Metalloporphyrins thin films sensors array equipped with backpropagation network for vapor recognition

This work reports the fabrication of an array of sensors system equipped with a pattern recognition system to classify four types of vapor samples; 2-propanol, ethanol, acetone and cyclohexane. The array comprises eight metalloporphyrins derivatives thin films as sensing element. A backpropagation artificial neural network was used as pattern classifier. The presentation of a vapor sample towards the sensing elements produced the response pattern which was considered as the vapor finger print. A library of the vapor pattern which has been introduced to the sensing elements was built up. The pattern was then labeled and introduced to the neural network. After proper learning, the network was tried to recognize the vapor pattern. The recognition results indicated that the system was able to recognize the sample with the overall system performance is 0.75.