Electrochimica Acta 140 (2014) 250–257

Contents lists available at ScienceDirect

Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta

Electron transfer through solid-electrolyte-interphase layers formed

on Si anodes of Li-ion batteries
L. Benitez c , D. Cristancho a,b , J.M. Seminario a,b,c,∗ , J.M. Martinez de la Hoz a,b ,
P.B. Balbuena a,b,∗
a
Department of Chemical Engineering Texas A&M University, College Station, TX 77843
b
Department of Materials Science and Engineering Texas A&M University, College Station, TX 77843
c
Department of Electrical and Computer Engineering Texas A&M University, College Station, TX 77843

a r t i c l e i n f o a b s t r a c t

Article history: Solid-electrolyte interphase (SEI) films are formed on the electrode surfaces due to aggregation of prod-
Received 26 December 2013 ucts of reduction or oxidation of the electrolyte. These films may grow to thicknesses in the order of
Received in revised form 2 May 2014 50-100 nm and contain a variety of organic and inorganic products but their structure is not well defined.
Accepted 3 May 2014
Although in some cases the films exert a passivating role, this is not always the case, and these phenomena
Available online 10 May 2014
are particularly more complex on Silicon anodes due to swelling and cracking of the electrode during
lithiation and delithiation. Since the driving force for SEI growth is electron transfer, it is important to
Keywords:
understand how electron transfer may keep occurring through the heterogeneous film once the bare
solid-electrolyte interphase
electron transfer
electron surface is covered. Here we introduce a novel approach for studying electron transfer through
density functional theory model films and show preliminary results for the analysis of electron transfer through model composite
ab initio molecular dynamics interfacial systems integrated by electrode/SEI layer/electrolyte. Ab initio molecular dynamics simula-
Green’s function tions are used to identify deposition of SEI components, and a density functional theory/Green’s function
approach is utilized for characterizing electron transfer. Three degrees of lithiation are modeled for the
electrodes, the SEI film is composed by LiF or Li2 O, and the ethylene carbonate reduction is studied. An
applied potential is used as driving force for the leakage current, which is evaluated as a function of the
applied potential. Comparative analyses are done for LiF and Li2 O model SEI layers.
© 2014 Elsevier Ltd. All rights reserved.

1. Introduction
Silicon has been proposed as a promising material for negative
electrode of lithium ion batteries. This is due to its much higher
capacity to store Li, which is about one order of magnitude higher
(4212 mAhg−1 )[1] than graphite (372 mAhg−1 ) [2]. Also, Si is an
abundant material that can form mixtures with up to 4.4 lithium
atoms/silicon atom (Li4.4 Si) [2]. However, Li4.4 Si expands 300% in
volume during lithiation leading to change in the surface area and
mechanical stresses[3] [4]. Typical electrolytes consist of mixtures
of alkyl carbonates such as ethylene carbonate (EC), diethyl carbon-
ate (DEC), dimethyl carbonate (DMC), propylene carbonate (PC),
and ethyl methyl carbonate (EMC) [2]. During discharge, lithium
ions flow from the anode to the cathode through the electrolyte
and the anode shrinks as the lithium ions are ejected. The opposite
∗ Corresponding authors.
E-mail addresses: seminario@tamu.edu (J.M. Seminario),
balbuena@tamu.edu (P.B. Balbuena).
process occurs during charge where the Si electrode changes vol-
ume swelling dramatically due to lithiation.[2] Starting during the
first cycle, the electrolyte decomposes electrochemically and a solid
electrolyte interphase (SEI) [5],[6] develops on the anode surface.
The SEI layer is formed by a variety of products which result from
several decomposition reactions. [7,8] Depending on its thickness
and on the type of products the SEI film may block electron transfer
between electrolyte and anode during charging, thus passivating
the electrode and preventing further SEI reactions. [9] However,
the film needs to be ionically conductive to avoid Li retention which
leads to irreversible capacity loss that affects power cyclability.[10]
Some additives have shown to be useful to develop a more stable
SEI and therefore to enhancing the cycle life in silicon electrodes.
[11] Fluoroethylene carbonate (FEC) is one of the most used sol-
vents that ameliorate the capacity retention of the battery.[12].
Several characterization studies[13–16] have found that SEI lay-
ers formed on silicon electrodes is alike to those formed on carbon
electrodes; however additional products may be generated on sil-
icon by conversion of silicon oxide (SiO2 ) to lithium silicon oxides
(Lin SiOm ) [10]. Most of the reported surface analyses[10,13–16]

http://dx.doi.org/10.1016/j.electacta.2014.05.018
0013-4686/© 2014 Elsevier Ltd. All rights reserved.
 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257
agree that the SEI layer may be formed by fluorinated species,
such as lithium fluoride (LiF), Lix PFy and PFy , as well as Li2 CO3 ,
alkyl carbonates, ethers, phosphates, and Li2 O, depending on the
cycling stage, electrolyte composition, but also electrode nano-
structure and morphology.[15] Additionally, multiscale modeling
has been used to characterize the initial stages of SEI formation.
Quantum chemical studies[7,17–19] include detailed density func-
tional theory (DFT) analyses of solvent decomposition mechanisms
in Li/solvent complexes, as well as ab initio molecular dynam-
ics including the description of the surface chemistry[8,20,21].
However, much less information is available regarding the growth
SEI stages, where issues such as aggregation, polymerization, and
nucleation of the heterogeneous SEI film are critical. An important
and not-yet answered question is how the electron transfer pro-
ceeds through such film that contains several insulator species. For
example, both LiF and Li2 O are wide-gap electrical insulators[22],
with reported bulk experimental band gaps of 14.2 eV[23] and 7.99
eV[24] respectively. Although some research has focused on ionic
conduction[22], much less is known about electronic conduction
through SEI blocks containing such species, although we note that
interfacial capacitances can also be obtained from first-principles
calculations, and have been reported by others.[25] In this work we
report a novel approach to evaluate electron transfer through com-
posite interfacial films. The reported atomistic simulations focus
on the chemistry of the SEI reactions and on the electron transfer
events during SEI growth. They are not designed to investigate the
swelling phenomena of Si anodes. Other coarse-grained models are
much better suited to tackle that problem, and we will address that
aspect in future reports.
Here first we briefly review and discuss the possible formation
of SEI products such as LiF and Li2 O. We then introduce our model
composite films formed by electrode/SEI/solvent species and eval-
uate the electron currents through such systems for a number of
interfacial configurations. The results of this study contribute new
insights to two important problems: 1) How the SEI layer is formed
on Si anodes; 2) How the SEI layer may grow based on the electronic
conductivity assessed in representative SEI model compounds (LiF
and Li2 O). The advantage of ab initio approaches is that they are
free of fitting parameters; therefore the results are such that can
provide a “theoretical microscope” additional tool to obtain insights
of phenomena in most cases not accessible to experiments.
Fig. 1. The studied model represents a piece of the composite interface integrated
by electrolyte/model SEI layer/model Lix Siy electrode. Both ends of the composite
interface are connected to nano-gold contacts, which are used to apply an external
applied potential (V). Such applied potential is the driving force for the leakage that
would occur during charge of the battery. The extent of electron transfer through the
system for various SEI layer compositions and degrees of lithiation of the electrode is
measured using the same nano-gold contacts. The red arrow indicates the direction
of electron flow during the SEI layer nucleation and growth.
251
2. Methodology
Fig. 1 illustrates the overall idea of the interfacial system we aim
to investigate. A composite interfacial system is built that is com-
posed by a model Lix Siy electrode covered by a multicomponent
model SEI layer which is in contact with an electrolyte layer.
Such electrolyte layer in principle may contain some solvents,
salt and/or additives. An applied potential is imposed between
two gold nanoelectrodes representing the leakage current that
would occur during battery charge. The constituent SEI molecules
that we discuss in this paper include LiF and Li2 O. The calcu-
lated systems are discussed below. In this section we explain the
methodology used to arrive at the formulation of the employed
models.
2.1. AIMD simulations
First, we determine the possible SEI layer constituents using
static quantum density functional theory (DFT) calculations[21]
and ab initio molecular dynamics (AIMD) simulations[8] of Lix Siy
surfaces covered by an electrolyte. In our previous reports we have
examined the decompositions of EC and FEC on representative
Lix Siy surfaces and clusters[8,21]. Here we report AIMD simula-
tions of the decomposition of the salt LiPF6 on Li13 Si4 slabs. As
done in our previous work[8], optimizations were performed using
the Vienna ab initio simulation package VASP,[26–30] with the
Perdew-Burke-Ernzerhof functional (GGA-PBE)[31] and the pro-
jector augmented wave (PAW) pseudopotentials provided in the
VASP databases describing electron-ion interactions.[32,33] The
plane wave was expanded up to a cutoff energy of 400 eV. The
convergence criteria for ionic relaxation loop and for electronic
self-consistent iteration were set to 10−3 and 10−4 eV respec-
tively. A Gaussian smearing with a width of 0.05 eV was employed
and a 4 × 4x1 k-points Monkhorst–Pack[34] mesh sampling was
used in the surface Brillouin zone. The AIMD simulations were
carried out on the optimized LiSi surfaces in contact with liquid-
phase EC (density = 1.32 g/cm3 ), using the NVT ensemble at 450 K
and a time step of 1 femtosecond. The Nose thermostat was
used to control the temperature oscillations during the simula-
tion with a Nose-mass parameter of 0.5, which gives a frequency
of oscillation corresponding to 176 time steps. A -point Bril-
louin zone sampling was applied in this case with a planewave
energy cutoff of 300 eV. In every case, EC molecules were placed
randomly in the liquid phase and a molecular dynamics minimiza-
tion was performed using the consistent valence force field (CVFF)
as implemented in the DISCOVER simulation software.[35] CVFF
is considered good for modeling small organic crystals and gas
phase structures.[36] For the minimization a smart algorithm was
used combining steepest descent, conjugate gradient, and New-
ton methods. The maximum force among all the atoms in the
system required for convergence was set to 0.001 kcal mol−1 Å−1 .
Two different initial configurations were used for each of the sur-
faces studied. Bader charge analysis was used to perform charge
calculations.[37,38] Within this method, the total electronic charge
of an atom is approximated by the charge enclosed within the
Bader volume defined by zero flux surfaces. The electrolyte uti-
lized in the AIMD simulations is composed by a 1.1 M solution of
LiPF6 and EC solvent molecules at their liquid density of 1.32 g/cm3 .
Additives were not tested in these simulations. These molecules
are represented by their full electronic structure, which gives
the correct dipole moment, and other higher order multipoles.
For this reason, although we have not calculated the electric
permittivity, we expect the experimental value to be accurately
reproduced.
252 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257

Fig. 2. AIMD simulations of a solution of one LiPF6 molecule and 32 EC molecules in contact with a Li13 Si4 surface. Left: Initial configuration. Center: After 2.83 ps the salt
anion is completely decomposed. Right: At 12.4 ps one EC molecule has been reduced (see text). Color code: Li: purple; Si: yellow; O: red; H: white; C: grey; P: orange; F:
light blue.

2.2. Electron transfer analyses Where e is the charge of the electron, h is the Planck constant,
T(E,V) is the transmission function, E is the energy and Ef is the
Silicon molecular clusters (Si22 ) were optimized until local min- Fermi level of the molecular junction.
ima using the B3PW91/6-31G(d) level of theory, [39] and the The transmission function is calculated using the following
3-21G basis set was used for LiSi (Li11 Si11 ) and Li22 cluster struc- expression (see for instance [53]):
tures. For the SEI model, LiF and Li2 O molecular geometries were +
T (E, V ) = Tr(T1 (E, V )GM (E, V )T2 (E, V )GM (E, V )) (2)
obtained from the Inorganic Crystal Structure Database (ICSD)
(http://icsd.fiz-karlsruhe.de/icsd/) and located between the model − +
where Tk = i( k − k
) (k = 1, 2) is the coupling between the elec-
anode and one EC molecule. Gold atoms representing the con- trodes and the molecule, GM is the retarded Green function and GM +
tacts of the I-V measurement device were attached at each side is its adjoint. GM is obtained using:
of the composite interfacial system. We note that the addition
of these gold atoms as well as the information of the bulk Au GM (E, V )
density of states is important since this is a nano-scale system    −1
where the presence of the materials of the measurement instru- = ESMM (E, V ) − HMM (E, V ) − (E, V ) − (E, V ) (3)
1 1
ment must be included. Once the complete interfacial system is
assembled, a single point self-consistent field (SCF) calculation is 
performed to obtain the Hamiltonian matrix elements needed to k
(E, V ) = HMk gk (E)HkM for k = 1,2 is the self-energy from the
compute the current as explained below. The optimizations and contacts. SMM and HMM are the overlap and Hamiltonian sub-
single point SCF calculations were carried out using the program matrices corresponding to the composite interfacial system; they
Gaussian 09[40]. Next, current-applied potential (I-V) characteris- are obtained from the global Schrödinger equation
tics were obtained using the Green’s function formalism with the H(V )ϕ(V ) = (V )S(V )ϕ(V ) (4)
help of the generalized electron nanointerface program (GENIP).
[41] This computational method has been used before to obtain once the wave function ϕ(V) is obtained in a SCF manner.
I-V characteristics of polypeptides in alpha-helix conformation,
[42] nanosensors for fissile materials,[43] cobalt phthalocyanine 3. Results and discussion
complexes, [44] among others. The density of states of bulk gold
is calculated using the CRYSTAL program[45–47]. The current- 3.1. SEI nucleation on lithiated Si anodes
applied potential (I-V) curves are calculated using the transmission
function T(E,V) using the Landauer equation [48–52] In previous work we have reported reduction mechanisms of EC
in contact with Si anodes at low [21], as well as intermediate and
 Ef +V2 high degrees of lithiation [8]. The analyses were done examining the
2e
I≈ T (E, V )dE (1) effects of different exposed surface facets and various surface ter-
h Ef +V1 minations (O, OH, and H) of the low-saturated Si surface sites. It was

Fig. 3. Details of the reduction process highlighted in Fig. 2. Left: an intact EC molecule becomes close to an adsorbed CO−2 radical anion, product of the reduction of another
EC molecule. Center: The intact EC molecule becomes nucleophilically attacked by the adsorbed radical anion CO−2 forming a CC bond. Right: The attacked EC molecule
becomes reduced and its ring is opened (see Fig. 4). Color code: Li: purple; Si: yellow; O: red; H: white; C: grey; P: orange; F: light blue.
 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257 253

Fig. 4. Ring opening due to a nucleophilic attack. Left: The EC molecule shown in Fig. 3 (right) is attacked by the adsorbed radical anion CO−2 forming a CC (C31-C13) bond.
C31, the carbon atom of the carbonyl atom of the attacked molecule bonds with the C13 attached to the surface, while the C31-O33 bond of the attacked molecule is broken.
Right: Time evolution of the two bonds involved in the reduction of the EC molecule. Color code: Li: purple; Si: yellow; O: red; H: white; C: grey; P: orange; F: light blue.

Fig. 5. Current-potential curves obtained for electron transport through Li2 O (red),
(Li2 O)2 (green), and (Li2 O)3 (blue) in two different configurations between the
EC molecule and the model Si electrode. Top: Units attached via chemical bonds.
Bottom: Units separated by Van der Waals distances. The inset includes the current-
potential result (black symbols) obtained for the system EC/Si electrode. Note the
change of scale of the inset.
Fig. 6. Current-potential curves obtained for electron transport through Li2 O (red),
(Li2 O)2 (green), and (Li2 O)3 (blue) in two different configurations between the EC
molecule and the model LiSi electrode. Top: Units attached via chemical bonds.
Bottom: Units separated by Van der Waals distances. The inset includes the current-
potential result (black symbols) obtained for the system EC/LiSi electrode. Note the
change of scale of the inset.

concluded that both one and two-electron mechanisms are possible
at low degrees of lithiation, whereas higher lithiated surfaces may
induce multiple reduction mechanisms[8]. The initial decomposi-
tion products include gases: C2 H4 , CO, CO2 , and numerous organic
fragments such as O(C2 H4 )OCO2- , EC- , CO3 2- , CO2- , O(C2 H4 )O2− .
Some of these initial decomposition products are adsorbed on the
surface and may further react through aggregation or polymeriza-
tion reactions starting nucleation of various species and formation
of a mosaic film. Similar species are derived from the other linear
(DMC, DEC) and cyclic (i.e. PC) solvents and additives such as VC or
FEC.
Another essential electrolyte component is the salt. One of the
most used salt components is LiPF6 whose decomposition yields
one of the most important SEI products, LiF. To understand its
decomposition process, we carried out AIMD simulations contain-
ing 1 LiPF6 molecule and 11 EC molecules; this corresponds to
approximately 1.1 M solution of LiPF6. The salt concentration was
chosen based on typical values used in Li-ion cells. These val-
ues are optimized to maximize ionic conduction.[54] Higher salt
254 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257
Fig. 7. Current-potential curves obtained for electron transport through Li2 O (red),
(Li2 O)2 (green), and (Li2 O)3 (blue) in two different configurations between the
EC molecule and the model Li electrode. Top: Units attached via chemical bonds.
Bottom: Units separated by Van der Waals distances. The inset includes the current-
potential result (black symbols) obtained for the system EC/Li electrode. Note the
change of scale of the inset.
concentrations would result in less salt reduction but also would
decrease ionic conduction. In contact with the Li13 Si4 surface, the
salt molecule becomes promptly dissociated and the PF5 anion
reduction is detected after 2.83 ps. The reduction leads to a total
decomposition of the anion into its atomic components: the F atoms
get closer to the surface and pair with lithium ions as illustrated in
Fig. 2. Typical LiF distances are between 0.172 and 0.2 nm, suggest-
ing the initial nucleation of LiF bonds. Continuation of the AIMD
simulation allows observation of the next steps. Within the follow-
ing 10 ps, we didn’t observe any EC reduction happening over the
surface section containing the LiF fragments. However, the area that
was not covered by LiF kept reducing EC molecules and some of the
EC decomposition fragments described above appeared readily on
the surface. In particular, Fig. 2 (right) illustrates the adsorption of
an opened EC molecule over a CO fragment adsorbed on the surface.
The EC reduction process is depicted in Fig. 3. Initially an EC
molecule (not shown) was reduced on the surface leaving as one
of the decomposition products a CO−2 fragment. This adsorbed
anion nucleophilically attacks an intact EC molecule forming a CC
bond, and the open molecule becomes adsorbed to the CO fragment
attached to the surface.
The ring opening is also followed by the AIMD simulations. Fig. 4
illustrates the steps of CO breaking (ring opening) and CC bond
formation occurring almost simultaneously as shown by the graph
that follows the CO (red) and CC (blue) distances as functions of
time. The onset of the CC bond forming is manifested by a constant
decrease of the CC distance, and∼ 300 fs before the bond is formed
there is a sharp decrease of the CC distance. At the end of such
Fig. 8. Current-potential curves obtained for electron transport through LiF (red),
(LiF)2 (green), and (LiF)3 (blue) in two different configurations between the EC
molecule and the model Si electrode. Top: LiF units attached via chemical bonds
forming a linear chain. Bottom: LiF units separated by Van der Waals distances. The
inset includes the current-potential result (black symbols) obtained for the system
EC/Si electrode. Note the change of scale of the inset.
decrease the CC bond between the C of the carbonyl group and the
adsorbed C is formed, and at the same time the CO bond from the
EC molecule starts to break and completely breaks in the following
∼300 fs.
Another inorganic SEI product is Li2 O, which may result from
reduction of SiO2 present after exposure of the Si electrode to air
during their fabrication, developing thin films of ∼1-3 nm on silicon
nanowires[55]. However, after the electrodes covered by the thin
film oxide are inserted in the electrochemical cell, the native oxide
is basically removed during the first cycles, which is attributed to
either electrochemical reaction 4Li + SiO2 →2 Li2 O + Si, or to chemi-
cal reaction by HF attack. HF can exist as a result of reaction of LiPF6
with traces of water present as impurities[15].
3.2. Electron transfer through model SEI layers
SEI layers were represented by a cluster modeled by one compo-
nent (Li2 O or LiF). Electrodes are represented by Si22 , Si11 Li11 , and
Li22 clusters respectively. Calculations were done for various thick-
nesses and configurations of the SEI layer. Gold tips were attached
at each end of the interfacial system and a bias applied potential
was used to evaluate electron transfer using a DFT/Green’s function
method. Optimized cluster structures were located in an interfacial
arrangement together with model SEI layers represented by Li2 O,
(Li2 O)2 , and (Li2 O)3 , and by LiF, (LiF)2 and (LiF)3 . In separate con-
figurations, the units of each system (Li2 O and LiF) were connected
forming chemical bonds or separated at Van der Waals distances
in order to simulate initial nucleation stages. These configurations
were based on the AIMD observations (Fig. 2) showing that the
 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257
Fig. 9. Current-potential curves obtained for electron transport through LiF (red),
(LiF)2 (green), and (LiF)3 (blue) in two different configurations between the EC
molecule and the model LiSi electrode. Top: LiF units attached via chemical bonds
forming a linear chain. Bottom: LiF units separated by Van der Waals distances. The
inset includes the current-potential result (black symbols) obtained for the system
EC/LiSi electrode. Note the change of scale of the inset.
initial stages of LiF formation do not contain a LiF crystal; instead
LiF fragments are detected. In all cases the current-potential curves
were computed in a range of applied potential from -5.5 to 5.5 V.
Due to the geometry of the system, the results are almost symmet-
ric. However, on the basis of our model (Fig. 1), the electron flow of
interest corresponds to the positive applied potential values.
Figs. 5–7 illustrate the effect of the presence of intercalated Li2 O
units on the transfer of electrons from the electrode structure to
the EC molecule. First, we can recognize the effect of the nature
of the electrode by comparing the magnitude of the current pass-
ing through the various systems. The maximum current for the Si
electrode in the presence of the attached SEI fragment is ∼1.5 ␮A
(Fig. 5), whereas it reaches 12 and 15 ␮A for LiSi and Li electrodes
respectively (Figs. 6 and 7) at the highest potentials (∼4 V). Note
also how these currents are significantly reduced with respect to
those in absence of the SEI fragment (black curves in each figure).
Secondly, each of the figures illustrates the effect of the SEI con-
figuration (bonded vs. disconnected SEI molecules). In the three
cases the non-bonded SEI configurations yield lower currents i.e.
higher resistance to electron transfer than the bonded cases. Finally,
increasing the width of the SEI layer also increases the resistance
to electron transfer as observed in the three model electrodes.
Similarly, Figs. 8–10 illustrate the effect of the presence of LiF
units on the transfer of electrons from the electrode structure to
the EC molecule. Again the effect of the nature of the electrode is
revealed by comparison of the maximum currents: ∼0.4 ␮A for the
Si electrode (Fig. 8), that contrasts with 1.9 and 3.2 ␮A for LiSi and
Li electrodes respectively (Figs. 9 and 10) obtained at the highest
255
Fig. 10. Current-potential curves obtained for electron transport through LiF (red),
(LiF)2 (green), and (LiF)3 (blue) in two different configurations between the EC
molecule and the model Li electrode. Top: LiF units attached via chemical bonds
forming a linear chain. Bottom: LiF units separated by Van der Waals distances. The
inset includes the current-potential result (black symbols) obtained for the system
EC/Li electrode. Note the change of scale of the inset.
applied potentials (∼ 4 V). Note that these values are much smaller
than those shown in Figs. 5–7, which demonstrates that at this
high applied potential the LiF system introduces a much higher
resistance to electron transfer compared to Li2 O. At lower applied
potentials the currents are comparable, suggesting a similar resis-
tance offered by Li2 O and LiF. By comparison of the current
maxima obtained at the highest applied potential for each electrode
type and different SEI configurations (bonded vs. non-bonded SEI
molecules) it is again concluded that the non-bonded SEI configura-
tions yield lower currents i.e. higher resistance to electron transfer
than the bonded cases. Also, increasing the thickness of the linear
LiF chain in most cases increases the resistance to electron transfer
as observed in the three model electrodes.
One of the goals of this study is to determine the mechanism
of electron transfer that allows continuous SEI layer growth. In all
the results shown in Figs. 5–10, a common feature is that the pres-
ence of the model SEI layer creates a clear resistance to electron
transfer, as observed by looking at the difference in magnitude of
the currents in the insets where the black curve corresponds to
the system without SEI layer. This result agrees with the gener-
ally accepted idea that the SEI layer blocks electron transfer [56].
In order to test how the SEI thickness would affect the transfer of
electrons we computed the electron current through the systems
as a function of SEI thickness. Fig. 11 shows the results for Li2 O
and LiF evaluated at a constant applied potential of 2 V. In both
cases, the results were fitted to a function of the form f (x) = p1 .exp
(-x/p2 ) +p3 + p4 x, where x is the SEI thickness in Å, and the p4 coeffi-
cient was found ∼0. For the Si electrode, the current decays to zero
256 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257
Fig. 11. Calculated current (in ␮A) at 2 V as a function of the SEI layer thickness (LiF and Li2 O), measured in Å. The calculated points (shown as symbols) are well correlated
with an exponential function. The colors of the curves correspond to those used in the schemes on top describing each system.
very rapidly, at distances in the order of 3 Å (p3 coefficient ∼0). In
contrast, for the lithiated Si electrodes, finite although much lower
currents than those found in absence of the SEI layer, are kept up
to distances in the order of 10 Å suggesting that electron transfer
may occur through the composite interface. Although this distance
(1 nm) is still small compared to typical SEI thicknesses (order of
20-100 nm), the result suggests that a lower rate steady current
may continue through relatively longer distances.
4. Conclusions
A novel first-principles computational approach is used to
characterize the electron transfer through two model composite
interfacial systems: electrode/LiF/EC and electrode/Li2 O/EC. Elec-
trodes are modeled through three degrees of lithiation: Si, LiSi, and
Li. AIMD simulations demonstrate the rapid formation of LiF on the
surface of lithiated Si anodes having an irregular structure in the
initial stages of nucleation, and show the possible continuation of
the EC reduction on parts of the surface that are not covered by LiF.
Based on the structure obtained from the AIMD simulations, model
SEI layers made of LiF fragments are constructed in various thick-
nesses and configurations, and by analogy, Li2 O model films are also
built. It is found that at the same thicknesses and similar configura-
tions, the LiF films offer a higher resistance to electron transfer than
Li2 O ones. In each case, the resistance of the films is also larger if the
fragments are separated at Van der Waals distances. As expected,
increasing the thickness increases the film resistance; however,
on the lithiated electrodes covered by SEI film thicknesses in the
order of 1 nm there is still a finite although small current. Thus, our
analysis proves useful for testing electron transfer through various
SEI configurations. Future reports will address the effect of the SEI
cluster geometry and composition, and those of the interfaces SEI
film/electrode and SEI/electrolyte.
Acknowledgements
This work was supported by the Assistant Secretary for
Energy Efficiency and Renewable Energy, Office of Vehicle Tech-
nologies of the U.S. Department of Energy under Contract No.
DE-AC02-05CH11231, Subcontract No. 7060634 under the Batter-
ies for Advanced Transportation Technologies (BATT) Program.
References
[1] Y. Yin, L.G.Y. Wan, Silicon-based nanomaterials for lithium-ion batteries, Chi-
nese Science Bulletin 57 (2012) 32.
[2] J.R. Szczech, S. Jin, Nanostructured silicon for high capacity lithium battery
anodes, Energy & Environmental Science 4 (2011) 56.
[3] M.J. Chon, V.A. Sethuraman, A. McCormick, V. Srinivasan, P.R. Guduru, Real-
Time Measurement of Stress and Damage Evolution during Initial Lithiation of
Crystalline Silicon, Physical Review Letters 107 (2011) 045503.
[4] R. Benedek, M.M. Thackeray, Lithium reactions with intermetallic-compound
electrodes, Journal of Power Sources 110 (2002) 406.
[5] C. Becker, K. Strawhecker, Q.L.C. McAllister, In Situ Atomic Force Microscopy of
Lithiation and Delithiation of Silicon Nanostructures for Lithium Ion Batteries,
ACS Nano 7 (2013) 9173.
[6] P.B. Balbuena, Y.X. Wang, Lithium Ion Batteries, Solid Electrolyte Interphase,
London, 2004.
[7] Y. Wang, S. Nakamura, M. Ue, P.B. Balbuena, Theoretical studies to understand
surface chemistry on carbon anodes for lithium-ion batteries: Reduction mech-
anisms of ethylene carbonate, J. Am. Chem. Soc. 123 (2001) 11708.
[8] J.M. Martinez de la Hoz, K. Leung, P.B. Balbuena, Reduction Mechanisms of
Ethylene Carbonate on Si Anodes: Effects of Degree of Lithiation and Nature of
Exposed Surface, ACS Applied Materials & Interfaces 5 (2013) 13457.
[9] M. Nie, D.P. Abraham, Y. Chen, A. Bose, B.L. Lucht, Silicon Solid Electrolyte
Interphase (SEI) of Lithium Ion Battery Characterized by Microscopy and Spec-
troscopy, Journal of Physical Chemistry C 117 (2013) 13403.
[10] B. Philippe, R. Dedryvere, J. Allouche, F. Lindgren, M. Gorgoi, H. Rensmo, D.
Gonbeau, K. Edstrom, Nanosilicon Electrodes for Lithium-Ion Batteries: Inter-
facial Mechanisms Studied by Hard and Soft X-ray Photoelectron Spectroscopy,
Chemistry of Materials 24 (2012) 1107.
[11] M. Nie, D. Chalasani, D.P. Abraham, Y. Chen, A. Bose, B.L. Lucht, Lithium Ion
Battery Graphite Solid Electrolyte Interphase Revealed by Microscopy and
Spectroscopy, Journal of Physical Chemistry C 117 (2013) 1257.
[12] R. Elazari, G. Salitra, G. Gershinsky, A. Garsuch, A. Panchenko, D. Aurbach, Li Ion
Cells Comprising Lithiated Columnar Silicon Film Anodes, TiS2 Cathodes and
Fluoroethyene Carbonate (FEC) as a Critically Important Component, Journal of
the Electrochemical Society 159 (2012) A1440.
[13] C. Pereira-Nabais, J. Swiatowska, A. Chagnes, F. Ozanam, A. Gohier, P. Tran-Van,
C.S. Cojocaru, M. Cassir, P. Marcus, Interphase chemistry of Si electrodes used
as anodes in Li-ion batteries, Applied Surface Science 266 (2013) 5.
[14] B. Philippe, R. Dedryvere, M. Gorgoi, H. Rensmo, D. Gonbeau, K. Edstrom, Role of
the LiPF6 Salt for the Long-Term Stability of Silicon Electrodes in Li-Ion Batteries
- A Photoelectron Spectroscopy Study, Chemistry of Materials 25 (2013) 394.
[15] C.K. Chan, R. Ruffo, S.S. Hong, Y. Cui, Surface chemistry and morphology of the
solid electrolyte interphase on silicon nanowire lithium-ion battery anodes,
Journal of Power Sources 189 (2009) 1132.
 L. Benitez et al. / Electrochimica Acta 140 (2014) 250–257
[16] D.E. Arreaga-Salas, A.K. Sra, K. Roodenko, Y.J. Chabal, C.L. Hinkle, Progression
of Solid Electrolyte Interphase Formation on Hydrogenated Amorphous Silicon
Anodes for Lithium-Ion Batteries, Journal of Physical Chemistry C 116 (2012)
9072.
[17] Y. Wang, S. Nakamura, K. Tasaki, P.B. Balbuena, Theoretical studies to under-
stand surface chemistry on carbon anodes for lithium-ion batteries: How does
vinylene carbonate play its role as an electrolyte additive? J. Am. Chem. Soc.
124 (2002) 4408.
[18] K. Leung, Two-electron reduction of ethylene carbonate: A quantum chemistry
re-examination of mechanisms, Chemical Physics Letters 568 (2013) 1.
[19] J.M. Yu, P.B. Balbuena, J. Budzien, K. Leung, Hybrid DFT Functional-Based Static
and Molecular Dynamics Studies of Excess Electron in Liquid Ethylene Carbon-
ate, Journal of the Electrochemical Society 158 (2011) A400.
[20] K. Leung, J.L. Budzien, Ab initio molecular dynamics simulations of the initial
stages of solid-electrolyte interphase formation on lithium ion battery graphitic
anodes, Physical Chemistry Chemical Physics 12 (2010) 6583.
[21] K. Leung, S.B. Rempe, M.E. Foster, Y. Ma, J.M. Martinez-delaHoz, N. Sai, P.B. Bal-
buena, Modeling Electrochemical Decomposition of Fluoroethylene Carbonate
on Silicon Anode Surfaces in Lithium Ion Batteries, J. Electrochem. Soc. 161
(2014) A213.
[22] Y.C. Chen, C.Y. Ouyang, L.J. Song, Z.L. Sun, Electrical and Lithium Ion Dynam-
ics in Three Main Components of Solid Electrolyte Interphase from Density
Functional Theory Study, Journal of Physical Chemistry C 115 (2011) 7044.
[23] F. Tran, P. Blaha, Accurate Band Gaps of Semiconductors and Insulators with a
Semilocal Exchange-Correlation Potential, Physical Review Letters 102 (2009)
226401.
[24] Y. Ishii, J. Murakami, M. Itoh, Optical spectra of excitons in lithium oxide, Journal
of the Physical Society of Japan 68 (1999) 696.
[25] B. Lee, C.K. Lee, S. Han, J. Lee, C.S. Hwang, First-principles calculation of capac-
itance including interfacial effects, J. Appl. Phys. 103 (2008) 024106.
[26] G. Kresse, J. Furthmüller, Efficient iterative schemes for ab initio total-energy
calculations using a plane-wave basis set, Physical Review B 54 (1996) 11169.
[27] G. Kresse, J. Furthmuller, Efficiency of ab-initio total energy calculations for
metals and semiconductors using a plane-wave basis set, Computational Mate-
rials Science 6 (1996) 15.
[28] G. Kresse, J. Hafner, Ab initio molecular dynamics for open-shell transition
metals, Physical Review B 48 (1993) 13115–13118.
[29] G. Kresse, J. Hafner, Ab initio molecular dynamics for liquid metals, Physical
Review B 47 (1993) 558.
[30] G. Kresse, J. Hafner, Ab initio molecular-dynamics simulation of the liquid-
metal–amorphous-semiconductor transition in germanium, Physical Review B
49 (1994) 14251.
[31] J.P. Perdew, K. Burke, M. Ernzerhof, Generalized gradient approximation made
simple, Phys. Rev. Lett. 77 (1996) 3865.
[32] P.E. Blöchl, Projector augmented-wave method, Physical Review B 50 (1994)
17953.
[33] G. Kresse, D. Joubert, From ultrasoft pseudopotentials to the projector
augmented-wave method, Physical Review B 59 (1999) 1758.
[34] H.J. Monkhorst, J.D. Pack, Special points for Brillouin-zone integrations, Physical
Review B 13 (1976) 5188.
[35] Materials Studio v5 5.0.0, Accelrys Software Inc, 2010.
[36] C. Song, W. Wu, A. Al-Ostaz, GeoCongress 2008, American Society of Civil Engi-
neers, 2008, pp. 1012–1019.
[37] G. Henkelman, A. Arnaldsson, H. Jónsson, A fast and robust algorithm for Bader
decomposition of charge density, Computational Materials Science 36 (2006)
354.
257
[38] E. Sanville, S.D. Kenny, R. Smith, G. Henkelman, Improved grid-based algorithm
for Bader charge allocation, Journal of Computational Chemistry 28 (2007)
899.
[39] K. Raghavachari, Perspective on Density functional thermochemistry. III. The
role of exact exchange - Becke AD (1993) J Chem Phys 98:5648-52, Theoretical
Chemistry Accounts, 103 (2000) 361.
[40] M.J.T. Frisch, G. W.; Schlegel, H. B.; Scuseria, G. E.;, M.A.C. Robb, J. R.; Scalmani,
G.; Barone, V.; Mennucci, B.;, G.A.N. Petersson, H.; Caricato, M.; Li, X.; Hratchian,
H. P.;, A.F.B. Izmaylov, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;, M.T. Ehara,
K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, Y.K. T.; Honda, O.; Nakai,
H.; Vreven, T.; Montgomery, J.; J. A., J.E.O. Peralta, F.; Bearpark, M.; Heyd, J. J.;
Brothers, E.; Kudin, V.N.K. K. N.; Staroverov, R.; Normand, J.; Raghavachari, K.;,
A.B. Rendell, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.J.K. N.; Millam,
M.; Knox, J. E.; Cross, J. B.; Bakken, V.;, C.J. Adamo, J.; Gomperts, R.; Stratmann,
R. E.; Yazyev, O.;, A.J.C. Austin, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;,
K.Z. Morokuma, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.;
Daniels, A. D.; O. Farkas, Foresman, J. B.; D , J.V.C. Ortiz, J.;, Fox, D. J. Gaussian 09,
revision A.02; CT, W, Gaussian, Inc., 2009.
[41] J.M. Seminario, L.M. Yan, Ab initio analysis of electron currents in thioalkanes,
International Journal of Quantum Chemistry 102 (2005) 711.
[42] D. Cristancho, J.M. Seminario, Polypeptides in alpha-helix conformation per-
form as diodes, Journal of Chemical Physics 132 (2010) 065102.
[43] N. Kumar, J.M. Seminario, Design of Nanosensors for Fissile Materials in Nuclear
Waste Water, Journal of Physical Chemistry C 117 (2013) 24033.
[44] G.I. Cardenas-Jiron, P. Leon-Plata, D. Cortes-Arriagada, J.M. Seminario, Electrical
Characteristics of Cobalt Phthalocyanine Complexes Adsorbed on Graphene,
Journal of Physical Chemistry C 115 (2011) 16052.
[45] R. Dovesi, M. Causa, R. Orlando, C. Roetti, V.R. Saunders, Abinitio Approach to
Molecular-Crystals - a Periodic Hartree-Fock Study of Crystalline Urea, J Chem
Phys 92 (1990) 7402.
[46] R. Dovesi, V.R. Saunders, C. Roetti, R. Orlando, C.M. Zicovich-Wilson, F. Pascale,
B. Civalleri, K. Doll, N.M. Harrison, I.J. Bush, P. D’Arco, M. Llunell, CRYSTAL06
User’s Manual, University of Torino, Torino, 2006.
[47] C. Roetti, in: C. Pisani (Ed.), Quantum-Mechanical Ab-initio Calculation
of the properties of Crystalline Materials, Springer-Verlag, Berlin, 1996,
pp. 125–137.
[48] M. Buttiker, Y. Imry, R. Landauer, S. Pinhas, Generalized many-channel con-
ductance formula with application to small rings, Physical Review B 31 (1985)
6207.
[49] R. Landauer, Spatial Variation of Currents and Fields Due to Localized Scatterers
in Metallic conduction, IBM J. Res. Dev. 1 (1957) 223.
[50] R. Landauer, Electrical resistance of disordered one-dimensional lattices, Philos.
Mag. 21 (1970) 863.
[51] R. Landauer, Comment on Lodder exact electromigration theory, Solid State
Commun 72 (1989) 867.
[52] R.W. Landauer, Irreversibility and Heat Generation in the Computing Process,
IBM J. Res. Dev. 5 (1961) 183–191.
[53] S. Datta, Electronic Transport in Mesoscopic Systems, Cambridge University
Press, Cambridge, UK, 1995.
[54] K. Xu, Nonaqueous liquid electrolytes for lithium-based rechargeable batteries,
Chemical Reviews 104 (2004) 4303.
[55] D.D.D. Ma, C.S. Lee, F.C.K. Au, S.Y. Tong, S.T. Lee, Small-diameter silicon nanowire
surfaces, Science 299 (2003) 1874.
[56] J. Christensen, J. Newman, A mathematical model for the lithium-ion negative
electrode solid electrolyte interphase, Journal of the Electrochemical Society
151 (2004) A1977.