English

• The excessive consumption of widely used fossil fuels has led to a series of environmental issues and an energy crisis, prompting the rapid development of renewable energy sources [1-4]. The intermittent and stochastic features of renewable energy sources hinder their large-scale applications, which, in turn, spurs the blooming of electrochemical energy storage and conversion devices due to their remarkable efficiency [5-7]. Among various electrochemical devices, proton exchange membrane fuel cells (PEMFCs) have been considered promising candidates due to their ultra-low emissions, high power density, and fast start-up characteristics [8-10]. As a critical component in PEMFCs, PEM must exhibit high proton conductivity. Consequently, developing functional materials with efficient proton transfer channels is of great significance [11-13].

  In recent years, numerous functional materials have been developed to exhibit high proton conductivity, including metal-organic frameworks (MOFs), porous coordination polymers (PCPs) [14-17], and covalent organic frameworks (COFs) [18], whereas they suffer from the issues of stability. Polyoxometalates (POMs) are a type of metal-oxygen clusters typically composed of high-oxidation state early transition metals (primarily V, Mo, W, Nb, and Ta), whose oxygen-rich surface can offer a greater number of proton transfer sites and enhance proton transport efficiency [19,20]. In particular, research on POM-based proton conductive materials has gained increasing attention since the discovery that Keggin-type hetero-polyacid, specifically 12-tungstophosphoric acid hydrate, exhibits a high proton conductivity of up to 0.18 S/cm at room temperature [21]. Subsequently, many POMs have been demonstrated to exhibit desirable proton conductivity performance [22-24].

  Under specific conditions, Keggin-type POM clusters can transform into metastable vacancy units with varying numbers of vacancies by losing one or more MO₆ polyhedra, inducing 3d transition metal ions to form clusters at their vacancies as multidentate oxygen ligands, leading to the construction of novel structures with unique properties [25]. Currently, abundant reports exist on POMs substituted with 3d transition metal ions [26-31]. However, there are few reports on POMs co-substituted with 3d transition metal ions and 4p metal ions [32,33]. Therefore, exploring 3d-5p substituted POMs with high proton conductivity represents a significant and challenging research opportunity.

  Here, we report a 3d-5p substituted polyoxotungstate (POT), H₂₉Na₉(H₂O)₂₁{Ca(H₂O)₂@Sb₁₂O₁₈[Ni₂(OH)(A-α-SiW₁₀O₃₇)]₃}₂·40H₂O (1), which consists of three [A-α-SiW₁₀O₃₇]¹⁰⁻ ({A-SiW₁₀}) cluster units surrounding an 18-nuclear 3d-5p heterometallic cluster [Ni₆(OH)₃Sb₁₂O₁₈] ({Ni₆Sb₁₂}), with a Ca²⁺ ion captured on the window of the {Ni₆Sb₁₂} cluster. The powder form of 1 exhibits a high proton conductivity of 3.00×10⁻² S/cm, while the single-crystal form of 1 shows an ultra-high proton conductivity of 1.11×10⁻¹ S/cm in the [110] direction and 1.04×10⁻¹ S/cm in the [100] direction, indicating that the anisotropic proton conductivity is closely related to its structural orientation. This work will provide valuable insights into the design of high proton conductivity materials and pave the way for their practical application in advanced electrochemical technologies.

  Compound 1 is a high-nuclear 3d-5p heterometallic cluster-substituted core-shell antimony-rich POT. Its asymmetric unit (Fig. S1 in Supporting information) comprises two identical isolated trimer units, crystallized in the trigonal crystal system and P-3 space group (Table S1 in Supporting information). The Sb shell of the compound consists of 12 Sb atoms arranged in four Sb₃O₆ ({Sb₃}) triangles. Each triangle is formed by three SbO₃ trigonal pyramids sharing oxygen atoms, and these four triangles are interconnected by six oxygen atoms at their vertices to form a T_d symmetric Sb₁₂O₁₈ ({Sb₁₂}) cage. The {Sb₁₂} cage features four hexagonal and four triangular windows, creating a cup-shaped structure with approximate dimensions of 3.89 × 7.25 × 6.59 ų, and an interior spherical space with a diameter of approximately 7.25 Å (Fig. 1a). The Sb-O bond lengths range from 1.935(16)-2.033(19) Å Notably, the {Sb₁₂} cup-shaped cage encapsulates an 8-coordinate Ca²⁺ ion (Fig. 1b) within one of the hexagonal windows through coordination with its O atoms, forming an enclosed Ca(H₂O)₂@Sb₁₂O₁₈ ({Ca@Sb₁₂}) cup-shaped cage (Fig. 1c). The Ca²⁺ ion exhibits a hexagonal bipyramidal coordination mode, involving six μ₃-O atoms from the {Sb₁₂} cage and two coordinated water molecules, with Ca-O bond lengths ranging of 2.36(7)-2.606(17) Å.

  Figure 1

  

  Figure 1.  (a–d) View of different structural motifs in 1. (e) The structure of 1. CaO₈: purple; WO₆: red; SiO₄: yellow; NiO₆: green.

  DownLoad: Full-Size Img PowerPoint

  On the other hand, the nickel shell consists of six Ni²⁺ ions. Each Ni²⁺ ion adopts a six-coordinated octahedral geometry, involving four O atoms from the dilacunary {A-SiW₁₀} units, one μ₃-O atom from the {Sb₁₂} cup-shaped cage, and one μ₂-OH group. Every two Ni²⁺ ions are linked by a shared μ₂-OH group to form a di-nuclear Ni₂O₁₀(OH) ({Ni₂}) unit, which occupies vacancies in {A-SiW₁₀}, resulting in three Ni-substituted POT cluster units [Ni₂O₂(OH)(A-α-SiW₁₀O₃₇)] ({Ni₂(A-SiW₁₀)}, Fig. 1d). The Ni-O bond lengths range of 1.961(15)-2.149(17) Å. Bond valence sum (BVS) calculations indicate that the oxidation states of all Sb and Ni ions in the structure are +3 and +2, respectively. The dilacunary {A-SiW₁₀} units are derived from the initial trilacunary precursor [A-α-SiW₉O₃₄]¹⁰⁻ ({A-SiW₉}) through reassembly and transformation during the reaction (Fig. S2 in Supporting information).

  Furthermore, the three {Ni₂(A-SiW₁₀)} units are capped by six Ni-O-Sb bridges to the remaining three open hexagonal windows of the {Ca@Sb₁₂} cup-shaped cage, forming a trimer with C₃ symmetry that contains a nineteen-nuclear heterometallic cluster {Ca(H₂O)₂@Sb₁₂O₁₈[Ni₂(OH)(A-α-SiW₁₀O₃₇)]₃} ({Ca@Sb₁₂[Ni₂(A-SiW₁₀)]₃}, Fig. 1e). Overall, the trimer exhibits a multi-shell configuration (Ca @Sb shell @Ni shell @W shell ≡ ({Ca}@{Sb₁₂}@{Ni₂}₃@{A-SiW₁₀}₃)), with charge-matched interactions showing a unique positive@zero@positive@negative pattern. Its stability is attributed to the stabilizing effect of the outermost highly negatively charged {A-SiW₁₀} cluster block: The dilacunary {A-SiW₁₀} cluster tends to capture the {Ni₂} unit into its vacancies to form more stable saturated Keggin-type cluster block, and the resulting {Ni₂(A-SiW₁₀)} cluster unit acts as a template to induce the formation of the {Sb₁₂} cup-shaped cage, which can capture alkaline earth metals. Therefore, a Ca²⁺ ion can be encapsulated in its hexagonal window.

  In addition, crystal 1 forms a folded two-dimensional double-layer structure through the connection of Na⁺ ions (Fig. 2a). Viewed along the c-axis, this two-dimensional double-layer structure reveals a channel with a size of approximately 0.85 nm (Fig. 2b). The formation of 1 highlights the potential and vast possibilities of synergistically assembling s-block alkaline earth metal ions, p-block Sb-O clusters, d-block Ni-O clusters and W-O clusters to create new compounds. These four components, each with different structural characteristics and properties, have not been previously reported to coexist within the same cluster unit. This multi-component synergistic assembly approach offers exciting opportunities for the synthesis of new structures.

  Figure 2

  

  Figure 2.  2D structure diagrams of 1 (a) in a direction and (b) in c direction. CaO₈: purple; WO₆: red; SiO₄: yellow; NiO₆: green; Sb: blue; Na: yellow; O: red.

  DownLoad: Full-Size Img PowerPoint

  The purity of sample 1 was assessed using powder X-ray diffraction (PXRD, Fig. S3 in Supporting information). In addition, the sample was characterized by infrared spectroscopy (IR), UV–visible spectrum, and thermal analysis (Figs. S4–S6 in Supporting information). The metal content in compound 1 was determined by inductively coupled plasma atomic emission spectrometry (ICP-AES), as detailed in Table S2 (Supporting information), and was found to be consistent with the values calculated from single-crystal X-ray diffraction data. Notably, the boron element from the buffer solution does not become part of the structure. However, replacing the borate buffer solution with other buffer solutions of the same pH did not yield the crystal with desired structure. Therefore, the borate buffer solution not only primarily plays a role in regulating the pH value of the reaction, stabilizing it within a range suitable for compound growth, but also is an indispensable raw material for synthesis.

  The structural stability, presence of counter cations (H⁺, Ca²⁺), and the oxygen-rich surface of compound 1 suggest its potential as a proton conducting material. Therefore, we conducted a preliminary study on the proton conductivity of the powder form of 1. Initially, we tested the humidity-dependent conductivity at 25 ℃. As shown in Fig. 3a, the conductivity at 55% RH was 1.86×10⁻⁴ S/cm. Increasing the relative humidity from 55% to 98% resulted in a fifteenfold increase in conductivity, reaching 2.83×10⁻³ S/cm. This indicates that, under high humidity conditions, increased water molecule infiltration into the structure enhances proton transport. Subsequently, we evaluated the temperature-dependent conductivity while maintaining the humidity at 98% RH (Fig. 3b). The conductivity increased by an order of magnitude to 3.00×10⁻² S/cm as the temperature rose from 25 ℃ to 85 ℃. This enhancement in conductivity with increasing temperature is attributed to the accelerated transfer of protons through the channels. The conductivity of 1 at 85 ℃ and 98% RH is comparable to that of some reported high-performance POM-based proton conductive materials [34-36]. According to Arrhenius equation (σT = σ₀exp(-E_a/k_bT), the activation energy E_a was estimated to be 0.38 eV, indicating that the proton conduction follows the Grotthuss mechanism (< 0.4 eV), where proton migration occurs through a hydrogen bond network (Fig. 3c). The excellent proton conductivity is attributed to the abundant proton carriers within the structure of 1, including water molecules and terminal oxygen atoms, which facilitate the formation of a hydrogen-bonded proton "jumping" network [37-39]. The PXRD analysis confirms that the structure of 1 remains intact after proton conductivity testing (Fig. 3d).

  Figure 3

  

  Figure 3.  Nyquist plots for 1 (a) at different RHs and T = 25 ℃; (b) at different temperatures and 98% RH. (c) Arrhenius plots and the linear fit of 1 at 98% RH. (d) The PXRD patterns of the sample before and after proton conductivity test for 1.

  DownLoad: Full-Size Img PowerPoint

  To further elucidate the relationship between crystal structure and performance, we conducted the single-crystal proton conductivity measurements by carefully attaching silver wires to the hexagonal prism crystal of 1 using conductive silver paste. Alternating current (AC) impedance data of the single-crystal proton conductivity for 1 along the [001], [110], and [100] directions are collected (Fig. S7 in Supporting information).

  We first estimate the single-crystal humidity-dependent conductivity along different directions (Figs. 4a–c). The proton conductivity along [001] direction reaches 6.61×10⁻⁵ S/cm at 25 ℃ and 55% RH (Fig. 4a), which is significantly higher than the value along [110] and [100] directions, measured at 1.51×10⁻⁶ (Fig. 4b) and 2.26×10⁻⁵ S/cm (Fig. 4c), respectively. When the RH increased from 55% to 98%, the conductivity along the [001] direction only increased 26-fold, reaching 1.72×10⁻³ S/cm (Fig. 4a). Surprisingly, the values along the [110] and [100] directions increased nearly 6000-fold and 200-fold, exhibiting excellent proton conductivity of 8.82×10⁻³ and 4.16×10⁻³ S/cm, respectively (Figs. 4b and c). These results indicate that water molecules play a crucial role in enhancing proton conductivity under high RH conditions and crystal 1 is a highly anisotropic proton conductive material. Moreover, the higher proton conductivity suggests that the proton diffusion is favored along the [110] and [100] directions.

  Figure 4

  

  Figure 4.  (a–c) Nyquist plots at different RHs and T = 25 ℃. (d–f) Nyquist plots at different temperatures and 98% RH (inset: proton conduction device) for single-crystal of 1 along [001], [110], and [100] directions, respectively.

  DownLoad: Full-Size Img PowerPoint

  We further investigated the temperature-dependent proton conductivity of single-crystal 1 along different directions (Figs. 4d–f). The proton conductivities of single-crystal 1 increase along all directions with rising temperature, which is attributed to the accelerated transfer of protons within the channels. Specifically, along the [001] direction, the proton conductivity of single-crystal 1 increases from 1.72×10⁻³ S/cm to 1.53×10⁻² S/cm as the temperature rises from 25 ℃ to 85 ℃ at 98% RH (Fig. 4d), reflecting an enhancement of only one order of magnitude (about nine times). This result indicates that the temperature has a lesser impact on conductivity enhancement compared to humidity along the [001] direction. The proton conductivities reach 1.11 × 10⁻¹ and 1.04 × 10⁻¹ S/cm, respectively, at 85 ℃ and 98% RH (Figs. 4e and f). These values are roughly ten times greater than that of single-crystal 1 along the [001] direction under the same conditions, indicating that the proton diffusion is more favorable along the [110] and [100] directions. Furthermore, the proton conductivities of single-crystal 1 along [110] and [100] directions increase 12 and 25 times as temperature rises from 25 ℃ to 85 ℃, suggesting that the proton diffusion of 1 along the [100] direction is the most temperature-sensitive. The two-dimensional layer structures of crystal 1 along the [001], [110], and [100] directions are illustrated in Figs. S8–S10 (Supporting information), which are connected by Na⁺ ions.

  Surprisingly, the above test results reveal that single-crystal 1 achieves proton conductivity along the [110] and [100] directions at 85 ℃ and 98% RH, reaching 1.11×10⁻¹ and 1.04×10⁻¹ S/cm, respectively (Table S3 in Supporting information). These values are higher than the proton conductivity along the [001] direction (1.53×10⁻² S/cm) by an order of magnitude, indicating that 1 is a highly anisotropic proton conducting single-crystal. Few stable single-crystal materials have been reported to exceed this value (Table S4 in Supporting information).

  According to the Arrhenius equation, the activation energy E_a along the [001], [110] and [100] directions were estimated to be 0.33, 0.38, and 0.49 eV, respectively (Fig. 5). This indicates that the proton conduction along the [001] and [110] directions belong to the Grotthuss mechanism (< 0.4 eV), where proton migration relies on a hydrogen bond network. However, the proton conduction process along the [100] direction belongs to the Vehicular mechanism (> 0.4 eV) [40], where proton transfer relies on carrier transport.

  Figure 5

  

  Figure 5.  Arrhenius plots and the linear fit of single-crystal 1 along [001], [110], and [100] directions at 98% RH, respectively.

  DownLoad: Full-Size Img PowerPoint

  In conclusion, we have successfully synthesized an 18-nuclear 3d-5p cage-like cluster consisting of nickel and antimony atoms capped by a Ca²⁺ ion. Compound 1 exhibits exceptionally high single-crystal proton conductivities, reaching 1.11×10⁻¹ and 1.04×10⁻¹ S/cm along the [110] and [100] directions at 85 ℃ and 98% RH, respectively. These results demonstrate that it is a promising candidate for proton conductive materials. This work may offer a feasible approach for advancing the development of practical proton conducting materials.

  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

  Yi Chen: Writing – original draft, Formal analysis, Data curation. Hua-Pan Wu: Software, Investigation. Zheng-Wei Guo: Methodology. Xin-Xiong Li: Methodology, Investigation. Ping-Wei Cai: Writing – review & editing, Validation, Funding acquisition, Conceptualization. Shou-Tian Zheng: Writing – review & editing, Supervision, Funding acquisition.

  Acknowledgments

  We gratefully acknowledge the financial support from the National Natural Science Foundation of China (Nos. 22109164 and 22371046) and the Key Program of the Natural Science Foundation of Fujian Province (No. 2021J02007).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2024.110552.

  
  
• 1. [1]

     M.C. Luo, Z.L. Zhao, Y.L. Zhang, et al., Nature 574 (2019) 81–85. doi: 10.1038/s41586-019-1603-7

  2. [2]

     R.Z. Sun, J.X. Chen, S.L. Huang, et al., CCS Chem. 7 (2025) 1216–1226. doi: 10.31635/ccschem.024.202404140

  3. [3]

     Y.Y. Liu, J.J. Wu, K.P. Hackenberg, et al., Nat. Energy 2 (2017) 17127.

  4. [4]

     X.S. Sun, L. Li, et al., eScience 3 (2023) 100095.

  5. [5]

     M.Z. Jacobson, W.G. Colella, D.M. Golden, Science 308 (2005) 1901–1905. doi: 10.1126/science.1109157

  6. [6]

     X.M. Yin, W. Sun, K. Chen, et al., Adv. Sci. 11 (2024) 2402343.

  7. [7]

     Y.S. Wei, X.P. Hu, Z. Han, et al., J. Am. Chem. Soc. 139 (2017) 3505–3512. doi: 10.1021/jacs.6b12847

  8. [8]

     F. Xiao, Y.C. Wang, Z.P. Wu, et al., Adv. Mater. 33 (2021) 2006292.

  9. [9]

     R. Haider, Y.C. Wen, Z.F. Ma, et al., Chem. Soc. Rev. 50 (2021) 1138. doi: 10.1039/d0cs00296h

  10. [10]

      Y.X. Ye, L.S. Gong, S.C. Xiang, et al., Adv. Mater. 32 (2020) 1907090.

  11. [11]

      Y.X. Ye, W.G. Guo, L.H. Wang, et al., J. Am. Chem. Soc. 139 (2017) 15604–15607. doi: 10.1021/jacs.7b09163

  12. [12]

      Y. Guo, Z.Q. Jiang, W. Ying, et al., Adv. Mater. 30 (2018) 1705155.

  13. [13]

      J. Chen, Q.Q. Mei, Y.L. Chen, et al., J. Am. Chem. Soc. 144 (2022) 11969–11974. doi: 10.1021/jacs.2c04900

  14. [14]

      D.W. Lim, H. Kitagawa, Chem. Soc. Rev. 50 (2021) 6349. doi: 10.1039/d1cs00004g

  15. [15]

      X.M. Li, Y.M. Wang, Y.B. Mu, et al., ACS Appl. Mater. Interface. 14 (2022) 9264–9271. doi: 10.1021/acsami.2c00500

  16. [16]

      L. Qin, Y.Z. Yu, P.Q. Liao, et al., Adv. Mater. 28 (2016) 10772–10779. doi: 10.1002/adma.201603381

  17. [17]

      F.M. Zhang, L.Z. Dong, J.S. Qin, et al., J. Am. Chem. Soc. 139 (2017) 6183–6189. doi: 10.1021/jacs.7b01559

  18. [18]

      W.W. Zou, G.X. Jiang, W.F. Zhang, et al., Adv. Funct. Mater. 33 (2023) 2213642.

  19. [19]

      S. Herrmann, M. Kostrzewa, A. Wierschem, C. Streb, Angew. Chem. Int. Ed. 53 (2014) 13596–13599. doi: 10.1002/anie.201408171

  20. [20]

      Y. Chen, Z.W. Guo, X.X. Li, et al., CCS Chem. 4 (2022) 1305–1314. doi: 10.31635/ccschem.021.202100774

  21. [21]

      O. Nakamura, T. Kodama, I. Ogino, Y. Miyake, Chem. Lett. 8 (1979) 17–18. doi: 10.1246/cl.1979.17

  22. [22]

      L. Yang, J. Lei, J.M. Fan, et al., Adv. Mater. 33 (2021) 2005019.

  23. [23]

      J.M. Lin, N. Li, S.P. Yang, et al., J. Am. Chem. Soc. 142 (2020) 13982–13988. doi: 10.1021/jacs.0c06582

  24. [24]

      A. Martinelli, J.M. Otero-Mato, M.N. Garaga, et al., J. Am. Chem. Soc. 143 (2021) 13895–13907. doi: 10.1021/jacs.1c06656

  25. [25]

      Z.W. Guo, L.H. Lin, J.P. Ye, et al., Angew. Chem. Int. Ed. 62 (2023) e202305260.

  26. [26]

      B.S. Bassil, M. Ibrahim, R. Al-Oweini, et al., Angew. Chem. Int. Ed. 50 (2011) 5961–5964. doi: 10.1002/anie.201007617

  27. [27]

      Z.W. Guo, Y. Chen, Z.H. Chen, et al., Chin. Chem. Lett. 35 (2024) 109124.

  28. [28]

      C.H. Zhan, J.M. Cameron, J. Gao, et al., Angew. Chem. Int. Ed. 53 (2014) 10362–10366. doi: 10.1002/anie.201402932

  29. [29]

      S.S. Mal, U. Kortz, Angew. Chem. Int. Ed. 44 (2005) 3777–3780. doi: 10.1002/anie.200500682

  30. [30]

      L. Huang, S.S. Wang, J.W. Zhao, et al., J. Am. Chem. Soc. 136 (2014) 7637–7642. doi: 10.1021/ja413134w

  31. [31]

      X.B. Han, Y.G. Li, Z.M. Zhang, et al., J. Am. Chem. Soc. 137 (2015) 5486–5493. doi: 10.1021/jacs.5b01329

  32. [32]

      J. Cai, X.Y. Zheng, J. Xie, et al., Inorg. Chem. 56 (2017) 8439–8445. doi: 10.1021/acs.inorgchem.7b01104

  33. [33]

      S.R. Li, H.Y. Wang, H.F. Su, et al., Small Method. 5 (2021) 2000777.

  34. [34]

      Z. Li, X.X. Li, T. Yang, et al., Angew. Chem. Int. Ed. 56 (2017) 2664–2669. doi: 10.1002/anie.201612046

  35. [35]

      P. Yang, M. Alsufyani, A. -H. Emwas, et al., Angew. Chem. Int. Ed. 57 (2018) 13046–13051. doi: 10.1002/anie.201806086

  36. [36]

      M.H. Zhu, T. Iwano, M.J. Tan, et al., Angew. Chem. Int. Ed. 61 (2022) e202200666.

  37. [37]

      A. Shigematsu, T. Yamada, H. Kitagawa, J. Am. Chem. Soc. 133 (2011) 2034–2036. doi: 10.1021/ja109810w

  38. [38]

      S. Ram, Acta Mater. 49 (2001) 2297–2307.

  39. [39]

      S. Ram, Mater. Sci. Eng. A 304-306 (2001) 923–927.

  40. [40]

      H.Y. Wang, S.R. Li, X. Wang, et al., Sci. China Chem. 64 (2021) 959–963. doi: 10.1007/s11426-021-9986-1

• 

  Figure 1  (a–d) View of different structural motifs in 1. (e) The structure of 1. CaO₈: purple; WO₆: red; SiO₄: yellow; NiO₆: green.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  2D structure diagrams of 1 (a) in a direction and (b) in c direction. CaO₈: purple; WO₆: red; SiO₄: yellow; NiO₆: green; Sb: blue; Na: yellow; O: red.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Nyquist plots for 1 (a) at different RHs and T = 25 ℃; (b) at different temperatures and 98% RH. (c) Arrhenius plots and the linear fit of 1 at 98% RH. (d) The PXRD patterns of the sample before and after proton conductivity test for 1.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (a–c) Nyquist plots at different RHs and T = 25 ℃. (d–f) Nyquist plots at different temperatures and 98% RH (inset: proton conduction device) for single-crystal of 1 along [001], [110], and [100] directions, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Arrhenius plots and the linear fit of single-crystal 1 along [001], [110], and [100] directions at 98% RH, respectively.

  下载: 全尺寸图片 幻灯片
•