Pigmented Tannin and Derivative Analysis by a Complimentary Suite of Mass Spectrometric Techniques
Jonathan Cave and Andrew Waterhouse*
*University of California, Davis, One Shields Ave, Davis, CA
95616 (alwaterhouse@ucdavis.edu)
A combinatorial prediction of the pigmented tannin in red wine was made using known proanthocyanidin and anthocyanin precursors, then following established reaction mechanisms for pigmented tannin such as pyranoanthocyanins, pinotins, portisins, and similar compounds, creating a database of postulated compounds. Formulae were derived to limit the list to only those compounds that were distinguishable by mass spectrometry. Analysis of 1.1 million compounds below 5000 Daltons showed that all could be distinguished by Fourier Transform Ion Cyclotron Resonance Mass Spectrometry (FTICR), but less-precise techniques would leave some adjacent signals indistinguishable. Pigmented tannin isolates from red wines spanning more than a decade were purified on Sephadex LH20, then analyzed by FTICR and separately by a custom-made diol (normal phase) separation with a Quadrupole Time of Flight tandem mass spectrometric analyzer (nano-HPLC-QTOF). FTICR analysis obtained several thousand signals that could be matched to the database of postulated pigmented tannin derivatives and the unknowns were cataloged for future MSn investigation. Nano-HPLC-QTOF successfully resolved several hundred peaks from the previously indistinguishable hump that results from red wine phenolic chromatography and fragment analysis provided isomeric discrimination. Both FTICR and QToF offer high sensitivity and selectivity, providing greater signal density for the samples analyzed. With FTICR contributing very high mass accuracy for molecular formula determination and QToF fragmentation allowing discrimination of isomers, the pairing of these two techniques reveals extraordinarily detailed insights into the composition of red wine phenolics. This complimentary suite of mass spectrometric techniques provides a framework for continued evaluation of real wine samples with some of the most exhaustive qualification to date and a basis for advancing the structural analysis of unknown compounds at the trace level.
Funding Support: American Vineyard Foundation