Autori: C. Luca, A.M. Grigoriu, R.M. Diaconescu, M.S. Secula
Editorial: Revista de Chimie -Bucharest, 62 (10), p.1033-1038, 2011.
This study presents the neural network modeling of monochlorotriazinyl- B-cyclodextrin (MCT-B-CD) grafting on paper supports. Neural networks are very efficient in predicting the evolution of a process, providing a good approximation ot the studied grafting process, further used for the interpolation and extrapolation of the results beyond the experimental range. The grafting was realized by a pad-dry-cure treatment and the mathematical model has quantified the influence of the real independent variables (i.e concentrations of MCT-B-CD and of the catalyst temperature) on the grafting degree as goal, function. The treated paper samples were analyzed by FT-IR-ATR spectroscopy and SEM, in order to prove the grafting of the cellulosic support.
Cuvinte cheie: reţele neuronale artificiale, modelare şi simulare, grefare // artificial neural network, modeling and simulation, grafting, nanocoated paper