The first millikelvin cryocooler (mKCC): Design and performance

Abstract

The design and performance of the first millikelvin cryocooler (mKCC) is presented. The mKCC is based upon a tandem Adiabatic Demagnetisation Refrigerator (ADR) that uses two single ADRs operated out of phase and connected to a common cold stage to provide continuous cooling. Development of this mKCC is part of an on-going research program to ultimately achieve sub-100 mK continuous cooling and builds upon our previous research, with each single ADR in the mKCC being a fast thermal response miniature ADR using a single crystal tungsten magnetoresistive heat switch as detailed in our 2015 publication [1]. With the mKCC operating from a 3.6 K bath temperature, the goal is to achieve continuous cooling at 250 mK (sub-100 mK is not possible without additional pre-cooling). The mKCC has dimensions of 120 × 56 × 228 mm and a mass of 4.67 kg. It can operate on very fast timescales – each superconducting magnet can be ramped to 2 Tesla in 30 s and the Chromium Potassium Alum pills have a measured sub-second thermal response, resulting in each miniature ADR being recycled in minutes. Unconventionally, the mKCC uses single crystal tungsten magnetoresistive heat switches.We present the performance of the first version of the fully automated mKCC (from a 3.6 K bath temperature), which has been determined by undertaking a range of tests analysing the cool down from 3.6 K to the operating temperature, the baseline performance, the thermal stability at the continuous stage, the reliability and repeatability in performance and the cooling power at a range of operating temperatures. The base temperature has been measured to be 750 mK and we have demonstrated that the mKCC can be operated at any temperature between 750 mK and 3 K, with the program-controlled transition between operating temperatures taking approximately 60 s. The cooling power of the mKCC (in addition to parasitic load) has been measured at a range of temperatures between 800 mK and 3 K by applying a heat load to the continuous stage via a heater; the maximum cooling power at 800 mK is 6 µW, increasing to 32 µW at 1 K and 412 µW at 3 K. In addition we conducted a six-week continuous test during which each ADR undertook 5,498 5.5 min cycles with no significant variation in performance detected. To conclude, we compare the measured performance of the mKCC to the expected performance based on mathematical thermal modelling and the performance of the miniature ADR. Whilst the measured performance does not meet the expected performance in terms of base temperature or cooling power, we have identified the limiting factors and discuss them here.