What the COA for a curcumin batch actually contains

The one-page sheet

This is the structure of the COA that accompanies each batch. The parameter list is fixed; the results change. Anything your QA team needs that is not on this list can usually be added as a separate test — tell us before the batch is packed, not after.

Batch results are recorded on the COA supplied with each shipment. A sample COA for the current batch is available on request.

What "95%" is measured against

"95% curcumin" is one of the most quietly ambiguous specifications in the botanical trade. It can mean two different materials.

• 95% total curcuminoids — the sum of curcumin, demethoxycurcumin and bisdemethoxycurcumin by HPLC. This is the standard reading, and the one our specification uses.
• 95% curcumin — the curcumin peak alone at 95%, with the other two curcuminoids correspondingly lower. The total is still around 95–97%, but the colour load per gram of "curcumin" is not the same.

In practice, a natural 95% curcuminoid extract sits at roughly 70–75% curcumin, 15–20% demethoxycurcumin and 3–5% bisdemethoxycurcumin. If a supplier quotes a profile outside that band, ask whether the material is natural or standardized by blending — that is a legitimate question, and it changes the colour load in your formula.

The method column

A number without a method is not a specification. Two laboratories can differ by one to two percentage points on the same powder because of column, mobile phase, wavelength and standard. When you compare quotations, compare the method as well as the number.

• HPLC — C18 column, acetonitrile/water with acid, detection around 425 nm, calibrated against a curcuminoid reference standard. USP and EP monographs describe the approach; the exact conditions belong in the TDS.
• ICP-MS — lead, cadmium, mercury and arsenic, after microwave digestion. This is the method behind the heavy-metal numbers on the COA.
• GC-MS/MS and LC-MS/MS — the pesticide panel. Ask how many compounds and against which MRL list; "pesticides: conforms" without a panel size is not auditable.
• GC-HS — residual solvents, headspace. Ethanol and ethyl acetate are Class 3 under ICH Q3C; if your process uses hexane, that is a Class 2 solvent with a much lower limit and should be stated.

Release limit vs batch result

The release limit is the window we promise. The batch result is what this specific drum measured. Both belong on the COA, and the difference between them is where most purchasing arguments start.

If your incoming inspection uses a tighter window than the release limit, send us that window before the order. It is much easier to select or adjust a batch before packing than to argue about a deviation after arrival.

What changes between batches

• Colour — from yellow-orange to reddish-orange, driven by the curcuminoid ratio and particle size. If colour is critical (food, cosmetics), specify a range and ask for a retention sample.
• Assay overshoot — a 95% specification often lands at 96–97%. That is normal; it is not a premium grade.
• Loss on drying — 1–3% is common; above 4% is worth a question about storage before shipment.
• Particle size distribution — the same "80 mesh" can hide a wide distribution. Ask for D10/D50/D90 if flow or dispersion matters.
• Pesticide results — depend on the rhizome lot and the season, not on the extraction process. This is the parameter most likely to fail an EU import check.

Third-party testing

We recommend independent testing for the first shipment into a new market, and for any batch that will be audited by your customer. The usual panels:

Turnaround depends on the panel and the laboratory's queue; we confirm it in the quotation rather than quoting a generic number here.

Matching your specification

To confirm that a batch will meet your incoming specification, send us four things:

1. The destination market and the final dosage form — the limits that apply depend on both.
2. Your internal specification, or your customer's COA template if you have one.
3. Any parameter where your limit is tighter than the release limit above.
4. The quantity and the date you need it — some tests add days to the lead time.

We reply with the current batch COA, the TDS and a statement of which parameters can be met as-is and which need a dedicated batch.