Ordered micro/macro porous K-OMS-2/SiO2 nanocatalysts: Facile synthesis, low cost and high catalytic activity for diesel soot combustion

A series of novel oxide catalysts, which contain three-dimensionally ordered macroporous (3DOM) and microporous structure, were firstly designed and successfully synthesized by simple method. In the as-prepared catalysts, 3DOM SiO2 is used as support and microporous K-OMS-2 oxide nanoparticles are supported on the wall of SiO2. 3DOM K-OMS-2/SiO2 oxide catalysts were firstly used in soot particle oxidation reaction and they show very high catalytic activities. The high activities of K-OMS-2/SiO2 oxide catalysts can be assigned to three possible reasons: macroporous effect of 3DOM structure for improving contact between soot and catalyst, microporous effect of K-OMS-2 for adsorption of small gas molecules and interaction of K and Mn for activation of gas molecules. The catalytic activities of catalysts are comparable to or even higher than noble metal catalyst in the medium and high temperature range. For example, the T50 of K-OMS-2/SiO2-50, 328 °C, is much lower than those of Pt/Al2O3 and 3DOM Au/LaFeO3, 464 and 356 °C,respectively. Moreover, catalysts exhibited high catalytic stability. It is attributed to that the K+ ions are introduced into the microporous structure of OMS-2 and stabilized in the catalytic reaction. Meanwhile, the K+ ions play an important role in templating and stabilizing the tunneled framework of OMS-2.


Results
Structural Features of As-prepared Catalysts. The XRD patterns results of as-prepared catalysts are shown in the Fig. 1. The Fig. 1a shows the diffraction peaks of pure SiO 2 , which the 23.5° is corresponding to the amorphous SiO 2 . As shown in Fig. 1b-h, the feature peaks, which 2θ degree are located at 12.6°, 28.6°, 37.4°, 41.6°, are corresponding to cryptomelane-M (OMS-2) with a space group of I2/m(12) (JCPDS Card No.  and they increased with increasing of K-OMS-2 loading amounts (as "▲ " marked in the Fig. 1b-h). Figure 1i suggests that the cryptomelane-M is also synthesized on the surface of silica gel support. Different active components supported on 3DOM SiO 2 show various feature peaks. The feature peaks (as "■ " marked in the Fig. 1j) could be readily indexed to manganese oxides 13 . Those feature peaks, which marked by the "★ " in the Fig. 1k, are corresponding to the diffraction peaks of KNO 3 . Different XRD patterns of as-prepared catalysts indicate that the potassium has been doped into the tunnel structure of OMS-2. The above results illustrate that we have successfully prepared a novel ordered micro/macro catalyst, which contains 3DOM SiO 2 and microporous structure in K-OMS-2 NPs.
The SEM images (Fig. 2) clearly demonstrate that all 3DOM K-OMS-2/SiO 2 catalysts have long range ordered macroporous with average diameter of ca. 310 ± 20 nm. As shown in the Fig. 2b-h, more and more visible nanoparticles can be obtained on the skeleton of 3DOM SiO 2 when the loadings of K-OMS-2 NPs are increased. Especially, when the loading of K-OMS-2 is higher than a certain value (K-OMS-2/SiO2-60), the surface of 3DOM SiO 2 is completely covered by K-OMS-2 NPs and the pores of 3DOM SiO 2 were also filled (Fig. 2h). Combined with the XRD results ( Fig. 1a-h), the K-OMS-2 NPs, which supported on the surface of 3DOM SiO 2 , would be contributed to enhancing the intensity of diffraction peaks for cryptomelane-M. In order to compare the influence of different active components and supports, the SEM images of K-OMS-2/silica gel, MnO x /SiO 2 -50 and KNO 3 /SiO 2 -50 are also exhibited in the Supporting information. As shown in the Fig. S2, K-OMS-2/silica gel is constituted by nanoparticles (Fig. S2a). SEM images of MnO x /SiO 2 -50 and KNO 3 /SiO 2 -50 ( Fig. S2b and c) indicate that 3DOM SiO 2 is stable when different active components supported on the surface of 3DOM structure.
TEM images of 3DOM K-OMS-2/SiO 2 catalysts are shown in the Fig. 3. From the Fig. 3a, no any K-OMS-2 NPs are covered on the skeleton of 3DOM SiO 2 . However, as shown in Fig. 3b-h, the surface of 3DOM SiO 2 is successfully decorated with well-dispersed K-OMS-2 NPs and no larger agglomerated particles is observed. The average K-OMS-2 NPs sizes are estimated to be 20-25 nm for 3DOM K-OMS-2/SiO 2 catalysts with different K-OMS-2 loadings (see in the Fig. S5). The densities of K-OMS-2 NPs on the skeleton of 3DOM SiO 2 are increased with the enhancement of loading dosages of active components. TEM images of Fig. 3g and h exhibit that the surface of 3DOM SiO 2 is completely covered by K-OMS-2 NPs, which is agreed with the SEM results (Fig. 2). Figure S3 shows TEM images of catalysts with different supports and active components. It indicates that the silica gel is composed of SiO 2 particles, and no obvious particles of K-OMS-2 are observed on the silica gel (Fig. S3a). The reason for this phenomenon is that the precursor solution of KNO 3 and Mn(NO 3 ) 2 is infiltrated into the inner of silica gel through accumulation pores. Figure S3b shows the TEM image of MnOx/SiO 2 -50. The surface of 3DOM SiO 2 is covered by nanoparticles of MnO x , which the particle size is about 22 nm (Fig. S5h). As shown in the Fig. S3c, no particles of KNO 3 appeared on the surface of 3DOM SiO 2 .
The HRTEM images of as-prepared catalysts with different supports and active components are shown in Fig. 4. HRTEM image of K-OMS-2/SiO 2 -50 shows that the lattice fringes of K-OMS-2 NPs are clearly observed (Fig. 4a), e.g. the interplanar spacing is measured to be 0.463 nm indexed as (002) planes of K-OMS-2. As shown in the Fig. S4a-g, clear lattice fringes can be observed on the K-OMS-2/SiO 2 with various K-OMS-2 loadings. The lattice fringes are corresponding to the crystal of K-OMS-2. Those results indicate that K-OMS-2 nanoparticles contain a microporous structure (0.46 nm) arising from edge sharing of 2 × 2 [MnO 6 ] octahedral chains to form one-dimensional tunnel structures (as shown insert of Fig. 4a) 20 . Because the tunnel cavity of OMS-2 is as large as 0.46 nm, the K + cations are inevitably introduced into the tunnel and stabilized in the catalytic reaction. From the Fig. 4b, the irregular K-OMS-2 particles are appeared on the surface and inner of silica gel. And the lattice fringes was measured to be 0.465 nm indexed as (002) planes of K-OMS-2. As shown in the Fig. 4c, the distinct lattice fringes (0.489 nm) are assigned to the (200) planes of manganese oxides. Figure 4d shows an interesting result of KNO 3 /SiO 2 -50, no particles or sheets of KNO 3 can be observed on the surface of SiO 2 , indicating that the KNO 3 may be adhered on the surface of SiO 2 with molten state during the calcined process. After finishing of calcined process, the KNO 3 are uniform distributed on the skeleton of SiO 2 .
In order to get the distributed information of K, Mn, O and Si elements in the as-prepared catalysts, the HAADF-STEM image and EDS elemental mapping of K-OMS-2/SiO 2 -50 are shown in the Fig. 5. The EDX elemental mappings suggest that the O and Si are uniformly distributed in the in the α orts of 3DOM SiO 2 . Due to O element in the K-OMS-2 NPs, the O element can be found more riches in the K-OMS-2 NPs while the Si element cannot be obviously observed in the K-OMS-2 NPs. As shown in the Fig. 5, the Mn element is aggregated in the K-OMS-2 NPs owing to Mn is the main components for K-OMS-2 NPs. In addition, the K element is also focused on the K-OMS-2 NPs. However, the intensity of K element is seemed weaker than corresponding that of Mn element. The possible reason for this phenomenon is that the K elements are doped into the one-dimensional tunnel structures of 2 × 2 [MnO 6 ] octahedral chains. TGA-DSC Analyses of As-prepared Catalysts. The weight loss and heat flow of K-OMS-2/SiO 2 -50 samples (none calcined and calcined) observed during thermogravimetry-differential thermal analyses (TGA-DSC) in an oxygen atmosphere are displayed in Fig. 6. As shown by the TGA curve of none calcined K-OMS-2/SiO 2 -50 ( Fig. 6A), weight losses occurred in three temperature ranges of 50-250, 250-600 and 600-1000 °C. The first significant weight loss of 31.3% is mainly attributed to the decomposition of KNO 3 and Mn(NO 3 ) 2 . In this process, the K and Mn ions are formed impermanent species of K-OMS-2, so that the NO 3 − in the KNO 3 is separated and  decomposed. Certainly, desorption of physisorbed and chemisorbed water is also contributed to the first significant weight loss. The second significant weight loss of 4.2% in the temperature range of 250-600 °C is generally due to further dehydration, lattice oxygen loss or part nitrate decomposition of K-OMS-2. The third slight weight loss of 2.1% is started at 600 °C and ended at 1000 °C. In this temperature range, the collapse and the change of K-OMS-2 to hausmannite by the transformation of Mn 4+ to Mn 3+ are the possible reasons for lattice oxygen release 30 . As you can see from the DSC curve of none calcined K-OMS-2/SiO 2 -50, a small exothermic peak appeared in the range of 120-180 °C with the highest exothermic value at ca. 160 °C. This exothermic peak reveals the thermal decomposition of nitrate, which came from the preparation process. With increasing of temperature, as a result of dehydration and crystal phase transformation, an endothermic process appears in the DSC curve when the temperature is over 250 °C. Another small exothermic peak at 830 °C is corresponding to crystal change of K-OMS-2 31 . Figure 6B shows the TGA-DSC curves of calcined K-OMS-2/SiO 2 -50. Compared with TG curve of none calcined K-OMS-2/SiO 2 -50, the calcined K-OMS-2/SiO 2 -50 has less weight loss owing to decomposition of nitrate in the calcination process. The weight loss also could be divided into three parts in the range of 50-1000 °C. The temperatures of the first weight loss are in the range of 50-250 °C. Because the nitrate has been decomposed in the calcination process, the first weight loss about 2.4% can be attributed to desorption of physisorbed and chemisorbed water. The second weight loss is occurred in the temperature range of 250-600 °C and the ratio of weight loss is 0.5%. Similarly, further dehydration and lattice oxygen loss of K-OMS-2 may be contributed to this weight loss. The third weight loss is 2.9% and temperature range is 600-1000 °C 32 . The reasons for this weight loss are similar to that of none calcined K-OMS-2/SiO 2 -50. As shown in Fig. 6B, compared with DSC curve in the Fig. 6A, the DSC curve of calcined K-OMS-2/SiO 2 -50 indicates that no exothermic peak appears in the range of 120-180 °C due to no nitrates in the calcined sample. However, the curve of calcined K-OMS-2/SiO 2 -50 is the same as that of none calcined K-OMS-2/SiO 2 -50 when the temperature is over 250 °C. An overview of TG-DSC characterizations indicates that the as-prepared catalysts are stable under the calcination temperature of 600 °C. If the calcination temperature is over 600 °C, the crystal form of K-OMS-2 would be changed. The content of exchangeable active oxygen in the range of 250-600 °C may affect the physicochemical performances of as-prepared catalysts.
Raman spectra of As-prepared Catalysts. In an effort to investigate the influence of K doping on the microscopic structure and vibrational properties of MnO x , Raman spectra of the K-OMS-2/SiO 2 with different K-OMS-2 loading amounts and 3DOM SiO 2 with different active components were measured, and the results are shown in Fig. 7. For the pure 3DOM SiO 2 (Fig. 7a)    H 2 -TPR Results of As-prepared Catalysts. Catalytic combustion of soot is a complicated gas-solid (soot)-solid (catalyst) multi-phase reaction. The intrinsic redox properties of catalysts play a key role in the combustion of soot. Therefore, temperature-programmed reduction (TPR) by H 2 was used to measure these characteristics in the present work. As shown in Fig. 8, the peak position and types are similar for 3DOM K-OMS-2/ SiO 2 with different K-OMS-2 loadings, while the intensity of H 2 consumption increases with the increasing of K-OMS-2 loading. For 3DOM SiO 2 (Fig. 8a), no reduction peak is observed. In the H 2 -TPR profiles of 3DOM K-OMS-2/SiO 2 catalysts, there are three overlapping peaks (ranging from 250 to 700 °C), corresponding to a three-step reduction process. Reduction peaks are observed in three temperature ranges of 250-380 °C, 380-500 °C and 500-650 °C. Assuming that MnO is the final state in the reduction of OMS-2 Mn species 38 . The peak at 200-350 °C could be assigned to the reduction of MnO 2 /Mn 2 O 3 to Mn 3 O 4 , and the peak at 380-500 °C may be assigned to the reduction of Mn 3 O 4 to MnO. The results indicate that substantial amount of Mn 4+ and Mn 3+ in K-OMS-2 can be reduced to Mn 2+ below 500 °C, which is consistent with the previous reports 39 . The third reduction peak at 500-650 °C may be assigned to the reduction of K species. When the mass ratio of KNO 3 to SiO 2 is over 25%, the intensity of K species reduction peak is obviously observed, and increases with increasing of K-OMS-2 loading amounts. The H 2 -TPR profiles of K-OMS-2/silica gel, 3DOM MnO x /SiO 2 -50 and KNO 3 /SiO 2 catalysts are presented in Fig. 8B. The peak position and types vary among the catalysts indicate different redox performance of as-prepared catalysts. Because of poor redox performance of SiO 2 supports and the same active components, the reduction peaks of K-OMS-2/silica gel-50 ( Fig. 8i) are similar to that of 3DOM K-OMS-2/SiO 2 -50. As for MnO x /SiO 2 -50, there are two main peaks at 309 °C and 423 °C, while no reduction peak is observed at relatively high temperature. It is apparent that the TPR peaks of 3DOM MnO x /SiO 2 -50 are present at lower temperature, indicating higher reducibility of manganese in MnO x than the corresponding 3DOM K-OMS-2/ SiO 2 -50. The TPR profile of KNO 3 /SiO 2 -50 catalyst presents two important reduction signals at 530-720 °C and 720-830 °C assigned to the potassium nitrate reduction. The potassium nitrate can be reduced with H 2 to generate KNO 2 , NO and NH 3 . The presence of the H 2 consumption peak for KNO 3 /SiO 2 -50 indicates the presence of main KNO 3 species on the surface of 3DOM SiO 2 even after calcination at 550 °C. It is in good agreement with the XRD results of 3DOM KNO 3 /SiO 2 -50 (Fig. 1k).
XPS spectra of As-prepared Catalysts. XPS studies are conducted to gain insight into the oxidation state, surface composition, and atomic environment of K, Mn, O and Si species in the different samples, and the results are shown in Fig. 9 and Table 1. As shown in Fig. 9A, the K 2p spectrum consists of a spin-orbit split doublet composed of two peaks with an intensity ratio between K 2p 3/2 and K 2p 1/2 at about 2:1. The binding energies (BEs) of 292.4-292.8 eV and 295.2-295.6 eV are assigned to K 2p 3/2 and K 2p 1/2 (Table S2), respectively, which the separation of K 2p 3/2 and K 2p 1/2 peaks is about 2.6 eV 40 . The various binding energies of K indicates that environment of K is different in the as-prepared catalysts. Figure 9B displays the Mn 2p XPS of as-prepared catalysts. The Mn 2p spectra are significantly broadened and show some asymmetry towards both Mn 2p 3/2 and Mn 2p 1/2 peaks. Due to this feature, it is worth noting that the binding energies of components of as-prepared catalysts are in good agreement with the literature data reported for Mn 3+ and Mn 4+ 41 . As shown in Fig. 9B, the binding energies of the XPS Mn 2p 3/2 peaks are found to be in the range 640.0-645.0 eV. Two kinds of Mn species including Mn 3+ (ca. 641.5 eV), and Mn 4+ (644.5 eV) are presented on the surface of as-prepared catalysts. Meanwhile, the Mn 2P 1/2 peak also shows two kinds of Mn species in the BEs range of 650-655 eV (Table S2). The detailed peak fitting results of Mn 2p features for the two Mn ions are listed in Table 1. Apparently, the K-OMS-2 loadings have a slight impact on the surface Mn 3+ /Mn 4+ molar ratio. Table S2 indicates that the binding energies of Mn element decreased with increasing of K-OMS-2 loadings. The small change of BEs of Mn indicates the electronic densities of Mn atoms increase, suggesting that the low-valence Mn species increased 42 . Quantitative analysis of the surface components also proves that Mn species in the high loading of K-OMS-2 have lower average oxidation state than that of low loading catalyst 43 . Among those catalysts, because of the same K-OMS-2 loadings, the K-OMS-2/ SiO 2 -50 and K-OMS-2/silica gel-50 show similar surface Mn 3+ /Mn 4+ molar ratio (1.93 and 1.97, respectively), whereas MnO x /SiO 2 -50 is the lowest in terms of surface Mn 3+ /Mn 4+ molar ratio (1.65). More Mn 3+ sites may be originated a weak Mn-O bond and formed more active oxygen species, which enhanced catalytic activity for redox reaction 42 .
The corresponding spectra of oxygen species are presented in  Catalytic Performances for Soot Combustion. The catalytic performances of 3DOM K-OMS-2/SiO 2 catalysts for soot oxidation were evaluated and the results are listed in Table 2. For comparison, the soot combustion reactions of without catalyst and over 3DOM SiO 2 were also estimated under the same reaction conditions. For the pure soot, the T 10 , T 50 and T 90 are 482, 564 and 609 °C, respectively. 3DOM SiO 2 also shows somewhat catalytic activity for soot combustion, and the T 10 , T 50 and T 90 are 354, 503 and 550 °C, respectively. This result indicates that the 3DOM structure of SiO 2 may enhance the contact area between soot and reaction gas. 3DOM K-OMS-2/SiO 2 catalysts show high catalytic activities for soot combustion. The catalytic activity enhanced with increasing of K-OMS-2 loading and reached maximum at a certain value (K-OMS-2/SiO 2 -50). Further increasing the K-OMS-2 loadings, the catalytic activity almost keep constant. Compared the Δ T 10 , Δ T 50 and Δ T 90 of K-OMS-2/SiO 2 catalysts in the Table 2, the Δ T 10 of 3DOM K-OMS-2/SiO 2 -50 catalyst is 199 °C, which is the highest among the 3DOM K-OMS-2/SiO 2 catalysts (i.e. the K-OMS-2/SiO 2 -50 catalyst has the lowest initiation temperature). As shown in Table 2, the temperature of T 90 decreases with increasing mass ratio of KNO 3 and SiO 2 and it is lower than 400 °C when the K loading is reached a certain value (K-OMS-2/SiO 2 -30). Because of temperature range of 175-400 °C for diesel exhaust, soot particles over the K-OMS-2/SiO 2 catalysts can be completely burnt off when the diesel engines are normal running (i.e. the temperature of diesel exhaust is ca. 400 °C). In addition, the K-OMS-2/SiO 2 catalysts show higher CO 2 selectivity for soot combustion than that of  Table 2. Catalytic performances of as-prepared catalysts for soot combustion. ∆ T 10 : The difference value of T 10 between pure soot and catalysts, ∆ T 50 : The difference value of T 50 between pure soot and catalysts, ∆ T 90 : The difference value of T 50 between pure soot and catalysts.
pure soot combustion. All surpassed 90%. The selectivity to CO 2 increased with increasing of K-OMS-2 loading and the CO 2 selectivity value of K-OMS-2/SiO 2 -50 catalyst is as high as 96.7%. In order to deeply investigate the interaction effect (K and Mn) and macropore effect for soot combustion, the catalytic activities of K-OMS-2/ silica gel-50, 3DOM MnO x /SiO 2 -50 and 3DOM KNO 3 /SiO 2 -50 were tested and the results are shown in Table 2.
Compared with 3DOM K-OMS-2/SiO 2 -50, K-OMS-2/silica gel-50 shows much lower catalytic activity indicating that the 3DOM structure is favorable for soot combustion. It is attributed to that K-OMS-2 may be infiltrated into the inner of silica gel through accumulation porous (see the TEM image of Fig. S3), so that most of K-OMS-2 contact soot particle more difficultly than 3DOM K-OMS-2/SiO 2 -50. 3DOM MnO x /SiO 2 -50 and KNO 3 /SiO 2 -50 show higher activity than K-OMS-2/silica gel-50 but lower than 3DOM K-OMS-2/SiO 2 -50, indicating that the interaction effect between K and Mn has positive influence on soot combustion. In order to compare the catalytic performances of 3DOM as-prepared catalysts, the catalytic activities of previous other catalysts reported in literature are summarized in Table 3. When a comparison is made on catalysts for soot combustion, it must be noted that experimental conditions may be deviated from each other concerning reaction gas, catalyst dosage and measurement system. Thus, fine activity comparison of different catalysts is usually challenging. Nevertheless, a rough comparison, including the temperature for soot combustion, is still feasible. As shown in Table 3, the 3DOM K-OMS-2/SiO 2 -50 catalyst showed lower temperature for T 50 than the previous reported catalysts except the 3DOM Pt 0.08 /Ce 0.8 Zr 0.2 O 2 . From the Table 3, the K-OMS-2/SiO 2 -50 gives the lowest T 90 , i.e., the highest catalytic activity than other 3DOM metal oxides at high reaction temperature (i.e. comparison of T 90 ). In addition, the T 90 of K-OMS-2/SiO 2 -50 is much lower than 400 °C, indicating that the soot can be completely burnt off blew 400 °C. The temperature for complete catalytic combustion of soot is lower than the highest temperature of exhaust (175-400 °C), which is important for practical application. More importantly, the 3DOM K-OMS-2/SiO 2 catalysts were prepared by simple method and took very cheap KNO 3 and Mn(NO 3 ) 2 materials as active components. Obviously, non-noble metal catalysts show similar or even higher catalytic activities than noble metal catalysts in the medium and high temperature rage (see the Table 3). However, the as-prepared catalysts show lower catalytic activities than noble metal catalysts in the low temperature rage (temperature of T 10 ). The reason for this phenomenon is that the K-OMS-2 NPs have relatively lower redox property than noble metal catalysts for activation of oxygen species at low temperature (< 300 °C). When the temperature is over 300 °C, a mass of oxygen can be activated and then the catalytic activities show great enhancement. This result is also well agreed with characterization of H 2 -TPR (Fig. 8). Considering those factors, the catalytic activity of current 3DOM K-OMS-2/ SiO 2 catalysts are significantly enhanced for soot combustion especially K-OMS-2/SiO 2 -50 (T 50 and T 90 at ca. 328 °C and 363 °C). Therefore, 3DOM K-OMS-2/SiO 2 catalysts are promising for the practical application due to high catalytic activity and low cost.

Catalytic Performance of K-OMS-2/SiO 2 -50 with Different Concentrations of NO in the Reactant Gases.
In this work, in order to test and research the catalytic performance of as-prepared catalysts more deeply, the catalytic activities of K-OMS-2/SiO 2 -50 with different concentrations of NO were also measured and the results are displayed in the Fig. 10. As shown in Fig. 10A, the profiles of CO 2 concentration over the K-OMS-2/SiO 2 -50 show different tendencies under the condition of different concentration of NO. In the absence of catalyst, the temperature of highest CO 2 concentration is located at 580 °C. However, the temperature shifts to 450 °C when catalyst is presented. Moreover, the CO 2 profiles are higher and narrower than the case without NO. It suggested that the catalytic activities of K-OMS-2/SiO 2 are strongly affected by the NO gas. In order to more clearly describe the differences of catalytic activities under various concentrations of NO, the T 10 , T 50 , T 90 and CO 2 selectivity are calculated and shown in Fig. 10B Table 3. Comparison between reported catalysts in referees and as-prepared catalyst for soot combustion under loose contact conditions. a mass of soot. b mass of catalyst.
result without NO is also included in Fig. 10B. Its T 10 , T 50 and T 90 are 356, 431 and 490 °C, respectively. The result indicates that NO gas plays an important role in the soot combustion when its concentration is located at low value while high concentration has less assistance. As the same as reported in the previous literature 23 , NO acts as an efficient mobile oxidizing agent, and the NO could be oxidized to NO 2 by oxygen during the reaction when catalysts are used. Since the oxidizing ability of NO 2 is stronger than that of O 2 , the removal of soot particles by NO 2 is one of the main ways in soot combustion. Therefore, the soot combustion subtly changed solid (soot)-solid (catalyst) contact into solid (soot)-gas (NO 2 )-solid (catalyst) contact. This altering reaction path greatly promotes soot combustion so that the catalysts show higher activities than that without NO 16 . More deeply explains will discuss in the following discussion.

Stability of 3DOM K-OMS-2/SiO 2 -50
Catalyst. The stability of catalysts is one of the most important performances, especially in practical application. Therefore, the stability of 3DOM K-OMS-2/SiO 2 -50 catalyst in five cycles running was examined and the results are shown in Fig. 11. 3DOM K-OMS-2/SiO 2 -50 maintained high catalytic activity and CO 2 selectivity after five-cycle reaction under the condition of loose contact between catalysts and soot particles, the numerical values of T 10 , T 50 and T 90 were 283 ± 5, 328 ± 6 and 363 ± 5 °C, respectively.  Meanwhile, the CO 2 selectivity value was more than 95% after reaction for five cycles. In this work, the 3DOM structure of SiO 2 support and K-OMS-2 nanoparticles play important roles in the catalytic combustion of soot. The used 3DOM K-OMS-2/SiO 2 -50 catalyst was characterized by measurements of SEM, TEM and XRD and the results are listed in the Fig. 12. According to the SEM and TEM images of used 3DOM K-OMS-2/SiO 2 -50 catalyst, 3DOM structure is not destroyed after five cycles reaction, and the mean size (23 ± 6 nm) of K-OMS-2 nanoparticles on the surface of 3DOM SiO 2 support is not remarkably changed in comparison with that of the fresh sample (24.25 ± 3.37 nm). The XRD similarity between used and fresh catalysts suggests that the crystal form and one-dimensional tunnel structure (0.46 nm) of K-OMS-2 are stable during the process of soot combustion. The characterization results of used catalyst indicate that 3DOM K-OMS-2/SiO 2 catalysts have good thermal stability and activity stability during soot combustion reaction.

Discussion
Macropores Effect of 3DOM SiO 2 . It is well-known that the catalytic combustion of soot is a gas-solidsolid reaction, which the contact between soot and catalysts is one of the most important factors for improving the catalytic performance. In fact, many studies have demonstrated that the tight contact between soot and catalysts shows excellent catalytic performances for soot combustion 23 . However, loose contact between soot particles and catalysts is a typical way of contact in the process of after-treatment for diesel engine exhaust. Therefore, it is extremely important to study and design the active catalysts, which can improve the contact efficiency between the catalysts and soot particles under loose contact conditions. In order to enhance the contact efficiency, 3DOM materials with uniform pore size and periodic voids interconnected through open windows are designed and synthesized by colloidal crystal template method. As shown in Fig. S1a, the average diameter of ordered macroporous is about 310 nm and diameter of interconnected open windows is about 100 nm. Soot particles could easily across those pores and reach the active sites on the inner wall of 3DOM materials with the help of gas-flow Thus, 3DOM structures could enhance the contact between soot particles and catalysts. In this work, in order to demonstrate the contact efficiency, the soot and 3DOM K-OMS-2/SiO 2 -50 catalyst was studied under the same conditions for TPO reaction. In this confirmatory experiment, the reaction temperature of soot and 3DOM catalyst was programmed to 280 °C, which means the soot was not ignited. As we all know, the catalytic soot combustion is a gas-solid-solid reaction, which soot particles can enter the inner pores of 3DOM catalyst with the help of the reaction gas flow (O 2 , NO and Ar) during the reaction process. More importantly, the rising reaction temperature may contribute to separating the agglomerate soot particles. Under the influence of gas flow and rising temperature, the soot particles can easily enter into the 3DOM structure and contact the inner active sites of 3DOM catalyst. As shown in Fig. 13a, the macropores of catalyst contacts with soot particles, indicating that 3DOM structure is a desirable feature for diesel soot combustion. This is the direct evidence that the perfect macroporous structure provides the ideal reaction place to solid reactants (diesel soot). As shown in HRTEM images (Fig. 13b and c), the soot and K-OMS-2 nanoparticles are well contact each other in the inner pores of 3DOM structure. Due to this contact efficiency, the number of available active sites of catalysts can be maximized for utilization, especially inner active sites of catalysts. More active sites would result in higher catalytic activity. The uniform macroporous network allows easy mass transfer and less diffusional resistance when the large size materials such as soot particles go through the catalyst structure. TEM results have intuitionally demonstrated that soot particles can easily enter the interior of 3DOM catalysts with the help of the airflow in the reaction process under the loose contact conditions, and have less resistance to go through the catalyst structure. In fact, our group has synthesized a series of 3DOM materials and they show higher catalytic activities than the corresponding particle materials 10,11,14 . Therefore, it is significantly important to study and design 3DOM structure for soot combustion.

Possible Reaction Mechanism for Soot Combustion on K-OMS-2/SiO 2 Catalyst.
Based on all the results and analyses above, the potential mechanism schemes for soot combustion under presence/absence of NO are proposed and the possible mechanisms are shown in Fig. 14. When the NO is absent in the reaction system, soot oxidation occurred on the interface of K-OMS-2 nanoparticles and soot by the active oxygen species, which can be continuously supplemented by gaseous O 2 through the oxygen vacancies. Firstly, the gaseous O 2 is adsorbed on the active sites (i.e. oxygen vacancies on the surface of K-OMS-2), and then the adsorbed O 2 is  decomposed by redox reaction of Mn 3+ /Mn 4+ and formed active oxygen species(AOSs). Due to containing of K + in the tunnels of K-OMS-2, the electron clouds of K + may enhance the redox performance of Mn 3+ /Mn 4+ and lead to high production of AOSs. Thirdly, the AOSs are released from the surface of K-OMS-2 and migrated to the surface of soot. At last, the AOSs would be reacted with soot 47 . In this section, long distance between soot and K-OMS-2 would result in inactivation of AOSs. However, because of macropores effect of 3DOM structure, most of AOSs are fully exploited due to good contact between soot and K-OMS-2. As mentioned in the previous discussion, the reaction path of soot combustion has changed when NO is presented. The reaction mechanisms for the soot-NO-O 2 system can be proposed or summarized as follows. (1) The NO is oxidized to NO 2 by AOSs (released from the surface of K-OMS-2) in the gaseous atmosphere, and then strong oxidizing NO 2 migrated to the surface of soot and oxidized the soot. This way may be not played a major role in the system of soot-NO-O 2 due to inactivation of AOSs in the diffusion process; (2) Adsorption of gas phase O 2 and NO on the surface of catalysts. As revealed by XPS in the Fig. 9, due to changing valence state of Mn species (Mn 3+ /Mn 4+ ) in the K-OMS-2, the oxygen vacancies are formed on the surface of catalysts, and then the active oxygen can be easily generated on the vacancies sites and formed chemisorbed oxygen 48 . Correspondingly, the NO was also interacted with surface active oxygen and formed bidentate/monodentate nitrates on the Mn n+ sites. Those bidentate/monodentate nitrates species were desorbed by formation of NO 2 from K-OMS-2, which acts as oxidant species for soot combustion via a spillover mechanism when temperature is increased. In this section, the doping of K + has significant influence on oxygen vacancies and NO adsorption. The intercalated K + leads to the mixed/averaged valence state of Mn 4+ and Mn 3+ are randomly distributed and easily transformed between Mn 4+ and Mn 3+ . The interaction between K + and Mn n+ contributes to forming AOSs and enhance catalytic activity 36 . (3) The special channel structure of K-OMS-2, which is larger than 0.46 nm, allows a lot of gas molecules to insert the channels. In this work, the dynamic diameters of NO and O 2 are 0.317 and 0.346 nm, respectively. Thus, the channel size of K-OMS-2 is suitable for adsorption of NO and O 2 . In the channel of K-OMS-2, O 2 can be activated by the redox reaction of Mn 3+ /Mn 4+ and forms AOSs, and then the NO will be oxidized by AOSs to form NO 2 . After that, the NO 2 is spread out from the tunnel structure and oxidizes the soot. In this section, a mass of gaseous NO molecules are spread into the tunnel structure owing to the adsorption of K + . Therefore, a great number of gaseous NO molecules can be transformed into NO 2 to enhance the catalytic activity. More importantly, the K + ions are trapped in the 2 × 2 channels of K-OMS-2 49 . This structure guaranteed the stabilization of K + in the soot combustion. In the current work, as discussed above, the doping K + in the manganese oxides plays an essential role in the high catalytic activity for soot combustion.

Conclusions
A series of novel ordered micro/macro porous K-OMS-2/SiO 2 oxide catalysts, which contain 3DOM structure in the SiO 2 support and microporous structure in the nano-K-OMS-2, were firstly designed and successfully prepared by a very simple method. The average diameter of 3DOM SiO 2 is about 310 nm and microporous K-OMS-2 is about 0.46 nm. More interestingly, the microporous K-OMS-2 oxide nanoparticles with 20-25 nm are well dispersed and supported on the inner wall of the uniform macropores of SiO 2 .
The as-prepared catalysts were firstly used in soot combustion reaction and they show high catalytic activities. Especially, the catalytic activities of as-prepared K-OMS-2/SiO 2 catalysts are similar to or even higher than those of the expensive noble metal catalysts in the medium and high temperature range. For example, the T 50 of K-OMS-2/SiO 2 -50, 328 °C, is much lower than those of Pt/Al 2 O 3 and 3DOM Au/LaFeO 3 , 464 and 356 °C, respectively. Moreover, the catalysts exhibited high catalytic stability for soot combustion. The macroporous effect of 3DOM structure is responsible for increasing the contact efficiency, the microporous effect of 2 × 2 tunnels of K-OMS-2 for adsorption of gas molecules and interaction of K and Mn for the activation of gas molecules, which are favorable for enhancing the catalytic activity for soot combustion. The K + ions are inevitably introduced into the microporous structure of OMS-2 and stabilized in the catalytic reaction. Meanwhile, the K + ions play an important role in templating and stabilizing the tunneled framework of OMS-2. The characterization results also prove that the catalytic activity of ordered hierarchical micro/macro porous K-OMS-2/SiO 2 catalyst is similar to that of fresh catalyst after five-cycle reaction.
Via the strategy of changing of K-OMS-2 loading and comparison of different active components, we successfully prepared a low cost, environmentally friendly, and highly active catalyst for soot combustion. The as-prepared catalysts are comparable to the expensive noble metal catalysts. This work should aid the rational design and facile preparation of highly efficient oxidation catalysts through the incorporation of other different cations into the tunnel of K-OMS-2. More importantly, the as-prepared micro/macro porous K-OMS-2/SiO 2 catalysts are promising for practical applications in the catalytic oxidation of diesel soot particles due to easy synthesis, low cost, high activity and stability.

Methods
Synthesis of Highly Well-defined PMMA Microspheres. The PMMA microspheres were synthesized by a modified emulsifier-free emulsion polymerization method 14 . The detailed synthesis steps are listed in the Supporting Information. (SEM images of PMMA microspheres and PMMA colloidal crystal templates are shown in Fig. S1).
Synthesis of 3DOM SiO 2 . 3DOM SiO 2 was synthesized by colloidal crystal template (CCT) method with PMMA arrays as template and using tetraethyl orthosilicate (TEOS) as precursors 50 . In a typical procedure, 4.16 g TEOS was dissolved into the mixture of 2.5 mL water, 5 mL alcohol and 2.5 mL HCl aqueous solution (2 mol/L). After that, the hydrolyzation was proceeded in a water bath at 35 °C for 4 h. Then, 3 g PMMA arrays were impregnated into the above solution for 2 h. After complete impregnation, the PMMA arrays with the precursor solution were separated by vacuum filter and dried at 30 °C for 24 h. The dried samples were calcined to remove the CCT Scientific RepoRts | 7:43894 | DOI: 10.1038/srep43894 in a tube furnace with an air flow (80 ml min −1 ). The temperature-rising rate was 1 °C min −1 from room temperature to 600 °C, and the temperature of calcination at 600 °C was kept for 4 h, and then 3DOM SiO 2 supports were obtained.

Synthesis of 3DOM SiO 2 -supported Microporous K-OMS-2 NPs. 3DOM K-OMS-2/SiO 2 catalysts
were synthesized by incipient wetness impregnation method. In a typical procedure, a certain amount of KNO 3 and Mn(NO 3 ) 2 aqueous solution(50 wt%) were dissolved into deionized water, and then the above mixed aqueous solution was added into 3DOM SiO 2 . In this step, the volume of KNO 3 and Mn(NO 3 ) 2 mixed aqueous solution should be equal to the pore volume of 3DOM SiO 2 . After that, the impregnated sample was dealt with ultrasonic for 10 min and dried at 80 °C for 24 h. Then, the sample was calcined at 550 °C for 4 h in tube furnace and 3DOM K-OMS-2/SiO 2 catalysts were obtained. In order to obtain different K-OMS-2 loading amounts for K-OMS-2/ SiO 2 catalysts, the weight ratio of K-OMS-2 to SiO 2 was changed. The as-prepared catalysts were defined as K-OMS-2/SiO 2 -10, K-OMS-2/SiO 2 -20 and K-OMS-2/SiO 2 -30 for corresponding mass ratios of KNO 3 to SiO 2 were 5%, 10% and 15%, respectively. The dosages of raw materials are listed in the Table S1. Physical and Chemical Characterization. The characterization methods of X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), thermogravimetric analysis and differential scanning caborimetry (TG-DSC), Raman spectroscopy, H 2 temperature-programmed reduction (H 2 -TPR) and X-ray photoelectron spectroscopy (XPS) are described in the Supporting information.

Synthesis of 3DOM MnO
Activity Measurements. The catalytic performances of all the catalysts were evaluated with a temperature-programmed oxidation reaction (TPO) on a fixed-bed tubular quartz reactor (Φ = 8 mm), and each TPO run from 150 to 650 °C at a 2 °C min −1 rate. The model soot was Printex-U particulates (diameter 25 nm, purchased from Degussa). Elemental analysis of Printex-U particulates showed its carbonaceous nature with 92.0% C, 0.7% H, 3.5% O, 0.1% N, 0.2% S and 3.5% others 14 . The catalyst (100 mg) and soot (10 mg) were mixed at a weight ratio of 10:1 with a spatula in order to reproduce the loose contact mode. Reactant gases (50 mL min −1 ) contain 10% O 2 and 0.2% NO balanced with Ar. The outlet gas compositions were analyzed with an on-line gas chromatograph (GC, Sp-3420, Beijing) by using FID detectors. Before entering the FID detector, CO and CO 2 were fully converted to CH 4 by a convertor with Ni catalyst at 380 °C. The catalytic activity was evaluated by the values of T 10 , T 50 and T 90 , which were defined as the temperatures at 10%, 50% and 90% of soot conversion, respectively. The selectivity to CO 2 formation (S CO2 ) was defined as that the CO 2 outlet concentration (C CO2 ) divided by the sum of the CO 2 and CO outlet concentration, i.e., S CO2 = C CO2 /(C CO + C CO2 ). S m CO2 was denoted as S CO2 at the maximum temperature corresponding to the soot-burnt rate was the highest. In all TPO experiments, the reaction was not finished until the soot was completely burnt off.