Did you perform that run in nitrogen? The only possible explanation is that the bottom of the crucible (or the sensor surface) was dirty, hence, a hi-temp run in air or oxygen may help.

Baking out in your oven below oxygen at max temperature (1100 or 1600°C depending on the oven). If this is not working: clean the balance arm in a bunsen burner.

I agree with the high-temp oxidation idea, that's probably the easiest. If you can remove the sample carrier (the thing the crucible rests on) you can try putting it in some acid, nitric maybe, for awhile and seeing if that helps. Be careful if there are thermocouples on the carrier, they could be damaged by the acid. 

I've had this happen with a Netzsch DSC/TGA before, sometimes you can get them off with these methods, sometimes you can get them off with a little elbow grease (which will bend the carrier, but it's usually still functional)...and sometimes you can't get them off at all and have to send them back to the manufacturer and be ready for a few k$ in repair bills.

Good luck!

heat at highest possible temperature (safety temp and must be grater then 1000C). if possible remove the crucible carefully at that temperature . this is the condition for pure alumina but if u have already some sample in it then it was better to call the company people.

Thank you all for your replies! I already tried high temp (1000C) under oxygen or nitrogen but nothing. Eventually I just got it out by hand trying not to bend the carrier. I believe that the acid treatment would work too. The hight temp oxidation maybe is causing more problems, since bonds can be created between alumina and the carrier which is also ceramic.  Anyway,thanks again for your time!

I really miss the TA Instruments! :)

Once, one of mine PhD students did a mistake and he tried to remove the crucible mechanically. The result was that he destroyed the arm of the balance... expencive...

Finally, we removed that by a chemical solution related to the substance that was inside the crucible (in a quantity really more superior than the one recommended.

A chemical solution of acids is also able to remove substances adsorbited in the surface of the crucible. (or nanoparticles that moved and stucked on the surface of the crucible!)

Sorry for digging up an old question.

We just experienced this as well with our PerkinElmer TGA. Our sensor ended up breaking apart, with its top part (I think platinum) sticking to the bottom of the crucible (alumina) and separating from the sensor stem. The service engineer is recommending using alumina powder to prevent this adhesion. I am very hesitant to follow his recommendation because I think it might affect the accuracy of the results by adding material between the pan and the sensor.

Do you see the sense in his recommendation? Lastly, do you know what could possibly cause the pan to stick to the sensor aside from having a dirty bottom? Thanks a lot!

@Albert Yes. Placing a material between the crucible and sensor (balance) is recommended as soon as the material has much higher melting temperature than you instrument can reach. Since I faced the problems mentioned above, I've been using tiny pieces of Ni underneath the crucible and the results are just fine (under N2 analysis). Of course, I first checked the TG curve of Ni at the temperature range of interest before trying that, just to be sure that I won't have any effects from the Ni (very unlikely). Don't use alumina powder cause it's difficult to handle. Also as it's the same material as the crucible therefore it might also stick at high temp. As about the factors that this is happening.. well as you said: dirty crucible or sensor (dirt is not always visible lol)

I just experienced it a few days ago. The machine used was the Mettler Toledo TGA / DSC 1. We dripped concentrated HCl around the crucible and left it overnight. The next day the crucible could be removed easily.

Adi Darmawan I found out that just few drops of water work quite good.

I had this problem with a TGA (Mettle Toledo TGA/DSC 1) years ago with organic samples. Once the crucible sticks to the TGA arm, the arm may be lost if chemical cleaning methods do not work; however, this is preventable.

The crucible sticks to the arm when organic samples have been repeatedly pyrolyzed. Pyrolysis methods are often presented in the literature as a single time-temperature ramp in the absence or oxygen; however, the literature rarely mentions what happens to the sample after the temperature ramp is complete. Many times with organic samples the TGA curve does not flatten completely at the end of a time-temperature ramp, which means the sample may still be smoldering. A smoldering sample releases smoke, which is mostly steam, but some tar. Tar is sticky and can condense on the arm and the crucible if temperature is set to drop at the end of the temperature ramp. Though the tar from a single pyrolyzed sample is not likely to stick a crucible, tar from multiple samples will eventually cause the crucibles to stick.

To prevent the crucibles from sticking, I always leave the last crucible in the auto-sampler empty. The empty crucible is set to run a cleaning method which ramps to 950°C under a high flow of zero grade air to volatilize and burn away any condensed tars.

If a multi-gas sampler is not available, it may be possible to change the method by adding a time-temperature hold at the end to give the pyrolysis time to complete or set a high flow rate as the sample cools to sweep away the tars before it condenses in the furnace. As long as the method is designed so nothing can condense on the crucible or the arm, then the crucible is not likely to stick.

If I am doing a proximate analysis with a TGA, I alternate between volatile and ash methods. The ash method is a combustive method which will clean any residue left behind by the pyrolytic volatile method.

Cleaning dirty crucibles in a kiln before using the TGA helps with removing or burning away any residues on the crucible which may cause it to stick.

On generic methods for acid cleaning of sintered alumina (yet for tubes); cf. my post at: https://www.researchgate.net/post/How_to_wash_the_alumina_tubeWhich_chemical_is_best_for_washing_the_alumina_tube_so_that_no_cracks_form_on_surface_of_tube

We have a Netzch TGA and we had experienced a lot of problems with the sample carrier. I would like to know if some of you have had the same bad experience with Netzsch. Thanks.

Victoria Padilla I am not using Netzsch but I or someone here could give some good advise if you could share more details about the issues you are facing.

George Pappas may I know how to you remove the stuck Alumina crucible ?

Lingaraj Ryavanaki I put one drop of water or 5 % v/v HCl aqueous solution around/outside the crucible and leave it for a couple of hours. Then carefully trying to remove the crucible.

George Pappas Thank you , I appreciate your idea. I will try to remove.

It was very easy I just used lighter (blaze) and heated up the crucible holder untill it turns to red and then removed the broken crucible easily.