Observing Formation of the Solid Electrolyte Interface By Operando Neutron Reflectivity

Abstract

To realise the widespread implementation of future generations of lithium ion batteries (LIBs) in electric vehicles, significant improvements in the lifetime and power density are required. One of the most important aspects in achieving a long battery lifetime is the formation of an effective solid electrolyte interphase (SEI) layer on the anode. This layer forms mostly within the first few charge/discharge cycles by the decomposition of electrolyte components at the electrode/electrolyte interface. Although the importance of the SEI is well accepted, much dispute remains regarding its chemical composition, thickness, morphology and formation/decomposition mechanisms. This is largely due to a lack of characterisation techniques that can accurately probe the electrode-electrolyte interface with nm-scale surface sensitivity under realistic cycling conditions. As a result, most efforts to characterise the SEI have involved ex situ samples, the preparation of which may alter the SEI layer composition. Therefore, in order to gain the most accurate information, there has been a great drive to perform in situ and operando measurements.Neutron techniques offer several advantages for characterising the processes occurring in LIBs, particularly as the negative scattering length of Lithium gives a high sensitivity to changes in the lithium environment, which is not the case with X-ray techniques. Neutron reflectometry (NR) is one of the few techniques with which the nm-scale structure of buried electrode-electrolyte interfaces in a cell can be observed during electrochemical cycling.1 A number of previous studies have investigated thin film silicon or carbon electrodes, but these have typically been carried out across a limited potential range due to concerns about restructuring due to significant lithiation that could otherwise hinder interpretation of the neutron reflectivity data.2-4 To overcome this limitation we have recently used a nickel working electrode (anode) which, given its zero capacity for lithium insertion, does not alter in structure during electrochemical cycling. Here we directly observe the growth of the SEI layer on a nickel anode as a function of potential (2.00-0.05 V) during the first cycle by using our operando cell developed at ISIS Neutron and Muon Source. During the first lithiation cycle (decreasing potential), we observe the growth of a lithium-rich layer directly in contact with the nickel and another, more organic, layer on top of this lithium-rich layer. We are able to track the evolution of these layers with potential, and characterise their chemical composition using additional X-ray Photoelectron Spectroscopy measurements performed on both cycled nickel and commercial graphite electrodes for comparison.