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
/
Nanomaterials for energy storage devices and electronic/optoelectronic devices
(USC Thesis Other)
Nanomaterials for energy storage devices and electronic/optoelectronic devices
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
NANOMATERIALS FOR ENERGY STORAGE DEVICES AND ELECTRONIC/OPTOELECTRONIC DEVICES By Chenfei Shen _________________________________________________________________ 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 2018 Copyright 2018 Chenfei Shen ii Dedication Dedicated to my beloved parents and my wife Xiaochen Cao iii Acknowledgement First of all, I would like to express my deep gratitude to my research advisor Prof. Chongwu Zhou. He has provided me the opportunity to pursue my passion in research. During the past four and half years, he not only taught me professional knowledge and skills, more importantly, his mentorship equips me with problem solving skills, which helps me not only be an experimentalist and researcher, but also be an independent thinker. Without his encouragement and support, I would never make these achievements. I would also thank my dissertation committee members Prof. Edward Goo, Prof. Wei Wu, and Prof. Jongseung Yoon for their helpful suggestions and comments, as well as Prof. Andrew Armani, Prof. Han Wang, and Prof. Pin Wang for serving on my qualifying exam committee. My appreciation also extends to my collaborators Dr. Chongmin Wang and Dr. Langli Luo in Pacific Northwest National Laboratory, Prof. Fengnian Xia and Mr. Shaofan Yuan in Yale University, Dr. Jiangbin Wu in USC, Prof. Wenzhuo Wu and Mr. Yixiu Wang in Purdue University, Dr. Matthew Mecklenburg in the Center for Electron Microscopy and Microanalysis in USC. Moreover, I would like to express my appreciation to my group members. I would like to thank Dr. Mingyuan Ge, Dr. Jiepeng Rong, and Dr. Xin Fang for their helpful guidance during my junior year. I also want to thank my current and former group members: Anyi iv Zhang, Yihang Liu, Chi Xu, Qingzhou Liu, Fanqi Wu, Zhen Li, Hongyu Fu, Dr. Bilu Liu, Dr. Gang Liu, Dr. Luyao Zhang, Dr. Ahmad Abbas, Dr. Yu Cao, Dr. Sen Cong, Dr. Hui Gui, Dr. Haitian Chen, Dr. Yuqiang Ma, Dr. Yuchi, Che, Dr. Maoqing Yao, Dr. Xuan Cao, Dr. Liang Chen, Dr. Nappadol Aroonyadet, Dr. Shelley Wang, I really enjoyed working with you. Finally, and the most importantly, I would like to thank my beloved parents and my wife Xiaochen Cao who have given me endless love, support, understanding, and encouragement. I would never have all these achievements today without you. v Table of Contents Dedication ........................................................................................................................... ii Acknowledgement ............................................................................................................. iii List of Tables .................................................................................................................... vii List of Figures .................................................................................................................. viii Abstract ............................................................................................................................ xvi Chapter 1: Introduction ....................................................................................................... 1 1.1 Introduction to lithium-ion battery ........................................................................ 1 1.1.1 Background .................................................................................................. 1 1.1.2 Operating principle of lithium-ion batter .................................................... 2 1.1.3 Silicon as lithium-ion battery anode ............................................................ 4 1.2 Introduction to two-dimensional materials ............................................................ 6 1.2.1 Background .................................................................................................. 6 1.2.2 Photodetector based on two-dimensional materials .................................... 7 1.3 References ............................................................................................................ 10 Chapter 2: In situ and ex situ transmission electron microscopy study of lithiation behaviors of porous silicon nanostructures ....................................................................... 13 2.1 Introduction .......................................................................................................... 13 2.2 Material synthesis ................................................................................................ 15 2.2.1 Synthesis of ball-milled silicon nanoparticles and porous silicon nanoparticles ....................................................................................................... 15 2.2.2 Synthesis of solid silicon nanowires and porous silicon nanowires .......... 16 2.3 In situ transmission electron microscopy characterization of lithiation of ball- milled silicon nanoparticles and porous silicon nanoparticles .................................. 18 2.4 In situ transmission electron microscopy characterization of lithiation of solid silicon nanowires and porous silicon nanowires ....................................................... 23 2.5 Mechanism of lithiation behaviors of porous silicon nanostructures .................. 26 2.6 Ex situ transmission electron microscopy characterization of ball-milled silicon nanoparticles and porous silicon nanoparticles .............................................. 31 2.7 Summary .............................................................................................................. 36 2.8 References ............................................................................................................ 37 Chapter 3. Silicon(lithiated)-sulfur full cells with porous silicon anode shielded by Nafion against polysulfides to achieve high capacity and energy density........................ 39 3.1 Introduction .......................................................................................................... 39 3.2 Failure mechanism of silicon(lithiated)-sulfur full cells and functionality of Nafion ........................................................................................................................ 41 3.3 Preparation of silicon anode and sulfur cathode .................................................. 49 3.4 Characterization of silicon anode and sulfur cathode .......................................... 53 3.5 Electrochemical test and discussion .................................................................... 58 3.6 Summary .............................................................................................................. 67 3.7 References ............................................................................................................ 69 Chapter 4: Synthesis of hierarchical carbon-coated ball-milled silicon and its applications in lightweight free-standing electrodes and high-voltage full cells .............. 72 4.1 Introduction .......................................................................................................... 72 4.2 Synthesis of hierarchical carbon-coated ball-milled silicon ................................ 74 vi 4.3 Electrochemical test and discussion .................................................................... 80 4.4 Lightweight and free-standing electrodes based on hierarchical carbon-coated ball-milled silicon ...................................................................................................... 89 4.5 High-voltage full cells based on hierarchical carbon-coated ball-milled silicon ................................................................................................................................... 94 4.6 Summary .............................................................................................................. 97 4.7 References ............................................................................................................ 99 Chapter 5: Air-stable room-temperature mid-infrared photodetectors based on hBN/black arsenic phosphorus/hBN heterostructures .................................................... 102 5.1 Introduction ........................................................................................................ 102 5.2 Material characterization of black arsenic phosphorus ..................................... 104 5.3 Electrical characterization of black arsenic phosphorus device ........................ 108 5.4 Photoresponse of black arsenic phosphorus phototransistor ............................. 111 5.5 Summary ............................................................................................................ 116 5.6 References .......................................................................................................... 117 Chapter 6: Conclusions and future work ........................................................................ 120 6.1 Conclusions........................................................................................................ 120 6.2 Future research ................................................................................................... 121 6.2.1 LixSi as lithium-ion battery anode ........................................................... 121 6.2.2 Two-dimensional tellurium photodetector .............................................. 122 6.3 References .......................................................................................................... 127 vii List of Tables Table 1.1 Electronic property of different layered TMDCs.. ............................................. 7 Table 5.1 The crystallographic information of b-As0.83P0.17. ......................................... 107 viii List of Figures Figure 1.1 Comparison of the different battery technologies in terms of volumetric and gravimetric energy density. .............................................................................. 2 Figure 1.2 Schematic representation and operating principles of Li batteries. ................... 4 Figure 1.3 Schematic representation of the failure mechanism of silicon nanoparticles during cycling. .................................................................................................. 5 Figure 1.4 2D materials covering a broad spectral range. (a) Electromagnetic spectrum. Applications that utilize the different spectral ranges are presented in the top portion of the panel. NIR, MIR and FIR indicate near-, mid- and far-infrared, respectively. The atomic structures of hBN, MoS2, BP and graphene are shown in the bottom of the panel, left to right. The crystalline directions (x and y) of anisotropic BP are indicated. The possible spectral ranges covered by different materials are indicated using colored polygons. (b-e) Band structures of single- layer hBN (b), MoS2 (c), BP (d) and graphene (e). .......................................... 8 Figure 2.1 Characterization of ball-milled Si nanoparticles and porous Si nanoparticles. (a,b) TEM images of ball-milled Si nanoparticles at different magnifications. (c,d) TEM images of porous Si nanoparticles at different magnifications. One domain of the porous Si nanoparticle was marked by the red dotted line in (d). ........................................................................................................................ 16 Figure 2.2 Characterization of solid silicon nanowires and porous Si nanowires. (a,b) TEM images of solid Si nanowires at different magnifications. (c,d) TEM images of porous Si nanowires at different magnifications. One domain of the porous Si nanowire was marked by the red dotted line in (d). ....................................... 18 Figure 2.3. In situ TEM observation of the lithiation process of a ball-milled Si particle. The test was carried out using a nanobattery configuration with a ball-milled Si particle attached to a Cu rod as the working electrode, Li as the reference electrode, and Li2O as the solid electrolyte. (a) Schematic of the in situ TEM nanobattery. (b) TEM image of the ball-milled Si particle before lithiation. (c- g) Time series of the lithiation of the ball-milled Si particle, which illustrates the crack nucleation and fracture of the particle. After the Li2O/Li electrode contacted the ball-milled Si, a potential of -2 V was applied to the Cu electrode with respect to Li electrode to initiate the lithiation process. (h,i) Selected area electron diffraction (SAED) patterns of the ball-milled Si particle before (h) and after lithiation (i). ..................................................................................... 20 Figure 2.4. In situ TEM observation of the lithiation process of a porous Si particle. (a) TEM image of the porous Si particle with diameter up to 1.52 μm before lithiation. (b-f) Time series of the lithiation of the porous Si particle, which ix illustrates the volume expansion of the particle without crack formation. (g,h) SAED patterns of the porous Si particle before (g) and after lithiation (h). .. 22 Figure 2.5 In situ TEM observation of the lithiation process of a porous Si particle with high magnification to show the lithiation front. (a) TEM image of the porous Si particle before lithiation. (b-d) Time series of the lithiation of the porous Si particle showing the propagation of the lithiation front, which is indicated by the red dotted line. .......................................................................................... 22 Figure 2.6 In situ TEM observation of the lithiation process of a typical Si nanowire. (a) TEM image of the Si nanowire before lithiation. (b-f) Time series of the lithiation of the Si nanowire. (g,h) SAED patterns of the Si nanowire before (g) and after lithiation (h). .................................................................................... 24 Figure 2.7 In situ TEM observation of the lithiation process of a porous Si nanowire bundle. (a) TEM image of the porous Si nanowire bundle before lithiation. (b-g) Time series of the lithiation of the porous Si nanowire bundle. The single nanowire beside the bundle provides lithium diffusion path. (h,i) SAED patterns of the Si nanowire bundle before (h) and after lithiation (i). .................................... 25 Figure 2.8 Schematic diagram illustrating the lithiation manners of ball-milled Si and porous Si nanoparticles. (a-d) Schematic diagram showing the surface-to- center lithiation manner of ball-milled Si particle. (e-h) Schematic diagram showing the end-to-end lithiation manner of porous Si particle. ................... 27 Figure 2.9 First-principle molecular dynamic simulation to study the structure stability of a nanosized c-Li15Si4 particle. (a) The modeled structure of c-Li15Si4. (b-e) Atomic structure and morphology of the Li15Si4 particle at different simulation stages. (f) Si-Si radial pair distribution function at different stages of the simulated process. The appearance and increasing intensity of the peak at 2.5 Å indicate the intermixing of Si and Li to form an amorphous phase. .......... 29 Figure 2.10 Classical molecular dynamic simulation to study the structure stability of c- Li15Si4 particles with different sizes. Atomic structures and morphologies of the Li15Si4 particles with diameter of 6 nm (a), 8 nm (b), 10 nm (c), and 12 nm (d) after 400 fs simulation. The insets in (a-d) are the enlarged images showing the surface and core of the corresponding particles. ...................................... 31 Figure 2.11 Ex situ TEM characterization of ball-milled Si and porous Si after different charge-discharge cycles and comparison of their cycling performances. The Si electrodes were cycled in Li-Si cells in the voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g and then disassembled at the delithiated state before TEM observation. (a-d) TEM images of ball-milled Si before cycling (a), after cycling for 1 cycle (b), 10 cycles (c), and 50 cycles (d). (e-h) TEM images of porous Si before cycling (e), after cycling for 1 cycle (f), 10 cycles (g), and 50 cycles (h). The insets in (a-h) are the corresponding SAED patterns. (i-l) Pore size distributions of porous Si before cycling (i) and the x comparison of ball-milled Si and porous Si after cycling for 1 cycle (j), 10 cycles (k), and 50 cycles (l). (m) Cycling performances of Li-Si cells using ball- milled Si and porous Si as working electrode, respectively. The galvanostatic charge-discharge test was carried out in the voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g. ....................................................... 35 Figure 3.1 Cyclic performance of Si-S full cell using lithiated bare porous Si as anode and S-C-G as cathode in the voltage window of 1.2-2.7 V (vs. Li/Li + ) at current density of 0.1 C. The electrolyte is LITFSI electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO3). The specific capacity of the Si-S full cell is calculated based on the mass of sulfur. ....................... 43 Figure 3.2 Functionality of Nafion in shielding Si from reaction with polysulfides. (a) Schematic diagram showing the reaction of a bare Si wafer with polysulfides electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 1 M Li2S4). After the reaction, a thin layer of Li-Si-S compound is formed on the surface of the wafer. (b) SEM image of Si wafer after being immersed in polysulfides electrolyte for 12 hours and washed by DI-H2O. The rough surface indicates the reaction between Si and polysulfides. (c) Schematic diagram showing a thin layer of Nafion coating on Si wafer can shield Si from reaction with polysulfides. (d) SEM image of Nafion-coated Si wafer after being immersed in polysulfides electrolyte for 12 hours and washed by DI-H2O, which shows clean and smooth surface. ......................................................... 45 Figure 3.3 EDS and XPS analyses of the reaction product of Si wafers and polysulfides electrolyte. (a-c) EDS spectra of a Si wafer after reaction with polysulfides electrolyte (a), a bare Si wafer (b), and polysulfides electrolyte (c). Insets show the SEM images of corresponding regions for collecting the EDS spectra. (d-f) Li 2s (d), S 2p (e), and Si 2p (f) XPS spectra of a Si wafer after reaction with polysulfides electrolyte. (g,h) Li 2s (g) and S 2p (h) XPS spectra of polysulfides electrolyte. (i) Si 2p XPS spectrum of a bare Si wafer. Hollow circles: experimental data; gray lines: background; black lines: overall fit; colored lines: fitted individual components. ......................................................................... 47 Figure 3.4 Characterization of porous Si and Nafion-coated porous Si (Si-N). (a,b) TEM images of a typical porous Si particle at different magnifications. The pores are uniformly distributed throughout the particle, with pore size of 10-15 nm. (c,d) TEM images of Si-N at different magnifications. The porous structure is not as clear as it shows in (a), mainly due to the filling of Nafion into the pores. A thin layer of Nafion can be found in (d) as indicated by the dotted line. (e) Another TEM image of Si-N particle. (f-h) Energy filtered TEM images of (e) to map out the distribution of Si (f), Nafion (g), and their superposition (h). ............ 55 Figure 3.5 FTIR spectra of Si, Si-N, and Nafion. ............................................................. 55 Figure 3.6 (a) SEM and (b) TEM images of S-C-G. (c) Cyclic performance and (d) Charge-discharge curves of S-C-G. ................................................................ 56 xi Figure 3.7 (a) SEM and (b) TEM images of Si-C-N-G. (c) Cyclic performance and (d) Charge-discharge curves of Si-C-N-G. .......................................................... 57 Figure 3.8 Electrochemical performance of Li-Si half cells and Si-S full cells to demonstrate the functionality of Nafion. (a) Comparison of Li-Si half cells using bare porous Si particles and Nafion-coated porous Si particles (Si-N) as working electrode, respectively. The galvanostatic charge-discharge test was conducted in the voltage window of 0.01-2 V at a current density of 400 mA/g and the electrolyte is polysulfides electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 1 M Li2S4). (b) Comparison of Si-S full cells using lithiated carbon-coated porous Si (Si-C) and lithiated carbon-coated porous Si with Nafion coating (Si-C-N) as anode, respectively. The Si-based anodes were first cycled in Li-Si half cells and then disassembled at lithiated state before coupling with S-C-G cathodes to assemble full cells. The galvanostatic charge-discharge test of full cells was conducted in the voltage window of 1.2-2.7 V at a current density of 0.1 C (1 C=1600 mA/g based on the mass of sulfur) and the electrolyte is LITFSI electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO3). The specific capacity is calculated based on the mass of sulfur. Charge capacity: solid circles and solid squares; discharge capacity: hollow circles and hollow squares. ................................. 60 Figure 3.9 Electrochemical performance of Si-S full cell with LITFSI electrolyte using lithiated Si-C-N-G as anode and S-C-G as cathode. The Si-S full cell is cycled in the voltage window of 1.2-2.7 V and the specific capacity is calculated based on the mass of sulfur. (a) Cyclic performance of Si-S full cell at a current density of 0.1 C. (b) Charge-discharge curves of Si-S full cell at different cycles. (c) Cyclic voltammetry curves of Si-S full cell at different cycles. The test is conducted at the scan rate of 0.2 mV/s in the voltage window of 1.2-2.7 V. (d) Cyclic performance of Si-S full cell at different current rates. ...................... 62 Figure 3.10 Nyquist plots of Si-C-G (red curve) and Si-C-N-G (blue curve) in Li-Si half-cell configuration before cycling (a) and after 5 cycles (b). ................... 64 Figure 3.11 Evaluation of cyclic performance of Si-S full cell with different S:Si mass loading ratios. (a) Cyclic performance of Si-S full cell at S:Si mass loading ratio of 0.33, 1.43, 2.22, and 4.24. The galvanostatic charge-discharge test was conducted in the voltage window of 1.2-2.7 V at a current density of 0.1 C and the specific discharge capacity is calculated based on the mass of sulfur. (b) Calculated specific discharge capacity of Si-S full cell after 100 cycles at different S:Si mass ratios, based on the mass of S only (red curve) and mass of S+Si (blue curve). For comparison, the theoretical capacity of graphite-LiCoO2 full cell at different mass ratios is demonstrated as black curve. ................... 67 Figure 4.1 Morphological and elemental characterization of HCC-M-Si and HCC-M-Si(2). (a) SEM and (b) TEM image of M-Si. (c) HRTEM image of M-Si. The inset is the corresponding SAED pattern. (d, e) TEM images of HCC-M-Si. (f) TEM xii image of HCC-M-Si(2). (g) STEM image of HCC-M-Si particle. (h) EELS mapping of elements C and Si in the HCC-M-Si particle. (i) EELS spectrum of the HCC-M-Si particle. .................................................................................. 77 Figure 4.2. Morphological characterization of P-Si and W-Si. (a) TEM image of P-Si. (b) HRTEM image of P-Si. The inset is the corresponding SAED pattern. (c) TEM image of W-Si. (d) HRTEM image of W-Si. The inset is the corresponding SAED pattern. ................................................................................................ 78 Figure 4.3. XRD patterns of metallurgical Si, polycrystalline Si, and Si wafer before (a) and after (b) ball-milling. ............................................................................... 79 Figure 4.4 Electrochemical performance of HCC-M-Si, HCC-P-Si, and HCC-W-Si. (a) Galvanostatic charge-discharge profiles of HCC-M-Si. (b) Cycling performance comparison of HCC-M-Si and HCC-M-Si(2) at a current density of 0.4 A/g. (c) Rate capability comparison of HCC-M-Si and M-Si. (d) Long- term cycling performance of HCC-M-Si, HCC-P-Si, and HCC-W-Si. The current density was 0.4 A/g for the first 10 cycles, and 2 A/g for the following cycles. The cycling performance of HCC-M-Si(2) and M-Si are provided for comparison with HCC-M-Si. ......................................................................... 82 Figure 4.5 Cyclic voltammetry curves of HCC-M-Si. ...................................................... 83 Figure 4.6. Nyquist plots of HCC-M-Si and M-Si electrodes before cycling (a) and after 20 cycles (b). ....................................................................................................... 84 Figure 4.7 Evaluation of Si anodes in terms of production cost, capacity, cycle number, and current density. The color scheme represents the cycle number; the symbol represents the current density; the number next to each symbol refers to the reference index. .............................................................................................. 86 Figure 4.8 Post-cycling analysis of HCC-M-Si. (a) Schematic illustration showing the morphology evolution of M-Si and HCC-M-Si after cycling. (b-d) TEM images of M-Si after 1 cycle (b), 10 cycles (c), and 50 cycles (d). (e) STEM image of HCC-M-Si particle after 50 cycles. (f) EELS mapping of elements C and Si in the HCC-M-Si particle. (g) EELS spectrum of the HCC-M-Si particle......... 88 Figure 4.9. Low-magnification STEM image of HCC-M-Si after 50 cycles. .................. 89 Figure 4.10 Characterization and electrochemical performance of CNFs. (a) Low- magnification TEM image of CNFs. (b) TEM image of the open end of one CNF. (c) HRTEM of the area which is marked by red square in (b). It shows the graphitic nature of the CNF. (d) SEM image of the surface of HCC-M- Si/CNF free-standing electrode. (e) Electrochemical performance of free- standing CNF electrode at a current density of 0.4 A/g. ................................ 91 xiii Figure 4.11 Characterization and electrochemical performance of lightweight and free- standing HCC-M-Si/CNF electrode. (a) Schematic illustration of the fabrication process of free-standing HCC-M-Si/CNF electrode. (b) Photograph of the free-standing HCC-M-Si/CNF electrode and the assembled coin cell. (c) Cross-sectional SEM image of HCC-M-Si/CNF electrode and corresponding EDS mapping of elements C and Si. (d) SEM image of the area which is marked by red square in (c). It shows the inside mixture of HCC-M-Si/CNF electrode where HCC-M-Si particles are dispersed in CNF network. The inset in (d) shows the high-resolution SEM image of the area which is marked by green square. (e) Cycling performance comparison of free-standing HCC-M-Si/CNF and M-Si/CNF electrodes. .............................................................................. 92 Figure 4.12 Electrochemical performance of full cells based on HCC-M-Si. (a) Schematic illustration of the full cell using HCC-M-Si as anode and LiCoO2 or LiNi0.5Mn1.5O4 as cathode. (b) Galvanostatic charge-discharge profiles of HCC-M-Si/LiCoO2 full cell. The inset shows the HCC-M-Si/LiCoO2 full cell can power blue and white LEDs. (c) Galvanostatic charge-discharge profiles of HCC-M-Si/LiNi0.5Mn1.5O4 full cell. The inset shows the HCC-M- Si/LiNi0.5Mn1.5O4 full cell can power blue and white LEDs. (d) Cycling performance comparison of HCC-M-Si/LiCoO2 and HCC-M-Si(2)/LiCoO2 full cells. (e) Cycling performance of HCC-M-Si/LiNi0.5Mn1.5O4 full cell. All the full cells investigated in the figure are cycled at a current density of 0.4 A/g based on the mass of HCC-M-Si and the specific capacity of the full cell is based on the mass of HCC-M-Si. ................................................................... 97 Figure 5.1 EDS spectrum of b-AsxP1-x. The elemental composition is summarized in the inset table. ..................................................................................................... 106 Figure 5.2 Crystal structure and infrared extinction characterizations. (a) X-ray diffraction pattern of the as-synthesized crystal with Miller indices labeled for peaks. The broad peak at 18° is from the Kapton tape, which is used to encapsulate the b- As0.83P0.17. (b) The orthorhombic puckered honeycomb crystal structure of b- As0.83P0.17 alloy. The lattice parameters, extracted from the X-ray diffraction profile, are a = 3.561(3) Å, b = 10.803(9) Å, c = 4.493(4) Å. (c) Polarization- resolved IR extinction spectra of b-As0.83P0.17 alloy. The optical image of the investigated flake is shown in the inset. ....................................................... 106 Figure 5.3 The thickness profile of the exfoliated b-As0.83P0.17 flake used for extinction spectra characterization acquired using atomic force microscope. .............. 108 Figure 5.4 Device structure and electrical characterizations. (a) The cross-sectional schematic of the as-fabricated hBN/b-As0.83P0.17/hBN heterostructure photodetector. (b) Left: the cross-sectional view of the device by transmission electron microscope (TEM). Right: the elemental analysis mapping by electron energy loss spectroscopy (EELS). In the EELS mapping, red, blue, and green color denote O, N, and As elements respectively. Only part of the silicon oxide layer is shown. The encapsulated b-As0.83P0.17 layer is 37 nm, free from xiv oxidation even several months after its fabrication. (c) Transfer characteristic of the as-fabricated phototransistor. The optical image of the as-fabricated device is shown in the inset. The gate bias was swept in both directions, with numbers denoting the sweeping sequence. The charge neutral point is at Vbg=- 3 V the hysteresis of the transfer curve is small. (d) Output characteristics of the b-As0.83P0.17 phototransistor. The output curves were measured with different Vbg , and VN represents the gate bias, at which the charge neutral point of the device is achieved (VN = -3 V in this case). ....................................... 110 Figure 5.5 Photoresponse of hBN encapsulated b-As0.83P0.17 phototransistor. (a) Polarization-resolved photocurrents of the b-As0.83P0.17 phototransistor for 3.4, 5.0, and 7.7 µm light excitation at Vds= 0.3 V when the device is biased at charge neutral point. (b) Vbg dependence of the photocurrent in the b- As0.83P0.17 phototransistor at Vds= 0.3 V. (c) The photoresponse as a function of source-drain bias Vds . In this measurement, the device works at charge neutral point. Transparent colored lines are the linear fitting in the range of | Vds | < 0.4 V. The photocurrent in (b) and (c) was measured when the polarization of incident light was aligned with the armchair direction. ....... 112 Figure 5.6 Power and frequency dependence of the photocurrent and noise characteristics. (a) Photocurrent as a function of incident power at 3.4 µm when the device works at charge neutral point and Vds = 0.3 V. The laser polarization is aligned with the armchair direction. (b) Photo-response as a function of incident light intensity modulation frequency from 1 to 10 kHz, showing no sign of response roll-off. (c) Noise equivalent power at Vds = 1 V for 3.4, 5.0, and 7.7 µm incident light when the device works at charge neutral point and the light polarization is aligned with the armchair direction. ..................................... 115 Figure 6.1 Schematic representation of the fabrication process of LixSi. ...................... 122 Figure 6.2 (a) TEM image of the synthesized 2D Te. (b) HRTEM image of Te. .......... 123 Figure 6.3 Device characteristics of Te FETs. (a) Typical output (Ids-Vds) and (b) transfer characteristics (Ids-Vg) of the Te FETs. ........................................................ 123 Figure 6.4 Photocurrent of Te filed-effect transistor under 520 nm laser. (a) Polarization- dependent photocurrent. (b) Power-dependent photocurrent. (c) Power dependence of responsivity and EQE. (d) Photocurrent amplitude versus modulation frequency under various incident power. .................................. 124 Figure 6.5 Photocurrent of Te filed-effect transistor under 1.55 µm laser. (a) Polarization- dependent photocurrent versus Vg. (b) Polarization-dependent photocurrent summarized in polar coordinate. (c) Power-dependent photocurrent. (d) Frequency-dependent photocurrent. ............................................................. 125 Figure 6.6 Photocurrent of Te filed-effect transistor under 3.39 µm laser. (a) Polarization- dependent photocurrent versus Vg. (b) Polarization-dependent photocurrent xv summarized in polar coordinate. (c) Power-dependent photocurrent. (d) Frequency-dependent photocurrent. ............................................................. 126 xvi Abstract Nanomaterials have been receiving great attention in the past decade due to their wide applications in numerous areas. Among all the applications, energy storage devices and electronic/optoelectronic devices are two of the most appealing topics. During my Ph.D. study, I conduct research on the two major topics: 1. Study of novel silicon (Si) nanostructures and their application as lithium-ion battery anode materials; 2. Electronic and optoelectronic device study of novel two-dimensional (2D) materials. This dissertation starts from a brief introduction to lithium-ion battery and 2D electronics/optoelectronics in Chapter 1. Chapter 2 to 5 report the work we have done in lithium-ion battery and 2D electronics/optoelectronics. Chapter 6 discusses the future work in both Si anode direction and 2D material photodetector direction. Chapter 2 reports the study of the lithiation behaviors of both porous Si nanoparticles and porous Si nanowires by in situ and ex situ transmission electron microscopy (TEM) and compare them with solid Si nanoparticles and nanowires. The in situ TEM observation reveals that the critical fracture diameter of porous Si particles reaches up to 1.52 μm, which is much larger than the previously reported 150 nm for crystalline Si nanoparticles and 870 nm for amorphous Si nanoparticles. After full lithiation, solid Si nanoparticles and nanowires transform to crystalline Li15Si4 phase while porous Si nanoparticles and nanowires transform to amorphous LixSi phase, which is due to the effect of domain size on the stability of Li15Si4 as revealed by the first-principle molecular dynamic simulation. Ex situ TEM characterization is conducted to further investigate the structural evolution of xvii porous and solid Si nanoparticles during the cycling process, which confirms that the porous Si nanoparticles exhibit better capability to suppress pore evolution than solid Si nanoparticles. The investigation of structural evolution and phase transition of porous Si nanoparticles and nanowires during the lithiation process reveal that they are more desirable as lithium-ion battery anode materials than solid Si nanoparticles and nanowires. Chapter 3 reports our study on Si-S full cells. The challenges in the Si-S full cell integration is discussed, and a failure mechanism of Si-S full cell is proposed, which is due to the spontaneous reaction between Si (and lithiated Si) and polysulfides. On this basis, we report one prototype of Si-S full cells using lithiated Nafion-coated porous Si as anode and sulfur as cathode, and our study on the functionality of Nafion in shielding Si from reaction with polysulfides. With optimized mass ratio between sulfur and silicon, the full cell yields specific capacity of 330 mAh/g and energy density of 590 Wh/kg after 100 cycles based on the total mass of sulfur and silicon. The achieved energy density is more than 2 times higher than commercially available lithium-ion batteries. The investigation of issues in Si-S full cell research and the proposed full cell prototype will shed light on the development of next-generation lithium-ion batteries. Chapter 4 reports the synthesis of low-cost hierarchical carbon-coated (HCC) Si using ball- milled Si as the starting material. The obtained particles prepared from different Si sources all show excellent cycling performance of over 1000 mAh/g after 1000 cycles. Interestingly, we observed in situ formation of porous Si and it is well confined in the carbon shell based on post-cycling characterization of the hierarchical carbon-coated metallurgical Si (HCC- xviii M-Si) particles. In addition, lightweight and free-standing electrodes consisting of the HCC-M-Si particles and carbon nanofibers were fabricated, which achieved 1015 mAh/g after 100 cycles based on the total mass of the electrodes. Compared with conventional electrodes, the lightweight and free-standing electrodes significantly improve the energy density by 745%. Furthermore, LiCoO2 and LiNi0.5Mn1.5O4 cathodes were used to pair up with the HCC-M-Si anode to fabricate full cells. With LiNi0.5Mn1.5O4 as cathode, energy density up to 547 Wh/kg was achieved by the high-voltage full cell. After 100 cycles, the full cell with LiNi0.5Mn1.5O4 cathode delivers 46% more energy density than that of the full cell with LiCoO2 cathode. The systematic investigation on low-cost Si anodes together with their applications in lightweight free-standing electrodes and high-voltage full cells will shed light on development of high-energy Si-based lithium-ion batteries for real applications. Chapter 5 studies the novel 2D material black arsenic phosphorus (b-AsxP1-x) formed by introducing arsenic into black phosphorus, which can significantly extend the operational wavelength range of photonic devices. The as-fabricated b-AsxP1-x photodetector sandwiched within hexagonal boron nitride (hBN) shows peak extrinsic responsivity of 190, 16, and 1.2 mA/W at 3.4, 5.0, and 7.7 m at room temperature, respectively. Moreover, the intrinsic photoconductive effect dominates the photocurrent generation mechanism due to the preservation of pristine properties of b-AsxP1-x by complete hBN encapsulation and these b-AsxP1-x photodetectors exhibit negligible transport hysteresis. The broad and large photo-responsivity within mid-infrared resulting from the intrinsic photoconduction, together with the excellent long-term air stability, makes black arsenic xix phosphorus a promising alternative material for mid-infrared applications, such as free space communication, infrared imaging, and biomedical sensing. Chapter 6 proposes future research topics. To avoid the capacity loss of Si-based lithium- ion batteries in the first lithiation process, synthesis of LixSi is highly desired. We propose a one-step large-scale synthesis method of LixSi, which produces LixSi and electrode slurry in a single step. Another research topic is Te photodetector. The application of black phosphorus and black arsenic phosphorus photodetector is limited because they are air- sensitive. As an air-stable two-dimensional material, Te also possesses similar electronic/optoelectronic properties as black phosphorus, which makes it promising candidate as medium-wavelength infrared photodetector. A systematic study of Te photodetector has been conducted and the underlying photocurrent generation mechanism is under investigation. 1 Chapter 1: Introduction 1.1 Introduction to lithium-ion battery 1.1.1 Background Nowadays the world faces two major energy challenges: 1. Shifting electricity production from burning fossil fuels to sustainable energy sources; 2. Moving ground transportation towards electrical propulsion, namely, using electric vehicles (EVs) instead of cars driven by internal combustion engines 1 . The sources of sustainable energy fluctuate during the day and night. For this reason, the use of sustainable energy for electricity production requires the availability of suitable energy storage devices, namely, batteries. Although impressive progress has been made recently in the development of technology for harvesting sustainable energy, e.g., better wind turbines 2 , photovoltaic cells 3-4 , and photothermal receivers 5 , the development of energy storage devices is still lagging far behind. Hence, the development of batteries that can store sustainable energy with long term stability, very prolonged cycle life and meeting environmental constraints is an important challenge for modern electrochemistry. During past several decades, various kinds of batteries have been developed, such as lead-acid, nickel-cadmium, nickel-metal-hydride, zinc-bromine, zinc-air, vanadium redox, sodium-sulfur, sodium-nickel-chloride, lithium-ion, etc 6 . Among all of them, lithium-ion battery is believed to be the most feasible alternatives from the viewpoint of current technical maturity and economic consideration. Figure 1.1 provides a comparison of various energy storage devices in terms of gravimetric and volumetric energy densities, which are two key parameters for energy storage devices, in addition to cycle life, safety and cost 7 . As shown in Figure 1.1, lithium-ion batteries 2 have relatively high energy densities due to the fact that lithium is the lightest (0.53 g cm -3 ) and the most electropositive (-3.04 V vs. standard hydrogen electrode) among all the metals 7 . Compared to other batteries, the advantages of lithium-ion batteries also include long cycle life, low self-discharge rate, rapid charge capability, high coulombic and energy efficiency, and no memory effect. For these reasons, lithium-ion batteries have been regarded as one of the most promising energy storage devices and development of advanced lithium-ion batteries has attracted significant attention. Figure 1.1 Comparison of the different battery technologies in terms of volumetric and gravimetric energy density. 1.1.2 Operating principle of lithium-ion battery A battery is composed of several electrochemical cells that are connected in series and/or in parallel to provide the required voltage and capacity. Each cell consists of a positive and a negative electrode separated by an electrolyte solution containing dissociated salts, which enable ion transfer between the two electrodes. Once these electrodes are connected externally, the chemical reactions proceed in tandem at both electrodes, thereby liberating 3 electrons and enabling the current to be tapped by the user. A typical commercial lithium- ion battery is schematically shown in Figure 1.2. The lithium-ion battery comprises of a layered lithium metal oxide cathode material and graphite anode material. They are separated by a porous membrane separator, which is soaked in a lithium-ion conducting electrolyte. The battery is also named as a “rocking-chair” battery as the lithium ions “rock” back and forth between the cathode and the anode when the battery is charged and discharged 8 . During charging process, an external voltage source pulls electrons from the cathode through an external circuit to the anode and causes lithium-ions to move from the cathode to the anode by transport through the liquid electrolyte. The discharge is a reversed process. Lithium-ions move from the anode to the cathode through the electrolyte while electrons flow through the external circuit from the anode to the cathode and produce power 9 . It can be clearly seen that lithium-ion batteries operate based on the reversible shuttling of lithium-ions between the cathode materials and the anode materials. As a consequence, the inherent properties of electrode materials will strongly restrict the performance of batteries. 4 Figure 1.2 Schematic representation and operating principles of Li batteries. 1.1.3 Silicon as lithium-ion battery anode Compared with carbonaceous anodes (372 mAh/g for LiC6) used in commercial lithium- ion batteries, silicon (Si) has a large theoretical gravimetric capacity of ~4200 mAh/g and volumetric capacity of ~8500 mAh/cm, and therefore has been considered as one of the most promising anode materials for the next-generation lithium-ion batteries 10-11 . However, Si experiences a dramatic volume change (>300%) during the lithium alloying/dealloying processes, and for crystalline Si (c-Si) this large volume expansion is accompanied with dramatic anisotropic expansion 12-14 . In addition, unstable surface electrolyte interphase (SEI) films will form during the lithiation and delithiation processes. These not only cause severe pulverization of the material but also induces electrical disconnection of the active material from the current collector as shown in Figure 1.3, resulting in performance degradation of the battery if Si is used as the anode. To minimize the extent of volume change, tremendous efforts have been made on the synthesis of novel nanostructured Si 5 materials, such as nanowires 15-16 , nanotubes 17-19 , hollow spheres, and core-shell structures 20-21 . Recently, three-dimensional porous structured Si has attracted significant attention. The pre-formed nanopores in the Si can provide a large space to accommodate the volume expansion, and therefore help to maintain the structure integrity when lithium alloys with Si. Moreover, this three-dimensional porous structure provides large surface area of the material to be accessible to the electrolyte and thus a short diffusion length for lithium ions to transport from electrolyte to Si, which facilitates the lithium alloying/dealloying processes at high current rates 22-27 . Figure 1.3 Schematic representation of the failure mechanism of silicon nanoparticles during cycling. In the following chapters, we will report the recent progress we have achieved in silicon anode and future research plan will also be discussed. Specifically, in Chapter 2, we will talk about in situ and ex situ transmission electron microscopy study of lithiation behaviors of porous silicon nanostructures. In Chapter 3, full cells based on porous Si anode and sulfur cathode are studied. The failure mechanism of the full cell is investigated and a prototype of the full cell is proposed, which achieves high capacity and energy density. In Chapter 4, we report the low-cost synthesis of hierarchical carbon-coated ball-milled 6 silicon and study its applications in lightweight free-standing electrodes and high-voltage full cells. In Chapter 6, future research plan is provided. 1.2 Introduction to two-dimensional materials 1.2.1 Background Layered materials have been studied scientifically for more than 150 years 28 . However, only recently researchers realize that, when layered material is thinned to its physical limits, it exhibits novel properties different from its bulk counterpart 29 . This finding realizes the true potential of these systems for advanced technological applications and these materials are referred to as two-dimensional (2D) materials. The most well-known 2D material is graphene because of its exceptional electronic, optoelectronic, electrochemical and biomedical applications. Beyond graphene, there is a very wide range of 2D electronic materials that range from insulators to semiconductors to metals and even to superconductors. 2D materials research initially focused on graphene since the work reported by Novoselov and Geim 30 . Later on, a family of novel semiconducting 2D materials, known as transition-metal dichalcogenides (TMDCs) attract researchers’ attention. The generalized formula of TMDCs is MX2, where M is a transition metal of groups 4-10 and X is a chalcogen 31 . The properties of bulk TMDCs are diverse, which ranges from insulators such as HfS2, semiconductors such as MoS2 and WS2, semimetals such as WTe2 and TiSe2, to true metals such as NbS2 and VSe2. A few bulk TMDCs such as NbSe2 and TaS2 exhibit low-temperature phenomena including superconductivity, charge density and Mott transition. The electronic property of different layered TMDCs are summarized in Table 1.1. The chemistry of MX2 compounds thus offers opportunities 7 for going beyond graphene and opening up new fundamental and technological pathways for inorganic 2D materials. Table 1.1 Electronic property of different layered TMDCs 31 . 1.2.2 Photodetector based on two-dimensional materials Compared with traditional 3D photonic materials such as gallium arsenide (GaAs) and Si, 2D materials exhibit many exceptional properties for photodetectors: 1. quantum confinement in the direction perpendicular to the 2D plane leads to novel electronic and optoelectronic properties that are distinctively different from their bulk materials 32-35 ; 2. their surfaces are naturally passivated without any dangling bonds, which makes it easy to integrate 2D materials with photonic structures 36-37 . It is also possible to construct vertical 8 heterostructures using different 2D materials without the conventional lattice mismatch issue 38 ; 3. despite being atomically thin, many 2D materials interact strongly with light 39 ; 4. 2D materials cover a very wide range of the electromagnetic spectrum because of their diverse electronic properties as shown in Figure 1.4 40 . Figure 1.4 2D materials covering a broad spectral range. (a) Electromagnetic spectrum. Applications that utilize the different spectral ranges are presented in the top portion of the panel. NIR, MIR and FIR indicate near-, mid- and far-infrared, respectively. The atomic structures of hBN, MoS2, BP and graphene are shown in the bottom of the panel, left to right. The crystalline directions (x and y) of anisotropic BP are indicated. The possible spectral ranges covered by different materials are indicated using colored polygons. (b-e) Band structures of single-layer hBN (b), MoS2 (c), BP (d) and graphene (e). In the following Chapter 5, we will report the study of a novel 2D material black arsenic phosphorus as medium-wavelength infrared photodetector and even long-wavelength 9 infrared photodetector. In addition, future research plan and preliminary results of a new air-stable 2D material tellurium and its application as wide-band photodetector is also provided. 10 1.3 References 1. Etacheri, V.; Marom, R.; Elazari, R.; Salitra, G.; Aurbach, D. Challenges in the development of advanced Li-ion batteries: a review. Energy Environ. Sci. 4, 3243-3262 (2011). 2. Lu, L.; Yang, H. X.; Burnett, J. Investigation on wind power potential on Hong Kong islands-an analysis of wind power and wind turbine characteristics. Renew. Energy 27, 1- 12 (2002). 3. Tachan, Z.; Ruhle, S.; Zaban, A. Dye-sensitized solar tubes: A new solar cell design for efficient current collection and improved cell sealing. Sol. Energy Mater. Sol. Cells 94, 317-322 (2010). 4. Nazeeruddin, M. K.; Baranoff, E.; Gratzel, M. Dye-sensitized solar cells: A brief overview. Sol. Energy 85, 1172-1178 (2011). 5. Chow, T. T. A review on photovoltaic/thermal hybrid solar technology. Appl. Energy 87, 365-379 (2010). 6. Choi, N. S.; Yao, Y.; Cui, Y.; Cho, J. One dimensional Si/Sn-based nanowires and nanotubes for lithium-ion energy storage materials. J. Mater. Chem. 21, 9825-9840 (2011). 7. Tarascon, J. M.; Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 414, 359-367 (2001). 8. Xu, G. L.; Wang, Q.; Fang, J. C.; Xu, Y. F.; Li, J. T.; Huang, L.; Sun, S. G. Tuning the structure and property of nanostructured cathode materials of lithium ion and lithium sulfur batteries. J. Mater. Chem. A 2, 19941-19962 (2014). 9. Bruce, P. G.; Scrosati, B.; Tarascon, J. M. Nanomaterials for rechargeable lithium batteries. Angew. Chem. Int. Ed. 47, 2930-2946 (2008). 10. Boukamp, B. A.; Lesh, G. C.; Huggins, R. A. All-Solid Lithium Electrodes with Mixed- Conductor Matrix. J. Electrochem. Soc. 128, 725-729 (1981). 11. Larcher, D.; Beattie, S.; Morcrette, M.; Edstroem, K.; Jumas, J. C.; Tarascon, J. M. Recent findings and prospects in the field of pure metals as negative electrodes for Li-ion batteries. J. Mater. Chem. 17, 3759-3772 (2007). 12. 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). 13. Liu, X. H.; Zheng, H.; Zhong, L.; Huan, S.; Karki, K.; Zhang, L. Q.; Liu, Y.; Kushima, A.; Liang, W. T.; Wang, J. W.; Cho, J. H.; Epstein, E.; Dayeh, S. A.; Picraux, S. T.; Zhu, T.; Li, J.; Sullivan, J. P.; Cumings, J.; Wang, C. S.; Mao, S. X., et al. Anisotropic Swelling and Fracture of Silicon Nanowires during Lithiation. Nano Lett. 11, 3312-3318 (2011). 14. Liu, X. H.; Zhong, L.; Huang, S.; Mao, S. X.; Zhu, T.; Huang, J. Y. Size-Dependent Fracture of Silicon Nanoparticles During Lithiation. ACS Nano 6, 1522-1531 (2012). 15. Chan, C. K.; Peng, H. L.; Liu, G.; McIlwrath, K.; Zhang, X. F.; Huggins, R. A.; Cui, Y. High-performance lithium battery anodes using silicon nanowires. Nat. Nanotechnol. 3, 31-35 (2008). 16. Cui, L. F.; Ruffo, R.; Chan, C. K.; Peng, H. L.; Cui, Y. Crystalline-Amorphous Core- Shell Silicon Nanowires for High Capacity and High Current Battery Electrodes. Nano Lett. 9, 491-495 (2009). 17. Park, M. H.; Kim, M. G.; Joo, J.; Kim, K.; Kim, J.; Ahn, S.; Cui, Y.; Cho, J. Silicon Nanotube Battery Anodes. Nano Lett. 9, 3844-3847 (2009). 11 18. Song, T.; Xia, J. L.; Lee, J. H.; Lee, D. H.; Kwon, M. S.; Choi, J. M.; Wu, J.; Doo, S. K.; Chang, H.; Il Park, W.; Zang, D. S.; Kim, H.; Huang, Y. G.; Hwang, K. C.; Rogers, J. A.; Paik, U. Arrays of Sealed Silicon Nanotubes As Anodes for Lithium Ion Batteries. Nano Lett. 10, 1710-1716 (2010). 19. Hertzberg, B.; Alexeev, A.; Yushin, G. Deformations in Si-Li Anodes Upon Electrochemical Alloying in Nano-Confined Space. J. Am. Chem. Soc. 132, 8548-8549 (2010). 20. Yao, Y.; McDowell, M. T.; Ryu, I.; Wu, H.; Liu, N. A.; Hu, L. B.; Nix, W. D.; Cui, Y. Interconnected Silicon Hollow Nanospheres for Lithium-Ion Battery Anodes with Long Cycle Life. Nano Lett. 11, 2949-2954 (2011). 21. Liu, N.; Wu, H.; McDowell, M. T.; Yao, Y.; Wang, C. M.; Cui, Y. A Yolk-Shell Design for Stabilized and Scalable Li-Ion Battery Alloy Anodes. Nano Lett. 12, 3315-3321 (2012). 22. 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 (2012). 23. Ge, M.; Fang, X.; Rong, J.; Zhou, C. Review of porous silicon preparation and its application for lithium-ion battery anodes. Nanotechnology 24, (2013). 24. Li, X. L.; Gu, M.; Hu, S. Y.; Kennard, R.; Yan, P. F.; Chen, X. L.; Wang, C. M.; Sailor, M. J.; Zhang, J. G.; Liu, J. Mesoporous silicon sponge as an anti-pulverization structure for high-performance lithium-ion battery anodes. Nat. Commun. 5, (2014). 25. Ge, M. Y.; Lu, Y. H.; Ercius, P.; Rong, J. P.; Fang, X.; Mecklenburg, M.; Zhou, C. W. Large-Scale Fabrication, 3D Tomography, and Lithium-Ion Battery Application of Porous Silicon. Nano Lett. 14, 261-268 (2014). 26. Lu, Z. D.; Liu, N.; Lee, H. W.; Zhao, J.; Li, W. Y.; Li, Y. Z.; Cui, Y. Nonfilling Carbon Coating of Porous Silicon Micrometer-Sized Particles for High-Performance Lithium Battery Anodes. ACS Nano 9, 2540-2547 (2015). 27. Shen, C. F.; Ge, M. Y.; Zhang, A. Y.; Fang, X.; Liu, Y. H.; Rong, J. P.; Zhou, C. W. Silicon(lithiated)-sulfur full cells with porous silicon anode shielded by Nafion against polysulfides to achieve high capacity and energy density. Nano Energy 19, 68-77 (2016). 28. Golden, J.; McMillan, M.; Downs, R. T.; Hystad, G.; Goldstein, I.; Stein, H. J.; Zimmerman, A.; Sverjensky, D. A.; Armstrong, J. T.; Hazen, R. M. Rhenium variations in molybdenite (MoS2): Evidence for progressive subsurface oxidation. Earth Planet. Sci. Lett. 366, 1-5 (2013). 29. Bhimanapati, G. R.; Lin, Z.; Meunier, V.; Jung, Y.; Cha, J.; Das, S.; Xiao, D.; Son, Y.; Strano, M. S.; Cooper, V. R.; Liang, L. B.; Louie, S. G.; Ringe, E.; Zhou, W.; Kim, S. S.; Naik, R. R.; Sumpter, B. G.; Terrones, H.; Xia, F. N.; Wang, Y. L., et al. Recent Advances in Two-Dimensional Materials beyond Graphene. ACS Nano 9, 11509-11539 (2015). 30. Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric field effect in atomically thin carbon films. Science 306, 666-669 (2004). 31. Chhowalla, M.; Shin, H. S.; Eda, G.; Li, L. J.; Loh, K. P.; Zhang, H. The chemistry of two-dimensional layered transition metal dichalcogenide nanosheets. Nat. Chem. 5, 263- 275 (2013). 32. Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Katsnelson, M. I.; Grigorieva, I. V.; Dubonos, S. V.; Firsov, A. A. Two-dimensional gas of massless Dirac fermions in graphene. Nature 438, 197-200 (2005). 12 33. Zhang, Y. B.; Tan, Y. W.; Stormer, H. L.; Kim, P. Experimental observation of the quantum Hall effect and Berry's phase in graphene. Nature 438, 201-204 (2005). 34. Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Atomically Thin MoS 2: A New Direct-Gap Semiconductor. Phys. Rev. Lett. 105, (2010). 35. Splendiani, A.; Sun, L.; Zhang, Y. B.; Li, T. S.; Kim, J.; Chim, C. Y.; Galli, G.; Wang, F. Emerging Photoluminescence in Monolayer MoS2. Nano Lett. 10, 1271-1275 (2010). 36. Liu, M.; Yin, X. B.; Ulin-Avila, E.; Geng, B. S.; Zentgraf, T.; Ju, L.; Wang, F.; Zhang, X. A graphene-based broadband optical modulator. Nature 474, 64-67 (2011). 37. Furchi, M.; Urich, A.; Pospischil, A.; Lilley, G.; Unterrainer, K.; Detz, H.; Klang, P.; Andrews, A. M.; Schrenk, W.; Strasser, G.; Mueller, T. Microcavity-Integrated Graphene Photodetector. Nano Lett. 12, 2773-2777 (2012). 38. Geim, A. K.; Grigorieva, I. V. Van der Waals heterostructures. Nature 499, 419-425 (2013). 39. Eda, G.; Maier, S. A. Two-Dimensional Crystals: Managing Light for Optoelectronics. ACS Nano 7, 5660-5665 (2013). 40. Xia, F. N.; Wang, H.; Xiao, D.; Dubey, M.; Ramasubramaniam, A. Two-dimensional material nanophotonics. Nat. Photonics 8, 899-907 (2014). 13 Chapter 2: In situ and ex situ transmission electron microscopy study of lithiation behaviors of porous silicon nanostructures 2.1 Introduction To develop advanced Si-based lithium-ion batteries, it is essential to understand lithiation behaviors of the Si materials so that rational design of Si nanostructures can be achieved. To understand the lithiation/delithiation process of Si, it is of importance to directly observe the structural and chemical evolution during the process and thus correlate with the battery properties. Over the past few years, tremendous progress has been made toward developing methodologies for in situ observation of structural and chemical evolution of electrodes used for LIBs. Among them, in situ transmission electron microscopy (TEM) has been particularly informative and has revealed important features of the lithiation/delithiation process of Si nanoparticles and nanowires on phase transition, structural evolution, and lithiation kinetics 1-10 . Specifically, both c-Si nanoparticles and nanowires are reported to transform to amorphous LixSi (a-LixSi) via electrochemical- driven solid-state amorphization. With further lithiation, a-LixSi transforms to crystalline Li15Si4 (c-Li15Si4) 2-4, 8 . The fracture behavior of c-Si nanoparticles during the first lithiation is reported to be particle-size-dependent. The critical fracture diameter is 150 nm, below which cracks do not form, and above which surface cracking and particle fracture takes place upon lithiation 2 . In comparison, the critical fracture diameter of amorphous Si (a-Si) particles is reported to be up to 870 nm. In addition, the lithiation reaction velocity of a-Si is approximately constant and does not slow as in c-Si, which suggests different stress evolution during lithiation and implies that a-Si may be a more desirable active material than c-Si 9 . These studies have led to fundamental understanding of the 14 lithiation/delithiation process of Si nanoparticles and nanowires; however, these studies cannot provide direct explanation of better electrochemical performance achieved by newly reported nanostructured Si than solid Si nanoparticles and nanowires. Moreover, most studies only focus on the first several lithiation/delithiation cycles of Si, but do not look into post-cycling analysis of the structural evolution of Si. In this work, we study the phase transition and structural evolution of both porous Si nanoparticles and porous Si nanowires by in situ and ex situ TEM. The in situ TEM observation of lithiation process of porous Si nanoparticles reveals that the lithiation proceeds in an end-to-end manner, which is different from the surface-to-center manner for solid Si nanoparticles under the same experimental condition. In addition, much larger critical fracture diameter is achieved in porous Si particle than previously reported for c-Si and a-Si particles. Another interesting feature in the lithiation process of porous Si nanoparticles and nanowires is that a-LixSi does not transform to c-Li15Si4 even after full lithiation, which is distinct from that of solid Si nanoparticles and nanowires. The distinct lithiation behaviors of porous Si nanoparticles and nanowires are attributed to their interconnected three-dimensional porous structure, which is built up by numerous small domains. First-principle molecular dynamic simulation was conducted to investigate the effect of domain size on the phase stability of c-Li15Si4, which confirms the effect of nanostructure on phase transition. Moreover, structural evolution of porous and solid Si nanoparticles under successive lithiation/delithation cycles are compared through ex situ TEM, which confirms that porous Si is a more desirable anode material for LIBs than solid Si. 15 2.2 Material synthesis 2.2.1 Synthesis of ball-milled silicon nanoparticles and porous silicon nanoparticles To synthesize ball-milled silicon particles, Metallurgical Si particles were ground to fine powder using ball-milling operated at grinding speed of 1200 rpm for 5 hours. The Si powder was then washed with HF and DI-H2O successively to remove surface oxide layer. After drying at 90 o C in air for 6 hours, the particles were collected for further use. To synthesize porous silicon nanoparticles, metallurgical Si particles were ground to fine powder using ball-milling operated at grinding speed of 1200 rpm for 5 hours 11-12 . After that, the Si particles were soaked in a ferric etchant containing 0.03 M Fe(NO3)3 and 5 M HF under continuous stirring for 2 hours. The precipitates containing porous Si particles were then collected and washed with ethanol and DI-H2O. After drying at 90 o C in air for 6 hours, the particles were collected for further use. The TEM images of ball-milled silicon particles are shown in Figure 2.1 a and b, which demonstrate broad particle size distribution and the morphology of the particles is irregular. The TEM images of porous silicon particles are shown in Figure 2.1 c and d, which demonstrate numerous pores distribute uniformly throughout the whole porous Si particle after electroless etching of the ball-milled silicon. One domain of the porous silicon particle is marked by the red dotted line. Based on Figure 2.1 c and d, the domain size in porous silicon particles is 10-20 nm. 16 Figure 2.1 Characterization of ball-milled Si nanoparticles and porous Si nanoparticles. (a,b) TEM images of ball-milled Si nanoparticles at different magnifications. (c,d) TEM images of porous Si nanoparticles at different magnifications. One domain of the porous Si nanoparticle was marked by the red dotted line in (d). 2.2.2 Synthesis of solid silicon nanowires and porous silicon nanowires To synthesize solid silicon nanowires, Si wafers without doping were immersed in an etchant solution containing 0.02 M AgNO3 and 5 M HF for 3 h. After being washed with 17 DI-H2O, concentrated HNO3, and DI-H2O again, sequentially, solid Si nanowires were collected by scratching the wafers using a blade 11 . To synthesize porous silicon nanowires, boron-doped Si wafers (resistivity <5 mΩ·cm) were immersed in an etchant solution containing 0.02 M AgNO3 and 5 M HF for 3 h. After being washed with DI-H2O, concentrated HNO3, and DI-H2O again, sequentially, porous Si nanowires were collected by scratching the wafers using a blade 11 . The TEM images of solid silicon nanowires are shown in Figure 2.2 a and b, which high aspect ratio of the obtained solid silicon nanowires. The TEM images of porous silicon nanowires are shown in Figure 2.2 c and d, which demonstrate numerous pores distribute uniformly throughout the whole nanowires. One domain of the porous silicon nanowire is marked by the red dotted line. Based on Figure 2.2 c and d, the domain size in porous silicon nanowire is also 10-20 nm, which is similar to that of porous silicon nanoparticles. 18 Figure 2.2 Characterization of solid silicon nanowires and porous Si nanowires. (a,b) TEM images of solid Si nanowires at different magnifications. (c,d) TEM images of porous Si nanowires at different magnifications. One domain of the porous Si nanowire was marked by the red dotted line in (d). 2.3 In situ transmission electron microscopy characterization of lithiation of ball-milled silicon nanoparticles and porous silicon nanoparticles The in situ TEM nanobattery setup is schematically shown in Fig. 2.3a. Figure 2.3b-g show the lithiation process of a ball-milled Si particle with largest diameter of ~950 nm and 19 smallest diameter of ~630 nm. The ball-milled Si was prepared by ball-milling metallurgical Si and then being washed with HF and deionized water (DI-H2O) to remove the surface oxide layer. After the Li2O/Li electrode contacted the ball-milled Si, a potential of -2 V was applied to the Cu electrode with respect to Li 2O/Li electrode to initiate the lithiation process. As shown in Fig. 2.3c, a bump (indicated by the red arrow) comes out from the particle after lithiation for only 28 s, which is due to anisotropic expansion of Si particles. Further lithiation results in the change of contrast of the particle as shown in Fig. 2.3d. The gray LixSi shell and dark Si core indicates that lithium ions flow from surface to center of the particle in the radial direction. As the particle size is well above the reported critical fracture diameter of c-Si (150 nm), cracks (indicated by the blue arrows) start to form in the particle after lithiation for only 120 s (Fig. 2.3d). After 468 s of lithiation, the particle fractures into several pieces (Fig. 2.3g). The selected area electron diffraction (SAED) pattern in Fig. 2.3h exhibits rings made up of discrete spots, indicating nanosized polycrystalline nature of the ball-milled Si particle before lithiation. The ball-milled Si particle is made up of nanosized single crystalline Si particles, which results in the anisotropic expansion of the ball-milled Si particle during lithiation process. After full lithiation, the particle transforms to polycrystalline Li15Si4 as indicated by Fig. 2.3i. The Li2O phase in Fig. 2.3i is from the Li2O/Li electrode in contact with the particle. 20 Figure 2.3. In situ TEM observation of the lithiation process of a ball-milled Si particle. The test was carried out using a nanobattery configuration with a ball-milled Si particle attached to a Cu rod as the working electrode, Li as the reference electrode, and Li 2O as the solid electrolyte. (a) Schematic of the in situ TEM nanobattery. (b) TEM image of the ball-milled Si particle before lithiation. (c-g) Time series of the lithiation of the ball-milled Si particle, which illustrates the crack nucleation and fracture of the particle. After the Li2O/Li electrode contacted the ball-milled Si, a potential of -2 V was applied to the Cu electrode with respect to Li electrode to initiate the lithiation process. (h,i) Selected area electron diffraction (SAED) patterns of the ball-milled Si particle before (h) and after lithiation (i). Figure 2.4 shows the lithiation behavior of a typical porous Si particle. As illustrated in Fig. 2.4a numerous pores distribute uniformly throughout the whole porous Si particle after electroless etching of the ball-milled Si. To investigate the fracture behavior of porous Si particle during the lithiation process, we chose a large particle with diameter up to 1.52 μm for in situ TEM observation. Figure 2.4b-f demonstrate the TEM images of the porous Si particle during the lithiation process. From the TEM images, the volume expansion of the particle initiates in the lower right corner and then proceeds to the top left corner of the particle. This indicates that the lithium ions flow in an end-to-end manner, which is distinct from the surface-to-center lithiation manner observed in both crystalline and amorphous Si particles 2, 9 . To clarify the lithium propagation manner of porous Si particle, the lithiation behavior of another porous Si particle was characterized by in situ TEM with higher 21 magnification as shown in Figure 2.5. The lithiation front is marked by the red dotted line in Figure 2.5b-d, which propagates from lithium source to the other end of the particle. This observation is consistent with Fig. 2.4 and confirms the end-to-end lithiation manner of porous Si particle. After lithiation for 1121 s, the volume expansion of the particle almost ended (Fig. 2.4e). To ensure full lithiation of the particle, the -2 V potential was applied to the Cu electrode for another ~200 s and no obvious volume expansion of the particle was observed during this period. After lithiation for 1335 s, no crack was observed in the particle and the diameter of the particle increased to 2.05 μm, corresponding to a volume expansion of 145% (Figure 2.4f). The volume expansion is far less than the theoretical 300% for solid Si particles after full lithiation. This is attributed to the porous structure of the particle, which provides large space to accommodate the volume expansion by possible inward expansion during the lithiation process. The SAED patterns of the particle were obtained before lithiation and after full lithiation as shown in Figure 2.4g and h, respectively. Before lithiation, the porous Si particle is polycrystalline as shown in Figure 2.4g. After full lithiation (Figure 2.4h), the SAED pattern indicates that a-LixSi (marked by the blue arc) and c-Li15Si4 (indicated by the green arrow) coexist 1, 13 . The rings from Li2O/Li electrode are marked by the yellow arcs. This observation contrasts the SAED pattern of the fully lithiated ball-milled Si, which exhibits only c-Li15Si4 phase as shown in Figure 2.3i. We note that the porous Si and ball-milled Si particles are prepared from the same starting material. Taking into account that the most distinguishable difference between the porous Si and the ball-milled Si is their microstructures, we believe that the porous structure helps to prevent the formation of c-Li15Si4 phase during the first lithiation process, and we will discuss it in detail later. 22 Figure 2.4. In situ TEM observation of the lithiation process of a porous Si particle. (a) TEM image of the porous Si particle with diameter up to 1.52 μm before lithiation. (b-f) Time series of the lithiation of the porous Si particle, which illustrates the volume expansion of the particle without crack formation. (g,h) SAED patterns of the porous Si particle before (g) and after lithiation (h). Figure 2.5 In situ TEM observation of the lithiation process of a porous Si particle with high magnification to show the lithiation front. (a) TEM image of the porous Si particle before lithiation. (b-d) Time series of the lithiation of the porous Si particle showing the propagation of the lithiation front, which is indicated by the red dotted line. A brief summary of the lithiation behaviors of solid and porous Si particles reveal that lithiation proceeds in a surface-to-center manner for solid Si particles while in an end-to- end manner for porous Si particles. In addition, c-Li15Si4 phase forms in solid Si particles 23 after full lithiation while a-LixSi phase forms in porous Si particles after full lithiation. In order to further demonstrate whether the microstructure or the starting material of porous Si would affect its lithiation behavior, we prepared solid Si nanowires and porous Si nanowires and studied their lithiation behaviors. 2.4 In situ transmission electron microscopy characterization of lithiation of solid silicon nanowires and porous silicon nanowires Figure 2.6 shows the lithiation behavior of a typical solid Si nanowire with diameter of ~120 nm and length of ~600 nm. The Si nanowire before lithiation is shown in Fig. 2.6a. After lithiation for 103 s (Figure 2.6d), the gray shell and dark core of the nanowire reveal that the lithiation of Si nanowire occurs through the formation of a-LixSi shell and Si core structure, which is due to the faster lithium diffusion rate on the nanowire surface than that in the center. After lithiation for 150 s, the volume expansion of the nanowire almost ended (Figure 2.6e). To ensure full lithiation of the nanowire, the -2 V potential was applied to the Cu electrode for another ~130 s and no obvious volume expansion of the nanowire was observed during this period. After lithiation for 285 s, no crack was observed in the nanowire (Figure 2.6f). This is in agreement with a previous report, which demonstrates that the critical diameter for pulverization of Si nanowire is in the regime of 220-260 nm 14 . The SAED pattern of the Si nanowire before lithiation (Figure 2.6g) reveals its polycrystalline nature. After lithiation for 285 s, the SAED pattern of the nanowire (Figure 2.6h) indicates that it has transformed to the c-Li15Si4 phase, which is similar to the result of the ball-milled Si particle in Figure 2.3i. 24 Figure 2.6 In situ TEM observation of the lithiation process of a typical Si nanowire. (a) TEM image of the Si nanowire before lithiation. (b-f) Time series of the lithiation of the Si nanowire. (g,h) SAED patterns of the Si nanowire before (g) and after lithiation (h). The lithiation behavior of a porous Si nanowire bundle consisting of several porous Si nanowires was also examined by in situ TEM as demonstrated in Figure 2.7. As shown in Figure 2.7a the nanowires obtain highly porous structure with pore diameter and wall thickness of ~8 nm before lithiation. The single nanowire beside the nanowire bundle in Figure 2.7a acts as the lithium diffusion path during the lithiation process. Figure 2.7b-g demonstrate the lithiation process of the porous Si nanowire bundle, from which we can find that the contrast of the nanowires from the porous structure becomes obscure and uniform during the process. This indicates that the a-LixSi expands into the void space in the nanowires, which helps to minimize the volume expansion of the nanowires. The lithiation front is marked by the red dotted line in Figure 2.7b-e, which also indicates the end-to-end lithiation manner similar to that of porous Si particle. After lithiation for 823 s, lithium was observed to diffuse out of the nanowire bundle as indicated by the red arrow in Figure 2.7f, indicating that the lithiation process was complete. To ensure full lithiation, 25 the -2 V potential was applied to the Cu electrode for another ~140 s and no obvious volume expansion of the nanowire bundle was observed during this period. After lithiation for 964 s, no crack was observed in the nanowire bundle (Figure 2.7g). Figure 2.7h shows the SAED pattern of the nanowire bundle before lithiation, which reveals its polycrystalline nature. After lithiation for 964 s, the SAED pattern of the porous Si nanowire bundle (Figure 2.7i) demonstrates that it has transformed to a-LixSi (marked by the blue arc). This observation contrasts the SAED pattern of the fully lithiated solid Si nanowire, which exhibits only c-Li15Si4 phase as shown in Figure 2.6h. Figure 2.7 In situ TEM observation of the lithiation process of a porous Si nanowire bundle. (a) TEM image of the porous Si nanowire bundle before lithiation. (b-g) Time series of the lithiation of the porous Si nanowire bundle. The single nanowire beside the bundle provides lithium diffusion path. (h,i) SAED patterns of the Si nanowire bundle before (h) and after lithiation (i). 26 2.5 Mechanism of lithiation behaviors of porous silicon nanostructures A brief summary of the lithiation behaviors of solid and porous Si nanostructures reveal that lithiation proceeds in a surface-to-center manner for solid Si while in an end-to-end manner for porous Si. Figure 3 schematically illustrates the different lithiation manners of solid and porous Si particles. As lithium diffuses faster in the surface of Si than that in the bulk, a-LixSi shell will form in ball-milled Si particle once lithiation occurs (Figure 2.8b). As the a-LixSi shell thickens, cracks will form on the surface of the particle (Figure 2.8c), which lead to final pulverization of the ball-milled Si particle as shown in Figure 2.8d. The situation is different in porous Si particle, which is made up of numerous small domains (Figure 2.8e) and possesses complex surface topological feature. The large and complex surface of porous Si lags the propagation of lithium in the whole particle. As a result, lithium tends to proceed from the lithium source and propagate through the whole particle in an end-to-end manner, even though lithium may proceed in a surface-to-center manner in each domain as shown in Figure 2.8f and g. Because each domain in porous Si is in several nanometers, which is much smaller than the critical fracture diameter of solid Si, no crack will form during lithiation process (Figure 2.8h). In addition, the porous structure provides large space to accommodate the volume expansion by possible inward expansion of each domain, leading to smaller volume change of the porous Si particle than solid Si particle. 27 Figure 2.8 Schematic diagram illustrating the lithiation manners of ball-milled Si and porous Si nanoparticles. (a-d) Schematic diagram showing the surface-to-center lithiation manner of ball-milled Si particle. (e-h) Schematic diagram showing the end-to-end lithiation manner of porous Si particle. Another difference in lithiation behaviors of porous Si and solid Si nanostructures is that solid Si nanostructures transform to c-Li15Si4 while porous Si nanostructures transform to a-LixSi after full lithiation. The porous Si nanostructures are made up of small Si domains, while the domain of solid Si nanoparticle or solid Si nanowire is the whole nanoparticle or whole nanowire due to their solid structures. We believe that the different sizes of the domains of porous Si and solid Si lead to their different phase transition behaviors. To further illustrate the effect of domain size on the resultant phase after lithiation, first- principle molecular dynamic simulation was performed to study the structure stability of nanosized c-Li15Si4 particle. The simulated nanoparticle was constructed by 2×2×2 Li15Si4 crystalline supercells, which is composed of 128 Si atoms and 480 Li atoms, and corresponds to the size of 2 nm in three dimensions. Periodic boundary condition is applied in the simulation, and the empty space between Li 15Si4 particles is set larger than 1 nm to 28 exclude the mutual interaction of atoms from neighboring particles. First-principle calculations were performed using the VASP code density functional theory (DFT) calculations in generalized gradient approximation (GGA) with the Perdew-Burke- Ernzerhof (PBE) function used to calculate the force among atoms 15-16 . Molecular dynamic simulation was carried out at 300 K with a time step of 1 fs interval. Figure 2.9a-e show the structural evolution of Li15Si4 nanoparticle from the initial crystal to a disordered structure after 400 fs simulation. The yellow atoms are Si, and blue atoms are Li. At the early stage of the simulation (e.g. 100 fs), it is clear to see that the surface atoms are the first to deviate from their original positions due to the lack of symmetric force potential at the particle surface (Figure 2.9b). In the following simulation, cascaded breakdown of the periodic force potential leads to the structure disordering from outer surface to the inner part of particle. After 400 fs simulation, the particle turns to an amorphous structure (Figure 2.9e). To semi-quantify the structure amorphization, the radial distribution function (RDF) of Si-Si pairs was calculated and shown in Figure 2.9f. At the initial stage (0-100 fs), the sharp peaks in RDF illustrate the well-defined crystal structure. However, after 400 fs simulation, peaks at large Si-Si distance are largely smoothed, indicating the disappearance of ordered atomic arrangement. The small peak showing up at 2.5 Å corresponds to the distance of Si-Si in the amorphous Si structure, which further demonstrates the destroying of crystalline Li15Si4 structure. 29 Figure 2.9 First-principle molecular dynamic simulation to study the structure stability of a nanosized c-Li15Si4 particle. (a) The modeled structure of c-Li15Si4. (b-e) Atomic structure and morphology of the Li15Si4 particle at different simulation stages. (f) Si-Si radial pair distribution function at different stages of the simulated process. The appearance and increasing intensity of the peak at 2.5 Å indicate the intermixing of Si and Li to form an amorphous phase. Due to the constrained computation resource for first-principle molecular dynamic simulation of large-size particles, we adopted classical molecular dynamic simulation to characterize the structure stability of c-Li15Si4 particles with the same initial crystal structure as 2 nm particle (Figure 2.9a) while with larger diameter of 6 nm, 8 nm, 10 nm, and 12 nm. After 400 fs simulation, the atomic structures and morphologies of the Li15Si4 particles are illustrated in Figure 2.10. The yellow atoms are Si, and blue atoms are Li. Periodic boundary condition was used with particle-to-particle distance larger than 5 nm to eliminate the mutual interaction. Simulation were conducted by using the LAMMPS software code 17 , and a second nearest neighbor (2NN) modified embedded atom method 30 (MEAM) potential was used to account for the atomic interaction in Li-Si system 18 . Based on the comparison of enlarged images in Figure 2.10a-d, it is found that in 6 nm Li15Si4 particle (Figure 2.10a), the surface of the particle is in amorphous structure and the core atoms have lost their initial crystalline arrangement. In 8 nm particle, however, the crystallinity of the core increases compared with that of 6 nm particle even though the surface atoms in the 8 nm particle still rearrange in amorphous structure (Figure 2.10b). Similar trend is observed in 10 nm particle (Figure 2.10c) and when particle size increases to 12 nm, the crystallinity of the core is the highest and the crystalline volume is the largest among four particles even though the surface of 12 nm particle still tends to be amorphous (Figure 2.10d). Generally speaking, the trend is that as the particle size increases, the crystallinity in the core of the particles and the crystalline volume in the particles also increase. However, due to lack of symmetric force potential in the surface, the surface atoms in the particles always tend to deviate from their original positions and thus rearrange in amorphous structure. This simulation result further supports our conclusion that the small domains in porous Si nanostructures help to suppress c-Li15Si4 formation during the first lithiation process. Besides, this explains why some diffraction spots of c-Li15Si4 show up in Figure 2.4h, which may be due to the formation of c-Li15Si4 in the cores of some large-size domains in the porous Si particle after full lithiation. For porous Si nanowire, however, only a-LixSi forms after full lithiation (Figure 2.7i), which is possibly due to the smaller size of domains in porous Si nanowires than that in porous Si nanoparticles as we compare the domains marked in Figure 2.1d and Figure 2.2d. The formation of c-Li15Si4 during the first lithiation process is reported to be detrimental to the cycle life of Si-based LIBs and a cutoff voltage higher than 0.05 V is usually selected to suppress the formation 31 of c-Li15Si4 at low potential 19 . Here, we report that in addition to the low cutoff voltage, the nanoporous structure can also suppress the formation of c-Li15Si4 during first lithiation process due to the effect of domain size, which helps to achieve the excellent cycling performances of porous Si nanostructures. Figure 2.10 Classical molecular dynamic simulation to study the structure stability of c- Li15Si4 particles with different sizes. Atomic structures and morphologies of the Li15Si4 particles with diameter of 6 nm (a), 8 nm (b), 10 nm (c), and 12 nm (d) after 400 fs simulation. The insets in (a-d) are the enlarged images showing the surface and core of the corresponding particles. 2.6 Ex situ transmission electron microscopy characterization of ball- milled silicon nanoparticles and porous silicon nanoparticles To further characterize the structural evolution of ball-milled Si and porous Si, ex situ TEM images and corresponding SAED patterns of the two samples were obtained before cycling and after being charge-discharged for 1 cycle, 10 cycles, and 50 cycles in Li-Si cells in the 32 voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g as shown in Figure 2.11a-h. According to previous reports, the cutoff voltage range plays an important role to induce pore evolution in Si. Specifically, a large voltage window of 0.05-1 V would lead to porous structure of Si while Si cycled in a small voltage window of 0.17-0.6 V retains its original structure well after cycling 20 . In this work, we cycled the Li-Si cells in large voltage window of 0.01-2 V (vs. Li/Li + ) so that we can study the capability of porous Si and ball-milled Si to suppress pore evolution during the cycling process. Before cycling, both ball-milled Si and porous Si are polycrystalline as indicated by the inset SAED patterns in Figure 2.11a and Figure 2.11e, respectively. After being charge-discharged for different cycles, both ball-milled Si and porous Si transform to amorphous structure as indicated by the inset SAED patterns in Figure 2.11b-d and Figure 2.11f-h. To quantitatively investigate the pore evolution processes of the two samples, pore size distributions were obtained based on statistical analysis of TEM images. Before cycling, the surface of ball-milled Si is smooth as shown in Figure 2.11a. For porous Si particles, the pores are clearly resolved by the contrast in the image in Figure 2.11e, and the mean diameter is 9.9 ± 0.1 nm based on the pore size distribution diagram in Figure 2.11i. After being charge-discharged for only 1 cycle, nanopores are observed to form on the periphery of the ball-milled Si as indicated by the dark/light contrast in Figure 2.11b, which is due to inelastic deformation of Li/Si during the lithiation/delithiation process 20 . On the contrary, the porous Si particle retains its original porous structure well as shown in Figure 2.11f. This is confirmed by the pore size distribution diagram of two samples in Figure 2.11j. The mean diameter of newly-formed pores in the ball-milled Si is 3.8 ± 0.1 nm. While for porous Si, the mean diameter of pores is 10.9 ± 0.1 nm, which is close to its original value 33 before cycling. After cycling for 10 cycles, the surface of the ball-milled Si particles gets much rougher (Figure 2.11c), while the pore size increase of porous Si is still not significant (Figure 2.11g). As shown in Figure 2.11k, the mean diameter of the pores in the ball-milled Si increases drastically to 20.1 ± 0.1 nm, corresponding to a 429% increase compared with that after 1 cycle. Similar to Ostwald ripening in which particles agglomerate to reduce surface energy, this increase of pore size with cycling is equivalent to agglomeration of pores so that the surface energy of the particle can be reduced 21 . In contrast to the significant pore size increase in ball-milled Si, the mean diameter of pores in porous Si is only 12.8 ± 0.1 nm after 10 charge-discharge cycles, corresponding to only 29% increase compared with that before cycling. After cycling for 50 cycles, pores in both ball-milled Si and porous Si increase in size as shown in Figure 2.11d and h. According to Figure 2.11l, the mean diameter of pores for ball-milled Si is 41.8 ± 0.1 nm. However, the mean diameter of pores for porous Si is 36.0 ± 0.1 nm, which is still smaller than that of ball-milled Si. As the pore evolution is due to inelastic deformation of Li/Si during lithiation/delithiation process, for solid Si nanoparticles, the inelastic deformation is severe due to its large volume change during lithiation/delithiation process. For porous Si nanoparticles, however, the domains of the particle are observed to expand into the void space in the particle based on the observation in Figure 2.5a-d that the contrast of the particle from the porous structure becomes obscure and uniform during the lithiation process. This lithiation behavior of porous Si particle results in smaller volume change of the particle and stress relaxation in each domain. The stress relaxation prevents the stress in porous Si nanostructures from exceeding the elastic limit of Si, and thus suppresses the pore evolution in porous Si nanostructures. 34 The formation and size increase of pores in ball-milled Si particles would cause significant volume expansion of particles as they transform from solid particles to totally porous structure. However, with pre-formed pores, the volume change of porous Si particles before and after cycling is much less significant than that of ball-milled Si particles. This difference in particle volume change during cycling results in the different cycling performance of the two electrodes. Figure 2.11m shows the cycling performances of ball- milled Si and porous Si electrodes tested in the voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g. As shown in the figure, the capacity of ball-milled Si decays rapidly in the initial 10 cycles and then decreases in constant rate. This corresponds to the TEM observation that pore formation in ball-milled Si particles takes place in early cycles, which causes the particles to lose electrical contact from the current collector and thus leads to loss of active materials for capacity contribution. The volume change of porous Si during cycling is much less than that of ball-milled Si; however, the relatively large surface area of porous Si as compared to ball-milled Si would lead to more severe solid electrolyte interface (SEI) formation on porous Si particles, which would also cause capacity decay in the initial cycles. For this reason, it is essential to apply coating on porous Si particles (e.g. carbon coating) to mitigate the SEI formation and thus to further improve the cyclability of porous Si electrodes 12, 22-23 . 35 Figure 2.11 Ex situ TEM characterization of ball-milled Si and porous Si after different charge-discharge cycles and comparison of their cycling performances. The Si electrodes were cycled in Li-Si cells in the voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g and then disassembled at the delithiated state before TEM observation. (a-d) TEM images of ball-milled Si before cycling (a), after cycling for 1 cycle (b), 10 cycles (c), and 50 cycles (d). (e-h) TEM images of porous Si before cycling (e), after cycling for 1 cycle (f), 10 cycles (g), and 50 cycles (h). The insets in (a-h) are the corresponding SAED patterns. (i-l) Pore size distributions of porous Si before cycling (i) and the comparison of ball-milled Si and porous Si after cycling for 1 cycle (j), 10 cycles (k), and 50 cycles (l). (m) Cycling performances of Li-Si cells using ball-milled Si and porous Si as working electrode, respectively. The galvanostatic charge-discharge test was carried out in the voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g. 36 2.7 Summary In conclusion, we have applied in situ and ex situ TEM to study the structural evolution and phase transition of porous Si nanoparticles and nanowires and have compared their behaviors with solid Si nanoparticles and nanowires. The critical fracture diameter of porous Si particles reaches up to 1.52 μm, which reveals its better capacity to accommodate volume expansion during the lithiation process. In addition, the porous Si nanoparticles and nanowires transform to the a-LixSi phase after full lithiation in contrast to the c-Li15Si 4 phase for solid Si nanoparticles and nanowires, which is due to small Si domains in porous Si nanoparticles and nanowires as revealed by the first-principle molecular dynamic simulation. Finally, ex situ TEM observation of porous Si nanoparticles and solid Si nanoparticles reveal that porous Si nanoparticles obtain better capability to suppress pore evolution than solid Si nanoparticles during the cycling process. The better capabilities of porous Si nanostructures to accommodate volume expansion, to suppress c-Li15Si4 formation during the first lithiation process, and to suppress pore evolution during cycling make them more desirable lithium-ion battery anode materials than solid Si nanostructures. 37 2.8 References 1. Liu, X. H.; Zheng, H.; Zhong, L.; Huan, S.; Karki, K.; Zhang, L. Q.; Liu, Y.; Kushima, A.; Liang, W. T.; Wang, J. W.; Cho, J. H.; Epstein, E.; Dayeh, S. A.; Picraux, S. T.; Zhu, T.; Li, J.; Sullivan, J. P.; Cumings, J.; Wang, C. S.; Mao, S. X., et al. Anisotropic Swelling and Fracture of Silicon Nanowires during Lithiation. Nano Lett. 11, 3312-3318 (2011). 2. Liu, X. H.; Zhong, L.; Huang, S.; Mao, S. X.; Zhu, T.; Huang, J. Y. Size-Dependent Fracture of Silicon Nanoparticles During Lithiation. ACS Nano 6, 1522-1531 (2012). 3. Liu, X. H.; Zhang, L. Q.; Zhong, L.; Liu, Y.; Zheng, H.; Wang, J. W.; Cho, J. H.; Dayeh, S. A.; Picraux, S. T.; Sullivan, J. P.; Mao, S. X.; Ye, Z. Z.; Huang, J. Y. Ultrafast Electrochemical Lithiation of Individual Si Nanowire Anodes. Nano Lett. 11, 2251-2258 (2011). 4. Wang, C. M.; Li, X. L.; Wang, Z. G.; Xu, W.; Liu, J.; Gao, F.; Kovarik, L.; Zhang, J. G.; Howe, J.; Burton, D. J.; Liu, Z. Y.; Xiao, X. C.; Thevuthasan, S.; Baer, D. R. In Situ TEM Investigation of Congruent Phase Transition and Structural Evolution of Nanostructured Silicon/Carbon Anode for Lithium Ion Batteries. Nano Lett. 12, 1624-1632 (2012). 5. Gu, M.; Li, Y.; Li, X. L.; Hu, S. Y.; Zhang, X. W.; Xu, W.; Thevuthasan, S.; Baer, D. R.; Zhang, J. G.; Liu, J.; Wang, C. M. In Situ TEM Study of Lithiation Behavior of Silicon Nanoparticles Attached to and Embedded in a Carbon Matrix. ACS Nano 6, 8439-8447 (2012). 6. McDowell, M. T.; Ryu, I.; Lee, S. W.; Wang, C. M.; Nix, W. D.; Cui, Y. Studying the Kinetics of Crystalline Silicon Nanoparticle Lithiation with In Situ Transmission Electron Microscopy. Adv. Mater. 24, 6034-6041 (2012). 7. McDowell, M. T.; Lee, S. W.; Wang, C. M.; Cui, Y. The effect of metallic coatings and crystallinity on the volume expansion of silicon during electrochemical lithiation/delithiation. Nano Energy 1, 401-410 (2012). 8. Gu, M.; Wang, Z. G.; Connell, J. G.; Perea, D. E.; Lauhon, L. J.; Gao, F.; Wang, C. M. Electronic Origin for the Phase Transition from Amorphous LixSi to Crystalline Li15Si4. ACS Nano 7, 6303-6309 (2013). 9. McDowell, M. T.; Lee, S. W.; Harris, J. T.; Korgel, B. A.; Wang, C. M.; Nix, W. D.; Cui, Y. In Situ TEM of Two-Phase Lithiation of Amorphous Silicon Nanospheres. Nano Lett. 13, 758-764 (2013). 10. Luo, L. L.; Wu, J. S.; Luo, J. Y.; Huang, J. X.; Dravid, V. P. Dynamics of Electrochemical Lithiation/Delithiation of Graphene-Encapsulated Silicon Nanoparticles Studied by In-situ TEM. Sci. Rep. 4, 3863 (2014). 11. 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 (2012). 12. Ge, M. Y.; Lu, Y. H.; Ercius, P.; Rong, J. P.; Fang, X.; Mecklenburg, M.; Zhou, C. W. Large-Scale Fabrication, 3D Tomography, and Lithium-Ion Battery Application of Porous Silicon. Nano Lett. 14, 261-268 (2014). 13. Liu, X. H.; Wang, J. W.; Huang, S.; Fan, F. F.; Huang, X.; Liu, Y.; Krylyuk, S.; Yoo, J.; Dayeh, S. A.; Davydov, A. V.; Mao, S. X.; Picraux, S. T.; Zhang, S. L.; Li, J.; Zhu, T.; Huang, J. Y. In situ atomic-scale imaging of electrochemical lithiation in silicon. Nat. Nanotechnol. 7, 749-756 (2012). 14. Ryu, I.; Choi, J. W.; Cui, Y.; Nix, W. D. Size-dependent fracture of Si nanowire battery anodes. J. Mech. Phys. Solids 59, 1717-1730 (2011). 38 15. 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). 16. Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson, M. R.; Singh, D. J.; Fiolhais, C. Atoms, Molecules, Solids, and Surfaces: Applications of the Generalized Gradient Approximation for Exchange and Correlation. Phys. Rev. B 46, 6671-6687 (1992). 17. Plimpton, S. Fast Parallel Algorithms for Short-Range Molecular-Dynamics. J. Comput. Phys. 117, 1-19 (1995). 18. Cui, Z. W.; Gao, F.; Cui, Z. H.; Qu, J. M. A second nearest-neighbor embedded atom method interatomic potential for Li-Si alloys. J. Power Sources 207, 150-159 (2012). 19. 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). 20. Hu, L. B.; Wu, H.; Gao, Y. F.; Cao, A. Y.; Li, H. B.; McDough, J.; Xie, X.; Zhou, M.; Cui, Y. Silicon-Carbon Nanotube Coaxial Sponge as Li-Ion Anodes with High Areal Capacity. Adv. Energy Mater. 1, 523-527 (2011). 21. Choi, J. W.; McDonough, J.; Jeong, S.; Yoo, J. S.; Chan, C. K.; Cui, Y. Stepwise Nanopore Evolution in One-Dimensional Nanostructures. Nano Lett. 10, 1409-1413 (2010). 22. Li, X. L.; Gu, M.; Hu, S. Y.; Kennard, R.; Yan, P. F.; Chen, X. L.; Wang, C. M.; Sailor, M. J.; Zhang, J. G.; Liu, J. Mesoporous silicon sponge as an anti-pulverization structure for high-performance lithium-ion battery anodes. Nat. Commun. 5, (2014). 23. Lu, Z. D.; Liu, N.; Lee, H. W.; Zhao, J.; Li, W. Y.; Li, Y. Z.; Cui, Y. Nonfilling Carbon Coating of Porous Silicon Micrometer-Sized Particles for High-Performance Lithium Battery Anodes. ACS Nano 9, 2540-2547 (2015). 39 Chapter 3. Silicon(lithiated)-sulfur full cells with porous silicon anode shielded by Nafion against polysulfides to achieve high capacity and energy density 3.1 Introduction Despite the encouraging progress made with Si anode, the inherent low capacity of traditional cathode materials significantly compromises the utilization of Si in achieving high capacity and energy density in practical batteries. For instance, a large family of cathode materials, including lithium metal oxide (LiMO2, M=Co, Ni, Mn) 1-4 and lithium metal phosphate (LiMPO4, M=Fe, Co, Ni, Mn) 5 , generally have capacities around 150 mAh/g. According to Equation (1), integrating Si with these cathode materials leads to highest theoretical specific capacity of 144 mAh/g, C full cell = C Si ×C LiMO 2 C Si +C LiMO 2 (1) in which 𝐶 𝑆𝑖 is the specific capacity of Si anode (assumed to be 3600 mAh/g), and 𝐶 𝐿𝑖𝑀𝑂 2 is the specific capacity of LiMO2 cathode (assumed to be 150 mAh/g). It is important to note that Equation (1) only calculates the highest theoretical specific capacity of the full cell when the capacities of anode and cathode are equal. When the capacities of anode and cathode are not equal, the theoretical specific capacity of the full cell is lower than the highest theoretical specific capacity and it is calculated using the lower capacity among anode and cathode divided by the total mass of anode and cathode active materials. Here, the calculated highest theoretical specific capacity of Si-LiMO2 system is only 36% higher than that of graphite-LiCoO2 system, which is 106 mAh/g if we assume graphite has a capacity of 360 mAh/g. In addition, in order to optimize the loading of electrode materials, 40 excessive amount of cathode is required to balance the loading of Si, which would bring technological difficulties to coat thick and stable layer of cathode material on electrode substrate. It is therefore highly desired to find cathode replacement with higher specific capacity. Recently, sulfur (S) cathode has attracted great attention, notable for its high theoretical specific capacity (1675 mAh/g) and reduced cost compared with traditional cathode materials. It therefore holds great promise to investigate the Si-S (lithiated Si and/or lithiated S) full cell, as integration of Si anode and S cathode can theoretically deliver specific capacity more than 7 times and energy density more than 3 times higher than both graphite-LiMO2 and the potential Si-LiMO2 battery systems 6-7 . Due to lack of lithium in Si anode and S cathode, one electrode or both electrodes need to be lithiated before assembled into full cells, and therefore the full cells can be made of Si(lithiated)-S, Si-S(lithiated), or Si(lithiated)-S(lithiated). For simplicity, we call all three types Si-S full cells and when talking about a specific type of full cell, we will mention which electrode is lithiated in the following discussion. In order to achieve high capacity and energy density of Si-S full cell, the electrochemical performance of both the Si anode and S cathode should be optimized. For Si anode, as mentioned above, the main problem arising from electrode pulverization can be tackled by fabricating nanostructured Si particle such as porous Si. For S cathode, the main challenge is to reduce internal redox shuttle between dissoluble polysulfides anions (Li2S4-6) which leads to pronounced capacity fading and low Coulombic efficiency of electrode 8 . The most popular approach to tackle the problems is to infiltrate sulfur into various host materials to encapsulate the dissolved polysulfides, 41 with aim to alleviate the shuttle effect. Carbon-based materials such as porous carbon, intertwined carbon nanotubes, and graphene are the most widely used sulfur capture matrix 9-12 . Embedding sulfur into other inorganic porous structures is also receiving great attention. Research has demonstrated that the cyclability of Li-S battery can be significantly improved by confining sulfur into TiO2 13 and MnO2 14 nanostructures, which not only confines the sulfur species spatially, but also provides weak chemical interactions to trap the dissolved polysulfides with the presence of functional groups at the surface of these oxide materials. However, none of the sulfur-confinement approaches are capable of eliminating the polysulfides dissolution completely. Recently, research has been directed to a different type of approach. By saturating the electrolyte with added lithium polysulfides, dissolution of polysulfides from sulfur cathode during electrochemical cycling can be effectively alleviated 15 . Therefore, polysulfides dissolution in electrolyte to some extent is inevitable in Li-S battery and if it is coupled with Si anode, spontaneous reaction between Si (and lithiated Si) and polysulfides would take place, which would lead to capacity decay of the full cell. To shed light on Si-S full cell research development, in this work, we will discuss the failure mechanism of Si-S full cell. On this basis, we will provide some effective solutions to tackle the problem, and finally present a prototype of Si(lithiated)-S full cell. 3.2 Failure mechanism of silicon(lithiated)-sulfur full cells and functionality of Nafion In our experiment, we use porous Si particles with multiple protective coatings as anode, and S infiltrated in a mixture of carbon black, carbon nanofiber, and graphene as cathode 42 (S-C-G). To choose electrolyte for full cells, tests of Li-Si and Li-S half cells using two kinds of electrolytes were conducted and results indicate that both electrolytes which are commonly used for Si (1 M LiPF6 in dimethyl carbonate (DMC)/fluoroethylene carbonate (FEC), 1:1 by volume) and S (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO3) deliver similar results for Li-Si system. However, Li-S cells fail to operate in DMC/FEC based electrolyte, mainly due to the reaction between polysulfides and carbonate-based electrolytes via a nucleophilic addition or substitution reaction, leading to sudden capacity fading 16 . With the choice of DME/DOL based electrolyte, the full cell using lithiated bare porous Si as anode and S-C-G as cathode, however, experiences severe capacity degradation after a few cycles despite of their good performance in Li-Si and Li-S half cells (Figure 3.1). The poor performance of Si-S full cell is attributed to the spontaneous reaction between dissolved polysulfides and Si (and lithiated Si). Based on the chemical potential, the open circuit voltage of Si is around 2.4-2.8 V (vs. Li/Li + ); the lithiation/delithiation voltages are around 0.2-0.8 V (vs. Li/Li + ) for lithiated Si (LixSi), and 1.7-2.4 V (vs. Li/Li + ) for polysulfides (Li2Sy), respectively. When Si is in contact with polysulfides, chemical reaction can take place between Si and polysulfides. This reaction would consume Si and Li irreversibly and thus reduce the available capacity of the electrode. During cycling process, Si will be lithiated and the formed LixSi will have lower potential than polysulfides, which will lead to reaction between LixSi and polysulfides as following: Li x Si+ Li 2 S y → Li x−∆ Si+ Li 2+∆ S y (2) This lithium ion transfer leads to shuttle effect in the Si(lithiated)-S full cell, which will result in capacity degradation during cycling process. 43 Figure 3.1 Cyclic performance of Si-S full cell using lithiated bare porous Si as anode and S-C-G as cathode in the voltage window of 1.2-2.7 V (vs. Li/Li + ) at current density of 0.1 C. The electrolyte is LITFSI electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO3). The specific capacity of the Si-S full cell is calculated based on the mass of sulfur. To circumvent the shuttle effect resulted from direct attack by polysulfides, a protective coating layer on the surface of Si is highly desired. During the past years, fluoropolymer-copolymer based on sulfonated tetrafluoroethylene, known as Nafion, has received considerable attention as a proton conductor for proton exchange membrane fuel cells 17 . The sulfonate functional group (-SO3 - ) can effectively prevent the approaching of negative-charged anions due to electrostatic repulsion force. Inspired by the impermeability of Nafion for anions, we have come up with the idea of conformally coating a thin layer of Nafion on Si, and then evaluating its functionality to shield Si (and lithiated Si) from spontaneous reaction with polysulfides in Si-S full cell. To elucidate the function of Nafion, one piece of bare Si wafer and another piece of Si wafer with Nafion coating were immersed in the same polysulfides electrolyte (1 M LITFSI 44 in DME/DOL, 1:1 by volume, with addition of 1 M Li2S4). Figure 3.2 schematically shows the concept of the experiments. As shown in Figure 3.2a, when a bare Si wafer is immersed into the electrolyte, spontaneous reaction between Si and polysulfides would occur. After reaction for 12 hours, the wafer was then washed with DI-H2O, during which process the formed Li-Si-S compound reacted with water. As a result, the original clean surface of Si wafer would become rough. Figure 3.2b shows a SEM image of the bare Si wafer after the reaction and then being washed by DI-H2O. The SEM image clearly shows a rough surface of the wafer, indicating the etching of wafer due to the spontaneous reaction. To shield Si wafer from reaction with polysulfides, the other wafer was coated with a thin layer of Nafion by spin coating (Figure 3.2c), and then treated in the same way as the bare wafer. From the SEM image shown in Figure 3.2d, the clean and smooth surface of wafer confirms that the reaction between Si and polysulfides is effectively circumvented due to the protection by Nafion. 45 Figure 3.2 Functionality of Nafion in shielding Si from reaction with polysulfides. (a) Schematic diagram showing the reaction of a bare Si wafer with polysulfides electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 1 M Li2S4). After the reaction, a thin layer of Li-Si-S compound is formed on the surface of the wafer. (b) SEM image of Si wafer after being immersed in polysulfides electrolyte for 12 hours and washed by DI-H2O. The rough surface indicates the reaction between Si and polysulfides. (c) Schematic diagram showing a thin layer of Nafion coating on Si wafer can shield Si from reaction with polysulfides. (d) SEM image of Nafion-coated Si wafer after being immersed in polysulfides electrolyte for 12 hours and washed by DI-H2O, which shows clean and smooth surface. To further identify the product of the reaction between Si wafer and polysulfides electrolyte, EDS and XPS analyses were performed as shown in Figure 3.3. In the experiment, one piece of bare Si wafer was immersed in polysulfides electrolyte for 12 h. The wafer was then washed with DME/DOL and dried in Ar-protected environment before characterization. Figure 3.3a shows the EDS spectrum of the Si wafer after reaction with polysulfides electrolyte, in which Si, S, and O signals are clearly resolved. Compared with the EDS spectrum of a bare Si wafer (Figure 3.3b), where only a peak corresponding to Si is observed, the presence of S signal in Figure 3.3a confirms the reaction between the Si wafer and polysulfides electrolyte, which results in the formation of complex Si-S 46 compound. For comparison, we prepared a “polysulfides electrolyte sample” by dropping polysulfides electrolyte onto a Cu foil and dried it in vacuum oven so that the solutes in the electrolyte would be left on the Cu foil for EDS measurement. Compared with the EDS spectrum of pure polysulfides electrolyte (Figure 3.3c), the absence of F, N, and C signals in Figure 3.3a confirms that the polysulfides electrolyte is completely washed away from the Si wafer so that the S signal in Figure 3.3a is from the reaction product of Si wafer and polysulfides electrolyte. The O content in the electrolyte based on EDS spectrum (Figure 3.3c) is 28.3%, which is higher than 14.9% calculated based on chemical formulas of the solutes in the electrolyte. The excess O is believed to result from the oxidation of polysulfides during sample transferring process from the Ar-protected environment to the scanning electron microscope. Similarly, the O signal in Figure 3.3a is also due to oxidation of reaction product of Si wafer and polysulfides electrolyte during the sample transferring process. 47 Figure 3.3 EDS and XPS analyses of the reaction product of Si wafers and polysulfides electrolyte. (a-c) EDS spectra of a Si wafer after reaction with polysulfides electrolyte (a), a bare Si wafer (b), and polysulfides electrolyte (c). Insets show the SEM images of corresponding regions for collecting the EDS spectra. (d-f) Li 2s (d), S 2p (e), and Si 2p (f) XPS spectra of a Si wafer after reaction with polysulfides electrolyte. (g,h) Li 2s (g) and S 2p (h) XPS spectra of polysulfides electrolyte. (i) Si 2p XPS spectrum of a bare Si wafer. Hollow circles: experimental data; gray lines: background; black lines: overall fit; colored lines: fitted individual components. Figure 3.3d-f show the Li 2s, S 2p, and Si 2p XPS spectra of a Si wafer after reaction with polysulfides electrolyte. For comparison, the Li 2s and S 2p XPS spectra of polysulfides 48 electrolyte are shown in Figure 3.3g and 3.3h, and Si 2p XPS spectrum of a bare Si wafer is shown in Figure 3.3i. Firstly, we compare Figure 3.3d and g. The Li 1s peak at 55.9 eV in Figure 3.3g is assigned to LixSOy, which demonstrates the oxidation of polysulfides electrolyte during the sample transferring process 18 . This is consistent with the EDS result in Figure 3.3c, which demonstrates high O content in the electrolyte. In comparison, the LixSO y peak (blue curve) also shows up in the Li 1s spectrum of Si wafer after reaction with polysulfides electrolyte (Figure 3.3d). Since the electrolyte is completely washed away from the wafer before characterization, it is believed that the LixSOy in Figure 3.3d is from the reaction product between the Si wafer and polysulfides, which is consistent with the EDS result in Figure 3.3a. In addition, the peak at 52.9 eV in Figure 3.3d is assigned to LixSi, indicating that the reaction product also contains Li-Si bond 19 . From this comparison, we conclude that the reaction product contains both Li-Si and Li-S bonds. Further confirmation comes from the comparison of S 2p spectra for reaction product (Figure 3.3e) and polysulfides electrolyte (Figure 3.3h). In Figure 3.3h, the S 2p peak (composed of 2p3/2 and 2p1/2 due to spin-orbit coupling) at 169.6 eV (green curves) corresponds to LITFSI, and the peaks in the range of 166-160 eV are assigned to Li2Sx species in the polysulfides electrolyte 20 . The peak at 167.9 eV (magenta curves) may result from the formation of Li2SO3 or Li2SO4, and therefore we assign it to LixSOy 18, 21 . This LixSOy peak indicates the oxidation of polysulfides during the sample transferring process, which is consistent with the EDS result in Figure 3.3c and XPS result in Figure 3.3g. By comparing Figure 3.3e and h, it is observed that the peaks which correspond to LITFSI and Li2Sx disappear in Figure 3.3e, confirming that the polysulfides electrolyte is completely washed away after reaction. On the contrary, a broad peak in the range of 172-162 eV is 49 observed and can be deconvoluted to four sub-peaks. As shown in Figure 3.3e, the peak at 168.1 eV (red curves) may belong to -SO3 or -SO4 18, 21 , and the peak at 165.6 eV (blue curves) may belong to -S2O6 21 . It should be noted that due to the complex oxidation states of the S element, and possible effect of Si element on the bonding environment in the reaction product, the 172-162 eV broad peak may contain other sub-peaks with low intensity (not shown in the figure), which results in the relatively large peak splitting of 2p3/2/2p1/2 doublets in Figure 3.3e, compared to the LixSO y 2p3/2/2p1/2 doublet in Figure 3.3h. For these reasons, we assign the 4 sub-peaks in Figure 3.3e as -SOx. Finally, we have compared Figure 3.3f and 3.3i. As shown in Figure 3.3i, the Si 2p peak is composed of 2p3/2 (99.4 eV) and 2p1/2 (100.0 eV) spin-orbit components 22 . In contrast, the main peaks in Figure 3.3f located at 98.8 eV and 99.4 eV are assigned to LixSi 2p3/2 and 2p1/2 components, respectively 23-24 . In addition, a broad peak centered at 101.5 eV is observed, which is assigned to LixSiOy. The existence of the LixSi and LixSiO y peaks in Figure 3.3f provides further evidence that the reaction product contains Si element, which is consistent with the XPS result in Figure 3.3d. In summary, based on the EDS and XPS analyses, we conclude that the reaction product of Si wafers and polysulfides electrolyte is composed of Li, Si, and S. 3.3 Preparation of silicon anode and sulfur cathode Synthesis of porous Si particles: Porous Si particles were synthesized according to our previous report 25 . Specifically, metallurgical Si particles were first ground to fine powder using ball-milling operated at grinding speed of 1200 rpm for 5 hours. The Si powder was washed in diluted hydrofluoric acid (HF) and deionized water (DI-H2O) successively to 50 remove surface oxide layer. The Si power was then soaked in a ferric etchant containing 30 mM Fe(NO3)3 and 5 M HF under continuous stirring. After 2 hours of reaction, precipitates containing porous Si particles were collected and washed using ethanol and DI-H2O. The washed particles were dried for further use. Nafion coating on porous Si particles: Nafion solution (5 wt.% in ethanol) was bought from Sigma-Aldrich. Porous Si particles were soaked in Nafion solution for 12 hours. After that, the particles were centrifuged to remove excessive Nafion, and further washed with DI-H2O before drying to get powder. In the paper, this Nafion-coated Si is denoted as Si-N. Carbon coating on porous Si particles: Chemical vapor deposition (CVD) was used to coat a thin layer of carbon on the surface of porous Si particles. Specifically, porous Si particles were loaded into a tube furnace, and gradually elevated the furnace temperature to 860 o C in Ar-protected environment. At 860 o C, diluted ethylene (C2H4:Ar = 1:10 by volume) was fed through, and the tube is kept in ambient pressure. After 15 min reaction, furnace was naturally cooled down to collect the carbon-coated Si. In the paper, this carbon-coated Si is denoted as Si-C. Nafion coating on Si-C particles: Nafion coating on Si-C particles was followed by the same procedure as used to coat Nafion on Si particles. In the paper, this Nafion-coated Si-C is denoted as Si-C-N. 51 Graphene oxide coating on Si-C particles: Graphene oxide (GO) was prepared according to modified Hummers method 26 . Si-C particles and GO (10:1 by weight) were mixed in ethanol under continuous stirring for 10 min. Droplets of hydrazine and ammonia were added into the mixed solution, and kept at 90 o C for 1 hour. GO-coated particles were collected by centrifuge and then washed by DI-H2O twice. The particles were dried to get powder. In the paper, this GO-coated Si-C is denoted as Si-C-G. Graphene oxide coating on Si-C-N particles: GO coating on Si-C-N particles was followed by the same procedure as used to coat GO on Si-C particles. In the paper, this GO-coated Si-C-N is denoted as Si-C-N-G. Preparation of S cathode material: Elemental sulfur was mixed with carbon black and carbon nanofiber with mass ratio of 10:4:1. The mixture was annealed at 155 o C for 4 hours to infiltrate S into the carbon matrix. The mixture was then coated with GO following the same procedure as used to prepare Si-C-G. In the paper, this S-based composite is denoted as S-C-G. Preparation of Si-based electrode: The active material can be selected from Si, Si-N, Si-C, Si-C-N, Si-C-G, and Si-C-N-G. To prepare electrode, active Si material was first mixed with carbon black and alginic acid sodium salt with mass ratio of 7:2:1.5 in water to form uniform slurry. The slurry was coated on copper foil and then dried at 90 o C in air for 6 hours. 52 Preparation of S electrode: To prepare S electrode, S-C-G was mixed with polyvinylidene fluoride with mass ratio of 9:1 in N-methyl-pyrrolidone to form uniform slurry. The slurry was coated on aluminum foil and then dried at 90 o C in air for 6 hours. For Li-Si and Li-S half cell measurements, CR2032 coin cells were assembled using lithium foil as counter/reference electrode and Celgard 2400 as separator. The prepared Si-based electrodes or S electrodes were used as working electrodes. Two kinds of electrolytes were used in different experiments: 1. Polysulfides electrolyte: 1 M lithium bis(trifluoromethanesulfonyl)imide (LITFSI) in 1,2-dimethoxyethane (DME)/1,3-dioxolane (DOL), 1:1 by volume, with addition of 1 M Li2S4; 2. LITFSI electrolyte: 1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO 3. The galvanostatic charge-discharge test was carried out in the voltage window of 0.01-2 V (vs. Li/Li + ) at current density of 400 mA/g for Li-Si half cells and in the voltage window of 1.7-2.7 V (vs. Li/Li + ) at current density of 0.1 C (1C=1600 mA/g) for Li-S half cells. The electrochemical impedance spectra (EIS) of Li-Si half cells were collected with an AC voltage of 5 mV amplitude in the frequency range of 1000 kHz to 10 mHz. To make Si(lithiated)-S full cells, Si-based electrodes were first assembled in Li-Si half-cell configuration using LITFSI electrolyte. After being charged and discharged for 5 cycles in the voltage window of 0.01-2 V (vs. Li/Li + ) at a current density of 400 mA/g, Si-based electrodes were disassembled at lithiated state, and then washed carefully with DME/DOL in Ar-protected environment. After drying in Ar-protected environment, the lithiated Si-based electrodes were assembled into full cells with S-C-G cathode using 53 CR2032 coin cells. The separator is Celgard 2400 and electrolyte is LITFSI electrolyte. The galvanostatic charge-discharge test for full cells was carried out in the voltage window of 1.2 -2.7 V (vs. Li/Li + ) at current densities from 0.1 C to 0.8 C (1C=1600 mA/g based on the mass of sulfur). The cyclic voltammetry (CV) of full cells was conducted at scan rate of 0.2 mV/s in the voltage window of 1.2-2.7 V. Following conventions in literature, for all the Li-Si half cells in this paper, we define the lithiation process to be charge and delithiation process to be discharge, and the Coulombic efficiency of the Li-Si half cells is defined as the discharge capacity divided by the preceding charge capacity. For all the Li-S half cells, we define the lithiation process to be discharge and delithiation process to be charge, and the Coulombic efficiency of the Li-S half cells is defined as the discharge capacity divided by the following charge capacity. For all the Si(lithiated)-S full cells, we define the lithiation of the S-C-G cathode process to be discharge and the delithiation of the S-C-G cathode process to be charge, and the Coulombic efficiency of the Si(lithiated)-S full cells is defined as the discharge capacity divided by the preceding charge capacity 27-28 . 3.4 Characterization of silicon anode and sulfur cathode To assemble full cells, porous Si particles are used as the anode material, because they have demonstrated good cyclic performance in Li-Si half cells and the preparation method is cost-effective and scalable as reported in our previous work 25 . Specifically, metallurgical Si was milled to submicron particles, and then got etched in a ferric etchant, leaving a highly porous structure. As-synthesized porous Si particles were then dispersed in Nafion 54 solution (5 wt.% in ethanol) overnight to coat a thin layer of Nafion on particle surface (the product is named Si-N). Figure 3.4a shows the TEM image of a typical pristine porous particle. Numerous pores are uniformly distributed throughout the whole particle with pore size of around 10-15 nm, which can be clearly resolved in the high resolution TEM (HRTEM) image as shown in Figure 3.4b. After Nafion coating, there is no significant change in the particle morphology (Figure 3.4c); however, the porous feature is not as clear as it shows in pristine Si particles, mainly due to the filling of Nafion into the porous structure. From the HRTEM image of Nafion-coated Si in Figure 3.4d, it is easy to identify the amorphous layer on the periphery of particle with a thickness of approximately 2-4 nm as indicated by the dotted line. To further confirm the existence of Nafion and rule out the possibility of native silicon oxide in the amorphous layer, we have collected energy filtered electron signals for specific Si and S elements to map out the Si and Nafion distribution (S element is from Nafion). TEM image in Figure 3.4e shows the region of interest, and Figure 3.4f-h show the elemental distribution of Si, S, and their superposition. From Figure 3.4h, signals from Si and S are well overlapped, which confirms the uniform coating of Nafion on the surface of Si. FTIR spectra of Si, Si-N, and Nafion were also obtained to confirm the coating of Nafion on Si (Figure 3.5). The characteristic peaks of Nafion at 1226, 1148, 1060, and 982 cm -1 correspond to asymmetric stretching of CF2 group, symmetric stretching of CF2 group, SO group, and CFRCF3 group, respectively 29 . These peaks are also observed in Si-N with much lower intensity, indicating that very thin layer of Nafion is coated onto Si particles, which is consistent with the TEM observation in Figure 3.4d. 55 Figure 3.4 Characterization of porous Si and Nafion-coated porous Si (Si-N). (a,b) TEM images of a typical porous Si particle at different magnifications. The pores are uniformly distributed throughout the particle, with pore size of 10-15 nm. (c,d) TEM images of Si-N at different magnifications. The porous structure is not as clear as it shows in (a), mainly due to the filling of Nafion into the pores. A thin layer of Nafion can be found in (d) as indicated by the dotted line. (e) Another TEM image of Si-N particle. (f-h) Energy filtered TEM images of (e) to map out the distribution of Si (f), Nafion (g), and their superposition (h). Figure 3.5 FTIR spectra of Si, Si-N, and Nafion. 56 Figure 3.6a and b show the SEM and TEM images of S-C-G. From Figure 3.6a and b, carbon nanofiber and graphene are clearly identified. The small particles are carbon black, and the bright region in SEM (Figure 3.6a) and the dark region in TEM (Figure 3.6b) indicate the presence of sulfur. Figure 3.6c shows the cyclic performance of Li-S half cell in the voltage window of 1.7-2.7 V (vs. Li/Li + ) at current density of 0.1 C using S-C-G as working electrode. Figure 3.6d shows the charge-discharge curves of Li-S half cell using S-C-G as working electrode. The electrolyte is LITFSI electrolyte. Figure 3.6 (a) SEM and (b) TEM images of S-C-G. (c) Cyclic performance and (d) Charge-discharge curves of S-C-G. 57 Figure 3.7a and b show the SEM and TEM images of Si-C-N-G. From the images, it is clear to observe that the Si particles are well wrapped with graphene layers. Figure 3.7c shows the cyclic performance of Li-Si half cell in the voltage window of 0.01-2 V (vs. Li/Li + ) at current density of 400 mA/g using Si-C-N-G as working electrode. Figure 3.7d shows the charge-discharge curves of Li-Si half cell using Si-C-N-G as working electrode. The electrolyte is LITFSI electrolyte. Figure 3.7 (a) SEM and (b) TEM images of Si-C-N-G. (c) Cyclic performance and (d) Charge-discharge curves of Si-C-N-G. 58 3.5 Electrochemical test and discussion Before assembling Si into full cells, galvanostatic charge-discharge test of porous Si particles with and without Nafion coating were first conducted in Li-Si half-cell configuration in the voltage window of 0.01-2 V (vs. Li/Li + ), and the results are shown in Figure 3.8a. Here, polysulfides electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 1 M Li2S4) is used as electrolyte to test the functionality of Nafion and the current density is 400 mA/g. Despite a gradual capacity loss, capacity of Nafion-coated porous Si (Si-N) remains 670 mAh/g after 100 cycles. On the contrary, the cell using bare Si particles drops to almost zero capacity within 20 cycles. The fast capacity degradation results from side reaction between LixSi anode and polysulfides, as denoted in Equation (2). In Figure 3.8b, full cells were fabricated to further demonstrate the functionality of Nafion. The anode materials are porous Si with carbon coating (Si-C), and porous Si with carbon coating and Nafion coating (Si-C-N), respectively. Here the porous Si particles were first coated with carbon, because conventionally, a thin layer of carbon coating is found to be beneficial to improve the Coulombic efficiency of Li-Si half cells, as the carbon layer is helpful to form stable SEI layer and reduce side reaction of Si with electrolyte. Before assembling into full cells, the prepared Si anodes were first charged and discharged for 5 cycles in Li-Si half cells. After cycling, Si electrodes were disassembled at lithiated state, and then washed carefully with DME/DOL. After drying in Ar-protected environment, the Si electrodes were assembled into full cells with S-C-G cathode using LITFSI electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO3). We note that 59 there are some recent reports on stable LixSi in air, which may facilitate the pre-lithiation process for Si-S full cells 30-31 . The galvanostatic charge-discharge test of full cells was conducted in the voltage window of 1.2-2.7 V (vs. Li/Li + ) at a current density of 0.1 C (1 C=1600 mA/g based on the mass of sulfur) and the capacity is calculated based on the mass of sulfur. As shown in Figure 3.8b, it is clear to notice the improved cyclability when Nafion is coated. After 200 cycles, the full cell with Si-C anode shows a fast capacity drop to 80 mAh/g. On the contrary, the capacity of the full cell with Si-C-N anode is well above that of the full cell with Si-C anode during 200 cycles. After 200 cycles, the full cell with Si-C-N anode still retains capacity of 170 mAh/g. In addition, Coulombic efficiency of Si-C-N full cell is much higher than that of Si-C full cell, indicating that Nafion layer can effectively reduce the polysulfides shuttle effect. We note that the Coulombic efficiency is higher than 100% for the initial several cycles. Our observation is consistent with previous reports 32-33 , and the mechanism is that in initial several cycles of Li-S half cells, the utilization of sulfur in cathode increases gradually during discharge process as the sulfur in the cathode gradually becomes exposed to the electrolyte. Upon charge process, however, only long chain polysulfides are formed, leading to lower charge capacity than discharge capacity and therefore the Coulombic efficiency is higher than 100%. This explanation can also be applied for our Si(lithiated)-S full cells, which explains why the Coulombic efficiency of the initial several cycles in Figure 3.8b is higher than 100%. 60 Figure 3.8 Electrochemical performance of Li-Si half cells and Si-S full cells to demonstrate the functionality of Nafion. (a) Comparison of Li-Si half cells using bare porous Si particles and Nafion-coated porous Si particles (Si-N) as working electrode, respectively. The galvanostatic charge-discharge test was conducted in the voltage window of 0.01-2 V at a current density of 400 mA/g and the electrolyte is polysulfides electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 1 M Li2S4). (b) Comparison of Si-S full cells using lithiated carbon-coated porous Si (Si-C) and lithiated carbon-coated porous Si with Nafion coating (Si-C-N) as anode, respectively. The Si-based anodes were first cycled in Li-Si half cells and then disassembled at lithiated state before coupling with S-C-G cathodes to assemble full cells. The galvanostatic charge-discharge test of full cells was conducted in the voltage window of 1.2-2.7 V at a current density of 0.1 C (1 C=1600 mA/g based on the mass of sulfur) and the electrolyte is LITFSI electrolyte (1 M LITFSI in DME/DOL, 1:1 by volume, with addition of 5% LiNO3). The specific capacity is calculated based on the mass of sulfur. Charge capacity: solid circles and solid squares; discharge capacity: hollow circles and hollow squares. To further improve the full-cell performance, Nafion-coated Si particles were wrapped with graphene (Si-C-N-G) because graphene can improve the overall electric conductivity of the electrode and it is also helpful to hold particles together without losing their electric contacts during cycling. Figure 3.9 shows the electrochemical performance of the Si-S full cell with LITFSI electrolyte using lithiated Si-C-N-G as anode and S-C-G as cathode. The Si-S full cell is cycled in the voltage window of 1.2-2.7 V and the specific capacity is calculated based on the mass of sulfur. Figure 3.9a shows the cyclic performance at a current density of 0.1 C. After 100 cycles, the charge capacity is 610 mAh/g, which is 80% 61 of its initial capacity, and the Coulombic efficiency is maintained at 92%. The Si-S full cell is determined to have two discharge voltage plateaus of 2.0 V and 1.7 V, as shown in Figure 3.9b, which corresponds well with the difference of voltage plateaus in Li-S (two voltage plateaus at 2.4 and 2.1 V (vs. Li/Li + )) and Li-Si (voltage plateau at approximately 0.4 V (vs. Li/Li + )) half cells, as illustrated in Figure 3.6d and Figure 3.7d. We note that the sloped curves of both charge and discharge branches in Figure 3.9b are mainly due to the gradually changed voltage profile of amorphous Si anode (Figure 3.7d). Figure 3.9c shows the CV curves of Si-S full cell under different cycles. From the CV curves, it is interesting to note that the anodic peak gradually shifts to low potential when the cell is under repeated cycling. This shift is mainly due to the decrease of electron and ion diffusion resistance of the cell, as indicated by the impedance tests shown in Figure 3.10. However, in the cathodic branch, we note that the high voltage peak (1.9-2.0 V) also shifts to low potential along with cycling, suggesting the gradual dissolution of elemental sulfur and transformation from high-order polysulfides (e.g. Li2S8) to relative lower-order polysulfides (e.g. Li2S6). In contrast, the low voltage peak (~1.6 V), which is mostly related to reactions among solid sulfide species (e.g. Li2S and Li2S2), does not change peak position significantly 9, 34 . Figure 3.9d shows the cyclic performance of Si-S full cell tested at current rates of 0.1 C, 0.2 C, 0.4 C, and 0.8 C for 10 cycles each. The capacities are stabilized at about 750 mAh/g, 600 mAh/g, 500 mAh/g, and 400 mAh/g, respectively. After switching back to 0.2 C, the capacity retains 550 mAh/g, which implies the good stability of Si-S full cell under different operation rates. 62 Figure 3.9 Electrochemical performance of Si-S full cell with LITFSI electrolyte using lithiated Si-C-N-G as anode and S-C-G as cathode. The Si-S full cell is cycled in the voltage window of 1.2-2.7 V and the specific capacity is calculated based on the mass of sulfur. (a) Cyclic performance of Si-S full cell at a current density of 0.1 C. (b) Charge-discharge curves of Si-S full cell at different cycles. (c) Cyclic voltammetry curves of Si-S full cell at different cycles. The test is conducted at the scan rate of 0.2 mV/s in the voltage window of 1.2-2.7 V. (d) Cyclic performance of Si-S full cell at different current rates. To further elucidate the functionality of Nafion, we have conducted impedance measurements on both Si-C-G and Si-C-N-G in Li-Si half-cell configuration. The EIS were collected with an AC voltage of 5 mV amplitude in the frequency range of 1000 kHz to 10 mHz. Figure 3.10a shows the Nyquist plots of the cells after being rested for 24 hours. 63 Conventionally, a typical Nyquist plot is composed of two semicircles at high frequency region and one linear diffusion drift at low frequency region. The high frequency semicircles are related to interfacial impedance due to the formation of SEI 35 and interphase electronic contact resistance (charge transfer resistance Rct) 36 . However, in Figure 3.10a, only the semicircle corresponding to Rct is observed in the high frequency region (>1 kHz) for both Si-C-G and Si-C-N-G before cycling, which is attributed to the negligible formation of SEI layer 37 . The value of Rct can be read from the intercept of the semicircle with the Z-real axis. From Figure 3.10a, it is observed that the Nafion-coated sample (Si-C-N-G) has slightly larger Rct than that of the sample without Nafion coating (Si-C-G). It is mainly due to the additional resistance from Nafion coating. However, as the Nafion coating is rather thin (Figure 3.4d), the resistance increase is not significant. Figure 3.10b shows the Nyquist plots of both electrodes after 5 charge-discharge cycles (electrodes are at delithiated state for EIS measurements). For Si-C-G electrode, a new semicircle appears in the high-frequency region of the Nyquist plot, indicating the formation of thick SEI layer. Moreover, the radius of the semicircle corresponding to Rct becomes larger compared with that before cycling, suggesting larger charge transfer resistance of the electrode due to poor electric contact between Si-C-G particles 38 . It is supposed to result from the increased electric resistance of a thick and insulating SEI layer formed on Si-C-G particle surface. On the contrary, the Si-C-N-G electrode still shows one semicircle and its radius is smaller compared with its pristine states before cycling, indicating reduced Rct of the electrode after cycling. This is because the Nafion coating on Si in Si-C-N-G electrode can prevent direct formation of SEI on Si surface. In addition, the loose structure of Nafion polymer provides 3-dimentional interconnected channels, which can significantly facilitate Li + ion 64 diffusion from electrolyte to Si. In conclusion, Nafion coating not only helps to prevent Si (and lithiated Si) from reacting with polysulfides, but also reduces the electrode-electrolyte interface resistance, which contributes to good electrochemical performance of the Si-S full cell. Figure 3.10 Nyquist plots of Si-C-G (red curve) and Si-C-N-G (blue curve) in Li-Si half-cell configuration before cycling (a) and after 5 cycles (b). To make battery practical, the amount of cathode and anode loading is critical to achieve high overall specific capacity and energy density. Unbalanced loading would significantly lower the specific capacity and increase the production cost because of the waste of excessive amount of cathode or anode materials. Currently, Li-S battery has attracted great attention due to its high capacity. However, it is important to note that excessive amount of Li is used in the cell, which not only lowers the Li-S cell capacity if we take into account the weight of Li, but also brings special concern on battery safety. Recently, great progress has been made to achieve safe Li anode by coating a protective layer on Li, such as single layer boron nitride 39 and hollow carbon layer 40 . Other strategies include electrodeposition 65 of Li on porous substrate 41 , or use specialized electrolyte additive such as LiF and LiBr to suppress the lithium dendrite formation 42 . However, previously reported approaches are still far from practical usage, and need further investigation on precise control of lithium used for Li-S battery, as well as side effects from the additives. For example, F - is found detrimental to most of cathode materials, and the effect from Br2 precipitation during cell operation is not clear at this moment 42 . In this context, we believe using lithiated Si as anode has great potential and deserves more research effort, because lithiated Si is safe (there is no dendrite formation) and it is easy to control the amount of loading. Here, we have investigated the battery performance of Si-S full cells with different mass ratios between S-C-G and Si-C-N-G. Figure 3.11a shows the cyclic performance of Si-S full cells with S:Si mass ratio of 0.33, 1.43, 2.22, and 4.24. The galvanostatic charge-discharge test was conducted in the voltage window of 1.2-2.7 V at a current density of 0.1 C and the specific discharge capacity is calculated based on the mass of sulfur. We find that in case of large S loading (S:Si = 4.24), the capacity drops from 180 mAh/g to almost 0 after 100 cycles, the low capacity is mainly due to the insufficient lithium provided by the small amount of lithiated Si anode. By decreasing sulfur loading to S:Si=2.22, the initial capacity increases to 690 mAh/g, and the cell is reasonably stable as the capacity is retained at 280 mAh/g after 100 cycles. Further decreasing the S loading to S:Si=1.43 raises the initial capacity up to 1000 mAh/g, and the capacity stays above 560 mAh/g for 100 cycles, which follows a trend similar to S-C-G cathode in Li-S half-cell configuration (Figure 3.6c). When S loading further decreases to S:Si=0.33, however, the full cell capacity decreases. We believe this is due to the small amount of sulfur used in the 66 cell, and dissolution of sulfur into electrolyte leads to loss of active sulfur in the cathode. We note that it is possible to decrease the amount of electrolyte used in battery to minimize the inevitable sulfur dissolution; however, certain amount of electrolyte is required to wet both cathode and anode. Figure 3.11b summarizes the specific discharge capacities of the full cells after 100 cycles calculated based on mass of S and mass of S+Si, respectively. The highest capacity achieved in our test is 330 mAh/g when the total mass of S and Si is considered. For comparison, the theoretical capacity of graphite-LiCoO2 full cell at different mass ratios is demonstrated in Figure 3.11b as black curve. If we assume LiCoO2 has a capacity of 150 mAh/g and graphite has a capacity of 360 mAh/g, the highest theoretical capacity of graphite-LiCoO2 full cell is 106 mAh/g when the mass ratio of LiCoO2 to graphite is 2.4. The 330 mAh/g achieved in our test is more than three times higher than the highest theoretical capacity of graphite-LiCoO2 full cell. If we use an average operation voltage of 1.8 V for Si(lithiated)-S full cell (Figure 4b), the estimated energy density is 590 Wh/kg. For graphite-LiCoO2 full cell, if we use an average operation voltage of 3.9 V, the highest theoretical energy density is 410 Wh/kg. However, the actual energy density achieved in industry is ~250 Wh/kg according to literature 27, 43 . The energy density of 590 Wh/kg achieved by our work is therefore more than 2 times higher than that of commercially available lithium-ion batteries and 43% higher than that of the highest theoretical energy density of graphite-LiCoO2 full cell. 67 Figure 3.11 Evaluation of cyclic performance of Si-S full cell with different S:Si mass loading ratios. (a) Cyclic performance of Si-S full cell at S:Si mass loading ratio of 0.33, 1.43, 2.22, and 4.24. The galvanostatic charge-discharge test was conducted in the voltage window of 1.2-2.7 V at a current density of 0.1 C and the specific discharge capacity is calculated based on the mass of sulfur. (b) Calculated specific discharge capacity of Si-S full cell after 100 cycles at different S:Si mass ratios, based on the mass of S only (red curve) and mass of S+Si (blue curve). For comparison, the theoretical capacity of graphite-LiCoO2 full cell at different mass ratios is demonstrated as black curve. 3.6 Summary In summary, we have addressed the critical issue in the implantation of Si-S full battery. Spontaneous reaction between Si (and lithiated Si) and dissolved polysulfides causes significant shuttle effect, which leads to severe capacity degradation. Nafion coating provides an effective way to shield Si from direct contact with polysulfides, and thus diminish the undesirable side reaction between Si (and lithiated Si) and polysulfides. We have demonstrated the Si-S full cell using lithiated Nafion-coated porous Si as anode and S as cathode. With optimized mass loading ratio of S to Si, high capacity of the full cell has been achieved. The capacity is 560 mAh/g based on the mass of sulfur, and 330 mAh/g based on the total mass of S and Si after 100 cycles. The estimated energy density of the Si-S full cell is 590 Wh/kg, which is more than 2 times higher than that of commercially 68 available lithium-ion batteries. We believe the reported various issues involved in Si-S full cell and the approach we have taken to address the issues can open up the door to further optimization of the cell, and lead to a significant step towards the design of new generation of batteries by taking advantages of high-capacity anode and cathode. 69 3.7 References 1. 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). 2. Armstrong, A. R.; Bruce, P. G. Synthesis of layered LiMnO2 as an electrode for rechargeable lithium batteries. Nature 381, 499-500 (1996). 3. Tarascon, J. M.; Armand, M. Issues and challenges facing rechargeable lithium batteries. Nature 414, 359-367 (2001). 4. Whittingham, M. S. Lithium batteries and cathode materials. Chem. Rev. 104, 4271- 4301 (2004). 5. Padhi, A. K.; Nanjundaswamy, K. S.; Goodenough, J. B. Phospho-olivines as positive- electrode materials for rechargeable lithium batteries. J. Electrochem. Soc. 144, 1188-1194 (1997). 6. Yang, Y.; McDowell, M. T.; Jackson, A.; Cha, J. J.; Hong, S. S.; Cui, Y. New Nanostructured Li2S/Silicon Rechargeable Battery with High Specific Energy. Nano Lett. 10, 1486-1491 (2010). 7. 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). 8. Mikhaylik, Y. V.; Akridge, J. R. Polysulfide shuttle study in the Li/S battery system. J. Electrochem. Soc. 151, A1969-A1976 (2004). 9. 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). 10. Zheng, G. Y.; Yang, Y.; Cha, J. J.; Hong, S. S.; Cui, Y. Hollow Carbon Nanofiber- Encapsulated Sulfur Cathodes for High Specific Capacity Rechargeable Lithium Batteries. Nano Lett. 11, 4462-4467 (2011). 11. Cheng, X. B.; Huang, J. Q.; Zhang, Q.; Peng, H. J.; Zhao, M. Q.; Wei, F. Aligned carbon nanotube/sulfur composite cathodes with high sulfur content for lithium-sulfur batteries. Nano Energy 4, 65-72 (2014). 12. Wang, H. L.; Yang, Y.; Liang, Y. Y.; Robinson, J. T.; Li, Y. G.; Jackson, A.; Cui, Y.; Dai, H. J. Graphene-Wrapped Sulfur Particles as a Rechargeable Lithium-Sulfur Battery Cathode Material with High Capacity and Cycling Stability. Nano Lett. 11, 2644-2647 (2011). 13. Seh, Z. W.; Li, W. Y.; Cha, J. J.; Zheng, G. Y.; Yang, Y.; McDowell, M. T.; Hsu, P. C.; Cui, Y. fSulphur-TiO2 yolk-shell nanoarchitecture with internal void space for long- cycle lithium-sulphur batteries. Nat. Commun. 4, (2013). 14. Liang, X.; Hart, C.; Pang, Q.; Garsuch, A.; Weiss, T.; Nazar, L. F. A highly efficient polysulfide mediator for lithium-sulfur batteries. Nat. Commun. 6, (2015). 15. Qiu, Y. C.; Li, W. F.; Zhao, W.; Li, G. Z.; Hou, Y.; Liu, M. N.; Zhou, L. S.; Ye, F. M.; Li, H. F.; Wei, Z. H.; Yang, S. H.; Duan, W. H.; Ye, Y. F.; Guo, J. H.; Zhang, Y. G. High- Rate, Ultra long Cycle-Life Lithium/Sulfur Batteries Enabled by Nitrogen-Doped Graphene. Nano Lett. 14, 4821-4827 (2014). 16. Yim, T.; Park, M. S.; Yu, J. S.; Kim, K. J.; Im, K. Y.; Kim, J. H.; Jeong, G.; Jo, Y. N.; Woo, S. G.; Kang, K. S.; Lee, I.; Kim, Y. J. Effect of chemical reactivity of polysulfide toward carbonate-based electrolyte on the electrochemical performance of Li-S batteries. Electrochim. Acta 107, 454-460 (2013). 70 17. Kreuer, K. D. On the development of proton conducting polymer membranes for hydrogen and methanol fuel cells. J. Membr. Sci. 185, 29-39 (2001). 18. Diao, Y.; Xie, K.; Xiong, S. Z.; Hong, X. B. Insights into Li-S Battery Cathode Capacity Fading Mechanisms: Irreversible Oxidation of Active Mass during Cycling. J. Electrochem. Soc. 159, A1816-A1821 (2012). 19. Arreaga-Salas, D. E.; Sra, A. K.; Roodenko, K.; Chabal, Y. J.; Hinkle, C. L. Progression of Solid Electrolyte Interphase Formation on Hydrogenated Amorphous Silicon Anodes for Lithium-Ion Batteries. J. Phys. Chem. C 116, 9072-9077 (2012). 20. Demir-Cakan, R.; Morcrette, M.; Gangulibabu; Gueguen, A.; Dedryvere, R.; Tarascon, J. M. Li-S batteries: simple approaches for superior performance. Energy Environ. Sci. 6, 176-182 (2013). 21. Aurbach, D.; Pollak, E.; Elazari, R.; Salitra, G.; Kelley, C. S.; Affinito, J. On the Surface Chemical Aspects of Very High Energy Density, Rechargeable Li-Sulfur Batteries. J. Electrochem. Soc. 156, A694-A702 (2009). 22. Dalchiele, E. A.; Aurora, A.; Bernardini, G.; Cattaruzza, F.; Flamini, A.; Pallavicini, P.; Zanoni, R.; Decker, F. XPS and electrochemical studies of ferrocene derivatives anchored on n- and p-Si(100) by Si-O or Si-C bonds. J. Electroanal. Chem. 579, 133-142 (2005). 23. Radvanyi, E.; De Vito, E.; Porcher, W.; Larbi, S. J. S. An XPS/AES comparative study of the surface behaviour of nano-silicon anodes for Li-ion batteries. J. Anal. At. Spectrom. 29, 1120-1131 (2014). 24. Xia, F.; Kwon, S.; Lee, W. W.; Liu, Z.; Kim, S.; Song, T.; Choi, K. J.; Paik, U.; Park, W. I. Graphene as an Interfacial Layer for Improving Cycling Performance of Si Nanowires in Lithium-Ion Batteries. Nano Lett. 15, 6658-6664 (2015). 25. Ge, M. Y.; Lu, Y. H.; Ercius, P.; Rong, J. P.; Fang, X.; Mecklenburg, M.; Zhou, C. W. Large-Scale Fabrication, 3D Tomography, and Lithium-Ion Battery Application of Porous Silicon. Nano Lett. 14, 261-268 (2014). 26. Marcano, D. C.; Kosynkin, D. V.; Berlin, J. M.; Sinitskii, A.; Sun, Z. Z.; Slesarev, A.; Alemany, L. B.; Lu, W.; Tour, J. M. Improved Synthesis of Graphene Oxide. ACS Nano 4, 4806-4814 (2010). 27. Lee, S. K.; Oh, S. M.; Park, E.; Scrosati, B.; Hassoun, J.; Park, M. S.; Kim, Y. J.; Kim, H.; Belharouak, I.; Sun, Y. K. Highly Cyclable Lithium-Sulfur Batteries with a Dual-Type Sulfur Cathode and a Lithiated Si/SiOx Nanosphere Anode. Nano Lett. 15, 2863-2868 (2015). 28. Agostini, M.; Scrosati, B.; Hassoun, J. An Advanced Lithium-Ion Sulfur Battery for High Energy Storage. Adv. Energy Mater. 5, (2015). 29. Liang, Z. X.; Chen, W. M.; Liu, J. G.; Wang, S. L.; Zhou, Z. H.; Li, W. Z.; Sun, G. Q.; Xin, Q. FT-IR study of the microstructure of Nafion((R)) membrane. J. Membr. Sci. 233, 39-44 (2004). 30. Zhao, J.; Lu, Z. D.; Liu, N. A.; Lee, H. W.; McDowell, M. T.; Cui, Y. Dry-air-stable lithium silicide-lithium oxide core-shell nanoparticles as high-capacity prelithiation reagents. Nat. Commun. 5, (2014). 31. Zhao, J.; Lu, Z. D.; Wang, H. T.; Liu, W.; Lee, H. W.; Yan, K.; Zhuo, D.; Lin, D. C.; Liu, N.; Cui, Y. Artificial Solid Electrolyte Interphase-Protected LixSi Nanoparticles: An Efficient and Stable Prelithiation Reagent for Lithium-Ion Batteries. J. Am. Chem. Soc. 137, 8372-8375 (2015). 71 32. Elazari, R.; Salitra, G.; Garsuch, A.; Panchenko, A.; Aurbach, D. Sulfur-Impregnated Activated Carbon Fiber Cloth as a Binder-Free Cathode for Rechargeable Li-S Batteries. Adv. Mater. 23, 5641-5644 (2011). 33. Su, Y. S.; Manthiram, A. Lithium-sulphur batteries with a microporous carbon paper as a bifunctional interlayer. Nat. Commun. 3, (2012). 34. Bruce, P. G.; Freunberger, S. A.; Hardwick, L. J.; Tarascon, J. M. Li-O2 and Li-S batteries with high energy storage. Nat. Mater. 11, 19-29 (2012). 35. Barsoukov, E.; Kim, J. H.; Kim, J. H.; Yoon, C. O.; Lee, H. Effect of low-temperature conditions on passive layer growth on Li intercalation materials-In situ impedance study. J. Electrochem. Soc. 145, 2711-2717 (1998). 36. Guo, J. C.; Sun, A.; Chen, X. L.; Wang, C. S.; Manivannan, A. Cyclability study of silicon-carbon composite anodes for lithium-ion batteries using electrochemical impedance spectroscopy. Electrochim. Acta 56, 3981-3987 (2011). 37. Ogumi, Z.; Abe, T.; Fukutsuka, T.; Yamate, S.; Iriyama, Y. Lithium-ion transfer at interface between carbonaceous thin film electrode/electrolyte. J. Power Sources 127, 72- 75 (2004). 38. Gaberscek, M.; Moskon, J.; Erjavec, B.; Dominko, R.; Jamnik, J. The importance of interphase contacts in Li ion electrodes: The meaning of the high-frequency impedance arc. Electrochem. Solid State Lett. 11, A170-A174 (2008). 39. Yan, K.; Lee, H. W.; Gao, T.; Zheng, G. Y.; Yao, H. B.; Wang, H. T.; Lu, Z. D.; Zhou, Y.; Liang, Z.; Liu, Z. F.; Chu, S.; Cui, Y. Ultrathin Two-Dimensional Atomic Crystals as Stable Interfacial Layer for Improvement of Lithium Metal Anode. Nano Lett. 14, 6016- 6022 (2014). 40. Zheng, G. Y.; Lee, S. W.; Liang, Z.; Lee, H. W.; Yan, K.; Yao, H. B.; Wang, H. T.; Li, W. Y.; Chu, S.; Cui, Y. Interconnected hollow carbon nanospheres for stable lithium metal anodes. Nat. Nanotechnol. 9, 618-623 (2014). 41. Liang, Z.; Zheng, G. Y.; Liu, C.; Liu, N.; Li, W. Y.; Yan, K.; Yao, H. B.; Hsu, P. C.; Chu, S.; Cui, Y. Polymer Nanofiber-Guided Uniform Lithium Deposition for Battery Electrodes. Nano Lett. 15, 2910-2916 (2015). 42. Lu, Y. Y.; Tu, Z. Y.; Archer, L. A. Stable lithium electrodeposition in liquid and nanoporous solid electrolytes. Nat. Mater. 13, 961-969 (2014). 43. Van Noorden, R. A Better Battery. Nature 507, 26-28 (2014). 72 Chapter 4: Synthesis of hierarchical carbon-coated ball-milled silicon and its applications in lightweight free-standing electrodes and high-voltage full cells 4.1 Introduction A variety of Si nanostructures have been developed to accommodate the large volume change of Si during charge-discharge processes. However, since cost per watt-hour ($/Wh) is a crucial criterion in battery industry, the sophisticated synthesis process of Si nanostructures increases the cost of Si-based batteries and thus hinders Si from real applications. Another challenge to achieve real applications of Si anode is to develop Si-based full cells with low cost per watt-hour. As the energy density of batteries can be calculated from ∫ 𝑉 (𝑞 )𝑑𝑞 /𝑤𝑡 𝑄 0 , the improvement of energy density can be accomplished by either increasing voltage, increasing capacity, or reducing weight 1 . Since the working potential of Si anode has almost reached the working potential of lithium metal, using a high-voltage cathode is an effective way to increase the working voltage of Si-based full cells, which can lead to higher energy density 2 . LiNi0.5Mn1.5O4 is one of the most promising high- voltage cathode materials due to the flat potential profile at high voltage and high theoretical capacity 3-5 . Compared with the working voltage of 4.0 V and 3.4 V for commercial LiCoO2 and LiFePO4, the high working voltage of 4.7 V and theoretical capacity of 146.7 mAh/g of LiNi0.5Mn1.5O4 can theoretically provide 20% and 30% higher energy density than LiCoO2 and LiFePO4 for the full cells, respectively 6-7 . 73 To further enhance the energy density, reducing the weight of the electrodes can be an effective and essential strategy. Conventional electrodes are prepared in a slurry-casting method by which the polymer binder, conductive additive, and active material are cast onto metal foils. The non-electrochemically active binders and current collector can account for a large mass fraction of the electrode, which dramatically reduces the overall energy density of the battery. Recently, conductive carbon matrices such as carbon nanotubes and graphene have been reported to replace metal foils as lightweight current collectors 8-10 . To fabricate Si-based free-standing batteries with carbon matrices, the major challenge is to maintain good contact between Si and the conductive matrix because the volume expansion rate of Si during cycling process is up to 300%, which is much larger than most of other battery materials. Usually bottom-up synthesis methods of Si such as silane chemical vapor deposition (CVD) lead to better contact between Si and the matrix 11 . However, the high- temperature pyrolysis of toxic and expensive silane precursor (∼$50,000/ton) increases the cost of the electrodes and thus hinders it from real applications. For Si produced by top- down synthesis methods, the poor physical contact between Si and carbon matrix together with the large volume change of Si during cycling process result in poor cycling performance of the batteries. To solve this problem, non-electrochemically active binders are usually added 12 , which adds extra weight to electrodes and reduces the conductivity. Taking into account the above-mentioned challenges, we made use of low-cost Si available from the semiconductor industry, such as metallurgical Si ($1,000/ton), polycrystalline Si ($1,500/ton), and waste Si wafers as starting materials. After ball-milling the Si sources, 74 thin carbon shells were created around the particles, leaving void space between Si particles and carbon shells to allow for volume change of Si during charge-discharge process, thus maintaining good contact between Si and the current collector through the carbon shells. More importantly, post-cycling characterization reveals in situ formation of porous Si inside the carbon shell during the cycling process. The hierarchical carbon-coated (HCC) structure helps to make full use of excellent electrochemical performance of the porous Si. The obtained hierarchical carbon-coated metallurgical Si (HCC-M-Si), polycrystalline Si (HCC-P-Si), and waste Si wafers (HCC-W-Si) all achieve over 1000 mAh/g after 1000 cycles at a current density of 2 A/g. Taking advantage of the good contact between the HCC Si and current collector, lightweight, binder-free, and free-standing electrodes consisting of HCC-M-Si and low-cost commercially available carbon nanofibers (CNFs) were prepared. At a current density of 0.4 A/g, the HCC-M-Si/CNF electrode achieves over 1000 mAh/g after 100 cycles based on total mass of the electrode. The lightweight and free-standing HCC-M-Si/CNF electrodes significantly improve the energy density of the electrodes by 745% compared with conventional electrodes. Moreover, full cells based on HCC-M-Si were studied using both commercial LiCoO2 and synthesized high-voltage cathode LiNi0.5Mn1.5O4 as cathodes. The HCC-M-Si/LiNi0.5Mn1.5O4 full cells deliver a high average working voltage of 4.4 V, which helps to increase the energy density by 46% compared with HCC-M-Si/LiCoO2 full cells. 4.2 Synthesis of hierarchical carbon-coated ball-milled silicon To achieve large-scale and low-cost fabrication of Si anode, ball-milling was applied to metallurgical Si, polycrystalline Si, and waste Si wafers, which was operated at grinding 75 speed of 1200 rpm for 5 h. The obtained samples are denoted as M-Si, P-Si, and W-Si, respectively. The obtained powders were washed in 5 M hydrofluoric acid (HF) and deionized water (DI-H2O) successively to remove surface oxide layer, and dried at 60 ℃ in air for 12 h for further use. To synthesize HCC-M-Si, M-Si was first coated with SiO2. Specifically, 400 mg M-Si was dispersed in 400 mL DI-H2O followed by ultrasonication for 30 min. 4 mL ammonium hydroxide and 3.2 g tetraethoxysilane was then added to the solution under continuous stirring. After reaction for 24 h, the particles were centrifuged at 9000 rpm for 8 min, washed with DI-H2O and ethanol, and dried at 60 ℃ in air for 12 h. 200 mg of the obtained particles are dispersed in 100 mL DI-H2O followed by addition of 10 mg trimethylammonium bromide and 0.2 mL ammonium hydroxide. After stirring for 1 h, 40 mg resorcinol and 50 µL formaldehyde solution were added to the solution. The reaction was left at room temperature under stirring for 12 h. The particles were then centrifuged at 9000 rpm for 8 min and washed with ethanol. After being dried at 60 ℃ in air for 12 h, the particles were annealed at 800 ℃ in Ar atmosphere for 2 h so that the resin coating on the particles is carbonized. To create the void space between M-Si and carbon coating, the obtained particles were dispersed in 5 M HF for 30 min to remove the SiO2 interlayer. The particles were then washed with DI-H2O and ethanol, and dried at 60 ℃ in air for 12 h. To synthesize HCC-M-Si(2), same procedure was followed except that 1.6 g tetraethoxysilane was used to obtain thinner void space between M-Si and carbon coating. To synthesize HCC-P-Si and HCC-W-Si, same procedure was followed except that M-Si was replaced by P-Si and W-Si. 76 Figure 4.1a and b are the scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of the M-Si particles, which show broad particle size distribution and the morphology of the particles are irregular. The broad size distribution can help efficient occupation of the electrodes by filling interstitial space with smaller Si particles. The high-resolution TEM (HRTEM) image and the inset selected area electron diffraction (SAED) pattern in Figure 4.1c indicate the polycrystalline nature of the M-Si. The TEM image of P-Si is shown in Figure 4.2a, which indicates that its particle size is similar to M-Si. The HRTEM image and corresponding SAED pattern in Figure 4.2b indicate the polycrystalline nature of the P-Si. Similarly, the particle size and crystallinity of W-Si are also similar to M-Si as shown in Figure 4.2c and d. X-ray diffraction (XRD) patterns of metallurgical Si, polycrystalline Si, and waste Si wafer before and after ball- milling are shown in Figure 4.3a and b. The main diffraction peaks of (111), (220), (311), (400), and (331) planes in the diffraction patterns of metallurgical Si and polycrystalline Si before and after ball-milling indicate their unchanged polycrystalline nature 13 . The broadening of the main peaks after ball-milling also indicates the shrinking of the particle size. On the contrary, the waste Si wafer transforms from single-crystalline before ball- milling to polycrystalline after ball-milling. This demonstrates that ball-milling works as an efficient way to unify different crystalline Si sources into polycrystalline particles, which makes it suitable for large-scale fabrication. 77 Figure 4.1 Morphological and elemental characterization of HCC-M-Si and HCC-M-Si(2). (a) SEM and (b) TEM image of M-Si. (c) HRTEM image of M-Si. The inset is the corresponding SAED pattern. (d, e) TEM images of HCC-M-Si. (f) TEM image of HCC- M-Si(2). (g) STEM image of HCC-M-Si particle. (h) EELS mapping of elements C and Si in the HCC-M-Si particle. (i) EELS spectrum of the HCC-M-Si particle. 78 Figure 4.2. Morphological characterization of P-Si and W-Si. (a) TEM image of P-Si. (b) HRTEM image of P-Si. The inset is the corresponding SAED pattern. (c) TEM image of W-Si. (d) HRTEM image of W-Si. The inset is the corresponding SAED pattern. 79 Figure 4.3. XRD patterns of metallurgical Si, polycrystalline Si, and Si wafer before (a) and after (b) ball-milling. HCC-M-Si was synthesized using obtained M-Si as stating material. Specifically, M-Si was first coated with SiO2 as hard template for the subsequent coating of resin. The resin was then carbonized after heat treatment. After etching away SiO2, void space was created between Si particle and the carbon shell to accommodate Si volume expansion, which guarantees stable electrode structure and therefore good contact between the Si active material and the current collector. For comparison, HCC-M-Si with smaller void space between Si particle and carbon shell was also synthesized and it is denoted as HCC-M- Si(2). The morphological characterizations of HCC-M-Si and HCC-M-Si(2) are shown in Figure 4.1d-i. Figure 4.1d is low-magnification TEM image of HCC-M-Si, which demonstrates uniform carbon shell coating around Si nanoparticles. The TEM images with higher magnification in Figure 4.1e and f compare morphology of HCC-M-Si and HCC- M-Si(2). The carbon shells in both samples coat the Si well, retaining similar shape as the Si particles inside. With thicker SiO2 coated around M-Si particles as hard template, the void space in HCC-M-Si is larger compared with that in HCC-M-Si(2). The scanning 80 transmission electron microscopy (STEM) image of HCC-M-Si and corresponding electron energy loss spectra (EELS) mapping of C (green color) and Si (red color) elements are shown in Figure 4.1g and h, respectively, which clearly confirms the HCC Si structure. Figure 4.1i shows the EELS spectrum, in which the C-K core-loss edge and Si-K core-loss edge are shown after appropriate background subtraction. A small O-K core-loss edge was also detected, which is from native oxide on the Si particle. 4.3 Electrochemical test and discussion To characterize the electrochemical performance of HCC-M-Si as LIB anode material, CR2032 coin cells were assembled with HCC-M-Si as working electrode and lithium foil as counter/reference electrode. Figure 4.4a shows the representative galvanostatic charge(lithiation)-discharge(delithiation) profiles of HCC-M-Si for the 1st, 2nd, 5th, 20th, 50th, and 100th cycle with a current density of 0.4 A/g between 0.01 and 1 V (vs. Li/Li + ). The cyclic voltammetry (CV) curves of HCC-M-Si at the scan rate of 0.2 mV/s in the voltage window of 0.01-1 V (vs. Li/Li + ) is shown in Figure 4.5. During the first charge process in Figure 4.4a, a long flat plateau at ~0.05 V is attributed to the reaction of crystalline Si with Li, which is consistent with the reductive peaks below 0.1 V in the CV plots 14 . In subsequent cycles, the voltage plateaus located at ~0.5 V during the discharge processes are also consistent with the oxidative peaks observed in the CV plots. The cyclic performances of HCC-M-Si and HCC-M-Si(2) were evaluated by galvanostatic charge- discharge measurements at a current density of 0.4 A/g between 0.01 V and 1 V vs. Li/Li + (Figure 4.4b). It can be seen that the reversible capacity of HCC-M-Si at the 1st cycle is 2773 mAh/g, which is retained at 2372 mAh/g after 100 cycles. It corresponds to a capacity 81 retention value of 85.5%. Meanwhile, the Coulombic efficiency of HCC-M-Si increases from 73% in the 1st cycle to over 98% in the 10th cycle and maintains a value over 99% in the 100th cycle, demonstrating a stable reversible electrochemical reaction. While for HCC-M-Si(2), the 1st reversible capacity is 2798 mAh/g, which is retained at 1928 mAh/g after 100 cycles, corresponding to a capacity retention value of only 68.9%. The higher capacity retention of HCC-M-Si than HCC-M-Si(2) is due to the larger void space in HCC- M-Si, which provides more space to accommodate the volume change of Si particles during charge-discharge processes. 82 Figure 4.4 Electrochemical performance of HCC-M-Si, HCC-P-Si, and HCC-W-Si. (a) Galvanostatic charge-discharge profiles of HCC-M-Si. (b) Cycling performance comparison of HCC-M-Si and HCC-M-Si(2) at a current density of 0.4 A/g. (c) Rate capability comparison of HCC-M-Si and M-Si. (d) Long-term cycling performance of HCC-M-Si, HCC-P-Si, and HCC-W-Si. The current density was 0.4 A/g for the first 10 cycles, and 2 A/g for the following cycles. The cycling performance of HCC-M-Si(2) and M-Si are provided for comparison with HCC-M-Si. 83 Figure 4.5 Cyclic voltammetry curves of HCC-M-Si. The rate capability of HCC-M-Si and M-Si were investigated under various current densities from 0.4 A/g to 8 A/g as shown in Figure 4.4c. Compared with HCC-M-Si, the M-Si anode showed much more rapid capacity decay with increasing rates. Specifically, at a current density of 0.4 A/g, 0.8 A/g, 2 A/g, and 4 A/g, HCC-M-Si delivers reversible capacities of 2900 mAh/g, 2700 mAh/g, 2200 mAh/g, and 1500 mAh/g, respectively. On the contrary, M-Si delivers only 2000 mAh/g, 1800 mAh/g, 1000 mAh/g, and 100 mAh/g, respectively. At a current density of 8 A/g, HCC-M-Si can still retain a reversible capacity of 950 mAh/g while M-Si exhibits almost no capacity. This suggests that the hierarchical carbon coating can effectively reduce the interparticle resistance of ball-milled Si and significantly boost the capacity at high current density. When the current density is reversed back to 0.8 A/g from 8 A/g, HCC-M-Si almost fully recovered to the initial capacity at 0.8 A/g, and maintained a better cycling stability than M-Si in the following cycles. The good conductivity of HCC-M-Si was further confirmed by the electrochemical impedance spectra (EIS) shown in Figure 4.6a and b, which demonstrate that HCC-M-Si 84 possesses both smaller interphase electronic contact resistance and charge transfer resistance than those of M-Si. Figure 4.6. Nyquist plots of HCC-M-Si and M-Si electrodes before cycling (a) and after 20 cycles (b). Long-term cycling performance of HCC-M-Si, HCC-P-Si, and HCC-W-Si are investigated in Figure 4.4d. The electrodes were run at 0.4 A/g for the first 10 cycles and 2 A/g for the following cycles. The cycling performance of HCC-M-Si(2) and M-Si are also provided to compare with that of HCC-M-Si. It is observed that HCC-M-Si maintains a better cycling stability than HCC-M-Si(2) over 1000 cycles, which is consistent with the cycling performance comparison at lower current density observed in Figure 4.4b. After 1000 cycles, HCC-M-Si retains a reversible capacity of 1031 mAh/g, while HCC-M-Si(2) retains only 540 mAh/g. For comparison, the M-Si exhibits poor cycling performance, where the capacity drops to almost zero after only 100 cycles. Long-term cycling test of HCC-P-Si and HCC-W-Si demonstrate they achieve cycling performance similar to that of HCC-M- Si, where over 1000 mAh/g is retained after 1000 cycles at a current density of 2 A/g. This 85 further demonstrates that the ball-milling process can unify different Si sources into similar quality and thus achieve similar electrochemical performance as LIB anodes. In addition, the hierarchical carbon coating further helps to maintain the intrinsic capacity of Si particles by confining the particles inside without losing contact with other particles, conductive additive, binder, and current collector. Figure 4.7 compares the electrochemical performance and the cost of Si anodes reported in this work and some recent literatures 15-33 . Different symbols represent the current densities and different symbol color represents the number of cycles reported. Among the five categories of synthesis methods compared in the figure, silane CVD is the only bottom- up synthesis method of Si. Due to the small dimension of the synthesized Si and good contact between the Si and the conductive matrix, silane CVD usually leads to excellent cycling performance. However, the high cost of silane hinders it from real application. Similarly, the other synthesis methods such as Si wafer etching, magnesiothermic reduction, and usage of commercial Si nanoparticles all lead to higher preparation cost of Si than M- Si. In addition, the findings reported in this paper that ball-milling can unify different low- cost Si into similar quality, and the hierarchical carbon coating can help the ball-milled Si achieve over 1000 mAh/g after 1000 cycles at a current density of 2 A/g, will lead to a solid step forward for large-scale production of low-cost and high-capacity Si anode materials. 86 Figure 4.7 Evaluation of Si anodes in terms of production cost, capacity, cycle number, and current density. The color scheme represents the cycle number; the symbol represents the current density; the number next to each symbol refers to the reference index. The merits of this HCC Si structure such as good contact and high conductivity depend on the structure integrity during the cycling process. In order to confirm the HCC-M-Si can maintain the hierarchical structure during cycling process, both HCC-M-Si and M-Si were characterized by TEM after cycling in the voltage window of 0.01-1 V (vs. Li/Li + ) at a current density of 0.4 A/g for certain cycles. As the conventional LIB prepared by slurry- casting method would bring difficulty to post-cycling characterization of carbon shell in HCC-M-Si due to the binder which also contains carbon, here we used a free-standing HCC-M-Si/CNF electrode, which will be discussed in detail later, for the characterization of HCC-M-Si after cycling. Figure 4.8a presents the schematic illustration of the morphology change of M-Si and HCC-M-Si after cycling. Figure 4.8b-d show the TEM images of M-Si after being charge-discharged for 1 cycle, 10 cycles, and 50 cycles, respectively. It is observed that nanopores form on the periphery of M-Si after being charge-discharged for only 1 cycle, as indicated by the dark/light contrast in Figure 4.8b. After 10 cycles, the surface of M-Si becomes even rougher (Figure 4.8c), and the M-Si 87 becomes totally porous after 50 cycles (Figure 4.8d). The pore evolution of M-Si during the cycling process is due to inelastic deformation of Li/Si during the lithiation/delithiation processes 34-35 . Compared with solid Si, the porous Si structure is more favorable because the large surface area could facilitate lithium ion diffusion and the thin walls between the pores could shorten the diffusion path for lithium ions and electrons, which leads to improved electrochemical performance 36 . However, here the porous structure is formed during the cycling process accompanied with significant volume expansion of particles, which would destruct the contact between Si particles and current collector as schematically shown in Figure 4.8a. The question now is whether the hierarchical carbon coating structure can help to confine the in situ formed porous Si and take advantage of its improved electrochemical performance. As shown in Figure 4.8e, the STEM image of HCC-M-Si after 50 cycles shows that the particle retains smooth surface. The EELS elemental mapping of C (green color) and Si (red color) in Figure 4.8f and EELS spectrum showing C-K core-loss edge and Si-L core-loss edge in Figure 4.8g further confirm that even though the in situ formed porous Si particle takes up the original void space between Si particle and carbon shell, the whole porous Si particle is still well confined within the carbon shell. In this way, the carbon shell can help the Si particle to maintain good electrical contact with other particles, conductive additive, binder, and therefore forms electron pathway to the current collector. As shown in Figure 4.9, which is a low-magnification STEM image of Figure 4.8e, the particle is actually well attached to a CNF. This directly proves that the HCC Si structure can help to confine the in situ formed porous Si particle and also maintain good electrical contact with the current collector. 88 Figure 4.8 Post-cycling analysis of HCC-M-Si. (a) Schematic illustration showing the morphology evolution of M-Si and HCC-M-Si after cycling. (b-d) TEM images of M-Si after 1 cycle (b), 10 cycles (c), and 50 cycles (d). (e) STEM image of HCC-M-Si particle after 50 cycles. (f) EELS mapping of elements C and Si in the HCC-M-Si particle. (g) EELS spectrum of the HCC-M-Si particle. 89 Figure 4.9. Low-magnification STEM image of HCC-M-Si after 50 cycles. 4.4 Lightweight and free-standing electrodes based on hierarchical carbon-coated ball-milled silicon To further improve the energy density of HCC-M-Si electrodes, lightweight and low-cost CNF network was used to replace copper foil current collector, polymer binder, and conductive additive used in conventional LIBs and free-standing HCC-M-Si/CNF electrode was fabricated by vacuum filtration method. The TEM images of the CNFs are shown in Figure 4.10a-c. The low-magnification TEM image (Figure 4.10a) and the high- magnification TEM image (Figure 4.10b) demonstrate the tubular structure of the CNFs with outer diameter of 100-200 nm and wall thickness of 20-30 nm. Due to high- temperature treatment, the CNFs exhibit highly graphitic nature as shown in Figure 4.10c, which is the enlarged image of the area marked by red square in Figure 4.10b. This graphitic nature guarantees good conductivity of the CNFs. To entrap the HCC-M-Si inside the CNF network instead of falling out from the surfaces of the electrodes, the top and 90 bottom surfaces of the HCC-M-Si/CNF electrodes are covered by extra thin layer of CNF network. The preparation process is schematically illustrated in Figure 4.11a and details of the preparation method can be found in the Methods section. A small amount of CNF was first filtrated to obtain a thin layer of CNF on the bottom, after which the HCC-M-Si/CNF mixture was filtrated on top of it and works as the main part of the electrode. Finally, another thin layer of CNF was filtrated on top of the electrode to fully encapsulate the HCC-M-Si/CNF mixture and the weight percentage of HCC-M-Si in the whole HCC-M- Si/CNF electrode is ~60%. The left side in Figure 4.11b is the photograph of the obtained HCC-M-Si/CNF electrode, which has been punched into circular shape for assembling into coin cells as shown in the right side of the figure. The SEM image of the surface of the electrode is shown in Figure 4.10d and it is observed that the surface is fully covered by CNFs. The length of the CNFs is over 100 µm and they entangle with each other to form a network, which helps to entrap the HCC-M-Si particles inside. 91 Figure 4.10 Characterization and electrochemical performance of CNFs. (a) Low- magnification TEM image of CNFs. (b) TEM image of the open end of one CNF. (c) HRTEM of the area which is marked by red square in (b). It shows the graphitic nature of the CNF. (d) SEM image of the surface of HCC-M-Si/CNF free-standing electrode. (e) Electrochemical performance of free-standing CNF electrode at a current density of 0.4 A/g. 92 Figure 4.11 Characterization and electrochemical performance of lightweight and free- standing HCC-M-Si/CNF electrode. (a) Schematic illustration of the fabrication process of free-standing HCC-M-Si/CNF electrode. (b) Photograph of the free-standing HCC-M- Si/CNF electrode and the assembled coin cell. (c) Cross-sectional SEM image of HCC-M- Si/CNF electrode and corresponding EDS mapping of elements C and Si. (d) SEM image of the area which is marked by red square in (c). It shows the inside mixture of HCC-M- Si/CNF electrode where HCC-M-Si particles are dispersed in CNF network. The inset in (d) shows the high-resolution SEM image of the area which is marked by green square. (e) Cycling performance comparison of free-standing HCC-M-Si/CNF and M-Si/CNF electrodes. 93 Figure 4.11c shows the cross-sectional SEM image of the free-standing HCC-M-Si/CNF electrode and corresponding energy dispersive X-ray spectroscopy (EDS) mappings of elements C (red color) and Si (yellow color). The EDS mappings of C and Si indicate that HCC-M-Si particles are uniformly distributed in the CNF network. As shown in Figure 4.11d and its inset, by zooming in a small area in the middle part of the cross-sectional SEM image, HCC-M-Si particles are observed to be entrapped in the CNF network. In this way, the continuous three-dimensional CNF network can provide efficient electron pathway to the HCC-M-Si particles. In addition, the tubular structure of the CNFs can facilitate infiltration of electrolyte, which further provides efficient ion pathway. The cycling performance of HCC-M-Si/CNF free-standing electrode cycled at 0.4 A/g is exhibited in Figure 4.11e and the capacity is calculated based on the total mass of the electrode. The first reversible capacity is 1595 mAh/g, which is retained at 1015 mAh/g after 100 cycles. In our conventional HCC-M-Si electrode produced by slurry-casting method, mass loading of HCC-M-Si active material is 0.5 mg/cm 2 , which takes up 70 wt.% of the pasted electrode material. The current collector is copper foil, which weights 9.2 mg/cm 2 . So this conventional electrode weights 9.9 mg/cm 2 as a whole. After 100 cycles at a current density of 0.4 A/g, the HCC-M-Si retained reversible capacity of 2372 mAh/g based on the mass of HCC-M-Si (Figure 4.4b) and the capacity calculated based on the total mass of the electrode is only 120 mAh/g. Since the free-standing HCC-M-Si/CNF electrode and the conventional HCC-M-Si electrode have similar working potential, the comparison of their energy density is thus merely the comparison of their capacity, which are 1015 mAh/g and 120 mAh/g respectively calculated based on total mass of the electrodes after 100 cycles at a current density of 0.4 A/g. This comparison demonstrates 94 that the lightweight free-standing HCC-M-Si/CNF electrodes significantly improve the energy density by 745% compared with conventional electrodes. The M-Si/CNF free- standing electrode was also fabricated following the same procedure as HCC-M-Si/CNF and the cycling performance is also provided for comparison. It is found that the reversible capacity decreases dramatically from 1078 mAh/g in the 1st cycle to 349 mAh/g in the 15th cycle. This is due to the pore evolution of Si particles during cycling and consequent large volume change of the particles as shown in Figure 4.8b-d, which leads to loss of contact of the particles with CNF network. Without binder used in the conventional LIBs, the volume change of M-Si particles leads to even faster capacity degradation in the free- standing electrode than that in the conventional electrode (Figure 4.4d black curve). On the contrary, the void space between Si core and carbon shell in HCC-M-Si can well accommodate the volume change of Si during cycling process and helps to maintain the contact between HCC-M-Si and CNF network so that the capacity of Si can be retained. The free-standing HCC-M-Si/CNF electrode not only provides a strategy to further improve the energy density of HCC-M-Si, but also confirms the functionality of the hierarchical carbon coating structure in providing good contact between Si and current collector. 4.5 High-voltage full cells based on hierarchical carbon-coated ball- milled silicon Considering the advantages of HCC-M-Si discussed above, prototype of full cells with HCC-M-Si as anode and commercial LiCoO2 as cathode were built to demonstrate the commercial viability of HCC-M-Si in LIBs. On this basis, high-voltage LiNi0.5Mn1.5O4 was 95 synthesized and used to pair up with HCC-M-Si to further improve the energy density of the full cells. The full cells were cycled at a current density of 0.4 A/g based on the mass of HCC-M-Si and the specific capacity of the full cells shown in Figure 4.12 is based on the mass of HCC-M-Si. The fabricated HCC-M-Si/LiCoO2 full cell and HCC-M- Si/LiNi0.5Mn1.5O4 full cell were schematically shown in Figure 4.12a. Following conventions, the lithiation process of HCC-M-Si is defined to be charge and delithiation process is defined to be discharge. Figure 4.12b shows the galvanostatic charge-discharge profiles of the HCC-M-Si/LiCoO2 full cell in the 1st, 2nd, and 5th cycle over the potential range from 3 V to 4.2 V (vs. Li/Li + ). The average voltage is observed to be 3.7 V, which is comparable to that of the commercial graphite/LiCoO2 LIB system. To demonstrate the stable working voltage of the full cell, the fabricated coin cells were used to power commercial blue and white light emitting diodes (LEDs), which have working voltage of 3.2 V and 3.5 V, respectively (insets in Figure 4.12b). The galvanostatic charge-discharge profiles of the HCC-M-Si/LiNi0.5Mn1.5O4 full cells are shown in Figure 4.12c over the potential range of 3 V to 4.8 V (vs. Li/Li + ). Due to the high-voltage cathode LiNi0.5Mn1.5O4, the average working voltage of the HCC-M-Si/LiNi0.5Mn1.5O4 full cell reaches up to 4.4 V, and the plateaus are observed to be more flat than that in Figure 4.12b. With this higher working voltage, the HCC-M-Si/LiNi0.5Mn1.5O4 full cells can also power blue and white LEDs as shown in the insets of Figure 4.12c. Figure 4.12d compares the cycling performance of HCC-M-Si/LiCoO2 full cell and HCC-M-Si(2)/LiCoO2 full cell. The reversible capacity of HCC-M-Si/LiCoO2 full cell in the first cycle is 2760 mAh/g calculated based on the mass of HCC-M-Si and 131 mAh/g based on the total mass of HCC-M-Si and LiCoO2. The energy density based on the total mass of HCC-M-Si and 96 LiCoO2 is calculated to be 486 Wh/kg. After 100 cycles, the HCC-M-Si/LiCoO2 full cell retains a reversible capacity of 931 mAh/g based on the mass of HCC-M-Si, and the corresponding energy density based on total mass is calculated to be 164 Wh/kg. For comparison, the HCC-M-Si(2)/LiCoO2 full cell exhibits energy density of only 383 Wh/kg in the 1st cycle and 90 Wh/kg in the 100th cycle based on the total mass, which further confirms the functionality of having sufficient void space between the Si core and the carbon shell to improve the cycling performance of the batteries. The cycling performance of HCC-M-Si/LiNi0.5Mn1.5O4 high-voltage full cell is demonstrated in Figure 4.12e. Even though the initial reversible capacity of HCC-M-Si/LiNi0.5Mn1.5O4 full cell is 2615 mAh/g, which is slightly lower than that of HCC-M-Si/LiCoO2 full cell, the energy density of the HCC-M-Si/LiNi0.5Mn1.5O4 full cell is boosted up to 547 Wh/kg based on the total mass due to the higher average working voltage than that of HCC-M-Si/LiCoO2 full cell. After cycling for 100 cycles, the reversible capacity retains 1152 mAh/g based on the mass of HCC-M-Si and energy density retains 241 Wh/kg based on the total mass. The energy density of HCC-M-Si/LiNi0.5Mn1.5O4 full cell after 100 cycles is improved 46% compared with HCC-M-Si/LiCoO2 full cell, which is attributed to the higher working voltage of LiNi0.5Mn1.5O4 than that of LiCoO2. This high-voltage Si-based full cells will shed light on further improvement of energy density of Si-based LIBs. 97 Figure 4.12 Electrochemical performance of full cells based on HCC-M-Si. (a) Schematic illustration of the full cell using HCC-M-Si as anode and LiCoO2 or LiNi0.5Mn1.5O4 as cathode. (b) Galvanostatic charge-discharge profiles of HCC-M-Si/LiCoO2 full cell. The inset shows the HCC-M-Si/LiCoO2 full cell can power blue and white LEDs. (c) Galvanostatic charge-discharge profiles of HCC-M-Si/LiNi0.5Mn1.5O4 full cell. The inset shows the HCC-M-Si/LiNi0.5Mn1.5O4 full cell can power blue and white LEDs. (d) Cycling performance comparison of HCC-M-Si/LiCoO2 and HCC-M-Si(2)/LiCoO2 full cells. (e) Cycling performance of HCC-M-Si/LiNi0.5Mn1.5O4 full cell. All the full cells investigated in the figure are cycled at a current density of 0.4 A/g based on the mass of HCC-M-Si and the specific capacity of the full cell is based on the mass of HCC-M-Si. 4.6 Summary In summary, we have demonstrated low-cost and scalable synthesis of hierarchical carbon- coated Si using ball-milled low-cost Si as starting materials. Electrochemical tests of the obtained particles prepared from different Si sources all show excellent cycling performance of over 1000 mAh/g after 1000 cycles at a current density of 2 A/g. Post- 98 cycling characterization of hierarchical carbon-coated metallurgical Si indicates that in situ formed porous Si is well confined in the carbon shell after cycling. Lightweight and free- standing electrodes consisting of hierarchical carbon-coated Si and carbon nanofibers were fabricated, which achieved 1015 mAh/g after 100 cycles based on the total mass of the electrode. The lightweight and free-standing electrodes significantly improve the energy density by 745% compared with conventional electrodes. LiCoO2 and LiNi0.5Mn1.5O4 were used to pair up with hierarchical carbon-coated metallurgical Si to fabricate full cells. With LiNi0.5Mn1.5O4 as cathode, the energy density of the full cell was boosted up to 547 Wh/kg. After 100 cycles, 46% more energy density was achieved by the full cell with LiNi0.5Mn1.5O4 cathode than that of the full cell with LiCoO2 cathode. We believe the reported systematic study on the contact issue in Si anode, the synthesis of low-cost Si anodes together with their applications in lightweight free-standing electrodes and high- voltage full cells can open up the door to further optimization of low-cost Si anodes with high energy density and lead to their real applications in the future. 99 4.7 References 1. Goodenough, J. B.; Kim, Y. Challenges for rechargeable batteries. J. Power Sources 196, 6688-6694 (2011). 2. Goodenough, J. B.; Kim, Y. Challenges for Rechargeable Li Batteries. Chem. Mater. 22, 587-603 (2010). 3. Amine, K.; Tukamoto, H.; Yasuda, H.; Fujita, Y. Preparation and electrochemical investigation of LiMn2-xMexO4 (Me : Ni, Fe, and x=0.5, 1) cathode materials for secondary lithium batteries. J. Power Sources 68, 604-608 (1997). 4. Zhong, Q. M.; Bonakdarpour, A.; Zhang, M. J.; Gao, Y.; Dahn, J. R. Synthesis and electrochemistry of LiNixMn2-xO4. J. Electrochem. Soc. 144, 205-213 (1997). 5. Arrebola, J. C.; Caballero, A.; Cruz, M.; Hernan, L.; Morales, J.; Castellon, E. R. Crystallinity control of a nanostructured LiNi 0.5Mn1.5O4 spinet via polymer-assisted synthesis: A method for improving its rate capability and performance in 5 V lithium batteries. Adv. Funct. Mater. 16, 1904-1912 (2006). 6. Mizushima, K.; Jones, P. C.; Wiseman, P. J.; Goodenough, J. B. LixCoO2 (0<x≤1): a New Cathode Material for Batteries of High-Energy Density. Mater. Res. Bull. 15, 783- 789 (1980). 7. Kang, B.; Ceder, G. Battery materials for ultrafast charging and discharging. Nature 458, 190-193 (2009). 8. Luo, S.; Wang, K.; Wang, J. P.; Jiang, K. L.; Li, Q. Q.; Fan, S. S. Binder-Free LiCoO2/Carbon Nanotube Cathodes for High-Performance Lithium Ion Batteries. Adv. Mater. 24, 2294-2298 (2012). 9. Li, N.; Chen, Z. P.; Ren, W. C.; Li, F.; Cheng, H. M. Flexible graphene-based lithium ion batteries with ultrafast charge and discharge rates. Proc. Natl. Acad. Sci. U.S.A. 109, 17360-17365 (2012). 10. Wen, L.; Li, F.; Cheng, H. M. Carbon Nanotubes and Graphene for Flexible Electrochemical Energy Storage: from Materials to Devices. Adv Mater 28, 4306-4337 (2016). 11. Fu, K.; Yildiz, O.; Bhanushali, H.; Wang, Y. X.; Stano, K.; Xue, L. G.; Zhang, X. W.; Bradford, P. D. Aligned Carbon Nanotube-Silicon Sheets: A Novel Nano-architecture for Flexible Lithium Ion Battery Electrodes. Adv. Mater. 25, 5109-5114 (2013). 12. Wang, Z. H.; Xu, C.; Tammela, P.; Huo, J. X.; Stromme, M.; Edstrom, K.; Gustafsson, T.; Nyholm, L. Flexible freestanding Cladophora nanocellulose paper based Si anodes for lithium-ion batteries. J. Mater. Chem. A 3, 14109-14115 (2015). 13. Lu, M.; Zhang, H. A. Controllable synthesis of spherical silicon and its performance as an anode for lithium-ion batteries. Ionics 19, 1695-1698 (2013). 14. Ji, J. Y.; Ji, H. X.; Zhang, L. L.; Zhao, X.; Bai, X.; Fan, X. B.; Zhang, F. B.; Ruoff, R. S. Graphene-Encapsulated Si on Ultrathin-Graphite Foam as Anode for High Capacity Lithium-Ion Batteries. Adv. Mater. 25, 4673-4677 (2013). 15. Park, E.; Yoo, H.; Lee, J.; Park, M. S.; Kim, Y. J.; Kim, H. Dual-Size Silicon Nanocrystal-Embedded SiOx Nanocomposite as a High-Capacity Lithium Storage Material. ACS Nano 9, 7690-7696 (2015). 16. Li, X. L.; Gu, M.; Hu, S. Y.; Kennard, R.; Yan, P. F.; Chen, X. L.; Wang, C. M.; Sailor, M. J.; Zhang, J. G.; Liu, J. Mesoporous silicon sponge as an anti-pulverization structure for high-performance lithium-ion battery anodes. Nat. Commun. 5, (2014). 100 17. Cho, W. C.; Kim, H. J.; Lee, H. I.; Seo, M. W.; Ra, H. W.; Yoon, S. J.; Mun, T. Y.; Kim, Y. K.; Kim, J. H.; Kim, B. H.; Kook, J. W.; Yoo, C. Y.; Lee, J. G.; Choi, J. W. 5L- Scale Magnesio-Milling Reduction of Nanostructured SiO2 for High Capacity Silicon Anodes in Lithium-Ion Batteries. Nano Lett. 16, 7261-7269 (2016). 18. Lin, D. C.; Lu, Z. D.; Hsu, P. C.; Lee, H. R.; Liu, N.; Zhao, J.; Wang, H. T.; Liu, C.; Cui, Y. A high tap density secondary silicon particle anode fabricated by scalable mechanical pressing for lithium-ion batteries. Energy Environ. Sci. 8, 2371-2376 (2015). 19. Liu, N.; Lu, Z. D.; Zhao, J.; McDowell, M. T.; Lee, H. W.; Zhao, W. T.; Cui, Y. A pomegranate-inspired nanoscale design for large-volume-change lithium battery anodes. Nat. Nanotechnol. 9, 187-192 (2014). 20. Lee, S. J.; Kim, H. J.; Hwang, T. H.; Choi, S.; Park, S. H.; Deniz, E.; Jung, D. S.; Choi, J. W. Delicate Structural Control of Si-SiOx-C Composite via High-Speed Spray Pyrolysis for Li-Ion Battery Anodes. Nano Lett. 17, 1870-1876 (2017). 21. Chen, S. Q.; Shen, L. F.; van Aken, P. A.; Maier, J.; Yu, Y. Dual-Functionalized Double Carbon Shells Coated Silicon Nanoparticles for High Performance Lithium-Ion Batteries. Adv. Mater. 29, (2017). 22. Li, B.; Yang, S. B.; Li, S. M.; Wang, B.; Liu, J. H. From Commercial Sponge Toward 3D Graphene-Silicon Networks for Superior Lithium Storage. Adv. Energy Mater. 5, (2015). 23. Xia, F.; Kwon, S.; Lee, W. W.; Liu, Z.; Kim, S.; Song, T.; Choi, K. J.; Paik, U.; Park, W. I. Graphene as an Interfacial Layer for Improving Cycling Performance of Si Nanowires in Lithium-Ion Batteries. Nano Lett. 15, 6658-6664 (2015). 24. Xiao, Q. F.; Gu, M.; Yang, H.; Li, B.; Zhang, C. M.; Liu, Y.; Liu, F.; Dai, F.; Yang, L.; Liu, Z. Y.; Xiao, X. C.; Liu, G.; Zhao, P.; Zhang, S. L.; Wang, C. M.; Lu, Y. F.; Cai, M. Inward lithium-ion breathing of hierarchically porous silicon anodes. Nat. Commun. 6, (2015). 25. Piper, D. M.; Evans, T.; Xu, S. S.; Kim, S. C.; Han, S. S.; Liu, K. L.; Oh, K. H.; Yang, R. G.; Lee, S. H. Optimized Silicon Electrode Architecture, Interface, and Microgeometry for Next-Generation Lithium-Ion Batteries. Adv. Mater. 28, 188-193 (2016). 26. Zhu, B.; Jin, Y.; Tan, Y. L.; Zong, L. Q.; Hu, Y.; Chen, L.; Chen, Y. B.; Zhang, Q.; Zhu, J. Scalable Production of Si Nanoparticles Directly from Low Grade Sources for Lithium-Ion Battery Anode. Nano Lett. 15, 5750-5754 (2015). 27. Ko, M.; Chae, S.; Ma, J.; Kim, N.; Lee, H. W.; Cui, Y.; Cho, J. Scalable synthesis of silicon-nanolayer-embedded graphite for high-energy lithium-ion batteries. Nature Energy 1, (2016). 28. Jin, Y.; Li, S.; Kushima, A.; Zheng, X. Q.; Sun, Y. M.; Xie, J.; Sun, J.; Xue, W. J.; Zhou, G. M.; Wu, J.; Shi, F. F.; Zhang, R. F.; Zhu, Z.; So, K. P.; Cui, Y.; Li, J. Self-healing SEI enables full-cell cycling of a silicon-majority anode with a coulombic efficiency exceeding 99.9%. Energy Environ. Sci. 10, 580-592 (2017). 29. Zuo, X. X.; Xia, Y. G.; Ji, Q.; Gao, X.; Yin, S. S.; Wang, M. M.; Wang, X. Y.; Qiu, B.; Wei, A. X.; Sun, Z. C.; Liu, Z. P.; Zhu, J.; Cheng, Y. J. Self-Templating Construction of 3D Hierarchical Macro-/Mesoporous Silicon from OD Silica Nanoparticles. ACS Nano 11, 889-899 (2017). 30. Zhang, X. H.; Qiu, X. Y.; Kong, D. B.; Zhou, L.; Li, Z. H.; Li, X. L.; Zhi, L. J. Silicene Flowers: A Dual Stabilized Silicon Building Block for High-Performance Lithium Battery Anodes. ACS Nano 11, 7476-7484 (2017). 101 31. Xu, Q.; Li, J. Y.; Sun, J. K.; Yin, Y. X.; Wan, L. J.; Guo, Y. G. Watermelon-Inspired Si/C Microspheres with Hierarchical Buffer Structures for Densely Compacted Lithium- Ion Battery Anodes. Adv. Energy Mater. 7, (2017). 32. Kim, S. J.; Kim, M. C.; Han, S. B.; Lee, G. H.; Choe, H. S.; Kwak, D. H.; Choi, S. Y.; Son, B. G.; Shin, M. S.; Park, K. W. 3D flexible Si based-composite (Si@Si3N4)/CNF electrode with enhanced cyclability and high rate capability for lithium-ion batteries. Nano Energy 27, 545-553 (2016). 33. Tian, H. J.; Tan, X. J.; Xin, F. X.; Wang, C. S.; Han, W. Q. Micro-sized nano-porous Si/C anodes for lithium ion batteries. Nano Energy 11, 490-499 (2015). 34. Shen, C. F.; Ge, M. Y.; Luo, L. L.; Fang, X.; Liu, Y. H.; Zhang, A. Y.; Rong, J. P.; Wang, C. M.; Zhou, C. W. In Situ and Ex Situ TEM Study of Lithiation Behaviours of Porous Silicon Nanostructures. Sci. Rep. 6, (2016). 35. Hu, L. B.; Wu, H.; Gao, Y. F.; Cao, A. Y.; Li, H. B.; McDough, J.; Xie, X.; Zhou, M.; Cui, Y. Silicon-Carbon Nanotube Coaxial Sponge as Li-Ion Anodes with High Areal Capacity. Adv. Energy Mater. 1, 523-527 (2011). 36. Ge, M.; Fang, X.; Rong, J.; Zhou, C. Review of porous silicon preparation and its application for lithium-ion battery anodes. Nanotechnology 24, (2013). 102 Chapter 5: Air-stable room-temperature mid-infrared photodetectors based on hBN/black arsenic phosphorus/hBN heterostructures 5.1 Introduction Semiconducting two-dimensional (2D) materials including various transition-metal dichalcogenides (TMDCs) and black phosphorus (BP) 1-3 have been extensively explored for photonic and electronic applications. The bandgaps of most semiconducting TMDCs, such as molybdenum disulfide and tungsten diselenide, are close to or larger than 1 eV, which limits their optical applications to ultraviolet (UV), visible, and near infrared (NIR) regions 4 . Black phosphorus, which has a bandgap of 0.33 eV in thin film (> 10 layers) or bulk form, can cover the short-wavelength infrared (SWIR) (0.9-2.5 µm), and part of the medium-wavelength infrared (MWIR) (3-5 µm) 5 . Various studies on BP photodetection in UV, visible 6-9 , NIR 10 , SWIR 11-12 , and MWIR (up to 3.4 µm) 13 were reported in the past few years, demonstrating the potential of BP in wideband optoelectronic applications. Extending the operational range of BP photodetectors to even longer wavelength will significantly enhance its optical functionalities in various applications, such as thermal imaging and sensing. For example, black-body radiation of objects with temperatures ranging from 300 to 1000 K has a peak intensity from 9.7 to 2.9 µm, which is crucial for industrial thermal imaging and infrared tracking 14 . Additionally, free-space optical communication at the first atmosphere window (3-5 µm) avoids significant signal attenuation for the wavelength-dependent nature of Rayleigh and Mie scattering, and at the same time keeps the unique advantages, such as higher data rate and enhanced transmission 103 security 15-17 . Currently, the most commonly used material for photodetector applications in MWIR and long-wavelength infrared (LWIR) is Hg1-xCdxTe (MCT) 18 . MCT is a popular choice due to its tunable bandgap through the control of the composition, and its high absorption in MWIR and LWIR 18 . However, besides its toxicity, its bulk nature makes its integration with other optical structures and components challenging 19 . Recently, synthesis of layered black arsenic phosphorus (b-AsxP1-x) alloy was reported, which has a similar crystal structure to that of orthorhombic BP 20 . The synthesized b-AsxP1- x alloy has a tunable bandgap down to ~0.15 eV, with As concentration of 83% 21 . In a recent study, IR-detector based on the b-AsxP1-x alloy was reported with a decent responsivity for MWIR light 22 . In this device, the interplay among photovoltaic, photo- thermo-electric, and photogating effects leads to complicated photocurrent generation mechanisms at different incident power levels and operational speed. Moreover, the trap states on the b-AsxP1-x/substrate interface play an important role in such b-AsxP1-x photodetectors, and accordingly the photo-response time is long (0.5 millisecond), making its fast operation challenging. A very recent work also reports b-AsxP1-xalloy based photodetectors 23 . In that work, the b-As0.91P0.09 photodetectors show impressive responsivity under mid-infrared excitation up to the wavelength of around 4.6 µm. However, given the reported long carrier lifetime (above 1µs), most likely the carrier trapping also plays a role in those devices. In this study, we fully characterize the structural and optical properties of synthesized b-As0.83P0.17, and then utilize the narrow bandgap of b-As0.83P0.17 to fabricate room temperature MWIR photodetectors operational in the intrinsic photoconduction regime, covering a wavelength range from 3.4 to 7.7 µm. In 104 particular, to demonstrate the long-term stability of b-As0.83P0.17, which can be oxidized under exposure to air, we leverage the hexagonal boron nitride (hBN)/ b-As0.83P0.17/hBN heterostructure for the effective protection. The as-fabricated photodetector shows an extrinsic responsivity of 190, 16, and 1.2 mA/W for 3.4, 5.0, and 7.7 µm incident light, respectively. We further study the photocurrent dependence on the applied gate (Vbg) and drain (Vds) biases, from which we conclude that the photocurrent originates from the intrinsic photoconductive effect. Subsequent investigations of power and frequency dependence confirm the photoconduction mechanism. Moreover, due to the in-plane anisotropy in the crystal structure of b-As0.83P0.17 21 , a polarization-dependent photocurrent is observed. Our demonstration represents an important step for the future applications of b-AsxP1-x in broad mid-infrared wavelength range, in particular for high-speed operations. 5.2 Material characterization of black arsenic phosphorus The b-As0.83P0.17 used for device demonstration in this work was prepared from a mixture of gray arsenic (Chempur, 99.9999%) and red phosphorus (Chempur, 99.999 + %) with a molar ratio 83:17. Lead (II) iodide (PbI2, 12 mg per 625 mg batch) was added as the mineralizing agent. The chemicals were enclosed in evacuated silica glass ampoules during reaction (length: 100 mm, inner diameter: 10 mm). Synthesis was performed in a Nabertherm furnace (L3/11/P330). The samples were heated up to 550 °C within 8 hours, held at this temperature for 24 hours, and then cooled down to room temperature within 20 hours. 105 The energy dispersive X-ray spectroscopy (EDS) spectrum of the synthesized b-AsxP1-x in Figure 5.1 confirms that the compound is composed of ~ 83 atomic percent (at. %) As and ~17 at. % P. The powder X-ray diffraction (XRD) pattern of b-As0.83P0.17 is shown in Figure 5.2a. All the diffraction peaks are indexed to corresponding planes. In b-As0.83P0.17, the interlayer distance is larger than that of grey arsenic at an arsenic concentration ρ As ≈ 10% and it increases monotonically as ρ As further rises up to 83% 20 , which shows that b-AsxP1- x crystal structure is very different from that of grey arsenic. The diffraction pattern obtained in b-As0.83P0.17 here is consistent with the previous work 20 , and the interlayer distance is observably larger than that of grey arsenic. Detailed investigation of the crystal structure of b-As0.83P0.17 was done with single-crystal XRD. The single-crystal XRD analysis indicates that the b-As0.83P0.17 has an orthorhombic structure with a puckered honeycomb lattice as shown in Figure 5.2b (the atoms in Figure 5.2b represent either As or P) 21, 24 . The lattice parameters of the b-As0.83P0.17 are a=3.561(3) Å, b=10.803(9) Å, and c=4.493(4) Å, where a, b, and c are the lattice constants along zig-zag, stacking, and armchair direction. The details of the crystallographic data of b-As0.83P0.17 are provided in Table 5.1. 106 Figure 5.1 EDS spectrum of b-AsxP1-x. The elemental composition is summarized in the inset table. Figure 5.2 Crystal structure and infrared extinction characterizations. (a) X-ray diffraction pattern of the as-synthesized crystal with Miller indices labeled for peaks. The broad peak at 18° is from the Kapton tape, which is used to encapsulate the b-As0.83P0.17. (b) The orthorhombic puckered honeycomb crystal structure of b-As0.83P0.17 alloy. The lattice parameters, extracted from the X-ray diffraction profile, are a = 3.561(3) Å, b = 10.803(9) Å, c = 4.493(4) Å. (c) Polarization-resolved IR extinction spectra of b-As0.83P0.17 alloy. The optical image of the investigated flake is shown in the inset. 107 Table 5.1 The crystallographic information of b-As0.83P0.17. Empirical formula As0.83P0.17 Space group Cmca Crystal system orthorhombic Unit cell dimensions a=3.561(3) Å α=90˚ b=10.803(9) Å β=90˚ c=4.493(4) Å γ=90˚ Volume 172.8(3) Å 3 Z 8 Calculated density 5.759 g/cm 3 Transmission ratio (max/min) 0.3950/0.1310 Absorption coefficient 38.049 mm -1 F(000) 264 θ range 3.77˚-30.30˚ Total number of reflections 763 Independent reflections 147 (Rint = 0.0467) Data/restraints/parameters 147/0/7 Goodness-of-fit on F 2 1.101 Final R indices [I > 2σ(I)] R1 = 0.0334 wR2 = 0.0760 R indices (all data) R1 = 0.0364 wR2 = 0.0775 Largest diff. peak and hole 1.239 and -2.215 eÅ -3 Figure 5.2c shows the polarization-resolved infrared extinction spectra of a typical exfoliated b-As0.83P0.17 flake on zinc sulfide (ZnS) substrate (shown in the inset), where T and T 0 are the optical transmission of b-As0.83P0.17 flake on ZnS substrate and the bare ZnS substrate, respectively. The measured extinction of the b-As0.83P0.17 flake shows an absorption edge around 1000 cm -1 , corresponding to a photon energy of 0.124 eV. The thickness of the investigated flake is measured to be 188 nm using atomic force microscope (AFM), and the height profile of the flake edge can be found in Figure 5.3. We employ the Lambert-Beer’s Law: A(ω) = 1-e −α(ω)Z to estimate the absorption constant α(ω) of bulk b-As0.83P0.17, where A(ω) is the absorption, ω is the frequency of incident light, and Z is the flake thickness. In the calculation, we assume the absorption constant α(ω) is the same 108 for the bulk material Z >> 5 nm. The absorption constant can be slightly overestimated by about 5% in this approach, considering the reflection at the b-As 0.83 P 0.17 surface. For incident light wavelength λ = 3.39 μm (υ = 2950 cm −1 ), the absorption constant α is calculated to be 3.4 × 10 4 cm −1 , using the 47.6% absorption in the 188-nm-thick flake. Figure 5.3 The thickness profile of the exfoliated b-As0.83P0.17 flake used for extinction spectra characterization acquired using atomic force microscope. 5.3 Electrical characterization of black arsenic phosphorus device After fully characterizing b-As0.83P0.17 flakes, we fabricated photodetectors based on hBN/b-As0.83P0.17/hBN heterostructures. The bottom hBN flakes were first exfoliated on a Si substrate covered by a 90-nm-thick SiO2 layer, and then the substrate was annealed in Argon/Hydrogen (Ar/H 2 ) atmosphere at 600 ℃ for 6 hours to remove possible polymer residues in the exfoliation. The b-As0.83P0.17 flake and the top hBN were transferred onto the bottom hBN flakes subsequently using a polymer-free dry transfer method 25 . All the transfer processes were performed in argon-filled glovebox to minimize oxidation. After a 6-hour annealing in Ar/H 2 atmosphere at 300 ℃, the defined metal contact region was 109 etched to expose b-As0.83P0.17 and then electrodes (3/47nm Chromium/Gold) were deposited 26 . Similar to BP 27-28 , b-AsxP1-x can also be oxidized in air, which makes the naked devices unstable in the long run. The encapsulation provides reliable protection for the b-As0.83P0.17 alloy against the oxidation and degradation. Figure 5.4a shows the cross-sectional schematic of the hBN/b-As0.83P0.17/hBN heterostructure photodetector on a SiO2/Si substrate. It is based on a field-effect transistor scheme with interdigitated source and drain electrodes to maximize the photocurrent collection efficiency. The cross-sectional view of the device and the elemental analysis by transmission electron microscope (TEM) are shown in Figure 5.4b. The 37-nm-thick b-As0.83P0.17 is well preserved in hBN and free from oxidation several months after its fabrication as shown in the elemental analysis in Figure 5.4b. The relatively thick b-As0.83P0.17 flake is chosen to enhance the absorption of IR light. The thicknesses of the top and bottom hBN layers are 20 and 22 nm, respectively, determined from the cross-sectional TEM image. 110 Figure 5.4 Device structure and electrical characterizations. (a) The cross-sectional schematic of the as-fabricated hBN/b-As0.83P0.17/hBN heterostructure photodetector. (b) Left: the cross-sectional view of the device by transmission electron microscope (TEM). Right: the elemental analysis mapping by electron energy loss spectroscopy (EELS). In the EELS mapping, red, blue, and green color denote O, N, and As elements respectively. Only part of the silicon oxide layer is shown. The encapsulated b-As0.83P0.17 layer is 37 nm, free from oxidation even several months after its fabrication. (c) Transfer characteristic of the as-fabricated phototransistor. The optical image of the as-fabricated device is shown in the inset. The gate bias was swept in both directions, with numbers denoting the sweeping sequence. The charge neutral point is at Vbg=-3 V the hysteresis of the transfer curve is small. (d) Output characteristics of the b-As0.83P0.17 phototransistor. The output curves were measured with different V bg , and V N represents the gate bias, at which the charge neutral point of the device is achieved (VN = -3 V in this case). The transfer characteristics of the as-fabricated device are plotted in Figure 5.4c. The inset shows its optical image. The device exhibits rather symmetrical ambipolar property compared to previous studies 13, 22 , probably due to the hBN encapsulation, which leads to the preservation of the intrinsic properties of b-As0.83P0.17. When Vbg is swept in both 111 directions, the encapsulated b-As0.83P0.17 transistors exhibit negligible hysteresis, due to the clean hBN/b-As0.83P0.17 interfaces. The electron and hole mobilities can be extracted using μ = ( dI ds dV bg )(L/W)(1/V ds C g ), where I ds is the drain current, C g is the gate capacitance, and L = 1.5 μm and W = 30 μm are the length and the total width of the transistor channel. Here both the hBN and SiO2 are taken into account for the gate capacitance calculation, with a relative dielectric constant of 3.1 and 3.9, respectively 29 . Based on this approach, the extracted effective electron and hole mobilities are about 83 and 79 cm 2 /(V ∙ s), respectively. We note that the intrinsic carrier mobilities could be higher than the extracted effective mobilities above, which do not exclude the contribution of the contact resistance. Figure 5.4d shows the output curves of the as-fabricated device at various back gate biases. The linear relation between Ids and Vds indicates that the Ohmic contact is achieved between metal electrodes and b-As0.83P0.17. 5.4 Photoresponse of black arsenic phosphorus phototransistor Then we characterized the photo-response of b-As0.83P0.17 devices for MWIR light at 3.4, 5.0, and 7.7 µm. The incident IR light was mechanically chopped at certain frequency (3 kHz, if not specified), and the alternating photocurrent (I ph ) at corresponding frequency was acquired using a lock-in amplifier. Since b-AsxP1-x has intrinsic anisotropic optical properties, the photocurrent naturally depends on the incident light polarization. The polarization dependent photocurrent was measured when the device was biased at the charge neutral point and V ds = 300 mV. In Figure 5.5a, the photocurrent is plotted in a polar coordinate as a function of the polarization for 3.4 , 5.0, and 7.7 µm incident light. The photocurrent anisotropy, defined as the photocurrent ratio along the armchair and zig-zag 112 directions, are ~12, 14, and 1.3 for 3.4, 5.0, and 7.7 µm lasers respectively. We notice that the anisotropy of photocurrent is stronger than that observed in the infrared extinction curves as shown in Figure 5.2c. In infrared extinction spectra, a large area (~ 30 × 30 × 0.188 μm) flake was used to accommodate the incidence light from Fourier transform infrared spectrometer (FTIR), while the device was fabricated on a much smaller flake (~ 8 × 8 × 0.037 μm). Accordingly, the measured anisotropy in extinction spectra is the average over larger area, which can contain multiple crystalline domains. As a result, the infrared extinction spectra show much less anisotropy. The weak photocurrent and infrared extinction at the band edge may be due to the emergence of a new less anisotropic valence band at high arsenic concentration, as suggested by the recent calculations based on the density function theory 30 . Figure 5.5 Photoresponse of hBN encapsulated b-As0.83P0.17 phototransistor. (a) Polarization-resolved photocurrents of the b-As0.83P0.17 phototransistor for 3.4, 5.0, and 7.7 µm light excitation at V ds = 0.3 V when the device is biased at charge neutral point. (b) V bg dependence of the photocurrent in the b-As0.83P0.17 phototransistor at V ds = 0.3 V. (c) The photoresponse as a function of source-drain bias V ds . In this measurement, the device works at charge neutral point. Transparent colored lines are the linear fitting in the range of |V ds | < 0.4 V. The photocurrent in (b) and (c) was measured when the polarization of incident light was aligned with the armchair direction. The photocurrent dependence on Vbg was measured at a light polarization along the armchair direction. For incident laser with different photon energies, the photocurrent 113 always peaks at the charge neutral point, as shown in Figure 5.5b. This is the typical feature of photoconductive effect, since the higher doping away from the charge-neutral point reduces the photocarrier lifetime, leading to the reduction in photocurrent. At the charge neutral point, the photocurrent scales almost linearly with Vds when |Vds| < 0.4 V, as shown in Figure 5.5c. The saturation of photocurrent at large Vds can be attributed to the Joule heating effect, which sets the limit of the internal amplification for the decreasing photocarrier lifetime at higher temperature 31 . The extrinsic responsivity is estimated using R ex = I ph /(P in ξ), where P in is the power of incident light, ξ is the ratio of which the incident light is impinging on the device, considering the Gaussian distribution of light intensity and the reflection of metal electrodes. When the device is biased at V ds = 1 V and V bg = −3 V (corresponding to the charge-neutrality point), the corresponding extrinsic responsivity for 3.4, 5.0, and 7.7 μm laser is calculated to be ~190, 16, and 1.2 mA/W, respectively. To further elucidate the photocurrent generation mechanism, the power and frequency dependence of the photocurrent were also measured. As shown in Fig. 5.6a, in our hBN/b- As0.83P0.17/hBN photodetectors, the photocurrent scales approximately linearly with input power. This is distinctively different from the previously demonstrated b-As0.83P0.17 photodetector 22 , which shows a sublinear power-dependent photocurrent. The decreasing responsivity at higher intensity light in the previous work can be attributed to trap states on the b-As0.83P0.17/SiO2 interface. At higher illumination condition, surface trap states are densely occupied, leading to the saturation of photogating effect and decreasing responsivity. In our device, the intrinsic photoconduction leads to a photocurrent linearly 114 scaled with the input power, since more electron-hole pairs can be excited proportionally by higher intensity light. Besides the distinguished power dependency, our hBN- encapsulated b-As0.83P0.17 MWIR photodetector also shows excellent response speed. The photocurrent remains constant when the chopped frequency of incident light ranges from 1 to 10 kHz, which indicates that the 3 dB cutoff frequency of the device is well beyond 10 kHz (Figure 5.6b). Here we use photocarrier lifetime to estimate the intrinsic 3 dB cutoff frequency of the device. With the absorption constant α(ω) calculated from extinction spectra, we estimate the absorption (η) of the 37-nm-thick flake to be 12% at 3.4 µm using Lambert-Beer’s Law. Then the product of internal quantum efficiency (IQE) and photoelectric current gain G at the charge neutral point and V ds = 1 V is calculated, with IQE · G = (R ex /η)(E ph /e), to be around 57%, where E ph is the photon energy and e is the elementary charge. On the other hand, the transition time τ tr = L 2 /(μV ds ) for electrons and holes are calculated to be 27 and 29 ps, respectively. Considering both carrier transition time and IQE · G , the photocarrier lifetime at the charge neutral point ( τ life = τ tr e τ tr h /(τ tr h + τ tr e )·IQE·G) is estimated to be around 8 ps, where τ tr e/h is the transition time for electrons or holes. The calculated photocarrier lifetime in this work suggests a potential 3-dB cutoff frequency (f 3dB = 1/(2πτ life )) of the photoconductor at around 19 GHz. In terms of the photocurrent generation mechanism, the observed fast photo-response also eliminates the probability of photogating effects in our hBN-encapsulated b-As0.83P0.17 photodetectors 13 . 115 Figure 5.6 Power and frequency dependence of the photocurrent and noise characteristics. (a) Photocurrent as a function of incident power at 3.4 µm when the device works at charge neutral point and V ds = 0.3 V. The laser polarization is aligned with the armchair direction. (b) Photo-response as a function of incident light intensity modulation frequency from 1 to 10 kHz, showing no sign of response roll-off. (c) Noise equivalent power at V ds = 1 V for 3.4, 5.0, and 7.7 µm incident light when the device works at charge neutral point and the light polarization is aligned with the armchair direction. Since the as-fabricated b-As0.83P0.17 photodetectors intrinsically operate at high frequencies, the frequency-independent shot noise and thermal noise play the major role in the detection. The contribution from 1/f flicker noise is significant only at low operating frequencies (<10 kHz), and hence is out of scope in this work. We estimated the detection noise and corresponding noise equivalent power (NEP) at high frequencies considering the shot noise, thermal noise, and generation-recombination noise in the device. The current noise power spectral density function S S of the shot noise, generated from the discrete nature of electrons, is given by S S = 2eI ds . For thermal noise, it originates from random thermal motion of electrons and is expressed as S T = 4kT R , where k is Boltzmann constant, T is the device working temperature and R is the device resistance. The generation and recombination of photocarriers is another generation source of noise, but we estimate the generation-recombination noise, S GR = 4eI ph (τ life /τ tr )/(1 + (2πfτ life ) 2 ), to be smaller than the shot noise by more than one order and contributes little to the NEP in our case. With considerations above, we estimate the noise current density (δi = √S S + S T ) and the 116 noise equivalent power (NEP = δi/R ex ) at V ds = 1 V at charge neutral points for 3.4, 5.0, and 7.7 m incident light when the light polarization is aligned along the armchair direction. The operating dark current is 117 µA and the associated noise current density is 6.3 pA/Hz 1/2 . Then we calculate the NEP of our photodetector to be 28, 370, and 4300 pW/Hz 1/2 for 3.4, 5.0, and 7.7 m incident light as shown in Figure 5.6c. 5.5 Summary In conclusion, we investigate the structural and optical properties of the synthesized b- As0.83P0.17, and further demonstrate the MWIR photodetectors, which cover a broad mid- IR wavelength range from 3.4 to 7.7 m. Most importantly, the hBN encapsulation provides long-term air stability for the device and eliminates trap states at the interfaces. In terms of the photocurrent generation mechanism, on the contrary to previously demonstrated BP and b-As0.83P0.17 photodetectors, the b-As0.83P0.17 photodetectors demonstrated in this work operate in intrinsic photoconduction mode and exhibit extrinsic photoresponsivity of 190, 16, and 1.2 mA/W for 3.4, 5.0, and 7.7 m incident light, respectively. The 3-dB cutoff frequency for MWIR detection can potentially be as high as 19 GHz. 117 5.6 References 1. Li, L. K.; Yu, Y. J.; Ye, G. J.; Ge, Q. Q.; Ou, X. D.; Wu, H.; Feng, D. L.; Chen, X. H.; Zhang, Y. B. Black phosphorus field-effect transistors. Nat. Nanotechnol. 9, 372-377 (2014). 2. Xia, F. N.; Wang, H.; Jia, Y. C. Rediscovering black phosphorus as an anisotropic layered material for optoelectronics and electronics. Nat. Commun. 5, (2014). 3. Liu, H.; Neal, A. T.; Zhu, Z.; Luo, Z.; Xu, X. F.; Tomanek, D.; Ye, P. D. Phosphorene: An Unexplored 2D Semiconductor with a High Hole Mobility. ACS Nano 8, 4033-4041 (2014). 4. Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S. Electronics and optoelectronics of two-dimensional transition metal dichalcogenides. Nat. Nanotechnol. 7, 699-712 (2012). 5. Li, L. K.; Kim, J.; Jin, C. H.; Ye, G. J.; Qiu, D. Y.; da Jornada, F. H.; Shi, Z. W.; Chen, L.; Zhang, Z. C.; Yang, F. Y.; Watanabe, K.; Taniguchi, T.; Ren, W.; Louie, S. G.; Chen, X. H.; Zhang, Y. B.; Wang, F. Direct observation of the layer-dependent electronic structure in phosphorene. Nat. Nanotechnol. 12, 21-25 (2017). 6. Wu, J.; Koon, G. K. W.; Xiang, D.; Han, C.; Toh, C. T.; Kulkarni, E. S.; Verzhbitskiy, I.; Carvalho, A.; Rodin, A. S.; Koenig, S. P.; Eda, G.; Chen, W.; Neto, A. H. C.; Ozyilmaz, B. Colossal Ultraviolet Photoresponsivity of Few-Layer Black Phosphorus. ACS Nano 9, 8070-8077 (2015). 7. Huang, M. Q.; Wang, M. L.; Chen, C.; Ma, Z. W.; Li, X. F.; Han, J. B.; Wu, Y. Q. Broadband Black-Phosphorus Photodetectors with High Responsivity. Adv. Mater. 28, 3481-+ (2016). 8. Buscema, M.; Groenendijk, D. J.; Steele, G. A.; van der Zant, H. S. J.; Castellanos- Gomez, A. Photovoltaic effect in few-layer black phosphorus PN junctions defined by local electrostatic gating. Nat. Commun. 5, (2014). 9. Miao, J. S.; Zhang, S. M.; Cai, L.; Wang, C. Black Phosphorus Schottky Diodes: Channel Length Scaling and Application as Photodetectors. Adv. Electron. Mater. 2, (2016). 10. Buscema, M.; Groenendijk, D. J.; Blanter, S. I.; Steele, G. A.; van der Zant, H. S. J.; Castellanos-Gomez, A. Fast and Broadband Photoresponse of Few-Layer Black Phosphorus Field-Effect Transistors. Nano Lett. 14, 3347-3352 (2014). 11. Engel, M.; Steiner, M.; Avouris, P. Black Phosphorus Photodetector for Multispectral, High-Resolution Imaging. Nano Lett. 14, 6414-6417 (2014). 12. Youngblood, N.; Chen, C.; Koester, S. J.; Li, M. Waveguide-integrated black phosphorus photodetector with high responsivity and low dark current. Nat. Photonics 9, 247-252 (2015). 13. Guo, Q. S.; Pospischil, A.; Bhuiyan, M.; Jiang, H.; Tian, H.; Farmer, D.; Deng, B. C.; Li, C.; Han, S. J.; Wang, H.; Xia, Q. F.; Ma, T. P.; Mueller, T.; Xia, F. N. Black Phosphorus Mid-Infrared Photodetectors with High Gain. Nano Lett. 16, 4648-4655 (2016). 14. Luong, M. P. Infrared Thermographic Scanning of Fatigue in Metals. Nucl. Eng. Des. 158, 363-376 (1995). 15. Majumdar, A. K. Free-Space Laser Communications: Principles and Advances; Springer: New York, (2008). 118 16. Ghassemlooy, Z.; Boucouvalas, A. C. Indoor optical wireless communication systems and networks. Int. J. Commun. Syst. 18, 191-193 (2005). 17. Capasso, F.; Paiella, R.; Martini, R.; Colombelli, R.; Gmachl, C.; Myers, T. L.; Taubman, M. S.; Williams, R. M.; Bethea, C. G.; Unterrainer, K.; Hwang, H. Y.; Sivco, D. L.; Cho, A. Y.; Sergent, A. M.; Liu, H. C.; Whittaker, E. A. Quantum cascade lasers: Ultrahigh-Speed operation, optical wireless communication, narrow linewidth, and far- infrared emission. IEEE J. Quantum Elect. 38, 511-532 (2002). 18. Dhar, N. K.; Dat, R.; Sood, A. K. Advances in Infrared Detector Array Technology. Optoelectronics-Advanced Materials and Devices 149-190 (2013). 19. Rogalski, A.; Adamiec, K.; Rutkowski, J. Narrow-Gap Semiconductor Photodiodes; SPIE Press: Bellingham, (2000). 20. Osters, O.; Nilges, T.; Bachhuber, F.; Pielnhofer, F.; Weihrich, R.; Schoneich, M.; Schmidt, P. Synthesis and Identification of Metastable Compounds: Black ArsenicuScience or Fiction? Angew. Chem. Int. Ed. 51, 2994-2997 (2012). 21. Liu, B. L.; Kopf, M.; Abbas, A. N.; Wang, X. M.; Guo, Q. S.; Jia, Y. C.; Xia, F. N.; Weihrich, R.; Bachhuber, F.; Pielnhofer, F.; Wang, H.; Dhall, R.; Cronin, S. B.; Ge, M. Y.; Fang, X.; Nilges, T.; Zhou, C. W. Black Arsenic-Phosphorus: Layered Anisotropic Infrared Semiconductors with Highly Tunable Compositions and Properties. Adv. Mater. 27, 4423- 4429 (2015). 22. Long, M. S.; Gao, A. Y.; Wang, P.; Xia, H.; Ott, C.; Pan, C.; Fu, Y. J.; Liu, E. F.; Chen, X. S.; Lu, W.; Nilges, T.; Xu, J. B.; Wang, X. M.; Hu, W. D.; Miao, F. Room temperature high-detectivity mid-infrared photodetectors based on black arsenic phosphorus. Sci. Adv. 3, (2017). 23. Amani, M.; Regan, E.; Bullock, J.; Ahn, G. H.; Javey, A. Mid-Wave Infrared Photoconductors Based on Black Phosphorus-Arsenic Alloys. ACS Nano 11, 11724-11731 (2017). 24. Kopf, M.; Eckstein, N.; Pfister, D.; Grotz, C.; Kruger, I.; Greiwe, M.; Hansen, T.; Kohlmann, H.; Nilges, T. Access and in situ growth of phosphorene-precursor black phosphorus. J. Cryst. Growth 405, 6-10 (2014). 25. Castellanos-Gomez, A.; Buscema, M.; Molenaar, R.; Singh, V.; Janssen, L.; van der Zant, H. S. J.; Steele, G. A. Deterministic transfer of two-dimensional materials by all-dry viscoelastic stamping. 2D Mater. 1, (2014). 26. Chen, X. L.; Wu, Y. Y.; Wu, Z. F.; Han, Y.; Xu, S. G.; Wang, L.; Ye, W. G.; Han, T. Y.; He, Y. H.; Cai, Y.; Wang, N. High-quality sandwiched black phosphorus heterostructure and its quantum oscillations. Nat. Commun. 6, (2015). 27. Favron, A.; Gaufres, E.; Fossard, F.; Phaneuf-L'Heureux, A. L.; Tang, N. Y. W.; Levesque, P. L.; Loiseau, A.; Leonelli, R.; Francoeur, S.; Martel, R. Photooxidation and quantum confinement effects in exfoliated black phosphorus. Nat. Mater. 14, 826-832 (2015). 28. Wood, J. D.; Wells, S. A.; Jariwala, D.; Chen, K. S.; Cho, E.; Sangwan, V. K.; Liu, X. L.; Lauhon, L. J.; Marks, T. J.; Hersam, M. C. Effective Passivation of Exfoliated Black Phosphorus Transistors against Ambient Degradation. Nano Lett. 14, 6964-6970 (2014). 29. Kim, K. K.; Hsu, A.; Jia, X. T.; Kim, S. M.; Shi, Y. M.; Dresselhaus, M.; Palacios, T.; Kong, J. Synthesis and Characterization of Hexagonal Boron Nitride Film as a Dielectric Layer for Graphene Devices. ACS Nano 6, 8583-8590 (2012). 119 30. Zhu, Z.; Guan, J.; Tomanek, D. Structural Transition in Layered As1-xPx Compounds: A Computational Study. Nano Lett. 15, 6042-6046 (2015). 31. Freitag, M.; Low, T.; Xia, F. N.; Avouris, P. Photoconductivity of biased graphene. Nat. Photonics 7, 53-59 (2013). 120 Chapter 6: Conclusions and future work 6.1 Conclusions In this dissertation, I have presented my systematic study on the lithiation behavior of porous Si nanostructures. On this basis, full cells based on porous Si anode and sulfur cathode are studied with discussion of the failure mechanism of the Si-S full cells and proposed solution. To achieve production of low-cost Si anode with high energy density, hierarchical carbon-coated Si was synthesized by using low-cost ball-milled Si as starting material. On this basis, lightweight and free-standing electrode based on the hierarchical carbon-coated Si was fabricated, which significantly improves the energy density by 745%. In addition, full cells with LiCoO2 and LiNi0.5Mn1.5O4 cathodes are studied. With LiNi0.5Mn1.5O4 as cathode, energy density up to 547 Wh/kg was achieved by the high- voltage full cell and the full cell with LiNi0.5Mn1.5O4 cathode delivers 46% more energy density than that of the full cell with LiCoO2 cathode after 100 cycles. My Si anode research provides insights into the factors that determine the electrochemical performance of Si anode and provides effective solution to develop low-cost and high-performance Si anode for lithium-ion batteries. For the 2D material photodetector research, novel 2D material black arsenic phosphorus was systematically studied in material characterization, device fabrication, and photocurrent measurement. The hBN encapsulation of the device leads to its intrinsic transport behavior and photoconduction. Extrinsic responsivity of 190, 16, and 1.2 mA/W at 3.4, 5.0, and 7.7 m is achieved. The research provides effective solution to obtain 121 intrinsic electrical properties of air-sensitive 2D material and will therefore shed light on the development of other 2D material photodetector. 6.2 Future research 6.2.1 LixSi as lithium-ion battery anode In the first lithiation process of Si anode, irreversible formation of solid electrolyte interphase takes place on the surface of Si. As a result, the first lithiation process consumes a fraction of the lithium ions, giving rise to a net loss of storage capacity 1 . Such first-cycle irreversible capacity loss is usually compensated by additional loading of cathode materials in current lithium-ion batteries. However, current commercially available lithium metal- oxide cathodes have much lower specific capacity (mostly less than ~200mAh/g) than anodes. Excessive loading of cathode material causes appreciable reduction of battery specific energy and energy density. For this reason, it is necessary to develop an alternative method to suppresses this irreversible capacity loss and consequently increases the first- cycle Coulombic efficiency of Si-based lithium-ion batteries. Prelithiation of Si anode before pairing up with cathode to assemble full cells is an effective approach to solve the above-mentioned problem 2-3 . However, since LixSi is air-sensitive, the preparation of LixSi should be done in inert environment such as Ar-filled glovebox, which brings difficulty for large-scale fabrication of LixSi. To solve these problems, we propose a single-step and large-scale fabrication method of LixSi in which ball-milling of Li and Si in glovebox is applied. As shown in Figure 6.1, Li, Si, together with carbon black, binder, and solvent are ball-milled, after which a uniform slurry is obtain. By pasting the 122 slurry onto Cu foil and drying, electrodes will be obtained directly. Systematic investigation on the obtained electrodes will be conducted. Figure 6.1 Schematic representation of the fabrication process of LixSi. 6.2.2 Two-dimensional tellurium photodetector With small direct bandgap, BP has demonstrated photoresponse under medium-wavelength infrared and b-As0.83P0.17 has even demonstrated photoresponse under long-wavelength infrared. However, the air-sensitive nature of BP and b-As0.83P0.17 makes it difficult to preserve the intrinsic properties of the fabricated devices 4 . In this sense, it is highly desirable to explore novel 2D materials which exhibit similar electronic/optoelectronic properties as BP/b-As0.83P0.17 and also air-stable. 2D tellurium (Te) is a newly-developed material which exhibits similar electronic/optoelectronic properties as BP 5 . The TEM images of the synthesized Te are shown in Figure 6.2, which demonstrates its single-crystalline nature and large-size. 123 Figure 6.2 (a) TEM image of the synthesized 2D Te. (b) HRTEM image of Te. Field-effect transistors (FETs) based on Te were fabricated as shown in the inset of Figure 6.3a. High on-state current up to 2.6 mA was achieved and the mobility of the device was calculated to be 325 cm 2 /V·s. Figure 6.3 Device characteristics of Te FETs. (a) Typical output (Ids-Vds) and (b) transfer characteristics (Ids-Vg) of the Te FETs. 124 The photoresponse of Te FETs under 520 nm laser is shown in Figure 6.4. The photodetector exhibits polarization-dependent photocurrent and the responsivity is up to 383 A/W. Figure 6.4 Photocurrent of Te filed-effect transistor under 520 nm laser. (a) Polarization- dependent photocurrent. (b) Power-dependent photocurrent. (c) Power dependence of responsivity and EQE. (d) Photocurrent amplitude versus modulation frequency under various incident power. The photoresponse of Te FETs under 1.55 µm and 3.39 µm laser are shown in Figure 6.5 and 6.6 respectively. It is observed that the photodetector exhibits different photocurrent generation mechanism under 1.55 µm and 3.39 µm infrared and the underlying mechanism is under investigation. 125 Figure 6.5 Photocurrent of Te filed-effect transistor under 1.55 µm laser. (a) Polarization- dependent photocurrent versus Vg. (b) Polarization-dependent photocurrent summarized in polar coordinate. (c) Power-dependent photocurrent. (d) Frequency-dependent photocurrent. 126 Figure 6.6 Photocurrent of Te filed-effect transistor under 3.39 µm laser. (a) Polarization- dependent photocurrent versus Vg. (b) Polarization-dependent photocurrent summarized in polar coordinate. (c) Power-dependent photocurrent. (d) Frequency-dependent photocurrent. 127 6.3 References 1. Zhao, J.; Lu, Z. D.; Liu, N. A.; Lee, H. W.; McDowell, M. T.; Cui, Y. Dry-air-stable lithium silicide-lithium oxide core-shell nanoparticles as high-capacity prelithiation reagents. Nat. Commun. 5, (2014). 2. Zhao, J.; Lu, Z. D.; Wang, H. T.; Liu, W.; Lee, H. W.; Yan, K.; Zhuo, D.; Lin, D. C.; Liu, N.; Cui, Y. Artificial Solid Electrolyte Interphase-Protected LixSi Nanoparticles: An Efficient and Stable Prelithiation Reagent for Lithium-Ion Batteries. J. Am. Chem. Soc. 137, 8372-8375 (2015). 3. Zhao, J.; Lee, H. W.; Sun, J.; Yan, K.; Liu, Y. Y.; Liu, W.; Lu, Z. D.; Lin, D. C.; Zhou, G. M.; Cui, Y. Metallurgically lithiated SiOx anode with high capacity and ambient air compatibility. Proc. Natl. Acad. Sci. U.S.A. 113, 7408-7413 (2016). 4. Wood, J. D.; Wells, S. A.; Jariwala, D.; Chen, K. S.; Cho, E.; Sangwan, V. K.; Liu, X. L.; Lauhon, L. J.; Marks, T. J.; Hersam, M. C. Effective Passivation of Exfoliated Black Phosphorus Transistors against Ambient Degradation. Nano Lett. 14, 6964-6970 (2014). 5. Huang, X. C.; Guan, J. Q.; Lin, Z. J.; Liu, B.; Xing, S. Y.; Wang, W. H.; Guo, J. D. Epitaxial Growth and Band Structure of Te Film on Graphene. Nano Lett. 17, 4619-4623 (2017).
Abstract (if available)
Abstract
Nanomaterials have been receiving great attention in the past decade due to their wide applications in numerous areas. Among all the applications, energy storage devices and electronic/optoelectronic devices are two of the most appealing topics. During my Ph.D. study, I conduct research on the two major topics: 1. Study of novel silicon (Si) nanostructures and their application as lithium-ion battery anode materials
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Nanostructured silicon for lithium-ion battery anode
PDF
Modification of electrode materials for lithium ion batteries
PDF
Nanostructured silicon anode and sulfur cathode for lithium rechargeable batteries
PDF
Nanomaterials for macroelectronics and energy storage device
PDF
One-dimensional nanostructures for chemical sensing, transparent electronics, and energy conversion and storage devices
PDF
Novel nanomaterials for electronics, optoelectronics and sensing applications
PDF
One-dimensional nanomaterials for electronic and sensing applications
PDF
Nanostructure design of sulfur cathodes and lithium metal anodes for lithium-ion batteries
PDF
Understanding the role of electrode design in determining the electrochemical performance of high-energy/high-power lithium-ion and lithium-sulfur batteries
PDF
GaAs nanowire optoelectronic and carbon nanotube electronic device applications
PDF
Zero-dimensional and one-dimensional nanostructured materials for application in photovoltaic cells
PDF
Nanostructured III-V photoelectrodes for high performance, high durability solar water splitting
PDF
New materials and device structure for organic light-emitting diodes
PDF
Silicon micro-ring resonator device design for optical interconnect systems
PDF
Hot carrier enhanced photocatalysis in plasmon resonant metal grating systems
PDF
Optoelectronic, thermoelectric, and photocatalytic properties of low dimensional materials
PDF
On the role of polyhedral rotations in mediating ion insertion processes for energy storage materials
PDF
Low-dimensional material based devices for neuromorphic computing and other applications
PDF
Cathode and anode materials for sodium ion batteries
PDF
Nonlinear optical nanomaterials in integrated photonic devices
Asset Metadata
Creator
Shen, Chenfei
(author)
Core Title
Nanomaterials for energy storage devices and electronic/optoelectronic devices
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Materials Science
Publication Date
04/11/2018
Defense Date
03/19/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
lithium-ion battery,OAI-PMH Harvest,silicon, anode, full cell, two-dimensional materials, electronics, photodetector
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Zhou, Chongwu (
committee chair
), Goo, Edward (
committee member
), Wu, Wei (
committee member
), Yoon, Jongseung (
committee member
)
Creator Email
cfshen1989@gmail.com,chenfeis@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c89-3625
Unique identifier
UC11671702
Identifier
etd-ShenChenfe-6222.pdf (filename),usctheses-c89-3625 (legacy record id)
Legacy Identifier
etd-ShenChenfe-6222.pdf
Dmrecord
3625
Document Type
Dissertation
Rights
Shen, Chenfei
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
lithium-ion battery
silicon, anode, full cell, two-dimensional materials, electronics, photodetector