Gradient Baseline problems

[HPLCissues2014]

Im looking for some advisc on how to make my baseline more stable for gradient analysis please.

Some background: Im using a buffer with sodium dihyrogen orthophosphate (ph to 3.1 with ortho acid), my needle washes are varients of Organic and Water. Its not a great method but it has worked fine in the past, my issue is the second half of my 4 minute injection gets very noisy and the peak tends to either split when it comes off with the background noise, or else have massive tailing. The first half of the injection is fine, its just goes very wrong during the second half. I have checked and triple checked the method so my gradient and wavelength is right. I have swapped over mobile, washes, and columns from a back up machine which is working fine but the problem was still there. The flow cell was washed out with water, acid and flushed again with organic, the problem cleaned up a bit but not to the extent i need it to. Is there anything else that can be done to improve the baseline. Note there are no leaks on system.

The only thing is that I usually dont clean out the columns with organic etc before use, i just start using it straight way could this be a factor?

[tom jupille]

Do you see the same problem when you run a blank?
http://chromforum.org/viewtopic.php?f=31&t=19085

[danko]

At one point during the gradient – at the high B-eluent % - the mixture (A + B eluents) becomes cloudy due to solubility limits of the phosphate in organic solvent and that is what the detector translate to absorbance hence steep rising baseline and much noise.
Maybe the buffer is a little bit more concentrated than normally – due to weighing discrepancy or something like that – and the critical line is quite narrow.

My advice: Try and weigh 10% less phosphate or just dilute the buffer a little bit with water and try again.
Best Regards

[HPLCissues2014]

To answer the first post- yes, the noise is in the blank too which suggests its machine-based, and thanks for the link, its has given me a few possibilities as to what is causing it. Thing is, it works perfectly well on another machine so Im trying to eliminate the causes. To the second poster, I never really thought of buffer amount. I would tend to go a bit over rather than under i.e. if i need 5g i would do 5.2 or 5.3g. I will try the 10% to see if it makes a difference. Can different HPLC machines be very sensitive to the same method such that these changes would make a difference? I wouldnt have thought so but Im not experienced enough either way to say so.

All I know is the flow cell has been thoroughly cleaned out with acid, and washed afterwards with Acetonitrile and Water. I have used a brand new column as well, maybe the machine still has some rubbish in it from the acid wash and needs a longer prime/equilibration? Any other suggestions are welcome because its still not perfect for what I need it for, thanks.

[bunnahabhain]

Can different HPLC machines be very sensitive to the same method such that these changes would make a difference? I wouldnt have thought so but Im not experienced enough either way to say so.

You don't believe what can make a difference in HPLC...
I have once seen gradient problems on an Agilent 1100 using 50 mM K-Pi buffer and acetonitrile. When switching channels A-B to B-A, everything was fine. The thing is, that channel A sits below channel B. Our problem was, we used phosphate buffer in the lower channel, ACN in the upper channe. This led to problems most likely due to micro-precipitation resulting from poor mixing performance. As soon as we switched phosphate to the upper channel, everything was OK because the precipitate redissolved during it's travel downwards.
Anyway, it is always a good idea to flush columns with high organic before starting your day...

Regards
Jörg

[danko]

To HPLCissues2014,
Jörg has a good story to tell and I’ve seen strange things happen too – when changing to another system.
Usually I wouldn’t expect any difference in this context, between 2 systems especially if they are of the same brand and model. But as mention earlier the line between clear and cloudy mixture of the eluents you describe, is very narrow indeed.
So, mixing discrepancy, temperature etc. could make a difference.

Best Regards

[HPLCissues2014]

Hi everyone and thanks for the advice here is an update, I weighed out ten per cent less buffer double checked my methods to make sure I was missing nothing and then performed a warm water followed by organic wash down of all the lines to remove any potential buffer. Problem still remained I have a broad peak coming off at the exact same time as my peak of interest and this peak is on the blank the standard the samples so does this mean some old traces of the analyte remain in the system somewhere? My hot water and organic wash should have removed it from the lives and now the detector is fully washed out so that leaves the needle/washdown as the source, perhaps stubborn analyte.

So I inceased the organic in needle wash and also injected a few acetonitrile vials on to the system to maybe dissolve it upstream but it has only made a tiny difference can anyone advise as to the best way to remove (what seems like) carryover of old analyte or what the best thing going forward is? Not even washing and priming the needle several times helped. There is another substance which my analyte is fully soluble in if I need to use that.

[Hollow]

Can you provide more details (system, gradient, eluents etc.) and maybe a picture of good and bad chromatograms?
do you record the system pressure as well? How do they compare on good and bad system? What does the pressure do when the noise starts?
- maybe some issue with check valves?
- degassing air at a certain point in gradient?
- have you switched the channels for test purpose e.g A<>B or even A/B>C
- you're saying it occurs also in the blank, so maybe try pure solvents (without any buffer salts) to check if effect is still there to exclude precipitation effects