Artificial neural networks modeling of electrospinning of polyethylene oxide from aqueous acid acetic solution

The artificial neural networks (ANNs) were used to provide a model for investigating the relationships of the electrospinning parameters with the diameter of polyethylene oxide (PEO) nanofibers from acid acetic aqueous solution. The effects of four parameters including PEO concentration, acetic acid concentration, applied voltage, and temperature of the electrospinning media on the nanofibers mean diameter were investigated. To train, test, and valid the model, three datasets of the input variables with random values were prepared and the mean diameters obtained were taken as the output for the network. The datasets were analyzed by ANNs software and the correlation coefficient, R-squared (R²), between the predicted values of the nanofibers mean diameter and actual amount were obtained. The results demonstrate the capability of the ANNs model for predicting the nanofibers diameter. The 3-D plots generated from the model show complex and nonlinear relationships between the parameters and nanofibers diameter. From the model, increasing the PEO concentration above a critical point leads to a sharp increase in the nanofibers mean diameter. The effects of applied voltage and temperature are mainly dependent on the PEO concentration. The acetic acid concentration, in general shows a direct relation with the nanofibers mean diameter. The plots also show that to produce nanofibers with the lowest diameter, both the PEO concentration and AcOH concentration should be at lowest values regardless the applied voltage and temperature. In contrast, highest nanofibers diameters are obtained when the PEO concentration and AcOH concentration are at their high values. © 2012 Wiley Periodicals, Inc. J Appl Polym Sci, 2012