Chromatography Forum

Hello all,

we are observing strange up/down steps at seemingly irregular intervals in our HPLC chromatograms. We are not sure what could cause this. A periodic seal wash is one idea but there is no such setting in the method for this.
The magnitude is about 0.0003 AU.

it is an isocratic method. The steps can disturb the auto integration if they coincide with a peak of interest.


and


Thanks for any input!

I don't have any advice, took all my brainpower to get to the chromatograms to show. We never used such level stuff.

Thank you for helping out with the chromatograms. I was unsure if the image hosting service was right. Maybe the next person will have some advice.

I was unsure if the image hosting service was right.

the picture hosting should work well. Use the provided BBCode from the ibb-page instead of only pasting the URL in [img]-tags
You can edit your initial post and replace the neat URLs with the BBCodes.
(for me, the pics from CPG are just the small images and not showing the whole pictures.

On your problem I don't have a clue neither but I will share my thoughts for troubleshooting.

First, please give more info about the system and method. Few of us have a crystal ball on the desk...

Some questions:
- is the mobile phase premixed and 100% of one line is used? (I guess so)

- is it a PDA or a Tuneable Wavelength detector?
-- if PDA or DualChannel: are the steps also seen in (all) other wavelengths?
-- > maybe the use of a reference wavelenght could then help
-- > or it would direct the problem to the detector itself, I guess.

- what about fluctuations in the electric current/voltage, that may in-/decrease the brightness of the lamp? Or a cooling fan or something like that? Maybe some air-conditioning or heatings of your whole building?
-- > can you put the detector on an stabilised power supply, e.g. on an UPS should filter fluctuations?
-- > can you run and record a prolonged run e.g. 3-4 hours to get an idea of the real frequency of the issue, not interupted by multiple injections that always autozero the detector?

- is there any battery on the mainboard of your detector/hplc, that need to be replaced?
- Or maybe some other capacitors on the mainboard are gone, which normally would filter these low frequency cycles

- are the bumps also seen with pure solvent?

As told, just guesses. Not sure how probable they are.

I dont know if you have found an answer yet, but I have seen similar baseline steps when a VWD UV lamp is beginning to fail. We had an isocratic system that showed the same stepping as you pictured, and a lamp replacement corrected the problem.

We never really figured out what the problem was, so thank you for coming in. We’ll change the lamp and I’ll report back.

Can also be caused by heavy placebo/matrix loads injected or contaminations left from previous runs.
Is the column dedicated to a single analysis? Is there a problematic matrix?
You might perform a multiple injection without performing autozero at injection to get an idea to what level the baseline returns. If you receive overlapping main peaks you probably have a contamination on column causing the baseline disturbance.