Exploration of Ti-based Anodes for Rechargeable Alkali-ion Batteries And Hybrid-ion Capacitors

Abstract

From 600 B.C. till 1700 A.D., electricity remained the least understood form of energy and was considered "Magical." In the 18th century A.D., the electricity that mysteriously struck Franklin as he arduously flew a kite directly into a thunderstorm was also thought to be "indwelling" inside all living beings by Galvani when he saw a dead frog's leg jolt upon scalpel touch. In this period, for the first time, electricity was bottled in its transient static (electronic) form inside a Leyden jar and in permanent (ionic) form between a "pile" of a brine-soaked copper-zinc discs, designed by a pragmatic Volta. Improved "pile"s enabled major fundamental findings ranging from water splitting to numerous element discoveries in the 20th century before Plante introduced the lead-acid rechargeable batteries in 1859. Surprisingly, these rechargeable batteries were used in clean, silent, and swift starting electric cars as early as 1900 by Edison, the inventor of the electric bulb. Gasoline took over as the energy source with the advent of post-world war V8 internal combustion engines by Rolls Royce and the electric starters by Kettering. Recently, electric cars have returned to limelight highlighting the need for economic and high-performance batteries to topple the incessant usage of gasoline-powered automobiles. Weber and Kummer's sodium-sulfur battery (1967) followed by Arab's oil embargo (1973) slingshot original research activity on rechargeable non-aqueous batteries by Huggins, Schollhörn, Murphy, Hagenmuller, and other pioneers. The key idea was to dump fossils and sustainably collect and (electrochemically) store energy in fuel cells, batteries, or capacitors from renewable sources like solar, wind, ocean, and others that are scattered in space and unpredictable/intermittent in time. This sense of urgency bore pivotal output by Whittingham and Armand on intercalation, Goodenough on LiCoO2 cathode, Yazami and Yoshino on carbon anodes. These enabled SONY's first commercial Lithium-Ion Battery (LIB) in 1991, which powered the portable industry's take-off. LIBs, even to date, remain the undisputed power source for high energy portable electronics, safe automotive, and cheap gridlevel applications that have penetrated our lives. SONY's LIB gift kickstarted arduous research and industry activity in energy storage. While the cathode and the electrolyte that underlie the maximum cost in a battery are feverishly researched, it was the anode that blocked the gateway to the first commercial LIB. Carbon as amorphous coal may be a polluting energy source, but as layered graphite is the cheapest and most widely used anode material in the LIBs. The energy of lithium does not significantly drop (unstable) inside the layered graphite anode compared to a vast energy drop (stable) in the cathode. Although this energy difference of lithium results in the highest energy density batteries when graphite is used, it is far from being the holy grail of anode materials. For example, while graphite is the preferred anode in Li- and K-ion batteries, it is inactive in sodium-ion batteries. Further, its low voltage operation risks fire hazards during accidental overcharge or high current conditions. It is a severe setback in the present scenario where a global uprising towards using earth-abundant materials has catapult research and industry activity in the Sodium-Ion Battery (NIB) space. Following this, spearheaded by the 1.55 V spinel Li4Ti5O12 (LTO), Ti-based anodes are safe and can be used in fast-charge applications. Along with the LiFePO4 (LFP) safe cathode, the LTO anode results in a 2 V battery that lasted for 30000 cycles at 15C fast charging conditions while retaining over 90% capacity. Building on the titanium chemistry, three stories of Ti-based anodes for rechargeable batteries and Hybrid-Ion Capacitors (HIC) are hereby discussed. They are inspired by spinel Li4Ti5O12, mineralogy, and potentially new directions in Ti-chemistry to design new anodes.

Chapter 1 reinforces the need for Ti-based anodes. Under the backdrop of the terawatt challenge, different primary sources of energy (e.g., fossil, oil, and coal) are discussed, emphasizing the need for renewable energy generation and storage in batteries in the (electro)chemical form. The evolution of the batteries and their current demand in automobiles is portrayed. Different anode operation mechanisms and their hybrid combination are outlined. Pressing topics relating to anodes like metal anode, risks with the use of graphite, the reason for its inactivity for NIB, formation of the Solid Electrolyte Interphase (SEI), and the anode roadblock to commercialization of LIB are discussed. Finally, some critical works relating to HIC and existing Ti-based anodes operating on different charge storage mechanisms are portrayed, setting the stage for the thesis work.

Chapter 2 focuses on the experimental procedures used for carrying out the thesis research. The preparative methods discussed include ball milling, dry, wet, and ion exchange routes. Different physicochemical, spectroscopy, microscopy, and diffraction techniques have been described. The preparation of electrode coatings, cell assembly, and different electrochemical testing techniques (galvanostatic, potentiodynamic, impedance etc.) are outlined. A description of high-temperature bulk ionic conductivity measurements, calculation of lithium migration pathway, and their energy barrier by Bond Valence Site Energy (BVSE) method is provided.

Chapter 3 is inspired from spinel LTO and revolves around a closely related group of titanates, MLi2Ti6O14 (M = 2Na, Sr, Ba, Pb) (MLTO), first reported by the Amine and Shu groups. The relation of MLTO to popular battery materials having a different dimensionality of (non)mobile structural units is discussed. BVSE calculations show how M influences the Li+ migration pathways and their energy barriers. Bulk high-temperature conductivity measurements indicate the transition points of the dimensionality of lithium flow. The electrochemical performance of combustion synthesized MLTO family of compounds is outlined. MLTO inserts lithium involving a Ti(IV)/Ti(III) redox change in a flat profile at 1.2-1.4 V that is slightly lower than Li4Ti5O12. Divalent MLTOs deliver up to four electron capacities (100 – 160 mAh/g) with an additional low voltage slope region, while NaLTO shows less than 80 mAh/g (2 electrons). The lower capacity of NaLTO is due to filled interlayer space, while it is half-filled for the divalent MLTOs. Hence, the MLTO structure's capacity becomes space-constrained, and the theoretical capacity (220 – 283 mAh/g) for all six-lithium transfer is out of reach. Following the battery study, MLTOs were tested versus Activated Carbon (A.C.) for the first time in asymmetric Lithium-Ion Capacitors (LIC) (100 mAh/g, 1-3 V). The role of the M-ion is emphasized while understanding the differences in electrochemical performance of MLTO. MLTOs were first loaded with lithium (discharged) before assembling them in MLTOs/AC LIC hybrids. A.C. cathode rapidly adsorbs PF6- ions with a capacitor like slopy profile (40 mAh/g), while MLTOs (de)intercalate Li+ (100 mAh/g, 1-2 V) at Ti redox potential. This chapter confirms MLi2Ti6O14 (M = 2Na, Sr, Ba, Pb) titanates as versatile electrodes for both secondary batteries and hybrid-ion capacitors.

Chapter 4 attempts to unravel new Ti-based anode materials using mineralogy, beyond conventional exploration using lab-based synthetic or computational databases. Three Ti-based minerals, (i) Narsarsukite Na2TiOSi4O10, (ii) Freudenbergite NaMTi3O8 (M = Al, Fe), and (iii) Priderite Na1.7Cr1.7Ti6.3O16, could insert lithium and sodium thereby working as anodes for metal-ion batteries and capacitors. Solid-state synthesized Narsarsukite (de)inserts one lithium (at C/2) using a Ti(IV)/Ti(III) redox giving near theoretical capacity (70 mAh/g) with 80% capacity retention after 200 cycles. Due to its unique tetragonal I4/m structure consisting of open tunnels of rigid silicate units, Narsarsukite could deliver 50% capacity at a high 50C rate. The Freudenbergites have a theoretical capacity of 220-250 mAh/g. Pure solution combustion synthesized phases were tested for lithium and sodium insertion (0.01 to 2.5 V). While AlFreudenbergite showed low capacities for the first-time lithium insertion, Fe-Freudenbergite delivered 200 mAh/g and 70 mAh/g for the unannealed and annealed phases, respectively. Unannealed Fe- Freudenbergite showed moderate Na de-insertion capacity (140 mAh/g) while annealed Fe-Freudenbergite yielded high capacity (180 mAh/g till 100 cycles). Mössbauer spectroscopy was performed at different states of charge, hoping to see Fe(0) in order to reinforce a conversion reaction and refute the earlier proposed sodium intercalation reaction. This claim is based on initial cues from work by the Greenblatt, Cava and Murphy group at the Bell Labs in the 1980s. They found a two-lithium insertion limit for alkali-free FeV3O8, which is a crystallographic rigid shear structure exactly like NaFeTi3O8. Obviously, NaFeTi3O8 already hosting structural sodium cannot topotactically insert three more sodiums. Further, the Fe-freudenbergite was tested vs. A.C. in sodium-ion capacitor (80 mAh/g at 50 mA/g, 0-3 V).[NaxCrxTi(8-x)O16 (x=1.7)] (NCTOq) Priderite mineral was prepared using solid-state synthesis by quenching at 1350 oC. It crystallizes in a hollandite-type (I4/m) structure with three types of sodium present in the tunnels, one of which uniquely and favorably sits at the closest spaced oxygen site bottlenecks. These structure types are usually stabilized by larger cations and do not occur for sodium compounds. Furnace-cooled samples (1350 oC) result in a mixture of Rutile, Freudenbergite, and Priderite phases, while low temperature (900 oC) annealing forms pure Freudenbergite phase. Electrochemical performance of ball-milled NCTOq samples was evaluated. Capacity values of 130 mAh/g and 60 mAh/g were observed vs. lithium and sodium (0.01 V to 2.5 V, 10 mA/g) after correcting for charge storage in binder and carbon additives. XPS analysis confirmed the presence of Ti(IV)/Ti(III) redox activity. The NCTOq samples were pre-lithiated, balanced, and were used in Li-ion and Na-ion capacitors with Activated Carbon. Based on the active material weight, ~100 Wh/kg energy density was observed in halfcell and hybrid-ion capacitor configurations. On a broader note, this chapter highlights possible exploitation of vast mineralogical database to design potential Ti-based anode materials.

Chapter 5 explores two new directions in Ti-based anodes bearing tremendous promise; one opens research gateways to innumerable structures containing an alloying element, another unveils a new lithium insertion mechanism for battery materials. Venturing beyond the unstable lead halide perovskites famous in solar cells, stable and ubiquitous lead titanate and lead zirconate perovskites were tested as battery anodes for the first time. Just four years after SONY's introduction of Li-ion battery (circa 1991), in a 1995 patent, Fuji Photo Film Co. announced a Sn-based Amorphous Tin Composite Oxide (ATCO) glass delivering four times volumetric and two times gravimetric anode energy density than graphite. Enlightened by this,xii there was a deep dive in research activity over the next decade into all materials having the Sn alloying center, including some having the perovskite-type ABO3 structure. Focusing on Pb alloying center in perovskite frameworks, PbTiO3 and PbZrO3 are presented as potential anode materials for the first time as a glimpse into many potential ABO3 battery materials. Following the structural breakdown of PbTiO3 in the first irreversible conversion cycle, Pb alloying and TiO2 insertion led to reversible capacities up to 400 mAh/g for Li/Na (nearly 4e-/mol) and 180 mAh/g for K (nearly 2e-/mol) in the first cycle. De-alloyed Pb particles were observed from TEM images after de-lithiation at the end of the charge. Perovskite family (and their derivatives) can be put on anvil to develop high-capacity battery anode materials. In the second new direction, following the lowest 0.3 V Na2Ti3O7 NIB anode, we explore PbTi3O7 titanate anode for the first time. Capacities of 300-400 mAh/g (in LIB/NIB) were observed via different electrochemical mechanisms. Large Na+ irreversibly converts PbTi3O7 to Pb and TiO2. These reversible alloys (NaxPb) convert (Pb/PbO2) and intercalate (TiO2/NaxTiO2) like PX-PbTiO3. On the contrary, small Li+ intercalates, displacing Pb as metal, in PbTi3O7 at low voltages like previously studied reversible Cu/Ag extrusion reactions studied in the 2000s by Thackeray and Tarascon. Pb metal stores more lithium forming various alloys. All reactions are reversible leading to parent PbTi3O7 structure at the end of the charge. It shows the first example of reversible conversion-alloying-displacement reaction in Ti-based anode materials. Chapter 6 provides a summary of the entire thesis work outlining possible future directions to expand the rich playground of titanium-based battery anode materials.