View PDF VersionPrevious ArticleNext Article
 
DOI: 10.1039/D5QI00961H (Research Article) Inorg. Chem. Front., 2025, 12, 5016-5028

Mixed addenda polyoxometalates by cooperative self-assembly and modulation of their optoelectronic properties†

Ganga Singh , Ruchika Choudhary and Debaprasad Mandal *
Department of Chemistry, Indian Institute of Technology Ropar, Rupnagar, Punjab 140001, India. E-mail: dmandal@iitrpr.ac.in

Received 15th April 2025 , Accepted 28th June 2025

First published on 2nd July 2025

Abstract

Orbital engineering through the cooperative effect of different transition metals (TMs) is a powerful and versatile approach for modulating the chemistry of polyoxometalates (POMs), either by introducing novel POM structures or by designing POM hybrids for more effective catalytic applications. Here, we present a cooperative mixed-metal strategy for the synthesis of mixed-addenda (Mo/W) sandwich POMs with varying compositions, denoted by the general formula [(TMi)2(TMe)2(H2O)2(XMoxW9−xO34)2]n− (TMe(2+) = Fe/Co/Ni/Zn and TMi(3+) = Mn/Fe, X = Zn/Co/Fe). Using this cooperative mix-metal strategy, overall, we report 24 new POMs, including 8 mixed-addenda, 12 W-based sandwich POMs, and 4 POM-based 1-D coordination frameworks. Structural analyses reveal that Mo-addenda incorporation into the POM framework, alongside W (Mo/W), is strongly influenced by the variation of the transition metal composition at the sandwich core, their oxidation states, and the pH of the reaction media. Electrospray ionization mass spectrometry (ESI-MS) and energy dispersive X-ray (EDAX) analysis confirm the detailed POM compositions, while UV-vis spectroscopy and complementary density functional theory (DFT) analysis provide insights into orbital engineering via distinctive charge transfer processes. Theoretical and electrochemical studies demonstrate that electron transfer modulation occurs through both mixed-addenda incorporation and mixed-metal substitution at the sandwich position. This is further elucidated by enhanced oxygen evolution (OER) activity, where the cooperative mixed-metal and mixed-addenda POMs exhibit significantly improved performance, with an overpotential of 500 mV at 1 mA cm−2, compared to 570 mV in a pH 7.1 buffer. Additionally, this cooperative mixed-metal, mixed-addenda strategy extends to the formation of 1-D polyoxometalate coordination frameworks (POMCFs), where the oxidation state of precursor metals plays a vital role in determining the overall structural attributes.

Introduction

Polyoxometalates1 (POMs), with their diverse structural architectures and properties, are applied across diverse fields such as catalysis, nanotechnology, magnetism, materials science, and medicine.2,3 Our particular focus is on exploring the binding affinity of various transition metals to lacunary POMs4–8 to form transition metal substituted sandwich POMs (TMSPs); specifically, tetrasubstituted TMSPs with the formula [(TM)4(H2O)2(XW9O34)2]n− (where TM = transition metals, X = P/Zn/Co, etc.).9,10 By selecting appropriate addenda atoms and substituting transition metals, the electronic structures of these TMSPs can be tuned, resulting in distinct positions for the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO). These TMSPs with transition metals at the sandwich core and W as addenda atoms exhibit robust structural diversity.11,12 Incorporation of other addenda atoms (Mo and V) into TMSPs remains challenging, possibly due to higher kinetic lability. Recently, using a protecting group strategy, a Mo-based tetrasubstituted TMSP was isolated in acetonitrile.5 Conversely, a mixed addenda approach by the inclusion of small amounts of Mo/V/Nb into W-based Keggin and Dawson POMs is reported for modulating the HOMO–LUMO levels, which significantly impacts molecular properties, such as improving absorption in the visible region.13–16 The Cronin group has reported mixed addenda (Mo, W) TMSP, [Co4(H2O)2(PMoxW9−xO34)2]n−, thus overcoming the sluggish kinetics in the oxygen evolution reaction (OER).17

Recently, we have also demonstrated fine-tuning of the optical and redox properties of P-based mixed addenda sandwich POMs, [(TM)4(H2O)2(PMoxW9−xO34)2]n−, which opens up a wide range of possibilities for the synthesis of these TMSPs with desired properties.18 Conversely, another set of TMSPs with d-block elements as heteroatoms [(TM)4(H2O)2(XW9O34)2]n− (where TM = X = Mn/Fe/Ni/Cu) and [W(TM)3(H2O)2(XW9O34)2]n− (where TM = X = Co/Zn), with W addenda atoms have been reported.19–22 Despite these advantages, a mixed addenda POM with the incorporation of Mo, V, or Nb alongside W in these tetrasubstituted TMSPs remains unexplored, which holds potential for fine-tuning the optical and redox properties. Alternatively, a mixed transition metal approach is widely used for the synthesis of W-only Keggin and Dawson-based tetrasubstituted sandwich, banana, cap, and tetramer TMSPs with improved catalytic activity. For example, [Ni2Co2(H2O)2(PW9O34)2]10− and [Fe2Co2(H2O)2(CoW9O34)2]14− have been reported as highly active OER catalysts.23–27 However, a systematic study of transition metal variations at different positions, such as the heteroatom position and sandwich (internal and external) positions, and the possible role of mixed metals in Mo addenda incorporation into the framework, and its impact on their band gap modulation, is still lacking. Also, the effect of the charge densities of lacunary POMs (overall charge divided by the number of atoms, Table 1) on binding with transition metals remains unexplored in these TMSPs, as observed in Nb-based POMs.28,29

Table 1 Charge densities of lacunary POMs with different heteroatoms (bold)
Formula Charge (−ve) (non-H) atoms Charge density (charge/atoms)
ZnW9O34 or ZnMoxW9−xO34 12 44 0.27
CoW9O34 or CoMoxW9−xO34 12 44 0.27
FeW9O34 or FeMoxW9−xO34 11 44 0.25
PW9O34 or PMoxW9−xO34 9 44 0.20
ZnW12O40 6 53 0.11
PW12O40 3 53 0.05

Thus, we propose TMSPs with mixed addenda to leverage the combined advantages of W and Mo along with various transition metals at the sandwich core and heteroatom position (Scheme 1). In this context, our efforts focus on the systematic approach for the synthesis of mixed metal, mixed-addenda TMSPs featuring late 3d transition metals (Mn, Fe, Co, Ni, Cu, and Zn) with the general formula [(TMe)2(TMi)2(H2O)2(XMoxW9−xO34)2]14− (TMe = X = Zn/Co/Fe, TMi = Mn/Fe). Moreover, the inclusion of mixed addenda in TMSPs alters the electron density of the framework, which can serve as a template for the construction of POM-based 1-D coordination frameworks30–33 (POMCFs). All these POMs are characterized using single crystal X-ray diffraction (SC-XRD), ESI-MS, Fourier transform infrared (FT-IR) and Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), thermogravimetric analysis (TGA), EDAX, powder X-ray diffraction (PXRD) and their role in dioxygen activation34–38 has been emphasized. Furthermore, electronic and optical properties are elucidated through electrochemical, UV-vis spectroscopy and DFT studies.39–42


image file: d5qi00961h-s1.tif
Scheme 1 Transition metal based tetrasubstituted sandwich POMs (W only and Mo/W based) indicating the role of transition metals at internal sandwich (TMi), external sandwich (TMe), and heteroatom positions (X).

Experimental section

Materials and methods

Deionized water used in the experiments was from a Millipore system (>18 MΩ cm−1). The transition metal salts used in the synthesis were obtained from Alfa Aesar (purity > 99.9%) and used without further purification. All polyoxometalates were recrystallized from an aqueous solution before characterization. All the synthetic procedures were optimized in terms of pH and the different ratios of Mo to W salt as mentioned in our previously reported article.18 The optimization in the case of mixed addenda POMs was done regarding the maximum Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio in the final compound using a pH dependent study. The yields reported are from the first crystallization only. The second and subsequent crystallizations may result in impurities in such complex mixtures.

Single crystal X-ray diffraction

The molecular structures of all the sandwich POMs were revealed by SC-XRD analysis.43,44 Diffraction quality crystals were coated with Paratone-N oil, and mounted on a cryo-loop, and single-crystal X-ray data were collected using graphite mono-chromated Mo Kα radiation (λ = 0.71073 Å) on a Bruker D8 SMART APEX4 CMOS diffractometer at 298 K. Data integration was performed using scalable automated integrations of reflections (SAINT) and APEX-4 software. Routine Lorentz and polarization corrections were applied for all structures and absorption corrections (multi-scan/numerical) were performed using SADABS. The Olex2 suite graphical user interface was used to solve all the structures by employing intrinsic phasing methods with the SHELXT program, and refinement was performed using SHELXL by the full matrix least-squares method. The differentiation of Mo/W in the crystal structure is done based on the electron density difference map while refinement is performed using Olex2 software. The electron density contour map along with refinements (variation in R1/wR2 values) was used to assign the crystallographic mixed addenda (Mo/W) position in the framework. Crystal data, data collection parameters, and refinement statistics have been deposited at the Cambridge Crystallographic Data Centre; CCDC numbers 2330238, 2330239, 2330240, 2330242, 2330243, 2330244, 2330245, 2330246, 2330247, 2330248, 2330249, 2330250, 2330252, 2330253, 2330254, 2330255, 2330256, 2330257, 2330258, 2330261.†

Density functional theory (DFT) calculations

DFT calculations were performed using Gaussian 09 programs with Becke's three-parameter hybrid exchange functional and the Lee–Yang–Parr correlation functional (B3LYP). The double-ζ basis set of Hay and Wadt (LanL2DZ) with an effective core potential (ECP) was used for W, Mo, Zn, Cu, Ni, Co, Fe and Mn to represent the innermost electrons of these atoms, and the O atom was described using the 6-31G+(d, p) basis sets. The calculations were performed considering the solvation effects of water in a conductor-like screening (CPCM) model. The closed-shell geometry optimization calculations in all the reported POMs were performed using the atomic coordinates provided by the X-ray structures of the respective anions except a few structures (mentioned as auxiliary structures in the text). The DFT-optimised structure of these POMs reveals the singly occupied molecular orbitals (SOMOs) lying between the highest occupied molecular orbitals (HOMOs, located at bridging oxygens) and the lowest unoccupied molecular orbitals (LUMOs, situated at W/Mo).

Electrochemical studies

Cyclic voltammograms were recorded on a Metrohm Autolab M204 and Nova 2.1.5 software using a standard three-electrode set-up equipped with a 3 mm diameter glassy carbon working electrode (WE), a Pt wire counter electrode (CE) and Ag/AgCl/3 M KCl as the reference electrode (RE) using a 0.2 mM POM at a 50 mV s−1 scan rate. The catalytic oxygen evolution reaction was studied using a three-electrode setup consisting of Ag/AgCl/3 M KCl (reference electrode), Pt wire (counter electrode), and Cs salt of POMs coated on glassy carbon (working electrode) in 25 mL phosphate buffer (50 mM) of pH 7.1 at a scan rate of 50 mV s−1 (detailed electrode preparation is given in the ESI†).

Results and discussion

TMSPs with TMs as heteroatoms

TMSPs with the same transition metals in the sandwich as well as the heteroatom position, [(TM)4(H2O)2(XW9O34)2]n−, are well known11 when TM = X = Mn/Fe/Ni/Cu; however, analogous mixed addenda of these structures are still unknown. In this context, first we synthesized mixed addenda Na14[Fe4(H2O)2(FeMoxW9−xO34)2]·20H2O (1-M, TM = X = Fe) analogous to Na14[Fe4(H2O)2(FeW9O34)2]·21H2O (1) obtained from FeCl2. 1-M has been synthesized via microwave heating at 85 °C at pH 6.6; increasing the pH results in the formation of 1, while lowering the pH does not yield the desired product (details in Fig. S2–S10, ESI†). ESI-MS of 1 (aq. solution) shows the isotopic single envelope of peaks at m/z 1580 (z = –3), which shifts to isotopic multiple envelopes of peaks in the range of m/z 1400–1550 (z = –3) for 1-M with different Mo/W ratios (Fig. S3 and S4, ESI†), thus indicating the incorporation of Mo addenda into the POM framework. The most abundant anion of 1-M has the composition of Mo[thin space (1/6-em)]:[thin space (1/6-em)]W (3[thin space (1/6-em)]:[thin space (1/6-em)]15) with the formula H7[Fe4(H2O)2(FeMo1.5W7.5O34)2]3− while the Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio varies from 1[thin space (1/6-em)]:[thin space (1/6-em)]17 to 6[thin space (1/6-em)]:[thin space (1/6-em)]12 in the cluster. Bulk composition and phase purity were further confirmed by EDAX and PXRD analyses, respectively (Fig. S7, ESI†). Due to Mo addenda incorporation, FT-IR (W = Oterm at 931 cm−1 in 1 and W/Mo = Oterm at 911 cm−1 in 1-M) peaks shift to lower values (Fig. S8a, ESI†). A SC-XRD study confirms that light yellow-colored crystals of both 1 and 1-M crystallize in the P21/n space group. Bond valence sum (BVS) calculations45 indicate that Fe is in a +3 oxidation state in 1-M, which was further confirmed by XPS analysis (Fig. S5 and S6, ESI†). Mixed-addenda sandwich POMs are mixtures of POMs with different Mo/W compositions, which have been referred to as mixed-addenda in the manuscript.

However, Fe is in a +2 oxidation state at external sandwich positions (TMe) and +3 at the heteroatom and internal sandwich position (TMi) in 1. The characteristic Fe–O bond lengths are between 1.8 and 2.1 Å with X-Oμ4-TMi (X = TMi = Fe) distances found to be 3.87 Å. FeCl2 plays a crucial role in incorporating mixed addenda into these POMs as FeCl3 as a precursor favors precipitation after some time, which could not be analyzed further. FeCl2 over FeCl3 not only facilitates crystallization but also maintains the required solution pH; i.e., FeCl3 drastically decreases the pH to less than 4 from 6.6, whereas the pH decreases to 5.0 in FeCl2. The crystallization of 1-M is slow and takes approximately 15 days with KCl. The use of Fe3+ instead of Fe2+ results in fast precipitation rather than gradual crystallization.

Effect of the TM at the internal sandwich position

Interestingly, sandwich POMs with the formula [(TM)4(H2O)2(XW9O34)2]n−, TM = X = Mn/Fe/Ni/Cu are well known, but when TM = X = Co/Zn, e.g., [Zn4(H2O)2(ZnW9O34)2]n− and [Co4(H2O)2(CoW9O34)2]n− are unknown. Instead of the (TM)4 sandwich core, one W(+6) atom incorporation at the sandwich position for TM = X = Co/Zn, e.g., [W(TM)3(H2O)2(XW9O34)2]12−, was always observed, which must be due to the cooperative effect/interplay between the transition metals at the sandwich core and heteroatom.46–48 In several sandwich POMs having 3d transition metals as the heteroatom, we observed that the X–Oμ4–TMi bond distances connected through Oμ4 (Fig. 1a) are generally between 3.87 and 4.0 Å, and the structures with an X–Oμ4–TMi bond length >4.0 Å are found to be distorted as observed in reported structures of Bi-POMs (Table S5, ESI†).49,50
image file: d5qi00961h-f1.tif
Fig. 1 (a) Combined polyhedral and ball-stick representation of 2 (left) and 2-M (right) with μ2, μ3, and μ4 oxygen atoms (counter cations and hydrogen atoms are omitted for clarity); ESI-MS envelope of peaks with the assigned formula table for 2-M (full ESI-MS and simulated spectra in Fig. S14, ESI†). (b) Frontier molecular orbitals from the DFT optimized structure with the theoretically calculated band gap and (c) solid-state UV-vis spectra and (d) corresponding optical band-gap derived from the Kubelka–Munk (K–M) function in 2 and 2-M.

Furthermore, DFT optimization of the auxiliary structure of [Zn4(H2O)2(ZnW9O34)2]16− suggests that the X–Oμ4–TMi distance of 4.03 Å with one of the Zn–O bond length (3.01 Å) elongation results in the distortion of the structure (Fig. S11, ESI†). Moreover, it was observed that the introduction of W(VI) in the sandwich position of these POMs not only enhances the structural stability but also facilitates the substitution of TMs at the external sandwich position within these sandwich POMs.46,51,52 Similarly, other higher oxidation state transition metals at the internal sandwich position also enhance the structural stability in these mixed metal TMSPs. For example, Fe3+ and Ru4+ substituted TMSPs, [Co2Fe2(H2O)2(CoW9O34)2]14− and [Zn2Ru2(H2O)2(ZnW9O34)2]12−, are reported to be highly stable under the strong oxidizing conditions of the OER.25,53 The good structural stability and TM substitution possibilities at the external sandwich position in Zn and Co-based TMSPs motivated us to exploit the potential impact of mixed TMs, particularly with those with a high oxidation state TM at the sandwich position. In this context, different transition metals such as V4+, Cr3+, Mn3+, Fe3+, Y3+, and Zr4+ were tested for sandwich POMs with Zn/Co as the heteroatom, out of which Mn3+ and Fe3+ showed excellent stability when taken as mixed metal precursors along with Zn/Co salts. Other TMs either lead to an insoluble precipitate or form [Zn3W(H2O)2(ZnW9O34)2]12− under our conditions (pH range: 6–8). Four compounds with the general formula [(TMe)2(TMi)2(H2O)2(XW9O34)2]14− (where TMi = Fe3+/Mn3+, X = TMe = Zn/Co) were obtained in situ from a mixture of Zn/Co, Fe/Mn and W salt solution after microwave heating at 80 °C for 45 min at pH 7.8. Except for 3,25 all these POMs are new and the obtained molecular formulas are Na14[(Zne)2(Fei)2(H2O)2(ZnW9O34)2]·23H2O (2); Na14[(Coe)2(Fei)2(H2O)2(CoW9O34)2]·15H2O (3); Na14[(Zne)2(Mni)2(H2O)2(ZnW9O34)2]·24H2O (4); Na14[(Coe)2(Mni)2(H2O)2(CoW9O34)2]·22H2O (5).

SC-XRD indicates that all these TMSPs crystallized in 2 h as monoclinic (P21/n) lattices except for 2, which crystallized in a triclinic (P[1 with combining macron]) lattice (details, Fig. S12–S46†). BVS calculations indicate that all TMi atoms are in +3 and TMe/X is in a +2 oxidation state with the characteristic TMi–O bond length average found to be 2.04 Å. In the case of 4/5, two TMi–O bond lengths are found to be longer (2.30 Å) than the average of the other four TMi–O bond lengths (1.94 Å) due to z-out Jahn Teller distortions in 3d4 {Mn3+} configurations. The X–(Oμ4)–TMi bond distances were found to be 3.86–3.93 Å. These POMs were then characterized using FT-IR (W = Ot at 900–930 cm−1 and W–Oc–W at 835–840 cm−1) and ESI-MS confirms the fragments of H11[2] at m/z 1594 (z = –3), H11[3] at 1585 (z = –3), H11[4] at 1593 (z = –3), and H11[5] at 1584 (z = –3). From the SC-XRD of 2, it was observed that Fe(+3) is present at the internal sandwich position, while Zn is located at the heteroatom position as well as the external sandwich position. This was further confirmed by ESI-MS, which shows only a single peak at m/z 1594 (z = −3) ascribed to the presence of H11[Zn2Fe2(ZnW9O34)2]3−. The FT-IR study of O2-activated 2/3 indicates dioxygen activation with two characteristic peaks at 1149 and 1203/1205 cm−1, which was further confirmed by Raman spectra (peaks at 722 and 1063 cm−1) (Fig. S17a/S27b, ESI†).

This cooperative mixed-metal approach seems useful for the synthesis of mixed-metal sandwich POMs, and we have extended it for the synthesis of mixed addenda with Zn/Co as the heteroatom. Using Fe3+ along with Zn2+/Co2+ as mixed transition metals results in the formation of mixed metal-mixed addenda sandwich POMs with the formula Na14[Zn2Fe2(H2O)2(ZnMoxW9−xO34)2]·23H2O (2-M) and Na14[Co2Fe2(H2O)2(CoMoxW9−xO34)2]·19H2O (3-M). The SC-XRD of 2-M (light yellow-colored) and 3-M (purple-colored) confirms the structures with the space group P21/n (Fig. 2). The distribution of the Mo addenda shows preferential binding with μ2 oxygen connected to substituted transition metals. The higher binding affinity of Mo to μ2 oxygens with lower charge density compared to other oxygens (μ3 and μ4) can be attributed to its lower effective nuclear charge and higher π-acceptor ability compared to W. Shifts in characteristic FT-IR peaks in 2/3 indicate Mo incorporation into the POM framework (Fig. S17a/S27c, ESI†).


image file: d5qi00961h-f2.tif
Fig. 2 Combined polyhedral and ball-stick representation of SC-XRD structures and corresponding optical images of 2, 2-M, 3, 3-M, 4, 4-M, 5, and 5-M (green – pure W, yellowish green – varying Mo/W ratios, and counter cations and hydrogen atoms are omitted for clarity).

In 2-M, Fe competes with Zn at the heteroatom position. BVS indicates Fei in +3, Zne in +2, and Zn at the heteroatom in +2 with 0.5 discrepancies indicating the presence of Fe(+3) along with Zn in 2-M, which was further confirmed by ESI-MS (Tables S10 and S13, ESI†) and 57Fe Mössbauer spectroscopy showing the presence of two types of Fe in the structure (two singlets) (Fig. S17d, ESI†). In ESI-MS, the isotopic single envelope of peaks at m/z 1594 (z = –3) for 2 shifts to isotopic multiple envelopes of peaks ranging from m/z 1340 to 1565 (z = –3) for 2-M due to the incorporation of Mo addenda atoms. Similarly, the isotopic single envelope of peaks at m/z 1585 (z = –3) for 3 shifts to isotopic multiple envelopes of peaks ranging from m/z 1320 to 1550 (z = –3) in 3-M. The most abundant anion in 2-M gave the composition of Mo[thin space (1/6-em)]:[thin space (1/6-em)]W (2[thin space (1/6-em)]:[thin space (1/6-em)]7) with the formula H9[Zn2Fe2(H2O)2(Fe2Mo4W14O34)]3−, while the Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio varies from 1[thin space (1/6-em)]:[thin space (1/6-em)]17 to 8[thin space (1/6-em)]:[thin space (1/6-em)]10 for z = –3. EDAX and PXRD further confirm the bulk purity of 2-M (Fig. S16, ESI†). In 3-M, the most abundant anion has a Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio of 5[thin space (1/6-em)]:[thin space (1/6-em)]13 with the formula H9[Co2Fe2(H2O)2(Fe2Mo5W13O34)]3−, while the Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio varies from 3[thin space (1/6-em)]:[thin space (1/6-em)]15 to 7[thin space (1/6-em)]:[thin space (1/6-em)]11 for z = –3 (Fig. S25, ESI†).

We have also performed pH-dependent studies, which indicate that the Mo/W ratio increases for the most abundant peak with decreasing pH from 7.5 to 6.0 for 2-M and 3-M. However, a further decrease in pH results in the formation of the Keggin framework, indicating the crucial role of pH in obtaining maximum Mo incorporation into the framework (details, Fig. S15/S26, ESI†). BVS indicates Fei in +3, Coe in +2, and Co at the heteroatom in +2 with 0.5 discrepancies, supporting the presence of Fe(+3) along with Co2+ in 3-M similar to 2-M, which was confirmed by XPS analysis, otherwise difficult to predict from SC-XRD and ESI-MS. The UV-visible spectra show the LMCT, d–d transition and small intensity peaks (at 473/480 nm) due to metal to POM charge transfer (MPCT) in 2/3 and 2-M/3-M. The band gap calculated from the Kubelka–Munk plot was found to be less in the case of 2-M/3-M when compared to 2/3 (e.g., 3.03 eV in 2 to 2.68 eV in 2-M; see Fig. 1c and S25, ESI†). The DFT optimization of mixed addenda TMSPs shows that the LUMO lies at Mo atoms rather than W (Fig. 1a and b, detailed discussion on 2/2-M, ESI†), resulting in the reduction of the band gap, which is also observed from UV-visible spectra.

Using a similar approach, mixed addenda were introduced into Zn/Mn3+ or Co/Mn3+ based sandwich POMs. SC-XRD and ESI-MS confirm the molecular formula Na14[(Zne)2(Mni)2(H2O)2(ZnMoxW9−xO34)2]·22H2O (4-M) and Na14[(Coe)2(Mni)2(H2O)2(CoMoxW9−xO34)2]·19H2O (5-M), with only a little Mo addenda incorporation (the highest Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio obtained: 2[thin space (1/6-em)]:[thin space (1/6-em)]16). SC-XRD reveals Mo at only two addenda positions connected with the transition metal at the external sandwich position through μ2 oxygen (Fig. S32/S44, ESI†). Thus, we observed changing the TM at the internal sandwich position, i.e., when we go from 2 to 4 (3–5, Tables 2-A and 3-A), the optical band gap decreases, which was also observed in the DFT study, indicating the significant effect of the change in the transition metal at the internal sandwich position. Also, going from 2-M to 4-M, it was observed that the TM at the internal sandwich position impacts the Mo incorporation into the framework.

Table 2 Effect of Mo incorporation, charge density of lacunary POMs, position and oxidation states of TMs towards the optical band gap and Mo/W ratio
POMs (reference code) Optical band gap from DRS (eV) Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio
Abundant Mo[thin space (1/6-em)]:[thin space (1/6-em)]W Maximum Mo[thin space (1/6-em)]:[thin space (1/6-em)]W
A. Effect of changing the transition metal at internal sandwich positions
2 vs. 4 3.03 vs. 2.31 — —
3 vs. 5 1.98 vs. 1.37 — —
2-M vs. 4-M 2.68 vs. 1.65 4[thin space (1/6-em)]:[thin space (1/6-em)]14 → 1[thin space (1/6-em)]:[thin space (1/6-em)]17 7[thin space (1/6-em)]:[thin space (1/6-em)]11 → 2[thin space (1/6-em)]:[thin space (1/6-em)]16
3-M vs. 5-M 1.73 vs. 1.29 5[thin space (1/6-em)]:[thin space (1/6-em)]13 → 1[thin space (1/6-em)]:[thin space (1/6-em)]17 7[thin space (1/6-em)]:[thin space (1/6-em)]11 → 2[thin space (1/6-em)]:[thin space (1/6-em)]16
B. Effect of changing the transition metal at external sandwich positions
2a → 2b → 2c 1.80 → 2.26 → 2.41 — —
3a → 3 → 3b → 3c 1.65 → 1.98 → 2.59 → 2.66 — —
C. Effect of change in heteroatoms
2 vs. 3d 3.03 vs. 2.02 — —
2b vs. 3 2.26 vs. 1.98 — —
D. Effect of charge density (sandwich to 1-D framework)
3 vs. 7 1.98 vs. 2.07 — —
2-M vs. 6-M 2.68 vs. 2.43 4[thin space (1/6-em)]:[thin space (1/6-em)]14 → 1[thin space (1/6-em)]:[thin space (1/6-em)]17 7[thin space (1/6-em)]:[thin space (1/6-em)]11 → 2[thin space (1/6-em)]:[thin space (1/6-em)]16
E. Effect of changing the heteroatom along with external sandwich positions (using the same metal)
2 vs. 3 3.03 vs. 1.98 — —
4 vs. 5 2.31 vs. 1.37 — —
2-M vs. 3-M 2.68 vs. 1.73 4[thin space (1/6-em)]:[thin space (1/6-em)]14 → 5[thin space (1/6-em)]:[thin space (1/6-em)]13 7[thin space (1/6-em)]:[thin space (1/6-em)]11 → 7[thin space (1/6-em)]:[thin space (1/6-em)]11
4-M vs. 5-M 1.65 vs. 1.29 1[thin space (1/6-em)]:[thin space (1/6-em)]17 → 1[thin space (1/6-em)]:[thin space (1/6-em)]17 2[thin space (1/6-em)]:[thin space (1/6-em)]16 → 2[thin space (1/6-em)]:[thin space (1/6-em)]16
F. Effect of Mo incorporation into the framework
1 vs. 1-M 2.24 vs. 1.82 3[thin space (1/6-em)]:[thin space (1/6-em)]15 (1-M) 6[thin space (1/6-em)]:[thin space (1/6-em)]12 (1-M)
2 vs. 2-M 3.03 vs. 2.68 4[thin space (1/6-em)]:[thin space (1/6-em)]14 (2-M) 7[thin space (1/6-em)]:[thin space (1/6-em)]11 (2-M)
3 vs. 3-M 1.98 vs. 1.73 5[thin space (1/6-em)]:[thin space (1/6-em)]13 (3-M) 7[thin space (1/6-em)]:[thin space (1/6-em)]11 (3-M)
4 vs. 4-M 1.84 vs. 1.65 1[thin space (1/6-em)]:[thin space (1/6-em)]17 (4-M) 2[thin space (1/6-em)]:[thin space (1/6-em)]16 (4-M)
5 vs. 5-M 1.37 vs. 1.29 1[thin space (1/6-em)]:[thin space (1/6-em)]17 (5-M) 2[thin space (1/6-em)]:[thin space (1/6-em)]16 (5-M)

Table 3 Effect of Mo incorporation, charge density of lacunary POMs, positions and oxidation states of TMs on the electronic band gap and position of the highest SOMO and LUMO (from DFT)
POMs Band gap (eV) LMCT (O → W/Mo)/MPCT (SOMO → W/Mo) Atom/group of atoms contributing to LUMO position (eV)
Highest SOMO LUMO
A. Effect of changing the transition metal at the internal sandwich position
2 vs. 4 4.62/4.09 vs. 4.01/3.78 Fe3+vs. Oc; Mn3+ W vs. Mn3+ −1.51 vs. −2.09
B. Effect of the transition metal at the external sandwich position on MPCT
2a → 2b → 2c → 2d 3.35 < 3.76 < 3.85 < 4.16 Mn/Fe3+; Fe3+/Zn; Fe3+; Fe3+ W only for all −1.57 → −1.52 → −1.57 → −1.62
3a → 3 → 3b → 3c → 3d 3.26 < 3.64 = 3.64 < 3.76 < 3.98 Mn/Fe3+; Fe3+/Co; Fe3+/Co; Fe3+/Co; Fe3+/Zn W only for all −1.58 → −1.58 → −1.58 → −1.64 → −1.53
C. Effect of change in the heteroatom on MPCT
2 vs. 3d 4.09 vs. 3.98 Fe3+; Fe3+/Zn W only −1.51 vs. −1.53
3 vs. 2b 3.64 vs. 3.76 Fe3+/Co; Fe3+/Zn W only −1.58 vs. −1.52
D. Effect of change in the heteroatom along with the transition metal at the external sandwich position
2 vs. 3 4.62/4.09 vs. 4.70/3.64 Fe3+ for both W for both −1.51 vs. −1.58
2-M vs. 3-M 4.12/3.73 vs. 4.32/3.29 Fe3+ for both Mo for both −1.94 vs. −1.95
E. Effect of Mo incorporation: W-only vs. W/Mo TMSPs
2 vs. 2-M 4.62/4.09 vs. 4.12/3.73 Fe3+ for both W vs. Mo −1.51 vs. −1.94
3 vs. 3-M 4.70/3.64 vs. 4.32/3.29 Fe3+ for both W vs. Mo −1.58 vs. −1.95
4 vs. 4-M 4.01/3.78 vs. 3.97/3.36 Oc + Mn3+; Oc + Mn3+ W vs. Mn −2.09 vs. −2.07
6 vs. 6-M 4.63/4.12 vs. 4.32/3.83 Fe3+ for both W vs. Mo −2.09 vs. −2.41

Effect of the precursor oxidation state/charge density of lacunary POMs

Additionally, we observed that the oxidation state of the TM precursors significantly influences the charge density of the bridge and terminal oxygens, subsequently impacting these mixed addenda sandwich POMs. To understand the effect, Fe2+/Zn2+ were taken as precursors instead of Fe3+/Zn2+ in 2-M. Strikingly, SC-XRD and ESI-MS reveal a much lower Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio in the POM framework compared to 2-M. The SC-XRD structure confirms the formation of POMCFs, [Fe4(O)2(ZnMoxW9−xO34)2]n− (6-M), where the oxygen linked to the external Fee is bridged to a tetrahedral molybdate (MoO4), thus extending the link through a Zn atom to form POMCFs with a ZnMo2O8 linker (Fig. 3a).
image file: d5qi00961h-f3.tif
Fig. 3 (a) Different views of the 1-D framework of 6-M and (b) 1-D framework of 7 (counter cations and hydrogen atoms are omitted for clarity).

BVS calculations indicate that all Fe atoms are in +3, Zn in +2 and Mo/W in a +6 oxidation state. Using Fe2+/Zn2+ precursor salts instead of Fe3+/Zn2+ in W-only POM, 2, resulted in rapid precipitation of microcrystalline POMCFs, which are difficult to characterize via SC-XRD. A similar POMCF formation was observed when Co2+/Fe2+ precursor salts were used instead of Co2+/Fe3+ in W-only POM, 3, and when Zn2+/Mn2+ salts replaced Zn2+/Mn3+ in 4. For example, 1-D POMCFs were obtained for [Co2Fe2(H2O)2(CoW9O34)2]14− (7) with Fe2+ as a linker and [ZnWMn2(H2O)2(ZnW9O34)2]12− (8) with Mn2+ as a linker between two POM frameworks (details in Fig. S47–S62, ESI†). BVS calculations indicate Fe3+ occupies the sandwich position, Fe2+ outside the POM framework, Co and Mn in +2, and W in +6 oxidation states. Additionally, the cation exchange of 3 with excess Fe2+ led to the formation of an insoluble crystalline powder with matching PXRD (Fig. S56d, ESI†), indicating POMCF formation.

To explore the influence of high negative charge density on the POM structure resulting from the low oxidation state TM (Fe2+), we conducted the DFT optimization of the auxiliary structures [(TMe)2(TMi)2(H2O)2(ZnW9O34)2]16− (TMe = Zn2+ or Fe2+, TMi = Fe2+). The optimized structures of [Fe4(H2O)2(ZnW9O34)2]16− (TMe = TMi = Fe2+, X = Zn2+) and [Zn2Fe2(H2O)2(ZnW9O34)2]16− (TMe = X = Zn2+, TMi = Fe2+) revealed an elongation of the TMe–H2O bond due to increased charge density (Fig. S60, ESI†). This elongation further increased in the case of mixed addenda POMs (representative structure of 6-M; [Fe4(H2O)2(ZnMo4W5O34)2]16−, Fig. S60c, ESI†). These findings indicate the crucial role of the precursor's oxidation states/charge density of lacunary POMs in influencing the self-assembly towards POMCFs as well as Mo incorporation; e.g., when we go from 2-M to 6-M (Table 2-D), Mo incorporation decreases from 7[thin space (1/6-em)]:[thin space (1/6-em)]11 to 2[thin space (1/6-em)]:[thin space (1/6-em)]16.

Effect of the TM at the external sandwich position

Our interest was to further extend the mixed addenda scope by synthesizing trimetallic mixed metal sandwich POMs. We explored the substitution of transition metals at external sandwich positions in [WZn3(H2O)2(ZnW9O34)2]12− since the presence of W6+ at the sandwich core (WZn3) renders the external TMs susceptible to substitution with a 3d TM, resulting in the formation of [(TMe)2WZn(H2O)2(ZnW9O34)2]12− (TMe = Mn/Fe/Co/Ni/Cu).46,48 While an analogous mixed addenda structure with WZn3 at the sandwich core [WZn3(H2O)2{ZnMoxW9−xO34}2]12− was DFT optimized (Fig. S63, ESI†), however, the POM synthesis was unsuccessful. Moreover, side substitution with a 3d TM in Zn2+/Fe3+-based sandwich POMs 2 results in a mixture of products as characterized by ESI-MS (Fig. S64, ESI†).46 However, using mixed transition metal (Zn/Fe3+/TM and Co/Fe3+/TM) precursors in stoichiometric ratios, we could successfully synthesize trimetallic sandwich POMs with the general formula Na14[(TMe)2Fe2(H2O)2(XW9O34)2]·nH2O (TMe = Mn/Co/Ni/Cu/Zn; 2a/2b/2c/2d/2, respectively (Fig. 4a), when X = Zn2+, and 3a/3/3b/3c/3d when X = Co2+), which were characterized by SC-XRD, FT-IR, XPS, Raman, TGA and ESI-MS (Fig. S65–S84, ESI†) analyses. ESI-MS for Zn/Fe3+/TM based POMs exhibited peaks at m/z 1586.41, 1589.48, 1589.52, 1591.52, and 1593.43 due to the fragment H11[(TMe)2Fe2(H2O)2(ZnW9O34)2]3−, where TMe = Mn/Co/Ni/Cu/Zn; (2a/2b/2c/2d/2), (Fig. 4b), respectively. A similar trend of m/z values at 1583.51, 1585.17, 1585.14, 1587.51, and 1589.48 was observed for Co-heteroatom-based POMs due to the fragment H11[(TMe)2Fe2(H2O)2(CoW9O34)2]3− (Fig. S75, ESI†).
image file: d5qi00961h-f4.tif
Fig. 4 (a) SC-XRD structure of 2 (left) and representative trimetallic sandwich POMs 2a/2b/2c/2d (right) (counter cations and hydrogen atoms are omitted for clarity). (b) Optical images and ESI-MS with a single envelope of peaks along with the assigned formula table in 2a/2b/2c/2d/2 (full range ESI-MS spectra, zoomed part of the highest intensity peak of 2 (m/z = 1593.43 with z = –3) and its corresponding simulated spectra in Fig S65, ESI†).

Mixed addenda of these trimetallic sandwich POMs were also synthesized in situ directly from their precursors, e.g., Na14[Ni2Fe23+(H2O)2(ZnMoxW9−xO34)2]·12H2O (2c-M) and Na14[Ni2Fe23+(H2O)2(CoMoxW9−xO34)2]·28H2O (3b-M), which were synthesized from Ni/Fe/TM (TM = Zn/Co) and Mo/W salts. The Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio was observed to be 5[thin space (1/6-em)]:[thin space (1/6-em)]13 in the most abundant peak of 3b-M in ESI-MS (Fig. S79 and Table S33, ESI†), which is similar to 3-M, thus indicating that the transition metal at external positions has a negligible effect on mixed addenda in these trimetallic sandwich POMs. The UV-vis spectra exhibit LMCT, MPCT and d–d transitions in the solid state (Fig. S68c, ESI†). This trimetallic sandwich POM approach was also applied to Mn3+ (Zn/Mn3+/TM) for the synthesis of Na14[(Fee)2(Mni)2(H2O)2(ZnW9O34)2]·25H2O (4a) (Fig. S85–S88, ESI†) and its mixed addenda analogue [(Fee)2(Mni)2(H2O)2(ZnMoxW9−xO34)2]14− (4a-M), where only a negligible Mo[thin space (1/6-em)]:[thin space (1/6-em)]W ratio (1[thin space (1/6-em)]:[thin space (1/6-em)]17) was observed in ESI-MS (Fig. S87, ESI†). From this, we observed that the transition metal at the external sandwich position impacts the optical band gap; however, the effect is comparatively less than that at the internal sandwich position (Tables 2-B and 3-B).

From the synthesis of these mixed-metal based W-only/mixed-addenda TMSPs, it was observed that two factors play a crucial role in the self-assembly: (a) the charge density of the trivacant lacunary part due to the heteroatom and (b) the cooperative effects of transition metals (TMs) at the sandwich position. This cooperative effect is especially prominent in TMSPs with higher charge density lacunary POMs with a transition metal at the heteroatom, which also allows Mo addenda incorporation into these TMSPs. In terms of Mo incorporation into these TMSPs, the cooperative effect of mixed metals like Zn/Fe3+ or Co/Fe3+ was found to be superior.

This is summarized in Fig. 5 as a diagram, and the following few essential points are discussed as the synthetic strategy for these TMSPs: (1) TMs as heteroatoms (Zn and Co) form high charge density trivacant lacunary POMs (e.g., ZnW9O34), while the P-heteroatom forms relatively low charge density trivacant lacunary POMs (e.g., PW9O34) (Table 1). (2) High charge density lacunary POMs with TMs as heteroatoms preferentially bind with TMs of high oxidation states (e.g., W6+, Ru4+, and Fe3+) at internal sandwich positions to form TMSPs. Meanwhile, using a low oxidation state TM with high charge density lacunary POMs results in the formation of 1-D frameworks instead of mixed metal at the sandwich position. For example, Zn as the heteroatom (i.e., ZnW9O34) forms TMSPs with Fe3+ in 3, while it forms a 1-D framework with Fe2+ in 7. (3) The P-heteroatom forms low charge density lacunary POMs preferentially binding with TMs of low oxidation states at both internal and external sandwich positions (e.g., Zn2+, Co2+, etc.).18 (4) In TM-heteroatom based TMSPs, the TM at internal sandwich positions has a pronounced impact on Mo incorporation into the mixed addenda as compared to the external sandwich position and heteroatom position. For example, Fe3+ favors more Mo incorporation into 2 than Mn3+ in 4 at the internal sandwich position.


image file: d5qi00961h-f5.tif
Fig. 5 (a) Different representative mixed-metal mixed-addenda sandwich POMs and 1-D framework; (b) classified synthesis scheme of mixed-metal mixed-addenda TMSPs (oxygen and H2O are omitted for clarity); (c) impact of pH on the Mo/W ratio of mixed addenda sandwich POMs; 1-M, 2-M, and 3-M. (d) Bar graph of the maximum Mo/W ratio and the corresponding Mo/W ratio in the most abundant peak of mixed addenda sandwich POMs; symbolic lines showing the pH of solution for W only and mixed addenda TMSP synthesis and (e) μ4-oxygen atom (Oμ4) connection with heteroatom (X) and internal transition metals (TMi) in TMSPs.

Following the synthesis of various mixed-metal, mixed-addenda TMSPs and theoretical analysis of these POMs using DFT, it is evident that orbital engineering and electronic bandgap align with the optical bandgap, as summarized in Table 2. The data in Table 2 indicate that modifying a transition metal at sandwich or heteroatom positions within these TMSPs or incorporating Mo addenda significantly impacts bandgap modulation. For example, when moving from right to left across a row of 3d transition metals (Zn to Mn), altering the transition metal at the individual sandwich (whether internal or external) or heteroatom position decreases the band gap (factor 1), a trend also supported by theoretical calculations (Table 3). Furthermore, a significant decrease in the bandgap was observed due to the combined effects of the heteroatom and the external sandwich position while moving from TMSP 2 to 3. Notably, the incorporation of Mo into mixed addenda POMs not only decreases the bandgap (factor 2) but also changes the LUMO position; e.g., the LUMO, which is located at W in TMSP 2, shifts entirely to Mo in 2-M. While both factors reduce the bandgap, the internal sandwich position (Fe to Mn) has a more pronounced impact on lowering the bandgap and providing a lower-lying LUMO compared to Mo incorporation, as supported by theoretical calculations (Table 3). The heteroatom effect is most pronounced with Co as the heteroatom, while the external sandwich position has comparatively less impact on the bandgap.

Following the trend of decreasing band gap, TMSPs with Co as the heteroatom and Mn at the internal sandwich position along with mixed addenda reveal the lowest band gap, which was indeed experimentally observed in 5-M (1.29 eV) among all the reported TMSPs here due to the combined effect of Co, Mn3+, and W/Mo. Similarly, 3-M shows a lower bandgap with Co as the heteroatom but higher than 5-M due to Fe instead of Mn at the sandwich position. Although the optical band gap is slightly lowered (by 0.2 eV) in 4-M compared to 4, DFT shows that the LUMO of 4-M remains entirely on Mn for both 4 and 4-M, thereby minimizing the effect of Mo incorporation (bandgap of −2.09 vs. −2.07 eV from DFT).

Electrochemistry and OER activity

Considering the significant Mo incorporation and the cooperative effect of transition metals in orbital engineering – lowering LUMO energy, as well as the HOMO–LUMO band gap in 2-M and 3-M, we investigated the impact of mixed-metal mixed-addenda on the catalytic activity of these POMs. First, we examined their influence on electrochemical properties through cyclic voltammetry (CV) for 2 and 2-M in an aqueous solution at pH 2.5, as these POMs do not exhibit redox peaks at pH > 5.5 and show only feeble peaks at pH 3.0. At even lower pH (typically pH < 1.5), these POMs show hydrogen evolution due to proton coupled electron transfer.

These POMs exhibit quasi-reversible redox couples in CV, characterized by their half-wave potentials (E1/2) and peak separations (ΔEp = |Epc − Epa|) (Fig. 6 and Table S38, ESI†). The CV of 2 shows quasi-reversible redox peaks of W, mostly on the negative side, while the 3d transition metal (Fe) at the sandwich position exhibits quasi-reversible peaks on the positive side, corresponding to the stepwise oxidation of the Fe2+/Fe3+ and Fe3+/Fe4+ couples vs. Ag/AgCl/3 M KCl. In the mixed-addenda analogue, peak assignment for Mo and W is challenging due to their similar redox potentials and multiple influencing factors. Mo-related peaks were distinguished by comparing the CV of W-only POMs (2) with that of mixed addenda (2-M). A comparison of CVs of 2 and 2-M (Fig. 6a/b and Table S38, ESI†) revealed that 2-M exhibits quasi-reversible redox peaks on the negative side, similar to 2. However, the redox peaks are broad in 2-M (possibly due to different Mo/W ratios) and appear slightly less negative than those in 2, indicating more facile electron transfer in 2-M due to lower reduction potential.


image file: d5qi00961h-f6.tif
Fig. 6 Cyclic voltammograms of 0.2 mM solutions of (a) 2 and (b) 2-M in 0.5 M Na2SO4 solution at pH 2.5 (pH of the solution was adjusted with 2 M H2SO4 and the 2nd cycle is shown); glassy carbon (WE): Pt wire, (CE): Ag/AgCl/3 M KCl (RE) (inset: zoomed part of the CV). LSV curves of (c) Cs[2] and Cs[2-M], (d) Cs[3], Cs[3-M], and Cs[3b] in 50 mM phosphate buffer solution (pH 7.1), 0.5 M Na2SO4.

To further understand orbital modulation and the influence of electron transfer in mixed-addenda as well as mixed-metal POMs, we investigated the electrocatalytic oxygen evolution reaction (OER) activity of 2, 2-M, 3, 3b, and 3-M. Cesium salts of these POMs (Cs[2], Cs[2-M], Cs[3], Cs[3b], and Cs[3-M]) were obtained through ion-exchange metathesis (details in the ESI†), and their OER activity was checked by coating onto a glassy carbon electrode (electrode preparation details in the ESI†). Linear sweep voltammetry (LSV) demonstrated well-defined OER responses for all the POMs, with overpotentials at 1 mA cm−2 of 600, 540, 570, and 530 mV vs. RHE for Cs[2], Cs[2-M], Cs[3], and Cs[3-M], respectively (Fig. 6c/d). These results clearly indicate that mixed-addenda POMs, Cs[2-M], and Cs[3-M], show better OER activity in terms of overpotential than their W-only counterparts, Cs[2] and Cs[3]. Furthermore, Cs[3b] demonstrated much improved OER activity, with an overpotential of 500 mV at 1 mA cm−2, compared to Cs[3] (570 mV) (Fig. 6d). This suggests that OER activity is modulated by a trimetallic mixed-metal W-only POM (Ni/Co/Fe combinations at the sandwich position/hetero) or by a mixed-addenda POM approach. These OER overpotentials are comparable to similar reported POMs with OER activity under neutral conditions.17 Given the vast potential for trimetallic mixed-metal substitutions along with mixed-addenda modifications in these sandwich POMs, they present a promising avenue for future catalytic applications.

Conclusion

This study highlights the strategic role of different transition metals in the rational design of one-pot self-assembly, leading to the formation of various mixed-metal mixed-addenda sandwich POMs with diverse structural attributes and varying Mo/W compositions. The incorporation of Mo, combined with the cooperative effects of transition metals, plays a crucial role in orbital engineering by lowering the LUMO energy and reducing the HOMO–LUMO band gap. The strategic integration of mixed addenda and transition metals in sandwich POMs demonstrates the versatility of this approach across diverse sandwich POM frameworks, including their applicability in POM-based coordination frameworks while highlighting the impact of charge density. The incorporation of Mo as a mixed addenda element using mixed transition metals proves to be an effective orbital engineering strategy, influencing the molecular properties, absorption characteristics, and overall electronic structure, which was further supported by DFT calculations. The enhanced electron transfer capabilities of sandwich POMs by orbital modulation are effectively confirmed by their improved electrocatalytic performance in the OER. Overall, structural modifications aimed at orbital engineering in POMs have successfully enabled fine-tuning of the optical band gap and the electronic band structure. These advancements open new possibilities for the application of sandwich POMs in various fields, such as photovoltaics, semiconductors, and energy conversion devices, highlighting their potential as next-generation functional materials.

Author contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflicts of interest

The authors declare no competing financial interest.

Data availability

The data supporting this article have been included as part of the ESI.†

Crystallographic data for all compounds has been deposited at the CCDC under [2330238–2330240, 2330242–2330250, 2330252–2330258 and 2330261†].

Acknowledgements

This study was supported by the SERB (CRG/2023/004666). G.S. thanks CSIR for the fellowship (09/1005(0022)/2018-EMR-I). R.C. thanks IIT Ropar for the fellowship. We also thank the IIT Ropar Mass Facility (SR/FST/CS-I/2018/55) and IIT Mandi ICP-MS facility.

References

  1. M. T. Pope, Y. Jeannin and M. Fournier, Heteropoly and isopoly oxometalates, Springer, 1983 Search PubMed .
  2. N. Li, J. Liu, B. X. Dong and Y. Q. Lan, Polyoxometalate–based compounds for photo–and electrocatalytic applications, Angew. Chem., Int. Ed., 2020, 59, 20779–20793 CrossRef CAS PubMed .
  3. D. E. Katsoulis, A survey of applications of polyoxometalates, Chem. Rev., 1998, 98, 359–388 CrossRef CAS .
  4. L.-L. Liu, L. Wang, X.-Y. Xiao, P. Yang, J. Zhao and U. Kortz, Structural overview and evolution paths of lacunary polyoxometalates, Coord. Chem. Rev., 2024, 506, 215687 CrossRef CAS .
  5. C. Li, A. Jimbo, K. Yamaguchi and K. Suzuki, A protecting group strategy to access stable lacunary polyoxomolybdates for introducing multinuclear metal clusters, Chem. Sci., 2021, 12, 1240–1244 RSC .
  6. K. Azmani, M. Besora, J. Soriano-Lopez, M. Landolsi, A.-L. Teillout, P. de Oliveira, I.-M. Mbomekallé, J. M. Poblet and J.-R. Galán-Mascarós, Understanding polyoxometalates as water oxidation catalysts through iron vs. cobalt reactivity, Chem. Sci., 2021, 12, 8755–8766 RSC .
  7. N. I. Gumerova and A. Rompel, Polyoxometalates in solution: speciation under spotlight, Chem. Soc. Rev., 2020, 49, 7568–7601 RSC .
  8. S. Mulkapuri, A. Ravi, R. Nasani, S. K. Kurapati and S. K. Das, Barrel-shaped-polyoxometalates exhibiting electrocatalytic water reduction at neutral pH: A synergy effect, Inorg. Chem., 2022, 61, 13868–13882 CrossRef CAS PubMed .
  9. H. T. Evans, C. M. Tourné, G. F. Tourné and T. J. Weakley, X-Ray crystallographic and tungsten-183 nuclear magnetic resonance structural studies of the [M4(H2O)2(XW9O34)2]10− heteropolyanions (M = CoII or Zn, X = P or As), J. Chem. Soc., Dalton Trans., 1986, 2699–2705 RSC .
  10. T. J. Weakley and R. G. Finke, Single-crystal X-ray structures of the polyoxotungstate salts K8.3Na1.7[Cu4(H2O)2(PW9O34)2]. cntdot. 24H2O and Na14Cu[Cu4(H2O)2(P2W15O56)2]. cntdot.53H2O, Inorg. Chem., 1990, 29, 1235–1241 CrossRef CAS .
  11. S.-T. Zheng and G.-Y. Yang, Recent advances in paramagnetic-TM-substituted polyoxometalates (TM = Mn, Fe, Co, Ni, Cu), Chem. Soc. Rev., 2012, 41, 7623–7646 RSC .
  12. H. N. Miras, J. Yan, D.-L. Long and L. Cronin, Engineering polyoxometalates with emergent properties, Chem. Soc. Rev., 2012, 41, 7403–7430 RSC .
  13. K. Suzuki, N. Mizuno and K. Yamaguchi, Polyoxometalate photocatalysis for liquid-phase selective organic functional group transformations, ACS Catal., 2018, 8, 10809–10825 CrossRef CAS .
  14. M. Chi, Y. Zeng, Z.-L. Lang, H. Li, X. Xin, Y. Dong, F. Fu, G.-Y. Yang and H. Lv, Molecular Orbital Engineering of Mixed-Addenda Polyoxometalates Boosts Light-Driven Hydrogen Evolution Activity, ACS Catal., 2024, 14, 5006–5015 CrossRef CAS .
  15. S. Li, S. Liu, S. Liu, Y. Liu, Q. Tang, Z. Shi, S. Ouyang and J. Ye, {Ta12}/{Ta16} cluster-containing polytantalotungstates with remarkable photocatalytic H2 evolution activity, J. Am. Chem. Soc., 2012, 134, 19716–19721 CrossRef CAS .
  16. H. Chen, Y. Sun, M. Zhang, P. Ma, J. Wang and J. Niu, Self-assembly of cis-trans Ta/W mixed-addendum POMs based on hexalacunary [H2P2W12O48]12− building blocks, Chin. Chem. Lett., 2024, 110329 Search PubMed .
  17. M. Martin-Sabi, J. Soriano-López, R. S. Winter, J.-J. Chen, L. Vilà-Nadal, D.-L. Long, J. R. Galán-Mascarós and L. Cronin, Redox tuning the Weakley-type polyoxometalate archetype for the oxygen evolution reaction, Nat. Catal., 2018, 1, 208–213 CrossRef CAS .
  18. G. Singh and D. Mandal, Modulation of the Band Gap and Redox Properties by Mixed Addenda in Sandwich Polyoxometalates, Inorg. Chem., 2023, 62, 19648–19663 CrossRef CAS PubMed .
  19. J. D. Compain, P. Mialane, A. Dolbecq, I. M. Mbomekallé, J. Marrot, F. Sécheresse, E. Rivière, G. Rogez and W. Wernsdorfer, Iron Polyoxometalate Single–Molecule Magnets, Angew. Chem., Int. Ed., 2009, 48, 3077–3081 CrossRef CAS PubMed .
  20. E. Limanski, M. Piepenbrink, E. Droste, K. Burgemeister and B. Krebs, Syntheses and X-ray Characterization of Novel [M4(H2O)2(XW9O34)2]n−(M = CuII, X = CuII; and M = FeIII, X = FeIII) Polyoxotungstates, J. Cluster Sci., 2002, 13, 369–379 CrossRef CAS .
  21. C. M. Tourné, G. F. Tourné and F. Zonnevijlle, Chiral polytungstometalates [WM3(H2O)2(XW9O34)2]12−(X = M = Zn or Co II) and their M-substituted derivatives. Syntheses, chemical, structural and spectroscopic study of some D, L sodium and potassium salts, J. Chem. Soc., Dalton Trans., 1991, 143–155 RSC .
  22. J. Wang, P. Ma, Y. Shen and J. Niu, A novel polyoxotungstate [Ni4(H2O)2(α-NiW9O34)2]16−based on an old structure with a new component, Cryst. Growth Des., 2007, 7, 603–605 CrossRef CAS .
  23. Z. Wang, X. Xin, M. Zhang, Z. Li, H. Lv and G.-Y. Yang, Recent advances of mixed-transition-metal-substituted polyoxometalates, Sci. China:Chem., 2022, 65, 1515–1525 CrossRef CAS .
  24. M. Tao, Q. Yin, A. L. Kaledin, N. Uhlikova, X. Lu, T. Cheng, Y.-S. Chen, T. Lian, Y. V. Geletii and D. G. Musaev, Structurally Precise Two-Transition-Metal Water Oxidation Catalysts: Quantifying Adjacent 3d Metals by Synchrotron X-Radiation Anomalous Dispersion Scattering, Inorg. Chem., 2022, 61, 6252–6262 CrossRef CAS .
  25. J. Soriano-López, F. W. Steuber, M. Mulahmetović, M. Besora, J. M. Clemente-Juan, M. O'Doherty, N.-Y. Zhu, C. L. Hill, E. Coronado and J. M. Poblet, Accelerating water oxidation–a mixed Co/Fe polyoxometalate with improved turnover characteristics, Chem. Sci., 2023, 14, 13722–13733 RSC .
  26. M. Ibrahim, Y. Lan, B. S. Bassil, Y. Xiang, A. Suchopar, A. K. Powell and U. Kortz, Hexadecacobalt(II)–containing polyoxometalate–based single–molecule magnet, Angew. Chem., Int. Ed., 2011, 50, 4708–4711 CrossRef CAS .
  27. J. Jia, Y. Niu, P. Zhang, D. Zhang, P. Ma, C. Zhang, J. Niu and J. Wang, A Monomeric Tricobalt(II)-Substituted Dawson-Type Polyoxometalate Decorated by a Metal Carbonyl Group:[P2W15O56Co3 (H2O) 3 (OH) 3Mn (CO) 3] 8−, Inorg. Chem., 2017, 56, 10131–10134 CrossRef CAS .
  28. M. Nyman, Polyoxoniobate chemistry in the 21st century, Dalton Trans., 2011, 40, 8049–8058 RSC .
  29. M. Nyman and P. C. Burns, A comprehensive comparison of transition-metal and actinyl polyoxometalates, Chem. Soc. Rev., 2012, 41, 7354–7367 RSC .
  30. Z. W. Wang, C. A. Chen and G. Y. Yang, A sandwiched Co-added polyoxometalate for efficient visible light-driven hydrogen evolution, Dalton Trans., 2024, 53, 9812–9818 RSC .
  31. A. Dolbecq, J.-D. Compain, P. Mialane, J. Marrot, E. Rivière and F. Sécheresse, Water substitution on iron centers: from 0D to 1D sandwich type polyoxotungstates, Inorg. Chem., 2008, 47, 3371–3378 CrossRef CAS .
  32. Z. Zhang, J. Liu, E. Wang, C. Qin, Y. Li, Y. Qi and X. Wang, Two extended structures constructed from sandwich-type polyoxometalates functionalized by organic amines, Dalton Trans., 2008, 463–468 RSC .
  33. L.-X. Shi, W.-F. Zhao, X. Xu, J. Tang and C.-D. Wu, From 1D to 3D Single-Crystal-to-Single-Crystal Structural Transformations Based on Linear Polyanion [Mn4(H2O)18WZnMn2(H2O)2(ZnW9O34)2]4−, Inorg. Chem., 2011, 50, 12387–12389 CrossRef CAS PubMed .
  34. R. Neumann and M. Dahan, A ruthenium-substituted polyoxometalate as an inorganic dioxygenase for activation of molecular oxygen, Nature, 1997, 388, 353–355 CrossRef CAS .
  35. R. Neumann and M. Dahan, Molecular oxygen activation by a ruthenium-substituted “sandwich” type polyoxometalate, J. Am. Chem. Soc., 1998, 120, 11969–11976 CrossRef CAS .
  36. I. A. Weinstock, R. E. Schreiber and R. Neumann, Dioxygen in polyoxometalate mediated reactions, Chem. Rev., 2017, 118, 2680–2717 CrossRef CAS PubMed .
  37. D. Barats, G. Leitus, R. Popovitz-Biro, L. J. Shimon and R. Neumann, A Stable “End–On” Iron(III)–Hydroperoxo Complex in Water Derived from a Multi–Iron(II)–Substituted Polyoxometalate and Molecular Oxygen, Angew. Chem., Int. Ed., 2008, 47, 9908–9912 CrossRef CAS .
  38. R. Neumann, Activation of molecular oxygen, polyoxometalates, and liquid-phase catalytic oxidation, Inorg. Chem., 2010, 49, 3594–3601 CrossRef CAS .
  39. X. López, C. Bo and J. M. Poblet, Electronic properties of polyoxometalates: Electron and proton affinity of mixed-addenda Keggin and Wells− Dawson anions, J. Am. Chem. Soc., 2002, 124, 12574–12582 CrossRef PubMed .
  40. E. N. Glass, J. Fielden, Z. Huang, X. Xiang, D. G. Musaev, T. Lian and C. L. Hill, Transition metal substitution effects on metal-to-polyoxometalate charge transfer, Inorg. Chem., 2016, 55, 4308–4319 CrossRef CAS PubMed .
  41. E. N. Glass, J. Fielden, A. L. Kaledin, D. G. Musaev, T. Lian and C. L. Hill, Extending Metal–to–Polyoxometalate Charge Transfer Lifetimes: The Effect of Heterometal Location, Chem. – Eur. J., 2014, 20, 4297–4307 CrossRef PubMed .
  42. X. López, J. J. Carbó, C. Bo and J. M. Poblet, Structure, properties and reactivity of polyoxometalates: a theoretical perspective, Chem. Soc. Rev., 2012, 41, 7537–7571 RSC .
  43. G. M. Sheldrick, SHELXT–Integrated space-group and crystal-structure determination, Acta Crystallogr., Sect. A:Found. Adv., 2015, 71, 3–8 CrossRef .
  44. G. M. Sheldrick, Crystal structure refinement with SHELXL, Acta Crystallogr., Sect. C:Struct. Chem., 2015, 71, 3–8 Search PubMed .
  45. I. D. Brown and D. Altermatt, Bond-valence parameters obtained from a systematic analysis of the inorganic crystal structure database, Acta Crystallogr., Sect. B:Struct. Sci., Cryst. Eng. Mater., 1985, 41, 244–247 CrossRef .
  46. G. Singh, S. Das Adhikary and D. Mandal, Physico-and Electrochemical Properties of First-Row Transition-Metal-Substituted Sandwich Polyoxometalates, Inorg. Chem., 2023, 62, 8551–8564 CrossRef CAS PubMed .
  47. G. Singh, S. D. Adhikary and D. Mandal, Stabilization and activation of polyoxometalate over poly (vinyl butylimidazolium) cations towards electrocatalytic water oxidation in alkaline media, Chem. Commun., 2023, 59, 4774–4777 RSC .
  48. R. Neumann and M. Gara, The manganese-containing polyoxometalate,[WZnMnII2 (ZnW9O34) 2] 12-, as a remarkably effective catalyst for hydrogen peroxide mediated oxidations, J. Am. Chem. Soc., 1995, 117, 5066–5074 CrossRef CAS .
  49. F. Evangelisti, P.-E. Car, O. Blacque and G. R. Patzke, Photocatalytic water oxidation with cobalt-containing tungstobismutates: tuning the metal core, Catal. Sci. Technol., 2013, 3, 3117–3129 RSC .
  50. M. Rafieeshad, N. I. Gumerova, E. Tanuhadi, G. Giester, H. Čipčić-Paljetak, D. Verbanac and A. Rompel, Synthesis, Characterization, and Antibacterial Activity of Ni-Substituted Krebs-type Sandwich-Tungstobismuthates Functionalized with Amino Acids, Inorg. Chem., 2023, 62, 9484–9490 CrossRef CAS .
  51. A. M. Morris, O. P. Anderson and R. G. Finke, Reinvestigation of a Ru2-Incorporated Polyoxometalate Dioxygenase Precatalyst,”[WZnRu2III (H2O)(OH)(ZnW9O34) 2] 11−”: Evidence For Marginal,≤ 0.2 Equivalents of Ru Incorporation Plus Faster Catalysis by Physical Mixtures of [RuII (DMSO) 4Cl2] and the Parent Polyoxometalate [WZn3 (H2O) 2 (ZnW9O34) 2] 12−, Inorg. Chem., 2009, 48, 4411–4420 CrossRef CAS .
  52. R. Neumann and A. M. Khenkin, Noble Metal (RuIII, PdII, PtII) Substituted” Sandwich” Type Polyoxometalates: Preparation, Characterization, and Catalytic Activity in Oxidations of Alkanes and Alkenes by Peroxides, Inorg. Chem., 1995, 34, 5753–5760 CrossRef CAS .
  53. A. R. Howells, A. Sankarraj and C. Shannon, A diruthenium-substituted polyoxometalate as an electrocatalyst for oxygen generation, J. Am. Chem. Soc., 2004, 126, 12258–12259 CrossRef .

Footnote

† Electronic supplementary information (ESI) available: Experimental details and characterization (SC-XRD structure and details, optical images, ESI-MS, TGA, PXRD, EDAX, FT-IR, Raman, and UV-visible spectroscopy with the Kubelka–Munk plot, XPS, ICP-MS, and BVS calculations and cyclic voltammetry), DFT study (HOMO, LUMO, and SOMO representation), and molecular oxygen binding/activation for 1/1-M, 2/2-M, 3/3-M, 4/4-M, 5/5-M, 6/6-M, 7, 8, 2-series (2a/2b/2c/2d/2c-M), 3-series (3a/3b/3c/3d/3b-M), and 4-series (4a/4a-M) and electrochemistry of sandwich POMs (2/2-M). CCDC 2330238, 2330239, 2330240, 2330242, 2330243, 2330244, 2330245, 2330246, 2330247, 2330248, 2330249, 2330250, 2330252, 2330253, 2330254, 2330255, 2330256, 2330257, 2330258, 2330261. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d5qi00961h
This journal is © the Partner Organisations 2025
Click here to see how this site uses Cookies. View our privacy policy here.