A method for reconfigurable optoelectronic memory logic operation of ferroelectric capacitor type memory

By adjusting the initial capacitance state of the ferroelectric capacitive memory and using electrical and optical signals to coordinately control the polarization state, reconfigurable optoelectronic in-memory logic operations of NAND gates and NOT gates are realized. This solves the problem of the lack of in-memory logic operations in ferroelectric capacitive memories and achieves low-power, high-computing-power integrated sensing-storage-computing.

CN117133328BActive Publication Date: 2026-11-17XIDIAN UNIV +1
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202311100722.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-08-29
Publication Date
2026-11-17
Estimated Expiration
2043-08-29

AI Technical Summary

Technical Problem

Existing ferroelectric capacitive memories do not have in-memory logic operation capabilities, which limits the development of integrated sensing-storage-computing technology, and also have static power consumption issues.

Method used

By regulating the initial capacitance state of the ferroelectric capacitor memory, and using electrical and optical signals to coordinately regulate the polarization state of the ferroelectric dielectric layer, reconfigurable optoelectronic logic operations of NAND gates and NOT gates are realized, giving the memory optical sensing capabilities.

Benefits of technology

It achieves low-power, high-computing-power integrated sensing-storage-computing, improves the functional density of devices and the flexibility of circuit design, expands the scaling space of circuits, and is suitable for high-efficiency, low-power integrated circuit manufacturing.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117133328B_ABST
    Figure CN117133328B_ABST
Patent Text Reader

Abstract

This invention provides a reconfigurable optoelectronic memory logic operation method for ferroelectric capacitive memory, belonging to the field of in-memory computing. The ferroelectric capacitive memory includes, from bottom to top, a bottom electrode, a semiconductor layer, a ferroelectric dielectric layer, and a top electrode; the top electrode is preceded by a C-shaped structure before the application of input voltage signal A and optical signal B. acc Let S be the initial capacitance state of the ferroelectric capacitive memory. In this ferroelectric capacitive memory, when the initial capacitance state S = 1, the output Y is a NAND gate logic; when the initial capacitance state S = 0, the output Y is a NOT gate logic. By adjusting the initial capacitance state S, the reconfigurable optoelectronic logic functions of the NAND and NOT gates in the ferroelectric capacitive memory are realized. The reconfigurable optoelectronic logic function implementation method provided by this invention enables ferroelectric capacitive memories to be applied to the development of high-efficiency, low-power, high-computing-power integrated memory and computing devices and chip technologies.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of in-memory computing technology, and specifically relates to a method for reconfigurable optoelectronic memory logic operation of ferroelectric capacitor type memory. Background Technology

[0002] Traditional von Neumann computing architectures have played a dominant role in complementary metal-oxide-semiconductor (CMOS) systems, where data processing and storage are separated. However, with the rapid development of information technology, especially the advent of the intelligent Internet of Things (IoT) era, mainstream computing architectures are facing the von Neumann bottleneck and memory wall caused by the separation of storage and processing. The increasingly frequent data transfers between memory and processors lead to significant power consumption, hindering the improvement of information processing performance. One existing approach to addressing the von Neumann bottleneck and memory wall is near-in-memory computing, which involves placing a large amount of memory around the processor in a 3D stack to increase memory access bandwidth. However, this approach is essentially still a storage-computation separation architecture and cannot fundamentally solve the problems faced by traditional computing architectures in the face of ever-increasing computing demands. Therefore, in response to the needs of sensing, storing, and computing massive amounts of complex information, developing new energy-efficient, low-power, and high-computing-power "non-von Neumann" architecture integrated sensing-memory-computing devices and their chip technologies has become a key bottleneck technology driving the expansion of related technological boundaries. In-memory logic technology with sensing capabilities is considered a promising solution. In this method, information sensed by the memory can be directly used for logical calculations within the in-memory logic units, while simultaneously storing the calculation results, thus achieving a fusion of sensing, storage, and computation. Currently, memories with in-memory logic capabilities mainly include phase-change RAM (PCRAM), ferroelectric RAM (FeRAM), magnetic random access memory (MRAM), and resistive random access memory (RRAM). These memories essentially store data through differences in conductivity. During operation, the unavoidable read current and trailing current lead to additional static power consumption, severely reducing the energy efficiency of information processing.

[0003] To address the energy efficiency issues of existing resistive random access memories (RRAMs), researchers have proposed ferroelectric capacitive memories (FCRAMs). These memories feature a single-capacitor structure, avoiding static power consumption and thus achieving low power consumption. Recently, researchers have proposed a photo-controlled capacitive ferroelectric memory, revealing its significant potential in inductively coupled computing (ICC) applications. However, current research on FCRAMs is extremely limited, and existing FCRAMs lack in-memory logic operation capabilities, severely hindering their application in ICC technology and limiting the development of ICC technologies with high energy efficiency, low power consumption, and high computing power. Summary of the Invention

[0004] To overcome the shortcomings of the prior art, the present invention aims to provide a reconfigurable optoelectronic memory logic operation method for ferroelectric capacitive memory. The method uses an external electrical signal and an optical signal as input signals to the ferroelectric capacitive memory, and the final capacitance state of the ferroelectric capacitive memory as the output signal. By adjusting the initial capacitance state of the ferroelectric capacitive memory, the logic function of the device can be changed, and finally, reconfigurable optoelectronic memory logic functions of NAND gates and NOT gates can be realized on a single ferroelectric capacitive memory device.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is as follows:

[0006] A method for reconfigurable optoelectronic memory logic operation of a ferroelectric capacitor type memory, wherein the ferroelectric capacitor type ferroelectric memory includes a semiconductor layer and a ferroelectric dielectric layer in contact, the ferroelectric dielectric layer is connected to a top electrode, and the semiconductor layer is connected to a bottom electrode; characterized in that:

[0007] The series capacitance (C) of the semiconductor layer capacitor and the ferroelectric dielectric layer capacitor acc The high capacitance state or low capacitance state represents the two types of stored information in the ferroelectric memory.

[0008] Record C before the voltage and light signals are applied to the top electrode. acc The initial capacitance state S of the ferroelectric capacitive memory; the capacitance C after applying voltage and optical signals. acc To output Y;

[0009] C in high capacitance state acc When the value is 1, C is in the low capacitance state. acc When the value is 0, the output Y is a NAND gate logic when S=1, and a NOT gate logic when S=0. Thus, by adjusting the initial capacitor state S, the internal logic operation of the NAND gate and NOT gate photoelectric memory can be reconfigured, and by applying voltage and light signals to the top electrode, the high capacitance state or the low capacitance state can be controlled.

[0010] In one embodiment, the voltage signal applied by the top electrode is denoted as A, and the optical signal applied by the top electrode is denoted as B; when the output Y is a NAND gate logic, the logic operation result is... When the output Y is a NOT gate logic, the result of the logical operation is:

[0011] In one embodiment, when light is applied above the top electrode, the light signal is 1 and B = 1; when no light is applied above the top electrode, i.e., when the device is in a dark environment, the light signal is 0 and B = 0.

[0012] In one embodiment, the semiconductor layer is selected as a P-type semiconductor; a positive voltage is applied to the top electrode, the voltage signal is 1, and A = 1; a negative voltage is applied to the top electrode, the voltage signal is 0, and A = 0.

[0013] In one embodiment, the voltage signal is a pulse voltage.

[0014] In one embodiment, the semiconductor layer is selected as a P-type semiconductor; the initial capacitance state S is set to a high capacitance state by applying a negative voltage to the top electrode under illumination; the initial capacitance state S is set to a low capacitance state by applying a positive voltage to the top electrode under illumination.

[0015] In one embodiment, the bottom electrode and top electrode are made of any one of ITO, tungsten, titanium, copper, aluminum, platinum, iridium, ruthenium, tungsten nitride, titanium nitride, tantalum nitride, iridium oxide, ruthenium oxide, tungsten carbide, titanium carbide, tungsten silicide, titanium silicide, and tantalum silicide; the ferroelectric dielectric layer is made of any one of HZO, HAO, HSO, HYO, BFO, AlScN, PZT, BST, ZrO2, Al2O3, and ZnSnO3; and the semiconductor layer is made of any one of Si, Ge, SiGe, GaN, AlN, GaAs, and SiC.

[0016] In one embodiment, the thickness of the depletion layer of the semiconductor layer ranges from 1 nm to 1000 nm.

[0017] The present invention also claims protection for a ferroelectric capacitive memory that uses the described ferroelectric capacitive memory to reconfigure the internal logic operation method of the optoelectronic memory.

[0018] The present invention further claims protection for an efficient inductive-in-memory computing artificial vision system using the aforementioned ferroelectric capacitive memory.

[0019] Compared with the prior art, the beneficial effects of the present invention are:

[0020] The beneficial effects of this invention are: This invention provides a reconfigurable optoelectronic memory logic operation method for ferroelectric capacitor type memory, which utilizes the input electrical signal and optical signal to coordinately control the polarization state of the ferroelectric dielectric layer, thereby controlling the photogenerated minority carriers in the semiconductor or the majority carrier holes in the semiconductor layer that respond to the polarization charge, and realizing the control of the output C accEditing the high and low states of the capacitor enables ferroelectric capacitive memory to possess in-memory logic functions, thus solving the problem that existing ferroelectric capacitive memory does not have in-memory logic operation capabilities. Secondly, this invention utilizes optical signals as one of the input signals to endow the in-memory logic functions of the ferroelectric capacitive memory with optical sensing capabilities, enabling its application in sensing-storage-computing integrated technology, which is beneficial to the development of high-efficiency, low-power, and high-computing-power sensing-storage-computing integrated technology. Furthermore, this invention utilizes the high or low initial capacitance state S to control the output logic. When the initial capacitance state S = 1, the output Y outputs NAND gate logic; when the initial capacitance state S = 0, the output Y outputs NOT gate logic. This allows a single ferroelectric capacitive memory to possess reconfigurable optoelectronic in-memory logic functions of NAND and NOT gates, improving the functional density of a single device, enhancing the flexibility of circuit design and integration, expanding the scaling space of the circuit, and opening up new avenues for overcoming the bottlenecks of integrated circuit size scaling and von Neumann architecture in the post-Moore's Law era.

[0021] This invention also provides a method for implementing the in-memory logic function of a ferroelectric capacitive memory with photoelectric sensing and reconfigurable logic. It can be used to develop high-efficiency, low-power, and high-computing-power integrated sensing-memory-computing devices and chip technologies. Ferroelectric capacitive memories with photoelectric reconfigurable in-memory logic function are expected to become one of the ideal device structures integrating efficient data sensing, storage, and computation in the era of big data. Moreover, the structure and materials of this device are all integrated circuit process compatible materials, which can be compatible with integrated circuit silicon-based processes and can be compatible with the existing integrated circuit manufacturing industry. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the ferroelectric capacitor type memory of the present invention.

[0023] Figure 2 This is a schematic diagram illustrating the implementation of the opto-reconfigurable internal logic function of the ferroelectric capacitor memory of the present invention.

[0024] Figure 3 This is the actual test result of the present invention on the opto-reconfigurable internal logic function of ferroelectric capacitor type memory.

[0025] Figure 4 This is a truth table for the opto-reconfigurable internal logic function of the ferroelectric capacitor type memory of the present invention.

[0026] In the diagram: 1. Bottom electrode, 2. Semiconductor layer, 3. Ferroelectric layer, 4. Top electrode. Detailed Implementation

[0027] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention are within the scope of protection of the present invention.

[0028] Those skilled in the art should understand that, in the disclosure of this invention, the terms "longitudinal," "lateral," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the above terms should not be construed as limiting this invention.

[0029] It is understood that the term "a" should be understood as "at least one" or "one or more", that is, in one embodiment, the number of an element can be one, while in another embodiment, the number of the element can be multiple, and the term "a" should not be understood as a limitation on the number.

[0030] like Figure 1 As shown, Figure 1 This is a schematic diagram of a ferroelectric capacitor memory. The ferroelectric capacitor memory includes a semiconductor layer 2 and a ferroelectric dielectric layer 3 in contact. The ferroelectric dielectric layer 3 is connected to a top electrode 4, and the semiconductor layer 2 is connected to a bottom electrode 1. In a specific embodiment, the bottom electrode 1, semiconductor layer 2, ferroelectric dielectric layer 3, and top electrode 4 can be arranged sequentially from bottom to top, such as... Figure 1 As shown.

[0031] In this invention, the series capacitance C of the semiconductor layer 2 capacitor and the ferroelectric dielectric layer 3 capacitor is... acc The high capacitance state or the low capacitance state represent two types of stored information in the ferroelectric memory.

[0032] In this invention, C is defined as the value before the voltage and light signals are applied to the top electrode 4. acc Let S be the initial capacitance state of the ferroelectric capacitive memory; let C be the capacitance state after applying voltage and light signals. acc The output is Y.

[0033] In this invention, C is in a high capacitance state. acc When the value is 1, C is in the low capacitance state. accWhen S = 1, the output Y is a NAND gate logic; when S = 0, the output Y is a NOT gate logic. Thus, by adjusting the initial capacitor state S, the internal logic operations of the NAND and NOT gates in the ferroelectric memory can be reconfigured. Furthermore, by applying voltage and light signals to the top electrode 4, the high-capacitance state or the low-capacitance state can be controlled, thereby realizing two types of information storage in the ferroelectric memory.

[0034] In the following embodiments of the present invention, the principle and method are described in detail using a P-type semiconductor as an example of semiconductor layer 2. Obviously, the present invention is not limited to P-type semiconductors. When an N-type semiconductor is used, its principle and control method are similar to those of P-type semiconductors. For those skilled in the art, such conversion and understanding do not require creative effort.

[0035] The series capacitance C of the capacitor in semiconductor layer 2 and capacitor in ferroelectric dielectric layer 3 of this invention acc The high capacitance state or low capacitance state represents two types of stored information in the ferroelectric memory. Specifically, when semiconductor layer 2 is a P-type semiconductor, if the side of semiconductor layer 2 closest to the ferroelectric dielectric layer 3 is in a majority carrier accumulation state and there is no depletion layer, then C at this time... acc The capacitance is equal to that of a single ferroelectric layer 3, which is a high-capacitance state; if the side of semiconductor layer 2 closest to ferroelectric layer 3 is in a majority carrier depletion state, and a depletion layer exists, then C at this time... acc The capacitance C is equal to the total capacitance of the series connection of the depletion layer capacitance of semiconductor layer 2 and the capacitance of ferroelectric layer 3. According to the series and parallel capacitance law, C at this point... acc It is in a low capacitance state.

[0036] When a positive voltage is applied to the top electrode 4, the number of charge carriers in the semiconductor layer 2 is modulated by the light signal. The non-equilibrium minority carriers generated by the light interact with the polarization charges in the ferroelectric layer 3, assisting in the reversal of the polarization charges in the ferroelectric layer 3 and changing the depletion layer capacitance state of the semiconductor layer 2, thereby changing the series capacitance C between the depletion layer of the semiconductor layer 2 and the ferroelectric layer 3. acc state.

[0037] Figure 2 This is a schematic diagram illustrating the implementation of the opto-reconfigurable internal logic function of the ferroelectric capacitor memory of the present invention. The two input signals are the voltage signal A applied by the top electrode 4 and the optical signal B applied by the top electrode 4, respectively. The output is Y. The initial capacitor state S is used to switch the logic function of the device.

[0038] (1) When the initial capacitor state S = 1, the output Y is a "NAND gate" logic, and the result of the logic operation is:

[0039] (2) When the initial capacitor state S = 0, the output Y is a NOT gate logic, and the result of the logic operation is:

[0040] The present invention is configured as follows:

[0041] When light is applied above the top electrode 4, the light signal is 1, i.e., B = 1; when no light is applied above the top electrode 4, i.e., when the device is in a dark environment, the light signal is 0, i.e., B = 0.

[0042] A positive voltage is applied to the top electrode 4, and the voltage signal is set to 1, i.e., A = 1; a negative voltage is applied to the top electrode 4, and the voltage signal is set to 0, i.e., A = 0. The applied voltage signal can be a pulse voltage.

[0043] By applying a negative pulse voltage to the top electrode 4 under illumination, the initial capacitance state S can be set to a high capacitance state. In this state, there is no depletion layer in semiconductor layer 2, and the polarization direction in ferroelectric layer 3 is upward. By applying a positive pulse voltage to the top electrode 4 under illumination, the initial capacitance state S can be set to a low capacitance state. In this state, there is a depletion layer in semiconductor layer 2, and the polarization direction in ferroelectric layer 3 is downward. When the initial capacitance state S is high capacitance, S = 1; when the initial capacitance state S is low capacitance, S = 0. When the output capacitance state is high capacitance, Y = 1; when the output capacitance state is low capacitance, Y = 0.

[0044] Figure 3 The following are the actual test results of the opto-reconfigurable in-memory logic function of the ferroelectric capacitor type memory according to this invention:

[0045] (1) When the initial capacitance state is high, i.e., S=1, the output Y is a "NAND gate" logic, and the result of the logic operation is: The initial capacitor state setting method is as follows: A negative pulse voltage is applied to the top electrode 4, and majority carrier holes accumulate on the side of the semiconductor layer 2 near the ferroelectric dielectric layer 3. The polarization charge in the ferroelectric dielectric layer 3 is completely reversed, and the output C... acc In the high capacitance state, the initial capacitance state S is set.

[0046] When the input voltage signal A = 0 and the input light signal B = 0, the output Y = 1. The principle is as follows: When the initial capacitance state is high capacitance, a negative voltage is applied to the top electrode 4 in darkness. The semiconductor layer 2 is in a majority carrier accumulation state with no depletion layer. Because there are enough holes inside the semiconductor layer 2 to respond to the polarization charges inside the ferroelectric dielectric layer 3, the polarization charges inside the ferroelectric dielectric layer 3 are fully flipped, and the polarization direction is upward, the same as the initial state. At this time, the C value is read. acc Same as the initial capacitor state, i.e., the output C acc In the high capacitance state, Y = 1.

[0047] When the input voltage signal A = 0 and the input light signal B = 1, the output Y = 1. The principle is as follows: When the initial capacitance state is high, a negative voltage is applied to the top electrode 4 under illumination. The semiconductor layer 2 is in a majority carrier accumulation state with no depletion layer. Since the semiconductor layer 2 has enough holes to respond to the polarization charges inside the ferroelectric layer 3, no additional photogenerated carriers are needed. The polarization charges in the ferroelectric layer 3 can fully flip, with the polarization direction upwards, the same as the initial state. At this time, the C value read... acc Same as the initial capacitor state, i.e., the output C acc In the high capacitance state, Y = 1.

[0048] When the input voltage signal A = 1 and the input light signal B = 0, the output Y = 1. The principle is as follows: When the initial capacitance state is high, a positive voltage is applied to the top electrode 4 in darkness. Because the P-type semiconductor cannot provide enough charge carriers (electrons) to respond to the polarization charges in the ferroelectric dielectric layer 3, most of the polarization charges in the ferroelectric dielectric layer 3 cannot reverse, the polarization state does not change, and the polarization direction remains upward, the same as the initial state. At this time, the C value is read... acc Same as the initial capacitor state, i.e., the output C acc In the high capacitance state, Y = 1.

[0049] When the input voltage signal A = 1 and the input light signal B = 1, the output Y = 0. The principle is as follows: When the initial capacitance state is high, a positive voltage is applied to the top electrode 4 under illumination. Although the P-type semiconductor cannot provide enough charge carriers (electrons) to respond to the polarization charge in the ferroelectric layer 3, the non-equilibrium charge carriers generated by illumination can assist in reversing the polarization charge in the ferroelectric layer 3, changing the polarization state downwards. At this time, the semiconductor layer 2 is in a majority carrier depletion state, and the read C... acc It is in a low capacitance state, that is, the output C acc In the low capacitance state, Y = 0.

[0050] (2) When the initial capacitance state is low, i.e., S = 0, the output Y is a NOT gate logic, and the result of the logic operation is: The initial capacitance state setting method is as follows: Under illumination, a positive pulse voltage is applied to the top electrode 4. The polarization charge in the non-equilibrium minority carrier-assisted ferroelectric layer 3 generated by illumination is completely reversed, and a depletion layer exists in the semiconductor layer 2. acc The initial capacitance state S is set to low capacitance.

[0051] When the input voltage signal A = 0 and the input light signal B = 0, the output Y = 1. The principle is as follows: Initially, the capacitance is low. In darkness, a negative voltage is applied to the top electrode 4, and the semiconductor layer 2 is in a majority carrier accumulation state with no depletion layer. Because the semiconductor layer 2 has enough holes to interact with the polarization charges inside the ferroelectric layer 3, the polarization charges in the ferroelectric layer 3 fully reverse, changing the polarization state to upward. At this point, the C value is read... acc It is in a high capacitance state, that is, the output C acc In the high capacitance state, Y = 1.

[0052] When the input voltage signal A = 0 and the input light signal B = 1, the output Y = 1. The principle is as follows: When the initial capacitance state is low, a negative voltage is applied to the top electrode 4 under illumination. The semiconductor layer 2 is in a majority carrier accumulation state with no depletion layer. Since there are still enough holes inside the semiconductor layer 2 to respond to the polarization charges inside the ferroelectric layer 3, no additional photogenerated carriers are needed. The polarization charges inside the ferroelectric layer 3 can fully flip, changing the polarization state to upward. At this time, the C value is read... acc It is in a high capacitance state, that is, the output C acc This is a high capacitance state, i.e., Y = 1.

[0053] When the input voltage signal A = 1 and the input light signal B = 0, the output Y = 0. The principle is as follows: When the initial capacitance state is low, a positive voltage is applied to the top electrode 4 in darkness. Because the P-type semiconductor cannot provide enough charge carriers (electrons) to respond to the polarization charges in the ferroelectric layer 3, most of the polarization charges in the ferroelectric layer 3 cannot reverse, the polarization state does not change, and the polarization direction is downward. At this time, the C value read... acc The initial capacitor state remains consistent, i.e., the output C acc It is in a low capacitance state, i.e., Y = 0.

[0054] When the input voltage signal A = 1 and the input light signal B = 1, the output Y = 0. The principle is as follows: When the initial capacitance state is low, a positive voltage is applied to the top electrode 4 under illumination. At this time, the non-equilibrium carriers generated by the illumination can assist in the reversal of the polarization charge in the ferroelectric layer 3. The polarization direction after the reversal is consistent with the initial direction, pointing downwards. The semiconductor layer 2 is in a majority carrier depletion state, consistent with the initial state. At this time, the C value is read... acc It is in a low capacitance state, that is, the output C acc It is in a low capacitance state, i.e., Y = 0.

[0055] Figure 4 This is the truth table for the opto-reconfigurable internal logic function of the ferroelectric capacitor type memory of the present invention.

[0056] This invention utilizes the coordinated modulation of the polarization state of the ferroelectric dielectric layer by input electrical and optical signals, thereby controlling the photogenerated minority carriers in the semiconductor or the majority carrier holes in the semiconductor layer that respond to the polarization charge, to achieve control over the output C acc Editing the high and low states of the capacitor enables the ferroelectric capacitive memory to possess in-memory logic functions, thus solving the problem that existing ferroelectric capacitive memories lack in-memory logic operation capabilities. Secondly, this invention utilizes optical signals as one of the input signals, endowing the in-memory logic functions of the ferroelectric capacitive memory with optical sensing capabilities, enabling its application in sensing-storage-computing integrated technology, which is beneficial to the development of high-efficiency, low-power, and high-computing-power sensing-storage-computing integrated technology. Furthermore, this invention utilizes the high or low initial capacitance state S to control the output logic. When the initial capacitance state S = 1, the output Y outputs a NAND gate logic; when the initial capacitance state S = 0, the output Y outputs a NOT gate logic, thereby enabling a single ferroelectric capacitive memory to possess both NAND and NOT gate logic. The reconfigurable optoelectronic internal logic function improves the functional density of a single device, enhances the flexibility of circuit design and integration, and expands the scaling space of the circuit, opening up new avenues for overcoming the bottlenecks of integrated circuit size scaling and von Neumann architecture in the post-Moore's Law era. This invention also provides a method for implementing the internal logic function of a ferroelectric capacitive memory with optoelectronic sensing and reconfigurable logic. This method can be used to develop high-efficiency, low-power, and high-computing-power integrated sensing-memory-computing devices and chips. Ferroelectric capacitive memories with optoelectronic reconfigurable internal logic functions are expected to become one of the ideal device structures integrating efficient data sensing, storage, and computation in the era of big data. Furthermore, the structure and materials of this device are all integrated circuit process compatible materials, compatible with silicon-based integrated circuit processes, and compatible with the existing integrated circuit manufacturing industry. Therefore, ferroelectric capacitive memories using the reconfigurable optoelectronic internal logic operation method can be used in artificial vision systems with high-efficiency in-sensor computation, meeting the demand of the Internet of Things era for integrated sensing-memory-computing devices and chips with "high energy efficiency, low power consumption, and high computing power."

[0057] In embodiments of the present invention, both the bottom electrode 1 and the top electrode 4 are made of metallic materials. Specifically, the materials of the bottom electrode 1 and the top electrode 4 can be any one of ITO, tungsten, titanium, copper, aluminum, platinum, iridium, ruthenium, tungsten nitride, titanium nitride, tantalum nitride, iridium oxide, ruthenium oxide, tungsten carbide, titanium carbide, tungsten silicide, titanium silicide, and tantalum silicide. The thickness of the bottom electrode 1 and the top electrode 4 is generally in the range of 5 nm to 500 nm.

[0058] The material of the ferroelectric layer 3 can be any one of HZO, HAO, HSO, HYO, BFO, AlScN, PZT, BST, ZrO2, Al2O3, and ZnSnO3. Among them, HZO is zirconium-doped hafnium oxide, HAO is aluminum-doped hafnium oxide, HSO is silicon-doped hafnium oxide, HYO is yttrium-doped hafnium oxide, BFO is bismuth ferrite, AlScN is scandium-doped aluminum nitride, PZT is lead zirconate titanate, ZrO2 is zirconium dioxide, Al2O3 is aluminum oxide, ZnSnO3 is zinc stannate, and BST is barium strontium titanate.

[0059] When semiconductor layer 2 is a P-type semiconductor, its material can be any one of Si, Ge, SiGe, GaN, AlN, GaAs, and SiC. The thickness of the depletion layer of semiconductor layer 2 is generally in the range of 1 nm to 1000 nm.

[0060] To make the above features and advantages more apparent and understandable, specific implementation examples are provided below, along with detailed explanations in conjunction with the accompanying drawings.

[0061] This invention uses HZO with a zirconium doping composition of 50% to fabricate the ferroelectric dielectric layer 3, a p-type Si substrate as the semiconductor layer 2, and tungsten metal as the material for the bottom electrode 1 and the top electrode 4. The light source is a microscope light source: a halogen fiber optic lamp, model: Motic, MLC-150C, with an illuminance of 0.3 W / cm². 2 .

[0062] refer to Figure 2 , Figure 3 , Figure 4 In this invention, input A is a pulse voltage with an amplitude of 5V and a pulse width of 1ms. When the pulse voltage is positive, input A is recorded as 1; when the pulse voltage is negative, input A is recorded as 0. Input B is an optical signal; when B = 1, the light intensity is 0.3W / cm². 2 When B=1, the light intensity is 0W / cm². 2 The output Y of this invention is obtained through CV testing. The CV testing frequency is 1MHz, the AC small signal amplitude is 30mV, and the voltage scan range is from 0.5V to -1V. The CV testing is performed in darkness. By applying +5V to the top electrode under illumination and a 1ms pulse sets the initial capacitor state S to 0, the logic function of the ferroelectric capacitor memory is switched to "NOT gate". By applying -5V to the top electrode under illumination and a 1ms pulse sets the initial capacitor state S to 1, the logic function of the ferroelectric capacitor memory is switched to "NAND gate".

[0063] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Anyone can modify or make equivalent substitutions to the technical solutions of the present invention without departing from the spirit and scope of the present invention. The protection scope of the present invention shall be determined by the claims.

Claims

1. A method for reconfigurable optoelectronic memory logic operation, wherein the ferroelectric capacitor type ferroelectric memory includes a semiconductor layer (2) and a ferroelectric dielectric layer (3) in contact, the ferroelectric dielectric layer (3) being connected to a top electrode (4), and the semiconductor layer (2) being connected to a bottom electrode (1); characterized in that: a series capacitance of the capacitance of the semiconductor layer (2) and the capacitance of the ferroelectric dielectric layer (3) C acc a high capacitance state or a low capacitance state, representing two kinds of storage information of the ferroelectric memory the top electrode (4) before the voltage signal and the light signal are applied C acc S is the initial capacitance state of the ferroelectric capacitor type memory after applying the voltage signal and the light signal C acc Y = output When in high capacitance state C acc A value of 1 indicates a low capacitance state. C acc When the value is 0, the output Y is a NAND gate logic when S=1, and a NOT gate logic when S=0. Thus, by adjusting the initial capacitor state S, the internal logic operation of the NAND gate and NOT gate photoelectric memory can be reconfigured, and by applying voltage and light signals to the top electrode (4), the high capacitor state or the low capacitor state can be controlled. Let the voltage signal applied to the top electrode (4) be A, and let the light signal applied to the top electrode (4) be B; when the output Y is NAND logic, the logic operation result is ; when the output Y is NOR logic, the logic operation result is ; The semiconductor layer (2) is a P-type semiconductor; the initial capacitance state S is set to a high capacitance state by applying a negative voltage to the top electrode (4) under illumination; the initial capacitance state S is set to a low capacitance state by applying a positive voltage to the top electrode (4) under illumination.

2. The reconfigurable optoelectronic memory logic operation method according to claim 1, characterized in that, When light is applied above the top electrode (4), the light signal is 1 and B=1; when no light is applied above the top electrode (4), i.e. the device is in a dark environment, the light signal is 0 and B=0.

3. The method according to claim 1 or 2, wherein the ferroelectric capacitor type memory reconfigurable photoelectric in-memory logic operation method is characterized by, The semiconductor layer (2) is selected as a P-type semiconductor; a positive voltage is applied to the top electrode (4), the voltage signal is 1, and A=1; a negative voltage is applied to the top electrode (4), the voltage signal is 0, and A=0.

4. The reconfigurable optoelectronic memory logic operation method according to claim 1, characterized in that, The voltage signal is a pulse voltage.

5. The reconfigurable optoelectronic memory logic operation method according to claim 1, characterized in that, The bottom electrode (1) and top electrode (4) are made of any one of ITO, tungsten, titanium, copper, aluminum, platinum, iridium, ruthenium, tungsten nitride, titanium nitride, tantalum nitride, iridium oxide, ruthenium oxide, tungsten carbide, titanium carbide, tungsten silicide, titanium silicide, and tantalum silicide; the ferroelectric dielectric layer (3) is made of any one of HZO, HAO, HSO, HYO, BFO, AlScN, PZT, BST, ZrO2, Al2O3, and ZnSnO3; and the semiconductor layer (2) is made of any one of Si, Ge, SiGe, GaN, AlN, GaAs, and SiC.

6. The reconfigurable optoelectronic memory logic operation method according to claim 1 or 5, characterized in that, The thickness of the depletion layer of the semiconductor layer (2) ranges from 1 nm to 1000 nm.

7. A ferroelectric capacitor type memory, and a method for reconfiguring the internal logic operation of the optoelectronic memory using the ferroelectric capacitor type memory according to any one of claims 1 to 6.

8. An integrated sensing, storage, and computing artificial vision system using the ferroelectric capacitive memory described in claim 7.

Citation Information

Patent Citations

  • Logic operation circuit and operation method

    CN106374912A

  • Light-operated capacitive ferroelectric memory and preparation method thereof

    CN115589774A