Stabilizing Ni catalysts in biogas reforming via in situ carbon deposit removal by CeO₂ oxygen vacancies

Abstract

Ni-based catalysts face significant challenges of carbon deposition during biogas reforming to syngas. Herein, we propose leveraging the in situ consumption of carbon deposits generated during biogas reforming through oxygen vacancies (O_v) on the surface of CeO₂, ensuring enhanced activity and stability of the Ni₃Fe/CeO₂-rod structure catalyst. The catalyst exhibits an initial CH₄ conversion of 80.4% and a CO₂ conversion of 97.2%, with a mere 3.9% decline in CO₂ conversion after 720 min.

---

The rapid population growth and drastic industrialization have led to an exponential increase in energy demand and a significant amount of CO₂ emissions. Efficient utilization of renewable resources and carbon dioxide has thus become a crucial pillar for future energy systems. Biogas, produced from various biomass-based feedstocks such as crop residue, household waste, and lumber, with typical compositions of 55–70% CH₄ and 30–45% CO₂, is a promising renewable energy source.¹ The high proportion of CO₂ in biogas makes it an ideal feedstock for dry reforming of methane (DRM). DRM can transform two major greenhouse gases into syngas (H₂/CO), a valuable chemical feedstock used for producing a wide range of chemicals and fuels.² This process contributes to the reduction of greenhouse gas emissions and facilitates the achievement of a carbon-neutral energy cycle. Despite the environmental and economic advantages of DRM in biogas, a significant challenge in its industrial application is the development of catalysts with high catalytic activity and resistance to carbon deposition.³

Ni-based catalysts are the most commonly investigated for biogas reforming due to their high activity and low cost.⁴ However, the issues of carbon deposition and sintering significantly hinder their activity and cycling stability.⁵ A common strategy to address these issues is to introduce a second metal to modulate the geometric and electronic structure of Ni-based catalysts. Among various candidates, Fe is often selected as a promoter due to its higher oxygen affinity compared to Ni, allowing it to convert carbon deposition into CO through the formation of FeO_x.^6,7 Fe also aids in reducing the size of Ni nanoparticles and suppressing sintering. However, the addition of a second metal to enhance catalyst performance often comes at the cost of reduced activity.

To overcome this trade-off, optimizing the interactions between the loaded metal and the support (metal–support interactions, MSI) and enhancing the vacancy sites on the support surface can improve catalyst performance without sacrificing activity. In this context, CeO₂ is a promising support candidate. As a reducible metal oxide, CeO₂ can modulate interactions with the loaded metal and the O_v by fine-tuning the surface concentrations of Ce⁴⁺ and Ce³⁺.⁸ Moreover, lattice distortion induced by the incorporation of lower-valence metal ions (such as Ni²⁺) into the host ceria matrix can generate additional O_v. These surface O_v on CeO₂ serve as optimal sites for CO₂ activation, accelerating the DRM process.⁹ However, the O_v structure can also be covered by carbon deposition, resulting in a decline in the activity of the catalysts. Therefore, ensuring the stability of the O_v structure and utilizing it to mitigate carbon deposition in biogas reforming reactions is an effective strategy, yet it has not been widely studied.

Herein, we proposed and elucidated the in situ mechanism of O_v consumption leading to carbon deposition in Ni–Fe/CeO₂ by designing and preparing catalysts with varying concentrations of O_v. The Ni₃Fe/CeO₂-rod structure (regarded as Ni₃Fe/CeO₂-RS) catalyst, prepared via the precipitation hydrothermal method, exhibited the highest concentration of O_v, resulting in an initial CH₄ conversion of 80.4% and an initial CO₂ conversion of 97.2%, with only a 3.9% decline in CO₂ conversion over 720 min. TEM, XRD, XPS, Raman, and TGA reveal that a higher concentration of O_v not only enhances the DRM activity but also engages in situ consumption with carbon deposits from the decomposition of CH_x species, thereby suppressing carbon formation. This work offers a viable strategy for the in situ consumption of carbon deposits via O_v for designing biogas reforming catalysts with enhanced activity and stability.

Three catalysts, Ni₃Fe/CeO₂, Ni₃Fe/CeO₂-nano squares (regarded as Ni₃Fe/CeO₂-NS), and Ni₃Fe/CeO₂-RS, were prepared via impregnation, impregnation and hydrothermal, and hydrothermal methods (details in Section Preparation of catalysts of the ESI†). The theoretical loadings of Ni and Fe in the catalysts were 7.5 wt% and 2.5 wt%, respectively, and the actual metal loadings were measured by inductively coupled plasma optical emission spectrometer (ICP-OES) (Table S1, ESI†). The lattice spacings of 0.20 nm and 0.32 nm observed in Fig. 1a and b and Fig. S1 (ESI†) correspond to the NiFe (111) and CeO₂ (111) facets, respectively. EDS mapping images (Fig. 1c and Fig. S2, S3, ESI†) show minimal aggregation of Ni or Fe elements across all examined areas, suggesting a homogeneous distribution of Ni and Fe obtained. In the X-ray diffraction (XRD) spectra, in addition to the diffraction peaks of the CeO₂ (pdf # 04-0593), the NiFe alloy structure was observed, and the diffraction peaks at 44.3° and 51.5° are associated with the (111) and (200) crystallographic planes of the NiFe alloy, respectively, as indicated by PDF no. 38-0419,¹⁰ in agreement with the transmission electron microscopy (TEM) results (Fig. 1b). The Brunauer–Emmett–Teller (BET) results indicate that Ni₃Fe/CeO₂-RS possesses the largest mesoporous structure (average pore diameter: 42.11 nm) and pore volume (0.227 cm³ g⁻¹), which may facilitate the mass transfer diffusion of substrates within the catalyst (see Fig. S4 and Table S1, ESI†).

Fig. 1 Characterizations of the fresh catalysts. (a)–(c) TEM images of Ni₃Fe/CeO₂-RS, (d) XRD patterns within the 2θ range of 20–80°, and (e) Ce 3d spectra.

The surface chemical states of Ce, Ni, and Fe species on CeO₂ were investigated using X-ray photoelectron spectroscopy (XPS). In the Ce 3d spectra (Fig. 1e), the u and v peaks represent spin–orbit cleavage of Ce 3d^5/2 and Ce 3d^3/2, respectively. The spectra of Ce 3d at 881.8, 888.4, and 897.9 eV (labelled v, v′′, and v′′′, respectively) and at 900.3, 907.3, and 916.3 eV (labelled u, u′′, and u′′′, respectively) show six strong peaks attributed to the Ce⁴⁺ species. Four peaks at 878.8 eV, 884.8 eV, 897.5 eV, and 903.5 eV (labelled v⁰, v′, u⁰, u′, respectively) are attributed to Ce³⁺ species.^11,12 O_v on the CeO₂ surface are generated by Ce³⁺ species, and the concentration of O_v can be determined using the ratio Ce³⁺/(Ce³⁺ + Ce⁴⁺).¹³ Compared to Ni₃Fe/CeO₂ and Ni₃Fe/CeO₂-NS, Ni₃Fe/CeO₂-RS exhibits a higher Ce³⁺ content (33.1%), implying the most abundant O_v. In the O 1s spectra (Fig. S5, ESI†), the peaks at 529.1 eV (α), 530.5 eV (β), and 532.3 eV (γ) are attributed to lattice oxygen, oxygen species linked to O_v, and chemisorbed oxygen species, respectively.¹⁴ The β peak is commonly identified as O_v, and the concentration of O_v is often estimated by the relative intensity of the β peak.¹⁵ The results indicate that Ni₃Fe/CeO₂-RS exhibits the highest concentration of O_v (23.2%), consistent with the data from the Ce 3d spectra (Fig. 1e). The Ni 2p^3/2 spectra of both samples can be deconvoluted into three main peaks (Fig. S6, ESI†) at approximately 852.4, 855.4, and 860.8 eV in Ni₃Fe/CeO₂-RS, which can be indexed to the presence of Ni⁰, Ni²⁺ and Ni sat.¹⁶ The Fe 2p^3/2 spectra of the three catalysts are composed of Fe²⁺ and Fe³⁺ species, with binding energies around 716.7 eV and 732.2 eV, respectively (Fig. S7, ESI†).¹⁷ The XPS results demonstrate that the electronic structures of Ni and Fe across the three catalysts are comparable. However, the notable differences in the proportion of Ce³⁺ in CeO₂ and the intensity of the β peak suggest that the concentration of O_v generated by Ce³⁺ is a critical factor influencing the catalytic performance.

To elucidate the impact of O_v concentration on the catalyst stability in biogas reforming, the stability of the three catalysts was evaluated under conditions of 800 °C, atmospheric pressure, with a gas mixture of CH₄ : CO₂ : N₂ = 3 : 2 : 5 and a maximum gas hourly space velocity (GHSV) of 60 000 mL g⁻¹ h⁻¹ (details for performance assessment are in Section Catalytic evaluations of the ESI†).

As illustrated in Fig. 2a, the Ni₃Fe/CeO₂-RS catalyst exhibited the highest catalytic stability, with CH₄ and CO₂ conversion decreasing from 80.4% to 67.7% and from 97.2% to 96.4%, respectively, over the initial 100 min. The Ni₃Fe/CeO₂-NS catalyst demonstrated slightly reduced overall stability, with CH₄ and CO₂ conversion declining from 87.4% and 96.7% to 65.8% and 95.9%, respectively. In contrast, the Ni₃Fe/CeO₂ catalyst showed the poorest performance, with significant decreases in CH₄ and CO₂ conversion from 67.7% to 56.5% and from 92.6% to 87.0% within the first 100 min, further dropping to 54.1% and 84.8% after 720 min. Fig. 2b reveals that the H₂/CO ratio exhibited a similar trend to the conversion, decreasing sharply in the initial 100 min before stabilizing at values of 0.90, 0.95, and 0.95. The results underscore the exceptional stability of Ni₃Fe/CeO₂-RS, likely attributable to its high concentration of O_v, which enhances the catalytic stability for C–H activation.

Fig. 2 Biogas reforming performance of different catalysts at T = 800 °C for 720 min. (a) CH₄ and CO₂ conversion, and (b) H₂/CO ratio and C balance. Reaction conditions: atmospheric pressure, CH₄/CO₂/N₂ = 3/2/5, m_cat. = 0.1 g, GHSV = 60 000 mL g⁻¹ h⁻¹.

The activity decline curves of these catalysts are consistent with the downward trends observed in other bimetallic catalysts reported in the literature.^18–23 To further investigate the correlation between O_v concentration and stability, we adjusted the O_v concentration of the catalyst and assessed its stability, deriving fitted curves that illustrate a linear correlation between the decline in CH₄ conversion and O_v concentration, whereas the decline in CO₂ conversion demonstrates a logarithmic correlation with O_v (Table S2 and Fig. S8, ESI†). These correlations imply that a higher O_v concentration leads to enhanced stability. To better comprehend the structural changes of the catalysts during biogas reforming and their impact on stability, we conducted a comprehensive series of characterizations and analyses on the spent samples.

The structures of the spent Ni₃Fe/CeO₂, Ni₃Fe/CeO₂-NS, and Ni₃Fe/CeO₂-RS were analyzed using XRD and TEM techniques. Fig. 3a and b and Fig. S9–S12 (ESI†) present the TEM images of these samples, which serve to visually analyze the structural alterations within the samples. Clearly, a significant quantity of carbon filaments formed on Ni₃Fe/CeO₂-RS (Fig. 3a), compromising the sample's original structure. Meanwhile, Fig. 3b shows NiFe particles situated atop or within these filaments, indicating a clustering of metal atoms. Fig. 3c displays the XRD patterns of the spent catalysts. While the Ni₃Fe/CeO₂ sample retains peaks corresponding to the NiFe alloy, the intensity of these peaks in Ni₃Fe/CeO₂-NS and Ni₃Fe/CeO₂-RS decreases, due to carbon deposition on the catalyst surface. Additionally, the XRD patterns show a notable small diffraction peak at 2θ of 25.8°, indicative of crystalline carbon species (002), in stark contrast to the fresh samples. Fig. 3d illustrates the XPS Ce 3d spectra of the spent samples. It is noteworthy that the Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratio has reduced compared to the fresh catalysts. This indicates a progressive oxidation of Ce³⁺ to Ce⁴⁺ during the reaction, resulting in a reduction of O_v on the catalyst surface. Among the samples, Ni₃Fe/CeO₂-RS exhibits the most significant reduction in O_v throughout the process, likely due to the depletion of O_v in combination with surface carbon species, which aids in carbon removal and stabilizes the catalyst. BET analysis showed that the pore sizes of the used Ni₃Fe/CeO₂, Ni₃Fe/CeO₂-NS, and Ni₃Fe/CeO₂-RS decreased by 5.75, 4.10, and 2.16 nm, respectively. The slightest reduction of pore sizes in Ni₃Fe/CeO₂-RS implies that its larger pores have mitigated the impact of carbon deposition, thereby improving its stability (shown in Fig. S13, ESI†). Concurrently, carbon compounds adhered to the sample's surface, resulting in significantly reduced Ni 2p and Fe 2p signals from the spent catalysts compared to the fresh catalysts (Fig. S14, ESI†).

Fig. 3 Characterizations of the spent samples. (a–c) TEM images of spent Ni₃Fe/CeO₂-RS, (d) XRD patterns with in 2θ range of 20–80°, and (e) Ce 3d spectra.

To further investigate the overall morphology, structure, and quantity of carbon species deposited on the catalysts, Raman spectroscopy and thermogravimetry analysis (TGA) measurements were performed, with the results presented in Fig. 4. The Raman spectra of the spent catalysts (Fig. 4a) display two distinct peaks at 1348 and 1582 cm⁻¹, corresponding to the D band (graphite imperfections) and G band (graphite layers) of carbon deposition, respectively.²² The intensity ratio of the D to G bands (I_D/I_G) is commonly used to evaluate the crystallization properties of carbon materials.²⁴ A higher I_D/I_G ratio signifies a lower degree of crystalline order within carbonaceous materials.¹⁰ Following the reactions of Ni₃Fe/CeO₂, Ni₃Fe/CeO₂-NS, and Ni₃Fe/CeO₂-RS, the I_D/I_G ratios recorded were 0.86, 0.41, and 0.98, indicating that Ni₃Fe/CeO₂-RS has the lowest degree of graphitized carbon, which is favorable for the removal of carbon deposition.

Fig. 4 (a) Raman spectra, and (b) TGA profiles of the spent catalysts after biogas reforming.

The quantity of carbon on the catalyst surface after the reactions was assessed using TGA analysis, as depicted in Fig. 4b. The carbon deposition on Ni₃Fe/CeO₂ is 22.6%, the lowest among the samples mentioned. Consequently, the carbon deposition rates for Ni₃Fe/CeO₂, Ni₃Fe/CeO₂-NS, and Ni₃Fe/CeO₂-RS are 1.29, 2.51, and 2.53 mg_c g_cat⁻¹ h⁻¹, respectively. The weight loss of the samples predominantly occurred between 520 °C and 720 °C, aligning with the properties of filamentary carbon.²⁵ Unexpectedly, Ni₃Fe/CeO₂-RS exhibited the highest carbon deposition rate, yet it demonstrated the most impressive biogas reforming activity and stability. Integrating the insights from XPS and TGA analyses, we can tentatively conclude that O_v engages in in situ consumption with carbon species generated from C–H cleavage, facilitating the prompt removal of carbon deposits from the catalyst surface. This elucidates why Ni₃Fe/CeO₂-RS showcases exceptional performance in stability assessments.

To further validate the mechanism by which O_v can in situ consume deposited carbon and to assess the recyclability of the catalyst in biogas reforming (see Section Catalytic evaluations of ESI†). Fig. 5a demonstrates that Ni₃Fe/CeO₂-RS maintained nearly consistent biogas reforming across three testing cycles, achieving 63.1% for CH₄ and 91.3% for CO₂ conversion by the end of the testing. In contrast, Ni₃Fe/CeO₂-NS exhibited a divergent trend, with conversion decline as the number of cycles increased. The H₂/CO ratios displayed a similar trend to the conversion, with final ratios of 0.91 for Ni₃Fe/CeO₂-NS and 0.99 for Ni₃Fe/CeO₂-RS (Fig. 5b). During the biogas reforming process, carbon formation arises from the dissociation of CH₄, due to the cleavage of C–H bonds to form CH_x* species.²⁶ These intermediates then react with O_v on the catalyst surface, resulting in a reduced proportion of Ce³⁺ as shown in Fig. 3d. When CO₂ is introduced, it participates in the Boudouard reaction with the deposited carbon, facilitating the conversion of carbon to CO while simultaneously regenerating O_v, which aids in the restoration of catalyst performance (as shown in Fig. S15, ESI†). The conclusion is corroborated by the XPS characterization of the catalyst following the first cycle (Fig. S16, ESI†), which shows that the O_v concentration in Ni₃Fe/CeO₂-RS stays at the original level, while the O_v concentration in Ni₃Fe/CeO₂-NS diminishes. As shown in Fig. 4, Raman spectroscopy characterization has previously indicated that the carbon deposited on Ni₃Fe/CeO₂-RS is attributed to lowly stable graphitic carbon, facilitating its participation in the Boudouard reaction with CO₂. This ultimately results in an excellent stability of the cycling performance of Ni₃Fe/CeO₂-RS (Fig. 4a).

Fig. 5 Cycling performance of catalysts for the biogas reforming reaction at T = 800 °C for 3 cycles. (a) CH₄ and CO₂ conversion, and (b) H₂/CO ratio and C balance. Reaction conditions: atmospheric pressure, CH₄/CO₂/N₂ = 3/2/5, m_cat. = 0.1 g, GHSV = 60 000 mL g⁻¹ h⁻¹. Reaction 180 min, CO₂ reaction 60 min.

In summary, an innovative in situ consumption of carbon deposition strategy through O_v is proposed in this study to significantly enhance the activity and stability of Ni–Fe alloy supported CeO₂ catalysts in biogas reforming. Ni₃Fe/CeO₂-RS demonstrated the highest concentration of O_v, achieving an initial conversion of 80.4% for CH₄ and 97.2% for CO₂ at 800 °C, with a mere 3.9% decline in CO₂ conversion over 720 min. Comprehensive characterizations reveal that the surface O_v in Ni₃Fe/CeO₂-RS can engage in in situ consumption reactions with carbon deposits to eliminate them, while the consumed O_v can be effectively regenerated by CO₂, further enhancing the catalyst's cycling stability. This study opens new perspectives for the utilization of O_v in biogas reforming and lays the foundation for the development of Ni-based catalysts for the efficient conversion of biogas into syngas.

This work was supported by the National Natural Science Foundation of China (No. 52276214) and the National Key R&D Program of China (2022YFE0117300).

Data availability

The data supporting this article have been included as part of the ESI.†

Conflicts of interest

There are no conflicts to declare.

Notes and references

1. Z. Wu, W. Qiao, Y. Y. Li, J. Yao, Y. Sun and R. Dong, Chemosphere, 2023, 335, 139075.
2. P. Kiani, M. Meshksar and M. R. Rahimpour, Int. J. Hydrogen Energy, 2023, 48, 6262–6274.
3. A. Braga, M. Armengol-Profitós and L. Pascua-Solé, ACS Appl. Nano Mater., 2023, 6, 7173–7185.
4. E. Enkhtuvshin, S. Yeo and H. Choi, Adv Funct Mater., 2023, 33, 2214069.
5. M. Kaviani, M. Rezaei, S. M. Alavi and E. Akbari, Fuel, 2024, 355, 129533.
6. X. Yao, Q. Cheng and Y. Attada, Appl. Catal., B, 2023, 328, 122479.
7. D. Ayillath Kutteri, I. W. Wang, A. Samanta, L. Li and J. Hu, Catal.: Sci. Technol., 2018, 8, 858–869.
8. F. Carraro, A. Fapohunda and M. C. Paganini, ACS Appl. Nano Mater., 2018, 1, 1492–1501.
9. Y. Long, X. Wang, H. Zhang and K. Wang, JACS Au, 2024, 4(7), 2462–2473.
10. Y. Li, Q. Wang and M. Cao, Appl. Catal., B, 2023, 331, 122669.
11. F. Jiang, S. Wang, B. Liu, J. Liu, L. Wang, Y. Xiao, Y. Xu and X. Liu, ACS Catal., 2020, 10, 11493–11509.
12. J. Li, Z. Liu, D. A. Cullen, W. Hu, J. Huang, L. Yao, Z. Peng, P. Liao and R. Wang, ACS Catal., 2019, 9, 11088–11103.
13. G. I. Siakavelas, N. D. Charisiou, S. AlKhoori and A. A. AlKhoori, Appl. Catal., B, 2021, 282, 119562.
14. N. Wang, K. Shen, L. Huang, X. Yu, W. Qian and W. Chu, ACS Catal., 2013, 3, 1638–1651.
15. H. Liu, Z. Fan and C. Sun, Appl. Catal., B, 2019, 244, 671–683.
16. G. S. Dhillon, G. Cao and N. Yi, Catalysts, 2023, 13, 1171.
17. S. A. Theofanidis, V. V. Galvita, M. Sabbe, H. Poelman, C. Detavernier and G. B. Marin, Appl. Catal., B, 2017, 209, 405–416.
18. K. Hong, J. Min, G. Min, Y. Bae and J. Hong, Chem. Eng. J., 2023, 462, 142334.
19. S. M. Kim, P. M. Abdala and T. Margossian, J. Am. Chem. Soc., 2017, 139, 1937–1949.
20. Z. Xie, B. Yan and S. Kattel, Appl. Catal., B, 2018, 236, 280–293.
21. Y. Tang, Y. Wei and Z. Wang, J. Am. Chem. Soc., 2019, 141, 7283–7293.
22. M. Yusuf, A. S. Farooqi and A. A. Al-Kahtani, Int. J. Hydrogen Energy, 2021, 46, 27044–27061.
23. S. Shah, S. Sayono and J. Ynzunza, AIChE J., 2020, 66, e17078.
24. Z. Y. Lim, J. Tu, F. Zhou, B. Chen and K. L. Choy, Appl. Catal. B, 2024, 350, 123891.
25. S. A. Theofanidis, J. A. Z. Pieterse and H. Poelman, Appl. Catal., B, 2020, 267, 118691.
26. H. Wang, G. Cui and H. Lu, Nat. Commun., 2024, 15, 3765.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4cc04453c

---