Capillary gas chromatography is a very powerful tool when doing synthesis, moreso than most researchers realize. It requires that you are working with “volatile” compounds, but you’d probably be surprised how much falls into that category. Of course, anything you can smell is about 100% certain to be OK, but many things without odor are fine too. For example, cholesterol is an odorless solid (bp 359 C) but goes through a GC in about 10 minutes at an oven temperature of 300 C. Of course, sugars, proteins and other polymers, salts, strong acids or bases are no good; be sure you have none of these (or water) in your samples.
A properly set-up capillary GC should give peaks that are around 0.04 min wide or less at half-height. Some polar materials, esp, H-bonders (c. acids, some alcohols, maybe amines), can give broader peaks, especially on aged columns. But with those exceptions, your peaks should be nice and sharp. If not, make sure your split ratio is at least 40:1, and that your column is 0.25 mm ID or less (0.32 is OK, but never the 0.53 mm “megabore” columns). If you are measuring theoretical plates, the value should be roughly in the vicinity of 100,000, depending on the length of your column.
About using GC in synthesis: Reactions are often easy to follow by GC, because virtually every organic compound present is very likely to be separated. You do have to consider if things are present (acids, bases, catalysts, water) that are no-nos for GC, so you might have to do a mini- water workup followed by drying, but these are easy. Mixtures are typically much better analyzed by GC than by NMR, because NMR gives many peaks per compound and GC gives one. GC is more sensitive than NMR, and you usually don’t need an MS detector; using the rule “bigger compounds take longer to elute” and vise-versa, you can often easily interpret the chromatogram. I have even analyzed column* fractions by GC when TLC wasn’t working well.
*Better than column chromatography on smallish scale (up to one gram) is radial chromatography. This is faster, easier and gives a “TLC-like” separation, better than flash chromatography. The only real competition is the automated, cartridge-based systems which are much more expensive and I believe only give a “column-like” separation. I hope to contribute to the revival of radial chromatography, and even for chiral separations.
About not getting “carry-over” (where residue from the previous sample gets into the next sample): you can entirely avoid this very easily. First start with a clean syringe. Make sure it is picking up pure solvent well, then take up 1 microliter of pure solvent, followed by one microliter of air, then one microliter of your sample, the pull the plunger back a few more microliters so you can see the sample plug. All this is done simply by plunger location; sample won’t be visible until the end. Wipe off the needle before injection. When you inject, the pure solvent washes out any residue of the sample. The critical feature here is that the sample never contacts the plunger directly. Otherwise cleaning is much more involved and carry-over is more likely.
About chiral separations: In my experience on achiral GC columns, if two compounds show hardly any separation after being retained, say, 10 minutes, there was no hope of achieving a baseline resolution no matter how long you left the compounds on the column. But I was surprised to discover that chiral separations are quite sensitive to ramp rate/time left on the column. For example, the two enantiomers of a lactone were barely separated (by 0.05 min) at 12 minutes with a 5 degree/min ramp on a commercial cyclodextrin derivative column, but approached baseline resolution (separation = 0.18 min) with a 2 degree/min ramp at 23 minutes.