Journal of The Electrochemical

Society

OPEN ACCESS You may also like

- (Invited) Latest Advances in 2D and 3D
Lithium-Ion Cells in Automotive Applications: Tesla Microstructural Characterization of Fuel
Cell Components
4680 Cylindrical Cell Teardown and Jasna Jankovic

- (Digital Presentation) Develop Cell Internal

Characterization Resistance Measurement Tool As a Lib
Cell Characterization Technology
Jaesik Chung, Gunho Kwak, Kwang Jung
To cite this article: Manuel Ank et al 2023 J. Electrochem. Soc. 170 120536 et al.

- Active Nano-Characterization and

Technology
Daisuke Fujita

View the article online for updates and enhancements.

This content was downloaded from IP address 103.145.109.166 on 19/07/2024 at 07:25

 Journal of The Electrochemical Society, 2023 170 120536

Lithium-Ion Cells in Automotive Applications: Tesla 4680

Cylindrical Cell Teardown and Characterization
Manuel Ank, 1,z Alessandro Sommer,2 Kareem Abo Gamra,1 Jan Schöberl, 1

Matthias Leeb,2 Johannes Schachtl,2 Noah Streidel,2 Sandro Stock, 2

Markus Schreiber,1 Philip Bilfinger,1 Christian Allgäuer,1 Philipp Rosner,1
Jan Hagemeister,2 Matti Rößle,1 Rüdiger Daub,2 and Markus Lienkamp1
1
Technical University of Munich (TUM), School of Engineering & Design, Department of Mobility Systems Engineering,
Institute of Automotive Technology, Garching, D-85748, Germany
2
Technical University of Munich (TUM), School of Engineering & Design, Department of Mechanical Engineering, Institute
for Machine Tools and Industrial Management, Garching, D-85748, Germany

Battery research depends upon up-to-date information on the cell characteristics found in current electric vehicles, which is
exacerbated by the deployment of novel formats and architectures. This necessitates open access to cell characterization data.
Therefore, this study examines the architecture and performance of first-generation Tesla 4680 cells in detail, both by electrical
characterization and thermal investigations at cell-level and by disassembling one cell down to the material level including a three-
electrode analysis. The cell teardown reveals the complex cell architecture with electrode disks of hexagonal symmetry as well as
an electrode winding consisting of a double-sided and homogeneously coated cathode and anode, two separators and no mandrel. A
solvent-free anode fabrication and coating process can be derived. Energy-dispersive X-ray spectroscopy as well as differential
voltage, incremental capacity and three-electrode analysis confirm a NMC811 cathode and a pure graphite anode without silicon.
On cell-level, energy densities of 622.4 Wh/L and 232.5 Wh/kg were determined while characteristic state-of-charge dependencies
regarding resistance and impedance behavior are revealed using hybrid pulse power characterization and electrochemical
impedance spectroscopy. A comparatively high surface temperature of ∼70 °C is observed when charging at 2C without active
cooling. All measurement data of this characterization study are provided as open source.
© 2023 The Author(s). Published on behalf of The Electrochemical Society by IOP Publishing Limited. This is an open access
article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY, http://creativecommons.org/licenses/
by/4.0/), which permits unrestricted reuse of the work in any medium, provided the original work is properly cited. [DOI: 10.1149/
1945-7111/ad14d0]

Manuscript submitted November 7, 2023; revised manuscript received December 7, 2023. Published December 29, 2023.

Since its commercialization by Sony in 1991, lithium-ion battery contrast, academic lithium-ion battery (LIB) research is typically
technology has undergone a development with significant conducted on a laboratory scale or utilizing pilot production lines
advances.1,2 Due to their high energy density, robustness and using smaller cell formats, easy-to-make cell designs, and manual or
efficiency, the global electric vehicle stock consisted in 2022 of semi-manual production processes.15 Therefore, research-grade
more than 26 million vehicles, marking an increase of 60 % LIBs inherently differ from commercially produced battery cells
compared to the previous year.3 However, a diverse set of possible and can thus only serve as a reference for few properties.15–18
approaches regarding lithium-ion battery technology, consisting of The described trend toward larger cell formats and the precarious
different battery chemistries, formats and implementation concepts, situation regarding the availability of LIB production data substantiates
exists. These range from common 18650 cylindrical cells with the scientific community’s interest in studying the characteristics of
around 3 Ah to large-format pouch or prismatic cells with capacities large-format cylindrical cells. Previously published articles highlighting
upwards of 100 Ah.4 Depending on the utilized concept, different and characterizing state-of-the-art cells further demonstrate the great
advantages and disadvantages arise with regard to cycle life, thermal appeal of such scientific manuscripts, as the data can be used, e.g. for
properties, gravimetric and volumetric energy density as well as battery modeling and further research.11,15,18–20 Since the combination
safety. of a novel cell format and a tabless architecture governs the actual cell
The 4680 cell format (46 mm diameter, 80 mm axial length) was behavior, this study provides the basis for a improved understanding of
introduced for the Model Y electric vehicle at Teslaʼs Battery Day these characteristics.
20205,6 and promises higher energy content and power capability
due to the larger format compared to frequently used 18650 or
Previous research.—Due to the short period of availability and
21700 cylindrical cells7 and a “tabless” design. Compared to a
limited procurement options from series-production vehicles, only
21700 cell, the new format increases volume by a factor of 5.5.8
comparatively few studies on the 4680 cylindrical cell format have
Large-format cylindrical cells are expected to reduce production
been published to date. Frank et al.21 used an experimentally
costs in pack assembly, as fewer individual cells need to be
validated multidimensional multiphysics model describing a high
assembled and interconnected to achieve the same battery pack
energy NMC811/Si-C cylindrical lithium-ion battery to evaluate the
characteristics. In addition to Tesla, the manufacturer BMW
effects of tabless design and cooling topologies for, among others,
announced the launch of its new electric vehicle platform called
4680 cell formats under varying charging protocols. The results
“Neue Klasse”, also utilizing cylindrical cells with a diameter of
show that a tabless design contributes to improving the charging
46 mm (with two different axial lengths).9
performance by reducing polarization and ohmic heating with an
Battery cell manufacturers usually only provide limited informa-
improved cell homogeneity and thus enables format-independent
tion about cell characteristics and do not disclose their highly
scalability of fast-charging performance with a tab-cooling topology.
complex, interdependent manufacturing processes.8,10–12 In addition,
However, these results were obtained purely by simulation and using
cell chemistry and cell design vary depending on the cell and
hypothetical layouts. Mayer et al.22 used data from Frank et al.21 to
original equipment manufacturers as well as customer requirements,
identify a processing strategy for high repetition ultrafast laser
which leads to manufacturer-specific production chains.13,14 In
ablation scenarios. Tranter et al.23 numerically explored the electro-
chemical and thermal behavior of a 4680 cylindrical cell and
revealed a mitigation of ohmic losses occurring around the jelly-
z
E-mail: [email protected] roll current collectors, which are considerable in the common tabbed
 Journal of The Electrochemical Society, 2023 170 120536

case, thus resulting in higher efficiency and capacity as well as lower discharges and galvanostatic electrochemical impedance spectro-
heat generation. As with Frank et al.,21 this study presented entirely scopy (EIS) measurements of multiple cells.
simulative results. Pegel et al.24 developed geometrical models Hybrid pulse power characterization of one cell: Investigation of
describing the properties of the cells’ structural components with the resistance behavior of one cell using a hybrid pulse power
regard to certain loadcases and explore the volumetric efficiency and characterization (HPPC) measurement employing 10 s charge and
energy density as a function of cell dimensions and housing material discharge pulses at C-rates of 0.5C to 1.5C in 10 % state-of-charge
choices. Once again, no actual cell with a diameter of 46 mm was (SOC) increments from 10 % to 90 % SOC.
investigated in the study, with the models being based on a 10 Ah Open source provision of all measurement data: All raw
cylindrical reference cell. Shen et al.25 used a distributed 3D coupled measurement data, including cell teardown images, SEM/EDX
electro-thermal equivalent circuit network model of a 4680 cell to recordings, and test procedures are provided as open source along-
study different thermal management options and demonstrate that side the article.

Figure 1. Overview of the lithium-ion cells examined, their cell ID, condition upon arrival and classification as well as tests conducted.

there does not need to be a trade-off between cell size and thermal Experimental
performance, as long as cylindrical cells are designed tabless. To the
The cells under study and the experimental procedures of cell
best of the authors’ knowledge, the first academic article with actual
teardown as well as electrical characterization and thermal investi-
4680 cells from Tesla was presented by Baazouzi et al.,8 in which
gation are described hereinafter.
the authors investigated 19 cylindrical lithium-ion battery cells from
four cell manufacturers in four formats (18650, 20700, 21700, 4680)
Overview of tested cells.—All cells under investigation in this
with respect to their design features, such as tab design and quality
study originate from a Tesla Model Y (year of manufacture 2022) and
parameters. The results show that the Tesla 4680 cell architecture is
were procured in late 2022 and early 2023, thus belonging to the first
not fully tabless since only the jelly roll design is engineered in a
generation of 4680 cells. Eight of the nine cells were obtained via an
tabless manner. Approx. 25 % of the cathode and 30 % of the anode
engineering and manufacturing consulting firm in the United States
feature no notches, thus resulting in long thermal and electrical
(Munro & Associates, USA), with an additional cell (herein titled as ID
transport paths. However, the authors focus on the comparison with
xxx/828) made available by industry which still was attached to parts of
additional cells of different formats and not on an extensive
the side cooling system. No precise exposure history is available for any
characterization of the 4680 cell or its production processes.
of the cells, but in the case of the cell from a different source in
According to communication with the authors, the investigated
particular, previous investigations are to be assumed, which is why this
cell appears to be from the identical vehicle as most test specimens
cell is excluded from any further examinations and thus only serves as a
investigated in this paper, allowing for comparative observations.
visual comparator. Upon receipt of the goods, all cells were first
This teardown study ties in methodically with the characteriza-
visually inspected revealing several dents near the top assembly on four
tion research of Günter et al.11 as well as Stock et al.,15 who
of the cells, presumably resulting from the forceful disassembly process
examined an automotive 78 Ah pouch cell from the Volkswagen
from the vehicle. All residual structural foam was removed using
ID.3 and a 161.5 Ah prismatic hardcase lithium iron phosphate
acetone. An overview of the cells under study and their condition at
(LFP) cell from a Tesla Model 3 (Standard Range), respectively.
incoming inspection is given in Fig. 1. The Figure also provides an
overview of the assignment of the cells to the experimental test
Contributions.—This characterization article extensively ana- procedures performed in each case, since the different characterizations
lyzes cylindrical 4680 lithium-ion cells from the so-called “first were performed in parallel. The cell ID nomenclature was adopted from
generation” that were extracted from a state-of-the-art Tesla Model Y the supplier and refers to the number of the cell in the Tesla Model Y
(model year 2022, manufactured in Austin, USA). The following battery pack. According to the supplier, there are 828 cells arranged in
contributions are particularly noteworthy: four modules.26,27 No official data sheet was provided upon delivery by
In-depth cell architecture and electrode material analysis: the manufacturer or supplier.
Detailed exploration of the cell architecture, analysis of the material
composition using scanning electron microscopy (SEM) and energy- Cell teardown and material extraction.—One cell (ID 083/828)
dispersive X-ray spectroscopy (EDX) and investigation of electrode was selected for teardown, discharged completely to 2.5 V, and
properties over the full length of the anode and cathode sheet. brought into a glove box specially designed for teardown analysis
Three-electrode analysis: Assembly of 1/2″ T-cells with gold wire (MBraun, Germany). Filled with an argon atmosphere, the values for
as a reference electrode and subsequent investigation of anode and H2O and O2 were constantly monitored and regulated to below
cathode characteristics using pseudo open-circuit voltage (pOCV) and 0.1 parts per million (ppm). The setup has been used for previous
galvanostatic electrochemical impedance spectroscopy (EIS). cell teardown studies11,28 and is illustrated in Fig. 2.
Thermal investigation during charging: Analysis of the heat To open the casing of the cell, a milling spindle (Proxxon,
generation at cap, center and bottom of one cell’s can during Germany) with a cutting disc with a diameter of 32 mm (Dremel,
charging rates of up to 2C utilizing a thermal imaging camera. USA) was used within the glove box, ensuring that the cutting disc
Electrical characterization of several cells: Extensive performance would only penetrate the casing and not the electrodes. First, a cut
characterization at the cell level using constant current (CC) charges/ was performed around the perimeter of the top of the cell, so that the
 Journal of The Electrochemical Society, 2023 170 120536

of the electrode using isopropanol and a scalpel, after which the

substrate was measured. The density of the electrodes’ active material
was calculated by subtracting the current collector foil, based on the
areal mass loading and the thickness of the electrodes. Scanning
Electron Microscopy (SEM) was conducted utilizing a JEOL IT-200
instrument (JEOL, Japan) operating at 15 kV, equipped with a Low
Energy Ionization (LEI) detector, and varying magnifications were
employed (as depicted in the images). The identical apparatus was also
employed for energy-dispersive X-ray spectroscopy (EDX) measure-
ments. For the SEM and EDX analysis, one sample each of the
cathode and the anode were also taken with the punch.

Three-electrode analysis.—A three-electrode setup consisting of

two original electrodes and an additional gold wire as a reference
electrode enables the investigation of single-sided electrodes.15,29 In
this case, T-cells (1/2”, Swagelok, USA) were used for the three-
electrode setup of the disassembled cell (ID 083/828). The elec-
trodes were delaminated on one side using acetone and isopropanol
for the cathode and anode, respectively. The anode in particular
required intensive delamination due to the primer needed for dry
coating. The electrodes were washed with diethyl carbonate (DEC)
for 24 h, leading to delithiation of the anode and potential removal of
parts of the solid electrolyte interphase (SEI).30 Electrodes with a
diameter of 10.95 mm were used for the T-cells, which were taken
with a hand punch (Nogami, Japan). In the assembly, the gold wire
(Goodfellow, England) with a diameter of 50 μm and an insulation
of 7 μm of polyimide was placed between two separators. These
separators (type 691, VWR International, USA) measure 250 μm in
thickness and 11 mm in diameter. Before assembling the cell, all
components were vacuum dried at 100 °C for 24 h to remove
residual water, which could cause damage to the cells. An argon-
Figure 2. Overview of the test setup for cell teardown. filled glove box (MBraun, Germany) was used for assembly. For the
electrolyte, 60 μL of LP572 (BASF, Germany) was used. All
subsequent activation and test protocols were carried out in a climate
lid could be removed. Since at this point, the lid was still attached to chamber at 25 °C using a potentiostat (VSP-3e, Bio-Logic, France).
the anode current collector, it was removed by applying a slight After assembly, the gold wire was lithiated with the cathode using a
force with pliers thus exposing the jelly roll. Due to the radial current of 150 nA for one hour. In addition, an open circuit phase of
pressure of the jelly roll acting on the cell housing, it is not possible 2 h allowed complete wetting of the electrodes. To reform the SEI
to remove the jelly roll simply by pulling on it. It was therefore after the washing process, three activation cycles with C/10 were
necessary to perform an additional vertical incision on the side wall carried out. Based on the determined capacity of the full cell and the
of the casing. Then, the metal casing was peeled back using a needle measured cathode area, the specific capacity is calculated to be
nose plier and the jelly roll was freed from the casing for further 4.9 mAh/cm2 (used for determining all test currents). A pOCV
analysis. Upon extraction of the jelly roll, no surplus electrolyte was measurement at C/50 in a constrained voltage range of 2.9 V to
extractable for analysis. Measurements of the cell and all of its 4.2 V and a galvanostatic electrochemical impedance spectroscopy
components in combination with images were taken throughout the (EIS) measurement from 100 kHz to 100 mHz (10 mV excitation
entire teardown process and are used throughout this work. amplitude (potential), 6 points per decade (ppd), 3 average measures
Additionally, the DataMatrix code located on the bottom side of per frequency (amf)) at 50 % SOC were carried out. A total of six
the cell can could be read using a handheld scanner (DS3608, Zebra cells were built to exclude outliers due to manufacturing defects. The
Technologies, USA). Cell teardown images can be found in the raw measurement data and the activation and test protocol can be
supplementary material. The teardown process of the cell is found in the supplementary material, while the electrical test
illustrated in Fig. 3. protocol is provided in Table A·I.

Structural and elemental analysis.—After opening the cell (ID Electrical characterization.—To evaluate the resistance of the
083/828), the electrode assembly was entirely removed from the full cell under test at different SOCs, a HPPC measurement was
casing and geometrically measured. The current collector discs were performed. For this purpose, the cell (ID 186/828) was subjected to
removed from both the cathode and anode sides of the jelly roll, and 10 s charge and discharge pulses at C-rates of 0.5 C, 1 C and 1.5 C
their dimensions along with the assembly structure were measured after 1 h of relaxation, each. The analysis was performed in 10 %
(cf. Fig. 3). Subsequently, the adhesive tapes of the jelly roll were SOC steps from 10 % to 90 % SOC. After each pulse, a pause of
removed, and the electrode assembly was unwound, resulting in 300 s was added, before the charge was reset using coulomb
separate cathode, anode, and the two separators. The thickness of all counting and the cell was relaxed for 1 h before the next pulse.
components was measured using a Micromar 40 EWRi (Mahr, The HPPC test protocol is presented in Table A·II.
Germany). Lengths and widths of the components were measured Capacity characterization protocols, which included a C/3 CC
using calipers and a folding rule. To determine the areal mass discharge capacitance determination and C/20 pOCV measurements
loading of the electrodes, samples were extracted using Nogami in charge and discharge direction, were applied to eight cells (ID
punches (Nogami, Japan) with a diameter of 14 mm for cathodes and 083/828, 131/828, 186/828, 444/828, 536/828, 549/828, 550/828,
15 mm for anodes. These samples were then weighed using a high- 601/828) and are presented in Table A·III.
precision scale. Starting from the core of the roll, samples were Three cells (ID 131/828, 186/828, 549/828) were subsequently
taken and measured every 50 cm up to the ends. For determining the subjected to galvanostatic EIS (10 kHz to 10 mHz, 10 ppd (loga-
current collector thickness, the coating was removed from both sides rithmic spacing), 500 mA excitation amplitude, 5 amf, drift
 Journal of The Electrochemical Society, 2023 170 120536

Figure 3. Display of the teardown process of one cell (ID 083/828) as well as depiction of partial measurements on the opened cell. (a)–(d) Outer cell
measurements and cell opening, with weld spot of the housing visible in (b); (e) Notched and folded anode tabs (f) Anode disk; (g) Notched and folded cathode
tabs; (h) Cathode disk; (i) Dismanteled jelly roll; (j)–(o) Inner cell architecture and component measurements. Additional cell teardown images can be found in
the supplementary material.

correction) at 80 %, 50 % and 20 % SOC in a separate test setup

optimized for impedance measurements (no adjacent experiments,
twisted pair cable feed, shielding in separate test chamber). The
corresponding test protocol is provided in Table A·IV.
For all test procedures at cell level a lower cutoff voltage of 2.5 V
was selected, with 4.2 V as the upper limit (voltage range based on
Baazouzi et al.8 and publicly available battery characteristics31). A
high precision battery cycler (BCS-815, Bio-Logic SAS, France)
was used to perform the electrical capacity characterization proce-
dures and the galvanostatic EIS measurements, with cell holders
model KSR-150-OS (Voltavision, Germany) to constrain and con-
nect the cells using a 4-wire-measurement. The HPPC measurement
was performed using a separate test system (XCTS50, Basytec,
Germany) to allow for the required higher currents. The HPPC tests
and the galvanostatic EIS measurements were performed in a IPP110 Figure 4. Overview of the test setup for electrical characterization (except
climate chamber (Memmert, Germany) at a constant temperature of where noted otherwise).
25 °C ± 0.2 °C, while all other electrical tests on cell level were
performed in a test chamber model KB400 (BINDER, Germany) at voltage and capacity signal and is plotted against the capacity. In this
identical setpoint temperature. Furthermore, the capacity determina- work, the DVA was normalized by the nominal capacity converting
tions of the cells with ID 083/828 and 601/828 were performed using the unit to Volts and shifting the regime of the data to around 1 V.
a similar test setup (BCS-815 cycler and KB400 test chamber), Forming the derivative of a measured discrete signal induces noise,
whereby however another cell holder (Battery Holder 60A— so that both the voltage and capacity signal were smoothed by a
Universal Cylindrical Cell, ARBIN Instruments, USA) was utilized mean filter with the window size of 73 points before derivation. The
due to logistical reasons. The raw measurement data as well as all DVA was again filtered by a mean filter with window size of 363
corresponding test protocols of the electrical characterizations can be from the left to the right, and vice versa, to compensate for induced
found in the supplementary material. The general test setup for the shifts. Generally, care must be taken when filtering the DVA, as
electrical characterizations is shown in Fig. 4. information can be lost or misinterpreted.36,37
To gain deeper insights for revealing transition phases of the cell
chemistry, non-invasive electro-chemical voltage spectroscopy Thermal investigation.—One cell (ID 536/828) was selected for
methods such as differential voltage analysis (DVA) and incremental the evaluation of heat generation during charging. The cell was
capacity analysis (ICA) were calculated. Hereby, the peaks in the cycled with an XCTS50 test system (Basytec, Germany) starting
DVA and ICA provide information about e.g. phase transitions in the from a discharged state with a CC charge profile with charge rates of
electrodes.32–35 The DVA is extracted by the ratio of the derived 1C and 2C up to the upper cutoff voltage of 4.2 V. The charging
 Journal of The Electrochemical Society, 2023 170 120536

Figure 6. Exploded view of the parts of the 4680 cell. The components are
drawn based on the measured dimensions of the cell, current collector disks,
and the jelly roll.

tapes. For the internal connection of the cell poles a disk tab design
Figure 5. Overview of test setup for thermal investigations. (a) View in the
direction of the thermal imaging camera; (b) View from thermal imaging
with notched and folded electrode tabs is utilized.8 The cathode disk
camera. is made out of aluminum and the anode disk out of copper with both
having a thickness of 0.2 mm. The dimensions of the two disks are
depicted in Fig. 7. Both disks exhibit hexagonal symmetry, but differ
phase was followed by a 30 min rest period to observe the
in the position of the joint to the leaf-shaped connector. The
subsequent cooling behavior. The cell was then discharged to the
connector of the cathode disk is joined at the outer ring whereas
cutoff voltage of 2.5 V at a discharge rate of C/3. The test protocol is
the connector of the anode disk is joined toward the center. The outer
provided in Table A·V. Analogously to the electrical characteriza-
ring of the anode disk is connected to the cell housing, whereas for
tion, the cell was secured in a KSR-150-OS cell holder (Voltavision,
the cathode disk the center is ultrasonically welded to the positive
Germany). A thermal imaging camera (A70, FLIR Systems, USA)
terminal of the cell. Therefore, the two disks can act like compen-
was used to record the axial temperature distribution at the cell can.
sating elements or springs when attaching the jelly roll with the
This camera has a measurement range of −20 °C to 175 °C with a
housing. Each connector is joined to the notched and folded
measurement accuracy of ±2 % and a resolution of 640 × 480 pixels
electrodes via five laser-welded zig-zag seams with a width of about
at a frame rate of 30 Hz. The focus of the camera was set to 1.2 m
0.5 mm. The lengths of the seams are depicted in Fig. 7.
and the camera was placed at the same distance from the cell surface.
Furthermore, a plastic disk, measuring 0.65 mm in thickness, serves
The cell can was covered with an emissive tape (ϵ = 0.95) to avoid
as an insulating barrier between the cathode disk and the cell
measurement inaccuracies due to reflective surfaces and unknown
housing, which is at anode potential. This arrangement facilitates a
emissivity. Temperature data was recorded and analyzed using FLIR
connection solely to the positive terminal.
Research Studio version 2.1.0 software. The measurement points in
The manufacturing process of the cell can be reconstructed as
the temperature profile at the cap, center, and bottom of the cell were
follows. The jelly roll with the folded electrode tabs are produced.
evaluated in order to evaluate the temperature spread across the cell
On the anode side, the copper disk is laser-welded to the copper
can during the charging process. In addition, the temperature
foils. On the cathode side, the aluminum disk with a previously
measurement at the center was validated with a PT100 temperature
ultrasonically welded cell tab is laser-welded to the aluminum foils.
sensor at the same location to exclude deviations compared to other
The plastic disk is placed onto the cathode disk and afterwards the
temperature measurement methods. The experimental setup was
assembly is inserted with the cathode side first into the cell can. On
placed in an EBU-3EA0PT-40 temperature chamber (ESPEC,
the anode side a cap with the filling hole is placed on top and the cell
Japan), and the ambient temperature was kept constant at
can is crimped and sealed. After the filling process, the filling hole is
25 ± 0.5 °C throughout all of the tests. No active cooling of the
sealed with a copper rivet. The DataMatrix code on the bottom side
cell was conducted. The raw measurement data and the corre-
of this cell can contains the information P1551370-10-D:
sponding test protocols of the thermal investigation can be found in
STRR221370808AK. It is assumed to be a unique identification
the supplementary material. The test setup for the thermal investiga-
for this specific cell, used for tracking and tracing applications in the
tion is depicted in Fig. 5.
manufacturing process and in vehicle deployment.40 The traceability
of battery cells offers a crucial means to ensure quality, safety, and
Results and Discussion
efficiency in battery production and usage.
The results of the experiments are presented hereinafter.
Particular attention is paid to the cells’ characteristics and archi- Electrode design.—The electrode winding consists of a double-
tecture. sided coated cathode and anode, as well as two separators. The
lengths and widths of these components can be seen in Fig. 8. Since
Cell and jelly roll architecture.—Figure 6 shows an exploded the component widths are different in size, they are depicted in an
view of all relevant parts within the cell. The cylindrical part of the enlarged form.
cell has an outer diameter of 46 mm and a height of 80 mm. The The structure of the electrodes resembles that of conventional
positive terminal with a diameter of 16 mm adds another 1 mm to the round cell designs.8 The total length of the anode is 3403 mm, which
total height of 81 mm. On the negative terminal side, a copper rivet is 136 mm longer than the cathode. In the assembled state, the
which seals the filling hole is placed in the center. The thickness of cathode is completely wrapped by the anode. Furthermore, the two
the housing is measured to be 0.5 mm which leads to a cylindrical separators are longer than the electrodes, with the longer separator 2,
inner volume of 125.6 cm3 (sealings and three-dimensional form on in the assembled state, completely wrapping the electrode assembly
the anode side not considered). The wall thickness is increased of cathode, anode, and separator 1 once again and being secured with
compared to common commercial cells in 18650 or 21700 format, blue tape. The current collector flags, which enable internal cell
which may be due to the alleged structural use of the cells in the contact, are not located along the entire length of the electrode but
vehicle design, yet it reduces the energy density.38,39 After removing are symmetrically positioned relative to the center of the electrode
the cathode and anode disks, the jelly roll has a height of 71 mm and length. In Fig. 9, the measured thicknesses and the calculated areal
a diameter of 44.5 mm. There is no winding mandrel in the core of mass loadings along the electrodes are displayed. A length of 0
the roll, but rather an empty space with a diameter of 5 mm. The roll corresponds to the electrode position at the core of the inner
is secured and held together at both ends with two 10 mm wide blue winding.
 Journal of The Electrochemical Society, 2023 170 120536

Figure 8. Dimensions of the electrodes and separators. The schematic

enlargement of the cathode (a) can be seen in (a*) and of the anode (b) in
(b*).

On the cathode side, spherical active material particles were

found. They have diameters ranging from 3 μm to 16 μm. These
Figure 7. Dimensions of (a) cathode and (b) anode disk including a close-up characteristics are typical for NMC materials, which can be
of the leaf-shaped connectors with length and shape of the laser weldings. confirmed by EDX measurements available in the supplementary
Drawings are not to scale. material. The element distribution was taken from an EDX measure-
ment taken from the surface of the cathode. NMC was used as the
active material, featuring a mass ratio of 81.8 % nickel, 12.1 %
It can be observed that, at both the cathode and anode, the electrode cobalt, and 6.1 % manganese, considering only the mass ratios of the
thickness at the beginning of the winding is highest at the core and then three elements. The EDX analysis of fluorine showed accumulations
decreases significantly. Beyond a 1 m electrode length, the electrode between the active material particles and a total amount of 7.9 %.
thickness increases again. In contrast, the electrode loadings weighed This led to the conclusion that polyvinylidene fluoride (PVdF) was
from the electrode samples remain almost constant throughout their used as a binder as it is one of the most commonly used binders in
length. The non-uniform electrode thickness can be explained by the cathodes.41 Small traces (<0.5 % each) of phosphorus and sulfur
following approach: Since the cell to be opened was already in were also detected by the EDX which are residues of lithium salts
operation, the electrodes underwent volume changes due to charging used in the electrolyte. Examples for commonly used lithium salts
and discharging processes. Because the electrodes are differently containing phosphorus and sulfur are lithium hexafluorophosphate
compressed due to the jelly roll winding structure, this results in (LiPF6) and lithium bis(fluorosulfonyl)imide (LiTFSI).41 The pre-
subsequent and varying changes in thickness. As no mandrel was used sence of minor cracks on the sides of larger particles may indicate
in the core of the cell, it is plausible to assume that the electrodes the application of substantial forces during the calendaring step.42
toward the center (length = 0) have more space for volume change and However, only the particles in the top layer are affected by this
resulting thickness changes than those in the middle of the winding. phenomenon as can be seen in the side profile. The side profile also
shows the aluminum current collector and that the two sides of the
Material characterization.—For a more in-depth analysis of the double-sided coating have a horizontal offset of approx. 140 μm. No
electrodes, SEM images were captured, with Fig. 10 displaying both elevation of the edge can be seen. On the edge of the cathode
top and side profile views of the cathode and anode. coating, an extra coating layer can be identified. It consists of a finer,
 Journal of The Electrochemical Society, 2023 170 120536

cathode coating, it can be inferred that it was applied subsequently to

the primary coating process and serves as an electrically insulating
or protective layer.
The anode consists of flake graphite particles.43 The image shows
that the particles have diameters of up to 35 μm, and the EDX analysis
confirms the exclusive use of graphite as the active material, with no
detectable presence of silicon in the measurements. A dominant carbon
content of 90.5 % was detected, consistent with graphite’s composition,
and also consistent with the typical composition of electrically
conductive additives such as carbon black and most binders.
Furthermore, fluorine is present at a significant quantity of 7.9 %,
which may suggest the presence of a binder containing fluorine, such as
PVdF as found in the cathode, or polytetrafluoroethylene (PTFE). The
utilization of PTFE could indicate a solvent-free anode fabrication
method.44 Similarly to the cathode, trace amounts of phosphorous and
sulfur are detected (<0.5 % each), supporting the hypothesis that they
originate from lithium salts in the electrolyte. Although copper, as a
heavy metal element used in the current collector foil, is evident
throughout the coating and appears in the EDX measurements, its
quantity is minimal, constituting less than 1 %. Between the coating and
the copper collector, a black primer coating was observable. It exhibited
a notably distinct structure from the actual anode coating and extended
approximately 1.5 mm beneath the coating. EDX analysis revealed that
it primarily comprises carbon and fluorine, leading to the inference that
the same binder as in the coating, along with carbon black for electrical
conductivity, was used. This primer coating strongly suggests the
utilization of a solvent-free coating process. The advantage here is
the elimination of the need for harmful solvents and the avoidance of
the energy-intensive production step of drying the anodes.44

Three-electrode analysis.—Using the three-electrode setup, the

pOCV for the full cell, anode, and cathode can be extracted with a
current of C/50, shown in Fig. 11a. During the C/50 cycle, all cells
achieved coulombic efficiencies above 99 %, indicating representative
results. As can be seen in Fig. 11, the anode potential shows
characteristic staging behavior during the lithiation at around 0 V vs
Li/Li+, which is also reported by Nitta et al.45 Considering the potential
of the gold wire after lithiation, three distinct potential plateaus are
visible at 210 mV, 120 mV, and 85 mV. Through the steps, the phase
transition of the graphite is visible with LiC24, LiC12, and LiC6,
respectively. During the transition, the potentials are almost constant.
This behavior is characteristic of a pure graphite anode, which leads also
to the conclusion that the anode does not contain any silicon. In addition,
the measured cathode potential shows the characteristic pattern of the
NMC chemistry. As shown in Fig. 11a, the charging and discharging
Figure 9. Electrode properties over the full length of the anode and cathode curves reveal overpotentials of around 0.5 V even at a small current of
sheet. (a) Electrode thickness; (b) Loadings; (c) Density. Loadings were C/50, indicating a high internal cell resistance. This can be caused by the
determined by punching out coins and thicknesses by a micrometer screw three-electrode setup, were two separators are being used.
using measurement points every 50 cm. The density was derived from the The results of the galvanostatic EIS measurement are shown at
electrode thickness and loading. 50 % SOC. In Fig. 11b, overall, a higher impedance of the anode can
be seen in comparison to the cathode. Therefore, the spectrum of the
crystal-like structure. The EDX analysis revealed that aluminum and full cell is dominated by the anode (in accordance with15). This
oxygen make up the extra layer which lead to the conclusion that an kinetical limitation is either due to material properties, degradation
Al2O3 was used. Given that it partially covers a portion of the or interdependencies with the dissassembly and washing procedure.

Figure 10. SEM images of the cathode from the (a) top and (b) side view, and the anode from the (c) top and (d) side view. The one-sided delamination of the
coating visible in (d) was presumably caused by the punching process. Additional SEM images can be found in the supplementary material.
 Journal of The Electrochemical Society, 2023 170 120536

Figure 11. Measurement data of the T-cells. (a) pOCV curves of full cell,
Figure 12. Electrical properties of the measured cells. (a) Measured capacity
anode, and cathode at C/50 as well as charge and discharge curves of full cell
values at C/3 (discharge) and C/20 (charge and discharge) with cell condition
potential in the constrained voltage range of 2.9 V to 4.2 V; (b) Half cell and
specified from Fig. 1; (b) Impedance spectra at a constant 25°C ± 0.2°C at
full cell impedance spectra of a T-cell setup measured at 50 % SOC. The
80 % SOC, 50 % SOC and 20 % SOC of cells with ID 131/828, 186/828 and
frequencies were sampled identically for all measurements.
549/828. The frequencies were sampled identically for all measurements.

In addition, the charge transfer resistance RCT and the contact Capacity and impedance.—The capacities of the individually
resistance RCont. between the anode and the current collector can be measured full cells at different C-rates are shown in Fig. 12a. A
extracted according to the model from Landesfeind et al.46 In the mean value of 22.078 Ah with a standard deviation of
anode EIS measurement, two semi-circles representing these over- σ = 186.2 mAh is obtained at a C-rate of C/3 in the discharge
lapping resistances can be seen. By fitting two semi-circles to the direction among the cells considered to be OK (Fig. 1). At C/20,
anode spectrum, the according values can be defined. RCont. is mean capacities of 22.411 Ah (σ = 199.9 mAh) in the charging
exceptionally high at 109 Ω, which could be due to the existing direction and 22.311 Ah (σ = 199.7 mAh) in the discharging direction
primer in the anode or due to the fact that PTFE is not stable at this are obtained among the measured cells labeled as OK. The depicted
anode potential and therefore the anode lost mechanical and NOK cells vary more in all capacity measurements (standard deviations
electrical integrity during handling.47,48 This can also lead to higher of 236.5 mAh at C/3, 245.2 mAh at C/20 discharge and 223.1 mAh at
overpotentials, as visible in Fig. 11a. In addition, RCT is remarkably C/20 charge), but do not show pronounced abnormalities in the overall
high with a value of 32.2 Ω, which can be explained through the values.
aging of the cell prior to teardown.49 Despite the utmost care, the Based on the volumetric cell data of the cell with ID 083/828 (as
washing and delamination processes cannot be excluded from shown above), a mean cell voltage of 3.7 V, the capacity value from
influencing the T-cell measurements. Due to the high resistance the C/3 discharge test and a measured weight of 355.9 g (residual
values and the complex washing process, the transferability of the foam on the cell was removed), energy densities of 622.4 Wh/L and
results of the impedance measurements from T-cells to full cells is 232.5 Wh/kg are calculated. The comparatively low energy density
only possible to a limited extent. One approach to determine the values compared to cylindrical cells of smaller formats optimized
half-cell impedance with as little intervention as possible would be over several years (see8,38), indicate a conservative design of this
to insert a reference electrode without prior disassembly.50 first-generation 4680 cells, still awaiting optimizations in e.g.
electrode chemistry, cell architecture and package structure. A
Electrical cell-level characterization.—The results of the elec- calculated dead volume of 16.7 cm3 (approx. 13.3 %; comprising
trically characterized cells are presented below, categorized by the the cell-internal contacting) of the internal volume, the measured
measurement methodology. wall thickness and the conventional cell chemistry (as shown above)
 Journal of The Electrochemical Society, 2023 170 120536

Figure 14. Pulse resistances (evaluated in 10 s pulses) of one cell (ID 186/828)
at different C-rates and SOCs in charge and discharge direction.

anode side the closely spaced local maxima in the low SOC area are
distinct for graphite anodes, respectively resembling LiC24 in the
liquid-like (L) lithiation stages 4L, 3L and 2L, which are assumed to
have no in-plane order.56 Hereby, the number represents empty
graphene layers, between two full layers. The peak at around 14 Ah
depicts a graphite stage 2 peak with a phase transition from LiC12 to
LiC6 composition and in-plane order.56–60 There is no evidence for
the characteristic silicon peak in the low SOC area61 in accordance
with the findings from the anode half cell potential (see Fig. 11). For
the cathode, the local minima in the low SOC area presumably
depicts the H1-phase, the saddle point (at around 12 Ah) the
monolithic M-phase and the local maximum at around 18 Ah the
H2-phase of a NMC811 cell chemistry.60,62,63
Figure 13. pOCV, DVA and ICA investigations of one cell (ID 536/828). The ICA is defined as the inverse of the DVA and is plotted
(a) C/20 pOCV measurement in charging direction; (b) Derived DVA against the voltage. The same filtering configuration was applied to
normalized by the nominal capacity; (c) ICA. the ICA as for the DVA. Following Jung et al.,63 the phase
transitions can be interpreted as follows: Peak ① defines the
underline this assumption. However, note that the cells are used in a transition in the anode from C6 → LiCx, peak ② the transition
so-called “structural battery” and could therefore be a structural from H1 → M and peak ③ from M → H2. Strong evidence for an
component in the battery system, requiring a more robust design assumed NMC811 in the cathode can be given due to peak ④, which
compared to conventional cylindrical cells. In addition, the tabless only occurs in this cell chemistry, showing the phase-transitions
design is apparently not trimmed for energy density, but rather for from H2 → H3.63
power capability. In summary, the peaks in the DVA and ICA of the cell propose a
The impedance spectra of three cells (ID 131/828, 186/828, NMC811 chemistry at the cathode and pure graphite without any
549/828) are shown superimposed in Fig. 12b at three different SOC silicone in the anode. This is in accordance with the material
levels. The impedance trend of all three investigated cells with characterization shown above.
decreasing SOC is in good agreement with published
literature51–54—with a shift in the positive Re(Z) direction with HPPC.—The resulting resistances of one cell (ID 186/828) as
decreasing SOC and an increase of the dominant semi-circle calculated according to Ohm’s law are shown in Fig. 14. Similarly to
especially at 20 % SOC. The shift of the Nyquist plot in the positive other battery cells, a characteristic SOC dependency can be
Re(Z) direction of the cell with ID 131/828—equivalent to a higher observed, showing increases in the low SOC region, as well as an
internal resistance of the cell—might be the result of internal SOC region between 50 % and 70 %, which coincides with the
damages impeding electron and ion transportation.55 The visible graphite stage transitions as seen in Fig. 13.64,65 Only a small
dent in the upper part of the casing (therefore classified as NOK in dependency of the resistance on the C-rate can be observed, leading
Fig. 1) reinforces this hypothesis. The calculated parameters of a to decreasing values for higher currents in charge and discharge
simple equivalent circuit model (ECM) (R0, R1, C1) fitted to the three direction. This matches previous observations, except for the low
measured cells behavior are provided in the supplementary material. SOC region in discharge direction, which shows increasing resis-
tances for higher C-rates.64 While previous literature has seldom
Pseudo OCV, DVA and ICA.—The C/20 pOCV curve of one cell examined the impact of current rate on battery impedance, the
(ID 536/828) in the charging direction is presented in Fig. 13a. The relationship is most likely driven by the nonlinear behavior of the
cell could be charged with 22.65 Ah before the upper cutoff voltage charge transfer reaction.66 Previous cell teardown analyses indicate
condition of 4.2 V was met. No constant voltage (CV) phase was this effect being especially pronounced in low SOC regions.67
applied.
The calculated DVA is presented in Fig. 13b, with the corre- Thermal cell-level investigation.—The measurement points ex-
sponding ICA plotted in Fig. 13c. By investigating the half-cell data tracted from the thermal imaging data used to evaluate the
shown above, the inheritance of most of the peaks can unambigu- temperature spread at the cell can during the charging process are
ously be assigned to either the cathode or anode. This enables more depicted in Fig. 15, while Fig. 16a shows the applied charging and
conclusive insights of the underlying electrode compositions. On the discharging profiles.
 Journal of The Electrochemical Society, 2023 170 120536

From the analysis of the overvoltages, the irreversible losses

during 2C charging can be calculated up to 15 W (low SOC region)
as a function of SOC and cell temperature. As can be seen in
Fig. 16b, this leads to an average surface temperature of the cell (ID
536/828) of ∼70 °C for 2C charging at a climate chamber
temperature of 25 °C. In comparison with data from literature, the
Samsung INR21700-50E or LG INR18650-MJ1 reach surface tem-
peratures of only around 42 °C during a 2C discharge in a climate
chamber at 22 °C, while the latter reaches an inside temperature of
57.6 °C.8 This clearly shows the challenge of cooling large-format
cylindrical cells due to the lower surface-to-volume ratio.
As shown in Fig. 16c, a temperature gradient of about 10 K
between Tcap and Tbottom is observed in axial direction at a charge rate
of 2C. This could be caused by the one-sided electrical contacting of
both the anode and the cathode on the cap. Wassiliadis et al.68 showed
that the electrical contact resistance of the cell holder can generate a
significant heat input, particularly at high currents. For example, an
assumed contact resistance of 0.5 mΩ per terminal at a current of 44 A
would result in an ohmic power loss of approx. 2 W. These contact
resistances may be lower in the vehicle depending on the selected
welding process, but due to the one-sided contacting they can provide
a significant heat input and thus lead to an inhomogeneous current
Figure 15. Extraction of the temperatures at the cap, center and bottom of distribution in the cell’s axial direction.
the cell can from the thermal imaging data.
Conclusions
This characterization study comprehensively examines the first-
generation Tesla 4680 cylindrical lithium-ion cells from a state-of-
the-art Tesla Model Y both at the cell level using electrical and
thermal investigations and at the component level by means of a cell
teardown to address the dearth of transparency in automotive
lithium-ion cell development and production. Several cells were
electrically characterized using pOCV, EIS and HPPC methods. One
of the cells was studied with respect to its temperature development
during charging rates of up to 2C. Additionally, one cell was
subjected to a complex cell teardown involving the determination of
the material composition using SEM and EDX, the analysis of the
cell architecture and the subsequent evaluation of three-electrode
cells. The key findings in the respective domains can be summarized
as follows:
Cell teardown: The anode and cathode disks of hexagonal
symmetry of the tabless design can act like compensating elements
or springs when connecting the jelly roll with the housing. The
thickness of the casing is high in comparison to conventional
cylindrical cells, which are not used as a structural component.
The electrode winding consists of a double-sided coated cathode and
anode as well as two separators, while no mandrel is present in
the cell core. The electrode loadings remain almost constant across
the electrodes, but the electrode thickness varies. Spherical active
material particles were found for the cathode side, while the anode
consists of flake graphite particles. EDX confirms the exclusive use
of graphite as the active material, with no detectable presence of
silicon. The utilization of PTFE could indicate a solvent-free anode
fabrication method, while the primer between the coating and the
copper collector coating strongly suggests the utilization of a
solvent-free coating process. The three-electrode analysis reveals
characteristic potentials for the identified cell chemistry and a full
cell impedance spectra dominated by the anode.
Characterization at cell level: Using data of one cell, energy
densities of 622.4 Wh/L and 232.5 Wh/kg are calculated, indicating
a conservative design of this first-generation 4680 cells. The pOCV
analysis (DVA and ICA) confirm a NMC811 chemistry and a pure
graphite anode. HPPC measurements reveal a characteristic SOC
dependency with resistance increases both in the low and mid SOC
Figure 16. Heat generation of the cell (ID 536/828) during 1C and 2C CC
charges at a constant ambient temperature of 25 °C. (a) Current profile; (b)
region. A relatively high surface temperature is observed during a
Absolute temperatures at the cap, center and bottom of the cell can; (c) 2C charge in a free convection setup—requiring an adequate cooling
Temperature gradient between Tcap and Tbottom. Temperature values during system in the battery pack assembly.
discharges are hidden. The data was filtered with moving-average filter, Overall, this characterization study highlights current develop-
unfiltered data is displayed in gray in the background. ments regarding large-format cylindrical lithium-ion cells for
 Journal of The Electrochemical Society, 2023 170 120536

electric vehicles and provides a fundamental basis for future Table A·II. (Continued).
industrial and academic research toward the optimization of
lithium-ion battery performance. # Command Parameter Skip crit. Registration

Acknowledgments 19 Pause — t > 300 s t = 20 ms

20 Discharge CC @ C/3 Q > Qpulse t = 10 s
This research received funding by the German Federal Ministry 21 Pause — t > 3600 s t = 10 s
of Education and Research (BMBF) within the projects “OptiPro” 22 Set Qpulse = 0
under grant number 03XP0364B as part of the competence cluster 23 Pause — t>5s t = 20 ms
“InZePro” and “TUBE” under grant number 03XP0425. 24 Discharge CC @ 1C t > 10 s t = 20 ms
Furthermore, this work was partly funded by the Bavarian 25 Pause — t > 300 s t = 20 ms
Ministry of Economic Affairs, Regional Development and Energy 26 Charge CC @ C/3 Q > Qpulse t = 10 s
within the project “charge.COM” under the grant number DIK-0262/ 27 Pause — t > 3600 s t = 10 s
02 and by the German Federal Ministry for Economic Affairs and 28 Set Qpulse = 0
Climate Action (BMWK) within the project “ultraBatt” under grant 29 Pause — t>5s t = 20 ms
number 01MV21015D. We would like to thank Jannis Wachter for 30 Charge CC @ 1.5C t > 10 s t = 20 ms
support with the measurements at the Institute for Machine Tools 31 Pause — t > 300 s t = 20 ms
and Industrial Management. Furthermore, we would like to thank 32 Discharge CC @ C/3 Q > Qpulse t = 10 s
Florian Biechl and the staff of the electric lab of the Institute of 33 Pause — t > 3600 s t = 10 s
Automotive Technology for their support. 34 Set Qpulse = 0
35 Pause — t>5s t = 20 ms
Appendix: Test Procedures 36 Discharge CC @ 1.5C t > 10 s t = 20 ms
The test protocols for the T-cells, the electrical characterization 37 Pause — t > 300 s t = 20 ms
and the thermal investigation are provided in Tables A·I to A·V. 38 Charge CC @ C/3 Q > Qpulse t = 10 s
39 Pause — t > 3600 s t = 10 s
40 Set Qpulse = 0
41 Charge CC @ C/3∩ Q > 0.1C t = 30 s∪
Table A·I. Activation and test protocol of the T-cells. 42 Pause — t > 3600 s
43 Loop to # 4 9 times
# Command Parameter Skip crit. Registration
C-rates are based on a nominal capacity of 22 Ah ∪ and U = 5 mV
1 Pause — t > 1800 s t = 10 s ∩
and CV @ 4.2 V with I < C/20 if U = 4.2 V reached
2 Charge CC @ C/10 U > 4.2 V t = 10 s
3 Discharge CC @ C/10 U < 2.9 V t = 10 s
4 Loop to # 2 2 times Table A·III. Capacity determination test protocol.
5 Charge CC @ C/50 U > 4.2 V t = 32 s
6 Discharge CC @ C/50 U < 2.9 V t = 32 s # Command Parameter Skip crit. Registration
7 Charge CC @ C/10 t>5h t = 16 s
8 EIS 100 kHz–100 mHz, 10 mV excitation ∪ Start of protocol in fully charged state of the cell
9 Charge CC @ C/10 U > 4.2 V t = 16 s 1 Pause — t > 1800 s∩ t=1s
10 Discharge CC @ C/10 U < 2.9 V t = 16 s 2 Discharge CC @ C/20 U < 2.5 V t = 1 s∪
3 Pause — t > 1800 s∩ t=1s
C-rates are based on a specific capacity of 4.9 mAh/cm2 4 Charge CC @ C/20 U > 4.2 V t = 1 s∪
∪
galvanostatic EIS, 6 ppd, 3 amf 5 Pause — t > 1800 s∩ t=1s
6 Discharge CC @ C/3 U < 2.5 V t = 1 s∪
7 Discharge CV @ 2.5 V I < C/50 t = 1 s∪
Table A·II. Hybrid pulse power characterization test protocol. 8 Pause — t > 600 s∩ t=1s

# Command Parameter Skip crit. Registration C-rates are based on individually determined C/3 disch. cap.
∩
or alternatively ∣dU/dt∣ < 0.2 mVmin−1
∪
Start of protocol in discharged state of the cell and U = 2 mV
1 Pause — t > 3600 s t = 10 s
2 Charge CC @ C/3 Q > 0.1C t = 30 s∪ Table A·IV. Electrochemical impedance spectroscopy test protocol.
3 Pause — t > 3600 s t = 10 s
4 Set Qpulse = 0 # Command Parameter Skip crit. Registration
5 Pause — t>5s t = 20 ms
6 Charge CC @ C/2 t > 10 s t = 20 ms Start of protocol in fully charged state of the cell
7 Pause — t > 300 s t = 20 ms 1 Pause — t > 600 s t=1s
8 Discharge CC @ C/3 Q > Qpulse t = 10 s 2 Discharge CC @ C/3 Q > 0.2C t=1s
9 Pause — t > 3600 s t = 10 s 3 Pause — t > 3600 s t=1s
10 Set Qpulse = 0 4 EIS 10 kHz–10 mHz, 500 mA excitation ∪
11 Pause — t>5s t = 20 ms 5 Pause — t > 600 s t=1s
12 Discharge CC @ C/2 t > 10 s t = 20 ms 6 Charge CC @ C/3 U > 4.2 V t=1s
13 Pause — t > 300 s t = 20 ms 7 Charge CV @ 4.2 V I < C/20 t=1s
14 Charge CC @ C/3 Q > Qpulse t = 10 s 8 Pause — t > 600 s t=1s
15 Pause — t > 3600 s t = 10 s 9 Discharge CC @ C/3 Q > 0.5C t=1s
16 Set Qpulse = 0 10 Pause — t > 3600 s t=1s
17 Pause — t>5s t = 20 ms 11 EIS 10 kHz–10 mHz, 500 mA excitation ∪
18 Charge CC @ 1C t > 10 s t = 20 ms 12 Pause — t > 600 s t=1s
 Journal of The Electrochemical Society, 2023 170 120536

Table A·IV. (Continued). Supplementary Material

Data Availability. We want to give any researcher access to our
# Command Parameter Skip crit. Registration results without any limits. The raw measurement data of the cell
characterization of the individual experiments, cell teardown images
13 Charge CC @ C/3 U > 4.2 V t=1s as well as additional SEM/EDX recordings and processing scripts
14 Charge CV @ 4.2 V I < C/20 t=1s are therefore provided as open source alongside the article at https://
15 Pause — t > 600 s t=1s doi.org/10.14459/2023mp1725661, provided by mediaTUM.
16 Discharge CC @ C/3 Q > 0.8C t=1s
17 Pause — t > 3600 s t=1s ORCID
18 EIS 10 kHz–10 mHz, 500 mA excitation ∪
Manuel Ank https://orcid.org/0000-0002-2060-9521
C-rates are based on a nominal capacity of 22 Ah Alessandro Sommer https://orcid.org/0000-0002-2390-1528
∪
galvanostatic EIS, 10 ppd, 5 amf, drift correction Kareem Abo Gamra https://orcid.org/0000-0001-9739-8710
∩
or alternatively ∣dU/dt∣ < 0.2 mV min−1 Jan Schöberl https://orcid.org/0009-0004-7601-0325
Matthias Leeb https://orcid.org/0000-0003-1156-9078
Table A·V. Test protocol for 1C (α = 1) and 2C (α = 2) charging
Johannes Schachtl https://orcid.org/0009-0007-9108-5327
investigations. Noah Streidel https://orcid.org/0009-0004-2515-2059
Sandro Stock https://orcid.org/0000-0003-2107-6031
# Command Parameter Skip crit. Registration Markus Schreiber https://orcid.org/0000-0002-6530-0077
Philip Bilfinger https://orcid.org/0000-0002-9100-9689
Start of protocol in discharged state of the cell Christian Allgäuer https://orcid.org/0009-0007-8037-895X
1 Pause — t>3h t = 10 s Philipp Rosner https://orcid.org/0000-0001-6481-532X
2 Charge CC @ αC U > 4.2 V t = 30 s∪ Jan Hagemeister https://orcid.org/0000-0001-8307-6614
3 Pause — t > 1800 s t = 10 s∩ Matti Rößle https://orcid.org/0009-0007-5407-6798
4 Discharge CC @ C/3 U < 2.5 V t = 30 s∪ Rüdiger Daub https://orcid.org/0000-0002-5120-4228
5 Discharge CV @ 2.5 V I < C/20 t = 30 s∪ Markus Lienkamp https://orcid.org/0000-0002-9263-5323
6 Pause — t > 3600 s t = 10 s
References
C-rates are based on a nominal capacity of 22 Ah 1. A. Yoshino, “The birth of the lithium-ion battery.” Angew. Chem. Int. Ed., 51, 5798
∪
and U = 5 mV; (2012).
∩
and U = 2 mV 2. M. Li, J. Lu, Z. Chen, and K. Amine, “30 years of lithium-ion batteries.” Adv.
Mater., 30, 1800561 (2018).
3. IEA—International Energy Agency,Global EV outlook 2023: Catching up with
climate ambitions, (2023). https://iea.org/reports/global-ev-outlook-2023.
Authors’ Contributions 4. S. Link, C. Neef, and T. Wicke, “Trends in automotive battery cell design: a
Manuel Ank: Conceptualization, Methodology, Formal analysis, statistical analysis of empirical data.” Batteries, 9, 261 (2023).
Investigation, Software, Validation, Data curation, Writing—ori- 5. E. Musk, D. Baglino, Presentation deck (2020). https://tesla-share.thron.com/
content/?id=96ea71cf-8fda-4648-a62c-753af436c3b6&pkey=S1dbei4.
ginal draft, Writing—review & editing, Visualization; Alessandro 6. J. Frazelle, “Battery day: A closer look at the technology that makes portable
Sommer: Methodology, Formal analysis, Investigation, Validation, electronics possible.” Queue, 18, 5 (2020).
Data curation, Writing—original draft, Writing—review & editing, 7. T. Yang, D. Luo, Y. Liu, A. Yu, and Z. Chen, “Anode-free sodium metal batteries
Visualization; Kareem Abo Gamra: Methodology, Formal analysis, as rising stars for lithium-ion alternatives.” iScience, 26, 105982 (2023).
8. S. Baazouzi, N. Feistel, J. Wanner, I. Landwehr, A. Fill, and K. P. Birke, “Design,
Investigation, Writing—original draft , Writing—review & editing; properties, and manufacturing of cylindrical li-ion battery cells—a generic over-
Jan Schöberl: Methodology, Formal analysis, Investigation, view.” Batteries, 9, 309 (2023).
Writing—original draft , Writing—review & editing; Matthias 9. BMW Group, Press release (2022). https://press.bmwgroup.com/deutschland/ar-
Leeb: Methodology, Formal analysis, Investigation, Writing—ori- ticle/detail/T0403470DE/mehr-leistung-co2-reduzierte-produktion-kosten-deutlich-
reduziert-die-bmw-group-setzt-in-der-neuen.
ginal draft, Writing—review & editing, Visualization; Johannes 10. A. Kwade, W. Haselrieder, R. Leithoff, A. Modlinger, F. Dietrich, and K. Droeder,
Schachtl: Methodology, Formal analysis, Investigation, Writing— “Current status and challenges for automotive battery production technologies.”
original draft, Writing—review & editing, Visualization; Noah Nat. Energy, 3, 290 (2018).
Streidel: Methodology, Formal analysis, Investigation, Writing— 11. F. J. Günter and N. Wassiliadis, “State of the art of lithium-ion pouch cells in
automotive applications: Cell teardown and characterization.” J. Electrochem. Soc.,
original draft, Writing—review & editing, Visualization; Sandro 169, 030515 (2022).
Stock: Methodology, Formal analysis, Investigation, Writing— 12. M. Ank, T. Kröger, M. Schreiber, and M. Lienkamp, “Experimental analysis of
original draft, Writing—review & editing, Visualization;Markus lithium-ion cell procurement: quality differences, correlations, and importance of
Schreiber: Formal analysis, Investigation, Writing—original draft , cell characterization.” Journal of Energy Storage, 66, 107430 (2023).
13. R. Schmuch, R. Wagner, G. Hörpel, T. Placke, and M. Winter, “Performance and
Writing—review & editing; Philip Bilfinger: Methodology, Formal cost of materials for lithium-based rechargeable automotive batteries.” Nat. Energy,
analysis, Investigation, Writing—original draft , Writing—review & 3, 267 (2018).
editing; Christian Allgäuer: Investigation, Writing—original draft , 14. Y. Ding, Z. P. Cano, A. Yu, J. Lu, and Z. Chen, “Automotive li-ion batteries:
Writing—review & editing; Philipp Rosner: Validation, Writing— Current status and future perspectives.” Electrochemical Energy Reviews, 2, 1
(2019).
review & editing; Jan Hagemeister: Methodology, Investigation, 15. S. Stock et al., “Cell teardown and characterization of an automotive prismatic LFP
Writing—review & editing, Visualization; Matti Röle: battery.” Electrochimica Acta, 471, 143341 (2023).
Methodology, Investigation, Software, Validation, Writing—review 16. M. Keppeler, S. Roessler, and W. Braunwarth, “Production research as key factor
& editing; Rüdiger Daub: Supervision, Resources, Writing—re- for successful establishment of battery production on the example of large-scale
automotive cells containing nickel-rich LiNi0.8Mn0.1Co0.1O2 electrodes.” Energy
view & editing; Markus Lienkamp: Supervision, Resources, Technology, 8, 2000183 (2020).
Writing—review & editing. 17. M. Keppeler, H.-Y. Tran, and W. Braunwarth, “The role of pilot lines in bridging
the gap between fundamental research and industrial production for lithium-ion
Declaration of Competing Interest battery cells relevant to sustainable electromobility: A review.” Energy Technology,
The authors declare that they have no known competing financial 9, 2100132 (2021).
18. D. Schreiner et al., “Comparative evaluation of lmr-ncm and nca cathode active
interests or personal relationships that could have appeared to materials in multilayer lithium-ion pouch cells:I. production, electrode character-
influence the work reported in this paper. ization, and formation.” J. Electrochem. Soc., 168, 030507 (2021).
 Journal of The Electrochemical Society, 2023 170 120536

19. J. E. Harlow et al., “A wide range of testing results on an excellent lithium-ion cell 44. Y. Zhang, F. Huld, S. Lu, C. Jektvik, F. Lou, and Z. Yu, “Revisiting polytetra-
chemistry to be used as benchmarks for new battery technologies.” J. Electrochem. fluorethylene binder for solvent-free lithium-ion battery anode fabrication.”
Soc., 166, A3031 (2019). Batteries, 8, 57 (2022).
20. L. Kraft, T. Zünd, D. Schreiner, R. Wilhelm, F. J. Günter, G. Reinhart, 45. N. Nitta, F. Wu, J. T. Lee, and G. Yushin, “Li-ion battery materials: present and
H. A. Gasteiger, and A. Jossen, “Comparative evaluation of lmr-ncm and nca future.” Mater. Today, 18, 252 (2015).
cathode active materials in multilayer lithium-ion pouch cells: Part II. rate 46. J. Landesfeind, D. Pritzl, and H. A. Gasteiger, “An analysis protocol for three-
capability, long-term stability, and thermal behavior.” J. Electrochem. Soc., 168, electrode li-ion battery impedance spectra: Part I. analysis of a high-voltage positive
020537 (2021). electrode.” J. Electrochem. Soc., 164, A1773 (2017).
21. A. Frank, J. Sturm, M. Steinhardt, A. Rheinfeld, and A. Jossen, “Impact of current 47. Q. Wu, J. P. Zheng, M. Hendrickson, and E. J. Plichta, “Dry process for fabricating
collector design and cooling topology on fast charging of cylindrical lithium-ion low cost and high performance electrode for energy storage devices.” MRS Adv., 4,
batteries.” ECS Advances, 1, 040502 (2022). 857 (2019).
22. A. Meyer, Y. Sterzl, and W. Pfleging, “High repetition ultrafast laser ablation of 48. R. Tao et al., “High-throughput and high-performance lithium-ion batteries via dry
graphite and silicon/graphite composite electrodes for lithium-ion batteries.” processing.” Chemical Engineering Journal, 471, 144300 (2023).
Journal of Laser Applications, 35, 042036 (2023). 49. R. Morasch, J. Keilhofer, H. A. Gasteiger, and B. Suthar, “Methods—understanding
23. T. G. Tranter, R. Timms, P. R. Shearing, and D. J. L. Brett, “Communication— porous electrode impedance and the implications for the impedance analysis of li-
prediction of thermal issues for larger format 4680 cylindrical cells and their ion battery electrodes.” J. Electrochem. Soc., 168, 080519 (2021).
mitigation with enhanced current collection.” J. Electrochem. Soc., 167, 160544 50. L. Somerville, S. Ferrari, M. J. Lain, A. McGordon, P. Jennings, and R. Bhagat,
(2020). “An in situ reference electrode insertion method for commercial 18 650-type cells.”
24. H. Pegel, D. Wycisk, and D. U. Sauer, “Influence of cell dimensions and housing Batteries, 4, 18 (2018), An in situ reference electrode insertion method for
material on the energy density and fast-charging performance of tabless cylindrical commercial 18650-type cells.
lithium-ion cells.” Energy Storage Materials, 60, 102796 (2023). 51. N. Meddings et al., “Application of electrochemical impedance spectroscopy to
25. S. Li, M. W. Marzook, C. Zhang, G. J. Offer, and M. Marinescu, “How to enable commercial li-ion cells: A review.” Journal of Power Sources, 480, 228742 (2020).
large format 4680 cylindrical lithium-ion batteries.” Applied Energy, 349, 121548 52. T. Stanciu, D.-I. Stroe, R. Teodorescu, and M. Swierczynski, “Extensive eis
(2023). characterization of commercially available lithium polymer battery cell for
26. Munro Live, Tesla’s 4680 Current Collector Layout (2022). https://youtu.be/ performance modelling.” 2015 17th European Conference on Power Electronics
UhAkjXZzjCk. and Applications (EPE’15 ECCE-Europe) 1–10 (2015).
27. Munro Live, 4680 battery pack: What we found under the foam! (2022) https:// 53. O. Alao and P. Barendse, “Characterization of lithium ncm and sealed lead acid
youtu.be/ozesI3OZEG0. batteries using electrochemical impedance spectroscopy.” 2018 IEEE PES/IAS
28. J. Hagemeister, S. Stock, M. Linke, M. Fischer, R. Drees, M. Kurrat, and R. Daub, PowerAfrica, 1, 166 (2018).
“Lean cell finalization in lithium-ion battery production: determining the required 54. L. Wildfeuer, P. Gieler, and A. Karger, “Combining the distribution of relaxation
electrolyte wetting degree to begin the formation.” Energy Technology, 11, times from eis and time-domain data for parameterizing equivalent circuit models
2200686 (2023). of lithium-ion batteries.” Batteries, 7, 52 (2021).
29. S. Solchenbach, D. Pritzl, E. J. Y. Kong, J. Landesfeind, and H. A. Gasteiger, “A 55. F. Xiao, B. Xing, L. Kong, and Y. Xia, “Impedance-based diagnosis of internal
gold micro-reference electrode for impedance and potential measurements in mechanical damage for large-format lithium-ion batteries.” Energy, 230, 120855 (2021).
lithium ion batteries.” J. Electrochem. Soc., 163, A2265 (2016). 56. S. Schweidler, L. de Biasi, A. Schiele, P. Hartmann, T. Brezesinski, and J. Janek,
30. L. Somerville, J. Bareño, P. Jennings, A. McGordon, C. Lyness, and I. Bloom, “The “Volume changes of graphite anodes revisited: A combined operando X-ray
effect of pre-analysis washing on the surface film of graphite electrodes.” diffraction and in situ pressure analysis study.” J. Phys. Chem. C, 122, 8829 (2018).
Electrochimica Acta, 206, 70 (2016). 57. M. Winter, J. O. Besenhard, M. E. Spahr, and P. Novák, “Insertion electrode
31. Batemo GmbH, Batemo Cell Explorer—Tesla Model Y (4680) (2023).https:// materials for rechargeable lithium batteries.” Adv. Mater., 10, 725 (1998).
batemo.com/products/batemo-cell-explorer/unknown-manufacturer-tesla-model-y- 58. M. He? and P. Novák, “Shrinking annuli mechanism and stage-dependent rate
4680/. capability of thin-layer graphite electrodes for lithium-ion batteries.”
32. I. Bloom, A. N. Jansen, D. P. Abraham, J. Knuth, S. A. Jones, V. S. Battaglia, and Electrochimica Acta, 106, 149 (2013).
G. L. Henriksen, “Differential voltage analyses of high-power, lithium-ion cells.” 59. J. Guo, Y. Li, J. Meng, K. Pedersen, L. Gurevich, and D.-I. Stroe, “Understanding
Journal of Power Sources, 139, 295 (2005). the mechanism of capacity increase during early cycling of commercial NMC/
33. I. Bloom, J. Christophersen, and K. Gering, “Differential voltage analyses of high- graphite lithium-ion batteries.” Journal of Energy Chemistry, 74, 34 (2022).
power lithium-ion cells.” Journal of Power Sources, 139, 304 (2005). 60. J. Schmitt, M. Rehm, A. Karger, and A. Jossen, “Capacity and degradation mode
34. I. Bloom, J. P. Christophersen, D. P. Abraham, and K. L. Gering, “Differential estimation for lithium-ion batteries based on partial charging curves at different
voltage analyses of high-power lithium-ion cells.” Journal of Power Sources, 157, current rates.” Journal of Energy Storage, 59, 106517 (2023).
537 (2006). 61. J. Schmitt, M. Schindler, and A. Jossen, “Change in the half-cell open-circuit
35. I. Bloom, L. K. Walker, J. K. Basco, D. P. Abraham, J. P. Christophersen, and potential curves of silicon-graphite and nickel-rich lithium nickel manganese cobalt
C. D. Ho, “Differential voltage analyses of high-power lithium-ion cells. 4. cells oxide during cycle aging.” Journal of Power Sources, 506, 230240 (2021).
containing NMC.” Journal of Power Sources, 195, 877 (2010). 62. O. Dolotko, A. Senyshyn, M. Mühlbauer, K. Nikolowski, and H. Ehrenberg,
36. Y. Li, M. Abdel-Monem, R. Gopalakrishnan, M. Berecibar, E. Nanini-Maury, “Understanding structural changes in NMC li-ion cells by in situ neutron
N. Omar, P. van den Bossche, and J. Van Mierlo, “A quick on-line state of health diffraction.” Journal of Power Sources, 255, 197 (2014).
estimation method for li-ion battery with incremental capacity curves processed by 63. R. Jung, M. Metzger, F. Maglia, C. Stinner, and H. A. Gasteiger, “Oxygen release
gaussian filter.” Journal of Power Sources, 373, 40 (2018). and its effect on the cycling stability of LiNixMnyCozO2 (NMC) cathode materials
37. M. Dubarry and D. Anseán, “Best practices for incremental capacity analysis.” for li-ion batteries.” J. Electrochem. Soc., 164, A1361 (2017).
Frontiers in Energy Research, 10, 1023555 (2022). 64. N. Wassiliadis et al., “Quantifying the state of the art of electric powertrains in
38. T. Waldmann, R.-G. Scurtu, K. Richter, and M. Wohlfahrt-Mehrens, “18650 vs battery electric vehicles: range, efficiency, and lifetime from component to system
21 700 li-ion cells—a direct comparison of electrochemical, thermal, and geome- level of the Volkswagen ID.3.” eTransportation, 12, 100167 (2022).
trical properties.” Journal of Power Sources, 472, 228614 (2020). 65. L. Wildfeuer, N. Wassiliadis, C. Reiter, M. Baumann, and M. Lienkamp,
39. X. Zhang and T. Wierzbicki, “Characterization of plasticity and fracture of shell “Experimental Characterization of Li-Ion Battery Resistance at the Cell, Module
casing of lithium-ion cylindrical battery.” Journal of Power Sources, 280, 47 and Pack Level.” 2019 Fourteenth International Conference on Ecological Vehicles
(2015). and Renewable Energies (EVER), IEEE, Piscataway, NJ, 1, 1–12 (2019).
40. A. Sommer, M. Leeb, S. Haghi, F. J. Günter, and G. Reinhart, “Marking of 66. J. Zhu, M. Knapp, M. S. D. Darma, Q. Fang, X. Wang, H. Dai, X. Wei, and
electrode sheets in the production of lithium-ion cells as an enabler for tracking and H. Ehrenberg, “An improved electro-thermal battery model complemented by
tracing.” Procedia CIRP, 104, 1011 (2021), LIV CIRP CMS 2021—Towards current dependent parameters for vehicular low temperature application.” Applied
Digitalized Manufacturing 4.0. Energy, 248, 149 (2019).
41. D. Das, S. Manna, and S. Puravankara, “Electrolytes, additives and binders for 67. L. Wildfeuer, N. Wassiliadis, A. Karger, F. Bauer, and M. Lienkamp, “Teardown
NMC cathodes in li-ion batteries; a review.” Batteries, 9, 193 (2023). analysis and characterization of a commercial lithium-ion battery for advanced
42. F. Huttner, A. Diener, T. Heckmann, J. C. Eser, T. Abali, J. K. Mayer, P. Scharfer, algorithms in battery electric vehicles.” Journal of Energy Storage, 48, 103909
W. Schabel, and A. Kwade, “Increased moisture uptake of NCM622 cathodes after (2022).
calendering due to particle breakage.” J. Electrochem. Soc., 168, 090539 (2021). 68. N. Wassiliadis, M. Ank, L. Wildfeuer, M. K. Kick, and M. Lienkamp,
43. A. D. Jara, A. Betemariam, G. Woldetinsae, and J. Y. Kim, “Purification, “Experimental investigation of the influence of electrical contact resistance on
application and current market trend of natural graphite: A review.” International lithium-ion battery testing for fast-charge applications.” Applied Energy, 295,
Journal of Mining Science and Technology, 29, 671 (2019). 117064 (2021).