Synthesis and Properties of Cobalt/Nickel-Iron-Antimony(III, V)-Oxo Tartrate Cluster-Based Compounds

Two types of isostructural iron-cobalt/nickel-antimony-oxo tartrate cluster-based compounds, namely (H3O)(Me2NH2)[M(H2O)6]2[FeII2SbIII12(μ4-O)3(μ3-O)8(tta)6]·6H2O (M = Co (1); Ni (3)), H5/3[Co2.5FeII4/3FeIII3(H2O)13SbV1/3FeIII2/3(μ4-O)2(μ3-O)4SbIII6(μ3-O)2(tta)6]·2H2O (2) and H2[Ni2.25FeII1.5FeIII3(H2O)14SbV0.25FeIII0.75(μ4-O)2(μ3-O)4SbIII6(μ3-O)2(tta)6]·2H2O (4) (H4tta = tartaric acid) were synthesized via simple solvothermal reactions. All the clusters in the structures adopt sandwich configurations, that is, bilayer sandwich configuration in 1 and 3 and monolayer sandwich configuration in 2 and 4. Interestingly, the monolayer sandwiched compounds 2 and 4 represent rare examples of cluster-based compounds containing mixed-valence Sb(III, V), whose center of the intermediate layer is the co-occupied [FexSbV1−x]. This is different from that of previously reported sandwich-type antimony-oxo clusters in which the center position is either occupied by a transition metal ion or a Sb(V) alone. Thus, the discovery of title compounds 2 and 4 makes the evolution of center metal ion more complete, that is, from M, MxSbV1−x to SbV. All the title compounds were fully characterized, and the photocatalysis, proton conduction and magnetism of compounds 2 and 4 were studied.


Introduction
In recent years, antimony oxides have been used in a wide range of applications such as nonlinear optics and catalysis [1][2][3][4][5][6].Crystalline Sb(III)-oxo-cluster-based compounds possess a precise structure and thus their performance can be effectively optimized through function-oriented structural design from the atomic point of view [1,2].The design and synthesis of organic-inorganic hybrid compounds based on Sb(III)-oxo clusters have drawn increasing attention recently.The organic components may play two roles in these hybrids, that is, acting as charge-balancing cations or as ligands entering or linking the Sb(III)oxo-based clusters [7][8][9][10][11][12][13][14][15].In the synthesis of these hybrids, potassium antimony tartrate has been widely favored by scholars for its feature of undergoing decomposition and rearrangement in water to form {Sb 3 (µ 3 -O)(tta) 3 } scaffolds with three bridging oxygen groups, which further capture other metal ions to form a cluster-based structure [13,16].
Figure 1 illustrates the synthesis methods for compounds 1, 2, 3 and 4. When only Co(NO 3 ) 2 •6H 2 O or Ni(NO 3 ) 2 •6H 2 O is added to the reaction vessel, compounds 1 and 2 or 3 and 4 can be obtained together.However, when 0.05 g of thiophene-2,5-dicarboxylic acid or pyrazole-3,5-dicarboxylic acid was added with Co(NO 3 ) 2 •6H 2 O or Ni(NO 3 ) 2 •6H 2 O at the same time as the starting materials of two syntheses, it was found that compound 2 as well as compound 4 could not be obtained, and only a small amount of compounds 1 and 3 could be obtained, respectively, and the quality of the crystals deteriorated compared to that from the original reaction.But when the reaction solvents were changed from 3 mL H 2 O as well as 1 mL DMF to 4 mL H 2 O, all of the above-mentioned crystals could not be obtained.In summary, we speculate that compounds 2 and 4 are synthesized under sensitive conditions and that minor pH changes may affect their synthesis [14].
Molecules 2024, 29, x FOR PEER REVIEW 3 of 13 H2O as well as 1 mL DMF to 4 mL H2O, all of the above-mentioned crystals could not be obtained.In summary, we speculate that compounds 2 and 4 are synthesized under sensitive conditions and that minor pH changes may affect their synthesis [14].

Description of the Structures
Single-crystal structure analysis indicates that compounds 1 and 3 are isostructural; so are compounds 2 and 4. Therefore, we chose compounds 1 and 2 as examples for the structural descriptions.The compound (H3O)(Me2NH2)[Co(H2O)6]2[Fe II 2Sb III 12(µ4-O)3(µ3- O)8(tta)6]•6H2O (1) crystallizes in the hexagonal crystal system, P6322 space group; its asymmetric unit comprises one-third of the molecular unit, i.e., one-third of the anionic cluster [Fe II 2Sb III 12(µ3-O)8(µ4-O)3(tta)6] 6− , two-thirds of the cationic [Co(H2O)6] 2+ complexes, one-third of H3O + cations and one-third of [Me2NH2] + cations, as well as two lattice water molecules.Its anionic cluster portion [Fe II 2Sb III 12(µ3-O)8(µ4-O)3(tta)6] 6− presents a bilayer sandwich configuration.The top and bottom layers of the cluster are two {Sb III 3(µ3-O)(tta)3} scaffolds (Figure 2c,e).The middle part is in a bilayer sandwich configuration consisting of two {Fe II (µ3-O)3(µ4-O)3Sb III 3} layers sharing three µ4-O (11) atoms (Figure 2a,b,d).The center of the {Fe II (µ3-O)3(µ4-O)3Sb III 3} layer is an [Fe II (µ3-O)3(µ4-O)3] octahedron, which is surrounded by three {SbO4} structural units that connect to the central {FeO6} via edgesharing a µ3-O (10) atom and a µ4-O (11).The average bond length of Fe-O and Sb-O in the layer is about 2.050 and 2.087 Å, respectively.Further, intermediate sandwich portions are connected by µ3-O atoms to the top and bottom layers {Sb III 3(µ3-O)(tta)3}, respectively, and together they assemble into the anionic cluster of [Fe II 2Sb III 12(µ3-O)8(µ4-O)3(tta)6] 6− (Figure 2f).The counterions of [Co(H2O)6] 2+ complexes, H3O + and [Me2NH2] + as well as lattice water molecules, wrap around the anionic cluster (Figure S1a,b).Compound 3 differs from compound 1 only in that the peripheral cation is changed from a [Co(H2O)6] 2+ complex to a [Ni(H2O)6] 2+ complex (Figure S1c,d).3b).In previously reported structures, the transition metals (e.g., Fe(II), Mn(II), Cu(II), etc.) occupy the center of the sandwich in majority, while occasionally the Sb(V) occupies the center [19,21].By contrast, in compound 2 the center of the sandwich layer is occupied by the {Sb V 1/3 Fe III 2/3 (µ 4 -O) 2 (µ 3 -O) 4 } octahedron, with an average Sb-O and Fe-O bond length of about 2.004 and 2.002 Å, respectively.This is the first time that Sb(V) is co-occupied with the transition metal Fe(III) in transitional-metal-antimony-oxo tartrate cluster-based compounds.More interestingly, the successful preparation of compounds 2 and 4 makes the transformation process of the central metal of this type of monolayer sandwich from transition metal ions to Sb(V) ions clearer and more complete.3b).In previously reported structures, the transition metals (e.g., Fe(II), Mn(II), Cu(II), etc.) occupy the center of the sandwich in majority, while occasionally the Sb(V) occupies the center [19,21].By contrast, in compound 2 the center of the sandwich layer is occupied by the {Sb V 1/3Fe III 2/3(µ4-O)2(µ3-O)4} octahedron, with an average Sb-O and Fe-O bond length of about 2.004 and 2.002 Å, respectively.This is the first time that Sb(V) is co-occupied with the transition metal Fe(III) in transitional-metal-antimony-oxo tartrate cluster-based compounds.More interestingly, the successful preparation of compounds 2 and 4 makes the transformation process of the central metal of this type of monolayer sandwich from transition metal ions to Sb(V) ions clearer and more complete.
As shown in Figure 3b, two [Fe III   Then the central layer of {Co2Fe II 4/3Fe III 3(H2O)10Sb V 1/3Fe III 2/3(µ4-O)2(µ3-O)4} connects to the top and bottom {Co0.25(H2O)1.5[SbIII 3(µ3-O)(tta)3]} layers via four µ3-O atoms and two µ4-O atoms to further assemble a sandwich structure (Figure 3d).Interestingly, the partially occupied [Co(2)0.25O6]as well as [Fe(5)0.5O6]polyhedra play the role of linkers, and thus the state of existence of the two groups affects the structural dimensions of the compounds.When none of the [Co(2)O6] octahedra are present, there is no way to further expand the anionic cluster portion, regardless of the presence or absence of [Fe(5)O6], at which point a discrete structure formed (Figure S2a); when only one of the [Co(2)O6] octahedra is present above or below, the discrete anionic cluster portion extends through [Co(2)O6] interconnections to form a 1-D chain structure (Figure S2b); when both [Co(2)O6] octahedra are present simultaneously, the 1-D chain is further interconnected through the [Co(2)O6] octahedra to form a 2-D layer structure (Figure S2c); however, when both [Co(2)O6] and [Fe(5)O6] polyhedra are present in the structure, the 2-D layer is further extended into a three-dimensional microporous structure by sharing O(28B) atoms through the [Co(2)O6] and [Fe(5)O6] polyhedra (Figure 3e).Regardless of the spatial stacking forms of compound 2, hydrogen ions as well as lattice water molecules are always located around the anion clusters.
Compounds 2 and 4 are isostructural: the difference lies in that the cluster in compound 2 consists of Fe-Co-Sb while that in 4 is Fe-Ni-Sb, and the contents of the three elements Fe, Sb and M(Co/Ni) are slightly different (Figure S3).In compound 4, the sandwich center is occupied by an [Sb V 0.25Fe III 0.75(µ4-O)2(µ3-O)4] octahedron; the content of Fe at the positions of Fe(4) and Fe(5) also differs slightly, with a slight increase in the Fe content at the position of Fe (5), which becomes [Fe II 0.715(5)O6], and an even smaller amount of Fe at the position of Fe(4), which becomes [Fe II 0.07(4)O6]; the [Ni(1)O6] octahedron and   5) connects the neighboring anionic cluster to form the dimeric units, and CoCd(6) does not serve as connecting units (Figure 4b).In addition, the two anionic clusters are connected and extended in different ways; compound 2 extends the cluster into a 3-D microporous structure by sharing O(28B) through partially occupied Fe (5) with Co(2) suspended at the periphery of the cluster (Figure 4c); whereas, the anionic cluster unit of {Cd[SbCo]} connects to each other through partially occupied Co (7) and Co(4) to form an 1-D chain, which in turn forms an 1-D belt-like structure through the bridging of two CoCd(5) compounds (Figure 4d).Therefore, it is expected that one can intentionally regulate the composition of the intermediate layer and connecting units at the periphery of the sandwich-typed antimony-oxo tartrate cluster to further construct more novel structures in terms of cluster size/composition, structural dimensionality and functionality, etc.
hanging on the periphery are not connected to each other in an edge-sharing manner as that for the [Co0.25(2)O6]octahedra; instead, they hang separately in the top or bottom layer of the anionic cluster (Figure S4).

Thermal Stability and UV-Vis Analysis
Prior to testing, all compounds were subjected to powder diffraction characterization and compared to simulated diffraction peaks obtained from single-crystal diffraction data to demonstrate the purity of the compounds (Figure S5).We ground the crystals of

Thermal Stability and UV-Vis Analysis
Prior to testing, all compounds were subjected to powder diffraction characterization and compared to simulated diffraction peaks obtained from single-crystal diffraction data to demonstrate the purity of the compounds (Figure S5).We ground the crystals of compounds 2 or 4 into powder and immersed them in water for 1 day, the powder diffraction results showed that both compounds were stable in water for 1 day (Figure S5b,d).The thermal stability of all compounds was also investigated (Figure S6).As shown in Figure S6, compounds 1 and 3 lost eighteen water molecules (six lattice and twelve coordination water molecules) and one H 3 O + as well as one dimethylammonium cation in the range of room temperature to 275 • C, respectively (calcd.12.43%, found 12.02% (1) and calcd.12.44%, found 12.59% (3)); compounds 2 and 4 lost fifteen (thirteen lattice and two coordination water molecules) and sixteen (fourteen lattice and two coordination water molecules) water molecules, respectively (calcd.10.92%, found 11.22% (2) and calcd.11.60%, found 11.62% (4)), in the temperature range of RT to 250 • C. In addition, we performed solid-state-UV-diffuse-reflectance measurements for all compounds, and the band gaps of 2.69 eV (1), 2.29 eV (2), 2.58 eV (3) and 2.32 eV (4) were ascertained from the fitting results of compounds 1-4, respectively, which are consistent with the colors of the corresponding compounds (Figure S7).

X-ray Photoelectron Spectroscopy (XPS) Analysis of 2 and 4
When resolving the structures of compounds 2 and 4, we determined the valence of Fe as well as Sb in the structures based on reported compounds with the same type.To further verify the accuracy of our results, we analyzed these two compounds via X-ray photoelectron spectroscopy (XPS) and bond valence calculations.The valence states calculation results show that all the Sb atoms in compounds 2 and 4 have a valence state of +3 except for the Sb( 4) atom which has a valence state of +5; the valence states of Fe(1), Fe(2) and Fe(3) atoms are +3, while the valence states of Fe(4) and Fe( 5) atoms are +2 (Table S2) [13,16,17,22].XPS results of compounds 2 and 4 further support our conclusions (Figure 5).In the survey spectra of compounds 2 and 4 (Figure 5a,d), the characteristic peaks of Fe, Co, Sb, C, O and Fe, Ni, Sb, C, O elements have been detected, respectively, which are consistent with the compositions of these two compounds.As shown in Figure 5b,e, the high-resolution XPS spectra of Sb 3d can all be fitted to two pairs of peaks; the split peaks at 528.98 and 538.43 eV (2) as well as 528.92 and 538.43 eV (4) correspond to Sb III 3d 5/2 and Sb III 3d 3/2 , respectively, while the peaks at 530.06 and 539.51 eV (2) and 529.92 and 539.76 eV correspond to Sb V 3d 5/2 and Sb V 3d 3/2 , respectively.Similarly, the high-resolution XPS pattern of Fe 2p can be fitted to two pairs of peaks, and the characteristic peaks located at 708.26 and 722.57(2) eV as well as 708.81 and 722.47 eV (4) correspond to Fe II 2p 3/2 and Fe II 2p 1/2 , respectively, while the split peaks located at 710.96 and 725.51 eV (2) and 711.75 and 725.62 eV (4) correspond to Fe III 2p 3/2 and Fe III 2p 1/2 , respectively (Figure 5c,f).The above XPS results are in agreement with those reported in the literature, which further confirms the coexistence of +3 and +5 valence Sb atoms and +2 and +3 valence Fe atoms in 2 and 4 [23,24].Furthermore, to the best of our knowledge, this is one of the few examples of heterometallic antimony-oxo clusters containing mixed valence Sb(III, V) [19,21].

Photodegradation Performance of Compounds 2 and 4
The vigorous development of printing, leather and other industries has generated a large amount of wastewater containing organic dye pollutants.Thus, it is crucial to solve the environmental pollution problems caused by these species [25][26][27].In recent years, a number of scholars have utilized metal oxides with precise atomic structures as catalysts

Photodegradation Performance of Compounds 2 and 4
The vigorous development of printing, leather and other industries has generated a large amount of wastewater containing organic dye pollutants.Thus, it is crucial to solve the environmental pollution problems caused by these species [25][26][27].In recent years, a number of scholars have utilized metal oxides with precise atomic structures as catalysts for photodegradation of organic dye wastewater [8, 28,29].On account of this, we investigated the photocatalytic degradation performance of compounds 2 and 4 as inhomogeneous reaction photocatalysts using methylene blue (MB) as a simulated pollutant.The photodegradation performance of MB solution after different times of light exposure in absence of a photocatalyst was measured with a UV-Vis spectrophotometer.According to the regression equations of the standard solution curves (Figure S8a), the degradation rate of the MB solution was only 23.07%(Figure S8b).Under the dark condition, the degradation rate of MB was less than 10% even in the presence of compounds 2 or 4 (Figure S8c-f).However, in the presence of compounds 2 and 4, the intensity of the characteristic absorption peak at 667 nm of the MB solution continuedly decreased with prolongation of light time (Figure 6).The degradation rates of MB solution via compounds 2 and 4 after 8 h of light exposure are calculated to be 93.58%(2) and 98.30% (4), respectively.The above results together prove that compounds 2 and 4 can act as the visible-light-driven photocatalysts to photocatalyze the degradation of MB solution.Furthermore, X-ray powder diffraction results also prove that the structures of 2 and 4 remained stable during the degradations (Figure S5b,d).

Photodegradation Performance of Compounds 2 and 4
The vigorous development of printing, leather and other industries has generated a large amount of wastewater containing organic dye pollutants.Thus, it is crucial to solve the environmental pollution problems caused by these species [25][26][27].In recent years, a number of scholars have utilized metal oxides with precise atomic structures as catalysts for photodegradation of organic dye wastewater [8, 28,29].On account of this, we investigated the photocatalytic degradation performance of compounds 2 and 4 as inhomogeneous reaction photocatalysts using methylene blue (MB) as a simulated pollutant.The photodegradation performance of MB solution after different times of light exposure in absence of a photocatalyst was measured with a UV-Vis spectrophotometer.According to the regression equations of the standard solution curves (Figure S8a), the degradation rate of the MB solution was only 23.07%(Figure S8b).Under the dark condition, the degradation rate of MB was less than 10% even in the presence of compounds 2 or 4 (Figure S8cf).However, in the presence of compounds 2 and 4, the intensity of the characteristic absorption peak at 667 nm of the MB solution continuedly decreased with prolongation of light time (Figure 6).The degradation rates of MB solution via compounds 2 and 4 after 8 h of light exposure are calculated to be 93.58%(2) and 98.30% (4), respectively.The above results together prove that compounds 2 and 4 can act as the visible-light-driven photocatalysts to photocatalyze the degradation of MB solution.Furthermore, X-ray powder diffraction results also prove that the structures of 2 and 4 remained stable during the degradations (Figure S5b,d).

Proton Conduction of Compounds 2 and 4
In previous work, we have demonstrated that such antimony tartrate cluster-based compounds have the potential to act as proton conducting materials [15,19,21].Given that compounds 2 and 4 can be stabilized in water, we evaluated the proton conductivity of columnar powder samples of the two compounds at different temperatures with a relative humidity of 98%.As shown in Figure 7, the resistance of the two examples of compounds decreases with increasing temperature, indirectly indicating that the proton conductivity of the compounds also increases.The proton conductivity was measured to be 1.81 × 10 −5 S•cm −1 for 2 and 1.39 × 10 −5 S•cm −1 for 4, respectively, when the temperature was 25 • C and the relative humidity was 98% (Figures S9 and S10).The proton conductivity continued to increase as the temperature raised and reached a maximum of 3.86 × 10 −4 S•cm −1 for 2 and 2.12 × 10 −4 S•cm −1 for 4 (Table S3), respectively, when the temperature reached 85 • C at 98% RH.The increase in temperature makes the water molecules move at a faster rate which may be responsible for the increase in proton conductivity.Compared to our previously reported antimony tartrate-based clusters of the same type, these two compounds show slightly lower proton conductivities, which may be due to the different contents of lattice water molecules.and 2.12 × 10 −4 S•cm −1 for 4 (Table S3), respectively, when the temperature reached 85 °C at 98% RH.The increase in temperature makes the water molecules move at a faster rate which may be responsible for the increase in proton conductivity.Compared to our previously reported antimony tartrate-based clusters of the same type, these two compounds show slightly lower proton conductivities, which may be due to the different contents of lattice water molecules.Based on the proton conductivity at different temperature, we calculated the activation energies regarding the proton conduction process of the two compounds, which are 0.31 eV and 0.28 eV, respectively (Figure 7c).According to previous studies, when the activation energy is less than 0.5 eV, the conduction mechanism is a Grothuss mechanism [30][31][32].In addition, the powder diffraction results after the proton conduction test show that the structures of the compounds remained stable (Figure S5b,d).
Figure 8 illustrates temperature dependence of magnetic susceptibility and the corresponding reciprocal one obtained at H = 0.1 T for 2 and 4. Both compounds exhibited analogous magnetic behavior.Their susceptibility increased with decreasing temperature Based on the proton conductivity at different temperature, we calculated the activation energies regarding the proton conduction process of the two compounds, which are 0.31 eV and 0.28 eV, respectively (Figure 7c).According to previous studies, when the activation energy is less than 0.5 eV, the conduction mechanism is a Grothuss mechanism [30][31][32].In addition, the powder diffraction results after the proton conduction test show that the structures of the compounds remained stable (Figure S5b,d).

Materials and Methods
All reagents for synthesis were purchased from commercial sources and used without further purification.
Synthesis of compounds 1 and 2: A mixture of K2Sb2(tta)2•3H2O (100.0 mg and 0.15   O 6 ] to [Sb V O 6 ], and thus completed the process of this change.We have also carried out studies related to photocatalytic degradation of MB solutions, proton conduction and magnetic properties for compounds 2 and 4, and the experimental results have shown that they are a class of multifunctional materials.In the future, we will adjust the experimental conditions to combine compounds 2 or 4 with organic carboxylic acid ligands to construct 2-D layered or 3-D framework structures with pores and then explore their potential in proton conductivity and ion exchange.

Figure 1 .
Figure 1.Schematic illustration of the synthesis of the title compounds.Reaction conditions as follows: heated at 100 °C for 4 days.Color scheme: S, yellow; N, blue; O, red and C, gray.For clarity, the hydrogen atoms are omitted.

Figure 1 .
Figure 1.Schematic illustration of the synthesis of the title compounds.Reaction conditions as follows: heated at 100 • C for 4 days.Color scheme: S, yellow; N, blue; O, red and C, gray.For clarity, the hydrogen atoms are omitted.

Figure 5 .
Figure 5. (a) The survey XPS spectrum of compound 2. High-resolution XPS spectra of Sb 3d (b) and Fe 2p (c) for compound 2. (d) The survey XPS spectrum of compound 4. High-resolution XPS spectra of Sb 3d (e) and Fe 2p (f) for compound 4.

Figure 5 .
Figure 5. (a) The survey XPS spectrum of compound 2. High-resolution XPS spectra of Sb 3d (b) and Fe 2p (c) for compound 2. (d) The survey XPS spectrum of compound 4. High-resolution XPS spectra of Sb 3d (e) and Fe 2p (f) for compound 4.

Figure 6 .
Figure 6.Liquid UV-visible absorption spectra of methylene blue (MB) solution containing compounds 2 (a) and 4 (b) over the varying illumination time.The initial concentration of MB solution was 30 ppm and the test temperature was 25 • C. The insets in (a,b) are the photos of MB solution photodegraded with compounds 2 and 4 for 8 h.

Figure 7 .
Figure 7. Nyquist plots from AC impedance data of 2 (a) and 4 (b) at 98% RH and varied temperatures between 35 and 85 °C.(c) Arrhenius plots of proton conductivity of compounds 2 and 4 at different temperatures under 98% RH.

Figure 7 .
Figure 7. Nyquist plots from AC impedance data of 2 (a) and 4 (b) at 98% RH and varied temperatures between 35 and 85 • C. (c) Arrhenius plots of proton conductivity of compounds 2 and 4 at different temperatures under 98% RH.

Figure 8 .
Figure 8. Plots of the χ and χ −1 versus T for compounds 2 (a) and 4 (b) in the 2~300 K temperature range in an applied field of H = 0.1 T. The red line indicates Curie-Weiss fitting.

Figure 8 .
Figure 8. Plots of the χ and χ −1 versus T for compounds 2 (a) and 4 (b) in the 2~300 K temperature range in an applied field of H = 0.1 T. The red line indicates Curie-Weiss fitting.
Figure S1.Structure of compounds 1 (a) and 3 (c) and packing diagrams of compounds 1 (b) and 3 (d).For clarity, hydrogen atoms, H 3 O + and [Me 2 NH 2 ] + cations are omitted.Color scheme: Fe, brown; Sb, turquoise; Co, pink; Ni, green; O, red and C, gray; Figure S2.Packing of 0-D isolated clusters in 2. The 1-D chain (b) and 2-D layer (c) in 2. For clarity, hydrogen atoms are omitted.Color scheme: Fe, brown; Sb, turquoise; Co, pink; O, red and C, gray; Figure S3.Structure of compound 4.For clarity, hydrogen atoms are omitted.Color scheme: Fe, brown; Sb, turquoise; Ni, green; O, red and C, gray; Figure S4.Packing diagram of compound 4.For clarity, hydrogen atoms and H 3 O + cations are omitted; Figure S5.Simulated and observed PXRD patterns of compounds 1 (a), 2 (b), 3 (c) and 4 (d); Figure S6.Thermogravimetric curves for compounds 1 (a), 2 (b), 3 (c) and 4 (d); Figure S7.Solid-state UV-Vis absorption spectra of compounds 1 (a), 2 (b), 3 (c) and 4 (d); Table S2.The bond valence sum calculations for compounds 2 and 4. Figure S8.(a) Linearity of absorbance (A) versus standard concentration (C) of MB solution.(b) Curve of C t /C 0 of MB solution with time (t) in the absence of photocatalyst.Liquid UV-visible absorption spectra of MB solution containing compound 2 (c) and compound 4 (d) under dark conditions at different times.C t /C 0 curves of MB solution for compounds 2 (e) and 4 (f) as a function of time (t) under dark (black line) or light (blue line) conditions; Figure S9.The Nyquist plot from AC impedance data of 2 at 98% RH and 25 • C; Figure S10.The Nyquist plot from AC impedance data of 4 at 98% RH and 25 • C; Table S3.Proton conductivity (σ) values of 2 and 4 under 98% RH at different temperatures; Figure S11.EDS results of compounds 1 (a), 2 (b), 3 (c) and 4 (d); Table ing to note that the center portions of the intermediate layer of 2 and 4 are occupied by [Sb V 1/3 Fe III 2/3 (µ 4 -O) 2 (µ 3 -O) 4 ] and [Sb V 0.25 Fe III 0.75 (µ 4 -O) 2 (µ 3 -O) 4 ] octahedrons, respectively, demonstrating that these two compounds appear to have a rare mixed valence state of Sb(III,V).In addition, we have captured the intermediate [Fe III x Sb V 1−x O 6 ] of the center metal from the simple transition metal [Fe III