GC-FID active sites Jet?

[MSCHemist]

I took my own advice and scrubbed the inlet again with a saturated citric acid solution to passivate the metal. I took out the seal and reducing nut, kept the inlet at 80 and put citric acid soln down and scrubbed with a brush. I then rinsed the heck out of it with water, then methanol, put it back together and it made no bleeping difference. I also put the column back a bit higher in the inlet.

The inlet looked a bit shinier but the GC chromatograms did not change one bit. I am still waiting for the jet.

[Yama001]

It may help to rinse the inlet / purge lines and rinse or replace the split line.

[MSCHemist]

The split line is band new. I nearly lost the valves a few months back on my 5890and put a new long length 1/8" copper to shield them in the future so stuff has more room to condense. THough I'd assume any analytes going down the split line are out of the picture.

[Peter Apps]

The split line is band new. I nearly lost the valves a few months back on my 5890and put a new long length 1/8" copper to shield them in the future so stuff has more room to condense. THough I'd assume any analytes going down the split line are out of the picture.

Not so sure about that - there have been several threads about poor peak shapes that were improved by cleaning split lines.

If the extra 1/8 inch pipe has a volume greater than the inlet itself, then the pressure pulse form vaporizing the solvent will push a pulse of sample into the line and at 10 ml/min in 1/8 inch pipe you have a linear velocity of only 5 cm/s which would allow back diffusion even without any bulk backflow.

Increasing the split ratio and sample concentration will test for this.

Peter

[MSCHemist]

I tried injecting with a 10:1 split followed by a 100:1 split. The 100:1 was of course 10X smaller but had the same shape as the 10:1 awfully tailed.

I tried scrubbing the fid parts with citric acid then washing with water and meOH. The steel parts got shinnier but it also made no noticeable difference.

The new FID jet came with a bent stem so it will be another few days until I can try that.

[Peter Apps]

It never rains but it pours.

While you wait for the new jet, try splitless injections with a 1 min splitless time and a 100:1 split once the vent opens.

Peter

[Peter Apps]

Just occurred to me; make sure that you do not have a part of the column touching the oven wall, or shielded from the air circulation in the oven.

Peter

[Yama001]

If the extra 1/8 inch pipe has a volume greater than the inlet itself, then the pressure pulse form vaporizing the solvent will push a pulse of sample into the line and at 10 ml/min in 1/8 inch pipe you have a linear velocity of only 5 cm/s which would allow back diffusion even without any bulk backflow.

Increasing the split ratio and sample concentration will test for this.

Peter[/quote]

Thanks for that, Peter. I have long known this could happen, but never understood how.

[MSCHemist]

FID jet made no difference. I can only guess now it is the column. I consulted Phenomenex what liners they use for the amino acid kit they sell which is identical to my method and am buying a set so I will try that.

The only thing else I am leary of is the FID capillary adapter and the graphite vespel at the bottom where the column seals and the big 1/4" one that seals it to the bottom of the detector. I've always wonder why they don't just use a 1/4" SS swagelok fitting.

I don't think the 1/8" Cu line makes any difference as that is what the 6890 uses for the split line.

[Yama001]

A metal ferrule at the detector adaptor is fine, it is just that if it developes problems from over tightening you have to replace the whole adapter.

I am not sure if you tried rinsing or replacing the inlet line assembly (I am never sure what this is called - its the part that has the large octaganol nut - one line is the gas flow to the inlet, the other is the septum purge). When I am out of ideas I give this a try, it has helped more often than I would think it should . This is less of an issue on 5890s than on the later models - 6890 onwards seem to have significantly more problems with upstream plumbing getting fouled, particularly in splitless mode.

[aldehyde]

What is your baseline signal on the FID when you have the oven at 50 C and what is it when it is up at your max oven temp?

Do you have any other samples you can inject? I would be interested to know how things look if you inject something similar to the Agilent FID sample (hexadecane in isooctane.) Something more inert would give us an idea of whether this is a reactivity issue, a leak, or something else.

[MSCHemist]

At 50 deg my signal is 5-6pA and at 280 it goes up to 12 pA. The peaks are very symmetrical except for those two (serine threonine) and some other hydroxyl compounds and I have issues with the hydroxy compounds literally disappearing at low levels and quadratic calibrations.

I'm at a loss I tried a brand new liner, new jet, I polished and passivated the heck out of the inlet with citric acid, I'm using a deactivated base seal, the column is relatively new, I cleaned and passivated the rest of the FID parts. It now goes from badly tailed to unintegratable. New septa. The split line is very new. The carrier and septa purge are a long shot I doubt there is much there. Blanks look fine.

[Peter Apps]

A brand new column sounds like the only other option; "relatively new" for a column can mean that it has had the one injection of death that generates active sites.

Peter

[MSCHemist]

I'm trying a solvent rinse using my solvent pump on the column. I'm trying MeCl2 followed by methanol followed by UFDI water followed by methanol again.

[Peter Apps]

I'm trying a solvent rinse using my solvent pump on the column. I'm trying MeCl2 followed by methanol followed by UFDI water followed by methanol again.

If the adsorptive sites are on soluble small molecules that might help, but if the silicone in the stationary phase is oxidised or hydrolysed the only column treatment option is a gas phase deactivation - which I have never known to do any good at all.

Peter