Microwave Synthesis of Cu-Viologen-MOFs and Their Electrochemical Redox Capabilities

Abstract

1. Introduction

Metal-Organic Frameworks (MOFs) are porous hybrid materials consisting of jungle-gym-like assemblies of metals (called connectors) and organic molecules (called linkers). Owing to t their regularly porous structure, their properties can be tuned by post-synthetic modifications, as they are expected to be functionalized for gas adsorption / separation, catalysis, and batteries.[1]

Although majority of MOFs are redox inactive due to high ionicity of the metal-organic bonds, there have been a few examples to claim reversible redox reactions by employing redox active constituents such as transition metal nodes and aromatic organic linkers. It certainly deserves to seek for electrochemically active MOFs and understand the rules to make them useful for electrochemical applications.

In this study, we synthesized novel MOFs by combination of redox active Cu and viologens. While reversible redox of Cu in a MOF structure has been reported [2], viologens are also well known for their reversible redox reactions. For their connection, Lewis basic carboxylates were introduced as N, N'-bis(2-carboxymethyl)-4,4'-bipyridinium dibromide (CM₂V) and N, N'-bis(2-carboxyethyl)-4,4'-bipyridinium dichloride (CE₂V) (Fig. 1). Microwave solvolysis has been employed as a method of synthesis of their MOFs, which were characterized for their redox capabilities.

1. Experiment

Copper nitrate trihydrate (600 mg), the bromide salt of CM₂V, or 100 mg of the chloride salt of CE₂V were dissolved in 20 ml of methanol and heated by subjecting microwave to obtain the precipitate at 100°C for 30 min. To eliminate unreacted organic ligands, an excess amount of copper source was added. The obtained precipitate was washed several times with methanol and separated in a centrifuge at 3500 rpm, 20 min.

After that, the samples were vacuum dried at 70°C for 12 hours to obtain yellow and pale blue powder samples, respectively. The crystal structure and composition of each sample were identified by XRD, TG-DTA, FT-IR, UV-vis, and BET measurements. The redox function was evaluated by electrochemical measurements of the electrodes coated on the FTO substrate with DMF solution dispersed with MOF and PVDF binder.

1. Results and discussion

Two types of samples (Cu-CM₂V-MOF and Cu-CE₂V-MOF) were isolated from the precursor solution of dissolved copper and various linkers. XRD and SEM showed that Cu-CM₂V-MOF and Cu-CE₂V-MOF were crystalline samples with a particle size of about 5 µm and a shape of spherical or plate-like Cu crosslinked by each molecule. From the XRD profile, all the peaks of the prepared samples were different from both copper nitrate and each viologen, and a diffraction peak was observed on the low angle around 10°, which suggests that a compound with a wide interlayer spacing was formed.

Furthermore, TG-DTA measurements of each sample showed combustion peaks originating from each organic molecule with 54% and 58% weight loss at around 375°C for Cu-CM₂V-MOF and 230°C for Cu-CE₂V-MOF, respectively, indicating that the obtained sample was a composite material consisting of copper and viologen organic ligand.

XRD patterns of the sample after calcination of Cu-CM₂V-mof under air showed the formation of copper (II) bromide, and the weight loss of TG-DTA indicated that the ratio of copper to viologen in the fabricated sample was 1:2.

The UV-vis spectra of each samples showed absorption at 279 nm and 273 nm from viologen in Cu-CM₂V-MOF and Cu-CE₂V-MOF, respectively, and broad absorption from 600 nm to 800 nm corresponded with the d-d transition of copper, which supports the above results.

On the other hand, FT-IR measurements of the CM₂V and CE₂V monomers revealed carboxylic acid-derived stretching vibration peaks at 1742 cm^-1 and 1719 cm^-1, respectively, while these peaks were not observed in the prepared samples. The carbonyl ν_{as coo-} peaks were observed at 1610 cm^-1 and 1630 cm^-1, and ν_{s coo-} peaks at 1414 cm^-1 and 1401 cm^-1, suggesting that the samples were synthesized by bridging copper atoms with viologen.

The BET surface areas of Cu-CM₂V-MOF and Cu-CE₂V-MOF were measured to be 24.46 and 11.12 (m²/g), respectively, while the crystal sizes were micro-ordered, suggesting that the samples were porous MOFs.

When the prepared MOF electrodes were immersed in 0.5 M KCl solution, the Cu-CE₂V-MOF electrode faded and dissolved, while the Cu-CM₂V-MOF electrode did not dissolve.  Therefore, we focused on the Cu-CM₂V-MOF electrode to evaluate its redox function by cyclic voltammetry (Fig. 2).

The potential range of -0.85 V ~ +0.5 V (vs. Ag/AgCl) was measured for 300 cycles at a sweep rate of 50 mV/s. The redox behaviour assumed as the first reduced species of CM2V described as CM2V ⇄ CM2V- + e-(E1/2 = -0.61 V vs. Ag/AgCl) was observed up to 150 cycles. The redox behaviours expected as Cu0 ⇄Cu+ +e- (E1/2 = -0.21 V vs. Ag/AgCl) and Cu+ ⇄Cu2+ +e- (E1/2 = +0.23 V vs. Ag/AgCl) were appeared gradually, and the redox currents decreased from around 250 to 300 cycles. At the 150th cycle, when the peak currents of viologen and copper reached their maximum, the reversibility ratio derived from the ratio of the total reduction current to oxidation current was estimated to be more than 90%. The redox activity ratio, which is the ratio of the theoretical active amount of MOFs on the electrode to the amount of chemical species involved in the actual oxidation reaction, was estimated to be 1.3% at the maximum (150 cycles). It is clear that only the MOF particles grounded on the FTO electrode surface could reversibly redox.

These results indicate that the redox of viologen MOFs originating from the connector and linker, respectively, can occur within the same structure. Although there is still a lot of issues for improvement in the electrical connection of the dead MOFs on the electrode, we believe that elucidating the details of the redox mechanism will lead to the development of batteries.