only get one good injection

[sjmarin]

I have a separation that is just a mess. I am trying to do a separation of opiates. I have Waters 2790 and Quattro micro. I am trying to use an Agilent SB phenyl, 1.8 um, 50 x 3.0 mm. I am doing a gradient separation. No guard column for this column, they recommend using a 0.2 um frit in a holder.

I am doing a gradient separation. I make one injection, things look OK. I make a second injection, my early eluting peaks spread out and start to split, but later eluting peaks look OK. This gets worse and worse until I change the frit, and/or flush the column, then I get one more nice injection. Also, early eluting peaks are about twice as wide as later eluting peaks. This seems odd to me.

I have made fresh mobile phase and flushed everything out. Mobile phase is ACN:water:100 mM ammonium formate pH 3 (yes I know it would be better to mix the buffer in with the water and ACN and not have the pump do it, but I inherited this set-up). We have about a dozen other isocratic assays that use this configuration and they are fine.

I hold at 5:93:2 ACN:H2O:buffer for two minutes, ramp to 25% ACN over two min, hold 0.5 min, then ramp to 95% ACN and hold for 1 min, then back to beginning, with 3.5 min re-equil time. I have a 600 uL pre-column volume, which is the dumb name Waters uses for gradient delay (sorry Waters people, it took me two days to figure that out).

Any ideas what is going wrong?

[sjmarin]

OK, I have been thinking about this. We have a Nanopure water system. We don't filter the water before use, just whatever final filter it goes through out of the system. Is it possible that because I have gone to a smaller column and smaller filter that there is stuff in the water that is causing problems that we don't see with the larger particle size columns? Or because my other assays all use guard columns that are changed regularly?

[Kostas Petritis]

It can be injection solvent miss-match (what is your injection solvent) or maybe re-equilibration issues (what is your buffer and how much do you wait between injections)?

[sassman]

Kostas is right. Because you are using such a weak initial mobile phase (5% ACN), the chromatography will be very sensitive to having too much organic solvent in the mobile phase. It doesn't happen the first injection because the pre-run equilibration is not the same before the first run. If this is the case, you can either choose to increase the initial ACN concentration or use less organic solvent in your sample to make the sample better match the mobile phase.

Another possibility could be particulates from your sample. Are these samples filtered or otherwise clear of particles?

[sjmarin]

Thanks for the reply. Samples are prepared in Nanopure water. I also tried prepping them is 95:5 water:ACN. Same result. I was having trouble with split peaks because my autosampler syringe purge solvent was ACN. I changed that to water. I have a five minute re-equlibration time at 0.6 mL/min. Shouldn't be sufficient for a 3.0 x 50 mm column?

We do SPE to prepare the samples, elute with 25:5:70 IPA:NH4OH:EtOAc then evap and re-constitute with water. We don't filter before injection. We don't do that for any of our other assays, but our other assays use columns with larger particles, and all have guard columns. I have never tried using a 1.8 um column and just a frit before.

[sassman]

That should be enough time for re-equilibration. It is interesting that the early eluting peaks are affected more than the later peaks. This still makes me suspect a problem with the injection solvent or sample. Do you evaporate the sample COMPLETELY after SPE. Any residual ethyl acetate or NH4OH could cause problems, but I'm not sure why it wouldn't happen in the first injection.