Authenticating icewine: closer, if not quite close enough

Scenario #1 – You’re sitting next to your fire after dinner, relaxed, with a few ounces of fine Canadian or German icewine, maybe a few slices of blue cheese and a ripe comice pear, and the current evening reading book. You enjoy all three for an hour or so and retire, happy and sleepy, to bed.

Scenario #2 – You’re sitting next to your fire after dinner with a few ounces of icewine and an active mind in search of a target, maybe two active minds if you have a companion. Conversation turns to the wine, how desperate those first Germans must have been to salvage their inadvertently frozen grapes and how arduous and expensive repeating the process on purpose now is. You speculate that cutting real icewine with something else must be mighty tempting, and the gaze you cast on your glass turns wary. And then you cast your gaze on Google and find this new article in the American Journal of Enology and Viticulture on a new strategy for testing the authenticity of icewine.

Icewine production is very expensive and no International Body of Icewine Authenticators polices producers to ensure that they’re doing it right or in good faith. Canada produces the bulk of the world’s stock (though I also enjoyed some fine examples in the Finger Lakes, not too far south of Ontario), and the Canadian Vintner’s Quality Alliance (VQA) legislates use of the term: a Canadian bottle with “icewine” or “ice wine” on the label must be made from approved varieties, from grapes harvested during “sustained” temps of at least -8°C, naturally frozen on the vine, coming in at at least 35°Brix, with no added sugar or alcohol, all overseen by a VQA representative. European producers employ similar standards, but the Asian sweet wine market is apparently well-populated with “Iced wine” and other unauthorized and fraudulent variations on the theme. Having a reliable means to verify that an “icewine” is really icewine made from frozen grapes seems prudent.

Per Armin Hermann’s new research, tracking oxygen isotopes could be that way. The idea is clever and conceptually simple. When grapes freeze, water partitions unequally between the part that turns to ice and the part that remains liquid. That’s the point of icewine: more water freezes, leaving sugars and other dissolved molecules concentrated in the syrupy liquid that remains. The naturally occurring isotope 18O, present in the water, will also distribute into the frozen and the unfrozen parts unequally. Since the frozen ice is more or less excluded from what ends up in a bottle of true icewine, then, icewines will contain a characteristic amount of 18O. All we need to do is determine — theoretically, using mathematical equations, and empirically, by measuring a bunch of icewines — what the “icewine” versus the “not icewine” 18O ranges are. Simple, elegant, and probably effective.

The plots of 18O measurements Hermann created show what looks like reasonably convincing separation between the ice- and non-icewine samples (understanding that judging how convincing is outside my expertise). BUT, there are two important caveats. First, the comparison was lab-frozen grape musts against the unfrozen originals. Again, it’s simple: “Frozen grapes, when pressed, will produce a must that is always depleted in 18O relative to its marc and also to their unfrozen counterparts.” The study didn’t include creating a database of icewine samples from various regions to establish reasonable 18O ranges. That’s solvable in theory, though the success of the whole method still depends on finding good, clear separation between real live ice and non-icewines.

Second, the method provides no way of determining how the wine was frozen. The 18O-depleted wine could have just as easily been frozen after harvest, in the winery, illegally. So, no matter how successful that empirical database is, the method won’t perfectly solve the how-do-we-detect-fakery problem. It is, as Hermann notes, an “additional” means giving a “strong indication” of authenticity. I wonder: is there a detectable chemical difference between the kind of slow freezing that would happen naturally on a grapevine in a cold Ontario winter and fast winery cryofreezing? Until then, looking for the Canadian VQA mark on the bottle — and avoiding anything labeled “iced wine” — remains the safest option, North American privilege notwithstanding.