This study explores low molecular weight polymers with the same chemical repeat structure as high molecular weight poly acrylic acid (PAA) and lithium poly acrylic acid (LiPAA) binders which act as a dispersant to improve the silicon-graphite electrode performance. The electrodes which utilize LiPAA as a dispersant perform, on average, the best with respect to the maximum capacity (compared to theoretical), rate performance, and capacity retention with time. These electrodes also have the poorest dispersion of binder, indicating that the poor binder dispersion is essential to electrode performance. In contrast, this study shows that electrodes formulated using PAA had good dispersion and performs the worst in cycling. Finally, this work demonstrates that the binder is not uniformly distributed in the electrode, but rather resides in local regions. The results indicate these regions accommodate volume expansion during cycling.

As lithium-ion battery (LIB) active material and cell manufacturing costs continue to drop with wider adoption of electric vehicles, electrode and cell processing costs remain too high in terms of reaching the ultimate U.S. Department of Energy (DOE) cell cost target of $80/kWh. This paper primarily covers major materials chemistry advancements made over the last 10 years at Oak Ridge National Laboratory (ORNL) in the space of advanced manufacturing science for LIBs with the aim of simultaneously meeting the ultimate cost target, 500 Wh/kg gravimetric energy density, 10-15-min fast charge times, and 1000 deep discharge cycles. Aqueous electrode processing with a variety of active anode and cathode materials is now a standard procedure at the DOE Battery Manufacturing R&D Facility at ORNL (BMF), including the latest processes developed for Ni-rich cathodes. New results on cobalt-free LiNi0.8Fe0.1Al0.1O2 (NFA 811) are also included and discussed in an electrode processing advantage context. In addition, colloidal processing advancements have been made for Si/C composite anodes for achieving >600 mA h/g capacities. Optimization of electrode coating parameters and drying protocols have been completed, which has elucidated how key processing variables need to be changed when parameters such as slurry solids loading, solvent type, and wet electrode thickness are changed. ORNL has also increased the line speeds at which thick cathodes can be processed using ultra-fast electron beam (EB) curing and substantially decreased formation cycling times to <1 day. Key details of these advancements are discussed in the context of materials chemistry and process-property-performance relationships.

Cobalt content in Li-ion battery cathodes has become a top concern due to its price volatility and limited source availability. Low-cobalt, Ni-rich active materials are promising candidates for next-generation cathodes due to their high capacities, and water-based processing of these materials can further reduce both cost and environmental impact. We systematically evaluated the water compatibility of four different LiNixMn1-x-yCoyO2 (NMC) powders with increasing nickel contents. Comprehensive characterization verified there is no major change to their bulk structures, and only slight surface modifications related to the removal of contaminant species. For the first time, we demonstrate that LiNi0.8Mn0.1Co0.1O2 (NMC 811) cathodes can be formulated in water and cycled 1000 times in full pouch cells with excellent capacity retention (~70% compared to ~76% for NMP-processed cells). When implemented in future battery production lines, aqueous processing of Ni-rich NMC will simultaneously enable cost reductions and higher cell energy densities.

We demonstrate a process to prepare model electrodes of the Ni-rich layered compound Li-Ni(0.6)Mno(0.2)Co(0.2)O(2). These thin-film cathodes are compared with the composite materials to demonstrate the system is a viable platform for isolating interfacial phenomena between the electrolyte and active material without the influence of binders and conductive additives. The appropriate choice of heterolayers was found to influence the preferential orientation of the (101) and (104) planes relative to the (003) plane of the layered R-3m crystal structure, enhancing Li+ diffusion and improving electrochemical performance. The addition of a Co interlayer between the Pt current collecting layer and alumina substrate increased the (101) and (104) texturing of the 500 nm Ni-rich film and allowed cells to deliver greater than 50% of their theoretical capacity. This work provides an architecture for isolating complex mechanisms of active materials that suffer from surface reconstruction and degradation in electrochemical cells.

Gas formation during lithium-ion battery (LIB) cycling impacts the stability and safety of these batteries, especially for those containing Ni-rich NMC cathodes. In this paper, the cycling performance and gassing behavior of NMC811/graphite full cells with 4.2 and 4.4 V upper cutoff voltages were first compared. Cells with a 4.2 V upper cutoff voltage had good cycling stability, exhibiting a capacity retention of 96.8% after 100 cycles and generated little gas. On the other hand, cells with a 4.4 V upper cutoff voltage lost over 25% of initial capacity after 100 cycles and generated large amounts of gas in the first 10 cycles. Electrochemical cycling of anode and cathode symmetric cells was implemented to isolate gases formed at the electrode. Gas chromatography-mass spectrometry, Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy, and scanning transmission electron microscopy were used to characterize the gas formation and associated material surfaces and structural properties. It was found that CO2 and fluorinated alkanes were the dominant gases evolved on the cathode side during cycling to 4.4 V. Gas crossover to the anode led to the depletion of gaseous products, which stabilized the cell performance to some extent. However, the growing surface reconstruction layer at the cathode, the thickening of the solid electrolyte interphase layer at the anode, and the gradual depletion of lithium inventory collectively contributed to the continuous capacity loss of full cells cycled to 4.4 V.

Glyme solvents are a promising avenue of research in novel electrolytes, due to their thermal and chemical stability, large electrochemical window, and tunable properties. This work investigates the transport properties of LiPF6 and NaPF6 salts in Monoglyme (G1), Diglyme (G2), and Tetraglyme (G4). Self-Diffusion coefficients, and degrees of ion association were found through spectroscopic techniques. Raman and Fourier-Transform Infrared (FTIR) spectroscopy exhibited an absorbance peak at 740 cm(-1), corresponding to the a(1g) mode; known to signify the presence of solvent-separated and contact ion pairs. The increased intensity of this peak for G1 compared to G4 suggests that the electrolyte exhibits stronger association with decreasing solvent size. Nuclear Magnetic Resonance Diffusometry measurements yield self-diffusion coefficients for Li-7, Na-23, and F-19 across all samples to be on the order of 10(-9) m(2)s(-1) in G1, to 10(-11) m(2-)s(-1) in G4. Comparison of conductivities calculated from the measured diffusivities and the Nernst-Einstein equation with conductivity measurements deduced from Electrochemical Impedance Spectroscopy (EIS) determined the ion association degree; the electrolytes were shown to exhibit stronger ion pairing with increased temperature which is attributed to a decrease in the dielectric constant of the solvents with increasing temperature. Additionally, the ion association was shown to decrease with increasing solvent molecular size, consistent with FTIR findings. (C) 2019 Elsevier Ltd. All rights reserved.

Sodium-based energy storage systems are promising candidates for electric vehicles and grid-level energy storage applications. The advancement of sodium-based energy storage systems relies on the development of high performance sodium-ion conducting electrolytes and membranes that exhibit high ionic conductivity and mechanical stability. A crosslinked poly(ethylene oxide) based polymer electrolyte was developed that demonstrates high ionic conductivity, as well as excellent mechanical stability over a wide temperature range. Ionic conductivities up to 2.0 x 10(-4) S/cm at 20 degrees C and 7.1 x 10(-4) S/cm at 70 degrees C are achieved for the plasticized membrane, almost four orders of magnitude greater than that of the non-plasticized membrane. The membranes are mechanically robust, and the storage modulus of the membrane is maintained at similar to 1 MPa from -20 to 180 degrees C even with the addition of plasticizer. This study provides a synthesis approach towards the design of highly ion conducting, mechanically robust gel polymer electrolytes for Na-ion batteries, non-aqueous flow batteries, and many other applications.

Understanding the aprotic solution structures at the immediate vicinity of solid/liquid interface (SLI) is critically important for next generation lithium ion battery development. Yet, it is still challenging to investigate the carbonate chemical profiles close to the diffuse layer (about 10 nm) of the electrical double layer at SLI due to the lack of a ultrahigh surface sensitive tool. In this work, we demonstrate the structures of commonly used carbonate solvents (ethylene carbonate (EC) and diethyl carbonate (DEC)) and a carbonate additive (fluoroethylene carbonate (FEC)) in a commercial Li-ion battery electrolyte can be determined at similar to 17 nm above the electrode surface. This is only enabled by a nanogap surface-enhanced Raman spectroscopy (SERS) technique based on a monolayer gold nanoparticle (Au NP) ensemble. The SERS enhancement factor (EF) of those carbonates was found to depend on the molecular polarizability, with the maximum EF at similar to 10(5) found for EC and FEC. Despite their alike chemical structures, this monolayer Au NP SERS substrate is fully capable of discrimiating the different Raman finger prints of EC and FEC. Compared to EC, several vibration modes in FEC, such as C-C skeletal deformation, ring breathing band and C=O stretching band, shift to higher frequencies because of the displacement of a hydrogen atom by a much heavier fluorine atom in a methylene bridge. This counterintuitive observation against the commonly used "ball and spring" model in vibrational spectroscopy is mostly due to the increased bond strength in the FEC ring versus that of EC. A second order empirical polynomial best describes the correlation between the SERS band integration of EC or DEC molar concentration. Our findings open up new opportunities for in-depth understanding of the electrolyte molecular vibrational behaviors at direct solid/liquid interface and developing advanced electrolytes for next generation lithium-ion batteries. (C) The Author(s) 2019. Published by ECS.

We employ tip-enhanced Raman spectroscopy (TERS) to study model amorphous silicon (a-Si) thin film anodes galvanostatically cycled for different numbers. For the 1× cycled a-Si, TERS shows good correlation between solid electrolyte interphase (SEI) topography and chemical mapping, corresponding to distribution of lithium ethylene dicarbonate (LEDC) and poly (ethylene oxide) (PEO)-like oligomer species. Subsequent electrochemical cycling makes the SEI relatively thick and rough with the chemical composition heavily dominated by LEDC monomer-dimer for 5× cycled a-Si. For 20× cycled a-Si, the TERS signal is dominated by carboxylate (RCO2Li) compounds of various conformations and fluorinated species (LixPOyFz). A nanomosaic-multilayer hybrid SEI model on top of the a-Si anode is proposed. The significance of this work is applicable not only to silicon, where SEI plays a dominant role in determining the cycle life performance and reversibility, but also for a number of other relevant battery chemistries such as Na-ion and multivalent redox systems.

High voltage high energy cathode materials for Li-ion battery are highly needed. High nickel content NMC (e.g. NMC811) is believed to be a promising candidate. When raised the upper cutoff voltage of NMC811 cathode to 4.5 V vs. Li+/Li, the cathode materials could deliver a capacity above 215 mAh g-1, which is very attractive to the electric vehicle application. However, raising the upper cutoff voltage will also greatly accelerate the capacity fade of the full cell. Of the various causes proposed by the battery community, a large volume of gases generated at high voltage is believed to be one of the major issues. In this study, we constructed full pouch cells composed of graphite anode, NMC811 cathode and standard GEN2 electrolyte which is 1.2M LiPF6 in EC/EMC (3/7 by weight). We also constructed anode and cathode symmetric pouch cells to investigate the gas evolution and crosstalk at specific sides. Using Archimedes’ Principle, we measured the generated gas volume along cycles within the pouch cells with different cutoff voltages. We found out that the full cells using 4.2 V cutoff voltage barely generated gases along the course of 100 cycles while the ones using 4.4 V cutoff voltage generated a large amount of gases in the first 20 cycles, but the gases would be gradually consumed in the subsequent cycles. By using GC-MS, we determined the main gas species generated in the full cell to be carbon dioxide and alkanes. Moreover, the relative amount of carbon dioxide decreased along cycles while the amount of alkanes would increase. In cathode symmetric cells, CO2 remained the major component after 100 cycles and its signal was much stronger than in full cells. Interestingly, some fluorinated alkanes not seen in the full cells were also identified in cathode symmetric cells, suggesting potential involvement and loss of LiPF6 from electrolyte due to the parasitic side reactions on the cathode. In anode symmetric cells, hydrocarbons attributed to electrolyte decomposition were the main products. Unlike CO2 and fluorinated alkanes formed on the cathode, these species were relatively stable and remained in the full cell after extensive cycling. It was also shown that gas crossover to the anode leads to the depletion of gaseous products during cycling, which could stabilize the cell cycle performance. Although some inactive surface reconstruction layer was formed on the cathode at high voltage operation, impedance buildup and active lithium inventory loss contributed to the significant capacity loss in the full cell. Our study revealed the important role of crossover effect in stabilizing Ni-rich layered oxide cathode, and provided a better understanding on capacity fading to take more effective measures for improving the full cell cycle life.

Ni-rich, or low-Co, layered active materials are promising candidates for next-generation cathodes for lithium-ion batteries. However, these materials present processing and performance challenges such as compatibility with water during aqueous electrode formulation, unoptimized SEI/CEI formation conditions during cell assembly, and unstable capacity fade when cycled to upper cutoff voltages above 4.3 V vs. Li/Li+. The DOE Battery Manufacturing R&D Facility at ORNL (BMF) has recently moved towards low-Co-containing and high-energy LiNi0.8Mn0.1Co0.1O2 (NMC 811) as a new internal cathode baseline and away from LiNi0.5Mn0.3Co0.2O2 NMC 532 for conducting cathode water stability and formation protocol studies. These XRD, Raman spectroscopy, XPS, and TEM studies verified that there was no change in bulk structure of the NMC 811 with long-term water exposure, and only small changes in the surface chemistry. Capacity retention under 0.33C/-0.33C USABC long-term cycling for NMC 811 aqueous-processed single-layer pouch cells was obtained and compared to: 1) the NMP/PVDF processed standard; and 2) NMC 811 exposed to water for 4 hours before processing with the standard NMP/PVDF formulation. The water exposed NMC 811 processed in NMP showed a similar capacity fade to the aqueous processed case. However, all three cases showed excellent capacity retention through 600 cycles, and the aqueous processed cells and NMP processed cells exhibited ~75% and ~80% capacity retention through 1000 cycles, respectively. It was also observed that the differences in capacity fade for all three cases occur within the first ~100 cycles, and the capacity fade slopes were similar from that point on. It is thought that this slight difference in early capacity retention is due to surface chemistry changes of the NMC 811 during aqueous processing and could be remedied with shorter mixing times or a surface protective coating. To decrease the high manufacturing cost associated with long formation times for low-Co cathodes, five different formation protocols were studied using graphite anodes where the total formation time varied from 10 to 86 h. Electrochemical characterization and post mortem analysis showed that longer formation times do not necessarily improve long-term performance while extremely short formation protocols result in lithium plating and poorer electrochemical performance. It was found that the optimum formation protocol is intermediate in length to minimize impedance growth, improve capacity retention, and avoid lithium plating. This presentation will focus on recent ORNL advancements in these areas, where aqueous processing conditions (mixing times, water exposure, binder type, etc.) and fast formation protocols are being developed for NMC 811 and that will be extended to other low-Co and Co-free cathodes.

Redox flow batteries (RFBs) are promising energy storage devices for grid-level applications due to their long cycle life and the ability to independently scale their energy and power densities. The energy density of conventional RFBs is dictated by their capacity (which is directly related to the solubility of the redox species in the electrolyte) and operating potential. Aqueous RFBs generally have low operating potentials ca. 1.5 V, resulting in poor energy densities (25 – 30 Wh/kg for an all vanadium RFB), whereas systems containing organic electrolytes with wider electrochemical windows have moderately higher energy densities. Our team recently demonstrated proof-of-concept for a revolutionary approach which uses mediated electrochemical reactions in a RFB configuration to drive reversible Na storage in a red P anode. Extremely high capacities ~1,000 mAh/g P have been demonstrated using this method. In this configuration, the anion radical mediators are recycled several times throughout the cell stack during a single charge/discharge cycle, effectively decoupling the RFB’s energy density from the redox species’ solubility in the electrolyte. By pairing this mediated red P anode with a sulfide-based cathode, energy densities up to 200 Wh/kg (~10x that of conventional RFBs) can potentially be achieved. This presentation will describe our recent progress developing polymer membranes and high energy density cathodes/catholytes for RFBs. The preparation and characterization of ionically conductive, mechanically robust poly(ethylene oxide) (PEO)-based membranes which are chemically resistant to and prevent crossover of the radical mediators will be discussed. The synthesis and electrochemical properties of a new class of high energy density sodium thiophosphate cathodes for RFBs will also be provided. Acknowledgements This research is supported by Dr. Imre Gyuk, Manager, Energy Storage Program, Office of Electricity Delivery and Reliability, U.S. Department of Energy and the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U.S. Department of Energy.

High-voltage Ni-rich cathodes have been studied as a possible way to achieve high energy density in Li-ion batteries. However, capacity fade due to structural changes at high voltages has limited their applications. This study identifies 4.5 V (vs. graphite) as the optimum upper cutoff voltage (UCV) for a Ni-rich NCA cathode [LiNi0.8Co0.15Al0.05O2]. At this UCV, NCA delivers a 12 % increase in reversible capacity (when discharged to 2.5 V) and retains 92 % of its initial capacity after 100 cycles at 1C/-1C cycling when compared to 4.2 V as UCV. By increasing UCV to 4.7 V, the discharge capacity can be raised to >200 mAh/g. However, the rate of capacity fade is greater when compared to 4.5 V as UCV. This increased rate of capacity fade, at higher UCV, is related to irreversible lattice contractions that leads to structural rearrangement at charged states during high-voltage cycling. Our results show a change in transition metal oxidation states and an onset of structural ordering occurs when the UCV is 4.7 V.

Recent achievements in high-energy batteries have been made by using Ni-rich NMC cathodes (LiNixMnyCo1-x-yO2 with x > 0.5) in conjunction with higher cell voltages. However, these gains have come at a cost of fast capacity fade and poor rate performace. In our previous study, we showed that Al2O3 ALD coatings on LiNi0.8Mn0.1Co0.1O2 (NMC811) and LiNi0.8Co0.15Al0.05O2 (NCA) cathodes prevented surface phase transitions, reduced impedance, and extended cycle life in high voltage cells. Here, neutron diffraction (ND), X-ray photoelectron spectroscopy (XPS), and electron energy loss spectroscopy (EELS) are used to fully investigate the mechanism by which ALD surface coatings mitigate NMC811 cathode degredation. Refinement of ND patterns indicated no changes in the bulk crystal structure of cycled cathodes-with or without the Al2O3 coating. Rather, the improved performance of ALD-coated cathodes is clearly due to surface stabilization. EELS established that all three transition metal oxidation states were reduced at the surface of the uncoated cathode after cycling, whereas the coated cathode showed no changes in surface oxidation states relative to the bulk. The surface coatings also prevented transition metal dissolution and crossover. XPS analysis of the anode harvested from cycled cells with uncoated cathodes showed significant amounts of Mn deposited within the SEI. In contrast, no Mn could be detected on the anodes cycled with coated cathodes. These results affirm that ALD coatings can effectively reduce the reactivity of the NMC surface and prevent detrimental side reactions that shorten battery cycle life.

The Cover Feature illustrates extended-battery electric vehicle (EV) driving range achieved through deep understanding of degradation mechanisms of lithium-ion cathodes charged to high voltage. More information can be found in the Article by L. David et al. on page 5571 in Issue 22, 2019 (DOI: 10.1002/celc.201901338).

An innovative approach to improve the energy density of redox flow batteries (RFBs) through electrochemically mediated reactions is demonstrated. Soluble anion radical species (biphenyl and pyrene) mediate reversible sodium storage in a red phosphorus (P) anode located in an external packed bed reactor in the absence of binders or conductive additives. Because the anion radical species can be recycled several times through the cell stack during a single charge/discharge cycle, a mediated RFB effectively decouples the battery's energy density from the redox species' solubility in the electrolyte. The present study demonstrates the highest capacity reported to date for a mediated anode (up to 800 mAh/g p via reversible formation of NaxP). This approach represents a paradigm shift in the field and can conceivably achieve energy densities exceeding 200 Wh/kg, which is similar to 10 times greater than that of conventional RFBs (e.g., 25 Wh/kg for an aqueous vanadium system).