Ultrasonic extraction of moist solids

[MichaelVW]

I'm looking for tips / suggestions.

Trying to use hex/ace to extract some moist samples by 3550C for GC/MS semis analysis. The amount of sodium sulfate necessary to dry 10g of sample to the point where I could call it a "free flowing powder" brings the total mass of material to be sonicated up to 50-100g. It seems to me that this is too much to sonicate; there's poor mixing of the material, and large sections don't seem to move at all. Spike recoveries are fine, but I'm not sure that accurately represents extraction efficiency. Using a wider bottom beaker seems to help a little, since the material is spread out over a larger surface.

If I decrease the amount of sodium sulfate mixed with the soil, but filter through enough sodium sulfate I can get most of the water out of the extract. But still, when evaporating it down, I see a water layer forming in the solvent. At that point, I'd guess polar solvents could have gone back into the water layer - so I can't just ignore it. Is it legitimate to add more sodium sulfate at this stage? Or would, say, phenol get absorbed along with the water into the sodium sulfate? Am I overthinking this?

I appreciate the feedback as always.

[ajc92]

Hi Michael,
Have you tried using Hydromatrix as a drying agent during the sonication process? It creates a much less 'blocky' material, and it removes water really well. I always preferred to use it vs sodium sulfate when I did 3550 for this reason. I would still use sodium sulfate for the filtration step - are you using filter paper or glass wool to hold it when you filter?
I cant speak to recoveries putting the salt in at the very end, I've sprinkled in sodium sulfate to a concentrated extract on occasion for PCBs/pesticides, but for full list semis my experience is extracting with DCM so I'd usually scoop any water off the top with filter paper.

A possibly less strenuous idea could be to extract a smaller amount of sample and calibrate lower to achieve the same RLs.

Hope this helps!