Page 1 of 1
impurity separation
Posted: Tue Nov 17, 2009 5:22 pm
by mskishore13
Hi all,
I am working in separation of Impurities for cortocosteroid molecule. I used RP-C18 (4.6X250mm,5microns). I have one inpurity that elutes immediately after the Analyte. I could separate but not to baseline. The peak elutes at 2 units height of analyte (right side of the peak). I stated at 10 height and the maxminum that i could obtain was 2 units. Can anyone know what the specifications for impurity separation, is it base to base?
what are the other thoughts?
Thanks,
K
Posted: Tue Nov 17, 2009 5:47 pm
by grzesiek
"Can anyone know what the specifications for impurity separation, is it base to base? " - if you mean baseline separation, there is no such specification, unless you made your own
"what are the other thoughts? " - change selectivity
Posted: Tue Nov 17, 2009 6:01 pm
by JGK
Most chromatogaphers would try for a peak resolution value (R) of 2 (baseline resolution occurs at R = 1.5)
Posted: Tue Nov 17, 2009 6:17 pm
by grzesiek
"baseline resolution occurs at R = 1.5" - rarely true as some tailing usually occurs
R=2.1
Posted: Wed Nov 18, 2009 1:23 am
by mskishore13
Thanks for reply.
I got the R=2.1 nut the peak is not base to base separation.
Thanks,
K
Posted: Wed Nov 18, 2009 8:04 am
by Mattias
If you have small impurities eluting close to the main peak, it is preferable to elute them before the main peak, due to tailing (as mentioned). But that is not always possible to achieve. Look if you have differences in acidic groups of the molecules, then usually pH of the mobile phase is the most powerful parameter to play around with, rather than changing column brand.
Posted: Wed Nov 18, 2009 10:34 pm
by tom jupille
We get into this all the time in our courses.
The "Rs = 1.5 is baseline" is not strictly accurate, as has been pointed out. In fact, that value was originally called "99% baseline resolution". Over the years, the "99%" part has been dropped off. The origins are that at Rs = 1.5, there is just less than 1% overlap between two equal-sized, Gaussian peaks.
If you have a big disparity in peak sizes, then Rs=1.5 will be inadequate even for symmetrical peaks. If your large peak tails (which is often the case when you overload sufficiently to get detectability for the impurities), you will need even more resolution. It's not unusual to require Rs values of 4 or more to get good quantitation in such cases. There is no simple rule for how much resolution is required. The "proof" comes when you validate the method, if you can get adequate precision and accuracy for your purpose.
Posted: Thu Nov 19, 2009 1:19 am
by Bryan Evans
Below are steroid separations on Cadenza CL-C18 (250x4.6mm, 3um):
http://www.imtaktusa.com/site_media/fil ... TI324E.pdf
http://www.imtaktusa.com/site_media/fil ... TI183E.pdf
You may need more horsepower (meaning switching from 5um to 3um)
Imtakts 250x4.6mm (3um) column can generate 50,000 plates on conventional HPLC. The system pressure
will be below 25 MPa.