Chromatography Forum

I thought I knew something about HPLC -PDA, but this one has got me.

Can someone please guide me in the general direction of this problem.

I have a HPLC attached to a PDA detector.
When I run a isocratic run with 65%MetOH I get perfect baseline, everthing is perfect.

As soon as i start running a gradient of low organic to high organic (10% MetOH -> 90%Metoh) I get huge baseline drift into -12000 mAU as soon as I introduce the High organic solvent. I know organic can do that because of absorbance but this much???

I've tried zero'ing the detector with 100%MetOH as instructed in the manual. But now it does the complete opposite. Huge baseline drift up to + 12000mAU.

the column I use was bought second hand but tested in our lab (Isocratic 65%MetOH) when we bought it and it seemed fine. Can this be the problem? bleed??
I Equilibrated the column before I tried running the gradient.

thanks in advance for your help.

Is the baseline rise due to the solvent effects at short wavelengths? If so, try collected data over a narrower wavelength range, maybe starting at 230 nm for example.

I really don't remember any problems using PDA attached to HPLC, maybe this is the problem, try attaching them this way

but seriously i remember sth simmilar a long time ago, but don't remember what was wrong - something simple

1. check if gradient is ok - A and B lines mistakenly changed with eachother?
2. check pda settings
3. change your solvents - purity issue?
4. i don't think the column has anything to do with it but if you like then run a gradient without the column
5. call your local support and cry

that should do it

Check the basline at various wavelengths.

If the drift is linear and worse at short wavelength, it's most likely a solvent purity problem

If the drift is non-linear (and especially if it comes down a bit at the end to look like a "hump") and is worse at longer wavelength, then it's most likely a flow cell alignment problem.

This kind of gradient drift problem is actually common enough that we've put together a "mini-seminar":
http://www.lcresources.com/more_resourc ... hp?f=3&t=5

You will have to register (free) to view the seminar.

thanks for the replies.

need to do other analyses for now but will play with it and keep you posted

help is much appreciated!

This kind of gradient drift problem is actually common enough that we've put together a "mini-seminar":
http://www.lcresources.com/more_resourc ... hp?f=3&t=5

Tom, you have a great resouce centre here, really helped a lot! altough i tend to loose track of time, working through them. kept me up till 1 am the other night!

Flattery will get you *everywhere*

Glad they were useful!