If the mixture is soluble in a solvent, liquid chromatography will do. If their chemical formula is enough different, mass spectrometry coupled with TGA may also allow separating the different contributions to weight loss in the same tempearture range.
Probably, Alain's idea of tandem TGA is the easiest (if the equipment is available). However, the analysis can be sometimes performed even without degradation (or dissolution). If the components are known and give rise to characteristic signals, NMR may help. If the components chemical composition is well different, even element analysis can be a starting point.
My point is that there is no must to isolate the components from the mixture if you merely want to quantify the composition.
Thank you for the answer from Alain, Taslim,and Evgeny.
To Alain, is there a minimum percentage of each component to be detected by liquid chromatography technique ? for example, one component is less than 1%.
HPLC is highly sensitive (hence the HP for High Performance), so detecting less than 1% is quite possible. The operator, nevertheless, needs to use suitable experimental conditions for allowing a good quantification, which are known by any specialist of the method.
Do you know the UV spectra of each of the components? Do you know the FTIR of all molecules? Do you know about their solubility?
UV-vis: Just extract the components in a god solvent for all of them, and if they have separate peaks, you should be able to quantify all of them by the intensity of the peak (use of calibration curves).
FTIR: You only need to know one specific peak which is separate enough from the others, as well as from the polymer matrix, and with enough intensity to be detectable. Then the area under the peak should be related to the concentration (use of calibration curves or standard addition)
NMR: Find a good solvent for all of them, but not for the polymer matrix, and run NMR. You should be able to determine the concentration by means of the area of characteristic protons for each compound (no need of calibration curves)
MS, HPLC, and other techniques are complicated, not available and sometimes expensive.
You are welcome. If you need more assistance, please write me.
Concerning the DSC... I do not believe that this thermal method would be suitable for determining the concentration of 2 to 3 components into a polymer matrix in the range of ppm of below 1%. Of course first and second order transitions might vary depending on the amount of mols of "impurities" present in the mixture. Forget DSC, please.