Logo
 
MoreWineMaking!

Benchmarking of SO2 Analysis Instruments and Methods in Wine Applications

01/31/2014

By Daniel Pambianchi

Abstract: Free sulfur dioxide (SO2) is a key parameter monitored throughout the winemaking process and at bottling to ensure wine is adequately protected from enzymatic and chemical oxidative effects and microbial spoilage. The aim of this study was 1) to benchmark accuracy and precision of various instruments and methods, i.e. aeration–oxidation (AO) and Ripper, available on the market for measuring free SO2 levels in wine, and 2) to determine any impacts from ascorbic acid and tannins as these may interfere with test chemistry. The AO methods measured free SO2 levels most accurately though some results were outside error margins. Titrets measured free SO2 levels most accurately and precisely even though they have a high error; however, these cannot be used in red wine due to the high polyphenol content that interferes with the tests. The Vinmetrica SC-300 had good precision; its accuracy was within error margins. The Hanna 84500 unit had variable accuracy and precision. The Quick Tests results were difficult to interpret and therefore their accuracy is uncertain, but tests are precise. Only the AO methods were relatively unaffected by the presence of ascorbic acid.

Key words: sulfur dioxide, sulfite, aeration-oxidation, Ripper method, Hanna, Vinmetrica, Accuvin

 

Introduction

 
Sulfur dioxide (SO2) has long been used in winemaking to protect wine from enzymatic and chemical oxidative effects and microbial spoilage. It can be added in gaseous form or, most common, from a sulfite salt, such as potassium metabisulfite. In aqueous solutions, SO2, bisulfite (HSO3–) and sulfite (SO32–) ions exist in equilibrium as per the equation:
 
SO2•H2O ↔ H+ + HSO3– ↔ 2 H+ + SO32–
 
     The sum of SO2, HSO3– and SO32– concentrations is referred to as free SO2, or FSO2, and is the active form that affords protection in wine.
     At wine pH, usually in the range 3–4, HSO3– is the most abundant form representing about 94–99% of the total, the rest being SO2; SO32– is negligible.
     FSO2 diminishes over time as SO2 is lost to the atmosphere via tank or barrel headspace or through bottle corks, as HSO3– binds with carbonyl (e.g. acetaldehyde and ketone acids) and phenolic compounds (e.g. anthocyanins and tannins), and as HSO3– reduces o-quinones back to their phenol forms. During alcoholic fermentation, S. cerevisiae yeast produces small amounts of FSO2, in the order of 10 mg/L, but, some strains have also been shown to be able to metabolize HSO3– and reduce it into hydrogen sulfide (H2S) although this trait appears to be rare (Linderholm and Bisson 2005). Winemakers therefore need to add more sulfite to maintain a nominal level based on pH, according to the following relationship, while ensuring that total SO2 (the sum of free and bound SO2) never exceeds the maximum set by regulatory agencies, where applicable.
 
     For example, a red wine with a pH of 3.20 and to be protected with 0.5 mg molecular SO2/L would require approximately 13 mg FSO2/L. FSO2 should never be allowed to drop below 8–9 mg/L (Stelzer et al. 2005).  
     Various apparatus and methods are available for measuring FSO2 in wine. Although there are several variants, these operate on one of two principles: Ripper chemistry and Monier–Williams method (Zoecklein et al. 1999; Pegram et al. 2013), which is based on aeration–oxidation (AO) chemistry.
     The Ripper determination of SO2 is based on the oxidation–reduction reaction (Ough and Amerine 1988):
 
SO2 + I3– + H2O → SO3 + 3 I¯ + 2 H+
 
     The wine sample is first acidified to reduce the oxidation of polyphenols by iodine, then titrated with iodine to a starch endpoint. This method works well with white wines; however, tannins and anthocyanins in reds cause iodine reduction and false results.
     A variation of this method generates iodine from an iodate solution, which is more stable, instead of iodine; the reactions are:
 
5 I¯ + IO3¯ + 6 H+ → 3 I2 + 3 H2O
I2 + SO2 + H2O → SO3 + 2 I¯ + 2 H+
 
     The AO method involves acidifying the wine sample with phosphoric acid to help volatize the SO2. A stream of air is passed through the acidified sample and the freed SO2 is collected and oxidized in a hydrogen peroxide (H2O2) solution to produce sulfuric acid (H2SO4) as per the reaction:
 
SO2 + H2O2 → SO3 + H2O → 2 H+ + SO42¯
 
     The sulfuric acid solution is then titrated with a base (e.g. 0.01N NaOH) to a known endpoint. This method, however, causes ascorbic acid to oxidize to H2O2, which then reacts with free SO2 and therefore yields false results if excessive amounts of the acid are used.
     In both methods, FSO2 in mg/L is determined by the relationship (Ough and Amerine 1988):
 
V is the volume (mL) of titrant used, N is the normality of the titrant and v is the volume (mL) of the wine sample.
 
     The purpose of this study was to benchmark six different kinds of apparatus and methods for accuracy (the degree of closeness of measurements of a quantity to that quantity's actual [true] value) and precision (the degree to which repeated measurements under unchanged conditions show the same results) and any impacts from ascorbic acid and tannins: AO method using classic laboratory apparatus and a second method using a scaled-down (home winemaking) version, Ripper-method Titrets® that use an iodide-iodate titrant vacuum-sealed in a bulb, Vinmetrica SC-300™ and Hanna® 84500 titrator units that measure conductivity during Ripper titration with iodate, and Accuvin Quick Tests™ that use a proprietary dye that reacts with SO32– in the treated sample.
 

Materials and Methods

 
     Test Equipment. R&D Apparatus for SO2 Determination by Aeration-Oxidation purchased from Research & Development Glass Products & Equipment Inc., Berkeley, CA; Vinmetrica SC-300 SO2 & pH/TA Analyzer Kit purchased from MoreWine! Concord, CA.; Hanna HI 84500 Sulphur Dioxide in Wine Titrator purchased from Hanna Instruments, Laval, Québec, Canada (via Prolab Scientific, Laval, Québec); MT140 Economy Aeration–Oxidation Free SO2 Test Kit purchased from MoreWine! Concord, CA; CHEMetrics Sulfite in Wine Titrets Kit purchased from Vines to Vintages, Niagara, Ontario; and Quick Tests Free SO2 purchased from Accuvin, Napa, CA.
     Instrumentation. Syringes and other volumetric apparatus supplied with the instruments were substituted for high-accuracy pipettes to minimize sample errors. Test samples were obtained using the same pipette or pipettes of similar accuracy.
    Test equipment was calibrated prior to testing. Reagents were purchased fresh or prepared fresh. The 0.01N NaOH titrant was standardized against a potassium acid phthalate solution. Potassium metabisulfite (KMS) was purchased fresh. Accuracy and resolution were recorded for all instrumentation.
     Model Solutions. Three model solutions with 35 mg FSO2/L were prepared using a volumetric flask, a 10% sulfite solution and distilled water acidified with tartaric acid to a pH of approximately 3.3. One was a control solution to benchmark FSO2 measurements and to compare results with a second solution containing 20 mg/L of ascorbic acid, typical of use in white winemaking, and a third solution containing 2 g/L of grape tannins, which represents a highly tannic wine. Solutions were immediately transferred to a sufficient number of 60-mL bottles, fully topped and capped, to run 5 tests for each instrument or method. All samples were held at ambient temperature of approximately 21⁰C (70⁰F).
     Test Procedure. Each instrument or method was tested by measuring FSO2 in a sulfited sample and then repeated for a total of 5 times. The tests were repeated using sulfited samples with ascorbic acid. The tests were again repeated using sulfited samples with grape tannins.
     Test Errors. Errors on all instrumentation were recorded and factored into test results where possible. Errors that could not be quantified are discussed below.
 

Results and Discussion

 
Refer to the data in Tables 1, 2 and 3 and the corresponding graphs in Figures 1, 2 and 3.
 
Aeration–Oxidation Method.  Both the R&D and MT140 test units provided measurements with good precision and accuracy within the margin of error. The tests were relatively unaffected by the presence of ascorbic acid. The results for both units were consistent although they measured lower FSO2 levels in the presence of grape tannins. As the AO method was not expected to be affected by grape tannins, it is assumed that the lower FSO2 levels were due to the binding action between SO2 and tannins.

 

Figure 1 Distribution of free SO2 measurements (circles) and average (black dashes) per instrument/method using a solution with 35 mg FSO2/L (gray rectangle, which includes FSO2 error).

Ripper Method. The Vinmetrica SC-300 unit and Titrets provided measurements with good precision and accuracy within the margin of error, whereas the Hanna unit was less precise. The tests were affected by the presence of ascorbic acid.

Figure 2 Distribution of free SO2 measurements (circles) and average (black dashes) per instrument/method using a solution with 35 mg FSO2/L (gray rectangle, which includes FSO2 error) and 20 mg ascorbic acid/L.

     The Vinmetrica and Hanna units measured lower FSO2 levels in the presence of grape tannins, but this is assumed to be due to the binding action between SO2 and tannins. Titrets were significantly affected by the presence of grape tannins, recording measurements in excess of the 100 mg/L test limit. A test using only 200 mg/L of tannins instead of 2 g/L had negligible impact; therefore, white wines with low levels of tannins can still be measured with Titrets.
Test Errors. Test results include instrumentation errors, except for the sample size used with the Quick Tests. For the AO methods, the 0.01N NaOH solution was standardized. Instrumentation errors have also been factored into the preparation of the 10% SO2 solution and model solutions.
     The manufacturers’ specs on errors for FSO2 in the 35 mg/L range are: Hanna 84500 (3%), Titrets (±5 mg/L), and Quick Tests (±4 mg/L). No error data was available for the Vinmetrica SC-300 unit; a 2% error was assumed in the calculations.
     AO methods, Titrets and Quick Tests rely on color changes to determine the titration endpoint. Test errors can be significant with inexperienced users. Titrets can have an additional error from sampling model solutions as the solution is drawn in by vacuum in the bulbs and cannot be controlled precisely.
     The AO methods proved to be the most error prone if solutions are not fresh or if the tests are not performed carefully. Leaks in the aspiration sample may result in a loss of free SO2 and skewed test results. The aspiration flow rate is also important; an inadequate rate may cause the loss of SO2 or SO2 that could not be dissolved in the H2O2 solution. The methods should be performed using a flow rate of 1 L/min measured with a flowmeter (Iland et al. 2000). The tests here were not performed using a flowmeter instead relying on user experience. Test results using the model solution containing grape tannins were impacted by the test delays as SO2 immediately starts binding with grape tannins when the solution is prepared.
Figure 3 Distribution of free SO2 measurements (circles) and average (black dashes) per instrument/method using a solution with 35 mg FSO2/L (gray rectangle, which includes FSO2 error) and 2 g grape tannins/L.
 

Conclusions

 
     The AO methods measured free SO2 levels most accurately though some results were outside error margins. Titrets measured free SO2 levels most accurately and precisely even though they have a high error; however, these cannot be used in red wine due to the high polyphenol content that interferes with the tests. The Vinmetrica SC-300 had good precision; its accuracy was within error margins. The Hanna 84500 unit had variable accuracy and precision. The Quick Tests results were difficult to interpret and therefore their accuracy is uncertain, but tests are precise. Only the AO methods were relatively unaffected by the presence of ascorbic acid.
     Although test results are well outside of the error margins of the model solutions, a free SO2 error of ±5 mg/L is considered acceptable in the 35 mg/L range at which these tests were executed. As a possible future study, the same tests can be performed in model solutions with, for example, 10–15 mg/L, to assess errors where lower free SO2 levels may be more of a concern to those wanting to minimize sulfite use in wines.
     Another buying consideration is cost effectiveness. Approximate suggested retailer prices (in $US) are: AO R&D Apparatus ($420) can perform free and total SO2 tests but requires all reagents to be purchased separately; MT-140 kit can perform 2–3 free SO2 tests, then more reagents must be purchased; Vinmetrica SC-300 ($350) can perform up to 50 free and total SO2 tests as well as 30 TA/pH tests; Hanna HI 84500 ($850) can perform up to 50 free and total SO2 tests; disposable CHEMetrics Titrets ($19) can perform 10 free SO2 tests; and disposable Accuvin Quick Tests ($65) can perform 20 (4 high-range plus 16 low-range) free SO2 tests.
 
Sample AO R&D AO MT140      Vinmetrica SC-300 Hanna 84500 CHEMetrics Titrets Accuvin Quick Tests

1

2

3

4

5

Avg

35.2 ± 0.7

35.2 ± 0.7

33.6 ± 0.7

35.2 ± 0.7

33.6 ± 0.7

34.6

33.6 ± 0.7

33.6 ± 0.7

35.2 ± 0.7

33.6 ± 0.7

35.2 ± 0.7

34.2

32.3 ± 1.7

33.8 ± 1.7

34.3 ± 1.7

34.2 ± 1.7

34.2 ± 1.7

33.8

34.5 ± 1.9

37.4 ± 1.9

38.0 ± 1.9

36.4 ± 1.9

29.1 ± 1.9

35.1

35 ± 5

36 ± 5

36 ± 5

34 ± 5

35 ± 5

35

28 ± 4

28 ± 4

28 ± 4

28 ± 4

28 ± 4

28

Table 1 Free SO2 measurements in mg/L with errors and average per instrument or method using a solution with 35 mg FSO2/L.

Sample AO R&D AO MT140      Vinmetrica SC-300 Hanna 84500 CHEMetrics Titrets Accuvin Quick Tests

1

2

3

4

5

Avg

35.2 ± 0.7

33.6 ± 0.7

35.2 ± 0.7

33.6 ± 0.7

33.6 ± 0.7

34.2

33.6 ± 0.7

33.6 ± 0.7

35.2 ± 0.7

33.6 ± 0.7

33.6 ± 0.7

33.9

38.6 ± 1.7

40.0 ± 1.7

40.4 ± 1.7

40.0 ± 1.7

39.6 ± 1.7

39.7

40.0 ± 1.9

40.0 ± 1.9

40.0 ± 1.9

40.0 ± 1.9

40.0 ± 1.9

40.0

45 ± 5

45 ± 5

40 ± 5

37 ± 5

36 ± 5

41

34 ± 4

34 ± 4

34 ± 4

34 ± 4

34 ± 4

34

Table 2 Free SO2 measurements in mg/L with errors and average per instrument or method using a solution with 35 mg FSO2/L with 20 mg ascorbic acid/L.

Sample AO R&D AO MT140      Vinmetrica SC-300 Hanna 84500 CHEMetrics Titrets Accuvin Quick Tests

1

2

3

4

5

Avg

30.4 ± 0.7

28.8 ± 0.7

28.8 ± 0.7

30.4 ± 0.7

30.4 ± 0.7

29.8

28.8 ± 0.7

28.8 ± 0.7

30.4 ± 0.7

30.4 ± 0.7

28.8 ± 0.7

29.4

31.2 ± 1.7

31.6 ± 1.7

31.6 ± 1.7

31.4 ± 1.7

31.6 ± 1.7

31.5

32.9 ± 1.9

35.5 ± 1.9

32.0 ± 1.9

34.8 ± 1.9

35.8 ± 1.9

34.2

>100

>100

>100

>100

>100

>100

34 ± 4

34 ± 4

34 ± 4

34 ± 4

34 ± 4

34

Table 3 Free SO2 measurements in mg/L with errors and average per instrument or method using a solution with 35 mg FSO2/L with 2 g grape tannins/L.

Literature Cited

 
Iland, P., A. Ewart, J. Sitters, A. Markides and N. Bruer. 2000. Techniques for chemical analysis and quality monitoring during winemaking. Patrick Iland Wine Promotions, Campbelltown (Australia).
 
Linderholm, A.L. and L.F. Bisson. 2005. Eliminating formation of hydrogen sulfide by Saccharomyces. Practical Winery & Vineyard Journal. Nov-Dec 2005: 65-76.
 
Ough, C.S. and M.A. Amerine. 1988. Methods for Analysis of Musts and Wines. John Wiley & Sons, New York (NY).
 
Pegram, Z., M.T. Kwasniewski and G.L. Sacks. 2013. Simplified Method for Free SO2 Measurement Using Gas Detection Tubes. Am. J. Enol. Vitic. 64:405-410.
 
Stelzer, T., J. Grosset, M. Brajkovich, J. Forrest and B. Rankine. 2005. Taming the screw: a manual for winemaking with screw caps. Wine Press, Brisbane (Australia).
 
Zoecklein, B.W., K.C. Fugelsang, B.H. Gump and F.S. Nury. 1999. Wine Analysis and Production. Aspen Publishers, Gaithersburg (MD).
 
**MoreWine! would like to thank Daniel Pambianchi for his time and knowledge dedicated to this article.  If you'd like to read more from Daniel, please visit his website at:
 
 
To view more SO2 testing devices and information, click here.

Copyright © 2014 by Daniel Pambianchi. All rights reserved.

Related Products

All contents copyright 2024 by MoreFlavor Inc. All rights reserved. No part of this document or the related files may be reproduced or transmitted in any form, by any means (electronic, photocopying, recording, or otherwise) without the prior written permission of the publisher.