Journal of The Electrochemical

Society

Effect of the Slurry Mixing Process on the Structural Properties of the

Anode and the Resulting Fast-Charging Performance of the Lithium-Ion
Battery Cell
To cite this article: Desiree Grießl et al 2022 J. Electrochem. Soc. 169 020531

View the article online for updates and enhancements.

This content was downloaded from IP address 192.12.14.1 on 19/12/2022 at 21:45

 Journal of The Electrochemical Society, 2022 169 020531
1945-7111/2022/169(2)/020531/13/$40.00 © 2022 The Electrochemical Society (“ECS”). Published on behalf of ECS by IOP Publishing Limited

Effect of the Slurry Mixing Process on the Structural Properties of

the Anode and the Resulting Fast-Charging Performance of the
Lithium-Ion Battery Cell
Desiree Grießl,1,2,z Alexander Adam,1,2,z Korbinian Huber,1,2 and Arno Kwade2,3
1
BMW Group, Battery Cell Competence Centre, 80788 Munich, Germany
2
TU Braunschweig, Institute for Particle Technology, 38104 Braunschweig, Germany
3
Battery LabFactory Braunschweig, TU Braunschweig, 38106 Braunschweig, Germany

Besides the performance of a lithium-ion battery cell, the manufacturing costs are crucial for the success of battery electric
vehicles. To enable cost efficient yet well performing battery cells, an optimization of the manufacturing steps with respect to the
cell properties is necessary. The slurry mixing process, being the initial step of the lithium-ion battery cell manufacturing process,
is well known to affect the structure of the electrode coating (e.g. porosity, tortuosity or the distribution of the binder and
conductive additive), which is further connected to its electrical and ionic resistances. Therefore, a variation of the formulation
strategy or mixing device can affect the performance of the lithium-ion battery cell. In this study, several variations of the slurry
mixing process are investigated with respect to its effect on the fast-charge capability of the lithium-ion battery cell. Properties of
the slurry, the electrode and the resulting lithium-ion battery cell are characterized for each variation to detect interdependencies
and derive process-structure-property relations.
© 2022 The Electrochemical Society (“ECS”). Published on behalf of ECS by IOP Publishing Limited. [DOI: 10.1149/1945-7111/
ac4cdb]

Manuscript submitted November 17, 2021; revised manuscript received January 4, 2022. Published February 10, 2022.

The electrification of vehicles represents one of the most evident by optimized electrode structures, which reduce the effective
trends in the automotive industry and is mainly driven by the diffusion resistance of the lithium-ions in the electrolyte.
European Commission’s demand to reduce the average CO2 con- An optimization of the electrode structure with respect to the
sumption of vehicle fleets.1 Besides the performance of the battery ionic diffusion resistance can either be achieved by a change of the
cell, the costs are decisive for their application. The still high costs material components (e.g. spherical particles allow a better diffusion
of a lithium-ion-battery constitute to about 70% to material costs. than flake-like particles15) and the material composition (e.g.
But the remaining 30% are attributable to the battery production. reduction of the binder content16) or by optimizing the manufac-
Factors such as the number of processing steps and process times turing process of the battery electrode (e.g. change of porosity due to
have to be considered.2 However, a key requirement for the different compression). Regarding the latter, BILLAUD et al.17
customer acceptance of battery electric vehicles is the fast-charge presented a method to fabricate anodes with an out-of-plane aligned
capability of the lithium-ion battery. Various factors can limit the architecture by using a low external magnetic field. An alignment of
fast-charge capability, e.g. maximum charging power of the charging flake-like graphite particles reduces the tortuosity of the anode and
station, the electrical architecture of the vehicle, the cooling system thus, decreases the ionic diffusion resistance. Furthermore, LIU
of the battery cells or the intrinsic fast-charge capability of the et al.18 and CHEN et al.19 presented a method to manufacture
battery cells.3–6 cathodes with multiple coating layers, which improved their
It is well known that the diffusion of lithium-ions in the electrical and ionic properties. This method could also be transferred
electrolyte is the most relevant factor that limits the intrinsic fast- to the anode manufacturing process in order to reduce the ionic
charge capability of the battery cell.7 During fast-charging, a high diffusion resistance. HABEDANK et al.20,21 presented a method to
current is applied to the battery cell, which forces the diffusion of a structure the anode coating with laser ablation. Micrometer thick
high concentration of lithium-ions through the porous structure of channels in vertical direction to the coating are ablated to reduce the
the anode coating. Beside electrolyte related properties like the ionic diffusion resistance. In comparison to the methods mentioned
conductive salt content or the transference number, the ionic before, the laser structuring of anodes leads to a reduced areal
diffusion is impeded by the porous structure of the anode, resulting capacity resulting from the ablation process. Therefore, the question
in an enhanced polarization of the anode. If the anode potential arises, if an increase of the coating thickness to balance the areal
locally drops below 0 V vs Li / Li + due to polarization, lithium is capacity of the laser structured electrode to the pristine electrode
metallically deposited on the anode surface instead of the desired compensates the benefit of the ablated vertical channels.
intercalation of lithium-ions into the graphite host structure. This Besides these methods to consciously modify the structure of the
unwanted parasitic side reaction is referred to as lithium-plating and anode, each step in the electrode manufacturing process potentially
causes a loss of capacity and potential safety issues.8,9 affects the electrical and ionic properties of the anode. The mixing
Two approaches can be distinguished in reducing the charging process as the initial manufacturing step is a very decisive proces-
time of a battery cell: the first approach is based on optimized sing step because it might affect the product properties resulting
charging profiles (e.g. multi-step constant current profiles,10,11 from all further process steps. Additionally, the mixing process can
constant anode potential profiles,3,12 etc.), which are designed to change the cell properties without affecting the cell design, e.g., a
decrease the charge current before reaching a critical polarization of better conductive additive distribution enhances the performance of
the anode. This results in the shortest possible charging time and a battery cell at constant conditions regarding recipe and electrode
exploits the full intrinsic fast-charge capability of a battery cell. The design. The mixing process mainly aims to generate a slurry with
second approach attempts to enhance the intrinsic fast-charge finely dispersed particles. The slurry is supposed to be stable over a
capability of a battery cell by optimizing the diffusion process of certain time regarding aspects like reagglomeration or sedimentation
the lithium-ions in the electrolyte. This can either be implemented by of particles, as well as to be further processable in the subsequent
advanced electrolytes with a higher transference number and thus a coating step.
better ionic conductivity of the lithium-ions in the electrolyte,13,14 or Conventionally during mixing, the binder (B) is dissolved in a
solvent and the active material (AM) and conductive additives, such
as carbon black (CB), are finely dispersed and homogenized. Input
z
E-mail: [email protected]; [email protected] of stress e.g., by a rotating mixing device causing fluid shear or
 Journal of The Electrochemical Society, 2022 169 020531

particle-particle interactions, is decisive for the degree of homo- viscosity η,36–38 which can further be correlated with the tip speed vt
genization and dispersing of the components.22 Especially the of the mixing equipment.
formation of a conductive additive network to provide electric
conductivity throughout the whole electrode is often mentioned to be SI ∝ τ = 2.5 · γ · η [ 1]
crucial.23–27 As already mentioned, subjection of CB to mechanical 39,40
stress is needed for sufficient dispersing. At the same time, According to KREKEL, the SN depends on the angular velocity of
irreversible damage of the AM by excessive shearing needs to be a spherical particle ωparticle , or, respectively, the shear rate, and the
prevented. Hence, pre-dispersing of CB prior to AM addition is often dispersing time t:
applied to yield a so-called premix.28 Moreover, significant effects
of pre structuring dry powders prior to the actual mixing process on ωparticle ·t γ ·t
SN = = [2]
the electrode structure and performance of cathodes has been shown 2·π 4·π
by BOCKHOLT et al.29,30 Stepwise addition of the solvent during the
mixing process and thus a gradually decreasing solids content results In order to take the geometry of the mixing device into considera-
in enhanced CB deagglomeration.31,32 Considerable potential of the tion, the ratio of the highly active mixing volume Vact to the total
mixing process on the electrode structure can be concluded. The volume Vtot of the mixing device has to be considered as well.
mixing process can either be conducted batch-wise in e.g., planetary Besides the slurry characteristics, also the process duration and
mixers or via continuous processing, which was recently presented scalability of the mixing process are highly relevant for a possible
with the aid of a twin-screw extruder.33 industrialization. Nevertheless, scientific literature regarding those
Aiming to describe the complexity of the interactions amongst topics is scarce. Being able to validate whether changes in the mixing
the process chain and resulting (intermediate) product characteris- process (e.g., a scale-up or shortening of the process duration) have an
tics, BOCKHOLT and INDRIKOVA et al.34 present process-structure- influence on the fast-charge capability is of great interest. In a recently
property relations together with applications in a mixing and published study of our group, we focused on the electrode design (layer
calendering study. The importance of such a description is evident thickness and porosity) and presented a process—structure—property
for a holistic understanding of the whole process chain. Figure 1 relation with respect to the electrode compression (resulting from the
shows a schematic description of a process-structure-property calendering process).41 It is based on an extensive electrode character-
relation for battery cell manufacturing. Each intermediate or ization including the quantification of the effective electrode tortuosity
the final product (e.g., slurry, electrode structure or cell) in the by means of electrochemical impedance spectroscopy as described by
manufacturing of battery cells is affected by the choice of the LANDESFEIND et al.42,43 The tortuosity is a central criterion to understand
parameter setting of the process step (e.g., mixing time, drying effects of process parameters on the electrode structure. Additionally,
temperature during coating or line load during calendering) and the the recently developed DCV-method by ADAM et al.44 is used to
properties of the previous intermediate product (e.g., slurry viscosity quantify the fast-charge capability of a cell, which finally enables the
or electrode porosity and tortuosity). The target of this study is a determination of relations between electrode and cell properties. The
correlation between process parameters, structural characteristics method is based on the occurrence of a characteristic feature in the
and the final product properties (in this case the properties of the voltage signal during charging, which is correlated with the onset of
cell). This is relevant in order to understand the process—structure— lithium-plating.
property relation of the battery cell manufacturing, which is an Based on these findings and the developed methods, this study
essential element for further battery cell optimization (e.g., fast- focusses on the effect of the mixing process in the manufacturing of
charging). lithium-ion battery cells on structural properties of the electrode and
The mixing process can mainly be described by the stress the resulting fast-charging properties of the cell. Different formula-
mechanism as well as the two measures stress intensity (SI) and tion strategies considering the mixing process in terms of varying
stress number (SN). Depending on the mixing device and conditions, solids content or predispersing of carbon black are investigated
such as laminar or turbulent flow, the SI depends on the shear stress extensively. Furthermore, the effects of shortening the mixing time
τ.35 According to RUMPF and RAASCH, the SI is given in Eq. 1 for and also the scalability from laboratory scale mixing to pilot size is
laminar flow. It depends on the shear rate γ and the dynamic examined.

Figure 1. Schematic description of a process-structure-property relation for battery cell manufacturing.

 Journal of The Electrochemical Society, 2022 169 020531

Figure 2. Schematic overview of the methodology of this study.

Experimental adding PVDF to N-methyl-2-pyrrolidon (NMP) as solvent and

mixing it with a tip-speed of 2.6 m s−1 until PVDF is fully dissolved.
Methodology.—Figure 2 shows a schematic overview of the
Afterwards, CB was added, and the mixture was stirred for 50 min at
methodology, which is used in this study. The superior focus is on
a tip-speed of 5.2 m s−1 with a toothed disc (50 mm diameter). After
the understanding of the process—structure—property relation in the
the addition of AM to the premix, a butterfly stirrer was mounted
battery cell manufacturing with respect to the mixing process.
instead of the toothed disc and the slurry was mixed at a tip-speed of
Therefore, the intermediate products (slurry and electrode) as well
3.9 m s−1 for 45 min. The resulting suspension was degassed prior to
as the final product (battery cell) are analyzed with different methods
the coating step.
to characterize their structure and properties. A step-by-step analysis
aims to reveal effects and interdependencies between parameters of
the process steps and properties of the intermediate and final Electrode manufacturing.—The slurries were coated on the
products. The whole manufacturing process of the battery cell as current collector foils with a doctor blade film applicator. Copper
well as each analysis method is explained in this chapter. foil (10 μm, Schlenk, Germany) was used on the anode side,
whereas aluminium foil (15 μm, Norsk Hydro, USA) was used for
Slurry preparation.—The anode consists of 96 wt% artificial the cathodes. After the coating at room temperature, the electrodes
graphite as active material and 1 wt% carbon black (CB) as conductive were dried at 60 °C for at least 2 h. The areal mass loading of the
additive. As binder, 1.5 wt% of carboxymethyl cellulose (CMC) and anodes is between 14.87 and 15.29 mg cm−12, whereas the cathode
1.5 wt% of styrene-butadiene rubber (SBR) are used. The cathode mass loading ranges from 27.36 to 28.08 mg cm−1. Both, anode and
contains 94 wt% of active material LiNi0.6 Mn 0.2 Co0.2 O2 (NMC 622), cathode sheets were compressed using a calendering machine (CA9,
4 wt% of CB and 2 wt% polyvinylidenfluorid (PVDF) as binder. All Sumet, Germany). For pouch cell fabrication, porosities of anodes
materials are commercially available products. and cathodes were adjusted to 33 ± 2%.
A planetary mixer in a pilot-plant scale with a nominal volume
capacity of 10 l (PMH 10, Netzsch-Feinmahltechnik GmbH, Germany) Characterization of the slurry and electrode properties.—This
was used to produce anode slurries with different mixing protocols. section describes the methods to characterize the structural proper-
During mixing, the vessel was tempered at 20 °C. Differentiation of a ties of the slurry and electrode.
low-speed motor (cross arm stirrer) and a high-speed motor (either
toothed disc or butterfly tool mounted) is relevant for the mixing Rheology of the slurry.—A rotational rheometer (Anton-Paar,
protocol. Processing of slurries with high viscosity is possible in MCR 102), equipped with cone (1°) and plate (50 mm) geometry,
planetary mixers due to the setup that includes a wall scraper. As was used to conduct rheological measurements. Measurements were
illustrated in Fig. 3, a variation of the total solids content (TSC) during performed directly after sampling at a temperature of 20 °C. For the
the kneading step was reached via delayed addition of a certain part of shear rate tests, viscosities at 21 different shear rates between 0.1 to
the binder solution (respectively premix). The TSC during the kneading 1000 1/s using a logarithmic ramp without pre-shearing were
step is referred to as kneading concentration (KC) in the following. Pre measured with a duration of 5 s for each shear rate.
dispersing of CB in the binder solution prior to AM addition was
conducted as an alternative to simultaneous adding CB and graphite. Particle size distribution in the slurry.—The particle size
Apart from that, the duration of the kneading phase was varied. Within distributions (PSD) of anode slurries were measured by laser
the framework of pre-trials and, most important, evaluation of the diffraction analysis (LA-950, Horiba Cop.) following guidelines
scalability of the mixing process, the same anode recipe was processed described by Dreger et al.43 0.5 g of each sample were mixed with
in a laboratory dissolver (Dispermat CA, VMA-Getzmann). The 0.25 g of a stabilizing additive (Tego Dispers 760 W, Evonik,
dissolver was equipped with a vessel that has a nominal volume Germany) and 12 g of deionised water prior to the measurement in
capacity of 500 ml and a diameter of 80 mm. The vessel was a fraction cell. Volume-based particle size distributions were
temperature-controlled at 20 °C with a jacket cooling during the entire determined for all samples.
mixing procedure. Further details concerning the mixing protocol are
given in Fig. 3. Cross-sectional images of the electrode coating using scanning
A standard cathode is prepared with a dissolver (Dispermat CA, electron microscope.—A scanning electron microscope, (ZEISS
VMA-Getzmann). As a first step, a binder solution was produced by GeminiSEM 450, Zeiss, Germany), equipped with a field emission
 Journal of The Electrochemical Society, 2022 169 020531

Figure 3. Schematic presentation of the different mixing protocols, aiming to investigate the influence of kneading concentration, pre-dispersing of CB, mixing
time and mixer type on the resulting product properties.

gun operated at an accelerating voltage of 2 kV, was used for Herein, Rion is the effective ionic resistance of the electrolyte in
scanning electron microscopy (SEM). An Inlens secondary electron the porous structure of the electrode with porosity ε, A is the
detector was used. For cross-section analysis, anodes were prepared chemical active area of the electrode, κ is the ionic (bulk)
using an ion beam milling system (Ar-Blade 5000, Hitachi, Japan) conductivity of the electrolyte and d is the thickness of the electrode
operating at an accelerating voltage of 8 kV and an Ar gas stream of coating. The ionic resistance is measured with impedance spectro-
0.12 cm3 min−1 for 1 h in a width of 25 mm of the electrode cross scopy, which was performed in a symmetrical cell setup (T-cells;
section. two identical electrodes, no reference electrode) at 25 °C. The
electrolyte contains a 3:7 wt% mixture of ethylene carbonate (EC)
Electrical resistances of the electrode coating.—The electrical and ethyl methyl carbonate (EMC) with 50 mM tetrabutylammonium
resistances of the anodes are measured with a resistance meter perchlorate (TBAClO4, Sigma Aldrich). Tetrabutylammonium-ions
(RM2610, HIOKI, Japan) at room temperature. A constant current is are used as a substitute to lithium-ions because they cannot intercalate
applied to the electrode sheet, the occuring potential distribution on the into the graphite electrode. This results in a purely capacitive behavior
electrode surface is measured via multiple microscopic probes, resulting on the electrode surface (so-called blocking conditions) and enables
in values for the interface resistance and volume composite resistivity.46 the determination of the ionic resistance via impedance measurement.
1.7 ·10−6 Ω· cm is assumed as the resistivity of the copper current The conductivity of the electrolyte is measured to be 1.14 ± 0.01 mS/
collector. cm at 25 °C. The coating thicknesses are measured with a dial gauge
(Digico 505MIE, Tesa Swiss, accuracy of ± 2 μm) and coated masses
Effective tortuosity of the electrode coating.—The determination are determined with a lab balance (Cubis analytical balance, Sartorius,
of the effective tortuosity is performed as described in a previous accuracy of ±0.00001 g). Three cells were measured for each
publication of our group.44 The effective tortuosity is an empirically parameter.
determined value, which describes the ratio of the ionic resistance
resulting from the lithium-ion diffusion in the electrolyte within the Quantification of the fast-charge capability of lithium-ion
porous electrode structure to the lithium-ion diffusion in the pure battery cells.—The DCV-method, described by Adam et al.,44 is
electrolyte. Please note, that there are different terms in literature used to quantify the fast-charge capability of lithium-ion battery
used for these kind of tortuosity (e.g., tortuosity factor, ionic cells. Therefore, a charge rate test is performed to determine the C-
tortuosity, effective tortuosity), which all refer to the same defini- rate necessary to cause lithium-plating and to ascertain the state-of-
tion. Eq. 3 is used to calculate the effective tortuosity τeff : charge at which lithium-plating occurs. The first derivative of the
voltage signal during charging is used to locate a characteristic local
Rion *ε*A*κ maximum, which correlates with the onset of lithium-plating. A
τeff = [3]
2*d change of the local maximum, in our case resulting from a change of
 Journal of The Electrochemical Society, 2022 169 020531

the anode structure, is used to evaluate the fast-charge capability of

the cell. If the local maxima at a specific charge rate shifts to lower
state-of-charges, in our case resulting from a change of the anode
structure, lithium-plating occurs at shorter charging time, i.e., the
fast-charging capability is inferior.

Battery cells.—Custom-made lithium-ion battery cells containing

the respective anodes to study their fast-charge capability and
standard cathodes are manufactured. In the present study, single-
layer pouch cells with a nominal capacity of 116 ± 1 mAh are used
to quantify the fast-charge capability. All cells contain a polyolefin
separator from Celgard with 25 μm thickness. The electrolyte
consists of a mixture of EC and EMC in a gravimetric ratio of 3:7
wt% with 2% vinylene carbonate (VC) and 1 mol/l LiPF6. All pouch
cells are dried in a vacuum drying cabinet for at least 12 h at 60 °C,
before they are filled with the electrolyte. The cell formation cycling
consists of one cycle with a C/10 rate (CC-CV charging, CC
discharging) with a lower and upper voltage limit of 2.8 V and
4.2 V, respectively. The subsequent CV-phase at a voltage of 4.2 V
stops if a current of C/50 is reached. The cells are clamped between
two pressure plates with a screw torque of 1 Nm in a custom-made
cell holder.

Charge rate test.—The fast-charge capability of each battery cell

is investigated with a charge rate test (see also44), which is depicted
in Fig. 4. A capacity check is applied before and after the charge rate
test in order to determine the capacity retention of each cell. The
actual capacity of the cell, which is determined in the first capacity
check, is used to calculate the charge rates of the following charge
rate test. The capacity check consists of three constant current—
constant voltage (CC-CV) cycles. For each cycle, the cells charge
with a rate of C/10 to a voltage of 4.2 V and the subsequent CV
phase has a cut-off current limit of C/50, which is followed by a CC Figure 4. Overview of the charge rate test: (a) the voltage signal of the cell
and (b) the corresponding C-rates. Please note that for clarity a negative C-
discharge phase with a C/10 rate to a voltage of 2.8 V. The charge rate indicates a discharge phase.
throughput of the discharge step in the third cycle is used to define
the individual cell capacity.
The charge rate test consists of ten consecutive cycles. In the different anode slurries during the mixing process, for three
charge phase, the cell is charged with increasing CC charge rates for representative kneading concentration KCs. The specific energy is
every cycle to a voltage of 4.2 V. The following order of charge rates derived from the rotational speed of the mixer, the corresponding
is used: C/10, C/3, 2 C/3, 1 C, 1.5 C, 2 C, 2.5 C, 3 C, 4 C and 5 C. torque, mixing duration and slurry mass.35 For a solids content of
After each charge phase, a subsequent pause of 30 min is carried out 55 wt% during the main mixing step, the highest specific energy is
and then the cell is discharged with a C/2 rate to a voltage of 2.8 V. observed. For the maximum kneading concentration of 65 wt%,
which is considered during this work, the energy input shows a
Results and Discussion lower value, because the lack of solvent results in a partly still
powder-like behavior, also remarkable by the eye, of the sample
Investigating the impact of the slurry mixing process on the during this kneading phase. Nevertheless, a high number of particle-
resulting fast-charge capability of the battery cell requires a deep particle interactions take place and one can expect the highest
understanding of each process step and product (slurry, electrode amount of particle collisions for the highest KC. A total solids
and cell) in order to derive process-structure-property relations. Due content of 50 wt% yields a free-flowing slurry with comparably low
to that reason, slurry and electrode characteristics are described prior viscosity, reducing the applicable shear stress during mixing.
to the characterization on cell level. The following chapter is Slurry properties are described by rheological data in Fig. 5a and
separated according to the variations in the slurry mixing process, particle size distributions in Fig. 5b. Note that the final solids content
such as: of all slurries is within a small range (49.6 ± 0.9%, Table AI in the
Appendix). The three different kneading concentrations result in
• Influence of the solids content during mixing significantly different viscosity functions (viscosity as a function of
• Influence of a pre-mixing step the applied shear rate). At small shear rates of up to 10 1/s, slurries
• Influence of the mixing time with a kneading concentration of 50 wt% (light blue) and 55 wt%
• Influence of the mixing device (medium blue) show a slight shear thickening behavior, whereas a
shear thinning behavior is detected for 65 wt% KC (dark blue). A
shear thinning behavior is typical for suspensions and also well
Influence of the solid content during mixing.—A higher solids known for anode slurries31,32 due to the disentanglement of CMC
concentration at the beginning and stepwise dilution of the slurry is polymer chains as well as an orientation of graphite particles with
expected to enhance particle deagglomeration during the mixing increasing shear rate.47 Still, during the kneading step, CMC chains
process.31 The total solids content of a slurry strongly accounts for are assumed to adsorb on the graphite surface. That’s why shear
the viscosity and, therefore, the applied shear stress (resembling the thinning behavior is less pronounced in the slurries with high amount
stress intensity) in the mixer, which can be further correlated to the of CMC in the kneading phase. Absence of the so-called “transient
stress intensity. At the same time, higher solids concentrations lead viscosity hill”47 has been proven by further measurements for those
to a higher stress frequency due to a higher number of particle- samples. For the higher kneading concentrations, 46% of CMC
particle interactions.35 Table I shows the specific energy, applied to solution for KC 65 and 18% for KC 55 are added after the kneading
 Journal of The Electrochemical Society, 2022 169 020531

Table I. Specific energy input during the mixing process and relative slurry with a low kneading concentration of 50 wt% (light blue) does
span factor RSF (see Eq. 4) of the particle size distribution. not show a separate aggregate peak at all. In this case, the applied
stress is not sufficient to deagglomerate CB agglomerates, which are
KC [wt%] Specific energy [Wh/kg] RSF [-] kept together by weak van der Waals forces and electrostatic
interactions.48 CB agglomerates in that case simply overlap with
65 11.5 1.16 the graphite signal. The higher the kneading concentration, the better
55 23.3 1.00 the deagglomeration of CB and the more intense is the signal for
50 11.5 0.95 aggregates. Concomitant with the deagglomeration of CB, fragmen-
tation of graphite appears for high shear stresses during the kneading
step. This can be seen by the broadening of the graphite signal,
which forms of a shoulder towards smaller particle sizes (down to
step. The variant with 50 wt% KC entails 100% of CMC solution ∼1.3 μm). The relative span factor RSF is an indicator for the
during the whole mixing process. Thus, distinct shear thickening uniformity or, respectively, polydispersity of a particulate system.
behavior at lower shear rates is most likely visible, due to a high The particle sizes for certain cumulative volume percentage, x10 , x50
degree of CMC adsorption on particle surfaces. In the case of adding and x 90 , are considered to calculate the RSF:49
46% of CMC after the kneading, a high amount of entangled
molecular chains is available which cause the already mentioned x 90 − x10
RSF = [4]
shear thinning curve. Considering shear rates >10 1/s, shear thinning x 50
applies to all variations. Orientation of polymer chains and particles
takes place with comparable extent. Reduced viscosities for higher Values for the RSF are given in Table I, attesting the broadening of
kneading concentrations are found. Higher polydispersity of the the distribution for increasing kneading concentrations. The max-
particle size can be connected to a decreased viscosity. imum packing density Φmax of the particulate system increases with
In general, the particle size distribution (Fig. 5b) shows a main the broadening, which is directly linked to the viscosity of the
peak around 18 μm corresponding to graphite particles and a second suspension according to Krieger and Dougherty.50 The Krieger-
peak around 400 nm, which is typical for CB aggregates. The anode Dougherty model states that a decrease of the viscosity follows from

Figure 5. Analysis on slurry, electrode and cell level for anodes, processed with different kneading concentrations (50; 55; 65% KC) during the mixing step. (a)
Viscosity of the anode slurries plotted against the shear rate, (b) Volume-based PSD, (c) effective tortuosity at different porosities, (d) electrical interface
resistance, (e) electrical composite volume resistivity shown over a broad porosity range and (f) the derivative of the cell potential U with respect to the SOC
plotted against the SOC (porosity of anode 33 ± 2%).
 Journal of The Electrochemical Society, 2022 169 020531

an increasing maximum packing density, which coincides well with believe that small particles can fill voids at the interface that are
the viscosity values for high shear rates. Moreover, the agglomerates formed if larger particles, especially unaligned flake-like particles,
become more compact and less fluffy without significant change of are attached to the current collector. Also, the connection of binder
specific surface. and CB might be enhanced, intensifying the conductivity. Contrary,
Each slurry is further coated on copper foil, whereby different the absence of CB aggregates and small graphite fractures using
electrode structures are to be expected if slurry properties differ due to 50 wt% KC (light blue) limits the contact area between the particles
variations of the mixing process. Higher packing densities of anodes and the current collector and, therefore, results in higher interface
after coating (i.e., small porosities of uncalendered electrodes) owing to resistance. Thus, a design strategy to minimize the current collector
a broader particle-size distribution PSD are found if the slurry is mixed interface resistance (without considering other properties) for con-
with higher kneading concentrations. This can be rationalized consid- ductive AM particles such as graphite might be the choice of a
ering that abraded graphite fragments fill pores between larger particles. graphite with large polydispersity, including small particles. For the
The initial porosity after coating and drying is measured to be 61.1% for composite volume resistivity, a local maximum at a porosity of
the low kneading concentration of 50 wt%, while it is 55.2% for the high roughly 50% is detected. Even though not fully understood, the
solids content during kneading (65 wt%). Note that the initial porosity increase in the resistivity suggests loosening of contact within the
for all electrode samples is displayed in Table AI in the Appendix. SEM conductive network for slight compression. Similar findings have
images, which are shown in Fig. 6 verify those findings. While surface been published by WESTPHAL et al.52 Increasing compression and
views show mostly pristine particles with sharp edges for the low thus decreasing porosity coincide with decreasing composite volume
kneading concentration (50 wt%), more fragments and round-shaped resistivity. Decreasing pore volume and denser packing of the
graphite particles can be found for higher kneading concentrations. SEM particles provide a higher number of particle-particle contacts and
cross-sections clearly demonstrate the differences in packing density and concomitant an enhanced percolation network of conductive path-
likewise initial porosity. ways. For lower porosities, no significant trend regarding the
The effect on the ionic diffusion in the electrolyte resulting from different kneading concentrations can be detected. Since graphite
changes of the electrode structure can be described with the effective particles are in sufficient contact even without a high number of
tortuosity τeff . However, τeff describes not only the detour of ions due to small fractures within the coating, stronger deagglomeration and
the tortuous electrode structure (geometric tortuosity), but also includes fracturing does not show further impact.
other mass transport impeding effects such as non-constant cross section In order to evaluate the fast-charge performance of the anodes, a
areas of pores (“bottlenecks”). Figure 5c shows the effective tortuosity charge rate cell-test is performed, and the newly developed DCV-
of the differently mixed anodes plotted against the electrode porosity. method is applied in order to detect the onset of lithium-plating at a
Apparently, the effective tortuosity increases, if the slurry is mixed at specific charge rate (further information in experimental section).
higher kneading concentrations. We believe that small, fragmented The onset of lithium-plating is used as the criterion to quantify the
particles clog pores and result in the presence of bottlenecks and, thus, fast-charge capability of the cells. The DCV-method is based on the
impede the diffusion flux and increases the effective tortuosity. localization of a peak in the first derivative of the cell voltage dU/d
Additionally, surface abrasion of the graphite particles, which result (SOC) with respect to the state-of-charge (SOC) of the cell, which is
from particle-particle interactions during mixing, can retard the inter- related to the onset of lithium-plating. If two cells are compared and
calation process and further results in a higher ionic diffusion resistance. both cells experience lithium plating during charging, the cell with
At lower kneading concentrations (light blue), a decrease of the anode an inferior fast-charge performance shows an earlier plating onset
porosity results in an increase of τeff , which is in agreement with the and the peak occurs at lower SOC. Figure 5f depicts the first
literature41 and theoretical considerations such as the Bruggeman derivative of the cell voltage with respect to the SOC during
relation.51 However, at higher kneading concentrations (dark blue) a charging with a rate of 1 C at 25 °C ambient temperature plotted
local minimum of τeff at a porosity of about 40% is found. Although we against the SOC. All cells, regardless of how the slurry of the
have no explanation for this behavior, this phenomenon seems to be containing anode is mixed, show the occurrence of a plating-related
typical at anodes with a broad particles size distribution and was also peak at SOC > 40%. An increase of the kneading concentration
seen recently in other experiments of our group. Potentially displace- (corresponding to an increase of the stress intensity) during slurry
ment of particles and, therefore, opening of pores due to slight mixing results in a shift of the peak to lower SOC, hence, the fast-
calendering might be a reason. charge capability of these cells decreases. While the electric inter-
Apart from the ionic resistance, also the electric resistance face resistance reduces with increasing stress intensity during mixing
depends on the electrode structure. The electric resistance of the (suggesting a better fast-charge performance), the tortuosity in-
anodes on electrode level is divided into two measures by an creases and, hence, implies a reduced fast-charge performance.
electrode resistance meter (HIOKI RM2610): Accordingly, in agreement with previous publications,41,44 the ionic
contribution dominates the overall fast-charge performance.
Summarizing, a higher stress intensity through a higher kneading
concentration during slurry mixing results in a change of the
• The electrode interface resistance [Ω·cm2], describing the electrode structure, which increases the effective tortuosity and
resistance at the interface of current collector (Cu-foil) and electrode reduces the fast-charge capability of the cell. Undesirable fragmen-
coating. tation of graphite appears together with increased CB deagglomera-
• The composite volume resistivity [Ω·cm], quantifying the tion, which is favorable. Although, those effects lead to enhanced
electrical resistivity of the coated layer itself. electrical conductivity, the ionic conductivity is the predominating
factor for fast charging capability in this case.
Both measures are plotted against the electrode porosity in
Fig. 5d and e, respectively. The interface resistance is monotonically Influence of a pre-mixing step.—Pre-dispersing of CB prior to
decreasing with decreasing porosity for all samples as with higher the actual mixing step is a practical measure to reach finely dispersed
compression of the electrode, more contact points between coating conductive additives while avoiding damage to the potentially shear-
and current collector are formed. Especially for high porosities, a sensitive active material.28 As already shown in the previous section,
remarkable difference between the three samples is visible. Higher abrasion of the active material takes place at high stress during the
kneading concentrations during mixing result in a lower interface kneading step, which negatively affects the fast-charge capability of
resistance. According to the PSD, a higher share of small CB the cell and should therefore be avoided. However, premix-manu-
aggregates and graphite fractures has been formed for those settings, facturing is time- and energy-consuming and, thus, should be
allowing a higher contact area towards the current collector. We avoided from an economic (and ecological) point of view. This
 Journal of The Electrochemical Society, 2022 169 020531

Figure 6. SEM images of the anodes’ surfaces and cross-sections. All anodes are uncompressed and have a similar mass loading.

elucidates the complexity of the right choice of formulation strategy, converge to a minimum. Further reduction may not to be possible via
which should be based on product properties and the resulting cell formulation strategies. As already discussed in the previous section, the
characteristics but also include economic considerations. formulation strategy does not impact the composite volume resistivity
Figure 7 shows the slurry, electrode and resulting cell characteristics due to the already high conductivity of the graphite particles.
to compare the manufacturing with and without pre-dispersion for two At cell level, the fast-charge capability behaves according to the
different kneading concentrations. In agreement with the previous detected effective tortuosity. The local maxima in the derivative
section, all findings related to different kneading concentrations are dU/dSOC of the voltage signal in Fig. 7e, which indicates the onset
confirmed. For a kneading concentration of 50 wt%, the flow curves are of lithium-plating, show no variation, if premixing is performed or
identical if a premix is prepared or not (see Fig. 7a). Whereas not. Thus, the fast-charge capability stays constant regardless
manufacturing of a premix does not cause any differences in the whether the formulation strategy includes a premix or not. Only
rheological behavior for the low kneading concentration (light blue and the different kneading concentrations, as discussed in the previous
yellow), slight differences for 65 wt% KC (dark blue and brown) are section, affect the effective tortuosity and, thus, the fast-charge
detected. The slurry with pre-dispersed CB at 65 wt% KC (brown) is capability of cell. Although the interface resistance decreases at
less viscous, since the matrix consisting of CMC and CB is significantly higher kneading concentrations (dark blue and brown), the fast-
more stressed during premix production, which leads to a certain charge capability is deteriorated, which is evident from the onset of
disentanglement of the CMC and a decrease of viscosity. The variant lithium-plating at a lower SOC. Hence, these results show again that
without premix preparation (dark blue) implies addition of CMC ionic resistance dominates the fast-charge capability.
solution with a higher degree of entanglements, leading to a higher In summary, the premixing step shows no effect on the fast-
viscosity. As expected, a higher share of CB aggregates is detected charge capability of the cell. At both kneading concentrations,
for the anode slurries with premixing step (yellow and brown). As similar results of the electrode and resulting cell properties can be
already shown in the previous section, there are no CB aggregates at KC recognized with and without premix preparation. Although, the
50 wt% (light blue). Overall, the premixing step yields a higher degree of presence of more small CB aggregates, which result from the pre-
CB dispersion, but still without graphite fragmentation. Similar findings, mixing step, improve the interface resistance, no change can be
namely a higher amount of CB aggregates, are detected at 65 wt% KC. observed in the effective tortuosity and further the onset of lithium-
However, the difference is less pronounced in this case since the plating. Noteworthy, these findings cannot be directly transferred to
kneading step at high solid content leads to deagglomeration anyways. other active materials, especially electrically non-conductive cathode
At the electrode level, the effective tortuosity in Fig. 7c shows, except for active materials such as NMC, without further verification.
high KC and high porosities, no difference, if a premix at both kneading
concentrations is prepared or not. Similar to the results of the previous
section, τeff increases at higher kneading concentrations (dark blue and Influence of the mixing time.—The net-mixing time has a high
brown) due to the abrasion of the present active material. Concerning the economic relevance with respect to the process time and energy
different formulation strategies with and without a premixing step, consumption. Hence, shortening the mixing time is very desirable,
almost no distinction can be recognized. The premixing step results in but also affects the number of stress events and thus, the product
significantly lower interface resistances (Fig. 7d) in the case of low quality. In this study, kneading times of 30 and 60 min are compared
kneading concentrations at 50 wt% (light blue and yellow). Hence, the to demonstrate the effect of different number of stress events.
presence of CB aggregates strongly impacts the interface resistance, Figure 8a indicates that the mixing time affects the properties of
since small aggregates significantly increase the contact area between the slurry. At lower shear rates, a shortening of the mixing time leads
coating and the copper-foil. Considering the anodes with high kneading to a slightly increased viscosity, which results from a lower degree
concentration and also with KC 50 wt% including premixing step of binder disentanglements. At higher shear rates, the viscosity of the
(yellow, brown and dark blue), the interface resistance is comparable slurry is similar at both mixing times. We assume, that the high shear
in a low range. Even though there are slight differences in the amount of rate enforces a disentanglement of the binder, which lowers the
CB aggregates (PSDs shown in Fig. 7b), the resistance seems to viscosity and overlaps the effects of the mixing time. As expected,
 Journal of The Electrochemical Society, 2022 169 020531

Figure 7. Analysis on slurry, electrode and cell level for anodes, processed with and without premix for two different kneading concentrations. (a) Viscosity of
the anode slurries plotted against the shear rate (light blue and yellow curve are very similar), (b) Volume-based PSD, (c) effective tortuosity at different
porosities, (d) electrical interface resistance, (e) electrocal composite volume resistivity shown over a broad porosity range and (f) the derivative of the cell
potential U with respect to the SOC plotted against the SOC (porosity of anode 33 ± 2%).

the PSD in Fig. 8b indicates, that a lower amount of CB aggregates tortuosity is the dominant property of the electrode, which affects the
and less graphite fragmentation results from the lower number of risk of lithium-plating of the cell. At a shorter mixing time of the
stress events at a shorter mixing time (grey). However, the difference slurry, the onset of lithium-plating appears at a higher SOC, which
in the particle size distribution is comparably small, because after a indicates an optimized fast-charge capability of the cell.
critical fineness is reached further dispersing without increasing Summarizing, the variation of the mixing time has a similar
stress intensity does not change the fineness.35,53 At the electrode effect on the fast-charge capability of the cell as the variation of the
level, the lower mixing time results in a decreased effective kneading concentration. A shorter mixing time of the slurry results
tortuosity of the anode, which is shown in Fig. 8c and can be in a lower number of stress events, which leads to less active
explained by less fragmentations of the graphite resulting in a lower material fragmentations. Similar to the section “Influence of the
number of stress events similar to the observations of the first solid content during mixing”, these fragmentations affect the ionic
section. Especially at a porosity between 30% to 40%, similar values diffusion of the lithium-ions in the electrolyte of the porous
as in the variation of the kneading concentration can be recognized: electrode, which is indicated by the effective tortuosity.
τeff is approximately 5.5 at a low KC of 50% (light blue in Fig. 5c) Furthermore, a decrease of the effective tortuosity results in an
and τeff is around 7.5 at a high KC of 65% (dark blue in Fig. 5c) optimized fast-charge capability.
compared to a short mixing time of 30 min with τeff = 6.2 (grey) and
a long mixing time of 60 min with τeff = 7.7 (brown). Even though, Influence of the mixing device.—Scale-up from lab-scale to
the interface resistance is higher at longer mixing times (brown), all production machinery is a major challenge, willing to transfer
values are in a similar range, which indicates no major distinction. knowledge from research and development to actual production.
Similar to the previous results, the variation of the mixing time does Often, different working principles of mixers are used, matching the
not impact the composite volume resistivity because of the relatively needs and technical options for the considered applications. In this
high conductivity of the graphite particles. case, a common dissolver with a single, centered drive shaft
At the cell level, the variation of the slurry mixing time shows an (equipped with either toothed disk or butterfly tool) is used for
effect on the fast-charge capability of the cell. Again, the effective laboratory scale experiments. In comparison to this, a planetary
 Journal of The Electrochemical Society, 2022 169 020531

Figure 8. Analysis on slurry, electrode and cell level for anodes, with different mixing duration (30 min: grey; 60 min: dark red). (a) Viscosity of the anode
slurries plotted against the shear rate, (b) Volume-based PSD, (c) effective tortuosity at different porosities, (d) electrical interface resistance, (e) electrocal
composite volume resistivity shown over a broad porosity range and (f) the derivative of the cell potential U with respect to the SOC plotted against the SOC
(porosity of anode 33 ± 2%).

mixer, equipped with wall scraper, cross stirrer and additional drive operated at the respective settings prevents particle fragmentations.
shaft is used as pilot plant equipment. For both cases, settings such This does not mean, that the mixing devices cannot be operated in
as tip speed vt , mixing time t and the ratio of the highly active such way, that the performance does vary (e.g., by using strongly
mixing volume to the total volume Vact/Vtot, have been chosen to deviating KCs).
result in comparable values for the combination of stress intensity
and stress number. To demonstrate the quality of the process
Conclusions
transfer, a comparison of the (intermediate-) product characteristics
on slurry, electrode and cell level of both mixing devices is shown in A comprehensive study on the influence of the mixing process of
Fig. 9. graphite anodes on slurry, electrode and cell properties is presented
Figure 9a shows a vertical shift but similar form of the viscosity in this study. Especially the fast-charging capability of the battery
curves of the anode slurries. The offset of the viscosity can be cell is in the focus with respect to the electrochemical properties.
explained by the final solids content of both slurries, which depends Summarizing, broadening of the PSD is found to be a central aspect
on the mixing process of each device. For the laboratory dissolver, for graphite anodes, considering the impact on the fast-charging
high solid contents are difficult to reach due to unsteady flowing capability of the battery cell. Increasing kneading concentration is
behavior with high solid contents and therefore high viscosities. identified as major impact on the broadening. Furthermore, exten-
Planetary mixers such as the one used here allow higher solid sion of the mixing time also slightly broadens the distribution. A
contents during mixing. The anode slurry manufactured in the premixing step is stated to be an effective measure to well disperse
dissolver with a total solids content TSC of 48.7 wt% results in CB, while avoiding graphite fragmentation. Nevertheless, the
lower viscosity values (black curve) than the one produced with the resulting decrease of interface resistance does not affect the fast-
planetary mixer (orange curve) with a TSC of 50.5 wt%. The charging capability. This confirms that electric conductivity is not
viscosity shift due to a more diluted slurry for the dissolver remains the limiting factor for graphite anodes in terms of fast charging. In
the only difference on slurry, electrode, and cell level. Hence, a the end, the established procedure is used to attest the transfer of the
successful transfer of the mixing process from a dissolver to a mixing process from a laboratory dissolver to a planetary mixer.
planetary mixer can be assumed. Please notice, that a transfer is only Similar results in all stages could be reached.
possible at equal conditions in both mixing devices. In this case we All in all, the coherent results allow to derive very clear process-
showed that a low number of stress events for both mixing devices structure-property relations and, therefore, requirements for
 Journal of The Electrochemical Society, 2022 169 020531

Figure 9. Analysis on slurry, electrode and cell level for anodes, produced with two different mixer types (dissolver: clack; planetary mixer: orange). (a)
Viscosity of the anode slurries plotted against the shear rate, (b) Volume-based PSD, (c) effective tortuosity at different porosities, (d) electrical interface
resistance, (e) electrical composite volume resistivity shown over a broad porosity range and (f) the derivative of the cell potential U with respect to the SOC
plotted against the SOC (porosity of anode 33 ± 2%).

intermediate product properties. Focusing on the onset of lithium-

plating as an indicator for the fast-charge capability, the effective the viscosity function is a first indication for the performance
tortuosity appears to be the most relevant electrode property. The deteriorating effect of high shear stresses during mixing of graphite
lower the effective tortuosity, the higher is the SOC at which slurries.
lithium-plating occurs when plating is deliberately triggered by fast Finally, this study presents a valid contribution to a better
charging. Since the electrical conductivity in terms of electrical understanding of the process-structure-property relations with
interface resistance and electrical volume composite resistivity respect to the mixing process in the battery cell manufacturing.
behave in the opposite direction to the effective tortuosity, one can However, further investigations regarding additional process steps or
clearly conclude that the ionic diffusion processes limit the fast- process parameters are necessary to describe the full impact of the
charging of the graphite anode. The increase of the effective manufacturing process on the electrochemical behavior of the
tortuosity can be ascribed to a broadening of the PSD, which results battery cell. Further research is therefore recommended especially
from a CB deagglomeration and graphite fragmentation. Smaller for the drying process, new coating techniques (e.g., dry coating) or
pore sizes in the electrode structures, resulting from the fragmented the formation procedure.
particles clogging the pores of the anode structure, are meant to be
the main reason for the higher ionic diffusion resistance. Apart from Acknowledgments
the effective tortuosity, the electric resistance is also depending on
the electrode structure coming from the PSD. A broadening of the The authors wish to thank the German Federal Ministry for
distribution not only leads to a higher packing density, but also Economic Affairs and Energy and the Bavarian Ministry of
causes a higher contact area between anode coating and current Economic Affairs, Regional Development and Energy for funding
collector, resulting in lower interface resistance. On slurry level, the this work in the BMW project “IPCEI on Batteries” (16BZF102).
broadening of the PSD can be linked to decreasing viscosity values,
validating the Krieger-Dougherty model for this application. Thus, Appendix
 Journal of The Electrochemical Society, 2022 169 020531

Table A·I. Information on the measured final TSC and initial porosity after coating and drying of the anodes.

Mixer Premix KC [wt%] tkneading/min TSC measured [%] initial porosity [%]

Dissolver yes 50 45 48.7 60.8 ± 0.5

PMH10 no 50 60 49.9 61.1 ± 0.4
PMH10 yes 50 60 50.5 62.1 ± 0.4
PMH10 no 55 60 50.4 59.0 ± 0.5
PMH10 no 65 60 49.5 55.2 ± 0.6
PMH10 yes 65 30 49.9 53.6 ± 0.6
PMH10 yes 65 60 50.3 55.1 ± 0.6

ORCID 21. J. B. Habedank, J. Endres, P. Schmitz, M. F. Zaeh, and H. P. Huber, “Femtosecond

laser structuring of graphite anodes for improved lithium-ion batteries: Ablation
Desiree Grießl https://orcid.org/0000-0002-5877-168X characteristics and process design.” J. Laser Appl., 30, 032205 (2018).
Alexander Adam https://orcid.org/0000-0003-2737-5307 22. V. Wenzel, H. Nirschl, and D. Nötzel, “Challenges in lithium-ion-battery slurry
preparation and potential of modifying electrode structures by different mixing
processes.” Energy Technology, 3, 692 (2015).
References 23. Z. Liu and P. P. Mukherjee, “Microstructure evolution in lithium-ion battery
electrode processing.” J. Electrochem. Soc., 161, E3248 (2014).
1. Z. P. Cano, D. Banham, S. Ye, A. Hintennach, J. Lu, M. Fowler, and Z. Chen, 24. G. Liu, H. Zheng, S. Kim, Y. Deng, A. M. Minor, X. Song, and V. S. Battaglia,
“Batteries and fuel cells for emerging electric vehicle markets.” Nat. Energy, 3, 279 “The effects of various conductive additive and polymeric binder contents on the
(2018). performance of a lithium-ion composite.” J. Electrochem. Soc., 155, A887 (2008).
2. A. Kwade, W. Haselrieder, R. Leithoff, A. Modlinger, F. Dietrich, and K. Droeder, 25. G. Liu, H. Zheng, X. Song, and V. S. Battaglia, “Particles and polymer binder
“Current status and challenges for automotive battery production technologies.” interaction: a controlling factor in lithium-ion electrode performance.”
Nat. Energy, 3, 290 (2018). J. Electrochem. Soc., 159, A214 (2012).
3. A. Adam, J. Wandt, E. Knobbe, G. Bauer, and A. Kwade, “Fast-Charging of 26. A. M. Gaikwad and A. C. Arias, “understanding the effects of electrode formulation
automotive lithium-ion cells: in-situ lithium-plating detection and comparison of on the mechanical strength of composite electrodes for flexible batteries.” ACS
different cell designs.” J. Electrochem. Soc., 167, 130503 (2020). Appl. Mater. Interfaces, 9, 6390 (2017).
4. S. Ahmed et al., “Enabling fast charging—A battery technology gap assessment,.” 27. W. Haselrieder, S. Ivanov, H. Y. Tran, S. Theil, L. Froböse, B. Westphal,
J. Power Sources, 367, 250 (2017). M. Wohlfahrt-Mehrens, and A. Kwade, “Influence of formulation method and
5. M. Keyser et al., “Enabling fast charging—Battery thermal considerations.” related processes on structural, electrical and electrochemical properties of LMS/
J. Power Sources, 367, 228 (2017). NCA-blend electrodes.” Prog. Solid State Chem., 42, 157 (2014).
6. A. Meintz et al., “Enabling fast charging—Vehicle considerations.” J. Power 28. D. Grießl, K. Huber, R. Scherbauer, and A. Kwade, “Dispersion kinetics of carbon black
Sources, 367, 216 (2017). for the application in lithium-ion batteries.” Adv. Powder Technol., 32, 2280 (2021).
7. X.-G. Yang, Y. Leng, G. Zhang, S. Ge, and C.-Y. Wang, “Modeling of lithium 29. H. Bockholt, W. Haselrieder, and A. Kwade, “Intensive dry and wet mixing
plating induced aging of lithium-ion batteries.” J. Power Sources, 360, 28 influencing the structural and electrochemical properties of secondary lithium-ion
(2017). battery cathodes.” ECS Trans., 50, 25 (2013).
8. T. Waldmann, B.-I. Hogg, and M. Wohlfahrt-Mehrens, “Li plating as unwanted side 30. H. Bockholt, W. Haselrieder, and A. Kwade, “Intensive powder mixing for dry
reaction in commercial Li-ion cells—A review.” J. Power Sources, 384, 107 dispersing of carbon black and its relevance for lithium-ion battery cathodes.”
(2018). Powder Technol., 297, 266 (2016).
9. N. Legrand, B. Knosp, P. Desprez, F. Lapicque, and S. Raël, “Physical 31. B. Bitsch, N. Willenbacher, V. Wenzel, S. Schmelzle, and H. Nirschl, “Einflüsse
characterization of the charging process of a Li-ion battery and prediction of Li der mechanischen Verfahrenstechnik auf die Herstellung von Elektroden für
plating by electrochemical modelling.” J. Power Sources, 245, 208 (2014). Lithium-Ionen-Batterien.” Chem. Ing. Tech., 87, 466 (2015).
10. T. Amietszajew, E. McTurk, J. Fleming, and R. Bhagat, “Understanding the limits 32. B. Bitsch, J. Dittmann, M. Schmitt, P. Scharfer, W. Schabel, and N. Willenbacher,
of rapid charging using instrumented commercial 18650 high-energy Li-ion cells.” “A novel slurry concept for the fabrication of lithium-ion battery electrodes with
Electrochim. Acta, 263, 346 (2018). beneficial properties.” J. Power Sources, 265, 81 (2014).
11. F. An, R. Zhang, Z. Wei, and P. Li, “Multi-stage constant-current charging protocol 33. M. Haarmann, D. Grießl, and A. Kwade, “Continuous processing of cathode slurry
for a high-energy-density pouch cell based on a 622NCM/graphite system.” RSC by extrusion for lithium‐ion batteries.” Energy Technology, 9, 2100250 (2021).
Adv., 9, 21498 (2019). 34. H. Bockholt, M. Indrikova, A. Netz, F. Golks, and A. Kwade, “The interaction of
12. U. R. Koleti, C. Zhang, R. Malik, T. Q. Dinh, and J. Marco, “The development of consecutive process steps in the manufacturing of lithium-ion battery electrodes
optimal charging strategies for lithium-ion batteries to prevent the onset of lithium with regard to structural and electrochemical properties.” J. Power Sources, 325,
plating at low ambient temperatures,.” Journal of Energy Storage, 24, 100798 140 (2016).
(2019). 35. C. Schilde, I. Kampen, and A. Kwade, “Dispersion kinetics of nano-sized particles
13. Z. Du, D. L. Wood, and I. Belharouak, “Enabling fast charging of high energy for different dispersing machines.” Chem. Eng. Sci., 65, 3518 (2010).
density Li-ion cells with high lithium ion transport electrolytes.” Electrochem. 36. J. Raasch, “Das Verhalten suspendierter Feststoffteilchen in Scherströmungen
Commun., 103, 109 (2019). hoher Zähigkeit.” J. Appl. Math. Mech., 41, 147 (1961).
14. E. R. Logan and J. R. Dahn, “Electrolyte design for fast-charging li-ion batteries.” 37. H. Rumpf, “Beanspruchungstheorie der prallzerkleinerung.” Chem. Ing. Tech.,
Trends in Chemistry, 2, 354 (2020). 31323-337 (1959).
15. M. Ebner, D.-W. Chung, R. E. García, and V. Wood, “Tortuosity Anisotropy in 38. H. Rumpf and J. Raasch, Desagglomeration in Strömungen, European symposium
Lithium-Ion Battery Electrodes.” Adv. Energy Mater., 4, 1301278 (2014). on size reduction (1st: Frankfurt aM:), pp. 151–159, 1962.
16. J. Landesfeind, A. Eldiven, and H. A. Gasteiger, “Influence of the Binder on 39. J. Krekel, “Zerkleinerung von Agglomeraten in Scherströmungen mit besonders
Lithium Ion Battery Electrode Tortuosity and Performance.” J. Electrochem. Soc., hoher Schubspannung.” Chem. Ing. Tech., 38, 229 (1966).
165, A1122 (2018). 40. R. Wengeler, F. Ruslim, H. Nirschl, and T. Merkel, “Dispergierung feindisperser
17. J. Billaud, F. Bouville, T. Magrini, C. Villevieille, and A. R. Studart, “Magnetically agglomerate mit mikro-dispergierelementen.” Chem. Ing. Tech., 76, 659 (2004).
aligned graphite electrodes for high-rate performance Li-ion batteries.” Nat. 41. A. Adam, K. Huber, E. Knobbe, D. Grießl, J. Wandt, and A. Kwade, “Development
Energy, 1 (2016). of an innovative workflow to optimize the fast-charge capability of lithium-ion
18. D. Liu, L.-C. Chen, T.-J. Liu, W.-B. Chu, and C. Tiu, “Improvement of Lithium-Ion battery cells.” J. Power Sources, 512, 230469 (2021).
Battery Performance by Two-Layered Slot-Die Coating Operation,.” Energy 42. J. Landesfeind, J. Hattendorff, A. Ehrl, W. A. Wall, and H. A. Gasteiger,
Technology, 5, 1235 (2017). “Tortuosity Determination of Battery Electrodes and Separators by Impedance
19. L.-C. Chen, D. Liu, T.-J. Liu, C. Tiu, C.-R. Yang, W.-B. Chu, and C.-C. Wan, Spectroscopy.” J. Electrochem. Soc., 163, A1373 (2016).
“Improvement of lithium-ion battery performance using a two-layered cathode by 43. J. Landesfeind, M. Ebner, A. Eldiven, V. Wood, and H. A. Gasteiger, “Tortuosity of
simultaneous slot-die coating.” Journal of Energy Storage, 5, 156 (2016). battery electrodes: validation of impedance-derived values and critical comparison
20. J. B. Habedank, L. Kraft, A. Rheinfeld, C. Krezdorn, A. Jossen, and M. F. Zaeh, with 3D tomography.” J. Electrochem. Soc., 165, A469 (2018).
“Increasing the Discharge Rate Capability of Lithium-Ion Cells with Laser- 44. A. Adam, E. Knobbe, J. Wandt, and A. Kwade, “Application of the differential
Structured Graphite Anodes: Modeling and Simulation.” J. Electrochem. Soc., charging voltage analysis to determine the onset of lithium-plating during fast
165, A1563 (2018). charging of lithium-ion cells.” J. Power Sources, 495, 229794 (2021).
 Journal of The Electrochemical Society, 2022 169 020531

45. H. Dreger, M. Huelsebrock, L. Froboese, and A. Kwade, “Method development for 50. I. M. Krieger and T. J. Dougherty, “A Mechanism for non‐newtonian flow in
quality control of suspensions for lithium-ion battery electrodes.” Ind. Eng. Chem. suspensions of rigid spheres.” Trans. Soc. Rheol., 3, 137 (1959).
Res., 56, 2466 (2017). 51. D.-W. Chung, M. Ebner, D. R. Ely, V. Wood, and R. Edwin García, “Validity of the
46. E. E. Hioki, CORPORATION (2022) [Online]. https://www.hioki.com/global/ bruggeman relation for porous electrodes.” Model. Simul. Mater. Sci. Eng., 21,
products/resistance-meters/resistance/id_6740. 074009 (2013).
47. T. Metzger, The Rheology Handbook: For Users Of Rotational And Oscillatory 52. B. G. Westphal, N. Mainusch, C. Meyer, W. Haselrieder, M. Indrikova, P. Titscher,
Rheometers (Vincentz, Hannover) (2020). H. Bockholt, W. Viöl, and A. Kwade, “Influence of high intensive dry mixing and
48. A. Kraytsberg and Y. Ein-Eli, “Conveying advanced li-ion battery materials into calendering on relative electrode resistivity determined via an advanced two point
practice the impact of electrode slurry preparation skills.” Adv. Energy Mater., 6, approach.” Journal of Energy Storage, 11, 76 (2017).
1600655 (2016). 53. J. K. Mayer, L. Almar, E. Asylbekov, W. Haselrieder, A. Kwade, A. Weber, and
49. A. P. Chauhan and K. Chawla, “Comparative studies on graphite and carbon black H. Nirschl, “Influence of the Carbon Black Dispersing Process on the Microstructure
powders, and their dispersions.” J. Mol. Liq., 221, 292 (2016). and Performance of Li‐Ion Battery Cathodes,’.” Energy Technology, 8, 1900161 (2020).