Elsevier

Electrochimica Acta

Volume 174, 20 August 2015, Pages 488-493
Electrochimica Acta

Expanding the Operational Limits of the Single-Point Impedance Diagnostic for Internal Temperature Monitoring of Lithium-ion Batteries

https://doi.org/10.1016/j.electacta.2015.06.003Get rights and content

Highlights

  • Single-point impedance diagnostic technique demonstrated for lithium-ion batteries

  • Correlation between imaginary impedance and internal temperature determined

  • Instantaneous monitoring of commercial lithium-ion battery internal temperature

  • Expanded temperature range from −10°C up to 95°C

  • Non-invasive method useful for practical temperature monitoring of commercial cells

Abstract

Instantaneous internal temperature monitoring of a commercial 18650 LiCoO2 lithium-ion battery was performed using a single-point EIS measurement. A correlation between the imaginary impedance, –Zimag, and internal temperature at 300 Hz was developed that was independent of the battery’s state of charge. An Arrhenius-type dependence was applied, and the activation energy for SEI ionic conductivity was found to be 0.13 eV. Two separate temperature-time experiments were conducted with different sequences of temperature, and single-point impedance tests at 300 Hz were performed to validate the correlation. Limitations were observed with the upper temperature range (68°C < T < 95°C), and consequently a secondary, empirical fit was applied for this upper range to improve accuracy. Average differences between actual and fit temperatures decreased around 3-7°C for the upper range with the secondary correlation. The impedance response at this frequency corresponded to the anode/SEI layer, and the SEI is reported to be thermally stable up to around 100°C, at which point decomposition may occur leading to battery deactivation and/or total failure. It is therefore of great importance to be able to track internal battery temperatures up to this critical point of 100°C, and this work demonstrates an expansion of the single-point EIS diagnostic to these elevated temperatures.

Introduction

Temperature monitoring of lithium-ion batteries, particularly in large format battery packs, is crucial for battery management systems. Specifically, foreign object debris (FOD) and manufacturing defects are issues that can occur and lead to internal short circuits, causing self-heating and severe drops in battery voltage [1], [2], [3], [4], [5]. Overheating to high temperatures (130-200°C) can result in unwanted exothermic reactions, such as decomposition of the solid-electrolyte interphase (SEI), electrolyte and/or electrodes, which can potentially lead to separator shutdown and a catastrophic thermal runaway event [6], [7], [8], [9], [10], [11], [12]. Extremely low temperatures can be equally problematic as well, particularly during charging, where mass transport limitations through the viscous electrolyte causes plating of lithium metal and high interfacial resistances which contribute to decreased capacity retention and possible battery failure [7], [13], [14], [15], [16], [17], [18]. Typical methods used for battery temperature monitoring involve surface-mounted thermocouples, which can be an issue since the internal temperature of battery is far more critical and relevant with respect to determining onset of a failure event. During normal operation, it can be assumed that the battery temperature is relatively homogeneous [14]; however, during a rapid overheating event (resulting from a battery failure or nearby heat source) there can be a radial discrepancy as high as 40-50°C between the surface and internal temperatures [19], [20], [21], [22], [23], [24], [25]. Therefore, there is a clear need for additional diagnostic techniques to enable monitoring of the internal temperature of lithium-ion batteries to supplement surface temperature measurements in battery management systems.

One approach to internal temperature monitoring is to physically insert thermocouples into an active battery, which was successfully demonstrated by Forgez et al. [26] to develop a thermal model and obtain thermophysical properties of a safe, thermally-stable LiFePO4 lithium-ion battery [27], [28]. This method is largely impractical for commercial and military applications, however, as it is invasive, requires destructive modification of active cells, and potentially introduces new locations where the structural integrity of the cell could be compromised if exposed to extreme conditions. Consequently, non-invasive diagnostic techniques for lithium-ion batteries have recently garnered considerable attention, particularly those using electrochemical impedance spectroscopy (EIS). EIS-based analyses have traditionally been used for determining internal resistances and capacitances, thermophysical properties, and/or for modeling thermal behavior of batteries [22], [29], [30], [31]. With respect to battery diagnostics, researchers have exhibited EIS-based correlations to monitor both state of health (SOH) [32], [33], [34], [35], [36], [37] and internal temperature [10], [15], [16], [21], [37], [38], [39], [40]. These diagnostic techniques are particularly attractive for several reasons: (i) as previously-mentioned, EIS measurements are non-invasive and therefore do not require battery disassembly or physical insertion of thermocouples or other instrumentation; (ii) they are virtually instantaneous (<1 second) in comparison to the cell’s thermal response time; and (iii) they can be applied to any lithium-ion battery chemistry or cell geometry.

SOH determinations via EIS have been performed to detect abuse and/or capacity fade of commercial lithium-ion batteries through various mechanisms, including overcharge [32], [33], overdischarge [34], and aging effects [35], [36]. Among these types of battery problems, overcharging is perhaps the most concerning with regard to large format battery packs due to a battery management system’s inability to detect unbalanced battery voltages or inaccurate state of charge measurements [33]. Love and coworkers have investigated this issue at length using EIS to probe the SOH of both single- and multi-cell lithium-ion batteries and battery packs subjected to intentional overcharge abuse tests [32], [33]. Interestingly, this technique was able to detect an overcharged cell whether EIS was performed on individual batteries or on the entire pack as a whole, illustrating the promise and versatility of these types of diagnostic tests.

Internal temperature monitoring using EIS has been successfully demonstrated for a wide range of lithium-ion battery cell chemistries and geometries. The novel and pioneering work of researchers like Srinivasan [10], [20], Schmidt [21], [22], [31], Suresh [15] and others [16], [37], [38], [39], [40] over the past decade have shed light on the connections between battery internal temperature and various EIS-derived parameters, including phase angle, real and imaginary impedances, and resistance. These correlations typically take advantage of the impedance spectra corresponding to the anode SEI layer due to the critical observation that the SEI, since it stores no charge itself, is largely unaffected by the battery’s overall state of charge (SOC) [10], [15], [16], [17], [21], [29], [30], [38]. Therefore, as long as a stable SEI exists at the anode of an active lithium-ion battery, an accurate estimation of the internal temperature via EIS probing can conceivably be obtained. Decomposition of the SEI is generally reported to begin around 100°C, after which additional exothermic reactions can begin to cascade and may lead to catastrophic thermal runaway [7], [9], [11], [21], [41]. It is therefore of great importance to be able to quickly and repeatedly determine the internal temperature of a lithium-ion battery during operation in order to identify the onset of a potentially dangerous failure event.

One of the defining characteristics demonstrated by Love, Srinivasan and others in their EIS diagnostic studies is the ability to use a single point at a designated frequency, rather than requiring an EIS scan over the full range of frequencies, enabling virtually instantaneous data acquisition and diagnostics. However, for internal temperature monitoring, the upper limit of detection in these reports was only around 50-65°C, leaving the higher temperatures of the SEI’s thermal stability window (65°C < T < 100°C) undetectable. Therefore, in this study we expand upon these previous works and demonstrate a rapid, single-point, EIS-based diagnostic technique for monitoring the internal temperature of a commercial lithium-ion battery up to 95°C.

Section snippets

Experimental

For all tests, commercial 18650 LiCoO2 lithium-ion batteries were used (Tenergy Corporation) with a 3.7 V nominal voltage and 2.6 Ah capacity. The cell voltage was kept within the manufacturer’s suggested limits at all times (between 3.0 and 4.2 V). All C-rate calculations were based on the battery’s rated capacity (1C = 2.6 A), and SOC estimations were thereby determined via current counting. Multiple batteries were used for all tests to ensure accurate capacity and SOC approximations.

Results and Discussion

Nyquist plots were produced for each battery temperature (between −10°C and 95°C in 5°C increments) by averaging the real and imaginary impedances (ZR and –Zimag, respectively) for all SOCs from 0 to 100%, and these results are shown in Fig. 2A for selected temperatures with error bars representing standard deviation across all SOCs (see Supplementary Data, Figs. S2-S5 for additional temperature Nyquist plots). Typical lithium-ion battery behavior was observed in Fig. 2A: inductance in the high

Conclusions

A semi-empirical, single-point diagnostic test based on electrochemical impedance spectroscopy was developed and applied to a commercial 18650 LiCoO2 lithium-ion battery to instantaneously monitor internal temperature over a wide temperature range. Impedance spectra were obtained over the entire range of states of charge, from 0 to 100%, between temperatures of −10°C and 95°C, and a correlation between internal temperature and imaginary impedance was demonstrated which was independent of state

Acknowledgments

The authors would like to thank the Office of Naval Research (award number 40001414WX20004) for financial support of this work.

References (48)

  • E. Darcy

    J. Power Sources

    (2007)
  • S. Santhanagopalan et al.

    J. Power Sources

    (2009)
  • D. Lee et al.

    J. Power Sources

    (2012)
  • G.-H. Kim et al.

    J. Power Sources

    (2012)
  • W. Cai et al.

    J. Power Sources

    (2011)
  • N.S. Spinner et al.

    J. Power Sources

    (2015)
  • P.G. Balakrishnan et al.

    J. Power Sources

    (2006)
  • R. Srinivasan et al.

    Electrochim. Acta

    (2011)
  • S. Tippmann et al.

    J. Power Sources

    (2014)
  • S.S. Zhang et al.

    Electrochim. Acta

    (2004)
  • S.J. Drake et al.

    J. Power Sources

    (2014)
  • R. Srinivasan et al.

    J. Power Sources

    (2013)
  • J.P. Schmidt et al.

    J. Power Sources

    (2013)
  • J.P. Schmidt et al.

    J. Power Sources

    (2011)
  • D. Belov et al.

    Solid State Ionics

    (2008)
  • C. Forgez et al.

    J. Power Sources

    (2010)
  • E. Barsoukov et al.

    J. Power Sources

    (2002)
  • J.P. Schmidt et al.

    J. Power Sources

    (2011)
  • C.T. Love et al.

    J. Power Sources

    (2014)
  • M. Kassem et al.

    J. Power Sources

    (2012)
  • B. Stiaszny et al.

    J. Power Sources

    (2014)
  • J.G. Zhu et al.

    J. Power Sources

    (2015)
  • X. Lin et al.

    J. Power Sources

    (2014)
  • L.H.J. Raijmakers et al.

    J. Power Sources

    (2014)
  • Cited by (72)

    View all citing articles on Scopus
    View full text