Fabrication and formation mechanism of Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts and application in low-temperature NO reduction with NH₃

Abstract

Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts were synthesized via a redox strategy, and presented 58–85% NO conversion at 80–180 °C. The 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalyst displayed the optimal activity, which may be owing to the generation of amorphous mixed metal-oxide catalysts, and higher contents of Ce³⁺ and surface oxygen (O_y). The formation mechanism of the catalysts was proposed.

---

Nitric oxides (NO and NO₂) originating from stationary sources can lead to photochemical smog, the greenhouse effect, ozone depletion, and acid rain.^1,2 Given this, some technologies for NO abatement have been studied. Among them, selective catalytic reduction of NO with NH₃ (SCR) has been used in practiced applications, and the core of the SCR reaction is a catalyst.³ However, the operating temperature window of the commercialized V₂O₅ + WO₃(MoO₃)/TiO₂ catalyst is high (300–400 °C).⁴ Additionally, it is also susceptible to deactivation by SO₂ and dust, which requires it to be located downstream of the desulfurizer and particle control device, where the flue gas temperature is generally below 200 °C.^5,6 Therefore, it is significant to develop a SCR catalyst with excellent low-temperature (<200 °C) catalytic activity for the control of NO.

Carbon nanotubes (CNTs) are of interest as catalyst carriers due to their unique tubular structure, and remarkable physical and chemical performances. Previous studies reported that nitric oxides and ammonia could be adsorbed by CNTs.⁷ Besides, NO could be decomposed and reduced by CNTs.⁸ Therefore CNTs are considered as a promising candidate for catalyst supports. A series of CNTs-based catalysts have been prepared and used in the SCR reaction, such as TiO₂@MnO_x–CeO_x/CNTs,⁹ MnO_x/CNTs,¹⁰ CeO₂/CNTs,¹¹ and so on. Nevertheless, these catalysts still have one thing in common-high working temperature window of 200–300 °C, which will limit the application to after the desulfurizer and electrostatic precipitator. Hence, development of highly active SCR catalysts below 200 °C is of importance.

Previous studies reported that Cu-based catalysts, including CuO_x-carbonaceous,¹² Cu/MCM-41,¹³ and MnO_x–Cu–CeO₂,¹⁴ etc., exhibited better SCR activities. Nonetheless, high-temperature calcination method was usually used to prepare these catalysts, which was a little complicated. Therefore, it is essential to adopt a simple method for the fabrication of SCR catalyst. Based on the above issues, a mild redox method was firstly initiated to prepare Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts, and the catalysts presented outstanding low-temperature catalytic activities in SCR reactions.

Fig. 1 shows the NO conversion of the as-prepared catalysts. Compared to Ce–Cu–MnO_x/CNTs-IM catalyst, the NO conversions over Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts were higher, and achieved 58.0–85.1% during the test temperature rang. It is noteworthy that 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalyst attained the first-rate catalytic efficiency, which reached 66–85% at 80–180 °C at a weight hourly space velocity of 280 000 mL g_cat⁻¹ h⁻¹. The BET surface area (Table S1, ESI†) data indicates that the BET surface area over 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs (45.2 m² g⁻¹) was smaller than that of over Ce–Cu–MnO_x/CNTs-IM (70.7 m² g⁻¹), whereas its low-temperature SCR activity was the optimum among all samples, revealing that BET surface area was not a decisive factor for catalytic activity.¹⁵ Besides, the N₂ adsorption–desorption and pore size distribution curves of the as-synthesized catalysts were listed in Fig. S1 (ESI†), these two kinds of catalysts all had a typical type IV isotherm together with a type-H4 loop, validating the formation of mesopore catalysts.^16,17 Plus, the pore size distributions of the catalysts (inset of Fig. S1, ESI†) were between 2 and 10 nm, further demonstrating the generation of mesopore catalysts. For pore volume, the acid-treated CNTs present the largest pore volume. After being supported by mixed metal oxides catalysts, the pore volumes of the as-prepared catalysts generally declined until the loading was more than 8%. The above results may be ascribing to nano metal oxide catalysts entry into the pore of CNTs, and then the pore volumes were decreased. As for 8% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalyst, its higher pore volume and BET surface area may be benefited from some bigger mesopores transformation into smaller mesopores, which was in good agreement with the result of Table S1.†

Fig. 1 NO conversion versus temperature for the as-synthesized catalysts. Reaction conditions: [NO] = [NH₃] = 420 ppm, [O₂] = 5%, 150 mg catalyst, WHSV = 280 000 mL g_cat⁻¹ h⁻¹.

XRD patterns of CNTs-based samples were given in Fig. 2. Evidently, all CNTs-based samples display four diffraction peaks at 26.2°, 42.8°, 53.9°, and 77.9°, respectively, corresponding to the typical graphite peaks.¹⁸ It should be pointed out that only weak Ce₂O₃ peak (PDF#74-1145) could be observed in Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts until the loading was greater than or equal to 6%, while the diffraction peaks of CeO₂, CuO, and MnO₂ could not be detected, proving the generation of amorphous metal oxide catalysts. For Ce–Cu–MnO_x/CNTs-IM catalyst, obvious CeO₂ (PDF#89-8436), CuO (PDF#89-2529), and Mn₃O₄ (PDF#86-2337) peaks could be found, indicating the formation of high crystalline metal oxide catalysts. Compared to crystalline catalysts, amorphous catalysts were more conducive to the SCR reaction, which was in line with the conclusions of NO conversion.¹⁹

Fig. 2 XRD patterns over acid-treated CNTs and the as-obtained catalysts: (a) acid-treated CNTs, (b) 2% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs, (c) 4% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs, (d) 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs, (e)8% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs, and (f) Ce–Cu−MnO_x/CNTs-IM.

The morphologies of CNTs-based samples were investigated via FE-SEM. As shown in Fig. 3A, acid-treated CNTs had smooth external surface. However, they became coarse after being supported by catalysts (Fig. 3B), suggesting the formation of CNTs-based catalysts. Again, element mappings [Fig. 3D–H] could find the signals of Ce, Cu, Mn, C, and O elements, validating that it has generated Ce, Cu, and Mn mixed-metal oxide catalysts. Regarding Ce–Cu–MnO_x/CNTs-IM catalyst (Fig. 3C), for one thing the catalysts were only partly anchored on CNTs, for another some aggregate catalysts dispersed among CNTs. It is known that agglomerate catalysts were bad for the SCR activity, which was in good agreement with the results of Fig. 1.

Fig. 3 FE-SEM images of CNTs-base samples: (A) acid-treated CNTs, (B) 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs, (C) Ce–Cu–MnO_x/CNTs-IM, and (D–H) element mappings of 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs (coming from the red region of B).

TEM, HR-TEM, and EDS were displayed in Fig. 4. For acid-treated CNTs (Fig. 4A), their surface were clear, while they became rough after being loaded by catalysts (Fig. 4B), confirming the generation of catalysts on CNTs. Furthermore, nanoflaky species corresponding to the catalysts could be observed at the edge of CNTs (Fig. 4C, HR-TEM), also validating the formation of catalysts on CNTs. Meanwhile, Ce, Cu, Mn, C, and O elements could be discovered in Fig. 4D (EDS spectrum), further evidencing the generation of Ce, Cu, and Mn mixed-metal oxide catalysts on CNTs, which were consistent with the conclusions of FE-SEM (Fig. 3). Moreover, apparent lattice fringe due to the metal oxide catalysts could not be found in HR-TEM, revealing that metal oxide catalysts with an amorphous phase have decorated on CNTs, which was in consistent with the results of XRD (Fig. 2). It is worthy to mention that amorphous catalysts were in favor of the SCR performance,²⁰ which was in good accordance with the results of Fig. 1.

Fig. 4 TEM images of (A) acid-treated CNTs, (B) 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs, and (C) HR-TEM of (B), as well as (D) EDS spectrum from the circle region of (B).

The STEM images and element mappings were adopted to further illustrate the morphologies of 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs. It is clear from Fig. 5A that some bright spots dispersed on CNTs, validating the presence of heavy metal elements on CNTs. In addition, highly-dispersed bright dots could be observed clearly from the high-resolution STEM (Fig. 5B), also demonstrating the existence of heavy metal elements on CNTs. It is worth noting that cylindrical element distribution images corresponding to Mn, Ce, Cu, O, and C elements could be discovered from element mappings [Fig. 5C–K], further proving that Mn, Ce, and Cu mixed-metal oxide catalysts have been formed and well dispersed on CNTs, which was in line with the conclusions of FE-SEM and TEM.

Fig. 5 (A and B) STEM images and (C–K) element mappings of 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs (B coming from the red region of A).

The valence states of metal oxide catalysts were evaluated by XPS and the results were displayed in Fig. 6. The XPS full spectrum of 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs (Fig. 6A) presents the signals of Ce, Cu, Mn, C, and O elements, indicating the formation of Ce, Cu, and Mn mixed-metal oxide catalysts, which was in line with the results of FE-SEM, TEM, and STEM. For Mn 2p, the binding energies of 642.2 and 653.8 eV were indexed to Mn 2p_3/2 and Mn 2p_1/2, respectively, along with a energy separation of 11.6 eV, suggesting the formation of MnO₂.²¹ It should be noted that MnO₂ exhibited the optimum SCR activity among all manganese oxides,²² which was correlated with the conclusions of Fig. 1. As for Ce 3d spectra (Fig. 6C), the peaks of M2 and N2 corresponding to the electronic state of 3d¹⁰4f¹ was due to Ce³⁺ (Ce₂O₃), whereas other six peaks were owing to Ce⁴⁺ (CeO₂),²³ revealing the co-existence of Ce³⁺ and Ce⁴⁺. Moreover, Table S2 (ESI†) displays that the Ce³⁺/(Ce³⁺ + Ce⁴⁺) ratio over 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs (0.162) was higher than that of over Ce–Cu–MnO_x/CNTs-IM (0.138). And higher ratio of Ce³⁺/(Ce³⁺ + Ce⁴⁺) could lead to more unsaturated chemical bands and vacancies, and thereby, increase the content of chemisorbed oxygen,²⁴ which was conducive to the catalytic activity.

Fig. 6 XPS spectra of (A) full spectrum, (B) Mn 2p, (C-a) Ce 3d, (D) Cu 2p, (E) Cu auger, and (F-a) O 1s for 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs; (C-b) and (F-b) for Ce–Cu–MnO_x/CNTs-IM.

Regarding Cu 2p (Fig. 6D), the binding energies at 954.6 and 934.6 eV was indexed to Cu 2p_1/2 and Cu 2p_3/2, respectively, together with two obvious satellites, indicating the formation of CuO.²⁵ Furthermore, a energy gap of 20.0 eV was attributed to the Cu 2p_1/2 and Cu 2p_3/2, also revealing the presence of CuO.²⁶ Meanwhile, the kinetic energy of Cu auger was 915.8 eV, further proving the generation of CuO.²⁷ In the case of O 1s (Fig. 6F), it could be divided into three peaks. The peaks at around 529.4 eV were indexed to lattice oxygen (labeled as O_x), and the binding energies between 529.4 and 536.5 eV were assigned to surface oxygen (designated as O_y). Besides, the ratio of O_y/(O_x + O_y) (Table S2, ESI†) shows that 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs (0.799) was higher than that of Ce–Cu–MnO_x/CNTs-IM (0.773). And O_y possessed higher mobility, which was conducive to NO oxidation to NO₂, and thus, could improve the SCR reaction.²⁸ According to the XPS data, it could conclude the generation of Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalyst, which was correlated with the results of FE-SEM, TME, and STEM.

On the basis of XRD and XPS results, the Ce₂O₃, CeO₂, CuO, and MnO₂ have been formed in the preparation procedure of catalysts. Therefore, the formation mechanism of Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts was proposed as follows: Ce³⁺ and Cu²⁺ were firstly absorbed on CNTs owing to the electrostatic force between acid-treated CNTs and metal cations. Subsequently, CeCl₃ and Cu(NO₃)₂ were partly hydrolyzed to Ce(OH)₃, HCl, Cu(OH)₂, and HNO₃ on CNTs.²⁹ Afterwards, Cl₂ and O₂ were generated due to the reaction of HCl and HNO₃ with KMnO₄, respectively, which would facilitate the hydrolyzation process. Meanwhile, KMnO₄ was reduced to MnO₂ on CNTs, and the Ce(OH)₃–Cu(OH)₂–MnO₂/CNTs samples were successfully synthesized. Finally, the Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts could be obtained via a heating dehydration process.^30,31 And the relevant reaction equations were given as follows:

(1) + (2) → Total reaction equation

Conclusions

In a word, Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalysts were firstly prepared by a redox approach, and showed 58–85% NO conversions between 80 and 180 °C under a weight hourly space velocity of 280 000 mL g_cat⁻¹ h⁻¹. Note that the 6% Ce₂O₃–CeO₂–CuO–MnO₂/CNTs catalyst was possessed of amorphous Ce₂O₃, CeO₂, CuO, and MnO₂, as well as high Ce³⁺ and O_y contents, which endowed it with the first-rate low-temperature SCR activity in the whole temperature range.

Acknowledgements

This work was supported by Scientific and Technological Program-Funded Project of Fuzhou City (Grant No. 2015H0016). I also appreciated the help of Mr Xinqi Zhang from the Testing Center of Fuzhou University.

Notes and references

1. X. Wang, X. Li, Q. Zhao, W. Sun, M. Tade and S. Liu, Chem. Eng. J., 2016, 288, 216.
2. Z. Liu, H. Su, J. Li and Y. Li, Catal. Commun., 2015, 65, 51.
3. N. Usberti, M. Jablonska, M. D. Blasi, P. Forzatti, L. Lietti and A. Beretta, Appl. Catal., B, 2015, 179, 185.
4. Z. Ma, X. Wu, Y. Feng, Z. Si and D. Weng, Catal. Commun., 2015, 69, 188.
5. Y. Zhang, Y. Zheng, X. Wang and X. Lu, Catal. Commun., 2015, 62, 57.
6. G. Qi and R. T. Yang, Appl. Catal., B, 2003, 44, 217.
7. D. Zhang, L. Zhang, C. Fang, R. Gao, Y. Qian, L. Shi and J. Zhang, RSC Adv., 2013, 3, 8811.
8. C. Fang, D. Zhang, L. Shi, R. Gao, H. Li, L. Ye and J. Zhang, Catal. Sci. Technol., 2013, 3, 803.
9. L. Zhang, D. Zhang, J. Zhang, S. Cai, C. Fang, L. Huang, H. Li, R. Gao and L. Shi, Nanoscale, 2013, 5, 9821.
10. L. Wang, B. Huang, Y. Su, G. Zhou, K. Wang, H. Luo and D. Ye, Chem. Eng. J., 2012, 192, 232.
11. X. Chen, S. Gao, H. Wang, Y. Liu and Z. Wu, Catal. Commun., 2011, 14, 1.
12. Q. Li, H. Yang, Z. Ma and X. Zhang, Catal. Commun., 2012, 17, 8.
13. J. Qiu, K. Zhuang, M. Lu, B. Xu and Y. Fan, Catal. Commun., 2013, 31, 21.
14. G. Qi, R. T. Yang and R. Chang, Appl. Catal., B, 2004, 51, 93.
15. Y. Zhang, Y. Zheng, X. Wang and X. Lu, RSC Adv., 2015, 5, 28385.
16. L. Schill, S. Putluru, R. Fehrmann and A. Jensen, Catal. Lett., 2014, 144, 395.
17. K. Sing, D. Everet, R. Haul, L. Moscou, R. Pierotti, J. Rouquerol and T. Siemieniewska, Pure Appl. Chem., 1985, 57, 603.
18. H. Xia, Y. Wang, J. Lin and L. Lu, Nanoscale Res. Lett., 2012, 7, 1.
19. R. Guo, Q. Chen, H. Ding, Q. Wang, W. Pan, N. Yang and C. Lu, Catal. Commun., 2015, 69, 165.
20. Z. Wu, B. Jiang, Y. Liu, W. Zhao and B. Guan, J. Hazard. Mater., 2007, 145, 488.
21. M. He, Y. Zheng and Q. Du, Mater. Lett., 2013, 104, 48.
22. F. Kapteijn, L. Singoredjo, A. Andreini and J. Moulijn, Appl. Catal., B, 1994, 3, 173.
23. X. Yao, F. Gao, Q. Yu, L. Qi, C. Tang, L. Dong and Y. Chen, Catal. Sci. Technol., 2013, 3, 1355.
24. B. Guan, H. Lin, L. Zhu and Z. Huang, J. Phys. Chem. C, 2011, 115, 12850.
25. C. Seo, B. Choi, H. Kim, C. Lee and C. Lee, Chem. Eng. J., 2012, 191, 331.
26. F. Li, X. Liu, Q. Zhang, T. Kong and H. Jin, Cryst. Res. Technol., 2012, 47, 1140.
27. Y. C. Zhang, J. Y. Tang, G. L. Wang, M. Zhang and X. Y. Hu, J. Cryst. Growth, 2006, 294, 278.
28. M. Pourkhalil, A. Z. Moghaddam, A. Rashidi, J. Towfighi, K. J. Jozani and H. Bozorgzadeh, Catal. Lett., 2013, 143, 184.
29. M. U. Anu Prathap, R. Srivastava and B. Satpati, Electrochim. Acta, 2013, 114, 285.
30. C. Lu, L. Qi, J. Yang, D. Zhang, N. Wu and J. Ma, J. Phys. Chem. B, 2004, 108, 17825.
31. X. Wang, Y. Zheng, Z. Xu, Y. Liu and X. Wang, Catal. Sci. Technol., 2014, 4, 1738.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental section, BET surface area, relative contents of Ce³⁺, Ce⁴⁺, O_x, and O_y, as well as N₂ adsorption–desorption and pore size distribution. See DOI: 10.1039/c6ra10482g

---