A quick, complete, and permanent solution to smoke taint
Since I first saw bushfires in Australia in the early 2000s, I've been working personally on a solution which is quick, complete, and permanent with no collateral impact on the wine. My R&D team at Mavrik has perfected that.
You are probably thinking about smoke taint all wrong
As the list of compounds analyzed for expands, one thing remains clear: nobody knows what compounds actually impart the flavors, aromas, and mouthfeel. In addition, different fires cause widely varying smoky characteristics--but with little to no correlation to the mix of volatile phenols on resulting analyses. This hasn’t stopped speculation over the years about what matters, but, fortunately, a handful of recent research at the very least should put an end to speculation.
The earliest analyses conveniently chose analytes for which testing procedures already existed. Fortunately, those correlate quite well with perceived smoky character. Unfortunately, these are merely markers, and not the actual cause(s) of smoky aromas and flavors. Each and every one of the metabolites are also derived through oak contact. Obviously, aging wine on oak does not result in smoke taint-like flavors. Clearly there are one or more unknown things that actually influence smoky flavors, aroma, and mouthfeel.
So a bunch of new compounds were later added to analyses. Yet, those are still derived from oak (and/or from microbes such as Brettanomyces and Dekkera). Are those the cause of smoky aromas and flavors? Disappointingly, no.
Kennison et al (see endnotes) determined that the sensory threshold for these analytes in wine range primarily from 50 to 500 ppb--way beyond the levels caused by smoke. Only the ortho and meta forms of cresol have lower thresholds (13-19ppb) which are occasionally crossed by smoke tainted wine.
An early theory was “synergy” among these components “caused” smoke taint. But if so, toasted barrels tainted with Brett would give us smoke tainted wines. In reality “synergy” was just a shrug of the shoulders.
But we like to measure things and know things. So then even more things were added to lab analyses--“bound precursors” of the already tested for oak-derived volatile phenols were added to the panels. But these were just a miniscule amount of the same oak-derived compounds, bound to disaccharides. The hypothesis was that these tiny amounts (in dry wine) of bound precursors will continue to be released, through acid hydrolysis, over time. Or by an enzyme (amylase) in saliva.
So, bound precursors cause smoke taint to return?
No.
Ristic et al. (see endnotes) demonstrated that, as was long suspected, during bottle aging, smoke tainted wines release a little bit of the glycosidically bound volatile phenols and that the smoky flavors and aromas of the wine worsened. However, during bottle aging, the control wines--not exposed to smoke--also released the same (and same amount of) volatile phenols previously bound to disaccharides just like the smoke tainted wines did. Nonetheless, the controls did not develop smoky aromas or flavors. The worsening smoke flavors and aromas observed in the smoke tainted wines was caused by a decline in fruity flavors and aromas during bottle aging, not acid hydrolysis of glycosidically bound precursors.
Still not convinced?
Szeto et al. (see endnotes) used misters in the fruit zone during smoke exposure. This resulted in a significant reduction in the volatile phenols in the subsequent wine. This was in comparison to wine made from grapes exposed to the same smoke, but without fruit zone misting. However, the sensory evaluation of the two wines were indistinguishable. Reducing the grapes’ uptake of volatile phenols (the ones that are tested for in smoke taint panels) did not reduce smoky flavors and aromas.
But could amylase in saliva break down the glycoside bonds and release those flavors in the mouth? Amylase breaks starches down into disaccharides. The disaccharides for bound volatile phenols are rutinose and gentiobiose. However, amylase does not break down disaccharides nor free them from their aglycones. Aglycones are what the volatile phenols are called when bound to a disaccharide. Even if there were starch in wine, salivary amylase only breaks down about 30% of the starch ingested. And, finally, salivary amylase is de-activated by an acidic pH like that found in wine. So, no.
And then there is the peculiar case of Syrah.
Non-smoke exposed Syrah grapes have (glycosidically bound) levels of these volatile phenols equal to, or often greater, than those that show in smoke tainted wines. The same is true for wine made from these grapes, including the forms bound to the scant remaining disaccharides in dry wine. And, although Syrah descriptors sometimes include things like “flinty” and “bacon” they never taste remotely like smoke-tainted wines.
The urge to know and thus control smoke taint has led some over the years to hypothesize that things like guaiacol and, subsequently, cresols or syringol, and/or their glycosylated forms might be responsible for smoky flavors and aromas. But, alas, this has been proven not to remotely be the case.
The takeaway from this is that the markers are only markers--no matter how many you test for. They are not the cause of smoke flavors and aromas. Their analysis is not useful, beyond confirming that a) the grapes have indeed taken up smoke flavors and b) giving a general idea of how much of those flavors and aromas were taken up. Although volatile phenols may (or may not) play a small role in smoky aromas and flavors, they are, at best, minor players. The major player(s) remain unknown. Detailed analyses of nearly a dozen analytes, and their bound “precursors” is fruitless. It tells you nothing about the nature of the problem nor how to deal with it, that you do not learn from a simple analysis of guaiacol and 4-methylguaiacol.
You are going to have to rely on your palate when it comes to smoke. Lab analyses do not lead directly to a resolution like they do with other things, like VA for example.
As a final thought, consider this. Our targeted membrane separation and smoky flavor/aroma removal results in a dramatic change. One of the early clients of the new approach remarked: “There is a dramatic improvement on the finish of this wine. Once all that junk was removed, the finish became fruity again. I’m very pleased." Yet, when you run a smoke taint panel after treatment, the analytes show little change. And this is a good thing--you pay good money for oak to impart those things to the wine. The last thing you want is some indiscriminate process removing those expensive flavors and more. You want us to remove the smoky flavors, but not the desirable ones.
Selective or Permissive Membranes: How to get rid of it?
It is a simple choice: allow only smoke taint to be removed, which is slow but effective OR allow desirable flavors to be removed at the same time as smoke flavors, which is fast and effective.
Across the three decades we have offered smoke taint removal, each time we have tested permissive membranes we have been confronted with a grim reality: although they remove smoky flavors and aromas a little faster, they rapidly strip a lot of desirable things out of the wine too. Things like oak flavors and aromas, fruity esters, even tannin and anthocyanins in some iterations.
So, Selective Membranes, and Then What?
Using selective membranes quickly reveals the wine's native fruit. Just after that, the smoky flavors and aromas disappear. But then selective membranes run into a time issue: It takes a while to eliminate the harsh mouthfeel once and for all.
So, is this last step worth it? If you want your solution to be complete and permanent, it used to be your only option. But that is no longer the only option. Non-traditional fining agents have created a shortcut to solving the palate issues.
Non-traditional fining
Traditional fining agents—things like animal proteins, synthetic polymers, clays and carbons—are all linear and non-discriminating. By linear, I mean the more you add of them, the more flavors they will remove from your wine. Non-discriminating simply means that it removes a wide spectrum of things—including desirable things—from wine. You are familiar with this.
Non-traditional fining is not linear—it has a limit, beyond which adding more doesn't do much. It can also have narrower affinities. Some non-traditional fining agents don't even seem to have much effect on wines which are not smoke tainted. We have identified two which work really well to finish off the mouthfeel once and for all—without removing important wine flavors that traditional fining agents remove. These are very effective when used after our membrane treatment. If used before our membrane treatment, they will typically not show dramatic results.
What are the alternatives?
Aside from permissive and selective membranes, what are the other options?
Some people have claimed success with fining and adjuncts on very low levels of smoke taint without too much negative impact on flavors and aromas.
An alternative commercial approach involves vacuum stripping the wine's aromas off and then subjecting the entire wine to an indiscriminate adsorption media before reintroducing the aromas back to the wine.
Similarly, there is an approach of pumping whole wine directly through indiscriminate polymeric fining agents (“resins” that are made of plastics like PVPP and PET).
These two broad spectrum approaches, in addition to being very damaging to wine are very expensive as well. The alternatives amount to some form of a scorched earth approach (pardon the metaphor).
Frequently Asked Questions
Does the process require high pressures like other RO based processes?
No. All of Mavrik's processes are low pressure—well below the point at which volatile esters degrade. In most cases that is 40-60% lower than alternatives.
Can I trial it?
Sure. We are currently doing small lot trials as small as one barrel. We can process smaller vessels so long as the opening is large enough--contact us for details.
How much does it cost?
Treatment costs vary by lot volume, lot quantity, and level of taint. Most wines will cost between 40 and 80 cents per gallon to treat with our selective membranes.
What will I get for my money?
The equivalent of 2-4 passes of your wine through our selective membranes. This will remove the smoky aromas and flavors and reveal the wine’s fruit. Fining with non-traditional fining agents may subsequently be necessary to remove the last elements of smoke mouthfeel.
When does smoke taint show up?
It can be tasted at various stages. But around 6 weeks post-ML flavors and analytes stabilize.
What should I test for?
We can gauge the need for processing off of the two simple markers guaiacol and 4-methyl-guaiacol. Nothing else is informative for the remediation process. Testing after treatment is also uninformative, because the markers are derived from oak as well as smoke, and even occur naturally in wine. We purposefully avoid removing oak related flavors.
Do I need a test after treatment to tell me whether it worked or not?
No. You just need your palate. The markers being tested for are NOT smoke taint. Analysis afterwards will only tell you if a more permissive or invasive process did as much damage to desirable flavors/aromas as your palate tells you it did.
What about bound precursors?
There is no benefit from testing for these things. See endnote.
What about enzyme treatments?
Lab analysis for smoke taint markers before and after treatment with various enzymes show no change to free or bound markers. There is no benefit from their use. Even if they did release glycosidically bound precursors, they will not cause smoke flavors and aromas. See footnote.
What about oak?
No problem! Just tell us what kind of oak exposure (barrel, stave, chip, new/used/neutral) and how long before the analyzed samples were taken. We can approximate the influence of oak and separate that from your analysis. And then we will leave your hard earned oak influence alone while removing the smoke taint.
When should I treat my wine?
The ideal time to treat your wine is as late as is practical. Ideally, you do it after un-barreling and before making blending decisions. But if you need to do so earlier, there is no real detriment. The ideal time to get analysis is 6 weeks after ML and before barreling down. Remember to keep detailed records of oak contact.
Should I be worried about the smoke taint returning?
Not if you follow these guidelines. The membrane process followed by non-traditional fining removes the smoky flavors and aromas as well as any remaining colloids capable of holding smoke flavors and aromas. This will allow a natural aging trajectory for your wine.
ENDNOTES
Kennison, K.L., Wilkinson, K.L., Williams, H.G., Smith J.H., & Gibberd, M.R. Smoke Derived Taint in Wine: Effect of Post Harvest Smoke Exposure of Grapes on the Chemical Composition and Sensory Characteristics of Wine. Journal of Agricultural and Food Chemistry 55, 10897-10901.
Established a definitive range of sensory thresholds for a full spectrum of smoke taint markers.
Ristic, R., van der Hulst, L., Capone, D.L., Wilkinson, K.L. Impact of Bottle Aging on Smoke-Tainted Wines from Different Grape Cultivars. J Agric Food Chem. 2017 May 24; 65(20):4146-4152.
Ristic et al found that glycosidically bound oak aromas did indeed hydrolyze slowly during bottling due to wine's weakly acidic conditions. This resulted in increasing levels of volatile phenols. However, bottle aging of non smoke-tainted wines showed the same acid hydrolysis of the same volatile phenols over time in the bottle. The non smoke-tainted bottles did not develop smoky aromas and flavors, but the smoke-tainted bottles did. Thus, acid hydrolysis of glycosylated precursors could not be the cause of smoky flavors and aromas. Rather, the cause was diminished fruit flavor and aroma.
Szeto, C., Rsitic, R., Capone, D., Puglisi, C., Pagay, V., Culbert, J., Jiang, W., Herderich, M., Tuke, J., & Wilkinson, K. Uptake and Glycosylation of Smoke-Derived Volatile Phenols by Cabernet Sauvignon Grapes and their subsequent fate during winemaking. Molecules. 2020 Aug; 25(16):3720.
Determined that misting fruit zone during exposure to smoke resulted in a reduction in volatile phenol uptake but did not result in a reduction of smoky flavors and aromas.