Dear Anna N. Matveyeva ,

I faced some problems before with TGA curves that did not match before. From the graph you show, it seems that you've done the test twice? Maybe try doing a couple more. When I faced this kind of problems, I would do 3-5 different test to see if any curves could match up, and usually, two of them get pretty close.

I don't think it's you moisture content giving you difficulties because, for all of your pictures you've included, the first part of the curve match up well. Therefore I believe they have similar moisture content and that water is being evaporated at the same rate and same quantity. But if you would like to check the moisture content, there is moisture analyzer in which you put in a sample, and it gives you precise measurements.

What is your sample made of? I know you said polyethylene, but you must have added other "ingredients" because I think polyethylene has a one-step degradation curve (but correct me if I am wrong). I believe it starts degrading around 400 C and shows a smooth curve. For your graph, I would put the temperature on the x-axis if possible; it makes it easier to evaluate than the time. Maybe try to look at your different composite and see at what temperature they degrade, and try to match the different step to their substance. You might find that it is one of your substance that gives you the variation as the two curves often cross one another. Also, under what atmosphere are you conducting your test (air, nitrogen, oxygen...)? I'm not an expert at the different type of gas and what they can help you accomplish, but this research gate question maybe could help you figure out if you're using the right one

(https://www.researchgate.net/post/How_does_the_type_of_gas_atmosphere_in_TGA_affect_thermal_residue_detailed_and_specific_questions_included).

At what rate are you increasing your temperature? I have faced problems in the past that at a certain rate (or range of rates) that funky stuff would happen to my sample, such as weird drops that weren't present at the other rates. Maybe try a different rate and see what happens.

I hope that some of this helps you figure out what's wrong with your graphs.

Good luck,

Céleste

If it is the same film and you store at the similar temperature, you do expect a same curve. I guess your film moisture content is different or residual solvent is there or heating ramp is different. You check your sample very well or check the moisture content before loading into the instrument and also the methodology.

Yes, it was always the same film and the heating rate .

How to check the moisture content?

I have attached the results here. Every time was performed 2 similar experiments. Only sample loading way was differed.

The following experimental clues should greatly improve repeatability/reproducibility:

(i) constitute your sample from particles previously classified and characterized by size (e.g. sieved);

(ii) in case that may be found convenient, select particles from well-defined size range;

(iii) ensure comparable shape (aspect ratio) for the particles that constitute the sample by means of judiciously selecting the size reduction method;

(iv) the specific surface area of distinct samples should closely match;

(v) the sample weight should match;

(vi) pack the sample in pan/crucible by some comparable procedure, to ensure reproducible bulk density of the packed particles;

(vii) alternatively, you may embed the sample within well-defined (weighted) amount of loosely packed dry α-alumina powder (inert) ─ you may consider this option if you think it may improve thermal conduction and/or sample packing reproducibility;

(viii) while adopting last option, you may consider to calibrate the instrument by running a blank essay, also filling the sample pan with the same amount of inert, but without the sample;

(ix) if thermal analysis uses air atmosphere, its absolute humidity and mass flow rate should be maintained;

(x) thermal analysis of semicrystalline polymers is often preceded by (or includes) a previously annealing (and cooling) stage, perhaps even heating the sample up to melting (assuming that melts) ─ if adopted, reproducibility should also be ensured for such stage.

Carlos Araújo Queiroz

This will not help me at all.

Dear Anna N. Matveyeva ,

I faced some problems before with TGA curves that did not match before. From the graph you show, it seems that you've done the test twice? Maybe try doing a couple more. When I faced this kind of problems, I would do 3-5 different test to see if any curves could match up, and usually, two of them get pretty close.

I don't think it's you moisture content giving you difficulties because, for all of your pictures you've included, the first part of the curve match up well. Therefore I believe they have similar moisture content and that water is being evaporated at the same rate and same quantity. But if you would like to check the moisture content, there is moisture analyzer in which you put in a sample, and it gives you precise measurements.

What is your sample made of? I know you said polyethylene, but you must have added other "ingredients" because I think polyethylene has a one-step degradation curve (but correct me if I am wrong). I believe it starts degrading around 400 C and shows a smooth curve. For your graph, I would put the temperature on the x-axis if possible; it makes it easier to evaluate than the time. Maybe try to look at your different composite and see at what temperature they degrade, and try to match the different step to their substance. You might find that it is one of your substance that gives you the variation as the two curves often cross one another. Also, under what atmosphere are you conducting your test (air, nitrogen, oxygen...)? I'm not an expert at the different type of gas and what they can help you accomplish, but this research gate question maybe could help you figure out if you're using the right one

(https://www.researchgate.net/post/How_does_the_type_of_gas_atmosphere_in_TGA_affect_thermal_residue_detailed_and_specific_questions_included).

At what rate are you increasing your temperature? I have faced problems in the past that at a certain rate (or range of rates) that funky stuff would happen to my sample, such as weird drops that weren't present at the other rates. Maybe try a different rate and see what happens.

I hope that some of this helps you figure out what's wrong with your graphs.

Good luck,

Céleste

Hello Céleste Robert ,

Thank you so much for your detailed and thoughtful answer!

''I would do 3-5 different test to see if any curves could match up''. Yes it can work. But what type of packaging seems right? I am inclined to think that the need to pack by the layers of circles.

Yes, I thought also that this problem is not in moisture.

I don't have full information about this sample, unfortunately. This is a test sample from Japan.

This analysis was performed in the air atmosphere. In the different atmosphere will be of course an other picture. Heating rate was 5 degrees/min it should not be too fast.

Best regards,

Anna.

Hi Anna N. Matveyeva,

I am not sure about the packaging of the sample into the pan, I have never worked with films before. I would believe that as long as you have a similar weight, the packing shouldn't matter too much. But be careful to always put a similar weight, I have experience difference in the graphs because of the weight differences between samples, but usually, that is shown by a small shift in where the curve is, not a whole different curve.

If I were you, I would ask for some more info on the samples if possible. You might not be testing what you think. If you can't get more information, maybe an FTIR analysis or something like that would be beneficial to see what you samples are and to make sure that from one to another you have the same product. If your samples were different in concentration because of bad dispersion or different processing methods (or different batch), then you might see a difference in weight loss from one curve to another.

Have you performed other test on those samples? If yes, have you observed differences in results in those test too? To get curves like that when it's supposed to be polyethylene, I don't think it's very normal.

Respectfully,

Céleste

Céleste Robert ,

I have never worked with film samples before too.

I know from the manual that the film-like samples are recommended to punch with a punch or cork borer the same diameter and to stack several pieces to adjust the weight. But even this case there is bad contact between pieces. So I have tried to melt it, but anyway it was not success method.

Yeah, I agree that the weight is important.

The sample always was the same, because it is the monolite material.

Best regards,

Anna

If you judge the measured curves carefully, you detect deviations of the linear temperature increase (green line) where exothermic peaks are visible. This indicates, that there are rather strong exothermic events. As you do the TG experiment under air, the events which don't match (in the temperature region between 250 and 400 °C) obviously are oxydations. The exothermic reaction rises the temperature locally at the surface of the sample drastically which accelerates the oxydation process. Will say, the sample is never heated linearly, the temperature of the sample varies from run to run in a different way and consequently the oxydation process changes arbitrarily and the DTA signal also. This can not be avoided if you use air as purge gas. If you would use an oxygen-free purge gas (nitrogen, argon) you may get much better reproducibility results. God luck!

Günther W. H. Höhne ,

Thank you for the recommendations!

Thus you think that impossible to achieve the reproducibility in these conditions? I wanted to carry out an experiment under a lid (with a small hole) in a dynamic air atmosphere.

Best regards,

Anna