For this temperature range, you will indeed need IR thermomether, but many low clas thermometers measure the temperature of large area 10 cm at 20 cm distance and 40 cm round area at 1 meter, so if the heated substrate is inclined it would be more diffcult. You will have to check the thermometer wavelenghts and then check the transmitance of the glass on the viewpoint. For higher temperatures you can check with almost every camera the spectral distribution in the visible range.

Wireless might be an option but I would NOT test battery in the vacum chamber. 

If you have spare wires coming out of the chamber, you can use thermocouple. As the usual lab thermocouples are encapsulated, I suggest home made.  We used homemade thermocouples in vacuum thermal deposition mashine and they worked fine.

When measuring the temperature of glass with an infrared thermometer, both reflectance and transmittance must be considered. By carefully selecting the wavelength, it is possible to measure temperature of both the surface and at a depth of your vacuum chamber. When taking measurements below the surface, a sensor for 1.0, 2.2, or 3.9 μm wavelength need to be chosen. You can try with the fluke non contact type thermometers where the above conditions will satisfy.  

Dear Fredrick,

Thanks for your answer. However, I am not trying to measure the temperature of glass. Rather I have to measure the temperature of a Si wafer over which I am depositing a-Si layer by CVD process inside a vacuum chamber. Somewhere and sometime ago I have seen some researchers measuring the temperature of their samples inside a vacuum chamber through a viewing port and by using an IR thermometer kept outside the chamber. They have similar KF-50 flange. But I am sure the window material was not glass. Any comment about the right window material (could be sapphire or something else) would be appreciated.

Thanks

You need to find a material that is compatible with the wavelength your IR sensor uses. Sapphire has 80% transmittance from 300-5000 nm. If you are using a longer wavelength detector (probably not, given your temperature of interest) you will need to source a different window material. The company supplying the viewports should have the spectral ranges for the different materials.

Thanks

Zinc selenide (ZnSe), or Zinc sulfide (ZnS) looks like promising candidates as my temperature of interest is 50 Degrees Celsius to 300 degrees Celsius. But are these windows transparent in optical range? 

u can try to sense the temperature with  a sensor and transmit the information through wireless link to a receiver in air

For this temperature range, you will indeed need IR thermomether, but many low clas thermometers measure the temperature of large area 10 cm at 20 cm distance and 40 cm round area at 1 meter, so if the heated substrate is inclined it would be more diffcult. You will have to check the thermometer wavelenghts and then check the transmitance of the glass on the viewpoint. For higher temperatures you can check with almost every camera the spectral distribution in the visible range.

Wireless might be an option but I would NOT test battery in the vacum chamber. 

If you have spare wires coming out of the chamber, you can use thermocouple. As the usual lab thermocouples are encapsulated, I suggest home made.  We used homemade thermocouples in vacuum thermal deposition mashine and they worked fine.

I am not thinking about using thermocouple as my Si wafer is kept on a moving carrier and it would be cumbersome to make a thermocouple contact over a moving substrate inside a vacuum chamber. 

We have a furnace which is used for High Temperature (around 2000 C). This furnace's inside temperature is measured by Pyrometer. There is an ZnS glass material just after this equipment towards to the inner part of the furnace. ZnS is transparent in both necessary IR wavelength and visual spectrum.

As this is interesting question/solution, I checked what is common in internet. There are several manufactuererrs of "IR thermocouple" ( infra red sensor, that emulates thermocouple). There are 2 and 4 wire sensors. In the data sheet of one manufacturer, it was written, that customer had performed tests in vacuum and strong microwave field, but high pressure and air/liquid coolingnis also possible.

Just search the internet. It is intresting solution.

Another concern with optical pyrometry is calibration. The emissivity of your surface will most likely change as you deposit material. A two-color ratio pyrometer can cancel out most of the effect of emissivity changes [Khan, M. A., Allemand, C. & Eagar, T. W. Noncontact temperature measurement. II. Least squares based techniques. Rev. Sci. Instrum. 62, 403–409 (1991).], although you still need a way to determine an initial reference temperature. In my lab we use the melting point of the sample, but that probably isn't a solution for you. Perhaps you can aim the pyrometer at an area in good thermal contact with the sample but not being deposited on? Then you could use a temporary thermocouple set-up to calibrate the pyrometer ahead of time.

I once used a disappearing filament optical pyrometer to measure the temperature of Hot Ti electrode over which I was melting gold.I was able to measure the temperature pretty closely with the disappearing filament optical pyrometer. But in that case The temperature was more than M.P of gold (T> 1064 C). At that temperature objects glow and based on the color and emissivity you can determine the temperature of the hot object. For my current application my temperature of interest is 50 C to 300 C. At that temperature optical pyrometer is not an option. Also I can't use low melting standards like Indium inside the vacuum chamber as it will contaminate the chamber.