Thermoelectric conversion element, thermoelectric conversion module, and thermoelectric conversion method

Abstract

A thermoelectric conversion element 200 according to the present invention comprises: a thermoelectric conversion section 202 having non-linear thermoelectric response characteristics with respect to an input of a heat flow H along a first direction; and electrode sections 207a, 207b for extracting a thermoelectric signal JN generated in the thermoelectric conversion section 202 due to the heat flow H input to the thermoelectric conversion section 202 from the outside along the first direction.

Description

Thermoelectric conversion element, thermoelectric conversion module, and thermoelectric conversion method

[0001] The present invention relates to a thermoelectric conversion element, a thermoelectric conversion module, a thermoelectric conversion method, and a fluctuation detection sensor.

[0002] Energy harvesting (environmental power generation) technology, which collects various forms of energy from the environment and converts it into electrical energy, has been attracting attention. Known examples of such technology include thermoelectric conversion elements that utilize the Seebeck effect as described in Patent Document 1 and thermoelectric conversion elements that utilize the spin Seebeck effect as described in Non-Patent Documents 1 and 2.

[0003] Also, detectors and sensors are known that detect failures, accidents, or deterioration of objects (e.g., device failures, building deterioration, or abnormalities in the health of living organisms) by measuring physical quantities such as specific frequencies, temperatures, or current values ​​of substances. For example, Patent Document 2 describes a fire detection device that detects the outbreak of a fire. This fire detection device includes a microphone that collects sounds generated in a monitored area as acoustic data, a frequency analysis unit that calculates a frequency spectrum by frequency analyzing the acoustic data collected by the microphone, and a flame detection unit that determines whether the frequency spectrum calculated by the frequency analysis unit contains 1 / f fluctuation characteristics in a frequency band below standing waves, and detects the outbreak of a flame in the monitored area if it is determined that the 1 / f fluctuation characteristics are contained.

[0004] JP 2015-144212 A Japanese Patent No. 7376660

[0005] K. Uchida, E. Saito, "Spin Seebeck effect and its application to thermoelectric conversion", Applied Physics, Applied Physics Society, 2013, Vol. 82, No. 11, pp. 928-931 K. Uchida, et. al., "Observation of the spin Seebeck effect", Nature, 2008, Vol. 455, No. 9, pp. 778-781

[0006] Conventional thermoelectric conversion elements utilizing the Seebeck effect or spin Seebeck effect convert the temperature gradient generated in a material into electrical energy. When the temperature gradient reverses, the polarity of the electric field generated within the element also reverses. Therefore, a constant temperature gradient on the macroscale is required for power generation. Therefore, it has been virtually impossible to generate thermoelectric power from a heat source in which the direction and magnitude of the heat flow change randomly, i.e., from temperature fluctuations on the microscale.

[0007] Furthermore, conventional detectors and sensors for detecting breakdowns, accidents, or deterioration, etc., can only detect the occurrence of breakdowns, accidents, or deterioration after they actually occur in the target object, due to their mechanism of measuring physical quantities such as specific frequencies, temperatures, and current values ​​of materials. Therefore, conventional detectors and sensors were unable to detect the signs of these problems and prevent them from occurring. It is believed that if energy, such as vibrations and electromagnetic noise emitted from a target object, which normally repeats a certain regular movement, can be detected to begin to move irregularly, it would be possible to detect the signs of breakdowns, accidents, or deterioration of the target object. However, to do this, it is necessary to detect fluctuations in such energy, which change randomly on a microscale. However, sensors capable of detecting microscale energy fluctuations emitted from such targets have not been known until now.

[0008] The present invention has been made in consideration of the above-mentioned problems, and aims to provide a thermoelectric conversion element, a thermoelectric conversion module, and a thermoelectric conversion method that are capable of thermoelectric conversion from microscale temperature fluctuations, or to provide a sensor that is capable of detecting microscale energy fluctuations emitted from an object.

[0009] In order to solve the above-mentioned problems, the thermoelectric conversion element of the present invention comprises a thermoelectric conversion unit having a nonlinear thermoelectric response characteristic with respect to the input of a heat flow along a first direction, and an electrode unit for extracting a thermoelectric signal generated in the thermoelectric conversion unit due to the heat flow input to the thermoelectric conversion unit from the outside along the first direction.

[0010] The present inventors have found that, in a thermoelectric conversion element according to the present invention, a thermoelectric signal generated in a thermoelectric conversion unit by an inflowing heat flow is nonlinear with respect to the direction of the heat flow due to the nonlinear thermoelectric response characteristics of the thermoelectric conversion unit, and that such a nonlinear thermoelectric signal can be electrically extracted by an electrode unit. Therefore, even from a heat flow whose direction and magnitude change randomly, a thermoelectric signal can be generated in a fixed direction within the thermoelectric conversion unit, making it possible to perform thermoelectric conversion from temperature fluctuations on a microscale.

[0011] Furthermore, in order to solve the above-mentioned problems, in the thermoelectric conversion element according to the present invention, the thermoelectric conversion portion exhibits nonreciprocal conduction based on the breaking of spatial inversion symmetry, and the electrode portion can include a pair of electrodes spaced apart from each other on the thermoelectric conversion portion for extracting a nonreciprocal thermoelectric signal.

[0012] The inventors of the present application have found that the thermoelectric signal generated in the thermoelectric conversion unit by the incoming heat flow is nonreciprocal with respect to the direction of the heat flow due to the nonreciprocal conduction of the thermoelectric conversion unit, and that such a nonreciprocal thermoelectric signal can be electrically extracted using a pair of electrodes. Therefore, even from a heat flow whose direction and magnitude change randomly, it is possible to generate a thermoelectric signal in a fixed direction within the thermoelectric conversion unit, making it possible to perform thermoelectric conversion from temperature fluctuations on a microscale.

[0013] Furthermore, in the thermoelectric conversion element according to the present invention, it is preferable that the thermoelectric conversion part is a thermoelectric conversion layer including a ferromagnetic metal layer and a paramagnetic metal layer stacked on each other along a second direction perpendicular to the first direction, and the pair of electrodes are provided on the thermoelectric conversion layer along the first direction in an in-plane direction of the thermoelectric conversion layer, spaced apart from each other. This increases the thermoelectric conversion voltage based on the breaking of the spatial inversion symmetry of the thermoelectric conversion layer, thereby particularly increasing the thermoelectric conversion efficiency.

[0014] Furthermore, in the thermoelectric conversion element according to the present invention, the distance between the pair of electrodes can be set to 0.1 μm or more and 1000 μm or less. By setting the distance in this range, the thermoelectric conversion voltage becomes large, and therefore the thermoelectric conversion efficiency becomes high.

[0015] Furthermore, in the thermoelectric conversion element according to the present invention, it is preferable that the paramagnetic metal layer is composed of a single layer or a plurality of laminated layers, and each of the single layer or the plurality of laminated layers of the paramagnetic metal layer is composed of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy. This particularly increases the nonreciprocity of conduction due to the breaking of the spatial inversion symmetry of the thermoelectric conversion layer, thereby particularly increasing the thermoelectric conversion efficiency.

[0016] Furthermore, in the thermoelectric conversion element according to the present invention, it is preferable that the ferromagnetic metal layer is composed of a single layer or a plurality of laminated layers, and each of the single layer or the plurality of laminated layers of the ferromagnetic metal layer is composed of a Ni—Fe alloy, Fe, CoNi, Gd, CoFeB alloy, or (Ga,Fe)Sb alloy. These are materials that exhibit ferromagnetism at room temperature, and realize nonreciprocity of conduction at room temperature, thereby increasing thermoelectric conversion efficiency.

[0017] Furthermore, a thermoelectric conversion module according to the present invention includes a plurality of thermoelectric conversion elements according to any one of the above, and the plurality of thermoelectric conversion elements are electrically connected to each other so that thermoelectric signals from the respective thermoelectric conversion elements can be superimposed with the same polarity, thereby enabling a large non-reciprocal thermoelectric signal to be extracted.

[0018] In addition, in order to solve the above-mentioned problems, the thermoelectric conversion method of the present invention includes the steps of: preparing a thermoelectric conversion element including a thermoelectric conversion layer exhibiting nonreciprocal conduction based on the breaking of spatial inversion symmetry; and a pair of electrodes spaced apart from each other on the thermoelectric conversion layer for extracting nonreciprocal thermoelectric signals, wherein the thermoelectric conversion layer includes a ferromagnetic metal layer and a paramagnetic metal layer stacked on each other, and the pair of electrodes are spaced apart from each other on the paramagnetic metal layer in the in-plane direction of the paramagnetic metal layer; and performing thermoelectric conversion while applying a magnetic field of 1 T or less to the ferromagnetic metal layer of the thermoelectric conversion layer.

[0019] According to the thermoelectric conversion method of the present invention, when a magnetic field of 1 T or less is applied and a heat flow flows into a stack of magnetized ferromagnetic and paramagnetic metal layers, a spin current is driven in the paramagnetic metal layer, and spin accumulation is formed at the interface between the paramagnetic and ferromagnetic metal layers. The polarity of this spin accumulation becomes parallel or antiparallel to the magnetization direction of the ferromagnetic metal layer depending on the direction of the heat flow. The scattering of electrons changes depending on the relative orientation between the polarity of this spin accumulation and the magnetization direction, resulting in a thermoelectric signal generated in the thermoelectric conversion layer that is dependent on the direction of the heat flow, i.e., nonreciprocity. This is a nonreciprocal thermoelectric signal, and such a nonreciprocal thermoelectric signal can be electrically extracted using a pair of electrodes. Therefore, even if a heat flow whose direction and magnitude change randomly, a thermoelectric signal can be generated in a fixed direction within the paramagnetic metal layer, enabling thermoelectric conversion from temperature fluctuations on a microscale.

[0020] In addition, in order to solve the above-mentioned problems, the thermoelectric conversion method according to the present invention includes the steps of applying a temperature gradient including an AC component to a thermoelectric conversion unit that exhibits nonreciprocal conduction based on the breaking of spatial inversion symmetry, and the step of the thermoelectric conversion unit thermoelectrically converting the temperature gradient including the AC component to generate a nonreciprocal thermoelectric signal.

[0021] According to the thermoelectric conversion method of the present invention, the thermoelectric conversion unit thermoelectrically converts an applied temperature gradient containing an AC component to generate a thermoelectric signal. At this time, the thermoelectric signal becomes nonlinear with respect to the direction of the temperature gradient due to the nonreciprocal conduction characteristics of the thermoelectric conversion unit, resulting in a nonreciprocal thermoelectric signal. As a result, temperature fluctuations on a microscale can be regarded as a temperature gradient containing an AC component, making it possible to perform thermoelectric conversion from temperature fluctuations on a microscale.

[0022] In addition, in order to solve the above-mentioned problems, the fluctuation detection sensor of the present invention comprises a nonlinear response unit having a nonlinear electrical response characteristic to the input of external energy along a first direction, and an electrode unit for extracting an electrical response signal generated in the nonlinear response unit due to energy fluctuations input from the outside to the nonlinear response unit along the first direction.

[0023] The inventors have found that, with the fluctuation detection sensor according to the present invention, the electrical response signal generated in the nonlinear response unit by energy fluctuations input from outside is nonreciprocal with respect to the direction of energy flow due to the nonlinear electrical response characteristics of the nonlinear response unit, and that such a nonreciprocal electrical response signal can be electrically extracted by the electrode unit. Therefore, even from external energy whose direction and magnitude change randomly, it is possible to generate a nonlinear voltage signal, which is an electrical response signal that is nonreciprocal in a fixed direction, within the nonlinear response unit, making it possible to detect microscale energy fluctuations emitted from an object.

[0024] Furthermore, in the fluctuation detection sensor according to the present invention, the external energy can be an electric current and / or a heat flow, and the energy fluctuation can be an electric current fluctuation and / or a heat flow fluctuation. In this case, it becomes possible to detect an electric current fluctuation and / or a heat flow fluctuation emitted from an object.

[0025] Furthermore, in the fluctuation detection sensor according to the present invention, the external energy can be a current and the energy fluctuation can be a current fluctuation, in which case it is possible to detect a current fluctuation emitted from an object.

[0026] Furthermore, in the fluctuation detection sensor according to the present invention, the nonlinear response section can have a configuration in which spatial inversion symmetry is broken along a second direction orthogonal to the first direction.

[0027] Furthermore, in the fluctuation detection sensor according to the present invention, the nonlinear response unit may include a ferromagnetic layer and a paramagnetic layer stacked on top of each other along the second direction.

[0028] Furthermore, in the fluctuation detection sensor according to the present invention, the ferromagnetic layer can be configured to have spontaneous magnetization and / or high coercive force.

[0029] Furthermore, in the fluctuation detection sensor according to the present invention, the ferromagnetic layer is a ferromagnetic metal layer, and the L1 layer is made of a magnetic metal and a paramagnetic metal. ０The paramagnetic layer may be a paramagnetic metal layer, which is made of an ordered alloy of the type mentioned above, thereby increasing the magnitude of the coercive force that the ferromagnetic layer can have.

[0030] According to the present invention, a thermoelectric conversion element, a thermoelectric conversion module, and a thermoelectric conversion method capable of thermoelectric conversion from microscale temperature fluctuations are provided, or a sensor capable of detecting microscale energy fluctuations emitted from an object is provided.

[0031] FIG. 1 is a perspective view showing a thermoelectric conversion element according to a first embodiment; FIG. 2 is a cross-sectional view of the thermoelectric conversion element according to the first embodiment; FIG. 3 is a process chart showing steps of a thermoelectric conversion method according to the first embodiment; FIG. 4 is a perspective view showing the configuration of a thermoelectric conversion module using a thermoelectric conversion element; FIG. 5 is a schematic view showing an experimental method for measuring a non-reciprocal thermoelectric signal; FIG. 6 is a diagram showing measurement results of an experiment for measuring a non-reciprocal thermoelectric signal; FIG. 7 is a perspective view showing a fluctuation detection sensor according to a second embodiment; FIG. 8 is a cross-sectional view of the fluctuation detection sensor according to the second embodiment; FIG. 9 is a process chart showing steps of an energy conversion method according to the second embodiment; FIG. 10 is a perspective view showing the configuration of a fluctuation detection sensor module using a fluctuation detection sensor; Ｎ１ 1 is a schematic diagram showing an experimental method for measuring a non-reciprocal voltage signal J in Example 4. Ｎ１ 1 is a schematic diagram showing an experimental method for measuring the amount of hydroxylase in the sample of Example 4. FIG.

[0032] Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In each drawing, the same reference numerals are used for the same elements where possible. Furthermore, the dimensional ratios within and between elements in the drawings are arbitrary for ease of viewing.

[0033] First Embodiment FIG. 1 is a perspective view showing a thermoelectric conversion element according to a first embodiment, and FIG. 2 is a cross-sectional view of the thermoelectric conversion element according to the first embodiment.

[0034] As shown in FIGS. 1 and 2, a thermoelectric conversion element 200 according to this embodiment includes a substrate 201, a thermoelectric conversion layer 202 as a thermoelectric conversion portion, and a pair of electrodes 207a and 207b as an electrode portion.

[0035] The substrate 201 is a plate- or film-shaped member, for example, as shown in FIG. 1 , and has a substantially flat surface 201S. The surface 201S is rectangular when viewed from the Z-axis direction, for example, as shown in FIG. 1 . The substrate 201 is a member for laminating a thermoelectric conversion layer 202 on the surface 201S. Examples of materials that can be used to form the substrate 201 include metals such as Au, Ag, Cu, and Al, silicon, sapphire, SiC, GaN, yttria-stabilized zirconia (YSZ), and resins such as polyimide, polyethylene terephthalate (PET), and polyethylene naphthalate (PEN).

[0036] 1 and 2 show an orthogonal coordinate system C, in which the Z axis is set in the thickness direction of the substrate 201, and the X axis and Y axis are set in the long side and short side directions, respectively, of the rectangle of the surface 201S of the substrate 201. The orthogonal coordinate system C is also shown in the subsequent figures as needed.

[0037] The thermoelectric conversion layer 202 has a single-layer ferromagnetic metal layer 203 and a single-layer paramagnetic metal layer 205 stacked on top of each other along the Z-axis direction (second direction), and in this embodiment, the ferromagnetic metal layer 203 is stacked on the surface 201S of the substrate 201, and the paramagnetic metal layer 205 is stacked on the ferromagnetic metal layer 203. The ferromagnetic metal layer 203 is made of a metal that exhibits ferromagnetism at the temperature in the usage environment of the thermoelectric conversion element 200. The ferromagnetic metal layer 203 is preferably made of a metal that exhibits ferromagnetism at room temperature (approximately 300 K), and from this perspective, it is preferably made of Fe, Co, Ni, Gd, or an alloy containing at least one of them, and is particularly preferably made of a Ni—Fe alloy, Co, or CoFeB.

[0038] The thermoelectric conversion layer 202 has, for example, a rectangular shape in plan view (i.e., when viewed from the Z-axis direction) with long sides along the X-axis direction and short sides along the Y-axis direction. The thickness of the ferromagnetic metal layer 203 is not particularly limited, but can be, for example, 1 nm or more and 100 nm or less.

[0039] The paramagnetic metal layer 205 is made of a metal that exhibits paramagnetism at the temperature of the environment in which the thermoelectric conversion element 200 is used, and is preferably made of a metal that exhibits paramagnetism at room temperature (300 K). The paramagnetic metal layer 205 is preferably made of a paramagnetic metal with high spin-orbit interaction in order to enhance the spin-Nernst effect in the paramagnetic metal layer 205 (described below). From this perspective, the paramagnetic metal layer 205 is preferably made of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy, and is particularly preferably made of Pt. The thickness of the paramagnetic metal layer 205 is not particularly limited, but can be, for example, 1 nm or more and 100 nm or less.

[0040] The thermoelectric conversion layer 202 exhibits nonreciprocal conduction due to the breaking of spatial inversion symmetry. Specifically, a paramagnetic metal layer is provided on the surface of the ferromagnetic metal layer 203, but a similar paramagnetic metal layer is not provided on the back surface of the ferromagnetic metal layer 203. Also, a ferromagnetic metal layer is provided on the back surface of the paramagnetic metal layer 205, but a similar ferromagnetic metal layer is not provided on the surface of the paramagnetic metal layer 205. Therefore, the thermoelectric conversion layer 202 has a structure in which spatial inversion symmetry is broken along the Z-axis direction. The thermoelectric conversion layer 202 essentially has spatial inversion symmetry along the XY plane. Furthermore, due to the breaking of spatial inversion symmetry along the Z-axis direction, the thermoelectric conversion layer 202 exhibits nonreciprocity in terms of electrical conductivity in the in-plane (XY plane) direction. As a result, the thermoelectric conversion layer 202 has nonlinear thermoelectric response characteristics with respect to the input of a heat flow H along one of the in-plane (XY plane) directions (in this embodiment, the X-axis direction as the first direction), as described below.

[0041] The pair of electrodes 207a, 207b are provided on the thermoelectric conversion layer 202 so as to be spaced apart from each other in the in-plane (XY plane) direction of the thermoelectric conversion layer 202. The pair of electrodes 207a, 207b converts a thermoelectric signal J generated in the thermoelectric conversion layer 202 due to a heat flow H input from the outside to the thermoelectric conversion layer 202 along one of the in-plane (XY plane) directions, as will be described later. ＮIn this embodiment, the pair of electrodes 207 a, 207 b are provided on the paramagnetic metal layer 205 so as not to contact the ferromagnetic metal layer 203, and are provided on a surface 205S of the paramagnetic metal layer 205 opposite the ferromagnetic metal layer 203. The pair of electrodes 207 a, 207 b are spaced apart from each other along the X-axis direction, which is one of the in-plane directions (in the XY plane) of the paramagnetic metal layer 205. The pair of electrodes 207 a, 207 b are made of a material having conductivity that allows a nonreciprocal thermoelectric signal to be extracted from the thermoelectric conversion layer 202, as will be described later. Examples of such materials include metal materials such as Cu, Ag, Au, Pt, Ni, Al, constantan, Cr, In, Pd, Fe, Cu alloys, Ti / Au stacks, and Cr / Au stacks, and conductive oxides such as indium tin oxide (ITO) and zinc oxide (ZnO). Cu, Ag, Au, Pt, Ni, Al, constantan, and Cu-copper alloys are preferred, and Cu, Au, Ag, Pt, Ni, Ti / Au stacks, and Cr / Au stacks are particularly preferred.

[0042] The thickness of the pair of electrodes 207 a, 207 b is not particularly limited, but may be, for example, 10 nm or more and 1 μm or less. The separation distance D207 between the pair of electrodes 207 a, 207 b along the X-axis direction is preferably 0.1 μm or more and 1000 μm or less.

[0043] Furthermore, the thermoelectric conversion element 200 of this embodiment may have a layer made of an electrically insulating magnetic material that is arranged in contact with the ferromagnetic metal layer 203, or may not have such a layer made of an insulating magnetic material.

[0044] To manufacture such a thermoelectric conversion element 200, for example, a substrate 201 is prepared, and a material for forming the ferromagnetic metal layer 203 and a material for forming the paramagnetic metal layer 205 are deposited in this order on the surface 201S of the substrate by physical vapor deposition such as DC magnetron sputtering to form a laminate, and the laminate is patterned into a rectangular shape in a plan view by photolithography and lift-off to form the thermoelectric conversion layer 202. Thereafter, a metal material for forming the pair of electrodes 207a, 207b is deposited by physical vapor deposition such as DC magnetron sputtering in a region including the surface 205S of the paramagnetic metal layer 205 to form a metal layer, and the metal layer is patterned into a predetermined shape by photolithography and lift-off to form the pair of electrodes 207a, 207b.

[0045] Next, a thermoelectric conversion method according to this embodiment will be described. Fig. 3 is a process diagram showing the steps of the thermoelectric conversion method according to this embodiment. As shown in Fig. 3, the steps of the thermoelectric conversion method according to this embodiment include step S11 of preparing a thermoelectric conversion element, step S12 of inputting a heat flow, and step S13 of performing thermoelectric conversion while applying an external magnetic field. These steps will be described again with reference to Figs. 1 and 2.

[0046] In step S11 of preparing a thermoelectric conversion element, the above-described thermoelectric conversion element 200 as shown in Figures 1 and 2 is prepared. Subsequently, in step S12 of inputting a heat flow, a heat flow H is applied to a thermoelectric conversion layer 202 that exhibits nonreciprocal conduction based on the breaking of spatial inversion symmetry. Specifically, in step S12 of inputting a heat flow, the heat flow H from the outside, which is the target of thermoelectric conversion, is input to the thermoelectric conversion layer 202 so as to flow along the X-axis direction, which is one of the in-plane directions (in the XY plane) of the thermoelectric conversion layer 202.

[0047] The heat flow H in this embodiment corresponds to the above-mentioned "temperature fluctuations on a microscale." The heat flow H is a temperature gradient containing one or more AC components. Here, an "AC component" is a component in which the magnitude of heat periodically fluctuates between positive and negative at a predetermined frequency. When the heat flow H contains multiple AC components, the frequencies of the AC components are different from each other. As described above, the "temperature fluctuations on a microscale" correspond to a heat source in which the direction and magnitude of the heat flow change randomly. However, since a single AC component or multiple AC components with different frequencies contained in the temperature fluctuations can be determined by Fourier transform, the above-mentioned "temperature fluctuations on a microscale" can be regarded as a temperature gradient containing one or more AC components.

[0048] Next, step S13 of performing thermoelectric conversion while applying an external magnetic field is performed. In step S13 of performing thermoelectric conversion while applying an external magnetic field, the thermoelectric conversion layer 202 thermoelectrically converts the heat flow H to generate a thermoelectric signal J. Ｎ Specifically, in step S13 of performing thermoelectric conversion while applying an external magnetic field, an external magnetic field B generated by a magnetic field generating unit provided outside the thermoelectric conversion element 200 is applied to the ferromagnetic metal layer 203. The direction of application of the external magnetic field B is an in-plane direction (within the XY plane) of the thermoelectric conversion element 200, and is the negative Y-axis direction, which is one of the directions perpendicular to the direction of the flow of the heat flow H. As a result, the ferromagnetic metal layer 203 is magnetized in the direction of the external magnetic field B, and has magnetization 203M oriented in the direction of the external magnetic field B. The application of such an external magnetic field B is maintained while thermoelectric conversion of the heat flow H is being performed. The magnitude of the external magnetic field B can be, for example, 1 T or less.

[0049] In step S13, when the heat flow H flows into the paramagnetic metal layer 205, the spin current J Ｓ is generated along the Z-axis direction and flows into the ferromagnetic metal layer 203. Ｓ The polarity of the spin current J changes to the positive or negative Y-axis direction depending on the direction of the heat flow H. Therefore, when the direction of the heat flow H is reversed, the spin current J Ｓ The polarity of the spin current J Ｓand the direction of the magnetization 203M of the ferromagnetic metal layer 203 are reversed. Ｓ The relative orientation of the polarity of the magnetization 203M and the direction of the magnetization 203M can also be reversed.

[0050] Spin current J Ｓ flows into the ferromagnetic metal layer 203, electron scattering changes depending on the relative direction between the polarity of the spin current and the direction of magnetization 203M of the ferromagnetic metal layer 203. Since the polarity of the spin current near the junction interface between the ferromagnetic metal layer 203 and the paramagnetic metal layer 205 changes depending on the direction of the heat current H, this electron scattering changes depending on the direction of the heat current H. This causes the thermoelectric signal generated in the thermoelectric conversion layer 202 to depend on the direction of the heat current H, i.e., to be non-reciprocal.

[0051] The thermoelectric signal J generated in the thermoelectric conversion layer 202 Ｎ is non-reciprocal with respect to the direction of the heat flow H due to the non-reciprocal conduction of the thermoelectric conversion layer 202 as described above (i.e., the thermoelectric signal J Ｎ changes nonlinearly with respect to the direction of the heat flow H), and the thermoelectric signal J Ｎ The direction of the non-reciprocal thermoelectric signal J is the positive direction of the X axis. Ｎ The inventors of the present application have found that the thermoelectric conversion element 200 according to this embodiment can extract the non-reciprocal thermoelectric signal J in a fixed direction within the thermoelectric conversion layer 202, even if the direction and magnitude of the heat flow H change randomly. Ｎ Since it is possible to generate thermoelectric power, it is possible to perform thermoelectric conversion from temperature fluctuations on a microscale.

[0052] Furthermore, by reversing the direction of the external magnetic field B to the positive Y-axis direction and reversing the direction of the magnetization 203M of the ferromagnetic metal layer 203 to the positive Y-axis direction, a non-reciprocal thermoelectric signal J Ｎ The direction of can be reversed to the negative X-axis direction.

[0053] In the thermoelectric conversion layer 202, a non-reciprocal thermoelectric signal J is extracted by a pair of electrodes 207a and 207b.Ｎ is the magnitude of the voltage V, the electric field generated in thermoelectric conversion layer 202 corresponding to V is E, the nonlinear Seebeck coefficient of thermoelectric conversion layer 202 is S, the distance between pair of electrodes 207a and 207b is L (D207 in FIG. 2), the temperature gradient within the range of distance L in thermoelectric conversion layer 202 is ∇T, and the temperature difference within the range of distance L in thermoelectric conversion layer 202 is ΔT, then electric field E=nonlinear Seebeck coefficient S×(temperature gradient ∇T) ２ By multiplying both sides of this equation by the distance L, the voltage V = nonlinear Seebeck coefficient S × (temperature difference ΔT) ２ / L, the smaller the distance L between the pair of electrodes 207a and 207b, the larger the voltage V. Therefore, by setting the separation distance D207 between the pair of electrodes 207a and 207b in the X-axis direction to a small value, preferably 1000 μm or less, as described above, the non-reciprocal thermoelectric signal J extracted by the pair of electrodes 207a and 207b can be reduced. Ｎ becomes larger.

[0054] The thermoelectric conversion element 200 may also include a heat conduction portion that is provided in contact with or close to the end face on the positive side of the X-axis and / or the end face on the negative side of the X-axis, and that facilitates the flow of heat flow H into the thermoelectric conversion layer 202. Such a heat conduction portion may be made of, for example, aluminum, copper, carbon fiber, sapphire, alumina, silicon with a thermal oxide film, or a high molecular weight polymer.

[0055] 4 is a perspective view showing the configuration of a thermoelectric conversion module using the above-described thermoelectric conversion elements. As shown in FIG. 4, a thermoelectric conversion module 300 according to this embodiment includes a plurality of thermoelectric conversion elements 200, and in this embodiment, includes four thermoelectric conversion elements 200a, 200b, 200c, and 200d. The four thermoelectric conversion elements 200a, 200b, 200c, and 200d generate non-reciprocal thermoelectric signals J from the respective thermoelectric conversion elements. Ｎ are electrically connected to each other so as to overlap with each other with the same polarity.

[0056] Specifically, in the thermoelectric conversion module 300, thermoelectric conversion elements 200a and 200c are arranged with respect to the Cartesian coordinate system C in a manner similar to that of the thermoelectric conversion element 200 shown in Figures 1 and 2, and thermoelectric conversion elements 200b and 200d are arranged with respect to the Cartesian coordinate system C in a manner obtained by rotating the thermoelectric conversion element 200 shown in Figures 1 and 2 by 180 degrees around the Y axis, and are alternately provided along the Y axis direction. The electrode 207a of the thermoelectric conversion element 200a is electrically connected to the electrode 207b of the thermoelectric conversion element 200b, the electrode 207a of the thermoelectric conversion element 200b is electrically connected to the electrode 207b of the thermoelectric conversion element 200c, and the electrode 207a of the thermoelectric conversion element 200c is electrically connected to the electrode 207b of the thermoelectric conversion element 200d. 1 and 2 , in the same manner as the thermoelectric conversion element 200 shown in Figures 1 and 2, and then the stacking order of the ferromagnetic metal layer 203 and the paramagnetic metal layer 205 of the thermoelectric conversion layer 202 in the thermoelectric conversion elements 200b and 200d can be reversed. In this case, the electrode 207a of the thermoelectric conversion element 200a and the electrode 207a of the thermoelectric conversion element 200b are electrically connected, the electrode 207b of the thermoelectric conversion element 200b and the electrode 207b of the thermoelectric conversion element 200c are electrically connected, and the electrode 207a of the thermoelectric conversion element 200c and the electrode 207a of the thermoelectric conversion element 200d are electrically connected.

[0057] Moreover, the ferromagnetic metal layer 203 of each of the four thermoelectric conversion elements 200a, 200b, 200c, and 200d is magnetized in the negative Y-axis direction by the external magnetic field B, and has a magnetization 203M oriented in the negative Y-axis direction.

[0058] The thermoelectric conversion module 300 also includes heat conduction units 301 and 302 that are provided in contact with or in close proximity to the end faces on the negative side of the X-axis and the end faces on the positive side of the X-axis of the four thermoelectric conversion elements 200a, 200b, 200c, and 200d. This makes it easier for the heat flowing in the X-axis direction to flow into the ferromagnetic metal layers 203 of the four thermoelectric conversion elements 200a, 200b, 200c, and 200d. The thermoelectric conversion module 300 does not necessarily have to include the heat conduction units 301 and 302.

[0059] The heat flow flows into the ferromagnetic metal layer 203 of each of the four thermoelectric conversion elements 200a, 200b, 200c, and 200d, and the nonreciprocal thermoelectric signals generated in each thermoelectric conversion element 200 are superimposed with the same polarity and can then be extracted by the electrode 207b of the thermoelectric conversion element 200a and the electrode 207a of the thermoelectric conversion element 200d. This makes it possible to extract a large nonreciprocal thermoelectric signal.

[0060] (Example) Next, to clarify the effects of the invention according to the first embodiment, an example will be described. An element corresponding to the thermoelectric conversion element 200 was fabricated as follows. First, a 0.5 mm-thick Si substrate with a thermal oxide film was prepared as the substrate 201. On this substrate, a 5 nm-thick Ni—Fe alloy layer as the ferromagnetic metal layer 203 and a 5 nm-thick Pt layer as the paramagnetic metal layer 205 were deposited in this order by DC magnetron sputtering to form the thermoelectric conversion layer 202. Then, by photolithography and lift-off, the thermoelectric conversion layer 202 was patterned into a rectangular shape in plan view, with a width in the Y-axis direction of 5 μm and a length in the X-axis direction of 50 μm. Thereafter, a Ti / Au layer was deposited on the thermoelectric conversion layer 202 by RF magnetron sputtering, and the Ti / Au layer was patterned into a predetermined shape by photolithography and lift-off to form a pair of electrodes corresponding to the pair of electrodes 207a, 207b, thereby producing an element corresponding to the thermoelectric conversion element 200.

[0061] Next, the element is subjected to step S12 of inputting a heat flow at room temperature (300 K) and step S13 of performing thermoelectric conversion while applying an external magnetic field, and a non-reciprocal thermoelectric signal J is output from the pair of electrodes 207 a and 207 b. Ｎ An experiment was conducted to measure the non-reciprocal thermoelectric signal J. Ｎ5 is a schematic diagram showing an experimental method for measuring the temperature of a thermoelectric conversion element 200. As shown in Fig. 5, a first resistance heater 210 was provided adjacent to the end of the element corresponding to the thermoelectric conversion element 200 in the negative direction of the X-axis, and a second resistance heater 211 was provided adjacent to the end of the element in the positive direction of the X-axis. A first power supply 220 and a second power supply 221 were electrically connected to the first resistance heater 210 and the second resistance heater 211, respectively. A lock-in amplifier 230 was also electrically connected between a pair of electrodes 207a and 207b.

[0062] Then, while the magnitude of the external magnetic field B applied to the thermoelectric conversion element 200 in the negative direction of the Y axis is changed from 5 mT to −5 mT in 0.1 mT intervals, an AC temperature gradient ΔT that changes sinusoidally with an angular frequency ω in the direction along the X axis of the thermoelectric conversion element 200 by the first resistance heater 210 and the second resistance heater 211 is generated. ＡＣ Then, among the thermoelectric signals generated by the thermoelectric conversion element 200, the AC temperature gradient ΔT ＡＣ ΔV is the y component of the change in the thermoelectric signal corresponding to twice the frequency of ２ω、ｙ was detected by a lock-in amplifier 230.

[0063] AC temperature gradient ΔT ＡＣ More specifically, when ω is the angular frequency and t is time, the first power supply 220 supplies V to the first resistance heater 210. ｄｃ +V ａｃ An AC voltage of sin(ωt) is applied, and V ｄｃ -V ａｃ An AC voltage of sin(ωt) was applied to the second resistance heater 211 by the second power supply 221. As a result, a voltage ΔT was applied to the thermoelectric conversion element 200 in the direction along the X axis. ＡＣ = (√2) ΔT ｒｍｓ An AC temperature gradient expressed as sin(ωt) is applied (ΔT ｒｍｓ is the effective value of the AC temperature gradient). Therefore, the AC temperature gradient ΔT ＡＣcorresponds to a heat flow H containing a single AC component. A voltage source device WF1968 manufactured by NF Corporation was used as the first power supply 220 and the second power supply 221, and a lock-in amplifier LI5640 manufactured by NF Corporation was used as the lock-in amplifier 230. The frequency f (=ω / 2π) of the AC voltage applied by the first power supply 220 and the second power supply 221 was set to 13.723 Hz, and the above ΔT ｒｍｓ The above V ａｃ The lock-in conditions of the lock-in amplifier 230 were set to low dynamic reserve, 500 nV measurement sensitivity, 1 second time constant, and −24 dB slope.

[0064] FIG. 6 shows the thermoelectric signal ΔV obtained as a result of the above experiment. ２ω、ｙ 1 is a graph showing the dependence of ΔT on the external magnetic field B. ＡＣ Since varies sinusoidally, when viewed on a macro scale, a temperature gradient is not substantially applied to the thermoelectric conversion element 200, and ΔT ＡＣ corresponds to the temperature fluctuation on a micro scale. If the thermoelectric signal output from the thermoelectric conversion element 200 is ＡＣ If the polarity of the AC temperature gradient ΔT ＡＣ The thermoelectric signal ΔV corresponding to twice the frequency of ２ω、ｙ is the AC temperature gradient ΔT ＡＣ Since the thermoelectric signals corresponding to the positive and negative phases of the AC temperature gradient ΔT cancel each other out, the result is zero. ＡＣ When the polarity of the thermoelectric signal is nonlinear (non-reciprocal), ２ω、ｙ is a finite value.

[0065] In this experiment, as shown in FIG. 6, when an external magnetic field B is applied, a finite value of ΔV ２ω、ｙ is detected, and when the polarity of the external magnetic field B is reversed, ΔV ２ω、ｙThis indicates that the thermoelectric conversion element used in this experiment is capable of generating a thermoelectric signal nonlinearly with respect to the direction of the temperature gradient, and therefore is capable of performing thermoelectric conversion from temperature fluctuations on a microscale where a temperature gradient does not substantially exist on a macroscale.

[0066] The present invention is not limited to the first embodiment described above, and various modifications are possible.

[0067] For example, in the first embodiment described above, the thermoelectric conversion element 200 includes the substrate 201 (see FIGS. 1 and 2), but the thermoelectric conversion element 200 does not necessarily include the substrate 201. In this case, the thermoelectric conversion layer 202 can be formed as a free-standing layer.

[0068] In the first embodiment described above, the external magnetic field B is applied to the ferromagnetic metal layer 203 in a direction within the plane of the thermoelectric conversion element 200 (in the XY plane) and perpendicular to the direction of the flow of the heat flow H (see FIGS. 1 and 2 ), but it may also be applied in a direction within the plane of the thermoelectric conversion element 200 (in the XY plane) and intersecting the direction of the flow of the heat flow H at an angle other than 90 degrees. However, the external magnetic field B and the heat flow H are set so as not to be parallel or anti-parallel to each other.

[0069] Furthermore, in the first embodiment described above, thermoelectric conversion is performed while applying an external magnetic field B in step S13 (see FIGS. 1 and 2 ), but if the ferromagnetic metal layer 203 has magnetization 203M even without applying the external magnetic field B, thermoelectric conversion can be performed in step S13 without applying the external magnetic field B. Methods for achieving this include, for example, forming the ferromagnetic metal layer 203 from a material with a large coercive force so that the ferromagnetic metal layer 203 has magnetization 203M as remanent magnetization, and providing the thermoelectric conversion element 200 with a layer or member that applies a static magnetic field or an exchange coupling magnetic field to the ferromagnetic metal layer 203. For example, the thermoelectric conversion element 200 may further include an antiferromagnetic layer such as NiO, IrMn, FeMn, NiMn, PtMn, PdMn, MnAu, FeRh, or MnSb laminated on the side of the ferromagnetic metal layer 203 opposite to the paramagnetic metal layer 205 so as to apply an exchange coupling magnetic field to the ferromagnetic metal layer 203. This makes it possible to perform thermoelectric conversion without applying an external magnetic field B.

[0070] Furthermore, in the first embodiment described above, the pair of electrodes 207 a, 207 b are provided on the surface 205S of the paramagnetic metal layer 205 so as not to contact the ferromagnetic metal layer 203 and so as to be spaced apart from each other along the X-axis direction (see Figures 1 and 2), but they may also be provided on the side surfaces of the paramagnetic metal layer 205 in the positive and negative X-axis directions so as not to contact the ferromagnetic metal layer 203 and so as to be spaced apart from each other along the X-axis direction.

[0071] Furthermore, although the ferromagnetic metal layer 203 in the first embodiment described above was a single layer (see FIGS. 1 and 2 ), the ferromagnetic metal layer 203 may be composed of multiple stacked ferromagnetic metal layers. In this case, it is preferable that each of the multiple layers be composed of a Ni—Fe alloy, Fe, Co, Ni, Gd, a CoFeB alloy, or a (Ga,Fe)Sb alloy. Such multiple stacked ferromagnetic metal layers may be, for example, a laminate in which layers made of multiple types of ferromagnetic metals are periodically stacked on top of each other (e.g., a laminate in which three types of layers are periodically stacked, such as Layer A, Layer B, Layer C, Layer A, Layer B, Layer C, ..., where Layer A is a layer made of a first ferromagnetic metal, Layer B is a layer made of a second ferromagnetic metal, and Layer C is a layer made of a third ferromagnetic metal), or a laminate in which layers made of multiple types of ferromagnetic metals are randomly stacked on top of each other (e.g., a laminate in which three types of layers are randomly stacked, such as Layer A, Layer B, Layer C, Layer B, Layer A, Layer C, ...).

[0072] Although the paramagnetic metal layer 205 in the first embodiment described above was a single layer (see FIGS. 1 and 2 ), the paramagnetic metal layer 205 may instead be composed of a plurality of stacked paramagnetic metal layers. In this case, each of the plurality of layers is preferably composed of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy. Examples of such stacked paramagnetic metal layers include a laminate in which layers made of a plurality of types of paramagnetic metals are periodically stacked on top of each other (e.g., a laminate in which three types of layers are periodically stacked, such as Layer D, Layer E, Layer F, Layer D, Layer E, Layer F, ..., where Layer D is a layer made of a first paramagnetic metal, Layer E is a layer made of a second paramagnetic metal, and Layer F is a layer made of a third paramagnetic metal), or a laminate in which layers made of a plurality of types of paramagnetic metals are randomly stacked on top of each other (e.g., a laminate in which three types of layers are randomly stacked, such as Layer D, Layer E, Layer F, Layer E, Layer F, ...). The stacked paramagnetic metal layers preferably have spin Nernst angles of the same sign, because this prevents nonreciprocal thermoelectric signals resulting from the paramagnetic metal layers from canceling each other out during thermoelectric conversion in the thermoelectric conversion layer 202, thereby increasing the overall nonreciprocal thermoelectric signal.

[0073] Furthermore, in the first embodiment described above, the thermoelectric conversion layer 202 includes only one paramagnetic metal layer (paramagnetic metal layer 205) (see FIGS. 1 and 2 ). However, the thermoelectric conversion layer 202 may include a first paramagnetic metal layer and a second paramagnetic metal layer. In this case, the thermoelectric conversion layer 202 is configured by stacking the first paramagnetic metal layer, the ferromagnetic metal layer 203, and the second paramagnetic metal layer in this order along the Z-axis direction, with the ferromagnetic metal layer 203 interposed between the first paramagnetic metal layer and the second paramagnetic metal layer. The preferred materials for the first paramagnetic metal layer and the second paramagnetic metal layer are the same as those for the paramagnetic metal layer 205, and the configurations (constituent materials, film thickness, etc.) of the first paramagnetic metal layer and the second paramagnetic metal layer are selected so that the thermoelectric conversion layer 202 has a structure in which spatial inversion symmetry is broken along the Z-axis direction.

[0074] In this case, it is preferable that the first paramagnetic metal layer and the second paramagnetic metal layer have spin Nernst angles of opposite signs (for example, this can be achieved by configuring the first paramagnetic metal layer from Pt and the second paramagnetic metal layer from Ta). This is because, during thermoelectric conversion in the thermoelectric conversion layer 202, spin currents having the same polarity are injected from the first paramagnetic metal layer and the second paramagnetic metal layer into the ferromagnetic metal layer 203, thereby increasing the nonreciprocal thermoelectric signal.

[0075] Furthermore, in this case, either or both of the first and second paramagnetic metal layers may be composed of a plurality of paramagnetic metal layers stacked as described above. In this case, it is preferable that the plurality of paramagnetic metal layers constituting either or both of the first and second paramagnetic metal layers have spin Nernst angles of the same sign, for the reasons described above.

[0076] Furthermore, in the first embodiment described above, the pair of electrodes 207 a, 207 b are provided on the surface 205S of the paramagnetic metal layer 205 (see Figures 1 and 2), but the pair of electrodes 207 a, 207 b may also be provided on the back surface (the surface on the negative side of the Z axis) of the ferromagnetic metal layer 203 so as not to contact the paramagnetic metal layer 205 and so as to be spaced apart from each other along the X axis direction.

[0077] Furthermore, in the first embodiment described above, the thermoelectric conversion layer 202 is formed by stacking the ferromagnetic metal layer 203 and the paramagnetic metal layer 205 in this order on the substrate 201 (see FIGS. 1 and 2 ), but the thermoelectric conversion layer 202 may also be formed by stacking the paramagnetic metal layer 205 and the ferromagnetic metal layer 203 in this order on the substrate 201. In this case, the pair of electrodes 207 a, 207 b can be provided on the front surface of the ferromagnetic metal layer 203 (the surface on the positive side of the Z axis) or the back surface of the paramagnetic metal layer 205 (the surface on the negative side of the Z axis) so as to be spaced apart from each other along the X axis direction.

[0078] The thermoelectric conversion element, thermoelectric conversion module, and thermoelectric conversion method disclosed herein have the following configurations: [1] A thermoelectric conversion element comprising: a thermoelectric conversion unit having a nonlinear thermoelectric response characteristic with respect to an input of a heat flow along a first direction; and an electrode unit for extracting a thermoelectric signal generated in the thermoelectric conversion unit due to a heat flow input from the outside to the thermoelectric conversion unit along the first direction. [2] The thermoelectric conversion element according to [1], wherein the thermoelectric conversion unit exhibits nonreciprocal conduction based on breaking of spatial inversion symmetry, and the electrode unit comprises a pair of electrodes spaced apart from each other and provided on the thermoelectric conversion unit for extracting the nonreciprocal thermoelectric signal. [3] The thermoelectric conversion element according to [2], wherein the thermoelectric conversion unit is a thermoelectric conversion layer including a ferromagnetic metal layer and a paramagnetic metal layer stacked on each other along a second direction orthogonal to the first direction, and the pair of electrodes are spaced apart from each other and provided on the thermoelectric conversion layer along the first direction in an in-plane direction of the thermoelectric conversion layer. [4] The thermoelectric conversion element according to [2] or [3], wherein the distance between the pair of electrodes is 0.1 μm or more and 1000 μm or less. [5] The thermoelectric conversion element according to [3] or [4], wherein the paramagnetic metal layer is composed of a single layer or a plurality of laminated layers, and each of the single layer or the plurality of laminated layers of the paramagnetic metal layer is composed of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy. [6] The thermoelectric conversion element according to any of [3] to [5], wherein the ferromagnetic metal layer is composed of a single layer or a plurality of laminated layers, and each of the single layer or the plurality of laminated layers of the ferromagnetic metal layer is composed of a Ni—Fe alloy, Fe, Co, Ni, Gd, a CoFeB alloy, or a (Ga,Fe)Sb alloy. [7] A thermoelectric conversion module comprising a plurality of thermoelectric conversion elements according to any one of [1] to [6], wherein the plurality of thermoelectric conversion elements are electrically connected to each other so that thermoelectric signals from the respective thermoelectric conversion elements can be superimposed with the same polarity.[8] A thermoelectric conversion method comprising: a step of preparing a thermoelectric conversion element including a thermoelectric conversion layer exhibiting nonreciprocal conduction based on the breaking of spatial inversion symmetry, and a pair of electrodes spaced apart from each other and provided on the thermoelectric conversion layer for extracting nonreciprocal thermoelectric signals, wherein the thermoelectric conversion layer includes a ferromagnetic metal layer and a paramagnetic metal layer stacked on each other, and the pair of electrodes are provided on the thermoelectric conversion layer spaced apart from each other in an in-plane direction of the thermoelectric conversion layer; and a step of performing thermoelectric conversion while applying a magnetic field of 1 T or less to the ferromagnetic metal layer of the thermoelectric conversion layer. [9] A thermoelectric conversion element comprising: a thermoelectric conversion layer exhibiting nonreciprocal conduction based on the breaking of spatial inversion symmetry; and a pair of electrodes spaced apart from each other and provided on the thermoelectric conversion layer for extracting a nonreciprocal thermoelectric signal, wherein the thermoelectric conversion layer comprises a paramagnetic metal layer and a ferromagnetic metal layer stacked on each other, and the pair of electrodes are provided on the paramagnetic metal layer spaced apart from each other in an in-plane direction of the paramagnetic metal layer (excluding a thermoelectric conversion element comprising a layer made of an electrically insulating magnetic material provided in contact with the ferromagnetic metal layer).

[10] A thermoelectric conversion method comprising: a step of applying a temperature gradient including an AC component to a thermoelectric conversion unit exhibiting nonreciprocal conduction based on the breaking of spatial inversion symmetry; and a step of the thermoelectric conversion unit thermoelectrically converting the temperature gradient including the AC component based on the nonreciprocal conduction based on the breaking of spatial inversion symmetry to generate a nonreciprocal thermoelectric signal.

[11] A thermoelectric conversion method using a thermoelectric conversion unit including a first layer and a second layer, comprising: a step of applying a temperature gradient including an AC component to the thermoelectric conversion unit; and a step of thermoelectrically converting the temperature gradient including the AC component in the first layer to generate a non-reciprocal thermoelectric signal based on the fact that electron scattering of a spin current flowing into the first layer differs depending on the polarity direction of the spin current.

[12] A thermoelectric conversion element comprising: a thermoelectric conversion unit; a first heat conduction unit provided at one end of the thermoelectric conversion unit in a first direction; a second heat conduction unit provided at the other end of the thermoelectric conversion unit in the first direction; and a pair of electrodes provided on the thermoelectric conversion unit and spaced apart from each other in the first direction, wherein when a temperature gradient including an AC component is applied between the first heat conduction unit and the second heat conduction unit, a DC electrical signal is generated in the thermoelectric conversion unit, and the DC electrical signal can be extracted by the pair of electrodes.

[13] A thermoelectric conversion module including at least a first thermoelectric conversion element and a second thermoelectric conversion element, wherein the first thermoelectric conversion element and the second thermoelectric conversion element each include a thermoelectric conversion unit and a pair of electrodes spaced apart from each other on the thermoelectric conversion unit for extracting a non-reciprocal thermoelectric signal, and one of the pair of electrodes of the first thermoelectric conversion element and one of the pair of electrodes of the second thermoelectric conversion element are electrically connected to each other so that the non-reciprocal thermoelectric signal from the first thermoelectric conversion element and the non-reciprocal thermoelectric signal from the second thermoelectric conversion element can be superimposed with the same polarity.

[14] A thermoelectric conversion module including at least a first thermoelectric conversion element and a second thermoelectric conversion element, comprising: a first heat conducting portion provided at one end of the first thermoelectric conversion element and the second thermoelectric conversion element in a first direction, and a second heat conducting portion provided at the other end of the first thermoelectric conversion element and the second thermoelectric conversion element in the first direction.

[15] The thermoelectric conversion module according to

[13] or

[14] , for converting an applied temperature gradient including an AC component into a DC electrical signal.

[16] A thermoelectric conversion module including at least a first thermoelectric conversion element and a second thermoelectric conversion element, wherein the first thermoelectric conversion element and the second thermoelectric conversion element each include a thermoelectric conversion unit and a pair of electrodes for extracting a non-reciprocal thermoelectric signal, the pair of electrodes being spaced apart from each other on the thermoelectric conversion unit; and wherein a surface on which one of the pair of electrodes of the first thermoelectric conversion element is provided and a surface on which one of the pair of electrodes of the second thermoelectric conversion element is provided overlap in an area including the edges of the first thermoelectric conversion element and the second thermoelectric conversion element, thereby electrically connecting the one of the pair of electrodes of the first thermoelectric conversion element and the one of the pair of electrodes of the second thermoelectric conversion element.

[0079] Second Embodiment FIG. 7 is a perspective view showing a fluctuation detection sensor according to a second embodiment, and FIG. 8 is a cross-sectional view of the fluctuation detection sensor according to the second embodiment.

[0080] As shown in FIGS. 7 and 8, a fluctuation detection sensor 400 according to this embodiment includes a substrate 401, a nonlinear response layer 402 as a nonlinear response section, and a pair of electrodes 407a and 407b as an electrode section.

[0081] The substrate 401 is a plate- or film-shaped member, for example, as shown in FIG. 7 , and has a substantially flat surface 401S. The surface 401S is rectangular when viewed from the Z-axis direction, for example, as shown in FIG. 7 . The substrate 401 is a member for laminating a nonlinear response layer 402 on the surface 401S. Examples of materials that can be used to form the substrate 401 include metals such as Au, Ag, Cu, and Al, silicon, sapphire, SiC, GaN, yttria-stabilized zirconia (YSZ), and resins such as polyimide, polyethylene terephthalate (PET), and polyethylene naphthalate (PEN).

[0082] 7 and 8 show an orthogonal coordinate system C, in which the Z axis is set in the thickness direction of the substrate 401, and the X axis and Y axis are set in the long side and short side directions, respectively, of the rectangle of the surface 401S of the substrate 401. The orthogonal coordinate system C is also shown in the subsequent figures as needed.

[0083] The nonlinear response layer 402 has nonlinear electrical response characteristics with respect to input of external energy along the X-axis direction, which is a first direction. The external energy is, for example, an electric current and / or a heat flow. In this embodiment, the nonlinear response layer 402 includes a ferromagnetic layer 403 and a paramagnetic layer 405 stacked on each other along the Z-axis direction, which is a second direction. The ferromagnetic layer 403 is stacked on the substrate 401, and the paramagnetic layer 405 is stacked on the ferromagnetic layer 403. The ferromagnetic layer 403 is made of a material that exhibits ferromagnetism at the temperature in the environment in which the fluctuation detection sensor 400 is used, and is preferably made of a material that exhibits ferromagnetism at room temperature (300 K). In this embodiment, the ferromagnetic layer 403 is made of a hard magnetic material and may have spontaneous magnetization 403M. Furthermore, the ferromagnetic layer 403 preferably has a high coercive force, for example, a coercive force of 50 mT or more.

[0084] The ferromagnetic layer 403 is preferably a ferromagnetic metal layer. The ferromagnetic metal layer is an L1 layer made of a magnetic metal that exhibits ferromagnetism at room temperature (300 K) and a paramagnetic metal that exhibits paramagnetism. ０ The magnetic metal is preferably made of at least one of iron, nickel, cobalt, gadolinium, dysprosium, and terbium, and the paramagnetic metal is preferably made of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy. ０ By being composed of an ordered alloy of this type, the ferromagnetic layer 403 can have a particularly large coercive force, so that the energy conversion of current fluctuations and / or heat flow fluctuations by the fluctuation detection sensor 400 described below can be efficiently performed at room temperature in a magnetic field-free environment.

[0085] The nonlinear response layer 402 has, for example, a rectangular shape in plan view (i.e., when viewed from the Z-axis direction) with its long side along the X-axis direction and its short side along the Y-axis direction. The thickness of the ferromagnetic layer 403 is not particularly limited, but can be, for example, 1 nm or more and 100 nm or less.

[0086] The paramagnetic layer 405 is made of a material that exhibits paramagnetism at the temperature of the environment in which the fluctuation detection sensor 400 is used, and is preferably made of a material that exhibits paramagnetism at room temperature (300 K). Furthermore, the paramagnetic layer 405 is preferably a paramagnetic metal layer. The paramagnetic metal layer is preferably made of a paramagnetic metal with high spin-orbit interaction in order to enhance the spin-Nernst effect and spin-Hall effect within the paramagnetic metal layer, as described below. From this perspective, the paramagnetic layer is preferably made of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy, and is particularly preferably made of Pt. The thickness of the paramagnetic layer 405 is not particularly limited, but can be, for example, 1 nm or more and 100 nm or less.

[0087] In this embodiment, the nonlinear response layer 402 exhibits nonreciprocal conduction based on the breaking of spatial inversion symmetry. Specifically, a paramagnetic layer is provided on the surface of the ferromagnetic layer 403, but a similar paramagnetic layer is not provided on the back surface of the ferromagnetic layer 403. Also, a ferromagnetic layer is provided on the back surface of the paramagnetic layer 405, but a similar ferromagnetic layer is not provided on the surface of the paramagnetic layer 405. Therefore, the nonlinear response layer 402 has a structure in which spatial inversion symmetry is broken along the Z-axis direction. The nonlinear response layer 402 has substantial spatial inversion symmetry in the direction along the XY plane. Due to the breaking of spatial inversion symmetry along the Z-axis direction, the nonlinear response layer 402 of this embodiment exhibits nonreciprocity in the in-plane (XY plane) electrical conductivity.

[0088] The pair of electrodes 407a, 407b are provided on the nonlinear response layer 402 and spaced apart from each other in the in-plane (XY plane) direction of the nonlinear response layer 402. In this embodiment, the pair of electrodes 407a, 407b are provided on the paramagnetic layer 405 so as not to contact the ferromagnetic layer 403, and are provided on the surface 405S of the paramagnetic layer 405 opposite to the ferromagnetic layer 403. The pair of electrodes 407a, 407b are spaced apart from each other along the X-axis direction, which is one of the in-plane (XY plane) directions of the paramagnetic layer 405.

[0089] The pair of electrodes 407a, 407b are made of a material having conductivity that allows a non-reciprocal voltage signal to be extracted from the nonlinear response layer 402, as described below. Examples of such materials include metal materials such as Cu, Ag, Au, Pt, Ni, Al, constantan, Cr, In, Pd, Fe, Cu alloys, Ti / Au laminates, and Cr / Au laminates, and conductive oxides such as indium tin oxide (ITO) and zinc oxide (ZnO). Cu, Ag, Au, Pt, Ni, Al, constantan, and Cu copper alloys are preferred, and Cu, Au, Ag, Pt, Ni, Ti / Au laminates, and Cr / Au laminates are particularly preferred.

[0090] The thickness of the pair of electrodes 407 a, 407 b is not particularly limited, but may be, for example, 10 nm or more and 1 μm or less. The separation distance D407 between the pair of electrodes 407 a, 407 b along the X-axis direction is preferably 0.1 μm or more and 1000 μm or less.

[0091] Furthermore, the fluctuation detection sensor 400 of this embodiment may further include a layer made of an electrically insulating magnetic material that is arranged in contact with the ferromagnetic layer 403, or may not include such a layer.

[0092] When manufacturing such a fluctuation detection sensor 400, for example, a substrate 401 is prepared, and a material for forming the ferromagnetic layer 403 and a material for forming the paramagnetic layer 405 are deposited in this order on the surface 401S of the substrate by physical vapor deposition such as DC magnetron sputtering to form a laminate, and the laminate is patterned into a rectangular shape in plan view by photolithography and lift-off to form the nonlinear response layer 402. Thereafter, a metal material for forming the pair of electrodes 407a, 407b is deposited by physical vapor deposition such as DC magnetron sputtering in an area including the surface 405S of the paramagnetic layer 405 to form a metal layer, and the metal layer is patterned into a predetermined shape by photolithography and lift-off to form the pair of electrodes 407a, 407b.

[0093] Next, the energy conversion method according to this embodiment will be described. Fig. 9 is a process diagram showing the steps of the energy conversion method according to this embodiment. As shown in Fig. 9, the energy conversion method according to this embodiment includes a step S11 of preparing a fluctuation detection sensor having a nonlinear response layer, a step S12 of inputting current fluctuations or heat flow fluctuations to the nonlinear response layer, and a step S13 of converting the current fluctuations or heat flow fluctuations into energy. These steps will be described again with reference to Figs. 7 and 8.

[0094] In step S11 of preparing a fluctuation detection sensor having a nonlinear response layer, the above-mentioned fluctuation detection sensor 400 is prepared as shown in Figures 7 and 8. Next, in step S12 of inputting current fluctuations or heat current fluctuations to the nonlinear response layer, with the ferromagnetic layer 403 having spontaneous magnetization 403M, external current fluctuations Jf or heat current fluctuations Hf, which are the target of energy conversion, are input to the nonlinear response layer 402 so as to flow along the X-axis direction, which is one of the in-plane directions (in the XY plane) of the nonlinear response layer 402.

[0095] In this specification, "current fluctuation" refers to a phenomenon in which current fluctuates over time and space, and specifically refers to a situation in which a current in a certain direction (e.g., the X-axis direction) and a current in the opposite direction (e.g., the -X-axis direction) overlap each other, causing the direction of the net current to fluctuate over time and space. "Heat flow fluctuation" refers to a phenomenon in which heat flow fluctuates over time and space, and specifically refers to a situation in which a heat flow in a certain direction (e.g., the X-axis direction) and a heat flow in the opposite direction (e.g., the -X-axis direction) overlap each other, causing the direction of the net heat flow to fluctuate over time and space.

[0096] In step S12, when a current fluctuation Jf is input to the paramagnetic layer 405, a spin current J is generated by the spin Hall effect. Ｓ１ is generated along the Z-axis direction, and when the heat flow fluctuation Hf is input to the paramagnetic layer 405, a spin current J Ｓ１ is generated along the Z axis.

[0097] In step S13 of converting the energy of the current fluctuation or the heat current fluctuation, the spin current J generated as described above is Ｓ１flows into the ferromagnetic layer 403 in a state of having spontaneous magnetization 403M. Ｓ１ The polarity of changes in the positive direction of the Y-axis or the negative direction of the Y-axis depending on the direction of the current caused by the current fluctuation Jf (i.e., the net current direction of the current fluctuation Jf at a certain time and place) or the direction of the heat flow caused by the heat flow fluctuation Hf (i.e., the net heat flow direction of the heat flow fluctuation Hf at a certain time and place). Therefore, when the direction of the current or heat flow is reversed, the spin current J Ｓ１ The polarity of the spin current J Ｓ１ The relative orientation between the polarity of the magnetization 403M and the direction of the spontaneous magnetization 403M of the ferromagnetic layer 403 is reversed. Ｓ１ The relative orientation of the polarity of the magnetization 403M and the direction of the spontaneous magnetization 403M can also be reversed.

[0098] Spin current J Ｓ１ flows into the ferromagnetic layer 403, the spin current J Ｓ１ The electron scattering behavior changes depending on the relative orientation of the polarity of the spin current J in the vicinity of the junction interface between the ferromagnetic layer 403 and the paramagnetic layer 405. Ｓ１ Since the polarity direction of changes depending on the direction of the electric current or heat flow, the mode of this electron scattering changes depending on the direction of the electric current or heat flow, which causes the voltage signal as the electric response signal generated in the nonlinear response layer 402 to depend on the direction of the electric current or heat flow, i.e., nonreciprocity.

[0099] In this way, the voltage conversion signal J as an electrical response signal generated in the nonlinear response layer 402 Ｎ１ is non-reciprocal with respect to the direction of the electric current or heat flow due to the non-reciprocal conduction of the non-linear response layer 402 as described above (i.e., the voltage conversion signal J Ｎ１ (changes in proportion to the quadratic of the current fluctuation Jf or heat flow fluctuation Hf), and regardless of whether the direction of the current or heat flow is the positive direction of the X axis or the negative direction of the X axis, the voltage conversion signal J Ｎ１ The direction of the non-reciprocal voltage conversion signal J Ｎ１The inventors of the present application have found that the non-reciprocal voltage signal J can be electrically extracted via the pair of electrodes 407 a, 407 b. Therefore, according to the fluctuation detection sensor 400 of this embodiment, even if the direction or magnitude of the current caused by the current fluctuation Jf or the heat flow caused by the heat flow fluctuation Hf changes randomly, the non-reciprocal voltage signal J can be generated in a fixed direction within the non-linear response layer 402. Ｎ１ Therefore, it is possible to detect energy fluctuations from current fluctuations Jf or heat flow fluctuations Hf on a micro-scale.

[0100] When a current fluctuation Jf is input to the nonlinear response layer 402, a nonreciprocal voltage signal J is extracted by a pair of electrodes 407a and 407b in the nonlinear response layer 402. Ｎ１ is the magnitude of the voltage V, the electric field generated in the nonlinear response layer 402 corresponding to V is E, the distance between the pair of electrodes 407a and 407b is L (D407 in FIG. 8), and the current density in the range of the distance L in the nonlinear response layer 402 is j, then electric field E=nonlinear electric transport coefficient γ×electric resistivity ρ×(current density j) ２ By multiplying both sides of this equation by the distance L, the voltage V = L x γ x ρ x j ２ Therefore, the greater the distance L between the pair of electrodes 407a and 407b, the greater the voltage V. Therefore, by setting the separation distance D407 between the pair of electrodes 407a and 407b in the X-axis direction to a large value, preferably 1000 μm or more, as described above, the non-reciprocal voltage signal J extracted by the pair of electrodes 407a and 407b can be reduced. Ｎ１ becomes larger.

[0101] Similarly, when heat flow fluctuations Hf are input to the nonlinear response layer 402, a nonreciprocal voltage signal J is extracted by the pair of electrodes 407a and 407b in the nonlinear response layer 402. Ｎ１ is the magnitude of the voltage V, the electric field generated in the nonlinear response layer 402 corresponding to V is E, the nonlinear Seebeck coefficient of the nonlinear response layer 402 is S, the distance between the pair of electrodes 407a and 407b is L (D407 in FIG. 8), the temperature gradient within the range of the distance L in the nonlinear response layer 402 is ∇T, and the temperature difference within the range of the distance L in the nonlinear response layer 402 is ΔT, then electric field E=nonlinear Seebeck coefficient S×(temperature gradient ∇T)２ By multiplying both sides of this equation by the distance L, the voltage V = nonlinear Seebeck coefficient S × (temperature difference ΔT) ２ Therefore, the smaller the distance L between the pair of electrodes 407a and 407b, the larger the voltage V. Therefore, by setting the separation distance D407 between the pair of electrodes 407a and 407b in the X-axis direction to a small value, preferably 1000 μm or less, as described above, the non-reciprocal voltage signal J extracted by the pair of electrodes 407a and 407b can be reduced. Ｎ１ becomes larger.

[0102] The fluctuation detection sensor 400 may also include a heat conduction portion provided in contact with or close to the end face on the positive side of the X-axis and / or the end face on the negative side of the X-axis, for facilitating the flow of the current fluctuations Jf or the heat current fluctuations Hf into the nonlinear response layer 402. Such a heat conduction portion may be made of, for example, aluminum, copper, carbon fiber, sapphire, alumina, silicon with a thermal oxide film, or a high molecular weight polymer.

[0103] 10 is a perspective view showing the configuration of a fluctuation detection sensor module using the fluctuation detection sensor described above. As shown in FIG. 10, a fluctuation detection sensor module 500 according to this embodiment includes a plurality of fluctuation detection sensors 400, and in this embodiment, the fluctuation detection sensor module 500 includes four fluctuation detection sensors 400a, 400b, 400c, and 400d. The four fluctuation detection sensors 400a, 400b, 400c, and 400d receive non-reciprocal voltage signals J from the respective nonlinear electron transport voltage conversion elements. Ｎ１ are electrically connected to each other so as to overlap with each other with the same polarity.

[0104] Specifically, in the fluctuation detection sensor module 500, fluctuation detection sensors 400a and 400c are arranged with respect to the Cartesian coordinate system C in a manner similar to the fluctuation detection sensor 400 shown in Figures 7 and 8, and fluctuation detection sensors 400b and 400d are arranged with respect to the Cartesian coordinate system C in a manner obtained by rotating the fluctuation detection sensor 400 shown in Figures 7 and 8 by 180 degrees around the Y axis, and are arranged alternately along the Y axis. Then, the electrode 407a of the fluctuation detection sensor 400a and the electrode 407b of the fluctuation detection sensor 400b are electrically connected, the electrode 407a of the fluctuation detection sensor 400b and the electrode 407b of the fluctuation detection sensor 400c are electrically connected, and the electrode 407a of the fluctuation detection sensor 400c and the electrode 407b of the fluctuation detection sensor 400d are electrically connected. 7 and 8 with respect to the Cartesian coordinate system C, similar to the fluctuation detection sensors 400a and 400c, and the stacking order of the ferromagnetic layer 403 and the paramagnetic layer 405 of the nonlinear response layer 402 in the fluctuation detection sensors 400b and 400d can be reversed. In this case, the electrode 407a of the fluctuation detection sensor 400a is electrically connected to the electrode 407a of the fluctuation detection sensor 400b, the electrode 407b of the fluctuation detection sensor 400b is electrically connected to the electrode 407b of the fluctuation detection sensor 400c, and the electrode 407a of the fluctuation detection sensor 400c is electrically connected to the electrode 407a of the fluctuation detection sensor 400d.

[0105] The fluctuation detection sensor module 500 also includes heat conduction units 501 and 502 that are provided in contact with or close to the end faces on the negative side of the X-axis and the end faces on the positive side of the X-axis of the four fluctuation detection sensors 400a, 400b, 400c, and 400d. This makes it easier for the heat flow in the X-axis direction to flow into the ferromagnetic layers 403 of each of the four fluctuation detection sensors 400a, 400b, 400c, and 400d. The fluctuation detection sensor module 500 does not necessarily have to include the heat conduction units 501 and 502.

[0106] The current fluctuations or heat flow fluctuations flow into the ferromagnetic layers 403 of the four fluctuation detection sensors 400a, 400b, 400c, and 400d, and the non-reciprocal voltage signals generated in each fluctuation detection sensor 400 are superimposed with the same polarity and then extracted by the electrode 407b of the fluctuation detection sensor 400a and the electrode 407a of the fluctuation detection sensor 400d, thereby enabling the extraction of a large non-reciprocal voltage signal. The fluctuation detection sensor module 500 may also include more fluctuation detection sensors 400, in which case a larger non-reciprocal voltage signal may be extracted.

[0107] (Example 3: Input of Current Fluctuation) Next, to clarify the effect of the invention according to the second embodiment, examples 3 and 4 will be used for explanation. In example 3, an element corresponding to the fluctuation detection sensor 400 was fabricated as follows, corresponding to step S11 of preparing a fluctuation detection sensor having a nonlinear response layer. First, a magnesium oxide substrate having a thickness of 0.5 mm and a (111) orientation was prepared as the substrate 401, and a ferromagnetic layer 403 was formed on this substrate by co-sputtering using DC and RF magnetron sputtering methods, with Co at DC 37 W and Pt at RF 109 W at 0.046 nm / s, and a 7.5 nm-thick L1 ０ A CoPt alloy layer, which is a regular alloy of the type, was formed on the CoPt alloy layer. Furthermore, a 7.5 nm thick Pt layer was deposited on the CoPt alloy layer as a paramagnetic layer 405 to form the nonlinear response layer 402. Thereafter, photolithography and lift-off were used to pattern the nonlinear response layer 402 into a Hall element shape in plan view, that is, into a shape having a rectangular portion with a width of 10 μm in the Y-axis direction and a length of 40 μm in the X-axis direction in plan view, and two protrusions protruding from each end of the rectangular portion in the Y-axis direction. Thereafter, a Ti / Au layer is deposited on the nonlinear response layer 402 by RF magnetron sputtering, and the Ti / Au layer is patterned into a predetermined shape by photolithography and lift-off to form a pair of electrodes corresponding to the pair of electrodes 407a, 407b on the protrusion and a pair of electrodes 408a, 408b for an AC power source 418 described below at both ends of the rectangular portion in the X-axis direction, thereby producing an element corresponding to the fluctuation detection sensor 400.

[0108] Next, for the element thus fabricated, experiments were carried out corresponding to step S12 of inputting current fluctuations or heat flow fluctuations to the nonlinear response layer and step S13 of converting the energy of the current fluctuations or heat flow fluctuations. Specifically, a current fluctuation was input to the element at room temperature (300 K) in a manner corresponding to step S12, and a non-reciprocal voltage signal J was output from the pair of electrodes 407 a and 407 b. Ｎ１ An experiment was conducted to measure the non-reciprocal voltage signal J Ｎ１ 11 is a schematic diagram showing an experimental method for measuring fluctuation. As shown in Fig. 11, an AC power supply 418 was electrically connected between a pair of electrodes 408a and 408b provided on the fluctuation detection sensor 400. In addition, a lock-in amplifier 430 was electrically connected between a pair of electrodes 407a and 407b.

[0109] Then, while the magnitude of the external magnetic field B applied to the fluctuation detection sensor of Example 3 in the negative direction of the Y axis was changed from 2 T to -2 T, an AC current that sinusoidally changed with an angular frequency ω was applied in the direction along the X axis. Of the voltage signals generated by the fluctuation detection sensor of Example 3, ΔV ２ω、ｙ was detected by a lock-in amplifier 430.

[0110] More specifically, when ω is the angular frequency and t is the time, the AC power supply 418 generates an AC current I in the direction along the X axis of the fluctuation detection sensor 400. ＡＣ = (√2) I ｒｍｓ An alternating current expressed as sin(ωt) was applied (I ｒｍｓ means the effective value of AC current). ｒｍｓ When the current is 0.01mA, ｒｍｓExperiments were conducted for each of the cases where the current was 2.77 mA. A 6221-type AC / DC current source manufactured by Tektronix, Inc. in the United States was used as the AC power supply 418, and a lock-in amplifier LI5640 manufactured by NF Corporation was used as the lock-in amplifier 430. The frequency f (=ω / 2π) of the AC voltage applied by the AC power supply 418 was set to 13.423 Hz. The lock-in conditions for the lock-in amplifier 430 were a time constant of 1 second and a slope of -24 dB.

[0111] FIG. 12 shows the voltage signal ΔV obtained as a result of the above experiment in Example 3. ２ω、ｙ 10 is a graph showing the dependence of the external magnetic field B on the fluctuation detection sensor of Example 3. ＡＣ Since varies sinusoidally, the time average of the current flowing through the fluctuation detection sensor of Example 3 is zero (i.e., <I ＡＣ >= 0), and I ＡＣ corresponds to the current fluctuation Jf on a micro-scale. If the voltage signal output from the fluctuation detection sensor of Example 3 is an AC current I ＡＣ If the polarity of the AC current I ＡＣ A voltage signal ΔV corresponding to twice the frequency of ２ω、ｙ is the AC current I ＡＣ Since the voltage signals corresponding to the positive and negative phases of the AC current I cancel each other out, the voltage signal output from the fluctuation detection sensor of Example 3 becomes zero. ＡＣ When the voltage signal ΔV ２ω、ｙ is a finite value.

[0112] In this experiment, as shown in FIG. ｒｍｓ When the external magnetic field B is applied, a finite value of ΔV ２ω、ｙ is detected, and when the polarity of the external magnetic field B is reversed, ΔV ２ω、ｙ As shown by the arrow in FIG. 12, the polarity of the voltage signal ΔV ２ω、ｙ has hysteresis, and the voltage signal ΔV ２ω、ｙIt was revealed that the fluctuation detection sensor of Example 3 is capable of generating a voltage signal non-reciprocally relative to the direction of the current, and therefore it is possible to convert energy from the current fluctuation Jf to a non-reciprocal voltage signal even when the time average of the current is zero, even under conditions where an external magnetic field is not applied.

[0113] (Example 4: Input of heat flow fluctuation) Next, an experiment was carried out for Example 4. In Example 4, an element corresponding to the fluctuation detection sensor 400 was fabricated as follows, corresponding to step S11 of preparing a fluctuation detection sensor having a nonlinear response layer. First, a magnesium oxide substrate having a thickness of 0.5 mm and a (111) orientation was prepared as the substrate 401, and a ferromagnetic layer 403 was formed on this substrate by co-sputtering using DC 37 W and RF 109 W at 0.046 nm / s, and a L1 layer having a thickness of 7.5 nm was formed. ０ A CoPt alloy layer, which is a regular alloy of the CoPt type, was formed on the CoPt alloy layer. Furthermore, a 5 nm-thick Pt layer was deposited on the CoPt alloy layer as a paramagnetic layer 405 to form a nonlinear response layer 402. Then, photolithography and lift-off were used to pattern the nonlinear response layer 402 into a rectangular shape in plan view, with a width of 7.5 μm in the Y-axis direction and a length of 60 μm in the X-axis direction. A Ti / Au layer was then deposited on the nonlinear response layer 402 by RF magnetron sputtering, and the Ti / Au layer was patterned into a predetermined shape by photolithography and lift-off to form a pair of electrodes corresponding to the pair of electrodes 407 a and 407 b, thereby producing an element corresponding to the fluctuation detection sensor 400.

[0114] Next, for the element thus fabricated, experiments were carried out corresponding to step S12 of inputting current fluctuations or heat flow fluctuations to the nonlinear response layer and step S13 of converting the energy of the current fluctuations or heat flow fluctuations. Specifically, in step S12, heat flow fluctuations Hf were input to the element at room temperature (300 K), and a non-reciprocal voltage signal J was output from the pair of electrodes 407 a and 407 b. Ｎ１An experiment was conducted to measure the fluctuation. As shown in Fig. 13, a first resistive heater 410 was provided adjacent to the end of an element corresponding to the fluctuation detection sensor 400 of Example 4 in the negative X-axis direction, and a second resistive heater 411 was provided adjacent to the end of the element in the positive X-axis direction. A first power supply 420 and a second power supply 421 were electrically connected to the first resistive heater 410 and the second resistive heater 411, respectively. A lock-in amplifier 430 was also electrically connected between the pair of electrodes 407a and 407b.

[0115] Then, while the magnitude of the external magnetic field B applied to the fluctuation detection sensor of Example 4 in the negative direction of the Y axis is changed from 2 T to −2 T, an AC temperature gradient ΔT that changes sinusoidally with an angular frequency ω in the direction along the X axis is generated by the first resistance heater 410 and the second resistance heater 411. ＡＣ Then, among the thermoelectric signals generated by the fluctuation detection sensor of Example 4, the AC temperature gradient ΔT ＡＣ ΔV is the y component of the change in the thermoelectric signal corresponding to twice the frequency of ２ω、ｙ was detected by a lock-in amplifier 430.

[0116] AC temperature gradient ΔT ＡＣ More specifically, when ω is the angular frequency and t is the time, the first power supply 420 supplies V to the first resistance heater 410. ｄｃ +V ａｃ An AC voltage of sin(ωt) is applied, and V ｄｃ -V ａｃ An AC voltage of sin(ωt) was applied to the second resistance heater 411 by the second power supply 421. As a result, a voltage ΔT was applied to the fluctuation detection sensor 400 in the direction along the X axis. ＡＣ = (√2) ΔT ｒｍｓ An AC temperature gradient expressed as sin(ωt) was applied (ΔT ｒｍｓis the effective value of the AC temperature gradient). A voltage source WF1948 manufactured by NF Corporation was used as the first power supply 420 and the second power supply 421, and a lock-in amplifier LI5640 manufactured by NF Corporation was used as the lock-in amplifier 430. The frequency f (=ω / 2π) of the AC voltage applied by the first power supply 420 and the second power supply 421 was set to 3.723 Hz. The lock-in conditions for the lock-in amplifier 430 were a time constant of 1 second and a slope of -24 dB.

[0117] FIG. 14 shows the voltage signal ΔV obtained as a result of the above experiment in Example 4. ２ω、ｙ 10 is a graph showing the dependence of ΔT on the external magnetic field B. ＡＣ Since varies sinusoidally, when viewed on a macro scale, the fluctuation detection sensor of Example 4 is not substantially subjected to a temperature gradient, and ΔT ＡＣ corresponds to the temperature fluctuation on a micro scale. If the voltage signal output from the fluctuation detection sensor of Example 4 is ＡＣ If the polarity of the AC temperature gradient ΔT ＡＣ A voltage signal ΔV corresponding to twice the frequency of ２ω、ｙ is the AC temperature gradient ΔT ＡＣ Since the thermoelectric signals corresponding to the positive and negative phases of the temperature gradient ΔT are cancelled out, the voltage signal output from the fluctuation detection sensor of Example 4 is zero. ＡＣ When the voltage signal ΔV ２ω、ｙ is a finite value.

[0118] In this experiment, as shown in FIG. 14, when an external magnetic field B is applied, a finite value of ΔV ２ω、ｙ is detected, and when the polarity of the external magnetic field B is reversed, ΔV ２ω、ｙ The polarity of the voltage signal ΔV ２ω、ｙIt was revealed that the fluctuation detection sensor of Example 4 is capable of generating a voltage signal non-reciprocally with respect to the direction of the temperature gradient, and therefore it is possible to convert energy from heat flow fluctuations Hf at a microscale where a temperature gradient does not substantially exist at a macroscale into a non-reciprocal voltage signal.

[0119] The present invention is not limited to the second embodiment described above, and various modifications are possible.

[0120] For example, in the second embodiment described above, the fluctuation detection sensor 400 includes the substrate 401 (see FIGS. 7 and 8), but the fluctuation detection sensor 400 does not need to include the substrate 401. In this case, the nonlinear response layer 402 can be formed as a free-standing layer.

[0121] Furthermore, in the second embodiment described above, in step S12 of inputting current fluctuations or heat flow fluctuations to the nonlinear response layer, either current fluctuations Jf or heat flow fluctuations Hf are input to the nonlinear response layer 402 (see Figures 7 and 8), but both current fluctuations Jf and heat flow fluctuations Hf may be input to the nonlinear response layer 402 in a superimposed manner.

[0122] Furthermore, in the second embodiment described above, step S12 of inputting current fluctuations or heat flow fluctuations to the nonlinear response layer and step S13 of converting the energy of the current fluctuations or heat flow fluctuations are performed when the ferromagnetic layer 403 has spontaneous magnetization 403M. However, these steps may also be performed by forming the ferromagnetic layer 403 from a soft magnetic material and applying an external magnetic field to the ferromagnetic layer 403 to magnetize the ferromagnetic layer 403 to the same extent as the spontaneous magnetization 403M.

[0123] Furthermore, in the second embodiment described above, the pair of electrodes 407 a, 407 b are provided on the surface 405S of the paramagnetic layer 405 so as not to contact the ferromagnetic layer 403 and to be spaced apart from each other along the X-axis direction (see FIGS. 7 and 8 ). However, the electrodes may also be provided on the side surfaces of the paramagnetic layer 405 in the positive and negative X-axis directions so as not to contact the ferromagnetic layer 403 and to be spaced apart from each other along the X-axis direction.

[0124] Although the paramagnetic layer 405 in the second embodiment described above was a single layer (see FIGS. 7 and 8 ), the paramagnetic layer 405 may instead be composed of multiple stacked layers of paramagnetic metals. In this case, each of the multiple layers is preferably composed of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy. Examples of such stacked multiple paramagnetic metal layers include a laminate in which layers of multiple types of paramagnetic metals are periodically stacked on top of each other (e.g., a laminate in which three types of layers are periodically stacked, such as Layer D, Layer E, Layer F, Layer D, Layer E, Layer F, etc., where Layer D is a layer of a first paramagnetic metal, Layer E is a layer of a second paramagnetic metal, and Layer F is a layer of a third paramagnetic metal), or a laminate in which layers of multiple types of paramagnetic metals are randomly stacked on top of each other (e.g., a laminate in which three types of layers are randomly stacked, such as Layer D, Layer E, Layer F, Layer E, Layer F, etc.). The stacked multiple paramagnetic metal layers preferably have spin Nernst angles of the same sign, because this prevents nonreciprocal voltage signals resulting from the multiple paramagnetic metal layers from canceling each other out during energy conversion in the nonlinear response layer 402, thereby increasing the overall nonreciprocal voltage signal.

[0125] Furthermore, in the second embodiment described above, the nonlinear response layer 402 includes only one paramagnetic layer (paramagnetic layer 405) such as a paramagnetic metal layer (see FIGS. 7 and 8 ). However, the nonlinear response layer 402 may include a first paramagnetic layer such as a paramagnetic metal layer and a second paramagnetic layer such as a paramagnetic metal layer. In this case, the nonlinear response layer 402 is configured by stacking the first paramagnetic layer, the ferromagnetic layer 403, and the second paramagnetic layer in this order along the Z-axis direction, with the ferromagnetic layer 403 interposed between the first and second paramagnetic layers. The preferred materials for the first and second paramagnetic layers are the same as those for the paramagnetic layer 405; however, it is preferable to select the configurations (materials, film thickness, etc.) of the first and second paramagnetic layers so that the nonlinear response layer 402 has a structure in which spatial inversion symmetry is broken along the Z-axis direction.

[0126] In this case, it is preferable that the first paramagnetic layer and the second paramagnetic layer have spin Nernst angles of opposite signs (for example, this can be achieved by configuring the first paramagnetic layer from Pt and the second paramagnetic layer from Ta). This is because, during energy conversion in the nonlinear response layer 402, spin currents of the same polarity are injected from the first paramagnetic layer and the second paramagnetic layer to the ferromagnetic layer 403, thereby increasing the nonreciprocal voltage signal.

[0127] Furthermore, in this case, either or both of the first and second paramagnetic layers may be made of a plurality of paramagnetic metal layers stacked as described above. In this case, it is preferable that the plurality of paramagnetic metal layers constituting either or both of the first and second paramagnetic layers have spin Nernst angles of the same sign for the reasons described above.

[0128] Furthermore, in the second embodiment described above, the pair of electrodes 407 a, 407 b are provided on the surface 405S of the paramagnetic layer 405 (see Figures 7 and 8), but the pair of electrodes 407 a, 407 b may also be provided on the back surface (the surface on the negative side of the Z axis) of the ferromagnetic layer 403 so as not to contact the paramagnetic layer 405 and to be spaced apart from each other along the X axis direction.

[0129] Furthermore, in the second embodiment described above, the ferromagnetic layer 403 and the paramagnetic layer 405 are laminated in this order on the substrate 401 to form the nonlinear response layer 402 (see FIGS. 7 and 8 ), but the nonlinear response layer 402 may also be formed by laminating the paramagnetic layer 405 and the ferromagnetic layer 403 in this order on the substrate 401. In this case, the pair of electrodes 407 a, 407 b can be provided on the front surface of the ferromagnetic layer 403 (the surface on the positive side of the Z axis) or the back surface of the ferromagnetic layer 405 (the surface on the negative side of the Z axis) so as to be spaced apart from each other along the X-axis direction.

[0130] The fluctuation detection sensor disclosed herein has the following configuration. [1] A fluctuation detection sensor comprising: a nonlinear response unit having a nonlinear electrical response characteristic with respect to an input of external energy along a first direction; and an electrode unit for extracting an electrical response signal generated in the nonlinear response unit due to energy fluctuations input to the nonlinear response unit from the outside along the first direction. [2] The fluctuation detection sensor according to [1], wherein the external energy is an electric current and / or a heat flow, and the energy fluctuations are current fluctuations and / or heat flow fluctuations. [3] The fluctuation detection sensor according to [1] or [2], wherein the external energy is an electric current, and the energy fluctuations are current fluctuations. [4] The fluctuation detection sensor according to any of [1] to [3], wherein the nonlinear response unit has a configuration in which spatial inversion symmetry is broken along a second direction orthogonal to the first direction. [5] The fluctuation detection sensor according to any of [1] to [4], wherein the nonlinear response unit includes a ferromagnetic layer and a paramagnetic layer stacked on top of each other along the second direction. [6] The fluctuation detection sensor according to any one of [1] to [5], wherein the ferromagnetic layer is configured to have spontaneous magnetization. [7] The ferromagnetic layer is a ferromagnetic metal layer, and L1 is made of a magnetic metal and a paramagnetic metal. ０ The fluctuation detection sensor according to any one of [1] to [6], wherein the fluctuation detection sensor is made of an ordered alloy of the type, and the paramagnetic layer is a paramagnetic metal layer.

[0131] 200... thermoelectric conversion element, 202... thermoelectric conversion layer, 203... ferromagnetic metal layer, 205... paramagnetic metal layer, 207a, 207b... pair of electrodes.

Claims

1. A thermoelectric conversion element comprising: a thermoelectric conversion unit having a nonlinear thermoelectric response characteristic to a heat flow input along a first direction; and an electrode unit for extracting a thermoelectric signal generated in the thermoelectric conversion unit due to a heat flow input from the outside to the thermoelectric conversion unit along the first direction.

2. A thermoelectric conversion element as described in claim 1, wherein the thermoelectric conversion unit exhibits nonreciprocal conduction based on the breaking of spatial inversion symmetry, and the electrode unit comprises a pair of electrodes spaced apart from each other on the thermoelectric conversion unit for extracting a nonreciprocal thermoelectric signal.

3. A thermoelectric conversion element as described in claim 2, wherein the thermoelectric conversion section is a thermoelectric conversion layer including a ferromagnetic metal layer and a paramagnetic metal layer stacked on top of each other along a second direction perpendicular to the first direction, and the pair of electrodes are provided on the thermoelectric conversion layer in an in-plane direction of the thermoelectric conversion layer, spaced apart from each other along the first direction.

4. A thermoelectric conversion element according to claim 2 or 3, wherein the distance between the pair of electrodes is 0.1 μm or more and 1000 μm or less.

5. A thermoelectric conversion element as described in claim 3, wherein the paramagnetic metal layer is composed of a single layer or a plurality of laminated layers, and each of the single layer or the plurality of laminated layers of the paramagnetic metal layer is composed of Pt, Pd, W, an AuW alloy, Ta, a CuIr alloy, a CuBi alloy, a BiSb alloy, or a BiSe alloy.

6. A thermoelectric conversion element according to claim 3 or 5, or claim 4 when dependent on claim 3, wherein the ferromagnetic metal layer is composed of a single layer or a plurality of laminated layers, and each of the single layer or the plurality of laminated layers of the ferromagnetic metal layer is composed of a Ni-Fe alloy, Fe, Co, Ni, Gd, a CoFeB alloy, or a (Ga,Fe)Sb alloy.

7. A thermoelectric conversion module comprising a plurality of thermoelectric conversion elements according to any one of claims 1 to 6, the plurality of thermoelectric conversion elements being electrically connected to each other so that thermoelectric signals from each thermoelectric conversion element can be superimposed with the same polarity.

8. A thermoelectric conversion method comprising the steps of: applying a temperature gradient including an AC component to a thermoelectric conversion unit exhibiting nonreciprocal conduction based on the breaking of spatial inversion symmetry; and causing the thermoelectric conversion unit to thermoelectrically convert the temperature gradient including the AC component to generate a nonreciprocal thermoelectric signal.

Images