Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Physics, 1981. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND SCIENCE. / Vita. / Includes bibliographical references. / by Paolo Rava. / Ph.D.

Batteries are complex, multidisciplinary, electrochemical energy storage systems that are crucial for powering our society. During operation, all battery technologies suffer from voltage losses due to energetic penalties associated with the electrochemical processes (i.e., ohmic resistance, kinetic barriers, and mass transport limitations). A majority of the voltage losses can be attributed to processes occurring on/in the battery electrodes, which are responsible for facilitating the electrochemical reactions. A major challenge in the battery field is developing strategies to mitigate these losses. To accomplish this, researchers must i) identify the processes limiting the performance of the electrode, ii) characterize the main, performance-limiting processes to understand the underlying mechanisms responsible for the poor performance, and iii) mitigate the voltage losses by developing strategies which target these underlying mechanisms. In this thesis, three studies are presented which highlight the role of electrochemical engineers in alleviating the performance limiting processes in battery electrodes. Each study is focused on a different step of the research approach (i.e., identification, characterization, and mitigation) and analyzes an electrode from a different battery system. The first part of the thesis is focused on identifying the processes limiting the capacity in nanocomposite lithium-magnetite electrodes. To accomplish this, the mass transport processes and phase changes occurring within magnetite electrodes during discharge and voltage recovery are investigated using a combined experimental and modeling approach. First, voltage recovery data are analyzed through a comparison of the mass transport time-constants associated with different length-scales in the electrode. The long voltage recovery times are hypothesized to result from the relaxation of concentration profiles on the mesoscale, which consists of the agglomerate and crystallite length-scales. The hypothesis was tested through the development of a multi-scale mathematical model. Using the model, experimental discharge and voltage recovery data are compared to three sets of simulations, which incorporate crystal-only, agglomerate-only, or multi-scale transport effects. The results of the study indicate that, depending on the crystal size, the low utilization of the active material (i.e., low capacity) is caused by transport limitations on the agglomerate and/or crystal length-scales. For electrodes composed of small crystals (6 and 8 nm diameters), it is concluded that the transport limitations in the agglomerate are primarily responsible for the long voltage recovery times and low utilization of the active material. In the electrodes composed of large crystals (32 nm diameter), the slow voltage recovery is attributed to transport limitations on both the agglomerate and crystal length-scales. Next, the multi-scale model is further expanded to study the phase changes occurring in magnetite during lithiation and voltage recovery experiments. Phase changes are described using kinetic expressions based on the Avrami theory for nucleation and growth. Simulated results indicate that the slow, linear voltage change observed at long times during the voltage recovery experiments can be attributed to a slow phase change from α¬-LixFe3O4 to β¬-Li4Fe3O4. In addition, simulations for the lithiation of 6 and 32 nm Fe3O4 suggest the rate of conversion from α¬-LixFe3O4 to γ-(4 Li2O + 3 Fe) decreases with decreasing crystal size. The next part of the thesis presents a study aimed at characterizing the formation of PbSO4 films on Pb in H2SO4, which has been previously identified as a performance-limiting process in lead-acid batteries. Transmission X-ray microscopy (TXM) is utilized to monitor, in real time, the initial formation, the resulting passivation, and the subsequent reduction of the PbSO4 film. It is concluded with support from quartz-crystal-microbalance experiments that the initial formation of PbSO4 crystals occurs as a result of acidic corrosion. Additionally, the film is shown to coalesce during the early stages of galvanostatic oxidation and to passivate as a result of morphological changes in the existing film. Finally, it is observed that the passivation process results in the formation of large PbSO4 crystals with low area-to-volume ratios, which are difficult to reduce under both galvanostatic and potentiostatic conditions. In a further extension of this study, TXM and scanning electron microscopy are combined to investigate the effects of sodium lignosulfonate on the PbSO4 formation and the initial growth of PbSO4 crystals. Sodium lignosulfonate is shown to retard, on average, the growth of the PbSO4 crystals, yielding a film with smaller crystals and higher crystal densities. In addition, an analysis of the growth rates of individual, large crystals showed an initial rapid growth which declined as the PbSO4 surface coverage increased. It was concluded that the increase in PbSO4 provides additional sites for precipitation and reduces the precipitation rate on the existing crystals. Finally, the potential-time transient at the beginning of oxidation is suggested to result from the relaxation of a supersaturated solution and the development of a PbSO4 film with increasing resistance. The final part of the thesis presents a study aimed at mitigating the ohmic losses during pulse-power discharge of a battery by the adding a second electrochemically active material to the electrode. Porous electrode theory is used to conduct case studies for when the addition of a second active material can improve the pulse-power performance. Case studies are conducted for the positive electrode of a sodium metal-halide battery and the graphite negative electrode of a lithium-ion battery. The replacement of a fraction of the nickel chloride capacity with iron chloride in a sodium metal-halide electrode and the replacement of a fraction of the graphite capacity with carbon black in a lithium-ion negative electrode were both predicted to increase the maximum pulse power by up to 40%. In general, whether or not a second electrochemically active material increases the pulse power depends on the relative importance of ohmic-to-charge transfer resistances within the porous structure, the capacity fraction of the second electrochemically active material, and the kinetic and thermodynamic parameters of the two active materials.

Batteries are ubiquitous in modern society, from the portable devices we use daily to the yet-to-be realized integration of batteries into the electrical grid and electrical vehicle markets. One of the primary roles of batteries to date has been to enable portability of devices, and as chemical energy storage becomes more affordable, batteries will play a larger role in how society cares for the environment by enabling technologies that are poised to decrease greenhouse gas emissions. Low cost and environmentally conscious materials are pivotal for the economic feasibility and widespread integration of batteries into new markets. Batteries operate far from equilibrium and may operate under extreme stress and varying loads, therefore, for a material to be successful in an operational battery it must meet multiple design criteria. Here, an in-depth analysis of magnetite, a low cost and abundant iron oxide studied for use as an electrode material in lithium-ion batteries, is presented. In the second Chapter, an in-depth analysis into how magnetite accepts lithium into the solid state at low depths of discharge is examined with density functional theory and a mechanistic understanding of a phase change from the parent spinel to a rocksalt-like material is presented. When magnetite is used as an electrode material in a lithium-ion battery, lithium must enter into and eject from the solid state of the host material, where the direction of lithium movement is a function of the current in the battery. In many electrode materials, magnetite included, large structural rearrangements can occur in the host material as lithium moves into and out of the lattice. These structural rearrangements can be irreversible and can contribute to overpotentials, decreasing efficiency and lifecycle for the battery. The structural rearrangements in bulk magnetite occurring due to lithium insertion are found to be driven primarily by Coulombic interactions. Additionally, the energetics and structural rearrangements for lithium insertion into defective magnetite and maghemite are examined, as these derivative structures commonly co-exist with magnetite, especially when the material is nanostructured. It is found that defective magnetite and maghemite accept lithium by a different mechanism, one that does not initially result in substantial structural rearrangement, as is the case in magnetite. In Chapter three, the effects of nanostructuring magnetite on the reversible potential are examined as a function of nanoparticle size. Due to solid-state mass-transport resistances, active electrode materials in batteries are commonly nanostructured. When a material is nanostructured, the bulk properties are often replaced due to interesting phenomena that can occur as a result of stark differences between the nanostructured material and the bulk counterpart. These differences are often attributed to surface area to volume ratios, the exaggerated role of surface energies, lattice defects, and the variation in electronic behavior, all properties which change between a bulk and nanostructured material. The reversible potential is found to be particle size dependent, and this dependence is explained, in part, by the cationic defective surfaces in the particles and the differences in surface area to volume ratio between varying particle sizes. Evidence for these defects is presented with materials characterization techniques such as XRD and EELS studies. Finally, the reversible potential at low lithiation states is predicted theoretically and found to match well to the experimentally measured potential. A study of the DFT predicted potentials and XRD characterization for multiple metastable pathways is examined in the fourth Chapter. Room temperature and long-time scale persistence of metastable phases is a pervasive phenomenon in nature. Magnetite is known to undergo both phase change and conversion reactions upon lithiation. Due to large mass transport and kinetic resistances, multiple phase changes are often observed in parallel during discharge, resulting in heterogenous phase formation in particles which can have large local lithium concentration variations. Phases which form during discharge can become kinetically trapped and the equilibrium state can therefore follow a metastable pathway. Theoretical potentials and XRD patterns are compared to the experimental patterns taken following 600 hours of relaxation following discharge at the slow rate of C/600. The evidence presented supports a metastable pathway occurring on the first voltage plateau. In the fifth Chapter, the methodologies for the density functional theory calculations are presented in full detail. This includes various studies on the more subtle electronic properties of magnetite and its lithiated derivates studied herein. These studies include examination of the charge and orbital ordering problem related to the Verwey transition in magnetite, the charge and magnetic order in the rocksalt-like lithiated magnetite, and a full theoretical description of the various phases in the Li-Fe-O ternary phase diagram that were calculated to make the relevant conclusions in Chapters 2-4. Finally, corrections to DFT predicted formation energy and volume are presented. The aim of this thesis is to use theoretical techniques to examine the lithiation of magnetite on the atomic scale and make meaningful connections to the experimentally observed electrochemical behavior of the material. To accomplish this, magnetite and the structural derivatives of magnetite that co-exist with the material under physically realistic conditions must be treated theoretically. In this thesis, ties between phenomena occurring on the atomic scale and the measurable properties of the macroscopic system, such as voltage, will be related. It will be illustrated that as a function of nanoparticle size, the magnetite system can vary in its atomic structure and the resultant electrochemistry and phase change characteristics are both affected. The findings indicate the relevance of the atomic properties and nanostructure for magnetite to the observed and measured electrochemical properties of the material.

Com o avanço da tecnologia aplicada em aparelhos celulares, são lançados no mercado modelos menores, mais leves e com maior rapidez em seu sistema operacional. Tudo isso atrai muito os consumidores, que por sua vez, trocam seus antigos aparelhos celulares por novos e modernos. Essas adesões e trocas freqüentes de aparelhos celulares geram um descarte significativo de todos os seus componentes e dentre eles a bateria. Assim, pelo fato do cobalto estar presente nesse tipo bateria e ser um metal com alto valor agregado, ele faz parte do estudo do presente trabalho. Este trabalho tem por objetivos estudar as etapas de condicionamento físico e de lixiviação como fases iniciais do processo de reciclagem de baterias de íons de lítio. Para a caracterização das baterias, as mesmas foram desmanteladas manualmente para a separação dos componentes. Os eletrodos foram caracterizados por espectrofotometria de absorção atômica, difração de raios-X e microscopia eletrônica de varredura com analise de microrregiões. Os ensaios de lixiviação foram feitos usando-se os seguintes parâmetros: pH entre 3 e 5, temperaturas de 25 e 50ºC, relação sólido/líquido de 1/5, tempos de 1 a 4h. Foram utilizados como meio lixiviante soluções de acido sulfúrico, clorídrico e nítrico. Peróxido de hidrogênio foi adicionado ao acido sulfúrico como agente oxidante. Os resultados alcançados mostraram que entre os moinhos de martelos, de facas e de bolas o que apresentou o melhor desempenho para a moagem de baterias de íons de lítio foi o moinho de facas. O bombardeamento com ultra-som faz com que haja a liberação do material ativo dos eletrodos que fica aderido aos suportes de cobre e alumínio mesmo apos moagem. A diminuição do pH de 5 para 3 e o aumento da temperatura de 25 para 50ºC causam o aumento da velocidade de lixiviação em meio sulfúrico do oxido de cobalto. A presença de agente oxidante na lixiviação acida faz com que diminua o tempo de lixiviação do oxido de cobalto. A lixiviação com acido nítrico e com acido sulfúrico com adições de peróxido de hidrogênio são os melhores meios de lixiviação quando comparados ao acido clorídrico e ao acido sulfúrico sem oxidante para as mesmas condições de pH e de temperatura. Nas etapas de extração liquido-liquido e reextração foram utilizados os seguintes parâmetros: relação orgânico-aquoso de 1/1, temperatura de 50ºC, pH 4 e tempo de 5 minutos na etapa de extração e como solução aquosa na fase da reextração o acido sulfúrico 2M. Nas duas etapas foi utilizado como extratante o Cyanex 272 diluído em querosene. Os resultados alcançados mostraram que nos primeiros dois contatos das fases orgânicas e aquosas já se obtém as melhores porcentagens de extração e reextração. A eficiência global dos quatro contatos na etapa de extração foi de 94% e na etapa de reextracao foi de 98%. / The fast changes in technologies applied to mobile phones causes an incredible appearance of new and even better models each day. As a consequence, each year increase the amounts of waste of electronic and electric equipments including batteries that should be disposed of. The goal of the present work is to study methods of physical conditioning and acid leaching of Li-ion batteries. Hand disassembling Li-ion batteries was performed to identify and characterize the components of scrap Li-ion batteries. Materials extracted form the electrodes were characterized using X-ray diffraction, atomic absorption spectrophotometry and scanning electron microscopy coupled with EDS micro-probe. Leaching tests were carried out using the following parameters: pH (3 - 5), temperature (25 and 50ºC), solid/liquid ratio equal to 1/5, leaching time (1 to 4h). Sulfuric acid, chloridric acid and nitric acid were tested as leaching media. Hydrogen peroxide was tested as an oxidizing agent during leaching tests using sulfuric acid. The results obtained showed that knives mill presented better results to grind the scraps in comparison to hammer and balls mills. Ultrasonic treatment was effective to release the active cell materials from copper and aluminum. Decreasing pH from 5 to 3, and increasing the temperature from 25 to 50ºC cause the increasing of the leaching rate of cobalt oxide. Oxidizing conditions also increase the rate of cobalt oxide leaching. Nitric acid and sulfuric acid plus hydrogen peroxide leaching results in better leaching rates compared to chloridric acid and sulfuric acid.

Orientador: Ana Beatriz Lopes de Sousa Jobbour / Banca: Lara Bartocci Liboni / Banca: Daniel Jugend / Resumo: Esta pesquisa teve como objetivo analisar como os stakeholders primários, de uma empresa focal em uma cadeia de suprimentos do setor de baterias automotivas, influenciam a adoção de práticas de green supply chain management (GSCM). Foi adotado o método de pesquisa de estudo de caso. A pesquisa contribuiu para identificar o papel dos stakeholders primários na adoção de práticas de GSCM de uma empresa focal de uma cadeia de suprimentos do setor de baterias automotivas. Foi possível identificar que os stakeholders clientes e governo exercem pressões ambientais sobre a empresa focal e que isso influencia a adoção de algumas práticas de GSCM. As pressões ambientais recebidas principalmente de seu stakeholder cliente, fazem com que a empresa focal gere um grupo de pressões, semelhantes às exercidas pelo seu cliente, sobre seu principal fornecedor. Essas pressões também fazem com que o fornecedor da empresa focal também adote práticas de GSCM. A propagação da pressão ambiental também identificada e explicada pelo isomorfismo do ambiente institucional, pelas características do setor de baterias automotivas e, também pelo fenômeno green bullwhip effect / Abstract: This research aimed to analyze how primary stakeholders of a focal company in a supply chain of the automative battery industry, influence the adoption of green supply chain management practices (GSCM). The method of case study research was adopted. The research helped identify the role of primary stakeholders in the adoption of a focal company's GSCM practices of a supply chain of the automotive battery industry. It was possible to identify that stakeholders and government customers exert environmental pressures on the focal company and that this influences the adoption of some GSCM practices. Environmental pressures have mainly received his client stakeholder, make the focal company manages a group of pressures, similar to those performed by your client on thier main supplier. This pressure also causes the focal company provider also adopt GSCM practices. The spread of environmental pressure was also identified and explained by the institutional environment isomorphism, the characteristics of automotive batteries industry and also the green bullwhip effect phenomenon / Mestre

The main aim of this thesis is to reveal the chemical structures of the solid-liquid interphase in lithium ion batteries by NMR spectroscopy in order to understand the working mechanism of electrolyte additives for achieving stable cycling performance. In the first part, a combination of solution and solid-state NMR techniques, including dynamic nuclear polarization (DNP) are employed to monitor the formation of the solid electrolyte interphase (SEI) on next-generation, high-capacity Si anodes in conventional carbonate electrolytes with and without fluoroethylene carbonate (FEC) additives. A model system of silicon nanowire (SiNW) electrode is used to avoid interference from the polymeric binder. To facilitate characterization via one- and two-dimensional NMR, ^13C-enriched FEC was synthesized and used, ultimately allowing a detailed structural assignment of the organic SEI. FEC is found to first defluorinated to form soluble vinylene carbonate (VC) and vinoxyl species, which react to form both soluble and insoluble branched ethylene-oxide-based polymers. In the second part, the same methodology is applied to study the decomposition products of pure FEC or VC electrolytes containing 1 M LiPF_6. The pure FEC/VC system simplifies the electrolyte solvent formulation and avoids the interaction between different solvent molecules. Polymeric SEIs formed in pure FEC or VC electrolytes consist mainly of cross-linked PEO and aliphatic chain functionalities along with additional carbonate and carboxylate species. The presence of cross-linked PEO-type polymers in FEC and VC correlates with good capacity retention and high Coulombic efficiencies of the SiNWs anode. Using ^29Si DNP NMR, the interfacial region between SEI and the Si surface was probed for the first time with NMR spectroscopy. Organosiloxanes form upon cycling, confirming that some of the organic SEI is covalently bonded to the Si surface. It is suggested that both the polymeric structure of the SEI and the nature of its adhesion to the redox-active materials are important for electrochemical performance. Finally, the soluble decomposition products of EC formed during electrochemical cycling have been thoroughly analyzed by solution NMR and mass spectrometry, in order to explain the capacity-fading of Si anodes in a conventional EC-based electrolyte and address questions that arose when studying the additive-containing electrolytes. The detailed structures for the EC-degradation products are determined: a linear oligomer consist of ethylene oxide and carbonate units is observed as the major degradation product of EC.

As baterias automotivas de chumbo-ácido são componentes, que em grande parte dos fabricantes de automóveis nacionais, estão garantidas (ou dentro do período de garantia) por um ano. Mesmo sendo um período considerado curto para garantia de um veículo nos dias atuais, a bateria tem uma contribuição expressiva para os custos de garantia nas empresas por este período. Com o intuito de conhecer a confiabilidade deste componente e verificar a influência do período de armazenagem do veículo produzido na confiabilidade da bateria, foi elaborado um estudo com dados coletados em campo por um determinado fabricante de automóveis de passeio. Os parâmetros de entrada destes dados são o tempo de pátio do veículo, o tempo em que uma falha na bateria foi detectada no período de um ano de garantia e a quantidade de falhas observadas no período. Os dados permitiram análises em função do tempo, de forma que pelo método de análise paramétrica foram traçadas as curvas de confiabilidade do produto representadas por uma distribuição de Weibull de dois parâmetros, bem como, da densidade de probabilidade de falha e ainda da taxa de falha. As análises forneceram uma estimativa da confiabilidade da bateria em função do tempo, da qual foi possível extrair algumas conclusões que serão descritas neste trabalho, tais como: o comportamento de falha por desgaste das baterias automotivas e a diminuição da confiabilidade de baterias de acordo com o tempo de pátio. / The guarantee for the automotive batteries at the majority of the national vehicle manufactures is given for 1 year. Even considering this as a short period actually, this component has an expressive contribution to the guarantee costs of the companies. With the intention to know the reliability of the automotive batteries and to evaluate the vehicle storage time influence, it was carried out a study case, which had as inputs collected data from a particular vehicle manufacturer. The selected input parameters for the analysis were the vehicle storage time, time to failure and the failures amount detected in a period of 1 year. This data provided information to analyze the time domain and, supported by the reliability parametric methods, estimate the reliability of the automotive batteries. The two parameter Weibull distribution is used to model the probability density function, the failure rate analysis and reliability providing information for the conclusion of this study, like the batteries behavior of failure by wear and the reliability decrease according to the storage time.

Silicon has the highest theoretical capacity of any known anode material, and silicon coated carbon nanotubes (Si-CNTs) have shown promise of dramatically increasing battery capacity. However, capacity fading with cycling and low rate capability prevent widespread use. Here, three studies on differing aspects of these batteries are presented. Here, three studies on differing aspects of these batteries are presented. The first examines the rate capability of these batteries. It compares the cycling of electrodes hundreds of microns thick with and without ten micron access holes to facilitate diffusion. The holes do not improve rate capability, but thinner coatings of silicon do improve rate capability, indicating that the limiting mechanism is the diffusion through the nanoscale bulk silicon. The second attempts to enable stable cycling of anodes heavily loaded with silicon, using a novel monolithic scaffolding formed by coating vertically aligned carbon nanotubes (VACNTs) with nanocrystalline carbon. The structure was only able to stabilize the cycling at loadings of carbon greater than 60% of the electrode by volume. These electrodes have volume capacities of ~1000 mAhr/ml and retained over 725 mAhr/ml by cycle 100. The third studies the use of an encapsulation method to stabilize the solid electrolyte interphase (SEI) and exclude the electrolyte. The method was only able to stabilize cycling at loadings below 5% silicon, but exhibits specific capacities as high as 3000 mAhr/g of silicon after 20 cycles.

<p>The lucrative characteristics of high energy and power density from lithium-ion batteries have also become drawbacks when they are not handled appropriately. The reactive and flammable materials present within the cell raise safety concerns which need to be addressed. Aging of the cell components occurs in a natural way due to continuous cycling. Constant intercalation/deintercalation of Li-ions into the active materials induces stresses that in the long-term cycling mechanically modify the electrodes in an irreversible way. Also, electrode/electrolyte side reactions diminish the Li-ion inventory reducing the cell capacity and lifetime. Along with cell aging, intentional/unintentional abuse tests can occur at the hands of the final user. Improper handling and operation may lead the Li-ion cell to failure and possibly going into thermal runaway. This condition represents a threat to safety not only for cell integrity but also for user safety. Failure event can occur not only in brand new cells but also in aged cells. Current degradation studies focus either on the long-term aging degradation mechanisms or on fresh new cells’ abuse test. And few of them focused on the combination of both of them. </p>In this work, the degradation of Li-ion cells is investigated at different levels. First, at the electrode level, the effect of electrode processing and the intercalation properties of an anode and cathode materials is investigated. Then, at the cell level, abuse conditions such as external short, overcharge and overdischarge are studied in fresh and aged cells with different levels of degradation. Last but not least, the cells are assembled in a module configuration to investigate how a minor difference from one cell to another can affect the long-term performance. The aim is accomplished via a controlled lab test approach in order to get insights about the electrochemical, thermal and morphological changes that take place when the cell degrades.<br>

Stable radicals have enjoyed widespread use in a variety of fields including synthetic chemistry, materials chemistry, energy storage, and biochemistry. This thesis outlines our investigations of cyclopropenium-based stable radicals and their application as redox mediators, redox-active ligands, catalysts, and materials for energy storage. The first chapter gives a brief overview of the use of radicals in synthetic chemistry. The principles that govern the stability of radicals is discussed and notable examples are highlighted. The second section of the first chapter reviews the aromatic platforms that have been developed by the Lambert group and how they might be converted into stable radical species. The second chapter details our study of 2,3-diaminocyclpropenones as stable radicals. These electron rich cyclopropenium derivatives undergo facile oxidation to yield a radical cation species. The origin of the stability of this oxygen-centered radical was elucidated by density functional theory calculations and analysis of the crystal structure. Diaminocyclopropenones were also found to be effective neutral L-type ligands in Ce(IV) complexes. EPR and UV-VIS experiments revealed that these complexes exhibited reversible homolytic dissociation of their diaminocyclopropenone ligands. The third chapter describes the use of trisaminocyclopropeniums as catholytes for nonaqueous redox flow batteries. A newly designed trisaminocyclopropenium structure could be accessed in large quantities and showed long lasting stability in its oxidized state. A new composite polyionic material was developed for use as a membrane suitable for organic solvent and high voltages. Cycling in combination with a perylenediimide anolyte yielded a 1.7 V battery that exhibited excellent coulombic efficiency and capacity retention. Using a spiro-bis(phthalimido) anolyte afforded a battery with an open circuit voltage of 2.8 V. The fourth chapter details how our battery studies with trisaminocyclopropenium radical dications led us to discover their photoinduced reactivity. We developed an electrophotocatalytic platform using trisaminocyclopropeniums as a species capable of being activated by both photochemical and electrochemical energy. The excited state oxidation potential of the doubly activated species was found to be +3.33 V, which was capable of effecting oxidative coupling reactions using both arenes and ethers as substrates. Density functional theory calculations and spectroscopic experiments revealed that the photoreactivity was due to a SOMO-inversion event. The trisaminocyclopropenium radical dication could be prepared on scale via direct electrolysis and subsequently used in high throughput screening.