Appliction of Tosted Ok nd Micro-oxygention to Ageing of Cernet Suvignon Wines Dr. Jeff McCord StVin Inc., Cliforni, USA Introduction For mny yers, tosted ok in the form of rrels hs een used for wine storge nd to influence the geing of white nd red wines. Alterntives to rrels, most notly ok chips, hve lso een utilised to enhnce the chrcter of ulk wines fermented nd ged in stinless steel (SS) tnks. However, these wines hve rrely mtched the complex tstes nd roms imprted y rrels. The use of tosted stves in SS tnks for geing of wines hs provided gret improvement, yielding more rrel-like flvour profile. This pproch, however, is still not le to mtch the geing chrcteristics tht rrels imprt through slow lekge of ir into wines. In the mid 1980 s, Dr. Michel Moutounet egn reserch on technique now known s micro-oxygention. This technique involves the ddition of very smll mounts of pure oxygen into wine over time. This process, while thought to e new nd revolutionry, is in relity refinement of the rrel geing process. There is slow, very slow, permetion of ir through ok stves nd joints of the rrel; however, most ir contct with wine in rrels rises through the removl of the ung. The process of ung removl to top up rrels, dditions of SO 2 nd rcking, ll llow oxygen to e sored into wines. How these processes re performed gives the winemker some control over the mount of ir the wine sors, nd therefore the wine style. The use of micro-oxygention devices hs given the winemker level of control not previously ville in the production of wine. Furthermore, the use of these devices, in comintion with tosted ok in SS tnks, cn encourge mturtion nd complexity in wines similr to those mde using rrels. There hs een considerle controversy surrounding the use of microoxygention since its introduction s new winemking tool. Most of this controversy is result of the sujective nture of trils conducted to dte; this highlights need for lrge-scle ojective trils. The im of this study ws to determine the effect of tosted ok products with nd without micro-oxygention on the geing of Cernet suvignon wine. A second im ws to evlute vrious nlyticl methods for their ility to provide n effective nd useful tool to monitor the mturtion of wine undergoing micro-oxygention. At present, the most effective tool winemkers hve is their own experience in the production of their wines, nd their judgement of their wine s ility to ge. Mterils nd Methods Cernet suvignon fruit ws hrvested from the Edn Vlley region of Sn Luis Oispo County, Cliforni in 2001. Approximtely 500 tons ws hrvested t n verge of 22.5? Brix in erly Novemer. Grpes were divided into fermenttion vessels vrying from 10 ton to 20 ton in size with n ddition of 50 ppm SO 2 t the crusher. Fermenttion trils were conducted to determine the effects of enzymes, tosted ok 1
ddition nd type of fermenttion, compring punch down with pump over techniques. All fermenttions hd enough Chrdonny concentrte dded to ring the finl lcohol concentrtion to 14%. The must ws inoculted with yest CY3079 nd ws pressed off t 0? Brix. After pressing, ll the wine ws comined together nd then divided into duplicte lots for the geing trils: no ok / no oxygen (control) to two 3,000 gllon (11,360 L) tnks; stves only / no oxygen to two 3,000 gllon (11,360 L) tnks; segments only / no oxygen to two 3,000 gllon (11,360 L) tnks; oxygen only / no ok to two 12,000 gllon (45,420 L) tnks; stves nd oxygen to one 6,000 gllon (22,710 L) tnk nd one 12,000 gllon (45,420 L) tnk; nd lstly, segments nd oxygen to one 6,000 gllon (22,710 L) tnk nd one 12,000 gllon (45,420 L) tnk. Tosted ok ws dded t 3 g/l in the form of StVin French Ok Medium Plus stves or StVin French Ok Medium Plus segments. The comined wine, efore division into tnks for individul tretments, ws inoculted with Oeno strin of mlolctic cteri. At completion of mlolctic fermenttion, ll tnks were clen rcked nd SO 2 ws dded t 30 ppm. Microoxygention tretments were dosed with oxygen for one month t 10 ml O 2 /L/month then reduced to 5 ml O 2 /L/month for the reminder of the tril, using two 4-chnnel Ox Boxes from StVin Inc. Dosing of oxygen egn 15 Jnury 2002 nd ws hlted on 15 June 2002. Wines were ottled from their respective tnks in August of 2002. Anlysis Wines were nlysed y ETS Lortories in St. Helen, Cliforni for 48 chemicl ttriutes: ethnol, protein rective tnnins, trns-ok lctone, cis-ok lctone, vnillin, furfurl, 5-methylfurfurl, guicol, 4-methylguicol, eugenol, isoeugenol, gllic cid, ctechin, epictechin, polymeric phenols, cftric cid, cffeic cid, quercetin glycosides, quercetin glycone, mlvidin glucoside, polymeric nthocynins, totl nthocynins, monomeric nthocynins, hydrogen sulfide, ethyl mercptn, methyl mercptn, diethyl sulfide, dimethyl sulfide, diethyl disulfide, dimethyl disulfide, other sulfides, 4- ethylguicol, 4-ethylphenol, free sulfur dioxide, ph, titrtle cidity, totl sulfur dioxide, voltile cidity, sornce t 420 nm, sornce t 520 nm, the sum of sornces t 420 nm nd 520 nm, nd the rtio of sornces 420 nm/520 nm; tristimulus L,,, hue ngle, nd chrom vlues. Further nlysis ws performed y Felipe Lurie, under the direction of Dr. Andrew Wterhouse, Deprtment of Viticulture nd Enology, University of Cliforni, Dvis. This nlysis mesured the Adms ssy, tnnin, totl colour, lrge polymeric pigments, smll polymeric pigments nd monomeric pigments. Norml phse HPLC ws used for nlysis of monomers, oligomers nd polymers soring t 280 nm, nd for monomers, oligomers nd polymers soring t 520 nm. Reverse phse HPLC ws used for nlysis of ctechin, epictechin, gllic cid, cffeic cid nd quercetin. Anlysis of vrince ws performed on the dt using MiniT sttisticl pckge, Minit Inc., Stte College, PA, USA. Results nd Discussion Of ll the chemicl ttriutes mesured y ETS nd t UC Dvis, only those found to differ etween tretments t the 95% confidence level will e discussed. 2
Totl nd free SO 2 Figure 1 illustrtes the difference etween oxygented nd non-oxygented wines for free nd totl SO 2 t eight months. Although Figure 1 only gives the 8 month snpshot of the wine, the mesurement of free nd totl SO 2 hs een found to e n effective tool for monitoring micro-oxygention of wines if done on weekly sis nd chrted. The rte of decrese of free SO 2 cn e indictive of how the wine is responding to microoxygention. Too rpid rte of decrese indictes too high rte of oxygention. Too low or no decrese would indicte tht the rte of oxygention could e incresed. Importnt points to e mde with monitoring free nd totl SO 2 concentrtions re the following: free SO 2 levels should not fll elow 10 ppm, to help ensure proper protection of the wine; decrese in totl SO 2 ppers to indicte tht the free SO 2 hs dropped elow 10 ppm leving the wine prone to severe oxidtion; nd totl SO 2 mesurements will not e dditive fter doing SO 2 dditions to mintin free SO 2 levels ove 10 ppm. For exmple with this study there were three 30 ppm dditions of SO 2. Free mercptns nd sulfides Figure 2 demonstrtes tht ll of the wines with micro-oxygention hd significntly lower level of oth methyl nd ethyl mercptn. The effect ws sufficient to lower the concentrtion of these two compounds elow their rom thresholds of 1 pp. Much concern hs een rised over the possile production of disulfides during the oxygention of wines contining ove rom threshold levels of mercptns. This study demonstrted no such formtion of disulfides; mesured levels of disulfides were not significntly different in ll tretments. Dimethyl sulfide (DMS) is usully ssocited with ottle geing, especilly in white wines, generting chrcteristic rom of cremed corn. Norml levels in red wines pproch 50 ppm nd re ssocited with rom complexity. Non-iologicl production of DMS is thought to occur s slow oxidtive process. Therefore, we expected n increse of DMS concentrtion in ll oxygented tretments. However, the results clerly show there ws not n increse of DMS concentrtion with the ddition of oxygen. Interestingly, significnt decrese in DMS level is seen in ll tretments with dded tosted ok. The cuse of this decrese is not understood. Spectrl effects Figures 3 & 4 demonstrte differences in spectrl chrcteristics etween tretments. All micro-oxygented tretments incresed in sornce, with n sornce mximum t 528 nm nd n sornce minimum out 420 nm. No shift in sornce wvelength ws detected t the mximum. There ws slight shift of sornce minimum from out 428 nm to 421 nm etween tretments. This shift my help to explin why certin wines do not pper to increse in rown colour s much s expected with microoxygention. Figure 5 is simpler illustrtion of the effect of oxygen nd ok on red colour (mesured t 520 nm) nd rown colour (mesured t 420 nm). In ll cses microoxygention incresed sornce t 520 nd 420 nm, incresing the pprent colour in the wines. 3
Reverse Phse HPLC Anlysis of Phenolics The wines were nlysed for numerous types nd clsses of phenolic compounds. Figure 6 demonstrtes the results using the Price HPLC method to determine protein rective tnnin, polymeric phenols, totl nthocynins nd mlvidin glucoside. Except for protein rective tnnin, only those compounds or clsses of compounds tht provided significnt differences re shown. The Price HPLC method identifies polymeric mteril s the re under the lst eluting pek tht is detected t 280 nm. The re under tht sme pek, when detected t 520 nm, is referred to s the polymeric nthocynin pek, s shown in Figure 7. Protein rective tnnin is mesured s the difference etween the polymeric pek of wine efore nd fter ddition of geltin. The ide eing tht the polymeric compounds tht rect with the geltin nd precipitte my e correlted with stringency or prticulr type of stringency, eg. hrsh, corse, or soft. This mesure ws thought to e potentilly useful indictor of mouth feel. However, cross the tretments we sw no significnt difference in protein rective tnnin t the 95% confidence level. However, there did pper to e trend towrds higher level of protein rective tnnin in the micro-oxygented tretments; thus this nlysis my wrrnt further investigtion. The polymeric phenol levels, s shown in Figure 6, were significntly different. The mount incresed with micro-oxygention nd lso with ok dded in the form of segments. This increse my correlte with decrese in monomeric procynidins, nd it will e discussed lter. This my indicte tht oth oxygen nd tosted ok encourge formtion of tnnin polymers. However, the type nd size of these polymers is very much in question s none of the methods used in this study were le to yield this informtion. Also shown in Figure 6 re the levels of totl nthocynin nd mlvidin glucoside, s mesured y HPLC. It is clerly demonstrted tht micro-oxygention decreses oth mlvidin glucoside nd totl nthocynin. Mlvidin glucoside lso demonstrted lrger decrese in the presence of tosted ok nd oxygen thn with either tosted ok lone or oxygen lone. This my indicte the occurrence of n interction etween oxygention of the wine nd compounds extrcted from tosted ok. At this point we cn hypothesise tht nthocynins re more effectively linked to oligomeric nd polymeric procynidins in the presence of tosted ok nd oxygen. To provide further support for this process, Figure 7 demonstrtes clerly n inverse reltionship etween monomeric nthocynins nd polymeric nthocynins. Agin, oxygen plus tosted ok, especilly the segment form, showed the most significnt polymeristion of nthocynins. Figure 8 compres two nlyticl methods tht mesure the overll tnnin concentrtion of wines, the Folin-Cioclteu method nd the newer Adms tnnin ssy. Here, the Folin method did not demonstrte ny cler difference etween tretments. However, the Adms method showed cler trend etween tretments. In the sence of oxygen, tosted ok stves slightly incresed the tnnin level, compred to the control; it ws further incresed with tosted segments. Higher tnnin levels occurred with ll oxygented smples, with or without tosted ok. Thus, the Adms method does pper to e cple of discerning differences etween the treted wines. Unfortuntely, sufficient work hs not een conducted to correlte these higher Adms tnnin vlues to the sensory perception of the wines. 4
Norml Phse HPLC Anlysis of Phenolics Norml phse nlysis of wine phenolics cn etter distinguish nd quntify moleculr weights of these clsses of compounds. The detected compounds were grouped into three clsses: monomers, oligomers (up to 7 ctechin units) nd polymers (greter then 7 ctechin units). Using two sornce wvelengths, 280 nm nd 520 nm, differences etween polymers with nd without covlently onded nthocynin cn e distinguished. Figure 9 depicts the results etween tretments t 520 nm. Figure 10 depicts similr trends t 280 nm. While error rs re shown in Figures 9 nd 10, the significnce of the dt ws not clculted. So only generl oservtions cn e mde for these results. The results gree with those shown in figures 6 & 7, derived y reverse phse HPLC nlysis, in tht monomers with 520 nm sornce tend to decrese with oxygention. It ws hoped tht significnt difference in quntity of oligomers nd polymers might e seen etween oxygented nd non-oxygented wines. Unfortuntely, there ws no such trend, nd norml phse HPLC nlysis did not yield usle winemking informtion. Other Phenolic sed compounds Three other compounds nlysed y the Steve Price reverse phse HPLC method were: cftric cid, useful compound for indicting the impct of oxidtion on white wines; quercetin glycone, potentilly itter compound; nd epictechin, uilding lock of tnnins. Figure 11 shows tht the tretment using tosted ok segments without oxygention mintined the highest concentrtion of cftric cid, indicting etter ility of this tretment to protect the highly oxidizle compound. Quercetin glcone showed significnt decrese in concentrtion in ll tretments with micro-oxygention. This my e prtil explntion for the perceived smoothness of oxygented wines, lthough this hs not yet een proven with quntittive sensory nlysis. Also seen is decrese in epictechin cross ll tretments with dded oxygen. This decrese in concentrtion grees with work (Vidl et l. 2002) using wine-like model solution, tht hs shown tht flvnol monomers, such s epictechin, rect with products of the rekdown of polymeric pronthocynidins. This leds to reduction in the verge size of the pronthocynidin polymers, nd tht chnge of polymer size my influence the perceived stringency. This my help to explin frequent decrese of perceived hrshness in wines with micro-oxygention. Figure 12, confirms the findings of the Price reverse phse HPLC method. These results, from the Wterhouse Lortory, used norml phse HPLC nd show very similr trends for these smll moleculr weight phenolic compounds. Effects on Ok Extrctives Figure 13 shows the undnce of six compounds known to e extrcted from ok nd found in ll tretments t less then 100 pp. All tretments hd detectle mounts of these six compounds, even though some tretments hd no dded tosted ok, ecuse the geing tril ws performed on comintion of wines tht received vrious tretments during fermenttion, including some tosted ok ddition. No significnt tretment effect ws found for trns- nd cis-ok lctone, guicol, 4 methylguicol or eugenol. A slight effect ws noted for isoeugenol. Both eugenol nd isoeugenol hve 5
clove like rom, nd isoeugenol my e slightly more prone to oxidtion then eugenol. The tretments with dded tosted ok showed significnt increse in ll compounds over tretments with no ok ddition. Tosted ok segment tretments, with or without micro-oxygention, were very similr. The tosted ok stve tretment showed lower concentrtions for most of these six compounds, indicting lower or slower extrctions rtes. Figure 14 shows the undnce of ok-derived compounds detected ove 100 pp in some tretments: 5 methylfurfurl, vnillin nd furfurl. As expected, oth tretments without dded tosted ok showed low concentrtions of these three compounds. However, the tretment with tosted ok stves lso showed low concentrtions of the three compounds. Interestingly, the oxygented tretment with ok stves showed similr concentrtions to the tosted ok segment tretments. Knowing tht exctly the sme mount of tosted stves ws dded to oth the non-oxygented nd oxygented tretment led us to speculte tht there my e n interction etween stves nd oxygen. One possile explntion is tht oxygention of the wine produces stronger driving force to extrct these compounds from the stves. Why this is not seen with the segments my e due to the higher exposure of end grin, which is known to increse extrction in tosted ok products, equlizing extrction rtes for this product. Further work needs to e done to clrify the phenomenon nd its effect on the resulting wine. Conclusions This study hs provided conclusive evidence tht micro-oxygention of wine during geing leds to n increse of colour intensity nd concurrent decrese in free mercptn concentrtions without n increse of disulfide concentrtions. We hve lso demonstrted n interction etween the form of tosted ok nd micro-oxygention. Further work is eing conducted to relte the effects seen in this study with the sensory impct of the resulting wines. We hve lso shown direct reltionship etween the decrese in monomeric nthocynins nd the increse in polymeric nthocynins with the ddition of tosted ok, with micro-oxygention, or with the comintion of the two. This my led to softening of the wine when monomeric nthocynins re ound to polymers, s demonstrted in the tlk erlier y Dr. Leigh Frncis. This study hs lso shown tht micro-oxygention led to decrese in monomeric procynidins, epictechin, nd n increse in polymeric phenolics. Both of these oservtions tend to support the perceived softening of wine tht hs een micro-oxygented. The quest for strightforwrd nlyticl methods to monitor the effects of microoxygention nd to id the winemker in the use of micro-oxygention hs still to e met. This study did show tht mny complex nlyticl methods re not dequte lone to monitor oxygention. However, certin simpler methods, such s the Adms tnnin method do show promise s simple tool. Much more work should e done with the Adms method to relte results to tretment effects on wine nd, most of ll, to sensory implictions. Acknowledgments Christin Roguennt, Orcutt Rod Cellrs Gordon Burns, Dr. Steve Price nd Eric Herve, ETS Lortories Felipe Lurie nd Dr. Andrew Wterhouse, Dept. Viticulture nd Enology, UC Dvis Leslie Norris, Flvorsense Reecc Bleium, Trgon Corportion 6
References Vidl, S.; Crtlde, D.; Souquet, J.-M.; Fulcrnd, H. nd Cheynier, V. (2002). Chnges in pronthocynidin chin length in winelike model solutions. J. Agric. Food Chem 50(8): 2261-2266. Adms,D.O.nd J.F. Hrertson. Use of lkline phosphtse for the nlysis of tnnins in grpes nd wine. AJEV. 1999, 50, 247-252. Hrertson,J.F., J.A. Kennedy nd D.O. Adms. Tnnins in skins nd seeds of Cernet Suvignon, Syrh, nd Pinot Noir erries during ripening. AJEV, 2002, 53, 54-59. Wterhouse, A.L., S. Ignelzi, nd J. R. Shirley. A comprison of methods for quntifying oligomeric pronthocynidins from grpe seed extrcts. AJEV, 2000, 383-389. Wterhouse, A.L., S.F. Price, J.D. McCord. Reversed-Phse High-Performnce Liquid Chromtogrphy Methods for Anlysis of Wine Polyphenols. Methods in Enzymology,199,8 299: 113-121. 7
60 50 free sulfur dioxide totl sulfur dioxide 40 ppm 30 20 10 0 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 Figure 1. Anlysis of free nd totl SO 2 t 8 months of geing. 8
1.2 50 1 48 46 ug/l 0.8 0.6 44 42 40 ug/l 38 0.4 36 34 0.2 32 0 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 ethyl mercptn methyl mercptn dimethyl sulfide 30 Figure 2. Anlysis of thiols nd sulfides. 0.7 0.6 Asornce 0.5 0.4 0.3 0.2 0.1 0 380 394 408 422 436 450 464 478 492 506 520 534 548 562 576 590 604 618 632 646 660 674 688 702 716 730 744 758 No O2 No Ok St No O2 Seg No O2 O2 No Ok St O2 Seg O2 Wvelength Figure 3. Spectrl comprisons etween tretments. 9
0.6 0.55 0.5 Asornce 0.45 0.4 0.35 0.3 0.25 400 420 440 460 480 500 520 540 560 580 No O2 No Ok St No O2 Seg No O2 O2 No Ok St O2 Seg O2 Figure 4. Enlrged 400 580 nm region of spectrl comprisons etween tretments. 10
0.6 0.55 0.5 sornce 420 sornce 520 Asornce 0.45 0.4 0.35 0.3 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 Figure 5. Asornce t 420 nd 520 nm of ll tretments. 11
800 700 600 c c c 500 mg/l 400 300 200 c c c 100 0 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 Protein Rective Tnnins polymeric phenols totl nthocynins mlvidin glucoside Figure 6. Reverse phse HPLC results for mjor phenolic clsses. 12
54 53 c 390 52 51 c c c 370 50 350 mg/l 49 330 mg/l 48 310 47 46 290 270 45 44 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 250 polymeric nthocynins monomeric nthocynins Figure 7. Comprison of polymeric nthocynins vs monomeric nthocynins. 13
2900 960 Gllic Acid Equivlents (mg) 2800 2700 2600 2500 2400 2300 2200 Control Stves Only Segments O2 only Stves + O2 Segments + O2 950 940 930 920 910 900 890 880 870 860 850 Tnnic Acid Equivlents (mg) Folin-Cioclteu Adms Tnnin Figure 8. Stndrd tnnin nlysis comprison for tretments, Folin Cioclteu ssy nd the Adms tnnin ssy 14
450 400 350 300 250 200 150 100 50 0 Control Stves Only Segments O2 only Stves+O2 Segm.+O2 Monomers Oligomers Polymers Figure 9. Norml phse HPLC of monomers, oligomers nd polymers soring t 520 nm. 15
4500 8000 4000 3500 3000 2500 2000 1500 1000 500 7500 7000 6500 6000 5500 5000 4500 0 Control Stves Only Segments O2 only Stves+O2 Segm.+O2 4000 Monomers Oligomers Polymers Figure 10. Norml phse HPLC of monomers, oligomers nd polymers soring t 280 nm. 16
30 25 20 mg/l 15 10 5 0 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 cftric cid quercetin glycone epictechin Figure 11. HPLC nlysis of cftric cid, quercetin glycone, nd epictechin. 17
mg/l 45.00 40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00 Control Stves Only Segments O2 only Stves + O2 Segments + O2 200.00 180.00 160.00 140.00 120.00 100.00 80.00 60.00 40.00 20.00 0.00 mg/l Ctechin Epictechin Gllic cid Cffeic cid Quercetin Mv-3-glc Figure 12. Norml phse HPLC nlysis of ctechin, epictechin, gllic cid, cffeic cid, quercetin, nd mlvidin-3-glucoside 18
60 c c 50 40 c ug/l 30 c c 20 10 0 c e d d Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 trns-ok Lctone cis-ok Lctone guicol 4-methyl guicol eugenol isoeugenol Figure 13. Anlysis of minor ok extrcted compounds. 19
160 1000 140 900 800 ug/l 120 100 80 60 40 20 700 600 500 400 300 200 100 ug/l 0 Control Ave Ave St no O2 Ave Seg no O2 Ave O2 only Ave St + O2 Ave Seg + O2 5-methyl-furfurl vnillin furfurl 0 Figure 14. Anlysis of mjor ok extrcted compounds. 20