sub-micron Wires

Subscriber access provided by AUSTRALIAN NATIONAL UNIV

Article

Synthesis and Characterization of CuFe2O4 Nano/sub-micron WiresCarbon Nanotube Composites as Binder-free Anodes for Li-ion Batteries Lei Wang, David C. Bock, Jing Li, Eric A. Stach, Amy C. Marschilok, Kenneth J. Takeuchi, and Esther S. Takeuchi ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.8b00244 • Publication Date (Web): 20 Feb 2018 Downloaded from http://pubs.acs.org on February 20, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Synthesis and Characterization of CuFe2O4 Nano/sub-micron Wires-Carbon Nanotube Composites as Binder-free Anodes for Li-ion Batteries Lei Wang1, David C. Bock2, Jing Li3, Eric A. Stach4,5, Amy C. Marschilok1,2,3*, Kenneth J. Takeuchi1,3*, and Esther S. Takeuchi1,2,3* Email: [email protected], [email protected], [email protected] 1

Department of Chemistry, State University of New York at Stony Brook, Stony Brook, NY 11794-3400

2

Energy Sciences Directorate, Interdisciplinary Sciences Building, Building 734, Brookhaven National Laboratory, Upton, NY 11973

3

Department of Materials Science and Chemical Engineering, State University of New York at Stony Brook, Stony Brook, NY 11794-2275

4

Center for Functional Nanomaterials, Brookhaven National Laboratory, Building 480, Upton, NY 11973

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

5

Page 2 of 49

Department of Materials Science and Engineering, University of Pennsylvania, Philadelphia, PA 19104

Keywords: copper ferrite, nano/submicron wires, binder-free electrode, Li-ion batteries, X-ray micro-fluorescence mapping, X-ray absorption spectroscopy

Abstract

A series of one-dimensional CuFe2O4 nano/sub-micron wires possessing different diameters, crystal phases, and crystal sizes have been successfully generated using a facile template-assisted coprecipitation reaction at room temperature, followed by a short post-annealing process. The diameter and the crystal structure of the resulting CuFe2O4 (CFO) wires were judiciously tuned by varying the pore size of the template and the post-annealing temperature, respectively. Carbon nanotubes (CNTs) were incorporated to generate CFO-CNT binder-free anodes, and multiple characterization techniques were employed with the goal of delineating the relationships between electrochemical behavior and the properties of both the CFO wires (crystal phase, wire diameter, crystal size) and the electrode architecture (binder-free vs. conventionally prepared approaches). The study reveals several notable findings. First, the crystal phase (cubic or tetragonal) did not influence the electrochemical behavior in this CFO system. Second, regarding crystallite size and wire diameter, CFO wires with larger crystallite sizes exhibit improved cycling stability, while wires possessing smaller diameters exhibiting higher capacities. Finally, the electrochemical behavior is strongly influenced by the electrode architecture, with CFO-CNT


2

Page 3 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


binder-free electrodes demonstrating significantly higher capacities and cycling stability compared to conventionally prepared coatings. The mechanism(s) associated with the high capacities under low current density but limited electrochemical reversibility of CFO electrodes under high current density were probed via x-ray absorption spectroscopy (XAS) mapping with sub-micron spatial resolution for the first time. Results suggest that the capacity of the binderfree electrodes under high rate is limited by the irreversible formation of Cu0, as well as limited reduction of Fe3+, to Fe2+ not Fe0. The results (1) shed fundamental insight into the reversibility of CuFe2O4 materials cycled at high current density and (2) demonstrate that a synergistic effort to control both active material morphology and electrode architecture is an effective strategy for optimizing electrochemical behavior.


3


Page 4 of 49

1. Introduction As one of the most widely deployed types of the batteries, rechargeable lithium ion batteries (LIBs) have achieved commercial success in a number of applications, especially mobile devices, due to their favorable electrochemical characteristics, such as high energy density, high voltage profile, and long cycle life.1 Conventional LIBs utilize carbon-based anode materials (typically graphite), which offer a limited theoretical capacity of 372 mAh/g. Another limitation of commercially utilized graphite anodes results from the deleterious lithium dendrite formation during overcharge process, owing to the Li-intercalation potential of the graphite anode which approaches 0 V versus Li/Li+.2-4 Therefore, designing new anode materials with higher capacity, long cycling stability, and opportunity for enhanced safety is an important challenge for battery research.5 Spinel metal ferrites have been proposed as promising alternatives to the graphite anode.6-7 The charge and discharge reaction mechanism occurring in these materials differs from conventional Li ion insertion/extraction, as it involves the formation and decomposition of lithium oxide (Li2O), along with the reduction and oxidation of metal nanoparticles.8 It is anticipated that greater reversible capacity, improved cyclability, and higher rate capability compared to pure iron oxides can be achieved via the suitable combination of different metal species in the metal ferrite structure.9 Moreover, the electrochemical performance of metal ferrites can be potentially controlled or further improved by judiciously tuning their morphology and structure. Among these materials, CuFe2O4 has generated high research interest because of its high theoretical capacity (895 mA h/g), low toxicity, as well as natural abundance.10 Several CuFe2O4 materials possessing different morphologies have been previously investigated as anodes for LIBs. Jin et al. synthesized hollow CuFe2O4 spheres encapsulated in


4

Page 5 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


carbon shells, which delivered a capacity of 480 mAh g−1, 91% retention of the initial capacity cycling at 100 mA g−1, after 50 cycles at a variable current density from 100 to 1600 mA g−1.11 Luo et al. used a combination of electrospinning and calcination processes to generate CuFe2O4 nanofibers, which delivered 536 mAh g−1, 454 mAh g−1, and 355 mAh g−1 at 200, 500 and 1000 mA g−1 after 50 cycles, respectively.12 Cu incorporated CuFe2O4 hexagonal platelets and graphene (Cu–CuFe2O4/G) composites were prepared through an one-pot and scalable hydrothermal strategy, and demonstrated a capacity of 672 mA h g-1 after 200 cycles at a current density of 1000 mA g-1.10 Wang and co-workers13 synthesized copper and copper ferrite nanoparticles anchored on a reduced graphene oxide (Cu/CuFe2O4@rGO) electrode which exhibited both high rate capability (560 mAh g−1 at 3200 mA g−1) and cycling stability (835 mAh g−1 over 100 cycles at 200 mA g−1), indicating a promising prospect for application in energy storage devices. Both cubic and tetragonal CuFe2O4 nanoparticles were selectively fabricated via a facile one-step solid state reaction. The cubic CuFe2O4 synthesized at 400 ◦C with smaller particle size and larger surface area exhibited superior discharge capacities, better cycling performance (950 mAh g−1 at 100 mA g−1 after 60 cycles), and higher rate capability.14 In contrast with previous literature, in this study we report the novel synthesis of onedimensional CuFe2O4 nano/sub-micron wires through a facile coprecipitation reaction at room temperature using polycarbonate (PC) membranes as templates. The method used is similar to that previously reported by Mcbean et al. for 200 nm perovskite LaNiO3 nanorods.15 Herein, the diameter and the crystal structure of the synthesized CuFe2O4 wires are judiciously tuned by varying the pore size of the PC membranes, and the annealing temperature, respectively. Compared with the zero-dimensional (0D) nanoparticles and two-dimensional (2D) nanoplatelets reported previously, the one-dimensional (1D) CuFe2O4 nanowires possess i) better strain


5


Page 6 of 49

relaxation, ii) large surface area, and iii) a continuous electron transport pathway, which can facilitate more efficient charge transport. Furthermore, in comparison with previously reported 1D CuFe2O4 electrospun motifs,12, 16 our method possesses more flexibility to tune the diameter and crystal structure of the resulting CuFe2O4 nanowires and does not require high voltage for preparation. We use the method to prepare CuFe2O4 wires with i) two different diameters, i.e. 50 nm and 200 nm, ii) two different crystal phases, i.e. cubic and tetragonal, and iii) three different crystallite sizes, i.e. 9, 14, and 20 nm. While our synthetic approach can be used to easily tune the crystal phase, size, and morphology, of CuFe2O4 nanowires, the material still suffers from inherently poor conductivity, severe electrode pulverization and electrical disconnection from the current collector owing to the volume change during the lithiation/delithiation, thus limiting its practical application in LIBs.10, 17 To mitigate these issues, one strategy is to eliminate the passive binder and current collector by anchoring the redox-active material on a conductive fibrous network.18-19 Carbon nanotubes (CNTs) are particularly promising conductive additives for improving the electrochemical behavior of CuFe2O4 composites as a result of their outstanding electrical conductivity and exceptional mechanical strength.20-23 In this report, we generate a CuFe2O4CNT binder-free self-standing anode, with the CNTs serving as not only the conductive network to facilitate electron and ion transport within the anchored CuFe2O4 wires, but also as a platform to maintain the structural integrity and stabilize the electrode during the cycling process.24 Through utilizing a series of CuFe2O4 nano/submicron wires incorporated into CFO-CNT binder-free anodes, this report investigated the relationship between electrochemical behavior and the properties associated with the CFO wires (crystal phase, wire diameter, crystal size) as well as the electrode architecture (binder-free vs. conventionally prepared approaches). XAS


6

Page 7 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


mapping with sub-micron spatial resolution was conducted to probe the oxidation states of the iron and copper centers after high rate discharge/charge. The results presented here demonstrate that control of the active material morphology and electrode architecture is critical for optimizing electrochemical behavior and the data shed fundamental insights into the reversibility of CuFe2O4 materials cycled at high current density.

2. Experimental 2.1 Synthesis of CuFe2O4 nano/sub-micron wires In a typical synthesis, track-etched polycarbonate (PC) membranes (Whatman) with pore size of 50 nm or 200 nm were immersed in Cu(NO3)2▪ 2.5H2O (J. T. Baker, 98%) and Fe(NO3)3▪ 9H2O (Sigma-Aldrich, 98%) aqueous stock solutions and were further sonicated for 2 min to allow saturation of the stoichiometric Cu2+ and Fe3+ ions within the inner pores of the PC template. Subsequently, the membranes were soaked in 2M NaOHaq (BDH) solution for 5 min, to form the Cu-Fe-OH precursors. The reacted membranes containing precursor wires were collected and subsequently sonicated to remove the superfluous, unwanted Cu-Fe-OH particles on the surface. The membranes were then dissolved and washed with N-methyl-2-pyrrolidone (NMP) (Sigma-Aldrich, 99%) and ethanol multiple times to collect the resulting Cu-Fe-OH products for further structural characterizations. To obtain the final CuFe2O4 products, the precursor wires were annealed within a temperature range of 500ff to 700 ff for 10 min. 2.2 Synthesis of CuFe2O4-CNT binder-free electrodes The pristine MWNTs (Cheap Tubes Inc., >95 wt%) were well-dispersed in ethanol by sonication. As-prepared CuFe2O4 wires were sonicated in ethanol before being added into the MWNT suspension and further sonicated. The final product was collected by vacuum filtration,


7


Page 8 of 49

washed with deionized water and ethanol, and dried to obtain the resulting CuFe2O4 -CNT electrode. Circular, free self-standing electrodes with 16 mm diameter were formed directly after the filtration process. The amount of CNT in the electrodes were controlled to be either 50% or 20% by weight. 2.3 Structural Characterizations X-ray Diffraction (XRD). XRD data of the various CuFe2O4 samples were acquired using a Rigaku SmartLab X-ray powder diffractometer. Cu Kα radiation was utilized with a BraggBrentano focusing geometry. The full width at half maximum (FWHM) of the (311) peak was determined using the Peak Fit software. Crystallite sizes were obtained using the Scherrer equation25 after correcting for instrumental broadening using a lanthanum hexaboride (LaB6) standard. Thermogravimetric analysis (TGA). To test the thermal behavior, TGA data were collected with a TGA Q500 instrument over the temperature range of 30 to 600 ff under an air atmosphere by using a heating rate of 5ff/min. Electron microscopy. The structure, morphology, and size of various CuFe2O4 wires and the representative CuFe2O4-CNT binder-free electrodes were probed using an analytical highresolution SEM (JEOL 7600F) instrument, operating at an accelerating voltage of 10 kV. To prepare samples for SEM characterization, fixed quantities of the CuFe2O4 wires were dispersed in ethanol and sonicated for ~1 min, prior to their deposition onto an underlying silicon (Si) wafer. The CuFe2O4-CNT binder-free electrodes samples were directly attached to a conductive carbon tape for subsequent image acquisitions. High resolution TEM characterization results, including data associated with morphology and selected area electron diffraction, were acquired with a JEOL JEM 2100F TEM instrument, equipped with a field-emission electron gun operating


8

Page 9 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


at 200 kV. For sample preparation, as-synthesized CuFe2O4 wires were dispersed and deposited onto a 300 mesh Cu grid coated with a lacey carbon film. Synchrotron-based X-ray microfluorescence (µ-XRF) mapping and X-ray absorption spectroscopy (XAS) measurements. µ-XRF mapping images of CuFe2O4-CNT electrodes recovered from coin cells were acquired using Beamline 5-ID of the National Synchrotron Light Source II (NSLS II) at Brookhaven National Laboratory. Post (dis)charge, electrodes were sealed between 50 µm thick polyimide film under Ar atmosphere to prevent environmental contamination from air and water. This beamline uses Kirkpatrick-Baez (KB) mirrors to produce a focused spot (0.5 µm x 0.5 µm) of hard X-rays with tunable energy achieved via Si (111) and Si (311) horizontal double crystal monochromator crystals.

For µ-XRF imaging, the

monochromator was calibrated using iron and copper metal foils and was set to a fixed energy of 9.8 keV to excite the Fe and Cu K-edges. Samples were oriented approximately 90° to the incident beam and rastered in the path of the beam by an XY stage while X-ray fluorescence was detected with a 3-element Vortex ME3 silicon drift detector (SDD) positioned at 90° to the incident beam. Coarse elemental maps were collected from a 100 µm x 100 µm sample area using a step size of 4 µm and a 1s acquisition time; high resolution maps were collected from a 50 µm x 50 µm sample area using a step size of 1 µm and a 1s acquisition time. Data acquisition and visualization were performed using Python-based beamline software developed for NSLS-II Beamline 5-ID. From the elemental maps, 0.5 µm x 0.5 µm areas were selected for XAS measurements with sub-micron spatial resolution using fluorescence geometry at both the Fe and Cu K-edges. In the pre-edge region (-250 to -15 eV below the edge) the incident beam energy was scanned using 10 eV steps and in the post edge region (30 eV to 850 eV above the edge), the step size was variable


9


Page 10 of 49

and gradually increased from 1 eV to 5.5 eV. Across the edge, a 0.35 eV step size was used for enhanced resolution. A 1s acquisition time was used at each data point. Fe and Cu metal foils were utilized for initial energy calibration. The extended X-ray absorption fine structure (EXAFS) spectra were aligned, averaged and normalized using Athena.26-27 The standard AUTOBK algorithm was used to remove the background below 1.0 Å. Both the Fe and Cu Kedge measurements were fit in Artemis with theoretical models generated from known crystal structures of CuFe2O4,28 Cu metal,29 and rock-salt FeO30 using FEFF6.26, 31-32 All spectra were fit using a k range of 2-8 Å-1 using a Hanning Fourier transform window with dk = 3 and were fit simultaneously using k, k2, and k3 weighting. An R-range of 1-3.0 Å was used, though it was extended to 1.0 – 3.5 Å to fully encompass the first and second coordination shells in the undischarged material. 2.4 Electrochemical Methods. Preparation of CuFe2O4 electrodes The binder-free electrodes were prepared by the aforementioned vacuum filtration processes. The electrodes contain either 50% CuFe2O4 wires and 50% CNTs, or 80% CuFe2O4 wires and 20% CNTs by weight. Conventional tape cast CuFe2O4 control electrodes were prepared on copper foil using a combination of 80% active material (CuFe2O4 wires), 10% carbon, and 10% polyvinylidene fluoride (PVDF) binder by weight. A CNT-only binder-free control sample was prepared by the same vacuum filtration process without the presence of CuFe2O4 wires. Electrochemical testing The cathodes prepared as noted above were used to assemble two-electrode stainless-steel experimental-type coin cells. The cells were assembled using a lithium foil anode and an


10

Page 11 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


electrolyte containing 1.0 M of lithium hexafluorophosphate (LiPF6) in ethylene carbonate and dimethyl carbonate (30/70, v/v). Cycling tests were conducted using a Maccor Battery Tester at 30°C. Lithium/copper-ferrite (Li/ CuFe2O4) cells were discharged and charged in a voltage window between 3.0-0.01V. A rate capability test was conducted with discharge and charge rates applied in the sequence of 200, 400, 800, 1600 mA/gCFO for 10 cycles each, and the 200 mA/gCFO used again for the next 10 cycles. To further assess the cycling stability, the various cells were cycled at 400 mA/gCFO for additional 100 cycles. Cyclic voltammetry (CV) data were collected using a two-electrode configuration wherein the reference and counter electrodes were both lithium metal. Voltage limits for the CV test were 0.01 V and 3.0 V at a scan rate of 0.1 mV/s. Electrochemical impedance spectroscopy (EIS) data were collected over a frequency range of 100 kHz to 10 mHz with a 10 mV amplitude at 30 ff. Analysis of the impedance measurements was conducted by using ZView software.

3. Results and Discussion 3.1 Crystal Structure Insights into 1D CuFe2O4 Nano/Submicron Wires. The purity and crystallinity of our as-prepared 1D CuFe2O4 wires were initially characterized using XRD. The intermediate Cu-Fe-OH precursor obtained after the template-directed precipitation reaction evidenced rather poor crystallinity, as depicted in Figure S1C (supporting information). The thermogravimetric analysis (TGA) profile (Figure S2) corresponding to the Cu-Fe-OH precursor indicated that it can be transformed into the crystalline CuFe2O4 phases by removing surface and crystal water molecules after heat treatment above 400 ˚C. Such data served as guidance and validation of our chosen annealing temperatures for the precursor wires.


11


Page 12 of 49

Specifically, Figure 1 displays the XRD profiles of the CuFe2O4 wires with 200 nm diameter annealed at either 500, 600, or 700˚C, respectively, as well as the 50 nm analogue annealed at 500 ˚C. These wires are labeled as CFO-200-500, CFO-200-600, CFO-200-700, and CFO-50500, respectively, in the following discussion, for abbreviation purpose. The XRD patterns suggested that the CFO-200-500, CFO-200-600, and CFO-50-500 possessed similar crystal structure, with peaks at 2θ = 18.3, 30.0, 35.5, 37.0, 43.4, 53.4, 57.3, and 62.6° ascribed to the (101), (220), (311), (222), (400), (422), (511), and (440) crystal planes of the CuFe2O4 with inverse cubic spinel structure (JCPDS 77-0010, space group Fd-3m).33 With higher annealing temperature of 700˚C, the crystal phase of CFO-200-700 was determined to be tetragonal CuFe2O4 (JCPDS 34-0425, space group I41/amd) structure.7,

34

No observable impurities were

detected in the as-prepared wires with either crystal phase, which indicated the feasibility of our synthesis method in terms of tuning the crystal phase of CuFe2O4 products through a one-step annealing process. Not surprisingly, the constituent crystallite size, as estimated by the Debye−Scherrer equation, increased with increasing annealing temperature and was calculated to be 9, 14, and 20 nm, for wires annealed at 500, 600, or 700˚C, respectively.


12

Page 13 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. XRD patterns of as-prepared 200 nm CuFe2O4 wires annealed at 500 ˚C (A), 600 ˚C (B), or 700˚C (C), as well as 50 nm CuFe2O4 wires annealed at 500 ˚C (D).


13


Page 14 of 49

3.2 Size and Morphology of the 1D CuFe2O4 Nano/Submicron Wires and the CuFe2O4CNT binder-free electrodes. The morphology and structure of the various as-synthesized CuFe2O4 wires were investigated by SEM. Typical SEM images of the Cu-Fe-OH precursor wires prepared using 200 nm and 50 nm PC templates were displayed in Figure SI 1A-B. The Cu-Fe-OH wires prepared using 200 nm possessed a diameter of 203 ± 16 nm, and a length of 2-10 µm, while the 50 nm analogue measured on average 82 ± 9 nm in diameter and 0.5-3 µm in length. The dimensions were obtained on the basis of the statistical measurements of several tens of wires pertaining to each of our samples. The representative SEM images of the final CuFe2O4 wires obtained after calcination process are illustrated in Figure 2A-D. The recurring, uniform 1D morphology of the final products was found to be independent of the annealing temperature used during the synthesis process. After annealing, the wires were measured 175 ± 30 nm, 167 ± 24 nm, 164 ± 22 nm, and 53 ± 8 nm in diameter; 3-7 µm, 4-8 µm, 2-6 µm, and 0.7-2.5 µm in length, for the CFO-200-500 (Figure 2A), CFO-200-600 (Figure 2B), CFO-200-700 (Figure 2C), and CFO-50500 wires (Figure 2D), respectively. It is evident that the final CuFe2O4 wires possessed narrower diameters as compared to those of the Cu-Fe-OH precursors before annealing (Figure SI 1A-B), possibly due to the loss of crystal water within the structures. In addition, the CFO-50500 nanowires evidenced much shorter length as compared with that of the wires prepared using the 200 nm template, possibly due to the hindered and interfered wire growth confined in the smaller 50 nm pores.


14

Page 15 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Typical SEM images of as-prepared CuFe2O4 wires, annealed at (A) 500°C, (B) 600°C, (C) 700°C using 200 nm template, as well as (D) CuFe2O4 wires derived from 50 nm template, followed by annealing process at 500ºC, respectively. Inset figures are images of the same materials obtained at higher magnifications. All wires maintained the overall 1D morphology after annealing. All the as-synthesized CuFe2O4 wires were further characterized using TEM to obtain additional insights into the nature of the morphology and crystallographic structure, as shown in Figure 3. The overall morphology (A-D) was confirmed to be nanowire-shape for four different CFO samples. The associated SAED patterns suggested that the CFO-200-500 (E), CFO-200600 (F), and CFO-50-500(H) samples were cubic strucures with space group Fd-3m, while the


15


Page 16 of 49

crystal structure was identified as tetragonal with space group I41/amd for the CFO-200-700 (G) sample.

Figure 3. Characterization of as-synthesized CuFe2O4 nanowires (left to right: CFO-200-500, CFO-200-600, CFO-200-700 and CFO-50-500): (A-D) overall morphology; (E-H) SEAD patterns and (I-L) their corresponding HRTEM images. The morphology and structure of the typical CFO-CNT binder-free electrode, generated using CFO-200-500 wires, were displayed in Figure 4. It is noted that the CFO wires were dispersed within the interconnected CNT network, both in the top-view images (Figure 4A-B), and in the side-view images (Figure 4C-D). The actual photographs of the electrode were displayed as inset figures in Figure 4A. The overall diameter of the electrode measured on average to be ~16 mm


16

Page 17 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(left inset figure), and the electrodes can be bent 90° without fragmentation (right inset figure), indicating a rather high structural flexibility. The average thickness and weight of the electrode was measured to be 0.15 ± 0.01 mm and 8.0 ± 0.2 mg.

Figure 4. Typical top-view (A-B) and side-view (C-D) SEM images of the resulting CFO-CNT binder-free electrodes prepared using CFO-200-500 wires. Inset figures are actual photographs of the flat and bent electrodes. 3.3 Electrochemical behaviors of the CFO-CNT binder-free electrodes. Cyclic voltammetry. Representative cyclic voltammograms of the four cell groups containing CFO-200-500, CFO-200-600, CFO-200-700, and CFO-50-500 active materials, for the first and second cycles are displayed in Figure 5, respectively. In the 1st cycle, the cathodic process was


17


Page 18 of 49

characterized by two major reduction peaks. The first peaks at 0.81-0.85 V is associated with the reductions of Cu2+ →Cu0 and Fe3+ →Fe0, as well as the formation of amorphous Li2O.11, 13 The second peak was located at 0.63-0.68V, which was not observed after the 1st cycle, indicating an irreversible loss of capacity commonly observed for metal oxide anodes, attributed to the formation of a solid electrolyte interface (SEI) layer and the decomposition of the electrolyte.35 It is worth noting that a distinctive peak at 1.26V was evidenced in the CFO-50-500 electrode, which had been previously identified as the transition of Cu2+ to Cu+, during the preceding intercalation process, corresponding to the reaction of CuFe2O4 + nLi+ + ne- → LinCuFe2O4.10, 36 In the 1st charge process, two broad anodic peaks appeared at 1.62-1.69V and 1.89-1.94V, which can be ascribed to the stepwise oxidation of Fe0/Fe3+ and Cu0/ Cu2+, respectively.14 After the 1st cycle, both the cathodic and anodic peaks shift to the higher voltage ranges due to the possible polarization of the electrodes, and the decrease of the cathodic peak intensity suggests the existence of a certain degree of irreversible redox reaction. During the subsequent cycles, only one cathodic peak located at 0.91-0.93V was observed, which might be the merged reduction processes of Fe3+ and Cu2+ into one peak.14 The successive anodic peaks are analogous to the initial scan with only minimal shift of the two peaks to higher voltages to 1.64-1.72 V, and 1.921.98V, respectively. It was noted that both the intensity and position of the redox peaks after 1st cycle were maintained, suggesting the discharge/charge process is reversible. The CV profile of the CNT-only control electrode was displayed in Figure S3. The voltammogram demonstrated the typical process associated to the intercalation of the Li+ in the graphitic structure of carbon materials, occurring at lower potential, i.e. between 0.3- 0.01V, and the process related to the formation of the SEI layer at the carbon surface, centered at about 0.7 - 0.6V.37


18

Page 19 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 5. CV results of various CFO-CNT binder-free electrodes for cycles 1–4 at a scan rate of 0.1 mV/s using active materials of: (A) CFO-200-500, (B) CFO-200-600, (C) CFO-200-700, and (D) CFO-50-500. Voltage profiles. Figure 6 depicts the representative discharge-charge voltage profiles of the various CFO-CNT electrodes acquired at a current density of 200 mA/gCFO within a cutoff voltage window of 0.01–3 V. During the first discharge, the binder-free electrodes with less crystalline active materials, i.e. CFO-200-500 (A) and CFO-50-500 (D), showed a smooth voltage decrease to 1.3 V first, then a short plateau-like step at about 1.3-1.2V, followed by a second voltage drop ending at 0.95V and a third voltage drop ending at 0.8V. The voltage drop


19


Page 20 of 49

to 1.2 V was more abrupt, followed by a more distinct flat plateau at ~1.24 V, in the other two samples with high crystallinity, i.e. CFO-200-600 (B) and CFO-200-700 (C). Three voltage plateaus at 1.24, 0.95, and 0.8 V can be clearly identified, corresponding to the reduction reaction of Cu2+ to Cu+, Fe3+ to Fe0, and Cu+ to Cu0, respectively,10 which was consistent with the aforementioned CV discussion. Such difference in the smoothness of the voltage curves with varying crystallite size of the active materials was observed previously in studies on Fe3O4 anode materials, where discharge slopes were smoother for the samples with smaller crystallite sizes while a distinct plateau existed in the samples with larger crystallite sizes.38-39 The large quantity of surface defects present in the samples with smaller crystallite size induce a reduction the band gap. In fact, the Fermi energy is gradually changing during the Li+ insertion process in the samples possessing smaller crystallite size, rather than an abrupt change as it would happen in the samples with larger crystallite size.40 According to the previous reports,11, 14 the preceding intercalation process at ~1.24V corresponded to the reaction of CuFe2O4 + nLi+ + ne- → LinCuFe2O4, where the XRD structure of the electrode materials remained to be cubic when discharging to 1.1V, corresponding to a formation of an insertion phase such as LinCuFe2O4 during discharge.14 No distinct difference was observed in the voltage profiles of CFO-200-600 (Figure 6B) and CFO-200-700 (Figure 6C), corresponding to the cubic and tetragonal CFO phases, respectively. Such observation was in agreement with the finding in a previous report, where similar oxidation/reduction processes were noted for both the cubic and tetragonal CFO particles.14 It was noted that the MWNTs contributed additional capacity to the initial discharge at 0.85V (Figure 6E), wherein the large irreversible capacities measured for the four CFO-CNTs (Figure 6A-D) at cycle 2 were consistent with the formation of a solid electrolyte interface (SEI) at low voltage. Amongst all four CFO-CNT binder-free electrodes prepared, the CFO-50-500


20

Page 21 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


sample delivered the highest initial capacity at ~2500 mAh/g and ~1360 mAh/g in cycle 2. By contrast, the analogous capacities in the 1st and 2nd cycles of CFO-200-500, CFO-200-600, and CFO-200-700 binder-free electrodes were 2400 and 1226 mAh/g, 2345 and 1180 mAh/g, as well as 2330 and 1170 mAh/g. For comparison, the voltage profile of the electrode generated by conventional coating method, using CFO-200-500 as active materials, was depicted in Figure 6F. Such electrode demonstrated significant capacity decrease in cycle 2, delivering only 377 mA/g.

Figure 6. 1st and 2nd discharge and charge voltage profiles of (A) CFO-200-500, (B) CFO-200600, (C) CFO-200-700, (D) CFO-50-500, and (E) CNT-only binder-free electrodes, as well as (F) CFO-200-500 electrode generated using conventional coating method.


21


Page 22 of 49

Rate capability, cycling stability and EIS. Rate capability measurements were carried out on the various electrodes at a series of current densities (Figure 7A and C). A total of 50 cycles using sequential rates of 200, 400, 800, 1600, and 200 mA/g were tested. With respect to the discharge (lithiation) capacity, the CFO-50-500 sample delivered 1360 mAh/g in cycle 2 and 1223 mAh/g in cycle 10 (90% retention). The other analogous samples yielded lower capacities. For example, the CFO-200-500 sample delivered 1242 mAh/g in cycle 2 and 1067 mAh/g in cycle 10 (86% retention). The CFO-200-600 and CFO-200-700 gave rise to similar capacities, which were ~1208 mAh/g and 1033 mAh/g (86% retention), as well as 1185 mAh/g and 1026 mAh/g (87% retention), in cycle 2 and cycle 10, respectively. With the discharge and charge rates increased to 400 mA/g, the CFO-50-500 sample possessing CFO wires with narrower diameters exhibited higher capacity, which was 1010 mAh/g at cycle 20, as compared to those of the other three binder-free analogues containing wires with larger diameters. In fact, our 1D CFO wires with an average diameter of 50 nm delivered much higher capacities at current densities of 200 mA/g and 400 mA/g, when compared with the 0D cubic CFO nanoparticle analogue with similar size (~50-100 nm), which demonstrated capacities of 960, 880 and 810 mAh/g at current densities of 100, 200, and 500 mA/g.14 The enhanced capacity can be attributed to the synergistic effect of i) the unique 1D morphology of the CFO active materials, which provided a continuous electron transport pathway, facilitating more efficient charge transport, and ii) the CNT-based binder-free electrode design, which lead to a stable conductive platform without the inert polymer binders. When the discharge/charge rates were increased to 800 mA/g and 1600 mA/g, the CFO-50-500 delivered similar capacities as those of the CFO-200-600 and CFO-200-700, while the CFO-200-500 gave rise to the lowest capacities. When changing the rate back to 200 mA/g, the capacity of CFO-50-500 was recovered to be 853 mA/g (63% retention from cycle 2-


22

Page 23 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


50). The CFO-200-500, CFO-200-600, and CFO-200-700 samples delivered 663 mAh/g (53% retention), 741 mAh/g (61% retention), and 760 mAh/g (64% retention), respectively. In comparison, the CFO-200-500 coating control sample delivered 957 mAh/g in the 1st cycle, and faded to 314 mAh/g in cycle 5 (33% retention). It is clear that the binder-free electrode unambiguously demonstrated better capacity retention as compared with the analogue generated using conventional coating method. The CNT-only binder-free control sample exhibited an initial capacity of 1435 mAh/g, which faded to 319 mAh/g in cycle 10 at a current density of 100 mA/gCNT. The CNT capacities at current rates of 200 mA/gCNT and 400 mA/gCNT were 206 mAh/g and 132 mAh/g, respectively, suggesting the capacity contributions from CNTs at these rates are small. The corresponding charge capacity of the various electrodes were displayed in Figure 7C, which demonstrated a similar overall trend as that of the discharge capacity presented in Figure 7A, suggesting that the CFO-50-500 binder-free electrode (BFE) delivered the highest charge capacities amongst all the electrodes at current densities of 200 mA/gCFO and 400 mA/g CFO.

It is noted that the initial coulombic efficiency of all the BFEs is rather low, denoting 44%

for CFO-200-500, CFO-200-600, as well as CFO-200-700, and 46% for CFO-50-500 BFEs, which can be attributed to the large irreversible capacity loss during the first charging– discharging process and the resulting low initial coulombic efficiency (21%) of the CNT component in the BFEs, mainly deriving from the trapped Li+ ions within the closed micropores in the CNT matrix and solid electrolyte interphase (SEI) layer formation.41-42 Starting from the second discharge-charge cycle, the coulombic efficiency dramatically increased to 83% for all four BFE electrodes, and subsequently stabilized at ~91% at cycle 10 and beyond. The initial coulombic efficiency of the CFO-200-500 conventional coating electrode was 54%, which was higher than that of the BFE analogue, due to the absence of the CNTs. The irreversible capacity


23


Page 24 of 49

loss herein was mainly attributed to the irreversibility of Cu0 upon delithiation process, which has been confirmed in a previous study on CFO particles.7 To further investigate the cycling stability of the binder-free electrodes, various electrodes containing different CFO materials were subsequently cycled for additional 100 cycles after the rate capability test, at a constant current density of 400 mA/gCFO. The four electrodes demonstrated decrease in both discharge and charge capacities during cycle 51-90. In terms of the discharge capacities (Figure 7B), after 150 cycles, the CFO-50-500, CFO-200-500, CFO200-600, and CFO-200-700 electrodes maintained capacities of 325 mAh/g, 206 mAh/g, 289 mAh/g, and 286 mAh/g, respectively, which corresponded to capacity retentions of 48%, 46%, 46%, 45% from cycle 51 to 150, as well as retentions of 32%, 24%, 33%, 32% from cycle 20 to 150. In terms of the charge (delithiation) capacity (Figure 7D), similar retention rates were observed as that of the aforementioned discharge (lithiation) capacity, with the corresponding coulombic efficiency of 87%, 79%, 83%, and 83% for the CFO-50-500, CFO-200-500, CFO200-600, and CFO-200-700 electrodes, respectively.

Our findings in electrochemical

characterizations of various CFO binder-free electrodes can be summarized as follows: 1) with the same wire diameter of 200 nm, the CFO-200-600 and CFO-200-700 samples possessing larger crystallite sizes demonstrated better cycling stability after extended cycles, although the low-crystalline CFO-200-500 sample delivered higher initial capacities at low discharge rates. Such crystallite size-dependent electrochemical behaviors were noted elsewhere in separate but relevant studies on LiV3O843-44 and MgMn2O445 electrode materials with different crystallite sizes, where the more crystallized materials usually delivers less capacity during initial discharge-charge cycle, but with better capacity retention over extended cycling when compared to the less crystallized materials. It is believed that reducing crystallite size of the electrode


24

Page 25 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


materials plays a major role in the initial capacity, where smaller crystallite size leads to shortened charge diffusion distances,46 however, cycling performance improvement of the electrode material is strongly dependent on the structural stability, where the more crystalline materials with larger crystallite sizes tend to maintain relatively intact structures during the extended (de)lithiation processes;47 2) with wires possessing similar crystallinity, the smaller wires, i.e. CFO-50-500, gave rise to both higher capacities and better cycling stability when compared with the larger analog, i.e. CFO-200-500, at all discharge/charge rates; and 3) the wires with different crystal phases, i.e. cubic CFO-200-600 and tetragonal CFO-200-700, showed no appreciable differences in terms of their electrochemical behavior. To provide better understanding of the significant capacity decrease of the binder-free electrodes at high discharge rates, such as 1600 mA/gCFO, more detailed discussion is provided in the following section 3.4.


25


Page 26 of 49

Figure 7. Specific discharge (lithiation) (A) and charge (delithiation) (C) capacity versus cycle number for binder-free CFO-200-500 (black), CFO-200-600 (red), CFO-200-700 (blue), CFO50-500 (pink), CNT-only-control (green), and CFO-200-500-coating-control (orange) materials


26

Page 27 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


measured at 200, 400, 800, 1600, and 200 mA/gCFO discharge/charge current density. (B and D) Additional discharge and charge cycling stability measurements of the four binder-free electrodes at a discharge/charge current density of 400 mA/gCFO. (E and F) Nyquist plot of the alternating-current impedance response of CFO-200-500 (black), CFO-200-600 (red), CFO-200700 (blue), CFO-50-500 (pink), and CFO-200-500-coating-control (orange) cells (E) before discharge and (F) after 50 cycles. In order to gain further insights into the reaction kinetics, EIS data were collected before and after 50 galvanostatic cycles, as displayed in Figure 7E and F. All impedance spectra were fit to the equivalent circuit model displayed in Figure S4A. In this treatment, R1 represents ohmic resistances, with R2 representing the charge transfer resistance Rct; CPE is the constant phase element, corresponding to the electrode double-layer capacitance; and Wo is the Warburg impedance. In the low frequency region, the Warburg coefficient (σw), which is inversely proportional to the ion-diffusion coefficient, was determined from the slope of Z’ versus ω−1/2. All of the fitted results are summarized in Table S1 in the supporting information. Before discharge, the CFO-200-500 coating materials displayed the lowest Rct of 7.7 Ω amongst all samples, while the four binder-free electrodes displayed similar Rct ranging from 9.9 Ω to 16.3 Ω, with the CFO-50-500 material possessing the lowest Rct value. However, after 50 cycles, the Rct value drastically increased to 1070 Ω for the CFO-200-500 coating control material, which resulted in the drastic capacity fading after 5 cycles depicted in Figure 7A. By contrast, all the binder-free electrodes only showed slight increase in the Rct values, denoting 31 Ω, 85 Ω, 19 Ω, 26 Ω, which confirmed that our binder-free electrodes indeed facilitated faster charge transfer owing to the enhanced dispersion of CFO wires in the CNT network, as well as the elimination of insulating binder materials. With respect to the calculated σw values before cycling, the CFO-


27


Page 28 of 49

200-500 coating material gave rise to the highest value of 6.06 Ωs-0.5, indicating a much slower effective Li-ion diffusion. All the binder-free electrodes demonstrated similar σw values, with the CFO-50-500 binder-free electrode possessing 0.60 Ωs-0.5, which was only one tenth of that calculated for the coating control sample (6.06 Ωs-0.5), suggesting a much-enhanced Li-ion transport behavior. After 50 cycles, the CFO-50-500 still maintained the lowest value of the σw amongst all the samples (Table S1), indicating the efficient ion-transport were retained. In addition, the porosities of the various CFO electrodes were calculated using the formula displayed in the supporting information, in order to probe the dependence of ion diffusion on the electrode porosities. It is worth noting that the CFO-conventional coating electrode evinced a porosity of 48.7%, while the binder-free electrodes demonstrated a much higher porosity of 89.9%, unambiguously suggesting that the ion diffusion process is more efficient within a porous binder-free electrode. The effect of CFO/CNT ratios. In order to further assess the role of CNTs in the resulting binder-free electrodes, a CFO/CNT weight ratio of 4:1, different from the 1:1 ratio used in the samples in the previous discussion, was used to generate another binder-free electrode using CFO-50-500 as active materials. Such electrode possessed a slightly smaller porosity of 88.4% than 89.9% of the previous BFEs containing 50wt% CNTs. Both rate capability and EIS tests were conducted. The corresponding data were presented in Figure S5 in the supporting information. By decreasing the amount of CNTs to 20% in the binder-free electrode, the discharge capacities dropped to 748, 596, 335, 102, and 524 mAh/g at cycle 10, 20, 30, 40 and 50 (Figure S5A) with a capacity retention rate of 61% from cycle 2 to 50, which were lower than those observed from the electrode containing 50% CNTs in Figure 7A, delivering 1223, 1010, 594, 316, and 854 mAh/g at cycle 10, 20, 30, 40 and 50, with 63% capacity retained from cycle 2


28

Page 29 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


to 50. An initial coulombic efficiency of 60% was observed for this electrode, which further increased to 88% in cycle 2, and subsequently stabilized at 93% from cycle 10 and beyond. A slight increase in the coulombic efficiency was noted in the electrode containing less CNTs, possibly due to the lower amount of CNT content, which was the primary origin of the low initial coulombic efficiency. The impedance spectra before and after 50 cycles were displayed in Figure S5B, and were fit to the equivalent circuit model presented in Figure S4B. The fitted results were summarized in Table S1 in the supporting information. The Rct value increased from 17 Ω to 452 Ω after 50 cycles, both values were larger than those of the analogue with higher CNT content, indicating slower charge transfer behaviors. In terms of the ion-transport behaviors, the σw values were rather similar for the CFO-50-500 4:1 with the CFO-50-500 1:1 sample before cycling. After 50 cycles, the σw value was about 2.5 times higher than that of the CFO-50-500 1:1 sample, suggesting that the Li-ion transport kinetics were much slower with the lower content of CNTs in the electrode. Such data suggested that although the CFO-50-CNT 4:1 electrode with lower CNT ratio demonstrated a slightly enhanced coulombic efficiency, the smaller porosity, larger charge transfer resistance, and slower Li-ion diffusion rate, especially after the rate capability test for 50 cycles, collectively lead to the lower delivered capacity and worse cycling performance at various current densities tested, highlighting the importance of the sufficient amount of CNT in the resulting binder-free electrodes. 3.4 Understanding the delivered capacity of the binder-free electrodes at high current density. It was noted in the rate capability tests (Figure 7A) that the binder-free electrodes delivered high capacity at a low current density such as 200 mA/g, but showed significantly lower capacity at higher current density, i.e. 1600 mA/g. To provide a better understanding of the significant


29


Page 30 of 49

capacity decrease of the binder-free electrodes at high discharge rates, several CFO-200-500CNT binder-free electrodes were tested for one cycle within the voltage window of 3.0-0.3V at a discharge/charge current density of either 200 mA/g (low-rate condition) or 1600 mA/g (highrate condition). The corresponding voltage profiles of these cells are provided in Figure S6 in the supporting information. These electrodes were recovered from coin cells to obtain SEM images, in order to probe the structural evolution and any possible morphological alterations of the active materials after the lithiation and delithiation processes at different current densities. In addition, the recovered electrodes tested at the current density of 1600 mA/g were further investigated by synchrotron based x-ray fluorescence elemental mapping in combination with X-ray absorption spectroscopy (XANES, EXAFS) with sub-micron resolution. Structural evolution after discharge/charge processes. The CFO-200-500 wires within the electrode before cycling, displayed in Figure 8A, possessed a relatively smooth surface. After discharging to 0.3V at low current density of 200 mA/g, the wires were wrapped by a uniform, gel-like film, which could be attributed to the formation of SEI layers (Figure 8B). By contrast, a minimal amount, if any, of the SEI layer was observed in the same electrode discharged at high current density of 1600 mA/g (Figure 8C), when compared to that of the analogue discharged at 200 mA/g (Figure S6). It was noted that after charging the electrode to 3V at 1600 mA/g, the surface of the wires remained rough, with nanoparticles uniformly distributed on top (Figure 8D), which could result from a certain degree of Cu extrusion reaction occurring during the discharge.48


30

Page 31 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 8. Typical SEM images of (A) CFO-200-500 binder-free electrodes before cycling, and CFO-200-500 binder-free electrodes after discharging at 200 mA/g for one cycle (B), discharging at 1600 mA/g for one cycle (C), as well as discharging and charging back to 3V at 1600 mA/g for one cycle (D), respectively. Inset figures are images of the same materials obtained at higher magnifications. The Cu nanoparticles are highlighted in Figure 8D inset with red circles.

Ex-situ electrode characterization – XRF mapping and sub-micron resolution X-ray absorption spectroscopy (XANES, EXAFS). The CFO-200-500-CNT binder-free electrode during higher rate cycling was further investigated by (dis)charging Li/CFO cells at a current density of


31


Page 32 of 49

1600 mA/g and analyzing the recovered electrodes by synchrotron based x-ray fluorescence elemental mapping in combination with X-ray absorption spectroscopy (XANES, EXAFS) with sub-micron resolution.

These techniques provide oxidation state and structural evolution

information as a function of spatial position. Copper K-edge elemental maps of the recovered electrodes are shown in Figure 9, and corresponding iron K-edge maps are shown in Supplemental Figure S7. The maps for all three electrode states (undischarged, discharged, charged) show regions of higher and lower fluorescence signal intensity within the analyzed areas. Using the density of the CFO within the composite electrodes (~0.13 g CFO /cm3), the calculated transmission at the maximum electrode thickness (150 micron) at the Fe and Cu absorption edge energies (7112 eV and 8979 eV for Fe and Cu, respectively) as well as K(alpha) fluorescence energies (6405 eV and 8046 eV for Fe and Cu, respectively) are all >50%.49 Thus, the signal observed in the fluorescence maps is representative of CFO particles throughout the entire electrode, rather than just at the electrode surface. The signal intensity varies by up to 100% from low to high intensity regions, indicating that the loading of the CFO wires integrated into CNTs is not homogeneous as a function of spatial location. SEM images of the binder-free electrodes shown in Figure 4 support this hypothesis – although the CFO wires are dispersed amongst the CNTs, there are distinct regions of higher and lower CFO concentration.


32

Page 33 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 9. (a, c, e) Coarse resolution and (b, d, f) fine resolution copper elemental maps of 200500-CNT binder-free electrodes which were (a, b) undischarged, (c, d) discharged to 0.3 V at 1600 mA/g and (e, f) discharged and recharged to 3.0 V at 1600 mA/g. White box insets in the coarse maps indicate the area selected for the fine resolution map. Corresponding iron elemental maps of the same electrode regions are included in Figure S7 in the supporting information. Arrows signify 0.5 µm x 0.5 µm spots where µ-XAS measurements were collected; black arrows


33


Page 34 of 49

indicate selected areas at the center of the high intensity (high concentration of CFO) regions within the binder-free electrode, white arrows indicate measurements collected at the surface of higher intensity regions. X-ray absorption spectroscopy measurements at both the Cu and Fe K-edges were collected at various positions from the elemental maps of the electrodes, indicated by the arrows in Figure 9 and Figure S7, in order to assess possible differences in oxidation state and active material structure as a function of spatial location. Measurements were taken at both the center (as indicated by black arrows) as well as the surface (as indicated by white arrows) of high intensity regions within the composite. Corresponding XANES data are presented in Figure 10a and b. At the Cu edge, the XANES spectra of the undischarged electrode material has an edge energy (defined as the maximum of the first derivative of xµ(E)) of 8994 eV, in good agreement with a previous XAS study of CuFe2O4.7 In the discharged state, for both center and surface locations, there is a significant shift in the edge energy to 8982 eV. The similarity in both edge position and profile of the electrode and the Cu metal powder reference signify that in the discharged state, the Cu2+ ions in CFO wires have converted to Cu metal. Upon recharge to 3.0 V, there is minimal shift in the edge position, thus the Cu atoms remain in the Cu metal state and do not contribute to electrochemical reversibility of the active material. No significant difference in oxidation state is observed between measurements collected at the center vs. surface of the high intensity regions from the XRF elemental maps, suggesting that the phase transition to Cu0 occurred homogenously throughout the binder-free electrode.


34

Page 35 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 10. XANES spectra of CFO-200-500-CNT binder-free electrode in the undischarged state and during discharge (lithiation) and charge (delithiation) at both the (a) Cu edge and (b) Fe edge. The undischarged, discharged, and charged states are indicated as black, red, and blue lines, respectively.

At the Fe K-edge, XANES spectra show that the undischarged electrode has an edge position of 7126.5 eV, in agreement with a previous report.7 When the electrode is reduced, the edge position shifts ca. 2.5 eV to lower energy, indicating partial reduction of Fe3+ ions in the starting structure. Although discharge to 0.3V is expected to result in conversion of the Fe3+ ions to Fe metal at low rates, it is clear from comparison to a Fe metal powder standard that full reduction to Fe metal did not occur at the high rate used to discharge the electrode. This finding is in good agreement with the galvanostatic cycling results at 1600 mA/g (Figure 7A), where the electrodes exhibit significantly lower capacity than the theoretical value. The low capacity utilization indicates insufficient lithium ion and electron transport through the active CFO wires at the high rate. Upon oxidation to 3.0 V, the edge position shifts back to ca. 7126 eV, signifying that the Fe ions return to a 3+ oxidation state consistent with the undischarged material. There is a slight


35


Page 36 of 49

variation in edge position (ca. 0.5 eV) between measurements collected at the surface vs. the center of the high intensity regions on the XRF map in the charged state, with the surface slightly more oxidized compared to the center. This result suggests that delithiation may be more effective in electrode regions with reduced CFO concentration such as the surface of aggregates. Since no reversibility is observed in the Cu edge XANES, the transition between Fe2+ on reduction and Fe3+ on oxidation is hypothesized to be the primary redox reaction that occurs under high rate cycling (1600 mA/g) of the electrode, resulting in approximately 2 electron equivalents (ca. 225 mAh/g) of reversible capacity. Additional capacity observed in Figure 7A is likely due to (de)lithiation of the CNT component of the electrode (Figure S6). The extended X-ray abosorption fine structure (EXAFS) region of the XAS spectra were also analyzed to provide structural evolution information complimentary to the oxidation state information provided by the XANES. EXAFS results are presented in Figure 11a and b, as Rspace plots (Fourier transforms of k2|χ(R)|). Distances indicated are not corrected for phase shifts, thus peak positions are ca. 0.3 Å shorter than the phase-corrected interatomic distances determined from modeling. The spectra of the as-prepared electrodes before cycling at both the Fe edge and Cu edge exhibit two coordination shells, with the first shell resulting from contributions from neighboring oxygen atoms, and the second shell resulting primarily from neighboring Cu and Fe atoms. Upon discharge, the EXAFS spectra at the Cu edge (Figure 11a) consist of a single broad peak centered at 2.1 Å, representing conversion to the Cu metal phase. In agreement with the XANES result, the Cu edge does not reconvert to an oxide like structure upon oxidization, though a small level of oxidation is apparent from the appearance of a shoulder at ca. 1.6 Å. In contrast to the EXAFS spectra at the Cu edge, at the Fe edge (Figure 11b), the electrode does not convert to a metal structure upon discharge, but rather remains as an oxide


36

Page 37 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


phase with a broadened first shell and significantly reduced amplitude in the second shell. The atomic structure of the iron phase after charge is similar to that after discharge. For each state measured, EXAFS spectra are similar for measurements collected at the center vs. the surface of the high intensity regions on the XRF map, indicating the (dis)charge processes are homogenous within the measured area.


37


Page 38 of 49

Figure 11. (a and b) k2 weighted |χ(R)| of CFO electrodes in the undischarged state and during discharge (lithiation) and charge (delithiation) at both the (a) Cu edge and (b) Fe edge. The undischarged, discharged, and charged states are indicated as black, red, and blue lines, respectively. Distances indicated are not corrected for phase shifts, thus peak positions are ca. 0.3 – 0.4 Å shorter than the phase-corrected interatomic distances determined from modeling. (c and d) Selected interatomic distances determined from EXAFS modeling results for CuFe2O4 electrodes in the undischarged state and during discharge (lithiation) and charge (delithiation) at both the (c) Cu edge and (d) Fe edge. The EXAFS spectra were modeled using Artemis to quantify structural changes at the atomic level. Results are presented as interatomic distances for selected paths in Figure 11c and d. Due to the similarity in the fitted results between measurements collected in center vs. surface regions, only fitting results for center position measurements are shown in the charts. Full fitting results for all spectra are presented in the supporting information, Tables S2 – S8. For undischarged electrodes, a theoretical model of cubic phase CFO was utilized and resulted in Rfactors < 2.0 for fits at both the Fe and Cu edges. Upon discharge, the Cu edge spectra were fit using metallic Cu with a primary scattering path at ca. 2.5Å; contributions from neighboring O and Fe atoms were no longer observed, suggesting that the Cu metal was completely removed from the starting structure. At the Fe edge, contributions from Feoctahedral-Cu and FeoctahedralFetetrahedral were no longer observed (evident in the loss of second shell amplitude in Figure 11b) and a general iron oxide model based on a FeO (rock-salt) crystal structure30 was utilized to fit the data. The model includes a Fe-O contribution at ca. 1.9 Å and a Fe-Fe contribution at ca. 3.0 Å, with these distances similar to the Fe-O and Fe-Fe distances from the octahedrally coordinated Fe central atom in the initial spinel structure. The EXAFS fits suggest that the


38

Page 39 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


discharged iron oxide phase is highly nanocrystalline, with coordination numbers of ca. 2.4 and ca. 1.6 for O and Fe near neighbors respectively, compared to an expected value of 6 for bulk FeO. On oxidation to 3.0 V, modeling clearly indicates that the initial spinel structure was not recovered.

Cu atoms remain in the Cu metal phase, though a small amount of copper is

coordinated to oxygen on recharge; this is likely due to surface oxidation of the Cu metal particles since XANES analysis shows no appreciable shift in the oxidation state. At the Fe edge, Fe atoms remain in octahedral-like geometry with single Fe-O and Fe-Fe path contributions. The number of oxygen near neighbors increases slightly, from ca. 2.4 to ca. 3.4, in good agreement with XANES results which shows oxidation state reversibility between discharged and charged states. Our results agreed well with a previous study on CuFe2O4 particles7, where XAS measurements were used to demonstrate that upon charge, the original spinel structure was not restored, and the copper remained in its reduced form, while the iron can be oxidized back to near its as-synthesized +3 valence state when not fully reduced to Fe metal. Thus, the XAS results indicate that the limited electrochemical reversibility of the binder-free electrodes at high rate (1600 mA/g) arises from (1) the irreversibility of the Cu, which does not convert back from Cu metal after discharge, and (2) the limited reversibility of Fe atoms, which do not fully reduce to Fe metal and at this discharge level, instead transition between Fe3+ and Fe2+ while maintaining a nanocrystalline iron oxide phase.

4. Conclusion


39


Page 40 of 49

In this report, we described the synthesis and characterization of a series of CuFe2O4 nano/submicron wires derived from a template-directed method, as well as the resulting CFO-CNT composites as binder-free anodes. The synthesis protocol presented herein demonstrates outstanding synthetic flexibility wherein monodispersed CFO wires with tunable diameters, crystal phases, and crystal size can be correspondingly generated simply by adjusting the pore size of the templates, as well as the post-annealing conditions. With the objective of determining the relationships between electrochemical behavior and the properties of both the CFO wires (crystal phase, wire diameter, crystal size) as well as the electrode architecture (binder free vs. conventionally prepared approaches), the study reveals several significant findings: (1) crystal phase is not critical to the electrochemical behavior in this CFO system, as both cubic and tetragonal CFO wires showed no appreciable differences in delivered capacity; (2) larger crystallite sizes demonstrate better cycling stability after extended cycles; (3) smaller diameter wires result in both higher capacities when compared with the larger analog due to improved electrolyte ion access; and (4) electrochemical behavior is strongly influenced by the electrode architecture, with the binder free electrodes exhibiting higher capacities and better cycling stability at all tested rates, as compared with the control sample generated using conventional coating method. Rate capability testing also shows that while the binder-free electrodes deliver high capacity at a low current density, reversibility is significantly reduced at higher cycling rates. To understand the mechanism resulting in the limited electrochemical reversibility of the binder-free electrodes under high current density, XAS mapping with sub-micron spatial resolution was conducted in the charged and discharged state for binder free electrodes cycled at 1600 mA/g. Notably, results show that at high discharge rates, the Cu2+ metal center is fully reduced to Cu0 during lithiation,


40

Page 41 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


while in contrast Fe3+ ions only reduce to Fe2+. Furthermore, upon delithiation, the initial spinel structure is not recovered, Cu atoms remain primarily as Cu0 metal, and Fe2+ is oxidized to a Fe3+ consistent with the undischarged material. Thus, the XAS results suggest that the limited electrochemical reversibility of CuFe2O4 at high rate is due to both (1) the irreversibility of the Cu metal on delithiation and (2) the limited reversibility of Fe atoms which transition between Fe3+ and Fe2+ while maintaining a nanocrystalline iron oxide phase.

5. Supporting Information Additional

structural

characterizations

of

Cu-Fe-OH

intermediates,

electrochemical

characterization data of CNT-only as well as CFO-50-500 4:1 binder-free electrodes, electrode porosity calculation, iron XRF mapping, and EXAFS modeling results of all the CFO electrodes analyzed. 6. Acknowledgements All of the work described in these studies was funded as part of the Center for Mesoscale Transport Properties (m2M), an Energy Frontier Research Center supported by the U.S. Department of Energy, Office of Science, Basic Energy Sciences, under award #DE-SC0012673. Experimental research characterizations were carried out in part at the Center for Functional Nanomaterials, Brookhaven National Laboratory, an Office of Science User Facility, which is supported by the U.S. Department of Energy, Office of Basic Energy Sciences, under Contract No. DE-SC0012704. This research used beamline 5-ID of the National Synchrotron Light Source II, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Brookhaven National Laboratory under Contract No. DE-SC0012704.


41


Page 42 of 49

References 1.

Wu, F.; Li, X.; Wang, Z.; Guo, H., Petal-Like Li4Ti5O12–TiO2 Nanosheets as High-

Performance Anode Materials for Li-Ion Batteries. Nanoscale 2013, 5, 6936. 2.

Zhu, G. N.; Wang, Y. G.; Xia, Y. Y., Ti-Based Compounds as Anode Materials for Li-

Ion Batteries. Energy Environ. Sci. 2012, 5, 6652-6667. 3.

Tang, Y. F.; Yang, L.; Fang, S. H.; Qiu, Z., Li4Ti5O12 Hollow Microspheres Assembled

by Nanosheets as an Anode Material for High-Rate Lithium Ion Batteries. Electrochim. Acta 2009, 54, 6244-6249. 4.

Liu, W. J.; Shao, D.; Luo, G. E.; Gao, Q. Z.; Yan, G. J.; He, J. R.; Chen, D. Y.; Yu, X.

Y.; Fang, Y. P., Mesoporous Spinel Li4Ti5O12 Nanoparticles for High Rate Lithium-Ion Battery Anodes. Electrochim. Acta 2014, 133, 578-582. 5.

Shi, Y.; Zhou, X. Y.; Zhang, J.; Bruck, A. M.; Bond, A. C.; Marschilok, A. C.; Takeuchi,

K. J.; Takeuchi, E. S.; Yu, G. H., Nanostructured Conductive Polymer Gels as a General Framework Material to Improve Electrochemical Performance of Cathode Materials in Li-Ion Batteries. Nano Lett. 2017, 17, 1906-1914. 6.

Guo, H. Y.; Zhang, Y. M.; Marschilok, A. C.; Takeuchi, K. J.; Takeuchi, E. S.; Liu, P., A

First Principles Study of Spinel ZnFe2O4 for Electrode Materials in Lithium-Ion Batteries. Phys. Chem. Chem. Phys. 2017, 19, 26322-26329.


42

Page 43 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


7.

Cama, C. A.; Pelliccione, C. J.; Brady, A. B.; Li, J.; Stach, E. A.; Wang, J.; Wang, J.;

Takeuchi, E. S.; Takeuchi, K. J.; Marschilok, A. C., Redox Chemistry of a Binary Transition Metal Oxide (AB2O4): A Study of the Cu2+/Cu0 and Fe3+/Fe0 Interconversions Observed Upon Lithiation in a CuFe2O4 Battery Using X-Ray Absorption Spectroscopy. Phys. Chem. Chem. Phys. 2016, 18, 16930-16940. 8.

Indhrajothi, R.; Prakash, I.; Venkateswarlu, M.; Satyanarayana, N., Binder Effect on the

Battery Performance of Mesoporous Copper Ferrite Nanoparticles with Grain Boundaries as Anode Materials. RSC Adv. 2014, 4, 44089-44099. 9.

Su, Q. M.; Wang, S. X.; Yao, L. B.; Li, H. J.; Du, G. H.; Ye, H. Q.; Fang, Y. Z., Study on

the Electrochemical Reaction Mechanism of ZnFe2O4 by in Situ Transmission Electron Microscopy. Sci. Rep. 2016, 6, 28197. 10. Dong, Y. C.; Chui, Y. S.; Ma, R. G.; Cao, C. W.; Cheng, H.; Li, Y. Y.; Zapien, J. A., One-Pot Scalable Synthesis of Cu-CuFe2O4/Graphene Composites as Anode Materials for Lithium-Ion Batteries with Enhanced Lithium Storage Properties. J. Mater. Chem. A 2014, 2, 13892-13897. 11. Jin, L. M.; Qiu, Y. C.; Deng, H.; Li, W. S.; Li, H.; Yang, S. H., Hollow CuFe2O4 Spheres Encapsulated in Carbon Shells as an Anode Material for Rechargeable Lithium-Ion Batteries. Electrochim. Acta 2011, 56, 9127-9132. 12. Luo, L.; Cui, R. R.; Qiao, H.; Chen, K.; Fei, Y. Q.; Li, D. W.; Pang, Z. Y.; Liu, K.; Wei, Q. F., High Lithium Electroactivity of Electrospun CuFe2O4 Nanofibers as Anode Material for Lithium-Ion Batteries. Electrochim. Acta 2014, 144, 85-91.


43


Page 44 of 49

13. Wang, J. Y.; Deng, Q. L.; Li, M. J.; Jiang, K.; Zhang, J. Z.; Hu, Z. G.; Chu, J. H., Copper Ferrites@Reduced Graphene Oxide Anode Materials for Advanced Lithium Storage Applications. Sci. Rep. 2017, 7, 8903. 14. Xing, Z.; Ju, Z. C.; Yang, J.; Xu, H. Y.; Qian, Y. T., One-Step Solid State Reaction to Selectively Fabricate Cubic and Tetragonal CuFe2O4 Anode Material for High Power Lithium Ion Batteries. Electrochim. Acta 2013, 102, 51-57. 15. McBean, C. L.; Liu, H.; Scofield, M. E.; Li, L.; Wang, L.; Bernstein, A.; Wong, S. S., Generalizable, Electroless, Template-Assisted Synthesis and Electrocatalytic Mechanistic Understanding of Perovskite LaNiO3 Nanorods as Viable, Supportless Oxygen Evolution Reaction Catalysts in Alkaline Media. ACS Appl. Mater. Interfaces 2017, 9, 24634-24648. 16. Zhao, J. X.; Cheng, Y. L.; Yan, X. B.; Sun, D. F.; Zhu, F. L.; Xue, Q. J., Magnetic and Electrochemical Properties of CuFe2O4 Hollow Fibers Fabricated by Simple Electrospinning and Direct Annealing. CrystEngComm 2012, 14, 5879-5885. 17. Shi, Y.; Zhou, X. Y.; Yu, G. H., Material and Structural Design of Novel Binder Systems for High-Energy, High-Power Lithium-Ion Batteries. Accounts Chem. Res. 2017, 50, 2642-2652. 18. Poyraz, A. S.; Huang, J.; Wu, L.; Bock, D. C.; Zhu, Y.; Marschilok, A. C.; Takeuchi, K. J.; Takeuchi, E. S., Potassium-Based Α-Manganese Dioxide Nanofiber Binder-Free SelfSupporting Electrodes: A Design Strategy for High Energy Density Batteries. Energy Technol. 2016, 4, 1358-1368. 19. Shi, Y.; Zhang, J.; Bruck, A. M.; Zhang, Y. M.; Li, J.; Stach, E. A.; Takeuchi, K. J.; Marschilok, A. C.; Takeuchi, E. S.; Yu, G. H., A Tunable 3d Nanostructured Conductive Gel


44

Page 45 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Framework Electrode for High-Performance Lithium Ion Batteries. Adv. Mater. 2017, 29, 1603922. 20. Choi, J.-H.; Ryu, W.-H.; Park, K.; Jo, J.-D.; Jo, S.-M.; Lim, D.-S.; Kim, I.-D., MultiLayer Electrode with Nano-Li4Ti5O12 Aggregates Sandwiched between Carbon Nanotube and Graphene Networks for High Power Li-Ion Batteries. Sci. Rep. 2014, 4, 7334. 21. Zhang, W. D.; Xu, B.; Jiang, L. C., Functional Hybrid Materials Based on Carbon Nanotubes and Metal Oxides. J. Mater. Chem. 2010, 20, 6383-6391. 22. Marschilok, A.; Lee, C. Y.; Subramanian, A.; Takeuchi, K. J.; Takeuchi, E. S., Carbon Nanotube Substrate Electrodes for Lightweight, Long-Life Rechargeable Batteries. Energy Environ. Sci. 2011, 4, 2943-2951. 23. Wang, L.; Zhang, Y. M.; McBean, C. L.; Scofield, M. E.; Yin, J. F.; Marschilok, A. C.; Takeuchi, K. J.; Takeuchi, E. S.; Wong, S. S., Understanding the Effect of Preparative Approaches in the Formation of "Flower-Like" Li4Ti5O12 -Multiwalled Carbon Nanotube Composite motifs with Performance as High-Rate Anode Materials for Li-Ion Battery Applications. J. Electrochem. Soc. 2017, 164, A524-A534. 24. Xia, G. F.; Li, N.; Li, D. Y.; Liu, R. Q.; Wang, C.; Li, Q.; Lu, X. J.; Spendelow, J. S.; Zhang, J. L.; Wu, G., Graphene/Fe2O3/SnO2 Ternary Nanocomposites as a High-Performance Anode for Lithium Ion Batteries. ACS Appl. Mater. Interfaces 2013, 5, 8607-8614. 25. Scherrer, P., Nachr. Ges. Wiss. Gottingen 1918, 96. 26. Ravel, B.; Newville, M., Athena, Artemis, Hephaestus: Data Analysis for X-Ray Absorption Spectroscopy Using Ifeffit. J. Synchrotron Rad. 2005, 12, 537-541.


45


Page 46 of 49

27. Newville, M., Ifeffit: Interactive Xafs Analysis and Feff Fitting. J. Synchrotron Rad. 2001, 8, 322-324. 28. Prince, E.; Treuting, R. G., The Structure of Tetragonal Copper Ferrite. Acta Crystallogr. 1956, 9, 1025-1028. 29. Otte, H. M., Lattice Parameter Determination with an X-Ray Spectrogoniometer by the Debye-Scherrer Method and the Effect of Specimen Condition. J. Appl. Phys. 1961, 32, 15361546. 30. Jette, E. R.; Foote, F., An X-Ray Study of the Wuestite (FeO) Solid Solutions. J. Chem. Phys. 1933, 1, 29-36. 31. de Leon, J. M.; Rehr, J. J.; Zabinsky, S. I.; Albers, R. C., Abinitio Curved-Wave X-RayAbsorption Fine-Structure. Phys. Rev. B 1991, 44, 4146-4156. 32. Rehr, J. J.; de Leon, J. M.; Zabinsky, S. I.; Albers, R. C., Theoretical X-Ray Absorption Fine-Structure Standards. J. Am. Chem. Soc. 1991, 113, 5135-5140. 33. Zhang, Q. Y.; Verde, M. G.; Seo, J. K.; Li, X.; Meng, Y. S., Structural and Electrochemical Properties of Gd-Doped Li4Ti5O12 as Anode Material with Improved Rate Capability for Lithium-Ion Batteries. J. Power Sources 2015, 280, 355-362. 34. Xu, H. Y.; Wang, Y. P.; Zheng, L.; Duan, X. H.; Wang, L. H.; Yang, J.; Qian, Y. T., Preparation of Polypyrrole-Coated CuFe2O4 and Their Improved Electrochemical Performance as Lithium-Ion Anodes. J. Energy Chem. 2014, 23, 354-357.


46

Page 47 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


35. Cheng, Y. H.; Chen, G.; Wu, H. B.; Zhu, M. F.; Lu, Y. F., Use of Regenerated Cellulose to Direct Hetero-Assembly of Nanoparticles with Carbon Nanotubes for Producing Flexible Battery Anodes. J. Mater. Chem. A 2017, 5, 13944-13949. 36. Wang, J.; Liu, Y. C.; Wang, S. Y.; Guo, X. T.; Liu, Y. P., Facile Fabrication of PomponLike Hierarchical CuO Hollow Microspheres for High-Performance Lithium-Ion Batteries. J. Mater. Chem. A 2014, 2, 1224-1229. 37. Agostini, M.; Brutti, S.; Hassoun, J., High Voltage Li-Ion Battery Using Exfoliated Graphite/Graphene Nanosheets Anode. ACS Appl. Mater. Interfaces 2016, 8, 10850-10857. 38. Menard, M. C.; Takeuchi, K. J.; Marschilok, A. C.; Takeuchi, E. S., Electrochemical Discharge of Nanocrystalline Magnetite: Structure Analysis Using X-Ray Diffraction and X-Ray Absorption Spectroscopy. Phys. Chem. Chem. Phys. 2013, 15, 18539-18548. 39. Menard, M. C.; Marschilok, A. C.; Takeuchi, K. J.; Takeuchi, E. S., Variation in the Iron Oxidation States of Magnetite Nanocrystals as a Function of Crystallite Size: The Impact on Electrochemical Capacity. Electrochim. Acta 2013, 94, 320-326. 40. Bruck, A. M.; Cama, C. A.; Gannett, C. N.; Marschilok, A. C.; Takeuchi, E. S.; Takeuchi, K. J., Nanocrystalline Iron Oxide Based Electroactive Materials in Lithium Ion Batteries: The Critical Role of Crystallite Size, Morphology, and Electrode Heterostructure on Battery Relevant Electrochemistry. Inorg. Chem. Front. 2016, 3, 26-40. 41. Guerin, K.; Fevrier-Bouvier, A.; Flandrois, S.; Simon, B.; Biensan, P., On the Irreversible Capacities of Disordered Carbons in Lithium-Ion Rechargeable Batteries. Electrochim. Acta 2000, 45, 1607-1615.


47


Page 48 of 49

42. Beguin, F.; Chevallier, F.; Vix-Guterl, C.; Saadallah, S.; Bertagna, V.; Rouzaud, J. N.; Frackowiak, E., Correlation of the Irreversible Lithium Capacity with the Active Surface Area of Modified Carbons. Carbon 2005, 43, 2160-2167. 43. Zhang, Q.; Bruck, A. M.; Bock, D. C.; Li, J.; Sarbada, V.; Hull, R.; Stach, E. A.; Takeuchi, K. J.; Takeuchi, E. S.; Marschilok, A. C., Visualization of Structural Evolution and Phase Distribution of a Lithium Vanadium Oxide (Li1.1V3O8) Electrode Via an Operando and in Situ Energy Dispersive X-Ray Diffraction Technique. Phys. Chem. Chem. Phys. 2017, 19, 14160-14169. 44. Xu, H. Q.; Zhang, H. L.; Zhang, T.; Pan, Q. Y.; Gui, Y. H., Influence of Heat-Treatment Temperature on Crystal Structure, Morphology and Electrochemical Properties of LiV3O8 Prepared by Hydrothermal Reaction. J. Alloy Compd. 2009, 467, 327-331. 45. Yin, J. F.; Brady, A. B.; Takeuchi, E. S.; Marschilok, A. C.; Takeuchi, K. J., MagnesiumIon Battery-Relevant Electrochemistry of MgMn2O4: Crystallite Size Effects and the Notable Role of Electrolyte Water Content. Chem. Commun. 2017, 53, 3665-3668. 46. Wang, L.; Yue, S. Y.; Zhang, Q.; Zhang, Y. M.; Li, Y. R.; Lewis, C. S.; Takeuchi, K. J.; Marschilok, A. C.; Takeuchi, E. S.; Wong, S. S., Morphological and Chemical Tuning of HighEnergy-Density Metal Oxides for Lithium Ion Battery Electrode Applications. ACS Energy Lett. 2017, 2, 1465-1478. 47. He, H. N.; Jin, G. H.; Wang, H. Y.; Huang, X. B.; Chen, Z. H.; Sun, D.; Tang, Y. G., Annealed NaV3O8 Nanowires with Good Cycling Stability as a Novel Cathode for Na-Ion Batteries. J. Mater. Chem. A 2014, 2, 3563-3570.


48

Page 49 of 49 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


48. Li, J.; Cama, C.; Marschilok, A. C.; Takeuchi, K. J.; Takeuchi, E. S.; Stach, E. A., Visualization of Phase Evolution of Ternary Spinel Transition Metal Oxides (CuFe2O4) During Lithiation. Microsc. Microanal. 2017, 23, 2022-2023. 49. Henke, B. L.; Gullikson, E. M.; Davis, J. C., X-Ray Interactions: Photoabsorption, Scattering, Transmission, and Reflection at E = 50-30,000 Ev, Z = 1-92. At. Data Nucl. Data Tables 1993, 54, 181-342.

TOC figure


49

sub-micron Wires

Recommend Documents