Tikalon Blog by Dev Gualtieri

Tikalon Header

Magnetocaloric & Electrocaloric Effects

April 25, 2011

It's quite easy to turn electricity into heat , as your hair dryer, toaster and soon-to-be-outlawed incandescent light bulbs prove. It would be strange, indeed, to apply a voltage to a material and expect it to cool. Also strange would be to get similar cooling upon application of a magnetic field . There are some materials that demonstrate these two effects, called the electrocaloric effect and the magnetocaloric effect.

First, it's important to distinguish the electrocaloric effect from the Peltier effect . Peltier materials are effectively heat pumps that transfer heat from a heat source to a heat sink. They're useful for cooling electronic circuitry is some critical applications, but they need a way to extract the pumped heat in real time, usually by air or water cooling. The electrocaloric effect will work in the absence of a heat sink, albeit for one thermodynamic cycle . Then the material needs to be heated, and it will provide cooling again in another cycle.

The electrocaloric effect is considered to be the inverse of the pyroelectric effect that I reviewed in a previous article (Pyroelectric Energy Harvesting, October 15, 2010). Pyroelectric materials generate a temporary voltage when heated or cooled, and no temperature difference is required for this reaction. In pyroelectrics, temperature change causes a shift in relative atomic positions. This leads to a change in polarization , as when you squeeze a piezoelectric crystal .

The principle behind the electrocaloric effect is similar, since it involves differences in energy for different arrangements of atoms in crystal structures. Application of a voltage to electrocaloric materials will cause them to change phase to another crystal structure. This causes a change in entropy , and the sign of that change determines whether we'll see heating or cooling.

The electrocaloric effect wasn't very useful until about five years ago. Up to that time, the reported highest cooling was 2.5 °C upon application of 750 volts - Interesting, but certainly nothing to write home about. In March, 2006, a paper was published in Science by researchers at Cambridge and Cranfield Universities in the UK who reported a much better effect for PZT (lead-zirconate-titanate) thin films.

These were 350 nanometer thick PbZr_0.95Ti_0.05O3 films near the material's ferroelectric Curie temperature . The cooling was 7 °C for 25 volts, which is fairly good, but the effect happened at 222 °C, not at room temperature. What the authors call a "giant" electrocaloric effect had a conversion constant of 0.48 K per volt. Could greater cooling be had at higher voltage? This isn't possible because of the dielectric strength of the material.

This 2006 paper may have generated a renewed interest in this research topic, since it was followed quickly by a 2008 paper on a ferroelectric polymer with 12 °C cooling near room temperature. The Penn State materials scientists who authored this paper found coolings of about 12 °C in poly(vinylidene fluoride-trifluoroethylene), a ferroelectric polymer. This was observed at temperatures above 70 °C, which is the ferroelectric-paraelectric transition temperature of this material. This cooling was a consequence of an isothermal entropy change of about 55 J /kg/K.

The magnetocaloric effect has been known for more than a hundred years. This effect was discovered by the German physicist, Emil Warburg , who is known principally as the father of Otto Heinrich Warburg , who won the 1931 Nobel Prize in Physiology or Medicine ; and as the "academic father" of James Franck (1925 Nobel physics prize ) and Hans von Euler-Chelpin (1929 Nobel chemistry prize ).

In 1880, Warburg discovered a minuscule effect in pure iron of the order of 0.5 - 2 K/T , which is just 5 - 20 x 10^-5 °C/gauss. Not surprisingly, Thomas Edison independently rediscovered this effect in 1892, but he wasn't able to find a practical use for it. William Giauque (1949 Nobel Prize in Chemistry) and D.P. MacDougall put the magnetocaloric effect to good use in 1933 by demonstrating cryogenic cooling to 0.25 K.[4]

Figure caption

Emil Gabriel Warburg (1846-1931).

Photograph via Wikimedia Commons

The mechanism of the magnetocaloric effect is easy to understand. A magnetic material with electron spins aligned is in a lower entropy state than one with a random alignment of spins. As applied to magnetic refrigeration , the essential thermodynamic cycle is to magnetize the material, which heats it. The heat is removed by contact with a cooling medium while the magnetic field is still present. The magnetic field is then slowly (i.e., adiabatically ) decreased, and this cools the material, since thermal entropy is converted to spin entropy. The cooled material is then placed in contact with whatever environment you want cooled.

It was only in 1997 that this effect started to appear useful when Karl A. Gschneidner, Jr. demonstrated a proof of concept magnetic refrigerator operating at room temperature.[5] Gschneidner is a world expert in rare earth materials at Iowa State University and Ames Laboratory , and rare earth alloys are excellent magnetocaloric materials at room temperature, as you can see from the data in the following table that lists the Curie temperature (T_C), a magnetic field change (ΔH), and the entropy associated with the field change (ΔS_M).[6]

Magnetic Material	T_C(K)	ΔH(T)	ΔS_M(Jkg^-1K^-1)
Gd	294	5.0	10.2
Gd_0.5Dy_0.5	230	5.0	10.2
Gd_0.74Tb_0.26	280	5.0	11.5
Gd₇Pd₃	323	5.0
Gd₅(Si_xGe_1-x)₄ at x=0.43	247	5.0	39.0
Gd₅(Si_1.985Ge_1.985Ga_0.03)₂	290	5.0
Ni_52.6Mn_23.1Ga_24.5	300	5.0	18.0
MnAs_0.9Sb_0.1	286	5.0	30.0
MnFeP_0.45As_0.35	300	5.0	18.0
Gd	294	1.5	3.8
Gd	294	6.0	11.4
La_1-xCa_xMn0₃ at x=0.33	267	3.0	6.4
La_0.9K_0.1Mn_0.3	283	1.5	1.5
La_0.75Ca_0.15Sr_0.1Mn_0.3	327	1.5	2.8
La_0.677(Ca,Pb)_0.333Mn_0.3	296	7.0	7.5

At the Curie temperature , these materials undergo a second-order paramagnetic -ferromagnetic phase transition that enables a large magnetocaloric cooling. The magnetocaloric effect is maximum near the Curie temperature. With the current problem of rare earth supply, it's nice that there are magnetocaloric materials that don't require a rare earth component. Of course, the real problem is generating the high magnetic fields. You can generate about two Tesla with permanent magnets , but the best of these are made from the rare earths.

References: