Characterization of Next Generation Lithium-ion Battery Materials Through Electrochemical, Spectroscopic, and Neutron-Based Methods
DISSERTATION
Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University
By
Danny Xin Liu
Graduate Program in Chemistry
The Ohio State University
2015
Dissertation Committee:
Anne C. Co, Advisor
Prabir K. Dutta
Sherwin J. Singer
Harris P. Kagan
Copyright by
Danny Xin Liu
2015
Abstract
The development of a real-time quantification of Li transport using a non-destructive
neutron method to measure the Li distribution upon charge and discharge in a Li-ion cell
is reported here. Using in situ neutron depth profiling (NDP), we probed the onset of
lithiation in high capacity Sn and Al anodes and visualized the enrichment of Li atoms on
the surface which is followed by their propagation into the bulk. The de-lithiation
process shows the removal of near surface Li, leading to a loss in coulombic efficiency assigned to trapped Li within the intermetallic material. In situ NDP developed in this
work provides temporal and spatial measurement of Li transport within the battery material with exceptional sensitivity. Direct application of Fick’s Laws allowed for the effective lithium diffusion coefficient to be calculated from the lithium concentration profiles. This diagnostic tool opens up possibilities of understanding rates of Li transport
and their distribution to guide materials development for efficient storage mechanisms.
In addition, in situ NDP was employed to explore the feasibility of utilizing Al as the anode current collector. The results indicate that an Al anode current collector can be
employed as a strategy to improve energy density while reducing cost, provided that the
surface of the Al is not in direct contact with Li+ or the voltage is limited to a value above
the Al lithiation redox voltage. Our observations provide important mechanistic insights
to the design of advanced battery materials.
ii
Dedication
This document is dedicated to my family.
iii
Acknowledgments
I am very grateful to The Ohio State University (OSU) and the Department of
Chemistry and Biochemistry for the resources and opportunity to pursue a doctoral degree. I am especially indebted to my advisor, Professor Anne C. Co, and my colleagues within the Co Research Group for their patience, support, and guidance throughout my time at OSU. Prof. Co has been an influential mentor establishing an environment which fosters independent thought, innovative approaches, and collaborative strategies to address relevant technical challenges. I am very fortunate to have worked with Dr. Jennifer M. Black and Dr. Eric J. Coleman – two outstanding scientists that I consider dear friends.
Of the numerous Faculty members that have contributed towards my development as a scientist. I would like to thank Prof. Prabir K. Dutta for teaching the Advanced
Analytical Chemistry course – that provided the foundation upon which I have refined myself as an experimental scientist. I am forever indebted to Prof. Sherwin J. Singer for affording me the opportunity and support to develop as a physical chemistry student. I want to thank Prof. Walter R. Lempert, Prof. Heather C. Allen, and Prof. Sherwin J.
Singer for serving as my First Year Oral Exam committee members. Additionally, I am grateful for the insights and critiques provided by my Candidacy Exam committee members, Prof. Prabir K. Dutta, Prof. Heather C. Allen, Prof. Yiying Wu. Finally, I
iv
would like to express my sincere gratitude to my Dissertation committee members listed
on the title page.
I have had the privilege of working with outstanding staff members within the
Department Chemistry and Biochemistry. My teaching responsibilities have been
primarily supervised by Dr. Steve Kroner. I thank Dr. Kroner for his guidance, support,
and advice on countless academic, professional, and personal topics. I would like to
acknowledge Jerry Hoff, Larry Antal, and Ryan Shea from the machine shop for their
time, knowledge, and expertise in numerous projects and designs that have contributed
towards my research efforts. I would like to thank Eric Jackson, Eric Kesselring, and
John Sullivan from the electronics shop for sharing their knowledge in electronics and
device fabrication. I want to thank Lisa Hommel for sharing her knowledge and
experience in surface chemistry and training me on the operations of the x-ray photoelectron spectrometer (XPS). I would like to acknowledge Spencer Porter, Tricia
Meyer, and Andrew Sharits for their maintenance and troubleshooting efforts regarding the x-ray diffractometer. Special thanks is reserved for the administrative staff especially
Judy Brown, Jennifer Hambach, Kelly Burke and Thomas Hyle for organizing departmental events, disseminating crucial deadlines, and all of the necessary behind-the- scenes daily tasks and documentation that keep the department functional. Additionally,
I would like to acknowledge Barbara Bennett (B2) from the computer support staff for her assistance during technical difficulties related to the computer network.
I am grateful for the opportunities that resulted in collaborative relationships leading to many of the findings within this dissertation. Prof. Lei R. Cao and Dr. Jinghui
v
Wang have been instrumental collaborators who introduced me to the field of neutron depth profiling (NDP). Special recognition is reserved for Dr. R. Gregory Downing, a staff scientist at the National Institute of Standards and Technology (NIST) Center for
Neutron Research (NCNR). I thank Dr. Downing for his experience, insights, and patience throughout the collaborative experiments. Additionally, I have very grateful to have been granted access to the Cold NDP system at NCNR, which have led to pioneering findings and the advancement of in situ NDP within the energy storage community. The vast majority of my experience and knowledge of neutron depth profiling is a direct result of working with Dr. Downing.
I would like to express my sincere appreciation to Rachelle L. Speth for her friendship, devotion, love and support throughout my graduate career. Returning her
“misplaced” mitten was one of the best decisions I could have ever made.
Finally, no words could express my gratitude for my parents. They have always encouraged me in my academic pursuits, challenged my rationalizations, provided guidance and supported me through turbulent times. They have had undeniably the greatest influence on my personal development and it is their experiences and insurmountable sacrifices that have shaped my outlook and afforded me the opportunities to pursue my dreams. I am and will forever be indebted to them.
vi
Vita
2004...... Woodbury High School
2008...... B.S. Chemistry, University of Minnesota –
Twin Cities
2010 to present ...... Graduate Teaching/Research Associate,
Department of Chemistry and Biochemistry,
The Ohio State University
Publications
Liu, D.; Wang, J.; Ke, P.; Qiu, J.; Canova, M.; Cao, L.; Co, A. “In Situ Quantification and Visualization of Lithium Transport with Neutrons.” Angew. Chem. Int. Ed., 53 (2014) 9498-9502.
Wang, J.; Liu, D.; Canova, M.; Downing, R. G.; Cao, L.; Co, A. “Profiling lithium distribution in Sn anode for lithium-ion batteries with neutrons.” J Radioanal Nucl Chem, 301, 1, (2014) 277-284.
Tan, C.; Leung, K. Y.; Liu, D.; Canova, M.; Downing, R. G.; Co, A.; Cao, L. “Gamma radiation effects on Li-ion battery electrolyte in neutron depth profiling for lithium quantification.” J Radioanal Nucl Chem. 305, 2, (2015) 675-680.
Fields of Study
Major Field: Chemistry
vii
Table of Contents
Abstract ...... ii
Dedication ...... iii
Acknowledgments...... iv
Vita ...... vii
Table of Contents ...... viii
List of Tables ...... xi
List of Figures ...... xii
Chapter 1. Introduction ...... 1
Chapter 2. In Situ Quantification and Visualization of Lithium Transport with
Neutrons ...... 6
2.1. Introduction ...... 6
2.2. Neutron Depth Profiling ...... 9
2.3. Methods ...... 11
2.4. Results and Discussion ...... 19
2.5. Conclusions ...... 24
2.6. Tables ...... 25
viii
2.7. Figures ...... 27
Chapter 3. Thermodynamics and Kinetics of Electrochemically formed Li-Al ...... 39
3.1. Introduction ...... 39
3.2. Methods ...... 44
3.3. Electrochemical and Microscopic Investigations ...... 48
3.4. In Situ Neutron Depth Profiling of Al Lithiation and De-lithiation ...... 51
3.5. In Situ X-ray Diffraction of Al Lithiation ...... 56
3.6. Gibbs Phase Rule and the Voltage Plateau ...... 60
3.7. Conclusions ...... 63
3.8. Figures ...... 65
Chapter 4. Aluminum Anode Current Collector in Lithium-ion Batteries ...... 89
4.1. Introduction ...... 89
4.2. Methods ...... 92
4.3. Lithiation of Al foil ...... 93
4.4. Lithiation and De-lithiation of 500 nm Sn thin film on Al foil ...... 94
4.5. Lithiation of Sn slurry on Al foil ...... 98
4.6. Conclusions ...... 100
4.7. Figures ...... 101
Appendix A: Tabulated literature Li diffusion coefficients ...... 125
ix
Appendix B: Derivation of Gibbs Phase Rule ...... 132
References ...... 134
x
List of Tables
Table 2.1:1Triton (3H) energy as a function of penetration depth through 12.5 μm Sn and
7.5 μm of Kapton film...... 25
Table 2.2:2Triton (3H) energy as a function of penetration depth through 17 μm Al and
7.5 μm of Kapton film...... 26
Table A.1:3Literature lithium diffusion coefficient in β-LiAl at various temperatures. 126
Table A.2:4Literature lithium diffusion coefficient in α-LiAl at various temperatures. 130
xi
List of Figures
Figure 2.1:1(A) A schematic representation of the battery components with 7.5 μm
Kapton film, 12.5 μm Sn foil, 25 μm porous “Celgard® 2004” separator containing the electrolyte, and 300 μm Li metal; (B) Illustration of the NDP chamber at NCNR; and (C)
Three snap shots of the in situ NDP spectra showing Li transport during charging/discharging a battery...... 27
Figure 2.2:2Lithium concentration profiles within a 12.5 m Sn foil as a function of time.
(dashed) before electrochemical lithiation, (A) Lithiation spectra plotted every 60 min
interval from 20 min to 740 min at 0.4 V vs. Li/Li+(reaching approx. 200 mAh/g); (B)
De-lithiation spectra plotted every 60 min interval from 10 min to 200 min de-lithiation at
1.0 V vs. Li/Li+...... 28
Figure 2.3:3The NDP spectrum of the energy standard (Boron SRM 93a) showing the
step-shaped peaks that are used for energy calibration...... 29
Figure 2.4:4Schematic of charge particle escaping path to the detector...... 30
Figure 2.5:5Energy calibration curve using the energy standard (Boron SRM 93a)...... 31
Figure 2.6:6NDP spectrum of Boron-10 with peaks representing the emitted charged
particle reaction products...... 32
Figure 2.7:7Lithium diffusion coefficient, within aluminum, obtained through application of Fick’s first law and second law during the second lithiation. The error bars represent an estimated 40% uncertainty resulting from the variation of electrode composition and is xii
calculated based on the percent difference between the depths that a triton charge particle
could traverse through in Al vs. LiAl...... 33
Figure 2.8:8Li concentration profile with a segment around 4 um (3 – 5um) fitted with a
second-order polynomial. The resulting fitted equation was differentiated twice to obtain
the curvature. The same process is repeated using a linear fit to obtain the concentration
gradient...... 34
Figure 2.9:9Lithium concentration at various regions of the battery as a function of
lithiation (0 to 740 min) and removal of the applied voltage (740 – 1150 min), de-
lithiation (830-1050 min) and removal of applied voltage (1050-1140 min)...... 35
Figure 2.10:10Li diffusion coefficient as a function of lithiation time. Values calculated
from Fick’s first law using Li concentrations from 1.2 to 5 m (black squares) and 5 to
10 m (blue triangles) and approximation of the Li flux from the electrochemical current density. The error bars represent the 10% upper boundary in the overall uncertainty of
the NDP spectra resulting from the different stopping powers of pure Sn versus Li2Sn5. 36
Figure 2.11:11Total integrated Li within the electrode (blue square) and the change in Li
concentration for every 10 min interval (red triangles)...... 37
Figure 2.12:12Ratio of electrochemical charge passed to the charge equivalent to the integrated Li (blue square) and the ratio of the change in electrochemical charge passed
and the corresponding change in the Li concentration for every 20 min interval (red
triangles)...... 38
Figure 3.1:13Schematic representation of a custom built 3-electrode Swagelok Cell.
Figure created by Amy Casaday...... 65
xiii
Figure 3.2:14XPS survey spectrum of pristine Al foil, 300, and 600 seconds of Ar+ etch.
...... 66
Figure 3.3:15Typical cyclic voltammogram of aluminum foil collected using a sweep rate
of 50 μV/s. Asterisk represents location of the voltammogram where representative SEM
image was taken (Figure 3.5)...... 67
Figure 3.4:16Typical galvanostatic (30 μA; 24.4 μA/cm2) lithiation of aluminum foil with
inset displaying voltage minimum. Asterisk represents location of the voltammogram
where representative SEM image was taken (Figure 3.6)...... 68
Figure 3.5:17Scanning electron micrograph (SEM) reflects surface morphology at the CV
reduction peak of 0.13 V vs Li/Li+ indicated by asterisk on Figure 3.3...... 69
Figure 3.6:18Scanning electron micrograph (SEM) reflects the surface morphology at the
end of the voltage plateau after passing 250 mC of charge indicated by asterisk on Figure
3.4...... 70
Figure 3.7:19SEM of pristine Al foil...... 71
Figure 3.8:20Lithium concentration profiles obtained through in situ NDP. First lithiation (0 – 60 mins) of a 16 μm Al foil...... 72
Figure 3.9:21Lithium concentration profiles obtained through in situ NDP. OCP period
(65 – 145 mins) of a 16 μm Al foil...... 73
Figure 3.10:22Lithium concentration profiles obtained through in situ NDP. De-lithiation
(150 – 215 mins) of a 16 μm Al foil...... 74
Figure 3.11:23Lithium concentration profiles obtained through in situ NDP. Second
lithiation (220 – 335 mins) of a 16 μm Al foil...... 75
xiv
Figure 3.12:24Al 2p region of the pristine Al foil prior to etching...... 76
Figure 3.13:25Electrochemical characterization during in situ NDP. The voltage transient is collected during the OCP resting period immediately following the first lithiation...... 77
Figure 3.14:26Lithium concentration, at various depths, as a function of elapsed time.
The inset highlights the first lithiation, OCP, and de-lithiation periods...... 78
Figure 3.15:27NDP integrated charge and its first derivative plotted as a function of elapsed time. The inset highlights the first lithiation, OCP, and de-lithiation periods. ... 79
Figure 3.16:28Echem Q / NDP equivalent Q ratio plotted as a function of elapsed time. 80
Figure 3.17:29In situ Cu Kα x-ray diffraction patterns, collected at 30 minute intervals, for the lithiation of aluminum foil. Inset highlights the appearance of the Al (111) peak followed by the onset of the LiAl (200) peak. The asterisk (*) indicate reflections contributed from experimental setup (casing components)...... 81
Figure 3.18:30Cu Kα x-ray diffraction background pattern of pristine Al, 90 mins of lithiation, and OCP resting period – focusing on the Al (200) region. For comparison, the peak position for a reference Al powder sample is indicated by a vertical dashed line. .. 82
Figure 3.19:31Cu Kα x-ray diffraction background pattern of pristine Al, 90 mins of lithiation, and OCP resting period – focusing on the Al (220) region. For comparison, the peak position for a reference Al powder sample is indicated by a vertical dashed line. .. 83
Figure 3.20:32The aluminum (200) reflection exhibited shifts towards higher 2θ values during lithiation...... 84
xv
Figure 3.21:33The shifted Al(200) is fitted as having contributions from pristine Al (red
trace) and solid solution α-phase (blue trace)...... 85
Figure 3.22:34Ex situ XRD pattern of lithiated aluminum foil (0.2 V, 23.5 hours) with
reference peak positions. Due to the numerous, low intensity peaks of the reference
patterns, only peaks with normalized intensities of 10% or greater are plotted. The
asterisk (*) indicate reflections contributed from experimental setup (casing components).
...... 86
Figure 3.23:35Reference XRD patterns of all thermodynamically stable lithium-
aluminum intermetallic alloys with experimentally collected ex situ lithiation of Al foil at
0.2 V for 23.5 hours. Due to the numerous, low intensity peaks of the reference patterns, only reference peaks with normalized intensities of 10% or greater are plotted...... 87
Figure 3.24:36Mechanism of aluminum lithiation. Lithium exists as ions in electrolyte
(a). The Li-ions are reduced at the surface of the Al (b). As more Li-ions are reduced the
solid-solution α-phase forms (c). Once the Li concentration reaches the solubility limit,
the β-LiAl intermetallic forms (d) and propagates towards the bulk (e)...... 88
Figure 4.1:37Typical cyclic voltammogram of Al foil in 1M LiPF6 EC:DMC (1:1 vol)
starting at 3.0 V to 0.1 V at a scan rate of 0.05 mV/s. Inset reflects the reduction peak
and the lithiation onset voltage at 0.2 V...... 101
Figure 4.2:38In situ NDP lithium concentration profiles via potentiostatic lithiation of Al
foil at 0.2 V. The time values in the legend allows for relative comparison of the elapsed
time within each plot...... 102
xvi
Figure 4.3:39A 0.2 V potentiostatic current transient during in situ NDP of 500 nm PVD
Sn on 16 μm Al foil. The sample was assembled in an Ar-filled glovebox (O2 & H2O <
0.5 ppm) using a modified 2032 coin cell, with lithium metal as the counter/reference electrode, Celgard® 2400 in 1M LiPF6 in EC:DMC (1:1 wt), and a working electrode area of 1.2 cm2...... 103
Figure 4.4:40A 0.5 V potentiostatic current transient during in situ NDP of 500 nm PVD
Sn on 16 μm Al foil. The sample was assembled in an Ar-filled glovebox (O2 & H2O <
0.5 ppm) using a modified 2032 coin cell, with lithium metal as the counter/reference electrode, Celgard® 2400 in 1M LiPF6 in EC:DMC (1:1 wt), and a working electrode area of 1.2 cm2...... 104
Figure 4.5:41A 1.0 V potentiostatic current transient during in situ NDP of 500 nm PVD
Sn on 16 μm Al foil. The sample was assembled in an Ar-filled glovebox (O2 & H2O <
0.5 ppm) using a modified 2032 coin cell, with lithium metal as the counter/reference electrode, Celgard® 2400 in 1M LiPF6 in EC:DMC (1:1 wt), and a working electrode area of 1.2 cm2...... 105
Figure 4.6:42In situ lithium neutron depth profiles during 0.2 V potentiostatic measurement. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 106
Figure 4.7:43In situ lithium neutron depth profiles during 0.5 V potentiostatic measurement. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 107
xvii
Figure 4.8:44In situ lithium neutron depth profiles during 1.0 V potentiostatic
measurement. The time values in the legend allows for relative comparison of the
elapsed time within each plot...... 108
Figure 4.9:45The first cycle cyclic voltammogram (CV) of 500 nm Sn on 16 μm Al foil
collected starting at 1.0 V scanning towards 0.1 V at a scan rate of 0.2 mV/s. CVs are
labeled with numbers to designate the corresponding lithium concentration profiles. ... 109
Figure 4.10:46The second cycle cyclic voltammogram (CV) of 500 nm Sn on 16 μm Al
foil collected starting at 1.0 V scanning towards 0.1 V at a scan rate of 0.2 mV/s. CVs are
labeled with numbers to designate the corresponding lithium concentration profiles. ... 110
Figure 4.11:47In situ lithium neutron depth profiles during the first lithiation segment of
the first cycle CV (1.0 V scanning towards 0.1 V at a scan rate of 0.2 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
CV. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 111
Figure 4.12:48In situ lithium neutron depth profiles during the first de-lithiation segment
of the first cycle CV (1.0 V scanning towards 0.1 V at a scan rate of 0.2 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
CV. The time values in the legend allows for relative comparison of the elapsed time
within each plot...... 112
Figure 4.13:49In situ lithium neutron depth profiles during the first lithiation segment of
the second cycle CV (1.0 V scanning towards 0.1 V at a scan rate of 0.2 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
xviii
CV. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 113
Figure 4.14:50In situ lithium neutron depth profiles during the first de-lithiation segment
of the second cycle CV (1.0 V scanning towards 0.1 V at a scan rate of 0.2 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
CV. The time values in the legend allows for relative comparison of the elapsed time
within each plot...... 114
Figure 4.15:51The first cycle cyclic voltammogram (CV) of 500 nm Sn on 16 μm Al foil
collected starting at 1.0 V scanning towards 0.05 V at a scan rate of 0.1 mV/s. CVs are
labeled with numbers to designate the corresponding lithium concentration profiles. ... 115
Figure 4.16:52The second cycle cyclic voltammogram (CV) of 500 nm Sn on 16 μm Al
foil collected starting at 1.0 V scanning towards 0.05 V at a scan rate of 0.1 mV/s. CVs
are labeled with numbers to designate the corresponding lithium concentration profiles.
...... 116
Figure 4.17:53In situ lithium neutron depth profiles during the first lithiation segment of
the first cycle CV (1.0 V scanning towards 0.05 V at a scan rate of 0.1 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
CV. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 117
Figure 4.18:54In situ lithium neutron depth profiles during the first de-lithiation segment
of the first cycle CV (1.0 V scanning towards 0.05 V at a scan rate of 0.1 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
xix
CV. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 118
Figure 4.19:55In situ lithium neutron depth profiles during the first lithiation segment of
the second cycle CV (1.0 V scanning towards 0.05 V at a scan rate of 0.1 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
CV. The time values in the legend allows for relative comparison of the elapsed time within each plot...... 119
Figure 4.20:56In situ lithium neutron depth profiles during the first de-lithiation segment
of the second cycle CV (1.0 V scanning towards 0.05 V at a scan rate of 0.1 mV/s)
measurement. The numbers within the legend designate the corresponding points on the
CV. The time values in the legend allows for relative comparison of the elapsed time
within each plot...... 120
Figure 4.21:57Scanning electron micrograph (SEM) of Sn nano-particles with conductive
carbon in a PVDF matrix slurry coated onto 16 μm Al foil...... 121
Figure 4.22:58Energy dispersive spectrum (EDS) indicating representative elemental composition of the slurry sample...... 122
Figure 4.23:59Voltage profile of 0.5 mA constant current lithiation (C/6) during in situ
NDP...... 123
Figure 4.24:60In situ NDP Li concentration profiles of slurry coated Sn nano-particles on
Al current collector...... 124
xx
Chapter 1. Introduction
The driving force for renewable energy storage research is propelled by society’s
heavy reliance on fossil fuel sources as the major energy source sustaining the
transportation and industrial sectors. The finite fossil fuel reserves will eventually be depleted, prompting great interest in renewable energy research on wind and solar
sources. The mismatch between the intermittent ability of these sources and the variable
consumer demand is the major obstacle for large scale integration. Among the leading
contenders, electrochemical energy storage technologies offer numerous advantages
including efficient energy storage, scalability, and modularity to meet power grid and
electrified vehicle requirements1.
With exceptional specific power and specific energy densities, lithium-ion
batteries have established its dominance as the portable energy storage device of choice,
owing to lithium’s low molecular weight, small ionic radius, and low redox voltage1, 2.
However, the energy densities afforded by current lithium-ion battery technologies
cannot meet the requirements of sustained, long-range operation of electrified vehicles and, in addition, is too cost-prohibitive for grid applications. Thus, directions for the next generation of lithium-ion battery technology must include research on intrinsic properties of new materials for which improvements in energy density, power density, safety, lifetime, and cost may be realized.
1
Improvements in the energy density of next generation lithium-ion technology
require (1) increasing the voltage difference between the anode/cathode, (2) maintaining
the amount of available lithium for lithiation, (3) and/or decreasing mass of the active
material per electron transferred3. Decreasing the mass of the active material per electron
transferred necessitates implementation of new electrodes with higher lithium packing
density per unit mass of the electrode material. Alternatively, reducing the mass
contribution of inactive components to the overall battery weight can also improve
energy density. Thus, research on suitable electrode materials and inactive components
for the next generation of lithium-ion batteries has been the focus of many investigators4.
Lithium-ion (Li-ion) batteries have been revolutionary in the advancement of
portable electronic devices. The first commercialized primary (non-rechargeable) Li-ion
batteries were introduced in the 1970s and featured elemental lithium as the anode, and a
transition metal sulfide (or oxide) as the cathode5. Initial developments of secondary
(rechargeable) Li-ion batteries with elemental lithium anodes proved to be challenging
due to the nucleation and subsequent growth of lithium dendrites upon charge/discharge cycles6. The lithium dendrites eventually penetrate the battery separator, reaching the
cathode, and causing short circuits. The rapid temperature increase from the short circuit
current combined with the flammable electrolyte leads to thermal runaway reactions, resulting in fires/explosions7. Improvements in safety and prolonged cycle life led to the
5 commercialization of graphite-LiCoO2 Li-ion batteries . During battery charging, current
is forced through the external circuit, from the cathode towards the anode, as lithium is
concurrently removed from the LiCoO2 cathode and enters the electrolyte. In a concerted
2
fashion, electrochemical reduction (insertion) of the lithium, from the electrolyte, occurs at the graphite anode. The reverse process (lithium oxidation at the anode, transition
metal reduction at the cathode) occurs during battery discharge.
The current graphite anode in Li-ion batteries offers numerous advantages
including good cycle life at relatively low cost. However, graphite anodes are limited by
their theoretical specific capacity of 375 mAh/g, ultimately limiting the energy density of
the battery. One approach towards increasing the energy density of Li-ion batteries is to
use higher capacity electrode materials. Among other candidates, aluminum (Al) and tin
(Sn), with their relatively high theoretical specific capacities of 993 (at LiAl β-phase) and
959 (at Li17Sn4) mAh/g, respectively, may prove to be promising anode materials.
During the alloying/de-alloying process with Li, aluminum suffers from volumetric
expansions of ~100%, forming an intermetallic LiAl β-phase alloy. Similarly, tin also
experience volumetric expansions on the order of ~300%, which is attributed to the
formation of thermodynamically stable intermetallic phases of Li2Sn5, LiSn, Li7Sn3,
Li5Sn2, Li13Sn5, Li7Sn2, Li17Sn4. This significant volumetric change causes mechanical
strain leading to particle pulverization of the active material and believed to be a
contributing factor to capacity fade. Fundamental understanding of the alloying/de-
alloying (i.e. lithiation/de-lithiation) and associated processes are necessary in order to
address capacity fade and provide strategies to optimize the chemical and structural
parameters of the next generation of Li-ion battery electrodes.
The dynamic nature of battery operation requires in situ methodologies to capture
the transient physical, chemical, and electrochemical processes. Ideal in situ techniques
3
would allow for non-destructive, non-invasive, and quantitative characterization of the battery with spatial and temporal resolution, enabling direct monitoring of the evolution of surface and bulk processes. Recent research efforts employing in situ x-rays, scanning probe, optical, magnetic resonance, and neutron-based techniques have improved the understanding of battery processes8. These techniques have been employed to study the
electrode surface, electrode/electrolyte interface, and bulk properties and the associated
processes with resolution capabilities ranging from the atomic to micron (μm) scale. By
probing the changes in local geometrical and electronic structure, crystallographic phases,
chemical compositions, and morphologies, fundamental insights of dynamic processes,
with particular emphasis on probing lithium transport, can be achieved. A limited
number of non-destructive techniques, including energy dispersive x-ray diffraction
(EDXRD) and nuclear magnetic resonance (NMR), allow for characterization as a
function of depth, where a series of depth profiles can yield insights to the material
structure and composition throughout battery operation9, 10, 11. However, diffraction
methods are limited to probing crystalline systems where periodic, long-range order is required and NMR is limited as a local probe. Thus, a generalized method that can be employed to study the representative macroscopic lithium transport phenomena for any
Li containing component within the battery is desired.
In this work, in situ neutron depth profiling was developed as a general technique
to visualize and quantify the lithium transport during lithiation/de-lithiation of Sn and Al.
The lithium concentration was resolved spatially and temporally through the electrolyte and throughout the bulk of the Sn and Al electrodes. Effective lithium diffusion
4
coefficients within Sn and Al were calculated from the concentration profiles through direct application of Fick’s Laws. Comparing the electrochemical and NDP results allowed for the quantification of parasitic reactions and evaluation of coulombic efficiencies. These results provide insights in addressing issues related to transport, materials selection, structural integrity, and failure/degradation mechanisms.
Phase evolution during Al lithiation was characterized using in situ XRD. Results suggest that the lithiation mechanism of Al includes the formation of a solid-solution α- phase that precedes and facilities intermetallic LiAl β-phase nucleation. The voltage plateau observed during galvanostatic lithiation of Al is discussced through the application of the Gibbs Phase Rule.
Additionally, in situ NDP was employed to explore the viability of using Al as the anode current collector. Sn thin films (500 nm) and slurry samples were deposited on Al foil. Results indicated preferential and reversible lithiation/de-lithiation of the Sn, without lithiation of the Al foil. Isolating Al from lithium-ions or limiting the voltage to a value positive of the Al lithiation voltage are necessary conditions for using Al as the anode current collector.
5
Chapter 2. In Situ Quantification and Visualization of Lithium Transport with
Neutrons
2.1. Introduction
Lithium-ion batteries (LIBs) are the most promising rechargeable energy storage system and because of their high gravimetric and volumetric energy capacity, they have been widely used in portable devices such as laptops, mobile phones and electronic devices. Owing to their success and promising properties, LIBs are now being considered for storage and distribution of renewable energy sources, and in the automotive industry, for powering hybrid and plug-in vehicles. The increasing demands on batteries has stimulated the development of advanced materials, diagnostic tools and battery models that could lead to better performance in terms of storage capacity, kinetics, and long-term cyclability. A real-time quantification of Li transport using a non-destructive neutron method to measure the Li distribution upon charge and discharge of a Li-ion cell is reported here. As a neutral particle exhibiting selective interaction cross sections with different isotopes, neutrons are ideal for probing Li atoms deep inside the battery. We have developed an in situ technique based on neutron depth profiling (NDP) to provide temporal and spatial measurement of Li within a material and to visualize its transposition during dynamic charging and discharging. Our experimental design also allows measurement in low vapor pressure “wet” electrolyte inside the NDP chamber, which opens up possibilities for probing a variety of energy storage materials and guide
6 the materials development for efficient storage. In this work, we probed the onset of lithiation in a model Sn electrode using in situ NDP and visualized the enrichment of Li atoms on the surface while monitoring the propagation of Li into the bulk. The methods developed here are highly valuable in providing mechanistic insights into the design of advanced battery components and in the evaluation of transport of light elements in other electronic devices.
Non-destructive in situ methodologies for investigating reaction dynamics, transient processes in non-equilibrium reactions and materials transformations, such as those in an operating battery, are crucial in advancing the understanding of reaction processes, structural changes in materials, and failure mechanisms under operating conditions for improving next generation materials. Real-time visualization and quantification of Li ions in an energy storage material can be a powerful tool for the understanding of the Li storage properties and guide the development of durable, cyclable and efficient storage materials. Especially as Li diffusion through solids is often the rate limiting step in battery processes 12, 13. A direct measurement of Li distribution within a material in real time can provide valuable information for modeling the lithium transport at the fundamental and systems level.
In situ X-ray diffraction (XRD), Mössbauer spectroscopy, nuclear magnetic resonance, transmission electron microscopy, scanning probe microscopy (AFM, STM), synchrotron x-ray tomography 14, 15, 16, 17, 18 have been used to quantify crystal structure evolution and visualize the expansion/contraction of electrode materials during lithiation and de-lithiation processes, each with their respective advantages and limitations. In situ
7
neutron techniques such as NDP 19, 20, while relatively new for applications in battery
materials research, can be quite powerful due to the neutron penetration and the high
neutron reaction cross section to certain light elements (e.g., 6Li, 10B, 14N, 7Be). This
selective nature of neutrons results in spectra that allows for direct tracking and counting
of Li atoms without resorting to inferring Li transport phenomenon from indirect observations. Here we demonstrate that in situ NDP is an effective tool for the quantification and visualization of Li transport in real time, under battery charge and discharge operations.
8
2.2. Neutron Depth Profiling
NDP is an analytical technique for quantitative measurements of elements that
have a high reaction cross section of absorbing neutrons, as a function of depth into a
solid material (Figure 2.1). 6Li absorbs neutrons, with reaction cross section of 938 barns
for neutron energy at 25 meV and 2355 barns at 4 meV, governed through the following nuclear reaction: