1 Introduction

During the last few decades, rare earth materials have been widely investigated in the field of heterogeneous catalysis to improve the reactivity, stability and electron conductibility of solid catalysts. In the field of automotive catalysis, in particular, ceria-based catalysts have been the most frequently studied materials [1,2,3,4,5,6,7]. The extensive investigations into ceria hark back to its early application in the three-way catalytic converter [8, 9], a catalytic technology that benefits from the unique ceria redox properties. In fact, CeO2-based systems participate in the oxidation reaction via redox cycles [10] and provide a high oxygen storage capacity (OSC) [4, 11].

Over the years, several studies have shown that ceria nanoparticles may exhibit different morphologies and structures [2, 7, 12,13,14]. In several previous works, well-defined ceria nanocubes and nanorods have been successfully prepared via a hydrothermal synthesis under various aging temperatures [2, 15, 16]. The hydrothermal synthesis, indeed, allows to obtain ceria-based nanoparticles with different morphologies, structural and textural properties, depending on the synthesis conditions (i.e. aging temperature, molarity of the solution, …) [2, 17].

The transmission electron microscopy has shown that a single CeO2-based nanocube consisted of six well-defined (100)-type crystalline planes, while a CeO2 nanorod was predominantly constituted of (110)-type planes [2, 17]. In terms of catalytic activity, both the nanocubes and nanorods have been found to outperform the “conventional” polycrystalline CeO2 (obtained by the solution combustion synthesis) for both the CO oxidation and soot combustion reactions [2, 6, 7, 17, 18]. This finding is most likely due to the lower energy for the formation of oxygen vacancies and higher oxygen reactivity with the low index facets. In fact, the oxygen reactivity among low-index facets is believed to follow the order: (100) > (110) > (111), while the energy for the formation of vacancies is (110) < (100) < (111) [19, 20]. Finally, (111)-type planes were found to be more abundant for polycrystalline ceria.

As is known, the physico-chemical properties of ceria can be modified by introducing foreign metals (e.g., Zr, Pr, Cu, etc.) into the ceria lattice. Depending on the nature of the metal, these additional elements may lead to more structural defects in the ceria framework, higher OSC and better thermal stability [18, 21,22,23,24,25]. Among the frequently investigated foreign metals, Cu and Mn are of particular interest. Specifically, CuO–CeO2 systems have been particularly effective towards both the CO and soot oxidation reactions [23, 26,27,28,29]. A synergistic effect was ascribed to the coupled redox cycles Ce3+|Ce4+ and Cu2+|Cu+ ions, the enhanced number of defective sites (e.g., oxygen vacancies) and the higher surface reducibility [26, 30, 31]. In our previous work, a copper-cerium mixed oxide sample (the Ce0.95Cu0.05 sample prepared through the solution combustion synthesis, Cu/Ce atomic ratio = 1/19) was found to be the best combination for the soot combustion reaction [26]. Likewise, the soot oxidation, NO oxidation and CO oxidation carried out over CeO2–MnOx catalysts also benefit from the coupled redox cycles Mn2O3|MnO2 and CeO2|Ce2O3 [32,33,34]. On the other hand, little attention has been paid to the ternary CeO2–MnOx–CuOx systems, which turn out to give higher catalytic activity towards the CO oxidation and toluene oxidation than the conventional binary oxides [35,36,37].

The present work aims to study a set of ceria-based materials doped with copper and manganese. All the catalysts were synthesised via the hydrothermal synthesis. Four systems were investigated: CeO2 (pure ceria), Ce0.95Mn0.05 (Mn/Ce atomic ratio = 1/19), Ce0.95Cu0.05 (Cu/Ce atomic ratio = 1/19) and Ce0.95Mn0.025Cu0.025 (Mn/Cu/Ce atomic ratio = 1/1/38). CO and soot oxidations were used as probe reactions. Complementary characterisations, such as the powder XRD, N2 physisorption at − 196 °C, FESEM, XPS, micro-Raman spectroscopy, CO-TPR and Soot-TPR analysis were performed to investigate the physico-chemical properties of the prepared catalysts.

2 Experimental

2.1 Catalyst Preparation

Four ceria-based samples were prepared by the hydrothermal synthesis: CeO2, Ce0.95Mn0.05 (Mn/Ce atomic ratio = 1/19), Ce0.95Cu0.05 (Cu/Ce atomic ratio = 1/19) and Ce0.95Mn0.025Cu0.025 (Mn/Cu/Ce atomic ratio = 1/1/38). Subscripts indicate the atomic fractions of the metal cations. Briefly, stoichiometric quantities of Ce(NO3)3·6H2O (Sigma-Aldrich) and Cu(NO3)2·5H2O (Sigma-Aldrich) and/or Mn(NO3)2·4H2O (Sigma-Aldrich) were dissolved in 10 ml of deionised water (solution A). Then, 48 g of NaOH were dissolved in 70 ml of deionized water (solution B). The salt precursor solution (A) was then added drop by drop into the hydroxide solution (B). The final mixture (A + B) was stirred at room temperature for 1 h. The final slurry was then poured into a 150-ml Teflon liner, that was later secured in an autoclave. The autoclave was heated in an oven for an aging treatment at 180 °C for 24 h. The resulted precipitate was washed with ethanol and deionised water and then dried overnight at 60 °C. Finally, the dry powder was calcined at 550 °C for 4 h.

2.2 Catalyst Characterization

Powder X-ray diffractograms of the catalysts were recorded on an X’Pert Philips PW3040 diffractometer using Cu Kα radiation (2θ range = 20°–70°; step = 0.05° 2θ; time per step = 0.2 s). The diffraction peaks were indexed according to the Powder Data File database (PDF 2000, International Centre of Diffraction Data, Pennsylvania). Crystallite size [D c, (nm)] was estimated by the Scherrer’s equation, D c = 0.9 λ/(β cos θ), where 0.9 is the shape factor for sphere, λ is the Cu Kα wavelength (nm), β is the full-width at half maximum [FWHM (rad)] and θ is the Bragg angle (rad). Prior to the calculation, the FWHM data were corrected with respect to those of lanthanum hexaboride (instrument calibration standard).

BET specific surface area (S BET), total pore volume (V p) and pore diameter (Dp) of the catalysts were measured via the N2 physisorption at − 196 °C (Micrometrics ASAP 2020). Prior to the analysis, the catalysts were outgassed at 200 °C for 2 h.

Catalyst morphology was analysed via a field emission scanning electron microscope (FESEM Zeiss MERLIN, Gemini-II column).

Raman measurements were carried out on a Renishaw InVia Reflex micro-Raman spectrometer with laser excitation at 514.5 nm in backscattering light collection. The O v/F 2g value, representative of oxygen vacancies density, was calculated as the ratio between the integrated areas of the oxygen vacancies related Raman bands (O v) and the main vibrational band related to the ceria matrix (F 2g) obtained by averaging the spectra recorded on three different areas of the sample.

The oxidation states for the elements of interest (namely, Ce, Mn, Cu and O) were investigated via X-ray photoelectron spectroscopy, performed in a PHI Versa probe apparatus using a band-pass energy of 187.85 eV, a 45° take-off angle and 100 μm diameter X-ray spot size.

CO-TPR and Soot-TPR measurements were carried out for evaluating the reducibility of the prepared catalysts. CO-TPR is basically an anaerobic catalytic CO oxidation, that is the reaction whose oxygen source solely comes from the catalyst surface. The reduction was performed in a U-tube reactor, whose fixed catalytic bed contains 45 mg of catalyst and 150 mg of silica. The analysis began with a catalyst pre-treatment with 200 ml min−1 of N2 at 100 °C for 1 h. Then, the reducing gas (2000 ppm-vol of CO in N2) was introduced into the reactor at a rate of 200 ml min−1. The ramp heating by a PID-controlled electric furnace (rate = 10 °C min−1) started upon the gas introduction. The test ended when the reactor temperature reached 700 °C.

Similarly, Soot-TPR was additionally carried out to evaluate catalyst reduction in the presence of soot. A bed mixture containing 5 mg of commercial soot (Printex-U), 45 mg of powder catalyst and 150 mg of silica was prepared in “tight contact” and put inside the same fixed-bed reactor.

2.3 Catalyst Performance

A typical temperature-programmed combustion (TPC) setup was used during the test of catalytic performances. The apparatus was comprised of a quartz U-tube fixed-bed reactor (ID = 4 mm, OD = 6 mm), a PID-controlled electric furnace, a K-type thermocouple placed in such a way that the sensor tip was as close as possible to the reactor bed. Temperatures corresponding to 10, 50, 90% conversion (denoted as T10%, T50%, T90%, respectively) were taken as indices of the activity of the tested catalysts. Evolution of CO and CO2 was monitored by a non-dispersive infrared analyser (ABB Uras 14). CO and soot oxidations were used as probe reactions.

2.3.1 CO Oxidation Reaction

The reactor bed contained 100 mg of catalyst. The reactor was initially pre-treated by air (flow rate = 50 ml min−1) for 30 min at 150 °C. Then, the reactor was naturally cooled down until 50 °C while the airflow was still maintained. The test started with gas introduction (1000 ppm-vol of CO and 10%-vol O2 in N2) into the reactor at a rate of 50 ml min−1 at 50 °C. The reactor temperature was then increased manually by 50 °C. After each increment, the reactor was kept isothermal for a certain period to allow better concentration stabilisation, possibly affected by adsorption–desorption phenomena. The test ended when 100% of CO conversion was reached.

2.3.2 Soot Oxidation Reaction (Under “Tight” Contact Condition)

The reactor bed contained an intimate mixture of 5 mg of model soot (Printex-U by Degussa), 45 mg of powdered catalyst and 150 mg of silica, made in a ball-miller (Giuliani Tecnologie Srl) operated at 290 rpm for 15 min. The latter condition guarantees a “tight” contact between the catalyst and soot. This method maximizes the number of contact points and, although it is less representative of the real contact conditions that occur in a catalytic trap, it is able to discriminate better the role of catalyst morphology. The reactor was initially pre-treated by air (flow rate = 100 ml min−1) for 30 min at 100 °C. The test began with gas introduction (50%-vol air and 50%-vol N2) into the reactor at a rate of 100 ml min−1. The reactor temperature was then increased from 100 to 700 °C with a programmed heating rate of 5 °C min−1.

3 Results and Discussion

3.1 Textural Properties

The powder X-ray diffractograms of the prepared catalysts are summarised in Fig. 1. As a whole, they exhibit the typical pattern of the fluorite structure of ceria, that is known to be characterized by the peaks related to the (111), (200), (220), (311) and (222) planes [2, 17, 18]. The inset of Fig. 1 shows the magnification of the (111) peak in the 2θ range (from 28° to 29°). The inset shows that the peaks of the mixed oxide samples are shifted to higher 2θ values. This is most likely due to ceria lattice shrinkage, as the ionic radii of Ce [r ion of Ce3+/Ce4+ (CN = 6) is 1.01/0.87 Å] are generally higher than those of Mn and Cu [r ion of Mn3+/Mn4+ (CN = 6) is 0.58/0.53 Å and r ion of Cu+/Cu2+ (CN = 6) is 0.77/0.73 Å] [38]. The Ce0.95Mn0.05 sample shows the biggest peak shift (ca. 0.25°) as Mn has the smallest ionic radii among the investigated metal cations. As expected, lower shifts are encountered in the diffractograms of Ce0.95Cu0.05 and Ce0.95Mn0.025Cu0.025 samples (0.07° and 0.09°, respectively). In Fig. 1 the intensity of the (200) and (220) peaks, relative to the (111) peak, may shed light upon the relative abundance of low-index planes. Table 1 summarises these two ratios, denoted as (200)/(111) and (220)/(111), respectively.

Fig. 1
figure 1

X-ray diffractograms for the prepared samples. (Inset) Magnification of the (111) peak for the 2θ shift observation

Full size image
Table 1 Textural properties of the prepared catalysts obtained from X-ray diffraction and N2 physisorption
Full size table

As a whole, the addition of the Cu and Mn cations into the ceria framework leads to more abundant low-index planes, particularly in the presence of Cu species. Specifically, the Ce0.95Cu0.05 sample has the highest (220)/(111) and (200)/(111) ratios.

Figure 1 shows that the peaks of CeO2 and Ce0.95Mn0.05 samples appear sharper and narrower than those of Ce0.95Cu0.05 and Ce0.95Mn0.025Cu0.025, thus suggesting bigger crystallite sizes. Then, the Scherrer’s equation was used to estimate the crystallite size (D c), whose calculated values for all catalysts are also summarised in Table 1. The CeO2 sample exhibits larger D c (ca. 135 nm) than the other mixed oxide samples (in the range of 23 to 39 nm). Interestingly, the Cu-containing catalysts have the smallest D c (23–34 nm), thus suggesting a significant change in the CeO2 lattice. Results from N2 physisorption are also summarised in Table 1. The BET specific surface area (S BET) of pure CeO2 is the lowest in the series (9 m2 g−1). On the other hand, the surface areas of mixed oxide samples are generally higher than the one of CeO2. Interestingly, the Cu-containing catalysts (including the ternary oxide) have the surface area almost five times higher than that of CeO2. This suggests one more time the contribution of Cu doping to the modification of the microstructure. As previously suggested from XRD results, Cu doping leads to smaller crystallite size. The decrease in crystallite size may increase the number of interparticle voids in the catalyst framework, and so higher surface area can be observed.

The total pore volume of the catalysts (V p) also follows the trend of surface area. Despite the high variability in S BET, the average pore diameter (D p) seems comparable among the catalysts (in the range of 11–21 nm).

Figure 2 shows the FESEM micrographs for the prepared catalysts. The CeO2 sample, as previously reported in the literature, exhibits a well-defined nanocubic morphology rich in (100)-type facets [2, 17]. The average size of the cubes ranges from 100 to 300 nm. On the other hand, the cubic structure is vanished when copper and manganese are introduced into ceria lattice. Both the Ce0.95Cu0.05 and Ce0.95Mn0.05 samples are comprised of small polyhedra (between 20 and 40 nm), with abundant low-index facets [7, 16, 21]. Similarly, the ternary sample exhibits clusters of small nanoparticles, in agreement with the XRD data (vide supra).

Fig. 2
figure 2

FESEM micrographs for the prepared samples

Full size image

3.2 Surface Oxidation States and Defective Sites

Surface oxygen species of the prepared samples were analysed via the deconvoluted O 1s X-ray photoelectron spectra (Fig. 3a). In general, two major species can be distinguished:

  • the so-called “Oβ” species or commonly known as lattice oxygen species, marked by an intense peak at lower binding energy (BE) values (528.8–529.1 eV) [39, 40]

  • the surface capping oxygen species (the “Oα”) that appears at higher BE values (531.2–531.6 eV). The species can be in the form of hydroxyls (OH−), carbonates (CO3 2−) and/or other anions, such as peroxo (O2 2−) and superoxo species (O2 −) [2, 40, 41]

Fig. 3
figure 3

Deconvoluted X-ray photoelectron spectra of a CeO2, b Ce0.95Mn0.05, c Ce0.95Cu0.05 and d Ce0.95Mn0.025Cu0.025 in the A O 1s and B Ce 3d core levels

Full size image

Table 2 summarises the relative abundances of these two oxygen species for the prepared catalysts. The abundance of Oα in the binary oxide samples appears similar to that for the CeO2 sample. However, when the three cations exist in the catalysts (ternary oxide) the Oα abundance increases. Such a phenomenon has also been observed for the Ce–Pr–Zr ternary oxides [42]. A richer population of Oα species on the catalyst surface may foster oxygen spillover coming from bulk dioxygen to the surface of ceria as it may enhance the electron mobility on the surface. It is widely understood that the spillover mechanism requires a dissociative adsorption of molecular oxygen step, during which electron transfers take place from metal sites to the antibonding π-orbital of the oxygen molecule until cleavage of the O–O bond is reached. This mechanism can be expressed as follows [43, 44]:

$${O_{2,{\text{ads}}}}\left( {+{{\text{e}}^ - }} \right) \leftrightarrow {{\text{O}}_{2,{\text{ads}}}}^{ - }\left( {+{{\text{e}}^ - }} \right) \leftrightarrow {{\text{O}}_{2,{\text{ads}}}}^{{2 - }} \leftrightarrow 2{{\text{O}}^ - }_{{2,{\text{ads}}}}\left( {+2{e^ - }} \right) \leftrightarrow 2{{\text{O}}^{2 - }}$$
Table 2 Relative abundances of Oα, Oβ and Ce3+ species (in atomic percentage) derived from deconvoluted XP spectra
Full size table

Figure 3b shows the deconvoluted Ce 3d spectra of the prepared catalysts. By default, eight sub-peaks have been deconvoluted in the spectra: the v quadruplets in the Ce 3d 5/2 region and the u quadruplets in the Ce 3d 3/2 region. Two sub-peaks v 1 and u 1 (marked by areas with darker shades on the figure) are the unique markers for Ce3+ species while the rest of the sub-peaks are ascribed to Ce4+ species [39, 45, 46]. Table 2 also summarises the abundances of the Ce3+ species (in atomic %) relative to the Ce4+ species of the prepared catalysts. The Ce0.95Mn0.05 sample has a similar surface Ce3+ abundance with the CeO2 sample, while the Ce0.95Cu0.05 sample has a higher abundance of Ce3+ species. This suggests that the Cu insertion in the CeO2 framework promotes the creation of Ce3+ species.

For the sake of brevity, the Cu 2p and Mn 2p spectra of the doped catalysts are not shown. In fact, due to their low concentrations in the solid solutions (2.5 and 5.0 at.% for ternary and binary oxides, respectively), the characteristic peaks are very low in intensity and hence hardly observable in the XPS spectra. Therefore, peak deconvolutions were not performed as they may lead to unreliable interpretations.

The Raman spectra of the four catalysts are reported in Fig. 4a. Indeed, Raman spectroscopy can provide additional insight into the influence of doping on the structural features of ceria and highlight the formation of new defective sites [47, 48]. All the spectra were normalised to the main Raman peak of ceria-based materials, the F2g mode, ascribed to the symmetric vibration of the Ce–O bonds of the fluorite crystal unit [49]. The position of this peak is sensitive to the incorporation of the dopants as witnessed by its marked red-shift in the binary and ternary oxides (Fig. 4b). The displacement of the F2g band centre towards lower Raman shifts is usually explained in terms of expanded lattice parameters due to the incorporation of bigger cations combined to an increased defectiveness and strain effects [48].

Fig. 4
figure 4

a Raman spectra of the prepared catalysts; b enlarged view of the F2g peak; c enlarged view of the defect bands region

Full size image

While the first phenomenon can be ruled out based on the ionic radius of the dopants and on the XRD measurements, both the increased content of oxygen vacancies and strains induced by the shrinkage of the unit cell are likely to contribute to the red shift. The higher strain due to the replacement of Ce4+ with the smallest cation in the Ce0.95Mn0.05 sample is in agreement with the strongest shift observed for the F2g mode. Together with the change in position, a broader linewidth is detected for the doped samples. An increased peak width may depend on a higher defectiveness [50], which can be assessed through the analysis of the spectral region between 500 and 650 cm−1, where disorder-induced bands can be identified. A single peak located at 595.9 cm−1 characterises the Raman pattern of pure ceria and, according to the work of Wu et al. [51] was previously assigned to oxygen vacancies part of a Frenkel anion pair. The defect band becomes broader and significantly more intense for the doped materials. Indeed, different defect sites can feature the ceria matrix. Usually, they are distinguished in intrinsic and extrinsic defects. The former consist of (i) oxygen vacancies due to the presence of reduced Ce3+ cations (DI Raman peak around 550–560 cm−1) [51] and (ii) Frenkel type anion defects (DII, 595–600 cm−1) [52]. Instead, extrinsic defects can be introduced by the incorporation of doping elements. In detail, oxygen vacancies created to compensate the lower positive charge of trivalent dopant cations are believed to provide vibrational bands close to the DI species (EI) [53], while the presence of substitutional cations in eightfold coordination has been also suggested to produce a defect-related band around 600 cm−1 (EII) [54]. Moreover, the dopants incorporation can furtherly promote the formation of intrinsic defects in ceria. A closer inspection of the defect spectral range, shown in Fig. 4c, allows to resolve a new component at higher Raman shifts for the Cu and Mn containing catalysts, in addition to the Frenkel defects mode around 595 cm−1. According to the deconvolution of the broad defect band, such peak is found at 628 cm−1 for the Ce0.95Cu0.05 sample, 637 cm−1 for the Ce0.95Mn0.05 sample and at an intermediate Raman shift value for the ternary oxide. The dependence of its spectral position on the dopant type suggests its classification as an extrinsic defect band (EIII). It should be noted, that besides this new component, also the peak at 595 cm−1 grows in intensity in presence of dopants, contributing to the overall increase in the intensity of the oxygen vacancies related band compared to pure ceria. In particular, the oxygen vacancy content for each catalyst was estimated by means of the calculated ratio between the total integrated area of the DII and EIII bands and the one of the F2g peak (Ov/F2g). Such values are summarised in Table 3, from which it can be inferred that the Cu-containing catalysts are featured by the highest Ov/F2g ratios. This could be ascribed to an increased defect density, taking into account that Ce0.95Cu0.05 and Ce0.95Mn0.025Cu0.025 are characterised by the highest specific surface area.

Table 3 Calculated Ov/F2g ratios for the different catalysts
Full size table

3.3 Reducibility

The employment of carbon monoxide as the reducing agent in a TPR analysis benefits from its direct correlation with CO oxidation, one of the probe catalytic reactions investigated in the present work. Under anaerobic conditions, CO consumes the surface and bulk oxygens that come from the solid catalyst. The oxygen species are recognisable from the temperatures at which their corresponding CO2 production peaks occur.

Figure 5 shows the evolution of CO2 (%) as a function of temperature during the CO-TPR analysis. It can be observed that at low temperatures (from 100 to 200 °C), the reduction lights off over doped-ceria catalysts. The Cu-containing catalysts (Ce0.95Cu0.05 and Ce0.95Mn0.025Cu0.025) demonstrate the earliest oxygen consumption by CO, followed by the Ce0.95Mn0.05 catalyst. The low-temperature reduction peak is most likely due to the weakening of Ce–O bonds near the Ce-dopant (Cu/Mn) interface on the surface. On the other hand, in the previous study with ceria-praseodymia catalysts, the reduction-peak occurred at higher temperatures (ca. 300 °C) [18]. This suggests that Mn/Cu perturbs the structure of ceria more predominantly on the surface than in the lattice, while the opposite has been observed for ceria-praseodymia catalysts.

Fig. 5
figure 5

CO-TPR profiles: CO2 evolution as a function of temperature

Full size image

To better quantify the reducibility at low temperature, the specific reaction rate for the CO oxidation under anaerobic conditions (r CO) has been used. The latter can be defined as the rate of CO conversion normalised with respect to specific surface area of the catalyst at a certain temperature (mmolCO m−2 h−1). Table 4 summarises the r CO values at 140 °C for the prepared catalysts. While r CO is the lowest for the ceria sample (5.11 mmolCO m−2 h−1), the metal doped-ceria catalysts somewhat demonstrate higher CO oxidation rates. In fact, the Ce0.95Mn0.05 sample has a similar rate of the Ce0.95Mn0.025Cu0.025, despite the reduction lights off at a higher temperature. Conversely, the Ce0.95Cu0.05 sample exhibits higher CO oxidation rate (22.70 mmolCO m−2 h−1) than the other mixed oxides. As a whole, the trend for the light-off temperatures among the catalysts follows this order:

$${\text{Ce}}{{\text{O}}_2}>{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}}>{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}>{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}},$$

while the trend for the BET surface areas is the reverse

$${\text{Ce}}{{\text{O}}_2}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}}<{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}.$$
Table 4 Summary of the catalytic performances obtained from the tests with CO-TPR, CO-oxidation and soot oxidation
Full size table

Interestingly, there is a direct correlation between the specific reaction rate of CO oxidation under anaerobic conditions and the abundance of structural defects on the catalyst surfaces, as evaluated by the Raman spectroscopy. Then, there is a broad activity region between ca. 200 and 450 °C, in which the reduction progressively increases for both the binary and ternary oxides. Conversely, the reduction for CeO2 become significant at around 400 °C.

Above 400 °C, further reduction peaks can be observed for all the catalysts. The ternary oxide has the most intense peak centred at around 530 °C, thus confirming the beneficial role of both Cu and Mn species to reduce the ceria-based catalyst in this temperature range. On the other hand, similar reduction profiles can be observed for the binary oxide catalysts. As expected, the least intense reduction peak occurs for the CeO2. At high temperature (550–700 °C) all catalysts show high activity, although the production of CO2 seems constant. This phenomenon may be ascribed to the consumption of bulk oxygens from the ceria lattice [18, 42].

The reducibility of the catalysts was also examined by the Soot-TPR analysis, in which the samples were exposed to soot under the increasing temperature and in absence of gas-phase oxygen (anaerobic condition). This new characterization technique allows to evaluate the overall oxygen content in the solid, the reaction selectivity for CO/CO2 and the intrinsic activity of the catalyst [42].

As is known, the catalytic soot oxidation takes place by the oxygen of the solid catalyst according two key steps: (i) catalyst reduction at low temperature (in the range of 300 to 500 °C) that produces mainly CO2 and (ii) high-temperature reduction (above 500 °C) in which CO formation occurs. These temperature ranges are comparable with the CO-TPR profiles of the catalysts (Fig. 6): the first reduction peak (CO2 formation) appears between 200 and 400 °C followed by another reduction peak at higher temperature (Fig. 6a). On the other hand, the CO production occurred only above 600 °C, as shown in Fig. 6b (especially for the CeO2 sample).

Fig. 6
figure 6

Soot-TPR profiles: CO2 (a) and CO concentrations (b) as a function of temperature

Full size image

The reducibility of the catalysts by soot in the 200–400 °C temperature range increases as follows:

$${\text{Ce}}{{\text{O}}_2} \approx {\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}}<{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}$$

On the other hand, above 400 °C a reverse trend can be observed

$${\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}<{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}}<{\text{Ce}}{{\text{O}}_2}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}}$$

thus suggesting that the Ce0.95Mn0.05 is the best soot oxidation catalyst under anaerobic conditions.

3.4 Catalytic Activity

3.4.1 CO Oxidation Reaction

The catalytic behavior towards the CO oxidation (%) was studied in the 50–600 °C temperature range. Figure 7 shows the CO conversion as a function of temperature over the prepared catalysts, along with the uncatalysed CO oxidation (dash dot curve). Likewise, Table 4 summarises the catalytic performances, in terms of T conv(%) (°C) and specific reaction rates (CO oxidation under aerobic conditions).

Fig. 7
figure 7

CO conversion (%) as a function of temperature over the prepared catalysts

Full size image

The former parameter (T10%, T50% and T90%) is defined as the temperature at which the reaction reaches 10, 50 and 90% of CO conversion. The latter (rCO) has been calculated at 140 °C, in agreement with the rCO values derived from the CO-TPR analysis (anaerobic CO oxidation). As a whole, the CO oxidation activity (i.e. in terms of light-off temperature) correlates with the CO-TPR profiles (vide Fig. 5). Specifically, the Ce0.95Cu0.05 sample exhibits the most remarkable catalytic activity (rCO = 6.76 mmolCO m−2 h−1) and a CO conversion of 50% already occurred at 162 °C. The Ce0.95Mn0.025Cu0.025 catalyst has a slightly lower catalytic activity (rCO = 4.59 mmolCO m−2 h−1), whereas the Ce0.95Mn0.05 is evidently less active than both the Cu-containing catalysts (rCO = 1.66 mmolCO m−2 h−1). Finally, the least performing catalyst is the CeO2 (rCO = 1.30 mmolCO m−2 h−1). Therefore, the specific reaction rates follow the increasing order:

$${\text{Ce}}{{\text{O}}_2}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}<{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}}$$

The rates for the aerobic CO oxidation reaction over doped-ceria catalysts are more discriminable from each other than those derived from the CO-TPR analysis. It seems, indeed, that in the absence of gas-phase oxygen, the oxidation activity depends on the lattice oxygens, whereas for the CO oxidation reaction under aerobic condition both the surface and lattice oxygens play a role. It appears that the surface defects evaluated by the Raman spectroscopy (vide supra) promotes the CO oxidation activity under either anaerobic and aerobic conditions. In fact, a positive correlation can be observed between the Ov/F2g and the rCO values (Tables 3, 4, respectively). This correlation is higher for the CO oxidation reaction under anaerobic conditions (CO-TPR analysis). In order to evaluate the stability of the prepared samples, several test–retest cycles have been performed (data not reported for the sake of brevity). As shown in Fig. 8, the most performing catalyst (Ce0.95Cu0.05) provided a good stability for two successive runs up to 700 °C. Catalytic activities of the fresh and aged sample (the latter was obtained by heating the powder at 700 °C for 8 h) are comparable, thus suggesting a good stability of the catalyst.

Fig. 8
figure 8

CO conversion (%) over the sample Ce0.95Cu0.05: a comparison between the fresh and the aged sample

Full size image

3.4.2 Soot Oxidation

Figure 9 shows the profiles of soot conversion and COx evolution achieved over the catalysts during the soot oxidation reaction. Table 4 also summarises two performance parameters, T conv(%) (°C) and r soot (mmolsoot g.cat−1 h− 1). The latter is defined as the rate of disappearance of soot normalised with respect to the catalyst weight at a certain temperature (320 °C was selected as the reference).

Fig. 9
figure 9

Soot conversion, CO and CO2 concentration as a function of temperature during the soot oxidation tests

Full size image

From the soot conversion profiles, it can be noticed that the soot oxidation values appear higher for the Ce0.95Mn0.05 and CeO2 samples rather than for the Ce0.95Cu0.05 and Ce0.95Mn0.025Cu0.025. The well-defined nanocubic structure that constitutes CeO2 apparently still plays the key role in soot oxidation, thanks to its highly reactive low-index surfaces, even if the presence of dopants promotes the reaction. Therefore, the following conversion trend appears:

$${\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}<{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}}<{\text{Ce}}{{\text{O}}_2} \approx {\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}}$$

It is also worth noticing that this activity trend correlates with the Soot-TPR profiles above 400 °C (in terms of CO2 production). This suggests that a high catalytic activity towards the soot oxidation takes place by a good lattice-to-surface oxygen mobility. On the other hand, a different trend is observed for the specific rates calculated at 320 °C (data in Table 4). Indeed, the following order can be drawn:

$${\text{Ce}}{{\text{O}}_2}<{\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.05}} \approx {\text{C}}{{\text{e}}_{0.95}}{\text{M}}{{\text{n}}_{0.025}}{\text{C}}{{\text{u}}_{0.025}}<{\text{C}}{{\text{e}}_{0.95}}{\text{C}}{{\text{u}}_{0.05}}$$

This finding is probably due to the higher surface areas for both the Cu-containing samples. At low temperatures, indeed, surface oxygens may play a major role on the oxidation activity. Figure 9 also shows the profiles of CO2 and CO as a function of temperature during the soot oxidation. Then, the presence of the solid catalyst not only lowers the reaction temperature but also enhances the selectivity to CO2. Finally, Fig. 10 shows two successive catalytic cycles (up to 650 °C) for the fresh and aged Ce0.95Mn0.05 sample, the “most effective” catalyst for this reaction. The latter exhibits a very good stability in terms of soot conversion and COx selectivity.

Fig. 10
figure 10

Soot oxidation over the Ce0.95Mn0.05 sample: comparison between the fresh and aged sample

Full size image

4 Conclusions

In the present work, a set of ceria-based catalysts doped with Mn and Cu was synthesised via a hydrothermal method. The prepared catalysts were characterized by nanocubes in the case of CeO2 and by nanopolyhedra for binary (Ce0.95Mn0.05 Ce0.95Cu0.05) and ternary oxides (Ce0.95Mn0.025Cu0.025).

The CO-TPR analysis has confirmed that the low-temperature reducibility follows the order: CeO2 < Ce0.95Mn0.05 < Ce0.95Mn0.025Cu0.025 < Ce0.95Cu0.05. A similar trend has also been encountered for the surface defective sites, analysed via the Raman spectroscopy. These findings confirm the beneficial role of dopants (Mn, Cu) into the CeO2 framework, since both the structural defects and the surface reducibility improve. Then, both the properties promote the CO oxidation activity. In fact, the Ce0.95Cu0.05 was the most effective catalyst for the CO oxidation reaction.

On the other hand, the Ce0.95Mn0.05 sample exhibited the best performances (in terms of overall soot conversion) for the soot oxidation. The following order was achieved: Ce0.95Mn0.025Cu0.025 < Ce0.95Cu0.05 < CeO2 ≈ Ce0.95Mn0.05, which is in agreement with the reduction profiles obtained by the Soot-TPR above 400 °C.