English

• Metal-organic cages (MOCs) are individual supramolecular units created from metal nodes and organic connectors. Despite their assembly and structure being similar to metal-organic frameworks (MOFs), the MOCs, structured at the nanoscale level, exhibit unique attributes in terms of size, topology, solubility, characterization, and stability [1]. So far, a variety of MOCs have been documented, including those derived from rhodium (Rh), iron (Fe), cobalt (Co), titanium (Ti), and zirconium (Zr) [2-6]. The high bond energies of metal-oxygen bonds impart considerable structural stability to these MOCs. In particular, the Zr-O bond, with its energy of 766.1 ± 10.6 kJ/mol, results in Zr-MOCs exhibiting greater stability compared to MOCs based on other metals. Nevertheless, there have been few reported instances of Zr-MOCs, with research into their application performances also being greatly limited [7,8]. During the synthesis of Zr-MOCs, the susceptibility of Zr to oxidation poses a challenge. Zr being a reactive transition metal, readily reacts with oxygen to form oxides, especially under conditions of elevated temperature, humidity, or exposure to air. This tendency contributes to the scarcity of reported Zr-MOCs structures, thereby limiting their applications in performance.

  To expand the performance research of Zr-MOCs, functionalizing modifications of the ligands of MOCs can be employed. Numerous studies have shown that incorporating functional groups into the organic ligands of MOCs can enhance their applications in various aspects [1,9-11]. Introducing Lewis basic functional groups (such as hydroxyl, amino, carboxyl, cyano, carbonyl, and fluorine) can enhance the selectivity and catalytic efficiency of specific organic reactions. Fluorine-functionalized MOCs exhibit characteristics such as fluorophilicity, electrostatic interactions, and anion-π stacking. Moreover, due to the high electronegativity of fluorine, the C-F bond possesses high bond energy (110 kJ/mol), leading to the high chemical stability of C-F bonds [12-17]. Furthermore, the high electronegativity of fluorine results in greater electron density toward the fluorine side in covalent bonds. This induces a shift in the electron energy levels within the molecular orbitals (MOs), elevating energy levels for atoms connected to fluorine and lowering energy levels for fluorine atoms. This facilitates redox reactions, making MOCs more widely applicable. Among the reported Zr-MOCs, the most classic are ZrT-1, 2, 3, 4 reported by Yuan in 2013 [18]. Over the past decade, various research groups have conducted numerous functionalization modifications on the cages reported by Yuan [18]. For example, Su’s group reported amino modifications on ligands, enabling their application in photocatalytic hydrogen evolution reactions [19]. However, their application in photocatalytic organic reactions is relatively limited.

  Building upon the achievements of Yuan in 2013 [18], where the classical ZrT-3 was employed, we fluorinated the organic ligands to synthesize H₂bpdc-8F as a fluorine-rich linker (Fig. 1). Using solvent-thermal synthesis, we then combined H₂bpdc-8F with Zirconocene dichloride (Cp₂ZrCl₂) to produce a tetrahedral-shaped Zr-MOCs: ZrT-3-F, with a molecular formula of {[Cp₃Zr₃μ₃-O(μ₂-OH)₃]₄(bpdc8F)₆}·4Cl·nS (where S = noncoordinated solvent molecule). It is noteworthy that the fluorinated cage exhibited higher stability and enhanced light absorption capabilities. Utilizing these advancements, we applied the fluorinated cage for the first time in the oxidation of benzyl sulfide to the sulfoxide reaction. In photocatalytic reaction experiments, we observed that the fluorinated cage demonstrated a selectivity and conversion rate of over 95% for sulfoxide, significantly surpassing the performance of the non-fluorinated cage.

  Figure 1

  

  Figure 1.  Contrast sample: ZrT-3 (left), ZrT-3-F (right), bridging ligands and trinuclear zirconium clusters.

  DownLoad: Full-Size Img PowerPoint

  The compound ZrT-3-F is produced via a solvent-thermal procedure, combining Cp₂ZrCl₂ and H₂bpdc-8F in a mixed solvent environment of dimethylacetamide (DMA), trichloromethane (CHCl₃), and H₂O. The exact synthesis route can be found in the supplementary materials provided. Utilizing X-ray single crystal diffraction analysis, we determined that ZrT-3-F crystallizes in the cubic Fm-3 m space group. The compound forms a distinctive V₄E₆ type cage, as previously documented by Yuan. The structure of ZrT-3-F consists of four secondary building units (SBUs), namely Cp₃Zr₃μ₃-O(μ₂-OH)₃, and six H₂bpdc-8F linkers. The Cp₃Zr₃μ₃-O(μ₂-OH)₃ unit acts as the vertex (V), while H₂bpdc-8F functions as the edges (E) of the tetrahedron. Due to the high symmetry of ZrT-3-F, the linking ligands exhibit significant disorder. The interspace between two SBUs measures 15.387 Å, resulting in an approximate internal volume of the cage of 139.2 ų, as derived from calculations.

  To evaluate the high photocatalytic activity potential of ZrT-3-F, a thorough investigation into its optical and electronic properties was conducted. Additionally, a comparative analysis between ZrT-3-F and ZrT-3 was carried out to elucidate key disparities in band structure and charge carrier behavior. UV/visible diffuse reflectance spectroscopy (DRS) measurements indicated that ZrT-3-F exhibited nearly full-spectrum absorbance, including visible light absorption between 400 nm and 480 nm, unlike ZrT-3 which primarily absorbed in the UV region (Fig. 2a). Bandgap energies were computed using the Kubelka-Munk function equation: αhv = A(hv – E_g)^1/2. The direct bandgap energies of ZrT-3-F and ZrT-3 were estimated to be 2.41 and 3.32 eV, respectively, based on transformed differential spectra (Fig. 2b). Electronic band structures were determined by estimating conduction band (CB) potentials via Mott-Schottky (M-S) plots at frequencies of 1000, 1500, and 2000 Hz in the dark. Both ZrT-3-F and ZrT-3 M-S plots exhibited positive slopes across all frequencies, consistent with n-type semiconductor characteristics (Figs. S9 and S10 in Supporting information) [20]. Flat band potentials (V_FB) were extrapolated to be –0.757 and –0.886 V vs. Ag/AgCl for ZrT-3-F and ZrT-3, respectively. Consequently, the CB potentials for ZrT-3-F and ZrT-3 were –0.557 and –0.686 V, respectively. Using the formula E_CB = E_VB – E_g [21], VB positions for ZrT-3-F and ZrT-3 were calculated as 1.853 and 2.634 V, respectively, corresponding to the band structure arrangement (Fig. 2c). Furthermore, photocurrent response tests revealed a substantial increase in current under light irradiation compared to dark conditions. Notably, the photocurrent of ZrT-3-F consistently exceeded that of ZrT-3, irrespective of light exposure (Fig. 2d).

  Figure 2

  

  Figure 2.  (a) UV–vis DRS spectra of ZrT-3-F (red) and ZrT-3 (black). (b) Band gaps of ZrT-3-F (red) and ZrT-3 (black). (c) Electronic band structure. (d) Photocurrent responses of ZrT-3-F (red) and ZrT-3 (black).

  DownLoad: Full-Size Img PowerPoint

  The process of converting sulfides into sulfoxides via aerobic oxidation finds extensive usage in sectors such as fossil fuel desulfurization, medicinal synthesis, and wastewater management [22,23]. In conventional synthesis methods, the elevated oxidation potentials of sulfides may induce over-oxidation reactions, generating undesired by-products [24-26]. Consequently, an efficacious photocatalyst is required for the successful aerobic oxidation of sulfides to sulfoxides [27,28].

  An investigation on the photocatalytic performance of ZrT-3-F was carried out, utilizing the aerobic oxidation of sulfides to sulfoxides reaction. To determine the type of reaction products, a gas chromatography-mass spectrometry (GC–MS) was employed and quantification was conducted using a 50-bit ALS gas chromatograph. Since the main light absorption of ZrT-3-F lies in the ultraviolet region, UV light was adopted as the light source for the catalytic reaction. Initially, the choice of the best solvent was required, using oxygen as the oxidant. The selected solvents: N,N-Dimethylformamide (DMF), methanol (CH₃OH), isopropanol (IPA), and acetonitrile (CH₃CN) were put under consideration. The conversion and selectivity of CH₃OH and CH₃CN are the highest with the catalyst being homogeneous and heterogeneous in CH₃OH and CH₃CN respectively. From the standpoint of recyclability, the heterogeneous catalyst is more appropriate and hence CH₃CN was selected as the reaction solvent. Subsequently, it was necessary to elect the right oxidant: even though O₂, hydrogen peroxide (H₂O₂), and tert-butyl hydroperoxide (TBHP) all achieved conversion rates up to 99%, O₂ proved to be the most optimal oxidant regarding product selectivity. Upon conducting experiments with varying masses of the catalyst, the results indicate that the optimal catalyst loading is 10 mg.

  The optimal experimental conditions are as follows: 0.1 mmol of substrate, 3 mL of CH₃CN, and 10 mg of catalyst are added into a 10-mL quartz tube. The photocatalytic reaction is then performed for 6 h under 1 atm of oxygen and ultraviolet light irradiation. When ZrT-3-F is used as the catalyst, the conversion rate reaches up to 99.51% and selectivity up to 95.16%. In contrast, when ZrT-3 is used as the catalyst in the experiment, the conversion rate of benzyl sulfide is only 21.44%, and the selectivity is only 66.7%. When the organic ligand H₂bpdc-8F is used, the conversion rate is 52.67%, and the selectivity is 96.67%. When Cp₂ZrCl₂ is used as the catalyst, the conversion rate is 51.48%, and the selectivity is 58.29%. In summary, in this experimental system, the catalyst ZrT-3-F has the highest value (Table 1).

  Table 1

  

  Table 1.  Photocatalytic oxidation of thioanisole catalyzed by the ZrT-3-F.

  DownLoad: CSV

  Next, the wide-ranging suitability of ZrT-3-F as a heterogeneous photocatalyst was put under scrutiny, assessing the transformation rates and selectivity amongst various sulfides. As shown in the table, all the chosen sulfides demonstrated over 90% selectivity and conversion to sulfoxides. Table 2 shows a marked drop in the conversion performance when the para-position on the sulfur is electron-absorbing, confirming that electron-donating para-substituents are more conducive to the photocatalytic oxidation of thioanisole to its corresponding sulfoxide under aerobic conditions.

  Table 2

  

  Table 2.  Photocatalytic selective oxidation of sulfides to sulfoxide over ZrT-3-F.

  DownLoad: CSV

  The photostability of ZrT-3-F was evaluated through several centrifugations processed and washed with CH₃CN. Upon conducting three repeated experiments with ZrT-3-F (Fig. 3a), it was found that the reactive capacity of the catalyst did not decrease significantly. The catalyst harvested after the cycle was put through Fourier transform infrared (FT-IR) spectroscopy (Fig. S16 in Supporting information), with the spectral results demonstrating negligible differences between the pre-reaction and post-reaction spectrums. This indicates that the structure remained unaffected. According to these findings, ZrT-3-F demonstrates exceptional stability as a heterogeneous photocatalyst.

  Figure 3

  

  Figure 3.  (a) Stability of the ZrT-3-F photocatalyst for the selective aerobic oxidation of thioanisole under ultraviolet light. (b) Effect of scavengers on the photocatalytic oxidation of thioanisole. Reaction conditions: thioanisole (0.1 mmol), ZrT-3-F (10 mg), CH₃CN (3.0 mL), O₂ (1 atm), 300 W Xe lamp (λ = 200–400 nm), room temperature, 6.0 h. EPR detection of in situ formed (c) ¹O₂ using TEMP as a spin-trapping agent and (d) ^•O₂^‾ radical using DMPO as radical trapping agent.

  DownLoad: Full-Size Img PowerPoint

  To delve into the fundamental reaction mechanism responsible for the photocatalytic oxidation of thioanisole into sulfoxide under aerobic conditions, a sequence of radical quenching experiments was performed (Fig. 3b). Utilizing 2,6-di-tert-butyl-p-cresol (BHT) as a radical quenching agent, the production of reaction outputs was inhibited to a significant extent. This finding denotes the involvement of free radicals in this reaction. The subsequent introduction of copper sulfate (CuSO₄), acting as an electron inhibitor, caused a noticeable reduction in catalytic conversion efficiency. The implementation of potassium iodide (KI) as a hole inhibitor induced a state wherein the substrate underwent nearly no conversion, attesting to the occurrence of a photo-induced electron-hole (e^‾-h⁺) separation process in the reaction [29,30]. To detect the presence of reactive oxygen species (ROS) generated during the reaction, tert-butyl alcohol was added to quench any potential hydroxyl radicals (^•OH). The lack of any conspicuous effect on conversion rates post its addition provided the evidence that ^•OH radicals are not produced within this reaction framework. When p-benzoquinone and 2,2,6,6-tetramethylpiperidin-1-oxyl (TEMPO) were incorporated for trapping superoxide anion radical (^•O₂^‾) and singlet oxygen (¹O₂) respectively, a sharp decline in conversion rates was observed. This observation conclusively established that the ROS involved in this reaction are ^•O₂^‾ and ¹O₂ [30,31].

  To further validate the existence of ROS in the reaction, electron paramagnetic resonance (EPR) experiments were implemented. The spin trap agent selected for the ^•O₂^‾ was 5,5-dimethyl-1-pyrroline N-oxide (DMPO), while 2,2,6,6-tetramethylpiperidine (TEMP) was utilized for trapping singlet oxygen ¹O₂. As depicted in the accompanying Figs. 3c and d, both ^•O₂^‾ and ¹O₂ demonstrated a significant response under photonic exposure, while no noticeable response was registered in dark conditions.

  In conclusion, the conceivable mechanism for the photolytic conversion of phenyl methyl sulfide to sulfoxide (Fig. 4) involves UV light irradiation on ZrT-3-F, instigating the production of photo-generated electrons and holes. These photo-generated electrons swiftly disperse towards the surface of the catalyst and into the conduits where O₂ is reduced by the photoelectrons, resulting in ^•O₂^‾. Neutral exciton-forming entities possessing robust Coulomb interactions undergo a transition from the singlet excited state (S₁) to the low-energy triplet state (T₁) via inter-system crossing (ISC) [32]. In this scenario, the energy is ceded to the attached O₂ molecules, leading to the production of ¹O₂ and phosphorescence. Upon acquisition of a proton, the sulfide transforms into the free radical cation CH₃SC₆H₅H⁺. ^•O₂^‾ and CH₃SC₆H₅H⁺ then initiate a nucleophilic reaction [33,34]. Concurrently, ¹O₂ instigates an electrophilic reaction with the sulfide to create a sulfur peroxide intermediate. The reaction of the sulfide with the sulfur peroxide intermediate paves the way for the synthesis of sulfoxides. Notably, the end product of both mechanisms is sulfoxide [35].

  Figure 4

  

  Figure 4.  Possible photocatalytic mechanisms.

  DownLoad: Full-Size Img PowerPoint

  Fluorination of the classical Zr-MOCs, such as ZrT-3, results in enhanced light absorption compared to the non-fluorinated cages. The fluorinated ZrT-3 exhibits a stronger photocurrent response, thereby presenting advantages in photocatalytic reactions. Therefore, we employed ZrT-3-F for the oxidation of benzyl sulfide to sulfoxide. In photocatalytic reactions, ZrT-3-F serves as a heterogeneous catalyst under aerobic conditions, and achieves conversion rates and selectivity in oxidizing benzyl sulfide to sulfoxide. The conversion rate and selectivity are significantly higher compared to the performance of ZrT-3 in catalytic reactions, demonstrating the excellent versatility and chemical stability of the catalyst throughout the reaction process. Importantly, this work demonstrates the applicability of fluorinated Zr-MOCs in organic catalysis.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Jia-Ru Li: Data curation, Investigation, Methodology, Writing – original draft, Writing – review & editing. Ning Li: Funding acquisition, Supervision, Writing – review & editing. Li-Ling He: Formal analysis, Validation. Jun He: Funding acquisition, Resources, Supervision.

  Acknowledgments

  This work is supported by National Natural Science Foundation of China (Nos. 22201046, 22371054), Local Innovation Research Team Project of Guangdong Pearl River Talent Plan (No. 2017BT01Z032).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.109934.

  
  
• 1. [1]

     E.M. El-Sayed, Y.D. Yuan, D. Zhao, et al., Acc. Chem. Res. 55 (2022) 1546–1560. doi: 10.1021/acs.accounts.1c00654

  2. [2]

     S. Furukawa, N. Horike, M. Kondo, et al., Inorg. Chem. 55 (2016) 10843–10846. doi: 10.1021/acs.inorgchem.6b02091

  3. [3]

     I. Papadaki, C.D. Malliakas, T. Bakas, et al., Inorg. Chem. 48 (2009) 9968–9970. doi: 10.1021/ic901145d

  4. [4]

     S.T. Wang, X.H. Gao, X.X. Hang, et al., J. Am. Chem. Soc. 138 (2016) 16236–16239. doi: 10.1021/jacs.6b11218

  5. [5]

     Y.P. He, L.B. Yuan, G.H. Chen, et al., J. Am. Chem. Soc. 139 (2017) 16845–16851. doi: 10.1021/jacs.7b09463

  6. [6]

     Z.F. Ju, G.L. Liu, Y.S. Chen, et al., Chem. Eur. J. 23 (2017) 4774–4777. doi: 10.1002/chem.201700798

  7. [7]

     M.G. Sullivan, H.K. Welgama, M.R. Crawley, et al., Chem. Mat. 36 (2024) 567–574. doi: 10.1021/acs.chemmater.3c02775

  8. [8]

     B.X. Jiang, H. Wang, Y.T. Zhang, et al., Polyhedron 243 (2023) 116569. doi: 10.1016/j.poly.2023.116569

  9. [9]

     H.Y. Lin, Z.F. Xiao, K.N. Le, et al., Angew. Chem. Int. Ed. 61 (2022) e202214055. doi: 10.1002/anie.202214055

  10. [10]

      J.J. Jiao, C.X. Tan, Z.J. Li, et al., J. Am. Chem. Soc. 140 (2018) 2251–2259. doi: 10.1021/jacs.7b11679

  11. [11]

      W.H. Xing, H.Y. Li, X.Y. Dong, et al., J. Mater. Chem. A 6 (2018) 7724–7730. doi: 10.1039/C8TA00858B

  12. [12]

      S. Lee, H. Jeong, D. Nam, et al., Chem. Soc. Rev. 50 (2021) 528–555. doi: 10.1039/D0CS00443J

  13. [13]

      Q. Tian, S. Chen, M. Shi, et al., Sens. Actuator. B: Chem. 404 (2024) 135309. doi: 10.1016/j.snb.2024.135309

  14. [14]

      H.T. Chen, T. Zhang, S.R. Liu, et al., Inorg. Chem. 61 (2022) 11949–11958. doi: 10.1021/acs.inorgchem.2c01686

  15. [15]

      Z.A. Qiao, S.S. Brown, J. Adcock, et al., Angew. Chem. Int. Ed. 51 (2012) 2888–2893. doi: 10.1002/anie.201107812

  16. [16]

      D.K. Dogutan, R. McGuire, D.G. Nocera, J. Am. Chem. Soc. 133 (2011) 9178–9180. doi: 10.1021/ja202138m

  17. [17]

      S.S. Wang, L. Wang, M. Dakovic, et al., ACS Catal. 2 (2012) 230–237. doi: 10.1021/cs200501n

  18. [18]

      G.L. Liu, Z.F. Ju, D.Q. Yuan, et al., Inorg. Chem. 52 (2013) 13815–13817. doi: 10.1021/ic402428m

  19. [19]

      M. Sun, Q.Q. Wang, C. Qin, et al., Chem. Eur. J. 25 (2019) 2824–2830. doi: 10.1002/chem.201805663

  20. [20]

      H.F. Ye, R. Shi, X. Yang, Y. Chen, et al., Appl. Catal. B: Environ. 233 (2018) 70–79. doi: 10.1016/j.apcatb.2018.03.060

  21. [21]

      X. Li, Y. Qi, G.Z. Yue, et al., Green Chem. 21 (2019) 649–657. doi: 10.1039/C8GC03295E

  22. [22]

      Y.M. Li, M. Wang, X.F. Jiang, ACS Catal. 7 (2017) 7587–7592. doi: 10.1021/acscatal.7b02735

  23. [23]

      A.K. Clarke, A. Parkin, R.J.K. Taylor, et al., ACS Catal. 10 (2020) 5814–5820. doi: 10.1021/acscatal.0c00690

  24. [24]

      Q. Li, X.W. Lan, G.Y. An, et al., ACS Catal. 10 (2020) 6664–6675. doi: 10.1021/acscatal.0c00290

  25. [25]

      X.Y. Wang, M.H. Liu, Y.X. Liu, et al., J. Am. Chem. Soc. 145 (2023) 26900–26907. doi: 10.1021/jacs.3c09699

  26. [26]

      X. Liang, Z.F. Guo, H.X. Wei, et al., Chem. Commun. 54 (2018) 13002–13005. doi: 10.1039/C8CC07585A

  27. [27]

      M. Forchetta, F. Sabuzi, L. Stella, et al., J. Org. Chem. 87 (2022) 14016–14025. doi: 10.1021/acs.joc.2c01648

  28. [28]

      X.N. Zou, D.S. Zhang, T.X. Luan, et al., ACS Appl. Mater. Interfaces 13 (2021) 20137–20144. doi: 10.1021/acsami.1c03083

  29. [29]

      H.X. Wei, Z.F. Guo, X. Liang, et al., ACS Appl. Mater. Interfaces 11 (2019) 3016–3023. doi: 10.1021/acsami.8b18206

  30. [30]

      C.L. Su, R. Tandiana, B.B. Tian, J. Su, K.P. Loh, et al., ACS Catal. 6 (2016) 3594–3599. doi: 10.1021/acscatal.6b00443

  31. [31]

      X.Y. Gu, X. Li, Y.H. Chai, et al., Green Chem 15 (2013) 357–361. doi: 10.1039/c2gc36683e

  32. [32]

      Q.Q. Li, P.H. Pan, H. Liu, et al., Inorg. Chem. 62 (2023) 17182–17190. doi: 10.1021/acs.inorgchem.3c02212

  33. [33]

      J. Jiang, R.C. Luo, X.T. Zhou, et al., Adv. Synth. Catal. 360 (2018) 4402–4411. doi: 10.1002/adsc.201800730

  34. [34]

      H.D. She, H. Zhou, L.S. Li, et al., ACS Sustain. Chem. Eng. 6 (2018) 11939–11948. doi: 10.1021/acssuschemeng.8b02217

  35. [35]

      Q.S. Zhu, H.Y. An, T.Q. Xu, et al., Appl. Catal. A: Gen. 662 (2023) 119283. doi: 10.1016/j.apcata.2023.119283

• 

  Figure 1  Contrast sample: ZrT-3 (left), ZrT-3-F (right), bridging ligands and trinuclear zirconium clusters.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (a) UV–vis DRS spectra of ZrT-3-F (red) and ZrT-3 (black). (b) Band gaps of ZrT-3-F (red) and ZrT-3 (black). (c) Electronic band structure. (d) Photocurrent responses of ZrT-3-F (red) and ZrT-3 (black).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (a) Stability of the ZrT-3-F photocatalyst for the selective aerobic oxidation of thioanisole under ultraviolet light. (b) Effect of scavengers on the photocatalytic oxidation of thioanisole. Reaction conditions: thioanisole (0.1 mmol), ZrT-3-F (10 mg), CH₃CN (3.0 mL), O₂ (1 atm), 300 W Xe lamp (λ = 200–400 nm), room temperature, 6.0 h. EPR detection of in situ formed (c) ¹O₂ using TEMP as a spin-trapping agent and (d) ^•O₂^‾ radical using DMPO as radical trapping agent.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Possible photocatalytic mechanisms.

  下载: 全尺寸图片 幻灯片

  Table 1.  Photocatalytic oxidation of thioanisole catalyzed by the ZrT-3-F.

  下载: 导出CSV

  Table 2.  Photocatalytic selective oxidation of sulfides to sulfoxide over ZrT-3-F.

  下载: 导出CSV
•