HPLC System Cleaning

[ChromatographyChris]

Hi Everybody,

I've been having some massive problems with system peaks in my HPLC (A1100):

C18 column
A = 0.2% TFA v/v aq
B = MeOH
35-95% B over 30 minutes
10 minute equilibration
Column thermostatted at 30oC
UV detection at 270nm

The peaks are late eluting at around 20-30 minutes.

I've carried out extensive investigations looking at the water, TFA and MeOH quality. I've used different columns and I've done 0uL blank injections from empty vials with no septa to rule out vial artefacts. None of this has allowed the cause to be identified.

Moving to different a system gives a different profile of artefact peaks. This suggests that it's system contamination and that there may be similar, but not identical contamination on each system. Moving to a system in a different lab gives clean baseline.

I've flushed the system with neat MeOH, MeCN, IPA and with 1/1/1 Cy/MeOH/IPA. I've also flushed it with water and with 10% HCL and 10% NaOH. None of this made a significant difference.

I've run blank injections with the column switch valve and peltiers isolated and also with the injector isolated (ie: plumbing the pump outlet directly into the column) and the peaks were still seen.

I've swapped my aqueous and organic lines round and I've also cleaned my mobile phase reservoirs to ensure there's no organic contaminants in there.

I can only think of a few further things:

1) Replace solvent inlet glass frits/PTFE solvent lines
2) Contamination within the degasser (possibly microbial?)
3) Flush with an even stronger organic solvent (DCM)
4) Passivate/flush with HNO3 aq

The systems have been used with phosphate aq at pH 7 in the past so microbial growth is a real possibility.

Any thoughts?

I'd really appreciate any advice on this. Also, if anybody knows of a proven protocol for cleaning an A1100 up from either organic or microbial contamination or both I'd be very grateful too.

Thanks in advance!!

[Chromatographer2010]

6N nitric acid will clean up bacterial contamination.

Make absolutely sure to not have the nitric acid contact organic solvent as it will most likely explode. Also make sure that the nitric does't contact any PEEK if you have it in your system.

You will need to flush the system with lots of water before and after. You can monitor the pH of the waste to determine when the nitric is out of the system.


One other suggestion is to make sure that your TFA is from ampules and make up batches of mobile phase individually. Also take a look at your pressure traces to see if the peaks correlate to your pump strokes.

Also when you used the system in the other lab did you bring your potentially contaminated mobile phases and bottles with you?

You can also look at your pressure trace to see if the ghost peaks correlate to pump strokes.

[Uwe Neue]

Before going thru this giant washing exercise, do one more thing: get the mobile phases from the clean lab and run them on your system. If the peaks disappear, the source is somehwere in your lab, and not the HPLC system.

[LC_labrat]

I've recently had a similar issue with TFA using a C8 column. The TFA was not prepared from ampoules and the peaks were present, but after preparing new mobile with ampoules the peaks disappeared.

[Bluesman75]

The purge valve frit on the Agilents is a good source of contamination (I don't think you included this in your list of things to try). The pump manual will tell you how to replace this.
The quality of the water and organic solvents will also be critical.
Is the noise only seen when running this method? The flow cell windows and lenses in the detector optics can produce noise (the lenses become fogged), but this is often seen over the length of the entire chromatogram.
My other ideas have already been suggested.
Good luck in finding a solution

[orcicdejan]

Our Agilent representative suggested us to use the following mixture if serious system cleaning is required: 50 % isopropyl alcohol, 25 % acetonitrile, 15 % cyclohexane, 10 % methylene chloride. Just run it overnight at a low flow rate, and without column.