Zn-based eutectic mixture as anolyte for hybrid redox flow batteries

Developing greener batteries with new chemistries is a formidable challenge, and a major focus for years to come. Redox flow batteries are receiving increasing research interest for grid-scale electrochemical energy storage owing to their unique architecture. However, challenges still remain by their low energy density as well as corrosive and/or toxic electrolytes. An anolyte based on aprotic Zn deep-eutectic-solvent, which uses low cost, abundant and environmentally benign materials, exhibits a utilizable concentration of Zn2+ ca. 1.7 M, resulting in a reversible volumetric capacity of ca. 90 A h·L−1. Combined with high efficiencies and relatively low redox potential of −1.12 V vs. Ag/AgCl, such an anolyte provides an alternative way to explore a family of anolytes using new chemistries for rechargeable Zn batteries that meet the criteria for grid-scale electrical energy storage.

Zn electrodes and are also able to suppress the self-corrosion of Zn and eliminate carbonation 15,16 . However, the solubility of Zn salts is much lower in aprotic electrolytes compared with the aqueous counter parts.
Herein, we report a Zn-based, aprotic deep-eutectic-solvent as anolyte potentially applicable to hybrid RFBs. Deep-eutectic-solvents are systems formed from a eutectic mixture of Lewis or Brønsted acids and bases which can contain a variety of anionic and/or cationic species. Compared with other aprotic solvents or ionic liquids used in RFBs, deep-eutectic-solvents are usually biodegradable, and readily available from inexpensive compounds. In this demonstration, ZnCl 2 and acetamide have been used to form eutectic solvent based on the considerations: i) eutectic solvents usually exhibit a low vapor pressure, relatively wide liquid-range, and nonflammability, as well as small surface tension, which suggest good wettability with electrode 17 ; and ii) Zn 2+ /Zn shows a low redox potential of −0.98 V vs. Ag + /Ag as well as relatively low overpotential upon electroplating/ electrostripping processes. Hence, such a Zn-based eutectic solvent that has self-contained electroactive species may serve as an ideal anolyte for the pursuit of greener RFBs with high volumetric capacities.

Experimental Section
Synthesis. To prepare Zn eutectic solvent, acetamide (99%, Alfa Aesar) and ZnCl 2 (anhydrous, 99.99%, Alfa Aesar) with different molar ratio were mixed under constant stirring at 60 °C to accelerate the formation of Zn eutectic solvent until a transparent liquid formed. To decrease viscosity and enhance ionic conductivity, EC/ DMC (3/7, v/v, water content < 30 ppm) and bis(trifluoromethane)sulfonamide lithium salt (LiTFSI, anhydrous, 99.95%, Sigma Aldrich) were added into of the as-prepared Zn eutectic solvent. The density of Zn eutectic solvents with ZnCl 2 molar fraction of 13.7%, 16.7% and19.4% are 2.02, 2.16, and 2.29 g·cm −3 , respectively. The concentration of the anolytes used for evaluating Zn electroplating/electrostripping efficiency is adjusted by EC/ DMC solvent. The ZnCl 2 concentration in the Zn-based eutectic solvent are 4, 4.25, 4.5, 4.75, 5, 5.25, 5.5, 5. (40 mmol, 97%, Alfa Aesar) was added into the solution and maintained at 90 °C for another 2 hrs. Afterwards, the precipitation was collected by centrifuging and washed with a small amount of frozen deionized water for several times to remove NH 4 Cl content. The solid was obtained after solvent removal by spin steaming, followed by filtering with silica gel, and washed with acetonitrile. Finally, Further purification was carried out by drying in vacuum to remove the residual water.
Characterizations. DSC (Perkin-Elmer DSC-7 calorimeter) measurement was carried out by calibrating with indium and flushed with N 2 and scanning at a rate of 10 °C·min −1 from −70 °C to 100 °C using about 10 μL of the Zn eutectic solvent packed in standard aluminum pans. Fourier Transform infrared spectroscopy (FTIR, Perkin-Elmer spectrophotometer) spectra were collected in the frequency range of 4000-400 cm −1 at room temperature with a resolution of 4 cm −1 . The viscosity and conductivity of the Zn eutectic mixture were determined by a rotary viscometer (NDJ-1 Shanghai Yueping Scientific Instrument Corporation Limited, China) and conductivity meter with four channel conductivity electrodes (STARTER 300 C), respectively. Scanning electron microscopy (SEM, Zeiss Supra 55) image showing the surface structure of Zn foil was recorded after washing with pure ethylene carbonate and dimethyl carbonate. Cyclic voltammetry tests were performed on a potentiostat (CHI760E, CH Instruments) with a three-electrode configuration of with glassy carbon as the working and counter electrodes, and silver chloride electrode as the reference electrode. Galvanostatic cycling performance was performed on a potentiostat (BT-2043, Arbin Instruments) with current density of 0.12 mA·cm −2 in the potential range of 1-2.3 V and 0.5-2.5 V for Zn eutectic mixture | [Fe(phen) 3 ](BF 4 ) 2 and Zn eutectic mixture|LiI cells, respectively. Rotation disk electrode (RDE, ALS RRDE-3A) voltammetry investigation was carried out using glassy-carbon as working electrode, Pt wire as counter electrode and silver chloride electrode as reference electrode. The electrolyte was composed of 2. Calculation details. The first principles calculations are performed using the Gaussian 09 software based on DFT combined with multiwfn software 18,19 . The structure optimization of [Fe(phen) 3 ] 2+ and [Fe(phen) 3 ] 3+ is performed using the B3lyp as the functional and 6-31 g(d,p)/lanl2dz as the mixed basis sets for C, N, H and Fe, respectively 20 . The structure optimization of Zn (II) complexes is performed using the B3lyp/def2tzvp level of theory 21 . The B3lyp/def2tzvp level of theory was selected to optimize the structure of Zn (II) complexes without imaginary frequencies. The hydrogen bond interaction energy was calculated using B3lyp/6-311 g*. The interaction energy (ΔE) between different molecules was calculated as the difference between the total energy of the complex and the individual energy according to where E is the single point energy after structure optimization without imaginary frequencies, E total is the single point energy of acetamide and x (x stands for acetamide, EC or DMC), E act is the single point energy of acetamide, and E x is the single point energy of acetamide, EC or DMC. The dissociation energy was calculated to: where ΔE is the dissociation energy; E total (y) is the total energy of [ZnCl(acetamide)] + , [ZnCl(acetamide) 2 ] + , and [ZnCl(acetamide) 3 ] + , respectively; E Zn , E Cl , and E act are the energy of Zn 2+ , Cl − and acetamide, respectively; n is the number of acetamide contained in the Zn(II) complexes. The activation energy was calculated according to the dynamical electron transfer theory using ΔG = 0.25 (Δ r G + λ) 2 /λ, where ΔG is the Gibbs energy of activation, Δ r G is the standard reaction Gibbs energy for electron transfer process: [ZnCl(acetamide) n ] + + 2e − → Zn + Cl − + n(acetamide), and λ is the reorganization energy 23 .

Results
The Zn-based eutectic solvent was readily prepared by simply mixing ZnCl 2 with acetamide under constant stirring at 60 °C. The mixing process was quick that a transparent liquid formed within 15 min (Fig. 1a). The lower and upper molar fraction limits of ZnCl 2 for the formation of a stable eutectic solvent were determined to be 13.7% and 19.4%, respectively (Fig. 1b) 3 ] + was likely to be triangle and tetrahedral structure. These Zn(II) complexes should not formed through a hydrogen bond, but more probably through an oxygen bond between the acetamide and Zn 2+ center. Such a coordination geometry suggested that the upper limit of ZnCl 2 at which the Zn eutectic solvent could be held should depend upon the stability of acetamide molecule. The Zn eutectic solvent would not decompose but rather lose the organic component at elevated temperatures close to the boiling point of the pure component.
Power-compensated differential scanning calorimetry (DSC) measurement indicated that the as-prepared Zn eutectic solvent showed a lowest freezing temperature of −38 °C at a ZnCl 2 molar ratio of 16.7%, and no phase change was observed up to 100 °C (Fig. 2a). Such a wide temperature range should be superior to most aqueous and aprotic electrolyte used for RFBs. However, it's necessary to point out that, as found in other eutectics 25,26 , the as-prepared Zn eutectic solvent exhibited relatively high viscosity of 10-25 cP in the ZnCl 2 concentration rage of 4.5-6.5 M (Fig. S2, Supplementary material), which should be resulted from the charge delocalization occurred through the hydrogen bonding between halide ion and hydrogen-donor moiety 25 . Decrease the viscosity would allow sufficient mass transport and thus enhance the conductivity, because the molar conductivity of an electrolyte should be inversely proportional to the viscosity of the solvent according to the Stokes' law 27 . According to the calculation results, by adding ethylene carbonate/dimethyl carbonate (EC/DMC, 3:7, v-v) into the Zn eutectic solvent, interaction among the solvent molecules could be weakened by means of tuning differing degrees of hydrogen bonding among acetamide, EC and DMC molecules (Fig. S3, Supplementary material). The resultant Zn eutectic mixture showed decreased viscosity of 1-16 cP with equivalent Zn 2+ concentration ranging from of 4 to 7 M (Fig. 2b) and ionic conductivity of 10 −3 S·cm −1 (Fig. 2c), comparable to the nonaqueous electrolytes for Li-ion batteries 28 . It should be noted that the participation of EC and DMC molecules lead to weakened molecular interactions rather than affected the coordination geometries of Zn 2+ as evident by Fourier Transform infrared (FTIR) spectra (Fig. 2d). On one hand, the obvious red-shifted broad band between 3600 and 3100 cm −1 could be assigned to the hydrogen bond forming between amine and carbonyl groups of acetamide 29 . This broad band showed a redshift after the adding of EC/DMC solvent, indicating that molecular interaction by hydrogen bonding was weakened, consistent with aforementioned calculation results. On the other hand, pure acetamide showed a characteristic band at ca. 1670 cm −1 is assigned to the vibration of carbonyl group, which showed a blue shift to ca. 1650 cm −1 after the formation of Zn eutectic solvent by resulting from the formation of metal oxygen bond between Zn 2+ and C=O group of acetamide. This band remained at the same wavenumber after the adding The cell used to evaluate the electrochemical performance of Zn-based eutectic mixture as anolyte was carried out in a home-built cell similar to previous reports 30 . The cell was composed of carbon electrode and Zn foil as current collector in cathode and anode side, respectively, Zn-based eutectic mixture as anolyte, a Li + -ion conductive membrane as separator, and other solutions containing redox species with proper redox potential as catholyte (Fig. S4, Supplementary material). In this demonstration, LiI and [Fe(phen) 3 ](BF 4 ) 2 (phen = 1,10-phenanthroline) were used as the active material in the catholyte because [Fe(phen) 3 ] 3+ /[Fe(phen) 3 ] 2+ redox couple shows good structural stability both in oxidative and reductive states, fast redox kinetics with rate constant on the order of 10 −3 cm·s −1 at room temperature, high redox potential of 0.68 V vs. Ag/AgCl, and stable cyclability as well as cycling efficiencies and in aprotic electrolyte (Figs S5-S7, Supplementary material), while I 3 − /Iredox couple shows reliable reversibility as well as high solubility in aqueous electrolyte [31][32][33] . The as-fabricated cell was in its discharged state. Upon charging, [Fe(phen) 3 ] 2+ was oxidized into [Fe(phen) 3 ] 3+ on the carbon electrode in the cathode chamber, while Zn(II) complex in the eutectic mixture was electroplated onto the Zn foil in the anode chamber. Simultaneously, the charge carrier, Li + , diffused from the catholyte into the anolyte to balance the charge.
The electrochemical properties of Zn eutectic mixture were firstly investigated to address some general concerns about its capability for practical battery application. The electrochemical stability of Zn eutectic mixture was first investigated using a conventional three-electrode configuration with glassy carbon as the working and counter electrodes, and silver chloride electrode as the reference electrode. Cyclic voltammogram profile of the Zn eutectic mixture (Fig. 3a) demonstrated that no obvious current fluctuation was observed in the potential range of −0.8 to 1.3 V vs. Ag/AgCl. Potential lower than −0.8 V vs. Ag/AgCl resulted in Zn 2+ electroplating, while potential higher than 1.3 V vs. Ag/AgCl led to the decomposition of acetamide. The Zn eutectic mixture showed an impressive electrochemical stable window up to 1.3 V vs. Ag/AgCl, which is comparable with most organic electrolyte used in Li + -ion batteries 34 . Moreover, the utilization ratio of Zn eutectic mixture with variable ZnCl 2 concentration was examined in a half-cell (5 M Zn eutectic mixture | separator | Li). Zn eutectic mixture showed a high electroplating/electrostripping ratio close to 100% when the concentration of ZnCl 2 between 4.5 and 6.25 M, compared with that obtained when the concentration of ZnCl 2 below 4.5 or above 6.25 M (Fig. 3b). This phenomenon could be attributed to phase separation of Zn eutectic mixture according to the general phase diagram of eutectics 26 . Furthermore, short term cycling test (Fig. 3c) revealed that the half-cell showed stable charge/discharge potential around 2.33/2.12 V vs. Li + /Li. The potential gap was ca. 0.25 V at the first a few cycles but gradually decreased to ca. 0.2 V in the subsequent cycles, which should be attributed to decreased internal resistance owing to the enhanced wettability between Zn eutectic mixture and Zn foil, as well as improved permeation of Li + -ion through the separator. It's important to note that Zn deposited from deep-eutectic-solvents tends to have a compact microcrystalline structure, in contrast to Zn deposited from aqueous electrolytes which tends to have a dendritic structure in the absence of strong base additives 16 . The current collector after electroplating for sufficient time (Fig. S8, Supplementary material) was examined by Ex-situ SEM, which confirmed a dendrite-free  (Fig. S10, Supplementary material). Galvanostatic cycling profiles of the full cell demonstrated stable cycling performance (Fig. 4a). During discharge, the voltage decreased monotonically until reaching the 1 V cutoff. While upon charging, the voltage jumped to 1.8 V, and then increased monotonically until reaching the 2.4 V cutoff. The average overpotentials were similar during the discharge and recharge processes, suggesting that the electrochemical steps involved in the forward and reverse cell reactions exhibited similar activation energies. The Coulombic and energy efficiency maintained ca. 100% and 78%, respectively (Fig. 4b). During the first 30 cycles, the cell showed a steady increase in capacity, which eventually stabilized to ca. 3 A h·L −1 based on the total volume of the anolyte and catholyte. To further demonstrate the capability of the Zn eutectic mixture, excess concentrated LiI aqueous solution was used as the catholyte (5 M Zn eutectic mixture | separator | 3 M LiI). The cell was able to deliver a volumetric capacity of ca. 90 A h·L −1 based on the volume of the anolyte (Fig. 4c), corresponding to a utilizable concentration ca. 1.7 M out of 6.25 M ZnCl 2 in the Zn eutectic mixture could be reversibly used. The Coulombic efficiency of the full cell was nearly 100% in the initial cycle, which was in consistent with that achieved in the half cell with ZnCl 2 concentration between 4.5 M and 6.25 M. In consideration of two-electron redox reaction of Zn 2+ , such a high concentration exceeded that of the commonly used stability limit (<1.7 M) for RFBs 36 .

Discussion
The mechanism of Zn electroplating/electrostripping in the Zn eutectic mixture was studied using DFT-based computational simulation (Fig. 5). The calculation results (  3 ] + being the most stable geometry; on the other hand, Zn electroplating/electrostripping processes were not likely to occur through direct dissociation of Zn(II) complexes in the Zn eutectic mixture due to such a high dissociation energy. Further analysis on the activation energy of the three coordination geometries revealed that [ZnCl(acetamide)] + and [ZnCl(acetamide) 3 ] + were responsible for Zn electroplating and electrostripping, respectively. In the electroplating process, [ZnCl(acetamide)] + showed an activation energy of 0.12 kcal·mol −1 , compared with those of 9.08 kcal·mol −1 and 13.54 kcal·mol −1 for [ZnCl(acetamide) 2 ] + and [ZnCl(acetamide) 3 ] + , respectively, indicating that the electroplating process should be more thermodynamically preferable for [ZnCl(acetamide)] + being electrochemically reduced to Zn than the  While in the electrostripping process, however, the activation energy of [ZnCl(acetamide) 3 ] + was found to be 148.063 kcal·mol −1 , lower than those of the other two geometries of 188.483 and 163.536 kcal·mol −1 for [ZnCl(acetamide)] + and [ZnCl(acetamide) 2 ] + , respectively. Therefore, it's likely Zn electrostripping occurred preferably following one major path via reaction (4)

Conclusions
we present here a successful demonstration of the utilization of Zn-based deep-eutectic-solvent that can be readily prepared using inexpensive and abundant materials as anolyte for hybrid RFBs. The new chemistry applied here is inherently environmental-friendly and elementally abundant. The Zn eutectic mixture is capable to utilize 1.7 M electroactive Zn 2+ thus delivers a volumetric capacity ca. 90 A h·L −1 with good cycling efficiencies. Zn electroplating/electrostripping in this anolyte shows good reversibility and no sign of dendrite formation. Zn electroplating/electrostripping process is investigated by DFT-based computational simulation, based on which the Zn redox reaction mechanism has been proposed. The cell architecture is compatible with traditional hybrid RFBs and can be readily scale-up. Together with its low cost and environmental benignity, such an anolyte may find its application in stationary electric energy storage with the rapidly growing quest for greener and more sustainable batteries.