Identity versus purity: two different questions

A certificate of analysis usually reports several results, and two of them are easy to blur together. One answers the question “is this the substance named on the label?” The other answers “how much of what is in the container is that substance?” The first is identity. The second is purity, or, in a stricter sense, assay. They are measured by different techniques, they fail in different ways, and a passing result on one says nothing about the other.

Identity: is it the right thing?

An identity test compares the sample with a known reference. The comparison is usually spectroscopic or chromatographic. An infrared spectrum of the sample is laid over the spectrum of a reference standard, and the absorption bands must match. A mass spectrum shows whether the molecular mass and fragmentation pattern agree with the expected structure. In chromatography, the main peak in the sample must appear at the same retention time as the reference, run on the same system on the same day.

Identity tests are pass or fail. They are designed to tell the expected substance apart from plausible alternatives, which is why a single technique is not always enough. Two compounds can share a retention time on one column. Closely related structures can give similar infrared spectra. For that reason, specifications often call for two independent identity tests, and a well-designed identity test is one whose specificity has been demonstrated against the substances it most needs to exclude.

Purity: how much of it is there?

Purity is a quantity, not a yes or no. The most familiar way to report it is chromatographic purity by area percent: the area of the main peak divided by the total area of all the peaks the detector sees. As an example of the arithmetic, a result of 99.2 percent by area would mean that, of everything that produced a signal, the main component produced 99.2 percent of it.

The phrase “that produced a signal” is the limitation. An ultraviolet detector sees only compounds that absorb at the chosen wavelength. Water does not show up. Inorganic salts do not show up. Residual solvents usually do not show up. A sample can therefore read very high by area percent and still contain a meaningful fraction of material the method cannot see.

Assay closes that gap. An assay measures the amount of the substance against a reference standard of known content and reports it as a mass fraction of the sample, either as received or on a stated basis such as anhydrous or solvent-free. Because the comparison is to a weighed standard rather than to the sample’s own peaks, assay accounts for everything in the container, visible to the detector or not.

Why the two diverge

Consider the combinations a certificate can describe:

• Correct identity, high purity, high assay: the material is what it says, and the container holds mostly that.
• Correct identity, high chromatographic purity, lower assay: the main compound is right and its organic impurities are few, but water, salts or solvent account for a share of the mass.
• Correct identity, low purity: the right compound is present alongside related impurities, such as degradation products or synthesis by-products.
• Failed identity: no other result on the certificate matters, because it describes the wrong material.

The second case is the one most often misread. A high area-percent figure is sometimes presented as if it were the content of the container. It is not. It describes the composition of what the detector saw.

Supporting results that complete the picture

Because no single method sees everything, a complete certificate reports several results that together account for the mass. Water content is commonly measured by Karl Fischer titration. Residual solvents are measured by gas chromatography. Inorganic content may be estimated as residue on ignition. Counter-ions, where the material is a salt, are measured separately. When these figures are reported alongside chromatographic purity, a reader can check whether they add up to something close to the whole.

This is the logic behind the term orthogonal: two methods that measure the same property by different physical principles, so that the weakness of one is not shared by the other. A result confirmed by orthogonal methods is far harder to get wrong by accident.

Reading a certificate with both questions in mind

Start with identity. Check which technique was used, whether the sample was compared against a reference standard, and whether the result is stated as conforming. Then look for how content is reported. If only area percent appears, the certificate has answered a narrower question than it may seem to. If an assay appears, check its basis — as is, anhydrous or dried — and look for water and solvent results that explain any difference between purity and assay. The two numbers are not meant to agree. They are meant to be explainable.