Soot particulate emitted from diesel engines is one of the most hazardous pollutants that threatens the health of people and harms the environment [1, 2]. For the elimination of soot particles, the combination of diesel particulate filter and oxidation catalysts [3, 4] is the most efficient after-treatment technique developed to date. An efficient soot combustion catalyst should typically have a certain surface area and porous structure that permits adequate interactions with the soot particulates in the form of a solid-solid-gas reaction. Three-dimensionally ordered macroporous (3DOM) materials [5-8] with big pore sizes (> 50 nm) permit solid reactants to enter the pores and develop sufficient contact with the active sites. Recently, it was reported that 3DOM materials such as Co₃O₄ [9], Ce-Zr-O_x [10], LaFeO₃ [11], and LaMnO₃ [12] showed good catalytic activity for several heterogeneous catalytic reactions, including soot combustion [13-16]. Zhao et al. [11] reported that 3DOM Au/LaFeO₃ catalysts prepared with three-dimensional structural supports and containing nanoparticle active sites might be useful in fundamental research on metal-support synergisetic effect, and also has potential for applications in the catalytic oxidation of solid particles. Dai et al. [17] found that three-dimensional macroporous perovskites exhibited excellent catalytic activity for methane combustion, because the unique nanovoid 3D structure of the catalysts can provide larger surface areas and pore volumes, uniform pore sizes, higher accessible surface oxygen concentrations, and better low-temperature reducibility in comparison to regular samples.

SnO₂ is an n-type semiconductor metal oxide having both reactive surface-deficient oxygen and lattice oxygen. Furthermore, its melting point is as high as 1630 ℃, proving it has good chemical and thermal stability. All these factors indicate that it is a promising and potential catalytic material [18]. Therefore, over the past five years, our group has systematically investigated the catalytic chemistry of SnO₂ for various reactions [19-21]. It was found that SnO₂ with a rod shape showed the reaction behavior of supported precious metals because of preferentially exposed (110) facets [19]. Moreover, doping the lattice of SnO₂ nano-rods with a small amount of In, Cr, and Al cations can further enhance its activity, as evaluated for the complete oxidation of toluene. An Al-doped Sn₉₉Al₁ rod shows the highest activity among all the catalysts owing to its optimized surface-active oxygen species and acidic sites [20]. If the crystal matrix of SnO₂ is modified by suitable amounts of Nb and Mn cations to form pure rutile tetragonal structured solid solutions, the best activity can be generally obtained for CO and CH₄ oxidation [22]. Based on our work, we have developed an easy and reliable X-ray diffraction (XRD) extrapolation method for the first time to evaluate the lattice capacity of a solute cation in a solvent metal oxide lattice [23].

Some other researchers have used SnO₂-based materials for different catalytic reactions. Li et al. [24] reported that a Ce-Sn binary oxide catalyst revealed promising activity for the selective reduction of NO by NH₃. The major reason for this is the synergetic effect between the Ce and Sn oxides, which not only enhances the redox property of the catalyst but also increases its Lewis acidity, thus promoting the adsorption and activation of the NH₃ reactant. In addition, it was revealed by another group that Ce-doped SnO₂ catalysts were also active for methane combustion. The surface Sn⁴⁺ cations are found to be the active sites, and the surface lattice oxygen species plays an important role in determining the activity [25]. Our studies also testified that the catalytic performance of SnO₂ could be enhanced by substituting a fraction of the Sn⁴⁺ ions in the crystal lattice with other cations such as Ce⁴⁺ [26], Mn³⁺ [22], and Cu²⁺ [27] to form solid solutions having the tetragonal rutile structure, which improves oxygen mobility, thermal stability, and eventually, the activity for CO and CH₄ oxidation.

In recent years, 3DOM SnO₂ materials have been used for gas sensing [28-31], anodes for Li-ion batteries [32], photoanodes for dye-sensitized solar cells [33], electrodes for degrading benzoic acid [34]and as catalysts for Suzuki cross coupling reactions [35]. For instance, O'Dwyer et al. [35] reported that Pd nanoparticles highly dispersed on ordered SnO₂ IO supports showed superior catalytic performance for liquid-phase Suzuki coupling reactions and allow the easy removal of the catalyst substrate after the reaction. Compared to the commercialized Pd/C catalysts, a higher mass electrocatalytic activity is also demonstrated over SnO₂-Pd IOs for formic acid oxidation, which is due to better access of the reactants to the active sites. However, the study on using 3DOM-SnO₂-based materials for soot combustion has not been reported yet. It is reasonably presumed that doping the lattice matrix of SnO₂ with other transition metals such as Cu²⁺, Mn³⁺, or Ce⁴⁺ to form a solid solution and then developing them into a 3DOM structure could enhance the activity of the resulting catalysts for soot combustion. Therefore, in this study, a series of Cu-, Mn-, or Ce-doped SnO₂ solid solution catalysts with 3DOM structures have been prepared and used for soot particulate combustion. Indeed, it has been revealed that with the modification of the SnO₂ matrix by Ce⁴⁺, Mn³⁺, or Cu²⁺ and the formation of the 3DOM structure, the soot combustion activity of the prepared catalysts evidently improved; among these, the Cu-modified catalyst exhibits the best promotional effect. Using different characterization techniques, the influence of the 3DOM solid solution structure on the properties of SnO₂ catalysts has been analyzed and elucidated, and correlated with their soot combustion activities.

Volumes of 115 mL of PMMA (Guoyao, CP) monomer and 1300 mL of distilled deionized (DDI) water were added to a 3000 mL three-necked round-bottomed flask. A N₂ flow rate of 100 mL min^–1 through the flask was used to remove the air inside it, and the mixture was constantly stirred magnetically. The temperature of the reaction system was controlled at 70 ℃; 50 mL of a potassium persulfate solution (0.03 mol L^–1) was added to the mixture to initiate the polymerization reaction, and held at this temperature for 60 min. Afterwards, it was cooled to room temperature. The solution mixture containing monodispersed PMMA microspheres was then centrifuged at the rate of 4500 r min^–1 for 50 min to obtain the PMMA precipitate, which was then dispersed in DDI water and dried using an 80 ℃ water bath to obtain the final highly ordered PMMA colloidal crystal templates (Fig. 1(A)).

The unmodified 3DOM-SnO₂ sample was prepared via the thermal decomposition of the Sn precursor and the PMMA template mixture. In detail, SnCl₄·5H₂O (Guoyao, AR) was dissolved in a certain amount of DDI water to form a stable solution (0.1 mol L^–1). Then, 5 g PMMA templates were permeated into 5 mL of this Sn precursor solution for 2 h, and the solution was filtered and dried at room temperature. Subsequently, the dried PMMA colloidal crystal templates containing the Sn precursor were calcined, with the temperature increasing linearly at 1 ℃ min^–1 to 300 ℃ and kept for 2 h in N₂ flow, before cooling to room temperature. After this, the sample was calcined again by increasing the temperature at the rate of 1 ℃ min^–1 to 700 ℃ and maintaining it for 4 h in air flow.

Ce⁴⁺-, Mn³⁺-, and Cu²⁺-doped 3DOM-SnO₂ were synthesized by similar processes, except that the SnCl₄·5H₂O precursor was dissolved with a certain amount of Ce (NO₃)₃·6H₂O (Aladdin, AR), Mn (NO₃)₂ (Guoyao, 50 wt%), or Cu (NO₃)₂·3H₂O (Guoyao, AR) in DDI water. The Ce/Sn, Mn/Sn, or Cu/Sn molar ratio for the different catalysts was always controlled at 1/9. The final catalysts were named 3DOM-Ce₁Sn₉, 3DOM-Mn₁Sn₉, and 3DOM-Cu₁Sn_9, respectively, according to the chemical compositions.

Scanning electron microscopy (SEM) images were recorded on a Hitachi S-4800 field emission scanning electron microscope. Similarly, transmission electron microscopy (TEM) images were taken using a TecnaiTM F30 transmission electron microscope. Elemental phase mapping and surface scans were also obtained by energy-dispersive spectroscopy (EDX) using the TecnaiTM F30 microscope equipped with an Oxford EDX detector operated at 300 keV.

Nitrogen adsorption-desorption experiments of the samples were performed at –196 ℃ on an ASAP2020 instrument. The specific surface areas of the catalysts were calculated using the Brunauer-Emmett-Teller method in the relative pressure (p/p₀) range 0.05–0.25. The pore size distributions of the samples were calculated using the Barrett-Joyner-Halenda method, and the pore sizes were obtained from the peak positions of the distribution curves. The total pore volume was accumulated at p/p₀ = 0.99.

Powder XRD patterns were obtained using a Bruker AXS D8Focus diffractometer operating at 40 kV and 30 mA with Cu K_α irradiation (λ = 1.5405 Å). The scans were performed in the 2θ range 10° to 90° using a step size of 2° min^–1. To keep the data comparable, all the samples were tested continuously.

Raman spectra of the catalysts were recorded on a Renishaw inVia Raman spectrometer equipped with an argon laser excitation source operating at 532 nm and a Renishaw RenCam CCD detector. The scanned Raman shift range was from 200 to 800 cm^–1.

Hydrogen temperature-programmed reduction (H₂-TPR) experiments were carried out on a FINESORB 3010C instrument. Generally, 0.01 g of the catalyst was used for the tests. Before the experiments, the catalysts were recalcined in high-purity air flow at 120 ℃ for 30 min to remove any possible surface impurities. After purging with ultrahigh purity Ar flow at room temperature for 30 min, the temperature was then increased from room temperature to 800 ℃ at the rate of 10 ℃ min^–1 with 10% H₂/Ar gas mixture flow (30 mL min^–1). A thermal conductivity detector (TCD) was employed to monitor the H₂ uptake. To quantify the amount of H₂ consumed, a 99.99% CuO sample was used as the calibration standard.

Temperature-programmed desorption (O₂-TPD) measurements were carried out on a Huasi DAS-7000 adsorption instrument equipped with a TCD detector. Typically, 0.05 g of the sample was placed in a quartz reactor pretreated in 30 mL min^–1 ultrahigh purity Ar flow at 300 ℃ for 60 min. Afterwards, the sample was cooled to 50 ℃ and saturated in 10% O₂ + Ar flow (30 mL min^–1), which was followed by purging with ultrahigh purity Ar flow (30 mL min^–1) for 60 min to remove any physically adsorbed O₂. The TPD experiments were then carried out from 50 to 800 ℃ at the heating rate of 10 ℃ min^–1 in ultrahigh purity Ar flow (30 mL min^–1).

X-ray photoelectron spectroscopy (XPS) tests were performed on a PerkinElmer PHI1600 system using a single Mg Kα X-ray source operating at 300 W and 15 kV. The spectra were obtained at ambient temperature using ultrahigh vacuum. The binding energies were calibrated using the C 1s peak of graphite at 284.6 eV as the standard.

Temperature-programmed oxidation technique (TPO) was used to evaluate the activity of the prepared catalysts for soot combustion. The tests were carried out in a fixed-bed quartz tube microreactor with an inner diameter of 6 mm. To evaluate the activity of the catalyst under loose contact condition, typically 5 mg of soot (Printex-U, diameter of 25 nm, purchased from Degussa) were mixed with 50 mg of the catalyst powder and stirred for 2 min to obtain a loose contact mixture. Before loading onto the reactor, the prepared mixture was diluted with 100 mg of inert silica containing a soot/catalyst/silica weight ratio of 1/10/20 to avoid the formation of hotspots during the activity test. A K-type thermocouple was used to monitor the temperature of the catalyst bed accurately, with the thermocouple head point touching the bed. To measure the reaction behaviors of the catalysts, all the data were collected by increasing the temperature to 800 ℃ at the rate of 10 ℃ min^–1. The volume composition of the feed gas was 10% O₂ balanced by high-purity Ar, with the flow rate (30 mL min^–1). To accurately quantify the amount of O₂ consumed, prior to entering the TCD, the formed CO₂ was removed thoroughly using soda lime trap. The catalytic activity was evaluated based on the ignition temperature (T_i) and the highest combustion temperature (T_p) reflected in the TPO profiles. In order to obtain the selectivity of the products on the samples, soot combustion on the catalysts was performed again without the soda lime trap, and the products were monitored through online SHP8400PMS mass spectrometry. The m/z = 44 signal corresponds to CO₂; and the m/z = 28 signal obtained by subtracting the fragment from CO₂ is of CO. The selectivity towards CO₂ production (S(CO₂)) is obtained according to the equation S(CO₂) = C (CO₂)/(C(CO) + C(CO₂)). Here, C (CO₂) and C (CO) are the outlet concentrations of CO₂ and CO, respectively.

The morphologies and macroporous structures of all the samples have been investigated by using SEM and TEM. Fig. 1(A) shows that the obtained PMMA template consists of highly uniform colloidal microspheres with narrow size distributions that are closely packed in a well-organized 3D arrangement, with the average size being around 350 nm. Fig. 1(B) (SEM) and Fig. 1(C) (TEM) of the 3DOM-SnO₂ sample show that it has a well-defined 3DOM structure, and visible SnO₂ nanoparticles are present on the skeleton of the framework. With the addition of Ce⁴⁺, Mn³⁺, and Cu²⁺ cations, all the samples still have a well-defined 3DOM structure, as shown in Fig. 1(D)–(F) in sequence. The average pore size of the 3DOM-SnO₂-based samples is about 300 nm, and the voids interconnected through the open windows are around 95 nm in diameter. Apparently, both modified and unmodified SnO₂ samples with well-defined 3DOM structures have been successfully synthesized. This structure could be favorable for mass transfer and the contact of the soot particles with active sites and oxygen because of the presence of abundant macropores and continuous pore connectivity [36].

The nanoparticles visible on the skeleton of the 3DOM samples were further studied by HRTEM, with the corresponding images shown in Fig. 2. Irregular microspheres are observed for all the samples, both modified and unmodified. In addition, both the exposed (110) and (101) facets are observed in all the samples, indicating that the addition of secondary Ce⁴⁺, Mn³⁺, and Cu²⁺ ions with a M/Sn molar ratio of 1/9 did not evidently influence the crystallization process or the growth direction of SnO₂. As listed in Table 1, all the samples have an average particle size of around 6.0 nm. With the addition of the secondary metal cations, the average particle size only slight decreases, which might imply that Ce⁴⁺, Mn³⁺, and Cu²⁺ could have been incorporated into the SnO₂ lattice to form a solid solution structure [22, 26, 37] that impedes the aggregation of the SnO₂ crystallites.

The N₂ adsorption-desorption technique was used to discern the textural properties of the catalysts, with the isotherms and pore diameter distribution profiles displayed in Fig. 3 and the quantification results in Table 1. All the catalysts exhibit the typical type Ⅱ isotherm, and the hysteresis loop in the relative pressure (p/p₀) range 0.7–1.0 belongs to H3-type, which is typical of the interstice mesoporous structure formed by the nanoparticle assembly. As listed in Table 1, after calcination at 700 ℃, the 3DOM-SnO₂ catalyst has a surface area of 52 m² g^-1, an average pore volume of 0.28 cm³ g^-1, and an average pore size of 17.97 nm, which are much higher than those of the regular SnO₂ nanoparticles prepared by traditional methods [21]. In comparison, the addition of the secondary metal ions increases the surface areas and average pore volumes apparently, but decreases the average pore sizes, testifying that the modified catalysts have a larger amount of smaller pores in the structure, which could be favorable for the diffusion of reactants and products. In summary, the addition of the secondary metal cations can evidently alter the texture properties of 3DOM-SnO₂, which might produce a porous structure that benefits the soot combustion activity.

The soot combustion activity over the prepared catalysts has been tested in the loose contact condition to simulate the real application process, and the TPO profiles are shown in Fig. 4. For better clarity, the combustion of pure soot and over regular SnO₂ nanoparticles (NP-SnO₂) are also evaluated under the same conditions, and the results are presented in Fig. 4(A) and Table 2.

Without any catalyst, the combustion of pure soot starts at 530 ℃ (T_i) and the highest rate is obtained at 630 ℃ (T_p). In addition, the selectivity towards CO₂ is only 68%, indicating that the oxidation of the soot is incomplete. On NP-SnO₂, the corresponding T_i and T_p values drop, and the CO₂ selectivity improves to 86%. On the 3DOM-SnO₂ catalyst, T_i and T_p further decrease by 80 and 40 ℃, respectively, though the CO₂ selectivity has only slightly increased, suggesting that the formation of the 3DOM structure is favorable for the soot combustion activity. The incorporation of the secondary cations more or less improves the soot combustion activity of the catalysts. It is observed that Cu²⁺ addition results in much better promotional effects for the activity and CO₂ selectivity compared to the other two dopants. Over 3DOM-Cu₁Sn₉, with the combustion starting at 370 ℃ (T_i) and the highest rate being obtained at 545 ℃ (T_p), the CO₂ selectivity reaches 95% even under the loose contact condition. In summary, the soot combustion activities of all the catalysts follow the sequence 3DOM-Cu₁Sn₉ > 3DOM-Mn₁Sn₉ > 3DOM-Ce₁Sn₉ > 3DOM-SnO₂ > NP-SnO₂. It is apparent that both the formation of the 3DOM structure and the incorporation of the secondary metal cations can benefit the soot oxidation activities of the catalysts.

The stability of 3DOM-Cu₁Sn₉, the best catalyst observed in this study, has been tested to evaluate its application potential by performing four continuous TPO experiments, and the results are shown in Fig. 4(B). It is observed that from the first to the last test, T_p increases from 545 to 570 ℃, indicating that the catalyst has experienced some deactivation.

The phase compositions and crystalline structures of all the samples have been analyzed by XRD, with the patterns shown in Fig. 5. For unmodified 3DOM-SnO₂, tetragonal rutile SnO₂ is the only crystalline phase detected, as evidenced by the typical diffraction peaks at (110) = 26.659°, (101) = 33.979°, and (211) = 52.177° (JCPDS no. 41–1445). As listed in Table 3, for the unit cell of 3DOM-SnO₂, the lattice parameters are a = b = 4.7106 Å, c = 3.1810 Å, and all the intersection angles are 90°, which is typical for a cell with tetragonal structure.

Interestingly, with the incorporation of the secondary Ce⁴⁺, Mn³⁺, and Cu²⁺ into SnO₂, all the samples still display the typical diffraction characteristics of the tetragonal rutile SnO₂ phase, without the detection of any diffraction peaks related to the corresponding secondary oxides; however, the diffraction peaks of the SnO₂ phase become less intense. Compared with the unmodified 3DOM-SnO₂ sample, the lattice parameters of the 3DOM-Ce₁Sn₉, 3DOM-Mn₁Sn₉, and 3DOM-Cu₁Sn₉ samples distinctly change (Table 3), which proves that the secondary cations entered the lattice matrix of SnO₂ to form a solid solution structure with lattice distortion and structural defects [23, 38, 39]. As reported, the radii of Ce⁴⁺, Mn³⁺, and Cu²⁺ with a coordination number of 6 are 0.87, 0.65, and 0.73 Å in that order, which are close to that of the Sn⁴⁺ ion (0.69 Å) with the same coordination number. Theoretically, Ce⁴⁺, Mn³⁺, and Cu²⁺ can enter the matrix of the SnO₂ lattice to form stable non-continuous solid solutions with certain lattice capacities [22, 23, 40]. It has previously been found that lattice distortion and charge imbalance could arise with the formation of a non-continuous solid solution structure, thus generating more mobile oxygen species. Furthermore, the formation of the solid solution structure can also inhibit the crystallization process, thus stabilizing the surface areas of the catalysts [22, 23]. The smaller crystallite sizes listed in Table 3 and the higher surface areas observed in Table 1 for the modified 3DOM-SnO₂ catalysts prove this hypothesis. In summary, the XRD results are in good agreement with the N₂ adsorption-desorption results, indicating that both the presence of the 3DOM structure and the doping of the secondary metal cations into the SnO₂ lattice are effective ways to enhance the soot combustion activities of the catalysts.

To further confirm that the doped cations have been effectively incorporated into the lattice matrix of the tetragonal rutile SnO₂ for the three-dimensionally ordered samples to form solid solution structures, 3DOM-Cu₁Sn₉, the most active catalyst obtained in this study, has been subjected to a STEM mapping study. As shown in Fig. 6, in the mapped zone, elemental Cu and Sn are distributed uniformly along with O, which provides further proof of the XRD results and demonstrates that the Cu²⁺ cations replace a fraction of the lattice Sn⁴⁺ to form a homogeneous solid solution structure.

Raman technique has been used to study the structures of the catalysts, and the spectra are shown in Fig. 7. For purposes of comparison, the spectrum of regular NP-SnO₂, without the 3DOM structure, which was calcined at 700 ℃ for 4 h, has also been collected and included. Fig. 7 shows three Raman bands at about 773, 633, and 476 cm^–1, which can be assigned to the B_2g, A_1g, and E_g vibration modes of tetragonal rutile SnO₂ of D_4h¹⁴ group, based on group theory [41]. Interestingly, 3DOM-SnO₂, a pure SnO₂ sample having a 3DOM structure, did not reveal any Raman peak that could be detected. The 3DOM structure is believed to be responsible for the disappearance of the peaks, since the abundant surface macropores can scatter and weaken the Raman signals. Therefore, for 3DOM-Ce₁Sn₉, a strong peak assigned to the F_2g vibration mode of CeO₂ is observed at 449 cm^–1, and another peak observed at 617 cm^–1 can be attributed to ceria surface oxygen vacancies [42, 43]. For 3DOM-Mn₁Sn₉, an intense peak at 627 cm^–1 is observed, which is assigned to the out-of-plane bending mode of Mn₂O₃ [44]. For 3DOM-Cu₁Sn₉, a peak centered at 609 cm^–1 is observed, corresponding to the B_g vibration mode of CuO [45]. In addition, for 3DOM-Mn₁Sn₉ and 3DOM-Cu₁Sn₉, a peak at 495 cm^–1 is observed, which is assigned to the A_2uTO mode due to surface defects such as oxygen vacancies (O_Vs) and lattice disorder [46].

In summary, it is apparent that the addition of the secondary metal cations into the lattice matrix of tetragonal rutile SnO₂ to form solid solutions can improve the lattice disorder and create more surface oxygen vacancies, thus altering the surface properties of the prepared catalysts.

The H₂-TPR technique was used to investigate the redox properties of the catalysts, and the profiles are shown in Fig. 8. Pure 3DOM-SnO₂ shows a major reduction peak at 626 ℃, which can be assigned to the reduction of SnO₂ to metallic Sn [23]. With the addition of Ce⁴⁺, Mn³⁺, or Cu²⁺, the temperature corresponding to the major reduction peak shifts to a lower value. This indicates the alteration of the chemical environment of Sn⁴⁺ in different samples that accompanies the formation of the more reducible lattice oxygen species, which is generally beneficial for soot oxidation activity. The 3DOM-Cu₁Sn₉ sample reveals the lowest temperature for SnO₂ reduction, which may be one of the reasons for its highest catalytic activity for soot combustion. It must be mentioned that the quantified O/Sn atomic ratios of all the samples are around 2.0/1, indicating that the Sn species has fully oxidized to Sn⁴⁺, in agreement with the XRD and XPS results.

More interestingly, for all the catalysts, a reduction peak below 450 ℃ is observed, as indicated by the enlarged profiles in Fig. 8(B). It has been reported that the surface of SnO₂ contains a lot of oxygen vacancies, which results in a certain amount of active deficient oxygen species [23]. With the addition of Ce⁴⁺, Mn³⁺, or Cu²⁺, a larger amount of this type of oxygen species is formed in all the samples, except for the case of 3DOM-Ce₁Sn₉, as listed in Table 4 for the amounts of H₂ consumed for the peak below 450 ℃.

It is noted here that owing to the reduction of CuO_x to metallic Cu in the 3DOM-Cu₁Sn₉ sample, as evidenced by the intense peak at 198 ℃, an extremely high H₂ consumption amount below 450 ℃ is observed [37]. The quantified results indicate that the Cu⁺/Cu²⁺ atom ratio is 0.19, proving that the Cu²⁺ cations predominantly exist in the bulk of 3DOM-Cu₁Sn₉. The result also confirms that the Cu²⁺ cations substitute a fraction of the Sn⁴⁺ cations in the SnO₂ crystalline lattice, which is consistent with the theoretical analysis discussed above that suggested that some Sn⁴⁺ cations can be replaced by Cu²⁺ in the SnO₂ lattice to form a solid solution structure.

The surface compositions and in particular the surface oxygen properties of the catalysts have been analyzed by XPS, and the corresponding spectra are shown in Figs. 9 and 10 and the quantified results listed in Table 5. The spectrum of 3DOM-SnO₂ (Fig. 9(A)) displays two Sn 3d peaks at 486.4 and 494.9 eV, which are assigned to Sn 3d_5/2 and 3d_3/2, respectively, and attributed to Sn⁴⁺ [21], indicating that Sn is fully oxidized in all the samples. As listed in Table 5, with the addition of the secondary metal ions (Ce⁴⁺, Mn³⁺, Cu²⁺), the Sn 3d peaks shift and the △E between the two peaks experiences only a slight change, revealing the change in the chemical environment of the Sn cations.

For 3DOM-Ce₁Sn₉, two sets of peaks corresponding to Ce 3d_3/2 and Ce 3d_5/2 of the Ce cation with different valence states are observed (Fig. 9(B)). The peaks at 907.9 and 883.7 eV are assigned to the surface Ce³⁺ species [47], while all the other peaks are ascribed to the surface Ce⁴⁺ species [24]. The quantified Ce³⁺/Ce⁴⁺ molar ratio is 0.06, suggesting that the surface of 3DOM-Ce₁Sn₉ is enriched by the Ce⁴⁺ cations. It has been reported that surface Ce³⁺ can induce the formation of oxygen vacancies and unsaturated chemical bonds, which is in line with the Raman results of 3DOM-Ce₁Sn₉, and improve the activity of the catalyst for soot combustion [38, 48].

For 3DOM-Mn₁Sn₉, two XPS Mn 3p_3/2 peaks at 641.6 and 645.4 eV are observed (Fig. 9(C)), which are assigned to Mn³⁺ and Mn⁴⁺ [49, 50], respectively. The Mn 3p_1/2 peak at 653.7 eV is also characteristic of Mn³⁺. The calculated Mn³⁺/Mn⁴⁺ molar ratio is 1.86, proving the surface of 3DOM-Mn₁Sn₉ is enriched by the Mn³⁺ cations, which is in accordance with the fact that Mn₂O₃ is the stable manganese oxide under normal conditions and that the Mn³⁺ cations enter the SnO₂ lattice matrix to form a solid solution structure [21]. It has been reported that the presence of the Mn⁴⁺ species is beneficial for the activity towards soot combustion [51].

For the 3DOM-Cu₁Sn₉ sample, the clear split in the Cu 2p spectrum (Fig. 9(D)) indicates the mixed oxidation state of Cu. The two peaks at 954.5 eV for Cu 2p_1/2 and 934.5 eV for Cu 2p_3/2 are attributed to the Cu²⁺ species; and the two peaks at 952.6 eV for Cu 2p_1/2 and 932.8 eV for Cu 2p_3/2 are attributed to the Cu⁺ species [38, 52, 53]. Interestingly, the calculated surface Cu⁺/Cu²⁺ molar ratio is 3.70, suggesting Cu⁺ enrichment on the surface of the 3DOM-Cu₁Sn₉ sample, where the concentration is different from the bulk compositions identified by H₂-TPR. This testifies that the formation of the 3DOM solid solution structure favors the formation of surface Cu⁺. According to previous studies on SiO₂(CuO_xSnO_y) and Cu_4–xMo₃O₁₂ catalysts [53, 54], Cu⁺ is the most active species capable of initiating soot combustion [54]. Therefore, the presence of abundant surface Cu⁺ species could be an important reason for the highest soot combustion activity of the 3DOM-Cu₁Sn₉ catalyst.

The oxygen species on the surface of the catalysts are particularly investigated, with the results displayed in Fig. 10 and Table 6. Two peaks at about 530.0 and 532.0 eV are observed for all the samples, which are assigned to the surface lattice oxygen (O_latt) and loosely adsorbed surface oxygen (O_ads), respectively [21]. Interestingly, the incorporation of different doping cations not only shifts the binding energies but also alters the O_ads/O_latt molar ratios of the catalysts (Table 6). Furthermore, the O_ads/O_latt ratios increase in the order 3DOM-SnO₂ < 3DOM-Ce₁Sn₉ < 3DOM-Mn₁Sn₉ < 3DOM-Cu₁Sn₉, indicating increasing abundance of mobile surface oxygen species in that sequence. According to the literature [21], the loosely bonded surface oxygen species could play a vital role in redox reactions such as soot combustion. To depict this clearly, the T_p on the catalysts has been plotted against the O_ads/O_latt molar ratio in Fig. 10(B). It is evident that the T_p decreases with increasing O_ads/O_latt molar ratio, testifying the formation of more abundant mobile oxygen species, which is another important reason for the high soot oxidation activity.

O₂-TPD experiments have been performed to further study the oxygen properties of the SnO₂ samples, and the results are shown in Fig. 11 and Table 7. For all the samples, three oxygen desorption peaks are observed below and above 200 ℃. For the convenience of discussion, the peaks are divided into two groups: the low-temperature peak at around 100 ℃ is named the α peak, and the two overlapping high-temperature peaks above 200 ℃ are named the β peak. The α peak is assigned to the desorption of loosely bonded surface oxygen, and the β peak to the desorption of mobile surface lattice oxygen [20, 21]. For better clarity, the two groups of peaks for each sample have been integrated and normalized to the sample amount, and the results are listed in Table 7. While the amounts of loosely bonded surface oxygen follow the order 3DOM-Ce₁Sn₉ > 3DOM-Cu₁Sn₉ > 3DOM-Mn₁Sn₉ > 3DOM-SnO₂ and the amounts of surface lattice oxygen follow 3DOM-Cu₁Sn₉ > 3DOM-Mn₁Sn₉ > 3DOM-SnO₂ > 3DOM-Ce₁Sn₉, the total amounts of oxygen desorbed follow the order 3DOM-Cu₁Sn₉ > 3DOM-Mn₁Sn₉ > 3DOM-Ce₁Sn₉ > 3DOM-SnO₂. Compared with the activity results, it is apparent that if a catalyst is more active, it contains a larger amount of total desorbed oxygen species. In brief, if the H₂-TPR results demonstrate that more reducible oxygen species has been formed in a more active catalyst, the O₂-TPD results prove directly that it contains a higher amount of active surface oxygen species, which is in agreement with the Raman and XPS results.

Indeed, the formation of the more facile oxygen species through the formation of 3DOM solid solution structures with secondary cations is believed to be favorable in enhancing the soot combustion activities of the prepared catalysts. In addition, the anchoring of the soot particulates in the macropores of the 3DOM catalysts, which ensures that the contact of the soot particles with the active sites on the 3DOM skeleton is achieved more easily, is another major reason for the enhancement in the activity of the catalysts.

To improve the reaction performance of SnO₂-based catalysts for soot combustion, a colloidal crystal templating method has been adopted to prepare a series of 3DOM catalysts. SEM, TEM, and HRTEM results have revealed that interconnected networks of the 3DOM structure can be successfully achieved using all the catalysts. The formation of the 3DOM structure can anchor the soot particulates in the macropores, which ensures that the contact of the soot particles with the active sites on the 3DOM skeleton is established more easily, thus benefiting the target reaction. XRD and STEM mapping results have confirmed that all the doped cations are incorporated into the crystal lattice of SnO₂ to form a non-continuous solid solution structure that impedes the crystallization process and improves the surface areas and pore volumes of the catalysts. Raman, H₂-TPR, XPS, and O₂-TPD results testify that a more abundant mobile oxygen species has also been formed, which is important for the soot combustion activity. The formation of the 3DOM solid solution structure and the more abundant mobile oxygen species are the critical reasons for the activities observed of the catalysts. 3DOM-Cu₁Sn₉ contains the largest amount of mobile oxygen species, thus revealing the highest activity among all the catalysts.