A purity percentage is a summary statistic calculated from a chromatogram. Like most summary statistics it discards information, and some of what it discards is the information you would most want. Where the full trace is published, it is worth spending a minute on it.
Baseline
A flat, quiet baseline indicates a clean system and a well-prepared sample. Drift — a baseline that climbs or falls across the run — is common in gradient methods and not in itself a problem, but a noisy baseline makes small peaks unresolvable and can inflate an apparent purity by hiding what should have been integrated.
Peak shape
- Symmetrical and narrow — the expected result for a well-behaved compound on a suitable column.
- Tailing — often an interaction with the stationary phase; can conceal a closely eluting impurity in the tail.
- Fronting — usually column overload; the sample was too concentrated for the method as run.
- Shouldered — a strong indication of a co-eluting species that the integration has absorbed into the main peak.
Where the impurities sit
Not all impurities are equally interesting. Material eluting far from the principal peak is typically unrelated — solvent, reagent, an artefact of the injection. Material eluting close to the principal peak is the concerning kind, because structural similarity is exactly what produces similar retention. A 99.2% result with everything else clustered at the shoulder deserves more scrutiny than a 98.6% result with impurities scattered across the run.
What the trace cannot tell you
It cannot tell you what the principal peak is. However clean the chromatogram, identity remains a separate question requiring a separate instrument. The trace establishes that there is one dominant substance; the mass spectrum establishes which.
