ISSN 00231584, Kinetics and Catalysis, 2015, Vol. 56, No. 6, pp. 741–746. © Pleiades Publishing, Ltd., 2015. Original Russian Text © D.S. Krivoruchenko, N.S. Telegina, D.A. Bokarev, A.Yu. Stakheev, 2015, published in Kinetika i Kataliz, 2015, Vol. 56, No. 6, pp. 729–734.
Mn–Ce/Beta “Bifunctional” Catalyst for the Selective Catalytic Reduction of Nitrogen Oxides with Ammonia D. S. Krivoruchenko, N. S. Telegina, D. A. Bokarev, and A. Yu. Stakheev* Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Moscow, 119991 Russia *email:
[email protected] Received January 29, 2015
Abstract—The properties of the Mn–Ce/Beta zeolite catalyst in the selective catalytic reduction (SCR) of NOx have been investigated. The introduction of Ce leads to a marked increase in the NOx conversion at 100– 250°C. The data of this study are consistent with the “bifunctional” pathway of SCR suggested for Mn/Beta, which consists of two stages—NO oxidation to NO2 over the oxide component and “fast” SCR over the zeo lite. The increased activity of Mn–Ce/Beta at the NOtoNO2 oxidation stage is due to the formation of MnCeOx mixed oxides enhancing the mobility of lattice oxygen. The determining role is played by the activ ity of the zeolite component in the “fast” SCR reaction. Keywords: bimetallic catalysts, Mn, Ce, SCR, zeolite catalyst DOI: 10.1134/S0023158415060051
INTRODUCTION Nitrogen oxide (NOx) emissions from stationary sources (thermal power stations) and mobile sources (automotive engines) pose a serious environmental problem. This is due to the fact that nitrogen oxides participate in ozone depletion, in the greenhouse effect, and in the formation of photochemical smog and acid rain [1]. One of the most efficient ways of removing nitrogen oxides is their selective catalytic reduction (SCR) with ammonia: 4NH3 + 4NO + O2 → 4N2 + 6H2O. (I) Presentday industries use V2O5/TiO2based cata lysts promoted with tungsten and molybdenum oxides in this process. However, because of the high toxicity of vanadium and because of the necessity of lowering the working temperature of the catalyst to 150–200°С, a typical temperature of the exhaust of highefficiency engines, search for alternative catalytic systems is still of importance. Particularly promising are zeolite cata lysts promoted with transition metals (Fe, Cu, Mn, Ce, Ni) [2–4] and oxide catalytic systems, such as MnOx/Al2O3 [5–8] and CuO/TiO2 [9]. Ironcontaining zeolites (Fe/Beta and Fe/ZSM5) [10, 11] are particularly attractive, for they are highly active and stable. The advantages of these systems include their resistance to the action of SO2. However, they are insufficiently active at reaction temperatures below 250°С. We demonstrated earlier [12] that the activity of FeBeta at 150–250°С can be significantly enhanced by promoting the zeolite with manganese. We think that
the observed effect is due to the “bifunctional” mech anism consisting of the following two stages: oxidation of NO into NO2 over the oxide compo nent MnOx, 2NO + O2 → 2NO2 (II) “fast” SCR over the zeolite component, 2NH3 + NO + NO2 → 2N2 + 3H2O. (III) This mechanism is in good agreement with the lit erature [13, 14]. This study, which continued a series of earlier stud ies in this area, was aimed at solving two problems. The first was to further increase the activity of Mn/FeBeta in the SCR of NOx with ammonia at 150–250°С by promoting the Mn component and to increase the oxi dizing power of the catalyst, i.e., its activity in reaction (II). The second problem was to determine the contri bution from the zeolite component to the total activity of the catalyst in the SCR of NOx with ammonia. In order to solve the first problem, the Mn compo nent was modified with CeO2, which possesses redox properties [15] and is capable of storing and releasing oxygen in the redox process Ce+4 Ce+3 [3, 16, 17]. Mn–Ce binary systems on oxide supports were inves tigated in detail [17, 19–22]. It was demonstrated in a number of studies that the introduction of cerium dioxide raises the activity of the Mn component in the oxidation of NO into NO2 [6, 18]. As was mentioned above, in an earlier work [12] we suggested a “bifunctional” mechanism for the SCR reaction over Mn/FeBeta. For verifying this mecha nism, it is necessary to discriminate the contributions from the “bifunctional” mechanism (reactions (II)
741
742
KRIVORUCHENKO et al.
and (III)) and “standard” SCR taking place on the zeolite component FeBeta (reaction (I)) to the overall process. For this purpose, we compared the catalytic properties of Mn–Ce systems supported on zeolites HBeta and FeBeta. Earlier [23], we demonstrated that zeolite HBeta, which is structurally similar to FeBeta, is ineffective in reaction (I), but its activity in the “fast” SCR reaction (III) is practically equal to the activity of the Fecontaining zeolite. In view of this fact, a simple comparison of the catalytic characteris tics of Mn–Ce/HBeta and Mn–Ce/FeBeta would provide a fairly correct estimate of the contribution from “standard” SCR to the overall SCR of nitrogen oxides with ammonia over the Mn–Ce/Beta catalysts. EXPERIMENTAL Catalyst Preparation Catalysts were synthesized using commercial zeo lite Beta in Hform (Si/Al = 12.5; residual iron con tent of ~0.02 wt %, according to elemental analysis data) and in Feform (Si/Al = 12.5; residual iron con tent of ~0.9 wt %, according to elemental analysis data), both received from Zeolist International. Cata lysts containing Mn or Ce alone (16 wt %) were pre pared by incipientwetness impregnation of the initial zeolite with a Mn(NO3)2 or Ce(NO3)3 solution, respectively. Bimetallic samples were obtained by incipientwetness coimpregnation. The impregnated samples were dried at room temperature for 24 h and were then calcined at 550°С for 4 h in flowing air (~300 mL/min per gram of catalyst). The 0.2–0.4 mm particle size fraction was used in catalytic experiments. Physicochemical Characterization of Catalysts Temperatureprogrammed reduction of catalysts with hydrogen (H2TPR) was carried out using a semi automated flowthrough setup consisting of a Ushaped quartz reactor, a water vapor trap, a ther malconductivity detector, and a data acquisition and processing unit. Prior to a TPR run, the sample (100 mg) was purged with argon at 325°С for 1 h and was then cooled to room temperature. Reduction was carried out in a flowing 5 vol % H2/Ar gas mixture (30 mL/min) while elevating the temperature to 820°С at a constant rate of 10°C/min. For removing the water yielded the reduction reaction from the gas phase, a trap cooled to –70°С with a dry ice–ethanol mixture was placed between the reactor and the detec tor. The detector was calibrated against TPR data for CuO (SigmaAldrich Chemie GmbH, Germany, 99%, 0.08–11 mg sample). TPR peaks were decom posed into components using the Ekokhrom program. The microstructure of materials was studied by field emission scanning electron microscopy (FESEM) on a Hitachi SU8000 microscope (Hita chi, Japan). The sample to be examined was placed on the surface of an aluminum stage 25 mm in diameter,
secured using a conductive adhesive tape, and covered with a 7nmthick conductive metal layer (Au/Pd, 60/40) by magnetron sputtering. Micrographs were obtained in the secondary electron imaging mode at an accelerating voltage of 2 kV and a working distance of 4–5 mm. In morphological studies, we applied a correction for the surface effects of conductive layer deposition. Catalytic Tests Catalytic tests were carried out in a quartz flow reactor with an inner diameter of 6 mm. The gas mix ture (600 ppm NH3, 500 ppm NO, 10% O2, 6% Н2О, N2 to balance) was supplied at a rate of 300 mL/min (GHSV = 270000 h–1), and the catalyst weight was 0.040 g. Reaction products were analyzed using an FTIR gas analyzer (GASMET Dx4000N, Gasmet Technologies Oy, Finland). Experiments on NO oxidation into NO2 The were performed using the following gas mixture: 530 ppm NO, 10% O2, 6% Н2О, N2 to balance. The catalyst weight was 0.040 g, and GHSV was 270000 h–1. Reac tion products were analyzed on a luminescence ana lyzer CLD822Mh, EcoPhysics, Switzerland). NOx conversion ( X NO x ) and the conversion of NO into NO2 ( X NO−NO 2 ) were calculated via the following formulas:
X NO x =
C in, NO − (Cout, NO + Cout, NO2 + 2Cout, N2O) ; (1) C in, NO
X NO−NO2 =
Cin, NO − Cout, NO , Cin, NO
(2)
where Cin and Cout are gas concentrations at the inlet and outlet of the reactor. RESULTS AND DISCUSSION Scanning Electron Microscopy Data The SEM images of the HBeta and 16% Mn– 16% Ce/HBeta samples (Fig. 1) show distinct micro crystallites of the initial zeolite HBeta (Fig. 1a), whose size is ~0.4 µm. Modifying the zeolite with cerium or manganese yields much smaller oxide phase parti cles on the outer surface of the zeolite. As the amount of supported phase is increased (16%Mn– 16%Ce/HBeta, Fig. 1b), the number of oxide particles increase markedly and the zeolite microcrystallites become densely decorated. Note that the formation of large oxide particles is not observed, the morphology remains unchanged, and the MnOx–CeOx particles are in intimate contact with the zeolite surface. KINETICS AND CATALYSIS
Vol. 56
No. 6
2015
Mn–Ce/Beta “BIFUNCTIONAL” CATALYST
500 nm
(а)
743
500 nm
(b)
Fig. 1. SEM images of the (a) HBeta and (b) 16%Mn–16%Ce/HBeta samples.
Effect of Ce on the NO Oxidation Activity of the Catalyst Figure 2 plots the conversion of NO into NO2 as a function of the reaction temperature for the 16%Ce/HBeta, 16%Mn/HBeta, and 16%Mn– 16%Ce/HBeta catalysts. The introduction of cerium markedly enhances the NO oxidation activity of Mn/HBeta. For example, the NO conversion over 16%Mn–16%Ce/HBeta at 200°С is 10%, while the NO conversion over the unmodified catalyst 16%Mn/HBeta at the same temperature is only 1.5%. The highest NO conversion is attained with 16%Mn– 16%Ce/HBeta (47% at 330°С) and 16%Mn/HBeta (35% at 395°С). Note that the activity of 16%Ce/HBeta in NO oxidation is low, so the observed increase in catalytic activity cannot be due to the con tribution from the Ce component. Since the rate of NO oxidation into NO2 decreases with an increasing temperature, the NO conversion decreases after pass ing through a maximum and approaches the equilib rium NO/NO2 ratio. Thus, cerium plays the role of a promoter, enhancing the oxidizing activity of MnOx by a factor of about 2.
MnO2 to Mn2O3, and the peak at 425°С is due to the reduction of Mn2O3 to MnO [28, 29]. The total hydro gen uptake is 2.69 mmol/gCat, indicating that the for mal composition of manganese oxide is MnO1.92. A comparison of the H2TPR spectra of Mn– Ce/FeBeta and Mn/FeBeta shows that the introduc tion of Ce lowers the Mn reduction temperature from 425 to 395°С and diminishes the total hydrogen uptake to 2.3 mmol/gCat. The shift of the reduction peak is possibly due to the increase in oxygen mobility as a result of modifying the catalyst with cerium, which leads to a decrease in the Mn3+ → Mn2+ transi tion temperature [30]. In addition, the spectrum of Mn–Ce/FeBeta shows a shoulder at 200°С, which X NO–NO2 , % 60
40
1 2 3 4
TPR Data The H2TPR method is widely used in the determi nation of oxygen mobility in mixed oxide systems. It provides means to see how the promoter affects the redox properties of the oxide component. The H2TPR spectrum of the 16%Ce/FeBeta cata lyst (Fig. 3) shows a hydrogen uptake peak at 225°С, which is assignable to the reduction of surface oxygen [24]. The hydrogen uptake value is 0.104 mmol/gCat. The TPR spectrum of the 16%Mn/FeBeta catalyst exhibits peaks at 330 and 425°С. According to the lit erature [18, 25–27], they characterize the consecutive reduction reactions Mn+4 → Mn+3 → Mn+2. The peak at 330°С arises from the reduction of amorphous KINETICS AND CATALYSIS
Vol. 56
No. 6
2015
20
0 100
200
300
400
500 T, °C
Fig. 2. (1–3) Temperature dependence of the NO conver sion into NO2 for the (1) 16%Mn/HBeta, (2) 16%Mn– 16%Ce/HBeta, and (3) 16%Ce/HBeta catalysts. (4) Equi librium NO/NO2 ratio.
744
KRIVORUCHENKO et al. X NOx , %
I, mV 15 12 9
100 1 2 3
80
60
6
1
40
2
3
3
20 0 100 150 200 250 300 350 400 450 500 550 T, °C Fig. 3. H2TPR spectra of the (1) 16%Mn/FeBeta, (2) 16%Mn–16%Ce/FeBeta, and (3) Ce/FeBeta samples.
also indicates an increase in oxygen mobility in the Cecontaining sample. According to the literature [31–34], the observed decrease in the oxide reduction temperature can be due to the formation of a MnOx–CeOx solid solution and, as a consequence, an increase in the mobility of lattice oxygen. It was demonstrated [35, 36] that mod ifying with cerium increases the NO oxidation activity of the Mn–Ce/HBeta catalyst. There is evidence [37– 39] that the introduction of cerium enhances the activity of Mn–Ce/HBeta in the total oxidation of hydrocarbons and carbon monoxide [40, 41]. Effect of Ce on the Activity of Mn/FeBeta in the SCR of NOx with Ammonia The introduction of Mn increases the lowtemper ature activity of FeBeta (Fig. 4), which is in agreement with the results of our earlier study [12]. For example, the 50% conversion temperature (T50) decreases from ~230°С for FeBeta to ~205°С for the catalyst contain ing 16% Mn. However, the NOx conversion declines at temperatures above 350°С, which is hypothetically due to the following competing ammonia oxidation reaction, which, on the one hand, diminishes the amount of reductant and, on the other hand, generates additional amounts of NO: 4NH3 + 5O2 → 4NO + 6H2O. (IV) The introduction of Ce further enhances the low temperature activity of the catalyst. The 50% conver sion temperature decreases from ~205°С for 16%Mn/FeBeta to ~145°С for the catalyst addition ally containing 16% Ce. A nearly complete NOx con version for the 16%Mn–16%Ce/FeBeta catalyst is reached at ~200°С.
0 100
200
300
400
500 T, °C
Fig. 4. Temperature dependence of the NOx conversion for the (1) 16%Mn–16%Ce/FeBeta, (2) 16%Mn/FeBeta, and (3) FeBeta catalysts.
The decrease in NOx conversion above 350°С for the 16%Mn–16%Ce/FeBeta catalyst is more signifi cant than for 16%Mn/FeBeta. This is consistent with the observed increase in the oxidizing activity of the Mn–Ce component. On the whole, the above data are in good agree ment with the hypothesized “bifunctional” mecha nism of the reaction. In the framework of this hypoth esis, the observed increase in catalytic activity can be attributed to the increase in the activity of the Mn–Ce oxide component in NO oxidation (Fig. 1), which increases the overall rate of SCR proceeding via the “bifunctional” pathway. Contribution from the Zeolite Component to the Activity of the Mn–Ce/Beta Catalyst It follows from the data presented in Fig. 5 that the HBeta sample is inactive below 300°С, while the Fecontaining zeolite shows a considerable activity in NOx SCR. Therefore, at reaction temperatures below 300°С the contribution from the initial HBeta to the total activity of the 16%Mn–16%Ce/HBeta catalyst is negligibly small. Nevertheless, the activities of 16%Mn–16%Ce/FeBeta and 16%Mn–16%Ce/HBeta at 100–250°С are equal, although their initial zeolites differ markedly in NOx SCR activity. These results correlate with the find ing that the activity of zeolite HBeta containing traces of Fe in the “standard” SCR reaction is very low [23]. However, its activity in “fast” SCR is equal to the activity of FeBeta in a wide tempera ture range. KINETICS AND CATALYSIS
Vol. 56
No. 6
2015
Mn–Ce/Beta “BIFUNCTIONAL” CATALYST
Mn–Ce/Beta catalyst. The contribution from the zeolite component to the total activity of Mn– Ce/Beta is insignificant. The results of this study might be helpful in devel oping new catalytic systems that would be highly active at 100–250°С and in elucidating the mechanism of the SCR of NOx with ammonia over zeolite and oxide catalysts.
X NOx , % 100
80 1 2 3 4 5
60
40
ACKNOWLEDGMENTS This study was supported by the Russian Founda tion for Basic Research, grant no. 150307802 A. D.S. Krivoruchenko is grateful to Haldor Topsøe A/S for a grant in the framework of the Haldor Topsøe Ph.D. Scholarship program.
20
0 100
200
300
500 T, °C
400
Fig. 5. Temperature dependence of the NOx conversion for the (1) 16%Mn–16%Ce/FeBeta, (2) 16%Mn– 16%Ce/HBeta, (3) FeBeta, (4) HBeta, and (5) 16%Ce/HBeta catalysts.
Note that the Ce/HBeta catalyst shows a notice able SCR activity only above 250°С and is much less active than FeBeta. This fact also disproves the assumption that the activity of the zeolite catalysts increases as a results of their modification with Ce. Thus, the fact that 16%Mn–16%Ce/FeBeta and 16%Mn–16%Ce/HBeta show the same behavior at 100–250°С indicates that the SCR of NOx over Mn– Ce/Beta proceeds mainly via the “bifunctional” path way, since, for the reaction to proceed via this pathway, it is crucial that the zeolite component be active in “fast” SCR, while its activity in “standard” SCR is inessential. Thus, this study demonstrated that the promotion of the Mncontaining zeolites with cerium substan tially enhances their lowtemperature (100–250°C) activity in the SCR of NOx with ammonia. The results of this study [12] are in agreement with our hypothesis that the SCR reaction proceeds via the “bifunctional” pathway consisting of two stages, namely, NO oxidation into NO2 on MnOx–CeOx par ticles and “fast” SCR on the zeolite component: 2NO + O2 → 2NO2 (on MnOx–CeOx), (V) 2NH3 + NO + NO2 → 2N2 + 3H2O (on zeolite). (VI) As a result of the introduction of Ce, the Mn–Ce component shows a higher NO oxidation activity, which is likely due to the formation of a MnOx–CeOx solid solution and the consequent increase in the mobility of lattice oxygen. The increase in NO oxida tion activity leads to an increase in the rate of the reac tion proceeding via the “bifunctional” pathway. It was also established that the “bifunctional” path way is the main way the SCR of NOx occurs over the KINETICS AND CATALYSIS
Vol. 56
No. 6
745
2015
REFERENCES 1. Bosch, H. and Janssen, F., Catal. Today, 1988, vol. 2, p. 369. 2. Zhu, Z., Liu, Z., Niu, H., and Liu, S., J. Catal., 1999, vol. 187, p. 245. 3. Parvulescu, V.I., Grange, P., and Delmon, B., Catal. Today, 1998, vol. 46, p. 233. 4. Carja, G., Kameshima, Y., Okada, K., and Madhuso odana, C.D., Appl. Catal., B, 2007, vol. 73, p. 60. 5. Yang, S.J., Li, J.H., Wang, C.Z., Chen, J.H., Ma, L., Chang, H.Z., Chen, L., Peng, Y., and Yan, N.Q., Appl. Catal. B, 2012, vol. 117, p. 73. 6. Qi, G.S., Yang, R.T., and Chang, R., Appl. Catal. B, 2004, vol. 51, p. 93. 7. Singoredjo, L., Korver, R., Kapteijn, F., and Moulijn, J., Appl. Catal. B, 1992, vol. 1, p. 297. 8. Kijlstra, W.S., Brands, D.S., Poels, E.K., and Bliek, A., Catal. Today, 1999, vol. 50, p. 133. 9. Ramis, G., Yi, L., Busca, G., Turco, M., Kotur, E., and Willey, R.J., J. Catal., 1995, vol. 157, p. 523. 10. Huang, H., Long, R.Q., and Yang, R.T., Appl. Catal. A, 2002, vol. 235, p. 241. 11. Kumar, M.S., Schwidder, M., Grunert, W., and Brück ner, A., J. Catal., 2004, vol. 227, p. 384. 12. Krivoruchenko, D.S., Kucherov, A.V., Telegina, N.S., Bokarev, D.A., Selvam, P., and Stakheev, A.Yu., Russ. Chem. Bull., 2014, vol. 63, p. 389. 13. Salazar, M., Becker, R., and Grünert, W., Appl. Catal. B, 2015, vol. 165, p. 316. 14. Stakheev, A.Yu., Baeva, G.N., Bragina, G.O., Tele guina, N.S., Kustov, A.L., Grill, M., and Thogersen, J.R., Top. Catal., 2013, vol. 56, p. 427. 15. Zhang, L., Pierce, J., Leung, V.L., Wang, D., and Epling, W.S., J. Phys. Chem. C, 2013, vol. 117, p. 8282. 16. Qi, G.S. and Yang, R.T., J. Catal., 2003, vol. 217, p. 434. 17. Qi, G.S. and Yang, R.T., J. Phys. Chem. B, 2004, vol. 108, p. 15738. 18. Cao, F., Su, S., Xiang, J., Wang, P., Hu, S., Sun, L., and Zhang, A., Fuel, 2015, vol. 139, p. 232.
746
KRIVORUCHENKO et al.
19. Hong, W.J., Iwamoto, S., and Inoue, M., Catal. Today, 2011, vol. 164, p. 489. 20. Liu, Y., Gu, T.T., Weng, X.L., Wang, Y., Wu, Z.B., and Wang, H.Q., J. Phys. Chem. C, 2012, vol. 116, p. 16582. 21. Wu, Z.B., Jiang, B.Q., Liu, Y., Zhao, W.R., and Guan, B.H., J. Hazard. Mater., 2007, vol. 145, p. 488. 22. Machida, M., Uto, M., Kurogi, D., and Kijima, T., Chem. Mater., 2000, vol. 12, p. 3158. 23. Mytareva, A.I., Baeva, G.N., Bragina, G.O., Selvam, P., Bokarev, D.A., and Stakheev, A.Yu., Mendeleev Com mun., 2014, vol. 24, p. 311. 24. Zhang, R., Zhang, B., Shi, Z., Liu, N., and Chen, B., J. Mol. Catal. A: Chem., 2015, vol. 398, p. 223. 25. You, Z., Balint, I., and Aika, K., Appl. Catal., B, 2004, vol. 53, p. 233. 26. Nie, J., Wu, X., Ma, Z., Xu, T., Si, Z., Chen, L., and Weng, D., Chin. J. Catal., 2014, vol. 35, p. 1281. 27. Sun, M., Lan, B., Yun, L., Ye, F., Song, W., He, J., Diao, G., and Zheng, Y., Mater. Lett., 2012, vol. 86, p. 18. 28. Lou, X., Liu, P., Li, J., and He, K., Appl. Surf. Sci., 2014, vol. 307, p. 382. 29. Liu, J., Yu, C.Y., Zhao, P.Q., and Chen, G.X., Appl. Surf. Sci., 2012, vol. 258, p. 9096.
30. Tang, X., Li, Y., Huang, X., Xu, Y., Zhu, H., Wang, J., and Shen, W., Appl. Catal., B, 2006, vol. 62, p. 265. 31. Machida, M., Catal. Surv. Jpn., 2002, vol. 5, p. 91. 32. Shi, L., Chu, W., Qu, F., and Luo, S., Catal. Lett., 2007, vol. 113, p. 59. 33. Wu, Z., Jin, R., Liu, Y., and Wang, H., Catal. Commun., 2008, vol. 9, p. 2217. 34. Xingyi, W., Qian, K., and Dao, L., Appl. Catal., B, 2009, vol. 86, p. 166. 35. Qi, G., Yang, R.T., and Chang, R., Catal. Lett., 2003, vol. 87, p. 67. 36. Tang, X., Hao, J., Yi, H., and Li, J., Catal. Today, 2007, vol. 126, p. 406. 37. Delimaris, D. and Ioannides, T., Appl. Catal., B, 2008, vol. 84, p. 303. 38. Li, W.B., Wang, J.X., and Gong, H., Catal. Today, 2009, vol. 148, p. 81. 39. Saqer, S.M., Kondarides, D.I., and Verykios, X.E., Appl. Catal., B, 2011, vol. 103, p. 275. 40. Liu, X., Lu, J., Qian, K., Huang, W., and Luo, M., J. Rare Earths, 2009, vol. 27, p. 418. 41. Zou, Z.Q., Meng, M., and Zha, Y.Q., J. Phys. Chem. C, 2009, vol. 114, p. 468.
Translated by D. Zvukov
KINETICS AND CATALYSIS
Vol. 56
No. 6
2015