DSC Metal hysteresis (melt/cryst)

Question

So we were analyzing Polymers using DSC and calibrated the system using Indium. I expected it to be a single transition only as Indium only has 1 known crystal structure and as I never worked with DSC nor seen a real thermogram before I was quite surprised to a relatively large hysteresis between metling and recrystallizing. I expected it to be a big gap for polymers but for something as uniform as a metal with only one crystal structure it forms I thought the point of melting and crystallizing should be quite close together.

I remember the lecture on solid state chemistry about DSC's of crystals where Ehrenfest and displacive/reconstructive changes in for example quartz-phases had hysteresis and I think displacive ones didn't have any but it was never really explained what caused this gap. I always imagined that it takes more time to form a symmetry like going against entropy and this is why more energy has to be taken out of the system.

But is there a better explanation for pure metals? Does this gap tell me anything about the structure in the end? I imagine that structures which tend to me rather amorphous might have a smaller gap because it's not so much symmetry involved in the process which would then also imply why it is big (relatively, I haven't compared it to any other compound yet) for Indium.

Answer 1

Disclaimer: the following is based on using DSC data for organic materials.

Is the melting temperature of Indium (156 C) just at the temperature where your sample exhibits a phase transition / decomposition? In general, I suggest to use the flat indium pellets and to record their melting for calibration of the DSC, rather than their recrystallisation. This is likely described in the manual of your DSC. (I does not hurt to have a look at DOI: 10.1007/s10973-015-5186-8, or similar, either.)

Because of its transient character, it is worth to keep an eye on the heating rate / cooling rate while recording DSC data. Provided enough sample material, sometimes a quick screening with 10 K/min is followed by a second recording at a rate of 4 K/min (or slower). While the endothermic peaks related to melting may be sharp, the ones for the recrystallisation are often shifted towards lower values because without proper seeds of crystallisation, supercooling may occur, especially organic samples and polymers.

I recommend to consider cyclic DSC, instead of a simple single-heating experiment, as this offers you additional insight about the sample in question. It is applied to characterize pure metals (like DOI: 10.1098/rspa.1996.0113), and may shed a light on the degree of crystallinity and processes of reorganisation of your polymer sample, too. In the following, a DSC of polylactic acid:

(source)

Sometimes these differences about recording the melting point (pristine sample, after 1st / 2nd heating cycle) are spot on small molecule materials, too. They may appear even if the sample did not incorporate solvent molecules.