Quattro premier possible charging issues!

[MHANLEY]

I think I’m having problems with charging (at least I think it’s charging) on my LCMS system (Acquity UPLC, Quattro premier, Waters HSS T3 (C18) Column, H2O/MeCN 0.1 % F.A.) but am baffled as to how to solve or even further diagnose the issue .
A few months ago I started observing my internal standard reducing in intensity over the first 16-24 samples by around 90 %, then after perhaps 50 + samples the signal would start to come back. (retention times remain very steady)
After a weekend of being left in standby the signal comes back to full strength and takes about 25 samples before the sensitivity starts to drop with a rate of decline is lower than if the MS gets used on consecutive days.
I’ve cleaned the source and ion tunnel numerous times (Sonicated in 50 % MeOH 20 min, followed by 100 % MeOH for 20, dried under N2) I’m careful with what I inject ensuring the divert valve only switches from waste when the analytes elute. I’ve been running the same types of sample for a long while now without seeing any change in sensitivity (apart from the usual cone contamination over time) and have not otherwise seen a gradual decline of sensitivity over a longer period.
I have a daily standard consisting of 4 compounds of varying polarity/lipophilicity (caffeine, Nalidixic acid, Glafenine, Fenofibrate) and find the same issue when running these so it’s not down to a specific compound or property.
Whilst the water is refreshed fairly often the organic phase comes from a 2.5 L bottle of MeCN which has been there for quite a while. The air inlet is filtered so there shouldn’t be anything contaminating the solvent but is there anything that could have been absorbed from the air that would cause this?? Clutching at straws I know but I’m not sure where else to look.
I’ve run a Gas collision cleanliness test with reserpine monitoring at 609.8>195.5, 1000>195.5 and 2000>195.5 with 20+ direct injections and found absolutely no cross talk between channels.
If there is contamination on the MS 1 quad (looking very likely) can this charge gradually then dissipate that charge whilst still remaining adherent? Do lipids behave like this?
Any insight or opinions would be gratefully received!

Thank you

Marcus

[Rndirk]

1) Do you measure in positive or negative ionization? If both, do you see the same effects?

2) I've had problems with acetonitrile bottles sitting on the instrument (above room temperature, sunlight,..) before, giving weird results. I think you should keep the stock in a dark, dry and cool place and replace the ACN on the instrument every couple of days. Moreover, formic acid concentration can go down as a function of time, it's relatively volatile.

[James_Ball]

When was the last time the roughing pump oil was changed?

I have seen problems where moisture gets into vacuum pump oil and causes loss of vacuum, then after sitting over a weekend it drys out and returns to normal.

[MHANLEY]

@rndirk
I only use +ve ion mode. Mid way through a test run with repeat injections of my daily standard I've switched to -ve ion mode for a few minutes then resumed the run but have not observed a significant change in sensitivity.
I think also I'll reduce the volume of my MeCN reservoir significantly and replace more the solvent more often.

@James
I ballast the oil pump once a week to try and avoid those issues with moisture in the oil, but yes, an oil change wouldn't hurt

[Rndirk]

@rndirk
I only use +ve ion mode. Mid way through a test run with repeat injections of my daily standard I've switched to -ve ion mode for a few minutes then resumed the run but have not observed a significant change in sensitivity.

I was going to suggest this. Switching to negative after compounds have eluted and flow directed to waste, during reconditioning, is something we do to prevent charging. Even if it doesn't help, it doesn't hurt.

Can you recall any incident around the time you had these problems appearing? Maybe it's time to have a thorough maintenance performed by Waters.

You could run tests where you compare the first and last quad performance in full scan mode, to check if there's a significant difference.

[mhr311]

A/D board has a DC polarity switch for each set of rods. I believe its SW8, top=MS1, bottom=MS2 (it should be labelled). That will change the actual rod DC polarity on each rod set and may desorb potential contaminants. Note that the filters are there to clean rather then the entire quad assembly. YMMV