I cant see Tm of two peaks for these PP & PA6. Did you use compatibilizer?

Hi, you can calculate the chemical composition from EGA @ Gerlinde Rusu

Dear Gerlinde Rusu,

a calculation of the composition of the blend from the melting peaks integral is impossible. The melting peak not only depends on the composition but on the degree of crystallinity as well. In reality you cannot have a crystallinity of 100%, moreover the degree of crystallinity depends strongly on the polymerstructure, therefore PA6 and PP in your mixture can differ in their degree of crystallinity. A calculation might be possible from a quantitative 1H-NMR spectrum.

First, take the DSC of pure PP under the same thermal history in DSC and measure the heat of melting of pure PP (j/g). Then calculate the composition of the blend. It is better to use a second run. Heat both pure PP and the blend up to 300 C with an appropriate heating rate. Then carry out cooling with the same heating rate and wait for about 10 minutes, then carry out the second run.

Another method: Find the heat of melting (j/g) of pure PP from literature. Calculate the PP content of the blend from the DSC graph.

Dear Gerlinde Rusu

I afraid that the DSC output you presented is more probably for PA66 blend with PP rather than for AP6 with PP. However, the quantitative analysis method is based on the principle that the melting enthalpies of the components in the blend are proportional to their amount. This means that the influence of the crystallization of blend components be neglected.

Now, there are more than one investigator treated this point as the Askar Karami [1] and M. Kisiel et al. [2]. They developed a linear correlation for allowing eliminating discrepancies between calculated and actual quantitative blend composition. Now if I take the sample as it is in the question detail, i.e., PA6 contaminated with PP, then go through fig4-20, fig4-21c and fig4-27 in [1] which shows a plot of the fit with 95% confidence intervals for PP/PA-6 blends using DSC analysis. Now, applying the function in fig4-27 or the equation 4-8 page 82 for 8.189j/g PP melting enthalpy and then you get amount of 11.9% PP in the blend.

Hope my contribution is helpful and any feedback is welcome.

Best regards…

[1] https://tspace.library.utoronto.ca/bitstream/1807/13261/1/NQ41184.pdf

[2]https://www.semanticscholar.org/paper/Quantitative-analysis-of-the-polymeric-blends-Kisiel-Mossety-eszczak/9afe8471cddbb74200ca07e6bdec70f8276c4475

Quantification of composition by thermal methods is likely to be misleading, even for those with high linearity. The root cause is that it has some intrinsic issues with respect of thermal measurements, not only those points discussed above regarding crystallinity of the tested materials, etc..

For instance, interactions between components, not only chemically but also physically, is neither predictable nor controllable. This would result in unreliable quantitative results. In fact, most linear quantufication by DSC methods were built up by errors compensated by another type of errors, making it with perfict R2 values > 0.99! The method is, however, has no values, albeit generate some gigo publications!

Yes, I agree. DSC gives approximate value. Accurate ratio would be probably obtained the microanalysis. FTIR can also be used. However, one should spent a lot of time to get a calibration curve.

More accurate quantifications of components can be acheived by Raman Micro-Analysis, and/or XRPD depending on the tested materials. Subject to the materials, the former can generally detect/quantify the content better than 0.01%(w/w), whilst the latter ~1 to 5%(w/w).

FTIR can also be an option for detection/quantification. I personally prefer Raman as its peaks are Gaussian types, whilst IR signals mostly noise shapes except higher intensity. This is the cause of less certainty of LOD and/or LOQ by FTIR.

By thermal methods, you can actually perform an appropriate quantitative analysis, but strictly subject to the materials. You have to know in depth the nature or thermal properties of your materials and learn how to control your measurements, e.g. sample weight, initial temperature, scan speed, and the particularly the event that can be used for quantification. My experience is combination of TGA and DSC could greatly improve the certainty of quantification results and accuracy.