WO2025047008A1 - Hybrid horizontal thermoelectric temperature modulation element and temperature modulation method using same - Google Patents

Abstract

The present invention provides a horizontal thermoelectric conversion element having a thermoelectric temperature modulation performance that can be on par with that obtained through the conventional Peltier effect, in comparison. The present invention is provided with: a first electrode and a second electrode which are disposed so as to face each other; and a laminate which is held between the first and the second electrodes and which is electrically connected to both the first and second electrodes. The laminate has a structure in which layers of a first thermoelectric material and layers of a second thermoelectric material are alternately laminated, and the value of at least one of Peltier coefficient, electrical conductivity, or thermal conductivity is different between the first thermoelectric material and the second thermoelectric material. At least one of the first and second thermoelectric materials exhibits a vertical magnetic thermoelectric effect or a horizontal magnetic thermoelectric effect. The laminate surfaces of the first layers and the second layers are inclined with respect to the direction in which the first electrode and the second electrode face each other. By having a current supplied between the first and second electrodes and having a magnetic field applied in a direction in which the magnetic thermoelectric effect is to be generated, a temperature difference is generated in a direction perpendicular to the current.

Description

Hybrid lateral thermoelectric temperature modulation element and temperature modulation method using the same

The present invention relates to a hybrid lateral thermoelectric temperature modulation element that combines the magneto-thermoelectric effect with lateral thermoelectric conversion derived from an anisotropic laminated structure, and a temperature modulation method using the same.

Thermoelectric conversion is the direct conversion of thermal energy to electrical energy in a material.
When a potential difference is applied to a thermoelectric material by an external power source, a temperature difference can be generated across both ends of the thermoelectric material. This is called the Peltier effect, and thermoelectric cooling and temperature modulation are achieved (see Non-Patent Document 1).
When a temperature difference is applied to a thermoelectric material, the kinetic energy of the carriers (electrons and holes) is greater at the high temperature end, so they are biased toward the low temperature end due to thermal diffusion, generating a potential difference between both ends of the thermoelectric material. This is called the Seebeck effect, and is used in thermocouples and thermoelectric power generation.

The Peltier effect is a phenomenon in which heat is absorbed and released at the junction surface when different conductors are joined and a voltage and current are applied, which is the opposite phenomenon to the Seebeck effect (see Non-Patent Document 1). In a thermoelectric conversion module using the Peltier effect, n-type and p-type thermopiles are arranged in a matrix and connected in series to increase the amount of thermoelectric cooling and heating and to achieve a large surface area, which is the basic conventional structure.
However, conventional Seebeck/Peltier elements have a complex three-dimensional module structure (Π-type structure),
・Easily broken,
・High cost
- The large number of junctions between different materials increases the interfacial electrical and thermal resistance, preventing the module from achieving the efficiency expected from the material's performance index.
There was a problem.

On the other hand, a horizontal thermoelectric conversion element has been developed in which the input voltage and current are perpendicular to the output temperature gradient direction. The horizontal thermoelectric conversion element does not require a Π-type module structure, and can solve the above problems (see Non-Patent Document 2).
However, conventional horizontal thermoelectric conversion elements have a problem in that their thermoelectric cooling/heating performance is still low compared to Peltier elements.

Next, as one type of horizontal thermoelectric conversion element, an artificial multilayer structure in which two types of conductors are alternately and obliquely laminated (hereinafter referred to as an artificially inclined multilayer structure) is known (see Non-Patent Document 2).
In an artificially graded multilayer laminate, even if each conductor exhibits isotropic conduction characteristics, anisotropic conduction of conduction electrons and holes occurs, and the off-diagonal terms of the thermoelectric transport tensor become finite, so that it functions as a lateral thermoelectric conversion (see Non-Patent Document 3). Lateral thermoelectric conversion in such artificial composite materials has been studied for a long time, and it is possible to design the lateral thermoelectric power and the dimensionless figure of merit ZT by selecting appropriate constituent materials and optimizing the cutting angle and element shape. Most research on artificially graded multilayer laminates has been carried out on sinter-bonded bulk materials, and ZTs exceeding 0.2 have been achieved. Lateral thermoelectric conversion has been demonstrated not only in macro-scale sintered bodies but also in nanometer-scale superlattices with similar graded structures, and is often called a (p×n)-type multilayer film (see Non-Patent Document 4).

As a multilayer structure for thermoelectric conversion, for example, a multilayer structure of Bi and metal ( _Bi2Te3 ) is known, as disclosed in Patent Document 1 (see paragraph [0081]). Patent Document 2 discloses a _graded superlattice structure (InAs/GaSb) that drives lateral thermoelectric power generation, and lists the Nernst-Ettingshausen effect and a composite laminate device as background art (see paragraph [0008]).
However, Patent Documents 1 and 2 make no mention of hybridizing the magneto-thermoelectric effect originating from magnetism in an artificially inclined multilayer laminate.

WO2008/056466 publication US Patent No. 8829324

Katsushi Fukuda, "Practical Use of Peltier Modules", Heat Transfer, May 2005, pp. 18-21 (J. HTSJ, Vol. 44, No. 186) Japan Society of Applied Physics Home Page > GX: Applied Physics Challenging Green Transformation > Horizontal Thermoelectric Conversion / Kenichi Uchida https://www.jsap.or.jp/columns/gx/e3-5 H. J. Goldsmid, Introduction to Thermoelectricity (Springer-Verlag, Berlin Heidelberg, 2010). Y. Tang, B. Cui, C. Zhou, and M. Grayson, “p×n-Type Transverse Thermoelectrics: A Novel Type of Thermal Management Material,” J. Electron. Mater. 44, 2095 (2015).

　Previously known horizontal thermoelectric conversion elements had the problem of inferior thermoelectric temperature modulation performance compared to Peltier elements, which are common thermoelectric elements.

The present invention aims to solve the problems of the conventional technology described above, and to provide a horizontal thermoelectric conversion element that has thermoelectric temperature modulation performance equivalent to or higher than that of a Peltier element, and a temperature modulation method using the same.

The inventors of the present invention believed that if a material exhibiting a magneto-thermoelectric effect is incorporated into an artificially inclined multilayer laminate, hybrid thermoelectric conversion would be possible through the magnetic field-independent contribution from the conventional anisotropic multilayer structure and the magnetic field/magnetization-dependent magneto-thermoelectric effect, and that it would be possible to provide a lateral thermoelectric conversion element with thermoelectric temperature modulation performance comparable to that of existing Peltier elements or even higher performance. This led to the invention.
[1] The hybrid lateral thermoelectric temperature modulation element of the present invention is, as shown in, for example, FIG. 3A, FIG. 3C, and FIG. 9A, a thermoelectric temperature modulation element including a first electrode and a second electrode arranged opposite to each other, and a laminate sandwiched between the first and second electrodes and electrically connected to both the first and second electrodes,
The laminate has a structure in which layers of a first thermoelectric material and layers of a second thermoelectric material are alternately stacked, the first thermoelectric material and the second thermoelectric material have different values for at least one of the Peltier coefficient, the electrical conductivity, and the thermal conductivity, at least one of the first and second thermoelectric materials exhibits a vertical magneto-thermoelectric effect or a horizontal magneto-thermoelectric effect, the stacking surfaces of the layers of the first thermoelectric material and the layers of the second thermoelectric material are inclined with respect to the opposing direction of the first electrode and the second electrode, and a temperature difference is generated in a direction perpendicular to the opposing direction in the thermoelectric temperature modulation element by a current supplied between the first and second electrodes and a magnetic field applied in a direction generating the magneto-thermoelectric effect.

[2] In the hybrid transverse thermoelectric temperature modulation element [1] of the present invention, preferably, the vertical magneto-thermoelectric effect exhibits a thermoelectric power change of 5 μV/K or more, more preferably 10 μV/K or more, and is a magnetic Peltier effect derived from an external magnetic field, or an anisotropic magnetic Peltier effect derived from spontaneous magnetization, and the transverse magneto-thermoelectric effect exhibits a thermoelectric power change of 5 μV/K or more, more preferably 10 μV/K or more, and is a normal Etchingshausen effect derived from an external magnetic field, or an abnormal Etchingshausen effect derived from spontaneous magnetization. If the thermoelectric power of the vertical magneto-thermoelectric effect and the transverse magneto-thermoelectric effect is less than 5 μV/K, there is a disadvantage that the enhancement effect on the transverse thermoelectric effect derived from the inclined stack structure is insufficient, but there is no problem with the operation itself.
[3] In the hybrid lateral thermoelectric temperature modulation element [1] or [2] of the present invention, preferably, the first thermoelectric material layer is a Bi _100-x Sb _x layer (0≦x≦50), and the second thermoelectric material layer is a Bi _2-y Sb _y Te ₃ layer (0≦y≦2). More preferably, the first thermoelectric material layer is a Bi _100-x Sb _x layer (5≦x≦25), and the second thermoelectric material layer is a Bi _2-y Sb _y Te ₃ layer (1.5≦y≦1.95). If x exceeds 50, there is a disadvantage that the magneto-thermoelectric effect decreases. Since the optimal characteristics of the second thermoelectric material depend on what is used for the first thermoelectric material, there is an advantage that the optimal composition can be selected according to the thermoelectric characteristics of the first thermoelectric material.
[4] In the hybrid lateral thermoelectric temperature modulation element [1] or [2] of the present invention, preferably, the first thermoelectric material layer is a Co ₂ MnGa layer, and the second thermoelectric material layer is a Bi _2-y Sb _y Te ₃ layer (0≦y≦2). Alternatively, the second thermoelectric material layer may be a Bi ₂ Te ₃ layer (y=0). Since the optimal properties of the second thermoelectric material depend on the first thermoelectric material, there is an advantage that an optimal composition can be selected according to the thermoelectric properties of the first thermoelectric material.
[5] In the hybrid lateral thermoelectric temperature modulation element [1] or [2] of the present invention, preferably, the external magnetic field is applied to the first thermoelectric material layer and the second thermoelectric material layer by a permanent magnet, an electromagnet, or a permanent magnet and an electromagnet.
[6] In the hybrid lateral thermoelectric temperature modulation element [1] of the present invention, preferably, the first thermoelectric material layer is made of a material exhibiting a magneto-thermoelectric effect, and the second thermoelectric material layer is made of a permanent magnet material, and a magnetic field can be applied to the first thermoelectric material by the spontaneous magnetization of the permanent magnet without applying an external magnetic field, so that the magneto-thermoelectric effect of the first thermoelectric material can be expressed.
[7] In the hybrid lateral thermoelectric temperature modulation element [6] of the present invention, preferably, the first thermoelectric material layer is a Bi _100-x Sb _x layer (0≦x≦50), and the second thermoelectric material layer is one or more types of permanent magnet material selected from the group consisting of SmCo ₅ magnets, Sm ₂ Co ₁₇ magnets, Nd ₂ Fe ₁₄ B magnets, alnico magnets, and ferrite magnets. More preferably, the first thermoelectric material layer is a Bi _100-x Sb _x layer (5≦x≦25). If x exceeds 50, there is a disadvantage that the magneto-thermoelectric effect decreases.
[8] In the hybrid lateral thermoelectric temperature modulation element [1] of the present invention, preferably, the inclination angle θ of the lamination surface with respect to the above-mentioned direction is 10° or more and 80° or less. When the inclination angle θ is less than 10° or more than 80°, the lateral thermoelectric effect derived from the inclined lamination may not be fully exhibited.
[9] In the hybrid lateral thermoelectric temperature modulation element [1] of the present invention, it is preferable that the figure of merit of the hybrid lateral thermoelectric temperature modulation element is 0.1 or more. If the figure of merit is less than 0.1, there is a possibility that the hybrid thermoelectric conversion element does not exhibit sufficient performance.
[10] In the hybrid lateral thermoelectric temperature modulation element [1] of the present invention, preferably, the magnetic field applied in the direction in which the magneto-thermoelectric effect is generated is 0.1 T or more.

[11] A temperature modulation method for a horizontal thermoelectric conversion element of the present invention is a temperature modulation method using a horizontal thermoelectric conversion element, which comprises supplying a current to the horizontal thermoelectric conversion element and applying a magnetic field in a direction that generates a magneto-thermoelectric effect, thereby generating a temperature difference in a direction perpendicular to the current supply direction of the horizontal thermoelectric conversion element,
The horizontal thermoelectric conversion element is
a first electrode and a second electrode arranged opposite to each other;
a laminate sandwiched between the first and second electrodes and electrically connected to both the first and second electrodes;
The laminate has a structure in which layers of a first thermoelectric material and layers of a second thermoelectric material are alternately stacked, the first thermoelectric material and the second thermoelectric material have different values for at least one of a Peltier coefficient, an electrical conductivity, and a thermal conductivity, at least one of the first and second thermoelectric materials exhibits a vertical magneto-thermoelectric effect or a horizontal magneto-thermoelectric effect, and the stacking surfaces of the layers of the first thermoelectric material and the layers of the second thermoelectric material are inclined with respect to the opposing direction of the first electrode and the second electrode, and a current is supplied via the first and second electrodes, and a magnetic field is applied in a direction that generates a magneto-thermoelectric effect, thereby generating a temperature difference in a direction perpendicular to the opposing direction in the horizontal thermoelectric conversion element.

[12] In the temperature modulation method for a horizontal thermoelectric conversion element [11] of the present invention, preferably, the vertical magneto-thermoelectric effect exhibits a thermoelectric power change of 5 μV/K or more, more preferably 10 μV/K or more, and is a magnetic Peltier effect derived from an external magnetic field, or an anisotropic magnetic Peltier effect derived from spontaneous magnetization, and the horizontal magneto-thermoelectric effect exhibits a thermoelectric power change of 5 μV/K or more, more preferably 10 μV/K or more, and is a normal Etchingshausen effect derived from an external magnetic field, or an anomalous Etchingshausen effect derived from spontaneous magnetization. If the thermoelectric power due to the vertical magneto-thermoelectric effect and the horizontal magneto-thermoelectric effect is less than 5 μV/K, there is a disadvantage that the enhancement effect on the horizontal thermoelectric effect derived from the inclined stack structure is insufficient.
[13] In the temperature modulation method of the horizontal thermoelectric conversion element of the present invention [11] or [12], preferably, the first thermoelectric material layer is a Bi _100-x Sb _x layer (0≦x≦50), and the second thermoelectric material layer is a Bi _2-y Sb _y Te ₃ layer (0≦y≦2). More preferably, the first thermoelectric material layer is a Bi _100-x Sb _x layer (5≦x≦25), and the second thermoelectric material layer is a Bi _2-y Sb _y Te ₃ layer (1.5≦y≦1.95). If x exceeds 50, there is a disadvantage that the magneto-thermoelectric effect decreases. Since the optimal characteristics of the second thermoelectric material depend on what is used for the first thermoelectric material, there is an advantage that the optimal composition can be selected according to the thermoelectric characteristics of the first thermoelectric material.
[14] In the temperature modulation method for a lateral thermoelectric conversion element of the present invention [11] or [12], preferably, the first thermoelectric material layer is Co ₂ MnGa, and the second thermoelectric material layer is Bi _2-y Sb _y Te _3- layer (0≦y≦2). Alternatively, the second thermoelectric material layer may be Bi ₂ Te _3- layer (y=0). Since the optimal properties of the second thermoelectric material depend on the first thermoelectric material, there is an advantage that an optimal composition can be selected depending on the thermoelectric properties of the first thermoelectric material.
[15] In the temperature modulation method for a horizontal thermoelectric conversion element of the present invention [12], the external magnetic field is preferably applied to the first thermoelectric material layer and the second thermoelectric material layer by a permanent magnet, an electromagnet, or a permanent magnet and an electromagnet.
[16] In the temperature modulation method for a horizontal thermoelectric conversion element [11] of the present invention, preferably, the first thermoelectric material layer is made of a material exhibiting a magneto-thermoelectric effect, and the second thermoelectric material layer is made of a permanent magnet material, so that a magnetic field can be applied to the first thermoelectric material without applying an external magnetic field by spontaneous magnetization of the permanent magnet, and the magneto-thermoelectric effect of the first thermoelectric material can be expressed.
[17] In the temperature modulation method for a horizontal thermoelectric conversion element [16] of the present invention, preferably, the first layer is a Bi _100-x Sb _x layer (0≦x≦50), and the permanent magnet layer is one or more types of permanent magnet material selected from the group consisting of SmCo ₅ magnets, Sm ₂ Co ₁₇ magnets, Nd ₂ Fe ₁₄ B magnets, alnico magnets, and ferrite magnets. More preferably, the layer of the first thermoelectric material is a Bi _100-x Sb _x layer (5≦x≦25). If x exceeds 50, there is a disadvantage that the magneto-thermoelectric effect decreases.
[18] In the temperature modulation method for a lateral thermoelectric conversion element of the present invention [11], it is preferable that the figure of merit of the lateral thermoelectric conversion element is 0.1 or more. If the figure of merit is less than 0.1, there is a possibility that the hybrid thermoelectric conversion element does not exhibit sufficient performance.
[19] In the temperature modulation method for a horizontal thermoelectric conversion element of the present invention [11], preferably, the magnetic field applied in the direction in which the magneto-thermoelectric effect is generated is 0.1 T or more.

According to the hybrid lateral thermoelectric temperature modulation element of the present invention, the layer of the first thermoelectric material exhibiting the magneto-thermoelectric effect and the layer of the second thermoelectric material are alternately laminated, and the first layer is a Bi _100-x Sb _x layer (0≦x≦50) and the second layer is a Bi _2-y Sb _y Te ₃ layer (0≦y≦2), so that the layer of the first thermoelectric material can improve the lateral thermoelectric power generation characteristics as the applied magnetic field increases due to at least one of the magnetic Peltier effect and the normal Etchingshausen effect. When the _{Bi 100-x} Sb _x layer is replaced with a magnetic material, the lateral thermoelectric power generation characteristics can be improved along with the magnetization process due to at least one of the anisotropic magnetic Peltier effect and the anomalous Etchingshausen effect.
In addition, according to the hybrid lateral thermoelectric temperature modulation element of the present invention, the first thermoelectric material layer exhibiting the magneto-thermoelectric effect and the second thermoelectric material layer having the remanent magnetization and capable of driving the magneto-thermoelectric effect in the absence of a magnetic field are alternately laminated, and the first thermoelectric material layer is a Bi _100-x Sb _x layer (0≦x≦50), and the second thermoelectric material layer is one or more types of permanent magnets selected from the group consisting of SmCo ₅ magnets, Sm ₂ Co ₁₇ magnets, Nd ₂ Fe ₁₄ B magnets, alnico magnets, and ferrite magnets. Therefore, the lateral thermoelectric power generation characteristics can be improved by the magnetic Peltier effect and the normal Etchingshausen effect in the magnetized state compared to the unmagnetized state of the second thermoelectric material layer. When the Bi _100-x Sb _x layer is replaced with a magnetic material, the magnetization process and the lateral thermoelectric power generation characteristics can be improved due to at least one of the anisotropic magnetic Peltier effect and the anomalous Etchingshausen effect.

FIG. 1 is a diagram showing a systematic classification of magneto-thermoelectric effects in simple materials. FIG. 2A is a configuration diagram of a lock-in thermography measurement device suitable for measuring the thermoelectric effect by focusing on the cooling and heating phenomena. FIG. 2B is a waveform diagram of the input current and the output temperature change signal in the lock-in thermography measurement. FIG. 2C is an observation diagram of the amplitude and phase measured by a lock-in thermography measurement device. FIG. 3A is an explanatory diagram of visualization of lateral thermoelectric conversion in an artificially inclined multilayer laminate in a zero magnetic field, showing a perspective view of the sample configuration and a steady-state temperature image of the imaging surface. FIG. 3B is an explanatory diagram of the visualization of lateral thermoelectric conversion in an artificially inclined multilayer stack in a zero magnetic field, showing the observed amplitude and phase when a net heat flow in the lateral direction is generated, as measured by a lock-in thermography measurement device. FIG. 3C is an explanatory diagram of visualization of lateral thermoelectric conversion in an artificially inclined multilayer laminate in zero magnetic field, showing a perspective view of the sample configuration and a steady-state temperature image of the imaging surface. FIG. 3D is an explanatory diagram of visualization of lateral thermoelectric conversion in an artificially inclined multilayer stack in zero magnetic field, and shows an observed diagram of amplitude and phase when a net heat flow is generated flowing from the top surface to the bottom surface, as measured by a lock-in thermography measurement device. FIG. 4 shows the separation of the component showing an even dependence on the magnetic field (symmetric component) and the component showing an odd dependence on the magnetic field (antisymmetric component). FIG. 5 illustrates the contribution of the magnetic Peltier effect, which shows an even dependence on the magnetic field. FIG. 6 shows the contribution of the normal Etchingshausen effect, which shows an odd dependence on the magnetic field. FIG. 7 shows the magnetic field dependence of the average temperature modulation. FIG. 8 represents steady state thermoelectric cooling. FIG. 9A is a perspective view showing the configuration of a permanent magnet-based artificially inclined multilayer stack for horizontal thermoelectric conversion. FIG. 9B is an observation of the amplitude and phase of temperature change in a permanent magnet-based artificially tilted multi-layer stack for lateral thermoelectric conversion. FIG. 9C shows the magnetic properties and lateral thermoelectric conversion properties of the permanent magnet layers in a permanent magnet-based artificially inclined multilayer stack for lateral thermoelectric conversion. FIG. 10A shows the magnetic field dependence of _the thermoelectric power generation output _{of a Bi88Sb12} / _{Bi0.2Sb1.8Te3} artificially inclined multilayer stack, and shows the results when a temperature difference ΔT of 7.9 K was applied to sample A, which has an inclination angle θ of 45° with respect to the bottom surface of the multilayer stack. FIG. 10B shows the magnetic field dependence of _the thermoelectric power output of _{the Bi88Sb12} / _{Bi0.2Sb1.8Te3} artificially inclined multilayer stack, and shows the results when a temperature difference ΔT _of 9.8 K is applied to sample B, which has an inclination angle θ of 21° with respect to the bottom surface of the multilayer stack. FIG. 11A is a diagram showing the configuration of a horizontal thermoelectric element cut into a rectangle, showing an example of an artificially inclined multilayer laminate. FIG. 11B is a perspective view showing a configuration for measuring the magnetic field dependence of the thermoelectric power output of a Bi ₈₈ Sb ₁₂ /Bi _0.2 Sb _1.8 Te ₃ artificially gradient multilayer laminate. FIG. 11C is a diagram showing the dependency of the external magnetic field H in the correlation diagram between the temperature difference ΔT and the electromotive force V in the device of FIG. 11B. FIG. 11D is a diagram showing the temperature difference ΔT dependency in the correlation diagram between the external magnetic field H and the electromotive force V in the device of FIG. 11B. FIG. 12A is a perspective view showing a configuration for measuring the magnetic field dependence of the thermoelectric power output of a Co ₂ MnGa-based artificially inclined multilayer laminate. FIG. 12B shows the results of measuring the magnetic field dependence of the transverse electromotive force when the temperature difference ΔT of the Co ₂ MnGa/Bi _0.2 Sb _1.8 Te ₃ artificially graded multilayer laminate is changed. FIG. 12C shows the results of measuring the magnetic field dependence of the transverse electromotive force when the temperature difference ΔT of the Co ₂ MnGa/Bi ₂ Te ₃ artificially graded multilayer stack is changed. FIG. 12D shows the results of measuring the temperature gradient dependence of the lateral electric field of Co ₂ MnGa/Bi _0.2 Sb _1.8 Te ₃ and Co ₂ MnGa/Bi ₂ Te ₃ artificially graded multilayer stacks.

(Definitions of technical terms used in thermoelectric conversion)
First, the technical terms used in this specification are defined.
Thermoelectric conversion is a direct conversion phenomenon between thermal energy and electrical energy in a material. The input and output are merely switched between the power generation phenomenon and the cooling/heating (temperature modulation) phenomenon, and it can be evaluated by the dimensionless figure of merit ZT. That is, when a module using thermoelectric conversion is applied to a place where there is a temperature difference, electric power can be obtained by the electromotive force derived from the Seebeck effect. On the other hand, when a module using thermoelectric conversion is connected to an external power source and a direct current is applied, a temperature difference occurs due to the absorption and generation of heat derived from the Peltier effect. Here, ZT is expressed as follows using the electrical resistivity ρ, thermal conductivity κ, and Seebeck coefficient S, which are physical properties specific to the material:
ZT=S ² T/(ρκ) (1)
Here, the Seebeck coefficient S represents the thermoelectromotive force per 1 K temperature difference in the Seebeck effect. When materials with different Seebeck coefficients are joined and a temperature difference is applied, a thermoelectromotive force is generated in each material, and the difference voltage between the electromotive forces in both materials is generated as the open circuit voltage. (See Masayuki Murata, "Fundamentals of Thermoelectric Conversion," Advances in Micro-Nano Thermal Engineering, Part 2 "Micro-Nano Phenomena in Thermal Engineering," Chapter 9 "Thermoelectric Conversion," pp. 301-309 (2021))

Next, thermoelectric conversion elements are roughly classified into vertical thermoelectric conversion elements and horizontal thermoelectric conversion elements.
A longitudinal thermoelectric conversion element is one in which the thermoelectric conversion phenomenon occurs in the same direction as the electric current and heat flow, whereas a transverse thermoelectric conversion element is one in which the thermoelectric conversion phenomenon occurs in perpendicular directions as the electric current and heat flow.
(Definition of technical terms related to magneto-thermoelectric effect)
Here, the magneto-thermoelectric effect is a thermoelectric effect that depends on an external magnetic field or spontaneous magnetization of a magnetic material. Figure 1 is a diagram that systematically classifies magneto-thermoelectric effects, where (A) is the electron transport phenomenon resulting from an external magnetic field, and (B) is the electron transport phenomenon resulting from spontaneous magnetization, divided into longitudinal and transverse effects, input is divided into electric current and heat current, and output is divided into electric current and heat current, with the name of the phenomenon given to each column. (See Kenichi Uchida, "Conversion Phenomena of Heat Current, Electric Current, and Spin Current," Progress in Micro-Nano Thermal Engineering, Part 1 "Basic Theory of Micro-Nano Thermal Engineering," Chapter 3 "Energy Transport Phenomena in Solids," pp. 62-71 (2021))

In electron transport phenomena caused by an external magnetic field, the thermoelectric conversion phenomenon in which the input is heat flow and the output is electric current due to the longitudinal effect is called the magnetic Seebeck effect, the thermoelectric conversion phenomenon in which the input is electric current and the output is heat flow due to the longitudinal effect is called the magnetic Peltier effect, the thermoelectric conversion phenomenon in which the input is heat flow and the output is electric current due to the transverse effect is called the normal Nernst effect, and the thermoelectric conversion phenomenon in which the input is electric current and the output is heat flow due to the transverse effect is called the normal Ettingshausen effect.
In the electron transport phenomenon resulting from spontaneous magnetization, the thermoelectric conversion phenomenon in which the input is a heat flow and the output is a current due to the longitudinal effect is called the anisotropic magnetic Seebeck effect, the thermoelectric conversion phenomenon in which the input is a current and the output is a heat flow due to the longitudinal effect is called the anisotropic magnetic Peltier effect, the thermoelectric conversion phenomenon in which the input is a heat flow and the output is a current due to the transverse effect is called the anomalous Nernst effect, and the thermoelectric conversion phenomenon in which the input is a current and the output is a heat flow due to the transverse effect is called the anomalous Ettingshausen effect.

In FIG. 1, even in cases where the input and output are the same electric current or heat flow, rather than the thermoelectric conversion phenomenon, the names of the phenomena are given in each column.
In electron transport phenomena caused by external magnetic fields, the phenomenon in which the input is electric current and the output is also electric current due to the longitudinal effect is called the magnetoresistance effect, and the phenomenon in which the input is heat flow and the output is also heat flow due to the longitudinal effect is called the magnetothermal resistance effect (or the McGehee-Leguier-Duc effect). The phenomenon in which the input is electric current and the output is also electric current due to the transverse effect is called the normal Hall effect, and the phenomenon in which the input is heat flow and the output is also heat flow due to the transverse effect is called the thermal Hall effect (or the Leguier-Duc effect).
In electron transport phenomena resulting from spontaneous magnetization, the phenomenon in which the input is electric current and the output is also electric current due to the longitudinal effect is called the anisotropic magnetoresistance effect, the phenomenon in which the input is heat flow and the output is also heat flow due to the longitudinal effect is called the anisotropic magnetothermal resistance effect, the phenomenon in which the input is electric current and the output is also electric current due to the transverse effect is called the anomalous Hall effect, and the phenomenon in which the input is heat flow and the output is also heat flow due to the transverse effect is called the anomalous thermal Hall effect.

Next, focusing on the thermoelectric generation phenomenon, the magnetic Seebeck effect will be explained as the longitudinal effect of the electron transport phenomenon resulting from the external magnetic field in Fig. 1A, and the anisotropic magnetic Seebeck effect will be explained as the longitudinal effect of the electron transport phenomenon resulting from the spontaneous magnetization in Fig. 1B. In addition, the normal Nernst effect will be explained as the transverse effect of the electron transport phenomenon resulting from the external magnetic field in Fig. 1A, and the anomalous Nernst effect will be explained as the transverse effect of the electron transport phenomenon resulting from the spontaneous magnetization in Fig. 1B.
The magneto-Seebeck effect is an electron transport phenomenon caused by an external magnetic field, in which the thermoelectric power generated in the same direction as the temperature gradient changes depending on the magnetic field strength. See Felix Spathelf et al., “Magneto-Seebeck effect in bismuth”, Physical Review B 105, 235116 (2022). It usually shows an even dependence on the magnetic field direction.
The anisotropic magneto-Seebeck effect is an electron transport phenomenon resulting from spontaneous magnetization, in which the thermoelectric power generated in the same direction as the temperature gradient varies depending on the relative angle between the temperature gradient and magnetization. See Minoru Hashizume et al., “Anisotropic magneto-Seebeck effect in the antiferromagnetic semimetal FeGe ₂ ”, Physical Review B 104,115109 (2021), Ken-ichi Uchida et al., “Observation of anisotropic magneto-Peltier effect in nickel”, Nature 558, 95 (2018). It usually shows an even dependence on the magnetization direction.

The ordinary Nernst effect (ONE) is an electron transport phenomenon resulting from an external magnetic field. It is caused by the Lorentz force acting on conduction electrons and holes, and generates a thermoelectromotive force in the cross product direction of the temperature gradient and the external magnetic field. The ordinary Nernst effect in typical metals generates only a very small thermoelectromotive force, but it is known that Bi-based semimetals and the like exhibit high thermoelectric conversion performance due to the ordinary Nernst effect (for example, in BiSb alloys, ZT>0.3 is reached at 100-200K when an external magnetic field of 1T is applied). However, the drawback is that an external magnetic field must be applied in order to utilize thermoelectric power generation due to the ordinary Nernst effect. It usually shows an odd dependence on the magnetic field direction.

The anomalous Nernst effect (ANE) is an electron transport phenomenon resulting from spontaneous magnetization, in which a thermoelectric power is generated in a magnetic material in the direction of the cross product of the temperature gradient and magnetization. If the magnetization is aligned in one direction, it will work without applying an external magnetic field. The mechanism of the anomalous Nernst effect is different from that of the normal Nernst effect, and originates from a virtual magnetic field derived from the band structure of electrons and impurity scattering dependent on spin. The transverse thermoelectric power (anomalous Nernst coefficient) due to the anomalous Nernst effect in pure metals Fe, Ni, and Co that exhibit ferromagnetism is only about 0.1 μV/K, but in magnetic Heusler alloys such as Co ₂ MnGa, a transverse thermoelectric power of over 6 μV/K derived from the topological electronic structure has been observed in the temperature range including room temperature. It usually shows an odd dependence on the magnetization direction.

Next, focusing on cooling and heating phenomena, the magnetic Peltier effect will be explained as the longitudinal effect of the electron transport phenomenon resulting from the external magnetic field in Fig. 1A, and the anisotropic magnetic Peltier effect will be explained as the longitudinal effect of the electron transport phenomenon resulting from the spontaneous magnetization in Fig. 1B. In addition, the normal Etchingshausen effect will be explained as the transverse effect of the electron transport phenomenon resulting from the external magnetic field in Fig. 1A, and the anomalous Etchingshausen effect will be explained as the transverse effect of the electron transport phenomenon resulting from the spontaneous magnetization in Fig. 1B.
The Magneto-Peltier effect is an electron transport phenomenon caused by an external magnetic field. The heat flow that occurs in the same direction as the input current changes depending on the strength of the magnetic field. It usually shows an even dependence on the magnetic field direction.
The anisotropic magneto-Peltier effect is an electron transport phenomenon resulting from spontaneous magnetization, in which the heat flow that occurs in the same direction as the input current varies depending on the relative angle between the current and magnetization. See Ken-ichi Uchida et al., “Observation of anisotropic magneto-Peltier effect in nickel”, Nature 558, 95 (2018). It usually shows an even dependence on the magnetization direction.

The ordinary Ettingshausen effect (OEE) is an electron transport phenomenon resulting from an external magnetic field, in which a heat flow occurs in the direction of the cross product of the input current and the external magnetic field. The ordinary Ettingshausen effect in typical metals generates only a very small temperature gradient, but it is known that Bi-based semimetals and the like exhibit high thermoelectric conversion performance due to the ordinary Ettingshausen effect (for example, in the case of BiSb alloys, ZT>0.3 is reached at 100-200K when an external magnetic field of 1T is applied). However, the drawback is that an external magnetic field must be applied in order to utilize the ordinary Ettingshausen effect for cooling and heating phenomena. It usually shows an odd dependency on the magnetic field direction.

The anomalous Ettingshausen effect (AEE) is an electron transport phenomenon resulting from spontaneous magnetization, in which a heat flow occurs in the cross product direction of the input current and magnetization in a magnetic material. If the magnetization is aligned in one direction, it will work without applying an external magnetic field. The mechanism of the anomalous Ettingshausen effect is different from that of the normal Ettingshausen effect, and originates from a virtual magnetic field derived from the band structure of electrons and impurity scattering dependent on spin. A large anomalous Ettingshausen effect has been observed in magnetic Heusler alloys such as _Co2MnGa and permanent magnets such as _SmCo5 . It usually shows an odd dependence on the magnetization direction.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.
2A is a block diagram of a lock-in thermography measurement device suitable for measuring the thermoelectric effect by focusing on the cooling and heating phenomena. In the figure, the lock-in thermography measurement device includes a sample 10 whose thermoelectric effect is to be measured, an electromagnet 20, a current source 30, an infrared camera 40, and a signal processing system 50.
The sample 10 is an object to measure the thermoelectric effect, and has a roughly rectangular parallelepiped shape, and its specific shape is an artificially tilted multilayer laminate having different tilt angles in zero magnetic field as shown in Fig. 9A. In a three-dimensional orthogonal coordinate system xyz, the x-axis direction is the longitudinal direction of the sample 10, the y-axis direction is the width direction of the sample 10, and indicates the direction in which the external magnetic field H is applied by the electromagnet 20, and the z-axis direction is the thickness direction of the sample 10, and indicates the imaging direction by the infrared camera 40.
The electromagnet 20 applies a magnetic field H in the width direction y of the cross section of the sample 10, and has a core made of a magnetic material and a winding made of a conductive material wound around the core. Note that a permanent magnet may be used instead of the electromagnet 20, or a permanent magnet may be used together with the electromagnet 20.
The current source 30 applies a current Jc in the longitudinal direction x of the sample 10. In the lock-in thermography measurement, the current source 30 is used to apply an alternating square wave current.
The infrared camera 40 photographs the sample 10 located between the north and south poles of the electromagnet 20 from the z-axis direction. For example, an infrared camera such as ELITE manufactured by FEI or ImageIR manufactured by InfraTec may be used.
The signal processing system 50 performs Fourier analysis using the thermal image signal sent from the infrared camera 40 and the reference signal sent from the current source 30, and extracts the lock-in amplitude signal A and the lock-in phase signal φ. The signal processing system 50 can use a computer as the hardware.

Figure 2B shows the waveforms of the input current and output temperature change signal. The input signal is a square wave periodic current with a time period (1/f) that is the inverse of frequency f, an amplitude Jc, and a duty ratio of 0.5. The output signal is a thermoelectric effect signal synchronized with the input current period, with a finite Joule heat signal that does not change with time superimposed on it. Lock-in thermography allows for the imaging and measurement of only the thermoelectric effect signal by extracting only the temperature change component synchronized with the frequency of the input signal.

Figure 2C is an observation diagram of the amplitude and phase of temperature change measured by a lock-in thermography measurement device. In the figure, the lock-in amplitude signal image A and the lock-in phase signal image φ are planar images of the sample 10, which is the imaging surface of the infrared camera 40. The shading of the planar image of the sample 10 indicates the strength of the amplitude and the degree of phase advance or delay. The repeating pattern in the shading of the planar image is due to the repeated appearance of the material junction interface in the artificially inclined multilayer laminate.

3A is an explanatory diagram of visualization of lateral thermoelectric conversion in a zero magnetic field in an artificially inclined multilayer stack, in which the longitudinal cross section is rectangular. The multilayer stack is stacked at an inclination in the xy plane of the sample 10.
FIG. 3A(a) is a perspective view of the sample. In the three-dimensional orthogonal coordinate system xyz, the x-axis direction is the current axis direction in which an external current is supplied, the y-axis direction is the magnetic axis direction in which an external magnetic field H is applied, and the z-axis direction is the thickness direction of the sample. In this arrangement, the net heat flow direction generated by the transverse thermoelectric effect originating from the inclined structure is the y-axis direction. The magnetic field direction in which the external magnetic field H is applied is not limited to the y-axis direction of the three-dimensional orthogonal coordinate system xyz, and does not need to strictly match the y-axis direction as long as it is a direction in which temperature modulation appears due to the magneto-thermoelectric effect, and may be within an appropriate tolerance, for example, within a range of ±30° with respect to the y-axis direction.

The multilayer stack is stacked at an inclination in the xy plane of the sample 10. In the xz plane of the sample 10, the stack is stacked in parallel. In Fig. 3A(a), sample A is used, in which the inclination angle θ of the multilayer stack in the xy plane is 45°. Indium electrodes are provided on both ends of the multilayer stack in the current axis direction x.
3A(b) shows a steady-state temperature image that appears on the imaging surface (xy plane) when a driving current is passed between a first electrode and a second electrode that are provided as a pair in the x-axis direction of the sample 10. In the xy plane, ₁₂ layers of Bi ₈₈ Sb as the first thermoelectric material and ₃ layers of Bi _0.2 Sb _1.8 Te as the second thermoelectric material are laminated, and when a current of 1 A is steadily passed, the temperature distribution of the Bi ₈₈ Sb ₁₂ layers is 293.5 K to 296 K, and the temperature distribution of the Bi _0.2 Sb _1.8 Te ₃ layers is 298 K to 302 K.

Figure 3B(c) is a lock-in amplitude signal image of the sample, showing the frequency f of the square wave periodic current of 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz, in that order. Areas where the amplitude signal is large (dark areas) indicate large temperature changes, while areas where the amplitude signal is small (light areas) indicate small temperature changes. It can be seen that the lower the frequency f, the greater the temperature change inside the sample.

Figure 3B(d) shows the lock-in phase signal image of the sample, with the frequency f of the square wave periodic current being 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz, in that order. In the steady state, a phase difference of 0° with the reference signal indicates heat generation, and a phase difference of 180° indicates heat absorption. At high frequencies, the phase difference with the reference signal changes spatially and continuously due to the effects of thermal diffusion, while the lower the frequency f, the closer it is to the steady state and the smaller the spatial distribution of the phase signal image φ becomes. At 0.1 Hz, the effects of thermal diffusion remain slight, but it can be seen that cooling occurs in the upper part of the region and heat is generated in the lower part of the region. Furthermore, a thermal distribution reflecting the tilt angle of the tilted layered structure of the element is formed in approximately the center of the element.

Fig. 3B(e) shows the values of the gray-scale signal shown in Fig. 3B(c) near the side surface of the sample, with the horizontal axis representing the position in the longitudinal direction of the sample and the vertical axis representing the amplitude value A, where the frequency f of the square-wave periodic current is 10 Hz, 1 Hz, and 0.1 Hz, respectively. For example, the maximum value of the temperature change amplitude appearing in _{the Bi88Sb12} layer is 0.9 K at 10 Hz, 2.8 K at 1 Hz, and 4.0 K at 0.1 Hz.
Fig. 3B(f) shows the values of the grayscale signal shown in Fig. 3B(d) near the side surface of the sample, with the horizontal axis representing the position in the longitudinal direction of the sample and the vertical axis representing the phase value φ, with the frequency f of the square wave periodic current being 10 Hz, 1 Hz, and 0.1 Hz in order. At 10 Hz and 1 Hz, it can be seen that the phase difference from the reference signal changes continuously due to the influence of thermal diffusion that has not yet reached a steady state. At 0.1 Hz, the phase difference is 180° over almost the entire region, indicating that heat absorption is occurring.

Figure 3B(g) shows the Aave(K) value, which is the average amplitude value for each frequency f shown in Figure 3B(e) within the range of one bonded pair of different materials. The average amplitude values _Aave (K) are _0.2K at 10 Hz, 0.3K at 5 Hz, 0.5K at 2 Hz, 1.0K at 1 Hz, 1.3K at 0.5 Hz, 1.6K at 0.2 Hz, and 1.6K at 0.1 Hz.
Fig. 3B(h) shows the average value φ _ave (deg) of the phase value for each frequency f shown in Fig. 3B(f) within the range of one pair of junctions of different materials. The average phase value φ _ave (deg) is 260° at 10 Hz, 260° at 5 Hz, 250° at 2 Hz, 240° at 1 Hz, 220° at 0.5 Hz, 200° at 0.2 Hz, and 190° at 0.1 Hz.

3C and 3D are explanatory diagrams visualizing horizontal thermoelectric conversion in an artificially inclined multilayer stack in zero magnetic field, showing the case where a net heat flow Jq is generated flowing from the top surface to the bottom surface in the z-axis direction.
3C(i) is a perspective view of the sample. In the three-dimensional orthogonal coordinate system xyz, the x-axis direction is the current axis direction to which an external current is supplied, the y-axis direction is the magnetic axis direction indicating the magnetic field direction to which an external magnetic field H is applied, and the z-axis direction is the thickness direction of the sample. In this arrangement, the net heat flow direction generated by the thermoelectric effect is the z-axis direction. The magnetic field direction to which the external magnetic field H is applied is not limited to the y-axis direction of the three-dimensional orthogonal coordinate system xyz, and does not need to strictly match the y-axis direction as long as it is a direction in which temperature modulation appears due to the magneto-thermoelectric effect, and it is sufficient if it is within an appropriate tolerance range, for example, within ±30° with respect to the y-axis direction.
The multilayer stack is stacked at an inclination in the xz plane of the sample 10. In the xy plane of the sample 10, the stack is stacked in parallel. In Fig. 3C(i), sample A is used, in which the inclination angle θ with respect to the bottom surface of the multilayer stack is 45°. Indium electrodes are provided on both ends of the multilayer stack in the current axis direction x.

3C(j) shows a steady-state temperature image that appears on the imaging surface (xy plane) in the heat flow direction when a driving current is passed between a first electrode and a second electrode that are provided as a pair in the x-axis direction of the sample 10. In the xy plane, a _12- layer Bi ₈₈ Sb as a first thermoelectric material and _{a 3-} layer Bi _0.2 Sb _1.8 Te as a second thermoelectric material are laminated, and when a current of 1 A is steadily passed, the temperature distribution of the Bi ₈₈ Sb _12- layer is 295 K to 296 K, and the temperature distribution of the Bi _0.2 Sb _1.8 Te _3- layer is 298 K to 300 K.

Fig. 3D(k) is a lock-in amplitude signal image of the sample, which shows the frequency f of the square wave periodic current of 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz in order. It can be seen that the lower the frequency f, the larger the temperature change in the sample.
Fig. 3D(l) shows the lock-in phase signal image φ of the sample, which shows the square wave periodic current frequency f of 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz in order. The lower the frequency f, the smaller the change in the phase signal image φ becomes, and at 0.1 Hz, the region around 180° is dominant. This indicates that the entire sample surface is cooled in a state close to the steady state.

3D(k) and 3D(l), the position where the maximum value of the temperature change amplitude appears in the Bi ₈₈ Sb ₁₂ layer is examined. The maximum value appears in the Bi ₈₈ Sb ₁₂ layer in a region where the phase signal image φ is near 180°. That is, in the Bi ₈₈ Sb ₁₂ layer, the maximum value of the temperature change amplitude occurs in a part of the region where the phase signal image φ in the current axis direction x is near 180°, and for example, the maximum value is reached in a region of one-quarter to one-half of the width of the Bi ₈₈ Sb ₁₂ layer in the current axis direction x. On the other hand, no dependency of the temperature change distribution is observed in the magnetic axis direction y axis.

Fig. 3D(m) shows the gray-scale signal shown in Fig. 3D(k) with the horizontal axis representing the position in the longitudinal direction of the sample and the vertical axis representing the amplitude value A, where the frequency f of the square-wave periodic current is 10 Hz, 1 Hz, and 0.1 Hz, respectively. For example, the maximum value of the temperature change amplitude appearing in _{the Bi88Sb12} layer is 1.0 K at 10 Hz, 3.5 K at 1 Hz, and 5.5 K at 0.1 Hz.
Fig. 3D(n) shows the grayscale signal shown in Fig. 3D(l) with the horizontal axis representing the longitudinal direction of the sample and the vertical axis representing the phase value φ, and shows the frequency f of the square wave periodic current of 10 Hz, 1 Hz, and 0.1 Hz in order. At 10 Hz and 1 Hz, it can be seen that the phase difference from the reference signal changes continuously due to the influence of thermal diffusion that has not yet reached a steady state. At 0.1 Hz, the phase difference is 180° over almost the entire region, indicating that heat absorption is occurring.

Figure 3D(o) shows the average Aave(K) of the amplitude values for each frequency f of the current shown in Figure 3D(m) within the range of one pair of junctions of different materials. The average amplitude values _Aave (K) are 0.2K at 10Hz, 0.4K at 5Hz, 0.7K at 2Hz, 1.2K at 1Hz, 1.8K at 0.5Hz, 2.1K at 0.2Hz _, and 2.5K at 0.1Hz.
Fig. 3D(p) shows the average value φ _ave (deg) of the phase value for each frequency f shown in Fig. 3D(n) within the range of one pair of junctions of different materials. The average phase value φ _ave (deg) is 260° at 10 Hz, 260° at 5 Hz, 250° at 2 Hz, 240° at 1 Hz, 220° at 0.5 Hz, 200° at 0.2 Hz, and 190° at 0.1 Hz.

The samples shown in Figures 3A(a) and 3C(i) were fabricated by the spark plasma sintering method using _{alternating Bi88Sb12} / _{Bi0.2Sb1.8Te3 multilayer} laminates, with each material having a thickness of approximately 1 mm _and an applied current amplitude of 1 A (except for Figure 8).
Lock-in thermography reveals how localized heat release and absorption at the interface between different materials acts as a lateral thermoelectric conversion.

The modulation due to the magneto-thermoelectric effect is measured by performing lock-in thermography measurement with the application of a magnetic field H. Figure 4 shows the results of separating a lock-in thermography image into symmetric and antisymmetric magnetic field components, where Fig. 4(a) shows the original image when the applied magnetic flux density μ ₀ H is +0.8 T, Fig. 4(b) shows the original image when the applied magnetic flux density μ ₀ H is −0.8 T, Fig. 4(c) shows the symmetric magnetic field component when the applied magnetic flux density strength μ ₀ |H| is 0.8 T, and Fig. 4(d) shows the antisymmetric magnetic field component when the applied magnetic flux density strength μ ₀ |H| is 0.8 T, and each shows a pair of lock-in amplitude signal images and lock-in phase signal images of the sample. Note that the frequency f of the periodic square wave current is 1.0 Hz.

The magnetic field symmetric components are composed of an amplitude value A _even (K) and a phase value φ _even (deg) and are extracted by complex-summing the lock-in thermography image signal at a negative applied magnetic field with the lock-in thermography image signal at a positive applied magnetic field and dividing the summed image signal by two.

The magnetic field antisymmetric component is composed of an amplitude value A _odd (K) and a phase value φ _odd (deg) and is extracted by complex subtracting the lock-in thermography image signal at a negative applied magnetic field from the lock-in thermography image signal at a positive applied magnetic field and dividing the subtracted image signal by two.

FIG. 4(a) shows the amplitude value A(K) and phase value φ(deg) of the original image in which the magnetic field symmetric component and the magnetic field antisymmetric component are not sharply distinguished. The image of the amplitude value A(K) of the original image shows that the maximum value of the temperature change amplitude appearing in the Bi ₈₈ Sb ₁₂ layer is 4.0 K, similar to the case of 1 Hz in FIG. 3D(k). The maximum value occurs in a region from one-quarter to one-half of the width of the Bi ₈₈ Sb ₁₂ layer in the current axis direction x, but the temperature change occurs in the entire region of the Bi ₈₈ Sb ₁₂ layer. On the other hand, no dependency of the temperature change distribution is observed in the magnetic axis direction y. The image of the phase value φ(deg) of the original image shows that the phase signal image φ alternates between regions of about 0° and about 180°, similar to the case of 1 Hz in FIG. 3D(l).
FIG. 4B shows the amplitude value A (K) and phase value φ (deg) of an original image in which the magnetic field symmetric component and the magnetic field antisymmetric component are not sharply distinguished, and is similar to FIG. 4A.

FIG. 4(c) shows the amplitude value A _even (K) and the phase value φ _even (deg), which are the symmetric components of the magnetic field, and is similar to FIG. 4(a).
4(d) shows the amplitude value A _odd (K) and phase value φ _odd (deg) of the magnetic field antisymmetric component, and the maximum value of the temperature change amplitude appearing in the Bi ₈₈ Sb ₁₂ layer is 0.3 K. Also, the temperature change occurs in the entire region of the Bi ₈₈ Sb ₁₂ layer, and the region where the maximum value of the temperature change amplitude occurs is about half the width of the Bi ₈₈ Sb ₁₂ layer in the current axis direction x. On the other hand, no dependency of the temperature change distribution is observed in the magnetic axis direction y. The value of the phase signal image φ appearing in the _{Bi 88} Sb ₁₂ layer is near 180°.

FIG. 5 is an explanatory diagram of the contribution of the magnetic Peltier effect, and the area surrounded by a white rectangle in the thermal image corresponds to the boundary between Bi ₈₈ Sb ₁₂ and Bi _0.2 Sb _1.8 Te _3. The magnetic field symmetric component shows the contribution of the magnetic Peltier effect in the artificially inclined multilayer laminate. FIG. 5(a) is an amplitude signal image of the magnetic field symmetric component, FIG. 5(b) is a phase signal image, and FIG. 5(c) is a graph showing the magnetic field dependence of the amplitude value A _even (K) of the magnetic field symmetric component, where the horizontal axis is the magnetic field μ ₀ |H| and the vertical axis is the amplitude value A _even (K). FIG. 5(d) is a graph showing the magnetic field dependence of the phase value φ _even (deg) of the magnetic field symmetric component, where the horizontal axis is the magnetic flux density μ ₀ |H| and the vertical axis is the phase value φ _even (deg).

FIG. 5(a) is an amplitude signal image of the symmetric components of the magnetic field, and is a reprint of FIG. 4(c).
FIG. 5(b) is a phase signal image of the symmetric components of the magnetic field, and is a reprint of FIG. 4(c).
In FIG. 5(c), when the magnetic field μ ₀ |H| is 0.0T, 0.2T, 0.4T, 0.6T, 0.8T, and 1.0T, the corresponding amplitude values A _even (K) are 3.3K, 3.4K, 3.5K, 3.7K, 3.8K, and 3.9K, respectively.
In FIG. 5D, when the magnetic field μ ₀ |H| is 0.0T, 0.2T, 0.4T, 0.6T, 0.8T, and 1.0T, the corresponding phase value φ _even (deg) is approximately the same value of 210°.
Thus, _the temperature _modulation signal near _{the Bi88Sb12} / _{Bi0.2Sb1.8Te3} interface increases with increasing applied magnetic field due to the magnetic Peltier effect.

Fig. 6 is an explanatory diagram of the contribution of the normal etching Shausen effect, and Fig. 6(a) is a lock-in amplitude signal image A _odd (K) of a sample, which shows the frequency f of the square wave periodic current of 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz in order. The lower the frequency f, the higher the temperature amplitude. For example, the maximum value of the temperature change amplitude appearing in the Bi ₈₈ Sb ₁₂ layer is 0.090 K at 10 Hz, 0.13 K at 5 Hz, 0.25 K at 1 Hz, 0.3 K at 0.5 Hz, and 0.4 K at 0.1 Hz.
Figure 6(b) shows the lock-in phase signal image φ _odd (K) of the sample, with the frequency f of the square wave periodic current being 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz, in that order. As the frequency f decreases, the image approaches a steady state and the spatial distribution of the phase signal image φ decreases. At 0.1 Hz, the effect of thermal diffusion remains slight, but the phase is close to 180° over the entire sample surface, indicating that the sample is cooled.

6(c) is a graph showing the frequency dependence of the amplitude value A _odd (K) of the antisymmetric magnetic field component, where the horizontal axis is frequency f and the vertical axis is amplitude value A _odd (K). The amplitude value A _odd (K) is 0.07 K at 10 Hz, 0.1 K at 5 Hz, 0.15 K at 2 Hz, 2.0 K at 1 Hz, 0.24 K at 0.5 Hz, 0.28 K at 0.2 Hz, and 0.32 K at 0.1 Hz.
6(d) is a graph showing the frequency dependence of the phase value φ _odd (deg) of the antisymmetric magnetic field component, with the horizontal axis being frequency f and the vertical axis being phase value φ _odd (deg). The phase value φ _odd (deg) is 240° at 10 Hz, 240° at 5 Hz, 230° at 2 Hz, 220° at 1 Hz, 210° at 0.5 Hz, 200° at 0.2 Hz, and 190° at 0.1 Hz.

6(e) is a graph showing the magnetic field dependence of the amplitude value A _odd (K) of the antisymmetric magnetic field component, where the horizontal axis is the magnetic field μ ₀ |H| and the vertical axis is the amplitude value A _odd (K). When the magnetic field μ ₀ |H| is 0.2T, 0.4T, 0.6T, 0.8T, and 1.0T, the corresponding amplitude values A _odd (K) are 0.12K, 0.17K, 0.20K, 0.21K, and 0.22K, respectively.
6(f) is a graph showing the magnetic field dependence of the phase value φ _odd (deg) of the antisymmetric magnetic field component, where the horizontal axis is the magnetic field μ ₀ |H| and the vertical axis is the phase value φ _odd (deg). When the magnetic field μ ₀ |H| is 0.2T, 0.4T, 0.6T, 0.8T, and 1.0T, the corresponding phase value φ _odd (deg) shows approximately the same value of 210°.

The magnetic field antisymmetric component indicates the contribution of the normal Etchingshausen effect in the artificially graded multilayer stack. The contribution of the normal Etchingshausen effect appears mainly in the Bi ₈₈ Sb ₁₂ layer and increases with increasing magnetic field H. The magnetic field dependence of the normal Etchingshausen effect contribution is consistent with the magnetic field dependence of the thermoelectric power due to the normal Nernst effect of Bi ₈₈ Sb ₁₂ (see Fig. 6(e)).

7(a) is a graph showing the average amplitude value A ave (K) for sample A, in which the inclination angle θ with respect to the bottom surface of the multilayer laminate is 45°, when the frequency f of the square wave periodic current is 0.1 Hz, where the horizontal axis is _the magnetic field μ ₀ |H| and the vertical axis is the average amplitude value A _ave (K). When the magnetic field μ ₀ |H| increases from -1.0 T to 1.0 T, the corresponding amplitude value A _ave (K) increases from 2.2 K to 2.68 K.
7(b) is a graph showing the rate of change δ of temperature change due to horizontal thermoelectric conversion caused by application of a magnetic field when the frequency f of the square wave periodic current is 0.1 Hz for sample A, in which the inclination angle θ with respect to the bottom surface of the multilayer laminate is 45°, and sample B, in which the inclination angle θ is 21°, with the horizontal axis being the magnetic field μ ₀ |H|. When the magnetic field μ ₀ |H| increases from -1.0 T to 1.0 T, the corresponding rate of change δ for sample A increases from -8% to 12%, and the corresponding rate of change δ for sample B increases from -9% to 15%.

Here, the rate of change δ of temperature change due to horizontal thermoelectric conversion caused by application of a magnetic field is calculated from the ratio of the amplitude value A _ave when a zero magnetic field is applied to the value when a magnetic field of 0.8 T is applied, and is expressed by the following formula.
δ = [ A _ave (0.8T) - A _ave (0T) ] / A _ave (0T)
The average amplitude A _ave of the Bi ₈₈ Sb ₁₂ /Bi _0.2 Sb _1.8 Te ₃ artificially graded multilayer stack can be enhanced by applying a magnetic field due to the superposition of the structure-derived signal, the magnetic Peltier effect, and the normal Etchingshausen effect.
The rate of change δ of the magnetic field-dependent current-induced temperature difference can be adjusted by changing the tilt angle θ, and its value also depends on the magnetic field.

8(a) is an explanatory diagram of thermoelectric cooling in a steady state, with the horizontal axis representing the constant current I applied in the direction of the current axis x of the sample, and the vertical axis representing the average temperature change ΔT in the steady state. For sample B, when the magnetic field μ ₀ |H| is -0.8 T and the constant current I is 4 A, the maximum average temperature change ΔT is -4 K. When the magnetic field μ ₀ |H| is +0.8 T and the constant current I is 5 A, the maximum average temperature change ΔT is -6 K.
8(b) is an explanatory diagram in which the horizontal axis represents the constant current |I| applied in the current axis direction x of the sample, and the vertical axis represents the temperature change ΔT _TE due to the thermoelectric effect. In the thermoelectric effect of sample B, when the magnetic field μ ₀ |H| is +0.8 T and the constant current I is 7 A, the maximum average temperature change ΔT _TE is -19 K. When the magnetic field μ ₀ |H| is -0.8 T and the constant current I is 7 A, the maximum average temperature change ΔT _TE is -16 K. In the thermoelectric effect, the temperature change ΔT _TE increases in proportion to the increase in the constant current I.
8(c), the horizontal axis is the constant current |I| applied to the sample in the current axis direction x, and the vertical axis is the temperature change ΔT _J due to Joule heating. For sample B, Joule heating does not depend on the sign of the magnetic field, and when the magnetic field μ ₀ |H| is ±0.8 T, a temperature change ΔT _J of +14 K is obtained when the constant current I is 7 A. With Joule heating, the temperature change ΔT _J increases in proportion to the square of the constant current I.

Here, the measurement points in the figure are indicated by ◆ for the steady-state temperature change ΔT when the magnetic flux density μ ₀ H is +0.8 T, and by △ for the steady-state temperature change ΔT when the magnetic flux density μ ₀ H is −0.8 T. The following equations hold for the steady-state temperature changes ΔT _TE and ΔT _J from room temperature due to the thermoelectric effect or Joule heat.
Here, ΔT(+I) is the average temperature change ΔT when a positive current is applied, and ΔT(−I) is the average temperature change ΔT when a negative current is applied.

Steady-state thermoelectric cooling is _possible with _Bi88Sb12 / _{Bi0.2Sb1.8Te3} artificially _graded multilayer stacks. Steady-state cooling can be strengthened by applying a positive magnetic field, and as a result of the competition between thermoelectric cooling and Joule heating, a maximum temperature change _of about -6K can be obtained at Jc = +5A. Since the magnitude relationship between thermoelectric cooling and Joule heating depends on the thermal boundary conditions of the element, it is possible to obtain greater thermoelectric cooling even with the same element by optimizing the thermal design.

In the embodiment shown in FIG. 9, _three layers of Bi _0.2 Sb _1.8 Te are replaced with a permanent magnet, Nd ₂ Fe ₁₄ B-based magnet, and the sample is an artificially inclined multilayer laminate of ₁₂ layers of Bi ₈₈ Sb as the first thermoelectric material and a Nd ₂ Fe ₁₄ B-based magnet layer as the second thermoelectric material.
9A shows a perspective view of the structure of a permanent magnet-based artificially inclined multilayer stack for horizontal thermoelectric conversion. The artificially inclined multilayer stack is inclined at an inclination angle θ in the xz plane of the sample 10. The generated heat flow Jq is in the z-axis direction, which is a direction that flows from the top surface to the bottom surface. The magnetization M is the magnetization direction of the Nd ₂ Fe ₁₄ B magnet layer, and is oriented in the normal direction of the stacking surface of the artificially inclined multilayer stack. Due to the magnetization M of the _{Nd 2} Fe ₁₄ B magnet layer, a magnetic field H is applied to the Bi ₈₈ Sb ₁₂ layer without using an electromagnet or the like.

Figure 9B shows the amplitude and phase of the current-induced temperature change in the permanent magnet-based artificially inclined multi-layer stack for horizontal thermoelectric conversion. Figure 9B(c) shows the side surface of Figure 9A, and Figure 9B(d) shows the top surface of Figure 9A.
The upper and lower left figures in Fig. 9B(c) are lock-in amplitude signal images of _{the Bi88Sb12} /NdFeB magnet artificially inclined multilayer laminate sample for the side surface in Fig. 9A, showing the results when the frequency f of the square wave periodic current is 10 Hz and 1.0 Hz, respectively. The maximum temperature change amplitude appearing in _{the Bi88Sb12} layer is 120 mK at 10 Hz and 220 mK at 1 Hz.
The upper right and lower right drawings in Fig. 9B(c) are images of the lock-in phase signal of _{the Bi88Sb12} /NdFeB magnet artificially inclined multilayer laminate with respect to the side surface of Fig. 9A, showing the frequency f of the square wave periodic current of 10 Hz and 1.0 Hz in order. Due to the effect of thermal diffusion, the phase difference with the reference signal changes spatially and continuously.

9B(c), the temperature change amplitude is maximum near _{the Bi88Sb12} /NdFeB magnet interface. In the low frequency region, the temperature amplitude is small in the approximate center of the inclined region in the xy plane of the sample 10, and is large near both edges of the y axis.

The upper and lower left figures in Fig. 9B(d) are images of the lock-in amplitude signal of the _Bi88Sb12 /NdFeB magnet artificially inclined multilayer laminate against the upper surface of Fig. _9A , showing the square wave periodic current frequency f of 10 Hz and 1.0 Hz, respectively. The maximum value of the temperature change amplitude appearing in _{the Bi88Sb12} layer is 120 mK at 10 Hz and 220 mK at 1 Hz.
The upper right and lower right drawings in Fig. 9B(d) are images of the lock-in phase signal of _{the Bi88Sb12} /NdFeB magnet artificially inclined multilayer laminate with respect to the upper surface of Fig. 9A, showing the frequency f of the square wave periodic current of 10 Hz and 1.0 Hz in order. Due to the effect of thermal diffusion, the phase difference with the reference signal changes spatially and continuously.

9B(d), the temperature change amplitude shows a maximum value near _{the Bi88Sb12} /NdFeB magnet interface. For example, the maximum temperature change is obtained in a region from one-quarter to one-half of the width of _{the Bi88Sb12} layer in the current axis direction x. On the other hand, no dependence of the temperature change distribution is observed in the magnetic axis direction y.

FIG. 9C shows the magnetic properties and horizontal thermoelectric conversion properties of a Bi ₈₈ Sb ₁₂ /Nd ₂ Fe ₁₄ B-based artificially inclined multilayer magnet for horizontal thermoelectric conversion.
9C(e) is a hysteresis curve for _the same _Nd2Fe14B magnet used in the artificially inclined multilayer laminate, with the horizontal axis representing the applied magnetic field _μ0H and the vertical axis representing the magnetization _μ0M of the sample, _and a loop curve is drawn in which the saturation magnetization is ±1.2 T within a range of ± _2.5 T for the applied magnetic field _μ0H . Thus, it can be seen that the _Bi88Sb12 / _Nd2Fe14B magnet artificially inclined multilayer laminate for horizontal thermoelectric conversion has a large coercive force and residual magnetization.
FIG. 9C(f) shows the average amplitude value of the temperature change in the demagnetized and magnetized state of the Nd ₂ Fe ₁₄ B permanent magnet. The magnetization direction is perpendicular to the lamination direction of the artificially inclined multi-layer laminate as shown in FIG. 9A. The temperature change in FIG. 9C(f) is the result of measuring the upper surface of FIG. 9A, and the horizontal axis shows the frequency (Hz) and the vertical axis shows the average amplitude value (mK) of the temperature change. All the results shown in FIG. 9 show the case where no magnetic field is applied, that is, the applied magnetic flux density μ ₀ H is 0 T. For the frequencies f of the square wave periodic current of 10 Hz, 5 Hz, 2 Hz, 1.0 Hz, and 0.5 Hz, the average amplitude value (mK) of the temperature change in the unmagnetized state is 25, 38, 70, 105, and 130 mK, respectively. In contrast, the results for are shown in order. The average amplitude values (mK) of temperature change in the magnetized state were 26, 40, 75, 110, and 145 mK, respectively.
9C(g), the horizontal axis shows frequency ( _Hz ) and the vertical axis shows the rate of change δ of the average amplitude of a _Nd2Fe14B permanent magnet in a magnetized state and in a generated state. When the frequency f of the square wave periodic current is 10 Hz, 5 Hz, 2 Hz, 1.0 Hz, and 0.5 Hz, the rate of change δ is 8, 8, 8, 8, and 9%, respectively.

9A to 9C are based on a combination of Nd ₂ Fe ₁₄ B magnet and Bi ₈₈ Sb ₁₂ , and since Bi ₈₈ Sb ₁₂ shows a large magneto-thermoelectric effect, even though measurements were taken without applying a magnetic field when Nd ₂ Fe ₁₄ B was demagnetized and magnetized, the signal was clearly enhanced in the magnetized state due to the magneto-thermoelectric effect, realizing hybrid horizontal thermoelectric conversion. From Fig. 9C(g), it can be seen that the temperature change signal increased by about 8% due to magnetization.
By constructing a permanent magnet-based artificially inclined multilayer stack for lateral thermoelectric conversion in this way, it is possible to enhance the output by utilizing magneto-thermoelectric effects (such as the magnetic Peltier effect and the normal and anomalous Etchingshausen effects) while still taking advantage of the benefits of incorporating permanent magnets.

Next, a method for fabricating a hybrid magneto-thermoelectric element based on an artificially graded multi-layer stack is described.
Both the Bi ₈₈ Sb ₁₂ /Bi _0.2 Sb _1.8 Te ₃ artificially inclined multilayer laminate and the Bi ₈₈ Sb ₁₂ /Nd ₂ Fe ₁₄ B-based magnet artificially inclined multilayer laminate used in the examples were produced by spark plasma sintering. Here, spark plasma sintering (SPS) is a processing method that sinters, joins, and synthesizes workpieces by mechanical pressure and pulse current heating. In addition to the thermal and mechanical energy used in general sintering, electromagnetic energy by pulse current, self-heating of the workpiece, and discharge plasma energy generated between particles are used as the driving force for sintering in a composite manner.

In the above embodiment, a temperature change signal is generated by inputting a current and measuring the temperature change signal. However, a hybrid horizontal thermoelectric power generation can also be performed by inputting a temperature gradient according to a similar principle.
10 _shows the results _of measurements _of the load current I _load dependence of the thermoelectromotive force V _{out and} the output power P _out of the _{Bi88Sb12./Bi0.2Sb1.8Te3} artificially graded multilayer stack at various magnetic field values, where (a) shows the measurement results for sample A at ΔT of 7.9 K and (b) shows the measurement results for sample B at ΔT of 9.8 K. The output power P _out was estimated from the following equation:
It was shown that the output power P _out was increased by applying a positive magnetic field, and the power generation output could also be improved by the magneto-thermoelectric effect.

In the above embodiment, the hybrid lateral thermoelectric temperature modulation element structure in the artificially inclined multilayer laminate is described, in which the first thermoelectric material layer is a Bi _100-x Sb _x layer (0≦x≦50) and the second thermoelectric material layer is a Bi _2-y Sb _y Te ₃ layer (0≦y≦2). However, the present invention is not limited to the above embodiment, and other combinations of materials may be used.
In this case, the vertical magneto-thermoelectric effect should show a thermoelectric change of 5 μV/K or more, for example, Bi and Bi-based alloys. Also, the horizontal magneto-thermoelectric effect should show a thermoelectric change of 5 μV/K or more, for example, Bi, Bi-based alloys, Co ₂ MnGa, Co ₂ MnAl, and SmCo _5. As a result of combining the above, it is preferable that the figure of merit as a hybrid horizontal thermoelectric temperature modulation element is 0.1 or more.

Here, the individual thermoelectric power, electrical resistivity ρ, and thermal conductivity κ of the first thermoelectric material layer and the second thermoelectric material layer used in the hybrid lateral thermoelectric temperature modulation element are measured as follows.
Regarding the measurement of thermoelectric power, for example, the measurement method of thermoelectric power specified in JIS R 1650-1 Measurement method of fine ceramic thermoelectric materials Part 1: Thermoelectric power or the measurement method disclosed in Japanese Patent No. 6202580 (WO2015/025586A1) may be used.
The electrical resistivity ρ may be measured using a DC four-terminal method or an AC two-terminal method.The thermal conductivity κ may be measured using, for example, a steady-state method in which a steady temperature gradient is applied to the sample and the temperature at each position on the sample is measured using a thermocouple or thermography, or a method in which the thermal diffusivity, specific heat capacity, and density are measured using a laser flash method, a differential scanning calorimeter, and an Archimedes method, respectively, and then calculated by taking the first-order product.
The thermal conductivity κ may be calculated, for example, by the following formula described in the section on calculation of thermal conductivity in JIS R 1650-3, Measurement methods for fine ceramic thermoelectric materials, Part 3: Thermal diffusivity, specific heat capacity, and thermal conductivity.

The thermoelectric power and figure of merit of the artificially graded multilayer laminate are calculated as follows, without considering the magneto-thermoelectric effect. The Seebeck tensor is expressed as follows when the basis vectors are taken in the layer-parallel direction (//) and the layer-perpendicular direction (⊥), which are mutually orthogonal. At this point, the off-diagonal terms of the Seebeck tensor are 0.
This material is cut into a rectangle of length l and thickness d so that the angle between the layered structure is θ, as shown in Figure 11A, to create an artificially inclined multilayer laminate. The directions of each side are the x-axis and z-axis, respectively, and when a temperature difference ΔT is applied across the thickness d in the z-axis direction, the electromotive force generated between both ends of the element in the x-axis direction is ΔV _x . When the Seebeck tensor is rewritten along the x-axis and z-axis, which are the directions of input and output, its transverse thermoelectric power S _xz is expressed by the following equation and takes a finite value.
Now, if we calculate the electromotive force using equation (1), we get
It becomes.

The dimensionless figure of merit for the transverse thermoelectric effect in an artificially graded multilayer stack is defined as follows:
Here, the electrical resistivity ρ _xx and the thermal conductivity κ _zz may be estimated in accordance with, for example, the formulas (3) to (9) described in H. Zhou et al., “Geometrical Optimization and Transverse Thermoelectric Performances of Fe / Bi ₂ Te _2.7 Se _0.3 Artificially Tilted Multilayer Thermoelectric Devices” Micromachines 13, 233 (2022).

<Giant anisotropy in BiSb/BiSbTe laminates>
The first thermoelectric material (BiSb: _Bi88Sb12 ) _and the second thermoelectric material (BiSbTe: _{Bi0.2Sb1.8Te3} ) used in this study are polycrystalline, and each material itself is isotropic. However, when a laminate is formed as shown in Figure 11A and an equivalent circuit is calculated, the macroscopic thermoelectric properties of the laminate become anisotropic in _the layer parallel direction ( _// ) and layer perpendicular direction (⊥) as shown below.

Here, t is the thickness ratio of BiSbTe, _tBiSbTe /( _tBiSb + _tBiSbTe ). The thermoelectric properties of BiSb and BiSbTe used here are significantly different from each other, as shown in Table 1. This results in a large anisotropy in the thermoelectric properties of the laminate.
Here, σ, κ, S _S , S _N , Z _S T, and Z _N T are electrical conductivity, thermal conductivity, Seebeck coefficient, Nernst coefficient, figure of merit due to Seebeck effect, and figure of merit due to Nernst effect, respectively. Onsager's reciprocity theorem holds between the Seebeck coefficient and the Peltier coefficient, and between the Nernst coefficient and the Ettingshausen coefficient, and the Peltier coefficient and the Ettingshausen coefficient are obtained by multiplying the Seebeck coefficient and the Nernst coefficient by the absolute temperature, respectively.

In the hybrid lateral thermoelectric temperature modulation element of the present invention, the thermal conductivity of the entire artificially inclined multilayer laminate can be calculated from formula (16), and the electrical conductivity σ, which is the reciprocal of the electrical resistivity ρ, can be calculated from formula (15). It is only necessary to consider the magnetic field and magnetization dependence of the thermal conductivity and electrical conductivity, and the formulas themselves are the same.
On the other hand, for the transverse thermoelectric power S _xz , the contribution of the magnetic Seebeck-Peltier effect can be taken into account as the magnetic field/magnetization dependency of the Seebeck-Peltier coefficient in the formula as it is, but the contribution of the Nernst-Ettingshausen effect is not included, making it difficult to calculate analytically. Therefore, in calculating the figure of merit of the hybrid transverse thermoelectric temperature modulation element, only S _xz needs to be measured directly.

Fig. 11B shows a perspective view of the configuration when measuring the magnetic field dependence of the thermoelectric power output of the _Bi88Sb12 / _{Bi0.2Sb1.8Te3} artificially inclined multilayer stack. _A heater is provided on the _xy plane that is the top _of the artificially inclined multilayer stack to generate a temperature gradient in the z-axis direction. The electromotive force generated in the x-axis direction of the artificially inclined multilayer stack is measured by a voltmeter V. An external magnetic field H is applied in the y-axis direction.
11C is a diagram showing the dependency of the external magnetic field H in the correlation diagram between the temperature difference ΔT and the electromotive force V in the device of FIG. 11B. When the temperature difference ΔT is 1.4K, 3.6K, 5.7K, and 7.9K, when the external magnetic field H is 0T, the electromotive force V is −0.45mV, −1.1mV, −1.8mV, and −2.5mV, respectively. When the external magnetic field H is 0.8T, the electromotive force V is −0.5mV, −1.2mV, −1.95mV, and −2.7mV, respectively. When the external magnetic field H is −0.8T, the electromotive force V is −0.4mV, −1.0mV, −1.65mV, and −2.3mV, respectively. In this way, the electromotive force V increases or decreases depending on the applied external magnetic field H. The dependence of the electromotive force V on the temperature difference ΔT is linear, and the transverse thermoelectric power S _xz can be calculated from the slope and the sample size.
FIG. 11D is a diagram showing the temperature difference ΔT dependency in the correlation diagram between the external magnetic field H and the electromotive force V in the device of FIG. 11B.

The results of measuring the lateral thermoelectric power generation output of the Co ₂ MnGa-based artificially inclined multilayer laminate will be described.

Artificially graded multilayer stacks consisting of Co ₂ MnGa/Bi _0.2 Sb _1.8 Te ₃ and Co ₂ MnGa/Bi ₂ Te ₃ were fabricated by alternately stacking Co ₂ MnGa disks prepared by SPS and Bi-Sb-Te powder or Bi-Te powder, and bonding them by SPS.

Fig. 12A shows a perspective view of the configuration when measuring the magnetic field dependence of the thermoelectric power output of a _Co2MnGa -based artificially inclined multilayer laminate. A heater is provided on the top plane of the artificially inclined multilayer laminate to generate a temperature gradient in the thickness direction. A magnetic field H is applied in the short axis direction of the sample perpendicular to the temperature gradient ∇T, and the electromotive force _VT generated in the long axis direction of the sample perpendicular to the temperature gradient ∇T and the magnetic field H is measured.

12B shows the results of measuring the magnetic field H dependence of the transverse electromotive force V _T when the temperature difference ΔT of the Co ₂ MnGa/Bi _0.2 Sb _1.8 Te ₃ artificially graded multilayer laminate is changed. When the temperature difference ΔT is 1.8K, 3.7K, 5.6K, 7.6K, and 9.5K, when the external magnetic field H is 0.8T, the electromotive force V _T is -0.68mV, -1.37mV, -2.08mV, -2.80mV, and -3.55mV, respectively. When the external magnetic field H is -0.8T, the electromotive force V _T is -0.70mV, -1.43mV, -2.17mV, -2.91mV, and -3.68mV, respectively. In this way, the electromotive force V _T increases or decreases depending on the applied external magnetic field H. The dependence of the electromotive force V _T on the temperature difference ΔT is linear, and the transverse thermoelectric power S _xz can be calculated from the slope and the sample size.

12C shows the results of measuring the magnetic field H dependence of the transverse electromotive force V _T when the temperature difference ΔT of the Co ₂ MnGa/Bi ₂ Te ₃ artificially graded multilayer laminate is changed. When the temperature difference ΔT is 1.7K, 3.5K, 5.2K, 7.0K, and 8.8K, when the external magnetic field H is 0.8T, the electromotive force V _T is 0.22mV, 0.44mV, 0.67mV, 0.90mV, and 1.13mV, respectively. When the external magnetic field H is -0.8T, the electromotive force V _T is 0.19mV, 0.39mV, 0.60mV, 0.80mV, and 1.00mV, respectively. In this way, the electromotive force V _T increases or decreases depending on the applied external magnetic field H. The dependence of the electromotive force V _T on the temperature difference ΔT is linear, and the transverse thermoelectric power S _xz can be calculated from the slope and the sample size.

Fig. 12D shows the results of measuring the temperature gradient ∇T dependence of the transverse electric field E _T of Co ₂ MnGa/Bi _0.2 Sb _1.8 Te ₃ and Co ₂ MnGa/Bi ₂ Te ₃ artificially graded multilayer laminates. The electromotive force V _T generated in the long axis direction was measured while sweeping the magnetic field H applied in the short axis direction of the sample perpendicular to the temperature gradient ∇T, and the transverse electric field E _T = V _T /l was obtained by normalizing it by the length l of the sample. The transverse electric field E _T obtained from the electromotive force V _T when the applied magnetic field H is +0.8T and -0.8T is proportional to the temperature gradient ∇T, and its gradient, i.e., the transverse thermoelectric power S _T , changes depending on the applied magnetic field H. The dimensionless figure of merit ZT was calculated from the transverse thermoelectric power S _T obtained from actual measurements and the electrical conductivity σ and thermal conductivity κ obtained by simulation. At an absolute temperature of 300 K, the values were 0.088 and 0.095 for the Co ₂ MnGa/Bi _0.2 Sb _1.8 Te ₃ artificially graded multilayer laminate when the magnetic field H was +0.8 T and -0.8 T, respectively, and 0.010 and 0.008 for the Co ₂ MnGa/Bi ₂ Te ₃ artificially graded multilayer laminate when the magnetic field H was +0.8 T and -0.8 T, respectively.
The modulation of the dimensionless figure of merit ZT of lateral thermoelectric conversion in a _Co2MnGa -based artificially inclined multilayer laminate by application of an external magnetic field H is several times larger than the highest value (up to 7 × ^10-4 ) of the dimensionless figure of merit ZT due to the anomalous Nernst effect in a single magnetic material reported so far.
As described above, by incorporating the anomalous Nernst effect into an artificially inclined multilayer stack and configuring it as a hybrid lateral thermoelectric temperature modulation element, it is possible to induce modulation of performance several times greater than the thermoelectric conversion performance of the anomalous Nernst effect alone.

In the examples shown in Figures 3 to 9, the temperature modulation phenomenon (cooling and heating phenomenon) of the hybrid lateral thermoelectric temperature modulation element of the present invention is directly measured, and the magnetic Peltier effect and normal etching shausen effect are measured through observation of the lock-in thermography image signal.

Furthermore, the temperature modulation phenomenon of the hybrid horizontal thermoelectric temperature modulation element of the present invention has been proven in the case of the magnetic Peltier effect or the normal Echingshausen effect, but in the case of electron transport phenomena resulting from spontaneous magnetization, it can be proven through the measurement of the temperature modulation phenomenon (cooling/heating phenomenon) of the thermoelectric conversion element by measuring the anisotropic magnetic Peltier effect or the anomalous Echingshausen effect. The hybrid horizontal thermoelectric temperature modulation due to the anomalous Echingshausen effect has been proven because the contribution of the anomalous Nernst effect, which is a reciprocal phenomenon, was confirmed in the example shown in Figure 12.

The hybrid lateral thermoelectric temperature modulation element of the present invention alternately laminates a first thermoelectric material layer exhibiting a magneto-thermoelectric effect and a second thermoelectric material layer, or a first thermoelectric material layer exhibiting a magneto-thermoelectric effect and a second thermoelectric material layer having residual magnetization and capable of driving the magneto-thermoelectric effect in the absence of a magnetic field, so that the temperature modulation phenomenon can be enhanced by an increase in the applied magnetic field or a magneto-thermoelectric effect such as the magnetic Peltier effect or the normal/abnormal Etchingshausen effect in a magnetized state. Since there is no complex three-dimensional joint structure like in conventional Peltier modules, it is effective in improving the durability of the element, reducing costs, and increasing the density of heat absorption and heat generation.

10 Sample 20 Electromagnet 30 Constant current source 40 Infrared camera 50 Signal processing system

Claims (19)

a first electrode and a second electrode arranged opposite to each other;
a laminate sandwiched between the first and second electrodes and electrically connected to both the first and second electrodes,
the laminate has a structure in which layers of a first thermoelectric material and layers of a second thermoelectric material are alternately laminated, the first thermoelectric material and the second thermoelectric material have at least one different value of a Peltier coefficient, an electrical conductivity, or a thermal conductivity, and at least one of the first and second thermoelectric materials exhibits a vertical magneto-thermoelectric effect or a horizontal magneto-thermoelectric effect;
a lamination surface of the first thermoelectric material layer and the second thermoelectric material layer is inclined with respect to a direction in which the first electrode and the second electrode face each other;
A temperature difference is generated in a direction perpendicular to the opposing direction in the thermoelectric temperature modulation element by a current supplied between the first and second electrodes and a magnetic field applied in a direction that generates a magneto-thermoelectric effect.
Hybrid lateral thermoelectric temperature modulation element. The vertical magneto-thermoelectric effect exhibits a thermoelectric change of 5 μV/K or more, and is a magneto-Peltier effect derived from an external magnetic field or an anisotropic magneto-Peltier effect derived from spontaneous magnetization;
The transverse magneto-thermoelectric effect exhibits a thermoelectric change of 5 μV/K or more, and is a normal Etchingshausen effect caused by an external magnetic field, or an anomalous Etchingshausen effect caused by spontaneous magnetization.
The hybrid lateral thermoelectric temperature modulation element according to claim 1 . the first layer of thermoelectric material is a _{Bi100-xSbx layer} (0≦x≦50);
the second layer of thermoelectric material is _a Bi2 _{- ySbyTe3} layer (0≦y≦2);
3. The hybrid lateral thermoelectric temperature modulation element according to claim 1 or 2. the first layer of thermoelectric material is a _Co2MnGa layer;
the second layer of thermoelectric material is _a Bi2 _{- ySbyTe3} layer (0≦y≦2);
3. The hybrid lateral thermoelectric temperature modulation element according to claim 1 or 2. The hybrid lateral thermoelectric temperature modulation element of claim 2, wherein the external magnetic field is applied to the first layer of thermoelectric material and the second layer of thermoelectric material by a permanent magnet, an electromagnet, or a permanent magnet and an electromagnet. the second thermoelectric material layer is a permanent magnet layer having residual magnetization, capable of driving the magneto-thermoelectric effect in the first thermoelectric material layer in the absence of a magnetic field, and improving horizontal thermoelectric conversion characteristics by the magneto-thermoelectric effect;
The hybrid lateral thermoelectric temperature modulation element according to claim 1 . the first layer of thermoelectric material is a _{Bi100-xSbx layer} (0≦x≦50);
The second thermoelectric material layer is one or more permanent magnet materials selected from the group consisting of _{SmCo5 -} based magnets, _{Sm2Co17 -} based magnets, _Nd2Fe14B -based magnets, alnico magnets, and ferrite magnets.
The hybrid lateral thermoelectric temperature modulation element according to claim 6 . The hybrid lateral thermoelectric temperature modulation element according to claim 1, wherein the inclination angle θ of the stacking surface with respect to the direction is 10° or more and 80° or less. The hybrid lateral thermoelectric temperature modulation element according to claim 1, wherein the figure of merit of the thermoelectric temperature modulation element is 0.1 or more. The hybrid lateral thermoelectric temperature modulation element according to claim 1, wherein the magnetic field applied in the direction that generates the magneto-thermoelectric effect is 0.1 T or more. A temperature modulation method using a horizontal thermoelectric conversion element, comprising: supplying a current to the horizontal thermoelectric conversion element and applying a magnetic field in a direction perpendicular to the current supply direction, thereby generating a temperature difference in a direction perpendicular to the current supply direction and the magnetic field application direction of the horizontal thermoelectric conversion element,
The horizontal thermoelectric conversion element is
a first electrode and a second electrode arranged opposite to each other;
a laminate sandwiched between the first and second electrodes and electrically connected to both the first and second electrodes;
the laminate has a structure in which layers of a first thermoelectric material and layers of a second thermoelectric material are alternately laminated, the first thermoelectric material and the second thermoelectric material have at least one different value of a Peltier coefficient, an electrical conductivity, or a thermal conductivity, and at least one of the first and second thermoelectric materials exhibits a vertical magneto-thermoelectric effect or a horizontal magneto-thermoelectric effect;
a lamination surface of the first thermoelectric material layer and the second thermoelectric material layer is inclined with respect to a direction in which the first electrode and the second electrode face each other;
A temperature modulation method using a horizontal thermoelectric conversion element, comprising: supplying a current through the first and second electrodes and applying a magnetic field in a direction perpendicular to the current supply direction, thereby generating a temperature difference in a direction perpendicular to the opposing direction in the horizontal thermoelectric conversion element. The vertical magneto-thermoelectric effect exhibits a thermoelectric change of 5 μV/K or more, and is a magneto-Peltier effect derived from an external magnetic field or an anisotropic magneto-Peltier effect derived from spontaneous magnetization;
The transverse magneto-thermoelectric effect exhibits a thermoelectric change of 5 μV/K or more, and is a normal Etchingshausen effect caused by an external magnetic field, or an anomalous Etchingshausen effect caused by spontaneous magnetization.
A temperature modulation method using the horizontal thermoelectric conversion element according to claim 11. the first layer of thermoelectric material is a _{Bi100-xSbx layer} (0≦x≦50);
the second layer of thermoelectric material is _a Bi2 _{- ySbyTe3} layer (0≦y≦2);
A temperature modulation method using the horizontal thermoelectric conversion element according to claim 11 or 12. the first layer of thermoelectric material is a _Co2MnGa layer;
the second layer of thermoelectric material is _a Bi2 _{- ySbyTe3} layer (0≦y≦2);
A temperature modulation method using the horizontal thermoelectric conversion element according to claim 11 or 12. The temperature modulation method using a horizontal thermoelectric conversion element according to claim 11 or 12, wherein the external magnetic field is applied to the first thermoelectric material layer and the second thermoelectric material layer by a permanent magnet, an electromagnet, or a permanent magnet and an electromagnet. the second layer of thermoelectric material is a permanent magnet layer that enhances the output of the lateral thermoelectric conversion element by a magneto-thermoelectric effect in a magnetized state compared to an unmagnetized state;
A temperature modulation method using the horizontal thermoelectric conversion element according to claim 11. the first layer of thermoelectric material is a _{Bi100-xSbx layer} (0≦x≦50);
The second thermoelectric material layer is one or more permanent magnet materials selected from the group consisting of _{SmCo5 -} based magnets, _{Sm2Co17 -} based magnets, _Nd2Fe14B -based magnets, alnico magnets, and ferrite magnets.
A temperature modulation method using the horizontal thermoelectric conversion element according to claim 16. The temperature modulation method using the horizontal thermoelectric conversion element according to claim 11, wherein the figure of merit of the horizontal thermoelectric conversion element is 0.1 or more. 12. The temperature modulation method using a horizontal thermoelectric conversion element according to claim 11, wherein the magnetic field applied in the direction in which the magneto-thermoelectric effect is generated is 0.1 T or more.

Images