Concentrated Actives Delivery: Waterless Serum Actives Loading & Penetration Data

Overview #

Waterless formulation is not just a sustainability story. It is a delivery architecture decision. When you remove water as the continuous phase, you fundamentally change how actives load, how they penetrate, and what concentration ceilings are actually achievable before stability breaks down. We work on these systems daily — anhydrous serums, oil-in-oil dispersions, concentrated balms — and the penetration data we generate in-house consistently outperforms equivalent aqueous formats at the same nominal active concentration. The question brand partners should be asking is not “can we go waterless?” but “which actives actually benefit from it, and what does the clinical evidence say?”

Why Waterless Changes the Penetration Equation #

The stratum corneum is lipophilic. Water-based vehicles have to fight that barrier. Anhydrous vehicles — silicone carriers, squalane, C12-15 alkyl benzoate blends — are already chemically closer to the skin’s own lipid matrix, so the partition coefficient works in your favor from the first second of contact.

In our lab, we measure this using Franz diffusion cell assays with excised porcine skin. Across more than 40 formulation runs over the past three years, anhydrous formats consistently show 1.4× to 2.1× higher cumulative permeation at 24 hours compared to matched aqueous controls at identical active concentrations. That range is wide because it depends heavily on the active, the carrier blend, and the molecular weight of the penetrant. Not every active benefits equally. Vitamin C derivatives and retinoids show the biggest gains. Peptides are more complicated — molecular weight above 1,000 Da and you’re largely relying on follicular penetration regardless of vehicle.

The other thing waterless does is remove the primary degradation pathway for oxidation-sensitive actives. No water means no hydrolysis. For ascorbic acid, that alone extends functional shelf life from roughly 3 months in a standard aqueous serum to 18+ months in an anhydrous format — and that is not a marketing claim, that is what our accelerated stability data at 40°C/75% RH shows across multiple batches.

One thing we are still working through: anhydrous systems can create a false sense of security on preservation. No water does not mean no microbial risk. We have seen gram-positive contamination in anhydrous balms that contained botanical extracts with residual moisture. We now require water activity (Aw) testing on every botanical raw material before it enters an anhydrous batch. Aw above 0.6 is a rejection criterion for us.

Clinical Evidence by Active: What the Data Actually Shows #

This is where most brand briefs go wrong. Brands cite ingredient supplier data sheets as clinical evidence. Supplier data is useful, but it is not the same as an independent study, and regulators — especially in the EU — are increasingly distinguishing between the two.

Retinol in Anhydrous Vehicle

The most relevant head-to-head data we reference internally comes from a double-blind, randomized, vehicle-controlled trial (n=36, 12 weeks, twice-daily application) that measured fine line depth via optical profilometry. The anhydrous retinol group (0.3% retinol in a squalane-silicone base) showed a 34% reduction in average wrinkle depth versus 19% in the aqueous retinol control at the same concentration. What the study does not capture — and what we see in our own stability work — is that the anhydrous group maintained retinol potency above 90% of label claim at week 12, while the aqueous group had degraded to approximately 61%. So part of the clinical delta is simply that the anhydrous product was still delivering active at week 12 and the aqueous product largely was not.

We stabilize retinol in anhydrous systems at a target pH-equivalent activity of 5.0–5.5 (measured in a 10% aqueous dilution of the finished product), using tocopherol at 0.5% and BHT at 0.02% as antioxidant co-stabilizers. Nitrogen blanketing during manufacturing is non-negotiable. One batch we ran without proper nitrogen purge showed 18% retinol degradation within 6 weeks at ambient storage. That batch was rejected.

For more on our retinoid stabilization protocols, see our Retinoid Technology formulation documentation.

Ascorbic Acid (Pure L-AA) in Anhydrous Serum

Pure L-ascorbic acid in aqueous format is notoriously unstable. Most brands end up using derivatives — ascorbyl glucoside, 3-O-ethyl ascorbic acid — because they are easier to stabilize. In anhydrous format, pure L-AA becomes viable again, and the clinical performance difference matters.

A split-face, investigator-blinded study (n=22, 8 weeks, once-daily application) compared a 15% L-AA anhydrous serum (silicone-free, squalane/jojoba base) against a 15% ascorbyl glucoside aqueous serum. Melanin index reduction at week 8: 12.4% for the anhydrous L-AA group versus 6.1% for the ascorbyl glucoside group. The researchers attributed the gap to both higher bioavailability of free ascorbic acid and better maintained potency over the study period. We are not fully convinced the study design controlled for all confounders — the vehicle difference was significant — but the direction of the result aligns with what our Franz cell data shows.

The formulation challenge with anhydrous L-AA is particle size and dispersion stability. We use micronized L-AA (D90 < 15 µm) suspended in a structured silicone network. If particle size drifts above 25 µm, you get grittiness on application and uneven skin contact. We have had two batches fail sensory QC for exactly this reason when a supplier changed their milling process without notifying us. We now specify D90 and D50 in every L-AA raw material purchase order.

Niacinamide in Concentrated Anhydrous Format

Niacinamide is water-soluble, which creates an interesting challenge in anhydrous systems. We use a propylene glycol / glycerin co-solvent approach to dissolve niacinamide at up to 10% before incorporating into the anhydrous base. Above 10%, we see crystallization on cooling that is very difficult to reverse at production scale.

A randomized, double-blind, placebo-controlled study (n=50, 12 weeks) using a 5% niacinamide anhydrous concentrate showed a 23% reduction in sebum excretion rate and a 17% improvement in skin barrier function (TEWL reduction) versus vehicle control. The sebum result is consistent with what we see in consumer perception studies we run for brand partners — typically 70–80% of panelists report visible pore improvement by week 6. The TEWL improvement is the more interesting finding for us, because it suggests the anhydrous vehicle itself may be contributing to barrier support independent of the niacinamide.

For brands building brightening or barrier-focused SKUs around niacinamide, our Brightening & Whitening formulation documentation covers the full active interaction matrix including niacinamide-vitamin C compatibility in anhydrous systems.

Evidence Strength Comparison: Anhydrous vs. Aqueous Delivery #

The bakuchiol entry deserves a note. We include it because it is one of the most-requested actives in waterless briefs right now, and the evidence base is genuinely thin. One open-label study with 20 subjects and no blinding is not strong clinical substantiation. We tell brand partners this directly. The ingredient has a good safety profile and the in-vitro retinoid receptor activity data is interesting, but if you are planning EU claim substantiation based on clinical evidence, bakuchiol is not ready for that yet.

Where Most Brands Get the Loading Limits Wrong #

There is a ceiling on how much active you can load into an anhydrous system before something breaks — either stability, sensory, or both. Brands consistently push past it.

Retinol above 1.0% in anhydrous format: we have done it, but three out of five batches at that concentration show yellowing by week 10 at 40°C even with full antioxidant protection. Most brands do not need 1.0% retinol. They need stable 0.3% retinol that is still potent at month 12. That is a harder formulation problem and a better product.

Vitamin C above 20% in anhydrous suspension: sensory failure before stability failure, in our experience. The gritty, powdery skin feel at application becomes a consumer complaint driver. We cap our standard anhydrous L-AA serums at 15% for this reason. Some brands push to 20% for on-pack claims. Honestly, the penetration data does not support a meaningful efficacy difference between 15% and 20% in anhydrous format — the skin’s uptake capacity is the limiting factor, not the concentration gradient above a certain threshold.

Scale-up is where this gets real. We had a concentrated peptide-niacinamide anhydrous serum that performed perfectly at 2 kg lab scale. At 150 kg production, the co-solvent phase did not homogenize uniformly because our high-shear mixer geometry was optimized for aqueous emulsions, not viscous anhydrous blends. The result was niacinamide crystallization in approximately 30% of the batch. We lost that batch. We now run anhydrous scale-up trials at 20 kg intermediate scale before committing to full production, and we specify mixing speed, temperature ramp rate, and addition sequence in the manufacturing record with zero tolerance for deviation.

This is usually where projects go sideways — not in the lab, but in the first production run.

Claim Substantiation Guidance: EU, US, and NMPA #

Getting the clinical data is one problem. Knowing what that data actually allows you to say on-pack is a different problem entirely. The rules are not the same across markets, and we see brand partners make expensive mistakes here.

EU Market

Under EU Cosmetics Regulation 1223/2009, cosmetic claims must be substantiated by evidence that is “adequate and verifiable.” The EU Common Criteria (Regulation 655/2013) require that claims be truthful, evidenced, honest, fair, and not misleading. In practice, this means your clinical study needs to be conducted on the finished product — not the raw ingredient — and the study design needs to be appropriate for the claim being made. A supplier-funded in-vitro study does not substantiate an on-pack efficacy claim in the EU. We have seen brands receive non-conformance notices from notified bodies for exactly this.

The SCCS Scientific Opinion framework is also relevant for any active with a concentration above established safe-use thresholds. Retinol above 0.3% in face products, for example, is currently under SCCS review, and the regulatory position may tighten. We are watching this closely.

US Market

The FDA Cosmetics Guidelines framework is more permissive on claim substantiation than the EU, but the drug/cosmetic boundary is the critical line. Claims that imply physiological change — “increases collagen production,” “repairs DNA damage” — cross into drug territory under the FD&C Act. “Visibly reduces the appearance of fine lines” stays on the cosmetic side. The distinction sounds obvious but it is not always clear in practice, especially for actives like retinol and niacinamide where the mechanism of action is well-characterized.

Substantiation standard in the US is “competent and reliable scientific evidence” — which the FTC enforces for advertising claims. Two well-designed human studies is the informal benchmark the FTC has applied in enforcement actions, though this is not codified.

NMPA (China)

NMPA Cosmetic Regulation is the most demanding of the three for functional claims. China’s 2021 Cosmetic Supervision and Administration Regulation (CSAR) introduced a “special cosmetics” category that includes whitening, anti-hair loss, and sunscreen products — these require pre-market registration with clinical data submitted to NMPA. For general cosmetics with efficacy claims, the 2022 Efficacy Claim Evaluation Guidelines require human efficacy testing conducted by a NMPA-recognized testing institution. Foreign clinical data is accepted in principle but scrutinized heavily. In practice, most of our brand partners targeting China run a separate China-specific efficacy study at a recognized domestic CRO.

One thing that catches brands off guard: NMPA requires that the efficacy testing be conducted on the product as it will be sold in China — same formula, same packaging. If you reformulate for China (which sometimes happens for ingredient compliance reasons), you need new efficacy data. This is not a small cost.

For stability documentation requirements across all three markets, the ICH Stability Guidelines provide the technical framework we align our stability protocols to, even for cosmetics where ICH is not formally required — it gives regulators a familiar reference point.

Formulation Notes for Brand Partners #

What market? What are you expecting on-pack? Those are the first two questions we ask when a waterless serum brief comes in, because the answers determine everything from active selection to study design to packaging format.

If you are targeting EU with efficacy claims, budget for a finished-product clinical study — minimum n=30, 8–12 weeks, with an appropriate instrumental endpoint. That is a real cost line item, typically $15,000–$40,000 USD depending on study design and CRO. If you are targeting US only with appearance-based claims, supplier data plus a well-designed consumer perception study (n=50+) is often sufficient. China requires its own track entirely.

On the formulation side: tell us your target active concentration before we discuss vehicle. If you want 0.5% retinol in an anhydrous serum, we will tell you that is achievable with the right antioxidant system and packaging. If you want 2.0% retinol, we will tell you we have not solved that stability problem reliably at production scale and we are not going to take your money pretending we have.

Packaging is not an afterthought in waterless systems. Airless pump is almost always the right call for oxidation-sensitive actives — it adds $0.40–$0.80 per unit at MOQ 3,000, which most indie brands can absorb, but it is a conversation we need to have at brief stage, not after the formula is locked. We have had projects where the brand selected packaging first and the formula second, and we ended up reformulating twice because the headspace oxygen exposure in the chosen jar format was incompatible with the retinol load.

Start with what you want the consumer to feel and see at week 8. Work backwards from there.

Frequently Asked Questions #

Q: We want to put “20% Vitamin C” on the pack for a waterless serum — is that actually stable and does it perform better than 15%?

A: Stability at 20% L-AA in anhydrous format is achievable, but sensory is the real problem — most consumers find the texture gritty above 15%. Our Franz cell data does not show a meaningful penetration difference between 15% and 20% in anhydrous systems. Honestly, “15% stabilized L-AA” with a good stability story is a stronger product than “20% L-AA” that feels unpleasant on skin.

Q: Can we skip the clinical study for EU if we use published ingredient supplier data?

A: No. EU Regulation 655/2013 requires substantiation on the finished product. Supplier ingredient data supports your technical dossier but does not substitute for finished-product evidence. For appearance-based claims, a well-designed consumer perception study (n=50, 4 weeks minimum) is the practical minimum. For functional claims, you need instrumental measurement.

Q: What’s the maximum retinol concentration you can reliably stabilize in an anhydrous serum?

A: In production, we reliably hit 0.5% with full antioxidant protection (tocopherol 0.5%, BHT 0.02%), nitrogen blanketing, and airless packaging. We have done 1.0% but the batch failure rate at that concentration is high enough that we do not recommend it for commercial production. Above 1.0%, we redirect brands to encapsulated retinol systems — see our Encapsulation Technology documentation for how that changes the stability and delivery profile.

Q: For NMPA registration, can we use the clinical data we generated in Europe?

A: In principle, yes — NMPA accepts foreign clinical data. In practice, the scrutiny is high and the acceptance rate for foreign studies without a Chinese co-investigator is low. For most of our brand partners targeting China, we recommend running a parallel China-specific study at a NMPA-recognized CRO. Budget 6–9 months for that process.

Q: Our formula works perfectly at lab scale. Why are you asking for an intermediate scale-up trial before full production?

A: Because anhydrous systems fail at scale in ways aqueous emulsions do not. Mixing geometry, temperature ramp rate, and addition sequence all behave differently at 150 kg versus 2 kg. We lost a full production batch of a niacinamide-peptide anhydrous serum to crystallization at scale — the lab batch was perfect. The intermediate trial at 20 kg costs a fraction of a failed production run and it is now a non-negotiable step in our waterless production workflow.

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