Random loss of retention

[HW Mueller]

This low wavelength "absorbance" of Fig. 2 I have seen many times, but sometimes it leveled off above the baseline up to ~300 nm (I never go much above 300). Even though I have pointed out that one shouldn´t forget about absorption tails of solvents, I can´t imagine that one can claim this for air above 200 nm. But maybe this is also a wrong impression? I have seen similar things for He (discussed before), that can certainly not have an absorbance there.
Now I don´t remember the details, but I have seen "fluorescence" that coincides with the "absorbance". A huge false fluorescence (was certainly scattering) was observed once just before I saw a foamy plug leave the fluorescence spectrometer (I was doing some sort of degassing). Single visible air bubbles give spikes in both UV and fluorescence, or if formed in the detection cell one can get saw tooth patterns. Gas spikes can be seen without visible (naked eye) bubbles. So I wonder whether there can be a state of gas in liquids which is neither a real solution, nor a gas state (at HPLC conditions), but a state which causes scattering. I think it´s Rayleigh scattering that has a power^4 dependance on wavelenth, ... difficult to veryfy with HPLC equipment.
If I see this correctly, the light beam would have to be slightly off from parallel (to the cell path) in order to see refractive index variations ..... another possibility.

[danko]

If I see this correctly, the light beam would have to be slightly off from parallel (to the cell path) in order to see refractive index variations ..... another possibility.

I forgot to address that situation: If the air bubble is clearly visible i.e. larger than the light beam, the observed signal will be due to refraction of the beam, originating from the interface between different substances (could be air/fluid, glass/fluid, glass/air, different concentrations of a solute etc). So this is indeed a third case of “artificial absorbance" signal. Air bubble in the flow cell is an excellent example of that – though it often starts its “lifeâ€

[HW Mueller]

danko, thanks for the wavelength info.
The sawtooth pattern I mentioned was generally accompanied by a bubble emerging from the detector just after the absorbance dropped sharply. The builtup of the absorbance was relative slow and curved. The frequency was sometimes surprisingly constant, other times changing rapidly in one direction. The time for one peak ranged from minutes to maybe 10 seconds. The conclusion was that a bubble formed in the cell and increased until it was swept out, the process repeating immediatly. This was especially observed after air leaks in the plumbing (ferrules) on the pressure (~200 bar) side. Now we discussed this before, and if I remember correctly there were others that have seen the sawteeth in the mentioned connection, as well as Venturi air "intake".

[unmgvar]

i had a hunch something else might be the reason here. couldn't say what,
but anyway the post has turn into a HWMueller and danko discussion so i am staying clear for a while

[danko]

but anyway the post has turn into a HWMueller and danko discussion so i am staying clear for a while

Now, that would be a shame – and I mean it. If it was meant as a joke, then it’s OK, but if you hold your thoughts back in order not to interrupt others discussions then the beauty of the forum will be gone in a split second.
And now that you mention it, I’m really intrigued to hear your hunch. Would you please share it with us?
Actually, I was under the impression that the initial case was closed – Mattias solved the mystery. Or?
So, we kind of took it away and brought in some general thoughts. I don’t even think we were in disagreement

Anyway, air in the system and in this case the detector can be “pain in the flow cellâ€

[unmgvar]

Yes Danko it was a joke.

the reason i first asked Mattias to relook the behavior of his chromatogram is because a random change in RT can be caused by:
#check valves- air bubbles or dirt
#clooged cappilaries- sample loop, needle, injection port, from the port to the inlet of the column. generally bits of septa or dirty samples
#unheaven leakage (due to the buffer cristallising) in a connection- i saw it happen on a faulty purge valve.

[danko]

Hi Unmgvar,

Yes Danko it was a joke

#check valves- air bubbles or dirt
#clooged cappilaries- sample loop, needle, injection port, from the port to the inlet of the column. generally bits of septa or dirty samples
#unheaven leakage (due to the buffer cristallising) in a connection- i saw it happen on a faulty purge valve.

I agree that all these can introduce retention time irregularities.

Best Regards