ARTIFICIAL NEURAL NETWORK BASED ANALYSIS OF OPERATING PARAMETERS IN METHYL ORANGE PHOTODEGRADATION USING TiO₂ As A PHOTOCATALYST

Abstract

Important operating parameters such dye concentration, light intensity, pH, and reaction duration have a big impact on photodegradation efficiency. An artificial neural network (ANN) model was used in this investigation to examine how these characteristics affected the degradation of methyl orange. An artificial neural network (ANN) was trained on 141 data points, optimizing the distribution of 33 neurons. With a mean squared error (MSE) value of 0.001438, the proposed model produced predictions that were correct. 3D surface plots were used to assess the fractional conversion of methyl orange and show how reaction time and crucial operating parameters relate to each other. According to the findings, fractional conversion falls as dye concentration rises, with lower MO concentrations (20–40 μM) showing the highest efficiency across longer time periods (100–160 minutes). While lower intensities by themselves do not achieve high degradation efficiency, higher light intensities (>80 mW/cm2) greatly increase conversion rates. Furthermore, pH has a significant impact on degradation performance; environments that are severely acidic (pH<5) or basic (pH>10) decrease efficiency, whereas neutral to slightly basic circumstances (pH 7–9) encourage the highest rates. The ANN model proved to be highly reliable in predicting deterioration trends, with a strong regression coefficient (R²) of 0.98 and a relative error of less than 10%. The simulation results also highlighted how important it is to optimize the operating parameters, as this plays a key role in accurately controlling the efficiency of the photodegradation process.