Green synthesis optimization with artificial intelligence studies of copper–gallic acid metal–organic framework and its application in dye removal from wastewater

Abstract

The subfield of artificial intelligence (AI) known as machine learning promises to advanced science and technology by enabling better quality, performance, and predictive capabilities. In this research, a new method was proposed to extend the application of AI to optimize the green synthesis of copper-gallic acid metal–organic framework (CuGA MOF) and explore the complex hyperdimensional relationship between the synthesis parameters and the resulting product qualities. The deep learning (DL) method has shown higher accuracy, with a coefficient of determination (R2) of 0.881. Various synthesis conditions, including the molar ratio of NaOH to GA, temperature, and reaction time, were conducted to obtain the optimal yield and crystallinity percentage. The Particle Swarm Optimization (PSO) algorithm suggests inputs of 1.8, 108 °C, and 1.5 h for parameters of the molar ratio of NaOH to GA, temperature, and reaction time, respectively. The optimum results indicate a yield percentage of 60.61 % and a crystallinity percentage of 59.2 %. Furthermore, CuGA MOF was used for basic red 9 (BR9) dye removal from aqueous solutions. The adsorption capacity of this MOF for the removal of BR9 was high, up to 115.08 mg/g. The performance of CuGA 90–2.2 remained high (>90 %) even after 3 adsorption–desorption cycles, making it a promising reusable adsorbent. In addition, the pre-adsorption and post-adsorption of CuGA MOF were characterized by various analytical techniques, including FTIR, XRD, SEM, EDS, and XPS.