Chromatography Forum

Hi:
Recently I submitted a papper about the analysis of active compounds in medicinal herb by HPLC method.
Because the compouds could not be purchased, so we prepared them by ourselives. All these compounds were identified by direct comparison of their spectral data (UV, IR, NMR and MS) with those reported in the literature. The purity of the standards was checked by HPLC/DAD at three different wavelengths (210, 252, 300 nm), and the results suggested their purity to be above 98%.
But the reviewer gave us some advices as below:
"it indicated the purity of four standards (prepared by authors) was checked by HPLC/DAD and got a result of >98% purity. To the best of our known, similarity test of spectra at various places in a peak can be employed to calculate the peak purity, but it is imprecise, and there are many other methods of purity test on HPLC/DAD reported. So, the author should give some explanation on the calculation method of peak purity test, and some literatures involved expected to be cited."

Could you give some literatures or some good method for how to identified the purity of srandard substances?

You actually try to define a compound purity without reference standard. We have to distinguish the chemical purity and chromatographic purity. When we use DAD to determine the purity, I am not sure you can assign a number accurately. You may be able to determine whether there is a peak co-eluted under the peak of interest or not.

Once you have a purified compound, you need to run an HPLC to determine a chromatographic purity. It will be a gradient run and with the capability to quantify, hopefully, all the impurity peaks at 0.1% level or lower. Then the resulting % of your major peak (by peak area normalization) is the % of chromatographic purity. If we have a reference standard, we can inject both standard and sample onto HPLC column to determine the chromatographic purity.

The chemical purity is the composition of the purified compound. You have to use one of the absolute analytical methods to determine the purity, like DSC, gravimetric, titration, coulometric, etc. The compound may contain moisture, organic solvent, metal etc and those components cannot be determined by HPLC. My recommendation is to confirm the chromatographic purity first and then to determine the chemical purity.

Spectra yield only the identity of a compound but not the purity. Peak purity with the DAD gives some hints but not more. If you prepare (what do you mean with "we prepared"? was it synthesis or isolation?) compounds you should perform elemental analysis; this technique would show if there are water, metals, other impurities etc. Of course, a 100% match with the elemental composition is no proof of purity (there can be other compounds with the same composition such as isomers) but the opposite is a proof of non-purity.

Elemental is good, but can do quantitative analysis by spectroscopy without a standard of the compound.

Quantitative proton NMR is done routinely at our company using an internal standard. Usually pick an internal standard with a very simple spectrum that doesn't interfer with an anlyte or react with it. Easy to calculate the weight percent purity from the amount of the internal standard added since a mole of protons is independent of compound if not exchanging.

Can also do carbon NMR, but have to be much more careful with experimental parameters.

Do a google search for internal standard quantitative NMR for some examples.

I agree with cctying. First of all, since you have done the spectral identity check by UV,MS, NMR and IR, then you can do other chemical purity check for moisture, residue solvent, element, enantiomer etc. Finally, use HPLC to determine the purity.