What Happens Between the Vial and the Autosampler

By the time a sample reaches the detector it has been dissolved, possibly diluted, possibly filtered, and held in a vial for some period. Every one of those steps can change what the chromatogram shows, and none of them usually appears on a certificate.

The dissolution solvent is a decision

A lyophilised peptide has to be dissolved in something, and the choice is not neutral. Water alone works for many sequences and fails for others: strongly acidic peptides dissolve poorly at low pH, strongly basic ones poorly at high pH, and amphipathic ones may need a small proportion of organic solvent or a brief period of gentle agitation.

Forcing dissolution with sonication or heat introduces its own risk. Sonication can raise local temperature substantially and generate reactive species; heat accelerates every degradation route the sequence is subject to. A sample that was difficult to dissolve and was made to dissolve may not be the same sample afterwards.

Injection solvent mismatch

This is the most common preparation error with a visible signature. If the solvent the sample is dissolved in is stronger — more organic — than the mobile phase at the start of the gradient, the analyte does not focus at the head of the column. It begins travelling immediately, spread over the width of the injection band.

The result is a peak that fronts, splits, or broadens, and the effect scales with injection volume. Diluting the sample into the starting mobile phase composition before injection removes it. A distorted peak that improves when the sample is diluted in weaker solvent was a preparation artefact, not a property of the material.

Adsorption between the vial and the column

Peptides adsorb to surfaces, particularly cationic ones to glass and some plastics. At high concentration the loss is negligible. At low concentration it can be a large fraction of what was weighed.

Two consequences follow. Recovery may be poor without any degradation having occurred, so a low apparent content is not necessarily a content problem. And the loss is not uniform across species, so if an impurity adsorbs more or less than the main component, the measured ratio shifts. Low-binding vials and adding a small amount of organic solvent both reduce it.

Filtration

Samples are often filtered before injection to protect the column. Filter membranes bind protein and peptide, and the amount bound depends on the membrane chemistry, the surface area and the concentration.

The standard mitigation is to discard the first portion of filtrate, which saturates the binding sites so that what is collected afterwards is unchanged in composition. A sample filtered without that step, particularly a dilute one, can lose a measurable proportion of its analyte, and can lose species selectively.

Centrifugation avoids the problem entirely where the purpose is only to remove particulates.

Time in the autosampler

A sequence of injections may run for many hours, and samples near the end of it have been sitting in solution at the autosampler’s temperature the whole time.

For a peptide with a rapid degradation route — a free thiol oxidising, an N-terminal glutamine cyclising, a labile asparagine deamidating — that is long enough to matter. The first and last injections of the same solution can give different impurity profiles, and the difference is real rather than instrumental.

A chilled autosampler slows it. An injection sequence that brackets the run with repeats of the same sample detects it: if the bracketing injections differ, something changed in the vial during the run.

Concentration and column overload

Injecting too much analyte saturates the stationary phase locally, and the peak fronts and broadens. Resolution between the main peak and a nearby impurity degrades exactly where it is needed.

Overload has a clear diagnostic: it is concentration-dependent. Halving the injected amount and re-running should halve the area and improve the shape. If the shape does not improve, the peak was not overloaded and the broadening has another cause.

Why none of this is on the certificate

Sample preparation is usually treated as internal method detail rather than reportable result, so a document gives the column and gradient and says nothing about how the vial was dissolved or whether the sample was filtered.

That is conventional and not in itself a criticism. It is worth knowing because it is one of the places where two laboratories analysing the same lot diverge without either doing anything wrong, alongside the mobile-phase and integration differences discussed in why certificates disagree on purity.

What this means practically

An unexpected chromatogram is worth re-preparing before it is worth re-interpreting. Dissolve fresh, dilute into weak solvent, skip or bracket the filtration, and inject promptly. If the result changes, the preparation was the variable. If it does not, the question moves back to the material. Related reading: choosing a reconstitution solvent.