Search

Flow batteries are an attractive technology for energy storage of grid-scale renewables. However, performance issues related to ion-exchange membrane (IEM) fouling and crossover of species have limited the success of flow batteries. In this work we propose the use of the solid-state sodium-ion conductor NaSICON as an IEM to fully eliminate active species crossover in room temperature, aqueous, neutral pH flow batteries. We measure the room temperature conductivity of NaSICON at 2.83–4.67 mS cm−1 and demonstrate stability of NaSICON in an aqueous electrolyte with conductivity values remaining near 2.5 mS cm−1 after 66 days of exposure. Charge and discharge of a full H-cell battery as well as symmetric cycling in a flow battery configuration using NaSICON as an IEM in both cases demonstrates the capability of the solid-state IEM. Extensive analysis of aged cells through electrochemical impedance spectroscopy (EIS) and UV–vis spectroscopy show no contaminant species having crossed over the NaSICON membrane after 83 days of exposure, yielding an upper limit to the permeability of NaSICON of 4 × 10−10 cm2 min−1. The demonstration of NaSICON as an IEM enables a wide new range of chemistries for application to flow batteries that would previously be impeded by species crossover and associated degradation.

The objectives of this report are as follows: elucidate degradation mechanisms, decomposition products, and abuse response for next generation silicon based anodes; and Understand the contribution of various materials properties and cell build parameters towards thermal runaway enthalpies. Quantify the contributions from particle size, composition, state of charge (SOC), electrolyte to active materials ratio, etc.

Anion receptors that bind strongly to fluoride anions in organic solvents can help dissolve the lithium fluoride discharge products of primary carbon monofluoride (CFx) batteries, thereby preventing the clogging of cathode surfaces and improving ion conductivity. The receptors are also potentially beneficial to rechargeable lithium ion and lithium air batteries. We apply Density Functional Theory (DFT) to show that an oxalate-based pentafluorophenyl-boron anion receptor binds as strongly, or more strongly, to fluoride anions than many phenyl-boron anion receptors proposed in the literature. Experimental data shows marked improvement in electrolyte conductivity when this oxalate anion receptor is present. The receptor is sufficiently electrophilic that organic solvent molecules compete with F– for boron-site binding, and specific solvent effects must be considered when predicting its F– affinity. To further illustrate the last point, we also perform computational studies on a geometrically constrained boron ester that exhibits much stronger gas-phase affinity for both F– and organic solvent molecules. After accounting for specific solvent effects, however, its net F– affinity is about the same as the simple oxalate-based anion receptor. Lastly, we propose that LiF dissolution in cyclic carbonate organic solvents, in the absence of anion receptors, is due mostly to the formation of ionicmore » aggregates, not isolated F– ions.« less

Entropy profiles of lithium cobalt oxide (LiCoO2) electrodes were measured at various stages in the cycle life to examine performance degradation and cycling-induced changes, or lack thereof, in thermodynamics. LiCoO2 electrodes were cycled at C/2 rate in half-cells (vs. lithium anodes) up to 20 cycles or C/5 rate in full cells (vs. MCMB anodes) up to 500 cycles. The electrodes were then subjected to entropy measurements (∂E/∂T, where E is open-circuit potential and T is temperature) in half-cells at regular intervals over the approximate range 0.5 ≤ x ≤ 1 in LixCoO2. Despite significant losses in capacity upon cycling, neither cycling rate resulted in any change to the overall shape of the entropy profile relative to an uncycled electrode, indicating retention of the basic LiCoO2 structure, lithium insertion mechanism, and thermodynamics. This confirms that cycling-induced performance degradation in LiCoO2 electrodes is primarily caused by kinetic barriers that increase with cycling. In the case of electrodes cycled at C/5, there was a subtle, quantitative, and gradual change in the entropy profile in the narrow potential range of the hexagonal-to-monoclinic phase transition. The observed change is indicative of a decrease in the intralayer lithium ordering that occurs at these potentials, and itmore » demonstrates that a cyclinginduced structural disorder accompanies the kinetic degradation mechanisms.« less

Isothermal microcalorimetry is a promising characterization tool to both enhance understanding of degradation processes of silicon anodes in lithium-ion batteries, as well as to measure the impacts of efforts to improve the performance of these electrode materials. Isothermal microcalorimetry is a thermal characterization technique in which a sample is held in isothermal conditions while heat flow into / out of the sample to maintain a constant temperature is measured. This heat flow is determined by endothermic or exothermic reactions within the prepared sample. Isothermal titration calorimetry is a similar technique in which controlled amounts of a reactant are introduced to the sample during measurement, allowing capture of the reaction energies between materials. Both techniques are employed in our study of silicon anodes. Silicon has long been pursued as a next generation anode material for lithium-ion batteries due to its very high theoretical capacity relative to the currently employed graphite anode. So far, however, the practical realization of silicon anodes has been limited by both mechanical degradation caused by volume change during lithiation / delithiation as well as electrode surface instability that combine to shorten battery life and reduce capacity. Mitigation of these issues is being pursued through advanced binder materials, electrolyte additives, nano-scale architectures, and composite active materials.1 Previous microcalorimetry work on graphite anodes has demonstrated the ability of the technique to discern different lithiation reactions, measure the severity of parasitic losses, and identify the points of electrolyte degradation.2 Sandia National Laboratories is employing isothermal microcalorimetry to study new composite anodes of 15 wt.% silicon combined with standard graphite active material. Our microcalorimetry characterization approach uses a TA Instruments TAM IV with resolution to ± 200 nW to analyze samples of silicon-composite electrodes and compare their thermal signatures to baseline graphite samples to offer an understanding of the impact of silicon during electrode processing and fabrication, initial active material passivation upon electrolyte exposure, and electrochemical cycling of the battery. Our initial work focuses on the impacts of processing various silicon materials using a commonly employed LiPAA binder in an aqueous solvent. While silicon has a naturally passivating surface oxide layer, heat and gas generation can be observed during slurry mixing with the sub-micron sized silicon materials used in lithium-ion batteries. This reactivity can serve to rob capacity and increase material impedance before the electrode is even built into a battery. Later work will focus on characterizing the composite silicon electrodes during electrochemical cycling. When cycled in situ, the microcalorimeter can capture heat flow related to both surface electrolyte reactions and lithiation reactions of the silicon electrodes as well as the impacts of new binders, additives, or surface passivation approaches to mitigate these effects. References Yang J., Wang B. F., Wang K., Liu Y., Xie J. Y., Wen Z. S., Solid-State Lett., 6, A154 (2003). Krause L. J., Jensen L. D., Dahn J. R., Electrochem. Soc., 159, A937 (2012). Acknowledgment: Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

This article will briefly discuss the genesis of the Li-ion chemistry and its meteoric rise to prominence, supplanting aqueous rechargeable batteries such as NiCd and NiMH. The principal intent of this article is to discuss the issues with thermal instability of common Li-ion electrolytes, which detract from the positive attributes of this chemistry. The development of an innovative low-cost non-flammable electrolyte will greatly improve the safety and reliability of lithium batteries, a key technological hurdle that must be overcome for the wider application of this chemistry. This article will also include the advancements made in combating/mitigating solvent flammability through the addition of fire retardants, fluoro-solvents, ionic liquids etc. The scope of the article will be limited to the flammability of non-aqueous solvents and will not include the thermal instability issues of anodes and cathodes. We will elaborate using examples from our in-house research aimed at mitigating solvent flammability by using hydrofluoro ethers (HFEs) as a cosolvent. Additionally, we will describe in-house capabilities for prototyping 18,650 cells and in-house thermal abuse test capabilities that allow us to evaluate materials and their thermal responses in actual cell configurations.

The optimum combination of high energy density at the desired power sets lithium-ion battery technology apart from the other well known secondary battery chemistries. However, this is besieged by thermal instability of the electrolyte. This “Achilles heel” still remains a significant safety issue and unless this propensity is improved the promise of widespread adoption of Li-ion batteries for Transportation application may not be realized. With this in mind we launched a systematic study to evaluate fluoro solvents that are known to be nonflammable, for thermal and electrochemical performances. We investigated hydro-fluoro-ethers (HFE) (1) 2-trifluoromethyl-3-methoxyperfluoropentane {TMMP} and (2) 2-trifluoro-2-fluoro-3-difluoropropoxy-3-difluoro-4-fluoro-5-trifluoropentane {TPTP} in Sandia-built cells. Thermal properties under near abuse conditions that exist in thermal runaway environment and the electrochemical characteristics for these electrolytes were measured. In the thermal ramp (TR) measurement, EC:DEC:TPTP-1 M LiBETI (or TFSI or LiPF 6 ) electrolytes exhibited no ignition/fire. Similar behavior was observed for the EC:DEC:TMMP-1 M LiBETI. Further, in ARC studies the HFE electrolytes generated less gas by 50% compared to the EC:EMC-1.2 M LiPF 6 {CAR-1} electrolyte. Although in all cases the HFEs generated less gas, the onset of gas generation appears to depend on the salt. For the LiBETI and TFSI containing HFEs the onset is pushed out by ∼80 °C and for the LiPF 6 the onset is comparable to that of the CAR-1. The solution ionic conductivity of these HFE electrolytes was lower (4–5 times) than that of the CAR-1 electrolyte however, the electrochemical performance was comparable. For example, full cells in 2032 type coin cells containing LiMN 0.33 Ni 0.33 Co 0.33 O 2 cathode and carbon anode showed around 5 mA h capacity and the computed specific capacity was ∼154 mA h for all the electrolytes. In half-cells against lithium the cathode and anode gave specific capacity on the order of 170 mA h and 340 mA h respectively. These electrolytes when tested in 18,650 cells containing the above cathode and anode also showed comparable capacity. Further, the voltage stability window was not compromised by the HFEs. ARC measurements on 18,650 full cells showed less gas generation for the HFE electrolytes compared to CAR-1 electrolyte.

Lithium-ion cell field failures due to internal short circuits are a significant concern to the entire lithium-ion cell market from consumer electronics to electric vehicles. While the probability of these failure events occurring is estimated to be very low (1 in 5–10 million), the consequences of a cell failure due to an internal short in a high energy battery system have the potential to be catastrophic. The statistical probability of one of these events is very low and they are difficult to predict and simulate in a laboratory using some external test; which makes cell failure due to an internal short circuit a unique challenge to overcome. Several of the experiments designed to simulate internal shorts have been adopted as testing protocols across the industry; in general, they do not accurately simulate an internal short. This work highlights our efforts to experimentally trigger an internal short circuit in a lithium-ion cell.

Recent advances in electrode fabrication and tailoring electrolyte properties for numerous applications have generated wide-spread interest in (CFx)n chemistry since it has the highest theoretical capacity and hence longer life of the four well known Li-primary chemistries. We are applying "Li-ion technology" electrode fabrication methodologies and preparing thin film (CFx)n electrodes in-house for evaluation. In this program we have evaluated 4 different (CFx)n materials for performance in coin cells in the temperature regime -55 to 72oC. We continue to evaluate the top performer in 18650 cell configuration and obtained ~3.6 Ahrs capacity at a C/400 rate. We also evaluated cathodes of different lengths for uniformity of loading. For example, a 36" long x 4.2 mil thick electrode gave 3.6 Ahrs while a 29" long x 4.2 mil thick electrode gave 2.9 Ahrs of capacity. We also measured impedance at different voltages and thermal abuse response of the 18650 cells.

(CFx)ncathode reaction during discharge has been investigated usingin situX-ray diffraction (XRD). Mathematical treatment of thein situXRD data set was performed using multivariate curve resolution with alternating least squares (MCR–ALS), a technique of multivariate analysis. MCR–ALS analysis successfully separated the relatively weak XRD signal intensity due to the chemical reaction from the other inert cell component signals. The resulting dynamic reaction component revealed the loss of (CFx)ncathode signal together with the simultaneous appearance of LiF by-product intensity. Careful examination of the XRD data set revealed an additional dynamic component which may be associated with the formation of an intermediate compound during the discharge process.

Although (CFx)n has the highest theoretical specific energy of all lithium primary batteries known, the practical specific energy (including packaging) is very small on the order of ∼10% for small capacity cells ( 100 Ah). Even these can be achieved only at a very low discharge rate

As lithium-ion battery technologies mature, the size and energy of these systems continues to increase (> 50 kWh for EVs); making safety and reliability of these high energy systems increasingly important. While most material advances for lithium-ion chemistries are directed toward improving cell performance (capacity, energy, cycle life, etc.), there are a variety of materials advancements that can be made to improve lithium-ion battery safety. Issues including energetic thermal runaway, electrolyte decomposition and flammability, anode SEI stability, and cell-level abuse tolerance continue to be critical safety concerns. This report highlights work with our collaborators to develop advanced materials to improve lithium-ion battery safety and abuse tolerance and to perform cell-level characterization of new materials.

Organosilicon electrolytes exhibit several important properties for use in lithium carbon monofluoride batteries, including high conductivity/low viscosity and thermal/electrochemical stability. Conjugation of an anion binding agent to the siloxane backbone of an organosilicon electrolyte creates a bi-functional electrolyte. The bi-functionality of the electrolyte is due to the ability of the conjugated polyethylene oxide moieties of the siloxane backbone to solvate lithium and thus control the ionic conductivity within the electrolyte, and the anion binding agent to bind the fluoride anion and thus facilitate lithium fluoride dissolution and preserve the porous structure of the carbon monofluoride cathode. The ability to control both the electrolyte conductivity and the electrode morphology/properties simultaneously can improve lithium electrolyte operation.

A kinetic expression for the chemical reaction of lithium metal with thionyl chloride is presented that is consistent with calorimetric measurements of the heat generation from a thionyl chloride cell. The kinetics expression is incorporated into a well-established electrochemical model for the discharge behavior, and then used to estimate the life of the battery under an intermittent discharge so as to assess the importance of lithium corrosion. The model predicts that, under the conditions examined, there is no danger of depleting the lithium anode and so introducing a safety hazard.

The seminal research by Wright et al. on polyethylene oxide (PEO) solid polymer electrolyte (SPE) generated intense interest in all solid-state rechargeable lithium batteries. Following this a number of researchers have studied the physical, electrical and transport properties of thin film PEO electrolyte containing Li salt. These studies have clearly identified the limitations of the PEO electrolyte. Chief among the limitations are a low cation transport number (t + ), high crystallinity and segmental motion of the polymer chain, which carries the cation through the bulk electrolyte. While low t + leads to cell polarization and increase in cell resistance high Tg reduces conductivity at and around room temperatures. For example, the conductivity of PEO electrolyte containing lithium salt is

At Sandia National Laboratories, we have built 18650 Li-ion cells with Li reference electrode for in situ characterization of electrodes including impedance and other electrochemical properties. At a 200 mA (∼C/5 rate) discharge, the cell gave ∼900 mAh. Impedance measurements indicate that the anode dominates the cell impedance. For example, at 0 °C, the anode and cathode impedances at 10 mHz were around 149 and 53 mΩ, respectively, and the total cell impedance at 10 mHz was ∼203 mΩ. The three-electrode configuration also permits measurement of individual electrode voltages during charge and discharge. During discharge, while the cell voltage drops from 4.1 to 3 V, the cathode and the anode voltages change from 4.1 to 3.7 and from ∼0 to 0.7 V, respectively. During charge, the cathode and anode voltages trace back to their initial values before discharging. The voltage swing for the anode is higher than that for the cathode. This also indicates that the impedance for the anode is higher than for the cathode. Pulse measurements on the cells indicate the voltage drop of the full-cell is equal to the sum of the anode and cathode voltage drops for a 2 A discharge pulse.

Li ion cells are being developed for high-power applications in hybrid electric vehicles, because these cells offer superior combination of power and energy density over current cell chemistries. Cells using this chemistry are proposed for battery systems in both internal combustion engine and fuel cell-powered hybrid electric vehicles. However, the safety of these cells needs to be understood and improved for eventual widespread commercial applications. The thermal-abuse response of Li ion cells has been improved by the incorporation of more stable anode carbons and electrolyte additives. Electrolyte solutions containing vinyl ethylene carbonate (VEC), triphenyl phosphate (TPP), tris(trifluoroethyl)phosphate (TFP) as well as some proprietary flame-retardant additives were evaluated. Test cells in the 18,650 configuration were built at Sandia National Laboratories using new stable electrode materials and electrolyte additives. A special test fixture was designed to allow determination of self-generated cell heating during a thermal ramp profile. The flammability of vented gas and expelled electrolyte was studied using a novel arrangement of a spark generator placed near the cell to ignite vent gas if a flammable gas mixture was present. Flammability of vent gas was somewhat reduced by the presence of certain additives. Accelerating rate calorimetry (ARC) was also used to characterize 18,650-size test cell heat and gas generation. Gas composition was analyzed by gas chromatography (GC) and was found to consist of CO2, H2, CO, methane, ethane, ethylene and small amounts of C1–C4 organic molecules.

Battery life is an important, yet technically challenging, issue for battery development and application. Adequately estimating battery life requires a significant amount of testing and modeling effort to validate the results. Integrated battery testing and modeling is quite feasible today to simulate battery performance, and therefore applicable to predict its life. A relatively simple equivalent-circuit model (ECM) is used in this work to show that such an integrated approach can actually lead to a high-fidelity simulation of a lithium-ion cell's performance and life. The methodology to model the cell's capacity fade during thermal aging is described to illustrate its applicability to battery calendar life prediction.

We present an equivalent-circuit-based battery model, capable of simulating charge and discharge behavior of lithium-ion batteries (LiB). The model, although simple in concept, can simulate complex discharge behavior with high fidelity, as validated by experimental results.

All-solid-state thin film micro-batteries comprised of a lithium anode, lithium phosphorus oxy-nitride (LiPON) solid electrolyte and Li x CoO 2 cathode were evaluated at different temperatures from −50 to 80 °C for electrical behavior and impedance raise. The cell dimensions were ∼2 cm long, ∼1.5 cm wide and ∼15 μm thick. The rated capacity of the cells was about 400 μAh. The cells were cycled (charge/discharge) at room temperature over 100 times at a 0.25 C rate. The charge and discharge cut-off voltages were 4.2 and 3.0 V, respectively. The cells did not show any capacity decay over 100 cycles. The measured capacity was 400 μAh. The coulombic efficiency was 1, which suggests that the cell reaction is free from any parasitic side reactions and the lithium intercalation and de-intercalation reaction is completely and totally reversible. These cells also have good high-rate performance at room temperature. For example, these cells discharged at a 2.5 C rate delivered ∼90% of the capacity at a 0.25 C rate. However, the delivered capacities even at a 0.25 C rate at 80 and −50 °C were much lower than the room temperature capacity. Cells soaked at −50 °C were not damaged permanently as seen by the near normal behavior when returned to room temperature. However, cells heated to 80 °C were permanently damaged as seen by the lack of normal performance back at room temperature. Cell impedance was measured before and after cycling at different temperatures. The high-frequency resistance (generally ascribed to the electrolyte and other resistances in series with the electrolyte resistance) decreased with decreasing temperature. However, the interfacial resistance increased significantly with decreasing temperature. Further, the electrolyte resistance accounted for ∼2% of the total cell resistance. The cycled cells showed higher impedance than the uncycled cells.

A statistically designed accelerated aging experiment was conducted to investigate the effects of aging time, temperature, and state-of-charge (SOC) on the performance of lithium-ion cells. In this experiment, a number of cells were stored in a variety of static aging environments ranging from 25 °C and 60% SOC to 55 °C and 100% SOC. The power output of each cell was monitored regularly over the course of 44 weeks via a low current level hybrid pulse power characterization test. A single empirical model of power fade, involving two concurrent degradation processes, was found to be applicable over a wide range of experimental conditions. The first degradation process is relatively rapid (nearly complete within 4 weeks) and is accelerated by temperature with unknown kinetics. The second degradation process (accelerated by temperature and SOC) is less rapid and exhibits time3/2 kinetics.

Sandia National Laboratories has been studying calendar and pulse discharge life of prototype high-power lithium-ion cells as part of the Advanced Technology Development (ATD) Program. One of the goals of ATD is to establish validated accelerated life test protocols for lithium-ion cells in the hybrid electric vehicle application. In order to accomplish this, aging experiments have been conducted on 18650-size cells containing a chemistry representative of these high-power designs. Loss of power and capacity are accompanied by increasing interfacial impedance at the cathode. These relationships are consistent within a given state-of-charge (SOC) over the range of storage temperatures and times. Inductive models have been used to construct detailed descriptions of the relationships between power fade and aging time and to relate power fade, capacity loss and impedance rise. These models can interpolate among the different experimental conditions and can also describe the error surface when fitting life prediction models to the data.

Thermal and electrochemical properties of LiPF6 and LiN(SO2C2F5)2, known as LiBETI, are compared in binary mixtures of ethylene carbonate (EC) and ethyl methyl carbonate (EMC). Thermal stability of these electrolytes was tested between 50 and 120 °C at 10 °C intervals. The electrolytes were kept at temperature for 1 week and every other day they were taken out of the oven and checked visually for any discoloration. Visual inspection showed discoloration of the LiPF6-containing electrolyte even at 60 °C (straw brown to dark brown at 120 °C) while the electrolyte containing LiBETI showed no or very little discoloration even at 120 °C. In lithium-ion “Tee” cells the electrolyte with LiBETI cycled equally well or better than the electrolyte containing LiPF6 at room temperature. Tee cells containing LiBETI showed lower cathode impedance than and comparable anode impedance to the cells containing electrolytes with LiPF6. In a full cell the electrodes performed equally well.

A series of cylindrical 18650 lithium-ion cells with an MAG-10|1.2 M LiPF 6 ethylene carbonate (EC):ethyl methyl carbonate (EMC) (w:w=3:7)|Li x Ni 0.8 Co 0.15 Al 0.05 O 2 configuration were made and tested for power-assist hybrid electric vehicle (HEV) applications under various aging conditions of temperature and state-of-charge (SOC). The cells were intermittently characterized for changes in power capability, rate capacity, and impedance as aging progressed. The changes of these properties with temperature, as depicted by Arrhenius equations, were analyzed. We found that the degradation in power and capacity fade seems to relate to the impedance increase in the cell. The degradation follows a multi-stage process. The initial stage of degradation has an activation energy of the order of 50–55 kJ/mol, as derived from power fade and C 1 capacity fade measured at C /1 rate. In addition, microcalorimetry was performed on two separate unaged cells at 80% SOC at various temperatures to measure static heat generation in the cells. We found that the static heat generation has an activation energy of the order of 48–55 kJ/mol, similar to those derived from power and C 1 capacity fade. The correspondence in the magnitude of the activation energy suggests that the power and C 1 capacity fades were related to the changes of the impedance in the cells, most likely via the same fading mechanism. The fading mechanism seemed to be related to the static heat generation of the cell.

Improved interfacial resistance was observed in lithium cells by the use of new additives. The additives, nitrile sucrose and nitrile cellulose and their lithium salts, were evaluated in polyvinylidene difluoride (PVDF) thin-film gelled electrolytes containing a mixture of ethylene carbonate (EC) and diethyl carbonate (DEC). The electrochemical properties of the films with and without the additives were measured as a function of temperature and compared. The interfacial resistance ( R in ) of the films with the additives was significantly lower than that without the additives, especially at sub-ambient temperatures. For example, the R in at −20°C for the films with additives was around 7000 Ω cm 2 and that for the films without the additives was >20,000 Ω cm 2 . Results obtained from using the additives in lithium-ion (Li-ion) cells show significant improvements in the low frequency resistance of the cells.

Low temperature electrical performance characteristics of A&T, Moli, and Panasonic 18650 Li-ion cells are described. Ragone plots of energy and power data of the cells for different temperatures from 25 °C to −40 °C are compared. Although the electrical performance of these cells at and around room temperature is respectable, at temperatures below 0 °C the performance is poor. For example, the delivered power and energy densities of the Panasonic cells at 25 °C are ∼800 W l−1 and ∼100 Wh l−1, respectively, and those at −40 °C are

Two- and three-electrode impedance measurements were made on 18650 Li-ion cells at different temperatures ranging from 35°C to −40°C. The ohmic resistance of the cell is nearly constant in the temperature range studied although the total cell impedance increases by an order of magnitude in the same temperature range. In contrast to what is commonly believed, we show from our three-electrode impedance results that, the increase in cell impedance comes mostly from the cathode and not from the anode. Further, the anode and cathode contribute to both the impedance loops (in the NyQuist plot).

The lithium/thionyl chloride battery (Li/SOCl 2) has received considerable attention as a primary energy source due to its high energy density, high operating cell voltage, voltage stability over 95% of the discharge, large operating temperature range ( 255 to 708C), long storage life, and low cost of materials. 1,2 However, a loss in performance may occur after periods of prolonged storage at high and low temperatures or when exposed to intermittent use. This loss in performance may result in reduced capacity or even worse, catastrophic failure, especially when operated at high discharge rates. High discharge rates and high temperatures promote thermal runaway, which can result in the venting of toxic gases and explosion. 2 Mathematical models can be used to tailor a battery design to a specific application, perform accelerated testing, and reduce the amount of experimental data required to yield efficient, yet safe cells. Models can also be used in conjunction with the experimental data for parameter estimation and to obtain insights into the fundamental processes occurring in the battery. Previous investigators 3,4 presented a one-dimensional mathematical model of the Li/SOCl2 battery. They used porous electrode theory 5 to model the porous cathode and concentrated solution theory 6 for the electrolyte solution to study the effect of various design and operational parameters on the discharge curves. The model equations were written under the assumption that the excess electrolyte was in a reservoir between the separator and the porous cathode. The result was that the electrolyte replenished the porous cathode through the front face of the electrode. The theoretical results showed similar qualitative trends to those observed experimentally. However, a lack of experimental data and unknown values for many of the kinetic and transport parameters as a function of temperature prevented quantitative comparisons. Evans and White 7 presented a parameter estimation technique and used it in conjunction with the one-dimensional mathematical model presented earlier. 4 However, the comparison between simulated and experimental discharge curves was done only for partially discharged cells at ambient temperature. This paper presents a one-dimensional mathematical model for the Li/SOCl2 cell, with model equations similar to those presented previously.3,4 The exception is the modification to the material balance in the porous cathode that accounts for electrolyte replenishment through the top rather than the front of the porous cathode. 8 The model is used to predict discharge curves at low-to-moderate discharge rates (discharge loads #10 V, corresponding to current densities less than 2 mA/cm 2 for a D-size cell). Previous thermal models of Li/SOCl2 cells have shown that under these operating conditions thermal runaway is not a problem. 9,10 Therefore, it is also assumed here that the temperature of the cell is uniform throughout but allowed to change during discharge. 3,4 The model is then used in conjunction with experimental data to obtain estimates for the transference number, diffusion coefficient, and kinetic parameters for the reactions at the anode and cathode as a function of temperature. Using the estimated parameters, the model predictions show good agreement with the experimental data over a wide temperature (255 to 498C) and load range (10 to 250 V). Finally, the model is used to study the effect of cathode thickness on cell performance as a function of operating temperature and load to illustrate the application in optimization studies. Experimental

Electrochemical properties of cylindrical (18 650, 17 500) and prismatic (48.3×25.4×7.6 mm) Li-ion cells from different manufacturers including A&T, Panasonic, Polystor, Sanyo, and Sony were studied. Impedance and pulse characteristics of these cells were evaluated for three open circuit voltages (OCVs): 4.1 V (fully charged), 3.6 V (partially discharged), and 3.1 V (nearly completely discharged) in the temperature regime +35°C to −40°C. The cell ohmic resistance was nearly constant from +35°C to −20°C, but increased by 2–3 times at −40°C. For example, the cylindrical Sony cells showed an average bulk resistance of ∼80 mΩ between 35°C and −20°C and ∼290 mΩ at −40°C for the three OCVs studied. The cell ohmic resistance remained nearly constant with OCV. The NyQuist plot (real vs. imaginary impedance) showed, at high frequencies (2.7–65 kHz), an inductive segment characteristic of a porous electrode and/or a jelly-roll cell design. The NyQuist plots also showed two ill-defined loops, a smaller loop at higher frequencies attributed to the anode electrolyte interface and a larger loop at lower frequencies due to the cathode electrolyte interface. A smaller charge transfer resistance (Rct) at the anode is indicated and the performance of the cell may be improved by reducing the interfacial resistances, in general. Ragone plots, relating energy density and power density or specific power and specific energy, were also constructed to compare the performance characteristics of these cell types. In the current range studied (20–1000 mA), the energy/power performance of both A&T and Panasonic cells is better than the rest. For these two cells, the power (W/kg or W/l) didn't reach a plateau in the current range studied. These data should be considered, however, in the context that the A&T and Panasonic cells may be newer (later generation) than the other cells used in this study. However, at higher currents (>2 A) and at lower temperatures, for the Panasonic cells, power reaches a plateau. This behavior is also true for the A&T cells. The cells were pulsed at different temperatures both as a function of OCV and current pulse amplitude. The cell voltage drop is almost linear with pulse current at ambient and slightly subambient temperatures. However, at lower temperatures, the voltage drop is nonlinear with pulse current, suggesting that the contribution of charge transfer resistance to the overall cell impedance under load is nontrivial. In addition, the cell voltage drop for a given current pulse increases with depth of discharge. For example, for a 500 mA pulse at 4.1 V, 3.6 V, and 3.1 V, the Panasonic cells showed an average voltage drop of 94 mV, 130 mV, and 200 mV, respectively, at room temperature. A similar observation was made for the other Li-ion cells.

The primary current‐collector materials being used in lithium‐ion cells are susceptible to environmental degradation: aluminum to pitting corrosion and copper to environmentally assisted cracking. Localized corrosion occurred on bare aluminum electrodes during simulated ambient‐temperature cycling in an excess of electrolyte. The highly oxidizing potential associated with the positive‐electrode charge condition was the primary factor. The corrosion mechanism differed from the pitting typically observed in aqueous electrolytes because each site was filled with a mixed metal/metal‐oxide product, forming surface mounds or nodules. Electrochemical impedance spectroscopy was shown to be an effective analytical tool for characterizing the corrosion behavior of aluminum under these conditions. Based on X‐ray photoelectron spectroscopy analyses, little difference existed in the composition of the surface film on aluminum and copper after immersion or cycling in electrolytes made with two different solvent formulations. Although Li and P were the predominant adsorbed surface species, the corrosion resistance of aluminum may simply be due to its native oxide. Finally, copper was shown to be susceptible to environmental cracking at or near the lithium potential when specific metallurgical conditions existed (work hardening and large grain size). © 1999 The Electrochemical Society. All rights reserved.

We describe below the electrochemical performance characteristics of cylindrical (18 650) and prismatic (48.26×25.4×7.62) lithium-ion cells at ambient and sub-ambient temperatures. Ragone plots of power and energy data for these cells are compared, and indicate that at room temperature, the ∼500 mAh prismatic lithium-ion cells exhibit higher specific power and power density than the ∼1100 mAh cylindrical cells. Over the temperature range from 35 to −20°C, the cell impedance is almost constant for both cell types. These cells at 100% state-of-charge (s.o.c.) show very little voltage drop for current pulses up to 1 A.

The application space for lithium ion cells is rapidly expanding to include transportation, grid storage, space system, and military power needs. Increasingly demanding mission requirements in these applications has necessitated higher energy and higher power energy storage systems. As the size and capacity of these power systems increases, so do the consequences of an off normal battery event (thermal runaway, short circuit, mechanical crush, etc.). Significant scientific efforts have been made to understand the consequences and requirements for runaway, understand policy decisions pertaining to battery safety, understand safety response in packs, and model thermal abuse of cells. Despite great progress, the overall technology still poses safety concerns for the public which, along with other considerations, have hindered the widespread adoption of many lithium based storage technologies. Several high-profile safety occurrences in recent history have brought the subject of lithium ion battery safety into the forefront of discussion within the community. While many solutions have been proposed, there has yet to be a robust and cost effective solution for the issue of battery safety, particularly in high capacity storage systems. Additionally, next generation materials such as silicon graphite/graphene anodes and high voltage cathodes are posed to increase the electrochemical performance of future systems, but little is known with regards to the abuse response of these materials. Our research efforts have focused on developing understanding and strategies to develop intrinsically safe lithium ion based systems. This material science based strategy to ensure battery safety offers benefits over engineered solutions to battery safety. As next generation materials mature, it becomes increasingly important to apply similar methodologies to understand and optimize the abuse response and safety envelope for these batteries. Intrinsically safe battery research topics have focused on materials development and coating efforts for anode and cathode safety, electrolyte flammability and reactivity reduction efforts, separator instability considerations, and solutions targeted at the mitigating the energy released in the case of a cell abuse. The understanding of next generation silicon carbon composites will be discussed to compare with current graphite based systems. Industrially relevant 18650 cell were built using silicon carbon composite anodes versus NMC 523 cathodes to understand both electrochemical and abuse performance, see Figure 1. Initial experiments show similar performance for specific capacity and rate capability for these cells. However, these next generation anode materials appear to have some potential differences in the degradation mechanisms and overall response during abuse conditions. These results may help to inform future development efforts, particularly as industry pushes towards higher silicon content in anode materials or towards significantly higher energy dense materials. Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000. Figure 1

Silicon has long been sought as a replacement to graphite anodes in next generation lithium-ion batteries due largely to it’s extremely high theoretical capacity of ~3500 mAh g-1 (for the formation of Li15Si4). This full value has yet to be realized in a practical electrode system due to both mechanical degradation caused by volume change during lithiation / delithiation and electrode surface instability that combine to cause rapid loss of capacity. However, advances in binder materials, materials synthesis of nano-silicon, and electrolyte additives have enabled the development of graphite and nano-silicon (Graphite + nSi) composite anodes that possess greater specific capacity that typical graphite electrodes and can be practically incorporated into improved lithium-ion cells. Sandia National Labs is conducting both materials scale and battery scale safety tests to determine the effects of nano-silicon incorporation on the safety performance of lithium-ion batteries. This study will focus on the effects of varying the silicon content in Graphite + nSi electrodes, using different binder materials, and using different electrolytes or additives on the failure mechanisms and characteristics of lithium-ion battery materials. Performance of Graphite + 15 % nSi Electrodes Electrodes of both graphite and Graphite + 15 % nSi composite were prepared using an aqueous LiPAA binder and 1.2 M LiPF6 in EC:EMC (3:7) electrolyte1. The electrodes possessed areal capacities of approximately 1.6 mAh cm-2 and were cycled in both half-cell configuration vs. lithium foil and in full cell configuration vs. an NCM 523 cathode. The voltage vs. capacity curve for the two electrodes, see Figure 1a, show several plateaus below 0.2 V arising from graphite activity with lithium while the 15 % nSi electrode also has significant capacity at higher voltages arising from lithiation of silicon which gives Graphite + 15 % nSi a specific capacity of 610 mAh g-1 compared to 319 mAh g-1 for graphite only. The differential capacity plot in Figure 1b shows more clearly the capacity contribution from the silicon peaks which occur at roughly 0.45 V and 0.25 V during charge and discharge. Figure 1d shows the cathode capacity retention and coulombic efficiency for the composite anode tested in full cell configuration, demonstrating the continued need for performance improvement for greater capacity retention. Figure 2 shows post cycling DSC analysis carried out on Graphite + 15 % nSi electrodes extracted from cycled full cells. After several formation cycles the cells were discharged to a prescribed voltage corresponding to a given state-of-charge (SOC). As expected the higher SOC electrodes generated more heat during testing. The various voltages were also selected for the given lithiation state of the anode: at 3.1 V the anode is fully delithiated, showing minimal thermal response; at 3.4 V the graphite is delithiated and the silicon is lithiated, making it the primary reactant and contributor to the thermal response at this state; at 4.1 V both graphite and silicon are lithiated so the resulting DSC will be a combined response of the two lithiated materials. As such, by looking at the 3.4 V DSC curve it can be seen that silicon has two strong exothermic peaks at ~ 200 °C and 260 °C, while the 4.1 V curve shows additional exothermic peaks at both 150 °C and 300 °C corresponding to lithiated graphite. So far this data agrees with previous results comparing graphite to silicon electrodes. References Klett M., Gilbert J. A., Trask S. E., Polzin B. J., Jansen A. N., Dees D. W., Abraham D. P., J. Electrochem. Soc., 163, A875 (2016). Acknowledgment: Sandia Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation for the U. S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL 85000. Figure 1

We are currently evaluating large capacity (20–40 Ah) Bluestar (cylindrical) and Yardney (prismatic) lithium-ion cells for their electrical and electrochemical performance characteristics at different temperatures. The cell resistances were nearly constant from room temperature down to −20°C, but increased by over 10 times at −40°C. The specific energies and powers, as well as the energy densities and power densities are high and did not reach a plateau even at the highest discharge rates tested. For example, the prismatic lithium-ion cells gave close to 280 Wh dm−3 from a 4 A discharge and 249 Wh dm−3 at 20 A, both at room temperature. For the same current range the specific energy values were 102 Wh kg−1 and 91 Wh kg−1. Cycle life and other electrical and electrochemical properties of the cells will be presented.

This LDRD project attempts to use novel electrochemical techniques to understand the reaction mechanism that limits the discharge reaction of lithium CF{sub x} chemistry. If this advanced component development and exploratory investigations efforts are successful we will have a High Energy Density Li Primary Battery Technology with the capability to double the run time in the same volume, or provide the same energy in a much smaller volume. These achievements would be a substantial improvement over commercial Li/Thionyl chloride battery technology. The Li(CF{sub x}){sub n} chemistry has the highest theoretical energy (and capacity) and hence very attractive for long life battery applications. However, the practical open circuit voltage (OCV) is only 3.2 V which is {approx}1.3 V lower than the thermodynamic cell voltage (for an in depth explanation of the voltage depression refer to 'Introduction'). The presence of intermediate has been invoked to explain the lower OCV of the cell. Due to the reduction in cell voltage the cell out put is reduced by {approx}40%. To account for the initial voltage loss a mechanism has been proposed which involves the formation of a ternary compound (like C(LiF){sub x}). But neither its presence nor its nature has been confirmed. Our work will seek to develop understanding of the voltage depression with a goal to produce a primary battery with improved properties that will have significant impact in furthering advancements.

Anion receptors that bind strongly to fluoride anions in organic solvents can help dissolve the lithium fluoride discharge products of primary carbon monouoride (CFx) batteries, thereby preventing the clogging of cathode surfaces and improving ion conductivity. The receptors are also potentially beneficial to rechargeable lithium ion and lithium air batteries. We apply Density Functional Theory (DFT) to show that an oxalate-based pentauorophenyl-boron anion receptor binds as strongly, or more strongly, to uoride anions than many phenyl-boron anion receptors proposed in the literature. Experimental data shows marked improvement in electrolyte conductivity when this oxalate anion receptor is present. The receptor is sufficiently electrophilic that organic solvent molecules compete with F - for boron-site binding, and specific solvent effects must be considered when predicting its F - affinity. To further illustrate the last point, we also perform computational studies on a geometrically constrained boron ester that exhibits much stronger gas-phase affinity for both F - and organic solvent molecules. After accounting for speci_c solvent effects, however, its net F - affinity is about the same as the simple oxalate-based anion receptor. Finally, we propose that LiF dissolution in cyclic carbonate organic solvents, in the absence of anion receptors, is due mostly to the formation of ionic aggregates, not isolated F - ions. Sandia National Laboratories is a multiprogram laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corpo ration, for the U.S. Deparment of Energy's National Nuclear Security Administration under contract DE-AC04-94AL85000. Figure 1

The Advanced Technology Development Program is currently evaluating the performance of the second generation of Lithium-ion cells (i.e., Gen 2 cells). The 18650-size Gen 2 cells consist of a baseline chemistry and one variant chemistry. These cells were distributed over a matrix consisting of three states-of-charge (SOC) (60, 80, and 100% SOC), four temperatures (25, 35, 45, and 55°C), and three life tests (calendar-, cycle-, and accelerated-life). The calendar-life cells are clamped at an opencircuit voltage corresponding to 60% SOC and undergo a once-per-day pulse profile. The cycle-life cells are continuously pulsed using a profile that is centered around 60% SOC. The accelerated-life cells are following the calendar-life test procedures, but using the cycle-life pulse profile. Life testing is interrupted every four weeks for reference performance tests (RPTs), which are used to quantify changes in capacity, resistance, and power. The RPTs consist of a C1/1 and C1/25 static capacity tests, a low-current hybrid pulse power characterization test, and electrochemical impedance spectroscopy at 60% SOC. Capacity-, power-, and electrochemical impedance spectroscopy-based performance results are reported.

Introduction: Thermal instability of LiPF6 at elevated temperatures is well known. Thus there is an imminent need to develop thermally stable salt to replace LiPF6. A variety of new salts have been studied in Li-ion batteries without much success. One of the salts we are currently evaluating is LiF. This is a low molecular weight salt that has high voltage and thermal stability. However, it doesn’t dissolve in common battery solvents. This limitation can be solved with the addition of anion binding agents (ABAs). For example, Mehta and Fujinami1 proposed boroxine (tri coordinate boron ring) which is a Lewis Acid, to solubilize salt in common battery electrolyte. This compound traps the anion of the salt and frees up the Li+ for conduction. A similar idea was developed by McBreen etal 2-4 where they synthesized boron containing additives to dissolve LiF in organic solvents. However, most of the ABAs investigated are heavy with molecular weight exceeding 500 g/mole. Bulky ABAs impede access of the anion of the salt to the binding site (boron center) thus leading to poor dissolution of the salt5, 6. Inspired by McBreen’s approach we launched a program and developed lower molecular weight (~250 g/mole) ABAs. We quickly learned that the as prepared ABAs do have DMSO solvent bound to the boron (this is designated as impure ABA). We purified the impure ABA by reacting with LiF in acetone which yielded a pure ABA+LiF salt. The purified salt gave a clear solution whereas the impure ABA+LiF salt gave a cloudy solution in EC:EMC (Photo-1). We have prepared a variety of ABAs and tested them for electrochemical performance. Most of them could solubilize LiF only poorly and showed very low conductivity. However, two of the ABAs are promising and the results on one of them will be described here. Electrochemical Performance: Figure 2 compares conductivity for the pure and impure Oxalic-ABA+LiF in EC:EMC We have prepared a variety of ABAs and tested them for electrochemical performance. Most of them could solubilize LiF only poorly and showed very low conductivity. However, two of the ABAs are promising and the results on one of them will be described here. Electrochemical Performance: Figure 2 compares conductivity for the pure and impure Oxalic-ABA+LiF in EC:EMC solvent blend at different temperatures. The pure ABA shows higher conductivity than the impure. Electrochemical measurements and cell building was performed with the purified salt. This showed stability up to 4.5V vs. Li+/Li but unstable close to Li voltage, which was eliminated by adding low concentration of LiPF6 to the solution. Figure 3 shows formation of an 18650 cell containing carbon anode, NMC (523) cathode in EC:EMC(3:7 w%)-ABA+LiF_10mM LiPF6. The reversible capacity is ~ 1.03Ahrs.

Li-ion batteries have generated widespread interest as batteries of choice for a number of applications, because of their attractive energy and power densities. However, the performance at sub-ambient temperatures is very limited due to the increase in cell impedance at sub-ambient temperatures. AC impedance measurements seem to indicate that the ohmic resistance-which includes electrolyte resistance, electrodes bulk resistance, current collector and tab resistances of the cells increases only marginally at sub-ambient temperatures. Most of the increase in cell impedance comes from the electrode/electrolyte interfacial impedance. Sandia National Laboratories and others have shown that the increase in impedance comes mostly from the cathode electrolyte interface and not from the anode electrolyte interface. The question then is "where exactly does the increase in cathode impedance come from?" Whether it comes from the interfacial resistance (charge transfer+the solid-electrolyte-interphase resistance) or from diffusion of Li/sup +/ inside the cathode. To answer the question, we are investigating the impedance characteristics of the LiCoO/sub 2/ and other metal oxide cathodes at different temperatures and cell voltage. The data seem to indicate that the interfacial impedance of the cathode/electrolyte interface mostly accounts for the cell impedance.

Advanced rechargeable lithium-ion batteries are presently being developed and commercialized worldwide for use in consumer electronics, military and space applications. The motivation behind these efforts involves, among other things, a favorable combination of energy and power density. For some of the applications the power sources may need to perform at a reasonable rate at subambient temperatures. Given the nature of the lithium-ion cell chemistry the low temperature performance of the cells may not be very good. At Sandia National Laboratories, we have used different electrochemical techniques such as impedance and charge/discharge at ambient and subambient temperatures to probe the various electrochemical processes that are occurring in Li-ion cells. The purpose of this study is to identify the component that reduces the cell performance at subambient temperatures. We carried out 3-electrode impedance measurements on the cells which allowed us to measure the anode and cathode impedances separately. Our impedance data suggests that while the variation in the electrolyte resistance between room temperature and -20"C is negligible, the cathode electrolyte interracial resistance increases substantially in the same temperature span. We believe that the slow interracial charge transfer kinetics at the cathode electrolyte may be responsible for the increase in cell impedance and poor cell performance.

Thermal instabilities were identified in SONY-type lithium-ion cells and correlated with interactions of cell constituents and reaction products. Three temperature regions of interaction were identified and associated with the state of charge (degree of Li intercalation) of the cell. Anodes were shown to undergo exothermic reactions as low as 100°C involving the solid electrolyte interface (SEI) layer and the LiPF6 salt in the electrolyte (EC:PC:DEC/LiPF6). These reactions could account for the thermal runaway observed in these cells beginning at 100°C. Exothermic reactions were also observed in the 200°C-300°C region between the intercalated lithium anodes, the LiPF6 salt, and the PVDF. These reactions were followed by a hightemperature reaction region, 300°C-400°C, also involving the PVDF binder and the intercalated lithium anodes. The solvent was not directly involved in these reactions but served as a moderator and transport medium. Cathode exothermic reactions with the PVDF binder were observed above 200°C and increased with the state of charge (decreasing Li content). This offers an explanation for the observed lower thermal runaway temperatures for charged cells.

Advanced rechargeable lithium-ion batteries are presently being developed and commercialized worldwide for use in consumer electronics, military and space applications. At Sandia National Laboratories, the authors have used different electrochemical techniques such as impedance and charge/discharge at ambient and subambient temperatures to probe the various electrochemical processes that are occurring in Li-ion cell. The purpose of this study is to identify the component that reduces the cell performance at subambient temperatures. Their impedance data suggest that while the variation in the electrolyte resistance between room temperature and -20/spl deg/C is negligible the anode electrolyte interfacial resistance increases by an order of magnitude in the same temperature regime. They believe that the solid electrolyte interface (SEI) layer on the carbon anode may be responsible for the increase in cell impedance. They have also evaluated the cells in hybrid mode with capacitors. High-current operation in the hybrid mode allowed full usage of the Li-ion cell capacity at 25/spl deg/C and showed a factor of 5 improvement in delivered capacity at -20/spl deg/C.

not Available.

not Available.

not Available.

not Available.

not Available.