Research Note
Determination of Sulfur Residues on Grapes Using Flame Atomic Absorption Spectroscopy
by
Barry Gump, Vickie Wahlstrom and Robert Pham
CATI Publication #950402
© Copyright October 1995, all rights reserved
INTRODUCTION
Elemental sulfur has traditionally been used for powdery mildew control in California vineyards. It is well recognized as an effective and economical treatment. However, sulfur residues remaining on grapes at harvest have the potential for forming hydrogen sulfide and other sulfur compounds during fermentation.1,2,7,10 Park, et al, have identified a number of volatile sulfur compounds in California wines with "sulfur-like" problems.6 Among these compounds dimethyl sulfide and ethanethiol were the most frequently found, with significant occurrence of diethyl disulfide and some methanethiol.
There is some question, as to whether "usual" cultural applications of elemental sulfur in the vineyard leave enough sulfur residue on the berries to potentially cause significant amounts of sulfur compounds in a wine. In companion vineyard11 and laboratory12 studies it was found that while sulfur residues on harvested wine grapes were higher on dusted grapes than on sterol demethylation inhibitor (DMI) fungicide-treated grapes, the typical sulfur residue levels found did not correlate with hydrogen sulfide production during fermentation.
The laboratory study with sulfur added to the fermentation medium did demonstrate some significant yeast strain-sulfur interactions. It has also been shown that sulfate3 can be the major source of H6S in fermentations with certain yeast strains in musts with nitrogen deficiencies. In the studies reported in this paper there was some evidence that delicate white wines produced from grapes with sulfur residues greater than 2 mg/kg berry weight could be detected relative to a control in a sensory test.9
In spite of the uncertainty surrounding any correlation between grape berry sulfur residue level and resultant sulfur problems with a wine, it would be prudent for a producer to be able to determine the level of sulfur residues on grapes being brought into the winery.
Typical analytical methods used for determining elemental sulfur residues remaining on grapes at harvest utilize (1) a vacuum monochrometer inductively coupled plasma instrument (ICP),11, or (2) the AOAC gravimetric procedure4. There are some difficulties with both of these methods. The ICP method requires an expensive instrument that is not commonly found in wine industry laboratories. The gravimetric procedure requires that one precipitate the sulfur as barium sulfate and then wash, filter, dry, and weigh this precipitate. This procedure would require either a very large sample, or the ability to accurately weigh minute amounts of the precipitate (less than 0.4 mg of barium sulfate precipitate for each milligram of sulfur residue per kilogram of berry weight).
We have developed a modification to the precipitation method in which we measure barium solution concentrations (by atomic absorption spectroscopic analysis, AAS) before and after some of the barium is precipitated with any elemental sulfur present. As a result we can quantitate sulfur residues on grapes by difference without having to filter and weigh small amounts of precipitate.
PROCEDURE
Random clusters of wine grapes, weighing approximately 100 g, are selected from various vineyard treatments, and placed in 100 mL of a 0.2 % surfactant solution. [Both Triton and Tween 80 surfactants have been used for these samples.] The samples are agitated on a shaker table for 30 minutes to wash elemental sulfur residues off the surface of the berries and into the surfactant solution. The clusters are removed, and the solutions analyzed for elemental sulfur.
Sample preparation prior to Atomic Absorption analysis is accomplished using a variation of the AOAC digestion procedure, 22.050, for Total Sulfur4. From each solution, 25 mL aliquots are pipeted into porcelain casserole dishes, followed by additions of 1 gram of magnesium oxide (MgO), 1 gram of Sucrose and 10 mL concentrated nitric acid (HNO3). A blank is prepared using 25 mL deionized water and the same amounts of reagents.
The casseroles are placed in a heated sand bath, and the solution allowed to evaporate to paste. Following digestion, the samples are placed in a cold muffle furnace, and heated (≤525°C) until all nitrogen dioxide (NO2) fumes are driven off and the organic matter is completely decomposed (2-3 hours).
After cooling, the ash is dissolved with a few drops of 6 M HCl. It is then neutralized with potassium hydroxide (potassium added to act as an ionization suppressor in the AAS technique), and neutralization checked with pH Litmus paper. The solution is transferred to a 10 mL volumetric flask with deionized water, 2 mL of 150 ppm Barium standard is added (see below), and the sample brought to volume with deionized water. In order to facilitate barium sulfate (BaSO4) precipitation, samples are capped and allowed to stand overnight.
The solution is filtered into storage vials through a 0.45 micrometer syringe filter and analyzed for the remaining Ba2+ concentration by Atomic Absorption Spectroscopy.
A 1000 mg/L (ppm, as barium) stock solution can be prepared from 1.7785 grams of Barium chloride dihydrate (BaCl2H2O) dissolved in one liter of deionized water. From this, 10.00, 20.00, 30.00, and 40.00 mg/L standards are prepared for a calibration curve (add 1.00, 2.00, 3.00, and 4.00 mL of the stock solution to a 100-mL volumetric flask and dilute to volume), and a 150 mg/L Ba solution is prepared for precipitating the sulfur residues in the samples (add 15.00 mL of the stock solution to a 100-mL volumetric flask and dilute to volume). The parameters for the AAS are: lamp current 20 mA, slit width 0.5 nm, wavelength 553.6 nm, flame nitrous oxide/acetylene.
To calculate an answer from the atomic absorption spectroscopic data:
1. Prepare calibration curve from known barium standards of 10, 20, 30, and 40 mg Ba2+ /Liter.
2. Read concentration of barium remaining in treated sample solution after precipitation of barium sulfate and filtering.
3. Subtract final barium concentration from 30 mg Ba / Liter to obtain the mg/L of barium that was precipitated by the sulfur in the sample. [Note: 2 mL of a 150 mg Ba/L standard was added to the sample in a 10 mL volumetric flask. Due to the 2/10 dilution factor the barium concentration is now 30 ppm.]
4. Calculate original sulfur residue level in "mg sulfur residue per kilogram berry weight" as follows: Divide the mg Ba / L that was precipitated by the berry cluster weight in grams. Multiply this by a factor of 9.34 (to convert mg/L to mg/kg and to account for sample size and dilution factors).
example:
RESULTS
In a series of vineyard trials using elemental sulfur (dust, wettable, and micronized) and DMI fungicides elemental sulfur concentrations were measured on French colombard and Zinfandel wine grapes (see Table 1). Since the prime purpose of these analyses was to develop the atomic absorption spectroscopic method, some of the solutions from the wine grapes were analyzed by both ICP and AAS (see Tables 2 and 3). Also presented in these tables are the gross amounts of elemental sulfur applied over the grwoing season. The different amounts of sulfur are due to varying application rates and schedules.
There is generally a good correlation between the analytical results by AAS and ICP for the wine grape washings. A paired t-test on the ICP and AAS data indicated no significant difference between the two methods at the 95 % confidence limit. To better present a comparison of the ICP and AAS analytical values the data in Tables 2 and 3 were plotted as bar graphs (Figures 1 and 2). From Figure 1 it is readily apparent that both analytical procedures provided essentially the same answer; with one exception. Only with the "sulfur dust only" treatment do the answers differ (by a little more than one mg/kg). In Figure 2 one can see that several answers by the two methods differ, the largest difference being for the control. It is presumed that the ICP answer is an aberration; one would not expect 6 mg/kg sulfur residue on a control sample. Most of the differences are less than 1 mg/kg in magnitude. Recall that 1 mg/kg of residue on a sample only gives the analyst 0.1 mg of elemental sulfur to recover from 100 grams of grapes.
The atomic absorption barium line at 553.6 nm is used for solutions containing from 10-40 mg/L of barium. With a typical calibration curve 30 mg/L (ppm) barium had an absorbance signal of approximately 0.180. Assuming a 100 gram berry sample taken for analysis in 100 mL of surfactant solution, each 1mg sulfur residue per kilogram (1000 grams) of berries (1 ppm) would reduce the barium concentration ~ 18 mg/L resulting in a change in absorbance of ~ 0.1 absorbance units. Under typical instrumental operation conditions residue levels of ~ 0.5 mg of sulfur per kilogram of berry weight should be readily discernable.
It was found that the choice of surfactant for the grape washings could be important. Initially, the washings done with Tween 80 caused the burner on the AAS to clog after 4-5 samples had been aspirated. This required shutting the instrument down to clean the burner since absorbance readings would begin to decay once the burner was clogged. It was found that samples subjected to a longer ashing period (2-3 hours) did not tend to clog the burner. Those samples where Triton was the surfactant did not clog the burner even when the minimum ashing time was used.
An important consideration in sample preparation is the ratio of surfactant volume to berry weight. Since the amount of sulfur on grapes is relatively low (from < 1mg/kg to 4-5 mg/kg), using a 1:1 ratio of surfactant to fruit (i.e. 100 g of grapes to 100 mL of solution) produces more barium sulfate precipitate. This makes filtering the samples prior to AAS analysis easier. Samples where smaller ratios of berry weight to solution volume are used often lead to no visible barium sulfate precipitate.
Barium is a more difficult element to analyze with atomic absorption than other alkaline earth metals such as calcium or magnesium. Due to low sensitivity for the element, the nitrous oxide-acetylene flame is required. In this hotter flame the population of ground-state atoms, measured at a wavelength of 553.6 nm, can be depleted by ionization of the barium atoms (Ba --> Ba2+ + 2e-). A large excess of easily-ionizable potassium is added to suppress this ionization. Another factor affecting sensitivity for barium is the flame conditions. The flame that is ideally used for barium is a reducing flame with a red cone about 2 to 3 cm high.
As a follow up to our original work, the atomic absorption analysis was used as a supplemental detection system for sulfur dioxide fumigation residues on table grapes. Typically these analyses are conducted using the Optimized Monier-Williams method.5 In the Monier-Williams procedure 100 grams of table grapes are boiled with hydrochloric acid and the sulfur dioxide residues are distilled off into a peroxide solution. Final analysis is made by titrating the sulfuric acid (produced through the reaction between sulfur dioxide and peroxide) with sodium hydroxide. Typical titration volumes for early-season sulfur dioxide residues levels are 0.25 to 0.70 mL of the sodium hydroxide (corresponding to sulfur dioxide residue levels of 0.4 mg SO2/kg fruit to 1.1 mg S2/kg fruit [ppm]). When 10 mL samples of these same peroxide solutions were analyzed by AAS, changes in the barium concentration of 1 to 3 mg/L (ppm) were obtained, allowing one to calculate the original sulfur dioxide residue levels (see Table 4). As before, statistical treatment of the data indicate no significant difference in the values from the titration and AAS procedures obtained for these low-residue samples.
An interesting feature of this work was the use of an alternative spectroscopic analytical line of 455.4 nm for measuring the barium concentration.8 Barium possesses two useful absorbing lines, the resonance line at 553.6 nm used in the sulfur residue work described above, and an ionic line at 455.4 nm. When samples do not contain any easily ionizable element, such as the alkali metals, sodium and potassium, then there is a very stable and slightly more sensitive response for barium obtained at the shorter wavelength. Since the sulfur dioxide is distilled into hydrogen peroxide in the Monier-Williams procedure, no significant concentrations of ionizable metals would be present.
CONCLUSION
The results of this study indicate that the AAS is a viable method for the analysis of elemental sulfur on grapes. By measuring excess barium with AAS one is able to detect sulfur residue concentrations of less than 1 mg/kg berry weight. As mentioned, choices of surfactant and/or ashing times are important variables, and we would recommend a 2-3 hour ashing time. The berry weight to surfactant volume ratio is also an important consideration, and minimum 100 gram samples of fruit are recommended. The atomic absorption analysis procedure also can be applied as a measuring tool for sulfur dioxide residues on fumigated table grapes.
ACKNOWLEDGEMENTS
Our thanks to the California Agricultural Technology Institute (CATI) and the American Vineyard Foundation (AVF) for funding this study
REFERENCES
1. Acree, T.E., E.P. Sonoff, and D.F. Splitstoesser (1972). Effect of yeast strain and type of sulfur compound on hydrogen sulfide production. Am. J. Enol. Vitic. 23, 6-9.
2. Eschenbruch, R. (1978). Sulfide and sulfite formation by wine yeasts. n: Proc. Fifth Int. Oenol. Symp., Auckland, NZ, 267-274.
3. Giudici, P., and R. F. Kunkee (1994). The effect of nitrogen deficiency and sulfur containing amino acids on the reduction of sulfate to hydrogen sulfide by wine yeasts. Am. J. Enol. Vitic. 45 (1), 107-112.
4. Official Methods of Analysis, 12th Edition (1975). Method 22.050. Association of Official Analytical Chemists, Washington, D.C.
5. Official Methods of Analysis, 15th Edition (1990). Association of Official Analytical Chemists, Washington, D.C.
6. Park, S.K., R. B. Boulton, E. Barta, and A. C. Noble (1994). Incidence of volatile sulfur compounds in California wines. A preliminary survey. Am. J. Enol. Vitic. 45 (3), 341-344.
7. Rankin, B.C. (1963). Nature, origin and prevention of hydrogen sulfide aroma in wines. J. Sci. Food Agric. 14 , 79-91.
8. Reynolds, R.J., and K. Aldous (1970). Atomic Absorption spectroscopy; A Practical Guide, Barnes & Noble, Inc., New York, NY.
9. Sawyer Ostrom, G. and V.E. Petrucci (1995). Sulfur Residues on Wine Grapes. In Proceedings of the Int. Soc. Hort. Sci. Int Wkshp, Congeliano, Italy, July 9-12.
10. Schutz, M., and R. E. Kunkee (1977). Formation of hydrogen sulfide from elemental sulfur by wine yeast. Am. J. Enol. Vitic., 28, 137-144.
11. Thomas, C.S., W. D. Gubler, M. W. Silacci, and R. Miller (1993). Changes in elemental sulfur residues on Pinot noir and Cabernet Sauvignon grape berries during the growing season. Am. J. Enol. Vitic. 44 (2), 205-210.
12. Thomas, C.S., R. B. Boulton, M. W. Silacci, and W. D. Gubler (1993). The effect of elemental sulfur, yeast strain, and fermentation medium on hydrogen sulfide production during fermentation. Am. J. Enol. Vitic. 44 (2), 211-216.
{ page top }
{ CATI , VERC , VERC - Research Publication }
Copyright © 1995. All rights reserved.
CALIFORNIA AGRICULTURAL TECHNOLOGY INSTITUTE - CATI
College of Agricultural Sciences and Technology
California State University, Fresno
