Lindqvist@Nanoporous MOF-Based Catalyst for Effective Desulfurization of Fuels

An effective and sustainable oxidative desulfurization process for treating a multicomponent model fuel was successfully developed using as a heterogeneous catalyst a composite material containing as an active center the europium Lindqvist [Eu(W5O18)2]9− (abbreviated as EuW10) encapsulated into the nanoporous ZIF-8 (zeolitic imidazolate framework) support. The EuW10@ZIF-8 composite was obtained through an impregnation procedure, and its successful preparation was confirmed by various characterization techniques (FT-IR, XRD, SEM/EDS, ICP-OES). The catalytic activity of the composite and the isolated EuW10 was evaluated in the desulfurization of a multicomponent model fuel containing dibenzothiophene derivatives (DBT, 4-MDBT and 4,6-DMDBT) with a total sulfur concentration of 1500 ppm. Oxidative desulfurization was performed using an ionic liquid as extraction solvent and aqueous hydrogen peroxide as oxidant. The catalytic results showed a remarkable desulfurization performance, with 99.5 and 94.7% sulfur removal in the first 180 min, for the homogeneous active center EuW10 and the heterogeneous EuW10@ZIF-8 catalysts, respectively. Furthermore, the stability of the nanocomposite catalyst was investigated by reusing and recycling processes. A superior retention of catalyst activity in consecutive desulfurization cycles was observed in the recycling studies when compared with the reusing experiments. Nevertheless, the nanostructure of ZIF-8 incorporating the active POM (polyoxometalate) was shown to be highly suitable for guaranteeing the absence of POM leaching, although structural modification was found for ZIF-8 after catalytic use that did not influenced catalytic performance.


Introduction
The combustion of transportation fuels has been one of the biggest global concerns in the last decades, due to environmental and health risks. In particular, sulfur compounds present in fuels can cause equipment corrosion, catalyst deactivation and, more importantly, the emission of sulfur oxides (SOx) [1,2]. The dangers associated with the emission of SOx, such as air pollution and acid rain, have led to the implementation of restrictive legislation limiting the sulfur content in commercialized transportation fuels to 10-15 ppm in Europe and the USA [3,4]. Hydrodesulfurization (HDS) is the conventional method used in refineries, because it is very effective in the removal of thiols, sulfides and dissulfides. However, it requires high temperatures (300-400 • C) and high hydrogen pressure to achieve the production of ultra-low-sulfur fuels [5,6]. The removal of certain sulfur compounds, such as benzothiophene derivatives, is more difficult to achieve by HDS, demanding even harsher conditions [7][8][9]. Therefore, alternative and/or complementary methods to current desulfurization technology have been developed. Oxidative desulfurization (ODS) is one of the most promising methods, reconciling sustainability with cost-efficiency when compared [1,4,11]. Oxidative desulfurization occurs in two main steps: extraction and cataly dation [2]. In the first step, a polar solvent is required in order to extract the sulf pounds from the fuel [3,6,[12][13][14]. Ionic liquids have demonstrated high desulfu efficiency for extracting sulfur compounds from fuels and also for promoting hi fective catalytic oxidation [13][14][15][16][17][18][19]. However, the main key to achieving ultra-low fuels is the adequate choice of catalyst and oxidant with which to oxidize the sulf pounds into sulfoxides and sulfones [20][21][22][23][24]. Hydrogen peroxide has been widely ODS due to its low environmental impact, since the only resulting sub-product i [13,14,[25][26][27][28][29][30][31][32][33].
The high solubility of POMs in polar solvents makes for difficult recovery an cling of catalysts. The immobilization of active POM centers in solid porous su such as metal-organic frameworks (MOFs), allows the heterogenization of POMs reusability [44,45]. MOFs are hybrid materials obtained from the covalent connec tween inorganic metal nodes using organic linkers [46]. ZIF-8 (zeolitic imidazolate work) is a widely studied MOF with structural simplicity, ease of synthetic repro ity, and potential for scale-up [47,48]. The existence of pores of adequate internal d (11.6 Å) and aperture (3.4 Å) allows the encapsulation of guests in the cavities wh venting leaching from the cavities [33,49,50].
In previous works from our research group, various Keggin-type POMs wer porated into different porous MOFs to prepare active heterogeneous catalysts fo tive desulfurization [1,30,33,39,46,[51][52][53][54][55][56][57][58]. To the best of our knowledge, no report found in the literature presenting Lindqvist-type isopolyanions incorporated int supports for use as heterogeneous catalysts. The use of Lindqvist-type materials europium metal center supports the preparation of a more stable bridged-polyo alate, which incorporates a high number of tungsten octahedrons, promoting the tion of the oxidant, and thus increasing the desulfurization efficiency of the cataly work reports the successful preparation of EuW10 isopolyanion encapsulated in an priate MOF structure (Scheme 1), the Zeolitic Imidazolate Framework, ZIF-8, and plication of the novel composite EuW10@ZIF-8 to desulfurize a multicomponent fue sustainable conditions. Scheme 1. Representation of the preparation of the EuW10@ZIF-8 composite material. Ro perature is abbreviated as rt. Scheme 1. Representation of the preparation of the EuW 10 @ZIF-8 composite material. Room temperature is abbreviated as rt. Fourier-transformed infrared (FT-IR) spectra were acquired using the attenuated total reflectance (ATR) operation mode of a PerkinElmer FT-IR System Spectrum BX spectrometer, and all the representations are shown in arbitrary units of transmittance. Elemental analysis for W was performed by ICP-OES on a Perkin-Elmer Otima 4300 DV instrument at the University of Santiago de Compostela, Spain. Powder X-ray diffraction (PXRD) patterns were obtained at room temperature on a Rigaku Geigerflex diffractometer operating with a Cu radiation source (λ 1 = 1.540598 Å; λ 2 = 1.544426 Å; λ 1 /λ 2 = 0.500) and in a Bragg-Brentano θ/2θ configuration (45 kV, 40 mA) at the Analytical Laboratory of the Faculty of Sciences and Technology of the NOVA University Lisbon, Portugal. Intensity data were collected by a step-counting method (step 0.026 • ) in continuous mode in the 3 ≤ 2θ ≤ 50 • range, and all representations are shown in arbitrary units of intensity. Scanning electron microscopy (SEM) and electron dispersive X-ray spectroscopy (EDS) analysis were performed at "Centro de Materiais da Universidade do Porto" (CEMUP, Porto, Portugal). The studies were performed in an FEI Quanta 400 FEG ESEM high-resolution scanning electron microscope equipped with an EDAX Genesis X4M spectrometer working at 15 kV. Samples were coated with an Au/Pd thin film by sputtering using SPI Module Sputter Coater equipment. Catalytic reactions were periodically monitored by GC-FID analysis carried out in a Bruker 430-GC-FID chromatograph. Hydrogen was used as carrier gas (55 cm·s −1 ) and fused silica Supelco capillary columns SPB-5 (30 m × 0.25 mm i.d.; 25 µm film thickness) were used. The ZIF-8 was prepared following an experimental procedure adapted from previously reported methods [61]. An initial solution of Zn(NO 3 ) 2 ·6H 2 O (900 mg, 3 mmol) in MeOH (30 mL) was magnetically stirred for 15 min. Then, a solution of 2-methylimidazole (2-MIM, 995 mg, 12 mmol) in MeOH (10 mL) was slowly added. The resulting reactional mixture was magnetically stirred for 24 h at ambient temperature, after which the resulting material was isolated by centrifugation, washed three times with MeOH, and dried at 60 • C under vacuum for 16 h.

Materials Preparation
The preparation of the composite was performed through the immobilization of the EuW 10 in the porous ZIF-8 support through an impregnation procedure for POM@MOF composites [52,53,62]. The procedure consisted of the preparation of an ethanolic solution of EuW 10 (90.2 mg; 0.031 mmol), followed by the addition of previously dried ZIF-8 (150 mg) and stirring of the mixture for 96 h at room temperature. The product was recovered by centrifugation, washed with MeOH, and dried in a desiccator over silica gel.

Oxidative Desulfurization
Desulfurization studies were performed with a model fuel prepared with DBT, 4-MDBT and 4,6-MDBT, in decane (500 ppm of each sulfur compound). The reactions were carried out under air in a closed borosilicate vessel with a magnetic stirrer and immersed in a thermostatically controlled liquid paraffin bath at 70 • C. Catalytic oxidative reactions were performed in a biphasic system with a 1:1 model fuel/extraction solvent with the ionic liquid [BMIM]PF 6 . In a representative experiment, a certain amount of the heterogeneous catalyst equivalent to 3 µmol of active POM (EuW 10 ) was added to [BMIM]PF 6 . After adding the model fuel, the system was stirred for 10 min at 70 • C. The oxidative catalytic step was then initiated with the addition of aqueous H 2 O 2 30% (75 µL) to the reaction mixture. Tetradecane was used as a standard in the periodical monitorization of the sulfur content by GC analysis.

Materials Characterization
The resulting EuW 10 @ZIF-8 composite material was characterized using several techniques, including Fourier transform infrared spectroscopy (FT-IR), powder XRD, thermogravimetric analysis (TGA), inductively coupled plasma-optical emission spectrometry (ICP-OES) and scanning electron microscopy (SEM) coupled with energy dispersive X-ray spectroscopy (EDS). The FT-IR spectrum of EuW 10 @ZIF-8 ( Figure 1) exhibits the typical bands arising from the ZIF-8 framework, namely the bands associated with the stretching modes of the C=N and C-N bonds located 1577 and 1144 cm −1 , respectively. The bands located at 993 and 752 cm −1 are ascribed to the bending modes of C-N and C-H bonds, respectively, while the band at 692 cm −1 arises from the out-of-plane bending of the ring. The band located at 420 cm −1 is assigned to the stretching vibration of the Zn-N confirming the coordination between the metal center and organic ligand [63,64]. The presence of additional bands can be observed in the spectrum of the composite when compared with the spectrum of ZIF-8. These bands are located at 937 cm −1 , which can be assigned to the terminal ν as (W-O d ) stretching mode, as well as at 887 and 835 cm −1 , ascribed to the corner-sharing ν(W-O b -W) stretching modes of EuW 10 , respectively [60,[65][66][67]. Moreover, the presence of the europium POM in the composite material was further confirmed by elemental analysis (ICP-OES). The results reveal a tungsten amount of 28.6 wt%, leading to a EuW 10 loading of 0.156 mmol per g, which confirms the successful immobilization of the POM onto the porous ZIF-8 structure.
Nanomaterials 2022, 12, x FOR PEER REVIEW 5 of 15 a EuW10 loading of 0.156 mmol per g, which confirms the successful immobilization of the POM onto the porous ZIF-8 structure. The crystalline structures of EuW10@ZIF-8 composite and the solid support were investigated by powder XRD. Figure 2 exhibits the patterns of both materials in the 5 < 2θ < 50° range. The results reveal highly crystalline materials with nearly identical diffraction patterns composed by the characteristic diffraction peaks of the ZIF-8 crystalline structure. The main diffraction peaks located at 2θ = 7.4, 10.5, 12.8, 14.9, 16.6, 18.2, 24.7 and 26.8° can be indexed to the (011), (002), (112), (022), (013), (222), (233) and (134) reflections. The high similarity between the diffraction patterns points out to the stability of the MOF by retaining its crystalline structure after the EuW10 incorporation process. Although an additional low-intensity peak can be detected at 2θ = 8.7°, the main diffraction peaks of EuW10 are absent from the EuW10@ZIF-8 pattern, suggesting its homogenous distribution within the MOF framework [68,69]. The high similarity between the diffraction patterns points out to the stability of the MOF by retaining its crystalline structure after the EuW 10 incorporation process. Although an additional low-intensity peak can be detected at 2θ = 8.7 • , the main diffraction peaks of EuW 10 are absent from the EuW 10 @ZIF-8 pattern, suggesting its homogenous distribution within the MOF framework [68,69].
The morphology and chemical composition of the MOF support and the EuW 10 @ZIF-8 composite material were studied by SEM/EDS techniques. The SEM image of ZIF-8 ( Figure 3A) reveals uniform particles with a morphology of rhombic dodecahedrons, which is in good agreement with the reported literature [70][71][72]. SEM images were used to estimate the average particle size (ImageJ v1.53e; https://imagej.nih.gov/ij/; accessed on 5 February 2022; 50 measurements), leading to a mean particle size of 554 ± 20 nm. The EDS spectrum of ZIF-8 ( Figure S1A) shows the presence of zinc as the main constituent as well as nitrogen from the organic ligand (2-methylimidazole). The SEM micrograph obtained for the EuW 10 @ZIF-8 composite material ( Figure 3B) also exhibits rhombic dodecahedronshaped particles with an average particle size of 567 ± 16 nm, suggesting that the MOF structure remained intact in the final composite. The EDS spectrum ( Figure S1B) shows, besides the Zn and N elements from the ZIF-8 support, the presence of W which further confirms the successful immobilization of EuW 10 in the composite material. The morphology and chemical composition of the MOF support and the 8 composite material were studied by SEM/EDS techniques. The SEM image ure 3A) reveals uniform particles with a morphology of rhombic dodecahe is in good agreement with the reported literature [70][71][72]. SEM images were mate the average particle size (ImageJ v1.53e; https://imagej.nih.gov/ij/; acc February 2022; 50 measurements), leading to a mean particle size of 554 ± 20 spectrum of ZIF-8 ( Figure S1A) shows the presence of zinc as the main const as nitrogen from the organic ligand (2-methylimidazole). The SEM microgr for the EuW10@ZIF-8 composite material ( Figure 3B) also exhibits rhombic do shaped particles with an average particle size of 567 ± 16 nm, suggesting t structure remained intact in the final composite. The EDS spectrum (Figure besides the Zn and N elements from the ZIF-8 support, the presence of W w confirms the successful immobilization of EuW10 in the composite material.  The morphology and chemical composition of the MOF support and the EuW10@ZIF-8 composite material were studied by SEM/EDS techniques. The SEM image of ZIF-8 (Figure 3A) reveals uniform particles with a morphology of rhombic dodecahedrons, which is in good agreement with the reported literature [70][71][72]. SEM images were used to estimate the average particle size (ImageJ v1.53e; https://imagej.nih.gov/ij/; accessed on 5th February 2022; 50 measurements), leading to a mean particle size of 554 ± 20 nm. The EDS spectrum of ZIF-8 ( Figure S1A) shows the presence of zinc as the main constituent as well as nitrogen from the organic ligand (2-methylimidazole). The SEM micrograph obtained for the EuW10@ZIF-8 composite material ( Figure 3B) also exhibits rhombic dodecahedronshaped particles with an average particle size of 567 ± 16 nm, suggesting that the MOF structure remained intact in the final composite. The EDS spectrum ( Figure S1B) shows, besides the Zn and N elements from the ZIF-8 support, the presence of W which further confirms the successful immobilization of EuW10 in the composite material.  The thermal stabilities of the support and the EuW 10 @ZIF-8 composite material were evaluated by thermogravimetric analysis (TGA) up to 1000 • C. In the TGA obtained for ZIF-8 ( Figure S2), it is possible to observe a mass loss (11.8%) in the range 150-400 • C, which is most likely due to the evaporation of solvent molecules of methanol from inside the pores. Nevertheless, the most pronounced mass loss (69.7%) starts at ca. 550 • C and can be attributed to the decomposition of the organic ligand (2-methylimidazole), thus resulting in the collapse of the structure [73].
The TGA of EuW 10 @ZIF-8 ( Figure S2) reveals a superior thermal stability of the composite when compared with the ZIF-8 support. In fact, the first mass loss (4.4%) is only detected at approximately 390 • C which should correspond to the evaporation of adsorbed solvent molecules. Afterwards, the main mass loss (51.8%) occurs at ca. 590 • C exhibiting a similar profile to the step observed in the ZIF-8 support and can be assigned to the degradation of the organic ligand together with the EuW 10 anions.

Catalytic Desulfurization Studies
The catalytic performance of the composite EuW 10 @ZIF-8 was investigated for the desulfurization of a multicomponent model fuel prepared with the most representative refractory sulfur compounds in real diesel, namely dibenzothiophene (DBT), 4-methyldibenzothiophene (4-MDBT) and 4,6-dimethyldibenzo-thiophene (4,6-DMDBT). Catalytic studies were performed at 70 • C in a biphasic liquid-liquid system based on equal volumes of the extraction solvent 1-butyl-3-methylimidazole hexafluorophosphate ([BMIM]PF 6 ) and the model fuel. Aqueous hydrogen peroxide was used as oxidizing agent. The desulfurization process occurred in two main steps: extraction and catalytic oxidation. During the first 10 min and before the addition of oxidant, an initial extraction occurs to transfer certain amount of the sulfur compounds from fuel to the [BMIM]PF 6 phase. The initial extraction equilibrium is reached, and the second step is initiated by adding hydrogen peroxide oxidant to the system. The metal-catalyzed oxidation of the extracted sulfur compounds is initiated (i.e., the oxidation of sulfur compounds to the corresponding sulfoxides and/or sulfones presented in the extraction solvent). The sulfur oxidation promotes a continued sulfur extraction from the diesel phase by decreasing the concentration of non-oxidized sulfur compounds in the extractant phase since no oxidation products were detected in the model diesel phase. The mechanism involved in the oxidation of benzothiophene and dibenzothiophene derivative compounds involved an initial formation of peroxo-POM active species via nucleophilic attack of the oxidant on the W VI atoms of EuW 10 . Oxidation of sulfur compounds into sulfoxides occurs by transfer of an oxygen atom from the intermediate active species, which are then regenerated into the initial W VI species. Subsequent nucleophilic attack on sulfoxides leads to the formation of the corresponding sulfones [9,25,27,41,43]. The desulfurization efficiency was determined by the periodical analysis of the model fuel phase by gas chromatography. Pristine ZIF-8 was tested as catalyst showing no activity in the studied reaction. Initially, the catalytic activity of the EuW 10 @ZIF-8 composite was compared with the homogeneous EuW 10 ( Figure 4). The desulfurization profile reveals a considerable extraction of sulfur compounds during the initial extraction step (50.2% and 68.3% of desulfurization for the heterogeneous EuW 10 @ZIF-8 and the homogeneous EuW 10 , respectively). Both catalysts were also demonstrated to be highly efficient for the desulfurization of the sulfur multicomponent model fuel. After 3 h of the process, near-complete desulfurization was achieved using the EuW 10 @ZIF-8 composite (94.7%). The homogeneous EuW 10 was shown to be slightly more efficient than the composite, since after only 1 h, a desulfurization of 98.9% was achieved. The individual desulfurization for each sulfur compound in the presence of EuW 10 @ZIF-8 is depicted in Figure 5. The difficulty to remove the sulfur compounds from the model fuel increases according to the following order: DBT > 4-MDBT > 4,6-DMDBT, which is in good agreement with previously reported studies [4,74,75]. This tendency is correlated with the steric hindrance introduced by the methyl groups that hinders the nucleophilic attack, delaying the oxidation of the sulfur compounds to sulfoxides and/or sulfones. In fact, the removal of DBT from the model fuel was greatly promoted, with complete desulfurization being achieved after only 2 h. As for 4-MDBT, nearly total desulfurization (97.2%) was achieved after three hours of reaction, while only 87.9% was achieved for 4,6-DMDBT. Despite the superior catalytic activity of the homogeneous catalyst, a major drawback is its inability to be removed from the reactional medium and recycled in consecutive catalytic cycles.
with complete desulfurization being achieved after only 2 h. As for 4-MDB desulfurization (97.2%) was achieved after three hours of reaction, while on achieved for 4,6-DMDBT. Despite the superior catalytic activity of the hom alyst, a major drawback is its inability to be removed from the reactional recycled in consecutive catalytic cycles.    and/or sulfones. In fact, the removal of DBT from the model fuel was greatly pro with complete desulfurization being achieved after only 2 h. As for 4-MDBT, nea desulfurization (97.2%) was achieved after three hours of reaction, while only 87. achieved for 4,6-DMDBT. Despite the superior catalytic activity of the homogene alyst, a major drawback is its inability to be removed from the reactional medi recycled in consecutive catalytic cycles.  The recycling capacity of the solid EuW 10 @ZIF-8 catalyst and the reuse ability of the EuW 10 @ZIF-8/[BMIM]PF 6 system were evaluated by performing five consecutive desulfurization cycles. The reuse method consisted of the preservation of the composite and the ionic liquid-phase EuW 10 @ZIF-8/[BMIM]PF 6 during five consecutive cycles. The sustainability of this procedure is higher, since the same portion of extraction solvent and catalyst were used in these consecutive cycles. After each cycle, the model fuel was removed from the system and a new desulfurization cycle was performed by adding a novel portion of sulfurized model fuel as well the H 2 O 2 oxidant, maintaining all the experimental conditions. In the case of the recycling procedure, the solid catalyst was recovered, washed with water and acetonitrile, dried, weighed, and used again in a new desulfurization cycle while maintaining the same experimental conditions. Figure 6 displays the results obtained after 4 h for the reusing and recycling processes. It is noteworthy that the recycling process of EuW 10 @ZIF 8 demonstrates the retention of the desulfurization efficiency during the various cycles, with near complete desulfurization after 4 h. On the other hand, the reuse process of the EuW 10 @ZIF 8 /[BMIM]PF 6 system presented a decrease in catalytic efficiency after the first cycle. Between the 2nd and the 4th cycles, the reuse process caused an accumulation of oxidized products (mainly sulfones) in the [BMIM]PF 6 extraction phase. This caused a decrease in sulfur transfer from the fuel to the ionic liquid phase, and consequently a decrease in the desulfurization efficiency. This behavior has been attributed to the accumulation and consequent saturation of the extraction phase during the reuse process with the oxidized sulfur compounds (sulfones), which prevents further transfer of sulfur compounds from the non-polar phase (model diesel) to the ionic liquid phase, decreasing the desulfurization efficiency of the system [27,76]. After the 4th cycle, the solid catalyst was recovered, washed, and used in a new desulfurization cycle. This promoted an increase in desulfurization efficiency, which was observed for the 5th cycle of the reuse process. Therefore, the results show that the saturation of the extraction phase prevents the reutilization of the solid catalyst for longer desulfurization cycles, since the oxidative reaction should occur in the extraction phase (sulfones and/or sulfoxides were observed in the model fuel phase), and a clean medium promotes a higher diffusion of reactants to the catalytic active center located at the ZIF-8 framework.
Nanomaterials 2022, 12, x FOR PEER REVIEW 9 The recycling capacity of the solid EuW10@ZIF-8 catalyst and the reuse ability o EuW10@ZIF-8/[BMIM]PF6 system were evaluated by performing five consecutive desu rization cycles. The reuse method consisted of the preservation of the composite and ionic liquid-phase EuW10@ZIF-8/[BMIM]PF6 during five consecutive cycles. The susta bility of this procedure is higher, since the same portion of extraction solvent and cat were used in these consecutive cycles. After each cycle, the model fuel was removed f the system and a new desulfurization cycle was performed by adding a novel portio sulfurized model fuel as well the H2O2 oxidant, maintaining all the experimental co tions. In the case of the recycling procedure, the solid catalyst was recovered, washed water and acetonitrile, dried, weighed, and used again in a new desulfurization c while maintaining the same experimental conditions. Figure 6 displays the results tained after 4 h for the reusing and recycling processes. It is noteworthy that the recyc process of EuW10@ZIF8 demonstrates the retention of the desulfurization efficiency du the various cycles, with near complete desulfurization after 4 h. On the other hand reuse process of the EuW10@ZIF8/[BMIM]PF6 system presented a decrease in catalytic ciency after the first cycle. Between the 2nd and the 4th cycles, the reuse process cau an accumulation of oxidized products (mainly sulfones) in the [BMIM]PF6 extrac phase. This caused a decrease in sulfur transfer from the fuel to the ionic liquid phase, consequently a decrease in the desulfurization efficiency. This behavior has been tributed to the accumulation and consequent saturation of the extraction phase during reuse process with the oxidized sulfur compounds (sulfones), which prevents fur transfer of sulfur compounds from the non-polar phase (model diesel) to the ionic li phase, decreasing the desulfurization efficiency of the system [27,76]. After the 4th c the solid catalyst was recovered, washed, and used in a new desulfurization cycle. promoted an increase in desulfurization efficiency, which was observed for the 5th c of the reuse process. Therefore, the results show that the saturation of the extraction p prevents the reutilization of the solid catalyst for longer desulfurization cycles, since oxidative reaction should occur in the extraction phase (sulfones and/or sulfoxides w observed in the model fuel phase), and a clean medium promotes a higher diffusio reactants to the catalytic active center located at the ZIF-8 framework. The mechanism involved in the oxidation of benzothiophene and dibenzothiophene derivative compounds catalyzed by POMs using H 2 O 2 as oxidant has been well described in the literature [9,52,75,77]. The initial formation of peroxo-POM active species occurs by nucleophilic attack of the oxidant on the W VI atoms of EuW 10 . Oxidation of sulfur compounds into sulfoxides occurs by transfer of an oxygen atom from the intermediate active species, which are then regenerated into the initial W VI species. Subsequent nucleophilic attack on sulfoxides lead to the formation of the corresponding sulfones.

Comparison with Other Catalysts
Recently, our research group reported the desulfurization of a multicomponent model fuel using homogeneous Lindqvist POMs (including EuW 10 ) and compared their activities with Keggin-type POM (Table 1) [43]. The superior efficiency of the Lindqvist-type structure was observed, requiring a smaller oxidant dosage to reach complete desulfurization. In the present work, the heterogenization of the active Lindqvist EuW 10 was strategically envisaged through its encapsulation into the framework of ZIF-8. The selection of this MOF as a host for the POM anion was due to the match between POM dimensions (near 10 Å) and the ZIF-8 pore size (11.6 Å), with a small pore window (3.4 Å) [78]. The small pore window prevents POM leaching, and at the same time can promote the diffusion of reactants towards the EuW 10 catalytic active center. Scarce work can be found in the literature reporting POMs@ZIF-8 catalysts for oxidative desulfurization, and the majority only reports their application in single-component DBT model fuel [50,79]. The only work presenting the application of a POM@ZIF-8 catalyst using a multicomponent model fuel was reported by our research group, and involved the incorporation of an imidazolium-based Keggin-type polyoxomolybdate ([BMIM] 3 PMo 12 O 40 ) into a ZIF-8 framework ([BMIM] 3 PMo 12 O 40 @ZIF-8) [33]. Under identical experimental conditions, near-complete desulfurization was achieved after 1 h; however, a structural alteration was observed after the first use of the catalyst. For this reason, the stability of the EuW 10 @ZIF-8 composite was investigated by using it in consecutive desulfurization cycles.

Catalyst Material Stability
The stability of the heterogeneous catalyst was evaluated by recovering the composite after catalytic use (EuW 10 @ZIF-8-ac). The catalyst was separated by centrifugation, washed with water and acetonitrile, and dried in a desiccator over silica gel. The powder XRD characterization of EuW 10 @ZIF-8-ac reveals that the composite undergoes structural transformation after catalytic use ( Figure S3). In fact, the ZIF-8 is known to undergo thermally and pressure-induced phase transformations as previously reported by our research group and others [33,80]. In this case, the main diffraction peaks in the powder XRD pattern of EuW 10 @ZIF-8-ac suggest the structural transformation of ZIF-8 into the ZIF-L polymorph as a consequence of the applied experimental conditions [81,82]. The morphology and chemical composition of EuW 10 @ZIF-8-ac were assessed using SEM/EDS techniques ( Figure S4). The SEM images showed a higher degree of aggregation between the parti-cles, which was most likely due to remaining solvent with high viscosity in the recovered material, as suggested by the presence of P and F from [BMIM]PF 6 in the EDS spectrum. The morphology of the particles also seems to have changed after catalytic use, most likely as a consequence of the previously discussed phase transformation. Nevertheless, the EDS spectrum of EuW 10 @ZIF-8-ac still exhibits the presence of Zn, N and W, with similar relative intensities to the spectrum of the as-prepared composite. This result, together with the retention of the high desulfurization results along the recycling studies suggests the minimal occurrence or even absence of EuW 10 leaching.

Conclusions
The heterogenization of the active Lindqvist isopolyoxometalate EuW 10 was performed in this work following an impregnation procedure to prepare the EuW 10 @ZIF-8 nanocomposite. The oxidative catalytic performance of this nanocomposite was compared with the homogeneous Lindqvist [Eu(W 5 O 18 ) 2 ] 9− (abbreviated as EuW 10 ) for the desulfurization of a multicomponent model fuel. Only a slightly superior catalytic efficiency of the homogeneous was observed during the first 2 h of the process and practically complete desulfurization was achieved after 3 h. The reutilization and recycling processes of the solid catalyst were investigated over five consecutive desulfurization cycles. Due to the high saturation of the [BMIM]PF 6 extraction phase after the first cycle, a decrease in desulfurization efficiency was detected. However, the activity was maintained for a number of consecutive cycles if the solid catalyst was recycled by separation and solvent washing was performed between cycles. Stability was further confirmed by different characterization techniques. The nanoporous ZIF-8 was shown to be highly suitable for promoting the high catalytic activity of the Lindqvist active center.