Elastocaloric Effect
January 6, 2020
Household refrigerators, and most other refrigerators, use the
expansion of a
gas such as
freon (1,1,1,2-Tetrafluoroethane, R-134a) to cause cooling by the well-known
thermodynamic process of
free expansion. Freon R-134a has the low
boiling point of −26.3
°C and a
latent heat of vaporization at its boiling Point of 51.9
kcal/
kg. This is nearly as large as the
heat of fusion of
water at its
freezing point, 79.7 kcal/kg. In the
1990s, when I was doing
research on the
magnetic and
hydrogen storage properties of
alloys of the
rare earth elements with other
metals, some of my
colleagues were doing research on using these materials for a different type of refrigeration,
magnetic refrigeration.
Magnetic refrigeration is based on the
magnetocaloric effect which was discovered in 1881 by
German physicist,
Emil Warburg. Warburg found that
iron subjected to an
applied magnetic field of 10,000
gauss would cool about a
degree Celsius. For iron, the effect is small, considering that the
Earth's magnetic field is about half a gauss. The
thermodynamics of the
refrigeration cycle are shown in the figure, and it's essentially a way of using the
entropy associated with the alignment of the magnetic moments of the atoms in a solid as a
heat pump.
The thermodynamic cycle of a magnetic refrigerator.
The magnetocaloric effect uses entropy as a means of extracting heat from a substance. The applied magnetic field aligns the magnetic moments of the atoms in a solid, and this results in a temperature change.
Magnetic refrigeration was first used as a means of cooling small volumes to temperatures near absolute zero, a method developed by chemist and Nobel laureate, William Giauque (1895-1982).
(Click for larger image.)
The
efficiency of a magnetic refrigerator is dependent on the
magnetic material, and there are many materials that demonstrate a better effect than iron. The magnetocaloric effect is greatest near the vicinity of a magnetic
phase transition, such as the
Curie temperature at which the
ferromagnetic order is lost and entropy becomes large. Some alloys of
gadolinium have a Curie temperature near
room temperature, and they exhibit a "giant magnetocaloric effect." Here's a table of properties of some of these materials, as taken from a
previous article (Magnetocaloric & Electrocaloric Effects, April 25, 2011).[2] In this table, T
C is the Curie temperature in
kelvin, and ΔS
M is the
transition entropy in
Joule-kg
-1K
-1.
Magnetic Material |
TC(K) |
ΔSM(Jkg-1K-1) |
Gd |
294 |
10.2 |
Gd0.5Dy0.5 |
230 |
10.2 |
Gd0.74Tb0.26 |
280 |
11.5 |
Gd7Pd3 |
323 |
|
Gd5(SixGe1-x)4 at x=0.43 |
247 |
39.0 |
Gd5(Si1.985Ge1.985Ga0.03)2 |
290 |
|
Ni52.6Mn23.1Ga24.5 |
300 |
18.0 |
MnAs0.9Sb0.1 |
286 |
30 |
MnFeP0.45As0.35 |
300 |
18.0 |
La1-xCaxMnO3 at x=0.33 |
267 |
6.4 |
La0.9K0.1Mn0.3 |
283 |
1.5 |
La0.75Ca0.15Sr0.1Mn0.3 |
327 |
2.8 |
La0.677(Ca,Pb)0.333Mn0.3 |
296 |
7.5 |
The following figure shows the precipitous drop in
magnetization of gadolinium at its Curie temperature of 294 K, which is quite close to
room temperature. There's a related large change in entropy over a small range of temperature below the Curie temperature.
Ideal magnetization curve of gadolinium as a function of temperature.
The function shown gives the temperature dependence of magnetization of a ferromagnetic material according the the mean-field approximation.
(Created using Gnumeric.)
There's another effect that's the
electric analog of the magnetocaloric effect. In the
electrocaloric effect, materials show a
reversible temperature change in response to an
applied electric field, just as magnetic materials have a reversible temperature change in response to a magnetic field. There's a change of the system entropy as the
electric dipoles in the material align themselves with the applied field. The
piezoelectric material,
lead zirconate titanate (PZT) demonstrates a cooling of more than 12 °
C when a field of 480
kilovolts per
centimeter is applied at 215 °C.[2]
Barium titanate (BaTiO3) is another electrocaloric material.[3]
There are other "caloric" effects besides magnetocaloric and electrocaloric, one of which is the
elastocaloric effect. Like the other "caloric" effects, a rapid change of an external field alters the system entropy, and for elastocalorics it's the
stress field that changes, with its effect manifest in the material
strain.[5] In 2015, a research team from the
Technical University of Denmark (Roskilde, Denmark) examined the elastocaloric effect in wires of the
superelastic alloy
NiTi, which is also known as a
shape-memory alloy.[4-5] Motivations for the study were the greater amount of
latent heat released during the elastocaloric effect than the magnetocaloric effect and its potentially higher
power density.[5]
At a
cycle rate of 2
Hz, NiTi elastocaloric devices in
wire form had a
specific cooling power that was 70 times better than gadolinium magnetocaloric devices. The specific cooling power was 7 kWh/kg, as compared with 0.1 kWh/kg for gadolinium. The elastocaloric devices, however, have a short
fatigue life, and they need large
tensile forces to operate.[5] The NiTi devices function by a
martensitic transformation, a reversible
solid-state phase transformation between two different
crystal phases that happens when a mechanical stress of about 15,000
psi (about 100
MPa) is applied. This martensitic transformation is also responsible for the temperature-induced shape memory effect and stress-induced superelasticity.[5] When stress causes the material in its
austenitic phase to transform to martensite, it heats. If the heat is extracted, and the stress removed, there's cooling during the reverse transformation.[5]
The NiTi wires were first subjected to 400 training cycles in which they were stressed (loaded), and then unstressed (unloaded), at various temperatures to stabilize their superelastic behavior.[4] The largest measured
adiabatic temperature change during loading was 25 °C, with a corresponding 21 °C change during unloading (at 48.85 °C).
Experiments showed that there are two sources of temperature irreversibility; namely, the
hysteresis in the
stress-strain curve, and a temporary
residual strain left after unloading. The residual strain produces a temporary
bending of the wire and a reduced temperature change.[4] The
Danish researchers propose guidelines about the required material properties for an efficient elastocaloric cooling device.[4] They point out the this refrigeration
technology can be used in
space applications, since it's
independent of
gravity and could be highly reliable.[5] The next step is to address the problem of limited fatigue resistance.[5]
While the NiTi devices depend on a phase transformation in their elastocaloric effect, a recently
published article in
Science utilizes just an entropy change in the same effect.[6-8] The research, on something as simple as a
twisting a
rubber band (see figure), was performed by a huge team of
scientists from
Nankai University (Tianjin, China), the
University of Texas at Dallas (Richardson, Texas),
Wuhan University (Wuhan, China), the
Tsinghua Shenzhen International Graduate School (Shenzhen, China),
Tsinghua University (Beijing, China), the
State University of Campinas (Campinas, Brazil),
Georgia Southern University (Statesboro, Georgia),
MilliporeSigma (Milwaukee, Wisconsin),
Tianjin University of Technology (Tianjin, China),
Lintec of America (Richardson, Texas), the
University of Science and Technology Liaoning (Anshan, China), and the
China Pharmaceutical University (Nanjing, China).
Twisted rubber band with an illustration of how stretch affects the molecular chains and system entropy. (Created using Inkscape and portions of screenshots from a University of Texas at Dallas video.)
As a result of the entropy change, stretched rubber bands will extract heat from their
surroundings as they are allowed to relax.[6] The research team's experiments went beyond simple stretching to twisted,
coiled, and supercoiled
fibers of
natural rubber, nickel titanium wires, and
nylon and
polyethylene fishing line.[6-7] Supercoiling is when the twisted fibers coil around themselves and coil around the coils.[7,9] For each material, 3-centimeter lengths were pulled taut, and then wound with a
rotary tool.[7]
The different fibers warmed up by as much as 15°C. When allowed to unwind, the fibers cooled by the same amount.[7] To properly assess their cooling potential, fibers were twisted and untwisted in the
water bath of a
calorimeter, and the rubber fibers produced about 20
joules of
heat energy per
gram, which was about eight times more energy than the rotary tool expended in their twisting.[7] The other materials had similar performance, which was comparable to the efficiency of standard refrigerants.[7]
To examine the
molecular basis of the effect, the research team used
X-ray analysis of the
molecular structure of the material fibers.[7] An elastocaloric system based on twist is more compact than one based on stretch.[7] Rubber needs to be stretched to seven times its length to get the same results.[7] A small refrigerator, about the size of a
ballpoint pen cartridge, created with twisted nickel titanium wires, was built, and this cooled a small
volume of water by 8°C in a few seconds.[7]
References:
- Engin Gedik, Muhammet Kayfeci, Ali Kecebas and Hüseyin Kurt, "Magnetic Refrigeration Technology Applications On Near-room Temperature," Fifth International Advanced Technologies Symposium (IATS'09), May 13-15, 2009, Karabuk, Turkey. (This reference appears to be no longer available online.)
- A. Mischenko, Q. Zhang, J.F. Scott, R.W. Whatmore, and N.D. Mathur, "Giant Electrocaloric Effect in Thin-Film PbZr0.95Ti0.05O3," Science, vol.311, no. 5765 (March 3, 2006) pp. 1270-1271. Also appears as A. Mischenko, Q. Zhang, J.F. Scott, R.W. Whatmore, and N.D. Mathur, "Giant electrocaloric effect in thin film Pb Zr_0.95 Ti_0.05 O_3," arXiv, November 19, 2005.
- Lluís Mañosa, Antoni Planes, and Mehmet Acet, "Advanced materials for solid-state refrigeration," J. Mater. Chem. A, 2013, 1, pp. 4925-4936, DOI: 10.1039/C3TA01289A; Also appears as Lluis Manosa, Antoni Planes, and Mehmet Acet, "Advanced materials for solid-state refrigeration," arXiv, March 15, 2013.
- J. Tušek, K. Engelbrecht, L. P. Mikkelsen and N. Pryds, "Elastocaloric effect of Ni-Ti wire for application in a cooling device,"
J. Appl. Phys., vol. 117 (2015), article no. 124901, http://dx.doi.org/10.1063/1.4913878.
- Squeeze to remove heat: Elastocaloric materials enable more efficient, 'green' cooling, American Institute of Physics Press Release, March 25. 2015.
Public Release: 24-Mar-2015
- Run Wang, Shaoli Fang, Yicheng Xiao, Enlai Gao, Nan Jiang, Yaowang Li, Linlin Mou, Yanan Shen, Wubin Zhao, Sitong Li, Alexandre F. Fonseca, Douglas S. Galvão, Mengmeng Chen, Wenqian He, Kaiqing Yu, Hongbing Lu, Xuemin Wang, Dong Qian, Ali E. Aliev, Na Li, Carter S. Haines, Zhongsheng Liu, Jiuke Mu, Zhong Wang, Shougen Yin, Márcio D. Lima, Baigang An, Xiang Zhou, Zunfeng Liu, and Ray H. Baughman, "Torsional refrigeration by twisted, coiled, and supercoiled fibers," Science, vol. 366, no. 6462 (October 11, 2019), pp. 216-221, DOI: 10.1126/science.aax6182.
- George Musser, "A fridge made from a rubber band? Twisted elastic fibers could cool your food," Science, October 10, 2019, doi:10.1126/science.aaz8133.
- “Twistocaloric” effect could usher in new wave of cooling technology free of greenhouse gases, YouTube Video by the University of Texas at Dallas, October 10, 2019.
- A colleague of mine who had an interest in model aircraft published an article on the twisting of rubber bands. Robert Morris, "Twist and Writhe near Max Turns in a Rubber Motor," Free Flight Quarterly, vol. 39 (April 2011), pp. 20-22. Part 2 of the article is in vol. 52 (July, 2014).