State-of-the-art of polyoxometalate-based catalysts for catalytic removal of vesicants

2.1 POMs in firsthand catalytic decontamination of vesicants

An’s group designed four carboxylic acid modified POMs in the form of Cs₄ [M(H₂O)₄][PMo₆O₂₁(PABA)₃]₂·nH₂O (M = Co, Mn, Ni, Zn, PABA = p-aminobenzoic acid) (CoPMo ₆ O ₂₁ , MnPMo ₆ O ₂₁ , NiPMo ₆ O ₂₁ , ZnPMo ₆ O ₂₁ ) for selective catalytic oxidation of thioethers including CEES (Figure 2). ⁴³ These four hybrid POMs displayed similar dimeric isostructural architectures. The synthesis of corresponding POMs using K⁺, Cs⁺, and Rb⁺ were attempted, but only Cs worked. Further, different metal cations and organic ligands were used to synthesize different POMs, and only Co²⁺, Mn²⁺, Ni²⁺, and Zn²⁺ and PABA were suitable. Four POMs were crystallized in space group P1 constructed from polyoxoanion units {PMo₆O₂₁}. Using methyl phenyl sulfide model, CoPMo ₆ O ₂₁ could efficiently catalyse its oxidation into the corresponding sulfoxide with the selectivity of 98 % in 20 min, which displayed the highest activity of the catalysts. Minimal methyl phenyl sulfide was converted by both PABA and CoCl₂, and Cs₃ [PMo₆O₂₁(PABA)₃]·nH₂O could catalyse 72 % of methyl phenyl sulfide with a selectivity of 89 %, far inferior to CoPMo ₆ O ₂₁ . This result showed the synergistic effect of Co site and PABA-modified POMs together was responsible for the rapid reaction rate and excellent selectivity. Similar results were found using {CoAsMo₆(PABA)₃} and {CoTeMo₆(PABA)₃}, indicating that the central heteroatoms affected the selectivity (in the order of PMo > AsMo > TeMo). Red shifts and a new band at 760 nm in the UV–vis absorption spectrum of the interaction between CoPMo ₆ O ₂₁ and H₂O₂ indicated the generation of active peroxomolybdenum and peroxocobalt species, which attacked the S atom of the sulfides to yield the corresponding product. The conversion of various organic sulfides did not obviously decrease with the electronic and steric-hindrance changes. The oxidation reaction by CoPMo ₆ O ₂₁ , MnPMo ₆ O ₂₁ , NiPMo ₆ O ₂₁ , and ZnPMo ₆ O ₂₁ were first-order reactions with corresponding kinetics constants of 0.23842, 0.1624, 0.20015, and 0.17483 min⁻¹, respectively. When methyl phenyl sulfide was replaced with the mustard gas simulant, CEES, 98.8 % was oxidized by CoPMo ₆ O ₂₁ into the nontoxic CEESO with 99.0 % selectivity in 12 min. The recyclability and stability of this system was also well demonstrated during the selective oxidation of CEES.

Figure 2:

The schematic diagram of catalytic oxidation of thioethers including CEES by MPMo₆O₂₁ (M = co, Mn, Ni, Zn). Reproduced with permission from ref. 43]. Copyright 2020 The royal society of Chemistry.

Zhou’s group employed four nanopolyoxometalate clusters Na₈K₁₄(VO)₂ [K₁₀⊂{Mo(Mo)₅O₂₁(H₂O)₃(SO₄)}₁₂ {(VO)₃₀(H₂O)₂₀] (denoted {Mo ₇₂ V ₃₀ }), ⁴⁴ ([{Na₁₀(H₂O)₁₂⊂{Mo₇₂Cr₃₀O₂₅₂(CH₃COO)₁₉(H₂O)₉₄}]·ca.120·H₂O (denoted {Mo ₇₂ Cr ₃₀ }), ⁴⁵ [Mo₇₂Fe₃₀O₂₅₂(CH₃COO)₁₀{Mo₂O₇(H₂O)}{H₂Mo₂O₈(H₂O)}₃(H₂O)₉₁]·ca.140·H₂O (denoted {Mo ₇₂ Fe ₃₀ }), ⁴⁶ [{Mo(Mo)₅O₂₁(H₂O)₄CH₃COO}₁₂{MoO(H₂O)}₃₀]·ca. 150·H₂O (denoted {Mo ₁₀₂ }) ⁴⁷ to catalyse the decontamination of HD and CEES. ⁴⁸ The conversion rate of CEES (0.042 mmol, 5 μL) reached 99.45 % in 10 min using 10.0 mg (0.53 μmol) of {Mo ₇₂ Cr ₃₀ } at approximately 20 °C. The optimal molar ratio of H₂O₂: CEES was determined to be 1:1. Within the reaction time of 1.0 min, the conversion of CEES was 51.42 %, 45.65 %, 38.38 %, and 34.95 % for {Mo ₇₂ V ₃₀ }, {Mo ₇₂ Cr ₃₀ }, {Mo ₇₂ Fe ₃₀ }, and {Mo ₁₀₂ }, respectively. A good linear relationship between ln (Ct_/C ₀ ) (C₀ and C_t was the concentrations of CEES at the start and time t) and the reaction time indicated first-order reaction kinetics with the kinetic constants of {Mo ₇₂ V ₃₀ }, {Mo ₇₂ Cr ₃₀ }, {Mo ₇₂ Fe ₃₀ }, and {Mo ₁₀₂ } of 0.56, 0.53, 0.47, and 0.44 min⁻¹, respectively and the following catalytic activity order: V > Cr > Fe > Mo. The reversible oxidation peak potentials with the addition of H₂O₂ and the further addition of the substrate CEES demonstrated that the intermediates generated from nodal metal interaction with H₂O₂ implemented the selective oxidation of CEES to CEESO. The unchanged Mo^VI3d X-ray photoelectron spectroscopy (XPS) peaks of {Mo ₇₂ V ₃₀ } indicated that Mo^VI did not participate in the catalytic reaction. Conversely, the binding energies of V 2p at 516.48 and 523.96 eV assigned to V^IV 2p_3/2 and V^IV 2p_1/2 were shifted to 516.40 eV (V^IV 2p_3/2), 517.49 eV (V^V 2p_3/2), 523.37 eV (V^IV 2p_1/2), and 524.71 eV (V^V 2p_1/2) with H₂O₂ addition, which showed that V was oxidized to a higher valence state by H₂O₂. ⁴⁹ With the simultaneous addition of H₂O₂ and CEES to the solution of {Mo ₇₂ V ₃₀ }, XPS peaks of V 2p were recovered, demonstrating that the catalyst recovered its original state. A similar situation was observed with the three other catalysts. The desired fast degradation of CEES by four catalysts benefitted from the large oil−water interface area due to agitation and the full exposure of active sites in {Mo ₇₂ M ₃₀ }. All four catalysts exhibited outstanding reusability, and the conversion rates of CEES still reached 93.41 %, 90.50 %, 90.13 %, and 89.70 % for {Mo ₇₂ V ₃₀ }, {Mo ₇₂ Cr ₃₀ }, {Mo ₇₂ Fe ₃₀ }, and {Mo ₁₀₂ }, respectively after five cycles.

Zhou’s group prepared an organic functional POM [C₄₄H₂₈N₄Cl₄]_1·5 [H₂PMo₁₀V₂O₄₀]·2C₂H₆O (H ₂ TClPP-H ₂ PVMo) by ion-exchange between 5,10,15,20-tetra(4-chlorine)phenylporphyrin (TClPP) and H₅PV₂Mo₁₀O₄₀ (H ₅ PVMo) for photocatalytic degradation of CEES . ⁵⁰ A higher singlet oxygen generation capacity was observed for H ₂ TClPP-H ₂ PVMo than for TClPP, which laid solid foundation for the selective oxidization of CEES to nontoxic CEESO. The red-shift of characteristic bands of P-Oa and the blue-shift of characteristic bands of Mo-Ob-Mo and Mo-Oc-Mo in the Fourier-transform infrared (FTIR) spectrum of H ₂ TClPP-H ₂ PVMo indicated the formation of hybrid H ₂ TClPP-H ₂ PVMo, which was further verified by the ³¹P solid-state nuclear magnetic resonance (NMR) spectrum. Spectrometric titration results of H₅PVMo solution by TClPP showed a nonequality to the simple sum of the UV spectrum of TClPP and H ₅ PVMo, which indicated that both had strong interactions and TClPP was converted to a deprotonated salt. An obvious turning point indicated that a stable hybrid was formed between TClPP and H ₅ PVMo at the stoichiometric ratio of 1:1.5; elemental analysis and thermogravimetric analysis (TGA) results confirmed this result. Within 180 min, the degradation rate of CEES by H ₂ TClPP-H ₂ PVMo in methanol could reach 99.52 % with irradiation and 28.24 % without irradiation, showing that light significantly accelerated the decontamination of CEES by H ₂ TClPP-H ₂ PVMo. At the same catalytic conditions, the degradation rates of CEES were approximately 20 % and 55.31 % for H ₅ PVMo and TClPP, respectively. In view of the hydrolysis effect on HD and CEES, the degradation of CEES by H ₂ TClPP-H ₂ PVMo in methanol/water (1:1, v/v) was also explored. The degradation rate of CEES by H ₂ TClPP-H ₂ PVMo reached 99.14 % with irradiation and 91.05 % within 90 min (under irradiation). The degradation rate of 98.1 % in methanol/water (1:1, v/v) was still achieved after the fifth in situ circulation, indicating the good repeatability of H ₂ TClPP-H ₂ PVMo. Under irradiation but no catalyst in air, CEES could be alcoholised to 2-methoxyethyl ethyl sulfide in methanol, and hydrolysed to ethyl 2-hydroxyethyl sulfide (HEES) because of a small amount of water in methanol. Active oxygen generated from H ₂ TClPP-H ₂ PVMo selectively oxidized CEES and HEES to CEESO and ethyl hydroxyethyl sulfoxide (HEESO) under irradiation. In addition, 2-methoxyethane and ethyl sulfide were possibly produced in methanol/water (1:1, v/v) via the alcoholysis and hydrolysis of CEES. ^{51 , 52} Active oxygen trapping experiments using benzoquinone (BQ) or NaN₃, revealed that CEES was degraded by O₂^·−and ¹O₂ producing from H ₂ TClPP-H ₂ PVMo in methanol with visible light irradiation. The ¹O₂ quantum yield of TClPP was significantly increased from 0.35 to 0.73 of H ₂ TClPP-H ₂ PVMo in methanol. The fluorescence intensity of H ₂ TClPP-H ₂ PVMo was greatly reduced because it dissipated more energy to complete energy transfer from the singlet state to the excited triplet state under irradiation. ⁵³ This material could degrade CEES under visible light irradiation.

Yang’s group successfully prepared a novel 12-Ti-substituted POM K₁₀H₁₅ [{K₂Na(H₂O)₃}@{(Ti₂–O)₂(Ti₄O₄)₂ (A-α-1,3,5-GeW₉O₃₆)₂ (A-α-2,3,4-GeW₉O₃₆)₂}]·45H2O (TiGeWO) with the key characteristic of chiral trivacant Keggin-type GeW₉O₃₆ (A-α-1,3,5-GeW ₉ O ₃₆ and A-α-2,3,4-GeW ₉ O ₃₆ ) units. ⁵⁴ Two A-α-1,3,5-GeW ₉ O ₃₆ and two A-α-2,3,4-GeW ₉ O ₃₆ were connected via two Ti₂O and two Ti₄O₄ alternate bridging cluster cores, which was the first time for the coexistence of Ti₂O and Ti₄O₄ cores as the bridges in POM. Each Ti atom in the Ti₄O₄ tetrad was coordinated with three μ₂-O, one μ₄-O in the GeW₉O₃₆ unit, and two μ₂-O in the Ti₄O₄ core. The two Ti atoms of the Ti₂O core were coordinated by four μ₂-O, one μ₄-O in the GeW₉O₃₆ unit, and one μ₂-O in the Ti₂O core. Although the trivacant A-α-1,3,5-GeW ₉ O ₃₆ and A-α-2,3,4-GeW ₉ O ₃₆ segments were chiral enantiomers, they coexisted in TiGeWO, resulting in a non-chiral overall structure. TiGeWO has been applied in the heterogeneous catalytic oxidation of various aromatic thioethers with H₂O₂ as an oxidant in acetonitrile at 60 °C. TiGeWO could efficiently catalyze the oxidation of various substrates into sulfoxides or sulfones. Substrates with electron-donating groups on the aromatic ring exhibited a higher conversion rates than those with electron-withdrawing groups. Additionally, larger steric hindrance could reduce the conversion and selectivity. The catalytic decontamination of CEES by TiGeWO was investigated under H₂O₂ as oxidizer at 30 °C. Nearly 100 % of CEES could be oxidized to CEESO in 90 min. The linear relation between ln (C_t/C₀) suggested that this was a first-order reaction with a kinetic constant k₁ of 0.05197 min⁻¹ and a half-life of approximately 13 min. After three cycles, the high catalytic conversion of aromatic thioethers and CEES was retained, and the unanimous IR spectra of TiGeWO after three cycles demonstrated its good reusability.

Overcoming the inactivity of oxoboron clusters and the repulsion between negative polyoxotungstates and oxoboron clusters has been a challenge in the creation of B-containing polyoxotungstates. recently, Zheng and Yang’s group developed B₃₀-incorporated polyoxotungstates [(B₁₈Si₃Ln₆O₃₆(OH)₁₄){B₄Ni₄O₁₀(OH) (A-α-SiWO₃₄)}₃]⁴⁴⁻ (1Ln) [(B₁₉Si₂Ln₇O₃₅(O–H)₁₅(H₂O)){B₄Ni₄O₁₀(OH)₄ (A-α-SiW₉O₃₄)}₂{B₃Ni₄O₉ (OH)₃ (A-α-SiW₉O₃₄)}]⁴¹⁻ (2Ln, Ln = Gd, Tb and Dy), and B₂₂-incorporated POTs [(B₂₂O₄₂){LnNi₃(OH)₃(B-α-SiW₉O₃₄)}₄]³⁴⁻ (3Ln; Ln = Sm, Gd and Tb) through multicomponent charge and symmetry matching among Ln³⁺, TM²⁺, borates, and polyoxotungstates. ⁵⁵ To date, 1Ln and 2Ln were the largest B-containing polyoxotungstates. B-rich polyoxotungstates 1Ln, 2Ln and 3Ln were prepared using NaBO₃·4H₂O, Li₂B₄O₇ and K₂B₁₀O₁₆·8H₂O as the sources of B, respectively. 1Gd, 2Gd, and 3Gd were discussed in detail as representatives (Figure 3). In 1Gd, 30 B atoms were dispersed over two B₉O₁₅(OH)₆ ({B₉}) nonamers, three B₂O₄(OH)₂ ({B₂}) dimers, and six BO₃ monomers with a new B–O nonamer appearing in {B₉}. A unique trigonal-prism cage [B₁₈Si₃O₃₆(OH)₁₂]¹⁸⁻ ({B₁₈Si₃}) was formed by two {B₉} assemblies in a foot-to-foot orientation and three SiO₄ tetrahedra connected by six B–O–Si bridges. The anionic, hollow {B₁₈Si₃} cage was encapsulated by two cationic planar triangular [Gd₃ (μ₃-OH)]⁸⁺ units in its inner cavities through coordination with O from the {B₁₈Si₃} cage, and forming the [B₁₈Si₃Gd₆O₃₆(OH)₁₄]²⁻ ({B₁₈Si₃Gd₆}) cluster. In this cluster, the innermost Gd-shell Gd₆O₁₂(OH)₂ ({Gd₆}) with trigonal prismatic shape was formed. Six BO₃ units in 1Gd served as bridges, bonding to both Ni²⁺ and Gd³⁺ ions. For the 12 Ni²⁺ ions in the Ni-shell, every four Ni²⁺ ions were arranged in a pyramidal array, and two μ₃-OH groups, two BO₃ units, and one {B₂} dimer are bridged by adjacent Ni²⁺ ions to form B₄Ni₄O₁₀(OH)₄ ({B₄Ni₄}). Three Ni-substituted polyoxotungstate motifs [B₄Ni₄O₁₀(OH)₄](A-α-SiW₉O₃₄) ({B₄Ni₄(A-SiW₉)}) were formed after the introduction of {B₄Ni₄} clusters in the vacant sites of [A-SiW₉]. The B30-incorporated 1Gd was obtained with the {B₁₈Si₃Gd₆} cluster surrounded by three {B₄Ni₄(A-SiW₉)} motifs. Owing to the multicomponent cooperative charge-matching interactions and matched geometries between components, a multi-shell configuration of Ln-shell@B-shell@Ni-shell@W-shell ({Gd₆}@{B₂₄Si₃}@{Ni₄}₃@{SiW₉}₃) was existed in 1Ln polyoxotungstates series. The robust coordination ability and flexible coordination modes of Gd³⁺ ions used as templates possessed fewer structural restrictions on the formation of oxoboron motifs and their architectures. The long Gd–O bond, high coordination number and positive charge of Gd³⁺ ion were crucial for constructing larger B–O clusters and reducing the outer steric hindrance of the polyoxotungstate. Similar structures were observed in 2Ln. The tetrameric 3Gd exhibited a Gd-shell@B-shell@Ni-shell@W-shell ({Gd₄}@{B₂₂}@{Ni₃}₄ @{B–SiW₉}₄) configuration featuring a cage-like 22-nuclearity Td-symmetry B–O cluster [B₂₂O₄₂]¹⁸⁻ ({B₂₂}). This was the first observation of the {B₂₂} cluster topology in oxoboron cluster chemistry. Within the {B₂₂} cluster, six B₃O₇ three-rings are located in the tetrahedral arrays defined by four Gd³⁺ ions, forming a hollow tetrahedral cage [Gd₄B₂₂O₄₂]⁶⁻ ({Gd₄}@{B₂₂}≡{Gd₄B₂₂}) with a diameter of approximately 1.2 nm through the connections of the Gd- and B-shells. A slightly larger tetrahedral cluster [Ni₁₂Gd₄B₂₂O₄₆(OH)₁₂]⁶⁻ {Ni₁₂Gd₄B₂₂} was formed by capping each corner of the tetrahedral {Gd₄B₂₂} cage with three edge-shared NiO₆ octahedra (Ni₃O(OH)₃ core). Under the precondition of the isomerization from {A-SiW₉} to [B-α-SiW₉O₃₄]¹⁰⁻ ({B–SiW₉}), four {B–SiW₉} attached to four corners of {Ni₁₂Gd₄B₂₂} cluster, resulting in the Td-symmetrical 3Gd. 1Ln, 2Ln, and 3Ln possessed 39, 21, and 24 protons, respectively, which were delocalized on the overall structures. All these POMs were soluble in water, but not in many solvents. 1Gd was tested for the catalytic decontamination CEES, where it could selectively convert about 96 % of CEES into CEESO with 97 % selectivity in 2 h and a half-life of approximately 10 min. In a control experiment without 1Gd, only around 66 % CEES was converted. Consistent IR and powder X-ray diffraction (PXRD) experimental results along with the maintained high activity of 1Gd in six catalytic reactions confirmed its structural stability and recyclability.

Figure 3:

Crystal structures of POMs 1Gd, 2Gd and 3Gd. Reproduced with permission from ref. 55]. Copyright 2021 The Chinese chemical Society.

Our group designed and synthesized a series of Ln-containing polytungstate ([H₂N(CH₃)₂]₈Na₈Cs₄H₉ [Ln₂ -SeW₄O₁₁(OH) (H₂O)₄(SbW₉O₃₃) (SeW₉O₃₃) (Se_1/2Sb_1/2W₉O₃₃)]₂·32H₂O [Ln = Tb (1), Dy (2), Ho(3), Er (4)] (Figure 4) ⁵⁶ and Cs₁₈Na₈H₂₀ [Ce₃(H₂O)₁₀W₈Bi₄O₂₈(B-α-BiW₉ O₃₃)₄]₂·64H₂O{SeO2 (OH)} (5) ⁵⁷ for catalytic oxidation CEES. POMs 1–4 were in the monoclinic space group P2₁/c and the hexameric architectures. In POMs 1–4, two equivalent trimeric subunits Ln₂W₄O₉(H₂O)₄(SbW₉O₃₃) (SeW₉O₃₃) (Se_1/2Sb_1/2W₉O₃₃) were linked through two μ₂-{SeO₂(OH)} bridges. Three different types of trivacant Keggin fragments {SbW₉}, {SeW₉}, and {Se_1/2Sb_1/2W₉} were existed in compounds 1–4, and {SeW₉} was formed in situ, not from the raw materials. It was a rare occurrence that POM was constructed from different fragments with the central heteroatoms of trigonal pyramidal Sb^III and Se^IV. From the bond lengths of Se1/2Sb1/2−O located in the scope of 1.892–1.924, which were between that of Se–O and Sb–O bonds, indicating that 1/2 S b and 1/2 Se atoms co-occupied the central position of Se_1/2Sb_1/2W₉O₃₃ fragment. POM three was selected as a representative to investigate their catalytic oxidation ability of CEES. POM three exhibited that the remarkable oxidation activity of CEES with 99 % percent conversion of CEES into CEESO and a high selectivity of 95 % within 30 min. Without 3, only 30 % conversion was observed. In addition, POM three exhibited a desirable heterogeneous catalytic behavior, and remained stable conversion and selectivity of CEES after five cycles. POM five was in monoclinic C2/c space group and contained, and it was the first case that tungstobismuthate building blocks self-assembled into unprecedented octameric Ln-containing POM([Ce₃(H₂O)₁₀W₈Bi₄O₂₈(B-α-BiW₉ O₃₃)₄]₂⁴⁶⁻) with an extended structure. The octameric [Ce₃(H₂O)₁₀W₈Bi₄O₂₈(B-α -BiW₉O₃₃)₄]₂⁴⁶⁻ was constructed by two tetrameric [Ce₃(H₂O)₁₀W₈Bi₄O₂₈(B-α-BiW₉O₃₃)₄]²³⁻ via two Ce–O–W bonds as linker, which was formed by a heterometallic cluster core Ce₃(H₂O)₁₀W₈Bi₄O₂₈ surrounded with four B-α-BiW9O33 fragments. POM five also demonstrated its.

Figure 4:

The crystal structures of ([H₂N(CH₃)₂]₈Na₈Cs₄H₉ [Ln₂SeW₄O₁₁(OH) (H₂O)₄(SbW₉O₃₃) (SeW₉O₃₃) (Se_1/2Sb_1/2W₉O₃₃)]₂·32H₂O [ln = Tb, dy, Ho, Er]. Reproduced with permission from ref. 56]. Copyright 2021 American chemical Society.

brilliant catalytic oxidation performance of CEES. The conversion of CEES and selectivity to CEESO could reach 98 % and 96 % within 35 min, and the pseudo first-order dynamic reaction rate constant k ₁ was 0.11632 min⁻¹ with a half-life of about 6 min. After five cycles, POM five still rrretained high catalytic activity and selectivity to CEESO, and meaning that POM five could be as a robust heterogeneous catalyst.

2.2 POMs-based MOFs for catalytic decontamination of vesicants

Liu’s group designed two polyoxovanadate-based MOFs ([Co(bib)]{V₂O₆} (V–Co-MOF) and [Ni(en) (bib)]{V₂O₆}·2H₂O (V–Ni-MOF) (bib = 1,4-bis(1H-imidazoly-1-yl)benzene, en = ethanediamine) for two-site synergetic catalytic decontamination of CEES. ⁵⁸ One-dimensional (1D) {V₂O₆}^2- polyanion chains (abbreviated as {V₂O₆}) were formed into two 2D bimetallic oxide layers, {CoV₂O₆} and {NiV₂O₆}, using {CoN₂O₂} tetrahedra and {NiN₄O₂} octahedra as linkers, and further assembled into 3D frameworks. The four-coordinated Co (II) ions in the V–Co-MOF were unsaturated metal sites, in contrast to the saturated six-coordinated Ni (II) ions of V–Co-MOF. The catalytic activities of both MOFs for CEES oxidation were studied. CEES was selectively oxidized to CEESO without 2-chloroethyl ethyl sulfone (CEESO₂) byproduct of. V–Co-MOF exhibited a higher catalytic activity than V–Ni-MOF. A pseudo-first order reaction of V–Co-MOF and V–Ni-MOF was deduced with the corresponding rate constants k₁ and k₂ of 0.27044 and 0.06273 min⁻¹, and the half-lives of 2.6 and 11.0 min, respectively. Through evaluating the catalytic activities of en, bib, NiCl₂, CoCl₂, NaK{V₂O₆}, NaK{V₂O₆} + NiCl₂, NaK{V₂O₆} + CoCl₂, V–Ni-MOF and V–Co-MOF, V–Ni-MOF possessed the highest catalytic activity, possibly owing to the generation of active peroxovanadium from the reaction of V^v with H₂O₂ and the CEES + via the coordination of Co(II) and S. V–Co-MOF could rapidly capture CEES with a maximum absorption of 14 mg g⁻¹, then activating CEES. ⁵⁹ The binding energy of Co 2p in the XPS of CEES-absorbing V–Co-MOF showed a notable downward shift, indicating electron transfer from CEES to Co ions and a significant increase in electron density of Co ions in V–Co-MOF. Similar experimental results were observed with V–Ni-MOF. The formation of O–O in the peroxovanadium was verified by new peaks at 860 cm⁻¹ and 870 cm⁻¹ in the FTIR and Raman spectra. Throughout the reaction, the weight and structure of V–Co-MOF remained largely unchanged, demonstrating its high stability and recyclability. The possible mechanism was speculated based on the Experimental results and shown in Figure 5. This work was the first example of a dual-active-site POM-based MOF catalyst for mustard gas detoxification.

Figure 5:

The proposed catalytic mechanism of CEES by V–Co-MOF. Reproduced with permission from ref. 58]. Copyright 2020 The royal society of Chemistry.

Another POV-based MOF [Ni(bib)₂]{V₂O₆} ({V ₆ }-MOF) was successfully synthesized by Liu’s group for the catalytic decontamination of CEES. ⁶⁰ A hexanuclear vanadium anion cluster, {V₆O₁₈}⁶⁻ ({V₆}) in {V ₆ }-MOF presented an unprecedented hollow Lindqvist-like structure compared with reported {V₆} clusters. ⁶¹ Bimetallic decanuclear cluster {Ni₄–V₆} was generated through the coordination between the {V₆} cluster and four Ni atoms. Through alternating {V₆} clusters and octahedral Ni atoms, a 2D bimetallic oxide layer of {Ni₃V₆O₁₈} emerged, which was further assembled into a 3D framework using bib ligand as the linker (Figure 6). {V ₆ }-MOF exhibited excellent thermal stability up to 450 °C and solvent stability, which was crucial to its sustainability. The catalytic activity of {V ₆ }-MOF on the oxidative detoxification of CEES was evaluated, whereby CEES was completely converted into nontoxic CEESO in 40 min with a half-life of 10.5 min. This was significantly faster than the 240 min reaction time of Lindqvist-type POV-based MOF catalyst also containing hexanuclear vanadium clusters, which was attributed to the more accessible V sites in {V ₆ }-MOF {V ₆ }-MOF compared to the Lindqvist-type {V₆} cluster. Feasible ratios of {V ₆ }-MOF: CEES and CEES: H₂O₂ was determined to be 0.05:1 and 1.0:1.2, respectively. Various raw materials and intermediates were used as catalysts in a series of control tests to verify the active components of {V ₆ }-MOF. {V₆}-mIM could completely catalyze the conversion of CEES to CEESO with 100 % selectivity, but bib and [Ni^II(H₂O)₄(L)₂] exhibited no activity, verifying that the main active site of {V ₆ }-MOF was the {V₆} cluster. A new Raman spectrum peak at 870 cm⁻¹ assigned to the O–O stretch of {V ₆ }-MOF treated with H₂O₂ indicated the production of active peroxo species. These species attacked S atom of CEES, generating nontoxic CEESO. The conversion and selectivity were retained after four successive runs. The consistent PXRD patterns, FTIR and XPS spectra of the cycled catalyst indicated that {V ₆ }-MOF could be applied as a highly effective catalyst for the detoxification of mustard gas due to its excellent stability and sustainability.

Figure 6:

The crystal structure of {V₆}-MOF. Reproduced with permission from ref. 60]. Copyright 2021 American chemical Society.

Liu’s group developed a hollow POM@MOF with a pseudo-homoepitaxial structure [Cu₂(BTC)₄/₃(H₂O)₂]₆ [H₃PW₁₂O₄₀]·(C₄H₁₂N)₂ (NENU-3a) for promoting the adsorption and catalytic performance toward CEES. ⁶² H₃PW₁₂O₄₀ (PW ₁₂ ) as guest with ideal catalytic activity was perfectly encapsulated in the [Cu₃BTC₂] (BTC = 1,3,5-benzenetricarboxylate) host through supramolecular assembly and formed, which exhibited a perfectly matched lattice/structure for [Cu₃BTC₂] and PW ₁₂ , yielding desirable catalytic performance. Scanning electron microscopy (SEM) images of NENU-3a showed the clear transits from cuboctahedral to octahedral morphology, indicative of epitaxial growth. XRD patterns and FTIR peaks of NENU-3a confirmed the presence of [Cu₃BTC₂] and PW ₁₂ in NENU-3a. Formation of an internal mesoporous structure was evident due to the noticeable interior collapse of the crystalline particle, leading to a clear hollow structure with a diameter of 700 nm after 1.5 min. Energy dispersive X-ray (EDX) elemental mapping revealed the distribution of Cu and W, confirming the hollow structure and the uniform epitaxial growth of NENU-3a. The stable existence of [Cu₃BTC₂] in the PW ₁₂ /Cu(NO₃)₂ solution was attributed to the inhibition of PW ₁₂ dissociation in weak polar chloroform/methanol and lowering H⁺ concentration. HNO₃, released from reactions between H₃BTC and Cu(NO₃)₂ in the NENU-3a assembly contributed to the formation of the hollow structure via in-diffusion etching. The concentration difference drove HNO₃ enriched in the confined space to diffuse inwardly through hydrogen bonding, etching the [Cu₃(BTC)₂] cores. Without this, HNO₃ directly diffused to the exterior owing to the lack of hydrogen bonds, and relatively dilute HNO₃ could not etch the inside [Cu₃(BTC)₂]. Transmission electron microscopy (TEM) images and the concentration-dependent etching experiment of HNO₃ on [Cu₃(BTC)₂] both confirmed the proposed internal diffusion mechanism in Figure 7. Etching occurred on the [Cu₃(BTC)₂] surface only when the HNO₃ concentration increased to about 70 mM, which indicated that the interior was more readily etched than the exterior. Compared to [Cu₃(BTC)₂] and solid NENU-3a, the intermediate Brunauer–Emmett–Teller (BET) surface area of hollow NENU-3a indicated that micropores were present only in the hollow layer of the surface, with some [Cu₃(BTC)₂] remaining inside. Similar thermal stability was observed on hollow NENU-3a, mainly due to the stabilization and protective effect of its outer shell layers for the interior [Cu₃(BTC)₂], in addition to its robust single-crystalline nature. Another POM@MOF ([Cu₁₂(BTC)₈][PMo ₁₂ ]) (NENU-5a) was constructed using the pseudo-homoepitaxial growth strategy, demonstrating the versatility of the strategy. The catalytic oxidation performance of CEES was tested on hollow NENU-3a with H₂O₂ as an oxidant. Hollow NENU-3a exhibited a faster catalytic rate than solid NENU-3a. The CEES conversions were almost 100 % and 86 % for hollow and solid NENU-3a, respectively, in 30 min. Moreover, hollow NENU-3a showed the higher selectivity toward CEESO than the solid form due to the rapid departure of CEESO from the shell structure, preventing over-oxidation. Hollow NENU-3a retained its catalytic activity, selectivity, morphology and cavity structure upon reuse.

Figure 7:

The proposed pseudo-homoepitaxial growth strategy of hollow POM@MOF. Reproduced with permission from ref. 62]. Copyright 2022 Chinese chemical Society.

Wang’s group synthesized a novel and multifunctional 2D POMOF [Co^II(bcbpy)₂ (θ-Mo₈O₂₆)_0.5]·5H₂O (BHU-1) through hydrothermal reactions of HbcbpyCl (1-(4-carboxybenzyl)-4,4′-bipyridinium chloride [Mo₈O₂₆]⁴⁻, and Co²⁺. ⁶³ BHU-1 was the first 2D [θ-Mo₈O₂₆]⁴⁻-based cobalt–viologen framework with a triclinic system and the space group P-1. The consistency of experimental and simulated PXRD confirmed the purity of BHU-1. ⁶⁴ BHU-1 exhibited two broad absorption bands, from 200 to 497 nm, and from 497 to 1,621 nm and one absorption peak at 1,663 nm, which were assigned to Mo–O ligand-to-metal charge transfer (LMCT), Co-bcbpy LMCT and the phenyl C–H stretching overtone band, respectively. Mott–Schottky (MS) electrochemical tests revealed that BHU-1 was a typical n-type semiconductor with the Fermi level (E_f) of −1.15 eV versus Ag/AgCl. According to the valence band (VB)-XPS method, the gap between E_f and the VB potential of BHU-1 (0.73 eV, vs. NHS) was found to be 1.88 eV. Under visible light irradiation, Nyquist experiments indicated that BHU-1 exhibited the fast separation and high transfer efficiency of photogenerated charge. BHU-1 catalyzed the oxidation of 98 % of CEES with 97 % selectivity within 5 min in ethanol under visible light. Under full-spectrum and near-infrared (NIR) light irradiation, the conversions of CEES were 99 % and 93 %, respectively, and the selectivities were both 96 % within 5 min, which meant that the catalytic activity was retained under different light regimes. Using the raw materials as homogeneous catalysts, CEES conversions of 17 %, 21 %, and 18 % were achieved with CoCl₂·6H₂O, HbcbpyCl and Na₂MoO₄·2H₂O, respectively, indicating that the photocatalytic oxidation of CEES was enhanced by the synergy of the three components in BHU-1. The excellent cycling stability and durability was confirmed by recycling experiments and consistent PXRD and IR spectra. The catalytic reaction was simplified to a pseudo-first-order reaction using the Langmuir–Hinshelwood model. ⁶⁵ The photogenerated electrons were separated from the hole with the help of C₂H₅OH in catalytic process of CEES, and the OH radical was easily produced through the reaction of H₂O₂ with photo-induced holes. The conversion rate of CEES was decreased by the introduction of BQ or tert-butyl alcohol (TBA), which indicated that CEES was oxidized to CEESO by O²⁻ and OH generated from the reaction between photogenerated electrons with oxygen. H₂O₂ could also facilitate O₂ circulation: H₂O₂ + ·O₂⁻ → O₂ + e⁻ → O₂⁻ + CEES → peroxide intermediate + CEES → CEESO.

To leverage the complementary advantages of MOFs and POMs and achieve excellent catalytic performance, Zhou’s group designed a composite material by encapsulating H₅PV₂Mo₁₀O₄₀ in MIL-101(Cr) for efficient CEES decontamination. ⁶⁶ The α-Keggin H₅PV₂Mo₁₀O₄₀ showed promise for the amient decontamination of HD. ⁶⁷ Using the impregnation method, the maximum loading amount of H₅PV₂Mo₁₀O₄₀ by MIL-101(Cr) was approximately 33.28 wt% even with additional H₅PV₂Mo₁₀O₄₀. Given the large 34 Å cages of MIL-101(Cr) and the diameter of 13 Å and volume of 2,250 ų of PV₂Mo₁₀O₄₀⁵⁻ anion, four H₅PV₂Mo₁₀O₄ per cage was speculated for the maximum loading sample. Analysis via IR spectra, XRD, TG, and N₂ adsorption/desorption experiments confirmed the successful preparation and intact crystalline structure of the MOF. Furthermore, the absence of characteristic peaks of H₅PV₂Mo₁₀O₄₀ in the XRD pattern of the composite and the elemental mapping analysis (Cr, O, P, Mo, and V) in the EDX spectra indicated a uniform distribution of H₅PV₂Mo₁₀O₄₀ in MIL-101(Cr). The maximum decontamination efficiency of CEES reached 90 % using 20 mg MIL-101(Cr). For the composite materials, the decontamination efficiency gradually increased with the loading amount of H₅PV₂Mo₁₀O₄₀ increasing. The maximum decontamination efficiency of HD was 97.39 % by the composite with the loading amount of 120 mg H₅PV₂Mo₁₀O₄₀. The decontamination reaction of HD exhibited pseudo-first order kinetics with a half-life of 26.38 min by theoretical fitting. Compared to H₅PV₂Mo₁₀O₄₀, MIL-101(Cr), or their physical mixtures, the composite showed superior performance, highlighting a synergetic effect. No products from the suspension in the decontamination process of MIL-101(Cr) were detected by gas chromatography/mass spectrometry (GC/MS) except for CEES, implying the decontamination was adsorption. However, in the composite, CEES and CEESO were detected with the relative amounts of 64.93 % and 32.16 %, indicating a combined adsorption and oxidation degradation. Additionally, the composite outperformed activated clay, nano-Al₂O₃ and nano-MgO under the same conditions. Its decontamination efficiency remained at 93.83 % and its structure was intact even after four consecutive recycles using the composite. Notably, the negligible leakage of H₅PV₂Mo₁₀O₄₀ from MIL-101(Cr) in the recycle process was attributed to the insolubility of highly polar H₅PV₂Mo₁₀O₄₀ in nonpolar petroleum ether.

2.3 POMs composite for catalytic decontamination of vesicants

Three composites Zn₂Cr-LDH-PW₁₁M(H₂O)O₃₉ were constructed through exfoliation-reassembly of mono-transition metal (Ni, Co, Cu)–substituted POMs [PW₁₁M(H₂O)O₃₉]^5- (PW ₁₁ M) into layered double hydroxides (LDHs) Zn ₂ Cr-LDH-NO ₃ by Chi’s group. ⁶⁸ Increased in spacing layers (1.06 nm matching the diameter of cluster) observed in PXRD patterns of the composites indicated the replacement of small nitrate with large PW ₁₁ M and the intercalation of PW ₁₁ M into the LDH. Broad hump diffraction peaks (2θ = ∼32-42°) and FTIR characteristic peaks of composites confirmed the preservation of the LDH lamellar structure and the integrity of POMs. SEM and TEM images showed that the intrinsic LDH nanosheet morphology retained in the composite. The loading amount of PW ₁₁ Ni was 52 wt% by inductively coupled plasma atomic emission spectroscopy (ICP-AES) analyses. Solid state ³¹P NMR of composite Zn ₂ Cr-LDH-PW ₁₁ Ni(H ₂ O)O ₃₉ still exhibited the characteristic peak of PW₁₁Ni (−11.3 ppm). The BET surface area and average pore size of composite Zn ₂ Cr-LDH-PW ₁₁ Ni(H ₂ O)O ₃₉ was enlarged to 24.92 m²/g and 42.48 nm, respectively, from 13.56 m²/g and 14.14 nm of the LDH host. CEES was used to evaluate the catalytic abilities of the composites, and Zn ₂ Cr-LDH-PW ₁₁ Ni exhibited the best catalytic performance with 98 % conversion. Similar selectivity (94 %) for nontoxic CEESO was observed for all three composites. At the same conditions, the catalytic conversions of CEES were 81 %, 78 %, 73 % and 30 % for PW ₁₁ Ni, PW ₁₁ Co, PW ₁₁ Cu, and Zn ₂ Cr-LDH, respectively, highlighting POM as the primary catalytic active species and the cooperative effect between POM and LDHs. The low conversions of 21 % and 58 % for plenary PW ₁₂ and lacunary K₇PW₁₁O₃₉, respectively, revealed the important roles of tungstophosphate and substituted transition metals in the catalytic reaction. The catalytic activity order of the composites was consistent with that of PW ₁₁ M. CEES conversion was increased with the amount of the composite Zn ₂ Cr-LDH-PW ₁₁ Ni and oxidant increasing. Initially, toxic CEESO₂ was formed within the first 30 min, and remained stable until the reaction was complete. Zn ₂ Cr-LDH-PW ₁₁ Ni maintained the good stability over nine cycles. Introducing several radical scavengers had no effect on the CEES conversion, indicating a non-radical reaction process for catalytic decontamination.

Due to the large accessible areas and workability of polymers with intrinsic microporosity (PIMs), Farha’s group constructed PIM-based hybrid materials for CEES degradation using PIMs carriers for supporting POM (H₃PW₁₂O₄₀, PW ₁₂ ) (Figure 8). ⁶⁹ Three kinds of functionalized PIMs with cyano (PIM-1), carboxy (PIM-1-COOH) and amidoxime (PIM-1-AO) groups were prepared and used to evaluate their POM loading performances. The characteristic stretching vibration peaks of POM appearing in the FTIR spectrum of PW ₁₂ @PIM-1-AO were not observed in that of PW ₁₂ @PIM-1 and PW ₁₂ @PIM-1-COOH, indicating that PIM-1-AO was a promising POM support. ICP-OES analysis indicated that about 80 % of POM was supported on PIM-1-AO, but only ∼1 wt % on the others, revealing a strong interaction between POMs and PIM-1-AO. ³¹P solid state NMR showed a chemical shift at −15.7 ppm, and ¹H NMR displayed –OH and –NH₂ shifts at 9.4 and 5.8 ppm, indicating the presence of free PW ₁₂ and interactions between POMs and the amidoxime of PIM-1-AO. The BET surface area of PW ₁₂ @PIM-1-AO decreased with the POM loading increasing. SEM images with EDX mapping confirmed uniform POM distribution in PIM-1-AO. Two XPS peaks of 4f_7/2 and 4f_5/2 at 36.3 and 38.5 eV in the W 4f region of PW ₁₂ were slightly shifted to lower binding energies for PW ₁₂ @PIM-1-AO. Combined with a broader peak at 399.8 eV in the N 1 s region and two O 1 s peaks at 532.8 and 530.1 eV, it was concluded that POMs were attached to PIM-1-AO via electrostatic interactions between the protonated amine groups and anionic POMs. ⁷⁰ Uniform fibers with an average diameter of 1.2 μm were prepared from a mixed solution of POM and PIM-1-AO in DMF (40 % w/v) via the electro-spinning. Analyses of the ³¹P cross-polarization magic-angle-spinning (CP/MAS) NMR spectrum, FTIR spectrum and EDX mappings of the composite fiber suggested that the POM was intact and uniform distributed within the fiber. PW ₁₂ @PIM-1-AO composite fibers exhibited complete catalytic conversion of CEES within 60 min with a selectivity of 86 % toward CEESO. However, full conversion of CEES by PW ₁₂ @PIM-1-AO composite powders required 90 min with 87 % selectivity toward CEESO. Compared with the full conversion and 95 % selectivity by homogeneous PW ₁₂ in 60 min, the trapping of CEESO in the pores of the composite possibly slightly reduced its selectivity. At the same conditions, negligible conversion of CEES occurred without an active catalyst or with PIM-1-AO fiber. The composite fibers with variable POM loading amounts could be prepared by immersing PIM-1-AO fibers into POM solution with different concentrations. Marginal conversion in the leaching tests further demonstrated the heterogeneous nature of the catalytic process. After three cycles, the conversion of CEES reached 72 ± 3.3 % after 60 min.

Figure 8:

The synthesis routine of PIM-based composite. Reproduced with permission from ref. 69]. Copyright 2022 American chemical Society.

Wang’s group designed a graphene supported polyoxotungstate K_1·5Cs_5.5 [γ-SiW₁₀O₃₉Cu₂(N₃)₂] (SiW ₁₀ Cu ₂ (N ₃ ) ₂ ) composite through a “click” reaction between the azido group of POM and the alkynyl groups on functionalized graphene, which boasted several advantages redox properties, higher thermal stability, electronic band structure, special hydrophobicity, and monodisperse sites (Figure 9). ⁷¹ Alkyne group-functionalized graphene (GO−C ≡ CH) was obtained through the reaction between graphene functionalized with a –COOH group and 4-(2-propynyloxy)aniline, which then reacted with SiW ₁₀ Cu ₂ (N ₃ ) ₂ through a “click” reaction. The Raman spectrum of SiW ₁₀ Cu ₂ /GO with the peaks at 935 and 717 cm⁻¹ confirmed the retention of the SiW ₁₀ Cu ₂ structure. XPS analysis of SiW ₁₀ Cu ₂ /GO showed binding energy peaks at ∼935 and 932 eV corresponding to Cu²⁺ and Cu⁺, respectively. This facilitated the electron transfer between graphene and SiW ₁₀ Cu ₂ and promoted the dismutation of H₂O₂ to O₂ and the oxidation reaction. TEM and EDX images of SiW ₁₀ Cu ₂ /GO revealed a single molecular layer of particles, approximately 1.2 nm in size (close to 10 Å of POMs) on the surface of SiW ₁₀ Cu ₂ /GO samples. SiW ₁₀ Cu ₂ /GO was used to catalyze the oxidation of CEES with H₂O₂. The conversion of CEES without SiW ₁₀ Cu ₂ /GO was 34.0 % with the selectivities of 50.9 % and 46.2 % toward CEESO and CEESO₂, respectively, after 60 min. A similar result was observed with the coexistence of graphene and H₂O₂. The addition of POM increased the conversion to 64.5 % with 100 % selectivity toward CEESO. The turnover frequency (TOF) increased from 0.748 min⁻¹ for SiW ₁₀ Cu ₂ (N ₃ ) ₂ to 51.556 for SiW ₁₀ Cu ₂ /GO. Under SiW ₁₀ Cu ₂ /GO catalysis, CEES was almost completely and selectively oxidized to CEESO. Radical scavenger experiments indicated no OH or superoxide free radicals in this catalytic system. The time-dependent absorbance decrease at 410 nm of 1,3-diphenylisobenzofuran (DPBF) (50 μM) with SiW ₁₀ Cu ₂ /GO catalyst (1 mg) and H₂O₂ (1.5 μM) indicated the emergence of singlet oxygen in this catalytic reaction. The CEES conversion of 72.0 % from the mixture of SiW ₁₀ Cu ₂ (N ₃ ) ₂ and graphene after 60 min highlighted synergistic effects. Silence in the UV−vis band at 300−400 nm for SiW ₁₀ Cu ₂ (N ₃ ) ₂ with H₂O₂ suggested that no peroxidation occurred, except for H₂O₂ dismutation. A decomposition efficiency of 35.9 % at 5 min and 61.5 % at 30 min for H₂O₂ on SiW ₁₀ Cu ₂ /GO was observed, with a decomposition rate of 1.35 × 10⁻¹ μmol s⁻¹ for H₂O₂, exceeding the 1.89 × 10⁻² μmol s⁻¹ rate of SiW ₁₀ Cu ₂ (N ₃ ) ₂  at 30 min. This efficiency benefited from SiW₁₀Cu₂ being bonded as single molecules to the graphene surface, unlike the assembly of SiW ₁₀ Cu ₂ (N ₃ ) ₂ . XPS result revealed that the molar ratio of Cu²⁺ and Cu⁺ in SiW ₁₀ Cu ₂ /GO was 1.3:1, which possibly enhanced the dismutation of H₂O₂ to O₂. A slightly higher reaction rate for heterogeneous CuCl compared to CuCl₂ indicated a higher activity for Cu⁺. SiW ₁₀ Cu ₂ (N ₃ ) ₂ achieved a 48.1 % conversion of CEES at 6 h through aerobic oxidation. For SiW ₁₀ Cu ₂ /GO, the conversion of CEES was significantly increased with increasing concentrations of O₂ and CEES around the graphene interface. The adsorption amount of SiW ₁₀ Cu ₂ /GO for CEES and O₂ was much larger than that of SiW ₁₀ Cu ₂ (N ₃ ) ₂ . The hydrophobic environment of SiW ₁₀ Cu ₂ /GO readily adsorbed more oleophilic molecules, but not polar molecules. When oleophilic CEES was.

Figure 9:

The schematic diagram of simultaneous dismutation of H₂O₂ to O₂ and oxidation of CEES by SiW₁₀Cu₂/GO catalyst. Reproduced with permission from ref. 71]. Copyright 2020 American chemical Society.

oxidized to polar CEESO, CEESO was repelled and escaped from SiW ₁₀ Cu ₂ /GO, preventing peroxidation to CEESO₂, as confirmed by XPS and IR experiments. Electron transfer and synergistic effects between SiW ₁₀ Cu ₂ and graphene promoted the aerobic oxidation of CEES. A selectivity of 100 % toward CEESO was still achieved with physically mixed hybrid SiW ₁₀ Cu ₂ (N ₃ ) ₂ /GO, although the conversion was only 49.0 % at 6 h. ⁷² Near-perfect CEES conversion and minimal leaching were observed after 10 cycles of SiW ₁₀ Cu ₂ (N ₃ ) ₂ /GO. The maintained structure and morphology confirmed by IR, ²⁹Si MAS NMR, XPS and TEM images of SiW ₁₀ Cu ₂ /GO after the cycle experiment demonstrated its potential as a catalyst for mustard gas analogue removal via aerobic oxidation at room temperature without additives.

Farha’s group prepared a composite material PW ₁₂ @NU-1000 via the direct impregnation method for catalytic oxidation of CEES leveraging the advantages of MOF NU-1000 and POM H₃PW₁₂O₄₀ (PW ₁₂ ), such as small triangular channels (12 Å) and larger hexagonal channels (31 Å) for fast substrate diffusion, the suitable size and high catalytic activity. ⁷³ About 0.8 [PW₁₂O₄₀]^3- was loaded per Zr₆ node in NU–1000 as confirmed by ICP-OES. SEM images with EDS mapping showed the intact size and morphology of NU-1000 and the uniform dispersion of PW ₁₂ within the MOF, not just on the surface. Compared to bare NU-1000, the BET surface area and the length of the large hexagonal pores of PW ₁₂ @NU-1000 were greatly decreased from 2,100 m²/g and 31 Å to 850 m²/g and 25 Å, respectively. The absence of non-H bonded–OH and H₂O characteristic peaks on the NU-1000 node and the presence of H-bonded H₂O and–OH peaks in diffuse reflectance infrared fourier-transform spectroscopy (DRIFTS) of PW ₁₂ @NU-1000 indicated that the non-H-bonded H₂O and hydroxyl ligands were replaced by POM. PW ₁₂ @NU-1000 exhibited good stability in HCl solution (pH 2–7) and only a 13 % loss of POMs after one week at pH 6, with less lost at lower pH values. PW ₁₂ @NU-1000 effectively catalyzed the oxidation of CEES with a TOF of 10.4 min⁻¹ in acetonitrile (1 mL) at 45 °C. In contrast, PW ₁₂ and NU-1000 alone had lower TOFs of 9.3 and 3.1 min⁻¹, respectively. For instance, full conversion of CEES was obtained by PW ₁₂ within 90 min, whereas PW ₁₂ @NU-1000 needed only 20 min under the same conditions. The cooperative effect between the POM and MOF endowed PW ₁₂ @NU-1000 with a faster initial rate and higher conversion efficiency than either component alone. However, the selectivity of PW ₁₂ to CEESO was superior to that of PW ₁₂ @NU-1000. The 10 % conversion and absence of a phosphorous NMR signal after the filtration of PW ₁₂ @NU-1000 suggested negligible leaching of POMs from the MOF during catalysis. The selectivity and reactivity were maintained upon recycling under these catalytic conditions, and the integrity of the composite material was unchanged in the cyclic catalytic process.

Farha’s group synthesized PW ₁₂ @NU-1000-scCO ₂ and PW ₁₂ @NU-1000-120 ^o C by the incorporation of H₃PW₁₂O4₀ (PW ₁₂ ) into the MOF, NU-1000, and the influence of the POM on catalytic performance was studied. ⁷⁴ PW ₁₂ @NU-1000-scCO ₂ was obtained via activation by supercritical CO₂ drying, and PW ₁₂ @NU-1000-120 ^o C was synthesized in 120 °C oven for over 1 h. ICP-OES analysis indicated a maximum loading of 0.8 POMs per Zr₆ node in both composites, with near uniform PW ₁₂ distribution (Figure 10). In situ variable temperature PXRD patterns showed that PW ₁₂ @NU-1000-scCO ₂ was changed into PW ₁₂ @NU-1000-120 ^o C via heating to and maintaining at 120 °C, and the process was not reversible. Through differential envelope density (DED) analyses, POMs of PW ₁₂ @NU-1000-scCO ₂ were located in the hexagonal mesoporous channels, as opposed to the triangular microporous channels of PW ₁₂ @NU-1000-120 ^o C. N₂ adsorption isotherms and DFT results indicated a reduction in mesopore volume for PW₁₂@NU-1000-scCO₂ and micropore volume for PW ₁₂ @NU-1000-120 ^o C, with the BET surface area decreasing from 1,020 m² cm⁻³ to 700, and 850 m² cm⁻³, respectively. The difference in BET was due to 5 wt% water present during the activated process of PW ₁₂ @NU-1000-scCO ₂ . At 170 °C, and a slight exothermic change in the differential scanning calorimetry (DSC) curve of PW ₁₂ @NU-1000-scCO ₂ indicated the loss of water H-bonded to the acidic protons of PW ₁₂ @NU-1000-scCO ₂ , correlating to POM movement from mesopore to micropore under the increasing van der Waals forces. In the micropore, POM was encircled by three pyrene linkers, as opposed to one pyrene linker in the mesopore. The broader and shifted of ³¹P signal indicated the loss of phosphorous symmetry and that POM strongly interacted with the MOF. The shift of reduction events to more positive potentials in the cyclic voltammograms (CVs) of PW ₁₂ @NU-1000-120 ^o C from H₃PW₁₂O₄₀ also indicated a stronger interaction between PW ₁₂ and pyrene, making it more easily reducible than POM alone. Fluorescence quenching of bare NU-1000 by POM about 95–99 % indicated the electron transfer from pyrene linkers to POM. Catalytic oxidation of CEES to CEESO was used to evaluate the catalytic performance of both composites. The half-lifes of catalytic oxidation of CEES were 13, 5, 3, and 1 min for NU-1000, PW ₁₂ , PW ₁₂ @NU-1000-120 ^o C, and PW ₁₂ @NU-1000-scCO ₂ , respectively. The selectivity of PW ₁₂ @NU-1000-scCO ₂ toward CEESO was 90.5 %, which was higher than that of PW ₁₂ @NU-1000-120 ^o C (59 ± 7 %). Similar to its selectivity toward CEESO, the TOF of PW ₁₂ @NU-1000-scCO ₂ was about three times that of PW ₁₂ @NU-1000-120 ^o C. POMs situated in the mesopores facilitated sulfide diffusion to POM and the near-exclusive generation of singly-oxidized product. Leach tests confirmed the heterogeneity of the reaction and the stability of POMs in the composite during catalysis.

Figure 10:

Visual structures of PW₁₂@NU-1000-scCO₂ and PW₁₂@NU-1000-120 °C. Reproduced with permission from ref. 74]. Copyright 2018 The royal society of Chemistry.

Compared with PW ₁₂ @NU-1000, PV ₂ Mo ₁₀ @NU-1000 was designed for catalytic degradation of CEES through immobilising decamolybdodivanado phosphoric acid (H₅PV₂Mo₁₀O₄₀, PV ₂ Mo ₁₀ ) in NU-1000. ⁷⁵ ICP-OES revealed a loading of 1.3 PV ₂ Mo ₁₀ per Zr₆ node. SEM images and EDS scans showed that the NU-1000 crystallite was retained, with POM homogeneously distributed throughout the crystallite. The BET surface areas of PV ₂ Mo ₁₀ @ NU-1000-scCO ₂ and PV ₂ Mo ₁₀ @NU-1000-80 °C slightly increased to 1,160 and 1,190 m² cm⁻³ from the original 1,020 m² cm⁻³ of NU-1000. The mesoporous peak was reduced and the microporous peak was retained for PV ₂ Mo ₁₀ @NU-1000-scCO ₂ . The opposite phenomenon was observed for PV ₂ Mo ₁₀ @NU -1,000-80 °C. These results suggested that POMs were located in the mesopores of PV ₂ Mo ₁₀ @NU -1000-scCO ₂ and in the micropores of PV ₂ Mo ₁₀ @NU-1000-80 °C, which was verified by comparison of PXRD pattern of NU-1000 and PV ₂ Mo ₁₀ @NU-1000-scCO ₂ . ⁷⁶ In situ temperature-dependent PXRD experiments revealed that POM migration from mesopores to micropores occurred earlier in PV ₂ Mo ₁₀ @NU-1000-scCO ₂ than in PW ₁₂ @NU-1000. The ³¹P NMR spectra of both composites broadened with slight shifts, indicating slight distortion of POM due to its interaction with the MOF framework. Compared to PMo ₁₂ @NU-1000, vanadium atoms were found to play a crucial role in aerobic activity, consistent with previous reports. ⁷³ CEESO₂ was almost the only product even after CEES consumption. Despite optimising the reaction conditions, over-oxidized CEESO₂ was still formed and the reaction rate was severely decreased. Although approximately 20 % POM leached from both composites, 100 % CEES conversion by PV ₂ Mo ₁₀ @NU-1000 was achievable in a second catalytic trial. In contrast, 5 % conversion was obtained using recycled H₅PV₂Mo₁₀O₄₀ due to POM decomposition. The in situ PXRD patterns of the composite, either as a dry powder or as a suspension in cyclohexane, indicated that solvent molecules could hinder migration of POM from mesopore to micropore. The MOF support strengthened the stability of POM in the composite.

An’s group designed two novel POMOFs {[ε-PMo₈^VMo₄^VIO₃₇(OH)₃][Zn₂(C₁₀N₂H₈) (H₂–O)₂]₂}₂·8H₂O (PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β) for the selective catalytic oxidation of multitudinous sulfides in which POMs was well dispersed at the molecular level and more active sites were exposed. ⁷⁷ PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β were prepared at a pH = 5.3 ± 0.1 and 4.7 ± 0.1, respectively, containing four Mo^V and two Mo^VI in the asymmetric unit. One-dimensional inorganic chains were initially formed by linking one hexacoordinated Zn₄-ε-Keggin POM unit with two adjacent Zn₄-ε-Keggin POM units perpendicular to the ab plane, creating a 3D framework through coordination with bpy ligands along the c axis, parallel to the ab plane. The difference between PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β was the rotation of approximately 39° clockwise acround the b axis. The XPS spectra differences for P 2p and Zn 2p in PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β indicated that P⁵⁺ and Zn⁺² were in two distinct POMs. According to the deconvolution of the spectra, the ratio of Mo^V/Mo^VI was approximately 2:1 for POMOFs PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β. The broad absorption band at 400–800 nm indicated the electronic transition of reduced molybdenum ions. PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β were used as heterogeneous catalysts in the catalytic oxidation of CEES. Both exhibited outstanding catalytic oxidation performance of CEES under conditions of n_CEES/n_pre-catalyst/n_oxidant = 1:0.005:1.08, at room temperature, in 0.5 mL of methanol. Over 99 % of CEES was oxidized into CEESO with more than 99 % selectivity by PMo ₁₂ O ₃₇ Zn ₄ -α in 15 min and PMo ₁₂ O ₃₇ Zn ₄ -β in 18 min (Figure 11). The catalytic efficiency decreased slifhtly after four consecutive trials. The structure remained unchanged before and after the catalytic reaction, demonstrating remarkable stability and recyclability. The catalytic oxidation reactions showed quasi-first order kinetics, with kinetic constants of 0.1561 and 0.2484 min⁻¹ for PMo ₁₂ O ₃₇ Zn ₄ -α and PMo ₁₂ O ₃₇ Zn ₄ -β, respectively. The synergistic effect of the raw materials along with their unique structure significantly contributed to the excellent catalytic activity and stability. The appearance of a new absorption peak at 237 nm and Raman spectrum peak at 760 cm⁻¹ of a PMo ₁₂ O ₃₇ Zn ₄ -α solution treated with H₂O₂ assigned to O–O stretching indicated the formation of peroxo species in the catalytic process, which attacked the electron-rich S atom of the sulfide to produce the target product. The addition of multiple scavengers did not effectively prevent the oxidation reaction, further confirming the generation of peroxo species in the catalytic process.

Figure 11:

The catalytic performance of CEES by PMo ₁₂ O ₃₇ Zn ₄ -α (a) and PMo ₁₂ O ₃₇ Zn ₄ -β (b). Reproduced with permission from ref. 77]. Copyright 2021 American chemical Society.

Yang’s group designed four polyoxomolybdate-based inorganic−organic hybrid materials for the catalytic oxidation of HD stimulant [Mn (TMR4A) (H₂O)₄][Mo₆O₁₉]·0.5CH₃CH₂OHH₂O [Ni(TMR4A) (H₂O)₄][Mo_6–O₁₉]·0.5CH₃CH₂OHH₂O [Zn (TMR4A) (H₂O)₄][Mo₆O₁₉]·0.5CH₃CH₂OHH₂O, and Co₂(TMR4A)₂(H₂O)₄ (β–Mo₈O₂₆)]·CH₃CN·12H₂O using a self-assembly strategy (Figure 10). ⁷⁸ Synthesis in different solvents, these hybrids exhibited distinct crystal structures. The first three samples contained [Mo₆O₁₉]²⁻ anions and the last one contained [β-Mo₈O₂₆]⁴⁻ anions. All four materials remained stable in various organic solvents and H₂O (pH = 0–12) for 12 h. Without catalyst, only 12 % of CEES was converted to CEESO and CEESO₂ at 40 °C for 25 min, using 0.25 mmol CEES, 4 ml CH₃OH and 1 mmol H₂O₂. With the addition of [Mn (TMR4A) (H₂O)₄][Mo₆–O₁₉]·0.5CH₃CH₂OHH₂O (1.25 μM), the conversions of CEES could reach 93 %and 99 % with 97 % selectivity at 30 and 40 °C, respectively, after 25 min, demonstrating its superiority as a catalyst for CEES. No catalytic activity was observed after filtering the catalyst, and the absence of Mn^II cations in the filtrate indicated a heterogeneous reaction. FTIR spectra and PXRD patterns of the catalyst after four cycles confirmed its structural integrity. Catalytic experiments using multiple components indicated that polyoxomolybdate was the primary catalytic active site.

Jiao’s group fabricated four nylon-6/POMs composite nanofibrous membranes (H₃PW₁₂O₄₀·xH₂O, H₄PMo₁₁VO₄₀·xH₂O, H₅PMo₁₀V₂O₄₀·xH₂O and H₆PMo₉V₃O₄₀·xH₂O, abbreviated as nylon-6/PW, PMoV1, PMoV2 and PMoV3) by electrospinning. ⁷⁹ FTIR absorption peaks at 1,080–1,060 cm⁻¹, ca. 960 cm⁻¹, and 860–900 cm⁻¹, 780 cm⁻¹ assigned to the corresponding stretching vibrations of P–O, Mo=O and Mo–O–Mo bonds, confirmed the successful preparation of PMoV1, PMoV2 and PMoV3. SEM images showed that the diameter of nylon-6/PW nanofibers was 500–700 nm, whereas that of nylon-6/PMoV1, nylon-6/PMoV2 and nylon-6/PMoV3 nanofibers ranged from 50 to 100 nm. The higher viscosity of nylon-6/PW precursor solution led to the larger fibers. The nylon-6/POMs membranes exhibited decent thermal stability below 200 °C. Compared with the other composite membranes, nylon-6/PMoV3 exhibited the highest HD degradation rate, which was 41.55 % in 6 h. Increasing the V content in POMs was beneficial for redox catalysis and resulted in the appearance of more reactive lattice oxygen associated with the Mo–O–V species. For nylon-6/PMoV1, the relative lower degradation efficiency was attributed to reduced mobility due to water loss. Considering the aerosol form of HD at ambient conditions, the filtration behaviors of these composite membranes were evaluated. Using nylon-6/PMoV3 as a typical example and PP nonwoven fabric as a support the aerosol filtration efficiency increased significantly to 98.35 % in 3 h, much greater than that of PP nonwoven fabric alone (11.45 %). The mass per square meter of the nylon-6/PMoV3 fibrous membrane was about 2.6 times that used in typical HD degradation experiments. Most HD was blocked by the membrane, and penetrating HD was rapidly degraded, indicating the potential of these composite nanofibrous membranes in soft protecting materials for in situ degradation of HD.

Babri’s group prepared H₃PW₁₂O₄₀/TiO₂ and H₆P₂W₁₆O₆₂/TiO₂ via a sol-gel technique for photocatalytic degradation of HD. ⁸⁰ To understand solvent effect, photocatalytic oxidation of HD by illuminated TiO₂ (150 mg) was investigated in relatively low viscosity solvents: hexane, acetonitrile, and acetone. After 90 min of irradiation, degradation rates of HD in the above solvents were 95 %, 90 %, and 88 %, respectively. The lower degradation rates in acetonitrile and acetone were due to reduced adsorption of the agent on TiO₂ surfaces, caused by the good solubility of HD in these solvents. Photocatalytic degradation of HD by POMs/TiO₂ was carried out in hexane because the physically adsorbed HD on the nTiO₂ surface could not be completely washed by hexane and hexane was inert and transparent. However, POMs/TiO₂ did not exhibit the ideal photocatalytic decontamination efficiency toward HD, indicating that POMs embedded in TiO₂ did not significantly enhance the catalytic effect. Under no radiation filtering, HD was possibly photolysed by the radiation. By only UV irradiation, the decontamination efficiency of HD could reach to 90 % via direct photolysis in 360 min. However, nearly 95 % HD was eliminated by nanoTiO₂ in 90 min. GC-MS analysis revealed that sulfoxide was initially formed, followed by sulfone, and elimination reactions occurred on both oxidation products, producing monovinyl and divinyl sulfoxides and sulfones. During photocatalytic degradation, no products or intermediates were detected, suggesting that the degradation products were adsorbed on the TiO₂ surface. These products were detected when acetone was used as the extracting solvent for the TiO₂ surface. Due to the presence of hydroxyl groups on the surface of nanoTiO₂, hydrolytic reactions still occurred, but the incomplete hydrolysis products were less toxic than HD itself.