(63a) Controlling Deactivation of ZSM-5 Catalysts in the Methanol-to-Hydrocarbons Reaction By Co-Feeding Water

The methanol-to-hydrocarbons (MTH) reaction yields a wide range of hydrocarbons of great interest in the petrochemical industry such as ethylene, propylene, aromatics and a gasoline-like mixture. This reaction takes place in the presence of a shape-selective acid catalyst, such as ZSM-5 or SAPO-34 zeolites. This kind of catalysts presents a rapid deactivation due to the formation and deposition of coke that blocks the access of reactants to the catalyst active sites. However, the catalyst can recover its activity after being subjected to a regeneration process by burning the coke deposited on it. Thus, the MTH catalyst may operate in reaction-regeneration cycles. Many efforts have been made with the purpose of improving the catalytic performance of the MTH reaction with special emphasis on attenuating catalyst deactivation. One common strategy is co-feeding water, which has been proved to be satisfactory for increasing the lifetime of SAPO-34 catalysts with an important impact for the industrial implementation of the methanol-to-olefins process.¹ However, this effect has not been studied with such detail for ZSM-5 catalysts, whose major research efforts are focused on zeolite modifications. Indeed, the modifiers improve the overall catalytic performance of ZSM-5 catalysts, such as delaying the catalyst deactivation that allows longer operational periods.^2,3 In an earlier stage of our research, we observed that the modification of ZSM-5 zeolites with ZnCl₂ improves the catalytic performance of the MTH reaction by reducing the formation of coke that leads to delay the catalyst deactivation. Now we aim to further improve the ZSM-5 catalyst performance in the MTH reaction by co-feeding water based on the experience reported for its counterpart SAPO-34.

In this work, we investigated the effect of co-feeding water with methanol as a strategy to control the deactivation of a Zn-modified ZSM-5 catalyst in the MTH reaction. The MTH experiments were performed in a fixed-bed reactor at 400 ºC, using different water contents in the feed (0, 10, 20 and 33 wt%), 100 mg of catalyst and a methanol flowrate of 0.1 (mol of carbon) h^-1. Each MTH experiment was run until the catalyst was highly deactivated (conversions under 5%) and the run time was reported as the catalyst operational time. After each run, the catalyst was taken out and subjected to thermogravimetric analysis (TGA) and soluble coke extraction to gain information about the coke retained. Additional MTH experiments were conducted in UV-VIS-NIR and FTIR spectroscopic operando reactors in order to gain more information on the coke formation kinetics.

Our results show that co-feeding water decreases the initial conversion of the methanol and dimethyl ether in this order: 98% conversion (0 wt% H₂O) < 80% (10 wt%) < 29% (20 wt%) < 11% (33 wt%). At the same time, the addition of any water content in the feed prolongs the catalyst operational time from 19 to 26 h on stream (see Figure 1). We further analyzed the catalyst deactivation in terms of the evolution of the conversion against time on stream for each MTH run. This analysis reveals that the conversion drop is slower as the water content in the feed increases. Also, the product distribution analysis in terms of the yield of different products against conversion shows no significant differences regardless of the water content used in the feed. We only observed that the yield of secondary products such as methane and carbon dioxide decreases with increasing water contents in the feed. The content of coke in the spent catalysts decreases as the content of water in the feed increases, following this order: 47 mg g^-1 (0 wt%) > 26 mg g^-1 (10 wt%) > 17 mg g^-1 (20 wt%) > 11 mg g^-1 (33 wt%). This is congruent with the decrease of the conversion capacity and the catalyst deactivation rate with increasing water contents in the feed.

Figure 1. Conversion evolution against time on stream for different water contents in the feed in the MTH reaction at 400 ºC.

We further studied the formation of retained species in the catalyst by means of UV-VIS-NIR and FTIR spectroscopies in operando reactors under different water contents in the feed. Both techniques allowed us to identify retained species that are reaction intermediates and coke precursors. The results show insignificant differences in the nature of retained species regardless of co-feeding water but significant differences in their concentration and kinetics: lower amounts and slower kinetics with increasing water content. For example, we observed the raise of a UV-VIS band at 394 nm attributable to polyalkylated aromatics⁴ for three different water contents in the feed with the following maximum absorbance at 750 s on stream: 0.28 (0 wt%) > 0.24 (33 wt%) > 0.22 (50 wt%). We observed a similar behavior between polyakylated aromatics and polyaromatics in the operando reactors, which shows that the formation of intermediates and coke is attenuated by co-feeding water. These results concord quite well with the observations obtained in the fixed-bed reactor. Thus, the lower conversion is linked with a slower catalyst deactivation and slower coke formation with increasing water contents in the feed.

The use of FTIR spectroscopy also allowed us to correlate the polyakylated aromatic and polyaromatic species with the level of accessibility of acid sites. In fact, hydroxyls groups interacting with Si (silanol), are the most affected ones among the different acid sites. This observation suggests that the extra water co-fed strongly interacts with these sites and this hinders the adsorption and further reaction of reactive hydrocarbon species. Thus, our results prove that the competitive adsorption of water lowers the catalyst surface coverage by any other hydrocarbon species and this decreases the rate of any kinetic event related to the MTH reaction.

In summary, we demonstrated that co-feeding water attenuates the deactivation of the Zn-modified ZSM-5 catalyst in the MTH reaction and this is mainly due to the competitive adsorption of the extra water co-fed. This decreases the amount and formation rate of coke. We also verified that co-feeding water has insignificant effects on the MTH chemistry, as the product distribution against conversion and the nature of intermediates are quite similar regardless of the water content in the feed. From a process engineering perspective, the addition of a small amount of water (10 wt%) in the feed may represent a positive outcome to the development of the MTH process using a Zn-modified catalyst. Under this scenario, the reactor may operate for a longer period reducing the reaction-regeneration cycles and keeping a high production level of primary products.

References

1. P. Tian, Y. Wei, M. Ye and Z. Liu, ACS Catal., 2015, 5, 1922–1938.
2. H. E. H. E. van der Bij and B. M. B. M. Weckhuysen, Chem. Soc. Rev., 2015, 44, 7406–7428.
3. X. Niu, J. Gao, Q. Miao, M. Dong, G. Wang, W. Fan, Z. Qin and J. Wang, Microporous Mesoporous Mater., 2014, 197, 252–261.
4. K. De Wispelaere, C. S. Wondergem, B. Ensing, K. Hemelsoet, E. J. Meijer, B. M. Weckhuysen, V. Van Speybroeck and J. Ruiz-Martínez, ACS Catal., 2016, 6, 1991–2002.