Walther Meißner Institut -

For most LT experiments resistance thermometers (RT) are used. Different types of thermometers and suitable electronics are commercially available. RTs are convenient to use as they are small and can easily be attached to experimental gadgets. On the other hand, reproducibility problems, thermal contact problems, electrical grounding problems and internal heating problems arise and become more critical the lower the temperature to be measured is. The magnetoresistance of RTs has to be taken into account, also. Usually, resistance thermometry becomes quite tricky below 20 mK.

At the WMI in most cases homemade RTs are utilized for ULT purposes. Certain commercial RuO chip resistors make excellent LT thermometers. In Fig.1 we show an example of a thermometer calibration curve; the exponential fit function (which is reminiscent of Mott’s theory) extends over a temperature range of 3 orders of magnitude. Several types of LT thermometers are depicted in Fig. 2.

To calibrate RTs and for general ULT thermometry, a ³He melting curve thermometer is available. The pressure of the solid-liquid phase transition of ³He is strongly temperature dependent due to the very different physical properties of liquid and solid ³He, and so this transition has been used for many years for ULT thermometry. In fact, a new low temperature scale for a temperature range from 1 K to 1 mK was established (PLTS-2000, Phys. Tech. Bundesanstalt, Berlin) which converts melting pressures measured to temperatures. In the photograph (Fig. 3) parts of a ³He pressure cell are shown. The cell body (bottom of Fig. 2) is from silver, and a silver sponge within the cell provides the thermal contact of the solid-liquid ³He sample with the cell. The top plate of the cell center of photo) consists of a membrane, so the pressure in the cell can be monitored by measuring the bulge of the membrane with a capacitor (top of photo) attached to the membrane.

Several fixed points of the ³He melting curve are important for the pressure-capacitance calibration. Sensitivity and reproducibility of the thermometer are excellent, ease of use is not.

To measure our lowest temperatures, especially those below 1 mK, we use pulsed NMR thermometry. With this thermometer use is made of the Curie-law behavior of the nuclear magnetization of certain metals; most often a sample of pure platinum wires or platinum powder immersed in liquid ³He is utilized. Conduction electrons provide the thermal contact between the nuclear spin system and the sample lattice system. Pulsed Pt NMR thermometers are usable for temperatures as low as 10 µK, possibly lower. The pulsed NMR thermometer has become the standard thermometer for temperatures below 1 mK. Although there is a commercial model available, we use our own electronics and Pt samples prepared in our lab.

Just recently a commercial CMN (cerium magnesium nitrate)-thermometer was tested at mK temperatures. CMN is a paramagntic salt whose susceptibility follows a Curie-law behavior over a wide temperature range; therefore it is frequently used for ULT thermometry. In Fig. 4, the magnetization of our thermometer is plotted versus T-1. For reliable thermometry we used a superconducting fixed-point-device which was purchased from NIST (National Institute for Standards and Technology). It contains 5 metal samples which undergo a transition to superconductivity at different temperatures (AuIn2, AuAl2, Ir, Be, W) in a temperature range of 200 mK to 15 mK. The transition temperatures given by NIST were checked and "fine-tuned" with the 3He melting curve thermometer described above where we applied the PLTS-2000 temperature scale.

ULT Thermometry