Baseline Drop on one system only

[Shouldknowbetter]

Hi Everyone, this is my first post so maybe you can help

I am currently running a gradient method on my LC that essentially runs from a mobile phase containing an aqueous buffer and acetonitrile to one that consists purely of acetonitrile. Detection is UV (above 260nm)

The problem is this. Normally the baseline for this method is either totally flat or we see a slight rise. (this has been so on three different instruments) When trying to use this method on another instrument we are seeing a decline in the baseline due to the gradient.

So far I have tried using different batches of mobile phase that I know give "normal" results on another system, different lines for the aqueous and acetonitrile phases (its a quaternery pump) and even making everything up from scratch but to no avail.

Any suggestions?

[Bruce Hamilton]

My first guess would be that your detector is seeing a different part of the spectrum, or lower energy. Maybe the bandwidth is wider, the reference is set incorrectly, the wavelength calibration is wrong, the lamp is old, or the cell is dirty.

My second guess, less credible, would be that, if you are using a vacuum degasser on the instrument, the channels have been contaminated with something that is mobilised by the buffer.

You can probably check that by allowing your instrument to running at the final mobile phase composition for an hour or two, then quickly revert to initial phase, and immediately perform two runs in succession, and look at the baseline slope. If it changes, you have contamination, if it's constant, you have an instrument setting differece.

Please keep having fun,

Bruce Hamilton

[HW Mueller]

Maybe you just got your plumbing mixed up and are starting with ACN and going to H2O?

[Sentenced]

I had the same problem before once...
It only happened to me with an old Perkin Elmer HPLC, and the problem was the binary pump (a PE250).
In this particular case, the only way to solve it was with a higher flow where the slope begins. For example, if it was 100% A for 15 minutes and then a 10 minutes gradient to 0% A, everything at flow 1.0, and the slope begins at 17 minutes, try with a 1.2 flow for the first 2 minutes of the gradient, and then back to 1.0 or 0.8 to compensate...
I know, it´s not the best way, but sometimes, some pumps needs a little pressure "punch" to "wake up"

Good luck with that, see you!