JPH0758367A - Superconducting element - Google Patents

Abstract

(57) [Summary] [Formation] An incident electrode 3 for injecting a plane wave electron wave and a normal conducting electrode 18 are provided on the Si semiconductor 2 so as to face each other, and superconducting conductive electrodes 4 and 5 are provided between them. It has a structure in which a gate electrode 6 is provided on a semiconductor surrounded by superconducting conductive electrodes 4 and 5 via a gate insulating film so as to face each other so as to be orthogonal to the traveling direction of waves. The width W of the electron waveguide 13 is set to the gate electrode 6
The number of modes of the passing electron wave is continuously modulated by changing the voltage applied to. According to the present invention, it is possible to provide a superconducting element which operates at high speed and consumes less power. Further, since the size of one element is minute and the operation can be performed with a small number of elements, high integration can be realized.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a superconducting device which has functionality and can be highly integrated, and operates at high speed and with low power consumption.

[0002]

2. Description of the Related Art A typical quantum effect device using the interference of electron waves is described in Physical Review Letter, (Phys. Rev. Let.) 55, 2344 (1985) by S. Datta et al. Some are being discussed. This device consists of island-shaped AlGaAs between electron waveguides made of GaAs.
Is provided, the electron waveguide is branched into two, and the interference of electron waves propagating through the respective electron waveguides is detected.

In addition, the quantum effect device using a superconductor is
There is one disclosed in JP-A-1-49830. This device is basically based on a structure in which an electron wave incident electrode 3 and a superconducting conductive electrode 1 are provided on a semiconductor 2 as shown in FIG. The electron wave that spreads from the incident electrode 3 into the semiconductor undergoes Andreev reflection at the boundary between the superconducting conductive electrode 1 and the semiconductor 2,
It bounces back like the light reflected on the concave boundary and returns to the original path and converges on the incident electrode. As shown in FIG. 11, when the gate electrode 4 is provided on the semiconductor between the incident electrode 3 and the superconducting electrode 1, the carrier concentration in the semiconductor near the gate electrode changes due to the change in the gate voltage. As a result, the superconducting proximity effect causes the superconducting electrode 1 to the semiconductor 2
Since the size of the region of the superconducting pair potential penetrating therethrough changes, the position where the electron wave is reflected by Andreev moves to change the path difference. Therefore, the interference resulting from the superposition of electron waves can be controlled.

[0004]

In the above-mentioned conventional quantum effect device using a semiconductor, it is not possible to configure the width or length and the impurity concentration of the bisected electron waveguide to be the same, and There is a problem that the phase of the electron wave varies. Also, GaAs and A corresponding to the wall of the electron waveguide
The influence of elastic scattering due to the roughness of the 1 GaAs boundary surface was large, and it was difficult to improve the coherence. Furthermore, when the two electron waveguides merge, scattering occurs between the electron waves, so it is necessary to devise a method for converging the electron waves.

Further, it has been necessary to suppress the spatial spread of the electron wave in the electron wave of the quantum effect element using the conventional superconductor to enhance the coherence.

A first object of the present invention is to collimate an electron wave to limit the number of modes of the electron wave, and further to set the number of modes to 0.
It is to provide a superconducting element that can be continuously changed from 1 to n.

A second object of the present invention is to provide a superconducting device which enhances the coherence of electron waves and operates at high speed and low power consumption with a simple structure of one device.

A third object of the present invention is to provide a superconducting element which controls the propagation direction of electron waves by an electric field or a current and operates at high speed with low power consumption.

A fourth object of the present invention is to provide a multiplier capable of performing parallel arithmetic processing and operating at high speed and low power consumption.

[0010]

A first object of the present invention is to use a semiconductor as a conductor of an electron wave, and to provide two superconductors and a boundary surface of the semiconductor in parallel to a traveling direction of the electron wave traveling in the semiconductor. This is achieved by constructing an electronic waveguide. Further, the width of the waveguide surrounded by the interface between the two superconductors and the semiconductor can be changed by the voltage applied to the gate electrode provided on the semiconductor via the insulating film, and as a result, the electron The number of wave modes can be continuously changed from 0 to n.

A second object is to provide an island-shaped superconductor in a conductor made of a semiconductor so as to divide an electron waveguide of an electron wave traveling in the semiconductor into two, and to provide each of the electron conductors. The waveguide is composed of two boundaries of a superconducting conductive electrode and a semiconductor, and a superconducting element having a function for giving a phase difference between electron waves propagating in respective electron waveguides and a function of joining the two electron waveguides. Achieved by

A third object is to provide a plurality of superconducting electrodes or superconducting electrodes for supplying electron waves into the semiconductor with a plurality of incident electrodes facing each other and a plurality of superconducting electrodes so as to face each other. A diffraction grating having a periodically changing potential, which is composed of a superlattice in which a plurality of superconductors and semiconductors are alternately and periodically arranged at right angles to the traveling direction of the electron wave, and the electrons are applied by a voltage applied to the diffraction grating. It is easily achieved by modulating the direction of wave travel.

A fourth object is to provide a plurality of incident electrodes for supplying electron waves into the semiconductor and a superconducting conductive electrode so as to face each other on the semiconductor, and to arrange the electrons in the semiconductor on the semiconductor between the incident electrode and the superconducting electrode. This is easily achieved by including a plurality of means for controlling the phase of the wave for one incident electrode, and surrounding the electron waveguide of the electron wave incident from the incident electrode with a superconducting conductive electrode.

[0014]

The superconducting proximity effect and a special quantum phenomenon called Andreev reflection occur at the interface between the superconductor and the semiconductor, so that the electron waves in the semiconductor are less susceptible to elastic scattering due to the roughness of the interface. This is because the effective boundary surface between the semiconductor and the superconductor is determined by the superconducting proximity effect.

The superconducting proximity effect and Andreev reflection will be described with reference to the energy diagram near the boundary between the superconductor and the semiconductor shown in FIG. When the superconductor and the semiconductor are brought into contact with each other, the superconductor pair penetrates from the superconductor side to the semiconductor side, and quasi-particles permeate from the semiconductor to the superconductor side. As a result, the electron pair density is reduced in the superconductor, and the electron pairs are present on the semiconductor side. This is the superconducting proximity effect. In the figure, the spatial change in superconductivity is represented by the pair potential Δ. The pair potential is the product of the probability amplitude of the superconducting pair of electrons and the interaction potential between the electrons, and is, so to speak, an index of superconductivity. The range of exudation of the pair potential into the semiconductor corresponds to the coherence length in the semiconductor, and this length changes in proportion to the third power of the carrier concentration in the semiconductor.

At the boundary, Andreyev reflection occurs in the process of conversion of electrons in the semiconductor into superconducting electron pairs.
Regarding this reflection, Soviet Physics J.T.P. (Vol. 19), 1964
It is described on page 228. The Andreev reflection is a phase conjugate reflection in quantum mechanics. If the excitation energy of the electrons entering the superconductor from the semiconductor side based on the energy corresponding to the Fermi level is smaller than the pair potential of the superconductor, there is no state corresponding to the excitation energy in the superconductor, No single electron can enter the superconductor from the boundary. However, if an incident electron condenses with another electron having opposite momentum and spin by Bose condensation, it can enter the superconductor as a superconductor pair. Along with this, holes having a momentum in the opposite direction to the incident electrons return in the semiconductor along the same path as the incident electrons.

FIG. 13 shows the structure of an electron waveguide showing the operation of the present invention. In FIG. 13, 1 and 5 are superconducting electrodes, 2
Is a semiconductor and 3 is an incident electrode. The conductor of the electron waveguide 13 is composed of the semiconductor 2, and the wall of the waveguide is constituted by the boundary surface between the superconducting conductive electrodes 1 and 5 and the semiconductor 2. When an electron wave having energy lower than the gap energy of the superconducting electrode is incident from the incident electrode 3 at an incident angle in the semiconductor 2, the boundary surface between the superconducting electrode 1 and the semiconductor 2, the superconducting electrode 5 and the semiconductor 2 will be described.
Then, it receives Andreev reflection at the boundary surface of and returns to the incident electrode 3. That is, only the electrons that do not collide with the superconducting electrode 1 or 5 can pass through the electron waveguide. Therefore, the number of modes of the electron wave is limited, and a collimated electron wave having a small spatial spread can be taken out.

Further, when the electron wave passes through the electron waveguide 13 having a width approximately equal to its wavelength, the conductance of the waveguide is quantized and takes an integral multiple of e ² / π _h . This corresponds to the quantum conductance. That is, as the width of the electron waveguide 13 increases, the number of modes of electrons that can pass increases. In the electron waveguide of FIG. 13, due to the superconducting proximity effect, the pair potential leaks from the superconducting conductive electrodes 1 and 5 to the semiconductor 2 side. When the carrier concentration in the semiconductor is changed by applying an electric field, the coherence length in the semiconductor changes in proportion to the third power of the carrier concentration in the semiconductor, so that the effective width of the electron waveguide is Change. Therefore, a quantum wire can be realized by applying an electric field, and the number of electron modes can be controlled. At this time, the conductance of the electron waveguide 13 changes stepwise, and the size of the step is the minimum unit of quantum conductance.

By using two electron waveguides, it is possible to cause interference of electron waves. In this case, the electron waves are collimated, and elastic scattering due to roughness at the boundary surface of the superconductor semiconductor is also suppressed, so that the coherence of the electron waves can be enhanced.

On the other hand, when superconductors and semiconductors are alternately arranged as shown in FIG. 14 to form a superlattice, electron waves traveling in a direction parallel to the superlattice are diffracted. This is because the pair potential in the direction orthogonal to the traveling direction of the electron wave changes periodically. Degree of modulation of pair potential P
Is determined by the change in height of the pair potential in each superconductor and the semiconductor and the pitch of the lattice (width shown by C and D in FIG. 14). The modulation factor P changes the diffraction rate of the electron wave. FIG. 15 shows the relationship between the potential modulation degree P and the maximum diffraction index. A superconducting element can be realized by controlling this diffraction phenomenon with an electric field. The modulation degree P of the potential can be changed by the above-mentioned superconducting proximity effect. That is, the magnitude of the pair potential in the semiconductor sandwiched by the superconductor can be changed by changing the carrier concentration in the semiconductor. Therefore, it is possible to realize an element that modulates the diffraction index by the electric field.

The multiplier capable of parallel processing is composed of a plurality of incident electrodes for inputting a plurality of input signals, and a gate electrode for performing weighting W _ai and W _bi twice corresponding to each input signal. ing. Since the multiplication of each input signal is performed in the electron waveguide surrounded by the boundary surface between the superconducting conductive electrode and the semiconductor, parallel processing is possible. Further, by connecting a superconducting transistor, this can be used as a threshold portion, compared with a predetermined threshold value, and an output signal can be output as a voltage based on the comparison result. Further, since the threshold characteristic is differentiable, the backpropagation method used as a learning method of the neuron element can be used, and as a result, the weight can be controlled. Further, the threshold value can be easily changed by an external voltage.

Since the superconducting element shown above operates at a voltage of several millivolts corresponding to the gap energy of the superconducting electrode, the power consumption can be reduced. Further, since the device uses the control of interference of electron waves, high speed operation is possible. Furthermore, it is possible to realize the functions of the circuit and the system, which cannot be performed by the conventional element, with a simple configuration of one element.

In order to efficiently control the above-mentioned reflection phenomenon, the incident electrode must be devised. In order to suppress carrier scattering when an electron wave is incident, it is necessary to provide the incident electrode 3 on the semiconductor via a minute junction having a radius of 10 times or less than the mean free path of carriers in the semiconductor. The distance traveled by the electron wave in the semiconductor is preferably selected to be 10 times or less the average free path of carriers in the semiconductor. Therefore, since the superconducting element has a fine size, it can be highly integrated.

In order to operate the superconducting material even at high temperature,
It is desirable that the oxide superconducting material has a high superconducting transition temperature. Superconducting material having n-type conductivity, (N
It is preferable to use a superconductor having a composition of d, Ce) ₂ CuO _4, but in addition to this, the Nd portion of the above composition is replaced with Pr, Pm,
It is replaced by at least one element selected from the group of Sm, Eu, Gd and Er, or the part of Ce is replaced by at least one element selected from the group of Th, Tl, Pb and Bi. May be. p
It is preferable to use a Y-based oxide superconductor which is a superconducting material having type conductivity, but in place of this, a La-based oxide superconductor, a Bi-based oxide superconductor, and a Tl-based oxide superconductor are used. Can be used to achieve the object of the present invention.

Further, a metallic superconducting material, specifically N
b, Pb, Pb alloy, or Nb metal compound may be used.

Si is preferably used as the material of the semiconductor, but instead of this, Ge, GaAs, InSb, InA
The object of the present invention can also be achieved by using s, InP. Further, the same effect can be obtained by using a material having a semiconducting property, which is different in composition ratio from the oxide superconductor but has a similar crystal structure.

[0027]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 1 shows a planar structure of a superconducting element which is a first embodiment of the present invention. In this element, an incident electrode 3 made of Nb for injecting a plane wave electron wave and a normal conduction electrode 18 are provided on the Si semiconductor 2 so as to face each other, and superconducting conductive electrodes 4 and 5 are provided between them. It has a structure in which a gate electrode 6 made of polycrystalline Si is provided on the semiconductor surrounded by the superconducting conductive electrodes 4 and 5 via a gate insulating film so as to face each other so as to be orthogonal to the traveling direction.

The distance between the incident electrode 3 and the normal conducting electrode 18 is shorter than the mean free path of the semiconductor 2. Therefore, the electron waveguide is in the ballistic conduction state. The region of the semiconductor 2 surrounded by the boundary surface between the superconducting conductive electrode 4 and the semiconductor 2 and the boundary surface between the superconducting conductive electrode 5 and the semiconductor 2 constitutes an electron waveguide 13 having a width W. An electron wave that is incident on the semiconductor 2 at a constant incident angle from the incident electrode 3 with energy equal to or lower than the gap energy of the superconducting electrode is subjected to Andreev reflection at the boundary surface between the superconducting electrode 4 or 5 and the semiconductor 2, and the incident electrode 3 Return to. That is, only the electrons that do not collide with the superconducting conductive electrode 4 or 5 can pass through the electron waveguide. Therefore, the collimated electron wave having a small spatial spread of the electron wave advances toward the normal conduction electrode 18. Therefore, the spatial distribution of the phase of the electron wave can be suppressed. Further, at the boundary surface between the superconducting conductive electrode 4 or 5 and the semiconductor 2, the effective boundary surface is flattened by the superconducting proximity effect. This allows
Electron waves in semiconductors are less susceptible to elastic scattering such as interface roughness.

Further, the width W of the electron waveguide 13 can be changed by the voltage applied to the gate electrode 6.
Due to the superconducting proximity effect, a pair potential leaks out from the superconducting electrodes 4 and 5 into the electron waveguide 13 on the semiconductor 2 side. When the voltage applied to the gate electrode 6 is increased and the carrier concentration in the electron waveguide 13 is increased, the coherence length L in the semiconductor is increased. As a result, the effective width W of the electron waveguide changes to (W-2L). On the other hand, when the width W of the electron waveguide becomes about the same as the wavelength of the electron wave in the semiconductor 2, the conductance of the waveguide is quantized. In the case of a Si semiconductor, W is quantized when it becomes about 0.1 μm or less, and the number of modes of electrons that can pass decreases as the width of the electron waveguide decreases.

FIG. 2 shows the electron waveguide 13 with respect to the gate voltage.
Shows the change in conductance of. Since the coherence length L in the semiconductor changes depending on the gate voltage, the mode number of the electron wave can be continuously changed from 0 to n. That is, the conductance changes stepwise, and the size of the step is the minimum unit of quantum conductance.

FIG. 3 shows current-voltage characteristics between the incident electrode 3 and the normal conducting electrode 18. The electron wave that has entered the semiconductor 2 from the incident electrode 3 passes through the electron waveguide 13, and only the electron wave that does not collide with the superconducting conductive electrode 4 or 5 reaches the normal conducting electrode 18. A current flows between the incident electrode 2 and the normal conduction electrode 18, but when the gate voltage increases, the number of modes of the electron wave passing through the electron waveguide 13 changes, so that the current flowing between the electrodes also varies stepwise depending on the gate voltage. Changes to. By using this phenomenon, it is possible to realize an analog / digital converter that converts an analog signal into a digital signal.

According to the present embodiment, the device current can be greatly changed by a small change of the gate voltage, and the quantized change of the conductance in the minimum unit is used. Therefore, a superconducting device with low power consumption can be realized. You can Further, since the size of one element is minute, high integration can be achieved.

FIG. 4 shows a sectional structure of a superconducting element which is a second embodiment of the present invention. In this device, a gate electrode 17 made of polycrystalline Si and having a film thickness of 100 nm is provided between an output electrode 15 and an incident electrode 3 made of a normal conductor or a superconductor for injecting a plane wave electron wave onto a Si semiconductor 2. Gate insulating film 16, N made of a thermally oxidized Si film and having a thickness of 10 nm
b, a superconducting electrode 12 having a thickness of 100 nm, an electron waveguide 14 made of a Si semiconductor and having a thickness of 70 nm, an island-shaped superconducting electrode 11 made of Nb having a thickness of 50 nm, and Si.
It has a structure in which an electron waveguide 13 made of a semiconductor and having a thickness of 70 nm and a superconducting conductive electrode 10 made of Nb and having a thickness of 100 nm are sequentially stacked. This is an element that utilizes the interference phenomenon of two electron waveguides surrounded by a boundary surface between a superconductor and a semiconductor.

An electron wave incident on the semiconductor from the incident electrode 3 at a constant incident angle with an energy equal to or less than the gap energy of the superconducting electrode is directed to the two electron waveguides 13 and 14 by the island-shaped superconducting electrode 11. Branched. Since the electron waves propagating in the respective electron waveguides are surrounded by the superconducting conductive electrodes, only electrons that do not collide with the walls of the superconducting conductive electrodes 10, 11, 12 can pass through the electron waveguides.
Further, at the boundary surface between the superconducting electrode and the semiconductor, elastic scattering due to the roughness of the interface is less likely to occur. Therefore, a collimated electron wave with little spatial spread of the electron wave is output electrode 15
Proceed toward. Therefore, the spatial distribution of the phase of the electron wave can be suppressed. The branched electron waves are merged again and converged on the output electrode 15. Where the gate electrode 1
By applying a voltage to 7, the electron waveguides 13 and 14
The electrostatic potential between these electron waves changes, and a phase difference occurs. Therefore, the interference of electron waves is controlled by the gate voltage.

FIG. 5 shows current-voltage characteristics between the incident electrode 3 and the output electrode 15. Due to interference, the current changes periodically with respect to the voltage. When the gate voltage is further changed, the peak position of this current changes. That is, the currents show the same value under different control conditions, and this can be made to correspond to multivalued logic of two or more values to operate.

According to this embodiment, the device current can be largely changed by a small change in the gate voltage, so that a superconducting device with low power consumption and high speed can be realized.
Moreover, since the electron waves are collimated, the interference effect can be enhanced. Further, since the size of one element is minute, high integration can be achieved.

FIG. 6 shows a planar structure of a superconducting element which is a third embodiment of the present invention. This device has the same characteristics and effects as the device shown in the second embodiment, but the semiconductor 2
On the other hand, two electron waveguides 13 and 14 are formed in a plane. An incident electrode 3 made of Nb for injecting a plane wave electron wave on the Si semiconductor 2 and a superconducting conductive electrode 1 are opposed to each other, and a superconducting conductive electrode 10 made of three Nb is interposed between the electrodes.
By providing 11 and 12 with a distance of 0.1 μm, a semiconductor electron waveguide 13 surrounded by superconducting electrodes 10 and 11 and a semiconductor electron waveguide 14 surrounded by superconducting electrodes 11 and 12 are formed. is doing. Further, by providing the gate electrode 6 on one electron waveguide 14 via the gate insulating film, the carrier concentration in the electron waveguide 14 can be changed.

An electron wave incident on the semiconductor from the incident electrode 3 at a constant incident angle with an energy equal to or less than the gap energy of the superconducting pole is supplied by the superconducting pole 11 to the two electron waveguides 1.
It is branched to 3,14. Since the electron waves propagating in the respective electron waveguides are surrounded by the superconducting conductive electrodes, only electrons that do not collide with the walls of the superconducting conductive electrodes 10, 11, 12 can pass through the electron waveguides. Further, at the boundary surface between the superconducting conductive electrode and the semiconductor, the electron wave in the semiconductor is less susceptible to elastic scattering such as the roughness of the interface. Therefore, a collimated electron wave with a small spatial spread of the electron wave travels toward the superconducting electrode 1. At the interface between the superconducting electrode 1 and the semiconductor 2, Andreev reflection occurs, and it is focused on the electron wave incident electrode 3. Here, when a voltage is applied to the gate electrode 6, the wavelength of the electron wave is modulated by the change of the carrier concentration in the electron waveguide 14, and a phase difference is generated between the electron waves of the electron waveguides 13 and 14. Therefore, the interference of electron waves is controlled by the gate voltage. As a result of the interference phenomenon, the current changes periodically with respect to the voltage. When the gate voltage is further changed, the peak position of this current changes. That is, the currents show the same value under different control conditions, and this can be made to correspond to multivalued logic of two or more values to operate.

According to this embodiment, the device current can be largely changed by a small change in the gate voltage, so that a superconducting device with low power consumption and high speed can be realized.
Moreover, since the electron waves are collimated, the interference effect can be enhanced. Further, since the size of one element is minute, high integration can be achieved.

FIG. 7 shows a planar structure of a superconducting element which is the fourth embodiment of the present invention. This device includes an incident electrode 3 for injecting a plane wave electron wave onto a Si semiconductor 2, an Nb / Si / Nb / Si / Nb ...
The lateral superlattice 29, 2 of the diffraction grating consisting of the combination of ...
It is composed of superconducting electrodes 21 and 23 and a superconducting electrode 22. Below the superlattice 29, the gate electrode 17 is provided via a gate insulating film. The distance from the incident electrode to the superlattice 29 is equal to or less than the phase coherence length of the electron wave, and the superlattice 29 is in the direction orthogonal to the traveling direction of the electron wave.
The periodic potential of changes. The superlattice 29 is an artificially arranged scattering source, and the potential is spatially a periodic potential structure consisting of a pair potential ΔNb of Nb and a pair potential ΔSi exuded from Nb into Si by the superconducting proximity effect. It is formed and acts as a diffraction grating for incident electron waves.

When the electron wave in the semiconductor 2 passes through the superlattice 29, the propagation direction is deflected by 45 degrees due to diffraction. The deflection efficiency corresponds to the maximum diffraction efficiency, and its change is shown in FIG. The deflection efficiency changes depending on the modulation degree P of the potential of the superlattice 29. The modulation degree P of the potential is determined by the height change of the pair potential in each superconductor and the semiconductor and the pitch of the lattice. The modulation degree P of this potential can be changed by the voltage applied to the gate electrode 17. This is because the magnitude of the pair potential induced by the superconducting proximity effect in the semiconductor sandwiched by the superconductors can be changed by changing the carrier concentration in the semiconductor with the gate voltage. Therefore, the modulation degree P of the potential changes with the gate voltage, and the deflection rate can be modulated.

For example, when the gate voltage changes by 0.05V, the diffraction efficiency changes from 10% to 80%. Therefore, the two groups of superconducting electrodes 21 and 23 and the superconducting electrode 22 are spatially arranged so that diffracted electrons and non-diffracted electrons are discriminated and collected.
Thereby, the current flowing between the superconducting electrodes 21 and 23 or between the superconducting electrode 22 and the incident electrode 3 is controlled by the gate voltage.

The characteristics of this element are shown in FIG. The horizontal axis of FIG. 8 represents the gate voltage, and the vertical axis represents the current flowing between the superconducting conductive electrode 22 and the incident electrode 3 and the current flowing between the superconducting conductive electrode 21 and the incident electrode 3. When the gate voltage changes from 0 to 0.05 V, the current flowing between superconducting electrode 22 and incident electrode 3 changes from 20 mA to 3 mA, and the current flowing between superconducting electrode 21 and incident electrode 3 changes from 3 mA to 20 mA.

According to the present embodiment, the device current can be largely changed with a small change in the control potential, so that a superconducting device with low power consumption can be realized. Further, since it is a superconducting element whose propagation is controlled by using the diffraction phenomenon of a plane wave, the propagation direction of the electron wave can be switched without impairing the high speed of the electron wave, and a high-speed element can be realized.

FIG. 9 shows a planar structure of a superconducting device which is a fifth embodiment of the present invention. This device includes an incident electrode 3 for injecting a plane wave electron wave onto a Si semiconductor 2, an Nb / Si / Nb / Si / Nb ...
The horizontal superlattice 29 of the diffraction grating is composed of a combination of ... And superconducting transistors 24, 25, and 26 of two groups.

This embodiment is different from the element shown in the fourth embodiment in that a superconducting transistor is used for detecting an output signal, and the operation principle and effect of the element are the same. By coupling the superconducting transistor, it can be used as a threshold and compared with a predetermined threshold value, and an output signal can be output as the voltage of the superconducting transistor based on the comparison result. Further, since the threshold characteristic is differentiable, the backpropagation method used as a learning method for the neuron element can be used, and the threshold value can be easily changed by an external voltage.

According to the present embodiment, the device current can be largely changed with a small change in the control potential, so that a superconducting device with low power consumption can be realized. Further, since it is a superconducting element whose propagation is controlled by using the diffraction phenomenon of a plane wave, the propagation direction of the electron wave can be switched without impairing the high speed of the electron wave, and a high-speed element can be realized.

In the fourth and fifth embodiments, a combination of a superconductor and a semiconductor is used as the superlattice 29. Instead of this, a superlattice composed of a plurality of superconductors having different superconducting transition temperatures is used. However, the object of the present invention can be achieved.

In the fourth and fifth embodiments, the deflection rate is modulated by the gate voltage, but the object of the present invention can be achieved by using a method of directly injecting a current into the superlattice.

FIG. 10 shows a planar structure of a superconducting element which is a sixth embodiment of the present invention. This element is composed of a plurality of electron wave incident electrodes 3 facing a superconducting electrode 1 made of Nb on a Si semiconductor 2, and Nb to which an input signal X _i is inputted.
And an input electrode 20 made of polycrystalline Si, and weighting parts 31 and 32 made of polycrystalline Si via an insulating film for weighting W _ia and W _ib corresponding to the respective input signals. It is performed in an electron waveguide surrounded by a superconducting electrode.

For example, the electron wave incident from the leftmost incident electrode 3 is surrounded by the superconducting electrodes 33 and 34, and the electron waveguide 35.
Is calculated by. That is, the input signal X ₁ has the weighting unit 31,
After being weighted twice at W _1a and W _1b at 32, the superconducting pole 1 is reached. The weighting is realized by changing the carrier concentration in the semiconductor by the voltage applied to the weighting units 31 and 32 and giving a phase difference to the electron waves. That is, the output result is the multiplication result of X _1x W _1ax W _1b .

Since each electron waveguide is surrounded by superconducting electrodes, they do not interfere with each other. Therefore, since the calculation is simultaneously performed for each input signal X _i in each electron waveguide, massively parallel processing is possible. Further, by forming the second superconducting electrode 30 facing the other end of the superconducting electrode 1 and providing the gate electrode 6 between the superconducting electrodes with the insulating film interposed therebetween, the superconducting transistor 25 is coupled to do this. It can be used as a threshold and compared with a predetermined threshold value, and an output signal can be output as a voltage between the superconducting electrode 1 and the superconducting electrode 30 based on the comparison result. In addition, since the threshold characteristic is differentiable, the back propagation method used as a learning method for the neuron element can be used, and the threshold is set to the superconducting transistor 2
It can be easily changed with a gate voltage of 5.

According to this embodiment, each multiplication is performed at a voltage below the gap energy of the superconductor.
This has the effect of realizing a multiplier that operates with low power consumption. Further, since the size of one element is minute and parallel processing of multiplication can be performed with a simple configuration of one element, high integration and high speed can be achieved.

In this embodiment, a superconductor is used as the electrode for injecting electrons, but the same effect can be obtained by using a normal conductor. Further, even if the means for injecting carriers is configured by using a superconducting electrode via a tunnel barrier layer, the same effect can be obtained.

[0055]

According to the present invention, it is possible to provide a superconducting element which operates at high speed and consumes less power. Further, since the size of one element is minute and the operation can be performed with a small number of elements, high integration can be realized.

[Brief description of drawings]

FIG. 1 is a plan view of a superconducting element according to a first embodiment of the present invention.

FIG. 2 is a characteristic diagram of the superconducting element according to the first embodiment of the present invention.

FIG. 3 is a characteristic diagram of the superconducting element of the first embodiment of the present invention.

FIG. 4 is a sectional view of a superconducting element of a second embodiment of the present invention.

FIG. 5 is a characteristic diagram of the superconducting element of the second embodiment of the present invention.

FIG. 6 is a plan view of a superconducting element according to a third embodiment of the present invention.

FIG. 7 is a plan view of a superconducting element according to a fourth embodiment of the present invention.

FIG. 8 is a characteristic diagram of a superconducting element according to a fourth embodiment of the present invention.

FIG. 9 is a plan view of a superconducting device according to a fifth embodiment of the present invention.

FIG. 10 is a plan view of a superconducting device according to a sixth embodiment of the present invention.

FIG. 11 is an explanatory view of a superconducting element according to a conventional technique.

FIG. 12 is an energy diagram showing the operation of the present invention.

FIG. 13 is an explanatory view of a superconducting element showing the operation of the present invention.

FIG. 14 is an explanatory view of a superconducting element showing the operation of the present invention.

FIG. 15 is a characteristic diagram showing the operation of the present invention.

[Explanation of symbols] 4, 5 ... Superconducting electrode, 18 ... Normal conducting electrode, 2 ... Semiconductor,
3 ... Incident electrode, 6 ... Gate electrode.

Claims (15)

[Claims] 1. A superconducting device in which one or a plurality of incident electrodes made of a superconductor or a normal conductor for supplying an electron wave into the semiconductor and a normal conducting electrode or a first superconducting electrode face each other on a semiconductor, An electron waveguide including two boundary surfaces of the superconductor and the semiconductor is formed on the semiconductor between the incident electrode and the normal electrode or the first superconducting electrode in parallel with the traveling direction of the electron wave. As described above, the second superconducting electrode is provided so as to face each other. 2. The width of the electron waveguide formed at the boundary between the second superconducting conductive electrode and the semiconductor provided opposite to each other is equal to or less than the mean free path of conduction electrons in the semiconductor. Smaller superconducting element. 3. The superconducting element according to claim 2, wherein a width of the electron waveguide is controlled by a voltage applied to a gate electrode provided on the semiconductor. 4. A structure is provided in which an island-shaped superconductor is provided inside a conductor made of a semiconductor to divide an electron waveguide of an electron wave traveling in the semiconductor into two, and a boundary between the respective electron waveguides. Is composed of two superconducting conductive electrodes and a semiconductor, and has a means for giving a phase difference between the electron waves propagating in the respective electron waveguides, and a function of joining the two electron waveguides after branching. Characteristic superconducting element. 5. The superconducting element according to claim 4, comprising means for applying an electric field to at least one of the electron waveguides as means for providing the phase difference. 6. The superconducting element according to claim 4, comprising means for applying a magnetic field as means for applying the phase difference. 7. A plurality of incident electrodes made of a superconductor or normal conductor for supplying electron waves into the semiconductor and a plurality of superconducting electrodes are provided on a semiconductor so as to face each other, and between the incident electrodes and the superconducting electrodes. Providing a diffraction grating in which a plurality of superconductors and semiconductors are alternately and periodically arranged so that the pair potential changes periodically at a right angle to the traveling direction of the electron wave, and the electron wave travels by the voltage applied to the diffraction grating. A superconducting element characterized by modulating the direction. 8. The superconducting element according to claim 7, wherein the diffraction grating comprises a superlattice in which a plurality of superconductors having different superconducting transition temperatures are alternately arranged. 9. A plurality of incident electrodes made of a superconductor or normal conductor for supplying electron waves into the semiconductor and a plurality of superconducting electrodes are provided on a semiconductor so as to face each other, and between the incident electrodes and the superconducting electrodes. Providing a diffraction grating in which a plurality of superconductors and semiconductors are alternately and periodically arranged so that the pair potential changes periodically at a right angle to the traveling direction of the electron wave, and the electron wave travels by the current injected into the diffraction grating. A superconducting element characterized by modulating the direction. 10. The superconducting element according to claim 7,
A superconducting element in which a second superconducting electrode is provided facing the superconducting electrode on the semiconductor via an insulating film. 11. The superconducting element according to claim 7,
A superconducting device comprising a second superconducting electrode facing the superconducting electrode on a semiconductor, and comprising a gate electrode provided between the superconducting electrodes via an insulating film. 12. A plurality of incident electrodes for supplying an electron wave into the semiconductor and a superconducting electrode are provided on a semiconductor so as to face each other, and the phase of the electron wave in the semiconductor is provided on the semiconductor between the incident electrode and the superconducting electrode. 2. A superconducting element, comprising a plurality of means for controlling a plurality of incident electrodes, wherein an electron waveguide of an electron wave incident from the incident electrode is surrounded by a superconducting conductive electrode. 13. The superconducting element according to claim 12, wherein the means for controlling the phase of the electron wave in the semiconductor comprises a gate electrode for applying an electric field to the semiconductor to change the carrier concentration. 14. The superconducting device according to claim 12, wherein a plurality of gate electrodes are provided for one incident electrode, and a change in voltage to the incident electrode is used as an input signal, and each of the gate electrodes is provided with the gate electrode. A superconducting device that processes the multiplication of input signals in parallel by simultaneously performing the operation of applying a plurality of weights in each electron waveguide by applying a voltage. 15. The gate electrode according to claim 12, wherein the superconducting element is provided with a second superconducting electrode facing the superconducting electrode on the semiconductor, and an insulating film is provided between the superconducting electrodes. A superconducting element composed of.