Carbon-Silicon Based Composite Anodes For Li and Na-Ion Batteries - Final

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 66

COMENIUS UNIVERSITY IN BRATISLAVA

FACULTY OF NATURAL SCIENCES

Carbon-silicon based composite anodes for Li and Na-ion batteries


Written thesis for dissertation examination

2022 M.Sc. Alper Güneren


COMENIUS UNIVERSITY IN BRATISLAVA
FACULTY OF NATURAL SCIENCES

Carbon-silicon based composite anodes for Li and Na-ion batteries


Written thesis for dissertation examination

Study program : Inorganic chemistry


Study field : dAOX_UACH
Department : Institute of Inorganic Chemistry, Slovak Academy of Sciences
Supervisor : Doc. Ing. Zoltán Lenčéš, PhD.

Bratislava 2022 M.Sc. Alper Güneren

1
Contents

List of abbreviations .................................................................................................................................. 4

List of figures .............................................................................................................................................. 5

List of tables ................................................................................................................................................ 8

1. Introduction ....................................................................................................................................... 9

2. Lithium and sodium-ion batteries ................................................................................................ 11

2.1. Fundamental storage mechanisms and anode reactions ................................................. 12


2.1.1. Carbon-based anode materials ................................................................................ 15
2.1.2. Silicon-based anode materials................................................................................. 18
2.2. Solid electrolyte interface (SEI) formation .................................................................... 21
2.2.1. Effect of electrolyte................................................................................................. 23
2.2.2. Effect of binder ....................................................................................................... 24
3. Carbon-silicon based composite anodes ..................................................................................... 27

3.1. Effect of powder characteristics on electrochemical properties ..................................... 33


3.2. Effect of electrode characteristics on electrochemical properties .................................. 36
3.2.1. Electrode composition ............................................................................................ 36
3.2.2. Electrode thickness ................................................................................................. 37
3.2.3. Electrode porosity ................................................................................................... 38
3.3. Effect of testing parameters............................................................................................ 39
4. Aim of the work ............................................................................................................................. 41

5. Experimental studies...................................................................................................................... 41

5.1. Powder preparation ........................................................................................................ 41


5.1.1. Design of experiment .............................................................................................. 41
5.1.2. X-ray diffraction analysis ....................................................................................... 43
5.1.3. Particle size distribution.......................................................................................... 43
5.1.4. Contamination measurement and oxygen content .................................................. 43
5.1.5. Scanning electron microscopy ................................................................................ 44

2
5.1.6. X-ray photoelectron spectroscopy .......................................................................... 44
5.2. Slurry and electrode preparation .................................................................................... 44
5.2.1. Viscosity measurement ........................................................................................... 45
5.3. Battery assembling and testing....................................................................................... 45
6. Experimental results ...................................................................................................................... 46

6.1. Results of ball milling experiments ................................................................................ 46


6.1.1. Correction experiments ........................................................................................... 48
6.2. Results of slurry and coating trials ................................................................................. 51
6.3. Results of electrochemical tests ..................................................................................... 52
7. Conclusions ..................................................................................................................................... 55

8. Future works…………………………………………………………………………………………………………………………..... 56
References ................................................................................................................................................. 57

3
List of abbreviations

Alg Alginate LiTFSI lithium bis


trifluoromethanesulfonyl-
BTP Ball to powder ratio
imide
CE Coulombic efficiency LUMO Lowest occupied molecular
orbit
CMC Carboxymethyl cellulose
MTP Medium to powder ratio
CNT Carbon nanotubes
N/P Ratio Negative/Positive electrode
CV Cyclic voltammetry ratio
CVD Chemical vapor deposition Na+ Sodium ion
DMC Dimethyl carbonate NIBs Sodium-ion batteries
DOE Design of experiment NMC LiNi0.5Mn0.3CO0.2O2
EC Ethylene carbonate PAA Poly (acrylic acid)
EIS Electrochemical impedance PC Propylene carbonate
spectroscopy
PVDF Poly (vinylidene fluoride)
ESS Energy storage systems
RPM Revolutions per minute
EV Electric vehicles
SBR Styrene butadiene rubber
FEC Fluoroethylene carbonate
SEI Solid electrolyte interface
PHEV Plug-in hybrid electric
Vehicles SEM Scanning electron
microscopy
HOMO Highest occupied molecular
orbit SHE Standard hydrogen electrode
ICE Initial coulombic efficiency SHP Self-healing
polymers/binders
Li+ Lithium ion
Si Silicon
LiBOB Lithium bis (oxalate)borate
C/Si Carbon/silicon composite
LiDFOB Lithium difluoro(oxalate)
borate Gr/Si Graphtite/silicon composite
LIBs Lithium-ion batteries SNW Silicon nanowires
LiFSI lithium bis-fluorosulfonyl- VC Vinylene carbonate
imide
XRD X-ray diffraction
LiPF6 Lithium hexafluorophosphate
KIBs Potassium-ion batteries

4
List of figures

Figure 1. Most important anode materials for a) LIBs, and b) NIBs .............................................. 9
Figure 2. Ion storage mechanisms ................................................................................................ 14
Figure 3. Schematic diagram showing the diffusion pathway of Li+ from micro to macroscopic
scale............................................................................................................................................... 15
Figure 4. a) Illustration of the graphite crystal structure, and b) changing of inter-layer distance of
graphite during charging and discharging..................................................................................... 16
Figure 5. Intercalation stages of Li+ into graphene layers and galvanostatic charging and
discharging cycle of graphite with lithium metal as counter electrode ........................................ 17
Figure 6. a) Schematic diagram of capacity vs. charging rate, b) intercalation models of different
carbon structures ........................................................................................................................... 18
Figure 7. a) Scanning transmission electron microscopy images of silicon particles from an
uncycled electrode and after full de-lithiation at 0.01-1.25 V vs. Li/Li+ and, b) schematic
illustration of failure mechanism of Si anode ............................................................................... 19
Figure 8. Theoretical specific capacity and volume expansion ratio of different anode materials for
NIBs .............................................................................................................................................. 20
Figure 9. a) Formation energies of AxSi (A = Li, Na), b) average voltages of AxSi, c) volume
expansion ratios of AxSi ................................................................................................................ 21
Figure 10. Schematic view of the mechanisms occurring at the surface of the silicon nanoparticles.
Formation of the SEI at the beginning of discharge. Formation of the Li−Si alloy upon further
discharge, together with Li2O and LixSiOy interfacial phases. ...................................................... 22
Figure 11. SEI formation potentials a) at different scanning rates, b) at different current densities
....................................................................................................................................................... 23
Figure 12. Schematic illustration of the lithiation of Si anode in EC/DEC and EC/DEC + 10 vol%
FEC electrolyte ............................................................................................................................. 24
Figure 13. a) Influence of the kind of binder on the cycling stability of C/Si anode, and b) rate
capability of Gr/Si anode at different LiPAA/CMC ratios ........................................................... 25
Figure 14. Schematic illustration of the design and behaviour of a conventional silicon electrode
and stretchable self-healing electrode ........................................................................................... 26
Figure 15. Critical factors for designing graphite/silicon composite anodes................................ 27

5
Figure 16. The relationship between the amount of Si in the anode and the specific volumetric
capacity, where the electrode includes 5 wt% binder and no conductive materials. Copyright:
Nature Publishing Group, 2016 .................................................................................................... 28
Figure 17. a) Differential capacity (dQ/dV)/Q vs. voltage plots for different Si ratios, and b) for 20
%wt Si ........................................................................................................................................... 28
Figure 18. a) Capacity and b) specific capacity of the Gr and Si components. The percentages
shown in plot (a) are the estimated expansions of Si particles, exceeding 300 %........................ 29
Figure 19. Magnified cross-sectional views of Gr/Si anodes with different morphologies before
cycling and after 50 cycles ............................................................................................................ 30
Figure 20. Initial coulombic efficiencies of various Si-based anodes a) produced by different
techniques and, b) by structural design ......................................................................................... 31
Figure 21. a) First cycle coulombic efficiency as a function of surface area for different Si
structures, and b) irreversible capacity loss after different milling times ..................................... 33
Figure 22. a) Specific discharge capacity vs. cycle number plots at various current densities for
ball milled Gr/Si nanocomposite half-cells, b) upper image: detachment of the coarser active
material from collector, bottom image: fine grained active material does not lose the contact with
collector after long cycles ............................................................................................................. 34
Figure 23. Composites were cycled at C/15, and the curves for cycles 1, 2, 5, 10, 20, 50, and 100
of the composites prepared from different particle size D90 of a) 1.85 μm, b) 1.26 μm, and c) 0.79
μm are shown. The influence of milling time is shown in: d) 1 h, e) 24 h, and f) 48 h are shown
....................................................................................................................................................... 35
Figure 24. Differential capacity curves of the a) 20 wt% silicon electrode, b) 50 wt% silicon
electrode plotted as a function of the silicon graphite potential (V vs. Li/Li+), and c) CV curves of
graphite electrode .......................................................................................................................... 36
Figure 25. a) Effective diffusion coefficient of Li+ (DLi) extracted at 3.74 V vs Li/Li+ as a function of
electrode thickness, and b) comparison of the discharge curves for different thickness electrodes
at the same current density of 5C .................................................................................................. 38
Figure 26. Differential capacity curves for non-porous and porous Si films................................ 39
Figure 27. a) Specific resistance values according to electrode thickness and porosity, b) Gr/Si
electrode coating thickness in pristine state and after 60 cycles ................................................... 40
Figure 28. Parts of the PAT-Cell ................................................................................................. 45

6
Figure 29. Contour plots showing the effect of parameters on weighted crystallite size of Gr/Si
powders ......................................................................................................................................... 47
Figure 30. Contour plots showing the effect of parameters on contamination level in grams ..... 48
Figure 31. Estimated and obtained results in 95 % prediction interval ........................................ 49
Figure 32. a) Correlation between oxygen content and crystallite size, and b) XPS analysis of
milled powders .............................................................................................................................. 49
Figure 33. SEM analysis of DOE14-DOE17-DOE18, and C3 powders at magnifications 50,
150, 5000 combined with the EDS mapping for graphite and silicon ...................................... 50
Figure 34. a) Optimal viscosity region to obtain proper coating, and b) contour plot showing the
combined effect of weighted crystallite size and solid ratio (content) of the slurry on viscosity at 5
s-1 shear rate .................................................................................................................................. 51
Figure 35. a) Effect of crystallite size on solid ratio, material mass loading and porosity, b)
electrode characteristics according to wet film thickness ............................................................. 52
Figure 36. Specific discharge capacity values at different current densities according to the mass
of active material........................................................................................................................... 53
Figure 37. a) Differential capacity plot, and b) phase transformation voltages of DOE14 sample
....................................................................................................................................................... 54

7
List of tables

Table 1. Lithium versus sodium characteristics ............................................................................ 12


Table 2. Interval of phase transformation voltages for graphite anodes ....................................... 16
Table 3. Effect of ball milling parameters on powder characteristics .......................................... 32
Table 4. Design of experiment table ............................................................................................. 43
Table 5. Results of design of experiment...................................................................................... 46
Table 6. Pareto charts for effective milling parameters and R-sq values of regression models ... 47
Table 7. Results of correction experiments and optimized parameters ........................................ 48
Table 8. Characteristics of active material and material mass loadings of investigated samples for
the anodes of composition 60:30:10 (active material:conductive carbon:binder). ....................... 52

8
1. Introduction

People’s growing demands for new eco-friendly technologies and electronic devices draw
attention to the lithium-ion batteries (LIBs) that have been widely used in the last decades due to
their advantages such as long cycle life and non-memory effect [1]. Nowadays, cathode materials
for LIBs have been developed rapidly, however, the choice of excellent anode materials is still
limited. Among all candidate materials, graphite, as the most popular anode material (372 mAh/g
capacity, LiC6) [2] with layered structure is eligible for intercalation of Li ions (Li+) and shows
excellent cycling ability with relatively low cost. On the other hand, the lack of Lithium resources
in earth crust and the problems of sustainibility make the Na-ion batteries (NIBs) more important
[3]. While graphite is a good option for LIBs, it is well known that electrochemical insertion of
sodium ion (Na+) into graphite is significantly hindered by the insufficient interlayer spacing [4]
and Na+ can only form NaC64 compound during sodiation resulting in the limited capacity of 35
mAh/g [5].

Fig. 1. Most important anode materials for a) LIBs [6], and b) NIBs [5].

Increasing demand for electric vehicles (EV), plug-in hybrid electric vehicles (PHEV),
together with a greater political awareness toward the importance of energy storage and batteries
[7] is pushing the scientists to develop new-generation LIBs and NIBs with a higher energy-storage
capacity which mostly depends on the design of active materials (Fig. 1). As an ideal candidate for
high capacity anode materials for LIBs, silicon has become prominent due to its high theoretical

9
capacity (3579 mAh/g Li15Si4 [8], compared to 725 mAh/g for Na0.76Si [9]), eco-friendly
properties, and abundance on earth. Nano or micron-sized uniformly dispersed Si in the active or
inactive materials has many advantages, on the other hand, the large volume expansion (up to 300
% for Li+, 114 % for Na+ [9], [10]) and stain/stress generation is accompanied with Li+/Na+
insertion and extraction process of Si anode, result in many drawbacks such as adhesion problems,
structure pulverization, and loss of electrical contact and poor cycling stability [11], [12]
Carbon/Silicon composite anodes (such as graphite-Si, graphene-Si, disordered carbon-Si, carbon
nanotubes-Si, and carbon aerogel-Si) with yolk-shell designs, core-shell designs or micro-sized
C/Si with a 3D carbon host [13]–[16] have been demonstrated as an efficient approach for
enhancing the electrochemical performance of anodes. The carbon layer can improve the electronic
conductivity of the composite and it can accommodate the volume changes of Si during the cycling
at the same time [17]. Therefore, these C/Si composites can be regarded as double-phased materials
where both phases are active towards Li+/Na+ within the same potential window. While high
content of Si in the electrode tends to lead to a lower operation voltage and brings higher
gravimetric energy, it decreases the volumetric energy density and electrode stability [18].
Therefore, it is important to determine the best ratio for the material design of the carbon/silicon
composite. It should be noted that bulk Si is not a promising anode material for NIBs, because Si
can only uptake one Na per Si atom and exhibits poor Na diffusion kinetics. However, structurally
modified Si may demonstrate better electrochemical performance such as in amorphous state due
to more favorable binding between Na and Si [19].

Besides the content of the active material of composite anode, also the powder characteristics
such as crystallite size, crystallinity, average particle size, particle size distribution, inactive
phases, and contaminations are also vital parameters to optimize the battery performance. In
contrast to the previous studies about carbon-silicon based anodes which mainly focused on the
production techniques or development of new approaches, in this study the effect of the powder
and electrode characteristics on LIBs/NIBs performance will be considered statistically.

10
2. Lithium and sodium-ion batteries

The term `battery` refers to a device that extracts electrical energy from the chemical energy
through two or more electrochemical cells. Primary batteries extract the energy once via
electrochemical reactions and cannot extract any more energy after the active material has reacted
completely. Here, the `active material` means the electrode material which has the chemical energy
[20]. Besides that, in secondary batteries (or rechargeable batteries) such as lithium-ion batteries
(LIBs), sodium-ion batteries (NIBs), or potassium ion batteries (KIBs) reverse reactions occur
when the electrical current is applied. A secondary battery also can be named a storage battery.

The modern LIBs mainly consist of four parts; electrodes (cathode and anode), separator,
electrolyte, and current collectors [21], [22]. Typically, lithium metal oxide or spinel cathodes
which act as the positive terminal, and graphite anode which acts as the negative terminal is
preferred [7]. These electrodes are separated by a porous separator immersed in a non-aqueous
liquid electrolyte. LiPF6 in a mixture of ethylene carbonate (EC) and at least one linear carbonate
selected from dimethyl carbonate (DMC), diethyl carbonate (DEC), ethyl methyl carbonate (EMC)
and several additives are used in order to provide ion transfer [21], [22]. Conventional LIBs can
be converted to SIBs by simply replacing the lithium ions with sodium ions. However, if the LIBs
electrode materials are directly used in SIBs, the crystal structure may be maintained or most often
the crystals may also deform due to the different sizes of the two ions [23] as given in Table 1.

The anode, where the oxidation takes place, is mainly composed of the active material, the
conducting material, and the binder. Traditionally, the active material is usually prepared with
graphite, the conducting material is a carbon material like acetylene black, and the conventional
binder is non-aqueous polyvinylidene fluoride (PVDF) dispersed in an N-methyl-pyrrolidone
(NMP) solvent [13]. Because of the environmentally harmful effects of NMP, water-soluble
binders are taking attention to eco-friendly future applications since the last decade [24].

To evaluate the performance of the battery systems typical terms such as specific capacity,
specific energy, power density, coulombic efficiency (CE), lifetime, impedance, internal resistance
(IR) drop, state of health (SoH), state of charge (SoC) are used. Mostly, the gravimetric/volumetric
capacity and the specific energy are arbitrary to determine the field of usage. While the specific
capacity (mAh/g) describes the amount of electric charge per gram that can be withdrawn from the

11
cell under specified conditions, the energy (Wh/kg) is calculated as the product of capacity and
average discharge voltage. Other vital parameters for evaluating the performance are; coulombic
efficiency and capacity retention which represent the fraction of the full capacity available from
the battery at a certain number of the cycles [6].

Table 1. Lithium vs. sodium characteristics [19], [25].


Category Lithium Sodium
Atomic weight (g/mol) 6.9 23
Cation radius (Å) 0.76 1.06
Capacity (mAh/g) 3829 1165
Standard hydrogen electrode, SHE (V) -3.02 -2.71
Coordination preference octahedral and tetrahedral octahedral and prismatic
Anode current collector Cu Cu, Al
Cost/ton*(€) 40000 3000
Global abundance (%) 0.0017 2.3
*carbonates

2.1. Fundamental storage mechanisms and anode reactions

Galvanic reactions are thermodynamically favorable (ΔG < 0) and occur spontaneously when
two materials of different positive standard reduction potentials are connected by an electronic
load [26]. LIBs and NIBs charge and discharge through a process of lithiation (Li+ insertion) and
de-lithiation (Li+ extraction) or sodiation and de-sodiation respectively. While charging, ions are
released from the positive electrode (cathode), move toward the negative electrode (anode).
Contrary, during discharging, ions move from the anode to the cathode. Both the electrodes act as
lithium-ion hosts [27]. A separator is located between the electrodes to avoid a short circuit while
the electrolyte supplies ions. Meanwhile, electrons are transferred between electrodes through
current collectors [7], [21].

Lithium metal is the lightest metal and possesses a high specific capacity (3860 mAh/g) and
an extremely low electrode potential (−3.04 V relative to SHE) [13]. The state-of-the-art LIBs use
graphite as the anode because of its high conductivity, good reversibility, and relatively low cost
without safety problems [21]. Most often Li metal oxides (LiCoO2, LiMn2O4, LiFePO4, NMC,
etc.) are used as the cathode. Likewise, for NIBs in lab-scale applications, metal alloys or

12
carbonaceous materials as anode and a high-potential oxide containing sodium ions, phosphides
[23], NASICON (Na super ionic conductor) structures are implemented as the cathode.

In spite of frequent usage in portable devices, smartphones, or laptops, LIBs are increasingly
being investigated for use in EVs. Due to the typical LIBs specific energy of 150-200 Wh/kg and
limited charging potential to about 4.2 V [28], LIBs powered EVs are still not comparable to the
driving range of internal combustion engine vehicles [15]. According to promulgated roadmaps
for the future, specific energy for EVs must be around 350 Wh/kg [29]. To reach high energy
density, the anode materials must combine high specific storage capacity and Coulombic
efficiency [30]. Since capacity is a function of the current density, high current densities are
desirable for certain high-performance applications of batteries [7]. C/n is the current rate which
is the amount of current provided such that the charge (C) stored in 1 g of the anodic material will
take n hours to discharge. Current density is also indicated in units of A/g or A/cm2.

Ions can be stored in an anode or cathode via three fundamentally different mechanisms:
intercalation, conversion, and alloying [19], [31], as can be seen below in Fig. 2. Such conversion
and alloying reaction materials are known to deliver high capacities, however suffer from huge
volume expansion of the host materials due to the continuous self-pulverization of the electrode
materials, which causes permanent capacity loss [19].

Capacity also depends on the current that applied during the charging/discharging process.
When a cell is measured at a high current density, its capacity is often lower compared to the
measurements at lower current densities, because a higher current rate is due to polarization
induced by the internal resistance of the cell [32]. While at low current rates, the electrochemical
performance is related to the specific ohmic, charge transfer, and contact resistance, at high current
rates the ion diffusion in the electrolyte becomes the limiting process [33].

The schematics of the diffusion pathways of Li+ or Na+ are shown in Fig. 3. Ions diffuse not
only into the crystal structure, but also penetrate inside the primary, secondary particles, and into
the porous (preferably nanoporous) electrode regions. In this perspective, it gives us an idea about
the whole charging/discharging process and limiting factors. These factors and the sources of the
internal resistances (R) can be summarized in the following 5 sections.

13
Fig. 2. Schematics of the ion storage mechanisms in LIBs [34].

▪ Electronic resistance of the electrode (Re)


▪ Transport of ions in the electrolyte to the active material particle surface (Rs)
▪ Diffusion of ions through the solid electrolyte interphase (SEI) film (RSEI)
▪ Diffusion of ions at the electrode/electrolyte interface (Rct)
▪ Diffusion of ions within the bulk electrode (crystal structure, primary and secondary) Rdiff
[32]
Based on the literature information which is given above, the following parameters should be
optimized to obtain better electrochemical performance and to achieve high specific capacity and
energy batteries.

▪ Decreasing the internal stresses through decreasing the pathways for ion diffusion
▪ Increasing electrical conductivity through providing better adhesion with current collector
and good conductive network inside the electrode matrix
▪ Decreasing the polarization during charging and discharging
▪ Increasing the contact area between electrode and electrolyte without forming thick SEI
▪ Synthesize high capacity and high energy materials for both anode and cathode side [34]

14
Fig. 3. Schematic diagram showing the diffusion pathway of Li+ from micro to macroscopic scale
[35].

2.1.1. Carbon-based anode materials

Graphite is the best candidate for intercalation or insertion mechanism because it is composed
of hexagonally stacked and bonded sheets of carbon, which are connected by weak van der waals
forces with approximately 3.35 Å (0.335 nm) interlayer spacing [36]. This structure allows Li+
diffusion into the layers (mostly from edge planes) and they occupy an interstitial site between the
graphene layers [22]. Interlayer spacing increases to about 3.5 Å – 3.7 Å when lithium ions are
inserted, as can be seen in Fig. 4. Relatively small expansion (5-20 %) [7], [14], [27] enables the
graphite to keep its structural integrity, and therefore maintain its charge capacity after many
charge-discharge cycles. Although graphite is well investigated and established material for LIBs,
its application in Na-ion batteries is more complicated, because the electrochemical insertion of
Na+ is significantly hindered by the insufficient interlayer spacing [4]. Studies on doping the
graphite structure with the aim to expand layer distances or developing new materials are quite
important for future NIB applications and also for LIBs to increase their storage capacities [36].

15
Fig. 4. a) Illustration of the graphite crystal structure [37], and b) variation of inter-layer distance of
graphite during charging and discharging [38].

Graphite anodes have low specific, areal and volumetric capacity respectively 372 mAh/g (at
room temperature for LiC6) [22], [30], [39], 3.3 mAh/cm2, and 570 mAh/cm3 [14] with ~95 %
coulombic efficiency. It should be noted that the first charging process of these carbons is
accompanied by the huge, irreversible capacity loss [40] because of internal stresses and SEI
formation which is explained in detail in section 2.2.

While charging (intercalation/lithiation), a reduction reaction occurs on the anode side


(negative electrode), therefore, electrons and cations are transferred through the current collector
and electrolyte to the cathode (positive electrode). Half cell reactions take place as follow;

Intercalation C6 + xLi+ + xe- → LixC6 (1)

Deintercalation LixC6 → xLi+ + C6 + xe- (2)

Table 2. Interval of phase transformation voltages for graphite anodes [24], [36], [41], [42].

Intercalation
De-intercalation
Stages Phases voltages vs. Li/Li+ SoC (%)
voltages vs. Li/Li+ (V)
(V)
Stage 1L LiC72
Stage 4 LiC36 0.20 - 0.16 0 - 25 0.23
Stage 3 LiC27
Stage 2L LiC18
0.16 - 0.13 30 - 50 0.15
Stage 2 LiC12
Stage 1 LiC6 0.08 - 0.07 100 0.11

16
Intercalation of non-solvated lithium in graphite occurs at low electrochemical potential below
0.2 V vs. Li/Li+ , and continues until 0 V [36]. However, it can start at slightly higher potentials in
disordered or porous carbons [37]. Inversely, graphite is active in the range of voltage 0.11 - 0.23
V for Li+ de-intercalation [39]. In the case of the staged intercalation compounds, the stage index
denotes the number of planes of the host material (here, constituent graphene layer) in-between
two nearest planes of the intercalant layers (here, Li layer) (Fig. 5) [36].

Fig. 5. Intercalation stages of Li+ into graphene layers and galvanostatic charging and discharging cycle
of graphite with lithium metal as counter electrode [6].

Depending on the structural design, the electrochemical properties may change especially at
low and high current densities, as shown in Fig. 6. While the non-uniform interlayer spacing causes
hindrance towards Li+ transport in-between the graphene layers, the uniform interlayer spacing
results in enhanced electronic conductivity, decreased hysteresis [40], lowered transport resistance,
hence providing better rate capability [36]. On the other hand, an additional capacity increase can
be seen due to the filling of inner nano or micro porosities by the adsorption mechanism [43]. It
has been reported that hard carbon is more favorable instead of graphitic structure, particularly for
NIBs.

17
Fig. 6. a) Schematic diagram of capacity vs. charging rate, b) intercalation models of different carbon
structures [6].

2.1.2. Silicon-based anode materials

Among the alloying materials silicon (Si) has the largest theoretical capacity of 4008 mAh/g,
corresponding to the formation of the Li22Si5 alloy [30] and 3579-3590 mAh/g based on the fully
alloyed form of Li15Si4 [21], [29] at high temperature (400°C) and room temperature (20°C)
respectively [39]. These values are ten times higher than for graphite (372 mAh/g) [27]. It has a
relatively low discharge potential of approximately 0.37 V (vs. Li/Li+) [16], [44] which finds a
good balance between retaining reasonable open-circuit voltage and avoiding adverse lithium
plating process [45]. The initial Coulombic efficiency (ICE) of silicon anode is typically in the
range of 65–85 %, far below that of commercial graphite anodes (90–94 %) [21]. On the other
hand, the slow lithium diffusion kinetics (diffusion coefficient between 10−14 and 10−13 cm2/s), low
intrinsic electric conductivity (10−5 – 10 −3 S/cm), and unstable SEI [39] affect the utilization of Si
anodes commercially [27].

Although silicon has advantages over graphite in terms of capacity, a huge volume change
during alloying and de-alloying (V > 300 %) generates large stresses and strains on the silicon
particles [13]. The leads to morphological changes in atomic level and pulverization at electrode
level. The swelling of silicon anode could also squeeze the electrolyte filling pores of the separator,
which blocks the transportation of Li ions [14]. After severe particle pulverization by cycling, the
freshly exposed surface consumes more electrolytes [16], [46]. The formation of unstable and
excessive SEI increases the internal resistance of the cell and causes loss of electrical contact [30]

18
as it is shown in Fig. 7. All of these structural problems result in capacity fading and limited cycle
life [21], [46].

Fig. 7. a) Scanning transmission electron microscopy images of silicon particles from an uncycled electrode
and after full de-lithiation at 0.01-1.25 V vs. Li/Li+ [46], and b) schematic illustration of failure mechanism
of Si anode [14].

It was shown that decreasing the silicon particles size to nano level helps to mitigate the
pulverization [47]. Critical particle size was determined as 150 nm [21], so below this value no
crack or pulverization was detected. Even though it has a positive influence on stability, there is
still conflict about the effect of decreasing the particle size on overall performance. While some
researchers claim that the nanostructures also provide a good solution for improved capacity [21],
[48] and capacity decay performance, the others indicated that it leads to high irreversible capacity
due to the massive side reactions [16], [30] and continuously growth of SEI [46]. There is also a
possibility of pulverization of anode because of the aggregation of nanoparticles after long cycling
[26]. Beyond any doubt, it can be said that nanostructures have disadvantages due to the high
surface area and causes low coulombic efficiency, low tap density, low areal, and volumetric
capacity. Also, complex and expensive synthesis procedures should be considered.

Similarly to graphite anodes, half-cell reactions take place as shown below and one atom of
silicon can accommodate on average 3.75 atoms of lithium at room temperature [47];

Alloying Sic + xLi+ + xe- → LixSi (3)

De-alloying LixSi → xLi+ + Sia + xe- (4)

19
In the first cycle, the crystalline Si (Sic) structure transforms into an amorphous Li-poor a-Li2Si
phase through alloying with Li at 0.23 V vs. Li/Li+ [39]. The formation of the Li-rich c-Li15Si4
phase, upon further lithiation [49] at ~0.08 V (3578 mAh/g) [24], has a detrimental effect on the
battery performance and leads to high internal stresses, even to particle cracking and thus to
capacity fades [16], [46], [50]. For avoiding the adverse effect of full lithiation of Si, the cutoff
voltage can be adjusted above formation voltage.

The initial lithiation for the amorphous Si requires less energy than for crystalline Si due to the
less defective structure of the latter [47] thus lithiation occurs at the higher voltage. In addition, Li
should preferentially diffuse into the grain boundary, because the diffusion energy of Li into the
boundary is lower than that of into the crystallite [50]. Two de-lithiation plateaus are observed at
higher potential between 0.28-0.33V for Li2Si and 0.45-0.48V for c-Li15Si4 [39], [50], [51]. In the
dQ/dV dependence the peak intensity of c-Li15Si4 phase decreases by cycling and disappears after
a few cycles [8], [52]. After full de-lithiation only amorphous silicon (Sia) remains in the electrode
[16]. It can be concluded that the fast capacity fade of Gr/Si blended electrode during the initial
cycles is strongly related to the Si nanoparticles.

Fig. 8. Theoretical specific capacity and volume expansion ratio of different anode materials for NIBs
[10].

One of the most prominent anode materials is Si for NIBs because of relatively low volume
expansion in comparison with Sn, Sb, Pb, and P (Fig. 8). On the other hand, Si can only uptake 1

20
Na per atom and exhibit poor Na diffusion kinetics as it is shown in Fig. 9. However, structurally
modified Si was predicted to demonstrate better electrochemical performance such as in
amorphous Si due to more favorable binding between Na and Si. Reasonable activation barriers
~0.4 eV is required for Na+ migration into amorphous Si [19].

Fig. 9. a) Formation energies of AxSi (A=Li and Na), b) average voltages of AxSi, and c) volume
expansion ratios of AxSi [9] .

2.2. Solid electrolyte interface (SEI) formation

The solid electrolyte interface (SEI) forms on the surface of the electrode during the first
insertion phase and works as a barrier between the electrolyte and the electrode. This inorganic
layer which is electronically insulating but ionically conducting protects the active material and
prevents further side chemical reactions. The particle size, pore size, degree of crystallinity,
crystallite size, basal-to-edge-plane ratio (for graphite electrodes), surface chemical composition

21
[37], and working conditions (current density and voltage window) affect the SEI properties.
Besides that, additives of electrolyte and binder have a vital impact on layer formation and their
characteristics.

When the potential of the anode is below ~1 V vs. Li/Li+, the anode surface becomes covered
with electrolyte interphase [21]. The major SEI formation begins at ~0.8 V [37] and continues
throughout the process. During cycling, continuous consumption of lithium ions occurs. Therefore,
additional layers form on the top of SEI. This increase causes a bigger mass transfer resistance
which results in a higher electrical resistance [6]. It accelerates the collapse of the electrode [16],
[45] and furthermore decreases the capacity [26] and cyclability [13].

As it mentioned in the previous section, the huge expansion and shrinkage of Si anodes create
a quite different case in terms of SEI formation and growth mechanism Commonly, the SEI with
a thickness of layer between 10−35 nm crushes almost each cycle [53] and exposed fresh electrode
surface tend to forms new and thicker layer which causes low initial coulombic efficiency and
constant capacity drop [21]. In addition, native oxide film is always present on Si particles, so
another oxide interlayers (LixSiOy) form below the SEI, as it is shown in Fig. 10.

Fig. 10. Schematic view of the mechanisms occurring at the surface of the silicon nanoparticles. Formation
of the SEI at the beginning of discharge. Formation of the Li−Si alloy upon further discharge, together with
Li2O and LixSiOy interfacial phases [54].

The formation voltage mainly affects the film formation reaction path, whereas the current
density of the formation impacts the rate of the film formation reaction [55]. The value of film-
forming potential is shifted to the lower potential with increase of the scanning rates (Fig. 11),
whereas the magnitude of peak current increases (polarization take place). Formation at 0.02 C
provides thicker and stable SEI.
22
Fig. 11. SEI formation potentials a) at different scanning rates, b) at different current densities [55].

2.2.1. Effect of electrolyte

The electrolyte between cathode and anode provides a sufficient source of Li+ or Na+ for the
electrode reactions while maintaining electrical insulation [56]. The working window of an
electrolyte is determined by its lowest and highest occupied molecular orbit (LUMO/HOMO),
which should be higher than the electrochemical potential of the anode and lower than the
electrochemical potential of the cathode [22]. In LIBs the most commonly used electrolytes are
LiPF6, LiAsF6, LiClO4, LiBF4, and new generation LiTFSI (lithium bis-trifluoromethanesulfonyl-
imide) or LiFSI (lithium bis-fluorosulfonyl-imide). For NIBs equivalent salts have been developed
including NaPF6, NaClO4, NaBF4, NaTFSI, and NaFSI [56]. By interaction between electrolyte
and electrode SEI layer is formed, as mentioned above. SEI should be strong enough to prevent
the crack formation and should be passive enough to avoid non-uniform thick layer growth. The
improvement of cycle life is possible by using organic or cosolvents with electrolytes [13] such as
ethylene carbonate (EC), dimethyl carbonate (DMC), fluoroethylene carbonate (FEC), lithium
bis(oxalate)borate (LiBOB), lithium difluoro(oxalate) borate (LiDFOB), etc. [21], [29]. For
instance, amorphous and more elastomeric SEI can be formed with 10 vol% FEC additive (Fig.
12) and the lifetime of the anode is expended. Except of the type of additive, also the amount of
electrolyte is also important. Fewer electrolyte results in higher ohmic resistance, larger cell-to-
cell capacity variation, and greater capacity fade. Contrary, SEI thickness increases with the

23
increasing volume of electrolyte. The ratio of the electrolyte volume to total pore volume should
be at least 3.1 [53].

Fig. 12. Schematic illustration of the lithiation of Si anode in EC/DEC and EC/DEC + 10 vol% FEC
electrolyte [49].

2.2.2. Effect of binder

Polymer binder is necessary for the preparation of slurry-based electrodes to hold together the
particulate components in the anode that is critical for the stable electrochemical performance of
the battery [21]. However, it should be noted that some binder-free electrodes have also been
developed for advanced trials. While poly-vinylidene fluoride (PVDF) is commonly using in
graphite anodes, this binder cannot accommodate the huge volume expansion of the silicon anodes
because of low elasticity (Young's modulus 650 MPa). In this sense, traditional binders should be
replaced by new types which have higher elastic modules such as carboxymethyl cellulose (CMC),
styrene butadiene rubber (SBR), poly-acrylic acid (PAA), or alginates (Alg) with elastic modules
close to 4 GPa [29]. Due to the higher elastic modulus these binders are able to withstand the force
of the silicon's volume expansion, maintains better contact with all anode materials, and reduces
the effect of pulverization. On the other hand, PAA and CMC are relatively brittle binders,
meaning that they have minimal elongation and cannot function well as elastomers. Therefore,
also a mixtures of binders were studied and it was reported that the electrode containing 1% CMC
+ 1% SBR had the same cycle stability as the electrode with 10 % PVDF binder [13]. Si-C-Alg
electrodes which exhibited reversible capacities of 3440 mAh/g and coulombic efficiencies of

24
60.1 %, while Si electrodes with PAA-C display reversible capacities of 3500 mAh/g and
coulombic efficiencies of 71.2 % [29]. Hamzelui et al. [39] reported that the combination of
different binders helps to improve the rate capability (Fig. 13). Another advantage of these binders
is their solubility in water, contrary to PVDF, which dissolves only in NMP. In general, according
to the European Union regulations an eco-friendly preparation of batteries should be developed,
including the type of binders and other substances used during processing.

Fig. 13. a) Influence of the kind of binder on the cycling stability of C/Si anode [6], and b) rate capability
of C/Si anode at different LiPAA/CMC ratios [39].

In the last decade the self-healing polymers (SHP) and conductive binders are gaining high
attention for the preparation of LIBs and NIBs. In general, SHP have both mechanical and
electrical healing capabilities, which heal the damages caused by the fracture of silicon
microparticles during battery cycling, as illustrated in Fig. 14 [21]. On the other hand, the
conductive binders contribute to the dual functionality of the electrode, because they eliminate the
issue of electrical connectivity and reduce the mass required for conducting material. Research
activities on these two areas may provide significant improvement for the LIBs and NIBs batteries.

25
Fig. 14. Schematic illustration of the design and behaviour of a conventional silicon electrode and
stretchable self-healing electrode [88].

26
3. Carbon-silicon based composite anodes

While graphite-based anodes have a low specific capacity with high structural stability, silicon-
based anodes show high specific capacity with much lower structural stability due to the huge
expansion/shrinkage problem during charging/discharging [15]. The combination of carbon and
silicon in different ratios in the anode with the aim to achieve superior electrochemical properties
and higher specific capacity was an important task in battery research field in the last decade.
Although for the maximum accessible capacity the Si content should be 10-20 % [57], in the
industrially adopted applications the Si content is much lower, in the range of 5-8 % [39]. Some
problems such as low ICE and poor cycling performance should be surpassed in order to
commercialize C/Si anodes [27]. The material properties listed in Fig. 15 and particularly the
silicon content determine the gravimetric and volumetric capacity as it is shown in Fig. 16. The
post-processing of anodes such as calendaring (pressing the layers together) can improve the
volumetric energy density due to the slippery and deformable properties of graphite, which allow
the C/Si anode to be easily calendared for high electrode density, while preventing the severe
damage of brittle silicon [14]. However, sometimes it is difficult to get a good adhesion between
the electrode and current collector by applying calendaring without peeling off the electrode from
collector.

Fig. 15. Critical factors for designing graphite/silicon composite anodes [14].

27
Fig. 16. The relationship between the amount of Si in the anode and the specific volumetric capacity,
where the electrode includes 5 wt% binder and no conductive materials. Copyright: Nature Publishing
Group, 2016 [14].

Swelling of the Si particles upon repeated de-lithiation and Si content make changes the
intercalation/de-intercalation behaviour and thus, the phase transformation voltages are shifted, as
shown in Fig. 17. The peaks in the plot characterize the different redox reactions, enlisting the
structural transformation linked to the SEI formation, the alloying/de-alloying of Si,
intercalation/de-intercalation of graphite, etc., during the charge/discharge process [58]. Owing to
these problems and behaviours which are mainly depending on Si (Fig. 18), the composition of
the composite anode should be optimized to obtain the required performance.

Fig. 17. a) Differential capacity (dQ/dV)/Q vs. voltage plots for different Si ratios, and b) for 20 wt% Si
[39].

28
Fig. 18. a) Capacity and b) specific capacity of the Gr and Si components. The percentages shown in plot
(a) are the estimated expansions of Si particles, exceeding 300 % [59].

The latest synthesis methods for the production of C/Si anodes has the aim to obtain highly
efficient production with low cost. While co-utilizing these materials, the properties such as
morphology, specific surface area, crystallite size, particle size distributions, and crystallinity
should be considered as mentioned before. For instance, to reach high capacity rate and cycle
stability [60], graphite particles should be as small as possible, however, the large surface area and
poor layered structure affect negatively the specific capacity and the irreversible capacity [14].
Likewise, compensation of Si volume changes is only possible with smaller particles and well-
designed composites, but residues and oxidations from longer treatments creates electrochemically
inactive phases. Moreover, the inherent combination and distribution of the materials in the active
material need to be taken into account. Nonetheless, the results obtained so far evidence that no
radical changes in the battery structure are required, but rather that significant advantages in the
battery performances can be ensured by a proper design of the material structure and judicious
engineering of the electrode. For example, it can be seen in Fig. 19, covering graphite with silicon
instead of distributing randomly changes the swelling behaviour of the electrode extremely.

29
Fig. 19. Magnified cross-sectional views of Gr/Si anodes with different morphologies before cycling and
after 50 cycles [14].

C/Si based structures are generally classified into three main categories;
▪ Core–shell and Yolk–shell structures (C@Si or Si@C)
▪ Porous C/Si structures
▪ Embedded or blended C/Si structures [29], [30]

In these structures preferentially micro or nano-sized materials are used. However, mostly
zero-dimensional 0D (nanoparticles), 1D (nanowires, nanotubes, nanospheres, nanorods,
nanosheets, graphene, etc.) [27], and 2D (thin film) [21] carbon and silicon nanostructures have
been implemented to improve the structural and electrochemical performance of anodes. Together
with structural design and electrolyte additives, some other new approaches such as pre-lithiation

30
have been developed for enhancing the performance of LIBs, especially for higher coulombic
efficiency (Fig. 20).

For producing nanostructures there are two main approaches classified as either bottom-up
processes based on liquid or gas phase synthesis such as chemical vapor deposition (CVD) [27],
atomic layer deposition (ALD) or top-down processes including mechanical milling as one of the
possible production routes which is more economical and simpler than other methods [60]–[63].
It was reported that ball milling of micrometric silicon (in argon) shows better performance than
nanosized commercial silicon [47]. However, Chae et al. [14] mentioned that damaging the
graphite structure, oxidizing the Si surface, and increasing the surface area, through the high
energy ball milling process, brings lower initial CE and irreversible capacity loss.

Fig. 20. Initial coulombic efficiencies of various Si-based anodes a) produced by different techniques
[21], and b) by structural design [27].

There are three main processes during milling which nanoparticles are obtained from coarser
particles; i) deformations and dislocations are introduced into bulk materials during collisions
between balls and bulk materials, ii) nanoscale small grains are formed due to the accumulation,
recombination or rearrangement of dislocations, and iii) the grain orientation became random.
Very often the small grains at the edge of bulk material are peeled off [64]. Ball milling can be
carried out either dry or wet. Shear force domination is an important factor for exfoliation of
graphite layers and decreasing particle size, thus liquid medium can work as a lubricant between
the impacting/shearing balls and partially convert the impact force into sheering force during wet

31
ball milling. A liquid medium may also behave as a cooling agent to reduce the local temperature
which may prevent the agglomeration or non-desirable side reactions throughout the process [65].

The yield of milling depends on several parameters like total volume, milling speed (RPM -
revolutions per minute), milling time, ball type, ball size, ball to powder ratio (BTP), medium type,
medium to powder ratio (MTP), etc., as given in Table 3. However, most often the effect of these
parameters on milling efficiency has been investigated separately or in limited interactions. All
these important parameters should be investigated together with a statistical perspective to find out
the main effect and interactions.

Table 3. Effect of ball milling parameters on powder characteristics.

RPM Crystallite size and particle size decreases with increasing RPM with lower [62] [66]
[67]
crystallinity.
Time X-ray diffraction peak intensities reduce and full-width at half-maximum (FWHM) [12][64]
[68][67]
broaden after long ball milling. Increasing the FWHM with the milling time
[69][70]
indicates the reduction in the average crystalline size. Further excessive grinding
generally causes high temperature therefore agglomeration of particles and the
collapse of the crystalline structure into the amorphous phase.
The relationship between grain size (D) and milling time (t) can be expressed:
D = k t -2/3 (5)
Total Powder loading increases the viscosity of the slurry and then reduces the collision [63][70]
volume
velocity of the balls in the jar. Therefore, the milling efficiency is affected negatively
and the average particle size increases.
Medium Liquid medium with appropriate surfactant reduces the slurry viscosity and helps to [66]

enhance milling efficiency.


Ball type Type: Heavier balls (e.g. WC, ZrO2) are more effective. Size: Large balls are [61][71]
[72]
and size
effective on a crash, small balls are effective on shear. Decreasing the ball size
resulted in a decrease of crystallinity, although the breakage mechanisms with
smaller grinding media are less efficient.
Also, the smaller the ball size the larger the aggregates are formed.
Ball to Increasing the BTP decreases the crystallite size up to one point. Higher ball ratio [62][68]
[73]
powder
may cause cold welding and adds undesirable contamination to the powders due to
ratio
high impact energy. Especially at high BTP, the median particle diameter increased
at the longer milling times.

32
3.1. Effect of powder characteristics on electrochemical properties

In this section, the effect of crystallite size and particle size distribution on anode performance
will be summarized. Capacity fading at the first cycle and structural integrity of composite anode
are closely connected with the characteristics of the silicon, such as the size of the crystallites or
phase structure (crystalline/amorphous), as mentioned in the previous section. Domi et al. [50]
investigated Li+ concentration during lithiation of Si anode and they showed that Li-poor and Li-
rich phases at the grain boundaries determine the structural integrity. The Li-rich phase which
segregates mostly at grain boundaries, increases the stress during lithiation due to the high
diffusion rate, and therefore crystal structure of silicon which has larger grains disrupts easily.
Hence, the rate performance of smaller grain sizes is low because of narrow boundaries but cycling
life is longer. However, in another study it was claimed that nanomaterials have higher
charge/discharge rates because of shorter Li pathways and higher diffusivity [26]. There are still
some unclear points concerning the influence of the crystallite size of C/Si on the performance of
anodes due to the limited investigations on this topic. Most of the research was concentrated on
the effect of particle size or surface area on coulombic efficiency and cell stability. For example,
the extension of milling for 5 hours contributed to the increase of the first cycle irreversible
capacity and electrolyte loss, as it is shown in Fig. 21. The underlying reason of this problem is
the high electroactive surface area of active material [36], [74].

Fig. 21. a) First cycle coulombic efficiency as a function of surface area for different Si structures [47],
and b) irreversible capacity loss after different milling times [74].

33
The reduction of particle size (increase of specific surface area) results in low tap density and
high interparticle resistance due to more space in between particles [21]. The influence of all these
factors results in an electrode with a low volumetric capacity [75], meaning one anode must be
produced thicker [12]. The high surface-to-volume ratio for small particles results in low van der
Waals forces between graphene layers and higher tension during ion insertion because of changes
in the interlayer distance. To withstand the expansion of the Si nanoparticles becomes more
difficult [76][77], hence the stability of the anode is affected negatively. Gauthier et al. [75]
claimed that better performance can be obtained with milled-Si compared to nano-Si. They
explained this situation with better electronic wiring associated with the larger size of the particles
in the milled material which can more easily be connected to the current collector. Contrary to
that, the composite electrode prepared from larger particles is not stable over the cycles, and even
after a few cycles there are severe morphological changes. Moreover, the active material loses the
electrical contact with the current collector as shown in the upper side of Fig. 22(b).

Fig. 22. a) Specific discharge capacity vs. cycle number plots at various current densities for ball milled
Gr/Si nanocomposite half-cells [81], b) upper image: detachment of the coarser active material from
collector, bottom image: fine grained active material does not lose the contact with collector after long
cycles [18].

Preventing agglomeration and obtaining a well-distributed (homogeneous) microstructure are


other vital conditions to build a high-performance battery [78]. Generally, particle sizes are
reduced to values smaller than 1 μm, aggregation may take place, thus contrasting breakage
phenomena. Van der Waals forces play a significant role only at particle size lower than 1 μm [79].

34
Inhomogeneity and large agglomerations of Si up to 10 μm are thought to be detrimental to the
electrode performance because these sites can behave as local hot spots for stress and they may
affect adjacent regions as well, which in turn would promote electrode disintegration and loss of
electrical contact. Cabello et al. [77] indicated the properties for satisfactory performance of the
rate capability as suitable distribution of particles for good contact and the presence of small Si
nanoparticles (<50 nm) for shorter diffusion path for ions. In addition, It was found that 150 nm is
the critical limit for preventing cracks and fracture [80].

Röder et al., and Capone et al. [82], [83] investigated the effect of particle size distribution on
capacity and cycle life for composite anodes and concluded that smaller particles (D90) and narrow
distribution enhance the performance (Fig. 23) because big particles prolong the ion diffusion way
and cause internal resistance. Results show that breaking of particles bigger than 2 μm plays an
important role in increasing the cycle life of the composite.

Fig. 23. Composites were cycled at C/15, and the curves for cycles 1, 2, 5, 10, 20, 50, and 100 of the
composites prepared from different particle size D90 of a) 1.85 μm, b) 1.26 μm, and c) 0.79 μm are shown.
The influence of milling time is shown in: d) 1 h, e) 24 h, and f) 48 h [83].

35
3.2. Effect of electrode characteristics on electrochemical properties

The most important characteristics of the electrode are mass loading, composition thickness,
porosity, and drying conditions. Cycling behaviours are highly dependent on the preparation of
electrodes, test cells and other parts, assembling sequences, and testing conditions. The energy
density of the electrode could be improved theoretically by engineering designs including
increasing electrode thickness, reducing electrode porosity, and decreasing the content of inactive
materials for EV and PHEV applications [32]. However, there are still limiting factors for
production.

3.2.1. Electrode composition

Main components of electrodes are active materials, conductive agents, and binders. The ratio
of active materials, here Gr/Si, changes the intercalation/alloying behaviour as it is shown in Fig.
24(a-b) and determines the theoretical capacity of electrode which can be calculated by formula
[57]:

Qcomposite = QGr.xGr +QSi.xSi (6)

Fig. 24. Differential capacity curves of the a) 20 wt % silicon electrode, b) 50 wt % silicon electrode plotted
as a function of the silicon graphite potential (V vs. Li/Li+) [8], and c) CV curves of graphite electrode [40].

36
Composition and powder characteristics are determinants for the solid ratio of slurry which
changes the mass loading of an electrode (Total mass loading (mareal) and active material mass
loading (mareal-Gr/Si) is expressed by mg/cm2 unit). Besides that, it determines the adhesion
performance that supports the electrical conductivity and drying conditions. As it is mentioned in
section 2.2.2., the type of binder has also an effect on capacity retention, therefore further studies
on the optimization of composition are needed to achieve superior features.

3.2.2. Electrode thickness

It is well known that the electrode thickness has a clear influence on electrochemical
performance [84] because it is directly correlated with the internal resistance [32], the ion diffusion
coefficient [35], [44], the porosity of the electrode, and SEI formation [36]. As mentioned in
section 2.1., the limiting factors such as Re, Rs are affected by the thickness of the electrode directly.
According to equation Re = L/σ, the increase of thickness (L) definitely contributes to the increase
of the electronic resistance of the electrode. Another important issue is losing integrity during the
process due to high-stress accumulation and SEI formation. However, the magnitude of RSEI is not
considerably affected by electrode thickness [32]. The electrolyte resistance increases with the
increasing thickness of the electrode due to the higher mass loading and the associated longer
transport path for Rs [33]. It can be seen in Fig. 24(c) that the internal resistance shifts the peaks to
lower voltage for intercalation/alloying and to higher voltage for de-intercalation/de-alloying and
causing higher polarization.

The capacity loss with increasing electrode thickness is not mainly coming from the cell
polarization, but from the limitation of ion diffusion within the electrode. Ion diffusion is the most
important factor affecting the rate performance of the electrode. It is shown in Fig. 25 that diffusion
effectiveness [35] and gravimetric capacity decrease with increasing thickness [85] and diffusion
time [32], [44].

37
Fig. 25. a) Effective diffusion coefficient of Li+ (DLi) extracted at 3.74 V vs Li/Li+ as a function of electrode
thickness [35], and b) comparison of the discharge curves for different thickness electrodes at the same
current density of 5C [32].

3.2.3. Electrode porosity

Internal resistance, particularly Re and Rs, are affected by the porosity of electrodes similar to
the thickness. Resistance decreases with increasing porosity due to the larger cross-sectional area
available for charge transport [33]. The porosity can be calculated according to the formula:

(7)
where mareal, L, ω and ρ are the mass loading, the layer thickness, the mass fractions and the density
of composites (AM: active material, B: binder, CA: conductive agent).

As it is shown in differential capacity plots (dQ/dV) in Fig. 26, the positions of the charging and
discharging peaks move towards more negative and positive potentials respectively, because of
increased resistance. That is why high capacity fading and stability problem occurs in non-porous
anode, Fig. 26a [86]. On the other hand, when the pores are in nanosize, the results clearly show
that the discharge curves of the porous and non-porous films are almost identical, demonstrating
that the introduction of nano-pores has no detrimental effect on the rate capability.

38
Fig. 26. Differential capacity curves for non-porous and porous Si films [86].

While the effect of porosity is small at low currents, the significant impact of the thickness and
porosity becomes obvious for higher charge/discharge rates, therefore higher porosity is required
for high power applications due to the need for higher ionic transfer [57]. It was also mentioned
that densification of electrodes results in lower initial coulombic efficiency and poor capacity
retention [87]. For example, reducing the porosity from 45 to 34 % decreased the attainable
capacity at 1C from 106 to 64 mAh/g [33]. Pores also behave like buffer region and accommodate
volume expansion of silicon and provide structural stability during cycling.

3.3. Effect of testing parameters

As much as electrode properties, testing conditions are also essential for battery performance,
and it is directly correlated with the level of intercalation or alloying. Voltage window (cut-off
voltages) and current density are the critical parameters for charge-discharge performance and they
play an important role for the formation of the phases. For instance, in Si-based anodes, potentials
above 0.05 V vs. Li/Li+ avoids the formation of the metastable crystalline Li15Si4 phase and results
in a better cycling performance due to the absence of the two-phase reaction of crystalline Li15Si4
to amorphous Li2Si [46]. The changes of coating thickness with different applied cutoff potentials
are shown in Fig. 27(b). Not only the voltage window, but also the current density plays an
important role to determine battery the performance, especially when a thicker film is used.

39
Fig. 27. a) Specific resistance values according to electrode thickness and porosity, and b) Gr/Si
electrode coating thickness in pristine state and after 60 cycles [46].

40
4. Aim of the work

The main objective of the work is the preparation of carbon-silicon based composite anode
with improved gravimetric/volumetric capacity, rate performance and longer lifetime. This
objective can be reached by:
▪ Creating regression models in order to find main effects and interactions of milling
parameters.
▪ Optimization of the milling process and obtaining non-contaminated C/Si composite
powders with fine particles by regression models.
▪ Optimization of the rheological properties of slurries used for the tape casting of anodes.
▪ Optimization of the anode properties such as the composition (C/Si ratio), the thickness,
and the porosity.
▪ Investigating the electrochemical performance of anodes with different binder and
electrolyte systems.

5. Experimental studies

The experimental studies are composed of five main parts: i) preparation of powder which will
be used as the active material of anode, ii) preparation of slurry with different compositions and
additives, iii) preparation of the electrode by casting, iv) battery assembling, and v)
electrochemical testing.

5.1. Powder preparation


5.1.1. Design of experiment

In order to elaborate the effect of ball milling on powder characteristics and evaluate the
experimental studies systematically, a design of experiment (DOE) was prepared based on the
milling speed (RPM), time, medium to powder ratio (MTP), and ball to powder ratio (BTP). 4
factor - 2 level full factorial design with 3 center points were selected as configuration. Graphite
(Imerys Graphite & Carbon, <6 µm particle size) and silicon (SicoMill, 2C grade) powders were

41
mixed as pristine material with the weight ratio of 80:20, total 5 g. Ethanol medium, 50 ml tungsten
carbide (WC) jar and ø2 mm WC balls were used as constant parameters. Ball milling was carried
out in a planetary ball mill (Retsch PM100) under the conditions listed in Table 4. The outputs of
milling were provided from X-ray diffraction (XRD) analysis, contamination measurements and
particle size distribution (PSD) analysis. All statistical analyses were carried out by Minitab
software.

Table 4. Design of experiment table.

Run Order RPM Time MTP BTP


1 100 12.0 5 20.0
2 400 12.0 3 5.0
3 400 12.0 5 5.0
4 100 1.0 5 5.0
5 100 12.0 3 20.0
6 250 6.5 4 12.5
7 400 12.0 5 20.0
8 400 1.0 3 20.0
9 250 6.5 4 12.5
10 100 1.0 3 20.0
11 100 12.0 3 5.0
12 400 1.0 5 5.0
13 250 6.5 4 12.5
14 400 12.0 3 20.0
15 400 1.0 5 20.0
16 100 12.0 5 5.0
17 100 1.0 3 5.0
18 400 1.0 3 5.0
19 100 1.0 5 20.0

42
5.1.2. X-ray diffraction analysis

X-ray diffraction analysis (Bruker AXS D8 Discover, CuK radiation, 40 mA-45 kV, 2 =
10°- 80°) was done in order to find the main (C, Si) and minor phases, together with the evaluation
of the changes of diffraction peak intensities at different milling conditions. The investigation was
carried out using both HighScore and OriginLab software to determine the full width at half
maximum (FWHM) value and peak positions. Crystallite size (D) of powders was calculated using
Scherrer equation with 100 % matched peaks for both graphite (~26.5°) and silicon (~28.4°):

D = κ λ/β.cos(θ) (8)

where κ stands for a shape factor (cubic shape 0.8, spherical crystallites 1.2) λ is wavelength, β is
the FWHM width of peak, and θ is the diffraction angle.

5.1.3. Particle size distribution

Particle size distribution (PSD) analyses were done at Institute of Materials and Machine
Mechanics SAS by Fritsch ANALYSETTE 22 NanoTec, which is based on a laser diffraction
system. Samples were measured while ultrasonic mixing was applying in the water tank. Each
measurement was repeated 3 times and average D10, D50, and D90 values were obtained. The span
value, which represents the distribution range, was calculated using the following formula:

Span = (D90-D10) / D50 (9)

5.1.4. Contamination measurement and oxygen content

To investigate the contamination level, weight of milling balls was measured before and after
for all experiments. The oxygen content of selected powders (DOE2, DOE14, DOE17, DOE18
and C3) were measured by HORIBA EMGA 830 Oxygen, Nitrogen, Hydrogen Analyzer and the
tests were repeated 3 times for each sample.

43
5.1.5. Scanning electron microscopy

Three different samples (DOE14, DOE17, DOE18) were chosen for scanning electron
microscopy (SEM) investigations using JSM-7600F SEM (Jeol, Japan). Energy dispersive
spectroscopy (EDS) from Oxford Instruments (England) was used for quantitative chemical
analysis and chemical element mapping of the carbon, silicon, tungsten, and oxygen (operating at
voltage of 10 kV).

5.1.6. X-ray photoelectron spectroscopy

X-ray photoelectron spectroscopy (XPS) analysis was applied to provide quantitative


elemental information from the surface of the powder mixture after ball-milling experiments. XPS
signals were recorded using a Thermo Scientific K-alpha XPS system (Thermo Fisher Scientific,
UK) equipped with a micro-focused, monochromatic Al K X-ray source (1486.68 eV). An X-ray
beam of 400 m size was used at 6 mA and 12 kV. The spectra were acquired in the constant
analyzer energy mode with pass energy of 200 eV for the survey. Narrow regions were collected
with pass energy of 50 eV.

5.2. Slurry and electrode preparation

Graphite-silicon (powder ratio 80:20) slurries, consisting of active material (prepared by ball
milling using different conditions), carbon black (Vulcan PF, CS Cabot), PVDF binder (~Mw: 534
000 g/mol) were prepared in NMP solution with different compositions and mixed at a speed of
600 rpm for 24 hours. The resulting slurry was cast onto Cu foil (thickness 6µm, MTI) using a Dr.
Blade (Zehntner ZAA 2300) tape caster and a universal applicator (ZUA 2000.60) with different
thicknesses at 10 mm/s speed. Films were dried in a vacuum dryer (Vacucell BMT) for 24 hours
at 100°C. From the films 18 mm discs were punched, before being transferred into an argon-
filled JACOMEX glove box (O2 and H2O concentration < 0.5 ppm). Dry thicknesses of electrodes
were measured by Mitutoyo micrometer at least in 3 points of the anodes. From the Cu foil and
film thicknesses the mass loading and porosity of the electrodes were calculated.

44
5.2.1. Viscosity measurement

Rheological tests were done through Physica MCR 501-Anton Paar device and PP25
measuring plate. The control parameters were adjusted as 400 measuring points for 400 seconds
in `pre-shear/hold and linear shear ramp stepwise` mode.

5.3. Battery assembling and testing

Prepared anodes were dried at 120°C in a glove box with PAT-CELL (Fig. 28) parts such as
PEEK insulator, upper and lower plunges, and Whatman® glass microfiber Grade GF/A separator
for overnight before assembling. Li metal was used as counter electrode and LiPF6 in EC/EMC
(volumetric ratio 50:50) commercial electrolyte was added. Galvanostatic charge-discharge tests
were performed by 8X-PAT Channel device and EL-CELL Software at different current densities
(0.2 A/g to 4 A/g) between 0.05-1.5 V Li/Li+ voltage window.

Fig. 28. Parts of the PAT-Cell.

45
6. Experimental results
6.1. Results of ball milling experiments

Ball milling experiments (19 in total) were carried out in order to obtain active material
powders at different crystallite sizes for graphite-silicon (Gr/Si) composite anode and determine
the effect of milling parameters on powder characteristics, and consequently on electrochemical
properties (Table 5). Regression models were created by Minitab software and adjusted R-squared
values were calculated (Table 6) for crystallite size (93.00 %), span (90.52 %), and contamination
level (91.67 %). It was found that the RPM, time, and interaction of rpm-time parameters were the
most effective inputs for milling system together with the ball to powder ratio.

Table 5. Results of design of experiment.

Run Speed Time MTP BTP Weighted crystallite size D10 D90 Span Cont. Oxygen
order (RPM) (h) Gr/Si (nm) (m) (m) (%) (%)
1 100 12.0 5 20.0 34.9 1.28 7.88 1.77 0.3 -

2 400 12.0 3 5.0 28.6 1.20 9.84 2.33 3.8 2.8

3 400 12.0 5 5.0 27.9 1.21 9.88 2.42 0.6 -

4 100 1.0 5 5.0 38.3 1.72 7.62 1.58 0.0 -

5 100 12.0 3 20.0 37.5 1.63 6.89 1.51 0.1 -

6 250 6.5 4 12.5 27.7 1.20 8.08 1.92 0.8 -

7 400 12.0 5 20.0 21.9 1.12 12.85 3.42 1.0 -

8 400 1.0 3 20.0 31.5 1.10 7.26 1.82 0.9 -

9 250 6.5 4 12.5 29.8 1.21 8.59 1.98 1.0 -

10 100 1.0 3 20.0 39.3 1.70 7.67 1.63 0.0 -

11 100 12.0 3 5.0 39.6 1.62 7.04 1.55 0.1 -

12 400 1.0 5 5.0 38.0 1.37 6.94 1.64 0.2 -

13 250 6.5 4 12.5 32.2 1.08 7.38 1.83 0.7 -

14 400 12.0 3 20.0 24.1 1.15 11.82 2.86 4.6 4.3

15 400 1.0 5 20.0 32.9 1.20 7.07 1.71 1.7 -

16 100 12.0 5 5.0 38.0 1.46 5.74 1.38 0.1 -

17 100 1.0 3 5.0 38.3 1.73 8.11 1.70 0.0 0.6

18 400 1.0 3 5.0 36.9 1.37 7.61 1.78 0.1 1.1

19 100 1.0 5 20.0 37.5 1.61 7.00 1.53 0.2 -

Deviation error is ±1.5 for crystallite size, ±0.05 for D10, ±1 for D90, ±0.15 for span, ±0.1 for contamination, ±0.1 for oxygen

46
Table 6. Pareto charts for effective milling parameters and model summaries.

Crystallite size (nm) Span Contamination (g)

*Parameters named as A: RPM, B:TIME, C:MTP, D:BTP

According to the model summary, R-sq(pred) values are also acceptable which means the
regression can estimate the output values in 95 % prediction interval for crystallite size, span, and
contamination. Contour plots of weighted crystallite size (Fig. 29) and contamination level (Fig.
30) show the optimal regions based on interactions of inputs. The black dot represents the hold
values which are the middle points of the design of experiment, here 250 rpm, 6.5 hours, 4 MTP,
and 12.5 BTP. Hence, through the contour plots, desirable outputs can be estimated.

Fig. 29. Contour plots showing the effect of parameters on weighted crystallite size of Gr/Si powders.

47
Fig. 30. Contour plots showing the effect of parameters on contamination level in grams.

The following requirements were determined for the powder properties to reach the desired
electrochemical properties:
1. Minimize the contamination (below 0.03 g, i.e. <1 %) to prevent the presence/formation
of non-desirable phases.
2. Minimize the crystallite size (d < 40 nm) for longer life-time of battery.
3. Keep the D10 value between 1.3-1.5 µm for preventing the agglomeration of powder and
reach longer life-time of battery.
4. Minimize the D90 value (up to 8.1 µm) for longer life-time of battery.
5. Minimize the D10-D90 span (below 2) for homogenous distribution of particles and
increase the capacity of anode.

6.1.1. Correction experiments

The created regression models were used to estimate the optimal milling conditions with the
goal to reach the required powder properties (crystallite size, D10, D90, span, contamination).

Table 7. Results of correction experiments and optimized milling parameters.

Run order Speed Time MTP BTP Weighted D10 D90 Span Cont. Oxygen
(RPM) (h) crystallite size (m) (m) (%) content
Gr/Si (nm) (%)

C1 251 4.2 5 20.0 31.7 1.23 8.74 2.02 1.0 -

C2 305 3.0 5 15.0 31.4 1.19 8.00 1.93 1.0 -

C3 320 12.0 5 9.5 31.0 1.24 10.90 2.60 0.3 2.8

Deviation error is ±1.5 for crystallite size, ±0.05 for D10, ±1 for D90, ±0.15 for span, ±0.1 for contamination, ±0.1 for oxygen

48
Fig. 31. Estimated and obtained results in 95 % prediction interval.

Three different correction trials (C1, C2, C3) were performed according to regression models.
The milling conditions and obtained powder characteristics are given in Table 7. It can be seen in
Fig. 31 that the obtained results for the level of contamination and crystallite size are between the
upper and lower estimated limits for these properties, which means the model worked properly.
Parameters can adjust in order to optimize system and obtain non-contaminated and fine crystallite
powders.

The elemental oxygen content analysis of the powders showed a rather strong negative
correlation between the oxygen content and crystallite size (Fig. 32a). Moreover, XPS analysis
showed that SiO2 peak intensity (at 103 eV) is the highest for DOE14 and oxide formation is
relatively low on the surfaces of other samples (Fig. 32b). It can be concluded that long and high-
speed milling increases the oxygen content by the formation of native passivation silica layer on
active materials.

Fig. 32. a) Correlation between oxygen content and crystallite size, and b) XPS analysis of milled
powders.

49
Detailed SEM and EDS analyses were done in order to investigate the morphology and
distribution of Si and graphite particles and are shown in Fig. 33. Similar to XRD and XPS analyses
it was found that fine Si particles (bright colour) are homogenously distributed in graphite matrix
(dark colour), when high energy ball milling was applied (DOE14). However, coarse Si particles
(20 m) were detected in DOE17 powder with sharp graphite flakes. The morphology of optimized
C3 powder bears a resemblance to DOE18 powder, however, more fine Si particles, smaller than
1 µm were detected.

Fig. 33. SEM analysis of DOE14, DOE17, DOE18, and C3 powders at magnifications 50, 150, 5000
combined with the EDS mapping for graphite and silicon.

50
6.2. Results of slurry and coating trials

The viscosity of the slurry is as important as the powder properties on the electrochemical
performance of the prepared anode. The slurry properties have a strong influence on material
loading and porosity of the electrode. In the range of 2500-3500 mPa·s measured at 5 s-1 shear
rate ensured a good coating quality as it is shown in Fig. 34. After viscosity measurements the
regression model was created by using Minitab software. A high positive correlation between the
weighted crystallite size of Gr/Si plus conductive carbon and solid ratio was observed (adj. R2 =
99.4 %).

Fig. 34. a) Optimal viscosity region to obtain proper coating and b) contour plot showing the combined
effect of weighted crystallite size and solid ratio (content) of the slurry on viscosity at 5 s-1 shear rate.

The slurries were prepared by using DOE2, DOE14, DOE17, DOE18 and C3 powders and
coated onto copper foil (section 5.2). Electrodes with 0.8-1.6 mg/cm2 total mass loading and 70 -
85 % porosity were obtained. The correlation between solid ratio and total material mass loading
is shown in Fig. 35. It is obvious that a higher solid content causes less porosity due to the low
amount of solvent inside the wet film. On the other hand, different anodes prepared by using the
same solid content and composition showed that increasing thickness and mass loading causes
higher porosity.

51
Fig. 35. a) Effect of crystallite size on solid ratio, material mass loading and porosity, b) electrode
characteristics according to wet film thickness.

6.3. Results of electrochemical tests

Three different anodes were prepared (Table 8) and galvanic charge/discharge tests were
performed as mentioned in section 5.3. Differential capacity dQ/dV graphics were plotted to find
out phase transformation, SEI formation, and plateau peak voltage values.

Table 8. Characteristics of active material and material mass loadings of investigated samples for the
anodes of composition 60:30:10 (active material:conductive carbon:binder).
Powder Weighted D10 D90 Oxygen Solid Electrode Active material
crystallite (µm) (µm) content ratio mass loading mass loading
size (nm) (%) (%) (mg∙cm-2) (mg∙cm-2)

DOE14 24.1±1.5 1.16 ±0.05 11.82±1.3 2.81±0.14 14.0 0.82 0.47

DOE17 38.3±1.5 1.73±0.03 8.11±0.6 0.62±0.04 21.0 0.82 0.50

DOE18 36.9±1.5 1.37±0.05 7.61±0.6 1.19±0.02 20.7 1.17 0.71


*crystallite size: D-Out of plane graphite and silicon (weighted average 8:2)

52
Fig. 36. Specific discharge capacity values at different current densities according to the mass of active
material.

The specific charge and discharge capacities were calculated by using the active material mass.
It was found that anodes with smaller crystallite size show greater capacity (sample DOE14, Fig.
36). However, the first cycling coulombic efficiency is lower than for the others because of the
high surface area and broader SEI formation. According to dQ/dV plots, SEI formation was
observed around 0.75 V for all anodes (Fig. 37a). Lithiation peaks can be seen at 0.3 V and 0.24
V for silicon (enlarged part in Fig. 37a) and crystalline silicon (Sic) transforms into the amorphous
structure (LixSi). Below 0.2 V, Li+ ions start to intercalate into graphite layers and diffusion
continues during charging until the cut-off voltage. Furthermore, de-lithiation of graphite starts at
0.10 V, just before silicon and other de-lithiation peaks appears at 0.14 V, 0.19 V, and 0.23 V. At
0.45 V amorphous LixSi phase transforms into amorphous Sia and disappears after first cycling.
The increasing cycling number also causes the lower intensity of all differential capacity peaks
and it can be concluded that the fast capacity decay of Gr/Si blended electrode during the initial
cycles is strongly related to the Si particles as mentioned in the literature.

53
Fig. 37. a) Differential capacity plot, and b) phase transformation voltages of DOE14 sample.

At higher current densities the deintercalation peaks are shifted to higher voltage due to V= i.R
equation (cathodic intercalation reaction -i current, anodic deintercalation reaction +i current),
while the intercalation voltages are shifted to lower voltages. The average specific discharge
capacities (Qdis) were taken after capacity decay and equations were created at different current
densities (x1 is the crystallite size, x2 is the current density) for 80:20 Gr/Si anodes in 60:30:10
composition.

Table 9. Specific discharge capacity models according to crystallite size and current density.

Current density (A/g) Model equation (mAh/g)


0.2 Qdis = 585.8 – 6.5 x1
0.5 Qdis = 462.3 – 6.5 x1
1 Qdis = 381.8 – 6.5 x1
2 Qdis = 320.5 – 6.5 x1
4 Qdis = 283.0 – 6.5 x1

General equation
Qdis(mAh/g) = 594.70 - 6.50 x1 - 223.50 x2 + 36.67 x22

54
7. Conclusions

Producing high-performance anode material with low cost is possible by ball milling however,
the milling parameters should be adjusted carefully to meet all requirements at the same time.
Although high energy input has an advantage to obtain homogeneous distribution and smaller
particle size, agglomeration and contamination may create problems in terms of capacity and cycle
life of anode. With the aim to minimize the number of experiments for the optimization of milling
conditions 19 different milling experiments have been carried out in order to obtain active material
powders at different crystallite sizes and determine the effect of milling parameters on powder
characteristics. From the analyzed data regression equations were obtained to calculate/estimate
the contamination level, crystallite size and particle size distribution (span). The model milling
experiments showed that the regression equations give good results.

From the milled powders stable slurries were prepared which were used for the tape casting of
anodes. The batteries were assembled and tested.

The preliminary experimental studies can be summarized as follows:

▪ Higher milling speed, longer milling time, and higher ball to powder ratio increases the
energy input to the system, therefore smaller crystallite size (24±1.5 nm), smaller particle
size (D10 ≈ 1.2 m), and homogeneous distribution was obtained together with low
contamination of milled powder with WC and oxygen.
▪ High correlation was found between obtained crystallite size and oxygen content: smaller
the crystallite size, higher the oxygen content is.
▪ Powder loading and viscosity of medium are also important parameters for determining the
milling efficiency, because the collision of balls mostly depends on medium properties.
Low viscosity gives a more chance to the balls for free movement which decreases the
contamination.
▪ Non-contaminated powders with the weighted average crystallite size of 30±1.5 nm were
obtained by using the optimized parameters.
▪ Regression model was created related to crystallite size and solid ratio (content) of the
slurry. Optimal viscosity values were found in the range of 2500-3500 mPa∙s for the good
coating application (tape casting).

55
▪ Galvanostatic charge-discharge tests were performed at different current densities with
different anodes which were prepared using different powders.
▪ Equations for the estimation of specific discharge capacity in dependence on crystallite
size and current density were created. It was found that smaller crystallite size shows
greater capacity particularly at low current densities.
▪ Phase transformation potentials for both charging and discharging were indicated by
differential capacity plots.
▪ It can be concluded that the fast capacity decay of Gr/Si blended electrode during the initial
cycles is strongly related to the Si particles.

8. Future works

Next experiments will be carried out to improve the coulombic efficiency and to prevent the
capacity decay in the beginning of the cycles, as follows:

▪ Water-soluble binders will be implemented to enhance adhesion between current


collector and anode.

▪ Investigation will be done about the high surface area electric conductive substrate which
contains silicon with both self-healing and interface adhesive behaviour.

▪ Calendaring will be applied to the anodes after drying to find out the effect of electrode
integrity on impedance behaviour, capacity retention, and the specific capacity.

▪ Electrolyte with FEC addition will be used to obtain more elastomeric SEI layer.

56
References

[1] L. Qian, J. Le Lan, M. Xue, Y. Yu, and X. Yang, “Two-step ball-milling synthesis of a
Si/SiOx/C composite electrode for lithium ion batteries with excellent long-term cycling
stability,” RSC Adv., vol. 7, no. 58, pp. 36697–36704, 2017, doi: 10.1039/c7ra06671f.
[2] X. Zhang, L. Jin, X. Dai, G. Chen, and G. Liu, “A record-high ion storage capacity of T-
graphene as two-dimensional anode material for Li-ion and Na-ion batteries,” Appl. Surf.
Sci., vol. 527, no. May, p. 146849, 2020, doi: 10.1016/j.apsusc.2020.146849.
[3] U. Arrieta, N. A. Katcho, O. Arcelus, and J. Carrasco, “First-principles study of sodium
intercalation in crystalline Na x Si24 (0 ≤ x ≤ 4) as anode material for na-ion batteries,”
Sci. Rep., vol. 7, no. 1, pp. 1–8, 2017, doi: 10.1038/s41598-017-05629-x.
[4] Y. Wen et al., “Expanded graphite as superior anode for sodium-ion batteries,” Nat.
Commun., vol. 5, no. May, pp. 1–10, 2014, doi: 10.1038/ncomms5033.
[5] W. Zhang, F. Zhang, F. Ming, and H. N. Alshareef, “Sodium-ion battery anodes: Status
and future trends,” EnergyChem, vol. 1, no. 2, p. 100012, 2019, doi:
10.1016/j.enchem.2019.100012.
[6] R. Korthauer, Lithium-ion batteries: Basics and applications. 2018.
[7] C. De Las Casas and W. Li, “A review of application of carbon nanotubes for lithium ion
battery anode material,” J. Power Sources, vol. 208, pp. 74–85, 2012, doi:
10.1016/j.jpowsour.2012.02.013.
[8] M. Wetjen, D. Pritzl, R. Jung, S. Solchenbach, R. Ghadimi, and H. A. Gasteiger,
“Differentiating the Degradation Phenomena in Silicon-Graphite Electrodes for Lithium-
Ion Batteries,” J. Electrochem. Soc., vol. 164, no. 12, pp. A2840–A2852, 2017, doi:
10.1149/2.1921712jes.
[9] S. C. Jung, D. S. Jung, J. W. Choi, and Y. K. Han, “Atom-level understanding of the
sodiation process in silicon anode material,” J. Phys. Chem. Lett., vol. 5, no. 7, pp. 1283–
1288, 2014, doi: 10.1021/jz5002743.
[10] S. L. Chou, Y. Pan, J. Z. Wang, H. K. Liu, and S. X. Dou, “Small things make a big
difference: Binder effects on the performance of Li and Na batteries,” Phys. Chem. Chem.
Phys., vol. 16, no. 38, pp. 20347–20359, 2014, doi: 10.1039/c4cp02475c.
[11] L. Liu, X. Li, G. He, G. Zhang, G. Su, and C. Fang, “SiO@C/TiO2 nanospheres with dual

57
stabilized architecture as anode material for high-performance Li-ion battery,” J. Alloys
Compd., vol. 836, p. 155407, 2020, doi: 10.1016/j.jallcom.2020.155407.
[12] H. F. Andersen et al., “Silicon-Carbon composite anodes from industrial battery grade
silicon,” Sci. Rep., vol. 9, no. 1, pp. 1–9, 2019, doi: 10.1038/s41598-019-51324-4.
[13] A. Casimir, H. Zhang, O. Ogoke, J. C. Amine, J. Lu, and G. Wu, “Silicon-based anodes
for lithium-ion batteries: Effectiveness of materials synthesis and electrode preparation,”
Nano Energy, vol. 27, pp. 359–376, 2016, doi: 10.1016/j.nanoen.2016.07.023.
[14] S. Chae, S. H. Choi, N. Kim, J. Sung, and J. Cho, “Integration of Graphite and Silicon
Anodes for the Commercialization of High-Energy Lithium-Ion Batteries,” Angew.
Chemie - Int. Ed., vol. 59, no. 1, pp. 110–135, 2020, doi: 10.1002/anie.201902085.
[15] T. Chen, J. Wu, Q. Zhang, and X. Su, “Recent advancement of SiOx based anodes for
lithium-ion batteries,” J. Power Sources, vol. 363, pp. 126–144, 2017, doi:
10.1016/j.jpowsour.2017.07.073.
[16] Y. Qi, G. Wang, S. Li, T. Liu, J. Qiu, and H. Li, “Recent progress of structural designs of
silicon for performance-enhanced lithium-ion batteries,” Chem. Eng. J., vol. 397, no.
December 2019, p. 125380, 2020, doi: 10.1016/j.cej.2020.125380.
[17] R. Maddipatla, C. Loka, W. J. Choi, and K. S. Lee, “Nanocomposite of Si/C anode
material prepared by hybrid process of high-energy mechanical milling and carbonization
for Li-ion secondary batteries,” Appl. Sci., vol. 8, no. 11, pp. 1–12, 2018, doi:
10.3390/app8112140.
[18] N. Dimov, S. Kugino, and M. Yoshio, “Mixed silicon-graphite composites as anode
material for lithium ion batteries - Influence of preparation conditions on the properties
of the material,” J. Power Sources, vol. 136, no. 1, pp. 108–114, 2004, doi:
10.1016/j.jpowsour.2004.05.012.
[19] J. Y. Hwang, S. T. Myung, and Y. K. Sun, “Sodium-ion batteries: Present and future,”
Chem. Soc. Rev., vol. 46, no. 12, pp. 3529–3614, 2017, doi: 10.1039/c6cs00776g.
[20] J. M. Kato, Yoshiaki_ Ogumi, Zenpachi_ Perlado Martín, Lithium-ion batteries _
overview, simulation, and diagnostics, vol. 53, no. 9. 2013.
[21] Y. Jin, B. Zhu, Z. Lu, N. Liu, and J. Zhu, “Challenges and recent progress in the
development of Si anodes for lithium-ion battery,” Adv. Energy Mater., vol. 7, no. 23, pp.
1–17, 2017, doi: 10.1002/aenm.201700715.

58
[22] J. Xie and Y. C. Lu, “A retrospective on lithium-ion batteries,” Nat. Commun., vol. 11,
no. 1, pp. 9–12, 2020, doi: 10.1038/s41467-020-16259-9.
[23] J. H. Kim, M. J. Jung, M. J. Kim, and Y. S. Lee, “Electrochemical performances of
lithium and sodium ion batteries based on carbon materials,” J. Ind. Eng. Chem., vol. 61,
pp. 368–380, 2018, doi: 10.1016/j.jiec.2017.12.036.
[24] J. Asenbauer, T. Eisenmann, M. Kuenzel, A. Kazzazi, Z. Chen, and D. Bresser, “The
success story of graphite as a lithium-ion anode material-fundamentals, remaining
challenges, and recent developments including silicon (oxide) composites,” Sustain.
Energy Fuels, vol. 4, no. 11, pp. 5387–5416, 2020, doi: 10.1039/d0se00175a.
[25] M. D. Slater, D. Kim, E. Lee, and C. S. Johnson, “Sodium-ion batteries,” in Advanced
Functional Materials, vol. 23, no. 8, 2013, pp. 947–958.
[26] K. E. Aifantis, S. A. Hackney, and R. V. Kumar, High Energy Density Lithium Batteries.
2010.
[27] X. Liu, X. Zhu, and D. Pan, “Solutions for the problems of silicon–carbon anode materials
for lithium-ion batteries,” R. Soc. Open Sci., vol. 5, no. 6, 2018, doi: 10.1098/rsos.172370.
[28] G. E. Blomgren, “The Development and Future of Lithium Ion Batteries,” J. Electrochem.
Soc., vol. 164, no. 1, pp. A5019–A5025, 2017, doi: 10.1149/2.0251701jes.
[29] X. Shen et al., “Research progress on silicon/carbon composite anode materials for
lithium-ion battery,” J. Energy Chem., vol. 27, no. 4, pp. 1067–1090, 2018, doi:
10.1016/j.jechem.2017.12.012.
[30] M. L. Terranova, S. Orlanducci, E. Tamburri, V. Guglielmotti, and M. Rossi, “Si/C hybrid
nanostructures for Li-ion anodes: An overview,” J. Power Sources, vol. 246, pp. 167–
177, 2014, doi: 10.1016/j.jpowsour.2013.07.065.
[31] C. Erk, T. Brezesinski, H. Sommer, R. Schneider, and J. Janek, “Toward silicon anodes
for next-generation lithium ion batteries: A comparative performance study of various
polymer binders and silicon nanopowders,” ACS Appl. Mater. Interfaces, vol. 5, no. 15,
pp. 7299–7307, 2013, doi: 10.1021/am401642c.
[32] H. Zheng, J. Li, X. Song, G. Liu, and V. S. Battaglia, “A comprehensive understanding
of electrode thickness effects on the electrochemical performances of Li-ion battery
cathodes,” Electrochim. Acta, vol. 71, pp. 258–265, 2012, doi:
10.1016/j.electacta.2012.03.161.

59
[33] C. Heubner et al., “Understanding thickness and porosity effects on the electrochemical
performance of LiNi0.6Co0.2Mn0.2O2-based cathodes for high energy Li-ion batteries,”
J. Power Sources, vol. 419, no. February, pp. 119–126, 2019, doi:
10.1016/j.jpowsour.2019.02.060.
[34] Y. Wu, X. Huang, L. Huang, and J. Chen, “Strategies for Rational Design of High-Power
Lithium-ion Batteries,” Energy Environ. Mater., vol. 4, no. 1, pp. 19–45, 2021, doi:
10.1002/eem2.12088.
[35] H. Gao, Q. Wu, Y. Hu, J. P. Zheng, K. Amine, and Z. Chen, “Revealing the Rate-Limiting
Li-Ion Diffusion Pathway in Ultrathick Electrodes for Li-Ion Batteries,” J. Phys. Chem.
Lett., vol. 9, no. 17, pp. 5100–5104, 2018, doi: 10.1021/acs.jpclett.8b02229.
[36] F. J. Sonia, M. Aslam, and A. Mukhopadhyay, “Understanding the processing-structure-
performance relationship of graphene and its variants as anode material for Li-ion
batteries: A critical review,” Carbon N. Y., vol. 156, pp. 130–165, 2020, doi:
10.1016/j.carbon.2019.09.026.
[37] S. J. An, J. Li, C. Daniel, D. Mohanty, S. Nagpure, and D. L. Wood, “The state of
understanding of the lithium-ion-battery graphite solid electrolyte interphase (SEI) and
its relationship to formation cycling,” Carbon N. Y., vol. 105, pp. 52–76, 2016, doi:
10.1016/j.carbon.2016.04.008.
[38] J. Park, S. S. Park, and Y. S. Won, “In situ XRD study of the structural changes of graphite
anodes mixed with SiOx during lithium insertion and extraction in lithium ion batteries,”
Electrochim. Acta, vol. 107, pp. 467–472, 2013, doi: 10.1016/j.electacta.2013.06.059.
[39] N. Hamzelui, G. G. Eshetu, and E. Figgemeier, “Customizing Active Materials and
Polymeric Binders: Stern Requirements to Realize Silicon-Graphite Anode Based
Lithium-Ion Batteries.,” J. Energy Storage, vol. 35, no. January, p. 102098, 2021, doi:
10.1016/j.est.2020.102098.
[40] D. Aurbach, B. Markovsky, I. Weissman, E. Levi, and Y. Ein-Eli, “On the correlation
between surface chemistry and performance of graphite negative electrodes for Li ion
batteries,” Electrochim. Acta, vol. 45, no. 1, pp. 67–86, 1999, doi: 10.1016/S0013-
4686(99)00194-2.
[41] A. Missyul, I. Bolshakov, and R. Shpanchenko, “XRD study of phase transformations in
lithiated graphite anodes by Rietveld method,” Powder Diffr., vol. 32, no. S1, pp. S56–

60
S62, 2017, doi: 10.1017/S0885715617000458.
[42] C. Sole, N. E. Drewett, and L. J. Hardwick, “Insitu Raman study of lithium-ion
intercalation into microcrystalline graphite,” Faraday Discuss., vol. 172, pp. 223–237,
2014, doi: 10.1039/c4fd00079j.
[43] A. Shellikeri et al., “Investigation of Pre-lithiation in Graphite and Hard-Carbon Anodes
Using Different Lithium Source Structures,” J. Electrochem. Soc., vol. 164, no. 14, pp.
A3914–A3924, 2017, doi: 10.1149/2.1511714jes.
[44] S. Das, J. Li, and R. Hui, “Simulation of the impact of Si shell thickness on the
performance of Si-coated vertically aligned carbon nanofiber as Li-Ion battery anode,”
Nanomaterials, vol. 5, no. 4, pp. 2268–2278, 2015, doi: 10.3390/nano5042268.
[45] X. Zuo, J. Zhu, P. Müller-Buschbaum, and Y. J. Cheng, “Silicon based lithium-ion battery
anodes: A chronicle perspective review,” Nano Energy, vol. 31, no. November 2016, pp.
113–143, 2017, doi: 10.1016/j.nanoen.2016.11.013.
[46] M. Wetjen, S. Solchenbach, D. Pritzl, J. Hou, V. Tileli, and H. A. Gasteiger,
“Morphological Changes of Silicon Nanoparticles and the Influence of Cutoff Potentials
in Silicon-Graphite Electrodes,” J. Electrochem. Soc., vol. 165, no. 7, pp. A1503–A1514,
2018, doi: 10.1149/2.1261807jes.
[47] S. Y. Lai, K. D. Knudsen, B. T. Sejersted, A. Ulvestad, J. P. Mæhlen, and A. Y. Koposov,
“Silicon Nanoparticle Ensembles for Lithium-Ion Batteries Elucidated by Small-Angle
Neutron Scattering,” ACS Appl. Energy Mater., vol. 2, no. 5, pp. 3220–3227, 2019, doi:
10.1021/acsaem.9b00071.
[48] M. Wagemaker and F. M. Mulder, “Properties and promises of nanosized insertion
materials for li-ion batteries,” Acc. Chem. Res., vol. 46, no. 5, pp. 1206–1215, 2013, doi:
10.1021/ar2001793.
[49] W. Huang et al., “Dynamic Structure and Chemistry of the Silicon Solid-Electrolyte
Interphase Visualized by Cryogenic Electron Microscopy,” Matter, vol. 1, no. 5, pp.
1232–1245, 2019, doi: 10.1016/j.matt.2019.09.020.
[50] Y. Domi, H. Usui, K. Sugimoto, and H. Sakaguchi, “Effect of Silicon Crystallite Size on
Its Electrochemical Performance for Lithium-Ion Batteries,” Energy Technol., vol. 7, no.
5, pp. 1–4, 2019, doi: 10.1002/ente.201800946.
[51] J. Sung et al., “Subnano-sized silicon anode via crystal growth inhibition mechanism and

61
its application in a prototype battery pack,” Nat. Energy, vol. 6, no. 12, pp. 1164–1175,
2021, doi: 10.1038/s41560-021-00945-z.
[52] J. Müller et al., “Si-on-Graphite fabricated by fluidized bed process for high-capacity
anodes of Li-ion batteries,” Chem. Eng. J., vol. 407, no. May 2020, p. 126603, 2021, doi:
10.1016/j.cej.2020.126603.
[53] S. J. An et al., “Electrolyte Volume Effects on Electrochemical Performance and Solid
Electrolyte Interphase in Si-Graphite/NMC Lithium-Ion Pouch Cells,” ACS Appl. Mater.
Interfaces, vol. 9, no. 22, pp. 18799–18808, 2017, doi: 10.1021/acsami.7b03617.
[54] B. Philippe et al., “Nanosilicon electrodes for lithium-ion batteries: Interfacial
mechanisms studied by hard and soft X-ray photoelectron spectroscopy,” Chem. Mater.,
vol. 24, no. 6, pp. 1107–1115, 2012, doi: 10.1021/cm2034195.
[55] T. Zhu et al., “Manipulating the Composition and Structure of Solid Electrolyte
Interphase at Graphite Anode by Adjusting the Formation Condition,” Energy Technol.,
vol. 7, no. 9, pp. 1–9, 2019, doi: 10.1002/ente.201900273.
[56] E. Bekaert, L. Buannic, U. Lassi, A. Llordés, and J. Salminen, Electrolytes for Li- and
Na-Ion Batteries: Concepts, Candidates, and the Role of Nanotechnology. Elsevier Inc.,
2017.
[57] D. J. Pereira, J. W. Weidner, and T. R. Garrick, “The Effect of Volume Change on the
Accessible Capacities of Porous Silicon-Graphite Composite Anodes,” J. Electrochem.
Soc., vol. 166, no. 6, pp. A1251–A1256, 2019, doi: 10.1149/2.1211906jes.
[58] A. L. Michan, M. Leskes, and C. P. Grey, “Voltage Dependent Solid Electrolyte
Interphase Formation in Silicon Electrodes: Monitoring the Formation of Organic
Decomposition Products,” Chem. Mater., vol. 28, no. 1, pp. 385–398, 2016, doi:
10.1021/acs.chemmater.5b04408.
[59] K. P. C. Yao, J. S. Okasinski, K. Kalaga, J. D. Almer, and D. P. Abraham, “Operando
Quantification of (De)Lithiation Behavior of Silicon–Graphite Blended Electrodes for
Lithium-Ion Batteries,” Adv. Energy Mater., vol. 9, no. 8, 2019, doi:
10.1002/aenm.201803380.
[60] D. Wang, M. Gao, H. Pan, J. Wang, and Y. Liu, “High performance amorphous-
Si@SiOx/C composite anode materials for Li-ion batteries derived from ball-milling and
in situ carbonization,” J. Power Sources, vol. 256, pp. 190–199, 2014, doi:

62
10.1016/j.jpowsour.2013.12.128.
[61] B. E. Nilssen and R. A. Kleiv, “Silicon Powder Properties Produced in a Planetary Ball
Mill as a Function of Grinding Time, Grinding Bead Size and Rotational Speed,” Silicon,
vol. 12, no. 10, pp. 2413–2423, 2020, doi: 10.1007/s12633-019-00340-0.
[62] R. Shashanka and D. Chaira, “Optimization of milling parameters for the synthesis of
nano-structured duplex and ferritic stainless steel powders by high energy planetary
milling,” Powder Technol., vol. 278, pp. 35–45, 2015, doi:
10.1016/j.powtec.2015.03.007.
[63] H. Shin, S. Lee, H. Suk Jung, and J. B. Kim, “Effect of ball size and powder loading on
the milling efficiency of a laboratory-scale wet ball mill,” Ceram. Int., vol. 39, no. 8, pp.
8963–8968, 2013, doi: 10.1016/j.ceramint.2013.04.093.
[64] K. Wu, D. Su, J. Liu, R. Saha, and J. P. Wang, “Magnetic nanoparticles in nanomedicine,”
arXiv, no. November, 2018.
[65] J. Zhang et al., “In Situ synthesis of SiC-graphene core-shell nanoparticles using wet ball
milling,” Ceram. Int., vol. 44, no. 7, pp. 8283–8289, 2018, doi:
10.1016/j.ceramint.2018.02.012.
[66] F. L. Zhang, M. Zhu, and C. Y. Wang, “Parameters optimization in the planetary ball
milling of nanostructured tungsten carbide/cobalt powder,” Int. J. Refract. Met. Hard
Mater., vol. 26, no. 4, pp. 329–333, 2008, doi: 10.1016/j.ijrmhm.2007.08.005.
[67] H. N. Kim, J. W. Kim, M. S. Kim, B. H. Lee, and J. C. Kim, “Effects of ball size on the
grinding behavior of talc using a high-energy ball mill,” Minerals, vol. 9, no. 11, 2019,
doi: 10.3390/min9110668.
[68] H. Ashrafizadeh and M. Ashrafizaadeh, “Influence of processing parameters on grinding
mechanism in planetary mill by employing discrete element method,” Adv. Powder
Technol., vol. 23, no. 6, pp. 708–716, 2012, doi: 10.1016/j.apt.2011.09.002.
[69] E. Ganjeh, H. Khorsand, and S. Shahsavar, “Study of mechanical milling mechanisms in
Al-Si eutectic system,” Mater. Lett., vol. 143, no. March 2015, pp. 144–147, 2015, doi:
10.1016/j.matlet.2014.12.111.
[70] R. Schmidt, H. Martin Scholze, and A. Stolle, “Temperature progression in a mixer ball
mill,” Int. J. Ind. Chem., vol. 7, no. 2, pp. 181–186, 2016, doi: 10.1007/s40090-016-0078-
8.

63
[71] C. Teng, D. Xie, J. Wang, Z. Yang, G. Ren, and Y. Zhu, “Ultrahigh Conductive Graphene
Paper Based on Ball-Milling Exfoliated Graphene,” Adv. Funct. Mater., vol. 27, no. 20,
2017, doi: 10.1002/adfm.201700240.
[72] N. Hlabangana, G. Danha, and E. Muzenda, “Effect of ball and feed particle size
distribution on the milling efficiency of a ball mill: An attainable region approach,” South
African J. Chem. Eng., vol. 25, pp. 79–84, 2018, doi: 10.1016/j.sajce.2018.02.001.
[73] S. Pazesh, J. Gråsjö, J. Berggren, and G. Alderborn, “Comminution-amorphisation
relationships during ball milling of lactose at different milling conditions,” Int. J. Pharm.,
vol. 528, no. 1–2, pp. 215–227, 2017, doi: 10.1016/j.ijpharm.2017.05.043.
[74] C. Natarajan, H. Fujimoto, A. Mabuchi, K. Tokumitsu, and T. Kasuh, “Effect of
mechanical milling of graphite powder on lithium intercalation properties,” J. Power
Sources, vol. 92, no. 1–2, pp. 187–192, 2001, doi: 10.1016/S0378-7753(00)00528-0.
[75] M. Gauthier et al., “A low-cost and high performance ball-milled Si-based negative
electrode for high-energy Li-ion batteries,” Energy Environ. Sci., vol. 6, no. 7, pp. 2145–
2155, 2013, doi: 10.1039/c3ee41318g.
[76] M. Ratynski, B. Hamankiewicz, M. Krajewski, M. Boczar, D. Ziolkowska, and A.
Czerwinski, “Impact of natural and synthetic graphite milling energy on lithium-ion
electrode capacity and cycle life,” Carbon N. Y., vol. 145, pp. 82–89, 2019, doi:
10.1016/j.carbon.2019.01.019.
[77] M. Cabello, E. Gucciardi, A. Herrán, D. Carriazo, A. Villaverde, and T. Rojo, “Towards
a High-Power Si@graphite Anode for Lithium Ion Batteries through a Wet Ball Milling
Process,” Molecules, vol. 25, no. 11, pp. 1–14, 2020, doi: 10.3390/molecules25112494.
[78] X. Tie, Q. Han, C. Liang, B. Li, J. Zai, and X. Qian, “Si@sioX/graphene nanosheets
composite: Ball milling synthesis and enhanced lithium storage performance,” Front.
Mater., vol. 4, no. January, pp. 1–7, 2018, doi: 10.3389/fmats.2017.00047.
[79] S. Fadda, A. Cincotti, A. Concas, M. Pisu, and G. Cao, “Modelling breakage and
reagglomeration during fine dry grinding in ball milling devices,” Powder Technol., vol.
194, no. 3, pp. 207–216, 2009, doi: 10.1016/j.powtec.2009.04.009.
[80] X. H. Liu, L. Zhong, S. Huang, S. X. Mao, T. Zhu, and J. Y. Huang, “Size-dependent
fracture of silicon nanoparticles during lithiation,” ACS Nano, vol. 6, no. 2, pp. 1522–
1531, 2012, doi: 10.1021/nn204476h.

64
[81] M. Ashuri, Q. He, Y. Liu, and L. L. Shaw, “Investigation towards scalable processing of
silicon/graphite nanocomposite anodes with good cycle stability and specific capacity,”
Nano Mater. Sci., vol. 2, no. 4, pp. 297–308, 2020, doi: 10.1016/j.nanoms.2019.11.004.
[82] F. Röder, S. Sonntag, D. Schröder, and U. Krewer, “Simulating the Impact of Particle
Size Distribution on the Performance of Graphite Electrodes in Lithium-Ion Batteries,”
Energy Technol., vol. 4, no. 12, pp. 1588–1597, 2016, doi: 10.1002/ente.201600232.
[83] I. Capone, K. Hurlbutt, A. J. Naylor, A. W. Xiao, and M. Pasta, “Effect of the Particle-
Size Distribution on the Electrochemical Performance of a Red Phosphorus-Carbon
Composite Anode for Sodium-Ion Batteries,” Energy and Fuels, vol. 33, no. 5, pp. 4651–
4658, 2019, doi: 10.1021/acs.energyfuels.9b00385.
[84] E. Hosseinzadeh, J. Marco, and P. Jennings, “Electrochemical-thermal modelling and
optimisation of lithium-ion battery design parameters using analysis of variance,”
Energies, vol. 10, no. 9, 2017, doi: 10.3390/en10091278.
[85] X. Lu et al., “Microstructural Evolution of Battery Electrodes During Calendering,”
Joule, vol. 4, no. 12, pp. 2746–2768, 2020, doi: 10.1016/j.joule.2020.10.010.
[86] J. Sakabe, N. Ohta, T. Ohnishi, K. Mitsuishi, and K. Takada, “Porous amorphous silicon
film anodes for high-capacity and stable all-solid-state lithium batteries,” Commun.
Chem., vol. 1, no. 1, pp. 1–9, 2018, doi: 10.1038/s42004-018-0026-y.
[87] F. Jeschull et al., “Graphite Particle-Size Induced Morphological and Performance
Changes of Graphite–Silicon Electrodes,” J. Electrochem. Soc., vol. 167, no. 10, p.
100535, 2020, doi: 10.1149/1945-7111/ab9b9a.
[88] C.Wang et al., “Self-healing chemistry enables the stable operation of silicon microparticle
anodes for high-energy lithium-ion batteries,” Nature Chemisty, vol. 5, no, 12, pp. 1042-
1048, doi: 10.1038/nchem.1802

65

You might also like