Ferroelectric structure and manufacturing method thereof, true random number generator using the same, solar cell using the same, and non-volatile RF switch using the same

Abstract

To provide a ferroelectric structure including a two-dimensional ferroelectric material capable of implementing a true random number generator through a single device, and a method for manufacturing the same.SOLUTION: A method for operating a ferroelectric structure comprises: preparing a ferroelectric structure including a ferroelectric layer that includes a substrate 100, a lower electrode 200 disposed on the substrate, a ferroelectric layer disposed on the lower electrode and including a two-dimensional ferroelectric material, and an upper electrode 400 disposed on the ferroelectric layer; applying an operating voltage to the ferroelectric structure such that polarization occurs in the ferroelectric layer; applying a read voltage to the ferroelectric structure to obtain a current value reflecting a polarization state of the ferroelectric layer; and applying an erase voltage to the ferroelectric structure such that the ferroelectric layer is restored to a state before the polarization; thereby obtaining a current value of random intensity for the same voltage being applied.SELECTED DRAWING: Figure 5

Description

Technical Field

[0001] The present invention relates to a ferroelectric structure and a method for manufacturing the same, and more particularly, to a ferroelectric structure having a structure in which a ferroelectric layer is disposed between an upper electrode and a lower electrode, and a method for manufacturing the same.

[0002] The ferroelectric structure and the method for manufacturing the same according to the present invention are applicable to a True Random Number Generator, a solar cell, and a non-volatile RF switch.

Background Art

[0003] Due to the rapid growth of the Internet of Things (IoT), an enormous amount of data is being generated and exchanged. As a result, concerns about data security, particularly confidentiality, integrity, and authentication, are also increasing. However, there is no integrated security standard for the spread of IoT devices, and such IoT devices are vulnerable to attacks, raising serious security issues.

[0004] Encryption keys have traditionally been used to protect data, but are becoming increasingly vulnerable to machine learning (ML) and physical attacks. To solve such problems, a Random Number Generator (RNG) has attracted attention.

[0005] More specifically, regarding the random number generator, a pseudo random number generator (PRNG) model based on CMOS elements is used. However, existing CMOS-based random number generator models have limitations in that they are systems implemented with a complex circuit structure and not perfect random numbers, but rather complex algorithms for mimicking random numbers. This has limitations in directly applying to the IoT industry where a huge amount of data is generated, and has a fatal problem that if one can even analyze the complex algorithm structure, the encryption can be hacked.

[0006] As described above, different from the software-based RNG (Pseudo Random Number Generator, PRNG) that extends the initial seed to a bit sequence using a mathematical algorithm, a true random number generator (TRNG) utilizes the physical attributes of a random and unpredictable hardware system to generate statistically independent bits. Therefore, the true random number generator is currently a major candidate group applicable to data encryption technologies that occur on a huge scale and requires in-depth research.

[0007] For example, in 2016, Z. Wei; Y. Katoh; S. Ogasahara; Y. Yoshimoto; K. Kawai; Y. Ikeda; K. Eriguchi; K. Ohmori; S. Yoneda, “True random number generator using current difference based on a fractional stochastic model in 40-nm embedded ReRAM”, 2016 IEEE International Electron Devices Meeting (IEDM) disclosed a true random number generator using tantalum oxide (Ta2O5), which is a resistive change material.

[0008] Also, for example, Bo Liu; Jing Ma; Han Hsiang Tai; Dharmendra Verma; Mamina Sahoo; Ying-Feng Chang; Hanyuan Liang; Shiwei Feng; Lain-Jong Li; Tuo-Hung Hou; Chao-Sung Lai, “Memristive True Random Number Generator with Intrinsic Two-Dimensional Physical Unclonable Function”, ACS Appl. Electron. Mater., 2023, 5, 2, 714-720 discloses a true random number generator using aluminum oxide (AlO x ) as a resistive change material.

Prior Art Documents

Non-Patent Documents

[0009]

Non-Patent Document 1

Non-Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0010] An object of the present invention is to provide a ferroelectric structure including a two-dimensional ferroelectric material and a method for manufacturing the same.

[0011] Another object of the present invention is to provide a ferroelectric structure and a method for manufacturing the same, in which a current value with a random intensity can be obtained for the same applied voltage.

[0012] Still another object of the present invention is to provide a ferroelectric structure and a method for manufacturing the same, which embody a true random number generator (TRNG) with a single element.

[0013] Still another object of the present invention is to provide a method for deriving an ideal operating voltage and a readout voltage so that the ferroelectric structure has ideal random number generation performance.

[0014] Still another object of the present invention is to provide a ferroelectric structure that can be easily applied to data encryption and encryption key generation for data security.

[0015] Still another object of the present invention is to provide a ferroelectric structure that can be easily applied to simulation data generation using random sampling of repetitive data.

[0016] Another object of the present invention is to provide a solar cell to which the above-described ferroelectric is applied.

[0017] Another object of the present invention is to provide a non-volatile RF switch to which the above-described ferroelectric is applied.

[0018] The object of the present invention is not limited to the above.

Means for Solving the Problems

[0019] The present invention for achieving the above object provides a ferroelectric structure. The ferroelectric structure includes a substrate, a lower electrode disposed on the substrate, a ferroelectric layer disposed on the lower electrode and containing molybdenum disulfide (MoS2), and an upper electrode disposed on the ferroelectric layer, and is characterized in that a current value of random intensity is obtained for the same applied voltage.

[0020] When the same voltage is applied multiple times to the ferroelectric structure, current values of different intensities are generated for each of the applied voltages.

[0021] When a voltage is applied to the ferroelectric structure, polarization occurs in the ferroelectric layer.

[0022] Due to the polarization generated in the ferroelectric layer, a current value of random intensity is obtained for the same applied voltage in the ferroelectric structure.

[0023] The ferroelectric structure is applied to a True Random Number Generator.

[0024] According to another embodiment, the ferroelectric structure includes a substrate, a lower electrode disposed on the substrate, a ferroelectric layer disposed on the lower electrode and containing a two-dimensional ferroelectric material, and an upper electrode disposed on the ferroelectric layer, and is characterized in that a current value of random intensity is obtained for the same applied voltage.

[0025] When the same voltage is applied multiple times to the ferroelectric structure, current values with different intensities are generated for each of the applied voltages.

[0026] When a voltage is applied to the ferroelectric structure, polarization occurs in the ferroelectric layer.

[0027] When a voltage is applied to the ferroelectric structure, polarization in the vertical direction (Out-of-plane, OOP) occurs in the ferroelectric layer.

[0028] The two-dimensional ferroelectric material contains compounds of different elements. When a voltage is applied, the position of any one of the different element ions contained in the two-dimensional ferroelectric material changes within the ferroelectric layer.

[0029] In the ferroelectric structure, the crystal structure of the ferroelectric layer changes due to the element ions whose positions change within the ferroelectric layer.

[0030] The ferroelectric layer changes from a symmetric crystal structure to an anti-symmetric crystal structure.

[0031] The ferroelectric structure is applied to a true random number generator.

[0032] In order to achieve the above-described object, the present invention provides a method for manufacturing a ferroelectric structure.

[0033] The method for manufacturing the ferroelectric structure includes the steps of preparing a substrate, forming a lower electrode on the substrate, forming a ferroelectric layer containing a two-dimensional ferroelectric material on the lower electrode, and forming an upper electrode on the ferroelectric layer.

[0034] Also, in order to achieve the above-described object, the present invention provides an operation method for a ferroelectric structure.

[0035] The operating method of the ferroelectric structure includes the steps of preparing a ferroelectric structure including a substrate, a lower electrode disposed on the substrate, a ferroelectric layer disposed on the lower electrode and containing a two-dimensional ferroelectric material, and an upper electrode disposed on the ferroelectric layer; applying an operating voltage to the ferroelectric structure so that polarization is generated in the ferroelectric layer; applying a read voltage to the ferroelectric structure to obtain a current value reflecting the polarization state of the ferroelectric layer; and applying a removal voltage to the ferroelectric structure so that the ferroelectric layer returns to a state before polarization generation.

[0036] The steps of applying the operating voltage, obtaining the current value, and applying the removal voltage are sequentially repeated, and the same operating voltage is repeatedly applied.

Advantages of the Invention

[0037] The ferroelectric structure according to the present invention includes a substrate, a lower electrode disposed on the substrate, a ferroelectric layer disposed on the lower electrode and containing a two-dimensional ferroelectric material, and an upper electrode disposed on the ferroelectric layer, and a current value with a random intensity is obtained for the same applied voltage.

[0038] Accordingly, when the ferroelectric structure according to the present invention is used, a true random number generator can be implemented with a single element. Further, the ferroelectric structure can be easily applied to data encryption for data security, encryption key generation, simulation data generation using random sampling of repetitive data, and the like.

Brief Description of the Drawings

[0039]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

Figure 18

Figure 19

Figure 20

Figure 21

Figure 22

Figure 23

Figure 24

Figure 25

Figure 26

Figure 27

Figure 28

Figure 29

Figure 30

Figure 31

Figure 32

Figure 33

Figure 34

Figure 35

Figure 36

Figure 37

Figure 38

Figure 39

Figure 40

Figure 41

Figure 42

Figure 43

Figure 44

Figure 45

Figure 46

Figure 47

Figure 48

Figure 49

Figure 50

Figure 51

Figure 52

Figure 53

Figure 54

Figure 55

Figure 56

Figure 57

Figure 58

Figure 59

Figure 60

Figure 61

Figure 62

Figure 63

Figure 64

Figure 65

Figure 66

Figure 67

Figure 68

Figure 69

Figure 70

Figure 71

[0040] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the technical idea of the present invention is not limited to the embodiments described herein, and can be embodied in other forms. The embodiments introduced here are provided so that the disclosed content is thorough and complete, and that the idea of the present invention is sufficiently conveyed to those skilled in the art.

[0041] In this specification, when a certain component is said to be on another component, it means that it can be directly formed on the other component, or a third component can be sandwiched between them. Also, in the drawings, shapes and sizes are exaggerated for an effective explanation of the technical content.

[0042] Also, in various embodiments of this specification, terms such as first, second, third, etc. are used to describe various components, but these components should not be limited by such terms. These terms are merely used to distinguish one component from another. Thus, what is referred to as the first component in one embodiment can also be referred to as the second component in another embodiment. Each embodiment described and illustrated here also includes its complementary embodiments. Also, in this specification, "and / or" is used to mean including at least one of the components listed before and after.

[0043] In the specification, singular expressions include plural expressions unless the context clearly indicates otherwise. Also, terms such as "including" or "having" are intended to specify the presence of the features, numbers, steps, components, or combinations thereof described in the specification, and should not be construed as excluding the presence or possibility of addition of one or more other features, numbers, steps, components, or combinations thereof. Further, in this specification, "connecting" is used in the sense of including both indirectly connecting and directly connecting a plurality of components.

[0044] Also, in describing the present invention, when it is determined that a detailed description of related known functions or configurations would unnecessarily obscure the gist of the present invention, that detailed description will be omitted.

[0045] (Ferroelectric structure, method for manufacturing the same, and operation method) FIG. 1 is a flowchart for explaining a method for manufacturing a ferroelectric structure according to an embodiment of the present invention, FIG. 2 is a schematic diagram for explaining step S110 among the methods for manufacturing a ferroelectric structure according to an embodiment of the present invention, FIG. 3 is a schematic diagram for explaining step S120 among the methods for manufacturing a ferroelectric structure according to an embodiment of the present invention, FIG. 4 is a schematic diagram for explaining step S130 among the methods for manufacturing a ferroelectric structure according to an embodiment of the present invention, FIG. 5 is a schematic diagram for explaining step S140 among the methods for manufacturing a ferroelectric structure according to an embodiment of the present invention, and FIG. 6 is a flowchart for explaining an operation method of a ferroelectric structure according to an embodiment of the present invention.

[0046] As shown in FIGS. 1 and 2, a substrate 100 is prepared (S110). According to one embodiment, the substrate 100 is a silicon semiconductor substrate. More specifically, the substrate 100 is a substrate having a silicon oxide (SiO2) layer 120 formed on a silicon (Si) substrate 110. According to other embodiments, the substrate 100 is a compound semiconductor substrate. According to other embodiments, the substrate 100 is a glass substrate. According to still other embodiments, the substrate 100 is a plastic substrate. The type of the substrate 100 is not limited.

[0047] As shown in FIGS. 1 and 3, a lower electrode 200 is formed on the substrate 100 (S120). According to one embodiment, the lower electrode 200 contains a metal. For example, the lower electrode 200 contains gold (Au). According to one embodiment, the lower electrode 200 is formed by electron beam lithography and electron beam deposition methods. The formation method and material of the lower electrode 200 described above are merely examples, and the types of the formation method and material of the lower electrode 200 are not limited.

[0048] As shown in FIGS. 1 and 4, a ferroelectric layer 300 is formed on the lower electrode 200 (S130). According to one embodiment, the ferroelectric layer 300 contains a two-dimensional ferroelectric material. Also, the two-dimensional ferroelectric material contains a compound of different elements. For example, the ferroelectric layer contains any one of molybdenum disulfide (MoS2), hafnium oxide (HfO2), and CuInP2S6 (CIPS).

[0049] According to one embodiment, the ferroelectric layer 300 is formed by a method of dry-transferring a two-dimensional ferroelectric material mechanically exfoliated from a bulk onto the lower electrode 200. The formation method of the ferroelectric layer 300 is not limited.

[0050] As shown in FIGS. 1 and 5, an upper electrode 400 is formed on the ferroelectric layer 300 (S140). Thereby, the ferroelectric structure according to the embodiment is manufactured. According to one embodiment, the upper electrode 400 contains a metal. For example, the upper electrode 400 contains gold (Au). According to one embodiment, the upper electrode 400 is formed by electron beam lithography and electron beam evaporation methods. The above-described formation method and material of the upper electrode 400 are merely illustrative, and the types of the formation method and material of the upper electrode 400 are not limited.

[0051] When a voltage is applied to the ferroelectric structure, any one of the different elements contained in the two-dimensional ferroelectric material of the ferroelectric layer 300 changes its position within the ferroelectric layer 300. For example, when the ferroelectric layer 300 contains CuInP2S6 (CIPS) as the two-dimensional ferroelectric material, the position of copper ions (Cu+) in the ferroelectric layer 300 changes when a voltage is applied to the ferroelectric structure.

[0052] In addition, the crystal structure of the ferroelectric layer 300 changes due to the element ions whose positions change within the ferroelectric layer 300. For example, when the position of copper ions (Cu+) in the ferroelectric layer 300 changes, the crystal structure of the ferroelectric layer 300 changes from a symmetric crystal structure to an anti-symmetric crystal structure.

[0053] In addition, as the crystal structure of the ferroelectric layer 300 changes, polarization occurs in the ferroelectric layer 300. For example, in the ferroelectric layer 300 containing CuInP2S6 (CIPS), out-of-plane (OOP) polarization occurs.

[0054] That is, when a voltage is applied to the ferroelectric structure, any one of the different elements contained in the two-dimensional ferroelectric material (for example, CIPS) of the ferroelectric layer 300 (for example, Cu+ ions) changes its position within the ferroelectric layer 300, and the crystal structure of the ferroelectric layer 300 changes (for example, changes from a symmetric crystal structure to an asymmetric crystal structure). As a result, polarization occurs in the ferroelectric layer 300 (for example, polarization occurs in the vertical direction).

[0055] In describing the present invention, "polarization" means a phenomenon in which the positions of negative charges and positive charges are separated in an electric field and a dipole moment is formed. Also, "in-plane (IP) polarization" means a phenomenon in which the positions of negative charges and positive charges are separated along a direction parallel to the upper or lower surface of the ferroelectric layer 300 and a dipole moment is formed. Also, "out-of-plane (OOP) polarization" means a phenomenon in which the positions of negative charges and positive charges are separated along a direction perpendicular to the upper or lower surface of the ferroelectric layer 300, that is, along the thickness direction of the ferroelectric layer 300, and a dipole moment is formed.

[0056] As described above, in the process in which the crystal structure of the ferroelectric layer 300 changes and polarization occurs due to the change in the position of specific element ions within the ferroelectric layer 300, random domain changes occur in the ferroelectric layer 300. More specifically, when partial polarization occurs in the ferroelectric layer 300, random domain changes occur in the ferroelectric layer 300.

[0057] In the description of the present invention, "partial polarization" means a state in which dipoles with different array directions are mixed. For example, when polarization in the vertical direction (OOP) occurs in the ferroelectric layer 300, dipoles arrayed in the upper (up) direction (for example, arranged such that positive charges face the upper surface and negative charges face the lower surface) and dipoles arrayed in the lower (bottom) direction (for example, arranged such that negative charges face the upper surface and positive charges face the lower surface) are mixed, which is defined as the partial polarization state. Accordingly, the occurrence of random partial polarization means that the position distribution of the dipoles arrayed in the upper direction and the dipoles arrayed in the lower direction in the ferroelectric layer 300 is random, and the ratio of each of the dipoles arrayed in the upper direction and the dipoles arrayed in the lower direction in the ferroelectric layer 300 changes randomly.

[0058] That is, even if the same voltage is repeatedly applied to the ferroelectric structure, different domain states and different polarization states are formed in the ferroelectric layer 300 for each applied voltage (random domain state formation and random polarization state formation). As a result, a current value with random intensity is obtained for the same voltage applied to the ferroelectric structure. That is, even if the same voltage is applied to the ferroelectric structure a plurality of times, current values with different intensities can be generated for each applied voltage. Due to such characteristics, the ferroelectric structure can be applied to a true random number generator.

[0059] According to one embodiment, a power supply application unit (not shown) and a random number generation unit (not shown) are connected to each of the input end and the output end of the ferroelectric structure. For example, when the lower electrode 200 is used as the input end and the upper electrode 400 is used as the output end, the power supply application unit is connected to the lower electrode 200, and the random number generation unit is connected to the upper electrode 400. Differently, when the upper electrode 200 is used as the input end and the lower electrode 400 is used as the output end, the power supply application unit is connected to the upper electrode 200, and the random number generation unit is connected to the lower electrode 400. The power supply application unit applies an input voltage less than the critical voltage, and the random number generation unit can generate true random numbers according to the magnitude of the output current corresponding to the input voltage.

[0060] According to one embodiment, the operation method of the ferroelectric structure for obtaining random current data by the ferroelectric structure is performed in the procedure as shown in FIG. 6.

[0061] Specifically, referring to FIG. 6, a ferroelectric structure is prepared (S10). The ferroelectric structure prepared in step S10 is the same as the ferroelectric structure manufactured by the method described with reference to FIGS. 1 to 5.

[0062] An operating voltage (V program ) is applied to the ferroelectric structure so that polarization occurs in the ferroelectric layer 300 (S20). According to one embodiment, the operating voltage can be selected in a range equal to or higher than the coercive voltage and equal to or lower than the saturation voltage. According to one embodiment, the coercive voltage is defined as the voltage at the moment when the movement of specific ions (for example, Cu+ ions) in the ferroelectric layer 300 begins to occur. According to one embodiment, the saturation voltage is defined as the voltage at the moment when, after the movement of specific ions (for example, Cu+ ions) in the ferroelectric layer 300 has occurred, complete movement has been performed and no further movement of specific ions (for example, Cu+ ions) occurs.

[0063] For example, the operating voltage can be selected within a range that is equal to or higher than the step-down voltage and equal to or lower than the saturation voltage, and can be selected as the voltage that moves 50% of specific ions (for example, Cu+ ions). In contrast, for example, the operating voltage can be selected within a range that is equal to or higher than the step-down voltage and equal to or lower than the saturation voltage, and can be selected as the voltage that forms 50% of the domains in the ferroelectric layer 300.

[0064] After the step (S20) of applying an operating voltage to the ferroelectric structure, a read voltage (V read ) is applied to obtain a current value reflecting the polarization state of the ferroelectric layer 300 (S30). Thereafter, a removal voltage (V erase ) is applied to the ferroelectric structure so that the ferroelectric layer 300 returns to the state before polarization generation (S40).

[0065] The step (S20) of applying the operating voltage, the step (S30) of obtaining the current value, and the step (S40) of applying the removal voltage are sequentially repeated, and the same operating voltage is repeatedly applied. By obtaining a current value for each applied operating voltage, random current data can be obtained by the ferroelectric structure.

[0066] Also, according to one embodiment, the present invention can derive an ideal operating voltage and read voltage so that the ferroelectric structure has ideal random number generation performance.

[0067] More specifically, among the operating methods of the ferroelectric structure, an average current value is derived for the plurality of current values obtained in step S30, and based on the derived average current value, each of the plurality of current values obtained in step S30 is binarized to generate bit data. For example, current values greater than or equal to the average current value are binarized to the state of "1", and current values less than the average current value are binarized to the state of "0" to generate bit data.

[0068] After generating bit data by binarization, the generated bit data is classified into a plurality of keys. For example, the generated 784 bits are composed of 28 bits for one key and are classified into a total of 28 keys.

[0069] After the keys are classified, an evaluation is performed on the classified keys, and an operating voltage and a read voltage that satisfy the corresponding evaluation are derived. For example, for the classified keys, an operating voltage and a read voltage are derived such that the uniformity value calculated by the following Equation (1) has 0.5 and the entropy value calculated by the following Equation (2) has 1. In the case of an ideal random number generator, since the occurrence probability of the 0 state and the occurrence probability of the 1 state are equivalent to 50%, the operating voltage and the read voltage having the uniformity value of 0.5 and the entropy value of 1 can be derived as ideal values.

[0070]

Equation

[0071]

Equation

[0072] As a result of deriving the ideal operating voltage and read voltage for a ferroelectric structure including a CuInP2S6 (CIPS) ferroelectric layer by the method as described above, the operating voltage is derived to be 4V and the read voltage is derived to be 1.5V. That is, by applying an operating voltage of 4V and a read voltage of 1.5V to a ferroelectric structure including a CuInP2S6 (CIPS) ferroelectric layer, a true random number generator having ideal random number generation performance can be realized.

[0073] As a result, the ferroelectric structure according to an embodiment of the present invention includes a substrate 100, a lower electrode 200 disposed on the substrate 100, a ferroelectric layer 300 disposed on the lower electrode 200 and containing a two-dimensional ferroelectric material, and an upper electrode 400 disposed on the ferroelectric layer 300, and includes obtaining a current value with a random intensity for the same applied voltage.

[0074] Thereby, when the ferroelectric structure according to the embodiment is used, a true random number generator is implemented with a single element. Further, the ferroelectric structure can be easily applied to data encryption for data security, encryption key generation, simulation data generation using random sampling of repetitive data, and the like.

[0075] As described above, the ferroelectric structure according to an embodiment of the present invention, its manufacturing method, and its operation method have been described. Hereinafter, other field application examples of the ferroelectric structure (metal / CIPS / metal) according to an embodiment of the present invention will be described.

[0076] According to one embodiment, the ferroelectric structure (metal / CIPS / metal) according to an embodiment of the present invention is used as a semiconductor layer of a solar cell. More specifically, it can be applied to a solar cell having a structure in which a CuInP2S6 (CIPS) semiconductor layer is disposed between an upper electrode and a lower electrode, and converts light energy into electrical energy using the photovoltaic effect. As described above, a solar cell using CuInP2S6 (CIPS) as a semiconductor layer exhibits a higher current production amount (JSC / Power) compared to a solar cell using a substance different from CIPS as a semiconductor layer for the same applied power.

[0077] According to another embodiment, the ferroelectric structure (metal / CIPS / metal) according to the embodiment of the present invention is applied to a non-volatile RF switch. More specifically, it can be applied to a non-volatile RF switch that uses a resistive change non-volatile memory with a capacitor structure to allow an RF signal to pass well with low resistance in the ON state and shield the RF signal well with high resistance and capacitance in the OFF state.

[0078] As described above, various application examples of the ferroelectric structure according to the embodiment of the present invention have been explained. Hereinafter, various modifications of the ferroelectric structure according to the embodiment of the present invention will be described.

[0079] (First modification example: Ferroelectric layer stacking structure) FIG. 7 is a diagram for explaining a ferroelectric structure according to a first modification example of the present invention.

[0080] As shown in FIG. 7, the ferroelectric structure according to the first modification example of the present invention includes a substrate 100, a lower electrode 200 disposed on the substrate 100, a first ferroelectric layer 310 disposed on the lower electrode 200, an intermediate electrode 400 disposed on the first ferroelectric layer 310, a second ferroelectric layer 320 disposed on the intermediate electrode 400, and an upper electrode 500 disposed on the second ferroelectric layer 320. That is, the ferroelectric structure according to the first modification example has a structure in which the second ferroelectric layer 320 is stacked on the first ferroelectric layer 310.

[0081] As described above, when the second ferroelectric layer 320 is stacked on the first ferroelectric layer 310, two random currents can be obtained by inputting an operation voltage (V program ) and a reading voltage (V read ). Thereby, the throughput of random number generation is improved.

[0082] According to one embodiment, the first ferroelectric layer 310 and the second ferroelectric layer 320 include the same two-dimensional ferroelectric material. For example, the first ferroelectric layer 310 and the second ferroelectric layer 320 include CuInP2S6 (CIPS).

[0083] According to another embodiment, the first ferroelectric layer 310 and the second ferroelectric layer 320 include different two-dimensional ferroelectric materials. For example, the first ferroelectric layer 310 includes CuInP2S6 (CIPS), while the second ferroelectric layer 320 includes CuCrP2S6 or CuInP2Se6.

[0084] Also, according to one embodiment, operation voltage scaling is performed by using thickness control of the first ferroelectric layer 310 and the second ferroelectric layer 320. More specifically, the thicker the thicknesses of the first ferroelectric layer 310 and the second ferroelectric layer 320 are, the higher the voltage required for polarization generation becomes. Thus, in an encryption situation where HPC (High-Performance Computing) is required, a high operation voltage is needed, so the thickness is increased, and in an encryption situation where edge computing is required, a low operation voltage is needed, so the thickness is decreased to perform operation voltage scaling.

[0085] (Second Modified Example: Ferroelectric Layer Split Structure) FIG. 8 is a diagram for explaining a ferroelectric structure according to a second modified example of the present invention.

[0086] As shown in FIG. 8, the ferroelectric structure according to the second modified example of the present invention includes a substrate 100, a lower electrode 200 disposed on the substrate 100, a first ferroelectric layer 310 and a second ferroelectric layer 320 disposed on the lower electrode 200, and an upper electrode 400 disposed on the first ferroelectric layer 310 and the second ferroelectric layer 320.

[0087] According to one embodiment, the first ferroelectric layer 310 and the second ferroelectric layer 320 are disposed on the lower electrode 200 and arranged side by side along a direction parallel to the upper surface of the lower electrode 200. Also, the lower surface of the first ferroelectric layer 310 and the lower surface of the second ferroelectric layer 320 are both arranged to be in contact with the lower electrode 200, and the upper surface of the first ferroelectric layer 310 and the upper surface of the second ferroelectric layer 320 are arranged to be in contact with the upper electrode 400. That is, compared with the ferroelectric structure according to the embodiment, the ferroelectric layer 300 of the ferroelectric structure according to the second modification has a structure in which the ferroelectric layer 300 is separated into the first ferroelectric layer 310 and the second ferroelectric layer 320.

[0088] As described above, when the ferroelectric layer 300 has a structure in which the ferroelectric layer 300 is separated into the first ferroelectric layer 310 and the second ferroelectric layer 320, domain randomness occurs in each of the first ferroelectric layer 310 and the second ferroelectric layer 320. Thereby, by generating random numbers from the first ferroelectric layer 310 and the second ferroelectric layer 320 respectively, the efficiency of overall random number generation is improved.

[0089] According to one embodiment, the first ferroelectric layer 310 and the second ferroelectric layer 320 include the same two-dimensional ferroelectric material. For example, the first ferroelectric layer 310 and the second ferroelectric layer 320 include CuInP2S6 (CIPS).

[0090] According to another embodiment, the first ferroelectric layer 310 and the second ferroelectric layer 320 include different two-dimensional ferroelectric materials. For example, the first ferroelectric layer 310 includes CuInP2S6 (CIPS), and the second ferroelectric layer 320 includes CuCrP2S6 or CuInP2Se6. As described above, when the first ferroelectric layer 310 and the second ferroelectric layer 320 include different two-dimensional ferroelectric materials, it becomes difficult to expect the same result in elements having the same structure, so the problem with element replication is solved and the randomness is also enhanced.

[0091] (Third modification: upper electrode split structure) FIG. 9 is a schematic cross-sectional view for explaining a ferroelectric structure according to a third modification of the present invention, and FIG. 10 is a schematic plan view for explaining the ferroelectric structure according to the third modification of the present invention.

[0092] As shown in FIGS. 9 and 10, the ferroelectric structure according to the third modification of the present invention includes a substrate 100, a lower electrode 200 disposed on the substrate 100, a ferroelectric layer 300 disposed on the lower electrode 200, a first upper electrode 410 disposed on the ferroelectric layer 300, a second upper electrode 420, a third upper electrode 430, and a fourth upper electrode 440.

[0093] According to one embodiment, the first upper electrode 410, the second upper electrode 420, the third upper electrode 430, and the fourth upper electrode 440 are disposed on the ferroelectric layer 300 and are spaced apart from each other. That is, compared with the ferroelectric structure according to the embodiment, the upper electrode 400 of the ferroelectric structure according to the third modification has a structure in which the upper electrode 400 is separated into the first upper electrode 410, the second upper electrode 420, the third upper electrode 430, and the fourth upper electrode 440.

[0094] As described above, when the upper electrode 400 has a structure in which the upper electrode 400 is separated into the first upper electrode 410, the second upper electrode 420, the third upper electrode 430, and the fourth upper electrode 440, the random selection of the first upper electrode to the fourth upper electrodes 410, 420, 430, 440 and the randomness of the domains of the ferroelectric layer 300 improve the efficiency of random number generation.

[0095] According to one embodiment, the first upper electrode to the fourth upper electrodes 410, 420, 430, 440 contain the same metal. For example, the first upper electrode to the fourth upper electrodes 410, 420, 430, 440 contain gold (Au).

[0096] According to one embodiment, the first upper electrode to the fourth upper electrodes 410, 420, 430, 440 have different sizes of area (the size of the area in contact with the ferroelectric layer). As a result, the number of cases for electrode selection increases, so randomness is enhanced, and for the same input signal, in the process of generating an encryption, the usage area of the material can be efficiently adjusted, thus improving the lifespan of the element.

[0097] As described above, various modifications of the ferroelectric structure according to the embodiments of the present invention have been explained. Hereinafter, specific experimental examples and characteristic evaluation results of the ferroelectric structure according to the embodiments of the present invention will be described.

[0098] (Experimental Example 1: Fabrication and Structure Confirmation of Ferroelectric Structure) On a SiO2 / Si substrate, a gold (Au) electrode was deposited as the lower electrode by electron beam lithography and electron beam evaporation. Thereafter, CuInP2S6 (CIPS) was secured by physical exfoliation method, and the secured CuInP2S6 (CIPS) was transferred onto the lower electrode by a dry transfer method to form a ferroelectric layer. Finally, a gold (Au) electrode was deposited as the upper electrode on the ferroelectric layer by electron beam lithography and electron beam evaporation.

[0099] FIG. 11 is a diagram showing a schematic diagram and an optical image of a ferroelectric structure according to an experimental example of the present invention.

[0100] FIG. 11 shows a schematic diagram (right) and an optical image (left) of the ferroelectric structure according to the experimental example. It can be seen from FIG. 11 that a CuInP2S6 (CIPS) ferroelectric layer was formed between the upper electrode and the lower electrode.

[0101] FIG. 12 is a schematic diagram of AFM measurement for the ferroelectric layer of the ferroelectric structure according to an experimental example of the present invention, and FIG. 13 is an image showing the thickness and surface roughness of the ferroelectric layer measured by AFM.

[0102] As shown in FIGS. 12 and 13, for the ferroelectric layer of the ferroelectric structure according to the experimental example, the thickness and surface roughness were confirmed by measuring with an atomic force microscope (AFM). More specifically, scanning was performed within a 30×30 μm^2 region of the ferroelectric layer. As shown in FIG. 13, scanning was performed along the red line, and it was confirmed from the height (nm) profile confirmed by the scanning that the ferroelectric layer had a thickness of about 80 nm.

[0103] (Experimental Example 2: Evaluation of Electrical Polarization Characteristics of Ferroelectric Structure) FIG. 14 is a schematic diagram of the LGD double-well model for the CIPS material used in the ferroelectric layer according to the experimental example of the present invention.

[0104] FIG. 14 shows a schematic diagram of the LGD double-well model showing stable fully polarized states in different directions due to the position change of copper ions (Cu+).

[0105] The CIPS material is a crystalline substance capable of spontaneous polarization that changes with an external electric field. The lattice structure of the CIPS material includes ionic bonds of the [P2S6]4- anion with Cu+ and In2+ cations. In each single layer, the positional deviation of Cu+ ions breaks the symmetry of the lattice structure and shows a ferroelectric effect in which the polarized state is maintained even when the external electric field disappears.

[0106] As can be seen from FIG. 14, the LGD double-well model shows two thermodynamically equivalent polarized states represented by the CIPS material (a state in which all dipoles are arranged in the downward direction, or a state in which all dipoles are arranged in the upward direction). Therefore, it can be seen that in order to change the position of copper ions (Cu+) and reach other polarized states of the LGD double-well model, it is necessary to supply external energy in the form of an electric field.

[0107] FIG. 15 is a schematic diagram of the P-V relationship of the ferroelectric layer according to the experimental example of the present invention.

[0108] Fig. 15 shows a schematic P-V relationship diagram representing the degree of polarization (P, μC / cm 2 ) generated in the ferroelectric layer by the voltage (Voltage, V) applied to the ferroelectric structure according to the experimental example.

[0109] From Fig. 15, it can be confirmed that when external voltages of 10 V and -10 V are applied, a saturated polarization (Ps) occurs, which means the moment of reaching different stable polarization states from each other in the LGD double-well model. Also, it can be confirmed that the two results under the condition of remnant polarization (Pr) without applying a further electric field have values lower than the saturated polarization intensity.

[0110] On the other hand, regarding the coercive voltage (Vc) at which the alignment of dipoles inside the ferroelectric layer starts due to the programmed electric field, it can be confirmed that the coercive voltages corresponding to both directions are slightly different. This is caused by various external factors (such as electrodes, thermal fluctuations, etc.) and acts as the reason for the asymmetric manifestation of the I-V characteristics of the ferroelectric layer described later.

[0111] Fig. 16 is a diagram for explaining the PFM analysis result of the ferroelectric structure according to the experimental example of the present invention in a state where no voltage is applied. Fig. 17 is a diagram for explaining the PFM analysis result of the ferroelectric structure according to the experimental example of the present invention in a state where a voltage of 3 V is applied. Fig. 18 is a diagram for explaining the PFM analysis result of the ferroelectric structure according to the experimental example of the present invention in a state where a voltage of 4 V is applied. Fig. 19 is a diagram for explaining the PFM analysis result of the ferroelectric structure according to the experimental example of the present invention in a state where a voltage of 5 V is applied. Fig. 20 is a diagram for explaining the PFM analysis result of the ferroelectric structure according to the experimental example of the present invention in a state where a voltage of 10 V is applied.

[0112] As shown in FIGS. 16 to 20, in order to confirm the partial polarization phenomenon occurring in the ferroelectric layer due to an external electric field (voltage application), polarization analysis was performed using PFM (Piezoresponse Force Microscopy) within a 30×30 μm^2 region of the ferroelectric layer. Specifically, FIGS. 16 to 20(a) show schematic diagrams of the state in which domains grow in the ferroelectric layer due to the applied voltage, FIGS. 16 to 20(b) show phase (Phase, deg) profiles according to the longitudinal position (Length, nm) of the ferroelectric layer, FIGS. 16 to 20(c) show PFM images of the ferroelectric layer with a voltage applied, and FIGS. 16 to 20(d) show intensity (Intensity, a.u.) profiles according to the phase (deg).

[0113] It can be seen from FIGS. 16 to 20 that as the applied voltage increases, the domain formation ratio increases. Here, the light and dark hue contrast appearing in the PFM image indicates a 180-degree phase difference, which means the dipole arrangement in the upper and lower directions or domain formation.

[0114] As shown in FIG. 16, it can be seen that in the ferroelectric layer (CIPS), all dipoles are aligned in the upward direction with a 100% probability under the condition that no external voltage is applied. Different from this, as can be seen from FIG. 17, when a voltage of 3V greater than the step-down voltage is applied, a change in the dipole arrangement begins to occur within the ferroelectric layer (CIPS), and it can be seen that the dipoles aligned in the upward direction exist with a 64% probability, and the dipoles aligned in the downward direction exist with a 36% probability. Different from this, as shown in FIG. 18, as a result of increasing the applied voltage to 4V, it can be seen that the dipoles aligned in the upward direction exist at 52%, and the dipoles aligned in the downward direction exist at 48%. Different from this, as can be seen from FIG. 19, as a result of increasing the applied voltage to 5V, it can be seen that the dipoles aligned in the upward direction exist at 37%, and the dipoles aligned in the downward direction exist at 63%. Different from this, as can be seen from FIG. 20, as a result of increasing the applied voltage to 10V, it can be confirmed that all dipoles are aligned in the downward direction with a 100% probability.

[0115] As a result, the results confirmed from FIGS. 16 to 20, similar to the polarization-voltage (P-E) results confirmed in FIG. 15, show a saturation voltage state where all dipoles are completely polarized in the same direction at a voltage of 10V. Also, such results suggest that all domains do not change by a single instant (one voltage application).

[0116] (Experimental Example 3: Evaluation of Current Characteristics of Ferroelectric Structure and Binarization and Key Classification Based Thereon) FIG. 21 is a diagram showing the current-voltage characteristics of the ferroelectric structure according to the experimental example of the present invention.

[0117] As shown in FIG. 21, voltages in different ranges (±3V, ±4V, ±5V) are applied to the ferroelectric structure according to the experimental example, and the measurement results of current values (I, A) for each case are shown.

[0118] From FIG. 21, a distinct reversal of the polarization state can be confirmed under the positive (+) voltage condition. From this, it can be inferred that due to the external electric field applied to the ferroelectric structure, a change in the position of copper ions (Cu+) in the ferroelectric layer (CIPS) occurs, thereby changing the dipole arrangement.

[0119] Also, it can be confirmed that a sharp increase in current intensity starts to occur at a point of about +2V. From this, it can be seen that a conversion from a high resistance state to a low resistance state due to the generation of polarization has occurred starting from about +2V.

[0120] FIG. 22 is a diagram for explaining the voltage application process for acquiring a random current signal by the ferroelectric structure according to the experimental example of the present invention.

[0121] In FIG. 22, the operating voltage (V program ) for forming a random polarization state in the ferroelectric structure according to the experimental example, the read voltage (V read ) for reading the current signal corresponding thereto, and thereafter, the removal voltage (V for returning to the initial conditionerase ) shows the process of applying. Each voltage is applied in a pulse shape, and the time interval between pulses is set to 150 ms. Repeated removal voltage application and operating voltage application are the main sources of entropy for implementing a true random number generator. For the quantitative evaluation of random current signals due to dipole polarization, a repeated cycle of operating voltage application → readout voltage application → current value acquisition → removal voltage application was performed 784 times in this order.

[0122] Figure 23 is a diagram showing the current values measured while a 3V operating voltage and a 0.5V readout voltage were applied during 784 cycles, and Figure 24 is a diagram showing the current distribution histogram based on the results of Figure 23.

[0123] In Figure 23, the change in the current value measured during the process of inputting a constant operating voltage (V prog = 3V) and a constant readout voltage (V read = 0.5V) for 784 times is shown. In Figure 24, the current values (Current, nA) obtained in Figure 23 are classified by the number of occurrences (Count), and a current distribution histogram approximated by a Gaussian function is shown.

[0124] From Figures 23 and 24, it can be confirmed that random current values are obtained during the process of inputting a constant operating voltage (3V) and a constant readout voltage (0.5V) for 784 times, and the average value of the obtained current values was confirmed to be 0.434 nA.

[0125] Figure 25 is a diagram showing the current values measured while a 3V operating voltage and a 1.0V readout voltage were applied during 784 cycles, and Figure 26 is a diagram showing the current distribution histogram based on the results of Figure 25.

[0126] In Figure 25, a constant operating voltage (V prog = 3V) and a constant readout voltage (V readThe change in the measured current value during the process in which a constant operating voltage (3V) and a constant read voltage (1.0V) are input 784 times is shown. Fig. 26 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Fig. 25 by the number of occurrences (Count).

[0127] From Figs. 25 and 26, it can be confirmed that random current values are obtained during the process in which a constant operating voltage (3V) and a constant read voltage (1.0V) are input 784 times, and the average value of the obtained current values was confirmed to be 0.612 nA.

[0128] Fig. 27 is a diagram showing the current values measured while a 3V operating voltage and a 1.5V read voltage are applied during 784 cycles, and Fig. 28 is a diagram showing a current distribution histogram based on the results of Fig. 27.

[0129] In Fig. 27, the change in the measured current value during the process in which a constant operating voltage (V prog = 3V) and a constant read voltage (V read = 1.5V) are input 784 times is shown. Fig. 28 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Fig. 27 by the number of occurrences (Count).

[0130] From Figs. 27 and 28, it can be confirmed that random current values are obtained during the process in which a constant operating voltage (3V) and a constant read voltage (1.5V) are input 784 times, and the average value of the obtained current values was confirmed to be 0.746 nA.

[0131] Fig. 29 is a diagram showing the current values measured while a 4V operating voltage and a 0.5V read voltage are applied during 784 cycles, and Fig. 30 is a diagram showing a current distribution histogram based on the results of Fig. 29.

[0132] In Fig. 29, a constant operating voltage (V prog = 4V) and a constant read voltage (V readThe change in the measured current value during the process in which a constant operating voltage (4 V) and a constant read voltage (0.5 V) are input 784 times is shown. Fig. 30 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Fig. 29 by the number of occurrences (Count).

[0133] From Figs. 29 and 30, it can be seen that random current values are obtained during the process in which a constant operating voltage (4 V) and a constant read voltage (0.5 V) are input 784 times, and the average value of the obtained current values was confirmed to be 0.953 nA.

[0134] Fig. 31 is a diagram showing the current values measured while a 4 V operating voltage and a 1.0 V read voltage are applied during 784 cycles, and Fig. 32 is a diagram showing a current distribution histogram based on the results of Fig. 31.

[0135] In Fig. 31, the change in the measured current value during the process in which a constant operating voltage (V prog = 4 V) and a constant read voltage (V read = 1.0 V) are input 784 times is shown. Fig. 32 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Fig. 31 by the number of occurrences (Count).

[0136] From Figs. 31 and 32, it can be seen that random current values are obtained during the process in which a constant operating voltage (4 V) and a constant read voltage (1.0 V) are input 784 times, and the average value of the obtained current values was confirmed to be 0.993 nA.

[0137] Fig. 33 is a diagram showing the current values measured while a 4 V operating voltage and a 1.5 V read voltage are applied during 784 cycles, and Fig. 34 is a diagram showing a current distribution histogram based on the results of Fig. 33.

[0138] In Fig. 33, a constant operating voltage (V prog = 4 V) and a constant read voltage (V readshows the change in the measured current value during the process in which a constant operating voltage (4V) and a constant read voltage (1.5V) are input 784 times. Figure 34 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Figure 33 by the number of occurrences (Count).

[0139] From Figures 33 and 34, it can be confirmed that random current values are obtained during the process in which a constant operating voltage (4V) and a constant read voltage (1.5V) are input 784 times, and the average value of the obtained current values was confirmed to be 1.007 nA.

[0140] Figure 35 shows the current values measured while a 5V operating voltage and a 0.5V read voltage are applied during 784 cycles. Figure 36 shows a current distribution histogram based on the results of Figure 35.

[0141] Figure 35 shows the change in the measured current value during the process in which a constant operating voltage (V prog = 5V) and a constant read voltage (V read = 0.5V) are input 784 times. Figure 36 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Figure 35 by the number of occurrences (Count).

[0142] From Figures 35 and 36, it can be seen that random current values are obtained during the process in which a constant operating voltage (5V) and a constant read voltage (0.5V) are input 784 times, and the average value of the obtained current values was confirmed to be 0.831 nA.

[0143] Figure 37 shows the current values measured while a 5V operating voltage and a 1.0V read voltage are applied during 784 cycles. Figure 38 shows a current distribution histogram based on the results of Figure 37.

[0144] Figure 37 shows a constant operating voltage (V prog = 5V) and a constant read voltage (V readThe change in the measured current value during the process in which a constant operating voltage (5 V) and a constant read voltage (1.0 V) are input 784 times is shown. Fig. 38 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Fig. 37 by the number of occurrences (Count).

[0145] From Figs. 37 and 38, it can be seen that random current values are obtained during the process in which a constant operating voltage (5 V) and a constant read voltage (1.0 V) are input 784 times, and the average value of the obtained current values was confirmed to be 1.011 nA.

[0146] Fig. 39 is a diagram showing the current values measured while a 5 V operating voltage and a 1.5 V read voltage are applied during 784 cycles, and Fig. 40 is a diagram showing a current distribution histogram based on the results of Fig. 39.

[0147] In Fig. 39, the change in the measured current value during the process in which a constant operating voltage (V prog = 5 V) and a constant read voltage (V read = 1.5 V) are input 784 times is shown. Fig. 40 shows a current distribution histogram approximated by a Gaussian function, classifying the current values (Current, nA) obtained in Fig. 39 by the number of occurrences (Count).

[0148] From Figs. 39 and 40, it can be seen that random current values are obtained during the process in which a constant operating voltage (5 V) and a constant read voltage (1.5 V) are input 784 times, and the average value of the obtained current values was confirmed to be 1.401 nA.

[0149] The average current values confirmed in Figs. 23 to 40 are as shown in Table 1 below.

[0150]

Table 1

[0151] Fig. 41 is a diagram explaining the binarization and key classification process for the random current values obtained from the ferroelectric structure according to the experimental example of the present invention.

[0152] FIG. 41 shows the process of binarization and the key classification process for the current values measured while a constant operating voltage and a constant read voltage are applied during 784 cycles.

[0153] More specifically, the current values were measured while a constant operating voltage and a constant read voltage were applied during 784 cycles, and the average current value for the measured current values during 784 cycles was calculated. Also, based on the calculated average current value, the current values greater than or equal to the average current value were binarized to the state of "1", and the current values less than the average current value were binarized to the state of "0". Through the above-described process, 784 bits were generated, and for the generated 784 bits, 28 were composed of one key, and classified into a total of 28 keys.

[0154] FIG. 42 is a diagram showing the binarization and key classification results for the case where an operating voltage of 3V is applied in an image, FIG. 43 is a diagram showing the binarization and key classification results for the case where an operating voltage of 4V is applied in an image, and FIG. 44 is a diagram showing the binarization and key classification results for the case where an operating voltage of 5V is applied in an image.

[0155] FIGS. 42 to 44 show the results expressed as 28×28 pixel color images after binarization and key classification were performed by the method described in FIG. 41. It can be seen from FIGS. 42 to 44 that different random images are represented depending on the applied operating voltage.

[0156] FIG. 45 is a diagram showing the histograms of the current values obtained at different read voltages.

[0157] Figure 45 shows a histogram of the number of occurrences (Counts) versus the current values (I, nA) obtained when read voltages (Vread, V) of 0.5 V, 1.0 V, and 1.5 V are applied after applying an operating voltage of 4 V. From Figure 45, it can be seen that the variation of the histogram depends greatly on the conditions of the read voltage, and the stochasticity of observing the current signal can be confirmed under read voltage conditions lower than the step-down voltage.

[0158] Figure 46 is a diagram showing a time-difference plot of current fluctuations for a read voltage of 0.5 V, Figure 47 is a diagram showing a time-difference plot of current fluctuations for a read voltage of 1.0 V, and Figure 48 is a diagram showing a time-difference plot of current fluctuations for a read voltage of 1.5 V.

[0159] From Figures 46 to 48, the degree of current fluctuation (time lag plot, TLP) for read voltages of 0.5 V, 1.0 V, and 1.5 V is shown. From Figures 46 to 48, an independent correlation between the intensity (In, nA) of the nth current signal and the intensity (In+1, nA) of the (n + 1)th current signal can be confirmed.

[0160] Figure 49 is a diagram showing the magnitude of current fluctuations according to the intensity of the read voltage.

[0161] As shown in Figure 49, the magnitudes of current fluctuations (I, nA) for read voltages (Vread, V) of 0.5 V, 1.0 V, and 1.5 V are shown. Variability or uniformity acts as an important mediator variable for identifying ideal operating conditions.

[0162] From Figure 49, it can be seen that when the operating voltage is 4 V and the read voltage is 0.5 V, the variability (σ) is calculated to be 7.5%, when the operating voltage is 4 V and the read voltage is 1.0 V, the variability (σ) is calculated to be 6.3%, and when the operating voltage is 4 V and the read voltage is 1.5 V, the variability (σ) is calculated to be 5.5%.

[0163] (Experimental Example 4: Derivation of Ideal Operating Voltage and Readout Voltage for the Ferroelectric Structure to Have Ideal Random Number Generation Performance) The uniformity value was derived by Equation 1 below, and the entropy value was derived by Equation 2 below. In the case of an ideal random number generator, since the occurrence probability of the 0 state and the occurrence probability of the 1 state are equivalent to 50%, the uniformity value has 0.5, and the entropy value has 1. Also, in explaining Experimental Example 4, binarization and key classification were performed as described in FIG. 41.

[0164]

Equation

[0165]

Equation

[0166] FIG. 50 is a diagram showing the uniformity value and the entropy value derived in a state where an operating voltage of 3V and a readout voltage of 0.5V are applied.

[0167] As shown in FIG. 50, for each of the 28 keys classified in a state where an operating voltage of 3V and a readout voltage of 0.5V are applied, the uniformity value (Uniformity) and the entropy value (Entropy) are measured and shown.

[0168] From FIG. 50, it can be seen that when an operating voltage of 3V and a readout voltage of 0.5V are applied, the uniformity value shows a large difference from 0.5, and the entropy value shows a large difference from 1. That is, it can be seen that the operating voltage of 3V and the readout voltage of 0.5V are not the ideal operating voltage and readout voltage.

[0169] FIG. 51 is a diagram showing uniformity values and entropy values derived in a state where an operating voltage of 3V and a read voltage of 1.0V are applied.

[0170] As shown in FIG. 51, for each of the 28 keys classified in a state where an operating voltage of 3V and a read voltage of 1.0V are applied, the uniformity value and the entropy value are measured and shown.

[0171] From FIG. 51, it can be confirmed that when an operating voltage of 3V and a read voltage of 1.0V are applied, the uniformity value shows a large difference from 0.5, and the entropy value shows a small difference from 1. That is, it can be seen that the operating voltage of 3V and the read voltage of 1.0V are not ideal operating voltage and read voltage.

[0172] FIG. 52 is a diagram showing uniformity values and entropy values derived in a state where an operating voltage of 3V and a read voltage of 1.5V are applied.

[0173] In FIG. 52, for each of the 28 keys classified in a state where an operating voltage of 3V and a read voltage of 1.5V are applied, the uniformity value and the entropy value are measured and shown.

[0174] From FIG. 52, it can be seen that when an operating voltage of 3V and a read voltage of 1.5V are applied, the uniformity value shows a large difference from 0.5, and the entropy value shows a large difference from 1. That is, it can be seen that the operating voltage of 3V and the read voltage of 1.5V are not ideal operating voltage and read voltage.

[0175] FIG. 53 is a diagram showing uniformity values and entropy values derived in a state where an operating voltage of 4V and a read voltage of 0.5V are applied.

[0176] For each of the 28 keys (key) classified in the state where an operating voltage of 4V and a read voltage of 0.5V are applied, the uniformity value and the entropy value are measured and shown in FIG. 53.

[0177] From FIG. 53, it can be seen that when an operating voltage of 4V and a read voltage of 0.5V are applied, the uniformity value shows a large difference from 0.5 and the entropy value shows a small difference from 1. That is, it can be seen that the operating voltage of 4V and the read voltage of 0.5V are not ideal operating voltage and read voltage.

[0178] FIG. 54 is a diagram showing the uniformity value and the entropy value derived in the state where an operating voltage of 4V and a read voltage of 1.0V are applied.

[0179] In FIG. 54, for each of the 28 keys (key) classified in the state where an operating voltage of 4V and a read voltage of 1.0V are applied, the uniformity value (Uniformity) and the entropy value (Entropy) are measured and shown.

[0180] From FIG. 54, it can be seen that when an operating voltage of 4V and a read voltage of 1.0V are applied, the uniformity value shows a large difference from 0.5 and the entropy value shows a small difference from 1. That is, it can be seen that the operating voltage of 4V and the read voltage of 1.0V are not ideal operating voltage and read voltage.

[0181] FIG. 55 is a diagram showing the uniformity value and the entropy value derived in the state where an operating voltage of 4V and a read voltage of 1.5V are applied.

[0182] In FIG. 55, for each of the 28 keys (key) classified in the state where an operating voltage of 4V and a read voltage of 1.5V are applied, the uniformity value (Uniformity) and the entropy value (Entropy) are measured and shown.

[0183] From FIG. 55, it can be seen that when an operating voltage of 4V and a read voltage of 1.5V are applied, the uniformity values for all keys (0 to 28 keys) appear close to 0.5, and the entropy values for all keys (0 to 28 keys) appear close to 1. That is, it can be seen that an operating voltage of 4V and a read voltage of 1.5V are ideal operating and read voltages.

[0184] FIG. 56 is a diagram showing the uniformity values and entropy values derived in a state where an operating voltage of 5V and a read voltage of 0.5V are applied.

[0185] In FIG. 56, the uniformity values (Uniformity) and entropy values (Entropy) are measured and shown for each of the 28 keys (key) classified in a state where an operating voltage of 5V and a read voltage of 0.5V are applied.

[0186] From FIG. 56, it can be seen that when an operating voltage of 5V and a read voltage of 0.5V are applied, the uniformity value shows a large difference from 0.5, and the entropy value shows a large difference from 1. That is, it can be seen that an operating voltage of 5V and a read voltage of 0.5V are not ideal operating and read voltages.

[0187] FIG. 57 is a diagram showing the uniformity values and entropy values derived in a state where an operating voltage of 5V and a read voltage of 1.0V are applied.

[0188] In FIG. 57, the uniformity values (Uniformity) and entropy values (Entropy) are measured and shown for each of the 28 keys (key) classified in a state where an operating voltage of 5V and a read voltage of 1.0V are applied.

[0189] From FIG. 57, it can be seen that when an operating voltage of 5V and a read voltage of 1.0V are applied, the uniformity value shows a large difference from 0.5, and the entropy value shows a large difference from 1. That is, it can be seen that an operating voltage of 5V and a read voltage of 1.0V are not ideal operating and read voltages.

[0190] FIG. 58 is a diagram showing uniformity values and entropy values derived in a state where an operating voltage of 5V and a read voltage of 1.5V are applied.

[0191] FIG. 58 shows the measurement of the uniformity value (Uniformity) and the entropy value (Entropy) for each of the 28 keys (keys) classified in a state where an operating voltage of 5V and a read voltage of 1.5V are applied.

[0192] From FIG. 58, it can be seen that when an operating voltage of 5V and a read voltage of 1.5V are applied, the uniformity value shows a large difference of 0.5, and the entropy value shows a small difference of 1. That is, it can be seen that the operating voltage of 5V and the read voltage of 1.5V are not ideal operating voltages and read voltages.

[0193] FIG. 59 is a diagram visualizing the uniformity value (Uniformity) at different operating voltages and read voltages, and FIG. 60 is a diagram visualizing the entropy value (Entropy) at different operating voltages and read voltages.

[0194] FIGS. 59 and 60 show the average uniformity value (average value of the uniformity values measured with 28 keys) and the average entropy value (average value of the entropy values measured with 28 keys) at different operating voltages (3V, 4V, 5V) and different read voltages (0.5V, 1.0V, 1.5V) in a color map image.

[0195] From FIG. 59, it can be seen that the average uniformity value measured at an operating voltage of 4V and a read voltage of 1.5V is close to 0.5, and from FIG. 60, it can be seen that the average entropy value measured at an operating voltage of 4V and a read voltage of 1.5V is close to 1.

[0196] As a result, it can be reconfirmed that the ideal operating voltage for the ferroelectric structure according to the experimental example to have ideal random number generation performance is 4V, and the read voltage is 1.5V.

[0197] (Experimental Example 5: Calculation of Additional Mediating Variables for Verifying the Randomness of the Ferroelectric Structure) Based on the ideal operating voltage of 4V and the read voltage of 1.5V derived from Experimental Example 4, the Hamming distance (HDintra) and the correlation coefficient (CCintra) that can determine the randomness of the output results of the ferroelectric structure according to the experimental example were calculated.

[0198] To evaluate the randomness between the binarized keys, the Hamming distance (HDintra) was used, which is defined as the number of bit substitutions required to convert one key to another key and was used to compare keys generated by the same device.

[0199] To form a key considered to be cryptographically secure, the Hamming distance must be 50% or more. For example, in the case of a key composed of 28 bits, the Hamming distance should be 14. In contrast, keys with Hamming distance values that are too low or too high are known to be relatively easily decodable by Brute Force Testing, which is a method of hacking.

[0200] Figure 61 is a diagram showing the Hamming distance histogram of the ferroelectric structure according to the experimental example of the present invention, and Figure 62 is a diagram visualizing the average Hamming distance values at different operating voltages and read voltages.

[0201] Figure 61 shows 378 Hamming distance histogram images of keys composed of 28 bits measured under the conditions of an operating voltage of 4V and a read voltage of 1.5V. It can be seen from Figure 61 that the average Hamming distance value of the ferroelectric structure according to the experimental example is 14.01, which is close to the ideal value, indicating that the ferroelectric structure according to the experimental example has excellent encryption ability.

[0202] Fig. 62 shows a color map image of the average Hamming distance value (Intra-HD) under different operating voltages and readout voltage conditions. From Fig. 62, it can be seen that under the operating voltage of 4V and the readout voltage condition of 1.5V, the average Hamming distance value appears close to 14.

[0203] Fig. 63 is a diagram showing an image in which the value of π is estimated using Monte-Carlo simulation.

[0204] In Fig. 63, the value of π was estimated using Monte-Carlo simulation. More specifically, the above-described method used the fact that the value obtained by dividing the area of a circle with a radius of r by the area of a square with a length of 2r is π / 4. To embody this, as shown in Fig. 63, a binarized color map image composed of 28×28 pixels was embodied using the binarized key obtained from the ferroelectric structure according to the experimental example. Here, dark-colored pixels represent bit "1", and bright-colored pixels represent bit "0". Next, the ratio of the number of bright-colored pixels inside the largest circle to the number of bright-colored pixels inside the square was calculated to estimate the value of π. As a result of the calculation, it can be seen that in the case of the ferroelectric structure according to the experimental example, the value of π shows 3.1 close to the actual value.

[0205] Fig. 64 is a diagram showing the self-correlation images of different keys of the ferroelectric structure according to the experimental example of the present invention.

[0206] As shown in Fig. 64, the self-correlation relationship is shown as a function of bit delay for each of the 28 keys with a 28-bit length obtained under the operating voltage of 4V and the readout voltage condition of 1.5V.

[0207] From Fig. 64, no spike of a high size is confirmed, and thus it can be seen that there is substantially no periodicity in the key. That is, it can be seen that the bit stream generated by the ferroelectric structure according to the experimental example is essentially a random point.

[0208] FIG. 65 is a diagram showing a correlation coefficient histogram of the ferroelectric structure according to the experimental example of the present invention, and FIG. 66 is a diagram visualizing the average correlation coefficient values at different operating voltages and read voltages.

[0209] As shown in FIGS. 65 and 66, a correlation coefficient (CCintra) verification was performed to verify the periodicity between different bits and determine the degree of correlation. The autocorrelation function (ACF) is located in the interval [-1, 1]. A value of -1 means semi-correlation, a value of 1 means correlation, and a value of 0 means no correlation between each other. Also, the results in FIGS. 65 and 66 show the autocorrelation relationships for each of the 28-bit keys measured under the conditions of an operating voltage of 4V and a read voltage of 1.5V.

[0210] From FIG. 65, it can be confirmed that the correlation coefficient value is about -0.02. Thus, it can be seen that the ferroelectric structure according to the experimental example has excellent random number generation ability as a true random number generator (TRNG). Also, from FIG. 62, it can be confirmed that the average correlation coefficient value appears close to -0.02 under the conditions of an operating voltage of 4V and a read voltage of 1.5V.

[0211] (Experimental Example 6: Evaluation of the International Standard Encryption Performance of the Ferroelectric Structure) Using the standard statistical test package (NIST sp 800-22 rev. 1a) developed by the National Institute of Standards and Technology in the United States, the randomness evaluation of the ferroelectric structure according to the experimental example as a true random number generator (TRNG) was performed.

[0212] The purpose of the aforementioned Experimental Example 6 is to determine whether the random number generator based on the ferroelectric structure according to the experimental example having a metal-CIPS-metal structure exhibits performance suitable for random number generation for security purposes. For this reason, a total of 15,680 bits of random current data were extracted, and this was evaluated against the null hypothesis for randomness / non-randomness, and the relevant data evaluation was performed using a test protocol that returns a P-value at a 99% confidence level. Here, a bit is regarded as truly random only when the P-value is greater than 0.001.

[0213] Figure 67 is a diagram for explaining the international standard encryption performance evaluation results according to Experimental Example 6.

[0214] As shown in Figure 67, it can be seen that all NIST tests were successfully passed without post-processing steps. From this, it can be seen that the suitability of the true random number generator based on the ferroelectric structure according to the experimental example for the field of encryption is high.

[0215] (Experimental Example 7: Verification of Random Number Generation Medium Variables between Different Ferroelectric Structures) Physical non-replicability is a basic requirement for a true random number generator. Specifically, it is to prevent anyone from refuting the random number generator because the random number results generated by one random number generator cannot be replicated by other random number generators.

[0216] In Experimental Example 7, to evaluate the non-replicability of the true random number generator based on the ferroelectric structure according to the experimental example, 10 ferroelectric structures according to the experimental example were prepared, and for each of them under the operating voltage condition of 4V and the readout voltage condition of 1.5V, 28 bit keys were generated.

[0217] Figure 68 is a diagram showing the Hamming distance histogram between 28 keys obtained from 10 pairs of different ferroelectric structures, and Figure 69 is a diagram visualizing the average Hamming distance values of 10 pairs of different ferroelectric structures.

[0218] As shown in FIG. 68, a histogram of Hamming distances (HDinter) between 28 keys obtained from 10 different ferroelectric structure pairs is shown, and as shown in FIG. 69, a color map image of the average Hamming distance (Mean inter-HD) values of 10 different ferroelectric structure pairs is shown. In FIG. 68, inter means the distance between key pairs generated using different structures. From FIGS. 68 and 69, it can be confirmed that the average Hamming distance value is close to the ideal value of 14, which is a result demonstrating that the random number results generated by different structures are unique.

[0219] FIG. 70 is a diagram showing a histogram of correlation coefficients between 28 keys obtained from 10 different ferroelectric structure pairs, and FIG. 71 is a diagram visualizing the average correlation coefficient values of 10 different ferroelectric structure pairs.

[0220] FIG. 70 shows a histogram of correlation coefficients (CCinter) between 28 keys obtained from 10 different ferroelectric structure pairs, and FIG. 71 shows a color map image of the average correlation coefficient (Mean inter-CC) values of 10 different ferroelectric structure pairs. From FIGS. 70 and 71, it can be confirmed that the average correlation coefficient value is close to the ideal value of 0, which is a result demonstrating that there is no correlation between the random number results generated by different structures.

[0221] As described above, the present invention has been described in detail using preferred embodiments. However, the scope of the present invention is not limited to specific embodiments and should be analyzed according to the appended claims. Also, those having ordinary knowledge in the relevant technical field will understand that many modifications and variations are possible without departing from the scope of the present invention.

Explanation of Reference Numerals

[0222] 100 Substrate 110 Si Substrate 120 SiO2 Layer 200 Lower electrode 300 Ferroelectric layer 310, 320 First ferroelectric layer, second ferroelectric layer 400 Upper electrode, intermediate electrode 410, 420, 430, 440 First to fourth upper electrodes 500 Upper electrode

Claims

1. A substrate; a lower electrode disposed on the substrate; A molybdenum disulfide (MoS) is disposed on the lower electrode. 2 ) a ferroelectric layer comprising a top electrode disposed on the ferroelectric layer; A ferroelectric structure comprising: a first dielectric layer and a second dielectric layer, the second dielectric layer and the third dielectric layer being arranged such that for the same applied voltage, random magnitude current values ​​are obtained.

2. 2. A ferroelectric structure according to claim 1, characterized in that when the same voltage is applied multiple times, current values ​​of different intensities are generated for each applied voltage.

3. 2. A ferroelectric structure according to claim 1, characterized in that polarization occurs in the ferroelectric layer when a voltage is applied.

4. 4. A ferroelectric structure according to claim 3, characterized in that the polarization occurring in the ferroelectric layer results in current values ​​of random magnitude for the same applied voltage.

5. 2. A ferroelectric structure according to claim 1, characterized in that it is applied to a true random number generator.

6. A substrate; a lower electrode disposed on the substrate; a ferroelectric layer disposed on the lower electrode and including a two-dimensional ferroelectric material; a top electrode disposed on the ferroelectric layer; A ferroelectric structure characterized in that for identical applied voltages, random magnitude current values ​​are obtained.

7. 7. A ferroelectric structure according to claim 6, characterized in that when the same voltage is applied multiple times, current values ​​of different magnitudes are generated for each applied voltage.

8. 7. A ferroelectric structure according to claim 6, characterized in that a polarization occurs in the ferroelectric layer when a voltage is applied.

9. 9. A ferroelectric structure according to claim 8, wherein, when a voltage is applied, a perpendicular (out-of-plane, OOP) polarization is generated in the ferroelectric layer.

10. The two-dimensional ferroelectric material includes compounds of different elements, 7. The ferroelectric structure of claim 6, wherein, when a voltage is applied, element ions of any one of the different elements contained in the two-dimensional ferroelectric material change position within the ferroelectric layer.

11. 11. A ferroelectric structure as claimed in claim 10, characterized in that elemental ions which change position within the ferroelectric layer alter the crystal structure of the ferroelectric layer.

12. 12. A ferroelectric structure according to claim 11, wherein the ferroelectric layer changes from a symmetric crystal structure to an anti-symmetric crystal structure.

13. A ferroelectric structure according to claim 6, characterized in that it is applied in a true random number generator.

14. Providing a substrate; forming a bottom electrode on the substrate; forming a ferroelectric layer on the bottom electrode, the ferroelectric layer comprising a two-dimensional ferroelectric material; and forming a top electrode on said ferroelectric layer.

15. providing a ferroelectric structure including a substrate, a bottom electrode disposed on the substrate, a ferroelectric layer disposed on the bottom electrode and including a two-dimensional ferroelectric material, and a top electrode disposed on the ferroelectric layer; applying an actuation voltage to the ferroelectric structure such that polarization occurs in the ferroelectric layer; applying a read voltage to the ferroelectric structure to obtain a current value reflecting the polarization state of the ferroelectric layer; applying an excision voltage to said ferroelectric structure such that said ferroelectric layer returns to its pre-polarization state.

16. The steps of applying the operating voltage, obtaining the current value, and applying the removal voltage are repeated in sequence; 16. A method for operating a ferroelectric structure according to claim 15, characterized in that the same operating voltage is repeatedly applied.

Images