Method for forming a resistive random access memory (RRAM) cell

The method of forming insulating spacers on the sidewalls of RRAM cell electrodes and dielectric layers simplifies the manufacturing process, improving yield and performance by reducing masking steps and preventing short-circuits and damage, thus enhancing the reliability of RRAM cells.

KR1020260161601APending Publication Date: 2026-11-09SILICON STORAGE TECHNOLOGY INC
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
KR1020267033248
Authority / Receiving Office
KR · KR
Patent Type
Applications
Current Assignee / Owner
Priority Date
2024-08-13
Filing Date
2024-08-15
Publication Date
2026-11-09

AI Technical Summary

Technical Problem

Existing methods for fabricating resistive random access memory (RRAM) cells are complex and result in low yield and performance issues.

Method used

A method for forming RRAM cells involves forming insulating spacers on the sidewalls of electrode and resistive switching dielectric material layers, using a self-aligning mask to simplify the manufacturing process and reduce the number of masking steps, thereby preventing short-circuits and damage to the RRAM structure.

Benefits of technology

This approach enhances the yield and performance of RRAM cell fabrication by reducing the number of masking steps and protecting the electrode and dielectric layers, ensuring precise alignment and preventing damage during the manufacturing process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure PCT00010_ABST
    Figure PCT00010_ABST
Patent Text Reader

Abstract

A method for forming a semiconductor device comprises the steps of: forming an insulating material on a semiconductor substrate; forming a conductive contact penetrating the insulating material; forming a lower electrode layer on the insulating material and electrically connected to the conductive contact; forming a resistive switching dielectric material (RSDM) layer directly on the lower electrode layer; forming an upper electrode layer directly on the RSDM layer; selectively removing portions of the upper electrode layer to form a block of the upper electrode layer; selectively removing portions of the RSDM layer to form a block of the RSDM layer disposed below the block of the upper electrode layer; forming insulating spacers on the lower electrode layer and extending along the sidewalls of the block of the upper electrode layer and the block of the RSDM layer; and selectively removing portions of the lower electrode layer to form a block of the lower electrode layer.
Need to check novelty before this filing date? Find Prior Art

Description

Technology Field

[0001] Related applications

[0002] This application claims the benefit of U.S. Provisional Application No. 63 / 644,433 filed on May 8, 2024 and U.S. Patent Application No. 18 / 803,363 filed on May 13, 2024.

[0003] Technology field

[0004] The present invention relates to non-volatile memory, and more specifically, to resistive random access memory (RRAM). Background Technology

[0005] Resistive Random Access Memory (RRAM) is a type of non-volatile memory. Generally, RRAM memory cells each comprise a layer of resistive switching dielectric material interposed between two conductive electrodes. The resistive switching dielectric material is typically insulating. However, by applying an appropriate voltage across the resistive switching dielectric material layer, a conductive path (generally referred to as a filament) can be formed through the resistive switching dielectric material layer, resulting in a lower resistance across the RRAM cell. Once the filament is formed, it can be "reset" (i.e., broken or ruptured—causing high resistance across the RRAM cell) and "set" (i.e., reformed—causing lower resistance across the RRAM cell again) by applying appropriate voltages across the resistive switching dielectric material layer. Low and high resistance states are utilized to indicate a digital state of "1" or "0" depending on the resistance state, thereby providing a reprogrammable non-volatile memory cell capable of storing a certain number of bits of information.

[0006] FIG. 1 illustrates a conventional configuration of an RRAM memory cell (1). The memory cell (1) includes a resistive switching dielectric material layer (2) interposed between two conductive material layers forming an upper electrode (3) and a lower electrode (4), respectively.

[0007] FIGS. 2a through 2d illustrate the switching mechanism of a resistive switching dielectric material layer (2). Specifically, FIG. 2a illustrates the resistive switching dielectric material layer (2) in its initial state after manufacturing, where the layer (2) exhibits a relatively high resistance. FIG. 2b illustrates the formation of a conductive filament (7) through the layer (2) by applying an appropriate voltage across the layer (2). The filament (7) is a conductive path through the layer (2), and thus the layer exhibits a relatively low resistance across its ends (due to the relatively high conductivity of the filament (7)). FIG. 2c illustrates the formation of a rupture (8) in the filament (7) caused by applying a "reset" voltage across the layer (2). The region of the rupture (8) has a relatively high resistance, and thus the layer (2) exhibits a relatively high resistance across its ends. FIG. 2d illustrates the recovery of the filament (7) in the region of the rupture (8), caused by applying a "set" voltage across the ends of the layer (2). The recovered filament (7) means that the layer (2) exhibits a relatively low resistance across its ends. The relatively low resistance of the layer (2) in the "forming" or "set" states of FIG. 2b and FIG. 2d, respectively, can represent a digital state (e.g., "1"), and the relatively high resistance of the layer (2) in the "reset" state of FIG. 2c can represent a different digital state (e.g., "0"). The reset voltage (which breaks the filament) may have a polarity opposite to that of the filament forming and set voltages, but it may also have the same polarity. The RRAM cell (1) can be repeatedly "reset" and "set," and thus it forms a reprogrammable non-volatile memory cell.

[0008] An improved methodology for fabricating RRAM cells is required to simplify manufacturing and increase yield and performance.

[0009] The aforementioned problems and requirements are solved by a method for forming a semiconductor device, the method comprising the steps of: forming an insulating material on a semiconductor substrate; forming a first contact hole extending through the insulating material; forming a first conductive contact within the first contact hole; forming a lower electrode layer on the insulating material and electrically connected to the first conductive contact; forming a resistive switching dielectric material layer directly on the lower electrode layer; forming an upper electrode layer directly on the resistive switching dielectric material layer; selectively removing portions of the upper electrode layer to form a block of the upper electrode layer; selectively removing portions of the resistive switching dielectric material layer to form a block of the resistive switching dielectric material layer disposed below the block of the upper electrode layer; forming insulating spacers extending on the lower electrode layer and along the sidewalls of the block of the upper electrode layer and the block of the resistive switching dielectric material layer; and selectively removing portions of the lower electrode layer to form a block of the lower electrode layer, wherein the block of the lower electrode layer is disposed below the insulating spacers, the block of the resistive switching dielectric material layer, and the block of the upper electrode layer.

[0010] Other purposes and features of the present disclosure will become apparent from a review of the specification, claims and the accompanying drawings. Brief explanation of the drawing

[0011] Figure 1 is a side cross-sectional view of a conventional RRAM memory cell. FIG. 2a is a side cross-sectional view of a conventional RRAM memory cell in its initial state. FIG. 2b is a side cross-sectional view of a conventional RRAM memory cell illustrating the formation of a conductive filament. FIG. 2c is a side cross-sectional view of a conventional RRAM memory cell illustrating the formation of a rupture portion of a conductive filament. FIG. 2d is a side cross-sectional view of a conventional RRAM memory cell illustrating the recovery of a conductive filament in a ruptured region. FIGS. 3a to 3h are cross-sectional views illustrating the formation of an RRAM memory cell. Specific details for implementing the invention

[0012] A method for forming a memory device having RRAM memory cells is illustrated in FIGS. 3a to 3g and begins by forming the structure illustrated in FIG. 3a. Specifically, a drain region (12) and a source region (14) having a first conductive type (e.g., n+) are formed within a semiconductor substrate (10) (e.g., silicon) having a second conductive type (e.g., p+) by injection, for example, and these define a channel region (16) extending between the drain region (12) and the source region (14) within the semiconductor substrate (10). A conductive gate (18) (e.g., made of a conductive material such as polysilicon or metal) is formed on the channel region (16) of the semiconductor substrate (10) and is insulated therefrom. The formation of a conductive gate may include forming an oxide layer (20), which is an insulating material (e.g., silicon oxide, silicon dioxide, hafnium oxide, and combinations thereof) disposed between a semiconductor substrate (10) and a conductive gate (18), then depositing a polysilicon or metal gate on the oxide layer (20), and then performing a photolithography and etching process (e.g., photoresist deposition, exposure, and selective removal followed by poly etching) to selectively remove the polysilicon layer excluding the portion constituting the conductive gate (18). Subsequently, an insulating material (22) (e.g., oxide) is formed on the semiconductor substrate (10). Contact holes (24) exposing a drain region (12) and a source region (14) are formed in the insulating material (22) by the photolithography and etching process. The contact holes (24) may be lined with an optional barrier layer (26), such as TaN / Ta. Subsequently, a metal material such as tungsten or copper is deposited, and then chemical mechanical polishing (CMP) is performed to fill the contact holes (24) with the metal material, thereby forming via contacts (28, 29) that are electrically connected to the exposed drain region (12) and source region (14), respectively. Additional insulating material is deposited to increase the height of the insulating material (22) (e.g., by oxide deposition).An insulating material (22) is placed over via contacts (28, 29) and patterned to form contact holes (30) that expose them. A metal material layer (e.g., copper) is deposited on the structure, and then a CMP process is performed to fill the contact holes (30) with the metal material to form conductive contacts (32, 33) that electrically contact each of the via contacts (28, 29). The resulting structure is illustrated in FIG. 3a. The drain region (12), source region (14), channel region (16), and conductive gate (18) form a transistor (34) for selectively operating the RRAM cell that is subsequently formed.

[0013] An insulating layer (36) (e.g., SiN / SiCN) is formed on the conductive contacts (32, 33) and the insulating material (22). An insulating material (38) (e.g., oxide) is formed on the insulating layer (36). The insulating material (38) and the insulating layer (36) are patterned to form a contact hole (40) that exposes the conductive contact (32). A metal barrier layer (42) (e.g., TaN / Ta) is formed on the structure, including the inside of the contact hole (40). A metal material layer (e.g., copper or tungsten) is deposited on the structure, and then a CMP process is performed to fill the contact hole (40) with the metal material to form a conductive contact (43) that electrically contacts the conductive contact (32). The resulting structure is illustrated in FIG. 3b.

[0014] A conductive layer (44) (e.g., TaN) is formed on the structure. A lower electrode layer (46) is formed on the conductive layer (44). The lower electrode layer (46) may be made of a conductive material such as TiN, TaN, HfN, TaAlN, Ti, Ta, Pt, iridium, or ruthenium. A resistive switching dielectric material layer (48) (also referred to herein as the RSDM layer (48)) is formed on the lower electrode layer (46). The RSDM layer (48) is a transition metal oxide (e.g., HfO2, Al2O3, TaO2).x , TiO x , WO x , VO x , CuO x It may be a single layer of a switching oxide such as ). The RSDM layer (48) may also include multiple sublayers of different oxides and metals. Non-limiting examples of sublayers that may be included within the RSDM layer (48) are transition metal oxides (e.g., HfO2, Al2O3, TaO2 x , TiO x , WO x , VO x , CuO x It may include a sublayer of an oxygen scavenger metal such as Ti or Ta on a sublayer of ), or a sublayer of HfO2 and a sublayer of Al2O3, or a sublayer of HfO2 and a sublayer of Hf and a sublayer of TaOx, or a sublayer of HfO2 and a sublayer of Ti and a sublayer of TiOx. An upper electrode layer (50) is formed on the RSDM layer (48). The upper electrode layer (50) may be made of a conductive material such as TiN, TaN, HfN, TaAlN, Ti, Ta, Pt, iridium, or ruthenium. A conductive layer (52) (e.g., TaN) is formed on the upper electrode layer (50). The resulting structure is illustrated in FIG. 3C.

[0015] The conductive layer (52) is patterned (i.e., covered with photoresist, and a portion of this photoresist is removed through photolithography, and portions of the exposed conductive layer (52) are selectively removed, for example, by etching). For example, photoresist is formed on the conductive layer (52), and then exposed and developed by a photolithography process so that portions of the photoresist, excluding blocks (53) of the photoresist, are selectively removed. Then, one or more etchings are used to selectively remove exposed portions of the conductive layer (52), the upper electrode layer (50), and the RSDM layer (48) (i.e. portions that are not protected from etching(s) because they are not under blocks (53) of the photoresist). By this process, a stack structure (S1) consisting of blocks of the conductive layer (52) placed on blocks of the upper electrode layer (50) placed on blocks of the RSDM layer (48) as shown in FIG. 3d remains.

[0016] After the block (53) of the photoresist is removed, insulating spacers (54) are formed on the sides of the stack structure (S1). The formation of the spacers is well known and involves forming a layer of material over the topology of the structure and performing etching that removes the material from the horizontal portions of the structure but leaves spacers of the corresponding material extending along the vertical surfaces of the structure. The insulating spacers (54) may be formed of silicon nitride (referred to as nitride herein) by nitride deposition and etching. The resulting structure is illustrated in FIG. 3e, where the block of the RSDM layer (48) is below the block of the upper electrode layer, the block of the upper electrode layer is below the block of the conductive layer, and the insulating spacers (54) extend along the sidewalls of these three blocks.

[0017] Next, etching is used to selectively remove exposed portions of the lower electrode layer (46) and the conductive layer (44) (i.e., blocks of the conductive layer (52), blocks of the upper electrode layer (50), blocks of the RSDM layer (48), and portions of these layers not located under the insulating spacers (54). The stack structure (S1) after this etching includes, as shown in FIG. 3f, blocks of the conductive layer (52), blocks of the upper electrode layer (50), blocks of the RSDM layer (48) (their side surfaces are in contact with the inner opposing side surfaces of the insulating spacers (54) facing each other), blocks of the lower electrode layer (46), and blocks of the conductive layer (44) (their side surfaces are aligned with the outer opposing side surfaces of the insulating spacers (54) facing in opposite directions). As illustrated in FIG. 3g, deposition and etching are performed to thicken the insulating spacers (54) so ​​that the insulating spacers (54) extend along the side surfaces of the blocks of the lower electrode layer (46) and the blocks of the conductive layer (44).

[0018] The structure is covered with an insulating material (56) (e.g., oxide), which is flattened to a block of the conductive layer (52). The insulating material (56) is patterned to form a contact hole (58) that extends to and exposes the conductive contact (33). A layer of metal material (e.g., copper) is deposited on the structure, and then a CMP process is performed to form a via contact (60) to electrically contact the conductive contact (33) by filling the contact hole (58) with the metal material. Additional insulating material is deposited to increase the height of the insulating material (56) (e.g., by oxide deposition). The insulating material (56) is patterned to form contact holes (62) on a block of the conductive layer (52) and the via contact (60). A metal material layer (e.g., copper) is deposited on the structure, and then a CMP process is performed to form a conductive contact (64) that is electrically in contact with a block of the conductive layer (52) and a conductive contact (65) that is electrically in contact with a via contact (60) by filling the contact holes (62) with the metal material. The resulting structure is illustrated in FIG. 3h and is a semiconductor device (66) having an RRAM cell (68) comprising a block of the upper electrode layer (50) on a block of the RSDM layer (48) on a block of the lower electrode layer (46).

[0019] The method of forming this semiconductor device has many advantages. For example, RRAM cell formation can be implemented as part of the metal BEOL (Back End Of Line) portion of the process flow, which is typically the rear end of the semiconductor device manufacturing process flow where metal interconnects are formed throughout the semiconductor device to electrically connect various components. By doing so, the number of masking steps for forming interconnections within the semiconductor device (66) as well as the RRAM cell (68) can be reduced. Even if the block of the lower electrode layer (46) has a first lateral dimension larger than the second lateral dimension of the block of the RSDM layer and the block of the upper electrode layer (50), only one patterning step is used to define the block of the lower electrode layer (46), the block of the RSDM layer (48), and the block of the upper electrode layer (50). Thus, redeposition of metal that could short-circuit them along the sidewalls of the blocks of these layers is prevented. In addition, excessive etching at the sidewalls of the blocks of these layers, which could damage the RSDM material, form rough sidewall surfaces, and degrade cell performance, is prevented. Insulating spacers (54) protect the sidewalls of the blocks of the upper electrode layer (50) and the RSDM layer (48) after they are formed, and the insulating spacers (54) are used as a self-aligning mask to etch the lower electrode layer (46) to form blocks of the lower electrode layer (46) having larger transverse dimensions than the blocks above.

[0020] It should be understood that the present disclosure is not limited to the example(s) described herein, but encompasses any and all variations falling within the scope of any claim. For example, reference to the disclosure or invention or embodiments in this specification is not intended to limit the scope of any claim or claim term, but instead merely refers to one or more features that may be covered by one or more claims. The materials, processes, and numerical embodiments described above are merely illustrative and should not be construed as limiting the claims. Furthermore, as is evident from the claims and the specification, not all method operations need to be performed in the exact order described or claimed, but rather may be performed in any order that enables the proper formation of the semiconductor device described herein (unless there is an explicit limitation on any order). A single layer of material may be formed from multiple layers of such material or similar material, and vice versa. For example, the conductive layer (52) may be omitted, the conductive layer (44) may be omitted, the insulating layer (36) may be omitted, or a combination thereof may be omitted. Finally, the terms “forming” and “being formed” as used herein will include material deposition, material growth, or any other technique in providing the material as disclosed or claimed.

[0021] It should be noted that, as used herein, both the terms "on" and "on" comprehensively include "directly on" (no intermediate material, element, or space placed between them) and "indirectly on" (intermediate material, element, or space placed between them). Likewise, the term "adjacent" includes "directly adjacent" (no intermediate material, element, or space placed between them) and "indirectly adjacent" (intermediate material, element, or space placed between them); "embedded in" includes "directly embedded in" (no intermediate material, element, or space placed between them) and "indirectly embedded in" (intermediate material, element, or space placed between them); and "electrically coupled" includes "directly electrically coupled to" (no intermediate material or element electrically connecting the elements together between them) and "indirectly electrically coupled to" (intermediate material or element electrically connecting the elements together between them). For example, forming an element "on a semiconductor substrate" may include not only forming the element directly on the semiconductor substrate without an intermediate material / element in between, but also forming the element indirectly on the semiconductor substrate with one or more intermediate materials / elements in between.

Claims

Claim 1 A method for forming a semiconductor device comprises the steps of: forming an insulating material on a semiconductor substrate; forming a first contact hole extending through the insulating material; forming a first conductive contact within the first contact hole; forming a lower electrode layer on the insulating material and electrically connected to the first conductive contact; forming a resistive switching dielectric material layer directly on the lower electrode layer; forming an upper electrode layer directly on the resistive switching dielectric material layer; selectively removing portions of the upper electrode layer to form a block of the upper electrode layer; selectively removing portions of the resistive switching dielectric material layer to form a block of the resistive switching dielectric material layer disposed below the block of the upper electrode layer; forming insulating spacers extending on the lower electrode layer and along the sidewalls of the block of the upper electrode layer and the block of the resistive switching dielectric material layer; A method for forming a semiconductor device, comprising the step of selectively removing portions of the lower electrode layer to form a block of the lower electrode layer, wherein the block of the lower electrode layer is disposed below the insulating spacers, the block of the resistive switching dielectric material layer, and the block of the upper electrode layer. Claim 2 A method for forming a semiconductor device according to claim 1, wherein selectively removing portions of the upper electrode layer and selectively removing portions of the resistive switching dielectric material layer comprises: forming a photoresist on the upper electrode layer and the resistive switching dielectric material layer; exposing and developing the photoresist to remove portions of the photoresist to form a block of the photoresist; and performing one or more etchings to selectively remove portions of the upper electrode layer and portions of the resistive switching dielectric material layer, wherein the block of the upper electrode layer and the block of the resistive switching dielectric material layer are disposed below the block of the photoresist. Claim 3 A method for forming a semiconductor device according to claim 1, wherein the insulating spacers include inner opposing side surfaces facing each other and outer opposing side surfaces facing in opposite directions; the block of the upper electrode layer and the block of the resistive switching dielectric material layer include side surfaces in contact with the inner opposing side surfaces of the insulating spacers; and the block of the lower electrode layer includes side surfaces aligned with the outer opposing side surfaces of the insulating spacers. Claim 4 A method for forming a semiconductor device according to claim 1, comprising: a step of forming a conductive layer on the upper electrode layer prior to selectively removing portions of the upper electrode layer and selectively removing portions of the resistive switching dielectric material layer; and a step of selectively removing portions of the conductive layer to form a block of the conductive layer; wherein, after selectively removing portions of the upper electrode layer to form a block of the upper electrode layer, the block of the conductive layer is disposed on the block of the upper electrode layer. Claim 5 A method for forming a semiconductor device according to claim 1, comprising the steps of: forming a transistor on the semiconductor substrate; and forming a second conductive contact electrically connected to the transistor and the first conductive contact. Claim 6 A method for forming a semiconductor device according to claim 5, wherein the second conductive contact comprises a metal material. Claim 7 A method for forming a semiconductor device according to claim 5, wherein the second conductive contact comprises copper or tungsten. Claim 8 A method for forming a semiconductor device according to claim 5, comprising: a step of forming a conductive layer on the insulating material; and a step of selectively removing portions of the conductive layer to form a block of the conductive layer disposed between a block of the lower electrode layer and a first conductive contact. Claim 9 A method for forming a semiconductor device according to claim 8, wherein the conductive layer comprises TaN. Claim 10 A method for forming a semiconductor device according to claim 8, wherein the blocks of the conductive layer include sidewalls aligned with the sidewalls of the blocks of the lower electrode layer. Claim 11 A method for forming a semiconductor device according to claim 1, comprising, after the step of forming a block of the lower electrode layer, the step of thickening the insulating spacers so that the insulating spacers extend along the sidewalls of the block of the lower electrode layer. Claim 12 A method for forming a semiconductor device according to claim 1, wherein the first conductive contact comprises a metal material. Claim 13 A method for forming a semiconductor device according to claim 1, wherein the first conductive contact comprises copper or tungsten. Claim 14 A method for forming a semiconductor device according to claim 1, wherein the first conductive contact comprises layers of TaN and Ta. Claim 15 A method for forming a semiconductor device according to claim 1, wherein the lower electrode layer and the upper electrode layer each comprise TiN, TaN, HfN, TaAlN, Ti, Ta, Pt, iridium, or ruthenium. Claim 16 A method for forming a semiconductor device according to claim 1, wherein the resistive switching dielectric material layer comprises HfO2, Al2O3, TaOx, TiOx, WOx, VOx, or CuOx. Claim 17 A method for forming a semiconductor device according to claim 1, wherein the resistive switching dielectric material layer comprises a sublayer of HfO2 and a sublayer of Al2O3. Claim 18 A method for forming a semiconductor device according to claim 1, wherein the resistive switching dielectric material layer comprises a sublayer of HfO2, a sublayer of Hf, and a sublayer of TaOx. Claim 19 A method for forming a semiconductor device according to claim 1, wherein the resistive switching dielectric material layer comprises a sublayer of HfO2, a sublayer of Ti, and a sublayer of TiOx.