Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1271

The O2 electrode performance in the Li-O2 battery JIA LIU

ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2015

ISSN 1651-6214 ISBN 978-91-554-9294-6 urn:nbn:se:uu:diva-259589

Dissertation presented at Uppsala University to be publicly examined in Häggsalen, Ångströmlaboratoriet, Lägerhyddsvägen 1, Uppsala, Friday, 25 September 2015 at 13:15 for the degree of Doctor of Philosophy. The examination will be conducted in English. Faculty examiner: Senior Scientist Poul Norby (Department of Energy Conversion and Storage, Technical University of Denmark, Denmark). Abstract Liu, J. 2015. The O2 electrode performance in the Li-O2 battery. Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1271. 73 pp. Uppsala: Uppsala universitet. ISBN 978-91-554-9294-6. Li-O2 batteries have been attracting increasing attention and R&D efforts as promising power sources for electric vehicles (EVs) due to their significantly higher theoretical energy densities compared to conventional Li-ion batteries. The research presented in this thesis covers the investigation of factors influencing the decomposition of Li2O2, the development of highly active electrocatalysts, and the design of low-cost and easy-operation binder-free O2 electrodes for LiO2 batteries. Being the main technique, SR-PXD was used both as a continuous light source to advance the electrochemical decomposition of Li2O2 under the X-ray illumination and an operando tool that allowed us to probe the degradation of Li2O2. Since XRD was intensively used in my thesis work, the effect of X-ray irradiation on the stability of Li2O2 was studied. The accelerating effect of X-rays on the electrochemical decomposition of Li2O2 was, for the first time, explored. The electrochemical decomposition rate of Li2O2 was proportional to the X-ray intensity used. It is proposed that the decomposition might involve a three-step reaction with [Li2O2]x+ and Li2-xO2* as intermediates, which followed pseudo-zero-order kinetics. Then, three electrocatalysts (Pt/MNT, Ru/MNT and Li2C8H2O6) were developed, which exhibited good electrocatalytic performances during the OER. Their activities were evaluated by following the Li2O2 decomposition in electrodes during the charging processes. In addition, the time-resolved OER kinetics for the electrocatalyst-containing LiO2 cells charged galvanostatically and potentiostatically was systematically investigated using operando SR-PXD. It was found that a small amount of Pt or Ru decoration on the MNTs enhanced the OER efficiency in a Li-O2 cell. The Li2O2 decomposition of an electrode with 5 wt % Pt/MNT, 2 wt% Ru/MNT or Li2C8H2O6 in a Li-O2 cell followed pseudo-zero-order kinetics. Finally, a novel binder-free NCPE for Li-O2 batteries was presented. It displayed a bird’s nest microstructure, which could provide the self-standing electrode with considerable mechanic durability, fast O2 diffusion and enough space for the discharge product deposition. The NCPE contained N-containing functional groups, which may promote the electrochemical reactions. Keywords: Li-oxygen battery, X-ray irradiation, Electrocatalyst, Synchrotron radiation powder X-ray diffraction, Time-resolved kinetics, Binder-free cathode, Bird’s nest microstructure. Jia Liu, Department of Chemistry - Ångström, Structural Chemistry, Box 538, Uppsala University, SE-751 21 Uppsala, Sweden. © Jia Liu 2015 ISSN 1651-6214 ISBN 978-91-554-9294-6 urn:nbn:se:uu:diva-259589 (http://urn.kb.se/resolve?urn=urn:nbn:se:uu:diva-259589)

To my grandparents and parents

List of Papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals. I

Accelerated electrochemical decomposition of Li2O2 under X-ray illumination Liu, J., Roberts, M., Younesi, R., Dahbi, M., Edström, K., Gustafsson, T., Zhu, J. The Journal of Physical Chemistry Letters, 4 (2013) 4045-4050.

II

Pt/α-MnO2 nanotube: A highly active electrocatalysts for Li-O2 battery Liu, J., Younesi, R., Gustafsson, T., Edström, K., Zhu, J. Nano Energy, 10 (2014) 19-27.

III Ru/α-MnO2 nanotube catalysts for electrochemical decomposition of Li2O2 in Li-O2 batteries Liu, J., Roberts, M., Ma, Y., Gustafsson, T., Edström, K., Zhu, J. In manuscript. IV An organic catalyst for Li-O2 batteries: Dilithium quinone-1,4dicarboxylate Liu, J., Renault, S., Brandell, D,. Gustafsson, T., Edström, K., Zhu, J. ChemSusChem, 8 (2015) 2198-2203. V

Binder-free nitrogen-doped carbon paper electrodes derived from polypyrrole/cellulose composite for Li-O2 batteries Liu, J., Wang, Z., Younesi, R., Nyholm, L., Edström, K., Zhu, J. In manuscript.

Reprints were made with permission from the respective publishers.

Papers not included in the thesis

I

Increased cycling efficiency of lithium anodes in dimethyl sulfoxide electrolytes for use in Li-O2 batteries Roberts, M., Younesi, R., Richardson, W., Liu, J., Gustafsson, T., Zhu, J., Edström, K. ECS Electrochemistry Letters, 3 (2014) A62-A65.

II

A Ru-Co hybrid material based on a molecular photosensitizer and a heterogeneous catalyst for light-driven water oxidation Wang, H-Y., Liu, J., Zhu, J., Styring, S., Ott, S., Thapper, A. Physical Chemistry Chemical Physics, 16 (2014) 3661-3669.

III Fluorine-doped tin oxide nanocrystal/reduced graphene oxide composites as lithium ion battery anode materials with high capacity and cycling stability Xu, H., Shi, L., Wang, Z., Liu, J., Zhu, J., Zhao, Y., Zhang, M., Yuan, S. In manuscript.

Comments on my contribution to the appended papers in this thesis: I. Involved in planning, major part of the experimental work, and writing of the manuscript. II. Involved in planning, major part of the experimental work, and writing of the manuscript. III. Major contribution to planning, major part of the experimental work, and writing of the manuscript. IV. Major contribution to planning, performed some characterizations and electrochemical measurements, and major contribution to writing of the manuscript. V. Major contribution to planning, major part of the experimental work, and writing of the manuscript.

Contents

1. Introduction.............................................................................................. 11 1.1 Renewable energies ........................................................................... 11 1.2 Li-O2 batteries.................................................................................... 12 1.2.1 A brief history of Li-O2 batteries............................................. 12 1.2.2 ORR and OER ......................................................................... 13 1.2.3 Electrolytes .............................................................................. 14 1.2.4 The O2 electrodes .................................................................... 16 1.2.4.1 O2 electrode materials ................................................. 16 1.2.4.2 Architecture of O2 electrodes ..................................... 18 1.2.5 Lithium electrodes ................................................................... 18 1.3 The scope of this thesis ...................................................................... 19 2 Experimental ............................................................................................ 21 2.1 Materials synthesis ............................................................................ 21 2.1.1 Preparation of MNT ................................................................ 21 2.1.2 Preparation of Pt/MNT and Ru/MNT ..................................... 21 2.1.3 Preparation of Li2C8H2O6 ........................................................ 22 2.1.4 Preparation of NCPE ............................................................... 22 2.2 Preparation of electrodes ................................................................... 24 2.2.1 Preparation of a commercial Li2O2-filled electrode used for the charge process.................................................................... 24 2.2.2 Preparation of the O2 electrode used for the discharge and charge processes ...................................................................... 25 2.3 Assembly and electrochemical testing of the Li-O2 cell.................... 25 2.4 Characterization ................................................................................. 26 2.4.1 XRD ........................................................................................ 26 3 Summary of the results and discussions .................................................. 31 3.1 Study of the electrochemical decomposition of Li2O2 under continuous X-ray illumination ........................................................... 31 3.2 Study of the highly active electrocatalysts for Li-O2 batteries .......... 37 3.2.1 Study of x wt% Pt/MNT (x=0, 1, 5, and 10) electrocatalysts ........................................................................ 38 3.2.2 Study of x wt% Ru/MNT (x=0, 0.5, and 2) electrocatalysts ........................................................................ 46 3.2.3 Study of a novel organic electrocatalyst (Li2C8H2O6) ............. 50 3.3 Study of a binder-free NCPE ............................................................. 53

4 Concluding remarks and further outlook ................................................. 60 5 Sammanfattning på Svenska .................................................................... 63 6 Acknowledgements.................................................................................. 65 7 References................................................................................................ 67

Abbreviations

BET DMSO EDX LiBOB Li2C8H2O6 LiClO4 LiPF6 MNT NCPE OER ORR PC Pt/MNT PVdF Ru/MNT SEI SEM SR-PXD TEM TG XPS XRD

Brunauer-Emmett-Teller Dimethyl sulfoxide Energy dispersive X-ray spectroscopy Lithium bis(oxalate)borate Dilithium quinone-1,4-dicarboxylate Lithium perchlorate Lithium hexafluorophosphate α-MnO2 nanotube Nitrogen-doped carbon paper electrode Oxygen evolution reaction Oxygen reduction reaction Propylene carbonate α-MnO2 nanotube deposited by Pt nanoparticles Polyvinylidene difluoride α-MnO2 nanotube deposited by Ru nanoparticles Solid electrolyte interphase layer Scanning electron microscopy Synchrotron radiation powder X-ray diffraction Transmission electron spectroscopy Thermogravimetry X-ray photoelectron spectroscopy X-ray diffraction

1. Introduction

1.1 Renewable energies With the growth of the population and parts of the world becoming more industrialized, experts predict a 56 % increase in the demand for usable energy by 20401. Most countries rely heavily on fossil fuels (e.g., oil, coal, petroleum, and natural gas) for their energy consumption. Fossil fuels are limited and non-renewable, and as such will eventually dwindle away. Some estimations claim that oil reserves will only last for another twenty years if the increase in consumption continues1. In addition, carbon dioxide (CO2) emissions from the combustion of fossil fuels are recognized to be the main factor for global warming, which is considered as one of the world’s most pressing challenges today. As alternatives to fossil fuels, a lot of development has gone into renewable energies from a wide variety of resources including solar, tidal, wind, hydro, biomass, and geothermal energies. Using renewable energy has several potential benefits, such as a decreasing dependency on fossil fuel, reduction in CO2 emissions, diversification of energy supplies, and stimulation of employment by creating jobs in the new “green” technologies. The EU has set out some directions for the development of renewable energies2. Sweden is one of the leading countries in using renewable energy2. With almost 50 % of the power generation derived from renewable resources, a growing number of energy strategies have been formulated in Sweden. The wind is, however, not always blowing and the sun is not always shining. For effective use of renewable energy it must be coupled to energy storage. Therefore, energy storage devices with high efficiencies are sorely needed. One such technology is a battery, where the stored chemical energy can be converted into the electrical energy through the electrochemical reactions. In 1800, the battery was firstly invented by Alessandro Volta who found that a continuous and stable current could be generated in a voltaic pile system3. A rechargeable battery has the advantage of repeated use, which has been developed for the green transportations (e.g., electric vehicles) with the aim of reducing the dependency for the fuels and environmental pollutions.

11

1.2 Li-O2 batteries The lithium-ion battery was commercialized in 1991 by Sony. This triggered a rapid development of portable electronics during the last two decades. Due to the relatively high energy density of a rechargeable battery, it is currently not only the prime choice as an energy storage device for electric and hybrid electric vehicles, it is now also being tested for large scale storage to support for renewable energy production. The participants in the electrochemical reactions of a lithium-ion battery are the positive and negative electrodes with an organic electrolyte providing a conductive medium for Li+ to transfer between the electrodes4. Up until now, many fundamental and applied advances have been achieved in this field. However, even though the theoretical capacity of the electrode material could be achieved, the energy density (energy per unit volume) and specific energy (energy per unit mass) of state-of-the-art lithium-ion batteries are still too low to meet the demands of all-electric vehicles in the long term5. Therefore, an increasing amount of research has recently been devoted to energy storage systems that can go beyond the limits of a Li-ion battery. One such device is the Li-O2 battery, which has attracted considerable attention in recent years due to its high specific energy6.

1.2.1 A brief history of Li-O2 batteries The major advantage of a Li-O2 battery is the extremely high energy density (the theoretical specific energy is as high as 11,430 W h·kg ), which could rival that of the traditional gasoline-powered engine. The first Li-O2 battery was introduced by Abraham and Jiang7. A typical Li-O2 battery consists of a lithium metal negative electrode, an organic electrolyte, and a porous O2 positive electrode (Figure 1). The electrochemical reactions are based on the oxidation of lithium at the negative electrode and reduction of oxygen at the positive electrode to induce the current flow. 2Li + O2 ⇄ Li2O2 is the key electrochemical reaction postulated to take place in a Li-O2 battery7,8. Shortly thereafter some researchers studied the influence of different electrolytes on the discharge capacities9-11. Interests in Li-O2 batteries increased substantially after 2006 when Ogasawara et al., reported the possibility of improved cyclability of a Li-O2 cell using a liquid organic electrolyte (1 M LiPF6/PC) and manganese dioxide (MnO2) as an electrocatalyst embedded in the porous cathode12. They claimed that the employment of electrocatalysts could lower the overpotential and improve the cyclability13. Since then, the Li-O2 battery has gained a lot of interest as a possible advanced electrochemical energy-storage technology with high energy density. At an industrial level, IBM launched the “Battery 500” project in 2009, aiming to develop a Li-O2 battery which could ensure a driving range of 500 miles. In 2012, Central Glass and Asahi Kasei joined this project. So far, the 12

scientists have realized that the Li-O2 battery is much more complex than initially expected. However, the excitement surrounding Li-O2 batteries has grown roughly exponentially due to the participation of hundreds of researchers8.

Figure 1. A schematic picture of the Li-O2 battery.

1.2.2 ORR and OER The ideal electrochemical reaction in a Li-O2 battery is shown in reaction 1 (R1): 2Li+ + O2 + 2e- ⇄ Li2O2

(E0 = 2.96 V vs. Li+/Li)

(R1)

with a discharge process named the ORR described by the forward direction and a charge process named the OER described by the reverse direction14,15. However, the oxygen electrochemistry in a real battery is much more complicated. For example, the electrochemistry involves multiple elementary reactions. For the ORR, it was commonly suggested that the O2 dissolved in the electrolyte is first reduced to the super oxide ion ( ·), which combines with Li+ to form LiO2 on the surface of the positive electrode (R2). Being an intermediate product, LiO2 further converts chemically and/or electrochemically to lithium peroxide (Li2O2), as shown in R3 and R416,17.

13

Li + O2 + e- → LiO2 2LiO2 → Li2O2 + O2 2LiO2 + Li+ + e- → Li2O2

(E0 = 3.0 V vs. Li+/Li) (chemically) (electrochemically)

(R2) (R3) (R4)

Equally, for the OER, as presented in R5, most researchers believe that the decomposition of Li2O2 occurs directly via a two-phase transition without participation of any intermediate. This reaction is limited by a de-nucleation process, which requires an overcoming of a significant reaction barrier, leading to a high charging voltage8,14; Li2O2 → 2Li+ + O2 + 2e-

(R5)

Besides the complex multiple electrochemical reactions, the ORR and OER can be influenced by the electrode materials, O2 pressure, and electrolyte18-20. In addition, the insulating and insoluble discharge products further complicate the Li-O2 electrochemistry, since these products could clog the surface of the porous electrode and hinder the OER process21,22. Furthermore, the highly reactive intermediates (i.e., LiO2 and/or Li2O) might react with the electrolyte and/or binder, resulting in a low coulombic efficiency and poor cyclability23,24.

1.2.3 Electrolytes The electrolyte has a profound influence on the electrochemical reactions in a Li-O2 battery. So far, despite many efforts devoted to electrolyte studies in Li-O2 batteries, no electrolyte has been able to satisfy all chemical criteria, such as stability during cell cycling. Thus, searching for a stable electrolyte is the most pressing challenge for the future of Li-O2 batteries. It has been well established that a good electrolyte candidate should have the following characteristic25: 1) High stability, especially in the presence of O2 and at a high charging potentials; 2) High O2 solubility and diffusivity; 3) Low volatility and vapour pressure to guarantee a long term cell operation; 4) Low viscosity to support fast Li+ transport. Here, several representative electrolytes in Li-O2 batteries will be introduced. Carbonate-based electrolyte solvents Organic carbonates (e,g., PC, diethyl carbonate (DEC), ethylmethyl carbonate (EMC) and ethylene carbonate (EC)) possess wide and stable electrochemical windows, which have been extensively applied in Li-ion batteries26,27. Inspired by this, PC was first employed by Ogasawara et al., in 200612. Their work triggered a rapid development of carbonate-based electrolytes for Li-O2 batteries. Within the following years, PC-based electrolyte was even used as a standard electrolyte. However, in 2010, Mizuno et al., reported that the oxygen radicals generated during the ORR could attack the 14

carbonate solvent, resulting in the decomposition of the electrolyte. It was found that in fact in this solvent lithium carbonate (Li2CO3) and lithium alkylcarbonates (RO-(C=O)-OLi) were formed as the discharge products rather than the expected Li2O228. After this finding, several studies were devoted to studying the stability of the electrolytes in association with the mechanism of Li-O2 batteries using a range of different techniques, such as Fourier transform infrared spectroscopy (FTIR)29, XRD30, Raman spectroscopy31, nuclear magnetic resonance (NMR)32, XPS33, and differential electrochemical mass spectrometry (DEMS)34. Ether-based electrolyte solvents As a result of the instability of carbonate-based electrolyte systems, attention was instead shifted to ether-based electrolytes. The advantages include stability to oxidation potentials up to 4.0 V vs. Li+/Li, low-cost, and safety. An ether-based electrolyte was first employed in the Li-O2 battery by Read in 200635. Subsequently, Bryantesev et al., demonstrated with density functional theory (DFT) that ether-based electrolytes were more suitable than the carbonate-based ones by comparing their electrochemical stabilities36,37. However, McCloskey et al., identified the decomposition of dimethoxy ethane (DME) electrolyte in a Li-O2 cell by studying the gases consumed and produced during battery cycling18. To date, although the growing experimental and theoretical evidence has revealed that ether-based electrolytes are not completely suitable18,23,38, they are still the most studied electrolytes in Li-O2 systems. DMSO solvent DMSO was investigated in 2012 by Peng et al., who reported that a Li-O2 cell containing 1 M LiClO4/DMSO as the electrolyte coupled with a nanoporous gold positive electrode exhibited a cyclability of over 100 cycles with low capacity fading39. Since then DMSO as electrolyte has gained much attention and is considered as the most stable electrolyte40-42. However, it should be noted that DMSO is not a definite champion, besides the instability of lithium metal in contact with DMSO, its decomposition and the reactivity with Li2O2 have been confirmed33,43. Other electrolytes Several other solvents have been studied in Li-O2 batteries, such as ionic liquids44, acetonitrile33,45, dimethylformamide46, N-methyl-2-pyrrolidone (NMP)47, and N, N-dimethylacetamide (DMA)48. So far, the commonly used electrolytes for Li-O2 batteries are flammable liquid electrolytes, which could lead to the risks of explosion and fire within the Li-O2 concept of combining metallic lithium and O2. Being a non-flammable and solid-state medium, solid electrolytes, are predicted to overcome the shortcomings of the liquid electrolytes mentioned above. They are thus, considered as a 15

promising alternative. Until now, there are a few reports on the use of solid polymer electrolyte49-52 and glass-ceramic electrolyte53,54 in Li-O2 batteries. Stability of salts Apart from providing the required conductivity, lithium salts play an important role in interface phenomena, such as SEI stability and conductivity, and current collector passivation. Several lithium salts including LiPF66,55, LiBF456, LiClO457, LiSO3CF358,59, LiB(CN)423, LiN(SO2CF3)231, and LiBOB19 have been investigated in Li-O2 batteries. Analogous to solvents, the main issue for the lithium salts is their degradations resulting from reactions with Li2O2 and/or intermediate products56. In summary, the stability of the electrolyte is a major challenge in Li-O2 batteries. Although great progress has been made, the work to find a suitable electrolyte is not over. Doubtless we need to explore the electrolytes further in the quest for higher stability.

1.2.4 The O2 electrodes So far, an O2 electrode in a Li-O2 battery is a porous electrode, providing the solid-liquid-gas tri-phase and electron transferring regions for the electrochemical reactions. Therefore, cell performances including energy storage capacity, rate capacity, and cycle life are strongly determined by the material and architecture of the O2 electrode60. 1.2.4.1 O2 electrode materials Typically, an O2 electrode consists of carbonaceous material, polymer binder, and electrocatalyst. Carbonaceous materials have been used as fillers to provide the porosity and electronic conductivity of electrode composites61. It has been well established that a good carbon candidate should have characteristics as follows: i) large surface area; ii) appropriate pore size/volume; and iii) high stability. A range of carbon materials have been systematically studied in Li-O2 systems, including commercial carbon powders62, onedimensional carbon nanomaterials63, two-dimensional graphene64, mesoporous carbon materials57, and carbon hybrids and composites65-67. However, the suitability of carbonaceous materials is currently under debate due to stability issues. It has been proved that carbon easily reacts with Li2O2 to form carbonate-like species (e.g., Li2CO3), which are difficult to oxidize during the charging process68-70. In addition, decomposition of the binder is also a serious concern. For example, the binder undergoes a dehydrofluorination reaction to produce LiF and/or LiOH, passivating the surface of electrode and leading to an increased charging potential as well as capacity losses71-73.

16

Eletrocatalysts for Li-O2 batteries As mentioned above, the pathways for the ORR and OER in Li-O2 batteries might be different, resulting in a large overpotential, which leads to a low coulombic efficiency17,74,75. Recently, it has been demonstrated that the employment of electrocatalysts could enhance the kinetics of the reactions (i.e., by lowering the overpotential, increasing the capacity, and improving the rate performance). The electrocatalysts in Li-O2 batteries can roughly be classified into the following categories: i) Noble metals/alloys, such as Pt76,77, Au78, Ru34,79, Pd31,80,81, PtxCoy alloy nanoparticles82, and Pt-Cu alloy83. Although these display superior electrocatalytic performance, their scarcity and high cost inevitably limit their practical application; ii) Metal oxides, mainly transition-metal oxides (e.g., Co3O422,29 and MnO26,84,85). The latter has been studied as the most favorable trade-off between cost and catalytic activity. Since a mono-component metal oxide cannot usually compete with a noble metal, efforts have been devoted to developing precious metal supported on metal oxides, which are expected to exhibit similar or even superior catalytic performances compared to pure precious metals66,86-88. In this respect, Pt/MNT and Ru/MNT were synthesized, which exhibited good catalytic activities during the OER (Papers II and III); iii) Carbon hybrids and composites are also a studied group since it is well-known that the porosity is one of the most important factors for gas diffusion in the cathode75. In order to provide the desired electronic conductivity and porosity, carbon is a critical material for the O2 electrode design. Strictly speaking, carbon itself is not an electrocatalyst. However, it is widely used as the electrocatalyst support. For example, noble metal particles loaded on carbonaceous materials89-91 and transition-metal oxide/carbon hybrids92,93 have demonstrated good catalytic performances in Li-O2 batteries; iv) Organic materials. So far, the commonly used electrocatalysts in Li-O2 batteries are inorganic materials. The electrocatalytic performance of inorganic materials are influenced by subtle changes in crystallite size, crystalline phase, and morphology, which not only affects the accurate evaluation of the electrocatalyst itself, but also results in poor stability and durability during battery operation. Being renewable, low-cost, and even self-repairing materials, organic electrocatalysts are predicted to overcome many of the shortcomings mentioned above. Recently, some organic-electrolyte-dissolved electrocatalysts have been reported which enhanced the electrochemical processes in Li-O2 batteries94,95. In comparison to these electrolyte-dissolved organic electrocatalysts, a solid organic electrocatalyst could be more useful for future practical applications, as it can be more versatile and adaptable in different electrolytes. However, a solid organic electrocatalyst for Li-O2 batteries hasonly been presented once96. As presented in Paper IV, Li2C8H2O6, which displayed a good catalytic performance in the Li-O2 battery, was prepared as a novel solid organic electrocatalyst via an original and low-polluting synthesis. So far, although substantial progress on development of electrocatalysts has 17

been achieved, the real role and catalytic mechanism of electrocatalysts in Li-O2 batteries is still under intense debate36,75. Thus, further investigation is worthwhile. 1.2.4.2 Architecture of O2 electrodes O2 electrode architecture also influences the performance of Li-O2 batteries. Owing to its practical application, the most common method of producing an O2 electrode is mixing the materials with a polymer binder and a solvent to make a slurry, which is then cast onto a foam or mesh based substrate58,97. Considering the instability of the binder materials, a novel “binder-free” design has been proposed as a promising strategy to avoid the negative influence of the binder on the long-term stability73,98-102. In this respect, we report a low-cost and versatile binder-free NCPE for use in the Li-O2 battery (Paper V). To sum up, in parallel with the importance of the electrolyte, investigations on the influence of the composition and architecture of the O2 electrode on the battery performance are still of great interest. Although the mechanism concerning reversible formation and degradation of Li2O2 is still controversial, this matter needs to be further studied by developing more stable O2 electrodes and more efficient electrocatalysts for Li-O2 batteries.

1.2.5 Lithium electrodes In all discussions of Li-O2 batteries, it is tacitly assumed that lithium metal is the negative electrode, since the theoretical specific energy is reduced to ~ 1000 W h·kg-1 using traditional Li-ion negative electrodes. Although lithium metal is considered as an ideal negative electrode, there are three main problems: i) Formation of dendrites during cycling, which could cause an internal short circuit and ultimately pose a serious safety hazard103; ii) Instability. It has been proved that the instability of lithium metal in contact with DMSO solvent is one of the reasons for the sudden death of the cell6,41. In addition, the conversion of metallic lithium to LiOH was observed by Liu et al.,104; iii) Formation of an unstable SEI. Lithium is thermodynamically unstable in contact with organic solvents and a SEI is instantaneously formed. Note that the stability of the SEI layer is the most important factor that determines the performance of lithium metal electrode as well as the coulombic efficiency of the cell. Although a silicon-based negative electrode has been attempted as an alternative105, obstacles with low coulombic efficiency and short lifetime have not been overcome. Indeed, serious drawbacks including O2 crossover, high charging potential, and instabilities of the electrolyte and O2 electrode in Li-O2 batteries need to be considered in the study of lithium metal electrodes.

18

1.3 The scope of this thesis This PhD project started in 2011 with the goal to study O2 electrode performance in the Li-O2 battery. Besides several issues such as the instability of electrolytes18,23,106, poor performance of lithium metal20,30 etc., one challenge regarding to the Li-O2 battery is to increase the electrode reaction kinetics, especially for the OER39. Another conundrum is that the decomposition mechanism of Li2O2 is not yet clear55. The research in this thesis has covered the investigation of factors influencing Li2O2 decomposition, the development of highly active electrocatalysts, and the design of binder-free O2 electrodes for Li-O2 batteries. Prior to 2013, the identification of chemical species was done using several techniques including XPS15, Raman107, FT-IR108, time-of-flight secondary ion mass spectrometry (ToF-SIMS)109, high-resolution TEM (HRTEM)69, and ex situ XRD110. While these ex situ studies provide valuable information on the electrochemistry of Li-O2 batteries, artifacts from post treatments were inevitably introduced into the measurements due to their post-mortem nature. This not only influences the accuracy of the result, but also leads to loss of important information. An important characterization method to follow the composition and structural changes in a cell during the cycling process is the analysis by means of in situ XRD (XRD measurement of a cell without unpacking it before and after the electrochemical process) or operando XRD (measuring XRD throughout the electrochemical process of a cell), which can give insight into the time-resolved kinetics throughout the whole electrochemical process in Li-O2 batteries111. In addition, the low scattering ability of oxygen and lithium and the limited X-ray intensity of normal in-house diffractometers lead to a low sensitivity for the detection of products. Synchrotron based XRD, SR-PXD, offers X-ray intensities several orders of magnitude higher, a better resolution and a wider tunability in wavelength permitting for different applications than the conventional research laboratory systems112. In this thesis, operando SR-PXD was carried out to study the Li2O2 decomposition during the OER in Li-O2 cells, which provides direct and real-time kinetic information during the whole electrochemical process. The research strategies employed here have primarily focused on three aspects: 1) Investigation of an accelerated effect of X-ray irradiation on the electrochemical degradation of Li2O2 in the Li-O2 cell (Paper I). In addition, a mechanism for the electrochemical decomposition of Li2O2 under X-ray irradiation was proposed; 2) Development of highly active electrocatalysts for Li-O2 batteries (Papers II-IV). X wt% Pt/MNT (x=0, 0.2, 1, 5 and 10) and x wt% Ru/MNT (x=0.5 and 2) were synthesized by a simple reduction and mechanical stirring method (Papers II and III). Li2C8H2O6 as a novel solid organic electrocatalyst was synthesized via an original and lowpolluting synthetic route (Paper IV). The catalytic activities of as-prepared 19

electrocatalysts were evaluated by following the Li2O2 decomposition during the charging process in Li-O2 cells using in situ XRD and operando SR-PXD measurements. In addition, the time-resolved OER kinetics for Li-O2 cells charged galvanostatically and potentiostatically were systematically investigated; 3) Investigation of a low-cost and easy-operation binder-free NCPE (Paper V).

20

2 Experimental

2.1 Materials synthesis 2.1.1 Preparation of MNT MNT has been successfully synthesized by a facile hydrothermal treatment of potassium permanganate (KMnO4) in a hydrochloric acid solution, as described by Luo et al.,113. KMnO4 powder (0.597 g, Merck) was first dissolved in deionized water (70 mL). Under vigorous stirring, hydrochloric acid (1 mL, HCl, 37%, Prolabo) was added drop-wise to the solution. After stirring for 30 min, the resultant precursor was then transferred to an autoclave, and maintained at 140 ◦C for 12 h. After the hydrothermal crystallization, the resulting black product, MNT, was filtered off and washed with deionized water several times, and then dried at 60 ◦C overnight.

2.1.2 Preparation of Pt/MNT and Ru/MNT The preparations of x wt% Pt/MNT (x=0.2, 1, 5 and 10) and x wt% Ru/MNT (x=0.5 and 2) were carried out by a simple reduction and mechanical stirring method. Sodium borohydride (0.4 M, NaBH4, Merck) was added drop-wise into a 0.1 wt% chloroplatinic acid hexahydrate (H2PtCl6·6H2O, Aldrich) or ruthenium (III) chloride hydrate (RuCl3·3H2O, Alfa Aesar) aqueous solution under the vigorous stirring at 60 ◦C. After stirring for additional 30 min, the mixture solution was cooled down to room temperature. As-prepared MNT (0.1 g) was added to the solution, which was kept under stirring for 24 h. The resultant product was filtered off and washed with deionized water several times, and then dried at 60 ◦C overnight. X wt% Pt/MNT (x=0.2, 1, 5 and 10) and x wt% Ru/MNT (x=0.5 and 2) were obtained by adjusting the dosages of NaBH4 and H2PtCl6·6H2O or RuCl3·3H2O. In order to compare to the catalytic activity of pure Pt or Ru nanoparticles, Pt and Ru nanoparticles were also synthesized under similar synthetic conditions (home-made Pt and Ru nanoparticles).

21

2.1.3 Preparation of Li2C8H2O6

Figure 2. Synthetic route to form Li2C8H2O6 3 obtained as a 1/1 mol/mol mixture with Li2O2114.

Li2C8H2O6 3 was prepared in a similar way to the previously reported dilithium (2,5-dilithium-oxy)-terephthalate (Li4C8H2O6) with the only difference that the final dismutation reaction was performed under dry and pure O2 in order to obtain the oxidized form of the electrocatalyst (Figure 2). Note that the dihydroxy-terephthalic acid 1 as starting material can be synthesized from D-glucose via eco-friendly reactions (water-based or solvent-less reactions with possible hydroquinone 4 recycling) as previously reported115, rendering the synthetic route sustainable and “green”. Deionized water/ ethanol solution (1:1 v/v, 20 mL) was added to a mixture of 2,5-dihydroxyterephthalic acid (396.3 mg, 2 mmol, Aldrich) and lithium carbonate (147.8 mg, 2 mmol, Merck). Under stirring at room temperature for 4 h, the solution was then dried in a ventilated oven (Heraeus LUT6050) at 90°C for 16 h, to produce a yellowish powder Li2C8H4O6 2 (420 mg, quant. yield). Li2C8H4O6 2 (362.1 mg, 1.72 mmol) was introduced into a vacuum oven (Buchi Glass Oven B-585) and was heated at 275°C for 20 h under dry and pure O2 and then under vacuum for additional 4 h to eliminate as-formed hydroquinone 4 and CO2. A dark green powder (190.4 mg) was obtained corresponding to a mixture of Li2C8H2O6 3 and Li2O2 in a 1/1 mol/mol ratio (yield = 44 % for a theoretical maximum of 50 %). Considering that Li2O2 is very difficult to separate from Li2C8H2O6 3, and that the electrocatalytic performance of Li2C8H2O6 3 was evaluated by following the degradation of Li2O2 in an electrode filled commercial Li2O2, which is not influenced by a small amount (~1.3 wt% of the total amount of electrode material) of Li2O2 co-existent with Li2C8H2O6 3, the mixture was used without any further purification. It was kept in an ultra-high purified argon-filled glove box (O2 and H2O < 1 ppm) before usage.

2.1.4 Preparation of NCPE Cladophora cellulose (200 mg, Baltic Sea) was dispersed in deionized water (40 mL) by sonication for 10 min with water cooling, using high energy ultrasonic equipment (Vibra-Cell 750) with a pulse length of 30 s and pulseoff duration of 30 s. The Cladophora nano-cellulose was extracted from the algae via grinding and acid hydrolysis as previously reported116. Pyrrole 22

(0.65 mL, Sigma-Aldrich) and HCl (0.5 M, 40 mL) were mixed with the described cellulose dispersion by stirring for 5 min. Polypyrrole (PPy) was formed on the Cladophora cellulose fibers by polymerization with iron (III) chloride nonahydrate (FeCl3·6H2O, 5.9 g, Sigma-Aldrich) dissolved in HCl (0.5 M, 40 mL) as an oxidant. The polymerization proceeded for 0.5 h under stirring, after which the composite was collected in a Büchner funnel connected to a suction flask and washed with HCl. In our previous work, the PPy/cellulose composite precursor has been characterized117-119. Carbon filaments (CFs, Goodfellow) were chopped with a kitchen herb-cutter knife on a wooden block so as to obtain a fluffy mass consisting of individual chopped carbon filaments (CCFs) with a maximum length of about 5 mm. The PPy/cellulose composite and CCFs (200 mg) were suspended in deionized water (200 mL) using a mechanical homogenizer (IKA T25 UltraTurrax) at 6200 rpm for 10 min. The mixture was drained and then dried to form a PPy/cellulose-CCFs paper sheet, which was annealed in a nitrogen atmosphere at 1000 °C for 1 h at a heating rate of 3°C/min to form NCPE. A schematic illustration of the NCPE synthetic procedure is shown in Scheme 1. The as-prepared NCPE was punched into a disc with a diameter of 12 mm for usage as an O2 electrode. The weight of each piece was 4.5-6.5 mg. In addition, in order to investigate the influence of doped N on the performance of the Li-O2 cell, a reference precursor of cellulose-CCFs paper sheet was also prepared without PPy under similar conditions. Thus an undoped carbon paper electrode was obtained after annealing.

Scheme 1. Schematic illustration of the synthetic procedure of NCPE.

23

2.2 Preparation of electrodes 2.2.1 Preparation of a commercial Li2O2-filled electrode used for the charge process Although the real role of the electrocatalyst is controversial and the catalytic mechanism is not yet clear18,120,121, it has been widely accepted that the ORR kinetics in a Li-O2 battery is catalytically insensitive, while the counterpart OER exhibits a different behaviour78,122. For example, the employment of an electrocatalyst can lower the overpotential, resulting in improvement of battery cyclability. Until now, most evaluations of electrocatalysts have depended on galvanostatic discharge-charge measurements, with no quantitative evidence of what kind of electrochemical reaction that occurs. Therefore, it is impossible to judge whether it is the genuine electrocatalytic effect related to the Li2O2 decomposition that takes place during the charging process. Model electrode assembled in its discharged state (i.e., a commercial Li2O2-filled electrode) is a useful tool for investigating the basic principles of electrocatalytic effects in Li-O2 batteries78,123,124. The preparation of commercial Li2O2-filled positive electrodes used for the charging process was performed inside the glove box. The electrodes were composed of a mixture of Super P carbon (lithium battery grade, Erachem Comilog), Kynar 2801 (a copolymer based on PVdF, Arkema), electrocatalyst (as-prepared x wt% Pt/MNT, x wt% Ru/MNT, home-made Pt or Ru nanoparticles, or Li2C8H2O6), Li2O2 (technical grade, 90 %, Sigma-Aldrich or 95 %, Acros Organics), and silicon powder (Si, 99.5 %, Alfa Aesar) in a weight ratio of 40:7:8:35:10. The effect of the amount of catalyst on the Li2O2 decomposition was also studied by adjusting the weight ratio of the components (44:7:4:35:10 or 46:7:2:35:10), as presented in Paper III. Since Si can provide stable and strong diffraction peaks in the XRD pattern, it was used as an inert external reference to quantify the amount of Li2O2 present in the electrode. Super P carbon, electrocatalyst, Si, and Li2O2 were mixed by the high energy ball-milling for 1-2 h (in order to prevent the Li2O2 oxidation in the air, the grinding bottle was sealed with a soft package (i.e., coffee bag) in the glovebox). Kynar and acetone solvent (≥ 99.0 %, Fluka) were then added into the mixture to prepare a slurry, which was handmilled for 30 min. The slurries were cast onto an aluminum mesh substrate with a diameter of 13 mm. After the acetone was evaporated, the obtained electrodes were transferred to a vacuum oven and further dried at 120 °C overnight.

24

2.2.2 Preparation of the O2 electrode used for the discharge and charge processes The preparation of the electrode for cell performance testing (discharge and charge process) was similar to that of electrodes used for the charge process described above, except that no commercial Li2O2 was added to the electrode. The electrode consisted of Super P carbon, electrocatalyst, and Kynar 2801 in a weight ratio of 70:20:10.

2.3 Assembly and electrochemical testing of the Li-O2 cell A pouch cell (i.e., a “coffee bag” cell) design was used in the study of the charging process (Figures 3a and b). A SwagelokTM cell design modified with an opening to allow pure and dry O2 to access through the electrode was used for the study of the cell cycling, which was set up in the O2-filled bottle (Figures 3c-e). The O2 pressure inside the bottle during cell operation was 1 atm, and the volume of O2 was ~300 mL, which was 100 times more than the consumption during the discharge process. All the Li-O2 cells were assembled in the glove box with the following components: a lithium metal as negative electrode, a double-layer Solupor separator soaked with the electrolyte (1 M LiPF6 (≥ 99%, Ferro)/PC (≥ 99%, Purolyte) or 1 M LiClO4 (≥ 99%, GFS)/DMSO (≥ 99.9%, Aldrich)), and a porous O2 electrode as the positive electrode as described above. Details regarding the applied currents, voltages, and capacities, which were different for the different studies, can be found in the appended papers.

Figure 3. Images of (a and b) cells sealed in "coffee bags", and (c-e) SwagelokTM designs modified with an airtight container with inlet and outlet valves.

25

2.4 Characterization The structure analysis for the as-prepared materials was made by XRD, Raman Spectroscopy, FT-IR and NMR. Morphology examinations were carried out using SEM, TEM and HRTEM. The surface area and porosity were determined by BET measurements. XPS and EDX measurements were performed to study the chemical state and elemental composition. TG measurements were utilized for the thermal analysis. Details regarding the measurements using the different techniques can be found in the appended papers. Along with these techniques, in situ XRD and operando SR-PXD were mainly used for probing the Li2O2 decomposition during the charging process.

2.4.1 XRD A number of powerful experimental techniques have been employed to study the material structure, and most of them involve diffraction. Diffraction occurs when the light encounters an obstacle or a slit. The light wave can either bend around the obstacle, or travel through the slit. If the light energy is in the X-ray range of 1-100 keV (wavelength range of 10-3 to 1 nm) the interaction of the incident ray with the electron cloud surrounding atoms in a crystalline solid will produce constructive interference pattern and a diffraction ray without energy transfer (elastic scattering) and time delay (coherence) will take place. This interaction is termed XRD or Bragg scattering125, which was first described by Von Laue and Bragg. To date, as a scientific method, XRD analysis is primarily used for determining the atomic structure of a solid material (identify chemical bonding and the distance between the atoms), the identification of crystalline materials (i.e., “fingerprints”), to distinguish between crystalline and amorphous materials, and quantification of the degree of crystallinity or stress in a compound. Here, a brief introduction to the XRD technique and its utilization in Li-O2 batteries is provided. X-ray generation X-rays are generated either by a cathode ray tube or as synchrotron radiation from accelerating electrons or positrons in a particle accelerator. In the X-ray tube, which is the primary X-ray source utilized in laboratory diffractometers, X-rays are produced when a focused electron beam accelerating across a high voltage field bombards the solid target. Common targets used in Xray tubes involve cobalt (Co), copper (Cu), and molybdenum (Mo), which emit 7, 8, and 14 keV X-rays with the corresponding wavelengths of 1.79, 1.54, and 0.8 Å, respectively. The energy of an X-ray photon (E) and the wavelength (λ) are related via equation 1 (Eq. 1) as follows: E = hc/λ

26

(1)

where h is Planck’s constant and c is the speed of light. Recently, synchrotron facilities have become widely used as a source for XRD measurements. Synchrotron radiation can be emitted by the electrons or positrons travelling at a nearly light speed in a circular storage ring. These powerful sources, which offer thousands to millions of times more intense X-rays than laboratory X-ray tubes, have been well-established and in use for a wide range of structural investigations126,127. Bragg's Law In XRD, an outgoing wave (diffracted wave) is generated by the interaction between the oscillating electric field of incident X-ray photons and the electrons. These diffracted waves interfere with each other and the resultant intensity distributions are strongly modulated by the interaction. Due to the periodic structure of a crystalline material, it can be described as a series of planes with an equal inter-planar distance (d). As an incident X-ray beam interacts with the lattice planes at an angel (θ), as shown in Figure 4, some of them will be directly diffracted away at the same angle while the remainder will travel deeper and interact with the second plane and will then be diffracted at an angle of θ. The processes are repeated for all the lattice planes in the crystals. Since X-ray beams travel different path lengths before the diffraction, the constructive wave interference occurs only when the path length difference is equal to an integer number of the wavelength (λ). As shown in Figure 4, the difference in the path lengths between the beam striking at the top and the bottom plane is equal to AB+AD. Therefore, these two diffracted beams could constructively interfere only when AB+AD = nλ. According to basic trigonometry AB = AD = d sinθ. Thus, the relationship between θ, d, and λ can simply be described using the following Eq. 2, which is the well-known Bragg's Law128. It is one of the most important laws used when interpreting XRD data.

Figure 4. Geometry for interference of a wave scattered from two planes separated by an inter-planar distance (d).

2d sinθ = nλ

(2)

27

Powder diffraction Powder XRD measurement is the only XRD technique used in this thesis, and it is briefly introduced here. As the name suggests, the sample is always in a powdery form. In fact, the term “powder” means that the crystalline domains have random orientations rather than only one orientation (i.e., single crystal), so the recorded 2D diffraction pattern displays concentric rings of scattering peaks corresponding to the various d values for the crystal lattices. Figure 5 shows diffraction from a single crystal and a powder, respectively.

Figure 5. Diffraction from a single crystal (to the left) and a powder (to the right).

The diffraction peaks can be collected using either the reflection or transmission mode, as shown in Figure 6. The positions and intensities of the recorded diffraction peaks are used for the identification of the underlying phase and structure of the sample. Additionally, crystallite size can be estimated by the Scherrer equation which relates the full width at half maximum (FWHM) of a peak to the crystallite size128.

Figure 6. The reflection (to the left) and transmission (to the right) mode for the collection of the diffracted beam.

In this thesis, XRD analyses of as-prepared samples were carried out by a Bruker D8 TwinTwin X-ray diffractometer, using Cu Kα radiation with λ = 1.5406 Å, operating at 40 kV and 40 mA. In situ XRD measurement in transmission mode was used to study the electrochemical degradation of Li2O2 in a Li2O2-filled electrode before and after the charging process. This was performed on an in-house STOE diffractometer with a position28

sensitive detector and Cu Kα radiation, operating at 45 kV and 40 mA (Figure 7a). To further study the real-time OER kinetics, operando SR-PXD was conducted using beamline I711 with λ = 0.996 Å at the MAX II synchrotron at the Max IV Laboratory in Lund, Sweden (Figures 7b and c)129. All the cells were mounted in the transmission mode, and the frames were collected with an Oxford diffraction Titan CCD area detector. Lanthanum hexaboride (LaB6, SRM-660) was employed as a reference material to calibrate the parameters of the instrument (e.g., sample-to-detector distance and tilt angle). In paper I, in order to investigate the effect of X-ray illumination on the electrochemical decomposition of Li2O2, the frames were collected under continuous X-ray irradiation during a standard experiment. In papers II-IV, the frames were collected every 15 or 30 min to avoid the influence of X-ray irradiation on the degradation of Li2O2. Note that a frame is integrated data, which is then converted to an intensity vs.2θ plot through circular integration performed by the program Fit2D130. The program FullProf was used to refine models to the in situ XRD and SR-PXD patterns and to analyze the weight fraction of reference Si and Li2O2131. The residual ratio of the Li2O2 (R) in the electrode before, after and during the charging process was calculated from Eq. 3 as follows:

× 100% =

R (%) = 100% =

/ /

× 100%

/ /

(

× 100% =

(

)/( )/(

) )

×

(3)

where W is the weight of Si in the electrode, which should not change during the charging process. W and W′ are the Li2O2 weights before and F are the Li2O2 and after the charging process, respectively. F and Si weight fractions before the charging process, respectively, while F′ and F′ are the corresponding weight fractions after the charging process, respectively.

29

Figure 7. Images of (a) in-house transmission XRD and (b) SR-PXD, and (c) schematic diagram of the beamline I711.

30

3 Summary of the results and discussions

In this chapter, the summary of the results and discussions is divided into three sections. The first section deals with the influence of X-ray illumination on the electrochemical degradation of Li2O2 during the charging process in a cell. In addition, fundamental information such as the limiting factors and involved steps of the OER were investigated. In the second section, three highly active electrocatalysts for Li-O2 batteries are presented. The time-resolved OER kinetics for the electrocatalyst-containing electrodes under galvanostatic and potentiostatic charging conditions was systematically studied by operando SR-PXD. In the last section, a binder-free NCPE is presented. The rate performance, coulombic efficiency, and cyclability of the Li-O2 batteries with NCPEs were investigated.

3.1 Study of the electrochemical decomposition of Li2O2 under continuous X-ray illumination The photoelectric effect was first discovered in 1839 by Becquerel. In the 1950’s, the possibility of the conversion of solar energy into electric power in a photo-electrochemical system was presented by researchers at Bell labs132. In 1972, Fujishima and Honda reported that UV light illumination can enhance the electrochemical decomposition of H2O using a TiO2 photocatalyst133. Their work triggered a rapid development of the photoelectrochemical systems134,135. In those systems, the composition of photoactive material normally does not change after the interactions, since it only plays a catalytic role in the reactions. Equally, in principle, the target materials could be electrochemically generated or decomposed under electromagnetic illumination. So far, UV and visible light have been used in this field136. However, being a photo source, the application of X-rays has been rarely reported in conjunction with electrochemical transformations of solid materials. In Paper I, the influence of X-ray irradiation on the electrochemical degradation of Li2O2 in a cell with a porous commercial Li2O2-filled electrode and 1M LiPF6/PC electrolyte was studied112. It was already known that the PC-based electrolyte is not stable during the discharging process due to the formation of intermediate superoxide species15,18,19,106,137. However, since we 31

only studied the OER and the oxidation potential of PC is higher than that of other solvent18,138, the 1M LiPF6/PC electrolyte was considered suitable in this work. SR-PXD was employed both as a continuous photo source to advance the electrochemical degradation of Li2O2 and as an operando technique to study the real-time OER kinetics. The effect of X-ray illumination on the Li2O2 decomposition in the OER of a Li-O2 cell Si

Li2O2

Charging

Li2O2

a 18

19

20

21

22

23

2 Theta (degree)

Residual ratio of Li2O2 (%)

100 90 80 70 60

b

50 0

20

40

60

80

100

Residual ratio of Li2O2 (%)

Time (min)

100

80

60

c 0

2

4

6

8

Time (h)

Figure 8. (a) The SR-PXD patterns of a Li2O2-filled electrode collected every 10 min during the charging process, (b) the Li2O2 decomposition curve in the electrode and (c) without any applied charged at a constant current density of 40 mA·g charging current under continuous X-ray irradiation.

Figure 8a shows a stacked plot of intensity vs. 2θ patterns collected for a Li2O2-filled electrode under continuous X-ray irradiation during charging at a constant current density of 40 mA·g . Si (JCPDS Card No: 04-0140211) and Li2O2 (JCPDS Card No: 04-013-3506) crystal phases were observed. Note that the absolute intensities of the Si peaks decreased slightly with time, which could be attributed to the gradual fading intensity of the synchrotron beam during the measurement. The decrease in the intensity of the Li2O2 peaks is much more rapid, however, indicating that Li2O2 degrada32

tion occurred during the charging process. The residual Li2O2 ratio in the electrode was 49 % after charging for 110 min (Figure 8b). It was reported that soft X-ray beam damage could lead to the Li2O2 decomposition to form Li2O139. However, as shown in Figure 8c, there was very little change in the amount of Li2O2 under continuous X-ray irradiation for 8 h without any applied charging current, indicating that the Li2O2 decomposition observed here cannot simply be attributed to the X-ray beam damage. Additionally, no Li2O peaks (JCPDS Card No: 00-012-0254) were observed in any of the collected patterns.

3

Si

Li2O2

Li2O2

2 1

18

19

20

21

22

23

2 Theta (degree)

Figure 9. The SR-PXD patterns of a Li2O2-filled electrode (1) before charging, after on a single spot (2) charging for 3 h at a constant current density of 40 mA·g with X-ray irradiation, and (3) without X-ray irradiation.

In order to study the influence of X-ray irradiation on the Li2O2 electrochemical decomposition, a Li2O2-filled electrode was charged for 3 h under continuous X-ray irradiation on a single spot with a diameter of 0.8 mm. Figure 9 shows the SR-PXD patterns of a Li2O2-filled electrode before charging, after charging for 3 h with X-ray irradiation, and in the absence of X-ray irradiation (by moving the measurement spot about 2 mm away from the previously used irradiation spot). Under constant irradiation, an obvious decrease in the intensity of the Li2O2 peaks can be seen compared to the peaks for the as-prepared electrode (Figure 9). The residual Li2O2 ratio was 47 % after charging for 3 h with X-ray irradiation. In contrast, the pattern corresponding to the case without X-ray irradiation showed only a slight reduction in the intensity of Li2O2 peaks as compared to that for the asprepared one. This corresponded to a residual Li2O2 ratio of 89 %, which agreed well with the theoretical value (89.8 %) calculated based on the charging capacity (i.e., pure electrochemical decomposition), as presented in Paper I. Therefore, the results confirm that the X-ray irradiation accelerates the electrochemical degradation of Li2O2 in the OER.

33

Investigation of the relationship between the X-ray intensity and the Li2O2 decomposition rate

4.4

Linear regression for curve fitting: Y = A + kX

90 80

I II R2 0.9919 0.9809 k -0.6704 -0.4586

70

II

60

a

50 0

4.0

Potential (V)

Residual ratio of Li2O2 (%)

100

I II

3.6

3.2

b

I 20

40

60

Time (min)

80

100

2.8

0

20

40

60

80

100

Time (min)

Figure 10. (a) The Li2O2 decomposition curves (−) and linear curve fittings (−·−) for under X-ray the electrodes charged at a constant current density of 40 mA·g irradiation with different intensities (I) Ia, and (II) Ib (Ib= 68 % of Ia), and (b) the charging profiles of both cells.

To further investigate this phenomenon, the influence of the X-ray intensity on the Li2O2 electrochemical decomposition was examined by following the Li2O2 decomposition during the charging process using X-ray irradiation with different intensities. The X-ray intensity was adjusted by inserting aluminum foils as filters in the path of the incident beam. As shown in Figure 10a, after charging for 110 min, the residual Li2O2 ratio decreased to 49 % with Ia and 66 % with Ib (Ib= 68 % of Ia), respectively, indicating that Li2O2 decomposition rate was influenced by the X-ray intensity. The oscillating charging profiles shown in Figure 10b may result from the fluctuating temperature in the SR-PXD room. Note that no significant difference between the charging profiles was observed. The explanation for this could be that the contribution of the X-ray on the whole electrode was less than 0.5 % as the diameter of the X-ray spot is only 0.8 mm. In addition, according to the sluggish decomposition of Li2O2 during the beginning of the charging process, it could be concluded that the X-ray irradiation cannot enhance the Li2O2 decomposition rate during the first 30 min of charging. However, the X-ray intensity largely affected the Li2O2 decomposition during the following period. It can be seen in Figure 10a that the Li2O2 decomposition curves were then approximately linear with respect to the charging time (30 min ≤ t ≤ 100 min). During this period, pseudo-zero-order kinetics was used to evaluate the corresponding decomposition rate constant (k) from Eq. 4 as following: ′

34

=

- kt

(4)

and ′ are the residual Li2O2 ratios for the electrode before where and during the charging time t (min), respectively, and k (min-1) is the pseudo-zero-order decomposition rate constant. The decomposition rate constant under the irradiation with Ib (kb = 0.46 min-1) was 68.6 % of that under the irradiation with Ia (ka= 0.67 min-1). This ratio is almost equal to the ratio (68 %) of their respective X-ray intensities (ka/kb≈ Ia/Ib), which indicates that a linear relationship exists between the decomposition rate constant and the applied X-ray intensity.

Residual ratio of Li2O2 (%)

Investigation of the effect of external potential on the electrochemical decomposition of Li2O2 under X-ray illumination a

100

b

c

80

d Linear regression for curve fitting: Y = A + kX c d b a R2 0.7943 0.8201 0.8704 0.9981 k -0.0040 -0.0091 -0.0153 -0.4127

60

40

0

120

240

360

480

Time (min)

Figure 11. The Li2O2 decomposition curves (−) and linear curve fittings (−·−) for an electrode charged at stepwise potentials of (a) 3.0, (b) 3.4, (c) 3.8, and (d) 4.2 V, respectively.

In an attempt to distinguish our findings from photolysis (i.e., beam damage), the influence of an applied external potential on the Li2O2 decomposition under a constant intensity of X-ray illumination was examined by charging a cell stepwise, at potentials between 3.0 and 4.2 V for 120 min at each potential, as shown in Figure 11. Pseudo-zero-order kinetics of Li2O2 decomposition does also apply here. Only 0.3, 1.2 and 2.6 % of the Li2O2 decomposed during charging for 120 min at 3.0, 3.4 and 3.8 V, respectively. However, more than 50 % of the Li2O2 decomposed when the cell was held at a charging potential of 4.2 V, which is much faster than for a cell charged at the same potential without X-ray illumination (see Paper I). The significant difference between the Li2O2 decomposition rates at different applied potentials demonstrates that the accelerated effect of X-rays on the Li2O2 degradation occurs only when a certain external potential is provided. This indicates that some involved steps might require a certain energy barrier. Based on these results, it can be well established that both X-ray irradiation

35

and the external potential play important roles in the electrochemical decomposition of Li2O2. A proposed mechanism for the electrochemical decomposition of Li2O2 under X-ray illumination

Figure 12. Scheme illustrating the electrochemical decomposition of Li2O2 under Xray illumination.

Based on the observations described above, a mechanism for the electrochemical decomposition of Li2O2 under continuous X-ray illumination is proposed based on the following R6-R8: Li2O2

[Li2O2]x+ + xe-

(R6)

[Li2O2]x+

Li2-xO2*+ xLi+

Li2-xO2*

+

(R7) -

(2-x) Li + O2↑ + (2-x) e

(R8)

In the first step, the X-rays eject photo electrons from the Li2O2 crystals to form a partly positive species ([Li2O2]x+). A higher X-ray intensity should produce more [Li2O2]x+, leading to a higher decomposition rate. Therefore, this step could be considered as a kinetically limiting step. The external potential plays an important role for driving the free electrons to the counter electrode. In the absence of an external potential, these free electrons can 36

recombine quickly with [Li2O2]x+. It has been demonstrated that X-ray irradiation can eject the electrons from the substance, but it cannot complete the whole electrochemical process140. In the second step, in order to keep electric neutrality, [Li2O2]x+ will lose Li+ and transform into Li2O2 with some Li vacancies (Li2-xO2*). Ong et al., reported that the formation energies of the Li vacancy in an intralayer and interlayer of Li2O2 are 3.8 and 4.1 eV, respectively141, which means that a certain driving force is needed for the formation of Li2-xO2*. This agrees well with our results. The mobile Li vacancy was proved to enhance the electronic conductivity of Li2-xO2*, which could promote the Li2-xO2* decomposition142,143. In the last step, the Li2-xO2* crystal structure completely collapses to form Li+ and O2. To sum up, X-ray promotes the kinetics in the first step, while the applied external potential provides the driving force to meet thermodynamics requirements of the second and/or last steps. A schematic illustrating the electrochemical decomposition of Li2O2 under X-ray illumination is shown in Figure 12. Brief summary of the results of the electrochemical decomposition of Li2O2 under continuous X-ray illumination  Being a photo energy, X-ray accelerated the electrochemical decomposition of Li2O2;  The Li2O2 decomposition rate was proportional to the X-ray intensity applied;  The electrochemical decomposition of Li2O2 in the Li2O2-filled electrode of a cell charged at a constant current under X-ray irradiation followed pseudo-zero-order kinetics;  A three-step reaction involving the [Li2O2]x+ and Li2-xO2* as intermediates was proposed in the decomposition process of Li2O2 under X-ray illumination;  X-ray irradiation could promote the kinetics, while the external potential could overcome the thermodynamic barriers.

3.2 Study of the highly active electrocatalysts for Li-O2 batteries One of the critical challenges for Li-O2 batteries is the low energy efficiency resulting from a large overpotential, due to the high charging potential17,74. So far, although the electrocatalytic mechanism is not yet clear and more fundamental research is still urgently needed, it has been generally believed that the employment of electrocatalyst could decrease the activation barrier in the electrochemical reactions, especially for the OER, leading to a lower overpotential144. Therefore, it is critical to develop highly active electrocatalysts, especially for the OER, in order to build a Li-O2 battery with satisfac37

tory energy efficiency and cyclability. In this thesis, three electrocatalysts were synthesized by facile routes and their catalytic performances were studied by following the Li2O2 decomposition for Li2O2-filled electrodes using in situ XRD and operando SR-PXD techniques. In addition, the time-resolved OER kinetics for the electrocatayst-containing electrodes was systematically studied under galvanostatic and potentiostatic charging conditions.

3.2.1 Study of x wt% Pt/MNT (x=0, 1, 5, and 10) electrocatalysts Pt (a highly stable and electrocatalytic precious metal) can facilitate oxygen evolution and lower the activation barrier in the OER76,77,145. However, the scarcity and cost of Pt inevitably limit its application in Li-O2 batteries. Supported noble metal catalysts have been used as potential candidates to replace noble metal catalysts. The catalytic performance of a material supported by Pt particles could even exceed that of pure Pt90,146, since the large size and/or aggregation of single Pt particles normally reduce its surface area, which can depress the electrocatalytic performance. In this respect, Pt deposited on Co3O4 (Pt/Co3O4)86, graphene (Pt/graphene)87, titanium nitride nanotubes (Pt/TiN)88, and carbon nitride (Pt/CN)66 have been explored as the electrocatalysts in Li-O2 batteries. Several parameters can affect the catalytic activity of the Pt including its synthesis method and the choice of support. Using a nanoscale support is a promising approach to stabilize a welldispersed nanostructured catalyst, in order to generate synergistic effects. It has also been reported that the implementation of a nanosize material can not only give a better electronic conductivity but also offer a robust support for the cathode catalyst147. MnO2 is a commonly used electrocatalyst in Li-O2 batteries, as it is considered to be the most favorable trade-off between cost and catalytic activity. The electrocatalytic activity depends on its surface area, crystal phase, and morphology75. Thackeray et al., demonstrated that MnO2 with α phase (αMnO2) displays a higher activity than other phases (i.e., β or γ phase) due to its intrinsic tunnel size and high electrical conductivity148. It has been reported that the α-MnO2 nanowire provides an optimized structure which exhibits a good catalytic performance85. MNT has been demonstrated to display much better catalytic activity than α-MnO2 nanowires due to the unique tunnel structure and higher surface area149. In addition, MnO2 as a catalyst support can be used in the Li-O2 battery81,150. In paper II, x wt% Pt/MNT (x=0, 0.2, 1, 5, and 10) were reported. The electrocatalytic performance of as-prepared samples was evaluated by following the Li2O2 degradation during the OER using operando SR-PXD. In addition, the effects of the deposited amount of Pt on the OER kinetic were investigated systematically. 38

Table 1. Characteristics of as-prepared x wt% Pt/MNT (x=0, 1, 5, and 10) x wt% Pt/MNT sample

calculated x

x obtained by

x obtained by

EDS

XPS

crystallite size of

particle size of Pt

Pt from XRD

from TEM (nm)

SBET (m2/g)

(nm)

MNT

0

0

0

-

-

17

1 wt% Pt/MNT

1

0.8

3.7

-

7

17

5 wt% Pt/MNT

5

4.5

19.7

11

7

19

10 wt% Pt/MNT

10

10.3

31.1

12

7

21

(311)

(220)

(521)

(200)

(411)

(121) (111) (301)

(310)

(200)

(110)

α-MnO2 Pt

Intensity (a.u.)

d c b

a 10

20

30

40

50

60

70

80

90

2 Theta (degree)

Figure 13. XRD patterns of (a) MNT, (b) 1 wt% Pt/MNT, (c) 5 wt% Pt/MNT, and (d) 10 wt% Pt/MNT.

The phase identification of as-prepared x wt% Pt/MNT (x=0, 1, 5, and 10) was performed by XRD analysis. As shown in Figure 13, the main peaks of the patterns of all samples could be perfectly indexed to α-MnO2 (JCPDS Card No. 04-007-2142). No characteristic peak corresponding to Pt in the 1 wt% Pt/ MNT sample was observed, which can be attributed to the low content of the deposited Pt. However, for 5 and 10 wt% Pt/MNT samples, the peaks corresponding to the (111), (200), (220), and (311) planes of Pt (JCPDS Card No. 01-087-0640) can be observed. The crystallite size for the deposited Pt of 5 and 10 wt% Pt/MNT was estimated by Scherrer’s formula according to the full width at half maximum of the most intense Pt peak of (111). As seen in Table 1, the crystallite size of Pt in 5 and 10 wt% Pt/MNT was 11 and 12 nm, respectively.

39

Figure 14. SEM micrographs of (a) MNT, (b) 1 wt% Pt/MNT, (c) 5 wt% Pt/MNT, and (d) 10 wt% Pt/MNT.

The morphologies of as-prepared x wt% Pt/MNT (x=0, 1, 5, and 10) can be seen in Figure 14. The MNTs displayed uniform “tube-like” shapes (Figure 14a). Note that all Pt deposited samples showed a similar morphology and MNT particle size (Figures 14b-d). For 1 and 5 wt% Pt/MNT samples, Pt nanoparticles were highly dispersed and attached on to the surfaces of the MNTs, as shown in Figures 14b and c. However, the Pt nanoparticles of 10 wt% Pt/MNT exhibited agglomeration (Figure 14d). EDS analysis was conducted to determine the chemical composition of the as-obtained samples, and the results are shown in Table 1. The contents of deposited Pt in all the final samples obtained from the EDS measurement agreed well with the values calculated based on their syntheses, indicating that the synthetic route of x wt% Pt/MNT was reliable and controllable.

40

Figure 15. TEM micrographs of (a) MNT, (b) 1 wt% Pt/MNT, (c) 5 wt% Pt/MNT, and (d) 10 wt% Pt/MNT., and (e) a statistical evaluation of the Pt particle size distribution estimated from at least 50 particles in the TEM micrographs.

TEM measurements were employed to study the detailed particle and crystal information of the as-prepared samples. As depicted in Figure 15a, the MNT showed a width of 80-100 nm, a length of 1-1.5 μm, and a tube inner diameter of 40-60 nm. Pt nanoparticles were distributed uniformly on the surfaces of 1 and 5 wt% Pt/MNT samples (Figures 15b and c). However, the Pt nanoparticles in 5 wt% Pt/MNT displayed serious aggregation (Figure 15d), which was also observed in the SEM micrographs. The nanoparticle size of the Pt in 1, 5, and 10 wt% Pt/MNT was about 7 nm, which was in good accordance with the results calculated based on the XRD data. The BET surface areas of x wt% Pt/MNT samples (x=0, 1, 5, and 10) are presented in Table 1.

41

Table 2. Binding energy from the XPS spectra of the MNT and 5 wt% Pt/MNT samples and relative contribution of each peak to deconvoluted O1s spectra. binding energy (eV) sample

Mn2p3/2

Pt4f7/2

OL 1s

OH 1s

(M-O)

(O-H)

O-C-O

OL (%)

OH (%)

OH/ OL

MNT

642.5

-

530

531.5

533.2

78.7

16.7

21.2

5 wt% Pt/MNT

642.3

71.7, 72.8,

529.9

531.4

533.2

74.6

19.7

26.4

and 74.7

Figure 16. (a) Mn2p, (b) Pt4f, and (c) O1s XPS core level spectra for the MNT and 5 wt% Pt/MNT samples.

XPS measurements were carried out to investigate the chemical states and surface composition of the as-obtained samples. Figure 16 shows Mn2p, Pt4f, and O1s spectra for the MNT and 5 wt% Pt/MNT samples. The Mn2p spectra for the MNT and 5 wt% Pt/MNT contain two peaks at 642.6 and 654.2 eV with a spin-orbit separation of 11.6 eV (Figure 16a), which were in good agreement with the binding energies of Mn2p3/2 and Mn2p1/2, respectively, indicating that the Mn mainly existed as Mn4+ 151. Note that the Pt4f spectrum of 5 wt% Pt/MNT sample reveals the presence of Pt in three different charge states, as shown in Figure 16b. The first peak which mainly contributed to the spectrum at 71.7 eV, indicates the presence of Pt as Pt0 152. The second and third peaks at 72.8 and 74.7 eV represent Pt2+ and Pt4+, respectively152. The relative amounts of the deposited Pt nanoparticles in the x wt% Pt/MNT samples obtained from the XPS measurements are given in Table 1. The XPS results provide the composition information of the surface of the as-prepared samples, while the EDS analyses reflect the average compositions. As seen in Table 1, the deposited amounts obtained from the XPS measurements were much higher than those calculated from the synthesis process and obtained from EDS, indicating that the Pt nanoparticles were highly dispersed on the surfaces of the MNTs. This could be favorable to the 42

catalytic reactions as catalysis is a surface process. The O1s spectra of these two samples in Figure 16c are asymmetric, indicating that there should be at least two chemical forms of oxygen present. The deconvolution of 5 wt% Pt/MNT shows the main contribution at a binding energy of 529.9 eV, which corresponds to oxygen in the MnO2 lattice (OL). The second peak at 531.4 eV corresponds to oxygen from chemisorbed H2O (OH) on the surface of the MnO2 while the third peak at 533.2 eV represents oxygen from a small contamination containing ether bonds (O-C-O) (Figure 16c and Table 2). The binding energies from the XPS spectra and the relative contribution of the peaks to the O1s spectra are listed in Table 2. As seen in Table 2, the amount of surface hydroxyl oxygen in the 5 wt% Pt/MNT sample was higher than that in the MNT, which may result from the increment of surface oxygen vacancies by Pt decoration. The oxygen vacancies are active groups, which easily could combine with H2O adsorbed on the sample surface to form surface hydroxyl oxygen153,154. Rossmeisl et al., proved by DFT calculations that the presence of surface hydroxyl oxygen and oxygen vacancies could facilitate the OER155. Therefore, as compared to the MNT, 5 wt% Pt/MNT with proper surface state modifications is expected to exhibit better catalytic activity.

43

a

MNT

Si

Li2O2

38

32

Residual ratio of Li2O2 (%)

80 60 40

b

34

36

38

32

40

theoretical value MNT commercial Pt 1 wt% Pt/MNT 5 wt% Pt/MNT 10 wt% Pt/MNT home-made Pt

34

36

38

40

2 Theta (degree)

2 Theta (degree)

2 Theta (degree)

100

before charging

before charging 40

c

4.5

4.0 Potential (V)

36

Li2O2 Li2O2 after 15 h charging

after 15 h charging

before charging 34

Si

Li2O2

Li2O2

after 15 h charging

32

5 wt% Pt/MNT

commercial Pt Si

Li2O2

3.5

MNT

commercial Pt 5 wt% Pt/MNT

home-made Pt

3.0

20 0

0

200

400

Capacity (mAh/g

600 )

Li2O2

Figure 17. (a) The in situ XRD patterns of Li2O2-filled electrodes with three representative catalysts (MNT, commercial Pt, and 5 wt% Pt/MNT) before and after , (b) the residual Li2O2 ratios charging for 15 h at a constant current of 40 mA·g for the electrodes with different catalysts after charging for 15 h, and (c) charging profiles of the cells with four representative catalysts (MNT, commercial Pt, homemade Pt, and 5 wt% Pt/MNT) during the charging for 15 h.

In order to investigate the effect of electrocatalytic activity of Pt deposited on MNTs, we also synthesized pure Pt particles (home-made Pt nanoparticles). The catalytic performance of as-prepared samples were evaluated by following the Li2O2 decomposition in the cells charged for 15 h at a constant . Figure 17a shows the in situ XRD patterns current density of 40 mA·g of Li2O2-filled electrodes with MNT, commercial Pt, and 5 wt% Pt/MNT samples before and after charging for 15 h. Although α-MnO2 was demonstrated to be a promising bifunctional (both for the ORR and the OER) electrocatalyst for Li-O2 batteries93,148,156, as shown in Figure 17b, there was a big difference between the residual Li2O2 ratio for the electrode with MNT catalyst (89 %) and the theoretical value (49.2 %) calculated based on the charging capacity. This reveals that MNT itself as an electrocatalyst cannot promote the OER. The residual Li2O2 ratio for the electrodes with 1, 5, and 10 wt% Pt/MNT catalysts were 52, 49, and 59 %, respectively, as compared to 51 % for an electrode with commercial Pt. Therefore, it can be established that even a small amount of Pt decoration on the MNT enhance the OER kinetics. 5 wt% Pt/MNT exhibited the highest catalytic activity due to its optimizing morphology (highly dispersed and uniform Pt deposition) and proper surface state modifications. In addition, the residual Li2O2 ratio of 49 % for an electrode with 5 wt% Pt/MNT catalyst agreed well with the theoretical value calculated based on the charging capacity (49.2 %), which indicates that the Li2O2 decomposition is the main reaction during the OER in this cell. The charging profiles of the cells using MNT, commercial Pt, 44

home-made Pt nanoparticles, and 5 wt% Pt/MNT catalysts are shown in Figure 17c. It can be seen that MNT decorated by the Pt nanoparticles give rise to lower charging potentials, i.e., reducing the overpotential. Note that no significant difference between the charging profiles of cells with homemade Pt nanoparticles and 5 wt% Pt/MNT was observed.

Li2O2 Charging

Li2O2

a 18

19

20

21

2 Theta (degree)

22

23

Residual ratio of Li2O2 (%)

100

Si

Linear regression for curve fitting: Y= A+kX

80

R2= 0.9901 k= -3.5829

60 40 20

b 0

5

10

15

20

25

30

Time (h)

Figure 18. (a) The operando SR-PXD patterns of an electrode with 5 wt% Pt/MNT catalyst collected every 30 min during charging for 27.5 h, and (b) the curve (−) and linear curve fitting (−·−) of the Li2O2 decomposition during the operando SR-PXD measurement.

The real-time OER kinetics in a Li-O2 cell with 5 wt% Pt/MNT was studied by operando SR-PXD measurements. Figure 18a shows the operando SRPXD patterns for an electrode with 5 wt% Pt/MNT collected during charging . After charging for 27.5 h almost all at a constant current of 40 mA·g Li2O2 in the electrode had decomposed, as shown in Figure 18b. This charging time was a little shorter than the theoretical one (29.5 h) calculated based on the needed charge capacity for all the Li2O2 decomposition in the electrode, which can be attributed to the X-ray accelerated effect on the Li2O2 decomposition112, and/or the low intensity produced by a small amount of Li2O2 left in the electrode at the final stage of charging. Moreover, since the Li2O2 decomposition curve was approximately linear with respect to the charging time (Figure 18b), pseudo-zero-order kinetics was also employed to determine the corresponding decomposition rate constant (k) from Eq. 4. Based on the curve fitting, the Li2O2 degradation in a cell with 5 wt% Pt/MNT followed pseudo-zero-order kinetics, and the decomposition rate constant k was 3.58 h-1.

45

Brief summary of the results of x wt% Pt/MNT (x=0, 1, 5, and 10) electrocatalysts  The Pt nanoparticles were dispersed uniformly on the surface of the MNT in as-prepared samples;  A small amount of Pt decoration on the MNT largely enhanced the OER efficiency in the Li-O2 batteries;  5 wt% Pt/MNT exhibited the highest catalytic activity due to its optimizing morphology (highly dispersed and uniform Pt decorations on the MNT support) and proper surface state (high presences of surface hydroxyl oxygen and oxygen vacancies);  The Li2O2 decomposition for an electrode using 5 wt% Pt/MNT as the catalyst during the charging process followed pseudo-zero-order kinetics.

3.2.2 Study of x wt% electrocatalysts

Ru/MNT

(x=0,

0.5,

and

2)

The real-time OER kinetics was studied by tracking the galvanostatic charging of a Li-O2 battery using operando SR-PXD6,114. Probing the timeresolved OER kinetics of a cell potentiostatically charged is also critical and urgently needed, however it has not yet been studied. In Paper III, the timeresolved OER kinetics of Li-O2 cells with the x wt% Ru/MNT catalysts charged either galvanostatically or potentiostatically was systematically investigated.

Figure 19. SEM micrographs of (a) MNT, (b) 2 wt% Ru/MNT, and TEM micrographs of (c) MNT and (d) 2 wt% Ru/ MNT, and (e) HRTEM micrograph of 2 wt% Ru/ MNT.

46

The SEM micrographs of the MNT and 2 wt% Ru/MNT samples are shown in Figures 19a and b. It can be seen that the MNTs displayed “tube-like” shapes with a length of 1-1.5 μm and a width of 80-100 nm (Figure 19a). Well-dispersed Ru nanoparticles were attached to the surface of the MNTs, as shown in Figure 19b. The Ru content (2.4 wt%) obtained from the EDS analysis was in good agreement with the value (2 wt%) based on its synthesis, indicating that the synthesis of the x wt% Ru/MNT from a simple reduction and mechanical stirring method was successful and controllable. Detailed particle information was obtained by TEM measurements (Figures 19c and d). As depicted in Figure 19c, the nanotube shape of the MNT can clearly be observed and the tube inner diameter was about 40-60 nm. Ru nanoparticles were uniformly distributed on the surface of the MNTs (Figure 19d), which is in agreement with the SEM micrograph. The Ru particle size in the 2 wt% Ru/MNT was about 5-10 nm. As presented in Figure 19e, the HRTEM micrograph of 2 wt% Ru/MNT displayed well-defined lattice fringes with distances of 0.485 and 0.242 nm, corresponding to the (200) lattice plane of the tetragonal-structure α-MnO2 (JCPDS Card No. 04-007-2142)157, and the (101) lattice plane of the hexagonal-structure Ru (JCPDS Card No. 06-0663)158, respectively, which provides direct evidence of a successful loading of the Ru nanoparticles on the surfaces of the MNTs. a

80

theoretical value MNT 0.5 wt% Ru/MNT 2 wt% Ru/MNT home-made Ru

Si

40

b

20

18

20

21

22

4.5

Linear regression for curve fitting: Y= A+kX R2= 0.9913 k= -3.3528

80

Potential (V)

Residual ratio of Li2O2 (%)

100

60

c 0

19

2 Theta (degree)

0

40

Li2O2

Li2O2

60 Charging

Residual Ratio of Li2O2 (%)

100

4.0

3.5

3.0

3

6 9 Time (h)

12

15

0

d 200 400 Capacity (mAh/gLi O ) 2

600

2

Figure 20. (a) The residual Li2O2 ratios for the electrodes with different catalysts (MNT, 0.5 wt% Ru/MNT, 2 wt% Ru/MNT and home-made Ru nanoparticles) after , (b) the operando SR-PXD charging for 15 h at a constant current of 40 mA·g patterns of an electrode with 2 wt% Ru/MNT collected every 30 min during charg, (c) the Li2O2 decomposiing for 15 h at a constant current density of 40 mA·g tion curve (−) and linear curve fitting (−·−), and (d) the charging profile of the same cell as in (b) and (c) during the operando SR-PXD measurement.

47

The electrocatalytic performance of as-prepared samples was evaluated by following the Li2O2 decomposition in Li-O2 cells charged for 15 h at a constant current density of 40 mA·g . Figure 20a shows the residual Li2O2 ratios for the electrodes when using different catalysts (MNT, 0.5 wt% Ru/MNT, 2 wt% Ru/MNT and home-made Ru nanoparticles) after charging for 15 h. The residual Li2O2 ratio for the electrode containing 0.5 wt% Ru/MNT decreased to 66 %, indicating that even a small amount of Ru decoration on the MNT enhanced the OER kinetics. In addition, the residual Li2O2 ratio of 51 % for the electrode containing 2 wt% Ru/MNT was close to 50 % obtained from electrode with home-made Ru nanoparticles and the theoretical value of 49.2 % calculated based on the charging capacity. Figure 20a also shows that the pure MNT sample has limited electrocatalytic activity (yielding a residual Li2O2 ratio of 89 %). Figure 20b displays the operando SR-PXD patterns of an electrode containing 2 wt% Ru/MNT collected during the charging process at a constant current of 40 mA·g . 49 % of the Li2O2 had decomposed after charging for 15 h (Figure 20c), which is close to the theoretical value (50.8 %) calculated based on the charging capacity. According to the curve fitting, it can be concluded that the Li2O2 decomposition in a cell using 2 wt% Ru/MNT catalyst followed pseudozero-order kinetics, and that the decomposition rate constant k was 3.35 h-1. Figure 20d shows the charging profile of the same cell during the operando SR-PXD measurement. Note that the oscillating curve was caused by the fluctuating temperature in the SR-PXD room. The influence of the amount of catalyst on the degradation rate of Li2O2 was also studied by adjusting the weight ratio of electrode composition, as presented in Paper III. Since the OER of a Li-O2 cell with 8 wt% catalyst exhibited the highest efficiency, it can be established that 8 wt% is the optimal weight ratio for the catalyst in the Li2O2-filled electrode.

48

MNT 0.5 wt% Ru/MNT 2 wt% Ru/MNT home-made Ru

2

Capacity (mAh/gLi O ) 2

100 50

4000

0

2000 0

2

4

6

MNT 0.5 wt% Ru/MNT 2 wt% Ru/MNT home-made Ru

a

0

2

4 Time (h)

100 80 60 40 20

c 0

1

2 Time (h)

3

4

100 80 60 40 20

e 0

2

4 Time (h)

600 400 200

b 0

6

6

Residual ratio of Li2O2 (%)

6000

Residual ratio of Li2O2 (%)

800

150

Residual ratio of Li2O2 (%)

Residual ratio of Li2O2 (%) Net current density (mA/gLi2O2)

8000

2

4 Time (h)

6

100 80 60 40 20

d 0

2

4 Time (h)

6

2

4 Time (h)

6

100 80 60 40

f

20 0

Figure 21. (a) Net current density vs. time (Inset: the enlarged part enclosed by the dashed arrow in (a)), and (b) capacity vs. time for Li-O2 cells with different catalysts (MNT, 0.5 wt% Ru/MNT, 2 wt% Ru/MNT and home-made Ru nanoparticles) charged at a constant potential of 4.3 V, and the curves (−) of Li2O2 decomposition calculated based on the charging capacity and the curves (−·−) of Li2O2 decomposition calculated according to the operando SR-PXD results in cells with (c) MNT, (d) 0.5 wt% Ru/MNT, (e) 2 wt% Ru/MNT, and (f) home-made Ru nanoparticles.

The OER kinetics in the Li-O2 cells with different catalysts (MNT, 0.5 wt% Ru/MNT, 2 wt% Ru/MNT and home-made Ru nanoparticles) under a potentiostatic charging condition was also systematically studied by following the decomposition of Li2O2 using operando SR-PXD. Considering that the oxidation potential of Li2O2 in a cell charged galvanostatically is above 4.2 V (Figure 20d), the cells were charged at a constant potential of 4.3 V. Figures 21a and b show the potentiostatic charging profiles of the cells with different catalysts during the charging for 4 or 7.25 h under the operando SR-PXD measurement. It can be seen that the net oxidation current densities obtained from the cells with 0.5 wt% Ru/MNT and 2 wt% Ru/MNT samples were much higher than that obtained from a cell with the MNT (Figure 21a), indicating that the decoration of Ru nanoparticles on the MNT can speed up the OER of a Li-O2 cell under a potentiostatic charging condition. As presented ) of a cell with 2 wt% in Figure 21b, the charge capacity (759 mAh/ 49

) of a Ru/MNT after 7.25 h of charging was close to that (787 mAh/ cell with the home-made Ru nanoparticles, which is in good agreement with the results from galvanostatic charging study. In addition, the OER kinetics in each cell was simultaneously investigated. As shown in Figures 21c-f, the residual Li2O2 ratios for different electrodes calculated based on the SR-PXD results were very close to those calculated based on the charging capacities. This indicates that the Li2O2 decomposition is the main reaction occurring during charging at 4.3 V. Based on the curve fittings, the Li2O2 decomposition in Li-O2 cells with as-prepared catalysts also followed pseudo-zeroorder kinetics (see Paper III). The Li2O2 decomposition kinetics in a cell with 2 wt% Ru/MNT (k2 wt% Ru/MNT = 10.21 h-1) was 2.6 times faster than that in a cell with MNT (kMNT = 2.84 h-1), further confirming that the Ru decoration can promote the OER kinetics. In addition, the promoting effect by suitable loading of Ru on the MNT was even larger than that for the home-made Ru (khome-made Ru = 9.29 h-1) under potentiostatic charging condition, which could be attributed to synergistic effects between the deposited Ru nanoparticles and the MNT support. Brief summary of the results of x wt% Ru/MNT (x=0, 0.5 and 2) electrocatalysts  The Ru nanoparticles were uniformly dispersed on the surface of the MNT in as-prepared samples;  Even a small amount of the Ru decoration on the MNTs largely enhanced the OER efficiency in Li-O2 batteries;  The Li2O2 decomposition in an electrode with x wt% Ru/MNT catalysts during galvanostatic and potentiostatic charging processes followed pseudo-zero-order kinetics;  The OER kinetics in a cell with 2 wt% Ru/MNT charged at a constant potential of 4.3 V was faster than that in a cell with home-made Ru nanoparticles, which could be attributed to the synergistic effects between the deposited Ru nanoparticles and the MNT support.

3.2.3 Study of a novel organic electrocatalyst (Li2C8H2O6) Although there is clearly a need for alternatives, a solid organic electrocatalyst for Li-O2 batteries has so far only been presented once96. In this work, we developed a novel solid organic electrocatalyst- Li2C8H2O6 via an original and low-polluting synthesis route, as shown in Paper IV. The identification of as-prepared Li2C8H2O6 was carried out by NMR, FTIR, and electrochemical techniques. The details can be found in Paper IV.

50

a

no catalyst

Si

Si

after 12.5 h charging

28

2 Theta (degree)

100 80

b

before charging

before charging

before charging

after 12.5 h charging 30 32 34

Li2O2

Li2O2

Li2O2

30

32

after 12.5 h charging

34

28

2 Theta (degree)

no catalyst theoretical value Li2C8H2O6

4.5

30

32

34

2 Theta (degree)

c

commercial Pt

4.0

60 40 20

Potential (V)

Residual ratio of Li2O2 (%)

Li2O2

Li2O2

Si

28

commercial Pt

Li2C8H2O6 Li2O2

3.5

no catalyst commercial Pt Li4C8H2O6

3.0 0

0

3

6

9

12

Time (h)

Figure 22. (a) The in situ XRD patterns before and after charging, (b) residual Li2O2 ratios after charging, and (c) charging curves of Li2O2-filled electrodes with and without Li2C8H2O6 as well as for commercial Pt during charging for 12.5 h at a . constant current density of 40 mA·g

The catalytic activity of Li2C8H2O6 was evaluated by following the Li2O2 decomposition in a Li-O2 cell charged for 12.5 h at a constant current density of 40 mA·g . Reference cells with commercial Pt and without catalyst were employed to determine the electrocatalytic performance of Li2C8H2O6. As seen in Figure 22b, there was a significant difference between the residual Li2O2 ratios for the electrodes with Li2C8H2O6 catalyst (58 %) and without catalyst (79 %) after charging for 12.5 h, which indicates that Li2C8H2O6 can promote the OER in a Li-O2 cell. This residual Li2O2 ratio of 58 % for the electrode with Li2C8H2O6 was very close to that of 57 % for the electrode with commercial Pt. In addition, this value is also closed to the theoretical value of 56.7 % calculated based on the charging capacity, indicating that the Li2O2 decomposition was the main reaction during the charging process. It can be seen that the employment of the Li2C8H2O6 catalyst decreased the charging potential, as compared to that for a cell without catalyst (Figure 22c). Note that no big difference between the charging profiles for the cells with Li2C8H2O6 and commercial Pt could be observed.

51

100

Charging

Li2O2

Li2O2

a 17

18

19

20

21

22

Residual ratio of Li2O2 (%)

Si

Linear regression for curve fitting: Y= A+kX

90

R2= 0.9984 k= -3.2401

80 70 60

b

0

23

2

4

6

8

10

12

Time (h)

2 Theta (degree)

Figure 23. (a) The operando SR-PXD patterns of a Li2O2-filled electrode containing Li2C8H2O6 collected every 30 min during charging for 12.5 h at a constant current , and (b) the curve (−) and linear curve fitting (−·−) of the density of 40 mA·g Li2O2 degradation during the operando SR-PXD measurement.

As shown in Figure 23b, 42 % of the Li2O2 had decomposed after charging for 12.5 h. From the curve fitting, it can be concluded that the Li2O2 decomposition in the Li-O2 cell with Li2C8H2O6 charged at a constant current densifollowed pseudo-zero-order kinetics, and that the decomty of 40 mA·g position rate constant k was 3.24 h-1. st

1 20th

2nd 30th

th

5 40th

a

th

10

4.0 3.5 0.86 V

3.0

1st

4.5 Potential (V)

Potential (V)

4.5

2nd

b

5th

4.0 3.5

1.19 V

3.0 2.5

20 Cycle No.

30

40

900

0

600

300

300

600

0 0

2

4

6

Cycle No.

8

900 10

(mAh/gcarbon)

d

Delivered Capacity from charge

10

200 400 600 800 Capacity (mAh/gcarbon)

900

(mAh/gcarbon)

600

Delivered Capacity from discharge

300

(mAh/gcarbon)

300

Delivered Capacity from charge

0

c

600

0 0

0

200 400 600 800 Capacity (mAh/gcarbon)

900

(mAh/gcarbon)

Delivered Capacity from discharge

2.5 0

Figure 24. The discharge and charge profiles of a cell using a constant current densiwithin a potential range of 2.2-4.5 V vs. Li+/Li (a) with ty of 70 mA· g Li2C8H2O6 catalyst and (b) without catalyst, and the delivered capacity from the discharge and charge vs. the cycle number (c) with Li2C8H2O6 catalyst and (d) without catalyst.

The stability of the electrolyte has been recognized as a big issue for Li-O2 batteries, since the coulombic efficiency and cyclability of a Li-O2 cell will inevitably be reduced by electrolyte degradation. Although this work was 52

focused on the development of electrocatalyst for the OER, the overpotential and cyclability of a Li-O2 cell containing Li2C8H2O6 were also studied. Figure 24a shows typical discharge and charge profiles of a cell using Li2C8H2O6 at a constant current density of 70 mA·g with a fixed ca. This cell exhibited an overpotential of 0.86 V pacity of 840 mAh/ on the 1st cycle, which was lower than 1.19 V obtained from a cell without catalyst (Figure 24b). Note that the overpotentials after the 2nd cycle increased, mainly due to the high charging potentials. This could be attributed to the influence of side products formed during the discharging process. When the polarization increased during the charging process and the charging potential exceeded 4.0 V, the alkaline carboxylate functional groups of Li2C8H2O6 gradually underwent a decarboxylation reaction known as Kolbe electrolysis159. Thereby, the catalyst lost its activity for the Li2O2 degradation. The cell exhibited 40 cycles with a setting capacity of 840 mAh/ (Figure 24c) as compared to 4 cycles for a cell without catalyst (Figure 24d), indicating that the use of Li2C8H2O6 could significantly improve the battery cyclability. Brief summary of the results of Li2C8H2O6 organic electrocatalyst  Li2C8H2O6 exhibited similar electrocatalytic activity as the benchmark material-commercial Pt for the Li2O2 decomposition during the OER;  The Li2O2 decomposition in an electrode with Li2C8H2O6 during the charging process followed pseudo-zero-order kinetics;  A Li-O2 cell containing Li2C8H2O6 catalyst displayed an overpotential of 0.86 V on the 1st cycle and a cyclability of over 40 cycles;

3.3 Study of a binder-free NCPE So far, carbonaceous materials have been used as fillers to provide the porosity and electronic conductivity for composite electrodes. To promote the electrochemical performance, incorporation of heteroatoms in a well-defined way has been regarded as a feasible strategy to modify the nature and chemical properties of pure carbon61. Among those heteroelements, N, having a comparable atomic size to carbon and five valence electrons for bonding with the carbon atom, is widely adopted98,160. N-doped carbon can not only improve the Li+ diffusion and transfer by generating defects and by withdrawing electrons from the carbon atom, but also increase the conductivity of the resulting carbon, which subsequently displays better battery performance than pure carbon materials161-164. Many approaches have been reported for the synthesis of N-doped carbon164-166. Those synthetic routes normally require multiple or complicated steps, with a high cost and low yield, which inevitably limit large-scale fabrication. Therefore, to obtain high performance N-doped carbon by an easy-to-operate and low-cost method is still 53

a challenge. Equally, the architecture of the O2 electrode also influences the battery performance. Recently, a binder-free O2 electrode, designed to prevent the negative influence of the binder, has been reported98-102. In Paper V, the results of using a binder-free NCPE were presented. The fabrication of this NCPE involves cheap raw materials (e.g., Cladophora sp. green algae) and easy operation (e.g., N doping by a carbonization of a N-rich polymer), which is especially suitable for large-scale production. The rate performance, coulombic efficiency, and cyclability of Li-O2 cells with this new designed electrode were investigated.

a

(002)

b

G D

(101)

10

20

30

40

50

60

1300

70

150 100 50

0.008

0.004

0.000

10

20

30

40

50

60

Pore width (nm)

adsorption desorption

0.2

0.4

0.6 P/P0

1600

d

80

60

PPy/cellulose PPy/cellulose-CCFs

40

0.0

1500

100

c

0.012

weight ratio(wt%)

Quantity (cm3g-1)

200

Incremental pore volume (cm3/g

250

1400

Raman shift (cm-1)

2 Theta (degree)

0.8

1.0

0

200

400

600

800

1000

o T ( C)

Figure 25. (a) XRD pattern, (b) Raman spectrum, (c) N2 adsorption/desorption isotherm curves (Inset: the pore size distribution) of NCPE, and (d) TG curves of the PPy/cellulose and PPy/cellulose-CCFs samples.

The XRD pattern of NCPE is shown in Figure 25a. The diffraction peaks at 2θ = 25.5° and 42.7° are indexed to the (002) and (101) reflections of graphitic carbon (JCPDS Card No.00-023-0064), respectively167. Figure 25b shows the Raman spectrum of NCPE. The band at 1373 cm-1 (D-band) originates from defect-induced mode161, while the band at 1575 cm-1 (G-band) corresponds to the formation of well-graphitized carbon163, which is in good agreement with the XRD result. Figure 25c exhibits the N2 adsorption/desorption isotherm curves for NCPE, which display type IV shapes (mesoporous structure) and yield a surface area of 107 m2·g-1. The NCPE has a pore size distribution in the range of 1-65 nm (Inset of Figure 25c). TG analysis not only characterizes the pyrolytic treatment, but also provides a feasible means for the quantitative analysis of the NCPE components. The NCPE is prepared by a simple carbonization of a PPy/cellulose-CCFs composite, in which CCFs remain as carbon residues after the pyrolytic treat54

ment. In order to quantify the content of N-doped carbon obtained after the pyrolysis of the PPy/cellulose precursor, we also synthesized PPy/cellulose (without CCFs) under similar conditions. The TG curves for the PPy/cellulose and PPy/cellulose-CCFs samples are shown in Figure 25d. According to the TG results, there is 64 wt% of N-doped carbon and 36 wt% of CCFs in the NCPE (detailed estimation can be found in Paper V).

Figure 26. (a) SEM micrograph of NCPE, (b-d) SEM micrographs of the region enclosed by the red frame in (a), (e) SEM micrograph of the region enclosed by the green frame in (a), EDS elemental mapping of (f) C, (g) N, and (h) O corresponding to (e).

55

Cellulose in the PPy/cellulose-CCFs composite acts as a substrate and structural template for the formation of the PPy coating168. The employment of CCFs can not only increase the mechanical strength and conductivity of the NCPE, but also support the binder-free electrode as a skeleton. Note that the sample without CCFs was very fragile. Figure 26 shows the SEM micrographs of NCPE and the corresponding EDS elemental mapping information for C, N, and O. CCFs were determined to have an average length of 1 mm, and are embedded in NCPE (Figure 26a). As shown in Figures 26b-d, the NCPE exhibited a bird’s nest microstructure with a rough surface of its branches (~70 nm in width). It has been proved that this bird’s nest architecture can provide self-standing electrodes with considerable mechanic durability, fast Li+ and O2 diffusion, and enough space for discharge product deposition98,169. The elemental information for NCPE was obtained by EDS elemental mapping. CCFs mainly consisted of C elements (Figure 26f), while the N-doped carbon was composed of well-dispersed C, N and O (Figures 26f-h), indicating that the N doping via the carbonization of a N-rich polymer composite was successful.

56

a

0.1 mA/cm2 0.2 mA/cm2 2 0.4 mA/cm 2 0.8 mA/cm

2.8 2.7 2.6

4.5

b

4.0

Potential (V)

Potential (V)

2.9

3.5 3.0 2.5

2.5

0

2000

4000

6000

0

8000

4.5

c

4.2

3.6

1st 5th th 10 th 20 th 30

3.3 0.68 V

3.0

Potential (V)

Potential (V)

4000

d

4.0

3.9

1st 5th th 10

3.5 0.85 V

3.0 2.5

2.7 0

100

200

300

400

0

500

Capacity (mAh/gN-doped carbon)

4.5

3.0

Potential (V)

st

1 5th

0.99 V

200

300

400

500

f

4.0

3.5

100

Capacity (mAh/gN-doped carbon)

4.5

e

4.0

Potential (V)

2000

Capacity(mAh/gN-doped carbon)

Delivered capacity(mAh/gN-doped carbon)

3.5 1st 3th

1.34 V

3.0 2.5

2.5 0

100

200

300

400

Capacity (mAh/gN-doped carbon)

500

0

100

200

300

400

500

Capacity (mAh/gN-doped carbon)

Figure 27. (a) The discharge profiles for Li-O2 cells with NCPEs at various current densities of 0.1-0.8 mA/cm2 and a cut off voltage of 2.4 V, (b) discharge/charge profiles for the 1st cycle of a Li-O2 cell with NCPE at a current density of 0.2 mA/cm2 within a potential range of 2.4-4.5 V vs. Li+/Li, and the cycling response of the Li-O2 cells with NCPEs at constant current densities of (c) 0.1, (d) 0.2, (e) 0.4 mA/cm2, and (f) the cycling response of a Li-O2 cell with a reference electrode made from the chopped NCPE powder and binder at a constant current density of 0.2 mA/cm2 within a potential range of 2.2-4.5 V vs. Li+/Li.

The capacities were calculated based on the weight of N-doped carbon in the NCPE, since the capacity of a reference cell with an undoped carbon paper electrode could be neglected (see Paper V). As presented in Figure 27a, the 1st discharge capacity at a current density of 0.1 mA/cm2 reached 8040 mAh/gN-doped carbon. Note that the achieved discharge capacity and operating potential decreased with increasing current density. Figure 27b shows the galvanostatic discharge/charge curves on the 1st cycle for a Li-O2 cell at a current density of 0.2 mA/cm2. The discharge and charge capacities were 4759 and 3865 mAh/gN-doped carbon, respectively, indicating that this cell exhib57

its a coulombic efficiency of 81 %. The cyclability of Li-O2 batteries with NCPEs under a specific capacity limit of 480 mAh/gN-doped carbon at various current densities (0.1-0.4 mA/cm2) were also investigated. A Li-O2 cell containing NCPE exhibited a cyclability of more than 30 cycles at a constant current density of 0.1 mA/cm2 (Figure 27c). Note that two potential stages were observed during the charging process. The first might correspond to the Li2O2 oxidation, while the other could be assigned to electrolyte decomposition25,170 and/or the LiOH oxidation171. LiOH could be generated due to the introduction of H2O into the system via DMSO, and/or the oxidation of DMSO by the reactive oxygen species (e.g., LiO2 and Li2O2)38,172. The overpotential increased gradually with cycle number, mainly resulting from an increased charging potential. In addition, the discharge capacity dropped after 30 cycles. Considering that DMSO is not completely stable, the instability of the interface between lithium metal and DMSO could be another reason for the sudden death of the cell6,39. The cells sustained 10 and 5 cycles with capacity limit of 480 mAh/gN-doped carbon at constant current densities of 0.2 and 0.4 mA/cm2, respectively (Figures 27d and e). The cell overpotential on the 1st cycle increased from 0.68 to 0.99 V when raising the current density from 0.1 to 0.4 mA/cm2 (Figures 27c-e). In an attempt to investigate the influence of the binder-free bird’s nest architecture of NCPE on the battery performance, we prepared a reference electrode made from chopped NCPE powder and binder in a weight ratio of 90:10. Figure 27f shows the cycling response of a cell with this reference electrode at a constant current density of 0.2 mA/cm2. It can be seen that the cell overpotential of 1.34 V on the 1st cycle was much higher than that (0.85 V) for a cell with NCPE. In addition, this cell only sustained 2 cycles with a capacity limit of 480 mAh/gN-doped carbon, and there was an large capacity fading on the 3rd cycle. Two possible reasons are given for this poor cyclability. One is that an obvious aggregation of the electrode materials (see Figure 28), resulting from the binder addition, leads to the decreasing rate of O2 diffusion and the limiting space for Li2O2 deposition. The other might be that binder-involved parasitic reactions interfered with electrochemical processes.

58

Figure 28. SEM micrograph of a reference electrode made from chopped NCPE powder and binder.

Brief summary of the results of the binder-free NCPE  The fabrication of NCPE involves cheap raw materials and easy operations, which is suitable for a large-scale production;  The NCPE exhibited a bird’s nest microstructure, which could provide the self-standing electrode with considerable mechanic durability, fast Li+ and O2 diffusion and enough space for discharge product deposition;  The NCPE was composed of N-containing functional groups, which might promote the electrochemical reactions;  Binder-free architecture designs can prevent binder-involved parasitic reactions.

59

4 Concluding remarks and further outlook

The aim of this thesis was to study the O2 electrode performance in the Li-O2 battery. The research in this work has covered the investigation of parameters influencing Li2O2 decomposition, the development of highly active electrocatalysts, and the design of low-cost and versatile binder-free O2 electrodes for Li-O2 batteries. As the main technique, SR-PXD was used both as a continuous light source to advance the electrochemical decomposition of Li2O2 under X-ray illumination and as an operando tool to probe the Li2O2 decomposition during battery charging. In the first track, the accelerating effect of X-rays on the electrochemical decomposition of Li2O2 was explored, as presented in Paper I. The decomposition rate of Li2O2 was proportional to the intensity of the X-ray beam. It is proposed that the electrochemical decomposition of Li2O2 under X-ray illumination might involve a three-step reaction with [Li2O2]x+ and Li2-xO2* as intermediates, exhibiting pseudo-zero-order kinetics. The X-ray interaction promoted the kinetics, while the external potential overcame the thermodynamic barriers. In the second track, three electrocatalysts were developed, which exhibited good electrocatalytic performances during the OER. Their activities were evaluated by probing the Li2O2 decomposition in commercial Li2O2-filled electrodes during the charging process. In addition, the time-resolved OER kinetics for the electrocatalyst-containing electrodes charged galvanostatically and potentiostatically were systematically investigated using operando SR-PXD. As presented in Papers II and III, the MNT supported by welldispersed Pt or Ru nanoparticles (Pt/MNT or Ru/MNT) was prepared by a simple reduction and mechanical stirring method. It was found that a small amount of decoration with Pt or Ru nanoparticles on the MNTs enhanced the OER efficiency in a Li-O2 cell. In Paper IV, a novel solid organic electrocatalyst, Li2C8H2O6, was prepared via a low-cost and low-polluting synthetic route, which was expected to overcome the shortcomings of inorganic catalysts during the cell cycling. The electrocatalytic activities of 5 wt% Pt/MNT, 2 wt% Ru/MNT or Li2C8H2O6 were similar to that of benchmark noble metals (Pt and Ru), indicating that the investigation of Pt/MNT, Ru/MNT or Li2C8H2O6 can be a promising strategy to enhance the OER efficiency and to decrease the cost of raw materials. Moreover, Li2O2 de60

composition in an electrode with 5 wt% Pt/MNT, 2 wt% Ru/MNT or Li2C8H2O6 in a Li-O2 cell followed pseudo-zero-order kinetics. In the last track, a binder-free NCPE for Li-O2 batteries was presented in Paper V. The fabrication of NCPE involved cheap raw materials (e.g., Cladophora sp. green algae) and easy operation (e.g., N doping via a carbonization of N-rich polymer), which is suitable for large-scale production. It displayed a bird’s nest microstructure, which could provide the self-standing electrode with considerable mechanic durability, fast Li+ and O2 diffusions and enough space for discharge product deposition. The NCPE was composed of N-containing functional groups, which might promote the electrochemical reactions. In this thesis, the experimental results with regard to the critical factors (e.g., X-ray illumination and electrocatalyst) influencing the OER kinetics and a binder-free NCPE for Li-O2 batteries have been discussed. Nonetheless, there are still many unresolved issues related to this fascinating and intriguing electrochemical system. In the following, a few are presented:  At the current stage of Li-O2 batteries, after screening all known electrolytes, parasitic reactions have been identified. Therefore, searching for a stable electrolyte is the priority in this field. Fundamentally, it determines whether a truly rechargeable Li-O2 battery can be built or not.  In parallel with the importance of the electrolyte, developing an O2 electrode with high stability and performance is also of great interest. Porous carbon with or without catalyst is the current choice of O2 electrode material, however, some observed side reactions have been attributed to carbon decomposition. In addition, although some efforts have been devoted to developing the electrocatalysts, the real role of the catalyst is controversial and the catalytic mechanism is not yet clear. Moreover, to guide the design of the electrocatalyst, a clear understanding of fundamental knowledge for the Li-O2 electrochemistry and a relative stable system are urgently needed.  The dendrite formation and poor stability in contact with the electrolyte in the presence of O2 and active products are the two main obstacles for the application of lithium metal anode. Considering that a convincing and highly efficient alternative has not yet been developed, using a solid membrane to separate Li metal from the electrolyte could be a suggestion here.  So far, all the reported work is based on cells using dry and pure O2. An obvious practical implication is the use of ambient air instead of O2 is more desirable. Impurities in air such as H2O and CO2 can affect the safety, cycle life and reliability of the battery. The goal might be achieved with use of O2 selective membranes in the future.  The cross-talk between each component of the Li-O2 battery complicates the studies. A suggested approach is to investigate each component of the battery individually before assembling the more complex full cell. 61

Although some start-up companies have embarked on R&D efforts to pave their industrial roads, some experts predict that Li-O2 batteries might not reach commercial application within the next 20 years, if the fundamental questions concerning their function remain unanswered.

62

5 Sammanfattning på Svenska

Litiumjonbatterierna kommersialiserades år 1991, vilket ledde till ett uppsving för portabel elektronik under de två följande årtiondena. På grund av en relativt hög energitäthet är det för nuvarande inte enbart förstahandsvalet för energilagring i el- och elhybridfordon, utan används även för storskalig energilagring till stöd för förnyelsebar energiomvandlare. Hittills har många fundamentala och tillämpade framsteg gjorts inom detta område. Men även om den teoretiska kapaciteten hos elektrodmaterialen kan erhållas, så är energitätheten (energi per volymenhet) samt den specifika energin (energi per massenhet) för de bästa tillgängliga litiumjonbatterierna för låga för att uppfylla kraven hos elfordon på lång sikt. Således har mer forskning nyligen ägnats åt energilagringssystem som kan överskrida litiumjonbatteriets gränser. Ett sådant system är litium-syrgas batteriet (förkortat Li-O2 batteriet), vilket har fått stor uppmärksamhet de senaste åren på grund av dess höga specifika energi. En schematisk figur av Li-O2 batteriet visas i Figur 1.

Figur 1. En schematisk figur av Li-O2 batteriet.

I ett idealiskt Li-O2-system, är den elektrokemiska reaktionen 2Li+ + O2 + 2e- ⇄ Li2O2, där batteriets urladdning beskrivs med reaktionen åt höger (ORR) medan laddningen av batteriet beskrivs av den omvända reaktionen (OER). Det har blivit välkänt att energilagringskapaciteten, den tillgängliga effekten samt livslängden bestäms av materialet och strukturen hos O2elektroden. Syftet med denna avhandling var att undersöka O2-elektrodens funktion i Li-O2 batterier. Forskningen i detta arbete har inkluderat studier av faktorerna som påverkar den elektrokemiska sönderdelningen av Li2O2, ut63

vecklingen av högaktiva elektrokatalysatorer samt designen av en billig och lätthanterlig O2-elektrod utan bindemedel för Li-O2 batteriet. SR-PXD, som var huvudtekniken, användes dels som kontinuerlig ljuskälla i syfte att undersöka den elektrokemiska sönderdelningen av Li2O2 under röntgenbelysning, dels som ett operando verktyg vilket möjliggjorde studier av nedbrytningen av Li2O2. Först utforskades den accelererande inverkan av röntgenstrålar på sönderdelningen av Li2O2. Hastigheten för nedbrytningen var proportionell mot intensiteten av röntgenstrålningen. Den elektrokemiska sönderdelningen av Li2O2 under påverkan av röntgenbelysning involverade en trestegsreaktion, med [Li2O2]x+ and Li2-xO2* som intermediåt, som kunde beskrivas som en pseudo-nollteordningens reaktion. Därefter utvecklades tre elektrokatalysatorer (Pt/MNT, Ru/MNT och Li2C8H2O6). Deras aktiviteter utvärderades genom att följa nedbrytningen av kommersiell Li2O2 under laddningsprocessen. Dessutom studerades kinetiken för OER tidsupplöst i Li-O2 celler innehållande de ouka elektrokatalysatorerna, som laddades med konstant ström eller med konstant spänning, systematiskt genom att använda operando SR-PXD. Det konstaterades att en liten mängd av Pt eller Ru nanopartiklar på MNT avsevärt kunde förbättra effektiviteten och selektiviteten för OER i en Li-O2 cell. Li2O2sönderdelningen för en elektrod med 5 wt% Pt/MNT, 2 wt% Ru/MNT eller Li2C8H2O6 i en Li-O2 cell som laddas med konstant ström eller med konstant spänning kunde beskrivas som en pseudo-nollteordningens reaktion. Slutligen presenteras en ny NCPE utan bindemedel för Li-O2 batteriet. Framställningen av NCPE utgår från billiga råmaterial (t.ex. Grönslick grönalger) och syntetiseras enkelt (t.ex genom att åstadkomma en N-dopning till följd av förkolning av en N-rik ploymer), vilket är särskilt tillämpbart för storskalig produktion. NCPE-elektrodernas struktur liknar ett fågelbo, och kan fungera som fristående elektroder med avsevärd mekaniskt hållbarhet, snabb O2 transport samt tillräckligt utrymme för reaktionsprodukterna som bildas vid urladdning av batteriet. N-dopningen av NCPE gav upphov till funktionella grupper, vilket kan främja den önskade elektrokemiska reaktionen. Även om Li-O2 batterier visar stora möjligheter samt att många framsteg redan har gjorts, så är huvudfrågorna som fortfarande kräver förbättringar följande: 1) Hur för man en djupare förståelse för den grundläggande elektrokemin; 2) Hur hittar man en långvarigt stabil elektrolyt med hög O2 löslighet och snabb diffusion; 3) Hur designar man en stabil O2 elektrod med utomordentliga elektrokatalysator egenskaper; 4) Hur förhindrar man av dendritbildning på ytan hos Li-elektroden; 5) Hur skyddar man litium elektroden mot CO2 och vatten.

64

6 Acknowledgements

First and foremost, I am deeply indebted to my main supervisor: Jiefang Zhu. Thank you for giving me the opportunity to work with this project and for your professional guidance, effort and patience during these four years. Kristina Edström, my co-supervisor, thank you for your trust, support, and encouragement. Without you, this work would never have been possible. Torbjörn Gustafsson, thank you for your refreshing inspirations and overall viewpoint on this project. I am grateful for your enormous efforts to help me with my experiments when I am lost. You are crucial to this project, I have benefited greatly from your relentless support. Leif Nyholm, thanks for your comments on this project and sharing your knowledge with me. Daniel Brandell, thanks for your quick replies for correcting our manuscript and your enthusiasm in the work. Special thanks to Matthew Roberts and Reza Younesi for introducing LiO2 batteries to me. Matt, thanks for your valuable comments and encouragements when I lost confidence. Reza, Hej Potatis! It has been a great help to receive your advices on XPS measurements and scientific comments on this project. I would like to acknowledge the staff at Maxlab I711 beamline for the kindness and patience. Olivier Balmes and Francisco Martinez (Willy), thanks for your helps, understandings, and kindly considerations when I asked for the beamtime extensions. Torbjörn Gustafsson, Mohammed Dahbi, Matthew Roberts, Reza Younesi, Rickard Eriksson, Andreas Blidberg, Adam Sobkowiak, Chenjuan Liu, and Viktor Renman, thank you for sharing beamtime with me at Maxlab. Håkan Rundlöf, thanks for your help with running the in-house diffractometer. Stéven Renault, thanks for your “super” organic catalyst and fruitful discussions. Zhaohui Wang, thanks for sharing your academic experiences, special food recipes and all funny moments. Chao Xu, thanks for your help and comments on XPS measurements. I am enjoying the “Las vegas” atmosphere at your home. Yue Ma, thanks for your help on TEM measurements and suggestions for travelling and shopping. William Brant, thanks for your kindly help for correcting the thesis. I hope I did not destroy your summer holidays too much. Andreas Blidberg, thanks for your big help on Sammanfattning på Svenska. Peng Zhang, thanks for your help on TG measurements. Bing Sun and Yuan Tian, thanks for your comments on thesis writing and your darings to travel with me. I will never forget those wonderful time. 65

Many thanks to all the colleagues and friends in the battery group: Alina, Anti, Bertrand, Cesar, Charlotte, David, Gabi, Fabian, Erik, Fredrik B., Fredrik L., Habtom, Jonas, Julia, Karima, Kasia, Mario, Matthew L., Nina, Sara Munktell, Sara Malmgren, Ruijun, Solveig, Shruti, Taha, Tim, and Wei for sharing the wonderful time at the lab, and during the conference trips, fika, dinners and parties. Thank you all for making these years enjoyable! A special thanks to Girma, it is a pleasure to share an office with you. Good luck for your defense next year. Special thanks to Henrik Eriksson for all the technical assistances at the lab. I cannot imagine how messy the lab would be without you. I would like to express my gratitude towards Eva Larsson, Kristina Israelsson (Tatti), and Diana Bernelind for the supports with all the administrative procedures, and to Anders and Peter for all technical supports. I am very grateful for financial support from China Scholarship Council during these four years. Gertrud Thelins, GradSAM 21, Anna Maria Lundins, ÅForsk and Liljewalchs foundations are acknowledged for the research and conference grants. Many thanks to my Chinese friends for friendship and support: Fang and Tao, I-Ming and Lina, Ocean, Shuyi, Shuainan and Peng, Song and Yue, Wei and Miao, Ying, Yu, Xia and Shuanglin, and Zhen. I will remember your smiles forever. Special thanks to Bing’s and Ji’s families for sharing happiness and sorrow together. It is great to have friends who have been ready for giving a hand. 感谢赵姨和叔叔的照顾，鼓励和支持。Zexi, tack för din hjälp på Sammanfattning på Svenska. Du är den bästa kusin i världen. Xiaolong, I hope you will have a wonderful future. Good luck for your PhD study in Canada. 感谢我远在中国的亲朋好友们的理解和关心。奶奶，爷爷，姥 爷，希望你们可以看到我拿到博士学位，我特别想念你们。姥姥，希 望你健康长寿，无论我在哪里你永远是我最牵挂和最爱的人。感谢刘 岩同学在我求学道路上给予的理解和支持。今年是我们认识的第十八 年，希望未来的 n 个十八年，我们继续前行。I will always support you, follow you, and love you. 最后，我要特别感谢我的父母，谢谢你们的培 养，理解，和支持，没有你们，我不可能完成学业。爸爸妈妈，我超 级爱你们 ！！！

Jia Liu 刘佳 66

7 References

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19) (20) (21) (22)

International Energy Outlook 2014, U.S. Energy Information Administration. http://www.energimyndigheten.se/en/About-us/Press-/Press releases/Swedenhas-the-highest-proportion-of-renewable-energy-in-the-EU/ 2014, Swedish Energy Agency. http://www.aps.org/publications/apsnews/200603/history 2015, American Physical Society. Miller, P. Johnson Matthey Technology Review 2015, 59, 4. Bruce, P. G.; Freunberger, S. A.; Hardwick, L. J.; Tarascon, J. M. Nature Materials 2012, 11, 19. Liu, J.; Younesi, R.; Gustafsson, T.; Edström, K.; Zhu, J. Nano Energy 2014, 10, 19. Jiang, K. M. A. a. Z. Journal of The Electrochemical Society 1996, 141, 1. Lu, J.; Li, L.; Park, J. B.; Sun, Y. K.; Wu, F.; Amine, K. Chemical Reviews 2014, 114, 5611. Read, J. Journal of The Electrochemical Society 2002, 149, A1190. Read, J.; Mutolo, K.; Ervin, M.; Behl, W.; Wolfenstine, J.; Driedger, A.; Foster, D. Journal of The Electrochemical Society 2003, 150, A1351. Xu, W.; Xiao, J.; Zhang, J.; Wang, D.; Zhang, J.-G. Journal of The Electrochemical Society 2009, 156, A773. Ogasawara, T.; Débart, A.; Holzapfel, M.; Novák, P.; Bruce, P. G. Journal of the American Chemical Society 2006, 128, 1390. Débart, A.; Bao, J.; Armstrong, G.; Bruce, P. G. Journal of Power Sources 2007, 174, 1177. Jian, Z.; Liu, P.; Li, F.; He, P.; Guo, X.; Chen, M.; Zhou, H. Angewandte Chemie 2014, 53, 442. Younesi, R.; Urbonaite, S.; Edström, K.; Hahlin, M. The Journal of Physical Chemistry C 2012, 116, 20673. Lu, Y-C.; Yang, S-H. The Journal of Physical Chemistry Letters 2013, 4, 93. Lu, Y-C.; Gallant, B. M.; Kwabi, D. G.; Harding, J. R.; Mitchell, R. R.; Whittingham, M. S.; Yang, S-H. Energy & Environmental Science 2013, 6, 750. McCloskey, B. D.; Bethune, D. S.; Shelby, R. M.; Girishkumar, G.; Luntz, A. C. The Journal of Physical Chemistry Letters 2011, 2, 1161. Younesi, R.; Hahlin, M.; Edström, K. ACS Applied Materials & Interfaces 2013, 5, 1333. Younesi, R.; Hahlin, M.; Roberts, M.; Edström, K. Journal of Power Sources 2013, 225, 40. Wang, H.; Yang, Y.; Liang, Y.; Zheng, G.; Li, Y.; Cui, Y.; Dai, H. Energy & Environmental Science 2012, 5, 7931. Lim, H.-D.; Gwon, H.; Kim, H.; Kim, S.-W.; Yoon, T.; Choi, J. W.; Oh, S. M.; Kang, K. Electrochimica Acta 2013, 90, 63.

67

(23) Younesi, R.; Hahlin, M.; Treskow, M.; Scheers, J.; Johansson, P.; Edström, K. The Journal of Physical Chemistry C 2012, 116, 18597. (24) Sergeev, A. V.; Chertovich, A. V.; Itkis, D. M.; Goodilin, E. A.; Khokhlov, A. R. Journal of Power Sources 2015, 279, 707. (25) Li, F.; Zhang, T.; Zhou, H. Energy & Environmental Science 2013, 6, 1125. (26) Svens, P.; Eriksson, R.; Hansson, J.; Behm, M.; Gustafsson, T.; Lindbergh, G. Journal of Power Sources 2014, 270, 131. (27) Xu, C.; Lindgren, F.; Philippe, B.; Gorgoi, M.; Björefors, F.; Edström, K.; Gustafsson, T. Chemistry of Materials 2015, 27, 2591. (28) Mizuno, F.; Nakanishi, S.; Kotani, Y.; Yokoishi, S.; Iba, H. Electrochemistry 2010, 78, 403. (29) Yang, Y.; Sun, Q.; Li, Y.-S.; Li, H.; Fu, Z.-W. Journal of Power Sources 2013, 223, 312. (30) Assary, R. S.; Lu, J.; Du, P.; Luo , X.; Zhang, X.; Ren, Y.; Curtiss, L. A.; Amine, K. ChemSusChem 2013, 6, 51. (31) Thapa, A. K.; Shin, T. H.; Ida, S.; Sumanasekera, G. U.; Sunkara, M. K.; Ishihara, T. Journal of Power Sources 2012, 220, 211. (32) Sharon, D.; Etacheri, V.; Garsuch, A.; Afri, M.; Frimer, A. A.; Aurbach, D. The Journal of Physical Chemistry Letters 2013, 4, 127. (33) Younesi, R.; Norby, P.; Vegge, T. ECS Electrochemistry Letters 2014, 3, A15. (34) Jung, H. G.; Jeong, Y. S.; Park, J. B.; Sun, Y. K.; Scrosati, B.; Lee, Y. J. ACS Nano 2013, 7, 3532. (35) Read, J. Journal of The Electrochemical Society 2006, 153, A96. (36) Bryantsev, V. S.; Giordani, V.; Walker, W.; Blanco, M.; Zecevic, S.; Sasaki, K.; Uddin, J.; Addison, D.; Chase, G. V. The Journal of Physical Chemistry A 2011, 115, 12399. (37) Bryantsev, V. S.; Faglioni, F. The Journal of Physical Chemistry A 2012, 116, 7128. (38) Luntz, A. C.; McCloskey, B. D. Chemical Reviews 2014, 114, 11721. (39) Peng, Z.; Freunberger, S. A.; Chen, Y.; Bruce, P. G. Science 2012, 337, 563. (40) Xu, D.; Wang, Z. L.; Xu, J. J.; Zhang, L. L.; Zhang, X. B. Chemical Communications 2012, 48, 6948. (41) Roberts, M.; Younesi, R.; Richardson, W.; Liu, J.; Zhu, J.; Edstrom, K.; Gustafsson, T. ECS Electrochemistry Letters 2014, 3, A62. (42) Gittleson, F. S.; Sekol, R. C.; Doubek, G.; Linardi, M.; Taylor, A. D. Pysical Chemistry Chemical Physics 2014, 16, 3230. (43) Sharon, D.; Afri, M.; Noked, M.; Garsuch, A.; Frimer, A. A.; Aurbach, D. The Journal of Physical Chemistry Letters 2013, 4, 3115. (44) Allen, C. J.; Mukerjee, S.; Plichta, E. J.; Hendrickson, M. A.; Abraham, K. M. The Journal of Physical Chemistry Letters 2011, 2, 2420. (45) Laino, T.; Curioni, A. New Journal of Physics 2013, 15, 095009. (46) Chen, Y.; Freunberger, S. A.; Peng, Z.; Bardé, F.; Bruce, P. G. Journal of the American Chemical Society 2012, 134, 7952. (47) Chen, Y-M., Zhu, Y. ChemElectroChem 2015, 2, 208 (48) Walker, W.; Giordani, V.; Uddin, J.; Bryantsev, V. S.; Chase, G. V.; Addison, D. Journal of the American Chemical Society 2013, 135, 2076. (49) Abraham, K. M.; Jiang, Z. J. Journal of The Electrochemical Society 1996, 143, 1. (50) Bonnet-Mercier, N.; Wong, R. A.; Thomas, M. L.; Dutta, A.; Yamanaka, K.; Yogi, C.; Ohta, T.; Byon, H. R. Scientific Reports 2014, 4, 7127. (51) Jung, K. N.; Lee, J. I.; Jung, J. H.; Shin, K. H.; Lee, J. W. Chemical communications (Cambridge, England) 2014, 50, 5458.

68

(52) Luo, W.-B.; Chou, S.-L.; Wang, J.-Z.; Kang, Y.-M.; Zhai, Y.-C.; Liu, H.-K. Chemical communications 2015, 51, 8269. (53) Jadhav, H. S.; Kalubarme, R. S.; Jang, S. Y.; Jung, K. N.; Shin, K. H.; Park, C. J. Dalton transactions 2014, 43, 11723. (54) Suo, L.; Hu, Y. S.; Li, H.; Armand, M.; Chen, L. Nature Communications 2013, 4, 1481. (55) Oh, S. H.; Black, R.; Pomerantseva, E.; Lee, J-H.; Nazar, L.F. Nature Chemistry 2012, 4, 1004. (56) Younesi, R.; Hahlin, M.; Björefors, F.; Johansson, P.; Edström, K. Chemistry of Materials 2013, 25, 77. (57) Kang, J.; Li, O. L.; Saito, N. Journal of Power Sources 2014, 261, 156. (58) Guo, K.; Li, Y.; Yang, J.; Zou, Z.; Xue, X.; Li, X.; Yang, H. Journal of Materials Chemistry A 2014, 2, 1509. (59) Zhang, J.; Zhao, Y.; Zhao, X.; Liu, Z.; Chen, W. Scientific Reports 2014, 4, 6005. (60) Shao, Y.; Ding, F.; Xiao, J.; Zhang, J.; Xu, W.; Park, S.; Zhang, J.-G.; Wang, Y.; Liu, J. Advanced Functional Materials 2013, 23, 987. (61) Trogadas, P.; Fuller, T. F.; Strasser, P. Carbon 2014, 75, 5. (62) Thotiyl, M. M. O.; Freunberger, S. A.; Peng, Z.; Bruce, P. G. Journal of the American Chemical Society 2013, 135, 494. (63) Fuentes, R. E.; Colón-Mercado, H. R.; Fox, E. B. Journal of Power Sources 2014, 255, 219. (64) Kim, S. Y.; Lee, H. T.; Kim, K. B. Physical Chemistry Chemical Physics 2013, 15, 20262. (65) Wang, S.; Zhao, X.; Cochell, T.; Manthiram, A. The Journal of Physical Chemistry Letters 2012, 3, 2164. (66) Lu, Y.; Wen, Z.; Jin, J.; Cui, Y.; Wu, M.; Sun, S. Journal of Solid State Electrochemistry 2012, 16, 1863. (67) Zhang, K.; Zhang, L.; Chen, X.; He, X.; Wang, X.; Dong, S.; Gu, L.; Liu, Z.; Huang, C.; Cui, G. ACS applied materials & interfaces 2013, 5, 3677. (68) McCloskey, B. D.; Scheffler, R.; Speidel, A.; Girishkumar, G.; Luntz, A. C. The Journal of Physical Chemistry C 2012, 116, 23897. (69) Gallant, B. M.; Mitchell, R. R.; Kwabi, D. G.; Zhou, J.; Zuin, L.; Thompson, C. V.; Shao-Horn, Y. The Journal of Physical Chemistry C 2012, 116, 20800. (70) Grande, L.; Paillard, E.; Hassoun, J.; Park, J. B.; Lee, Y. J.; Sun, Y. K.; Passerini, S.; Scrosati, B. Advanced materials 2015, 27, 784. (71) Black, R.; Oh, S. H.; Lee, J.-H.; Yim, T.; Adams, B.; Nazar, L. F. Journal of the American Chemical Society 2012, 134, 2902. (72) Nasybulin, E.; Xu, W.; Engelhard, M. H.; Nie, Z.; Li, X. S.; Zhang, J.-G. Journal of Power Sources 2013, 243, 899. (73) Riaz, A.; Jung, K. N.; Chang, W.; Shin, K. H.; Lee, J. W. ACS applied materials & interfaces 2014, 6, 17815. (74) Lee, J-S.; Tai Kim, S.; Cao, R.; Choi, N-S.; Liu, M.; Lee, K. T.; Cho, J. Advanced Energy Materials 2011, 1, 34. (75) Shao, Y.; Park, S.; Xiao, J.; Zhang, J-G.; Wang, Y.; Liu, J. ACS Catalysis 2012, 2, 844. (76) Barile, C. J.; Gewirth, A. A. Journal of the Electrochemical Society 2013, 160, A549. (77) Wang, J.; Li, Y.; Sun, X. Nano Energy 2013, 2, 443. (78) Lu, Y-C.; Gasteiger, H. A.; Parent, M. C.; Chiloyan, V.; Yang, S-H. Electrochemical and Solid-State Letters 2010, 13, A69.

69

(79) Guo, K.; Li, Y.; Yang, J.; Zou, Z.; Xue, X.; Li, X.; Yang, H. Journal of Materials Chemistry A 2014, 2, 1509. (80) Jung, K-N.; Riaz, A.; Lee, S-B.; Lim, T-H.; Park, S-J.; Song, R-H.; Yoon, S.; Shin, K-H.; Lee, J-W. Journal of Power Sources 2013, 244, 328. (81) Thapa, A. K.; Ishihara, T. Journal of Power Sources 2011, 196, 7016. (82) Su, D.; Kim, H.-S.; Kim, W.-S.; Wang, G. Journal of Power Sources 2013, 244, 488. (83) Lee, M.; Hwang, Y.; Yun, K.-H.; Chung, Y.-C. Journal of Power Sources 2015, 288, 296. (84) Qin, Y.; Lu, J.; Du, P.; Chen, Z.; Ren, Y.; Wu, T.; Miller, J. T.; Wen, J.; Miller, D. J.; Zhang, Z.; Amine, K. Energy & Environmental Science 2013, 6, 519. (85) Debart, A.; Paterson, A. J.; Bao, J.; Bruce, P. G. Angewandte Chemie International Edition in English 2008, 47, 4521. (86) Zhao, G.; Lv, J.; Xu, Z.; Zhang, L.; Sun, K. Journal of Power Sources 2014, 248, 1270. (87) Yang, Y.; Shi, M.; Zhou, Q-F.; Li, Y-S.; Fu, Z-W. Electrochemistry Communications 2012, 20, 11. (88) Dong, S.; Chen, X.; Wang, S.; Gu, L.; Zhang, L.; Wang, X.; Zhou, X.; Liu, Z.; Han, P.; Duan, Y.; Xu, H.; Yao, J.; Zhang, C.; Zhang, K.; Cui, G.; Chen, L. ChemSusChem 2012, 5, 1712. (89) Wu, J.; Park, H. W.; Yu, A.; Higgins, D.; Chen, Z. The Journal of Physical Chemistry C 2012, 116, 9427. (90) Zhou, C.; Wang, H.; Peng, F.; Liang, J.; Yu, H.; Yang, J. Langmuir 2009, 25, 7711. (91) Xu, J-J.; Wang, Z-L.; Xu, D.; Zhang, L-L.; Zhang, X-B. Nature Communications 2013, 4, 2438. (92) Zhang, G. Q.; Zheng, J. P.; Liang, R.; Zhang, C.; Wang, B.; Au, M.; Hendrickson, M.; Plichta, E. J. Journal of The Electrochemical Society 2011, 158, A822. (93) Cao, Y.; Wei, Z.; He, J.; Zang, J.; Zhang, Q.; Zheng, M.; Dong, Q. Energy & Environmental Science 2012, 5, 9765. (94) Bergner, B. J.; Schürmann, A.; Peppler, K.; Garsuch, A.; Janek, J. Journal of the American Chemical Society 2014, 136, 15054. (95) Sun, D.; Shen, Y.; Zhang, W.; Yu, L.; Yi, Z.; Yin, W.; Wang, D.; Huang, Y.; Wang, J.; Wang, D.; Goodenough, J. B. Journal of the American Chemical Society 2014, 136, 8941. (96) Weng, W.; Barile, C. J.; Du, P.; Abouimrane, A.; Assary, R. S.; Gewirth, A. A.; Curtiss, L. A.; Amine, K. Electrochimica Acta 2014, 119, 138. (97) Guo, Z.; Zhou, D.; Liu, H.; Dong, X.; Yuan, S.; Yu, A.; Wang, Y.; Xia, Y. Journal of Power Sources 2015, 276, 181. (98) Lin, X.; Lu, X.; Huang, T.; Liu, Z.; Yu, A. Journal of Power Sources 2013, 242, 855. (99) Zhang, W.; Zhu, J.; Ang, H.; Zeng, Y.; Xiao, N.; Gao, Y.; Liu, W.; Hng, H. H.; Yan, Q. Nanoscale 2013, 5, 9651. (100) Wang, Z-L.; Xu, D.; Xu, J-J.; Zhang, L-L.; Zhang, X-B. Advanced Functional Materials 2012, 22, 3699. (101) Cetinkaya, T.; Ozcan, S.; Uysal, M.; Guler, M. O.; Akbulut, H. Journal of Power Sources 2014, 267, 140. (102) Jian, Z.; Chen, Y.; Li, F.; Zhang, T.; Liu, C.; Zhou, H. Journal of Power Sources 2014, 251, 466. (103) Zhang, T.; Zhou, H. Nature communications 2013, 4, 1817.

70

(104) Shui, J. L.; Okasinski, J. S.; Kenesei, P.; Dobbs, H. A.; Zhao, D.; Almer, J. D.; Liu, D. J. Nature Communications 2013, 4, 2255. (105) Hassoun, J.; Jung, H-G.; Lee, D-J.; Park, J-B.; Amine, K.; Sun, Y-K.; Scrosati, B. Nano Letters 2012, 12, 5775. (106) Freunberger, S. A.; Chen, Y.; Peng, Z.; Griffin, J. M.; Hardwick, L. J.; Barde, F.; Novak, P.; Bruce, P. G. Journal of the American Chemical Society 2011, 133, 8040. (107) McCloskey, B. D.; Bethune, D. S.; Shelby, R. M.; Girishkumar, G.; Luntz, A. C. The Journal of Physical Chemistry Letters 2011, 2, 1161. (108) Shui, J. L.; Okasinski, J. S.; Zhao, D.; Almer, J. D.; Liu, D. J. ChemSusChem 2012, 5, 2421. (109) Jung, H-G.; Park, J-B.; Sun, Y-K.; Scrosati, B. Nature Chemistry 2012, 4, 579. (110) Harding, J. R.; Lu, Y. C.; Tsukada, Y.; Yang, S-H. Physical Chemistry Chemical Physics 2012, 14, 10540. (111) Lim, H.; Yilmaz, E.; Byon, H. R. The Journal of Physical Chemistry Letters 2012, 3, 3210. (112) Liu, J.; Roberts, M.; Younesi, R.; Dahbi, M.; Edström, K.; Gustafsson, T.; Zhu, J. The Journal of Physical Chemistry Letters 2013, 4, 4045. (113) Luo, J.; Zhu, H. T.; Fan, H. M.; Liang, J. K.; Shi, H. L.; Rao, G. H.; Li, J. B.; Du, Z. M.; Shen, Z. X. The Journal of Physical Chemistry C 2008, 112, 12594. (114) Liu, J.; Renault, S.; Brandell, D.; Gustafsson, T.; Edström, K., Zhu, J. ChemSusChem 2015 DOI: 10.1002/cssc.201500381. (115) Renault, S.; Gottis, S.; Barrès, A.-L.; Courty, M.; Chauvet, O.; Dolhem, F.; Poizot, P. Energy & Environmental Science 2013, 6, 2124. (116) Naseem, A.; Olliff, C. J.; Martini, L. G.; Lloyd, A. W. International Journal of Pharmaceutics 2004, 269, 443. (117) Wang, Z.; Tammela, P.; Zhang, P.; Huo, J.; Ericson, F.; Stromme, M.; Nyholm, L. Nanoscale 2014, 6, 13068. (118) Wang, Z.; Tammela, P.; Zhang, P.; Strømme, M.; Nyholm, L. Journal of Materials Chemistry A 2014, 2, 16761. (119) Nyholm, L.; Nyström, G.; Mihranyan, A.; Strømme, M. Advanced Materials 2011, 23, 3751. (120) Lu, Y-C.; Gasteiger, H. A.; Yang, S-H. Journal of the American Chemical Society 2011, 133, 19048. (121) McCloskey, B. D.; Scheffler, R.; Speidel, A.; Bethune, D. S.; Shelby, R. M.; Luntz, A. C. Journal of the American Chemical Society 2011, 133, 18038. (122) Yao, K. P. C.; Lu, Y-C.; Amanchukwu, C. V.; Kwabi, D. G.; Risch, M.; Zhou, J.; Grimaud, A.; Hammond, P. T.; Barde, F.; Yang, S-H. Physical Chemistry Chemical Physics 2014, 16, 2297. (123) Meini, S.; Tsiouvaras, N.; Schwenke, K. U.; Piana, M.; Beyer, H.; Lange, L.; Gasteiger, H. A. Physical Chemistry Chemical Physics 2013, 15, 11478. (124) Oh, S. H.; Pomerantseva, E.; Lee, J-H.; Nazar, L. F. Nature Chemistry 2012, 4, 1004. (125) Waseda, Y., E. M., Shinoda, K. X-Ray Diffraction Crystallography, SpringerVerlag Berlin Heidelberg 2011. (126) Graf-Zeiler, B.; Fink, R. H.; Tzvetkov, G. ChemPhysChem 2011, 12, 3503. (127) Knies, D. L.; Violante, V.; Grabowski, K. S.; Hu, J. Z.; Dominguez, D. D.; He, J. H.; Qadri, S. B.; Hubler, G. K. Journal of Applied Physics 2012, 112, 083510. (128) Fultz, B.; Howe, J. Transmission Electron Microscopy and Diffractometry of Materials, Springer 2013.

71

(129) Cerenius, Y.; Ståhl, K.; Svensson, L. A.; Ursby, T.; Oskarsson, Å.; Albertsson, J.; Liljas, A. Journal of Synchrotron Radiation 2000, 7, 203. (130) Hammersley, A. P.; Svensson, S. O.; Hanfland, M.; Fitch, A. N.; Hausermann, D. High Pressure Research 1996, 14, 235. (131) Rodriguez-Carvajal, J. LLB JRC 1997. (132) McDevitt, J. T. Journal of Chemical Education 1984, 61, 217. (133) Fujishima, A.; Honda, K. Nature 1972, 238, 37. (134) Grätzel, M. Nature 2001, 414, 338. (135) Pearson, A.; Jani, H.; Kalantar-zadeh, K.; Bhargava, S. K.; Bansal, V. Langmuir 2011, 27, 6661. (136) Alpatova, N. M.; Krishtalik, L. I.; Pleskov, Y. V. In Organolithium Compounds/Solvated Electrons; Springer Berlin Heidelberg.; 1987, 138, 149. (137) Veith, G. M.; Dudney, N. J.; Howe, J.; Nanda, J. The Journal of Physical Chemistry C 2011, 115, 14325. (138) Ryan, K. R.; Trahey, L.; Ingram, B. J.; Burrell, A. K. The Journal of Physical Chemistry C 2012, 116, 19724. (139) Qiao, R.; Chuang, Y-D.; Yan, S.; Yang, W. PLoS One 2012, 7, e49182. (140) Ma, Q.; Divan, R.; Mancini, D. C.; Rosenberg, R. A.; Quintana, J. P.; Keane, D. T. Applied Physics Letters 2006, 89, 083114. (141) Ong, S. P.; Mo, Y.; Ceder, G. Physical Review B 2012, 85, 081105. (142) Hummelshoj, J. S.; Blomqvist, J.; Datta, S.; Vegge, T.; Rossmeisl, J.; Thygesen, K. S.; Luntz, A. C.; Jacobsen, K. W.; Norskov, J. K. The Journal of Chemical Physics 2010, 132, 071101. (143) Zhong, L.; Mitchell, R. R.; Liu, Y.; Gallant, B. M.; Thompson, C. V.; Huang, J. Y.; Mao, S. X.; Shao-Horn, Y. Nano Letters 2013, 13, 2209. (144) Xu, J-J.; Wang, Z-L.; Xu, D.; Meng, F-Z.; Zhang, X-B. Energy & Environmental Science 2014, 7, 2213. (145) Lu, Y.-C.; Xu, Z.; Gasteiger, H. A.; Chen, S.; Hamad-Schifferli, K.; ShaoHorn, Y. Journal of the American Chemical Society 2010, 132, 12170. (146) Xiao, F.; Li, Y.; Zan, X.; Liao, K.; Xu, R.; Duan, H. Advanced Functional Materials 2012, 22, 2487. (147) Kraytsberg, A.; Ein-Eli, Y. Nano Energy 2013, 2, 468. (148) Trahey, L.; Karan, N. K.; Chan, M. K. Y.; Lu, J.; Ren, Y.; Greeley, J.; Balasubramanian, M.; Burrell, A. K.; Curtiss, L. A.; Thackeray, M. M. Advanced Energy Materials 2013, 3, 75. (149) Truong, T. T.; Liu, Y.; Ren, Y.; Trahey, L.; Sun, Y. ACS Nano 2012, 6, 8067. (150) Jung, K-N.; Riaz, A.; Lee, S-B.; Lim, T-H.; Park, S-J.; Song, R-H.; Yoon, S.; Shin, K-H.; Lee, J-W. Journal of Power Sources 2013, 244, 328. (151) Younesi, R.; Hahlin, M.; Björefors, F.; Johansson, P.; Edström, K. Chemistry of Materials 2012, 25, 77. (152) Xu, J. B.; Zhao, T. S.; Liang, Z. X. The Journal of Physical Chemistry C 2008, 112, 17362. (153) Liu, J.; Zhao, Y.; Shi, L.; Yuan, S.; Fang, J.; Wang, Z.; Zhang, M. ACS Applied Materials & Interfaces 2011, 3, 1261. (154) Ye, Q.; Zhao, J.; Huo, F.; Wang, D.; Cheng, S.; Kang, T.; Dai, H. Microporous and Mesoporous Materials 2013, 172, 20. (155) Rossmeisl, J.; Logadottir, A.; Nørskov, J. K. Chemical Physics 2005, 319, 178. (156) Mo, Y.; Ong, S. P.; Ceder, G. Physical Review B 2011, 84. (157) Li, H.; Wang, W.-l.; Pan, F.; Xin, X.; Chang, Q.; Liu, X. Materials Science and Engineering: B 2011, 176, 1054.

72

(158) Wen, M.; Sun, Y.; Li, X.; Wu, Q.; Wu, Q.; Wang, C. Journal of Power Sources 2013, 243, 299. (159) Kolbe, H. Justus Liebigs Ann. Chem. 1849, 69, 257. (160) Li, L.; Manthiram, A. Advanced Energy Materials 2014, 4, 130175. (161) Wang, Z.; Xiong, X.; Qie, L.; Huang, Y. Electrochimica Acta 2013, 106, 320. (162) Li, H.; Kang, W.; Wang, L.; Yue, Q.; Xu, S.; Wang, H.; Liu, J. Carbon 2013, 54, 249. (163) Mi, R.; Liu, H.; Wang, H.; Wong, K.-W.; Mei, J.; Chen, Y.; Lau, W.-M.; Yan, H. Carbon 2014, 67, 744. (164) Shin, W. H.; Jeong, H. M.; Kim, B. G.; Kang, J. K.; Choi, J. W. Nano Letters 2012, 12, 2283. (165) Wang, Z.; Qie, L.; Yuan, L.; Zhang, W.; Hu, X.; Huang, Y. Carbon 2013, 55, 328. (166) Chen, Z.; Yu, A.; Higgins, D.; Li, H.; Wang, H.; Chen, Z. Nano Letters 2012, 12, 1946. (167) Fu, T.; Liu, R.; Lv, J.; Li, Z. Fuel Processing Technology 2014, 122, 49. (168) Tammela, P.; Wang, Z.; Frykstrand, S.; Zhang, P.; Sintorn, I-M.; Nyholm, L.; Strømme, M. RSC Advances 2015, 5, 16405. (169) Li, F.; Chen, Y.; Tang, D.-M.; Jian, Z.; Liu, C.; Golberg, D.; Yamada, A.; Zhou, H. Energy & Environmental Science 2014, 7, 1648. (170) Thotiyl, M. M. O.; Freunberger, S. A.; Peng, Z.; Chen, Y.; Liu, Z.; Bruce, P. G. Nature Materials 2013, 12, 1051. (171) Shen, Y.; Sun, D.; Yu, L.; Zhang, W.; Shang, Y.; Tang, H.; Wu, J.; Cao, A.; Huang, Y. Carbon 2013, 62, 288. (172) Sharon, D.; Noked, M.; Garsuch, A.; Frimer, A. A.; Aurbach, D. The Journal of Physical Chemistry Letters 2013, 4, 3115.

73

Acta Universitatis Upsaliensis Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 1271 Editor: The Dean of the Faculty of Science and Technology A doctoral dissertation from the Faculty of Science and Technology, Uppsala University, is usually a summary of a number of papers. A few copies of the complete dissertation are kept at major Swedish research libraries, while the summary alone is distributed internationally through the series Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology. (Prior to January, 2005, the series was published under the title “Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology”.)

Distribution: publications.uu.se urn:nbn:se:uu:diva-259589

ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2015