Technical Note pubs.acs.org/ac
Angel’s Share Combats Wine Fraud: Opening the Bottle
14
C Dating of Wine without
Simon M. Fahrni,*,‡ Benjamin T. Fuller,§ and John R. Southon Department of Earth System Science, University of California, Irvine, Keck CCAMS Group, B321 Croul Hall, Irvine, California 92697, United States S Supporting Information *
ABSTRACT: The problem of fraud continues to plague the wine industry, and detecting cases where the original bottle has been refilled with an inferior vintage is especially difficult. A novel noninvasive procedure presented here relies on radiocarbon dating the so-called angel’s share: the trace amounts of ethanol and other gases that diffuse into and through the cork as bottled wine ages and matures. These are collected by applying a vacuum to the top of the bottle and cryo-trapping the extracted gas, leaving the liquid contents untouched. Vintage verification is therefore possible without exposing the liquid contents to the atmosphere, which may render a bottle costing tens of thousands of dollars worthless for later resale and consumption. The method also has a potential for stable isotope ratio as well as chemical analysis to find indications for fraud or spoilage of fine and rare wines as well as other liquids with cork stoppers.
F
raud is nothing new in the wine business,1 but the recent surge in prices of fine and rare wines has added incentive for forgery;2 it is estimated that 5% of the wine sold in secondary markets could be fraudulent.3 In the past, the ability to detect fraudulent wine or to check for spoilage markers has typically required opening the bottle to sample the liquid, a destructive process that can render the wine no longer storable and thus worthless for future resale. In response, several noninvasive methods for wine fraud detection and quality control have been developed and applied in recent years.4 For example, wines from the preatomic (before 1945) and the atomic age (1945−present) can be distinguished by detecting the characteristic γ-rays of the anthropogenic radionuclide 137Cs from wine in unopened bottles,5,6 but large and irregular variations in the spatial and temporal distribution of 137Cs7 can make it difficult to attribute the wine to a specific vintage year. Particle induced X-ray emission (PIXE) can be used to determine the elemental composition of the wine bottle’s glass,4,8 which can help determine the origin and in some cases the age range of the bottle, provided an authentic “fingerprint” is available. However, glass composition may remain constant over long periods, and since only the glass is analyzed, authentic bottles may mask fraudulent contents.9,10 The recently developed diamagnetic screening technique compares the frequency-dependent responses of the contents of a test bottle and a verified reference bottle to low frequency radio waves,11−13 and an analogous technique using optical absorption spectroscopy in the near-infrared (NIR) has also been applied.14−16 Nondestructive techniques that primarily focus on the detection of spoilage markers include NMR (nuclear magnetic resonance) spectroscopy for the detection of acetic acid17,18 and low pressure microextraction combined © XXXX American Chemical Society
with GC/MS for identifying trichloroanisole (TCA), the main compound responsible for cork-taint.19,20 These methods can provide very useful information on authenticity and/or quality of the wine, but they either require a comparative sample or are limited in the extent to which wine can be dated to a specific vintage year. In contrast to the previously described methods, the extraction of wine vapors from the cork potentially allows simultaneous authentication and measurement of a range of spoilage markers without the need for comparative samples of verified origin, vintage, and quality. Our extraction technique relies on applying a vacuum to the cork and is thus similar to the method of Lim et al.,19 but those researchers analyzed a cork-borne contaminant whereas the present work aims at testing the actual contents of the bottle. Most traditional wine closures are made from the bark of the cork oak tree (Quercus suber L.). Cork allows some exchange of gases and vapors between the ullage (the headspace between the liquid and the cork) and air. Exposure to atmospheric oxygen plays a role in the maturation of wine,21−23 and several uptake and transport mechanisms of compounds into and through cork have been described in the literature, including sorption,24 diffusion25−in particular through small channels between cells (plasmodesmata),26−or even diffusion along the cork-glass boundary.27 The same properties that allow the permeation of oxygen into the bottle also allow water, ethanol, and other components of wine to diffuse into and through the Received: May 28, 2015 Accepted: July 24, 2015
A
DOI: 10.1021/acs.analchem.5b01998 Anal. Chem. XXXX, XXX, XXX−XXX
Analytical Chemistry
■
cork. The vapor pressure and sorption properties of the species of interest as well as the cork structure and the resulting diffusion coefficients determine the outgassing rates of different species. However, since the closure is intended to preserve the bulk of the contents for many years or even decades, the permeability of cork is strictly limited, and a vacuum applied to the top of the bottle for a short period will typically only affect the top of the cork, leaving the larger fraction of the closure and the contents of the bottle unaffected. Absorbed gases and more importantly absorbed liquid vapors will volatilize under vacuum and can be measured directly or collected cryogenically for later analysis. In addition to radiocarbon dating, this method should be applicable to study the stable isotope ratios of wine ethanol (δD, δ13C, δ18O), which can be important tools for the verification of the region of origin28 and for chemical analysis of extracted volatiles to detect spoilage markers. However, the present proof of concept study focuses on radiocarbon measurements because sample sizes needed for radiocarbon analysis are tens of micrograms of carbon and are thus more demanding than typical requirements for other analyses. The advantage of 14C dating over the relative dating methods discussed above is that single bottles of wine can be dated almost independently of the origin and without the need of a verified reference sample. Accurate radiocarbon dating of modern vintages is a legacy of above ground nuclear weapons tests in the 1950s and early 1960s, which produced a global spike in radiocarbon in atmospheric CO2 followed by a decline as atmospheric CO2 exchanged with other carbon reservoirs (Figure 1).29 Dating of
Technical Note
RESULTS
Here, we demonstrate the feasibility of the method by dating known aged wines from vintages between 1934 and 2006, obtained from a private collection. All bottles were in optimal condition and showed no sign of abnormal cork leakage or damage, either before or after the extractions. Wine vapors were extracted for 20 to 180 min, and samples were converted to filamentous graphite for analysis by accelerator mass spectrometry (AMS). 14 C values of samples were plotted together with two Northern Hemisphere atmospheric 14C records35−37 for comparison in Figure 1 and converted into a calibrated age interval with the CALIBomb software.38 The results and comparisons of vintages with the measured ages are listed in Table 1, and two examples of calibrated radiocarbon values are shown in Figure S1 for illustration of the calibration precision. A total of 32 wines were tested, of which 23 (72%) were unambiguously identified to the correct vintage. Of the remaining 9 samples, 5 yielded gas extracts that were too small (2−8 μg of carbon) for accurate age determinations. This was possibly due to corks of low permeability and small diffusion coefficients, which is expected from the distribution of diffusion coefficients (spanning several orders of magnitude) compiled by Lequin et al.25 These ultrasmall samples were strongly affected by nonsample carbon contamination, and the required background corrections increased the overall uncertainty of the 14C values to the point where no useful age information was obtained (Table 1). The remaining four outliers all yielded sufficiently large samples for the radiocarbon method but returned anomalously low radiocarbon values, suggesting a systematic contamination with “old” carbon. In principle, such contaminants could be characterized by, e.g., chromatographic separation of the extracted gases, but this goes beyond the scope of the present study. Also, due to their age and value, the bottles were not opened to take bulk comparison samples. The origin of these outliers therefore remains unclarified, but since these wines were purchased firsthand within a few years of the vintage, when they were more readily available and relatively inexpensive,39 the possibility of forgery seems remote. The 2 oldest wines, the 1934 and 1945 Clos des Lambrays, were recorked in France in the early 1980s,39 which may be the reason for the below modern value of the 1945 wine. However, while topping up of wines with the original vintage during recorking is a standard procedure, it is unlikely that the 1945 Clos des Lambrays would have been refilled with 10−15% of radiocarbon-dead (i.e., fossil fuel based) ethanol required to yield the measured F14C value. In the case of forgery, refilling with cheap wine would have been much more likely, but topping up with a younger wine would have led to an increased F14C value due to bomb-14C in younger wines. As a more likely explanation, the low F14C values may reflect unusual variations in the volatility of fossil fuel based substances used in standard cork treatments, such as lubrication (e.g., paraffin wax),40 or inward diffusion of external old carbon contamination during cellaring. Nevertheless, the authenticity and integrity of the 1945 Clos des Lambrays remain questionable. Our results demonstrate that 14C can be used noninvasively to determine wine vintages, though in its current state of development the method can produce a percentage of false positive results, indicating potential fraud when bottles are in fact authentic. The technique is most accurate for wines bottled
Figure 1. 14C bomb curve between 1950 and 201035,36 depicted as the black line, and the data of the 32 analyzed wine bottles are shown as red circles. Before 1950, the regular radiocarbon calibration37 is shown for comparison with wines. Error bars indicate the uncertainties of the 14 C values after all background corrections. Asterisked samples were too small for reliable dating and thus yielded ambiguous results (F14C values indicated were outside of the graph area).
wines using this 14C bomb curve is not new, but past measurements30−33 required milligram or larger sample sizes taken from the liquid itself. However, measurement of 14C in samples of ∼20 μg of carbon or less is now quite routine (e.g., Santos et al.),34 and this allows for the dating of the trace amounts of vapors stored in the cork. B
DOI: 10.1021/acs.analchem.5b01998 Anal. Chem. XXXX, XXX, XXX−XXX
Technical Note
Analytical Chemistry Table 1. Compilation of the Sample Data, Sorted According to Vintage year 1934 1945 1961 1962 1966 1967 1968 1970 1970 1970 1974 1975 1976 1976 1977 1977 1979 1981 1981 1982 1983 1984 1985 1986 1987 1988 1989 1991 1997 2000 2002 2006
wine name Clos des Lambrays Clos des Lambrays Château Latour Château Lafite-Rothschild Château Mouton Rothschild Château Mouton Rothschild Conterno Monfortino Barolo Château Giscours Margaux Château Mouton Rothschild Château Haut-Brion Sebastiani Cabernet Sauvignon Robert Mondavi Cabernet Sauvignon Mayacamas Cabernet Sauvignon Mayacamas Cabernet Sauvignon Robert Mondavi Cabernet Sauvignon (Reserve) Robert Mondavi Cabernet Sauvignon (Reserve) Robert Mondavi Cabernet Sauvignon Robert Mondavi Cabernet Sauvignon Mayacamas Chardonnay Jordan Cabernet Sauvignon Stag’s Leap Wine Cellars Cabernet Sauvignon Stag’s Leap Wine Cellars Cabernet Sauvignon Cask 23 Sanctus Jacobus Piesporter Goldtröpfchen Riesling Duckhorn Sauvignon Blanc Stag’s Leap Wine Cellars Cabernet Sauvignon SLV Hawk Crest California Chardonnay Stag’s Leap Wine Cellars Sauvignon Blanc Stag’s Leap Wine Cellars Cabernet Sauvignon FAY Stag’s Leap Wine Cellars Cabernet Sauvignon FAY Stag’s LeapWine Cellars Sauvignon Blanc Château Haut-Brion Beringer Private Reserve Cabernet Sauvignon
region
country
sampling time
extracted carbon (μg)a
F14C
error
conformity
Burgundy Burgundy Bordeaux Bordeaux Bordeaux Bordeaux Barolo Bordeaux Bordeaux Bordeaux Sonoma Valley Napa Valley Napa Valley Napa Valley Napa Valley
France France France France France France Italy France France France USA USA USA USA USA
1h 30 min 1h 30 min 1h 1h 20 min 2h 1h 1h 2h 30 min 2h 2h 2h
348 42 75 >1230 40 31 >202 19 2 475 178 907 54 83 138
0.9695 0.8507 1.2232 1.3944 1.4751 1.6431 1.5548 1.3866 5.2437 1.5411 1.3889 1.3839 1.3541 1.3473 1.3344
0.0024 0.0097 0.0076 0.0026 0.0180 0.0262 0.0038 0.0278 10.5677 0.0040 0.0029 0.0024 0.0089 0.0024 0.0085
yesb noc yes yes no yes yes no ambiguous yes yes yes yes yes yes
Napa Valley
USA
2h
23
1.3337
0.0209
yes
Napa Valley Napa Valley Napa Valley Alexander Valley Napa Valley Napa Valley
USA USA USA USA
2 3 1 1
33 8 >1400 25
1.2950 1.1095 1.2601 1.2357
0.0135 0.2784 0.0022 0.0238
yes ambiguous yes yes
USA USA
2h 1h
56 194
1.2060 1.1879
0.0209 0.0031
yes yes
Mosel-SaarRuwer Napa Valley Napa Valley
Germany
1h
2
2.7147
5.0142
ambiguous
USA USA
30 min 2h
>808 25
1.19414 1.0548
0.00309 0.0136
yes no
Napa Valley Napa Valley Napa Valley
USA USA USA
1h 1h 2h
31 702 54
1.1336 1.1648 1.1243
0.0171 0.0030 0.0066
yes yes yes
Napa Valley
USA
1h
4
1.3591
0.2637
ambiguous
Napa Valley Bordeaux Napa Valley
USA France USA
50 min 1h 1h
6 23 15
0.7116 1.0904 1.0455
0.0788 0.0228 0.0364
no yes yes
h h h h
Amounts denoted with “>” are not precisely known because samples were cut down as they were too large to be combusted. bThe sample was too old for bomb calibration but is probably in agreement with the atmospheric signature at the time. cThe sample was too old for bomb calibration but the 14C value did not match the atmospheric signature at the time. a
between 1955 and 1990, as the progressive flattening of the bomb curve (Figure 1) makes individual year determination for recent vintages more challenging. As sample sizes are large enough for radiocarbon analyses, the extraction method also has the potential for stable isotope studies for geolocation and “fingerprinting” of wines,23 chemical analysis to detect spoilage markers, and noninvasive assessment of cork integrity or oxygen transmission rates (OTR)20 through pressure curves recorded over the cork. For stable isotope applications, potential isotopic fractionation due to diffusion needs to be examined in greater detail. An upper limit leakage through the cork during the sampling is estimated to be on the order of wine losses due to natural cork leakage in 6 days (Supporting Information). We therefore argue that the method is safe and does not harm the cork closure or the wine. For additional safety, corks can be bled up after the extraction procedure with inert, food-grade argon,
which will protect both cork and wine from anomalous intrusion of air. Improvements to the extraction line and sampling protocol should increase the reliability of the method and make it fully applicable in the fight to eradicate wine fraud. Concerning potential false negative results (a fake wine passing the vintage authentication), we note that deliberate “doping” of the cork with sufficient volatiles of appropriate 14C content to swamp the natural signal and circumvent the technique would most likely produce an anomalously high sample yield. Moreover, we found that previous wetting of the top of the cork (i.e., addition of volatiles) produces an unusual pressure vs time response during extraction that is easily detected (Figure S2).
■
DISCUSSION In the present study, it has been shown for the first time that vapors extracted from cork closures largely represent the bottled liquid. Authentication of vintage and a number of other C
DOI: 10.1021/acs.analchem.5b01998 Anal. Chem. XXXX, XXX, XXX−XXX
Technical Note
Analytical Chemistry analyses are therefore possible for wine and other liquids with cork based stoppers. Future work should focus on the improvement of the vacuum line and optimizing sampling strategies in order to improve the reliability of the method. Atmospheric CO2 or volatile organic carbons (VOCs) in the cork are likely contaminants that were not separated from the targeted ethanol in this study other than by pumping away an initial aliquot of extracted vapors prior to the actual sampling. Optimal pressures or timing for the start of the sampling and chromatographic separation of extracts may therefore hold the key for further improvements in the accuracy of the present method. However, low diffusion coefficients of a certain percentage of cork stoppers may always hinder collection of sufficient sample sizes and in these cases will deliver large measurement errors and thus ambiguous or false positive results. At the same time, it will be difficult for forgers to circumvent the dating method, as prior doping of the cork with ethanol of a certain age can likely be detected by an atypical pressure drop over the sampling period. Finally, a portable sampling device using a molecular sieve trap in conjunction with a thermoelectric cooler could render liquid nitrogen unnecessary and would allow the method to be applied to a broad range of wines that cannot be transported to a laboratory for sampling.
Figure 3. Three-dimensional illustration (left) and cross section (right) of the wine sampler.
sampler, and a vacuum was then applied to the bottleneck and the cork. After pumping away the first fraction of gas, the cryo trap (device 5 in Figure 2) was cooled down with liquid N2 and wine vapors were collected for between 20 min and 3 h depending on the pressure measured over the bottle (device 4, Figure 2). The cryo-trapped vapors were then transferred cryogenically to a prebaked quartz tube containing 60−70 mg of cupric oxide. The tube with the frozen vapors was sealed off with an oxyacetylene torch while under vacuum and baked at 900 °C for 3 h in a muffle furnace. After cooling, the resulting CO2 was released into a vacuum line and reduced to graphite with hydrogen over an iron catalyst.34 The graphitized sample was pressed into a sample holder and was measured by accelerator mass spectrometry to determine the isotopic ratios for 14C/12C and 13C/12C. Radiocarbon results were normalized and corrected for isotopic fractionation according to Stuiver and Polach41 and converted to “fraction Modern” units (F14C) according to Reimer et al.42 Both sampling and graphitization backgrounds were investigated by measuring small aliquots of reference and radiocarbon blank materials. More details concerning the experimental setup and procedures are provided in Supporting Information, Materials and Methods.
■
MATERIALS AND METHODS The vacuum line used is illustrated schematically in Figure 2. Pressures of 1 × 10−4 Torr or better were achieved over the
■
ASSOCIATED CONTENT
S Supporting Information *
Figure 2. Wine vacuum line as used for the proof of concept study. The schematic diagram shows the following: (1) wine bottle, (2) wine sampler, (3a) venting valve of the wine sampler, (3b, 3c) shut-off valves of the vacuum line, (4) Pirani pressure gauge with automated data acquisition, (5) cryo-trap for sample collection, cooled with liquid nitrogen, (6) 8 in. × 1/4 in. quartz tube with 60−70 mg of CuO for sample combustion, (7) Pirani pressure gauge, (8) turbo molecular pump, and (9) diaphragm roughing pump.
The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b01998. Wine sampler and vacuum line, sampling procedure, sample preparation and radiocarbon measurements, method validation, determination of radiocarbon blanks, calibration of radiocarbon data, detecting ways of circumventing the wine sampling method, and wine extracted from the cork vs the bottle. (PDF)
wine sampler (Figure 3) as well as near the pump. All tubing and valves were either 1/4 in. or 1/2 in. outer diameter in order to allow sufficient pumping over the wine samples and thus permit fast trapping of low vapor pressure components such as water and ethanol. The main component of the line is the sampler (Figure 3), which consists of a hollow stainless steel cylinder and an O-ring which makes a vacuum tight seal around the neck of the wine bottle. Its design is crucial to minimize dead volume and, more importantly, to guarantee small leak rates, which affect both yield and the measurement accuracy of the subsequent analysis. Samples were taken as described in the following: The neck of the wine bottle was wiped free of dust and fitted into the
■
AUTHOR INFORMATION
Corresponding Author
*Tel +41 44 633 77 01. Fax: +41 44 633 10 67. E-mail: fahrni@ phys.ethz.ch. Present Addresses
‡ S.M.F.: Laboratory of Ion Beam Physics, Otto-Stern-Weg 5, ETH Zurich, 8093 Zurich, Switzerland. § B.T.F.: University of Chinese Academy of Sciences, Department of Scientific History and Archaeometry, No19A Yuquan Road, Beijing, 100049 China.
D
DOI: 10.1021/acs.analchem.5b01998 Anal. Chem. XXXX, XXX, XXX−XXX
Technical Note
Analytical Chemistry Author Contributions
(23) Pasteur, L. In Etudes sur le vin, ses maladies, causes qui les provoquent, procédés nouveaux pour le conserver et pour le vieillir; Imprimerie Impériale: Paris, France, 1866; pp 83−91. (24) Karbowiak, T.; Mansfield, A. K.; Barrera-Garcia, V. D.; Chassagne, D. Food Chem. 2010, 122, 1089−1094. (25) Lequin, S.; Chassagne, D.; Karbowiak, T.; Simon, J. M.; Paulin, C.; Bellat, J. P. J. Agric. Food Chem. 2012, 60, 3348−3356. (26) Lopes, P.; Saucier, C.; Teissedre, P. L.; Glories, Y. J. Agric. Food Chem. 2007, 55, 5167−5170. (27) Faria, D. P.; Fonseca, A. L.; Pereira, H.; Teodoro, O. M. N. D. J. Agric. Food Chem. 2011, 59, 3590−3597. (28) Rossmann, A. Food Rev. Int. 2001, 17 (3), 347−381. (29) Hua, Q.; Barbetti, M.; Rakowski, A. Radiocarbon 2013, 55 (4), 2059−2072. (30) L’Orange, R.; Zimen, K. E. Naturwissenschaften 1968, 55 (1), 35−36. (31) Lopes, J. S.; Pinto, R. E.; Almendra, M. E. Agron. Lusit. 1975, 36 (3), 223−234. (32) Burchuladze, A. A.; Pagava, S. V.; Povinec, P.; Togonidze, G. I.; Usacev, S. Nature 1980, 287, 320−322. (33) Martin, G. J.; Thibault, J.; Bertrand, M. J. Radiocarbon 1995, 37, 943−954. (34) Santos, G. M.; Moore, R. B.; Southon, J. R.; Griffin, S.; Hinger, E.; Zhang, D. Radiocarbon 2007, 49 (2), 255−269. (35) Levin, I.; Kromer, B. Radiocarbon 2004, 46 (3), 1261−1272. (36) Levin, I.; Kromer, B.; Hammer, S. Tellus, Ser. B 2013, 65, 20092. (37) Reimer, P.; Bard, E.; Bayliss, A.; Beck, J.; Blackwell, P.; Bronk Ramsey, C.; Buck, C.; Cheng, H.; Edwards, R.; Friedrich, M.; Grootes, P.; Guilderson, T.; Haflidason, H.; Hajdas, I.; Hatté, C.; Heaton, T.; Hoffmann, D.; Hogg, A.; Hughen, K.; Kaiser, K.; Kromer, B.; Manning, S.; Niu, M.; Reimer, R.; Richards, D.; Scott, E.; Southon, J.; Staff, R.; Turney, C.; Van Der Plicht, J. Radiocarbon 2013, 55 (4), 1869−1887. (38) Reimer, P. J.; Reimer, R. W. CALIBomb software; http://calib. qub.ac.uk/CALIBomb/; accessed on April 2, 2015. (39) Fuller, J. personal communication, April 13, 2015. (40) Jackson, R. S. In Wine Science; Academic Press: San Diego, 2014; pp 535−676. (41) Stuiver, M.; Polach, H. A. Radiocarbon 1977, 19 (3), 355−363. (42) Reimer, P. J.; Brown, T. A.; Reimer, R. W. Radiocarbon 2004, 46, 1299−1304.
The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS The authors thank James Fuller for his generous donation of wines for the development and testing of the presented method. Furthermore, the authors thank Lee Moritz and his team for manufacturing several prototypes of the wine sampler tool.
■
REFERENCES
(1) Holmberg, L. Int. J. Wine Res. 2010, 2, 105−113. (2) Frank, M. Wine Spectator, 2006; http://www.winespectator.com/ webfeature/show/id/Right-Bottle-Wrong-Wine_3325; accessed April 9, 2015. (3) The Economist, 2011; http://www.economist.com/node/ 18836894; accessed April 9, 2015. (4) Médina, B.; Salagoïty, M. H.; Guyon, F.; Gaye, J.; Hubert, P.; Guillaume, F. In New Analytical Approaches for Verifying the Origin of Food; Woodhead Publishing Limited: Cambridge, UK, 2013; pp 149− 188. (5) Hubert, P. Nucl. Instrum. Methods Phys. Res., Sect. A 2007, 580, 751−755. (6) Hubert, P.; Perrot, F.; Gaye, J.; Médina, B.; Pravikoff, M. S. C. R. Phys. 2009, 10, 622−629. (7) Bossew, P.; Ditto, M.; Falkner, T.; Henrich, E.; Kienzl, K.; Rappelsberger, U. J. Environ. Radioact. 2001, 55, 187−194. (8) DeYoung, P. A.; Hall, C. C.; Mears, P. J.; Padilla, D. J.; Sampson, R.; Peaslee, G. F. J. Forensic Sci. 2011, 56 (2), 366−371. (9) Moore, M. The Telegraph; 2011; http://www.telegraph.co.uk/ foodanddrink/wine/8246212/Empty-wine-bottles-sell-for-300-inChina.html; accessed April 9, 2015. (10) Pierson, D. Los Angeles Times; 2012; http://articles.latimes. com/2012/jan/14/business/la-fi-china-counterfeit-wine-20120115; accessed April 9, 2015. (11) Augustine, M. P.; Harley, S. J.; Lim, V.; Stucky, P. US Patent application publication No.: 2011/0184681 A1, 2010. (12) Harley, S. J.; Lim, V.; Stucky, P. A.; Augustine, M. P. Talanta 2011, 85, 2437−2444. (13) Harley, S. J.; Lim, V.; Augustine, M. P. Talanta 2012, 89, 484− 489. (14) Cozzolino, D.; Kwiatkowski, M. J.; Waters, E. J.; Gishen, M. Anal. Bioanal. Chem. 2007, 387, 2289−2295. (15) Liu, L.; Cozzolino, D.; Cynkar, W. U.; Dambergs, R. G.; Janik, L.; O’Neill, B. K.; Colby, C. B.; Gishen, M. Food Chem. 2008, 106, 781−786. (16) Jeffress Engineering Pty Ltd, BevScan; http://www.jeffress.com. au/products_bs01.htm; accessed April 9, 2015. (17) Weekley, A. J.; Bruins, P.; Augustine, M. P. Am. J. Enol. Vitic. 2002, 53 (4), 318−321. (18) Weekley, A. J.; Bruins, P.; Sisto, M.; Augustine, M. P. J. Magn. Reson. 2003, 161, 91−98. (19) Lim, V.; Harley, S. J.; Augustine, M. P. Am. J. Enol. Vitic. 2011, 62 (3), 291−297. (20) Buser, H. R.; Zanier, C.; Tanner, H. J. Agric. Food Chem. 1982, 30 (2), 359−362. (21) Karbowiak, T.; Gougeon, R. D.; Alinc, J. B.; Brachais, L.; Debeaufort, F.; Voilley, A.; Chassagne, D. Crit. Rev. Food Sci. Nutr. 2009, 50 (1), 20−52. (22) Lopes, P.; Saucier, C.; Teissedre, P. L.; Glories, Y. J. Agric. Food Chem. 2006, 54, 6741−6746. E
DOI: 10.1021/acs.analchem.5b01998 Anal. Chem. XXXX, XXX, XXX−XXX
