Syntheses, Structures, and Properties of Coordination Polymers with 2,5-Dihydroxy-1,4-Benzoquinone and 4,4'-Bipyridyl Synthesized by In Situ Hydrolysis Method.

The coordination polymers (CPs) with binary ligands, including 2,5-dihydroxy-1,4-benzoquinone (H2DHBQ) and 4,4'-bipyridyl (bpy), were synthesized using in situ hydrolysis of 2,5-dimethoxy-1,4-benzoquinone (DMBQ). Three kinds of CPs were obtained depending on the metal ions. For M = Mn and Zn, a 1D zigzag chain structure with cis conformation ( cis-1D-M) was obtained, whereas Co, Ni, and Cu compounds afforded a 2D net structure with trans conformation (trans -2D-M) with a 1D pore. A linear chain structure was also obtained for M = Cu. Magnetic susceptibility (χM T) at 300 K in cis -1D-Mn and trans -2D-Co was evaluated to be 4.421 and 2.950 cm3 K mol-1, respectively, indicating that both compounds are in the high-spin state. According to the N2 adsorption isotherms at 77 K, trans -2D-Ni showed microporosity with the BET surface area of 177 m2 g-1, whereas the isomorphic trans -2D-Co rarely adsorbed N2 at 77 K. This phenomenon was explained by the difference of diffusion kinetics of the adsorbent molecules, which was supported by the CO2 adsorption isotherms at 195 K. The optical band gaps of cis -1D-Mn, cis -1D-Zn, trans -2D-Co, and trans -2D-Ni were estimated to be 1.6, 1.8, 1.0, and 1.1 eV, respectively, by using UV-vis-NIR spectroscopy.

Introduction

Coordination polymers (CPs) with a permanent porosity, including metal–organic frameworks (MOFs), have attracted attention as molecular porous materials because of their potential applications for gas storage,[1−7] gas separation,[8−11] sensing,[12] catalysis,[13−15] etc. In particular, redox-active CPs have been gaining a lot of interest because of their marked electronic properties. 2,5-Dihydroxy-1,4-benzoquinone (H2DHBQ, R = H in Figure ) and its derivatives are widely used as a redox-active ligand for CP syntheses, and remarkable magnetic[16−18] and conductive[19−21] properties have been reported. Though various structural motifs such as 1D chain,[22] 2D honeycomb sheet,[16−21] and 3D (10,3)-a net[23] have been reported using DHBQ and its derivatives, more structural diversity is required for further study of these CPs. One useful approach to afford structural diversity is to introduce auxiliary ligands. To date, no reports of DHBQ-based CPs have been reported with binary ligands, although some reports exist for chloranilic acid (R = Cl).[24−26] A plausible reason for this is due to the fast complexation of DHBQ ligand, which makes it difficult to obtain the phase-pure product or to obtain single crystals. In situ generation of the active ligand is a useful way to control the speed of crystallization. Abrahams and co-workers reported several single crystals of DHBQ-based CPs by using in situ generation of DHBQ by hydrolysis of 2,5-diamino-1,4-benzoquinone (DABQ).[23] Recently, such an in situ generation of the active ligand has also been used for introduction of functional groups in the CPs.[27,28] Herein, we utilized the in situ synthesis for introduction of the auxiliary ligands by using 2,5-dimethoxy-1,4-benzoquinone (DMBQ, Figure ) as a precursor of DHBQ. Using this method, we obtained three kinds of the structural motifs of binary CPs with DHBQ and 4,4′-bipyridyl (bpy), which are a 1D zigzag chain of cis-[M(DHBQ)(bpy)2] for M = Mn and Zn (-1D-M), a 3D porous framework composed of an interlocked 2D rectangular net of trans-[M(DHBQ)(bpy)]·n(H2O) for M = Co, Cu, and Ni (-2D-M), and a 1D linear chain of trans-[M(DHBQ)(bpy)2] for M = Cu (-1D-M). In this paper, we report their structure, thermogravimetry, spectroscopic properties, and porosity.

Figure 1

Molecular structures of (a) H2DHBQ derivatives and (b) DMBQ.

Experimental Section

Characterization and Instrumental Procedures

Mn(OAc)2·4H2O, Co(OAc)2·4H2O, Ni(OAc)2·4H2O, Cu(OAc)2·H2O, and Zn(OAc)2·2H2O were purchased from Fujifilm Wako Chemicals and used as received. 2,5-Dihydroxy-1,4-benzoquionone (H2DHBQ) was purchased from Acros Organics and used as received. 4,4′-Bipyridyl (bpy) was purchased from TCI chemicals and used as received.

Solid-state IR spectra from 600 to 4000 cm–1 were collected on a JASCO FT/IR 4200 Fourier transform infrared spectrometer by means of an attenuated total reflection (ATR) method with 4 cm–1 resolution. Solid-state UV–vis–NIR spectra from a diffuse reflection method were recorded on a Shimadzu UV-3100. Samples were diluted with BaSO4, and obtained diffuse reflectivity spectra were converted to absorption spectra using the Kubelka–Munk function. 1H NMR measurements were performed on a Bruker AV500 at RT. Thermogravimetry (TG) was measured on a SHIMADZU DTG-60/60H at a heating rate of 5 °C/min under a constant nitrogen flow (0.1 L/min). The N2, H2, and CO2 sorption isotherms were measured using MicrotracBEL BELSORP MAX. Density functional theory (DFT) calculations were performed using Gaussian16 software.[29] All geometry optimizations were performed at the UB3LYP/6-31G+(d,p) level of theory.[30−32] Elemental analyses were performed using the J-Science Lab JM-10 equipped in the Research and Analytical Center for Giant Molecules at Tohoku University. Single crystal X-ray crystal structure analyses were performed using a Rigaku Varimax diffractometer with a Mo Kα radiation source. A portion of the single crystals for the crystal structure analysis was directly pipetted from the reaction vials and transferred to a Petri dish. Appropriate single crystals were selected using an optical microscope and then scooped with a MiTeGen crystal mount. The data were collected with an ω-scan strategy under several φ-angles. Data reduction was performed with the CrysAlisPro package, and an empirical absorption correction was applied. The structures were solved by direct methods and refined by full-matrix least-squares on F2 with anisotropic displacement using the SHELXTL software package.[33] The non-H atoms were treated anisotropically, whereas the hydrogen atoms were placed in calculated, ideal positions and refined as riding on their respective carbon atoms. The electron density in the pore was assigned to oxygen atoms of the water molecules, and their site occupancies were optimized to minimize the R-value. Powder X-ray diffraction (PXRD) patterns for capillary-encapsulated samples were obtained at room temperature with a step of 0.02° using a Rigaku SmartLab diffractometer with Cu Kα radiation source. Magnetic susceptibility at 300 K was measured using Quantum Design MPMS-XL. Scanning electron microscope (SEM) images were acquired using the Hitachi S-4300.

Syntheses

2,5-Dimethoxy-1,4-Benzoquinone (DMBQ)

DMBQ was prepared following a procedure published earlier.[34] 2,5-Dihydroxy-1,4-benzoquinone (3.3 g, 24 mmol) was stirred in MeOH (850 mL) with 5 mL of 38% HCl (w/w) aq overnight at room temperature. The precipitate was filtered and then purified by recrystallization with N,N-dimethylformamide to obtain a goldish yellow crystalline solid. This crystalline solid was washed with MeOH until the filtrate became colorless and then dried in vacuo to afford 3.2 g (81%) of DMBQ. 1H NMR (500 MHz CDCl3) δ 5.87 (s, 2H), δ 3.85 (s, 6H). Found: C, 57.17; H, 4.86; N, 0.00. Calc. for C8H8O4: C, 57.18; H, 4.80; N, 0.00.

cis-Mn(DHBQ)(bpy)2 ()

Mn(OAc)2·4H2O (245 mg, 1.0 mmol), finely milled DMBQ (168 mg, 1.0 mmol), and bpy (304 mg, 2.0 mmol) were suspended in 25 mL of deionized water in a 30 mL screw-capped glass vial with a Teflon-coated rubber seal. The suspension was sonicated for several minutes. The vial was put in an electric oven and heated to 120 °C for 24 h. After cooling, the precipitate was collected by filtration. The solid was washed with water, then washed with acetone, and then dried in vacuo to obtain a deep brown solid. Yield: 340 mg (67%). Elemental analysis calc. for C26H18MnN4O4: C, 61.78%; H, 3.60%; N, 11.09%. Found: C, 61.81%; H, 3.77%; N, 10.83%.

cis-Zn(DHBQ)(bpy)2 ()

Zn(OAc)2·2H2O (220 mg, 1.0 mmol), finely milled DMBQ (168 mg, 1.0 mmol), and bpy (304 mg, 2.0 mmol) were suspended in 25 mL of deionized water in a 30 mL screw-capped glass vial with a Teflon-coated rubber seal. The suspension was sonicated for several minutes. The vial was put in an electric oven and heated to 120 °C for 24 h. After cooling, the precipitate was collected by filtration. The solid was washed with water, then washed with acetone, and then dried in vacuo to obtain a pink-red solid. Yield: 495 mg (96%). Elemental analysis calc. for C26H18N4O4Zn: C, 60.53%; H, 3.52%; N, 10.86%. Found: C, 59.87%; H, 3.70%; N, 10.87%.

trans-Co(DHBQ)(bpy)·2H2O ()

Co(OAc)2·4H2O (162 mg, 0.65 mmol), finely milled DMBQ (168 mg, 1.0 mmol), and bpy (152 mg, 1.0 mmol) were suspended in 25 mL of deionized water in a 30 mL screw-capped glass vial with a Teflon-coated rubber seal. The suspension was sonicated for several minutes. The vial was put in an electric oven and heated to 120 °C for 24 h. After cooling, the precipitate was collected by filtration. The solid was washed with water, then washed with acetone, and then dried in vacuo to obtain a deep brown solid. Yield: 250 mg (99%). Elemental analysis calc. for C16H14CoN2O6: C, 49.37%; H, 3.63%; N, 7.20. Found: C, 49.48%; H, 3.59%; N, 7.08%.

trans-Ni(DHBQ)(bpy)·2.25H2O ()

Ni(OAc)2·4H2O (249 mg, 1.0 mmol), finely milled DMBQ (168 mg, 1.0 mmol), and bpy (152 mg, 1.0 mmol) were suspended in 25 mL of deionized water in a 30 mL screw-capped glass vial with a Teflon-coated rubber seal. The suspension was sonicated for several minutes. The vial was put in an electric oven and heated to 120 °C for 24 h. After cooling, the precipitate was collected by filtration. The solid was washed with water, then washed with acetone, and then dried in vacuo to obtain a red-brown solid. Yield: 375 mg (95%). Elemental analysis calc. for C16H14.50NiN2O6.25: C, 48.8%; H, 3.71%; N, 7.12%. Found: C, 48.99%; H, 3.56%; N, 7.05%.

Results and Discussion

We synthesized CPs with the existence of metal acetate, DMBQ, and 4,4′-bipyridyl (bpy) under hydrothermal conditions. We obtained several CPs including DHBQ and bpy ligands. When we used manganese and zinc acetate as a metal source, we obtained 1D zigzag chains of cis-[M(DHBQ)(bpy)2] (abbreviated as -1D-M). -1D-Mn and -1D-Zn were isomorphic. The crystal structure of -1D-Mn is shown in Figure , and crystallographic parameters are summarized in Table S1. The molar ratio of Mn, DHBQ, and bpy in this compound is 1:1:2. Two DHBQ2– ligands are in cis position and each of them bridges the neighboring Mn2+ ions in a bis-bidentate manner, forming a 1D zigzag chain. Two bpy ligands coordinate to the Mn2+ in cis position. One nitrogen atom of the bpy ligand coordinates to Mn2+, whereas the other does not interact with any of the neighboring atoms. The distances of Mn–N1, Mn–O1, and Mn–O2 are 2.2487(13), 2.1670(11), and 2.1990(15) Å, respectively. The magnetic susceptibility of -1D-Mn (χMT = 4.421 cm3 K mol–1 at 300 K, namely μeff = 5.94 μB) was similar to the S = 5/2 spin-only value (4.38 cm3 K mol–1) and that of the representative Mn2+ high spin complexes,[35,36] showing that -1D-Mn is in high-spin state. There is a π–π interaction between DHBQ2– and the neighboring bpy (inter plane distance is 3.454 Å). It should be noted that when we mixed metal acetate, DMBQ, and 4,4′-bipyridyl (bpy) with a 1:1:1 ratio, we obtained the mixture of -1D-M and one-dimensional CP without bpy ligands (M(DHBQ)(H2O)2).[22]

Figure 2

Crystal structure of -1D-Mn. Hydrogen atoms are omitted for clarity. (a) ORTEP drawing of the local environment. (b,c) Two perspective views. DHBQ and bpy are shown in blue and red, respectively.

By mixing cobalt or nickel acetate, DMBQ, and bpy, 2D rectangular nets of trans-[M(DHBQ)(bpy)] (-2D-M) were obtained. The crystal structure of the Co complex is shown in Figure . The crystallographic parameters are summarized in Table S2. The molar ratio of Co, DHBQ, and bpy is 1:1:1, and the DHBQ ligands are in the trans position to form a 1D linear chain (···Co–DHBQ–Co···) along the c-axis. In addition, bpy bridges the adjacent 1D chains (···Co–bpy–Co···) along the a + b and a – b directions to form two independent (4,4) rectangular 2D nets, respectively (Figure b). Each 2D net interlocked with each other—more specifically, interlocked with parallel–parallel (p–p) mode[37]—to form a polycatenated 3D framework (Figure c). There are small 1D pores with ca. 4.2 Å diameter space between two 2D nets (Figure S8). The electron density in the pore was assigned to approximately 2.3 water molecules per formula unit, which is consistent with the elemental analysis. The distances of Co–N1 = 2.142(4) Å and Co–N2* = 2.136(4) Å are longer than that of the four Co–O bond distances [2.057(4), 2.058(4), 2.065(4), and 2.077(4) Å]. Magnetic susceptibility (χMT) at 300 K of -2D-Co was 2.950 cm3 K mol–1, namely μeff = 4.86 μB, which is larger than that expected for the spin-only (1.875 cm3 K mol–1 for S = 3/2 and g = 2.0) but in agreement with the values observed for the octahedral high spin Co2+ complexes with a significant first-order orbital contribution.[38] In the case of M = Ni, we concluded that -2D-Ni is isomorphic to -2D-Co according to the PXRD results (Figure ).

Figure 3

Structure of -2D-Co. The water atoms are omitted for clarity. (a) ORTEP drawing of the local structure. Perspective view of the (b) 2D net and (c) packing motif with inclined polycatenation.

Figure 4

Powder X-ray diffraction (PXRD) patterns. (a) -1D-Zn and -1D-Mn together with a simulated pattern of -1D-Zn and -1D-Mn. (b) -2D-Co and -2D-Co together with a simulated pattern of -2D-Co.

We performed the stability test against the solvents by soaking -2D-M in solvents (H2O, MeOH, EtOH, DMF, CH3CN, and pyridine) for 48 h at room temperature. The PXRD patterns of the compounds before and after soaking are shown in Figure S9. As for the -2D-Ni, the PXRD patterns did not change with soaking for any of the aforementioned solvents, indicating that -2D-Ni has high stability against the solvents and a rigid structure. In the case of -2D-Co, conversely, peak broadening or peak splitting was observed with soaking in DMF, CH3CN, or pyridine, while there was no change in H2O, MeOH, and EtOH. It should be noted that the PXRD patterns returned to original after standing in air for 3 days. Therefore, this change in the PXRD patterns is not due to the collapse of the structure but due to the subtle structural change of the framework by introducing the solvent into the pore. These findings indicate that the framework of the -2D-Co can be more flexible than -2D-Ni, though they are isostructural.

When we used copper acetate as a metal source, the obtained crystalline solid contained two kinds of crystal that could not be visually separated. One is mostly similar with that of -2D-Co (Figure S2 and Table S2), thus we named it -2D-Cu. The Cu–N distance is 2.0040(17) Å, whereas the Cu–O1 and Cu–O2 distances are 2.1595(17) and 2.0997(18) Å, respectively, indicating that the O1–Cu–O1* axis corresponds to the Jahn–Teller axis.[35] The other crystal structure was found only for M = Cu (Figure ). The molar ratio of Cu, DHBQ, and bpy is 1:1:2, and the DHBQ ligands are located at the trans position to form a 1D linear chain. One side of the nitrogen atom of the bpy ligand coordinates to Cu2+, whereas another side was free. Therefore, the complex is named trans-[Cu(DHBQ)(bpy)2] (-1D-Cu). The Cu–O2 [2.235(3) Å] and Cu–O4 [2.247(3) Å] distances are significantly longer than the Cu–O1 [1.997(3) Å], Cu–O3 [1.996(3) Å], Cu–N1 [2.037(4) Å], and Cu–N2 [2.034(4) Å] distances, indicating that the O2–Cu–O4 axis is the Jahn–Teller axis.[39] The shortest C–C distance between the neighboring pyridyl ring is 3.8 Å, indicating that there is a weak π–π interaction between pyridyl rings. The PXRD pattern of the Cu compound (Figure S3) was explained by the superposition of -2D-Cu and -1D-Cu.

Figure 5

Structure of -1D-Cu. (a) ORTEP drawing of the local structure. Two perspective views along (b) a- and (c) c-axes.

The coordination conformation and network motifs of the crystals in relation to the metal ions is summarized in Table . Here, we discuss the coordination conformation in relation to the metal ions. In the case of MnII and ZnII, cis conformation was formed, whereas CoII, NiII, and CuII afforded trans conformation. In order to estimate the thermodynamic stability, we performed DFT calculations. The discrete neutral complexes, cis- or trans-[M(HDHBQ)(py)2] (py = pyridyl), were used as model compounds to simplify the calculations. The hydration effect is considered with regards to the integral equation formalism of the polarizable continuum model (IEFPCM). The ΔG (≡ Gcis – Gtrans) is shown in Table . Our result indicates that the cis conformation is more stable, except for M = Cu. This is not consistent with the experimental result. So, we presumed that the kinetic factor plays a role for the preference. According to Shriver & Atkins’s textbook, the ligand-substitution rate of the aqua complexes is in the order of CuII > ZnII > MnII > CoII > NiII,[40] which means CuII has more chance to relax to the thermodynamically stable structure before crystallization than NiII does. Meanwhile, a thermodynamically controlled crystal tends to be densely packed compared to a kinetically controlled one.[41] From this point of view, because the crystal density of the cis complexes (ca. 1.6 g cm–3) is higher than that of the trans complexes (ca. 1.3 g cm–3), it is natural to consider that the more inert CoII and NiII preferred the kinetically controlled trans conformation, whereas more labile ZnII and MnII preferred the thermodynamically controlled cis conformation. In the case of CuII, the trans conformation is probably thermodynamically stable because of the Jahn–Teller effect.

Table 1

Summary of the Conformation and the Network Motifs of the Obtained Crystals

MII	coordination conformation	network motif
Mn, Zn	cis	1D zigzag chain
Co, Ni	trans	3D framework (interlocking of 2D rectangular nets)
Cu	trans	mixture of 1D linear chains and 3D framework (interlocking of 2D rectangular nets), not separable

Table 2

Difference of the Gibbs Free Energies (ΔG) between cis and trans Conformers

metal	ΔG /kcal mol–11
Mn	–3.90467
Co	–3.84079
Ni	–0.63212
Cu	1.25780
Zn	–1.19185

ΔG is defined as ΔG ≡ Gcis – Gtrans

Thermogravimetric analysis (TGA) of -2D-Co and -2D-Ni showed a weight decrease caused by removal of the crystal water in the pore, while -1D-Mn and -1D-Zn showed no weight loss until around 300 °C (Figure ). The amount of water estimated from TGA is 8.7% and 10.3% for -2D-Co and -2D-Ni, respectively, which corresponds to 2.5 and 3.0 water molecules per formula unit.

Figure 6

Thermogravimetric analysis. (a) -1D-Zn and -1D-Mn. (b) -2D-Co and -2D-Ni.

The N2 sorption isotherms were measured to confirm the porosity of -2D-Ni and -2D-Co. (Figure a). The N2 isotherms of -2D-Ni were classified as Type I, suggesting -2D-Ni has micropores. The N2 uptake was nearly saturated at P/P0 = 0.9. The further increase in the N2 uptake at P/P0 > 0.95 is probably due to adsorption on the crystalline surface. The N2 saturated volume (ca. 60 cm3(STP) g–1) is consistent with that of the roughly estimated volume (74.7 cm3(STP) g–1) from the crystal structure (see details in SI). The BET and Langmuir surface areas of -2D-Ni were estimated to be 177 and 183 m2g–1, respectively. However, -2D-Co showed almost no N2 uptake despite its porous structure. As for the H2 adsorption (Figure b), the Ni and Co compounds showed 90 and 20 cm3(STP) g–1, respectively, at 77 K and 1 atm. It is noteworthy that both compounds showed larger H2 uptake than N2 at 77 K and 1 atm. In particular, -2D-Co showed almost no N2 uptake, but moderate H2 uptake. Such a phenomenon is not common because the boiling point of N2 (77 K) is much higher than H2 (20 K), but it is reported in several MOFs with quite small pores.[42,43] In these MOFs, only H2 can enter the pore because H2 has a smaller kinetic diameter (2.89 Å) than N2 (3.64 Å).[44] Therefore, we presume that the unusual N2 and H2 uptake phenomenon is due to the small 1D pore of the present compounds. This presumption is supported by the hysteresis of the H2 isotherm observed in -2D-Co. Such a hysteresis is often observed when the kinetics of the adsorption are quite slow because of the slow diffusion in the small pores.

Figure 7

(a) N2 adsorption isotherms at 77 K, (b) H2 adsorption isotherms at 77 K, and (c) CO2 adsorption isotherms at 195 K in -2D-Co and -2D-Ni.

Here, we discuss the origin of the difference in the sorption properties between -2D-Co and -2D-Ni. Figure S10 shows scanning electron microscope (SEM) images of both compounds. -2D-Co crystals were needle shaped with a typical length of 10–30 μm, and we confirmed that the long axis of the crystal is parallel to the 1D pore by using SXRD measurement. Conversely, the crystal particle of -2D-Ni is much smaller than -2D-Co and typically 1 μm size. Therefore, in the case of -2D-Co, the adsorbent molecules have to diffuse a longer distance to reach the center of the crystal particle. This is one of the possible reasons for the difference. Another possible reason is in the flexibility of the framework. As we noted above, the framework of -2D-Co can be more flexible than that of -2D-Ni. Therefore, the structural change by the removal of solvent molecules should be larger in -2D-Co, which affords smaller pore size.

Such a diffusion issue is less dominant at higher temperature because of the greater lattice motion. Thus, we performed CO2 adsorption at 195 K. As a result, both compounds showed similar isotherms, as expected (Figure c). Therefore, we concluded that the difference in the adsorption isotherms between -2D-Co and -2D-Ni is due to the difference of the crystal shape or subtle differences in the pore structure, both of which affect large diffusion kinetics.

Figure shows the UV–vis–NIR spectra of -1D-M and -2D-M, which is evaluated by the Kubelka–Munk transformation of the raw diffuse reflectance spectra. The UV–vis–NIR spectrum of -1D-M shows mainly two bands at 360 and 520 nm. On the basis of the assignment of the spectrum in (H2NMe2)1.5Cr2(DHBQ)3,[45] we assigned these bands to a ligand-based π*−π* and π–π* transition, respectively. Alternatively, the UV–vis–NIR spectra of -2D-M exhibits three bands around 440, 600, and 1000 nm. The two higher energy transitions are assigned as a d–d transition involving a ligand-based π*−π* and π–π* transition, respectively. The broad peak in the mid-IR region is assigned as a ligand-to-metal charge transfer (LMCT) or metal-to-ligand charge transfer (MLCT). The sharp absorbances around 1160 and 1680 nm are probably due to overtones of contaminating water. By assuming the indirect band gap, we showed the Tacu plot in the inset of Figure . The optical band gaps of -1D-Mn, -1D-Zn, -2D-Co, and -2D-Ni were estimated to be 1.6, 1.8, 1.0, and 1.1 eV, respectively.

Figure 8

UV–vis–NIR spectra of -1D-M and -2D-M. Inset shows Tauc plot by assuming the indirect band gap.

Conclusion

In conclusion, the binary DHBQ-based CPs, cis-[M(DHBQ)(bpy)2] (M = Mn and Zn), trans-[M(DHBQ)(bpy)] (M = Co, Ni, and Cu), and trans-[Cu(DHBQ)(bpy)2] were synthesized using in situ hydrolysis of a DMBQ precursor. These were characterized with SXRD, PXRD, IR, TGA, H2, N2, and CO2 adsorption; magnetic susceptibility; and elemental analysis. This in situ hydrolysis method is beneficial not only to grow a good, single crystal but also to introduce auxiliary ligands into an MOF, which leads to the diversity of MOFs. Thus, the present synthetic approach should lead to the further application of redox-active MOFs, such as cathode materials of secondary batteries, electrode catalysts, etc.