Michal Afri

On the Challenge of Electrolyte Solutions for Li–Air Batteries: Monitoring Oxygen Reduction and Related Reactions in Polyether Solutions by Spectroscopy and EQCM

Polyether solvents are considered interesting and important candidates for Li-O2 battery systems. Discharge of Li-O2 battery systems forms Li oxides. Their mechanism of formation is complex. The stability of most relevant polar aprotic solvents toward these Li oxides is questionable. Specially high surface area carbon electrodes were developed for the present work. In this study, several spectroscopic tools and in situ measurements using electrochemical quartz crystal microbalance (EQCM) were employed to explore the discharge-charge processes and related side reactions in Li-O2 battery systems containing electrolyte solutions based on triglyme/lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) electrolyte solutions. The systematic mechanism of lithium oxides formation was monitored. A combination of Fourier transform infrared (FTIR), NMR, and matrix-assisted laser desorption/ionization (MALDI) measurements in conjunction with electrochemical studies demonstrated the intrinsic instability and incompatibility of polyether solvents for Li-air batteries.

Understanding the behavior of Li–oxygen cells containing LiI

Mankind has been in an unending search for efficient sources of energy. The coupling of lithium and oxygen in aprotic solvents would seem to be a most promising direction for electrochemistry. Indeed, if successful, this system could compete with technologies such as the internal combustion engine and provide an energy density that would accommodate the demands of electric vehicles. All this promise has not yet reached fruition because of a plethora of practical barriers and challenges. These include solvent and electrode stability, pronounced overvoltage for oxygen evolution reactions, limited cycle life and rate capability. One of the approaches suggested to facilitate the oxygen evolution reactions and improve rate capability is the use of redox mediators such as iodine for the fast oxidation of lithium peroxide. In this paper we have examined LiI as an electrolyte and additive in Li oxygen cells with ethereal electrolyte solutions. At high concentrations of LiI, the presence of the salt promotes a side reaction that forms LiOH as a major product. In turn, the presence of oxygen facilitates the reduction of I3− to 3I− in these systems. At very low concentrations of LiI, oxygen is reduced to Li2O2. The iodine formed in the anodic reaction serves as a redox mediator for Li2O2 oxidation.

Li–O2 cells with LiBr as an electrolyte and a redox mediator

After many years of successful and disappointing results, the field of Li–O2 research seems to have reached an equilibrium state. The extensive knowledge that has accrued through advanced analytical studies enables us to delineate the weaknesses of the Li–O2 battery. It is now clear that the instability of the cell components toward extreme conditions existing during cell operation leads to early cell failure as well. One serious challenge is the high oxidation potential applied during the charge process. Redox-mediators may reduce the over-potential and, therefore, improve the efficiency and cyclability of Li–O2 cells. Their use in Li–O2 cells is mandatory. We have previously shown that LiI can indeed behave in such a manner; however, it also promotes the formation of side products during cell operation. We have, therefore, embarked on a comprehensive study of lithium halide salts as electrolytes for use in Li–O2 cells. We examine herein the effect of other components in the cell, such as solvents and contaminants, on the lithium halide salt activity. Based on the electrochemical behavior and the identity of the final cell products under various conditions, we can glean substantial information regarding the detailed operation mechanisms for each specific case. We have concluded that low concentration of LiBr in diglyme solution can improve the cell performance with fewer side effects than LiI. With LiBr, only the desired Li2O2 is formed during discharge. During charge, the bromine redox couple (Br−/Br3−) can reduce the oxidation potential to only 3.5 V. Higher efficiency and better cyclability of cells containing LiBr demonstrate that the electrolyte solution is the key to a successful Li–O2 battery.

Catalytic Behavior of Lithium Nitrate in Li-O2 Cells.

The development of a successful Li-O2 battery depends to a large extent on the discovery of electrolyte solutions that remain chemically stable through the reduction and oxidation reactions that occur during cell operations. The influence of the electrolyte anions on the behavior of Li-O2 cells was thought to be negligible. However, it has recently been suggested that specific anions can have a dramatic effect on the chemistry of a Li-O2 cell. In the present paper, we describe how LiNO3 in polyether solvents can improve both oxygen reduction (ORR) and oxygen evolution (OER) reactions. In particular, the nitrate anion can enhance the ORR by enabling a mechanism that involves solubilized species like superoxide radicals, which allows for the formation of submicronic Li2O2 particles. Such phenomena were also observed in Li-O2 cells with high donor number solvents, such as dimethyl sulfoxide dimethylformamide (DMF) and dimethylacetamide (DMA). Nevertheless, their instability toward oxygen reduction, lithium metals, and high oxidation potentials renders them less suitable than polyether solvents. In turn, using catalysts like LiI to reduce the OER overpotential might enhance parasitic reactions. We show herein that LiNO3 can serve as an electrolyte and useful redox mediator. NO2(-) ions are formed by the reduction of nitrate ions on the anode. Their oxidation forms NO2, which readily oxidizes to Li2O2. The latter process moves the OER overpotentials down into a potential window suitable for polyether solvent-based cells. Advanced analytical tools, including in situ electrochemical quartz microbalance (EQCM) and ESR plus XPS, HR-SEM, and impedance spectroscopy, were used for the studies reported herein.

Mechanistic Role of Li+ Dissociation Level in Aprotic Li–O2 Battery

The kinetics and thermodynamics of oxygen reduction reactions (ORR) in aprotic Li electrolyte were shown to be highly dependent on the surrounding chemical environment and electrochemical conditions. Numerous reports have demonstrated the importance of high donor number (DN) solvents for enhanced ORR, and attributed this phenomenon to the stabilizing interactions between the reduced oxygen species and the solvent molecules. We focus herein on the often overlooked effect of the Li salt used in the electrolyte solution. We show that the level of dissociation of the salt used plays a significant role in the ORR, even as important as the effect of the solvent DN. We clearly show that the salt used dictates the kinetics and thermodynamic of the ORR, and also enables control of the reduced Li2O2 morphology. By optimizing the salt composition, we have managed to demonstrate a superior ORR behavior in diglyme solutions, even when compared to the high DN DMSO solutions. Our work paves the way for optimization of various solvents with reasonable anodic and cathodic stabilities, which have so far been overlooked due to their relatively low DN.

Determining radical penetration of lipid bilayers with new lipophilic spin traps

Predicting the susceptibility of lipid moieties to radical attack requires a determination of the depth of radical penetration into a lipid membrane. We thus synthesized three homologous series of lipophilic spin traps--DMPO analogs 2-alkanoyl-2-methyl-1-pyrroline N-oxides (11) and PBN derivatives 4-alkoxyphenyl N-tert-butylnitrones (18) and 4-alkoxyphenyl N-admantylnitrones (20). The intercalation depth of these spin traps within the liposomal bilayer was determined via the previously reported NMR technique, which correlates the chemical shift and the micropolarity (measured in ET(30) units) experienced by the pivotal nitronyl carbon. Hydroxyl and alpha-hydroxyalkyl radicals were generated in the extraliposomal aqueous phase and the lowest depth at which a radical could be spin trapped was determined. The ESR data indicate that these radicals can exit the aqueous phase, penetrate the lipid bilayer past the head groups (ET(30)=63 kcal/mol) and the glycerol ester (ET(30)=52 kcal/mol), and pass down to an ET(30) polarity of at least 44 kcal/mol. The latter depth presumably corresponds to the upper portion of the lipid slab. It is likely, if not probable, that having come this far they can abstract the allylic/diallylic hydrogens resident in the midslab at ET(30) values of >31 kcal/mol.

Solvatochromic Effects in the Electronic Absorption and Nuclear Magnetic Resonance Spectra of Hypericin in Organic Solvents and in Lipid Bilayers

Abstract The natural product hypericin was tested in recent years as a biological photosensitizer with a potential for viral and cellular photodamage. We thus studied extensively its spectroscopy and membrane partitioning. Absorption, fluorescence excitation and emission spectra of the sodium salt (HyNa) were measured in 36 protic and aprotic, polar and apolar, solvents. Electronic transition bands as well as vibrational progressions were identified. Aggregation in some nonpolar solvents and protonation in organic acids were demonstrated. Modeling solvatochromism was done by Lippert equation, by the ET(30) parameter and by the Taft multiparameter approach. In all cases, separation into protic and aprotic solvents gave much better fits to the models. 13C chemical shift data could also be correlated with solvent polarity. They correlated best with Lipperts Δf polarity measure, but tended to fall into two distinct solvent groups—each along different lines—corresponding to protic and aprotic media, respectively. This interesting phenomenon suggests that in the case of the charged and slightly water soluble HyNa, two mechanisms of solvation are involved, each resulting in its own line equation. In aprotic media, dipole–dipole interaction is the predominant solvation mechanism. In protic solvents, the most effective means of solvation is likely to be hydrogen bonding. When intercalated into the liposomal phospholipid bilayer, HyNa is oriented at an angle to the interface, thus experiencing a gradient of solvent polarities: a highly polar environment (similar to methanol) for C-2/5, suggesting that they lie not far from the interface; a moderately polar environment (similar to that of n-propanol) for C-6a/14a, which are somewhat deeper within the bilayer; and a more lipophilic environment (akin to n-hexanol) for C-10/11. The fluorescence excitation peak in liposomes also correlates with an aprotic medium of relatively high polarity, as might be excepted from a molecule in a shallow position in the bilayer.

Hierarchical activated carbon microfiber (ACM) electrodes for rechargeable Li–O2 batteries

Hierarchical activated carbon microfiber (ACM) and ACM/α-MnO2 nanoparticle hybrid electrodes were fabricated for high performance rechargeable Li–O2 batteries. Various oxygen diffusion channels present in these air-cathodes were not blocked during the oxygen reduction reactions (ORR) in triglyme–LiTFSI (1 M) electrolyte solution. ACM and ACM/α-MnO2 hybrid electrodes exhibited a maximum specific capacity of 4116 mA h gc−1 and 9000 mA h gc−1, respectively, in comparison to 2100 mA h gc−1 for conventional carbon composite air-electrodes. Energy densities of these electrodes were remarkably higher than those of sulfur cathodes and the most promising lithium insertion electrodes. In addition, ACM and ACM/α-MnO2 hybrid electrodes exhibited lower charge voltages of 4.3 V and 3.75 V respectively compared to 4.5 V for conventional composite carbon electrodes. Moreover, these binder free electrodes demonstrated improved cycling performances in contrast to the carbon composite electrodes. The superior electrochemical performance of these binder free microfiber electrodes has been attributed to their extremely high surface area, hierarchical microstructure and efficient ORR catalysis by α-MnO2 nanoparticles. The results showed herein demonstrate that the air-cathode architecture is a critical factor determining the electrochemical performance of rechargeable Li–O2 batteries. This study also demonstrates the instability of ether based electrolyte solutions during oxygen reduction reactions, which is a critical problem for Li–O2 batteries.

Superoxide organic chemistry within the liposomal bilayer, part II: a correlation between location and chemistry.

Coumarin ester derivatives 1, substituted at C-4 and/or C-12 with alkyl chains, were synthesized and intercalated within DMPC liposomal bilayers. By correlating the 13C chemical shift with medium polarity [E(T)(30)], the relative location of these substrates within the liposomal bilayer was determined. The length of the alkyl chain substituents clearly influences the lipophilicity of the substrates and their location and orientation within the liposome: Superoxide readily saponifies the C-12 esteric linkage of 1, when this reaction site lies in a polar region of the liposome (E(T)(30) > 45 kcal/mol), to give the corresponding 7-hydroxy coumarin derivatives 2. However, when C-12 lies deeper and is hence less available to O(2)(*-), the lactonic carbon C-2, which lies in a shallower region (E(T)(30) = 43-49), is the preferred site for superoxide-mediated cleavage. When coumarin 1 is disubstituted with long chains at both C-12 and C-4, these derivatives lie deep within the bilayer and react only slowly with O(2)(*-). These results indicate there is indeed a correlation between location within the bilayer and substrate reactivity. Contrary to the suggestion of Dix and Aikens (Chem. Res. Toxicol.6:2-18; 1993) superoxide can penetrate deep within the liposomal bilayer. Nevertheless, its concentration drops precipitously (to approximately 16% of what it is near the interface) below E(T) values of 38, thereby precluding substantial reaction with many highly lipophilic substrates. This work also confirms the findings of others that reactions of small oxy-radicals occur within cellular membranes and appear to be of significant biological importance.

Determining radical penetration into membranes using ESR splitting constants

Determination of the depth of radical penetration into a lipid membrane is critical to the understanding of the role membranes play in radical attack. We have previously studied radical penetration into lipid bilayers using novel lipophilic spin traps and a combination of NMR and ESR techniques. We now focus on erythrocyte ghost (EG) membranes. Based on a correlation between ESR beta-H splitting constants (a(beta-H)) and solvent polarity, we have been able to locate stable radicals such as doxyls 2-4 and spin adducts 6-8 intercalated within liposomal bilayers and EG membranes. As a rule, the more lipophilic a spin adduct, the deeper it is found in the bilayer; however, the depth of penetration also depends on the steric bulk of the intercalant and whether intercalation is effected by sonication or diffusion, with the former more energetic and more effective. Compared to simple liposomes, the head group region of the red blood cell membrane is more rigid and lipophilic because of the presence of cholesterol. Hence, the biomembrane head group filters out possible intercalants that are not sufficiently lipophilic. Steric bulk plays less of a role in the EG system, perhaps because the cholesterol introduces a greater element of disorder, attenuating the role played by lipid-lipid interactions.