Optimization of Production Process of Agro-Based Activated Carbon for Sustainable Cooling

This study investigates the experimental production and multi-objective optimization of the production process of microporous activated carbon (AC) derived from Raphia nut endocarp (RNE) for solid adsorption refrigerators (SAR) and related systems. Using phosphoric acid (H₃PO₄) and calcium chloride (CaCl₂) as activating agents, a comprehensive optimization technique integrating genetic algorithms (GA), Pareto optimality (PO), min-max normalization (MMN), and machine learning (ML) was applied to determine the optimal RNE-derived AC (RNEAC) properties, including surface area, carbon yield, and ash content. Linear regression was used as the ML algorithm to analyze the relationship between the production variables of the RNEAC. The experimental design varied parameters such as carbonization temperature, residence time, activating agent concentration, and impregnation ratio, to statistically evaluate their impacts. Results revealed that temperature and residence time significantly influence ash content, while impregnation ratio, temperature, and residence time optimize surface area. Similarly, carbon yield was affected by temperature, residence time, and impregnation ratio. The optimized RNEAC exhibits properties that include a high surface area, low ash content, and a promising methanol adsorption capacity, highlighting its suitability for application in SAR systems. This study will contribute to achieving a greener environment and the development of high-efficiency and sustainable adsorption cooling technologies for rural communities by valorizing agricultural wastes and turning them into excellent low-cost adsorbents for the next generation of SAR.