Is a 97.9% result from VendorInvestigate comparable to 97.1% from Janoshik?

Conditions: 97.9% · VendorInvestigate · 97.1% · Janoshik.

These are treated as interchangeable and I do not think they are.

If both are acceptable I would like to know that, so I can stop thinking about it.

What is the actual trade-off, and does it matter at the scale I am working at?

97.9 and 97.1 are 0.8 percentage points apart, which sounds small until you restate it as impurity. VendorInvestigate is leaving 2.1 per cent of the detected area unassigned and Janoshik 2.9 per cent — a factor of 1.38 between them. The impurity fraction is the quantity that moves when a method changes, and it is the one worth arguing about; the headline is just its complement. Comparable means same column chemistry, same gradient slope, same detection wavelength, same integration convention. Until you have those four from both laboratories, 97.9 and 97.1 are two measurements of slightly different quantities that happen to share a unit.

Identity and purity are orthogonal, and a high purity says almost nothing about whether the peak is actually what you think it is.

Temperature affects the dynamics of molecular conformation, and if a peptide has proline residues that interconvert on the chromatographic timescale, the peak will split or shoulder at low temperature and collapse at high temperature.

Tailing factor measures peak shape, and a badly tailing peak spreads into the region where small impurities live, forcing tangent-skim integration that assigns tail area to the main peak.

The ICH Q3A impurity thresholds and the relevant pharmacopoeial chapters all specify method validation requirements that almost no research-grade certificate claims to meet.

Worth noting that method standardisation is poor in the research-grade space compared to pharmaceutical work, so identical-looking methods can produce different results.

Compare purity within a single laboratory on the same method, never across laboratories.

edited 17 Nov 2024 by j_wierzbicki — fixed an arithmetic slip in the third paragraph

The method matters more than the vial, which is why specifying a method buys you far more than changing suppliers does.

Mobile phase additive choice affects ionisation and peak shape — TFA gives sharp peaks but suppresses mass spectrometry signal, formic acid gives worse peaks but preserves signal.

Integration of the shoulder — whether you use perpendicular drop or tangent skim — determines what area gets assigned to the main peak versus the impurity table.

Proline conformer interconversion kinetics are well-characterised and the half-life is of the same order as the chromatographic peak width at room temperature.

If you are ranking vendors, specify a method and have all samples tested at the same place.

Reporting threshold is convention and not chemistry, which is why two certificates with different thresholds disagree by a tenth of a point or more.

Column pore size affects mass transfer — a 100 Angstrom packing on a 5 kDa peptide restricts diffusion, broadening the peak and potentially hiding small impurities in the shoulders.

Mass on column affects detector linearity and peak overlap — overloading broadens peaks and hides neighbours, while underloading improves resolution but loses sensitivity.

The practical summary: ask for the chromatogram and the method, and ignore the headline number until you have both.