Strange Peak From Fluorescence Detector

[steve2]

Hello everyone.

I have a problem with the fluorescence detector from Waters that my company purchased a few months ago. It is connected in series (as a system) to a UV detector and 2690 Separations module. The flow from the 2690 is first through the UV detector (2487 dual wavelength absorbance detector) and, then, through the fluorescence detector (2475 Multi λ), which then leads out to waste.

I am running 100% MeOH as the mobile phase and using 100% MeOH as the test sample. I'm trying to minimize the variables within the run. The peak from the UV detector is as it should be, but the peak from the fluorescence detector is inverted (looks like an aqueous inversion). The two peaks (one from the UV detector and one from the fluorescence detector) are mirror images of each other.

What is going on? I've tried for a week to get rid of the inverted peak, but nothing seems to help. There is NO water running through the system.

The fluorescence detector's parameters are all defaults, except for the excitation and emission λ, which are:

ex λ: 254
em λ: 402

This is a recent development. It was fine until about a week ago. I've tried 3 different columns (all the same - from the same manufacturer). All 3 columns are:

Symmetry C18 RP
4.6 i.d.
75 mm length
3.5 um particles

Any help would be greatly appreciated.

[Uwe Neue]

You are injecting methanol into methanol. You should not get any peaks either in UV or fluorescence, unless there is something in your injector or your sample vial.

Since you get a peak in the UV, and this does not bother you, why does it bother you that you also get a peak in the fluorescence? The peak is negative, which corresponds to the fact that something in your sample is absorbing the excitation light. The mere fact that you get a negative peak either means that there is a small fluorescence background in your methanol or that you are seeing straylight. In either case, if the excitation light is absorbed, you get a negative peak.

[steve2]

Uwe, I was getting positive peaks for both UV and FL until about a week ago. But, from what you're stating the negative peak would be expected rather than a positive peak, or no peak at all. Now, I'm confused. It was stated to me by the Waters technicians that the peaks were normal and that they both should be positive. I'm still pretty new to HPLC, and I'm learning as I go.

I don't have anything, other than the MeOH, in the injector or the sample vial - at least that I'm aware of. I've been very careful about this. Everything running through the HPLC and the detectors should be 100% MeOH.

[Uwe Neue]

Injecting methanol into methanol, I would not expect any peaks at all. What sensitivity are you using? Occasionally, you may collect something from your vials etc...

Honestly, once you inject real samples, you will always have something eluting with the solvent front, and you will rarely worry about it any more.

[steve2]

Uwe, the sensitivity is defaulted to be fairly high. So, it might be that there is nothing of consequence, and I'm making more out of this than is there. I guess I'm a little edgy about making sure that everything is working as it should.

I'll run an anthracene standard to check the FL detector. That'll be a better indicator of problems anyway.

Thank you for your help.

[HW Mueller]

What happens if you just actuate your injection valve without injecting anything?

[steve2]

What happens if you just actuate your injection valve without injecting anything?

I haven't tried that, yet. What would this tell me?

I've run the standards for the FL and have gotten exactly what I'm suppose to get. The peak and RT for the standards are right where they are suppose to be.

[HW Mueller]

It lets you know how much of those peaks at ~to (tm) are due to the short flow stoppage of injection.
Also, if the negative peaks in fluorescence are large it should ring the alarm bells (you will get a "feeling for what is a negative peak if you watch things carefully). You can only get negative peaks if your mobile phase fluoresces (MeOH should not) or scatters light (everything does to varying extend, but relatively weakly with small molecules).
You may still get very good standards with badly upshifted fluorescence baselines, your robustness, sensitivity, and linearity may suffer, though.

[steve2]

Well, I actuated the injection valve, basically injecting nothing into the mobile the phase, and I get similar, but sightly smaller peaks. The peaks I was getting before were very small. Now that I've had a chance to think about this a little more, it seems these peaks may just be noise. These detectors seem to be more sensitive than I expect them to be.

I apologize for being overzealous.