Two polymolybdate-based complexes and their graphene composites with visible-light photo-responses

Abstract

Two polymolybdate-based complexes formulated as (H₂L1)₂(Mo₄O₁₃) (1) and (L2)₂(HL2)₂(Mo₈O₂₅)·7H₂O (2) have been hydrothermal synthesized and structurally characterized by single-crystal X-ray diffraction. In complex 1, one-dimensional (1D) [Mo^V₄Mo^VI₄O₂₆]_∞⁸⁻ chain is constructed by vertice-sharing {MoO₆} octahedra, and the uncoordinated protonated [H₂L1]²⁺ species link different [Mo₈O₂₆]_∞⁸⁻ chains via supramolecular forces. In complex 2, the [Mo₈O₂₅]²⁻ moieties are linked by the μ₃-[HL2]⁺ and μ₂-L2 ligands into 1D chain. Complex 2 possesses more narrow band gap than complex 1, and it shows stronger photocurrent response than complex 1 under the visible light (650 nm > λ > 350 nm). The different densities of states (DOS) of the two complexes are attributed to their different compositions and frameworks, thus giving rise to their different band gaps. With the addition of graphene into the two complexes, the electroconductivity of the complex/graphene composite material is much improved, and 2/graphene composite material shows much improved electrocatalytic activity for the H₂ evolution reaction (HER) than complex 2.

---

Introduction

Recently, metal complexes have received tremendous attention due to their powerful attributes of structural versatility and various potential applications.¹ Among their diverse functionalities, metal complexes can be used as promising semiconductor materials.² Recently, there has been an increasing number of studies demonstrating the potential of metal complexes as semiconductor materials for optoelectronic processes and devices.^3,4 Generally, determination of the band gap of a semiconductor material is beneficial for us to understand the origin of its specific properties, such as electrical conductivity, photovoltaics, and so forth. In our previous work, it was found that the ligands and the frameworks of metal complexes play important role in their band gaps, and the few-layered metal complex shows improved photoresponse than its bulk counterpart.⁵

Polyoxometalates (POMs), a large family of anionic metal oxide clusters of d-block transitional metals (W, Mo, V), are good electron reservoirs.⁶ They can undergo potential photo- or electron-induced electron- and proton-transfer processes without changing in structure, which makes them an interesting material in photonic and photovoltaic applications.⁷ The photoelectric properties of POMs can be adjusted by changing the structures and compositions.⁸ And a series of POMs including Keggin, Dawson, lacunary, and transitional metal trisubstituted structures have been investigated systematically to analyze their energy band structures.⁹

In the present work, in an attempt to investigate the band gaps and the photoelectrochemical behaviors of POM-based complexes, we synthesized two rigid N-donors, 4-(5-((2H-1,2,4-triazol-3-yl)-2H-1,2,4-triazol-3-yl)pyridine) (L1) and 5-[4-(1H-imidazol-1-yl)-phenyl]-2H-tetrazole (L2) (Scheme 1),^10,11 which possesses superior coordination ability due to the triazole, imidazole or tetrazole N atoms in the structure. Based on L1 and L2, two novel POM-based complexes formulated as (H₂L1)₂(Mo₄O₁₃) (1) and (L2)₂(HL2)₂(Mo₈O₂₅)·7H₂O (2) were hydrothermally synthesized and structurally characterized by single-crystal X-ray diffraction. Their cyclic voltammograms (CVs), UV-Vis absorption and diffuse reflectance spectra, band gaps and density of states as well as thermal stabilities have been investigated. In order to improve the electrical conductivities of the two POM-based complexes, their graphene composites have been synthesized and they are characterized by elemental analyses, powder X-ray diffraction (PXRD), Raman spectra and scanning electron microscopy (SEM). The electrochemical and photoelectrochemical behaviors of the complexes and their graphene composites have been measured.

Scheme 1 Schematic representation of L1 (a) and L2 (b).

Experimental

General considerations

All chemicals purchased were of reagent grade and used without further purification. Graphene powder was purchased from Nanjing XFNANO Materials Tech Co., Ltd. The melting point was determined using an uncorrected X-4 melting point apparatus of Beijing Kaifu Company. IR spectra were recorded as KBr pellets on a Nicolet iS50 FT-IR spectrometer. C Elemental analyses were performed on an Elementar Vario MICRO E III analyzer. The powder X-ray diffraction (PXRD) data were collected on a TD-3500 diffractometer using Cu Kα radiation. Raman spectra were conducted on a Nicolet iS50 Raman spectrometer. The morphologies of the samples were observed by scanning electron microscopy (SEM, JSM-7600F, JEOL). UV-Vis absorption and diffuse reflectance spectra were measured on a HITACHI U-4100 UV-Vis spectrophotometer. TGA was performed on a NETZSCH STA 449C thermogravimetric analyzer in flowing N₂ with a heating rate of 10 °C min⁻¹.

Electrochemical measurements

The electrochemical measurements were done in a three-electrode test cell with a saturated calomel electrode (SCE) and a platinum foil as the reference and counter electrode, respectively.

1 mg sample was ultrasonicated in a mixture of ethanol (1 mL) and Nafion (0.05 mL) solution, then was deposited on a glassy carbon electrode (GCE) and dried by an IR lamp to obtain the working electrode with a 0.2 cm² of the geometrical area. The electrodes were immersed in a 0.4 M acetic acid–sodium acetate buffer solution (pH 4.5, 50 mL) and a CHI660E electrochemical workstation was used in the electrochemical measurement.

Theoretical calculation

The band gaps and density of states (DOS) of complexes 1 and 2 were investigated by the plane-wave-pseudopotential approach based on the density functional theory (DFT). A standard Cambridge serial total energy package (CASTEP) code was used. The electron–core interaction was represented via ultrasoft pseudopotentials with a plane-wave basis cutoff energy of 500 eV for complex 1 and 260 eV for complex 2. The exchange–correlation potential is approximated by the generalized gradient approximation (GGA), and the Perdew, Burke, and Ernzerhof (PBE) GGA functional is used for the calculation. The self-consistent field (SCF) tolerance was 1.0 × 10⁵ eV per atom. The empty bands of 12 and k-points sets of (2 × 2 × 2) were used. In order to simplify the calculation, the uncoordinated solvent molecules were removed from the single crystal structural data of complex 2 for the energy calculation.

Photoelectrochemical measurements

Photoelectrochemical experiments were carried out in a conventional three-electrode setup with a Pt foil and a AgCl/Ag electrode as the counter electrode and the reference electrode, respectively. A fluorine doped tin oxide (FTO)-coated glass substrate (12–14 μS m⁻¹) was cleaned ultrasonically in detergent, acetone, ethanol and deionized water, respectively, then dried in vacuum. 2 mg sample was ultrasonicated in dehydrated methanol (2 mL), then 0.4 mL of the mixture was deposited on the FTO slice and dried by an IR lamp to obtain the working electrode with a 1.0 × 1.0 cm² of the light illumination area. And the supporting electrolyte was 0.4 M acetic acid–sodium acetate buffer solution (pH 4.5, 50 mL). The working electrode was irradiated with 300 W xenon lamp (CEL-HXF300, Beijing Aulight Co., 650 nm > λ > 350 nm) from the front side, and the photocurrent generation from the modified FTO electrode was detected by a voltammetric analyzer (model 660E, CH Instrument). Electrochemical impedance spectroscopy (EIS) measurements were carried out over the frequency range from 0.01 Hz to 1 MHz, and the amplitude of the potential perturbation was 0.005 V.

Synthesis of L1

L1 was prepared according to the literature method.¹⁰ Melting point: >250 °C.

Synthesis of L2

L2 was prepared according to the literature method.¹¹ Melting point: >250 °C.

Synthesis of (H₂L1)₂(Mo₄O₁₃) (1)

A mixture of L1 (0.06 mmol, 0.013 g), MoO₃ (0.09 mmol, 0.015 g), ZnSO₄·7H₂O (0.06 mmol, 0.017 g) and water (8 ml) was heated at 150 °C in Teflon-lined autoclaves for 3.5 days, then followed by slow cooling to room temperature. The pH value of the final mixture after hydrothermal synthesis was ca. 4.5. The resulting yellow block crystals were filtered off (yield: ca. 80% based on MoO₃). Elemental anal. found: C, 21.0%. Calcd for C₁₈H₁₈Mo₄N₁₄O₁₃: C, 21.2%. IR (KBr, cm⁻¹): 3483(m), 3250(m), 3100(s), 1894(w), 1638(s), 1531(m), 1508(s), 1440(m), 1367(m), 1293(m), 1239(s), 1183(m), 1096(m), 1028(m), 949(s), 915(s), 807(m), 714(s), 599(s).

Synthesis of (L2)₂(HL2)₂(Mo₈O₂₅)·7H₂O (2)

The reaction procedure was carried out in a similar manner to that of complex 1, but starting with a mixture of L2 (0.1 mmol, 0.021 g), (NH₄)₆Mo₇O₂₄·4H₂O (0.025 mmol, 0.031 g), CdCl₂·2.5H₂O (0.15 mmol, 0.034 g) and water (8 ml), and the mixture was heated at 180 °C in Teflon-lined autoclaves for 5 days. The pH value of the final mixture after the hydrothermal synthesis was ca. 4.5. The resulting yellow rod-like crystals were filtered off (yield: 85% based on Mo). Elemental anal. found: C, 22.1%. Calcd for C₄₀H₄₈Mo₈N₂₄O₃₂: C, 22.4%. IR (KBr, cm⁻¹): 3377(s), 3156(s), 3110(s), 3089(s), 2360(s), 2499(m), 1961(m), 1617(m), 1548(s), 1526(m), 1451(s), 1337(m), 1313(s), 1291(w), 1217(w), 1133(w), 1062(m), 1011(m), 884(w), 837(s), 762(s), 750(w), 654(w), 622(w), 531(w), 490(w).

Synthesis of the complex/graphene composite material

The synthesis of the complex 1 or 2/graphene composite material was carried out as described above for complex 1 or 2 except that 0.5 mg of graphene was extra added. The resulting black powder was filtered and characterized by elemental analysis, PXRD and SEM. The C% for complex 1/graphene and 2/graphene composite materials are 21.6 and 23.0%, respectively, indicating the content of the graphene in the two composite materials are ca. 0.4 and 0.6%, respectively.

X-ray crystallography

Single-crystal X-ray data for complexes 1 and 2 was collected on an Oxford XCalibur Eos diffractometer using graphite monochromated Cu Kα (λ = 1.54178 Å) radiation for complex 1 and Mo Kα (λ = 0.71073 Å) radiation for complex 2 at room temperature. Empirical absorption correction was applied. The structures were solved by direct methods and refined by the full-matrix least-squares methods on F² using the SHELXTL-97 software.¹² All non-hydrogen atoms were refined anisotropically. All of the hydrogen atoms were placed in the calculated positions. The crystal data and structure refinements for the two complexes are summarized in Table 1. Selected bond lengths and angles for the two complexes are listed in Table S1 in the ESI.† The CCDC reference numbers are the following: 1032253 for complex 1 and 1032252 for complex 2.†

Table 1 Crystal data and structure refinements for MOFs 1 and 2^a

a R1 = Σ‖F0| − |Fc‖/Σ|F0|; wR2 = Σ[w(F0^2 − FC^2)^2]/Σ[w(F0^2)^2]^1/2.
Complex	1	2
Empirical formula	C18H18Mo4N14O13	C40H48Mo8N24O32
M	1022.22	2144.54
Crystal system	Triclinic	Monoclinic
Space group	P1̄	C2/c
a/Å	7.9900(12)	16.6099(14)
b/Å	12.086(3)	14.2716(11)
c/Å	16.780(3)	26.463(3)
α/°	107.73(2)	90
β/°	91.705(14)	93.849(2)
γ/°	101.866(17)	90
V/Å^3	1503.0(5)	6259.0(9)
Z	2	4
Dcalcd/g cm^−3	2.259	2.276
μ/mm^−1	14.158	1.662
No. of unique reflections	524	5494
Reflections used [I > 2σ(I)]	382	4379
F(0 0 0)	992	4192
Goodness-of-fit on F^2	1.116	1.067
Final R indices [I > 2σ(I)]	R1 = 0.0749,	R1 = 0.0307
	wR2 = 0.2345	wR2 = 0.0645

---

Results and discussion

Crystal structure of (H₂L1)₂(Mo₄O₁₃)(1)

Single-crystal X-ray diffraction analysis reveals that complex 1 crystallizes in the triclinic space group P1̄ (Table S1†), with half [Mo₈O₂₆]⁸⁻ and two uncoordinated protonated [H₂L1]²⁺ ligands in the asymmetric unit. In complex 1, the [Mo₈O₂₆]⁸⁻ moiety has C_2h symmetry and is composed of six edge-sharing {MoO₆} octahedra and thus displays the characteristic β-octamolybdate arrangement (Fig. 1b).¹³ Valences sum calculations show that the crystallographically independent Mo(1), Mo(2) are in the +6 oxidation states [Mo(1)–O 1.67(17)–2.34(14) Å, Mo(2)–O 1.69(15)–2.35(14) Å], and Mo(3), Mo(4) are in the +5 oxidation states [Mo(3)–O 1.70(13)–2.33(14) Å, Mo(4)–O 1.78(19)–2.47(15) Å] (Table S1†).¹⁴ The [Mo^V₄Mo^VI₄O₂₆]⁸⁻ moiety is assembled into a one-dimensional (1D) chain constructed by vertice-sharing {MoO₆} octahedra (Fig. 1a).

Fig. 1 1D [Mo^V₄Mo^VI₄O₂₆]_∞⁸⁻ chain constructed by vertice-sharing {MoO₆} octahedra in complex 1 (a); 3D supramolecular architecture constructed by strong H bonds and π–π stacking interactions in complex 1 (H bonds denoted in pink dotted lines) (b).

The two protonated [H₂L1]²⁺ ligands are almost planar molecules with 1.7, 11.7, 0.7 and 9.1° of the dihedral angles between two neighboring aromatic rings. Strong H bonds (Table 2) and π–π stacking interactions (Table 3) are observed in complex 1, and complex 1 shows a three-dimensional (3D) supramolecular architecture (Fig. 1b).

Table 2 Distances (Å) and angles (°) of the selected hydrogen bonds in complexes 1 and 2^a

D	H	A	D⋯A distance	H⋯A distance	∠D–H⋯A
a Symmetry transformations used to generate the equivalent atoms: #1: −x + 2, −y + 1, −z + 2; #2: x − 1, y − 1, z; #3: x, y − 1, z; #4: −x + 1, y, −z + 1/2; #5: −x + 1, −y + 1, −z + 1.
Complex 1
N1	H1A	N13#1	2.794	2.013	151
N7	H7A	O1#2	2.816	1.963	171
N10	H10A	N5#1	2.745	1.959	151
N14	H14	O4#3	2.459	1.621	164
Complex 2
N4	H4	O12#4	2.948	2.101	168
N4	H4	O3#4	2.796	2.369	111
N10	H10	O4#5	2.693	1.861	162

---

Table 3 The centroid–centroid (CC) distance (Å) and perpendicular (P) distance (Å) involving π⋯π stacking interactions for complexes 1 and 2^a

Plane	Plane	CC distance	P distance
a Symmetry transformations used to generate equivalent atoms: A −1 + x,y,z, B 2 − x,1 − y,2 − z, D 1 − x,y,1/2 − z, E 1/2 + x,3/2 − y,−1/2 + z, F 1 − x,y,3/2 − z, G −1/2 + x,3/2 − y,1/2 + z.
Complex 1
N1–N3/C1/C2	N8A–N10A/C10A/C11A	3.7(2)	3.32(14)
N4–N6/C3/C4	N11A–N13A/C12A/C13A	3.6(2)	3.20(17)
N7/C5–C9	N14A/C14A–C18A	3.59(15)	3.40(11)
N8–N10/C10/C11	N11B–N13B/C12B/C13B	3.9(3)	3.07(18)
Complex 2
C1/N1–N4	C1D/N1D–N4D	3.919(2)	3.1889(18)
N5–N6/C8–C10	C12E–C17E	3.945(3)	3.799(2)
N11/N12/C18–C20	N11F/N12F/C18F–C20F	3.984(3)	3.530(2)
N11/N12/C18–C20	C2G–C7G	3.633(3)	3.467(2)

---

Crystal structure of (L2)₂(HL2)₂(Mo₈O₂₅)·7H₂O (2)

As described above, in complex 1, the ligand L1 is uncoordinated with the Mo center. In comparison, another POM-based complex with the Mo–N coordination bond was obtained. Complex 2 crystallizes in the monoclinic space group C2/c, its asymmetric unit contains half [Mo₈O₂₅]²⁻, one coordinated L2, one coordinated protonated [HL2]⁺ and 3.5 uncoordinated water molecules. One crystallographically independent [HL2]⁺ shows a tridentate coordination mode linking three Mo atoms with its three N atoms from the tetrazole group, leaving the imidazole N atoms uncoordinated. The crystallographically independent L2 acts as a bidentate ligand with two tetrazole N atoms coordinated with the Mo(2) and Mo(3) centers. Valences sum calculations show that all the Mo atoms in complex 2 are in the +6 oxidation states.¹⁴ The crystallographically independent Mo(1), Mo(2) and Mo(3) atoms both display {MoO₅N} octahedral coordination geometries [Mo(1)–O 1.693(3)–2.209(3) Å, Mo(1)–N 2.428(4) Å, Mo(2)–O 1.699(3)–2.203(3) Å, Mo(2)–N 2.456(4) Å, Mo(3)–O 1.698(3)–2.224(3) Å, Mo(3)–N 2.647(4) Å] (Table S1†). And the crystallographically independent Mo(4) atom exhibits a {MoO₄N₂} octahedral coordination geometry [Mo(4)–O 1.703(4)–1.9230(16) Å, Mo(4)–N 2.322(4)–2.443(4) Å] (Table S1†). The [Mo₈O₂₅]²⁻ moieties are linked by the μ₃-[HL2]⁺ and μ₂-L2 ligands into a 1D chain, in which edge-sharing or vertice-sharing {MoO₅N} and {MoO₄N₂} octahedra are observed (Fig. 2a). The dihedral angles between the two neighboring aromatic rings of the [HL2]⁺ ligand are 8.8 and 26.3°, respectively. And the dihedral angles in L2 ligand are 16.0 and 18.8°, respectively, indicating both of them are almost planar molecules. Strong H bonds (Table 2) and π–π stacking interactions (Table 3) are observed in complex 2, and complex 2 shows a 3D supramolecular architecture (Fig. 1b).

Fig. 2 1D chain constructed by [Mo₈O₂₅]²⁻, L2 and [HL2]⁺ in complex 2 (H atoms omitted for clarity) (a); strong π–π stacking interactions are observed in complex 2 (The L2 and [HL2]⁺ ligands coordinated to different 1D chains denoted in different colors; uncoordinated water molecules omitted for clarity) (b).

The electrochemical behaviors of complexes 1–2 and their graphene composite materials

As shown in Fig. S1a and b,† the experimental PXRD patterns for complexes 1 and 2 match well with their respective simulated one (generated on the basis of single crystal structure analysis), confirming the phase purities of the bulk materials. The electrochemical properties of complexes 1 and 2 were evaluated by cyclic voltammetry (CV) in a three-electrode test cell with a saturated calomel electrode (SCE) and a platinum foil as the reference and counter electrode, respectively. And we modify the sample on glassy carbon electrode (GCE) to obtain the working electrode (Fig. S2†). As described above, complexes 1 and 2 were synthesized at the pH value of 4.5. Because POMs are very sensitive to the pH value, herein, the acetic acid–sodium acetate buffer solution (pH = 4.5) was chosen as the electrolyte solution to keep the pH value invariable. The CVs are measured in the potential range from −1.8 to 2.5 V vs. SCE at different scan rates (Fig. S3–S5†).

As shown in Fig. 3 and S3,† the bare GCE shows a weak reduction peak at −0.86 V vs. SCE at a scan rate of 0.005 V s⁻¹, which is probably attributed to the impurity on the electrode. L1-modified GCE (L1-GCE) exhibits an irreversible reduction peak at −1.10 V vs. SCE (Fig. 3 and S4†). And similar reduction peak at −0.64 V vs. SCE is observed in the CV of complex 1 (1-GCE) (Fig. 3 and S5†), indicating it is probably ascribed to the redox of the L1 ligand in complex 1. And the potentials for the H₂ evolution reaction (HER)/oxygen evolution reaction (OER) at L1-GCE and 1-GCE are similar, they are −1.42/+1.30 and −1.39/+1.30 V vs. SCE, respectively (Fig. 3, S4 and S5†).

Fig. 3 CVs of the bare GCE (pink), L1-GCE (green), 1-GCE (red) and 1/graphene-GCE (blue) in a 0.4 M acetic acid–sodium acetate buffer solution (pH = 4.5, 50 mL) in the potential range from −1.8 to 2.5 V vs. SCE at a scan rate of 0.005 V s⁻¹.

Similarly, the electrochemical response of complex 2 was also investigated in the buffer solution (CH₃COOH–CH₃COONa, pH = 4.5, 50 mL) and it is expected that complex 2 can suffer the CV experiment in the potential range from −1.8 to 2.0 V vs. SCE. As shown in Fig. 4 and S6,† the CV of L2-modified GCE (L2-GCE) shows an irreversible reduction peak at −0.80 V vs. SCE at a scan rate of 0.005 V s⁻¹. And similar irreversible reduction peak is observed at −0.83 vs. SCE in the CV of complex 2-modified GCE (2-GCE), indicating it is probably attributed to the redox of the L2 ligand in complex 2 (Fig. 4 and S7†). The potentials for the HER/OER at L2-GCE and 2-GCE are observed at −1.60/+1.90 V and −1.30/+1.87 V vs. SCE, respectively (Fig. 4, S6 and S7†).

Fig. 4 CVs of L2-GCE (green), 2-GCE (red) and 2/graphene-GCE (blue) in a 0.4 M acetic acid–sodium acetate buffer solution (pH = 4.5, 50 mL) at a scan rate of 0.005 V s⁻¹.

Generally, POM-based complexes are not electroconductive. Whereas graphene possesses extraordinary electron conductivity, thermal and mechanical properties, which attracts our attention. In an attempt to improve the conductivities of complexes 1 and 2, in the present work, we synthesize their graphene composite materials to investigate their electrochemical properties for comparison. Herein, graphene was commercially purchased from Nanjing XFNANO Materials Tech Co., Ltd, which exhibits graphene's characteristic Raman signals at approximately 1342 cm⁻¹ (D peak), 1571 cm⁻¹ (Q peak) and an overtone peak near 2682 cm⁻¹ (2D peak) (Fig. S8†).¹⁵ The synthesis of the complex/graphene composite was carried out as described above for complex 1 or 2 except that 0.5 mg graphene was extra added. The resulting black powder was filtered and characterized by PXRD, which displays a PXRD pattern similar to that of complex 1 or 2 (Fig. S1a and b†), indicating the black sample is the corresponding complex/graphene composite. The result is also proved by the SEM images, as shown in Fig. S9.†

As shown in Fig. 3 and S10,† the CV of the complex 1/graphene composite-modified GCE (1/graphene-GCE) shows an irreversible reduction peak at −0.66 V vs. SCE at 0.005 V s⁻¹, which is similar to that (−0.64 V vs. SCE) observed at 1-GCE (Fig. 3 and S5†). However, the current density for this reduction wave at 1/graphene-GCE is a bit larger than that at 1-GCE (Fig. 3), indicating the conductivity of the 1/graphene composite material is a bit improved relative to the complex with the addition of graphene. The potentials for the HER and OER at 1/graphene-GCE are −1.30 and +1.28 V vs. SCE, respectively. And the HER potential at 1/graphene-GCE (−1.30 V vs. SCE) is a bit positively shifted relative to the HER potential at 1-GCE (−1.39 V vs. SCE), indicating the addition of the graphene into complex 1 can promote the HER to some extents. The phenomenon is more obviously observed at the complex 2/graphene composite-modified GCE (2/graphene-GCE).

As shown in Fig. 4 and S11,† the CV of 2/graphene-GCE at a scan rate of 0.005 V s⁻¹ reveals that the HER and OER potentials are −0.80 and +1.80 V vs. SCE, respectively. The HER potential at 2/graphene-GCE (−0.80 V vs. SCE) is remarkably positively shifted in comparison to the HER potential at 2-GCE (−1.30 V vs. SCE), indicating 2/graphene composite possesses much improved electrocatalytic activity for the HER than complex 2, which is probably due to the improved conductivity of the composite material with the addition of graphene into complex 2.¹⁶ And the expectation is convinced by the electrochemical impedance spectroscopy (EIS) at −1.0 V vs. SCE. In EIS, the radius of the arc on the Nyquist plot reflects the charge-transfer impedance and the reaction rate on the surface of the electrode.¹⁷ As shown in Fig. S12,† the radius of the arc in the Nyquist plot of 2/graphene-GCE is much smaller than that of 2-GCE, indicating with the addition of graphene into complex 2, the charge-transfer impedance is much decreased.

The UV-Vis absorption and diffuse reflectance spectra for the compounds

The UV-Vis absorption spectra of the free organic ligands, complexes 1–2 together with their graphene composite materials in the solid state at room temperature are shown in Fig. S13.† As shown in Fig. S13,†L1 exhibits absorption peak at 305 nm in the range of 240–650 nm, and L2 displays a strong absorption peak at 285 nm in the similar range, which may be ascribed to the n–π* or π–π* transition.¹⁸ Complex 1 displays one strong absorption peak at ca. 270 nm and one shoulder peak at 320 nm in the range of 240–550 nm. In the similar range, complex 2 exhibits absorption peak at ca. 295 nm. The absorption peaks for complexes 1 and 2 are different from those of L1 and L2, indicating they may be ascribed to the intraligand transition (ILCT) or metal-to-ligand charge-transfer transition (MLCT).¹⁸ And complex 1 shows absorption peaks at 655 and 702 nm in the range of 550–1000 nm (Fig. S13†), which are assigned to the visible d–d transition.

The UV-Vis absorption spectra of the graphene composite materials are a bit different from those of their corresponding complex. As shown in Fig. S13,† 1/graphene composite material shows absorption peaks at 320, 655 and 710 nm, and 2/graphene composite material displays absorption peaks at 270, 650 and 735 nm. The two graphene composite materials both exhibit absorption in the visible range of 500–1000 nm, which is because the black graphene can absorb the visible light.

The diffuse reflectance spectra (DRS) for complexes 1 and 2 are presented in Fig. S14.† According to the DRS analysis and using the Kubelka–Munk equation, F(R) = α = (1 − R)²/2R, where R is the percentage of reflectance and α is absorption coefficient,¹⁹ the band gap was calculated. The transformed Kubelka–Munk function, [F(R)hν]^1/2 = A(hν − E_g), relates the incident photon energy (hν) and the optical band gap energy (E_g), where A is the constant. From the plot of [F(R)hv]^1/2 versus hν, by extrapolating the linear fitted regions to [F(R)hv]^1/2 = 0, the band gap energy for complexes 1 and 2 was evaluated as 2.3 and 1.8 eV, respectively (Fig. S14†).

Band gaps and density of states (DOS) for complexes 1 and 2

In order to further verify the experimental result, the band gaps and density of states (DOS) for the two complexes are calculated using the Materials Studio 7.0. In an attempt to simplify the calculation, the uncoordinated water molecules are removed from the single crystal structural data of complex 2.

As shown in Fig. 5, the two complexes are both semiconductors, and the gaps between the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO) are calculated as 1.722 eV for complex 1 (Fig. 5a) and 1.268 eV for the dehydrated complex 2 (Fig. 5b), which are smaller than the experimental results (2.3 eV for complex 1 and 1.8 eV for complex 2) by ca. 0.6 eV. The tendency for the calculation to underestimate the band gap energy has been well documented in previous studies,¹⁹ which is an intrinsic feature of the methods based on DFT, namely not taking into account the discontinuity in the exchange–correlation potential.²⁰ However, the order of the calculated band gaps for the two complexes shows excellent agreement with the experimentally observed sequence, indicating the theoretical calculation is relatively accurate.

Fig. 5 Band gaps of complex 1 (a) and the dehydrated complex 2 (b).

The band structures can be assigned according to total (TDOS) and partial density of states (PDOS). As shown in Fig. 6a and b, the bottom of conduction bands (CBs) of both complex 1 and the dehydrated complex 2 are mainly made up of C 2p, N 2p, O 2p and Mo 4d states. However, the top of valence band (VB) of complex 1 is mostly formed from C 2p and N 2p states (Fig. 6a), while the VB maximum of the dehydrated complex 2 are mainly composed of C 2p, N 2p and O 2p states (Fig. 6b), indicating the O 2p state occupies the VB maximum of the dehydrated complex 2, but it hasn't contribution to the VB maximum of complex 1. The different TDOS and PDOS of complexes 1 and 2 are probably attributed to their different compositions and frameworks, thus giving rise to their different band gaps.

Fig. 6 Total (TDOS) and partial density of states (PDOS) as obtained from CASTEP calculations for complex 1 (a) and the dehydrated complex 2 (b). In the PDOS, blue, red and green lines represent s, p and d orbits, respectively.

Photoelectrochemical behaviors of the complexes and their graphene composites

As described above, both complexes 1 and 2 can be regarded as semiconductors, herein, their photoelectrochemical behaviors have been investigated. In comparison, the photocurrent responses of the free ligands were also conducted under the irradiation of the visible light (650 nm > λ > 350 nm). All the measurements were at E = 0 V vs. Ag/AgCl in the buffer solution (CH₃COOH–CH₃COONa, pH = 4.5, 50 mL) with a Pt foil and a Ag/AgCl electrode used as the counter and the reference electrodes, respectively.

As shown in Fig. 7, S15 and S16,† it is clearly observed that both complex 1-modified FTO and 2-modified FTO electrodes can exhibit obvious photocurrent responses under the visible light irradiation (650 nm > λ > 350 nm). In contrast, the photocurrent produced on the free ligand L1 almost can be neglected (Fig. 7). And the POM-based complex 2 also yields more remarkable photoresponse than the free ligand L2 (Fig. 7 and S16†). Furthermore, it is found complex 2 shows a stronger photocurrent response (4.5 × 10⁻⁵ mA cm⁻²) than complex 1 (2 × 10⁻⁵ mA cm⁻²) (Fig. 7), which is in agreement with the band gaps of the two complexes. As described above, complex 2 possesses more narrow band gap (1.8 eV) than complex 1 (2.3 eV), then it is more easily for the electrons from the top of VB in complex 2 to be excited into the bottom of CB under illumination with respect to those in complex 1.

Fig. 7 Photocurrent spectra of the free ligands, graphene, complexes 1–2 and their graphene composite materials under the visible light (650 nm > λ > 350 nm) at 110 mW cm⁻².

To study the charge separation and transfer process in detail, electrochemical impedance spectroscopy (EIS) was employed. The radius of the arc in the Nyquist plot of complex 1 or 2 with light irradiation (650 nm > λ > 350 nm) is obviously smaller than that without irradiation, suggesting that the charge migration and transfer is more fast under the light illumination than without illumination (Fig. S17†). Furthermore, the arc radius of complex 2-modified FTO electrode is smaller than that of 1-modified FTO electrode under the visible light (Fig. S17†), which indicates a more effective separation of the photogenerated electron–hole pair, as well as a faster interfacial charge transfer in complex 2 than in complex 1.

The photoelectrochemical behaviors of the graphene composite materials of the two complexes have also been measured under the similar condition. However, the complex/graphene composite material shows almost similar photocurrent response with respect to its corresponding complex, as shown in Fig. 7, S15 and S16.† As we know, graphene is electroconductive with zero of band gap, then the pure graphene can't yield photocurrent under the visible light irradiation (Fig. 7), and the charge transfer impedances on the pure graphene with and without light irradiation are almost the same (Fig. S18†). With the addition of graphene into complex 1 or 2, the electron conductivity of the complex/graphene composite material is much improved. And it has been proved by the Nyquist plots at 0 V vs. SCE, in which the arc radius of the complex/graphene composite is much smaller than that of the corresponding complex (Fig. S17†), indicating the charge transfer impedance of the composite material is much decreased with respect to the pure complex. However, the addition of graphene can decrease the charge transfer impedance of the composite material with and without the visible light irradiation (650 nm > λ > 350 nm) almost to the same extent, then the photoresponse of the complex/graphene composite is not obviously improved with respect to its corresponding complex. On the other hand, the phenomenon is probably also associated with the fact that no strong interactions exist between the complex and the graphene in the composite materials, as shown in their SEM images (Fig. S9†). The detailed mechanism is under way.

Thermal stabilities of complexes 1 and 2

In order to examine the thermal stabilities of the two complexes, thermogravimetric analyses (TGAs) were carried out. The sample was heated up to 750 °C in N₂. As shown in Fig. S19,† no weight loss is observed in the temperature range of 30–340 °C in the TGA curve of complex 1, which is in good agreement with the crystal structure of complex 1, in which no solvent is included. Decomposition of the organic component happened in the temperature range 340–630 °C with a loss of 57.5 wt% (calc. 57.9 wt%). And when the temperature was higher than 630 °C, complex 1 exhibited a weight gain, which might correspond to the phase change of the molybdenum oxide. As the blue curve shown in Fig. S19,† complex 2 released its uncoordinated water molecules in the range of 15–110 °C with a loss of 6.3 wt% (calc. 5.9 wt%). The anhydrous sample remained stable up to ∼280 °C without any weight loss. And the decomposition of the organic ligand began at 280 °C.

Conclusions

In conclusion, two POM-based complexes have been hydrothermally synthesized at the pH value of 4.5, and they are structurally characterized by single-crystal X-ray diffraction. In complex 1, 1D [Mo^V₄Mo^VI₄O₂₆]_∞⁸⁻ chain is constructed by vertice-sharing {MoO₆} octahedra, and the uncoordinated protonated [H₂L1]²⁺ species link different [Mo₈O₂₆]_∞⁸⁻ chains via supramolecular forces. In complex 2, the [Mo₈O₂₅]²⁻ moieties are linked by the dangling μ₃-[HL2]⁺ and μ₂-L2 ligands into 1D chain. Complex 2 possesses more narrow band gap than complex 1, and it shows stronger photocurrent response than complex 1 under the visible light (650 nm > λ > 350 nm). With the addition of graphene into complex 1 or 2, the electroconductivity of the complex/graphene composite material is much improved. And 2/graphene composite material shows much improved electrocatalytic activity for the HER than complex 2. However, the photocurrent response of the graphene composite material is not obviously enhanced with respect to its corresponding complex, which is probably because the addition of graphene can decrease the charge transfer impedance of the complex/graphene composite with and without the visible light irradiation almost to the same extent. And it is probably also associated with the fact that no strong interactions exist between the complex and the graphene in the composite material. The detailed mechanism is under way.

Acknowledgements

Financial supports from the National Natural Science Foundation of China (No. 21371184), the large-scale instrument and equipment open foundation in Chongqing University (No. 201606150053), and Chongqing Key Laboratory of Chemical Process for Clean Energy and Resource Utilization are gratefully acknowledged.

References

1. (a) O. M. Yaghi, M. O'Keeffe, N. W. Ockwig, H. K. Chae, M. Eddaoudi and J. Kim, Nature, 2003, 423, 705; (b) T. Uemura, N. Yanai and S. Kitagawa, Chem. Soc. Rev., 2009, 38, 1228; (c) P. Q. Liao, D. D. Zhou, A. X. Zhu, L. Jiang, R. B. Lin, J. P. Zhang and X. M. Chen, J. Am. Chem. Soc., 2012, 134, 17380.
2. (a) M. D. Allendorf, A. Schwartzberg, V. Stavila and A. A. Talin, Chem.–Eur. J., 2011, 17, 11372; (b) C. G. Silva, A. Corma and H. Garcia, J. Mater. Chem., 2010, 20, 3141; (c) L. M. Yang, P. Vajeeston, P. Ravindran, H. Fjellvag and M. Tilset, Inorg. Chem., 2010, 49, 10283; (d) F. X. Llabres i Xamena, A. Corma and H. Garcia, J. Phys. Chem. C, 2007, 111, 80.
3. (a) Y. Takahashi, R. Obara, K. Nakagawa, M. Nakano, J. Y. Tokita and T. Inabe, Chem. Mater., 2007, 19, 6312; (b) D. L. Turner, T. P. Vaid, P. W. Stephens, K. H. Stone, A. G. Dipasquale and A. L. Rheingold, J. Am. Chem. Soc., 2008, 130, 14; (c) D. Y. Lee, D. V. Shinde, S. J. Yoon, K. N. Cho, W. Lee, N. K. Shrestha and S. H. Han, J. Phys. Chem. C, 2013, 118, 16328.
4. (a) A. A. Talin, A. Centrone, A. C. Ford, M. E. Foster, V. Stavila, P. Haney, R. A. Kinney, V. Szalai, F. El Gabaly, H. P. Yoon, F. Léonard and M. D. Allendorf, Science, 2014, 343, 66; (b) W. Ki and J. Li, J. Am. Chem. Soc., 2008, 130, 8114; (c) Y. Li, A. Pang, C. Wang and M. Wei, J. Mater. Chem., 2011, 21, 17259.
5. (a) X. X. Yan, X. D. Qiu, Z. S. Yan, H. J. Li, Y. Gong and J. H. Lin, J. Solid State Chem., 2016, 237, 313; (b) P. Zhang, Y. Gong and J. H. Lin, Eur. J. Inorg. Chem., 2016, 3, 322; (c) P. G. Jiang, P. Zhang, Y. Gong and J. H. Lin, Dalton Trans., 2016, 45, 4603.
6. (a) R. M. Yu, X. F. Kuang, X. Y. Wu, C. Z. Lu and J. P. Donahue, Coord. Chem. Rev., 2009, 253, 2872; (b) H. N. Miras, J. Yan, D. L. Long and L. Cronin, Chem. Soc. Rev., 2012, 41, 7403.
7. (a) A. Dolbecq, E. Dumas, C. R. Mayer and P. Mialane, Chem. Rev., 2010, 110, 6009; (b) T. Tachikawa, S. Tojo, M. Fujitsuka and T. Majima, Chem.–Eur. J., 2006, 12, 3124; (c) D. Karimian, B. Yadollahi and V. Mirkhani, Dalton Trans., 2015, 44, 1709.
8. (a) H. Park and W. Choi, J. Phys. Chem. B, 2003, 107, 3885; (b) T. T. Lu, X. X. Xu, H. L. Li, Z. Y. Li, X. Zhang, J. Z. Ou and M. L. Mei, Dalton Trans., 2015, 44, 2267; (c) S. S. Xu, W. L. Chen, Y. H. Wang, Y. G. Li, Z. J. Liu, C. H. Shan, Z. M. Su and E. B. Wang, Dalton Trans., 2015, 44, 18553.
9. (a) M. Vasilopoulou, A. M. Douvas, L. C. Palilis, S. Kennou and P. Argitis, J. Am. Chem. Soc., 2015, 137, 6844; (b) J. S. Li, X. J. Sang, W. L. Chen, L. C. Zhang, Z. M. Zhu, T. Y. Ma, Z. M. Su and E. B. Wang, ACS Appl. Mater. Interfaces, 2015, 7, 13714; (c) S. Roy, S. Sarkar, J. Pan, U. V. Waghmare, R. Dhanya, C. Narayana and S. C. Peter, Inorg. Chem., 2016, 55, 3364.
10. (a) B. Nohra, H. E. Moll, L. M. R. Albelo, P. Mialane, J. Marrot, C. M. Draznieks, M. O'Keeffe, R. N. Biboum, J. Lemaire, B. Keita, L. Nadjo and A. Dolbecq, J. Am. Chem. Soc., 2011, 133, 13363; (b) L. Yang, S. Kinoshita, T. Yamada, S. Kanda, H. Kitagawa, M. Tokunaga, T. Ishimoto, T. Ogura, R. Nagumo, A. Miyamoto and M. Koyama, Angew. Chem., Int. Ed., 2010, 49, 5348; (c) X. L. Wang, N. Han, H. Y. Lin, A. X. Tian, G. C. Liu and J. W. Zhang, Dalton Trans., 2014, 43, 2052.
11. (a) V. Niraimathi, V. Vaidhyalingam, A. Aruna and A. Vadivelu, Acta Cienc. Indica, Chem., 2009, 35, 43; (b) Y. Gong, P. Zhang, P. G. Jiang and J. H. Lin, CrystEngComm, 2015, 17, 1056.
12. (a) G. M. Sheldrick, SHELXS 97, Program for Crystal Structure Solution, University of Göttingen, Göttingen, Germany, 1997; (b) G. M. Sheldrick, SHELXL 97, Program for Crystal Structure Refinement, University of Göttingen, Göttingen, Germany, 1997.
13. (a) R. Xi, B. Wang, K. Isobe, T. Nishioka, K. Toriumi and Y. Ozawa, Inorg. Chem., 1994, 33, 833; (b) L. D. Chen, Y. H. Wang, C. W. Hu, L. Y. Feng, E. B. Wang, N. H. Hu and H. Q. Jia, J. Solid State Chem., 2001, 161, 173.
14. N. E. Brese and M. O'keeffe, Acta Crystallogr., Sect. B: Struct. Sci., 1991, 47, 192.
15. (a) H. Liu, S. Ryu, Z. Y. Chen, M. L. Steigerwald, C. Nuckolls and L. E. Brus, J. Am. Chem. Soc., 2009, 131, 17099; (b) E. Pollak, B. S. Geng, K. J. Jeon, I. T. Lucas, T. J. Richardson, F. Wang and R. Kostecki, Nano Lett., 2010, 10, 3386; (c) X. L. Gao, Y. Gong, P. Zhang, Y. X. Yang, J. P. Meng, M. M. Zhang, J. L. Yin and J. H. Lin, CrystEngComm, 2014, 16, 8492.
16. X. L. Gao, Y. Gong, P. Zhang, Y. X. Yang, J. P. Meng, M. M. Zhang, J. L. Yin and J. H. Lin, CrystEngComm, 2014, 16, 8492.
17. (a) H. Liu, S. A. Cheng, M. Wu, H. J. Wu, J. Q. Zhang, W. Z. Li and C. N. Cao, J. Phys. Chem. A, 2000, 104, 7016; (b) W. H. Leng, Z. Zhang, J. Q. Zhang and C. N. Cao, J. Phys. Chem. B, 2005, 109, 15008; (c) X. X. Xu, X. Gao, T. T. Lu, X. X. Liu and X. L. Wang, J. Mater. Chem. A, 2015, 3, 198.
18. (a) S. Ohkoshi, H. Tokoro, T. Hozumi, Y. Zhang, K. Hashimoto, C. Mathonière, I. Bord, G. Rombaut, M. Verelst, C. C. Moulin and F. Villain, J. Am. Chem. Soc., 2006, 128, 270; (b) M. Stadler, F. Puntoriero, S. Campagna, N. Kyritsakas, R. Welter and J. M. Lehn, Chem.–Eur. J., 2005, 11, 3997; (c) J. H. Wang, Y. Q. Fang, L. Bourget-Merle, M. I. J. Polson, G. S. Hanan, A. Juris, F. Loiseau and S. Campagna, Chem.–Eur. J., 2006, 12, 8539.
19. L. L. Wang, H. M. Noh, B. K. Moon, S. H. Park, K. H. Kim, J. S. Shi and J. H. Jeong, J. Phys. Chem. C, 2015, 119, 15517.
20. (a) H. W. Eng, P. W. Barnes, B. M. Auer and P. M. Woodward, J. Solid State Chem., 2003, 175, 94; (b) H. Mizoguchi, H. W. Eng and P. M. Woodward, Inorg. Chem., 2004, 43, 1667; (c) M. W. Stoltzfus, P. M. Woodward, R. Seshadri, J. H. Klepeis and B. Bursten, Inorg. Chem., 2007, 46, 3839; (d) J. P. Perdew and M. Levy, Phys. Rev. Lett., 1983, 51, 1884; (e) L. M. Yang, P. Ravindran, P. Vajeeston and M. Tilset, Phys. Chem. Chem. Phys., 2012, 14, 4713.

---

Footnote

† Electronic supplementary information (ESI) available: Crystallographic data; PXRD patterns; images of the electrodes; CVs; Raman spectrum; SEM images; UV-Vis absorption and diffuse reflectance spectra; Nyquist plots; TG curve and other supplementary material are included. CCDC 1032252 and 1032253. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra19805h

---