Peptide purity testing

Almost every purity claim in this space rests on two techniques, and the difference between them is the difference between "mostly one thing" and "mostly the right thing".

HPLC: how much of it is one substance

High-performance liquid chromatography pushes a dissolved sample through a packed column under pressure. Components travel at different speeds depending on how strongly they interact with the packing, so they emerge separated in time. A detector at the outlet records what comes off and when, producing a chromatogram of peaks.

Purity is calculated from the areas: the target peak's area as a percentage of all peak area. A result of 99 percent means the main peak accounted for 99 percent of the total signal.

Three limits follow directly from how that works.

• It only sees what the detector responds to. UV detection at the usual wavelengths responds to particular bonds. A contaminant that does not absorb there contributes nothing and is invisible to the percentage.
• Co-elution. Two substances that travel at the same speed emerge together as one peak. A closely related impurity, a deletion sequence missing a single residue, can hide inside the main peak.
• It does not identify anything. Retention time is suggestive, not conclusive.

Mass spectrometry: which substance it is

Mass spectrometry ionises the sample and measures mass-to-charge ratio, yielding molecular weight to high precision. Since a peptide's sequence determines its exact mass, a measured mass matching the theoretical mass is strong evidence the molecule is the one claimed.

This is the identity test, and it catches what HPLC cannot: a sample that is 99 percent pure and 100 percent the wrong peptide would pass the first and fail the second.

Its own limits are worth stating. Isomers of identical mass are not distinguished by mass alone. Quantifying how much is present is not straightforward, because different molecules ionise with different efficiency, which is why mass spectrometry answers identity while HPLC answers proportion. Sequence confirmation, as opposed to mass confirmation, needs fragmentation analysis.

Why both, and what still is not covered

Run together the two answer complementary questions: what fraction of the sample is a single substance, and what that substance is. A certificate of analysis reporting only one of them has answered half the question, and a certificate reporting only HPLC is the common case.

Neither technique addresses several things that matter specifically for an injectable:

• Sterility, a microbiological test.
• Endotoxin, measured by LAL assay. It is not a peptide, produces no impurity peak, and is a genuine hazard in an injected product.
• Residual solvents from synthesis, needing gas chromatography.
• Net peptide content, the fraction of the powder that is actually peptide rather than counter-ions and water, which is what determines whether a vial labelled 5 mg holds 5 mg. See lyophilized.

Material can therefore be 99 percent pure by HPLC, confirmed by mass spectrometry, and still be non-sterile, carry endotoxin, and contain meaningfully less peptide than the label states. Each is a separate test that is separately present or absent from the certificate.

And none of it speaks to whether the compound does anything. That is a question about clinical evidence, which is what the compound library grades, and it is entirely independent of what is in the vial.

What a real certificate shows

Knowing the techniques makes a certificate readable rather than decorative. A chromatogram should be present, not just a number: the trace shows whether the main peak is clean or sitting on a shoulder, and whether the baseline is flat. A number with no trace behind it cannot be checked.

For the mass spectrum, the figure to look for is the observed mass against the theoretical mass for the sequence. They should agree closely. A certificate that states a molecular weight without saying what it was compared against has not made a claim you can evaluate.

Then the conditions: which column, which gradient, which wavelength. These determine what the run could have separated and what it could have missed, and their absence is itself informative.

Finally the identifiers. A batch or lot number matching the vial in hand, a date, and the name of the laboratory that ran it. Testing commissioned and paid for by the seller is not independent of the sale; independent third-party testing is the version that carries weight.

Related terms: certificate of analysis, lyophilized, research chemical.