The Influence of Cap Management and Fermentation Temperature Larry Lerno, Cristina Medina Plaza, Jordan Beaver, Konrad Miller, Siriwan Panprivech, Ravi Ponangi, Leanne Hearne, Tom Blair, Anita Oberholster, and David Block Driving innovation in grape growing and winemaking The Influence of Cap Management and Fermentation Temperature The Problem: Controlling phenolic extraction Winery studies Effects of cap and must temperature on phenolic extraction Effects of pumpover volume and frequency Examining phenolic gradients during fermentation Lab studies Adsorption and desorption of anthocyanins/tannins on grape cell wall material How do we connect the two? 1
The Problem: Controlling phenolic extraction Add a second temp sensor Cap/Skins Color Anthocyanin Monomeric Phenolics Tannin Juice T Closer look at the molecules extracted Polymeric flavan-3-ols (tannins) The most abundant class of phenolics in grapes Present in skins and seeds Anthocyanins Malvidin-3-glucoside is the predominant anthocyanin Found in the skin Hydroxycinnamates Ex: caftaric acid, caffeic acid, coumaric acid Found in the skin and pulp (Adams 26) 2
Closer look at the molecules extracted Grape tannins-oligomers catechin (C), epicatechin (EC), epigallocatechin (EGC), and epicatechin gallate (ECG) Differences between skin and seed tannins Mean degree of polymerization (mdp) for skin tannin are ~3; seed tannins are ~1 (Souquet et al. 1996) Proportion of ECG units is different in seeds (~3%) and skins (~5%) (Cheynier et al. 26) Hypothetical tannin tetramer (Adams 26) A molecular mechanism for cap extraction Epidermal Cell Layer Hypodermal Cells Mesocarp Cells Adsorption/Desorption Polymeric Phenolics Release Reaction Structural Carbohydrate Vacuole (containing phenolics) Cell Nucleus Monomeric and Polymeric Phenolics All steps are likely a function of temperature and EtOH 3
Experimental winemaking Grapes: Cabernet Sauvignon from Lodi, CA (example 212) 24.3 Brix ph = 3.85 T.A. = 3.8 g/l (adjusted to 5.97 g/l) YAN adjusted to 3 ppm, addition of ppm SO 2 Inoculated with S. cerevisiae strain Lalvin D254 Pressed after 14 days 7-9 days extended maceration Sampling: AM & PM till dry, then AM only Fermentations performed in triplicate using Cypress/UC Davis Research Fermentors (TJs) Effects of Cap and Must Temperature on Phenolic Extraction Driving innovation in grape growing and winemaking 4
Temperature treatments (212) Set Treatment Liquid Temp Cap Temp Cap Management Regime Control (.5) 25 C n/a PO.5 volume, 2/day A Control (1) 25 C n/a PO 1 volume, 2/day Control (2) 25 C n/a PO 2 volumes, 2/day 1 2 2 B 2 25 25 3 3 3 As needed to maintain temp 4 35 35 1 2 25 C 2 25 3 As needed to maintain temp 3 3 35 All treatments are in triplicate. Measured phenolic profile at various time points in each fermentation TJ Fermentors Cap temp probe Must temp probe IFCS Jacket temp probe Density meter Water outlet Water inlet Screen Pump sha 5
The effect of cap and liquid temperature 2 Average Malvidin 3 O glucoside Concentra on Mlv 3 O glu (mg/l) 1 1 48 96 144 192 24 288 336 Time (Hours) 2M2C 25M25C 3M3C 35M35C Effect of Cap and Liquid Temperature Average Ca aric Acid Concentra on Ca aric acid (mg/l) 4 3 2 1 48 96 144 192 24 288 336 Time (Hours) 2M2C 25M25C 3M3C 35M35C Average Catechin Concentra on Catechin (mg/l) 7 6 4 3 2 1 48 96 144 192 24 288 336 Time (Hours) 2M2C 25M25C 3M3C 35M35C 6
Is the main driver for extraction the liquid temperature or cap temperature? 7 Average Catechin Concentra on 6 Catechin (mg/l) 4 3 2 2M2C 25M25C 3M3C 35M35C 1 48 96 144 192 24 288 336 Time (Hours) Is the main driver for extraction the liquid temperature or cap temperature? 7 Average Catechin Concentra on 6 Catechin (mg/l) 4 3 2 2M2C 25M25C 3M3C 35M35C 2M25C 25M3C 3M35C 1 48 96 144 192 24 288 336 Time (Hours) 7
Learnings from temperature studies Extraction from skins happens early and seeds happens later Extraction of seeds is more temperature dependent Liquid temperature seems to be more important than cap temperature Effects of Pumpover Volume and Frequency Driving innovation in grape growing and winemaking 8
Pumpover treatments (212) Set Treatment Liquid Temp Cap Temp Cap Management Regime Control (.5) 25 C n/a PO.5 volume, 2/day A Control (1) 25 C n/a PO 1 volume, 2/day Control (2) 25 C n/a PO 2 volumes, 2/day 1 2 2 B 2 25 25 3 3 3 As needed to maintain temp 4 35 35 1 2 25 C 2 25 3 As needed to maintain temp 3 3 35 All treatments are in triplicate. Measured phenolic profile at various time points in each fermentation Effect of Pumpover Volume 2 Average Malvidin 3 O glucoside Concentra on Mlv 3 O glu (mg/l) 1 1 48 96 144 192 24 288 336 Time (Hours).5V2X 1V2X 2V2X 9
Effect of Pumpover Volume Average Ca aric Acid Concentra on Ca aric acid (mg/l) 4 3 2 1 48 96 144 192 24 288 336 Time (Hours).5V2X 1V2X 2V2X 6 Average Catechin Concentra on Catechin (mg/l) 4 3 2 1 48 96 144 192 24 288 336 Time (Hours).5V2X 1V2X 2V2X Examining pumpover frequency Set Treatment Liquid Temp Cap Temp Cap Management Regime 1 25 NC 1/2 vol 8x per day A 2 25 NC 1 vol 4x per day 3 25 NC 2 vol 2x per day 4 25 NC 4 vol 1x per day 1 1 day Cold Soak, 25 NC 2 vol 2x per day B 2 4 day Cold Soak, 25 NC 2 vol 2x per day 3 7 day Cold Soak, 25 NC 2 vol 2x per day 4 1 day Cold Soak, 25 NC 2 vol 2x per day For 213, pressed all batches at 8 days. 1
Pumpover frequency does not affect phenolic extraction 45 Ca aric acid concentra on at different pump over frequency 4 35 Ca aric acid (mg/l) 3 25 2 15 1 PO1X PO2X PO4X PO8X 5 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hours) 55 45 4 Catechin concentra on at different pump over frequency Catechin (mg/l) 35 3 25 2 15 1 5 PO1X PO2X PO4X PO8X 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hors) RP-HPLC Pumpover frequency does not affect phenolic extraction 16 Malvidin 3 glucoside concentra on at different pump over frequency 14 Malvidin 3 glucoside (mg/l) 12 1 8 6 4 PO1X PO2X PO4X PO8X 2 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hoours) 14 Tannin concentra on at different pump over frequency 12 1 Tannin (mg/l) 8 6 4 PO1X PO2X PO4X PO8X 2 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hours) RP-HPLC 11
No change in color Color density during fermenta on Absorbance (42+52+62) 9 8 7 6 5 4 3 2 1 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hour) PO1X PO2X PO4X PO8X Absorbance (42+52+62) 9 8 7 6 5 4 3 2 1 Color density of finished wine a a a PO1X PO2X PO4X PO8X *Pathlength (cm) = 1 a UV-VIS mg/l malvidin 3 glucoside equivalents 7 6 4 3 2 1 No effect on anthocyanin or tannin extraction Amount of Anthocyanins during fermenta on 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hour) PO1X PO2X PO4X PO8X mg/l malvidin 3 glucoside equivalents 6 4 3 2 1 Anthocyanins of finished wine a a PO1X PO2X PO4X PO8X a a mg/l catechin equivalents Amount of Tannins during fermenta on 4 4 3 3 2 2 1 1 15 3 45 6 75 9 15 12 135 1 165 18 195 Time (hour) PO1X PO2X PO4X PO8X mg/l catechin equivalents 4 4 3 3 2 2 1 1 Tannins of finished wine a a a PO1X PO2X PO4X PO8X UV-VIS/AH Correlation a 12
No effect on phenolic profile of finished wine RP-HPLC Examining pumpover frequency (214) Set Treatment Liquid Temp Cap Temp Cap Management Regime 1 25 NC 1/2 vol 2x per day A 2 25 NC 1/2 vol 1x per day 3 25 NC 1/4 vol 2x per day 4 25 NC No Pumpover 1 1 day Cold Soak, 25 NC 2 vol 2x per day B 2 4 day Cold Soak, 25 NC 2 vol 2x per day 3 7 day Cold Soak, 25 NC 2 vol 2x per day 4 1 day Cold Soak, 25 NC 2 vol 2x per day 13
Even lower levels of pumpovers do not have much effect on phenolic extraction 3 Malvidin-3-O-glucoside Tannin Mlv-3-O-glucoside (mg/l) 2 2 1 1 24 48 72 96 12 144 168 192 Time (Hours) Tannin (mg/l Catechin) 4 4 3 3 Half vol., 2X 2 Half vol., 1X 2 Quarter vol., 2X 1 No PO 1 24 48 72 96 12 144 168 192 Time (Hours) Half vol., 2X Half vol., 1X Quarter vol., 2X No PO Even lower levels of pumpovers do not have much effect on phenolic extraction 45 Caftaric acid 6 (+) Catechin Caftaric acid (mg/l caffeic acid) 4 35 3 25 2 15 1 5 24 48 72 96 12 144 168 192 Time (Hours) (+)-Catechin (mg/l) 4 Half vol., 2X 3 Half vol., 1X Quarter vol., 2X 2 No PO 1 24 48 72 96 12 144 168 192 Time (Hours) Half vol., 2X Half vol., 1X Quarter vol., 2X No PO 14
Even lower levels of pumpovers do not have much effect on final phenolic profile at 5 months Learnings from pumpover studies Pumpover volume does not make a difference in phenolic extraction (for volumes tested) Pumpover frequency does not make a difference in phenolic extraction 15
Is there spatial variation in phenolics during fermentation? Driving innovation in grape growing and winemaking Are there chemical gradients in red wine fermentations? Installed a curtain of 66 temperature sensors throughout the cross-section of a 2 L tank 15 sample extraction points Fermented 2 Tons of Cabernet Sauvignon Pumped-over 1 tank volume 2x per day Peristaltic pump on catwalk for sample extraction 16
Results Free Anthocyanin 1 2 3 4 6 7 24 hr 4 hr 48 hr 64 hr 72 hr 88 hr 96 hr 112 hr 12 hr 136 hr Units = mg/l Tannin 1 1 2 2 Units = mg/l 24 hr 4 hr 48 hr 64 hr 72 hr 88 hr 96 hr 112 hr 12 hr 136 hr 17
Mean Degree of Polymerization (mdp) 8 9 1 11 12 13 14 mdp decreases in cap over time Likely change from skin extraction to seed extraction 4 hr 64 hr 88 hr 112 hr 136 hr Percent Galloylation 4 hr 64 hr 88 hr 112 hr 136 hr 1 2 3 4 5 6 Percent galloyation is the percentage of ECG subunits of tannins (Seeds>Skins) Suggests increased extraction of tannins from the seeds at later times 18
Learnings from gradient studies Gradients in phenolics do exist Skin extraction is early, seed extraction is late Early cap samples may indicate final tannin level needs more work. Examining gradients in more detail may allow us to understand extraction at a more fundamental level How do we explain these results using a fundamental understanding? 19
A molecular mechanism for cap extraction Epidermal Cell Layer Hypodermal Cells Mesocarp Cells Adsorption/Desorption Polymeric Phenolics Release Reaction Structural Carbohydrate Vacuole (containing phenolics) Cell Nucleus Monomeric and Polymeric Phenolics All steps are likely a function of temperature and EtOH Adsorption Experiments to Measure Model Constants 7 8 A 5 8 O 2 4 C 3 2' O 3' B 6' O OH 4' 5' CH 2OH OH OH Purified Anthocyanins from Cabernet Sauvignon skins Purified Cell Wall Material from Thompson seedless grapes Analysis of Adsorption and Desorption Kinetics by LC-DAD/MS 2
UC DAVIS VITICULTURE AND ENOLOGY Adsorption decreases with Ethanol Adsorption Results concentration Adsorption decreases with Temperature Anthocyanin concentration % EtOH-No impact 15% EtOH- Key concentration above 1.5 mg/ml Plateau reached after 6 minutes Using fundamental knowledge to predict behavior at commercial scale Driving innovation in grape growing and winemaking 21
Developing a mathematical model based on fundamental knowledge Yeast growth model Chemical and physical properties of juice and Model to predict the spatial distribution of temperature and Combined ethanol fluid flow and heat transfer model solids Predictions of Temperature Gradients in Red Wine Fermentors L Tank T_max = 31.2 C T_ave = 27.3 C 16L Pilot Tank T_max = 35.2 C T_ave = 28.5 C 3L Commercial Tank T_max = 34.9 C T_ave = 28.3 C 22
Predictions of Ethanol Gradients in Red Wine Fermentors L Tank T_max = 31.2 C T_ave = 27.3 C 16L Pilot Tank T_max = 35.2 C T_ave = 28.5 C 3L Commercial Tank T_max = 34.9 C T_ave = 28.3 C Scales differ for the three tanks Developing a mathematical model based on fundamental knowledge Model to predict phenolic extraction as a function of temperature and EtOH Yeast growth model Chemical and physical properties of juice and solids Model to predict the Integrated Model to spatial Predict distribution Phenolic of Extraction temperature and Combined ethanol fluid flow and heat transfer model 23
Summary Temperature affects overall extraction of phenolics (more seed than skin) Pumpover volume and frequency do not seem to affect phenolic profile Adsorption and desorption of phenolics from cell wall material are functions of temperature and ethanol Modeling will connect all of this knowledge and allow prediction of the effects of processing and equipment design Acknowledgements Chik Brenneman Cyd Yonker Nick Dokoozlian T.J. Rodgers Cypress Semiconductor E. & J. Gallo Winery Ernest Gallo Endowed Chair in Viticulture and Enology 24