Boosting Catalytic Combustion of Ethanol by Tuning Morphologies and Exposed Crystal Facets of α-Mn₂O₃

Abstract

Three types of α-Mn₂O₃ catalysts with different well-defined morphologies (cubic, truncated octahedra and octahedra) and exposed crystal facets have been successfully prepared via hydrothermal processes, and evaluated for ethanol total oxidation with low ethanol concentration at low temperatures. The α-Mn₂O₃-cubic catalyst shows a superior catalytic reaction rate than that of α-Mn₂O₃-truncated octahedra and α-Mn₂O₃-octahedra under high space velocity of 192,000 mL/(g·h). Based on the characterization results obtained from XRD, BET, FE-SEM, HR-TEM, FT-IR, H₂-TPR, XPS, ethanol-TPD, and CO-TPSR techniques, the observed morphology-dependent reactivity of α-Mn₂O₃ catalysts can be correlated to the good low-temperature reducibility, abundant surface Mn⁴⁺ and adsorbed reactive oxygen species, which was originated from the exposed (001) crystal planes. Through tuning the morphology and exposed (001) crystal facet of α-Mn₂O₃, a highly active ethanol oxidation catalyst with high selectivity and excellent stability is obtained. The developed approach may be applied broadly to the development of the design principles for high-performance low-cost and environmentally friendly Mn-based oxidation catalysts.

1. Introduction

As atmospheric and photochemical contaminants, Volatile Organic Compounds (VOCs) emitted from fossil fuel combustion, transportation, automobile exhaust, and photochemical pollution have caused tremendous detrimental impact on the environment and human’s health [1,2]. As one of the typical gaseous pollutants among these VOCs, ethanol derived mainly from the residue of unburned ethanol in ethanol-fueled vehicles [3,4]. By comparing different technologies for burning ethanol into CO₂, H₂O, and other less hazardous by-products, the catalytic ethanol combustion has been recognized as one of the most viable and environmentally-friendly technologies at low concentration and lower temperature (200–600 °C) owing to its higher efficiency and yields, and lower cost than that of traditional physical/chemical adsorption and non-catalytic thermal oxidation technologies [5]. Thus, the development of the efficient ethanol combustion catalysts is of great significance in reducing ethanol emissions below the limits imposed by the regulation standards.

In the past few decades, noble metal-based catalysts such as Pt, Pd, Ir, Rh and Au [6,7,8,9,10], etc., have been reported to exhibit high performance of ethanol oxidation at low operation temperatures. However, the high cost of noble metals have caused bottlenecked, which limits the economic viability and impedes the widespread use in ethanol-fueled vehicles. Therefore, it is highly desirable to develop an alternative catalyst to substitute for the noble-metal-based catalyst in order to meet the ever-rigorous ethanol emission standards. Metal oxides, such as Cu/Al₂O₃ [11], CuO/Fe₂O₃ [3], Al/Mn-K [12], CoFe₂O₄ [13], Mn-Ce-Zr-O [14], Mn-Cu [15], etc., are considered as the most promising alternative materials, which have been validated via their good catalytic activity, low cost, resistant to poisonings and higher thermal stability. Catalyst composition, specific surface area, crystal and pore structure have been reported to be the key factors in determining the ethanol combustion efficiencies. Despite its excellent catalytic performance, most of the studies were performed to work the catalysts under lower space velocity, remaining insufficient for practical application.

During recent years, more and more researchers have paid their attention to nano/microsized metal oxides material with different morphologies, such as tube, rods, wires, spheres, cubic, and octahedra, etc. Generally, the catalytic reaction performance of catalyst particles can be finely tuned by their anisotropic morphology, which further results in different exposed crystal facets, to this end, the degree of coordination unsaturation of catalytically active atoms are vital for the correlation between structure and catalytic performance, as reported in the earlier literatures [16,17,18,19,20,21,22]. For instance, Shen et al. [21] investigated that the catalytic oxidation of CO over nanoscale catalytic particles of Co₃O₄ with controlled size and topology, and found that Co₃O₄ nanorod have highly structure-ordered, constituting 40% of the (110) crystal plane on the nanorod surface was able to oxidize CO at −77 °C, showing a 10 times higher activity than conventional Co₃O₄ catalysts. Trovarelli et al. [19] reported that soot combustion over ceria was a surface-dependent reaction and (100) surface for nanocubes; (100), (110) and partially (111) for nanorods, and enclosed octahedral possessed much higher catalytic performance for soot oxidation than ceria polycrystalline powders, which in agreement with the density functional theory (DFT) calculations that the formation energies of ceria surface oxygen vacancies decreased tendency were obtained: (111) < (100) < (110) [23]. Hence, morphology-controlled synthesis of nano/microsized metal oxides further brings up new opportunities for tuning the catalytic activity, selectivity, and stability of metal oxide catalysts via selectively exposing uniform and higher energy/reactive crystal facets.

For transition-metal oxides (TMOs), the Mars-Van Krevelen (MVK) mechanism is proverbially accepted to be responsible for the oxidation of VOCs; the gas-phase organic molecules are oxidized by the active oxygen species of TMOs, then re-oxidized by gas-phase oxygen molecules, which will regenerate or maintain the oxidation state of the metal cations [24,25]. The MVK mechanism is mainly caused by transition-metal cations which possess the ability of electron transport or lattice oxygen mobility for their d or f outer electrons. As an important TMOs, earth-abundant manganese oxides (MnO_x), such as MnO₂, Mn₂O₃, and Mn₃O₄, are considered as promising catalytic materials due to their potentially high catalytic performance, low cost, low toxicity, and durable for the catalytic oxidation of VOCs [1,26,27,28,29,30,31], soot [32,33], or CO [34,35,36]. The superior efficiency catalytic reactivity of MnO_x material mainly related to the existence of various valence states of manganese ions (Mn²⁺, Mn³⁺, and Mn⁴⁺ species) and lattice oxygen on the MnO_x materials, resulting in a facile and reversible Mn³⁺/Mn²⁺ or Mn⁴⁺/Mn³⁺ redox cycle [25,32,37,38,39]. For example, Gandhe et al. [39] revealed that the total oxidation performance of ethyl acetatethe over cryptomelane type octahedral molecular sieve (OMS-2) material was dependent on the existence of Mn⁴⁺/Mn³⁺ type redox couples and facile lattice oxygen on catalysts. Peluso et al. [40] reported that high concentration of Mn³⁺ was beneficial to weakening Mn-O bond and increasing the concentration of active oxygen species, which would enhance catalytic performance of MnO_x in the catalytic oxidation of ethanol. Kim et al. [37] reported that VOCs oxidation over manganese oxides catalysts, including Mn₃O₄, Mn₂O₃ and MnO₂ follows Mn₃O₄ > Mn₂O₃ > MnO₂ could be correlated to the surface area and oxygen mobility of samples. Hence, the multiple valence states and surface reactive oxygen species of MnO_x catalysts are the major determining factors for the catalytic activity of VOCs oxidation.

So far, much research efforts have been devoted to further develop high-performance MnO_x catalysts with excellent reactivity and selectivity by tailoring the shape of the catalysts [20,32,41,42]. Feng et al. [43] reported different manganese oxides with distinct morphologies (1D-Mn₃O₄ nanorod, 2D-Mn₃O₄ nanoplate, and 3D-Mn₃O₄ nano-octahedron) were synthesized by hydrothermal treatment. The Mn₃O₄ nanoplate catalyst exhibits a small crystal size, large surface area, more exposed (112) facets, abundant Mn⁴⁺, and defective structure, contributing to a superior catalytic performance at the high space velocity of 120,000 mL g⁻¹ h⁻¹. Wang et al. [41] reported shape-dependent activation of peroxy monosulfate by single crystal α-Mn₂O₃ (cube, octahedra and truncated octahedra) for catalytic phenol degradation in aqueous solution followed the order of catalytic activity of three α-Mn₂O₃ samples as α-Mn₂O₃-cubic > α-Mn₂O₃-octahedra > α-Mn₂O₃-truncated. Recently, [32] crystal facet-dependent reactivity of α-Mn₂O₃ microcrystalline catalyst for soot combustion was reported, in the rank of α-Mn₂O₃-cubic > α-Mn₂O₃-truncated octahedra > α-Mn₂O₃-octahedra. The origin of the superior performance of cubic α-Mn₂O₃ was correlated with the higher concentration of low-coordinated surface oxygen sites and improved surface redox properties on deliberately exposed (001) crystal facets.

To the best of our knowledge, the ethanol total combustion over different morphologies α-Mn₂O₃ materials with selectively exposed different crystal facets has not yet been investigated in the literature. In the presented work, three types of controllably synthesized three types of α-Mn₂O₃ catalysts with different morphology and exposed crystal facets were prepared by a facile hydrothermal route. The intrinsic properties of α-Mn₂O₃ catalysts are characterized by means of XRD, BET, FE-SEM, HR-TEM, FT-IR, H₂-TPR, XPS, Ethanol-TPD, and CO-TPSR techniques. Kinetic study was also performed to understand the morphology-dependent reactivity of α-Mn₂O₃ catalysts for ethanol total combustion. As demonstrated by various physicochemical characterizations, the morphology of α-Mn₂O₃ catalysts was identified to play a crucial role in exposing the different crystal facets and therefore dictate the catalytic performance of the ethanol total combustion.

2. Results and Discussion

2.1. Crystal Phase Structure and Morphology of Catalysts

The crystal phase of as-prepared manganese oxides catalysts was confirmed by XRD, whose patterns were presented in Figure 1. All catalysts showed the same main characteristic diffraction peaks with a remarkable crystallinity, agreeing well with the previous work [32,41,44,45]. The main characteristic diffraction peaks at 23.1°, 33.0°, 38.2°, 45.2°, 49.3°, 55.2°, and 65.9° (2θ values), corresponding to the (211), (222), (400), (332), (431), (440), and (622) (hkl) planes, can be well-indexed to the body centered cubic phase crystalline structure of α-Mn₂O₃ with the lattice parameter a = 0.9409 nm (JCPDS Card No:00-041-1442). From the diffraction profiles of α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O, one can observe the considerably sharpening of the characteristic diffraction peaks with the increasing relative intensity, indicating the enhanced crystallinity. The crystalline sizes of α-Mn₂O₃-C, α-Mn₂O₃-TO and α-Mn₂O₃-O were estimated to be 24.8 nm, 51.6 nm and 76.1 nm, respectively. No diffraction peaks of other impurities phase were observed via the XRD patterns, further implying that three catalysts possessed high phase purity.

The morphologies and crystallographic microstructures of α-Mn₂O₃ catalysts were investigated by using FE-SEM and HR-TEM. Figure 2 clearly showed that three catalysts possessed the same Mn₂O₃ crystal phase but presented completely different morphologies, including cubic, truncated octahedra, and octahedra, implying the morphologies can be controlled by adjusting hydrothermal temperatures and different solvents. The observations were quite similar to the previous works [32,41]. The α-Mn₂O₃-C presented well-regulated cubic morphology with amiable edges and particle sizes ranging between 0.7 and 2.0 μm (Figure 2a,b). Based on these FE-SEM images (Figure 2c–f), the α-Mn₂O₃-TO and α-Mn₂O₃-O catalysts clearly presented uniform distribution of truncated octahedra and octahedra morphologies with the smooth surface and sharp edges and particle sizes ranging from 1.0–2.2 μm and 0.8–2.5 μm, respectively. No particle with other morphologies was observed in the three samples with cubic, truncated octahedra, and octahedra morphologies (Figure 2a,c,e).

TEM characterizations were performed on various α-Mn₂O₃ catalysts to further investigate the morphologies and crystallographic features i.e., the exposed crystal facets. TEM and HR-TEM images (Figure 3) clearly showed that three α-Mn₂O₃ catalysts presented morphologies of cubic (Figure 3a), truncated octahedra (Figure 3c), and octahedra (Figure 3e), in consistent with the results of FE-SEM and previous reports [32,41]. Figure 3c showed that the lattice fringe of cubic α-Mn₂O₃ (004) facets was 0.23 nm, and α-Mn₂O₃-C mainly exposed the (001) crystal facets as the crystal (001) facets and (004) facets were parallel to each other. Figure 3f showed that α-Mn₂O₃-O mainly exposed the crystal (111) facets which stays parallel to the (222) facets. In addition, the truncated octahedra (Figure 3d) exposed truncated crystal (001) facets and crystal (111) facets, in good agreement with the previous works [32,41,44]. According to the literatures [46,47], crystal growth rates in the direction perpendicular to a high-index plane are usually much faster than those along the normal direction of a low-index plane, therefore high-index planes are rapidly eliminated during particle formation. Li et al. [44] studied the influence of preparation conditions on the morphology control of α-Mn₂O₃ and found out that the domain relied on the different growth rates of [001] and [111] crystallographic directions. Therefore, growing orientation along the [001] and [111] crystallographic directions can form both morphologies of cubic and octahedra with exposed (001) and (111) crystal facets, respectively.

FT-IR spectra of the α-Mn₂O₃ catalysts with different morphologies were shown in Figure S1. It can be seen that the different catalysts presented the analogical positions of characteristic peaks as in literature [32], which located around 481, 529, 572, 660, 1626, 3432 cm⁻¹, respectively. In the fingerprint region, the presence of peaks at around 529 cm⁻¹, 572 cm⁻¹, and 660 cm⁻¹ were assigned to the Mn-O-Mn band asymmetric and symmetric stretching vibration and the peak corresponded to metal-oxygen chains (Mn-O band) bending vibrations in α-Mn₂O₃ [48]. In the functional group region, the weak broad band at around 3432 cm⁻¹ was attributed to the stretching vibration modes of the hydroxyl functional group and the peak at around 1626 cm⁻¹ was mainly caused by the adsorbed water molecules on the α-Mn₂O₃.

2.2. Surface Area and Surface Chemical Properties

The N₂ physical adsorption-desorption isotherms and BJH pore-size distribution patterns of different α-Mn₂O₃ catalysts are shown in Figure S2 and their S_BET, D_p, and V_p are listed in

Table 1. Preparation parameters, BET specific surface areas (S_BET), pore volumes (V_p), Pore diameters (D_p), reduction peak temperatures, H₂ consumptions and surface element composition of three α-Mn₂O₃ catalysts.

Catalyst	SBET a (m2/g)	Vp (cm3/g)	Dp b (nm)	T (°C)		H2 Consumptions (mmol/g)			Surface Element Molar Ratio c
				Peak1	Peak2	Mn2O3→Mn3O4	Mn3O4→MnO	Total	Mn4+/Mn3+	Oads/Olatt
α-Mn2O3-C	30.5	0.212	22.1	287	383	2.5	3.7	6.2	1.27	0.53
α-Mn2O3-TO	2.5	0.009	13.5	409	458	3.6	2.5	6.1	1.15	0.43
α-Mn2O3-O	1.0	0.004	12.9	449	496	3.7	2.3	6.0	1.07	0.38

^a Specific surface areas were calculated by the BET method. ^b The data was calculated via the BJH method according to the N₂ adsorption-desorption isotherms. ^c The ratios were calculated based on the peak areas processed by the XPS-Peak software.

. All catalysts displayed a similar type IV isotherms, with H3 (P/P₀ = 0.6–1.0) hysteresis rings, indicating the presence of microporous and mesoporous in the α-Mn₂O₃ materials. Table 1 showed that α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O catalysts possessed surface area of 30.5, 2.5, and 1.0 m²/g, the pore diameter of 22.1, 13.5, 12.9 nm, pore volume of 0.212, 0.009, and 0.004 cm³/g, respectively. Comparing to α-Mn₂O₃-TO and α-Mn₂O₃-O catalyst, α-Mn₂O₃-C had much higher surface area, pore diameter, and pore volume, which is favorable for the activation and diffusion of the reactants, thereby enhancing the catalytic performance of the catalyst. From the adsorption-condition of N₂ molecules on the surface of porous material at 77 K, the surface area and pore volume of α-Mn₂O₃-TO and α-Mn₂O₃-O catalysts primarily relied on interparticle spaces, thus both catalysts were seen as less developed porous α-Mn₂O₃ catalysts. The morphologies and structures of catalysts would have great influence on pore size distributions, which is likely to be correlated with the distribution of the active sites [24,49]. From FE-SEM images, one can also observed that all catalysts were composed of solid single crystals with smooth planes, which brought about much smaller surface area, pore diameter, and pore volume.

The redox properties of various α-Mn₂O₃ catalysts were examined by using H₂-TPR technique. Figure 4 presented the reduction profiles of the α-Mn₂O₃ catalysts in the temperature range from 100 °C to 600 °C. Two typical peaks can be observed for all catalysts. The different H₂-TPR profiles of three α-Mn₂O₃ samples indicated the different reactivity of reduction of reactive oxygen species in different local environments [50,51]. The α-Mn₂O₃-O catalyst exhibited a large reduction peak centered at about 496 °C with a low temperature shoulder peak at 449 °C, corresponding to H₂ consumptions of 2.3 and 3.7 mmol·g⁻¹, respectively (Table 1). Comparing with α-Mn₂O₃-O catalyst with exposed crystal (111) facets, a similar reduction pattern was observed for the α-Mn₂O₃-TO catalyst with exposed (001) & (111) crystal facets, which had the primary reduction peak at 458 °C and the low temperature shoulder peak n at 409 °C, corresponding to H₂ consumptions of 3.6 and 2.5 mmol·g⁻¹, respectively (Table 1). The alteration of the reduction behavior might be caused by the presence of crystal (001) facets on truncated octahedra α-Mn₂O₃ sample. In the case of α-Mn₂O₃-C catalyst, two distinct lower temperature reduction peaks at 287 and 383 °C were observed, corresponding to H₂ consumptions of 2.5 and 3.7 mmol·g⁻¹, respectively (Table 1). The two reduction peaks of various α-Mn₂O₃ samples corresponds to a stepwise transformation under H₂ atmosphere. The first step peak represented the reduction of Mn₂O₃ to Mn₃O₄, whereas the second step peak corresponded to the reduction of Mn₃O₄ to the final state MnO, similar to the findings reported in the previous literatures [27,32,52,53,54,55,56]. The order of low-temperature reducibility for surface oxygen is α-Mn₂O₃-C with crystal (001) facets > α-Mn₂O₃-TO with crystal (001) & (111) facets > α-Mn₂O₃-O with crystal (111) facets, which is related to the fact that surface O atoms on (001) facet, due to its high flexibility, are energetically more facile for release and participation than those on (111) facet [32], suggesting a more facile surface oxygen reducibility for α-Mn₂O₃-C and morphology-dependence of redox behavior. The excellent reducibility of the α-Mn₂O₃ could be beneficial to the enhanced catalytic activity for the ethanol combustion.

XPS measurements were carried out to identify the surface element compositions, element valence states, and adsorbed oxygen species of α-Mn₂O₃ catalysts. In the XPS survey spectrum presented in Figure S3A, the peaks of carbon (C 1s), oxygen (O 1s), and manganese (Mn 2p) can be clearly detected. In Figure S3B, the binding energy (BE) obtain at 284.6 eV and 287.8 eV obtained were assigned to the C-C (nonoxygenated carbon) and C-O (oxidized carbon) species, respectively. Figure 5 displayed the Mn 2p and O1s spectra for different samples, respectively. The related results of surface element compositions in molar ratio were listed in Table 1. The XPS spectra of Mn 2p presented two contributions, assignable to spin-orbit splitting into Mn 2p_3/2 (641.6 eV) and Mn 2p_1/2 (653.2 eV), respectively. The BE separation between the two main peaks was 11.6 eV [57]. The Mn 2p spectrum of each catalyst could be further resolved into four peak components, attributable to the presence of surface Mn³⁺ species at BE = 641.2 and 652.9 eV and Mn⁴⁺ species at BE = 642.6 and 654.0 eV [58]. According to Table 1, the surface Mn⁴⁺/Mn³⁺ molar ratio of α-Mn₂O₃-C was significantly higher than the other two catalysts, indicating that α-Mn₂O₃-C with (001) crystal facets possessed more surface Mn⁴⁺ species. The differences in surface Mn⁴⁺/Mn³⁺ molar ratios of three catalysts may be explained by the difference in morphologies and the degree of coordination unsaturation of surface active atoms on exposed crystal facets.

As revealed in Figure 5B, the O1s spectra of different α-Mn₂O₃ catalysts were decomposed into two components at BE = 529.6 and 531.2 eV, corresponding to the surface lattice oxygen (O_latt: O²⁻) and O_ads such as O⁻, O₂⁻, O₂²⁻ and OH⁻ [37,58,59,60], respectively. The surface O_ads/O_latt molar ratios of α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O were 0.53, 0.43, and 0.38, respectively (Table 1). Obviously, the amount of surface adsorbed oxygen in α-Mn₂O₃-C with (001) facets is much higher than that of α-Mn₂O₃-TO with (111) & (001) and α-Mn₂O₃-O with (111) facets, implying the important role of exposed (001) crystal facet in oxygen vacancy formation. The increase in surface adsorbed oxygen species might have a significant effect in enhancing catalytic performance for total oxidation reactions mainly because the surface oxygen species had higher mobility than lattice oxygen in the presence of abundant oxygen vacancies [61,62,63].

Oxygen vacancy concentration is crucial in combustion reaction, and oxygen can be chemisorbed on the site of oxygen vacancy to become adsorbed reactive species under reaction atomosphere. Russo et al. [59] have directly detected the existence of oxygen vacancy by Raman characterization, agreeing with the XPS characterization results of surface adsorbed oxygen species.

In order to further unravel the origin of the morphology effect of α-Mn₂O₃, CO-TPSR was also performed by using CO as probe molecule over online MS. The curves of the CO₂ (m/z = 44) signals for three catalysts were displayed in Figure 6. The positions of CO₂ peaks can be related to the different reactivity of the surface reactive oxygen species to oxidize CO. Figure 6 clearly showed the different capability for CO₂ generation by surface active oxygen species over the samples, which was ranked as the decreased order of α-Mn₂O₃-C with (001) facets > α-Mn₂O₃-TO with (001) & (111) facets > α-Mn₂O₃-O with (111) facets. It was evident that theα-Mn₂O₃-C sample with exposed (001) could afford the much more abundant surface reactive oxygen species for CO₂ generation than the other two samples with exposed (111) and mixed (001) & (111) facets, agreeing well with the sequence of the H₂-TPR and XPS results discussed above. The results demonstrated the α-Mn₂O₃-C with (001) facets contains more lattice oxygen and surface oxygen species as activated oxygen species for CO oxidation than (111) facets.

H₂-TPR and CO-TPSR have provided a great deal of practical information on the reactivity of surface oxygen species towards hydrogen molecule. However, the information is indirect as to how strongly the surface and lattice oxygen held in α-Mn₂O₃ catalysts reacts with substrate molecule i.e., ethanol. Ethanol-TPD is a much more direct measure for the determination of surface oxygen species as compared to H₂-TPR. Hence, the catalytic nature of surface adsorbed reactive oxygen species and the desorption behavior of ethanol on the morphology-controlled α-Mn₂O₃ catalysts with different exposed crystallographic facets has been investigated by using ethanol-TPD technique. Figure 7 showed the ethanol-TPD profiles of different morphologies α-Mn₂O₃ catalysts. As the desorption temperature rises, the ethanol reactant as well as some important intermediates including ethanol, acetaldehyde, CO₂, and H₂O were detected on mass spectrometer, corresponding to the signals of m/z = 31, m/z = 29, m/z = 44, and m/z = 18, respectively. It was noticed that the dehydrogenation of ethanol occurred already over three samples in the low temperature range of 40–200 °C, as evidenced by the characteristic peak of acetaldehyde at m/z = 29. TPD profiles for α-Mn₂O₃ catalysts show an apparent morphology-dependent CO₂ desorption profile at temperature above 160 °C. The significantly higher intensity as well as the lowered desorption temperature of CO₂ over α-Mn₂O₃-C catalyst strongly suggest that the preferential exposure of (001) crystal facet by tailoring the catalyst morphology may favor the generation of more abundant surface oxygen reactive species and therefore facilitate CO₂ production.

2.3. Catalytic Performance of Different Shaped α-Mn₂O₃ Catalysts

The morphology-dependent catalytic performances of α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O catalysts were evaluated for ethanol total combustion and the results were shown in Figure 8, Figure S4 and

Table 2. The catalytic activity data of the α-Mn₂O₃ catalysts with different exposed crystal facets for ethanol oxidation under the same reaction conditions: 600 ppm ethanol, 20 vol.% O₂, N₂ as balance gas, SV = 192,000 mL/(g·h).

Catalysts	Catalytic Activity (°C)			T (°C)	Ethanol Conversion (%)	Normalized Rate (mmol·min−1·m−2) × 104	Ea (kJ/mol)	R2 for Ea
	T10	T50	T90
α-Mn2O3-C	90	140	178	90	10.1	2.84	55.3	0.99
α-Mn2O3-TO	151	208	259	90	0.50	1.53	63.7	0.99
α-Mn2O3-O	179	261	314	90	0.10	0.84	68.3	0.98

. Figure 8A shows a clear morphology-dependent catalytic performance for ethanol total combustion with the following order: α-Mn₂O₃-C (190 °C) > α-Mn₂O₃-TO (290 °C) > α-Mn₂O₃-O (340 °C). α-Mn₂O₃-C catalyst with exposed (001) facets exhibited the best catalytic performance on ethanol total oxidation, achieving the complete conversion temperature at 190 °C, which was 150 °C lower than α-Mn₂O₃-O catalyst with exposed crystal (111) facets. The same order of CO₂ yield was found in Figure 8B, exhibiting nearly 100% CO₂ yield at temperature of 240 °C over α-Mn₂O₃-C, 350 °C over α-Mn₂O₃-TO, and 400 °C over α-Mn₂O₃-O catalysts, respectively. Figure 8C showed the volcano-like plot of the acetaldehyde yield over three catalysts, presenting the maximum in acetaldehyde yield of 35.4% at 160 °C over α-Mn₂O₃-C, 45.8% at 230 °C over α-Mn₂O₃-TO, and 42.2% at 300 °C over α-Mn₂O₃-O catalyst, respectively, strongly suggesting that acetaldehyde is the primary intermediate species during ethanol total oxidation, in good accordance with previous work [5,25,28]. It was worth noting that acetaldehyde and CO₂ were the only detected carbon-containing products. On the basis of product distribution observed, we concluded that the cascade reaction pathway of ethanol total combustion via acetaldehyde as important intermediate over different crystal facets of α-Mn₂O₃. Figure 8D and Table 2 further listed the important catalytic data of different α-Mn₂O₃ catalysts with different exposed crystal facets in terms of T₁₀, T₅₀, and T₉₀, which were defined as the values of the reaction temperature corresponding to the ethanol conversions of 10%, 50%, and 90%, respectively. In this study, T₁₀ for α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O catalysts were 90, 151, and 179 °C; T₅₀ were 140, 208, and 261 °C; and T₉₀ were 178, 259, and 314 °C, respectively. In addition, the normalized ethanol combustion rate of α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O at 90 °C were about 2.84, 1.53, and 0.84 mmol·min⁻¹·m⁻²·10⁴, respectively. The superior ethanol combustion activity for α-Mn₂O₃-C catalyst with exposed (001) crystal facets than α-Mn₂O₃ with only (111) crystal facets can be closely related to the low-temperature reducibility and highly active surface mobile oxygen species over the α-Mn₂O₃ surface.

The BET surface area is an important parameter in determining the catalytic performance of heterogeneous catalysts [64,65,66]. Herein, to eliminate the effect of the BET surface areas, the kinetics of ethanol catalytic oxidation were investigated. And Arrhenius plots were obtained over α-Mn₂O₃ catalysts with different exposed crystal facets in the kinetically controlled regime. The apparent activation energy (E_a) was calculated via Arrhenius equation by normalizing reaction rate on per unit BET surface area. The results in Figure 9A and Table 2, showed that α-Mn₂O₃-C catalyst exhibited the much higher initial normalized ethanol reaction rate at low temperature compared with α-Mn₂O₃-TO, and α-Mn₂O₃-O catalysts, in the reverse order of the apparent reaction activation energy. The lower the E_a values indicate the easier oxidation of ethanol, and hence the better performance of a catalyst [1,31,32,34,67]. Under the same reaction conditions, (Table 2) the E_a values of α-Mn₂O₃-C, α-Mn₂O₃-TO, and α-Mn₂O₃-O catalysts were 55.3 kJ/mol, 66.7 kJ/mol, and 68.3 kJ/mol, respectively. The observed morphology effect of α-Mn₂O₃ on the catalytic activity is presumably related to the nature of exposed crystal facets, and density of surface Mn⁴⁺ and surface reactive oxygen species, which are responsible for the adsorption/activation of ethanol and oxygen molecules. Hence, α-Mn₂O₃-C exhibited superior catalytic performance than α-Mn₂O₃-TO and α-Mn₂O₃-O catalysts, and further illustrated cubic with (001) facets was more active followed by octahedra with (111) facets, in consistent with our XPS, CO-TPSR, and ethanol-TPD results discussed earlier.

As documented in the literature, the dissociation of oxygen molecules on can be seen as the first elementary step of catalytic oxidation over metal or oxide surfaces. Therefore, the activation capability of oxygen molecules play an important role in surfaces of oxidation catalyst. For metal oxide catalyst, oxygen molecules can be facilely adsorbed and activated in the presence of oxygen vacancies, forming surface adsorbed reactive oxygen species, which would participate in the oxidation of organic substrates. Recently, a combined experimental/theoretical study has been performed by our group [32] to investigate the morphology–dependent reactivity of α-Mn₂O₃ on soot combustion and gained the origin of the difference on their crystal facet-dependent reactivity. The density functional theory (DFT) calculations revealed that the oxygen atoms on top-layer of crystal (001) facets on α-Mn₂O₃ catalysts can flexibly move with the oxygen atoms of sub-layer. By contrary, such a transfer is energetically unfavorable over the crystal (111) facets of α-Mn₂O₃. Moreover, the formation energy of oxygen vacancy over (001) crystal facet and the coordination number (CN) of surface oxygen atoms is remarkably lower than that on (111) surface. As pointed by An et al. [68], the enhanced coordination unsaturation on specially exposed crystal facet may facilitate oxidation reactions. Accordingly, abundant surface Mn⁴⁺ ions and surface adsorbed reactive oxygen species over (001) surface of α-Mn₂O₃ catalysts may facilitate ethanol/oxygen activation at low temperature and boost ethanol combustion efficiency, in good accordance with our activity and characterization results.

2.4. Effects of SV, Ethanol, and Water Vapor Concentration

The influence of SV, ethanol, and water vapor concentration was investigated over the so-far best-performing α-Mn₂O₃-C catalysts and presented in Figure 10. The effect of SV in Figure 10A showed that the catalytic performance increased as the SV decreased, and the temperatures for complete ethanol oxidation to CO₂ also shifted from 230 °C down to 180 °C. The temperature values corresponding to the maximum acetaldehyde yield also decreased from 160 °C to 150 °C, indicating that the activity of the sample increased as the space velocity decreased. The influence of the initial concentration of ethanol was studied and depicted in Figure 10B. Under the reaction conditions of initial ethanol concentration = 1200 ppm, α-Mn₂O₃-C catalyst showed the catalytic activity of T₁₀ = 99 °C, T₅₀ = 156 °C, and T₉₀= 203 °C which was inferior to those (T₁₀ = 90 °C, T₅₀ = 140 °C, and T₉₀ = 178 °C) obtained from the initial ethanol concentration = 600 ppm as shown in

Table 3. The catalytic performances of the α-Mn₂O₃-C catalyst under different reaction conditions of ethanol oxidation.

Ethanol Con. Tem (°C)	600 ppm Ethanol SV of 96,000 mL/(g·h)	600 ppm Ethanol SV of 192,000 mL/(g·h)	1200 ppm Ethanol SV of 192,000 mL/(g·h)	600 ppm Ethanol, 6 vol.% H2O, SV of 192,000 mL/(g·h)
T10	79	90	99	150
T50	136	140	156	195
T90	172	178	203	218

and Figure 10B. The impact of steam was also investigated over α-Mn₂O₃-C catalyst. Figure 10C clearly showed that the oxidation rate of ethanol could be considerably suppressed by the presence of steam in the feed, in accordance with the inhibition effect of steam on VOCs combustion in the literatures [28,30,69]. This is not unexpected because the strong and competitive adsorption of water vapor on the catalyst surface may lead to a decrease in the available active sites towards VOCs and oxygen, and therefore inhibit ethanol combustion reaction. Under the reaction conditions of 6 vol.% H₂O in the feed, the catalytic activity (T₁₀ = 150 °C, T₅₀ = 195 °C, and T₉₀ = 218 °C) over α-Mn₂O₃-C catalyst became much worse than that under conditions without H₂O (T₁₀ = 90 °C, T₅₀ = 140 °C, and T₉₀ = 178 °C), as shown in Table 3 and Figure 10D.

2.5. The Stability of α-Mn₂O₃-C Catalyst

Figure 11 displayed the ethanol catalytic combustion performance as a function of time on stream over the best-performing α-Mn₂O₃-C catalyst with exposed (001) crystal facets under the conditions of reaction temperature 230 °C, SV 192,000 mL/(g·h), and feed composition of 600 ppm ethanol, 20 vol.% O₂, 6 vol.% H₂O, N₂ as balance gas. Apparently, with the addition of 6 vol.% H₂O, the α-Mn₂O₃-C catalyst exhibited outstanding reaction stability for ethanol oxidation was assessed with experiments lasting at over 230 °C. According to the stability data of successive measurements (Figure 11A), the ethanol conversion, acetaldehyde yield, and CO₂ yield remained almost unchanged for the test duration of more than 50 h, illustrating the excellent stability over the entire run of the stability test. Figure 11B presented almost the same XRD diffraction peaks on the used α-Mn₂O₃-C catalyst as the pattern on the fresh example. Furthermore, the FE-SEM inset in Figure 11B illustrated that the α-Mn₂O₃-C sample remained their original cubic morphology after the test duration of more than 50 h. Based on the stability test, it appeared that α-Mn₂O₃-C could be used as one of the promising manganese based catalysts for efficient ethanol combustion under practical conditions.

3. Materials and Methods

3.1. Catalyst Synthesis

Three types of morphology-tuned α-Mn₂O₃ catalysts, including morphology of cubic, truncated octahedra and octahedra, were synthesized via a low-temperature hydrothermal method according to the literatures earlier [32,41,44,45]. For truncated octahedra and octahedra α-Mn₂O₃ catalysts was prepared by solvothermal method with different solvents. The first catalyst α-Mn₂O₃-octahedra (refer as α-Mn₂O₃-O), for a typical synthesis, Mn(NO₃)₂·4H₂O (16 mmol) was dissolved in ethanol (52 mL) at room temperature (RT) under vigorous magnetic stirring for 20 min to form a homogeneous solution, which then then transferred to a 60 mL Teflon-lined stainless-steel autoclave. The autoclave was tightly sealed and heated for 10 h at 383 K. After hydrothermal reaction, the autoclave was cooled down to RT naturally. The resulting solid product was centrifuged, washed three times in distilled water and ethanol to eliminate impurity ions, and then placed into an oven at 373 K overnight to obtain the α-Mn₂O₃-O catalyst. Similarly, the synthetic method of the second catalyst, α-Mn₂O₃-truncated octahedra (referred to as α-Mn₂O₃-TO), merely changed the solvent replaced by 2-butanol. All catalysts should be calcined in air atmosphere for 2 h at 873 K. The third catalyst, α-Mn₂O₃-cubic (refer as α-Mn₂O₃-C), glucose (6 mmol) was added into a KMnO₄ solution (6 mmol of KMnO₄ was dissolved in 60 mL of distilled water) at RT under vigorous magnetic stirring for another 20 min to form a homogeneous solution, which then was transferred to a 100 mL autoclave, which was tightly sealed and heated for 10 h at 433 K, after natural cooling to RT, immediately followed via centrifugation, washing, and final drying to obtain the precursor MnCO₃. Prior to yielding cubic α-Mn₂O₃ catalyst, the precursor MnCO₃ should be calcined in air atmosphere for 2 h at 873 K with a ramp rate of 5 K/min. Finally, a series of α-Mn₂O₃ catalysts were crushed and sieved to 40–60 mesh for catalytic activity tests.

3.2. Catalyst Characterization

The crystalline structure of catalysts was recorded by the powder X-ray diffraction (XRD) on a Bruker D2 Phaser (Bruker Axs Gmbh, Karlsruhe, Germany) using Cu-Kα radiation. The 2θ of the XRD range from 15^o to 80° with the scan speed during analysis was 0.5 s/step. Field emission-scanning electron microscopy (FE-SEM) surface morphology of the catalyst on Hitachi S-4800 instrument (Hitachi, Ibaraki, Japan). High-resolution transmission electron microscopy (HR-TEM) images of the catalysts were taken on Hitachi JMF-2100 instrument (Hitachi, Japan). The Fourier transform infrared (FT-IR) spectrum of the catalyst was collected by using a Nicolet 380 spectrometer (Nicolet, Madison, WI, USA) in the range of 400–4000 cm⁻¹. The nitrogen adsorption−desorption isotherms were performed on a Micromeritics ASAP 2460 (Micromeritics, Norcross, GA, USA) analyzer at 77 K. By using Brunauer-Emmet-Teller (BET) and Barrett-Joyner-Halenda (BJH) models, specific surface areas (S_BET), pore diameter (D_P) and pore volume (V_P) of the catalysts were calculated from the adsorption branches of the isotherms. Prior to measurement it was degassed at least for 5 h at 393 K. The X-ray photoelectron spectroscopy (XPS) measurements of the catalyst was examined on ESCALAB 250 Xi (Thermo Fisher Scientific, Waltham, MA, USA) using Al-Kα X-ray source. The binding energy scale was corrected for surface charging by use of the C 1s peak of contaminant carbon as reference at 284.6 eV.

Hydrogen temperature-programmed reduction (H₂-TPR) measurements were carried out on a chemisorption analyzer (Micromeritics, Auto Chem II2920, Norcross, GA, USA) instrument, which was equipped with a thermal conductivity detector (TCD) to calibrate the H₂ consumption of metal valence reduction. Before H₂-TPR experiments started, the catalysts (50 mg, 40–60 mesh) were loaded into a U-shaped fixed-bed quartz micro-reactor and pretreated in an Ar atmosphere of 50 mL/min for 1 h at 573 K, and then cooled down to 373 K. The pretreated catalysts were reduced under a mixture-gas of 10 vol.% H₂–90 vol.% Ar flow with 30 mL/min and heated in the range of 373 K to 1173 K with a ramp rate of 10 k/min. Ethanol temperature-programmed desorption (Ethanol-TPD) measurements were carried out via using a Micromeritics AutoChem Ⅱ2920 instrument connected to mass spectrometry (MS) (Hiden, HPR20, Warrington, UK) instrument. The catalysts (50 mg, 40–60 mesh) were loaded into a U-shaped fixed-bed quartz micro-reactor and previously treated for 0.5 h at 373 K in pure He flow with 30 mL/min, cooling down to 313 K. The adsorption experiment of ethanol was carried out with a He flow of 30 mL/min. After saturation, the temperature was increased from 313 K to 873 K with a ramping rate of 10 K/min. The desorption substances coming from α-Mn₂O₃ catalysts were monitored by means of online MS apparatus.

3.3. Temperature-Programmed Surface Reactions (TPSR) and Mixed-Gas without Oxygen

The CO-TPSR measurements(Micromeritics, Auto Chem II2920, America ) to characterize the nature of the active sites of catalysts on α-Mn₂O₃ catalysts were performed in a fixed-bed tubular quartz system (Φ = 6.0 mm) via using MS, where masses (m/e: CO = 28, CO₂ = 44) were monitored. 50 mg of each catalyst were previously treated for 0.5 h at 773 K in a pure Ar flow of 50 mL/min, in order to remove physical adsorption oxygen, cooling down to 373 K. The mixture-gas of 5 vol.% CO-95 vol.% He with a total flow rate of 30 mL/min increasing the temperature from 373 K to 1173 K, and a ramping rate of 5 K/min.

3.4. Catalytic Performance Evaluation

The morphology-dependent catalytic performance of three α-Mn₂O₃ catalysts for ethanol total combustion were evaluated in a flow-through quartz micro-reactor (Φ = 6.0 mm), which was positioned in the center of the tube furnace at atmospheric pressure. The catalytic combustion temperature was monitored via a thermocouple, which was placed in a few millimeters above the bottom of the catalyst fixed-bed. 100 mg (40–60 mesh) of catalyst was held in place by using quartz wool at upper and lower ends [70,71]. The total flow rate of the uniformity mixed-gas (600 ppm ethanol/20 vol.% O₂/N₂ balance gas) passed through the catalyst fixed-bed remained at 320 mL/min, which was detected by using mass flow-meter. The space velocity (SV) of catalytic ethanol combustion was 192,000 mL/(g × h). Prior to catalytic evaluation, to exclude the effect of impurities, the catalysts should be pretreated for 30 min at 333 K in air atmosphere. The outflow gases of the oxidation reaction from the reactor were quantitatively analyzed online via a Gas Chromatograph (Shimadzu, GC-2014C, Kyoto, Japan) which was equipped with two detectors of a thermal conductivity detector (TCD) and a flame ionization detector (FID). And TCD was only responsible for the detection of CO₂, FID responsible for the detection of ethanol and acetaldehyde due to it can respond to almost all organic matter. In the present work, the final products of ethanol total combustion were CO₂ and H₂O. During the ethanol oxidation, the main products were acetaldehyde, CO₂ and other remaining trace products which can be ignored. The effect of performance for SV, ethanol and water vapor concentration were explored. The thermal stability of α-Mn₂O₃ catalyst was tested for 50 h under the condition of 6.5 mg catalyst was diluted with 100 mg of SiO₂, 600 ppm ethanol, 6vol.% H₂O, 20 vol.% O₂, N₂ as the equilibrium gas, and SV was 192,000 mL/(g × h). The conversion of ethanol was calculated based on the participation of ethanol as follows (1):

$$\begin{array}{c}\mathrm{X}=\left({(\mathrm{C}}_{\mathrm{acetaldehyde}\mathrm{out}}+\frac{{\mathrm{C}}_{{\mathrm{co}}_2\mathrm{out}}}{2})/({\mathrm{C}}_{\mathrm{acetaldehyde}\mathrm{out}}+\frac{{\mathrm{C}}_{{\mathrm{co}}_2\mathrm{out}}}{2}+{\mathrm{C}}_{\mathrm{ethanol}\mathrm{out}})\right)\text{×}100\%\\ {\mathrm{S}}_{{\mathrm{co}}_2}=\left(\frac{{\mathrm{C}}_{{\mathrm{co}}_2\mathrm{out}}}{2}/({\mathrm{C}}_{\mathrm{acetaldehyde}\mathrm{out}}+\frac{{\mathrm{C}}_{{\mathrm{co}}_2\mathrm{out}}}{2}+{\mathrm{C}}_{\mathrm{ethanol}\mathrm{out}})\right)\times 100\%\\ {\mathrm{S}}_{\mathrm{acetaldehyde}}={\mathrm{C}}_{\mathrm{acetaldehyde}\mathrm{out}}/\left({\mathrm{C}}_{\mathrm{acetaldehyde}\mathrm{out}}+(\frac{{\mathrm{C}}_{{\mathrm{co}}_2\mathrm{out}}}{2})\right)\times 100\%\end{array}$$

$${\mathrm{Y}}_{\mathrm{yield}}=\mathrm{conversion}\times \mathrm{selectivity}\times 100\%,$$

(1)

where X, ethanol conversion rate; C, concentration; S, selectivity of products; Y, yield of products.

3.5. Kinetic Analysis for Ethanol Total Oxidation

To obtain kinetics data, the kinetic research was performed under ethanol conversion below 15% with the condition of 600 ppm ethanol, 20 vol.%O₂, N₂ as the equilibrium gas, and SV was 192,000 mL/(g·h) for three α-Mn₂O₃ catalysts in the range of 333 K to 383 K. The catalytic activation energy (E_a) of the catalysts were calculated according to the Arrhenius Equation (2):

$$\mathrm{lnR}=-\mathrm{A}\exp (-\frac{{\mathrm{E}}_{\mathrm{a}}}{\mathrm{RT}}),$$

(2)

where E_a, apparent activation energy (kJ/mol); R, the reaction rate (mol·min⁻¹·m⁻²); T, the value of the reaction temperature (K).

4. Conclusions

In summary, a series of α-Mn₂O₃ catalysts with different morphologies, including cubic, truncated octahedra and octahedra, were controllably synthesized by using a facile hydrothermal process and systematically investigated for their morphology-dependent catalytic performance in ethanol total combustion. The catalytic activity according to each manganese oxide catalyst was in the order of α-Mn₂O₃-C > α-Mn₂O₃-TO > α-Mn₂O₃-O. The present study indicated that this morphology-dependent reactivity of α-Mn₂O₃ nanocrystal was originated from the chemical nature of the exposed (001) facets. As revealed by characterization results of HR-TEM, H₂-TPR, XPS, CO-TPSR, ethanol-TPD, the superior activity of α-Mn₂O₃-C sample can be well correlated with its enhanced low temperature reducibility, abundant surface Mn⁴⁺ species and surface reactive oxygen species governed by the specially exposed (001) facets. Moerover, the effect of space velocity and feed compsotion (ethanol and steam) were also investigated on α-Mn₂O₃-C catalyst under different reaction conditions. Furthermore, the α-Mn₂O₃-C catalyst exhibited good stability for 50 h at 230 °C in the presence of 6 vol.% H₂O, demonstrating that α-Mn₂O₃-C catalyst could be used as promising catalyst for efficent practical total oxidation of ethanol. This study presents a new strategy to design and develop the catalyst for efficient combustion of ethanol by morphological control of earth-abundant inexpensive Mn-based catalysts.