'Electronic Tongue' Can Tell Old Brandy From New

Researchers from St. Petersburg State University have designed a brandy analyzing device which can distinguish the ‘young’ drink from the seasoned one and even to distinguish among variants of brandy. This is of great help both to manufacturers and analysts on the hunt for counterfeits.

This ‘electronic tongue’ won't replace the cooper just yet but it is suited for routine work. The fun part - drinking - is still up to people.

The concept of ‘electronic noses’ and ‘electronic tongues’ are widespread in analyst circles but are sometimes confusing to laypeople. It means the device is intended for determination and/or identification of the objects of complicated composition. These devices do work on the principles of ‘human’ noses and tongues – that is, unlike ordinary analytical instrumentation it is not necessary to separate all components of a complicated raciness of the object’s taste or aroma so as to analyze them later individually.

The electronic tongue determines the taste ‘as a whole’ like a good cooper who recognizes gustatory sensations but has no idea about exact chemical compositions of the products. The electronic nose deals with scents in the same manner – it does not break them down into components but remembers and then determines at once as a whole – by the bouquet.

All these devices have one but a very significant drawback. They are never multi-purpose. For each task (should it be identification of rare earth elements in spent fuel or analysis of coffee, tea, juice, vodka, or dairy produce), an individual set of sensors should be developed to enable accurate reveal and identification of the aroma in question without ‘digressing’ to other foreign tastes and scents.

For brandy, the task was to ‘teach’ the device, firstly, to distinguish the ‘young’ unseasoned drink from the seasoned one. The St. Petersburg ‘electronic tongue’ has coped with this task more than successfully. This is quite obvious – the seasoned brandy contains tannin, the presence of which fundamentally changes the chemical composition of the drink.

Then there came a task to teach the device to distinguish variants of the drink, and to tell the brandy produced in laboratory environment from the one produced via industrial process. They have also succeeded in this task although not that brilliantly – expectancy of ‘hitting’ made approximately 3 out of 4. This is not that bad in principle – coopers also make mistakes.

source science20.com