Chromatography Forum

Just starting to learn to use MassLynx with Waters Acquity UPLC/MS/MS to develop methods for pesticide screening and quantification - already seeing significant day-to-day signal losses on a new instrument -any hints for troubleshooting this?

Make sure you have a reliable standard to check sensitivity day to day (so you know it's the instrument, not the standard, that is declining)
If you're paying Waters service contract, you will find they will be quite helpful about advice on maintenance - don't feel guilty about using them.
Keep the system clean: it's much easier to clean a dirty cone than to clean an ion-block, and if the dirt gets beyond the ion-block, you really really don't want to start cleaning stepwaves. Dirt washes further and further into an instrument with time, so catch it before it gets too far.
Make sure your drying gas flows are reliable, and that you don't find yourself using some strange tune-file with the wrong gas-flows. In fact I strongly recommend naming your tune-file in the sequence table.
Occasionally check that the spray looks like a healthy cone-shape coming out of the nozzle and not some horrible misdirected spray because your ESI needle has got blocked or damaged.

Thanks for the advice. I first thought I needed to do some maintenance on the interface other than just cleaning cones - they were fine. The problem was a slow vacuum leak. The Masslynx software doesn't have a warning or interlock for inadequate vacuum, so I was oblivious. After a week and a half of work on the instrument, the service engineer found the leak in one of the o-rings in the detector. The detector was completely replaced and the leak hasn't recurred.

He also recommended caffeine as a cheap sensitivity check instead of the waters check standards, which are expensive to replace. I also now log the counts for my most common analyte standard as a sensitivity performance record.

I also learned that you have to be careful when using Intellistart in infusion mode - it changes the desolvation gas settings, which then have to be changed back manually or by re-loading the original tune file - this made me think my sensitivity had dropped when I ran the next sample run with reduced desolvation gas and interface temperature.