How to read a certificate of analysis for a research reagent

A certificate of analysis (CoA) is the document that ties a specific lot of a research reagent to the tests that were run on it. It is easy to glance at the purity figure, see a pass, and file it. That habit misses most of what the document is for.

The practical question is simple: when a CoA lands on your desk, what do you read, in what order, and what should make you stop and ask the issuer a question? This guide walks through the parts of a typical certificate from top to bottom: the header block, the lot number, the test methods, the specifications and results table, and the signature block.

The header block: who issued it, and about what

The header tells you whose document this is and which material it describes. Before reading any numbers, confirm each of these fields:

• Issuing organization. The name and address of the laboratory or supplier that produced the certificate. If testing was done by an outside laboratory, the certificate should say so, and the testing laboratory should be named.
• Material description. The name of the reference material as it appears on the vial label and on the order, with any catalog number. The two should match exactly, character for character.
• Document identifier. Most certificates carry their own number, separate from the lot number. This is the reference you quote when you contact the issuer.
• Dates. The date of manufacture or fill, the date of analysis, the date the certificate was issued, and a retest or expiry date. The analysis date should fall after manufacture and before issue.
• Quantity or format. Fill amount per vial, physical form (for example, a lyophilized research peptide as a white powder), and container type.

A mismatch at this stage is the cheapest one to catch. If the catalog number on the certificate belongs to a different item, nothing further down the page applies to the vial in your hand.

The lot number: the thread that holds the record together

The lot (or batch) number identifies one production run. It is the key that connects the certificate, the vial label, the shipping paperwork, your receiving log and any retained sample. Everything else on the certificate is only meaningful because of it.

Three checks cover most problems:

1. Read the lot number off the vial, then off the certificate. Compare them directly, and record the lot in your receiving log at that moment.
2. Do not confuse the certificate number with the lot number. Some issuers print both in similar-looking fields near the top of the page. The certificate number is unique to each document, which makes it look like a reliable identifier, but it identifies the paperwork, not the material. If you file vials under the certificate number, two certificates for one lot will look like two lots, and one lot shared across several catalog items will go unnoticed.
3. Watch for one lot number across unrelated items. If certificates for different materials carry the same lot number, ask the issuer why. It may be a shared fill run with a valid explanation, or it may be a transcription error. Either way, it is a question worth asking before the material goes into use at the bench.

Test methods: how each number was produced

A result is only as useful as the method behind it. A good certificate names, for each test, the technique used and a reference to the method document. Common entries on a certificate for a research reagent include:

• Identity, often by mass spectrometry, reporting an observed molecular mass against the theoretical value.
• Purity, often by HPLC with UV detection, reported as a percentage of total peak area at a stated wavelength.
• Appearance, a visual description of color and form.
• Water content, often by Karl Fischer titration, and sometimes counterion or residual solvent content.

Look for enough detail to know what the number means. "Purity: 98.7%" without a method tells you little. "Purity by HPLC, area percent at 220 nm, method QC-012" tells you what was measured and gives you something to ask about. Area-percent purity describes how the analyte compares with other UV-absorbing species in the chromatogram; it is not the same thing as net content, which accounts for water, counterions and other non-absorbing material. A certificate that reports both is more informative than one that reports only the first.

Method references also let you judge whether the method is fit for purpose. The ICH Q2 guideline describes what validating an analytical procedure involves (specificity, accuracy, precision, linearity, range and related characteristics). You do not need the validation report to read a certificate, but you should be able to request it, and the issuer should know which method version was used.

Specifications versus results: the table that matters most

The core of the certificate is usually a table with three columns: the test, the specification (sometimes called the acceptance criterion), and the result for this lot. These two columns answer different questions, and reading them together is the whole point.

A few patterns deserve a second look:

• Specification equals result. If the specification column reads "98.7%" and the result reads "98.7%", the limit was probably written after the fact. A specification should be a range or a threshold.
• Results reported only as "pass" or "conforms". This is normal for appearance, but a quantitative test should report the measured value.
• Rounding that crosses a limit. A result of 97.96% reported as 98.0% against "not less than 98.0%" depends on the rounding rule. Ask what convention the issuer follows.
• Identical results across lots. Separate lots measured separately rarely produce the same figure to the last decimal on every test. Identical tables across certificates can mean a template was reused.

Where the certificate attaches raw data, such as a chromatogram or a mass spectrum, compare it with the table. The main peak should dominate the chromatogram, and the observed mass on the spectrum should match the value in the table. Raw pages often settle questions a summary table leaves open.

The signature block: who stands behind it

The bottom of the certificate states who reviewed and released the data. Look for the printed name and role of the person approving the certificate (often a QC or QA reviewer), a signature or an equivalent electronic approval, and the date of approval.

A certificate with no named reviewer is a data sheet, not a certificate. If the testing laboratory holds ISO/IEC 17025 accreditation, the certificate may carry the accreditation body's mark and a certificate or scope number. Accreditation applies to specific tests listed on the laboratory's scope, so a mark on the page does not mean every test on that page is covered. The accreditation body publishes the scope, and you can check the relevant test against it.

Also note any statement of intended use. A certificate for a research reagent should describe the material as for laboratory research use only, consistent with the label and the product page.

A reading checklist for incoming certificates

Use this at receiving, before a lot is released to the bench, and keep the completed record with the certificate:

1. Issuer, material name and catalog number match the order and the vial label.
2. Lot number on the certificate matches the vial, and is recorded as the lot, not the certificate number.
3. Manufacture, analysis, issue and retest dates are in a sensible order and current.
4. Each test names a method, and quantitative tests report a measured value with units.
5. Specifications are real limits, and every result falls inside its limit.
6. Any attached chromatograms or spectra agree with the table.
7. A named reviewer has signed and dated the certificate, and any accreditation mark covers the tests shown.
8. Questions for the issuer are logged with the certificate number, and the lot is held until they are answered.

For more on keeping these records usable over time, see the other posts on the ruogate blog.