Operando detection and suppression of spurious singlet oxygen in Li-O2 batteries

CÓRDOBA, DANIEL; BENAVIDES, LEANDRO N.; MURGIDA, DANIEL H.; RODRÍGUEZ, HERNAN B.; CALVO, ERNESTO J.

Faraday Discussions

Royal Society of Chemistry

Año: 2023 vol. 248 p. 190 - 209

1359-6640

The rechargeable lithium air (oxygen) battery (Li-O2) has very high energy density, comparable to that of fossil fuels (∼3600 W h kg−1). However, the parasitic reactions of the O2 reduction products with solvent and electrolyte lead to capacity fading and poor cyclability. During the oxygen reduction reaction (ORR) in aprotic solvents, the superoxide radical anion (O2˙−) is the main one-electron reaction product, which in the presence of Li+ ions undergoes disproportionation to yield Li2O2 and O2, a fraction of which results in singlet oxygen (1O2). The very reactive 1O2 is responsible for the spurious reactions that lead to high charging overpotential and short cycle life due to solvent and electrolyte degradation. Several techniques have been used for the detection and suppression of 1O2 inside a Li-O2 battery under operation and to test the efficiency and electrochemical stability of different physical quenchers of 1O2: azide anions, 1,4-diazabicyclo[2.2.2]octane (DABCO) and triphenylamine (TPA) in different solvents (dimethyl sulfoxide (DMSO), diglyme and tetraglyme). Operando detection of 1O2 inside the battery was accomplished by following dimethylanthracene fluorescence quenching using a bifurcated optical fiber in front-face mode through a quartz window in the battery. Differential oxygen-pressure measurements during charge-discharge cycles vs. charge during battery operation showed that the number of electrons per oxygen molecule was n > 2 in the absence of physical quenchers of 1O2, due to spurious reactions, and n = 2 in the presence of physical quenchers of 1O2, proving the suppression of spurious reactions. Battery cycling at a limited specific capacity of 500 mA h gC−1 for the MWCNT cathode and 250 mA gC−1 current density, in the absence and presence of a physical quencher or a physical quencher plus the redox mediator I3−/I− (with a lithiated Nafion® membrane), showed increasing cyclability according to coulombic efficiency and cell voltage data over 100 cycles. Operando Raman studies with a quartz window at the bottom of the battery allowed detection of Li2O2 and excess I3− redox mediator during discharge and charge, respectively.