Resources
Purity is not assay: reading HPLC figures correctly
Area percent purity and mass fraction assay answer different questions on different bases. How to read both without overstating what is in the vial.
Published 28 September 2026 · ImpurityLab technical team
Two different questions
"Purity" and "assay" are not two words for the same measurement. They answer different questions, and the questions have different bases.
- Purity asks: of everything the detector saw, what fraction was my analyte? The answer is a ratio of responses, and it is dimensionless.
- Assay asks: of the mass I weighed out, what fraction was my analyte? The answer is a mass fraction, and it can be expressed on a defined basis.
A chromatographic purity is almost always reported on the material as supplied. An assay is frequently reported on the anhydrous and solvent-free basis, or corrected to the free base. If the vial in question holds one and a half percent water, those two conventions alone separate the numbers by roughly that amount before anything interesting has happened.
Before treating a difference between purity and assay as a discrepancy, check the basis statement. Two figures that look inconsistent are often two correct measurements of the same material, taken on two different definitions of “the material”.
How each number is actually obtained
| Method | What it measures | What it can miss |
|---|---|---|
| HPLC purity, area percent | Ratio of the analyte peak area to the total peak area at the detection wavelength | Components with weak or no absorbance at that wavelength; anything that does not elute or does not retain |
| HPLC assay against an external standard | Response of the analyte relative to a standard of assigned content | Inherits the uncertainty of the standard's assigned content |
| Quantitative NMR | Mass fraction, from the ratio of analyte proton integrals to an internal standard of known purity | Anything without protons in the integration window; requires solubility and resolved signals |
| Mass balance | One hundred percent minus organic impurities, water, residual solvents and inorganic residue | Nothing in principle — but every subtracted term carries its own uncertainty, and the uncertainties add |
The area percent method is the one that gets misused. It assumes every component responds equally at the detection wavelength. That assumption is convenient and, for most impurity pairs, wrong. Its usefulness comes from consistency: the same method on the same system gives a comparable number batch to batch, which is what a specification needs. It does not give you a mass fraction, and it is not traceable.
Quantitative NMR is the opposite. It gives a mass fraction directly, traceable to a primary standard through the internal standard, and it is the reason a certificate can quote an assay with an uncertainty statement. The price is that it needs solubility, resolved signals and a suitably long relaxation delay, and it takes longer than an injection.
The mass balance approach is conceptually complete and practically demanding. Subtract the organic impurities, the water by Karl Fischer, the residual solvents by gas chromatography, and the inorganic residue. Each term is measured by a different technique with a different uncertainty. Add four terms each carrying half a percent of uncertainty and the combined uncertainty on the assay is not half a percent.
Why the gap widens on small quantities
The smaller the quantity, the more the two numbers diverge, and there are physical reasons rather than bookkeeping ones.
- Water is a larger share of a small vial. A hygroscopic solid picks up atmospheric moisture in proportion to its surface area, not its mass. On a one milligram quantity, water acquired during weighing and transfer can be a meaningful fraction of the total, and Karl Fischer needs a sample size that a one milligram vial cannot spare.
- Residual solvent survives purification. The last traces of the solvent a material was recrystallised or precipitated from do not always leave, and they are invisible to a chromatographic purity method that has no peak for them.
- Inorganic residue is invisible to both. A chromatographic purity method does not see inorganic salts, and neither does proton NMR if the counter-ion has no protons. A mass balance does.
- Handling losses are not random. Static, vial wall adsorption and transfer losses bias small quantities in a consistent direction rather than averaging out.
This is why a reference standard intended as a calibrant is normally supplied with an assigned content and an uncertainty, rather than with a chromatographic purity alone. You cannot build an accurate molarity on a ratio of peak areas.
Reading the numbers correctly
A worked comparison, using figures that are common on certificates:
| Statement on the certificate | What it is telling you | What you can do with it |
|---|---|---|
| Purity 98.5 % by HPLC, area percent, as supplied | 98.5 % of the detector response at the stated wavelength was the analyte | Compare batch to batch; confirm the material is the right order of quality |
| Assay 99.8 % by qNMR, corrected to the free base, anhydrous | 99.8 % of the anhydrous free-base mass was the analyte | Calculate a concentration |
| Water 1.4 % by Karl Fischer | What separates the two bases above | Convert one figure to the other, approximately |
The three statements are consistent. The chromatographic purity is on the material as supplied and is a response ratio. The assay is on the anhydrous free base and is a mass fraction. The water figure is the bridge. Read together, they describe one vial well. Read separately, the first two look contradictory.
Note also the direction of the correction. If you weigh the material as supplied and calculate using the assay corrected to the free base, you will overstate how much analyte is in the vial — and the error is systematic, so replicate injections will not reveal it.
What to write in your report
Four items, and any of them missing makes the number ambiguous.
- The basis. As supplied, anhydrous, anhydrous and solvent-free, or corrected to the free base. State the form the value refers to.
- The method and the detector. HPLC with the wavelength, quantitative NMR with the internal standard, or mass balance with the contributing techniques.
- The uncertainty, where one exists. An assigned content with an uncertainty statement is a different category of information from an area percent, and the report should not present them as though they were interchangeable.
- The identity of the material. Lot number, form as supplied, and the certificate version, so the number can be reproduced later.
Do not mix bases inside one table. A results table that lists a purity on the as-supplied basis beside an assay on the anhydrous basis, with no column for the basis, is a table that will be misread by whoever inherits it.
The short version
Purity is a response ratio. Assay is a mass fraction. They diverge by the water, the residual solvent, the inorganic content and the counter-ion — and by the basis each one was reported on. Use purity to judge whether the material is the right quality; use assay to calculate. And read the basis statement before deciding that two numbers disagree.