Thermal stability analysis using machine learning by integrating biochar-based phase change materials and graphene for thermal energy storage applications

Recently, great interests have been grown for use of biochar-based phase change material in the storage systems to improve the performance and thermal stability. Nevertheless, enhancement on the thermal degradation resistance and the flame-retardant properties of these materials is still a challenging task. This work reports the thermal properties (thermal stability and flammability behaviour) of polyethylene glycol (PEG1000), biochar-based PCM, as well as nano-enhanced biochar based PCM containing carbon-based nanoparticles i.e., graphene nano particles in a systematic manner using thermogravimetric analysis and flammability test. The activation energy of the synthesized samples is calculated based on model-free kinetic models, including Flynn–Wall–Ozawa, Kissinger–Akahira–Sunose, and Starink kinetic models. The results show that biochar addition greatly improves the thermal stability, and the average activation energy is increased by about 40 % when compared with pure PCM. The flammability study demonstrated a significant decline in burning rate and delayed ignition behaviour for nano-enhanced biochar PCM, which suggested that the fire resistance is enhanced compared to PEG1000. Machine learning models are used to predict the activation energy in which polynomial regression achieved a R² value of 0.98 and root mean square error (RMSE) of 0.58 kJ/mol, indicating excellent agreement between experimental and predicted degradation behaviour. As a whole, the present study demonstrates that nano-enhanced biochar PCM displays excellent thermal stability, low flammability and reliable prediction accuracy, suggesting its great potential for safe and efficient TES applications.