Page 2 of 2

Posted: Tue Sep 18, 2007 2:26 pm
by Mattias
Why not consider the 1.7 µm columns? If you can get the peak witdh down to seconds you will have a decent chance to get at least partial separation. You will prefarably need a UPLC, but you can get quite far with a standard LC (<400 bar) with optimised components (minimised dwell volumes, small volume flow cell).

With a 100 mm column you can typically use 0.4 ml/min without exceeding the pressure maximum.

Posted: Tue Sep 18, 2007 9:06 pm
by firework1
Ha,ha, you are talking 2 dimentional HPLC

Tons of research articles, but not easy to apply in real life

see this attached paper, it tells the major technical difficulties

http://web.ebscohost.com/ehost/pdf?vid= ... 4%40SRCSM2

Posted: Wed Sep 19, 2007 6:54 am
by HW Mueller
firework1, true, some are difficult to set up, but once you get it going it works when all attempts at single column or series columns fail. A real life example: J Chrom B, 678, 137 (1996).
Also, this is not really rare at all, as all methods using SPE for cleanup are two step chroms (two dimensional).

Posted: Thu Sep 20, 2007 4:16 pm
by juddc
Thanks for all of your comments.

Due to a serious lack of time lately to really get into the meat of such a separation during the day (I have validation protocols to write), I've been running my AMDS nightly on a series of samples containing mixtures of 6-8 or so analytes grouped roughly by retention time. The composition of each sample overlaps each "adjacent" one by at least 2 compounds on both ends. For instance, sample one contains the first 6 eluters, sample two contains 4-12, sample three contains 10-18, and so forth.

I'm using alot of solvent, but thus far results appear to be promising and it's automated, so what the heck.

With care and some tweaking, I hope to be able to string optimized gradients together, check the effects of column dimensions / particle size for further optimization, and come up with a reasonable method on a single column.

So far I have 13 well defined peaks (min Rs ca 1.7), except for the cis-trans isomers of one, with which I'm not terribly concerned.

If it works, I'll post a chromatogram. If it doesn't, I'll go the two column route...and maybe try the whole thing over again that way.