Reference standard, analytical standard, CRM: what the words mean

The same vial, four different labels

Open four catalogues and the same compound is sold as a reference standard, an analytical standard, a certified reference material and a working standard. The material may even be the same lot, repackaged. The words are not synonyms, and the difference is not marketing: each one implies a different level in a measurement hierarchy, a different set of documents, and a different claim you are entitled to make about a result obtained with it.

The confusion is worst between reference standard and analytical standard, because both are commercial labels rather than defined terms. The two that are defined are reference material and certified reference material, and everything else is a role someone is playing in a particular laboratory.

What each term actually commits to

• Reference material. A material that is sufficiently homogeneous and stable with respect to one or more specified properties, and that has been established as fit for its intended use in a measurement. That is the whole definition. It says nothing about how the property value was assigned, and nothing about uncertainty. A house standard characterised in a laboratory is a reference material.
• Certified reference material. A reference material with documentation from an authoritative body, giving one or more property values with an associated uncertainty and a stated traceability, established by valid procedures. The certification is a claim about the measurement, not about the packaging: it says the value is traceable and that its uncertainty was evaluated. Producers are assessed against ISO 17034.
• Pharmacopoeial reference standard. A standard established by a pharmacopoeia for use in the tests of its own texts. Its fitness is defined by those tests, and using it outside them is a judgement call, not a default. It occupies the role of an official standard rather than a commercially certified one.
• Analytical standard. A commercial description, not a metrological category. It usually means material intended for instrument calibration or peak identification, and it may be a reference material with a full characterisation or a neat compound with a purity figure. The label alone tells you nothing; the certificate does.
• Working standard. A laboratory's own secondary standard, calibrated against a primary or against a pharmacopoeial standard and then used day to day. Useful, necessary, and only as good as the calibration that created it and the re-calibration that maintains it.

Why the hierarchy matters more than the label

A measurement result inherits its uncertainty from the standard used to calibrate the instrument, along with everything else in the chain. When a laboratory reports an assay of 99.2 percent, the number rests on a value assigned to some material by some route, and that route is the traceability chain. A standard with no stated assignment route breaks the chain: the result still exists, but it can no longer be traced to anything, and in an audit that is the difference between a defensible figure and an assertion.

This is why two standards that look identical on the bench behave differently in a dossier. The one with a stated characterisation, an assigned content and its basis, homogeneity and stability data, and an uncertainty can support a claim. The one described only as "purity 98 percent" cannot, because area percent from a chromatogram is a measurement of the material's own impurity profile, not an assignment of content against a traceable reference.

Where each one is the right tool

• System suitability and peak identification: a reference material is enough. You need the peak to appear at the right time and the resolution to be demonstrated, not a certified content.
• Assay by external standard: the content value carries the result, so the assignment route and its uncertainty matter. A certified reference material, or a pharmacopoeial standard used inside its compendial scope, is the appropriate choice.
• Impurity limits near the reporting threshold: the standard's own purity limits what can be demonstrated. A standard containing the analyte at the level you are trying to control cannot support the method.
• Routine release testing: a working standard calibrated against the primary, with a defined re-calibration interval, is both adequate and economical. The risk is drift: a working standard that has never been re-checked gradually becomes an assumption.

Reading a supplier's claim without being misled

Four phrases are worth translating before a purchase order is raised.

• "Purity 99 percent" with no method. Almost always a chromatographic area percent. It describes the material as it eluted, and it is not a content assignment.
• "Certified" with no uncertainty and no traceability statement. Certification is a claim about a value; a certificate that does not state the value with its uncertainty is using the word decoratively.
• "Traceable to USP" on a non-pharmacopoeial product. This usually means the material was compared against an official standard at some point, which describes a calibration step, not a certified value.
• "For laboratory use only" combined with a content figure. It may be entirely honest material; the disclaimer tells you the producer is not standing behind a metrological claim.

Two certificates, same compound

A worked comparison makes the vocabulary concrete. Two suppliers offer the same impurity, both at a stated 98 percent.

The first certificate gives identity by proton and carbon NMR with the assigned shifts listed, a high-resolution mass with the calculated and found values side by side, a chromatographic purity of 98.4 area percent with the method and column named, water content by Karl Fischer, residual solvents by headspace gas chromatography, an assigned content of 97.1 percent by qNMR against a traceable calibrant with an expanded uncertainty, a homogeneity check across ten units, and a retest period supported by stability data. That is a certified material in everything but the word.

The second gives a chromatographic purity of 98.2 area percent, one method, no water content, no residual solvents, no assignment of content, and no uncertainty. It is a legitimate product and it will make a peak appear at the right retention time. It is not a calibrant for an assay, because nothing on the document establishes what the other 1.8 percent is or what the material weighs beyond its own peak area.

Both are called reference standards in their catalogues. The difference is not the number, which is nearly identical, but the presence of an assignment route with an uncertainty. That is the whole distinction the vocabulary is trying to carry, and it is the one worth checking before the purchase order rather than after the audit.

What to ask

• Which of these is it: reference material, certified reference material, or a pharmacopoeial standard, and under what scope?
• How was the content value assigned, and what is the uncertainty on it?
• What is the traceability statement, and to what does it trace?
• Was homogeneity assessed across the batch, and how?
• What is the retest or expiry basis, and what stability data supports it?
• If it is a working standard, against what was it calibrated and when?

The answers do not have to be elaborate. What matters is that each one has an answer, because a laboratory that cannot state the assignment route for its calibrant cannot state the uncertainty of its result. The certificate fields guide walks through where these answers should appear on the document, and the qNMR and mass balance guide covers the two assignment routes that most certificates rely on.