Baseline issue - negative/positive peaks

[ksharp]

Hi everyone,

We are running a RP method for 6 synthetic peptides on an Agilent 1100.

Method:
150 x 4.6 5um C18 column @ 50°C, UV (DAD) @ 215nm BW 4 (ref: 360nm BW 100), Flow is 1 mL/min from a quat pump using A and B channels
MPA is 10%ACN in H2O + 0.1%TFA, MPB is 40%ACN in H2O + 0.1%TFA
Gradient is 0-35 minutes, 20%-90%B

Normal system suitability standard comes out like this:


There is a intermittent problem where the baseline fluctuates to varying degrees, dropping and then rising and then settling back down (see 4-13 min.)


It occurs at different times of the method, without any predictability or certainty. It happens in both standards and samples. Both of these chromatograms came from the same vial, about one hour apart. I suspect it is the instrument. There is no pressure drop or leak correlating with the noise. The instrument was recently (March/2013) put through OQ/PV and all tests passed except for gradient composition (A/B), which was resolved after cleaning the MCGV.

I welcome any suggestions and advice

Thanks
KS

[tom jupille]

Check to see if a reference wavelength has been set. If so, turn it off. You can get negative peaks on a DAD if something elutes that absorbs at the reference wavelength. If the problem goes away, then try resetting the reference to a different wavelength.

[ksharp]

Hi Tom, glad to have you weighing in here.

Yes a reference wavelength of 360 nm with bandwidth 100 is set. Next run I can have the analyst repeat with referencing off as you suggest.

It seemed to me that the way it is fluctuating down and up resembles something you normally see at the dead volume (air peak? sample solvent?). That made me think it might be a solvent mixing issue in the pump or something...like the -/+ absorption is due to the difference in ACN% or refractive index. Any thoughts there?

Thanks again,

[sepscientologist]

Is the disturbance always near the dead volume?

Do you see it with running a blank?

Could you try varying the dwell time at starting aqueous to see if it is in there?
ie... immediately start gradient maybe you get no disturbance dwell at aqueous
big disturbance?

The reference off was my first thought too.

[ksharp]

Hi,

In the example shown earlier it is a mess from 4-13 minutes, but the issue has presented at all different times during a run. It popped up many other times during this sequence: once at 32-35 minutes, once at 18 minutes and 35 minutes...in one run I many instances of this kind of noise throughout the run - here is another look:



It also happens in blanks (mobile phase A), standards and samples. Seems independent of sample composition.

Thanks,

[mattmullaney]

Perhaps...could this be bubble(s), tiny ones that are arising from micro-leaks (system was recently serviced--maybe not all fitting are quite tight enough) and sticking in the detector cell to different degrees? Maybe putting a knot or pressure restrictor after the detector...taking care not to overpressure the cell, or simply reducing the ID of the tubing a bit leading from the detector and the waste container...after checking the tightness of the fittings, could clear up your troubles?

[tom jupille]

The fact that it's not consistent argues against a reference wavelength problem. Lamp fluctuations or gradient proportioning problems would be the next most likely suspects, but you've already eliminated those. Bubbles, per mattmullaney's suggestion, are a distinct possibility. Flow fluctuations (perhaps caused by air bubbles in the pump) are another. In that case you would expect to see retention time variability as well. You might want to check that your degasser is functioning.

Aside from that, I'm out of ideas.