What steps can be taken to avoid overloading on a gc-column WITHOUT sacrificing resolution? (Note: I don't want to inject less!!) I know wider bore columns do have higher capacities, but the efficiency of these columns is also poorer than small id columns (H=dc)

If the column is overloading, why would you have an issue with injecting less material? That doesn't really make sense to me. If you are worried about the relative volume error being higher with a smaller volume you can dilute the sample and inject the same volume at a lower concentration. Is there a specific reason for not injecting less even though the column is overloading.

Are you overloading the column, or the detector?

If the problem is the detector, you can help the problem, under some circumstances, by slowing the chromatography so that the peak elutes later, and is therefore wider. This may not help much, but if you are running very short times it might.

And, of course, if using a split injector, increase the split ratio so less gets to the column. If splitless injector, investigate split techniques.

More details would help.

Sorry for the lack of information. I have the following dilemma. I am interested in the identification of very minor peaks which elute very close (after) the solvent peak on a Wax column (DBWax, in my case). I have tried a 1ul splitless injection, but then the chromatography is horrible. An even worse separation is obtain on a non-polar column (PONA), because these minor compounds have the same boiling point as the solvent.
The main reason why I would like to get more material onto the column is that these minor compounds are visible on a FID, but their concentration is too low for an identification with mass spectrometry (a MSD, in my case).

Now it makes sense, you have a chromatography issue, not a problem with overloading the column. You will probably need to try a different phase and more than likely a column with different dimensions. If you go to a narrower id column you will have a more efficient separation, and may be able to separate the compounds of interest from the solvent. A longer column may also help. If you have valving on the GC (which I realize is not likely) a valve switching of the majority solvent to waste before the analyte peaks will help in the separation.

I agree with Anon 07:41am but going to a smaller ID will only give less sample onto the column (while it might improve the detection due to better separation as noted before).

You may improve separation and capacity of the column you are using now by adjusting the flow rate and temperature of the injector and oven very carefully.

One additional trick you can use is to butt connect a length of narrow bore tubing (1-5 meters) to the TAIL of your present column. This will simulate a longer column. You will have to increase pressure to achieve the same amount of flow if you do this.

Rodney George

So the real issue appears to be the identity of these small peaks? Since you are considering column loading issues, I am guessing you are not using long capillary columns on your FID system.

To address the separation issue on the FID system...
Use a different solvent, if possible, to make the separation a little easier. Perhaps methylene chloride will work, or iso-octane. You may be able to move the peaks in FRONT of the solvent peak using this trick, or by using an even more polar column, such as one of the Porapak (Supelco?) packed column phases (obviously, not for MSD use). It sounds like you are essentially doing solvent analysis, so look to the catalogs for column ideas.

Splitless injection should not give 'horrible' chromatography, if it is done correctly. It may not be easy to get it right, but amazing things can be done. Check the chromatography catalogs for examples to see if your system seems to be working correctly. I have run GC's that had unique injector problems that required creative solutions for certain situations. We have almost all HP machines here, and some of the older ones were tricky to set up. But just a couple weeks ago I pushed toluene out to 10-12 minutes while working with a messy hydrocarbon analysis (paint thinner). The injection solvent was carbon disulfide, and it worked great.

To address the identification issue ...
Dilute less, or somehow concentrate your samples to a concentration you think your MSD will see reliably. That would probably be in the 10 ng/uL range (you must base your guess on the response factors and concentrations of the known analytes). Put the longest, most polar column you have in your MSD, set up a slow temperature program, make sure splitless injection is working correctly, and run the samples. This will give you all the separation power readily available, and probably exhaust the technology available to you. Anything else is a research or method development project to isolate these compounds, and should be treated as such. It will take time, money, talent to solve the problem, if it really is that important.