flow cell contamination

[blakeyk]

We were having trouble with the baseline on our relatively new Agilent 1260 FLD HPLC. We run an isocratic method using 55% HPLC grade water and 45% HPLC grade methanol. We use this same HPLC, method, solvents, etc.. at 6 other locations. Not only was the baseline very noisy, but the absorbance units for our baseline on the Y axis had began ~20 LU (at installation) and had increased gradually to ~70 LU over a 12 month period. There were no visible leaks in the flow cell. I ruled out the column and frit. I cleaned the flow cell with IPA, MeOH and water. I also cleaned with Nitric Acid. Finally after working with Agilent I ordered a new flow cell. Upon installation of the new flow cell in March of 2017 the baseline was flat and the absorbance units for the baseline was back to ~20 LU. This was maintained for from 3/2017 to 6/5/2017. Suddenly on 6/5/17 the absorbance units shot back up from ~20 on 6/2/2017 when it was last ran to ~42 LU on the first injection of 6/5/17. The flow cell may have become contaminated but we can't figure out how or with what. I have done the cleaning procedures. I have tried different scenarios to see if I can get the LU to continue to increase. I have used mobile phases from other locations (that have had no issues) and injected blanks (100 injections), I've injected samples (25 injections), controls (25 injections) and standards (50 injections). The LU started ~42 and ended ~42, so no increase. Any ideas other than a contaminated flow cell on what could cause the absorbance unit and baseline noise issues. I can order a new cell but I do not want the same thing to happen again.

[JMB]

Possibly very non-polar, late-eluding material is being washed off the column.
Does your HPLC method include a high % MeOH wash after each injection ? It should.
The period June 2-5 is a weekend; do you have a very low flow ON , say 20-50 uL/min during this time ? You should.
Have you used, say MeCN/CH2Cl2/DMSO to clean the flow cell ?
Can you describe your sample type in general terms---biological matrix etc ?

[blakeyk]

I have only done the MeOH/IPA wash and the nitric acid. I can try some other solvents. I'll need to order them. We are testing ground nuts and some tree nuts. They are extracted and filtered. Only solvents used are water, methanol and a 1% acetic acid solution. The odd thing about this is that it is identical to all of our other sites with no issues. No to the low flow or the MeOH wash.

[James_Ball]

Thinking through this, since the LU is increasing, you have to either have a contaminate in the system that fluoresces (absorbance would have less light hitting the detector and lower LU) or the light source is becoming more intense, leading to a higher background fluorescence.

Have you run the diagnostics to measure the intensity of the light source and compared to the original values?

FLD works by emitting a light at one wavelength from the light source and measuring the amount of a different wavelength of light at 90 degrees to the initial beam path to reduce the chance of scattered light being detected as signal. The third reason for increased LU could be debris or a crack in the cell that causes a reflection at 90 degrees to the source light beam, but should not be detected as a positive LU since it would need to shift wavelength to be detected, unless you have the wavelength range set too wide and the emission and excitation wavelengths overlap, which I believe the Agilent software prevents.

One good source of fluorescing material is algae or bacterial growth. Have you flushed the mobile phase lines from bottle to detector with nitric acid? 1-5% should be concentrated enough. Also a peroxide solution can help remove organic growth if it is really bad. Any time there is no flow in the system you should not have 100% water in a line. A good shutdown procedure for the system you describe would be to place both mobile phase lines in the methanol bottle and flush the system until the water is replaced down to the pump.

It can be difficult to find the source of the contamination, but give flushing the entire system, especially the aqueous mobile phase line a shot.