Effect of aluminum distribution in ZSM-22 zeolite on ethylene oligomerization

Abstract

Olefin oligomerization is a crucial process for fuel and chemical production. Here, we report that, in ZSM-22 zeolite, paired aluminum sites are more effective at ethylene oligomerization than isolated aluminum sites, giving higher conversion and forming heavier oligomers. These results highlight the importance of modulating the organization of aluminum in zeolite for catalysis.

---

Zeolites are the catalytic workhorses of a variety of reactions, because of the inherent acidity, which typically arises from framework aluminum sites.^1,2 Recently, there has been growing interest from the academy in investigating the proximal Al sites or paired Al sites to tune zeolite catalytic performance.^3,4 To date, most studies have focused on three-dimensional or two-dimensional zeolites, including ZSM-5,^5,6 ferrierite,⁷ mordenite,⁸ beta,⁹ and SSZ-13.¹⁰ However, the non-uniform distribution of Al in the different zeolite channels could have an additional influence on catalytic performance.

One-dimensional (1D) zeolites should be more suitable for such investigations due to the presence of unique microporous and non-intersecting pore channels.¹¹ Moreover, 1D zeolites have been found to be effective at catalyzing isomerization and oligomerization reactions, due to their exceptional shape selectivity.^12,13 In this report, we chose ZSM-22 (1D) as a study subject. It exhibits a TON framework topology with one-dimensional 10-membered ring channels.¹⁴ We prepared a series of H-ZSM-22 zeolite samples with comparable Al contents, crystal sizes and morphologies, but varying the number of Al pairs. We then evaluated their performances in the ethylene oligomerization reaction. The results revealed a clear relationship between the abundance of aluminum pairs in this zeolite and its oligomerization performance.

Three zeolite samples, named HZ-25(19%), HZ-27(10%), and HZ-34(17%), were synthesized in our laboratory based on a previous literature report,¹⁵ and a fourth sample named HZ-35(12%) was purchased. X-ray diffraction analysis confirmed that all samples have a TON topology (Fig. 1). Samples were prepared in their H-form by carrying out a typical NH₄⁺ ion exchange followed by calcination. The Si/Al ratio of zeolite was determined by performing inductively coupled plasma optical emission spectroscopy (ICP-OES), and is indicated by “x” in the sample code HZ-x. In general, based on Si/Al ratios, the four samples can be classified into two groups: Group I with a ratio of about 25–27, and Group II with a ratio of about 34–35.

Fig. 1 Powder XRD patterns of indicated samples.

Fig. 2 shows the electron micrographs of all samples, which exhibit a typical rod-like morphology. All crystals were nanosized, showing similar lengths of ∼50 nm in the [100] and [010] directions but different lengths in the [001] direction, specifically 100–350 nm for the two samples of Group I and 150–500 nm for the two Group II samples. Their acidity was probed using pyridine-IR, and the quantities of Brønsted acid sites (BAS) and Lewis acid sites (LAS) were estimated and are listed in Table 1. BAS predominated in all the samples, suggesting a likely minimal influence of LAS on catalytic performance. For zeolites with similar Si/Al ratios, fairly similar BAS concentrations were found, for example, 179 and 161 µmol g⁻¹ for HZ-25(19%) and HZ-27(10%), respectively. No significant difference in their acid properties was observed based on temperature-programmed desorption (TPD) of NH₃ (Fig. S1). Also, these two samples did not differ much in external surface acidity (Fig. S2). Table S1 lists their elemental textural properties: HZ-25(19%) and HZ-27(10%) exhibited similar surface areas (∼160–170 m² g⁻¹), micropore volumes (∼0.05 cm³ g⁻¹) and total pore volumes (∼0.2 cm³ g⁻¹).

Fig. 2 Transmission electron microscopy images of (a) HZ-25(19%), (b) HZ-27(10%), (c) HZ-34(17%), and (d) HZ-35(12%) samples.

Table 1 Concentrations of BAS and LAS probed using pyridine-IR

	Samples	BAS (µmol g^−1)	LAS (µmol g^−1)	Si/Al	Al-pair percent (%)
Group I	HZ-25(19%)	179	7	25	19
	HZ-27(10%)	161	10	27	10
Group II	HZ-34(17%)	155	13	34	17
	HZ-35(12%)	136	17	35	12

---

To investigate the proximity of aluminum atoms, a titration method using Co²⁺ cations as the probe was employed.¹⁶ In principle, two negative AlO⁻ charges in close proximity are required to balance the positive charge of Co²⁺ at the cationic site. Therefore, only two paired Al sites located in one zeolite ring can accommodate the Co²⁺ hexaaqua complexes. In contrast, Al atoms located far from each other can accommodate neither Co²⁺ hexaaqua complexes nor their dehydrated bare Co²⁺ cations and should be regarded as isolated Al sites. The amount of Co introduced into the zeolite after the ion-exchange process with a cobalt acetate solution was measured by ICP-OES (Table 1). The ZSM-22 samples with similar Si/Al ratios exhibited different percentages of total Al sites present as paired Al sites. Note that this percentage is denoted as “y” in the sample code HZ-x(y). The feasibility of experimentally determined Co-exchangeable Al fraction was also confirmed using Monte-Carlo calculations.

After confirming the comparable compositions and textural properties but different abundances of aluminum pairs among the samples, two H-ZSM-22 zeolites, namely HZ-25(19%) and HZ-27(10%), were extensively examined and compared for their catalytic performance in ethylene oligomerization, owing to the pronounced difference in their Al pair percentages. The reaction was conducted at 350–425 °C and atmospheric pressure, with a weight hourly space velocity (WHSV) of 5 h⁻¹. Ethylene conversion decreased with reaction time for both samples, most likely as a result of coke formation. However, at all temperatures, HZ-25(19%) exhibited higher conversions than did HZ-27(10%), as shown in Fig. 3a and Fig. S3. To normalize the activity to that per Brønsted acid site, turnover frequency (TOF) was also calculated, with TOF defined as moles of ethylene converted per mole of BAS per second (Fig. 3b). The higher activity of BAS associated with HZ-25(19%) was also observed in this calculation. The main products of the reaction were C3–C6 olefins, and plots of their selectivities with time are shown in Fig. 3c and Fig. S4. Interestingly, a higher proportion of heavier olefins, particularly C5 and C6 olefins, was produced over the sample with more acid pairs (HZ-25(19%)), likely due to consecutive oligomerization steps occurring at paired sites. In contrast, a higher selectivity for C4 olefins was observed over HZ-27(10%), with these olefins produced in a one-step dimerization of two ethylenes. We may attribute the above differences in activity and selectivity to be related to the amount of paired Al sites. Such a trend was also observed for HZ-34(17%) and HZ-35(12%) samples (Fig. S5).

Fig. 3 (a) Rates of conversion of ethylene at 400 °C. (b) TOF values at indicated temperatures. (c) Product selectivities at 400 °C. Reaction conditions: 1 bar, WHSV = 5 h⁻¹.

To elucidate the activity difference between paired and isolated Al sites, we performed DFT calculations. We first considered the possible influence of acidic strength, as ethylene oligomerization is mainly catalyzed by BAS in zeolites. The deprotonation energy (DPE) is a widely used indicator for acidic strength, and is defined as the energy needed to remove a bonded proton from the anionic zeolite structure (Z⁻) to an infinite, non-interacting distance (ZH → Z⁻ + H⁺). Our calculation results suggested only a small difference between the DPE values for the isolated and paired acid modes (Fig. S6), consistent with previous reports showing no significant difference between the acid properties of isolated Al sites and Al pairs.^17,18

To further elucidate the reaction energetics, specifically to quantify adsorption and reaction energies along the ethylene-to-butene pathway, static calculations were employed. The models comprised the unoccupied zeolite with ethylene, the ethylene-adsorbed zeolite, the butene-adsorbed zeolite, and the regenerated zeolite with desorbed butene (Fig. 4). The results indicated the adsorption of two ethylene molecules to be more favorable in models with aluminum sites in close proximity, being more exothermic by ∼16–26 kJ mol⁻¹ and hence more feasible. After being adsorbed, the molecules would undergo the oligomerization reaction, resulting in the formation of a single butene molecule. A slight difference was observed between the total energy levels of isolated (−138.3 kJ mol⁻¹) and paired acid sites (−135.0 to −144.7 kJ mol⁻¹). Accordingly, we concluded that ethylene adsorption at the paired acid sites is stronger than that at isolated sites, with the enhanced adsorption promoting the reaction. We anticipate that oligomerization intermediates preferentially accumulate near paired acid sites, thereby enhancing the progression of subsequent reactions, as supported by experimental observations. These observations also validated a previous report, based on experimental work, of oligomerization of propene over proximal acid sites having a turnover rate per Al(H⁺) higher than that of propene over isolated acid sites in H-ZSM-5.¹⁹

Fig. 4 Energy profile diagrams for the conversion of ethylene to butene. The number in the legend represents the number of T-sites (Si atoms) between paired Al sites.

In summary, we have shown that in one-dimensional ZSM-22 zeolite, paired Al sites are more active for ethylene oligomerization than isolated Al sites. Our work also illustrates how controlling acid site distribution at the nanoscale can be used to develop high-performing zeolite catalysts for oligomerization of olefins.

Teng Li: conceptualization, validation, data curation, formal analysis, investigation, writing – original draft, writing – review & editing. Yazeed Alfawaz: validation, data curation, formal analysis, investigation. Stefan Nastase: validation, data curation, formal analysis, investigation. Juan Carlos Navarro de Miguel: validation, data curation. Luigi Cavallo: conceptualization, supervision, writing – review & editing, investigation. Javier Ruiz-Martinez: conceptualization, supervision, writing – original draft, writing – review & editing, funding acquisition, investigation.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

Data will be made available on request.

Additional information including sample preparation, characterization, catalytic tests, and DFT calculation results are provided in the supporting information. See DOI: https://doi.org/10.1039/d6cc02736a.

Acknowledgements

Funding for this work was provided by a baseline grant (BAS/1/1402-01-01) from King Abdullah University of Science and Technology (KAUST). The authors gratefully thank Dr Daniil Nazimov, Dr Cristina Queiros da Silva and Dr Lu Song for support in characterization.

References

1. A. Corma and A. Martinez, Adv. Mater., 1995, 7, 137–144.
2. B. M. Weckhuysen and J. Yu, Chem. Soc. Rev., 2015, 44, 7022–7024.
3. J. Dědeček, E. Tabor and S. Sklenak, ChemSusChem, 2019, 12, 556–576.
4. K. Chen, A. Zornes, V. Nguyen, B. Wang, Z. Gan, S. P. Crossley and J. L. White, J. Am. Chem. Soc., 2022, 144, 16916–16929.
5. M. Bernauer, E. Tabor, V. Pashkova, D. Kaucky, Z. Sobalik, B. Wichterlova and J. Dědeček, J. Catal., 2016, 344, 157–172.
6. V. Pashkova, S. Sklenak, P. Klein, M. Urbanova and J. Dědeček, Chem. – Eur. J., 2016, 22, 3937–3941.
7. J. Dědeček, Z. Sobalik and B. Wichterlova, Catal. Rev.: Sci. Eng., 2012, 54, 135–223.
8. J. Dědeček and B. Wichterlova, J. Phys. Chem. B, 1999, 103, 1462–1476.
9. J. Dědeček, L. Capek, D. Kaucky, Z. Sobalik and B. Wichterlova, J. Catal., 2002, 211, 198–207.
10. J. R. DiIorio and R. Gounder, Chem. Mater., 2016, 28, 2236–2247.
11. S. Ernst, J. Weitkamp, J. A. Martens and P. A. Jacobs, Appl. Catal., 1989, 48, 137–148.
12. M. W. Simon, S. L. Suib and C. L. Oyoung, J. Catal., 1994, 147, 484–493.
13. Z. Cui, Q. Liu, Z. Ma, S. Bian and W. Song, J. Catal., 2008, 258, 83–86.
14. https://www.iza-structure.org/databases/.
15. A. K. Jamil and O. Muraza, Micropor. Mesopor. Mat., 2016, 227, 16–22.
16. J. Dědeček, D. Kaucký, B. Wichterlová and O. Gonsiorová, PCCP, 2002, 4, 5406–5413.
17. P. Sazama, J. Dědeček, V. Gabova, B. Wichterlova, G. Spoto and S. Bordiga, J. Catal., 2008, 254, 180–189.
18. T. Li, S. Chung, S. Nastase, A. Galilea, Y. Wang, I. Mukhambetov, M. Zaarour, J. C. N. de Miguel, J. Cazemier, A. Dokania, L. Panarone, J. Gascon, L. Cavallo and J. Ruiz-Martínez, Chem. Catal., 2023, 3, 100540.
19. M. Bernauer, E. Tabor, V. Pashkova, D. Kaucký, Z. Sobalík, B. Wichterlová and J. Dědeček, J. Catal., 2016, 344, 157–172.

---

Footnote

† These authors contribute equally to this work.

---