Artificial Neural Network Modeling of Tetracycline Biosorption by Pre-treated Posidonia oceanica

Importance of the artificial intelligence in the chemical processes has been increased in the recent studies. Although biosorption is widely studied topic in chemistry, modelling of biosorption data is based on very old equations. However, use of artificial intelligence in the biosorption based studies can give important clues to researchers. For this purpose, the biosorption of tetracycline by using Posidonia oceanica from the Mediterranean Sea was studied in this study. According to classical evaluation, the data were well in line with pseudo-second order kinetic and Langmuir's isotherm. In the artificial neural network modelling, the best back propagation algorithm, optimum number of hidden neuron and optimum training: validation: testing ratio were found as Bayesian Regulation, 16 and 70: 10: 20, respectively. In conclusion, P. oceanica based marine waste can be used in the development of high performance biosorbents for environmental pollutants. However, it should not be forgotten that P. oceanica is a threatened species; therefore, only dead leaves accumulated in recreational area should be collected and evaluated based on the permissions of governmental authorities. The results also exhibited that artificial neural network can also be used in the modelling of the biosorption data in which it helps scientists to estimate biosorption ratio correctly under various conditions.