Baseline Rolling

[dwilliamson2]

In our laboratory, we run gradient methods with water / methanol or acetonitrile with no acid modifiers. At present, we are having problems with ripples in the baseline at the end of a 10-100% organic gradient method.

The problem occurs at the time point corresponding to 100% organic and consists of an abrupt negative baseline drift followed by a slight positive excursion ending with a stable baseline. The problem is almost eliminated by changing the final gradient composition to only 90% organic and is much smaller with 95% organic final composition. In addition, the observed ripples are much worse with a water / methanol system.

Is this related to pump seals or gradient proportioning valve? We are running an Agilent 1100 series binary pump with detection @ 210nm and 254nm.

Any and all comments are greatly appreciated. Thanks

[tom jupille]

Do you get the same amount of ripple at both wavelengths?

[dwilliamson2]

The effect is the same at both wavelengths, but, is more obvious at 254nm since the overall baseline rise is much lower and so these ripples (~2mAU) are much more obvious at this wavelength.

Thanks very much for your query.[/b]

[tom jupille]

Actually, that wipes out a couple of hypotheses.

If the effect had been more pronounced at longer wavelength, that would have pointed to RI effects and a mixing or proportioning problem. If more pronounced at shorter wavelength, I would have suspected a contaminant in the sample. That said, I still tend to agree with your initial supposition (pump hardware problem).

Do you see the same problems with "dummy" gradients (i.e., no sample injection)? That would at least let us rule out sample-related problems. And if you can trigger a dummy gradient without making the injection valve turn, that would let us rule out pressure shifts or garbage from the valve.

[Roy]

Have you been able to solve the problem?
Have you done the experiment Tom suggested?

Thanks,

[dwilliamson2]

Hi-

The methods we are running do not use the injection port valve which rules out injector peak contamination. The "peak" is more of an up and down wave immediately after the gradient reaches 100% organic. The wave is minimized or eliminated at 95 and 90% final organic compositions, respectively. I have played with sovent compression settings and pump cycle volume parameters and have had no success. For normal HPLC work the magnitude of the wave (2-3 mAU) would not be a problem, but, we are trying to characterize low - level impurities in solvents.

Thank you very much for all of your thoughtful responses!

[unmgvar]

dwilliamson2

what is the frequency of the waves?

i have seen on some instruments that you some time need to set a faster or slower opening/closing speed of the proportionnong valves.
it looks to me like what you describe is the same behaviour. check with agilent, maybe you need to change the firmware of your pumps.
i don't know what causes it but it seemed it was location related. one lab had the effect very strong, and another one, 200 km north of it with the same pump models the effect was insignifcant and necessited no firmware change.

hope this helps

[unmgvar]

dwilliamson2

what is the frequency of the waves?

i have seen on some instruments that you some time need to set a faster or slower opening/closing speed of the proportionnong valves.
it looks to me like what you describe is the same behaviour. check with agilent, maybe you need to change the firmware of your pumps.
i don't know what causes it but it seemed it was location related. one lab had the effect very strong, and another one, 200 km north of it with the same pump models the effect was insignifcant and necessited no firmware change.

hope this helps

[cafran]

We've seen this in our lab too. This seems to go away if we degass all of our mobile phase components by sonication under vacuum.