Thank you for your information, regarding to your first comment, yes, we verified the dispersion of the sample, and that's fine, but we can try another type of alcohol to improve this.
Regarding to your second comment, I forgot to report that we proceeded to shake the sample with hexane for 10 minutes at 200 rpm in a digital orbital shaker. The concentration of vitamin D3 in my product is 5000 IU / 100 g, and I weight 4 g (which represents a lot of volume), therefore, in my sample I only have 200 IU of vitamin D3.
The point of retro-extraction is interesting, i will try that. On the other hand, i already tried a RPLC method, the problem was the recovery of the sample and the precipitation of the sample into the vial, we try a lot of diluents (Methanol, Isopropanol, Acetonitrile, THF, etc.), and with every one of them we had the same problem. The risk of clogging the column or the HPLC was too high
I have to fulfill a recovery from the vitamin d3 between 75-125%. An the only way i obtain something near to that was with NPLC. I wish to try a solvent that retain the lanoline but not the vitamin D3, but I still can't find it.
Rndirk wrote:
In my opinion/experience, a saponification for this analysis is meaningful in a lipid matrix.
In step 2, did you check (visually) if the sample disperses in ethanolic KOH? We sometimes have issues with samples high in fat clumping together in this step. If this happens, we use an ethanolic KOH solution with lower water content.
In step 3-4, do you shake the samples after adding hexane?
If the concentration is high enough, I would repeat the procedure with less sample.
I can also recommend to perform a back-extraction after extracting your analyte to hexane: transfer (part of) the hexane phase to a new recipient, add fresh water/ethanol 50/50, shake for +-1min and settle. Continue with filtration and injection of hexane phase.
It is possible for vitamin D3 to perform RPLC if you have issues related to NPLC. Dry the hexane extract under nitrogen and dissolve in methanol. Elute on a conventional RPLC column with a gradient/isocratic program starting with high % organic.