@aretaon My question would be: when was the rotor seal exchanged and was this exchange done by a "professional"? Besides the dirt, is there any scratch on the surface? (Difficult to see on that shiny surface, I know...)

@kabalak Seals are regularly exchanged by us and we have been doing this for years (last exchange was 5-6 months ago). There are abrasions on the rotor seal (picture of the corresponding seal below) which we assume to result from the dirt accumulating and mechanically damaging the seal. We have seen this now multiple times on two LCs, so I assume this is rather a global problem than a faultily installed seal.

@aretaon Ok, and the solvents are the same, same LOT all the time? And what about your samples running on the LCs, proteomics samples (plasma, urine or other weird things?) - C18 cleaned and/or what is the installed sample loop. Injector things look still fine?
Very weird things can happen in that autosampler; especially if you use µL-pickup combined with the big vials to get the µL-pickup solutions into the system. Deteriorated metal caps' contamination can be seen - but there must be a leak.

@kabalak Yes solvents are the same, we didn’t record the LOTs so I cannot comment on that but the same manufacturer and product for sure. No strange samples only proteins.
Sample loop is a normal 20uL nanoViper loop. We use 100uL glass inserts in vials with PTFE seals in the autosampler but we didn’t check the needle yet.

@Ramses_X @kabalak Troubleshooting is still work in progress but our current suspect is the ultra pure water system. We changed to bottled water and have now a dedicated HPLC just running mock samples on the type of column that made most problems. We regularly check the valves for debris and hope for the best. Funnily, the conductivity readbacks for the water system are great, we have 0.055 uS/cm which is pretty much the physical limit at 25C. So what ever is potentially in the water would have to non-contributing to conductivity. Silica for example has this property but we will have to test for that to be sure. [1/2]

@Ramses_X @kabalak Peptides as such are not the issue we have been measuring proteomics samples with the identical setup for years without ever having precipitates in the valves. As you said it’s likely something in the solvents that we don’t know about.
We have been collecting the debris but it’s minute amounts and I don’t know yet how to analyse it. We’re even not sure yet how to dissolve it for analysis because it’s such a low amount (that is stuck to a cotton swab). [2/2]