Chromatography Forum

Hi,

I am running a lot of HILIC these days, but I have some problems with the columns. I have quite nice and pure samples, but on some columns I get peak distortion or peak splittning after only few injections. It seems to dependant on the individual column. I have columns that are still fine after many hundreds of injections.

I run only on silica columns (like Kinetex HILIC and Accucore HILIC). This is limiting the possibility to transfer the methods to QC.

Are HILIC columns harder to produce / more sensitive than standard C18 columns?

HILIC columns aren't harder to pack well than are reversed-phase columns (high-capacity ion-exchange columns can be harder to pack well, due to electrostatic repulsion effects). That doesn't mean that Phenomenex and Thermo Fisher always do a good job packing Kinetex and Accucore columns. In any case, we should take the opportunity to eliminate possible factors that can lead to peak splitting:

1) Does your sample solvent match the composition of the starting mobile phase?

2) Do your analytes have (-) charge under your running conditions? If so, then how much salt is there in the mobile phase (overall, not just in the aqueous portion)?

3) If you're using an automatic sample injector, then what's your needle wash solvent?

- Andy Alpert

HILIC columns are not "special". They are mostly just relabeled silica, amino or cyano columns. The manufacturers found a new market for this mode of chromatography and use new labels to get buyers to purchase more columns from them (marketing).

Silica columns are easily fouled with sample when running HILIC methods. We have found that the thin water layer gets fouled at the surface and this leads to lost resolution, peak splitting or broad peak shapes. The trick to making the columns last longer is to do the same things we always do when using silica columns. For example: Filter all samples. Make sure samples are 100% dissolved before injecting in the initial mobile phase. Following up each (or every few) analysis runs with a good wash solution that sweeps the water out of the column, then replaces it. Alcohols work best for this (We use a high percentage of IPA to do this when running any methods on silica phases and it works great at re-conditioning that thin water layer).

HILIC columns are not just a marketing gag!! DIOL or NH2 are also called HILIC, but there are several modifications only desinged for HILIC doing a good job. Silica based HILIC columns are not harder to pack and they are also sensitive to hard changes in MP composition or changes in pH. The questions posted by Andy are good questions. Good luck.

Thank you for your replies!

Regarding Andy's questions, I have typically 10-20 mM of ammonium acetate (total conc) pH 4-5 in the mobile phase (about 85% ACN in mobile phase). My molecules are basic and positively charged up to about pH 9.

Right now I am only running reference standards dissolved in mobile phase, so it is very "friendly" samples.

I have tested to implement wash protocols with 50 mM ammonium acetate in 50% ACN. But I cannot see any improvement. It was a good tip to try IPA instead, I will test that. But right now it seems as when the column is dead, it is dead for good. Change of precolumn can help to get maybe 10 more injections, but it is not a big benefit to use precolumn. I see no pressure changes over time either.

With small, highly charged analytes (example: arginine), you may need to use as much as 40 mM salt in the mobile phase and sample solvent to insure that all of the ionizable groups have the same counterions. Different ion pairs differ in their polarity and will migrate at different rates through a HILIC column.

Also, try running a column in the reverse direction before discarding it, in case the peak splitting is caused by particulate material that's accumulated on the outside surface of the inlet frit.

To the original question - C18 columns are, in my experience, The toughest packing you can use on an LC, so Every other packing is therefore easier to "break".

DR wrote:
To the original question - C18 columns are, in my experience, The toughest packing you can use on an LC, so Every other packing is therefore easier to "break".

I was thinking about it being bare silica, if that is what is being used. If you get any physical "crushing" of the particles, or any chemical "disolving" of the particles then you will end up with void space that would cause the peak splitting. C-18 would have the "thickest" layer of protection to help prevent such things from happening, and could protect the silica from chemical attack. The mobile phase doesn't seem as it would be attacking the silica chemically, so the next question would be what is the operating pressure during analysis and are there any times the column would undergo rapid changes in pressure?