Defining sulfur spray recommendations based upon the re-emergence of hydrogen sulfide off-aromas in wine post-bottling

2015 Annual Report for GNE14-082

Project Type: Graduate Student
Funds awarded in 2014: $12,249.00
Projected End Date: 12/31/2016
Grant Recipient: Cornell University
Region: Northeast
State: New York
Graduate Student:
Faculty Advisor:
Dr. Gavin Sacks
Cornell University

Defining sulfur spray recommendations based upon the re-emergence of hydrogen sulfide off-aromas in wine post-bottling


Hydrogen sulfide (H2S) is a leading cause of sulfurous off-aromas in wines.  H2S can increase during post-bottling storage under low oxygen conditions (i.e. screw cap closures), though the mechanism for this post-bottling evolution is not yet fully understood.  Our experiments show that elemental sulfur (S-0) residues on grapes not only cause increased H2S production during fermentation, but also result in continued H2S generation under reductive storage conditions.  Grape juice was spiked with various levels of S-0 (0-100 mg/L) and fermented to dryness.  Resulting wines were racked and sparged to remove H2S, followed by reductive bottling and storage.  After 3 months, free H2S was found to have increased in proportion to the original S-0 concentration, corresponding to the conversion of approximately 1% S-0 residue.  An assay for non-volatile H2S precursors based on (2-carboxyethyl) phosphine (TCEP) addition was developed.  According to this assay, H2S precursors derived from S-0 residues were determined to remain in the more polar fractions for a normal phase flash chromatography experiment. This is consistent with our hypothesis that the compound exists as one or more polysulfide derivatives. New S0 spray limitations were defined based upon the amount of latent H2S produced for wines fermented on 5 different yeast strains.

Objectives/Performance Targets

The objectives as described in the original plan of work are:



    1. Define new limits for use of S0 fungicides on the basis of H2S formed during storage rather than H2S formed during fermentation.


    1. Identify the wine soluble precursor compound responsible for H2S formed during storage.


Towards the first objective, an experiment was conducted to investigate the effect of S0 level in combination with yeast strain on “latent” H2S production in wine. Five commercial yeast strains were selected to represent a range of H2S producing capabilities. Fermentations were carried out in 100-mL volumes of grape juice, on either 0 mg/L or 100 mg/L of S0 for each yeast strain. All samples were prepared in triplicate, for a total of 30 fermentations. At the completion of fermentation, wines were racked and tested for free and “latent” H2S according to the assay developed in our lab using the reducing agent tris(2-carboxyethyl)phosphine (TCEP). In addition to the fermentation replicates, analytical replicates were performed in triplicate for a total of 90 assays. Data was analyzed by a Multiple Analysis of Variance (ANOVA) test with a Tukey HSD test. Statistical conclusions showed that a significant difference exists between “latent” H2S produced by different strains. The data, in combination with a previous experiment (detailed in the annual report for 2014) were used to define a maximum recommendation of 7 mg/L of elemental sulfur residues allowable.

Towards the second objective, a crude fraction of a polar, H2S-releasing species has been isolated. 50 mL of filtered wine was dry-loaded on acid-washed Celite, and flash chromatography was performed.on a Combiflash RF75 System using a linear gradient from 1% MeOH in DCM to 100% MeOH, on an underivatized 24-gram silica column. The TCEP assay for latent H2S was performed on reconstituted aqueous solutions from the fractions. The H2S-releasing species was found to exist only in the polar fractions, yet was spread over more than one fraction. This is consistent with the hypothesis that the unknown species could exist as more than one polysulfide derivative; for instance, a glutathione or cysteine polysulfide incorporating 3 or more sulfur atoms, as well as other symmetrical or asymmetrical polysulfide derivatives. The existence of multiple species, if that is the case, adds a challenge in that each compound must be present at lower concentration than if there existed only one compound. The ability to achieve a definitive conclusion on this matter is therefore limited by instrument and method sensitivity. An HPLC-MS/MS method was optimized for the detection of glutathione trisulfide and higher derivatives in a synthesized mixture; the stability of these compounds at wine pH was ascertained over a 6-month storage period at ambient temperature. However, these compounds were unable to be conclusively identified in wine due to the low concentration at which they are (or would be) present. Further experiments must be conducted with the goal of surmounting this obstacle.





    • American Society of Enology and Viticulture (ASEV) 2015 National Annual Conference, poster presentation


    • ASEV 2015 Eastern Section Annual Conference, poster presentation



“Precursors of Hydrogen Sulfide during Wine Storage – the Role of Elemental Sulfur Pesticide Residues”        

A poster was presented at the 2015 ASEV National Conference in Portland, Oregon entitled “Precursors of Hydrogen Sulfide during Wine Storage – the Role of Elemental Sulfur Pesticide Residues”, and again at the Eastern Section conference, accompanied by a “flash talk”. Both conferences enabled outreach to winemakers and growers, as well as discussion with other chemists regarding the mechanism under research. Work was also presented at the 2015 Cornell Global Summit. Furthermore, in the upcoming months of 2016, 15 hours per week will be devoted to extension activities in the form of webinars and outreach to industry members.

Impacts and Contributions/Outcomes

Previous recommendations for S0 spray limitations were given as 10 mg/L (Kwasniewski 2014). Our new data, based upon fermentations by 5 different yeast strains, representing a range of H2S-producing capabilities, resulted in a recommendation of 7 mg/L limit, which is very much in agreement with the previous study. These data are of critical importance to growers and winemakers wishing to maintain economically and environmentally sustainable practices while upholding the quality of their product.


Dr. Gavin Sacks

[email protected]
faculty advisor
Cornell University
152 Stocking Hall
Ithaca, NY 14850