For a polar compound you'll obviously want a more polar stationary phase, of which there are many available.  The actual stationary phase will depend on the level of retention for the compound, and the selectivity you require, both of which will require additional experiments to verify.  Something like an HP-50 (50% phenyl) is a good balanced phase, where a WAX phase (as you mention) is suited more to the polar side.

A similar statement can be made for the column dimensions.  The phase ratio of stationary phase to internal diameter is selected based on the retention characteristics (higher phase to diameter ratio to increase retention) and the magnitude is typically dependent on how much you are loading onto the column.  If you are looking in the low ppm range than a 30 m x 0.25 mm x 0.25 µm column might be a good place to start.

Using a more polar solvent like methanol may be problematic for a polar, potentially early eluting, compound like the one in question.  You can end up with a solvent effect which causes poor peak shape.  However, you may be able to work around that by optimizing the method parameters.

Thank you fo your answer! It's ppm-subppm range, I'd also start from the same dimensions. Since the samples would be seedlings, we can expect a tricky matrix, since they contain high levels of oils, and sugars. I think we will test multiple solvents for the compound includng methanol and ethyl-acetate, also I'm thinking about trying headspace, SPME fiber technique for injection. Any thoughts on this?

It may be difficult to get a good response for this compound in headspace if the matrix is somewhat polar (i.e. is aqueous to some extent) because it would have a high affinity for the matrix and not partition well into the headspace.  However, it would be worth a try  since headspace is the easiest prep technique.  Otherwise, as you say, you might need something else like SPME.  Regular SPE may also work well for getting rid of the sugars, but the oils may be a problem as would the recovery of this compound if it is as polar as you say it is (assuming a reverse phase mechanism).  I typically use ethyl acetate as my GC solvent and find it works well for most compounds so hopefully it will work well for your needs.

What is the purpose of this analysis? This will guide you and us to some specific solutions :)

Seeds will be treated with this compound by biologists, and they would like me to measure this for them from seedlings, or the leaves of very young plants, just yesterday they told me they obtained the standard reference material for 3,4,5-trimethylfuran-2(5H)-one measurement method and it's C isotope labelled counterpart, I'd have to measure the original compound, and the isotope labeled one too, the labelled material has +2 mass shift.

OK. So you will make an extraction of this compounds from plan material?

if you can select any column - take DB-624 it gives very good selectivity for polar and moderate polarity compounds.

The main problem will be to obtain accurate determination - because probably you will not obtain a 100% recovery during 1-stage extraction.

Somehow a standard addition method (obviously use the SIM mode) can help you with this issue.

If it is highly soluble in water you'll probably need a polar solvent. Acetone is the best for GC. Methanol is not a great GC solvent but for this compound it should work and methanol has some MS advantages (it has very low mass).

Use a 5% phenyl phase (like a Restek RXi-5-Sil MS). The furan will chromatograph nicely on it. No need to move to a highly polar phase since your compound is not highly polar, and the 5-MS phases are very stable. You should have lots of sensitivity using GC-MS, so you'll be able to use a split injection and get really nice chromatography.

Thank you very much! Maybe next month I can start this project on the GC-MS, and I will report back here the finalised method. In the meantime I would welcome any suggestions in sample prep. I won't have too much plant material avaible, maybe 250mg/sample tops.