Fluorescence detection: peak area issue

[LCFan]

Dear all,

since February, we use a fluorescence detector (Waters FLR).

I have reviewed the peak areas (corrected to the nominal concentration of the analyte):

Date / Area
08.02.2013 / 14755849
15.02.2013 / 13346913
20.02.2013 / 12779644
07.03.2013 / 12064618
28.05.2013 / 10574250

Is this a known phenomenon? And if not: where would you begin the troubleshooting process?

Thanks a lot for your help.

Florian

[mattmullaney]

Hi Florian,

This is kind of a long-shot, but this is similar behavior to what I've seen when the mirrors in the detector become "solarized" when they have aged to a point--there can be a precipitous drop in peak response almost all at once, seems like the response falls off of a cliff. If the detector is new, then obviously this guess is incorrect.

Another guess, could there be a relatively newly-developed leak somewhere between the column outlet and the flowcell, or within the cell?

Best of Luck with Your Troubleshooting.

[LCbob]

Hi

First off check the degasser on the LC is functioning correctly.

[LCFan]

Hello,

thank you for your help.

I will check for leakage and for functionality of the degasser.

By the way: how shall I do this degasser test.

Regards

Florian

[HippyLabRat]

Were you using the same sample prep for all of these injections?
(i.e. prepped once, injected on prep day or t=0, t=1 week post-prep... for a sample stability study)

Or are you preparing a new sample each time?

[LCFan]

Hello HippyLabRat,

after implementation of a new method, I set up an excel sheet in order to monitor the SST values (R, T, ...) and I also implemented a column for the peak area corrected to the nominal analyte concentration.

Thus, the 5 peak areas outlined do relate to 5 freshly prepared standard solutions.

(at each of these time points, sample quantitation led to the expected values = peak areas in the same range as outlined)

Regards

Florian

[mattmullaney]

For Waters or any other FLDs, if you open the detector, you'll be able to view the mirrors inside. If the silver within the mirror has "disappeared" and you can only see mostly the black color of the base that holds the mirror...this is what I meant by "solarized". The light impinging on the sample cell has finally driven off (ablated) most of the reflective portion of the mirror(s) to the point where the detector is just not responding the way it once did. If this is so, probably it would require a trip for the detector back to the manufacturer for a re-working of the optics, new mirror(s) and alignment.

My apologies...just figured that you'd have noticed by now a leak(s) and have evaluated your degasser.

In any event, Continued Good Luck with Your Troubleshooting!

[Mattias]

Have you considered the lamp?

The lifetime of a FL lamp is usually much shorter than for a standard UV-lamp. I think we need to replace ours (Dionex) after 500 hours. For UV the replacing interval is 2000 hours.

(These intervals are of course dependant on what equipment you have).

[Mattias]

... and solarization of mirrors happens over years, not weeks!

[mattmullaney]

Hi Matthias,

Point taken (that is why I suggested that this was a "long shot" above)...in either case, it seems there's an approximate 30% loss in response in a bit less than four months as measured by peak area. Could be a dying lamp as well as the onset of solarization--or leak(s) or an ailing pump/transducer (flow too fast and measured pressure not accurately measured). Not enough of a history in terms of peak area response as a function of time to tell. Time period of observed decrease in peak area response and age of detector as well as optics are both unknown.

The onset of confirmed solarization in the optical benches of two Waters 474 and 2475 detectors were similar--both happened after about four years of aging in both cases, both of these detectors were used fairly constantly (about 70% of total elapsed calendar time--these were both used for amino acid work, fairly high throughput). Generally we replaced the Xenon lamps every 1500 hours of operation or so...but this could be a "lemon lamp" in LCFan's case. Heck, I just saw a 1200 Agilent VWD lamp "burn out" after only 600 hours of operation, and those are generally fine for 1500-2000 hours of operation, to your point. Nearly anything can and will happen with any manufacturer's hardware.

Whatever the Case, LCFan, I hope your troubles are over and if so, please could you let us know what happened to the FLD? Curiosity is getting the Best of Me...

[elsico]

My opinion lamp. Solarization of mirrors? First time hear about FLD. Maybe new trick to take extra money for service from waters or agilent)
Watch about 30 fld detectors from 90's till now. My opinion most good successful models produce jasco.

[mattmullaney]

Hi Ivan,

Understood...I am certain then that as you are a student of the 1990's you'll recall that the Waters 474 Fluorescence Detector was manufactured by JASCO for Waters.

Light does cause silver ablation from a mirror surface over time. While it is popular, seemingly, these days to create the impression of antique mirrors via treatment with paint remover and nitric acid (for example), often mirrors that are truly antique arrived at that state through exposure to light over time. Fluorescence detectors employ high-intensity lamps (Xe in some detectors)...eventually the mirrors impinged upon from the light generated by such a source will create spots in the mirror where the silver has "evaporated" away.

No optics are proof to this phenomenon...and as Matthias points out, it takes years for it to happen even when the source of impinging light is a Xe lamp.

[LCFan]

Hello,

in the meantime, I checked my detector for leaks as advised. The flow cell is completely dry.

The lamp is new. Thus, its life time should not be an issue.

Do you have any advise how I shall check the functionality of my degasser.

Thanks a lot

Florian



BTW: I also contacted Waters directly with my problem today (irequest).

[mattmullaney]

Hang in there, Florian,

I'll be honest, I've not done this myself, check into the workings of a degasser. I suppose, though, you could remove the column from the flowpath, run a degassed eluent through the instrument (eluent of your choice, say 50:50 MeOH/water) with the degasser "on", and try to degas the "waste eluent" after the fluorescence detector.

If you observe lots and lots of bubbles when you attempt to degas the "waste eluent", then probably the degasser is on the fritz.

Best of Luck with Waters service!

[LCbob]

Hi

What degasser do you have on the system? Waters 2695 / 2795 have a feedback which will indicate if there is a problem.

How is your precision on the system day to day? There could be an issue inside the pump / autosampler causing the lowering of the areas.