I"m doing FAME work on watermelon seeds...I'm not sure how to go about creating a calibration...I understand that I need 4-5 levels for a good curve, but I'm not sure what concentrations each of my dilution solutions should be. I have a FAME stnd kit with a concentration of 33 mg/mL in 1mL of heptane, and a methyl ester internal stnd of 100 mg in liquid form, which needs to be added to the standards and analytes (in what concentration I also don't know). All of the literature on watermelon seeds i've found only gave % comp. for fatty acids, never actual concentrations, so I"m not sure what to expect and what concentrations my calibration should therefore be at. Can someone help get me rolling on this?

Wow, do you really need so many calibrations for something like this? I'd find out which fatty acids were present (either experimentally or through literature), order those in pure form from someone like NuChek Prep, make up a calibration standard of similar profile, then do my assay.

I've done a literature search on which fatty acids are present, and from the few I"ve found, I have a basic idea of what is going to be show up. However, the literature didn't completely agree...some works found more fatty acids than others. Furthermore, I'm expecting the fatty acid comp. to be different for the different seed varieties. I have a standard kit of FAMEs in pure form, but I was going to do a series of calibrations like i mentioned because as I understand it is necessary to do so in order to more accurately determine concentration of fatty acids in the analyte, since using a split injector will change the concentration/volume ratio on the chromatogram and since detector response can change. I could be way off base here...I'm just going on what I've read and my few years of GC experience..

There is going to be some trial and error involved. You need to prepare some watermelon seed extract to practice with. One way to approach this is to make a series of evenly spaced dilutions (arithmetic, not logrithmic) of your calibration mix to cover a large range of concentrations. Determine over which range your curve is linear. This will be quite a large range. Then I would prepare a test sample of your watermelon seed extract and begin analyzing dilutions until your samples fall in the middle of your chosen standard curve. Do not use the split ratio as part of your quantitation. Run all your samples at the same split ratio as your calibration standards.

After doing this you should be able to estimate the correct quantity of your internal standard to be added to your samples and calibration mix.

I'm not sure what you're saying with the comment that using a split injector will cause the concentration/volume ratio on the chromatogram to change. In a properly functioning system there shouldn't be injector discrimination, and if you use an FID detector as most FAME analyses do you should have good linearity over a wide range of concentrations, as stated above. In most cases a three point curve bracketing the expected concentrations should be enough.

In our lab we get equivalent results on FAMEs using GC-FID on either packed columns or wide-bore capillaries of same phase (either split or on 0.53mm capillaries). We've used a quantitative fatty acid methyl ester standard (Alltech mixture #19007) to document detector response and splitting efficiencies.