problem with peak interpretation

[zanio]

Hi all
In my work I investigate the content of flavonoids and phenolic acids in tomatoes by HPLC. The HPLC sistem is Agilent 1200, column Poroshel 120 ( 70x 4.6x 2,7 microns), UV-DAD detector, mobile phase A: 0,5% acetic acid in water and phase B : methanol. - gradient elution. In these case the standart compounds have separeted very well. But when I recieved the sample chromatogram I had a problem with interpretation of peaks. For example, at 18.6 minutes would have to appear quercetin. In the chromatogram of the sample there is a peak at the same time but it is not quercetin. The spectra of this peak is different. What I have to do? In these case I cannot present this peak as quercetin and the data will be wrong.

Thank you!

[tom jupille]

What I have to do?

1. Identify the peak (LC/MS? collect and run MS?)
2. Obtain a standard of that compound
3. Adjust the conditions to separate that compound from quercetin (and anything else you are interested in) while not hurting the separation(s) of your other peaks.

Welcome to the world of natural products chromatography!

[James_Ball]

We had a similar problem running Formaldehyde in tobacco products. We noticed that in actual samples the peaks were not completely symmetrical. We changed from a Luna C18 column to a Kinetics C18 column with smaller particle size and smaller ID and we were able to separate two compounds at that retention time.

You may also want to change phase, say from C18 to CN phase which can change the order of elution for many compounds which may allow you to separate the interfering peak from quercetin. We analyze extracts from onions for quercetin and now use the Kinetics C18 to give better peak separation.

For the Quercetin analysis I started with the method listed in this article: “HPLC-UV and GC-MS Characterization of the Flavonol Aglycons Quercetin, Kaempferol, and Myricetin in Tomato Pastes and Other Tomato-Based Products” O.Tokusoglu, M. K. Unal, and Z, Yildrium ACTA Chromatographica, No. 13, 2003

Then developed better separation using a gradient analysis starting with about 90% aqueous using a mobile phase of 0.2% Phosphoric acid as A and Acetonitrile as B. You may need to work with different mobile phases and gradients and longer or narrow bore columns to achieve the separation needed for your particular samples as one method does not fit all when it comes to analyzing natural products such as this.

[Gerhard Kratz]

Please google for this:
Quantitative Analysis of Quercetin in Natural Sources by
RP-HPLC
Ch. R. S. Phani*, Ch. Vinaykumar, K. Umamaheswara rao and G. Sindhuja
R. V. Labs, Guntur, India
__________________________________________________________________________________________
Google search was: Quercetin HPLC analysis
Good luck.

[lmh]

quercetin and its glycosides are amongst the most common and studied compounds in the plant kingdom, so there are certainly hundreds of published methods, possibly thousands. Every year there's another paper saying "we put some flavonoids down a C18 column and look! They all came out separate!"

The fundamental problem is that a crude extract of tomato contains a phenomenally large number of compounds that will absorb in the UV, so you will never get them all separated in a single LC run. If you believe that something like rutin is the most abundant compound in your mixture, then you may be able to come up with an LC run where the rutin peak is sufficiently separated from other biggish peaks, and contaminated only by tiny peaks, such that you trust the result, but HPLC is always going to be wobbly on crude extracts.

But James_Ball is right; the more resolution you have, the better your chances, so it may help a bit to change to a Kinetex-type column (solid core) so you can get the effect of smaller particles/longer column while keeping a back-pressure that's appropriate for your system. You can experiment with flattening the gradient over the region where the coeluting peaks appear (but this has diminishing returns; once you're flat enough, you get little by going flatter). If you can't resolve the unwanted peak from genuine quercetin, then ultimately a different mechanism of retention is good. James_Ball suggested CN, I would add something like phenyl-hexyl to the list, or if you're a Phenomenex user, PolarRP. I haven't tried the perfluorinated phenyl columns myself, but have heard amazing things about them. The point of a phenolic column (and is CN doing the same thing, James??) is that it will bind by pi-pi interactions, so it's good for phenolics, which quercetin is, but selectivity is very different to C18. But note: these columns need to be used with methanol rather than ACN as the pi electrons in ACN compete.

Personally, on flavonoids, I like to have MS as a back-up, but I know that's a luxury not everyone has. My feeling is that UV is more quantitative, and gets round the problem of standards for things that aren't commercially available, but if my UV peak areas don't correlate with MS peak areas, then my UV detector has been misled by a coeluting absorbing compound.