Enhanced Soot Oxidation Activity of a CuO-Doped CeO₂ Catalyst via Acid Etching

Abstract

Copper oxides tend to agglomerate on the surface of CeO₂, with a high amount of Cu. In this study, a CeO₂ catalyst with a high CuO doping amount was treated with nitric acid to improve its catalytic performance for soot oxidation. The effect of acid etching on the structural properties of the CuO-doped CeO₂ catalyst were elucidated. The characterization results indicated that aggregated CuO particles formed over CuCe. The acid etching resulted in a remarkable increase in the surface area of CuCe. Additionally, acid etching promoted the formation of surface-adsorbed oxygen species and oxygen vacancy, and reduced the content of CuO_x species with weak interaction with CeO₂. The soot temperature-programmed oxidation results show the acid etching of CuCe catalyst could reduce the T₅₀ from 443 to 383 °C. The isothermal reaction results also suggest that acid etching of CuCe leads to an increase in reaction rate from 16.2 to 46.0 μmol min⁻¹ g⁻¹.

1. Introduction

The soot particles emitted by diesel engines can trigger respiratory and cardiovascular diseases, leading to severe health problems [1,2]. The most common after-processing technology for hindering soot discharge is the use of diesel particulate filters (DPFs) to trap particulates. The diesel exhaust temperature (200–500 °C) cannot meet the ignition temperature of soot (>600 °C) [3]. Hence, oxidation catalysts must be applied for catalyzing soot oxidation at lower temperatures. In the past decades, a good diversity of catalysts such as precious metal catalysts [4], CeO₂-based oxides [5], and transition metal oxides [6] have been applied to soot removal for achieving DPFs regeneration. Because of the scarcity of precious metals, there is always an increasing impetus to exploit cost-effective and active alternatives to noble metal catalysts. In comparison with noble metals, inexpensive CeO₂ and transition metal oxides are promising catalysts for soot combustion [7,8].

It was reported that modification of transition metal oxides via wet chemistry or impregnation methods is an efficient way of improving the redox ability of CeO₂ catalysts [9]. Replacing Ce⁴⁺ with Fe^x+, Co^x+, and Mn^x+ can promote the formation of oxygen vacancies, thereby accelerating the diffusion of surface and bulk oxygen species and achieving improved soot oxidation activities [8,10]. Among transition metals, Cu mainly exists in three valence states of Cu⁰, Cu⁺ and Cu²⁺, which can promote the reduction/oxidation process of ceria via the Mars–van Krevelen (Mv-K) mechanism [11,12]. Liang et al. [13] suggested that Cu can improve the oxidation reaction by activating the O₂ molecules on the surface and regulating the oxygen spillover to the ceria support. Due to the unique catalytic performance of Cu, it has gained tremendous attention from researchers. For example, the binary CuO-CeO₂ system has been widely investigated in many catalytic reactions, such as soot oxidation [14], CO oxidation [15], and volatile organic compound oxidation [16].

Liang et al. [13] prepared CuO-doped CeO₂ catalysts for soot oxidation using the sol-gel method. The results showed that CuO_x nanoclusters highly dispersed on the surface of CeO₂ are responsible for the activation of lattice oxygen, resulting in the enhancement of the soot oxidation activity of CuO-CeO₂ (T_m = 356 and 522 °C in 10% O₂ under tight and loose contact conditions). Wang et al. [17] suggested that highly dispersed copper species favor the enhanced redox capability of three-dimensionally ordered macroporous CuO–CeO₂ catalysts at low temperatures, thus resulting in an enhanced soot oxidation (T_m = 551 and 427 °C in 5.0% O₂ and 2500 ppm NO/5.0% O₂ under loose contact conditions). Andana et al. [18] found that CuO/CeO₂ catalysts are active for soot oxidation at low temperatures under an NO atmosphere due to the early NO evolution (T₅₀ = 552 and 534 °C in 10% O₂ and 550 ppm NO/10% O₂ under loose contact conditions). Piumetti et al. [19] fabricated a set of Ce_1-xCu_xO₂ (x ranging from 0.05 to 0.95) oxide catalysts, and evaluated the soot oxidation activity. The results showed that the Ce₀.₉₅Cu₀.₀₅O₂ binary oxide catalyst exhibits the best catalytic activity (T_m = 384 °C in 50% air/50% N₂ under tight contact conditions). Cui et al. [20] synthesized CuO-doped CeO₂ nanosheets and evaluated their soot oxidation performance. The results demonstrated that CuO doping promoted the formation of oxygen species on the surface of CeO₂, and enhanced the soot oxidation activity (T₅₀ = 312 °C in 5% O₂ under tight contact conditions; T₅₀ = 360 °C in 1000 ppm/5% O₂ under loose contact conditions). However, excessive CuO doping would suppress the catalytic activity of CeO₂. Acid etching of catalytic materials can improve structural properties, thus influencing catalytic activity. For example, Zhao et al. [21] synthesized high-surface-area perovskite oxides LaMnO_3+δ via an acid-etching method, and favorable NO catalytic oxidation activity was observed on the etched catalysts. Liang et al. [22] also found that nitric acid etching of Ag@CeO₂ significantly increased the defect sites of CeO₂ and improved the electrocatalytic performance.

In this study, a CuO-doped CeO₂ catalyst was prepared via a hydrothermal method, and was then subjected to an acid-etching process. The intrinsic reactivity of the catalysts for soot oxidation was evaluated via temperature programmed oxidation (TPO) and an isothermal reaction model under tight contact conditions. The results showed that acid etching significantly improved the soot oxidation activity of the CuCe catalyst, and the corresponding mechanism of acid etching was discussed.

2. Results and Discussion

2.1. Catalytic Activities for Soot Oxidation

The intrinsic activities of the catalysts were evaluated by conducting a soot–TPO reaction under tight contact conditions. Soot conversion as a function of reaction temperature over the as-prepared catalysts is profiled in Figure 1a, and the corresponding T₅₀ data are listed in

Table 1. Reaction parameters of the catalysts for soot oxidation.

Catalyst	T50 (°C) a	SCO2 (%) a	Reaction Rate (μmol min−1 g−1) b	Ea (kJ/mol) c
CeO2	412	99.8	30.8	96
CuCe	443	99.0	16.2	110
CuCe-AE	383	99.5	46.0	83

^a Obtained from soot-TPO. ^b Determined by isothermal reaction at 310 °C. ^c Determined by Arrhenius equation.

. All the catalysts exhibit a high CO₂ selectivity (CO₂/CO_x > 99%). The T₅₀ values of CeO₂ and CuCe are 412 and 443 °C, respectively, revealing that a high Cu loading level suppresses soot oxidation over CeO₂. However, the CuCe-AE catalyst exhibits a much lower T₅₀ value of 383 °C, displaying remarkably enhanced soot oxidation activity. Such a catalytic performance is normal among the Ce-based catalysts reported in the literature, as compared in

Table 2. Comparison of activity of CuCe-AE with Ce-based catalysts previously reported in the literature under tight contact conditions.

Catalyst	Reactant Gas	Flow Rate (mL/min)	T50 (°C)	Reference
CuCe-AE	10%O2/N2	80	383	This work
Sn0.7Ce0.3O2	O2/N2	268	437	3
Ru/CeO2	10%O2/N2	80	386	4
Cu-Ce	10%O2/N2	500	356	13
CuO/Ce-DDA	Air	-	413	14
Ce0.95Cu0.05O2	10%O2/N2	100	384	19
Cu0.1Ce0.9O2-NF	5%O2/Ar	50	310	20

, in which some important reaction parameters are also listed.

To further characterize the intrinsic activities of the as-prepared catalysts, the isothermal reactions were performed under tight contact conditions at 310 °C. At a low conversion rate (<15%), soot oxidation can be considered a surface reaction, while soot oxidation at a high conversion rate is limited by diffusion. Figure 1b shows the soot conversion amount as a function of reaction time. The reaction rate can be obtained from the slope of the lines after linear fitting, and the results are listed in Table 1. Similar to the soot–TPO results, a high Cu loading level can inhibit the reaction rate of CeO₂. After the CuCe was subjected to acid etching, the reaction rate increased from 16.2 to 46.0 μmol min⁻¹ g⁻¹, which exceeds that of bare CeO₂. These data confirm that the acid etching prompts the intrinsic reactivity of the CuO-doped CeO₂ catalyst. The E_a of catalysts was also calculated using the Arrhenius method. The Arrhenius plots of lnr vs. 1/T are shown in Figure 1c, and the corresponding parameters are listed in Table 1. The E_a of the catalysts follows a sequence of CuCe-AE (83 kJ/mol) < CeO₂ (96 kJ/mol) < CuCe (110 kJ/mol), which is in accordance with the catalytic performance in Figure 1a,b. Nevertheless, the E_a differences are not large, indicating that they conform with an identical soot oxidation reaction mechanism. To determine the stability of CuCe-AE, four consecutive cycles under the same conditions were conducted, with the results shown in Figure 1d. An obvious increase of 18 °C in the T₅₀ value is observed for the second cycle, which is believed to be related importantly to the decrease in the surface area and sintering of the oxides in the used catalyst. There is no significant deactivation after the third cycle, although a small increase in the T₅₀ (approximately 1–4 °C) occurs, and the catalytic behavior of the cycled CuCe-AE catalyst is still better than those of the fresh CeO₂ and CuCe. These observations indicate the potential reusability of CuCe-AE catalyst in soot combustion.

2.2. Physicochemical Characterization of Catalyst

Table 3. Structural and textural properties of the catalysts.

Catalyst	Cu Content (wt.%) a	D (nm) b	SBET (m2/g) c	Vpore(cm3/g) c	ID/IF2g d
CeO2	/	10.5	104	0.30	0.32
CuCe	20.0	5.8	98	0.22	0.38
CuCe-AE	2.3	5.4	145	0.27	0.40

^a Determined by ICP-AES. ^b Crystallite size estimated via the Scherrer equation using the (111) peak of CeO₂. ^c Determined by N₂ sorption. ^d Obtained from Raman spectra.

summarizes the structural and textural features of the catalysts. As determined by ICP-AES, the Cu content is 20.0 and 2.3 wt.% for CuCe and CuCe-AE, respectively. Obviously, the acid etching significantly reduces the Cu content in the CuO-doped mixed oxides.

The TEM images were collected to observe the morphologies of the catalysts. As shown in Figure 2, CeO₂ prepared by alkaline hydrothermal method displays a typical nanorod-like morphology (Figure 2a). Upon the addition of Cu(NO₃)₂, the CuCe (Figure 2b) and especially CuCe-AE (Figure 2c) exhibit irregular shape. Aggregated CuO particles with a size larger than 15 nm can be clearly observed on CuCe. Figure 2d illustrates the XRD pattern of the as-prepared catalysts. The diffraction peaks at 2θ of 28.5°, 33.1°, 47.5°, 56.3°, 59.1°, 69.4°, 76.7° and 79.1° are indexed to the (111), (200), (220), (311), (222), (400), (331) and (420) planes of CeO₂ (JCPDS 34-0394). The half widths of the diffraction peak for CuO-doped catalysts are broader than pure CeO₂, indicative of a decrease in the crystallite size of the oxides. The crystallite sizes of the catalysts were calculated via the Scherrer equation, and the results are listed in Table 3. The crystallite sizes of CeO₂, CuCe and CuCe-AE are 10.5, 5.8 and 5.5 nm, respectively. In addition, the typical diffraction peaks of CuO particles appear at 2θ of 35.6 and 38.7^o for the CuCe catalyst [19], suggesting the agglomeration of CuO at a high doping level. However, the diffraction peaks of CuO are not identified for CuCe-AE, implying that acid etching can remove the agglomerated CuO particles, while the highly dispersed CuO_x clusters with a size below the XRD detection limit still exist.

To further identify Cu distribution on the catalyst surface, EDS-mapping analyses of the mixed oxides before and after acid etching were carried out, and the received images are compiled in Figure 3. The markedly agglomerated CuO particles are observed on CuCe, consistent with the XRD results. In these Cu-rich regions, a lower concentration of cerium occurs, indicating the non-uniform distribution of metal oxides. For CuCe-AE, the highly even Cu distribution is consistent with that of Ce, implying the removal of aggregated CuO particles and the reservation of homogeneously distributed mixed oxides.

The N₂ sorption isotherms and BJH pore size distribution of the catalysts are shown in Figure 4, and the BET specific surface area and pore volume results are given in Table 3. The N₂ sorption isotherms of all the catalysts show a type IV isotherm with H3 hysteresis loop, which is mainly caused by the irregular pores generated via particle accumulation. The measured pore volume is predominantly generated by the gap of primary particles. Notably, the specific surface area of CuCe-AE is about 40 m²/g higher than that of CuCe, suggesting that acid etching can enlarge the specific surface area of CuO-doped CeO₂. A similar phenomenon was also observed by Zhao and co-workers [21]. They found that the acid-etching method not only maintained the perovskite-type structure of the LaMnO_3+δ catalyst, but also improved its surface area. It is generally accepted that a large specific area facilitates the contact between the soot and catalyst, being therefore beneficial for soot oxidation [23].

XPS analysis was used to disclose the chemical state of Ce, O, and Cu on the catalyst surface. As listed in

Table 4. XPS results of the catalysts.

Catalyst	Surface Cu/Ce	Ce 3d		Ce3+/Ce4+ (%)	O 1s		Oα/Oβ (%)
		Ce3+ (%)	Ce4+ (%)		Oα (%)	Oβ (%)
CeO2	-	18.3	81.7	22.4	28	72	38.8
CuCe	0.95	16.6	83.4	19.9	25	75	33.3
CuCe-AE	0.31	19.2	80.8	23.8	38	62	61.3

, the atom ratios of Cu/Ce on the surface of CuCe and CuCe-AE catalysts are 0.95 and 0.31, respectively. The measured values are not only determined by the actual content of surface elements, but also affected by the dispersion of metal oxides. Figure 5 shows the XPS full spectra of the catalysts and the fine spectrum in the Ce 3d, O 1s and Cu 2p regions, and the deconvolution results are listed in Table 4. In the spectra of Ce 3d in Figure 5b, six peaks assigned to Ce⁴⁺ 3d (u_1–6: 882.4, 888.6, 898.3, 900.8, 907.5, and 916.7 eV) and two peaks attributed to Ce³⁺ 3d (v_1–2: 884.5 and 902.0 eV) are observed after deconvolution [24,25]. The ratio of Ce³⁺/Ce⁴⁺ decreases after CuO doping. This is due to the fact that the electronegativity of Cu (1.90) is stronger than that of Ce (1.12). After copper doping, electrons transfer from Ce to Cu (Cu²⁺ + Ce³⁺→Cu⁺ + Ce⁴⁺), resulting in a decrease in the ratio of Ce³⁺/Ce⁴⁺. A similar phenomenon was observed by Piumetti and co-workers [19]. It is generally accepted that surface oxygen vacancy is normally correlated to the content of Ce³⁺ [26]. The ratios of Ce³⁺/Ce⁴⁺ for CuCe and CuCe-AE catalysts are 19.9% and 23.8%, respectively, suggesting that acid etching could increase the amount of surface oxygen vacancy on the CuO-doped CeO₂. The increased Ce³⁺/Ce⁴⁺ ratio for CuCe-AE leads to an expansion of the ceria crystal cell compared with CuCe.

Figure 5c shows the Cu 2p spectra of CuCe and CuCe-AE. The two main peaks of CuCe at the binding energies of 953 and 933 eV correspond to Cu 2p_1/2 orbital and Cu 2p_3/2 orbital, respectively, accompanied with two satellite peaks at the binding energies of 942 and 962 eV [27]. This result is similar to the spectrum of CuO [28], which indicates that Cu species mainly exist in the form of CuO. The CuCe-AE catalyst demonstrates weak signals in the Cu 2p region, meaning that the surface Cu content is relatively low.

Two oxygen species can be observed in the O 1s spectra in Figure 5d. The binding energy of 529–530 eV is characteristic of lattice oxygen (denoted as O_β), and the binding energy in the 531–532 eV region is related to surface oxygen species (including -OH, ${\mathrm{O}}_2^{-}$ and ${\mathrm{O}}_2^{2-}$, denoted as O_α) [29,30]. The relative content of surface oxygen species can be roughly estimated by the ratio of O_α/O_β. The CuO doping results in a decrease in the proportion of O_α/O_β. This is because a great quantity of agglomerated CuO particles on the surface of CeO₂ leads to the surface oxygen species being covered. Notably, acid etching removes these aggregated CuO particles and recovers surface oxygen species to a great extent within the amount of CuO-doped CeO₂.

Raman spectroscopy of CeO₂ and CuO-doped catalysts was carried out to estimate the relative concentration of oxygen vacancies in CeO₂ lattice with an indicative at the wavelength of 532 nm [31]. As displayed in Figure 6a, the as-prepared catalysts exhibit a strong peak at ~456 cm⁻¹ and two weak peaks at ~600 and 1175 cm⁻¹, which can be attributed to the F_2g, defect-induced (D), and second-order longitudinal optical (2LO) modes of the fluorite phase, respectively [32,33]. The ratios of I_D/I_F2g, as a characteristic of bulk oxygen vacancy concentration, are listed in Table 3. In comparison with pure CeO₂, CuO-doped CeO₂ shows a higher I_D/I_F2g ratio. The enhanced ratio of bulk oxygen vacancy in CuO-doped CeO₂ can be ascribed to the incorporation of smaller copper ions into the CeO₂ lattice (r_Cu^x+ < r_Ce^x+). As a result, CuO doping could promote the formation of bulk oxygen vacancy in CeO₂. The I_D/I_F2g ratios of CeO₂, CuCe and CuCe-AE are 0.32, 0.38 and 0.40, respectively. The CuO doping increases the bulk oxygen vacancy concentration in the mixed oxides due to distortion of the ceria crystal cell, and the acid etching creates more bulk oxygen vacancies by removing aggregated CuO particles and strengthening the interaction of highly dispersed CuO_x clusters with ceria. As we know, surface oxygen vacancy is essential for soot oxidation reaction [34]; the conclusion that enhanced bulk oxygen vacancy leads to the boosted catalytic activity may be not strict. However, the increase in bulk oxygen vacancy can accelerate the utilization of lattice oxygen CeO₂, which is responsible for improved soot oxidation activity of CuCe-AE. When the surface active oxygen is consumed by the soot, the lattice oxygen will diffuse from the bulk of the oxides onto the surface, which is closely associated with the bulk oxygen vacancy [4,35].

The reducibility of the catalysts was investigated using H₂-TPR experiments, and the results are shown in Figure 6b. For pure CeO₂, two reduction peaks are observed at 510 and 750 °C, corresponding to the reduction of surface oxygen and bulk oxygen of CeO₂, respectively [36]. Two intense reduction peaks are observed at 100–200 °C for CuO-doped CeO₂, which is attributed to the reduction of CuO [37] and the promoted reduction of CeO₂ by spillover effect. The presence of multiple peaks implies the existence of different copper species with varied strength with CeO₂. The reduction peak at a low temperature (peak I) corresponds to the reduction of finely dispersed CuO_x clusters with strong interaction with CeO₂ [38,39]. The reduction peak at a high temperature (peak II) is attributed to the reduction of aggregated CuO particles with interaction with CeO₂ [39]. For the CuCe catalyst, an intense reduction peak of CuO interacting with CeO₂ at 174 °C can be clearly observed, which means that the substantial agglomerated CuO particles exist on its surface. Upon acid etching, the reduction peaks shift towards higher temperatures, and the content of aggregated CuO particles obviously decreases in peak intensity. The Gaussian fitting of the H₂-TPR curves was further conducted for CuO-doped catalysts. The area of peak I of CuCe-AE (7920) is almost identical to that of CuCe (8310), while the area of peak II decreases sharply from 44,800 to 8280 after acid etching. These results demonstrate that the acid etching almost does not affect the finely dispersed CuO_x, but removes the aggregated CuO particles significantly.

3. Experimental

3.1. Catalyst Preparation

CuCe mixed oxide catalyst with a Cu/Ce mole ratio of 1:1 was prepared using a hydrothermal method. Briefly, Ce(NO₃)₃·6H₂O (>99.9%, Nanjing Chemical Reagent, Nanjing, China) and Cu(NO₃)₃·3H₂O (>99%, Nanjing Chemical Regent) were added into deionized water. NaOH solution (>98%, Nanjing Chemical Regent) was dropped into the mixed solution slowly and magnetically stirred for 30 min. The pure CeO₂ was synthesized by a similar process but without Cu(NO₃)₃·3H₂O addition. The mixed solution was transferred into a Teflon-lined stainless-steel autoclave, which was subsequently heated at 100 °C for 24 h. After cooling to room temperature, the obtained material was washed with deionized water and alcohol for three times and dried at 80 °C overnight. Finally, the powders were calcined at 500 °C in a muffle oven for 2 h.

The CuCe sample was then washed with nitric acid to reduce the Cu content. Briefly, 1 g of CuCe was dispersed in 2 M HNO₃ solution and magnetically stirred for 12 h at room temperature. The mixture was filtered and washed with deionized water repeatedly until the filtrate was neutral. Then, the sample was dried at 80 °C in the oven overnight, and the obtained catalyst was denoted CuCe-AE.

3.2. Catalytic Activity Evaluation

Printex-U soot, which was consistently used as the model soot, was purchased from Degussa Company (Germany). In activity tests, the catalyst (50 mg) and soot (5 mg) were ground in a mortar for 5 min to achieve tight contact conditions. Tight contact maximizes the number of contact points between catalyst and soot, allows a rapid screening of catalysts, and facilitates the study of structure–performance relationships, although it is less representative of the real contact conditions that occur in a catalytic trap for diesel particulate removal [40]. To minimize the effect of hot spots caused by the exothermal nature of the reaction, the soot–catalyst mixture was diluted with silica pellets (150 mg), sandwiched by quartz wool, and placed in the vertical quartz reactor. Before each test, the mixture was pretreated in N₂ (50 mL/min) at 200 °C for 30 min to eliminate surface adsorbed species. A gas mixture of 10% O₂/N₂ (80 mL/min) was then introduced, and the mixture was heated at a ramping rate (5 °C/min) to 600 °C. The concentrations of CO_x (CO₂ and CO) in the outlet gas were measured using an infrared spectrometer (Thermo Nicolet iS10) [3,20,24]. The soot combustion efficiency was evaluated in terms of T₅₀, corresponding to the temperature at which 50% of soot conversion was obtained. The selectivity to CO₂ was calculated using the following equation.

S_CO2 (%) = A_CO2/(A_CO2 + A_CO) × 100%

(1)

A_CO2 and A_CO represent the total peak areas of produced CO₂ and CO, respectively.

To determine the apparent activation energy (E_a) of the catalysts, the isothermal reactions were conducted under tight contact with the 10% O₂/N₂ flow of 150 mL/min. The soot conversion was kept below 15% to ensure the reaction in the kinetic regime, and the apparent activation energy was calculated according to the Arrhenius equation.

$$\ln r=A-\frac{E_a}{R}·\frac{1}{T}$$

(2)

where A is the pre-exponential factor, R is the ideal gas constant, r is the specific soot oxidation rate (mol min⁻¹ m⁻²), and T is temperature (K).

3.3. Catalyst Characterization

The Cu content of catalysts was determined via inductively coupled plasma atomic emission spectrometry (ICP-AES, J-A1100, Jarrell-Ash, Waltham, MA, USA). The micromorphology of catalysts was characterized using a transmission electron microscope (TEM) (JEM-2100, JEOL, Tokyo, Japan). X-ray diffraction (XRD) spectra of samples were identified with an Ultima (Rigaku, Tokyo, Japan) machine operated at 40 kV and 40 mA, with a monochromator using Cu Kα radiation (λ = 0.15406 nm). The XRD diffractograms were collected in the 2θ range between 10° and 80°, with a step size of 10°/min. To check the distribution of Cu element in samples, an energy-dispersive spectrometry (EDS) analysis was carried out. The specific surface area of catalysts was measured via N₂ sorption at −196 °C using the Brunauer–Emmett–Teller (BET) method on an ASAP 2020 apparatus (Micromeritics, Unterschleissheim, Germany). The sample was vacuum-pretreated at 300 °C for 3 h prior to measurement. The H₂ temperature-programmed reduction (H₂-TPR) was carried out on a Finesorb-3010 instrument (Fantai, Hangzhou, China) equipped with a thermal conductivity detector (TCD). About 10 mg of the sample was placed in a U-shaped quartz tube and pretreated in N₂ (50 mL/min) at 200 °C for 1 h. After cooling to room temperature, 7.0 vol% H₂/Ar (50 mL/min) was introduced, and the sample was then heated from 25 to 800 °C with a ramp rate of 10 °C/min. Raman spectroscopy of samples was performed on a LabRAM Aramis (Horiba, Tokyo, Japan) with a 532 nm Ar⁺ laser beam. The valence states of surface O, Ce and Cu were acquired via X-ray photoelectron spectroscopy (XPS) (ULVAC-PHI5000, Kanagawa, Japan). The binding energy was calibrated with a C1s peak at 284.6 eV.

4. Conclusions

In this work, acid etching was applied to tune the loading amount and form of copper oxide species in ceria-based mixed oxides. An abnormal decrease in the soot oxidation activity of a CuO-doped CeO₂ catalyst was observed, with a T₅₀ increase of 31 °C. This was ascribed to the extra high Cu loading (20 wt.%), resulting in the formation of aggregated CuO particles in addition to highly dispersed CuO_x clusters. The strong interaction of CuO_x clusters with CeO₂ promoted the formation of bulk oxygen vacancy in CeO₂, but the amount of surface oxygen species decreased obviously due to coverage by agglomerated CuO particles. Correspondingly, the intrinsic reactivity of CuCe was reduced for soot oxidation.

Importantly, these aggregated CuO particles can be removed via acid etching, resulting in a remarkable increase in the catalyst surface area and restoration of surface oxygen vacancies. Meanwhile, highly dispersed CuO_x clusters facilitating the generation of bulk oxygen vacancy are mostly reserved in the etched catalyst. As a result, the etched mixed oxides show the lowest light-off temperature and apparent activation energy for soot oxidation via an enhanced Mv-K reaction mechanism.