Promising Routes to a High Li+ Transference ... - ACS Publications

Oct 6, 2017 - Eric J. McShane is currently a second year student pursuing his Ph.D. in .... WangVenkatasubramanian ViswanathanJay F. WhitacreKrzysztof...
0 downloads 0 Views 3MB Size
http://pubs.acs.org/journal/aelccp

Promising Routes to a High Li+ Transference Number Electrolyte for Lithium Ion Batteries Kyle M. Diederichsen, Eric J. McShane, and Bryan D. McCloskey* Department of Chemical and Biomolecular Engineering, University of California, Berkeley, California 94720, United States Energy Storage and Distributed Resources Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, United States S Supporting Information *

ABSTRACT: The continued search for routes to improve the power and energy density of lithium ion batteries for electric vehicles and consumer electronics has resulted in significant innovation in all cell components, particularly in electrode materials design. In this Review, we highlight an often less noted route to improving energy density: increasing the Li+ transference number of the electrolyte. Turning to Newman’s original lithium ion battery models, we demonstrate that electrolytes with modestly higher Li+ transference numbers compared to traditional carbonatebased liquid electrolytes would allow higher power densities and enable faster charging (e.g., >2C), even if their conductivity was substantially lower than that of conventional electrolytes. Most current research in high transference number electrolytes (HTNEs) focuses on ceramic electrolytes, polymer electrolytes, and ionomer membranes filled with nonaqueous solvents. We highlight a number of the challenges limiting current HTNE systems and suggest additional work on promising new HTNE systems, such as “solvent-in-salt” electrolytes, perfluorinated solvent electrolytes, nonaqueous polyelectrolyte solutions, and solutions containing anion-decorated nanoparticles.

C

Li alloys (Si and Sn being the most well-studied) or pure Li metal are being explored as high-capacity anode materials, and the so-called Li-rich transition metal oxides are being pursued as cathode materials.7−11 The focus of this Review is another, sometimes overlooked, strategy to improve energy density and charging rates of batteries, namely, the development of high Li+ transference number electrolytes (HTNEs), those in which the ionic current is carried predominantly by the Li+ rather than its counterion. The definition of the Li+ transference number in the dilute limit for a binary salt electrolyte in which both ions are univalent (a 1:1 electrolyte) relates the diffusion of Li+ and its counterion through the following simple relationship

onsumer lithium ion battery powered devices are ubiquitous in modern society, and significant growth in the electrification of vehicles, powered by advances in lithium ion batteries, has occurred in the last several years. Plugin hybrid (PHEV) and pure electric vehicles (EVs) accounted for 0.9% of total cars purchased globally in 2016, and sales of all plug-in vehicles increased 550% from 2012 to 2016, with nearly 780 000 PHEV/EV sales (160 000 in the United States) in 2016.1 However, current consumer electronics and EVs still suffer from a few major challenges related to their batteries: they are larger and heavier than desired, they take a long time to charge, and safety incidents continue to attract significant media attention. For lithium ion batteries to continue their impressive market penetration over the next few decades, breakthroughs in battery technology are required that continue to directly address these issues. For this reason, significant research has targeted increased energy density, higher charge and discharge rates, and improved safety through various advances in every component of the Liion battery.2,3 Electrolyte mixtures and separators have been identified that enable reasonably safe, long-term use,4,5 and cell designs and energy management technologies have enabled impressive gains in cycle life, safety, and energy density.6 Recent work has particularly emphasized improving the energy density of lithium ion electrode materials. As relevant examples, © 2017 American Chemical Society

t+ =

D+ D− + D+

(1)

where t+ is the Li+ transference number, D+ is the Li+ diffusion coefficient, and D− is the anion diffusion coefficient.12 In this limit, the transference number can be considered as simply the fraction of the total ionic conductivity that is carried by Li+. Received: August 24, 2017 Accepted: October 6, 2017 Published: October 6, 2017 2563

DOI: 10.1021/acsenergylett.7b00792 ACS Energy Lett. 2017, 2, 2563−2575

Review

Cite This: ACS Energy Lett. 2017, 2, 2563-2575

ACS Energy Letters

Review

topic.5,14 During the first few discharge−charge cycles, these additives will facilitate the formation of solid, ion-conducting layers on the anode surface that otherwise substantially suppress parasitic reactions between the anode and electrolyte.15 While commercial mixtures of additives have been realized after many years of research, understanding the influence of additives on electrode processes at a molecular level remains a poorly understood topic. Second, although current liquid electrolytes offer high conductivity across a wide temperature range (∼1−10 mS/ cm) and well dissociated ions with a solubility > 1 M Li+, they

Despite numerous advances in porous electrode materials, improvement of electrolyte properties, including stability, ionic conductivity, σ, and t+, remains one of the most important challenges to any further performance gains. Conventional electrolytes are based on a binary lithium salt and stabilizing additives dissolved in a mixture of carbonate liquid solvents.4 These electrolytes are then imbibed in the porous polyethylene or polypropylene separator and electrodes to create ionic contact while maintaining electronic insulation between the electrodes. This class of electrolytes imparts high lithium ion conductivity and has been successfully commercialized for many years but suffers from a couple of key drawbacks; we will briefly discuss the first drawback prior to focusing the remainder of this Review on the second. First, all liquid electrolytes possess a voltage stability window, typically between ∼1 and ∼4.5 V vs Li/Li+, that ultimately limits battery rechargeability, safety, and high-energy active electrode materials development.13 To address capacity fade during battery cycling, in combination with limiting the upper operating voltage to improve stability at the cathode− electrolyte interface, carbonate-based electrolytes typically contain small amounts of organic additives that are usually selected through empirical combinatorial analyses, although effort has been made to develop a deeper understanding of the

It is shown here that even modest improvements in t+, e.g., to t+ ≈ 0.7, would be beneficial, particularly allowing a higher attainable state of charge at high charge rates, where a large, constant current would be necessary to quickly charge the battery. have a Li+ transference number below 0.5, indicating that the majority of the total ionic conductivity is in fact the result of

Figure 1. (A) Schematic of a modeled dual lithium ion insertion cell consisting of a graphite anode, LiCoO2 cathode, and separator. Thicknesses and porosity (ε) are given for each cell component, and the model is explained in the Supporting Information. (B) Red dashed charge curve for standard binary salt liquid electrolyte (t+ = 0.40, σ = 10 mS/cm) and solid black charge curves for HTNE (t+ varies, σ = 6 mS/ cm) at 2C with a 4.2 V cutoff voltage. (C) Attainable SOC versus charge rate for electrolyte with σ = 10 mS/cm and variable t+. (D) Attainable SOC versus charge rate for standard liquid electrolyte (red dashed triangles) and HTNE (solid black squares) at 2C. HTNE σ = 5 mS/cm for this set of simulations. (E) Minimum σ and t+ values required to achieve 75 (red triangles) and 85% (black squares) SOC at 2C with a 4.2 V cutoff voltage. Our baseline binary salt liquid electrolyte (σ = 10 mS/cm and t+ = 0.4) achieved 75% SOC at 2C and a 4.2 V cutoff, as observed in (B). 2564

DOI: 10.1021/acsenergylett.7b00792 ACS Energy Lett. 2017, 2, 2563−2575

ACS Energy Letters

Review

trade-off between σ and t+; therefore, we chose to study charge behavior of cells where t+ of the electrolyte is varied between 0.4 and 1 and σ is varied between 1 and 10 mS/cm (i.e., higher t+ and lower σ than traditional liquid electrolytes). Typical charge profiles (with a voltage cutoff of 4.2 V) are shown in Figure 1B, and the total attainable state of charge (SOC) prior to reaching the cutoff voltage is shown as a function of charge rate for these cells in Figure 1C,D. Although little difference is observed in the attainable SOC at low current densities (