Combating Wine Fraud: A New Noninvasive Approach Shows Promise for Testing Authenticity of Wines Without Damage to the Bottle or Wine

Wine fraud is sadly not an uncommon occurrence in the industry, so much so that it has even caught the attention of the Federal Bureau of Investigation (FBI) in the United States.  With rare bottles reaching on average more than $5000, the incentive to create counterfeits is great.  Private collectors and auction houses alike have been implicated in wine fraud, which creates problems with supply and demand in the high-end wine market.  The first solution ever created to help combat wine fraud was to create special bottle labels that would distinguish high-end rare wines from the unauthentic wines.  However, just as a virus can adapt to its environment, the counterfeiters simply found a way around it by simply dumping out the contents of the authentic bottles and replacing it with a poorer quality wine.

The current approaches to analyzing and determining the authenticity of a wine bottle require breaking the cork and seal, which significantly devalue the wine in question, thus making it much less attractive to collectors and investors.  These invasive techniques include ultra-violet/visible infrared spectroscopy, electron and nuclear spin resonance, mass spectrometry, and X-ray absorption.  These methods can often suffer from bottle-to-bottle variation due to glass composition, the requirement of large sample sizes in order to overcome background “noise” issues, or requiring the use of large, bulky pieces of equipment that are not easily moved from place to place.

The study presented here today, which was recently published in the journal Analytica Chimica Acta by chemists at the University of California at Davis, aimed to describe a new method of identifying fraudulent wine bottles, one of which is minimally invasive and easily portable, compared to the current, sometimes unreliable methods.  To be more specific, the method described is based on the dielectric absorption of wine at low (  In theory, the frequency absorption by the various constituents in wine leads to a difference between an electric field that is applied to the wine and the resulting dielectric displacement of the wine.  Basically, what ends up happening is that once the electric field is applied, there is movement of atoms and other molecular ions, a change in orientation of neutral polar molecules, and the polarization of neutral non-polar molecules.  All of these changes are a function of the composition of the wine, which result in a unique “profile” for each wine tested.

The molecular composition of wine varies dramatically due to location, vintage, other local growing conditions, and winemaking techniques, all of which are unique on the molecular level for each and every wine.  Metal ions naturally accumulate in the grapes throughout the growing season, and will be different depending upon the exact vineyard location.  Pollution, sea spray, fertilizers, and pesticides can all introduce metal ions in the grapes as well, which can add to the uniqueness of the molecular signature of the grape and subsequent wine.  Specific vinification techniques can also change the molecular composition of the finished wine, including the particular storage/fermentation vessels, the use of flocculants, pH buffering, alcohol fermentation, malolactic fermentation, the use of glycerol, and the removal of potassium bitartarate from American versus European wines.  The presence of a large variety of molecules, ions, and compounds that vary depending upon growing conditions and vinification techniques change the dielectric response when undergoing analysis using the methods of the authors of this study result in a specific wine signature, or wine fingerprint, which may be used to determine the authenticity of any bottle of wine.

Harley et al, 2011doi: 10.1016/j.aca.2011.06.048

Basically (and when I say basic, I mean really basic), this method works by passing a dielectric frequency across a bottle of wine via a copper electrode, the dielectric response and absorption of which are monitored by another copper electrode at the bottom of the bottle and measured by a computer.  Figure 1 by the authors of this study illustrates the device used to measure this dielectric absorption of intact wine bottles.  The data measured and collected by the computer are then subject to PCA analysis, the results of which provide a unique fingerprint for each wine analyzed.  By knowing the unique fingerprint of authentic wines, and by comparing them to a potential counterfeit wines, it can be determined with accuracy whether or not the wines are authentic or an imposter.


Due to the intense mathematical, chemical, and physical science presented in the methods of this paper, I will be condensing them and only discussing the most pertinent pieces of information.  Should you require more details, please feel free to ask.  Dielectric measurements were calculated using a suite of equations, of which I’ll skip for now.

Before analysis of specific wines could be done, a set of standards had to be completed, to use as a reference for all other unknown samples.  To create these standards, 16 bottles of Charles Shaw (a.k.a. “Two Buck Chuck”) from Trader Joe’s were drained, rinsed multiple times, and refilled with various solutions of de-ionized water, ethyl alcohol, acetic acid, potassium sulfate, glycerol, and sucrose.  One bottle contained just de-ionized water, to serve as the zero standard, and all other bottles were created using different amounts of the aforementioned components, the means of which are all reported in existing wines.

The high-end wines used for determining their specific fingerprints and establishing a wine fingerprint library were two bottles each of the 1990 Grand Vin Chateau Latour, the 1958 Grand Vin Chateau Latour, the 2002 Opus One, the 1985 Opus One, the 1994 Petrus Pomerol, and the 1979 Petrus Pomerol.


The results of this study showed that for every wine in the library, there was a unique dielectric fingerprint that could potentially be used to compare against when testing potential counterfeit bottles.

Harley et al, 2011doi: 10.1016/j.aca.2011.06.048

Looking at Figure 4 from the paper described today, one can see the results of the study in action.  The closed and open squares represent the fingerprints of the 1985 and 2002 Opus One wines; the closed and open diamonds (“lozenges” in the figure heading) represent the fingerprints of the 1958 and 1990 Grand Vin Chateau Latour wines; the closed the open triangles represent the fingerprints of the 1979 and 1994 Petrus Pomerol.  The ‘x’ represents a randomly selected bottle measured on a different day, in order to determine the consistency of the procedure.  That bottle was the 1979 Petrus Pomerol, which due to its’ extreme proximity to the Petrus Pomerol signature acquired on an earlier day, indicates that the procedure is consistent and reliable for identifying the authenticity of wine.

What is required in order to use this procedure for determining the authenticity of wine?  Well, first and foremost, a library of known wines must be available, and the particular wine in question must have had an authentic bottle analyzed previously in order to have something to compare it against.  Most other methods of analyzing authenticity are much the same (i.e. a sommelier must have actually tasted the wine in the past to know if what they are tasting is in fact the same wine), so this is not seen as a drawback.

One of the results of this research was that this method appears to be temperature dependent.  For example, if the library wine was tested at room temperature, then the unknown wines must also be sampled at room temperature.  It must be noted in the library what temperature the wines were measured at, in order to determine the proper temperature of which to measure the unknown wines in question.

The next step from this research is to go ahead and create an extensive library with the fingerprints of as many authentic high-end wines as possible.  Since this method does not damage the wine bottle or the wine itself, this should be relatively straight forward, provided collectors are willing to loan their bottles to scientists.

In conclusion, the results of this study show that a new noninvasive, nondestructive technique for determining the authenticity of wine bottles has been established.  Provided the fingerprints are available for the authentic versions of the wine in question, this new method may prove to be the superior choice to analyzing potential counterfeit wines, thereby decreasing the frequency of wine fraud across the globe.

I’d love to hear what you all think about this topic!  Do you have any personal experience with wine fraud?  Feel free to leave any comments below!

Source:  doi: 10.1016/j.aca.2011.06.048
I am not a health professional, nor do I pretend to be. Please consult your doctor before altering your alcohol consumption habits. Do not consume alcohol if you are under the age of 21. Do not drink and drive. Enjoy responsibly!

7 comments for “Combating Wine Fraud: A New Noninvasive Approach Shows Promise for Testing Authenticity of Wines Without Damage to the Bottle or Wine

  1. September 29, 2011 at 3:20 pm

    The real question is, wouldn't "de-ionized water, ethyl alcohol, acetic acid, potassium sulfate, glycerol, and sucrose" be a few notches up from Two Buck Chuck?

  2. September 29, 2011 at 10:57 pm

    While that may be true, at least doing it their way ensures consistency 🙂

    I'm also curious about the "draining" method of the Two Buck Chuck. I wonder if it happened to coincide with any grad student socials…

  3. WineKnurd
    September 30, 2011 at 12:09 pm

    Forget the 2BC when you have all these other great wines around! I wonder if the adviser at Davis had any bottles of the Petrus missing 🙂

    The hard part is going to be finding "authentic" standards for this testing. If collectors have fakes already, then this test will do exactly what it was intended not to do, and that is to promote the fraud!

    There is also the possibility that some of the the older bottles have been topped off with newer wines at the winery, a practice done by most of the greats like Grange and Petrus, which might complicate things.

    But I think if we start at a realistic time point, like maybe the 90's where some wineries still have bottles they can confirm as authentic, then this will be a great asset to combat fraud for mega-vintages such as 2000 and 2005 Bordeaux, especially in newly emerging markets like China.

    Great article Becca!

  4. September 30, 2011 at 3:33 pm

    The individual lending these rare bottles to the scientists sure did have some trust in them 😉

    You are right though, it'll still be difficult to know for sure which bottles are authentic to even start the library to begin with. Starting a library now, however, with newer wines is a great idea. For those wines bottled before 1945 should also be relatively simple to add to the library, due to the fact that anything bottled after 1945 will have molecular signatures indicative of nuclear fallout (which will remain in the air for a long time (long being an understatement). If a bottle from say 1940 shows up as containing these signatures indicative of nuclear fallout, then we know it's bogus since those signatures do not appear until 1945 after the atomic bombs were tested and dropped. Provided there wasn't a lot of counterfeiting before 1945, we should be somewhat confident in the results of those very old bottles.

    Great comments!

  5. WineKnurd
    September 30, 2011 at 11:09 pm

    The radioactive Cs test can only be done if you open the bottle 🙁

  6. October 1, 2011 at 1:59 am

    True, though since each element or compound has its own dielectric fingerprint (or at least the combo of differing elements and compounds result in a unique fingerprint), one should still be able to get a unique signature with a wine from 1940 versus wine from the same vineyard/winery in 1945. I suppose one wouldn't really be able to tell if Cs was in it for sure, but at least the signature would be different. It's still a pretty cool test anyway, even if it doesn't work exactly the same way with this method! 🙂

  7. WineKnurd
    October 1, 2011 at 12:54 pm

    Good point Becca! Keep and eye out for any other articles of this nature, would love to see if the results from a 10+ year vertical, say Bordeaux or Burgundy, and see what the overall trend might be, outliers for good/bad vintages, overlapping data points, etc. Makes me want to go back to grad school (except then I would only be able to afford 2 buck chuck…)

Comments are closed.