In the last years, the group of Electroanalysis has been focused on the electroanalytical study of the interactions between heavy metal ions and molecules of environmental and biological interest. The great difficulties of this subject have required the development of a large deal of methodologies, both experimental and theoretical, and have considerably diversified the research lines of the group, which nowadays are the following:

Development of theoretical models and experimental methodologies for metal speciation based on electroanalytical measurements.

A totally different approach for heavy metal speciation is that based on the use of chemometric techniques of multivariate data analysis. This strategy, usually termed as soft modelling, tries to identify the speciation model exclusively from the data variation, without any previously assumed model.

In order to obtain such information, an adequate experimental design is required to allow a further combination of the results into a data matrix. The usual case is the measurement by differential pulse voltammetry (DPV) of a metal solution after successive additions of ligand or vice-versa. The application of factor analysis techniques to the resulting data allows one to obtain the concentration profiles and the pure signals of the formed species.

Chemometric techniques of multivariate data analysis have been extensively applied to spectrophotometric measurements but, in contrast, their application to electrochemical measurements has been scarce up to now.

One research topic of the group which has become especially important is, indeed, the application of chemometric techniques of multivariate data analysis to data obtained by means of electroanalytical techniques. This application is especially interesting when numerous overlapping signals are obtained which vary in number, shape and relative height during the experiment. This behaviour can be observed in the study of molecules of biochemical origin such as peptides or proteins.

With the aim of testing the possibilities of these techniques, several relatively simple systems were studied first and, later, more intricate systems were considered such as the metal complexes of metallothioneins (MT), phytochelatins (PC), glutathione (GSH) and other related compounds. The interest of such substances, besides the possibilities of electroanalytical measurement, is their role of bioregulators of metals in organisms (MT in animals and PC in plants).

The special characteristics of some of these systems made necessary the development of new restrictions in the algorithms of multivariate data analysis, such as the shape constraints (fitting of sigmoid or peak-shaped signals to a parametric function or fitting of the concentration profiles to a model of complexation equilibria). We have also considered the losses of linearity derived from signal movement along potential axis (e.g., due to fast equilibria) or from peak broadening (due to changes in electrochemical reversibility). To minimize these problems we have developed correction algorithms previous to or integrated into the multivariate analysis.

By other hand, the increasing data complexity has also demanded the use of complementary techniques for a better selection among the different complexation models compatible with voltammetric data. One of these techniques (circular dichroism, CD) allows the combination of spectroscopic and voltammetric data for an integrated multivariate analysis, and other techniques (isothermal titration calorimetry, ITC, and electrospray-ionisation mass spectrometry, ESI-MS) are essentially used to confirm the presence of species detected by voltammetry.

Development and application of modified electrodes.

Despite the excellent features of mercury electrodes, the toxicity of this element and the restrictive regulation about its use have stimulated the search of substitutive environmentally friendly materials such as bismuth. In the last years, the group has investigated the possibilities of conventional glassy carbon electrodes (GCE) and screen printed carbon electrodes (SPCE) both modified with bismuth films. This is made not only to obtain less toxic electrodes, but also to minimise the voltammetric signals generated by the oxidation of the electrode material to form complexes with the ligands under study (especially thiol-rich proteins).

Other modifications tested in the group where focusing more on the improvement of electrode kinetics than on mercury substitution. This is the case of gold nanoparticles and carbon nanotubes deposited as films in screen-printed electrodes.

Different fast scan techniques have been tested in order to obtain three-dimensional chromatograms (current matrices as a function of retention time and potential). They have been applied for the analysis of mixtures of catecholamines, easily oxidisable on glassy carbon electrodes. By now, the best results have been obtained with linear sweep voltammetry (LSV), which allows one to get ca. one potential sweep per second with an acceptable signal-to-noise ratio.

With the aim of testing the possibilities of multivariate analysis in this kind of data, measurements have been carried out under conditions that produce overlapping chromatographic peaks. The analysis of the data matrices generated in this way shows that, if the baseline is considered as an additional component, an acceptable resolution can be achieved for the coeluting species, provided their voltammetric signals are sufficiently different. For more involved systems, we are investigating the possibilities of multichannel detection by using serial or parallel electrodes.

Additionally, we are working on the optimisation of both the separation and the amperometric detection of synthetic mixtures of phytochelatins and their complexes with mercury, as well as on the analysis of natural phytochelatins extracts from plants grown in the presence of high levels of mercury.