Cerium-Copper Oxides Synthesized in a Multi-Inlet Vortex Reactor as Effective Nanocatalysts for CO and Ethene Oxidation Reactions

Abstract

In this study, a set of CuCeO_x catalysts was prepared via the coprecipitation method using a Multi-Inlet Vortex Reactor: the Cu wt.% content is 5, 10, 20, 30 and 60. Moreover, pure CeO₂ and CuO were synthesized for comparison purposes. The physico-chemical properties of this set of samples were investigated by complementary techniques, e.g., XRD, N₂ physisorption at −196 °C, Scanning Electron Microscopy, XPS, FT-IR, Raman spectroscopy and H₂-TPR. Then, the CuCeO_x catalysts were tested for the CO and ethene oxidation reactions. As a whole, all the prepared samples presented good catalytic performances towards the CO oxidation reaction (1000 ppm CO, 10 vol.% O₂/N₂): the most promising catalyst was the 20%CuCeO_x (complete CO conversion at 125 °C), which exhibited a long-term thermal stability. Similarly, the oxidative activity of the catalysts were evaluated using a gaseous mixture containing 500 ppm C₂H₄, 10 vol.% O₂/N₂. Accordingly, for the ethene oxidation reaction, the 20%CuCeO_x catalyst evidenced the best catalytic properties. The elevated catalytic activity towards CO and ethene oxidation was mainly ascribed to synergistic interactions between CeO₂ and CuO phases, as well as to the high amount of surface-chemisorbed oxygen species and structural defects.

1. Introduction

Over the last few years, the emission limits in the automotive field have become stricter. Several pollutants are produced in diesel and gasoline engines, such as carbon monoxide (CO) volatile organic compounds (VOCs), nitrogen oxides (NO, N₂O and NO₂, labelled as NO_x), unburned carbon-based compound (UHC) and particulate matter (PM) [1,2,3,4].

Carbon monoxide (CO) originates from the incomplete combustion of fuel inside the engine reaction chamber. The engine working conditions are directly connected with the CO production: a higher quantity of CO is generated when the air-to-fuel ratio is under-stoichiometric (λ ratio < 1, rich mixture) [4,5]. However, it is possible to form CO in lean conditions too (λ ratio > 1), due to kinetic effects [6]. The CO is a harmful substance for humans because it can create confusion and asphyxia over a certain concentration [7,8,9].

The volatile organic compounds (VOCs) are a set of various substances present in the atmosphere that are originated due to natural or human-related processes (i.e., biogenic or anthropogenic, respectively) [10]. These compounds may represent an important risk for humans’ health [11]. Moreover, they can participate in atmospheric photochemical processes that lead to the formation of other harmful substances (e.g., peroxyacyl nitrates or ozone) [11,12]. Accordingly, the emission of these substances, as well as the emission of CO, is regulated [13]. In previous research works [14], ethene has been often used as VOC probe molecule [15,16,17,18,19,20]. In this sense, ethene has shown higher resistance to oxidation respect to other VOCs (e.g., propylene) [16,18], whereas it can be oxidized at temperatures similar to those required by some aromatic compounds, e.g., toluene [16,20].

The catalytic systems used for the abatement of automotive pollutants typically contain noble metals. Among them, Pt is probably the most used metal, since it is very active towards several oxidation reactions [21,22,23,24,25,26,27]. However, there are limitations in the utilization of noble metals (i.e., their high cost and the poisoning phenomenon) [28]. In this work, Cu was chosen in order to enhance the ceria reactivity and to be proposed as a valid alternative to noble metal-based catalysts.

According to the literature, CeO₂-based catalysts are interesting materials for both the CO and ethene oxidations [17,18,29,30,31,32,33,34,35,36]. Ceria exhibits good redox properties due to the presence of the Ce³⁺–Ce⁴⁺ couple and it is able to store a high amount of oxygen. Moreover, the addition of foreign metals inside the CeO₂ framework allows to further improve its promising performances [37,38,39,40,41,42,43,44,45].

Among the foreign metals, Cu is of particular interest. In fact, copper can be found as Cu⁺ and Cu²⁺, which could create a synergistic effect in terms of coupled Ce and Cu redox pairs (i.e., Ce³⁺–Ce⁴⁺ and Cu²⁺–Cu⁺) during the oxidation reactions. Specifically, the presence of CuO enhances the redox mechanism at the catalysts surface, because CuO and CeO₂ can be reduced and oxidized at the same time [46,47,48]. Moreover, copper provides new active sites for the adsorption of reduced molecules. Hence, CuO–CeO₂ systems are highly effective for both the CO and ethene oxidation reactions [14,17,18,49,50,51].

In this work, a set of CuCeO_x catalysts was studied for the catalytic oxidation of CO and ethene. The latter was used as probe molecule for the oxidation of VOCs. The samples were also characterized by complementary techniques, such as X-Ray Diffraction (XRD), N₂ physisorption at −196 °C, Field Emission Scanning Electron Microscopy (FESEM), X-ray Photoelectron Spectroscopy (XPS), Fourier Transform Infra-Red spectroscopy (FT-IR), Raman spectroscopy and H₂ Temperature-Programmed Reduction (H₂-TPR) in order to unveil which physico-chemical properties determine the catalytic activity.

2. Materials and Methods

2.1. Synthesis of CuCeO_x Catalysts

A Multi-Inlet Vortex Reactor (MIVR) was used to synthesize the CuCeO_x catalysts. Five samples were prepared with different Cu wt.% content, namely 5, 10, 20, 30 and 60, and herein labeled as 5%CuCeO_x, 10%CuCeO_x, 20%CuCeO_x, 30%CuCeO_x and 60%CuCeO_x, respectively. Moreover, pure CeO₂ and CuO were prepared for comparison purposes. The synthesis procedure was adapted from reference [52].

Two 60 mL syringes were filled with a 0.1 M NaOH (Merk, Steinheim, Germany). Afterwards, a metal precursor solution (0.1 M) at the desired weight ratio was prepared with Ce(NO₃)₃•6H₂O and Cu(NO₃)₂•3H₂O (Merck, Steinheim, Germany). The metal precursor solution was put inside other two syringes (60 mL). Then, the four syringes were connected to the MIVR inlets (Ø_inlet = 1 mm), as shown in Scheme 1A and 1B. The syringes are driven by a KD Scientific KDS220 infusion syringe pump (Merck, Steinheim, Germany) with the total flow rate fixed to 20 mL min⁻¹ in order to have a turbulent flow in the MIVR reaction chamber (Ø_{reaction chamber} = 4 mm), according to previous fluid dynamic studies (Re = 8320) [52,53]. During the injection, the metal hydroxides nanoparticles (CuCeOH NPs) were formed in the reaction chamber, meanwhile the suspension was moving upward in a spiral way towards the outlet on the top of the reactor. Then, the suspension was collected into a beaker (Scheme 1C).

Afterwards, the thus-obtained suspension was stirred for 1 h at room temperature. Then, it was centrifuged (5000 rpm, 1 h) and the produced powder was washed three times with ultra-pure water (MilliQ, Direct Water Purification System, Merk, Steinheim, Germany). Finally, the powder was dried overnight at 60 °C and calcined at 650 °C for 4 h (heating rate of 10 °C min⁻¹).

2.2. Characterization Techniques

The X-ray diffraction (XRD) patterns of the powders were registered by means of an X’Pert Philips PW3040 diffractometer (Malvern Panalytical Ltd., Malvern, UK) and using Cu Kα radiation (2θ = 10–80°; step = 0.013° 2θ; time per step = 0.2 s). The data obtained was indexed using the Powder Data File database (PDF-2 1999, International Centre of Diffraction Data).

The measurement of the textural properties of the catalysts (i.e., specific surface area and total pore volume) was carried out by performing N₂ physisorption at −196 °C in a Micromeritics Tristar II 3020 (v1.03, Micromeritics Instrument Corp., Norcross, GA, USA). Before the analysis, the samples were outgassed at 200 °C for 2 h. The values of specific surface area (SSA) were calculated through the Brunauer–Emmett–Teller (BET) method. On the other hand, the pore volume (V_p) was assessed by analyzing the desorption phase using the Barrett–Joyner–Halenda (BJH) method.

The morphology of the catalysts was investigated with a field emission scanning electron microscope (FESEM Zeiss MERLIN, Gemini-II column, Oberkochen, Germany), in which Energy Dispersive X-Ray (EDX) analysis was performed as well.

The elemental composition of the samples was also investigated using an Inductively Coupled Plasma Mass Spectrometer (iCAP Q ICP-MS, ThermoFisher Scientific, Waltham, MA, USA). For this analysis, 100 mg of each sample were dissolved in an aqueous solution of hydrochloric acid (1 M), nitric acid (1 M) and ascorbic acid (0.5 M), which was then stirred for 8 h to completely solubilize the powder.

Raman spectroscopy was carried out with a Renishaw InVia micro-Raman spectrometer (Renishaw plc, Wotton-under-Edge, UK). The spectra were collected at room temperature using a 514.5 nm excitation wavelength, a 10 mW laser power, a 5× objective and a 225 s total acquisition time. For each sample, three spectra were collected in different points and averaged. The ratio between the area of the defect-related band (D band) and that of the ceria F_2g peak (D/F_2g ratio) was calculated as an estimation of defect abundance. The areas were obtained by deconvolution of the main bands using Lorentzian fitting curves.

FT-IR (Fourier transform infrared) spectroscopy was performed using a Bruker Invenio S spectrophotometer (Bruker Corporation, Billerica, MA, USA) equipped with a cooled MCT detector. Self-supporting wafers were prepared by pressing the catalyst powder and were analyzed in transmission mode in a quartz cell equipped with KBr windows. Prior to analysis, the wafers were outgassed at 50 or 150 °C for 1 h to remove water, using a standard vacuum frame. The spectra were collected at a 2 cm⁻¹ resolution and were normalized to the mass per unit area of each wafer.

The core-level spectra of the elements of interest were studied through the X-ray photoelectron spectroscopy (XPS) technique. The analyses were performed in a PHI Versa probe apparatus (Physical Electronics Inc. PHI, Chanhassen, MN, USA) using: band-pass energy: 187.85 eV; take-off angle: 45°; X-ray spot size diameter: 100 μm.

The reducibility of the catalysts was investigated performing temperature-programmed reduction analyses under a flow of H₂ (H₂-TPR). The analyses were carried out in a ThermoQuest TPD/R/O 1100 (ThermoFisher Scientific, Milan, Italy), comprising a thermal conductivity detector (TCD). Before the analysis, an oxidative pretreatment was carried out under a flow of O₂ (20 mL min⁻¹) at 550 °C for 1 h. Subsequently, the catalysts were put under a reducing gas mixture containing 5 vol.% of H₂/Ar (flowrate: 20 mL min⁻¹). During a typical analysis, the temperature was increased from 50 °C until 900 °C (heating rate 10 °C min⁻¹). At the end of the analysis, the temperature was held at 900 °C for 10 min. For the H₂-TPR analysis, the onset temperature (T_onset) was considered as a catalytic parameter. The T_onset is defined as the intersection of the tangent passing through the inflection point of the first reduction peak and the baseline.

2.3. Catalytic Activity Tests

The catalytic activity of the prepared powder catalysts was performed in a temperature-programmed oxidation (TPO) setup comprising a PID-controlled furnace. A quartz U-tube reactor (ID = 0.4 cm) hosted a fixed-bed containing the catalyst. The temperature was quantified using a K-type thermocouple set in the upper limit of the fixed-bed. The gas outlet from the reactor was analyzed by a non-dispersive infrared analyzer ABB Uras 14 (ABB S.p.A – PAMA, Milan, Italy).

2.3.1. CO Oxidation Reaction

The catalytic bed for the CO oxidation reaction was prepared by putting 100 mg of catalyst inside a quartz-U reactor. Then, the catalyst was pretreated with 50 mL min⁻¹ of N₂ for 30 min at 100 °C, in order to desorb impurities on the catalyst surface (i.e., H₂O). Then, the temperature was cooled down until 25 °C. After that, a gas mixture containing about 1000 ppm CO and 10 vol.% O₂ balanced with N₂ was sent to the reactor (total flow 50 mL min⁻¹). Then, the temperature was increased by 25 °C (heating rate 5 °C min⁻¹). The temperature was kept constant for 30 min in order to have a complete stabilization at each increment of the temperature and to avoid adsorption-desorption phenomena. The test finished when the CO conversion reached the 100%.

2.3.2. Ethene Oxidation Reaction

For the ethene catalytic oxidation tests, 100 mg of catalyst were introduced into the quartz-U reactor. Prior to the catalytic testing, a degassing pretreatment under a flow of N₂ (50 mL min⁻¹) at 150 °C was performed for 1 h. Afterwards, the temperature was set to 100 °C and the catalytic test started. During this test, a gaseous mix containing C₂H₄ (500 ppm), O₂ (10 vol.%) and N₂ (balance) was flowed (50 mL min⁻¹) in the reactor. In addition, the temperature was raised from 100 °C to 425 °C. Isothermal steps were performed every 25 °C until the conversion of the ethene (in terms of outlet CO and CO₂ concentrations) was stable.

3. Results

3.1. Structural and Textural Properties

The crystalline structure of the prepared samples was investigated by means of XRD analysis in the 10° < 2θ < 80° range. The diffractograms of the catalysts are reported in Figure 1. As a whole, all the samples (except for CuO) exhibit the typical CeO₂ cubic fluorite structure, characterized by the (1 1 1), (2 0 0), (2 2 0), (3 1 1), (2 2 2), (4 0 0), (3 3 1) and (4 2 0)-type planes [31,33,37,54,55]. The patterns of these samples were compared to reference code 96-900-9009 (CeO₂). The CuO (reference code 96-101-1195) has the main diffraction peaks at 35.6° ((0 2 2)-type plane) and 38.8° ((1 1 1)-type plane), which are ascribed to the typical monoclinic crystalline phase of CuO [56].

Notably, the samples 30%CuCeO_x and 60%CuCeO_x exhibit both CuO and CeO₂ patterns. This outcome evidences the presence of two different phases (segregates). Furthermore, the presence of weak reflection peaks of the CuO was detected for 10%CuCeO_x and 20%CuCeO_x too (Figure S1).

The crystallites dimensions (D_c) were evaluated by Scherrer’s equation and reported in

Table 1. Textural properties of the CuCeO_x catalysts derived by means of X-ray diffraction, N₂ physisorption and Raman spectroscopy.

Catalyst	Dc a (nm)		SSA b (m2 g−1)	Vp c (cm3 g−1)	D/F2g d
	CeO2	CuO
CeO2	15	-	45	0.07	0.06
5%CuCeOx	13	-	50	0.08	0.13
10%CuCeOx	11	7	55	0.10	0.16
20%CuCeOx	10	9	50	0.09	0.34
30%CuCeOx	14	26	22	0.03	0.13
60%CuCeOx	16	39	11	0.06	0.09
CuO	-	73	2	0.01	-

^a Crystallite size estimated by the Scherrer’s equation; ^b Specific surface area calculated by the Brunauer–Emmett–Teller (BET) method; ^c Total pore volume evaluated by the Barrett–Joyner–Halenda (BJH) method; ^d Ratio between the area of the defect band and that of the F_2g peak in the Raman spectra.

. As a whole, all the samples have CeO₂ crystallites with similar dimensions, in the range between 10 and 16 nm. Instead, the crystal size of the CuO phase gradually increases with the Cu content, from 7 nm (10%CuCeO_x) to 73 nm (pure CuO).

The SSA and the total pore volume were investigated through the N₂ physisorption analysis at −196 °C. The results are resumed in Table 1. The 10%CuCeO_x presents the highest SSA (55 m² g⁻¹, V_p = 0.10 cm³ g⁻¹), while the 5%CuCeO_x and 20%CuCeO_x have promising textural properties as well (SSA = 50 m² g⁻¹, V_p = 0.08–0.09 cm³ g⁻¹).

The morphology of the CuCeO_x catalysts was investigated by means of FESEM analysis and the micrographs are reported in Figure 2. The CeO₂, 5%CuCeO_x, 10%CuCeO_x and 20%CuCeO_x images suggest a good homogeneity of the particle size. Conversely, at higher Cu content, namely, 30 and 60 wt.%, big CuO particles were observed along with small CeO₂ nanoparticles. The CuO phase in 30%CuCeO_x and 60%CuCeO_x presents the typical octahedral or truncated-octahedral shape, confirmed by other research works [56,57].

The elemental composition of the different catalysts was evaluated through EDX and ICP analyses (

Table 2. EDX and ICP analysis of the CuCeO_x catalysts: the elemental content is expressed in percentage of Ce and Cu with respect to the total amount of cations (Ce + Cu).

Catalyst	Elemental Composition (wt.%)
	EDX *			ICP
	Ce	Cu	Tot.	Ce	Cu	Tot.
5%CuCeOx	97	3	100	97	3	100
10%CuCeOx	92	8	100	90	10	100
20%CuCeOx	82	18	100	79	21	100
30%CuCeOx	67	33	100	67	33	100
60%CuCeOx	32	68	100	37	63	100

* The values are estimated over three different areas.

and Table S1): for all the samples, both the techniques confirmed a copper content similar to the nominal one.

Figure 3A displays the Raman spectra of the different catalysts. Raman spectroscopy can indeed provide insight into the modifications induced in the catalyst structure and defectiveness by the increasing content of Cu. Pure ceria features an intense F_2g peak at 464 cm⁻¹ related to the symmetric stretching of Ce-O bonds in the fluorite lattice [58]. A much less intense band at about 595 cm⁻¹ (D band) is instead associated with the presence of defects and it is usually assigned to intrinsic Frenkel sites, in which a vacancy is generated by the displacement of an oxygen atom into an interstitial position [58,59].

The ceria-copper mixed oxides are characterized by analogous Raman spectra, although some differences are evident. The F_2g peak of doped ceria is broader and red-shifted to 448–451 cm⁻¹, evidencing structural distortion [60,61]. Moreover, the intensity of the defect band gradually increases upon copper addition and new defect-related components appear. In detail, a peak centered at 620–630 cm⁻¹ can be ascribed to substitutional sites containing a copper atom in 8-fold coordination [62,63], while a shoulder band around 550 cm⁻¹ arises as a consequence of oxygen vacancies generation [64,65,66]. These outcomes point out that copper is partially incorporated in ceria structure, causing lattice distortion and the formation of new extrinsic defects. At high Cu loading, the typical peaks of CuO starts to be visible in the Raman spectra of the mixed catalysts, as a consequence of the segregation of this oxide. These components, centered at about 280, 335 and 612 cm⁻¹, have been ascribed to one A_g and two B_g modes, respectively [67].

Spectral deconvolution allowed to estimate the abundance of defects in the cerium oxide structure, by calculating the ratio between the area of the defect (D) band and the area of the F_2g peak of ceria (D/F_2g) [63,68]. For highly doped ceria, a peak was set at 612 cm⁻¹ in order to take into account the presence of segregated CuO, and its contribution was not considered in the calculation of the D band area. The D/F_2g values are reported in Table 1 and their trend is depicted in Figure 3B. The amount of defects initially grows with increasing Cu doping and it is maximum for the 20%CuCeOx sample. However, further addition of copper results in the segregation of CuO particles rather than in the formation of structural defects in the ceria framework, in line with previous studies [17].

FT-IR spectroscopy was employed for studying the chemical species at the surface of the different catalysts. FT-IR spectra collected after outgassing the samples at 150 °C are reported in Figure 4. The thermal treatment allows to remove physisorbed water and to more clearly detect the vibrational modes of the surface species (for comparison purposes, spectra of samples outgassed at 50 °C are displayed in Figure S2).

The bands in the 3800–3000 cm⁻¹ region are related to the stretching of the hydroxyl groups. Different OH species are present at the surface of the ceria-rich materials: the band at 3710 cm⁻¹ can be ascribed to isolated OH groups while the intense peak at 3655–3650 cm⁻¹ was assigned to bridging OH [69,70]. The asymmetry of the latter component has been previously attributed to the presence of bridging OH in the proximity of an oxygen vacancy, giving rise to a shoulder at 3640–3630 cm⁻¹ [69]. The less intense band at about 3520 cm⁻¹ may instead be related to triply coordinated OH [69,70]. The presence of a variegated population of OH groups is an index of ceria hydrophilicity. According to previous studies, a hydrophilic support can promote the catalytic activity of metallic active sites [71]. All the hydroxyl species become less abundant at high copper loading and are no more detectable in pure copper oxide, suggesting that Cu doping can reduce ceria hydrophilicity.

The 1700–800 cm⁻¹ region is featured by several broad bands which signal the presence of a heterogeneous population of carbonate and formate species, including monodentate, bidentate, bridged and polydentate carbonates [58,70,72,73]. The carbonate fingerprint is quite similar for pure ceria and for the mixed oxides containing up to 30% of copper. Conversely, new components can be noticed in the FT-IR spectrum of the 60%CuCeO_x sample, probably linked to carbonate species at the surface of CuO particles. An almost flat profile was obtained for pure copper oxide, likely due to the much lower specific surface area of this material (see BET analysis in Table 1).

3.2. Chemical Properties

3.2.1. Surface Oxidation States

The oxidation state of the elements of interest and the surface species of the catalysts were investigated through XPS measurements. The Ce 3d, O 1s and Cu 2p spectra are reported in Figure 5, while the Cu LMM Auger spectra are shown in Figure S3.

The Ce 3d core levels are reported in Figure 5A, which shows the different u and v peaks corresponding to 3d_3/2 and 3d_5/2 spin orbital, respectively. The Ce³⁺ amount was estimated by the deconvolution of the u₁ and v₁ peaks, while the amount of Ce⁴⁺ was evaluated by the deconvolution of u₀, v₀, u₂, v₂, u₃ and v₃ peaks [17,18,31]. The relative amounts of Ce³⁺ and Ce⁴⁺ are reported in

Table 3. Relative amount of Ce species as derived from XP spectra of Ce 3d core level.

Catalyst	Ce3+ (at.%)	Ce4+ (at.%)
CeO2	12	88
5%CuCeOx	22	78
10%CuCeOx	25	75
20%CuCeOx	26	74
30%CuCeOx	19	81
60%CuCeOx	21	79

. The 10%CuCeO_x and 20%CuCeO_x samples exhibit the highest Ce³⁺ amount, namely 25 and 26 at.%, respectively.

The O 1s core levels are displayed in Figure 5B. Each O 1s spectrum is characterized by two peaks: the one at higher binding energy (BE) which is related to chemisorbed oxygen, O_α (O₂²⁻, carbonates, O⁻ or OH⁻), while the one at lower binding energy is due to lattice oxygen, O_β (O²⁻ of the Ce-O bonds) [17,31,54,74,75]. The relative abundance of the O_α and O_β species is reported in

Table 4. Relative amounts of the surface-adsorbed oxygen (O_α) and lattice oxygen (O_β) species and their respective ratios, as derived from the deconvolution of the O 1s core level.

Catalyst	Oα		Oβ (Cu-O)		Oβ (Ce-O)		Oα/Oβ
	BE (eV)	%-Atom	BE (eV)	%-Atom	BE (eV)	%-Atom
CeO2	531.2	26	-	-	529.0	74	0.35
5%CuCeOx	530.9	32	-	-	529.1	68	0.47
10%CuCeOx	531.4	33	-	-	529.4	67	0.49
20%CuCeOx	531.3	40	-	-	529.3	60	0.67
30%CuCeOx	531.4	31	529.9	36	528.8	33	0.45
60%CuCeOx	531.3	29	529.6	45	528.7	26	0.41
CuO	531.4	43	529.8	57	-	-	0.75

. Interestingly, when the Cu content increases, two different O_β species can be noticed, due to the simultaneous presence of two different oxides, CeO₂ and CuO. This is clearly evident for the 30%CuCeO_x and 60%CuCeO_x samples (Figure 5B, curves e and f): the two O_β peaks can be ascribed to Cu-O (529.6–529.9 eV) and Ce-O (528.7–528.8 eV). Moreover, when the Cu content increases, the O_β peak generally becomes more intense [74,75,76,77,78]. Furthermore, the O_α-to-O_β ratio was calculated and reported in Table 4. Interestingly, the 20%CuCeO_x sample exhibits the highest ratio, namely, 0.67, possibly due to the higher content of structural defects (vide Raman) [79]. The presence of abundant O_α may enhance the reactivity during the oxidation reactions since these oxygen species, namely, O₂²⁻, carbonates, O⁻ or OH⁻, are more reactive compared to bulk oxygen [17,54,74,75].

The Cu 2p spectra are reported in Figure 5C. The most intense peak can be deconvoluted into two different Cu²⁺ species: the peak at lower binding energy is related to Cu-O bonds in the lattice, while Cu-CO₃ or Cu-(OH)₂ species bonded at the surface give a contribution at higher binding energy. The Cu LMM Auger spectra are reported in Figure S3. The Auger peaks are centered in the range 917.0–917.7 eV, thus confirming the presence of Cu²⁺ species in the CuO phase and the absence of Cu metallic phase which could be ascribed to the presence of possible alloy [74,77,78].

3.2.2. Reducibility Tests

The H₂-TPR measurement was performed in order to investigate the reducibility of the CuCeO_x catalysts and the obtained profiles are reported in Figure 6A. Pure ceria is characterized by a side peak in the 300–400 °C range which can be attributed to the reduction of chemisorbed oxygen species (i.e., carbonates) that are feebly attached to the catalyst surface [80,81,82]. Furthermore, a main reduction peak is centered at 540 °C and it is due to the reduction of surface Ce⁴⁺ species [17,18,54,83]. Finally, the signal occurring at higher temperatures (T > 600 °C) can be attributed to the reduction of Ce⁴⁺ in the bulk [17,18,54,80,81,82,83].

The samples containing a Cu loading between 5 and 60 wt.% exhibit two convoluted reduction peaks in the 150–500 °C range, as shown in Figure 6A. As a whole, the reduction of these CuCeO_x catalysts occurs at lower temperatures comparing with pure CeO₂. Thus, the reducibility of these samples was enhanced by the presence of CuO alongside CeO₂ phase. According to the literature, the first low temperature peak was attributed to the reduction of Ce-O-Cu species, while the second one was assigned to the reduction of small CuO clusters interacting with CeO₂ [17,18,34,74,80,83,84]. Moreover, when the copper loading increases, the second peak shifts towards higher temperatures and this phenomenon may be related to the larger CuO particles. Figure 6B displays the trend of the position of the T_onset reduction peaks as a function of the Cu content. The following trend for the surface reducibility can be outlined: 10%CuCeO_x < 5%CuCeO_x < 20%CuCeO_x < 30%CuCeO_x < 60%CuCeO_x < CuO < CeO₂.

Table 5. Hydrogen consumption (mmol g⁻¹) evaluated by H₂-TPR.

	H2-Uptake Total	H2/Cu	H2-Uptake <200 °C	H2-Uptake 200–400 °C	H2-Uptake 400–500 °C	H2-Uptake >500 °C
CeO2	1.41	-	0.01	0.11	0.16	1.13
5%CuCeOx	1.33	3.46	0.16	0.52	0.05	0.60
10%CuCeOx	1.91	1.50	0.15	1.21	0.06	0.49
20%CuCeOx	3.59	1.34	0.09	2.52	0.30	0.68
30%CuCeOx	5.35	1.27	0.05	4.07	0.43	0.79
60%CuCeOx	8.83	1.11	0.02	5.86	1.99	0.96
CuO	12.57	1.00	0.00	4.37	5.92	2.28

reports the hydrogen uptakes measured during the H₂-TPR analysis. As a whole, the higher the Cu content, the higher the H₂ consumption. Moreover, the H₂-to-Cu consumption ratio was estimated and reported in Table 5. This value corresponds to the ratio between the H₂ consumed during the TPR analysis and the stoichiometric amount of H₂ required to reduce the Cu²⁺ species contained in the samples. Notably, the CuCeO_x catalysts present higher H₂-to-Cu consumption with respect to pure CuO, indicating that the Cu²⁺ reduction is accompanied by the reduction of Ce⁴⁺ to Ce³⁺ [17].

3.3. Catalytic Activity

3.3.1. CO Oxidation Reaction

The results of the CO oxidation tests are shown in Figure 7. The CO conversion (%) is reported as a function of the temperature and the catalytic tests were compared with the non-catalytic one (labeled as “no-catalyst”). The red graph displays a magnification in the range between 0 and 125 °C for a better comprehension. Moreover,

Table 6. Catalytic performances of the prepared samples in the catalytic oxidation of CO.

Catalyst	CO Oxidation Rate a (µmol h−1 g−1)	Tparameter (°C)
		T10%	T50%	T90%
CeO2	2.6 × 10−2	258	314	372
5%CuCeOx	171	46	69	98
10%CuCeOx	257	35	62	90
20%CuCeOx	375	30	58	75
30%CuCeOx	138	46	73	105
60%CuCeOx	2.7 × 10−4	104	139	171
CuO	3.7 × 10−4	151	179	197

^a Calculated at 50 °C.

reports the temperatures at which the catalysts convert the 10, 50 or 90% of CO (T_10%, T_50% and T_90%, respectively) and the CO oxidation rates (µmol h⁻¹ g⁻¹) evaluated at 50 °C.

As a whole, all the samples exhibit good catalytic performances comparing to the thermal CO oxidation (no catalyst). Interestingly, the most performing sample is 20%CuCeO_x: in fact, the CO complete oxidation occurs at 125 °C in the presence of this catalyst. On the other hand, pure CeO₂ presents the worst performance. In detail, as can be seen in Figure 7, the catalytic activity follows this increasing reaction order: CeO₂ < CuO < 60%CuCeO_x < 30%CuCeO_x < 5%CuCeO_x < 10%CuCeO_x < 20%CuCeO_x.

In order to explain the good catalytic performance of the 20%CuCeO_x sample, Figure 8 reports the values of O_α abundancy (at.%), Ce³⁺ abundancy (at.%), D/F_2g ratio and the CO oxidation rate (µmol h⁻¹ g⁻¹) as a function of the Cu content (wt.%). As a whole, a good correspondence was found among the trends of all these parameters. In particular, the Pearson correlation coefficients (reported in Table S2) point out that the catalytic activity for CO oxidation is primarily linked to the presence of reactive O_α species, followed by the quantity of structural defects and lastly by the abundance of Ce³⁺ ions.

Actually, defect sites and Ce³⁺ species are directly involved in the Mars–van Krevelen-type reaction mechanism. This mechanism implies the oxidation of CO by means of structural oxygen, leaving a defective site (i.e., an oxygen vacancy) that is then replenished (oxidized) by gas phase O₂ [65,85]. Moreover, the O_α species, consisting of chemisorbed oxygen on the catalyst surface, are also highly reactive during the oxidation processes. Hence, the higher the amount of defects, Ce³⁺ and O_α species, the higher the oxidative performance of the catalyst.

Consequently, the 20%CuCeO_x mixed oxide, which exhibits the largest quantity of defect sites, O_α and Ce³⁺ species, has the highest CO oxidation rate. The 5%CuCeO_x and 10%CuCeO_x samples exhibit similar reducibility properties to 20%CuCeO_x (see Figure 6), but their lower abundancy of O_α, Ce³⁺ and structural defects is likely responsible for their slightly lower catalytic activity.

The CO oxidation performances of the 20%CuCeO_x sample were compared with those of the most performing catalysts studied in references [17,86,87,88,89,90,91,92]. The latter research works deal with copper-cerium based catalysts synthesized by means of different co-precipitation techniques, such as Solution Combustion Synthesis (SCS) [17,86], Sol-gel [87,88], hydrothermal [89,90] and wet impregnation method [91,92]. The comparison, in terms of T_50% for CO oxidation, is shown in Figure 9.

The 20%CuCeO_x sample exhibits the lowest T_50% (58 °C) comparing with the other Cu–Ce mixed catalysts. Then, a last comparison was performed with Pt/CeO₂ catalysts studied in references [93,94,95] and it is possible to conclude the CuCeO_x samples show good low-temperature CO oxidation activity compared with ceria-supported noble-metal. Thus, the MIVR method represents a promising synthesis procedure, since it enables to obtain CuCeO_x nanocatalysts with very interesting physico-chemical and catalytic properties.

In order to investigate the effect of the co-presence of CeO_x and Cu_xO phases, the CO oxidation rates listed in Table 6 are plotted in Figure 10 as a function of the Cu weight content. A growing trend could be observed for Cu doping ranging from 5 to 20 wt.%, while a decreasing one was noticed upon further Cu addition. As previously described, the sample containing the 20 wt.% of Cu exhibits the highest reaction rate (375 µmol h⁻¹ g⁻¹).

Moreover, Figure 10 shows a black dashed line representing the theoretical reaction rate of a mixture of completely non-interacting CeO₂ and CuO phases. This hypothetical rate is defined by the following linear function (Equation (1)):

$$r_i=r_{{\mathrm{CeO}}_2}{x}_{{\mathrm{CeO}}_2}+r_{\mathrm{CuO}}{x}_{\mathrm{CuO}}$$

(1)

where r_i represents the theoretical reaction rate, r_CeO2 and r_CuO the reaction rates of cerium and copper oxide, respectively, while x_CeO2 and x_CuO are the mass fractions of cerium and copper oxide. In Figure 10, the red dashed line represents the actual trend of the reaction rates. Interestingly, for all the mixed samples, the real CO oxidation rates are above the theoretical ones (the black dashed line). Thus, a synergistic effect between the two active phases is occurring [96].

Similarly, the T_10% was reported in Figure S4 as a function of the Cu wt.% content. The theoretical T_i value is represented by the following equation (Equation (2)):

$$T_i=T_{{\mathrm{CeO}}_2}{x}_{{\mathrm{CeO}}_2}+T_{\mathrm{CuO}}{x}_{\mathrm{CuO}}$$

(2)

where T_i represents the theoretical T_10%, T_CeO2 and T_CuO the T_10% of cerium and copper oxide, respectively, while x_CeO2 and x_CuO are the mass fractions of cerium and copper oxide. It appears that the two oxide phases exhibit a positive collaboration in agreement with the results reported in terms of reaction rates (Figure 10).

In order to assess the stability of the most performing catalyst (20%CuCeO_x), two consecutive CO oxidation tests were carried out, and the results are reported in Figure 11A. The 20%CuCeO_x sample exhibits good catalytic stability since reproducible results were achieved after the two consecutive runs.

Moreover, another stability test was performed on this sample and reported in Figure 11B. During this test, the T_50% and T_95% were chosen and alternated for two consecutive cycles as follows:

• First cycle: the starting temperature was fixed to 89 °C, at which the catalyst exhibits a CO conversion of about 95% and this temperature was kept constant for 2 h. Then, the catalytic system was cooled down until 58 °C (T_50%) and it was kept constant for other 2 h. After that, the temperature was increased until 89 °C. Hence, the first cycle is characterized by a Time-on-Stream (TOS) of about 5 h.
• Second cycle: the same as the first one.

As a whole, the 20%CuCeO_x catalyst is able to reach, during the second cycle, the same conversion achieved during the first one. It is therefore possible to that the 20%CuCeO_x catalyst exhibits not only a notable activity but also good stability properties since, after two cycles, its performance is fully preserved.

3.3.2. Ethene Oxidation Reaction

The catalytic performances of the synthesized materials measured during the catalytic oxidation of ethene are displayed in Figure 12. The presence of the prepared catalysts allowed the conversion of ethene at lower temperatures, compared to those of the blank test (i.e., the test where no catalyst was present).

Consistently, the catalytic improvement is confirmed by the rates of the ethene oxidation reaction calculated at 175 °C, as included in

Table 7. Catalytic performances of the prepared samples in the catalytic oxidation of C₂H₄.

Catalyst	Ethene Oxidation Rate a (µmol h−1 g−1)	Tparameter (°C)
		T10%	T50%	T90%
CeO2	15	275	345	396
5%CuCeOx	25	194	238	273
10%CuCeOx	56	177	223	259
20%CuCeOx	62	175	221	252
30%CuCeOx	26	197	242	277
60%CuCeOx	21	217	290	357
CuO	3	237	287	328

^a Calculated at 175 °C.

. As a whole, considering the catalytic performances reported in Figure 12, a catalytic activity trend for the CuCeO_x samples can be outlined, as follows: 20%CuCeO_x > 10%CuCeO_x > 5%CuCeO_x > 30%CuCeO_x > 60%CuCeO_x.

The catalyst containing the lowest loading of Cu (i.e., 5%CuCeO_x) evidenced an overall improved catalytic performance compared to the parent CeO₂, thus highlighting the beneficial effect of incorporating Cu species in the ceria structure (as observed in XRD and Raman analyses). Interestingly, an enhancement of the reaction rate was observed for the prepared set of mixed oxide catalysts (see Figure 13).

As observed in Figure 13, the ethene oxidation rates were in all cases above the theoretical (linear) trend drawn between CeO₂ and CuO (see black dashed line). The mentioned baseline represents an ideal estimation of the reaction rate as if it depended only on the relative amounts of non-interacting CeO₂ and CuO in a mixed sample, as reported in Equation (1) of Section 3.3.1.

Consistently, the increased reaction rates verified for the whole set of mixed oxides (compared to the ideal behavior) evidence that beneficial interactions are taking place between the active components present in the catalysts. Moreover, this result is supported by the trend observed in Figure S5 for the temperatures needed for converting 10% of the ethene present in the gaseous stream (i.e., the T_10% parameter). In fact, this improvement coincides with the contemporary presence of small CuO crystallites along with CeO₂ in the catalysts (as demonstrated by means of XRD, vide supra). In this sense, the overall catalytic performance varies (in terms of ethene reaction rate and T_10%) as a function of the size of CuO crystallites (as reported in Figure S6) and evidenced the following: the smaller the size of CuO crystallites in contact with CeO₂, the better the catalytic performance in ethene oxidation.

Moreover, there is a positive correlation between the catalytic activity and some physico-chemical properties, namely, the amount of defect sites (vide Raman spectroscopy) and the O_α/O_β ratio (vide XPS). The catalytic behavior (in terms of T_X%), as a function of the aforementioned properties is reported in Figure S7. Accordingly, the trends evidence that the higher the amount of O_α species and lattice defects, the better the catalytic performance. As well, XPS revealed that the most active catalysts (i.e., the 20%CuCeO_x followed by the 10%CuCeO_x) contained elevated amounts of Ce³⁺ species (see Table 3).

This complex scenario evidences that synergistic interactions are most likely taking place between CuO and CeO₂ and leading to a marked improvement of the catalytic oxidation of ethene (Figure 13). As reported in the literature, smaller particles may contain a higher number of crystal edges and corners, and thus an increased number of structural defects and reactive sites that may improve the catalytic performance [97]. On the other hand, electrophilic O_α species are highly active in the oxidation of hydrocarbons [16,98,99]. Consistently, the elevated amounts of active O_α species (that can participate in spillover phenomena between CeO₂–CuO phases) could contribute to the catalytic activity of the ceria-based catalysts prepared in this work. Moreover, the results highlight the positive effect of defective sites for the ethene catalytic oxidation reaction.

Similarly to CO oxidation, the catalytic oxidation of VOCs over ceria-based materials could also be modeled as a redox process occurring at the catalytic surface known as Mars–van Krevelen reaction mechanism [100,101]. However, the literature highlights the complexity of VOCs oxidation mechanism over oxide catalysts, as a function of the nature of the reactant molecules and of the catalytic materials [100,101,102,103].

4. Conclusions

In this study, CuCeO_x catalysts prepared by the Multi-Inlet Vortex Reactor (MIVR), were carefully characterized and tested towards the CO and ethene oxidation reactions. The results pointed out that the MIVR procedure can be effective to synthesize mixed oxides with promising physico-chemical properties and catalytic oxidation activity.

The CuCeO_x mixed catalysts present much better CO oxidation performances with respect to pure CeO₂ due to the co-presence of CeO₂ and CuO phases that cooperate during the catalytic cycle. Interestingly, this set of samples is featured by outstanding catalytic activities comparing with the literature results of other copper-cerium based catalyst prepared by different co-precipitation methods and with Pt/CeO₂-type catalysts. The most promising catalyst is the 20%CuCeO_x sample, thanks to its optimized physico-chemical properties: good textural-structural properties (SSA, V_p and D_c), high amount of structural defects (such as oxygen vacancies and Frenkel sites), and great availability of Ce³⁺ and O_α species. This catalyst also exhibited a considerable thermal stability.

The catalytic activity for the oxidation of ethene was observed to vary according to the amount of copper present in the catalyst. Overall, the best catalytic performance for ethene oxidation was observed for the 20%CuCeO_x catalyst. The catalytic activity was ascribed to the contemporary presence of CeO₂ and CuO phases and to their possible synergistic interactions. Moreover, the catalytic performance in ethene abatement increased when the size of the CuO particles decreased, i.e., the smaller the size of CuO particles, the higher the catalytic activity. Finally, it was observed that the occurrence of elevated relative amounts of (i) chemisorbed O_α species and (ii) defect sites positively contribute to the catalytic performances in ethene oxidation.