Methods for fabricating memory and memory

By controlling the process parameters during the epitaxial layer growth process in 3D NAND memory, the top surface of the epitaxial layer in the channel hole is made flush, which solves the problem of inconsistent channel hole depth, ensures the consistency of electrical and storage performance of the memory, and expands the process window.

CN114400230BActive Publication Date: 2026-11-17YANGTZE MEMORY TECH CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202111652436.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-30
Publication Date
2026-11-17
Estimated Expiration
2041-12-30

AI Technical Summary

Technical Problem

As the number of stacked layers in 3D NAND memory increases, the poor consistency of the depth of the channel vias leads to inconsistent charge storage structure height, affecting the electrical and storage performance of the memory.

Method used

By controlling the process parameters during the epitaxial layer growth process, epitaxial layers of different heights are grown in channel holes of different depths, so that the top surfaces of the epitaxial layers in multiple channel holes are basically flush. After forming the charge storage structure, it is ensured that the height of the charge storage structure in channel holes of different depths is basically the same.

Benefits of technology

This invention enables memory cells corresponding to different channel vias in a memory to have essentially the same electrical parameters, ensuring the uniformity of the memory's electrical and storage performance and expanding the process window for channel vias.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114400230B_ABST
    Figure CN114400230B_ABST
Patent Text Reader

Abstract

This application discloses a method for fabricating a memory and a memory. The method for fabricating the memory includes: forming a stacked structure on a substrate; forming a plurality of channel holes that penetrate the stacked structure and extend into the substrate; wherein at least two of the plurality of channel holes have different depths; forming an epitaxial layer in the plurality of channel holes; wherein the top surface of the epitaxial layer in the plurality of channel holes is substantially flush.
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 more specifically, to a method for fabricating a memory and a memory itself. Background Technology

[0002] With the continuous development of 3D NAND technology, the number of layers that can be vertically stacked in three-dimensional memory is increasing, from 24 layers, 32 layers, 64 layers to more than 100 layers of high-order stacking structures, which can significantly increase storage density and reduce the price per unit storage cell.

[0003] In 3D NAND technology, vias determine the formation of the charge storage layer and its storage performance, making them the most critical process step. However, as the number of stacked layers increases, controlling the via process becomes increasingly difficult. Poor depth consistency of vias has an extremely negative impact on the architecture of 3D NAND memories; therefore, controlling the depth consistency of vias is of great importance. Summary of the Invention

[0004] According to one aspect of this application, a method for fabricating a memory is provided, comprising:

[0005] A stacked structure is formed on the substrate;

[0006] A plurality of channel holes are formed that penetrate the stacked structure and extend into the substrate; wherein at least two of the plurality of channel holes have different depths;

[0007] An epitaxial layer is formed within the plurality of channel holes; wherein the top surface of the epitaxial layer in the plurality of channel holes is substantially flush.

[0008] In some embodiments, each of the channel holes includes a first segment located within the substrate and a second segment located within the stacked structure; the epitaxial layer includes a first epitaxial layer and a second epitaxial layer, wherein the first epitaxial layer and the second epitaxial layer are an integral structure;

[0009] An epitaxial layer is formed within the plurality of channel holes; wherein the top surface of the epitaxial layer in the plurality of channel holes is substantially flush, including:

[0010] A first epitaxial layer is formed in each of the first hole segments by an epitaxial growth process; wherein the first epitaxial layers located in different first hole segments are substantially flush with the surface of the stacked structure.

[0011] Within each second hole segment, a second epitaxial layer of substantially the same height is formed on the first epitaxial layer through the epitaxial growth process, so that the top surfaces of the second epitaxial layers in the plurality of channel holes are substantially flush.

[0012] In some embodiments, forming a stacked structure on a substrate includes: sequentially forming a sacrificial layer and a dummy gate stacked structure on the substrate; wherein the dummy gate stacked structure includes alternately stacked insulating layers and dummy gate layers;

[0013] The preparation method further includes:

[0014] A charge storage structure is formed covering the sidewalls of the channel aperture and the epitaxial layer; wherein the charge storage structure includes a charge storage layer and a channel layer arranged sequentially along the radial direction of the channel aperture; the top surface of the epitaxial layer is located in the aperture segment corresponding to the sacrificial layer, so that the bottom of the charge storage structure is located in the aperture segment corresponding to the sacrificial layer;

[0015] Remove the substrate and the sacrificial layer to expose the bottom of the charge storage structure;

[0016] Remove the charge storage layer at the bottom of the charge storage structure so that the bottom surface of the remaining charge storage layer is substantially flush with the bottom surface of the stacked structure, and expose the channel layer;

[0017] A semiconductor layer is formed to cover the exposed channel layer.

[0018] In some embodiments, the sacrificial layer includes a first sacrificial layer and a second sacrificial layer stacked together, wherein the first sacrificial layer is located between the substrate and the second sacrificial layer, and the second sacrificial layer is an etch stop layer when the first sacrificial layer is removed;

[0019] The method of forming an epitaxial layer within the plurality of channel holes includes: forming the epitaxial layer by an epitaxial growth process; wherein the top surface of the epitaxial layer is located in the hole segment corresponding to the first sacrificial layer, so that the bottom of the charge storage structure is located in the hole segment corresponding to the first sacrificial layer.

[0020] In some embodiments, the material of the first sacrificial layer includes silicon oxide and / or silicon oxynitride; and / or,

[0021] The material of the second sacrificial layer includes at least one of metal, silicon nitride, silicon oxide, polycrystalline silicon, or doped polycrystalline silicon.

[0022] In some embodiments, the preparation method further includes:

[0023] The epitaxial layer is removed before a semiconductor layer covering the exposed channel layer is formed.

[0024] In some embodiments, the preparation method further includes:

[0025] Before forming the dummy gate stack structure, a via is formed through the sacrificial layer;

[0026] The vias are filled to form a first barrier layer;

[0027] After the charge storage structure is formed, a pseudo-channel structure is formed that penetrates the pseudo-gate stack structure and extends into the first barrier layer.

[0028] The step of removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure includes: removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure and the first barrier layer;

[0029] The semiconductor layer forming the exposed channel layer includes: a semiconductor layer forming the exposed channel layer and the first barrier layer.

[0030] In some embodiments, the preparation method further includes:

[0031] Before forming the dummy gate stack structure, a trench is formed through the sacrificial layer; wherein the location of the trench is different from the location of the via.

[0032] Fill the trench to form a second barrier layer;

[0033] After the charge storage structure is formed, a gate trench is formed that penetrates the pseudo-gate stack structure and extends into the second barrier layer;

[0034] The dummy gate layer is replaced with a gate layer by the gate spacer to obtain a gate stack structure;

[0035] An array of common sources is formed within the gate slot.

[0036] In some embodiments, the preparation method further includes:

[0037] A third barrier layer is formed covering the sacrificial layer and the first barrier layer; wherein the third barrier layer is located between the sacrificial layer and the dummy gate stack structure;

[0038] The formation of a pseudo-channel structure within the first barrier layer that penetrates the pseudo-gate stack structure and extends into the via includes: forming a pseudo-channel structure that penetrates the pseudo-gate stack structure and the third barrier layer and extends into the first barrier layer.

[0039] The step of removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure and the first barrier layer includes: removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure, the first barrier layer, and the third barrier layer;

[0040] The semiconductor layer forming the exposed channel layer and the first barrier layer includes: forming a semiconductor layer covering the exposed channel layer, the first barrier layer and the third barrier layer.

[0041] In some embodiments, the third barrier layer comprises at least one of polysilicon, doped polysilicon, or metal.

[0042] According to another aspect of this application, a memory is provided, comprising:

[0043] A semiconductor layer, a gate stack structure located on the semiconductor layer, the gate stack structure comprising alternately stacked insulating layers and gate layers; and,

[0044] Multiple memory pillars are disposed within the gate stack structure; wherein the multiple memory pillars extend to be substantially flush with the bottom surface of the semiconductor layer.

[0045] In some embodiments, the memory further includes:

[0046] The first barrier layer and the second barrier layer are both disposed in the semiconductor layer, and the positions of the first barrier layer and the second barrier layer are different;

[0047] A pseudo-channel structure is disposed within the gate stack structure and extends into the first barrier layer; and...

[0048] The array common source is disposed in the gate stack structure and extends into the second barrier layer.

[0049] In some embodiments, the storage column includes a charge storage layer and a channel layer arranged radially thereon, and the memory further includes:

[0050] A third barrier layer is located between the semiconductor layer and the gate stack structure, and the charge storage layer extends through the third barrier layer to the top surface of the semiconductor layer; the channel layer extends through the third barrier layer into the semiconductor layer.

[0051] In the memory fabrication method provided in this application embodiment, when the depths of multiple vias are inconsistent, by controlling the process parameters during the epitaxial layer growth process, epitaxial layers of different heights are grown in vias of different depths, making the top surfaces of the epitaxial layers in multiple vias essentially flush. After forming charge storage structures on the inner walls of the vias and the epitaxial layers, the heights of the charge storage structures in vias of different depths can be essentially the same, ensuring that the memory cells corresponding to different vias in the memory have essentially consistent electrical parameters, guaranteeing the uniformity of the memory's electrical and storage performance. Furthermore, by controlling the height of the charge storage structure within the vias by the position of the top surface of the epitaxial growth layer, the influence of the via depth on the height of the charge storage structure is reduced, thereby appropriately relaxing the requirements for the consistency of via depth and expanding the process window for forming the vias. Attached Figure Description

[0052] Figure 1 A cross-sectional schematic diagram of a channel hole obtained by a low-temperature etching process, provided for an embodiment of this application;

[0053] Figure 2 A schematic flowchart illustrating a method for fabricating a memory according to an embodiment of this application;

[0054] Figures 3a to 3h This is a cross-sectional schematic diagram of a memory obtained by a memory fabrication method provided in an embodiment of this application;

[0055] Figure 4 A schematic flowchart illustrating another method for fabricating a memory provided in this application embodiment;

[0056] Figure 5 A top view of a partial structure of the memory provided in an embodiment of this application;

[0057] Figures 6a to 6p This is a schematic cross-sectional view of a memory obtained by another memory fabrication method provided in the embodiments of this application;

[0058] Figures 7a to 7t This is a cross-sectional schematic diagram of a memory obtained by using another memory fabrication method provided in the embodiments of this application.

[0059] Explanation of reference numerals in the attached figures:

[0060] 100: Substrate; 200: Sacrificial layer; 201: First sacrificial layer; 202: Second sacrificial layer; 300: Pseudo-gate stacked structure; 301: Insulating layer; 3011: First insulating layer; 302: Pseudo-gate layer; 303: First dielectric layer; 304: Second dielectric layer; 300': Gate stacked structure; 302': Gate layer;

[0061] 400: Channel hole, 410: Epitaxial layer, 420: Charge storage structure, 421: Charge storage layer, 4211: Barrier layer, 4212: Charge trapping layer, 4213: Tunneling layer, 422: Channel layer;

[0062] 500: Through-hole, 510: First barrier layer, 530: Pseudo-channel structure, 600: Trench, 610: Second barrier layer, 620: Gate spacer, 630: Array common source, 631: Third dielectric layer; 632: Conductive layer;

[0063] 700: Contact plug; 800: Semiconductor layer; 810: Contact hole; 820: Fourth dielectric layer; 821: First opening; 830: Interconnect layer; 900: Third barrier layer; 910: First oxide layer; 920: Second oxide layer. Detailed Implementation

[0064] The technical solution of this application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0065] In the description of this application, it should be understood that the terms "length", "width", "depth", "upper", "lower", "outer", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this application.

[0066] In the fabrication process of 3D NAND flash memory, as the number of stacked layers increases, the process of etching deep holes using room-temperature (20 to 80°C) plasma dry etching becomes increasingly difficult and time-consuming. Low-temperature etching processes offer significant advantages in controlling deep hole etching. Low-temperature etching refers to low-temperature (< -30°C) plasma dry etching. Low-temperature deep hole etching offers faster etching speeds and a higher bottom critical dimension (CD) to top critical dimension ratio. However, its drawback is a lower selectivity for etching silicon nitride, silicon oxide, and silicon substrates, leading to inconsistent channel depths.

[0067] See here. Figure 1 , Figure 1 This is a schematic cross-sectional view of a channel hole obtained using a low-temperature etching process. (Example:) Figure 1 As shown, the depth of the via segments within the substrate is relatively large, and the bottom positions of the vias vary, resulting in poor consistency. Inconsistent via depth leads to inconsistent heights of the charge storage structures formed within the vias, which has an extremely negative impact on the memory's storage performance.

[0068] Based on this, embodiments of this application provide a method for fabricating a memory. For example... Figure 2 As shown, the preparation method includes the following steps:

[0069] S100: A stacked structure is formed on the substrate;

[0070] S200: Forming a plurality of channel holes that penetrate the stacked structure and extend into the substrate; wherein at least two of the plurality of channel holes have different depths;

[0071] S300: An epitaxial layer is formed in multiple channel holes; wherein the top surface of the epitaxial layer in the multiple channel holes is basically flush.

[0072] In this embodiment, when the depths of multiple vias are inconsistent, by controlling the process parameters during epitaxial layer growth, epitaxial layers of different heights are grown in vias of different depths. This ensures that the top surfaces of the epitaxial layers in multiple vias are essentially flush. Consequently, after forming charge storage structures on the inner walls of the vias and the epitaxial layers, the heights of the charge storage structures in vias of different depths are essentially the same. This allows memory cells corresponding to different vias in the memory to have essentially consistent electrical parameters, ensuring the uniformity of the memory's electrical and storage performance. Furthermore, by controlling the height of the charge storage structure within the vias by the position of the top surface of the epitaxial growth layer, the influence of the via depth on the height of the charge storage structure is reduced. This allows for a more relaxed requirement for consistent via depth, expanding the process window for forming the vias. The following section combines... Figures 3a to 3h The cross-sectional schematic diagram of the corresponding memory during the fabrication process provides a detailed description of a memory fabrication method provided in this application embodiment.

[0073] Please refer to Figure 3a A stacked structure is formed on the substrate 100. The substrate 100 can be a semiconductor substrate such as a silicon substrate, a germanium (Ge) substrate, a silicon germanide (SiGe) substrate, a silicon-on-insulator (SOI) substrate, or a germanium-on-insulator (GOI) substrate. In this embodiment, the substrate 100 is a silicon substrate.

[0074] The stacked structure may include a dummy gate stacked structure 300, which includes an insulating layer 301 and a dummy gate layer 302 alternately disposed sequentially. In subsequent processes, the dummy gate layer 302 may be removed and replaced with a conductive gate layer. Alternatively, in some embodiments, the stacked structure may include a gate stacked structure, in which an insulating layer 301 and a gate layer are alternately disposed sequentially.

[0075] like Figure 3a As shown, this embodiment of the present disclosure takes as an example a stacked structure including a pseudo-gate stacked structure 300, which includes alternating layers of insulating layer 301 and pseudo-gate layer 302.

[0076] Please refer to Figure 3b A plurality of channel holes 400 are formed through the stacked structure and extend into the substrate 100; wherein at least two of the plurality of channel holes 400 have different depths.

[0077] The channel via 400 can be formed in the stacked structure using methods such as room temperature plasma dry etching or low temperature plasma dry etching. In this embodiment, the channel via 400 can be formed using a low temperature plasma dry etching process.

[0078] Please refer to Figure 3c An epitaxial layer 410 is formed within a plurality of channel holes 400; wherein the top surfaces of the epitaxial layers 410 in the plurality of channel holes 400 are substantially flush. Here, the top surface of the epitaxial layer 400 refers to the surface of the epitaxial layer 400 relatively away from the substrate. In some embodiments, each channel hole 400 includes a first hole segment located within the substrate and a second hole segment located within the stacked structure, and the epitaxial layer 410 includes a first epitaxial layer and a second epitaxial layer. The step of forming the epitaxial layer 410 within the plurality of channel holes 400 includes:

[0079] A first epitaxial layer is formed in each first hole segment through an epitaxial growth process; wherein the first epitaxial layer located in different first hole segments is basically flush with the surface of the stacked structure.

[0080] Within each second hole segment, a second epitaxial layer of essentially the same height is formed on the first epitaxial layer through an epitaxial growth process, so that the top surfaces of the second epitaxial layers in the multiple channel holes are essentially flush.

[0081] Here, the first epitaxial layer and the second epitaxial layer can be an integral structure. The top surface of the second epitaxial layer is the surface of the second epitaxial layer that is relatively far away from the substrate 100.

[0082] The channel filling effect of the epitaxial layer 410 depends on the setting of process parameters during epitaxial growth. In some embodiments, by controlling the process parameters of epitaxial growth, the growth rate of the epitaxial growth material at the top of the first hole segment is made faster than the growth rate at other locations within the hole segment. This causes the epitaxial growth material on the top sidewall of the hole segment to fill the hole cross-section at the top of the hole segment earlier than at other locations within the hole segment, effectively sealing the first hole segment earlier. Exemplarily, the process parameters of epitaxial growth include, but are not limited to, at least one of temperature, gas flow rate, gas pressure, or gas composition.

[0083] It can be understood that although the depths of different first hole segments are not completely consistent, their hole diameters are basically the same. Therefore, by pre-sealing different first hole segments, the surfaces of the first epitaxial layers in different first hole segments that are relatively close to the stacked structure are basically flush.

[0084] When the obtained first epitaxial layer is flush with the surface relatively far from the substrate, a second epitaxial layer of substantially the same height is then formed on the first epitaxial layer, thereby making the top surfaces of the epitaxial layers 410 inside and outside the different channel holes 400 substantially flush. Here, the second epitaxial layer is provided on the first epitaxial layer so that the epitaxial layer 410 extends into the corresponding hole segment of the stacked structure to realize its corresponding storage function.

[0085] It should be noted that the first epitaxial layer and the second epitaxial layer can be completed in two separate process steps, that is, the first epitaxial layer is formed in the first process step and the second epitaxial layer is formed in the second process step.

[0086] In some embodiments, the first epitaxial layer and the second epitaxial layer can be completed in the same process step, that is, after the first epitaxial layer is completed in the same process step, the second epitaxial layer is grown immediately thereafter.

[0087] It is understandable that the second epitaxial layer can also be formed simultaneously with the formation of the first epitaxial layer. When the epitaxial growth material on the top sidewall of the first hole segment in the channel 400 grows radially along the hole segment to form the first epitaxial layer, epitaxial growth material will also grow along the axial direction of the hole segment on the already grown epitaxial growth material, thus forming the second epitaxial layer. That is, the formation processes of the first and second epitaxial layers overlap in time.

[0088] In some embodiments, such as Figure 3c As shown, voids exist in some parts of the first epitaxial layer. It is understood that this application utilizes the pre-sealing of the top of the first via segment to make the first epitaxial layer substantially flush with the surface of the stacked structure. However, in some deeper first via segments, when the top of the first via segment is sealed, the interior of the first via segment may not be completely filled, resulting in voids remaining within the first epitaxial layer. These voids cause the epitaxial layer 410 to have higher impedance when conducting between the substrate 100 and the channel layer, affecting the electrical and storage performance of the memory.

[0089] In view of this, this application proposes further optimization schemes. In some embodiments, such as Figure 4 As shown, step S100 includes: S110: sequentially forming a sacrificial layer and a dummy gate stack structure on the substrate; wherein, the dummy gate stack structure includes alternately stacked insulating layers and dummy gate layers;

[0090] The methods for fabricating memory also include:

[0091] S400: A charge storage structure is formed covering the sidewall of the channel hole and the epitaxial layer; wherein, the charge storage junction includes a charge storage layer and a channel layer arranged sequentially along the radial direction of the channel hole; the top surface of the epitaxial layer is located in the hole segment corresponding to the sacrificial layer, so that the bottom of the charge storage structure is located in the hole segment corresponding to the sacrificial layer;

[0092] S500: Remove the substrate and sacrificial layer to expose the bottom of the charge storage structure;

[0093] S600: Remove the charge storage layer at the bottom of the charge storage structure, so that the bottom surface of the remaining charge storage layer is substantially flush with the bottom surface of the stacked structure, and expose the channel layer;

[0094] S700: Forms a semiconductor layer covering the exposed surface of the channel layer.

[0095] See Figure 3a The stacked structure also includes a sacrificial layer 200, which is located between the substrate 100 and the dummy gate stacked structure 300. The layer of the dummy gate stacked structure 300 closest to the sacrificial layer 200 is an insulating layer 301, which is defined as the first insulating layer 3011.

[0096] For example, the insulating layer 301 may be made of silicon oxide. The dummy gate layer 302 may be made of silicon nitride. For example, the insulating layer 301 and the dummy gate layer 302 may be formed using physical vapor deposition (PVD), chemical vapor deposition (CVD), or atomic layer deposition (ALD) processes.

[0097] In some embodiments, the sacrificial layer 200 includes a first sacrificial layer 201 and a second sacrificial layer 202 stacked together, wherein the first sacrificial layer 201 is located between the substrate 100 and the second sacrificial layer 202, and the second sacrificial layer 202 is an etch stop layer when the first sacrificial layer 201 is removed.

[0098] For example, the material of the first sacrificial layer 201 may include silicon oxide and / or silicon oxynitride. The material of the second sacrificial layer 202 may include at least one of metal, silicon nitride, silicon oxynitride, polycrystalline silicon, or doped polycrystalline silicon, and the metal may include tungsten, copper, aluminum, or cobalt.

[0099] The second sacrificial layer 202 is an etching stop layer when the first sacrificial layer 201 is removed, so as to protect the first insulating layer 3011 when the first sacrificial layer 201 is removed and improve the thickness uniformity of the first insulating layer 3011.

[0100] See Figure 3dA charge storage structure 420 is formed covering the sidewalls of the channel via 400 and the epitaxial layer 410. Specifically, the charge storage structure 420 includes a charge storage layer 421 and a channel layer 422 arranged radially along the channel via 400. The charge storage layer 421 may include a barrier layer 4211, a charge trapping layer 4212, and a tunneling layer 4213 arranged radially along the channel via 400, wherein the tunneling layer 4213 is located between the charge trapping layer 4212 and the channel layer 422. Exemplarily, the barrier layer 4211 may be made of silicon oxide, the charge trapping layer 4212 may be made of silicon nitride, and the tunneling layer 4213 may be made of silicon oxide, thereby forming a charge storage layer 421 with an ONO (silicon oxide-silicon nitride-silicon oxide) structure. Exemplarily, the channel layer 422 may be made of polysilicon.

[0101] In some embodiments, a filling layer 423 is further formed within the channel hole 400 along the radial direction of the channel hole 400. A channel layer 422 is located between the charge storage layer 421 and the filling layer 423 along the radial direction of the channel hole 400. Exemplarily, the material of the filling layer 423 may include silicon oxide.

[0102] Furthermore, Figure 3d In this configuration, the top surface of the epitaxial layer 410 is located within the corresponding aperture of the sacrificial layer 200, thus ensuring that the bottom of the charge storage structure 420 is also located within the corresponding aperture of the sacrificial layer 200. This allows the bottom of the charge storage structure 420 to be exposed after the sacrificial layer is removed. When the sacrificial layer 200 comprises a stacked first sacrificial layer 201 and a second sacrificial layer 202, the top surface of the epitaxial layer 410 is located within the corresponding aperture of the first sacrificial layer 201, thus ensuring that the bottom of the charge storage structure 420 is also located within the corresponding aperture of the first sacrificial layer 201.

[0103] Next, the substrate 100 is removed to expose the sacrificial layer 200. Exemplarily, the substrate 100 can be removed using plasma dry etching, gas reactive dry etching (GAS Etch), wet etching, and chemical mechanical polishing (CMP).

[0104] In this embodiment, participants Figure 3d and Figure 3e The sacrificial layer 200 includes a first sacrificial layer 201 and a second sacrificial layer 202. After the substrate 100 is removed, the first sacrificial layer 201 is exposed. Here, chemical mechanical polishing is used to remove the substrate 100, simultaneously removing a portion of the epitaxial layer 410 located within the substrate 100. At this time, a portion of the epitaxial layer 410 located within the first sacrificial layer 201 is retained. Next, the sacrificial layer 200 is removed, exposing the bottom of the charge storage structure 420 and the remaining epitaxial layer 410. Exemplarily, a plasma dry etching process, a gas reactive dry etching process, or a wet etching process can be used to remove the sacrificial layer 200.

[0105] As described above, the bottom of the charge storage structure 420 is located in the aperture segment corresponding to the first sacrificial layer 201. Therefore, removing the first sacrificial layer 201 exposes the bottom of the charge storage structure 420. The second sacrificial layer 202 can be removed or retained.

[0106] In this embodiment, see Figure 3f The first sacrificial layer 201 is removed using a plasma dry etching process, exposing the bottom of the charge storage structure 420, the remaining epitaxial layer 410, and the second sacrificial layer 202. Here, the second sacrificial layer 202 serves as an etching stop layer during the removal of the first sacrificial layer 201, thereby protecting the first insulating layer 3011 and improving the thickness uniformity of the first insulating layer 3011.

[0107] This application does not impose any restrictions on the method of removing the second sacrificial layer 202. In this embodiment, the material of the second sacrificial layer 202 includes silicon nitride or silicon oxide nitride. Since the charge storage layer 421 has an ONO (silicon oxide-silicon nitride-silicon oxide) structure, the material of the charge storage layer 421 and the material of the second sacrificial layer 202 have a small etching selectivity ratio. Therefore, the second sacrificial layer 202 can be removed simultaneously when removing the charge storage layer 421 at the bottom of the charge storage structure 420.

[0108] For example, plasma dry etching, gas reactive dry etching, or wet etching can be used to remove the charge storage layer 421 at the bottom of the charge storage structure 420, exposing the channel layer 422.

[0109] In this embodiment, see Figure 3d and 3g The charge storage layer 421 and the second sacrificial layer 202 at the bottom of the charge storage structure 420 are removed by plasma dry etching process, exposing the sidewalls of the channel layer 422 and the first insulating layer 3011.

[0110] It should be noted that although the material of the first insulating layer 3011 includes silicon oxide, due to the relatively thick thickness of the first insulating layer 3011, a very small portion of the first insulating layer 3011 may be lost when the charge storage layer 421 and the second sacrificial layer 202 are removed simultaneously. However, the vast majority of the first insulating layer 3011 is retained, and therefore, it will not affect the storage performance of the memory.

[0111] For example, the thickness of the second sacrificial layer 202 can be set so that the second sacrificial layer 202 and the charge storage layer 421 are removed almost simultaneously. For instance, if the removal rate of the second sacrificial layer 202 and the removal rate of the charge storage layer 421 are basically the same, the thickness of the second sacrificial layer 202 can be set to be basically equal to the thickness of the charge storage layer 421. Then, when the charge storage layer 421 is completely removed, the second sacrificial layer 202 is simultaneously completely removed. In this way, the etching of the first insulating layer 3011 is minimal, and most of the first insulating layer 3011 is retained. Alternatively, the thickness of the second sacrificial layer 202 can be set to be greater than the thickness of the charge storage layer 421. Then, after the charge storage layer 421 is removed, the second sacrificial layer 202 is removed by the same thickness simultaneously, and then the removal time is appropriately extended to ensure that the second sacrificial layer 202 is completely removed. In this way, the etching of the first insulating layer 3011 is minimal, and most of the first insulating layer 3011 is retained.

[0112] It should also be noted that the plasma dry etching process is anisotropic. Therefore, a charge storage layer 421 remains on the bottom surface of the channel layer 422. However, since the sidewalls of the channel layer 422 have been exposed, the charge storage layer 421 remaining on the bottom surface of the channel layer 422 does not affect the contact connection between the channel layer 422 and the subsequently formed semiconductor layer.

[0113] In addition, in some embodiments, when the material of the second sacrificial layer 202 includes metal, polysilicon or doped polysilicon, since the etching options for metal, polysilicon or doped polysilicon and the material of charge storage structure 420 are relatively large, the second sacrificial layer 202 can be removed in steps first, and then the charge storage layer 421 at the bottom of the charge storage structure 420 can be removed.

[0114] See further Figure 3g After removing the charge storage layer 421 at the bottom of the charge storage structure 420 and the second sacrificial layer 202, the bottom surface of the remaining charge storage layer 421 is substantially flush with the bottom surface of the stacked structure, thus making the bottom surfaces of the charge storage layers 421 in different channel holes 400 substantially flush. This ensures that the memory cells corresponding to different channel holes in the memory have substantially consistent electrical parameters, guaranteeing the uniformity of the memory's electrical and storage performance. Here, the bottom surface of the stacked structure is the surface of the stacked structure exposed after removing the second sacrificial layer 202.

[0115] See Figure 3h A semiconductor layer 800 is formed on the surface of the stacked structure exposed after the removal of the second sacrificial layer 202, and the semiconductor layer 800 covers the exposed channel layer 422. Exemplarily, the semiconductor layer 800 may be made of polysilicon and / or doped polysilicon. The semiconductor layer 800 is used to form a doped well region, electrically connecting the doped well region and peripheral circuitry via a common-source contact structure.

[0116] In this embodiment, a semiconductor layer covering the bottom of the channel layer is formed to replace the epitaxial layer and achieve electrical connection with the channel layer, thereby improving the electrical and storage performance of the memory.

[0117] Furthermore, the method for fabricating the memory also includes forming a pseudo-channel structure and an array of common sources. In some embodiments, the method for fabricating the memory includes forming a pseudo-channel structure, specifically, including the following steps:

[0118] Before forming the pseudo-gate stack structure, a via is formed through the sacrificial layer;

[0119] Fill the vias to form a first barrier layer;

[0120] After the charge storage structure is formed, a pseudo-channel structure is formed that penetrates the pseudo-gate stack structure and extends into the first barrier layer.

[0121] Removing the substrate and sacrificial layer to expose the bottom of the charge storage structure includes: removing the substrate and sacrificial layer to expose the bottom of the charge storage structure and the first barrier layer;

[0122] The semiconductor layer forming the exposed channel layer includes: forming a semiconductor layer covering the exposed channel layer and a first barrier layer.

[0123] In some embodiments, the method for fabricating the memory includes forming an array of common sources, specifically including the following steps:

[0124] Before forming the pseudo-gate stack structure, a trench is formed that penetrates the sacrificial layer; wherein the location of the trench is different from the location of the via.

[0125] Fill the trenches to form a second barrier layer;

[0126] After the charge storage structure is formed, a gate trench is formed that penetrates the pseudo-gate stack structure and extends into the second barrier layer.

[0127] By replacing the dummy gate layer with the gate layer through gate trenches, a gate stack structure is obtained;

[0128] An array of common sources is formed within the gate spacer.

[0129] The following is combined with Figure 5 The top view of the memory shown, and Figures 6a to 6o The cross-sectional schematic diagram of the corresponding memory during the fabrication process provides a detailed description of another memory fabrication method provided in this application embodiment.

[0130] here, Figures 6a to 6o yes Figure 5A partial structural view of cross-sections along the Y and X directions. The Z direction is parallel to the thickness direction of the substrate, and the X and Y directions are perpendicular to each other and both perpendicular to the Z direction.

[0131] See Figure 6a A sacrificial layer 200 is formed on the substrate 100; wherein the sacrificial layer 200 includes a first sacrificial layer 201 and a second sacrificial layer 202 stacked together, the first sacrificial layer 201 is located between the substrate 100 and the second sacrificial layer 202, and the second sacrificial layer 202 is an etch stop layer of the first sacrificial layer 201.

[0132] For example, the material of the first sacrificial layer 201 may include silicon oxide and / or silicon oxynitride. The material of the second sacrificial layer 202 may include silicon nitride and / or silicon oxynitride.

[0133] See Figure 6b This forms a through hole 500 and a trench 600 that penetrate the first sacrificial layer 201 and the second sacrificial layer 202, wherein the position of the trench 600 is different from the position of the through hole 500.

[0134] Next, see Figure 6c and Figure 6d A first barrier layer 510 is formed by filling the via 500, and a second barrier layer 610 is formed by filling the trench 600. The material of the first barrier layer 510 has a large etching selectivity ratio with the materials of the first sacrificial layer 201 and the second sacrificial layer 202, so that when the first sacrificial layer 201 and the second sacrificial layer 202 are removed in subsequent processes, the first barrier layer 510 is not etched or is etched to a very small degree. The materials of the first barrier layer 510 and the second barrier layer 610 can be the same or different.

[0135] For example, the material of the first barrier layer 510 may include at least one of polycrystalline silicon, doped polycrystalline silicon, or a metal, wherein the metal may include tungsten, copper, aluminum, or cobalt. The material of the second barrier layer 610 may include at least one of polycrystalline silicon, doped polycrystalline silicon, or a metal, wherein the metal may include tungsten, copper, aluminum, or cobalt.

[0136] Here, the steps of forming the first barrier layer 510 and the second barrier layer 610 can be performed separately or simultaneously. In some embodiments, the first barrier layer 510 and the second barrier layer 610 are made of different materials; therefore, the first barrier layer 510 and the second barrier layer 610 are formed in steps. For example, a through hole 500 penetrating the sacrificial layer 200 is formed first, and the through hole 500 is filled to form the first barrier layer 510. Then, a trench 600 penetrating the sacrificial layer 200 is formed, and the trench is filled to form the second barrier layer 610.

[0137] In this embodiment, the first barrier layer 510 and the second barrier layer 610 are made of the same material; for example, the first barrier layer 510 and the second barrier layer 610 comprise polycrystalline silicon. Therefore, the steps of forming the first barrier layer 510 and the second barrier layer 610 are performed simultaneously. See also... Figure 6c A barrier material layer is formed, covering the sacrificial layer 200 and filling the vias 500 and trenches 600. Exemplarily, the barrier material layer can be formed using physical vapor deposition, chemical vapor deposition, or atomic layer deposition processes.

[0138] See Figure 6d Chemical mechanical polishing is used to remove the barrier material layer on the surface of the second sacrificial layer 202, forming a first barrier layer 510 filling the via 500 and a second barrier layer 610 filling the trench 600. This simultaneous formation of the first barrier layer 510 and the second barrier layer 610 simplifies the process, shortens the fabrication cycle, and saves on process auxiliary materials such as photomasks and photoresists, thus reducing costs.

[0139] Referring to 6e, a dummy gate stack structure 300 is formed on the sacrificial layer 200, wherein the dummy gate stack structure 300 includes alternately stacked insulating layers 301 and dummy gate layers 302. Next, a plurality of channel holes 400 are formed through the dummy gate stack structure 300 and the sacrificial layer 200 and extending into the substrate 100, wherein at least two of the plurality of channel holes 400 have different depths.

[0140] See Figure 6f An epitaxial layer 410 is formed within multiple channel holes 400. The top surfaces of the epitaxial layer 410 are substantially flush within the multiple channel holes 400.

[0141] See Figure 6g A charge storage structure 420 is formed covering the sidewalls of the channel hole 400 and the epitaxial layer 410, wherein the charge storage structure 420 includes a charge storage layer 421 and a channel layer 422 arranged radially along the channel hole 400 (see...). Figure 3d The top surface of the epitaxial layer 410 is located in the corresponding aperture of the sacrificial layer 200, so that the bottom of the charge storage structure 420 is located in the corresponding aperture of the sacrificial layer 200.

[0142] See also Figure 6g The method for fabricating the memory further includes: forming a pseudo-channel hole that penetrates the pseudo-gate stack structure 300 and extends into the first barrier layer 510.

[0143] For example, when multiple pseudo-channel holes are formed, different pseudo-channel holes can extend into different first barrier layers 510.

[0144] Since the etching selectivity of the first sacrificial layer 201 and the insulating layer 301 and the pseudo gate layer 302 in the pseudo gate stack structure 300 is relatively low, if multiple pseudo channel holes extending into the first sacrificial layer 201 are directly formed, it is difficult to control the bottom position of the pseudo channel holes, resulting in poor depth consistency of the multiple pseudo channel holes.

[0145] In this embodiment, a first barrier layer 510 is formed in the sacrificial layer 200, allowing the dummy channel vias to penetrate the dummy gate stack structure 300 and extend into the first barrier layer 510. Since the first barrier layer 510 has a high etch selectivity with the insulating layer 301 and the dummy gate layer 320, it increases the etch resistance at the bottom of the dummy channel vias within the first barrier layer 510, which helps improve the depth consistency of the multiple formed dummy channel vias and reduces the process difficulty of forming dummy channel vias with essentially uniform depths. This, in turn, helps to expand the process window for forming dummy channel vias. Furthermore, during the subsequent removal of the sacrificial layer 200, the first barrier layer 510 can serve as an etch stop layer to protect the dummy channel structure 530.

[0146] It should be noted that the pseudo-channel hole can be formed after the charge storage structure 420 is formed, or the pseudo-channel hole can be formed simultaneously with the channel hole 400.

[0147] In some embodiments, the dummy channel via is formed after the charge storage structure 420 is formed. Specifically, the fabrication method further includes: after forming the charge storage structure 420, forming a dummy channel via that penetrates the dummy gate stack structure 300 and extends into the first barrier layer 510. Further, the dummy channel via is filled to form a dummy channel structure 530. For example, the dummy channel structure 530 may include a silicon oxide layer.

[0148] In some other embodiments, the dummy channel vias are formed simultaneously with the channel vias 400. Specifically, the fabrication method further includes forming dummy channel vias that penetrate the dummy gate stack structure 300 and extend into the first barrier layer 510 while forming multiple channel vias. Furthermore, while forming the charge storage structure 420 within the channel vias, a charge storage structure is also formed within the dummy channel vias. It should be noted that the charge storage structure filling the dummy channel vias is used to support the dummy gate stack structure 300. In subsequent processes, the charge storage layer at the bottom of the charge storage structure within the dummy channel vias will not be removed, and the channel layer within the dummy channel vias will be electrically connected to the semiconductor layer through the first barrier layer 510.

[0149] In some embodiments, the substrate 100 includes an array region and a peripheral circuit region, the array region including a core region and a stepped region, the stepped region being located between the core region and the peripheral circuit region. The dummy gate stack structure 300 has multiple steps formed in the stepped region. The charge storage structure 420 is located in the core region, and the dummy channel structure 530 is located in the stepped region. The method for fabricating the memory further includes:

[0150] Remove part of the pseudo-gate stack structure 300 in the peripheral circuit area and retain the first insulating layer 3011 on the sacrificial layer 200;

[0151] A first dielectric layer 303 is formed, covering the step area and the peripheral circuit area;

[0152] A second dielectric layer 304 is formed covering the pseudo-gate stack structure 300 and the first dielectric layer 303.

[0153] For example, the material of the first dielectric layer 303 may include at least one of titanium nitride, silicon oxide, or aluminum oxide. The material of the second dielectric layer 304 may include silicon oxide.

[0154] See next Figure 6g After forming the pseudo-channel structure 530, a gate trench 620 is formed that penetrates the pseudo-gate stack structure 300 and extends into the second barrier layer 610.

[0155] For example, when multiple gate spacers 620 are formed, different gate spacers may extend into different second barrier layers 610.

[0156] Since the etching selectivity of the first sacrificial layer 201 and the insulating layer 301 and the pseudo gate layer 302 in the pseudo gate stack structure 300 is relatively low, if multiple gate trenches 620 extending into the first sacrificial layer 201 are directly formed, it is difficult to control the bottom position of the gate trenches 620, resulting in poor depth consistency of the multiple gate trenches 620.

[0157] In this embodiment, a second barrier layer 610 is formed in the sacrificial layer 200, so that the gate trench 620 penetrates the dummy gate stack structure 300 and extends into the second barrier layer 610. Since the second barrier layer 610 has a high etch selectivity with the insulating layer 301 and the dummy gate layer 302, it can increase the etch resistance of the bottom of the gate trench 620 in the second barrier layer 610, which is beneficial to improve the depth consistency of the multiple gate trenches 620 formed, and reduces the process difficulty of forming gate trenches 620 with basically consistent depth, thereby helping to expand the process window for forming the gate trenches 620.

[0158] For example, the gate spacer 620 can be formed using plasma dry etching, gas reactive dry etching, wet etching, or other similar processes.

[0159] See Figure 6g and 6h The dummy gate layer 302 is removed through the gate spacer 620, and a gate layer 302' is deposited in the gap formed after the removal of the dummy gate layer 302 to obtain a gate stack structure 300'. Exemplarily, the material of the gate layer 302' may include metal and / or polysilicon, and the metal may include tungsten.

[0160] Next, an array of common sources 630 is formed within the gate trench 620. In some embodiments, the step of forming the array of common sources 630 includes: forming a third dielectric layer 631 covering the sidewalls and bottom surface of the gate trench 620; and forming a conductive layer 632 on the third dielectric layer 631.

[0161] Here, the third dielectric layer 631 covering the bottom surface of the gate spacer 620 can be removed, while the third dielectric layer 631 covering the side surface of the gate spacer 620 is retained, so that the conductive layer 632 and the second barrier layer 610 can be contacted and connected, thereby electrically connecting certain components in the stacked structure to the second barrier layer 610.

[0162] Exemplarily, the material of the third dielectric layer 631 may include one or more combinations of silicon oxide, silicon nitride, silicon oxynitride, or high dielectric constant materials. Exemplarily, high dielectric constant materials include materials with a dielectric constant higher than that of silicon nitride, such as aluminum oxide, hafnium oxide, zirconium oxide, magnesium oxide, etc. The material of the conductive layer 632 may include a metal (e.g., tungsten, aluminum, copper, or cobalt) and / or polycrystalline silicon. In some embodiments, see [link to documentation]. Figure 6i The method for fabricating the memory further includes forming a contact plug (CT) 700 in the peripheral circuit region, penetrating the second dielectric layer 304, the first dielectric layer 303, and the first insulating layer 3011, and extending into the second sacrificial layer 202. Exemplarily, the contact plug 700 may be made of metal and / or polysilicon, and the metal may include tungsten, copper, aluminum, or cobalt.

[0163] See Figure 6j The substrate 100 is removed, exposing the first sacrificial layer 201. In this embodiment, chemical mechanical polishing is used to remove the substrate 100, thereby simultaneously removing a portion of the epitaxial layer 410 in the substrate 100.

[0164] See Figure 3d and Figure 6kThe first sacrificial layer 201 and the barrier layer 4211 at the bottom of the charge storage structure 420 are removed by a gas dry etching process or a wet etching process, exposing the charge trapping layer 4212 at the bottom of the charge storage structure 420, the remaining epitaxial layer 410, the first barrier layer 510, the second barrier layer 610 and the second sacrificial layer 202.

[0165] It is understood that in this embodiment, the first sacrificial layer 201 is made of silicon oxide or silicon oxynitride. The barrier layer 4211 in the charge storage layer 421 is made of silicon oxide. Therefore, when removing the first sacrificial layer 201 using a gas dry etching process or a wet etching process, the barrier layer 4211, made of the same material, can be removed simultaneously.

[0166] See Figure 3d and Figure 6l The charge trapping layer 4212 and tunneling layer 4213 in the charge storage layer 421 are removed using a gas dry etching process or a wet etching process to expose the channel layer 422, and the second sacrificial layer 202 is removed simultaneously to expose the bottom of the gate stack structure 300' and the contact plug 700. Specifically, after removing the second sacrificial layer 202, the first insulating layer 3011 of the gate stack structure 300' is exposed.

[0167] It is understood that in this embodiment, the second sacrificial layer 202 is made of silicon nitride and / or silicon oxide nitride. The charge trapping layer 4212 is made of silicon nitride, and the tunneling layer 4213 is made of silicon oxide. Since the etching selectivity of the second sacrificial layer 202, the charge trapping layer 4212, and the tunneling layer 4213 is relatively small, the second sacrificial layer 202 can be removed simultaneously when the charge trapping layer 4212 and the tunneling layer 4213 are removed using a gas dry etching process or a wet etching process.

[0168] Furthermore, it is understood that gas dry etching or wet etching processes are isotropic, thus enabling the removal of all charge storage layers 421 at the bottom of the charge storage structure 420, causing the remaining epitaxial layer 410 to fall off along with the removal of the charge storage layers 421. This exposes more of the channel layer 422, increasing the contact area between the channel layer 422 and the subsequent semiconductor layer 800, thereby improving its storage performance.

[0169] exist Figure 6k and Figure 6lIn the process steps shown, the first barrier layer 510 and the second barrier layer 610 have a higher etch selectivity than the first sacrificial layer 200 and the second sacrificial layer 200. Therefore, the first barrier layer 510 is used as an etch stop layer to protect the pseudo-channel structure 530 when the first sacrificial layer 200 and the second sacrificial layer 200 are removed, and the second barrier layer 610 is used as an etch stop layer to protect the array common source 630 when the first sacrificial layer 200 and the second sacrificial layer 200 are removed.

[0170] See Figure 6m A semiconductor layer 800 is formed on the surface of the exposed gate stack structure 300', and the semiconductor layer 800 covers the bottom of the exposed channel layer 422, the first barrier layer 510, the second barrier layer 610 and the contact plug 700.

[0171] Furthermore, in some embodiments, the method for fabricating the memory further includes:

[0172] See Figure 6n A contact hole 810 is formed through the semiconductor layer 800, exposing the contact plug 700. Here, the contact hole 810 is used to form a through-silicon contact (TSC).

[0173] See Figure 6o A fourth dielectric layer 820 is formed, covering the surface of the semiconductor layer 800 away from the gate stack structure 300' and the sidewalls of the contact hole 810. Exemplarily, a fourth dielectric material layer can first be deposited covering the surface of the semiconductor layer 800 away from the gate stack structure 300', and the sidewalls and topwalls of the contact hole 810. Then, a portion of the fourth dielectric material layer covering the topwall of the contact hole 810 is removed by an etching process, forming a layer as shown in the image. Figure 6o The fourth dielectric layer 820 is shown. Here, the material of the fourth dielectric layer 820 includes silicon oxide.

[0174] See also Figure 6o A first opening 821 is formed in the fourth dielectric layer 820 to expose the semiconductor layer 800.

[0175] See Figure 6p This forms an interconnect layer 830 that fills the first opening 821 and covers the top wall of the fourth dielectric layer 820 and the contact hole 810. Exemplarily, the interconnect layer 830 may be made of a metal, such as aluminum, copper, tungsten, or cobalt.

[0176] Here, the portion of the interconnect layer 830 located within the contact hole 810 forms a through-silicon contact, and the portion of the fourth dielectric layer located within the contact hole 810 forms a spacer (SPA) between the through-silicon contact and the semiconductor layer 800.

[0177] In some embodiments, the interconnect layer 830 is used to form contact pads for bonding to a wafer on which peripheral circuitry is formed.

[0178] The memory fabrication method provided in this application involves forming a first barrier layer and a second barrier layer in a sacrificial layer, allowing dummy channel vias to penetrate the dummy gate stack structure and extend into the first barrier layer, and allowing gate spacers to penetrate the dummy gate stack structure and extend into the second barrier layer. Because the first barrier layer, the second barrier layer, the insulating layer, and the dummy gate layer have a high etch selectivity, the etch resistance at the bottom of the dummy channel via in the first barrier layer is increased, facilitating the formation of dummy channel vias of uniform depth and expanding the process window for forming dummy channel vias, thereby obtaining a highly uniform dummy channel structure. Similarly, the etch resistance at the bottom of the gate spacer in the second barrier layer is increased, facilitating the formation of gate spacers of uniform depth and expanding the process window for forming gate spacers, thereby obtaining a highly uniform array common source. Furthermore, the first barrier layer can also act as an etch stop layer to protect the dummy channel structure when removing the first and second sacrificial layers, and the second barrier layer can also act as an etch stop layer to protect the array common source when removing the first and second sacrificial layers.

[0179] Furthermore, in some embodiments, the method for fabricating the memory further includes:

[0180] After the first barrier layer is formed, a third barrier layer is formed covering the sacrificial layer and the first barrier layer; wherein the third barrier layer is located between the sacrificial layer and the dummy gate stack structure;

[0181] Forming a pseudo-channel via in a first barrier layer that penetrates the pseudo-gate stack structure and extends into the via includes: forming a pseudo-channel via that penetrates the pseudo-gate stack structure and the third barrier layer and extends into the first barrier layer.

[0182] Removing the substrate and sacrificial layer to expose the bottom of the charge storage structure and the first barrier layer includes: removing the substrate and sacrificial layer to expose the bottom of the charge storage structure, the first barrier layer, and the third barrier layer;

[0183] Forming a semiconductor layer covering the exposed channel layer and the first barrier layer includes: forming a semiconductor layer covering the exposed channel layer, the first barrier layer and the third barrier layer.

[0184] The material of the third barrier layer has a large etching selectivity ratio with that of the sacrificial layer, and the material of the third barrier layer also has a large etching selectivity ratio with that of the charge storage layer. Therefore, when removing the sacrificial layer and the charge storage layer, the third barrier layer can act as an etching stop layer to protect the first insulating layer and ensure the thickness consistency of the first insulating layer.

[0185] The material of the third barrier layer can be the same as that of the first and second barrier layers, or it can be different from the materials of the first and second barrier layers.

[0186] For example, the material of the third barrier layer includes at least one of polycrystalline silicon, doped polycrystalline silicon, or a metal, wherein the metal may include tungsten, copper, aluminum, or cobalt. Doped polycrystalline silicon may include phosphorus-doped polycrystalline silicon.

[0187] The following example uses doped polysilicon as the third barrier layer, combined with... Figures 7a to 7t The cross-sectional schematic diagram of the corresponding memory during the fabrication process provides a detailed description of another memory fabrication method provided in this application embodiment.

[0188] See Figure 7a A sacrificial layer 200 is sequentially formed on the substrate 100. Here, the substrate 100 is a silicon substrate 100, and the material of the sacrificial layer 200 may include silicon oxide or silicon oxynitride.

[0189] See Figure 7b Through-holes 500 and trenches 600 are formed through the sacrificial layer 200 to expose the substrate 100, wherein the position of the trenches 600 is different from the position of the through-holes 500.

[0190] See Figure 7c A barrier material layer is formed, covering the sacrificial layer 200 and filling the vias 500 and trenches 600. The barrier material layer is made of polycrystalline silicon.

[0191] See Figure 7d Chemical mechanical polishing is used to remove the barrier material layer on the surface of the sacrificial layer 200, forming a first barrier layer 510 that fills the through hole 500 and a second barrier layer 610 that fills the trench 600.

[0192] Referring to 7e, a third barrier layer 900 is formed covering the sacrificial layer 200, the first barrier layer 510, and the second barrier layer 610. The material of the third barrier layer 900 includes doped polycrystalline silicon.

[0193] See Figure 7f A dummy gate stack structure 300 is formed on the third barrier layer 900; and a plurality of channel holes 400 are formed that penetrate the dummy gate stack structure 300, the third barrier layer 900 and the sacrificial layer 200 and extend into the substrate 100. The dummy gate stack structure 300 includes an insulating layer 301 and a dummy gate layer 302 that are alternately stacked.

[0194] See Figure 7gA first oxide layer 910 is formed on the sidewalls of the substrate 100 exposed by the channel via 400 and the third barrier layer 900 using a thermal oxidation process. The thickness of the first oxide layer 910 on the sidewalls of the third barrier layer 900 is greater than the thickness of the first oxide layer 910 on the substrate 100. Here, the first oxide layer 910 comprises silicon oxide.

[0195] As described above, the third barrier layer 900 is made of doped polysilicon, while the substrate 100 is a silicon substrate. Therefore, in subsequent processes, when a silicon epitaxial layer is grown on the substrate 100, if the sidewalls of the third barrier layer 900 are exposed through the channel via 400, the silicon epitaxial layer will also grow radially along the channel via 400. This would prevent the formation of the charge storage structure 420 within the channel via 400 corresponding to the sacrificial layer 200 in subsequent processes. Therefore, in this embodiment, a first oxide layer 910 is formed on the sidewalls of the third barrier layer 900 so that no silicon epitaxial layer grows on the sidewalls of the third barrier layer 900 when the silicon epitaxial layer is subsequently grown on the substrate 100.

[0196] It is understandable that when the first oxide layer 910 is formed on the sidewall of the third barrier layer 900, the first oxide layer 910 will also grow on the substrate 100. Therefore, in this embodiment, the thickness of the first oxide layer 910 on the sidewall of the third barrier layer 900 is greater than the thickness of the first oxide layer 910 on the substrate 100. Thus, when the first oxide layer 910 on the substrate 100 is subsequently removed, although the first oxide layer 910 on the sidewall of the third barrier layer 900 of the same thickness will be removed simultaneously, the portion of the first oxide layer 910 on the sidewall of the third barrier layer 900 that is thicker than the first oxide layer 910 on the substrate 100 will be retained. Therefore, when an epitaxial layer is subsequently formed on the substrate 100, the first oxide layer 910 will be formed on the sidewall of the third barrier layer 900, preventing the formation of an epitaxial layer on the sidewall of the third barrier layer 900.

[0197] Furthermore, it can be understood that the material of the third barrier layer 900 includes doped polycrystalline silicon, while the substrate 100 is a silicon substrate, that is, the material of the substrate 100 is monocrystalline silicon. Since the chemical potential energy of doped polycrystalline silicon is higher than that of monocrystalline silicon, under the same process parameters and the same time, the thickness of silicon oxide formed on doped polycrystalline silicon by thermal oxidation process is greater than the thickness of silicon oxide formed on monocrystalline silicon. In other words, the thickness of silicon oxide formed on the sidewall of the third barrier layer 900 is greater than the thickness of silicon oxide formed on the substrate 100.

[0198] For example, the first oxide layer 910 can be formed using either a dry oxidation process or a wet oxidation process. In this embodiment, a wet oxidation process is used to form the first oxide layer 910, thereby greatly increasing the rate at which the first oxide layer 910 is formed.

[0199] See Figure 7hThe first oxide layer 910 on the substrate 100 is removed, while a portion of the first oxide layer 910 on the sidewall of the third barrier layer 900 is retained. Here, the removal of the first oxide layer 910 on the substrate 100 is to allow the formation of an epitaxial layer 410 on the substrate 100 via an epitaxial growth process in subsequent processes. The retention of the first oxide layer 910 on the sidewall of the third barrier layer 900 is to prevent the growth of epitaxial material on the sidewall of the third barrier layer 900, thereby enabling the formation of a charge storage structure 420 within the hole segment of the channel hole 400 corresponding to the sacrificial layer 200 in subsequent processes.

[0200] For example, the first oxide layer 910 on the substrate 100 can be removed using plasma dry etching, gas dry etching, and wet etching processes.

[0201] See further Figure 7h As can be seen, during the removal of the first oxide layer 910 on the substrate 100, the sidewalls of the insulating layer 301 within the channel via 400 are etched, resulting in grooves between adjacent dummy gate layers 302. This is detrimental to the formation of the charge storage layer 421. Therefore, see... Figure 7i The dummy gate layer 302 is etched onto the sidewall of the channel hole 400 so that the dummy gate layer 302 and the insulating layer 301 are substantially flush with the sidewall of the channel hole 400.

[0202] For example, the dummy gate layer 302 can be etched using plasma dry etching, gas dry etching, and wet etching processes.

[0203] It should be noted that this embodiment uses doped polysilicon as an example to illustrate the fabrication method of the memory. Since the third barrier layer is doped polysilicon, when a silicon epitaxial layer is formed on the substrate, a silicon epitaxial layer will form on the sidewalls of the third barrier layer exposed through the channel vias. Therefore, a... Figure 7g-7i The steps shown involve forming a first oxide layer along the radial direction of the channel hole on the sidewall of the third barrier layer, without forming the first oxide layer on the substrate, and ensuring that the sidewall surfaces of the channel hole are flush. When the material of the third barrier layer is metal, steps 7g-7i are not necessary because a silicon epitaxial layer is not formed on the metal when a silicon epitaxial layer is formed on the substrate.

[0204] See Figure 7j An epitaxial layer 410 is formed within multiple channel holes 400. The top surfaces of the epitaxial layer 410 are substantially flush within the multiple channel holes 400.

[0205] See Figure 7kA charge storage structure 420 is formed covering the sidewall of the channel hole 400 and the epitaxial layer 410. The charge storage structure 420 includes a charge storage layer 421 and a channel layer 422 arranged radially along the channel hole 400. The top surface of the epitaxial layer 410 is located in the hole segment corresponding to the sacrificial layer 200, so that the bottom of the charge storage structure 420 is located in the hole segment corresponding to the sacrificial layer 200.

[0206] In some embodiments, the method for fabricating the memory further includes: after forming a charge storage structure 420 within the channel via 400, forming a pseudo-channel via penetrating the pseudo-gate stack structure 300 and the third barrier layer 900 and extending into the first barrier layer 510, and forming a pseudo-channel structure 530 within the pseudo-channel via. Exemplarily, the pseudo-channel structure 530 may include a silicon oxide layer.

[0207] In some embodiments, the substrate 100 includes an array region and a peripheral circuit region. The array region includes a core region and a step region. The step region is located between the core region and the peripheral circuit region. The portion of the dummy gate stack structure 300 located in the step region has multiple steps. The via 400 is located in the core region, and the dummy via is located in the step region.

[0208] The methods for fabricating memory also include:

[0209] Remove part of the pseudo-gate stack structure 300 in the peripheral circuit area and retain the first insulating layer 3011 on the sacrificial layer 200;

[0210] A first dielectric layer 303 is formed, covering the step area and the peripheral circuit area;

[0211] A second dielectric layer 304 is formed covering the pseudo-gate stack structure 300 and the first dielectric layer 303.

[0212] See also Figure 7k After forming the pseudo-channel structure 530, a gate trench 620 is formed that penetrates the pseudo-gate stack structure 300 and the third barrier layer 900 and extends into the second barrier layer 610.

[0213] See Figure 7l A second oxide layer 920 is formed on the substrate within the gate trench 620 and the sidewalls of the third barrier layer 900 using a thermal oxidation process. Here, the second oxide layer 920 comprises silicon oxide.

[0214] When the dummy gate insulating layer 302 is removed through the gate spacer 620, or when the gate layer 302' is removed from the gate spacer 620 after the gate layer 302' has been deposited, the second oxide layer 900 can serve as an etch stop layer to protect the substrate 100 and the third barrier layer 900.

[0215] See Figure 7mThe dummy gate layer 302 is removed through the gate spacer 620, and the gate layer 302' is deposited in the gap formed after the removal of the dummy gate layer 302 to obtain the gate stack structure 300'. Then, an array common source 630 is formed in the gate spacer 620.

[0216] In some embodiments, see Figure 7n The method for fabricating the memory further includes forming a contact plug (CT) 700 in the peripheral circuit region, penetrating the second dielectric layer 304, the first dielectric layer 303, and the first insulating layer 3011, and extending into the third barrier layer 900. The contact plug 700 may be made of metal or polysilicon, and the metal includes, but is not limited to, tungsten, copper, aluminum, or cobalt.

[0217] See Figure 7o The substrate 100 is removed, exposing the sacrificial layer 200. In this embodiment, chemical mechanical polishing is used to remove the substrate 100, thereby simultaneously removing a portion of the epitaxial layer 410 in the substrate 100.

[0218] See Figure 7p The sacrificial layer 200 is removed by plasma dry etching process, exposing the bottom of the charge storage structure 420, the remaining epitaxial layer 410, the first barrier layer 510, the second barrier layer 610 and the third barrier layer 900.

[0219] The third barrier layer 900 has a high etching selectivity ratio with the sacrificial layer 200. Therefore, the third barrier layer 900 is used as an etching stop layer when the sacrificial layer 200 is removed to protect the first insulating layer 3011 and make the thickness of the first insulating layer 3011 more uniform.

[0220] See Figure 7q The charge storage layer 421 is removed using a gas dry etching process or a wet etching process, exposing the channel layer 422. It is understood that the gas dry etching process or the wet etching process is isotropic; therefore, it can remove all of the charge storage layer 421 at the bottom of the charge storage structure 420, causing the remaining epitaxial layer 410 to fall off along with the removal of the charge storage layer 421. This exposes more of the channel layer 422, increasing the contact area between the channel layer 422 and the subsequent semiconductor layer 800, thereby improving its storage performance.

[0221] In addition, the third barrier layer 900 has a high etching selectivity with the charge storage layer 421. Therefore, the third barrier layer 900 is also used as an etching stop layer when removing the charge storage layer 421 to protect the first insulating layer 3011 and make the thickness of the first insulating layer 3011 more uniform.

[0222] See Figure 7rA semiconductor layer 800 is formed on the third barrier layer 900, and the semiconductor layer 800 covers the exposed channel layer 422, the first barrier layer 510 and the second barrier layer 610.

[0223] In some embodiments, see Figure 7s The methods for fabricating memory also include:

[0224] A contact hole 810 is formed that penetrates the semiconductor layer 800 and the third barrier layer 900, and the contact hole 810 exposes the contact plug 700;

[0225] A fourth dielectric layer 820 is formed covering the surface of the semiconductor layer 800 away from the gate stack structure 300' and the sidewalls of the contact hole 810;

[0226] A first opening 821 is formed in the fourth dielectric layer 820 to expose the semiconductor layer 800.

[0227] Here, the diameter of the contact hole 810 is larger than the diameter of the contact plug 700, so that the diameter of the contact hole 810 after the fourth dielectric layer 820 is formed on the sidewall is basically equal to the diameter of the contact plug 700. After the interconnect layer is formed subsequently, the contact area between the interconnect layer and the contact plug 700 is increased, thereby reducing the contact resistance and improving the electrical connection performance.

[0228] See Figure 7t An interconnect layer 830 is formed that fills the first opening 821 and covers the fourth dielectric layer 820 and the top wall of the contact hole 810.

[0229] The memory fabrication method provided in this embodiment uses a third barrier layer as an etch stop layer during the removal of the sacrificial layer and the charge storage layer at the bottom of the charge storage structure. This protects the first insulating layer in the gate stack structure, resulting in better uniformity of the thickness of the first insulating layer in the final memory. This improves the breakdown voltage (BV) of the bottom select gate (BSG), preventing breakdown failure and ensuring good electrical and storage performance of the memory. Here, the bottom select gate is the gate layer closest to the semiconductor layer in the gate stack structure, i.e., the gate layer in contact with the first insulating layer.

[0230] Based on the memory fabrication method provided in the embodiments of this application above, the embodiments of this application also provide a memory, such as... Figure 6p and 7t The memory includes a semiconductor layer 800, a gate stack structure 300' located on the semiconductor layer 800, and a plurality of memory pillars. The gate stack structure 300' includes an insulating layer 301 and a gate layer 302' that are alternately stacked. The plurality of memory pillars are disposed in the gate stack structure 300' and extend to the bottom surface of the semiconductor layer 800, which is substantially flush with the bottom surface.

[0231] For example, the semiconductor layer 800 is made of polycrystalline silicon and / or doped polycrystalline silicon.

[0232] Furthermore, the storage column includes a charge storage layer 421 and a channel layer 422 arranged sequentially along its radial direction.

[0233] In some embodiments, such as Figure 6p The memory also includes a first barrier layer 510, a second barrier layer 610, a pseudo-channel structure 530, and an array common source 630. The first barrier layer 510 and the second barrier layer 610 are both disposed in the semiconductor layer 800, and the positions of the first barrier layer 510 and the second barrier layer 610 are different. The pseudo-channel structure 530 is disposed in the gate stack structure 300' and extends into the first barrier layer 510. The array common source 630 is disposed in the gate stack structure 300' and extends into the second barrier layer 610.

[0234] For example, the material of the first barrier layer 510 may include at least one of polycrystalline silicon, doped polycrystalline silicon, or a metal, wherein the metal may include tungsten, copper, aluminum, or cobalt. The material of the second barrier layer 610 may include at least one of polycrystalline silicon, doped polycrystalline silicon, or a metal, wherein the metal may include tungsten, copper, aluminum, or cobalt.

[0235] In some embodiments, such as Figure 7t The memory also includes a third barrier layer 900, which is located between the semiconductor layer 800 and the gate stack structure 300'. A charge storage layer 421 extends through the third barrier layer 900 to the top surface of the semiconductor layer 800; a channel layer 422 extends through the third barrier layer 900 into the semiconductor layer 800.

[0236] For example, the material of the third barrier layer 900 includes at least one of polycrystalline silicon, doped polycrystalline silicon, or metal, and the metal may include tungsten, copper, aluminum, or cobalt.

[0237] In the storage fabrication method, the third barrier layer serves as an etch stop layer when removing the sacrificial layer and the charge storage layer at the bottom of the charge storage structure. It can protect the first insulating layer in the gate stack structure 300', resulting in better uniformity of the thickness of the first insulating layer in the final memory. This improves the breakdown voltage (BV) of the bottom select gate (BSG), prevents breakdown failure, and ensures that the memory has good electrical and storage performance.

[0238] In some embodiments, such as Figure 7tThe memory also includes a first oxide layer 910, which is located between the sidewalls of the charge storage layer 421 and the third barrier layer 900. Specifically, the charge storage layer 421 includes a barrier layer 4211, a charge trapping layer 4212, and a tunneling layer 4213 sequentially disposed radially along the storage pillar. The first oxide layer 910 is located between the sidewalls of the barrier layer 4211 and the sidewalls of the third barrier layer 900. Here, the first oxide layer 910 includes silicon oxide.

[0239] In some embodiments, such as Figure 7t The memory also includes a second oxide layer 920, which is located between the array common source 630 and the second barrier layer 610, and between the array common source 630 and the sidewall 900 of the third barrier layer. Here, the second oxide layer 920 comprises silicon oxide.

[0240] This application provides specific embodiments, but its scope of protection is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for producing a memory, characterized by, include: A stacked structure is formed on a substrate; the stacked structure includes a sacrificial layer located between a pseudo gate stacked structure or a gate stacked structure and the substrate; A plurality of channel holes are formed that penetrate the stacked structure and extend into the substrate; wherein at least two of the plurality of channel holes have different depths; An epitaxial layer is formed within the plurality of channel holes; wherein the top surface of the epitaxial layer in the plurality of channel holes is substantially flush, and the top surface of the epitaxial layer is located in the hole segment corresponding to the sacrificial layer; A charge storage structure is formed covering the sidewalls of the channel aperture and the epitaxial layer; wherein the charge storage junction includes a charge storage layer and a channel layer arranged sequentially along the radial direction of the channel aperture; Remove the substrate and the sacrificial layer to expose part of the charge storage structure; Remove a portion of the charge storage layer of the charge storage structure to expose the channel layer; A semiconductor layer is formed to cover the exposed channel layer.

2. The method of claim 1, wherein the step of depositing the memory material is performed by atomic layer deposition. Each of the aforementioned vias includes a first via segment located within the substrate and a second via segment located within the stacked structure; the epitaxial layer includes a first epitaxial layer and a second epitaxial layer, wherein the first epitaxial layer and the second epitaxial layer are an integral structure; An epitaxial layer is formed within the plurality of channel holes; wherein the top surface of the epitaxial layer in the plurality of channel holes is substantially flush, including: A first epitaxial layer is formed in each of the first hole segments by an epitaxial growth process; wherein the first epitaxial layers located in different first hole segments are substantially flush with the surface of the stacked structure. Within each second hole segment, a second epitaxial layer of substantially the same height is formed on the first epitaxial layer through the epitaxial growth process, so that the top surfaces of the second epitaxial layers in the plurality of channel holes are substantially flush.

3. The method for fabricating a memory according to claim 1, characterized in that, The method of forming a stacked structure on a substrate includes: sequentially forming a sacrificial layer and a dummy gate stacked structure on the substrate; wherein the dummy gate stacked structure includes alternately stacked insulating layers and dummy gate layers; The top surface of the epitaxial layer is located in the hole segment corresponding to the sacrificial layer, so that the bottom of the charge storage structure is located in the hole segment corresponding to the sacrificial layer; After removing the substrate and the sacrificial layer, the bottom of the charge storage structure is exposed; Remove the charge storage layer at the bottom of the charge storage structure so that the bottom surface of the remaining charge storage layer is substantially flush with the bottom surface of the stacked structure.

4. The method of claim 1, wherein the step of depositing the memory material is performed by atomic layer deposition. The sacrificial layer includes a first sacrificial layer and a second sacrificial layer stacked together, wherein the first sacrificial layer is located between the substrate and the second sacrificial layer, and the second sacrificial layer is an etch stop layer when the first sacrificial layer is removed; Forming an epitaxial layer within the plurality of channel holes includes: forming the epitaxial layer by an epitaxial growth process; wherein the top surface of the epitaxial layer is located in the hole segment corresponding to the first sacrificial layer, so that the bottom of the charge storage structure is located in the hole segment corresponding to the first sacrificial layer.

5. The method for fabricating a memory according to claim 4, characterized in that, The material of the first sacrificial layer includes silicon oxide and / or silicon oxynitride; and / or, The material of the second sacrificial layer includes at least one of metal, silicon nitride, silicon oxide, polycrystalline silicon, or doped polycrystalline silicon.

6. The method for fabricating a memory according to claim 1, characterized in that, The preparation method further includes: The epitaxial layer is removed before a semiconductor layer covering the exposed channel layer is formed.

7. The method for fabricating a memory according to claim 1, characterized in that, The preparation method further includes: Before forming the dummy gate stack structure, a via is formed through the sacrificial layer; The vias are filled to form a first barrier layer; After the charge storage structure is formed, a pseudo-channel structure is formed that penetrates the pseudo-gate stack structure and extends into the first barrier layer. The step of removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure includes: removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure and the first barrier layer; The semiconductor layer forming the exposed channel layer includes: a semiconductor layer forming the exposed channel layer and the first barrier layer.

8. The method for fabricating a memory according to claim 7, characterized in that, The preparation method further includes: Before forming the dummy gate stack structure, a trench is formed through the sacrificial layer; wherein the location of the trench is different from the location of the via. Fill the trench to form a second barrier layer; After the charge storage structure is formed, a gate trench is formed that penetrates the pseudo-gate stack structure and extends into the second barrier layer; The dummy gate layer is replaced with a gate layer by the gate spacer to obtain a gate stack structure; An array of common sources is formed within the gate spacer.

9. The method for fabricating a memory according to claim 7, characterized in that, The preparation method further includes: A third barrier layer is formed covering the sacrificial layer and the first barrier layer; wherein the third barrier layer is located between the sacrificial layer and the dummy gate stack structure; The formation of a pseudo-channel structure within the first barrier layer that penetrates the pseudo-gate stack structure and extends into the via includes: forming a pseudo-channel structure that penetrates the pseudo-gate stack structure and the third barrier layer and extends into the first barrier layer. The step of removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure and the first barrier layer includes: removing the substrate and the sacrificial layer to expose the bottom of the charge storage structure, the first barrier layer, and the third barrier layer; The semiconductor layer forming the exposed channel layer and the first barrier layer includes: forming a semiconductor layer covering the exposed channel layer, the first barrier layer and the third barrier layer.

10. The method for fabricating a memory according to claim 9, characterized in that, The third barrier layer includes at least one of polycrystalline silicon, doped polycrystalline silicon, or a metal.

11. A memory, characterized in that, The memory is prepared by the method of fabrication according to any one of claims 1 to 10, wherein the memory comprises: A semiconductor layer, a gate stack structure located on the semiconductor layer, the gate stack structure comprising alternately stacked insulating layers and gate layers; and, Multiple memory pillars are disposed within the gate stack structure; wherein the multiple memory pillars extend to be substantially flush with the bottom surface of the semiconductor layer.

12. The memory according to claim 11, characterized in that, The memory also includes: A first barrier layer is disposed in the semiconductor layer; A pseudo-channel structure is disposed within the gate stack structure and extends into the first barrier layer.

13. The memory according to claim 12, characterized in that, The memory also includes: A second barrier layer is disposed in the semiconductor layer, and the positions of the first barrier layer and the second barrier layer are different; The array common source electrode is disposed in the gate stack structure and extends into the second barrier layer.

14. The memory according to claim 12 or 13, characterized in that, The storage column includes a charge storage layer and a channel layer arranged radially therefrom, and the memory further includes: A third barrier layer is located between the semiconductor layer and the gate stack structure, and the charge storage layer extends through the third barrier layer to the top surface of the semiconductor layer; the channel layer extends through the third barrier layer into the semiconductor layer.

Citation Information

Patent Citations

  • Vertical-type memory device

    CN110391245A

  • Three-dimensional memory device and manufacturing method thereof

    CN110676258A

  • Three-dimensional memory and manufacturing method thereof

    CN112885840A

  • Three-dimensional memory and preparation method thereof

    CN112951841A