Ternary Gradients and Demixing

[hajdaei]

I was corresponding with someone last week (a knowledgable chromatographer) and they said that when using a gradient with more than two solvents there is "a possibility of eluant demixing and compressed peaks".

I've done HPLC for many years but I have to admit I have no idea what he's referring to.

Is anyone familiar with problems concerning gradient runs with more than two solvents.

Thank You

[tom jupille]

when using a gradient with more than two solvents there is "a possibility of eluant demixing and compressed peaks".

Certainly there is, but I don't see any reason why ternary gradients should be any different from binary gradients in this respect.

Some "compression" of peaks is inevitable in any gradient because the trailing edge of a peak sees a slightly stronger mobile phase than does the leading edge of the peak. It's not a problem, and only hard-core theorists worry about accounting for it.

"Demixing" refers to the fact that the stronger component of the mobile phase is preferentially sorbed by the stationary phase; this retards that stronger component and generates an anomalously steep segment in the gradient (which can anomalously compress peaks). I've encountered this in thin-layer chromatography (where it is a common phenomenon) but I'm not aware of it ever being significant in HPLC practice. That doesn't mean it can't happen, but it's not high on my list of concerns.

[goxy43]

Somebody did say that a fool will make things complicated and that the genius will make those same things simple. Well, I am the fool, but I feel good since my complicated and ternary gradient resulted with a nice increase of peptide IDs in comparison with the binary gradient. I did not observe decomposition or other phenomena. All separations (test runs) are very reproducible independently when the sample was injected.

[mohan_2008]

Solvent demixing may be seen " frequently" in Normal Phase Chromatography.

If you are running a IPA/Hexane mobile phase composition, for eg -

IPA - strong solvent
Hexane - weak solvent

Specially if you are using "Silica" columns (not the bonded phase ones).
In the initial gradient step, IPA tends to adsorb more strongly (or localize) onto the silica phase and tends to saturate the phase, while Hexane will simply be rinsing the column.

At a point, the silica phase becomes saturated with IPA - which tends to elute out suddenly pushing your retention times abruptly.

We don't see this affect in regular RP columns.