Lewis Acid-Catalyzed Ring-Opening Alcoholysis of Cyclohexene Oxide: The Role of Open Metal Sites in the Bi(III)-based Metal-Organic Framework SU-101

SU-101 was screened for the acid-catalyzed ring-opening alcoholysis of cyclohexene oxide. Results indicated access to open metal sites within SU-101, a fundamental requirement (Lewis acid Bi + 3 sites) for this reaction. In addition, SU-101 exhibited high chemical stability, demonstrated by retaining its crystalline structure after the reaction. The cyclohexene conversion was estimated to be 99.8, 96.8, and 14.3% at 40 ° C for methanol, ethanol, and propanol, respectively. Also, SU-101 demonstrated an outstanding catalytic cyclability performance for five cycles without losing catalytic activity.


Introduction
In the last century, heterogeneous catalysis has emerged as a promising route to accelerate different chemical conversions with the advantage of facile material separation (solid-state catalyst). [1] Commonly, solid-state catalysts are classified as metals and alloys, oxides and sulfides, and metal oxides and solid acids. [2] Metal complexes are widely used for hydrogenation (reduction) reactions, [3] and metal/sulfide oxides are employed for desulfurization reactions. [4] A common strategy to improve the catalytic activity of these refers to the active phases (metals) carefully deposited on the surface of porous matrices (supports), which considerably increases the catalyst's homogeneity, dispersion, and surface area. [5] Alumina (Al 2 O 3 ) and silica (SiO 2 ) are typically applied for this purpose, and these can also modify the pH of the catalyst. [6] Furthermore, based on the alumina/silica properties, different porous materials have been introduced into the field of catalysis. Zeolites have demonstrated relevant catalytic properties, particularly in the oil industry. However, these shown some drawbacks, such as deactivation (poisoning) and difficult and costly regeneration processes (high temperatures). [7,8] Thus, a new class of porous materials, metal-organic frameworks (MOFs), has been postulated as a new emerging class of functional material for multitasking applications. MOFs are constructed with metal centers and organic ligands, displaying, in some cases, high thermal and chemical stabilities. [9] Therefore, MOFs have been investigated in different areas, such as pollutant adsorption, [10] gas detection, [11] and catalysis. [12] Selected MOFs have been used as catalysts for several reactions exhibiting different kinds of active sites, [13] such as single atoms, [14] defect sites, [15] metal clusters, [16] functional groups, [17] and, recently, semi-open metal sites. [18] MOF materials have been applied for gas and liquid phase reaction catalysts with high conversion and successful selectivity. [19,20] One of the most promising candidates to catalyze chemical reactions in MOFs is open-metal sites (OMS). OMS is described as vacant Lewis acid sites (electrons acceptor) on the metal ions, generated by water/solvent de-coordination after an activation process (vacuum/temperature). [21] The ability to directly interact with guest molecules (reagents) via a Lewis acid center exhibits a straightforward pathway catalytic process. For example, the aerobic oxidation of cyclohexene using bimetallic MOF-74 displayed a 36 % of conversion, [22] while MIL-101(Cr) showed a characteristic open Cr + 3 site. This MOF material was employed for one-step methyl isobutyl ketone synthesis (60 % of conversion) [23] and a hydrogenation process (almost 100 % of conversion). [24] Also, a Cu(II)-MOF exhibited catalytic activity for alcoholysis with high conversion and selectivity. [25] Thus, OMS has demonstrated key features for catalytic applications, even for ring-opening reactions, where it is necessary for nucleophilic species. For example, the low nucleophilicity of methanol is complemented by a Lewis acid metal center from OMS. [26] Based on the outstanding OMS characteristics for catalytic applications, we investigated the catalytic properties of a Bi(III) ellagate-based MOF. SU-101 is a robust and biocompatible MOF displaying a green synthesis process. Highlighting its well-defined open Bi + 3 sites (Figure 1), which can act as Lewis acid centers for different catalytic applications. [27] The role of OMS within this MOF material was demonstrated in the transformation of H 2 S into polysulfides. [28] Additionally, SU-101 was applied for transfer hydrogenation, [29] CO 2 adsorption, [30] and pollutants adsorption. [31] Interestingly, nanorods of nonporous SU-101 were recently reported as a promising catalyst for CO 2 electroreduction, [32] showing robust chemical stability, high and selective catalytic activity, and exceptional retention of its catalytic activity for a long-term process.
Inspired by this work and by enhancing the attractive catalytic properties of this material, SU-101 with a rod-like morphology catalyst was synthesized at room temperature following a previously reported methodology. [32] The material was characterized by FTIR, PXRD, TGA, SEM, and N 2 adsorption.
SU-101 was evaluated in the ring-opening of epoxides by nucleophilic alcoholysis of cyclohexene oxide. Three different alcohols (methanol, ethanol, and propanol) were employed for the reaction. The reaction was monitored for three different temperatures (30,40, and 50°C) for each alcohol.

Results and Discussion
SU-101 was synthesized via the green hydrothermal method without using toxic solvents, following the methodology previously reported. [27] PXRD pattern ( Figure S1) displays the characteristic peaks for its crystalline structure, which is in good agreement with the crystal structure. [27] The well-defined peaks suggest a high crystallinity of SU-101. FTIR ( Figure S2) shows a broad band at 3554 cm À 1 attributed to the À OH stretching vibration. The chemical coordination between Bi + 3 and the hydroxyl groups from the linker is shown at 1326 cm À 1 . [33] To corroborate the thermal stability of SU-101, TGA analysis was implemented ( Figure S3). It observed a thermal resistance of up to 320°C, which is convenient for its use as a catalyst. SEM micrographics ( Figure S4) corroborated the rod-like morphology phase in the material. Also, BET surface area and pore volume were calculated from nitrogen isotherm at 77 K ( Figure S5). Displaying 35.5 m 2 g À 1 and 0.246 cm 3 g À 1 , respectively. Moreover, the values of BET surface area were reported as 412 [27] and 26 m 2 g À 1 . [32] Such different BET values are attributed to the synthetic routes. The low surface area could be associated with a lower amount of modulator (glacial acetic acid) in the synthesis.
SU-101 has been applied as a catalyst for the ring-opening of cyclohexene oxide using methanol, ethanol, and propanol (Scheme 1). Each reaction was conducted at 30, 40, and 50°C an atmospheric pressure of 30 h. These reactions are performed using Lewis acid catalysts. In this case, the open Bi + 3 sites are the main catalytic site to accomplish the reaction.
Prior to evaluating SU-101 as a catalyst, blank experiments were conducted (Table S1). The conversion was negligible (� 2.3 %), referring to the catalyst required for the reaction to proceed. Thus, SU-101 was utilized as a catalyst for this catalytic reaction and exhibited high catalytic activity. The cyclohexene oxide conversion for each experiment is illustrated in Table 1. For entries 1-3, methanol was used for the reaction, varying the reaction temperature. In the three cases, the conversion was higher than 99 %. A slight improvement in conversion was noticed with the increase in temperature. For entries 4-6, ethanol was employed for the reaction. Moreover, it was evidenced by the effect of temperature. In entry 4, the conversion was only 62.3 %. However, in entries 5 and 6, the conversion has increased considerably to 96.8 and 98.7 %, respectively. Clearly, this reaction is affected by thermodynamics. The change in the temperature gradient evidently enhanced the catalytic activity of SU-101. [34] For entries 7-9, propanol was applied for the reaction. Particularly, the higher conversion was 24.5 % at 50°C (entry 9). The conversion improved two-fold from 30 to 50°C, which is related to a further motion of the molecules within to access the catalytic sites. [35] The kinetic performances for each alcohol at 40°C are displayed in Figure 2a. A high conversion to the short-term reaction time for methanol and ethanol is observed. In these cases, the conversion upgrading increased systematically with the reaction time. However, in the case of propanol, the catalytic reaction was delayed. Furthermore, an effect on reaction rate is observed with different alcohols. A reaction equilibrium is achieved after 24 h of reaction with methanol and ethanol, showing almost a complete conversion. When propanol is used, it is observed that a longer time is necessary in order to obtain an equilibrium to boost conversion. These performances are related to the diffusion of the molecules to the external surface of the catalyst, where the active sites of the material are located, reaching the catalytic centers to achieve a chemical reaction finally. This is associated with the different size distribution of each alcohol (methanol < ethanol < propanol). [36] SU-101 exhibits different kinetic profiles for each alcohol at different temperatures ( Figures S6-S8). For methanol, an apparent increase in the reaction rate is observed. For example, at 7 h of reaction, the conversion raised from 70.3 at 30°C to 97.5 % at 50°C. Clearly, at a lower temperature, the reaction time is reduced. In the case of ethanol, the kinetic show a similar pattern for each temperature, with a slight increment to a higher value. On the other hand, when propanol is used, the reaction rate is slower. Moreover, the reaction is limited by the reagent diffusion from the solution to the external surface on SU-101. This is attributed to the geometry localization of the open metal site since Bi + 3 displays a hexacoordinated chemical environment. From this, the interaction between the reagent and the reaction site can be affected by the size of the reagent.
Furthermore, a leaching test was carried out to corroborate the heterogeneity of the reaction using SU-101 as a catalyst (Figure 2b). For this experiment, the same protocol using methanol at 40°C was performed. After 5 h of reaction, the catalyst was separated from the solution by centrifugation. It reached 66.7 % conversion. However, after removing the catalyst, the conversion only increased by 1 %. In this case, it is observed that no active species were leached during the catalytic reaction. We cycled 5 times this experiment, and we did not notice any difference in the catalytic conversion.
Additionally, based on the outstanding catalytic performance of SU-101 employing methanol, five catalytic cycles were carried out at 40°C. Remarkably, the catalytic activity remained unchanged ( Figure S9) since the conversion only dropped 2 % after the fifth cycle. Additionally, the crystalline structure of SU-101 was conserved after catalysis and during all the cycle process reactions, which was confirmed by PXRD ( Figure S10).
Catalytic activity of SU-101 is compared to conventional materials and different MOFs in Table S2. The well-studied mesoporous aluminosilicate only showed 69 and 96 % conversion for methanol and ethanol at 50°C, respectively. [37] Ga and Zn-based MOFs exhibited a conversion of around 92.1 % for methanol. However, the reaction is performed at 120°C. [38,39] A modified UiO-66 shows high conversion (98 %) after 30 h and 60°C. [40] Nevertheless, the synthesis difficulties  and high prices for these MOF materials negatively impact their use in these catalytic reactions. Moreover, SU-101 outperforms these materials for cyclohexene oxide alcoholysis due to its high cyclability. Thus, the mechanism for cyclohexene oxide alcoholysis is proposed based on previously reported literature for an acidcatalyzed ring-opening reaction and based on the results previously discussed. [41,42] (i) First, the diffusion of the reagents in the solution to the external surface of SU-101 is driven. (ii) Then, cyclohexene oxide is adsorbed on the unsaturated Bi + 3 sites (Lewis sites). This leads to an acid-base interaction between the Bi + 3 sites and the oxygen from the epoxide ring, enhancing the electrophilicity of the carbon atom, which is bonded to oxygen. This performance has been demonstrated in comparable systems, mainly for CO 2 cycloaddition. [43] (iii) Later, the nucleophile (methanol, ethanol, and propanol) attacks directly the carbon atom, establishing a proton transfer and forming the opening of the epoxide. (iv) Finally, the ringopened product is formed and then released from the active sites.

Conclusion
Summarizing, the catalytic activity of SU-101 in the acidcatalyzed ring-opening alcoholysis of cyclohexene oxide has been successfully demonstrated. The catalytic behavior is associated with the presence of open metal sites within the structure of SU-101. PXRD confirms the high stability of SU-101 after the catalysis and the cycling process. Moreover, the alcoholysis reaction was performed with three different alcohols and at three temperatures, using methanol, ethanol, and propanol, at 30, 40, and 50°C. Thus, SU-101 exhibits a high catalytic activity for this reaction, showing an outstanding conversion of 99.8 and 96.8 % for methanol and ethanol, respectively. Also, SU-101 remains catalytic activity after five consecutive catalytic cycles. In addition, the mechanism was proposed involving the Bi + 3 open sites from the SU-101.

SU-101 synthesis
SU-101 was synthesized following a previously reported procedure. [32] 15 mg of Ellagic acid and 38 mg of Bismuth (III) acetate were dissolved in 30 mL of water and 1 mL of acetic acid. First, the solution was stirred at room temperature for 48 h. Then, the powder was washed with water, ethanol, and acetone. After that, it was dry overnight at 80°C.

SU-101 characterization
The characterization technics were described in Section S1 from the Supporting Information.

Catalysis experiments
Prior to the catalysis test, SU-101 was solvent exchanged (5 mL) three times a day for two days with methanol, ethanol, and propanol. Then the catalyst was dry for 3 h at the same temperature as the reaction. The solvent exchanged was applied with the same alcohol as the reaction took place. Thus, the activation in our system occurs automatically with the reaction system without heat or being subject to a vacuum. The alcoholysis reaction was carried out in a 10 mL glass vial. A suspension containing 75 mmol of each alcohol (methanol, ethanol, or 1-propanol), 3 mmol of cyclohexene oxide, and 0.01 mmol of SU-101 to achieve an alcohol excess and a 300 : 1 relation between the limited reagent and catalyst. The solution was stirred vigorously (400 rpm) at different temperatures (40,50, and 60°C) for 30 h. Each reaction was measured in triplicate to calculate the error values. Also, an aliquot of 0.1 mL was taken and dissolved in 1 mL of acetonitrile for different time intervals to be analysed. For the cyclability test, the catalyst was separated by centrifugation (4000 rpm for 4 min) and washed with acetone. In the leaching test, the catalyst was separated by centrifugation (4000 rpm for 4 min) after 5 h of reaction. Then the solution was incorporated into the vial. Furthermore, reaction products were analyzed using a Clarus 480 PerkinElmer gas chromatograph equipped with a flame ionization detector (FID) and an Elite-1 capillary column. Toluene (0.3 mmol) was used as an internal standard.