Although dimethyl sulfoxide (DMSO) has emerged as a promising solvent for Li-air batteries, enabling reversible oxygen reduction and evolution (2Li + O2 ⇔ Li2O2), DMSO is well known to react with superoxide-like species, which are intermediates in the Li-O2 reaction, and LiOH has been detected upon discharge in addition to Li2O2. Here we show that toroidal Li2O2 particles formed upon discharge gradually convert into flake-like LiOH particles upon prolonged exposure to a DMSO-based electrolyte, and the amount of LiOH detectable increases with increasing rest time in the electrolyte. Such time-dependent electrode changes upon and after discharge are not typically monitored and can explain vastly different amounts of Li2O2 and LiOH reported in oxygen cathodes discharged in DMSO-based electrolytes. The formation of LiOH is attributable to the chemical reactivity of DMSO with Li2O2 and superoxide-like species, which is supported by our findings that commercial Li2O2 powder can decompose DMSO to DMSO2, and that the presence of KO2 accelerates both DMSO decomposition and conversion of Li2O2 into LiOH.

Understanding the oxygen reduction reaction kinetics in the presence of Na ions and the formation mechanism of discharge product(s) is key to enhancing Na− O 2 battery performance. Here we show NaO 2 as the only discharge product from Na− O 2 cells with carbon nanotubes in 1,2-dimethoxyethane from X-ray diffraction and Raman spectroscopy. Sodium peroxide dihydrate was not detected in the discharged electrode with up to 6000 ppm of H 2 O added to the electrolyte, but it was detected with ambient air exposure. In addition, we show that the sizes and distributions of NaO 2 can be highly dependent on the discharge rate, and we discuss the formation mechanisms responsible for this rate dependence. Micron-sized (∼500 nm) and nanometer-scale (∼50 nm) cubes were found on the top and bottom of a carbon nanotube (CNT) carpet electrode and along CNT sidewalls at 10 mA/g, while only micron-scale cubes (∼2 μm) were found on the top and bottom of the CNT carpet at 1000 mA/g, respectively.R echargeable metal-air (oxygen) batteries are receiving intense interest as possible alternatives to lithium-ion batteries, in particular due to their potential to provide higher gravimetric energies.1−5 While much attention has been focused on aprotic Li−O 2 batteries since their introduction in 1996 by Abraham et al.,6 substantial challenges must be addressed before widespread commercial exploitation is possible. These include the instability of aprotic electrolytes 7−9 and oxygen electrodes, 10,11 which contribute to low round-trip efficiency, poor rate capability, and poor cycle life. Recently, a metal-air battery in which lithium has been replaced by sodium has received increasing attention. Unfortunately, the factors responsible for dissimilar discharge product chemistry and morphologies are still unclear, and no correlation has been found between the type of air electrode or electrolyte and the discharge product formed. Because the discharge product crystal structure, morphology (e.g., shape and thickness), and distribution are important parameters that influence the voltage profile on discharge and charge, the rate capability, the discharge capacity, and the reversibility in metalair batteries, 23 it is critical to gain insights into the oxygen reduction kinetics in the presence of Na ions and the nucleation and growth mechanisms of discharge products.In this study, we investigate discharged Na−O 2 cells with carbon nanotube (CNT) carpet cathodes to understand the effect of discharge kinetics on the formation of the discharge product. We analyzed the chemical composition of the discharge product using X-ray diffraction (XRD) and Raman spectroscopy and found that it consisted solely of NaO 2 . The formation of Na 2 O 2 ·H 2 O was not readily detected with the addition of water to the solvent (<6000 ppm); however, the evolution of Na 2 O 2 ·2H 2 O was detected when we intentionally exposed the sample to the ambient environment. We also observed rate-dependent effects on the size and distribution of the discharge product. ...

No abstract

The performance of current Li–air batteries is greatly limited by critical obstacles such as electrolyte decomposition, high charging overpotentials, and limited cycle life. Thus, much effort is devoted to fundamental studies to understand the mechanisms of discharge/charge processes and overcome the above-mentioned obstacles. In particular, the search for new stable electrolytes is vital for long-lasting and highly cyclable batteries. The highly reactive lithium superoxide intermediate (LiO2) produced during discharge process can react with the electrolyte and produce a variety of byproducts that will shorten battery life span. To study this degradation mechanism, we investigated oxygen reduction reaction (ORR) in highly concentrated electrolyte solutions of lithium bis­(trifluoromethanesulfonyl)­amide (Li­[TFSA])/dimethyl sulfoxide (DMSO). On the basis of rotating ring disk electrode measurements, we showed that LiO2 dissolution can be limited by increasing lithium salt concentration over 2.3 mol dm–3. Our Raman results suggested that this phenomenon can be related to lack of free DMSO molecules and increasing DMSO–Li+ interactions with higher Li+ concentration. X-ray diffraction measurements for the products of ORR suggested that the side reaction of DMSO with Li2O2 and/or LiO2 could be suppressed by decreasing the solubility of LiO2 in highly concentrated electrolytes.

Understanding what controls Li-O2 battery discharge product chemistry and morphology is key to enabling its practical deployment as a low-cost, high-specific-energy energy conversion technology. Several studies have recently shown that the addition of substantial quantities (hundreds to thousands ppm) of water and weak acids to dimethoxyethane (DME)-based electrolytes can significantly increase Li-O2 battery discharge capacity, without substantially changing the discharge product chemistry, which remains Li2O2. The exact mechanisms behind these device-level improvements, however, are not yet understood. In this study, we show that the presence of water in a DME-based electrolyte decreases the rate of Li2O2 nucleation on the electrode surface during Li-O2 battery discharge, using potentiostatic electrochemical measurements, and direct, ex situ observations of Li2O2 particles. We also show that adding water to an acetonitrile (MeCN)-based electrolyte results in LiOH upon discharge, as opposed to only Li2O2. Using first principles calculations, we propose that this change in discharge product chemistry is attributable to increased proton availability, as shown by a lower pKa for water in MeCN than in DME. This study combines kinetic and morphological analyses with first principles calculations, and elucidates relationships among electrolyte composition, discharge product chemistry and growth mechanisms for the rational design of efficient metal-air batteries.