There is a debate going on at our company about the correct way to calculate the purity of a primary standard. For example, if you tested HPLC area%, ROI w/w%, water w/w% and residual solvents w/w% all from the same sample, would you calculate it:

= HPLC area% x (1-ROI-water-RS) or
= HPLC area% x (1-ROI%) x (1-water%) x (1-RS%) or
is there another way? Any comments?

Just my two cents, but I don't believe either equation is correct. Assume the organic compound of interest, A, contains 1% organic impurity, B (by HPLC), and 1% each ROI, water, and RS. The purity is then:

%A = 100-%B-%ROI-%water-%RS = 100-1-1-1-1 = 96%.

Your first equation assumes that the HPLC area % is determined on the sample after the other impurities have been removed, hence the HPLC area % is adjusted for the additional weight of the other impurities. The second equation assumes that each of the impurities are determined sequentially on the same sample, first RS then water then ROI then HPLC area %, and that each impurity has been removed before proceeding to the next determination. I don't see how this is possible since these are generally destructive tests.

We determine the percentage of each impurity independently on individual aliquots drawn from the bulk sample, then subtract each of their contributions from 100% as shown above.

Best Regards,

Michael Hinsberg
http://www.hinsbarlabs.com

I guess the only difference is that the HPLC impurity was calculated by area% not by w/w% as all the other tests were. The weight of the sample or the other impurities in the sample does not matter when calculating area%. What do you think?

Aha, I think I understand now. If none of the other impurities show up on the HPLC as peaks and you are relating the HPLC % of the impurity directly to that of the main component then, I believe, your first equation is correct. Assuming, of course, that the response factors of the main component and impurity are equal.

Best Regards,

Michael Hinsberg

http://www.hinsbarlabs.com