Nuclear refrigeration and thermometry at microkelvin temperatures

Abstract

The present status and problems of refrigeration and thermometry at micro. kelvin temperatures will be discussed. It will be shown that a better understanding of internal, time dependent heat leaks and of thermal boundary resistances as well as further progress in thermometry are necessary to reduce the present minimum temperature of about IO I~K to which matter has been refrigerated. 1. PRESENT STATUS OF NUCLEAR REFRIGERATION The ambition of physicists to investigate m~ttter under extreme conditions and to look for new phenomena under these conditions has led to considerable progress in the techniques of obtaining ultralow temperatures. In the millikelvin temperature range refrigeration by the continuous 3I-Ie-4He dilution process and thermometry by various techniques are well advanced and thoroughly understood. 1-3 For the microkelvin temperature range the only known method for refrigeration is adiabatic demagnetization of nuclear magnetic moments and the only thermometric technique applied at microkelvin temperatures is nuclear magnetic resonance on 195pt.l-s Even though both nuclear adiabatic demagnetization refrigeration as well as nuclear magnetic resonance thermometry have been substantially advanced during the last two decades, 3-1~ they have not been developed to their full potential and the experimentalist still encounters a number of difficulties in applying them. The schematic of the low temperature part of a typical nuclear refrigerator is shown in Fig. 1. The nuclear refrigeration stage, usually 10+ 100 moles of Cu are exposed to a magnetic field of typically 8 T and precooled by a 3He-4He dilution refrigerator to a starting temperature of about 10ink for the adiabatic demagnetization. The thermal contact between the dilution refrigerator and the Cu nuclear stage is provided by a superconducting heat switch. The starting conditions for the demagnetization are determined by the cooling power of the SHe-4He dilution