Certified reference materials and ISO 17034

In chemical and physical testing, certified reference materials provide traceability where calibration of an instrument alone does not. ISO/IEC 17025 clause 6.5 recognises certified values of CRMs from a competent producer as a route to traceability; ISO 17034 defines that competence.

In temperature and humidity work

• Fixed-point cells (water triple point, gallium, indium, tin, zinc) function as physical references with certified values.
• Saturated salt solutions for humidity have tabulated equilibrium RH values with stated uncertainties; commercially certified versions exist.
• Certified thermometers are not CRMs — they are calibrated instruments.

Whatever the reference, the rule is the same: the certificate must state the value, its uncertainty and the traceability route.

What makes a material certified

The word certified is doing specific work. A reference material is a material sufficiently homogeneous and stable with respect to a property, suitable for its intended use. A certified reference material adds a certificate giving the value of that property, its uncertainty, and a statement of metrological traceability — produced under a procedure that establishes all three. ISO 17034 is the standard that producers are accredited against, and it is the mark that separates a CRM from a material sold with a nominal value.

The distinction matters because ISO/IEC 17025 clause 6.5 recognises certified values from a competent producer as a route to traceability. A material without a certified value, an uncertainty and a traceability route is not that route, whatever the supplier's literature calls it.

Choosing a producer, and what this site will not tell you

There is no list of suppliers here, and there will not be one. A catalogue of reference material producers with their scopes and prices would be out of date within months, and a wrong accreditation scope copied from a web page into somebody’s traceability file is exactly the kind of error that is discovered during an assessment rather than before it. What follows is how to check a producer yourself, which does not go out of date.

• Accreditation to ISO 17034 specifically, as a reference material producer. ISO/IEC 17025 for a testing or calibration laboratory is a different accreditation and does not cover producing reference materials.
• The scope of that accreditation, read for your material, your property and your range — not just for the producer’s name appearing on an accreditation body’s register. A producer accredited for one matrix is not thereby accredited for yours.
• Whether the accreditation body is an ILAC MRA signatory, which is what makes the certificate mean anything outside its own country.
• What the certificate of analysis actually states: the property value, its uncertainty and the coverage factor, the basis of the assigned value, the period of validity, and the storage and handling the value depends on.
• Whether the certificate names the measurement procedure used to assign the value, and whether that procedure is traceable to the SI or to an agreed reference.

The last one separates a certified reference material from a material that merely has a number on the label. If the value was assigned by a method the certificate will not name, the traceability chain you are relying on has an undocumented link in it, and an assessor asking you to demonstrate that chain will find the same gap you did not.

Reading a CRM certificate

• The certified value and its uncertainty, with a coverage factor. A value with no uncertainty cannot form a link in a traceability chain.
• The traceability route: how the producer established the value, and against what.
• The period of validity, and the storage and handling conditions the certificate assumes. A CRM stored outside them is no longer described by its certificate.
• Homogeneity and stability statements, including any between-unit variation for a batch — that variation is a component in your budget.
• The producer's ISO 17034 accreditation and scope. As with calibration, accreditation is per scope rather than per organisation.

Salt solutions are the case where people go wrong most often. The equilibrium humidity above a saturated salt depends on temperature, so the value is only as good as the temperature control of the chamber — and using a tabulated figure without recording the temperature it applied at is not traceability.

One practical consequence for a temperature laboratory: a fixed-point cell is the only reference on the bench whose value does not depend on someone else having calibrated it correctly. That is why a triple-point-of-water cell is worth its cost in a laboratory claiming millikelvin uncertainties, and why the ice point remains the workhorse everywhere else.