Jintao Miao,
Nishan Paudyal
,
Rosa V. Melinda
and
Jing Zhou
*
Department of Chemistry, University of Wyoming, Laramie, WY 82071, USA. E-mail: jzhou2@uwyo.edu
First published on 15th September 2025
In this study, 5 wt% Ni catalysts over a series of Ce1−xTixO2−δ supports with controlled Ti dopant composition (x = 0–0.5) were synthesized by sol–gel and impregnation methods. Compositions, crystal structures, and surface properties were investigated to confirm the formation of Ce1−xTixO2−δ mixed oxides with low Ti compositions (e.g., Ce0.9Ti0.1O2−δ). Ce4+ in Ce–O–Ti shows a lower reduction temperature compared to bulk Ce4+ in CeO2−δ, and thus Ti-doped ceria exhibits better reducibility. TiO2 is also formed over Ce1−xTixO2−δ with high Ti compositions (e.g., Ce0.5Ti0.5O2−δ), suggesting the limited solubility of Ti in the ceria lattice. The amount of Ti in Ce1−xTixO2−δ plays a role in the formation of Ni species. NiO was found to be the major species over CeO2−δ and Ce0.9Ti0.1O2−δ. However, NiTiO3 was observed over Ce1−xTixO2−δ (x ≥ 0.2). Compared to Ni/CeO2−δ, Ni/Ce0.9Ti0.1O2−δ delivers better CH4 and CO2 conversions in DRM. This can be attributed to the enhanced reducibility of Ce0.9Ti0.1O2−δ and the stronger metal–support interaction by a small amount of Ti doping in ceria. The DRM activity of Ni decreases with the increased Ti composition in Ni/Ce0.5Ti0.5O2−δ. This can be correlated with the formation of NiTiO3, which produces significantly less metallic Ni as the active species for DRM compared to NiO that is formed over CeO2−δ and Ce0.9Ti0.1O2−δ. The TGA results of spent catalysts indicate a decrease in carbon deposition during DRM with increasing Ti composition in Ni/Ce1−xTixO2−δ. XRD data suggest the formation of a new Ce2Ti2O7 phase in spent Ni/Ce0.5Ti0.5O2−δ, which could better help remove carbon deposits. Doping Ti into the ceria lattice significantly helps mitigate the issue related to carbon deposition over the Ni catalyst during DRM. Similar behavior was also observed over Ce1−xTixO2−δ-supported Co catalysts. Our study clearly demonstrates that doping Ti in ceria can tune both the activity and stability of supported metal catalysts in DRM.
Doping of ceria with other metal cations can enhance the thermal stability as well as the redox properties of ceria and thus better improve the activity and stability of Ni in DRM.22,23 Various metal elements have been selected as dopants to prepare doped ceria, including Zr, La, Ti, and Mg.24–28 Ti was found to be a good candidate as a dopant based on both computational and experimental work.29,30 Compared to pure ceria, Ti-doped ceria has a lower formation energy of oxygen vacancies, which enhances the oxygen mobility in the ceria lattice.31,32 The Ce/Ti ratio was found to be an important parameter in tuning the redox properties of ceria. Efstathiou's group prepared Ce0.8Ti0.2O2−δ and Ce0.5Ti0.5O2−δ supports for Ni and Pt catalysts and examined their catalytic performance for DRM and water–gas shift (WGS) reactions.33–35 They reported a better carbon resistance behavior over Ni/Ce0.8Ti0.2O2−δ than over Ni/CeO2−δ and Ni/Ce0.5Ti0.5O2−δ, which is attributed to the effect of the particle size and supports. Since the effect of the particle size and support are usually coupled, in an attempt to decouple these two factors, Han and co-workers prepared Ni-based catalysts over various supports with controlled particle sizes. They suggested that a small particle size of Ni particles and their interaction with a basic metal oxide support are beneficial to the performance in DRM.36 In our previous study, we prepared well-ordered CeO2(111) thin films as well as Ce1−xTixO2−δ(111) thin films over Ru(0001) under ultra-high vacuum (UHV) conditions. We found that Ti-doped ceria can better anchor Ni as smaller particles and help inhibit its sintering with heating to 800 K than pure ceria.31 Sintering of Ni at high reaction temperatures also causes catalyst deactivation.16,36 Smaller metal particles not only provide more active sites for the reaction, but also minimize carbon formation.36,37 To extend the exciting observation from the model system of Ni/Ce1−xTixO2−δ(111) thin films, we synthesized powder materials of Ce1−xTixO2−δ with controlled Ti compositions by sol–gel methods and dispersed 5 wt% Ni over as-synthesized supports by impregnation methods. The composition of Ti was controlled between x = 0 and x = 0.5. Compared to previous studies, a smaller increment of the Ti composition (i.e., 0.1) was considered with a motivation to better tune the structure and redox properties of ceria and thus examine potential improvement of supported Ni catalysts towards the DRM reaction and carbon resistance.33,35 In our study, conventional and synchrotron X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), inductively coupled plasma-optical emission spectrometry (ICP-OES), Brunauer–Emmett–Teller (BET), H2-temperature-programmed reduction (H2-TPR), and H2 chemisorption were used to examine the composition, the crystal structure, the reducibility, and the surface properties of prepared Ni catalysts with respect to the Ti composition in ceria. XRD and thermogravimetric analysis (TGA) were used to examine the structure and the extent of carbon deposition over spent catalysts. This systematic study using combined spectroscopy and microscopy techniques allowed for the elucidation of the role of Ti doping in ceria in the activity, stability, and carbon resistance of supported Ni in DRM. It was found that, by doping Ti into the ceria lattice, significant enhancement in the reducibility of the supports as well as the metal–support interaction was observed. Furthermore, the amount of Ti dopant plays a role in the nature of Ni species including NiO and NiTiO3 formed over Ce1−xTixO2−δ, which exhibits a strong correlation with their physical properties and catalytic performance in DRM. Such behavior was also observed for Co catalysts over Ti-doped ceria.
![]() | (1) |
![]() | (2) |
![]() | (3) |
![]() | (4) |
Ce1−xTixO2−δ | Ti composition, x, by ICP-OES | Lattice constant (Å) | Crystallite size (Å) | BET surface area (m2 g−1) | Metal dispersion of supported Ni catalysts (%) |
---|---|---|---|---|---|
CeO2−δ | 0 | 5.413 | 298 | 13.7 | 1.2 |
Ce0.9Ti0.1O2−δ | 0.10 | 5.403 | 142 | 15.2 | 1.0 |
Ce0.8Ti0.2O2−δ | 0.22 | 5.403 | 127 | 17.0 | 0.3 |
Ce0.7Ti0.3O2−δ | 0.32 | 5.394 | 125 | 12.6 | 0.3 |
Ce0.6Ti0.4O2−δ | 0.42 | 5.392 | 119 | 8.8 | 0.4 |
Ce0.5Ti0.5O2−δ | 0.51 | 5.394 | 121 | 8.8 | 0.3 |
The reducibility of the same series of Ce1−xTixO2−δ supports was examined by performing H2-TPR experiments in a temperature range between 50 and 850 °C (Fig. 1c). The H2-TPR profile of a standard TiO2 powder sample (Sigma-Aldrich, ≥99.5%) is also shown, which exhibits little signal in this temperature range. To deconvolute the reduction features originating from different species formed over each sample, the peak fitting of the H2-TPR results was conducted using Voigt functions in the Fityk software and the detailed fitting information can be found in Fig. S2a and Table S1 in the SI. Reduction in two temperature ranges with peaks at ∼481 and 791 °C was observed for CeO2−δ, which can be attributed to the reduction of surface and bulk Ce4+, respectively.34,52,53 The surface area and crystallite size of CeO2 significantly affect the peak intensity and reduction temperature of both surface and bulk Ce4+. Rao reported the synthesis of CeO2 samples with the surface area values ranging from 1.5 to 130 m2 g−1 and the characterization of the reducibility using H2-TPR.53 For the sample with a surface area value of 1.5 m2 g−1, ignorable reduction of surface Ce4+ species was detected and only reduction of bulk Ce4+ was observed at 900 °C. When the surface area value was increased to 130 m2 g−1, surface reduction features between 200 and 600 °C were clearly observed and the bulk reduction temperature decreased to 800 °C. In our study, for the ceria support with a 10% Ti dopant amount, the reduction temperature of bulk Ce4+ in ceria decreases to ∼763 °C. This is consistent with the decrease in its crystallite size and surface area as well as the formation of reduced ceria as a result of Ti doping that exhibits enhanced reducibility.34 Additional reduction signals between 400 and 650 °C over Ce1−xTixO2−δ can be associated with Ce4+ in Ce–O–Ti that has a lower reduction temperature compared to bulk Ce4+ in CeO2−δ.54,55 Doping additional metal cations (e.g., Zr, Nb, and Ti) into the ceria lattice can modify the metal–oxygen bond length, which results in the formation of labile oxygen near Ce cations.34,56,57 Thus, doped ceria usually exhibits a lower reduction temperature and higher oxygen storage capacity compared to pure ceria.58 The intensity associated with the reduction of Ce4+ in Ce–O–Ti increases when the Ti composition (x) increases to 0.3. However, it shows no significant change with further increase in Ti composition to 0.5. A sharp reduction peak at around 800 °C was observed in Ce1−xTixO2−δ (x ≥ 0.3) supports and the intensity of this peak increases while further increasing Ti composition. This peak could suggest a change in the bulk structure of ceria. Smal and co-workers observed a decrease in the reduction temperature of ceria supports modified with Ti dopant and detected a sharp reduction peak at around 792 °C with high Ti concentrations in ceria. They ran H2-TPR on 5 wt% Ni over Ce0.55Ti0.45O2 and Ce0.65Ti0.35O2 and examined their crystal structures during the H2-TPR experiment as well as after cooling down to room temperature. The pyrochlore structure of Ce2Ti2O7 was detected after cooling down to room temperature.59 In our study, after reduction, Ce2Ti2O7 was also observed in Ni/Ce0.5Ti0.5O2−δ, while the fluorite structure of ceria was maintained for both Ni/CeO2−δ and Ni/Ce0.9Ti0.1O2−δ as shown in Fig. S3. It is known that ceria with a pyrochlore structure can have high oxygen mobility due to its disordered structure, which could assist in oxidizing deposited carbon during the DRM reaction.60
Pure ceria (CeO2−δ) and Ti-doped ceria with a low Ti composition of x = 0.1 and a high value of x = 0.5 (Ce0.9Ti0.1O2−δ and Ce0.5Ti0.5O2−δ) were selected as representative samples to further examine the effect of Ti doping using in situ synchrotron XRD. The XRD data (Fig. 2a–c) were collected with 2-theta values between 0.500° and 16.500° while the sample was heated from 50 °C up to 750 °C with a rate of 30 °C min−1 under a mixture gas flow of 5 mL min−1 H2 and 5 mL min−1 He. For comparison, the XRD data of CeO2−δ with heating under an inert environment of a 10 mL min−1 He flow were also collected. All three samples exhibit XRD patterns consistent with the fluorite structure of CeO2 (PDF #34-0394). In agreement with the results from conventional XRD experiments in Fig. 1a, the calculated lattice constant (Fig. 2d) decreases from 5.410 Å for CeO2−δ to 5.394 Å for Ce0.5Ti0.5O2−δ due to the substitution of Ti into the ceria lattice. With heating, the CeO2−δ support under an inert He flow shows a gradual increase in the lattice constant from 5.410 Å at room temperature to 5.457 Å at 800 °C, consistent with the thermal expansion of ceria.61–64 When switching the gas from He to H2He, an increase in the lattice constant of CeO2−δ with temperature due to thermal expansion was observed. Additionally, a more significant increase in the lattice constant was detected at around 675 °C. This is attributed to the reduction of Ce4+ to Ce3+ under a reducing environment as Ce3+ cations have a larger radius (1.14 Å) than Ce4+ (0.97 Å).27,42,65 Such behavior was also observed for Ce0.9Ti0.1O2−δ and Ce0.5Ti0.5O2−δ during heating under a H2He flow. However, the extensive change in the lattice constant associated with the Ce4+ reduction occurred at lower temperatures. This behavior agrees well with the above H2-TPR results as well as the XPS data shown in Fig. S1 in the SI, suggesting the enhanced reducibility of ceria with Ti doping.
The XRD patterns of 5 wt% Ni dispersed over the series of Ce1−xTixO2−δ supports (x = 0–0.5) were collected as shown in Fig. 1b. In addition to the reflections associated with ceria and titania, the peaks related to different Ni species were observed, demonstrating that the nature of the ceria supports plays a role in the formation of these Ni species. For Ni/CeO2−δ and Ni/Ce0.9Ti0.1O2−δ, the peaks at 37.2° and 43.3° were detected, corresponding to the (111) and (200) planes of NiO (PDF #47-1049). The peaks at 35.7°, 40.9°, and 54.0° that are associated with the (110), (113), and (116) planes of NiTiO3 (PDF #33-0960) were observed over Ni/Ce0.8Ti0.2O2−δ and became more pronounced with further increase of Ti composition to 0.5 in Ce1−xTixO2−δ. The formation of NiTiO3 is suggested due to the reaction of Ni with isolated domains of TiO2 during the metal dispersion and calcination process with a temperature typically higher than 550 °C.26,49,50,66–69
The H2-TPR profile of 5 wt% Ni/CeO2−δ (Fig. 1d) shows reduction peaks at 213 and 312 °C, which are attributed to the reduction of surface and bulk NiO, respectively.70,71 The detailed peak fitting results of H2-TPR can be found in Fig. S2b and Table S2. For Ni/Ce0.9Ti0.1O2−δ, reduction peaks of NiO shift to 330 and 453 °C, suggesting that Ti doping can enhance the metal–support interaction. For Ni/Ce1−xTixO2−δ with higher Ti compositions (x = 0.3–0.5), the major reduction peak was observed at around 650 °C. This temperature is higher than that for reduction of NiO over pure titania and is consistent with reduction of Ni2+ in NiTiO3 that takes place in a temperature range between 550 and 700 °C.49,50,68,69,72 The H2-TPR data are consistent with the XRD results, demonstrating that doping CexTi1−xO2−δ by Ti influences the formation of Ni species (e.g., NiO and NiTiO3). It seems like the nature of Ni species formed over the ceria support also affects the measured metal dispersion value of Ni as shown in Table 1. Despite the fact that the surface area values are not particularly high for CeO2−δ and Ce0.9Ti0.1O2−δ, the metal dispersions of Ni over these two supports in our study are comparable with previously reported data.73,74 However, the metal dispersion values of Ni decreased extensively with further increase in Ti composition to 0.3 or higher in CexTi1−xO2−δ. This is consistent with the observation that NiO is the major species over CeO2−δ and Ce0.9Ti0.1O2−δ, while NiTiO3 is formed over ceria supports with higher Ti compositions. During reduction, NiTiO3 is reduced to metallic Ni and TiO2, where Ni could be encapsulated by TiO2 that could inhibit the adsorption of active gases (e.g., H2 and CO) over Ni and thus result in a low metal dispersion measured by chemisorption.72,75 In addition to the formation of NiO over ceria, incorporation of Ni into its lattice to form a Ce1−xNixO2−δ solid solution was reported, the extent of which depends on synthesis methods.76–80 The XRD pattern of 5 wt% Ni/CeO2−δ clearly shows the diffraction peaks associated with the NiO phase (Fig. 1b), which is consistent with the intense reduction peak at 312 °C in the H2-TPR profile (Fig. 1d). In our study, there could be the possibility of incorporation of a small amount of Ni into the ceria lattice. A slight decrease in the lattice constant value of ceria from 5.413 Å for CeO2−δ to 5.409 Å for CeO2−δ with dispersed 5 wt% Ni was detected. Although this change in the lattice constant is within the range of the XRD resolution of the instrument, incorporation of Ni into the ceria lattice can cause the decrease of the lattice constant considering that Ni2+ has a smaller size (0.72 Å) than Ce4+ (0.97 Å).80 When comparing the H2-TPR profile of 5 wt% Ni/CeO2−δ to that of pure CeO2−δ (Fig. 1c and d), it shows small, enhanced signals between 350 and 650 °C. This could be due to the reduction of NiO that strongly interacts with ceria and/or the reduction of the ceria support.49,50,68,69,72,78 This could also be due to reduction of Ni in Ni–O–Ce by H2, which was observed at temperatures above 400 °C.77
To examine the effect of the Ti dopant on the activity of Ni in DRM, temperature-dependent tests were conducted for 5 wt% Ni catalysts supported over all prepared Ce1−xTixO2−δ supports. Prior to the studies, all samples were reduced at 750 °C in H2 for 1 h. A total flow rate of 30 mL min−1 (N2:
CH4
:
CO2 = 10
:
10
:
10) was used for DRM. As shown in Fig. 3a and b, the percent conversion values of CO2 and CH4 reactants for all catalysts increase with increasing reaction temperature. The results are consistent with the endothermic nature of the DRM reaction.3,6 A higher percent conversion value of CO2 was observed compared to that of CH4, which could be due to the reverse-water–gas shift (RWGS) reaction as a side reaction in DRM.6,16 The ignition temperature and activity of catalysts vary with the composition of Ti dopant in Ce1−xTixO2−δ supports. Ni over pure CeO2−δ and Ce0.9Ti0.1O2−δ exhibited a comparable DRM activity except during the temperature range between 550 and 750 °C. Compared to these two catalysts, Ni/Ce0.8Ti0.2O2−δ showed a slightly lower activity in DRM, and further increase in the Ti composition resulted in a further decrease in the activity of Ni/Ce1−xTixO2−δ. As it is known that metallic Ni is the active species in DRM, the lower activity of Ni over ceria with high Ti amounts (e.g., Ce0.5Ti0.5O2−δ) could be correlated with a significantly lower Ni metal dispersion value with a much lower metallic Ni amount present over the catalyst surface.81,82
5 wt% Ni catalysts over pure CeO2−δ as well as over doped ceria with a low Ti composition (Ce0.9Ti0.1O2−δ) and a high Ti composition (Ce0.5Ti0.5O2−δ) were selected for stability tests at 750 °C for 24 h. The results shown in Fig. 3c and d were collected with a gas mixture of N2, CH4, and CO2 with a flow rate of 10–10–10 mL min−1, corresponding to a GHSV value of 18000 mL g−1 h−1. There is around 5% difference in the conversion values reported at 750 °C in the stability tests compared to those in temperature-dependent DRM results. This is likely due to the additional stepwise heating treatments in temperature-dependent DRM tests during which all catalysts were heated from 300 to 800 °C with a temperature increment of 50 °C under the reaction stream and held at each temperature for 0.5 h during the data collection. For the stability tests, all catalysts were directly heated up to 750 °C in a temperate ramp of 20 °C min−1 under the reaction stream. As shown in Fig. 3c and d, all three catalysts showed good stability with less than 5% loss of activity after 24 h. It is known that Ni-based catalysts are prone to deactivation due to carbon deposits in DRM.15,83 Therefore, after the stability test, the TGA analysis (Fig. 3f) was carried out over spent catalysts for the investigation of carbon that was present over the catalyst surface, which was compared to that of reduced catalysts prepared prior to the DRM tests (Fig. 3g). In the temperature range between 400 and 600 °C, the weight loss in TGA data can be attributed to the oxidation of carbon to CO2.84 Our TGA results (Fig. 3f) in general suggest significantly less carbon deposits over Ni supported on Ti-doped ceria. The spent sample of Ni/CeO2−δ showed around 7% weight loss due to carbon deposition. The carbon exhibits fiber-like features in the SEM image (Fig. 3h). However, only about 2 wt% or even a negligible amount of weight loss was detected in spent samples of Ni/Ce0.9Ti0.1O2−δ and Ni/Ce0.5Ti0.5O2−δ. At around 350 °C, there is a weight increase of the catalysts, observed especially clearly for Ni/Ce0.5Ti0.5O2−δ and this is attributed to the oxidation of the reduced catalyst as shown in Fig. 3g.50 The same behavior was observed when we increased the GHSV value from 18
000 to 36
000 mL g−1 h−1 as shown in Fig. S4. The amount of carbon deposits over these samples followed the trend Ni/Ce0.5Ti0.5O2−δ ≪ Ni/Ce0.9Ti0.1O2−δ < Ni/CeO2−δ. As suggested by combined catalyst characterization and activity studies, Ni catalysts supported over Ti-doped ceria exhibited an enhanced reducibility and a stronger metal–support interaction, which could promote the removal of carbon deposits in DRM.85,86
The XRD patterns of spent samples of Ni catalysts over CeO2−δ, Ce0.9Ti0.1O2−δ, and Ce0.5Ti0.5O2−δ were obtained (Fig. 3e). Both samples of Ni/CeO2−δ and Ni/Ce0.9Ti0.1O2−δ maintained the fluorite structure of CeO2. A small intensity at 44.5° was attributed to metallic Ni, confirming that Ni0 is the active metal species during the DRM reaction. However, for the spent sample of Ni/Ce0.5Ti0.5O2−δ, the reflections due to CeO2 became much boarder, indicating a less crystalline fluorite structure. The patterns related to TiO2 were not clearly detected. Furthermore, new reflections at 21.3°, 23.2°, 27.8°, 30.1°, 32.3°, 40.3°, 43.3°, 48.4°, 52.0°, 55.2°, and 58.4° were observed which can be attributed to (210), (002), (400), (112), (410), (022), (420), (520), (114), (304) and (232) of Ce2Ti2O7 (PDF #47-0667). This indicates a reduction of Ce4+ to Ce3+ and the formation of a new crystal phase of Ce2Ti2O7 after sample reduction and the DRM reaction. Ruan and co-workers investigated the structure change between CeO2–TiO2 and Ce2Ti2O7 and observed the transformation of CeO2–TiO2 mixed oxides into Ce2Ti2O7, promoted by the dispersed Ni metal under reduction in CH4.48 Compared to pure ceria, Ce2Ti2O7 has a higher oxygen mobility due to the disorder of both cations and anions in the sublattice.87 Although Ni/Ce0.5Ti0.5O2−δ delivered the least activity in DRM among all three catalysts, it showed very little carbon deposit. Such behavior could also be correlated with the enhanced oxygen mobility and redox properties due to the formation of Ce2Ti2O7 in Ni/Ce0.5Ti0.5O2−δ.
Our study provides new insight into the role of Ti doping in the formation of Ni species over Ce1−xTixO2−δ and associated activity and stability in DRM. The effect of the Ti4+ dopant over 5 wt% Ni/Ce1−xTixO2−δ was previously studied by Damaskinos and co-workers.33 They chose the Ti stoichiometry of x = 0.2 and 0.5. The fluorite structure of CeO2 was maintained through the different concentrations of Ti4+ doping. After the doping of CeO2 with Ti with the stoichiometry of 0.5, the diffraction peaks shifted to higher 2-theta angles in XRD, indicating a decrease in the lattice constant of ceria due to incorporation of Ti4+ ions in the CeO2 lattice. Also, a broadening of peaks representing CeO2 was observed for Ti-doped samples indicating a decrease in crystallite size of the support which correlated well with our XRD data. Although not mentioned in their study, the XRD pattern of 5 wt% Ni/Ce0.5Ti0.5O2−δ seemed to suggest the formation of isolated titania domains and NiTiO3, which are similar to what have been observed in our studies. For the DRM reactivity studies, they observed an increase in the CH4 conversion and a significant decrease in carbon deposition for 5 wt% Ni/Ce0.8Ti0.2O2−δ compared to the 5 wt% Ni/CeO2 catalyst. However, the CH4 conversion was lower for the 5 wt% Ni/Ce0.5Ti0.5O2−δ. This in general agrees well with our results which showed that a small amount of Ti dopant can enhance both the catalyst activity and the resistance to carbon in DRM due to the participation of lattice labile oxygen from the support for gasification of C, leading to the formation of CO.33,35 With the control of the small increment of Ti dopant, our study demonstrated that 5 wt% Ni/Ce0.9Ti0.1O2−δ gave the highest conversion of CH4 compared to all the other catalysts, indicating that the small stoichiometry of 0.1 for Ti in Ce1−xTixO2−δ delivers a good enhancement of the catalytic performance of Ni. A major difference observed between our study and the one performed by Damaskinos et al. was in the C resistance for Ni/Ce0.5Ti0.5O2−δ. Our results (Fig. 3f) clearly show that the spent 5 wt% Ni/Ce0.5Ti0.5O2−δ catalyst had negligible C deposition compared to Ni/CeO2−δ and Ni/Ce0.9Ti0.1O2−δ. As shown in Fig. 3e and Fig. S3, 5 wt% Ni/Ce0.5Ti0.5O2−δ underwent a change in the crystal structure for CeO2 during the reduction with H2 at 750 °C and the DRM reaction, leading to the formation of Ce2Ti2O7. It is known that Ce2Ti2O7 has high O mobility that could facilitate the removal of deposited C.60 Damaskinos and co-workers concluded higher C deposition over spent Ni/Ce0.5Ti0.5O2−δ compared to 5 wt% Ni/Ce0.8Ti0.2O2−δ. They explained that this was due to the least participation of lattice labile O for removal of C in DRM as demonstrated in transient experiments using 18O2 isotope species.33 The exact nature for the difference between our results and the previous report is not clear, although there is a likelihood due to the differences in catalyst synthesis that could result in some variations in the size, structure, and composition of the prepared samples. In particular, the Ti doping level plays an important role in the nature of synthesized ceria supports, which in turn could influence the formation of Ni species (NiO and NiTiO3). Our group has confirmed the concentration of Ti for the prepared catalysts through ICP-OES studies with the stoichiometry of Ti matching the targeted values. The quantification of the amount of Ti dopant was not indicated in their study. Furthermore, it would be interesting to know if there was a composition and structure change for spent Ni/Ce0.5Ti0.5O2−δ.
To extrapolate the effect of Ti doping in ceria on the catalytic performance of other metals in DRM, our group carried out a study over Co catalysts considering that Co shows good DRM activity.64,88 Ce0.9Ti0.1O2−δ-supported Co catalysts with both 5 and 10 wt% loadings were prepared considering that among the studied series of Ce1−xTixO2−δ, Ni supported over Ce0.9Ti0.1O2−δ showed a good activity with a reasonable amount of carbon deposits during the DRM reaction. As a comparison, Co catalysts over pure ceria were also synthesized. As shown in Fig. 4a, the fluorite structure of ceria was observed in both ceria supports and all ceria-supported Co catalysts. No clear reflection peaks of Co species were observed for Co catalysts with a 5 wt% Co loading, and Co3O4 was identified with a peak located at 37.0° (PDF #42-1467) for the samples with a 10 wt% Co loading. This indicates that Co could maintain good dispersion as small particles over the support at a low weight loading (i.e., 5 wt%). With higher amounts of Co (i.e., 10 wt%), it can aggregate and form crystallite structures during calcination. Like the study of Ni as described above, a small decrease of 0.003 Å in the lattice constant of CeO2 was detected with dispersed 5 wt% Co, which increased to 0.011 Å with dispersed 10 wt% Co. This is consistent with the suggestion of incorporation of Co into the ceria lattice.89,90 The extent of formation of Co-doped ceria could be small when compared to the study by Yang and co-workers as they reported a reduction in the ceria lattice by ∼0.019 Å for 2% atom doping of Co in ceria (Ce0.98Co0.02O2−δ) and by 0.036 Å for 10% atom doping of Co in ceria (Ce0.90Co0.10O2−δ).89 Fig. 4b shows the H2-TPR profiles of both supports with and without 5 wt% Co. The reduction features between ∼200 and 450 °C with two peaks at 284 and 332 °C from 5 wt% Co/CeO2−δ were assigned to two reduction steps of Co3O4 to CoO and CoO to Co, respectively.91–93 These two peaks representing stepwise reduction of Co3O4 are also present over 5 wt% Co/Ce0.9Ti0.1O2−δ. Additionally, a small reduction peak at 535 °C was observed in the TPR profile for 5 wt% Co/Ce0.9Ti0.1O2−δ, which could be due to the reduction of Co3O4 that has a strong interaction with Ti-doped ceria.94 The temperature-dependent DRM results of Co over CeO2−δ and Ce0.9Ti0.1O2−δ are given in Fig. 4c and d. Prior to the DRM reaction, all catalysts were reduced at 750 °C under H2 for 1 h, which reduced Co3O4 to metallic Co through a CoO transition as demonstrated in previous reports.64,91,92,95,96 However, during the DRM reaction, it has been shown that the chemical state and/or structure of Co species along with the ceria support experience dynamic changes with respect to the temperature and gas species in the reaction stream.64,97 In DRM, CoO can reappear at 200 °C and metallic Co becomes predominant with the increase of the reaction temperature to 500 °C, which are active species for methane activation. This is consistent with our XRD patterns of all spent samples of 5 and 10 wt% Co catalysts over CeO2−δ and Ce0.9Ti0.1O2−δ (Fig. 4f), which show metallic Co as evident with a small intensity at 44.2° (PDF #15-0806). For a 5 wt% Co loading, a clear decrease in the ignition temperature was observed for Co/Ce0.9Ti0.1O2−δ compared to that of Co/CeO2−δ. With the increase of the Co loading to 10 wt%, the ignition temperatures for Co over both ceria supports were similar. For both 5 and 10 wt% Co loadings, Co over Ce0.9Ti0.1O2−δ showed similar activity in DRM. However, the increase in the cobalt loading to 10 wt% increased the activity of Co/CeO2−δ but at the cost of extensive formation of carbon deposits. TGA analysis indicated an increased weight loss from 5 to 16% due to carbon burn off as the weight loading of Co was increased from 5 to 10 wt% over the CeO2−δ support (Fig. 4e). On the other hand, both Ce0.9Ti0.1O2−δ-supported Co catalysts showed a weight loss of around 2%. As shown in the SEM images (Fig. 4g and h), there is little evidence of carbon deposition over the spent sample of 5 wt% Co/Ce0.9Ti0.1O2−δ. However, filamentous carbon-like features were clearly observed over 5 wt% Co/CeO2−δ. Similar to Ni, Ti doping in ceria has a promotional effect on the redox properties of ceria and the interaction of supported Co catalysts that help remove carbon deposits during the DRM reaction.
The authors confirm that the data supporting the findings of this study are available within the article and its SI.
This journal is © The Royal Society of Chemistry 2025 |