Soot Combustion over Cu–Co Spinel Catalysts: The Intrinsic Effects of Precursors on Catalytic Activity

In this work, a series of CuCo2O4-x (x = N, A and C) catalysts were synthesized using different metal salt precursors by urea hydrothermal method for catalytic soot combustion. The effect of CuCo2O4-x catalysts on soot conversion and CO2 selectivity in both loose and tight contact mode was investigated. The CuCo2O4-N catalyst exhibited outstanding catalytic activity with the characteristic temperatures (T10, T50 and T90) of 451 °C, 520 °C and 558 °C, respectively, while the CO2 selectivity reached 98.8% during the reaction. With the addition of NO, the soot combustion was further accelerated over all catalysts. Compared with the loose contact mode, the soot conversion was improved in the tight contact mode. The CuCo2O4-N catalysts showed better textural properties compared to the CuCo2O4-A and CuCo2O4-C, such as higher specific surface areas and pore volumes. The XRD results confirmed that the formation of a CuCo2O4 crystal phase in all catalysts. However, the CuO crystal phase only presented in CuCo2O4-N and CuCo2O4-A. The relative contents of Cu2+, Co3+ and Oads on the surface of CuCo2O4-x (x = N, A and C) catalysts were analyzed by XPS. The CuCo2O4-N catalyst displayed the highest relative content of Cu2+, Co3+ and Oads. The activity of catalytic soot combustion showed a good correlation with the order of the relative contents of Cu2+, Co3+ and Oads. Additionally, the CuCo2O4-N catalyst exhibited lower reduction temperature compared to the CuCo2O4-A and CuCo2O4-C. The cycle tests clarified that the copper–cobalt spinel catalyst obtained good stability. In addition, based on the Mars–van Krevelen mechanism, the process of catalytic soot combustion was described combined with the electron transfer process and the role of oxygen species over CuCo2O4 spinel catalysts.


Introduction
Soot particles are emitted from incomplete combustion of hydrocarbon fuels in diesel engines [1]. Mitigating soot particle emissions from diesel engines has raised much attention due to the negative impact of soot particles on human health and environmental protection [2,3]. A diesel particulate filter (DPF) is the most effective after-treatment technology for soot particles from diesel engines. However, a common issue of pore blockage over DPFs after a period of use should be addressed [4]. One of the most effective measures to improve soot oxidation is to coat the DPF with catalysts and form a catalyzed DPF (CDPF) [5]. CDPFs with highly active catalysts can accelerate soot oxidation and lower the ignition temperatures.
Various catalysts have been proposed and optimized to accelerate the catalytic soot combustion, while Pt-based catalysts have shown superb soot oxidation performance under practical conditions [6,7]. However, due to the high cost of noble metals, transition metalbased catalysts have been intensively studied for catalytic soot oxidation more recently [8]. Copper-cobalt catalysts play an important role as substitutes for noble metals in the field of catalytic soot combustion due to their outstanding redox properties and thermal stability [9]. 2

of 17
Zhang et al. reported that a spinel-type CuCo 2 O 4 catalyst showed a lower T 50 temperature (574 • C) of catalytic soot combustion than that of Co 3 O 4 catalysts (580 • C) due to its higher relative content of Co 3+ species [10]. Jampaiah et al. found that a CuCo-MnO 2 catalyst exhibited a lower T 90 temperature (485 • C) of catalytic soot combustion than that of NiCo-MnO 2 catalysts (513 • C) under the loose contact mode due to more oxygen vacancies and active sites on the catalyst surfaces [11]. Zhang et al. tested the stability of CuCo 2 O 4 spinel catalysts in four consecutive soot combustion cycles and found that the values of T 50 for catalytic soot combustion increased from 539 • C to 544 • C, 547 • C and 549 • C, respectively [12].
Recently, the effect of metal salt precursor on the textural and redox properties of catalysts was found to play a key role in catalytic activity and product selectivity [13]. Yang et al. found that the catalytic activity of selective acetylene hydrogenation at 50 • C over the Pd/Al 2 O 3 catalyst using acetate as the Pd precursor reached 78.6%, which was 31.1% higher than that using chloride precursors since Cl residuals existed in the form of PdCl 2− 4 , resulting in the decrease of the electron density of Pd atoms [14]. Wang et al. reported that the n-hexadecane conversion over Pt/ZSM-22 catalysts with Pt(NH 3 ) 4 Cl 2 precursors was~9.1% higher compared to that over catalysts using (Pt(NO 3 ) 2 and H 2 PtCl 6 as the precursors due to the higher platinum dispersion [15]. Yun et al. found that the Ni/Al 2 O 3 catalyst using acetate salt precursors achieved the highest activity (85.6%) in the steam reforming of acetic acid at 450 • C, while the catalytic activity of the catalyst prepared from chloride salts was only 38.7% [16]. These results indicated that the role of precursors should be preferentially considered during the designing and preparation of heterogeneous catalysts.
In this work, copper-cobalt spinel oxides were prepared with a urea hydrothermal method using different copper and cobalt precursors (nitrate, chloride and acetate). The physicochemical properties of these catalysts were characterized by N 2 adsorptiondesorption, X-ray diffraction (XRD), scanning electron microscopy (SEM), X-ray photoelectron spectroscopy (XPS), temperature-programed reduction of H 2 (H 2 -TPR) and temperature-programed desorption of O 2 (O 2 -TPD). The catalytic activity of soot combustion was investigated by temperature-programed oxidation (TPO) experiments under various reaction conditions.

Catalyst Preparation
In this work, a series of CuCo 2 O 4 -x catalysts was synthesized by urea hydrothermal methods with different precursors, where x represented the types of copper and cobalt precursors, e.g., the nitrates (N), acetate (A) and chloride (C), respectively. All chemicals used were of analytical reagent grade and purchased from Macklin Co., Ltd. (Shanghai, China). Taking CuCo 2 O 4 -N as an example, 0.02 mol Cu(NO 3 ) 2 ·3H 2 O, 0.04 mol Co(NO 3 ) 3 ·6H 2 O and 0.06 mol CH 4 N 2 O were dissolved in 100 mL of deionized water and stirred for 1 h at room temperature. Afterwards, the mixed solution was transferred to a polytetrafluoroethylene-lined high pressure reactor placed in a thermostat at 150 • C for hydrothermal reaction. After 12 h, the resulting solution was filtered to obtain a precipitate. The precipitate was then washed three times with deionized water and dried in an oven at 110 • C overnight. Finally, the precipitate was calcined in a muffle furnace at 700 • C for 6 h and sieved to 40-60 meshes. The prepared catalysts were denoted as CuCo 2 O 4 -N, CuCo 2 O 4 -A and CuCo 2 O 4 -C, respectively.

Catalyst Characterizations
N 2 adsorption-desorption analysis was conducted at 77 K to determine the specific surface area, pore size distribution and average pore diameter of the catalysts using TriStar II 3020 (Micromeritics, Norcross, GA, USA). The specific surface areas were calculated using the Brunauer-Emmett-Teller (BET) method. Total pore volume and pore size distribution were calculated via the Barrett-Joyne-Halenda (BJH) method at the relative pressure of p/p 0 = 0.99. The crystal structure of catalysts was observed by X-ray diffraction (XRD) using a diffractometer system (D-Max 2000, Rigaku, Tokyo, Japan) with Cu-Kα radiation operating at 40 kV and 30 mA. All catalysts were scanned in the range of 10 • to 80 • with a step size of 0.02 • . A scanning electron microscopy (SEM) instrument (JSM-7001F, JEOL, Tokyo, Japan) was used to observe the surface morphology of catalysts at an accelerating voltage of 10 kV. The X-ray photoelectron spectra (XPS) were measured with Thermo Escalab 250Xi equipment with monochromatic Al-Ka X-ray radiation at 150 W. All binding energies were calibrated using the C 1s photoelectron peak at 284.8 eV. The redox properties of all catalysts were analyzed using the temperature-programed reduction of the H 2 (H 2 -TPR) apparatus (Autochem II 2920, Micromeritics, Norcorss, GA, USA). To remove impurities, 16 mg catalysts were pretreated at 250 • C for 1 h and cooled down to room temperature before each test. Then, the catalysts were heated from room temperature to 800 • C at a heating rate of 10 • C·min −1 in a 30 mL·min −1 feeding gas flow (10 vol.% H 2 /Ar). The amount of H 2 consumption was calculated based on the H 2 -TPR profiles. The profiles of the temperature-programed desorption of O 2 (O 2 -TPD) were performed using the same apparatus as that for H 2 -TPR. Before the measurement, 200 mg of catalysts were pretreated in an He stream at 200 • C for 1 h and cooled down to room temperature. Then, the adsorption of O 2 was conducted at 70 • C for 1 h in a gas mixture of 3 vol.% O 2 /He (30 mL·min −1 ). Subsequently, the sample was purged by a flowing pure He stream to remove excessive and weakly adsorbed O 2 . Finally, the sample was heated to 800 • C with a heating rate of 10 • C·min −1 in a pure He flow (30 mL·min −1 ), and the desorption profile was recorded. Figure 1 shows a schematic diagram of the experimental setup. The catalytic activity of the CuCo 2 O 4 -x (x = N, C and A) catalysts was evaluated by temperature-programed oxidation (TPO) experiments. Printex-U (Degussa, with the size of 20-30 nm) was used as the model soot in this study. For each test, 180 mg of catalyst powder was mixed with 20 mg of soot in loose (mixing with a spatula for 5 min) or tight (grinding in an agate mortar for 5 min) contact mode. Then, the resulting soot-catalyst mixtures (200 mg) were mixed with 400 mg of inert silica (40-60 meshes) for another 5 min to avoid the formation of hot spots during the reaction. The experimental procedure was as follows: Firstly, the mixture of catalyst and inert silica after reaction was collected from the quartz tube and then placed in an agate mortar. Secondly, 20 mg of soot was mixed with the obtained mixture of the catalyst and inert silica after reaction in loose contact mode. Finally, the above obtained mixture of soot, catalyst and inert silica was filled back into the quartz tube for next catalytic soot combustion reaction cycle. The reaction temperature was increased from 50 • C to 700 • C at a heating rate of 5 • C·min −1 .

Experimental System
All gas streams from the gas cylinders were regulated by mass flow controllers (Sevenstars D07-B, Beijing, China). The mixed gases (10 vol.% O 2 with balanced N 2 ) were fed into the reactor at the flow rate of 200 mL·min −1 during the TPO experiment. In addition, 1000 ppm NO was added into the feeding gas to investigate the effect of NO on catalytic soot combustion when necessary. The outlet concentrations of CO and CO 2 were monitored online using an infrared (IR) gas analyzer (GXH-3010/3011AE, Huayun, Beijing, China) with the accuracy of ±3%. The temperatures at which 10%, 50% and 90% of the soot was oxidized (denoted as T 10 , T 50 and T 90 , respectively) were recorded as indicators of the catalytic activity. Soot conversion (denoted as α) and CO 2 selectivity (denoted as SCO 2 ) were calculated by integrating the CO and CO 2 concentration curves with time as follows: where [CO 2 ] out and [CO] out are the real-time concentrations of CO and CO 2 at the reactor outlet, respectively, and M is the weight of the initially packed soot.  (2) where [CO2]out and [CO]out are the real-time concentrations of CO and CO2 at the reactor outlet, respectively, and M is the weight of the initially packed soot.

Textural Properties of the Catalysts
The specific surface area (SBET), pore volume and pore size of CuCo2O4-x (x = N, A and C) catalysts are obtained with a N2 adsorption-desorption experiment ( Table 1). The CuCo2O4-N catalyst shows the largest specific surface area of 2.1 m 2 ·g −1 , followed by CuCo2O4-A (2.0 m 2 ·g −1 ) and CuCo2O4-C (1.2 m 2 ·g −1 ). The pore volumes of CuCo2O4-N, CuCo2O4-A and CuCo2O4-C catalysts are 4.5 mm 3 ·g −1 , 4.3 mm 3 ·g −1 and 2.5 mm 3 ·g −1 , respectively. The specific surface area and pore volume of the CuCo2O4 catalysts prepared with nitrate and acetate metal salt show no significant differences, while the specific surface area and pore volume of CuCo2O4-C catalyst dramatically decreases. Yu et al. also reported the formation of hydrochloric acid from chloride salts during calcination, which may inhibit the formation of developed pore systems and negatively affect the specific surface area [16]. Compared with conventional metal oxides and supported catalysts, the specific surface area, pore volume and average pore size of the prepared Cu-Co spinel catalysts are rather low. The results could be ascribed to the formation of the well-crystallized spinel structure under high calcination temperature (700 °C) [12,17].

Textural Properties of the Catalysts
The specific surface area (S BET ), pore volume and pore size of CuCo 2 O 4 -x (x = N, A and C) catalysts are obtained with a N 2 adsorption-desorption experiment ( Table 1). The CuCo 2 O 4 -N catalyst shows the largest specific surface area of 2.1 m 2 ·g −1 , followed by CuCo 2 O 4 -A (2.0 m 2 ·g −1 ) and CuCo 2 O 4 -C (1.2 m 2 ·g −1 ). The pore volumes of CuCo 2 O 4 -N, CuCo 2 O 4 -A and CuCo 2 O 4 -C catalysts are 4.5 mm 3 ·g −1 , 4.3 mm 3 ·g −1 and 2.5 mm 3 ·g −1 , respectively. The specific surface area and pore volume of the CuCo 2 O 4 catalysts prepared with nitrate and acetate metal salt show no significant differences, while the specific surface area and pore volume of CuCo 2 O 4 -C catalyst dramatically decreases. Yu et al. also reported the formation of hydrochloric acid from chloride salts during calcination, which may inhibit the formation of developed pore systems and negatively affect the specific surface area [16]. Compared with conventional metal oxides and supported catalysts, the specific surface area, pore volume and average pore size of the prepared Cu-Co spinel catalysts are rather low. The results could be ascribed to the formation of the well-crystallized spinel structure under high calcination temperature (700 • C) [12,17]. Figure 2 shows the XRD patterns of the CuCo 2 O 4 -x (x = N, A and C) catalysts. Sharp and intense diffraction peaks of copper-cobalt spinel phases are observed for all samples, suggesting the formation of well-crystallized structures at high calcination temperature of 700 • C. The diffraction peaks observed at 19 [20].  [19]. No distinct CuO phase is found on the CuCo2O4-C catalyst. Based on the characteristic peak of CuCo2O4 (3 1 1) crystal face the crystal size of the CuCo2O4-x catalysts were calculated using the Scherrer equation (Table 1). The crystal size of the CuCo2O4-N catalyst (38.0 nm) is slightly smaller than tha of CuCo2O4-A and CuCo2O4-C catalysts. Wen et al. also found that the Clcoordination anion was an important contributor to the formation of larger clusters of CuCo2O4 [20].  Figure 3 shows the representative SEM images of the CuCo2O4-x (x = N, A and C catalysts. The CuCo2O4-N catalyst exhibits a sheetlike morphology with tiny spherical par ticles on its surface. The stacking of spherical particles and its sheetlike morphologies pro mote the formation of porous structures (Figure 3a). The phenomenon of particle agglom eration is exhibited over the CuCo2O4-A and CuCo2O4-C catalyst (Figure 3c,e). As shown in Figure 3b, the spherical particles are dispersed on the sheetlike morphology and formed a dendritic structure. However, tiny spherical particles are hardly generated on the sur faces of the CuCo2O4-A and CuCo2O4-C catalysts (Figure 3d,f), while the agglomeration o bulk particles might have resulted in the blockage of the pore systems. These tiny spheri cal particles can improve the contact between the catalyst and soot particles, which can facilitate the utilization of the catalyst active sites for catalytic soot combustion. The par ticle sizes of CuCo2O4-A and CuCo2O4-C catalysts increased obviously compared to the CuCo2O4-N catalysts.

Redox Properties of the Catalysts
The Co 2p spectra of the CuCo2O4-x (x = N, A and C) catalysts are shown in Figure  4a. The peaks of Co 2p3/2 and Co 2p1/2 are observed in the range of 776.0-784.0 eV and 792.0-800.0 eV, respectively [21]. The peaks between 784.0-792.0 eV and 800.0-808.0 eV are attributed to the satellite peaks of Co 2+ [22,23]. The Co 3+ and Co 2+ signals are obtained after the deconvolution of the Co 2p3/2 and Co 2p1/2 spectra. The peaks centered at 779.7 eV and 794.8 eV correspond to the Co 3+ species, while the peaks located at 780.8 eV and 796.4 eV belong to the Co 2+ species [24]. The Cu 2p spectra of all catalysts are also given in Figure 4b. The peaks observed at 932.6 eV belong to the reduced Cu species (Cu + or Cu 0 ), while the prominent signals at 934.6 eV are ascribed to the Cu 2+ species [25]. Additionally, the satellite peaks between 937.0 eV to 946.0 eV also confirm the existence of divalent Cu species [19]. Figure 4c shows the Cu LMM Auger spectra of all CuCo2O4 catalysts. The peaks at the kinetic energy of 912.6 eV correspond to the Cu + species, while the peaks at 917.8 eV are attributed to the Cu 2+ species [26]. However, Cu 0 species are not observed on the Cu LMM Auger spectrum. These results suggest that the reduced copper species on the surfaces of CuCo2O4-x (x = N, A and C) catalysts mainly exists as Cu + . The de-convoluted XPS signals of O 1s are shown in Figure 4d. Two types of oxygen species present on the surface of the CuCo2O4 catalysts. The peaks around 529.8 eV correspond to the lattice oxygen species (Olatt), while the peaks around 531.4 eV are attributed to the adsorbed oxygen species (Oads) [27].

Redox Properties of the Catalysts
The Co 2p spectra of the CuCo 2 O 4 -x (x = N, A and C) catalysts are shown in Figure 4a. The peaks of Co 2p 3/2 and Co 2p 1/2 are observed in the range of 776.0-784.0 eV and 792.0-800.0 eV, respectively [21]. The peaks between 784.0-792.0 eV and 800.0-808.0 eV are attributed to the satellite peaks of Co 2+ [22,23]. The Co 3+ and Co 2+ signals are obtained after the deconvolution of the Co 2p 3/2 and Co 2p 1/2 spectra. The peaks centered at 779.7 eV and 794.8 eV correspond to the Co 3+ species, while the peaks located at 780.8 eV and 796.4 eV belong to the Co 2+ species [24]. The Cu 2p spectra of all catalysts are also given in Figure 4b. The peaks observed at 932.6 eV belong to the reduced Cu species (Cu + or Cu 0 ), while the prominent signals at 934.6 eV are ascribed to the Cu 2+ species [25]. Additionally, the satellite peaks between 937.0 eV to 946.0 eV also confirm the existence of divalent Cu species [19]. Figure 4c shows the Cu LMM Auger spectra of all CuCo 2 O 4 catalysts. The peaks at the kinetic energy of 912.6 eV correspond to the Cu + species, while the peaks at 917.8 eV are attributed to the Cu 2+ species [26]. However, Cu 0 species are not observed on the Cu LMM Auger spectrum. These results suggest that the reduced copper species on the surfaces of CuCo 2 O 4 -x (x = N, A and C) catalysts mainly exists as Cu + . The de-convoluted XPS signals of O 1s are shown in Figure 4d. Two types of oxygen species present on the surface of the CuCo 2 O 4 catalysts. The peaks around 529.8 eV correspond to the lattice oxygen species (O latt ), while the peaks around 531.4 eV are attributed to the adsorbed oxygen species (O ads ) [27].
The relative contents of Cu 2+ and Co 3+ on the surface of CuCo 2 O 4 -x catalysts are given in Table 2 Table 2). The adsorbed oxygen species were more chemically active than lattice oxygen in catalytic soot combustion, indicating a better soot conversion performance over the CuCo 2 O 4 -N catalyst with more O ads species [28].  For the CuCo2O4-N catalyst, two reduction peaks are observed ( Figure 5). The first reduction peak in the range of 150-200 °C is attributed to the reduction of aggregated CuO, while the second peak between 200 °C and 250 °C can be ascribed to the reduction of Cu-Co mixed oxides [29]. The CuCo2O4-A catalyst exhibits shoulder peaks at 200-300 °C, the first reduction peak (244 °C) corresponds to the reduction of Cu-Co mixed oxides, and the latter reduction peak at 278 °C represents the reduction of Co 3+ to Co 2+ [30]. The CuCo2O4-C catalyst also shows two major peaks between 250 °C and 350 °C. The weak reduction peak at 299 °C is ascribed to the reduction of Co 3+ to Co 2+ , while the reduction peak at 331 °C represents the reduction of Co 2+ to Co 0+ [31]. These results suggest that CuCo2O4-N catalyst has better reducibility at relatively low temperatures. The amount of H2 consumption was in the order of CuCo2O4-C > CuCo2O4-A > CuCo2O4-N. Although the CuCo2O4-C catalysts show the higher H2 consumption (14.5 mmol·g −1 ), the H2 consumption of the CuCo2O4-N catalyst is likely mainly concentrated in the low temperature range (150-250 °C). Therefore, the distribution of oxygen species was further investigated.  For the CuCo 2 O 4 -N catalyst, two reduction peaks are observed ( Figure 5). The first reduction peak in the range of 150-200 • C is attributed to the reduction of aggregated CuO, while the second peak between 200 • C and 250 • C can be ascribed to the reduction of Cu-Co mixed oxides [29]. The CuCo 2 O 4 -A catalyst exhibits shoulder peaks at 200-300 • C, the first reduction peak (244 • C) corresponds to the reduction of Cu-Co mixed oxides, and the latter reduction peak at 278 • C represents the reduction of Co 3+ to Co 2+ [30]. The CuCo 2 O 4 -C catalyst also shows two major peaks between 250 • C and 350 • C. The weak reduction peak at 299 • C is ascribed to the reduction of Co 3+ to Co 2+ , while the reduction peak at 331 • C represents the reduction of Co 2+ to Co 0+ [31]. These results suggest that CuCo 2 O 4 -N catalyst has better reducibility at relatively low temperatures. The amount of H 2 consumption was in the order of  and O -, labeled as α-O2) [32], while the oxygen desorption peaks between 600 °C and 850 °C belong to the lattice oxygen species (O 2− , labeled as β-O2) [33]. Figure 6b shows the enlarged O2-TPD profiles in the temperature range of 50 °C to 550 °C. The oxygen desorption peaks within 50-300 °C and 300-500 °C correspond to the physically adsorbed oxygen (α1-O2) and chemically adsorbed oxygen (α2-O2) species, respectively [34,35]. Generally, the adsorption of oxygen followed the procedure of O2→ O -2 →O -→O 2− [27]. The CuCo2O4-N catalyst shows an extra oxygen desorption peak in the temperature range of 200 °C to 250 °C compared to the CuCo2O4-A and CuCo2O4-C. The extra oxygen desorption peak could be attributed to physically adsorbed oxygen (α1-O2), indicating that the catalysts had better oxygen mobility [36]. Similarly, Li et al. also reported that the CuCo2O4 and NiCo2O4 catalysts showed an extra oxygen desorption peak in the temperature range of 200 °C to 250 °C compared to the ZnCo2O4. The catalytic performance of CuCo2O4 and NiCo2O4 catalysts were found to be far superior to that of ZnCo2O4 in toluene combustion [7]. The desorption amount of α1-O2 and α2-O2 species were calculated according to the desorption peaks in the O2-TPD profiles. As shown in Table 2, the amount of O2 desorption was in the order of CuCo2O4-N (41.2 μmol·g −1 ) > CuCo2O4-A (27.1 μmol·g −1 ) > CuCo2O4-C (25.6 μmol·g −1 ), which confirms the existence of more adsorbed oxygen species over the CuCo2O4-N catalyst.    and O -, labeled as α-O2) [32], while the oxygen desorption peaks between 600 °C and 850 °C belong to the lattice oxygen species (O 2− , labeled as β-O2) [33]. Figure 6b shows the enlarged O2-TPD profiles in the temperature range of 50 °C to 550 °C. The oxygen desorption peaks within 50-300 °C and 300-500 °C correspond to the physically adsorbed oxygen (α1-O2) and chemically adsorbed oxygen (α2-O2) species, respectively [34,35]. Generally, the adsorption of oxygen followed the procedure of O2→ O -2 →O -→O 2− [27]. The CuCo2O4-N catalyst shows an extra oxygen desorption peak in the temperature range of 200 °C to 250 °C compared to the CuCo2O4-A and CuCo2O4-C. The extra oxygen desorption peak could be attributed to physically adsorbed oxygen (α1-O2), indicating that the catalysts had better oxygen mobility [36]. Similarly, Li et al. also reported that the CuCo2O4 and NiCo2O4 catalysts showed an extra oxygen desorption peak in the temperature range of 200 °C to 250 °C compared to the ZnCo2O4. The catalytic performance of CuCo2O4 and NiCo2O4 catalysts were found to be far superior to that of ZnCo2O4 in toluene combustion [7]. The desorption amount of α1-O2 and α2-O2 species were calculated according to the desorption peaks in the O2-TPD profiles. As shown in Table 2, the amount of O2 desorption was in the order of CuCo2O4-N (41.2 μmol·g −1 ) > CuCo2O4-A (27.1 μmol·g −1 ) > CuCo2O4-C (25.6 μmol·g −1 ), which confirms the existence of more adsorbed oxygen species over the CuCo2O4-N catalyst.

Soot Conversion in O 2 /N 2
The catalytic activity of CuCo 2 O 4 -x (x = N, A and C) catalysts for soot combustion were studied using a TPO experiment under the loose contact mode. Firstly, the soot conversion was investigated under the carrier gases of 10 vol.% O 2 balanced with N 2 . The characteristic temperatures (T 10 , T 50 and T 90 ) of soot conversion are 530 • C, 586 • C and 614 • C, respectively in the absence of a catalyst, which is much higher than those in the presence of the CuCo 2 O 4 catalysts (Figure 7 and Table 3 The CuCo 2 O 4 -N and CuCo 2 O 4 -A catalysts exhibit larger specific surface area and pore volume compared to the CuCo 2 O 4 -C catalyst. It is widely recognized that a larger specific surface area could facilitate the contact between gaseous reactants (e.g., O 2 , NO and NO 2 ) and the catalyst [37,38]. SEM images further confirmed the generation of more developed pore structure systems of the CuCo 2 O 4 -N catalyst compared to the CuCo 2 O 4 -A and CuCo 2 O 4 -C catalysts. Fang et al. reported that a perovskite-type macro/mesoporous La 1-x K x FeO 3-δ catalysts with large specific surface area and pore volume could improve the utilization of catalytic sites in the soot combustion reaction [39]. Furthermore, the presence of an appropriate amount of single metal oxide CuO on the CuCo 2 O 4 surface could contribute to the crystal lattice distortion of the spinel phase and promote the formation of a defect structure, thereby improving the catalytic activity of the soot combustion reaction [40].

Soot Conversion in O2/N2
The catalytic activity of CuCo2O4-x (x = N, A and C) catalysts for soot combustion were studied using a TPO experiment under the loose contact mode. Firstly, the soot conversion was investigated under the carrier gases of 10 vol.% O2 balanced with N2. The characteristic temperatures (T10, T50 and T90) of soot conversion are 530 °C, 586 °C and 614 °C , respectively in the absence of a catalyst, which is much higher than those in the presence of the CuCo2O4 catalysts (Figure 7 and Table 3). The catalysts prepared with different metal salt precursors exhibit different catalytic activity in the soot conversion. Among them, the CuCo2O4-N catalyst achieves the highest soot conversion, while the values of T10, T50 and T90 are 451 °C, 520 °C and 558 °C, respectively. The T50 values for the CuCo2O4x (x = N, A and C) catalysts followed the order of CuCo2O4-N (520 °C) < CuCo2O4-A (539 °C) < CuCo2O4-C (550 °C) < no catalysts (586 °C). The CO2 selectivity is also improved remarkably from 81.8% over the CuCo2O4-C to almost 100% over the CuCo2O4-N catalysts.  The CuCo2O4-N and CuCo2O4-A catalysts exhibit larger specific surface area and pore volume compared to the CuCo2O4-C catalyst. It is widely recognized that a larger specific surface area could facilitate the contact between gaseous reactants (e.g., O2, NO and NO2) and the catalyst [37,38]. SEM images further confirmed the generation of more developed pore structure systems of the CuCo2O4-N catalyst compared to the CuCo2O4-A and CuCo2O4-C catalysts. Fang et al. reported that a perovskite-type macro/mesoporous La1-xKxFeO3-δ catalysts with large specific surface area and pore volume could improve the utilization of catalytic sites in the soot combustion reaction [39]. Furthermore, the presence of an appropriate amount of single metal oxide CuO on the CuCo2O4 surface could contribute to the crystal lattice distortion of the spinel phase and promote the formation of a defect structure, thereby improving the catalytic activity of the soot combustion reaction [40].  The redox properties of CuCo 2 O 4 -x catalysts played a crucial role in catalytic soot combustion. The higher relative content of Co 3+ /Co total (38.4%) and Cu 2+ /Cu total (51.7%) obtained over CuCo 2 O 4 -N catalysts were much higher compared to the CuCo 2 O 4 -A (35.1% and 48.2%) and CuCo 2 O 4 -C (33.8% and 47.8%), respectively. Spinel-type CuCo 2 O 4 catalysts possessed outstanding redox properties since the synergistic effects between Co 3+ and Cu 2+ species could enhance the adsorption-activation properties of oxygen species for soot conversion [41]. Abundant Co 3+ species would increase the anionic defects on catalyst surfaces, leading to the formation of more oxygen vacancies [42]. Zhang et al. reported that the presence of Cu 2+ species induced structural defects on cobalt oxide and weakened the Co-O bonds, which could facilitate the activation of oxygen and improve the reducibility of CuCo 2 O 4 [43]. In addition, the CuCo 2 O 4 -N catalysts showed the highest relative content of O ads /(O ads + O latt ) (46.8%) compared to the CuCo 2 O 4 -A (37.2%) and CuCo 2 O 4 -C (36.5%). The O ads species possessed better mobility than O latt species and could participate in catalytic soot combustion via the contact points between catalyst pellets and soot particulates [35].
The lower reduction peaks temperature proved that the CuCo 2 O 4 -N catalysts (173 • C) has excellent low temperature reduction performances compared to the CuCo 2 O 4 -A (184 • C) and CuCo 2 O 4 -C (247 • C). The total H 2 consumption amount of the CuCo 2 O 4 -N catalysts (13.4 mmol·g −1 ) was slightly lower than that of CuCo 2 O 4 -A (13.6 mmol·g −1 ) and CuCo 2 O 4 -C (14.5 mmol·g −1 ). The CuCo 2 O 4 -N catalyst possessed more adsorbed oxygen species (41.2 µmol·g −1 ) compared to the CuCo 2 O 4 -A (27.1 µmol·g −1 ) and CuCo 2 O 4 -C (25.6 µmol·g −1 ). It was suggested that the CuCo 2 O 4 -N catalysts could release more adsorbed oxygen species below 500 • C [44]. Compared with lattice oxygen species, adsorbed oxygen species were more important in soot conversion due to their better oxygen mobility [45]. He et al. also reported that the adsorbed oxygen species released at low temperatures were more important than lattice oxygen species since the lattice oxygen species could be activated and released only at high temperatures [35]. Lower reduction temperature and abundant adsorbed oxygen species may be crucial factors for soot conversion over the CuCo 2 O 4 -N catalyst at lower temperature.

Soot Conversion in NO/O 2 /N 2
NO is one of the main components in diesel engine exhaust and has a great effect on soot conversion. Therefore, soot conversion was investigated over the CuCo 2 O 4 -x (x = N, A and C) catalysts in the presence of 1000 ppm NO ( Figure 8 and Table 4). The addition of NO resulted in lower T 10 , T 50  Using CuCo2O4-N as an example, the T50 value decreases from 520 °C to 414 °C with NO addition, while the CO2 selectivity also decreases from 98.8% to 96.6%. NO could be oxidized with O2 to form NO2 in gas phase (Reaction (3)) [46]. NO could be also adsorbed on the surface of CuCo2O4-N catalyst and then oxidized by the active oxygen to form NO2  Using CuCo 2 O 4 -N as an example, the T 50 value decreases from 520 • C to 414 • C with NO addition, while the CO 2 selectivity also decreases from 98.8% to 96.6%. NO could be oxidized with O 2 to form NO 2 in gas phase (Reaction (3)) [46]. NO could be also adsorbed on the surface of CuCo 2 O 4 -N catalyst and then oxidized by the active oxygen to form NO 2 species, while could promote the formation of surface oxygen vacancies (Reaction (4)). NO 2 is a stronger oxidant and participate in catalytic soot combustion as a mobile gaseous oxidant [46]. Possible pathways for the reaction of NO 2 with soot are shown in Reaction (5) [47]. NO 2 could attack the soot on the surface, resulting in the generation of CO and NO, while CO was eventually oxidized by O 2 to CO 2 , with NO again participating in the NO xassisted catalytic soot combustion [37]. However, the slight decrease in CO 2 selectivity could be attributed to the rapid oxidation of soot by NO 2 , resulting in the incomplete oxidation of soot to CO [48].
where O* represents the active oxygen species, O v represents the surface oxygen vacancy and C(Soot) represents the soot particulate.

Effect of the Contact Mode
The contact mode between the catalyst and soot could greatly affect the activity of soot conversion [41]. The soot conversion and CO 2 selectivity were investigated under both loose and tight contact modes over the CuCo 2 O 4 -N catalyst ( Figure 9 and Table 5). Under the tight contact mode, soot conversion is accelerated at lower temperatures regardless of the presence of NO compared to the loose contact mode, while the CO 2 selectivity reaches up to 100%. The increase of catalyst-soot contact points in the tight contact mode could contribute to better utilization of active sites on the catalyst surface compared to the loose contact mode [4,27]. Machida et al. found that the utilization/transfer of lattice oxygen species was more efficient under the tight contact mode. Besides, the case of "O 2 slip" may be decreased in the tight contact mode, which increases the utilization of the released oxygen species to the soot combustion [49].

Stability Test
The stability test of CuCo2O4-N catalysts was conducted for four consecutive cycles of soot combustion. As shown in Figure 10, the characteristic temperatures (T10, T50 and T90) increased slightly with the increase in the number of cycles of soot combustion. For examples, the values of T50 for catalytic soot combustion for each cycle are 520 °C, 524 °C, 526 °C and 528 °C, respectively. Figure 11 shows the deconvoluted XPS spectra of O 1s for the CuCo2O4-N catalyst after four consecutive cycles of soot combustion. The relative content of Oads/(Oads + Olatt) of CuCo2O4-N catalyst decreases from 46.8% to 40.6% ( Table 6). The slight decrease in soot catalytic activity may be due to the minor attenuation of the relative content of Oads/(Oads + Olatt). Chen et al. also reported the catalytic soot performance of Ag/Co3O4 and found that the decrease in adsorbed oxygen species was an important reason for the decrease of catalytic soot combustion [50].

Stability Test
The stability test of CuCo 2 O 4 -N catalysts was conducted for four consecutive cycles of soot combustion. As shown in Figure 10, the characteristic temperatures (T 10 , T 50 and T 90 ) increased slightly with the increase in the number of cycles of soot combustion. For examples, the values of T 50 for catalytic soot combustion for each cycle are 520 • C, 524 • C, 526 • C and 528 • C, respectively. Figure 11 shows the deconvoluted XPS spectra of O 1s for the CuCo 2 O 4 -N catalyst after four consecutive cycles of soot combustion. The relative content of O ads /(O ads + O latt ) of CuCo 2 O 4 -N catalyst decreases from 46.8% to 40.6% ( Table 6). The slight decrease in soot catalytic activity may be due to the minor attenuation of the relative content of O ads /(O ads + O latt ). Chen et al. also reported the catalytic soot performance of Ag/Co 3 O 4 and found that the decrease in adsorbed oxygen species was an important reason for the decrease of catalytic soot combustion [50].

Stability Test
The stability test of CuCo2O4-N catalysts was conducted for four consecutive cycles of soot combustion. As shown in Figure 10, the characteristic temperatures (T10, T50 and T90) increased slightly with the increase in the number of cycles of soot combustion. For examples, the values of T50 for catalytic soot combustion for each cycle are 520 °C, 524 °C, 526 °C and 528 °C, respectively. Figure 11 shows the deconvoluted XPS spectra of O 1s for the CuCo2O4-N catalyst after four consecutive cycles of soot combustion. The relative content of Oads/(Oads + Olatt) of CuCo2O4-N catalyst decreases from 46.8% to 40.6% ( Table 6). The slight decrease in soot catalytic activity may be due to the minor attenuation of the relative content of Oads/(Oads + Olatt). Chen et al. also reported the catalytic soot performance of Ag/Co3O4 and found that the decrease in adsorbed oxygen species was an important reason for the decrease of catalytic soot combustion [50].    Table 6. XPS parameters of CuCo2O4-N catalyst before and after reaction.

Reaction Mechanisms for Catalytic Soot Combustion
The potential reaction pathways of catalytic soot combustion over CuCo2O4 were discussed. The excellent catalytic soot combustion of the copper-cobalt spinel catalyst performance depended on the redox properties of the catalysts [41]. The interactions between Cu and Co species in the CuCo2O4 catalyst played a crucial role in the redox reactions. Previous studies showed that the redox pairs of Cu 2+ /Cu + and Co 3+ /Co 2+ were involved in the electron transfer process from Co 3+ to Cu 2+ within the Cu 2+ -O-Co 3+ connections in the copper-cobalt spinel catalysts [31,51]. The Cu 2+ -O-Co 3+ connections could bridge the oxygen transfer within the structure and reduce the redox potential of the Cu species, which ensures the improvement of reducibility for both Cu and Co oxides in the Cu-Co catalysts (Reaction (6)), facilitating the NOx-assisted mechanism in the soot combustion reaction [46].
Catalytic soot combustion over the copper-cobalt spinel catalyst followed the Marsvan Krevelen (MvK) mechanism, in which the abundance of active oxygen species directly determined the performance of catalytic soot combustion [5]. At the beginning of soot conversion, plenty of active oxygen species on the catalyst surface would come into contact with the soot, resulting in soot combustion at a relatively low temperature and the subsequent formation of oxygen vacancies (Ov) (Reaction (7)) [49]. As the reaction proceeded, the consumed reactive oxygen species could be replenished in two ways. Firstly, the reduction of high-valence Cu 2+ and Co 3+ to low-valence Cu + and Co 2+ released the oxygen species and formed oxygen vacancies, making the surface of the catalyst ready for the oxygen species adsorption from the gas phase (O2(gas)) and accelerating the conversion of the gas phase oxygen species to the adsorbed oxygen species (Reactions (8) and (9)) [7]. The relative content of adsorbed oxygen species of the used CuCo2O4-N catalysts (40.6%) was significantly decreased compared to the fresh CuCo2O4-N catalysts (46.8%), and the activity of soot conversion was also decreased. The adsorbed oxygen species played a crucial role in catalytic soot combustion, which could be transformed to active oxygen species such as O2, O -2 and O − (Reactions (10)-(12)) [52]. Moreover, some lattice oxygen species took over the oxygen vacancies and were transferred into adsorbed oxygen species. The adsorbed oxygen species were spilled over to the soot particle surface and further

Reaction Mechanisms for Catalytic Soot Combustion
The potential reaction pathways of catalytic soot combustion over CuCo 2 O 4 were discussed. The excellent catalytic soot combustion of the copper-cobalt spinel catalyst performance depended on the redox properties of the catalysts [41]. The interactions between Cu and Co species in the CuCo 2 O 4 catalyst played a crucial role in the redox reactions. Previous studies showed that the redox pairs of Cu 2+ /Cu + and Co 3+ /Co 2+ were involved in the electron transfer process from Co 3+ to Cu 2+ within the Cu 2+ -O-Co 3+ connections in the copper-cobalt spinel catalysts [31,51]. The Cu 2+ -O-Co 3+ connections could bridge the oxygen transfer within the structure and reduce the redox potential of the Cu species, which ensures the improvement of reducibility for both Cu and Co oxides in the Cu-Co catalysts (Reaction (6)), facilitating the NO x -assisted mechanism in the soot combustion reaction [46].
Catalytic soot combustion over the copper-cobalt spinel catalyst followed the Mars-van Krevelen (MvK) mechanism, in which the abundance of active oxygen species directly determined the performance of catalytic soot combustion [5]. At the beginning of soot conversion, plenty of active oxygen species on the catalyst surface would come into contact with the soot, resulting in soot combustion at a relatively low temperature and the subsequent formation of oxygen vacancies (O v ) (Reaction (7)) [49]. As the reaction proceeded, the consumed reactive oxygen species could be replenished in two ways. Firstly, the reduction of high-valence Cu 2+ and Co 3+ to low-valence Cu + and Co 2+ released the oxygen species and formed oxygen vacancies, making the surface of the catalyst ready for the oxygen species adsorption from the gas phase (O 2(gas) ) and accelerating the conversion of the gas phase oxygen species to the adsorbed oxygen species (Reactions (8) and (9)) [7]. The relative content of adsorbed oxygen species of the used CuCo 2 O 4 -N catalysts (40.6%) was significantly decreased compared to the fresh CuCo 2 O 4 -N catalysts (46.8%), and the activity of soot conversion was also decreased. The adsorbed oxygen species played a crucial role in catalytic soot combustion, which could be transformed to active oxygen species such as O 2 , O − 2 and O − (Reactions (10)-(12)) [52]. Moreover, some lattice oxygen species took over the oxygen vacancies and were transferred into adsorbed oxygen species. The adsorbed oxygen species were spilled over to the soot particle surface and further contributed to the soot combustion reaction [35]. All these active oxygen species could directly participate in catalytic soot combustion reaction via the contact points between catalysts and soot particulates to form CO and CO 2 [37]. In addition, the released active oxygen species for the conversion of NO oxidation of NO 2 [42]. Moreover, the Cu + and Co 2+ species were reoxidized to Cu 2+ and Co 3+ due to the replenishment of oxygen vacancies and participation in catalytic soot combustion (Reaction (13)), and a portion of the gaseous NO molecules were also converted to NO 2 (Reaction (4)) [53]. Hence, the facilitated electron transfer process and abundant adsorbed oxygen species endowed the copper-cobalt spinel catalysts with outstanding soot oxidation activity.

Conclusions
To obtain a fundamental understanding on the catalytic soot combustion performance of the CuCo 2 O 4 catalysts prepared with different metal salt precursors, the relationships between the textural and redox properties of the catalysts and soot conversion were investigated.  Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.