[kamiyu]

Hi there

As a synthetic guy, I always have to run column. In some cases, unfortunately the polarity of a compound on a column is different suggested from TLC and this is very annoying because the column does not work at all and the Rf of a compound on the column can only be estimated by observing the elution concentration/volume.

I learn from books that this results from the different silica gel size used in the TLC and flash column chromatography.

Have you guys also come across this kind of situation? Any suggestion about how to work with it??

[TheUnassuming]

I have come across this.  The solution is to buy the same silica on the TLC plates as you will be using for your columns.   
If that isn't an option you should (after a while and a number of compounds with different polarity) be able to make an educated guess based on the TLC what the optimal column conditions will be.   Then just make sure to start at a lower polarity solvent mix and ramp up slowly until things move. 

[Johte]

Don't know if it helps your scenario, but I recently had difficulty resolving an amine product from leftover amine reagent. They were binding too strongly too the silica and I didn't have C18 or alumina columns so I used a third solvent line with a base additive. 

It may not be your issue but a third isocratic solvent can make a world of difference!