Expanded In Situ Aging Indicators for Lithium-Ion Batteries with a Blended NMC-LMO Electrode Cycled at Sub-Ambient Temperature

Abstract

An important step toward safer and more reliable lithium-ion battery systems is the improvement of methods for detection and characterization of battery degradation. In this work, we develop and track aging indicators over the life of 18650-format lithium-ion batteries with a blended NMC532-LMO positive electrode and graphite negative electrode. Cells are cycled until reaching 80% of their original capacity under combinations of four cycling conditions: ambient and sub-ambient temperatures (29 °C and 10 °C) and fast and mild rates (2.7 and 1.0C). Loss of lithium inventory dominates aging for all cases, with additional loss of NMC capacity under the combination of sub-ambient temperature and mild rate. A novel, easily acquired polarization factor complements capacity fade analysis; it correlates well with impedance and galvanostatic cycle life and indicates changes in active aging processes. These processes are further revealed by differential voltage analysis (DVA) and incremental capacity analysis (ICA). New indicators and aging scenarios are evaluated for these techniques and supported by post mortem analysis. From in operando cycling data and a single, slow discharge curve, these four methods (capacity fade, polarization factor, DVA, and ICA) comprise a simple, explanatory, and non-invasive toolbox for evaluating aging in lithium-ion battery systems.

Degradation in Ni-Rich LiNi1–x–yMnxCoyO2/Graphite Batteries: Impact of Charge Voltage and Ni Content

“…In the NMC811/graphite system, there are four distinct features reported at ≈3.5, 3.7, 4.0, and 4.2 V on charge . The first peak at 3.5 V likely relates to the removal of Li + ions from the graphite anode. − This is followed by the H1 to M phase conversion in the NMC cathode material and a further conversion from the M to H2 NMC structure at 3.7 and 4.0 V, respectively, Figure S3. The fourth d Q /d V peak at 4.2 V in NMC811, describes the H2 to H3 phase transition.…”

Ageing of High Energy Density Automotive Li-Ion Batteries: The Effect of Temperature and State-of-Charge

“…Test cell data already processed for ICI did not require additional smoothing for DVA. The other test cell data were smoothed using the strategy first suggested by Li et al 44 and applied in our earlier work, 45 with an initial moving average smoothing step over a 0.55% SoC span and a Gaussian filter over a 3.0% SoC span. The differential voltage parameter was obtained by numerical differentiation of the discharge curves (corresponding to the second complete discharge of each half-cell) and normalization by the discharge capacity (…”

Ageing of High Energy Density Automotive Li-ion Batteries: The Effect of Temperature and State-of-Charge

“…In the cases with higher cutoff voltages, the discharge Test cell data already processed for ICI did not require additional smoothing for DVA. The other test cell data were smoothed using the strategy first suggested by Li et al [43] and applied in our earlier work [44], with an initial moving average smoothing step over a 0.55% SoC span and a Gaussian filter over a 3.0% SoC span. The differential voltage parameter was obtained by numerical differentiation of the discharge curves (corresponding to the second complete discharge of each half-cell) and normalization by the discharge capacity (DV = 𝑄 𝑑 ⋅ dV/dQ).…”

“…The shift in the trough between peaks ➁×➊ and ➁×➋ to higher voltages is an indicator for LLI [44]. However, this might only be possible to capture by tracking intermediate stages (see MoL profiles in Figure 3) and the observation that the trough deepens between NCA reactions ➀ and ➁.…”

“…Degradation of NCA results in changes of both the position of the peaks 𝑄 𝑃𝑛 (where n is number of the peak) as well as its relative quantity. Such quantities have previously been defined and tracked over ageing for similar cells [12,44].…”