Abstract:Acid Brown 14 (AB14) and Acid Yellow 36 (AY36) are synthetic azo dyes extensively utilized in numerous industries, resulting in detrimental environmental consequences. This study aims to manufacture self-nitrogen-doped porous activated carbon (AC7-800) and investigate its effectiveness in removing the AB14 and AY36 dyes from water solutions. The AC7-800 was created by combining fish waste (with a protein composition of 60% as a nitrogen source), which served as a self-nitrogen dopant. An equal mass ratio (1:1:1) of sawdust, fish waste, and zinc chloride underwent a hydrothermal treatment at 180 °C for 5 h. Subsequently, the material underwent pyrolysis for 1 h in a continuous flow of nitrogen gas at 800 °C to produce AC7-800. The AC7-800 adsorbent was successfully tested and approved to eliminate colours from water in batch trials. The AC7-800 samples were analyzed using BET, SEM, EDX, XRD, FTIR, TGA, and DTA techniques. The results demonstrated the practical synthesis of AC7-800 with a nitrogen mass percentage concentration of 13.73%. The specific surface area, mean pore diameter and monolayer volume were measured to be 437.51 m2 g− 1, 2.01 nm, and 100.52 cm3 g− 1, respectively. The objective is to examine the elimination of AB14 and AY36 dyes from a water-based solution using various factors such as initial dye concentration, solution pH, AC7-800 dosage, and contact time. The efficacy of AC7-800 in removing AB14 and AY36 dyes was found to be dependent on the pH level. The highest elimination efficiency of 63.29% and 85.86% was achieved at pH 1.5 for AB14 and AY36 dyes, respectively. Additionally, the maximum adsorption capacity (Qm) for AB14 and AY36 dyes was determined to be 107.5 and 263.2 mg g− 1, respectively. The equilibrium data demonstrated a good association with the Langmuir model (LIM) for both dyes, although the best-fit kinetic model was the pseudo-second-order model (PSOM). Electrostatic interactions between the dye molecules and the charged spots on the AC7-800 surface cause both dyes to adsorb. The prepared AC7-800 can be considered a highly effective, accessible, and environmentally acceptable adsorbent for the adsorption of AB14 and AY36 dyes from simulated water. AB14 and AY36 dyes adsorption to AC7-800 was predicted by the response-surface methodology (RSM) and artificial neural networks (ANN) models. The ANN model was more effective in predicting AB14 and AY36 dyes adsorption than the D-optimal RSM, and it was highly applicable in the sorption process.