Anodic Oxidation of Conductive Carbon and Ethylene Carbonate in High-Voltage Li-Ion Batteries Quantified by On-Line Electrochemical Mass Spectrometry

Abstract

The anodic oxidation stability of battery components like the conductive carbon black (Super C65) and the co-solvent ethylene carbonate (EC) is of great relevance, especially with regards to high-voltage cathode materials. In this study, we use On-line Electrochemical Mass Spectrometry (OEMS) to deconvolute the CO and CO 2 evolution from the anodic oxidation of carbon and electrolyte by using a fully 13 C-isotope labeled electrolyte based on ethylene carbonate with 2 M LiClO 4 . We present a newly developed two-compartment cell, which provides a tight seal between anode and cathode compartment via a solid Li + -ion conducting separator, and which thus allows us to examine the effect of trace amounts of water on the anodic oxidation of carbon ( 12 C) and ethylene carbonate ( 13 C) at high potentials (> 4.5 V) and 10 to 60 • C. Moreover, we report on the temperature dependence of the water-driven hydrolysis of ethylene carbonate accompanied by CO 2 evolution. Finally, by quantifying the evolution rates of 12 CO/ 12 CO 2 and 13 CO/ 13 CO 2 at 5.0 V, we demonstrate that the anodic oxidation of carbon and electrolyte can be substantial, especially at high temperature and in the presence of trace water, posing significant challenges for the implementation of 5 V cathode materials. The requirements placed on Li-ion battery technology have changed from powering small portable electronics to applications demanding high energy and high power density, such as hybrid and plug-in electric vehicles.1,2 As a result, high-voltage cathode materials have been developed that raise the cell voltage from 3.7 V in the case of a traditional LiMO 2 (M = Co, Ni, Mn) cathode to 4.8 V in new cathodes such as the high-voltage spinel LiNi 0.5 Mn 1.5 O 4 (LNMO) or LiCoPO 4 (LCP).3-6 The state-of-the-art electrolyte, a mixture of ethylene carbonate (EC) and diethyl carbonate (DEC), dimethyl carbonate (DMC), and/or ethyl methyl carbonate (EMC) with dissolved LiPF 6 salt, tends to decompose on the surface of the delithiated cathode at potentials higher than 4.5 V vs. Li/Li + , especially at high temperature.7-12 Thus, the practical application of these high-voltage materials remains hindered by several obstacles: [13][14][15][16][17][18] (i) the limited anodic stability of electrolyte solvents and salts as well as of binders, (ii) the loss of active Li + -ions, (iii) the corrosion of the aluminum current collector, and (iv) the instability of conductive carbon additives due to anion intercalation and/or carbon oxidation.While the corrosion of conductive carbons is suggested by the frequently observed increase of the electrical resistivity of long-term cycled high-voltage cathodes, 15 quantitative measurement on the anodic decomposition of carbons have only been made in the context of Li-air battery research, making use of isotopically labeled battery components to decouple electrode and electrolyte related CO 2 evolution. For example, Thotiyl et al. used a 13 C carbon cathode in DMSO and tetraglyme-based electrolyte to study CO 2 evolution from...

Gas Evolution at Graphite Anodes Depending on Electrolyte Water Content and SEI Quality Studied by On-Line Electrochemical Mass Spectrometry

“…However, as was discussed previously, 24 additional interfacial resistance roughly doubles the overall electrode polarization.…”

“…This allows electrolyte access and gas diffusion from both sides of the graphite electrode, which is a prerequisite for measurements with the OEMS system. 23,24 Afterwards, the coating was dried overnight in ambient atmosphere on a hot plate held at 60…”

“…15 The mass fragment at m/z = 26 was selected to quantify C 2 H 4 evolution in order to avoid interference from the high ion current background at m/z = 28 and 27 from the EMC solvent. 24 Potentiodynamic formation procedure.-For the OEMS experiments, a potentiodynamic cyclic voltammetry (CV) formation procedure at a scan rate of 0.5 mV/s in a potential window of 0 -1.5 V Li (Series G300 potentiostat, Gamry) was chosen instead of the usual galvanostatic cycling. The former was chosen because: (i) the slow CV procedure allows to accurately resolve the potential region of interest (1.5 -0 V Li ) with a sufficiently high and constant time resolution of gas evolution vs. potential, contrary to a galvanostatic charging procedure, in which the potential of a graphite electrode at practical C-rates would rapidly drop to the lithiation plateau of ≈100 mV Li ; and, (ii) previous measurements revealed a relatively large iR-drop caused by the limited conductivity of the Ohara glass, due to which the cutoff voltage (normally 0.01 V Li for graphite half-cells) in a galvanostatic charge at C/10 (i.e., at 0.21 mA/cm 2 ) would be reached already after 10 -15% of the theoretical capacity.…”

“…The former was chosen because: (i) the slow CV procedure allows to accurately resolve the potential region of interest (1.5 -0 V Li ) with a sufficiently high and constant time resolution of gas evolution vs. potential, contrary to a galvanostatic charging procedure, in which the potential of a graphite electrode at practical C-rates would rapidly drop to the lithiation plateau of ≈100 mV Li ; and, (ii) previous measurements revealed a relatively large iR-drop caused by the limited conductivity of the Ohara glass, due to which the cutoff voltage (normally 0.01 V Li for graphite half-cells) in a galvanostatic charge at C/10 (i.e., at 0.21 mA/cm 2 ) would be reached already after 10 -15% of the theoretical capacity. 15,24 The purely iR-based voltage drop can be estimated by the areal resistance of the Ohara glass ≈190 cm 2 from the ionic conductivity of 0.8 · 10 −4 S/cm at a thickness of 150 μm;…”

Measuring Oxygen Release from Delithiated LiNixMnyCo1-x-yO2and Its Effects on the Performance of High Voltage Li-Ion Cells

“…[51][52][53] These various studies suggest that the oxidation of electrolyte is a major issue that increases significantly with cell voltage. Self et al recently noted that decomposition products from a sulfur-containing additive (prop-1-ene-1,3-sultone) and the organic carbonate solvent oxidation have been observed at lower electrode potentials (i.e., ∼4.2 -4.7 V vs. Li/Li + ) 42 than expected from voltammetric studies (>5.3 V vs. Li/Li + ) [54][55][56][57][58] and density functional theory (DFT) calculations (> 6.0 V vs. Li/Li + ). 56,59 However it is noted that there is not universal agreement for the onset potential for electrochemical oxidation of organic carbonates (e.g., Moshkovich et al reported oxidation at > 3.9 V vs. Li/Li + ).…”