Why does neomycin require gradient for peak?

[MestizoJoe]

I have been working on neomycin for a few weeks and have only been able to obtain a peak when a gradient is used. The simplest method was 0.1% HFBA on a C18 using an acetonitrile or methanol gradient. I have been trying to get it to elute using HILIC, so far with no success.

All published methods I have found use gradient. Can anyone please explain why?

I have seen this before with PHMB. I assumed that in the case of PHMB it was because the different species (PHMB is a mix of various lengths of polymer) were eluting at different mobile phase composition, making isocratic elution of all peaks either impossible or a very broad peak. Neomycin is not a polymer.

[tom jupille]

Interesting! Gradients with ion-pair reagents are always dicey because the equilibration between the IP and the system changes dynamically during the run. If this were my problem (and I had the time to pursue it), I would try a series of runs where I gradually narrow the gradient range (thus making the gradient more shallow) and see what happens to the peak shape and retention (does the peak just get broader, or is the retention critically dependent on % organic?).

[Andy Alpert]

Neomycin and related aminoglycoside antibiotics represent an extreme case of hydrophilicity and charge density ((+) charged analytes are the most hydrophilic, and these compounds can have as many as 5 (+) charged groups in four sugar residues). Consequently, they stick like glue in HILIC. In order to get them to elute reliably in HILIC:

1) Have 125 mM ammonium acetate or formate in the mobile phases;
2) Use a HILIC column with the lowest surface area available. Our customers found that our 1000-Å material worked well.

Contact me offlist if you'd like a pdf version of a paper in which these compounds were successfully analyzed by HILIC with detection via ESI-MS.

Andy Alpert
PolyLC Inc.
aalpert@polylc.com

[MestizoJoe]

Thanks for the responses.

I have to add a correction. I do not know that a gradient is needed using C18 and HFBA ion pairing. I am only referring to HILIC methods I have found in the literature.