Baseline Drift

[natemullo]

Hi there,

Im having issues with my baseline when running aqueous samples. My aqueous media contains a series of organics (for microbial degradation) as well as dissolved minerals. Prior to injection the solution is centrifuged then passed through a 0.2 micron pore size syringe filter. My baseline starts off at zero and fluctuations lightly between 0 and 30 microvolts. However, upon temperature ramp my baseline increases to over 300. Additionally, I am seeing peaks of my specific analytes (previously determined elution time) shift and they are also being distorted.
My specs are:
Varian 450 GC-FID
Helium carrier
flow rate - 2.5 ml/min
injector temp 300
Oven ramp - 35C (start) hold for 2 min (at two min mark is when drift occurs)
ramp @ 40C/min to 175C, hold for 1 min
FID temp = 300C
column = VF-5ms

Any help would be appreciated. I've posted an example photo (i took from online - not mine) which shows the trend, however there are sometimes discrete peaks when running an actual sample and when running pure reverse osmosis water, there are not.

[Peter Apps]

Welcome to the forum.

Injecting water is a thing to avoid if you can. It degrades the column and the inlet deactivation and causes pressure pulses that push contaminants into all sorts of unwelcome places. Depending what your target analytes are you will be better off doing a solvent extraction.

What is you inlet maintenance schedule, and how much work has the column done ?

What makes you so certain that the drift is not column bleed ?

Peter

[Rndirk]

Can you give the liner type, the injection volume and method (split, splitless?) and the column dimensions?

You will overload your inlet very quickly injecting water at 300°C.

Moreover, starting at 35°C, your water will condense on the column and might spread in droplets all over it.

Oven ramp - 35C (start) hold for 2 min (at two min mark is when drift occurs)
ramp @ 40C/min to 175C, hold for 1 min

How can this ever be a run of over 1 hour as is shown in the chromatogram?

Like Peter said I would avoid injecting water.

[natemullo]

Hi, thanks for the response.

I know injecting water is ultimately frowned upon and it can cause a whole host of issues including backflash. Unfortunately, the nature of my work involves a lot of water analysis and ultimately we try to reduce pre-processing as much as possible.

As for the column, last change of septum, o-ring and liner was in November, however i just changed it 2 days ago.
Column has been in use for probably 4 years (rough estimation, came into the lab and not much was passed down to me)
I am not certain its not column bleed, that is a good point. I just haven't seen this phenomena before running similar samples (I'm relatively new to GC).

Column is 5% phenylmethylpolysiloxane
I'm unsure of the liner type, I know it has glass wool of course but I just used the similar liners available that we had in stock.
I use a splitless injection.

How will the water overload the inlet at 300C? I'm not sure I know what you mean.


Thanks

[Rndirk]

I'm unsure of the liner type, I know it has glass wool of course but I just used the similar liners available that we had in stock.
I use a splitless injection.

How will the water overload the inlet at 300C? I'm not sure I know what you mean.

Assuming you inject 1µL of water in splitless at 300°C, it leads to > 1mL water vapor = backflash unless your liner/inlet can hold this volume.

The other effect: water (polar) will not spread nicely on your (apolar) column phase and can degrade it.

I think it might be possible to get it working if you avoid backflash and use the right column. I haven't done it but i guess there are columns, and maybe even liners tailored for water as a solvent.

[Bigbear]

I've injected water, but you are limited to 0.5 ul into an injection port at about 120C. This is assuming a 4mm liner with an internal volume of 800 ul. If you have the option of a pressure pulsed injection, this may help.

[Peter Apps]

Four years of splitless water injections will kill any column - you are seeing baseline drift due to bleed, and ghost peaks.

The peak distortion is due to solvent effects from the water condensing on the column - if you have to do 1 ul splitless water injections you need to keep the column above 100C to stop condensation. Keeping the water as vapour will slow down column deterioration a bit also.

You could try a retention gap with a polar deactivation, but if your water samples have non-volatiles in them you will need to change the gap very frequently.

Peter