Skip to main content
NCBI home page
Current Research in Food Science logoLink to Current Research in Food Science
. 2025 May 21;10:101089. doi: 10.1016/j.crfs.2025.101089

Dimethyl sulfide transfers through closure during accelerated model wine ageing: proof-of-concept & prospects

Rémi De La Burgade a, Lucas Suc a, Somaya Sachot a, Christine Le Guerneve a, Nicolas Galy b, Dimitri Tixador b, Christophe Loisel b, Nicolas Sommerer a, Aurélie Roland a,⁎
PMCID: PMC12152913  PMID: 40503522

Abstract

Dimethyl sulfide (DMS) is a volatile sulfur compound that plays a complex role in wine aroma, contributing both positive and negative sensory attributes depending on its concentration. While DMS is known to accumulate during bottle aging through the degradation of precursors such as S-methylmethionine (SMM), this study presents the first evidence that DMS can also be lost through wine closures via a permeation mechanism.

In practice, the permeation of DMS through closures was demonstrated using model wines spiked with DMS and aged under accelerated conditions at 35 °C. As DMS was detected only in the ®Tenax tubes placed above bottles containing the spiked model wines, we formally proved that DMS can permeate closures under these conditions and may account for 12 % of initial DMS. In Syrah wines, DMS concentrations increased during bottle aging due to the breakdown of SMM. However, wines sealed with more permeable closures exhibited lower DMS levels compared to those sealed with low-permeability closures, supporting findings from the model wine studies. This previously unreported phenomenon of permeation underscores the significant influence of closure permeability on the aromatic evolution of aged wines.

Keywords: Dimethyl sulfide, S-methylmethionine, Wine closure, Accelerated ageing, Permeability

Graphical abstract

Image 1

Open in a new tab

Highlights

  • •

    First evidence that dimethyl sulfide (DMS) can permeate through closure under accelerated ageing of a model wine.

  • •

    Closure permeability directly influences DMS permeation.

  • •

    Breakthrough in understanding the evolution of wine aroma profiles: migration vs. oxidation and/or acid catalyzed mechanisms.

1. Introduction

During bottle storage, the oxygen level plays an important role on the evolution of wine aroma and can significantly modify its quality. While wine evolution was established to depend on multiple factors such as temperature, bottle position, light or humidity (Furtado et al., 2021), oxygen appears to be the determinant factor during bottle ageing, with more than half of the flavour compounds involved in aroma evolution being affected by oxygen (Ugliano, 2013).

The transfer of small molecules such as oxygen through closures can be due to different mechanisms such as: diffusion through closure pores, diffusion in the interval between closure and bottle, and oxygen expulsion from the closure during compression (Crouvisier-Urion et al., 2018). It was established that wine closures had different behaviours towards oxygen content, with screwcaps and technical cork stoppers being the tightest closures, synthetic closures being the most permeable ones and natural cork closures having an intermediate behaviour (Silva et al., 2011), even though the latters presented very heterogenous permeabilities (Karbowiak et al., 2009). Technical closures, also known as micro-agglomerated closures (MAC), are produced with granulated cork and a food-grade binder moulded together (Azevedo et al., 2022). Like natural cork closures, micro-agglomerated closures are macroporous alveolar material composed of empty cells, where gas transfer kinetics is controlled by cork cell walls (Crouvisier-Urion et al., 2018). The limiting step for gas transfer through micro-agglomerated closures was determined to be the surface diffusion mechanism, due to polyurethane being used as a binder. Generally, to characterise the oxygen transfer through the closure, the oxygen transfer rate parameter (OTR; mg O2/year), defined as the mass flow between the 2 extremities of the closure, is often used. The higher the OTR value, the more permeable the closure is.

Cork closures can also modulate wine aroma profile by releasing volatile compounds, which could be either pleasant such as terpenes, or responsible for off-flavours, such as 2,4,6-trichloroanisole, involved in moldy and musty notes (Buser et al., 1982). In addition, recent investigations showed that closures could sorb flavour compounds at their surface (Silva et al., 2011). This phenomenon, known as flavour scalping, has been identified for a wide range of aroma compounds (Capone et al., 2003; Karbowiak et al., 2010). In general, synthetic closures can scalp more flavour compounds than cork closures (whether natural or micro-agglomerated) whereas screwcaps were shown to have no scalping effect (Capone et al., 2003). Interestingly, it has also been documented that some compounds, such as d5-2,4,6-trichloroanisole present in the atmosphere, could contaminate wine inside the bottle, meaning that closures were not completely hermetic to atmosphere conditions during ageing (Lopes et al., 2011). Nevertheless, to the best of our knowledge, no flavour losses through wine closure were ever identified during wine bottle ageing.

Dimethyl sulfide (DMS) is a well-known flavour compound found in fermented beverages such beer and wine (Landaud et al., 2008). DMS was described as black olive, truffle and undergrowth in Shiraz and Grenache wines originating from the Rhone Valley (Segurel et al., 2004). DMS could also be perceived as green olive in Shiraz and Xinomavro wines when its concentration was above 100 μg/L (Anocibar, 1998). In dearomatized Spanish Grenache wine, DMS was also considered as a faulty sulfurous flavour when its concentration reached up to 50 μg/L and above (Escudero et al., 2007). Other studies highlighted the role of DMS in enhancing wine fruity notes especially by exhausting ester families (Segurel et al., 2004; Escudero et al., 2007; Demora et al., 1987; Lytra et al., 2014). DMS was found in various types of wines, with concentrations varying from 35 to 84 μg/L in Xynomavro wines (Beloqui et al., 1996); 42–910 μg/L in Australian Cabernet Sauvignon wines (Demora et al., 1987); and from 3.2 to 46 μg/L and 3.4–15.6 μg/L in Rhone Valley Shiraz and Grenache noir wines, respectively (Segurel et al., 2004). DMS levels are usually above their perception threshold, which can vary from 27 μg/L in Bordeaux red wine (Beloqui et al., 1996) to 60 μg/L in a South Australian Cabernet Sauvignon (Demora et al., 1987).

Up to now, the origin of DMS in wine is only partially known. During fermentation, yeast could release DMS from various amino acids such as cysteine (De et al., 1986), but also from the reduction of dimethyl sulfoxide (DMSO) (Anocibar, 1998), although this precursor was not detected or at very low levels in wine (Ségurel et al., 2005), meaning DMSO could not completely explain the formation of DMS. Besides, DMS was found to accumulate during bottle ageing (Marais, 1979), from no DMS found after bottling in a Riesling wine up to 29 μg/L after 16 weeks of storage at 20 °C. It was thus prouved that DMS formation could originate from a pathway other than fermentation (Bekker et al., 2016; Loscos et al., 2008).

A methionine derivative present in grapes, S-methylmethionine (SMM), was confirmed to be a DMS precursor during wine ageing (Loscos et al., 2008). To this day, SMM is considered to be the main precursor of DMS during storage, which could account for 21 %–74 % of DMS formed, in a Sauvignon Blanc wine and in a Petit Manseng must, respectively (Bekker et al., 2018). Because all the DMS precursors in wine have not been fully identified up to now, it was important to use a global indicator to monitor DMS release in wine: the DMS potential (DMSP). Suggested a few years ago, DMSP corresponds to an analytical method based upon alkaline degradation of wine by NaOH and further SPME-GC-MS/MS analysis of released DMS (Segurel et al., 2004).

The aim of this study was to assess possible losses of DMS through wine closure, and then take a well known chemical mechanism occurring during storage, i.e. the degradation of SMM to DMS, to quantify the fate of DMS, whether through chemical mechanisms, flavour scalping or possible losses through wine closure during accelerated bottle ageing, in model wine and finally to apply the results in Syrah wine under regular bottle ageing.

2. Material and methods

2.1. Chemicals

L-methionine (≥98 %), iodomethane (≥98 %), iodomethane-d3 (≥98 %), thiophene (≥98 %), tartaric acid (≥99.5 %), sodium hydroxide (≥98 %), and magnesium sulfate heptahydrate (≥99 %) were purchased from Sigma-Aldrich (Saint-Quentin-Fallavier, France). Dimethyl sulfide (≥98 %) was supplied by Fluka (Charlotte, USA). Absolute ethanol (≥99.8 %) was purchased from VWR (Rosny-sous-Bois, France). Water LC-MS grade was provided by Biosolve (Dieuze, France).

S-methylmethionine (purity 99 %) and S-methylmethionine-d3 (purity 100 %) were synthesized according to a published method (Deed et al., 2019) and all the protocols are availble in supplementary information section.

2.2. Analytical methods

2.2.1. Quantification of S-methymethionine by LC-MS/MS

S-methylmethione was quantified by Stable Isotope Dilution Assay (SIDA) and LC-MS/MS as published by Deed and co-workers (Deed et al., 2019). The protocols are available in supplementary information section.

2.2.2. Quantification of DMSP by GC-MS/MS

Potential in DMS was quantified according to Segurel and co-workers method (Segurel et al., 2004).

2.2.3. Quantification of DMS by GC-MS/MS

DMS was analysed by SPME-GC-MS/MS according to our previous published method (De La et al., 2023a).

2.2.4. Untargeted analysis of aroma compounds by GC-MS

For each wine sample, MgSO4, 7H2O (2.5 g) was added to an SPME vial (20 mL) as described by Fedrizzi et al. (2007a). 10 mL of the model wine contained in the bottle sample were spiked with an internal standard solution (thiophene; 75 μg/L in absolute ethanol; 80 μL).

Analysis was performed by SPME-GC-MS/MS using extraction and GC-MS method developed by De La Burgade and co-workers (De La Burgade et al., 2023a) except for detection conditions, where acquisition was performed in full scan mode (27–350 uma).

2.3. Experimental design

2.3.1. Material

For the accelerated ageing experiment, 21 transparent Burgundy wine bottles (75 cL) were purchased from La Cavine Lattiere (Marquay, France), and 6 transparent cider bottles (75 cL) were supplied by Boboco (Gensac la Pallue, France).

Crown caps (CRW; 29 mm) were provided by Rolling Beers (Baillargues, France), and micro-agglomerated closures (MAC) by Diam Bouchage (Céret, France). Five types of closures were studied, with different OTR and so, different permeabilities, and were coded C1, C2, C3, C4 and C5. The dimensions for all MAC were 44,0 ± 0.2 x 24.2 ± 0.1 mm. OTR and OIR (Oxygen Initial Rate) were measured by DIAM Bouchage company: CRW had an OTR of 1.17 mg O2/year and no OIR, while C1 to C5 displayed both OIR and OTR. C1 had a very low OTR (0.19 ± 0.02 mg O2/year) and OIR (0.91 ± 0.04 mg O2). C2 and C3 had intermediate OTR (1.15 ± 0.40 mg O2/year and 1.07 ± 0.29 mg O2/year respectively) and OIR (1.98 ± 0.32 mg O2 and 1.92 ± 0.21 mg O2 respectively). C4 had a high permeability (OTR: 1.79 ± 0.36 mg O2/year; OIR: 2.31 ± 0.20 mg O2). C5 was the most permeable closure, with high OTR (3.56 mg O2/year) and OIR (3.18 mg O2), but, as C5 was a prototype, no standard deviation was measured on this closure.

To measure dissolved oxygen in model wine, SP-PSt3-NAU-D5-YOP planar oxygen-sensitive spots were purchased from PreSens (Regensburg, Germany).

2.3.2. Highlighting DMS permeation through closure during accelerated wine ageing

For this experiment, a model wine solution was prepared with 13 % ethanol in milliQ water and 5 g/L tartaric acid, and then pH was adjusted to 3.8 with a NaOH solution (32 %). 6 bottles were then filled as followed: 3 of them were filled with only model wine (V = 75 cL) and 3 of them with model wine (V = 75 cL) and spiked with DMS solution at 10 mg/L.

Each bottle headspace was inerted with argon and closed with C4 closures placed about 2.6 cm from the neck of the bottle and sealed with glue on the glass to avoid any movement from the closure during ageing. After conditioning, a tenax tube was inserted at the top of the bottle using a perforated rubber plug with Parafilm. Each bottle was then placed vertically at 35 °C for 1 month (Fig. 1).

Fig. 1.

Fig. 1

Open in a new tab

Highlighting DMS permeation through closure (A: picture of the setup; B: tenax tube chromatographs after 1 month of ageing for bottle with DMS (4, 5, 6) and control bottles without DMS (1, 2, 3).

After 1 month of ageing, the tenax tubes were analysed using GC-MS following a previous developed method (Jimenez-Lorenzo et al., 2021). All data were summarized in Fig. 1.

2.3.3. Modelling the fate of SMM and DMS under accelerated model ageing conditions

The aim of this experiment was to carry out a material assessment of the degradation of SMM in DMS during the accelerated ageing of a model wine and to quantify the permeation of DMS through the stopper.

In practice, 2 model wines with 2 different pH (MW1 and MW2) were prepared as a solution with 13 % ethanol in milliQ water and 5 g/L tartaric acid (final volumes: 15 L for MW1 and 8 L for MW2). The pH was then adjusted at 3.8 and 2.8 for MW1 and MW2 using either NaOH (32 %) of HCl solution (1M), respectively.

The bottles were then filled as followed: 3 cider bottles and 15 wine bottles were used for MW1, while 3 cider bottles and 6 wine bottles were used for MW2. For each bottle filled with 75 cL MW, the medium was spiked with SMM iodide at 10 mg/L (61 μmol/L) from a pure stock solution (7.5 g/L in MW). All bottles were inerted with argon before being closed with crown caps or wine closures. The general workflow of this experiment is displayed in Fig. S1. Each bottle was then placed vertically at 35 °C for 3 months to perform an accelerated ageing. After 3 months, all the wines were chemically characterized in term of SMM, DMS.

2.4. Hypothesis of the fate of DMS in Syrah wines during standard ageing

In order to validate the experimental results determined in model wine, DMS concentrations, but also SMM and DMSP levels were analysed in 6 Syrah wines, bottled and put for ageing in conventional conditions.

Six Syrah wines were collected from different wineries, four coming from Languedoc-Roussillon region (LR1, LR2, LR3 and LR4), and two from Côtes-du-Rhône (CR1 and CR2). LR1, LR3, LR4 and CR2 were aged into a tank, whereas LR2 and CR1 were aged in barrels. Each wine was bottled in March 2021 at a local winery (Vivelys, France) using conventional oenological practices. Bordelaise-type bottles (0.75 cL) were manually placed on a bottling machine. The empty bottles were first inerted with nitrogen and, to homogenise the wine bottles, the first and last bottles from each estate were not used. Once the bottles had been filled with 0.75 cL of wine, they were corked using a manual compressed-air corker, with closures C1 to C4. For each estate, the closures used for corking were used in random order. In the end, for each wine, 4 types of closures were used for a total of 60 bottles. Each modality (i.e. a specific wine bottled with a specific closure) was done in triplicate. A total of 366 bottles were filled and corked for this study. After bottling, 72 bottles were analysed to characterise initial DMS, SMM and DMSP concentrations. Out of the 366 bottles, 294 were stored in a cellar (Pech Rouge, France) at 18 °C for ageing. The bottles were stored horizontally. On all these bottles, half were aged for one year and the other half was aged for 2 years. DMS, SMM and DMSP were quantified at t0, t12 months and t24months on all the wines.

3. Statistical analyses

All the statistical analyses (Tukey test and Spearman test) were performed using XLStat software (Microsoft).

4. Results and discussions

4.1. Validation of DMS permeation through closure by sampling outside from the bottle

After accelerated ageing, the Tenax tubes placed at the top of the bottles were analysed by GC-MS. While Tenax tubes 1, 2 and 3, corresponding to control bottles with no DMS in the model wine showed no DMS peak on the chromatograms, Tenax tubes 4, 5 and 6 (bottles with DMS at 10 mg/L in the model wine) exhibited presence of DMS on the chromatograms.

These results showed that flavours such as DMS, like O2, could be submitted to permeation phenomenon through micro-agglomerated closures from the inside to the outside of the bottle, and could lead to a new way of flavour loss during bottle ageing.

To quantify these potential losses during wine ageing, another experiment was performed under model conditions.

4.2. Modelling the fate of SMM and DMS under accelerated model ageing conditions

The aim of this experiment was to quantify the losses due to DMS permeation through closure during accelerated mode wine ageing. In practice, model wines adjusted to 2 different pH (MW1: 2.8 and MW2: 3.8) were spiked with S-methylmethionine (53.6 μmol/L). Model wines were then bottled using different kinds of closures (crown cap, micro-agglomerated cork closures with different OTR) and placed in an oven (35 °C) for 4 months of accelerated ageing. At the end of ageing, both degraded SMM and DMS produced were quantified (Table 1), as well as other potential flavours that were investigated by untargeted GC-MS analysis to have a complete picture of DMS fate.

Table 1.

Concentration levels of S-methylmethionine (SMM) and dimethyl sulfide (DMS) after 4 months of accelerated ageing at 35 °C, depending on the type of closure and model wine (MW) pH.

SMM
Degraded SMM
Released DMSa
Scalped DMSb
Total DMSc
μmol/L μmol/L % μmol/L μmol/L μmol/L
pH 2.8
MW spikedd 53.57 ± 1.27 ae
CRW 29.54 ± 0.66 b 24.03 ± 0.66 a 45 % 25.43 ± 1.47 a 0.44 ± 0.08 b 25.87 ± 1.43 –
C1 29.55 ± 0.26 b 24.02 ± 0.26 a 45 % 19.94 ± 1.32 b 0.37 ± 0.01 a 20.31 ± 1.32 a
C5 30.07 ± 0.36 b 23.50 ± 0.36 a 44 % 18.24 ± 1.05 b 0.68 ± 0.09 c 18.92 ± 1.08 a
pH 3.8
MW spiked 53.58 ± 1.18 a
CRW 27.07 ± 0.17 b 26.51 ± 0.17 a 49 % 27.36 ± 1.41 a 0.47 ± 0.02 a 27.83 ± 1.39 –
C1 29.99 ± 0.77 b 23.58 ± 0.77 a 44 % 21.05 ± 0.87 a 0.47 ± 0.01 a 21.52 ± 0.88 a
C2 29.33 ± 1.07 b 24.25 ± 1.07 a 45 % 23.69 ± 3.00 a 0.38 ± 0.07 a 24.07 ± 2.98 a
C3 30.81 ± 2.90 b 22.77 ± 2.90 a 42 % 21.13 ± 1.97 a 0.54 ± 0.08 a 21.67 ± 2.05 a
C4 26.10 ± 3.46 b 27.48 ± 3.46 a 51 % 21.91 ± 0.43 a 0.61 ± 0.02 a 22.51 ± 0.45 a
C5 29.49 ± 0.81 b 24.08 ± 0.81 a 45 % 19.64 ± 0.57 a 0.88 ± 0.49 a 20.52 ± 0.63 a
Open in a new tab
a

Levels of DMS found in model wine.

b

Levels of DMS scalped at the surface of wine closure.

c

Sum of released and scalped DMS.

d

Levels of SMM that were spiked in model wine before bottling.

e

Significance levels at 5 % obtained via a Tukey test. For the “Total DMS” column the MAC closures were only compared between them.

4.2.1. SMM degradation

After accelerated ageing, SMM evolution was measured using the following equation: Cd=CSMM,t0−CSMM,t4, with CSMM,t0 the SMM concentration at bottling and CSMM,t4 the SMM concentration after accelerated ageing for each bottle. The percentage of SMM degradation ranged from 42 % to 52 % of the initial concentration (Table 1), and no significant differences were observed between bottles at either pH. These results are in agreement with previous studies, because while previous research showed that SMM degradation increased at high pH (Anness and Bamforth, 1982), wine pH difference was considered too low to influence the chemical reaction under our conditions. Indeed, since the degradation reaction is a nucleophilic substitution, a pH < 5 has no influence on either the reaction kinetics or yield notably in the case of paprika spice (Cremer and Eichner, 2000).

In addition, there was no significant difference between degraded SMM in MAC samples (Table 1), which is coherent because SMM degradation during wine ageing is only an acido-catalyzed reaction and not an oxidative process (Loscos et al., 2008).

4.2.2. DMS formation

In Table 1, DMS concentrations ranged from 18.24 μmol/L to 25.43 μmol/L at pH 2.8, and from 19.64 μmol/L to 27.36 μmol/L at pH 3.8 in our aged model wines. All samples produced DMS during ageing, which is consistent with previous studies (Marais, 1979; Bekker et al., 2018).

Flavour scalping of DMS was also analysed to have a complete picture of DMS in bottles according to our previously developped procedure (De La Burgade et al., 2023a). Scalped concentrations varied from 0.37 μmol/L to 0.88 μmol/L for all samples, confirming that DMS could indeed be subject to flavour scalping as already reported (Silva et al., 2012). Most closures showed no significant differences among them, except for sample C5 at pH 2.8, which exhibited a higher level of scalped DMS compared to the other two samples at the same pH (CRW and C1). In this study, since flavour scalping accounted for only 1 %–4 % of the fate of DMS in the bottle, the phenomenon was considered negligible compared to other mechanisms, which was consistent with our previous work (De La Burgade et al., 2023a).

The total DMS content in the bottle was calculated by summing for each sample the DMS content in model wine and scalped DMS for its corresponding closure (Table 1). At each pH, CRW model wines had a significant higher DMS concentration compared to Micro-Agglomerated Closure (MAC) samples (25.9 ± 1.4 μmol/L for CRW and 19.6 ± 1.3 μmol/L for MAC at pH 2.8; 27.8 ± 1.4 μmol/L for CRW and 21.9 ± 2.0 μmol/L for MAC at pH 3.8) suggesting some losses of DMS in MAC capped bottles by permeation mechanism during accelerated ageing.

Likewise, no significant differences in DMS levels were observed between bottles closed with the same type of closure at different pH (Table 1), validating the hypothesis that wine pH value had too low a variation to influence DMS formation by SMM under our conditions (Cremer and Eichner, 2000).

4.2.3. Chemical mass balance

To evaluate the amount of DMS that can escape through closure during ageing, we performed a chemical mass balance based upon Lavoisier's law. For that purpose, we used SMM degradation amounts (nSMM degraded with n: moles) into theoretical DMS formed by assuming that 1 mol of degraded SMM led to 1 mol of produced DMS (data not shown). In a second step, the value of theoretical DMS produced was compared to the actual DMS amounts in bottles (CDMS in wine + CScalped DMS+C by-products in wine) with CDMS in wine corresponding to the DMS in wine, CScalped DMS corresponding to the amount of DMS scalped on the closures, and C by-products in wine corresponding to the putative by-products of DMS in wines.

For the putative by-products formed ageing, we decided to perform a flavour profile by SPME-GC-MS on all aged model wines. The study of each chromatogram did not evidence any flavour compounds other than DMS (data not shown) in CRW and in C1 to C5 samples. Because oxygen was present in all the bottles, it was hypothetised that DMS could be oxidized into DMSO. None of the flavour profiles displayed any peak corresponding to potential DMSO in model wine, and no DMS oxidation into DMSO in wine was ever observed in the litterature to our knowledge. We can thus assume that DMS oxidation into DMSO under oenological conditions seems unlikely, especially since no DMSO was detected in wine (Anocibar, 1998; Ferreira et al., 2003). Consequently, the C by-products in wine was considered equal to zero for the rest of the calculations.

In terms of repartition, the CRW sample had most of its DMS content found in the wine (98 %), while the rest was accounted for by flavour scalping (2 %) (Fig. 2). By computing this material balance, we were able to demonstrate that the amounts of degraded SMM and DMS produced in the bottle were significantly equivalent, meaning that no loss of DMS occurred during ageing for CRW samples. For MAC samples, 86 % of DMS was contained into model wine, 2 % was sorbed at the surface of the closure, and the remaining 12 % were considered to be lost by permeation during (Table 1).

Fig. 2.

Fig. 2

Open in a new tab

Repartition of the fate of formed DMS (lost DMS: DMS escaping from the inside to the outside of the bottle through wine closure following permeation), depending on the type of closure (CRW: bottles closed with crown caps; MAC: micro-agglomerated closures)- Experiments were conducted in model wine.

DMS losses by permeation were also calculated for each cork closure (ranging from 0.18 ± 2.10 μmol/L to 4.97 ± 3.91 μmol/L), and no significant differences were observed whatever the cork closure or pH (Table 2). Since O2 is well known to move from the outside to the inside of the bottle through the closure, the opposite effect could also be possible for very low molecular weight compounds such as DMS. As oxygen, DMS in gas phase would then follow Fick's laws, meaning that, due to matter diffusion, DMS would migrate from the bottle headspace (high DMS amount) to the outside of the bottle (lacking DMS). These results could represent a breakthrough in the understanding of the fate of flavour compounds, because to our knowledge, this is the first time that flavour compounds, here DMS, are prouved to migrate outside the bottle following permeation. This could also bring to light a new tool for the monitoring of the ageing wine bouquet.

Table 2.

DMS losses in μmol/L (left) and μg/L (right) through permeation depending on the type of micro-agglomerated closure (Experiments were conducted in model wine.).

Permeation of DMSa,b
μmol/L μg/L Tukey testsc
pH 2.8
C1 3.71 ± 1.42 230 ± 88 a
C5 4.58 ± 0.75 284 ± 47 a
pH 3.8
C1 2.07 ± 1.65 128 ± 102 a
C2 0.18 ± 2.18 11 ± 130 a
C3 1.09 ± 4.82 68 ± 299 a
C4 4.97 ± 3.91 308 ± 242 a
C5 3.56 ± 1.44 221 ± 89 a
Open in a new tab
a

Permeation was calculated as the difference between consumed SMM and formed DMS contents in micro-agglomerated closure samples.

b

No permeation was observed in CRW samples.

c

Significance levels at 5 % obtained via a Tukey test. The letter a means that no significant differences were found between closures, for a given pH value.

From a sensory point of view, DMS permeation ranged from 11 to 308 μg/L, with concentrations that were almost always above its odour threshold (60 μg/L (De et al., 1987);), meaning that this mechanism could have an impact on wine aromatic profile.

The correlation between DMS in wine and OTR in MAC samples was calculated using a Spearman correlation test. Since there was no significant DMS levels differences between pH 2.8 and 3.8 for the same kind of closure, the complete set of data was used for the test. The p-value obtained was equal to 0.95 which was superior to signification level (α = 0.05), meaning that the distribution between DMS levels and OTR was random and so, no relation between these 2 parameters could be found.

Since DMS permeation was only observed on micro-agglomerated wine closures, this could mean that the DMS diffusion through the micro-agglomerated wine closures is not directly correlated to oxygen permeability and probably answers to others diffusion mechanisms which could be very interesting to investigate further.

4.3. DMS permeation through closure in the case of Syrah wine

All DMS, SMM and DMSP concentrations were monitored at t0, t12 months and t24 months during ageing (Table S1).

4.3.1. DMS evolution

DMS was already present in all wines at bottling, with amounts varying from 29.9 to 314.9 μg/L for CR1 and LR1 respectively. At t12 analysis, DMS levels varied form 36.6 (LR3) to 539.8 μg/L (LR1), and then from 68.2 μg/L for LR4 to 399.5 μg/L at t24. Whatever the sample, DMS concentration was superior to its odour threshold found in wine (60 μg/L) and could therefore impact wine flavour. These concentrations were higher compared to previous studies performed on Syrah wine (Segurel et al., 2004; Bekker et al., 2016), but other work focused on Australian Cabernet Sauvignon presented similar amounts as in our study (De Mora et al., 1987). However, since DMS is a varietal aroma, it could be considered delicate to compare these works because grape variety and vineyard management could vary from one experiment to another, resulting in a difficulty to compare absolute DMS values.

In our study, DMS levels increased from 10 % to 235 % after 2 years of bottle ageing according to Syrah wines estate. Only LR1 wine bottled with C4 closure had statistically equivalent amounts throughout bottle ageing. This increase was consistent with the literature, which has already demonstrated that DMS accumulated during bottle ageing (Segurel et al., 2004; Fedrizzi et al., 2007b), which could be partially due to the hydrolysis of residual SMM in wine. From the kinetics view, several behaviours were observed during bottle ageing depending on the wine. CR1 wine showed a significant increase in DMS concentration between bottling and t12, before remaining stable until t24. For CR2, LR3 and LR4 wines, DMS levels were significantly equivalent between t0 and t12 before increasing significantly the second year of ageing. LR1 wine did not show a significant evolution whatever the time ageing. And only LR2 revealed different evolutions according to the closure used.

DMS evolution according to the permeability of the closure has also been investigated. After 2 years of bottle ageing, for 4 out of the 6 wines studied, DMS levels were significantly higher in the bottles closed with the tightest closure (C1), compared to the most permeable one (C4). These results confirmed other results found in previous studies (De La Burgade et al., 2023b), with Syrah wines being submitted to accelerated ageing. The phenomenon seemed to be pronounced for accelerated ageing, which is coherent because a high temperature could be used to catalyse chemical reactions in wine, therefore leading to higher DMS formation compared to regular bottle ageing.

4.3.2. SMM evolution

Over bottle ageing, the amounts of SMM decreased significantly for all wines, with the exception of LR4. In average, SMM concentration decreased by 15 % between bottling and 2 years of ageing, with a maximum of 27 % for LR3 wine and a minimum of 7 % for CR1 wine. These results were consistent with previous studies that confirmed the degradation of SMM during bottle ageing (Loscos et al., 2008; Bekker et al., 2018). Generally, no effect of closure was observed on the SMM degradation during aging, confirming that hydrolysis is only depending on pH and not on oxygen ingress.

Therefore, the hydrolysis of SMM in DMS could not explain the different levels of DMS found in wines depending on the closure used. The accumulation of DMS in the bottles closed with the tightest closures (C1), could be explained either by the degradation of an unknown precursor such as the dimethylsulfonium propanoic acid (DMSPA) identified in several plants (Dickschat et al., 2015), or by DMS losses through closure via a permeation phenomenon. In addition, it seemed unlikely that dimethysulfoxide (DMSO), another DMS precursor, could be at the origin of this trend since to this day, DMSO was not quantified or at very low amounts in wine (Anocibar, 1998).

4.3.3. DMS potential (DMSP)

As defined in the literature, DMSP is an analytical indicator to find out the total amount of DMS precursors, and so to quantify the maximum DMS concentration that could be formed in wine (Segurel et al., 2005). At bottling DMSP concentrations varied from 119.0 to 958.4 μg eq. DMS/L for LR4 and LR1, respectively. At t12 analysis, the amounts were from 73.7 to 575.5 μg eq. DMS/L before reaching 41.3–387.9 μg eq. DMS/L at t24, for the same wines. Generally, these levels were considerably higher compared to amounts found in previous studies, like in Syrah and Grenache wines of vintage 1992 to 2002, where DMSP concentrations varied from 8.6 to 97.1 μg eq. DMS/L (Segurel et al., 2004).

All wines followed the same evolution over time, with PDMS levels decreasing significantly each year whatever the wine and type of closure considered. The decrease is in accordance with other studies that showed that DMSP concentration decreased during bottle ageing (Segurel et al., 2005). After 2 years of bottle ageing, DMSP decreased by 61 % in average compared to its initial value after bottling, with a maximum decrease of 71 % for LR3 wine and a minimum decrease of 53 % for CR1 wine.

For each wine, a comparison of DMSP degradation during bottle ageing depending on the closure used was investigated. Generally, no significant differences were observed for each wine bottle with the 4 different closures, whatever the year of ageing. Only LR4 wine had a slightly higher degradation for the wines bottled with C1 closure compared to closures C2, C3 and C4, but this difference was equal, at most, to 12.0 μg eq. DMS/L, which could not explain the differences of DMS levels found in wine.

Since no significant difference was observed depending on the type of closure, it is possible that the accumulation of DMS in bottles sealed with the most airtight closures may not primarily result from the degradation of its precursors but from the type of closure itself.

4.3.4. Analysis of the DMS-SMM-DMSP equilibrium

In this part, all calculations performed on DMS, SMM and DMSP were reported in Table S2.

First, DMS production amounts between bottling and t24 were compared to DMSP degradation levels during the same time of ageing. For DMS, a concentration between 23.0 and 255 μg/L were formed, except for LR1 wine closed with C4 closure, for which the concentration decreased by 18.4 μg/L. After 2 years of storage, DMSP concentrations decreased from 65.6 (LR4) to 578.8 (LR1) μg eq. DMS/L. By comparing DMS formation to DMSP degradation amounts (Fig. 3 Fig. 4), it appeared that DMS formed levels were significantly inferior to DMSP degraded levels, for all wines. This means that a large part of DMS was lost during bottle ageing. But since DMS is considered as a chemically stable compound in enological conditions, its degradation into another compound seems unlikely. These results proved the observations made in model wine, and seemed to demonstrate that DMS could undergo a permeation phenomenon by migrating from the bottle headspace to the outside of the bottle through the micro-agglomerated closure.

Fig. 3.

Fig. 3

Open in a new tab

Levels of formed DMS (t24-t0) and degraded DMSP (t0-t24) after 2 years of bottle ageing for each Syrah wine and depending on the closure (letters a and b corresponded to significant differences between formed DMS and degraded DMSP, obtained via the Tukey test at 5 %).

Fig. 4.

Fig. 4

Open in a new tab

DMS losses depending on degraded DMSP after 2 years of bottle ageing for the 6 Syrah wines.

After 2 years of bottle ageing, DMS losses were significantly different depending on the closure. For 4 out of the 6 studied wines (CR1, CR2, LR1 and LR4), these losses were significantly higher for the most permeable closure (C4) compared to the tightest closure (C1). This trend was particularly pronounced for high amounts of degraded DMSP, above approximately 300 μg eq. DMS/L. Therefore, DMS permeation could be dependent to closure permeability to oxygen, in classic ageing conditions and for some enological matrixes. The effect of closure permeability to oxygen was not observed in our model wine study, but the conditions of the experiment were strictly different since the trial was performed in model wine during accelerated ageing.

The ratio between DMS produced and DMSP degraded was calculated (Table S2). In average, 35 % of DMSP degraded could be explained by the accumulation of DMS in the bottle. This ratio seemed wine dependent, since it ranged from 47 % to 81 % for CR1, while it ranged from 44 % to −3 % for LR1. In addition, DMS levels lost following permeation were evaluated as followed: CPermeation=CDMSPdegraded−CDMSinwine. Therefore, DMS permeated could vary from 15.9 μg/L for CR1 to 592 μg/L for LR1. These losses seemed to be correlated to the quantity of degraded DMSP. Indeed, DMS losses depending on degraded DMSP for the 6 studied wine were represented in Fig. 4, and we observed a linear relation between the 2 parameters (R2 = 0.92). This could mean that the greater the degradation of DMSP, the more DMS accumulates in the bottle and becomes vulnerable to permeation. Finally, as degraded DMSP increased (above 300 μg eq. DMS/L), the 4 studied closures seemed to have different DMS permeation, with C4 (the most permeable to oxygen) having the highest losses and C1 (the tightest closure) having the least losses.

The proportion of hydrolysed SMM compared to degraded DMSP was also studied, with the percentage of SMM that led to DMS formation compared to total DMSP (Table S2). In the studied wines, SMM accounted for 7 %–35 % of DMSP. This meant that in Syrah wines, at least 65 % of DMS formed during bottle ageing came from unknown precursors which have yet to be identified, and constitute interesting scientific prospects.

5. Conclusion

For the first time, we have demonstrated that DMS could escape through micro-agglomerated corks during accelerated model wine ageing through permeation with nearly 12 % of the DMS lost over a 4-month ageing period at 35 °C. Then, depending on the aroma profile the winemaker is seeking, micro-agglomerated wine closures could be used to control DMS levels during bottle ageing in order to guarantee wine quality. This study highlights the dual role of closures in managing oxygen ingress and sulfur volatile compound retention. Low-permeability closures can help preserve DMS in wines where it enhances aromatic complexity, while higher-permeability closures might mitigate reductive off-flavors in wines prone to sulfur compounds. These findings may account for the reductive notes empirically observed in red wines sealed with highly impermeable closures compared to those bottled with more permeable alternatives. The latter could be advantageous by facilitating the permeation of light volatile sulfur compounds, such as H2S, MeSH, or EtSH.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

We acknowledged Diam Bouchage for funding this project.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.crfs.2025.101089.

Appendix A. Supplementary data

The following is the Supplementary data to this article:

Multimedia component 1
mmc1.docx (150.8KB, docx)

Data availability

Data will be made available on request.

References

  1. Anness B., Bamforth C.W. Dimethyl sulphide—a review. J. Inst. Brew. 1982;88(4):244–252. [Google Scholar]
  2. Anocibar Beloqui A. 1998. Les composés soufrés volatils des vins rouges. [PhD]: Victor Segalen Bordeaux II. [Google Scholar]
  3. Azevedo J., Lopes P., Mateus N., de Freitas V. Cork, a natural choice to wine? Foods. 2022;11(17):2638. doi: 10.3390/foods11172638. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bekker M.Z., Mierczynska-Vasilev A., Smith P.A., Wilkes E.N. The effects of pH and copper on the formation of volatile sulfur compounds in Chardonnay and Shiraz wines post-bottling. Food Chem. 2016;207:148–156. doi: 10.1016/j.foodchem.2016.03.060. [DOI] [PubMed] [Google Scholar]
  5. Bekker M.Z., Wilkes E.N., Smith P.A. Evaluation of putative precursors of key ‘reductive’compounds in wines post-bottling. Food Chem. 2018;245:676–686. doi: 10.1016/j.foodchem.2017.10.123. [DOI] [PubMed] [Google Scholar]
  6. Beloqui A.A., Kotseridis Y., Bertrand A. Determination of the content of dimethyl sulphide in some red wines. OENO One. 1996;30(3):167–170. [Google Scholar]
  7. Buser H.R., Zanier C., Tanner H. Identification of 2, 4, 6-trichloroanisole as a potent compound causing cork taint in wine. J. Agric. Food Chem. 1982;30(2):359–362. [Google Scholar]
  8. Capone D., Sefton M., Pretorius I., Høj P. 2003. Flavour'scalping'by Wine Bottle Closures-The'winemaking'continues Post Vineyard and Winery. [Google Scholar]
  9. Cremer D.R., Eichner K. Formation of volatile compounds during heating of spice paprika (capsicum a nnuum) powder. J. Agric. Food Chem. 2000;48(6):2454–2460. doi: 10.1021/jf991375a. [DOI] [PubMed] [Google Scholar]
  10. Crouvisier-Urion K., Bellat J.-P., Gougeon R.D., Karbowiak T. Gas transfer through wine closures: a critical review. Trends Food Sci. Technol. 2018;78:255–269. [Google Scholar]
  11. De Mora S.J., Eschenbruch R., Knowles S.J., Spedding D.J. The formation of dimethyl sulphide during fermentation using a wine yeast. Food Microbiol. 1986;3(1):27–32. [Google Scholar]
  12. De Mora S., Knowles S., Eschenbruch R., Torrey W. Dimethyl sulfide in some Australian red wines. Vitis. 1987;26(2):79–84. [Google Scholar]
  13. De La Burgade R., Nolleau V., Godet T., Galy N., Tixador D., Loisel C., et al. New insights on the scalping phenomenon of volatile sulphur compounds on micro-agglomerated wine closures. Molecules. 2023;28(13):5094. doi: 10.3390/molecules28135094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. De La Burgade R., Nolleau V., Godet T., Galy N., Tixador D., Loisel C., et al. Wine Cork closures impacts on dimethyl sulfide (DMS) and precursors (DMSP) equilibrium of different shiraz wines during accelerated bottle ageing. Beverages. 2023;9(1):15. [Google Scholar]
  15. Deed R.C., Pilkington L.I., Herbst‐Johnstone M., Miskelly G.M., Barker D., Fedrizzi B. A new analytical method to measure S‐methyl‐l‐methionine in grape juice reveals the influence of yeast on dimethyl sulfide production during fermentation. J. Sci. Food Agric. 2019;99(15):6944–6953. doi: 10.1002/jsfa.9983. [DOI] [PubMed] [Google Scholar]
  16. Demora S.J., Knowles S.J., Eschenbruch R., Torrey W.J. Dimethyl sulfide in some AUSTRALIAN red wines. Vitis. 1987;26(2):79–84. [Google Scholar]
  17. Dickschat J.S., Rabe P., Citron C.A. The chemical biology of dimethylsulfoniopropionate. Org. Biomol. Chem. 2015;13(7):1954–1968. doi: 10.1039/c4ob02407a. [DOI] [PubMed] [Google Scholar]
  18. Escudero A., Campo E., Farina L., Cacho J., Ferreira V. Analytical characterization of the aroma of five premium red wines. Insights into the role of odor families and the concept of fruitiness of wines. J. Agric. Food Chem. 2007;55(11):4501–4510. doi: 10.1021/jf0636418. [DOI] [PubMed] [Google Scholar]
  19. Fedrizzi B., Magno F., Moser S., Nicolini G., Versini G. Concurrent quantification of light and heavy sulphur volatiles in wine by headspace solid-phase microextraction coupled with gas chromatography/mass spectrometry. Rapid Commun. Mass Spectrom. 2007;21:707–714. doi: 10.1002/rcm.2893. [DOI] [PubMed] [Google Scholar]
  20. Fedrizzi B., Magno F., Badocco D., Nicolini G., Versini G. Aging effects and grape variety dependence on the content of sulfur volatiles in wine. J. Agric. Food Chem. 2007;55(26):10880–10887. doi: 10.1021/jf072145w. [DOI] [PubMed] [Google Scholar]
  21. Ferreira A.C.S., Rodrigues P., Hogg T., De Pinho P.G. Influence of some technological parameters on the formation of dimethyl sulfide, 2-mercaptoethanol, methionol, and dimethyl sulfone in port wines. J. Agric. Food Chem. 2003;51(3):727–732. doi: 10.1021/jf025934g. [DOI] [PubMed] [Google Scholar]
  22. Furtado I., Lopes P., Oliveira A.S., Amaro F., Bastos MdL., Cabral M., et al. The impact of different closures on the flavor composition of wines during bottle aging. Foods. 2021;10(9):2070. doi: 10.3390/foods10092070. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Jimenez-Lorenzo R., Bloem A., Farines V., Sablayrolles J.M., Camarasa C. How to modulate the formation of negative volatile sulfur compounds during wine fermentation? FEMS Yeast Res. 2021;21(5) doi: 10.1093/femsyr/foab038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Karbowiak T., Gougeon R.D., Alinc J.-B., Brachais L., Debeaufort F., Voilley A., et al. Wine oxidation and the role of cork. Crit. Rev. Food Sci. Nutr. 2009;50(1):20–52. doi: 10.1080/10408398.2010.526854. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Karbowiak T., Mansfield A.K., Barrera-García V.D., Chassagne D. Sorption and diffusion properties of volatile phenols into cork. Food Chem. 2010;122(4):1089–1094. [Google Scholar]
  26. Landaud S., Helinck S., Bonnarme P. Formation of volatile sulfur compounds and metabolism of methionine and other sulfur compounds in fermented food. Appl. Microbiol. Biotechnol. 2008;77(6):1191–1205. doi: 10.1007/s00253-007-1288-y. [DOI] [PubMed] [Google Scholar]
  27. Lopes P., Marques J., Lopes T., Lino J., Coelho J., Alves C., et al. Permeation of d5-2, 4, 6-Trichloroanisole via vapor phase through different closures into wine bottles. Am. J. Enol. Vitic. 2011;62(2):245–249. [Google Scholar]
  28. Loscos N., Ségurel M., Dagan L., Sommerer N., Marlin T., Baumes R. Identification of S-methylmethionine in Petit Manseng grapes as dimethyl sulphide precursor in wine. Anal. Chim. Acta. 2008;621(1):24–29. doi: 10.1016/j.aca.2007.11.033. [DOI] [PubMed] [Google Scholar]
  29. Lytra G., Tempere S., Zhang S., Marchand S., De Revel G., Barbe J.-C. Olfactory impact of dimethyl sulfide on red wine fruity esters aroma expression in model solution. Oeno One. 2014;48(1):75–85. [Google Scholar]
  30. Marais J. Effect of storage time and temperature on the formation of dimethyl sulfide and on white wine quality. Vitis. 1979;18(3):254–260. [Google Scholar]
  31. Segurel M.A., Razungles A.J., Riou C., Salles M., Baumes R.L. Contribution of dimethyl sulfide to the aroma of Syrah and Grenache noir wines and estimation of its potential in grapes of these varieties. J. Agric. Food Chem. 2004;52(23):7084–7093. doi: 10.1021/jf049160a. [DOI] [PubMed] [Google Scholar]
  32. Ségurel M.A., Razungles A.J., Riou C., Trigueiro M.G.L., Baumes R.L. Ability of possible DMS precursors to release DMS during wine aging and in the conditions of heat-alkaline treatment. J. Agric. Food Chem. 2005;53(7):2637–2645. doi: 10.1021/jf048273r. [DOI] [PubMed] [Google Scholar]
  33. Segurel M.A., Razungles A.J., Riou C., Trigueiro M.G., Baumes R.L. Ability of possible DMS precursors to release DMS during wine aging and in the conditions of heat-alkaline treatment. J. Agric. Food Chem. 2005;53(7):2637–2645. doi: 10.1021/jf048273r. [DOI] [PubMed] [Google Scholar]
  34. Silva M.A., Julien M., Jourdes M., Teissedre P.L. Impact of closures on wine post-bottling development: a review. Eur. Food Res. Technol. 2011;233(6):905–914. [Google Scholar]
  35. Silva M.A., Jourdes M., Darriet P., Teissedre P.L. Scalping of light volatile sulfur compounds by wine closures. J. Agric. Food Chem. 2012;60(44):10952–10956. doi: 10.1021/jf303120s. [DOI] [PubMed] [Google Scholar]
  36. Ugliano M. Oxygen contribution to wine aroma evolution during bottle aging. J. Agric. Food Chem. 2013;61(26):6125–6136. doi: 10.1021/jf400810v. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.docx (150.8KB, docx)

Data Availability Statement

Data will be made available on request.

Articles from Current Research in Food Science are provided here courtesy of Elsevier

RESOURCES