Non-volatile memory and methods of forming the same

By setting an erase structure between storage cells, the data erasure method is improved, the dependence on the tunneling oxide layer is reduced, the reliability problem of non-volatile memory caused by excessive tunneling is solved, and the reliability and lifespan of the memory are improved.

CN123075907APending Publication Date: 2026-11-10GUANGZHOU ZENGXIN TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611289486.4
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-24
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

In the prior art, the reliance of non-volatile memory on the tunneling oxide layer during data programming and erasure processes leads to insufficient reliability and easy degradation when there are too many tunneling cycles.

Method used

An erase structure is set between memory cells, including a first tunneling layer, a second tunneling layer and an erase gate. By applying a voltage to the erase gate, electrons in the floating gate migrate through the tunneling layer to the erase gate, reducing the dependence on the tunneling oxide layer and improving the data erasure method.

Benefits of technology

By reducing the number of times the tunneling oxide layer is penetrated during data erasure, the reliability of the tunneling oxide layer is improved, thereby enhancing the reliability and lifespan of the non-volatile memory.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123075907A_ABST
    Figure CN123075907A_ABST
Patent Text Reader

Abstract

Embodiments of the present application provide a non-volatile memory and a forming method thereof. The non-volatile memory comprises: a substrate layer; a first drain, a source and a second drain located in the substrate layer; a storage unit located on the surface of the substrate layer, and a projection of the storage unit on the substrate layer is located between the source and the first drain and / or between the source and the second drain; an erase structure located on the surface of the substrate layer and between two adjacent storage units, the erase structure comprising a first tunnel layer, a second tunnel layer and an erase gate located between the first tunnel layer and the second tunnel layer; the erase structure is configured to receive an erase voltage when a data erase operation is performed on the storage unit, so that the charge in the storage unit moves to the erase gate through the first tunnel layer or the second tunnel layer. Through the arrangement of the above technical features, the reliability of the non-volatile memory can be improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the field of semiconductor technology, and in particular to a non-volatile memory and a method for forming the same. Background Technology

[0002] Conventional stacked-gate or non-form flash memory employs a channel hot electron injection mechanism during data programming. Different forward high voltages are applied to the source, drain, and control gate, accelerating charge carriers within the channel to form high-energy hot electrons that tunnel through the substrate through the tunneling oxide layer into the floating gate. During data erasure, a FN (Fowler-Nordheim) tunneling mechanism is used. A strong reverse electric field is applied between the floating gate and the silicon substrate, allowing electrons stored in the floating gate to return to the substrate via the tunneling oxide layer through the quantum tunneling effect.

[0003] Since data programming and data erasure operations both require passing through the tunnel oxide layer, the quality requirements for the tunnel oxide layer in related technologies are very high. Otherwise, it is easy to degrade when the number of tunneling cycles is too high, resulting in insufficient programming and erasure cycles, thereby affecting the reliability of the memory. Summary of the Invention

[0004] In view of the above-mentioned deficiencies in the related technologies, the technical problem to be solved by this application includes at least how to improve the reliability of memory.

[0005] To address at least one of the aforementioned technical problems, this application proposes a non-volatile memory and a method for forming the same.

[0006] According to a first aspect of this application, a non-volatile memory is provided, comprising: Substrate layer; The first drain, the source, and the second drain are located in the substrate layer; A memory cell is located on the surface of a substrate, and the projection of the memory cell onto the substrate is located between the source and the first drain and / or between the source and the second drain. An erase structure is located on the surface of the substrate layer and between two adjacent memory cells. The erase structure includes a first tunneling layer, a second tunneling layer, and an erase gate, with the erase gate located between the first tunneling layer and the second tunneling layer. The erase structure is configured to receive an erase voltage when a data erase operation is performed on the memory cell, so that the charge in the memory cell moves to the erase gate via a first tunneling layer or a second tunneling layer.

[0007] In some embodiments, the storage unit includes: It then penetrates the oxide layer and is located on the surface of the substrate layer where the first drain electrode is disposed; The floating gate is located on the surface of the tunneling oxide layer away from the substrate layer; A composite dielectric layer is located on the surface of the floating gate away from the substrate layer; The control gate is located on the surface of the composite dielectric layer away from the substrate layer; An insulating oxide layer is located on the surface of the control gate away from the substrate layer.

[0008] In some embodiments, the storage unit further includes: The first sidewall is located on the side of the memory cell away from the erase structure. In the stacking direction of the memory cell, the first sidewall covers the floating gate, the composite dielectric layer, the control gate, and the insulating oxide layer. The second sidewall is located on the side of the memory cell closest to the erase structure. In the stacking direction of the memory cell, the second sidewall covers the control gate and the insulating oxide layer.

[0009] In some embodiments, the composite dielectric layer includes: a first oxide layer, a nitride layer, and a second oxide layer sequentially stacked; The first oxide layer is located on the surface of the floating gate away from the substrate layer.

[0010] In some embodiments, the control gate includes a first portion located on the surface of the composite dielectric layer and a second portion located on the first portion.

[0011] In some embodiments, the non-volatile memory further includes: The first metal lead is connected to the first drain or the second drain; The second metal lead connects to the control gate; The third metal lead connects to the erase gate; The fourth metal lead is connected to the source.

[0012] According to a second aspect of this application, a method for forming a non-volatile memory is provided, the method comprising: An initial stacked structure is provided, comprising a substrate layer, a first silicon oxide layer, a first polysilicon layer, a composite material layer, a second polysilicon layer, a conductive material layer, and a second silicon oxide layer stacked sequentially. Etching a second silicon oxide layer, a conductive material layer, a second polysilicon layer, a composite material layer, a first polysilicon layer, and a first silicon oxide layer forms phase-separated memory cells; Ion implantation is performed on the substrate layer between two adjacent memory cells to form a source. An erase structure is formed between two adjacent memory cells. The erase structure includes a first tunneling layer, a second tunneling layer, and an erase gate. Ion implantation is performed on the substrate layer on the side of the memory cell away from the erase structure to form a first drain and a second drain.

[0013] In some embodiments, etching a second silicon oxide layer, a conductive material layer, a second polysilicon layer, a composite material layer, a first polysilicon layer, and a first silicon oxide layer to form phase-spaced memory cells includes: Based on the first mask, a second silicon oxide layer, a conductive material layer and a second polysilicon layer are etched to form an insulating oxide layer and a control gate. The first mask has a first masking pattern corresponding to the memory cell. Based on the second mask etching of the composite material layer and the first polysilicon layer, an intermediate stacked structure is formed. The first mask has a second masking pattern corresponding to the memory cell and the area between the memory cells. A first silicon oxide material layer is deposited, which covers the first silicon oxide layer, the side of the insulating oxide layer away from the substrate layer, the sidewall of the insulating oxide layer, the sidewall of the control gate, the sidewall of the intermediate stack structure, and the portion of the intermediate stack structure exposed between two adjacent memory cells. A silicon nitride material layer is deposited, which is located on the side of the first silicon oxide material layer away from the substrate and covers the silicon oxide material layer; Etch a silicon oxide material layer and a silicon nitride material layer to form a first sidewall and a second sidewall; The intermediate stack structure exposed between two adjacent memory cells is etched to form a floating gate and a composite dielectric layer, with the first silicon oxide layer serving as a tunneling oxide layer.

[0014] In some embodiments, an erase structure is formed between two adjacent memory cells, including: A second silicon oxide material layer is deposited, which covers the substrate layer, the sidewall of the memory cell, and the side of the memory cell away from the substrate layer. Etching removes part of the second silicon oxide material layer, leaving the second silicon oxide material layer on the sidewall between two adjacent memory cells to form a first tunneling layer and a second tunneling layer; A polysilicon material layer is deposited, which fills the area between two memory cells to form an erase gate.

[0015] In some embodiments, the method further includes: A first metal lead, a second metal lead, a third metal lead, and a fourth metal lead are formed. The first metal lead is connected to the first drain or the second drain, the second metal lead is connected to the control gate, the third metal lead is connected to the erase gate, and the fourth metal lead is connected to the source.

[0016] Compared with related technologies, the technical solution provided in this application has the following beneficial effects: The non-volatile memory provided in this application has an erase structure including an erase gate, a first tunneling layer, and a second tunneling layer. The erase structure is positioned between two adjacent memory cells. During data erasure, an erase voltage is applied to the erase gate, allowing electrons remaining in the floating gate to tunnel through the first or second tunneling layer and enter the erase gate via quantum tunneling. Compared to related technologies, the non-volatile memory disclosed in this application eliminates the need for electrons to pass through the tunneling oxide layer to return to the substrate from the floating gate during data erasure. This reduces the number of tunneling passes through the tunneling oxide layer, thereby improving its reliability and ultimately enhancing the reliability of the non-volatile memory. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the direction of electron migration in related technologies; Figure 2 This is a schematic diagram of the structure corresponding to the non-volatile memory provided in the embodiments of this application; Figure 3 A top view of the structure corresponding to the non-volatile memory provided in the embodiments of this application; Figure 4 This is a flowchart illustrating the memory formation method provided in the embodiments of this application. Figure 5 This is a schematic diagram of the structure corresponding to step S1 provided in the embodiments of this application; Figure 6 Schematic diagram of the structure corresponding to step S21 provided in the embodiments of this application Figure 1 ; Figure 7 Schematic diagram of the structure corresponding to step S21 provided in the embodiments of this application Figure 2 ; Figure 8 Schematic diagram of the structure corresponding to step S22 provided in the embodiments of this application Figure 1 ; Figure 9 Schematic diagram of the structure corresponding to step S22 provided in the embodiments of this application Figure 2 ; Figure 10 This is a schematic diagram of the structure corresponding to step S25 provided in the embodiments of this application; Figure 11 This is a schematic diagram of the structure corresponding to step S3 provided in the embodiments of this application; Figure 12 Schematic diagram of the structure corresponding to step S4 provided in the embodiments of this application Figure 1 ; Figure 13 Schematic diagram of the structure corresponding to step S4 provided in the embodiments of this application Figure 2 ; Figure 14This is a schematic diagram of the structure corresponding to step S5 provided in the embodiments of this application; Figure 15 This is a schematic diagram of the structure corresponding to the formation of metal leads provided in the embodiments of this application.

[0018] Explanation of reference numerals in the attached figures: 1-Initial stack structure, 2-Intermediate stack structure, 3-Word line, 4-Bit line; 10 - Substrate, 11 - First drain, 12 - Second drain, 13 - Source; 20-Memory cell, 21-Tunnel oxide layer, 22-Floating gate, 23-Composite dielectric layer, 24-Control gate, 25-First part, 26-Second part, 27-Insulating oxide layer, 28-First sidewall, 29-Second sidewall; 30 - Erasure structure, 31 - First tunneling layer, 32 - Second tunneling layer, 33 - Erasure gate; 40 - First metal lead, 41 - Second metal lead, 42 - Third metal lead; 50 - First oxide layer, 51 - Second oxide layer, 52 - Nitride layer; 60 - First mask, 61 - First occlusion pattern, 62 - Sacrificial layer, 63 - Second mask, 64 - Second occlusion pattern; 70-First silicon oxide layer, 71-First polycrystalline silicon layer, 72-Composite material layer, 73-Second polycrystalline silicon layer, 74-Conductive material layer, 75-Second silicon oxide layer, 76-First silicon oxide material layer, 77-Silicon nitride material layer, 78-Second silicon oxide material layer, 79-Polycrystalline silicon material layer. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, such that a process, method, system, product, or device that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0020] In related technologies, the direction of electron migration during the data erasure and data programming processes of stacked gate or non-form flash memory is as follows: Figure 1 As shown, during data programming, electrons travel from the substrate beneath the tunneling oxide layer through the tunneling oxide layer to the floating gate. During data erasure, electrons travel from the floating gate above the tunneling oxide layer through the tunneling oxide layer to the substrate. Therefore, both data erasure and data programming require multiple passes through the tunneling oxide layer. If the quality of the tunneling oxide layer is poor, it may degrade after multiple passes, thus affecting the reliability of the memory.

[0021] In view of this, this application proposes a non-volatile memory and a method for forming the same. The following description will be provided in conjunction with the accompanying drawings.

[0022] like Figure 2 , Figure 10 , Figure 11 as well as Figure 13 As shown, a non-volatile memory may include a substrate 10, a first drain 11, a source 13, a second drain 12, a memory cell 20, and an erase structure 30.

[0023] In some embodiments, the first drain 11, the source 13, and the second drain 12 are located in the substrate layer 10. The first drain 11 and the source 13, or the second drain 12 and the source 13, form the two ends of a conductive channel, which can cooperate with the gates included in the non-volatile memory to complete operations such as data reading, data programming, and data erasure.

[0024] In one embodiment, the memory cell 20 is located on the surface of the substrate 10, and the projection of the memory cell 20 onto the substrate 10 is located between the first drain 11 and the source 13, and / or between the source 13 and the second drain 12.

[0025] For details, please refer to Figure 2 The projection direction of the memory cell 20 is perpendicular to the surface of the substrate layer 10. This positioning of the memory cell 20 allows each memory cell 20 in the non-volatile memory to correspond to a drain, and every two memory cells 20 to share a source 13, thereby enabling more memory cells 20 to be made on the same wafer area, thus reducing manufacturing costs.

[0026] In this embodiment, the memory cell 20 is located on the side surface of the substrate layer 10 where the first drain 11 is disposed. Figure 2 as well as Figure 11 As shown, the memory cell 20 may include a tunnel oxide layer 21, a floating gate 22, a composite dielectric layer 23, a control gate 24, and an insulating oxide layer 27.

[0027] In this embodiment, the tunneling oxide layer 21 is located on the side surface of the substrate layer 10 where the first drain electrode 11 is disposed. The substrate layer 10 may be a silicon substrate, and the material forming the tunneling oxide layer 21 may be silicon dioxide.

[0028] Under normal conditions, the tunnel oxide layer 21 isolates the substrate layer 10 and the floating gate 22, thereby preventing charge leakage. During data programming, hot electrons in the channel of the substrate layer 10 pass through the tunnel oxide layer 21 into the floating gate 22.

[0029] In this embodiment, the floating gate 22 is located on the surface of the tunnel oxide layer 21 away from the substrate layer 10. The floating gate 22 can store or release charge, thereby enabling data programming and data erasure. The material forming the floating gate 22 can be polycrystalline silicon.

[0030] In this embodiment, the composite dielectric layer 23 is located on the side surface of the floating gate 22 away from the substrate layer 10. The composite dielectric layer 23 may include a first oxide layer 50, a nitride layer 52 and a second oxide layer 51 stacked sequentially, with the first oxide layer 50 located on the side surface of the floating gate 22 away from the substrate layer 10.

[0031] In one implementation, the composite dielectric layer 23 can be an ONO (Oxide-Nitride-Oxide) three-layer composite dielectric, that is, the first oxide layer 50 and the second oxide layer 51 can be formed of silicon oxide, and the nitride layer 52 can be formed of silicon nitride. By introducing the composite dielectric layer 23 to achieve inter-gate isolation between the floating gate 22 and the control gate 24, the capacitive coupling efficiency between the control gate 24 and the floating gate 22 can be improved, and the upward migration of electrons stored in the floating gate 22 to the control gate 24 can be restricted.

[0032] In this embodiment, the control gate 24 is located on the surface of the composite dielectric layer 23 away from the substrate layer 10. For example... Figure 2 as well as Figure 11 As shown, the control gate 24 may include a first portion 26 located on the surface of the composite dielectric layer 23 and a second portion 25 located on the first portion 26. The material forming the first portion 26 may be polycrystalline silicon; the material forming the second portion 25 may be a metallic conductive material, such as tungsten silicide.

[0033] In this embodiment, the first part 26 can serve as the base conductive layer of the control gate 24, forming capacitive coupling with the floating gate 22. The injection and release of charge in the floating gate 22 are controlled by applying voltage. The second part 25 reduces resistance, enabling the memory cell 20 with the control gate 24 to operate quickly and stably.

[0034] In this embodiment, the insulating oxide layer 27 can be located on the surface of the control gate 24 away from the substrate layer 10. The insulating oxide layer 27 can be a silicon oxide layer. As the outermost insulating structure of the memory cell 20 as a whole, the insulating oxide layer 27 can isolate the control gate 24 from other memory structures not shown in the figure, such as upper contact plugs, to prevent electrode short circuits. At the same time, the insulating oxide layer 27 can smooth the wafer surface and prevent moisture and impurity ions from penetrating the internal gate structure and dielectric layer, delaying device aging and improving reliability.

[0035] In some embodiments, the storage unit 20 may further include: The first sidewall 28 is located on the side of the storage cell 20 away from the erase structure 30. In the stacking direction of the storage cell 20, the first sidewall 28 covers the floating gate 22, the composite dielectric layer 23, the control gate 24 and the insulating oxide layer 27, that is, the first sidewall 28 covers the sidewall surface of the floating gate 22, the composite dielectric layer 23, the control gate 24 and the insulating oxide layer 27.

[0036] The second sidewall 29 is located on the side of the storage cell 20 near the erase structure 30. In the stacking direction of the storage cell 20, the second sidewall 29 covers the control gate 24 and the insulating oxide layer 27, that is, the second sidewall 29 covers the sidewall surface of the control gate 24 and the insulating oxide layer 27.

[0037] Please continue reading. Figure 2 as well as Figure 10 In a direction parallel to the surface of the substrate layer 10, that is, in the extension direction of each layer of the memory cell 20, the first sidewall 28 is located on the side of the memory cell 20 away from the erase structure 30. Figure 10 as well as Figure 11 As shown, in the arrangement direction of the plurality of memory cells 20, the projection of the first sidewall 28 toward the composite dielectric layer 23, the control gate 24, the floating gate 22, and the insulating dielectric layer 27 can cover the side surface of the composite dielectric layer 23, the side surface of the floating gate 22, the side surface of the insulating oxide layer 27, and the side surface of the control gate 24. The side surface of the floating gate 22 is flush with the side surface of the composite dielectric layer 23, the side surface of the composite dielectric layer 23 is flush with the side surface of the control gate 24, and the side surface of the control gate 24 is flush with the side surface of the insulating oxide layer 27. The side surfaces of the floating gate 22, the composite dielectric layer 23, the control gate 24, and the insulating oxide layer 27 are all parallel to a standard plane. The standard plane can be a plane defined by the stacking direction of the insulating oxide layer 27 and the control gate 24, and the arrangement direction of the plurality of memory cells 20.

[0038] The first sidewall 28 isolates the memory cell 20 from other structures by covering the sidewalls of the floating gate 22, the composite dielectric layer 23, the control gate 24, and the insulating oxide layer 27. These other structures may be the metal leads mentioned below.

[0039] In a direction parallel to the surface of the substrate layer 10, that is, in the extending direction of each layer of the memory cell 20, the second sidewall 29 is located on the side of the memory cell 20 closer to the erase structure 30. For example... Figure 10 As shown, in the arrangement direction of the plurality of memory cells 20, the projection of the second sidewall 29 toward the insulating oxide layer 27 and the control gate 24 can cover the side surface of the insulating oxide layer 27 and the side surface of the control gate 24. The side surfaces of the insulating oxide layer 27 and the control gate 24 are flush and parallel to a standard plane. By covering the sidewalls of the control gate 24 and the insulating oxide layer 27, the second sidewall 29 achieves isolation between the memory cells 20 and the erase structure 30.

[0040] As one implementation, the material forming the first sidewall 28 can be silicon nitride and silicon oxide, and the material forming the second sidewall 29 can also be silicon nitride and silicon oxide. (See also...) Figure 10 as well as Figure 11It is important to note that the silicon nitride and silicon oxide here are not mutually doped, but rather layered. For example, the second sidewall 29 can include two layers: one formed of silicon nitride and the other of silicon oxide. Since the materials used to form the second sidewall 29 are the same as those used for the second oxide layer 51 and the nitride layer 52, and since electrons need to migrate from the floating gate 22 to the erase structure 30, the second sidewall 29 can cover the sidewalls of the control gate 24 and the insulating oxide layer 27, and does not need to extend further toward the substrate 10 in a direction perpendicular to the surface of the substrate layer 10.

[0041] Please continue reading. Figure 2 as well as Figure 13 The erase structure 30 is located on the surface of the substrate layer 10 and between two adjacent memory cells 20. The erase structure 30 may include a first tunneling layer 31, a second tunneling layer 32 and an erase gate 33, with the erase gate 33 located between the first tunneling layer 31 and the second tunneling layer 32.

[0042] The erase structure 30 is configured to receive an erase voltage when a data erase operation is performed on the memory cell 20, so that the charge in the memory cell 20 moves to the erase gate 33 via the first tunneling layer 31 or the second tunneling layer 32.

[0043] In one implementation, the erase structure 30 can be positioned between two adjacent memory cells 20, with each pair of memory cells 20 sharing one erase structure 30. The erase structure 30 includes a first tunneling layer 31, a second tunneling layer 32, and an erase gate 33. Each memory cell 20 corresponds to one tunneling layer; that is, the first tunneling layer 31 and the second tunneling layer 32 can each correspond to one memory cell 20, and two adjacent memory cells 20 share one erase gate 33. This structural arrangement reduces space occupation and manufacturing costs.

[0044] In some embodiments, the material forming the first tunneling layer 31 may be silicon dioxide, the material forming the second tunneling layer 32 may also be silicon dioxide, and the material forming the erase gate 33 may be polysilicon.

[0045] In one implementation, the erase structure 30 is configured to receive an erase voltage when a data erase operation is performed on the memory cell 20, so that the charge in the memory cell 20 moves to the erase gate 33 via the first tunneling layer 31 or the second tunneling layer 32. The erase voltage can be a positive voltage, for example, an 11V value.

[0046] Specifically, when performing data storage operations, after the erase gate 33 receives the erase voltage, the electrons stored in the floating gates 22 of the two corresponding storage cells 20 can migrate simultaneously toward the first tunneling layer 31 and the second tunneling layer 32, respectively. Alternatively, it can work with the control gate 24 to selectively erase a single storage cell 20, so that the electrons stored in the floating gate 22 of one of the erase cells migrate toward the first tunneling layer 31 or the second tunneling layer 32, thereby improving the erasure efficiency and adapting to different erasure scenario requirements.

[0047] The non-volatile memory disclosed in this application, when performing a data erase operation on the memory cell 20, eliminates the need to apply a negative high voltage to the control gate 24 and a positive high voltage to the substrate layer 10, compared to related technologies. Instead, a positive high voltage is applied only to the erase gate 33, thereby simplifying the corresponding peripheral circuit design. Furthermore, since no high voltage needs to be applied to the substrate layer 10, the inclusion of a deep N-well in the conventional stacked gate structure of the substrate layer 10 becomes optional. This optional design simplifies the setup of the substrate layer 10 and saves on process costs.

[0048] Meanwhile, in this embodiment, data programming requires passing through the tunneling oxide layer 21, while data erasure does not. Compared to related technologies where both data erasure and data programming require passing through the tunneling oxide layer 21, reducing the number of tunneling operations by half reduces damage to the tunneling oxide layer 21 caused by electron tunneling, increases the number of data programming and / or erasure operations, enhances the reliability of the tunneling oxide layer 21, and thus enhances the reliability and extends the lifespan of the non-volatile memory.

[0049] In addition, the erase grid 33 and the floating grid 22 are made of the same material, which can erase sharp corners during the data erasure process. Since the electric field strength is high at the sharp corners, the data erasure efficiency can be improved compared with related technologies.

[0050] As one implementation, the non-volatile memory may also include multiple metal leads. For example... Figure 2 As shown, the first metal lead 40 is connected to the first drain 11 or the second drain 12; the second metal lead 41 is connected to the control gate 24; and the third metal lead 42 is connected to the erase gate 33.

[0051] Specifically, although not shown in the figure, it can be understood that the non-volatile memory may also include a metal layer, with one end of multiple metal leads connected to the metal layer and the other end connected to the first drain 11, the second drain 12, the control gate 24, and the erase gate 33, respectively.

[0052] Furthermore, although not shown in the figure, it is understood that the non-volatile memory may include at least a fourth metal lead in addition to the first metal lead 40, the second metal lead 41, and the third metal lead 42. One end of the fourth metal lead is connected to the source 13, and the other end is connected to the metal layer.

[0053] By using multiple metal leads in conjunction with a metal layer, electrical signals can be transmitted between the storage unit 20 and the upper wiring.

[0054] In some embodiments, the non-volatile memory may further include multiple word lines 3 and multiple bit lines 4. For example... Figure 3 As shown, in a non-volatile memory, multiple memory cells 20 are arranged in an array to form a memory array. Word lines 3 can be electrically connected to control gates 24, and the control gates 24 of each memory cell 20 in the same row are interconnected to form word lines 3. Bit lines 4 can be electrically connected to the first drain 11 and / or the second drain 12, and the drains of each memory cell 20 in the same column are interconnected to form bit lines 4.

[0055] In some embodiments, the non-volatile memory may further include peripheral circuitry coupled to the memory array. This peripheral circuitry is configured to perform a data erasure operation, such that electrons, or charges, in the floating gate 22 migrate to the erase gate 33 via the first tunneling layer 31 and / or the second tunneling layer 32.

[0056] Correspondingly, this application also discloses a method for forming non-volatile memory, applicable to the non-volatile memory in any of the above embodiments, such as... Figure 4 As shown, the method includes: Step S1: Provide an initial stacked structure 1, which includes a substrate layer 10, a first silicon oxide layer 70, a first polysilicon layer 71, a composite material layer 72, a second polysilicon layer 73, a conductive material layer 74, and a second silicon oxide layer 75 stacked sequentially.

[0057] In one embodiment, the composite material layer 72 can be a three-layer structure, wherein the three layers are respectively made of silicon oxide, silicon nitride, and silicon oxide. For example... Figure 5As shown, a substrate layer 10, a first silicon oxide layer 70, a first polysilicon layer 71, a composite material layer 72, a second polysilicon layer 73, a conductive material layer 74, and a second silicon oxide layer 75 can be deposited sequentially to obtain an initial stacked structure 1. The substrate layer 10 can be a silicon substrate; the first silicon oxide layer 70 can be formed of silicon oxide material and is used to form a tunnel oxide layer 21 after subsequent etching or other processes; the first polysilicon layer 71 can be formed of polysilicon material and is used to form a floating gate 22 after subsequent etching or other processes; the composite material layer 72 is used to form a composite dielectric layer 23 after subsequent etching or other processes; the second polysilicon layer 73 can be formed of polysilicon material, and the conductive material layer 74 can be formed of tungsten silicide or a tungsten-silicon alloy. The second polysilicon layer 73 and the conductive material layer 74 are used to form a control gate 24 after subsequent etching or other processes; the second silicon oxide layer 75 can be formed of silicon oxide material and is used to form an insulating oxide layer 27 after subsequent etching or other processes.

[0058] Step S2: Etch the second silicon oxide layer 75, the conductive material layer 74, the second polysilicon layer 73, the composite material layer 72, the first polysilicon layer 71, and the first silicon oxide layer 70 to form phase-spaced memory cells 20.

[0059] In some embodiments, step S2 may include: Step S21: Based on the first mask 60, the second silicon oxide layer 75, the conductive material layer 74 and the second polysilicon layer 73 are etched to form the insulating oxide layer 27 and the control gate 24. The first mask 60 has a first masking pattern 61 corresponding to the memory cell 20.

[0060] As one implementation method, such as Figure 6 As shown, a first mask 60 is first formed on the side of the second silicon oxide layer 75 away from the substrate layer 10. The first mask 60 may include a plurality of first masking patterns 61. The first mask 60 may be formed of photoresist, and the plurality of first masking patterns 61 are used to define the region where the memory cell 20 is located on the initial stacked structure 1.

[0061] Furthermore, dry etching can be used to etch the second silicon oxide layer 75, the conductive material layer 74, and the second polysilicon layer 73 using multiple patterned blocks as masks, and the composite material layer 72 can be used as a stop layer to obtain the desired result. Figure 7 The insulating oxide layer 27 and control gate 24 are shown.

[0062] As one implementation method, after etching, the photoresist can be removed using an ASH (Ashing / Plasma Dry Resist Removal) process, and the photoresist can be further removed. Figure 7 The structure shown is then cleaned.

[0063] Step S22: Based on the second mask 63, the composite material layer 72 and the first polysilicon layer 71 are etched to form an intermediate stacked structure 2. The second mask has a second masking pattern 64 corresponding to the memory cell 20 and the area between the memory cells 20.

[0064] As one implementation method, such as Figure 8 As shown, before etching the composite material layer 72 and the first polysilicon layer 71, a sacrificial layer 62 can be deposited first. The sacrificial layer 62 covers the surface of the composite material layer 72 and the surface of the insulating dielectric layer. A photoresist layer is formed on the side of the sacrificial layer 62 away from the substrate layer 10. A preset pattern is transferred to the photoresist layer using a photomask as a mask to form a second mask 63. The second mask 63 includes at least one second masking pattern 64. The sacrificial layer 62 can be a SOC (Spin-On Carbon).

[0065] Furthermore, using the second masking pattern 64 as a mask, the sacrificial layer 62, the composite material layer 72, and the first polysilicon layer 71 can be dry-etched to remove residual photoresist and the etched sacrificial layer 62, forming a structure as shown below. Figure 9 The intermediate stacking structure 2 is shown.

[0066] Step S23: Deposit a first silicon oxide material layer 76. The first silicon oxide material layer 76 covers the first silicon oxide layer 70, the side surface of the insulating oxide layer 27 away from the substrate layer 10, and covers the sidewalls of the insulating oxide layer 27, the sidewalls of the control gate 24, and the sidewalls of the intermediate stack structure 2. The first silicon oxide material layer 76 also covers the surface of the intermediate stack structure 2 exposed between two adjacent memory cells 20.

[0067] Step S24: Deposit silicon nitride material layer 77, which is located on the side of the first silicon oxide material layer 76 away from the substrate and covers the first silicon oxide material layer.

[0068] In this embodiment, the silicon nitride material layer 77 conformally covers the first silicon oxide material layer.

[0069] Step S25: Etch the first silicon oxide material layer and the silicon nitride material layer 77 to form the first sidewall 28 and the second sidewall 29.

[0070] In some embodiments, a first silicon oxide material layer 76 is deposited, followed by a silicon nitride material layer 77. Then, a dry etching technique is used to remove the first silicon oxide material layer 76 and the silicon nitride material layer 77 from the surface of the first silicon oxide layer 70, to remove the first silicon oxide material layer 76 and the silicon nitride material layer 77 on the side of the insulating oxide layer 27 away from the substrate layer 10, and to remove the portion of the first silicon oxide material layer 76 and the silicon nitride material layer 77 exposed on the surface of the intermediate stack structure 2 between two adjacent memory cells 20, retaining the first silicon oxide material layer 76 and the silicon nitride material layer 77 covering the sidewalls, forming a first sidewall 28 and a second sidewall 29, resulting in... Figure 10 The structure shown.

[0071] Step S26: Etch the exposed intermediate stack structure 2 between two adjacent memory cells 20 to form a floating gate 22, a composite dielectric layer 23, and use the first silicon oxide layer 70 as a tunnel oxide layer 21.

[0072] In one implementation, the intermediate stacked structure 2 exposed between two adjacent memory cells 20 is etched to form a floating gate 22 and a composite dielectric layer 23, and the first silicon oxide layer 70 is directly used as a tunneling oxide layer 21 to obtain the memory cell 20. Specifically, for a memory cell 20, its tunneling oxide layer 21 is a part of the first silicon oxide layer 70, such as... Figure 11 The middle dashed box is used as an indicator.

[0073] Step S3: Ion implantation is performed on the substrate layer 10 between two adjacent memory cells 20 to form a source 13.

[0074] In one implementation, a source region is first defined. This source region can be disposed on the substrate layer 10 and can be located between two adjacent memory cells 20. Ion implantation, for example, of arsenic is performed into the source region to form a source 13, resulting in... Figure 11 The structure shown. The source 13 serves as a shared source for its two adjacent memory cells 20.

[0075] Step S4: An erase structure 30 is formed between two adjacent memory cells 20. The erase structure 30 includes a first tunneling layer 31, a second tunneling layer 32, and an erase gate 33.

[0076] In one implementation, step S4 may include: A second silicon oxide material layer 78 is deposited, which covers the substrate layer 10, the sidewall of the memory cell 20, and the side of the memory cell 20 away from the substrate layer 10.

[0077] Part of the second silicon oxide material layer 78 is etched away, while the second silicon oxide material layer 78 of the sidewall between two adjacent memory cells 20 is retained to form the first tunneling layer 31 and the second tunneling layer 32.

[0078] A polysilicon material layer 79 is deposited, which fills the area between two memory cells 20 to form an erase gate 33.

[0079] In a specific embodiment, such as Figure 12 As shown, a second silicon oxide material layer 78 is first deposited on the side of the memory cell 20 away from the substrate 10, on the sidewall of the memory cell 20, and on the surface of the substrate 10. The second silicon oxide material layer 78 is then etched to remove the second silicon oxide material layer 78 on the surface of the substrate 10, the second silicon oxide material layer 78 on the side of the memory cell 20 away from the substrate 10, and the second silicon oxide material layer 78 covering the first sidewall 28, leaving the second silicon oxide material layer 78 covering the other side of the memory cell 20, thus forming a first tunneling layer 31 and a second tunneling layer 32. The other side of the memory cell 20 is the surface where the second sidewall 29 is located.

[0080] Further, please refer to Figure 12 and Figure 13 After forming the first tunneling layer 31 and the second tunneling layer 32, a polysilicon material layer 79 is deposited. The polysilicon material layer 79 can also be treated using CMP (Chemical Mechanical Polishing) to maintain the surface flatness of the structure. The polysilicon material layer 79 is then etched, retaining the polysilicon material layer 79 located between two adjacent memory cells 20 to form an erase gate 33. The surface of the erase gate 33 away from the substrate layer 10 is flush with the surface of the insulating oxide layer 27 in the memory cell 20 away from the substrate layer 10. The polysilicon material layer 79 is then etched. Step S5: Ion implantation is performed on the substrate layer 10 located on the side of the memory cell 20 away from the erase structure 30 to form a first drain 11 and a second drain 12.

[0081] In one implementation, a drain region is first defined. The drain region is located on the substrate layer 10 and on the side of the memory cell 20 away from the source 13. Ion implantation, for example, implantation of arsenic or phosphorus elements, is performed into the drain region to form a first drain 11 and a second drain 12, resulting in... Figure 14 The structure shown.

[0082] As another implementation, to improve the reliability of the first sidewall 28, a layer of silicon nitride material can be deposited before defining the drain region. This silicon nitride material can cover the substrate layer 10, the first sidewall 28, the surface of the insulating dielectric layer away from the substrate layer 10, and the surface of the erase structure 30 away from the substrate layer 10. Then, the silicon nitride material covering the substrate layer 10, the silicon nitride material covering the surface of the insulating dielectric layer away from the substrate layer 10, and the silicon nitride material covering the surface of the erase structure 30 away from the substrate layer 10 are removed, leaving the silicon nitride material covering the first sidewall 28, forming a first sidewall 28 with a larger size in a direction parallel to the surface of the substrate layer 10. Figure 14 As shown, the first sidewall 28 may include a three-layer structure arranged in layers.

[0083] In some embodiments, the method further includes: A first metal lead 40, a second metal lead 41, a third metal lead 42, and a fourth metal lead are formed. The first metal lead 40 can be connected to the first drain 11 and / or the second drain 12. The second metal lead 41 is connected to the control gate 24. The third metal lead 42 is connected to the erase gate 33, and the fourth metal lead is connected to the source.

[0084] Specifically, the multiple metal leads may include at least a first metal lead 40, a second metal lead 41, a third metal lead 42, and a fourth metal lead, to connect the first drain 11, the second drain 12, the source 13, the control gate 24, and the erase gate 33 to the metal layer.

[0085] To prevent the tunneling oxide layer 21 from hindering current conduction between the metal leads and the drain, thus increasing contact resistance, the portion of the tunneling oxide layer 21 above the first drain 11 and the portion of the tunneling oxide layer 21 above the second drain 11 need to be removed before forming the metal leads. This allows for the formation of multi-metal leads to achieve the desired result. Figure 15 , Figure 2 The non-volatile memory shown.

[0086] As another implementation, but in order to further reduce the contact resistance and avoid direct contact between the first drain 11, the second drain 12 and the first metal lead 40 to form an ohmic contact, a metal silicide layer can be formed above the first drain 11 and the second drain 12. That is, although not shown in the figures, a metal silicide layer can also be provided between the first drain 11 and the first metal lead 40, and between the second drain 12 and the first metal lead 40.

[0087] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A non-volatile memory, characterized in that, include: Substrate layer; The first drain, the source, and the second drain are located in the substrate layer; A memory cell is located on the surface of the substrate layer, and the projection of the memory cell onto the substrate layer is located between the source and the first drain and / or between the source and the second drain. An erase structure is located on the surface of the substrate layer and between two adjacent memory cells. The erase structure includes a first tunneling layer, a second tunneling layer, and an erase gate, with the erase gate located between the first tunneling layer and the second tunneling layer. The erase structure is configured to receive an erase voltage when a data erase operation is performed on the memory cell, so that the charge in the memory cell moves to the erase gate via the first tunneling layer or the second tunneling layer.

2. The non-volatile memory according to claim 1, characterized in that, The storage unit includes: The oxide layer is then penetrated and located on the surface of the substrate layer; A floating gate is located on the surface of the tunneling oxide layer away from the substrate layer; A composite dielectric layer is located on the surface of the floating gate away from the substrate layer; The control gate is located on the surface of the composite dielectric layer away from the substrate layer; An insulating oxide layer is located on the surface of the control gate away from the substrate layer.

3. The non-volatile memory according to claim 2, characterized in that, The storage unit further includes: A first sidewall is located on the side of the storage cell away from the erase structure. In the stacking direction of the storage cell, the first sidewall covers the floating gate, the composite dielectric layer, the control gate, and the insulating oxide layer. The second sidewall is located on the side of the storage cell closest to the erase structure, and in the stacking direction of the storage cell, the second sidewall covers the control gate and the insulating oxide layer.

4. The non-volatile memory according to claim 2, characterized in that, The composite dielectric layer includes: a first oxide layer, a nitride layer, and a second oxide layer stacked sequentially; The first oxide layer is located on the surface of the floating gate away from the substrate layer.

5. The non-volatile memory according to claim 2, characterized in that, The control gate includes a first portion located on the surface of the composite dielectric layer and a second portion located on the first portion.

6. The non-volatile memory according to claim 1, characterized in that, The non-volatile memory also includes: A first metal lead is connected to either the first drain or the second drain. A second metal lead is connected to the control gate; A third metal lead is connected to the erase gate; A fourth metal lead is connected to the source electrode.

7. A method for forming a non-volatile memory, characterized in that, include: An initial stacked structure is provided, the initial stacked structure comprising a substrate layer, a first silicon oxide layer, a first polysilicon layer, a composite material layer, a second polysilicon layer, a conductive material layer, and a second silicon oxide layer stacked sequentially; The second silicon oxide layer, the conductive material layer, the second polysilicon layer, the composite material layer, the first polysilicon layer, and the first silicon oxide layer are etched to form phase-spaced memory cells; Ion implantation is performed on the substrate layer between two adjacent memory cells to form a source. An erase structure is formed between two adjacent memory cells, the erase structure including a first tunneling layer, a second tunneling layer and an erase gate; Ion implantation is performed on the substrate layer on the side of the memory cell away from the erase structure to form a first drain and a second drain.

8. The method for forming a non-volatile memory according to claim 7, characterized in that, Etching the second silicon oxide layer, the conductive material layer, the second polysilicon layer, the composite material layer, the first polysilicon layer, and the first silicon oxide layer to form phase-spaced memory cells includes: Based on the first mask, the second silicon oxide layer, the conductive material layer and the second polysilicon layer are etched to form an insulating oxide layer and a control gate. The first mask has a first masking pattern corresponding to the memory cell. The composite material layer and the first polysilicon layer are etched based on the second mask to form an intermediate stacked structure. The second mask has a second masking pattern corresponding to the memory cell and the area between the memory cell. A first silicon oxide material layer is deposited, which covers the first silicon oxide layer, the side of the insulating oxide layer away from the substrate layer, the sidewall of the insulating oxide layer, the sidewall of the control gate, the sidewall of the intermediate stack structure, and the portion of the intermediate stack structure exposed between two adjacent memory cells. A silicon nitride material layer is deposited, the silicon nitride material layer being located on the side of the first silicon oxide material layer away from the substrate, and covering the silicon oxide material layer; The silicon oxide material layer and the silicon nitride material layer are etched to form a first sidewall and a second sidewall; The intermediate stack structure exposed between two adjacent memory cells is etched to form a floating gate, a composite dielectric layer, and the first silicon oxide layer is used as a tunneling oxide layer.

9. The method for forming a non-volatile memory according to claim 7, characterized in that, The process of forming an erase structure between two adjacent memory cells includes: A second silicon oxide material layer is deposited, which covers the substrate layer, the sidewall of the memory cell, and the side of the memory cell away from the substrate layer. Etching removes part of the second silicon oxide material layer, leaving the second silicon oxide material layer on the sidewall between two adjacent memory cells to form the first tunneling layer and the second tunneling layer; A polysilicon material layer is deposited, which fills the area between two memory cells to form an erase gate.

10. The method for forming a non-volatile memory according to claim 7, characterized in that, The method further includes: A first metal lead, a second metal lead, a third metal lead, and a fourth metal lead are formed. The first metal lead is connected to the first drain or the second drain, the second metal lead is connected to the control gate, the third metal lead is connected to the erase gate, and the fourth metal lead is connected to the source.