Mostly the flavour chemistry, not the nicotine — and cooling and burning are different receptors, which is why a pouch can be arctic and barely irritating at the same time. Two independent laboratories have measured the compounds responsible, by brand and by flavour, in micrograms per pouch. The alkali everyone blames is real and well measured, but no study has ever shown it causes the burn.
The problem with most answers to this question is that they name one cause. The measured picture is four causes with wildly different amounts of evidence behind them, and the one with the strongest evidence is the one nobody mentions. This page separates them and is explicit about which are mechanism rather than demonstration. If what you want is to tell the burn apart from the dose, that is strength vs burn; if you are mid-pouch and it is unpleasant right now, what to do if your first pouch is too strong.
The burn is at least three different sensations
Your mouth has separate detectors for cold, for irritation and for heat, and the flavour industry knows exactly which is which. From a National Institutes of Health-funded review of chemosensory additives: “The cooling properties of menthol are mediated through activation of TRP melastatin 8 (TRPM8) receptors expressed throughout the mouth, nose, throat, and lungs”, and TRPM8 “is the principal detector of environmental cold”. Irritation is a different channel, TRPA1. Heat and pain are a third, TRPV1.
Menthol hits more than one. The same review notes that “Menthol is ‘promiscuous’ in its actions with sensory channels, interacting not only with TRPM8 and TRPA1, but also with nAChRs, TRPV1, and TRPV3”. The synthetic coolants do not: “Menthol’s irritant effects are mediated by TRPA1, a sensory irritant receptor. Compared to menthol, synthetic cooling agents (WS-3, WS-23) are only weak activators of TRPA1, explaining their lack of irritant effects.” And: “These agents are also less irritating than menthol that causes burning and painful sensations at higher concentrations.”
That is the mechanistic answer to the headline question, and it is worth stating baldly. A pouch can be intensely cold and barely irritating, or minty and sharply irritating, at exactly the same nicotine strength, because the two sensations run on different receptors and the formulator chooses which one to load.
The compounds, measured, by flavour — and humans can tell
Jabba and colleagues at Yale and Duke bought ZYN pouches in six flavours at both 3 and 6 mg, ran the extracts through calcium microfluorimetry on human receptors, and quantified the flavour chemicals by gas chromatography–mass spectrometry. Average pouch mass in their sample was “405±18 mg (n=12)”. What they found:
| Pouch | Flavour chemicals measured |
|---|---|
| Cool Mint | Menthol 5406±259 µg per pouch — the only flavourant detected |
| Menthol | Menthol 1390 µg per pouch — the only flavourant detected |
| Peppermint | WS-3 201±11 µg; menthol 1,925±140 µg; menthone 791±55 µg; menthyl acetate 105±9 µg |
| Spearmint | WS-3 209±15 µg; carvone 1180±185 µg; menthol 470±44 µg |
| Chill | WS-3 234±7 µg — the only major flavourant and the only observable peak besides nicotine |
| Smooth | Nothing detectable by their method |
Look at the spread: within one brand, on one shelf, Cool Mint carries roughly four times the menthol of Menthol and about twenty-three times the mass of active compound that Chill carries. The receptor data then lines up as the pharmacology predicts — “The sensory irritant receptor, TRPA1 was robustly activated by mint-flavoured ONP extracts (Zyn Cool Mint and Zyn Peppermint), yet TRPA1 activation was significantly weaker for Zyn Chill extracts” while “Zyn Chill ONP extracts robustly activated TRPM8, with much higher efficacy (39%–53%) than the mint-flavoured ONPs.”
An independent group at Roswell Park replicated the Chill measurement — “WS-3 was also present in the implicit flavor ‘Chill’ from ZYN 3 and 6 mg, consistent with previous findings (0.3 ± 0.1 vs. reported 0.2 ± 0.0 mg/pouch)” — and found that “The most common flavorings in ONPs were menthol (cooling) (60.0%), menthone (26.0%), benzyl alcohol (20.0%), and methyl salicylate (16.0%).” Their reading of why is the sharpest sentence in the literature: “Like e-cigarettes and conventional cigarettes, menthol may provide a cooling sensation even among products without explicit mint/cooling claims, to mask the harshness of nicotine, especially among those products with a high proportion of freebase nicotine.” Both studies were federally funded, by NIDA and FDA’s Center for Tobacco Products and by the National Cancer Institute respectively; one Roswell Park author discloses a Pfizer grant and Johnson & Johnson advisory work outside the work.
Two studies put real people on the scale. A double-blind randomised trial of twenty-two young adults using four ZYN products found that “‘Chill’ and ‘Classic’ ONPs contained WS-3 exclusively (0.24±0.02 mg/pouch)” while “WS-3 was not detected in ‘Original’ and ‘Smooth’. Also, WS-23 and menthol were not detected in any of the analyzed ONPs.” The ratings followed: “Compared to ‘Smooth’, human participants rated ‘Chill’ and ‘Classic’ as having stronger cooling sensations (p<0.05), while ‘Classic’ was rated as more minty flavored (p<0.005).”
And in a separate study of twenty-three young adults given eight pouches from six brands, four mint and four fruit, with five minutes of standardised use each, fruit-flavoured products “caused less burning” than mint, and “while pouch versus gum increased burning and harshness more for mint.” Mint burns measurably more than fruit, with a confidence interval around it.
Caveat both honestly: twenty-two and twenty-three people, young adults, exploratory analyses with many uncorrected comparisons. The mint-versus-fruit contrast is the robust part; brand rankings from a dozen people are not. And one precisely labelled hole sits inside the first of those studies: its questionnaire included “How strong was the burning sensation in your mouth?” and the results table reports appeal, cooling, mintiness and tingling only. The burning data were collected and never published.
The alkali: well measured, and not shown to be the cause
Every pouch is deliberately alkaline, because alkalinity is what frees nicotine to cross your gum. Germany’s Federal Institute for Risk Assessment puts the mechanism in one sentence: “In its uncharged state, nicotine can pass through biomembranes such as the oral mucosa more easily, which leads to improved oral nicotine absorption.”
The measurements are good and the range is enormous. CDC’s survey of thirty-seven pouch brands found products “varied in pouch content mass, moisture content (1.12%‒47.2%), alkalinity (pH 6.86‒10.1), and % free nicotine (7.7%‒99.2%).” A later CDC study of seventy products: “Nicotine pouches exhibited the highest pH and widest range, with values ranging from 7.64 to 9.49, corresponding to freebase nicotine ranging from 30.8% to 96.7%”. BfR’s own analysis of forty-four pouches bought online: “The median pH for aqueous extracts was 8.8 for the pouches examined by the BfR, with only one product exhibiting an acidic pH,” and “The median proportion of free-base nicotine was 86%”.
Two things to notice before you treat that as the answer.
First, the constant everybody divides by is not agreed. CDC calculates free base “by substituting the pKa value for nicotine, 8.02 at room temperature (25°C), and the measured pH value into the Henderson-Hasselbalch equation and multiplying by 100.” Roswell Park uses 8.01. Small, but it means the published percentages are arithmetic from an assumed constant, not a measurement of free base.
Second and much more important: there is no such thing as a published in-mouth pH. Every figure above is the pH of an extract made under a laboratory protocol, and the protocols differ — CDC mixes roughly a gram with 10 mL of deionised water, Roswell Park uses a modified CORESTA method, and Philip Morris’s own patent specifies a phosphate buffer at pH 7.4 warmed to body temperature. Nobody has measured the pH under a real lip.
And the decisive gap: the one experiment that ever deliberately varied pH in a smokeless product measured the wrong thing. Forty moist-snuff users “completed four crossover visits and used a single 2 g portion of Copenhagen Original Long Cut amended to 4 pH levels: 5.0, 7.7, 8.2, and 8.6 (free-base nicotine 0.1, 32, 60, and 79%) for 30 minutes.” The result was a pharmacokinetic one — “Nicotine PK substantially depends on its free-base proportion, with more than 4-fold increases in mean plasma nicotine maximum concentration and area under the curve”. In the published record of that study the words “irritation”, “burn”, “sting” and “harsh” do not appear at all. A clean sweep from 0.1% to 79% free base, and nobody asked whether it hurt. The full-text is on a blocked host, so I can only scope that to the published abstract and record.
So the alkali row below is a mechanism, not a demonstration. The snus version of the same chemistry is snus pH; for pouches, pouch pH and why it changes the hit.
What the nicotine itself does
Nicotine is a genuine chemical irritant where it touches you, and the mechanism is better established than most people assume — but it is not the one usually named. From the NIH-funded review: “When applied to the tongue or nasal cavity, nicotine causes pain and discomfort, which is reduced by the application of nAChRs antagonists”, indicating that the effect is “predominately mediated by nAChRs located on trigeminal nerve endings”. Nicotine does activate TRPA1 in cultured neurons, “However, the involvement of TRPA1 in nicotine irritation is not clear-cut”, and the review’s own conclusion is blunt: “Together these studies emphasize nAChRs as the primary mediator of nicotine’s irritant effects and not TRPA1.” It closes with “Taken together, the precise role of TRPA1 in nicotine irritation is in need of more definition.”
In short: do not write that nicotine burns via TRPA1. It burns via the nicotinic receptors on your trigeminal nerve endings, and the review names pH as a determinant of how strongly it does so, which is the one place where the alkali story and the nicotine story actually join up in a primary source.
The regulator’s own trial — and a correction to the record
BfR, with the tobacco dependence clinic at LMU Munich, ran a five-arm crossover in fifteen people: pouches at nicotine-free, 6, 20 and 30 mg held for twenty minutes, plus each participant’s own cigarette. It is the most-cited piece of evidence that nicotine is not the whole story, and it is routinely misread — including, until today, on this site.
BfR’s English-language opinion says: “While the tobacco cigarette rarely caused irritation to the oral mucosa, all nicotine pouch users reported oral mucosa irritation ranging from moderate (0–20 mg nicotine) to severe (30 mg).”
The German original of the same document says something different. It reads “kam es bei allen Nikotinbeuteln zu mittelstarken (0 – 20 mg Nikotin) bis starken (30 mg) Mundschleimhautirritationen” — bei allen Nikotinbeuteln, with all nicotine pouches. It is a claim about the four products, not about the fifteen people. BfR’s own English translation turned a statement about products into a statement about participants, and that mistranslation is where “it irritated every user” entered the literature.
The peer-reviewed paper gives the clinical endpoint, which is the number to use: “The oral mucosa was inspected before and 10 minutes after pouch use for redness or ulceration. A slightly increased redness was revealed in 1, 4, 3, and 5 cases after use of the pouches with 0, 6, 20, and 30 mg nicotine, respectively. An ulceration was visible in one case after use of the 6 mg pouch.” One case in fifteen on the nicotine-free pouch; five in fifteen on the 30 mg; and the single ulceration on the weak product, not the strong one.
The authors’ own reading is the fair one: “This suggests a tendency that nicotine may contribute to local effects. However, as the pouches with no nicotine or low amounts also induced local adverse effects, other substances are involved.”
Two further things nobody says about this study. Its four pouch arms came from four different manufacturers — nicotine-free from Swedish Match, 6 mg from Imperial Brands, 20 mg from British American Tobacco, 30 mg from a Polish maker — so the 1/4/3/5 ladder is confounded with brand and formulation and is not the clean dose-response it is usually quoted as. And it measured no pH at all: searching the full body for “pH”, “alkali”, “menthol”, “coolant”, “carbonate” and “TRP” returns zero hits for every one. It attributes the residual irritation only to unnamed “other substances”, and it was funded by BfR’s own budget with no industry money.
Our page on why nicotine pouches tingle carries the older, overstated form of this finding, and it is being corrected.
The formulation differences you cannot see
Three major brands use three different nicotine raw materials, and they say so themselves. ZYN lists “NICOTINE SALT (Nicotine bitartrate dihydrate): Pharmaceutical-grade nicotine” and, separately, “pH ADJUSTERS (Sodium carbonate and sodium bicarbonate): Minerals to help adjust pH levels.” Velo describes a “tobacco-free” pharmaceutical-grade nicotine polymer. Rogue states that “Rogue Pouches contain quality nicotine bound to a polymer resin to create nicotine polacrilex.”
A resin-bound nicotine does not need the same alkali load as a salt, and Philip Morris’s patent makes the trade-off explicit. It claims a “pH regulating agent in an amount between 0.01 and 15% by weight of the pouch composition” narrowing to “between 5 and 10% by weight”, targeting “a pH of at least 8.0, such as a pH of at least 9.0” in a standard buffer test — and then: “Providing nicotine in the free base form allows facilitates obtaining a higher pH in the pouch composition, without using too much alkaline pH adjusting agent.”
That is the sharpest thing in the patent literature here. Two pouches measuring the same pH can carry very different amounts of carbonate — and whether the carbonate itself irritates has never been tested. “irritat” returns zero hits across all four pouch patents I read. Pouch patents do not discuss irritation.
ZYN discloses the sodium carbonate and the sodium bicarbonate by name. Its FAQ mentions pH seven times and never once with a number; every mention is a category label. Search the same page for “WS-3”, “WS-23”, “cooling”, “coolant”, “capsaicin”, “irritat”, “sting” and “burn” and you get zero hits for all eight — the ingredient list names a “flavorings” category and discloses no individual flavour chemical, which is the gap the 234 µg of WS-3 that two independent laboratories measured in Chill sits in. No brand publishes a pH figure for its own product anywhere I could find. Velo goes the other way and brands the sensation, telling you to “Feel the burn” and attributing it to the nicotine working — a marketing claim that BfR’s nicotine-free arm and the “Original” tingling result both cut against.
The fleece first, because it is what people blame. It is “a dry-laid bonded nonwoven comprising viscose rayon fibres (i.e. regenerated cellulose) and an acrylic polymer that acts as binder”, and the manufacturer’s own description beats any paraphrase: “The viscose nonwoven material normally used for smokeless tobacco pouches is similar to the fabric used in tea bags.” Its tensile strength is specified to a standard test method. Its abrasiveness against your mucosa is measured nowhere — “abrasion” returns zero hits across all four patents I read. More on it in pouch fleece materials.
Size and pressure next. The pressure mechanism is presented by the company that makes them as a belief, not a measurement: “It is believed that the comparatively narrow width and the specified length-to-width ratio of the oral pouched product as disclosed herein… may mitigate pressure on the gum/teeth interface upon use of the oral pouched product, thereby reducing the risk for uncomfortable rubbing and gum recession.” Format comparisons are in slim vs mini vs large.
Moisture is a real category difference — snus portions run around 45% water and most pouches are far drier — but two CDC datasets disagree about the pouch range by a factor of seven, so there is no single honest figure to quote, and whether a dry pouch draws water out of your mucosa has never been measured. Placement is weaker still: one cross-sectional study of forty-two users found consistent placement associated with lower lesion severity, odds ratio 0.19, and the association “did not reach statistical significance (p=0.057)” — underpowered, and reverse causation wide open, since people who already have a sore spot move the pouch. Both belong in the unknown column.
The physical causes, which are mostly unmeasured — and whether it settles down
And does it settle down? Within a session, probably, by a mechanism established in a different tissue: “In human subjects, menthol applied to the tongue elicits a self-desensitization effect, where a repeated presentation, even 60 mins later, induces far less irritation”, and “Menthol also reduces the irritating effects of cinnamaldehyde, a TRPA1 agonist”. Both are tongue-application and vaping studies. No study addresses desensitisation in pouch users, so between-session tolerance is not established for this product at all. What is established is that the visible marks go: lesions from Swedish smokeless tobacco “are reversible, that is, clinical and histological tissue changes return to normal following cessation. However, the exact mechanisms behind these changes are unknown.” The six-week study showing a reduction when users switched to tobacco-free pouches lists Swedish Match AB as its funding source while its authors “declare that they have no conflicts of interest” — a company-funded study of that company’s own product.
The table
| Cause | What is actually measured, and in what | Strength of evidence |
|---|---|---|
| Menthol | 5406±259 µg per pouch in ZYN Cool Mint, 1390 in ZYN Menthol; present in 60% of pouch brands; extracts robustly activate TRPA1; humans rate mint as burning more than fruit | Strongest. Chemistry, receptor assay and human rating all agree, from three independent non-industry laboratories |
| Synthetic coolants WS-3 / WS-23 | 234±7 µg in ZYN Chill, replicated at 0.3±0.1 mg and 0.24±0.02 mg by two other groups; strong TRPM8, weak TRPA1; rated significantly cooler than an unflavoured product in a double-blind trial | Strong — and this is the “cold but doesn’t burn” mechanism |
| Alkaline pH and free-base nicotine | pH 6.86–10.1 and 7.7–99.2% free nicotine across 37 brands; BfR median 8.8 and 86%. All in water or buffer extracts, under protocols that differ | Exposure well measured; the link to burn is not. The one pH-manipulation trial measured plasma nicotine and no irritation endpoint. No in-mouth pH exists |
| Nicotine at the nerve ending | Nicotinic antagonists reduce nicotine-induced oral pain; a TRPA1 blocker does not. BfR rated irritation moderate at 0, 6 and 20 mg, strong at 30, n=15 | Mechanism strong, dose-response weak. The 0 mg arm still irritated, so nicotine is a contributor, not the cause |
| The carbonates and the nicotine raw material | Carbonates disclosed by name, no level; a patent claims 0.01–15% by weight. Three brands use three nicotine chemistries | Presence documented, contribution never tested. No study varies either at matched pH. Inference only |
| The fleece, the size and the pressure | Nonwoven viscose or lyocell with an acrylic binder; 0.3–3.5 g; tensile strength specified to a standard | Materials precisely specified; abrasiveness against mucosa measured nowhere. The maker hedges the pressure mechanism itself |
What this means for picking a tin
If a pouch burned and you want one that doesn’t, the lever with the most evidence behind it is the flavour, not the strength: a product cooled by a synthetic coolant rather than by menthol has been shown — chemically, pharmacologically and in blinded human ratings — to deliver the cold without much of the irritation, and fruit burned measurably less than mint in the one head-to-head that rated it. Changing brand at the same milligram figure changes more than you would expect, because the nicotine chemistry and the alkali load change with it. What the evidence does not support is reading the burn as a dose signal: see why mint pouches feel stronger than they are. If you are choosing a strength rather than a flavour, today’s companion page is what strength should a heavy smoker start on, and choosing your first strength is the general version. The liquid that appears while all this is going on is the drip.
Common questions
Does a stronger pouch burn more? A little, and less than you would guess. In a double-blind within-subject study, higher nicotine strength produced higher harshness ratings — about seven points on a hundred-point scale. In BfR’s crossover, rated mouth irritation was the same at 0, 6 and 20 mg.
Is the burn a sign the pouch is working? One manufacturer says so in its marketing. The nicotine-free arm of a regulator’s own clinical study says otherwise, and so does the finding that a ZYN product containing neither menthol nor WS-3 still rated significantly higher on tingling than another product containing neither.
Is a pouch that burns damaging me? Lesions at the placement site are well documented in snus users and are reversible on stopping. Whether the burn itself predicts them has not been studied — a real gap, not a reassurance.
Which brands burn least? On measured chemistry rather than on opinion, the products with no detectable menthol and either no coolant or WS-3 only. I have not tested any product, so there is no ranking here from experience; the master ranking is built from the specifications and documents available. For what goes into a pouch in the first place, see what is inside a nicotine pouch, and for the brand that invented the coolant-only category, the Velo brand guide.
Sources
- Jabba SV, Erythropel HC, Woodrow JG et al., Synthetic cooling agent in oral nicotine pouch products marketed as “Flavour-Ban Approved”, Tobacco Control, 2023. Supported by NIDA (R56DA055996, U54DA036151) and FDA’s Center for Tobacco Products; no industry funding.
- Page M, Leigh N, Block A et al., Comparison of nicotine and selected flavoring contents between tobacco-free and tobacco-containing oral pouches, Nicotine & Tobacco Research. Supported by the National Cancer Institute (U54 CA228110, P01 CA200512). One author discloses a Pfizer grant and Johnson & Johnson advisory work outside the work.
- Johnson NL, Patten T, Ma M, De Biasi M, Chemosensory contributions of e-cigarette additives on nicotine use, Frontiers in Neuroscience, 2022. NIH-funded; authors declare no commercial relationships.
- Pre- and post-flavor ban oral nicotine pouches: a chemical, sensory, and young adult appeal analysis, Tobacco Control, 2026. Double-blind, n=22. Supported by NCI, FDA CTP and NIDA.
- Peraza N, Han DH, Whaley RC et al., Appeal and sensory characteristics of oral nicotine products in young adults who vape e-cigarettes, Nicotine & Tobacco Research. n=23. NCI, FDA CTP, NIDA and NHLBI funded; no industry money. Full text is withheld by the publisher, so this is scoped to the published abstract.
- Mallock-Ohnesorg N et al., Nicotine pouches: a five-arm crossover study of nicotine delivery and acute effects, Frontiers in Pharmacology, 2024. n=15. “The study was funded by intermural funding from the German Federal Institute for Risk Assessment (BfR), grant No. BfR-CPS-08-60-0102-02.P606.”
- Bundesinstitut für Risikobewertung, Health risk assessment of nicotine pouches, Opinion 023/2022, and the German original. The translation discrepancy described above is between these two documents.
- Stanfill S et al., Characterization of total and unprotonated (free) nicotine content of nicotine pouch products, Nicotine & Tobacco Research, 2021. CDC authors. Full text is on a blocked host; the funding statement could not be read, and the PubMed record shows CDC support only.
- Tran H et al., Total and unprotonated (freebase) nicotine content in new types of oral “tobacco-free” nicotine products, Tobacco Control, 2026. CDC; no specific grant declared.
- Wilhelm J et al., pH of smokeless tobacco determines nicotine buccal absorption, Clinical Pharmacology & Therapeutics. n=40 randomised crossover. Battelle and FDA Center for Tobacco Products affiliations; full text blocked, so the absence of an irritation endpoint is scoped to the abstract and record.
- Lawler TS, Stanfill SB, Tran HT et al., Chemical analysis of snus products from the United States and northern Europe, PLOS ONE, 2020. “The authors have no support or funding to report.”
- Philip Morris Products S.A., EP 4 162 927 A1, Nicotine pouch composition; Swedish Match North Europe AB, EP 3 192 380 A1, Oral pouched product and EP 3 087 852 A1, from the European Patent Office publication server.
- Oral mucosal lesions and nicotine pouch use, Tobacco Induced Diseases, 2026. n=42, cross-sectional; “There was no source of funding for this research.”
- Oral mucosal lesions and inflammatory response in snus users switching to tobacco-free nicotine pouches, European Journal of Oral Sciences, 2022. Funding source listed as Swedish Match AB, Stockholm, Sweden; the authors “declare that they have no conflicts of interest.”
- ZYN, US FAQ and ingredient statement; VELO, ingredients page; Rogue, FAQ. All read 4 November 2026. Company sources.
Last verified 4 November 2026.
