HILIC Column Cleanup

[Douglas Moore]

We've been developing a HILIC method for analyzing amino acids in plasma. It worked fine but the peak shape degraded and retention time shifted rapidly after injecting samples prepared by protein crash. I suspect that the phase is bound with phospholipids, or perhaps the enzyme inhibitor added to the plasma. Anyone else encountered this problem? If so, do you have any idea what is the source of this problem, and how to regenerate a HILIC (or Silica) column after you've introduced plasma?

Many Thanks in Advance,
Doug

[HW Mueller]

I am having similar problems (drifting retention times, rt) in an even much less complicated system. Of course, I have to use EtOH wich is somewhat problematic, giving strange rt variations (very slow precipitation problems with some buffers). Dramatic rt changes obtained when I co-injected prolines with ionic solvent (1-butyl-3-methylimidazolium triflate, 10µL). Cleaning after that worked best with an acidic (to prevent SiO-) mobile phase, still it was a slow process. To prevent this type of thing you may have to pre-clean samples further.
HILIC appears quite touchy, at least in my hands (includes silica as well as zwitter ionic systems, some has been mentioned before).

[Uwe Neue]

There could be multiple sources to the problem, including insufficient protein precipitation. I would start looking at this...

If you do not find a reliable solution there, I recommend to use SPE as the sample cleanup tool. Consider an ion-exchange method for the sample cleanup using a mixed-mode ion-exchanger, followed by elution in high organic with injection onto the HILIC column.

Here is the idea: acidify your sample until you analytes are positively charged. Load this sample onto an Oasis MCX mixed-mode cation exchanger. Wash with methanol under acidic conditions to remove all neutral and oppositely charged interferences. Now elute the analytes changing the pH to alkaline with ammonia (potentially with some addition of water, if phospholipids are indeed the problem). Inject the now much cleaner solution...