If you have optical access to the sample, you can use a pyrometer.. Measurements at 350C should be possible but it's not a solution at temperatures close to RT.

If you have a rotating sample, it's difficult to have a thermocouple directly touching the sample backside. But you could think of some kind of spring-loaded system that pushed the TC against the sample backside if the sample is clamped sufficiently well. It may need some readjustment from time to time and it may also break if the mechanical forces become too big during rotation, so the system should be designed for easy exchange of the TC.

You can have a TC somewhere in the non-rotating part of the holder. In that case, the measured temperature will surely not be the same than on the sample surface although there may be some correlation. In that case, you can calibrate the surface temperature against the TC reading. You can use melting points of e.g. metals (In, Sn, ...) to do that but there are also special wafers/samples for this purpose.

In any case, you should make sure that the TC is not directly heated by the lamp. Probably the biggest errors between TC reading and true temperature can come from direct heating of the TC. If the TC sees the lamp, you should think of adding some heat shielding. As I understand your description, the TC is heated by the lamp. This is not ideal since the TC temperature will depend on heat conductivity (weak for thin wires) as well as heat capacity. The TC may behave very differently than the sample chuck, so you may not be able to get a good reading. If you want to work at 350C, I think that it should be better to have the TC be heated by the sample chuck (e.g. in contact or even by radiative heating), not by the lamp. The TC temperature will probably not be exactly the same than the sample surface but there should be a clear correlation,which could be calibrated if you care.

hi Mikhail,

I do not trust thermocouple measurements in vacuum if they do not have robust contact with the substrate. And if you have only radiation from a lamp (and in vacuum almost no conduction and no convection ) it might be that the thermocouple shows a big influence from that radiation. The simplest solution is to buy temprature stickers/strips. These you have in different temperature ranges. They have an adhesive and you can put them on flat substrates like Si, glass, etc and they are quite accurate to determine temperatures of a substrate in vacuum. A sticker contains several temperature indicators and if a certain temperature is reached it turns grey/black. If you want them to protect against the plasma you can put PI-tape over it.

So this is a very good method, simple and cheap and you don't need complicated vacuum feedthroughs. Only disadvantage is that you can only use them once. But for characterising a process is works great.

kind regards,

Peter

Thank you very much for your answer, Peter. We definitely should try the stickers however we need some feedback from the temperature measuring unit so we can maintain the temperature at the same value. May you say something about thermistors?

Hi Mikhail,

because of no difference in raman spectra you have a doubt about the temperature ? right?? do you expect a change in crystalinity or structure with the elevated (350C) temperature...in-situ mesurement of temperature using a pyrometer probably helpful...

best regards

dillip mishra

Hi Mikhail,

Thermistors might work just as good as thermocouple as long as they are in intimate contact with your product. A thermocouple placed in a baseplate or chuck in the vacuum will not give a good indication of the temperature of the substrate due to bad heat conduction (unless you physically clamp the substrate to this chuck).

best regards,

Peter