Improved Analytical Methods for ANDREW WATERHOUSE, TODD JENKINS DEPARTMENT OF VITICULTURE & ENOLOGY UC DAVIS RAVE 2018, NOV 30, 2018 Overview Currently Practiced Analysis Methods And the need for a more accurate method Chemistry Review New Method, Headspace Gas Detection Tubes (HS GDT) Adaptation of HS GDT to Headspace Gas Chromatography (HS GC) Enological Implications 1
Measuring in the Lab o Aeration / Oxidation o Ripper Method o Flow Injection o NIR o Colorimetric Dissociation + H 2 O HSO 3 + H + Molecular pka = 1.8 2.0 Bisulfite pka = 8.1 7.7 SO 3 2 + 2H + Sulfite o Antimicrobial (0.5 0.8 mg/l) o Antioxidant Quenches H 2 O 2 o Reacts with quinones o Binding/addition reactions o Inactivates polyphenoloxidase (PPO) o Negligible amounts in wine o Dissociation influenced by ethanol o Dielectric effect o ~13% ethanol o pka 1 (~2.0) (Butler, 1964) Ionic Equilibrium: Solubility and ph Calculations (Usseglio Tomasset, 1984) La prima costante di dissocianzione dell acido solforoso 2
Free and Bound Bisulfite anion binds with carbonyls, particularly acetaldehyde Truly Free Weakly Bound (metastable) Acetaldehyde Bound 3 SO 2 2 HSO 3 SO 2 3 3 SO 3 bound to phenolics, SO 3 bound colors, to sugars, acetaldehyde, proteins phenolics, SO 3 bound color, to sugars acetaldehyde Total Anthocyanins Pyruvate ketoglutarate Acetoin Galacturonic acid Sugars Proteins Diacetyl Strongly Β Damascenone Bound Β Ionone Hexanal (E) 2 Pentenal (E) 2 Nonenal An alternative approach Headspace ] (g) Binder HSO 3 (aq) K H Ka1 Ka2 (aq) HSO 3 (aq) SO 3 2 (aq) 3
New Method: Headspace Gas Detection Tubes (HS GDT) (Coelho et al. 2015) 1) Draw sample (10 ml) 2) Draw 50mL Headspace 3) Equilibrate for 5 min (RT) 4) Expel headspace through GDT 5) Record colorimetric change along tube 6) Perform back calculations using alcohol, ph, and temperature to obtain molecular and truly free. Headspace Gas Detection Tubes Strengths Fast (5 minutes) provided you already know alcohol and ph Robust Accessible to just about anyone (perfect for home winemakers) Can run multiple samples in batches Does not disturb equilibrium (no ph shift) Weaknesses Difficulty in endpoint determination Potential cross sensitivities No option for automation Difficulty in controlling temperature (water bath) 4
pa 3000 2500 2000 1500 1000 500 0 SCD1B,BackSignal(TJ180313A\TJ180313A2018-03-1309-05-18\15WHCHARDONNAY1.D) SCD1 B, Back Signal (TJ180313A\TJ180313A2018-03-1309-05-18\15 WH CHARDONNAY2.D) SCD1 B, Back Signal (TJ180313A\TJ180313A2018-03-13 09-05-18\15 WH CHARDONNAY3.D) 1 2 3 4 5 6 7 min Gas Chromatography (HS GC SCD) PAL RSI 85 Robotic Autosampler (With Static Headspace Syringe) Agilent 7890B Gas Chromatograph (GC) Agilent 8355 Sulfur Chemiluminescence Detector (SCD) HS GC Analysis Column: DB WAX Ultra Inert (30m x 0.25 mm ID x 0.25 um) Injection Volume: 0.500 ml (4:1 split ratio), inlet temp. 240 C Carrier Gas: Helium (constant flow 1.7 ml/sec.) Oven Program: Initial temp. 50 C (2 minutes) Ramp at 50 C/min to 220 C (hold for 2 minutes) (0.84 ppm molecular ) % RSD = 2.15% 1 ml gas tight headspace syringe (0.500mL) Total chromatographic time 7.9 minutes (10.5 minutes cycle time) EMS (ISTD) DMS H 2S Gas phase SO 2 SO SO LOD = 0.033 mg/l molec. SO 2 2 2 15mLs of wine SO LOQ = 0.067 mg/l molec. SO 2 SO 2 + 100ppb Ethyl Methyl Sulfide 2 SO (Internal Standard) 2 Aqueous Phase 5
2015 Viogner Central Coast 14.3% ethanol (true pka 1 = 2.20) 3.45 ph 25.0 C SCD1 B, Back Signal (TJ180313A\TJ180313A 2018-03-1309-05-18\15 WH CHARDONNAY1.D) SCD1 B, Back Signal (TJ180313A\TJ180313A 2018-03-1309-05-18\15 WH CHARDONNAY2.D) SCD1 B, Back Signal (TJ180313A\TJ180313A 2018-03-1309-05-18\15 WH CHARDONNAY3.D) pa 3000 2500 (0.84 mg/l molecular ) Henderson Hasselbalch 15.9 ppm Truly Free 3 analytical replicates % RSD = 2.15% 2000 1500 1000 500 EMS (ISTD) DMS H 2 S 0 1 2 3 4 5 6 7 min Apparent Free Truly Free Sample ID Wine Type A/O Ripper HS GC HS GDT "Weakly Percentage Bound" in (Metastable) Metastable SO2 complexes 2015 California Red Blend Red 35.6 44.6 15.4 14.3 20.2 57% 2014 Sonoma County Zinfandel Red 22.7 22.7 4.4 7.1 18.3 81% 2016 Monterey County Pinot Noir A Red 30.6 36.7 15.5 15.1 15.1 49% 2015 Paso Robles Blaufrankisch Red 13.3 16.3 2.2 1.9 11.2 84% 2015 California Cabernet Sauvignon Red 15.4 19.8 4.5 2.4 10.9 71% 2014 Napa Valley Merlot Red 10.0 15.5 0.0 0.0 10.0 100% 2013 Alexander Valley Zinfandel Red 11.9 15.2 2.8 4.4 9.2 77% 2015 California Chardonnay White 30.1 31.8 22.4 29.1 7.7 26% 2015 Central Coast Merlot Red 17.8 23.0 11.2 10.1 6.5 37% 2016 Central Coast Rose Rose 21.1 24.6 14.6 18.8 6.5 31% 2016 Monterey County Pinot Noir B Red 22.8 30.4 16.9 11.3 5.9 26% 2013 Paso Robles Merlot Red 11.7 15.9 6.1 0.0 5.6 48% 2015 Central Coast Pinot Noir Red 22.2 28.4 16.7 15.0 5.5 25% 2014 California Moscato White 8.8 9.5 3.5 3.4 5.4 61% 2015 Alexander Valley Fume Blanc White 16.2 17.5 11.2 14.6 5.0 31% 2014 Napa Valley Chardonnay White 20.9 25.7 15.9 18.2 4.9 24% 2013 California Zinfandel Red 8.0 10.6 3.5 0.0 4.5 56% N.V. Brut Sparkling White 27.1 25.7 25.2 25.9 1.8 7% 2014 Central Coast White Blend White 14.3 14.3 13.0 13.6 1.3 9% 2015 California Sauvignon Blanc White 23.3 24.8 22.8 23.1 0.5 2% 2015 Napa Valley Chardonnay White 17.2 18.1 16.8 15.3 0.4 2% 2015 Napa Valley Moscato White 17.1 17.3 16.8 13.2 0.3 2% 2014 Central Coast Viogner White 19.5 23.1 19.3 21.8 0.2 1% 2012 Napa Valley Port Red 0.0 6.4 0.0 0.0 0.0 0% 2016 Sonoma County Moscato White 7.4 10.5 7.5 7.3 0.0 0% 2014 Central Coast Chardonnay White 32.7 34.5 32.9 34.3 0.0 0% 2015 Central Coast Viogner White 15.2 17.3 15.9 11.5 0.0 0% Average Std. Dev. 0.8 0.7 0.4 1.4 Average % RSD 4.6% 3.8% 3.4% 11.8% HS GC Free SO2 HS GC Free SO2 35.0 30.0 25.0 20.0 15.0 10.0 5.0 HS GC Free SO2 vs. HS GDT Free SO2 y = 0.85x + 2.03 R² = 0.91 0.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 HS GDT Free SO2 A O Free SO2 vs. HS GC Free SO2 Whites R² = 0.88 Reds R² = 0.65 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 A O Free SO2 6
"Weakly Percentage Sample ID Wine Type Bound" in (Metastable) Metastable SO2 complexes 2015 California Red Blend Red 20.2 57% 2014 Sonoma County Zinfandel Red 18.3 81% 2016 Monterey County Pinot Noir A Red 15.1 49% 2015 Paso Robles Blaufrankisch Red 11.2 84% 2015 California Cabernet Sauvignon Red 10.9 71% 2014 Napa Valley Merlot Red 10.0 100% 2013 Alexander Valley Zinfandel Red 9.2 77% 2015 California Chardonnay White 7.7 26% 2015 Central Coast Merlot Red 6.5 37% 2016 Central Coast Rose Rose 6.5 31% 2016 Monterey County Pinot Noir B Red 5.9 26% 2013 Paso Robles Merlot Red 5.6 48% 2015 Central Coast Pinot Noir Red 5.5 25% 2014 California Moscato White 5.4 61% 2015 Alexander Valley Fume Blanc White 5.0 31% 2014 Napa Valley Chardonnay White 4.9 24% 2013 California Zinfandel Red 4.5 56% N.V. Brut Sparkling White 1.8 7% 2014 Central Coast White Blend White 1.3 9% 2015 California Sauvignon Blanc White 0.5 2% 2015 Napa Valley Chardonnay White 0.4 2% 2015 Napa Valley Moscato White 0.3 2% 2014 Central Coast Viogner White 0.2 1% 2012 Napa Valley Port Red 0.0 0% 2016 Sonoma County Moscato White 0.0 0% 2014 Central Coast Chardonnay White 0.0 0% 2015 Central Coast Viogner White 0.0 0% Metastable Bisulfite Complexes Metastable Bisulfite Complexes Anthocyanin vs. Metastable Bisulfite Complexes 25.0 20.0 15.0 10.0 5.0 0.0 y = 0.12x + 3.82 R² = 0.39 0.0 20.0 40.0 60.0 80.0 100.0 120.0 Monomeric Anthocyanin 2 Ketoglutarate vs. Metastable Bisulfite Complexes 25.0 20.0 15.0 10.0 5.0 0.0 R² = 0.16 R² = 0.00 0 20 40 60 80 100 Ketoglutarate Metastable Bisulfite Complexes Metastable Bisulfite Complexes 25.0 20.0 15.0 10.0 5.0 0.0 25.0 20.0 15.0 10.0 5.0 0.0 Acetaldehyde vs. Metastable Bisulfite Complexes R² = 0.01 R² = 0.04 0 20 40 60 80 100 Acetaldehyde Pyruvate vs. Metastable Bisulfite Complexes R² = 0.20 R² = 0.06 0.0 10.0 20.0 30.0 40.0 50.0 Pyruvate HS GC Strengths Fast (8 10 minutes) provided you know alcohol and ph AUTOMATED Repeatable (%RSD < 3%) No interferences Does not disturb equilibria truly free Weaknesses Skilled (GC) operator required Needs periodic maintenance ( is acidic and shortens the column life) Alcohol level, ph, temperature needed High up front cost ($70 120K) S detector, headspace autosampler 7
Implications for Wine Microbiology White Wine Red Wine A O GDT Molec. by HS GDT Molec. by A O Decrease in EC 1118 Cell Counts Truly Free by HS Gas Detection Tubes Free by Aeration Oxidation Similar losses in red, white wine Red wines have no protection * Howe, P.A., et al. 2018. Conventional Measurements of Sulfur Dioxide ( ) in Red Wine Overestimate Antimicrobial Activity. Am. J. Enol. Vitic. 68: 210. 8
Enological Implications Wine Microbiology Better predictor of microbial spoilage in red wines (Howe et al. 2018) Wine Oxidation Chemistry Is Truly Free a better predictor of shelf life? Development of better methods for iron speciation (Fe 2+ /Fe 3+ ) Potential Future Studies Compare prediction of wine oxidation Antimicrobial activity of bisulfite complexes on spoilage microorganisms (brettanomyces, lactobacillus, etc.) Accelerated and long term aging Sensory studies Acknowledgments UC Davis Waterhouse Lab Mauri Anderson, Nicolas Delchier, Annegret Cantu, Aude Watrelot, Lingjun Ma, Thi Nquyen, James Campbell Cary Doyle Food Safety Measurement Lab Chik Brenneman and the RMI Winery Staff Dr. Roger Boulton Collaborators: Dr. Gavin Sacks Dr. Patricia Howe Constellation Brands Funding/Scholarships American Vineyard Foundation Grant ID #2075 American Society of Enology and Viticulture Henry A. Jastro Shields Wine Spectator, Harold Berg, Mario Tribuno, Rusty Staub 9
Thanks for your attention Questions? 10
The Wine Oxidation Cascade Sample ID Wine Type 'Weakly Bound' SO2 Monomeric Anthocyanin Pyruvate 2 Ketoglutarate Acetaldehyde 2015 California Red Blend Red 25.2 107.2 15.6 43.3 11.1 2016 Monterey County Pinot Noir A Red 18.3 82.0 10.5 35.6 3.1 2014 Sonoma County Zinfandel Red 17.0 42.9 12.7 67.2 22.5 2015 California Cabernet Sauvignon Red 14.1 37.7 11.0 55.7 6.1 2015 Paso Robles Blaufrankisch Red 12.8 48.4 9.9 10.1 4.6 2014 Napa Valley Merlot Red 12.7 40.2 15.4 76.1 20.1 2016 Monterey County Pinot Noir B Red 12.5 83.1 11.8 39.3 10.8 2013 Paso Robles Merlot Red 10.8 28.1 13.8 6.1 3.1 2013 Alexander Valley Zinfandel Red 10.0 22.5 8.6 67.0 9.5 2015 Central Coast Merlot Red 9.8 49.5 17.7 28.5 12.2 2015 Central Coast Pinot Noir Red 9.5 49.5 17.7 41.4 9.8 2013 California Zinfandel Red 7.6 20.0 8.7 91.9 8.6 2014 Napa Valley Chardonnay White 6.2 0.0 31.2 39.4 43.9 2016 Central Coast Rose Rose 6.2 6.3 10.5 38.9 21.2 2014 California Moscato White 5.8 0.0 39.6 30.2 66.0 2015 California Chardonnay White 5.2 0.0 12.3 26.6 27.5 2015 Alexander Valley Fume Blanc White 4.0 0.0 14.1 30.9 49.5 2012 Napa Valley Port Red 3.2 11.1 46.0 56.8 13.3 2015 Central Coast Viogner White 2.6 0.0 13.1 26.6 27.5 2015 Napa Valley Moscato White 2.2 0.0 13.1 42.2 24.7 2015 Napa Valley Chardonnay White 1.6 0.0 15.7 37.9 54.2 2016 Sonoma County Moscato White 1.5 0.0 18.2 0.0 47.4 2015 California Sauvignon Blanc White 1.1 0.0 19.1 29.5 40.1 2014 Central Coast White Blend White 1.0 0.0 14.7 22.7 39.4 N.V. Brut Sparkling White 0.8 0.0 46.8 33.6 81.9 2014 Central Coast Viogner White 0.7 0.0 16.6 22.7 46.8 2014 Central Coast Chardonnay White 0.0 0.0 14.6 36.2 42.5 Monomeric Anthocyanin Anthocyanin vs. Metastable Bisulfite Complexes 150.0 100.0 y = 4.40x 8.38 50.0 R² = 0.70 0.0 0.0 5.0 10.0 15.0 20.0 25.0 30.0 Metastable Bisulfite Complexes 11
Sample ID Wine Type 'Weakly Bound' SO2 Monomeric Anthocyanin Pyruvate 2 Ketoglutarate Acetaldehyde 2015 California Red Blend Red 29.2 107.2 15.6 43.3 11.1 2016 Monterey County Pinot Noir A Red 21.2 82.0 10.5 35.6 3.1 2014 Sonoma County Zinfandel Red 18.4 42.9 12.7 67.2 22.5 2014 Napa Valley Merlot Red 15.5 40.2 15.4 76.1 20.1 2015 California Cabernet Sauvignon Red 15.3 37.7 11.0 55.7 6.1 2015 Paso Robles Blaufrankisch Red 14.2 48.4 9.9 10.1 4.6 2016 Monterey County Pinot Noir B Red 13.5 83.1 11.8 39.3 10.8 2013 Alexander Valley Zinfandel Red 12.5 22.5 8.6 67.0 9.5 2015 Central Coast Merlot Red 11.8 49.5 17.7 28.5 12.2 2015 Central Coast Pinot Noir Red 11.8 49.5 17.7 41.4 9.8 2016 Central Coast Rose Rose 10.0 6.3 10.5 38.9 21.2 2013 Paso Robles Merlot Red 9.8 28.1 13.8 6.1 3.1 2014 Napa Valley Chardonnay White 9.8 0.0 31.2 39.4 43.9 2015 California Chardonnay White 9.4 0.0 12.3 26.6 27.5 2013 California Zinfandel Red 7.1 20.0 8.7 91.9 8.6 2012 Napa Valley Port Red 6.4 11.1 46.0 56.8 13.3 2015 Alexander Valley Fume Blanc White 6.3 0.0 14.1 30.9 49.5 2014 California Moscato White 6.1 0.0 39.6 30.2 66.0 2014 Central Coast Viogner White 3.8 0.0 16.6 22.7 46.8 2016 Sonoma County Moscato White 3.0 0.0 18.2 0.0 47.4 2015 California Sauvignon Blanc White 2.0 0.0 19.1 29.5 40.1 2014 Central Coast Chardonnay White 1.7 0.0 14.6 36.2 42.5 2015 Central Coast Viogner White 1.4 0.0 13.1 26.6 27.5 2014 Central Coast White Blend White 1.4 0.0 14.7 22.7 39.4 2015 Napa Valley Chardonnay White 1.3 0.0 15.7 37.9 54.2 2015 Napa Valley Moscato White 0.5 0.0 13.1 42.2 24.7 N.V. Brut Sparkling White 0.4 0.0 46.8 33.6 81.9 Monomeric Anthocyanin Anthocyanin vs. Metastable Bisulfite Complexes 150.0 100.0 50.0 y = 4.03x 11.61 R² = 0.68 0.0 0.0 10.0 20.0 30.0 40.0 Metastable Bisulfite Complexes Other methods for Truly Free GC/Mass Spectrometry (Carrascon et al. 2017) Capillary Electrophoresis (Collins & Boulton, 1996) Carrascon, V., I. Ontañón, M. Bueno, and V. Ferreira. 2017. Gas chromatography mass spectrometry strategies for the accurate and sensitive speciation of sulfur dioxide in wine. J Chromatogr 1504:27 34. Collins, T., and R.B. Boulton. The analysis of free sulfur dioxide and sugars in juices and wines by capillary electrophoresis. In Proceedings of the 5th Symposium International d Oenologie. A. Lonvaud Funel (ed.), pp. 637 640. Proceeding of Oenologie. 12
Sulfur Chemiluminescence Reaction Principle Furnace: R S + O 2 + H 2 + other products of combustion + H 2 SO + H 2 0 Reaction Cell: SO + O 3 SO * 2 + O 2 * + UV Light Example (13% ethanol, ph 3.5) USING pka = 1.81 How much Free to achieve 0.8 molec.? USING pka 2.06 ( TRUE pka AT 13% ETHANOL) How much Free to achieve 0.8 molec.? F = M (1 + 10^(pH pka) ) F = 0.8 (1 + 10^(3.50 1.81) ) F required = 40.0 ppm F = M (1 + 10^(pH pka) ) F = 0.8 (1 + 10^(3.50 2.06) ) F required = 22.8 ppm 13
Sensory Implications 2.4 2.2 2 1.8 20 30 40 50 60 70 80 90 mg/l free SO2 Model wine 13% alcohol (true pka = 2.06) 1.6 Molecular SO2 1.4 1.2 1 0.8 10 0.6 0.4 0.2 0 2.80 2.90 3.00 3.10 3.20 3.30 3.40 3.50 3.60 3.70 3.80 3.90 4.00 ph 14
2015 Viogner Central Coast 14.3% ethanol (pka = 2.20) 3.45 ph 23.0 C SCD1 B, Back Signal (TJ180313A\TJ180313A 2018-03-13 09-05-18\15 WH CHARDONNAY1.D) SCD1 B, Back Signal (TJ180313A\TJ180313A 2018-03-13 09-05-18\15 WH CHARDONNAY2.D) SCD1 B, Back Signal (TJ180313A\TJ180313A 2018-03-13 09-05-18\15 WH CHARDONNAY3.D) pa 3000 2500 (0.84 ppm molecular ) HH 15.9 ppm Free 2000 1500 3 analytical replicates % CV = 2.15% 1000 500 H 2 S EMS (ISTD) DMS MeSH 0 1 2 3 4 5 6 7 min Implications for wine oxidation Truly Free a better predictive tool for wine oxidation? Godden, P., L. Francis, J. Field, M. Gishen, A. Coulter, P. Valente, P. HØJ, and E. Robinson. 2008. Wine bottle closures: physical characteristics and effect on composition and sensory properties of a Semillon wine 1. Performance up to 20 months post bottling. Australian Journal of Grape and Wine Research 7:64 105. 15
When you can measure what you are speaking about and express it in numbers, you know something about it; but when you cannot measure it, when you cannot express it in numbers, your knowledge is of a meager and unsatisfactory kind: it may be the beginning of knowledge, but you have scarcely, in your thoughts, advanced to the state of a science. Lord Kelvin, Electrical Units of Measurement 16