• Members of the previous forum can retrieve their temporary password here, (login and check your PM).

Thermal Analysis of Polymorphic Transitions of Pure DMT

Migrated topic.


Rising Star
Howdy Friends,

Been a while since I posted. I was running some thermal analysis on a recent batch and noticed some preeetty interesting behavior I thought I'd share with you all.

This last batch (Max Purity: 99.19% šŸ˜ ) demonstrated something I thought was strange on the DSC thermograms. I have been producing crystals (heh - with difficulty, out of amorphous extract) with a characteristic melting point [Tm(1) 68.9-69.8] similar to that described by Shulgin and Shulgin [Tm 67-68] (The amorphous starting material melts at ~58C). On the thermograms this is indicated by the large endotherm seen in the first heating cycle

During this first heating cycle, the crystalline material (Samples 1,2,4) exhibits a single melting peak at ~69.65C, and absence of exothermic event on the cooling.
During the 2nd heating cycle things get interesting - Sample 1 demonstrates double melting peaks without an intermediate cold crystallization peak at Tm(2): ~44-46C & Tm(3): ~56-58C, whereas Sample 2 demonstrates a single melting peak at 46C, without any apparent recrystallization peak anywhere.

Sample 1 was a single crystal, 8.4mg, (TZ Alum pan, crimped) which actually demonstrated some amorphous residue on the outside of the pan afte ranalysis
Sample 2 was very fine crystals which were recrystallized again, 5.27mg, no residue after.

Heating 1 was performed at 1K/min, Cooling at 2.5K/min, and 2nd Heating at 1K/min
Temperature cycle was (RT) to 100C, isotherm 2 min, ramp to -15C, hold 2 min, then ramp to 100C.

There is a lot to be said, but it appears to me:
1. 3 Different polymorphic forms of DMT are clearly evident (as described elsewhere)
2. As seen during reheating, there seems to be a polymorphic transition that occurs as the 'Form A' is actually melting. The melting of 'Form B' then kind of balances any exothermal events during this transition. Any other thoughts?
3. An interesting observation is that all of the endothermic peaks seen during the reheating cycles are extremely high purity, >99.95% in most cases. I wonder what is happening after the original crystal melts? Obviously the higher MP form is the more thermodynamically stable, but it looks like the energy barrier to nucleate again is very high?

Thought everyone would find this interesting as well.
Attached are all the DSC thermograms for 4 Samples.
Sample 1: Crystalline, 2 melting peaks on 2nd heating
Sample 2: Crystalline, 99.19% Purity, single melting peak on reheat
Sample 3: Amorphous, exothermal peak on cooling with no subsequent thermal events
Sample 4: Crystalline, single melting peak on reheat but also with preceding exotherm
At the end is a reprint of mine of various literature values for DMT MPs.

Be very well,


  • Sample+1%2C+DSC+Thermogram+(with+Purity+Analysis).png
    186.8 KB · Views: 0
  • Thermal+Analysis+of+Polymorphic+Phase+Transitions+in+DMT.pdf
    726 KB · Views: 0
  • Sample+2%2C+Close+Up+of+Single+Exo-Endothermal+Event+During+Reheating+(Focus+Endotherm).png
    156.4 KB · Views: 1
  • Sample+1%2C+Close+Up+of+Double+Melting+Peaks+During+2nd+Heating+Cycle.png
    248.3 KB · Views: 1
  • DSC+Thermogram+Comparison+to+Previous+Lot%2C+Showing+Single+Melting+Peak+Preceded+by+Exotherma...png
    210.5 KB · Views: 1
  • Sample+3+(Amorphous)%2C+DSC+Thermogram+of+HCH+Cycle+Showing+Melting+Peak+on+First+Heating.png
    111.9 KB · Views: 0
  • Sample+2%2C+DSC+Thermogram+(with+Purity+Analysis).png
    103.9 KB · Views: 0
Welcome back and thanks for posting your results. While I'm probably not best placed to comment on the results, it does seem to my eyes that after the first melting/cooling cycle that sample material has crystallised as a mixture of all three polymorphs.

It would be interesting to do a scan that slows down the temperature increase in the transition range for those polymorphs. Maybe a sample can be produced with a higher concentration of the lower melting forms by cooling a melt (much) more rapidly? Perhaps that would give a stronger exothermic event. Slower heating might give a chance for the intermediate exotherm to make an appearance on the chart too, but I suspect that depends on the transition temperature of the highest-melting polymorph.

Do you have the capacity for including pressure in your measurements? That might help tease out a few more of the specific regarding the physical behavior of this substance.

I'd be intrigued to know what our other chemical experts (the ones more competent than me, especially šŸ˜ ) think of this too.
Thank you kindly downwardsfromzero
What you say is the other possibility, that the sample re-crystallized into different polymorphs which are melting in succession on the reheat. If this is happening I can't figure out why, because the new melting peaks are not consistent, and the exothermic event is so small that I want to know what else is going on here.
From one of the crystallization papers I put up with the final solubility curves of the 3 polymorphs, it looks like one polymorph is enantiotropic and the other monotropic. I have never been able to figure out what I'm reading, though, converting between the three is difficult to control in practice.
The crystalline material (MP ~69C) is difficult to produce from the amorphous starting material (MP ~58C). The relaxation time is on the order of days to weeks, where small spherulites generally come up after a few days and gradually grow to cover the entire material, and only then can you get bulk transformation to crystals from hot heptane.

Since the kinetics are very slow, I think you're right about trying to slow down the temperature changes during the transformation periods. I do not know actually about the pressure, so I will have to check back on that. I did find a temperature program that can elucidate these events better, it involves stepping the temp up/down by 0.3C and holding for 10 min, then stepping up another 0.3C and holding another 10 min, etc.
Regarding the speed of the cooling, I want to do both next time, decrease the speed to try and resolve the processes better, and also increase it because I forgot to this time - I went to -15C because I still can't find my Tg but I didn't quench the sample from the melt like I should have to get a Tg on the reheat.

Appreciate your thoughts, thanks!

Update: I forgot to add, that maybe the reason why the first sample gave the double melting peaks like that because of the way the sample was prepared? It wasn't ground, so there could have been regions of inhomogeneity with different temperature gradients etc. Somehow that contributed to the events on reheat?
Thank you very much for sharing the results of your experiments.

I don't have anything to add really, just here curiously waiting for more experimentation results :)

Be well
Glad to contribute!

I have samples being prepared now. Several interesting ones really. I managed to crystallize some strangely insoluble material (one of DMT's many poly-amorphic quality's I'm finding..) that I have had for quite some time waiting to figure out what to do with them.
Going to explore some different DSC parameters too, which should really help us get a better view.

I'd expect to post the update within a week, two at the most. I'll be sharing soon :)
downwardsfromzero said:
It would be interesting to do a scan that slows down the temperature increase in the transition range for those polymorphs.
With a slower temperature ramp the heat flow will be reduced. Hopefully you can see something though.

Hope to see more, thank you! StepScan could maybe separate those pre-melt crystallization peaks (or whatever you call them) from the melting.
This reminds me of some of the tinkering around with the molecule I was doing some years ago. In the end it more or less convinced me that it has a mind of its own. On more rational days I support the conclusion that much of this baffling behavior is down to the various modes of oligomerisation on top of the polymorphism of the monomeric material.

How do you control for inorganic contamination, particularly stuff with a notable basicity? Presumably you've followed Loveall's recent work on low polymer methods?
Ha, downwardsfromzero, you reminded me of something I just read on polymorphs, the phrase is " 'God-only knows' seeds ".

I'm going to attach an article that I just started reading that you should check out too. It's the first article I've read that starts to put the more irrational findings to sense.
There are "concomitant polymorphs", "disappearing polymorphs", .. there is a section on Cross-nucleation, how polymorph crystallization can depend on degree of supersaturation, temperature of crystallization, interfacial tension, presumable degree of super-heating as well...

Gotta love crystals, they're just so ... enigmatic !
Hey guys,

I'm a little delayed in posting my latest results, apologies. But, nonetheless I present a much better summary.

This last time I had 4 different materials to test, which I broke into 8 samples (n=2 for all 4 samples), run under different conditions.
For the summaries, I know how important it is to know how everything was done and how it began etc., so I have the physical description of the samples, their weights, heating parameters, and summary alongside the DSC graphs.

The samples:

All were crystalline. The first 4 came from the same source, the 2nd set being the first couple pulls of heptane after DCM extraction, the 1st set being the later pulls. Those crystals (Samples 1-2) weren't dried and it shows in the thermograms. Their counterparts, 3-4, were dried.
And then the interesting thing was, I had this .. problem where over a year ago I provoked the material I was working with to become utterly insoluble in everything I was working with at the time (even toluene). It turned into this dark brown amorphous wet product which after time hardened. Finally was able to convert it into crystals which were very different . Samples 5-6 represent nice crystals which came out of the re-x after solubilization. These looked normal and had the same MP as the others. Samples 7-8 represent spherulites, (I call them that because of the way they look but maybe that's not technically true?), which work like: the heptane pull will evaporate into a yellow oily mass which is stable for several days. Then, a circular flat crystal will appear, and then another will nucleate the next day, and these expand in all directions eventually merging with the other growing crystals and you can see the interaction patterns in the final form.. These had a MP ~3C lower than average (69.282, n=14, RSD=0.76), with much less enthalpy than usual.

And then, this time for the thermal parameters,

I went to 80C instead of 100C to limit superheating, and then went either back down to 20C or only 40C. My thinking for the latter was maybe there'd be a better memory of the previous crystals, idk. The cooling rates went down to 1C or 0.5C.
A couple of times I quenched it from 80C by supercooling to -30 at 40C in order to get a Tg.

Brief summary:
No double melting points this time, but I did get some exothermic events finally.
The first 2 samples didn't have any 2o thermal events, probably because they weren't fully dried (?).
Samples 3-4: Interesting. Sample 3 was the most intriguing of all. The cooling rate was 0.5, the reheat was 0.5. The exothermic event on cooling actually peaked when the temperature started to heat up again, at ~20C. The crystallization continued coldly with a peak in the 30s. Sample 4 was cooled at a rate of 1C/min to 40C, but didn't have these events. Seems the kinetics are super slow, next time I'll have to try a rate of 0.1 and see if the process occurs at a higher temperature. Further, the endothermic peak afterwards was huge, with enthalpy of 41 J/g. Usually they are ~5 J/g or less if I remember. And, what stands out too is the purity of this 2nd melting peak matched the starting purity. In all the samples in this post the re-melting peaks were extremely pure (99.95%). This wasn't the case here. The MP was 56. Like the spherulites, this represents a lower MP than is seen when melting for the first time, which is usually 59.09 (n=3). Why the melting peaks are a bit lower in these cases vs. the starting MPs is curious.
Samples 5-6 both gave melting peaks on reheat, with MP of ~45 for both (Average MP for the low-melting form: 45.89, n=5)
No additional events from the Spherulites

I'll attach the summary slides, the PDF has a few more slides, showing the full DSC graph for the quench cooled sample instead of the closeups, and a more close up look at that 3rd sample.

Ah, when looking at the results, when I say samples 1-8 here I mean 1A,1B, ..., 4B as 1-8.

Next time I will definitely have to do significantly slower cooling and reheating rate.

I'm trying to be mindful of the length of this post, but I found some great articles that seem to be good in helping understand what's going on. Just found them so haven't looked too much but like I was saying to downwardsfromzero, this idea of "cross-nucleation" of polymorphs seems to be a feature here, as well as "concomitant polymorphism".

Anyways, enjoy :)


  • Concomitant+Polymorphs+(Bernstein+1999).pdf
    1.6 MB · Views: 0
  • A+practical+guide+to+pharmaceutical+polymorph+screening+%26+selection.pdf
    919.9 KB · Views: 0
  • Thermal+Analysis+Report+2%2C+20Jun2022.pdf
    1,011.4 KB · Views: 0
  • Sample+4.PNG
    288.9 KB · Views: 0
  • Sample+3.PNG
    283.5 KB · Views: 0
  • Sample+2.PNG
    289 KB · Views: 0
  • Sample+1.PNG
    260.8 KB · Views: 0
  • Disappearing+Polymorphs+(Dunitz+1995).pdf
    1 MB · Views: 0
Thanks for the update... I'll read through things eventually. A random thought: could an additional variable be reaction with the aluminum pan (assuming that's what you're using)? Alodine coated pans can be used for this purpose (+ hermetic sealing under nitrogen if you're really worried).

Just looking for an excuse to bump the post :p
Top Bottom