If you want general utility and MS detection, I would say neither HILIC nor silica (but it's really splitting hairs): ANP with a Cogent Diamond Hydride column. http://kb.mtc-usa.com/article/AA-00346/96/
We stopped using other columns about 2 years ago because this one completely surpassed the other options on the market for hydrophillic molecules. Maybe there are more competitive options now, but I doubt there's anything substantially superior.
Hi Fuad, we use an amide column run in HILIC mode to separate pyridoxine, riboflavin, ascorbic acid, folic acid and cyanocobalamin. The resolution is very good. Pyridoxine and especially folic acid resolve into multiple peaks.
Perhaps I can help bridge the gap between the two options presented here. I did a bit of research, and found water-soluble vitamins can be run on both HILIC and Reverse Phase C18!
In the case against Si(2) and HILIC, that's an easy answer... HILIC all the way. About six or seven years ago, we used to sell Si(2) columns ALL the time. Now a days, people only buy HILIC... because it is very reproducible, and you don't have the water pre-saturation problem you have with Si(2)
For your convenience, I have also provided a couple of applications for you. Does this answer your question?
for polar analytes, the choice would be any normal phase column. Off course, the efficiency of different NP columns have to be checked for this specific problem
I'm totaly agree with Mr Paul Karch. It also depends of the investissement amount you have to perform your analysis (and what the versatility of the presume column after your separation)
In general for polar water-soluble vitamins you want to use a reversed-phase (RP) column that can tolerate high aqueous content (even 100% aqueous) as well as acidic conditions. These columns are usually either polar-embedded RP columns or hyperbranched RP columns. The following link shows work in my lab by Dr. Deanna Hurum with a hyperbranched style column. Using either style of column lets you avoid ion-pairing reagents.