WO2021254830A1 - Neuronen und synapsen mit ferroelektrisch modulierten metall-halbleiter schottky dioden nebst verfahren - Google Patents

Neuronen und synapsen mit ferroelektrisch modulierten metall-halbleiter schottky dioden nebst verfahren Download PDF

Info

Publication number
WO2021254830A1
WO2021254830A1 PCT/EP2021/065297 EP2021065297W WO2021254830A1 WO 2021254830 A1 WO2021254830 A1 WO 2021254830A1 EP 2021065297 W EP2021065297 W EP 2021065297W WO 2021254830 A1 WO2021254830 A1 WO 2021254830A1
Authority
WO
WIPO (PCT)
Prior art keywords
layer
electrode
component according
ferroelectric
source electrode
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/EP2021/065297
Other languages
German (de)
English (en)
French (fr)
Inventor
Qing-Tai Zhao
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Forschungszentrum Juelich GmbH
Original Assignee
Forschungszentrum Juelich GmbH
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Forschungszentrum Juelich GmbH filed Critical Forschungszentrum Juelich GmbH
Priority to CN202180035992.XA priority Critical patent/CN115668226A/zh
Priority to JP2022577164A priority patent/JP2023534389A/ja
Priority to EP21731993.8A priority patent/EP4133419B1/de
Priority to US18/007,721 priority patent/US20230231030A1/en
Publication of WO2021254830A1 publication Critical patent/WO2021254830A1/de
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Classifications

    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10DINORGANIC ELECTRIC SEMICONDUCTOR DEVICES
    • H10D64/00Electrodes of devices having potential barriers
    • H10D64/60Electrodes characterised by their materials
    • H10D64/64Electrodes comprising a Schottky barrier to a semiconductor
    • H10D64/647Schottky drain or source electrodes for IGFETs
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/06Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/063Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using electronic means
    • G06N3/065Analogue means
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/21Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
    • G11C11/22Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using ferroelectric elements
    • G11C11/223Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using ferroelectric elements using MOS with ferroelectric gate insulating film
    • GPHYSICS
    • G11INFORMATION STORAGE
    • G11CSTATIC STORES
    • G11C11/00Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
    • G11C11/54Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using elements simulating biological cells, e.g. neuron
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10BELECTRONIC MEMORY DEVICES
    • H10B51/00Ferroelectric RAM [FeRAM] devices comprising ferroelectric memory transistors
    • H10B51/30Ferroelectric RAM [FeRAM] devices comprising ferroelectric memory transistors characterised by the memory core region
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10DINORGANIC ELECTRIC SEMICONDUCTOR DEVICES
    • H10D30/00Field-effect transistors [FET]
    • H10D30/60Insulated-gate field-effect transistors [IGFET]
    • H10D30/611Insulated-gate field-effect transistors [IGFET] having multiple independently-addressable gate electrodes influencing the same channel
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10DINORGANIC ELECTRIC SEMICONDUCTOR DEVICES
    • H10D30/00Field-effect transistors [FET]
    • H10D30/60Insulated-gate field-effect transistors [IGFET]
    • H10D30/701IGFETs having ferroelectric gate insulators, e.g. ferroelectric FETs
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10DINORGANIC ELECTRIC SEMICONDUCTOR DEVICES
    • H10D64/00Electrodes of devices having potential barriers
    • H10D64/20Electrodes characterised by their shapes, relative sizes or dispositions 
    • H10D64/23Electrodes carrying the current to be rectified, amplified, oscillated or switched, e.g. sources, drains, anodes or cathodes
    • H10D64/251Source or drain electrodes for field-effect devices
    • H10D64/258Source or drain electrodes for field-effect devices characterised by the relative positions of the source or drain electrodes with respect to the gate electrode
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10DINORGANIC ELECTRIC SEMICONDUCTOR DEVICES
    • H10D64/00Electrodes of devices having potential barriers
    • H10D64/60Electrodes characterised by their materials
    • H10D64/66Electrodes having a conductor capacitively coupled to a semiconductor by an insulator, e.g. MIS electrodes
    • H10D64/68Electrodes having a conductor capacitively coupled to a semiconductor by an insulator, e.g. MIS electrodes characterised by the insulator, e.g. by the gate insulator
    • H10D64/689Electrodes having a conductor capacitively coupled to a semiconductor by an insulator, e.g. MIS electrodes characterised by the insulator, e.g. by the gate insulator having ferroelectric layers
    • HELECTRICITY
    • H10SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10DINORGANIC ELECTRIC SEMICONDUCTOR DEVICES
    • H10D8/00Diodes
    • H10D8/60Schottky-barrier diodes 
    • GPHYSICS
    • G06COMPUTING OR CALCULATING; COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00Computing arrangements based on biological models
    • G06N3/02Neural networks
    • G06N3/06Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
    • G06N3/063Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using electronic means

Definitions

  • the invention relates to a synaptic component for a neural network.
  • a synaptic component is an electronic component that has the properties of a biological synapse. In the following, the electronic component is therefore also referred to as a synapse.
  • neural network is meant an artificial neural network.
  • a brain is a biological neural network. It comprises a large number of biological neurons, i.e. nerve cells.
  • a biological neuron can send information to another biological neuron via biological synapses. This is done with the help of an output channel of the biological neuron. Such an exit channel is called an axon.
  • the other biological neuron has an input channel. The entrance channel is called the dendrite. The other biological neuron can receive the information via the input channel.
  • the decision as to whether a biological neuron triggers an action potential via its axon and thus sends information depends on the entirety of the incoming signals.
  • the influence of one biological neuron on another biological neuron can change over time with their activity. Such an influence is called the “synaptic weight” of a biological neuron.
  • the properties and behaviors of a neural network are said to be similar to the properties and behaviors of a biological neural network.
  • a neural network should be able to recognize patterns using adaptive algorithms. Similar to the human brain, a neural network should not have to start over for every task or problem. It should have the opportunity to fall back on already acquired knowledge and experiences.
  • a neural network with memristors is known from document CN 208922326 U. Scaling, stability and blocking ability are problematic in a neural network with memristors.
  • the document WO 2019147859 A2 discloses an electronic component with a semiconductor channel. An input electrode adjoins one end of the semiconductor channel. An output electrode adjoins the other end of the semiconductor channel. Above the semiconductor channel there is a layer of dielectric material. A circuit made of semiconductor materials with an electrically charged carrier gas and used as a synapse is used in the Document EP 0 529 565 B1 described. An integrated circuit for providing a synapse is known from the document US 2019164597 A1.
  • a neural network can include ferroelectric field effect transistors (FeFET).
  • FeFET ferroelectric field effect transistors
  • Traditional ferroelectric materials such as lead zirconate titanate (PZT) are harmful to the environment and incompatible with established CMOS technology.
  • a synaptic component with the features of patent claim 1 serves to solve the problem.
  • the dependent claims relate to advantageous embodiments of the invention.
  • the independent claim relates to a method for operating the synaptic component.
  • Claim 1 relates to a synaptic component for a neural network.
  • the synaptic component comprises a layer made of a semiconductor.
  • the layer consisting of a semiconductor is called a semiconducting layer.
  • a source electrode is connected to the semiconducting layer.
  • a drain electrode is connected to the semiconducting layer.
  • the source electrode is spatially separated from the drain electrode.
  • the source electrode and the semiconducting layer form a Schottky diode.
  • the source electrode is separated from a first gate electrode by a ferroelectric material.
  • the drain electrode can be separated from a second gate electrode by a ferroelectric material.
  • the two gate electrodes are spatially separated from one another.
  • An electrode is separated from another electrode when there is a distance between the two electrodes. Spatially separated electrodes do not touch each other. There is therefore no electrically conductive connection between the two electrodes. With the aid of the ferroelectric material, an electrical voltage applied to the first gate electrode can be transferred to the source electrode. The same applies to the case that a second gate electrode is separated from the drain electrode by ferroelectric material. As a rule, one gate electrode is located on one side of the ferroelectric material and the other source or drain electrode is located on an opposite side of the ferroelectric material. The two electrodes can be directly connected to the ferroelectric material.
  • the invention uses the memory Properties of ferroelectric material to mimic the behavior of a biological synapse.
  • a Schottky diode can comprise a metal layer which is applied to a semiconducting layer such as a silicon layer, for example.
  • a metal layer is a layer made of metal.
  • a silicon layer is a layer made of silicon.
  • Silicon is a semiconductor, i.e. a chemical substance whose electrical conductivity is between that of electrical conductors and that of electrical insulators.
  • the silicon layer can be, for example, n-silicon, that is to say an n-type doped silicon layer. Electrons from the n-conducting silicon layer migrate to the metal layer. Because electrons from n-conducting silicon get into the metal layer more easily than vice versa, an electron-depleted area is created in the silicon layer. This area is called the Schottky barrier. This creates a barrier layer or a space charge zone. In addition, an electric field is created. The electric field counteracts the migration of the electrodes. If the electric field is large enough, electrons no longer migrate from the n-conducting silicon layer to the metal layer.
  • the silicon layer can, however, also be p-silicon, that is to say a p-type doped silicon layer.
  • the Schottky diode is switched in the direction of flow. Electrons flow from n-silicon into the metal layer. If the Schottky diode is switched in the forward direction, the space charge zone is cleared.
  • An electrode is an electrical conductor.
  • An electrode is basically made of metal.
  • a layer according to the invention can run in a straight line, that is to say along a plane.
  • a layer according to the invention can, however, also run round and / or run at an angle.
  • the component With the component with the features of claim 1, the behavior of a biological synopsis can be imitated.
  • the component can be manufactured with little technical effort.
  • the component can be made from environmentally friendly materials.
  • the source electrode and the drain electrode can be on one side of the semiconducting layer. Such a component can be manufactured in a technically simple manner.
  • the source electrode and the drain electrode can be produced on a surface of the semiconducting layer by deposition.
  • the source electrode and the drain electrode can be at opposite ends of the semiconducting layer. In this way, a suitable distance can be provided between the source electrode and the drain electrode.
  • the component can comprise a substrate.
  • the semiconducting layer is then applied above the substrate.
  • the semiconducting layer can be a thin layer.
  • the substrate can consist of silicon.
  • An electrically insulating layer can be located between the substrate and the semiconductor. The electrically insulating layer enables disruptive influences to be avoided.
  • Ferroelectric material of the component can be in the form of a layer.
  • a layer of ferroelectric material can therefore be applied to the electrode.
  • a layer of ferroelectric material can therefore be applied to the drain electrode.
  • a ferroelectric layer can be located partially on the semiconducting layer in order to enable a technically simple production.
  • a ferroelectric layer can be located partially on the source electrode or drain electrode. It is not absolutely necessary for two spatially separated layers of ferroelectric material to be present. It can be a single layer of ferroelectric material which is at least partially on the source electrode and at least partially on the drain
  • Electrode is in place. This can also facilitate manufacture.
  • a first and / or second gate electrode can be applied on the opposite side of such a ferroelectric layer.
  • the first gate electrode and / or the second gate electrode are located on one side of the ferroelectric layer.
  • Electrode and / or the drain electrode are located on the opposite side of the ferroelectric layer.
  • a drain electrode made of metal can serve as a substrate for the synaptic component.
  • Semiconductor material can first be applied to the substrate.
  • the metal for a source electrode can be applied above the semiconductor material.
  • Ferroelectric material can be applied to the source electrode.
  • metal can be applied for a gate electrode. A vertical construction is possible.
  • One or more electrically insulating layers can be provided during manufacture in order to manufacture a synaptic component in a technically simple manner.
  • an insulating layer with recesses can be applied to a semiconducting layer.
  • the semiconducting layer can be contacted through the recesses.
  • the recesses can be at least partially coated with metal in order to produce Schottky diodes, for example.
  • the insulating layer with the recesses contributes to the fact that the two electrodes are electrically separated from one another.
  • the production of one or more insulating layers can also be advantageous in the case of a vertical construction.
  • the electrically insulating layer can cover the semi-material as long as the semiconductor material is not to be contacted by metallic electrodes.
  • a multiplicity of synaptic components which are electrically connected in parallel and / or electrically in series with one another can be produced in a simple manner.
  • a so-called crossbar structure can be produced, which can be advantageous in neural networks.
  • the semiconductor material can be selected from: Si, Ge, SiGe, SiGeSn, GeSn, SiC.
  • the semiconductor material can be an III-V compound semiconductor, II-VI compound semiconductor.
  • the semiconductor material can be a 2D material, which consists of only one layer or only a few layers of breaths and molecules.
  • the semiconductor material can be a substrate or a semiconductor layer on a substrate, such as a "silicon-on-insulator" (SOI).
  • substrate is meant a self-supporting layer which can therefore serve as a support for other layers. For the production of the component, a substrate can therefore be assumed to which further layers can be applied.
  • the semiconductor material can be a semiconductor heterostructure with a plurality of semiconductor layers.
  • the ferroelectric material can be selected from: HfÜ2-based ferroelectric, perovskite ferroelectric and organic ferroelectric.
  • the ferroelectric material for example HfÜ2-based ferroelectric, can be doped.
  • One or more of the following elements can be doped: N, Al, Si, Sc, Ge, Y, Zr, Gd, Pr, Sr, Tb, La, Lu.
  • the metal for the Schottky diode can be selected from: Al, Ag, Au, Cu, Cr, Mo, Ni, Nb, Pt, Ti, Ni, TiN, TaN, and metal alloys.
  • the metal for the Schottky diode can be a metal semiconductor alloy such as silicides, germanides, metal-SiGeSn-alloys.
  • the invention makes it possible to provide artificial synaptic components based on metal / semiconductor Schottky barriers.
  • the polarization of the ferroelectric layers leads to a shielding charge at the metal / semiconductor interface, so that the effective strength of the Schottky barrier in the metal / semiconductor contact is modulated. Induce polarization switching in multi-domain systems
  • Multilevel charge distributions at the metal-semiconductor interface causing multi-level conduction of the diode. This can be used to mimic properties of synapses.
  • the present invention solves the problem of interface traps at the interface between ferroelectric and semiconducting material in FeFETs, because the ferroelectric material is primarily arranged on the metal.
  • the ferroelectric components according to the invention can be produced at relatively low temperatures ( ⁇ 800 ° C.) using the CMOS process. This means that almost all ferroelectric materials can be used.
  • the invention provides a method for the construction of artificial synapses, which can be produced according to the following pattern.
  • These can comprise two metal / semiconductor Schottky diodes connected on the rear side, which are contacted with a ferroelectric layer and a gate electrode.
  • the thickness of the ferroelectric layer on the two Schottky diodes can be identical or different. This also applies to the selection of the material for the ferroelectric layer.
  • One Schottky diode can act as a signal input, while the second Schottky diode can be used to regulate the synaptic weight. With a constant bias on the second Schottky diode, the synaptic weight can also be adjusted via the control voltage on the first Schottky diode.
  • a metal-semiconductor junction can be present, which acts like an ohmic contact.
  • the invention also relates to a method for operating a synaptic component according to the invention.
  • the first Schottky diode is switched in the reverse direction during operation.
  • An electrical voltage is applied in the form of a pulse to the first gate electrode.
  • the synaptic component then behaves like a biological synapse.
  • FIG. 1 shows the functional principle of a neural gate
  • FIG. 2a embodiment of a synapse
  • FIG. 2b symbol representation of the synapse according to FIG. 2a
  • FIG. 3 embodiment of a synapse
  • FIG. 4a modulation of a Schottky diode by a positive gate voltage
  • FIG. 4b Modulation of a Schottky diode by a negative gate voltage
  • FIG. 5a ribbon model of a Schottky diode with a positive gate voltage
  • FIG. 5b ribbon model of a Schottky diode with negative gate voltage
  • FIG. 6a current-voltage curve
  • FIG. 6b embodiment of a synapse
  • FIG. 6a current-voltage curve
  • FIG. 7 pulse-current diagram
  • FIG. 8a embodiment for a plurality of synapses
  • FIG. 8b sectional view from FIG. 8a
  • FIG. 9a embodiment of a synapse
  • FIG. 9b symbol representation of the synapse from FIG. 9a
  • FIG. 10 vertical arrangement of a synaptic component
  • FIGS. 11a, b symbol representations of a neural component
  • FIG. 12 neural network
  • FIG. 13 Real representation of a synapse
  • FIG. 14 NAND gate
  • Figure 15 AND gate.
  • FIG. 1 shows the functional principle of a neural gate, which consists of several input synapses Xi to x n and a neuron.
  • the input signals of the synapses x t with the weights w t are integrated with the following function:
  • S is the integration via the input signals with the associated weights and Q is an offset.
  • the neural function / (s) acts as a threshold function. As soon as S reaches a threshold value, the neuron is activated with the function / (s).
  • Fig. 2 shows a first embodiment of a synapse.
  • the synapse includes a layer 101 made of a semiconductor.
  • a source is located on the semiconducting layer 101
  • Electrode 102 and a drain electrode 103 are spatially separated from one another.
  • the two electrodes 102, 103 can be present adjacent to the end faces of the semiconducting layer 101.
  • the two electrodes 102, 103 are made of metal.
  • the two electrodes 102, 103 can be layered.
  • An electric potential Vs can be applied to the source electrode 102.
  • An electrical potential V ü can be applied to the drain electrode 103.
  • the electric potential Vs is different from the electric potential VD.
  • the metal of the source electrode 102 and the semiconductor material of the semiconducting layer 101 are selected such that the transition between the source electrode 102 and the semiconducting layer 101 is a Schottky contact. There is thus a first Schottky diode, which is formed by the source electrode 102 and the semiconducting layer 101.
  • the metal of the drain electrode 103 and the semiconductor material of the semiconducting layer 101 can be selected such that the transition between the drain electrode 103 and the semiconducting layer 101 is a Schottky contact.
  • the drain electrode 103 and the semiconducting layer 101 can thus form a second Schottky diode.
  • the first Schottky diode and the second Schottky diode are connected “back-to-back” in an equivalent circuit.
  • a potential difference between the electrical potential Vs and the electrical potential VD therefore has the consequence that one Schottky diode is switched in the reverse direction and the other Schottky diode is switched in the forward direction.
  • an electrical current can flow from the source electrode 102 to the drain electrode 103.
  • An electric current flows from the source electrode 102 to the drain
  • the electric current passes the first Schottky diode in the reverse direction and the second Schottky diode in the forward direction.
  • the reverse flow of electrical current can take place in that electrons tunnel through the barrier layer of the first Schottky diode. If the electrical current flows through the second Schottky diode in the forward direction, the second Schottky diode acts like an ohmic resistor.
  • a first ferroelectric layer 104a is located above the source electrode 102.
  • the first ferroelectric layer 104a may cover part of the source electrode 102.
  • the first ferroelectric layer 104 a overlaps with the semiconducting layer 101.
  • a second ferroelectric layer 104b is located above the drain electrode 103.
  • the second ferroelectric layer 104b covers part of the source electrode 103.
  • the second ferroelectric layer 104b overlaps with the semiconducting layer 101.
  • a first electrode 105a is located on the first ferroelectric layer 104a.
  • a second electrode 105b is located on the second ferroelectric layer 104b. Both electrodes 105a and 105b can be made of metal.
  • the synapse is set up in such a way that an electrical potential V gi can be applied to the first electrode 105a.
  • the Schottky barrier of the first Schottky diode can be modulated by an applied electrical potential V gi.
  • the synapse is set up in such a way that an electrical potential V g 2 can be applied to the second electrode 105a.
  • the Schottky barrier of the second Schottky diode can be modulated by an applied electrical potential V g 2.
  • the first electrical potential V gi can be understood as a synaptic input signal within a neural network.
  • the second electrical potential V g 2 can be used to regulate weights within a neural network.
  • the ferroelectric layers 104a and / or 104b can overlap regions of the electrodes 102 and 103 which are applied to the semiconducting layer 101. Production can thus be facilitated.
  • the Schottky barriers of the first and second Schottky diodes can also be modulated in a more controlled manner.
  • FIG. 2b shows an electrical symbol of the synapse shown in FIG. 2a with the source electrode S and the drain electrode D.
  • FIG. 3 shows a second embodiment of a synapse which, instead of two ferroelectric layers 204a and 204b, has only one continuous ferroelectric layer 204.
  • the synapse shown here has a layer 201 which consists of a semiconductor.
  • On the semiconducting layer 201 are as in the case of the figure 2a a source electrode 202 and a drain electrode 203.
  • the metal of the source electrode 202 and the semiconductor material of the semiconducting layer 201 are selected such that the junction between the source electrode 202 and the semiconducting layer 201 is a Schottky contact .
  • the metal of the drain electrode 203 and the semiconductor material of the semiconducting layer 201 can be selected such that the transition between the drain electrode 203 and the semiconducting layer 201 is a Schottky contact.
  • a first electrode 205a and a second electrode 205b are located on the ferroelectric layer 204.
  • the electrical symbol in FIG. 2b can also represent the synapse shown in FIG.
  • FIG. 4a shows the case in which a voltage or electrical potential applied to electrode 305 is positive. The polarization of the ferroelectric causes non-volatile memory effects, so that electrons are present in the semiconductor layer 301.
  • FIG. 4b illustrates the situation when a negative voltage V g is applied to the first or second electrode 305, positive charges (holes) being present in the semiconductor layer 301 due to polarization-related non-volatile memory effects.
  • the thickness of a Schottky barrier can consequently be modulated by applying potentials V g. This affects the flow of tunnel electrical currents.
  • FIG. 5 uses the ribbon model to illustrate a modulation of a Schottky barrier from the metal layer 302 shown in the figure to the n-type doped semiconducting layer 301. If a positive voltage V g is applied, as is shown in FIG. 4a, there are more Electrons are generated at the metal-semiconductor interface. The band curvature (dashed line) is increased. The result is a smaller Schottky barrier. This causes a higher tunnel current density along the arrow, which can be formulated as follows:
  • N D is the doping concentration on the surface of the semiconductor 301.
  • the Schottky barrier is 0 bp .
  • FIG. 6a presents an example of a measurement of an electrical current I D as a function of V gl from a source electrode 402 to a drain electrode 403. A potential of -0.5 volts was applied to the drain electrode.
  • the current-voltage characteristic I d - V gl in Figure 6a shows a counterclockwise hysteresis. This is typical for ferroelectric materials. The ferroelectric material thus achieves a memory effect desired for a synapse.
  • the change in voltage V g2 also affects the initial current. This demonstrates the weight function of V g2 of a synapse.
  • a thin p-doped silicon layer is separated from an Si substrate by an insulating layer. This is called a silicon-on-insulator and is abbreviated to SOI.
  • the insulating layer is formed by "buried silicon oxide", which is also called buried silicon oxide or BOX.
  • the source electrode 402 and the drain electrode 403 are formed from single crystal NiSh.
  • the layers 402 and 403 are applied to the thin, p-doped silicon layer 401 by siliciding on very thin Ni.
  • the layer 401 is 55 nm thick and is lightly p-doped.
  • HfZrO was deposited as a ferroelectric layer 404 by ALD (Atomic Layer Deposition).
  • the layer thickness of the ferroelectric layer 404 can be 3 to 30 nm.
  • the first electrode 405a and the second electrode 405b are made of TiN.
  • the first electrode 405a and the second electrode 405b serve as a gate
  • FIG. 7 shows an example of the measured synaptic characteristic curve of the component from FIG. 6 under the influence of a pulsed signal at the first gate, that is to say at the first gate electrode 405a.
  • the current ID is plotted against the number of pulses PN.
  • a voltage V gi was therefore applied in the form of a pulse to the first gate electrode.
  • long-term potentiation (LTP) is generated by positive pulsed signals.
  • FIGS. 8a and 8b A cross-sectional view from A to A 'from Figure 8a is shown in Figure 8b.
  • Two Schottky diodes of a synapse have a common semiconducting layer 501.
  • a metal layer 502a of the first Schottky diode serves as a source electrode.
  • a metal layer 502b serves as a drain electrode.
  • an electrically insulating layer 505 was first produced on the semiconducting layer in such a way that two accesses to the semiconducting layer remain.
  • the two layers 502a and 502b were then produced in such a way that they are electrically separated from one another.
  • the ferroelectric layers 502a and 502b shown with the gate electrodes 504a and 504b were then produced.
  • conventional production techniques can advantageously be used in order to produce a large number of synapses in the form of a crossbar structure in a technically simple manner.
  • the Schottky diode comprises a drain electrode 602 made of metal.
  • the drain electrode 602 made of metal is applied to a semiconductor layer 601.
  • the first Schottky diode can be modulated by a ferroelectric layer 604 with a gate electrode 605 with the aid of a voltage V g.
  • a second Schottky diode with the semiconducting layer 601 and the drain electrode 603 made of metal can be present. This can be operated with a bias voltage VD in the direction of flow.
  • the semiconducting layer 601 and the metal drain electrode 603 do not have to be selected in such a way that a second Schottky diode is present. Instead, the semiconducting layer 601 and the drain electrode 603 made of metal can be an ohmic contact.
  • the first Schottky diode 602, 601 is operated under reverse bias.
  • This synaptic component works like a single gate transistor, which can be represented by the symbol from FIG. 9b.
  • a layer 701 made of metal can serve as a substrate.
  • the layer 701 made of metal can, however, also have been applied to a substrate.
  • the layer 701 made of metal serves as a drain electrode.
  • the layer 701 made of metal is therefore set up in such a way that it can be connected to a voltage VD.
  • a semiconducting layer 702 is applied to the layer 701 made of metal.
  • the metal layer 701 and the semiconducting layer 702 can be a Schottky diode.
  • An insulator layer 703a, 703b has been applied to the semiconducting layer in such a way that access to semiconducting layer 702 has remained on top.
  • the access can also have been created subsequently after the deposition of the insulating layer 703a, 703b, for example by etching.
  • a metallic layer 704 was then deposited on the top. This makes contact with the semiconducting layer 702 through the access.
  • the metallic layer 704 serves as a source electrode. This can therefore be connected to a voltage Vs.
  • the metallic layer 704 and the semiconducting layer 702 form a Schottky diode.
  • a ferroelectric layer 705 was deposited on the metallic layer 704.
  • a metallic layer 706 was deposited on the ferroelectric layer 705.
  • the metallic layer 706 serves as a gate electrode. This can therefore be connected to a voltage V G.
  • FIG. 10 The structure shown in FIG. 10 can be rotated through 180 °.
  • This structure shown in FIG. 10 is rotationally invariant C2.
  • FIGS. 11a and 11b symbolize a neural component with a synapse according to the invention, a transistor and a resistor being present.
  • the transistor can be either a double gate transistor 801 or a single gate transistor 803. These transistors represent a synaptic component.
  • the resistors 802 and 804 act as pull-up / pull-down resistors. By adjusting the conductivity of the resistors, the activation of a neuron can be adjusted.
  • FIG. 12 shows a neural network with several synapses according to the invention and with conventional CMOS neurons.
  • the synapses are shown in the transistor representation Xi, Wi, ..., X n , W n on the left.
  • a conventional neural component is shown to the right of this.
  • the synapse can be a single gate component, as shown in FIG. 9 and FIG. 10.
  • the synapses can be produced using the CMOS neuron process in order to avoid the high-temperature treatment during the CMOS process.
  • Synapses according to the invention can be combined with conventional neurons, which is illustrated by FIG.
  • FIG. 13 outlines a synapse serving as a connection between neuron and neuron.
  • a signal 904 flows in one direction, namely from the presynaptic neuron 901 to the postsynaptic neuron 903.
  • a neuromodulator 902 is also shown.
  • the structure shown in FIG. 13 can be implemented by the invention.
  • the first gate (V gi ) from FIGS. 2 and 3, formed by 105a / 205a, can serve as a presynaptic neuron.
  • the source electrode serves as a postsynaptic neuron.
  • the second gate (V g 2) formed by 105b / 205b, serves as a neuromodulator. Processing, learning and modulation functions can be implemented at the same time.
  • FIG. 14 illustrates how the component shown in FIG. 11a can be used as a NAND gate with two inputs.
  • FIG. 15 shows how a component according to FIG. 2b can be used as an AND gate, where Iout is the current flowing off via D.
  • Iout is the current flowing off via D.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • General Health & Medical Sciences (AREA)
  • Neurology (AREA)
  • Molecular Biology (AREA)
  • Biophysics (AREA)
  • Theoretical Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Semiconductor Memories (AREA)
  • Thin Film Transistor (AREA)
  • Insulated Gate Type Field-Effect Transistor (AREA)
PCT/EP2021/065297 2020-06-16 2021-06-08 Neuronen und synapsen mit ferroelektrisch modulierten metall-halbleiter schottky dioden nebst verfahren Ceased WO2021254830A1 (de)

Priority Applications (4)

Application Number Priority Date Filing Date Title
CN202180035992.XA CN115668226A (zh) 2020-06-16 2021-06-08 神经元和具有铁电调制金属半导体肖特基二极管的突触及方法
JP2022577164A JP2023534389A (ja) 2020-06-16 2021-06-08 強誘電的に変調された金属-半導体ショットキーダイオードを有するニューロンおよびシナプス、ならびに方法
EP21731993.8A EP4133419B1 (de) 2020-06-16 2021-06-08 Neuronen und synapsen mit ferroelektrisch modulierten metall-halbleiter schottky dioden nebst verfahren
US18/007,721 US20230231030A1 (en) 2020-06-16 2021-06-08 Neurons and synapses with ferroelectrically modulated metal-semiconductor schottky diodes and method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
DE102020207439.9A DE102020207439A1 (de) 2020-06-16 2020-06-16 Neuronen und Synapsen mit ferroelektrisch modulierten Metall-Halbleiter Schottky Dioden nebst Verfahren
DE102020207439.9 2020-06-16

Publications (1)

Publication Number Publication Date
WO2021254830A1 true WO2021254830A1 (de) 2021-12-23

Family

ID=76421990

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2021/065297 Ceased WO2021254830A1 (de) 2020-06-16 2021-06-08 Neuronen und synapsen mit ferroelektrisch modulierten metall-halbleiter schottky dioden nebst verfahren

Country Status (6)

Country Link
US (1) US20230231030A1 (https=)
EP (1) EP4133419B1 (https=)
JP (1) JP2023534389A (https=)
CN (1) CN115668226A (https=)
DE (1) DE102020207439A1 (https=)
WO (1) WO2021254830A1 (https=)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023231144A1 (zh) * 2022-05-31 2023-12-07 复旦大学 一种极化场调控的正负光响应探测器及其制备方法和应用
JP2024068667A (ja) * 2022-11-08 2024-05-20 リサーチ アンド ビジネス ファウンデーション ソンギュングァン ユニバーシティ 接合構造素子、この製造方法及びこれを備えるインメモリーコンピューター素子
JP7620348B1 (ja) 2023-08-30 2025-01-23 高麗大学校産学協力団 シリコンゲーテッドダイオードを用いた二値化ニューラルネットワーク回路

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20220012576A1 (en) * 2020-07-09 2022-01-13 Korea Advanced Institute Of Science And Technology Charge trap based neuromorphic synaptic transistor with improved linearity and symmetricity by schottky junctions, and a neuromorphic system using it

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0529565B1 (en) 1991-08-23 1997-11-19 NEC Corporation Semiconductor synapse circuit and neuron device using the same
US20190164597A1 (en) 2017-11-29 2019-05-30 International Business Machines Corporation Synaptic crossbar memory array
CN208922326U (zh) 2018-10-29 2019-05-31 南京邮电大学 一种基于忆阻器阵列的神经网络突触觉结构
US20190198617A1 (en) * 2017-12-21 2019-06-27 International Business Machines Corporation Field effect transistor with controllable resistance
WO2019139598A1 (en) * 2018-01-11 2019-07-18 Intel Corporation Ferroelectric neurons and synapses with dual gate ferroelectric transistors
WO2019147859A2 (en) 2018-01-24 2019-08-01 The Regents Of The University Of California Synaptic resistors for concurrent parallel signal processing, memory and learning with high speed and energy efficiency

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9660067B2 (en) * 2014-03-25 2017-05-23 Intel Corporation III-N transistors with epitaxial layers providing steep subthreshold swing
DE102016111237B3 (de) 2016-06-20 2017-11-23 Namlab Ggmbh Rekonfigurierbarer Nanodraht-Feldeffekt-Transistor und dessen Herstellung sowie ein Nanodraht-Array und dessen Rekonfigurierung
US10510903B2 (en) * 2016-11-29 2019-12-17 Taiwan Semiconductor Manufacturing Co., Ltd. Impact ionization semiconductor device and manufacturing method thereof
CN106653837B (zh) * 2016-12-02 2019-09-13 电子科技大学 一种氮化镓双向开关器件
WO2019066959A1 (en) * 2017-09-29 2019-04-04 Intel Corporation FERROELECTRIC NEURONS AND SYNAPES
US10963776B2 (en) * 2018-08-24 2021-03-30 Namlab Ggmbh Artificial neuron based on ferroelectric circuit element

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0529565B1 (en) 1991-08-23 1997-11-19 NEC Corporation Semiconductor synapse circuit and neuron device using the same
US20190164597A1 (en) 2017-11-29 2019-05-30 International Business Machines Corporation Synaptic crossbar memory array
US20190198617A1 (en) * 2017-12-21 2019-06-27 International Business Machines Corporation Field effect transistor with controllable resistance
WO2019139598A1 (en) * 2018-01-11 2019-07-18 Intel Corporation Ferroelectric neurons and synapses with dual gate ferroelectric transistors
WO2019147859A2 (en) 2018-01-24 2019-08-01 The Regents Of The University Of California Synaptic resistors for concurrent parallel signal processing, memory and learning with high speed and energy efficiency
CN208922326U (zh) 2018-10-29 2019-05-31 南京邮电大学 一种基于忆阻器阵列的神经网络突触觉结构

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
TANG BIN ET AL: "Novel Type of Synaptic Transistors Based on a Ferroelectric Semiconductor Channel", APPLIED MATERIALS & INTERFACES, vol. 12, no. 22, 11 May 2020 (2020-05-11), US, pages 24920 - 24928, XP055840003, ISSN: 1944-8244, Retrieved from the Internet <URL:https://pubs.acs.org/doi/pdf/10.1021/acsami.9b23595> [retrieved on 20210913], DOI: 10.1021/acsami.9b23595 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2023231144A1 (zh) * 2022-05-31 2023-12-07 复旦大学 一种极化场调控的正负光响应探测器及其制备方法和应用
JP2024068667A (ja) * 2022-11-08 2024-05-20 リサーチ アンド ビジネス ファウンデーション ソンギュングァン ユニバーシティ 接合構造素子、この製造方法及びこれを備えるインメモリーコンピューター素子
JP7620348B1 (ja) 2023-08-30 2025-01-23 高麗大学校産学協力団 シリコンゲーテッドダイオードを用いた二値化ニューラルネットワーク回路
JP2025035998A (ja) * 2023-08-30 2025-03-14 高麗大学校産学協力団 シリコンゲーテッドダイオードを用いた二値化ニューラルネットワーク回路

Also Published As

Publication number Publication date
EP4133419A1 (de) 2023-02-15
US20230231030A1 (en) 2023-07-20
CN115668226A (zh) 2023-01-31
EP4133419C0 (de) 2025-03-26
JP2023534389A (ja) 2023-08-09
EP4133419B1 (de) 2025-03-26
DE102020207439A1 (de) 2021-12-16

Similar Documents

Publication Publication Date Title
EP4133419B1 (de) Neuronen und synapsen mit ferroelektrisch modulierten metall-halbleiter schottky dioden nebst verfahren
DE102018204283B4 (de) Programmierbare logikelemente und verfahren zum betreiben derselben
DE69018348T2 (de) Speicherbauelement aus organischem Halbleiter mit einer MISFET-Struktur und sein Kontrollverfahren.
DE4020007C2 (de) Nichtflüchtiger Speicher
DE69132469T2 (de) Elektrische Anordnung mit einem dotierten amorphen Siliziumkanal
DE112019001709T5 (de) Halbleitervorrichtung und multiply-accumulate-operations-vorrichtung
EP4062464B1 (de) Kapazitives synaptisches bauelement und verfahren zu dessen ansteuerung
DE2011794C3 (de) Halbleiterspeicheranordnung
DE69501381T2 (de) Halbleitergerät
DE1803035A1 (de) Feldeffekt-Halbleiterbauelement
DE102020100777A1 (de) Analoge nichtflüchtige Speichervorrichtung unter Verwendung eines polyferroelektrischen Films mit zufälligen Polarisationsrichtungen
EP0050773A2 (de) Steuerbares MIS-Bauelement
DE2363089C3 (de) Speicherzelle mit Feldeffekttransistoren
DE2201028B2 (de) Verfahren zum Betrieb eines Feldeffekttransistors und Feldeffekttransistor zur Ausübung dieses Verfahrens
DE102023213345A1 (de) Schaltungsanordnung für ein binarisiertes neuronales Netz unter Verwendung von Siliziumgatedioden
DE102016205860A1 (de) Verfahren und Mittel zum Betrieb eines komplementären analogen rekonfigurierbaren memristiven Widerstandsschalters sowie dessen Verwendung als künstliche Synapse
DE2124635A1 (de) Verfahren zur Steuerung eines Bereiches hoher elektrischer Feldstarke in Halblei tem
DE112021005740B4 (de) Gemischt leitendes flüchtiges speicherelement zum beschleunigten beschreiben eines nichtflüchtigen memristiven bauelements
DE69127585T2 (de) Schaltung für künstliche neurale Funktion
DE112006002377T5 (de) Halbleitervorrichtung und Verfahren zum Herstellen einer Halbleitervorrichtung
KR102392450B1 (ko) 시냅틱 트랜지스터 및 이의 제조 방법
DE2742936B2 (de) Nichtflüchtiger Langzeitspeicher
WO2009040375A1 (de) Elektronisches bauelement mit schalteigenschaften
DE102022104831A1 (de) Memristive struktur, memristives kreuzschienen-array und verfahren dafür
WO2012022342A1 (de) Halbleiterbauelement

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21731993

Country of ref document: EP

Kind code of ref document: A1

DPE1 Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101)
ENP Entry into the national phase

Ref document number: 2021731993

Country of ref document: EP

Effective date: 20221109

ENP Entry into the national phase

Ref document number: 2022577164

Country of ref document: JP

Kind code of ref document: A