Is ultrasound suited to probe electrochemical processes in batteries? An operando multiprobe investigation of NMC/graphite commercial cell

doi:10.21203/rs.3.rs-5111062/v1

Download PDF

Article

Is ultrasound suited to probe electrochemical processes in batteries? An operando multiprobe investigation of NMC/graphite commercial cell

https://doi.org/10.21203/rs.3.rs-5111062/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Ultrasound technique has recently been employed to probe the internal dynamics of closed-form battery cells, offering a cost-effective means to gain real-time insights into electrochemical processes. It therefore has the potential to be a core part of the cell development process. However, existing literature has primarily focused on establishing superficial correlations between electrochemical and ultrasonic parameters, thereby severely constraining the interpretability of alterations of the ultrasound transmission properties of a cell upon cycling. This study aims to enhance the interpretability of these variations through simultaneous operando measurements of ultrasound and synchrotron X-rays diffraction (XRD) during electrochemical cycling and relaxation measurements. The results are further correlated to mechanical and electrochemical changes obtained through operando nano-dilatometry measurement. Thereby, we demonstrate the influence of both NMC and graphite on ultrasound features upon cycling. This work marks a significant advancement of the application of synchrotron XRD analysis to establish an understanding of ultrasound interactions with commercial cells.

Physical sciences/Materials science/Materials for energy and catalysis/Batteries

Physical sciences/Chemistry/Energy

Commercial battery

ultrasound sensor

Operando XRD

dilatometry measurement

X-ray diffraction

Li-ion battery technology is currently dominating the market of energy storage systems but still suffers from limitations especially high costs, low energy densities and safety issues. Over aging, cells performance decrease due to several degradation reactions causing severe safety issues such as gas release or thermal runaway [1]. Those aging processes are accelerated by additional external parameters such as mechanical (vibration, compression, penetrations), thermal (high and low temperatures) or electrical (external short-circuit, overload, over discharge) processes [2]. Monitoring the State of Charge (SoC) and State of Health (SoH) of batteries during operation is of utmost importance to detect preliminary battery failure. Once detected, the battery management system (BMS) can properly disconnect the cell from a stack ensuring safety for the users [3]. One approach in SoH estimation passes through the plotting of differential curves dQ/dV and dV/dQ, where variations are observed with cell aging [4], [5], [6]. However, this method implies the presence of clear electrochemical processes. Recent studies have also used electrochemical impedance spectroscopy (EIS) for aging analysis by monitoring the frequency-dependent impedance of the cell [7]. Recently, optical fibres were introduced to probe mechanical insights in commercial cells [8], but also the thermal distribution in cylindrical cells [9]. Internal strain inside cylindrical cells was also probed using strain gauge sensors and a correlation with graphite swelling was evidenced [10]. The main issue with those sensors is their invasive nature and the eventuality of modifying the local electrochemical behaviour due to additional strain [11]. For this reason, the development and adaptation of non-invasive techniques on Li-ion batteries is crucial to further accelerate cell development and improve their operation.

Ultrasound technology, which has been already used in many industrial sectors such as nuclear [12] and medical [13], possesses many advantages such as a low price, high technology readiness level (TRL) and its non-invasive nature. Such sensors could be implemented and used to probe the electrochemical performance and degradation of batteries providing additional information about the mechanical dynamics [14]. Few studies have been carried out using ultrasound technique to detect abnormal battery behaviour during cycling. In 2019, Robinson et al. [15] employed ultrasound technique to probe a commercial cell (Samsung J5 smartphone battery composed of LiCoO₂/graphite electrodes) and demonstrated a direct link between density evolution of positive or negative electrode imposed by lithiation and delithiation processes and Time of Flight (ToF) response. Having successfully assessed the battery SoC as a function of the ToF signal, they demonstrated that shifts in the ToF signal could probe locally the lithium content in the active material. Similarly, Zhao et al.[16] used non-contact ultrasonic guided wave to probe and estimate the SoC and SoH of a NMC-LMO-LCO blend vs. graphite commercial cell. Whatever the wavelength used, as the SoC increases, the signal amplitude (SA) increases and, conversely, the ToF signal tends to decrease. More recently, Appleberry et al. [17] detected via ultrasound sensors some overcharging behaviour by collecting operando ultrasound data on a commercial battery. Ultrasounds can thus detect degradation processes, but, to date, the literature has yet to deduce a clear understanding of complex interaction of ultrasound with cell components. Despite current research in the field, a clear attribution of electrochemical effects to changes in ultrasound features remains an open research question.

In 2023, Liu et al. [17] were able to decorrelate the influence of SoC of the LiFePO₄ (LFP) positive electrode from that of the graphite negative electrode using ToF signal collected from ultrasound sensor. They evidenced that the signal is dominated by the evolution of the electrode Young’s modulus showing that this latter can be influenced by the lithiation/delithiation processes. Based on their investigation, they concluded that in a full cell LiFePO₄ vs. graphite, most of the ultrasound signal is coming from the graphite electrode. Pursuing this kind of investigation, Owen et al. demonstrated that thanks to ultrasound measurement, the internal temperature of a cell could be determined, as the temperature also impacts the Young's modulus [18].

In this study, we investigated a commercial NMC811 vs. graphite cell equipped with an ultrasound sensor, while simultaneously monitoring the structural evolution within the cell using operando X-ray diffraction (XRD) operated at large scale facilities (BM02 beamline at ESRF, Grenoble, France). To understand the signal monitored by the ultrasound sensor, we investigated the dynamics of the lithiation/delithiation using a rate of C/3 with relaxation protocol as well as performed a measurement close to the thermodynamic equilibrium (C/8 rate). The results obtained were further correlated to mechanical and electrochemical processes by a nano-dilatometric measurement, deconvoluting each electrode. First, we demonstrated that the ultrasound signal evolution is mainly driven by the graphite Young’s modulus evolution. Then, by correlating ToF evolution with the structural and electrochemical one, we demonstrated the sensitivity of ultrasound to the mechanism of lithium storage in both active materials.

A schematic of the experimental set-up is displayed in the upper part of Fig. 1a, where we can see the commercial cell (NMC811 vs. graphite) paired with an US emitter receptor. The cell was placed on a custom sample holder [20] developed by Lyonnard’s group (Fig. 1b) allowing lateral scans with the microbeam. The cycling protocol of the cell is displayed in Fig. 1c where three cycles were performed, a first one at C/3 rate, to verify the cells behaviour, a second one at C/8 rate and the last one cycled by galvanostatic intermittent titration “like” technique (GITT) using C/3 rate charging with 1 h pulses and 1 h relaxation steps.

Prior starting the operando measurement, we first verified that the measurement of the ultrasound signal was not impacted by temperature fluctuations of the hutch (see Note 1, Supporting Information) and that the measurement at the beamline is similar to the one obtained in the laboratory (Note 2, Supporting Information). We also checked that the cell did not suffer from X-ray damage (see Note 3, Supporting Information). Then, scalar quantities are extracted from the ultrasound signal as mentioned in Note 4 (Supporting Information), especially the time of flight (ToF), one of the most relevant parameters when it comes to ultrasound investigation in batteries. However, the interpretation of ToF’s evolution during cycling is far from trivial as it can be seen in Eq. 1. The ToF depends on constants such as the total mass and the section probed, and on two variables being the thickness and the average effective elastic modulus (Young’s modulus) of the entire medium crossed by ultrasound waves.

$$\:\begin{array}{c}ToF=\frac{d}{\sqrt{\frac{{E}_{eff}}{\rho\:}}}=\sqrt{\frac{m}{A}}\sqrt{\frac{d}{{E}_{eff}}}\:\#\left(\varvec{E}\varvec{q}\varvec{u}\varvec{a}\varvec{t}\varvec{i}\varvec{o}\varvec{n}\:1\right)\end{array}$$

with m (kg) being the mass probed, A (m²) the section, d (m) the thickness of the sample probed, E_eff (kg.m^− 1.s^− 2) the sample’s effective Young’s modulus and ρ (kg.m^− 3) the sample’s average density. Both variables depend on the sample’s chemical composition and environmental conditions, thus temperature and pressure.

To understand the ToF evolution and determine which parameter(s) dominate it, a measurement of the cell’s thickness and the global effective elastic modulus (referred to as Young’s modulus in the following) evolution is needed. For the former, several papers denote a thickness increase of several tens of micrometers during the charge of NMC vs. graphite cell (graphite lithiation and NMC delithiation) depending on the number of electrode layers [21], [22], [23]. During charge, NMC particles undergo an anisotropic variation in lattice parameters resulting in a global contraction of the lattice volume [24]. At the NMC electrode level, it is difficult to determine if the volume change of NMC particles has an impact in any of the three dimensions of the electrode, since the electrodes are porous. For graphite particles, the variations in lattice parameters are also anisotropic and induce a large volume change along the c-axis of the crystal structure (corresponding to graphene layers d-spacing), this volume increase reaching 14%. At the graphite electrode level, the increase of the interlayer d-spacing upon lithiation results in a unidirectional expansion, easily visible even at electrode level [25], [26]. Based on this discussion, the global expansion of a full cell NMC | graphite upon charge is attributed to the more dominant volume expansion, i.e. the graphite electrode. To verify our hypothesis, we dismounted, in an Ar-filled glovebox, the same commercial cell as the one brought to the synchrotron, washed the electrodes, gently removed a side of the double coated electrodes and mounted these single-side coated electrodes in a nano-dilatometer allowing the operando monitoring of the working electrode (graphite) expansion in a three electrodes set-up (NMC811 as counter electrode and Li metal as reference) (see Note 5, Supporting Information). In Fig. 2, we compare the full cell’s thickness evolution (evolution at 25 kPa constant pressure) reported from literature [23] with the graphite electrode expansion measured by operando dilatometric measurement as mentioned above. Figure 2a shows the full cell voltage and the negative electrode potential of the graphite vs. NMC cell. Qualitatively, we notice a very similar evolution of the full cell expansion (Fig. 2b) with the graphite electrode expansion (Fig. 2b), supporting the fact that graphite dominates the full cell thickness evolution in the low SoC regime (below 80%). Note that above this value, the positive electrode starts contributing with a negative expansion that is correlated to the shrinking of the NMC c-axis parameter at high voltage. At SoC > 80% it seems that graphite and NMC effects are counter balancing each other, yielding almost unchanged full cell thickness.

Based on Eq. 1, we need to consider the Young’s modulus evolution of the electrode materials (positive and negative). This metric is more complicated to assess, even more upon lithiation and delithiation of active materials, since generally, this kind of measurement should be performed without any current collector, which is far away from a realistic electrode and is hardly performed operando due to liquid electrolyte state. Generally, the Young’s modulus describes the elastic properties of cell components, so in the present case we have three Young’s modulus in total, two coming from each electrode and one coming from the separator, the last one could be attributed to cell housing (generally negligible). From the literature, graphite, NMC811, and most electroactive materials display a Young’s modulus that increases upon lithiation [27]. According to Davies et al., graphite displays an important variation of electrode sound velocity compared to LiCoO₂ (LCO) [28], explaining the larger impact of the graphite electrode on the average Young’s modulus of a full cell. A reason for this statement resides in the “softness” of graphite that can deform reversibly upon de/lithiation [29] compared to lithium oxide materials such as LCO [30] or NMC [27] that easily break upon mechanical strains during cycling.

Based on Eq. 1, an increase in cell thickness upon charge should lead to an increase in the ToF signal which is not consistent with our observation (Fig. 3) nor with the literature [22] showing a decrease in ToF upon charge.

The remaining parameter that evolves during cycling and could dominate ToF signal evolution is the effective Young’s modulus of the cell. During cycling, the Young’s modulus of graphite electrode becomes stiffer upon lithiation, while NMC’s Young’s modulus is decreasing upon delithiation which matches our results indicating that the ToF signal evolution of the full cell is dominated by the Young’s modulus evolution of graphite electrode. However, the non-monotonic evolution of ToF measurement highlights a complex correlation between the various cell components and additional correlation can be found with both electrode materials.

To deconvolute the ToF signal and attribute each process to a specific electrode phenomenon, we carried out a full structural investigation through operando XRD measurements while simultaneously collecting the ToF signal. As can be seen in Fig. 4, we can follow the evolution of the main NMC811 (003) Bragg reflection, as well as the one of graphite (002), both being properly separated in the q range. Using a home-made software for data correlation based on MatPlotLib Python library [31] and LMFIT package [32], we fitted both main reflections (raw data plotted in Note 6, Supporting Information) and obtained the average peak position to follow the structural changes through the evolution of the c-axis. The c-axis is indeed interesting for graphite electrodes since the de/lithiation processes modify the interlayer d-spacing [33]. For the NMC811 electrode, a-axis and c-axis are evolving. However, as shown by Quilty et al. [24], the c-axis is the one evolving the most, especially during the H2-H3 phase transition, justifying the choice of this axis too.

Although we can separately access the structures of the positive and negative active materials in given states of lithiation by XRD and access the c-parameter variations, the deconvolution between the positive electrode and the negative electrode reaction mechanisms is far from trivial since, in a full cell, the electrochemical signatures of both electrodes are merged. To address this issue, the operando nano-dilatometric measurement allowed a deconvolution of each electrode potential to understand the global electrochemical signature as well as correlate the volume change of one electrode at a time as a function of the cell’s voltage.

From the three-electrode set-up, the full cell voltage evolution is a linear combination of the two electrodes' potential variations vs. lithium metal as described in Eq. 2. However, the weight of each electrode potential variation in the full cell voltage evolution varies depending on the de/lithiation processes. Then, from the classical galvanostatic curves E = f(t) obtained, we can define weighting functions for both electrodes (Eq. 3), ω_gr and ω_NMC, which quantify the impact of the electrode potential variation on the full cell voltage. Those weighting functions are plotted in Fig. 5a.

$$\:\begin{array}{c}\frac{dE}{dt}\left(FC\right)=\:\frac{dE}{dt}\left(Gr\right)+\frac{dE}{dt}\left(NMC\right)\#\left(\varvec{E}\varvec{q}\varvec{u}\varvec{a}\varvec{t}\varvec{i}\varvec{o}\varvec{n}\:2\right)\end{array}$$

$$\:\begin{array}{c}{\omega\:}_{Gr}=\frac{dE}{dt}\left(Gr\right)/\frac{dE}{dt}\left(FC\right)\:and\:{\omega\:}_{NMC}=\frac{dE}{dt}\left(NMC\right)/\frac{dE}{dt}\left(FC\right)\#\left(\varvec{E}\varvec{q}\varvec{u}\varvec{a}\varvec{t}\varvec{i}\varvec{o}\varvec{n}\:3\right)\end{array}$$

Additionally, the dt/dE (analogue to Differential Voltage Analysis under galvanostatic conditions) evolution (Fig. 5b) provides information about the nature of the de/lithiation mechanism. For NMC811 material, lithium de/insertion takes place through a solid solution mechanism with some lattice parameter changes inducing specific variations in the dt/dE representation. For graphite, biphasic mechanisms generate peaks in the dt/dE curve, while solid solution ones do not impact significantly the curve evolution. Here, we should note that the three curves are linked through the following Eq. 4:

$$\:\begin{array}{c}\frac{dt}{dE}\left(FC\right)=\frac{1}{\frac{dE}{dt}\left(Gr\right)+\frac{dE}{dt}\left(NMC\right)}\#\left(\varvec{E}\varvec{q}\varvec{u}\varvec{a}\varvec{t}\varvec{i}\varvec{o}\varvec{n}\:4\right)\end{array}$$

With this representation, we can assign each phenomenon to a particular electrode [34]. The region A of the charging cell (Fig. 5) displays a voltage variation imposed by the graphite electrode as showed by the weighting function curve. However, in this region, no specific peak in the dt/dE representation is observed since solid solution like processes of lithiation/delithiation are taking place for both electrodes. In region B (Fig. 5), the full cell voltage variation is due to both electrodes’ potential variation but the similar shape of peak observed for the full cell and graphite curves in the bottom graph indicates that graphite undergoes its first biphasic phase transition (stage IV, formation of LiC₃₆ from LiC₇₂ [35]) while NMC811 is still delithiated through solid solution mechanism with no phase transition. A similar observation can be made in region C (Fig. 5), where the apparent peak of the full cell curve is attributed to the graphite stage III liquid-like transition. In region D (Fig. 5), two processes can be found, in the first one (depicted as D1) graphite undergoes the stage II phase transition (LiC₁₂ phase formation) while the hexagonal to monoclinic phase transition H1 to M of NMC811 takes place (common phase transition for the NMC811 compounds family). The second subregion (D2) contains still the graphite stage II formation, but NMC811 stops its specific phase transition. The last region E (Fig. 5) displays first a local minimum attributed to the change in biphasic mechanism of graphite from LiC₁₂ formation to LiC₆ formation (stage I) and the NMC electrode undergoes first the monoclinic to second hexagonal phase transition M to H2 in the E1 subregion. The second characteristic peak in subregion E2 correlates with the second to third hexagonal transition H2 to H3, involving Ni³⁺ and Co³⁺ oxidation into Ni⁴⁺ and Co⁴⁺ [36].

Now that we better understand the electrochemical response of a full cell (graphite vs. NMC811), we correlate the ToF signal to the electrochemical processes of NMC811 and graphite and to their structural evolutions obtained from operando XRD. Figure 6a and Fig. 6c plot the evolution of peak position from (002) Bragg reflection of graphite and (003) Bragg reflection of NMC811, respectively, while Fig. 6b plots the evolution of dt/dE from the full cell voltage variation obtained at C/8 correlated to the ToF evolution. Similarly, as for the dilatometric data, we divided the obtained correlated ToF-XRD data by the five same regions of interest, beginning by the cell discharge (graphite delithiation and NMC811 lithiation). It should be noted that the boundaries of potential are a bit shifted between Fig. 5 and Fig. 6 due to the high separator resistance used in the nano-dilatometric measurement. Starting with the region E₂’, we noticed a huge variation of ToF signal correlating with the peak attributed to the H3 ◊ H2 transition of NMC811, inducing an important change in lattice volume of NMC811 estimated at 4% by Quilty et al. [24], which is the most important variation in volume among all the transitions of NMC811. For the rest of the region E’, from 4.1 V – 3.8 V, a linear increase of ToF signal is observed and could be attributed to the LiC₆ phase disappearance. No obvious change in slope of ToF signal are noticed when NMC811 is undergoing the H2 ◊ M phase transition (E₁’ subregion). For the D₂’ region, we noticed a change in slope of ToF signal compared to the E₁’ region which correlates with the change in biphasic mechanism of graphite (from LiC₆ to LiC₁₂). When NMC811 undergoes the M ◊ H1 phase transition in the subregion D₁’, an increase in the slope intensity of ToF is observed. For the third region (C’), a period of pseudo-plateau in ToF signal correlates with liquid-like stage III of graphite. In this range of potential, very small amount of charges is exchanged leading to very small variation in global Young’s modulus of the cell. Concerning the B’ region, again, a specific slope in ToF increase is observed correlating here with the stage IV of graphite disappearance. For the last region of interest (A’), a kind of plateau in ToF signal correlates with very low amount of lithium stored through solid solution like mechanism in both materials.

Regarding the previous discussion, we observed that ToF signal is mainly driven by the Young’s modulus evolution of graphite, leading to a decrease in ToF during charge of the cell and an increase during the discharge. By deconvoluting all the electrochemical processes, we also demonstrated that most of the modulations in ToF are correlated to structural changes of the two electrode materials. We notice that, in general, the ToF changes linearly depend on the potential in the selected regions, with varying positive or negative slopes that might reflect a correlation between specific phase transitions and ultrasounds properties. This effect could be further investigated on simplified systems to obtain a quantified relationship between ToF and electrochemical processes, which is beyond our scope here.

Additionally, galvanostatic intermittent titrations “like” technique (GITT) was implemented to understand the impact of the relaxation in the structure of the material as well as in the ToF signal. This electrochemical characterization is widely used in the field of batteries [37], consisting in the application of regular pulses of low-intensity current. A long resting period is then left between those pulses, allowing the cell to relax and the electrode material to reach a state of quasi-equilibrium. Lithium concentrations are then equilibrated in the cell leading to homogeneous lithium distribution throughout the electrode volume. Here a current pulse equivalent to C/3 rate was selected to generate heterogeneities in the cell, to understand the sensitivity of the ultrasound technique.

As depicted in Fig. 7, ToF signal is sensitive to the relaxation processes occurring in the cell, since during the resting periods of the cell (R₁ – R₆ on Fig. 7) the ToF signal is not constant. Several parameters could contribute to the relaxation processes, here we tried to discriminate the one dominating this relaxation: i) the temperature, imposed by a cycling at C/3 rate, could generate local heating that the cell needs to dissipate and could impact the ToF signal. From previous experiment performed in the laboratory with similar commercial batteries, an extrapolated dependency of 26 ns/K was estimated. Considering most of the relaxation phases (expect R₃ which displays higher relaxation), an average decrease of 12 ns is observed in the ToF signal, corresponding to a temperature decrease of ~ 0.5 K, a realistic temperature change in such system at C/3 rate [38]. ii) Young’s modulus of graphite, which was the dominant factor for the overall evolution of ToF signal upon cycling, could also be sensitive to relaxation processes due to the local inhomogeneous reaction imposed by the C/3 rate. For this purpose, specific attention was given to the diffraction patterns of NMC811 and graphite evolution during relaxation periods (see Note 7, Supporting Information), specifically the last relaxation period related to high state of charge of the cell meaning the transition LiC₁₂ to LiC₆. Figure 8 represents the X-ray diffraction patterns of the two electrodes during the R₆ relaxation involving H3 phase of NMC811 and a mixture of LiC₁₂ (~ 1.78 Å⁻¹) and LiC₆ (~ 1.70 Å⁻¹) graphite phases. Here, we did not notice any visible shift of the peak position for NMC811 leading us to conclude that NMC is not presently affected by a relaxation process at C/3 rate. Concerning the graphite, no changes in LiC₆ peak intensity, nor in its shape (constant FWHM) are observed and the small variations in LiC₁₂ contribution are noise-related, indicating a constant structural state during the relaxation process. However, we may note that, since the numerous layers probed in the battery induce large diffraction peaks, small variations in local state of charge of both electrode material could be hard to detect in the present through-plane-averaged XRD setup. Therefore, we cannot strictly exclude an eventual relaxation in Young’s modulus that should impact the ToF. Nevertheless, our observations prove that Young’s modulus is not the main factor driving the ToF relaxation. iii) Mechanical relaxation of the electrode, coming from the electrode volume changes, could also impact the ToF signal. So far, the binder signal was not detectable due to its very small contribution (less than 2wt.% in each electrode). Then, in the present case, temperature relaxation due to ohmic relaxation takes part in the ToF relaxation, and we cannot exclude a contribution from mechanical relaxation, this one coming from the electrode binder or the separator.

Ultrasound technique can be used in battery management system to assess the state of health of battery but to date, the ultrasound signals are not easily correlated to any particular phenomenon. Based on our in-depth investigation correlating ultrasound time of flight signal to electrochemical response and structural evolution, obtained through operando XRD measurement, we demonstrated that the time-of-flight signal is mostly dominated by graphite negative electrode and in particular the evolution of its Young’s modulus during cycling. By using an operando dilatometric measurement in three-electrode setup on the same electrode materials, we were able to properly deconvolute the evolution of the NMC811 from the one of graphite. It allowed a complete correlation between the observed variation in ToF and the structural and electrochemical changes occurring upon charge and discharge, each electrode material being deconvoluted. At high voltage, where the phase transitions in both electrode materials are the most detrimental for long-term cycling, we demonstrated the sensitivity of ultrasound sensor to structural changes like the H2 to H3 phase transition of NMC811 electrode which is known to generate important stress on the particles. Quantitative links between the specific slopes observed in ToF variation with the different structural changes of electrode materials could be further established by setting repeated and reproducible experiments, including dedicated model systems or half-cell measurements. Our study demonstrated the interest of implementing ultrasound sensors in battery management system to track structural evolution and, potentially, provide reliable state of health evaluation through ultrasound modulations quantification.

Experimental set-up

The cell used for this study is an industrial KOKAM® pouch cell. It is composed of 16 double-coated and two single-coated sheets of NMC811 with a coating thickness of 65 µm on 11 µm (double coated) and 47 µm (single coated) aluminum current collector and 17 double-coated sheets of 80 µm graphite coated on an 8 µm copper current collector finally separated by a polyethylene film (15 µm) [23]. This cell has a nominal capacity of 11.6 Ah.

Ultrasound sensor

On the cell’s surface piezoelectric ceramic elements (manufactured from material PIC255 by PI Ceramic, LxBxT: 10 mm x 10 mm x 2 mm, thickness resonance frequency: 1 MHz) were glued in the center of both sides. Special care was taken to put the elements directly opposite of each other. At the synchrotron, an Olympus Epoch 650 was used to excite on of the piezoceramic elements with a single pulse (500 µs duration) and read out the other one’s delayed response. During additional measurements outside the beam, custom hardware was used to excite and read out the piezoceramic elements. The custom hardware further allows the excitation by a rectangular wave (1 MHz frequency) in contrast to a single pulse.

Cycling protocol

The cell was cycled at the synchrotron beamline as following. The cell was first discharged and charged at C/3 rate, followed by one cycle at C/8 rate, and a last one was done using a GITT “like” protocol with a cycling rate of C/3 and paused 1 hour for 1 hour.

XRD acquisition and integration

The XRD measurements have beam performed on the BM02-D2AM French CRG beamline at the ESRF. The beam energy has been set to 25 keV with a beam size of 100 * 100 µm². The detector used was a Cegitek Innovation detector, Si Hybrid pixel technology with a 130 µm² pixel size. The detector was mounted just after the sample to measure a large Q-range. The integration of the 2D images was made with the pyFAI library [39] after calibrating the geometry using LaB₆ samples. All integrations were corrected from the flat, the detector geometrical distortion and the incoming beam intensity.

XRD treatment

From the raw diffraction patterns obtained, specific ranges of q (Å⁻¹) were selected to isolate the peaks of interest (1.2 to 1.45 Å⁻¹ for 003 NMC Bragg reflection and 1.6 to 1.95 Å⁻¹ for 002 graphite Bragg reflection). Each diffractogram was fitted using the python LMFIT library [32] with a pseudo Voigt model, the peak position was then extracted, other metrics such as FWHM and amplitude were extracted but not relevant due to the high broadness of the peaks. Simpler metrics as the mass center of the diffractograms was extracted by a weight averaging of the peaks mean position.

Dilatometric measurement

The nano-dilatometric measurement was done with a EL-CELL ECD-3 Nano dilatometer set-up presented in Note 5 (Supporting Information). The dilatometer allows a measurement in three electrodes set-up, with one electrode to be probed in dilatometry. Here, graphite was used as working electrode and probed in dilatometry (measurement of the one-dimensional expansion along the electrode thickness), NMC811 was the counter electrode and metallic lithium was the reference. The assembly includes a T-frit as separator wetted by ~ 500 µL LP30 as electrolyte and an additional Whatman was placed between the T-frit and the counter electrode. All the dilatometer assembly was realized in an Ar-filled glovebox with less than 1 ppm of O₂ and H₂O, the cycling was then done in a thermoregulated enclosure at 25°C.

Declaration of Competing Interest

The authors declared that they have no conflicts of interest to this work.

Acknowledgment

Beamtime at the ESRF was granted within the Battery Pilot Hub MA4929 “Multi-scale Multi-techniques investigations of Li-ion batteries: towards a European Battery Hub”. We thank Dr. Samuel Tardif for his kind help at the beamline and the Region Auvergne Rhone-Alpes (Pack Ambition Recherche 2021 - Projet IsoBATT) for the financial support of Corentin Renais’s thesis. The authors are grateful for the financial support provided by the Federal Ministry of Education and Research as part of the research cluster Batterienutzungskonzepte through the project MADAM4Life (grant number 03XP0327C) and NUbase (grant number 03XP0322C). Open Access funding enabled and organized by Projekt DEAL.

Authors contribution

Corentin Renais: writing original draft, data treatment and acquisition and visualization/correlation; Benjamin Mercier-Guyon:writing original draft, data treatment and acquisition visualization/correlation; David Wasylowski: experimental design and acquisition, simulation, reviewing; Morian Sonnet: experimental design and acquisition, simulation, reviewing; Phillip Dechent: experimental design; Maxime Servajon: data treatment, software optimization; Nils Blanc: beamline management, experimental acquisition; Sandrine Lyonnard: experimental design, data acquisition, reviewing; Dirk Uwe Sauer: experimental design, reviewing; Claire Villevieille: experimental design, data treatment and acquisition, supervising original draft and reviewing, supervision of the project.

Feng X, Ren D, He X (2020) et M. Ouyang, « Mitigating Thermal Runaway of Lithium-Ion Batteries », Joule, vol. 4, n^o 4, pp. 743–770, avr. 10.1016/j.joule.2020.02.010
Liu K et al (2023) « Transfer learning for battery smarter state estimation and ageing prognostics: Recent progress, challenges, and prospects ». Adv Appl Energy 9:100117. 10.1016/j.adapen.2022.100117
Chen M, Ma G, Liu W, Zeng N, et, Luo X (2023) « An overview of data-driven battery health estimation technology for battery management system », Neurocomputing, vol. 532, pp. 152–169, mai 10.1016/j.neucom.2023.02.031
Berecibar M, Gandiaga I, Villarreal I, Omar N, Van Mierlo J, et, Van Den Bossche P (2016) « Critical review of state of health estimation methods of Li-ion batteries for real applications », Renew. Sustain. Energy Rev., vol. 56, pp. 572–587, avr. 10.1016/j.rser.2015.11.042
Zhou W, Lu Q, et, Zheng Y (2022) « Review on the Selection of Health Indicator for Lithium Ion Batteries », Machines, vol. 10, n^o 7, p. 512, juin 10.3390/machines10070512
Li X, Yuan C, Li X, et, Wang Z (2020) « State of health estimation for Li-Ion battery using incremental capacity analysis and Gaussian process regression », Energy, vol. 190, p. 116467, janv. 10.1016/j.energy.2019.116467
Eddahech A, Briat O, Bertrand N, Delétage J-Y, et, Vinassa J-M (nov. 2012) « Behavior and state-of-health monitoring of Li-ion batteries using impedance spectroscopy and recurrent neural networks ». Int J Electr Power Energy Syst 42:487–494. n^o 110.1016/j.ijepes.2012.04.050
Ge X et al (2024) « Revealing the electrochemical-mechanical correspondence between electrode films and 20 Ah prismatic Li-ion batteries via optical fiber monitoring », Chem. Eng. J., vol. 488, p. 150895, mai., 10.1016/j.cej.2024.150895
Guo Z, Briggs C, Vincent TA, Gulsoy B, Sansom JEH (2024) et J. Marco, « Global thermal image of cylindrical 21700 Li-ion batteries with distributed optical fibre sensor », J. Power Sources, vol. 592, p. 233980, févr. 10.1016/j.jpowsour.2023.233980
Zhu S et al (2021) « In operando measuring circumferential internal strain of 18650 Li-ion batteries by thin film strain gauge sensors ». J Power Sources 516:230669. 10.1016/j.jpowsour.2021.230669
Lyonnard S et al « Revealing How Internal Sensors in a Smart Battery Impact the Local Graphite Lithiation Mechanism », 31 mai 2024. 10.21203/rs.3.rs-4432970/v1
Juengert A, présenté à 44TH ANNUAL REVIEW OF PROGRESS IN QUANTITATIVE NONDESTRUCTIVE EVALUATION (2018) « Advanced ultrasonic techniques for nondestructive testing of austenitic and dissimilar welds in nuclear facilities », VOLUME 37, Provo, Utah, USA, p. 110002. 10.1063/1.5031581
Colchester RJ, Zhang EZ, Mosse CA, Beard PC, Papakonstantinou I (2015) Desjardins, « Broadband miniature optical ultrasound probe for high resolution vascular tissue imaging ». Biomed Opt Express 6:1502. n^o 410.1364/BOE.6.001502
Ladpli P, Liu C, Kopsaftopoulos F, et, Chang F-K « Estimating Lithium-ion Battery State of Charge and Health with Ultrasonic Guided Waves Using an Efficient Matching Pursuit Technique », in (2018) IEEE Transportation Electrification Conference and Expo, Asia-Pacific (ITEC Asia-Pacific), Bangkok, Thailand: IEEE, juin 2018, pp. 1–5. 10.1109/ITEC-AP.2018.8433297
Robinson JB, Maier M, Alster G, Compton T, Brett DJL, et, Shearing PR (2019) « Spatially resolved ultrasound diagnostics of Li-ion battery electrodes », Phys. Chem. Chem. Phys., vol. 21, n^o 12, pp. 6354–6361, 10.1039/C8CP07098A
Zhao G, Liu Y, Liu G, Jiang S, et, Hao W (2021) « State-of-charge and state-of-health estimation for lithium-ion battery using the direct wave signals of guided wave », J. Energy Storage, vol. 39, p. 102657, juill. 10.1016/j.est.2021.102657
Appleberry MC et al (2022) « Avoiding thermal runaway in lithium-ion batteries using ultrasound detection of early failure mechanisms ». J Power Sources 535:231423. 10.1016/j.jpowsour.2022.231423
Liu X et al (2023) « Decoupling of the anode and cathode ultrasonic responses to the state of charge of a lithium-ion battery ». Phys Chem Chem Phys 25:21730–21735. n^o 3210.1039/D2CP05948G
Owen RE et al (2022) « Operando Ultrasonic Monitoring of Lithium-Ion Battery Temperature and Behaviour at Different Cycling Rates and under Drive Cycle Conditions », J. Electrochem. Soc., vol. 169, n^o 4, p. 040563, avr. 10.1149/1945-7111/ac6833
Berhaut CL et al (2020) « Prelithiation of silicon/graphite composite anodes: Benefits and mechanisms for long-lasting Li-Ion batteries », Energy Storage Mater., vol. 29, pp. 190–197, août., 10.1016/j.ensm.2020.04.008
Aufschläger A, Kücher S, Kraft L, Spingler F, Niehoff P, et, Jossen A (2023) « High precision measurement of reversible swelling and electrochemical performance of flexibly compressed 5 Ah NMC622/graphite lithium-ion pouch cells », J. Energy Storage, vol. 59, p. 106483, mars 10.1016/j.est.2022.106483
Mohtat P, Lee S, Sulzer V, Siegel JB, et, Stefanopoulou AG (2020) « Differential Expansion and Voltage Model for Li-ion Batteries at Practical Charging Rates », J. Electrochem. Soc., vol. 167, n^o 11, p. 110561, janv. 10.1149/1945-7111/aba5d1
Laufen H et al (2024) « Correlation between Voltage, Strain, and Impedance as a Function of Pressure of a Nickel-Rich NMC Lithium‐Ion Pouch Cell », Adv. Mater. Technol., vol. 9, n^o 8, p. 2301965, avr. 10.1002/admt.202301965
Quilty CD, Bock DC, Yan S, Takeuchi KJ, Takeuchi ES (2020) et A. C. Marschilok, « Probing Sources of Capacity Fade in LiNi _0.6 Mn _0.2 Co _0.2 O ₂ (NMC622): An Operando XRD Study of Li/NMC622 Batteries during Extended Cycling », J. Phys. Chem. C, vol. 124, n^o 15, pp. 8119–8128, avr. 10.1021/acs.jpcc.0c00262
Michael H et al (2021) « A Dilatometric Study of Graphite Electrodes during Cycling with X-ray Computed Tomography », J. Electrochem. Soc., vol. 168, n^o 1, p. 010507, janv. 10.1149/1945-7111/abd648
Mohtat P, Lee S, Sulzer V, Siegel JB, et, Stefanopoulou AG (2020) « Differential Expansion and Voltage Model for Li-ion Batteries at Practical Charging Rates », J. Electrochem. Soc., vol. 167, n^o 11, p. 110561, janv. 10.1149/1945-7111/aba5d1
Xu R, Sun H, de Vasconcelos LS, et, Zhao K (2017) « Mechanical and Structural Degradation of LiNi. Mn y Co z O 2 Cathode Li-Ion Batteries: Experimental Study » J Electrochem Soc 164:A3333–A3341. n^o 1310.1149/2.1751713jes
Davies G et al (2017) « State of Charge and State of Health Estimation Using Electrochemical Acoustic Time of Flight Analysis », J. Electrochem. Soc., vol. 164, n^o 12, pp. A2746–A2755, 10.1149/2.1411712jes
Qi Y, Guo H, Hector LG, et, Timmons A (2010) « Threefold Increase in the Young’s Modulus of Graphite Negative Electrode during Lithium Intercalation », J. Electrochem. Soc., vol. 157, n^o 5, p. A558, 10.1149/1.3327913
Cheng EJ, Taylor NJ, Wolfenstine J, et, Sakamoto J (2017) « Elastic properties of lithium cobalt oxide (LiCoO ₂) », J. Asian Ceram. Soc., vol. 5, n^o 2, pp. 113–117, juin 10.1016/j.jascer.2017.03.001
The Matplotlib Development Team (2024) Zenodo. 10.5281/ZENODO.592536
Newville M, Stensitzki T, Allen DB, Ingargiola etA (2014) LMFIT: Non-Linear Least-Square Minimization and Curve-Fitting for Python21 septembre. Zenodo. 10.5281/ZENODO.11813
Dahn JR (nov. 1991) « Phase diagram of Li x C 6 ». Phys Rev B 44:9170–9177. n^o 1710.1103/PhysRevB.44.9170
Busà C, Belekoukia M, Loveridge etMJ (2021) « The effects of ambient storage conditions on the structural and electrochemical properties of NMC-811 cathodes for Li-ion batteries ». Electrochim Acta 366:137358. 10.1016/j.electacta.2020.137358
Levi MD et, Aurbach D (1997) « Simultaneous Measurements and Modeling of the Electrochemical Impedance and the Cyclic Voltammetric Characteristics of Graphite Electrodes Doped with Lithium », J. Phys. Chem. B, vol. 101, n^o 23, pp. 4630–4640, juin 10.1021/jp9701909
Liu S et al (nov. 2023) « Origin of Phase Separation in Ni-Rich Layered Oxide Cathode Materials During Electrochemical Cycling ». Chem Mater 35:8857–8871. n^o 2110.1021/acs.chemmater.3c00924
Dees DW, Kawauchi S, Abraham DP, et, Prakash J (2009) « Analysis of the Galvanostatic Intermittent Titration Technique (GITT) as applied to a lithium-ion porous electrode », J. Power Sources, vol. 189, n^o 1, pp. 263–268, avr. 10.1016/j.jpowsour.2008.09.045
Ma S et al (2018) « Temperature effect and thermal impact in lithium-ion batteries: A review », Prog. Nat. Sci. Mater. Int., vol. 28, n^o 6, pp. 653–666, déc. 10.1016/j.pnsc.2018.11.002
Kieffer J, Valls V, Blanc N, et, Hennig C (2020) « New tools for calibrating diffraction setups », J. Synchrotron Radiat., vol. 27, n^o 2, pp. 558–566, mars 10.1107/S1600577520000776

There is NO Competing Interest.

BM02SupportingInformationVF.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Is ultrasound suited to probe electrochemical processes in batteries? An operando multiprobe investigation of NMC/graphite commercial cell

Status:

Version 1

Abstract

Figures

Introduction

Results

Conclusion

Materials and Methods

Experimental set-up

Ultrasound sensor

Cycling protocol

XRD acquisition and integration

XRD treatment

Dilatometric measurement

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1