Partial Dealcoholization of Wine Comparing Methods for dealcoholization Dr. Andreas Blank Wine Eng Conference K+H process tec GmbH Why are we talking about alcohol reduction? Wine Riesling early late Harvest 17 Sep 13 Oct 2011 2011 alcohol control chaptalized control reduced Alcohol 12%vol 14%vol 14%vol 12%vol GC-Analysis aroma compounds Riesling 30 25 2011 20 15 Linalool 10 rel. peak area peak rel. 5 0 C C A B 12%vol 14%vol 14%vol 12%vol Control chaptalized Control alcohol reduced early harvest late harvest Post hoc Fisher LSD (5%) Sensorial effect of ethanol wine „burning, hot, alcoholic“ „over-ripe“ „green“ „tropical“ alcoholic full-bodied full-bodied style style style content chaptalization „green“ „tropical“ Alcohol- Alcohol elegant elegant reduction „sour“ style style green style „thin“ ripeness © Blank Technologies: Overview • Technical dealcoholization – Distillation / Spinning Cone Column – Membrane contactor – Reverse Osmosis – Combinations Distillation 100 ] 90 mol 80 [% 70 Ethanol is more volatile than water 60 50 Distillation separates 40 concentration 30 20 phase 10 Gas 0 What about aroma compounds? 0 10 20 30 40 50 60 70 80 90 100 Liquid phase concentration [%mol] z.B. MCCABE und THIELE, 1925 Volatility of Aroma Compounds 3-Isopropyl-3-methoxypyrazin 2,3-Butandiol 2-Aminoacetophenon Essigsäure 2-Phenylethanol Milchsäureethylester Hexansäure beta Damascenon Low Volatiliy 3-Mercaptohexan-1- ol Terpineol Furfural Decansäureethylester 4-Ethylphenol Diacetyl Bernsteinsäurediethylester Citronellol Guaiacol 4-Ethylguaiacol Linalool Ethanol Ethanol 3-Mercaptohexylacetat Isobutanol 2-Phenylethylacetat 3-Methylbutanol Medium Volatility 4-Mercapto-4-methylpentan-2-on Hexanol Octansäureethylester Ethylacetat Isoamylacetat Isobuttersäureethylester Buttersäureethylester Hexansäureethylester Ethanthiol Methanethiol Blank, 2015 High VolatilityVinylbenzol H2S Own Data calculated from -8 -7 -6 -5 -4 -3 -2 -1 0 1 Archemcalc.org 2012 log H IKARI und KUBO, 1976 Spinning Cone Column Distillation Feed Distillate Reboiler Dealcoholized Product Spinning Cone Column A Aroma concentrate r²=0,92 8000 6000 Initial Wine high 15%vol 4000 2000 Konzentration 0 low medium -5 -4 -3 -2 -1 SCC „Aroma“ Recovery Aromalog H High volatile fraction of aroma compounds B Aroma reduced wine r²=0,91 2% 98% 150 „Aroma“ – Reduced Aroma concentrate 100 low wine 65% vol medium 13,5 %vol 50 Konzentration 0 high -5 -4 -3 -2 -1 log H Fraction analysis C Dealcoholized fraciton r²=0,84 D Distillat r²=0,88 120 400 350 100 medium low 300 80 250 60 200 150 concentration 40 concentration 100 20 50 medium high low high 0 0 -5 -4 -3 -2 -1 -5 -4 -3 -2 -1 log H log H Recombination in Spinning Cone Column E „Aroma“ + Dealcoholized fraction • Low and high volatile aroma r²=0,43 compounds are completely recovered 120 • Medium volatile aroma compounds 100 low high are lost equal to ethanol reduction. 80 60 • Problem: 40 medium concentration Many aroma compounds have similar 20 volatitlity compared to Ethanol 0 -5 -4 -3 -2 -1 log H Reverse Osmosis Alcohol Reduction Initial Wine Reverse Osmosis 15%vol Alcohol Reduction Reverse Osmosis 85% 15% Concentrate Permeate 15,2%vol 14,8%vol Reduced alcohol Water wine 15% 13,0%vol Separation by molecular size RO Separation Membrane A „tight RO“ Membrane B „loose RO“ A 100 B 100 90 90 80 80 70 [%] 70 60 60 50 rejection 50 40 40 rejection [%] 30 30 20 20 10 10 0 0 0 50 100 150 200 250 0 50 100 150 200 250 Molekular weight [g/mol] Molecular weight [g/mol] RO Separation – Take home • RO Alcohol reduction 100 Separation as molecular Umkehrosmose 90 sieve Nanofiltration 80 Extract compounds can be 70 lost relative Konzentration relative 60 Ethanol 50 0 50 100 150 200 250 Molecular weigt [g/mol] Membrane contactor – Osmotic Distillation Porous Membrane of Polypropylen hydrophobic Main Purpose(CO2/O2) Application in Alcohol Management Osmotic distillation GABELMANN und HWANG, 1998 SENGUPTA et al., 1998 BLANK et al., 2013 Alcoholmanagement using Membrane Contactors 16 14 12 ] vol 10 Wine „Strip“ 8 Water 6 Alcohol [% 4 2 0 0 200 400 600 Mass exchange time [min] in gaseous form Osmotic Distillation: Analytical changes control reduced Sign. Analyse mean mean p Alcohol reduction by 2%vol. Ethanol g/L 114,5 99,1 *** Sugar g/L 3,9 4,1 ** Total number of wines: n=15 Sugar free Extrakt g/L 21,0 21,4 * Residual extrakt g/L 10,6 11,0 ** 2 factoral ANOVA Density Ethanol0,9906 is removed0,9930 *** Treatment x Wine pH 3,42 3,41 ns ConcentrationTitratable acidity g/L of Non5,0 -volatiles.5,2 2***-4 % No significant Interactions LossTartrate of Gase g/L SO2 und1,8 CO2 1,9 ns Volatile acidity g/L 0,38 0,39 ns Shown is significance of Potassium mg/L 764 815 ** treatment Magnesium mg/L 85,0 86,7 ns SO2 free mg/L 47,6 39,5 ** CO g/L 1,0 0,8 * ns=non significant 2 * p=0,05 E 420 nm/1cm 0,5 0,5 ns ** p=0,01 *** p=0,001 Total phenols mg/L 678 687 ns Aroma compounds Flüchtigkeit Compounds Mean loss log H A 40% Hexyl acetate 26% -1,44 35% Isoamyl acetate 23% -1,73 Buttersäureethylester 21% -1,58 30% Isobutanol 20% -3,14 25% r² = 0,619 loss Isobuttersäureethylester 18% -1,62 20% Phenylethylacetat 18% -2,72 Mean 15% Hexanol 17% -2,89 3-Methylbutanol 16% -3,05 10% Ethyl acetate 13% -1,45 5% Linalool 9% -3,42 0% Bernsteinsäurediethylester 9% -3,91 -5 -4 -3 -2 -1 4-Vinylphenol 9% -4,10 Volatility log H Geraniol 8% -3,69 Nerol 3% -3,69 Reduzierung um 2%vol. 2-Phenylethanol 3% -4,80 n=12 different wines Terpineol 2% -4,27 Flüchtigkeit log H berechnet mit archemcalc.com Alcohol reduction / Aroma attributes Riesling 2012 10,8%vol 11,8%vol 12,8%vol 13,8%vol 3.0 2.6 2.7 2.4 2.4 2.4 2.4 2.4 2.4 2.5 2.4 2.2 2.3 2.3 2.3 2.2 5] 2.2 2.1 - 2.0 1.5 1.0 Intensität [0 Intensität 0.5 0.0 Attribut GeruchsintensitätTotal intensity Greengrüner apple Apfel Orange Tropicaltropisch fruit Fisher LSD 5% A|A|A|A A|A|A|A B|A|AB|B A|A|A|A ANOVA 0,295 0,997 0,049 0,784 Alkoholreduzierung: Taste Attributes Riesling 2012 10,8%vol 11,8%vol 12,8%vol 13,8%vol 3.0 2.5 2.5 2.3 5] 2.2 - 2.0 2.0 1.9 2.0 1.8 1.7 1.7 1.8 1.51.6 1.6 1.6 1.5 1.5 1.21.2 1.3 1.0 1.0 0.9 Intensität [0 Intensität 0.5 0.0 Attribut SweetSüße SourSäure Bitter Brandig/AlkoholAlcoholic GeschmBody Fülle Fisher LSD 5% C|BC|AB|A A|A|A|A BC|C|AB|A B|B|A|A B|AB|A|A ANOVA 0,005 0,833 0,006 < 0,0001 0,005 Osmotic Distillation: Take home • Machine fo disolved gas management Alcohol reduction: • Treat small volume strong – blend back to desired content • No significant difference in Wine Sensory • Aroma loss dependant on volatility • Concentration of extract compounds • Losses of SO2 and CO2 have to be replaced Comparing Technologies Technologies Distillation Reverse Osmosis Membrane Combination Reverse Osmosis Distillation contactor RO + MC RO+H2O 2 wines each in Duplicate Linalool: Alcohol reduction by 2%vol. Linalool: Terpen Molecular weight: medium - large 154 g/mol Volatility similar to Ethanol 120 100.0 103.1 111.1 100 81.7 76.9 area 80 60 40 peak 20 0 ControlKontrolle Osmot. Osmot. UmkehrosmoseCombination & Umkehrosmose & elative elative Distillation Membrane RO + Water r untreated Destillation ContactorDestillation RO+MCOsmot. Osmosewasser Teilmenge Gesamtmenge Destillation A B B A A Tukey HSD 5% Isoamyl acetate: Alcohol reduction by 2%vol. Isoamyl acetate: Molecular weight: medium 130 g/mol High volatile aroma compound 100.0 100 74.9 73.8 80 67.2 60 52.7 peakarea 40 20 relative relative 0 ControlKontrolle Osmot.Distillation DestillationOsmot.Membrane Destillation UmkehrosmoseCombination & UmkehrosmoseRO + Water & untreated Teilmenge ContactorGesamtmenge Osmot.RO+MC Destillation Osmosewasser A B C B B Tukey HSD 5% Comparing technologies 120 OD Teilmenge 100100 • Alcoholreduction always comes together 120 Umkehrosmose with loss of some volatiles 90 80100 105 OD Nanofiltration • Strong treatment of small volume reduces 80 60 Sensorial80 effect at 2%vol. (trainedGesamtmengSpinningpanel Cone) Column aroma loss to an acceptable minimum 10060 e 70 40 • No statistical detectable difference relative Konzentration relative 40 20 • 95 Ni differenceEthanolin triangle tests • Distillation (vacuum and membrane MC) Konzentration relative 60 20 Ethanol Ethanol 0 Konzentration relative • 90 No difference in descriptive test – Separtation by volatility 0 50 -5 -4 -3 -2 -1 • Reverse Osmoses 0 85-550 -4100 -3150 200-2 250-1 loglog H Flüchtigkeit H Flüchtigkeit Molekulargewicht [g/mol] – Separation by molecular size 80 1,0 relative Konzentration relative • Spinning Cone Column SCC 75 0,8 – Double distillation 0,6 70 -4,5 0,4 • Kombination RO + Distillaiton -4,0 -3,5 -3,0 0,2 – Combined separation -2,5 -2,0 Rückhaltevermögen log H -1,5 0,0 -1,0 K+H VinZero • Inline alcohol adjustment: • Automated alcohol adjustment using RO and Membrane contacter • In line alcohol probe Conclusion Technologies for dealcoholization • Positive for wine style modification VinZero • Separation by physical means – Volatility CO2 Membrane System – Molecular sieve – Combination • In-Line Alcohol reduction – Combi Contact K+H process tec GmbH [email protected] Einordnung der Alkoholreduzierung: Reifeentwicklung Riesling am Beispiel 2012 15.09 30.09 14.10 30.10 Unreif „Frühe“ Lese „Späte“ Lese Überreif Alkoholreduzierung: Sensorischer Vergleich mit Lesezeitpunkt Riesling Wein Vergleich Lesezeitpunkt unabhängig vom Alkoholgehalt Früh Spät Lesezeitpunkt 17.09.2011 13.10.2011 84°Oe 97°Oe Kontrolle angereichert Kontrolle reduziert Alkohol 12%vol 14%vol 14%vol 12%vol Blank et al., 2013 Aromaintensität und Lesezeitpunkt 30 24.2 23.1 Riesling 25 20 15 8.5 2011 10 8.0 5 0 12%volAromasteigerung14%vol durch14%vol späte Lese12%vol Linalool Peakfläche Relative Kontrolle angereichert Kontrolle alkoholred.