Steeping, reviewed — what actually changes in the bottle, and when it stops

Ask a forum how long to steep a custard and you get a number — two weeks, four weeks, until the bottle goes the colour of weak tea. Ask why, and the register changes: the flavours “marry”, the liquid “mellows”. There is a genuine reaction underneath. It has been measured. It does not run to the timetable the folklore prescribes.

The reaction that steeping actually is

Most flavour in a sweet e-liquid rides on aldehydes — vanillin and ethylvanillin for bakery and custard, benzaldehyde for cherry and almond, cinnamaldehyde for cinnamon, citral for lemon. Propylene glycol is not inert to any of them: it reacts, forming an acetal, and the aldehyde you paid for stops being entirely present.

Erythropel and colleagues put numbers on that in Nicotine & Tobacco Research — online October 2018, in the 2019 volume — and the figures remain the most useful thing published about resting a bottle.

Flavour aldehyde to PG acetal — the published conversion

Cinnamaldehyde	About 92 percent converted, plateau inside the first day
Citral	About 88 percent converted, plateau inside the first day
Benzaldehyde	Above 95 percent mole fraction, equilibrium around day five
Ethylvanillin	About 50 percent mole fraction, equilibrium around day seven
Vanillin	About 40 percent mole fraction, equilibrium around day seven
Off-the-shelf bottles	15 to 30 percent of the aldehyde already sitting as its acetal
Carried into vapour	50 to 80 percent of the acetal's concentration in the neat liquid
Conditions	Room temperature; daylight and dark, no significant difference

Those figures come from the published report: a bottle that is a slow reactor, not a container, its clock set by which aldehyde is inside.

The timetable is days, not weeks

Read the table as a buying instruction and the folklore falls apart two ways.

A cinnamon or lemon blend is finished before you open it — close to ninety percent of the change done within a day of mixing, at the factory. Resting cinnamon Danish for a fortnight is theatre; what it tastes like on day fourteen it tasted like on day two.

A custard is the opposite case, the one where the advice is closest to right. Vanillin and ethylvanillin are the slow pair, still moving on day three and settling around day seven — and settling at 40 to 50 percent, not at completion. That is a mechanism for the thing bakery vapers insist on, and it explains why the same shelf that swears by steeping is disproportionately a bakery and custard shelf. It also caps the wait at about a week; the remaining three weeks of a month-long rest are storage, not steeping.

Two caveats. Acetal formation is the best-documented change in a resting bottle, not the only one; the other changes people report — nicotine oxidising, sweetener coming fully into solution — are not what this paper measured. And it measures compounds, not liking: no timetable for “tastes better” appears in this literature, and anyone quoting one is quoting a preference.

The 2026 addition: water

Several of the same authors came at it from the other end this year. Davis and colleagues, in Tobacco Control in May 2026, tested whether adding water suppresses acetal formation, and whether users notice.

In laboratory-made liquids the reductions ran 2.6 to 38-fold for vanillin and ethylvanillin, 1.3 to 7.5-fold for benzaldehyde and cinnamaldehyde. Commercial liquids were lopsided too. At 20 percent water a vanilla product’s vanillin and ethylvanillin acetals fell about 25-fold, while the cinnamon and cherry products moved less than 2-fold. Thirty users then rated the formulations, and water concentration had no main effect on rated irritation, flavour intensity or sweetness; only at 20 percent did appeal and reward drop significantly against the no-water and 5-percent conditions.

That is a formulation and policy proposal, not a kitchen tip; the trial used prepared liquids at controlled water fractions. The asymmetry is the interesting part for a tasting bench: the two aldehydes water helps most are the two the folklore is built around.

The ledger

Steeping, assessed as a practice

Pros

There is a measured chemical mechanism underneath the folklore, not just shelf superstition
The class of aldehyde predicts the timetable, so a buyer can stop guessing
Bakery, custard and vanilla blends are the ones where a week's rest has a documented basis
It costs nothing but shelf space and a week of patience
The published work is specific enough to falsify bad advice, rare in this category

Cons

The published equilibrium times are days; almost every steeping guide in circulation says weeks
Cinnamon and citrus blends reach equilibrium within a day of mixing, so a rest adds nothing day one has not already done
The chemistry measures compounds, not enjoyment — no study establishes when a liquid tastes its best
Steeping cannot rescue a thin, unbalanced or oversweetened blend; it alters composition, not craft
The same reaction yields adducts the 2018 paper's title credits with 'novel toxicological properties', and 15 to 30 percent of that conversion has happened before you buy
Bottle panels state neither the flavour-aldehyde class nor a rest time, so any schedule handed to you is inference rather than disclosure

The score is for the practice, not for any bottle. Steeping earns its keep on a narrow band of the shelf — vanillin-led builds, on a horizon of about a week — and nothing on the cinnamon and citrus blends where it is most confidently prescribed. What it never does is improve a badly made juice; the schedules that circulate for thin bottles often look like an excuse.

The fix is dull and available: publish the flavour-aldehyde class and a rest time, and stop letting forums write the instructions. Until then, treat any timetable — this one included — as inference from published chemistry, not a promise about your bottle.

Whatever is in the flavour, the bottles this bench covers carry nicotine, an addictive chemical, and nothing above changes that. This desk writes for adults of legal purchase age — 21 or over in the United States, 18 or over in most other markets — and for people who already use nicotine, not for anyone thinking of starting.