Pattern recognition approach to the study of the interactions between metalloporphyrin Langmuir-Blodgett films and volatile organic compounds

In recent years metalloporphyrins and their derivatives have been proposed as sensitive layer for mass variation based transducers for the detection of volatile compounds. Their sensitivity/selectivity properties have suggested their application in electronic nose systems. Despite the positive applications appeared so far, a systematic study of the kind of interactions ruling the sensing mechanism has not yet been conducted. The selectivity towards a variety of compounds suggests the presence of weak Van der Waals or hydrogen bond interactions, but also certain evidences about coordination interactions with the central metal ion coordinated to the porphyrin were reported. In this paper, a study oriented towards a deeper comprehension of such mechanisms is introduced. From the experimental point of view, a set of quartz microbalances, coated with Langmuir-Blodgett films of various metalloporphyrins, were exposed to a number of volatile compounds at various concentrations. Linear sorption energy relationship theory provides, in many cases, a satisfactory model for sorption sensors in which Van der Waals and hydrogen bonding are responsible for the interaction. Nevertheless, the application of such a model results inadequate in the case here discussed. A more satisfactory treatment of the data was obtained considering a pattern recognition approach based, in particular, on a self-organizing neural network. This type of analysis has shown the role played by the central metal in determining the sensor selectivity; furthermore, the results have allowed to hypotize the presence of a coordination mechanism, superimposed to the weak interactions, between volatile compounds and metalloporphyrin films. (C) 1999 Elsevier Science B.V All rights reserved.