HPLC and UHPLC instruments

Liquid chromatography separates because each component distributes itself between a flowing mobile phase and a stationary phase bonded to porous particles, and the components differ in how strongly they partition. A molecule that spends more time associated with the stationary phase moves down the column more slowly. Retention is the easy part of the method. The difficulty is that every band also spreads while it travels, and resolution is the contest between the separation gained and the spreading suffered.

Three ways a band spreads#

Band broadening decomposes into three contributions, the classical van Deemter terms. Molecules take paths of unequal length through the packed bed, arriving slightly out of step. They diffuse longitudinally along the column, which matters most when the flow is slow and they have time to do it. And they must diffuse into and out of the stagnant liquid inside the porous particle to interact with the stationary phase at all, which matters most when the flow is fast and they are swept downstream before equilibrating.

Because the second term shrinks with velocity and the third grows with it, there is an optimum flow rate. The important consequence is what happens when the particle diameter is reduced: the diffusion distance into the particle falls, the mass-transfer term collapses, the optimum shifts to higher velocity and the curve flattens. Smaller particles therefore give both better peaks and faster runs.

What the smaller particle costs#

Flow resistance through a packed bed rises as the inverse square of particle diameter. Moving from five-micrometre to sub-two-micrometre packing multiplies the pressure required by roughly an order of magnitude, which is the whole reason ultra-high-pressure systems exist: UHPLC is not a different separation principle but the plumbing consequence of the same equation. The route is not open-ended. Pumping viscous liquid through a restriction at very high pressure dissipates real power as frictional heat, producing a radial temperature gradient between the column core and its wall — and the resulting viscosity gradient broadens bands again, partially cancelling the gain. Superficially porous “core-shell” particles attack the same mass-transfer term by putting a thin porous layer on a solid core, shortening the diffusion path without the full pressure penalty.

Shrinking the column also shifts the bottleneck outside it. Volume contributed by injector, connecting tubing and detector cell no longer disappears against the column volume, so extra-column dispersion, not packing quality, often sets the achievable efficiency on small columns.

What the detector cannot see#

A chromatogram is only evidence about species the detector responds to. Ultraviolet absorbance at 280 nm reports on tryptophan, tyrosine and disulphide-bonded cystine; a peptide, excipient, sugar or lipid with no aromatic residue produces no signal at all and is invisible in a purity trace. Dropping to about 214 nm detects the peptide bond itself and is far more general, but almost every solvent, buffer and mobile-phase additive also absorbs there, so baseline and gradient artefacts increase sharply. Species with neither chromophore need charged-aerosol, evaporative light-scattering or mass-spectrometric detection.

Mode choice carries its own caveats. Size-exclusion chromatography is meant to be purely entropic, but residual interaction between protein and packing means salt must be added to suppress it — and that salt is incompatible with the mass spectrometer one might want downstream. Reversed-phase separation of proteins is performed under conditions that are, by design, denaturing.