How I Formulated a Laundry Detergent That Actually Works Without the Toxins
How I Formulated a Laundry Detergent That Actually Works Without the Toxins
By Kristina Braly, MD — Founder, AEMBR
The hardest product I've formulated isn't the one that smells the best or looks the most beautiful on a shelf. It's laundry detergent.
Laundry detergent is a performance product. Nobody buys it for the experience of using it — they buy it because their clothes need to be clean. That means there's no aesthetic margin to hide behind. If it doesn't clean, it fails. And cleaning well without the ingredients I'd already ruled out — SLS, SLES, 1,4-dioxane-generating surfactants, optical brighteners, undisclosed fragrance, synthetic dyes — turned out to be a genuinely difficult formulation problem.
Here's how I worked through it.
Starting With the Exclusion List
I wrote about the ten ingredients I banned from all AEMBR products in an earlier piece. For laundry detergent specifically, the most consequential exclusions are:
- SLS and SLES — the conventional workhorse surfactants. SLS is a potent skin irritant that stays on fabric after rinsing; SLES carries 1,4-dioxane contamination risk.
- Optical brighteners — purely cosmetic, persist on fabric against skin, not biodegradable.
- Synthetic fragrance with undisclosed components — the fragrance exemption shields exactly the ingredients I care most about evaluating.
- Phosphates — highly effective water softeners, but devastating to aquatic ecosystems and now banned in laundry detergent in most US states.
With those off the table, I needed to find surfactants that cleaned well, enzymes that worked at lower temperatures, a fragrance approach that was both effective and fully disclosable, and a format that made the formula stable without the preservative and filler infrastructure most conventional liquid detergents require.
Surfactant Selection: The Hardest Part
Surfactants are the active cleaning agents in any detergent — the molecules that reduce water's surface tension, lift soil from fabric, and keep it suspended in the wash water so it rinses away. Choosing the right surfactant blend for a clean-formulated detergent means finding ingredients that:
- Clean effectively across fabric types and soil loads
- Rinse fully — no residue on skin-contact fabric
- Are biodegradable
- Don't produce 1,4-dioxane as a manufacturing byproduct
- Aren't skin sensitizers at the contact concentrations that occur from residual detergent in fabric
The glucoside surfactants — decyl glucoside, coco glucoside, lauryl glucoside — checked most of those boxes. They're plant-derived, readily biodegradable, non-ethoxylated (so no 1,4-dioxane concern), and mild enough that residual levels in rinsed fabric are unlikely to cause irritation even in sensitive-skin populations.
The problem: glucoside surfactants alone don't produce the foam profile or the grease-cutting performance most people associate with laundry detergent. They're excellent on light-to-moderate soil loads. They struggle with heavy grease or protein-based stains without additional support.
The solution was a blended system: glucoside primary surfactants paired with sodium coco sulfate (SCS) at low concentration. SCS is structurally different from SLS — it's derived from whole coconut oil rather than isolated lauric acid, which produces a broader carbon chain distribution and a milder irritation profile. It's not completely equivalent to SLS in mildness, but it's meaningfully better, it's not ethoxylated, and at the concentrations needed to supplement glucoside performance it functions without the skin irritation pattern SLS produces.
It took several rounds of testing to land on the ratio. Too much SCS and the formula crossed back toward the irritation profile I was trying to move away from. Too little and the surfactant system underperformed on moderate-to-heavy soil. The working ratio is one I'm confident in — effective, mild, fully biodegradable.
Enzymes: The Performance Multiplier
Enzymes are what allow modern detergents to clean effectively at lower temperatures. Different enzyme classes target different soil types:
- Proteases break down protein-based stains: blood, sweat, grass, food.
- Amylases target starches: pasta, grains, baby food.
- Lipases address oils and fats: cooking grease, body oils, cosmetics.
- Cellulases work on cotton fiber surfaces, helping remove embedded soil and reducing fabric pilling over time.
Enzyme sourcing matters. The challenge isn't the enzyme class — it's the carrier and stabilizer system the enzyme supplier uses. Some commercial enzyme preparations include preservatives and stabilizers I wanted to avoid. Finding enzyme preparations with clean carrier systems that remained stable in the powder format required sourcing work before formulation work.
The AEMBR laundry formula uses a protease/amylase/lipase blend that performs across the soil categories most relevant to household use. The powder format actually helps here — enzymes are more stable in powder form than in liquid suspension, which means the formula retains enzymatic activity over the product's shelf life more reliably than liquid enzyme detergents typically do.
The Fragrance Problem
Laundry detergent fragrance is a specific challenge. The scent applied in the wash needs to survive mechanical agitation, hot water (in warm or hot wash cycles), and the rinse cycle — and still be perceptible on dry fabric. That requires fragrance compounds with specific volatility profiles and wash substantivity.
The fragrance compounds with the best wash substantivity — the ones most likely to survive and remain on fabric — tend to be the heavier, lower-volatility molecules. Some of these are synthetic musks. Musks have excellent substantivity, which is exactly why they're used so widely in laundry fragrance. They're also the category with the most documented environmental persistence and bioaccumulation concern.
I excluded synthetic polycyclic musks from the fragrance. That constrained the wash substantivity. The trade-off is that AEMBR laundry fragrance is lighter on dry fabric than detergents that use musk-heavy fragrance systems. I've chosen not to compensate with higher fragrance load, which would increase inhalation exposure during use. The result is a clean, present scent during the wash and a subtle, faded finish on dry fabric — not the heavy "freshness" most conventional detergents deliver via musk residue.
That's a deliberate formulation choice. I think it's the right one.
Format: Why Powder
Most premium and DTC laundry detergents launch as liquids. There are good marketing reasons for this — liquid is more intuitive to use, easier to pre-treat with, and visually elegant in a bottle. I chose powder anyway, for formulation reasons.
Liquid detergents require a significant water content — often 60–80% of the formula by weight is water. That water requires preservatives to prevent microbial growth, and it adds weight and packaging volume for what is essentially inert carrier. Powder concentrates the active ingredients without requiring preservative infrastructure. It's more stable over time, ships with a lower carbon footprint per load, and allows for a cleaner ingredient list because I'm not formulating around the constraints of liquid stability.
There's also an enzyme stability advantage I mentioned above. And a fragrance concentration advantage: because the formula is anhydrous (no water), the fragrance materials remain stable and consistent throughout the product's shelf life rather than slowly degrading in aqueous suspension.
The user experience trade-off — powder dissolves more slowly in cold water, some people find measuring less intuitive than a liquid cap — is real. I've addressed it with a fine-milled powder that dissolves well even in cold cycles. But I won't pretend the format choice was purely performance-driven. Powder was the right formulation decision, and the format convenience gap was a trade-off I made consciously.
What Didn't Work
Three versions of the formula preceded the current one.
Version 1 used glucoside surfactants exclusively, no SCS supplement. It cleaned light loads adequately. It failed on grass stains, cooking grease, and moderate protein stains. Not acceptable for a product I'd claim cleans effectively.
Version 2 added SCS at a higher concentration than the current formula. Performance improved significantly. Skin patch testing on sensitive-skin subjects showed mild irritation responses at the SCS level I'd used. I reduced the concentration until irritation responses cleared.
Version 3 had a fragrance approach that used higher concentrations of natural essential oil components to compensate for excluding synthetic musks. The result was a formula that smelled beautiful in the bag but produced significant top-note loss during the wash cycle — the high-volatility components evaporated during hot-water washing, and nothing was left on dry fabric. Reformulating the fragrance to use lower-volatility naturals with better wash substantivity produced the current scent profile.
The current formula is the fourth iteration. I'm confident in it. Not because it's perfect, but because I understand the trade-offs I made and they're trade-offs I'd make again.
What This Means for You
Understanding what's in laundry detergent and why those ingredients are there is the foundation for evaluating any formula — including ours. The AEMBR approach is: start with what we won't use, find alternatives that meet the performance bar, and be transparent about the trade-offs where they exist.
The AEMBR Laundry Powder is the product that came out of that process. If you have questions about specific formulation decisions, the full ingredient list is on the product page and I'll answer direct questions.
Further Reading
- Laundry Detergent Ingredients: What They Are and What They Do
- Physician-Formulated: What It Actually Means
Kristina Braly, MD, is the physician founder of AEMBR. Nothing in this article constitutes medical advice.
























































































































































































