Mercury Cadmium Telluride (MCT) detector in the range of 700−3200 cm⁻¹ with 32 scans at a resolution of 4 cm⁻¹. A high temperature reaction chamber from Harrick Scientific Products Inc. was used as an environmental chamber, and a dome with two pieces of KBr windows (thickness = 2 mm, diameter = 15 mm) and one glass observation window were used in the chamber. The sample was placed into a sample cell in an atmospherically controllable chamber for in situ DRIFTS analysis. Meanwhile, looping water kept the outside of the chamber cool when operating at a high temperature. Prior to each experiment, the sample was pre-treated at 300 ℃ for 1 h in a gas flow of N₂ to remove any adsorbed impurities and then cooled to a desired temperature. The background spectrum were collected under N₂ and automatically subtracted from the sample spectra at the identical temperature. Afterward, 500 ppm of toluene balanced with N₂ was introduced to the cell at a flow rate of 30 mL min⁻¹, and then in situ DRIFTS were recorded. After adsorbed toluene for 250 min, the weakly adsorbed toluene was removed by purging with N₂ for 30 min, subsequently, 20% O₂/N₂ with the flow rate of 30 mL min⁻¹ was introduced into the cell to react with pre-adsorbed toluene. It is necessary to emphasis that the temperature is identical during the adsorption and oxidation of toluene to avoid two factors including both desorption of toluene from the surface of sample and the signal of in situ DRIFTS effected by the temperature variation.

The catalytic activities of catalysts for toluene oxidation were carried out in a fixed-bed quartz tubular micro-reactor (i.d. = 6 mm, length = 500 mm) at atmospheric pressure under steady-state conditions. 100 mg of catalyst (40–60 mesh) mixed with 400 mg of quartz sands (40–60 mesh) minimizing the effect of hot spots was loaded in the quartz reactor with quartz wool packed at the both ends of the catalyst bed. A thermocouple was inserted inside the catalytic bed to measure the reaction temperature. A gas mixture composed of 1000 ppm of toluene, 20% O₂, and balance N₂ was fed into the micro-reactor, and the total flow rate was held at 80 mL min⁻¹ via a mass flow controller, equivalent to a weight hourly space velocity (WHSV) of 48000 mL g⁻¹ h⁻¹. After the toluene oxidation reaction was stabilized for 15 min, the outlet products were continuously monitored using a gas chromatograph at a given temperature for 8 times. Toluene conversions were monitored by a gas chromatograph (GC-2014C, Shimadzu) equipped with a flame ionization detector and the concentrations of CO₂ in the outlet gas was analyzed by other FID with a conversion furnace for converting CO₂ to CH₄. In the case of water vapor addition, 2.0% of H₂O was introduced via a mass flow controller, which was achieved by a saturated H₂O vapor generator at 40 ℃. The toluene conversion (η_toluene) was calculated as per the equations.

Herein, C_{toluene, in} (ppm) and C_{toluene, out} (ppm) are the concentrations of toluene in the inlet and outlet gas, respectively.

A typical procedure based on a two-step strategy, as illustrated in Scheme 1, was applied to synthesize the α-MnO₂@Co₃O₄ hybrid materials. Firstly, α-MnO₂ was fabricated via a reflux method. Thereafter, the ZIF-67 rhombic dodecahedron was grown in situ on MnO₂ (denoted as α-MnO₂@ZIF-67). The ZIF-derived Co₃O₄, grown in situ on α-MnO₂ with a constructed coupled interface (denoted as α-MnO₂@Co₃O₄), was obtained after calcination at 350 ℃ for 3 h in static air.

The XRD patterns of the as-obtained α-MnO₂ and α-MnO₂@Co₃O₄ samples are displayed in Fig. 1. The peaks displayed at 12.8°, 18.1°, 25.7°, 28.8°, 37.5°, 42.0°, and 49.9° are ascribed to the (110), (200), (220), (310), (211), (301), and (411) facets of the pure tetragonal α-MnO₂ catalyst (JCPDS card no. 44-0141), respectively [15]. After decorating Co₃O₄ on the α-MnO₂, the original diffraction peaks of MnO₂ were adequately maintained. In addition, several distinct peaks at 19.0°, 31.3°, 36.8°, 44.8°, 59.4°, and 65.2° are attributed to the (111), (220), (311), (400), (511), and (440) facets of the cubic spinel Co₃O₄ sample (JCPDS card no. 42-1467), respectively [3]. Moreover, the XRD patterns of the α-MnO₂@ZIF-67, ZIF-67, and Co₃O₄-b samples are displayed in Fig. S1. The Co/Mn molar ratio of the as-obtained α-MnO₂@Co₃O₄ is 0.44 (Table 1), as confirmed by inductively coupled plasma atomic emission spectroscopy (ICP-OES), which is lower than the nominal Co/Mn molar ratio owing to the low yield of ZIF-67.

The N₂ adsorption-desorption isotherms and pore-size distribution curves of the α-MnO₂, α-MnO₂@Co₃O₄, and Co₃O₄-b samples are shown in Fig. S2. The BET data are summarized in Table 1. The specific surface area of α-MnO₂@Co₃O₄ (54.9 m² g⁻¹) is slightly less than that of α-MnO₂ (60.6 m² g⁻¹). The BET result indicates that the specific surface area of the α-MnO₂@Co₃O₄ samples is determined by the physical structure of both α-MnO₂ and Co₃O₄-b, considerably.

The SEM images reveal the morphologies of the α-MnO₂ samples (before and after calcination at 350 ℃), α-MnO₂@ZIF-67, and α-MnO₂@Co₃O₄, as shown in Figs. 2(a), (b), 2(c), (d), and 2(e)-(f), respectively. It can be distinctly observed that the 1D α-MnO₂ nanowires have different lengths and diameters. As shown in Figs.2(a),(b), the length of the α-MnO₂ nanowire is roughly 0.75 μm. The diameter of α-MnO₂ would be discussed in the section presenting the TEM results. After calcination, the morphology of the α-MnO₂ samples remained intact. Notably, the α-MnO₂ nanowire is embedded in the ZIF-67 rhombic dodecahedron, α-MnO₂@ZIF-67 (highlighted by the faint yellow circles), as clearly shown in Fig. 2(c). Moreover, the schematic illustration of the α-MnO₂@ZIF-67 samples is depicted in Fig. 2(d). After calcination at 350 ℃ in air, the morphology of the as-obtained α-MnO₂@Co₃O₄ is partly destroyed due to the high calcination temperature. Nevertheless, it is noteworthy that the as-obtained α-MnO₂@Co₃O₄ with a hierarchical nanostructure maintained a rhombic dodecahedron morphology to some extent, as can be distinctly observed from Fig. 2(e). Moreover, the schematic illustration of α-MnO₂@Co₃O₄ is displayed in Fig. 2(f).

The representative TEM images of α-MnO₂ and α-MnO₂@Co₃O₄ in Fig. 3 present more details about their microstructures. The diameter of the α-MnO₂ nanowires is roughly 25 nm, as shown in Figs. 3(a) and (b). The HRTEM images (Fig. 3(c)) display that the interplanar spacing of the fringe is 0.31 nm, which is consistent with the (310) plane of α-MnO₂ [24, 25]. It can be observed that the α-MnO₂ nanomaterials are embedded in Co₃O₄ derived from ZIF-67, as illustrated in Figs. 3(d), (e). There are three typical lattice fringes of 0.48, 0.45, and 0.29 nm, corresponding well to the (200) plane of α-MnO₂ [25], and the (111) and (220) facets of Co₃O₄, respectively, as illustrated in Fig. 3(f). Furthermore, the blue dotted line (Fig. 3(f)), at which point Co₃O₄ touches the surface of the α-MnO₂ nanowire, denotes the coupled interface. The effect of the coupled interface would be explained by further characterizations. The HAADF-STEM images of α-MnO₂@Co₃O₄ with a hierarchical nanostructure are exhibited in Fig. 3(g), which also confirm that the MnO₂ nanomaterials are embedded in Co₃O₄ derived from ZIF-67. Noticeably, as shown in the elemental distribution maps (Figs. 3(h)–(j)), Co is distributed on the surface of the 1D MnO₂. The line-scan images (Fig. 3(k), (l)) further confirm the distribution of Co on the surface of the 1D MnO₂ nanowire. The HRTEM and HAADF-STEM images show that the ZIF 67-derived Co₃O₄ is successfully grown in situ on the α-MnO₂ nanomaterials with a constructed coupled interface.

To investigate the reduction behavior over the α-MnO₂@Co₃O₄ catalyst, H₂-TPR was employed using α-MnO₂ and Co₃O₄-b as reference samples, as illustrated in Fig. 4. Generally, the reduction of MnO₂ proceeds via the sequence: MnO₂ → Mn₃O₄ → MnO. For MnO₂, the lower temperature reduction peak is ascribed to the reduction of MnO₂ to Mn₃O₄, while the higher one is attributed to the reduction of Mn₃O₄ to MnO [11, 25]. However, the reduction of MnO₂ may occur at a certain temperature range, which would combine the reduction peaks. For α-MnO₂, the asymmetrical reduction peak manifests that the main reduction steps are processed together in the range of 200–370 ℃. In addition, the α-MnO₂@Co₃O₄ sample profile displays complex reduction peaks occurring at the range of 120–380 ℃, attributed to the reduction process of Co₃O₄ [26, 27]. It can be clearly observed that the reduction peaks of α-MnO₂@Co₃O₄ shift toward lower temperatures in comparison to those of pure α-MnO₂ and Co₃O₄-b (Fig. 4). According to the literatures [25, 28, 29], a catalyst shows a reduction peak at the lowest temperature, indicating that the highest amount of mobile oxygen species is present both at the surface and in the bulk. Therefore, α-MnO₂@Co₃O₄ shows higher oxygen mobility than both pure α-MnO₂ and Co₃O₄-b, which is attributed to the strong synergistic effect derived from the coupled interface constructed between α-MnO₂ and Co₃O₄ and is beneficial for toluene oxidation [30].

The type of oxygen species was determined by O₂-TPD, as depicted in Fig. 5. Generally, there are four types of oxygen species in a metal oxide [31, 32]: the adsorbed oxygen molecule (ads-O₂), adsorbed oxygen molecule anion (ads-O₂⁻), adsorbed oxygen anion (ads-O⁻), and lattice oxygen (latt-O²⁻). For the α-MnO₂@Co₃O₄ sample, the peak in the range of 60–200 ℃ is ascribed to the desorption of ads-O₂ and ads-O₂⁻; the peak between 300  and 650 ℃ is attributed to the ads-O⁻ and lattice oxygen species near to the surface, while the peak above 650 ℃ belongs to the desorption of bulk‐phase lattice oxygen species. At the range of 60–200 ℃, the peak intensity of α-MnO₂@Co₃O₄ is stronger than those of pure α-MnO₂ and Co₃O₄-b, suggesting that a relatively high number of active oxygen species can be more easily released from the surface of the α-MnO₂@Co₃O₄ catalysts [3, 32]. It is well known that the higher the number of surface oxygen vacancies, the easier the activation of the gaseous phase (O₂) to surface-active oxygen species (O_ads) [33]. In other words, the strong synergistic effect derived from the coupled interface constructed between α-MnO₂ and Co₃O₄ is favorable to the formation of active oxygen species.

Subsequently, XPS is applied to determine the oxygen species and valence states of the elements in the surface atomic layers of the catalysts. It is well known that the surface oxygen species play an important role in the catalytic oxidation [34, 35]. The XPS curves of O 1s for α-MnO₂, α-MnO₂@Co₃O₄, and Co₃O₄-b samples are fitted to distinguish the possible existence of different oxygen species, as illustrated in Fig. 6(c). The peak at about 529.7 eV is ascribed to the lattice oxygen (O_latt) species, and the shoulder peak at the higher binding energy of around 531.5 eV is assigned to the surface-adsorbed oxygen (O_ads) species, which are directly associated with the catalytic activity of toluene oxidation [36, 37]. In comparison with the cases of pure α-MnO₂ and Co₃O₄-b, the molar ratio of O_ads/O_latt in α-MnO₂@Co₃O₄ is notably increased by decorating Co₃O₄ on α-MnO₂, which is favorable to the catalytic oxidation of toluene over the α-MnO₂@Co₃O₄ catalysts.

As shown in Fig. 6(a), the peaks of Mn 2p_3/2 and Mn 2p_1/2 are centered at 642.2 and 653.9 eV, respectively [38]. The spin energy separation of 11.7 eV is in good agreement with the reported data of Mn 2p_3/2 and Mn 2p_1/2 in MnO₂ [39]. The three components divided from the asymmetrical Mn 2p_3/2 XPS profiles at 640.8, 642.2, and 643.7 eV are attributed to the surface Mn³⁺, Mn⁴⁺ species, and the satellite, respectively [40, 41]. Fig. 6(b) presents the Co 2p XPS profiles of the α-MnO₂@Co₃O₄ and Co₃O₄-b samples. The asymmetrical Co 2p_3/2 XPS profiles of all samples can be deconvoluted into two component peaks of Co³⁺ (779.5 eV) and Co²⁺ (780.8 eV). Generally, the cations of Mn³⁺ and Co²⁺ are considered to be involved in the formation of lattice oxygen, since the reversible valence transformation of cations (Mn³⁺ ⇋ Mn⁴⁺ + e and Co²⁺ ⇋ Co³⁺ + e) donates some electrons to activate oxygen during the catalytic reaction [42, 43]. Based on the peak areas, the relative contents of Mn³⁺/Mn⁴⁺ and Co³⁺/Co²⁺ molar ratios are summarized in Table 1. The Mn³⁺/Mn⁴⁺ and Co³⁺/Co²⁺ molar ratios are noticeably increased by the introduction of ZIF-67-derived Co₃O₄ on α-MnO₂.

From another perspective, the effect of valence change on the physicochemical properties is also explained. Chen et al. [43] recently reported that the interfacial effect existing in Ce-Mn composite oxides plays a key role in the modification of their physicochemical properties. The results of XPS indicated that the proportion of Ce³⁺ and Mn⁴⁺ cations is noticeably changed by the introduction of an appropriate amount of CeO_y. In other words, there is a strong synergistic interaction between Ce and Mn oxides. Moreover, the gaseous oxygen could be readily activated into active adsorbed oxygen species on the surface of α-MnO₂@Co₃O₄ by Co cations through the Zener exchange owing to the reversible electron transfer between Co³⁺ and Mn³⁺, thereby leading to the formation of a significantly high amount of active oxygen species on the surface of α-MnO₂@Co₃O₄ [44], which is consist with the result of O₂-TPD.

The catalytic activities of toluene oxidation over the α-MnO₂, α-MnO₂@Co₃O₄, and Co₃O₄-b catalysts are evaluated under the conditions of toluene concentration = 1000 ppm, WHSV = 48 000 mL g⁻¹ h⁻¹, and 20% O₂/N₂ as balance gas, as illustrated in Fig. 7. For pure α-MnO₂, the reaction temperatures for toluene conversions of 50% (T_50%) and 90% (T_90%) are 264 and 276 ℃, respectively, as listed in Table 1. Over pure Co₃O₄-b, the reaction temperatures for T_50% and T_90% are 245 and 257 ℃, respectively. Interestingly, α-MnO₂@Co₃O₄ is more active than pure α-MnO₂ and Co₃O₄-b. The T_50% and T_90% values of α-MnO₂@Co₃O₄ are 220 and 229 ℃, respectively, which are 44 and 47 ℃ lower than those of pure α-MnO₂, and 25 and 28 ℃ lower than those of Co₃O₄-b, respectively. The results suggest that the synergistic effect derived from the coupled interface constructed between the α-MnO₂ and Co₃O₄ catalysts significantly promotes the toluene oxidation catalytic activity. In addition, the catalytic activities of typical samples for toluene oxidation are summarized in Table 2. Evidently, the catalytic activity over the α-MnO₂@Co₃O₄ hybrid catalyst is considerably better than those previous reported for other catalysts, such as Mn10Cu1-S (T_90% = 258 ℃) [45], Mn-Co (1:1) (T_90% = 249 ℃) [46], 7.4Au/Co₃O₄ (T_90% = 250 ℃) [47], 1% Pt/Al₂O₃ (T_90% = 258 ℃) [48], 0.12Ag/Mn₂O₃-redn (T_90% = 250 ℃) [49], and Pd/Co₃AlO (T_90% = 230 ℃) [31].

Moreover, to investigate the effect of water vapor on the catalytic performance, we carried out a toluene oxidation experiment in the presence of 2% water vapor over the α-MnO₂@Co₃O₄ catalyst at 240 ℃, as shown in Fig. S8(a). It is distinctly observed that the incorporation of 2% water vapor at 240 ℃ led to a decrease in the toluene conversion by ca. 15%, which is ascribed to the competitive adsorption of H₂O and reactant molecules (oxygen molecules and toluene molecules) over the α-MnO₂@Co₃O₄ catalyst [50, 51]. When the water vapor is cut off, the toluene conversion is almost restored to the initial value, which indicates that α-MnO₂@Co₃O₄ possesses good water-resistant ability in toluene oxidation.

We subsequently extended the fabrication method to other 1D MnO₂ nanomaterials. β-MnO₂, β-MnO₂@ZIF-67, and β-MnO₂@Co₃O₄ were synthesized. The metal Co/Mn molar ratios of these β-MnO₂@Co₃O₄ materials were determined by ICP-OES. The characterizations by XRD, BET, SEM, TEM, HRTEM, EDS elemental analysis, H₂-TPR, O₂-TPD (Figs. S1–S6), and XPS (Fig. S7) generated a similar structural picture to that of the α-MnO₂@Co₃O₄ analog: 1D MnO₂ embedded in the Co₃O₄ rhombic dodecahedron (Figs. S4(d), (e)). The key properties of materials are summarized in Table 1. We tested the catalytic activity of β-MnO₂@Co₃O₄ toward toluene oxidation under the conditions of toluene concentration = 1000 ppm, WHSV = 48 000 mL g⁻¹ h⁻¹, and 20% O₂/N₂ as balance gas, as illustrated in Fig. 7. The activities of the pure β-MnO₂ and Co₃O₄-b catalysts were measured for comparison. The results of the catalytic activity tests of the catalysts are summarized in Table 1. In addition, β-MnO₂@Co₃O₄ displayed a good water-resistant ability in toluene oxidation, as illustrated in Fig. S8(b). The foremost conclusion from these results is that the structure of 1D MnO₂@Co₃O₄ dictates its catalytic ability, which underscores the utility of our synthetic methodology. The performance of the β-MnO₂@Co₃O₄ catalyst is comparable to that of the Co₃O₄-b catalyst because the Co content of β-MnO₂@Co₃O₄ is low (Table 1).

In situ DRIFTS experiments were performed to track the intermediates and investigate the reaction mechanisms of toluene oxidation over pure α-MnO₂ and α-MnO₂@Co₃O₄; the results are illustrated in Fig. 8. The samples were firstly pre-treated under N₂ atmosphere at 300 ℃ for 1 h, after which they were cooled to 180 ℃ under N₂. Subsequently, the samples were exposed under a flow of 500 ppm toluene/N₂ at 180 ℃, as shown in Figs. 8(a) and 8(c). For pure α-MnO₂ (Fig. 8(a)), it was evident that the intensities of the aromatic ring vibration peaks at about 1558 and 1595 cm⁻¹ [6, 60, 61] and the peak at 1450 cm⁻¹ associated with the asymmetric methyl bending vibrations of toluene [62] were increased with time, which indicated the accumulation of adsorbed toluene on the surface of α-MnO₂. Notably, the peak at approximately 1545 cm⁻¹ is assigned to the carboxylate group (–COOH); simultaneously, the peaks at 1414 and 1521 cm⁻¹ are ascribed to the symmetric C–O stretching vibration and antisymmetric C–O stretching vibration of benzoate species, respectively [61]. Further, the intensity of the bands (benzoate species) gradually ascends with time in the presence of 500 ppm toluene/N₂, indicating that the benzoate species are the key intermediates in the toluene oxidation over α-MnO₂. In addition, it is necessary to mention that the desorption temperatures of ads-O₂ and ads-O₂⁻ are in the range of 60–200 ℃, according to the result of O₂-TPD. Therefore, it is reasonable to speculate that the adsorbed oxygen species (ads-O₂ and ads-O₂⁻) may be involved in the transformation of adsorbed toluene into benzoate species without the gas-phase oxygen at 180 ℃. Moreover, the peaks at 2339 and 2362 cm⁻¹ are attributed to CO₂ [53]. It can be distinctly observed that the toluene molecules are partially oxidized to intermediates (benzoate species), even to CO₂, further suggesting that the adsorbed oxygen species (ads-O₂ and ads-O₂⁻) on the surface of the catalyst may participate in the deep oxidation of toluene. For α-MnO₂@Co₃O₄ (Fig. 8(c)), a distinct increasing trend can be observed for toluene (1601 and 1445 cm⁻¹) [6, 62], the benzoate species (1545 and 1402 cm⁻¹) [61], and CO₂ (2339 and 2362 cm⁻¹) [53] with time.

After the adsorption of toluene over the catalysts for 250 min, the weakly adsorbed toluene molecules were removed by purging with N₂ for 30 min; subsequently, 20% O₂/N₂ was flowed into the reaction chamber at 180 ℃. For pure α-MnO₂ (Fig. 8(b)), it was observed that the peak intensities did not decrease significantly even at 250 min, indicating that the adsorbed toluene (1450, 1558, and 1595 cm⁻¹) [6, 62] and the benzoate species (1414, 1521 and 1545 cm⁻¹) [61] were barely further oxidized at 180 ℃ in 20% O₂/N₂. Contrarily, for α-MnO₂@Co₃O₄ (Fig. 8(d)), the peak intensities of toluene (1445, 1560, and 1601 cm⁻¹) and the benzoate species (1402 and 1545 cm⁻¹) distinctly decreased at 50 min and even completely disappeared at 250 min. The reason is that the gaseous oxygen (O₂) could be more easily activated into active adsorbed oxygen species (O_ads) on the surface of α-MnO₂@Co₃O₄ according to the XPS and O₂-TPD results, which gives rise to the oxidation of toluene into key intermediate benzoate species and, finally, into CO₂ over the α-MnO₂@Co₃O₄ catalyst.

Based on the in situ DRIFTS characterization, we can speculate the possible reaction path over α-MnO₂@Co₃O₄, as follows [63, 64]: adsorbed toluene → benzoate species → alkanes containing oxygen functional group (such as maleic anhydride) → CO₂ and H₂O. Moreover, it was observed that toluene can be more easily completely oxidized over α-MnO₂@Co₃O₄ than over the pure α-MnO₂ catalyst at the same temperature, which is consistent with the results of the catalytic activity tests for toluene oxidation. The results further indicated that the synergistic effect between α-MnO₂ and Co₃O₄ in the α-MnO₂@Co₃O₄ catalyst is beneficial to the toluene oxidation.

Since the synergistic effect derived from the coupled interface constructed between MnO₂ and Co₃O₄ has a significant influence on the catalytic performance, the inner relationship is investigated by the combination of physico-chemical properties. It is recognized that a large specific surface area is beneficial for enhancing the adsorption and retention time of gaseous toluene. However, although α-MnO₂ possesses the highest specific surface area, it does not exhibit the best catalytic activity for toluene oxidation; this indicates that the specific surface area is not the primary factor in the enhancement of the catalytic activity. The chemical properties of the catalysts play important roles in enhancing the catalytic activity. The H₂-TPR result suggests that the oxygen mobility of α-MnO₂@Co₃O₄ is notably enhanced by the synergistic effect derived from the constructed coupled interface, which can accelerate the oxidation process. Combining the XPS and O₂-TPD results, it is established that synergistic effect owing to the constructed coupled interface distinctly facilitates the augmentation of absorbed oxygen species on α-MnO₂@Co₃O₄. Generally, a high number of absorbed oxygen species is beneficial for the oxidation of toluene, and absorbed oxygen species play key roles in the catalytic oxidation process [30, 64, 65]. In addition, the reversible valence transformation of cations (Mn³⁺ ⇋ Mn⁴⁺ + e and Co²⁺ ⇋ Co³⁺ + e) donates some electrons to activate oxygen during the catalytic reaction [42, 43]. Furthermore, the results of in situ DRIFTS demonstrated that the gaseous oxygen could be more easily activated into active adsorbed oxygen species on the surface of α-MnO₂@Co₃O₄ than on that of α-MnO₂. Therefore, the α-MnO₂@Co₃O₄ catalyst with the constructed coupled interface exhibits better catalytic activity for toluene oxidation than α-MnO₂.

In summary, we have designed a bottom-down approach to successfully synthesize ZIF-derived Co₃O₄ grown in situ on a 1D α-MnO₂ material, denoted as α-MnO₂@Co₃O₄. The enhanced catalytic activity over α-MnO₂@Co₃O₄ is attributed to the synergistic effect owing to the coupled interface constructed between α-MnO₂ and Co₃O₄. Compared to the pure α-MnO₂ nanowire and Co₃O₄-b obtained via thermolysis of ZIF-67, the resultant α-MnO₂@Co₃O₄ catalyst exhibits better catalytic activity for toluene oxidation, which is ascribed to the increased number of surface-adsorbed oxygen species, which accelerate the oxygen mobility and enhance the redox pairs of Mn⁴⁺/Mn³⁺ and Co²⁺/Co³⁺. In addition, the in situ DRIFTS result indicates that α-MnO₂@Co₃O₄ possessed a stronger ability to activate gaseous oxygen into adsorbed oxygen species than α-MnO₂. The reaction route over α-MnO₂@Co₃O₄ is as follows: toluene → benzoate species → alkanes containing oxygen functional group → CO₂ and H₂O. Furthermore, the benzoate species are key intermediates in toluene oxidation. Moreover, the α-MnO₂@Co₃O₄ catalyst shows outstanding stability and good water resistance in toluene oxidation. Furthermore, the fabrication method can be extended to other 1D MnO₂ materials. Our work provides a new avenue for the development of high-performance catalysts of practical significance.

Acknowledgments

Quanming Ren thanks for a scholarship from the Overseas Short-term Visiting Program of South China University of Technology.