Redox flow batteries (RFBs) are promising energy storage candidates for grid deployment of intermittent renewable energy sources such as wind power and solar energy. Various new redox-active materials have been introduced to develop cost-effective and high-power-density next-generation RFBs. Electrochemical kinetics play critical roles in influencing RFB performance, notably the overpotential and cell power density. Thus, determining the kinetic parameters for the employed redox-active species is essential. In this Perspective, we provide the background, guidelines, and limitations for a proposed electrochemical protocol to define the kinetics of redox-active species in RFBs.

Free-floating, and copper-supported, few-layer graphene sheets were spontaneously modified from an aqueous solution containing nitrobenzenediazonium ions. The infrared spectra of the chemically modified (copper etched) free-floating graphene were measured in transmission mode by manipulating the sheets onto a KBr disc. The major advantage to this method is the ability to release the graphene sheets off the disc to refloat on a water bath, allowing the graphene to be further modified or deposited onto a new substrate suitable for other analysis. In this study, graphene sheets were then mounted onto highly ordered pyrolytic graphite (HOPG) for atomic force microscopy imaging and electrochemical measurements. The results show there are at least two reaction pathways for spontaneous film grafting to graphene: the commonly accepted aryl radical leading to films containing −C−C− linkages and a direct reaction of the diazonium cation with graphene to give films containing −NN− linkages. The ability to manipulate modified graphene sheets onto electrodes with two orientations, with the film exposed to electrolyte solution or sandwiched between graphene and HOPG, leads to different estimates of the surface concentration of electroactive groups. When the film is sandwiched between graphene and HOPG, two electroreduction signals for the nitro group are seen and much larger surface concentrations are measured. This is the first account of such a signal and is tentatively attributed to different peak potentials for reduction of nitro groups at graphene and HOPG. The solution permeability through the graphene sheet and attached films has important electrochemical consequences for systems of this type employed in supercapacitor applications.

The quantum capacitance from large area, 6- to 7-layered graphene, and chemically modified graphene, was determined using electrochemical impedance spectroscopy measured at 100 Hz in a three-electrode electrochemical cell and aqueous acidic solution. Aryldiazonium chemistry was used to modify one side of the few-layered graphene sheets with methoxy- and iodo-phenyl groups. The graphene sheets were then mounted via an aqueous transfer method onto an epoxy substrate. It was found that both sides of the graphene sheets could be accessed by the electrolyte and thus the sheets were in a pseudo free-floating arrangement. The results show a complex quantum capacitance behavior with applied electrode potential, and that behavior was not stable with potential cycling. Chemically modified samples have similar quantum capacitance minimum to unmodified graphene, and all samples have a shallow U-shaped relationship with respect to applied electrode potential. The results show chemically modifying one side of few-layer graphene sheets using diazonium chemistry was not detrimental to the measured quantum capacitance minimum at the potential of zero charge, but the capacitance remained low over a large potential window which is not desirable for supercapacitor applications.

The utility of supercapacitors for both fixed and portable energy storage would be greatly enhanced if their energy density could be increased while maintaining their high power density, fast charging time and low cost. This study describes a simple, solution-phase and scalable modification of carbon materials by a covalently bonded "brush" of hydrogenterminated graphene ribbons (GRs) with layer thicknesses of 2-20 nm, resulting in a 20-100 times increase in the areal capacitance of the unmodified electrode surface. On a flat sp 2 carbon surface modified by GRs, the capacitance exceeds 1200 µF cm -2 in 0.1 M H 2 SO 4 due to a distinct type of pseudocapacitance during constant current charge/discharge cycling, with minor changes in voltammetry after 10,000 cycles between -0.2 and +0.8 V vs. Ag/AgCl. Modification of high surface area carbon black electrodes with GRs yields capacitances of 950-1890 F g -1 , power densities >40 W g -1 , and minimal change in capacitance during 1500 charge/discharge cycles at 20 A g -1 . A capacitance of 1890 F g -1 affords an energy density of 318 Wh kg -1 operating at 1.1 V and 590 Wh kg -1 at 1.5 V. The projected energy density of a hybrid GR/Carbon supercapacitor greatly exceeds the current 10 Wh kg -1 for commercial supercapacitors, and approaches that of lithium ion batteries.

The surface concentration of nitrophenyl (NP) films is commonly calculated from the charge associated with its redox reactions in aqueous solution. Reduction is usually carried out in acidic or basic solution and is assumed to proceed by either four electrons, or a mixed four and six electron process. Here, we reproducibly graft NP films of different thicknesses to glassy carbon electrodes from the corresponding diazonium ion and examine the reduction of the films in strongly acidic and basic solutions. We demonstrate that in most, if not all, cases, assumption of a four‐electron reduction of NP groups coupled with use of the redox reaction of the hydroxylamine product will lead to an underestimation of the surface concentration of NP groups. Use of acidic media with the assumption of a mixed four and six electron reduction of NP groups is shown to be an electrochemically and chemically sensible approach to estimating surface concentrations of NP groups.

We have recently reported that reversible electrowetting can be observed on the basal plane of graphite, without the presence of a dielectric layer, in both liquid/air and liquid/liquid configurations. The influence of carbon structure on the wetting phenomenon is investigated in more detail here. Specifically, it is shown that the adsorption of adventitious impurities on the graphite surface markedly suppresses the electrowetting response. Similarly, the use of pyrolysed carbon films, although exhibiting a roughness below the threshold previously identified as the barrier to wetting on basal plane graphite, does not give a noticeable electrowetting response, which leads us to conclude that specific interactions at the water–graphite interface as well as graphite crystallinity are responsible for the reversible response seen in the latter case. Preliminary experiments on mechanically exfoliated and chemical vapour deposition grown graphene are also reported.

Humidity- and temperature-dependent errors in concentrations reported by electrochemical sensors for atmospheric nitrogen dioxide significantly limit the reliability of the data. A basic understanding of the source of these errors has been missing. Empirical, software-based corrections are of limited reliability. The sensors feature a 40 wt % (≈4 molal) sulfuric acid electrolyte, and carbon working and quasi-reference (QRE) electrodes. We show that the sensor behaves as a truncated transmission line with resistance and capacitance elements varying with humidity. High-amplitude current fluctuations are due to humidity fluctuations, and are charging currents in response to fluctuations in interfacial capacitance. Baseline currents are due to very small differences in the open-circuit electrode potential between working and reference electrodes. We deduce that acid concentration changes in the meniscus within the porous electrode structure, in response to changes in the ambient temperature and humidity, cause both the capacitance fluctuations and the baseline changes. The open-circuit potential differences driving the baseline current variations are in part due to a difference in the liquid junction potential between the QRE and working electrode, dependent on humidity and temperature and caused by a gradient of acid concentration, and in part due to temperature- and acid-concentration-dependent variations in the rate of the potential-determining reactions. Based on the understanding obtained, we demonstrate a simple hardware change that corrects these unwanted errors.

The differential capacitance of 1–2 layered and 6–7 layered graphene (LG) was measured in aqueous 0.01, 0.1, 1.0, and 3 M perchloric and sulfuric acid solutions. The total measured capacitance was evaluated for approx. ±500 mV either side of the potential of zero charge to observe the contribution from the quantum capacitance and shielding effects on the measured capacitance. The experimental results were compared to the recent theoretical evaluations of similar electrode–electrolyte interfaces for supercapacitor applications. At 6–7 LG electrodes, the measured differential capacitance was dependent on the solution and electrical double layer structures, and although the 1–2 layered electrode showed far fewer differences upon changing solution conditions, it was not strictly independent. The concept of shielding effects within the graphene electrode and a dielectric capacitance as proposed by theory would account for these observations.