The Certificate of Analysis is the single most important document a buyer receives with a batch of herbal extract - and one of the most misread. It is the evidence that the material in the drum matches what was promised on the specification. This guide walks through a COA line by line, explains what each parameter means, which test methods to trust, and the quiet red flags that separate a serious supplier from a risky one.
1. What a COA Is - and What It Is Not
A Certificate of Analysis (COA) is a batch-specific record of the tests performed on one particular lot of material, showing the results measured against the agreed specification. It is not the same as a specification sheet, which simply states the target limits and is identical for every batch. A supplier can send the same specification a thousand times; a COA is a fingerprint of one manufacturing run.
The practical rule for buyers is simple: always ask for the batch COA, tied to the exact lot you are being offered or shipped. A specification sheet tells you what a product should be; the COA tells you what this batch actually is.
Quick check: If a document has no batch or lot number and no manufacturing date, it is a specification sheet, not a COA - regardless of what the title says.
2. The Identification Block
Every credible COA opens with a header that identifies precisely what was tested. Before you look at a single result, confirm these fields are present and consistent with your order:
- Product name and botanical (Latin) name - e.g. Ashwagandha extract, Withania somnifera. The Latin binomial removes ambiguity between look-alike species.
- Plant part - root, leaf, fruit, seed or whole herb. Root and leaf extracts of the same plant are different products with different actives.
- Batch / lot number - the anchor that links this certificate to physical drums.
- Manufacturing date and retest/expiry date - and the quantity the batch represents.
- Specification reference and issue date - which spec version the results are judged against.
3. Organoleptic & Identity Parameters
The first results usually cover appearance, colour, odour and taste, followed by identification tests. Identity is the assurance that the powder is what it claims to be - typically confirmed by TLC (thin-layer chromatography), HPLC fingerprint, or a positive chemical test for a characteristic constituent. A COA that reports assay figures but skips identity is telling you how much of something is present without confirming it is the right something.
4. Assay - the Number Everyone Looks At
The assay states the standardised active content - for example "Withanolides 5% by HPLC" or "Curcuminoids 95% by HPLC". This is the figure buyers negotiate on, and it is also the one most open to inflation, so the method matters as much as the number:
| Method | What it measures | Note for buyers |
|---|---|---|
| HPLC | Individual marker compounds, separated and quantified | Most specific; harder to inflate; preferred where a validated method exists |
| UV-Vis spectrophotometry | A broad class of compounds absorbing at one wavelength | Can read higher than HPLC; acceptable for some totals but less specific |
| Gravimetric | Total weight of a fraction (e.g. total saponins) | Measures a broad group, not a single marker; useful but non-selective |
| Titration | Chemical reaction to an endpoint | Traditional; may over-read versus a specific HPLC marker |
When comparing two suppliers' extracts, make sure you are comparing the same marker measured by the same method. A "40% saponins by gravimetric" and a "40% saponins by HPLC" are not the same claim, and the gap can be substantial.
5. Physico-Chemical Parameters
These describe the physical behaviour and purity of the powder, and they matter for formulation and stability as much as for quality:
- Loss on drying / moisture - high moisture invites microbial growth and dilutes actives; typically capped around 5%.
- Total ash and acid-insoluble ash - a proxy for mineral content and, importantly, for adulteration with sand or soil.
- Bulk and tapped density - affects capsule fill and blending.
- Particle size / mesh - e.g. 100% through 80 mesh.
- Solubility and pH - relevant for beverages and liquid formats.
- Extraction solvent and carrier - which solvent was used, and whether a carrier such as maltodextrin is present (which dilutes the native extract).
6. Residual Solvents
If the extract was made with an organic solvent (ethanol, methanol, acetone, hexane), the COA should report residual solvent levels, tested by gas chromatography against the ICH Q3C / pharmacopoeial classes. Class 1 solvents (such as benzene) should be absent; Class 2 and 3 solvents have defined limits. Water and hydro-alcoholic extracts carry the least residual-solvent risk. A missing residual-solvent line on a solvent-extracted product is a gap worth challenging.
7. Contaminants: Heavy Metals, Pesticides & Mycotoxins
This is where a COA proves the material is safe, not just potent. For botanical extracts destined for regulated markets, expect:
| Contaminant group | Typical test | What to confirm |
|---|---|---|
| Heavy metals (Pb, Cd, Hg, As) | ICP-MS | Individual element limits stated, not just "complies" |
| Pesticide residues | Multi-residue GC-MS/MS & LC-MS/MS | Screen against the destination market's MRLs (e.g. EU) |
| Aflatoxins & Ochratoxin A | HPLC-FLD / LC-MS | Essential for dried herbs and root materials |
| Ethylene oxide (ETO) | GC-MS (EO + 2-chloroethanol) | Critical for EU-bound botanicals; should be at LOQ |
Numeric values beat the word "complies" every time. "Lead: 0.42 ppm (limit 3.0)" is verifiable; "Heavy metals: complies" tells you nothing about the margin. For high-value or first orders, this is exactly the data worth confirming independently through a third-party laboratory such as Eurofins.
8. Microbiological Panel
The micro section confirms the material is free from harmful organisms and within total-count limits. A complete panel reports total plate count, yeast and mould, Enterobacteriaceae, and the absence of Salmonella, E. coli and often Staphylococcus aureus. Because ethylene oxide is banned as a sterilant in the EU, EU-bound material should reach these targets through validated steam or other non-ETO treatment - and the COA should make clear the micro figures were achieved without ETO.
9. The Signature - and Why It Matters
A COA should close with the name, title and signature of the quality authority who released the batch, plus the issue date. An unsigned, undated certificate carries little weight. Where testing was outsourced, the accredited lab's report (ideally NABL/ISO 17025 accredited) should be available on request to back the summary figures.
10. Red Flags on a COA
- No batch number, or a batch number that does not match the drums.
- Assay stated without the test method, or a method that flatters the number.
- "Complies" everywhere instead of measured numeric values.
- Identical results across supposedly different batches (a tell-tale of a template being reused).
- Missing contaminant lines - especially pesticides, ETO or heavy metals - for a botanical.
- No signature, no issue date, or a specification version that predates the batch.
This guide is general information for B2B buyers and does not constitute regulatory or analytical advice. Required parameters and limits vary by product, destination market and end-use; always confirm the specification and testing appropriate to your application.
Frequently Asked Questions
What is a Certificate of Analysis (COA)?
A Certificate of Analysis is a batch-specific document issued by a supplier that lists the tested properties of a material against its specification - identity, assay, physical parameters, contaminants and microbiology - together with the test methods used and the results obtained for that particular lot.
What is the difference between a COA and a specification sheet?
A specification sheet defines the acceptable limits a product must meet and is the same for every batch. A COA reports the actual measured results for one specific batch against those limits. You should always request the batch COA, not just the generic specification.
Which assay method is best for an herbal extract, HPLC or UV?
HPLC separates and quantifies individual marker compounds and is generally more specific, while UV-spectrophotometry measures a broader class of compounds and can read higher. Where a marker can be measured by HPLC, an HPLC assay is usually the more reliable and less easily inflated figure.
What contaminant tests should appear on a COA for a botanical extract?
A complete COA should report heavy metals (lead, cadmium, mercury, arsenic by ICP-MS), a pesticide multi-residue screen, aflatoxins and ochratoxin A for dried botanicals, residual solvents where solvent extraction is used, and a microbiological panel including total plate count, yeast and mould, and absence of Salmonella and E. coli.
How do I know a COA is genuine?
A trustworthy COA is batch-specific with a lot number and manufacturing/expiry dates, names the test method for every parameter, is signed by a named quality authority, and reports real numeric values rather than only "complies". For high-value orders, independent verification of key parameters by a third-party lab such as Eurofins removes any doubt.
Request a Sample COA from SV Botanica
Batch-specific certificates · Named test methods · Contaminant & micro panels · Third-party verification welcome