Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Nanostructured silicon for lithium-ion battery anode
(USC Thesis Other)
Nanostructured silicon for lithium-ion battery anode
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
NANOSTRUCTURED SILICON FOR LITHIUM-ION BATTERY ANODE
By
Mingyuan Ge
_________________________________________________________________
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(MATERIALS SCIENCE)
MAY 2015
Copyright 2015 Mingyuan Ge
ii
Dedication
Dedicated to my beloved parents and my wife Qian Huang.
iii
Acknowledgments
First and foremost, I would like to express my sincere gratitude to my academic advisor
Prof. Chongwu Zhou, for his continuous support during my PhD study. His enthusiasm
and knowledge in interdisciplinary areas are the great inspiration to my research career.
Without his encouragement and dedication, I could not make the achievement.
My sincere thanks also go to my dissertation committees Prof. Edward Goo and Prof. Wu
Wei for their invaluable suggestions, as well as Prof. Steven Cronin and Prof. Andrew
Armani for serving on my qualifying exam committee.
I extend my gratitude to our collaborators Dr. Matthew Mecklenburg in the Center for
Electron Microscopy and Microanalysis in USC, Dr. Peter Ercius in National Center for
Electron Microscopy (NCEM) at LBNL.
Moreover, many thanks to the battery team: Dr. Jiepeng Rong, Dr. Xin Fang, Mr. Anyi
Zhang, Mr. Chenfei Shen, and Mr. Yihang Liu. Also thanks to the former and current
group members: Dr. Bilu Liu, Dr. Gang Liu, Dr. Chuan Wang, Dr. Alexandar Badmaev,
Dr. Hsiao-Kang Chang, Dr. Anuj Madaria and Dr. Yi Zhang, Dr. Jialu Zhange, Yue Fu,
Dr. Haitian Chen, Dr. Yuchi, Che, Xue Lin, Zhen Li, Jing Qiu, Shelley Wang, Dr.
Maoqing Yao, Nappadol Aroonyadet, Luyao Zhang, Pyojae Kim, Rebecca Lee, Ning
iv
Yang, Liang Chen, Hui Gui, Pattaramon Vuttipittayamongkol, Ahmad Abbas, Sen Cong,
Yu Cao, Fanqi Wu, Yuqiang Ma, Xuan Cao, Qinzhou Liu, and Yilin Huang. I really
enjoyed working with you.
Finally I take pride in dedicating the current PhD thesis work to my beloved parents and
my wife Qian Huang who have given me endless love and encouragement. I would not
have all these achievements today without you.
v
Contents
Dedication ....................................................................................................................................... ii
Acknowledgments .......................................................................................................................... iii
List of Tables ................................................................................................................................. vii
List of Figures .............................................................................................................................. viii
Abstract ........................................................................................................................................ xiv
Chapter 1 Introduction to lithium-ion battery ................................................................................ 1
1.1 Status of lithium-ion batteries...................................................................................... 1
1.2 Electrochemistry of lithium-ion batteries .................................................................... 3
1.3 Fundamental properties of silicon as high capacity anode .......................................... 8
1.4 Chapter reference ....................................................................................................... 12
Chapter 2 Porous doped silicon nanowire as a lithium-ion battery anode ................................... 14
2.1 Mechanic properties of silicon anode ........................................................................ 14
2.2 Simulation of porous silicon during lithiation ........................................................... 16
2.3 Preparation of porous silicon nanowire ..................................................................... 22
2.4 Electrochemical test and discussion .......................................................................... 24
2.5 Conclusion ................................................................................................................. 33
2.6 Chapter reference ....................................................................................................... 34
Chapter 3 Scalable production of porous silicon nanoparticles for Li-ion battery anodes ........... 36
3.1 Motivation ................................................................................................................. 36
3.2 Preparation of porous silicon nanoparticles ............................................................... 37
3.3 Electrochemical tests and discussion ......................................................................... 43
3.4 Conclusion ................................................................................................................. 50
3.5 Chapter reference ....................................................................................................... 51
vi
Chapter 4 Cost-efficient fabrication of porous silicon from bulk metallurgical silicon for Li-ion
battery anodes ................................................................................................................................ 52
4.1 Motivation ................................................................................................................. 52
4.2 Preparation of nanoporous silicon from metallurgical silicon ................................... 55
4.3 Characterization of nanoporous silicon particle ........................................................ 58
4.4 Mechanism of nanoporous silicon formation ............................................................ 60
4.5 Electrochemical tests and discussion ......................................................................... 66
4.6 Discussion and conclusion ........................................................................................ 71
4.7 Chapter reference ....................................................................................................... 75
Chapter 5 Structure evolution in dealloyed Li-Si
x
Ge
1-x
ternary system ....................................... 78
5.1 Motivation ................................................................................................................. 78
5.2 Introduction of dealloying ......................................................................................... 80
5.3 Preparation of lithiated-SiGe nanoparticles and characterization methods ............... 83
5.4 Results and discussion ............................................................................................... 86
5.5 Conclusion ............................................................................................................... 104
5.6 Chapter reference ..................................................................................................... 105
Chapter 6 Future work: towards LiSi-S full battery ................................................................... 107
6.1 Introduction ............................................................................................................. 107
6.2 Experiments and results ............................................................................................ 115
6.3 Conclusion ............................................................................................................... 121
6.4 Chapter reference ..................................................................................................... 122
Bibliography ................................................................................................................................ 124
vii
List of Tables
Table 1.1 Electrode potential and capacity of some cathode and anode materials.
Table 2.1 Parameters used in simulation.
Table 2.2 Cyclic performance of porous silicon anode.
Table 4.1 summery of the preparation method and battery performanceof porous silicon.
Table 6.1 Comparison of battery systems.
7
17
28
54
108
viii
List of Figures
Figure 1.1 Comparison of different battery technologies in terms of gravimetric
energy density and volumetric energy density.
Figure 1.2 Schematic illustration of a typical lithium-ion battery where graphite and
LiFePO
4
are used as anode and cathode, respectively.
Figure 1.3 The volume per mole of host atoms (Li) for silicon as a function of
lithium content for phases occurring in its binary phase diagrams with
lithium.
Figure 1.4 Charge-discharge voltage profiles obtained with a silicon powder anode.
Another problem associated with silicon anode lies on the intrinsic low
diffusion rate of lithium in silicon, which is around 2 x 10
-11
cm
2
s
-1
. The
low diffusion rate of Li in Si limits the usage of Si anode for high power
committed applications that need to be operated at high
charging/discharging rate.
Figure 1.5 Structure of silicon nanowire before (a) and after lithiation (b), and its
cyclic performance (c). Structure of silicon hollow sphere before (d) and
after lithiation (e), and its cyclic performance. (f). Structure of 3D silicon
(a-b), and its cyclic performance (i)
Figure 2.1 (a) Schematic diagram of a porous silicon structure. (b) One unit of the
porous structure used for theoretical simulation and analysis. (c) Pore
size before and after lithiation and (d) corresponding maximum stress at
fixed pore-to-pore distance (l =12 nm). (e) Pore sizebefore and after
lithiation and (f) corresponding maximum stress v.s. initial pore size at
fixed pore/edge ratio (r/l=1/3)
Figure 2.2 (a) SEM and (b) TEM images of porous Si nanowires etched with
0.02M AgNO
3
. (c) and (d) HRTEM image of a nanowire in (b). (e)
SAED pattern of a single porous silicon nanowire. (f) Pore size
distribution of porous Si nanowires etched with 0.02 M and 0.04 M
AgNO
3
, respectively.
Figure 2.3 Electrochemical performance of battery using porous silicon nanowires
as anode and lithium metal as current collector. (a) Charge/discharge
profile within a voltage window of 0.01-2V vs. Li+/Li for the 1st cycle
2
4
9
9
10
19
23
27
ix
at a current rate of 0.4 A/.g, and the 50th, 100th, and 200th cycles at 2
A/g. (b) Cyclic voltammetry curves of porous silicon nanowire electrode
for the 1st and 2nd cycles using a voltage window 0.01-2V at rate of
0.1mV/s. (c) Charge/discharge capacity and Coulombic efficiency of
porous silicon nanowire electrode at current rates of 0.6, 1.2, 2.4, 3.6,
4.8, and 9.6 A/g. (d) Charge/discharge capacity of porous silicon
nanowire electrode at current rates of 2A/g, 4A/g, and 18A/g for 250
cycles. (e) Charge/discharge capacity of a porous silicon nanowire
electrode at current rates of 2A/g (0.5 C) and 4A/g (1 C) with additional
2000 cycles
Figure 2.4 TEM images of silicon nanowires before (a) and after (b) lithiation after
10 cycles at a current rate of 0.4 A/g. (c) Enlarged TEM image of (b)
showing the amorphous silicon structure. (d) Selected area electron
diffraction pattern showing black spots in (b) are crystalline silicon.
Figure 3.1 (a) Schematic diagram of the procedure to prepare porous silicon
nanoparticles. (b) Photographs of nonporous and porous Si
nanoparticles, illustrating the scalable nature of our porous silicon
nanoparticle preparation.
Figure 3.2 (a) TEM images of silicon nanoparticles after boron doping. (b) TEM
image showing porous silicon nanoparticles with large Ag nanoparticles
after electroless etching. (c) TEM image of porous Si nanoparticles after
washing with HNO
3
and H
2
O to remove Ag nanoparticles. (d) High
resolution TEM image showing the pores are uniformally distributed on
particle surface with size around 9 nm.
Figure 3.3 (a) Boron concentration of doped Si nanoparticles prepared using
different H
3
BO
3
: Si mass ratios, calibrated by ICP-AES. (b-d) TEM
images of porous silicon nanoparticles prepared with initial H
3
BO
3
: Si
mass ratios of 2:5 (b), 4:5 (c), and 8:5 (d), respectively
Figure 3-4. (a-b) Cyclic voltammetry curves of doped porous silicon nanoparticles
and undoped nonporous silicon nanoparticles for the first cycle (a), and
the second cycle (b), respectively. (c) Differential capacity curves of
doped porous silicon nanoparticles and undoped nonporous silicon
nanoparticles in the charge branch of the first cycle.
Figure 3.5 Characterization of porous Silicon nanoparticles with carbon coating and
reduced graphene oxide wrapping as lithium-ion battery anode. (a)
30
38
40
42
44
49
x
Charge/discharge capacity at current rates of 1/16 C, 1/8 C, 1/4 C, and
1/2 C, (1C = 4 Ah/g). (b) Cycling performance at current rates of 1/4 C
and 1/2 C.
Figure 4.1 Schematic diagram of synthetic route of porous silicon from
metallurgical silicon through ball-milling and stain-etching.
Figure 4.2 SEM image of silicon powder after ball-milling.
Figure 4.3 Schematic diagram of the synthesis and morphology of porous silicon
particles. A. Synthetic route of porous silicon from metallurgical silicon
through ball-milling and stain-etching. B. A STEM image of porous
silicon particles. Inset: a TEM image showing the pore locations. C.
Reconstructed structure of the particles from HAADF-STEM 3D
tomography. D-F. Projected views of orthogonal slices cut through the
center of the particle.
Figure 4.4 Morphology characterization of a porous silicon particle (Si(1)) before
(A) and after (B) indentation using an AFM tip. The protrusion found on
the surface after indentation (B) indicates the plastic-like deformation of
the porous structure.
Figure 4.5 Mechanism of stain-etching to prepare porous silicon. A. Part of atomic
model containing Fe-Al dopants in silicon used in first-principle
calculations. The whole atomic model is a 3 x 3 x 3 supercell of silicon
conventional cell. B. First-principle calculations of projected density of
states of Fe, Al, and Si atoms. C. Schematic diagram of band alignment
of silicon and different redox pairs. The etchant solution functions as a
hole reservoir and can inject holes into the impurity band of silicon. D.
Calculated spatial distribution of electrons with energy within ±0.5 eV of
Fermi energy. The yellow part indicates electron within this energy is
mostly localized around Fe atom.
Figure 4.6 Structure characterization of Si(1) and Si(2). HAADF-STEM images of
Si(1) and Si(2) are shown in A and G. Reconstructed structure from
HAADF-STEM tomography for Si(1) and Si(2) are shown in B-F and
H-L, respectively. C-F show the section planes of Si(1) at corresponding
slice positions shown in B, which demonstrate the porous structure
throughout the whole particle. I-L show the section planes of Si(2) at
corresponding slice positions shown in H. It clear shows porous
structure at surface and solid (non-porous) structure at inner part of the
56
57
59
61
64
56
xi
particles.
Figure 4.7 Morphology comparison (SEM images: A and E, TEM images: the rest)
of porous silicon etched using different etchant. A-D: SEM and TEM
images of Si(1) (etched in Fe(NO
3
)
3
/HF), which show the highly porous
structure of the particles, and individual pores can be discerned from
high-magnification TEM images (e.g. D). E-H: SEM and TEM images
of Si(3) (etched in Fe(NO
3
)
3
/HNO
3
/HF), which show similar structure to
Si(1).
Figure 4.8 Electrochemical measurement and structure characterization of porous
silicon nanoparticles. A. Cyclic performance of porous silicon of Si(1),
Si(2) and un-etched silicon at current rate of 0.2 C (1 C = 4 A/g, samples
are coated with carbon and GO). B. Ex-situ STEM images showing the
morphology of Sample 1after 1
st
lithiation, which shows the porous
feature of, and demonstrate the capability of porous structure to
accommodate volume expansion of the particle during the lithiation
process. C. EELS mapping of Li (red) and Si (blue) at the enlarged
region of B. The peripheral Li-rich region suggests the formation of solid
electrolyte interface (SEI). D. EELS spectroscopy indicates the signal
from Li and Si at the enlarged region in figure C. E. Ex-situ TEM image
showing the morphology of Si(1) after 10 cycles and being charged to
2V (de-lithiated state). The particle keeps integrated without crack.
Figure 4.9 Electrochemical measurements and structure characterization of porous
silicon nanoparticles with GO after annealing in Ar at 700
o
C. A.
Discharge capacity of the samples at different current rates. B. Cyclic
performance of the samples at current rates of 0.5 C and 1.0 C (samples
are pre-charged/discharged at 0.1 C for 10 cycles). The Coulombic
efficiency stables above 99% for all the cycles (the Coulombic efficiency
curve is for sample cycled at 0.5 C).
Figure 4.10 Evaluation of silicon anodes in terms of capacity, cycle number, current
rate, and production cost. The color scheme represents cycle number; the
symbol represents current rate; the numbers next to each symbol refer to
the reference index.
Figure 5.1 SiH
4
molar ration to GeH
4
during the synthesis versus silicon atomic
ratio in Si
x
Ge
1-x
alloy nanoparticles.
Figure 5.2 X-ray diffraction (XRD) pattern of as-synthesized Si
x
Ge
1-x
alloy particle
65
68
70
73
86
87
xii
with different composition.
Figure 5.3 TEM images of Si
x
Ge
1-x
nanoparticles with composition of Si
77
Ge
23
(a),
Si
54
Ge
46
(b), Si
23
Ge
77
(c), respectively. Scale bar is 30 nm for all images.
Figure 5.4 (a) TEM and (b) HAADF-STEM images of Si
77
Ge
23
nanoparticles at
different magnifications. The subtle contrast difference in the inset of b
indicates the non-homogeneity of Si-Ge distribution at atomic scale. c,
elemental distribution of Si (blue) and Ge (yellow) of the Si
77
Ge
23
sample. A schematic model in the inset of c illustrates the non-
homogeneity of Si-Ge. d, Raman spectrum of Si
77
Ge
23
using 514 nm
laser as incident beam. The Raman shifts around 300 cm
-1
, 400 cm
-1
and
510 cm
-1
correspond to the vibration of Ge-Ge, Si-Ge, and Si-Si,
respectively. e, XRD pattern indicates the single phase of Si
77
Ge
23
at
particle scale. Scale bar is 100 nm, 10 nm, 2 nm, and 20 nm in the
images of a, b, inset of b, and c, respectively.
Figure 5.5 Raman spectra of Si
x
Ge
1-x
with different composition.
Figure 5.6 Morphology and SiGe distribution after the 1
st
cycle for Si
77
Ge
23
,
Si
54
Ge
46,
and Si
23
Ge
77
, respectively. Energy-filtered TEM was used to
determine Si-Ge elemental distribution.
Figure 5.7 Morphology and SiGe distribution after the 5
th
cycle for Si
77
Ge
23
,
Si
54
Ge
46,
and Si
23
Ge
77
, respectively. Energy-filtered TEM was used to
determine Si-Ge elemental distribution.
Figure 5.8 TEM images of (a-b), (g-h), and (m-n) show the morphology of
Si
77
Ge
23
, Si
54
Ge
46
, and Si
23
Ge
77
after the 1
st
and 5
th
cycle, respectively. It
is clear the particles for all three samples are merged together. After
100
th
cycle, particles evolve into loose structures, as shown in the STEM
images (c, i, o). EDX mapping is performed within the dashed square
regions in STEM. EDX of Si (d, j, p) and Ge (e, k, q), and combination
of Si+Ge (f, l, r) demonstrate that Si and Ge have almost homogeneous
distribution after 100
th
cycles. Scale bar is 100 nm for all images
Figure 5.9 (a) measured capacity of Si
x
Ge
1-x
nanoparticles at difference current
rates. (b) cyclic test for additional 100 cycles at current rate of 0.8A/g
after rate test shown in (a).
Figure 5.10 Electrochemical performance of Si
x
Ge
1-x
nanoparticles as lithium-ion
90
88
91
93
93
95
96
96
xiii
battery anode. Capacities at 25
th
, 50
th
, 75
th
, and 100
th
cycle of Si
x
Ge
1-x
were compared at a current rate of 0.8A/g. Capacity retention of Si
x
Ge
1-x
nanoparticles defined as the ratio of capacity at 100
th
cycle with respect
to capacity at 50
th
cycle is shown as the square symbol in the figure.
Figure 5.11 (a) modeled structure of Li
15
(Si
0.5
Ge
0.5
)
4
to simulate the lithiated Li-SiGe
alloy, and non-uniform Si and Ge distribution was created to simulate the
non-homogeneity. (b) Ge-Si pair distribution function (PDF) at different
stages of simulated delithiation process. The appearance and intensity
increasing of peak at 2.5 Å indicate the inter-mixing of Si and Ge to
form a homogenized alloy. (c) atomic structure and morphology of the
particle at different simulation states.
Figure 5.12 Particle morphology of Si
54
Ge
46
after fast delithiation rate at different
locations and different magnifications.
Figure 6.1 Charge/discharge profile of Li-S battery and its typical battery
performance.
Figure 6.2 Various structures of S incorporated in carbon-based materials as
compound cathodes for Li-S battery. (a) S incorporated into mesoporous
CMK3 porous carbon, and its battery performance. (b) S incorporated
into mesoporous carbon spheres and its battery performance. (c) S
incorporated into carbon nanotubes and its battery performance.
Figure 6.3 TEM images of porous carbon (a) and after infiltration with sulfur (b).
Figure 6.4 Battery performance of Li-Si half cell (left) and Li-S half cell (right)
with LiPF6 and LiTFSI electrolyte, respectively.
Figure 6.5 (a) battery performance of Si(lithiated)-S full cell using LiTFSI
electrolyte. (b) battery performance of Li-Si half-cell with additional
Li
2
S
x
in LiTFSI electrolyte.
Figure 6.6 Molecular structure of nafion polymer.
Figure 6.7 TEM and FETEM images of nafion coated Si nanoparticles.
Figure 6.8 Cyclic performance of Si(lithiated)-S full cell.
100
101
109
111
116
117
118
119
119
120
xiv
Abstract
Lithium-ion battery has generated great impacts on portable electronics since its first
commercialization in 1990s. Existing lithium-ion batteries using graphite as anode have
already been widely used in mobile applications; however, it is still urgent and important
to develop new battery systems with larger specific capacity and higher power density for
applications in mobile devices, hybrid electric vehicles (HEVs), and plug-in hybrid
electric vehicles (PHEVs). For electrode material, silicon is a promising anode material
for the next generation lithium-ion batteries because of its high theoretical specific
capacity (4200 mAh/g), which is about ten times as large as currently used graphite (372
mAh/g). However, significant volume change of silicon during the repeated insertion and
extraction of lithium ions leads to severe electrode pulverization and capacity loss, which
has dramatically hindered its practical usage. We have developed a technology of using
porous silicon as a stable and long cyclic life anode material with high capacity.
The thesis will start from a brief introduction of lithium-ion battery and its fundamentals.
Address will be paid on the discussion of the pros-and-cons of silicon, followed by a
concise review of the recent development on silicon anode materials. Then in the
following chapters, we will report the achievements we have made on the evaluation of
porous silicon for use as lithium-ion battery anode.
xv
In chapter 2, start from an analytical analysis of mechanic properties of porous silicon
upon lithiation, we have demonstrated that silicon with embedded pores is beneficial to
release the stress and strain during the lithiation process, and helps to keep its structure
integrity, which is further confirmed by experimental test on porous silicon nanowire as a
prototype material as battery anode. Porous silicon nanowires have been tested for 2000
cycles with capacity larger than 1000 mAh/g. Details on the synthesis of porous silicon
nanowires and electrochemical measurements are provided therein.
In chapter 3, we introduce a novel method to synthesize porous silicon structure in bulk
quantity from commercially available solid silicon nanoparticles through a controlled
boron-doping and silver-assistant electroless etching process, to overcome the limit of
quantity yields of porous silicon nanowire that can be obtained as described in chapter 2.
Effect of boron-doping is discussed and it is found to have negligible effects to the
voltage profile of silicon anode. As-prepared porous silicon nanoparticles can achieve
1500 mAg/g at a charge/discharge rate of 2 A/g for over 250 cycles after mixing with
graphene oxide wrapping.
In chapter 4, we further have developed a facile and scalable synthetic protocol to
produce nanoporous silicon particles in a large scale and eliminated the use of expensive
silver precursor for the preparation of porous silicon in previous chapters. Industrial
xvi
grade metallurgical silicon is employed as starting material and going through eco-
friendly ball-milling and stainless etching (ferric etchant, Fe(NO
3
)
3
/HF) to produce
nanoporous silicon. Etching mechanism is discussed with the aids of density-function-
theory (DFT) calculation and 3D-tomogrphy analysis of particle under transmission
electron microscopy (TEM), which disclose the effects from the different choices of
etchant species to the particle morphology. Battery tests show that the porous silicon
particles can be cycled over 600 cycles with capacity larger than 1000 mAh/g.
In chapter 5, we will discuss an alternative strategy to improve the cyclic performance of
silicon anode by alloying silicon with germanium, which provides a complementary
method to stabilize cyclic performance of silicon anode, in additional to the currently
well-adopted strategy of engineering the silicon into nanostructures, such as the porous
structure mainly described in this thesis. It is found SiGe composition approaching to
Si
50
Ge
50
has the best capacity retention behavior.
Finally, in chapter 6, we have made an attempt to integrate silicon with sulfur into full
batteries, noting that sulfur is a promising cathode with high capacity (>1000 mAh/g v.s.
~150 mAh/g for traditional metal oxide cathode). Challenges faced in the full-battery
integration, as well as potential solutions are discussed, and preliminary results will be
presented.
1
Chapter 1 Introduction to lithium-ion battery
1.1 Status of lithium-ion batteries
Lithium-ion battery is a family member of rechargeable batteries, in which lithium ions
are reversibly shuttled between two electrodes to convert chemical energy into electrical
energy, and vice versa. Since their commercialization by Sony Corporation in 1991,
lithium-ion batteries present to be a key component in today’s portable electronics and
telecommunication equipment. Compared with other battery systems, lithium-ion
batteries outperform their counterparts, such as lead acid and nickel cadmium batteries, in
term of, e.g. the high energy density (either gravimetric energy density (Wh/kg), or
volumetric energy density (Wh/l)), as indicated in Figure 1-1
1
. Other distinguishing
features of today’s commercial Li-ion batteries are:
· High operating voltage: a single cell has an average operating voltage of 3.6 V .
· Fast charging potential: battery can be charged to 80-90% of full capacity in one
hour.
· Wide range of operating temperature: from -20 to 60
o
C.
· Long cycle life: life of a battery exceeds 500 cycles.
· Low self-discharge: only 8-12% per month.
· No memory-effect: can be recharged at any time.
2
· Environmentally friend: no toxic heave metals such as Pb, Cd, and Hg.
In the last decade, continuous efforts have been devoted to developing high performance
lithium-ion batteries to meet the vast need for clean energy and the ever-increasing
demand for powerful energy storage systems. There is a worldwide thirst for batteries
which can deliver higher capacity and larger energy density than ever before. Especially,
the advent of electric and hybrid vehicles at the beginning of this century has pushed the
requirement of batteries to an unprecedented high level. It is strongly desired to make
battery competitive with fossil fuels in powering up vehicles.
Figure 1.1 Comparison of different battery technologies in terms of gravimetric energy density and
volumetric energy density
1
.
3
1.2 Electrochemistry of lithium-ion batteries
1.2.1 Working principle
Like all the electrochemical batteries, the structure of lithium-ion battery is composed of
a cathode and an anode, separated by electrolyte solution which enables ion transfer
between the two electrodes but blocks electron transport, as schematically shown in
Figure 1.2.
When the lithium ions flow through electrolyte, electrons generated from chemical
reactions go through the outer circuit to do work. For a typical lithium-ion battery using
LiFePO
4
as cathode and graphite as anode, the following reactions take place:
charge
+-
4 1-x 4
discharge
charge
+-
x
discharge
Cathode: LiFePO Li FePO + xLi + xe
Anode: C + xLi + xe Li C
In the first charge process, along with the insertion of Li
+
ion into anode, a passivation
layer will form on the surface of both electrodes resulting from the reaction between
electrolyte and electrode material, which is called solid electrolyte interface (SEI). The
SEI is ionic conductive; a stable SEI layer is crucial to protect electrodes from further
reaction with electrolyte in the subsequent cycles. The formation of SEI consumes the Li
in the cathode and leads to an irreversible transportation of Li
+
ions in the first discharge
process; however, a good battery should have this irreversible lithium ion transportation
less than 0.1% in the following charge/discharge process.
4
Figure 1.2 Schematic illustration of a typical lithium-ion battery where graphite and LiFePO
4
are used
as anode and cathode, respectively.
5
1.2.2 Working voltage
For lithium-ion batteries, the cathode functions like a lithium ion sink, which has a high
potential to deliver lithium ions into anode via electrolyte. The difference of electrode
potential between cathode and anode determines the cell voltage, which is expressed as:
cell cathode anode
E E E
To achieve a large cell potential, positive electrode materials (cathode) need to have high
lithium atom binding energy and the negative electrode materials (anode) should have a
low lithium binding energy.
1.2.1 Capacity
Capacity is another important property of a battery, which is tightly related to the amount
of electricity involved in the electrodes reactions as indicated previously. Conventionally,
the capacity is quantified in unit of mAh, which is calculated as:
Q xnF
where x is the number of moles of a chosen electroactive component that takes place in
the reaction, n is the number of electrons transferred per mole of reaction, and F is
Faraday constant (F = 96485 C/mol). In order to characterize and compare capacities
among different electrode materials, specific capacity in unit of mAh/g is usually
calculated to take the mass of the electroactive component into accounts, which is a
characteristic property of different electrode materials. Table 1.1 compares the electrode
potential and capacity of some cathode and anode materials; either has been already
6
commercialized, or has aroused intense interests in research study.
Since the amount of energy of a battery that can deliver is directly linked to the electrode
capacity and cell voltage, both capacity and electrode potential are two key criteria to
evaluate the electrode material. During the years, developments in cathode has
remarkably boosted the energy storage capability, as different types of cathode materials,
e.g. spinel, stacked, or folded structure lithium metal oxides, lithium iron phosphate, and
polyanions lithium cathode have been intensively investigated, and many of them are
being used in batteries for various applications currently
2-4
. There are many insightful
review papers on cathode material related to both fundamental physics and the state-of-
art technologies
4,5
. On the other hand, for the anode part, graphite has been widely used
with reliable performance; however, the low specific capacity (372 mAh/g) of graphite
has led researchers to seeking for alternative anode materials with high capacity. A recent
announcement on anode material progress was made by Sony Company in 2005. They
use Sn-Co-C alloy as the anode material in their “Nexelion” battery, which improves the
capacity of conventional offerings by 30%. This suggests that other material candidates
may begin to replace graphite as anode material in the following lithium-ion battery
development. To find anode alternatives to increase the energy density of the battery is
the aim of this.
7
Table 1.1 Electrode potential and capacity of some cathode and anode materials
Material Structure
Plateau voltage
(vs. Li+/Li)
Capacity
(mAh/g)
Ref
Cathode
LiCoO
2
R-3m ~ 3.7 V 274
6
LiNiO
2
R-3m ~ 4.0 V 260
7
LiMnO
2
C2/m ~ 3.5 V 285
8
LiMn
2
O
4
Fd3m ~ 3.7 V 148
9
Li
4
V
3
O
8
P21/m ~ 3.7 V 280
10
LiFePO
4
Pnma ~ 3.7 V 170
11
LiMnPO
4
Pnma ~ 4.1 V 171
12
LiCoPO
4
Pnma ~ 4.8 V 170
13
Li
3
V
2
(PO
4
)
3
P21/n ~ 4.0 V 197
14
LiNi
0.5
Mn
0.5
O
2
R-3m ~ 4.8 V 160
15
Anode
Graphite P63mc ~ 0.1 V 372
16
Li
4
Ti
5
O
12
Fd-3m ~ 1.5 V 175
17
Silicon Fd-3m ~ 0.2 V 4200
18
Germanium Fd-3m ~ 0.4 V 1625
19
Tin I41/amd ~ 0.4 V 994
20
8
1.3 Fundamental properties of silicon as high capacity anode
1.3.1 Advantages of Si
Among the vast anode candidates, silicon (Si) has attracted the most attention. With its
ability to form a rich family of Li-Si alloy, silicon can store 4.4 Li atoms per 1 Si atom to
form an alloy of Li
22
Si
5
, corresponding to the highest theoretical capacity of 4200 mAh/g.
In addition, the low binding energy of Li to Si leading to a low discharge potential around
~0.2 V with respect to lithium metal makes silicon a suitable anode material for lithium-
ion battery.
1.3.2 Disadvantages of Si
However, silicon has been overlooked for a long time since the first discovery of Li
22
Si
5
in 1966
21
, mainly due to the fast capacity degradation for limited cycles. The poor
cyclability of silicon is attributed to the large volume change when lithium ions
intercalate into the silicon frame. Figure 1.3 shows the volume evolution of Si when
hosted with different amount of Li.
From Figure 1.3, it shows that there is around 400 % volume increase when silicon is
fully lithiated, which can cause severe pulverization of Si, loss of electric contact, and
eventually capacity fading
22
. Figure 1-4 shows a galvanostatic voltage profile of silicon
powders, from which, severe degradation is found after 10 charge/discharge cycles.
9
Figure 1.3 The volume per mole of host atoms (Li) for silicon as a function of lithium content for
phases occurring in its binary phase diagrams with lithium
23
.
Figure 1.4 Charge-discharge voltage profiles obtained with a silicon powder anode
24
.
Another problem associated with silicon anode lies on the intrinsic low diffusion rate of lithium in
silicon, which is around 2 x 10
-11
cm
2
s
-1
25,26
. The low diffusion rate of Li in Si limits the usage of Si
anode for high power committed applications that need to be operated at high charging/discharging
rate.
10
Recently, results using silicon thin film
27
and silicon nanowires
28
showed that significant
performance improvements could be achieved for silicon anode, by taking the advantages
of silicon nanostructures that can accommodate large volume expansion. Many new
silicon nanostructures were then designed and fabricated to enhance the cycling
performance. As Figure 1.5 indicates, diverse silicon nanostructures show promising
electrode performance.
Figure 1.5 Structure of silicon nanowire before (a) and after lithiation (b), and their cyclic
performance (c)
28
. Structure of silicon hollow sphere before (d) and after lithiation (e), and its cyclic
performance (f)
29
. Structure of 3D silicon (a-b), and its cyclic performance (i)
30
.
11
Among the diverse silicon nanostructures, three-dimensional porous structure is of great
interest. The porous structure can provide large space to accommodate volume expansion,
and therefore help to maintain structure integrity when lithium intercalates into silicon. In
addition, porous structure can provide large surface area that is accessible to electrolyte,
and short diffusion length for lithium-ions to transport from electrolyte to silicon, thus
facilitating the charge/discharge process at high current rates.
In the following chapters, we will report the recent progress we have achieved; future
research plan will be discussed at the end of this article. Specifically, in chapter 2, we will
talk about the virtue of porous silicon anode and experimentally examine the advantage
of porous structure using porous silicon nanowire as a prototype. In chapter 3, a scalable
way to produce porous silicon nanoparticle is provided. In chapter 4, a cost-efficient
production of porous silicon from metallurgical silicon is described, followed by a
discussion about mechanism study. In chapter 5, advantages of Si-Ge alloy particles are
iterated as the battery anode with stable cyclic performance. In chapter 6, future research
plan and preliminary results on the integration of silicon anode and sulfur cathode are
provided.
12
1.4 Chapter reference
1 Tarascon, J. M. & Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature
414, 359-367 (2001).
2 Gummow, R. J., Dekock, A. & Thackeray, M. M. Improved Capacity Retention in Rechargeable
4v Lithium Lithium Manganese Oxide (Spinel) Cells. Solid State Ionics 69, 59-67 (1994).
3 Liu, W., Farrington, G. C., Chaput, F. & Dunn, B. Synthesis and electrochemical studies of spinel
phase LiMn2O4 cathode materials prepared by the Pechini process. J Electrochem Soc 143, 879-
884 (1996).
4 Koksbang, R., Barker, J., Shi, H. & Saidi, M. Y . Cathode materials for lithium rocking chair
batteries. Solid State Ionics 84, 1-21 (1996).
5 Whittingham, M. S. Lithium batteries and cathode materials. Chem Rev 104, 4271-4301 (2004).
6 Antolini, E. LiCoO2: formation, structure, lithium and oxygen nonstoichiometry, electrochemical
behaviour and transport properties. Solid State Ionics 170, 159-171 (2004).
7 Arai, H., Okada, S., Sakurai, Y . & Yamaki, J. Reversibility of LiNiO2 cathode. Solid State Ionics
95, 275-282 (1997).
8 Ammundsen, B. et al. Formation and structural properties of layered LiMnO2 cathode materials. J
Electrochem Soc 147, 4078-4082 (2000).
9 Thackeray, M. M., Dekock, A. & David, W. I. F. Synthesis and Structural Characterization of
Defect Spinels in the Lithium-Manganese-Oxide System. Mater Res Bull 28, 1041-1049 (1993).
10 Benedek, R., Thackeray, M. M. & Yang, L. H. Lithium site preference and electronic structure of
Li4V3O8. Phys Rev B 56, 10707-10710 (1997).
11 Padhi, A. K., Nanjundaswamy, K. S., Masquelier, C., Okada, S. & Goodenough, J. B. Effect of
structure on the Fe3+/Fe2+ redox couple in iron phosphates. J Electrochem Soc 144, 1609-1613
(1997).
12 Kim, S. W., Kim, J., Gwon, H. & Kang, K. Phase Stability Study of Li1-xMnPO4 (0 <= x <= 1)
Cathode for Li Rechargeable Battery. J Electrochem Soc 156, A635-A638 (2009).
13 Bramnik, N. N., Nikolowski, K., Trots, D. M. & Ehrenberg, H. Thermal stability of LiCoPO(4)
cathodes. Electrochem Solid St 11, A89-A93 (2008).
14 Zhu, X. J. et al. Synthesis and characteristics of Li3V2(PO4)(3) as cathode materials for lithium-
ion batteries. Solid State Ionics 179, 1679-1682 (2008).
15 Reed, J. & Ceder, G. Charge, potential, and phase stability of layered Li(Ni0.5Mn0.5)O-2.
Electrochem Solid St 5, A145-A148 (2002).
16 Peled, E., Menachem, C., BarTow, D. & Melman, A. Improved graphite anode for lithium-ion
batteries - Chemically bonded solid electrolyte interface and nanochannel formation. J
13
Electrochem Soc 143, L4-L7 (1996).
17 Prosini, P. P., Mancini, R., Petrucci, L., Contini, V . & Villano, P. Li4Ti5O12 as anode in all-solid-
state, plastic, lithium-ion batteries for low-power applications. Solid State Ionics 144, 185-192
(2001).
18 Li, H., Huang, X. J., Chen, L. Q., Wu, Z. G. & Liang, Y . A high capacity nano-Si composite anode
material for lithium rechargeable batteries. Electrochem Solid St 2, 547-549 (1999).
19 Baggetto, L. & Notten, P. H. L. Lithium-Ion (De)Insertion Reaction of Germanium Thin-Film
Electrodes: An Electrochemical and In Situ XRD Study. J Electrochem Soc 156, A169-A175
(2009).
20 Jiang, T., Zhang, S. C., Qiu, X. P., Zhu, W. T. & Chen, L. Q. Preparation and characterization of
tin-based three-dimensional cellular anode for lithium ion battery. J Power Sources 166, 503-508,
doi:DOI 10.1016/j.jpowsour.2007.01.017 (2007).
21 Axel, H., Schafer, H. & Weiss, A. Zur Kenntnis Der Phase Li22si5. Z Naturforsch Pt B B 21, 115-
& (1966).
22 Kasavajjula, U., Wang, C. S. & Appleby, A. J. Nano- and bulk-silicon-based insertion anodes for
lithium-ion secondary cells. J Power Sources 163, 1003-1039, doi:DOI
10.1016/j.jpowsour.2006.09.084 (2007).
23 Obrovac, M. N., Christensen, L., Le, D. B. & Dahnb, J. R. Alloy design for lithium-ion battery
anodes. J Electrochem Soc 154, A849-A855 (2007).
24 Ryu, J. H., Kim, J. W., Sung, Y . E. & Oh, S. M. Failure modes of silicon powder negative
electrode in lithium secondary batteries. Electrochem Solid St 7, A306-A309 (2004).
25 Yoshimura, K., Suzuki, J., Sekine, K. & Takamura, T. Measurement of the diffusion rate of Li in
silicon by the use of bipolar cells. J Power Sources 174, 653-657 (2007).
26 Tritsaris, G. A., Zhao, K. J., Okeke, O. U. & Kaxiras, E. Diffusion of Lithium in Bulk Amorphous
Silicon: A Theoretical Study. J Phys Chem C 116, 22212-22216 (2012).
27 Ohara, S., Suzuki, J., Sekine, K. & Takamura, T. Li insertion/extraction reaction at a Si film
evaporated on a Ni foil. J Power Sources 119, 591-596, doi:Doi 10.1016/S0378-7753(03)00301-X
(2003).
28 Chan, C. K. et al. High-performance lithium battery anodes using silicon nanowires. Nat
Nanotechnol 3, 31-35, doi:DOI 10.1038/nnano.2007.411 (2008).
29 Yao, Y . et al. Interconnected Silicon Hollow Nanospheres for Lithium-Ion Battery Anodes with
Long Cycle Life. Nano Lett 11, 2949-2954, doi:Doi 10.1021/Nl201470j (2011).
30 Kim, H., Han, B., Choo, J. & Cho, J. Three-Dimensional Porous Silicon Particles for Use in High-
Performance Lithium Secondary Batteries. Angew Chem Int Edit 47, 10151-10154, doi:DOI
10.1002/anie.200804355 (2008).
14
Chapter 2 Porous doped silicon nanowire as a lithium-ion battery
anode
2.1 Mechanic properties of silicon anode
Despite the highest theoretical capacity of silicon anode, severe capacity degradation is
always the problem that handicaps the developing progress of silicon anode. Researches
show that the large volume change during the insertion and extraction of lithium ions in
silicon are responsible for the electrode pulverization and fast capacity fading. In the
lithiation process, a real-time measurement of stress evolution in silicon film indicated
that there was around 0.5 GPa compressive stress in the amorphous lithiated silicon layer
1
. The high stress accumulated in the amorphous silicon layer would initiate a crack and
then extend into underline crystalline silicon. Some recent work on in-situ TEM
observation of both crystalline and amorphous silicon nanoparticles during the lithiation
and delithiation process indicated that there was a critical size around 150 nm for
crystalline silicon
2
and 870 nm for amorphous silicon
3
, beyond that, particle could crack
into pieces. However, we should notice that the critical size of silicon nanoparticle is
derived from the observation when particles are presented in a constrain-free
environment. In the case of real electrodes where particles are densely stacked, the
critical size to prevent particles from cracking in the volume-constrained environment
might be much smaller than which is observed in the constrain-free environment due to
15
the additional interaction force exerted by neighboring particles. Along with the
experimental observations, recent theoretical simulations suggest that porous structured
silicon with pores in several nanometers, or hollow silicon spheres with thin shell, can
dramatically reduce the stress by providing additional free space for volume expansion
induced by lithium-ion insertion
4
. The methodology and implantation of the structure
simulation on porous silicon will be described below.
16
2.2 Simulation of porous silicon during lithiation
To simulate the structure evolution of silicon under lithium-ion insertion and then to
evaluate the stress generated due to volume expansion, a few coupled partial differential
equations need to be solved. On one hand, diffusion of lithium-ions into silicon would
generate large stress and strain; on the other hand, the strain in turn affects the diffusion
rate of lithium-ions. In a simplified model, it is assumed that the diffusion of lithium in
silicon is proportional to the lithium concentration gradient, and compromised by elastic
energy of the system. There are the two master equations describing the lithiation
process
5,6
:
0 , and ( )
h
cc
J J D c
t RT
where c is lithium ion concentration, D is lithium ion diffusion coefficient, is partial
molar volume, R is gas constant, T is temperature, and
h
is hydrostatic stress, defined as
(
11
+
22
+
33
)/3, where
i,j
are the stress tensor elements. The equations above can be
solved using Neumann boundary condition:
/ J n i F
where n is the surface normal vector, i is electric current density, and F is Faraday’s
constant. After lithiation, the stress and strain due to the insertion of lithium is expressed
as:
1
[(1 ) ]
3
ij ij kk ij ij
c
E
17
Where
ij
is strain component, and
ij
is the Kronecker delta. Parameters for silicon used
in simulation are listed in Table 2.1
Table 2.1 Parameters used in simulation
Name Symbol and unit value
Young's modulus E (GPa) 80 (Ref
4
)
Poisson's ratio v 0.22 (Ref
4
)
Diffusion coefficient D (m
2
/s) 1x10
-16
(Ref
7
)
Partial molar volume m
3
/mol) 1.2x10
-5
(Ref
7
)
Stoichiometric maximum
concentration
c
max
(mol/m
3
)
5.27x10
4
(Li
15
Si
4
)
8.87x10
4
(Li
22
Si
5
)
18
Figure 2.1a schematically shows the porous structure, and calculation and analysis were
carried out on one unit (Figure 2.1b) of the structure in Figure 2.1a. Insertion of lithium
would generate stress in the silicon matrix, and the strain induced by stress not only
deforms the structure (expansion), but also compromises the lithium diffusion. The pore
size evolution after lithium ion intercalation to Li
22
Si
5
and Li
15
Si
4
at fixed pore-to-pore
distance (l =12 nm) was presented in Figure 2.1c. It shows that the pore diameter after
lithium intercalation decreases with decreasing of initial pore size. The maximum stress
around the pore increases as we decrease the initial pore size (Figure 2-1d), which would
act as a source of fracture. In another case, we fix the porosity by means of fixing the
ratio of initial pore radius (r) and pore-to-pore distance (l), Figure 1e shows the
correlation of pore sizes before and after lithiation, and there is almost no change in the
maximum stress at different pore sizes (Figure 2-1f). Generally speaking, decreasing r/l
ratio to a low value (low porosity) would increase maximum stress, and smaller initial
pore results in higher maximum stress around the pore. Therefore, obtaining silicon with
high porosity and large pore size would help to stabilize the structure during the
charge/discharge process.
19
Figure 2.1 (a) Schematic diagram of a porous silicon structure. (b) One unit of the porous structure
used for theoretical simulation and analysis. (c) Pore size before and after lithiation and (d)
corresponding maximum stress at fixed pore-to-pore distance (l =12 nm). (e) Pore sizebefore and after
lithiation and (f) corresponding maximum stress v.s. initial pore size at fixed pore/edge ratio (r/l=1/3)
20
We note that the absolute value of the calculated stress in this simulation is larger than
previous experimental measurements. The discrepancy may come from the following two
reasons. On one hand, the periodic boundary conditions that have been applied on the
simulation structure restricts the expansion in all around surfaces, and therefore results in
high stresses. In reality, a porous structure with numerous pores both on the surface and
inside the body is free to expand in all directions. On the other hand, during the insertion
of lithium-ions, apart from the elastic expansion of silicon as it is assumed in the studied
model, some other studies from theoretical and experimental point of views have shown
that silicon also takes plastic deformation which leads to a much lower stress
8-10
. Another
limitation of the model lies on the overlook of the geometry effect (e.g particle size,
shape, etc) to the mechanic failure of silicon. As suggested by Ma el al, simulation shows
that the critical sizes of fracture are different, specifically, they are ~ 90 nm for particles,
~ 70 nm for nanowires, and ~ 33 nm for nanofilms, respectively
11
. The geometry effect
and the fracture process have not been considered in the model. However, it is difficult
and may not be practical to simulate the porous silicon structure precisely if taking all the
aspects into account, in additional to other issues such as surface roughness, percentage
of different crystal planes with anisotropic mechanic behavior which varies from sample
to sample, etc. Nevertheless, the simple model still serves as a good demonstration to
illustrate the merit of porous silicon structure: a porous structure with large pore size
effectively release the stress of silicon after lithiation, thus helps to keep the structure
integrity of silicon anode without losing electrical conductance during charge and
21
discharge process. In order to verify the assumption, we have synthesized porous silicon
nanowires with large pores to check their electrochemical performance.
22
2.3 Preparation of porous silicon nanowire
Porous Si nanowires were prepared according to previous reports
12,13
Briefly, boron
doped Si wafers (resistivity < 5 m cm) were immersed in an etchant solution containing
5 M hydrofluoric acid (HF) and 0.02 M silver nitrite (AgNO
3
) for 3 hours. Porous
nanowires were washed by de-ionized water (DI-H
2
O), concentrated nitric acid (HNO
3
),
and DI-H
2
O again sequentially, and then collected by scratching from the wafers using a
blade. We note that doping is necessary for getting porous structure, and without doping,
we can only get solid silicon nanowire. Figure 2.2 shows a scanning electron microscopy
(SEM) image (Figure 2.2a) and transmission electron microscopy (TEM) images (Figure
2.2b-d) of porous Si nanowires. The nanowires are highly porous at surface, with pore
diameter and wall thickness both around 8nm (Figure 2.2b-c). The high resolution TEM
image in Figure 2-2d shows the nanowires are crystalline with clear lattice fringes
corresponding to Si (111) lattice. The crystalline structure was also confirmed by the spot
pattern in selected area electron diffraction (SAED) taken on a single porous nanowire, as
shown in Figure 2-2e. The pore size could be tailored by reacting with AgNO
3
at different
concentrations. Etchants containing 0.02 M and 0.04 M AgNO
3
gave pores with mean
diameters of 7.8±0.1 nm and 10.5 ± 0.1 nm (Figure 2.2f), respectively, based on a statistic
analysis of TEM images. The boron dopants provide defective sites facilitating the
etching process which leaves holes on the silicon nanowire surface. The etching process
was described as two simultaneous electrochemical reactions
13
:
23
2
6
4Ag 4 4Ag, Si 6F [SiF ] 4 ee
Figure 2.2 (a) SEM and (b) TEM images of porous Si nanowires etched with 0.02M AgNO
3
. (c) and
(d) HRTEM image of a nanowire in (b). (e) SAED pattern of a single porous silicon nanowire. (f) Pore
size distribution of porous Si nanowires etched with 0.02 M and 0.04 M AgNO
3
, respectively.
24
2.4 Electrochemical test and discussion
2.4.1 Fabrication of porous silicon nanowire anode
To test the electrochemical performance of porous silicon nanowires, two-electrode coin
cells using porous silicon nanowires as anode and lithium metal as the counter electrode
were fabricated. The electrode was made by mixing the porous silicon nanowire with
SuperP conductive carbon black and alginic acid sodium salt (alginate binder, Sigma
Aldrich, viscosity ~ 2000 cP at 2wt. %) in water to form uniform slurry (mass ratio of
silicon : SuperP = 2 :1, alginate binder: 15wt. %) and then spread on a copper foil using a
stainless steel blade. The electrode was dried at 90 ℃ overnight in air. Then CR2032 coin
cells were assembled in an Ar-filled glovebox using the as-prepared porous silicon
nanowire anodes as working electrodes and lithium metal foil as counter electrodes. The
electrolyte was 1 M LiPF
6
dissolved in a 1:1 (weight ratio) mixture of ethylene carbonate
(EC) and diethyl carbonate (DEC).
2.4.2 Electrochemical test on porous silicon anode
Figure 2.3a shows the voltage profile in the charge (lithiation) and discharge
(delithiation) process in the potential window of 0.01-2.0 V v.s. Li
+
/Li. The 1
st
cycle at
current rate of 0.4 A/g shows charge and discharge capacities of 3354 mAh/g and 3038
mAh/g, respectively. Cycles from 20
th
and thereafter run at a current density of 2 A/g
showing capacity degradation only about 9% per 100 cycles. After 200 cycles, capacity is
25
still above 1960 mAh/g, indicating good structure stability of porous Si nanowires.
Transition from crystalline Si to amorphous structure during cycling was confirmed by
cyclic voltammetry (C-V) curves at the 1
st
and 2
nd
charge/discharge cycles (Figure 2-3b).
During the 2
nd
cycle, the peak at 0.15 V is absent in the 1
st
cycle at cathodic branch
(lithiation), which indicates the crystal-to-amorphous transition
14
.
Figure 2.3c shows the charge/discharge capacity and Coulombic efficiency at different
current rates. The capacity maintains above 3400, 2600, 2000, 1900, 1700, and 1300
mAh/g at current densities of 0.6, 1.2, 2.4, 3.6, 4.8 and 9.6 A/g. The Coulombic
efficiency is around 90% at first several cycles, which may be very well due to the large
surface area of porous Si that needs longer time to form stable solid electrolyte interface
(SEI) layer. After 20 cycles, the Coulombic efficiency reaches above 99.5% at each
different current rate cycling step. The current rate, charging/discharging time, and
average specific capacity at different current rates are summarized in Table 2.2. It
indicates that even at high current rate (2.4 C = 9.6 A/g at step 6), charging/discharging
finished within 10 minutes still gives capacity above 1300 mAh/g, equivalent to 38% of
capacity using 0.15 C (step 1). Figure 2.3d shows long cycle performance at
charging/discharging rate of 0.1 C for the 1
st
cycle, and 0.5 C, 1 C, and 4.5 C for
additional 250 cycles, which shows stable capacities around 2000, 1600, and 1100
mAh/g, respectively. Capacity degradation is almost negligible at each case,
demonstrating good stability of the porous silicon structure. The battery shown in the top
26
panel of Figure 3d has been re-tested with 0.5 C and then 1 C current rate after one
month. It shows that the capacity remains above 1000 mAh/g after additional 2000 cycles
(Figure 2.3e).
27
Figure 2.3 Electrochemical performance of battery using porous silicon nanowires as anode and
lithium metal as current collector. A. Charge/discharge profile within a voltage window of 0.01-2V vs.
Li
+
/Li for the 1st cycle at a current rate of 0.4 A/.g, and the 50th, 100th, and 200th cycles at 2 A/g. B.
Cyclic voltammetry curves of porous silicon nanowire electrode for the 1st and 2nd cycles using a
voltage window 0.01-2V at rate of 0.1mV/s. C. Charge/discharge capacity and Coulombic efficiency
of porous silicon nanowire electrode at current rates of 0.6, 1.2, 2.4, 3.6, 4.8, and 9.6 A/g. D.
Charge/discharge capacity of porous silicon nanowire electrode at current rates of 2A/g, 4A/g, and
18A/g for 250 cycles. E. Charge/discharge capacity of a porous silicon nanowire electrode at current
rates of 2A/g (0.5 C) and 4A/g (1 C) with additional 2000 cycles.
28
Table 2.2 Cyclic performance of porous silicon anode
Step number Number of cycle Charging time
Charging rate
(1C=4 A/g)
Average specific
capacity (mAh/g)
1 10 6h 0.15 C > 3400
2 20 2h 0.30 C > 2600
3 20 1h 0.60 C > 2000
4 20 40min 0.90 C > 1900
5 20 30min 1.20 C > 1700
6 20 10min 2.40 C > 1300
29
2.4.3 More discussions
Morphology change of porous Si nanowires is then checked. Batteries after 10 cycles
running at 0.1 C (0.4 A/g) were disassembled, and silicon anodes were washed with
acetonitrile and 0.5 M HNO
3
to remove SEI layer, and then dissolved in ethanol to make
samples for TEM observation. For clear comparison, Figure 2.4 shows the TEM image of
a porous Si nanowire before (Figure 2.4a) and after cycling (Figure 2.4b-c). Figure 4b
clearly shows that nanowire remains highly porous, and pore size does not change
significantly after cycling (compared to Figure 2.4a). This agrees well with theoretical
analysis showing that porous silicon with a large initial pore size and high porosity would
not change its structure significantly after lithiation. Here the initial pore diameter was
around 8 nm and the wall between adjacent pores had thickness about 6 nm (Figure 2.4a);
after cycling, the pore diameter and wall thickness are still around 7-8 nm. The porous
silicon nanowires are mostly amorphous (Figure 2.4c), with some dark dots less than 5
nm embedded in the amorphous matrix. SAED in Figure 2.4d confirms that the dots are
crystalline silicon. Similar observation of small portion of un-lithiated silicon were
reported before
15,16
, and the underlying mechanism is not fully understood and deserves
further investigation.
30
Figure 2.4 TEM images of silicon nanowires before (a) and after (b) lithiation after 10 cycles at a
current rate of 0.4 A/g. (c) Enlarged TEM image of (b) showing the amorphous silicon structure. (d)
Selected area electron diffraction pattern showing black spots in (b) are crystalline silicon.
31
The existence of un-reacted silicon provides evidences that lithiation and delithiation in
silicon is not homogenous, and therefore contributes to non-uniform stress distribution
even at low charging/discharging rate. At some locations, accumulated stress may be
large enough to break silicon into fragments. This is especially true for non-porous
structures, like silicon nanowires which are not able to sustain its capacity after long
cycling, since lithium ions can only intercalate into silicon from the very outer surface
and generate large concentration gradient from surface to inner core, thus inducing large
stress. In addition, low diffusivity of lithium ions in silicon generates lithium ion
concentration gradient that also compromises the capacity as discussed with simulation in
Figure 1: the larger the concentration gradient, the lower the charge/discharge capacity.
For porous silicon nanowires, electrolyte goes everywhere in the pores, and lithium
intercalation happens instantaneously wherever there is a contact between silicon and
electrolyte. Previous study shows that CVD grown silicon nanowire can also become
porous after lithiation and delithiation
17
; however, there is still capacity degradation after
some cycles for CVD silicon nanowires
18
. As our simulation in Figure 2-1d shows, the
stress in porous silicon after lithiation strongly depends on the porosity, which is
understandable. We suspect that when CVD grown silicon nanowires become porous
after several cycles, the resulting porosity and location of pores (surface of nanowires v.s.
bulk of nanowires) can be different from ours, which might lead to the observed
difference in the cycling performance. Here, we also believe that boron doping increases
electron conductivity in silicon, which might help to reach high capacity at high current
32
rate, and the alginate, due to its high viscosity, could further improve the structural
stability during cycling.
33
2.5 Conclusion
In this chapter, theoretical and experimental studies are carried out using porous doped
silicon nanowires for lithium ion battery applications. The simulation shows porous
silicon having large pore size and high porosity can maintain its structure after lithium
ion intercalation while having low stress, which is beneficial for getting high capacity and
long cycle retention. Porous silicon nanowires were produced by direct etching of boron-
doped silicon wafers, and exhibited superior electrochemical performance and long cycle
life as anode material in lithium ion battery with alginate used as binder. The capacity
remained stable above 2000, 1600, and 1100 mAh/g at current rates of 2 A/g, 4 A/g, and
18 A/g, respectively, even after 250 cycles. We believe the good cyclability mainly stems
from the use of porous silicon structure.
34
2.6 Chapter reference
1 Chon, M. J., Sethuraman, V. A., McCormick, A., Srinivasan, V. & Guduru, P. R. Real-Time
Measurement of Stress and Damage Evolution during Initial Lithiation of Crystalline Silicon. Phys
Rev Lett 107 (2011).
2 Liu, X. H. et al. Size-Dependent Fracture of Silicon Nanoparticles During Lithiation. Acs Nano 6,
1522-1531 (2012).
3 McDowell, M. T. et al. In Situ TEM of Two-Phase Lithiation of Amorphous Silicon Nanospheres.
Nano Lett 13, 758-764, doi:Doi 10.1021/Nl3044508 (2013).
4 Yao, Y. et al. Interconnected Silicon Hollow Nanospheres for Lithium-Ion Battery Anodes with
Long Cycle Life. Nano Lett 11, 2949-2954, doi:Doi 10.1021/Nl201470j (2011).
5 Park, J., Lu, W. & Sastry, A. M. Numerical Simulation of Stress Evolution in Lithium Manganese
Dioxide Particles due to Coupled Phase Transition and Intercalation. J Electrochem Soc 158,
A201-A206 (2011).
6 Zhang, X. C., Shyy, W. & Sastry, A. M. Numerical simulation of intercalation-induced stress in
Li-ion battery electrode particles. J Electrochem Soc 154, A910-A916 (2007).
7 Chandrasekaran, R. & Fuller, T. F. Analysis of the Lithium-Ion Insertion Silicon Composite
Electrode/Separator/Lithium Foil Cell. J Electrochem Soc 158, A859-A871, doi:Doi
10.1149/1.3589301 (2011).
8 Zhao, K. J. et al. Concurrent Reaction and Plasticity during Initial Lithiation of Crystalline Silicon
in Lithium-Ion Batteries. J Electrochem Soc 159, A238-A243 (2012).
9 Shenoy, V. B., Johari, P. & Qi, Y. Elastic softening of amorphous and crystalline Li-Si Phases
with increasing Li concentration: A first-principles study. J Power Sources 195, 6825-6830 (2010).
10 Sethuraman, V. A., Chon, M. J., Shimshak, M., Van Winkle, N. & Guduru, P. R. In situ
measurement of biaxial modulus of Si anode for Li-ion batteries. Electrochem Commun 12, 1614-
1617 (2010).
11 Ma, Z. S. et al. Critical silicon-anode size for averting lithiation-induced mechanical failure of
lithium-ion batteries. Rsc Adv 3, 7398-7402 (2013).
12 Qu, Y., Zhou, H. & Duan, X. Porous silicon nanowires. Nanoscale 3, 4060-4068,
doi:10.1039/c1nr10668f (2011).
13 Hochbaum, A. I., Gargas, D., Hwang, Y. J. & Yang, P. Single Crystalline Mesoporous Silicon
Nanowires. Nano Lett 9, 3550-3554, doi:10.1021/nl9017594 (2009).
14 Xu, C. X. et al. Low temperature CO oxidation over unsupported nanoporous gold. J Am Chem
Soc 129, 42-43, doi:Doi 10.1021/Ja0675503 (2007).
35
15 Kim, H., Han, B., Choo, J. & Cho, J. Three-Dimensional Porous Silicon Particles for Use in High-
Performance Lithium Secondary Batteries. Angew Chem Int Edit 47, 10151-10154,
doi:10.1002/anie.200804355 (2008).
16 Park, M. H. et al. Silicon Nanotube Battery Anodes. Nano Lett 9, 3844-3847, doi:Doi
10.1021/Nl902058c (2009).
17 Choi, J. W. et al. Stepwise Nanopore Evolution in One-Dimensional Nanostructures. Nano Lett 10,
1409-1413 (2010).
18 Chan, C. K. et al. High-performance lithium battery anodes using silicon nanowires. Nat
Nanotechnol 3, 31-35, doi:DOI 10.1038/nnano.2007.411 (2008).
36
Chapter 3 Scalable production of porous silicon nanoparticles for
Li-ion battery anodes
3.1 Motivation
We have demonstrated the advantage of using porous silicon nanostructure as lithium-ion
battery anodes both theoretically and experimentally in the last chapter; however, we note
that the synthesis of porous nanowire is limited to small scale, as only very top thin layer
of silicon wafer can be converted to porous nanowire, and the length of nanowire can
hardly exceed 100 m. In addition, the wafer after nanowire peeling off could not be used
to produce porous nanowire again unless additional polishing process is conducted to
restore the surface to be atomic smooth. This is due to the selective etching ability of
Ag
+
/Ag nanoparticles on different silicon crystalline. The etching speed along different
crystal direction follows the sequence: <100> is larger than <110> and is much larger
than <111>; therefore, good quality of [100] surface is the prerequisite for nanowire
formation. As limited the small quantity of porous silicon nanowire could be fabricated,
strategies to produce large quantity porous silicon nanostructure is preferable.
37
3.2 Preparation of porous silicon nanoparticles
Here, we introduce a new and simple synthetic route for the preparation of porous silicon
nanoparticles. By doping and then etching of commercially available silicon
nanoparticles, porous silicon nanoparticles can be synthesized in bulk quantity.
The synthesis of porous silicon nanoparticles was schematically presented in Figure 3-1a,
wherein silicon nanoparticles with particle size < 200 nm (Shanghai Chaowei Nano.,
Ltd.) were used as starting material. We first doped silicon with boron, and then etched
the boron-doped silicon in an etchant containing silver nitrate (AgNO
3
) and hydrofluoric
acid (HF) to get porous structure. In the doping process, typically, 1.0 g silicon
nanoparticles and various amount of boric acid (0.4 g, 0.8 g, and 1.6 g) were well mixed
in solution, and then dried to get powder and annealed at 900 ℃ in argon environment for
three hours. The powder was washed with 5% hydrofluoric acid (HF) solution and de-
ionized water (DI water) to remove byproducts such as B
2
O
3
and SiO
2
. As obtained
boron-doped silicon nanoparticles were then immersed into 50 ml etchant solution
containing 10 mM silver nitrite (AgNO
3
) and 5 M HF under mild stirring. Immediately,
the solution bubbled as an indication of etching.
38
Figure 3.1 (a) Schematic diagram of the procedure to prepare porous silicon nanoparticles. (b)
Photographs of nonporous and porous Si nanoparticles, illustrating the scalable nature of our porous
silicon nanoparticle preparation.
39
After one hour, reaction was stopped by adding more DI water, and the solution was
centrifuged at 8000 rpm for 10 minutes, followed by additional washing using DI water.
Figure 1b shows the photograph of commercial nonporous silicon nanoparticles (left),
and porous silicon nanoparticles after the treatment (right). The amount of porous silicon
after the treatment is about 1/3 of the original amount of nonporous silicon. We note that
this technique is highly scalable to large quantities, as the starting material, silicon
nanoparticles, is commercially available in bulk quantity, and the doping and etching
processes are also compatible with large-scale manufacturing.
Figure 3.2a shows the TEM images of silicon nanoparticles right after the boron doping.
We noticed that there was no significant change in the morphology of silicon
nanoparticles after boron doping (Figure 3-2a), neither second crystal growth, nor particle
agglomeration was found. After etching, silicon nanoparticles became porous like, as
shown in Figure 3.2b, and were surrounded by many Ag nanoparticles (dark particles
shown in Fig. 3.2b) with a broad size distribution from 10-100 nm. These large Ag
particles might come from the nucleation and growth of small Ag clusters during the
reaction. HNO
3
could be used to remove Ag, leaving neat porous silicon nanoparticles
(Figure 3.2c) with pores (around 9 nm) that are uniformly distributed on top of the
particle surface, as shown in Figure 3.2d.
40
Figure 3.2 (a) TEM images of silicon nanoparticles after boron doping. (b) TEM image showing
porous silicon nanoparticles with large Ag nanoparticles after electroless etching. (c) TEM image of
porous Si nanoparticles after washing with HNO
3
and H
2
O to remove Ag nanoparticles. (d) High
resolution TEM image showing the pores are uniformally distributed on particle surface with size
around 9 nm.
41
According to the previous analysis, boron doping plays an important role to etch silicon
nanoparticles into porous structure; therefore, it is interesting and important to examine
the effect of doping concentration to the final morphology of porous silicon
nanoparticles, which will provide us a guideline to synthesize porous silicon nanoparticle
with desired structure. We have observed that, different doping concentrations can be
well controlled and achieved by adjusting the ratio of boric acid and silicon. According to
a simplified one-dimensional diffusion model
1
:
where C(t) is the total boron concentration, C
s
is surface concentration of boron atoms,
and D is diffusion coefficient, a positive correlation between C(t) and C
s
is established. In
our experiments, three samples with different doping concentrations had been achieved,
and characterized by inductively coupled plasma atomic emission spectrum (ICP-AES).
Figure 3.3a shows that the boron doping concentration monotonically increases with the
increase of initial mass ratio of boric acid to silicon from 2:5 to 8:5. Figure 3.4b, c, and d
show the TEM images of porous silicon nanoparticles prepared with initial Si : H
3
BO
3
mass ratios of 5:2, 5:4, and 5:8, respectively. It is found that silicon nanoparticles with
higher doping concentration give rougher surface and larger pores after etching (Fig. 3.3d
compared with Figure 3.3b and Figure 3.3c). This phenomenon can be understood by the
fact that, higher dopant concentration in p-type silicon (lower Fermi level) can lower the
energy barrier for electron transferring from silicon to Ag, thus facilitating the etching
2
( ) erfc( )
2
ss
x
C t C C Dt
Dt
42
process to generate large pores. Specific surface area of porous silicon nanoparticles was
analyzed using Brunauer-Emmet-Teller (BET) method. The surface area is 61 m
2
/g, 69
m
2
/g, and 82 m
2
/g for porous silicon nanoparticles prepared with initial Si : H
3
BO
3
mass
ratios of 5:2, 5:4, and 5:8, respectively.
Figure 3.3 (a) Boron concentration of doped Si nanoparticles prepared using different H
3
BO
3
: Si
mass ratios, calibrated by ICP-AES. (b-d) TEM images of porous silicon nanoparticles prepared with
initial H
3
BO
3
: Si mass ratios of 2:5 (b), 4:5 (c), and 8:5 (d), respectively
43
3.3 Electrochemical tests and discussion
The electrodes were prepared as follow. Porous silicon nanoparticles were first coated
with carbon using chemical vapor deposition (CVD) to help to form stable solid-
electrolyte-interface (SEI) layer during electrochemical test. In addition, carbon coated
porous silicon nanoparticles were then mixed with reduced graphene oxide (RGO), which
can function as elastic and electric conductive matrix to hold particles, and prevent
particles from losing electrical conduction. Finally, the electrode was prepared with
active material (carbon-coated porous silicon nanoparticles with reduced graphene oxide
wrapping, 70%), carbon black (super-P, 20%), and alginic acid sodium salt as binder
(10%, Alfa Aesar, denoted as alginate binder), and then assembled in coin cells for further
tests.
Figure 3.4a and b present the measured cyclic voltammetry curves of porous silicon
nanoparticles for the first two cycles. Undoped silicon nanoparticles (coated with carbon,
and wrapped with reduced graphene oxide) were also tested as a control group to
understand the effect of boron doping to the battery voltage. It was found that both curves
in Figure 3-4a show typical behavior of silicon in the lithiation and delithiation process.
The peak located at 0.5-1.0 V of the charge branch in the first cycle are related to the
formation of solid electrolyte interface (SEI) layer for both doped and undoped silicon
nanoparticles. After the first cycle, as shown in Figure 3-4b, the peak located at 0.2 V in
44
the charge branch in both curves indicates the formation of amorphous silicon.
Figure 3.4. (a-b) Cyclic voltammetry curves of doped porous silicon nanoparticles and undoped
nonporous silicon nanoparticles for the first cycle (a), and the second cycle (b), respectively. (c)
Differential capacity curves of doped porous silicon nanoparticles and undoped nonporous silicon
nanoparticles in the charge branch of the first cycle.
45
To get a detailed understanding of the effect of boron doping to battery voltage,
differential capacity as a function of potential for the first cycle was examined and
presented in Figure 3-4c. It is found that the intercalation voltage, which is related to the
lithiation of crystalline silicon (c-Si + nLi a-Li
n
Si)
2
, is 0.15 V for undoped silicon and
0.12 V for doped porous silicon (Peak A in Figure 3-4c). The lower intercalation voltage
in doped porous silicon might be attributed to the relatively higher conductivity of
particles due to boron doping, and relatively lower intercalation energy of lithium in
doped silicon, as we derived from First-principle density function theory (DFT)
calculation.
Theoretically, the intercalation voltage is associated with the Gibbs free energy of the
intercalation reaction, which can be obtained by DFT calculation
3,4
.
For reaction:
the average voltage is determined by
(1)
where F is Faraday ’s constant, and is change in the Gibbs free energy change.
r
G
can be calculated using
(2)
where is the change of internal energy, and is around the order of 0.1-4.0 eV/Li
atom. The term is of the order of 10
−5
eV, and is of the order of the thermal
15 4
Si 15 4Li Li Si
V
15 4
r
G
F
r
G
r r r r
G E P V T S
r
E
r
PV
r
TS
46
energy, which is much smaller than
5,6
. Therefore, we can calculate the
electrochemical voltage using the following as a good approximation:
(3)
To make the first principle calculation feasible, atomic models using Li
480
Si
128
(supercell
of Li
15
Si
4
, which is the structure of lithiated Si at room temperature
5
), Si
128
(supercell of
Si), Li
480
Si
127
B, and Si
127
B as unit cells were used for energy calculation. Regarding
atomic models of Si
127
B, the B atom was added to replace one of the Si atoms, and the
exact location does not affect the calculation results as the supercells take periodic
boundary condition and all the silicon atoms are equivalent.
The first-principle calculation was performed using the VASP code
6
. Density functional
theory (DFT) calculations in generalized gradient approximation (GGA) with the
Perdew-Burke-Ernzerhof (PBE) functional were used to optimize all structures and
obtain ground-state energy
7,8
. The projector augmented wave potentials
9
and Monkhorst-
Pack
10
k-points sampling were employed. Cut-off energy was set to 250 eV, and k-points
sampling was set to 2 x 2 x 2.
Calculation led to:
15 4 127 128 1 128
[Li Si B ] 13.681
r
E eV,
127 128 1 128
[Si B ] 5.429
r
E eV,
15 4
[Li Si] 13.688
r
E eV, and
[Si] 5.414
r
E eV, respectively.
r
E
15 4
r
E
V
F
47
According to equation (3), the difference of lithium intercalation voltage in doped and
undoped silicon can be calculated as:
15 4 127 128 1 128 127 128 1 128 15 4
doped undoped
( [Li Si B ] [Si B ])-( [Li Si] [Si])
6 mV
15 4
r r r r
E E E E
VV
F
which is consistent with our experimental finding that the lithium intercalation voltage
for undoped silicon is higher than doped silicon. In addition, we also performed similar
calculation for higher doping concentrations by using supercells of Si
63
B, Si
31
B, and
Si
15
B. Each supercell has one B atom replacing one of the Si atoms. Our calculation also
indicates that higher dopant concentration (uniform doping) leads to larger voltage
difference; however, the difference does not exceed 35 mV even when the dopant
concentration is as high as 6.7 %
Nevertheless, the difference of intercalation voltage between doped porous silicon
nanoparticles and undoped silicon nanoparticles is small. We can therefore conclude that
boron doping has no significant effect on the battery working voltage.
The cyclic performance of porous silicon nanoparticles as lithium-ion battery anode is
presented in Figure 3-5. The voltage range is set to 0.01-2.0 V . The rate performance of
porous silicon nanoparticles is demonstrated in Figure 3-5a. It was found that the capacity
remained around 2500, 2200, 1400 and 1000 mAh/g at current rates of 1/16 C, 1/8 C, 1/4
C, and 1/2 C (1 C = 4 Ah/g), respectively. The high capacity could be retained at
extended cycles, in which the capacity remained around 1400 and 1000 mAh/g at 1/4 C
48
and 1/2 C after 200 cycles, as shown in Figure 3-5b. The capacity degradation over 200
cycles is as small as 13 % for 1/4 C (from 1622 mAh/g for the first few cycles to 1410
mAh/g for the 200
th
cycle), and 19 % for 1/2 C (from 1172 mAh/g for the first few cycles
to 945 mAh/g for the 200
th
cycle).
The performance of our graphene-wrapped porous silicon nanoparticle anode compares
favorably with results in recent publications. For instance, Lee et al. reported using
nonporous silicon nanoparticles/graphene composites as lithium-ion battery anodes
11
;
however, their anodes exhibited noticeable capacity degradation around 47 % at 1/4 C
(from ~1900 mAh/g for the first few cycles to ~1000 mAh/g after 120 cycles). In
comparison, our graphene-wrapped porous silicon nanoparticle anodes exhibit much
better cyclability and much less capacity degradation. In addition, Liu et al. used porous
silicon particles without graphene wrapping and achieved anode capacity of 2826 mAh/g
for the first cycle and 1022 mAh/g for the 50
th
cycle at relatively low charging rate of
1/10 C [26]. In comparison, our anodes are able to operate at much higher current rates of
1/4 C and 1/2 C, and have delivered very stable capacity of 1400 and 1000 mAh/g after
200 cycles. We attribute the enhanced rate performance and cyclability of our anodes to
the unique structure of graphene-wrapped porous silicon nanoparticles, as porous silicon
nanoparticles can accommodate large volume change during cycling and provide large
surface area accessible to electrolyte, while the reduced graphene oxide wrapping can
serve as an elastic and electrically conductive matrix, and therefore boost the overall
49
battery performance.
Figure 3.5 Characterization of porous Silicon nanoparticles with carbon coating and reduced graphene
oxide wrapping as lithium-ion battery anode. (a) Charge/discharge capacity at current rates of 1/16 C,
1/8 C, 1/4 C, and 1/2 C, (1C = 4 Ah/g). (b) Cycling performance at current rates of 1/4 C and 1/2 C.
50
3.4 Conclusion
In summary, we have reported a facile doping and electroless etching method to prepare
porous silicon nanoparticles using silicon nanoparticles as starting materials that are
available in bulk quantity. We have shown that the porous silicon nanoparticles can be a
potential anode material for lithium-ion battery. Our battery tests have demonstrated that
porous silicon nanoparticle anode is able to deliver capacities around 1400 mAh/g and
1000 mAh/g at current rates of 1/4 C and 1/2 C, respectively, and show stable operation
up to 200 cycles. With the scalable and cost-efficient preparation method we have
reported, porous silicon nanoparticles may stimulate further study on their fundamental
properties, and may find broad applications for lithium-ion batteries, biomedical imaging,
and thermoelectric devices.
51
3.5 Chapter reference
1 Silva, J. A. et al. Sprayed boric acid as a dopant source for silicon ribbons. Sol Energ Mat Sol C
91, 1948-1953 (2007).
2 Chan, C. K., Ruffo, R., Hong, S. S., Huggins, R. A. & Cui, Y. Structural and electrochemical study
of the reaction of lithium with silicon nanowires. J Power Sources 189, 34-39 (2009).
3 Kubota, Y ., Escano, M. C. S., Nakanishi, H. & Kasai, H. Crystal and electronic structure of
Li15Si4. J Appl Phys 102 (2007).
4 Courtney, I. A., Tse, J. S., Mao, O., Hafner, J. & Dahn, J. R. Ab initio calculation of the lithium-tin
voltage profile. Phys Rev B 58, 15583-15588 (1998).
5 Obrovac, M. N. & Christensen, L. Structural changes in silicon anodes during lithium
insertion/extraction. Electrochem Solid St 7, A93-A96 (2004).
6 Kresse, G. & Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations
using a plane-wave basis set. Phys Rev B 54, 11169-11186, doi:DOI 10.1103/PhysRevB.54.11169
(1996).
7 Perdew, J. P. et al. Atoms, Molecules, Solids, and Surfaces - Applications of the Generalized
Gradient Approximation for Exchange and Correlation. Phys Rev B 46, 6671-6687, doi:DOI
10.1103/PhysRevB.46.6671 (1992).
8 Perdew, J. P., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys
Rev Lett 77, 3865-3868, doi:DOI 10.1103/PhysRevLett.77.3865 (1996).
9 Blochl, P. E. Projector Augmented-Wave Method. Phys Rev B 50, 17953-17979, doi:DOI
10.1103/PhysRevB.50.17953 (1994).
10 Monkhorst, H. J. & Pack, J. D. Special Points for Brillouin-Zone Integrations. Phys Rev B 13,
5188-5192, doi:DOI 10.1103/PhysRevB.13.5188 (1976).
11 Lee, J. K., Smith, K. B., Hayner, C. M. & Kung, H. H. Silicon nanoparticles-graphene paper
composites for Li ion battery anodes. Chem Commun 46, 2025-2027 (2010).
52
Chapter 4 Cost-efficient fabrication of porous silicon from bulk
metallurgical silicon for Li-ion battery anodes
4.1 Motivation
In the previous chapter, we have discussed a method to produce porous silicon
nanoparticles from silicon nanoparticles through a boron-doping and Ag-assisted
electroless etching method. It presents to be a scalable approach since the raw materials
as silicon nanoparticle and silver nitrite are available commercially in large quantity.
However, it is still a noticeable high production cost for the synthesis of silicon
nanoparticle itself, which involves high-energy laser ablation of bulk silicon, or high
temperature pyrolysis of toxic and expensive silane precursor (~ $50,000 / ton, compared
to ~ 1000 /ton for graphite)
1-8
. In addition, AgNO
3
is not a cheap reagent when scaled up
to industrial manufacture, and the noble metal waste may course environmental problems
as well. Based on our knowledge, there still lacks a cost-efficient way to obtain
nanostructured silicon, especially porous silicon to satisfy the requirement for mass
production. Table 4.1 summarizes the approaches that have been used in preparing porous
silicon. My previous work on porous silicon nanowires (Chapter 2) and porous silicon
nanoparticles (Chapter 3) are also included. Based on the analysis, efficient methods that
are capable of producing silicon nanostructures with scalability and production cost
similar to graphite are highly desirable, which may greatly impact the adoption of porous
silicon materials as anodes for lithium batteries.
53
Here we have developed a low cost and scalable method to convert bulk metallurgical
silicon into porous particles, which have been utilized as lithium-ion battery with high
capacity, fast charging rate, and long cyclic lifetime. Metallurgical silicon is chosen
because it is the cheapest silicon material (~ $1000 / ton) available in large amount for
industrial applications. Specifically, large pieces of metallurgical silicon were first ground
into fine powder using ball-mill. Then the powder was treated in ferric etchant composed
of Fe(NO
3
)
3
/HF, named stain-etching, to prepare porous silicon. The use of ferric etchant
eliminates the use of expensive AgNO
3
, in sharp contrast with early reports
9-11
.
Transmission electron microscopy (TEM) characterization confirms the highly porous
nature of the agglomerated particles. Moreover, STEM tomography - an imaging method
allowing a three-dimensional reconstruction of the porous silicon particles - clearly
shows that not only the surface but also the whole body of the particles are porous under
proper etching conditions. Electrochemical tests demonstrate that the lithium-ion battery
anodes making use of nanoporous silicon particles can deliver a stable capacity of 2900
mAh/g at a charging rate of 400 mA/g (0.1 C) with 10 cycles measured, and a capacity
above 1100 mAh/g at 2000 mA/g (0.5 C) with extended 600 cycles measured,
significantly better than early reports on nonporous silicon structures
5
.
54
Table 4.1 summery of the preparation method and battery performanceof porous silicon
Category Starting
material
Preparation
method
Structure Yields
*
Anode
performanc
e
Ref.
Non-
template
1 Si wafer Electroless etch Porous Si
nanowires
S 1000 mAh/g,
1500 cycles
12
2 Si
nanoparticle
s
Doping +
electroless etch
Porous Si
nanoparticles
M 1400 mAh/g,
200 cycles
13
3 Crushed Si
wafer
Electroless etch Hierarchical
porous Si
particles
M 1500 mAh/g,
50 cycles
10
4 Micro Si
particles
Electroless etch Macroporous Si
particles
L 2000 mAh/g,
50 cycles
11
5 Si wafer Electrochemical
etch
Free standing
porous Si film
S 1260 mAh/g,
20 cycles
14
6 Si wafer Electrochemical
etch
Porous Si film
with gold
coating
S 2000 mAh/g,
20 cycles
15
7 Si wafer Electrochemical
etch
Porous Si film
converted to
SiNP
S 1000 mAh/g,
600 cycles
16
Template
assisted
Template
8 SiO
2
opal CVD Inverse opal S 2200 mAh/g,
145 cycles
17
9 Irregular
SiO
2
sphere
pattern
CVD Interconnected
hollow Si
spheres
S 1420 mAh/g,
700 cycles
2
10 Porous Ni CVD 3D porous Si
particles
S 1650 mAh/g,
120 cycles
18
11 Irregular
SiO
2
sphere
pattern
Si gel
infiltration +
annealing
3D porous Si
particles
M 2820 mAh/g,
100 cycles
19
12 SBA-15
mesoporous
SiO
2
Si gel
infiltration +
annealing
Mesoporous Si M 2738 mAh/g,
80 cycles
20
13 Silica
diatom
Magnesiothermi
c reduction
Silica diatom
replica
M
21
14 Mesoporous
SiO
2
(KLE
copolymer)
Magnesiothermi
c reduction
Mesoporous Si M
22
15 SBA-15
mesoporous
SiO
2
Magnesiothermi
c reduction
Mesoporous Si M 1500 mAh/g,
100 cycles
23
*Yields: S = small quantity; M = Moderate quantity; L = large quantity
55
4.2 Preparation of nanoporous silicon from metallurgical silicon
A schematic diagram of synthesizing porous silicon from bulk metallurgical silicon is
presented in Figure 4.1a. In a typical experiment, metallurgical-grade silicon (Elkem Inc.)
was used as received. The purity of the silicon is around 99.2%, where iron and
aluminum are the two major impurities, as indicated by the supplier. The large
metallurgical silicon particles were first ground to obtain fine powder using ball-mill
(MTI Inc.) operated at a grinding speed of 1200 rpm for 5 hours. The ground powder has
a dark-brown color. Impurities of Fe and Al after ball milling were calibrated using
inductively coupled plasma atomic emission spectrometer (ICP-AES) measurements.
Specifically, 0.5 g sample was put into polytetrafluoroethene (PTFE) beaker containing 5
ml concentrated nitric acid (HNO
3
). 5 ml hydrofluoric acid (HF, 50 wt. %) was then
added by droplets to dissolve the sample to get a clear solution. The solution was added
with additional 5 ml acid mixture (sulfuric acid (H
2
SO
4
): phosphorous acid (H
3
PO
3
) = 3 :
2 by volume) and kept at 130
o
C for 5 min. At this stage, all the silicon was converted to
silicon tetrafluoride (SiF
4
) and removed by volatilization, leaving only boron in the
solution. The solution was then cooled down, and diluted with DI water. Fe and Al
concentration measurements were performed using HORIBA Jobin Yvon ULTIMA-C
inductively coupled plasma emission spectrometer (ICP-AES). After ball-milling, atomic
percentage of Fe and Al in silicon powder are 1.9% and 0.47%, respectively. The feature
size of the silicon powder after ball-milling is around 1 m, as shown in Figure 4-2.
56
Then, the obtained silicon powder was soaked into ferric etchant containing 30 mM
Fe(NO
3
)
3
and 5 M HF. Continuous mixing is required since hydrogen-terminated silicon
is hydrophobic and floats on the solution surface to form a foam. We found that the use of
ethanol suppress foam formation. After 2 hours of reaction, precipitates containing
porous silicon particles were collected and washed for further characterization (samples
from this etching condition are denoted as porous Si(1). Morphologies of particles after
etching are characterized using STEM and TEM (Figure 4.3a), which shows the porous
feature of Si(1) with a pore size around 3-10 nm.
Figure 4.1 Schematic diagram of synthetic route of porous silicon from metallurgical silicon through
ball-milling and stain-etching.
Figure 4.2 SEM image of silicon powder after ball-milling.
57
Figure 4.3 Schematic diagram of the synthesis and morphology of porous silicon particles. A.
Synthetic route of porous silicon from metallurgical silicon through ball-milling and stain-etching. B.
A STEM image of porous silicon particles. Inset: a TEM image showing the pore locations. C.
Reconstructed structure of the particles from HAADF-STEM 3D tomography. D-F. Projected views
of orthogonal slices cut through the center of the particle.
58
4.3 Characterization of nanoporous silicon particle
4.3.1. 3D tomogram of nanoporous silicon
A 3D tomography is powerful tool to detect the inner structure of a particle at sub-micron
to nanometer scale. It can be realized by taking a serial of high angle annular dark field
(HAADF) STEM images at different rotation angle. In this study, we take the HAADF-
STEM images at a single rotation from -75
o
to 75
o
at 1
o
interval, and a total of 151
images were aligned using ETOMO to reconstruct the 3D structure of particles. Figure
4.3b show the reconstructed tomogram of Si(1), wherein the projected views of
orthogonal slice cutting through the center of the particles are shown in Figure 4.3d-f. By
checking the slice views at different positions of the particle, we demonstrate the
reconstruction of the interior and exterior of such silicon-based porous particles for the
first time.
4.3.2. Nano-indentation of nanoporous silicon
A multimode AFM (Veeco Nanoscope IV) was used to carry out the imaging and force-
indentation experiment in the tapping mode. The silicon substrate was used as reference
to determine the optical sensitivity. The AFM probes used for measurement were silicon
probes with spring constant k=40 Nm
-1
and a tip radius of 25 nm. The force indentation
was fitted with Hertz model to determine the Young’s modulus
24
. Nano-indentation on
Si(1) indicates a Young’s modular of 2.43±0.29 GPa, which is much lower than bulk
59
silicon. The small Young’s modular may arise from the plastic-like deformation of the
porous structure
25
. Figure 4.4 shows the morphology of Si(1) before and after
indentation.
Figure 4.4 Morphology characterization of a porous silicon particle (Si(1)) before (a) and after (b)
indentation using an AFM tip. The protrusion found on the surface after indentation (b) indicates the
plastic-like deformation of the porous structure.
60
4.4 Mechanism of nanoporous silicon formation
We attribute the success of making porous silicon particles through a simple etching
method to the unique properties of metallurgical silicon in terms of unintentional high
impurity densities, as well as the selection of etchant components. A mechanistic study of
the porous silicon formation is interesting and would be beneficial to optimizing the
porous structure for various applications.
4.4.1 Calculation of electronic states of Fe, Al-doped silicon
For the metallurgical silicon we used in this study, Fe and Al are the two major
impurities. With the aid of first principle calculation, projected density of states (PDOS)
of Fe, Al and Si were calculated based on the atomic model of Fe and Al-doped silicon
(part of the atomic model is shown in Figure 4.5a)
26
. It was found that the electrons from
both Fe and Al atoms form a continuous defect band around the Fermi level of pure
silicon, while Fe contributes more states than Al does in the middle of the band-gap
(Figure 4.5b). When the particles are immersed in the etchant, the etchant solution will
act as a hole reservoir, wherein both Fe
3+
and NO
3
-
will oxidize silicon by donating holes
to the defect bands, as illustrated in the schematic diagram in Figure 4.5c. It is worth
noting that the electrons corresponding to the defect bands are highly localized.
Calculations show that electrons possessing energy within ±0.5 eV around the Fermi
level are spatially localized around the Fe atoms (yellow part in Figure 4.5d), and
61
therefore, etchant species like Fe
3+
and NO
3
-
are energetically favorable to inject holes to
the defect sites to implement etching and leave void space on silicon.
Figure 4-5. Mechanism of stain-etching to prepare porous silicon. (a) Part of atomic model containing
Fe-Al dopants in silicon used in first-principle calculations. The whole atomic model is a 3 x 3 x 3
supercell of silicon conventional cell. (b) First-principle calculations of projected density of states of
Fe, Al, and Si atoms. (c) Schematic diagram of band alignment of silicon and different redox pairs.
The etchant solution functions as a hole reservoir and can inject holes into the impurity band of
silicon. (d) Calculated spatial distribution of electrons with energy within ±0.5 eV of Fermi energy.
The yellow part indicates electron within this energy is mostly localized around Fe atom.
62
4.4.2. Effects of Fe
3+
and NO
3
-
in etching
In the etching process, there are several reactions taking place simultaneously.
Oxidization of silicon with Fe
3+
is straightforward; however, NO
3
-
has a complex
behavior in the presence of H
+
. Following equations demonstrate the reaction happened
in the etching process. As we see, both Fe
3+
and NO
3
-
can oxidize Si. The reaction
between Fe
3+
and Si is shown as:
3+ - 2+ 2-
6
4Fe + Si + 6F 4Fe + [SiF ]
The reaction between NO
3
-
and Si is shown as (in the present of HF):
-+
33
+-
2 3 2 2
2+ 2-
2
2+ -
2 6 2 2
NO H HNO
H NO HNO 2NO H O
2NO Si Si 2NO
Si 2(OH) 6HF H SiF H O H
We have carried out additional experiments to isolate the roles that Fe
3+
and NO
3
-
played
during the etching process. In the control experiments, Fe(NO
3
)
3
was substituted by FeCl
3
in the etchant (samples from this etch condition are denoted as Si(2)). Figure 4.6
compares the morphologies of the porous silicon (Si(1) and Si(2)) obtained from the two
different etching conditions. HAADF-STEM images of Si(1) (Figure 4-6a) and Si(2)
(Figure 4.6g) show that both samples are porous particles. BET measurements of Si(2)
imply a rather low specific surface area of around 19 m
2
/g, much lower than that of the
Si(1), suggesting a low porosity of the Si(2) particles. Morphology of both samples was
reconstructed through STEM tomography, as shown in Figure 4-6b for Si(1) and Figure
63
4-6h for Si(2). Figure 4-6c-f and Figure 4-6i-l are projected views of the two samples at
different slice positions (1-4). It is clear that Si(1) has the porous nature throughout the
whole particles; on the contrary, for Si(2), although the outmost surface layer is highly
porous, it retains solid without much pore formation in the inner part of the particles.
The comparisons between Si(1) and Si(2) suggest that NO
3
-
also has a positive effect to
the porous structure formation. In the presence of HF, weakly dissociated H
+
can increase
the redox potential of NO
3
-
to assist the oxidization of silicon at the defect sites. Another
control experiment was conducted to help to identify the role that NO
3
-
played in the
etching process. In the experiment, similar to the etch condition as for Si(1), Fe(NO
3
)
3
was reduced by 20% and replaced by HNO3 to keep total concentration of NO
3
-
a
constant. This sample is denoted as Si(3). Figure 4-7 compares the morphology of Si(1)
and Si(3) using SEM and TEM. The structure of the two samples is very similar, and both
of them clearly show the porous feature. The specific surface area of Si(3) is 80 m
2
/g,
quite comparable to the surface area of Si(1) which is 70 m
2
/g. This might be attributed to
the strong oxidization ability of NO
3-
at high H
+
concentration, which leads to the etching
of silicon at and also away from the defect sites.
64
Figure 4.6 Structure characterization of Si(1) and Si(2). HAADF-STEM images of Si(1) and Si(2) are
shown in A and G. Reconstructed structure from HAADF-STEM tomography for Si(1) and Si(2) are
shown in (b-f) and (h-l), respectively. (c-f) show the section planes of Si(1) at corresponding slice
positions shown in (b), which demonstrate the porous structure throughout the whole particle. (i-l)
show the section planes of Si(2) at corresponding slice positions shown in (h). It clear shows porous
structure at surface and solid (non-porous) structure at inner part of the particles.
65
Figure 4.7. Morphology comparison (SEM images: (a) and (e), TEM images: the rest) of porous
silicon etched using different etchant. (a-d) SEM and TEM images of Si(1) (etched in Fe(NO
3
)
3
/HF),
which show the highly porous structure of the particles, and individual pores can be discerned from
high-magnification TEM images (e.g. (d)). (e-h) SEM and TEM images of Si(3) (etched in
Fe(NO
3
)
3
/HNO
3
/HF), which show similar structure to Si(1).
66
4.5 Electrochemical tests and discussion
The electrochemical characteristics of porous silicon as lithium-ion battery anodes were
examined using a half-cell configuration. All samples were coated with carbon and then
coated with graphene oxide (GO) to increase the electrical conductivity of electrodes
(denoted as Si@GO). Figure 4A compares the cyclic performance of Si(1), Si(2), and un-
etched silicon powders. There was a severe capacity degradation for the un-etched silicon
when the battery was tested at a current rate of 0.1 C. Furthermore, the capacity of un-
etched silicon quickly dropped to 200 mAh/g when the current rate was increased to 0.2
C. Porous silicon (e.g. Si(2), etched using FeCl3/HF) showed much improved cyclic
performance, as it could deliver stable capacity > 800 mAh/g after 160 cycles at a 0.2 C
rate.
In contrast, Si(1) showed even better performance than Si(2). Capacity of Si(1) retained >
1400 mAh/g at a current rate of 0.2 C for 160 cycles. It is not a surprise that silicon
particles with higher porosity present better performance, as large particle-electrolyte
interface area can facilitate fast intercalation of Li
+
ions into silicon to improve the rate
performance at high current density. In addition, particles with high porosity can provide
sufficient void space to accommodate the large volume expansion during the lithiation
process, and therefore, help to keep electrode integrity without losing electric contact.
With the comparison of battery performance among samples, the improved performance
67
is related to the high surface area of the samples that enables fast intercalation of Li
+
ions
into silicon. For example, the apparent Li
+
ion diffusion rates are 3.4 x 10
-13
m
2
/s for Si(1)
and 6.8 x 10
-14
m
2
/s for Si(2) using potentiostatic intermittent titration technique31. It is
worth noting that the calculated values are not the intrinsic diffusion rates of Li
+
ions in
silicon, but a reflectance of particle geometry effect, as particles with higher surface area
allows more Li
+
ions diffusing into the body at a given time interval.
We have performed ex-situ TEM studies of porous silicon particles Si(1) at lithiated and
de-lithiated states. From the STEM image of the particle’s lithiated state (Figure 4.8b),
the particles remained porous-like, and no obvious cracks were found. Distribution of Li
in silicon was mapped out using electron energy loss spectroscopy and the coexistence of
Li and Si indicates the formation of Li-Si alloy (Figure 4.8c-d). It is worth noting that
there is a Li-rich region located at particle’s periphery, which corresponds to the
formation of solid-electrolyte interface (SEI). At de-lithiated state, silicon particles turn
back into the highly porous structure without crack initiation (Figure 4.8e). The structure
integrity kept at both lithiated and de-lithiated states demonstrates the capability of
porous structure in accommodating large volume expansion, thus promoting the battery’s
cyclic performance.
68
Figure 4-8. Electrochemical measurement and structure characterization of porous silicon
nanoparticles. (a) Cyclic performance of porous silicon of Si(1), Si(2) and un-etched silicon at current
rate of 0.2 C (1 C = 4 A/g, samples are coated with carbon and GO). (b) Ex-situ STEM images
showing the morphology of Sample 1after 1
st
lithiation, which shows the porous feature of, and
demonstrate the capability of porous structure to accommodate volume expansion of the particle
during the lithiation process. (c) EELS mapping of Li (red) and Si (blue) at the enlarged region of B.
The peripheral Li-rich region suggests the formation of solid electrolyte interface (SEI). (d) EELS
spectroscopy indicates the signal from Li and Si at the enlarged region in figure (c). (e) Ex-situ TEM
image showing the morphology of Si(1) after 10 cycles and being charged to 2V (de-lithiated state).
The particle keeps integrated without crack.
69
The performance of Si@GO (Si from Si(1)) can be further improved by annealing the
samples before assembling into a battery, since conductivity of GO can be dramatically
increased when annealed at high temperatures comparing with using chemical reduction
methods
27
. As shown in Figure 4-9a, the samples annealed at 700
o
C show excellent
performance: the capacity stabilized at different current rates from 2900 mAh/g at 0.1 C
to 1000 mAh/g at 2 C. Figure 4-9b shows the discharge capacity of the same sample at
extended cycles. The capacity remained above 1100 mAh/g and 550 mAh/g at rates of 0.5
C and 1.0 C over 600 cycles, respectively, and the Coulombic efficiency is higher than
99% starting from the 6th cycle and remained stable for the rest of cycles.
70
Figure 4.9 Electrochemical measurements and structure characterization of porous silicon
nanoparticles with GO after annealing in Ar at 700
o
C. (a) Discharge capacity of the samples at
different current rates. (b) Cyclic performance of the samples at current rates of 0.5 C and 1.0 C
(samples are pre-charged/discharged at 0.1 C for 10 cycles). The Coulombic efficiency stables above
99% for all the cycles (the Coulombic efficiency curve is for sample cycled at 0.5 C).
71
4.6 Discussion and conclusion
Compared with other reported work on silicon anodes, our study offers many advantages
and may generate broad interests in and beyond the battery research field:
(i) Metallurgical silicon is the cheapest silicon material available with a comparable price
(~ $1000 / ton) to battery-grade graphite, and the abundance of metallurgical silicon
fulfills the prerequisite for scalable production.
(ii) Ball-mill as a mature technology employed in the process is simple, eco-friendly, and
energy efficient.
(iii) We used ferric nitrate/hydrofluoric acid (Fe(NO
3
)
3
/HF) as the etchant to prepare
porous silicon to eliminate the use of AgNO
3
in metal-assisted electroless etching, taking
the advantage that Fe
3+
has similar oxidant potential as Ag
+ 28
, thus, presenting a cost-
efficient manufacturing approach. In addition, this etching method can be extended to
prepare other nanoporous silicon with different starting materials, which might find
interest in multiple disciplines, e.g. catalysis, biosensing, drug delivery, etc..
(iv) We demonstrate the reconstruction of nanoporous silicon particles using STEM
tomography for the first time, which helps to build up the connection between etching
72
condition, material structure, and battery performance. In addition, the tomography
technique demonstrates itself as a powerful tool to visualize the inner structure of
material in nano-scale.
(v) Overall, the remarkable simplicity of the etching procedure and inexpensive etchant
in conjunction with the good electrode performance would lead to a significant progress
toward high-performance, low-cost lithium-ion batteries.
In conclusion, we have developed a low-cost and scalable approach to prepare porous
silicon particles from bulk metallurgical silicon. The impurities inside the silicon have a
profound effect on the formation of porous structure when using Fe(NO
3
)
3
/HF as etchant.
The STEM tomography reconstruction shows that the whole body of the as-prepared
silicon particles is porous. We have further demonstrated the outstanding performance of
using porous silicon as lithium-ion battery anode. Electrodes making use of porous
silicon particles can achieve a stable capacity of 2900 mAh/g at a charging rate of 0.1 C
with 10 cycles measured, and a capacity above1100 mAh/g at 0.5 C with extended 600
cycles measured. The overall performance is superior to the materials prepared by many
other synthetic approaches. We anticipate that this method will boost the development of
silicon anode towards real applications.
Figure 4.10 compares the Si anode performance from this work and some recent
73
progress
2-7,10,12,13,18,29-34
. The approach in this study compares more favorable than many
other approaches in the production cost. For example, while the reported capacity of
porous Si prepared through a disproportion reaction of SiO is comparable to our results,
SiO is of higher price than the metallurgical Si studied in this work, and the disproportion
of SiO at high temperature requires large thermal budget.
35,36
Figure 4.10 Evaluation of silicon anodes in terms of capacity, cycle number, current rate, and
production cost. The color scheme represents cycle number; the symbol represents current rate; the
numbers next to each symbol refer to the reference index.
74
We believe further optimization of porous structures will lead to an even higher capacity
and longer cyclic life. Study of the compatibility of Si anode with cathode materials such
as LiMn
2
O
4
, LiFePO
4
, and sulfur is also deserved further endeavor to harvest the full
potential of the porous Si anodes.
In conclusion, we have developed a low-cost and scalable approach to prepare porous Si
particles from bulk metallurgical Si. The impurities inside the Si have a profound effect
on the formation of porous structure when using Fe(NO
3
)
3
/HF as etchant. Extensive study
on the STEM tomography reconstruction reveals the highly porous feature of Si particles
under proper etching condition. Electrodes making use of porous Si particles have shown
encouraging battery performance, as they can achieve stable capacity of 2900 mAh/g at a
charging rate of 0.1 C with 10 cycles measured, and capacity above1100 mAh/g at 0.5 C
with extended 600 cycles measured. The overall performance is superior to the materials
prepared by many other synthetic approaches, especially in term of the production cost.
We anticipate that this method will boost the development of Si anode towards real
applications.
75
4.7 Chapter reference
1 Chan, C. K. et al. High-performance lithium battery anodes using silicon nanowires. Nat
Nanotechnol 3, 31-35, doi:DOI 10.1038/nnano.2007.411 (2008).
2 Yao, Y . et al. Interconnected Silicon Hollow Nanospheres for Lithium-Ion Battery Anodes with
Long Cycle Life. Nano Lett 11, 2949-2954, doi:Doi 10.1021/Nl201470j (2011).
3 Wu, H. et al. Engineering Empty Space between Si Nanoparticles for Lithium-Ion Battery Anodes.
Nano Lett 12, 904-909 (2012).
4 Wu, H. et al. Stable Li-ion battery anodes by in-situ polymerization of conducting hydrogel to
conformally coat silicon nanoparticles. Nature Communications 4, 6 (2013).
5 Chen, X. L. et al. Conductive Rigid Skeleton Supported Silicon as High-Performance Li-Ion
Battery Anodes. Nano Lett 12, 4124-4130 (2012).
6 Liu, N. et al. A Yolk-Shell Design for Stabilized and Scalable Li-Ion Battery Alloy Anodes. Nano
Lett 12, 3315-3321 (2012).
7 Chou, S. L. et al. Enhanced reversible lithium storage in a nanosize silicon/graphene composite.
Electrochem Commun 12, 303-306 (2010).
8 Zhou, X. S., Yin, Y . X., Wan, L. J. & Guo, Y . G. Facile synthesis of silicon nanoparticles inserted
into graphene sheets as improved anode materials for lithium-ion batteries. Chem Commun 48,
2198-2200 (2012).
9 Bang, B. M., Kim, H., Song, H. K., Cho, J. & Park, S. Scalable approach to multi-dimensional
bulk Si anodes via metal-assisted chemical etching. Energ Environ Sci 4, 5013-5019 (2011).
10 Zhao, Y ., Liu, X. Z., Li, H. Q., Zhai, T. Y . & Zhou, H. S. Hierarchical micro/nano porous silicon
Li-ion battery anodes. Chem Commun 48, 5079-5081, doi:Doi 10.1039/C2cc31476b (2012).
11 Bang, B. M., Lee, J. I., Kim, H., Cho, J. & Park, S. High-Performance Macroporous Bulk Silicon
Anodes Synthesized by Template-Free Chemical Etching. Adv Energy Mater 2, 878-883, doi:DOI
10.1002/aenm.201100765 (2012).
12 Ge, M. Y ., Rong, J. P., Fang, X. & Zhou, C. W. Porous Doped Silicon Nanowires for Lithium Ion
Battery Anode with Long Cycle Life. Nano Lett 12, 2318-2323, doi:Doi 10.1021/Nl300206e
(2012).
13 Ge, M. et al. Scalable production of porous silicon nanoparticles and their application in lithium-
ion battery anode. Nano Research 6, 174-181 (2012).
14 Thakur, M. et al. Freestanding Macroporous Silicon and Pyrolyzed Polyacrylonitrile As a
Composite Anode for Lithium Ion Batteries. Chem Mater 24, 2998-3003, doi:Doi
10.1021/Cm301376t (2012).
15 Thakur, M., Isaacson, M., Sinsabaugh, S. L., Wong, M. S. & Biswal, S. L. Gold-coated porous
76
silicon films as anodes for lithium ion batteries. J Power Sources 205, 426-432 (2012).
16 Thakur, M., Sinsabaugh, S. L., Isaacson, M. J., Wong, M. S. & Biswal, S. L. Inexpensive method
for producing macroporous silicon particulates (MPSPs) with pyrolyzed polyacrylonitrile for
lithium ion batteries. Sci Rep-Uk 2, doi:Artn 795 (2012).
17 Esmanski, A. & Ozin, G. A. Silicon Inverse-Opal-Based Macroporous Materials as Negative
Electrodes for Lithium Ion Batteries. Adv Funct Mater 19, 1999-2010 (2009).
18 Gowda, S. R. et al. Three-Dimensionally Engineered Porous Silicon Electrodes for Li Ion
Batteries. Nano Lett 12, 6060-6065 (2012).
19 Kim, H., Han, B., Choo, J. & Cho, J. Three-Dimensional Porous Silicon Particles for Use in High-
Performance Lithium Secondary Batteries. Angew Chem Int Edit 47, 10151-10154 (2008).
20 Kim, H. & Cho, J. Superior Lithium Electroactive Mesoporous Si@Carbon Core-Shell Nanowires
for Lithium Battery Anode Material. Nano Lett 8, 3688-3691 (2008).
21 Bao, Z. H. et al. Chemical reduction of three-dimensional silica micro-assemblies into
microporous silicon replicas. Nature 446, 172-175 (2007).
22 Richman, E. K., Kang, C. B., Brezesinski, T. & Tolbert, S. H. Ordered mesoporous silicon through
magnesium reduction of polymer templated silica thin films. Nano Lett 8, 3075-3079 (2008).
23 Jia, H. P. et al. Novel Three-Dimensional Mesoporous Silicon for High Power Lithium-Ion Battery
Anode Material. Adv Energy Mater 1, 1036-1039 (2011).
24 Sneddon, I. N. The relation between load and penetration in the axisymmetric boussinesq problem
for a punch of arbitrary profile. International Journal of Engineering Science 3, 47-57 (1965).
25 Charitidis, C. A., Skarmoutsou, A., Nassiopoulou, A. G. & Dragoneas, A. Nanomechanical
properties of thick porous silicon layers grown on p- and p plus -type bulk crystalline Si. Mat Sci
Eng a-Struct 528, 8715-8722 (2011).
26 Vankooten, J. J., Weller, G. A. & Ammerlaan, C. A. J. Electron-Paramagnetic Resonance on Iron-
Acceptor Pairs in Silicon. Phys Rev B 30, 4564-4570 (1984).
27 Pei, S. F. & Cheng, H. M. The reduction of graphene oxide. Carbon 50, 3210-3228 (2012).
28 Dudley, M. E. & Kolasinski, K. W. Stain Etching with Fe(III), V(V), and Ce(IV) to Form
Microporous Silicon. Electrochem Solid St 12, D22-D26 (2009).
29 Kovalenko, I. et al. A Major Constituent of Brown Algae for Use in High-Capacity Li-Ion
Batteries. Science 334, 75-79 (2011).
30 Wu, H. et al. Stable cycling of double-walled silicon nanotube battery anodes through solid-
electrolyte interphase control. Nat Nanotechnol 7, 309-314 (2012).
31 Yoo, J. K., Kim, J., Jung, Y . S. & Kang, K. Scalable Fabrication of Silicon Nanotubes and their
Application to Energy Storage. Adv Mater 24, 5452-5456 (2012).
32 Liu, N., Huo, K. F., McDowell, M. T., Zhao, J. & Cui, Y . Rice husks as a sustainable source of
nanostructured silicon for high performance Li-ion battery anodes. Sci Rep-Uk 3, 7 (2013).
77
33 Yi, R., Dai, F., Gordin, M. L., Chen, S. R. & Wang, D. H. Micro-sized Si-C Composite with
Interconnected Nanoscale Building Blocks as High-Performance Anodes for Practical Application
in Lithium-Ion Batteries. Adv Energy Mater 3, 295-300 (2013).
34 Yi, R., Dai, F., Gordin, M. L., Sohn, H. & Wang, D. H. Influence of Silicon Nanoscale Building
Blocks Size and Carbon Coating on the Performance of Micro-Sized Si-C Composite Li-Ion
Anodes. Adv Energy Mater 3, 1507-1515 (2013).
78
Chapter 5 Structure evolution in dealloyed Li-Si
x
Ge
1-x
ternary
system
5.1 Motivation
Previously, we have understood that the structure stability of silicon during the lithiation
and delithiation process is of central importance to enable high capacity and long cyclic
life of silicon anode. Porous silicon, due to its unique structure, well has the capability to
accommodate its large volume change, thus to maintain the structure stability.
At this moment, it will be of great interests to ask another question: apart from initial
structure design of silicon, whether there exist some other approaches to improve the
silicon electrode performance, for example, by alloying silicon with other elements? The
idea of alloying comes from a simple concept. For all material systems, the entropy of the
system will increase with addition of other elements due to inter-mixing. Thus, the silicon
anode may be stabilized with the mediation of additional atom species. In this chapter, I
will mainly discuss the structure evolution of Li-SiGe alloy nanoparticles under the
dealloying (or saying, delithiation) process.
The structure of this chapter is organized as following. First, I will give a brief
introduction of alloying, which overviews the study of dealloying in binary metal
systems, and the limitation of current research scope will be discussed. In the second part,
79
I will describe my experimental results and give detailed discussion on the morphology
evolution for lithiated Si-Ge alloy nanoparticles. Outlook of this research will be
provided at the end of this chapter.
80
5.2 Introduction of dealloying
Dealloying is a common corrosion process that involves a selective dissolution of active
components in an alloy. It has become an important technique in fabricating nanoporous
materials for a variety of applications including catalysis
1-3
, sensing
4
, and
supercapacitors
5
. For decades, substantial efforts have been devoted to understand the
mechanism behind the microscopic view of the structure evolution under the dealloying
process, which plays a pivotal role in developing new strategies to fine tune the
morphology and structure to optimize materials’ functionality for different purposes.
Binary alloy (in form of A
x
M
1-x
) is a model system due to its compositional simplicity.
The dealloying behavior of a binary alloy can roughly be divided into two categories,
depending on the atomic bond type between the component A and M. If A and M form
metallic bonds and have a significant difference in their metal ion/metal equilibrium
potentials, there exists a critical electrochemical potential (V
crit
)
6,7
, above which,
component A spontaneously dissolves, and a large family of metallic alloys such as
noble-metal systems (e.g. Au-Ag)
8,9
belong to this category. In the other case, when A
and M form mixed metallic and covalent bonds, dissolution of the more active
component A is always concomitant with significant solid-state diffusive transport and
may experience several phase transitions
10
. The morphology evolution for such A
x
M
1-x
with mixed metallic and covalent bonds under dealloying is much complicated, and many
factors such as particle size, dealloying rate, and composition form a complicated
81
interplay to the dealloyed structure
11
. For years, studies on different types of alloy
systems have accumulated essential knowledge to describe the dealloying behavior in
binary systems; however, multi-component alloy (e.g. ternary alloy) has rarely been
studied due to its complexity, yet is of significant importance, as adding additional
components to the alloy offers another degree of freedom to tune the dealloyed structure
and correlated properties. Up to now, understanding of the dealloying behavior for multi-
component alloy, especially for the alloy with mixed metallic and covalent bonds, is
incomplete in both morphological aspect and mechanistic point of view.
In noble-metal alloy systems, dealloying occurs at a critical electrochemical potential
when the more active component in an alloy (e.g. Ag in Au-Ag system) reaches a
threshold concentration to form anatomic-scale conduit. A continuous dissolution of the
active component results in a formation of nanoporous structure, which is attenuated by
the surface diffusion and accumulation of the more noble metal atoms (e.g. Au), behaving
like a Stranski–Krastanov growth. A continuum model proposed by Sieradzki and the
kinetic Monte-Carlo simulation based on Cahn-Hilliard diffusion equation, which
combines the normal diffusion and spinodal decomposition, well capture the dealloying
feature for noble-metal alloys
12-14
. In contrast, for A
x
M
1-x
systems with mixed metallic
and covalent bonds, there is considerable volume diffusion (lattice diffusion) and large
volume change accompanied with the dissolution of A. One example of this system
is Li-
Sn alloy
11
. Studies on the delithiation of Li-Sn system showed that Sn would evolve into
82
a bicontinuous structure even at a low Li concentration that is far below the required
threshold concentration for noble-metal alloys (such as Au-Ag system), and the
observation was attributed to the fast diffusion of the host species (Sn atoms) under
ambient-temperature dealloying. Close examination of the solid-state diffusion of the
host species in a multi-component alloy (e.g., Li-SiGe ternary alloy) is interesting, and
may serve as a first try to disclose the underlying physics governing the morphology
evolution in a general case.
Ternary alloy Li-SiGe was chosen as the studying material. The simplicity of this
prototypical alloy lies in the almost identical dealloying behavior of Li-Si and Li-Ge
binary systems, which make the ternary alloy an ideal choice to identify the effect from
the added components in the dealloying process. Analysis on the Si-Ge elemental
distribution also allows us to make assessments on the atomic diffusion at different
dealloying conditions. In Li-Si binary system, dissolution of Li from Li
15
Si
4
experiences
large volume shrinkage, resulting in an amorphous silicon structure. Despite the existence
of many crystalline Li
x
Si
y
phases in the phase diagram
15
, it is interesting that an nearly
isotropic amorphous phase is found under the extraction of Li via a single-phase
mechanism, as no Li concentration boundary is clearly detected based on in-situ TEM
observation of lithiated Si nanoparticles
16
. Under repeated lithiation/delithiation cycles,
Si would evolve into a porous structure with a ligament length around 7 nm
17
. Similar
structural and electrochemical properties were identified in Ge in previous reports
18-21
83
5.3 Preparation of lithiated-SiGe nanoparticles and characterization
methods
5.3.1 Sample preparation
Si
x
Ge
1-x
nanoparticles were synthesized using CO
2
laser pyrolysis of gas mixture as
reported previously
22,23
. The focused laser beam (wavelength 10.6 μm) was aligned to
cross the gas flow line perpendicularly within the reaction chamber. The gas mixtures
were jetted from two coaxially aligned tubes; reaction gas mixtures flew through the
inner tube and the He gas flow through the outer tube to confine reaction gases and
products. All the gas flows were regulated by mass flow controllers. SiH
4
(99.9999 %,
Daesung Industrial Gases Co., Ltd) and GeH
4
(99.999 %, 10 % diluted in H
2
, V oltaix
Inc.) gases were used as feed stocks for Si
x
Ge
1-x
nanoparticles, and SF
6
was used as
photosensitizer gas to promote dissociation of SiH
4
and GeH
4
. The flow of SiH
4
was
fixed to 25 standard cubic centimeter per minute (sccm) while the flow of GeH
4
was
varied to control the composition ratio of Ge within the Si
x
Ge
1-x
alloy nanoparticles. The
pressure of the reaction chamber was maintained to be 400 Torr during the synthesis
process. The nanoparticles were collected at a membrane filter installed in a tightly sealed
collector and transferred to the glove box without exposure to the air.
To produce Li-SiGe alloy, Si
x
Ge
1-x
particles were electrochemically lithiated in the
potential window of 0.01-2V vs. Li
+
/Li in a coin-cell configuration. After cycles, cells
84
were disassembled at charged state (charged to 2 V) to fully extract lithium out of
previously formed Li-SiGe alloy.
5.3.2 Electrode preparation
The Si
x
Ge
1-x
electrodes were fabricated by pasting the slurry (Si
x
Ge
1-x
: carbon-black :
alginic acid sodium salt = 7 : 2 : 1 by weight) on a copper foil and then dried at 90
o
C for
6 hours. Later, coin cells were assembled using a lithium foil as the counter/reference
electrode. 1 M LiPF
6
in dimethyl carbonate / fluoroethylene carbonate (1:1 by volume)
was used as electrolyte.
5.3.3 Microscopy
General STEM images and EDX data were acquired inside an aberration-corrected JEOL
JEM-ARM200CF STEM equipped with a 200 keV Schottky cold-field emission gun,
HAADF, and a EDX spectrometer (X-Max
N
, 100TLE Oxford Instruments). The STEM
provides spatial resolution of ~0.8 Å. Here, a 22-mrad-probe convergence angle was used
for all images and spectra. The HAADF images were acquired using a 90-mrad inner-
detector angle.
5.3.4 Raman spectra of Si
x
Ge
1-x
Raman spectra of Si
x
Ge
1-x
were collected at reflective micro-Raman station. A laser beam
with a wavelength of 514 nm and a spot size of 50 m was used as incident light source.
85
5.3.5 XRD spectra of Si
x
Ge
1-x
XRD was performed at Rigaku Ultima IV with a scanning rate of 5
degrees per minute.
5.3.6 Ab-initio molecular dynamic simulation (Ab-MD) and ground energy
calculation
Ab-initio molecular dynamic simulation were performed using the V ASP code
24
. The
ultrasoft peusdo-potentials were used to model electron–ion interactions. Structure
relaxation was carried out using the generalized gradient approximation (GGA) for
exchange-correlation function
25
. The k-space integration was done by summing over the
Monkhorst-Pack special points in the Brillouin zone. The Gaussian method, with a
smearing width of 0.05eV was used in determination of the Fermi level. The plane wave
basis set was restricted by cutoff energy of 425 eV . A periodic nanocubic containing
Li
480
Si
64
Ge
64
atoms was constructed with a vacant layer of 15 nm in 3 dimensions to
eliminate the interference from neighboring mirror structure. 1x1x1 k-point mesh was
used for energy calculations. To simulation the delithiation (dealloying) process of Li-
SiGe, Li atoms were removed from the outer most layers after every 50fs MD simulation.
Si-Ge radial distribution function (RDF) was calculated during the MD simulation.
86
5.4 Results and discussion
5.4.1 Characterization of Si
x
Ge
1-x
alloy nanoparticles
Atomic concentration of silicon in as-produced Si
x
Ge
1-x
alloy nanoparticles increases as
the molar ratio of SiH
4
gas to GeH
4
increases as shown in Figure 5.1 based on EDX
analysis. This implies that the composition ratio of Si
x
Ge
1-x
was successfully controlled
by regulating SiH
4
and GeH
4
gas mixtures.
Crystal structure of SiGe alloy nanoparticles were analyzed by XRD. Figure 5.2 shows
the diffraction pattern of alloy particles with different composition. The peak shifts
among samples indicate the composition difference. Pure silicon nanoparticles and
standard Si and Ge diffraction lines are also included for comprehensive comparison.
Figure 5.1 SiH
4
molar ratio to GeH
4
during the synthesis versus silicon atomic ratio in Si
x
Ge
1-x
alloy
nanoparticles.
87
Figure 5.2 X-ray diffraction (XRD) pattern of as-synthesized Si
x
Ge
1-x
alloy particle with different
composition.
88
The composition difference was further confirmed through the energy dispersive x-ray
analysis (EDX) performed with a transmission electron microscope (TEM). The atomic
Si
x
Ge
1-x
concentration in these nanoparticles was determined to be Si
100
Ge
0
, Si
77
Ge
23
,
Si
64
Ge
36
, Si
54
Ge
46
, Si
23
Ge
77
, Si
0
Ge
100
, respectively, which correlates with the molar ratio
of SiH
4
-to-GeH
4
gas precursor (as previously shown in Figure 5.1). Structures of all
samples with different Si
x
Ge
1-x
composition are quite similar, as all samples show
spherical morphology as illustrated in Figure 5.3.
Figure 5.3 TEM images of Si
x
Ge
1-x
nanoparticles with composition of Si
77
Ge
23
(a), Si
54
Ge
46
(b),
Si
23
Ge
77
(c), respectively. Scale bar is 30 nm for all images.
89
For detailed characterization, here we take the Si
77
Ge
23
sample as an example and present
its typical results. TEM images in Figure 5.4a-b show the particle morphology at different
magnifications. The particles are spherical with relatively uniform size around 30 nm in
diameter. The highly crystallized structure is revealed in the high-angle annular dark field
scanning TEM (HAADF-STEM) image in the inset of Figure 5.4b. It is interesting to
note that there exists subtle contrast difference of the atoms in the HAADF image,
indicating the non-homogeneity of Si-Ge distribution at atomic scale. In addition, the
bright region at the outer surface indicates Ge enrichment, which was further
characterized and clearly visualized in the Si-Ge elemental distribution map based on the
EDX image shown in Figure 5.4c, as enriched Ge can be found at the periphery of the
particles. The Ge enrichment (or segregation) at particle surface will lower the surface
energy to stabilize the SiGe nanostructure, which has been previously studied in the
epitaxial growed SiGe alloy
26
.
90
Figure 5.4 Structure characterization of as-synthesized Si
77
Ge
23
nanoparticles. a-b, TEM (a) and
HAADF-STEM (b) images of Si
77
Ge
23
nanoparticles at different magnifications. The subtle contrast
difference in the inset of b indicates the non-homogeneity of Si-Ge distribution at atomic scale. c,
elemental distribution of Si (blue) and Ge (yellow) of the Si
77
Ge
23
sample. A schematic model in the
inset of c illustrates the non-homogeneity of Si-Ge. d, Raman spectrum of Si
77
Ge
23
using 514 nm laser
as incident beam. The Raman shifts around 300 cm
-1
, 400 cm
-1
and 510 cm
-1
correspond to the
vibration of Ge-Ge, Si-Ge, and Si-Si, respectively. e, XRD pattern indicates the single phase of
Si
77
Ge
23
at particle scale. Scale bar is 100 nm, 10 nm, 2 nm, and 20 nm in the images of a, b, inset of
b, and c, respectively.
91
In the Raman scattering spectra, coexistence of Raman shifts of Si-Si (510 cm
-1
), Si-Ge
(400 cm
-1
) and Ge-Ge (300 cm
-1
) vibrations also indicate non-homogeneous feature of the
alloy sample at atomic scale (Figure 5.4d). At particle scale, however, Si
77
Ge
23
shows
homogenized feature, as illustrated from the single phase indexed in the powder X-ray
diffraction pattern in Figure 5.4f. Si
x
Ge
1-x
particles with other compositions are
characterized with similar morphology and structure. Comprehensive Raman
characterization can be found in Figure 5.5.
Figure 5.5 Raman spectra of Si
x
Ge
1-x
with different composition. Coexistence of Raman shifts of Si-Si
(510 cm
-1
), Si-Ge (400 cm
-1
) and Ge-Ge (300 cm
-1
) vibrations also indicate non-homogeneous feature
of the alloy sample at atomic scale.
92
5.4.2 Characterization of delithiated Li-Si
x
Ge
1-x
at different cyclic states
Figure 5.6 shows the morphology of delithiated samples (Si
77
Ge
23
, Si
54
Ge
46,
and
Si
23
Ge
77
) after the 1
st
cycle. The colored images indicate the elemental distribution of Si
(blue), Ge (yellow) and overlap of Si+Ge under energy-filtered TEM (EFTEM). It is
interesting to note that the individual nano particles before cycling (Figure 5.4) are
merged together and form a large structure after the 1
st
lithiation/delithiation cycle for
three samples, indicating a strong inter-particle interaction, e.g. atomic diffusion between
contacted neighboring particles. The phenomenon of particle merging is absent in the
previous reports on the in-situ TEM study of isolated Si nanoparticles or nanowires
27,28
,
suggesting a complicated collective behavior of electrode materials at confined
environment (e.g. particles are compacted in battery case) compared to that in interaction-
free environment. Figure 5.7 shows the delithiated particle after the 5
th
cycles. There is no
significant morphology change observed. It is noteworthy to mention that the elemental
distribution of Si and Ge is still inhomogeneous after the 1
st
and 5
th
cycle, as indicated in
Figure 5-6 and Figure 5-7.
93
Figure 5.6 Morphology and SiGe distribution after the 1
st
cycle for Si
77
Ge
23
, Si
54
Ge
46,
and Si
23
Ge
77
,
respectively. Energy-filtered TEM was used to determine Si-Ge elemental distribution.
Figure 5.7 Morphology and SiGe distribution after the 5
th
cycle for Si
77
Ge
23
, Si
54
Ge
46,
and Si
23
Ge
77
,
respectively. Energy-filtered TEM was used to determine Si-Ge elemental distribution.
94
However, it is interesting to note that after 100 cycles, all three samples have loose
structures (Figure 5.8e, i, and o. For clear comparison, samples after the 1
st
and 5
th
cycle
were also included in Figure 5.8). It is noticeable that the Si-rich sample (Si
77
Ge
23
, Figure
2e-f) and Ge-rich sample (Si
23
Ge
77
, Figure 2o-r) are fluffy and show similar morphology
with larger ligaments and enclosing larger void holes than that of the Si
54
Ge
46
sample.
It is worth noting that the difference among the three samples is reflected on their
capacity retention behavior as lithium-ion battery anodes. Figure 5.9 presents the
charge/discharge curve, and Figure 5.10 summarizes the electrode capacities for each
Si
x
Ge
1-x
composition at different cyclic states. If we define the capacity retention as the
percentage of the capacity at 100
th
cycle with respect to the capacity at 50
th
cycle, it is
found that the Si
x
Ge
1-x
alloy with composition ratio approaching 1:1 (Si
54
Ge
46
sample)
has higher capacity retention than both Si-rich and Ge-rich alloy.
95
Figure 5.8 TEM images of (a-b), (g-h), and (m-n) show the morphology of Si
77
Ge
23
, Si
54
Ge
46
, and
Si
23
Ge
77
after the 1
st
and 5
th
cycle, respectively. It is clear the particles for all three samples are merged
together. After 100
th
cycle, particles evolve into loose structures, as shown in the STEM images (c, i,
o). EDX mapping is performed within the dashed square regions in STEM. EDX of Si (d, j, p) and Ge
(e, k, q), and combination of Si+Ge (f, l, r) demonstrate that Si and Ge have almost homogeneous
distribution after 100
th
cycles. Scale bar is 100 nm for all images.
96
Figure 5.9 (a) measured capacity of Si
x
Ge
1-x
nanoparticles at difference current rates. (b) cyclic test
for additional 100 cycles at current rate of 0.8A/g after rate test shown in (a)
Figure 5.10 Electrochemical performance of Si
x
Ge
1-x
nanoparticles as lithium-ion battery anode.
Capacities at 25
th
, 50
th
, 75
th
, and 100
th
cycle of Si
x
Ge
1-x
were compared at a current rate of 0.8A/g.
Capacity retention of Si
x
Ge
1-x
nanoparticles defined as the ratio of capacity at 100
th
cycle with respect
to capacity at 50
th
cycle is shown as the square symbol in the figure.
97
It is reasonable to relate the different capacity retention behavior to their structure
difference for the various Si
x
Ge
1-x
compositions. We suggest for the Si-rich and Ge-rich
samples, the continuous structure-evolution from merged particles to the resultant fluffy
structure would expose more surface area to electrolyte than the Si
54
Ge
46
sample,
resulting a formation of larger portion of solid electrolyte interface (SEI), leading to the
successive capacity degradation as indicated in Figure 5.10. On the contrary, the Si
54
Ge
46
sample evolves into a porous structure under the same tested cycles. The porous structure
with balanced Si-Ge composition ratio may be more stable against the volume change
during the lithiation/delithiation process than other compositions, and prevent itself to
form the fluffy structure to lose capacity continuously.
Considering the similar chemical properties of Si and Ge species, the composition-
dependent morphology evolution in the dealloyed Li-SiGe system is unexpected. To
uncover the underlying physical mechanism, it is interesting to ask two questions:
1. For certain SixGe1-x composition, how does the Si-Ge atomic distribution change
during the structure evolution?
2. What is the driving force for structure evolution, e.g. why are particles merged
after cycling?
Figure 5.8 (d-f, j-l, p-r) illustrates the elemental distribution of Si and Ge after 100
th
cycle. It is found that Si and Ge are uniformly distributed in the amorphous porous
structures in all three samples, which is different from their original states (Figure 5.6)
98
and after few cycles (Figure 5.7). The homogenization of Si and Ge under the dealloying
process of Li-SiGe appears as a generic phenomenon with the Si-Ge compositions we
studied. Because the diffusion rates of Si and Ge are extremely low at room
temperature
29
, interdiffusion between Si and Ge is therefore inferred having minimal
effect to the homogenization. Here, we have applied ab-initio molecular dynamic (ab-
MD) technique to simulate the dealloying process at simulated temperature of 300 K.
Extraction of Li from SiGe host is found dramatically accelerate the mixing of Si and Ge.
The structure of Li
15
(Si
0.5
Ge
0.5
)
4
crystal nanoparticle showing in Figure 5.11 was
constructed to simulate the lithiated Li-SiGe alloy, and non-uniform Si and Ge
distribution was created to simulate the non-homogeneity. Li atoms were continuously
removed from the outmost layer of the particle to simulate the dealloying process. We
note that this simulated structure (2 nm in dimension) is smaller than the particles used in
this study, and is used to make simulation tractable. Pair distribution function (PDF) of
Si-Ge was calculated as an evaluation of the inter-mixing between Si and Ge (Figure
5.11). As shown in PDF, Si and Ge have a nearest atom-to-atom distance around 5 Å at
the early stage of the simulation (0-150 fs), corresponding to the distance of Si and Ge in
the lithiated Li
15
(Si
x
Ge
1-x
)
4
structure. The intensity of the peak continues to increase,
evincing an intensification of Si-Ge interdiffusion to form homogeneous alloy. As the
delithiation (dealloying) process proceeds, when the frontier of delithiation comes across
the Si-Ge interface, a small peak equivalent to a bonding length of 2.5 Å starts to show
99
up (e.g. 250 fs), which is inferred as the normal bonding length in SiGe alloy. The images
in Figure 4c display the particle structure at successive simulation stages, showing the
mixing of Si-Ge during the delithiation process, which eventually results in an
amorphous particle. The observed fast interdiffusion is striking considering the low
atomic diffusion rate of Si and Ge; however, with the aid of dealloying, extraction of Li
leaves a large space between Si and/or Ge. The instability of Si and/or Ge atoms staying a
long distance away from each other induces large diffusion rate. Based on the fast atomic
diffusion observed in simulation, it is reasonable to infer that when particles are contacted
with each other, inter-particle diffusion is plausible and will initiate particle merging, as
shown in Figure 5.6 in experiments. In addition, with the incorporation of Ge, the mixing
entropy, which thermodynamically characterizes the measure of material disorder is
responsible for structure stability, will increase with inter-mixing of Si-Ge. For Si
x
Ge
1-x
,
entropy can be calculated as
( ln (1 )ln(1 )) S R x x x x
, where R = 8.314 J mol
-1
K
-1
is the gas constant. The entropy reaches its maximum at the composition of x = 0.5, and
reaches its minimum at x = 0 and x = 1.0. We believe the large entropy of Si-Ge balanced
composition is helpful to stabilize the structure against repeated cycling. However, it is
worthy to mention that the delithiation rate would also affect the particle morphology,
especially for delithiation rate. Fast removal of lithium from Li-SiGe will make the SiGe
system deviate from its thermodynamic stable state, thus significantly change the particle
shape compared to its structure under normal delithiation rate. Figure 5.12 shows the
morphology of Si
54
Ge
46
after 5 cycles with the same lithiation rate (0.8A/g) but fast
100
delithiation rate (by directly setting the voltage to 2 V vs. Li
+
/Li). The particles not only
merge together to form large piece of structures, but also form porous structure.
Figure 5.11 Simulated delithiation process (a-e) of Li-SiGe and calculated energy diagram of
Si
x
Ge
1-x
. (a) modeled structure of Li
15
(Si
0.5
Ge
0.5
)
4
to simulate the lithiated Li-SiGe alloy, and non-
uniform Si and Ge distribution was created to simulate the non-homogeneity. (b) Ge-Si pair
distribution function (PDF) at different stages of simulated delithiation process. The appearance and
intensity increasing of peak at 2.5 Å indicate the inter-mixing of Si and Ge to form a homogenized
alloy. (c) atomic structure and morphology of the particle at different simulation states.
101
Figure. 5.12 Particle morphology of Si
54
Ge
46
after fast delithiation rate at different locations and
different magnifications.
102
One implication of the present work is to improve the stability of Si-based material as
high-capacity lithium-ion battery anode under repeated cyclic test. Compared to pure Si
nanoparticles, SiGe alloy nanoparticle can have stable performance if Si
x
Ge
1-x
composition is tuned to be around Si
50
Ge
50
, which is suggested as a meta-stable
composition which appears more stable than other compositions against structure
evolution during repeating charge/discharge cycling.
Finally, we note some unresolved issues. We have shown that individual nanoparticle will
merge to form a large particle after the 1
st
cycle and then evolve into porous structure, but
direct observation of the transition from solid to porous particle is absent, for example,
the question of whether different compositional Si
x
Ge
1-x
would debut to evolve into
porous particle after the same lithiation/delithiation cycles is not clear. According to MD
simulation, extraction of lithium is found to accelerate the rearrangement and
homogenization of Si-Ge, which is indicative that delithiation rate plays important role to
structure evolution. In the experiments, we have compared the delithiated results at one
fixed charge/discharge rate, thus, future research on the charge/discharge rate effect to the
particle morphology are suggested. In addition, Si-Ge homogenization and phase
separation are apparently competing process, which are related to the kinetic atomic
movement and thermodynamic structure stability. In-depth understanding of which is the
dominant process in the structure evolution is still lack of knowledge, especially for a
comprehensive research on the combined effects of particle composition, particle size,
103
charge/discharge rate, and charge/discharge history. Despite the preliminary results we
have in this work, we hope continuous effort will be devoted to understand the dealloying
behavior in Li-SiGe alloy comprehensively, and broad the research to general multi-
component systems.
104
5.5 Conclusion
We have shown how the Si
x
Ge
1-x
composition
and the alloying/dealloying process history
affect the dealloyed structure of Li-SiGe ternary systems. A slow kinetic process of Si-Ge
homogenization followed by Si-Ge phase separation was identified in the delithiated Li-
SiGe nanostructure in both experiments and theoretical analysis. We have shown that the
existence of meta-stable phase with Si-Ge composition closer to Si
50
Ge
50
gives relative
stable Si-Ge structure under the repeated lithiation/delithiation process, which is
suggested responsible for the high capacity retention of Si
0.54
Ge
0.46
nanoparticles as a
lithium-ion battery anode material. We believe our results will broaden the scope of our
understanding of dealloying in complex systems beyond binary alloys, and may shed
light on the rational design of nanostructures for new applications.
105
5.6 Chapter reference
1 Ding, Y ., Chen, M. W. & Erlebacher, J. Metallic mesoporous nanocomposites for electrocatalysis.
J Am Chem Soc 126, 6876-6877 (2004).
2 Xu, C. X. et al. Low temperature CO oxidation over unsupported nanoporous gold. J Am Chem
Soc 129, 42-43 (2007).
3 Strasser, P. et al. Lattice-strain control of the activity in dealloyed core-shell fuel cell catalysts. Nat
Chem 2, 454-460 (2010).
4 You, T. Y ., Niwa, O., Tomita, M. & Hirono, S. Characterization of platinum nanoparticle-
embedded carbon film electrode and its detection of hydrogen peroxide. Anal Chem 75, 2080-
2085 (2003).
5 Lang, X. Y ., Hirata, A., Fujita, T. & Chen, M. W. Nanoporous metal/oxide hybrid electrodes for
electrochemical supercapacitors. Nat Nanotechnol 6, 232-236 (2011).
6 Li, R. & Sieradzki, K. Ductile-Brittle Transition in Random Porous Au. Phys Rev Lett 68, 1168-
1171 (1992).
7 Sieradzki, K. et al. The dealloying critical potential. J Electrochem Soc 149, B370-B377 (2002).
8 Jia, F. L., Yu, C. F., Deng, K. J. & Zhang, L. Z. Nanoporous metal (Cu, Ag, Au) films with high
surface area: General fabrication and preliminary electrochemical performance. J Phys Chem C
111, 8424-8431 (2007).
9 Zhang, Z. H. et al. Generalized Fabrication of Nanoporous Metals (Au, Pd, Pt, Ag, and Cu)
through Chemical Dealloying. J Phys Chem C 113, 12629-12636 (2009).
10 Wen, C. J. & Huggins, R. A. Chemical Diffusion in Intermediate Phases in the Lithium-Tin
System. J Solid State Chem 35, 376-384 (1980).
11 Chen, Q. & Sieradzki, K. Spontaneous evolution of bicontinuous nanostructures in dealloyed Li-
based systems. Nat Mater 12, 1102-1106 (2013).
12 Sieradzki, K., Corderman, R. R., Shukla, K. & Newman, R. C. Computer-Simulations of
Corrosion - Selective Dissolution of Binary-Alloys. Philos Mag A 59, 713-746 (1989).
13 Erlebacher, J., Aziz, M. J., Karma, A., Dimitrov, N. & Sieradzki, K. Evolution of nanoporosity in
dealloying. Nature 410, 450-453 (2001).
14 Cahn, J. W. & Hilliard, J. E. Free Energy of a Nonuniform System .1. Interfacial Free Energy. J
Chem Phys 28, 258-267 (1958).
15 Okamoto, H. The Li-Si (Lithium-Silicon) System. Bulletin of Alloy Phase Diagram 11, 7 (1990).
16 McDonwell, M. T. et al. In Situ TEM of Two-Phase Lithiation of Amorphous Silicon
Nanospheres. Nano Lett 13, 758-764 (2013).
17 Choi, J. W. et al. Stepwise Nanopore Evolution in One-Dimensional Nanostructures. Nano Lett 10,
106
1409-1413 (2010).
18 Obrovac, M. N. & Christensen, L. Structural changes in silicon anodes during lithium
insertion/extraction. Electrochem Solid St 7, A93-A96 (2004).
19 Laforge, B., Levan-Jodin, L., Salot, R. & Billard, A. Study of germanium as electrode in thin-film
battery. J Electrochem Soc 155, A181-A188 (2008).
20 Liu, X. H. et al. Reversible Nanopore Formation in Ge Nanowires during Lithiation-Delithiation
Cycling: An In Situ Transmission Electron Microscopy Study. Nano Lett 11, 3991-3997 (2011).
21 Park, M. H. et al. Germanium Nanotubes Prepared by Using the Kirkendall Effect as Anodes for
High-Rate Lithium Batteries. Angew Chem Int Edit 50, 9647-9650 (2011).
22 Kim, S., Jeon, K., Lee, J. C., Swihart, M. T. & Yang, M. Enhanced Performance of a Polymer
Solar Cell upon Addition of Free-Standing, Freshly Etched, Photoluminescent Silicon
Nanocrystals. Appl Phys Express 5 (2012).
23 Kim, S., Lee, J. H., Swihart, M. T., Lee, J. C. & Kim, J. Y . Silicon nanoparticle size-dependent
open circuit voltage in an organic-inorganic hybrid solar cell. Curr Appl Phys 14, 127-131 (2014).
24 Kresse, G. & Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations
using a plane-wave basis set. Phys Rev B 54, 11169-11186 (1996).
25 Perdew, J. P. et al. Atoms, Molecules, Solids, and Surfaces - Applications of the Generalized
Gradient Approximation for Exchange and Correlation. Phys Rev B 46, 6671-6687 (1992).
26 Kelires, P. C. & Tersoff, J. Equilibrium Alloy Properties by Direct Simulation - Oscillatory
Segregation at the Si-Ge(100) 2x1 Surface. Phys Rev Lett 63, 1164-1167 (1989).
27 McDowell, M. T. et al. In Situ TEM of Two-Phase Lithiation of Amorphous Silicon Nanospheres.
Nano Lett 13, 758-764 (2013).
28 Lee, S. W., McDowell, M. T., Choi, J. W. & Cui, Y . Anomalous Shape Changes of Silicon
Nanopillars by Electrochemical Lithiation. Nano Lett 11, 3034-3039 (2011).
29 Silvestri, H. H., Bracht, H., Hansen, J. L., Larsen, A. N. & Haller, E. E. Diffusion of silicon in
crystalline germanium. Semicond Sci Tech 21, 758-762 (2006).
107
Chapter 6 Future work: towards LiSi-S full battery
6.1 Introduction
6.1.1 Motivation
In previous chapters, we have evaluated the properties and electrochemical performance
of silicon anode for lithium-ion battery application, along with a further attempt to
improve silicon’s structure stability by alloying silicon with germanium. Half-cell tests
have demonstrated superior high capacity and long cyclability when porous silicon is
coupled with lithium metal foil. Now, it is an appropriate time to integrate silicon into a
full battery by combining silicon anode with other cathode materials.
There are plenty of cathode choices, and sulfur is a potential interesting cathode material.
Sulfur has high capacity which is around 1600 mAh/g when coupled with lithium metal
to form Li-S battery. Compared with other battery system such as traditional
graphite/LiCoO
2
, integrating sulfur and silicon (lithiated) would boost the battery’s
energy density by a fold of 3. Table 6.1 has summarized the theoretical capacity and
calculated energy density for different anode/cathode combinations. Combination of
silicon (lithiated) and sulfur clearly demonstrate its advantages in terms of high capacity
and high energy density. In addition, the abundance and non-toxicity of sulfur and silicon
108
provide Si (lithiated)-S batteries with improved energy economy and environmental
friendliness.
Table 6.1 Comparison of battery systems
Anode/Cathode Graphite/LiCoO
2
Si/LiCoO
2
Li
3.75
Si/S
Anode/Cathode (mAh/g) 372 / 155 3600 / 155 1858 /1600
Full cell capacity (mAh/g) 109 148 860
Voltage (V) 3.75 3.4 1.7
Energy Density (Wh/kg) 410 505 1461
6.1.2 Characteristic property of sulfur cathode: Li-S battery
Rechargeable Li-S battery is operated by the reduction of sulfur at the cathode on
discharge to form various lithium-polysulphides to finally produce Li
2
S. Figure 6.1
demonstrates the operation principle of Li-S cell and typical battery performance. Despite
the promise of Li-S as high capacity battery, several intrinsic problems inherent the
progress of Li-S:
1. Shuttle effects
During the charge/discharge process, intermediate polysulphide Li
2
S
n
(3 ≤ n ≤ 6)
become soluble in electrolyte
1,2
. They are transported to the Li anode and get reduced to
lower polysulphides, which are then transport back to cathode. After reaching the
cathode, the lower polysulphides get re-oxidized and will transport to anode again. The
reduction and oxidization of polysulphides and repeating transportation between cathode
109
(S) and anode (Li) is named as shuttle effect
3
. In addition, when polysulphides reach the
Li anode, it is very possible to form non-soluble Li
2
S
2
or LiS
2
and get deposited on the
anode surface, which is in analogy to the formation of internal short circuit, consuming
both active cathode material (S) and anode material (Li), leading to fast capacity fading,
as indicated in Figure 6.1
4
.
Figure 6.1 Charge/discharge profile of Li-S battery and its typical battery performance
4
.
110
2. Poor conductivity
Insoluble sulfur and low order polysulphides such as Li
2
S and Li
2
S
2
are electrically
insulators. The insulating nature of these products limits the rate capability and induces
poor rechargeability of Li-S battery.
Many efforts have been devoted to circumvent the above challenges by designing a
conductive sulfur-contained nanoporous composite cathode that is capable to capture the
dissolved polysulphides species as well as providing enough conductive paths to reduce
the electric resistance. Improvement is obvious, for example, encapsulating sulfur in
carbon-based porous structure, such as porous carbon, hollow carbon sphere
5-9
, etc..
Figure 6.2 presents some typical results from reported work. Other approaches such as
coating a protective layer (e.g. TiO
2
)
10
, or wrapping graphene on the outer surface of
sulfur particles
11,12
, or inserting another porous carbon layer between sulfur cathode and
separator to capture the dissolved polysulphides are demonstrated to be positive to retard
the capacity fading
13
.
111
Figure 6.2 Various structures of S incorporated in carbon-based materials as compound cathodes for
Li-S battery. (a) S incorporated into mesoporous CMK3 porous carbon, and its battery performance
7
.
(b) S incorporated into mesoporous carbon spheres and its battery performance
9
. (c) S incorporated
into carbon nanotubes and its battery performance
14
.
112
On the other hand, it is also challenging to find proper electrolyte to combat the
irreversible capacity loss due to polysulphides dissolution. Systems as tetrahydrofuran
15
,
1,3-dioxolane
15,16
, dimethoxyethane
17
and tetra(ethylene glycol)-dimethyl ether
18
have
attracted great attention because of the absence of chemical attack from S
-
centers. In
addition, LiNO
3
is found as a good electrolyte additive to Li-S battery. LiNO
3
will
promote the formation of a stable passivation film on the surface of Li anode which
significantly suppresses the redox shuttle of lithium polysulphides
19,20
.
With the optimized design of electrode material and electrolyte, Li-S battery is reported
to be cycled for hundreds of cycles with small degradation. However, there is one thing
people always overlooked. The theoretical specific capacity of Li-S battery is calculated
based on a balanced mass loading of both anode and cathode. For example, if a cathode
has a theoretical capacity of x, and anode has a theoretical capacity of y, then the required
amount of cathode and anode should is correlated to be m
anode
: m
cathode
= x : y. In Li-S
battery, the amount of lithium is far more than required even we consider the essential
amount for overdesign. It is therefore of less meaningful to advocate the high capacity
and energy density of Li-S compared to traditional Li-ion battery system. And the large
amount of Li used in Li-S battery has posed potential environmental hazards as the
formation of lithium dendrites would penetrate the separator and cause short circuit.
113
Substituting lithium metal with silicon may be a good choice. In the following, I will
discuss the challenges of the development of Si-S batteries and propose potential solution
idea.
The major challenge faced by Si-S battery is the absent of lithium in the cell. Generally,
lithium can be pre-inserted into either the anode (Si Li
x
Si) side or cathode (S Li
2
S)
side. For example, Li
x
Si can be formed through an electrochemical prelithiation of Si in a
Li-Si half-cell
21
, and S can be converted to Li
2
S by reacting S with butyl-lithium in inert
(Ar protected) atmosphere
22
. Here, we choose the first approach to insert lithium in the Si
side based on two considerations:
1. In Li-Si half-cell, there is a large capacity drop between the first several cycles and
following cycles due to a formation of stable solid-electrolyte-interface (SEI), which is
essential to stabilize the Si anode for afterward charge/discharge process. In the formation
of SEI, considerable amount of lithium is irreversibly consumed from the Li metal in the
Li-Si half cell, which is in the same situation that a large amount of cathode (Li
2
S) will be
wasted in a Li
2
S-Si cell. It is therefore more preferable to use prelithiated silicon as anode
material.
2. To convert S into Li
2
S, organic lithium salts (e.g. n-butyl-lithium) are used. These
metal organic salts are usually air-sensitive and toxic, and can only be used in a well-
114
protected environment for safety concern. Another possible approach is directly using
Li
2
S as the cathode which is readily available from chemical venders. However, as-
obtained Li
2
S are usually very large particles; considering its insulating nature, it is not
suitable for battery application. In addition, strategies to crash the large Li
2
S particle into
nanometer or loosely sub-micron sized small particles are not well established, as special
manipulation is required because of the air-sensitive nature of Li
2
S as well. Therefore, we
decide to lithiate Si anode instead to lithiate S.
Another challenge faced by Si-S battery is the choice of electrolyte. Currently, the
optimized electrolyte for Si is LiPF
6
dissolved in EC-DMC-FEC mixed solvent, and for
Li-S battery is LITFSI dissolved in (DME-DOL) system. The question as “which type of
electrolyte is good to use, or whether it is required to develop new electrolyte system”
needs comprehensive research. Here, I will present some preliminary results about Si-S
full battery.
115
6.2 Experiments and results
Si anode is fabricated using nanoporous silicon particles (prepared as in Chapter 4). To
lithiate silicon, Li-Si half-cell is assembled and running the charge/discharge repeating
process for 10 cycles in the voltage window of 0.01-2V . After 10 cycles, cell is
discharged to 0.01V and then disassembled in glovebox for further use.
Sulfur cathode is fabricated using porous carbon-sulfur composite particles. In detail,
porous carbon is first produced using SiO
2
nanoparticles as template. Specifically, SiO
2
nanoparticles in size of 10 nm were dissolved in water, with additional amount of sucrose
as carbon source. Mixed solution was kept in an autoclave and heated up to 180
o
C for 4
hours. At this stage, sucrose was cross linked forming a polymeric shell on the surface of
SiO
2
sphere. The mixture was then transferred to a tube furnace, and heated up to 800
o
C
for 1 hour in Ar-protected atmosphere. After cooling down, the material was soaked into
HF solution to dissolve the SiO
2
template. Porous carbon was obtained after washing
with DI-H
2
O and dried to powder. To produce porous carbon-sulfur compound (in the
following, denoted as PCS), 2:1 of sulfur and porous C were mixed and annealed at
155
o
C in autoclave for 4 hours. Electrode was made by mixing PCS, carbon-black and
PVDF to form a slurry using DME as solvent, and then coated on an Al-foil. Figure 6.3
shows the typical morphology of porous carbon and PCS.
116
Figure 6.3 TEM images of porous carbon (a) and after infiltration with sulfur (b).
Li-Si and Li-S half cells were cross-tested using both LiPF
6
electrolyte (specified for Li-
Si) and LITFSI (specified for Li-S) as electrolyte. Figure 6.4 shows the cyclic
performance of both half-cells. According to Figure 6.4, both electrolytes works well for
the Li-Si half-cell without qualitative difference; however, LiPF
6
electrolyte fails for the
Li-S cell as almost no capacity retained after the 2
nd
cycle. The incompatibility of LiPF
6
in Li-S battery leaves us nearly no choice about the selection of electrolyte for Si-S full
cell. The underlying reason of the incompatibility is not clear at this moment. We will
temporary use the LITFSI electrolyte for Si-S full battery study before new proper
electrolyte is designed.
117
Figure 6.4 Battery performance of Li-Si half cell (left) and Li-S half cell (right) with LiPF6 and
LiTFSI electrolyte, respectively.
After Si lithiation, Li-Si cell was dissembled and Si electrode was washed with fresh
electrolyte before assembling the Si(lithiated)-S full cell. Figure 6.5a shows the cyclic
performance of the Si(lithiated)-S battery. The specific capacity is calculated based on the
amount of sulfur, because we intentionally have more lithiated Si then required to
guarantee sufficient amount of lithium inside the cell. LITFSI electrolyte was used. From
Figure 6.5a, the capacity of the first cycle is around 440 mAh/g, however, it drops
quickly from the second cycle and results in almost no capacity after few cycles. The fast
drop of capacity is due to the internal short circuit of transportation of dissolved
polysulfide to silicon anode. It is the same mechanism as the shuttle effect in Li-S battery.
One may be curious why Li-S battery as shown in Figure 6.1 and Figure 6.3 does not
encounter the fast capacity drop problem. This reason is there is more than enough Li in
the Li-S cell, and therefore, the shuttle effect in Li-S cell will only decreases the capacity
due to the loss of polysulfide, but will not have detrimental effect on the Li anode part.
118
However, in Si-S battery, due to the limited amount of lithiated Si, deposition of
polysulfide on Si consumes both the cathode and anode, leading to fast capacity drop. To
further confirm the point, we tested the Li-Si half-cell with additional small amount of
Li
2
S
x
in LITFSI electrolyte. Figure 6.5b clearly shows the fast drop of capacity, which is
in sharp contrast to the Li-Si cell without Li
2
S
x
in electrolyte (Figure 3).
Figure 6.5 (a) battery performance of Si(lithiated)-S full cell using LiTFSI electrolyte. (b) battery
performance of Li-Si half cell with additional Li
2
S
x
in LiTFSI electrolyte.
To circumvent the above issue, protecting the Si anode without direct contact with
polysulfide anion is highly desired. Here we choose to use a polymer called “Nafion” to
protect Si anode. Nafion is a sulfonated tetrafluoroethylene based fluoropolymer
copolymer discovered in the late 1960s by Walther Grot of DuPont. Nafion can be used
as a proton exchange membrane. Its unique ionic properties are a result of incorporating
perfluorovinyl ether groups terminated with sulfonate groups onto a tetrafluoroethylene
119
(Teflon) backbone, as shown in Figure 6.6. The chemical group of SO
3
H allows the
movement of cations but the membranes do not conduct anions or electrons, which is
ideal to block polysulfide anion to electrically contact silicon anode.
Figure 6.6 Molecular structure of nafion polymer.
We used silicon nanoparticle to test nafion coating. Figure 6.7 shows the morphology of
nafion coated silicon. We use EFTEM to identify of existence of elemental Si and S(come
from nafion). It could be seen that Si and S has a well overlapped region, indicating a
conformal coating of nafion on Si particles. A dramatic improvement was found when we
test the nafion-coated Si in a Li-Si half-cell with additional Li
2
S
x
in electrolyte, as shown
in Figure 6.5. It demonstrates successful protection of nafion on Si.
Figure 6.7 TEM and FETEM images of nafion coated Si nanoparticles.
120
We then applied the coating on porous Si particle, and did the same procedure to lithiate
Si and then assemble Si(lithiated)-S full cell. Figure 6.8 shows the battery test
performance. There is no fast capacity drop observed, and the battery can be run for 160
cycles with capacity higher than 200 mAh/g. Comparing the results in Figure 6.8 and in
Figure 6.5a (full cell without nafion coating on Si), we have demonstrated a prototype of
Si-S full cell.
Figure 6.8 Cyclic performance of Si(lithiated)-S full cell.
121
6.3 Conclusion
We have addressed some critical issue for building up a Si-S full battery. The shuttle
transport of polysulfide is found detrimental to the full battery, as it will form internal
short circuit with the direct contact of polysulfide and Si anode. Nafion was found as an
effective coating material which will block the polysulfide deposition on Si due to its
unique chemical structure. Full cell test has shown preliminary results of the Si-S full
battery with promising performance. However, there are still many unsolved issues such
as the continuing improvement of the capacity retention, which needs great efforts in
optimizing both electrode material as well as suitable electrolyte system.
122
6.4 Chapter reference
1 Rauh, R. D., Shuker, F. S., Marston, J. M. & Brummer, S. B. Formation of Lithium Polysulfides in
Aprotic Media. J Inorg Nucl Chem 39, 1761-1766 (1977).
2 Ryu, H. S., Guo, Z. P., Ahn, H. J., Cho, G. B. & Liu, H. K. Investigation of discharge reaction
mechanism of lithium vertical bar liquid electrolyte vertical bar sulfur battery. J Power Sources
189, 1179-1183 (2009).
3 Mikhaylik, Y . V . & Akridge, J. R. Polysulfide shuttle study in the Li/S battery system. J
Electrochem Soc 151, A1969-A1976 (2004).
4 Bruce, P. G., Freunberger, S. A., Hardwick, L. J. & Tarascon, J. M. Li-O-2 and Li-S batteries with
high energy storage. Nat Mater 11, 19-29 (2012).
5 Jayaprakash, N., Shen, J., Moganty, S. S., Corona, A. & Archer, L. A. Porous Hollow
Carbon@Sulfur Composites for High-Power Lithium-Sulfur Batteries. Angew Chem Int Edit 50,
5904-5908 (2011).
6 He, G. et al. Tailoring Porosity in Carbon Nanospheres for Lithium-Sulfur Battery Cathodes. Acs
Nano 7, 10920-10930 (2013).
7 Ji, X. L., Lee, K. T. & Nazar, L. F. A highly ordered nanostructured carbon-sulphur cathode for
lithium-sulphur batteries. Nat Mater 8, 500-506 (2009).
8 Schuster, J. et al. Spherical Ordered Mesoporous Carbon Nanoparticles with High Porosity for
Lithium-Sulfur Batteries. Angew Chem Int Edit 51, 3591-3595 (2012).
9 Zhang, B., Qin, X., Li, G. R. & Gao, X. P. Enhancement of long stability of sulfur cathode by
encapsulating sulfur into micropores of carbon spheres. Energ Environ Sci 3, 1531-1537 (2010).
10 Seh, Z. W. et al. fSulphur-TiO2 yolk-shell nanoarchitecture with internal void space for long-cycle
lithium-sulphur batteries. Nature Communications 4 (2013).
11 Wang, H. L. et al. Graphene-Wrapped Sulfur Particles as a Rechargeable Lithium-Sulfur Battery
Cathode Material with High Capacity and Cycling Stability. Nano Lett 11, 2644-2647 (2011).
12 Ji, L. W. et al. Graphene Oxide as a Sulfur Immobilizer in High Performance Lithium/Sulfur
Cells. J Am Chem Soc 133, 18522-18525 (2011).
13 Su, Y . S. & Manthiram, A. Lithium-sulphur batteries with a microporous carbon paper as a
bifunctional interlayer. Nature Communications 3 (2012).
14 Lu, S. T., Cheng, Y . W., Wu, X. H. & Liu, J. Significantly Improved Long-Cycle Stability in High-
Rate Li-S Batteries Enabled by Coaxial Graphene Wrapping over Sulfur-Coated Carbon
Nanofibers. Nano Lett 13, 2485-2489 (2013).
15 Yamin, H., Penciner, J., Gorenshtain, A., Elam, M. & Peled, E. The Electrochemical-Behavior of
Polysulfides in Tetrahydrofuran. J Power Sources 14, 129-134 (1985).
123
16 Tobishima, S. I., Yamamoto, H. & Matsuda, M. Study on the reduction species of sulfur by alkali
metals in nonaqueous solvents. Electrochim Acta 42, 1019-1029 (1997).
17 Peled, E., Sternberg, Y ., Gorenshtein, A. & Lavi, Y . Lithium-Sulfur Battery - Evaluation of
Dioxolane-Based Electrolytes. J Electrochem Soc 136, 1621-1625 (1989).
18 Ryu, H. S. et al. Self-discharge characteristics of lithium/sulfur batteries using TEGDME liquid
electrolyte. Electrochim Acta 52, 1563-1566 (2006).
19 Zhang, S. S. Role of LiNO3 in rechargeable lithium/sulfur battery. Electrochim Acta 70, 344-348
(2012).
20 Zhang, S. S. & Read, J. A. A new direction for the performance improvement of rechargeable
lithium/sulfur batteries. J Power Sources 200, 77-82 (2012).
21 Liu, N. A., Hu, L. B., McDowell, M. T., Jackson, A. & Cui, Y . Prelithiated Silicon Nanowires as
an Anode for Lithium Ion Batteries. Acs Nano 5, 6487-6493 (2011).
22 Yang, Y . et al. New Nanostructured Li2S/Silicon Rechargeable Battery with High Specific Energy.
Nano Lett 10, 1486-1491 (2010).
124
Bibliography
Ammundsen, B. et al. Formation and structural properties of layered LiMnO2 cathode materials. J
Electrochem Soc 147, 4078-4082 (2000).
Antolini, E. LiCoO2: formation, structure, lithium and oxygen nonstoichiometry, electrochemical behaviour
and transport properties. Solid State Ionics 170, 159-171 (2004).
Arai, H., Okada, S., Sakurai, Y . & Yamaki, J. Reversibility of LiNiO2 cathode. Solid State Ionics 95, 275-282
(1997).
Axel, H., Schafer, H. & Weiss, A. Zur Kenntnis Der Phase Li22si5. Z Naturforsch Pt B B 21, 115-& (1966).
Baggetto, L. & Notten, P . H. L. Lithium-Ion (De)Insertion Reaction of Germanium Thin-Film Electrodes: An
Electrochemical and In Situ XRD Study. J Electrochem Soc 156, A169-A175 (2009).
Bang, B. M., Kim, H., Song, H. K., Cho, J. & Park, S. Scalable approach to multi-dimensional bulk Si anodes
via metal-assisted chemical etching. Energ Environ Sci 4, 5013-5019 (2011).
Bang, B. M., Lee, J. I., Kim, H., Cho, J. & Park, S. High-Performance Macroporous Bulk Silicon Anodes
Synthesized by Template-Free Chemical Etching. Adv Energy Mater 2, 878-883 (2012).
Bao, Z. H. et al. Chemical reduction of three-dimensional silica micro-assemblies into microporous silicon
replicas. Nature 446, 172-175 (2007).
Benedek, R., Thackeray, M. M. & Yang, L. H. Lithium site preference and electronic structure of Li4V3O8.
Phys Rev B 56, 10707-10710 (1997).
Blochl, P . E. Projector Augmented-Wave Method. Phys Rev B 50, 17953-17979 (1994).
Bramnik, N. N., Nikolowski, K., Trots, D. M. & Ehrenberg, H. Thermal stability of LiCoPO(4) cathodes.
Electrochem Solid St 11, A89-A93 (2008).
Bruce, P . G., Freunberger, S. A., Hardwick, L. J. & Tarascon, J. M. Li-O-2 and Li-S batteries with high energy
storage. Nat Mater 11, 19-29 (2012).
Cahn, J. W. & Hilliard, J. E. Free Energy of a Nonuniform System .1. Interfacial Free Energy. J Chem Phys 28,
258-267 (1958).
Chan, C. K. et al. High-performance lithium battery anodes using silicon nanowires. Nat Nanotechnol 3, 31-
35 (2008).
125
Chan, C. K., Ruffo, R., Hong, S. S., Huggins, R. A. & Cui, Y . Structural and electrochemical study of the
reaction of lithium with silicon nanowires. J Power Sources 189, 34-39 (2009).
Chandrasekaran, R. & Fuller, T. F. Analysis of the Lithium-Ion Insertion Silicon Composite
Electrode/Separator/Lithium Foil Cell. J Electrochem Soc 158, A859-A871 (2011).
Charitidis, C. A., Skarmoutsou, A., Nassiopoulou, A. G. & Dragoneas, A. Nanomechanical properties of thick
porous silicon layers grown on p- and p plus -type bulk crystalline Si. Mat Sci Eng a-Struct 528, 8715-8722
(2011).
Chen, Q. & Sieradzki, K. Spontaneous evolution of bicontinuous nanostructures in dealloyed Li-based
systems. Nat Mater 12, 1102-1106 (2013).
Chen, X. L. et al. Conductive Rigid Skeleton Supported Silicon as High-Performance Li-Ion Battery Anodes.
Nano Lett 12, 4124-4130 (2012).
Choi, J. W. et al. Stepwise Nanopore Evolution in One-Dimensional Nanostructures. Nano Lett 10, 1409-
1413 (2010).
Choi, J. W. et al. Stepwise Nanopore Evolution in One-Dimensional Nanostructures. Nano Lett 10, 1409-
1413 (2010).
Chon, M. J., Sethuraman, V. A., McCormick, A., Srinivasan, V. & Guduru, P . R. Real-Time Measurement of
Stress and Damage Evolution during Initial Lithiation of Crystalline Silicon. Phys Rev Lett 107 (2011).
Chou, S. L. et al. Enhanced reversible lithium storage in a nanosize silicon/graphene composite.
Electrochem Commun 12, 303-306 (2010).
Courtney, I. A., Tse, J. S., Mao, O., Hafner, J. & Dahn, J. R. Ab initio calculation of the lithium-tin voltage
profile. Phys Rev B 58, 15583-15588 (1998).
Ding, Y ., Chen, M. W. & Erlebacher, J. Metallic mesoporous nanocomposites for electrocatalysis. J Am
Chem Soc 126, 6876-6877 (2004).
Dudley, M. E. & Kolasinski, K. W. Stain Etching with Fe(III), V(V), and Ce(IV) to Form Microporous Silicon.
Electrochem Solid St 12, D22-D26 (2009).
Erlebacher, J., Aziz, M. J., Karma, A., Dimitrov, N. & Sieradzki, K. Evolution of nanoporosity in dealloying.
Nature 410, 450-453 (2001).
126
Esmanski, A. & Ozin, G. A. Silicon Inverse-Opal-Based Macroporous Materials as Negative Electrodes for
Lithium Ion Batteries. Adv Funct Mater 19, 1999-2010 (2009).
Ge, M. et al. Scalable production of porous silicon nanoparticles and their application in lithium-ion
battery anode. Nano Research 6, 174-181 (2012).
Ge, M. Y ., Rong, J. P ., Fang, X. & Zhou, C. W. Porous Doped Silicon Nanowires for Lithium Ion Battery Anode
with Long Cycle Life. Nano Lett 12 (2012).
Gowda, S. R. et al. Three-Dimensionally Engineered Porous Silicon Electrodes for Li Ion Batteries. Nano Lett
12, 6060-6065, doi:Doi 10.1021/Nl302114j (2012).
Gummow, R. J., Dekock, A. & Thackeray, M. M. Improved Capacity Retention in Rechargeable 4v Lithium
Lithium Manganese Oxide (Spinel) Cells. Solid State Ionics 69, 59-67 (1994).
He, G. et al. Tailoring Porosity in Carbon Nanospheres for Lithium-Sulfur Battery Cathodes. Acs Nano 7,
10920-10930 (2013).
Hochbaum, A. I., Gargas, D., Hwang, Y . J. & Yang, P . Single Crystalline Mesoporous Silicon Nanowires. Nano
Lett 9, 3550-3554 (2009).
Jayaprakash, N., Shen, J., Moganty, S. S., Corona, A. & Archer, L. A. Porous Hollow Carbon@Sulfur
Composites for High-Power Lithium-Sulfur Batteries. Angew Chem Int Edit 50, 5904-5908 (2011).
Ji, L. W. et al. Graphene Oxide as a Sulfur Immobilizer in High Performance Lithium/Sulfur Cells. J Am Chem
Soc 133, 18522-18525 (2011).
Ji, X. L., Lee, K. T. & Nazar, L. F. A highly ordered nanostructured carbon-sulphur cathode for lithium-
sulphur batteries. Nat Mater 8, 500-506 (2009).
Jia, F. L., Yu, C. F., Deng, K. J. & Zhang, L. Z. Nanoporous metal (Cu, Ag, Au) films with high surface area:
General fabrication and preliminary electrochemical performance. J Phys Chem C 111, 8424-8431 (2007).
Jia, H. P . et al. Novel Three-Dimensional Mesoporous Silicon for High Power Lithium-Ion Battery Anode
Material. Adv Energy Mater 1, 1036-1039 (2011).
Jiang, T., Zhang, S. C., Qiu, X. P ., Zhu, W. T. & Chen, L. Q. Preparation and characterization of tin-based
three-dimensional cellular anode for lithium ion battery. J Power Sources 166, 503-508 (2007).
Kasavajjula, U., Wang, C. S. & Appleby, A. J. Nano- and bulk-silicon-based insertion anodes for lithium-ion
secondary cells. J Power Sources 163, 1003-1039 (2007).
127
Kelires, P . C. & Tersoff, J. Equilibrium Alloy Properties by Direct Simulation - Oscillatory Segregation at the
Si-Ge(100) 2x1 Surface. Phys Rev Lett 63, 1164-1167 (1989).
Kim, H. & Cho, J. Superior Lithium Electroactive Mesoporous Si@Carbon Core-Shell Nanowires for Lithium
Battery Anode Material. Nano Lett 8, 3688-3691 (2008).
Kim, H., Han, B., Choo, J. & Cho, J. Three-Dimensional Porous Silicon Particles for Use in High-Performance
Lithium Secondary Batteries. Angew Chem Int Edit 47, 10151-10154 (2008).
Kim, S. W., Kim, J., Gwon, H. & Kang, K. Phase Stability Study of Li1-xMnPO4 (0 <= x <= 1) Cathode for Li
Rechargeable Battery. J Electrochem Soc 156, A635-A638 (2009).
Kim, S., Jeon, K., Lee, J. C., Swihart, M. T. & Yang, M. Enhanced Performance of a Polymer Solar Cell upon
Addition of Free-Standing, Freshly Etched, Photoluminescent Silicon Nanocrystals. Appl Phys Express 5
(2012).
Kim, S., Lee, J. H., Swihart, M. T., Lee, J. C. & Kim, J. Y . Silicon nanoparticle size-dependent open circuit
voltage in an organic-inorganic hybrid solar cell. Curr Appl Phys 14, 127-131 (2014).
Koksbang, R., Barker, J., Shi, H. & Saidi, M. Y . Cathode materials for lithium rocking chair batteries. Solid
State Ionics 84, 1-21 (1996).
Kovalenko, I. et al. A Major Constituent of Brown Algae for Use in High-Capacity Li-Ion Batteries. Science
334, 75-79 (2011).
Kresse, G. & Furthmuller, J. Efficient iterative schemes for ab initio total-energy calculations using a plane-
wave basis set. Phys Rev B 54, 11169-11186 (1996).
Kubota, Y ., Escano, M. C. S., Nakanishi, H. & Kasai, H. Crystal and electronic structure of Li15Si4. J Appl Phys
102 (2007).
Laforge, B., Levan-Jodin, L., Salot, R. & Billard, A. Study of germanium as electrode in thin-film battery. J
Electrochem Soc 155, A181-A188 (2008).
Lang, X. Y ., Hirata, A., Fujita, T. & Chen, M. W. Nanoporous metal/oxide hybrid electrodes for
electrochemical supercapacitors. Nat Nanotechnol 6 (2011).
Lee, J. K., Smith, K. B., Hayner, C. M. & Kung, H. H. Silicon nanoparticles-graphene paper composites for Li
ion battery anodes. Chem Commun 46, 2025-2027 (2010).
128
Lee, S. W., McDowell, M. T., Choi, J. W. & Cui, Y . Anomalous Shape Changes of Silicon Nanopillars by
Electrochemical Lithiation. Nano Lett 11 (2011).
Li, H., Huang, X. J., Chen, L. Q., Wu, Z. G. & Liang, Y . A high capacity nano-Si composite anode material for
lithium rechargeable batteries. Electrochem Solid St 2, 547-549 (1999).
Li, R. & Sieradzki, K. Ductile-Brittle Transition in Random Porous Au. Phys Rev Lett 68, 1168-1171 (1992).
Liu, N. A., Hu, L. B., McDowell, M. T., Jackson, A. & Cui, Y . Prelithiated Silicon Nanowires as an Anode for
Lithium Ion Batteries. Acs Nano 5, 6487-6493 (2011).
Liu, N. et al. A Yolk-Shell Design for Stabilized and Scalable Li-Ion Battery Alloy Anodes. Nano Lett 12, 3315-
3321 (2012).
Liu, N., Huo, K. F., McDowell, M. T., Zhao, J. & Cui, Y . Rice husks as a sustainable source of nanostructured
silicon for high performance Li-ion battery anodes. Sci Rep-Uk 3, 7 (2013).
Liu, W., Farrington, G. C., Chaput, F. & Dunn, B. Synthesis and electrochemical studies of spinel phase
LiMn2O4 cathode materials prepared by the Pechini process. J Electrochem Soc 143, 879-884 (1996).
Liu, X. H. et al. Reversible Nanopore Formation in Ge Nanowires during Lithiation-Delithiation Cycling: An
In Situ Transmission Electron Microscopy Study. Nano Lett 11, 3991-3997 (2011).
Liu, X. H. et al. Size-Dependent Fracture of Silicon Nanoparticles During Lithiation. Acs Nano 6, 1522-1531
(2012).
Lu, S. T., Cheng, Y . W., Wu, X. H. & Liu, J. Significantly Improved Long-Cycle Stability in High-Rate Li-S
Batteries Enabled by Coaxial Graphene Wrapping over Sulfur-Coated Carbon Nanofibers. Nano Lett 13,
2485-2489 (2013).
Ma, Z. S. et al. Critical silicon-anode size for averting lithiation-induced mechanical failure of lithium-ion
batteries. Rsc Adv 3, 7398-7402 (2013).
McDonwell, M. T. et al. In Situ TEM of Two-Phase Lithiation of Amorphous Silicon Nanospheres. Nano Lett
13, 758-764 (2013).
McDowell, M. T. et al. In Situ TEM of Two-Phase Lithiation of Amorphous Silicon Nanospheres. Nano Lett
13, 758-764 (2013).
Mikhaylik, Y . V. & Akridge, J. R. Polysulfide shuttle study in the Li/S battery system. J Electrochem Soc 151,
A1969-A1976 (2004).
129
Monkhorst, H. J. & Pack, J. D. Special Points for Brillouin-Zone Integrations. Phys Rev B 13, 5188-5192
(1976).
Obrovac, M. N. & Christensen, L. Structural changes in silicon anodes during lithium insertion/extraction.
Electrochem Solid St 7, A93-A96 (2004).
Obrovac, M. N., Christensen, L., Le, D. B. & Dahnb, J. R. Alloy design for lithium-ion battery anodes. J
Electrochem Soc 154, A849-A855 (2007).
Ohara, S., Suzuki, J., Sekine, K. & Takamura, T. Li insertion/extraction reaction at a Si film evaporated on a
Ni foil. J Power Sources 119, 591-596 (2003).
Okamoto, H. The Li-Si (Lithium-Silicon) System. Bulletin of Alloy Phase Diagram 11, 7 (1990).
Padhi, A. K., Nanjundaswamy, K. S., Masquelier, C., Okada, S. & Goodenough, J. B. Effect of structure on
the Fe3+/Fe2+ redox couple in iron phosphates. J Electrochem Soc 144, 1609-1613 (1997).
Park, J., Lu, W. & Sastry, A. M. Numerical Simulation of Stress Evolution in Lithium Manganese Dioxide
Particles due to Coupled Phase Transition and Intercalation. J Electrochem Soc 158, A201-A206 (2011).
Park, M. H. et al. Germanium Nanotubes Prepared by Using the Kirkendall Effect as Anodes for High-Rate
Lithium Batteries. Angew Chem Int Edit 50, 9647-9650 (2011).
Park, M. H. et al. Silicon Nanotube Battery Anodes. Nano Lett 9, 3844-3847 (2009).
Pei, S. F. & Cheng, H. M. The reduction of graphene oxide. Carbon 50, 3210-3228 (2012).
Peled, E., Menachem, C., BarTow, D. & Melman, A. Improved graphite anode for lithium-ion batteries -
Chemically bonded solid electrolyte interface and nanochannel formation. J Electrochem Soc 143, L4-L7
(1996).
Peled, E., Sternberg, Y ., Gorenshtein, A. & Lavi, Y . Lithium-Sulfur Battery - Evaluation of Dioxolane-Based
Electrolytes. J Electrochem Soc 136, 1621-1625 (1989).
Perdew, J. P . et al. Atoms, Molecules, Solids, and Surfaces - Applications of the Generalized Gradient
Approximation for Exchange and Correlation. Phys Rev B 46, 6671-6687 (1992).
Perdew, J. P ., Burke, K. & Ernzerhof, M. Generalized gradient approximation made simple. Phys Rev Lett 77,
3865-3868 (1996).
Prosini, P . P ., Mancini, R., Petrucci, L., Contini, V. & Villano, P . Li4Ti5O12 as anode in all-solid-state, plastic,
130
lithium-ion batteries for low-power applications. Solid State Ionics 144, 185-192 (2001).
Qu, Y ., Zhou, H. & Duan, X. Porous silicon nanowires. Nanoscale 3, 4060-4068 (2011).
Rauh, R. D., Shuker, F. S., Marston, J. M. & Brummer, S. B. Formation of Lithium Polysulfides in Aprotic
Media. J Inorg Nucl Chem 39, 1761-1766 (1977).
Reed, J. & Ceder, G. Charge, potential, and phase stability of layered Li(Ni0.5Mn0.5)O-2. Electrochem Solid
St 5, A145-A148 (2002).
Richman, E. K., Kang, C. B., Brezesinski, T. & Tolbert, S. H. Ordered mesoporous silicon through magnesium
reduction of polymer templated silica thin films. Nano Lett 8, 3075-3079 (2008).
Ryu, H. S. et al. Self-discharge characteristics of lithium/sulfur batteries using TEGDME liquid electrolyte.
Electrochim Acta 52, 1563-1566 (2006).
Ryu, H. S., Guo, Z. P ., Ahn, H. J., Cho, G. B. & Liu, H. K. Investigation of discharge reaction mechanism of
lithium vertical bar liquid electrolyte vertical bar sulfur battery. J Power Sources 189, 1179-1183 (2009).
Ryu, J. H., Kim, J. W., Sung, Y . E. & Oh, S. M. Failure modes of silicon powder negative electrode in lithium
secondary batteries. Electrochem Solid St 7, A306-A309 (2004).
Schuster, J. et al. Spherical Ordered Mesoporous Carbon Nanoparticles with High Porosity for Lithium-
Sulfur Batteries. Angew Chem Int Edit 51, 3591-3595 (2012).
Seh, Z. W. et al. fSulphur-TiO2 yolk-shell nanoarchitecture with internal void space for long-cycle lithium-
sulphur batteries. Nature Communications 4 (2013).
Sethuraman, V. A., Chon, M. J., Shimshak, M., Van Winkle, N. & Guduru, P . R. In situ measurement of
biaxial modulus of Si anode for Li-ion batteries. Electrochem Commun 12, 1614-1617 (2010).
Shenoy, V. B., Johari, P . & Qi, Y . Elastic softening of amorphous and crystalline Li-Si Phases with increasing
Li concentration: A first-principles study. J Power Sources 195, 6825-6830 (2010).
Sieradzki, K. et al. The dealloying critical potential. J Electrochem Soc 149, B370-B377 (2002).
Sieradzki, K., Corderman, R. R., Shukla, K. & Newman, R. C. Computer-Simulations of Corrosion - Selective
Dissolution of Binary-Alloys. Philos Mag A 59, 713-746 (1989).
Silva, J. A. et al. Sprayed boric acid as a dopant source for silicon ribbons. Sol Energ Mat Sol C 91, 1948-
1953 (2007).
131
Silvestri, H. H., Bracht, H., Hansen, J. L., Larsen, A. N. & Haller, E. E. Diffusion of silicon in crystalline
germanium. Semicond Sci Tech 21, 758-762 (2006).
Sneddon, I. N. The relation between load and penetration in the axisymmetric boussinesq problem for a
punch of arbitrary profile. International Journal of Engineering Science 3, 47-57 (1965).
Strasser, P . et al. Lattice-strain control of the activity in dealloyed core-shell fuel cell catalysts. Nat Chem 2,
454-460 (2010).
Su, Y . S. & Manthiram, A. Lithium-sulphur batteries with a microporous carbon paper as a bifunctional
interlayer. Nature Communications 3 (2012).
Tarascon, J. M. & Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 414,
359-367 (2001).
Thackeray, M. M., Dekock, A. & David, W. I. F. Synthesis and Structural Characterization of Defect Spinels in
the Lithium-Manganese-Oxide System. Mater Res Bull 28, 1041-1049 (1993).
Thakur, M. et al. Freestanding Macroporous Silicon and Pyrolyzed Polyacrylonitrile As a Composite Anode
for Lithium Ion Batteries. Chem Mater 24, 2998-3003 (2012).
Thakur, M., Isaacson, M., Sinsabaugh, S. L., Wong, M. S. & Biswal, S. L. Gold-coated porous silicon films as
anodes for lithium ion batteries. J Power Sources 205, 426-432 (2012).
Thakur, M., Sinsabaugh, S. L., Isaacson, M. J., Wong, M. S. & Biswal, S. L. Inexpensive method for producing
macroporous silicon particulates (MPSPs) with pyrolyzed polyacrylonitrile for lithium ion batteries. Sci Rep-
Uk 2, doi:Artn 795 (2012)
Tobishima, S. I., Yamamoto, H. & Matsuda, M. Study on the reduction species of sulfur by alkali metals in
nonaqueous solvents. Electrochim Acta 42, 1019-1029 (1997).
Tritsaris, G. A., Zhao, K. J., Okeke, O. U. & Kaxiras, E. Diffusion of Lithium in Bulk Amorphous Silicon: A
Theoretical Study. J Phys Chem C 116, 22212-22216 (2012).
Vankooten, J. J., Weller, G. A. & Ammerlaan, C. A. J. Electron-Paramagnetic Resonance on Iron-Acceptor
Pairs in Silicon. Phys Rev B 30, 4564-4570 (1984).
Wang, H. L. et al. Graphene-Wrapped Sulfur Particles as a Rechargeable Lithium-Sulfur Battery Cathode
Material with High Capacity and Cycling Stability. Nano Lett 11, 2644-2647 (2011).
132
Wen, C. J. & Huggins, R. A. Chemical Diffusion in Intermediate Phases in the Lithium-Tin System. J Solid
State Chem 35, 376-384 (1980).
Whittingham, M. S. Lithium batteries and cathode materials. Chem Rev 104, 4271-4301 (2004).
Wu, H. et al. Engineering Empty Space between Si Nanoparticles for Lithium-Ion Battery Anodes. Nano
Lett 12, 904-909 (2012).
Wu, H. et al. Stable cycling of double-walled silicon nanotube battery anodes through solid-electrolyte
interphase control. Nat Nanotechnol 7, 309-314 (2012).
Wu, H. et al. Stable Li-ion battery anodes by in-situ polymerization of conducting hydrogel to conformally
coat silicon nanoparticles. Nature Communications 4, 6 (2013).
Xu, C. X. et al. Low temperature CO oxidation over unsupported nanoporous gold. J Am Chem Soc 129, 42-
43 (2007).
Yamin, H., Penciner, J., Gorenshtain, A., Elam, M. & Peled, E. The Electrochemical-Behavior of Polysulfides
in Tetrahydrofuran. J Power Sources 14, 129-134 (1985).
Yang, Y . et al. New Nanostructured Li2S/Silicon Rechargeable Battery with High Specific Energy. Nano Lett
10, 1486-1491 (2010).
Yao, Y . et al. Interconnected Silicon Hollow Nanospheres for Lithium-Ion Battery Anodes with Long Cycle
Life. Nano Lett 11, 2949-2954 (2011).
Yi, R., Dai, F., Gordin, M. L., Chen, S. R. & Wang, D. H. Micro-sized Si-C Composite with Interconnected
Nanoscale Building Blocks as High-Performance Anodes for Practical Application in Lithium-Ion Batteries.
Adv Energy Mater 3, 295-300 (2013).
Yi, R., Dai, F., Gordin, M. L., Sohn, H. & Wang, D. H. Influence of Silicon Nanoscale Building Blocks Size and
Carbon Coating on the Performance of Micro-Sized Si-C Composite Li-Ion Anodes. Adv Energy Mater 3,
1507-1515 (2013).
Yoo, J. K., Kim, J., Jung, Y . S. & Kang, K. Scalable Fabrication of Silicon Nanotubes and their Application to
Energy Storage. Adv Mater 24, 5452-5456 (2012).
Yoshimura, K., Suzuki, J., Sekine, K. & Takamura, T. Measurement of the diffusion rate of Li in silicon by the
use of bipolar cells. J Power Sources 174, 653-657 (2007).
You, T. Y ., Niwa, O., Tomita, M. & Hirono, S. Characterization of platinum nanoparticle-embedded carbon
133
film electrode and its detection of hydrogen peroxide. Anal Chem 75, 2080-2085, doi:Doi
10.1021/Ac026337w (2003).
Zhang, B., Qin, X., Li, G. R. & Gao, X. P . Enhancement of long stability of sulfur cathode by encapsulating
sulfur into micropores of carbon spheres. Energ Environ Sci 3, 1531-1537 (2010).
Zhang, S. S. & Read, J. A. A new direction for the performance improvement of rechargeable lithium/sulfur
batteries. J Power Sources 200, 77-82 (2012).
Zhang, S. S. Role of LiNO3 in rechargeable lithium/sulfur battery. Electrochim Acta 70, 344-348 (2012).
Zhang, X. C., Shyy, W. & Sastry, A. M. Numerical simulation of intercalation-induced stress in Li-ion battery
electrode particles. J Electrochem Soc 154, A910-A916 (2007).
Zhang, Z. H. et al. Generalized Fabrication of Nanoporous Metals (Au, Pd, Pt, Ag, and Cu) through Chemical
Dealloying. J Phys Chem C 113, 12629-12636 (2009).
Zhao, K. J. et al. Concurrent Reaction and Plasticity during Initial Lithiation of Crystalline Silicon in Lithium-
Ion Batteries. J Electrochem Soc 159, A238-A243 (2012).
Zhao, Y ., Liu, X. Z., Li, H. Q., Zhai, T. Y . & Zhou, H. S. Hierarchical micro/nano porous silicon Li-ion battery
anodes. Chem Commun 48, 5079-5081, doi:Doi 10.1039/C2cc31476b (2012).
Zhou, X. S., Yin, Y . X., Wan, L. J. & Guo, Y . G. Facile synthesis of silicon nanoparticles inserted into graphene
sheets as improved anode materials for lithium-ion batteries. Chem Commun 48, 2198-2200 (2012).
Zhu, X. J. et al. Synthesis and characteristics of Li3V2(PO4)(3) as cathode materials for lithium-ion
batteries. Solid State Ionics 179, 1679-1682 (2008).
Abstract (if available)
Abstract
Lithium-ion battery has generated great impacts on portable electronics since its first commercialization in 1990s. Existing lithium-ion batteries using graphite as anode have already been widely used in mobile applications
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Nanostructured silicon anode and sulfur cathode for lithium rechargeable batteries
PDF
Modification of electrode materials for lithium ion batteries
PDF
Nanomaterials for energy storage devices and electronic/optoelectronic devices
PDF
Nanostructure design of sulfur cathodes and lithium metal anodes for lithium-ion batteries
PDF
Cathode and anode materials for sodium ion batteries
PDF
Nanomaterials for macroelectronics and energy storage device
PDF
Chemical vapor deposition of graphene and two-dimensional materials: synthesis, characterization, and applications
PDF
Understanding the role of electrode design in determining the electrochemical performance of high-energy/high-power lithium-ion and lithium-sulfur batteries
PDF
Electrocatalysts for direct liquid-feed fuel cells and advanced electrolytes for lithium-ion batteries
PDF
Biophysical studies of nanosecond pulsed electric field induced cell membrane permeabilization
PDF
Lithium-ion batteries (LIB) industrial recycling and policies
PDF
Novel nanomaterials for electronics, optoelectronics and sensing applications
PDF
Semiconducting metal oxide nanostructures for scalable sensing applications
PDF
Nanostructured III-V photoelectrodes for high performance, high durability solar water splitting
PDF
Studies on lithium-ion battery electrolytes and three component Strecker reaction
PDF
Silicon micro-ring resonator device design for optical interconnect systems
PDF
New materials and device structure for organic light-emitting diodes
PDF
One-dimensional nanostructures for chemical sensing, transparent electronics, and energy conversion and storage devices
PDF
Substrate-induced effects in thermally cycled graphene & electron and thermoelectric transports across 2D material heterostructures
PDF
Silicon carbide ceramic membranes
Asset Metadata
Creator
Ge, Mingyuan
(author)
Core Title
Nanostructured silicon for lithium-ion battery anode
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Materials Science
Publication Date
01/29/2015
Defense Date
12/17/2014
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
anode,cost-efficient,Li-ion battery,OAI-PMH Harvest,porous silicon
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Zhou, Chongwu (
committee chair
), Goo, Edward K. (
committee member
), Wu, Wei (
committee member
)
Creator Email
gmysage@gmail.com,mingyuag@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c3-527343
Unique identifier
UC11297910
Identifier
etd-GeMingyuan-3148.pdf (filename),usctheses-c3-527343 (legacy record id)
Legacy Identifier
etd-GeMingyuan-3148.pdf
Dmrecord
527343
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Ge, Mingyuan
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
anode
cost-efficient
Li-ion battery
porous silicon