Steven C. DeCaluwe, Jeanette E. Owejan, Jon P. Owejan, Joseph A. Dura

Lithium-ion batteries boast the highest energy density and power density of any working electrical storage device and thus have the potential to revolutionize clean energy storage technologies, ranging in scale from personal electronics to automotive to smart-grid applications.  Today’s lithium-ion batteries are oversized and degrade rapidly due to a number of known mechanisms that lead to irreversible capacity loss, one of which is passive film formation via the Solid Electrolyte Interphase (SEI).  Despite many efforts to engineer SEI films that mitigate capacity fade and are less prone to degradation, direct study of the SEI is made difficult by the scarcity of in situ techniques suitable for observing these layers in operating batteries.   As such, most information regarding the nature of the SEI has been gained from ex situ, post-mortem, and indirect measurements.  We present here the use of in situ neutron reflectometry (NR) to measure the SEI that occurs over a Cu working electrode in a lithium half-cell.  NR is a technique that determines the depth profile of the scattering length density (SLD, which is related to composition) by fitting the intensity of reflected neutrons as a function of grazing angle from the surface.  NR provides several distinct advantages in that it is performed in situ, is not destructive in nature, and provides sub-angstrom resolution of features thicker than 1.5 nm. 

NR data was taken of the virgin Cu cathode in the electrolyte-filled cell at open circuit voltage.  Subsequently an SEI was produced by running cyclic voltammograms followed by potentiostatic holds at eight separate reducing and oxidizing potentials, during which NR data sets were taken.  Each had excellent fits, which clearly show how the SEI changes in both composition and thickness as a function of potentiostatic holds.  These observations will be presented in detail, as well as a more detailed investigation of SEI chemical composition by comparison of model SEI compositions (informed by post-mortem XPS) with the simultaneously collected NR and electrochemical data. Results represent the first NR measurements of critical SEI properties – such as thickness, porosity, layered structures and gradients, and chemical composition – on an operating Li battery, without confounding experimental artifacts associated with ex situ techniques.  Future implications of this study include improvements in Li-ion battery transport modeling and the capability to directly and quantitatively study SEI properties as a function of electrolyte composition (including additives), temperature, voltage, current, and cycling/time, etc.  Such studies will enable systematic engineering of the SEI properties, leading to future improvements for commercial device performance, affordability, and sustainability.