Buying and sourcing

Third-party testing: what HPLC and mass spectrometry each show

This page explains two analytical techniques as they appear on certificates. It recommends no laboratory, evaluates no product, and names no purity threshold as acceptable. Understanding a method makes a document readable; it does not make a result reassuring.

The one idea that makes both readable

A result is a number produced by a procedure. Change the procedure and the number changes. That is not a criticism of analytical chemistry; it is what analytical chemistry says about itself, which is why validation guidance exists at all and why it defines a method's characteristics with respect to a stated purpose.²

What HPLC physically does

In reversed-phase HPLC, a sample is pushed through a packed column under pressure while the composition of the mobile phase changes over the run. Components leave the column at different times depending on how strongly they interact with the stationary phase. A detector — most often ultraviolet absorbance at a chosen wavelength — records a signal as each one emerges, producing peaks.

Purity is then reported as the area of the main peak divided by the total area of all detected peaks. Read that definition slowly, because three limits are contained in it.

• Anything that does not respond to the detector never appears. A UV method sees what absorbs at that wavelength; salts and water do not.
• Anything that does not leave the column within the run time never appears. Strongly retained material simply is not in the denominator.
• Anything that emerges at the same time as the main component is counted as the main component. Co-elution is invisible by definition.¹

This is why two competent laboratories can report different purity figures for the same material and neither be wrong. They ran different procedures, and the percentage is a property of the procedure applied to the material, not of the material alone.¹²

What HPLC is good at

Separation. It is the technique that can resolve species a mass measurement cannot distinguish — including diastereomers and sequence variants that share a molecular formula — provided the method was developed to resolve them. The peptide-impurity literature catalogues the classes involved: deletion and insertion sequences, truncated and elongated chains, diastereomers, oxidation and deamidation products, protecting-group residues and aggregates.¹

What HPLC cannot tell you

What anything is. A chromatogram is a set of retention times and areas. Assigning an identity to a peak requires either a reference standard run under the same method or a detector that measures something structural — which is where mass spectrometry enters.

What mass spectrometry physically does

A mass spectrometer ionises molecules and measures their mass-to-charge ratio. For peptides this normally means electrospray ionisation, which produces multiply charged ions that are deconvoluted into an apparent molecular mass. On a certificate, the result is usually presented as a measured mass alongside the theoretical mass for the intended sequence.

That is a genuine and useful check. A mass consistent with the expected value is evidence against a grossly different molecule having been supplied. It is bounded evidence, in three specific ways.

Bound one: mass is not identity

Different molecules can share a mass. The textbook case inside peptide chemistry is leucine and isoleucine, which are isomers with identical mass and therefore indistinguishable by a mass measurement alone; distinguishing them reliably has required purpose-built methods and remains a recognised problem in de novo sequencing. Stereochemistry is worse: a D-amino acid substitution changes no mass at all.³

Bound two: mass is not quantity

Ionisation efficiency varies between species and with the matrix, so peak intensity in a mass spectrum is not a direct measure of how much is present. Quantitation requires calibration against a reference standard, usually with an internal standard, under a validated procedure. A certificate line reading "mass confirmed" contains none of that and asserts nothing about content.²

Bound three: what was not ionised was not seen

The same blind-spot logic as chromatography applies. Species that ionise poorly under the chosen conditions are under-represented or absent, and a spectrum showing the expected mass says nothing about what else is in the vial that did not ionise.

Why the two are usually run together

Coupling them — LC-MS — is the standard answer to both blind spots at once: the column separates, and the spectrometer identifies what came off it at each moment. That combination is what makes it possible to say not just that impurities are present at some percentage but what they are, which is the level at which regulators actually work.¹⁴

FDA's peptide work is a good illustration of the gap in ambition. Its material on assessing peptide impurities is oriented toward immunogenicity risk from specific impurity classes, and its revised draft product-specific guidances for generic peptides address impurity thresholds, higher-order structure and biological activity. A consumer certificate reporting one percentage and one mass is not a smaller version of that analysis. It is a different exercise.⁴⁶

Reading a method line honestly

• Which technique, named specifically enough to be looked up — not "HPLC" alone but the column chemistry, gradient, detector and wavelength.
• Whether the procedure is stated to have been validated for this material and this purpose, in the sense validation guidance defines.
• Whether a reference standard was used, and which one, for any result presented as a quantity.
• What the acceptance criterion was, if the laboratory was asked to judge the result against one.
• Whether the laboratory's accreditation scope covers the test on the page, checked in the accreditation body's register rather than against a logo.²⁵