Analysis of a complex milieu, containing a vast number of structurally diverse analytes, is increasingly demanded in various domains including quality controlling of foods and beverages, environment monitoring, crime preventing by detecting and monitoring explosive and harmful biological species, as well as medical diagnostics. For such applications, electronic noses/tongues have inherent advantages comparing to conventional biosensors and biochips based on specific interaction (the principle of lock-to-key). They are inspired by mammals' olfactory system and constructed using sensor arrays based on cross reactivity between receptors and analytes. The analysis is then realized based on pattern recognition.

Recently in our laboratory we have developed a novel approach in order to simplify the preparation of electronic tongue (eT) using new sensing materials coupled with surface plasmon resonance imaging as sensing system. Such an eT has been successfully used for analysis of pure analytes, for example, differentiation of proteins of different families or isoforms of same family. It is also very efficient for analysis of complex mixtures, for example, classification and identification of beers, wines and milk.

In this thesis, we propose to introduce more building blocks to increase diversity of sensing elements and to improve the efficiency of the eT system. Moreover, we propose to broaden its application domain, particularly to explore the potential in biomedical domain. The main challenges of the thesis are related to the choice of the sensing materials and the construction of an efficient, robust and reproducible eT for working with biological complex milieu. In the process of design and development of this kind of eT, a variety of fundamental physical and chemical phenomena will be studied. Analysis of results, particularly concerning a large number of data to treat, will be carried out in close collaboration with Group Théorie of laboratory SPrAM.