Chromatography Forum

My dear chromotographers,
We have developed a linear gradient HPLC method with pH 3.0 phophate-TEA buffer as Mobile Phase-A and Mixture of Methanol:Acn (80:20)v/v as Mobile Phase-B.The wave length is Uv-210 nm.We have facing severe noise problems in the initial gradient , where the max. composition reahes to 30% of B.After 30 min the noise drastically comes down and the base line is clean and neat.The problem is all known impurities elute before 30 mins.The noise levels are varying from system to system.We have found during validation a good base line and hence no issue.But while Technology transfer to QC the problem hit us again.PLease advice if any body has quick solution without altering method conditions?Waiting for your valuable advice.

First thing I'd try is to premix the mobile phases.

We have tried pre-mixing without affecting the linear gradient composition.That did not workout as expected.

UV cutoff for methanol can be above 210 nm. Also depending on the source of TEA phosphate noise can come from dirty TEA. I also would suggest to check how long UV lamp was in service.

Thanks for the suggestion,Vlad.We have other methods with this bufffer that r still working fine even at 205 nm.We r much worried as the prodcut time lines r badly affected because of this issue at tranfer level.Is there any inherent flaw in this method ? This is what troubling our minds very much.We have good baseline some times and the repeat analysis on the same day with same grade/make of chemicals using same system , the noise is more pronounced.This is what the current status my project.PLease do share if u have any experience in the past.

When pH of your TEA phosphate buffer is controlled maybe a temperature problem in the QC lab can cause such noice. When you run this application in the morning and noice is ok, and in the afternoon, for example, you see this noice, what happened in between? Was flow stopped, or flow rate reduced, or do you use a solvent "recycler"?

Is it really noise that you're seeing or is it system peaks coming off the column? You might want to check this by the gradient equillibration test often described here, there's a good explanation somewhere by Tom Jupille.
If system peaks indeed is your problem, TEA is the likely culprit. The amount of system peaks will heavily depend upon the gradient used, so the fact that this buffer works fine with other methods is no real clue.