Mechanical effects of carboxymethylcellulose binder in hard carbon electrodes

TL;DR

This study introduces a low-cost indentation test to measure the elasticity of hard carbon electrodes with varying amounts of carboxymethylcellulose binder, linking mechanical properties to electrochemical performance in sodium-ion batteries.

Contribution

It presents a practical method to evaluate binder elasticity and demonstrates how binder content influences electrode mechanical stability and capacity retention.

Findings

Elastic moduli increase with binder content.

Higher binder content improves capacity retention.

Mechanical testing correlates with electrochemical performance.

Abstract

Electrodes in sodium-ion batteries endure mechanical stress during production and application, which can damage these fragile coatings, causing performance inefficiencies and early failure. Binder material provides elasticity in electrode composites to resist fracture, but evaluating the effectiveness of binder is complicated by substrate dependency of these films, while conventional cell tests are beset by multiple electrochemical variables. This work introduces a practical low-cost indentation test to determine the elasticity of hard carbon electrodes containing standard carboxymethylcellulose binder. Using the proposed method, relative elastic moduli of hard carbon electrodes were found to be 0.079 GPa (1% binder), 0.088 GPa (2% binder), 0.105 GPa (3% binder) and 0.113 GPa (4% binder), which were validated using a computational model of film deflection to predict mechanical deformation under stress. Effects on the electrochemical performance of hard carbon anodes were also demonstrated with impedance spectroscopy and galvanostatic cycling of sodium half-cells, revealing 8-9% higher capacity retention of anodes with 4% binder compared with those containing 1% binder. These findings suggest binder content in hard carbon electrodes should be selected according to requirements for both cycle life and film flexibility during cell manufacturing.