Dear Forum,

With this forum's help, I created a method that retained a small polar peptide on a silica colum (see the discussion "HPLC conditions for basic compound?").

I was able to perform a study in monitoring the concentration in mouse serum using:

Absorbosphere Silica column 4.6mm * 250mm 5u
mobile phase isocratic 75% aqueous (0.1% acetic acid and 3mM CsCl) with 25% v/v MeOH

I packed the column into storage with chloroform for several months.

Upon unpacking and resuming some work, the column no longer retains the analyte. Upon purchasing a new Phenomenex Luna 2 silica column, the analyte still elutes in void volume.

I can get a little retention upon lowering the CsCl to 0.5mM without much tailing or reproducability issues.

I have tried several ways to activate the column. The most rigorous was with 1hr flow with water, ~2hr flow in water with 200mM CsCl and 0.1% acetic acid, followed by another water wash, and overnight at half flow (0.5mL/min) with 10mM CsCl and 25%MeOH. A gradient array of CsCl has not shown any improvement.

* Can someone assist me in how to activate a silica column for HILIC separations? Could there be another reason on why this isn't working on the old column or new one?

I am still using HPLC bottled water, and have started over with a new batch of analyte. The solvent frit is stainless steel and was replaced recently (is that a problem?).

thank you for any suggestions or information on how to solve this non-retention problem.

HILIC mobile phases typically have more organic solvent than aqueous stuff. Are you sure you wrote down the correct solvetn composition? Maybe it was 75% methanol and 25% aqueous. I would try this first before doing anything else.

thank you for replying, A. Mouse.

Yes, this is atypical. This method does use mostly water than methanol and is isocratic. Traditionally, it gives a retention time for the analyte of 7 minutes (internal standard 5.5minutes, mouse matrix 3.7 minutes). Higher organic content pushed the retention into long runs with peak widths of over 3 and 4 minutes. This can be countered with higher salt concentrations, however, the analyte is virtually insoluble in anything except water, and I am uncomfortable mixing higher salt conc with high organic...ppt??

The main concern is why did this work so well in March, but now everything elutes in the void volume; on both the previous column as well as a new one?

It is very unusual to get retention of a compound under the circumstances that you describe. It can not be the polarity of your compound - I do not know anything that would be retained on silica in this mobile phase by polarity. Your compound must have basic functions, i.e. it must be retained by ion-exchange. There should not be any magic in this.

I still bet on my first suspicion. It is easy to put the solvent lines into the wrong container. Have done so myself...

Yes, the peptide is ~700 Da and has 4 basic amine groups. Ion-exchange is what I'm attempting to use.

I've switched back to the old column and it is retaining the analyte: however the new column doesn't.

There's no probability that the mobile phase make-up is inversed. Since the method was isocratic, I would make the mobile phase up in one container and run 100% solvent A.

So now it will work with an old Absorbosphere 5u Silica column, but not with a new Phenomenex Luna 2 5u silica column.

I switched brands b/c I wanted a column that was known for it's robustness. My boss advises me to purchase a new Absorbosphere and try it with that.

NEW QUESTION: What is the main difference between Luna 2 silica and Absorbosphere Alltech silica columns? And, when a silica column breaks down with age and use, does it break down into something better suited for HILIC?

Could this old abused silica column have so much "bio-gunk" that's it's actually the contaminants that are doing the chromatography??? LOL!

Michael,

There is always the possibility of permenant modification of your stationary phase, lets say by permenant adbsorption of significant amounts of a compound. However, (unless it is a controled process like some kinds of ion-pairing chromatography) this leads to useless columns as you will never been able to reproduce your results (unless you trace back and understand how you modified your column etc).

HILIC in general needs more than 50% of organic solvent in order to work properly (but it might work with lower concentrations if your compound is too polar).

I would agree with A.Mouse that ion-exchange or even ion-pairing chromatography with some heptafluorobutyric acid in the mobile phase would work well for your aplication.

OK, here is what I read:
You had developed the separation on an old Adsorbosphere. You put that into chloroform for storage, and now it does not work any more.
You tried it on a new Luna, and it does not work.
Did you now get it to work again on the old Adsorbosphere (Sept 24, 7:25)? Or did I misunderstand...
Kostas may have the right idea... what did you do with the old Adsorbosphere before you got the separation going? Did you play with surfactants (=ion-pair reagents)? They may get stuck on the silica by hydrophobic interaction (don't laugh yet). Since your sample is a peptide, maybe you played with HFBA or TFA that was retained on the surface of the silica before you washed it off with chloroform.
I also don't think that biogunk did it, but the last mobile phase may be responsible for the retention.

update on this research:

Thank you all for the feedback. Just want to touch base on what's going on at the moment.

A. Mouse yes, you read correctly. Old adsorbosphere, worked, stored, didn't work, works again (w/ increase of CsCl from 3mM to 8mM). Yes, it has been exposed in it's history to TFA and surfactants.

new luna column doesn't retain under above mobile phase conditions (void volume elution).

I have had multiple experts in HILIC chromatography tell me that it sounds as if some other kind of ion exchange phenomena is occuring with the old adsorbosphere. Nothing should be retained with the mobile phase I'm using in HILIC.

I am pursuing a more traditional HILIC mobile phase with the Luna column. So far, 80% MeOH and 20% aq (.1% acetic acid, 10mM CsCl) is showing some promising results (k'~ 2.5)...though there is a peak tailing issue. I do believe I'll get this worked out.

Addtionally, a new adsorbosphere silica column is being shipped to verify if the old method is truly plausible.

I will post my results when they are completed.

again, thanks for the assistance and any other tips will be appreciated.

Well, it is ion-exchange, not old-fashioned HILIC. The old Adsorbosphere could be more of an ion-exchanger that the Luna, but this does not explain why the separation disappeared after storage in organic. I think that it is some ion-pair reagent that makes it work. Go through all the steps in detail that you did on the Adsorbosphere after you took it out of the organic and then got it to work again. Somewhere in there is the trick that makes it work. Maybe TFA, maybe a particular "buffer"....

the final results:

when the new adsorbosphere arrived, it was washed (2-proH, H2O), exposed to 500 mM NaCl with .1% TFA for 30 min at 1mL/min (to mimic old column history) and then equilibrated with the mobile phase 75% .1% acetic acid 8mM CsCl and 25% MeOH over night at .2mL/min.

the peptides were retained.

Increasing the CsCl from 8mM to 20mM and running a gradient to 40mM CsCl gives baseline resolution of mouse matrix, internal standard, and both peptide analytes.

With this 19 min program I have analyzed over 300 samples to date.

Exposing the Luna column to .1% TFA and 500mM NaCl and equilibrate did not improve retention. Method development discontinued.

Apparently not all silica columns are the same. The adsorbosphere is better for IC.

thank you all for your assistance.

Thanks for getting back to us with your results! We are learning too.