J. A. Barclay

Materials for magnetic refrigeration between 2 K and 20 K

Abstract As part of a research and development programme on magnetic refrigeration, paramagnetic materials suitable for magnetic refrigeration in the 2 to 20 K range have been studied. The bestgadolinium materials with known properties, as derived from a literature survey, are plotted. For one of those materials, gadolinium gallium garnet, we present the results of our field-dependent heat capacity measurements. Preparation and fabrication of gadolinium compounds are discussed briefly as well as the scope for further work for selecting and studying working materials for magnetic refrigeration.

A reciprocating magnetic refrigerator for 2–4 K operation: Initial results

The basic theory and design of a reciprocating magnetic refrigerator to pump heat from 2.2 to 4.2 K is presented. The results of initial experiments are shown. These results include conduction losses, eddy current losses, frictional losses, and mixing losses. Two cooling cycles were attempted and a net cooling power of 52 mW was observed at 1/60 Hz. The key problems in this design are identified and discussed.

Use of a ferrofluid as the heat‐exchange fluid in a magnetic refrigerator

The use of a ferrofluid is proposed as the heat‐exchange fluid in a wheel‐type magnetic refrigerator in order to avoid flow‐control problems. An equivalent‐circuit analysis of the ferrofluid flow path with several different magnetic field profiles indicates that it is possible to obtain the desired flow control in at least one case. Sample design calculations for the revised wheel‐type refrigerator are presented. In addition, the results of heat‐transfer measurements from kerosene to a stainless‐steel screen and from a kerosene‐base ferrofluid to the same screen in and out of an 8‐T magnetic field are described.

Low hysteresis materials for magnetic refrigeration: Gd1−xErxAl2

The series of ferromagnetic intermetallics Gd1−x ErxAl2 are candidate working materials for magnetic refrigeration because of their large moment, small anisotropy, and their wide range of Curie temperatures (14 to 164 K). We have measured the magnitude of the magnetic hysteresis as a function of temperature from 4 K to TC for polycrystalline samples of these materials by ac magnetic susceptibility and magnetization hysteresis loops. We found, as expected, virtually no hysteresis in GdAl2, and, more surprisingly, less than 3000 A/m hysteresis in ErAl2 and 2×104 A/m in Gd0.5 Er0.5Al2. Moreover, the hysteresis in Gd0.5 Er0.5Al2 was <100 A/m for T−TC< 20 K.

A 4 K to 20 K rotational-cooling magnetic refrigerator capable of 1mW to > 1W operation

Abstract The low-temperature, magnetic entropy of certain single-crystal paramagnetic materials, eg DyPO4 changes dramatically as the crystal rotates in a magnetic field. A new magnetic refrigerator design based on the anisotropic nature of such materials is presented. The key advantages of the rotational-cooling concept are: a single, rotary motion is required; magnetic field shaping is not a problem because the entire working material is in a constant field; and the size of the refrigerator is smaller than other comparable magnetic refrigerators because the working material is entirely inside the magnet at all times. The main disadvantage of the rotational-cooling concept is that small-dimension single crystals are required.

Specific heat of GdRh

GdRh, a ferromagnetic intermetallic, is of interest as a working material for magnetic refrigeration. The requirements for such materials have been discussed by Barclay et al. and it is in this context that we analyze our data. Magnetic measurements of Buschow et al. have indicated TC =24 K while specific heat measurements of Olijhoek et al. have shown a maximum at 20 K and have yielded a magnetic entropy of 16.7 J/mole K. No Debye temperature has been reported previously. We have measured the specific heat from 1.8 to 77 K and our data show a peak at 19.93 K that is 17% higher than in the previous work. Fitting of the data from 40 to 77 K to C=A/T2+D(T/θ{T}), where A and θ{T} are adjustable parameters and D is the Debye function, yields a Debye temperature of 198±2 K. For the magnetic entropy we obtain 17.4 J/K, very close to R ln 8.

Low temperature thermal conductivity of some composite materials

Abstract Thermal conductivity data are presented for several samples: Gd 2 ( SO 4 ) 3 8 H 2 O pressed powders with and without copper fibres, pressed powders of NaCl and KCl, cloth base phenolic sheet, epoxy resin-bonded glass fibre laminate, and AGOT graphite in the presence of helium gas. Measurements on all samples were made in the 1–4 K range and on the phenolic sheet and glass fibre laminate at 70 K as well.

Heat capacity of Gd0.06Er0.94Al2 in magnetic fields

We have measured the specific heat of an arc‐melted specimen of Gd0.06Er0.94Al2 in applied magnetic fields of 1, 3, 5, and 8 T from 4.5 to 40 K, and in zero field from 1.4 to 98 K. Inspection of entropy versus temperature with the lattice included shows that the maximum adiabatic cooling is 13 K and occurs from 30 K at 8 T to 17 K at zero field. With our previous magnetization measurements on Gd0.5Er0.5Al2, our data indicate that substitution of Gd for Er in this crystalline environment does not significantly increase the magnetocrystalline anisotropy. The solid solution is magnetically soft and most of the magnetic entropy changes occur below 30 K.

Low temperature conductivity of Gd2(SO4)38H2O and Dy2Ti2O7 as a function of magnetic field

Abstract The thermal conductivities of pressed powders of Gd 2 (SO 4 ) 3 8H 2 O and Dy 2 Ti 2 O 7 and a single crystal of Gd 2 (SO 4 ) 3 8H 2 O were measured in the range 3 to 9 K. At approximately 4 K the thermal conductivities were measured as a function of magnetic field from 0 to 1.8 T. The single crystal had better thermal conductivity than the pressed powders and no significant field dependence was detected in any sample.

Measurement of the adiabatic temperature change in ferromagnetic intermetallic compounds (abstract)

Study of the thermodynamics of magnetic refrigeration cycles that use the magnetocaloric effect in ferromagnetic materials requires, as a minimum, knowledge of the zero‐field heat capacity C0 and the equation of state of the magnetization, M= f(B,T), for each material. However, refrigerator design calculations are easier to make if the field‐dependent heat capacity CB and adiabatic temperature change ΔTS are known. The field‐dependent heat capacity can either be directly measured or be derived according to CB(B,T) =CB(B=0,T) + ∫B0  (∂2M/∂l2)dB. The adiabatic temperature change can be obtained in three ways: by direct measurement, from field and temperature dependent entropy curves, or from dT=−T ∫B0  (1/CB) (∂M/∂T)dB. The time required for measurement, the complexity of the apparatus, and the quality of the resultant data differ considerably among the above methods. We have measured the magnetization, heat capacity, and adiabatic temperature change for several ferromagnetic intermetallic gadolinium compou...