Size Exclusion Chromatography and Peptide Aggregate Detection

Reversed-phase chromatography separates by how strongly a molecule sticks to a surface. Size exclusion separates by how large it is in solution, and that difference is why the two methods see different problems. A dimer that co-elutes with its monomer on a reversed-phase column is resolved on a size exclusion column, because the two differ in size even when they behave identically on a hydrophobic surface.

What the column actually does

The packing is porous beads. Small molecules diffuse into the pores and take a long path through the column. Large molecules cannot enter and travel only through the spaces between beads, taking a short path. Large elutes first, small elutes last, and the whole separation happens within one column volume.

Nothing is retained in the reversed-phase sense. There is no binding step and no gradient — the mobile phase composition is held constant for the whole run. This is why size exclusion runs are short and why they are comparatively gentle on the sample.

The two volumes that bound every run

Two numbers define the usable window:

• The void volume is the elution volume of something too large to enter any pore. Everything above the column’s exclusion limit elutes at the void volume together, unresolved.
• The total permeation volume is the elution volume of something small enough to enter every pore freely. Everything below the lower limit elutes there together, also unresolved.

Separation exists only between those two points. A column chosen with the wrong pore size puts the peptide at one end or the other and resolves nothing, which is the most common reason a size exclusion run on a small peptide returns a single uninformative peak.

Calibration is relative, not absolute

Elution volume is converted to an apparent size by running standards of known molecular weight and fitting a curve. The result is an apparent molecular weight, and the qualifier matters: the separation responds to hydrodynamic radius, not to mass.

An extended or unstructured chain occupies more volume than a compact globular one of the same mass and elutes earlier, reporting an apparent weight well above its true value. Many peptides are unstructured in aqueous buffer, so an apparent weight from a globular-protein calibration curve is systematically high and should not be read as a mass measurement. Mass measurement is what the spectrometer is for, at the accuracy described in mass accuracy in parts per million.

The mobile phase is not a bystander

The ideal is that nothing interacts with the packing. In practice the packing surface carries residual charge, and peptides carry charge, so ionic interaction competes with the size mechanism. A basic peptide can be retained past the total permeation volume and appear smaller than it is; an acidic one can be repelled and appear larger.

The standard correction is salt in the mobile phase, typically a phosphate or acetate buffer at moderate ionic strength, sometimes with a small percentage of organic solvent to suppress hydrophobic interaction as well. Changing that buffer changes the answer, so a size exclusion result is only comparable against another run under the same conditions. The underlying charge behaviour is the same one discussed in isoelectric point and solubility.

What it detects that a purity chromatogram does not

This is the point of running it at all. Reversed-phase conditions are denaturing — organic solvent and acid pull most non-covalent assemblies apart before they reach the detector, so a reversed-phase trace reports the monomer content of a sample that may not be monomeric in the vial.

Size exclusion is run in aqueous buffer near neutral pH, so non-covalent dimers, oligomers and soluble aggregates survive the run and appear as earlier-eluting peaks. Covalent aggregates — a disulfide-linked dimer, for example — are seen by both methods, but the non-covalent population is visible only here. That population is what the solution behaviour described in peptide aggregation in solution produces. The two methods are a genuine pair in the sense set out in confirming identity with two independent methods, because neither shares the other’s blind spot.

Where the method misleads

• On-column dilution. The sample is diluted several-fold as it travels. A concentration-dependent, reversible oligomer can dissociate during the run and be under-reported or missed entirely.
• Adsorption losses. Aggregates can stick to the frit or the packing and never elute. Mass recovery should be checked against an injection onto a column blank; unaccounted mass is a result, not a nuisance.
• Insoluble aggregate is invisible. Anything filtered out or spun down before injection is not in the answer. A sample prepared by filtration reports the soluble fraction only, and should say so.
• Resolution is low by design. Separating species that differ by less than roughly a factor of two in size is unreliable on a single column.

Detectors, and what each adds

Ultraviolet absorbance is standard and shares the response-factor caveats set out in ultraviolet response factors — an aggregate can scatter light and inflate its own apparent absorbance. Refractive index responds more nearly to mass but is less sensitive.

Multi-angle light scattering placed in line changes the character of the measurement: it determines molar mass directly from the scattering, without reference to a calibration curve, which removes the shape assumption entirely. It requires an accurate concentration alongside it, of the kind obtained by the method in concentration by A280.

Why it is uncommon on a peptide certificate

For a short synthetic peptide, the impurities that matter are deletion and truncation sequences and residual synthesis reagents — all of which are close in size to the target and therefore invisible to size exclusion, while being exactly what a reversed-phase gradient resolves. Running size exclusion on a fifteen-residue peptide would usually produce one peak and no information.

It earns its place as material gets longer, as structure becomes relevant, and where aggregation state is a stability question rather than a purity question. Read as a companion to a reversed-phase trace it answers a question that trace is not built to answer: not how much of the material is the right molecule, but how much of the right molecule is travelling alone.