I feel that screening is usually the best approach, unless you have prior experience. In a previous job we had a column switcher with six columns, and tried 3 methods on all 6 columns. We loooked at the best result from those 18 runs, and optimzed from there. It sounds like a lot of time, and it was, but most enantiomers were separated on a Daicel AD-H or OD-H. IF you decide to go with this approach, I'd suggest a few different types or columns, like some cellulosic/amylosic columns, a pirkle column, a protein column, etc.
However, if you want to try as few columns as possible, you can contact the manufacturers (
http://www.chiraltech.com/ and
http://www.registech.com/ are a couple) and ask their applications people what they recommend for your structure (if you are allowed to send the structure out). Some of them even have application guides which you can sort by functional group type to find one similar to the one you are working on.
And I found out Regis offers a free screening service (
http://www.registech.com/Markets/Chroma ... ening.html) so you may want to check them out if you decide to go the chiral column route. Best,
Chris
Chris Singleton
DMPK
BiogenIdec