Temperature-programmed reduction method for stabilization of the inorganic framework of SAPO-37 materials: promising catalysts for MTBE production

Abstract

A temperature-programmed reduction method was developed to stabilize SAPO-37 and molybdenum-(oxy)-carbide-loaded SAPO-37 (SAP-37MoCR). The inorganic framework (Faujasite type SAPO-37) was retained for the first time even after treatment at 550 °C and was confirmed through powder XRD and SEM analysis. SAP-37MoCR exhibits strong acidic sites and is a promising catalyst for MTBE synthesis. Furthermore, the catalyst is stable and recyclable.

---

Zeolites and zeolite-based materials have a significant impact on science and technology, particularly in petrochemical, refining, environmental, and fine chemical processes. Large-pore rare earth-containing dealuminated zeolite Y (USY) is a commercial catalyst for the fluid catalytic cracking (FCC) process for gasoline production in the petroleum industry.¹ Similar to zeolites, silicoaluminophosphate possesses moderate acidity and strong coke resistivity, and is commercially used in the methanol-to-olefin (MTO) process using SAPO-34² and the dewaxing process using SAPO-11.³ SAPO-37 materials, similar to FCC commercial catalysts, exhibit mild acidity and coke resistivity, and are effective for various catalytic processes, such as the isomerization of n-alkanes and alkenes and isobutene/2-butene alkylation.⁴ However, their applications are limited because of their relatively poor structural stability in the presence of moisture and thermal treatment.^5,6 Several attempts have been made to stabilize SAPO-37 frameworks, including retaining TMA⁺ cations, solvent extraction of the template, and enhancing surface hydrophobicity by introducing organosilane. SAPO-37 demonstrates water resistance upon calcination at 400 °C, primarily by retaining the majority of TMA⁺ ions while selectively removing TPA⁺ ions from the faujasite structure.⁶ Calcination at 600 °C results in the formation of P–O–P bonds through the condensation of P–OH species, which are highly susceptible to hydrolysis at ambient temperature.^5,6 Partial template removal can also be achieved by solvent extraction using NaNO₃–ethanol mixtures.⁷ To the best of our knowledge, no reports exist on the structural stability of SAPO-37 after the complete removal of the template. The development of mildly acidic, coke-resistant, multifunctional, and stable SAPO-37-type materials is a topic of interest in the field of heterogeneous catalysis. Recently, transition metal carbides supported on zeolites have emerged as potential alternatives to noble metal-containing zeolites for various chemical processes.^8,9 Temperature-programmed reduction (TPR) is often adopted for the synthesis of metal carbides with large surface areas.¹⁰ Since SAPO-37 has limitations on structural stability upon removing the template, exploiting the same template for the carbon source for metal carbide synthesis will pave the way for a functional material with enhanced structural stability as well as catalytic properties.

In this regard, molybdenum carbide is an important potential coke-resistant catalyst alternative to noble metals and has become increasingly important in various chemical reactions, such as reforming⁸ and hydrogen production.⁹ Molybdenum carbide-based materials on various supports, such as alumina, silica, and zeolite surfaces, show promise for various catalytic and electrochemical processes. Therefore, to explore their potential catalytic application and advance the development of catalysts in the energy sector, it will be interesting to disperse molybdenum carbide-type species within the framework of SAPO-37 using an in situ TPR method and stabilize the framework. Herein, we report the stabilization of the SAPO-37 framework for the first time via TPR by introducing molybdenum blue precursor-type species in the form of isopolymolybdate {Mo₃₆} at pH ∼ 2,¹¹ using a post-synthetic method. The resulting stable molybdenum (oxy) carbide decorated SAPO-37 was employed to prepare methyl tertiary-butyl ether (MTBE), a fuel additive in gasoline. All the chemicals are used as received. The details of the synthesis procedure,^11–13 and characterization methods are summarized in the ESI.‡ The TPR-treated parent sample is represented as SAP-37R. SAPO-37, stabilized by the TPR method in the presence of molybdenum blue with different R/Mo ratios (2, 4, 5, and 6), are denoted as SAP-37MoC-2R, SAP-37MoC-4R, SAP-37MoC-5R, and SAP-37MoC-6R. The prepared materials were systematically characterized and subsequently explored for MTBE production (see ESI‡).

The FTIR spectra (Fig. S1, ESI‡) of the as-prepared SAPO-37 samples (SAP-37 as) exhibit vibrational bands at 533 and 565 cm⁻¹ corresponding to the double six-membered ring (D6R) of the AlPO framework, which is the secondary building unit (SBU) of faujasite-type SAPO-37. The band in the region 1100 cm⁻¹ corresponds to the asymmetric stretching frequency of T–O–T, where T – Al, P, or Si.⁷ Molybdenum blue-loaded SAPO-37 that underwent TPR exhibits a vibrational band corresponding to the D6R building unit, with a decrease in intensity owing to the filling of molybdenum species within the channels of the faujasite structure.

Fig. 1 compares the powder XRD patterns of parent SAPO-37 and those loaded with molybdenum blue dispersion having R/Mo ratios of 2, 4, 5, and 6, followed by TPR at 550 °C. The parent SAPO-37 exhibits reflections at 2θ = 6.1°, 10°, 15.5°, 18.6°, 20.4°, and 23.4°, corresponding to (111), (220), (331), (333), (440), and (533) planes, respectively, which are typical of faujasite-type SAPO-37 frameworks.⁷ The direct calcination of SAPO-37 at 450 °C results in the framework collapse owing to the breakage of P–O–P bonds in the faujasite structure.⁶ In contrast, the crystallinity of SAP-37R obtained after TPR remained intact, as evident from the XRD pattern (Fig. S2, ESI‡). Furthermore, to improve the crystallinity and catalytic activity of SAPO-37, molybdenum blue, a precursor of molybdenum carbide, was loaded with different R/Mo ratios followed by TPR at 550 °C. The XRD patterns of the samples with different R/Mo ratios (R/Mo = 2, 4, 5, and 6) exhibit a highly crystalline faujasite SAPO-37 framework. Furthermore, the XRD patterns indicate that the relative intensities of the diffraction peaks, such as (333) and (533), were considerably reduced, suggesting the presence of molybdenum species inside the SAPO-37 pores. The shift in 2θ values and significant broadening also indicate the incorporation of molybdenum species within the faujasite structure. However, the characteristic diffraction peaks of the molybdenum species were not distinguishable from the XRD data, indicating that the molybdenum species were well integrated into the SAPO-37 basic structure.

Fig. 1 Powder XRD pattern of (a) SAP-37, (b) SAP-37MoC-6R, (c) SAP-37MoC-5R, (d) SAP-37MoC-4R, and (e) SAP-37MoC-2R.

Molybdenum blue with R/Mo = 4 is the optimum ratio for the formation of monophase carbide¹⁴ and thus, for the further studies, SAP-37MoC-4R is used. The nature of the molybdenum species on SAP-37MoC-4R was determined using XPS analysis. XPS profiles for Mo 3d and C 1s are shown in Fig. 2. The Mo 3d_5/2 XPS profile shows a broad peak in the binding energy range of 228–238 eV. The deconvolution profile shows multiple peaks centered at 228.8 (FWHM 1.3), 230.1 (1.4), 231.7 (1.3) and 232.7 (1.2) eV, corresponding to the presence of molybdenum in its +2, +4, +5 and +6 states, respectively.¹⁵ The binding energies of each species are summarized in Table S1 (ESI‡). The peaks appearing at lower binding energy values (228.8 and 230.1 eV) show that significant carburization occurred in the SAPO-37 framework after TPR. Carburization was also evident from the C 1s XPS profile, which shows a broad spectrum with the main peak at a binding energy of 284.8 eV. The additional weak peaks at 283.9, 286.4 and 288.6 eV correspond to C–O–Mo bonding in (oxy) carbide form, C–O, and C=O bonding, respectively.¹⁶ The O 1s XPS spectrum shows a symmetrical peak at 532.3 eV, which can be attributed to the C–O species.

Fig. 2 XPS spectra of SAP-37MoC-4R.

Molybdenum oxide clusters present in the molybdenum blue precursor undergo partial carburization at 550 °C, resulting in the formation of molybdenum (oxy) carbide-type species (MoOC), which is also facilitated by the carbon content present in the SAPO framework. The proposed pathway for the formation of molybdenum (oxy) carbide species is shown in Scheme S1 (ESI‡). The loaded molybdate species may be present in a tetrahedral environment and interact strongly with the T atoms of the SAPO-37 framework. Subsequent heat treatment under a hydrogen atmosphere resulted in strong Mo–O–T (T – Si, Al, or P) species, which stabilized the open cage of the faujasite. This reduction facilitated the formation of molybdenum (oxy) carbide-type intermediate carbide species, which anchored the faujasite structure through strong covalent interactions, thereby promoting framework stabilization which is in agreement with the XRD results. Furthermore, the presence of molybdenum (oxy) carbide-type species is supported by the subsequent reduction of SAP-37MoC-4R at 650 °C which leads to the formation of molybdenum carbide¹⁷ in the SAPO-37 framework with short range ordering (Fig. S3, ESI‡). To understand the extent of template removal by TPR, the TGA profiles of the samples were analyzed (Fig. S4, ESI‡). The weight loss at each stage and the total weight loss for the reduced and parent samples are summarized in Table S2 (ESI‡).

The parent SAPO-37 exhibited a total weight loss of 32% (Fig. S4f, ESI‡). The first stage of weight loss (7%) below 200 °C can be attributed to physisorbed water molecules. The weight loss in the region of 200–400 °C can be attributed to the organic template TPA⁺ (16%), while above 400 °C, the weight loss corresponds to the decomposition of template TMA⁺. Fig. S4a–e (ESI‡) shows the results of the TGA of the reduced samples. The TPR treated samples exhibit a weight loss of approximately 3–5% above 200 °C, which suggests that heat treatment above 550 °C in the presence of hydrogen effectively removes the templates and converts molybdenum blue species to molybdenum (oxy) carbide. The presence of predominant physisorbed species (below 100 °C) in the TPR-treated samples indicates the removal of the template and the accessibility of the faujasite framework.

The average particle size of the parent SAPO-37 is 2–7 μm, while that of molybdenum (oxy) carbide incorporated SAPO-37 is 5–10 μm.

Furthermore, the high-resolution TEM images (Fig. 3) of SAP-37MoC-4R showed a uniform dark pattern with lattice fringes analogous to SAPO-37 and molybdenum (oxy) carbide. The electronic diffraction pattern (SAED) showed the presence of the (020) plane (0.205 nm) and (533) plane (0.36 nm) corresponding to molybdenum (oxy) carbide and Faujasite SAPO-37, respectively.^7,16 The chemical environments of Al, P, and Si were examined using solid-state MAS-NMR studies (Fig. S6, ESI‡). The ²⁹Si MAS-NMR spectra of SAP-37R and SAP-37MoC-4R reveal broad resonance peaks at −111 ppm and −115 ppm, respectively, corresponding to the presence of Si in a well-condensed tetrahedral environment.¹⁸ The additional peak appearing around −91 ppm for SAP-37R and −105 ppm for SAP-37MoC-4R corresponds to Si(3Al, 1Si) and (1Al, 3Si) environments. These results align with those of the powder XRD analysis, where TPR reduction facilitated the retention of the crystalline SAPO-37 framework. The ²⁷Al MAS-NMR spectrum of both samples shows a typical NMR signal of around 50 ppm, corresponding to the tetrahedrally coordinated framework of aluminium. An additional peak appears around 0 ppm, attributed to octahedrally coordinated aluminium derived from extra-framework aluminium, which is predominant in SAP-37MoC-4R.¹⁹ The ³¹P NMR spectrum of both samples exhibits a strong resonance at −27 ppm, corresponding to the tetrahedrally coordinated P(4Al). SAP-37MoC-4R showed an additional shoulder peak around −15 ppm, corresponding to P(Al)_x(OH₂)_y species. The robust and intense peak around −27 ppm indicates that the materials are crystalline, consistent with the powder XRD results. The incorporation of MoOC results in strong interaction with the SAPO-37 framework, leading to the broadening of the signal. The ICP-AES analysis (Table S3, ESI‡) showed a molybdenum concentration of about 5%, and the Si, Al, and P concentrations remained consistent.

Fig. 3 TEM image of the TPR-treated sample SAP-37MoC-4R.

The number and strength of the acid sites present in the TPR-treated SAPO-37 were determined via NH₃-TPD (Fig. S8, ESI‡). The TPR-treated SAPO-37 (SAP-37R) exhibits an acidity of 0.125 mmol g⁻¹, indicating weak acidity. The introduction of molybdenum species increased the surface acidity and acidic strength. An increase in the R/Mo ratio of SAPO-37 resulted in a decrease in surface acidity due to the porous carbon deposition. Among the TPR-treated SAPO-37 samples, SAP-37MoC-4R exhibits the highest surface acidity of 0.700 mmol g⁻¹. The total number of acid sites present in the TPR-treated SAPO-37 is summarized in Table S5 (ESI‡). The nature of the acid sites was further followed with the help of pyridine IR desorption studies. Fig. S9 (ESI‡) shows the absorption bands at 1540 and 1450 cm⁻¹ corresponding to pyridine adsorbed on Brønsted and Lewis acid sites, respectively.²⁰ The amount of strong Brønsted acidic sites is increased in the TPR reduced molybdenum modified SAPO-37 (SAP-37MoC-R) samples. Among the various samples, SAP-37MoC-4R showed strong Brønsted acidic sites evident from the strong pyridine sorbed peak at 1540 cm⁻¹, which was retained even after high desorption temperature (400 °C). The SAP-37MoC-6R derived from the molybdenum precursor with (R/Mo ratio of 6) showed strong Lewis acid sites, which may be due to excess R used depositing on the SAPO-37 framework.

The well-characterized SAPO-37R and SAP-37MoCR samples were used to synthesize methyl tertiary butyl ether (MTBE) from methanol (in excess) and tertiary butanol (TBA). Among these samples, SAP-37MoC-4R exhibited the highest acidity and surface area, which prompted us to optimize the reaction parameters (temperature (120–160 °C), duration (2–8 h), and methanol-to-TBA molar ratio (1 : 1 to 6 : 1)) using the SAP-37MoC-4R catalyst (Fig. S10, ESI‡). The optimal conditions for enhanced activity were found to be at 150 °C for 6 h with a methanol-to-TBA ratio of 3 : 1, resulting in an 83% TBA conversion with exclusive formation of MTBE. Unlike other reported catalysts (Table S6, ESI‡),¹¹ here in all cases, MTBE was the primary product, and only trace amounts of dimethyl ether (DME) and isobutene dimer were identified. The use of long chain alcohols (ethanol, propanol, and butanol) resulted in a steady decrease in the conversion, which may be attributed to steric factors. Subsequently, the reactions were examined using various SAPO-37 catalysts: SAPO-37R, SAP-37MoC-2R, SAP-37MoC-5R, and SAP-37MoC-6R. The activity was directly proportional to the concentration of Brønsted acidic sites and the surface area of the catalyst. Strong Brønsted acid sites and high surface area of SAP-37MoC-4R facilitated enhanced TBA conversion (83%) with exclusive MTBE formation. On the other hand, SAP-37MoC-6R, which has strong Lewis acidity with high surface area, showed a comparable activity of 70% TBA conversion. Furthermore, the catalytic activity remained intact for several cycles with a slight reduction in conversion (Fig. 4) due to fouling at the active sites.

Fig. 4 Effects of different SAPO-37-based catalysts and their recyclability in MTBE synthesis.

The nature of the catalyst after the reaction was determined through TGA, powder XRD, nitrogen sorption and TEM analyses (Fig. S11 and S12, ESI‡). Even after four cycles, the catalyst retained the faujasite framework, as evident from the powder XRD analysis. The TGA profile of the spent catalyst indicates a total weight loss of 7% (250–800 °C) owing to fouling of the reactant molecule and is responsible for the reduction in activity after four cycles. The catalyst regenerated by TPR treatment exhibits a TGA profile similar to that of the parent catalyst. The nitrogen sorption isotherm of the spent catalyst shows similar sorption behavior with a slight decrease in surface area, indicating that the catalyst remained stable over the catalytic cycles. Attempts were made to incorporate other metal ions, such as Ti, Zr, etc., and stabilize the SAPO-37 framework. The textural properties showed that the crystallinity remained intact after TPR treatment at 550 °C. The resultant materials are also active for MTBE preparation.

This is the first study to investigate the stabilization of SAPO-37 using the TPR method. The introduction of molybdenum (oxy) carbide-type species during TPR stabilization facilitated the introduction of strong acidic sites. Powder XRD and NH₃-TPD analyses confirmed the presence of a faujasite structure with strongly acidic sites. The resulting SAPO-37 materials showed great potential for MTBE synthesis under ambient conditions, and their activity remained stable.

Prof. A. Sakthivel (supervisor): conceptualization, methodology, formal analysis, writing, editing, funding acquisition. K. K. Shabana, N. P. Nimisha, Dr B. N. Soumya: methodology, formal analysis, data curation, writing, editing, Dr N. J. Venkatesha, and G. Sheetal: data curation and analysis.

The authors thank DST-SERB-CRG/2023/001107, CSIR 09/1108(15575)/2022-EMR-I, KSHECA3/344/Govt.Kerala-NKPDF/2022, and DST/INSPIRE/03/2019/000097 for financial assistance.

Data availability

All relevant data generated and analyzed during this study, which include experimental, spectroscopic, and analytical data, are included in this article and its ESI.‡

Conflicts of interest

There are no conflicts to declare.

References

1. A. Primo and H. Garcia, Chem. Soc. Rev., 2014, 43, 7548.
2. M. S. Ahmad, C. K. Cheng, P. Bhuyar, A. E. Atabani, A. Pugazhendhi, N. T. L. Chi, T. Witoon, J. W. Lim and J. C. Juan, Fuel, 2021, 283, 118851.
3. P. Zhang, H. Liu, Y. Yue, H. Zhu and X. Bao, Fuel Process. Technol., 2018, 179, 72.
4. H. B. Mostad, M. Stöcker, A. Karlsson and T. Rørvik, Appl. Catal., A, 1996, 144, 305.
5. G. N. Kalantzopoulos, F. Lundvall, K. Thorshaug, A. Lind, P. Vajeeston, I. Dovgaliuk, B. Arstad, D. S. Wragg and H. Fjellvåg, Chem. Mater., 2020, 32, 1495.
6. T. T. H. Dang, M. Bartoszek, M. Schneider, D. L. Hoang, U. Bentrup and A. Martin, Appl. Catal., B, 2012, 121, 115–122.
7. R. Yadav, M. Ahmed, A. K. Singh and A. Sakthivel, Sci. Rep., 2016, 6, 22813.
8. M. D. Porosoff, X. Yang, J. A. Boscoboinik and J. G. Chen, Angew. Chem., Int. Ed., 2014, 53, 6705.
9. Y. Ma, G. Guan, X. Hao, J. Cao and A. Abudula, Renewable Sustainable Energy Rev., 2017, 75, 1101.
10. S. T. Oyama, J. C. Schlatter, J. E. Metcalfe III and J. M. Lambert Jr, Ind. Eng. Chem. Res., 1988, 27, 1639.
11. M. A. Myachina, N. N. Gavrilova and V. V. Nazarov, Russ. J. Phys. Chem. A, 2018, 92, 2237–2241.
12. L. Sierra de Saldarriaga, C. Saldarriaga and M. E. Davis, J. Am. Chem. Soc., 1987, 109, 2686.
13. J. F. Knifton and J. C. Edwards, Appl. Catal., A, 1999, 183, 1.
14. N. Gavrilova, M. Myachina, V. Nazarov and V. Skudin, Nanomaterials, 2021, 11, 873.
15. W. Gruenert, A. Y. Stakheev, R. Feldhaus, K. Anders, E. S. Shpiro and K. M. Minachev, J. Phys. Chem., 1991, 95, 1323.
16. P. Delporte, C. Pham-Huu, P. Vennegues, M. J. Ledoux and J. Guille, Catal. Today, 1995, 23, 251.
17. Y. Zheng, Y. Tang, J. R. Gallagher, J. Gao, J. T. Miller, I. E. Wachs and S. G. Podkolzin, J. Phys. Chem. C, 2019, 123, 22281–22292.
18. C. Blackwell and R. Patton, J. Phys. Chem., 1988, 92, 3965.
19. Z. Li, J. Martínez-Triguero, P. Concepción, J. Yu and A. Corma, Phys. Chem. Chem. Phys., 2013, 15, 14670.
20. H. Song, J. Wang, Z. Wang and Z. Jin, J. Catal., 2014, 311, 257–265.

---

Footnotes

† This work is dedicated to Prof. Y. Sugi, Gifu University, Japan, for his 82nd birthday.

‡ Electronic supplementary information (ESI) available: Experimental details, Fig. S1–S12, Scheme S1 and Tables S1–S6 are available. See DOI: https://doi.org/10.1039/d4cc01839g

---