Chromatography Forum

Hi all,

I am currently working on developing an analysis method for Vitamin D2 and D3 in food. I am following the extraction procedure in AOAC method 2012.11, where I performed saponification of the sample (12 g sample + 6 mL Methanol and 4 mL 45% KOH and heat in 75C water bath for 30 mins) then extraction with 20:80 ether: petroleum ether mix, evaporate to dryness then reconstitute with 2 mL methanol. But I am not getting a good recovery. Any suggestion or advice?

HPLC condition,
Column: acclaim C18
Mobile phase: water/methanol (gradient, start with 20/80 water/methanol ramp up to 100% methanol)
UV-Vis detection at 265 nm
column temperature: 30 C

A common problem I can think of with regards to recoveries crops up when the calibration curve is based on standards that have been prepared in a "clean" solvent rather than spiked into a blank sample matrix. The lack of compensation for matrix effects can cause problems with regards to recovery.

If you are not already doing so, preparing calibration standards that are spiked into blank sample matrices might be worth a shot before moving onto other things.

Thank you for your reply! I have been preparing my calibration standards in methanol. I will try preparing them in a blank sample matrix. So, do I extract the blank sample (saponification with MeOH&KOH then extraction with Hexane) and then add in different amount of standard to prepare the calibration standards?

Typically you would add the known concentrations of standard into the blank sample matrix (these concentrations will be the ones that you want to plot on your calibration curve) then extract the standard back out of the matrix using your extraction technique. You would then use the resulting solutions to construct the curve.

The matrix itself really needs to be an untreated version of the food you are testing. If you can't obtain a genuinely blank version of your sample i.e. the vitamins of interest are always going to be present in the food, you might want to consider a standard additions method of calibration. Hope this is of use