Barrier repair sells. It also gets written into briefs as a single line, as if one active could carry it. In practice the two materials most often asked to do the job, ceramide and hyaluronic acid, are not competing for the same space in the formula. One rebuilds the mortar between skin cells. The other holds water at the surface. Swap them and you get a product that reads well on the front of the pack and underperforms in a corneometer reading twelve weeks later.
Hyaluronic acid is a humectant. It binds water, forms a light film, and improves the immediate feel of almost any emulsion. That immediate feel is why it appears in so many formulas at 0.1%, and why the ingredient has a reputation for being a marketing material as much as a functional one.
Molecular weight changes the story. High molecular weight grades, above roughly one million daltons, stay on the surface and give cushion and slip. Low molecular weight fractions penetrate the stratum corneum and are the ones cited in the deeper hydration literature, though the evidence is thinner than the marketing suggests. What a spec sheet must state is the molecular weight or intrinsic viscosity, because "hyaluronic acid" covers everything from 5 kDa to 2 MDa and the two behave nothing alike in a formula.
Watch the viscosity at your use level. A high molecular weight grade at 1% will turn a lotion into a stringy gel, and formulators then cut the level to 0.2% and wonder why the claim stopped showing up in measurements.
A ceramide on its own does not repair a barrier. It needs cholesterol and free fatty acids in roughly the ratio found in the skin, and it needs to be delivered in a lamellar structure rather than scattered through an oil phase. This is where most ceramide products quietly fail: the ingredient is present at 0.5% and the arrangement is wrong, so the measured effect is close to zero.
Practically, that means melting the ceramide into the oil phase above its own melting point, cooling under controlled shear, and not shocking the emulsion with a cold water addition that forces crystallisation into the wrong polymorph. Ceramide NP, AP and EOP behave differently here, and a blend is not automatically better than a single grade.
Dose is smaller than most people guess. Ceramide at 0.1 to 0.5% is normal; above 1% you are usually paying for a label claim rather than a measured improvement.
The physiological ratio of ceramide to cholesterol to fatty acid is often quoted as 3:1:1 or 1:1:1 depending on which paper you read, and buyers sometimes ask us to match it exactly. Our answer is that the ratio matters less than whether the three are in the same phase and whether the emulsifier tolerates them. A well-built 1:1:1 system outperforms a 3:1:1 system where the cholesterol has crystallised out. If a supplier cannot tell you how their blend behaves in a lamellar system, the ratio on the certificate is decoration.
Ceramides are made by fermentation or by synthetic routes, and the two are not identical in impurity profile or in cost. Fermentation-derived material brings the usual questions about residual media and solvent, and the answer belongs on the COA with numbers attached. Identity is worth checking by LC-MS on the first order, because ceramide is a class of molecules and a certificate that says "ceramide" without naming the species is not a specification.
For hyaluronic acid, ask about fermentation origin, protein content, and whether the grade is cosmetic or ophthalmic. Protein at the low end of the range matters for sensitive-skin positioning, and it is one of the few parameters where a small difference shows up in consumer complaints. Endotoxin is rarely relevant cosmetically but it tells you something about the purification discipline.
The combination works, and it works for a simple reason: HA gives the hydration signal in the first application, ceramide gives the barrier result that shows up in a TEWL measurement after four to eight weeks. A serum with both, positioned on barrier function, needs a study long enough to catch the second effect. Two-week instrumental data will show the humectant and nothing else, which is how a decent formula gets written off as a marketing exercise.
If you are building that study, measure transepidermal water loss at baseline, week four and week eight, and add a corneometry reading at each point. Report both. Buyers in the professional channel ask for the TEWL curve now, and a flat line with a good hydration number beside it tells them exactly what the product is.
For ceramide: species identity, assay, appearance, residual solvent, heavy metals, microbial limits, and a statement of the carrier system if it is a diluted grade. For hyaluronic acid: molecular weight or intrinsic viscosity, glucuronic acid assay, protein, loss on drying, heavy metals, and origin.
Insen Biotech supplies both, with the identity and molecular weight stated per batch rather than per product family. The ceramide and hyaluronic acid pages list current grades, and we will send trial quantities of both if you want to run the lamellar work in your own lab.
You can replace it with a blend that contains ceramide, which is a different decision. Check the actual ceramide content, not the percentage of the blend.
Sometimes, at higher levels. If the product is aimed at compromised skin, start with a majority of high molecular weight material and test before you push the low fraction up.
Send us the brief and the pack format and we will suggest which grade to start with, including the case where one active is enough. Ask for samples and batch documentation.
