MUST ACETIC ACID AND ETHYL ACETATE AS MOLD AND ROT INDICATORS IN GRAPES

MUST ACETIC ACID AND ETHYL ACETATE AS MOLD AND ROT INDICATORS IN GRAPES C. A. Crisn, C. S. Ough, H. W. Berg, and K. E. Nelsn Respectively Graduate Student, Prfessr f Enlgy, Prfessr Emeritus f Enlgy, and Prfessr f Viticulture, Department f Viticulture and Enlgy, University f Califrnia, Davis 95616. The authrs wish t acknwledge the supprt frm the State Department f Agriculture, cntract N. 9585.. Presented at the Annual Meeting f the American Sciety f Enlgists, July 29, 1978, San Dieg, Califrnia. Manuscript submitted August 4, 1978. Revised manuscript received Nvember 6, 1978. Accepted fr publicatin Nvember 6, 1978. The threshld values f acetic acid and ethyl acetate were evaluated in white and red table wines. Musts were frtified t varius levels with these tw chemicals in rder t determine their fate during yeast grwth and fermentatin. Frm these data the rejectin levels f the cmpunds, when present in musts, were established fr bth red and white table ABSTRACT wines. Grapes mlded with a number f fungi were fermented and the rejectin levels fr ethyl acetate verified. These must rejectin levels fr ethyl acetate in white and red grapes were 6 and 115 mg/1 respectively. Fr acetic acid in white and red musts the rejectin levels were 119 and 9 mg/1 respectively. The extent f mld and rt cntaminatin has always been hard t quantify in harvested grapes. Grwer, wineries, cnsumers and gvernment agencies alike are becming increasingly cncerned with the use f sund fruit in the wine industry. The Califrnia Agricultural Cde requires that, when the purchase price f grapes fr crush is based upn defects, inspectin fr, and measurement f, these substances must be made by the Department f Fd and Agriculture. There is bviusly a need fr a reliable, quantitative methdlgy and set f standards fr spilage inspectin. Micrbial metablites, e.g. diacetyl and acetylmethyl carbinl, have been used as indicatrs f unsund fruit in ther fd industries (6,7). A useful metablite shuld be easy t measure and be present in sufficient quantity t be detected. Sme mlds invade injured grapes and thers break the skins f intact grapes n infectin. In either case, the grwth f yeast and bacteria will accmpany mld invasin with the resultant prductin f ethyl acetate and acetic acid. The purpse f this study was t investigate these cmpunds as pssible indicatrs f mld and rt. Fermentatin studies were set up t determine the relatinship between must levels f ethyl acetate and acetic acid and levels present after fermentatin. In rder t set suggested rejectin levels, careful threshld studies were cnducted fr these cmpunds. In additin, intentinally mlded grapes were screened fr ethyl acetate and acetic acid as well as fr ff-drs. MATERIALS AND METHODS Fermentatins: The first set f fermentatins was designed t determine whether the presence f ethyl acetate affects changes in acetic acid cncentratins during fermentatin and vice versa. Duplicate neliter lts f Perlette juice frm Kearney (23.1 B) were fermented at 21 C (7 F) with an active dry yeast, Fermvin. Additins f ethyl acetate (EtOAc) and acetic acid (HOAc) in mg/1 were: 1) N additins 2) 45 EtOAc + N HOAc 3) N EtOAc + 16 HOAc 4) 45 EtOAc + 16 HOAc A secnd set f fermentatins was set up t determine the effects f fermentatin temperature and yeast strain n the changes in ethyl acetate and acetic acid levels during fermentatin. Duplicate tw-liter lts f Oakville Chardnnay (23 B) were fermented at tw temperatures, 11.7 C (53 F) and 21 C (7 F), using tw yeast strains, 522 (Mntrachet) and Fermvin with the fllwing additins (in mg/1); 1) N additins 2) 225 EtOAc + 53 HOAc 3) 45 EtOAc + 16 HOAc 4) 9 EtOAc + 212 HOAc There were 32 lts in all. Additinally, duplicate twliter lts f Davis Grenache (23 B) were fermented at 28 C (83 F) with strain 522 and the same additins. All musts were settled vernight, received an additin f 1 ppm SO2 and had 5 mg/1 bentnite added t prmte cmplete fermentatin. The lts were sampled immediately upn cmpletin f fermentatin and analyzed fr vlatile acid by distillatin and titratin (4) and fr ttal vlatile esters spectrphtmetrically (8). 13

ACETIC ACID THRESHOLD- 131 Threshlds: Tw clean wines, ne red and ne white, were selected as base wines. The white wine was predminantly Chenin blanc and the red was a blend f Davis and Oakville Cabernet Sauvignns. Varius amunts f redistilled ethyl acetate (calculated frm specific gravity) were added t the wines. These spiked samples were pair-tested against the base wines. A panel f trained tasters smelled ten pairs in randmized rder and was asked t select the highest cncentratin f ethyl acetate in each pair. The abve prcedure was repeated fr acetic acid in the red and white wines, except that the panel was asked t smell and taste the pairs and nly five pairs were tasted daily t avid fatigue. Vlatile acid and ttal vlatile ester determinatins were made n the base wines t determine the ttal ethyl acetate and acetic acid present in the spiked wines. Cncentratin vs. % crrect abve chance was pltted n lg-prbability paper t determine the threshld value. Threshld was defined as the level at which the spiked sample culd be distinguished 5% f the time ver chance. Mldy grapes: Lts f grapes were intentinally mlded with pure cultures and made int wine t determine levels f ethyl acetate and acetic acid in the musts and wines. Twenty-pund lts f sund Thmpsn Seedless grapes were immersed fr 3 secnds in an ethanl/ NaOH slutin (7%:1 g/l) at 71 C (16 F) t injure the skins. The grapes were then thrughly washed with water and placed in plastic bags. The bags were flded with a chlrine bleach/water slutin (1:1) t sterilize the grapes and rinsed with water. Bth ends f the bag were attached t tubes fitted with sterilized air filters and flushed with air fr 15 minutes t remve excess water. The bags f grapes were then inculated with a pure culture f mld spres by syringe and incubated at rm temperature fr 12 days. By that time, at least 9% f the grapes were mldy. These mldy grapes were crushed, pressed, sulfited (75 ppm) and inculated with yeast (522). Any unusual armas were nted. They were fermented t dryness at 15 C (6 F) with yeast strain 522 (Mntrachet). Apprximately tw mnths after cmpletin f fermentatin, the wines were submitted t an 11-member panel. Panel members were asked t smell the wines and rate the intensity (n a scale f 1-1) f three cmpnents: acetic acid, ethyl acetate and "mldiness". The musts and wines were analyzed fr ethyl acetate and acetic acid in rder t crrelate sensry ratings with analytical values. RESULTS AND DISCUSSION Fermentatins: The first set f fermentatins indicated that the presence f ethyl acetate had n effect n the change in the acetic acid levels during fermentatin, and vice versa. Accrding t Nrdstrm (9), the presence f acetic acid has little effect n the frmatin f ethyl acetate. Apparently, the mechanisms invlved in the changes in the levels f the tw cmpunds are independent. In the secnd fermentatin study, there was n significant difference in the ethyl acetate and acetic acid levels in the wines due t either yeast strain r grape variety. There was, hwever, a difference due t the temperature f fermentatin. The results are given in Table 1. Because there was n significant difference between 21 C (7 F) and 28 C (83 F), the values tabulated are averages f the tw. This is cnvenient because mst red wines are fermented in this temperature range. The lines f best fit fr these data were calculated and pltted in Figs. 1 and 2. The pints shwn are the actual data. These graphs shw that greater amunts f bth ethyl acetate and acetic acid remain in wines fermented at lwer temperatures. (3 Z I.L 16 12 8 "1" E 4 / i I i / 1 O 4 8 12OO I 6 OO 2 mg/i HAc ADDED Fig. 1. Acetic acid levels in the wines vs. amunts added t the musts. 8 6 D Z ~D4 u.. UJ \ C:~ E 2 J / / U I I I I i 2 4 6 8 mg/i EtAc ADDED Fig. 2. Ethyl acetate levels in the wines vs. amunts added t the musts. / /

132- ACETIC ACID THRESHOLD It can be seen frm Table 1 that, in mst cases, the level f acetic acid in the must actually drpped during fermentatin. Accrding t Amerine et al. (2), yeasts can use acetic acid as a carbn surce during the early stages f fermentatin. Durmishidze als reprted the rapid metablizatin f acetate (2). Using C14 labeling, he nted that acetate was cnverted t acetaldehyde, ethanl, 2,3-butylene glycl and glyclic and lactic acids. Nrdstrm suggested that acetic acid culd be activated t acetyl CA and fed int lipid, citrate, prtein and nucleic acid synthesis (9). The rise in the acetic acid cntent nted fr the 53 F lt with the lwest level added can be explained by the fact that cl fermentatins prduce mre acetic acid than warm nes (3) and less disappears. In all cases there are tw ppsing prcesses perating. Small amunts f acetic acid are prduced in all fermentatins (2-4 mg/1) and acetic acid can be metablized during fermentatin. Similar trends were nted fr ethyl acetate. Table 1. Effect f fermentatin n the acetic acid and ethyl acetate cntents f the musts. Added t juice Fund in wine a EtAc HOAc EtAc HOAc 53OF 7-83OF b 53 F 7-83OF b 225 53 28 252 663 51 45 16 44 374 943 878 9 212 743 691 1886 1826 amg per liter. baverage f 7 and 83 F fermentatin. Threshlds: The threshld values presented in Table 2 were taken frm the lg prbability plts f levels in the spiked samples vs. % crrect abve chance (Figs. 3,4). These threshld values agree with the limited threshld infrmatin in the literature. Amerine et al. (2) suggest that belw 2 mg/1 ethyl acetate may be desirable, but ver that, it imparts a spiled character. Peynaud, accrding t Amerine et al., fund that the level f ethyl acetate with a characteristic dr f acescence is 18-2 mg/1. Accrding t Amerine and Ressler (5), expert wine judges usually begin t detect acetic character at 7 mg/1. Table 2. Base and threshld levels f ethyl acetate and acetic acid in white and red table wines. Wine N. f Level in mg/i cmparisns Base Threshld Ethyl acetate White 8 76 17 Red 138 78 16 Acetic acid a White 5 38 113 Red a 4 281 74 b 47 225 84 a The tw red wines were averaged t give a threshld level f 79 mg/l. These data supprt the belief that it is ethyl acetate, and nt acetic acid, that is predminantly respnsible fr acescent spilage. It is surprising t nte that E a X O2 LIJ 7 5 IE uj ~- ~2 w b- -r 8 w - 6 X ~,7mg/, WHITE 113 mg/i 2 5 i.5 5 7 9 95 98 PERCENT CORRECT ABOVE CHANCE Fig. 3. Threshld plt. Spiked levels f ethyl acetate and acetic acid in white wines vs. % crrect abve chance. a I-- LLI \ UJIE I--- w I _J >" -r K)O b- RED ~~74m g/i, ~ ~ ~ ~ f ~ mg / I I 2,3 4 5 6 7 8 % CORRECT ABOVE CHANCE 9 95 Fig. 4. Threshld plt. Spiked levels f ethyl acetate and acetic acid in red wines vs. % crrect abve chance. the threshld cncentratin determined here fr acetic acid alne in white wines is actually abve the legal limit fr vlatile acid in Califrnia (.11 g acetic/1

ACETIC ACID THRESHOLD- 133 ml) (4). Althugh few wines will have such high acetic acid levels withut accmpanying excessive ethyl acetate cncentratins, this pints ut the inadequacies f using acetic acid levels as an index f spilage. Althugh these threshld values apply strictly nly t the specific wines tested, they shuld prvide a gd estimate f the values t be expected fr mst dry table wines. Given the threshld levels and fermentatin data, it is pssible t set suggested levels (Table 3). Wine rejectin levels were assumed t be the threshld values. The level set fr acetic acid in red wines is the average f the tw values in Table 2. Must rejectin levels were determined using Figs. 1 and 2. [Assuming that white grapes wuld be fermented between 1 C (5 F) and 155 C (6. F) and red grapes between 21 C (7 F) and 29 C (85 F) t find the cncentratins that wuld result in threshld values in wine.] Table 3. Suggested rejectin levels f ethyl acetate and acetic acid in wines and musts, a Wine type Wine Rejectin levels in mg/i Must Ethyl acetate White 17 6 Red 16 115 Acetic acid White 113 119 Red 79 9 a Based n white fermented at 5-6 F and red at 7-88 F. Mldy grapes: The ethyl acetate and acetic acid arma intensity scres fr the wines made frm mldy grapes are presented in Table 4 with the crrespnding empirical values. A plt f ethyl acetate arma intensity rating vs. ttal vlatile ester cncentratin, Fig. 5, shws a gd crrelatin between amunts perceived and amunts present. These data als supprt the threshld findings, as an upward trend in intensity respnse was nt fund until levels were abve abut 175 mg/1. The acetic acid cncentratins were, in all cases, well belw the threshld, but the ethyl acetate cncentratins, with ne exceptin, were abve threshld. On this basis, seven ut f eight f the mldy wines wuld be rejected. Amerine (1) analyzed 69 Califrnia wines (presumably made frm unmldy grapes) and fund the ttal vlatile esters t range frm 65 t 145 mg/1, all belw threshld. This wuld suggest that ethyl acetate might be a gd indicatr f mld and rt because the nrmal levels f ethyl acetate in wines are well belw the threshld. Even thugh the mlds were grwn under attempted sterility, sme secndary infectin by yeast and bacteria ccurred and was respnsible fr prducing easily detectable amunts f ethyl acetate. It is questinable if, under field cnditins, ethyl acetate remains in the mlded fruit. Besides prmting acescence, mlds can prduce ther ff-drs. Table 5 lists cmments n the juices, and the wine "mldiness" scres fr varius mld species. It can be seen that the nly mld that caused ntable ~mldy" armas was Penicillium expansum. This wine, t, wuld have been rejected n the basis f ethyl acetate levels. Table 4. Acetic acid and ethyl acetate in wines made frm grapes infected with varius mlds. Organism Vlatile acidity Acetic acid Vlatile esters Ethyl acetate mg HOAc/I dr a mg EtOAc/I dr Rhizpus stlnifer 35 2.34 245 3.8 Aspergillus niger 32 1.35 2 2.57 Rhizpus arrhizus 3 1.57 98 1.55 Btrytis cinerea 43 1.91 334 4.9 Penicillium expansum 26 1.73 195 1.91 Alternaria tenuis 45 1.8 319 5.23 Cladsprium herbarium 23 1.59 255 3.23 Fusarium mnilifrme 23 1.82 251 2.77 a Average f 11 judges smelling the samples twice using a 1-1 scre card with 1 lw and 1 high. 5 W Z'4.3 z2 I-- n," I Y= -.54-1- O.OI6X r=.9.31 / 2,b 2~ sb 4b TOTAL VOLATILE ESTERS mg/i Fig. 5. Ethyl acetate intensity, sensry rating, in mldy wines with varius levels f vlatile esters. Table 5. Mldiness in juices and wines frm grapes infected with varius mlds. Organism Cmments n juice Wine mldiness scre a Rhizpus stlnifer Aspergillus niger Rhizpus arrhizus Btrytis cinerea Penicillium expansum Alternaria tenuis Cladsprium herbarium Fusarium mnilifrme Acetic 3. Slightly mldy 3.52 Grapy 3.59 Grapy 2.73 Very mldy 7.2 Grapy 2.8 3.5 3.95 a Average f 11 judges smelling the samples twice using a 1-1 scre card with 1 lw and 1 high. CONCLUSION Fermentatin studies indicated that the changes in levels f ethyl acetate and acetic acid during fermentatin were independent f each ther. The tw yeast strains and the few grape varieties studied had n ef-

134- ACETIC ACID THRESHOLD fect n changes that ccur during fermentatin. There was a significant temperature effect. Higher levels f bth ethyl acetate and acetic acid remain after fermentatin at lwer temperatures. There was a linear relatinship between the amunt f ethyl acetate and acetic acid present in a must and the amunt present after fermentatin. This allws the predictin f wine levels frm must levels. With fermentatin data and threshld values, suggested rejectin levels were set at the must cncentratins that wuld result in threshld levels in wine. The mldy grape studies indicate that ethyl acetate was a useful indicatr f spilage, because levels in clean musts and wines were belw threshld and thse in the mldy wines tested were abve threshld. Levels f acetic acid were nt as useful because the wines made frm mldy grapes had nrmal vlatile acid levels. Further studies are prceeding t see if, under vineyard and winery cnditins, ethyl acetate and acetic acid are reliable indexes fr rt and mld spilage in grapes. LITERATURE CITED 1. Amerine, M. A. Determinatin f esters in wines-liquid-liquid extractin. Fd Res. 9(5):392-5 (1944). 2. Amerine, M. A., H. W. Berg, and W. V. Cruess. The Technlgy f Wine Making. Avi Publishing C., Inc., Westprt, Ct. (1972). 3. Amerine, M. A., and M. A. Jslyn. Table Wines, the Technlgy f Their Prductin. Univ. Calif. Press, Berkeley (1973). 4. Amerine, M. A., and C. S. Ough. Wine and Must Analysis. Jhn Wiley and Sns, New Yrk (1974). 5. Amerine, M.A., and E. B. Ressler. Wines, Their Sensry Evaluatin. W. H. Freeman and C., San Francisc, Ca. (1976). 6. Fields, M. L. Acetylmethyl carbinl and diacetyl as chemical indexes f micrbial quality f apple juice. Fd Technl. 18(2): 1224-8 (1964). 7. Hill, E., and M. L. Fields. Ethyl alchl as a chemical index t the micrbial quality f apple juice. Fd Technl. 2(1 ): 77-8 (1966). 8. Libraty, V. Ester determinatins and their applicatin t wine. M. S. thesis, Univ. Calif., Davis (1961). 9. Nrdstrm, K. Studies n the frmatin f vlatile esters in fermentatin with brewer's yeast. Svensk Kemisk Tidskrift 76(9): 1-34 (1964).