Direct Evidence of Solution-Mediated Superoxide Transport and Organic Radical Formation in Sodium-Oxygen Batteries.

Advanced large-scale electrochemical energy storage requires cost-effective battery systems with high energy densities. Aprotic sodium-oxygen (Na-O2) batteries offer advantages, being comprised of low-cost elements and possessing much lower charge overpotential and higher reversibility compared to their lithium-oxygen battery cousins. Although such differences have been explained by solution-mediated superoxide transport, the underlying nature of this mechanism is not fully understood. Water has been suggested to solubilize superoxide via formation of hydroperoxyl (HO2), but direct evidence of these HO2 radical species in cells has proven elusive. Here, we use ESR spectroscopy at 210 K to identify and quantify soluble HO2 radicals in the electrolyte-cold-trapped in situ to prolong their lifetime-in a Na-O2 cell. These investigations are coupled to parallel SEM studies that image crystalline sodium superoxide (NaO2) on the carbon cathode. The superoxide radicals were spin-trapped via reaction with 5,5-dimethyl-pyrroline N-oxide at different electrochemical stages, allowing monitoring of their production and consumption during cycling. Our results conclusively demonstrate that transport of superoxide from cathode to electrolyte leads to the nucleation and growth of NaO2, which follows classical mechanisms based on the variation of superoxide content in the electrolyte and its correlation with the crystallization of cubic NaO2. The changes in superoxide content upon charge show that charge proceeds through the reverse solution process. Furthermore, we identify the carbon-centered/oxygen-centered alkyl radicals arising from attack of these solubilized HO2 species on the diglyme solvent. This is the first direct evidence of such species, which are likely responsible for electrolyte degradation.

Estimating the Coherence of Noise in Quantum Control of a Solid-State Qubit.

To exploit a given physical system for quantum information processing, it is critical to understand the different types of noise affecting quantum control. Distinguishing coherent and incoherent errors is extremely useful as they can be reduced in different ways. Coherent errors are generally easier to reduce at the hardware level, e.g., by improving calibration, whereas some sources of incoherent errors, e.g., T_{2}^{*} processes, can be reduced by engineering robust pulses. In this work, we illustrate how purity benchmarking and randomized benchmarking can be used together to distinguish between coherent and incoherent errors and to quantify the reduction in both of them due to using optimal control pulses and accounting for the transfer function in an electron spin resonance system. We also prove that purity benchmarking provides bounds on the optimal fidelity and diamond norm that can be achieved by correcting the coherent errors through improving calibration.