Artificial Intelligence Meets Catalysis: A New Approach to Heavy Diesel Desulfurization Using Trimetallic Activated Carbon Catalyst in Central Oscillating Reactor

To meet the new strict environmental legislations about sulfur content in petroleum fuels and their harmful emissions for high quality fuels, deep desulfurization strategies, such as oxidation process, have become an interesting topic in the academic and industrial fields. Therefore, this work develops a new Trimetallic Activated Carbon (TAC) catalyst, which is synthesized by decorating activated carbon with magnetic-manganese active oxides and an alumina coating film, for continuous deep oxidative desulfurization (ODS). Furthermore, a novel central oscillating reactor (COR) is fabricated by developing central baskets, which are employed to pack catalyst particles for continuous oxidative desulfurization. This new design exhibits a practical solution for handling the solid catalytic materials in central basket baffles for continuous operation, compared to previous studies that utilized only a central baffle with dispersed catalyst particles in the reacting fluids through batch operation mode. Process efficiency is examined utilizing hydrogen peroxide oxidant under mild conditions: 1 atm, temperature (30–90) °C, liquid hour space velocity, LHSV (0.33–0.08) min-1, oscillation parameters (amplitude, A: 3–12 mm, and frequency, f = 0.5–2 Hz). Also, Support Vector Machine (SVM) model is examined as a new machine learning tool to predict the desulfurization model. ODS technology shows high performance through low oxidation time (12 min) by reducing sulfur in heavy diesel fuel from 8281 ppm to 294 ppm to achieve 96.45 % oxidation efficiency under 90 °C, LHSV = 0.08 min-1, A = 12 mm and f = 2 Hz. SVM model data performed excellent prediction at R2 of 0.9962, mean absolute error (MAE) of 0.0791, and mean squared error (MSE) of 0.0078. The SVM strategy results in a high-accuracy artificial intelligence model under minimal deviations between actual and predicted data. The new integrated COR-TAC system provides an efficient, practical approach to deep, cost-effective and eco-friendly oxidative desulfurization; it can be directly scaled up into the refining industry at the same process performance.