Purity check of impurity..

[Pegry]

Hi everybody,

I need to make the qualification of some impurities (identification and quantification) to use them as working standard.
Unfortunately there aren't CRS for this substances..
I'm going to make the identification of the substances by their mass spectra (LC-MS), create a library and use it for the identifications.
The problems come for the analysis of the purity of the compounds:
It would be possible to measure the purity of my substances by HPLC-UV at three different wavelenghts (and calculate the HPLC-purity by area normalization)?
Can I also measure the water content and the residual solvents content and take these in accounts for the calculation of the purity?

I'm not sure that the procedure is correct, could please someone tell me which are the correct step to make a qualification of impurity without certified reference standard (USP or EP)..

Thanks everyone for the answer..

Davide

[ntruong]

It would be possible to measure the purity of my substances by HPLC-UV at three different wavelenghts (and calculate the HPLC-purity by area normalization)?

You will have to pick one wavelength that gives you the best shot of your substance and its related coumpound. Use the area % method to quantify the related compounds, not the main peak.

Can I also measure the water content and the residual solvents content and take these in accounts for the calculation of the purity?

Yes, these are parts of the equation to assign the purity value for your compound. If you could determine the inorganic volatile (residue on ignition) of the compound, it would be nice. Then you could assign the purity value by subtracting it from 100 (100 - area % of related compound - water - residual solvents - inorganic volatile)

I'm going to make the identification of the substances by their mass spectra (LC-MS), create a library and use it for the identifications.

The LC-MS can only give you the mass, you still have to confirm the structure of this compound by NMR.

[DR]

Peak purity can be addressed as a function of multiple wavelengths (DADs are good at this, if nothing else). As long as you don't have co-eluters with highly similar UV spectra, DADs can confirm or refute the purity of a given peak (if that's what you were after).

[Pegry]

Thanks everybody for the answers..

Dear Ntruong the determination for the water and the residual solvents could be done with simple method like loss on drying or I need to measure the water by Karl Fisher and the solvents with gaschromatography?

In your experience is acceptable to use this procedure (assign the purity value by subtracting it from 100 (100 - area % of related compound - water - residual solvents - inorganic volatile) also in regulated enviroment (GxP)?

Thanks again

Davide

[ntruong]

the determination for the water and the residual solvents could be done with simple method like loss on drying

In theory, yes. However, I will have to disagree with you on the pratical side since I couldn't get it to work every time I try to prove it.

measure the water by Karl Fisher and the solvents with gaschromatography

Yes, This is the route that I would go with. Do you have a validated method for these?

In your experience is acceptable to use this procedure (assign the purity value by subtracting it from 100 (100 - area % of related compound - water - residual solvents - inorganic volatile) also in regulated enviroment (GxP)?

If the USP doesn't have the compound that you made, this is the only way to go for initial characterization. I believe it is acceptable in the GxP environment because we get audited by the FDA and external auditors all the time and they said nothing about it.

[jzt]

we qualify impurity standards based on their "intended use". if this standard is going to be used as quantitative standard to estalish the amount of such impurity in your sample, you will need to determine the "mass purity" taking into accounts of organic impurity (HPLC area%), water, residual solvents, and inorganic impurities (residual on ignition). the equation described by ntruong will work for relatively pure materials.

If this impurity standard is going to be used for identification purpose only (to mark which peak in the chromatogram is this known impurity), then getting purity number from the HPLC area% of the main peak is acceptable. The wavelength should be the same as in the method that this impurity standard is intended for.

in both cases, the identity of the standard must be proved, this usually required more than just LC-MS. we typically use NMR, MS and IR. the subsequent requalification can be based on MS or IR confirmation.