Hello all, I have some general questions on HILIC. I'm currently running RP-HPLC-MS on some samples, polar and nonpolar fractions containing a few knowns and lots of unknowns. Ionization mode is ESI negative. A polar fraction appears to contain lots of unknown material that is either unretained or poorly retained on a C18 column at 5% MeOH. It also contains a known and some unknowns which are sufficiently retained. I want to get some retention and separation on the presently unretained compounds in the hope of getting cleaner spectra, so I figured this would be a good opportunity to try HILIC to complement the RP data.

I don't presently have any columns in the lab that were specifically made for HILIC. I did have some polar bonded phases (cyano and amino) that had never been used, and could probably borrow an old silica column if need be. I started with the amino column by flushing with IPA to remove the hexane, then running with the following solvent system: A = acetonitrile; B = water w/ 0.1% acetic acid.

Initially I see lots of bleed from the column compared with the C18, and a gradient from A to B and back doesn't seem to eliminate it. In the sample in question, I see some compounds are retained and elute when I do a gradient from 5% to ~ 40% B. The problem is, my known (a small carboxylic acid) appears to be retained indefinitely at any combination of A and B, and this is confirmed with a pure standard. (An alternate explanation is that the bleed is causing such severe ion suppression that I don't see the known, but I think that's less likely.)

Should I change the pH of "B"? Would I get better results with a different polar phase, or with a column specifically sold for HILIC? Is bleed a common characteristic of polar bonded phases, or is this just a bad brand?

Thanks in advance,

Mike

Why not just use anion exchange? I guess you already are, if you are using an amino phase but you need approach it as an anion exchange separation rather than HILIC to get things working. First, try a more potent eluent anion such as formic acid. Also, you don't need so much acetonitrile in your eluent. If formic acid still won't elute your analyte try an even more potent eluent anion such as TFA.

Chris, I guess I've been leery of ion exchange because it isn't often done with MS-friendly buffers (or that may be an incorrect assumption on my part). I now have some questions on ion exchange.

If I'm using formic acid as my eluent, would I do a gradient from low to high concentration of formic acid? Can I have organic in my mobile phase (for improved ES ionization)? If so, is there a preferred concentration range, and does it matter whether the organic is methanol or acetonitrile?

Thanks,
Mike

What you are experiencing is the standard problem with amino columns in HILIC. When used for sugar separations, they bleed for weeks. My recommendation would be to buy a silica column that has been designed for HILIC (Atlantis HILIC for example). I dont have experience with MS detection (and MS bleed) with the ZIC-HILIC column, therefore I do not know if I can recommend it or not. On the other hand, you will have no bleed problem with silica.
To use the silica column will also get your carboxylic acid to elute without difficulties.
I do not recommend CN column for HILIC, for a multitude of reasons. The most important one is that they don't work well in HILIC. They are not polar enough.

Mike,

I understand that options are relatively limited when trying to use ion exchange in conjunction with MS, which is why I was suggesting use of volatile eluent ions such as formate or TFA (I'm not sure if TFA is volatile enough for your application). Of course, it is possible to use ion exchange with a suppressor device which removes the eluent ion and allows detection of analyte on a background of only solvent.

But, in answer to your questions above, if you decide to proceed with anion exchange on a weak anion exchange material, then yes you will want to perform a gradient from low to high concentrations of formic acid. Assuming you're analyte has a higher pKa than formic acid (which is most likely true since formic acid is a stronger acid than most carboxylic acids), this should be sufficient for eluting your analyte. The amount of solvent present is somewhat immaterial although one must keep in mind that both weak acids and weak bases are less ionized in low dielectric environments (i.e. when solvent percentages are high). It's difficult to predict the effect of solvent when the ionization of both the ion exchange sites and the analyte are simultaneously being affected by the solvent content, but it can be expected to be a useful tool in adjusting retention. However, I would recommend against a simultaneous gradient in both solvent and formic acid. Generally, ion exchange has a significantly higher focusing power and thus provides better chromatographic efficiency in gradient mode with a solvent level is kept constant.

What is MS Bleed ?

Satish,

MS is a very sensitive detection technique. If a ligand is not completely stable, its mass will show up as a background. This is not a problem when you use single ion monitoring, but it can be a problem if you scan.