Rapid Determination of Phenolic Compounds in Red Wine Fermentations

1
Rapid Determination ofPhenolic Compounds in Red
Wine Fermentations
Kirsten SkogersonGraduate ResearcherDepartment
of Viticulture EnologyUniversity of
California, Davis
2
Presentation Overview

• Phenolic compounds in red wine
• Project goals experimental plan
• Data collection analysis
• Conclusions

3
Phenolic Compounds in Red Wine
Tannin
Anthocyanins
Polymeric Pigment LPP / SPP
Non-Tannin Phenols

• Diverse class of compounds
• Dihydroxy functionality

4
Laboratory Methods for MeasuringPhenolic
Compounds in Red Wine

• Spectroscopic
• Chromatographic

5
Limitations of Current Methods

• Time-consuming
• Contract laboratories
• expensive and delayed
• Small sample number limits scope of research

6
Predictive Method

• Use multivariate statistics to correlate spectral
  data with traditional laboratory (reference)
  methods
• Develop model that predicts laboratory method
  values from spectra
• Once the model has been built, results are
  obtained instantaneously
• Used for analysis in many food beverage
  industries
• Increasingly used in grape and wine analysis

7
Spectroscopy-Based Predictive Method
Reference Method (GC)
Mathematical expression that relates EtOH to
Spectrum
Multivariate statistical program
Spectroscopy (NIR)
MODEL BUILDING
8
Project Goal

• To predict phenolic composition using UV-vis
  spectroscopy

9
UV-visible spectra of phenolic compounds
10
Project Overview Model Building
REFERENCE METHOD
MODEL GENERATION
SAMPLE COLLECTION
SPECTRA
11
Sample Collection 2006

• 600 fermentation aged wine samples

• Regions (14)
• Barossa Valley
• Riverland
• Waikerie
• Barmera
• McLaren Vale
• Langhorne Creek
• Clare
• Bordertown
• Blanchetown
• Loxton
• Padthaway
• Adelaide Hills
• Renmark
• Coonawarra . . .

• Cultivars (8)
• Shiraz
• Cabernet Sauvignon
• Merlot
• Tempranillo
• Pinot Noir
• Grenache
• Chardonnay
• Riesling

12
Reference Method

• Harbertson-Adams Assay
• Robust, comprehensive method
• Measures
• Anthocyanins
• Large and Small Polymeric Pigments
• Tannins
• Non-tannin phenols

13
Harbertson-Adams Assay Anthocyanins and Total
Phenols
Juice / wine sample
Dilute into SDS buffer
Take background measure at 510 nm
Add Fe Cl3
14
Harbertson-Adams Assay Polymeric Pigments and
Tannin
Juice / wine sample
15
Harbertson-Adams Assay Sample Data Set
Statistics
Phenolic Class
Range
Mean
S.D.
Anthocyanin (mg/L M3G)
0 - 1096
414
244
Tannin (mg/L CE)
0 - 798
217
189
Pigmented Polymers (AU)
0.015 - 6.6
1.98
1.27
Non-Tannin Phenols (mg/L CE)

18 - 1518
582
255
16
Spectral Analysis

• Harbertson-Adams assays spectral collection
  performed on same day

Metertech SP8001
17
Model Building Unscrambler

• Randomly select 314 of 414 samples for model
  construction
• Model built using PLS with cross-validation
• Test Model with remaining 100 samples

18
Anthocyanin
PREDICTED
r2 0.91
MEASURED
19
Pigmented Polymers
r2 0.85
PREDICTED
MEASURED
20
Tannin
r2 0.86
PREDICTED
MEASURED
21
Non-Tannin Phenols
r2 0.83
22
Comparing UV-vis vis-NIR Models
r2
Error
Anthocyanin
UV-vis
0.91
18
0.92
vis-NIR1
15
Pigmented Polymer
UV-vis
0.85
25
vis-NIR
0.87
26
Tannin
UV-vis
0.86
33
vis-NIR
0.80
38
Non-Tannin Phenols
UV-vis
0.83
18
vis-NIR
-
-
1Cozzolino et al. 2004. Analytica Chemica Acta
51373-80.
23
Project Overview Model Predictions
COLLECT SAMPLES
IMMEDIATE RESULTS
COLLECT SPECTRA
24
Anthocyanin Predictions
PREDICTED
r2 0.88
MEASURED
25
Pigmented Polymer Predictions
PREDICTED
r2 0.78
MEASURED
26
Tannin Predictions
r2 0.78
PREDICTED
MEASURED
27
Non-Tannin Phenol Predictions
r2 0.78
28
Conclusion

• Demonstrated potential of UV-vis predictive
  models.
• Predict anthocyanins, pigmented polymers,
  tannins, and non-tannin phenols.
• Effective for juices, fermenting musts, and wine.
  
• Predictions independent of cultivar and growing
  region of the samples.

29
Acknowledgements

• Roger Boulton
• Jerry Lohr
• T.J. Rogers
• Orlando Wyndham
• Yalumba

• Primary Industries Research, Victoria
• Mark Downey
• Marica Mazza

• Scholarship Support from
• ASEV
• Wine Spectator
• Temecula Valley Wine Society
• Orange County Wine Society
• AWSEF
• UC Travel Grant