by
ljc » Mon Sep 19, 2005 1:35 am
The HPLC method for Vit E including tocos and T3s is well documented, check the AOCS. There are minor variations in the literature, but it is a normal phase method using either a DIOL or silica column and FLD at 290 nm excitation/330 nm emission. We do this analysis routinely using the following conditions:
Supelco DIOL 250x4.6 mm 5 micron at 35 deg C (Supelco calls it a "diol-like" column, and I've found it to be very robust under these conditions, yet different than other DIOL columns that are out there)
Eluent: 0.5% isopropyl alcohol in heptane, isocratic at 1 mL/min
Inj volume: 20 uL (you can adjust of course, but you have to be careful not to go too big unless you're using straight heptane as the sample solvent).
Detector: FLD excitation: 295 nm emission 330 nm
The vitamers elute in less than 20 mins in the order alpha, alphaT3, gamma, gammaT3, delta, delta T3. The betas elute between the alphas and gammas I think, but are very low in concentration, at least in rice-derived oils. The glyderides elute before the alpha toco and don't interfere anyway using FLD (though it's nice to have pure peaks anyway)
Calibration: Here's the catch: the "standards" for tocos or T3s are hard to get (Merck, Calbiotech, etc) and are extremely expensive, and are in such small ampules (50 mg or so) that it's hard to make solutions unless you use the AOCS method. This calls for making a solution with a candidate standard in methanol and measuring its UV absorbance, then back-calculating the "true" concentration from published molar absorbtivities (which were obtained in MeOH). You then add an aliquot of the meoh solution, evap to dryness, and take the residue up in your normal phase solvent. Theoretically you know the mass in the MeOH aliquot from the calculated concentration, the final volume of the heptane you dissolved the residue in, and therefore its concentraton. This is tedious especially since std solutions have a limited lifetime. You can do this with inexpensive alpha and delta from Aldrich, just run a separation on them to be sure they're at least pure in the vitamer of interest. Finally, the accepted practice is to set the response factors of each T3 to be the same as their corresponding tocos. This makes sense since the chroman ring is the part that either absorbs or fluoresces.
Your sample prep for edible oils is trivial: just dissolve in heptane
The remaining details about dilution factor, how many levels to calibrate, etc have to be determined by you based on the types of samples you have.
There are also reports of reverse phase methods but with oils, the normal phase seems better since you have the perfect sample diluent already as the mobile phase, and the vitamers are separated isocratically very quickly.
