A machine learning approach for estimating supercapacitor performance of graphene oxide nano-ring based electrode materials

Abstract

This work utilizes a novel approach leveraging the machine learning (ML) technique to predict the electrochemical supercapacitor performance of graphene oxide nano-rings (GONs) as electrode nanomaterials. Initially, the experimental procedure was carried out to synthesize GO via a modified Hummers method, followed by GONs preparation using the water-in-oil (W/O) emulsion technique. High-resolution transmission electron microscopy (HRTEM) analysis reveals the formation of a typical two-dimensional GO nanosheet and multilayer-GO nano-rings. The X-ray diffraction (XRD), Raman spectroscopy, X-ray photoelectron spectroscopy (XPS), and Brunauer–Emmett–Teller (BET) analysis results show that the GONs possess similar structural and surface chemistry properties as of GO, with a slight reduction in oxygenous functionalities, enhancing the capacitive behaviours through facile electron migration at the electrode surface. The electrochemical assessment of GO and GONs samples indicates outstanding specific capacitances of 164 F g⁻¹ and 294 F g⁻¹ at 1 mV s⁻¹, showcasing capacitive retention of up to 63% and 60% after 2500 cycles. In addition, four different machine learning models were tested to estimate the role of electrochemical parameters in determining the specific capacitance of GONs.