In Situ Multilength-Scale Tracking of Dimensional and Viscoelastic Changes in Composite Battery Electrodes

Abstract

Intercalation-induced dimensional changes in a composite battery electrode (comprising a polymeric binder) are one of the major factors limiting electrode cycling performance. Since electrode performance is expressed by the quantities averaged over its entire surface area (e.g., capacity retention, Faradaic efficiency, rate capability), significant efforts have been made to develop a methodology allowing its facile mechanical diagnostics at the same areal scale. Herein we introduce such a generic methodology for a highly sensitive in situ monitoring of intrinsic mechanical properties of composite battery electrodes. The gravimetric, dimensional, viscoelastic, and adhesive changes in the composite electrodes caused by Li-ions intercalation are assessed noninvasively and in real time by electrochemical quartz-crystal microbalance with dissipation monitoring (EQCM-D). Multiharmonic acoustic waves generated by EQCM-D penetrate into thin porous electrodes comprising either rigid or a soft binder resulting in frequency and dissipation changes quantified by analytical acoustic load impedance models. As a first demonstration, we used a composite LiFePO (LFP) electrode containing either polyvinylidene dichloride (PVdF) or Na carboximethyl cellulose (NaCMC) as rigid and viscoelastic binders, respectively, in aqueous electrolytes. The intercalation-induced volume changes of LFP electrode were evaluated from a hydrodynamic correction to the mass effect of the intercalated ions for PVdF, and both components of the effective complex shear modulus (i.e., storage and loss moduli) in case of NaCMC binder have been extracted. The sliding friction coefficients for large particles bound at their bottom to the quartz crystal surface (a measure of the adhesion strength of binders) has also been evaluated. Tracking the mechanical properties of the composite electrodes in different environments and charging/cycling conditions in a self-consistent manner provides all necessary conditions for an optimal selection of the polymeric binders resistant to intercalation-induced volume changes of intercalation particles.

In Situ Acoustic Diagnostics of Particle-Binder Interactions in Battery Electrodes

“…Previously, we described in detail how this analysis is performed. 22,23 For the electrodes used in this work, the experimental frequency and dissipation changes of the electrode measured in air are presented in Figure S2, whereas the extracted model viscoelastic parameters are indicated in Table S1 (see also comment following Figure S2).…”

“…The details of this simple test have been reported previously. 22,23 The data for neat PVdF films in contact with 2 M aqueous solutions of Li 2 SO 4 and 1 M solutions of LiPF 6 and NaPF 6 /EC + DMC (with related comments) are presented in Figure S3. The major conclusion is that PVdF binder is stiff in both aqueous solutions, but swells and softens in the aprotic solutions we used (swelling and softening of the binder is easily recognized by this EQCM-D test).…”

“…The major conclusion is that PVdF binder is stiff in both aqueous solutions, but swells and softens in the aprotic solutions we used (swelling and softening of the binder is easily recognized by this EQCM-D test). 18,22 Single-Layer (Thin) Electrodes in Aqueous and Non-aqueous Solutions The behavior of single-layer (further termed thin) electrodes with a loading mass $35 mg/cm 2 is remarkably simple in all the four relevant electrolyte solutions used. Cyclic voltammetry (CV) curves for both LFP and NFP electrodes in aqueous and aprotic solutions (Figures 3A and 3D, respectively) are strictly quasi-reversible.…”

“…The higher harmonics, as expected, are more sensitive to the increase in the effective electrode softness. 15 The result of fitting of the two-layer viscoelastic model to the experimental shifts of Df/n and DD caused by the insertion of Na ions during the first, fifth, and 50 th cycles (the fitting routine was considered in full detail previously 22,23 ) is presented in Figure S5. The retrieved viscoelastic parameters are listed in Table S2.…”

“…Viscoelastic modeling for the contact of electrodes with air and with liquids was performed following our previously published papers and literature data. 22,23 The relevant necessary equations were implemented in the commercial software Qtools (3.1.25.604 from Biolin Scientific, Sweden) supplied together with an EQCM-D instrument. By fitting the viscoelastic models to the experimental DD and Df/n changes, one can evaluate the solid-layer density d s , the thickness h s , the elastic shear storage modulus G 0 (sometimes designated as m, especially for a quartz plate), and the shear loss modulus G 00 , which is linked to the solid shear viscosity, h S = G 00 /2pnf o .…”