By Anonymous on Monday, May 17, 2004 - 05:58 pm:
Dear all,
What should we do to validate an analytical method for natural products, ecpecially in accuracy or recovery?
Appreciate any advise.
By Jan on Monday, May 17, 2004 - 11:49 pm:
If your method consist of quantifying some tracer compounds, buy the standards and spike them to your sample. Then check if you recover them with your sample prep method; you should find 100 % although for trace amounts an inaccuracy of plusminus 10% could be accepted (depending on the compounds).
What compounds are you analysing ? In what matrix ?
Jan
By Anonymous on Tuesday, May 18, 2004 - 12:32 am:
We analyze tablet/capsule that contains plant extract. Sometime we need to determine total of certain groups of components, eg. total phenols, total alkaloids, etc, with using only one component among them. By using this one calib curve we should measure all other components with certain formula. How can we spike it? On one component or all of them?
By Jan on Tuesday, May 18, 2004 - 11:57 pm:
If you have all of the components as pure standards, of course it's better to test recovery for every single one of them. But mostly only your external standard - compound is available. If this compound is of the same chemical family as the components your are measuring, then this should be sufficient. We suppose then that if for instance rutine is extracted from your sample with 100 % recovery, other flavanoids "behave in the same way"...
Another possibility is to compare recovery from the pure extract with recovery from your formulated product (absence of influence on accuracy from tablet excipients)
Jan