Sample presentation, sources of error and future perspectives on the application of vibrational spectroscopy in the wine industry

Vibrational spectroscopy encompasses a number of techniques and methods including ultra-violet, visible, Fourier transform infrared or mid infrared, near infrared and Raman spectroscopy. The use and application of spectroscopy generates spectra containing hundreds of variables (absorbances at each wavenumbers or wavelengths), resulting in the production of large data sets representing the chemical and biochemical wine fingerprint. Multivariate data analysis techniques are then required to handle the large amount of data generated in order to interpret the spectra in a meaningful way in order to develop a specific application. This paper focuses on the developments of sample presentation and main sources of error when vibrational spectroscopy methods are applied in wine analysis. Recent and novel applications will be discussed as examples of these developments. © 2014 Society of Chemical Industry.