separation of vitamin E and its metabolites

[JL919]

I need help! Any suggestion is welcome, please!
I am working on the separation of four unknown compounds from EtOAc extract of human serum sample, which probably are tocopherol acetate and its metabolites.
Detector: AppliedBio Q-TOF, negative APCI mode
Column: Hypersil ODS C18 4.6*125, 5 um
Mobile phase A: 5% Methanol, 95% water with 0.1% formic acid; B: Methanol
Flow rate 1ml/min, injection volume: 10 ul; TEM: 35ºC
Gradient: 0 min 20% B, 18 min 100% B, hold to 21 min

The peak shape of these compounds is really bad, broad and tailing, and they always coming out together at about 16 min. I tried different gradient, temperature, acetonitrile as solvent B, but no improvement on the separation. I tried zorbax eclipse XDB C18 column too, same thing. What should I do?

[Uwe Neue]

Are you injecting the sample out of EtOAc? Then this is the issue. You need to inject from a more polar solvent composition. I usually recommend to dissolve the sample in the starting composition of the gradient. In your case, this may cause some solubility problems, and you may need to use an intermediate composition.

You can find an example of such a problem on page 360 in the troubleshooting section of my book on "HPLC Columns".

[Mark Tracy]

Are the samples dissolved in ethyl acetate? That would explain the poor peak shape. I've used EtOAc for a similar purpose, and you do pay a price in peak shape. (It was worth it to avoid the losses and effort of evaporation/reconstitution.) You might try starting the gradient at 90% methanol to reduce the contrast between the sample and mobile phase.

[HW Mueller]

Do you get a reasonable chromatogram of tocopherol acetate under your conditions?

[JL919]

Thank you all for the reply.
But no, my sample is not in EtOAc. I evaporate it to dryness and reconstitute with MeOH : IPA (9 :1).

[HW Mueller]

Sorry, but that´s too close to EtOAc as far as "washing in" is concerned (with a 95% aqu. starting mobile phase). Expect your standard in that medium to give equally disagreeable peaks.

[JL919]

The compounds I am trying to separate are the minor component in serum and they are quite hydrophobic. If the starting solvent is used, I won’t get them detected. My internal standards looked pretty good with the current method I am using (start with 20% B).

I searched this forum with “tocopherolâ€

[Bryan Evans]

If you get stuck on the C8 phase, you could always give our Unison UK Silica column a try (under normal phase conditions):

http://www.imtakt.com/TecInfo/TI166E.pdf

[Uwe Neue]

Methanol IPA is still a strong solvent. See if you can add some water without precipitating the sample.

The use of a C8 column is counterproductive.

Your gradient starting conditions are also part of the problem. The analytes won't elute until late organic.

Start your gradient at 50% organic (or higher), and dilute you sample with water (preferentially 1:1, if solubility permits). The peak shape will be fine, unless your column is dead.

[HW Mueller]

If the standard works you just need to clean up the samples so that they closely resemble the standard.