Lithium-based batteries are today’s most favourable systems to provide energy for battery-powered applications such as (Hybrid) Electric Vehicles ((H)EVs), laptops and smartphones. Generally, (commercial) Li-ion batteries are two-electrode systems. Therefore, only the battery potential or impedance can be measured across the negative and positive electrode. However, for research purposes and to design more sophisticated Battery Management Systems (BMS) it is of interest to distinguish between both electrodes by using reference electrodes (RE), making it possible to measure the electrochemical characteristics of the individual electrodes.

RE have already been introduced in many studies^1,2. However, Electrochemical Impedance Spectroscopic (EIS) measurements on three-electrode Li-ion battery systems are prone to measurement artefacts. The majority of the research on EIS measurement artefacts focuses on the cell geometry and/or the position of the RE^3,4. In the present contribution, new results are presented which show that EIS artefacts in the high frequency range can be compensated by averaging two distinctive EIS measurements. This strikingly results in artefact-free EIS measurements, especially in the high frequency range of the impedance spectra. This new method has been applied to pouch-type Li-ion batteries with integrated metallic lithium-based micro-reference electrodes^1,2.

Fig. 1 shows EIS measurements, revealing the high-frequency artefacts as well as the compensated EIS measurements. The EIS measurements of the battery (Bat), and those of the positive (P) and negative (N) electrode vs a lithium micro-reference electrode are shown in Fig. 1a. In addition, the impedance spectrum of the summation of P and N is shown (P+N), which obviously should end up with the same result as found for Bat in the entire frequency range. However, in the high frequency range a large deviation between Bat and P+N can be observed, indicating that the impedance measurements of the individual electrodes are not correct. This is because the generated excitation current induces a net voltage in the RE due to an unbalance in the measurement setup. Additional EIS measurements have been carried out with the measurement cables in a reversed connection. These results are shown in Fig. 1b and indicated with the subscript r. It can be observed that also these EIS measurements are deviating in the high frequency range but now in the opposite direction. However, averaging the measurements shown in Fig. 1a and b, results in correct impedance spectra for both the individual electrodes and added spectra, as indicated in Fig. 1c and, at a larger magnification, in Fig. 1d. It can indeed be seen that the impedance spectra of the individual electrodes are now artefact-free in the high frequency range and that the summation of P and N are now in perfect agreement with the total battery impedance in the entire frequency range.

It can be concluded that high frequency impedance measurement artefacts observed with micro-reference electrodes integrated in Li-ion batteries can perfectly be compensated by averaging two EIS measurements. This results in artefact-free impedance spectra of (commercial) three-electrode batteries without making complicated measurement setups, which can easily be facilitated by conventional electronic circuitry as will be shown in the near future.

References

1. J. Zhou and P. H. L. Notten, J. Electrochem. Soc., 151, A2173 (2004) http://jes.ecsdl.org/cgi/doi/10.1149/1.1813652.

2. E. McTurk, C. R. Birkl, M. R. Roberts, D. A. Howey, and P. G. Bruce, ECS Electrochem. Lett., 4, A145–A147 (2015) http://eel.ecsdl.org/cgi/doi/10.1149/2.0081512eel.

3. S. Klink, D. Höche, F. La Mantia, and W. Schuhmann, J. Power Sources, 240, 273–280 (2013) http://dx.doi.org/10.1016/j.jpowsour.2013.03.186.

4. S. Klink et al., Electrochem. commun., 22, 120–123 (2012) http://linkinghub.elsevier.com/retrieve/pii/S138824811200255X.