• Energy Research

Li[Ni0.5Mn0.3Co0.2]O2/graphite pouch cells were cycled using protocols that included 24 h spent at high voltage (≥ 4.3 V) under constant voltage or open circuit conditions to accelerate failure. Compared to traditional cycling, failure was reached up to 3.5 times faster. When this protocol was applied to cells containing low LiPF6 concentrations (≤ 0.4 M) failure was achieved up to 17.5 times faster than traditional cycling with normal LiPF6 concentrations. This represents a time improvement on the order of years and therefore can be used as a high-throughput screening method. Failure mechanisms for cells containing a range of LiPF6 concentrations undergoing these aggressive protocols were investigated using charge-discharge cycling, impedance spectroscopy (including symmetric cell analysis) and isothermal microcalorimetry. Long times at high voltage rapidly increase positive electrode impedance but do not seem to consume lithium inventory. The use of lower LiPF6 concentrations does not seem to introduce new failure mechanisms but makes cells less tolerant to positive electrode impedance growth. The utility of this method is demonstrated by screening cells with a variety of electrolyte additive combinations. Fewer than 3 months were required to distinguish cells containing 1% lithium difluorophospate as superior to cells with other additive combinations.

I ncreasing the energy density of lithium-ion batteries requires denser positive electrodes with higher voltage and/or capacity. Li-rich nickel-manganese-cobalt layered oxides (Li-rich NMCs), such as Li[Li 0.2 Ni 0.13 Mn 0.54 Co 0.13 ]O 2 , can deliver specific capacities above 270 mAh g-1 to reach 1,000 Wh kg-1 of specific energy at the material level. Despite their slightly lower crystalline density than today's Li-stoichiometric Ni-based layered oxides (Li NMCs and Li nickel-cobalt-aluminium oxides), Li-rich cathodes remain very promising for the long term, as we will need to move away from Ni-based towards Mn-based materials (Mn is inexpensive and environmentally benign) without compromising the energy density 1-3. The high capacity of Li-rich NMCs stems from cumulative anionic and cationic bulk redox processes 4-6. However, these electrodes currently fall short in other performance metrics 7 because of large voltage hysteresis 5,8 , sluggish kinetics 5,9,10 and gradual voltage fade 11. These issues are concomitant with anionic redox activity 7-the very same feature that enhances capacity. Therefore, further investigations are needed to fundamentally understand the overall anionic redox process, which constitutes not just electron removal from oxygen-based electronic states, but also the ensuing (local) struc-tural/bonding rearrangements 7. The undesirable issue of voltage hysteresis in rechargeable batteries leads to energy inefficiency, presumably dissipated as heat, consequently imposing an additional energy cost on the end users 12. Voltage hysteresis would also complicate the state of charge (SOC) and thermal management of such batteries. Li-rich NMCs show a relatively large difference between charge/discharge voltages (~400-500 mV after the first cycle and ~87% energy efficiency; Fig. 1). This gap persists over cycling, even at extremely low rates (C/300) 10 , at high temperatures (55 or 85 °C) 13 and after long relaxation periods (100 h) 10. Such behaviour therefore cannot be described by simple electro-chemical kinetics that fails to explain the observed path dependence and quasi-static hysteresis. Interestingly, and similar to Li-rich NMCs, many other newly discovered Li-and Na-based materials with reversible anionic redox also suffer from voltage hysteresis 7. These include layered Li 2 Ru 1-y Sn y O 3 9 (Fig. 1), Na 2/3 [Mn 1-y Mg y ]O 2 14 and Na 2 RuO 3 15 , as well as disordered Li 1.2 Mn 0.4 Ti 0.4 O 2 16 and Li 2 MnO 2 F 17. Only a few studies have attempted to understand the origin of voltage hysteresis in this class of cathodes. These include electro-chemical measurements in different voltage windows to identify correlated differential capacity (dQ/dV) peaks 5,8,11,18,19 , 6 Li nuclear magnetic resonance to observe path dependence in Li site occupation 20 , X-ray diffraction to claim back-and-forth (partially reversible) transition metal migration 21 , and bulk X-ray spectrosco-pies 5,22,23 to show the absence of hysteresis in the potentials at which transition metals show redox activity, unlike for the hysteretic oxygen redox process. A couple of phenomenological models assuming either an Li-driven phase change 10 or an asymmetry in transition metal migration 24 were also conceived. Despite such widespread efforts, the general thermodynamic mechanism behind voltage hys-teresis and its thermal effects remains unclear. Two questions need answering: (1) how exactly is the lost energy dissipated as heat so that it can be better managed/predicted? and (2) what is the underlying mechanism along with the corresponding thermochemical conditions that lead to hysteresis? In light of this, we adopt a different approach herein and perform isothermal calorimetry measurements during the cycling of a 'model' Li-rich layered cathode-Li 2 Ru 0.75 Sn 0.25 O 3 (LRSO) or Li[Li 0.33 Ru 0.5 Sn 0.17 ]O 2. This high-capacity (~250 mAh g-1 reversibly) material, although only suitable for niche applications without cost barriers (for example, space), shows remarkable structural and elec-trochemical similarities to the practically important Li-rich NMCs (Fig. 1) 9,25. Overall, it serves as a simplified 'model' compound for understanding the general properties of Li-rich layered electrodes. The commercialization of high-energy batteries with lithium-rich cathode materials exhibiting combined cationic/anionic redox processes awaits the elimination of certain practical bottlenecks. Among these, large voltage hysteresis remains the most obscure from a fundamental thermochemical perspective. Here, we study this issue by directly measuring, via isothermal calo-rimetry, the heat generated by Li/Li 2 Ru 0.75 Sn 0.25 O 3 (Li/LRSO) cells during various cycling conditions, with LRSO being a 'model' Li-rich layered cathode. We show how this heat thermodynamically relates to the lost electrical work that is crucial for practical applications. We further reveal that anionic redox on charging and discharging adopts different metastable paths having non-identical enthalpy potentials, such that the overall Li content no longer remains the unique reaction coordinate, unlike in fully path-reversible cationic redox. We elucidate how quasi-static voltage hysteresis is related to heat dissipated due to non-equilibrium entropy production. Overall, this study establishes the great benefits of isothermal calorimetry for enabling energy-efficient electrode materials in next-generation batteries.

We present a wide range of testing results on an excellent moderate-energy-density lithium-ion pouch cell chemistry to serve as benchmarks for academics and companies developing advanced lithium-ion and other "beyond lithium-ion" cell chemistries to (hopefully) exceed. These results are far superior to those that have been used by researchers modelling cell failure mechanisms and as such, these results are more representative of modern Li-ion cells and should be adopted by modellers. Up to three years of testing has been completed for some of the tests. Tests include long-term charge-discharge cycling at 20, 40 and 55°C, long-term storage at 20, 40 and 55°C, and high precision coulometry at 40°C. Several different electrolytes are considered in this LiNi0.5Mn0.3Co0.2O2/graphite chemistry, including those that can promote fast charging. The reasons for cell performance degradation and impedance growth are examined using several methods. We conclude that cells of this type should be able to power an electric vehicle for over 1.6 million kilometers (1 million miles) and last at least two decades in grid energy storage. The authors acknowledge that other cell format-dependent loss, if any, (e.g. cylindrical vs. pouch) may not be captured in these experiments.