Page 5 - UVSpecPaperBDASWhitePaperandBMGLabTechNote
P. 5

might be the case with other components changing       and usually no blank wells needed to be scanned; thus
         the overall fingerprint profile upon sample dilution;   maximizing sample testing. Three wells per sample are
         alcoholic beverages composed of several hundred        sufficient, in our opinion, once a library of data has been
         constituents providing a complex matrix for such       established for a product to assist in subsequent testing
         analyses. The issues with respect to sample dilution   of newer or suspect batches of the same beverage.
         with beer samples was also discussed elsewhere (12).   Carbonated beverages were first degassed. Plates were
         Nevertheless, either with or without sample dilution   then read in a BMGLABTECH SPECTROStar Nano
         it seems the method can be applied to a wide range     scanning spectrophotometer in 1 nm increments
         of alcoholic beverages and provide solid and reliable   from 220 to 1000 nm; path-length correction was set
         results as to quality – consistency, brand authentication   to “off”. Averaged data from the individual sets were
         and potential adulteration of the spirit, liqueur, wine or   evaluated and plotted using spreadsheet programs
         beer. The method may be used alone or in conjunction   (either internal or external to the manufacturers’
         with powerful chemometric evaluations of the spectral   software). Simple Z-score statistical evaluations were
         data  to  further  discriminate  different  brands  or   made  to  look  for  occasional  outlier  data  from  wells
         types/styles of beverages (1-4, 7, 9, and 10). Recent   measured from within a set. Deviation may be caused
         developments in chemometric analyses are, in our       by slight imperfections in the plastic micro-plates, if
         opinion, expected to play a greater role in this arena   the wells are soiled or otherwise contaminated, if the
         as more testing of alcoholic beverages takes place. Our   correct volume from multi-pipette additions are not
         lab in collaboration with a British University team is   even across the range of wells or if any air bubbles
         looking into the application of chemometrics to further   are present after  pipetting such  low volumes.  Only
         discriminate samples and map styles.                   occasionally do we see any such issues. The method
                                                                and testing should be amenable to newly trained non-
         Experimental Protocols and Systems Assessment          laboratory personnel (including beverages sales teams)
         Samples of most distilled spirits, 1:10 distilled water-  without the need to perform statistical analyses for
         diluted sherry samples or beer/malt beverages were     routine quality evaluations in the beverage plant or the
         loaded as 50 or 100 microliter additions into micro-   field.  While many instruments on the market are useful
         wells (Costar Corning #3635 UV transparent 96-well     for such analyses, as described herein, we believe that
         plates)  avoiding bubbles;  typically up  to eight wells   many benefits are offered through the use of a scanning
         per sample were examined in each run to test for any   plate-reading instrument. As the preferred instrument
         variance in data through optical transmission through   in our own facility (used in other studies, 18 and 19)
         the base of the plate and within the individual cells   we discuss such advantages in more detail elsewhere






























                          Figure 2: Five sherry samples with different aging periods and characteristics.


                                                              3
   1   2   3   4   5   6   7   8   9