I need to select a column for a secondary confirmation of a compound and its polar metabolites formed in soil. The matrix is difficult to deal with but I am able to quantitate using an SB-Phenyl in the reverse phase with acetonitrile and water and 0.1% formic acid. Does anyone have suggestions as to how I should proceed selecting a normal phase column and mobile phase? Any suggestions or direction will be appreciated.

There's a very good section on normal-phase in "Practical HPLC Method Development" by Snyder, Glajch, and Kirkland (Wiley, 1997), but a brief synopsis is:
- the most general starting point is a cyano bonded phase column. If you need more retention, try a diol column. If you want better isomer / shape selectivity, bite the bullet and go to bare silica.
- you can start with a gradient from 1 to 100% isopropanol in hexane (20 minutes @ 2 mL/min on a 150 x 4.6mm column). This will allow UV detection down to about 220 nm or so. If all of the peaks you are interested in elute within a 5 minute window, then you should be able to do the separation isocratically (since you have an isocratic reversed-phase separation, you will probably also be able to run isocratic normal phase). Look at the concentration of iPA at the midpoint of your bunch of peaks and cut that in half as a starting point for an isocratic method.
- tweak the iPA concentration until you get a reasonable range of retentions (usually, you don't want the k of early peaks to be below 1 and you want late peaks coming off before k = 15 or so).
- if you want to try different solvents, the choices are usually ethyl acetate, methylene chloride, MTBE. The column manufacturer should be able to point you toward equal solvent strength diagrams to let you approximate equivalent concentrations as you shift from one to the other.

There's obviously a lot more detail to the process (and there are lots of ways to go about it!) but this should give you an idea.

-- Tom Jupille / LC Resources Inc.