A common way to measure temperature on the scale of household objects and cooking is to use an infrared thermometer. The basic idea is to look at the thermal emission of infrared light from an object which is related to the temperature (like blackbody radiation).

I am interested in ways to use non-contact thermometers at low temperatures (below 50 Kelvin). Blackbody radiation quickly becomes very weak in total energy as you cool down, so I don't think that would work. My question is, are there non-contact thermometers that work at cryogenic temperatures (<50 Kelvin)?

For example, I know Raman spectroscopy can be used to measure temperature by looking at frequency shifts and the ratio between energy absorption and loss events (which scales as $e^{E/k_B T}$). I can't find any examples of this technique for low-temperature thermometry though.

As a note, I am not necessarily looking for a cheap or convenient way to do this measurement. I am more interested to know if such devices exist and how they work physically.

enter image description here

  • $\begingroup$ Is there a good reason to suspect that the device will be significantly out of thermal equilibrium with its immediate surroundings? $\endgroup$ – catalogue_number Jun 22 at 14:22
  • $\begingroup$ @catalogue_number, good question: yes. It is quite technical and unrelated to the question, so I won't dwell on that though. i suspect that in the end a diode or thermocouple will have to be used practically speaking, but the question still stands about non-contact thermometry at low temperature $\endgroup$ – KF Gauss Jun 22 at 20:24
  • $\begingroup$ It's a lot of pain to build a combined cryo-laser system. Better ask plainly your question (what is the specific set-up, what are the possible configurations), not to get lost in good answers to your implicit question, that may not be relevant to your actual case. Depending on the actual situation, you can place calibrated resistors in adjecency to the object (what is the object? is it a lithographic sample?), with substrate with decent heat conductivity $\endgroup$ – Alexander Jun 24 at 12:05
  • $\begingroup$ @Alexander, I decided to divorce the fundamental question with the application in mind. I know in reality the best route in practice is to integrate a thermocouple and that will eventually happen in our case. But I am interested to know if a "cryo-laser system" (whatever that means to you) actually exists and the physical principles involved. I've edited the question to emphasize that $\endgroup$ – KF Gauss Jun 24 at 12:29

I suppose that there are two fundamental limiting factors - the very low emitted blackbody power at low temperatures (resulting in an extremely poor signal-to-noise ratio if an approach similar to the infrared blackbody curve fitting that room-temperature thermometers use were tried) and the troublesome intermediate-temperature regime in which the radiation occurs.

The total scattered power from a blackbody at 50K is of the order of 1 µW/cm^2 in primarily radio wavelengths. The visible and infrared components of this would be extraordinarily small contributions to background intensity, and the approximation is questionable in cold metallic material.

Raman spectroscopy is the answer. One may exploit the asymmetric temperature dependence of Stokes and anti-Stokes phonon scattering intensities to infer temperature: $$ \frac{I_\text{Stokes}}{I_\text{Anti-Stokes}} \propto \exp \left(\frac{hc\bar{\nu}}{kT}\right) $$ Optical methods for measuring low temperatures have been published in this open access paper.

| cite | improve this answer | |
  • $\begingroup$ Thanks for the reference paper. One question: it appears the error bars on the Raman method are quite large. Is there a deeper reason for that? $\endgroup$ – KF Gauss Jun 26 at 0:24
  • $\begingroup$ To me, the temperatures in fig. 4c) seems to have fairly good error bars (brown boxes)- it looks like it depends mainly on choice of mode, but I'm by no means an expert in the technique. The paper suggested that accuracy of the intrinsic mechanical linewidth (with ~0.1% uncertainty, but order 1% long-term drift) was the limiting factor. $\endgroup$ – catalogue_number Jun 26 at 2:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.