Main volatile organic compounds (VOCs) are crucial precursors of secondary organic aerosol, ozone pollution, and photochemical smog [1-5]. Nitrogen-containing VOCs (NVOCs), such as acrylonitrile, n-butylamine and N,N-dimethylformamide, which are typical NVOCs, are most frequently in the emission of the petrochemical, pharmaceutical, coating industries [3, 6]. NVOCs will cause substantial pollution to the environment and exert strong toxic effects on the human body. For the elimination of NVOCs, the traditional combustion technology usually operates at high temperatures, which leads to highly energy consumption and secondary pollutant NO_x formation [7]. Conversely, selective catalytic oxidation (SCO) technology can realize selective generation of nitrogen and reduce the generation of nitrogen oxides [8]. Therefore, effective catalysts with a high catalytic conversion rate and N₂ selectivity in the decomposition of NVOCs must be explored.

A number of studies describing the catalytic degradation of NVOCs have been published. According to Nanba et al., Cu/ZSM-5 could completely convert acrylonitrile with an N₂ selectivity of at least 80% at a temperature greater than 350 ℃ [7-9]. Zhang et al. reported the nearly complete conversion of acetonitrile by Cu/SBA-15 that was associated with an N₂ selectivity of approximately 80% at T > 300 ℃, and thus Cu/SBA-15 was the most promising catalyst [10]. Zhou et al. investigated CrCe supporting on different pillared interlayered clays catalysts for the NVOCs (n-butylamine, ethylenediamine, and acetonitrile) oxidation [11]. Ma et al. discussed the high catalytic activity of Pd-confined materials, showed the complete conversion of n-butylamine at 280 ℃ and explored the structure activity relationship and the reaction mechanism [12]. However, it is rarely reported that low temperature (< 300 ℃) catalytic oxidation of NVOCs can be achieved on non-noble metal catalysts. In particular, the catalysts simultaneously exhibit high conversion and N₂ selectivity at low temperatures.

In recent years, as a rare earth metal oxide, CeO₂ has excellent oxygen storage performance, which was used to catalyze various reactions, such as nitric oxide reduction, carbon monoxide oxidation and NH₃ oxidation reactions [13-15]. Wang et al. observed that the CuO-CeO₂ sample exhibited highly dispersed CuO species, which possessed the highest activity in NH₃ oxidation [16]. The Ce-Zr mixed oxide has been used as a catalyst or support for NO_x selective catalytic reduction, because it combines the advantages of CeO₂ and ZrO₂. Si et al. impregnated CeO₂-ZrO₂ with nickel and sulfate were to improve the catalytic performance for removal of NO with NH₃ [17].

To take full advantage of the Ce-Zr mixed oxide and the promoting effect of Cu, CeZrO_x and CeCu_aZrO_x (a = 1%, 5%, 10%, and 15%) were synthesized using a precipitation method and tested for selective oxidation of n-butylamine in the present study. The catalysts showed high n-butylamine conversion rates and N₂ selectivity. Moreover, the structure, physical properties, oxygen species, and redox ability of catalysts were also characterized by various techniques, and the reaction mechanism underlying the catalytic degradation of n-butylamine was also discussed.

The structure and physical properties of CeCu_aZrO_x catalysts were characterized by X-ray diffraction (XRD), high resolution transmission electron microscopy (HR-TEM) and N₂ adsorption/desorption. The XRD results of CeCu_aZrO_x catalysts are shown in Fig. S1 (Supporting information). CeZrO_x displayed four main peaks at approximately 29.1°, 33.7°, 48.5° and 57.5°, consistent with the structure of Ce_0.6Zr_0.4O₂ (JCPDS No. 38-1439) and corresponding to the (111), (200), (220) and (311) planes. For CeCu_aZrO_x catalysts, characteristic diffraction peaks for the Ce_0.6Zr_0.4O₂ and the ZrO₂ phase (JCPDS No. 49-1642) were obtained, indicating the weak interaction between Ce and Zr, which resulted in the segregation of CeO₂ and ZrO₂ [18, 19]. For CeCu_aZrO_x samples, the intensity of peaks assigned to Ce_0.6Zr_0.4O₂ and ZrO₂ decreased significantly as Cu content increased. No obviously characteristic diffraction peaks of Cu species were presented in XRD results, indicating that Cu species were highly dispersed on the catalyst's surface. Additionally, the CeO₂ crystallite sizes were calculated using the Scherrer equation and shown in Table S1 (Supporting information), it decreased a little as the increasing of Cu content.

HR-TEM images of CeCu_10%ZrO_x are shown in Figs. S2a and b (Supporting information), the lattice spacing values of 0.269, 0.291 and 0.312 nm, which were attributed to the exposed Ce_0.6Zr_0.4O₂ (200), ZrO₂ (111) and (111) facets of Ce_0.6Zr_0.4O₂, respectively. The exposure of ZrO₂ (111) facets indicated the presence of phase separation in the CeCu_10%ZrO_x catalyst, consistent with the XRD result. Elemental mapping was performed by energy-dispersive X-ray spectroscopy (EDX) to confirm the elemental distributions of Ce, Cu, Zr, and O species in CeCu_10%ZrO_x. As shown in Figs. S2d-h (Supporting information), the O, Ce, Cu, and Zr species were homogeneously dispersed throughout CeCu_10%ZrO_x, verifying that Cu²⁺ was successfully incorporated into CeZrO_x.

N₂ adsorption/desorption isotherms and pore size distribution of CeCu_aZrO_x catalysts are shown in Fig. S3 (Supporting information). According to International Union of Pure and Applied Chemistry (IUPAC) classification, all the catalysts presented type Ⅳ isotherm curves, indicated that the samples are typical mesoporous materials. The special surface area (SSA), pore volume and pore size are calculated by N₂ physical adsorption results of CeZrO_x and CeCu_aZrO_x catalysts and shown in Table S1. The addition of CuO in CeZrO_x had a significant impact on the special surface area. As the contents of Cu increased from 1% to 15%, the SSA of the samples grew from 61 to 85 m²/g, CeCu_15%ZrO_x exhibited the largest surface area of 85 m²/g. The results showed that the introduction of Cu into the CeZrO_x could lead to the change of the structure, further result in the higher specific surface area [20], which is instrumental in the dispersion of Cu species.

The conversion of n-butylamine and characteristic activity data are shown in Fig. 1a and Table S1. Obviously, the addition of Cu increased the n-butylamine conversion in all samples. The temperature recorded at 50% and 90% n-butylamine conversion (T₅₀ and T₉₀) decreased in the order CeZrO_x > CeCu_15%ZrO_x > CeCu_10%ZrO_x > CeCu_5%ZrO_x > CeCu_1%ZrO_x. Interestingly, the values of T₁₀₀ decreased in the order CeZrO_x > CeCu_1%ZrO_x = CeCu_5%ZrO_x > CeCu_10%ZrO_x = CeCu_15%ZrO_x. The activities of CeCuZrO_x catalysts were higher than the values reported in the literature (Table S2 in Supporting information). As shown in Fig. 1a, the conversion vs. temperature plot of n-butylamine oxidation over CeZrO_x exhibited a relativity lower conversion rate, while complete conversion was achieved at approximately 350 ℃. Furthermore, n-butylamine oxidation activity in the presence of Cu in CeZrO_x was very similar, the conversion of n-butylamine progressively increased with the temperature rising, and a 100% n-butylamine conversion rate was achieved at about 300 ℃. The n-butylamine conversion over CeCu_aZrO_x samples is tightly related to the Cu contents. CeCu_10%ZrO_x and CeCu_15%ZrO_x exhibited the highest activity and achieved 100% n-butylamine conversion at 250 ℃.

The selectivity to CO₂ and CO over CeCu_aZrO_x catalysts during n-butylamine oxidation is shown in Figs. 1b and c. For CO₂ formation, as shown in Fig. 1b, the selectivity to CO₂ over CeZrO_x and CeCu_aZrO_x catalysts increased with the increasing of temperature and achieved 100% at a higher temperature. The shape of CO₂ selectivity showed the same trend as that of n-butylamine conversion rates over CeCu_aZrO_x, and the CO₂ selectivity increased with the increase of copper contents. For CO formation, as shown in Fig. 1c, the CO selectivity over CeZrO_x and CeCu_aZrO_x catalysts first increased and then decreased. Relative higher amount of CO was formed over CeZrO_x, and the formation of CO over CeCu_aZrO_x decreased with the increase of copper contents. Above all, CO was almost not produced over CeCu_10%ZrO_x and CeCu_15%ZrO_x samples at 100–400 ℃.

Figs. 1d–h present the selectivity of N-containing products (N₂, N₂O, NO, NO₂ and NH₃) of n-butylamine catalytic degradation over CeCu_aZrO_x. As seen from Fig. 1d, CeZrO_x exhibited very low N₂ selectivity at temperatures ranging from 100 ℃ to 300 ℃. The N₂ selectivity over CeCu_aZrO_x catalysts first increased and then decreased with the temperature rising. And the N₂ selectivity of CeCu_10%ZrO_x was apparently higher than other catalysts, increased with the reaction temperature and reached a maximum of 90% at 250 ℃. The increase in N₂ selectivity observed at T < 300 ℃ for CeCu_aZrO_x was probably because of copper doping. The incorporation of Cu in CeZrO_x might favor the generation of N₂. While at high temperatures from 300 ℃ to 400 ℃, the decrease in N₂ selectivity was caused by the formation of NO_x. For the degradation of NVOCs, the crucial point for NVOCs catalytic oxidation lies in the enhancement of the nitrogen generation and control of NO_x production, avoiding secondary pollution.

To get an intensive study into the catalytic behaviors of CeCu_aZrO_x catalysts, the undesirable byproducts (N₂O, NO, NH₃, NO₂) selectivity are respectively depicted in Figs. 1e–h. As noted, NO was the main N-containing byproducts for CeZrO_x and CeCu_aZrO_x catalysts. NO selectivity first increased to 8% and then decreased at T < 250 ℃. Besides, selectivity to NO over CeCu_aZrO_x increased with the raising of copper content at T > 250 ℃. The selectivity to NO over catalysts increased upon the rising temperature and achieved 50% at 400 ℃. Additionally, NO₂ was generated over CeCu_aZrO_x at 300 ℃, and the production of NO and NO₂ at high temperatures is the major cause for the decrease of N₂ selectivity. Fig. 1i shows N-containing product selectivity at T₁₀₀. CeCu_10%ZrO_x had the highest N₂ selectivity, as well as the lowest NO selectivity. Less NH₃ formation occurred with CeCu_5%ZrO_x, CeCu_10%ZrO_x and CeCu_15%ZrO_x than with CeCu_1%ZrO_x and CeZrO_x. N₂ was the main product generated by CeCu_10%ZrO_x, and thus this catalyst represents a promising candidate for n-butylamine selective catalytic oxidation. At temperatures greater than T₁₀₀, the nitrogen balance was more than 95% for CeCu_aZrO_x catalysts.

The catalytic degradation of exhaust containing water vapor may result in the deactivation of catalysts in practical application. Hence, the investigation of the effect of H₂O on the catalytic performance of CeCu_10%ZrO_x is worthwhile. As shown in Fig. S4 (Supporting information), the addition of 15 vol% H₂O exerted little inhibition effect on n-butylamine conversion. A 100% n-butylamine conversion rate was obtained at T > 250 ℃, which was the same as that of the reaction without H₂O. Moreover, N₂ selectivity was slightly decreased to 85% by the addition of H₂O. The decrease in N₂ selectivity in the reaction with CeCu_10%ZrO_x was accompanied by increases in NH₃ and NO_x formation. H₂O exists in the reaction atmosphere as a reactant, and there is competitive adsorption with reactant molecules n-butylamine and oxygen, resulting in the decrease of N₂ selectivity and the formation of other nitrogen-containing products. The formation of NH₃ may be due to the presence of H₂O. H₂O molecule will participate in the reaction after decomposition, and ^*NH₂ species are more likely to produce NH₃. Briefly, the CeCu_10%ZrO_x catalyst exhibited good resistance to H₂O and thus it is potentially useful for eliminating n-butylamine from the exhaust containing a certain amount of H₂O.

The capability of oxygen species transmission and migration on Ce-Zr based catalysts are vital factors in catalytic oxidation reactions, which were evaluated using O₂ temperature-programmed desorption (O₂-TPD), UV-vis diffuse reflectance spectra (UV-vis DRs), X-ray photoelectron spectra (XPS) and Raman. The O₂-TPD patterns of the CeCu_aZrO_x samples are presented in Fig. S5 (Supporting information) and could be divided into three areas. The desorption peak centered at T < 200 ℃, is characteristic of the desorption of ordinarily chemically adsorbed oxygen species. The area at the temperature range of 200-600 ℃ is ascribed to the desorption of oxygen chemically adsorbed on the oxygen vacancies [21]. And the third area appeared at high temperatures, which could be attributed to the bulk lattice oxygen [22]. The lattice oxygen of CeZrO_x and CeCu_1%ZrO_x catalysts did not appear in this temperature range, indicating that the desorption of lattice oxygen species in these samples required a higher temperature. The desorption temperatures of the first peaks decreased in the order that: CeZrO_x > CeCu_15%ZrO_x > CeCu_10%ZrO_x > CeCu_5%ZrO_x > CeCu_1%ZrO_x, which was related to the catalytic activity of n-butylamine at lower temperatures (T₅₀ and T₉₀). It indicating that chemisorbed oxygen species was more conducive to the low-temperature activation of pollutant molecules and further oxidation to produce CO₂ and H₂O.

The electronic states of CeCu_aZrO_x catalysts were obtained from UV-vis-DRs and shown in Fig. S6 (Supporting information). The band at ca. 218 nm could be assigned to tetravalent Zr species with eight-coordinated [23]. The band near 245 nm corresponded to the charge transfer transitions between O²⁻ and Ce³⁺, which indicated the existence of oxygen vacancy. The bands for all catalysts around 310 nm and 341 nm were attributed to the charge transfer and inter band transitions of O²⁻and Ce⁴⁺, respectively [24, 25]. Based on these results, both Ce³⁺ and Ce⁴⁺ species were present in these samples, consistent with the XPS results.

The surface chemical state of CeCu_aZrO_x catalysts was obtained from an XPS analysis, and the results are presented in Fig. 2. The surface compositions of the samples (Cu, Ce, Zr and O) are displayed in Table S3 (Supporting information). The XPS results for the Ce3d spectra of CeZrO_x and CeCu_aZrO_x catalysts are shown in Fig. 2a. The labels are used to identify Ce 3d peaks, where u and v refer to the 3d_3/2 and 3d_5/2 spin-obit components, respectively. The peaks referred to as u, u′′, v, and v′′ were assigned to Ce³⁺ species and the other six peaks (u′, u′′′, u′′′′, v′, v′′′ and v′′′′) represented the presence of Ce⁴⁺ [26]. Thus, Ce was in a partially reduced state on the surfaces of the catalysts, which might be ascribed to the interaction between Cu and Ce. Moreover, the presence of Ce³⁺ led to a charge imbalance that was assigned to the formation of oxygen vacancies, consistent with the charge compensation. The surface Ce³⁺ concentration of total Ce (Ce³⁺ + Ce⁴⁺) on the CeCu_10%ZrO_x was 0.40 (Table S3), which was higher than other samples and indicted the formation of more oxygen vacancies on CeCu_10%ZrO_x. The high Ce³⁺ content in CeCu_10%ZrO_x might result from the increased reducibility of CeO₂ caused by the synergistic effect of Cu species and Ce species. As shown in Table S3, the incorporation of Cu into these samples significantly increases the proportion of Ce³⁺/(Ce³⁺ + Ce⁴⁺). We believe that is due to the Cu species introduction and the transfer of electrons from Cu⁺ to Ce⁴⁺.

Fig. 2b displays the O 1s XPS spectra for CeCu_aZrO_x catalysts. The main peak was observed at around 529.4 eV and assigned to the lattice oxygen of catalysts. And the shoulder peak centered at 531.5 eV could be ascribed to the chemisorbed oxygen species [27]. According to the literature, in the oxidation reaction, the surface chemisorbed oxygen is more active than lattice oxygen because of its strong mobility [28]. Moreover, the percentages of O_ads/O_latt are calculated and shown in Table S3 (Supporting information). Compared with CeZrO_x, higher contents of chemisorbed oxygen were observed on the surface of CeCu_aZrO_x, implying that Cu introduction increased the concentration of activated adsorbed oxygen species. Additionally, the percentages of O_ads and O_latt in CeCu_10%ZrO_x (0.79) were relatively higher among all samples, suggesting that more chemisorbed oxygen species were presented on the surface of CeCu_10%ZrO_x. Hence, these results indicated well catalytic activity and N₂ selectivity for the degradation of n-butylamine over CeCu_10%ZrO_x.

The comparison of Raman spectra of CeZrO_x and CeCu_aZrO_x catalysts is shown in Fig. 2c. Raman spectra of all samples displayed an absorption band at around 470 cm⁻¹, corresponding to the characteristic F_2g vibration mode of fluorite CeO₂. The broad bands at 620 cm^-1 were assigned to oxygen vacancies and ascribed to the presence of defective structures in CeCu_aZrO_x samples. In addition, this finding also indicated the formation of Cu-Ce-Zr solid solution, which could enhance the formation of oxygen vacancies [29]. The weak band at 310 cm⁻¹ was ascribed to the substitution of oxygen atoms from the ideal lattice positions. Finally, a weak intensity Raman peak was observed at 130 cm^-1 and ascribed to Zr doping into the lattice of Ce, which led to the distortion of the framework symmetry. The intensity of the band at 470 cm^-1 decreased as the copper content increased and might be related to the optical absorption of copper [30]. In addition, the existence of Cu influenced the position and intensity of the F_2g band at 470 cm⁻¹. Copper incorporation into CeZrO_x led to a red shift (from 470 cm⁻¹ to 450 cm⁻¹) and the broadening of F_2g peaks, which was related to the formation of oxygen vacancies. Meanwhile, the band at 620 cm⁻¹ showed increasing intensity, further verifying the existence of oxygen vacancies caused by the incorporation of copper into the lattice [31]. The areas ratio of the peaks at 470 cm⁻¹ (A₄₇₀) and 620 cm⁻¹ (A₆₂₀) reflects the concentration of oxygen vacancies [29]. The A₆₂₀/A₄₇₀ value for CeCu_aZrO_x rapidly increased compared with CeZrO_x (Fig. 2d), suggesting that the addition of Cu resulted in a higher oxygen vacancy concentration in CeCu_aZrO_x samples, and we reached a similar conclusion from the theoretical calculations. Thus, the CeCu_10%ZrO_x catalyst could generate more oxygen vacancies among other samples, because of the value of A₆₂₀/A₄₇₀ and the shift of the F_2g peaks. Finally, this catalyst would promote the performance of n-butylamine selective oxidation.

In situ visible Raman spectra of CeCu_10%ZrO_x catalyst under a flow of n-butylamine and O₂ are shown in Fig. 2e. As shown in the spectra, the main peak at around 465 cm^-1 (F_2g) always dominated throughout the reaction, indicating that the catalysts maintained the major fluorite-like structure [32]. The position of the F_2g peak was slightly shifted toward a lower Raman peak as the reaction temperature increased from 30 ℃ to 300 ℃, due to the thermal expansion and phonon confinement. In some studies, the downward shift of F_2g possibly corresponds to the thermal-induced loss of lattice O that results in the sintering of CeO₂ crystals [33]. Additionally, in some cases, the shift was also in consequence of the lattice expansion and mode softening that occurred when oxygen vacancies are created [34]. The peak at 615 cm^-1 was attributed to the intrinsic oxygen vacancies in CeCu_10%ZrO_x, which were caused by the escape of interstitial oxygen from the lattice, thus leaving oxygen vacancies in original sites [33, 35]. As shown in Fig. 2f, the ratios of A₆₁₅/A₄₆₅ enhanced with the increase of temperature, indicating that more oxygen vacancies were generated in the bulk phase of CeCu_10%ZrO_x due to the temperature-induced structural changes [33], and the increase in the number of oxygen vacancies promote n-butylamine oxidation.

Brief summary, the better catalytic performance of CeCu_10%ZrO_x depended on the formation of surface oxygen vacancies and chemisorbed oxygen species. Mixed valences of Ce³⁺ and Ce⁴⁺ were present in the CeCu_aZrO_x catalyst and accompanied the generation of oxygen vacancies due to Cu introduction. The surface Ce³⁺ concentration on CeCu_10%ZrO_x was higher than in other samples, indicating the formation of a greater number of oxygen vacancies, and the higher A₆₂₀/A₄₇₀ ratio further confirmed the higher oxygen vacancy concentration of CeCu_10%ZrO_x. Furthermore, the O_ads/O_latt ratio in CeCu_10%ZrO_x was higher than in all other samples, suggesting that more chemisorbed oxygen species were presented on the surface of CeCu_10%ZrO_x. Hence, this catalyst displayed better performance in the catalytic oxidation of n-butylamine.

The chemical state of Cu species was investigated by UV-vis-DRs (Fig. S6). The CeCu_aZrO_x catalysts showed a weak band at around 450 nm, which could be assigned to a charge transfer band in Cu²⁺→O²⁻→Cu²⁺ [36]. In addition, the CeCu_aZrO_x catalysts also exhibited broad bands at around 700 nm, it corresponded to the d-d transitions of Cu²⁺ in octahedral symmetry. This band shifted from 650 nm to 750 nm, indicated that the coordination environment of copper species was influenced by the Cu content [37]. Furthermore, this band shifted to a longer wavelength as Cu content increased from 1 wt% to 15 wt%, which was attributed to less CuO dispersion. Cu 2p XPS spectra for CeCu_aZrO_x catalysts are presented in Fig. S7 (Supporting information). Two peaks centered at 932.2 and 952.1 eV were observed and potentially attributed to Cu 2p_3/2 and 2p_1/2 of Cu⁺, respectively, whereas two other peaks at 933.6 and 953.6 eV corresponded to the existence of Cu²⁺ [38, 39]. Meanwhile, the shake-up satellite feature centered at binding energies of around 942 and 962 eV was also observed. Both Cu⁺ and Cu²⁺ symmetrical peaks could be found in CeCu_aZrO_x catalysts. Moreover, the peak area of Cu²⁺ species was larger than that of Cu⁺, indicating that the characteristics Cu species predominantly present on the surface is Cu²⁺.

H₂-TPR measurements were performed to investigate the reduction behavior of the CeCu_aZrO_x catalysts, and the results are presented in Fig. S8 (Supporting information). The reduction profile of CeZrO_x revealed a weak reduction peak that started at approximately 450 ℃ and was centered at 600 ℃. This peak potentially corresponded to the reduction of Ce⁴⁺ to Ce³⁺, because the reduction of Zr⁴⁺ occurred at a temperature greater than 1000 ℃ [40, 41]. Additionally, the presence of a single peak for CeZrO_x suggested that the reduction of surface and bulk CeO₂ occurred concurrently because of the good oxygen mobility.

Obviously, the reduction behavior of CeZrO_x catalyst was altered by copper-doping. The peak at 600 ℃ disappeared, potentially because the reduction peak of CeO₂ shifted to low temperatures, which was promoted by Cu doping. The peak splitting of the H₂-TPR data was performed and presented in Fig. S8b (Supporting information). The profiles of samples consisted of three peaks, which were denoted as α, β and γ. According to the literature [32, 42], the attributions of the peaks are listed below. The first peak (α) is ascribed to the reduction of highly dispersed CuO species, including isolated Cu²⁺ ions, and the small dimension of the Cu-O-Cu clusters. The second peak (β) is assigned to the presence of large CuO clusters. The third peak (γ) is attributed to the presence of bulk-like CuO, as well as Cu⁺ and Ce⁴⁺ reduction [32]. Furthermore, values for H₂ consumption are present in Table S4 (Supporting information). For the CeCu_10%ZrO_x catalyst, the H₂ consumption of peak α was 0.71 mmol/g, which was the highest value observed among the samples and proved the best redox property.

Based on the analyses described above and combined with the catalytic activity data, the superior catalytic performance of CeCu_10%ZrO_x compared with CeCu_aZrO_x catalysts could be related to the fact that CeCu_10%ZrO_x displayed the highest redox ability, due to the presence of the reduction peak of highly dispersed copper oxide species at a lower temperature and the highest H₂ consumption of this peak among all the samples.

To identify surface intermediate species, we carried out in situ DRIFTs under n-butylamine and the condition of n-butylamine and O₂. Fig. 3a shows in situ DRIFTs during exposure of CeCu_10%ZrO_x to gas mixture containing n-butylamine with helium as the balance gas at room temperature. The absorption bands of n-butylamine were observed at 3223, 3125, 2962-2855, 1643, 1588, 1466, 1378, 1173, 1094 and 1071 cm^‒1. The bands at 3223 and 3125 cm⁻¹ were ascribed to the amino group of n-butylamine, the vibrations of v(sym)NH₂ and 2δNH₂, respectively. The band at 1587 cm⁻¹ was also attributed to the vibration of δNH₂, and the band observed at 1094 and 1072 cm^-1 were assigned to the vibration of v(CN) and t(CH₂(N)) [43]. The bands centered at 2963–2855, 1466 and 1378 cm^‒1 were assigned to asymmetric and symmetric bending vibrations of the C‒H bonds (v(as)CH₃, v(as)CH₂, v(s)CH₃, v(s)CH₂(N), δ(as)CH₃, δ(s)CH₃, respectively). Simultaneously, the band at 1173 cm^‒1 was also detected and attributed to the vibration of v(C-C). The band observed at 1643 cm^‒1 corresponded to the molecular adsorption of H₂O. The DRIFT spectra of CeCu_10%ZrO_x with an exposure to 375 ppm n-butylamine and 5% O₂ at room temperature are shown in Fig. 3b, it was found that adding the gaseous O₂ would not change the adsorption species over CeCu_10%ZrO_x.

The catalysts were exposed to reaction conditions (375 ppm n-butylamine and 5% O₂, with helium as the balance gas) to further ascertain the intermediate species of the reaction between n-butylamine and O₂. Fig. 3c illustrates in situ DRIFT spectra of the CeCu_10%ZrO_x catalyst in a flow of n-butylamine+O₂ at various temperatures. The band intensities of the adsorption spectra of n-butylamine decreased gradually as the temperature increased. While these band intensities decreased, the band intensity of ‒NH₂ (1583 cm^‒1) increased from 100 ℃ to 250 ℃. Additionally, some new bands appeared at a temperature above 100 ℃. The bands at 2177 and 2126 cm^-1 could be assigned to adsorbed N₂O [44]. The bands were observed at temperatures ranging from 150 ℃ to 300 ℃, but disappeared at temperatures above 300 ℃. The bands at 1555 and 1569 cm^‒1 were also detected and attributed to monodentate nitrate species and bidentate nitrate species, respectively [19, 45]. Meanwhile, ionic nitrate (1369 cm^-1) formed. Furthermore, the existence of a peak at 1434 cm^‒1 indicated the formation of acetate species [46]. Notably, the acetate species was detected at 200 ℃, and its quantity increased as the temperature increased up to 300 ℃. A weak band at 1344 cm^‒1 appeared at 150 ℃ was attributed to the chelation of bidentate carbonate [47]. An NH₃ species was observed at 1300 cm^‒1, and the band intensity increased monotonously as the temperature increased, but suddenly decreased at 300 ℃ [48].

Combining the results of DRIFT spectra and the distribution of products (Fig. 1), we were able to predict two period procedures in reaction from the formation of N₂O, N₂, NO, NO₂ and NH₃.

The following main reaction routes were proposed Eqs. 1 and 2:

(1)

(2)

As shown in Fig. 3, the intensity of the vibration peaks at 1583 and 1555 cm^‒1 increased as the temperature increased from 100 ℃ to 250 ℃. When the temperature increased from 100 ℃ to 250 ℃, some portion of NH₂^* was oxidized into nitrate species (1555 cm^‒1, monodentate nitrate) and then NH₂^* reacted with surface nitrate species, leading to the production of N₂ and N₂O. Within this temperature range, the formation of N₂ exceeded NO production. Therefore, we inferred that the reaction rate of Eq. 4 is greater than Eq. 3, and the formation of NO^* (Eq. 3) is a rate-controlled step (Eqs. 3-6).

(3)

(4)

(5)

(6)

At T > 250 ℃, the gaseous O₂ directly participated in the reaction, and then the NH₂^* was over oxidized into bidentate nitrate species. According to Fig. 3c, monodentate nitrate (1555 cm^‒1) appeared at temperatures from 100 ℃ to 250 ℃ and then was further oxidized to bidentate nitrate species (1569 cm^-1) at T > 250 ℃. Hence, we speculated that the reaction rate of Eq. 7 was higher than Eq. 8, most of NH₂^* was oxidized to stable nitrate species, and the concentration of NH₂^* was not sufficient for NO^* reduction. At this time, the rate-controlled step was the formation of N₂^* (Eq. 8). Moreover, NO₂ was generated through the reaction of NO with O₂ at high temperatures. These processes were described by Eqs. 7–9.

(7)

(8)

(9)

Theoretical calculations were performed to study the formation energy of oxygen vacancy, the adsorption of reactants and product molecules and the reaction mechanism. In general, oxygen vacancies play a vital role in catalytic oxidation reactions. In the present study, the formation energies of two different oxygen vacancies (labeled as O_vCe2Zr1 and O_vCe3) were calculated for the pure CeZrO₂(111), as shown in Fig. 4a. The optimized geometry of the structures of CeZrO₂(111) with O_vCe2Zr1 and O_vCe3 vacancies are presented in Figs. 4b and c, respectively. The calculated formation energy for O_vCe2Zr1 and O_vCe3 vacancies are 3.10 and 4.08 eV, respectively, indicating that the formation of O vacancy on pure CeZrO₂ has certain difficulties. Moreover, the structure of the Cu-doped CeZrO₂(111) was first investigated and the formation energies of four different oxygen vacancies (labeled as O_v1, O_v2, O_v3 and O_v4) were calculated, as shown in Fig. 4d. The doping of Cu promoted the formation of oxygen vacancies, as the calculated E_ov values near the doped Cu (O_v1, O_v2, O_v3 and O_v4) are much lower than the formation energy of O_vCe2Zr1 and O_vCe3 vacancies. In particular, the calculated E_ov value for O_v1 is negative, indicating that an oxygen vacancy forms spontaneously. Therefore, further computations in this study were performed using the model of a Cu-doped CeZrO₂(111) slab with an O_v1 vacancy.

The adsorption energies of reactants and products are presented in Fig. S9 (Supporting information). The pure facets of Cu-doped CeZrO₂(111) have a lower binding strength for O₂, and the adsorption energy is very small (‒0.01 eV), indicating the very weak physical adsorption of O₂ on this facet. Interestingly, as oxygen vacancies are introduced to the Cu-doped CeZrO₂(111), the adsorption energy of O₂ becomes higher, indicating that an oxygen vacancy is more conducive to the selective catalytic oxidation of n-butylamine. O₂ and n-butylamine molecules display relatively stronger binding strengths (‒0.37 and ‒0.67 eV, respectively), while N₂ and CO₂ present very weak adsorption strengths (‒0.04 and ‒0.12 eV, respectively). The distinct adsorption strengths will potentially improve the catalytic performance of Cu-doped CeZrO₂(111) in butylamine oxidation, as the reactant easily adsorbs, while the products will desorb rapidly.

The minimum energy pathway for n-butylamine oxidation was subsequently calculated to further verify the n-butylamine adsorption on the surface of the catalyst and the formation of intermediate species and N₂ during the reaction. As shown in Fig. 4e, n-butylamine in the gas phase was initially adsorbed on exposed Ce atoms near the O_v1 vacancy to form an adsorbed n-butylamine molecule. First, the C‒H bond was attacked by a neighboring O to produce MS (CH₃CH₂CH₂CHOHNH₂) species; this process should overcome an energy barrier of 1.57 eV. Subsequently, the C‒N bond was dissociated into ^*C₄H₉ and ^*NH₂ species, which will adsorb on the exposed Ce atom and the neighboring O atom, respectively.

For the selective catalytic oxidation reaction of n-butylamine, we mainly focused on the formation of N₂. The transition states and reaction pathways were calculated to further confirm the mechanism of N₂ generation from ^*NH₂ species, as evidenced by in situ DRIFTs. O₂ was adsorbed on oxygen vacancies (O_v1) and reacted with ^*NH₂ species to form ^*OONH₂ species with an energy of 0.39 eV, indicating that this reaction readily occurred. Subsequently, the O atom of ^*OONH₂ species was attacked by the H atom of ^*OONH₂ species, conquering a high energy barrier of 2.55 eV to produce the ^*NHOH species. Then, the O atom of the ^*NHOH species was again attacked by the H atom to form the ^*NO species with an energy barrier of 1.23 eV. At this time, ^*NO species occupied the position of the original oxygen vacancy (O_v1). The ^*NH₂ species was adsorbed on the ^*NO species, and should surmount an energy barrier of 1.68 eV to obtain the ^*NOHNH species. Finally, the H atom of the ^*NOHNH species attacked the O atom to generate the end product of N₂ after H₂O was stripped.

In short, ^*NH₂ was oxidized into N₂ through the following steps Eqs. 10 and 11:

(10)

(11)

The rate-determining step of the whole process is ^*OONH₂ = ^*NHOH + ^*O, with an energy barrier of 2.55 eV. Notably, the energy barriers were calculated without considering the effects of coverage, O₂ pressure, solvent, etc. Therefore, the actual reaction barriers would be much lower, indicating that the catalytic oxidation of n-butylamine through the proposed pathway may be feasible.

In conclusion, a series of mixed CeCu_aZrO_x oxides were successfully synthesized and tested for n-butylamine selective catalytic oxidation. The CeCu_aZrO_x sample with 10% Cu showed excellent performance, exhibited 100% conversion of n-butylamine and 90% N₂ selectivity at 250 ℃. CeCu_10%ZrO_x displayed the highest redox ability, the reduction peak of the highly dispersed copper oxide species was located at a lower temperature, and the H₂ consumption of this peak is maximum. The incorporation of Cu into CeZrO_x significantly increased the proportion of Ce³⁺/(Ce³⁺ + Ce⁴⁺). Additionally, more chemisorbed oxygen species and oxygen vacancies were presented in CeCu_10%ZrO_x. The changes in oxygen vacancies over CeCu_10%ZrO_x during reaction process were studied by in situ Raman spectra, and the number of oxygen vacancies gradually increased as the temperature increased. Theoretical calculations were performed to study the formation energy of oxygen vacancy, the doping of Cu could promote the formation of oxygen vacancy as the oxygen vacancies formation energy of CuCeZrO_x was much lower than that of pure CeZrO_x. Furthermore, the catalytic reaction mechanism underlying n-butylamine selective catalytic oxidation was proposed as described below. First, n-butylamine and O₂ were adsorbed on the oxygen vacancy (O_v1), and then n-butylamine was attacked by the neighbor O to produce MS (CH₃CH₂CH₂CHOHNH₂) species. Subsequently, the C-N bond was dissociated, forming CH₃-CH₂-CH₂-CH₂^* and NH₂^*. CH₃-CH₂-CH₂-CH₂^* was further oxidized into CO₂ and H₂O, and NH₂^* could be oxidized to NO/nitrate species through important intermediate species (^*OONH₂ and ^*NHOH). Finally, NH₂^* reacted with NO/nitrate species, leading to the production of N₂ and H₂O. The results of the present study are believed to provide a deeper understanding of the catalytic mechanisms and a high-performance removal of NVOCs.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work is financially supported by the R & D Program of Beijing Municipal Education Commission (No. KJZD20191443001), Beijing Municipal Science and Technology Commission (No. Z181100000118003), the Fundamental Research Funds for the Central Universities and Doctoral Research Start-up Fund Project of Taiyuan University of Science and Technology (No. 20202053).

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2021.09.056.