Peaks not showing up in analytical system

Hi community,

maybe someone has an idea what I could do, before a technician comes over.

During my absence my colleagues told me that our HPLC (1260 Vialsampler, 1260 Inf II Flex pump, TCC, FC-AS, DAD) does show a peak shift of about 3 to 10 minutes - when I came back after my leave apparently all peaks from mostly hydrophobic compounds just didn't show up at all. 

Flushing with pure ACN or MeOH doesn't seem to flush anything out. Also no ghost peaks were observed. So here is what I did so far:

Pump: Proper flushing, priming, conditioning over extended periods. Replaced the small frit in the outlet filter (PN: 5067-5716), replaced glass inlet solvent filters. Pressure seemed a bit fluctuating but was fine afterwards
Sampler: Replaced needle and seat, we had some issues with that. Replaced the switch valve seal (PN: 0101-1416)
No issues with TC-AS and the TCC
DAD: Replaced D2 lamp, it had over 2k hours and failed the intensity test, replaced the windows on the flow cells, the intensity ration was 0,03 (should be 0,3) and is now 0,49 after replacement of the quarz windows. However, even with the new lamp, the intensity count in the UV range is about 3000 counts but should be 5000 counts at least. All other tests pass.

What I tested:

Flow: Seemed maybe a bit too low but got better after priming and conditioning. I think it was about ~9mL (5mL/min, 2min) but I only had a cylinder at hand and no volumetric flask, might considering getting one and try that again.
Most of our methods are gradients so I suspected the MCGV and purged all channels with hot water for quite some time. No change, I think. Even if the mixing doesn't work and the gradient doesn't go into the organic direction, flushing with 100% organic should flush out something that is visible on the DAD. Also when you leave out the column and put in a long capillary, a peak is appearing.

I also observe that liquid is drawn from the channels when using the MCGV for online mixing.

I also injected a caffeine/theophyllin solution (4µL injection, c=0,5mg/mL for each substance, 270nm, AUC about 350 if I recall correctly, 25% MeOH in water premixed on channel D, no MCGV involved) and could observe peaks with a Rt of about 1,5 and 2,5 mins or so with the isocratic method and the premix. I also then injected the same with a online-MCGV mix of the same composition, the peaks eluted a bit later but were visible. 
Afterwards I ran my colleagues' method for his analyte which is Ciclesonide but could not observe any peak, also not with extended organic flushing.

We then took our column to our prep device (1260 binary pump 1-50mL/min, manual injector, MWD) and could observe the Ciclesonide peaks without any issues.

We also thought about that the substance is sticking in the column but we used 2 different ones to test this and as stated above, it works on the same column when we use the prep system.

What I ordered:

new MCGV, new seals for pump (although pressure tests all pass), new valves, new seal for the autosampler metering device, new outlet filter ... But these are a lot of variables and we will probably also get a maintenance by an Agilent guy. Still, my colleagues and I which have a lot of HPLC experience cannot think of more things that could cause this problem. It is just super weird. Maybe you have an idea of what I can do before I just replace everything.

Sorry for the long text but any help or advice is highly appreciated. Thanks a lot!

Parents Reply Children
No Data
Was this helpful?