Does HILIC have a future?

[Einar Ponten]

There are several good reasons why Alpert (PolyLC) was advocating HILIC for a long time and SeQuant introduction of ZIC®-HILIC in 2002 was followed by Waters and now by almost all column manufactures.

Simply, HILIC solves problems where other techniques fails or are limited. In addition, it provides faster and easier sample preparation and increased sensitivity for LCMS users.

The major problem is typically that the user try to use a HILIC column as it was a RP separation and since many chemist (including me) are lazy they sometimes don't adopt the sample pre-treatment for HILIC and the mobile-phase for the sample. Injection of water on a HILIC column require a low injection volume, just as acetonitrile would on a RP column. This is the most frequent mistake. Then it also may be an advantage to use another mobile phase pH for HILIC and thereby modulate the hydrophilicity of the analyte and control the selectivity.

In response to some posts the capacity of HILIC has been reported to be superior to RP in several cases and that is also our experience. Uwe mentioned sugars and that is alright, but there are also alternative techniques in frequent use, but if you look at all ion-pairing and derivatisation methods the only alternative is often HILIC. An attractive one since it opens the door for new sample prep and MS, ELSD, RI, UV etc

For sure, HILIC will continue to grow while still the need for qualified support is widespread. We try to contribute by providing "A Practical Guide to HILIC" free of charge to anyone requesting it at http://www.sequant.com/sn/p_notes.php?id=7

[Omar]

Carbohydrate Analysis Column from Waters

Uwe, can you please give the specifications, (packing type, dimensions ecc) of the column you mentioned above? Thanks very much.

[Uwe Neue]

It is a stabilized amino-type column, 3.9 mm x 300 mm, 4 micron. Can be found readily in the catalogue or on-line.

[SIELC_Tech]

We recently introduced a new concept of Liquid Separation Cell (LiSC). It covers two columns: one works in reverse phase mode and another one in HILIC mode. Our HILC column allows you to change ionization state of the stationary phase in HILIC/normal phase mode. This gives you additional tools to adjust selectivity. It also allows you to work at lower ACN concentration (70% vs. 90%) which is good in terms of solubility of polar compounds (limited in NP/HILIC) and reproducibility (easier to control composition of the mobile phase). Here is brochure for Liquid Separation Cell:

http://www.sielc.com/pdf/SIELC_Obelisc_Intro.pdf

Our next newsletter is devoted to comparison of different HILIC column (www.sielc.com)

Regards,

Vlad

[blackdrum]

I find an introduction of hilic mechanism from waters:1."Polar analyte partitions into and out of adsorbed water layer", 2."Charged polar analyte can undergo cation exchange with charged silanol groups". if these are true, comparing to RP, the effective surface should be small (without long chains), so the loadbility should be limited, is this interpretation true? I use a atlantis hilic silica column for preparation(19*150mm), the loadbility is really much lower than sunfire column with same size. any comment?

[Uwe Neue]

Actually, the surface area of an underivatized silica is higher than the surface area of a dense C18 on the same packing. Therefore I do not think that this is the reason for the lower loadability that you are observing. There could very well be experimental details that play a role. The solubility of analytes that undergo HILIC interactions is limited in the solvents used. If the dominant factor is ion-exchange, then indeed the loadability is low due to charge repulsion, and one needs to work out a method with a higher ionic strength.

[blackdrum]

my compound is soluble well in pure acetonitrile, so solubility isn't a reason. Uwe Neue, could you give some evidence to support this:"the surface area of an underivatized silica is higher than the surface area of a dense C18 on the same packing"? and I think we should consider the stereo structure of stationary phase and analyte, the partition efficiency should be different for bare balls and balls with long chains. I hope somebody gives much professional interpretation for hilic mechanism, after all, for now, it's very hard to find a good method on hilic than on RP, and we can't predict the retention as we are used to.

[Uwe Neue]

Re evidence for the change in surface area with bonding: How about measurements by nitrogen adsorption, which is the tool commonly used to measure surface area.

There are two dominant mechanism active in HILIC. The first one is a partioning between the mobile phase that is poor in water and a layer (or maybe even multilayer) of water on the polar surface. The polar surface can be silica, an amino bonded phase, a zwitterion bonded to the surface, an amide layer, either obtained by bonding or coating. Sugars are retained via this primary mechanism.

The second mechanism is ion-exchange: cation exchange with silica, and anion exchange with an amino phase. Many analyses of pharmaceuticals on silica are dominated by this second mechanism.

If solubility is not the problem, then I would need to understand the details of your method to comment further why the loadability is low in your case. What is the loadabililty? What is the nature of the analyte? What is the mobile phase? What is the ionic strength, if the mechanism is ion-exchange? Etc. ...

If you do not want to post this in public, you can contact me directly.

[Patrik Appelblad]

In my opinion, loadability in HILIC should be as good as in RPLC, at least that is my experience from HILIC prep work. Have a look at page 32 in this presentation: http://www.sequant.com/sn/ufiles/SeQuan ... r_2700.pdf.

There are though, differences among materials, different stereo structures etc wherefore the partition will differs as shown in recent studies. Tolerance to differences among sample matrix and mobile phase composition is also an issue among materials.
As explained above:
1) Two dominant mechanisms, partitioning between the mobile phase and a layer (or maybe multilayer) of stagnant water, and ion-exchange interaction, affect separations.
2) A majority of analytes are dominated by the ion-exchange interaction, and if you use cation exchange on silica, or anion exchange with an amino phase, then indeed the loadability is low due to charge repulsion.
3) The partition efficiency is different for bare particles and those with a bonded phase.

A viable alternative is to use a zwitterionic phase. With a permanent, yet neutral, charged stationary phase, the separation process will be controlled by weak electrostatic rather than strong electrostatic interactions on silica and amine phases. You will not need to increase the ionic strength of the mobile phase to improve retention as for those dominated by ion-exchange. pH is another important parameter that will affect the outcome, and a zwitterionic phase is independent in contrast to other phases.

[Victor]

Patrik,

This link did not work for me. It was therefore difficult to see which of your application files ( I could see them on your website) that you were particularly referring to?

"You will not need to increase the ionic strength of the mobile phase to improve retention as for those dominated by ion-exchange." I do not understand this statement. If you have bases retained by ion exchange e.g on silica, then an increase in ionic strength will reduce the retention?

[JA]

The Sequant hyperlink was corrupted by an additional "." at the end.

Try http://www.sequant.com/sn/ufiles/SeQuan ... r_2700.pdf

[HW Mueller]

I am loath to get into the fray again, but this matter of partition, as well as claims that ZIC-HILIC has no effects due to SiO- should, in my opinion, be considered highly controversial. A recent article is quite interesting in this respect: Y GUO, S Scrinivasan, S Gaiki, Chromatographia, 66, 223 (2007).

[Victor]

HWM-yes this is a good reference.

Patrik-perhaps you are talking about the analysis of acids? In this case increasing the ionic strength on a silica column reduces the repulsion that these solutes will have with ionized silanols?

I would also ask the same question as HWM. The bonded cationic group and anionic group on the ZILIC phase are supposed to be so close together that ionic interactions with each are minimized. But what about the ionic effects of the silanols on the underlying silica? Are you claiming that the bonded ligand also minimizes any ionic effects of these groups?

On another point, the company SIELC is marketing a HILIC phase which has cationic and anionic groups further apart on the bonded ligand. They claim this is beneficial since it increases the retention of either cations or anions which can interact separately with appropriately charged solutes. ie the claim is that such ionic interactions are no bad thing. This is the opposite of the aim of the ZILIC column. Can you comment on this?

[Victor]

sorry-that's ...appropriately charged column groups. And thanks JA for repairing the link

[SIELC_Tech]

Our approach is still HILIC approach. You can retain non-ionazable compounds (sugars) in HILIC mode and use 75-80% ACN. The advantage is that you have additional mechanism to control your retention and separation. In case of amino acids (same with acids and weak amines) you can control ionization state of analyte and stationary phase. You can easily separate amines, amino acids, sugars and organic acids in one run (fermentation mixtures)

In a lot of cases HILIC mixed-mode (Obelisc N) allows you to reduce amount of ACN by at least 10% and achieve longer retention, higher solubility of analytes and higher loadability. Also in HILIC mixed mode approach you can use methanol which can give you different selectivity. Here is newsletter with several examples:
http://www.sielc.com/pdf/SIELC_September_2007.pdf