HILIC frustration

[Mattias]

Experienced a lot of ion-exchange on the ZIC-HILIC column so I changed to the Waters XBridge HILIC column. But there seems to be a lot of ion-exchange on this one as well.

I run my samples with a water gradient (5 to 60% water), but some peaks never elute (mostly excipients such as methionne). That would be OK, but the column gets overloaded with excipients after a few injections, causing retention time drift and pressure increase.

After risning the column with 0.5M NaCl, the column is as new again. But to be able to use HILIC I need to run wash programs between every injection. The other option would be to use more buffer in the mobile phase, but the only option seems to be ammonium acetate (completely destroying the sensitivity at 220 nm).

Just some frustration written down... If you know of any HILIC column without ion-exchange or how to shut the ion-exchange down (in 95% acetonitrile) - please let me know

[Andy Alpert]

Mattias,

I've been writing about this issue for years; see my paper per the following link: http://pubs.acs.org/doi/pdf/10.1021/ac070997p

A number of companies see fit to market their columns for HILIC as "neutral" materials, and recommend mobile phases containing ~ 5 mM salt or so for their convenience. No material based upon silica is completely neutral. The ZIC-HILIC material has an additional problem. Evidently the amine group in the zwitterionic ligand is titrated with a counterion at a lower salt concentration than is the sulfonate group, for whatever reason. Accordingly, at salt concentrations < 20 mM, it acts as a cation-exchange material. Both Knut Irgum's group and David McCalley have published papers that demonstrate that at salt concentrations of 5 mM or less, the material has a higher cation-exchange capacity than does our PolySULFOETHYL A material that is synthesized as an SCX material! As the salt concentration is increased to 20 mM, these electrostatic effects are rapidly attenuated; there are quite a few papers in the literature that have graphs that demonstrate this effect with a number of commercially available HILIC materials. With 20 mM salt or more in the mobile phase overall, nearly all HILIC materials do act as neutral materials. The salt can be a volatile one, if that's important.

The worst electrostatic effects are seen with materials with no coating (e.g., uncoated silica) or thin coatings (e.g., diol type) or with materials where the functional group contains a charged ligand (ZIC-HILIC; click-chemistry ligands with a triazole ring; etc.) Materials with thick neutral coatings have less of a problem and require less salt; our PolyHYDROXYETHYL A, Tosoh's Amide-80, etc. Contact me off-list (aalpert@polylc.com) if you want references.

Per your requirement that the salt be soluble in 95% ACN: Use an organic salt. Examples are triethylammonium acetate or formate, triethylammonium phosphate, and sodium methylphosphonate. Be advised that selectivity can change dramatically with different salts, per Table 1 in the paper via the above link. Also, the salt in your sample solvent should match the salt in the starting mobile phase, per Fig. 14 in that paper.

Why did 0.5 M NaCl restore your column's performance? Because that supplied enough salt to create a layer of counterions for the charged groups in your stationary phase coating (the technical term is the electrical double layer). If you proceeded to use a mobile phase with less than 20 mM salt, then that's not enough to replace the ions that get leached away over time, and your coating became progressively less well titrated with counterions.

[Petrus Hemstrom]

If you absolutely has to start your gradient at 95% AcN then you are very limmited in what buffer salts you can use. Potassium phosphate can be used from about 85%. The organic phosphates and phosphonates Andy suggests works better but are often not HPLC grade so you could run into problems with ghost peaks.
Sodium perchlorate could perhaps be an alternative but perchlorates are explosive so take care should you try that. (I tried ordering ammonium perchlorate when at Uni for this purpose but there were some issues about transporting it )

[Mattias]

I found in another thread that isopropanol could be used to improve the solubility of ammonium acetate. I had no problem to dissolve 25 mM in a mix of 80% ACN, 10% IPA and 10% water. I will see shortly if I have any retention though...

HILIC appears to be a technique full of limitations, no wonder why it has never become big.

- Very hydrophilic samples must be forced in solution (at very high organic%)
- A lot of salt is needed in the mobile phase (under conditions where salts are not soluble). (Maybe not the case with columns Andy suggested!)
- Appears to be extremely sensitive to sample/mobile phase mismatch

[Petrus Hemstrom]

It has limmitation but so has RP (or you would not be trying HILIC). If you have very complex samples you want to analyse by UV I would split the sample in a C-18 SPE anything not retained you analyse by HILIC anything retained you analyse by RP.
With that you should be able to start your gradient at 85% AcN with 5-10mM phosphate buffer, you should have no problem with background at 220nm.