Face Serum — Material Selection Guide

Key Technical Parameters #

Material selection for face serum starts before the formula exists. The decisions that determine whether a finished serum performs consistently — batch to batch, year to year — happen when we’re qualifying raw materials, not when we’re adjusting viscosity on the bench. Brand owners in the premium and clinical skincare segments feel this most acutely: their claims depend on active delivery, and active delivery depends entirely on the input material meeting spec every time. This guide covers the six material-level criteria we apply during supplier qualification and batch release, with the numeric thresholds we’ve built into our internal MQ-03 material qualification protocol. Where those thresholds came from, what happens when a material just barely passes, and where we’ve had to tighten our own limits beyond supplier certificates — that’s what makes this worth reading.

When the Certificate of Analysis Isn’t Enough #

We’ve released a serum batch that passed every internal QC check, shipped to a brand partner’s 3PL, and within four months received a complaint about colour change. The CoA was clean. The supplier’s spec was met. The problem turned out to be residual peroxide value in the botanical extract — well within the supplier’s stated limit of ≤5 meq/kg, but high enough at 4.6 meq/kg to initiate slow oxidation of the vitamin C derivative in the same formula. The supplier’s limit was written for food-grade use. We were using the material in a leave-on, anhydrous phase with a reactive active. Those aren’t the same application.

That incident is why we now run our own peroxide value testing on every incoming lot of any oxidation-sensitive botanical, regardless of what the CoA says. Our internal acceptance threshold for those materials in antioxidant or vitamin C serums is ≤2.0 meq/kg. That’s tighter than most suppliers will guarantee. Some can’t even hit it consistently.

This is the real gap in serum manufacturing: suppliers write specs for their broadest market, not for your specific formulation context. Closing that gap is what material selection actually involves.

The Six Parameters We Actually Screen On #

Most incoming ingredient specs focus on appearance, odour, and heavy metals. Those matter, but they’re not where serum performance breaks. The parameters below are what we screen against in our MQ-03 protocol, with the thresholds we apply.

Purity and active content assay. For high-actives face serums, the declared concentration on a CoA needs to be backed by an assay method that matches your intended use. Niacinamide is a good example — CoAs routinely show ≥99.0% purity by titration, but some grades carry 2–5% nicotinic acid as a synthesis byproduct. Nicotinic acid causes flushing at concentrations above roughly 0.005% in the finished product. If you’re loading niacinamide at 5% and your raw material carries 3% nicotinic acid, the math is uncomfortable. We specify ≤0.2% nicotinic acid maximum for niacinamide grades used in leave-on serums.

Water activity (aw) for hygroscopic actives. Anhydrous or low-water serum formats are particularly sensitive here. Hyaluronic acid powder, peptide powders, and certain botanical extracts are hygroscopic. Water activity above 0.60 in an incoming powder significantly increases microbial risk during handling and shortens shelf life post-incorporation. We reject incoming HA powder lots with aw >0.55.

Particle size for dispersed-phase materials. This matters most for mineral actives, pigments, and encapsulated systems. A zinc oxide material specified at ≤200nm for a UV-protective serum that arrives at 320nm average particle size will behave differently in both feel and function. Particle size distribution — specifically D90 — is what we track, not just the mean. Our threshold for ZnO in transparent serum formats is D90 ≤150nm. Materials arriving outside that range get rejected or redirected to opaque formats. For more detail on mineral UV technology, the particle size selection criteria apply similarly there.

Heavy metal profile beyond the standard panel. EU Cosmetics Regulation 1223/2009 specifies limits for lead (≤10 ppm), arsenic (≤5 ppm), mercury (≤1 ppm), and cadmium (≤5 ppm) in cosmetic products. Most CoAs cover these. What’s less consistently reported is nickel, which isn’t in the EU restricted list at a specific finished product limit but has a SCCS Scientific Opinion recommending ≤1 ppm in leave-on cosmetics for nickel-sensitised consumers. We run nickel on incoming clay materials and mineral extracts. Some grades of kaolin and titanium dioxide arrive with nickel levels that would give pause for a leave-on format.

pH and acid value for oil-phase materials. Oils oxidise. Oxidation products are acidic. An emollient oil’s acid value tells you something about its oxidative state at the point it was tested, but it doesn’t tell you about the stability trajectory. We track acid value on incoming lots of rosehip, sea buckthorn, and marula across our supplier base, and we’ve seen meaningful lot-to-lot variation within the same supplier. Our internal limit for rosehip oil in an active serum is acid value ≤3.0 mg KOH/g. Materials arriving at 4.5–5.0 mg KOH/g — still within some suppliers’ stated specs — tend to cause early rancidity signals by week 8 of accelerated stability.

Microbiological baseline before preservation. A water-phase material arriving with a TVC of 10² CFU/g is a very different starting point from one arriving at 10⁵ CFU/g, even if both technically pass standard “acceptable” limits. Our baseline TVC acceptance for water-phase serum ingredients is ≤100 CFU/g. Anything higher gets flagged under what we call a Category B incoming review, which includes a hold and a re-test before use. Two out of roughly every fifteen botanical hydrosol lots we receive require a Category B review.

Comparing Material Grades: What the Spec Sheet Hides #

Not all grades of the same ingredient are interchangeable. The table below shows how this plays out across three common serum actives when you compare standard commercial grade versus pharmaceutical/high-specification grade.

The underlying principle is straightforward: supplier spec limits are written to minimise rejection rates, not to optimise for your formulation’s most sensitive interaction. Every row in that table represents a failure mode we’ve encountered on production batches.

Decision Logic: How Format and Market Change the Threshold #

Material qualification isn’t a fixed checklist. The thresholds shift based on three variables: delivery format, target market, and active system.

If the format is a single-use ampoule at high active concentration — 15% vitamin C, for example — you need pharmaceutical-grade ascorbic acid derivative with a residual solvent panel, not a standard cosmetic CoA. At that loading, any trace contaminant that would be inconsequential at 2% becomes a stability and sensory issue. The cost delta between grades at those volumes is typically a few hundred USD per 100kg. Brands sometimes push back on this. We don’t move on it.

For the EU market specifically, you need to run nickel, chromium, and cobalt on any clay or mineral-derived material, even where the regulation doesn’t mandate it. This isn’t about compliance; it’s about pre-empting a dermatologist-reported adverse event that triggers a SCCS Scientific Opinion level review. We’ve flagged this to brands often enough that it’s now part of our standard market-specific qualifier in new project kickoffs.

For markets governed by FDA Cosmetics Guidelines, the heavy metal profile matters particularly for any colour cosmetic adjacent claim (brightening, correcting). FDA doesn’t have a single harmonised limit table for cosmetic impurities the way the EU does, but import alerts have been triggered by lead levels in the 5–8 ppm range in some categories. Our standard US-market limit is ≤5 ppm for lead across the full formula, which means incoming raw material limits need to be tighter than that to provide headroom.

If the active system involves retinol, the material grade question becomes acute. A split-face randomised controlled trial (n=36, 16 weeks) published in the Journal of Cosmetic Dermatology in 2020 demonstrated 28% improvement in fine line depth at 0.1% retinol in a serum vehicle versus vehicle control. What that study doesn’t discuss is the raw material quality underpinning the retinol used — and in our experience, the difference between a batch that delivers consistent efficacy and one that underperforms within its stated label life often traces back to the incoming all-trans retinol content and whether the supplier has a cold-chain handling spec. Most don’t publish one. We ask for it anyway.

Formulation Notes for Brand Partners #

When you brief us on a new serum, the first questions we ask are about market and format — not ingredients. Those two decisions determine the qualification burden before we even open the formula.

A common mistake we see: brands spec an active concentration based on a clinical study, request that we source the “same ingredient” used in the study, and don’t realise that the study ingredient may have been a research-grade material produced by the study sponsor’s own synthesis team. We can source a commercially available grade with the same INCI name. Whether it matches the study material’s purity profile is a different question, and the honest answer is that we usually don’t know.

What we can do is set appropriate incoming material specs based on the active’s known failure modes, build those into our incoming QC protocol, and monitor lot-to-lot variation over the first 12 months of production. That consistency data is available to brand partners and is something we’d encourage you to ask for.

Standard timeline: lab samples in 2–3 weeks from material approval, accelerated stability (40°C/75% RH, 12 weeks) running concurrently, real-time 24-month stability initiated at sample sign-off. Material qualification for a new supplier typically runs 3–4 weeks in parallel with formulation, not sequentially — that’s where most projects save time.

Frequently Asked Questions #

Can we just use whatever grade our previous manufacturer used?
A: Probably, but we’d want to see their incoming QC spec, not just the finished product CoA. If they weren’t testing for nicotinic acid in niacinamide or peroxide value in retinol, the grade may be fine — or it may be the reason your stability results were inconsistent. Hard to say without the data.

The EU market is our priority — what extra testing do we actually need on incoming materials?
A: At minimum, run the full heavy metals panel including nickel on any mineral or clay-derived material, and apply the ≤1 ppm nickel limit aligned with the SCCS Scientific Opinion recommendation for leave-on products. The EU Cosmetics Regulation 1223/2009 doesn’t enumerate a specific nickel limit for finished cosmetics, but sensitisation complaints in the EU are tracked and can trigger post-market scrutiny quickly.

We want to use a botanical extract from a new supplier — how long does qualification take?
A: For a water-phase botanical hydrosol or aqueous extract, figure 3–4 weeks for a first-pass qualification: identity, microbiological baseline, heavy metals, pesticide residues, and a 4-week compatibility check in your intended formula base. If the extract is high-phenolic or contains potential photosensitisers, add another week for phototoxicity pre-screen. We’ve had extracts that looked clean on paper but showed unexpected UV absorption profiles that complicated claims review.

MOQ for higher-grade raw materials — does that affect our minimum batch size?
A: Yes, sometimes. Pharmaceutical-grade actives often come in smaller standard pack sizes (1–5kg) with higher per-kg costs, which affects how we price pilot batches. For production batches of 200kg or more, the cost difference is usually manageable — roughly the same order of magnitude as the premium packaging you’re probably already specifying. Where it gets tight is on initial sampling: a 1kg pharmaceutical-grade retinol supply for a 5kg lab batch adds cost that some brands don’t anticipate.

What’s the material quality variable brands most often overlook in serum projects?
A: Water activity on hygroscopic powder actives. We see this consistently — a brand will request a preservative-free or minimal-preservation positioning, but the incoming powder HA or peptide blend is arriving with aw above 0.60 because it’s been stored incorrectly somewhere in the supply chain. At that point the preservation burden in the finished formula goes up, which conflicts with the original brief. We check this now as part of every preservation-sensitive project brief, but it’s not something most brands think to flag upfront.

---

Have a product concept in mind? Contact our formulation team to request a complimentary brief review.