Retention Time Is Evidence, but Only Against a Standard

Retention time is the first thing anyone reads off a chromatogram and the most frequently over-interpreted number in peptide analysis. It is real evidence of identity, but only under conditions that are often not stated.

What retention time is a property of

It is not a property of the molecule. It is a property of the molecule in combination with a particular column, a particular mobile phase, a particular gradient, a particular temperature and a particular instrument’s plumbing.

Change the column to a different batch of nominally identical packing and it shifts. Change the gradient slope, the flow rate, the column temperature or the dwell volume of the pump and it shifts. Two laboratories running what they describe as the same method will not report the same retention time, and neither is wrong.

What it does establish

Run alongside an authentic reference standard on the same system in the same sequence, retention time is strong evidence. A sample peak that co-elutes with a standard shares whatever property the separation is sorting on, which for reversed-phase means hydrophobic surface.

That is a real constraint. A different compound would have to match the standard’s hydrophobicity closely to co-elute, and most do not.

What it does not establish

Co-elution is not identity. Two different molecules with similar hydrophobicity elute at the same time, and the whole category of diastereomers and isomers exists precisely because composition can be identical while structure is not.

Retention time also cannot distinguish a single compound from two compounds that happen to overlap. A peak is a region of the chromatogram where something absorbed; it is not a guarantee that only one thing was there. That question is addressed in co-elution and peptide purity.

Relative retention time travels better

Because absolute retention times do not transfer between systems, methods often report relative retention time instead: the retention of a peak divided by the retention of a reference peak in the same run.

Dividing cancels much of the system-to-system variation, since most of it affects both peaks similarly. A relative retention time of 0.87 for an impurity means something closer to the same thing in two laboratories than an absolute figure of 12.4 minutes does.

It is not perfectly transferable either — selectivity differences between columns affect the two peaks differently — but it is the more portable number, and a certificate that reports it is being more useful than one that reports minutes alone.

Why the void volume matters

Everything injected passes through the column. Anything with no affinity for the stationary phase passes through in the time it takes the mobile phase itself to traverse the system: the void time.

A peak at the void has not been separated from anything else at the void. This is the recurring problem with very polar analytes, and it means a retention time near the void carries almost no identity information. Retention only becomes evidence once there is retention.

What shifts retention without anything being wrong

• Column age. Stationary phase degrades with use and retention typically decreases over a column’s life.
• Temperature. Retention generally falls as temperature rises. A method without a stated column temperature has left out a variable that moves the number.
• Mobile phase preparation. Small differences in additive concentration change ion-pairing strength and therefore retention, particularly for charged peptides.
• Sample solvent. Injecting in a solvent stronger than the starting mobile phase distorts the front of the peak and shifts its apparent centre.
• Equilibration. A column not fully re-equilibrated after the previous gradient starts the next run in a different state.

Peak shape carries information retention time does not

A peak at the expected time with the wrong shape is worth more attention than a peak slightly off time with the right shape. Tailing indicates secondary interactions, fronting often indicates overload or a solvent mismatch, and broadening can indicate conformational exchange, as with proline-containing sequences.

None of that is visible in a retention time reported as a number. It is visible in the chromatogram, which is one of several reasons a certificate that includes the trace is more informative than one that includes a table.

How to read it on a certificate

A retention time is interpretable when the document also gives the column, the gradient, the flow rate, the temperature and whether a reference standard was run in the same sequence. Without those, it identifies a peak within that one chromatogram and nothing beyond it.

That is not a small thing — identifying peaks within a run is what the number is for — but it is less than it is often taken to mean. Related reading: gradient elution and reference standards and traceability.