Refrigeration through voltage-controlled entropy change
09574802 ยท 2017-02-21
Assignee
Inventors
Cpc classification
F25B21/00
MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
Y02B30/00
GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
F25B2321/002
MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
International classification
Abstract
A method for refrigeration through voltage-controlled entropy change includes applying a voltage signal to a piezoelectric material to generate strain in the piezoelectric material, generating strain in a magnetic material attached to the piezoelectric material, and generating a change in a temperature of the magnetic material in response to the strain in the magnetic material.
Claims
1. A cooling device, comprising: a first material that induces strain upon application of a voltage signal, in which a level of the strain varies in response to changes in the voltage signal; a second material coupled to the first material, in which strain is induced in the second material when the strain is induced in the first material, and in which an entropy and a temperature of the second material change when the strain is induced in the second material; and a voltage signal generator to provide the voltage signal, in which the voltage signal is configured to cause the second material to vary between higher and lower temperatures.
2. The cooling device of claim 1 in which the first material comprises a piezoelectric material.
3. The cooling device of claim 1 in which the second material comprises a magnetic material.
4. The cooling device of claim 3 in which the second material comprises a ferromagnetic material.
5. The cooling device of claim 3 in which the second material comprises a magnetocaloric material.
6. The cooling device of claim 3 in which the magnetic material comprises a LaSrMnO compound.
7. The cooling device of claim 1 in which the voltage signal generator is configured to generate a voltage signal having a voltage level that changes periodically.
8. The cooling device of claim 1, comprising an actuator to cause the second material to alternately move between a first position and a second position.
9. The cooling device of claim 8 in which when the second material is at the first position, heat flows from a first object or region to the second material, and when the second material is at the second position, heat flows from the second material to a second object or region.
10. The cooling device of claim 1 in which the first and second materials have grain structures, and the first and second materials are mixed and in contact with each other.
11. The cooling device of claim 1 in which the first material is configured as a first thin film, the second material is configured as a second thin film, and the first and second thin films form a layered structure.
12. The cooling device of claim 1 in which the first material is configured as columns that are surrounded by the second material.
13. The cooling device of claim 1 in which the second material is configured as columns that are surrounded by the first material.
14. The cooling device of claim 1 in which the first and second materials are tightly bonded to each other.
15. A method of cooling a device, the method comprising: applying a voltage signal to a first material to induce strain in the first material, in which the first material is selected to have a property such that the first material induces strain upon application of the voltage signal, and a level of the strain varies in response to changes in the voltage signal; inducing strain in a second material that is coupled to the first material; and generating a change in an entropy and a temperature of the second material in response to the strain in the second material.
16. The method of claim 15 in which applying the voltage signal to the first material comprises applying the voltage signal to a piezoelectric material.
17. The method of claim 15 in which inducing the strain in the second material comprises inducing the strain in a magnetic material.
18. The method of claim 17 in which inducing the strain in the second material comprises inducing the strain in a ferromagnetic material.
19. The method of claim 17 in which inducing the strain in the second material comprises inducing the strain in a magnetocaloric material.
20. The method of claim 17 in which inducing the strain in the second material comprises inducing the strain in a LaSrMnO compound.
21. The method of claim 15 in which applying the voltage signal comprises applying an alternating voltage signal having a voltage level that changes periodically.
22. The method of claim 21 in which applying the alternating voltage signal comprises applying a first voltage to the first material to induce a first strain to cause a reduction in the temperature of the second material, and applying a second voltage to the first material to induce a second strain to cause an increase in the temperature of the second material.
23. The method of claim 15, comprising alternately moving the second material between a first position and a second position.
24. The method of claim 23, comprising flowing heat from a first object or region to the second material when the second material is at the first position, and flowing heat from the second material to a second object or region when the second material is at the second position.
25. The method of claim 15 in which the first and second materials have grain structures, and the first and second materials are mixed and in contact with each other.
26. The method of claim 15 in which the first material is configured as a first thin film, the second material is configured as a second thin film, and the first and second thin films form a layered structure.
27. The method of claim 15 in which the first material is configured as columns that are surrounded by the second material.
28. The method of claim 15 in which the second material is configured as columns that are surrounded by the first material.
29. The method of claim 15 in which the first and second materials are tightly bonded to each other.
Description
BRIEF DESCRIPTION OF DRAWINGS
(1)
(2)
(3)
(4)
DETAILED DESCRIPTION
(5) In the past, researchers conducting research on magnetic field-induced caloric effects have focused on finding optimized magnetocaloric materials. Based on Equation 1, one way to obtain large entropy change is to have a large change in magnetic field. This can be achieved by moving a magnetic material in and out of a strong magnetic field generated by a permanent magnet. The inventor recognized that Equation 1 misleadingly suggests that the use of an external magnetic field is mandatory in order to achieve a sizeable isothermal entropy change. The description below shows that applied magnetic fields are not necessary to utilize the magnetocaloric effect. In addition, voltage-induced entropy change in magnetocaloric materials has significant advantages over conventional magnetic field-induced entropy change with the potential to revolutionize magnetic refrigeration technology. It is beneficial to combine the advantages of magnetocaloric and electrocaloric materials by utilizing the magnetocaloric effect through pure voltage control.
(6) The following describes advantage of voltage-control over magnetic field-induced entropy change. The ordinary path towards sizable isothermal entropy change relies on applying magnetic fields to ever higher final values, H.sub.f, until technical saturation of the magnetization is reached. This brute-force approach has practical limitations. When relying on the maximum achievable flux densities of 1-2 Tesla of modern permanent magnets (e.g., NdFeB or SmCo), the feasible adiabatic temperature changes remain below 10 K. Permanent magnetic flux densities of the order of 4 Tesla can be generated in Halbach cylinders, but logarithmic dependence of the field on the diameter of the cylinder makes such devices very heavy. Because permanent magnets can generate magnetic fields in an energy efficient manner, most of today's realizations of magnetocaloric refrigerators utilize a mechanism that moves the magnetocaloric material relative to the permanent magnet in order to create a sizable change in magnetic field, H=H.sub.fH.sub.i. The moving parts may generate noise, losses in friction, and wear-and-tear of components. The disadvantages of magnetic field-induced entropy changes can be avoided when employing voltage-controlled entropy change in the absence of electric currents.
(7) Referring to
(8) In some implementations, a voltage signal V1 is applied to the top ferromagnetic layer 106a, and a voltage signal V2 is applied to the bottom ferromagnetic layer 106d. The voltage difference V1V2=V generates an electric field between the top and bottom ferromagnetic layers 106, in which the electric field is applied across the piezoelectric thin films 104. In this example, the voltage signal generator 102 only needs to output two voltage signals having different voltage levels.
(9) The voltage signal 108 can have, e.g., a sinusoidal waveform or a sawtooth waveform. The voltage signal 108 can have an alternating waveform having a voltage level that varies periodically. When the voltage signal 108 is applied to the thin films, strain is induced in the piezoelectric thin films 104. Because the ferromagnetic thin films 106 are attached to the piezoelectric thin films 104, the strain in the piezoelectric thin films 104 causes strain to be induced in the ferromagnetic thin films 106. When strain is induced in the ferromagnetic thin films 106, the magnetization of the ferromagnetic material changes, the entropy of the ferromagnetic thin films 106 changes, and the temperature of the ferromagnetic thin films 106 changes.
(10) By controlling the voltage signal 108, the temperature of the ferromagnetic thin films 106 can be made to increase or decrease. For example, suppose the voltage signal 108 alternates between a high positive voltage level and zero. When the high voltage level is applied to the piezoelectric thin films 104, strain is induced in the piezoelectric thin films 104 and the ferromagnetic thin films 106, causing magnetization of the ferromagnetic thin films 106 to decrease (or increase depending on the bias strain the film has at zero voltage) from the original state (the state when no voltage is applied), thereby causing the entropy to increase in an isothermal situation and, in an adiabatic situation suitable for the refrigeration application, temperature to decrease. When the zero voltage level is applied to the piezoelectric thin films 104, strain is removed from in the piezoelectric thin films 104 and the ferromagnetic thin films 106, causing magnetization of the ferromagnetic thin films 106 to increase and return to the original state, thereby causing the entropy to decrease and temperature to increase.
(11) For example, each of the LSMO thin films can have a thickness in the order of a few nanometers to tens of nanometers, e.g., 20 nm. An advantage of using LSMO thin films instead of LSMO bulk material is that strain can be induced in a large portion (e.g., the entire portion) of the LSMO thin film, which increases the magnetocaloric effect, resulting in a larger temperature change. If LSMO bulk material is used, strain is induced only near the surface of the LSMO material that is in contact with the piezoelectric material so that the magnetocaloric effect is smaller (for a given amount of LSMO material).
(12) The following describes the mechanism for realization of voltage-activated entropy change using the magnetocaloric effect. In some implementations, piezoelectrically-induced strain can be used to control anisotropy and critical temperature of magnetic materials. For example, piezoelectrically-induced strain can be used to substantially change the magnetic anisotropy in iron (Fe) thin films or used to control the exchange-bias field in an exchange-bias magnetic heterostructure. Similarly, strain originating from stress can be induced via the inverse piezoelectric effect. The strain, when carried over into an adjacent magnetic thin film, can substantially change the magnetic Curie temperature, T.sub.C of the magnetic material. An external control parameter such as tensile or compressive strain can tune the degree of magnetic order in a magnetically long-range ordered system. Decrease (increase) in long-range magnetic order can be accompanied by a significant increase (decrease) in entropy. The transition from a paramagnetic into a long-range ordered magnetic state is accompanied by sizeable entropy reduction.
(13) For example, a compressively strained La.sub.0.7Sr.sub.0.3MnO.sub.3 (LSMO) film of 16 nm thickness on LaAlO.sub.3 (001) changes the Curie temperature of the former from 365 K (unstrained) to 270 K (strained through lattice mismatch). Voltage-controlled epitaxial strain in LSMO can be achieved when exploiting the inverse piezoelectric effect of PbMg.sub.1/3Nb.sub.2/3O.sub.3PbTiO.sub.3(001) (PMN-PT) substrates. When the temperature of a permanent magnet increases above a certain point, referred to as the Curie temperature, the permanent magnetism changes to paramagnetism. For a given temperature, the magnetization of the LSMO film increases when an electric field is applied to the compressively strained La.sub.0.7Sr.sub.0.3MnO.sub.3 (LSMO) film on LaAlO.sub.3 (001). In addition, when an electric field is applied to the compressively strained La.sub.0.7Sr.sub.0.3MnO.sub.3 (LSMO) film on LaAlO.sub.3 (001), the Curie temperature (or critical temperature) in the LSMO film tends to increase. Various parameters allow fine-tuning of the critical temperature at zero applied voltage. The tuning parameters include the Sr concentration of the LSMO compound, the film thickness, and the initial strain in zero-applied electric field.
(14) Changes of the critical temperature, T.sub.C, of ferromagnetic thin films such as the complex oxide LSMO can be achieved by pure voltage-control, in which the inverse piezoelectric effect is used in order to strain the LSMO film in an electrically controlled manner. Large magnetoelectric susceptibilities can be achieved, e.g., in complex oxides, when using the electric field effect. For example, sizable electric modulation of magnetization in a BaTiO.sub.3/LSMO heterostructure can be achieved. The electric field-controlled metal-insulator transition in the LSMO film produces a large magnetoelectric effect that can be used in the same way as the strain-induced change in magnetization.
(15) The following describes thermodynamic consequences of large magnetoelectric effects and their impact on voltage-controlled isothermal entropy change. Theoretical basis for the voltage-controlled entropy change is described below. The differential form of the Helmholtz free energy, F, of a magnetic system, can be represented as follows:
dF=SdT+.sub.0VHdM.(Equ. 2)
(16) When the applied magnetic field is zero, the variable H in Equation 2 can be considered to be the internal magnetic field, which depends on temperature T and the equilibrium magnetization M. The magnetization in turn is a function of a control parameter, which in this case is the electric field E. Therefore, the equation of state can be written as
H=H(T,M(E))(Equ. 3)
From Equation 2, we derive a Maxwell relation by identifying the mixed second-order derivatives of the Helmholtz free energy:
(17)
Integration of Equation 4 provides the expression:
(18)
for the isothermal entropy change. Note that Equation 5 differs from Equation 1, which is derived from a different Maxwell relation originating from the Gibbs free energy.
(19) Using thermodynamic identities,
(20)
we get
(21)
which leads after substitution into Equation 5 a formula for the isothermal entropy change:
(22)
where
(23)
is the magnetic susceptibility. From the fact that M depends on the control parameter, E, we substitute
(24)
to obtain
(25)
(26) The following describes numerical estimates for voltage-controlled entropy change in a heterostructure that includes a piezoelectric material PMN-PT and a magnetocaloric material LSMO. The values of the parameters can be estimates or values based on experiment data. To further explore the consequences of Equation 7, we consider the Landau expression F=F.sub.0(T)+AM.sup.2+BM.sup.4.sub.0VHM. Here, A=a.sub.0(TT.sub.C(E)), a.sub.0 and B are positive constants, and F.sub.0 is a regular temperature-dependent background. The Landau expression allows specifying M and x in terms of the expansion coefficients. This yields the entropy change for TT.sub.C(0)
(27)
One can replace the parameters of the Landau expansion using
(28)
which yields
(29)
(30) For example, using a density value for LSMO of =6600 kg/m.sup.3, we calculate a mass-specific isothermal entropy change which can now be used for comparison with current state-of-the-art magnetic field-induced, specific entropy changes. For example, the saturation magnetization of a pulsed laser deposited LSMO film of 30% Sr concentration can be M(T=0)0.45 MA/m. We use the mean-field expression,
(31)
which for T=2T.sub.C yields
(32)
where S3.5. Using further a Land g-factor of g2 in accordance with Ref.(.sup.i), T.sub.C(E=0)=279K and T.sub.C=T.sub.C(E=7 kV/cm)T.sub.C(0)=19 K, we estimate the value of the specific entropy change, which yields
(33)
(34) Note that S(T, E) and, therefore, Equation 8 can be directly calculated from the Landau free energy according to
(35)
This yields
(36)
(37) Equations 7 and 11 are both useful. Equation 7 is free from approximations. Also, and M(T,E) can be measured while the Landau free energy is a crude approximation of the underlying thermodynamic potential F and may not be experimentally accessible.
(38) Alternatively, we estimate the mass-specific isothermal entropy change at T=280 K from the magnetoelectric susceptibility,
(39)
of PMN-PT/LSMO. The numerical value of a is determined from the data adapted from C. Thiele, K. Drr, O. Bilani, J. Rdel, and L. Schultz, Phys. Rev. B 75, 054408 (2007). For E=0 and 7 kV/cm the magnetization data can be described by the functions
(40)
which yields
(41)
We use this value of magnetoelectric susceptibility for further estimates and neglect the details of the temperature dependence of . We use a rough estimate for
(42)
for LSMO by extrapolating M (T, E=0) towards T=0. The extrapolation of the Landau expression overestimates the magnetization at T=0 which, in turn, gives rise to an underestimation of the isothermal entropy change. We obtain
(43)
from the M (T, E=0) function. We assume a linear dependence of T.sub.C on the applied electric field which reads T.sub.C(E)=2.4310.sup.5 E Km/V+280 K. Using Equation 10 to quantify (T, T.sub.C(E)), we estimate the mass-specific isothermal entropy change from
(44)
It yields
(45)
in good agreement with the alternative approach based on Equation 9 and outlined above. The remaining difference in the numerical values of the specific isothermal entropy changes originates from differences in the assumptions and approximations. Equation 12 shows that entropy change can be achieved without applying a magnetic field.
(46) The following describes refrigerant capacity and implications for refrigeration applications. The voltage-induced entropy change estimated above is of respectable magnitude when compared, e.g., with the bulk giant magnetocaloric material Gd.sub.5Si.sub.2Ge.sub.2, which represents a benchmark for magnetocaloric materials. Gd.sub.5Si.sub.2Ge.sub.2 has an isothermal entropy change of approximately 4 J/kgK when an external magnetic field is ramped from zero to 1 Tesla. Although this value is still approximately twice the entropy change we estimate for the voltage-controlled effect in PMN-PT/LSMO, it is important to realize that for virtually all magnetocaloric materials and Gd.sub.5Si.sub.2Ge.sub.2 in particular, the isothermal entropy change strongly peaks at a given temperature, T.sub.max, and decreases for both higher and lower temperatures. This limits the refrigerant capacity (RC), which is the figure-of-merit of a refrigerator. The refrigerant capacity can be calculated from the temperature-dependence of the isothermal entropy change according to
(47)
Here T is the width of S(T) at half-maximum. Therefore, for the narrow Gaussian S vs. T behavior, the refrigerant capacity is limited despite potentially large values of S at the maximum of S vs. T.
(48) In the case of voltage-controlled entropy change, S vs. T will remain virtually constant for all T<T.sub.C(0) at the value given by Equation 9. The absolute value of the entropy change decreases linearly to zero for T>T.sub.C(0) and remains zero for TT.sub.C(E).
(49) Referring to
(50) Dashed line 128 indicates the critical temperature T.sub.C(0) of the ferromagnetic film in electric field E=0. Dashed line 128 indicates the critical temperature T.sub.C(E=E.sub.f) of the ferromagnetic film in electric field E=E.sub.f. The line 126 shows S, which is the difference between the entropy value when voltage is applied, and the entropy value when a voltage is applied. The line 126 (right axis) shows the temperature dependence of the isothermal entropy change. The line 126 indicates that a significant difference in entropy S exists (|S|0.19) for a wide range of temperature values, e.g., from less than 200K to about 278K. This indicates that the cooling device 100 can be used for cooling for a wide range of temperatures. The absolute value of the entropy difference |S| decreases from about 0.19 to 0 when the temperature increases from about 278K to about 298K. This indicates that the cooling device 100 is useful for room temperature cooling applications.
(51) The temperature independence of S vs. T for the temperatures T<T.sub.C(0) largely increases the refrigerant capacity to values potentially much higher than those reported in the literature. The emphasis here is to achieve sizeable entropy change in the absence of applied magnetic fields. Pure voltage-controlled entropy change broadens the range of potential applications when compact cooling solutions with little to no mechanical vibrations are required.
(52) The following describes realization of multiferroic materials for voltage-controlled entropy change. The bilayer system PMN-PT/LSMO shown in
(53) Another candidate of a magnetoelectric composite for magnetocaloric applications is a laminate composite of piezoelectric AlN and amorphous FeCoSiB which can be fabricated by sputtering methodology and has a high magnetoelectric effect at room temperature. Operation of the laminate in alternating current (AC) mode at resonance frequency and in the presence of a small biasing magnetic field (order of the Earth's magnetic field) gives rise to 1.4 10.sup.7 s/m, which may prove suitable as an alternative to the complex oxide composites analyzed here in more detail.
(54) The M-type hexaferrite SrCo.sub.2Ti.sub.2Fe.sub.8O.sub.19 shows a promising large magnetoelectric effect at room temperature, which is about 50 times higher than the maximum magnetoelectric susceptibility 4 10.sup.12 s/m of the archetypical magnetoelectric chromia. Although the bulk magnetoelectric susceptibility is orders of magnitude below the magnetoelectric response of LSMO and FeCoSiB composites, the bulk magnetoelectric multiferroics may be produced at a lower cost and can be alternatives to the composite materials that include nanolayers of LSMO.
(55) Referring to
(56) Referring to
(57) Referring to
(58) Referring to
(59) The concept of voltage-controlled entropy change in magnetocaloric materials for magnetic refrigeration applications has been described above. One of the key features of this approach is that the magnetocaloric effect is utilized without applying an external magnetic field. We estimate a specific isothermal voltage-controlled entropy change for the bilayer heterostructure PMN-PT/LSMO is larger than 1 J/kgK and serves as proof of principle for voltage-controlled magnetic refrigeration near room temperature.
(60) Referring to
(61) The foregoing description is intended to illustrate and not to limit the scope of the invention, which is defined by the scope of the appended claims. Other embodiments are within the scope of the following claims. For example, the thicknesses and the materials used for the piezoelectric thin films and the magnetic thin films can be different from those described above. The critical temperatures and the amount of entropy change in the materials can be different from those described above.
(62) Instead of using a two-phase composite material having distinct piezoelectric material component and ferromagnetic material component, a magnetoelectrically active material in bulk compound form having a high magnetoelectric susceptibility can also be used. For example, the bulk compound may have piezoelectric grains mixed with ferromagnetic grains. The magnetoelectrically active material in bulk compound form can be made at a lower cost (compared to using the piezoelectric thin films and ferromagnetic thin films). When a voltage difference is applied to electrodes across the magnetoelectrically active material, an electric field is generated across the magnetoelectrically active material, causing the magnetization of the magnetoelectrically active material to change, and in an adiabatic situation causing the temperature of the magnetoelectrically active material to change.
(63) In
(64) In