Strange baseline disturbance ionpair-RP method

[Mattias]

method:

isocratic RP method with ionpair (SDS)
mobile phase 50% MeOH in 25 mM SDS in phosphate buffer pH 2.

When I inject the same mobile phase that is running in the column, I get a stepwise increase of the baseline at approx k= 2.5 - that slowly goes back to normal baseline. This happens in every injection - and it is becoming a problem since I have peaks of interest in that region. The peak does not appear if I inject water - but my samples are dissolved in mobile phase...

Does anyone have a clue what is going on?

[HW Mueller]

Two things come to mind: a.The injected mp is different from your mp (air, evaporation, diff. prep.). b. Something wrong with your injection mechanism when inj. mp.

[Dan]

Mattias,

Two things I would consider:

1) Are you using a potassium phosphate salt for the buffer? Potassium + SDS = precipitate (in some instances). This may not be your problem as your mobile phase is your sample solvent, but, with the injection peak that you are seeing, it is something to consider.

2) Leave the SDS out of the sample solvent unless you need it there for a reason such as solubilizing the analyte. SDS in the sample solvent can mean problems/issues with foaming that makes the sample preparation troublesome (or just a plain old pain in the butt). Injecting a solvent containing SDS can also be problematic for some injectors.

Regards,
Dan

[mohan_2008]

I agree with Dan. Take out the SDS out of the diluent.

Question: What is your operating wavelength. It looks like the SDS is absorbing at the operating wavelength.

If so, change your ion pair to SHS or NAPS.

[Mattias]

Thanks all!

The method is quite fixed I am afraid - so I am hoping this is an artefact I can correct without changing the method.

The method uses sodium phosphate buffer, and the wavelength is 325 nm. The UV-spectrum of the peak show no UV max (highest at 190 nm, and then slowly falling over the wavelength scale).

I am also thinking about air - but I have not been able to increase the peak by bubbling air through my samples. I will try to inject air (empty vial) and see where it ends up.

[HW Mueller]

That´s exactly the spectrum I got when I had air problems. (Hopefully to prevent a repeat of an earlier discussion: This is not the spectrum of air, but something caused by air, not an air bubble). It is not necessarily easy to prove an air problem. If you inject too much air you may get spikes, etc. etc. At least in one session, when I got somewhat broad extra peaks, that looked like what I believed to be a dissolved air peak, I saw (in a teflon capillary) foam emerge from the detector each time I got this peak.

[lmh]

What's your needle wash? Does your system keep needle wash away from the sample that is injected?

[Mattias]

I use an Agilent 1100 system, and I have no (external) needle wash at all.

Injecting air was not a very good idea - I got a huge peak (over range) after a couple of minutes. Not easy to prove that this is my "artefact"..

Maybe I could try to increase the pressure over the cell? If it is air/airrelated, an increased pressure would reduce the problem? I will give it a go.

[HW Mueller]

The problem will not be reduced if the air is dissolved. I warned you . . . . wait until you think you got it, do it once more and have the peak change shape and/or position, or even disappear into to (tm). Another approach is to try to keep the air out of the injection process, also not always easy. I have introduced small amounts of air into the system by just pushing a filled syringe into a manual Rheodyne (can easily be eliminated by overloading, but syringes also have a tendency to mix in air with the sample).

[Bruce Hamilton]

Is your mobile phase premixed and degassed?. If not I would try that first, preferably using ultrasonics for degassing.

Is your column at ambient temperature?. I if it's above ambient, bubbles are more likely as gas solubility decreases.

If mobile phase is premixed and degassed, i would try slightly warming the sample vials and degassing them - just to see if they are the problem, rather than the mobile phase..

Please keep having fun,

Bruce Hamilton