Journal of Energy Storage 15 (2018) 181–190

Contents lists available at ScienceDirect

Journal of Energy Storage

journal homepage: www.elsevier.com/locate/est

Theoretical optimization of electrode design parameters of Si based

anodes for lithium-ion batteries
C. Heubnera,* , U. Langklotza , A. Michaelisa,b
a
Institute of Materials Science, TU Dresden, 01062 Dresden, Germany
b
Fraunhofer IKTS Dresden, 01277 Dresden, Germany

A R T I C L E I N F O A B S T R A C T

Article history:
Received 10 October 2017 Silicon is considered one of the most promising anode materials for next-generation lithium-ion
Accepted 9 November 2017 batteries. However, dramatic volume expansion during the lithiation of silicon complicates its practical
Available online 16 November 2017 implementation. Literature reports nanostructured electrodes, which are capable to accommodate the
volume expansion, reducing associated swelling, degradation and capacity fading. However, several
Keywords: phenomena associated with the volume expansion, such as the reduction of the electrode’s porosity, are
Battery inherent to the system and must be carefully considered for targeted engineering of high-energy lithium-
Silicon anode ion batteries. Herein, we determine design criteria of silicon based electrodes, taking into account the
Energy density
volume expansion during lithiation. A “deformation threshold” is defined signifying the minimum value
Volumetric
of the initial porosity that must be adjusted to avoid plastic deformation and dramatic reduction of the
Gravimetric
electrode’s porosity during charging. In addition, a “C-rate threshold” is defined, guaranteeing diffusion
limited currents not falling below a desired discharge rate. The impact of these theoretical limitations on
the electrochemical performance of silicon-based electrodes is analyzed from an engineering point of
view. The derived relations are used to optimize the electrode design parameters regarding maximum
gravimetric and volumetric capacity.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction volumetric changes, stress evolution, mechanical properties, and

In the recent years, silicon (Si) has attracted intense attention as
suitable anode material for next generation lithium-ion batteries
(LIBs). Si offers both a low electrode potential vs. Li/Li+ (0–1 V) and

1
a high theoretical gravimetric capacity (4200 mAh gSi ) in com-

1
parison to common graphite-based anodes (372 mAh gC ) [1–3].
However, several challenging issues emerged during the develop-
ment of Si based anodes. Si based electrodes can exhibit
significantly higher irreversible capacity and lower coulombic
efficiency compared to conventional graphite anodes, which may
lower the practical capacity and cycle life performance [4,5].
Moreover, the volume expansion of 300
400 % during the
lithiation of Si causes several challenging issues regarding
mechanical stress and degradation, and swelling of the battery
pack during operation, which complicates the practical imple-
mentation of Si as an anode material in LIBs [4,6–8]. Fundamental
studies that have been conducted to understand structural and
* Corresponding author.
E-mail address: christian.heubner@ikts.fraunhofer.de (C. Heubner).
fracture behavior of nanostructured Si anodes have recently been
reviewed [9]. Repeated volume changes during cycling cause loss
of electrical contact between individual active material units and
the current collector [10]. Furthermore, the volume expansion of Si
causes disruption and subsequent recreation of the solid electro-
lyte interphase layer [4,7], as well as fracture and pulverization of
Si/alloy particles [11], resulting in capacity fade.
However, due to the very promising properties of Si, tremen-
dous efforts have been made to tackle these critical issues. The
acquired approaches include e.g. the nanostructuring of Si [1,12–
14], targeted porosity tailoring [15,16], and graphite blending [17–
19] to accommodate the volume expansion. The resulting electro-
des reveal outstanding electrochemical performance regarding the

1
specific capacity, e.g. in mAh gSi , related to the mass of active
material used.
However, from the engineering point of view, we are interested
in the energy density, e.g. scalable in Wh kg
1, related to the total
mass of the electrode or cell. The mass of the electrode includes the
active material(s), the loading of the pores by liquid electrolyte,
binder and conductive additives and some kind of current
collector. Thus, the energy density of the electrode depends on

https://doi.org/10.1016/j.est.2017.11.009
2352-152X/© 2017 Elsevier Ltd. All rights reserved.
182 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190
the mass ratio of active to inactive materials used. Common
technological approaches to increase the energy density of porous
insertion electrodes include the increase of the electrode thickness
(active material/current collector ratio) and the decrease of the
porosity (electrolyte/active material ratio). However, both increas-
ing the electrode thickness and decreasing the porosity reduce the
power density due to increased limitations of lithium transport
inside the electrode [20,21]. Additionally, the degree of graphite
blending affects both the attainable capacity and the mean volume
expansion of the composite electrode [24]. Therefore, optimizing
these design parameters is a critical issue for the development of
high-energy and high-power LIBs [4].
In this theoretical study, we determine optimum design criteria
of porous Si based electrodes, taking into account the volume
expansion during lithiation, the degree of graphite blending and
different limits of swelling. For this purpose, a “deformation
threshold” is defined, signifying the minimum value of the initial
porosity that must be adjusted to avoid plastic deformation and
dramatic reduction of the electrode’s porosity during charging.
Furthermore, a “C-rate threshold” is defined, guaranteeing
diffusion limited currents not falling below a desired C-rate and
thus avoiding massive reduction of the capacity during rapid
charging. The impact of these design restrictions on the perfor-
mance parameters of Si based anodes is analyzed. Moreover, the
derived relations are used to optimize the electrode design
parameters regarding gravimetric and volumetric capacity.
Hereinafter we consider porous Si based electrodes of arbitrary
design blended with varying amounts of graphite. This includes
different nanostructures, e.g. nanorods, nanotubes, nanoparticles,
etc. on a two-dimensional current collector foil or some sort of
three-dimensional current collector, e.g. copper foam. Non-porous
electrodes, as e.g. Si thin films, are not considered.
2. Porosity
The volume expansion associated with the lithiation of Si
causes a swelling of the electrode and a decrease of the porosity. In
this section, we investigate these effects by straightforward
calculations of the state of charge (SoC) dependent porosity
considering the volume expansion and different limits of swelling.
The initial porosity, e0, of a porous insertion electrode of arbitrary
design is given by:
X additive). The range of initial porosities resulting in total
Vi
e0 ¼ 1
i ð1Þ
V
Fig. 1. a) Development of the porosity during lithiation of the electrode for different initial porosities. Relevant parameters used for the calculation are listed at the top of the
figure. b) Schematic illustration of the development of the porosity for different initial porosities.
with Vi and V being the initial volumes of the solid components and
the electrode. In our calculations, the solid components include
variable mass fractions of silicon (Si) and graphite (C) as well as
binder (B) and conductive additives (A). Assuming a linear volume
expansion by n-times the initial volume between SoC = 0 and
SoC = 1, the porosity of the electrode can be expressed as:
X
V i ð1 þ ni SoC Þ
eðSoCÞ ¼ 1
i ð2Þ
V ð1 þ nS SoC Þ
In Eq. (2), nS accounts for the permitted limit of swelling of the
electrode (typically approximately 10 %). Mixing Eqs. (1) and (2)
and introducing mass fractions gives:
ð1
e0 Þ X vi 
eðSoCÞ ¼ 1
Xv ð1 þ ni SoCÞ ð3Þ
ð1 þ nS SoCÞ i r i i ri
i
A detailed derivation of Eq. (3) can be found as Supplementary
material.
Fig. 1 shows the change of the electrode’s porosity during the
lithiation of the electrode for different initial porosities, computed
from Eq. (3). Relevant variables used for the calculations are listed
at the top of Fig. 1. Parameters kept constant throughout the study
are listed in Table 1.
As expected, the porosity clearly decreases with increasing SoC.
This effect is most pronounced at low initial porosities. For initial
porosities in the range of 20–40 %, which is typical for commercial
graphite electrodes, the porosity of a Si based electrode would
rapidly decrease to zero during charging. Such a process would
induce enormous mechanical stress inside the electrode, causing
pulverization of Si, loss of electrical contact, etc. and corresponding
capacity fade. In the case of medium initial porosities of 50–70 %,
the reduction of the porosity is less pronounced. However, even at
these relatively large initial porosities, when compared to
commercial graphite electrodes, the porosity decreases to zero
before reaching SoC = 1. Only for very high porosities, above 80 %,
the porosity at SoC = 1 is significantly larger than zero. This
behavior is illustrated schematically in Fig. 1b.
Fig. 2a shows the porosity at SoC = 1 as a function of the initial
porosity (SoC = 0) for different mass fractions of graphite and Si in
the composite, respectively. Please note that vSi = 95 % corresponds
to a pure Si electrode, whereas vSi = 0 % represents a pure graphite
electrode (each including 3 wt% binder and 2 wt% conductive
densification of the electrode during charging clearly expands
with increasing amount of Si. In the case of a pure graphite
 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190 183

Table 1
Nomenclature and physical properties used in this work

Symbol Explanation Value [unit]

a Area [m2]
c0 Initial Li-concentration in the electrolyte 1 [mol L
1]
D Li-diffusion coefficient in electrolyte 410
10 [m2 s
1][22]
F Faraday’s constant 96485[As mol
1]
I C-rate (charge-discharge rate) [h
1]
j Current density [A m
2]
L Electrode thickness [m]
LCC Current collector thickness [m]
z Valence 1 [
]
m Mass [kg]
ni Volume expansion factor of the component i [
] or [%]
(nSi = 3, nC = 0.1, nA = 0, nB = 0)
nS Permitted limit of swelling of the electrode [
] or [%]
SoC State of charge [
] or [%]
V Volume of the electrode [m3]
Vi Volume of the solid component i [m3]
Q Capacity [Ah kg
1]

Greek letters
e Porosity [
]
g Bruggeman coefficient 2.5 [
] [23]
ri Mass density of the component i (rSi = 2336, rC = 2200, rA = 2200, rB = 1800, rCC = 8920, rel = 1500) [kg m
3]
vi Mass fractions of solid components (vSi = variable, vC = 95 %
vSi, vA = 2 %, vB = 3 %) [
] or [%]

Subscripts
A Conductive additive
B Binder
C Graphite
CC Current collector
Si Silicon
0 Refers to initial conditions (SoC ¼ 0)
lim Diffusion limited
eff Effective parameter
min Minimum
m Refers to mass (gravimetric)
V Refers to volume (volumetric)

electrode (vSi = 0 %), the porosity at SoC = 1 is identical to the initial
porosity because the volume expansion factor of graphite (nC = 10
%) is identical to the permitted limit of swelling of the electrode
(nS = 10 %).
Fig. 2b shows the porosity at SoC = 1 as a function of the initial
porosity (SoC = 0) for different limits of swelling (0 %, 10 %, 20 %) and
three different mass fractions of graphite and Si in the composite,
respectively. As expected, the smaller the limit of swelling, the
wider the range of initial porosities resulting in total densification
of the electrode during charging. The magnitude of this effect
increases with increasing amount of graphite.
3. Electrolyte depletion
The Figs. 1 and 2 reveal that the porosity of Si based electrodes
decreases significantly during charging. Thus, the amount of

Fig. 2. a) Porosity at SoC = 1 as a function of the initial porosity (SoC = 0) for different mass fractions of graphite and Si in the composite. b) Porosity at SoC = 1 as a function of the
initial porosity (SoC = 0) for different limits of swelling (0 %, 10 %, 20 %) and three different mass fractions of graphite and Si in the composite. Relevant parameters used for the
calculations are listed at the top of the figures.
184 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190

electrolyte contained in the pores of the electrode decreases as concentration in the entire electrode rapidly decreases to zero,
well, which in turn will affect the electrochemical performance. causing irreversible side reactions and dramatic loss in capacity
During the lithiation reaction, lithium is removed from the (Fig. 3d). To avoid such a critical depletion of lithium in Si based
electrolyte and inserted into the active material(s). Consequently, anodes, the porosity at the fully lithiated state must be high
the lithium concentration of the electrolyte, contained in the pores enough to ensure sufficient transport properties at a desired C-
of the electrode, decreases. A concentration gradient forms rate.
throughout the cell, causing diffusion of lithium towards the To describe this behavior, diffusion limited C-rates are
anode. If the lithium concentration in the electrolyte decreases to computed as a function of the SoC by implementing the derived
zero, the insertion reaction stops. The maximum current that can SoC dependence of the porosity (Eq. (3)) into Eq. (5) (for details see
be achieved is the so-called diffusion limited current, jlim [25]: Supplementary material). The diffusion limited C-rate of a porous
c0 insertion electrode is a critical parameter as it defines the
jlim ¼ 2zFDeff ð4Þ minimum duration to charge and discharge the battery. The
L
numerical values presented in Fig. 4 denote the maximum C-rate
with z, F and Deff being the valence, Faraday’s constant and the that can be achieved from a Si based anode (L = 50 mm) with a
effective diffusion coefficient of lithium in the electrolyte. L and c0 certain initial porosity at particular SoC values, thus, characterizing
are the thickness of the electrode and the initial lithium the rate capability of the electrode. The rate capability decreases
concentration in the electrolyte. The effective diffusion coefficient with increasing SoC. For example, at an initial porosity of e0 = 80 %,
in porous media is commonly described by the Bruggeman the maximum C-rate that can be obtained from the electrode at
approximation [23]: SoC = 0 is 9.6 C but decreases to approximately 0.85 C at SoC = 1. In
Deff ¼ Deg ð5Þ the case of low initial porosities, the maximum C-rate tends to very
small values already at medium SoC. This indicates that such
where D and g are the diffusion coefficient of lithium in the electrodes cannot be charged rapidly, e.g. within 30 min (2.0 C). In
electrolyte and the Bruggeman coefficient. From the Eqs. 4 and 5 contrast to that, in the case of large initial porosities, the maximum
can be inferred that the limiting current depends on the porosity, e, C-rate hardly changes with the SoC. Electrodes with larger initial
of the electrode. Consequently, the reduction of the porosity,
associated with the volume expansion, decreases the diffusion
limited current and thus the rate capability and power density of Si
based anodes.
This effect is schematically illustrated in Fig. 3. For large
porosities (Fig. 3a), lithium transport in the electrode is sufficient
to keep the lithium concentration in the electrolyte close to the
initial value (dashed line). In this case, the lithium insertion
reaction can take place throughout the entire electrode without
significant limitations (e.g. diffusion overpotential). Reducing the
porosity at the same current density results in a distinct
concentration gradient within the electrode and a corresponding
decrease of the average lithium concentration in the electrolyte
(Fig. 3b). In this case, the lithium insertion reaction also takes place
throughout the entire electrode, but the reaction rate will vary over
the thickness of the electrode due to local differences of kinetic
limitations [26]. Further reduction of the porosity causes the total
depletion of lithium close to the current collector (Fig. 3c). In this
case, the effective transport properties of the porous electrode are
not capable to maintain a sufficient lithium concentration in the
electrolyte. At locations of total depletion, the lithium insertion Fig. 4. Development of diffusion limited C-rates during lithiation of the electrode
reaction cannot take place, which reduces the attainable capacity for different initial porosities (L = 50 mm). Relevant parameters used for the
of the electrode. In the case of very low porosities, the lithium calculations are listed at the top of the figure.

Fig. 3. Schematic illustration of electrolyte depletion at constant current for different porosities. a) Large porosity: lithium transport in the electrode is sufficient to keep the
lithium concentration in the electrolyte close to the initial value. b) Lowered porosity: distinct concentration gradient and corresponding decrease of the average lithium
concentration in the electrolyte. c) Low porosity: total depletion of lithium close to the current collector. d) Very low porosity: lithium concentration in the entire electrode
rapidly decreases to zero.
 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190
porosity have lower specific capacities. Consequently, the absolute
charge-discharge current is lower compared to electrodes with
reduced porosity at the same C-rate. For example, at an initial
porosity of e0 = 90 %, the maximum C-rate is almost unaffected
from the change in porosity due to both the larger initial porosity
and the lower absolute current.
According to the Eqs. (4) and (5), the diffusion limited current
depends on the electrode thickness and the porosity. Fig. 5a shows
diffusion limited C-rates at the fully lithiated state, SoC = 1, as a
function of the initial porosity for different electrode thicknesses
(20, 50, 100 mm). Again, the diffusion limited C-rate increases with
increasing initial porosity. Furthermore, the diffusion limited C-
rate at a given initial porosity decreases with increasing electrode
thickness. This is reasonable due to the longer diffusion path and
the increase of the corresponding diffusion time constant [27].
Additionally, the C-rate is related to the capacity of the electrode,
which increases with increasing electrode thickness as well as with
decreasing porosity. Consequently, the absolute charge-discharge
current that corresponds to a certain C-rate increases with
increasing electrode thickness and decreasing porosity. Very thick
and less porous electrodes are most severely affected by diffusion
limitations and the maximum C-rate that can be obtained at SoC = 1
is dramatically reduced. For example, at an initial porosity of 85 %,
the maximum C-rate that can be obtained at SoC = 1 is 6.3 C and
0.25 C for electrode thicknesses of 20 mm and 100 mm, respective-
ly.
Fig. 5b shows diffusion limited C-rates at the fully lithiated
state, SoC = 1, as a function of the initial porosity for different mass
fractions of graphite and Si in the composite, respectively.
Increasing the amount of graphite in the composite clearly
increases the rate capability of the electrode. This is caused by
both the lower absolute current and the smaller reduction of the
electrode’s porosity during lithiation (Fig. 2a).
4. Design thresholds
From the results shown above it can be concluded that the
volume expansion significantly affects the electrochemical perfor-
mance of Si based anodes. The reduction of the electrode’s porosity
results in mechanical stress and limitations of the rate capability.
The extent of these limitations depends on the design parameters
of the electrode. Increasing the porosity and decreasing the
thickness of the electrode reduces mechanical stress and increases
Fig. 5. Diffusion limited C-rates at SoC = 1 as a function of the initial porosity for a) different electrode thicknesses (20, 50, 100 mm) and b) different mass fractions of graphite
and Si in the composite. Relevant parameters used for the calculations are listed at the top of the figures.
185
the rate capability. However, all of these measures lead to a
reduction of the specific capacity of the electrode.
Therefore, it is necessary to identify design criteria that ensure:
 The avoiding of plastic deformation of the active material and
corresponding mechanical stress (“Deformation threshold” –
DFTH)
 A desired diffusion limited C-rate of the electrode at SoC = 1 (“C-
rate threshold” – CRTH)
and to optimize the electrode design parameters taking account
of these thresholds.
We define the DFTH by the porosity in hexagonal close packing
of equal spheres (e = 26 %). An electrode exceeding this value during
lithiation is expected to undergo substantial mechanical stress due
to mutual mechanical interactions between individual active
material units, e.g. particles or nanotubes. Literature reports
significant impact of mechanical stress associated with the volume
expansion during lithiation of Si on the electrochemical properties
of the electrode, e.g. accelerated degradation and associated
capacity fade. Therefore, the exceedance of the porosity of a
hexagonal close packing during lithiation of the electrode should
explicitly be avoided.
The CRTH is defined by the diffusion limited current not falling
below a desired C-rate. A detailed derivation of the DFTH and the
CRTH can be found as Supplementary material.
Initial porosities meeting the DFTH and the CRTH are computed
as a function of the electrode thickness (Fig. 6a). Relevant
parameters used for the calculations are listed at the top of
Fig. 6a. As expected, no correlation between the DFTH and the
thickness of the electrode is observed. In contrast to that, initial
porosities meeting the CRTH depend on both the electrode
thickness and the desired diffusion limited C-rate. The higher
the desired diffusion limited C-rate and the electrode thickness,
the higher the required initial porosity. For instance, for an
electrode thickness of 20 mm, initial porosities of 79, 81, and 84 %
are necessary to realize C-rates of 1.0, 2.0 and 5.0 C, respectively.
For an electrode thickness of 100 mm, initial porosities of 91, 94,
and 97 % are necessary to realize C-rates of 1.0, 2.0 and 5.0 C,
respectively.
Fig. 6b presents initial porosities meeting the DFTH and the
CRTH (2.0 C) as a function of the mass fraction of Si and graphite in
the composite, respectively. The required initial porosities clearly
186 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190

Fig. 6. Initial porosities meeting the DFTH and the CRTH. a) Plotted against the electrode thickness for different desired diffusion limited C-rates. b) Plotted as a function of the
mass fraction of Si and graphite in the composite for different electrode thicknesses and a desired diffusion limited C-rate of 2.0 C. Relevant parameters used for the
calculations are listed at the top of the figures.

decrease with increasing amount of graphite in the composite. In Detailed derivations of the Eqs. 6 and 7 are available as
the case of mid to high electrode thicknesses, the CRTH restricts the Supplementary material. Fig. 7 shows the gravimetric capacity
design freedom, whereas the DFTH limits for thin electrodes. as a function of the initial porosity for different electrode
thicknesses. The defined DFTH and CRTH are indicated in the
5. Electrode performance figures by the respective shadowed areas. As expected, the
theoretical value of Qm decreases with increasing porosity and
To evaluate the impact of the derived design criteria on the decreasing electrode thickness. From this point of view, thick Si
theoretical performance of Si based anodes the gravimetric (Qm) based anodes with low porosity would be beneficial for the
and volumetric (QV) capacity is computed as a function of the preparation of high-energy LIBs. However, the defined restrictions
initial porosity for different electrode thicknesses by: significantly narrow the design freedom and the attainable specific
X capacity of Si based anodes. The DFTH only permits porosities
v Q X vi 
i i m;i LCC

larger than approximately 75 % and corresponding specific
Qm ¼ ir
e0 r el þ r ð6Þ
1 þ 1
1e0 Þ i L CC capacities. The CRTH shows a complex nonlinear behavior as it
signifies a set of design parameters (porosity and electrode
thickness) guaranteeing the diffusion limited current not falling
X
ð1
e0 Þ i vi Q m;i below the indicated C-rate while maximizing Qm.
QV ¼ X   ð7Þ Thereby, the maximum value of Qm decreases with increasing
vi
ir
1 þ LLCC desired C-rate. This is reasonable since any increase of the rate
i

capability requires improving the effective transport properties,

The Eqs. (6) and (7) consider the Si active material, the
e.g. by reducing the electrode thickness and enlarging the porosity,
electrolyte contained in the pores of the electrode, and assume a
which in turn lowers the gravimetric capacity.
foil type copper current collector with a thickness of 10 mm.

Fig. 7. a) Gravimetric capacity as a function of the initial porosity for different electrode thicknesses (20, 50, 100 mm). The defined DFTH and CRTH are indicated by the
respective shadowed areas. Relevant parameters used for the calculations are listed at the top of the figure. b) Magnified view if the respective area of interest.
 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190
6. Optimization of electrode design parameters
From the magnified view in Fig. 7b can be seen that the CRTH
forms an optimum. This optimum signifies the design parameters
(porosity and electrode thickness) leading to maximum attainable
capacity for the indicated composition of the electrode in
consideration of the defined thresholds.
In a next step, we determine the optimum design parameters
for varying CRTH, limits of swelling, and mass fractions of Si and
graphite in the composite electrode. The optimization procedure is
described in detail in Supplementary material. Fig. 8a shows the
maximum attainable gravimetric capacity as a function of the CRTH
for different mass fractions of Si and graphite. Fig. 8b depicts the
corresponding design parameters. The maximum attainable
gravimetric capacity increases with decreasing CRTH and increas-
ing amount of Si in the composite. The obtained CRTH dependence
characterizes the well-known tradeoff between energy- and
power-density. Electrodes intended for high-energy applications
(low CRTH) have large electrode thicknesses and relatively low
porosity (Fig. 8b). This reduces the inactive to active material ratio
but diminishes the effective transport properties of the electrode,
resulting in reduced rate capability and power-density, respec-
tively. On the other hand, electrodes intended for high-power
applications (high CRTH) have thin electrodes and relatively large
porosity (Fig. 8b). This supports the effective transport properties
of the electrodes but diminishes the active to inactive material
ratio, resulting in reduced gravimetric capacity. The magnitude of
the tradeoff between energy- and power-density increases with
increasing amount of Si in the composite. This is caused by
increasing mean volume expansion of the composite and the
corresponding restrictions of design freedom (Fig. 6). Similar
trends are found for the volumetric capacity (see Supplementary
material).
Fig. 9a shows the maximum attainable gravimetric capacity as a
function of the mass fraction of Si in the composite for different
Fig. 8. a) Maximum attainable gravimetric capacity of Si based anodes as a function of the CRTH for different mass fractions of Si and graphite in the composite and b)
corresponding design parameters. Relevant parameters used for the calculations are listed at the top of the figures.
187
limits of swelling (CRTH = 2 C). Fig. 9b depicts the corresponding
design parameters. The maximum attainable gravimetric capacity
rises with increasing amount of Si in the composite and increasing
limit of swelling. The corresponding optimum design parameters
significantly vary with the mass fraction of Si and graphite,
respectively. The larger the amount of Si in the composite, the
thinner and more porous the electrode.
Fig. 10a compares maximum values of the gravimetric and
volumetric capacity as a function of the CRTH. The corresponding
design parameters are plotted in Fig. 10b. Again, the obtained CRTH
dependence characterizes the well-known tradeoff between
energy- and power-density, which is more pronounced in the
case of gravimetric capacity. The corresponding design parameters,
leading to maximum gravimetric and volumetric capacity, differ
significantly. Electrodes designed for maximum gravimetric
capacity are thicker and more porous compared to electrodes
design for maximum volumetric capacity. By implication, electro-
des intended for maximum gravimetric capacity have inferior
volumetric capacity and vice versa. The magnitude of this tradeoff
decreases with increasing amount of graphite in the composite,
however, at the cost of both gravimetric and volumetric energy
density.
7. Discussion
Table 2 shows optimum values of the gravimetric and
volumetric capacity of Si based electrodes (CRTH = 2 C) in
comparison to the intrinsic properties of Si and graphite. Both
the gravimetric and the volumetric capacity of Si based electrodes
are dramatically reduced in comparison to the properties of pure
Si. Considering the design criteria derived above, the maximum
values of Qm and QV of Si based anodes amount to approximately
11 % and 14 % of the respective properties of pure silicon. In the case
of graphite electrodes (vSi = 0 %), considering the same design
188 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190

Fig. 9. a) Maximum attainable gravimetric capacity of Si based anodes as a function of the mass fractions of Si and graphite in the composite and b) corresponding design
parameters. Relevant parameters used for the calculations are listed at the top of the figures.

Fig. 10. Comparison of a) maximum gravimetric and volumetric capacity as a function of the CRTH and b) corresponding design parameters. Relevant parameters used for the
calculations are listed at the top of the figures.

criteria, the maximum values of Qm and QV amount to approxi-
mately 44 % and 56 % of the respective properties of graphite.
As mentioned above, literature often reports outstanding
properties of novel Si based electrodes in terms of specific

1
capacity, e.g. in mAh gSi , related to the mass of active material
used. However, from an engineering point of view such statements
lack on information regarding the gravimetric and volumetric
energy density of the resulting electrodes, which are of crucial
importance for the operation of electric or hybrid-electric vehicles.
Our calculations reveal that the maximum attainable capacity of Si
based electrodes is substantially lower compared to the intrinsic
properties of Si when the total mass/volume as well as the volume
 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190 189

Table 2 electrode’s porosity during charging. Furthermore, a “C-rate

Maximum values of gravimetric and volumetric capacity of Si based electrodes
threshold” is defined, guaranteeing diffusion limited currents
capable to manage a 2.0 C charging rate and comparison to the intrinsic properties
of silicon and graphite. not falling below a desired C-rate. The derived relations are used to
optimize the porosity and the electrode thickness regarding
gravimetric volumetric
maximum gravimetric and volumetric capacity, depending on the
Qm [Ah kg
1] L [mm] e [%] QV [Ah kg
1] L [mm] e [%] composition (graphite blending) and accepted volume expansion
Silicon 4200 – – 9800 – – (swelling) of the electrode. It turns out, that the enormous volume
vSi ¼ 95 % 468 60 84 1418 28 79 expansion during lithiation significantly narrows the design
vSi ¼ 55 % 415 63 77 1259 29 69 freedom, attainable capacity and rate capability. The maximum
vSi ¼ 15 % 284 72 58 858 33 45
gravimetric and volumetric capacity of a Si based electrode,
vSi ¼ 0 % 161 87 39 479 39 22
Graphite 370 – – 850 – –
capable to manage a 2.0 C charging rate, is determined to
468 Ah kg
1 and 1418 Ah L
1, respectively. These values are clearly
inferior to the intrinsic properties of Si but still significantly larger
expansion of the electrode is taken into account. Nevertheless, the compared to state of the art graphite anodes.
maximum values of gravimetric and volumetric capacity of Si
based electrodes are clearly superior to state of the art graphite
Acknowledgment
electrodes. According to our calculations, a maximum improve-
ment of approximately 300 % of gravimetric and volumetric
This work has been funded by the German Federal Ministry of
capacity can be achieved, respectively.
Education and Research BMBF through the Excellent – Battery
Our analyses reveal that highest gravimetric and volumetric
WING center “Batteries – Mobility in Saxony” (Grant No. KZ
capacities of Si based electrodes are obtained for the maximum
03X4637 A and Grant No. KZ 03X4637 B).
amount of Si in the composite. However, it also turns out that there
exist several drawbacks accompanied with high mass fractions of
Si: Appendix A. Supplementary data

i The technological implementation of the corresponding opti-
mum design parameters is challenging, especially regarding
porosities > 85 % (Fig. 10). A promising approach might be the
application of highly porous 3D current collectors decorated
with nanostructured Si.
ii High mass fractions of Si in the composite are accompanied
with a pronounced tradeoff between energy- and power-
density. Electrodes intended for high-energy applications have
low power density and vice versa. However, both energy and
power density are still significantly larger compared to state of
the art graphite electrodes.
iii In the case of high mass fractions of Si in the composite, the
electrode parameters for maximum gravimetric and volumetric
capacity significantly differ (Fig. 10). Consequently, electrodes
designed for maximum gravimetric capacity have inferior
volumetric capacity and vice versa. The magnitude of this
tradeoff decreases with increasing amount of graphite in the
composite, however, at the cost of both gravimetric and
volumetric energy density.
It should be noted, that the presented derivation of design
parameters for maximum gravimetric and volumetric energy
capacity assumes impeccable electrochemical performance of the
electrodes. However, as mentioned above, currently several
challenges exist, regarding the practical limits of Si based
electrodes related to the loss of electrical contact, disruption
and subsequent recreation of the solid electrolyte interphase as
well as fracture and pulverization of Si particles, leading to
degradation of the electrochemical performance and capacity fade.
However, if these problems are solved, the derived design
parameters will yield optimum values of gravimetric and
volumetric energy density. In this sense, the derived quantities
represent the theoretical limits of Si based electrodes.
8. Conclusion
We studied design criteria and theoretical limits of Si based
anodes taking into account the volume expansion during lithiation.
A “deformation threshold” is defined, signifying the minimum
value of the initial porosity that has to be adjusted to prevent
deformation of the active material and dramatic reduction of the
Supplementary data associated with this article can be found, in
the online version, at https://doi.org/10.1016/j.est.2017.11.009.
References
[1] X. Su, Q. Wu, J. Li, X. Xiao, A. Lott, W. Lu, B.W. Sheldon, J. Wu, Silicon-Based
nanomaterials for lithium-ion batteries, Adv. Energy Mater. 4 (2014) 1300882.
[2] M. Ashuri, Q. He, L.L. Shaw, Silicon as a potential anode material for Li-ion
batteries: where size, geometry and structure matter, Nanoscale 8 (2016) 74–
103.
[3] X. Zuo, J. Zhu, P. Müller-Buschbaum, Y.-J. Cheng, Silicon based lithium-ion
battery anodes, Nano Energy 31 (2017) 113–143.
[4] M.N. Obrovac, V.L. Chevrier, Alloy negative electrodes for Li-ion batteries,
Chem. Rev. 114 (2014) 11444–11502.
[5] Z. Du, R.A. Dunlap, M.N. Obrovac, High energy density calendered Si Alloy/
Graphite anodes, J. Electrochem. Soc. 161 (2014) A1698–A1705.
[6] J.H. Ryu, J.W. Kim, Y.-E. Sung, S.M. Oh, Failure modes of silicon powder negative
electrode in lithium secondary batteries, Electrochem. Solid-State Lett. 7
(2004) A306.
[7] H. Wu, Y. Cui, Designing nanostructured Si anodes for high energy lithium ion
batteries, Nano Today 7 (2012) 414–429.
[8] M.A. Rahman, G. Song, A.I. Bhatt, Y.C. Wong, C. Wen, Nanostructured silicon
anodes for high-performance lithium-ion batteries, Adv. Funct. Mater. 26
(2016) 647–678.
[9] M.T. McDowell, S.W. Lee, W.D. Nix, Y. Cui, 25th anniversary article:
understanding the lithiation of silicon and other alloying anodes for
lithium-ion batteries, Adv. Mater. (Deerfield Beach, Fla.) 25 (2013) 4966–4985.
[10] L.Y. Beaulieu, T.D. Hatchard, A. Bonakdarpour, M.D. Fleischauer, J.R. Dahn,
Reaction of Li with alloy thin films studied by in situ AFM, J. Electrochem. Soc.
150 (2003) A1457–A1464.
[11] K. Rhodes, N. Dudney, E. Lara-Curzio, C. Daniel, Understanding the degradation
of silicon electrodes for lithium-Ion batteries using acoustic emission, J.
Electrochem. Soc. 157 (2010) A1354.
[12] C.K. Chan, H. Peng, G. Liu, K. McIlwrath, X.F. Zhang, R.A. Huggins, Y. Cui, High-
performance lithium battery anodes using silicon nanowires, Nature Nanotech
3 (2008) 31–35.
[13] R. Teki, M.K. Datta, R. Krishnan, T.C. Parker, T.-M. Lu, P.N. Kumta, N. Koratkar,
Nanostructured silicon anodes for lithium ion rechargeable batteries, Small
(Weinheim an der Bergstrasse, Germany) 5 (2009) 2236–2242.
[14] J.R. Szczech, S. Jin, Nanostructured silicon for high capacity lithium battery
anodes, Energy Environ. Sci. 4 (2011) 56–72.
[15] W. He, H. Tian, F. Xin, W. Han, Scalable fabrication of micro-sized bulk porous Si
from Fe–Si alloy as a high performance anode for lithium-ion batteries, J.
Mater. Chem. A 3 (2015) 17956–17962.
[16] H. Tian, X. Tan, F. Xin, C. Wang, W. Han, Micro-sized nano-porous Si/C anodes
for lithium ion batteries, Nano Energy 11 (2015) 490–499.
[17] G. Wang, J. Ahn, J. Yao, S. Bewlay, H. Liu, Nanostructured Si?C composite anodes
for lithium-ion batteries, Electrochem. Commun. 6 (2004) 689–692.
[18] H. Dong, X.P. Ai, H.X. Yang, Carbon/Ba–Fe–Si alloy composite as high capacity
anode materials for Li-ion batteries, Electrochem. Commun. 5 (2003) 952–957.
[19] I.-S. Kim, P.N. Kumta, High capacity Si/C nanocomposite anodes for Li-ion
batteries, J. Power Sources 136 (2004) 145–149.
190 C. Heubner et al. / Journal of Energy Storage 15 (2018) 181–190
[20] M. Singh, J. Kaiser, H. Hahn, Thick electrodes for high energy lithium ion
batteries, J. Electrochem. Soc. 162 (2015) A1196–A1201.
[21] K.G. Gallagher, S.E. Trask, C. Bauer, T. Woehrle, S.F. Lux, M. Tschech, P. Lamp, B.J.
Polzin, S. Ha, B. Long, Q. Wu, W. Lu, D.W. Dees, A.N. Jansen, Optimizing areal
capacities through understanding the limitations of lithium-ion electrodes, J.
Electrochem. Soc. 163 (2015) A138–A149.
[22] L.O. Valo ̸en, J.N. Reimers, Transport properties of LiPF[sub 6]-Based Li-Ion
battery electrolytes, J. Electrochem. Soc. 152 (2005) A882.
[23] A. Hess, Q. Roode-Gutzmer, C. Heubner, M. Schneider, A. Michaelis, M. Bobeth,
G. Cuniberti, Determination of state of charge-dependent asymmetric Butler–
Volmer kinetics for LixCoO2 electrode using GITT measurements, J. Power
Sources 299 (2015) 156–161.
[24] R. Dash, S. Pannala, Theoretical limits of energy density in silicon-Carbon
composite anode based lithium ion batteries, Sci. Rep. 6 (2016) 27449.
[25] P.M. Biesheuvel, M.Z. Bazant, Nonlinear dynamics of capacitive charging and
desalination by porous electrodes, Phys. Rev. E Stat. Nonlin. Soft. Matter. Phys.
81 (2010) 31502.
[26] C. Heubner, U. Langklotz, M. Schneider, A. Michaelis, Analysis of the counter-
electrode potential in a 3-electrode lithium ion battery cell, J. Electroanal.
Chem. 759 (Part 2) (2015) 91–94.
[27] D.T. Gillespie, E. Seitaridou, Simple Brownian Diffusion: An Introduction to the
Standard Theoretical Models, Oxford University Press, Oxford, 2012.