HPLC vs Mass Spectrometry: Why Peptide Testing Needs Both

The sample is pushed through a hydrophobic stationary phase under a gradient of increasing organic solvent. Species partition between the mobile and stationary phases according to hydrophobicity, so they leave the column at different times. A UV detector at 214 nm — the peptide bond absorbance — records what comes off and when.

The output is a chromatogram: peaks in time. Area-percent of the main peak is your purity number. What HPLC is excellent at is counting distinct species and quantifying their relative amounts. Deletion sequences, truncations, oxidised variants and most process impurities separate cleanly and show up as their own peaks.

A co-eluting impurity is invisible. If a contaminant happens to have essentially the same hydrophobicity as your target, it hides underneath the main peak and inflates your purity figure. Running a second, orthogonal gradient — different column chemistry or different pH — is the standard defence, because two different separation mechanisms are unlikely to co-elute the same pair.

More fundamentally, HPLC does not identify anything. A peak at 14.2 minutes is a peak at 14.2 minutes. Without a reference standard run under identical conditions, the chromatogram is consistent with your peptide and with an unknown number of other molecules.

Electrospray ionisation puts multiple charges on a peptide, producing a series of m/z peaks that deconvolute to a single molecular weight. Comparing that measured mass to the theoretical mass calculated from the sequence is a strong identity constraint — a deletion sequence missing one residue is off by that residue's mass, which is unmistakable.

Coupled to the chromatographic separation, LC-MS also tells you the mass of each impurity peak, which usually tells you what the impurity is and therefore where in the synthesis it came from. That is the difference between knowing a lot is 98 % pure and knowing why it is not 99.5 %.

Stereochemistry. A D-amino acid where an L-amino acid belongs produces a molecule with exactly the same mass. Ipamorelin, which contains two D-residues by design, is the standard example: an epimerised lot is mass-spectrometrically identical to a correct one and only chromatography against a reference standard separates them.

Isomerisation generally. Aspartimide formation, a common side reaction in aspartate-rich sequences, produces a same-mass rearrangement. Cyclisation state is nearly invisible at loose tolerances — 18 daltons on a 1,024 Da molecule is under 2 %.

Mass spectrometry is also poorly quantitative on its own. Ionisation efficiency varies between species, so peak intensity in a mass spectrum is not a reliable proxy for abundance. Quantification belongs to the UV detector.

HPLC quantifies but does not identify. Mass spectrometry identifies but neither quantifies reliably nor distinguishes isomers. Run together on the same injection, they answer 'how much of it is there' and 'what is it' with independent failure modes, so a lot that passes both is very unlikely to be wrong in a way either method alone would have missed.

Every Certus lot is released on both, plus a compound-specific third assay where the molecule has a known failure mode that neither catches — chiral chromatography for D-residue peptides, an aspartimide-resolving gradient for acidic sequences, ICP-MS for copper complexes, conjugation efficiency for maleimide-bearing analogues.