Three alpha hydroxy acids, all sold as exfoliating actives, all described with the same language, all behaving differently at the same number on the label. Glycolic, lactic and mandelic are not interchangeable, and the reason is mostly physical: the size of the molecule and the pH of the finished product together determine how much of the acid is actually free to act.
The single most common formulation error in this category is treating the label percentage as the dose. It is not. Two products both described as 10% glycolic can differ by a factor of several in the amount of free acid they deliver, and the customer feels the difference immediately.
Free acid value is the number that matters, not the label percentage
An acid in a water-based product exists in two states. The undissociated form is the one that can cross into the skin; the dissociated form carries a charge and behaves differently. The balance between them is set by the pH of the formula relative to the acid's pKa.
The consequence is that the useful measure is not how much acid you added, but how much is present in the undissociated form in the finished product. That quantity is the free acid value, and it is calculated rather than measured off the label.
The practical rule of thumb is that free acid rises sharply as pH falls below the pKa of the acid and falls off just as sharply as pH rises above it. A product at pH 3.5 with 10% acid and a product at pH 4.5 with 10% acid are not the same product, and the second one may be closer to a moisturiser with an acid in the ingredient list than to an exfoliating serum.
For a brand, this is the question to ask a supplier: what is the free acid value of the finished formula, and at what pH? A supplier who answers in label percentages has not understood the question.
Molecule size sets penetration
The three acids differ in molecular weight, and molecular weight governs how readily the molecule moves through the outer layer of the skin.
Glycolic acid is the smallest of the three, with the simplest structure - two carbons. Lactic acid has three carbons and a hydroxyl group on the second carbon. Mandelic acid has eight carbons and an aromatic ring.
Size alone produces three predictable differences. The smaller the molecule, the faster the uptake and the more immediate the sensation. The larger the molecule, the slower and more even the effect, and the gentler the experience tends to be. This is not a marketing distinction; it is the same relationship you would expect from any small molecule diffusing into a barrier.
There is a second structural factor that matters for mandelic acid specifically: the aromatic ring makes the molecule more lipophilic than its size alone would suggest, and it also means the material is supplied as a racemic mixture in most commercial grades. The two optical forms behave slightly differently. That is a specification detail worth asking about, and it is almost never discussed in marketing material.
Glycolic: the smallest and the fastest
Glycolic acid gives the most for a given level and the least tolerance for error.
Its small size makes uptake quick, which is why it is the acid most associated with an immediate, noticeable result and also the one most likely to produce stinging. It is highly water-soluble, so it fits an aqueous serum easily and needs no solubiliser.
Formulation constraints to be aware of:
- pH control is critical. Because the free acid value moves so steeply with pH, a small drift in a glycolic formula changes the product meaningfully. The buffer choice and the stability protocol both have to respect that.
- The sensory experience is sharp. On skin with any degree of compromise, the sensation is immediate. That is a function of the active rather than the base, and it cannot be formulated away without dropping the free acid value.
- It is not the right choice for a product that has to be used daily by everyone. If your positioning is a gentle everyday product, glycolic is the wrong plant to put in it.
Lactic: larger, and the one that suits drier skin
Lactic acid is one carbon larger and noticeably more forgiving. The uptake is slower, the peak sensation is lower, and the material has an additional property that makes it attractive in a moisturising context: it is itself a humectant.
That combination is why lactic acid appears so often in products positioned around smoother-looking skin on the drier side. The acid contributes to the exfoliating story and the same molecule helps the formula hold water, which is an unusually neat overlap.
It is also cheaper than glycolic in most grades and easier to handle in a base, since it is supplied as a liquid rather than a crystalline solid.
The caution is the same as for any acid: the label percentage is not the dose. A lactic serum at pH 4.0 behaves like a much milder product than the same percentage at pH 3.5, and both may be described as "10% lactic acid".
Mandelic: the aromatic acid with the slowest uptake
Mandelic acid is the largest of the three and the one with the least immediate sensation. Its eight-carbon structure with an aromatic ring means slower diffusion and a more gradual experience.
The commercial argument for mandelic acid is twofold. First, it suits customers who find the smaller acids too strong. Second, it is often positioned for skin tones where the appearance of post-inflammatory marks is a concern, because the slower action reduces the risk of a visible reaction that would make the marks look worse.
Two practical points. It is much less water-soluble than glycolic or lactic, so a high-load mandelic formula needs either a solubiliser or an alcohol content that changes the skin feel. And it is significantly more expensive, which matters because the level needed at a given pH is higher in weight terms.
The aromatised structure also means it has a faint characteristic odour. In a perfumed product this disappears. In a fragrance-free, minimal product it does not, and it has ended more than one sampling round in a reformulation.
pKa, pH and the buffering problem
Each acid has its own pKa, and the number describes the pH at which half the acid is in the undissociated form. Below the pKa, more of it is undissociated; above it, less.
Two formulation consequences follow.
First, the target pH for a given free acid value differs between the three acids. You cannot set one pH and expect the same behaviour from all of them.
Second, the buffer system has to hold the pH without introducing a large electrolyte load. A poorly chosen buffer can leave the formula with a pH that drifts during storage, and because the free acid value is steeply pH-dependent, a drift of a few tenths can move the product from a meaningful exfoliating serum to a mild toner. Our stability article covers the protocol that catches it.
Working out the free acid value
The calculation follows from the acid dissociation relationship. For a monoprotic acid, the fraction present in the undissociated form is a function of pH minus pKa:
- Two units below the pKa - almost all of the acid is undissociated. This is the high-activity zone.
- At the pKa - half of the acid is undissociated.
- Two units above the pKa - very little is undissociated, and the product is behaving as a salt solution with an acid in it.
The practical version: the same 10% of acid gives roughly ten times more free acid at pH 2.8 than at pH 4.8. That factor is the reason the two products feel completely different despite identical labels.
The calculation is straightforward and any competent formulator will run it. If the answer you get back is a restatement of the percentage you specified, ask again in writing.
The three acids compared
| Property | Glycolic | Lactic | Mandelic |
|---|---|---|---|
| Relative molecule size | Smallest | Small | Large |
| Uptake | Fastest | Moderate | Slowest |
| Sensation on skin | Strongest | Moderate | Mildest |
| Water solubility | Very high | Very high | Limited; needs solubilising |
| Extra benefit | None beyond the acid action | Humectant in its own right | Slow, even action; suited to mark-prone concerns |
| Cost per kg | Lowest | Low | Highest |
| Best fit | Positioning built around visible results | Drier skin; everyday products | Reactive or reactive-leaning skin; uneven-tone positioning |
Sensation, stinging and the sensory trade-off
The stinging sensation that customers describe is not a formula defect. It is a predictable consequence of an acid acting on skin, and it scales with the free acid value rather than with the label.
That gives you two levers and no third. You can lower the free acid value, which reduces both the sensation and the effect. Or you can choose a larger molecule at the same free acid value, which reduces the sensation while keeping a comparable effect. There is no way to get the fastest possible action with the mildest possible sensation, and a brief that asks for both is asking for something that does not exist.
What you can do is manage expectations honestly in the copy and control the base: a formula with a slightly higher pH and a slower acid can feel significantly more comfortable than a sharper product, and the customer will usually describe that as a better product rather than a weaker one.
For formats where the acid sits alongside a lower pH, the interaction with the preservative system matters as well - see our preservative article.
Neutralising and the risk of over-neutralising
In production it is common to add the acid to the water phase and then adjust the pH upward with a base. This is normal practice and it carries one recurring risk.
Overshooting the neutralisation point leaves the formula above the target pH. Because the free acid value falls steeply as pH rises, a small overshoot has a disproportionate effect on the finished product. Then, because the pH reading on the bulk is the number everyone trusts, the batch is reported as compliant against a specification that said "pH 4.0" rather than against a free acid value.
The fix is procedural rather than technical: specify both the pH and the free acid value, neutralise in small increments with a full mix between additions, and take the pH reading on a cooled sample rather than a warm one. Temperature affects the reading, and a warm sample read against a cool specification is the other half of this problem.
Packaging: container materials and closures
A low-pH product is a more demanding container environment than a neutral one.
The classic issue is metal. Acidic formulas can attack aluminium components, and the result is a product that develops an off-note or an unexpected colour over time. Pumps and closures with metal springs or discs in contact with the product need to be checked rather than assumed.
There is also a migration question with some plastics and liners, and a permeation question for volatile components at low pH. All of it falls under packaging compatibility testing, which should be run on the finished formula in the finished pack rather than on a generic simulant - see our packaging compatibility article.
For a single-dose capsule format, the shell is part of the packaging and the acid has to be compatible with it. That is a different test again, and it is worth running early because it can change the whole format decision.
Market limits and what they mean for the brief
Markets treat at-home acid products differently from professional-use products, and the difference is usually expressed as a combination of a maximum concentration and a minimum pH.
European rules restrict the level and the pH of at-home AHA products and require a warning on the label where the restriction applies. Other markets set their own thresholds, and some treat higher-concentration products as a separate category requiring professional application.
The practical implication for a brief is that the target market and the intended user - consumer or professional - have to be settled before the formula is designed. A formula built for one can be non-compliant in the other, and the fix is not a labelling change. Our reformulation article covers what changing markets after the fact actually involves.
Claims
Acid products support a well-established cosmetic claim set: the appearance of smoother skin, a more even-looking surface, the appearance of refined texture, and brighter-looking skin.
What they cannot claim is any treatment of a skin condition, any thinning or removal of skin, and any comparison to a professional procedure. Peeling language is particularly risky because it describes a physical process and it sits close to a medical framing. Our efficacy testing article sets out how the allowable versions are put together and substantiated, and our brightening article covers the adjacent claim boundary.
Testing to commission
- Assay on the acid at release and at stability points, so the free acid value stays true rather than being calculated once.
- pH at release and across the stability period, because drift is the failure mode that matters here.
- Free acid value reported as a number, not inferred from the pH.
- Preservative efficacy at the finished pH, since many preservatives are less effective in an acidic environment.
- Pack compatibility on the finished pack if any component is metallic.
- Consumer use testing that captures the sensation, because that is the variable most likely to drive returns.
Where our acid serum capability fits
We formulate glycolic, lactic and mandelic systems with the free acid value specified rather than implied, and with the pH and buffer system designed together. Formats include a facial essence capsule, a custom cosmetic softgel face serum capsule and a amino acid facial cleanser where the acid story sits in a rinse-off format instead.
We work with brands on private label (your brand on a proven base), OEM/ODM contract manufacturing, and full custom formula development - whichever matches where your brand is today.
Tell us the market, the intended user and the sensation you are prepared to stand behind, and we will come back with a pH and a free acid value rather than a percentage. Start at our inquiry page, email capsulebiotech@163.com, or message the right team on WhatsApp: Capsule Series (capsules only) +86 158-7655-5319, Cosmetics Series (serum, hair care, body care, fragrance) +86 158-7655-5319.
General manufacturing information, not regulatory or legal advice. Claim limits and concentration limits vary by market.