200-Fold Lifetime Extension of 2,6- Dihydroxyanthraquinone Electrolyte during Flow Battery Operation.

We study the capacity fade rate of a flow battery utilizing 2,6-dihydroxyanthraquinone (DHAQ) and its dependence on hydroxide concentration, state of charge, cutoff voltages for the discharge step and for the electrochemical regeneration (oxidation of decomposition compounds back to active species) step, and the period of performing the electrochemical regeneration events. Our observations confirm that the first decomposition product, 2,6-dihydroxyanthrone (DHA), is stable, but after electro-oxidative dimerization, the anthrone dimer decomposes. We identify conditions for which there is little time after dimerization until the dimer is rapidly reoxidized electrochemically to form DHAQ. Combining these approaches, we decrease the fade rate to 0.02%/day, which is 18 times lower than the lowest rate reported previously of 0.38%/day, and over 200 times lower than the value under standard cycling conditions of 4.3%/day. The findings and their mechanistic interpretation are expected to extend the lifetime and enhance the effectiveness of in situ electrochemical regeneration for other electroactive species with finite lifetimes.

In situ electrochemical recomposition of decomposed redox-active species in aqueous organic flow batteries.

Aqueous organic redox flow batteries offer a safe and potentially inexpensive solution to the problem of storing massive amounts of electricity produced from intermittent renewables. However, molecular decomposition represents a major barrier to commercialization-and although structural modifications can improve stability, it comes at the expense of synthetic cost and molecular weight. Now, utilizing 2,6-dihydroxy-anthraquinone (DHAQ) without further structural modification, we demonstrate that the regeneration of the original molecule after decomposition represents a viable route to achieve low-cost, long-lifetime aqueous organic redox flow batteries. We used in situ (online) NMR and electron paramagnetic resonance, and complementary electrochemical analyses to show that the decomposition compound 2,6-dihydroxy-anthrone (DHA) and its tautomer, 2,6-dihydroxy-anthranol (DHAL) can be recomposed to DHAQ electrochemically through two steps: oxidation of DHA(L)2- to the dimer (DHA)24- by one-electron transfer followed by oxidation of (DHA)24- to DHAQ2- by three-electron transfer per DHAQ molecule. This electrochemical regeneration process also rejuvenates the positive electrolyte-rebalancing the states of charge of both electrolytes without introducing extra ions.

Electrochemical Glucose Sensors Enhanced by Methyl Viologen and Vertically Aligned Carbon Nanotube Channels.

Freestanding, vertically aligned carbon nanotubes (VACNTs) were patterned into 16 µm diameter microchannel arrays for flow-through electrochemical glucose sensing. Non-enzymatic sensing of glucose was achieved by the chemical reaction of glucose with methyl viologen (MV) at an elevated temperature and pH (0.1 M NaOH), followed by the electrochemical reaction of reduced-MV with the VACNT surface. The MV sensor required no functionalization (including no metal) and was able to produce on average 3.4 electrons per glucose molecule. The current density of the MV sensor was linear with both flow rate and glucose concentration. Challenges with interference chemicals were mitigated by operating at a low potential of -0.2 V vs Ag/AgCl. As a comparison, enzymatic VACNT sensors with platinum nano-urchins were functionalized with glucose oxidase by covalent binding (1-ethyl-3-(-3-dimethylaminopropyl)carbodiimide/ N-hydroxysuccinimide) or by polymer entrapment [poly(3,4-ethylene-dioxythiophene)] and operated in phosphate buffered saline. With normalization by the overall cross-sectional area of the flow (0.713 cm2), the sensitivity of the MV, enzyme-in-solution, and covalent sensors were 45.93, 18.77, and 1.815 mA cm-2 mM-1, respectively. Corresponding limits of detection were 100, 194, and 311 nM glucose. The linear sensing ranges for the sensors were 250 nM to 200 µM glucose for the MV sensor, 500 nM to 200 µM glucose for the enzyme-in-solution sensor, and 1 µM to 6 mM glucose for the covalent sensor. The flow cell and sensor cross-sectional area were scaled down (0.020 cm2) to enable detection from 200 µL of glucose with MV by flow injection analysis. The sensitivity of the small MV sensor was 5.002 mA cm-2 mM-1, with a limit of detection of 360 nM glucose and a linear range up to at least 150 µM glucose. The small MV sensor has the potential to measure glucose levels found in 200 µL of saliva.