Magnetic refrigeration is a form of refrigeration that uses the magnetocaloric effect: a change in a magnetic material’s temperature or entropy when an applied magnetic field changes. By coordinating magnetization and demagnetization with heat exchange, a magnetic refrigerator absorbs heat from a cold region and releases it to a warmer one. The technique is established in cryogenics, particularly at temperatures below one kelvin, while systems operating near room temperature remain an area of materials research and engineering development.
Physical principle
Magnetic refrigeration depends on the coupling between a material’s magnetic state and its thermal behavior. In a conventional magnetocaloric material, applying a magnetic field tends to align magnetic moments, reducing the disorder associated with their orientations.
Two idealized processes explain how this produces cooling:
- Isothermal magnetization: When the material remains at a constant temperature while the field increases, its magnetic entropy decreases. To maintain that temperature, it releases heat to its surroundings.
- Adiabatic demagnetization: When the material is thermally isolated and the field decreases, its magnetic moments become less ordered. In a reversible process, total entropy remains constant, so the increase in magnetic entropy is accompanied by a reduction in the entropy of other degrees of freedom and a fall in temperature.
The reverse field changes ordinarily produce the reverse thermal responses. These descriptions apply to the conventional magnetocaloric effect. Some materials exhibit an inverse effect, in which applying a field increases entropy and can cause cooling under adiabatic conditions.
In thermodynamics, the effect is commonly characterized by the isothermal entropy change, , and the adiabatic temperature change, . For a reversible magnetic system, a Maxwell relation gives
where is specific entropy, is magnetic moment per unit mass, is magnetic field strength, and is the vacuum permeability. Other mechanical variables are assumed fixed. Consequently,
The corresponding differential expression for an adiabatic field change is
where is the specific heat capacity at constant field.
These relations connect cooling performance to the temperature dependence of magnetization and to heat capacity. They require particular care near irreversible, hysteretic transitions: applying equilibrium relations indiscriminately to measurements taken along incompatible transition paths can produce misleading entropy estimates.
Refrigeration cycles
A field change alone produces only a temperature change, not continuous refrigeration. Repeated cooling requires a cycle that transfers heat between a cold load and a warmer heat sink, together with an external supply of work.
An ideal magnetic Carnot cycle contains two isothermal and two adiabatic processes:
- The material is magnetized isothermally while rejecting heat to the hot reservoir.
- It is thermally isolated and demagnetized until it reaches the cold-reservoir temperature.
- It is demagnetized further while absorbing heat isothermally from the cold reservoir.
- It is isolated and magnetized back to the starting temperature.
Practical machines generally use arrangements that are easier to implement, including regenerative cycles. Their performance depends not only on the magnetic material but also on field strength, heat-transfer rates, cycle frequency, unwanted heat flow, and the power required to operate magnets, pumps, and moving components.
Active magnetic regeneration
Near room temperature, the temperature change produced by one practical field swing is often only a few kelvin. An active magnetic regenerator can establish a larger temperature difference between its hot and cold ends.
The regenerator contains magnetocaloric material formed into a bed of particles, plates, channels, or another structure that exchanges heat with a flowing fluid. In a typical cycle:
- Magnetization heats the solid.
- Fluid flows from the cold end toward the hot end, carrying heat toward a hot-side heat exchanger.
- Demagnetization cools the solid.
- Fluid flows back toward the cold end, enabling heat absorption from the cold-side load.
The bed acts both as the field-responsive working material and as a regenerative thermal store. Repeated operation establishes a temperature gradient along it. Layers of materials with different transition temperatures can extend the useful operating range.
Reciprocating systems move a bed or magnet back and forth; rotary systems repeatedly carry beds through high- and low-field regions. Fluid flow and magnetic-field changes must be synchronized. Although the active refrigerant is solid, these machines commonly contain moving parts and a liquid heat-transfer circuit.
Cryogenic magnetic refrigeration
Adiabatic demagnetization refrigeration uses magnetic materials to achieve very low temperatures. A typical single-stage operating sequence is:
- Precool the refrigerant using another refrigeration system.
- Apply a magnetic field while maintaining thermal contact with a heat sink.
- Open a thermal switch to isolate the refrigerant.
- Reduce the field to lower its temperature.
- Absorb heat from the experimental load, often while gradually reducing the field further to maintain a nearly constant temperature.
Paramagnetic salts are important refrigerants in this temperature range. Their relatively weak magnetic interactions allow substantial magnetic entropy to remain available at low temperatures. Magnetic ordering, residual internal fields, and heat leaks ultimately constrain the attainable temperature.
A single stage must periodically stop cooling and regenerate. Multistage arrangements can provide continuous cooling by coordinating stages that absorb heat, reject heat, or regenerate at different times.
Applications include low-temperature detectors, laboratory experiments, and astronomical instrumentation. Magnetic refrigeration is distinct from a dilution refrigerator, which relies on the thermodynamic behavior of helium-3 and helium-4 mixtures. At still lower temperatures, nuclear demagnetization exploits nuclear magnetic moments, usually after extensive precooling.
Magnetocaloric materials
Useful refrigerants must combine a substantial magnetic entropy change with suitable heat capacity, thermal transport, mechanical integrity, and chemical stability. Performance must be evaluated across the intended temperature range rather than at a single peak.
Near-room-temperature materials
Gadolinium is a widely studied benchmark because its Curie temperature is close to room temperature, approximately 294 K. Its continuous magnetic transition produces a conventional magnetocaloric effect with relatively low hysteresis. Cost, corrosion protection, and manufacturability affect its practical use.
Other major research families include:
- Lanthanum–iron–silicon compounds, including hydrogenated compositions whose transition temperatures can be adjusted.
- Manganese–iron–phosphorus-based compounds, often modified with silicon or related elements.
- Gadolinium–silicon–germanium compounds, associated with large effects involving coupled magnetic and structural changes.
- Heusler alloys and manganites, which offer composition-dependent conventional or inverse effects.
Composition, processing, and crystal structure influence transition temperature, hysteresis, corrosion resistance, and mechanical behavior.
Continuous and first-order transitions
Materials with continuous magnetic transitions generally provide smoother, more reversible responses over a comparatively broad temperature interval. Materials with first-order transitions may exhibit large entropy changes because magnetic ordering is coupled to a structural transformation.
A large peak effect does not necessarily yield the best refrigerator. First-order materials can exhibit hysteresis, incomplete transformation, and strain during cycling. These effects can dissipate work, reduce usable cooling, and damage components.
Material selection therefore considers several quantities together:
- Reversible entropy and adiabatic temperature changes.
- Useful operating-temperature range.
- Thermal conductivity and heat capacity.
- Hysteresis and required field strength.
- Mechanical and chemical durability.
- Cost, availability, and ease of fabrication.
Measures based on the height and width of an entropy-change peak are useful for screening materials, but they do not directly predict device cooling capacity or efficiency.
Historical development
In 1881, Emil Warburg reported a thermal effect associated with magnetizing iron, an early experimental observation of magnetocaloric behavior. Peter Debye in 1926 and William Giauque in 1927 independently proposed adiabatic demagnetization as a route to very low temperatures. Giauque and Duncan MacDougall demonstrated cooling by this method in 1933.
Magnetic refrigeration subsequently became an important technique in low-temperature physics. Its extension toward room temperature required materials with useful effects near ambient conditions and regenerative cycles capable of sustaining larger temperature spans.
In 1997, Vitalij Pecharsky and Karl Gschneidner Jr. reported a giant magnetocaloric effect in a gadolinium–silicon–germanium compound. This helped stimulate research into materials with coupled magnetic and structural transitions. Room-temperature demonstrators have explored different bed geometries, fluids, magnetic circuits, and operating cycles, but such demonstrations do not by themselves establish commercial competitiveness.
Performance and engineering limitations
The coefficient of performance of a refrigerator is
where is the rate of heat removal from the cold load and is total input power. A meaningful system comparison includes pumping, actuation, magnet operation where applicable, controls, and auxiliary equipment.
For reservoirs at absolute temperatures and , the reversible upper limit is
Magnetic refrigeration is subject to this same limit as other refrigeration technologies.
Major engineering constraints include:
- Magnetic-field generation. Permanent magnets avoid continuous electrical excitation but require material and carefully designed magnetic circuits. Electromagnets consume power; superconducting magnets require cryogenic support.
- Heat-transfer trade-offs. Thin channels and small particles improve thermal exchange but can increase pressure drop and pumping power.
- Regenerator losses. Axial heat conduction, fluid mixing, dead volumes, and imperfect synchronization reduce useful cooling.
- Cycle frequency. Faster cycling can increase power density, but only while the material and fluid exchange heat sufficiently quickly.
- Durability. Corrosion, particle breakdown, and transition-induced strain can degrade a bed.
- Integration. Magnetic forces, fringe fields, seals, moving assemblies, and heat exchangers affect size, cost, reliability, and noise.
The temperature span measured with little or no cooling load must be distinguished from the span maintained while delivering useful cooling power.
Applications and environmental considerations
Near-room-temperature magnetic refrigeration is investigated for refrigerators, air conditioning, process cooling, and heat pumps. Unlike vapor-compression refrigeration, its principal cooling mechanism does not require repeated evaporation and condensation of a circulating refrigerant.
This creates the possibility of avoiding volatile refrigerants with high global-warming potential. It does not make a magnetic cooling system automatically environmentally preferable. Lifecycle impacts depend on electricity consumption, magnet and refrigerant production, material supply chains, service life, and recycling. Heat-transfer fluids also require appropriate selection and containment.
Magnetic refrigeration also differs from thermoelectric cooling, which uses the Peltier effect. A magnetic refrigerator’s solid working material does not imply an entirely motionless device: most near-room-temperature regenerative designs still require fluid circulation, mechanical motion, or both.