PROMOTING ACCESS TO AFRICAN RESEARCH

Journal of Fundamental and Applied Sciences

Log in or Register to get access to full text downloads.

Remember me or Register



Synthesized and extending the Bidentate Schiff base complexes using multilayer feedforward neural network

A. M. Tajuddin, N. Ismail, K. Kassim, H. Bahron

Abstract


Complexes of Pd(II) and Ni(II) have been synthesized with general composition ML2X2 (M =
Pd(II), Ni(II); L = benzylsalicylideneimine and X = OCH3, F). All synthesized compounds
have been characterized using elemental analysis, magnetic susceptibility measurements,
infrared and NMR spectral studies that led to the conclusion that the ligands act as bidentate
manner to form square planar geometry for all complexes. As an extending work, the model
development of these complexes using multilayer feedforward neural network were
performed. NiL1d, PdL1d, NiL1c and PdL1c were fed to the training network as inputs and
bacteria as output. Levenberg Marquardt training algorithm was used during the network
training with 10 nodes in hidden layer. The results of testing network showed that the
regression, R is 1, indicating that the developed model is good. This is supported by the small mean square error (MSE) is 1.948x10-28 at epochs 5. The finding in this study is significant, thus contributed to the design of antibacterial agent especially to the bidentate Schiff base complexes.

Keywords: Schiff base, palladium(II), nickel(II), antibacterial, regression, neural network.




http://dx.doi.org/10.4314/jfas.v9i4S.57
AJOL African Journals Online