The elements that are not usually determined by ICP/OES fall into three basic categories. The ﬁrst category includes those elements that occur either as trace contaminants in the argon gas used in the ICP/OES ($C$ from $CO_2$), constituents of the sample solvent ($C$, $O$, $H$), or as contaminants from the environment or atmosphere ($N$ for example). The second category encompasses those elements that require high excitation energy, such as the halogens. These elements could be determined with poor LOD, however. The third category is the family of short-lived radioactive elements that are commonly determined by $\gamma$-ray spectrometry.
So, apparently, when I'm working with, say carbon-based fuels, and I want to determine $C$ contents, ICP-OES would not be the method of choice, at least not the standard variety. So, what else would be the typical choice? Or are there modifications to ICP-OES, such as using very pure $Ar$, or $He$, as the plasma gas and sealing the plasma chamber from the atmosphere?