S-Shaped Baseline

[FSE]

I am running a gradient with 0.2 ml/min from 100 % H2O to 100 % methanole in 10 minutes. I always see a S-shaped baseline (254+-4 and 280+-10), it is going down and then rising overproportionally until it falls back to the absorbance level of methanol. I get this baseline with and without column. If I run a slower gradient the baseline is just strectched but still exists.

Instrument:
Agilent 1100 quaternary pump + online degasser, autosampler, DAD-detector, LC/MSD (ESI, APPI)
solvents: glass distilled grade (changing them did not help)
A: MeOH B: Dichlormethane C: Water D: Acetonitrile

The instrument is 6 month old and the pump pistons and seals have been changed by agilent (They thought it is a pump problem) - but it did not help.
Manually degassing solvents, playing with the stroke or compressibility did not help.
I see a similar behavior with acetonitrile instead of methanole but it is much less of extent (MeOH apex approx 80 mAU ACN apex approx. 4 mAU)
I am running out of ideas therefore I will be very happy about any suggestions. THX

[tom jupille]

First, if possible, run the same gradient on another instrument of the same model. If both show about the same effect, then it's a design characteristic of the system, and there's not a whole lot you can do.

If the other system looks OK, then have Agilent come in and service your detector, with particular attention to the alignment of the flow cell with the optical bench. What you're describing is consistent with a refractive index shift (non-linear, more pronounced at longer wavelength, more pronounces with MeOH than with ACN). To a certain extent, it's an unavoidable side effect of trying to squeeze as many photons as possible through a long skinny aperture, but a misaligned cell exacerbates the problem.

[FSE]

Unfortunately I cannot run it at another Agilent 1100. but...
I just found out that my channel B (dichloromethane) is leaking into the A channel (methanole) when I am working at low flow rates.
Could that be also the cause of the problem?
Thank you very much for your input - I really appreciate that.

First, if possible, run the same gradient on another instrument of the same model. If both show about the same effect, then it's a design characteristic of the system, and there's not a whole lot you can do.

If the other system looks OK, then have Agilent come in and service your detector, with particular attention to the alignment of the flow cell with the optical bench. What you're describing is consistent with a refractive index shift (non-linear, more pronounced at longer wavelength, more pronounces with MeOH than with ACN). To a certain extent, it's an unavoidable side effect of trying to squeeze as many photons as possible through a long skinny aperture, but a misaligned cell exacerbates the problem.

[Kostas Petritis]

Maybe...

In any case this is a problem that needs to be taken care as soon or later it can give you all sort of problems with your analysis (i.e. reproducibility issues etc...).

A quick fix would be to replace the dichlomethane with Methanol just to test if this is the source of your problem...

[syx]

Perhaps seal compatibility to dichloromethane is the reason. Years ago I had ever run my Shimadzu using mobile phase that contains hexane. The new seals were leaked quickly during analysis. It was shortest seal lifetime in my experience.