Troubleshooting baseline in rp HPLC..

[sylf]

Hi all,

I am having some baseline noise in my HPLC experiments, and as I am not any kind of expert on HPLC, I hope I can ask some of you for advice. Basically, I am using a dC18 column to separate a sample of nucleosides in a potassium phosphate (pH 4.5) buffer using a methanol gradient (2.5-16%). The problem is that the baseline is not as good as it used to be, with lots of little random noise throughout the run. I highly suspect this is due to buildup of contaminant in the column, as if I do the run without the column in line, the baseline is somewhat better (but not perfect). The UV lamp is quite old (~1500hrs), but if I turn off the flow, the baseline is stable. The column is probably just over a year old, but we didn't have a guard column. I've tried washing the column through several times with 100% methanol, 50/50 methanol and water (maybe for 1/2hr each), but it hasn't really improved yet.

So, my question is 1) does anyone have any other ideas on what this might be caused by, and 2) what else could I try washing this column with?

Thank you for any help in advance

[Uwe Neue]

It would be good if you could show us a picture of the noise. Random noise in the base line is not likely to be a column problem, but a problem of the mobile phase or the detector. We should not always blame the column for everything...

[sylf]

Ok, here's what my injections used to look like:


And here's now it looks now (excuse that the scale isn't precisely the same, but you get the idea). Not very nice.



I usually filter my buffers, and I still make it up the same way. Unless my potassium phosphate is going off? O_o

[Kostas Petritis]

Are you sure that the baseline does not look similar if you scale the first chromatogram to 15 mAU max?

[vicou_chen]

If the problem dose not underlie within the chromatgram scale,I think it may be caused by your mobile phase or your pump state.So I want to know in which way your mobile phase is mixed,manually or by mechine-driving?In addition,there may be some trouble with your pump... so sophisticated!

[HW Mueller]

The second chromatogram appears to be a 5-fold expansion in scale, could be that the baseline is even better than in the first chromatogram?

On the side: phosphate at pH = 4.5 can not be considered a buffer, and integration is set wrong.

[danko]

Correction: The fist chrom. is zoomed to approx. 0.0075 AU, whilst the second is zoomed to 0.015 AU (i.e. if they’re normalized, the difference will be even more pronounced).

Best Regards

[HW Mueller]

Power of suggestion? On second look it is the first chrom. that is more expanded, so the second chrom. does have a noisier baseline! Sorry.
If all the chromatograms and corresponding injections are reviewed the cause for it might become apparent?

[unmgvar]

sylf,

are those pictures of the same type of sample?
it does not look the same at all. especially the dwell volunme area.
we cannot be sure that this is someting maybe from your sample preparation.

it is possible that your pump has gone bad and the gradient is not done correclty. do a gradient check.

[sylf]

@vicou - Well, I mix the mobile phase using a magnetic stirring bar in a beaker when I'm making it up, usually for about 10-15 min in 2L volume. Then, I vacuum suck it through a 0.2uM filter, which I presume also helps mix it up.

@ unmgvar - It's the same type of sample, yes. I know the retention time seems to have drifted, but there was a long time between these two runs (like, 6months). As for sample preparation... This is an enzymatic digest of DNA into nucleosides. I somewhat suspect I should have used a guard column now..


If it is the pump, as some of you think it is, how would I know?

[HW Mueller]

To reduce exposure to dust it is better to mix solvents in a closed (screw cap) bottle. In my hands microfiltration was not a good degasser (especially if the liquid was transferred to a different container after foltration), it also should not be used for mixing. Thus it might be worthwhile to properly degass your mobile phase.
Anyway, this sort of problem has to be approached systematically, my recommendation from above still stands. Also, if there was only partial improvement when running without the column you already know that you have two "dirt" sources, the column + ? (Caution: If air or dirt in the mp is the culprit the baseline can look different without a column, even if the column is/was clean). Normally, dirt in the mp doesn´t show if the system is at equilibrium and the mp well mixed, the rise in baseline is normally not seen because of auto-zero, etc. Problems with the pump are evident via pressure and flow fluctuations and retention time, etc. problems.

[tom jupille]

HW Mueller is right in that you have to approach the problem systematically. Since you are running a gradient, there is a strong possibility that the problem is coming from the system or your solvents rather than the sample.

If you run a blank gradient (no injection, or simply inject your initial mobile phase), what do you see? If you see a pattern of noise, drift, & garbage peaks similar to your second chromatogram, then you have an instrument problem. If the blank looks like your first chromatogram, then you have a sample problem.

If it's an instrument or solvent problem, look at your pressure, as HW suggested; if that's fluctuating, then you have a flow problem and your pump or degasser needs attention. If the pressure is OK, then run the "gradient garbage" test for contamination in the "A" solvent (it's described on our web site here: http://www.lcresources.com/resources/TSWiz/hs400.htm ). If it's not garbage in your solvent, the next place to look would be your detector lamp.