Enhanced Li‐O2 Battery Performance in a Binary “Liquid Teflon” and Dual Redox Mediators

Low capacity, poor rechargeability, and premature cell death are major setbacks in the operation of Li‐O2 battery, hindering its practical application. A promising approach of meeting those challenges is via the use of redox mediators (RMs), promoting Li2O2 solution phase formation upon cell discharge and an efficient oxidation on charging. The use of dual RMs decouples the electrochemical reactions at the cathode with formation/decomposition of Li2O2, resulting in improved discharge capacity, lower charge overpotential, and cycle stability. Although Li‐O2 cell performance is no longer mitigated by an insulating Li2O2, a major inherent barrier to implement viable and functioning Li‐air batteries lies in both limited O2 mass transport and pores clogging. Here, a record discharge capacity of 6 mAh cm−2 (60% increase), by combining dual RMs with “liquid Teflon” type perfluorocarbons binary system, is demonstrated. The combination of the two materials in the cell contributes to the enhanced cell performance manifested also in lower charge overpotential values throughout dozens of cycles. This is also attributed to the unique compact and an exceptionally smooth morphology of the Li2O2 deposit layers at both ends of the air cathode.

Li-O 2 battery is considered a promising energy storage device for future electric propulsion technology because of its exceptionally high theoretical specific energy of 3500 Wh kg −1 . [1,2] Successfully harvesting merely a third of the Li-O 2 theoretical specific energy could potentially deliver a battery that exceeds Li-ion technology. Li-O 2 cells usually comprise a lithium metal anode and porous carbonaceous O 2 -cathode, inner-sphere process, involving adsorption of the mediator on peroxide surface, leading to a lower charge overpotential in this advance cell, compared with PFC-free Li-O 2 cells.
Methods of analysis and measurements performed in this study as well as Li-O 2 cells construction are described in the Experimental Section (Supporting Information). In short, excess amount of LiFePO 4 was used as anode instead of lithium metal to avoid undesired reactions involving the latter. It should be noted that the potential of LiFePO 4 , 3.45 V versus Li + /Li, would not result in a practical cell voltage. A lithium anode protected by a Li + conducting membrane is required in practical Li-O 2 cells. [18,[31][32][33] A metal-free binder-free gas diffusion layer (GDL) served as the carbonaceous air cathode, while 15 × 10 −3 m DBBQ-15 × 10 −3 m TEMPO-1 m LiTFSI in tetraglyme was utilized as the electrolyte. The concentration of DBBQ and TEMPO was fixed at 15 × 10 −3 m due to the solubility of DBBQ in tetraglyme and to avoid the mediators being the rate determining factors.
In order to study the influence of dual mediators and the compatibility of DBBQ and TEMPO in the electrolyte, cyclic voltammetry (CV) tests were conducted on DBBQ, TEMPO, and DBBQ/TEMPO under Ar and O 2 , respectively (Figure 1). It is clear from Figure 1a that both the cyclic voltammograms for DBBQ and TEMPO at a gold electrode under Ar exhibit quasireversible behavior and does not affect each other. Under O 2 , DBBQ promoted the O 2 reduction and avoid the direct reduction of O 2 to Li 2 O 2 , in accord with previous studies. When TEMPO was added into the system with DBBQ, a negative shift is observed on oxidation potential, indicating a more efficient oxidation of Li 2 O 2 and a lower charge potential in the cell. The enhancement on peak current of DBBQ on reduction and TEMPO on oxidation under O 2 proves that both mediators serve as catalysts in formation and decomposition of Li 2 O 2 .
In order to investigate the contribution of PFC additive on the discharge and charge processes in terms of capacity, average operation voltage, and cycle stability, similar Li-O 2 cells with dual mediators (with and without PFC) and with PFC alone were discharged to 2.2 V and charged to 4.5 V versus Li + /Li at two different current densities of 50 and 500 µA cm −2 . The discharge load curves are presented in Figure 2 and the first cycling load curve in Figure S1 (Supporting Information).
In the absence of dual RMs, Li-O 2 cells containing the PFC additive exhibit a negligible discharge capacity and an early cell death. Cells with only dual RMs discharged under the same conditions show a typical discharge plateau around 2.7 V versus Li + /Li, a substantially higher discharge capacity and low charge potential on the order of 3.6 V versus Li + /Li, in accordance with recent observations, indicating Li 2 O 2 film suppression and a predominant solution growth mechanism. [17] When a binary system composed of both PFC and dual RMs was introduced, the Li-O 2 cell capacity at a discharge process, in a current density of 50 µA cm −2 , increases by slightly more than 60%, from 3.6 to 5.9 mAh cm −2 , while the effect was alleviated at higher current density of 500 µA cm −2 , with the capacity rising from 1.5 to 1.7 mAh cm −2 . A binary system with lithium anode protected by a Lithium SuperIonic CONductor disc (LiSICON, Ohara) was also discharged to compare the performance with LiFePO 4 ( Figure S2, Supporting Information). Considering the iR (current × resistance) loss across the LiSICON, the results are very similar to those with LiFePO 4 as the anode, indicating no shuttling effect between LiFePO 4 and the mediators. The abrupt, rather than gradual, voltage decrease observed for both dual RMs Li-O 2 systems indicates a similar failure mechanism of low oxygen diffusion rather than high resistance Li 2 O 2 layer. In nonaqueous Li-O 2 batteries, the electrolyte wets the carbonaceous electrode and floods the pores, and thus oxygen availability is determined by the concentration of O 2 dissolved in the electrolyte solution. As a result, Li 2 O 2 is mainly deposited at the electrode side, close to the oxygen reservoir, leaving a large portion of the electrode unused. Although GDL, as the porous air electrode, is an excellent gas diffusion layer electrode in aqueous medium, it fails to fulfill its role in nonaqueous electrolyte solutions, leading to substantially reduced current densities being driven from the cell. Introducing PFCs allows better oxygen distribution and more efficient air cathode utilization, as demonstrated in an extended discharge profile and enhanced capacity. However, mitigating the poor O 2 mass transport can only support a limited and quite restricted improvement in Li-O 2 battery performance.
By deliberately limiting the cell capacity to 1 mAh cm −2 at a current density of 500 µA cm −2 , a step that is being adopted previously to prevent pore clogging at the oxygen/electrode interface, [18] Li-O 2 cells cycled under 1 atm of O 2 with PFC show a clear and evident improvement in overall cycle performance, compared to the PFCfree cells, as shown in Figure 3. The discharge plateau, at ≈2.65 V versus Li + /Li, recorded during the first 10 cycles was hardly influenced by the addition of PFC, indicating that PFC additive did not affect the electrochemical reduction at the cathode surface. During the first 10 cycles, the charge voltage profile steadily increased toward a plateau of 3.7 V versus Li + /Li for both of the dual RMs cells due to oxidation of Li 2 O 2 by oxidized TEMPO, with a continuously increasing voltage between 3.7 and 4.5 V versus Li + / Li due to the oxidation of TEMPO in the absence of Li 2 O 2 at the final stage of charge, similar to the observations made in previous studies. [34] This trend is a symptom to a problem attributed to the electrochemical oxidation of carbon and/or electrolyte. [35]   During the proceeding cycles, a continuous and gradual increase on the charging potential from 3.7 V (10th cycle) to 4.1 V (25th cycle) and 4.35 V (50th cycle) was observed for the Li-O 2 cell that did not utilize PFC. Li-O 2 cells with PFC showed a much less charge overpotential during the first 50 cycles, from 3.7 V (10th cycle) to 3.75 V (25th cycle) and 4.1 V (50th cycle). Although TEMPO alone allows cycling with a reduced charging potential, the addition of PFC further improves both cycle performance and the round-trip efficiency, demonstrated by a plateau-like charging curve at a lower charging voltage. This, in turn, will potentially lead to less side products formation and accumulation at the air cathode. Lower oxidation voltage can be attributed to a more efficient oxidation of Li 2 O 2 due to its morphology and higher contact area between TEMPO and the discharge products. Nevertheless, high charging voltage of 4.1 V at the 50th cycle (Figure 3d) could not be avoided through the use of PFC, due to the accumulation of side products, associated with the reactivity of Li 2 O 2 , carbon, and the electrolyte degradation processes, leading to the increase in the electrode's resistance and thus, a rise in the overcharge potential. [36] In an effort to comprehend the morphology of the deposited Li 2 O 2 and better understand the associated mechanisms, identical cells (with or without PFC) were discharged to 2.2 V versus Li + /Li followed by an extraction of the air cathodes from the cells and their examination by a scanning electron microscopy (SEM). The SEM images presented in Figure 4 clearly demonstrate differences in Li 2 O 2 morphology as a function of PFC content and distance from the oxygen reservoir. In all cases, solution growth of Li 2 O 2 is promoted, attributed to the presence of DBBQ, showing little evidence of a film growth on the electrode surface. [37] In a PFC-free Li-O 2 cell containing dual RMs, toroidal Li 2 O 2 particles with an average diameter of 500 nm are sparsely distributed on the carbon fibers and in the pores of the electrode, whereas a denser ≈1 µm thick layer of products are intimately formed near the carbon electrode surface in the presence of PFC. As PFC has higher O 2 solubility, once DBBQ is reduced on the electrode surface, it will form Li 2 O 2 near the carbon surface, leading to the formation of a thick layer. It should be noted that in the presence of DBBQ in the electrolyte and PFC with high O 2 solubility, Li 2 O 2 formed as a thick layer near the electrode rather than on the electrode surface and the electrode surface remained clean.
Accumulation of discharge products in a dense-layer configuration is advantageous with respect to effectively utilizing the surface area for higher discharge capacity. Oxygen evolution reaction (OER) is also affected by the current density as well as the identity and morphology of the discharge products. [38,39] PFC dual RMs Li-O 2 cells potentially ease Li 2 O 2 oxidation, due to a larger contact area between the dissolved mediator and the discharge products. It is postulated that an increase in oxygen concentration due to PFC addition may have broken the delicate balance between O 2 diffusion, DBBQ, and TEMPO and their concentrations. Higher oxygen concentration, attributed to PFC introduction, may have shifted the limiting factor from mediator mass transport to O 2 mass transport. Albeit the thicklayered structure of Li 2 O 2 and its relatively efficient oxidation suggest otherwise, as it is found to be advantageous for solution-operating mediators.
Powder X-ray diffraction (PXRD) and infrared spectroscopy were obtained in order to inspect and evaluate the identity of the particles observed in Figure 4. PXRD pattern collected from the GDL after discharge with PFC exhibits peaks associated with Li 2 O 2 , as the primary discharge product and LiOH as a minor by-product ( Figure S3, Supporting Information). Although tetraglyme is considered as a relatively stable electrolyte, decomposition products (lithium carbonate, formate, and acetate) were reported. [37,40] Such products are undetectable in the XRD but clearly visible in infrared spectroscopy ( Figure S4, Supporting Information), where the presence of Li 2 O 2 is confirmed. In addition, minor by-products as Li-acetate, Li-formate, and Li 2 CO 3 are also detected. After charging, the electrodes were investigated by Fourier transform infrared (FTIR) and SEM ( Figures S4 and S5, Supporting Information, respectively). It is clear from the SEM that the thick layer disappeared after charging, suggesting the complete removal of Li 2 O 2 on charge. Some fine particles are observed on the charged electrode surface at the end of charge, which, according to FTIR, is lithium carbonate, and organic carbonate due to some side reactions between Li 2 O 2 , the electrolyte, and the electrode.
Further investigation on the impact of PFC during operation of a binary PFC dual RMs Li-O 2 cell was performed by the application of in situ differential electrochemical mass spectroscopy (DEMS), enabling the detection of gas consumption and evolution. No gas was detected, except only for O 2 during discharge in the dual RMs Li-O 2 cell with and without PFC, as shown in Figure 5 and Figure S6 (Supporting Information). The total O 2 consumed and total charge passed were calculated, and the integral provided a ratio of electrons passed per O 2 consumed of 2.04 e − /O 2 with PFC and 2.05 e − /O 2 without PFC, corresponding to a dominant two-electron reduction in both systems, leading to the formation of Li 2 O 2 . [34,41,42] Upon charging and as common in Li-O 2 cells without PFC, an evidence of both O 2 and CO 2 evolution was detected ( Figure S6, Supporting Information). During the transition from the first charging plateau at 3.7 V, corresponds to the oxidation of Li 2 O 2 by TEMPO, [34] to the second plateau above 4.0 V, a strong evolution of CO 2 was detected, associated with an oxidation of decomposition products. With the addition of PFC, it did not eliminate the side reactions between Li 2 O 2 , electrolyte, and carbon electrode, forming lithium acetate/formate and Li 2 CO 3 and subsequently did not eliminate their oxidation to CO 2 above 4.0 V ( Figure 5). However, the CO 2 evolved was halved comparing to that without PFC, suggesting a much less side reaction due to the contribution of PFC to Li 2 O 2 morphology enabling a favorable OER at lower potentials.
We studied the concept of binary dual RMs Li-O 2 cells by incorporating oxygen carrier compound, that is, PFC, soaked in the GDL air electrode. SEM images alongside XRD, infrared, and DEMS analysis demonstrate that the increase in capacity is directly governed by a compact and efficient Li 2 O 2 deposition. The formation of Li 2 O 2 via a mediated solution mechanism is promoted by DBBQ. Nevertheless, PFCs introduce new oxygen channels through the entire volume of the GDL electrode with continuous and even O 2 transportation routes. The binary combination results in a dense thick layer of Li 2 O 2 at both sides of the air electrode, as opposed to disconnected toroidal Li 2 O 2 products, clogging the pores at the O 2 /electrode interface in the absence of PFC. On charging, by introducing PFCs and changing the product morphologies, the charge voltage is reduced even further, and cycle life is improved accordingly. As the nature of oxidation process is highly influenced by the size and shape of Li 2 O 2 , new Li 2 O 2 morphology is observed upon PFC addition, promoting an efficient oxidation due to higher surface area and possibly a better adsorption of the mediator on the peroxide surface. [7] Such phenomenon leads to a lower charge overvoltage and occurs without triggering carbon decomposition, which is commonly enhanced due to intimate contact between the carbon and the discharge products.
Mediated formation and decomposition of lithium peroxide is an advantageous approach to fulfill Li-O 2 cell prospective as high specific energy power source. Nevertheless, Li-O 2 cell with dual RMs has been shown to shift the cell limitation from insulating and insoluble nature of Li 2 O 2 to O 2 mass transport and is associated with pores clogging. With the introduction of oxygen-carrier additive, namely, PFCs, the limitation of oxygen distribution throughout the entire volume cathode was mitigated, demonstrating over 60% increase in discharge capacity and improved cycling performance. The addition of the highly inert PFC compound directly to the GDL electrode supports the formation of artificial three-phase reaction zones, allowing continuous transportation of O 2 molecules to the entire electrode, without interrupting the process of mediating reduction and oxidation during discharge and charge, respectively. The improved capacity was attributed to product deposition in a denser and thicker layer of ≈1 µm, advantageous not only for dissolved RMs operation but also for an efficient utilization of the surface area for a maximal product deposition with a minimal risk of pores clogging and restricted O 2 transportation and availability.
Even though previously dismissed as a suitable cathode material due to instability, carbon as the substrate material and main constituent in the air cathode is still an attractive material when coupled with dual RMs due to improved stability attributed to solution-based mechanism. Here, we resolve the challenge of carbon-based air cathodes associated with a flooding of all air channels that prevents an effective utilization of GDL air electrode. The introduction of oxygen-carrier additive, in the form of PFC, would alter the Li 2 O 2 morphology, leading to a lower discharge/charge overpotential, also evident with a higher cycle number. Though electrolyte solution stability remains an acute challenge, the present work on the binary PFC dual RM cells, being a generic in its nature, allows the community to move a few steps closer to meet the market demand of highcapacity Li-O 2 battery with a sustained cycling. Discovery and implementation of PFCs as complementary to the operation of dual RMs can be easily implemented and adopted once stable electrolyte solutions are to be found. The search for stable electrolytes solutions is still an on-going task that eventually will allow improved Li-O 2 cell's performance and enhanced cycling stability. Once such solutions are recognized and established, systems utilizing a binary PFC-dual RMs system would provide higher capacity, superior rate, and a stable cycle performance Li-O 2 battery.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.