|
Magnetic refrigeration is based on a fundamental thermodynamic property of magnetic materials: the so-called magnetocaloric effect, which causes a temperature change if the material is subject to an applied magnetic field under adiabatic conditions. The magnetocaloric effect was discovered in 1881 in iron by the German physicist Emil Warburg. Usually the temperature increases when the field is applied (and decreases when the field is removed) and the process is reversible.
The magnetocaloric effect can qualitatively be understood as an interaction between the entropy (which is a measure of the disorder) associated with the magnetic moments of the atoms (spins) in a material and the entropy associated with the heat motion of the same atoms. If an external magnetic field is applied it will tends to order the spins, thus decreasing the magnetic entropy. If the material is isolated from its surroundings (i.e. its entropy is constant), the decrease in magnetic entropy must be compensated by an increase of the entropy of the heat motion and therefore an increase in temperature. The magnetocaloric effect is most pronounced in the vicinity of a magnetic phase transition of the material, e.g. from a non-ordered (paramagnetic) to a ferromagnetic state.
A magnetocaloric material can be used as the active element in a refrigeration apparatus. The apparatus can for instance be operated in a four step cycle:
1) The magnetocaloric material is magnetized by a magnet and the temperature increases.
2) The material cools by giving off heat to the surroundings through a heat exchanger.
3) The magnetic field is removed and the temperature of the material drops further.
4) The material takes up heat from the cold side heat exchanger ("the inside of the refrigerator") thus cooling it. Then the cycle starts over at step 1).
Such a magnetic refrigerator has a number of advantages compared to conventional refrigerators, e.g. environmentally hazardous refrigeration gasses such as HFC (hydroflourocarbons) or ammonia are avoided, and higher efficiencies are possible.
It has been recognized since the 1920s that magnetocaloric materials can be used for cooling purposes. In the laboratory magnetic refrigeration using paramagnetic salts is a standard technique for obtaining ultra low temperatures. However, for the purpose of using magnetic refrigeration near room temperature several problems arise:
- magnetocaloric materials are only active in a certain, material-specific temperature range significantly limiting the temperature range where the refrigerator should function
- the temperature change induced by the active material is only of the order of a few degrees, which is too small for practical purposes
To overcome these obstacles, the simple cycle illustrated above must be modified by the use of a regenerator. By having the magnetocaloric material act as a regenerator on the cooling fluid, the device can span a wider temperature range. This concept, called an active magnetic regenerator (AMR), was first introduced by J.A. Barclay and W.A. Steyert in the early 1980s. Work in particular by Ames Laboratory (University of Iowa) and Astronautics Corporation in the USA have elaborated on this idea, resulting in a number of prototype magnetic refrigerators using metals such as pure gadolinium and gadolinium alloys as the active materials. Combining these materials with high applied magnetic fields (applied by superconducting magnet coils), temperature spans in excess of 40 degrees may be reached.
The work at Risø focuses on the use of ceramics as the active materials and permanent magnets to generate the required magnetic fields. Ceramics are very stable at room temperature, can be compositionally tuned and do not corrode in water. The Fuel Cells and Solid State Chemistry Department has extensive experience in the synthesis and shaping of advanced ceramics.
Furthermore the department is putting a lot of work into developing computer models of the active magnetic regenerator and the design of the permanent magnets. This is done for several reasons, which include developing a firm platform for gaining knowledge about how an optimal regenerator should be designed as well as the more basic and fundamental knowledge about physical behavior of such systems. |