How to Read a Certificate of Analysis a Reference Standard CoA, Section by Section
A certificate of analysis is usually filed and forgotten. That is a shame, because almost every avoidable error traceable to a reference standard — a wrong potency correction, a superseded lot, a weighing below the minimum intake — is visible on the certificate before the vial is ever opened. Here is what each section actually obliges you to do.
1. Identity and lot
Name, catalogue number, lot or batch number, CAS number, molecular formula and weight, and the physical form supplied. Two checks: the lot on the certificate matches the vial in your hand, and the salt form matches what your calculation assumes. A standard supplied as a hydrochloride where your method works in free base needs a stoichiometric conversion factor - EDQM prints these on the leaflet where they apply, and they are easy to miss.
2. Intended use
The most under-read line on the document. Reference standards are established for a defined purpose: identification, system suitability, peak identification, or assay. A standard qualified for identification only carries no assigned value and cannot be used for quantitation.
USP is explicit that where no assigned value appears on the label or certificate, that lot was developed for qualitative compendial use. EDQM goes further: where a Ph. Eur. leaflet states no content, the value is simply not needed for the monograph procedure — and it must not be assumed to be 100%. The one exception is an impurity CRS used in a related-substances test.
3. Assigned value and its basis
Where a value is given, the basis matters as much as the number. "As-is", "on the anhydrous basis", "on the dried basis" and "anhydrous and solvent-free" all imply different weighings. USP labels as-is wherever the data support it, but some lots still require drying or a concurrent moisture correction, so read the label and the monograph rather than relying on habit.
A certificate reporting 99.2% on the anhydrous, solvent-free basis, with 0.35% water and 0.10% residual solvent, gives an as-is potency of 99.2 × 0.9955 = 98.75%.
Weigh 50.00 mg and you have 49.37 mg of analyte. Using 99.2% directly overstates the standard by 0.45% — material against a 98.0–102.0% specification.
4. Uncertainty
A genuine certified reference material states an expanded uncertainty with its coverage factor, usually k = 2 for roughly 95% confidence. That figure bundles characterisation, between-bottle homogeneity, and stability across the validity period. Materials sold as "reference standards" with a purity but no uncertainty statement are not CRMs, whatever the catalogue says. Using one does not make your uncertainty zero; it makes it unquantified.
5. How the value was assigned
Mass balance subtracts everything measurable — chromatographic impurities, water, residual solvents, inorganic residue — from 100%. Quantitative NMR measures the analyte directly against a certified internal standard. The two do not always agree, and mass balance is the more optimistic of the pair whenever an impurity is chromatographically invisible. A certificate that names its method lets you judge that. One that states "99.5%" and nothing else does not.
6. The supporting test data
Below the headline value sits the evidence: chromatographic purity, water by Karl Fischer, residual solvents by headspace GC, residue on ignition, and identity by IR, NMR or MS. Read this section for internal consistency rather than reassurance. If the certificate claims 99.8% but reports 0.6% water, the basis is doing work you need to understand. The identity data is also your comparator if a spectrum from your own laboratory later looks wrong.
7. Traceability and accreditation
Look for an explicit traceability statement — to the SI, to a NIST SRM, or to a pharmacopoeial standard — and for ISO 17034 accreditation of the producer, with the accreditation body and scope number named. Testing performed under ISO/IEC 17025 is a separate claim and does not by itself make the material a CRM. For pharmacopoeial standards, the establishment process is the traceability.
8. Dates and the trap
Commercial CRMs carry an expiry or retest date. Pharmacopoeial standards frequently do not, and this is where laboratories get caught. USP operates an official lot system with no printed expiry: the current lot, and any previous lot still within its valid use date, are published in an online catalogue updated daily. EDQM issues a Batch Validity Statement intended to be downloaded at the time of use.
In both cases, the certificate sitting in your file is not evidence that the lot is still current. Checking the catalogue on the day of analysis is the control, and it is the one most often missing from an SOP.
9. Storage, handling and minimum sample intake
Storage temperature, light protection and desiccation conditions, plus two instructions that routinely get skipped. Let a refrigerated container reach ambient temperature before opening, or condensation will inflate the weighing. And treat Ph. Eur. standards as single-use: once the container is opened the entire content is intended for immediate use, and further storage is not warranted.
Where a certificate specifies a minimum sample intake, weighing below it voids the homogeneity claim that underpins the assigned value. The number is not a suggestion.
The five-minute read
A CoA is a set of instructions, not a record. Confirm the intended use, the basis, the uncertainty and the lot validity before the first weighing; transcribe the assigned value into your system with a second-person check; and re-verify lot status on the day of analysis. Most standard-related deviations are caught in that five minutes, or not at all.
