Artificial neural networks modeling of the parameterized gold nanoparticles generation through photo-induced process

In this study, gold nanoparticulate patterns were generated in polymer thick film using a direct light writing method and were characterized using optic, spectroscopic and transmission electron microscopy investigations. Based on physical and chemical process parameters that have an important contribution on the generated gold nanoparticle size, an artificial neural network was developed to predict the localized surface plasmon absorption maxima and consequently, the gold nanoparticles corresponding dimension. Due to the excellent predicting capabilities supported by a high correlation factor and low relative errors, the trial and error approach for generating the desired gold nanoparticle dimension is no longer used and, in addition, the samples are no longer destroyed for transmission electron microscopy measurements. Furthermore, correlations between the predicted gold nanoparticles dimension and the citrate to gold(III) ratio, scanning velocity and radiation intensity are investigated. The results highlighted that the absorption maxima along with its associated gold nanoparticles dimension increased with decreasing the intensity and the citrate to gold(III) ratio as well as with increasing the scanning velocity. The radiation intensity was found to have the most important influence on the gold nanoparticle size, followed by the scanning velocity and the citrate to gold(III) ratio.