Chromatography Forum

Hi I could use some advice. I have a 5890-FID that I've been transfering methods to because my GC/MS is always tied up. I am having issues with disappearing hydroxy compounds at low levels. I have a derivatized amino acid method where serine and threonine dissappear, an acid method where succinic and fumaric work fine but citric, malic, and tartaric start to dissappear more quickly than they should at the bottom of the calibration range (~10ppm).

I've been through the inlet over and over you can see your reflection in it, it is so clean. I replaced the gold seal with a deactivated one, replaced the autosampler syringe, the liner is very well deactivated, the column is brand new and has very nice peak shape.

At this point I am wondering if the issue is at the FID jet? I've checked the column position and rehooked it up. I am wonder if I should replace the jet with a deactivated one. It is clean, but very very old. Does the FID jet usually make a difference?

How low is low ? ng per peak ?, and what column stationary phase are you using ?

In principle a jet could cause peak tailing, but it is very hot compared to the rest of the system, and very short compared to both the inlet and the column. I'm presuming that you have the column threaded up to just below the orifice.

Peter

1701 column 30m .25mm .25um

solvent injections 10:1 split 10ppm. I take the vial to the GC/MS and inject and it is fine on a 1701 column I have there which is far older and then inject it on the 5890 and it is almost entirely gone. I use the same type of liners, gold seals, syringes on both instruments so that is why I am leaning towards something with the FID.

I've definitely checked the position of the column in the FID jet and everything. I am wonder if compounds are just adsorbing to the metal surface.

I keep the FID at 280 because that is the max T for a 1701 column and the column runs through the FID.

Do the failing compounds give a quadratic fit curve across the rest of the calibration range? Adsorption would normally cause more bias at low concentrations and less at higher concentrations since it will normally work as a total absorption not just on a percentage scale, so that the percent adsorbed is greater with lower concentration standards.

Also, have you checked the attenuation to make sure you are able to see the response at the lower concentrations, it could be the sensitivity just isn't there or it is being attenuated out.

1701 column 30m .25mm .25um

solvent injections 10:1 split 10ppm. I take the vial to the GC/MS and inject and it is fine on a 1701 column I have there which is far older and then inject it on the 5890 and it is almost entirely gone. I use the same type of liners, gold seals, syringes on both instruments so that is why I am leaning towards something with the FID.

I've definitely checked the position of the column in the FID jet and everything. I am wonder if compounds are just adsorbing to the metal surface.

I keep the FID at 280 because that is the max T for a 1701 column and the column runs through the FID.

Certainly puzzling. With 1 ng on the column you should see a decent peak. How is the peak shape - tailed or symmetrical ?

Peter

I baked the column at 290 for an hour and managed to get a blob of serine and threonine back. They are tailed and overly wide definitely adsorbing somewhere and I've been over the inlet and liner many times. That could be an indication of a bad FID jet. I've read somewhere baking the column out puts column material onto the jet which masks active sites.

I am also seeing linearity issues with the active compounds all compounds with exposed -OH groups (or in the case of glutamine an amide).

I've gone ahead and ordered a silicosteel FID jet. The only other thing I could do at this point is buy a silicosteel inlet weldment but that should not be necessary. My GC/MS doesn't have one and the compounds work fine there.

I would be interested to hear if the Silcosteel jet makes a difference.

What column temperature are the derivatives eluting at ?

Peter

Hmm - a 3 pipe problem

Unless I have misread the posts,the only thing that you haven't tried/eliminated is moving your column from the GCMS (which worked) to the GCFID instrument

Regards

Ralph

It is rather difficult to swap column between the FID and GC/MS. The GC/MS has a quickswap and uses metal ferrules. The FID uses standard graphite/vespel at the detector. However the column is pretty new and is giving good peak shape even of polar compounds.

I've been thinking of putting some deactived tubing in the FID jet and have a union which uses the same ferrules as the GC/MS so I can swap out columns between the two instruments.

The amino acid derivatives elute in the range of 160-280. The hydroxy acids also 150-250 deg C. I esterfy the acids with isobutanol but citric, malic, lactic, and tartaric still have exposed hydroxyl groups which is why they are prone to activity/adsorption.

I would sooner suspect the inlet weldment before the FID. Agilent recommends " gun brushing" the inlet. I have done it to get endrin break down under control.
I don't "gun brush" any more, instead I replace the weldment with a clean one, and then clean the dirty one with the alumina powder used for MS cleaning.

I would sooner suspect the inlet weldment before the FID. Agilent recommends " gun brushing" the inlet. I have done it to get endrin break down under control.
I don't "gun brush" any more, instead I replace the weldment with a clean one, and then clean the dirty one with the alumina powder used for MS cleaning.

As above, a .32 caliber bronze bore cleaning brush on a gun cleaning rod works great for a quick clean on an Agilent split/splitless injection port. But we also do the swap on our semivolatiles instruments and clean then with chrome polish and a Dremel tool. The inside looks like a mirror when it is finished, but the only thing I don't like about that method is how long it takes to get all of the hydrocarbon base from the polish out of the weldment. I am thinking about having them switch to alumina powder/citranox mixture like we use on the MS source because it will wash out much easier.

Though believe me I've very thoroughly cleaned the inlet over and over. I've scrubbed it with water and multiple solvents ending with methanol and put in a fresh liner and deactivated base seal. It looks very very clean. I'm not sure what else I can do there but replace the inlet weldment with silicosteel which none of my instruments have.

Would ripping it out and soaking it in 20% nitric or citric acid help passivate it. That is the only thing I haven't tried yet

What style of inlet liner are you using, and does it have any glass wool in it ? In Agilent inlets the very short distance that the column protrudes into the liner makes me worry that sample is bouncing back and forward off the gold seal - and we know from several other posts that crud in the split vent line can affect peak shape.

Two diagnostic things you could try; move the column further into the inlet - say another 10 mm up, and increase the sample concentration and the split ratio by a factor of five or so - the faster gas flow through the constriction at the bottom of the liner should reduce back diffusion off the seal.

Peter

For split I use a deactivated (Restek Blue) glass liner with wool with a taper on the bottom to minimize gold seal exposure.