wavy baseline on Waters 2695, checked other forum refs

[rich hammond]

I am new to using a Waters 2695. The instrument has an inline degasser. I'm doing protein runs using a YMC protein-pac column. A=0.1% TFA in optima water, B=0.1% TFA in optima acetonitrile. TFA is newly purchased from pierce. Flow=1mL/min, 214nm detection. I'm getting a 0.5 mAU undulation in my baseline during the shallow part of my gradient, with a periodicity of 0.5min. Baseline is flat for 100%A (and B and C and D). I ran the GPV valve test and the proportioning is very good (the instrument just had a PM checkup). Undulations present for other combinations of mixing (eg A&D, B&C, when checked with the 0.1% acetone solution used for GPV test). I manually degassed solvents (vacuum sonication for 1min) and reflushed system but no improvement.

It looks to be a mixing issue. Another forum ref mentioned stroke volume to help. I checked and it is already set to maximum value (130uL). Another ref concerned backpressure after detector but the 2487 detector is set up per waters, with metal outlet tubing connected to waste via teflon tubing.

Is this just standard? I never had this problem with the 1100 system. Thanks for any advice?

[unmgvar]

Hello Rich,

have you tested your instrument with the same mobile phase composition but without the TFA?

do you start at 100% water and finish at 100% ACN?
if not then premix your initial and final solvent composition. this could help you a bit.

also don't forget that you are working at a very noise UV range for your solvents. have you tried your OQ test at the same wavelenght in some way? drift and noise in the UV differ from the UV range used.

[rich hammond]

I have only used TFA. The gradient starts at 100%A and goes to 90%B. As you say, if I perform the runs at 280nm the baseline ripples go away. Of course as you say, I was anticipating a rising baseline with TFA at 214nm (especially since the TFA concentration is same for both A and B), but I did not expect the ripples.

I'm wondering about a pump seal problem now. The ripples have ~ 0.5min periodicity at 1mL/min. The proportioning valve test was performed at 2mL/min and the period is less, around 0.3min. I do not see a ripple at 100%B, should the ripple only be noticeable when there is solvent mixing?

[unmgvar]

Rich,

It is possible that your proportion valves are adding noise to the baseline.
i don't remember if for your instrument the proportion valves stay open all the time when you use 1 channel only. you might want to check that.

if they do stay open, you might want to see the rate at which your valves open and close during a gradient run. there is a correlation between that and the noise added to the baseline.
see if youre instrument does not need a firmware upgrade on the matter.

also see if you get an improvement by having a final solution of 10% H2O with TFA :90% ACN with TFA. see if that also help improve the noise issue. if it does then it is most probably an issue of mixing within the instrument that is not performed well enough.

[rich hammond]

I believe the valve stays permanently open with no mixing; at least I never hear the valves clicking. I'll check out the valve opening frequency in the ~35%B range that my protein elutes.

When I have done the proportioning valve test the value is quite accurate, under 0.1% at 10%. This is per Waters testing. I should have also measured at other percentages, and I didn't yet check a changing the gradient in steps.

Thanks for the advice!

[Chris Pohl]

Rich,

It's definitely possible for the ripple you described to be caused by incomplete mixing. I've seen this problem on several instruments. You can demonstrate that the problem is caused by the impact of incomplete mixing on the amount of TFA in the mobile phase as a consequence of oscillating solvent content by running a blank gradient with a piece of capillary tubing sufficient to achieve decent back pressure during the gradient. With a blank capillary you will see no ripple if the cause of your problem is incomplete mixing. The reason that incomplete mixing can cause the ripple you describe is that the equilibrium concentration of TFA in the mobile phase relative to that absorbed on the stationary phase is dependent upon the solvent composition in the mobile phase. If the mobile phase content in terms of solvent composition is oscillating, the TFA content in the mobile phase will oscillate in concert with this solvent oscillation. Of course, the simple solution to this problem is to add a mixer or you might be able to minimize the problem by adding some solvent to your 100% aqueous eluent and by adding some water to your 100% acetonitrile eluent. Usually, mixing is improved when both components are already mixtures of one another.

[rich hammond]

Thanks Chris, I'll try that test. I am seeing the ripple when I have had capillary tubing in place for the proportioning valve test, but that was all with mobile phases with different composition of the UV absorber (ie 0.1% acetone in B).

So it is not unusual for the GPV test to be tight, but to have this minor oscillation in solvent mixing? It seems strange that it can reliably get to 99.9% mixed and have this oscillation. Any thoughts?

[rich hammond]

Further comment & question to my last post:

When I was checking the waters manual concerning the plunger stroke volume, it stated to keep this volume large for solvents requiring additional mixing such as TFA gradients.

What is the issue with how readily TFA mixes? Perhaps I should read all of the posts on TFA in this forum!

[Chris Pohl]

Rich,

Yes, there's quite a bit about this topic sprinkled throughout various different threads on the Chromatography Forum. It's been quite a number of years since I use the Waters pump so I'm not sure what you're referring to when you talk about the plunger stroke volume. In general, though, when you are using a mixer you need to make sure the volume of the mixer is greater than the sum of the displacement of both pistons (assuming you're using a dual piston pump).

[Uwe Neue]

Part of the problem with TFA is its own background absorbance, part of it is the fact that it is adsorbed on the column. Thus small differences in the mobile phase concetration are amplified by the column, and this is the fundamental issue with TFA.

Anything that improves the mixing will help solve the problem.