The Dilution Cryostate

We are using basically the same NMR electronics in the dilution cryostate as we are using in the flow cryostate. The main difference to the setup in the flow cryostate is the external pulse programmer, an RS-670 40 MHz Digital Word Generator and the external data acquisition in a Le Croy dual 300MHz oscilloscope. These devices have the advantage of being very accurate and easy to handle and the disadvantage of being very expensive and having quite a slow data transfer rate to the computer. We will change this setup to one similar to the flow cryostate soon. For that reason I don't want to go into technical details. The only thing I'd like to mention is the cryostate itself.

For all measurements below 20K we were using a TLM 400 dilution refrigerator. For a detailed description of the principles of a dilution refrigerator please see e.g. [#!lowtempbook!#]. A sketch is given in figure . The sample is placed in the glass tail (GT) in direct contact with the ³He-⁴He-mixture. This setup is advantageous because of its very good heat contact. With this setup the energy deposited in the sample after an rf-pulse is removed quickly even at low temperatures. Nevertheless the time to achieve equilibrium after a pulse is increasing with decreasing temperature such that there is a low temperature limit of around 35mK. The setup also enables us to use the dilution refrigerator like a flow refrigerator. Bypassing the booster pump and using only the mechanical pump for the circulation one can go up to temperatures of about 20K. Then it is no more the two phases in the mixing chamber that are cooling but a flow of gaseous He who forms the heat contact of the sample with the 1K pot. We usually even increased the pressure in the 1K pot. Having all the mixture in the gaseous phase required us to store 5/6 of it in the dumbs in order to reduce the pressure in the closed cycle. This procedure enables us to cover all the temperature range from 35mK to 20K with the dilution refrigerator, only having a little gap at around 2K where we have to take off the mixture. The direct contact and the fact that the cryostate is top-loading also enabled us to change the tank circuit (involves unmounting and remounting the probe) within one day.

  

Figure: The dilution refrigerator. It consists in a closed ³He-⁴He-cycle and a 1K pot for precooling the mixture. In regulating the pressure in the 1K pot with (PR) we can stabilize its temperature between 4.2K and 1.6K. The precooled mixture condenses in (C) and finally ends up in (MC)/(GT). The mixing chamber is pumped through the still (S) which can be heated a bit to increase the pressure locally. The pumping is performed either with a booster pump in combination with a mechanical pump or with only a mechanical pump. Image source: [#!juerg!#], p. 33

For the temperature control we used four RuO₂ thick film resistors for their low magnetoresistance. Two were placed at the probe head and two at the teflon tube in which the probe head is inserted. In each pair one of them had a room temperature resistance of 1k and the other of 100k. The 1k was used to control the range below 4K, the 100k was used for all temperatures above. To bring the mixture to the desired temperature we heated it in the teflon tube; the heater was regulated with the resistors at the teflon tube. All the resistance measurements -- also in the flow cryostate -- were 4-point measurements.