Nina Medding 2020

Journal of Power Sources 480 (2020) 228742
Contents lists available at ScienceDirect
Journal of Power Sources

journal homepage: www.elsevier.com/locate/jpowsour
Application of electrochemical impedance spectroscopy to commercial

Li-ion cells: A review
Nina Meddings a, 1, Marco Heinrich b, 1, Frédéric Overney c, Jong-Sook Lee d, ****, Vanesa Ruiz e,
Emilio Napolitano e, Steffen Seitz b, Gareth Hinds a, Rinaldo Raccichini a, *, Miran Gaberšček f, **,
Juyeon Park a, ***
a
National Physical Laboratory, Hampton Road, Teddington, TW11 0LW, United Kingdom
b
Physikalisch-Technische Bundesanstalt (PTB), Braunschweig, D 38116, Germany
c
Federal Institute of Metrology METAS, Bern, 3003, Switzerland
d
School of Materials Science and Engineering, Chonnam National University, Gwangju, 61186, South Korea
e
European Commission, Joint Research Centre (JRC), Directorate for Energy, Transport & Climate, Energy Storage Unit, Westerduinweg 3, NL-1755 LE, Petten, the
Netherlands
f
Department of Materials Chemistry, National Institute of Chemistry, Hajdrihova 19, 1000, Ljubljana, Slovenia
H I G H L I G H T S G R A P H I C A L A B S T R A C T
• Comprehensive evaluation of EIS-based

techniques for investigation of com
mercial LIBs.
• In-depth report and analysis of a large
number of papers dealing with com
mercial LIBs.
• Critical assessment of the state-of-the-
art for EIS calibration and measurement.
• Critical assessment of the state-of-the-
art for EIS interpretation and validation.
• Critical review of both standard and
new, physics-based equivalent circuit
models.
A R T I C L E I N F O A B S T R A C T
Keywords: Electrochemical impedance spectroscopy (EIS) is a widely applied non-destructive method of characterisation of
Lithium-ion battery Li-ion batteries. Despite its ease of application, there are inherent challenges in ensuring the quality and
EIS reproducibility of the measurement, as well as reliable interpretation and validation of impedance data. Here, we
Interpretation
present a focus review summarising best metrological practice in the application of EIS to commercial Li-ion
Validation
Metrology
cells. State-of-the-art methods of EIS interpretation and validation are also reported and examined to high
Degradation light the benefits and drawbacks of the technique.
* Corresponding author.
** Corresponding author.
*** Corresponding author.
**** Corresponding author.
E-mail addresses: [email protected] (J.-S. Lee), [email protected] (R. Raccichini), [email protected] (M. Gaberšček), [email protected]
(J. Park).
1
These authors contributed equally.
https://doi.org/10.1016/j.jpowsour.2020.228742
Received 29 May 2020; Received in revised form 16 July 2020; Accepted 3 August 2020
Available online 13 September 2020
0378-7753/© 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
N. Meddings et al. Journal of Power Sources 480 (2020) 228742
1. Introduction solid state lithium ion diffusion in the active material of the cell elec
trodes [52,55], although other diffusion aspects (e.g., diffusion in
The internal impedance of a battery is an important characteristic electrolyte-filled pores within the electrodes [56,57] and concentration
that has a direct effect on its operating voltage, rate capability and ef gradients within the separator [58]) have been considered. Interpreta
ficiency, and can also affect its practical capacity. In practical terms, tion of EIS spectra is discussed in more detail in section 3.2.
impedance is the opposition to current flow through the battery, leading Identifying the relative contribution of the different features in the
to overpotentials during charge and discharge. Impedance generally impedance spectrum as a function of operating conditions can aid in the
increases as a function of cell ageing due to degradation of the electrode design and development of batteries [59–64]. However, since EIS of
materials, electrolyte, and electrical contacts within the cell [1]. Char commercial cells relies on two-electrode measurements, detailed attri
acterising impedance, therefore, is integral in defining battery opera bution of the features can be ambiguous. Similar time constants for any
tional boundaries, estimating performance and tracking state of health of the various interfacial electrochemical processes occurring at the two
(SOH) and state of function (SOF) [2,3]. electrodes will result in merged mid-frequency arcs in the Nyquist plot
Commercial Li-ion batteries (LIBs) are complex systems consisting of [65], making it difficult to distinguish between them. Incorporation of a
many components: each cell contains a cathode, anode, current collec third, (micro-)reference electrode into modified commercial cells has
tors, separator and electrolyte, as well as electrical contacts. These been reported in order to deconvolute the responses of the anode and
components each give rise to various sources of impedance, including cathode [32,66–68], but requires great care in interpretation of data,
resistive, capacitive and inductive contributions, both electrical and and is not yet employed as a mass-production solution [69]. Insights into
electrochemical in nature [1]. Additional impedance sources and effects the contribution of charge transfer, SEI and particle-to-particle re
arise when cells are combined to form modules and packs, which has sistances can be gained by examining how spectra change as a function
been the subject of several studies [4–11]. However, since the present of SOC and temperature [48,70–72] (varying the electrolyte concen
review focuses on commercial cell-level impedance, modules and packs tration can also be helpful [73], but this is only possible in lab-scale
are not discussed here. The aim of this review paper is to report and cells). It can also be helpful to correlate impedance changes with dif
evaluate the state-of-the-art in how to measure, interpret and validate ferential capacity analysis [74]. Additionally, impedance measurements
impedance of commercial Li-ion cells. We would like to point out that, in in lab-scale three-electrode half cells [53,55,60,75–87], or symmetric
this paper, the terms battery and cell are used interchangeably to two-electrode cells [32,88–92], using electrode samples harvested from
describe a single unit of commercially available battery product. commercial cells, can be used to validate and guide interpretation of
Electrochemical impedance spectroscopy (EIS) is a powerful and commercial cell measurements. This is discussed in more detail in the
widely used non-invasive test method for characterisation of LIBs [12]. Postmortem analysis section (Section 3.3.2).
As a non-destructive technique, EIS can be used at various points Since typically the impedance spectrum of a cell, and the contribu
throughout a battery’s lifetime as a diagnostic or prognostic tool: for tion of each of the different processes described above, varies as a
quality assurance [13–16]; for state estimation [17] including (i) state of function of temperature, SOC and SOH, EIS is a powerful tool for un
charge (SOC) [18–27], (ii) SOH [8,25,28–34], (iii) SOF [35]; for moni derstanding and estimating battery performance under different oper
toring of internal temperature [36–42]; and for characterisation for ating conditions [74,93–99], as required for battery management and
second-life applications [30,43–45]. control strategies [100–104]. For example, EIS measurements can be
The measurement approach typically consists of applying a sinusoi used to determine the parameters of equivalent circuit models devel
dal current (galvanostatic mode, GEIS) or voltage (potentiostatic mode, oped to describe and predict the behaviour of LIBs [27,52,57,71,85,93,
PEIS) of a certain amplitude and frequency, and measuring the ampli 103,105–117]. Note that the effects on impedance of temperature, SOC
tude and phase shift of the output voltage or current, respectively [46].
This procedure is repeated for a number of frequencies, typically in the
kHz to mHz range, thereby generating a characteristic impedance
spectrum. In this way, it is possible to probe processes with different
time constants within the battery. Features on the resulting Nyquist plot
(where the imaginary part of the impedance is plotted against the real
part) and/or Bode plot (where either magnitude and phase, or real and
imaginary parts, are plotted as a function of frequency) enable param
eterisation of the resistance, capacitance, and inductance of the various
cell components and processes [47]. Fig. 1 shows an idealised Nyquist
plot for a commercial LIB. The inductive behaviour observed in the
upper frequency range of the spectrum can be an artefact of the mea
surement system (wires connecting the cell to the measuring device)
[48,49], but may also stem from the geometry of the cell and cell
windings [50]. The high-frequency intercept with the real axis, when the
inductance effect is properly subtracted [51], corresponds to the sum of
internal ohmic resistances, including the electrolyte, active material,
current collectors and electrical (metallic) contacts. The arcs appearing
in the mid-frequency range are primarily due to the electrochemical
processes occurring at the electrode/electrolyte interfaces inside the
cell, which combine resistive and capacitive effects. At each electrode,
lithium transport through the solid electrolyte interphase (SEI) occurs in
parallel with dielectric polarisation, and lithium (de-)intercalation oc
curs in parallel with double layer (dis)charging [52]. Thus there are
contributions from at least four different processes (anode charge
transfer, anode SEI, cathode charge transfer, cathode SEI). However, Fig. 1. Graphical representation of a typical Li-ion cell EIS measurement pre
some research works suggest a possible additional contribution of the sented in a complex impedance plot (i.e., Nyquist plot). f is the frequency and,
current collector/active material contacts in commercial cells [53,54]. the order of data acquisition refers to the order in which frequencies are typi
In the context of commercial cells, the low-frequency tail mainly reflects cally interrogated.
2
and SOH can, however, be complex, and decoupling them is a challenge collect data at the low frequencies required to probe diffusional pro
for the application of EIS measurements, requiring careful design of cesses within the cell may also be seen as a drawback of conventional
experiments, measurements and analysis [13,118–121]. In addition, EIS, particularly in a commercial setting, where testing is often restricted
relationships between impedance and operating parameters may change to DC internal resistance (DCiR) measurement, despite the more limited
depending on the level of ageing, which must be accounted for in information this provides. However, fast measurement of full impedance
modelling/estimation [4,94,122]. spectra can be enabled via the use of multi-sine methods [195], whilst
Clearly, the usefulness of EIS measurements for state estimation time domain methods also offer a potential solution [196]. These
depends on the exact relationships between the impedance spectra and methods are discussed in more detail in the Measurement section (Sec
the variable in question. For some cell chemistries, such as those with tion 3.1).
LiNixMnyCozO2-based (NMC) cathodes, there is a clear relationship be Beyond EIS, another useful impedance-based method of analysis is
tween impedance and SOC, for example; but for others, particularly electrical impedance tomography (EIT). EIT is a tomographic imaging
those with LiFePO4-based (LFP) cathodes, estimation can be much more method which reconstructs the spatial resistivity distribution in an ob
challenging [123]. The same may be true for SOH [124]. Impedance ject under test using voltage-current data measured at the surface [197].
contributions from individual cell components/processes, or at partic Widely used in biomedical and other applications, EIT has recently been
ular frequencies, may show a clearer/stronger correlation to SOC, applied to LIBs [197]. Although it has lower spatial resolution compared
SOH/SOF or remaining useful life than the overall cell impedance [35, to conventional tomographic imaging techniques such as computed to
120,123,125–129,133]. Individual impedance contributions may also mography (CT), it has the advantage of higher temporal resolution
correlate more strongly than total impedance with ‘sudden death’ (i.e. [197]. An additional interesting analogue to EIS is thermal impedance
rapid capacity fade) of cells [127,130–132]. The challenge in using spectroscopy, which has been reported as a cost-efficient method for
impedance data for sudden death prediction (or avoidance), however, is determining the thermal parameters of LIBs [198–202]. However,
whether early warning signals can be identified in the measurements, in despite the intriguing scientific concepts of these latter impedance-based
addition to impedance changes that happen concurrently with rapid methods, these techniques are beyond the scope of the present review.
capacity fade. Rapid capacity fade is often preceded by lithium plating As noted above, the aim of this review is to report and evaluate the
on the anode, which has been associated with certain changes in the state-of-the-art in how to obtain, interpret and validate impedance
impedance spectra; thus detection of plating via EIS may be helpful as an measurements on commercial LIBs. We conducted a literature search to
advance indicator for sudden death [134,135]. identify papers mentioning impedance in the context of commercial
With regard to cell ageing, EIS has been widely used to study LIBs, which yielded over 250 papers in which EIS measurements on
degradation and degradation mechanisms of commercial LIBs during commercial cells are reported. These papers were analysed in terms of
cycling and calendar ageing [70,75,90,98,122,129,136–147] and to the purpose/application of the EIS measurements and the type of cell
understand the impact of different cycling/storage conditions, including being measured. Fig. 2 presents the results of this analysis. Fig. 2a shows
temperature [79,80,127,148–155], rate of cell charge/discharge that the most common applications of EIS measurements for commercial
(C-rate) [81,125,156–162], charging protocol [100,163–165], over LIBs have been to characterise and understand degradation (degradation
charge/discharge [82,166–168], SOC/depth of discharge (DOD) mechanisms and/or factors affecting ageing/degradation) and perfor
[169–173] and other storage/operating conditions [174–186], as well as mance (for example, to identify the main contributions to impedance as
manufacturing factors such as electrode misalignment [54], time a function of SOC and/or temperature). In terms of cell size and capacity
required for cell formation [187,188], cell bracing [189], and applica (see Fig. 2b and c), there has been a focus on relatively small cells, with
tion of external pressure [190]. The information gained from these over half of studies investigating cells under 3 Ah capacity, and only a
studies is clearly relevant for several aspects, including design and fifth of studies measuring EIS of cells over 20 Ah capacity. Accordingly,
development of cells and materials, battery management and control almost half of studies have investigated cylindrical cells, whilst less than
strategies, and extrapolation of accelerated ageing tests to real-world 15 % have measured prismatic cells. In line with commercial applica
conditions. Generally, the number of reference tests in a cycle or cal tion, there has been an increasing focus in recent years on cells with
endar life test should be low enough to avoid additional stress caused by NMC chemistry, which were measured in over a third of studies con
the characterisation itself, but still provide sufficient performance data ducted since 2015; over the same period, LFP cells were measured in
to allow adequate evaluation of ageing effects [172,191]. Compared to around a quarter of studies (see Fig. 2d). Fig. 2e includes an analysis of
other, destructive, techniques, EIS allows degradation to be studied as it the types of impedance measurements conducted; these are discussed in
evolves, avoids the risk of spurious results arising from sample damage more detail in the Measurement section (Section 3.1). An interactive
during cell disassembly [192], and reduces the technical effort required, table containing more detailed information for each of the studies ana
thus reducing time and costs [65]. lysed is provided as supplementary material.
Whilst indispensable for a thorough understanding of cell imped The remainder of this review is structured as follows. Section 2
ance, conventional EIS measurements require specialised equipment outlines key concepts in impedance metrology, including traceability to
[12] and are beyond the capability of most on-board battery manage the international system of units. The sources of uncertainty associated
ment systems [193], although there is also great interest in on-board with LIB impedance measurements are also discussed. Section 3 focuses
impedance measurements for state estimation [194]. Approaches for on EIS measurement, interpretation and validation in relation to com
on-board impedance spectroscopy (for cells used in an electric vehicle) mercial LIBs. It begins with a discussion of various measurement tech
are generally based on time domain methods (see Section 3.1), involving niques for obtaining EIS spectra and some of the key considerations that
either actively generated current pulses or measured current and voltage should be borne in mind when making measurements. Next, the widely
signals that result from operating the vehicle. Consequently, the fre practiced method for interpreting EIS, namely equivalent circuit
quency range is intrinsically limited to frequencies within the frequency modelling, is critically reviewed in view of the validity of the underlying
spectrum of the current transient drawn from the cell [194]. physics, and its usefulness in the context of commercial LIBs. Finally,
The requirement for specialised equipment may be seen as a disad approaches for validation, of both the measurement themselves, and the
vantage of EIS measurements. In addition, the strong dependence of conclusions drawn from them, are discussed, including a review of
impedance on temperature (and other state variables), whilst useful on complementary analytical techniques. Section 4 summarises and pro
one hand, means that control, or at least knowledge, of the conditions vides an outlook for the application of EIS to commercial LIBs.
under which measurements are taken is important. Where control is not
possible, extrapolation to other conditions may be possible via model
ling, provided the test conditions are known [31]. The time required to
3
Fig. 2. Analysis of papers reporting EIS measurements on commercial LIBs according to: (a) purpose/application of the EIS measurements (note that individual
papers may be counted in more than one category); (b) maximum cell capacity investigated; (c) cell format investigated (where a paper measured more than one cell
format, the measurement of each format is counted as a separate ‘study’); (d) cell chemistry investigated (where a paper measured more than one type of cell
chemistry, the measurement of each chemistry is counted as a separate ‘study’); (e) type of impedance measurement conducted (where a paper applied more than
type of measurement, each type is counted as a separate ‘study’.).
2. Impedance metrology and calibration between the two terminals. The two terminal (2T) impedance of the
element is then defined by the ratio Z = V/I. If the current has an
The ability of EIS measurements to effectively characterise the per angular frequency ω (i.e. I = ̂Iejωt+φI , using phasor notation with j =
formance of LIBs (as described in the Introduction) largely relies on the √̅̅̅̅̅̅̅
− 1) and the response of the impedance network is linear, the voltage
accurate measurement of the battery’s impedance over a wide frequency ̂ jωt+φV ) and the
drop has the same angular frequency (i.e. V = Ve
range. The traceability of the impedance measurement is therefore a
impedance is then given by:
prerequisite to any interpretation of EIS results.
As represented in Fig. 3a,2, metrological traceability is “the property ̂ ejωt+φV
V
of a measurement result whereby the result can be related to a reference Z= = |Z|ejφ
̂I ejωt+φI (1)
through a documented unbroken chain of calibrations, each contrib
= Re(Z) + jIm(Z) = Z’ + jZ’’
uting to the measurement uncertainty” [205]. In short, a traceable
impedance measurement ensures that the quantity is measured in the where φ = φV − φI is the phase shift between the voltage and the cur
international system of units (SI) [206]. rent. The impedance is a complex number and can be represented either
In this section, the definitions and terminology related to electrical in polar representation or by its projection to a Cartesian coordinate
impedance are first briefly introduced. Then, a critical analysis of the system, as represented in Eq. (1). When the phase shift φ is close to zero,
realisation of the different impedance scales is presented and the the impedance is mainly resistive and the current flowing though it is
different Calibration and Measurement Capabilities (CMCs) are dissipating energy. When the phase shift is close to ±90◦ , the impedance
reviewed. Finally, the influencing factors involved in uncertainty eval is mainly reactive (purely inductive if φ = 90∘ and purely capacitive if
uation of impedance measurement of LIBs are examined. φ = − 90∘ ); almost no energy is dissipated by the current flow but some
As represented in Fig. 3b, when a current I flows from terminal 1 to reactive power loss can exist.
terminal 2 of a given impedance network, a voltage drop V is generated The electromagnetic field associated with the current and the voltage
extends around the impedance element and can interact with metallic
parts or with other currents flowing nearby, making the measured
2
impedance unstable and ill-defined. To avoid such a situation, the
Even where commercial instruments are identifiable this does not imply
impedance element is generally placed in a metallic box acting as an
recommendation or endorsement by the authors, nor does it imply that the
electromagnetic shield, or Faraday cage. DC heaters should be used for
equipment identified is necessarily the best available for the purpose.
4
Fig. 3. a) Typical traceability chain of impedance measurement. The impedance meter measuring impedance at industry level has to be calibrated regularly using
reference impedance standards. Values of the reference impedance standards are determined in terms of SI units realised at the level of the National Metrology
Institute (NMI). Typical relative uncertainties required at the different steps of the traceability chain are represented on the right. b) Definition of an impedance as a
two terminal (2T), two terminal-pair (2 TP) or four terminal-pair (4 TP) standard. c) Reactance chart representing the Calibration and Measurement Capabilities
(CMCs) of NMIs. Most of the CMCs apply to impedance values located along the axis of the impedance plane (blue lines) i.e. either pure resistance, pure capacitance
or pure inductance. The green rectangle represents the impedance and frequency ranges for which traceability has been improved through the AIM QuTE project
[203]. The red rectangle highlights the impedance and frequency ranges of interest for impedance spectroscopy of Li-ion cells (LiBforSecUse project [204]). d) Two
different examples of real-world measurement setups used for impedance spectroscopy of LIBs.
temperature control. Terminals 1, 2 and the shield form what is called a terminal-pairs to the network. This leads to the so-called four terminal-
two terminal-pair (2 TP) configuration [207,208] (see the middle part in pair definition (4 TP) [207,209] of impedance standards as represented
Fig. 3b). The impedance is then given by the ratio of the voltage in the bottom part of Fig. 3b. The impedance is then defined by the ratio
measured between terminal 1 and the shield to the current flowing out of the voltage measured at terminal-pair 3 (also called the high potential
of terminal 2 and returning to the shield under the condition that the port or HP) to the current flowing out of terminal-pair 2 (also called the
voltage between terminal 2 and the shield is zero. low current port or LC) with the condition of zero voltage at
In this configuration, the measured voltage drop V is not solely due to terminal-pair 4 (also called the low potential port or LP) and no current
the impedance element but also to the contact resistances inevitably flowing through ports HP and LP. There is no specific condition applied
present between the impedance element and the measuring instruments. to terminal-pair 1 (also called the high current port or HC). Therefore,
Therefore, this definition is only suitable for the measurement of rela
V3
tively high impedance values. Z4TP = | (2)
I 2 V4 = 0
When low impedance values have to be measured or high accuracy is I4 = 0, I3 = 0
required, the impedance must be defined by adding two supplementary
5
The four terminal-pair definition can be used for any value of to the measuring instrument. A home-made test fixture may need to be
impedance. Despite being conceptually complex and requiring specific used to connect the DUT to the measuring instrument. These measuring
active circuitry to guarantee the voltage and current conditions, the four conditions are specific to each situation (see Fig. 3d) and have to be
terminal-pair definition is commonly employed in commercial imped carefully considered to avoid the measured impedance deviating
ance meters when ultimate accuracy is targeted. When the measured significantly from the impedance of the DUT.
quantity is defined in different ways (e.g., two-terminal electronic There are some basic concepts that can be followed to avoid detri
components), a carefully designed fixture or adapter becomes necessary mental effects on the measurement accuracy. The general technique that
[210,211]. A supplementary uncertainty component then has to be is most widely used is the Open/Short/Load (OSL) compensation pro
added to account for the weakness of the impedance definition. cedure [226]. This procedure allows a significant reduction of the
Depending on the measuring setup, this component could become the measurement error related to the test fixture, test leads or other addi
dominant term of the whole uncertainty budget. tional measurement accessory [227]. However, this is only possible at
the expense of having to conduct three supplementary measurements:
2.1. Realisation of impedance scales
1. open admittance measurement, Yo = 1/Zo , carried out by replacing
The primary role of impedance metrology is the realisation of the the DUT with a zero admittance standard;
different units of impedance and their associated scales: the ohm (Ω), the 2. short impedance measurement, Zs , obtained by replacing the DUT
farad (F) and the henry (H). These three units are related to each other with a zero impedance standard;
through the second (s) [206]: 1 Ω = 1 H/s = 1 s/F and therefore, one unit 3. load measurement, Zm REF , performed with a known reference imped
can be chosen as starting point for the realisation of the other two using ance standard ZREF .
different dedicated AC bridges [212].
Over time, different units have been used as the starting point [213, The effective impedance of the DUT, ZDUT , is then obtained from its
214]. Early in the 20th century, a calculable mutual inductor [215] was measured value, Zm DUT , using the following equation [226]:
used to realise the henry. Later on, the calculable capacitor was ( m
)( m )
1 − ZREF Yo ZDUT − Zs
preferred [216] since its value can be deduced from a single mechanical ZDUT = ZREF ( )( ) (3)
dimension with great accuracy. Since the 1990s, with the advent of
m m
1 − ZDUT Yo ZREF − Zs
electrical quantum standards, a new representation of the volt and the In the specific case of commercial Li-ion cells the range of interest is
ohm has been based on conventional values of the Josephson and von mainly below 0.1 Ω at frequencies lower than a few tens of kHz. Open
Klitzing constants [217]. The farad and the henry can then be deduced ⃒ ⃒ ⃒ m ⃒
compensation is therefore less critical because ⃒Zm
REF Yo and ZDUT Yo are
⃒ ⃒ ⃒
from the conventional ohm. With the redefinition [218] of the Inter
usually much less than 1. Therefore, Eq. (3) can be reduced to
national System of Units, this representation of the resistance units has
underpinned the SI realisation of the ohm since May 20th, 2019. ZREF ( m )
ZDUT = ( ) ZDUT − Zs (4)
National Metrology Institutes (NMIs) have deployed large efforts to m
ZREF − Zs
develop and improve measuring instruments and secondary impedance
standards needed to maintain the different impedance scales [219]. The and the dominant correction term is the measured short impedance Zs .
resulting CMCs of selected NMIs are represented in Fig. 3c for the A value of |Zs | significantly different from zero may indicate the
impedance and frequency domain of interest for impedance spectros presence of magnetic coupling between the different parts of the
copy of Li-ion cells. Until recently, most of the CMCs concerned measuring setup. Indeed, measurement of the low impedance of com
impedance values located along one axis of the impedance plane, i.e. mercial LIBs implies the use of a relatively large current (order of 1 A or
either pure resistance, pure capacitance or pure inductance (represented higher) flowing along the high current (HC) and low current (LC) cables.
by the blue lines), leaving a large part of the impedance plane with no The magnetic field associated with this measuring current can interact
clear and accurate traceability. with other parts of the measuring setup and can especially induce
The situation is now changing and a new generation of impedance electromotive force in the loop formed by the high potential (HP) and
bridges [220–223] and impedance standards have been developed to low potential (LP) cables. If current equalised coaxial cables [228]
improve the traceability of impedance calibration over the whole com cannot be used, the undesirable magnetic field generated by the
plex plane. Recently, within the framework of the AIM QuTE project go-and-return currents can be minimised using twisted-pair wires for the
[203], a first comparison of impedance standards having phase angle of HC and LC cables [229]. Moreover, the size of the pick-up loop can be
either ±30∘ or ±60∘ have been performed [224]. Also, a new impedance minimised by also using twisted-pair wires for the HP and LP cables.
simulator [225] has been developed to calibrate LCR-meters over the The use of Eq. (4) also implies that the measurements of Zm REF and Zs
whole complex plane in the frequency range from 50 Hz to 20 kHz and in have to be carried out using the same setup as that used for the DUT
the impedance range from 1 Ω to 10 MΩ (green rectangle in Fig. 3c). measurement, i.e. the same cables in the same geometry. If the reference
Despite these recent improvements, there is still a large range of impedance is defined as a 4 TP standard but the measuring system does
impedance values that is not covered by any CMCs, and is therefore not not allow the realisation of the 4 TP definition, correction of the refer
properly traceable to SI units, especially in the impedance and frequency ence value, ZREF , may need to be applied and a supplementary uncer
ranges of interest for impedance spectroscopy of commercial Li-ion cells. tainty component added to the reference value uncertainty. Finally, if
One of the objectives of the LiBforSecUse project [204] is to address this the geometry of the setup has to be significantly modified for the
issue and to extend impedance traceability to low impedance values at different measurements, the effect of the variation of the geometry on
low frequencies (the red rectangle in Fig. 3c). the measured value has to be evaluated and uncertainty components
added accordingly.
2.2. Measuring impedance of LIBs
3. EIS measurement, interpretation and validation
When impedance measurements are performed in an industrial
environment, the ideal conditions required to obtain the highest level of 3.1. Measurement techniques
accuracy cannot always be realised. The device under test (DUT) may be
not configured as a four terminal-pair standard and shielding may not be As discussed in the Introduction, conventional EIS measurements
possible. The measuring system itself may not be designed with coaxial involve applying a sinusoidal perturbation in either potentiostatic or
cables and excessively long cables may be required to connect the DUT galvanostatic mode. Three basic requirements of valid EIS
6
measurements are linearity, causality and stationarity. Electrochemical 0.1–1 A range, and measuring a voltage response of a few mV, is well
systems are inherently non-linear, but application of a sufficiently small within the capability of most instruments. For this reason, GEIS may give
perturbation ensures that their response remains within the pseudo- more accurate EIS measurements of commercial cells [231], though, as
linear regime. This means that the response has the same frequency as noted, PEIS has frequently been used, and has been shown to produce
the perturbation, i.e. it has negligible contributions from higher har comparable results [119,232] (Fig. 4a).
monics which are typically generated in non-linear systems; then, as in Using a higher current perturbation (relative to cell capacity) can
the time domain, the frequency-dependent impedance can be calculated improve signal to noise ratio [2]; however, care is needed to ensure that
by using Ohm’s law [47] and PEIS and GEIS should yield identical re the response remains in the linear regime, and that the internal tem
sults [230]. For testing of many electrochemical systems, including perature of the cell is not affected significantly by larger current am
lab-scale cells, PEIS is typically used, with a voltage perturbation ca. plitudes, particularly for impedance measurements at low ambient
5–25 mV (peak), as this generally ensures linearity of the current temperatures [57,93]. For example, in Ref. [93], cell impedance and cell
response [47]. Indeed, as shown in Fig. 2e, over 40% of the studies on surface temperature (as an indicator of internal temperature) were
commercial Li-ion cells identified in our literature search also used PEIS, measured as a function of applied current perturbation in preliminary
typically with a perturbation of 5–10 mV (although often not stated tests to identify a suitable perturbation value.
whether a peak, peak-to-peak or rms value). However, commercial As already mentioned in Section 2, a well-defined measurement
batteries have much lower impedance than lab-scale cells, typically in setup is important to minimise variations between measurements. In
the mΩ range [99], due to the much larger electrode areas. This means Ref. [99], for example, a variation in the cell ohmic resistance of up to
that small input voltage perturbations will lead to large output current 10% was observed over ten EIS measurements involving different cell
flows. Voltage control in such a situation tests the specification of most holders and/or removal and reinsertion of the cell into the same holder.
commercial potentiostats [231] and may result in a poor signal-to-noise As discussed in the Introduction, EIS measurements at different values of
ratio [196]. On the other hand, applying a current perturbation in the cell SOC, temperature and number of cycles can yield important
Fig. 4. (a) Comparison of impedance spectra of a commercial 8 Ah prismatic lithium iron phosphate battery obtained with different potentiostatic or galvanostatic
excitation signal amplitudes in the 1 kHz - 10 mHz frequency range at 23 ◦ C. Inductive effects at high frequencies not shown in the original figure (Reproduced from
[119]). (b) Variation with SOC of the impedance spectrum of a Kokam LiCoO2–LiMn 0.33 Ni0.33Co0.33O2||carbon pouch cell (nominal capacity 53 Ah) collected at
25 ◦ C in the 2 kHz - 10 mHz frequency range with a 10 mV perturbation (Reproduced from [233]). (c) Variation with temperature of the impedance spectrum of a
Kokam LiCoO2–LiMn 0.33 Ni0.33Co0.33O2||carbon pouch cell (nominal capacity 53 Ah) at SOC 50 % (frequency range and voltage perturbation as in (b)) (Reproduced
from [233]). (d) EIS spectra (perturbation = 20 mV) of Sanyo’s UR14500P lithium cobalt oxide battery (nominal capacity 0.8 Ah) collected at SOC 50 % and 25 ◦ C in
the 1 kHz - 100 mHz frequency range, after selected cycles at 1C. (Reproduced from [34]). Note that the x- and y-axes in b, c, d have different scales; the arcs are
much more depressed than they appear.
7
information. Variation of impedance spectra as a function of these pa precautions outlined above, the battery under test fulfils the stationarity
rameters are illustrated in Fig. 4 b, c, and d respectively, for different condition for valid impedance measurements. On the other hand,
commercial LIBs. In the case of SOC, measurements are typically taken applying a DC current during GEIS allows the impedance of the battery
by first fully charging (or discharging) the cell using a constant-current during continuous charge or discharge to be investigated, which may be
constant-voltage (CCCV) protocol, then discharging (or charging) at useful as an operando impedance technique. Impedance measured dur
constant current to a desired value of SOC, measuring EIS and repeating ing DEIS may be lower than that measured during SEIS due to the non-
this process for each of the required SOC values. However, it is worth linearity of the charge transfer reactions (i.e. the Butler-Volmer rela
noting that in some cases the measured impedance of a cell can depend tionship between current and overpotential) (see Fig. 5A) [242]. Note
on whether the previous step was a charge or discharge, as a result of that the impedance measurement itself, which is obtained from the small
open circuit voltage (OCV) hysteresis [21,234]. In order to fulfil the AC perturbation superimposed on the DC bias, can, and should, still
condition of stationarity, a sufficient rest period should be allowed fulfill the linearity condition; however, the measured impedance value
following each partial discharge for the cell to reach steady state [234, may depend non-linearly on the applied DC current. Not only the
235]. Note that even after a steady OCV has been reached (defined in magnitude of the DC current but also its direction can have an influence
terms of a maximum rate of change in potential, e.g. 5 mV/h, or on the measured impedance due to asymmetry of the charge transfer
maximum residual current value during the constant voltage step if SOC reactions. Impedance has been found to be lower during charge than
setting is achieved via a CCCV protocol), continued relaxation (e.g. of during discharge (see Fig. 5A) [74,243,244]. During dynamic EIS,
slow diffusional processes in the cell) can lead to changes in the however, the SOC of the cell will be different at the start and end of the
measured impedance spectrum [236]. The time taken for the measured measurement [95,242], which can lead to significant errors in the low
impedance to stabilise may depend on the cell chemistry and format, as frequency components of the spectrum if unaddressed. One approach is
well as the SOC, temperature, current/protocol used for SOC setting, and to compensate deviations at low frequencies through the use of a 3D
region of interest in the impedance spectrum [196,236–238]. One study Nyquist plot, including a time axis in addition to real and imaginary
of several different cell types found that a rest period of 4 h was impedance. The impedance spectrum at a specific SOC point can then be
generally adequate (subsequent changes in measured EIS spectra were estimated by interpolation [245–247]. Another approach involves
within acceptable limits) [239]; others have found that much shorter continuous measurements at a single frequency across a full SOC range
(15–90 min) [93] or longer (up to 100 h) [94,234,236,237] rest times during charge or discharge. This process is repeated at other selected
are necessary. Although it is an essential part of the measurement pro frequencies and the impedance spectrum extracted by interpolation
tocol, only half of the studies identified in our literature search stated the [243]. Alternatively, changes in SOC during the course of the mea
allowed rest period after SOC setting. Amongst these, ca. 40% allowed surement can be minimised by reducing the measurement time, either
1 h or less, ca. 40% allowed 1–4 h, and ca. 20% allowed more than 4 h. In by using a multi-sine approach (as described later in this section), or
most cases, however, the basis for the choice was not explicitly stated. simply by omitting very low frequency measurements, although this of
A rest period is also required if the temperature of the cell is adjusted, course has implications for the ability to probe slow diffusional pro
to allow the cell to reach thermal equilibrium [39,71,74,93,232,240], cesses within the cell. Note that the differences between static and dy
though this rest period may be short if a slow temperature ramp rate has namic EIS will be more pronounced at lower test temperatures (due to
been applied [238]. It is important to remember that the internal cell greater sensitivity to temperature differences generated by Joule heat
temperature, which is the relevant one, may differ from that measurable ing), whilst for temperatures above 25 ◦ C they may be negligible [95].
at the cell surface. Another consideration is the order in which different Hybrid EIS is a modified form of GEIS in which the applied current is
temperatures are applied, in order to minimise degradation-induced continuously adjusted to match a user-specified value of the measured
changes in the impedance [190]. Moreover, the adjustment of SOC is potential. At each frequency, the cell voltage is measured and the cur
affected by temperature, since the capacity of a cell depends on tem rent is adjusted until the measured voltage agrees with the desired one.
perature. Thus, if investigating impedance as a function of both SOC and For low-impedance systems, this can improve the signal-to-noise ratio of
temperature, it may be advantageous to adjust the SOC of the cell first, impedance measurements [250]. However, hybrid GEIS has not been as
and measure impedance at different temperatures for a given SOC widely applied as PEIS or GEIS to commercial LIBs, with relatively few
(before returning to the original temperature, adjusting the SOC, and studies using this method [4,94,131,172,251].
repeating this process) [57,241]. Although this procedure is more time Total harmonic distortion (THD) analysis can be used to assess any
consuming than the alternative (adjusting the temperature and non-linearities in the measured system response. The THD factor for a
measuring impedance at different SOC values for a given temperature), measured signal is the ratio of the combined root mean square (RMS)
it improves the comparability of the impedance measurements. Other magnitude of all its nth-order harmonic components (Yn with n ≥ 2),
wise, SOC adjustment can be made with reference to separate discharge which are due to non-linearities in the response, to the RMS magnitude
curves measured at each of the temperatures in question. of the fundamental frequency component (Y1 ). The THD factor increases
In terms of the EIS measurement itself, consideration should be given with the non-linearity of the response. It is calculated for each frequency
to the number and spacing of sampled frequencies required to resolve of the applied sinusoidal signal (analogously to the impedance) and can
each of the features and to estimate parameter values reliably [93]; be plotted versus the frequency of the applied signal, indicating, for a
higher sampling rates, however, particularly in the lower frequency given amplitude of applied signal, the frequency range over which
range, will increase the duration of the test. Similarly, selection of the linearity can be assumed and EIS measurements are valid [248,
number of periods measured at each frequency, as well as the delay 252–254]. Note that the degree of non-linearity may vary depending on
allowed for a periodic steady-state to be achieved after a current or the SOC of the cell [255].
voltage perturbation is applied, involves a trade-off between accuracy Indeed, Non-Linear Frequency Response Analysis (NFRA) involves
and measurement time. Thus preliminary tests are indispensable for application of relatively large currents (e.g. 1.5–3C) to deliberately
establishing appropriate measurement parameters [93,238]. excite higher harmonics in the battery’s voltage response. The resulting
GEIS may be performed at zero net current condition (also called harmonic amplitudes (usually the second and third, individually or
stationary EIS or SEIS), or with AC current perturbations superimposed combined, as these provide the largest contributions) can be analysed as
on an applied DC current (known as dynamic EIS or DEIS). Similarly, a function of frequency to obtain additional dynamic information about
PEIS can be performed at open circuit voltage or with an applied DC the cell (see Fig. 5B) [256–258]. However, the measured response is
bias. The vast majority of the studies identified in our literature search always input amplitude-dependent and results can depend strongly on
took the former measurement approach, with no DC current or potential cell type and specification [259], restricting the technique to locally
bias. The main advantage of this approach is that, subject to the comparative analysis. In contrast, Non-Linear EIS (NLEIS) applies only
8
Fig. 5. (A) (a) impedance spectra of a commercial NMC 18650 cell (nominal capacity 2.2 Ah) at different SOC, measured using SEIS and DEIS during charge and
discharge. Data collected at 20 ◦ C in the 10 kHz - 100 mHz frequency range with a 10 mA perturbation (SEIS) and in the 1 kHz - 1 Hz frequency range with IDC = 1 A,
IAC = 10 mA (DEIS). Note that the imaginary part of impedance is artificially shifted by 0.1*SOC Ω as a function of SOC, for visualisation purposes. (b) Equivalent
circuit model used to fit the spectra. (c) Comparison of the fitted parameter Rct obtained from SEIS and DEIS measurements, as a function of SOC. (Reproduced from
[243]). (B) (a) Correlation of the non-linear frequency response spectrum (left) and impedance spectrum (right) of a NMC111||graphite pouch cell (nominal capacity
0.3 Ah). (b) Intensities of the fundamental frequency response Y1 (from EIS), and higher harmonic responses (Y2 and Y3 , from NFRA) for cells aged at 25 ◦ C (panels A
and B) and at − 10 ◦ C (panels C and D). An increase in the ratio Y3 /Y2 is observed in the NFRA results for cells aged at − 10 ◦ C, which was correlated via ex situ
analysis to lithium plating. In contrast, EIS results of cells aged at − 10 ◦ C and at 25 ◦ C show similar qualitative behaviour. Data collected at 25 ◦ C in the frequency
range 10 kHz - 0.02 Hz (NFRA) and 1 MHz - 0.02 Hz (EIS) with a current perturbation of 1.5C (NFRA) and 0.16C (EIS). (Reproduced from [248]). (C) Measured
impedance spectra (Bode plot) of an A123 Systems LFP 26650 cell (nominal capacity 2.3 Ah), obtained using single-sine EIS with a commercial measurement system
vs. excitation of a low cost electronic load and a motor test rig using multisine (a) and broadband noise (b) approaches, for online applications. Data collected at
ambient temperature (18 ◦ C–21 ◦ C) in the 2 kHz–1 Hz frequency range with an applied current of IDC = 150 mA, IAC = 130 mA (peak-to-peak). (Reproduced from
[249]). (D) (a) Schematic of a cell voltage response to a DC current pulse, highlighting different contributions at different pulse times. (b) Measured impedance
modulus/‘total resistance’ of a commercial LFP||graphite pouch cell (nominal capacity 20 Ah), obtained at 25 ◦ C using different AC and DC measurement techniques
and plotted vs. timescale of the measurement, showing that values (approximately) agree when measurement timescales match. EIS measured in the frequency range
100 kHz - 0.01 Hz with a current perturbation of C/20. Pulse tests carried out with a pulse amplitude of 5C and 10 s duration. Note that ‘multisine’ here refers to a
pulsed multisine method; see ref. [2] for details. (Reproduced and adapted from [2]).
moderately larger excitation currents than conventional EIS (e.g. C/3), avoided [260], though care should be taken to avoid damaging the cell
such that the battery’s response remains in the ‘weakly non-linear’ and enhancing ageing, especially at extreme SOC values and low tem
regime (THD ≪ 1) [244,252]. In this case, the response remains peratures [259]. Sensitivity of higher harmonic responses to underlying
input-independent and higher order harmonics in the measured signal electrochemical processes can provide richer information and improve
can be used to calculate meaningful “higher order impedance” spectra, model parameter identification [252]. However, this requires more so
akin to the first order (linear) impedance, from the same experimental phisticated analysis, and a lack of experience in these relatively new
data set. However, the small amplitude of the higher order harmonics techniques means there is currently no unique, standardised method
requires very good measurement capability, and may be distorted by ology and no common implementation in commercial instru
circuitry contributions at frequencies above a certain value (e.g. 10 Hz) ments/software [259–261].
[252]. Non-linear techniques have been used to elucidate degradation As discussed in the Introduction, EIS measurements are commonly
mechanisms [252] and in the identification of lithium plating (see made using the “single-sine” method in which individual frequencies are
Fig. 5B) and SOH estimation of LIBs [248,259]. With larger permissible measured sequentially. A disadvantage of this approach is the time it
current perturbations, problems with low signal-to-noise ratio may be takes to acquire a full spectrum, particularly if probing of solid-state
9
diffusion processes within the cell is required. These processes are probed [2,94] (see Fig. 5D). Values will also depend on the amplitude of
usually very slow and thus require time-consuming measurements at the applied current pulse/step, due to non-linearity of the electro
very low frequencies <1 mHz [262]. If measurements are averaged over chemical processes, as well as increased Joule heating and changes in
several periods at each frequency, measurement times to acquire a SOC during tests with higher amplitudes [2,94,278]. Pulse power
complete spectrum of several hours - even days in extreme cases - can characterisation (PPC) tests, with specified pulse durations and ampli
result. This can bring the steady state condition of the battery under test tudes, are commonly defined in standards and manuals [279–281].
into question, and there may also be current drift in the measurement Correlation between resistance values obtained from EIS measurements
equipment [263]. One way to address this is to measure multiple fre and from PPC tests will depend on all of the aforementioned factors (see
quencies at the same time, known as the “multi-sine” method [195,264]. Fig. 5D) [2,278].
Steps must be taken to ensure similar completion times for all fre
quencies and to keep the overall amplitude of the perturbation, given by 3.2. Interpretation
the sum of the individual sine wave amplitudes, small in order to
maintain good stability and linearity [195,265]. That said, methods to Even when reliable EIS measurements can be obtained, their inter
estimate linear response functions from non-linearly excited systems pretation, in a manner that can provide useful guidance on battery
have been proposed [255]. Compared to the single-sine method, design and performance prediction, remains a significant challenge. By
signal-to-noise ratio is typically lower for multi-sine EIS since the exci far the most widely used approach is Equivalent Circuit Modelling
tation is defined for all the frequencies and therefore each frequency is (ECM), whereby the response of the cell is analytically parameterised in
less excited than in single-sine mode [266] (see Fig. 5C). Odd random terms of electrical circuit elements representing different physical and
phase EIS, a multi-sine based technique designed to allow evaluation of chemical processes in the battery. This approach can rapidly generate
the linearity, stationarity and noise level of the measured impedance, broad insights into cell performance but suffers from the limitation that
has been applied to commercial NCA electrodes [121]. such models are difficult to validate with respect to experimentally
In multi-sine EIS, the time domain response of the system is measurable parameters and often lead to unnecessary over-
measured, and then the impedance spectrum, which is the frequency parameterisation.
response function of the system, is obtained through mathematical (e.g. Impedance has also been numerically simulated from the governing
Fourier) transformation. Other time domain and wideband methods physical equations such as the Butler-Volmer equation, drift-diffusion
have also been developed [196,267] in which the excitation signal is an equations, and conservation of mass and charge [282,283]. The use of
arbitrarily shaped current (e.g. step function, rectangular pulse/ this latter approach has increased in traction in recent years with the
Gaussian/sinc pulse [196], pseudo-random binary sequence [17,32,52, help of commercial tools [252,284–286], and can help understanding
268] or broadband noise [249], see Fig. 5C) that simultaneously excites and discovery of underlying physics in LIBs, as well as supporting
multiple frequencies. Fourier, Laplace or other transformations are used development of physics-based ECMs, which can be more conveniently
to obtain the impedance spectrum from the time domain response [34, implemented for LIBs.
40,237,269]. Achieving a sufficiently high signal-to-noise ratio can be In the following sections, we will discuss some physics-based con
an issue with time domain approaches involving a current step or pulse cepts implemented in conventional ECM studies. Finally, a fully inte
[17]; however, both the pulse duration and the current amplitude grated physics-based ECM, still simplified enough for practical battery
(typically on the order of 1C) can be tuned to optimise this, with the impedance analysis, is introduced and discussed.
trade-off being limits to the measurable frequency range and the line
arity of the system, respectively [270]. A combination of frequency and 3.2.1. Distribution of relaxation times
time domain measurements, covering high and low frequency ranges Attribution of specific features in an ECM to physical quantities
respectively, may be advantageous [55,263,271]. Time domain methods associated with different parts of the LIB requires an a priori knowledge
extend to passive impedance spectroscopy, which has been proposed as of the underlying processes in the LIB. Distribution of Relaxation Times
an online diagnostic tool for electric vehicles. Passive approaches are (DRT) has been recommended as a method that assists in developing a
based on current and voltage signals that result from operating the specific ECM that is physically meaningful and is less likely to be over-
vehicle, without the need for an additional excitation; a recent summary parameterised. This method of analysis transforms the impedance data
was provided by Liebhart et al. [194]. from the frequency to the time domain via the following equation:
AC impedance measurements at a single frequency (e.g. 1 kHz) are ∫∞
g(τ)
commonly used for quality assurance and cell matching during battery Zpol (ω) = Rpol dτ (5)
0 1 + jωτ
manufacture, along with DCiR measurement [2]. Such measurements
are rapid and can identify cell-to-cell variances and deviations from
where Rpol is the total ‘polarisation’ resistance of the considered
manufacturing thresholds. Single-frequency AC measurements have also impedance spectrum, g(τ) the distribution of relaxation times, j the
been applied for internal temperature estimation and health monitoring
imaginary unit, ω the angular frequency and τ the time constant of the
[41,173,272–276]. However, the particular frequency corresponding to exponential relaxation. Polarisation is used to represent different pro
any given feature in the overall impedance spectrum will be highly
cesses leading to the overpotentials during current flow in electro
dependent on the cell specification, and therefore the frequency needs to chemical cells, i.e. ohmic, activation, and concentration polarisations.
be chosen according to the particular battery; preliminary full-spectrum
Note that Z = R/(1 +jωτ) is the Laplace transform of the voltage expo
EIS measurements can be used to guide selection [2,50]. DCiR mea
nential relaxation in an R||C circuit, where the relaxation time constant
surements involve some form of step current (applied current pulse,
τ = RC. The relaxation time τ defines the time until the effect of an
current interrupt, or current switch) and the ohmic resistance of the cell
impulse on the system has subsided to 1/e of its initial value, with e
is given by the instantaneous voltage drop: ΔV/ΔI. The measured value,
being Euler’s number. Two sequential exponential relaxation processes
however, is dependent on the data sampling rate, and in the case of an
are represented by the second order Voigt-type circuit element [287]
applied current pulse, may also be affected by the rise times of the test
consisting of two series R||C circuits shown in Fig. 6a. The trans
equipment [2,277,278]. After the initial instantaneous drop, further
formation for the real processes is shown schematically in Fig. 6b where
voltage drop is due to resistance associated with charge transfer and, at capacitors are substituted by constant phase elements (CPEs).
longer times, diffusional processes in the cell. Measurements at different
In general, the DRT method is a suitable option for electrochemical
time points/pulse durations (analogous to AC measurements at different energy storage systems presenting [262]:
frequencies) will therefore give different values for “total” cell resis
tance, depending on the extent to which those processes are being
10
1) low impedance signal regularisation parameters [301]. Therefore, accurate selection of pa
2) a large number of possible rate limiting processes rameters is the basis of a proper DRT evaluation [262,302].
3) overlap of the anode and cathode contributions Even with suitable values for the regularisation parameters, the
outcome of an LIB DRT analysis can be affected by inductive and
Although the development of the DRT technique was carried out by diffusional components. Based on Eq. (5), the analysis is often limited to
Schweidler and Wagner in the early 20th century [288,289], nowadays the impedance arcs representing the charge transfer reaction together
DRT is considered as a valuable tool for the analysis of both fuel cells with double layer capacitance and SEI films [70,292]. Recently, Danzer
[290] and LIBs [55,237,251,291,292]. Another advantage for the use of proposed an alternative strategy of analysis using a more comprehensive
DRT is the free availability and usability of tools (e.g. Matlab® [293] DRT in which inductive and diffusional impedance signals are taken into
using the Optimization toolbox [294]) such as that developed by Wan account [303]. Concerning diffusion processes, Song and Bazant intro
et al. [295]. duced the mathematical principles for the Distribution of Diffusion
In general, considering that one of the most beneficial aspects of DRT Times (DDT), which they suggested for use in conjunction with DRT
is its broad applicability (independent of the system under analysis), where there is a strong contribution from diffusion processes [304].
some researchers have specifically highlighted that erroneous mea
surement data points could have a significant impact on the resulting 3.2.2. Conventional equivalent circuit modelling
DRT, while noise has a minor effect [262,298]. Therefore, it is recom ECMs are typically applied to EIS data by interconnecting electrical
mended to delete erroneous data points prior to DRT analysis [262]. circuit elements and adjusting their parameters in order to minimise the
Improved DRT mathematical algorithms for LIBs were also investigated difference between measured and simulated impedance spectra. ECMs
by several research groups [295,299]. Another improvement has been are nowadays frequently used because they are easy to set up and have
presented by Schmidt and Ivers-Tiffée, who combined DRT with a low computational cost compared to elaborate physico-chemical models
pulse-fitting method and have been able to analyse polarisation losses at (e.g., a Newman-type model) [112,252,282,284–286,305]. Therefore,
frequencies down to 0.01 mHz [237]. they are suitable for application in Battery Management Systems
For commercial Li-ion cells, DRT is mainly used to distinguish (BMSs). Furthermore, the analysis of various states of an LIB can be
overlapping processes in order to predefine a suitable ECM. Such pro easily carried out by evaluating the change of fitting parameters due to
cesses appear as separate peaks in the DRT plot (Fig. 6a and b). The changes in, for example, the SOC, SOH or temperature. To this end, some
distinct processes derived from DRT analysis can be correlated with SOC research groups have developed SOH and temperature dependent circuit
[53,55], temperature [70,300], SOH [56] or other system variables. In elements [102,233,306]. A good overview of the techniques used to
addition, knowledge of the number of underlying processes can be used identify parameters of ECMs is given by Fleischer et al. [307]. The main
to set up advanced physics-based ECMs of LIBs [70]. However, partic drawback of ECMs, and in particular those that are most elaborate, is the
ular care should be taken because the number of processes identified by lack of the exact correlation between the circuit elements and the
the DRT algorithm highly depends on the magnitude of the physical and chemical processes occurring in the LIB. This, in spite of the
Fig. 6. a) Second order Voigt-type circuit element

consisting of two R||C circuit elements for ideally
capacitive processes and corresponding DRT where
g(t) consist of Dirac pulses. b) Two fractional order
Voigt-type circuit elements or R||Q circuit elements
for real processes with DRT as general distribution
function (reproduced and adapted from [55]). c)
Randles circuit. The Warburg element, W, is some
times alternatively positioned in series as indicated
with dashed lines. d) TLM for Warburg element where
Rd = Σrd and Cd = Σcd and terminal impedance ZT . e)
Hierarchical ladder network with SEI capacitance in
parallel with Randles circuit and additional SEI
resistance as proposed in [111,243]. f) Bisquert TLM
for the electrochemical reaction of a porous electrode
connected to Warburg element [32,251]. g) TLM for
porous electrodes where Randles circuit is distributed
together with the electrolyte resistance along the
pores [47,296,297] (the area marked by the grey
outlined rectangle represents the Bisquert TLM re
ported in f).
11
apparently excellent fit to the data, can make the interpretation of the non-ideal behaviour [28,103,111,243], which can be described by a
identified parameters ambiguous and thus the entire analysis of sur resistor in parallel with the L element [103,111,243] (see Fig. 6e) or by a
prisingly limited use. In this review, therefore, the status of ECM ap non-ideal L (using a CPE with negative α deviating from − 1 [28]). It
proaches for LIBs is systematically and critically examined. should be pointed out that the R parallel to L does not contribute to the
The basis for the ECMs used today was established by the original DC cell resistance.
work of Kohlrausch [308] and Heaviside [309] at the end of the 19th In the mid frequency region of the LIB Nyquist plot, strongly over
century. In the same era, Warburg integrated Fick’s second law and lapping arcs represent charge transfer impedance, as shown in Fig. 1
identified the 45◦ phase angle which is distinctive for the impedance of (green area) and Fig. 4 (note that the x- and y-axes in Fig. 4b, c, d have
diffusion processes [310]. The Randles circuit (see Fig. 6c), introduced different scales; the arcs are much more depressed than they appear).
in 1947 [311], is probably the most used ECM in the LIB research field. These are attributed to the combined impedance from cathode, anode
The original Randles circuit represents the charge transfer reaction at and also the respective SEI layers, each of which could be modelled as an
the metal-electrolyte interface and the diffusion of a dilute faradaic RC parallel circuit based on the ideal physical descriptions of charge
reactant in the supporting liquid electrolyte. In 1980, Ho et al. [312] transfer resistance and double layer capacitance at a planar electrode-
applied the Randles circuit to model lithium diffusion in the solid state electrolyte interface, or a thin SEI film with constant specific conduc
for the first time. Another important development for the ECM was tivity and dielectric constant. In the case of poorly conductive active
introduced in the 1960s by de Levie through the adoption of a trans particles such as LFP, the mid frequency region response is also affected
mission line model (TLM) for the charge transfer reaction at the by the presence/absence and quality of electronically conductive coat
distributed metal-electrolyte interface in a porous electrode [313,314] ings [328,329]. In order to satisfactorily fit the overlapping arcs, how
(see the grey outline rectangle in Fig. 6g). The model is often known as ever, the capacitors are typically replaced with CPEs [18] (Fig. 6b). The
the Gerischer model as Gerischer used it to represent the CPE is a non-intuitive element that cannot be given a specific physical
diffusion-reaction process in homogeneous bulk electrolyte [315]. This interpretation in any particular context. It is parameterised by a quantity
TLM has been more recently actively developed by Bisquert for finite Q with units of capacitance and a phase angle α, and expressed as an
thickness electrodes [316,317] and thus is also named the Bisquert impedance Z∗Q = 1/Q(jω)α or complex capacitance C∗Q = Q(jω)α− 1 that
model [318–320] in commercially available software for graphing and reduces to a corresponding element R, C, L, and W (infinite length) for
analysis of impedance data [321]. The concept of the CPE was intro α = 0, 1, − 1, and 1/2, respectively. Since the simulated arc of a parallel
duced by Cole and Cole in 1941 [322], originally to represent non-ideal connection of R and CPE or R||Q connection is depressed compared to
R in the Debye dielectric response. the semicircle of a R||C connection, the parallel connection of resistor
The most common circuit elements used to simulate the impedance and CPE is called ZARC or ZARC which denotes an arc-shaped impedance
spectra of LIBs are inductors (L), resistors (R), capacitors (C), CPEs (Q) trace. Note that the respective impedance arcs are described by the three
and Warburg elements (W), where the latter may be either infinite- parameters, R, Q, and α, rather than numerous R||C with distributed τ
length or finite-length with an “open” terminal ZT (see Fig. 6d) [47, values represented by Eq. (5) of DRT analysis. Models including a par
323,324]. While R, C, L, and W have well-defined electrical circuit an allel connection of resistor and CPE are called fractional order circuit
alogues, the measured impedance spectrum of a commercial LIB cannot models [330]. The distribution is generally and rather vaguely ascribed
be described using these alone [298] and many research groups have to the various inhomogeneities associated with composite battery elec
proposed modified ECMs. Firstly, capacitors are generalised as CPEs trodes, far from the idealised case of the single planar interface.
with α < 1 (see below). Secondly, the selection, and position, of the To fit the diffusion processes in the low frequency range, the War
components may be different. The state-of-the-art ECM could consist of burg element W is conventionally used. It is either placed in series with
either complex variations of the Randles circuit or TLMs for porous R||C for the surface charge transfer reaction, which corresponds to
electrodes. Indeed, since different ECMs can produce the same imped simple geometrical consideration, or connected in series to R but in
ance spectrum, choosing a more physics-based ECM is essential for an parallel to C as in the Randles circuit (Fig. 6c). The effect of this
appropriate interpretation of impedance data. In general, to minimise simplification has been analysed by Huang et al. [331]. From a physical
the risk of over-fitting, it is strongly recommended to reduce to a min point of view placing the Warburg element in parallel with the double
imum the number of circuit elements used in the ECM. More or less layer capacitance is to be preferred because diffusion following the
physics-based and often non-trivial adjustments of ECMs have been charge transfer reaction at the interface is a parallel process with the
extensively reported. Different ECMs for LIBs have been overviewed by charging and discharging of double layers. A hierarchical ladder
Osaka et al. [32], Huang et al. [103], and Choi et al. [325]. network [47] has also been suggested where the capacitance of the SEI is
In general, ECMs for LIBs are constructed considering three different placed in parallel with the Randles circuit and SEI resistance as shown in
regions of the complex impedance spectrum [172,326,327] (see also Fig. 6e, for the study of SEI deterioration [111,243], which may be a
Fig. 1). Each region is assigned to a frequency range as follows: more physical representation of the processes. However, if impedance
contributions from various processes need to be separated (e.g., to
1) High frequency range above 1 kHz: Inductive effects of the con reduce the spectrum to its polarisation effects for DRT analysis accord
necting wires (and cell windings) modelled with an inductance (and ing to Eq. (5)), Voigt-type modelling of the respective processes (Fig. 6b)
often with a parallel resistor, see below) and a series resistor repre and placing the Warburg element in series is recommended [332]. Some
senting the resistance of electrolyte and electrodes i.e. (L||R)Rohm research groups generalise the Warburg element as a CPE with α
(see Fig. 6e). different from − 1/2, because the measured slope of the diffusion
2) Mid frequency range: Series connections of parallel networks of impedance often deviates from the ideal inclination angle of 45∘ . It
resistor and capacitor (or CPE) used to model charge transfer re should be noted, however, that the low frequency diffusion impedance
actions and double layer effects at anodes, cathodes and solid elec “tail” in the Nyquist plot often exhibits an inclination higher than 45∘
trolyte interphases (SEIs), which are straightforwardly related to (see Fig. 1), which cannot be described by CPE generalisation if a TLM is
DRT analysis (Fig. 6a and b). adopted. Oldenburger et al. [234] have explained this behaviour by the
3) Low frequency range: Diffusion processes modelled with Warburg effect of the hysteresis of the OCV curve on the differential capacity (see
elements, W, which in connection to the mid frequency responses below), which in turn affects the slope of the diffusion impedance. Other
modelled as R||C or R||Q, have two different variations as shown in physical explanations have been given by stating that the slope is
Fig. 6c. affected by anisotropic diffusion in the particles and by the particle size
distribution [333,334].
It has been noted that high frequency inductive effects often exhibit a
12
It should be noted that the Randles circuit and its variation with the components are recommended and they should be applied to fit the
Warburg diffusion element in series assumes charge transfer at a planar experimental data properly. However, generalisation of the capacitive
interface with one-dimensional diffusion into a homogeneous bulk. The components into CPEs, with arbitrary adjustment of α values, often al
one-dimensional diffusion (d) can be equivalently represented by an RC lows different ECMs to describe the experimental data equally well but
TLM, shown in Fig. 6d, where Rd = Σrd and Cd = Σcd , due to the analogy with different sets of parameters. A common model can also yield totally
between Fick’s 2nd law for one-dimensional diffusion and the telegra different fit results from automated multi-variable complex nonlinear
pher’s equations [309]. fitting, depending on the initial values, weighting methods and opti
However, in LIBs, the charge transfer reaction occurs at the surface of misation algorithms [321,338–340].
active material particles dispersed in the electrode layers, connected by
conducting binders, with electrolyte permeating through the pores. 3.2.3. Advanced physics-based modelling
Ionic transport through the electrode pores and also possibly electronic As pointed out in the previous section, in spite of the excellent fit of
transport through the solid network, with resistance rm (where ‘m’ many ECMs with experimental data, the results can be poorly utilised or
identifies the medium), giving a total resistance Rm = Σrm , causes dubiously interpreted due to over-parameterisation and the presence of
additional, non-negligible resistance with respect to the ohmic resis multiple CPEs. This aspect is a major drawback of rather phenomeno
tance of the electrolyte in the separator. This is accompanied by the logical ECMs in comparison with more fundamental physics-based
charge transfer reaction at the individual particles, represented by models, such as pseudo-two-dimensional (P2D) LIB models where the
rct ||cdl , of a total Rct = (Σrct
− 1
− 1
) and Cdl = Σcdl . Current passes through battery impedance response is numerically simulated using physico
the thickness of the electrode layer in a hierarchical manner, as repre chemical parameters in the governing equations, i.e. Butler-Volmer
sented in Fig. 6g (grey rectangle) which is known as the Bisquert model equation and drift-diffusion equations in different phases under mass
[316,321] or finite-length Gerischer model. The impedance response is and charge conservation conditions [252,282–286,305]. The essential
characterised by a slope-one high frequency trace, as if an Rm Cdl TLM of physics of these governing equations have been implemented in the ECM
√̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ of Fig. 6g, which was developed for LIBs almost two decades ago by
Z = 1/ jω⋅ Rm /Cdl , and a semi-circular low frequency response as if
Barsoukov et al. [47,105,296,341], (see Fig. 7a). In short, the Bisquert
Rct ||Cdl . The magnitude of the skewed impedance arc is
√̅̅̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ TLM for porous electrodes is combined with charging/discharging of
Rm Rct coth Rm /Rct . With Rm →0 the response becomes Rct ||Cdl (see
three-dimensionally dispersed particles in the manner of a Randles cir
Section 3.2.3). The individual particles may be considered to represent cuit. An in-depth explanation and discussion of the model is reported in
the discrete components in the ladder network when the numbers are the book edited by Barsoukov and Macdonald [47] as well as in the
sufficiently large, the response of which approaches the continuum limit manual of the “Multiple Electrochemical Impedance Spectra Parame
represented by the telegrapher’s equation. In common practice for TLMs terization” or MEISP program [339]. The MEISP program allows users to
for porous electrodes [32,251,335,336], Cdl is replaced by Qdl to modify or build models using analytical equations. The Barsoukov
describe the depressed semicircular low frequency arcs (from Rct ||Qdl ), model for spherical diffusion and some modified forms have been
as shown in Fig. 4. (It should be noted that Fig. 4b, c, d are not plotted recently implemented in commercial software for graphing and analysis
with isotropic axis scales.). With this modification, modelling predicts of impedance data such as ZView® (DX-28, 29 and, 30 models) [321].
√̅̅̅̅̅̅̅̅̅̅̅ √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
that the high frequency slope is reduced to Z = 1/ (jω)α ⋅ Rm /Qdl Both programs are based on Fortran-coded LEVM [338] by J. R. Mac
[337]. Due to the overlapping inductive effects, this feature cannot be donald. Recently, a Python-based analyzer, PyPhyEIS, has been devel
directly observed, and thus this model may not be supported by exper oped by T. L. Pham [340]. Unfortunately, the use of this advanced
imental data. physics-based ECM by the battery research community is still very
Although the Bisquert TLM is used for the electrochemical reaction of limited. This situation may be due to the widespread acceptance of the
LIBs [32,251,335], diffusion processes described in the Randles circuit customary practice for ECMs, as critically reviewed in the previous
are not fully incorporated. Rather, the Warburg element for section.
one-dimensional diffusion is placed outside the Randles circuit in series It is to be noted that the Barsoukov model [47,296,339] has theo
with the TLM, as for the one-dimensional diffusion into the bulk elec retical and practical limitations related to the generality of the detailed
trode, as shown in Fig. 6f; this is inconsistent. Randles circuits microscopic parameters in the numerical simulations [252,282–286,
combining solid-state diffusion with charge transfer should be posi 305]. Nevertheless, it integrates important LIB physics, i.e., ion-binding
tioned as distributed elements as shown in Fig. 6g [297,336]. Incorpo or trapping, or intercalation, representing the rate-limiting process in
rating the one-dimensional Warburg element into the TLM can be the solid state charge storage [342,343] (Barsoukov et al. named this a
applicable when the diffusion in the active particles is essentially ‘new phase formation’ process, which may cause some confusion in view
one-dimensional due to the particle geometry or crystallographic of materials science for battery materials [344]). This process can be
anisotropy [333,334]. More generally, three-dimensional diffusion into incorporated in the diffusion equation [342], but has not been consid
spherical active material particles should be considered. This will be ered in physics-based battery simulations so far [252,282–286,305]. The
discussed in detail in the Section 3.2.3. physics can, however, be succinctly represented by ECM as indicated by
It should be noted that the apparently excellent description of the zss in Fig. 7a. The implementation is to divide the overall differential
data by ECMs, regardless of the different circuit configurations, is due to battery capacity into cd representing faster diffusion and ci representing
the ubiquitous replacement of capacitors by CPEs, including those in a slower crystallisation or binding reaction with relaxation time τi =
TLMs. The α parameters in CPEs represent frequency dispersion over a ri ci , where in general in LIBs, ci ≫cd . Simulated results from this
wide frequency range and thus multiple CPEs with arbitrarily adjustable physics-based ECM are shown in Fig. 7c. The left panel represents the
α can describe experimental data almost perfectly, and do so equally well-known diffusion impedance for the different geometries shown in
well for somewhat different positioning of the elements, as long as the Fig. 7b [333,334,345] and the right panel shows the effect of including
data are Kramers-Kronig compliant (see Section 3.3.1). Thus, whilst the solid state reaction where the low frequency limiting capacitance
capacitance effects differing by several orders of magnitude allow the Cd + Ci corresponds to the differential capacity of the LIB as computed
respective processes to be deconvoluted in EIS of commercial LIBs, from the cell open-circuit voltage, V vs. SOC. As data in the low fre
capacitive parameters cannot be unequivocally derived from models quency limit are generally difficult to obtain using EIS, Barsoukov et al.
using CPEs. [47] suggested using the capacitance values from V vs. SOC as a fixed
The variety of ECMs used for LIBs can thus be summarised as parameter for ECM analysis. Such comprehensive interpretation of the
different connections of electrical elements which have respective EIS spectra can be beneficial for the detection of specific features in the
physical meanings. Physics-based arrangements or connections of these case of electrochemical testing and measurements of commercial LIBs.
13
Fig. 7. (a) Top, graphical representation of spherical active material particles distributed within a porous battery electrode where the blue and grey colours
schematically represent the network of electrolyte pores connected to the separator layer and electronically conducting network connected to the current collector.
Bottom, Barsoukov TLM [47,105,296,341] with addition of Warburg element as the terminal impedance, ZT = W, highlight green, of the parameter QW , which is CPE
1
coefficient with α = 1/2[319,320]. ECM for drift-diffusion in liquid electrolyte is also shown where R−ohm = R−+1 + R−− 1 . (b) Active material particle geometry for
3-dimensional (blue), 2-dimensional (red), and 1-dimensional (black outline) diffusion solution where r and l represent the respective diffusion length. (c) Complex
impedance spectra and capacitance Bode plots for solid state reaction, Zss for diffusion only (left column) and for diffusion-intercalation (right column) for spherical
(blue), cylindrical (red) and planar geometry (black) for a given amount of active material in the electrode. TLM for 1-dimensional case is indicated. (d) Full spectra
of Barsoukov battery impedance model for different diffusion geometry. (e) Full spectra as (d) with the terminal Warburg impedance and diffusion in separator. For
(d) and (e) top row presents magnified high frequency region. In (c), (d), and (e) the characteristic capacitive parameters are indicated. Adapted from [340].
Some lack of clarity [333,334] or confusion [47,296,339] exists in conductivity and separator geometry, and Rm depends on liquid elec
ECM parameterisation for different geometries of active material parti trolyte conductivity, porosity, and possibly the electronic network in the
cles. In a recent Python algorithm, PyPHyEIS [340], the impedance electrode composites [47,296,339]. Some of these properties are either
function is formulated so that the specific material quantities Cd and Ci unrelated or only indirectly related to the active materials. With the
in F/kg and Ri in Ω⋅kg, and thereby the derived diffusivity and inter common particle diffusion dimension (l or r) as shown in Fig. 7c, the
calation reaction constants, D = l2 /(Rd Cd ) or r2 /(Rd Cd ) and τi = Ri Ci , ratio of the specific surface area (cm2/g) is 1:2:3 for the planar, cylin
are consistently parameterised in a physically meaningful way for all drical, and spherical geometry, respectively, and thus correspondingly
diffusion geometries, as shown in Fig. 7c. varying values of Cdl and Rct (as 3:2:1 for the latter) can be used in the
Fig. 7d shows the response of the full Barsoukov TLM, where the simulation.
Bisquert TLM of the components rm , rct , and cdl , also indicated in Fig. 6f, As discussed in the previous section, for a finite medium resistance
is combined with the solid state processes zss of Fig. 7c. Note that the Rm , the mid-frequency impedance response exhibits a high frequency
impedance is given using specific gravimetric units to enable compari inclination of 45◦ but a semicircular low frequency trace, attributed to
son with commercial cells, although Rd , Rct , and Cdl depend on the size Cdl in the TLM of Fig. 7a. The presence of an ideal Cdl is indicated by the
of the particles (l or r), Rohm depends on the liquid electrolyte plateau (circled) in the capacitance Bode plot in Fig. 7d, bottom. (The
14
plateau magnitude is also dependent on Rm , Rct , Rohm [340].) As already general validity [349]. Nowadays, the following equations are used by
mentioned, however, real EIS responses recorded for commercial LIBs, many research groups and are widely referred to as the Kramers-Kronig
as in Fig. 1, exhibit often skewed and strongly depressed semicircular relations:
traces. This may be one of the reasons why the Barsoukov model has not ∫
2 ∞ xZ’’(x) − ωZ’’(ω)
been widely utilised since the earlier work [296]. Only recently was the Z’(ω) = Z’(∞) + dx (6)
π 0 x2 − ω2
model applied to interpret EIS measurements of modified commercial
pouch cells containing a lithiated lithium titanate reference electrode
and
[68]; the goodness-of-fit using an ideal Cdl was shown to be rather poor, ∫
compared to the conventional analyses using CPEs.
∞
2ω Z’(x) − Z’’(ω)
Z’’(ω) = − dx (7)
A modified Bisquert TLM has been proposed [319], which explains π 0 x2 − ω2
the high frequency 45◦ inclination, as observed clearly in experiment
The equations shown above can be used to calculate the real part of
[319] and requiring a well-defined Cdl component in the TLM, alongside
the impedance from its imaginary part and vice versa but they can also
the generally observed depressed feature. The depressed feature can be
be formulated in terms of the magnitude of the impedance and its phase.
simulated by introducing a terminal Warburg impedance ZT = W,
In both cases the calculated values can be compared to the measured
highlighted green in Fig. 7a. This interfacial impedance plays a unique
ones and the degree of convergence describes the quality of the mea
role as it is connected in series with Rm but in parallel with Rct and Cdl
surement. Barsoukov and Macdonald pointed out that measuring the
[316]. A Warburg-like interfacial impedance is suggested to be a generic
magnitude of the impedance and calculating its phase is easier and
feature of many electrochemical cells [319,320]. The physics associated
therefore possible with cheaper measurement devices [47]. Since
with this component should be further clarified. For example, it has
impedance spectra cannot be measured from 0 Hz to ∞ Hz direct use of
recently been proposed that instead of using W, one may add an addi
equations (6) and (7) is not recommended, but a calculation using direct
tional parallel current path with resistive elements that represent
and inverse Fourier transforms has been presented [350]. Usually, the
migration of non-active (i.e. non-lithium) mobile ions [346]. These
fact that simple ECMs fulfil the requirements of the Kramers-Kronig
resistive elements are then coupled to the active ions through an
relations is sufficient to avoid the issue of incomplete measurement
appropriate chemical capacitance of the liquid electrolyte, as indicated
data. If such a model can be fitted to experimental data, the data are
by r+ , r− and c± for the electrolyte in Fig. 7b. Thus, the physical meaning
assumed to be validated in terms of the Kramers-Kronig relations. This
of W may be linked to the coupled transport of active and non-active ions
approach, known as the linear Kramers-Kronig transform, was intro
in the electrode pores. It has been reported that the diffusion of the
duced by Agarwal and Orazem [351] and has been extended by Bou
electrolyte in the separator layer [346], due to a large capacitance c±
kamp who added inductive and capacitive behaviour to the model
(originating from the large amount of electrolyte contained within the
[352]. More recently, the linear Kramers-Kronig transform has been
separator), could overlap the low frequency response of
refined in order to avoid ambiguities due to under- and over- fitting
diffusion-intercalation in the solid state [286]. In Fig. 7e, compared to
[353]. In the same publication, the authors present a short derivation of
Fig. 7d, the low frequency diffusion and intercalation impedance of solid
the fundamental equations and describe how a suitable model order can
particles are shown to be substantially affected by the diffusion
be determined using a given impedance spectrum. Based on this work
impedance of the solution, indicating further complications in evalu
the linear Kramers-Kronig test has been automated and improved by
ating diffusivity from EIS of LIBs.
Lohmann et al. [269]. In research on impedance spectroscopy on com
The model with up to 11 parameters indicated in Fig. 7a is presently
mercial LIBs the Kramers-Kronig relations have only been applied in a
available for simulation and fitting in the PyPhyEIS GitHub platform
minority of publications until now, but their importance has been
[340]. It should be emphasised that this ECM succinctly and correctly
increasing since the advent of DRT, which is used in an increasing
represents underlying governing equations used in physics-based sim
number of studies and requires excellent data quality [55]. The devia
ulations. Physics underlying the intercalation process can be straight
tion between measured and calculated impedance should be well below
forwardly included in the ECM, even those not yet considered before in
1% [302] which is only possible by careful selection of amplitude for the
the Newman-type simulations.
sinusoidal excitation signal as a trade-off between good signal-to-noise
Overall, the physics-based Barsoukov TLM model is characterised by
ratio and approximately linear excitation of the non-linear system [354].
additive capacitance effects (Ci + Cd + Cdl ). Note that the Cdl contribu
tion may be especially significant in nanostructured active materials.
3.3.2. Postmortem analysis
The situation should be clearly distinguished from the additive imped
As discussed in the Introduction section, since EIS of commercial
ance effects underlying DRT analysis and Voigt-type models (Fig. 6a and
cells relies on two-electrode measurements, the interpretation of
b). Although the impedance spectra in Fig. 7d and e appear to be seg
impedance spectra in terms of the underlying contributions from
mentable into different processes, such segments do not exactly corre
particular cell components and processes can be ambiguous. Measure
spond to any R parameters of the ECM in Fig. 7a. Limiting values as
ments in lab-scale cells, using electrode samples harvested from com
indicated by Rm+ct , Rm+ct+d and Rm+ct+d+i in Fig. 7d can be found by
mercial cells, can be used to validate and guide interpretation of
simulation and are also given by somewhat complex analytical expres
commercial cell measurements. Harvesting of electrode samples can be
sions of R and also C parameters [340]. The individual parameters can
done before or after cell cycling/storage, and in the latter case, can be
be extracted by fitting and used to quantify specific LIB characteristics.
used to investigate ageing mechanisms, and validate conclusions about
these mechanisms drawn from EIS measurements. Ageing of Li-ion cells
3.3. Validation is a complex process which involves changes in battery parameters
related to their usage (cycling ageing) and/or storage (calendar ageing)
3.3.1. Kramers-Kronig relations which can be evaluated through EIS measurements. The deterioration of
Reliable interpretation of impedance measurement data is only the battery performance can be mainly attributed to the chemical re
possible if the data can be validated. Since EIS requires linearity, cau actions happening in the various battery components [355] and, despite
sality and stationarity, these conditions can be used to formulate its complexity, battery degradation has two primary effects at the
mathematical equations that can be applied to the validation of macroscale: capacity fade (e.g. due to the loss of active material able to
impedance data. Kramers and Kronig derived equations which describe store Li) and increased internal resistance/power fading (e.g. due to
the correlation between the real and imaginary parts of the impedance decreased particle-to-particle contact, growth of surface passivating
[347,348]. These first publications on this topic have been reviewed films, delamination of active materials from current collectors) [65]. To
with regard to whether or not the equations were intended to have
15
gain useful insights into Li-ion cell degradation and to validate the procedure is especially relevant for the comparison of “fresh” and
interpretation of EIS measurements, several physico-chemical methods “cycled” samples and the application of a relatively slow C-rate (e.g.
of analysis of battery materials obtained from aged cells can be C/20) is generally recommended to ensure comparable sample condi
employed. In this section we will focus only on the most relevant tech tions. Cell cutting can be carried out in various ways depending on the
niques used to validate the interpretation of EIS measurements; for a type of cell design; cylindrical cells can be opened by using a pipe tube
comprehensive review on postmortem analyses of LIBs we suggest the cutter, a carbide-tip saw tool [360] or, in the case of a more precise cut, a
article by Waldmann et al. [356]. lathe tool [361]. For soft cases such as pouch cells, ceramic scissors are
In order to perform reliable postmortem analysis of battery mate sufficient. However, for hard prismatic cases, drilling might even be
rials, particular attention should be paid to the procedure for sampling required [362] although technical challenges and risk for the operator(s)
materials from disassembled cells to obtain representative and reliable can arise. It is important in any case to entirely avoid short-circuiting
results. The various steps generally needed to carry out an exhaustive and bending/damaging of the cell to ensure the integrity of the har
postmortem analysis are reported in the flowchart of Fig. 8a. It should be vested electrodes. To achieve this, the use of specially designed tools is
pointed out that, when referring to the analysis of aged cells, an anal advised; a description of an in-house built tool developed for this pur
ogous assessment must be performed on fresh cells in order to establish a pose has been presented by Pfrang et al. [361]. Information obtained by
baseline for comparison purposes [82]. means of e.g. XCT as mentioned previously is very useful to avoid
short-circuiting the cell during disassembly. Opening or disassembling
3.3.2.1. Cell inspection and disassembly. In general, after long-term of the cell needs to be carried out under an inert environment to mini
ageing of an LIB, visual inspection of the specimen is recommended, mise potential safety risks and contamination. Typically an argon-filled
to observe possible mechanical deformation of the cell enclosure which glove box is necessary for this, to protect cell components from exposure
could be attributed to the generation of excessive amounts of gas and to O2, N2 and H2O.
growth of solid side-products during cycling. Extraction of cell components from their enclosure needs to be car
Although a small electrolyte leakage might be undetectable by mass ried out with care to avoid cross-contamination, mechanical damage,
loss measurements, visual inspection of the cell could give an indication contact with sharp edges, or contamination with metallic debris or
of electrolyte leakage as it can lead to corrosion of the cell casing or fragments generated during the cutting. The welding point of the tabs to
other components. In case of leakage, it is generally not recommended to the battery case may need to be cut with insulated tools (e.g. ceramic) or
disassemble the cell, but if deemed necessary, since leakage may evi to be removed manually. Following this step, anode, separator and
dence phenomena of interest, this should be done with extreme care by a cathode can be separated and harvested. To extract the electrolyte, the
trained (or authorised) operator to avoid any safety incidents, as the electrode can be immersed in a specific organic solvent (e.g., CH2Cl2)
presence of leakage elevates the risk of the procedure. Non-destructive [363]. This step must be done quickly as the solvents contained in the
investigations, such as X-ray computed tomography (XCT) and electrolyte are highly volatile.
neutron diffraction (ND) tomography, can be employed to reveal A washing step can also be considered in the case of electrode (anode
essential information regarding the cell design, position of safety vents, and cathode) and separator harvesting by using specific solvents (e.g.,
thickness of various layers, signs of internal failure, thickness of the dimethyl carbonate, DMC, or diethyl carbonate, DEC). A critical aspect
casing and cutting locations for cell opening [357,358]. It should be to ensure repeatability and comparability of postmortem analysis is the
mentioned that alternative non-destructive electrochemical methods, reporting procedure for the washing step which includes, e.g., the
such as capacity measurement, cyclic voltammetry and differential ca amount and type of solvent used, time of exposure applied, temperature.
pacity analysis [359] can also be used to gain knowledge about ageing Somerville et al. have found that washing with DMC removes not only
mechanisms and to validate EIS measurements. the electrolyte salt but also SEI components on the electrode surface
Prior to cell cutting, a discharge step is required (usually to 0% SOC [364]. Therefore, the need for a washing step should be evaluated
to ensure personnel safety). It is also important that the cell voltage carefully on a case by case basis. After washing, the drying procedure
remains inside the limits of standard operation (identified by the cell also needs to be reported (e.g. drying in the glove box antechamber or
manufacturer) to avoid undesired material changes beyond those solely vacuum drying in an oven).
assigned to ageing [356]. The exact definition of the discharge Certain techniques require additional steps in terms of sample
Fig. 8. a) Flow chart representation for postmortem analysis of Li-ion cells; b) Physico-chemical techniques used for postmortem analysis of LIB cells (Reproduced
from [356]); c) Schematic representation of a Li-ion pouch cell disassembly and symmetric cell reconstruction (Reproduced from [88]).
16
preparation; in some cases the “intact” electrodes (i.e. electrode samples A Cu sealed anode is the most common X-ray source used for general
directly extracted from the cells without any further manipulation) are materials characterisation. However, this source is not the best choice
sufficient, but in other cases the material should be collected by me for battery electrode materials [43,372–375]. Mn, Fe, Co and Ni are the
chanical abrasion of the electrode surface. For example, when per main chemical constituents of cathode materials in Li-ion cells, and their
forming electrochemical testing on harvested electrodes, usually interaction with Cu X-ray photons generates low quality diffraction
referred to as lab-scale cell reconstruction, the electrode needs to be cut patterns due to absorption and fluorescence issues [371,376], which
to an appropriate size/diameter, and one side of the coating needs to be results in several side effects such as: (i) high background noise, (ii)
removed (as commercial electrodes are double-side coated) using, for diminished X-ray penetration into the bulk and (iii) reduced
example, N-methyl-2-pyrrolidone (NMP) [154] and water for cathode peak-to-background ratio and detection limit. Limited improvements to
and anode respectively, by gently rubbing with a cotton bud. It is the collection of diffraction patterns can be obtained, e.g. by using a
important to highlight that, since degradation processes do not occur graphite monochromator or discriminating energy detectors in the path
uniformly through the whole electrode foil, multiple sections of the cell of the diffracted beam. Both approaches can only filter parasitic fluo
need to be harvested to obtain meaningful and representative data. An rescence and select X-ray diffracted photons reaching the detector.
example of this multiple sampling procedure is reported by Klett et al. However, the interaction between Cu X-ray photons and the sample is
[365]. where non-uniform electrode ageing (mainly of the graphite unchanged and so difficulties related to phase identification and results
electrodes) was observed, showing clear physical differences (even produced by Rietveld refinement [377,378] arise with a Cu-based X-ray
visible to the naked eye) between the central parts of the electrode source [376,379,380]. To solve this issue, Co, Mo and, Ag sealed tubes
compared to the outer parts. have been proposed as enhanced X-ray sources for the analysis of battery
Another important aspect, sometimes overlooked when postmortem materials [374,381,382]. Ex situ postmortem X-ray measurements
sampling procedures are described, is the time elapsed between cell should be carried out by careful sampling of electrode materials fol
disassembly and the actual analysis. This should be minimised to ensure lowed by gentle hand-milling action on the harvested material using an
that the electrode samples in contact with electrolyte do not continue agate mortar. These additional steps are strongly recommended to
reacting. Finally, it should be pointed out that the use of a specific improve X-ray data quality and interpretation, due to the reduction of
sample holder for transport from the glove box to the equipment of in preferential orientations and micro-absorption issues. The two latter
terest is essential to avoid air/moisture contamination of the samples. phenomena detrimentally affect the Rietveld method via deviation from
correct quantitative phase analysis, calculated density values of each
3.3.2.2. Ex situ analysis. The physicochemical techniques most polycrystalline phase and incorrect determination of crystallographic
frequently used for postmortem analysis to validate EIS measurements site occupancy by ions, with direct impact on assumptions of
are summarised below and represented in Fig. 8b. ion-migration paths [381,383].
Surface sensitive methods for electrode and separator analysis: Specific XRD techniques such as Wide-angle X-ray scattering
(WAXs), Small Angle X-ray scattering (SAXs), X-ray Reflectivity (XRR)
• Microscopy-based: Optical microscopy, Scanning electron micro and atomic Pair Distribution Function (PDF) methods are useful tools
scopy (SEM), Focused ion beam (FIB)-SEM, Transmission electron also for postmortem LIB material characterisation. WAXs measurement
microscopy (TEM) is the most widely used of these techniques for battery material char
• Spectroscopy-based: Energy dispersive X-ray spectroscopy (EDX), acterisation. This method can be performed in both reflection (flat
Depth profiling X-ray photoelectron spectroscopy (XPS), Secondary sample; Bragg-Brentano setting) and transmission (capillary or passing
ion mass spectroscopy (SIMS)/Time of Flight SIMS (TOF-SIMS). through a pouch cell) geometries. WAXs measurements can provide
Bulk methods for electrode analysis: useful crystallographic details and complementary information in sup
• Inductively coupled plasma optical emission spectrometry (ICP- port of EIS analysis, helping to identify specific phenomena (e.g. ion
OES), Nuclear magnetic resonance spectroscopy (NMR), X-ray migration) which can contribute to the increase of the resistance during
diffraction (XRD). battery cell ageing [383,384]. Crystallographic and electrochemical
Surface & bulk methods of analysis: investigation can be combined and enriched with SAXs measurements.
• Diffraction-based techniques, Fourier transform infrared spectros The SAXs technique is suited to providing information at the nano
copy (FTIR), Glow discharge optical emission spectroscopy depth particle and morphological level. Critical parameters such as nano
profiling (GD-OES). particle distribution, dimension, shape, nanostructure, surface area and
Chemical methods for electrolyte analysis: porosity can be extracted by means of SAXs measurements of battery
• NMR (liquid), Liquid Chromatography coupled with Fourier infrared electrode materials as reported in Refs. [385–387]. With the XRR
spectrometry (LC-FTIR), Gas Chromatography (GC) coupled with a technique, information about the correlation of the SEI with battery
thermal conductivity detector (GC-TCD), GC-Mass Spectrometry performance degradation can be acquired [388]. Indications about
(GC-MS), GC coupled with a flame ionisation detector (GC-FID) and thickness, surface-interface roughness and density of electrode materials
Ion Chromatography (IC) coupled with Electro Spray Ionisation can be also obtained through XRR analysis of electrode components
(ESI). harvested from disassembled cells. However, a special experimental
set-up is required to achieve nanometric resolution [388,389]. The
Diffraction-based techniques (e.g. XRD or ND) are most suited to atomic PDF technique can take into account the total scattered X-ray
better understand the solid-state transformations taking place in the information coming from a sample, in the form of small peaks and
active materials of the battery, both in the bulk and, using special bumps of the background signal. All of these small halons provide in
experimental set-ups, also in the surface/interface domain [366,367]. formation on short length ordering of atoms, such as coordination
XRD and ND techniques are complementary, although in some cases the environment, distance between neighbouring atoms and local distor
latter is preferred to the former, e.g. in targeting light elements, specific tions. This level of information is not restricted to the first sphere of
isotopes or to discriminate between adjacent elements in the periodic coordination, and is complementary to crystal structure information
table [368,369]. from traditional XRD measurements and Rietveld refinement. In the
For investigation of Li-ion cell materials, it is strongly recommended battery field, PDF methods are used to investigate phases under for
to select appropriate X-ray wavelengths for diffraction experiments mation or still in amorphous or high disordered states [389–392].
[370,371]. The choice should be made as a function of (i) chemical SEM techniques are also widely applied in postmortem analysis to
composition of the samples and (ii) targeted experimental results. obtain information about changes in electrode morphology upon
cycling. Typically, this provides more information on the anode side
17
than the cathode side [356,393]. It has been reported that, upon information.
degradation, anode materials (harvested from a small Li-ion pouch cell
with a capacity of 0.65 Ah) show an apparent roughness on the surface 4. Summary, conclusion and outlook
which can be attributed to a damaged SEI [82]. A thick deposited layer
on the anode at the midpoint location in the harvested electrode can be Any interpretation, prediction or evaluation based on impedance
also identified by SEM on cells degraded under constant current con spectra relies on the ability to trace the measured impedance values and
ditions [365]. SEM can also provide useful information on the separator. their uncertainty to a well established common reference: the interna
For example, Kwak et al. [62] reported a decrease in separator porosity tional system of units. The analysis of CMCs currently proposed by NMIs
with storage time. This research work suggests that by-products, origi clearly highlights the need to improve the traceability of impedance
nating from side-reactions in the Li-ion cell, can accumulate on the measurement, especially in the range of interest for commercial LIBs (i.
separator surface resulting in poor Li-ion transfer and increased internal e. below 0.1 Ω and below 10 Hz). Moreover, a careful uncertainty budget
resistance of the cell. has to be established to take into account measurement conditions (long
XPS measurements are also applied to both cathode and anode to single-wire cables, home-made test fixtures, etc.) that differ from the
identify the various products generated during cycling of the cell which ideal conditions required for impedance metrology. Many laboratories
could contribute to changes in the EIS spectra. The collection and and research projects are currently working to address these issues and
analysis of the XPS spectra of the various electrode samples generally to improve the reliability of EIS measurement of commercial LIBs [204].
reveal the presence of fluorine (mainly attributed to the polyvinylidene These aspects, if not properly considered, could limit the constructive
fluoride binder and LiF formed by decomposition of the electrolyte so impact of EIS in understanding the underlying phenomena occurring in
lution) and oxygen-containing species, attributed to carbon-oxygen commercial LIBs.
bonds in carbonates and carboxylates generated by decomposition of A number of measurement techniques are available to obtain the
electrolyte solvent during cell cycling [139]. In the case of LiFe impedance spectra of commercial LIBs. Conventional single-sine EIS is
PO4-based cells, XPS and EDX were able to identify Fe contamination in the most commonly applied method for offline impedance measure
the graphite, which occurs when LiPF6 is used as electrolyte salt at ments. Key considerations include the type (current or voltage) and size
elevated temperatures [394,395]. This Fe contamination in the anode of perturbation, in order to maximise signal-to-noise ratio whilst
has been correlated to increased impedance [396,397]. For this tech maintaining a linear system response, as well as the procedure for taking
nique, particular attention should be paid during sample preparation measurements at different temperatures and SOCs (including rest pe
and transportation, as well as in the interpretation of the XPS spectra riods allowed after adjustment) and the frequency range and resolution
[398]. for data collection. A well defined measurement system is required to
As mentioned previously, an electrochemical approach to validate ensure repeatability and reproducibility of results. Various other
the interpretation of commercial cell impedance measurements consists impedance measurement techniques have been proposed to provide
of reassembly of harvested cathode or anode electrodes in a symmetric additional information or to overcome some of the measurement chal
two-electrode cell configuration (see Fig. 8c) [32,89,129,251,399]. lenges of conventional EIS. Dynamic EIS, where the applied perturbation
Applying this cell set-up, the individual contribution of each electrode to is superimposed on a “baseline” DC current/voltage offset, can be used
the overall Li-ion cell impedance spectrum can be evaluated by a com to investigate dynamic effects of charge/discharge, provided the non-
parison of the commercial cell impedance with that of the symmetric stationarity of the system during the test is minimised/accounted for.
lab-scale cells made of harvested materials [400]. To obtain a proper Multi-sine EIS, whereby multiple frequencies are excited simulta
validation of ageing effects, it is suggested to perform multiple sym neously, can be useful for reducing measurement times, though gener
metric cell measurements at different conditions (e.g. various SOCs ally at a cost to the signal-to-noise ratio obtainable whilst keeping the
and/or temperatures) using electrodes harvested from aged cells [121, overall perturbation amplitude within the linear response regime. The
251]. same is true for other time domain techniques, based on non-sinusoidal
Although the use of symmetric cells (i) removes any possible EIS excitation signals. On the other hand, the non-linear response of the
distortions coming from a chemically distinct counter or reference battery, analysed through NFRA/NLEIS, can provide additional infor
electrode [121,401] and (ii) is beneficial in terms of qualitative evalu mation about electrochemical processes. However, this requires more
ation of the electrode contributions, reliable quantitative evaluation is sophisticated interpretation, which is at a relatively early stage of
not straightforward. Indeed, since symmetric cells are reconstructed development in the research field. Care is also needed to avoid cell
from harvested electrodes, fresh electrolyte and uncycled separator are damage with large current perturbations. Whereas conventional EIS
used to acquire the impedance measurements [88]. Moreover, because requires specialist equipment for signal generation, passive impedance
of the symmetric nature, SOC cannot be varied in this type of cell after spectroscopy has been proposed as an onboard diagnostic tool for
assembly [251] and, thus, precision of cell construction (e.g., in terms of electric vehicles, using only current and voltage signals that result from
sampling and application of harvested electrodes at the same SOC) is operating the vehicle, allowing the use of impedance measurements for
essential to obtain reliable validation procedures. Alternatively, the real-time state estimation and fault detection. Single-frequency EIS
three-electrode half-cell configuration (i.e. where the counter electrode measurements have also been proposed for state (temperature) estima
is metallic lithium and a separate, stable electrode is applied as reference tion, as well as being commonly used for quality assurance due to their
electrode) [402] is also used for EIS validation [403]. With this cell rapidity, but the choice of frequency depends strongly on cell design
configuration, the SOC of the electrode under study can be varied, (especially format). Overall, EIS can be highly effective for under
although the other limitations of the symmetric cell are present. As with standing and estimating performance of commercial LIBs under different
full cells, it is important to allow the reconstructed cell to equilibrate conditions. However, the measurement challenges listed above must be
after adjusting the SOC [76]. properly considered in order to advance the usefulness of impedance-
To summarise, performing postmortem analysis to validate the based techniques in the research field of commercial LIBs. Failure to
interpretation of EIS measurements of commercial Li-ion cells is a do so can significantly affect the reliability of EIS measurements and the
complex and non-trivial task and several factors can affect reliable data conclusions drawn from them. Moreover, tests should be properly
collection. The procedure for disassembly of the battery cells is crucial, designed and controlled in order to decouple the effects of different
and all the details need to be comprehensively reported in order to parameters.
ensure reproducibility and comparability of results. Although many A critical insight into the most commonly used methods of inter
characterisation techniques can be applied, the selection of a suitable set pretation, i.e., ECM and physics-based modelling, to evaluate impedance
of techniques must be done carefully in order to obtain meaningful data from LIBs has been given. Due to their easy application, ECMs are
18
currently the most widespread option to relate impedance spectra to authors thank Dr. Edmund Dickinson (National Physical Laboratory,
physical processes and parameters. However, a possible major drawback Teddington, UK) for productive scientific and linguistic discussions.
of applying ECMs to interpret EIS measurements concerns the lack of
direct correlation of the ECMs with the physical processes occurring in Appendix A. Supplementary data
the battery. Despite the higher level of complexity with respect to ECMs,
physics-based models (e.g., the Barsoukov model) are applied increas Supplementary data to this article can be found online at https://doi.
ingly often and are more likely to become the state-of-the-art in the org/10.1016/j.jpowsour.2020.228742.
future. In any case, before applying models, validation of the impedance
results should be carried out mathematically with the Kramers-Kronig References
relations in order to check whether the measurements adequately ful
fil the criteria of linearity, causality and stationarity. [1] K.W. Beard, Linden’s Handbook of Batteries, fifth ed., McGraw-Hill Education,
2019.
In terms of attribution of features in the impedance spectrum to [2] A. Barai, K. Uddin, W.D. Widanage, A. McGordon, P. Jennings, A study of the
particular cell components and processes, validation can be carried out influence of measurement timescale on internal resistance characterisation
to some extent via postmortem analysis. For such a reason, in this review methodologies for lithium-ion cells, Sci. Rep. 8 (1) (2018) 1–13, https://doi.org/
10.1038/s41598-017-18424-5.
we have presented some of the most relevant postmortem analysis [3] C. Fleischer, W. Waag, H.-M. Heyn, D.U. Sauer, On-line adaptive battery
techniques that can be used to validate, e.g., conclusions about LIB impedance parameter and state estimation considering physical principles in
ageing and degradation processes, through the study of the most rele reduced order equivalent circuit battery models part 2. Parameter and state
estimation, J. Power Sources 262 (2014) 457–482, https://doi.org/10.1016/j.
vant cell components (anode, cathode and electrolyte). From the anal jpowsour.2014.03.046.
ysis carried out, it is clear that selection of a specific technique must be [4] L. Wildfeuer, N. Wassiliadis, C. Reiter, M. Baumann, M. Lienkamp, Experimental
made wisely (targeting each particular case), taking into account that no characterization of Li-Ion battery resistance at the cell, module and pack level,
EVER vol. 2019 (2019), https://doi.org/10.1109/EVER.2019.8813578, 14th
single technique can provide all the information desired; only through a
International Conference on Ecological Vehicles and Renewable Energies 2019)
combination of several techniques can meaningful conclusions be 1–12.
drawn. Of particular importance is the procedure for disassembly of the [5] Y. Tao, J. Ye, X. Li, Q. Zhang, A. Pan, Q. Liao, X. Yang, Ageing study of 26650
cell itself. This step, which may be incorrectly regarded as trivial, is type li-ion batteries module at the charge-discharge rates of 1 C, AIP Conference
Proceedings 2154 (September) (2019), https://doi.org/10.1063/1.5125355.
actually critical. Careful planning of the sequence of steps requires [6] B.M. Huhman, J.M. Heinzel, L. Mili, C.T. Love, D.A. Wetze, Investigation into
proper attention and the reporting of steps undertaken needs to be state-of-health impedance diagnostic for 26650 4P1S battery packs,
sufficiently detailed to ensure repeatability and reliable interlaboratory J. Electrochem. Soc. 164 (1) (2017) A6401–A6411, https://doi.org/10.1149/
2.0631701jes.
comparison. The sampling must also consider the relative position [7] T. Yokoshima, D. Mukoyama, H. Nara, S. Maeda, K. Nakazawa, T. Momma,
within the cell, to assess the possible - and likely - distribution of local T. Osaka, Impedance measurements of kilowatt-class lithium ion battery
effects. Proper transfer of the harvested samples is essential and needs to modules/cubicles in energy storage systems by square-current electrochemical
impedance spectroscopy, Electrochim. Acta 246 (2017) 800–811, https://doi.
avoid contact with air and moisture, to ensure the system analysed is not org/10.1016/j.electacta.2017.05.076.
altered beyond its original harvesting condition. Although the use of [8] K. Takeno, M. Ichimura, K. Takano, J. Yamaki, S. Okada, Quick testing of
non-destructive techniques seems to be preferred to avoid these possible batteries in lithium-ion battery packs with impedance-measuring technology,
J. Power Sources 128 (1) (2004) 67–75, https://doi.org/10.1016/j.
complications, disassembly of the cell may be unavoidable in some jpowsour.2003.09.045.
cases. A clear example is represented by symmetric cell reconstruction [9] M.J. Brand, M.H. Hofmann, M. Steinhardt, S.F. Schuster, A. Jossen, Current
with harvested electrodes, whereby the specific processes occurring at distribution within parallel-connected battery cells, J. Power Sources 334 (2016)
202–212, https://doi.org/10.1016/j.jpowsour.2016.10.010.
the anode and cathode of the cell can be studied and validated. How
[10] X. Li, L. Zhang, Y. Liu, A. Pan, Q. Liao, X. Yang, A fast classification method of
ever, the recent development of operando and in situ techniques which retired electric vehicle battery modules and their energy storage application in
combine electrochemical and non-electrochemical method of analysis photovoltaic generation, Int. J. Energy Res. (2019) 1–8, https://doi.org/10.1002/
opens up a whole new realm of possibilities which have gained more er.5083.
[11] M. Baumann, L. Wildfeuer, S. Rohr, M. Lienkamp, Parameter variations within Li-
attention in the research field. Ion battery packs - theoretical investigations and experimental quantification,
In summary, the methods for measurement, interpretation and J. Energy Storage 18 (2018) 295–307, https://doi.org/10.1016/j.
validation examined in this review represent the state-of-the-art in the est.2018.04.031.
[12] A. Barai, K. Uddin, M. Dubarry, L. Somerville, A. McGordon, P. Jennings,
application of EIS to commercial LIBs. With proper consideration of the I. Bloom, A comparison of methodologies for the non-invasive characterisation of
requirements and limitations associated with specific experimental set- commercial Li-ion cells, Prog. Energy Combust. Sci. 72 (2019) 1–31, https://doi.
ups, different techniques employed, and proper attention to metrolog org/10.1016/j.pecs.2019.01.001.
[13] K. Rumpf, M. Naumann, A. Jossen, Experimental investigation of parametric cell-
ical practice, EIS could be considered as a powerful tool for characteri to-cell variation and correlation based on 1100 commercial lithium-ion cells,
sation of commercial cells. J. Energy Storage 14 (2017) 224–243, https://doi.org/10.1016/j.
est.2017.09.010.
[14] P.S. Attidekou, S. Lambert, M. Armstrong, J. Widmer, K. Scott, P.A. Christensen,
Declaration of competing interest A study of 40 Ah lithium ion batteries at zero percent state of charge as a function
of temperature, J. Power Sources 269 (2014) 694–703, https://doi.org/10.1016/
The authors declare that they have no known competing financial j.jpowsour.2014.06.064.
[15] F.J. Günter, J.B. Habedank, D. Schreiner, T. Neuwirth, R. Gilles, G. Reinhart,
interests or personal relationships that could have appeared to influence Introduction to electrochemical impedance spectroscopy as a measurement
the work reported in this paper. method for the wetting degree of lithium-ion cells, J. Electrochem. Soc. 165 (14)
(2018) A3249–A3256, https://doi.org/10.1149/2.0081814jes.
[16] X. Kong, G.L. Plett, M. Scott Trimboli, Z. Zhang, D. Qiao, T. Zhao, Y. Zheng,
Acknowledgement
Pseudo-two-dimensional model and impedance diagnosis of micro internal short
circuit in lithium-ion cells, J. Energy Storage 27 (2020) 101085, https://doi.org/
Nina Meddings, Marco Heinrich, Frédéric Overney, Vanesa Ruiz, 10.1016/j.est.2019.101085. July 2019.
Emilio Napolitano, Steffen Seitz, Gareth Hinds, Rinaldo Raccichini, [17] H. Piret, P. Granjon, N. Guillet, V. Cattin, Tracking of electrochemical impedance
of batteries, J. Power Sources 312 (2016) 60–69, https://doi.org/10.1016/j.
Miran Gaberšček and Juyeon Park acknowledge the financial support of jpowsour.2016.02.006.
the European Metrology Programme for Innovation and Research [18] S. Rodrigues, N. Munichandraiah, A.K. Shukla, AC impedance and state-of-charge
(EMPIR), co-financed by the Participating States and the European analysis of a sealed lithium-ion rechargeable battery, J. Solid State Electrochem. 3
(7–8) (1999) 397–405, https://doi.org/10.1007/s100080050173.
Union’s Horizon 2020 research and innovation programme, within the [19] S. Rodrigues, N. Munichandraiah, A.K. Shukla, Review of state-of-charge
project 17IND10-LiBforSecUse. Jong-Sook Lee acknowledges the sup indication of batteries by means of a.c. impedance measurements, J. Power
port of the National Research Foundation (NRF) of Korea funded by the Sources 87 (1) (2000) 12–20, https://doi.org/10.1016/S0378-7753(99)00351-1.
Ministry of Science and ICT (MSIT) (NRF-2018R1A5A1025224). All the
19
[20] J. Xu, C.C. Mi, B. Cao, J. Cao, A new method to estimate the state of charge of [43] H. Zhou, M.-A. Einarsrud, F. Vullum-Bruer, In situ X-ray diffraction and
lithium-ion batteries based on the battery impedance model, J. Power Sources electrochemical impedance spectroscopy of a nanoporous Li2FeSiO4/C cathode
233 (2013) 277–284, https://doi.org/10.1016/j.jpowsour.2013.01.094. during the initial charge/discharge cycle of a Li-ion battery, J. Power Sources 238
[21] A. La Rue, P.J. Weddle, M. Ma, C. Hendricks, R.J. Kee, T.L. Vincent, State-of- (2013) 478–484, https://doi.org/10.1016/j.jpowsour.2013.03.193. Supplement
charge estimation of LiFePO4-Li4Ti5O12 batteries using history-dependent C.
complex-impedance, J. Electrochem. Soc. 166 (16) (2019) A4041–A4046, [44] A. Podias, A. Pfrang, F. Di Persio, A. Kriston, S. Bobba, F. Mathieux, M. Messagie,
https://doi.org/10.1149/2.0221916jes. L. Boon-Brett, Sustainability assessment of second use applications of automotive
[22] A. Cuadras, O. Kanoun, Soc li-ion battery monitoring with impedance batteries: ageing of Li-ion battery cells in automotive and grid-scale applications,
spectroscopy, in: 6th International Multi-Conference on Systems, Signals and World Electric Vehicle Journal 9 (2) (2018), https://doi.org/10.3390/
Devices, SSD, 2009, pp. 1–5, https://doi.org/10.1109/SSD.2009.4956761, 2009 wevj9020024.
(2009). [45] Q. Liao, M. Mu, S. Zhao, L. Zhang, T. Jiang, J. Ye, X. Shen, G. Zhou, Performance
[23] J.L. Jespersen, A.E. Tønnesen, K. Nørregaard, L. Overgaard, F. Elefsen, Capacity assessment and classification of retired lithium ion battery from electric vehicles
measurements of Li-ion batteries using AC impedance spectroscopy, World for energy storage, Int. J. Hydrogen Energy 42 (30) (2017) 18817–18823,
Electric Vehicle Journal 3 (1) (2009) 127–133, https://doi.org/10.3390/ https://doi.org/10.1016/j.ijhydene.2017.06.043.
wevj3010127. [46] J. Li, C. Arbizzani, S. Kjelstrup, J. Xiao, Y.y. Xia, Y. Yu, Y. Yang, I. Belharouak,
[24] F. Leng, C.M. Tan, R. Yazami, M.D. Le, A practical framework of electrical based T. Zawodzinski, S.T. Myung, R. Raccichini, S. Passerini, Good practice guide for
online state-of-charge estimation of lithium ion batteries, J. Power Sources 255 papers on batteries for the Journal of Power Sources, J. Power Sources 452
(2014) 423–430, https://doi.org/10.1016/j.jpowsour.2014.01.020. (February) (2020) 227824, https://doi.org/10.1016/j.jpowsour.2020.227824.
[25] R. Mingant, J. Bernard, V. Sauvant-Moynot, A. Delaille, S. Mailley, J.L. Hognon, [47] E. Barsoukov, J.R. Macdonald (Eds.), Impedance Spectroscopy: Theory,
F. Huet, EIS measurements for determining the SoC and SoH of Li-ion batteries, Experiment, and Applications, third ed., Wiley, 2018.
ECS Transactions 33 (39) (2011) 41–53, https://doi.org/10.1149/1.3589920. [48] T. Momma, M. Matsunaga, D. Mukoyama, T. Osaka, Ac impedance analysis of
[26] Y. Sun, Y. Li, M. Yu, Z. Zhou, Q. Zhang, B. Duan, Y. Shang, C. Zhang, Variable lithium ion battery under temperature control, J. Power Sources 216 (2012)
fractional order - a comprehensive evaluation indicator of lithium-ion batteries, 304–307, https://doi.org/10.1016/j.jpowsour.2012.05.095.
J. Power Sources 448 (October 2019) (2020) 227411, https://doi.org/10.1016/J. [49] T.F. Landinger, G. Schwarzberger, A. Jossen, A novel method for high frequency
JPOWSOUR.2019.227411. battery impedance measurements, in: 2019 IEEE International Symposium on
[27] B. Wang, S.E. Li, H. Peng, Z. Liu, Fractional-order modeling and parameter Electromagnetic Compatibility, Signal and Power Integrity, EMC+SIPI, 2019,
identification for lithium-ion batteries, J. Power Sources 293 (2015) 151–161, pp. 106–110, https://doi.org/10.1109/ISEMC.2019.8825315, 2019.
https://doi.org/10.1016/j.jpowsour.2015.05.059. [50] S. Schindler, M.A. Danzer, Influence of cell design on impedance characteristics of
[28] R. Xiong, J. Tian, H. Mu, C. Wang, A systematic model-based degradation cylindrical lithium-ion cells: a model-based assessment from electrode to cell
behavior recognition and health monitoring method for lithium-ion batteries, level, J. Energy Storage 12 (2017) 157–166, https://doi.org/10.1016/j.
Appl. Energy 207 (2017) 372–383, https://doi.org/10.1016/j. est.2017.05.002.
apenergy.2017.05.124. [51] E.C. Shin, P.A. Ahn, H.H. Seo, J.M. Jo, S.D. Kim, S.K. Woo, J.H. Yu, J. Mizusaki, J.
[29] R. Arunachala, K. Makinejad, S. Athlekar, A. Jossen, J. Garche, Cycle life S. Lee, Polarization mechanism of high temperature electrolysis in a Ni-YSZ/YSZ/
characterisation of large format lithium-ion cells, 2013 World Electric Vehicle LSM solid oxide cell by parametric impedance analysis, Solid State Ionics 232
Symposium and Exhibition, EVS (2014) 1–9, https://doi.org/10.1109/ (2013) 80–96, https://doi.org/10.1016/j.ssi.2012.10.028.
EVS.2013.6914865, 2014. [52] S.M.M. Alavi, C.R. Birkl, D.A. Howey, Time-domain fitting of battery
[30] S.F. Schuster, M.J. Brand, C. Campestrini, M. Gleissenberger, A. Jossen, electrochemical impedance models, J. Power Sources 288 (2015) 345–352,
Correlation between capacity and impedance of lithium-ion cells during calendar https://doi.org/10.1016/j.jpowsour.2015.04.099.
and cycle life, J. Power Sources 305 (2016) 191–199, https://doi.org/10.1016/j. [53] T.P. Heins, N. Schlüter, S.T. Ernst, U. Schröder, On the interpretation of
jpowsour.2015.11.096. impedance spectra of large-format lithium-ion batteries and its application in
[31] X. Wang, X. Wei, H. Dai, Estimation of state of health of lithium-ion batteries aging studies, Energy Technol. (2019) 1900279doi, https://doi.org/10.1002/
based on charge transfer resistance considering different temperature and state of ente.201900279.
charge, J. Energy Storage 21 (2019) 618–631, https://doi.org/10.1016/j. [54] T.P. Heins, R. Leithoff, N. Schlüter, U. Schröder, K. Dröder, Impedance
est.2018.11.020. August 2018. spectroscopic investigation of the impact of erroneous cell assembly on the aging
[32] T. Osaka, D. Mukoyama, H. Nara, Review-Development of diagnostic process for of lithium-ion batteries, Energy Technol. (2019) 1900288, https://doi.org/
commercially available batteries, especially lithium ion battery, by 10.1002/ente.201900288.
electrochemical impedance spectroscopy, J. Electrochem. Soc. 162 (14) (2015) [55] J. Illig, J.P. Schmidt, M. Weiss, A. Weber, E. Ivers-Tiffée, Understanding the
A2529–A2537, https://doi.org/10.1149/2.0141514jes. impedance spectrum of 18650 LiFePO4-cells, J. Power Sources 239 (2013)
[33] M. Galeotti, L. Cinà, C. Giammanco, S. Cordiner, A. Di Carlo, Performance 670–679, https://doi.org/10.1016/j.jpowsour.2012.12.020.
analysis and SOH (state of health) evaluation of lithium polymer batteries [56] X. Zhou, J. Huang, Z. Pan, M. Ouyang, Impedance characterization of lithium-ion
through electrochemical impedance spectroscopy, Energy 89 (2015) 678–686, batteries aging under high-temperature cycling: importance of electrolyte-phase
https://doi.org/10.1016/j.energy.2015.05.148. diffusion, J. Power Sources 426 (2019) 216–222, https://doi.org/10.1016/j.
[34] C. Lyu, T. Zhang, W. Luo, G. Wei, B. Ma, L. Wang, SOH Estimation of lithium-ion jpowsour.2019.04.040. April.
batteries based on fast time domain impedance spectroscopy, in: Proceedings of [57] S. Gantenbein, M. Weiss, E. Ivers-Tiffée, Impedance based time-domain modeling
the 14th IEEE Conference on Industrial Electronics and Applications, ICIEA, 2019, of lithium-ion batteries: Part I, J. Power Sources 379 (2018) 317–327, https://
pp. 2142–2147, https://doi.org/10.1109/ICIEA.2019.8834119, 2019. doi.org/10.1016/j.jpowsour.2018.01.043. January.
[35] R.G. Jungst, G. Nagasubramanian, H.L. Case, B.Y. Liaw, A. Urbina, T.L. Paez, D. [58] J. Landesfeind, D. Pritzl, H.A. Gasteiger, An analysis protocol for three-electrode
H. Doughty, Accelerated calendar and pulse life analysis of lithium-ion cells, Li-ion battery impedance spectra: Part I. Analysis of a high-voltage positive
J. Power Sources 119–121 (2003) 870–873, https://doi.org/10.1016/S0378- electrode, J. Electrochem. Soc. 164 (7) (2017) A1773–A1783, https://doi.org/
7753(03)00193-9. 10.1149/2.0131709jes.
[36] R. Koch, A. Jossen, Temperature measurement of large format pouch cells with [59] J. Xia, K.J. Nelson, Z. Lu, J.R. Dahn, Impact of electrolyte solvent and additive
impedance spectroscopy, in: 28th International Electric Vehicle Symposium and choices on high voltage Li-ion pouch cells, J. Power Sources 329 (2016) 387–397,
Exhibition 2015, EVS, 2015, pp. 1–9, 2015. https://doi.org/10.1016/j.jpowsour.2016.08.100.
[37] J.P. Schmidt, S. Arnold, A. Loges, D. Werner, T. Wetzel, E. Ivers-Tiffée, [60] S. Schindler, M.A. Danzer, A novel mechanistic modeling framework for analysis
Measurement of the internal cell temperature via impedance: evaluation and of electrode balancing and degradation modes in commercial lithium-ion cells,
application of a new method, J. Power Sources 243 (2013) 110–117, https://doi. J. Power Sources 343 (2017) 226–236, https://doi.org/10.1016/j.
org/10.1016/j.jpowsour.2013.06.013. jpowsour.2017.01.026.
[38] L.H.J. Raijmakers, D.L. Danilov, J.P.M. Van Lammeren, M.J.G. Lammers, P.H. [61] B. Manikandan, C. Yap, P. Balaya, Towards understanding heat generation
L. Notten, Sensorless battery temperature measurements based on characteristics of Li-ion batteries by calorimetry, impedance, and potentiometry
electrochemical impedance spectroscopy, J. Power Sources 247 (2014) 539–544, studies, J. Electrochem. Soc. 164 (12) (2017) A2794–A2800, https://doi.org/
https://doi.org/10.1016/j.jpowsour.2013.09.005. 10.1149/2.1811712jes.
[39] J.G. Zhu, Z.C. Sun, X.Z. Wei, H.F. Dai, A new lithium-ion battery internal [62] G. Kwak, J. Park, J. Lee, S. Kim, I. Jung, Effects of anode active materials to the
temperature on-line estimate method based on electrochemical impedance storage-capacity fading on commercial lithium-ion batteries, J. Power Sources
spectroscopy measurement, J. Power Sources 274 (2015) 990–1004, https://doi. 174 (2) (2007) 484–492, https://doi.org/10.1016/j.jpowsour.2007.06.169.
org/10.1016/j.jpowsour.2014.10.182. [63] R. Petibon, J. Xia, L. Ma, M.K. Bauer, K.J. Nelson, J.R. Dahn, Electrolyte system
[40] X. Wang, X. Wei, Q. Chen, J. Zhu, H. Dai, Lithium-ion battery temperature on-line for high voltage li-ion cells, J. Electrochem. Soc. 163 (13) (2016) A2571–A2578,
estimation based on fast impedance calculation, J. Energy Storage 26 (2019), https://doi.org/10.1149/2.0321613jes.
https://doi.org/10.1016/j.est.2019.100952. July. [64] Y. Duan, H. Wu, L. Huang, L. Liu, Y. Zhang, Optimizing current terminals of 18
[41] H.P.G.J. Beelen, L.H.J. Raijmakers, M.C.F. Donkers, P.H.L. Notten, H.J. Bergveld, 650 lithium-ion power batteries under high discharge current, Energy Technol. 5
A comparison and accuracy analysis of impedance-based temperature estimation (9) (2017) 1619–1626, https://doi.org/10.1002/ente.201700008.
methods for Li-ion batteries, Appl. Energy 175 (2016) 128–140, https://doi.org/ [65] J. Vetter, P. Novák, M.R. Wagner, C. Veit, K.C. Möller, J.O. Besenhard, M. Winter,
10.1016/j.apenergy.2016.04.103. M. Wohlfahrt-Mehrens, C. Vogler, A. Hammouche, Ageing mechanisms in
[42] P. Haussmann, J. Melbert, Internal cell temperature measurement and thermal lithium-ion batteries, J. Power Sources 147 (1–2) (2005) 269–281, https://doi.
modeling of lithium ion cells for automotive applications by means of org/10.1016/j.jpowsour.2005.01.006.
electrochemical impedance spectroscopy, SAE International Journal of [66] H. Ge, J. Huang, J. Zhang, Z. Li, Temperature-adaptive alternating current
Alternative Powertrains 6 (2) (2017), https://doi.org/10.4271/2017-01-1215. preheating of lithium-ion batteries with lithium deposition prevention,
20
J. Electrochem. Soc. 163 (2) (2016) A290–A299, https://doi.org/10.1149/ to accelerated aging, J. Power Sources 449 (2019) 227369, https://doi.org/
2.0961602jes. 10.1016/j.jpowsour.2019.227369. July 2019.
[67] K. Amine, C.H. Chen, J. Liu, M. Hammond, A. Jansen, D. Dees, I. Bloom, [91] C.H. Chen, J. Liu, K. Amine, Symmetric cell approach towards simplified study of
D. Vissers, G. Henriksen, Factors responsible for impedance rise in high power cathode and anode behavior in lithium ion batteries, Electrochem. Commun. 3 (1)
lithium ion batteries, J. Power Sources 97–98 (2001) 684–687, https://doi.org/ (2001) 44–47, https://doi.org/10.1016/S1388-2481(00)00145-4.
10.1016/S0378-7753(01)00701-7. [92] R. Weber, A.J. Louli, K.P. Plucknett, J.R. Dahn, Resistance growth in lithium-ion
[68] S.P. Rangarajan, Y. Barsukov, P.P. Mukherjee, In operando impedance based pouch cells with LiNi0.80Co0.15Al0.05O2 positive electrodes and proposed
diagnostics of electrode kinetics in Li-ion pouch cells, J. Electrochem. Soc. 166 mechanism for voltage dependent charge-transfer resistance, J. Electrochem. Soc.
(10) (2019) A2131–A2141, https://doi.org/10.1149/2.1191910jes. 166 (10) (2019) A1779–A1784, https://doi.org/10.1149/2.0361910jes.
[69] R. Raccichini, M. Amores, G. Hinds, Critical review of the use of reference [93] D. Andre, M. Meiler, K. Steiner, C. Wimmer, T. Soczka-Guth, D.U. Sauer,
electrodes in Li-ion batteries: a diagnostic perspective, Batteries 5 (1) (2019) 12, Characterization of high-power lithium-ion batteries by electrochemical
https://doi.org/10.3390/batteries5010012. impedance spectroscopy. I. Experimental investigation, J. Power Sources 196
[70] T.P. Heins, N. Schlüter, U. Schröder, Electrode-Resolved monitoring of the ageing (12) (2011) 5334–5341, https://doi.org/10.1016/j.jpowsour.2010.12.102.
of large-scale lithium-ion cells by using electrochemical impedance spectroscopy, [94] W. Waag, S. Käbitz, D.U. Sauer, Experimental investigation of the lithium-ion
ChemElectroChem 4 (11) (2017) 2921–2927, https://doi.org/10.1002/ battery impedance characteristic at various conditions and aging states and its
celc.201700686. influence on the application, Appl. Energy 102 (2013) 885–897, https://doi.org/
[71] J. Gomez, R. Nelson, E.E. Kalu, M.H. Weatherspoon, J.P. Zheng, Equivalent 10.1016/j.apenergy.2012.09.030.
circuit model parameters of a high-power Li-ion battery: thermal and state of [95] F. An, L. Chen, J. Huang, J. Zhang, P. Li, Rate dependence of cell-to-cell variations
charge effects, J. Power Sources 196 (10) (2011) 4826–4831, https://doi.org/ of lithium-ion cells, Sci. Rep. 6 (2016) 4–10, https://doi.org/10.1038/srep35051.
10.1016/j.jpowsour.2010.12.107. [96] A.S. Kumar, T.V.S.L. Satyavani, M. Senthilkumar, Effect of temperature and
[72] P. Suresh, A.K. Shukla, N. Munichandraiah, Temperature dependence studies of a. charge stand on performance of lithium-ion polymer pouch cell, J. Energy Storage
c. impedance of lithium-ion cells, J. Appl. Electrochem. 32 (3) (2002) 267–273, 6 (2016) 239–247, https://doi.org/10.1016/j.est.2016.02.002.
https://doi.org/10.1023/A:1015565404343. [97] J.P. Fellner, G.J. Loeber, S.S. Sandhu, Testing of lithium-ion 18650 cells and
[73] J.-M. Atebamba, ž. Moškon, S. Pejovnik, M. Gaberšček, On the interpretation of characterizing/predicting cell performance, J. Power Sources 81–82 (1999)
measured impedance spectra of insertion cathodes for lithium-ion batteries, 867–871, https://doi.org/10.1016/S0378-7753(98)00238-9.
J. Electrochem. Soc. 157 (1) (2010) A1218–A1228, https://doi.org/10.1149/ [98] M. Ecker, P. Shafiei Sabet, D.U. Sauer, Influence of operational condition on
1.3489353. lithium plating for commercial lithium-ion batteries – electrochemical
[74] S.S. Zhang, K. Xu, T.R. Jow, Charge and discharge characteristics of a commercial experiments and post-mortem-analysis, Appl. Energy 206 (August) (2017)
LiCoO2-based 18650 Li-ion battery, J. Power Sources 160 (2006) 1403–1409, 934–946, https://doi.org/10.1016/j.apenergy.2017.08.034.
https://doi.org/10.1016/j.jpowsour.2006.03.037, 2 SPEC. ISS. [99] V. Muenzel, A.F. Hollenkamp, A.I. Bhatt, J. De Hoog, M. Brazil, D.A. Thomas,
[75] P. Svens, R. Eriksson, J. Hansson, M. Behm, T. Gustafsson, G. Lindbergh, Analysis I. Mareels, A comparative testing study of commercial 18650-format lithium-ion
of aging of commercial composite metal oxide - Li 4Ti5O12 battery cells, J. Power battery cells, J. Electrochem. Soc. 162 (8) (2015) A1592–A1600, https://doi.org/
Sources 270 (2014) 131–141, https://doi.org/10.1016/j.jpowsour.2014.07.050. 10.1149/2.0721508jes.
[76] I. Jiménez Gordon, S. Grugeon, A. Débart, G. Pascaly, S. Laruelle, Electrode [100] S.S. Zhang, The effect of the charging protocol on the cycle life of a Li-ion battery,
contributions to the impedance of a high-energy density Li-ion cell designed for J. Power Sources 161 (2) (2006) 1385–1391, https://doi.org/10.1016/j.
EV applications, Solid State Ionics 237 (2013) 50–55, https://doi.org/10.1016/j. jpowsour.2006.06.040.
ssi.2013.02.016. [101] L.R. Chen, S.L. Wu, T.R. Chen, W.R. Yang, C.S. Wang, P.C. Chen, Detecting of
[77] M. Mastali, M. Farkhondeh, S. Farhad, R.A. Fraser, M. Fowler, Electrochemical optimal li-ion battery charging frequency by using AC impedance technique, 2009
modeling of commercial LiFePO4 and graphite electrodes: kinetic and transport 4th IEEE Conference on Industrial Electronics and Applications, ICIEA 1 (2009)
properties and their temperature dependence, J. Electrochem. Soc. 163 (13) 3378–3381, https://doi.org/10.1109/ICIEA.2009.5138829, 2009.
(2016) A2803–A2816, https://doi.org/10.1149/2.1151613jes. [102] S. Guo, R. Xiong, K. Wang, F. Sun, A novel echelon internal heating strategy of
[78] P. Shafiei Sabet, G. Stahl, D.U. Sauer, Non-invasive investigation of predominant cold batteries for all-climate electric vehicles application, Appl. Energy 219 (5)
processes in the impedance spectra of high energy lithium-ion batteries with (2018) 256–263, https://doi.org/10.1016/j.apenergy.2018.03.052.
Nickel-Cobalt-Aluminum cathodes, J. Power Sources 406 (2018) 185–193, [103] Q.A. Huang, Y. Shen, Y. Huang, L. Zhang, J. Zhang, Impedance characteristics and
https://doi.org/10.1016/j.jpowsour.2018.10.024. November. diagnoses of automotive lithium-ion batteries at 7.5% to 93.0% state of charge,
[79] Q. Zhang, R.E. White, Calendar life study of Li-ion pouch cells, J. Power Sources Electrochim. Acta 219 (2016) 751–765, https://doi.org/10.1016/j.
173 (2) (2007) 990–997, https://doi.org/10.1016/j.jpowsour.2007.08.044. electacta.2016.09.154.
[80] P. Ramadass, B. Haran, R. White, B.N. Popov, Capacity fade of Sony 18650 cells [104] J. Fan, S. Tan, Studies on charging lithium-ion cells at low temperatures,
cycled at elevated temperatures: Part I. Cycling performance, J. Power Sources J. Electrochem. Soc. 153 (6) (2006), https://doi.org/10.1149/1.2190029.
112 (2) (2002) 606–613, https://doi.org/10.1016/S0378-7753(02)00474-3. A1081–A1092.
[81] Y. Zheng, Y.-B. He, K. Qian, B. Li, X. Wang, J. Li, S.W. Chiang, C. Miao, F. Kang, [105] E. Barsoukov, J.H. Kim, C.O. Yoon, H. Lee, Universal battery parameterization to
J. Zhang, Deterioration of lithium iron phosphate/graphite power batteries under yield a non-linear equivalent circuit valid for battery simulation at arbitrary load,
high-rate discharge cycling, Electrochim. Acta 176 (2015) 270–279, https://doi. J. Power Sources 83 (1–2) (1999) 61–70, https://doi.org/10.1016/S0378-7753
org/10.1016/j.electacta.2015.06.096. (99)00257-8.
[82] Y. Zheng, K. Qian, D. Luo, Y. Li, Q. Lu, B. Li, Y.B. He, X. Wang, J. Li, F. Kang, [106] T. Osaka, T. Momma, D. Mukoyama, H. Nara, Proposal of novel equivalent circuit
Influence of over-discharge on the lifetime and performance of LiFePO4/graphite for electrochemical impedance analysis of commercially available lithium ion
batteries, RSC Adv. 6 (36) (2016) 30474–30483, https://doi.org/10.1039/ battery, J. Power Sources 205 (2012) 483–486, https://doi.org/10.1016/j.
c6ra01677d. jpowsour.2012.01.070.
[83] V.J. Ovejas, A. Cuadras, Impedance characterization of an LCO-NMC/graphite [107] T.K. Dong, M. Montaru, A. Kirchev, M. Perrin, F. Lambert, Y. Bultel, Modeling of
cell: ohmic conduction, sei transport and charge-transfer phenomenon, Batteries lithium iron phosphate batteries by an equivalent electrical circuit: Part II - model
4 (3) (2018), https://doi.org/10.3390/batteries4030043. parameterization as function of power and State of Energy (SOE), ECS
[84] R. Premanand, A. Durairajan, B. Haran, R. White, B. Popov, Studies on capacity Transactions 35 (32) (2011) 229–237, https://doi.org/10.1149/1.3655706.
fade of spinel-based Li-ion batteries, J. Electrochem. Soc. 149 (1) (2002) A54, [108] T.K. Dong, A. Kirchev, F. Mattera, Y. Bultel, Modeling of lithium iron phosphate
https://doi.org/10.1149/1.1426399. batteries by an equivalent-electrical circuit: method of model parameterization
[85] R. Scipioni, P.S. Jørgensen, C. Graves, J. Hjelm, S.H. Jensen, A physically-based and simulation, ECS Transactions 25 (35) (2010) 131–138, https://doi.org/
equivalent circuit model for the impedance of a LiFePO4/graphite 26650 10.1149/1.3414011.
cylindrical cell, J. Electrochem. Soc. 164 (9) (2017) A2017–A2030, https://doi. [109] A. Farmann, D.U. Sauer, Comparative study of reduced order equivalent circuit
org/10.1149/2.1071709jes. models for on-board state-of-available-power prediction of lithium-ion batteries
[86] G. Sikha, P. Ramadass, B.S. Haran, R.E. White, B.N. Popov, Comparison of the in electric vehicles, Appl. Energy 225 (2018) 1102–1122, https://doi.org/
capacity fade of Sony US 18650 cells charged with different protocols, J. Power 10.1016/j.apenergy.2018.05.066. April.
Sources 122 (1) (2003) 67–76, https://doi.org/10.1016/S0378-7753(03)00027- [110] E. Samadani, S. Farhad, W. Scott, M. Mastali, L.E. Gimenez, M. Fowler, R.
2. A. Fraser, Empirical modeling of lithium-ion batteries based on electrochemical
[87] D.P. Abraham, J.L. Knuth, D.W. Dees, I. Bloom, J.P. Christophersen, Performance impedance spectroscopy tests, Electrochim. Acta 160 (2015) 169–177, https://
degradation of high-power lithium-ion cells—electrochemistry of harvested doi.org/10.1016/j.electacta.2015.02.021.
electrodes, J. Power Sources 170 (2) (2007) 465–475, https://doi.org/10.1016/j. [111] S.E. Li, B. Wang, H. Peng, X. Hu, An electrochemistry-based impedance model for
jpowsour.2007.03.071. lithium-ion batteries, J. Power Sources 258 (2014) 9–18, https://doi.org/
[88] T. Momma, T. Yokoshima, H. Nara, Y. Gima, T. Osaka, Distinction of impedance 10.1016/j.jpowsour.2014.02.045.
responses of Li-ion batteries for individual electrodes using symmetric cells, [112] T. Kalogiannis, M.S. Hosen, M.A. Sokkeh, S. Goutam, J. Jaguemont, L. Jin,
Electrochim. Acta 131 (2014) 195–201, https://doi.org/10.1016/j. G. Qiao, M. Berecibar, J. Van Mierlo, Comparative study on parameter
electacta.2014.01.091. identification methods for dual-polarization lithium-ion equivalent circuit model,
[89] G. Nagasubramanian, Two- and three-electrode impedance studies on 18650 Li- Energies 12 (21) (2019) 1–35, https://doi.org/10.3390/en12214031.
ion cells, J. Power Sources 87 (1) (2000) 226–229, https://doi.org/10.1016/ [113] M. Ecker, J.B. Gerschler, J. Vogel, S. Käbitz, F. Hust, P. Dechent, D.U. Sauer,
S0378-7753(99)00469-3. Development of a lifetime prediction model for lithium-ion batteries based on
[90] P. Shafiei Sabet, A.J. Warnecke, F. Meier, H. Witzenhausen, E. Martinez-Laserna, extended accelerated aging test data, J. Power Sources 215 (2012) 248–257,
D.U. Sauer, Non-invasive yet separate investigation of anode/cathode https://doi.org/10.1016/j.jpowsour.2012.05.012.
degradation of lithium-ion batteries (nickel–cobalt–manganese vs. graphite) due
21
[114] T.K. Dong, A. Kirchev, F. Mattera, J. Kowal, Y. Bultel, Dynamic modeling of Li-ion [136] C. Pastor-Fernández, T.F. Yu, W.D. Widanage, J. Marco, Critical review of non-
batteries using an equivalent electrical circuit, J. Electrochem. Soc. 158 (3) invasive diagnosis techniques for quantification of degradation modes in lithium-
(2011) 326–336, https://doi.org/10.1149/1.3543710. ion batteries, Renew. Sustain. Energy Rev. 109 (2019) 138–159, https://doi.org/
[115] B.Y. Liaw, G. Nagasubramanian, R.G. Jungst, D.H. Doughty, Modeling of lithium 10.1016/j.rser.2019.03.060. August 2018.
ion cells - a simple equivalent-circuit model approach, Solid State Ionics 175 [137] D. Mukoyama, T. Momma, H. Nara, T. Osaka, Electrochemical impedance
(1–4) (2004) 835–839, https://doi.org/10.1016/j.ssi.2004.09.049. analysis on degradation of commercially available lithium ion battery during
[116] M. Greenleaf, H. Li, J.P. Zheng, Application of physical electric circuit modeling chargedischarge cycling, Chem. Lett. 41 (4) (2012) 444–446, https://doi.org/
to characterize Li-ion battery electrochemical processes, J. Power Sources 270 10.1246/cl.2012.444.
(2014) 113–120, https://doi.org/10.1016/j.jpowsour.2014.07.083. [138] L. Zhang, J. Jiang, W. Zhang, Capacity Decay Mechanism of the LCO + NMC532/
[117] P. Schröer, H. van Faassen, T. Nemeth, M. Kuipers, D.U. Sauer, Challenges in Graphite cells combined with post-mortem technique, Energies 10 (8) (2017),
modeling high power lithium titanate oxide cells in battery management systems, https://doi.org/10.3390/en10081147.
J. Energy Storage 28 (2020) 101189, https://doi.org/10.1016/j.est.2019.101189. [139] R. Scipioni, P.S. Jørgensen, D.I. Stroe, R. Younesi, S.B. Simonsen, P. Norby,
November 2019. J. Hjelm, S.H. Jensen, Complementary analyses of aging in a commercial
[118] R. Gopalakrishnan, Y. Li, J. Smekens, A. Barhoum, G. Van Assche, N. Omar, J. Van LiFePO4/graphite 26650 cell, Electrochim. Acta 284 (2018) 454–468, https://
Mierlo, Electrochemical impedance spectroscopy characterization and doi.org/10.1016/j.electacta.2018.07.124.
parameterization of lithium nickel manganese cobalt oxide pouch cells: [140] J.H. Choi, H. Kim, J.H. Lim, S.J. Kwon, E.Y. Park, K.T. Lee, Thermal behavior of
dependency analysis of temperature and state of charge, Ionics 25 (1) (2019) Ni-rich layered oxide cathode materials during cycling of 20 Ah-scale Li-ion
111–123, https://doi.org/10.1007/s11581-018-2595-2. batteries, J. Electrochem. Soc. 165 (16) (2018) A3837–A3843, https://doi.org/
[119] X. Wang, X. Wei, H. Dai, Q. Wu, State estimation of lithium ion battery based on 10.1149/2.0501816jes.
electrochemical impedance spectroscopy with on-board impedance measurement [141] F. Hall, J. Touzri, S. Wußler, H. Buqa, W.G. Bessler, Experimental investigation of
system, in: IEEE Vehicle Power and Propulsion Conference (VPPC), IEEE, 2015, the thermal and cycling behavior of a lithium titanate-based lithium-ion pouch
pp. 1–5, https://doi.org/10.1109/VPPC.2015.7353021, 2015. cell, J. Energy Storage 17 (2018) 109–117, https://doi.org/10.1016/j.
[120] H. Dai, B. Jiang, X. Wei, Impedance characterization and modeling of lithium-ion est.2018.02.012.
batteries considering the internal temperature gradient, Energies 11 (1) (2018), [142] A. Matasso, D. Wong, D. Wetz, F. Liu, Correlation of bulk internal pressure rise
https://doi.org/10.3390/en11010220. and capacity degradation of commercial LiCoO2 cells, J. Electrochem. Soc. 161
[121] X. Zhu, L. Fernández Macía, J. Jaguemont, J. de Hoog, A. Nikolian, N. Omar, (14) (2014) A2031–A2035, https://doi.org/10.1149/2.0221414jes.
A. Hubin, Electrochemical impedance study of commercial [143] S. Saxena, Y. Xing, M. Pecht, A unique failure mechanism in the Nexus 6P lithium-
LiNi0.80Co0.15Al0.05O2 electrodes as a function of state of charge and aging, ion battery, Energies 11 (4) (2018), https://doi.org/10.3390/en11040841.
Electrochim. Acta 287 (2018) 10–20, https://doi.org/10.1016/j. [144] M. Petzl, M. Kasper, M.A. Danzer, Lithium plating in a commercial lithium-ion
electacta.2018.08.054. battery - a low-temperature aging study, J. Power Sources 275 (2015) 799–807,
[122] C. Pastor-Fernández, K. Uddin, G.H. Chouchelamane, W.D. Widanage, J. Marco, https://doi.org/10.1016/j.jpowsour.2014.11.065.
A comparison between electrochemical impedance spectroscopy and incremental [145] P.L. Moss, G. Au, E.J. Plichta, J.P. Zheng, Study of capacity fade of lithium-ion
capacity-differential voltage as Li-ion diagnostic techniques to identify and polymer rechargeable batteries with continuous cycling, J. Electrochem. Soc. 157
quantify the effects of degradation modes within battery management systems, (1) (2010) 1–7, https://doi.org/10.1149/1.3246001.
J. Power Sources 360 (2017) 301–318, https://doi.org/10.1016/j. [146] H. Ishikawa, O. Mendoza, Y. Sone, M. Umeda, Study of thermal deterioration of
jpowsour.2017.03.042. lithium-ion secondary cell using an accelerated rate calorimeter (ARC) and AC
[123] D.A. Howey, V. Yufit, P.D. Mitcheson, G.J. Offer, N.P. Brandon, Impedance impedance method, J. Power Sources 198 (2012) 236–242, https://doi.org/
measurement for advanced battery management systems, in: 2013 World Electric 10.1016/j.jpowsour.2011.09.067.
Vehicle Symposium and Exhibition, EVS, 2014, pp. 1–7, https://doi.org/ [147] P. Röder, B. Stiaszny, J.C. Ziegler, N. Baba, P. Lagaly, H.D. Wiemhöfer, The
10.1109/EVS.2013.6914960, 2014. impact of calendar aging on the thermal stability of a LiMn 2O4-Li(Ni1/3Mn1/
[124] P. Liu, J. Wang, J. Hicks-Garner, E. Sherman, S. Soukiazian, M. Verbrugge, 3Co 1/3)O2/graphite lithium-ion cell, J. Power Sources 268 (2014) 315–325,
H. Tataria, J. Musser, P. Finamore, Aging Mechanisms of LiFePO4 batteries https://doi.org/10.1016/j.jpowsour.2014.06.040.
deduced by electrochemical and structural analyses, J. Electrochem. Soc. 157 (4) [148] J. Li, J. Zhang, X. Zhang, C. Yang, N. Xu, B. Xia, Study of the storage performance
(2010), https://doi.org/10.1149/1.3294790. of a Li-ion cell at elevated temperature, Electrochim. Acta 55 (3) (2010) 927–934,
[125] J. Groot, M. Swierczynski, A.I. Stan, S.K. Kær, On the complex ageing https://doi.org/10.1016/j.electacta.2009.09.077.
characteristics of high-power LiFePO4/graphite battery cells cycled with high [149] K. Jalkanen, J. Karppinen, L. Skogström, T. Laurila, M. Nisula, K. Vuorilehto,
charge and discharge currents, J. Power Sources 286 (2015) 475–487, https:// Cycle aging of commercial NMC/graphite pouch cells at different temperatures,
doi.org/10.1016/j.jpowsour.2015.04.001. Appl. Energy 154 (2015) 160–172, https://doi.org/10.1016/j.
[126] K. Goebel, B. Saha, A. Saxena, J.R. Celaya, J.P. Christophersen, Prognostics in apenergy.2015.04.110.
battery health management, IEEE Instrum. Meas. Mag. 11 (4) (2008) 33–40, [150] W. Cao, J. Li, Z. Wu, Cycle-life and degradation mechanism of LiFePO4-based
https://doi.org/10.1109/MIM.2008.4579269. lithium-ion batteries at room and elevated temperatures, Ionics 22 (10) (2016)
[127] J. Zhu, M.S. Dewi Darma, M. Knapp, D.R. Sørensen, M. Heere, Q. Fang, X. Wang, 1791–1799, https://doi.org/10.1007/s11581-016-1703-4.
H. Dai, L. Mereacre, A. Senyshyn, X. Wei, H. Ehrenberg, Investigation of lithium- [151] R. Genieser, S. Ferrari, M. Loveridge, S.D. Beattie, R. Beanland, H. Amari, G. West,
ion battery degradation mechanisms by combining differential voltage analysis R. Bhagat, Lithium ion batteries (NMC/graphite) cycling at 80 ∘C: different
and alternating current impedance, J. Power Sources 448 (2020) 28–30, https:// electrolytes and related degradation mechanism, J. Power Sources 373 (2018)
doi.org/10.1016/j.jpowsour.2019.227575. December 2019. 172–183, https://doi.org/10.1016/j.jpowsour.2017.11.01.
[128] E. Cuervo-Reyes, C.P. Scheller, M. Held, U. Sennhauser, A unifying view of the [152] B.S. Haran, P. Ramadass, R.E. White, B.N. Popov, Capacity fade of Li-ion cells
constant-phase-element and its role as an aging indicator for Li-ion batteries, cycled at different temperatures, Proceedings of the Annual Battery Conference
J. Electrochem. Soc. 162 (8) (2015) A1585–A1591, https://doi.org/10.1149/ on Applications and Advances January (2002) 13–18, https://doi.org/10.1109/
2.0791508jes. BCAA.2002.986361.
[129] B. Stiaszny, J.C. Ziegler, E.E. Krauß, M. Zhang, J.P. Schmidt, E. Ivers-Tiffée, [153] Y. Zhang, C.Y. Wang, X. Tang, Cycling degradation of an automotive LiFePO4
Electrochemical characterization and post-mortem analysis of aged LiMn 2O4- lithium-ion battery, J. Power Sources 196 (3) (2011) 1513–1520, https://doi.org/
NMC/graphite lithium ion batteries part II: calendar aging, J. Power Sources 258 10.1016/j.jpowsour.2010.08.070.
(2014) 61–75, https://doi.org/10.1016/j.jpowsour.2014.02.019. [154] T. Rauhala, K. Jalkanen, T. Romann, E. Lust, N. Omar, T. Kallio, Low-temperature
[130] J.C. Burns, A. Kassam, N.N. Sinha, L.E. Downie, L. Solnickova, B.M. Way, J. aging mechanisms of commercial graphite/LiFePO4 cells cycled with a simulated
R. Dahn, Predicting and extending the lifetime of Li-ion batteries, J. Electrochem. electric vehicle load profile—a post-mortem study, J. Energy Storage 20 (2018)
Soc. 160 (9) (2013) 1–5, https://doi.org/10.1149/2.060309jes. 344–356, https://doi.org/10.1016/j.est.2018.10.007. June.
[131] A. Maheshwari, M. Heck, M. Santarelli, Cycle aging studies of lithium nickel [155] T. Matsuda, K. Ando, M. Myojin, M. Matsumoto, T. Sanada, N. Takao, H. Imai,
manganese cobalt oxide-based batteries using electrochemical impedance D. Imamura, Investigation of the influence of temperature on the degradation
spectroscopy, Electrochim. Acta 273 (2018) 335–348, https://doi.org/10.1016/j. mechanism of commercial nickel manganese cobalt oxide-type lithium-ion cells
electacta.2018.04.045. during long-term cycle tests, J. Energy Storage 21 (2019) 665–671, https://doi.
[132] Y. Liu, J. Xie, Failure study of commercial LiFePO4 cells in overcharge conditions org/10.1016/j.est.2019.01.009. March 2018.
using electrochemical impedance spectroscopy, J. Electrochem. Soc. 162 (10) [156] Y. Wu, P. Keil, S.F. Schuster, A. Jossen, Impact of temperature and discharge rate
(2015) A2208–A2217, https://doi.org/10.1149/2.0911510jes. on the aging of a LiCoO2/LiNi0.8Co0.15Al0.05O2 lithium-ion pouch cell,
[133] Y. Zhang, Q. Tang, Y. Zhang, J. Wang, U. Stimming, A.A. Lee, Identifying J. Electrochem. Soc. 164 (7) (2017) A1438–A1445, https://doi.org/10.1149/
degradation patterns of lithium ion batteries from impedance spectroscopy using 2.0401707jes.
machine learning, Nat. Commun. 11 (1) (2020) 1706, https://doi.org/10.1038/ [157] D.N. Wong, D.A. Wetz, J.M. Heinzel, A.N. Mansour, Characterizing rapid capacity
s41467-020-15235-7. fade and impedance evolution in high rate pulsed discharged lithium iron
[134] S.F. Schuster, T. Bach, E. Fleder, J. Mueller, M. Brand, G. Sextl, A. Jossen, phosphate cells for complex, high power loads, J. Power Sources 328 (2016)
Nonlinear aging characteristics of lithium-ion cells under different operational 81–90, https://doi.org/10.1016/j.jpowsour.2016.08.013.
conditions, J. Energy Storage 1 (1) (2015) 44–53, https://doi.org/10.1016/j. [158] Y. Leng, S. Ge, D. Marple, X.G. Yang, C. Bauer, P. Lamp, C.Y. Wang,
est.2015.05.003. Electrochemical cycle-life characterization of high energy lithium-ion cells with
[135] S. Schindler, M. Bauer, M. Petzl, M.A. Danzer, Voltage relaxation and impedance thick Li(Ni0.6Mn0.2Co0.2)O2 and graphite electrodes, J. Electrochem. Soc. 164
spectroscopy as in-operando methods for the detection of lithium plating on (6) (2017) A1037–A1049, https://doi.org/10.1149/2.0451706jes.
graphitic anodes in commercial lithium-ion cells, J. Power Sources 304 (2016) [159] B. Shrestha, P.M. Novak, D.A. Wetz, A.F. Matasso, Evaluation of high C rate cycle
170–180, https://doi.org/10.1016/j.jpowsour.2015.11.044. induced aging on low impedance lithium-ion batteries using in-situ
22
electrochemical impedance spectroscopy (EIS) analysis, ECS Transactions 58 (48) [183] A. Bessman, R. Soares, O. Wallmark, P. Svens, G. Lindbergh, Aging effects of AC
(2013) 207–222, https://doi.org/10.1149/05848.0207ecst. harmonics on lithium-ion cells, J. Energy Storage 21 (2019) 741–749, https://doi.
[160] S. Schindler, M. Bauer, H. Cheetamun, M.A. Danzer, Fast charging of lithium-ion org/10.1016/j.est.2018.12.016. December 2018.
cells: identification of aging-minimal current profiles using a design of experiment [184] K. Uddin, A.D. Moore, A. Barai, J. Marco, The effects of high frequency current
approach and a mechanistic degradation analysis, J. Energy Storage 19 (2018) ripple on electric vehicle battery performance, Appl. Energy 178 (2016) 142–154,
364–378, https://doi.org/10.1016/j.est.2018.08.002. March. https://doi.org/10.1016/j.apenergy.2016.06.033.
[161] D. Wong, B. Shrestha, D.A. Wetz, J.M. Heinzel, Impact of high rate discharge on [185] M.J. Brand, M.H. Hofmann, S.S. Schuster, P. Keil, A. Jossen, The influence of
the aging of lithium nickel cobalt aluminum oxide batteries, J. Power Sources 280 current ripples on the lifetime of lithium-ion batteries, IEEE Trans. Veh. Technol.
(2015) 363–372, https://doi.org/10.1016/j.jpowsour.2015.01.110. 67 (11) (2018) 10438–10445, https://doi.org/10.1109/TVT.2018.2869982.
[162] G. Ning, B. Haran, B.N. Popov, Capacity fade study of lithium-ion batteries cycled [186] B.V. Ratnakumar, M.C. Smart, L.D. Whitcanack, E.D. Davies, K.B. Chin,
at high discharge rates, J. Power Sources 117 (1) (2003) 160–169, https://doi. F. Deligiannis, S. Surampudi, Behavior of Li-ion cells in high-intensity radiation
org/10.1016/S0378-7753(03)00029-6. environments II. Sony/AEA/ComDEV cells, J. Electrochem. Soc. 152 (2) (2005)
[163] M. Abdel Monem, K. Trad, N. Omar, O. Hegazy, B. Mantels, G. Mulder, P. Van den 357–363, https://doi.org/10.1149/1.1848212.
Bossche, J. Van Mierlo, Lithium-ion batteries: evaluation study of different [187] C. Mao, S.J. An, H.M. Meyer, J. Li, M. Wood, R.E. Ruther, D.L. Wood, Balancing
charging methodologies based on aging process, Appl. Energy 152 (2015) formation time and electrochemical performance of high energy lithium-ion
143–155, https://doi.org/10.1016/j.apenergy.2015.02.064. batteries, J. Power Sources 402 (2018) 107–115, https://doi.org/10.1016/j.
[164] J. Xia, M. Nie, L. Ma, J.R. Dahn, Variation of coulombic efficiency versus upper jpowsour.2018.09.019.
cutoff potential of Li-ion cells tested with aggressive protocols, J. Power Sources [188] K.W. Knehr, T. Hodson, C. Bommier, G. Davies, A. Kim, D.A. Steingart,
306 (2016) 233–240, https://doi.org/10.1016/j.jpowsour.2015.12.013. Understanding full-cell evolution and non-chemical electrode crosstalk of Li-ion
[165] K.J. Nelson, J.E. Harlow, J.R. Dahn, A comparison of NMC/graphite pouch cells batteries, Joule 2 (6) (2018) 1146–1159, https://doi.org/10.1016/j.
and commercially available LiCoO2/graphite pouch cells tested to high potential, joule.2018.03.016.
J. Electrochem. Soc. 165 (3) (2018) A456–A462, https://doi.org/10.1149/ [189] M. Wünsch, J. Kaufman, D.U. Sauer, Investigation of the influence of different
2.0041803jes. bracing of automotive pouch cells on cyclic liefetime and impedance spectra,
[166] J. Liu, Q. Duan, M. Ma, C. Zhao, J. Sun, Q. Wang, Aging mechanisms and thermal J. Energy Storage 21 (2019) 149–155, https://doi.org/10.1016/j.
stability of aged commercial 18650 lithium ion battery induced by slight est.2018.11.019. October 2018.
overcharging cycling, J. Power Sources 445 (2020) 227263, https://doi.org/ [190] A. Barai, Y. Guo, A. McGordon, P. Jennings, A study of the effects of external
10.1016/j.jpowsour.2019.227263. July 2019. pressure on the electrical performance of a lithium-ion pouch cell, in:
[167] H. Maleki, J.N. Howard, Effects of overdischarge on performance and thermal International Conference on Connected Vehicles and Expo, ICCVE 2013 -
stability of a Li-ion cell, J. Power Sources 160 (2) (2006) 1395–1402, https://doi. Proceedings, 2013, pp. 295–299, https://doi.org/10.1109/ICCVE.2013.6799809,
org/10.1016/j.jpowsour.2006.03.043. 2013.
[168] D. Kong, R. Wen, P. Ping, R. Peng, J. Zhang, G. Chen, Study on degradation [191] Y. Olofsson, J. Groot, T. Katrašnik, G. Tavčar, Impedance spectroscopy
behavior of commercial 18650 LiAlNiCoO 2 cells in over-charge conditions, Int. J. characterisation of automotive NMC/graphite Li-ion cells aged with realistic
Energy Res. 43 (1) (2019) 552–567, https://doi.org/10.1002/er.4302. PHEV load profile, in: IEEE International Electric Vehicle Conference, IEVC 2014,
[169] Y. Zheng, Y.B. He, K. Qian, B. Li, X. Wang, J. Li, C. Miao, F. Kang, Effects of state 2014, https://doi.org/10.1109/IEVC.2014.7056095, 2015.
of charge on the degradation of LiFePO4/graphite batteries during accelerated [192] A.S. Mussa, M. Klett, M. Behm, G. Lindbergh, R.W. Lindström, Fast-charging to a
storage test, J. Alloys Compd. 639 (2015) 406–414, https://doi.org/10.1016/j. partial state of charge in lithium-ion batteries: a comparative ageing study,
jallcom.2015.03.169. J. Energy Storage 13 (2017) 325–333, https://doi.org/10.1016/j.
[170] A. Barai, K. Uddin, J. Chevalier, G.H. Chouchelamane, A. McGordon, J. Low, est.2017.07.004.
P. Jennings, Transportation safety of lithium iron phosphate batteries - a [193] M. Mathew, S. Janhunen, M. Rashid, F. Long, M. Fowler, Comparative analysis of
feasibility study of storing at very low states of charge, Sci. Rep. 7 (1) (2017) lithium-ion battery resistance estimation techniques for battery management
1–10, https://doi.org/10.1038/s41598-017-05438-2. systems, Energies 11 (6) (2018) 121693718, https://doi.org/10.3390/
[171] Y. Liu, Q. Liu, Z. Li, Y. Ren, J. Xie, H. He, F. Xu, Failure study of commercial en11061490.
LiFePO4 cells in over-discharge conditions using electrochemical impedance [194] B. Liebhart, L. Komsiyska, C. Endisch, Passive impedance spectroscopy for
spectroscopy, J. Electrochem. Soc. 161 (4) (2014) A620–A632, https://doi.org/ monitoring lithium-ion battery cells during vehicle operation, J. Power Sources
10.1149/2.090404jes. 449 (2020) 227297, https://doi.org/10.1016/j.jpowsour.2019.227297. October
[172] J. Schmitt, A. Maheshwari, M. Heck, S. Lux, M. Vetter, Impedance change and 2019.
capacity fade of lithium nickel manganese cobalt oxide-based batteries during [195] H. Zappen, F. Ringbeck, D.U. Sauer, Application of time-resolved multi-sine
calendar aging, J. Power Sources 353 (2017) 183–194, https://doi.org/10.1016/ impedance spectroscopy for lithium-ion battery characterization, Batteries 4 (4)
j.jpowsour.2017.03.090. (2018) 1–18, https://doi.org/10.3390/batteries4040064.
[173] C.T. Love, K. Swider-Lyons, Impedance diagnostic for overcharged lithium-ion [196] N. Lohmann, P. Weßkamp, P. Haußmann, J. Melbert, T. Musch, Electrochemical
batteries, Electrochem. Solid State Lett. 15 (4) (2012) 53–56, https://doi.org/ impedance spectroscopy for lithium-ion cells: test equipment and procedures for
10.1149/2.014204esl. aging and fast characterization in time and frequency domain, J. Power Sources
[174] R.P. Ramasamy, R.E. White, B.N. Popov, Calendar life performance of pouch 273 (2015) 613–623, https://doi.org/10.1016/j.jpowsour.2014.09.132.
lithium-ion cells, J. Power Sources 141 (2) (2005) 298–306, https://doi.org/ [197] X. Hong, N. Li, J. Feng, Q. Kong, G. Liu, Multi-electrode resistivity probe for
10.1016/j.jpowsour.2004.09.024. investigation of local temperature inside metal shell battery cells via resistivity:
[175] J. Zhu, Z. Sun, X. Wei, H. Dai, W. Gu, Experimental investigations of an AC pulse experiments and evaluation of electrical resistance tomography, Energies 8 (2)
heating method for vehicular high power lithium-ion batteries at subzero (2015) 742–764, https://doi.org/10.3390/en8020742.
temperatures, J. Power Sources 367 (2017) 145–157, https://doi.org/10.1016/j. [198] M. Fleckenstein, S. Fischer, O. Bohlen, B. Bäker, Thermal Impedance Spectroscopy
jpowsour.2017.09.063. - a method for the thermal characterization of high power battery cells, J. Power
[176] J. Jiang, H. Ruan, B. Sun, L. Wang, W. Gao, W. Zhang, A low-temperature internal Sources 223 (2013) 259–267, https://doi.org/10.1016/j.jpowsour.2012.07.144.
heating strategy without lifetime reduction for large-size automotive lithium-ion [199] E. Barsoukov, J.H. Jang, H. Lee, Thermal impedance spectroscopy for Li-ion
battery pack, Appl. Energy 230 (2018) 257–266, https://doi.org/10.1016/j. batteries using heat-pulse response analysis, J. Power Sources 109 (2) (2002)
apenergy.2018.08.070. August. 313–320, https://doi.org/10.1016/S0378-7753(02)00080-0.
[177] Q. Kellner, D. Worwood, A. Barai, W.D. Widanage, J. Marco, Duty-cycle [200] P. Keil, K. Rumpf, A. Jossen, Thermal impedance spectroscopy for Li-ion batteries
characterisation of large-format automotive lithium ion pouch cells for high with an IR temperature sensor system, World Electric Vehicle Journal 6 (3)
performance vehicle applications, J. Energy Storage 19 (2018) 170–184, https:// (2013) 581–591, https://doi.org/10.3390/wevj6030581.
doi.org/10.1016/j.est.2018.07.018. May. [201] M. Swierczynski, D.I. Stroe, T. Stanciu, S.K. Kær, Electrothermal impedance
[178] M. Reichert, D. Andre, A. Rösmann, P. Janssen, H.G. Bremes, D.U. Sauer, spectroscopy as a cost efficient method for determining thermal parameters of
S. Passerini, M. Winter, Influence of relaxation time on the lifetime of commercial lithium ion batteries: prospects, measurement methods and the state of
lithium-ion cells, J. Power Sources 239 (2013) 45–53, https://doi.org/10.1016/j. knowledge, J. Clean. Prod. 155 (2017) 63–71, https://doi.org/10.1016/j.
jpowsour.2013.03.053. jclepro.2016.09.109.
[179] M. Spielbauer, P. Berg, M. Ringat, O. Bohlen, A. Jossen, Experimental study of the [202] J.P. Schmidt, A. Weber, E. Ivers-Tiffée, A novel and precise measuring method for
impedance behavior of 18650 lithium-ion battery cells under deforming the entropy of lithium-ion cells: δs via electrothermal impedance spectroscopy,
mechanical abuse, J. Energy Storage 26 (November) (2019) 101039, https://doi. Electrochim. Acta 137 (1) (2014) 311–319, https://doi.org/10.1016/j.
org/10.1016/j.est.2019.101039. electacta.2014.05.153.
[180] M.J. Brand, S.F. Schuster, T. Bach, E. Fleder, M. Stelz, S. Gläser, J. Müller, [203] L. Palafox, F. Raso, J. Ku, F. Overney, L. Callegaro, P. Gournay, A. Zio, J. Nissilä,
G. Sextl, A. Jossen, Effects of vibrations and shocks on lithium-ion cells, J. Power G. Eklund, T. Lippert, AIM QuTE : automated impedance metrology extending the
Sources 288 (2015) 62–69, https://doi.org/10.1016/j.jpowsour.2015.04.107. quantum toolbox for electricity, in: 16th International Congress of Metrology
[181] J.M. Hooper, J. Marco, G.H. Chouchelamane, C. Lyness, Vibration durability 11001, 2013, pp. 1–3, https://doi.org/10.1051/metrology/201311001.
testing of nickel manganese cobalt oxide (NMC) lithium-ion 18,650 battery cells, [204] LiBforSecUse, Quality assessment of electric vehicle Li ion batteries for second use
Energies 9 (1) (2016) 1–27, https://doi.org/10.3390/en9010052. applications. https://www.ptb.de/empir2018/libforsecuse/home/.
[182] P.K.P. Ferraz, J. Kowal, A comparative study on the influence of DC/DC-converter [205] BIPM, IEC, IFCC, LAC, IUPAC, IUPAP, ISO, OIMLl, The international vocabulary
induced high frequency current Ripple on Lithium-Ion Batteries, Sustainability 11 of metrology—basic and general concepts and associated terms (VIM). www.
(21) (2019) 6050, https://doi.org/10.3390/su11216050. bipm.org/utils/common/documents/jcgm/JCGM_200_2012.pdf, 2012.
[206] L. Callegaro, Traceable measurements of electrical impedance, IEEE Instrum.
Meas. Mag. 18 (6) (2015) 42–46, https://doi.org/10.1109/MIM.2015.7335839.
23
[207] S.A. Awan, B. Kibble, J. Schurr, Coaxial Electrical Circuits for Interference-free [234] M. Oldenburger, B. Bedürftig, A. Gruhle, F. Grimsmann, E. Richter, R. Findeisen,
Measurements, IET Electrical Measurement Series, Institution of Engineering and A. Hintennach, Investigation of the low frequency Warburg impedance of Li-ion
Technology, London, Unitied Kingdom, 2011, https://doi.org/10.1049/ cells by frequency domain measurements, J. Energy Storage 21 (2019) 272–280,
PBEL013E. https://doi.org/10.1016/j.est.2018.11.029. December 2018.
[208] L. Callegaro, Electrical Impedance. Principles, Measurement, and Application, [235] M. Messing, T. Shoa, S. Habibi, Lithium-Ion Battery Relaxation Effects, ITEC 2019
Series in Sensors, Taylor & Francis, 2013, https://doi.org/10.1201/b13069. - 2019 IEEE Transportation Electrification Conference and Expo, 2019, https://
[209] R.D. Cutkosky, Four-terminal-pair networks as precision admittance and doi.org/10.1109/ITEC.2019.8790449.
impedance standards, IEEE Transactions on Communication and Electronics 83 [236] F.M. Kindermann, A. Noel, S.V. Erhard, A. Jossen, Long-term equalization effects
(70) (1964) 19–22, https://doi.org/10.1109/TCOME.1964.6539563. in Li-ion batteries due to local state of charge inhomogeneities and their impact
[210] B. Kibble, Four terminal-pair to anything else!, in: IEE Colloquium on impedance measurements, Electrochim. Acta 185 (2015) 107–116, https://
Interconnections from DC to Microwaves 1999, IEE, 1999, pp. 1–6, https://doi. doi.org/10.1016/j.electacta.2015.10.108.
org/10.1049/ic:19990102. [237] J.P. Schmidt, E. Ivers-Tiffée, Pulse-fitting - a novel method for the evaluation of
[211] B.P. Kibble, Proposals for extending traceable impedance measurements to higher pulse measurements, demonstrated for the low frequency behavior of lithium-ion
frequencies, Metrologia 35 (1) (1998) 17–20, https://doi.org/10.1088/0026- cells, J. Power Sources 315 (2016) 316–323, https://doi.org/10.1016/j.
1394/35/1/3. jpowsour.2016.03.026.
[212] B. Hague, T.R. Foord, Alternating Current Bridge Methods, Pitman Publishing, [238] Y. Fernández Pulido, C. Blanco, D. Anseán, V.M. García, F. Ferrero, M. Valledor,
1971. Determination of suitable parameters for battery analysis by electrochemical
[213] G.H. Rayner, The derivation of resistance, inductance, and capacitance from the impedance spectroscopy, Measurement, Journal of the International
NPL primary standard of mutual inductance, IRE Transactions on Instrumentation Measurement Confederation 106 (1) (2017) 1–11, https://doi.org/10.1016/j.
I- 7 (3 \& 4) (1958) 212–220, https://doi.org/10.1109/IRE-I.1958.5006792. measurement.2017.04.022.
[214] G.H. Rayner, B.P. Kibble, M.J. Swan, On obtaining the henry from the farad, NPL [239] A. Barai, G.H. Chouchelamane, Y. Guo, A. McGordon, P. Jennings, A study on the
Report DES 63 (1980). impact of lithium-ion cell relaxation on electrochemical impedance spectroscopy,
[215] A. Campbell, On a standard of mutual inductance, in: Proceedings of the Royal J. Power Sources 280 (2015) 74–80, https://doi.org/10.1016/j.
Society A: Mathematical, Physical and Engineering Sciences 79, 1907, jpowsour.2015.01.097.
pp. 428–435, https://doi.org/10.1098/rspa.1907.0053, 532. [240] Y. Troxler, B. Wu, M. Marinescu, V. Yufit, Y. Patel, A.J. Marquis, N.P. Brandon, G.
[216] A.M. Thompson, D. Lampard, A new theorem in electrostatics and its application J. Offer, The effect of thermal gradients on the performance of lithium-ion
to calculable standards of capacitance, Nature 177 (4515) (1956) 888, https:// batteries, J. Power Sources 247 (2014) 1018–1025, https://doi.org/10.1016/j.
doi.org/10.1038/177888a0. –888. jpowsour.2013.06.084.
[217] B.N. Taylor, T.J. Witt, New international electrical reference standards based on [241] X. Wang, H. Dai, B. Jiang, Research on impedance and nonlinear characteristics of
the Josephson and quantum Hall effects, Metrologia 26 (1) (1989) 47–62, https:// lithium-ion batteries, in: IEEE Vehicle Power and Propulsion Conference, VPPC
doi.org/10.1088/0026-1394/26/1/004. 2018 - Proceedings, 2018, pp. 1–6, https://doi.org/10.1109/
[218] M. Stock, R. Davis, E. de Mirandés, M.J.T. Milton, The revision of the SI—the VPPC.2018.8604963, 2019.
result of three decades of progress in metrology, Metrologia 56 (2) (2019), [242] J. Huang, H. Ge, Z. Li, J. Zhang, Dynamic electrochemical impedance
022001, https://doi.org/10.1088/1681-7575/ab0013. spectroscopy of a three-electrode lithium-ion battery during pulse charge and
[219] F. Overney, B. Jeanneret, Impedance bridges: from wheatstone to Josephson, discharge, Electrochim. Acta 176 (2015) 311–320, https://doi.org/10.1016/j.
Metrologia 55 (5) (2018) S119–S134, https://doi.org/10.1088/1681-7575/ electacta.2015.07.017.
aacf6c. [243] J. Huang, Z. Li, J. Zhang, Dynamic electrochemical impedance spectroscopy
[220] L. Callegaro, V. D’Elia, M. Kampik, Dan Bee Kim, M. Ortolano, F. Pourdanesh, reconstructed from continuous impedance measurement of single frequency
Experiences with a two-terminal-pair digital impedance bridge, IEEE Transactions during charging/discharging, J. Power Sources 273 (2015) 1098–1102, https://
on Instrumentation and Measurement 64 (6) (2015) 1460–1465, https://doi.org/ doi.org/10.1016/j.jpowsour.2014.07.067.
10.1109/TIM.2015.2401192. [244] S. Ernst, T.P. Heins, N. Schlüter, U. Schröder, Capturing the current-overpotential
[221] L. Callegaro, V. D’Elia, M. Ortolano, F. Pourdanesh, A three-arm current nonlinearity of lithium-ion batteries by nonlinear electrochemical impedance
comparator bridge for impedance comparisons over the complex plane, IEEE spectroscopy (NLEIS) in charge and discharge direction, Frontiers in Energy
Transactions on Instrumentation and Measurement 64 (6) (2015) 1466–1471, Research 7 (2019) 1–13, https://doi.org/10.3389/fenrg.2019.00151. December.
https://doi.org/10.1109/TIM.2015.2398953. [245] Z. Stoynov, Nonstationary impedance spectroscopy, Electrochim. Acta 38 (14)
[222] F. Overney, N.E. Flowers-Jacobs, B. Jeanneret, A. Rüfenacht, A.E. Fox, J. (1993) 1919–1922, https://doi.org/10.1016/0013-4686(93)80315-Q.
M. Underwood, A.D. Koffman, S.P. Benz, Josephson-based full digital bridge for [246] M. Itagaki, N. Kobari, S. Yotsuda, K. Watanabe, S. Kinoshita, M. Ue, In situ
high-accuracy impedance comparisons, Metrologia 53 (4) (2016) 1045–1053, electrochemical impedance spectroscopy to investigate negative electrode of
https://doi.org/10.1088/0026-1394/53/4/1045. lithium-ion rechargeable batteries, J. Power Sources 135 (1) (2004) 255–261,
[223] S. Mašláň, M. Šíra, T. Skalická, T. Bergsten, Four-terminal pair digital sampling https://doi.org/10.1016/j.jpowsour.2004.04.004.
impedance bridge up to 1MHz, IEEE Transactions on Instrumentation and [247] M. Itagaki, N. Kobari, S. Yotsuda, K. Watanabe, S. Kinoshita, M. Ue, LiCoO2
Measurement 68 (6) (2019) 1860–1869, https://doi.org/10.1109/ electrode/electrolyte interface of Li-ion rechargeable batteries investigated by in
TIM.2019.2908649. situ electrochemical impedance spectroscopy, J. Power Sources 148 (2005)
[224] M. Ortolano, L. Palafox, J. Kučera, L. Callegaro, V. D’Elia, M. Marzano, 78–84, https://doi.org/10.1016/j.jpowsour.2005.02.007.
F. Overney, G. Gülmez, An international comparison of phase angle standards [248] N. Harting, N. Wolff, U. Krewer, Identification of lithium plating in lithium-ion
between the novel impedance bridges of CMI, INRIM and METAS, Metrologia 55 batteries using nonlinear frequency response analysis (NFRA), Electrochim. Acta
(4) (2018) 499–512, https://doi.org/10.1088/1681-7575/aabf24. 281 (2018) 378–385, https://doi.org/10.1016/j.electacta.2018.05.139.
[225] F. Overney, B. Jeanneret, Calibration of an LCR-meter at arbitrary phase Angles [249] D.A. Howey, P.D. Mitcheson, S. Member, V. Yufit, G.J. Offer, N.P. Brandon,
using a fully automated impedance simulator, IEEE Transactions on Online measurement of battery impedance using motor controller excitation,
Instrumentation and Measurement 66 (6) (2017) 1516–1523, https://doi.org/ IEEE Transaction on Vehicular Technology 63 (6) (2014) 2557–2566, https://doi.
10.1109/TIM.2017.2652500. org/10.1109/TVT.2013.2293597.
[226] A. Technologies, Using OPEN/SHORT/LOAD correction application note 346-3, [250] P.T. Wojcik, P. Agarwal, M.E. Orazem, A method for maintaining a constant
Tech. rep (1998). https://literature.cdn.keysight.com/litweb/pdf/5091-6553E. potential variation during galvanostatic regulation of electrochemical impedance
pdf?id=1000030303:epsg:dow. (Accessed 25 August 2020). measurements, Electrochim. Acta 41 (7) (1996) 977–983, https://doi.org/
[227] J.M. Torrents, R. Pallàs-Areny, Compensation of impedance meters when using an 10.1016/0013-4686(95)00428-9.
external front-end amplifier, IEEE Transactions on Instrumentation and [251] B. Stiaszny, J.C. Ziegler, E.E. Krauß, J.P. Schmidt, E. Ivers-Tiffée, Electrochemical
Measurement 51 (2) (2002) 310–313, https://doi.org/10.1109/19.997829. characterization and post-mortem analysis of aged LiMn 2O4-Li(Ni0.5Mn0.3Co
[228] D.N. Homan, Applications of coaxial chokes to a-c bridge circuits, J. Res. Natl. 0.2)O2/graphite lithium ion batteries. Part I: cycle aging, J. Power Sources 251
Bur. Stand. J. Res. Natl. Bur. Stand., Section C: Engineering and Instrumentation (2014) 439–450, https://doi.org/10.1016/j.jpowsour.2013.11.080.
72C (2) (1968) 161, https://doi.org/10.6028/jres.072C.011. [252] M.D. Murbach, V.W. Hu, D.T. Schwartz, Nonlinear electrochemical impedance
[229] H.W. Ott, Noise Reduction Techniques in Electronic Systems, second ed., 1988. spectroscopy of lithium-Ion batteries: experimental approach, analysis, and initial
[230] BioLogic, EC-Lab, Application note #49 EIS measurements: potentio or galvano findings, J. Electrochem. Soc. 165 (11) (2018) A2758–A2765, https://doi.org/
mode?. https://www.biologic.net/documents/potentio-or-galvano-eis-electroch 10.1149/2.0711811jes.
emistry-battery-application-note-49/, 2013 accessed: 2020-05-27. [253] M. Kiel, O. Bohlen, D.U. Sauer, Harmonic analysis for identification of
[231] Gamry Application Note, EIS measurement of a very low impedance lithium ion nonlinearities in impedance spectroscopy, Electrochim. Acta 53 (25) (2008)
battery. www.gamry.com/application-notes/EIS/eis-measurement-of-a-very-l 7367–7374, https://doi.org/10.1016/j.electacta.2008.01.089.
ow-impedance-lithium-ion-battery/, 2016. [254] J.J. Giner-Sanz, E.M. Ortega, V. Pérez-Herranz, Total harmonic distortion based
[232] L. Deleebeeck, S. Veltzé, Electrochemical impedance spectroscopy study of method for linearity assessment in electrochemical systems in the context of EIS,
commercial Li-ion phosphate batteries : a metrology perspective, Int. J. Energy Electrochim. Acta 186 (2015) 598–612, https://doi.org/10.1016/j.
Res. (2020) 1–25, https://doi.org/10.1002/er.5350. electacta.2015.10.152.
[233] T. Stanciu, D.I. Stroe, R. Teodorescu, M. Swierczynski, Extensive EIS [255] Y. Firouz, R. Relan, J.M. Timmermans, N. Omar, P. Van den Bossche, J. Van
characterization of commercially available lithium polymer battery cell for Mierlo, Advanced lithium ion battery modeling and nonlinear analysis based on
performance modelling, in: 2015 17th European Conference on Power Electronics robust method in frequency domain: nonlinear characterization and non-
and Applications, EPE-ECCE Europe 2015, 2015, pp. 1–10, https://doi.org/ parametric modeling, Energy 106 (2016) 602–617, https://doi.org/10.1016/j.
10.1109/EPE.2015.7309443. energy.2016.03.028.
24
[256] N. Harting, N. Wolff, F. Röder, U. Krewer, Nonlinear frequency response analysis [280] Electrically propelled road vehicles, Test Specification for Lithium-Ion Traction
(NFRA) of lithium-ion batteries, Electrochim. Acta 248 (2017) 133–139, https:// Battery Systems – Part 2: High Energy Applications, International Standard,
doi.org/10.1016/j.electacta.2017.04.037. British Standards Institution, 2012.
[257] N. Wolff, N. Harting, M. Heinrich, U. Krewer, Nonlinear frequency response [281] J.P. Christophersen, Battery test manual for electric vehicles, ReVision 3 (2015),
analysis on lithium-ion batteries: process identification and differences between https://doi.org/10.2172/1186745.
transient and steady-state behavior, Electrochim. Acta 298 (2019) 788–798, [282] J.P. Meyers, M. Doyle, R.M. Darling, J. Newman, The impedance response of a
https://doi.org/10.1016/j.electacta.2018.12.107. porous electrode composed of intercalation particles, J. Electrochem. Soc. 147 (8)
[258] N. Wolff, N. Harting, M. Heinrich, F. Röder, U. Krewer, Nonlinear frequency (2000) 2930, https://doi.org/10.1149/1.1393627.
response analysis on lithium-ion batteries: a model-based assessment, [283] M. Doyle, J.P. Meyers, J. Newman, Computer simulations of the impedance
Electrochim. Acta 260 (2018) 614–622, https://doi.org/10.1016/j. response of lithium rechargeable batteries, J. Electrochem. Soc. 147 (1) (2000)
electacta.2017.12.097. 99, https://doi.org/10.1149/1.1393162.
[259] N. Harting, N. Wolff, F. Röder, U. Krewer, State-of-health diagnosis of lithium-ion [284] G. Sikha, R.E. White, Analytical expression for the impedance response of an
batteries using nonlinear frequency response analysis, J. Electrochem. Soc. 166 insertion electrode cell, J. Electrochem. Soc. 154 (1) (2007) A43, https://doi.org/
(2) (2019) A277–A285, https://doi.org/10.1149/2.1031902jes. 10.1149/1.2372695.
[260] F. Fasmin, R. Srinivasan, Review—nonlinear electrochemical impedance [285] G. Sikha, R.E. White, Analytical expression for the impedance response for a
spectroscopy, J. Electrochem. Soc. 164 (7) (2017) H443–H455, https://doi.org/ lithium-ion cell, J. Electrochem. Soc. 155 (12) (2008) A893, https://doi.org/
10.1149/2.0391707jes. 10.1149/1.2976359.
[261] M. Heinrich, N. Wolff, N. Harting, V. Laue, F. Röder, S. Seitz, U. Krewer, Physico- [286] J. Huang, J. Zhang, Theory of impedance response of porous electrodes:
chemical modeling of a lithium-ion battery: an ageing study with electrochemical simplifications, inhomogeneities, non-stationarities and applications,
impedance spectroscopy, Batteries and Supercaps 2 (1) (2019) 530–540, https:// J. Electrochem. Soc. 163 (9) (2016) A1983–A2000, https://doi.org/10.1149/
doi.org/10.1002/batt.201900011. 2.0901609jes.
[262] E. Ivers-Tiffée, A. Weber, Evaluation of electrochemical impedance spectra by the [287] A. Zaban, E. Zinigrad, D. Aurbach, Impedance spectroscopy of Li electrodes. 4. A
distribution of relaxation times, J. Ceram. Soc. Jpn. 125 (4) (2017) 193–201, general simple model of the Li-solution interphase in polar aprotic systems,
https://doi.org/10.2109/jcersj2.16267. J. Phys. Chem. 100 (8) (1996) 3089–3101, https://doi.org/10.1021/jp9514279.
[263] D. Klotz, M. Schönleber, J.P. Schmidt, E. Ivers-Tiffée, New approach for the [288] E.R.v. Schweidler, Studien über die Anomalien im Verhalten der Dielektrika, Ann.
calculation of impedance spectra out of time domain data, Electrochim. Acta 56 Phys. 329 (14) (1907) 711–770, https://doi.org/10.1002/andp.19073291407.
(24) (2011) 8763–8769, https://doi.org/10.1016/j.electacta.2011.07.096. [289] K.W. Wagner, Zur Theorie der unvollkommenen Dielektrika, Ann. Phys. 345 (5)
[264] R. Mingant, J. Bernard, V. Sauvant-Moynot, Novel state-of-health diagnostic (1913) 817–855, https://doi.org/10.1002/andp.19133450502.
method for Li-ion battery in service, Appl. Energy 183 (2016) 390–398, https:// [290] H. Schichlein, A.C. Müller, M. Voigts, A. Krügel, E. Ivers-tiffe, Deconvolution of
doi.org/10.1016/j.apenergy.2016.08.118. electrochemical impedance spectra for the identification of electrode reaction
[265] Gamry Application Note, ™ OptiEIS, A Multisine Implementation, 2016. mechanisms in solid oxide fuel cells, J. Appl. Electrochem. 32 (2002) 875–882,
[266] BioLogic, EC-Lab, Application note #19 2010 EIS measurements with multisine I https://doi.org/10.1023/A:1020599525160.
– introduction. https://www.biologic.net/documents/eis-multisine-electrochemis [291] J.P. Schmidt, P. Berg, M. Schönleber, A. Weber, E. Ivers-Tiffée, The distribution of
try-battery-corrosion-application-note-19/, 2010 accessed: 2020-05-27. relaxation times as basis for generalized time-domain models for Li-ion batteries,
[267] K. Takano, K. Nozaki, Y. Saito, K. Kato, A. Negishi, Impedance spectroscopy by J. Power Sources 221 (2013) 70–77, https://doi.org/10.1016/j.
voltage-step chronoamperometry using the Laplace transform method in a jpowsour.2012.07.100.
lithium-ion battery, J. Electrochem. Soc. 147 (3) (2000) 922–929, https://doi. [292] J. Illig, T. Chrobak, M. Ender, J.P. Schmidt, D. Klotz, E. Ivers-Tiffée, Studies on
org/10.1149/1.1393293. LiFePO4 as cathode material in Li-ion batteries, ECS Transactions 28 (30) (2010)
[268] J. Sihvo, T. Messo, T. Roinila, R. Luhtala, Online internal impedance 3–17, https://doi.org/10.1149/1.3505456.
measurements of Li-ion battery using PRBS broadband excitation and fourier [293] MATLAB, Version 7.10.0, The MathWorks Inc., Natick, Massachusetts, 2019,
techniques: methods and injection design, in: International Power Electronics 2019.
Conference, IPEC-Niigata - ECCE Asia 2018 vol. 0, 2018, pp. 2470–2475, https:// [294] Matlab Optimization Toolbox, the MathWorks, Natick, MA, USA, 2019.
doi.org/10.23919/IPEC.2018.8507565, 1) (2018. [295] T.H. Wan, M. Saccoccio, C. Chen, F. Ciucci, Influence of the discretization
[269] N. Lohmann, P. Haussmann, P. Wesskamp, J. Melbert, T. Musch, Employing real methods on the distribution of relaxation times deconvolution: implementing
automotive driving data for electrochemical impedance spectroscopy on lithium- radial basis functions with DRTtools, Electrochim. Acta 184 (2015) 483–499,
ion cells, SAE International Journal of Alternative Powertrains 4 (2) (2015) https://doi.org/10.1016/j.electacta.2015.09.097.
308–317, https://doi.org/10.4271/2015-01-1187. [296] E. Barsoukov, D.H. Kim, H.S. Lee, H. Lee, M. Yakovleva, Y. Gao, J.F. Engel,
[270] A. Mertens, I.C. Vinke, H. Tempel, H. Kungl, L.G.J. De Haart, R.A. Eichel, Comparison of kinetic properties of LiCoO2 and LiTi0.05Mg0.05Ni0.7Co0.2O2 by
J. Granwehr, Quantitative analysis of time-domain supported electrochemical impedance spectroscopy, Solid State Ionics 161 (1–2) (2003) 19–29, https://doi.
impedance spectroscopy data of Li-ion batteries: reliable activation energy org/10.1016/S0167-2738(03)00150-4.
determination at low frequencies, J. Electrochem. Soc. 163 (7) (2016) [297] J. Moškon, S.D. Talian, R. Dominko, M. Gaberšček, Advances in understanding Li
H521–H527, https://doi.org/10.1149/2.0511607jes. battery mechanisms using impedance spectroscopy, J. Electrochem. Sci. Eng. 10
[271] T. Tanaka, S. Ito, M. Muramatsu, T. Yamada, H. Kamiko, N. Kakimoto, Y. Inui, (2020), https://doi.org/10.2298/JSC100212057S.
Accurate and versatile simulation of transient voltage profile of lithium-ion [298] F. Dion, A. Lasia, The use of regularization methods in the deconvolution of
secondary battery employing internal equivalent electric circuit, Appl. Energy underlying distributions in electrochemical processes, J. Electroanal. Chem. 475
143 (2015) 200–210, https://doi.org/10.1016/j.apenergy.2015.01.028. (1) (1999) 28–37, https://doi.org/10.1016/S0022-0728(99)00334-4.
[272] N.S. Spinner, C.T. Love, S.L. Rose-Pehrsson, S.G. Tuttle, Expanding the [299] F. Ciucci, C. Chen, Analysis of electrochemical impedance spectroscopy data using
operational limits of the single-point impedance diagnostic for internal the distribution of relaxation times: a Bayesian and hierarchical Bayesian
temperature monitoring of lithium-ion batteries, Electrochim. Acta 174 (2015) approach, Electrochim. Acta 167 (2015) 439–454, https://doi.org/10.1016/j.
488–493, https://doi.org/10.1016/j.electacta.2015.06.003. electacta.2015.03.123.
[273] R. Srinivasan, B.G. Carkhuff, M.H. Butler, A.C. Baisden, Instantaneous [300] H. Qu, J. Kafle, J. Harris, D. Zheng, J. Koshina, D. Boone, A.M. Drake, C.
measurement of the internal temperature in lithium-ion rechargeable cells, J. Abegglen, D. Qu, Application of ac impedance as diagnostic tool – low
Electrochim. Acta 56 (17) (2011) 6198–6204, https://doi.org/10.1016/j. temperature electrolyte for a Li-ion battery, Electrochim. Acta 322 (2019),
electacta.2011.03.136. https://doi.org/10.1016/j.electacta.2019.134755.
[274] R. Srinivasan, P.A. Demirev, B.G. Carkhuff, Rapid monitoring of impedance phase [301] A.L. Gavrilyuk, D.A. Osinkin, D.I. Bronin, The use of Tikhonov regularization
shifts in lithium-ion batteries for hazard prevention, J. Power Sources 405 method for calculating the distribution function of relaxation times in impedance
(August) (2018) 30–36, https://doi.org/10.1016/j.jpowsour.2018.10.014. spectroscopy, Russ. J. Electrochem. 53 (6) (2017) 575–588, https://doi.org/
[275] R.R. Richardson, P.T. Ireland, D.A. Howey, Battery internal temperature 10.1134/S1023193517060040.
estimation by combined impedance and surface temperature measurement, [302] M. Hahn, S. Schindler, L.C. Triebs, M.A. Danzer, Optimized process parameters
J. Power Sources 265 (2014) 254–261, https://doi.org/10.1016/j. for a reproducible distribution of relaxation times analysis of electrochemical
jpowsour.2014.04.129. systems, Batteries 5 (2) (2019), https://doi.org/10.3390/batteries5020043.
[276] C.T. Love, M.B.V. Virji, R.E. Rocheleau, K.E. Swider-Lyons, State-of-health [303] M.A. Danzer, Generalized distribution of relaxation times analysis for the
monitoring of 18650 4S packs with a single-point impedance diagnostic, J. Power characterization of impedance spectra, Batteries 5 (3) (2019) 1–16, https://doi.
Sources 266 (2014) 512–519, https://doi.org/10.1016/j.jpowsour.2014.05.033. org/10.3390/batteries5030053.
[277] A. Chu, P. Braatz, Comparison of commercial supercapacitors and high-power [304] J. Song, M.Z. Bazant, Electrochemical impedance imaging via the distribution of
lithium-ion batteries for power-assist applications in hybrid electric vehicles: I. diffusion times, Phys. Rev. Lett. 120 (11) (2018) 116001, https://doi.org/
Initial characterization, J. Power Sources 112 (1) (2002) 236–246, https://doi. 10.1103/PhysRevLett.120.116001.
org/10.1016/S0378-7753(02)00364-6. [305] M.D. Murbach, D.T. Schwartz, Analysis of li-ion battery electrochemical
[278] H.G. Schweiger, O. Obeidi, O. Komesker, A. Raschke, M. Schiemann, C. Zehner, impedance spectroscopy data: an easy-to-implement approach for physics-based
M. Gehnen, M. Keller, P. Birke, Comparison of several methods for determining parameter estimation using an open-source tool, J. Electrochem. Soc. 165 (2)
the internal resistance of lithium ion cells, Sensors 10 (6) (2010) 5604–5625, (2018) A297–A304, https://doi.org/10.1149/2.1021802jes.
https://doi.org/10.3390/s100605604. [306] A. Eddahech, O. Briat, J.M. Vinassa, Performance comparison of four lithium-ion
[279] IEC 62660-1, Secondary Lithium-Ion Cells for the Propulsion of Electric Road battery technologies under calendar aging, Energy 84 (2015) 542–550, https://
Vehicles – Part 1: Performance Testing, International Standard, International doi.org/10.1016/j.energy.2015.03.019.
Electrotechnical Commission, Geneva, Switzerland, 2012. [307] C. Fleischer, W. Waag, H.M. Heyn, D.U. Sauer, On-line adaptive battery
impedance parameter and state estimation considering physical principles in
25
reduced order equivalent circuit battery models: Part 1. Requirements, critical [334] J. Song, M.Z. Bazant, Electrochemical impedance of a battery electrode with
review of methods and modeling, J. Power Sources 260 (2014) 276–291, https:// anisotropic active particles, Electrochim. Acta 131 (2014) 214–227, https://doi.
doi.org/10.1016/j.jpowsour.2014.01.129. org/10.1016/j.electacta.2014.03.013.
[308] F. Kohlrausch, W.A. Nippoldt, Über die Gültigkeit der Ohm’ schen Gesetze für [335] J. Illig, M. Ender, A. Weber, E. Ivers-Tiffée, Modeling graphite anodes with serial
Electrolyte und eine numerische Bestimmung des Leitungswiderstandes der and transmission line models, J. Power Sources 282 (2015) 335–347, https://doi.
verdünnten Schwefelsäure durch alternirende Ströme, Ann. Phys. 214 (10) (1869) org/10.1016/j.jpowsour.2015.02.038.
280–289. [336] J. Landesfeind, J. Hattendorff, A. Ehrl, W.A. Wall, H.A. Gasteiger, Tortuosity
[309] O. Heaviside, Electromagnetic Theory, Macmillan & Co., London, 1894. determination of battery electrodes and separators by impedance spectroscopy,
[310] E. Warburg, Ueber das Verhalten sogenannter unpolarisirbarer Elektroden gegen J. Electrochem. Soc. 163 (7) (2016) A1373–A1387, https://doi.org/10.1149/
Wechselstrom, Ann. Phys. 303 (3) (1899) 493–499, https://doi.org/10.1002/ 2.1141607jes.
andp.18993030302. [337] J. Bisquert, G. Garcia-Belmonte, F. Fabregat-Santiago, N.S. Ferriols, P. Bogdanoff,
[311] J.E.B. Randles, Kinetics of rapid electrode reactions, Discuss. Faraday Soc. 1 E.C. Pereira, Doubling exponent models for the analysis of porous film electrodes
(1947) 11–19, https://doi.org/10.1039/DF9470100011. by impedance. relaxation of tio2 nanoporous in aqueous solution, J. Phys. Chem.
[312] C. Ho, I. Raistrick, R. Huggins, Application of a-c techniques to the study of B 104 (10) (2000) 2287–2298, https://doi.org/10.1021/jp993148h.
lithium diffusion in tungsten trioxide thin films, J. Electrochem. Soc. 127 (2) [338] CNLS (Complex Nonlinear Least Squares), Immittance, Inversion, and Simulation
(1980) 343–350, https://doi.org/10.1149/1.2129668. Fitting Programs, J. R. MacDonald, 2020 accessed, https://jrossmacdonald.com/l
[313] R. De Levie, On porous electrodes in electrolyte solutions: I. capacitance effects, evmlevmw/, 05-27.
Electrochim. Acta 8 (10) (1963) 751–780, https://doi.org/10.1016/0013-4686 [339] MEISP 3.0:Multiple Electrochemical Impedance Spectra Parameterization, Korea
(63)80042-0. Kumho Petrochemical Co., http://impedance0.tripod.com/#3.
[314] R. De Levie, On porous electrodes in electrolyte solutions—iv, Electrochim. Acta 9 [340] Python-assisted physics-based electrochemical impedance spectroscopy. htt
(9) (1964) 1231–1245, https://doi.org/10.1016/0013-4686(64)85015-5. ps://github.com/thuylinhpham/PyPhyEIS.
[315] H. Gerischer, Wechselstrompolarisation von Elektroden mit einem [341] E. Barsoukov, J.H. Kim, K.S. Hwang, D.H. Kim, C.O. Yoon, H. Lee, Parametric
potentialbestimmenden Schritt beim Gleichgewichtspotential II, Z. Phys. Chem. analysis of electrical storage materials: new concept and application, Synth. Met.
201 (1) (1952) 55–67, https://doi.org/10.1515/zpch-1952-20105. 117 (1–3) (2001) 53–59, https://doi.org/10.1016/S0379-6779(00)00538-5.
[316] J. Bisquert, Influence of the boundaries in the impedance of porous film [342] J. Bisquert, V.S. Vikhrenko, Analysis of the kinetics of ion intercalation. two state
electrodes, Phys. Chem. Chem. Phys. 2 (18) (2000) 4185–4192, https://doi.org/ model describing the coupling of solid state ion diffusion and ion binding
10.1039/B001708F. processes, Electrochim. Acta 47 (24) (2002) 3977–3988, https://doi.org/
[317] J. Bisquert, F. Fabregat-Santiago, Impedance spectroscopy: a general introduction 10.1016/S0013-4686(02)00372-9.
and application to dye-sensitized solar cells, in: Dye-sensitized Solar Cells, EPFL [343] D. Aurbach, M.D. Levi, E. Levi, A review on the solid-state ionics of
Press, 2010, pp. 477–574. electrochemical intercalation processes: how to interpret properly their
[318] E.C. Shin, J. Ma, P.A. Ahn, H.H. Seo, D.T. Nguyen, J.S. Lee, Deconvolution of four electrochemical response, Solid State Ionics 179 (21–26) (2008) 742–751,
transmission-line-model impedances in Ni-YSZ/YSZ/LSM solid oxide cells and https://doi.org/10.1016/j.ssi.2007.12.070.
mechanistic insights, Electrochim. Acta 188 (2016) 240–253, https://doi.org/ [344] R. Huggins, Advanced Batteries: Materials Science Aspects, Springer Science &
10.1016/j.electacta.2015.11.118. Business Media, 2008, https://doi.org/10.1007/978-0-387-76424-5.
[319] H.T. Nguyen, T.L. Tran, D.T. Nguyen, E.-C. Shin, S.-H. Kang, J.-S. Lee, Full [345] T. Jacobsen, K. West, Diffusion impedance in planar, cylindrical and spherical
parametric impedance analysis of photoelectrochemical cells: case of a tio2 symmetry, Electrochim. Acta 40 (2) (1995) 255–262, https://doi.org/10.1016/
photoanode, J. Kor. Chem. Soc. 55 (3) (2018) 244–260, https://doi.org/10.4191/ 0013-4686(94)E0192-3.
kcers.2018.55.3.11. [346] S.D. Talian, J. Moškon, R. Dominko, M. Gaberšček, Impedance response of porous
[320] H. Seo, K. Jin, S. Park, K.H. Cho, H. Ha, K.G. Lee, Y.H. Lee, D.T. Nguyen, carbon cathodes in polysulfide redox system, Electrochim. Acta 302 (2019)
H. Randriamahazaka, J.S. Lee, K.T. Nam, Mechanistic investigation with kinetic 169–179, https://doi.org/10.1016/j.electacta.2019.02.037.
parameters on water oxidation catalyzed by manganese oxide nanoparticle film, [347] R. Kronig, On the theory of dispersion of X-rays, J. Opt. Soc. Am. 12 (6) (1926)
ACS Sustain. Chem. Eng. 7 (12) (2019) 10595–10604, https://doi.org/10.1021/ 547, https://doi.org/10.1364/JOSA.12.000547.
acssuschemeng.9b01159. [348] M.H.A. Kramers, La diffusion de la lumière par les atomes, Trans. Volta Centenary
[321] ZView 3.5, Impedance/gain Phase Graphing and Analysis Software, Scribner Ass. Congr 2 (1927) 545–557.
Inc., 2020 accessed, http://www.scribner.com/software/68-general-electroche [349] C.F. Bohren, What did Kramers and Kronig do and how did they do it? Eur. J.
mistr376-zview-for-windows/, 05-27. Phys. 31 (3) (2010) 573–577, https://doi.org/10.1088/0143-0807/31/3/014.
[322] K.S. Cole, R.H. Cole, Dispersion and absorption in dielectrics i. alternating current [350] C.W. Peterson, B.W. Knight, Causality calculations in the time domain: an
characteristics, J. Chem. Phys. 9 (4) (1941) 341–351, https://doi.org/10.1063/ efficient alternative to the Kramers–Kronig method*, J. Opt. Soc. Am. 63 (10)
1.1750906. (1973) 1238, https://doi.org/10.1364/josa.63.001238.
[323] P. Vyroubal, T. Kazda, Equivalent circuit model parameters extraction for lithium [351] P. Agarwal, M.E. Orazem, L.H. Garcia-Rubio, Measurement models for
ion batteries using electrochemical impedance spectroscopy, J. Energy Storage 15 electrochemical impedance spectroscopy: I. Demonstration of applicability,
(2018) 23–31, https://doi.org/10.1016/j.est.2017.10.019. J. Electrochem. Soc. 139 (7) (1992) 1917–1927, https://doi.org/10.1149/
[324] Q.K. Wang, Y.J. He, J.N. Shen, X.S. Hu, Z.F. Ma, State of charge-dependent 1.2069522.
polynomial equivalent circuit modeling for electrochemical impedance [352] B.A. Boukamp, A linear kronig-kramers transform test for immittance data
spectroscopy of lithium-ion batteries, IEEE Trans. Power Electron. 33 (10) (2018) validation, J. Electrochem. Soc. 142 (6) (1995) 1885–1901, https://doi.org/
8449–8460, https://doi.org/10.1109/TPEL.2017.2780184. 10.1149/1.2044210.
[325] W. Choi, H.-C. Shin, J.M. Kim, J.-Y. Choi, W.-S. Yoon, Modeling and applications [353] M. Schönleber, D. Klotz, E. Ivers-Tiffée, A method for improving the robustness of
of electrochemical impedance spectroscopy (eis) for lithium-ion batteries, Journal linear kramers-kronig validity tests, Electrochim. Acta 131 (2014) 20–27, https://
of Electrochemical Science and Technology 11 (1) (2020) 1–13, https://doi.org/ doi.org/10.1016/j.electacta.2014.01.034.
10.33961/jecst.2019.00528. [354] O. Alao, P. Barendse, Online condition monitoring of sealed lead acid lithium
[326] Y. Zhang, R. Zhao, J. Dubie, T. Jahns, L. Juang, Investigation of current sharing nickel-cobalt-manganese oxide batteries using broadband impedance
and heat dissipation in parallel-connected lithium-ion battery packs, ECCE 2016 - spectroscopy, IEEE Energy Conversion Congress and Exposition, ECCE 2018
IEEE Energy Conversion Congress and Exposition, Proceedings (2016) 1, https:// (2018) 2026–2032, https://doi.org/10.1109/ECCE.2018.8558352, 2018.
doi.org/10.1109/ECCE.2016.7855217. [355] X. Han, L. Lu, Y. Zheng, X. Feng, Z. Li, J. Li, M. Ouyang, A review on the key issues
[327] S. Buteau, J.R. Dahn, Analysis of thousands of electrochemical impedance spectra of the lithium ion battery degradation among the whole life cycle,
of lithium-ion cells through a machine learning inverse model, J. Electrochem. eTransportation 1 (2019) 100005, https://doi.org/10.1016/j.etran.2019.100005.
Soc. 166 (8) (2019), https://doi.org/10.1149/2.1051908jes. A1611–A1622. [356] T. Waldmann, A. Iturrondobeitia, M. Kasper, N. Ghanbari, F. Aguesse, E. Bekaert,
[328] H.C. Shin, W.I. Cho, H. Jang, Electrochemical properties of the carbon-coated L. Daniel, S. Genies, I.J. Gordon, M.W. Löble, E.D. Vito, M. Wohlfahrt-Mehrens,
LiFePO4 as a cathode material for lithium-ion secondary batteries, J. Power Review - post-mortem analysis of aged lithium-ion batteries: disassembly
Sources 159 (2) (2006) 1383–1388, https://doi.org/10.1016/j. methodology and physico-chemical analysis techniques, J. Electrochem. Soc. 163
jpowsour.2005.12.043. (10) (2016) A2149–A2164, https://doi.org/10.1149/2.1211609jes.
[329] J. Xia, F. Zhu, L. Wang, G. Wang, Y. Meng, Y. Zhang, In situ coating on LiFePO4 [357] A. Pfrang, A. Kersys, A. Kriston, D.U. Sauer, C. Rahe, S. Käbitz, E. Figgemeier,
with ionic liquid as carbon source for high-performance lithium batteries, Long-term cycling induced jelly roll deformation in commercial 18650 cells,
J. Nanoparticle Res. 20 (7) (2018), https://doi.org/10.1007/s11051-018-4298-z. J. Power Sources 392 (2018) 168–175, https://doi.org/10.1016/j.
[330] C. Zou, L. Zhang, X. Hu, Z. Wang, T. Wik, M. Pecht, A review of fractional-order jpowsour.2018.03.065.
techniques applied to lithium-ion batteries, lead-acid batteries, and [358] J. Lamb, C.J. Orendorff, Evaluation of mechanical abuse techniques in lithium ion
supercapacitors, J. Power Sources 390 (2018) 286–296, https://doi.org/10.1016/ batteries, J. Power Sources 247 (2014) 189–196, https://doi.org/10.1016/j.
j.jpowsour.2018.04.033. February. jpowsour.2013.08.066.
[331] J. Huang, Z. Li, B.Y. Liaw, J. Zhang, Graphical analysis of electrochemical [359] M. Dubarry, V. Svoboda, R. Hwu, B.Y. Liaw, A roadmap to understand battery
impedance spectroscopy data in Bode and Nyquist representations, J. Power performance in electric and hybrid vehicle operation, J. Power Sources 174 (2)
Sources 309 (2016) 82–98, https://doi.org/10.1016/j.jpowsour.2016.01.073. (2007) 366–372, https://doi.org/10.1016/j.jpowsour.2007.06.237.
[332] S. Buller, Impedance-Based Simulation Models for Energy Storage Devices in [360] D. Aurbach, B. Markovsky, A. Rodkin, M. Cojocaru, E. Levi, H.-J. Kim, An analysis
Advanced Automotive Power Systems, Shaker Verlag GmbH, 2003. of rechargeable lithium-ion batteries after prolonged cycling, Electrochim. Acta
[333] J. Song, M.Z. Bazant, Effects of nanoparticle geometry and size distribution on 47 (12) (2002) 1899–1911, https://doi.org/10.1016/S0013-4686(02)00013-0.
diffusion impedance of battery electrodes, J. Electrochem. Soc. 160 (1) (2013) [361] A. Pfrang, F.D. Persio, A. Kriston, N. Lebedeva, V. Ruiz, D. Dams, T. Kosmidou,
A15–A24, https://doi.org/10.1149/2.023301jes. J. Ungeheuer, I. Adanouj, L. Boon-Brett, Battery Testing Methods Assessed from a
26
Policy-Making Perspective: Battery Materials and Cell Performance Testing, Li1.2Ni0.15Mn0.55Co0.1O2–graphite lithium-ion cells, J. Electrochem. Soc. 160
EVS30 Symposium, 2017. (5) (2013) A3006–A3019, https://doi.org/10.1149/2.002305jes.
[362] N.P. Lebedeva, F.D. Persio, T. Kosmidou, D. Dams, A. Pfrang, A. Kersys, L. Boon- [384] F. Friedrich, B. Strehle, A.T.S. Freiberg, K. Kleiner, S.J. Day, C. Erk, M. Piana, H.
Brett, Amount of free liquid electrolyte in commercial large format Prismatic Li- A. Gasteiger, Capacity fading mechanisms of NCM-811 cathodes in lithium-ion
Ion battery cells, J. Electrochem. Soc. 166 (4) (2019) A779–A786, https://doi. batteries studied by X-ray diffraction and other diagnostics, J. Electrochem. Soc.
org/10.1149/2.1151904jes. 166 (15) (2019) A3760–A3774, https://doi.org/10.1149/2.0821915jes.
[363] A.W. Golubkov, D. Fuchs, J. Wagner, H. Wiltsche, C. Stangl, G. Fauler, G. Voitic, [385] D. Saurel, J. Segalini, M. Jauregui, A. Pendashteh, B. Daffos, P. Simon, M. Casas-
A. Thaler, V. Hacker, Thermal-runaway experiments on consumer Li-ion batteries Cabanas, A SAXS outlook on disordered carbonaceous materials for
with metal-oxide and olivin-type cathodes, RSC Adv. 4 (7) (2014) 3633–3642, electrochemical energy storage, Energy Storage Materials 21 (2019) 162–173,
https://doi.org/10.1039/C3RA45748F. https://doi.org/10.1016/j.ensm.2019.05.007.
[364] L. Somerville, J. Bareño, P. Jennings, A. McGordon, C. Lyness, I. Bloom, The effect [386] M. Povia, J. Sottmann, G. Portale, K.D. Knudsen, S. Margadonna, S. Sartori,
of pre-analysis washing on the surface film of graphite electrodes, Electrochim. Operando SAXS/WAXS on the a-P/C as the anode for Na-ion batteries, J. Phys.
Acta 206 (2016) 70–76, https://doi.org/10.1016/j.electacta.2016.04.133. Chem. C 122 (11) (2018) 5917–5923, https://doi.org/10.1021/acs.jpcc.7b12825.
[365] M. Klett, R. Eriksson, J. Groot, P. Svens, K. Ciosek Högström, R.W. Lindström, [387] T.-G. Jeong, J. Chun, B.-W. Cho, J. Lee, Y.-T. Kim, Enhanced performance of
H. Berg, T. Gustafson, G. Lindbergh, K. Edström, Non-uniform aging of cycled sulfur-infiltrated bimodal mesoporous carbon foam by chemical solution
commercial LiFePO4//graphite cylindrical cells revealed by post-mortem deposition as cathode materials for lithium sulfur batteries, Sci. Rep. 7 (1) (2017)
analysis, J. Power Sources 257 (2014) 126–137, https://doi.org/10.1016/j. 42238, https://doi.org/10.1038/srep42238.
jpowsour.2014.01.105. [388] J. McBreen, The application of synchrotron techniques to the study of lithium-ion
[366] H. Liu, Z. Li, A. Grenier, G.E. Kamm, L. Yin, G.S. Mattei, M.R. Cosby, P. batteries, J. Solid State Electrochem. 13 (7) (2009) 1051–1061, https://doi.org/
G. Khalifah, P.J. Chupas, K.W. Chapman, Best practices for operando depth- 10.1007/s10008-008-0685-1.
resolving battery experiments, J. Appl. Crystallogr. 53 (1) (2020) 133–139, [389] M. Hirayama, H. Ido, K. Kim, W. Cho, K. Tamura, J. Mizuki, R. Kanno, Dynamic
https://doi.org/10.1107/S1600576719016315. structural changes at LiMn2O4/electrolyte interface during lithium battery
[367] M. Morcrette, Y. Chabre, G. Vaughan, G. Amatucci, J.-B. Leriche, S. Patoux, reaction, J. Am. Chem. Soc. 132 (43) (2010) 15268–15276, https://doi.org/
C. Masquelier, J.-M. Tarascon, In situ X-ray diffraction techniques as a powerful 10.1021/ja105389t.
tool to study battery electrode materials, Electrochim. Acta 47 (19) (2002) [390] N. Yamakawa, M. Jiang, B. Key, C.P. Grey, Identifying the local structures formed
3137–3149, https://doi.org/10.1016/S0013-4686(02)00233-5. during lithiation of the conversion material, iron fluoride, in a Li ion battery: a
[368] P. Barpanda, G. Rousse, T. Ye, C.D. Ling, Z. Mohamed, Y. Klein, A. Yamada, solid-state NMR, X-ray diffraction, and pair distribution function analysis study,
Neutron diffraction study of the Li-ion battery cathode Li2FeP2O7, Inorg. Chem. J. Am. Chem. Soc. 131 (30) (2009) 10525–10536, https://doi.org/10.1021/
52 (6) (2013) 3334–3341, https://doi.org/10.1021/ic302816w. ja902639w.
[369] W.K. Pang, V.K. Peterson, A custom battery for operando neutron powder [391] B. Key, M. Morcrette, J.-M. Tarascon, C.P. Grey, Pair distribution function
diffraction studies of electrode structure, J. Appl. Crystallogr. 48 (1) (2015) analysis and solid state NMR studies of silicon electrodes for lithium ion batteries:
280–290, https://doi.org/10.1107/S1600576715000679. understanding the (De)lithiation mechanisms, J. Am. Chem. Soc. 133 (3) (2011)
[370] M. Balasubramanian, X. Sun, X.Q. Yang, J. McBreen, In situ X-ray diffraction and 503–512, https://doi.org/10.1021/ja108085d.
X-ray absorption studies of high-rate lithium-ion batteries, J. Power Sources 92 [392] D. Zeng, J. Cabana, J. Bréger, W.-S. Yoon, C.P. Grey, Cation ordering in Li
(1) (2001) 1–8, https://doi.org/10.1016/S0378-7753(00)00493-6. [NixMnxCo(1–2x)]O2-layered cathode materials: a nuclear magnetic resonance
[371] U.W. Arndt, D.C. Creagh, R.D. Deslattes, J.H. Hubbell, P. Indelicato, E.G.K. Jr, (NMR), pair distribution function, X-ray absorption spectroscopy, and
E. Lindroth, International Tables for Crystallography 4.2 (2006). electrochemical study, Chem. Mater. 19 (25) (2007) 6277–6289, https://doi.org/
[372] T. Zhang, H. Zhou, A reversible long-life lithium–air battery in ambient air, Nat. 10.1021/cm702241a.
Commun. 4 (1) (2013) 1817, https://doi.org/10.1038/ncomms2855. [393] J. Wu, M. Fenech, R.F. Webster, R.D. Tilley, N. Sharma, Electron microscopy and
[373] J. Zheng, M.H. Engelhard, D. Mei, S. Jiao, B.J. Polzin, J.-G. Zhang, W. Xu, its role in advanced lithium-ion battery research, Sustainable Energy and Fuels 3
Electrolyte additive enabled fast charging and stable cycling lithium metal (7) (2019) 1623–1646, https://doi.org/10.1039/c9se00038k.
batteries, Nature Energy 2 (3) (2017) 17012, https://doi.org/10.1038/ [394] M. Koltypin, D. Aurbach, L. Nazar, B. Ellis, More on the performance of LiFePO4
nenergy.2017.12. electrodes-The effect of synthesis route, solution composition, aging, and
[374] I. Buchberger, S. Seidlmayer, A. Pokharel, M. Piana, J. Hattendorff, P. Kudejova, temperature, J. Power Sources 174 (2) (2007) 1241–1250, https://doi.org/
R. Gilles, H.A. Gasteiger, Aging analysis of graphite/LiNi1/3Mn1/3Co1/3O2 cells 10.1016/j.jpowsour.2007.06.045.
using XRD, PGAA, and AC impedance, J. Electrochem. Soc. 162 (14) (2015) [395] N. Dupré, J.F. Martin, J. Degryse, V. Fernandez, P. Soudan, D. Guyomard, Aging
A2737–A2746, https://doi.org/10.1149/2.0721514jes. of the LiFePO4 positive electrode interface in electrolyte, J. Power Sources 195
[375] Y. Su, Y. Yang, L. Chen, Y. Lu, L. Bao, G. Chen, Z. Yang, Q. Zhang, J. Wang, (21) (2010) 7415–7425, https://doi.org/10.1016/j.jpowsour.2010.05.042.
R. Chen, S. Chen, F. Wu, Improving the cycling stability of Ni-rich cathode [396] K. Amine, J. Liu, I. Belharouak, High-temperature storage and cycling of C-
materials by fabricating surface rock salt phase, Electrochim. Acta 292 (2018) LiFePO4/graphite Li-ion cells, Electrochem. Commun. 7 (7) (2005) 669–673,
217–226, https://doi.org/10.1016/j.electacta.2018.09.158. https://doi.org/10.1016/j.elecom.2005.04.018.
[376] X. Zhang, W.J. Jiang, A. Mauger, Qilu, F. Gendron, C.M. Julien, Minimization of [397] K. Striebel, J. Shim, A. Sierra, H. Yang, X. Song, R. Kostecki, K. McCarthy, The
the cation mixing in Li1+x(NMC)1-xO2 as cathode material, J. Power Sources 195 development of low cost LiFePO4-based high power lithium-ion batteries,
(5) (2010) 1292–1301, https://doi.org/10.1016/j.jpowsour.2009.09.029. J. Power Sources 146 (1–2) (2005) 33–38, https://doi.org/10.1016/j.
[377] H.M. Rietveld, Line profiles of neutron powder-diffraction peaks for structure jpowsour.2005.03.119.
refinement, Acta Crystallogr. 22 (1) (1967) 151–152, https://doi.org/10.1107/ [398] G. Zampardi, F. La Mantia, Solid-Electrolyte Interphase at the Positive Electrodes
S0365110X67000234. in High Energy Li-ion Batteries: Current Understanding and Analytical Tools,
[378] H.M. Rietveld, A profile refinement method for nuclear and magnetic structures, Batteries & Supercaps 3 (8) (2020) 672–697, https://doi.org/10.1002/
J. Appl. Crystallogr. 2 (2) (1969) 65–71, https://doi.org/10.1107/ batt.201900177.
S0021889869006558. [399] J.Y. Song, H.H. Lee, Y.Y. Wang, C.C. Wan, Two- and three-electrode impedance
[379] F. Lin, I.M. Markus, D. Nordlund, T.-C. Weng, M.D. Asta, H.L. Xin, M.M. Doeff, spectroscopy of lithium-ion batteries, J. Power Sources 111 (2) (2002) 255–267,
Surface reconstruction and chemical evolution of stoichiometric layered cathode https://doi.org/10.1016/S0378-7753(02)00310-5.
materials for lithium-ion batteries, Nat. Commun. 5 (1) (2014) 3529, https://doi. [400] C.H. Chen, J. Liu, K. Amine, Symmetric cell approach and impedance
org/10.1038/ncomms4529. spectroscopy of high power lithium-ion batteries, J. Power Sources 96 (2) (2001)
[380] C.S. Johnson, N. Li, C. Lefief, J.T. Vaughey, M.M. Thackeray, Synthesis, 321–328, https://doi.org/10.1016/S0378-7753(00)00666-2.
characterization and electrochemistry of lithium battery electrodes: xLi2MnO3(1-x) [401] R. Raccichini, L. Furness, J.W. Dibden, J.R. Owen, N. García-Aráez, Impedance
LiMn0.333Ni0.333Co0.333O2 (0≤x≤0.7), Chem. Mater. 20 (19) (2008) 6095–6106, characterization of the transport properties of electrolytes contained within
https://doi.org/10.1021/cm801245r. porous electrodes and separators useful for Li-S batteries, J. Electrochem. Soc.
[381] E. Meza, J. Ortiz, D. Ruíz-León, J.F. Marco, J.L. Gautier, Lithium-nickel cobalt 165 (11) (2018) A2741–A2749, https://doi.org/10.1149/2.0631811jes.
oxides with spinel structure prepared at low temperature. XRD, XPS, and EIS [402] C.-H. Chen, F. Planella, K. O’Regan, D. Gastol, D. Widanagea, E. Kendrick,
measurements, Mater. Lett. 70 (2012) 189–192, https://doi.org/10.1016/j. Development of experimental techniques for parameterization of multi-scale
matlet.2011.11.108. lithium-ion battery models, J. Electrochem. Soc. 167 (2020), 080534, https://doi.
[382] D. Mohanty, S. Kalnaus, R.A. Meisner, K.J. Rhodes, J. Li, E.A. Payzant, D.L. Wood, org/10.1149/1945-7111/ab9050.
C. Daniel, Structural transformation of a lithium-rich Li1.2Co0.1Mn0.55Ni0.15O2 [403] B.N. Popov, A. Durairajan, Y. Podrazhansky, R.C. Cope, Capacity fade of Li-ion
cathode during high voltage cycling resolved by in situ X-ray diffraction, J. Power cells: comparison of DC and ENREV charging protocols, in: Proceedings of the
Sources 229 (2013) 239–248, https://doi.org/10.1016/j.jpowsour.2012.11.144. Annual Battery Conference on Applications and Advances 2000-Janua (Dmc),
[383] Y. Li, M. Bettge, B. Polzin, Y. Zhu, M. Balasubramanian, D.P. Abraham, 2000, pp. 185–191, https://doi.org/10.1109/BCAA.2000.838402.
Understanding long-term cycling performance of
27

Nina Medding 2020

Uploaded by

Copyright:

Available Formats

Nina Medding 2020

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nina Medding 2020

Uploaded by

Copyright:

Available Formats

Journal of Power Sources 480 (2020) 228742

Contents lists available at ScienceDirect

Journal of Power Sources

Application of electrochemical impedance spectroscopy to commercial

• Comprehensive evaluation of EIS-based

Fig. 6. a) Second order Voigt-type circuit element

You might also like