Material: Paraffin wax

Temperature: 30 to 600 degC

Heating Rate: 10 degC/min

Condintioning: N2 at 19.8 ml/min at 1.9 bar

Initial thought was the crucibles were not cleaned properly. But after doing an empty crucible heating to 900 degC, the same crucible was used for the first run the following day. Still there was considerable weight loss. This only happens in the first run. But after the first run, the following runs are more accurate irrespective of the crucible used.

In case you intend to refer to the problem of a negative percentage value found at the TGA plot (not shown), you may possibly find helpful to check my answer given to another question: https://www.researchgate.net/post/Can_TGA_show_negative_percentage_value

Carlos Araújo Queiroz Can it be a software issue? If it was a calibration issue, results would be inaccurate for all samples. As paraffin wax totally decomposes around 300 to 400 degC, the crucible can be reused as there are no leftover material. The same crucible provides -40% in the first run and during the second run, the same crucible gives accurate results with a final weight % at 0.023%. The sample used was the same for both the runs.

The most likely cause is an impurity into the crucible during a previous sample. On your first TG analysis, this impurity is removed

This type of problem often appeared when thermal equilibrium is not established between the crucibles, sensor and sample material. Switch-On all the functional spares including thermostate at least for 2h before to the start.

TARE will be very difficult under thermal fluctuation state.

Software part is OK; that is why weight loss value is correct from 2nd run and onward.

You are using paraffin wax for your research. Try to SET temperature range more than the required temperature. In this circumstances, the wax materials got vaporized (~700C) during the analysis process.

Winifred Obande Thank you for your suggestion. I will surely try it out. Narendra Yadav This is really interesting. I will try it out tomorrow. Will let you know if there is change in results. Nedelcho Nedelchev Initially, my thought was the same, but I used a clean crucible and still the results were not accurate.

Check if you start second and further runs at the same temperature as the first run. If not allowing enough cooling time between the samples, oven well can remain at 50 - 60C. That is, if your method supposes start temperature to be 30C, you should ensure it is indeed the case for all your consequitive runs (and not just for the first one). Depending on the chain length, paraffin waxes (crystalline n-paraffins) can be quite volatile. For C18 n-paraffin evaporation is quite detectable even at 40 -50C

Try to not start the heating as soon as the crucible is loaded into the furnace. Ensure that the nitrogen flow is uniform and there is no deviations in the mass loss before the heating is actually begun. The balance hangs by a very thin and sensitive platinum cable and any sort of distortions in the furnace atmosphere can affect the mass loss percentage. Also ensure that the Nitrogen you're using is pure.

Narendra Yadav As you mentioned, it seems the thermal equilibrium was not attained for the first run. The results after leaving the instrument switched ON for two hours was accurate.

Further runs will be conducted in the following days. If any changes or further challenges pops out, will let you all know. As of now the problem has been solved. Thank you all for putting in the effort solve this challenge.

It is interesting to heard that the current problem has been resolved. Thank you everyone for contributing their ideas to get approachable solution.

In my lab, we always firstly run a blank crucible to check the airflow in the channel. If the test curve does not fluctuate, then we will test the sample by the same method.