Learn
Extract guide
An extract is not defined by the name of the plant. It is defined by five data points that have to travel together, and by a test method that decides what the number means. This guide walks through them in the order you have to fix them.
What you are actually buying
«Turmeric extract» is not a specification: it is a family of products that can differ by a factor of thirty. For two people to be talking about the same material you need five data points, always all five.
Binomial name of the plant — Curcuma longa, not «turmeric» —, plant part, analytical marker, percentage of that marker and the method it is measured with. Change any one of the five and you change the material, the price and often what you are allowed to claim on the label.
The plant part is forgotten more often than you would expect, and it is not a detail: the root, the leaf and the fruit of the same species have different profiles. An ivy leaf extract and a stem extract are not substitutes even though the plant is the same.
Ratio or standardised
There are two ways to declare the concentration of an extract, and they serve different purposes.
The ratio — 4:1, 10:1 — says how many kilos of dried plant went into one kilo of extract. It describes the process, not the result: a 10:1 from a poor harvest can carry less active than a 4:1 from a good one. It is the traditional form and it remains valid when what you want is the whole profile of the plant.
Standardisation fixes the destination instead of the route: 95 % curcuminoids, 50 % EGCG, 80 % silymarin. The lot is adjusted to that figure, measured and certified. That is what you need when the label declares an amount of active, when the dose has to fit inside a capsule or when the product is registered with a health authority.
If your formulation declares milligrams of a compound, you need a standardised extract. A ratio does not let you derive that figure.
The test method changes the number
This is where most money is lost comparing offers, because two correct numbers can be incomparable.
Milk thistle illustrates it well: the same reference can be declared as 80 % silymarin by UV-VIS or as 30 % silybin by HPLC. Both values are true and describe the same material. UV measures a set of compounds absorbing in the same region; HPLC separates and quantifies one specific compound. A supplier offering «80 %» and one offering «30 %» may be selling exactly the same thing.
That is why a percentage is never accepted on its own. It always travels with its method, and when comparing two offers the first move is to bring both to the same method. If one of them does not declare it, there is no comparison: there is an assumption.
Solvent and residues
Which solvent was used decides what came along with it and which controls you have to demand afterwards.
Water and ethanol are class 3 solvents in the ICH classification: low toxicity, wide limits and accepted in practically every market. Acetone, methanol and hexane are class 2, with strict limits, and their use makes a residual solvent test mandatory on every lot.
Supercritical CO2 leaves no residue — it evaporates — and gives extracts clean of the lipid fraction, but it does not carry polar compounds well: for water-soluble polyphenols it is the wrong technique.
Always ask for the residual solvent declaration of the lot, not of the process. The process describes the intention; the lot describes what arrived.
Carrier, solubility and format
Almost no powdered extract is one hundred per cent extract. The active fraction is dried onto a carrier — maltodextrin, gum arabic, silicon dioxide, sometimes the milled plant itself — that gives it flow, stability and the ability to be dried at all.
The carrier is not hidden filler: it is declared on the sheet and it has to be counted. It affects capsule weight, carbohydrate load and, in some markets, what appears on the ingredient list.
Solubility is the other figure that decides whether a formula works. A fat-soluble extract in a clear beverage clouds the product no matter how good the specification is. Water-soluble grades exist precisely for that, and they usually carry a lower marker percentage: part of the weight is what makes them soluble.
The minimum documentation
Four documents. If one is missing, so is information that sooner or later will be asked of you.
Technical data sheet with the full specification and its limits. Certificate of analysis of the specific lot you receive, with real results and not with the limits copied across. Safety data sheet. And declarations of residual solvents, allergens, GMO and origin.
The certificate of analysis is the most forged and the easiest to verify: it must carry a lot number, analysis date, methods and numeric results. A certificate that repeats the specification limits in the results column is not a certificate, it is a cover page.
How to compare two offers that look identical
- Normalise the method: bring both percentages to the same assay before you look at the price.
- Check that the marker is the same compound, not one from the same family.
- Discount the carrier: compare active fraction against active fraction.
- Work out the cost per kilo of marker, not per kilo of extract. It is the only comparable number.
- Verify that the certificate of analysis belongs to the lot they are going to ship you.
- Test solubility against your real format, not against the data sheet.
- Ask about lot-to-lot variation over the last twelve months. The average matters less than the spread.
This guide describes how we work and how we read a specification. It does not replace the regulation of the market where you register the product, which always prevails.
Browse the catalogue and its data sheets