Chromatography Forum

Hi All,

I am measuring catecholamines using electrochemical detection (pump and EC both Waters). Over the past few weeks, the peak quality has dramatically decreased. I am trying to determine if it is a column or electrochemical issue. Both the columns I have tried are ~4 years old and produce similar results at this time.

I have done the following to no avial:
1. Polished the cell (glassy carbon working vs. ISAAC) and changed the spacer.
2. Made up new mobile phase.
3. Made up new standards.

The back pressure is stable and I have used the same mobile phase recipe and EC settigns for years.


Thanks for any help,

Jason

The drift could be a detector issue, but the peak shape problem is a textbook example of an "inlet flow profile" problem. All your peaks show the same kind of shoulder, which suggests that the problem occurred *before* separation. If the deterioration has been gradual, the most likely cause is a void space at the column inlet. It it came on abruptly, a partially plugged inlet frit is more likely.

Try turning the one of the columns around and backflushing it for about 15 minutes. If it's a blockage, that may clear it. If it's a void space, that may fix it temporarily, but it will return, in which case a new column is probably the best solution.

Turns out it has something to do with the guard column. I removed the guard column and the peaks look better than ever. The peaks are narrower with a much greater heght than I have seen for a given amount. Wierdly, when I reconnected a new guard column, with a clean holder the problem returned - not sure how that can be. I am considering running without a guard column full time and chucking the column as performance deteriorates - I always filter my samples anyways.

A few possibilities come to mind:

1. you may have gotten a bad batch of guard columns.
2. The cartridge holder may be tweaked so the cartridges are not fitting properly, giving a void space
3. The fittings on the cartridge holder (or the corresponding tube ends) may be the wrong "stop depth" so that you are getting a void space inside the connection.

Guard columns are intended to take out chemical contamination, so filtering the samples won't help in that case. Depending on what else is in your matrix (and how extensive your workup is), you may go through analytical columns fairly quickly.

Thanks much for the comments. They have been very helpful.