Draft — awaiting QA contact, page not indexed
How to Read a Certificate of Analysis for a Botanical Extract
A COA is the document that tells you whether a batch matches what you specified. What each section means, which figures matter, and the gaps that should stop a purchase order.
Before this can be published
- consider adding an annotated sample COA image using a real Alpha Nova certificate with the batch number redacted. This would substantially strengthen the piece and is the kind of asset that attracts links.
A certificate of analysis is a batch document. That is the first thing to understand about it, and the thing most often missed. A COA does not tell you what a supplier is generally capable of. It tells you what one specific lot tested at, on one specific date, by one specific method.
Which means a COA is only useful if you read it against something: your specification, or the supplier's own published specification. A COA on its own, with nothing to compare it to, is a page of numbers.
Here is what each part of it is doing.
The identification block
Product name, batch or lot number, manufacture date, expiry or retest date, and quantity. Boring, and the section where problems show up first.
Check that the batch number on the certificate matches the batch number on the drums that arrived. It sounds trivial. It is the single most common discrepancy in practice, and if they do not match you do not have a COA for your material, you have a COA for somebody else's.
Check the date logic too. A manufacture date after the shipping date, or a retest date that does not correspond to the stated shelf life, usually means the certificate was generated from a template rather than from the batch record.
Identity and assay
This is the section that answers whether the material is what it claims to be, and how much of the active constituent it contains.
For a standardised extract, the assay is the number you specified against. A liquorice extract standardised to a glycyrrhizin percentage, a green tea extract standardised to EGCG, a hibiscus extract standardised to anthocyanin content: the certificate should state the measured value, the specification range it is being measured against, and the method used.
That last element is where buyers get caught out. The same material assayed by different methods can produce different numbers, and a figure quoted without its method is not comparable to anything. If a certificate says a value but not how it was obtained, ask. A supplier who cannot tell you the method either did not run it or does not know what their contract lab ran.
Watch for a result quoted as a single value where you specified a range, and for a result sitting exactly on a specification limit batch after batch. Real analytical data scatters. Data that does not scatter is worth a question.
Physical characteristics
Appearance, colour, odour, solubility, particle size or mesh, loss on drying or moisture content, and bulk density where relevant.
Moisture is the one to read carefully. It affects the actual quantity of active material you are paying for, it affects microbiological stability in storage, and it is the specification most likely to drift between a sample and a production lot. For a hygroscopic powder, the moisture figure at the factory and the moisture figure after six weeks in a humid port are not the same number.
Particle size matters more than it appears to if the material is going into a process with defined dissolution or dispersion behaviour. A material that meets every chemical specification and fails on mesh will still cause problems on your line.
Purity and contaminants
Heavy metals, usually lead, arsenic, cadmium and mercury. Residual solvents where the extraction used them. Pesticide residues. Sometimes mycotoxins, particularly aflatoxins, for materials of plant origin from certain regions.
Two things to check. First, that limits are stated and not only results, so you can see what standard is being applied. Second, that the limits actually correspond to the regulation of the market you are selling into, because permitted levels differ between jurisdictions and a certificate written against one standard tells you nothing about compliance with another.
"Complies" as a result, with no figure, is weaker than a number. Sometimes it is legitimate, where a method is pass or fail. Often it means nobody measured.
Microbiological results
Total plate count, yeasts and moulds, and specified pathogens: typically E. coli, Salmonella, Staphylococcus aureus. For a botanical material these are not a formality. Plant material carries a natural microbial load and the extraction process does not always eliminate it.
Look at whether the testing is on the finished batch or carried over from an earlier stage, and at the sample size the pathogen result is based on. Absence in 25 g and absence in 10 g are different assurances.
The signature and the lab
A COA should identify who performed the testing and who is taking responsibility for the certificate. In-house testing is normal and not a problem in itself. Third-party testing at an accredited laboratory carries more weight, and for a new supplier relationship it is reasonable to ask for third-party verification of at least the first shipment.
An unsigned certificate, or one with no named laboratory, is not a controlled document.
What is missing is as informative as what is present
The most useful habit when reviewing a COA is to read it against the specification and notice the absences. A parameter you specified that does not appear on the certificate has not been demonstrated, whatever the rest of the document shows.
Common gaps worth challenging: no method stated for the assay, no specification limits alongside results, contaminant results given as "complies" with no figure, microbiological results with no sample basis, and a certificate covering a parameter with a date that predates the batch.
COA, specification sheet and SDS are three different documents
They get conflated, and they answer different questions.
The specification sheet is what the supplier commits to for the product generally. It is the standard.
The certificate of analysis is what one batch actually measured. It is the evidence against that standard.
The safety data sheet covers hazard, handling, storage and transport. It is a safety document and it says nothing about quality or conformity.
You need all three, and a supplier who offers one in place of another is worth a second look.
Requesting documentation from Alpha Nova
Specification sheets and sample certificates of analysis for every product we supply are available without a form. If you are qualifying us as a supplier, ask for the specification, a representative COA, and the relevant certifications together, and compare them against whatever you are currently buying.
If something on one of our certificates is not clear, ask us what method produced it. That is a reasonable question and any supplier should be able to answer it.