Semiconductor devices and their fabrication methods
By employing multiple etching processes and chemical atomic layer removal technology, the problem of uneven thickness of internal spacers in nanostructure transistors was solved, improving the yield and reliability of nanostructure transistors and ensuring the stability of gate structure replacement.
Patent Information
- Application Number
- CN202512053186.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2025-10-29
- Filing Date
- 2025-12-31
- Publication Date
- 2026-11-17
AI Technical Summary
When forming internal spacers in nanostructure transistors, the etching process can easily lead to rounded corners, resulting in uneven thickness of the internal spacers. This increases the risk of etching through the spacers into the source/drain regions, affecting device yield and reliability.
A multi-etching process is employed, including a first etching operation to form cavities and a second etching operation to refine corners. Chemical atomic layer removal (CAR) etching tools are used to precisely control the gas flow, reduce corner rounding, create a sharper transition, and ensure uniform thickness of internal spacers.
This improves the yield and reliability of nanostructured transistors, reduces the possibility of etching through internal spacers into the source/drain regions, and enhances the stability of the gate structure replacement process.
Smart Images

Figure CN123124426A_ABST
Abstract
Description
Technical Field
[0001] This disclosure relates to semiconductor devices and methods of forming them. Background Technology
[0002] As semiconductor device manufacturing advances and technology processing node sizes shrink, transistors can become susceptible to short-channel effects (SCE), such as hot carrier degradation, barrier reduction, and quantum confinement. Furthermore, with smaller transistor gate lengths at smaller technology nodes, source / drain (S / D) electron tunneling increases, which amplifies the transistor's cutoff current (the current flowing through the transistor channel when the transistor is in cutoff configuration). Nanostructured transistors (e.g., nanowire transistors, nanosheet transistors, gate-all-around (GAA) transistors, multi-bridge channel transistors, nanoribbon transistors, and / or other types of nanostructured transistors) are potential candidates for overcoming short-channel effects at smaller technology nodes. Nanostructured transistors offer a highly efficient structure that reduces SCE and increases carrier mobility relative to other types of transistors by providing greater gate control over charge carrier flow through the nanostructured transistor channel. Summary of the Invention
[0003] According to one aspect of this disclosure, a method of forming a semiconductor device is provided, comprising: forming a first plurality of semiconductor layers and a second plurality of semiconductor layers, the second plurality of semiconductor layers alternating with the first plurality of semiconductor layers in a direction substantially perpendicular to a semiconductor substrate of the semiconductor device; etching the first plurality of semiconductor layers, the second plurality of semiconductor layers, and the semiconductor substrate to define a fin substrate and a layer stack above the fin substrate; etching the layer stack and the fin substrate to form a source / drain recess, wherein etching the layer stack and the fin substrate to form the source / drain recess results in the formation of a mesa structure extending above the fin substrate and a channel layer stack above the mesa structure, wherein the channel layer stack includes a semiconductor interposer formed from the first plurality of semiconductor layers and a nanostructured channel formed from the second plurality of semiconductor layers; etching the semiconductor interposer to remove the semiconductor interposer between the nanostructured channels of the channel layer stack; and etching along the side of the source / drain recess. The process involves depositing a dielectric layer in the walls and spaces formed between the nanostructure channels by removing a semiconductor interposer; performing a first etching operation to etch the dielectric layer, thereby forming a dielectric interposer at the end of the dielectric interposer, wherein the first etching operation is performed using approximately constant flow rates of a first process gas and a second process gas, and resulting in the formation of a cavity at the end of the dielectric interposer; performing a second etching operation to etch the end of the dielectric interposer to modify the shape of the inner surface of the cavity, wherein the second etching operation is performed using a first variable flow rate of the first process gas and a second variable flow rate of the second process gas; after performing the second etching operation, forming an internal spacer in the cavity; and forming a source / drain region in the source / drain recess such that the source / drain region contacts the internal spacer, and such that the source / drain region includes a plurality of epitaxial layers comprising the same semiconductor material but with different dopant concentrations.
[0004] According to one aspect of this disclosure, a method of forming a semiconductor device is provided, comprising: forming a layer stack over a fin substrate of the semiconductor device, the layer stack including a plurality of semiconductor interposers and a plurality of semiconductor channel layers, the plurality of semiconductor channel layers alternating with the plurality of semiconductor interposers in a first direction, wherein the fin substrate and the layer stack extend in a second direction; forming a shallow trench isolation (STI) region extending in the second direction adjacent to the fin substrate, wherein the STI region has a non-planar top surface along a third direction, the third direction being substantially perpendicular to the second direction; etching the layer stack and the fin substrate to form a source / drain recess, wherein etching the layer stack and the fin substrate to form the source / drain recess results in the formation of a mesa structure extending over the fin substrate and a channel layer stack over the mesa structure, wherein the channel layer stack includes portions of the plurality of semiconductor interposers and from the plurality of semiconductor interposers... The process involves: forming a nanostructured channel using a channel layer; etching a portion of the semiconductor interposer to remove a portion of the semiconductor interposer from between the stacked nanostructured channels; depositing a dielectric layer material along the sidewalls of the source / drain recesses and in the spaces formed between the nanostructured channels by removing the semiconductor interposer; performing a first etching operation to etch the dielectric layer to form a dielectric interposer and a cavity at the end of the dielectric interposer, wherein after the first etching operation, the inner surface of the cavity has a first corner rounding amount in the third direction; performing a second etching operation to etch the end of the dielectric interposer, wherein after the second etching operation, the inner surface of the cavity has a second corner rounding amount in the third direction, wherein the second corner rounding amount is less than the first corner rounding amount; and forming an internal spacer in the cavity after performing the second etching operation.
[0005] According to one aspect of this disclosure, a semiconductor device is provided, comprising: a fin substrate; a shallow trench isolation (STI) region extending in a first lateral direction adjacent to the fin substrate; a channel layer stack located above a mesa region of the fin substrate, including a plurality of nanostructured channels; a gate structure surrounding the plurality of nanostructured channels and extending in a second lateral direction above the STI region, the second lateral direction being substantially perpendicular to the first lateral direction, wherein the gate structure includes a gate dielectric layer and one or more metal layers, and wherein at least one of the one or more metal layers includes a titanium-containing material; and an internal spacer adjacent to an end of a gate portion of the gate structure located in a vertical direction between adjacent nanostructured channels of the plurality of nanostructured channels, wherein the internal spacer has a first width in the first lateral direction at approximately its center in the second lateral direction, wherein the internal spacer has a second width in the first lateral direction at its endpoint in the second lateral direction, and wherein the difference between the first width and the second width is in the range of about 0.1 nanometers to about 5 nanometers. Attached Figure Description
[0006] The various aspects of this disclosure can be best understood from the following detailed description when read in conjunction with the accompanying drawings. It should be noted that, in accordance with standard practice in the industry, the various features are not drawn to scale. In fact, for clarity of discussion, the dimensions of the various features may be arbitrarily enlarged or reduced.
[0007] Figure 1A-1C This is a diagram illustrating an example implementation of the fin definition process described in this article.
[0008] Figure 2A and 2B This is a diagram illustrating an example implementation of the occupancy gate formation process described in this article.
[0009] Figure 3A and 3B This is a diagram illustrating an example implementation of the source / drain recess formation process described in this article.
[0010] Figure 4A-4G This is a diagram illustrating an example implementation of the mediator replacement process described in this article.
[0011] Figures 5A-5D This is a diagram illustrating an example implementation of the intermediate trimming process described in this article.
[0012] Figures 6A-6D This is a diagram illustrating an example implementation of the internal spacer forming process described in this article.
[0013] Figure 7This is a diagram illustrating an example implementation of the source / drain region formation process described in this article.
[0014] Figure 8 This is a diagram illustrating an example implementation of the dielectric layer formation process described in this article.
[0015] Figures 9A-9C This is a diagram illustrating an example implementation of the gate replacement process described in this article.
[0016] Figure 10 This is a flowchart of an example process associated with forming the semiconductor device described herein.
[0017] Figure 11 This is a flowchart of an example process associated with forming the semiconductor device described herein. Detailed Implementation
[0018] The following disclosure provides numerous different embodiments or examples for implementing various features of the provided subject matter. To simplify this disclosure, specific examples of components and arrangements are described below. These are, of course, merely examples and not limiting. For example, in the following description, forming a first feature on or over a second feature may include embodiments where the first and second features are formed in direct contact, and may also include embodiments where additional features may be formed between the first and second features such that the first and second features may not be in direct contact. Furthermore, this disclosure may repeat reference numerals and / or letters in various examples. Such repetition is for simplicity and clarity and does not in itself indicate a relationship between the various embodiments and / or configurations discussed.
[0019] Furthermore, spatially related terms (e.g., "below," "below," "lower," "above," "higher," etc.) may be used herein to readily describe the relationship of one element or feature shown in the figure relative to another element(s) or feature(s). In addition to the orientations depicted in the figures, these spatially related terms are also intended to encompass different orientations of the device during use or operation. The device may be oriented in other directions (rotated 90 degrees or in other orientations), and the spatially related descriptors used herein may be interpreted similarly accordingly.
[0020] Nanostructured transistors can be formed from stacked nanosheets comprising alternating semiconductor layers and interposers of nanoscale thickness. A vacant gate structure can be formed on the nanosheet stack and used to etch the nanosheet stack to form source / drain recesses, in which the source / drain regions of the nanostructured transistor are formed by epitaxial growth. The vacant gate structure and interposers can then be removed and replaced with a metal gate structure for the nanostructured transistor. The use of the interposer allows the metal gate structure to fill the area surrounding the semiconductor layer (e.g., the channel layer of the nanostructured transistor), such that the metal gate structure completely surrounds the channel layer, enhancing gate control over the flow of charge carriers through the channel layer.
[0021] In some cases, multiple sets of interposers are formed and then removed during the formation of nanostructured transistors for semiconductor devices. For example, a set of semiconductor interposers (e.g., silicon-germanium (SiGe)) can be formed together with semiconductor layers (e.g., silicon (Si) layers, which will be etched to define the nanostructured channels of the nanostructured transistors) as a portion of a nanosheet stack. The semiconductor interposers define the vertical spacing between the semiconductor layers and enable the nanosheet stack to be formed by epitaxial growth, thereby realizing the crystalline structure of the semiconductor layers.
[0022] Before forming the source / drain regions of the nanostructured transistor, the semiconductor interposer can be subsequently removed and replaced with a dielectric interposer. The semiconductor material of the semiconductor interposer, subjected to the high temperatures during epitaxial deposition of the source / drain regions, may readily diffuse into the nanostructured channel. This diffusion from the semiconductor layer can lead to phase mixing between the semiconductor materials of the semiconductor interposer and the nanostructured channel, potentially degrading the electrical properties of the nanostructured channel. The dielectric material of the dielectric interposer is better able to withstand this high-temperature process degradation, allowing the electrical properties of the nanostructured channel to be maintained.
[0023] Some nanostructured transistors include internal spacers between the source / drain regions and the gate structure. Internal spacers can provide various process and / or performance advantages, such as electrical isolation between the source / drain regions and the gate structure and / or protection of the source / drain regions from etching during replacement gate operations that replace the sacrificial nanostructure layer with the gate structure. The process of forming internal spacers may include forming a cavity at the end of a dielectric interposer, filling the cavity with a dielectric material, and then trimming the dielectric material to form the internal spacer.
[0024] However, the process of forming internal spacers can be challenging and may lead to defects in the internal spacers, thereby reducing their ability to provide electrical isolation and / or protect the source / drain regions from etching. For example, when forming a cavity in which an internal spacer for a nanostructured transistor is to be formed, the ends of the dielectric interposer may be rounded, and this rounding may result in a reduction in the lateral thickness of the top and bottom of the internal spacer. When performing a gate replacement operation, this reduction in the lateral thickness of the internal spacer may result in a lower thickness etch buffer or etch stop, and etching may occur through the internal spacer and into the source / drain regions, damaging the source / drain regions and / or causing an electrical short circuit between the source / drain regions and the gate structure. This can lead to failure of the nanostructured transistor and may reduce the yield of nanostructured transistors on semiconductor devices.
[0025] In some embodiments described herein, internal spacers are formed in nanostructured transistors of semiconductor devices in a manner that reduces the likelihood of etching through the internal spacers and into adjacent source / drain regions. Specifically, a multiple etch process described herein is performed to form cavities in which the internal spacers are to be formed. The multiple etch process includes one or more first etch operations to form the cavity, and one or more second etch operations to refine the corners of the cavity to reduce corner rounding. Due to the performance of one or more second etch operations, the corners of the cavity have a sharper transition between the cavity sidewalls and the inner surface (e.g., increased orthogonality). This results in smaller lateral thickness variations and increased uniformity in the internal spacers subsequently formed in the cavity. The increased uniformity of the lateral thickness of the internal spacers reduces the likelihood of etching through any particular portion of the internal spacer during a replacement gate operation that replaces the sacrificial nanostructure layer with the gate structure of the nanostructured transistor. Thus, the techniques described herein can reduce the likelihood of nanostructured transistor failure and / or can improve the yield of nanostructured transistors formed on semiconductor devices.
[0026] One or more second etching operations may include the use of a chemical atomic layer removal (CAR) etching tool comprising multiple gas pulse units capable of precisely controlling the process gas flow in one or more second etching operations. The gas pulse units can pulse the process gas flow into the processing chamber of the CAR etching tool, allowing the corners of the cavity to be precisely trimmed to achieve a sharper transition (e.g., increased orthogonality) between the cavity sidewalls and inner surface.
[0027] Figures 1A-1C This is an illustration of an example implementation 100 of the fin-defined process described herein. Example implementation 100 includes an example of forming a fin structure and associated shallow trench isolation (STI) region for a semiconductor device 105 described herein. Figures 1A-1C Cross-sectional views of the semiconductor device 105 are shown in the y-direction.
[0028] Semiconductor device 105 may be fabricated to include a plurality of transistors. The transistors may include nanostructured transistors, such as nanowire transistors, nanosheet transistors, gate-all-around (GAA) transistors, multi-bridge channel transistors, nanoribbon transistors, and / or other types of nanostructured transistors. Example implementation 100 includes examples of fin structures and associated STI regions forming the nanostructured transistors of semiconductor device 105.
[0029] like Figure 1A As shown, the processing of semiconductor device 105 is performed in conjunction with semiconductor substrate 110. A nanostructure layer stack 115 may be formed or disposed on semiconductor substrate 110. The nanostructure layer stack 115 comprises multiple nanosheets or nanostructure layers, each nanosheet or nanostructure layer having a z-axis thickness on the order of nanometers (e.g., about 10 nanometers or less). On top of the nanostructure layer stack 115, a hard mask layer 120 may be formed and subsequently used for fin definition processes.
[0030] Semiconductor substrate 110 may include a silicon (Si) substrate, a substrate formed of a material including silicon, a III-V compound semiconductor material substrate such as gallium arsenide (GaAs), a silicon-on-insulator (SOI) substrate, a germanium (Ge) substrate, a silicon-germanium (SiGe) substrate, a silicon carbide (SiC) substrate, or another type of semiconductor substrate.
[0031] The nanostructure layer stack 115 may include alternating layers arranged in a direction generally perpendicular to the semiconductor substrate 110 (e.g., the z-direction). For example, the nanostructure layer stack 115 may include a plurality of interposer layers 125-1 to 125-3 alternating with semiconductor layers 130-1 to 130-3 in the z-direction. Figure 1A The number of interposers and semiconductor layers in the nanostructure layer stack 115 shown is an example, and other numbers of interposers and semiconductor layers are also within the scope of this disclosure.
[0032] Intermediate layers 125-1 to 125-3 may be included in the nanostructure layer stack 115 to define the vertical distance between vertically adjacent nanostructure channels formed from semiconductor layers 130-1 to 130-3 in the nanostructure layer stack 115. Intermediate layers 125-1 to 125-3 may also serve as placeholder layers for the gate structure of the nanostructure transistor of the subsequently formed semiconductor device 105, the gate structure being formed around the nanostructure channels formed from semiconductor layers 130-1 to 130-3.
[0033] Interposers 125-1 to 125-3 may each comprise a first material component, and semiconductor layers 130-1 to 130-3 may each comprise a second material component. The first and second material components may be different material components to provide etch selectivity between interposers 125-1 to 125-3 and semiconductor layers 130-1 to 130-3. For example, interposers 125-1 to 125-3 may each comprise silicon germanium (SiGe), and semiconductor layers 130-1 to 130-3 may each comprise silicon (Si). Other combinations of materials used for interposers 125-1 to 125-3 and semiconductor layers 130-1 to 130-3 are within the scope of this disclosure.
[0034] One or more types of deposition tools can be used to deposit and / or grow alternating layers of the nanostructured layer stack 115. For example, deposition tools can be used to grow interposers 125-1 to 125-3 and / or semiconductor layers 130-1 to 130-3 via epitaxial growth, which can include epitaxial techniques such as molecular beam epitaxy (MBE), metal-organic chemical vapor deposition (MOCVD) processes, and / or another suitable epitaxial technique. Additionally and / or alternatively, interposers 125-1 to 125-3 can be deposited by chemical vapor deposition (CVD), physical vapor deposition, atomic layer deposition (ALD), and / or another suitable deposition technique. Deposition tools can be used to deposit material for the hard mask layer 120 via CVD, PVD, ALD, and / or another suitable deposition technique. In some implementations, one or more layers of the nanostructured layer stack 115 and / or hard mask layer 120 can be planarized using a planarization tool such as a chemical mechanical planarization (CMP) tool.
[0035] like Figure 1B As shown, a portion of the nanostructure layer stack 115 and the semiconductor substrate 110 can be etched to remove some portions of the nanostructure layer stack 115 and the semiconductor substrate. This results in the formation of a fin structure 135 extending over the semiconductor substrate. The fin structure 135 can extend in the semiconductor device 105 along the x-direction and can be aligned in the semiconductor device 105 along the y-direction.
[0036] The fin structure 135 includes a portion of a nanostructure layer stack 115 extending in the x-direction and a portion of a semiconductor substrate 110 extending in the x-direction. The semiconductor substrate 110 portion may be referred to as the fin substrate 140 of the fin structure 135, and the nanostructure layer stack 115 portion is included on the fin substrate 140.
[0037] The fin structure 135 can be formed by patterning the hard mask layer 120 and etching the nanostructure layer stack 115 and / or semiconductor substrate 110 based on the pattern formed in the hard mask layer 120. The hard mask layer 120 can be patterned using photolithography techniques, including dual patterning or multiple patterning techniques. Additionally and / or alternatively, multiple hard mask layers 120 can be patterned and used to form the fin structure 135. Etching tools can be used to etch the nanostructure layer stack 115 and / or semiconductor substrate 110 based on the pattern using dry etching techniques (e.g., reactive ion etching), wet etching techniques, and / or combinations thereof.
[0038] like Figure 1C As shown, an STI liner 145 and an STI region 150 can be formed between adjacent fin substrates 140 of the fin structure 135. In some implementations, each of the STI regions 150 may comprise a dielectric oxide material. For example, the STI region 150 may comprise a dielectric oxide material with a dielectric constant ranging from about 3 to about 5. For example, the STI region 150 may comprise a dielectric material with a dielectric constant of about 3.9, such as silicon oxide (SiO2). x (e.g., SiO2). As another example, the STI region 150 may include a dielectric material with a dielectric constant less than about 3.9. Such a dielectric material may be referred to as a "low-k" dielectric material, and examples include undoped silicate glass (USG), borosilicate glass (BSG), fluorosilicate glass (FSG), and / or organosilicon glass (OSG). Additionally and / or alternatively, the STI region 150 may include another dielectric material, such as silicon nitride (SiO2). x N y STI pad 145 may include similar dielectric materials such as silicon oxynitride (SiON) and / or silicon oxynitride (SiON).
[0039] In some implementations, the STI region 150 is formed as a multilayer structure comprising multiple dielectric layers 150a-150c. The dielectric layers 150a-150c can be formed from different dielectric materials.
[0040] To form the STI pad 145 and the STI region 150, the material of the STI pad 145 can be conformally deposited using a deposition tool, and a volumetric dielectric layer can be deposited (e.g., by CVD, PVD, ALD, and / or another suitable deposition technique) such that the dielectric layer completely fills the space between the fin structures 135 and extends over the top of the fin structures 135. A planarization tool can then be used to perform a planarization or polishing operation (e.g., a CMP operation) to planarize the dielectric layer such that the top surface of the dielectric layer is substantially coplanar with the top surface of the hard mask layer 120. The hard mask layer 120 can be used as a CMP stop layer in the planarization operation.
[0041] The STI pad 145 and dielectric layer can then be etched down to and / or below the fin substrate 140 using an etching tool to remove the STI pad 145 from the nanostructure layer stack 115 and form the STI region 150. The dielectric layer can be etched back such that the top surface of the STI region 150 is substantially coplanar with or below the top surface of the fin substrate 140. The etching operation can include a wet etching operation using a wet chemical etchant, a dry etching operation using a gas-based etchant and / or plasma and / or another type of etchant.
[0042] In implementations that perform dry etching operations to etch the dielectric layer to define the STI region 150, various types of process gases can be used. For example, etchant gases such as difluoromethane (CH2F2), trifluoromethane (CHF3), carbon tetrafluoride (CF4), hydrogen fluoride (HF), ammonia (NH3), and / or another type of etchant gas can be used to etch the dielectric layer to define the STI region 150. In some implementations, additional gases can be used for passivation and / or dilution during the dry etching operation. For example, gases such as nitrogen (N2), oxygen (O2), and / or carbon dioxide (CO2) can be used for passivation to promote etch selectivity between the dielectric material of the dielectric layer and the materials of the individual layers in the nanostructure layer stack 115. As another example, gases such as helium (He), argon (Ar), and / or nitrogen (N2) can be used to dilute the etchant gas to control the etching rate of the dielectric layer. The total flow rate of the gas entering the processing chamber of the etching tool for dry etching operations can be included in the range of about 20 standard cubic centimeters per minute (sccm) to about 3000 sccm. However, other values and ranges are also within the scope of this disclosure.
[0043] In some implementations, the dry etching operation is plasma-assisted. Plasma can be generated or supplied to the processing chamber of the etching tool to control the directionality and / or flow of the etchant during the dry etching operation. Specifically, plasma can be used to achieve highly directional etching of the dielectric layer to suppress lateral etching of the fin structure 135. In some implementations, the plasma power used in the dry etching operation can be in the range of about 100 watts to about 2000 watts. However, other values and ranges are also within the scope of this disclosure. In some implementations, the pressure in the processing chamber used for the dry etching operation can be in the range of about 1 millitor to about 800 millitors. However, other values and ranges are also within the scope of this disclosure.
[0044] like Figure 1CAs further shown, the etching operation may result in the STI region 150 having a non-planar top surface. For example, the STI region 150 may have a top surface having a first side portion 155a adjacent to the first fin base 140 in the y-direction, a second side portion 155b adjacent to the other fin base 140 in the y-direction, and a central portion 155c laterally located between the first side portion 155a and the second side portion 155b in the y-direction, wherein the first side portion 155a and the second side portion 155b are higher than the central portion 155c. In other words, the top surface of the STI region 150 may be concave. The height difference (or vertical position difference) between the central portion 155c and the first side portion 155a (and the second side portion 155b) of the STI region 150 is... Figure 1C The height difference is denoted as dimension D1. In some implementations, this height difference is less than about 10 nanometers and is included in the range of about 7 to about 8 nanometers. However, other values and ranges are also within the scope of this disclosure.
[0045] In some implementations, the central portion 155c may be approximately equidistant from the first side portion 155a and the second side portion 155b in the y-direction. Alternatively, the central portion 155c may be closer to the first side portion 155a or the second side portion 155b in the y-direction.
[0046] like Figure 1C As further shown, the central portion 155c of the STI region 150 can be recessed below the top surface of the adjacent fin base 140. The height difference (or vertical position difference) between the central portion 155c of the STI region 150 and the top of the fin base 140 of the fin structure 135 (or the bottom surface of the lowest intermediate layer 125-1) is in Figure 1C The middle part is represented as dimension D2.
[0047] like Figure 1C As further shown, a hard mask layer 160 can be formed on the STI region 150. To form the hard mask layer 160, a dielectric layer can be deposited on the semiconductor device 105, including deposition on the STI region 150 between the fin structure 135 and the fin structure 135, and the dielectric layer is etched to remove a portion of the dielectric layer, such that the remaining portion corresponds to the hard mask layer 160 on the STI region 150. The dielectric layer can include one or more dielectric materials. In some implementations, the hard mask layer 160 can include a silicon-based dielectric material. For example, the dielectric layer can include a nitrogen-containing silicon-based dielectric material, such as silicon nitride (Si). x N yMaterials such as Si3N4, silicon oxynitride (SiON), silicon carbonitride (SiCN), and / or silicon carbonitride (SiOCN) are used. In some implementations, the dielectric layer includes another type of dielectric material. In some implementations, the dielectric constant of the dielectric material in the dielectric layer can be greater than the dielectric constant of the STI region 150.
[0048] Deposition tools can be used to deposit the material of the dielectric layer using PVD, ALD, CVD, and / or another suitable deposition technique. Etching tools can be used to perform an etching operation to etch the dielectric layer to define the hard mask layer 160. The etching operation can include wet etching, dry etching, plasma-assisted etching, and / or another type of etching operation. Some material of the remaining portion of the dielectric layer may be consumed during the etching operation. As a result, the vertical (z-direction) thickness of the hard mask layer 160 may be less than the deposition thickness of the dielectric layer portion on the STI region 150. In some implementations, the vertical thickness of the hard mask layer 160 may be included in the range of about 16 nanometers to about 17 nanometers. However, other values and ranges are also within the scope of this disclosure. The top surface of the hard mask layer 160 may be substantially coplanar with or below the top surface of the fin substrate 140 of the fin structure 135 (or the bottom surface of the bottommost intermediate layer 125-1). In some implementations, the hard mask layer 160 may have a non-linear top or a curved surface. In particular, the top surface of the hard mask layer 160 can protrude along the y-direction.
[0049] As mentioned above, Figures 1A-1C Provided as an example. Other examples may be related to... Figures 1A-1C The descriptions differ from those in the text.
[0050] Figure 2A and 2B This is an illustration of an example implementation 200 of the vacant gate formation process described herein. Example implementation 200 includes an example of forming a vacant gate structure 205 for a nanostructured transistor of a semiconductor device 105. In some implementations, the operations described in connection with example implementation 200 are combined with... Figures 1A-1C The process described is performed afterward.
[0051] Figure 2A A cross-sectional view of a semiconductor device 105 having a centrally located gate structure 205 is shown in the y-direction. Figure 2B A cross-sectional view of a semiconductor device 105 having a centrally located gate structure 205 is shown in the x-direction.
[0052] like Figure 2AAs shown, a vacant gate structure 205 (also referred to as a virtual gate structure) is formed on some portions of the fin structure 135 such that the vacant gate structure 205 surrounds the side and top surfaces of the fin structure 135 in the y-direction. The vacant gate structure 205 may also extend in the y-direction on top of the STI region 150 between adjacent fin structures 135. The vacant gate structure 205 may include a stop layer 210 below the polysilicon layer 215 and hard mask layers 220, 225 above the polysilicon layer 215.
[0053] The stop layer 210 may include silicon oxide (e.g., SiO2). x (e.g., SiO2), silicon nitride (e.g., Si) x N y The stop layer 210 may be included between the polysilicon layer 215 of the vacancy gate structure 205 and the fin structure 135, and between the polysilicon layer 215 and the STI region 150, to protect the fin structure 135 and the STI region 150 from etching during the formation process of the vacancy gate structure 205.
[0054] The polysilicon layer 215 comprises polysilicon (polysilicon or PO) or other materials. The polysilicon layer 215 can serve as a temporary structure during the formation of other layers and / or structures in the semiconductor device 105. The polysilicon layer 215 of the vacant gate structure 205 can subsequently be removed in a later processing stage of the semiconductor device 105 and replaced by an alternative gate structure (e.g., a metal gate structure). The vacant gate structure 205 can be used as a self-aligned pattern to define a source / drain (S / D) recess in the fin structure 135, in which the source / drain regions of the nanostructure transistor are formed. Using the vacant gate structure 205 to form and / or define other layers and / or structures of the semiconductor device 105 allows the polysilicon layer 215 of the vacant gate structure 205 to withstand damage from various processes and operations that would otherwise be suffered by the alternative gate structure. Thus, the alternative gate structure can be subsequently formed after such processes and operations, reducing the likelihood (and / or amount) of damage to the alternative gate structure.
[0055] Hard mask layers 220 and 225 can be patterned and used to etch polysilicon layer 215 to form a placeholder gate structure 205. Hard mask layer 220 may include a nitride-containing dielectric material, such as silicon nitride (Si). x N y Silicon carbonitride (SiCN), silicon oxynitride (SiON), and / or silicon carbonitride oxynitride (SiOCN), etc. The hard mask layer 225 may include an oxide-containing dielectric material, such as silicon oxide (SiO2). x )wait.
[0056] like Figure 2B As shown, two or more vacant gate structures 205 can be arranged along the fin structure 135 in the x-direction. Therefore, the fin structure 135 can extend in the x-direction and be arranged in the y-direction, the STI region 150 can extend in the x-direction next to and between adjacent fin structures 135, and the vacant gate structures 205 can extend in the y-direction and be arranged in the x-direction, such that the vacant gate structures 205 are substantially perpendicular to the fin structure 135 and the STI region 150.
[0057] like Figure 2B As further shown, spacer layers 230a and / or 230b may be formed on opposite sides of the vacant gate structure 205 and on top of the vacant gate structure 205. Spacer layers 230a and / or 230b may also extend on the top and sidewalls of the exposed portion of the fin structure 135. Spacer layer 230a may be formed on the sidewalls of the vacant gate structure 205, on the fin structure 135, and on top across the STI region 150. Spacer layer 230b may be formed on spacer layer 230a. In some implementations, spacer layers 230a and 230b may be formed of similar or different materials. Bulk spacer layers 230a and 230b may each include one or more dielectric materials, such as silicon nitride (Si). x N y Materials such as silicon carbonitride (SiCN), silicon oxynitride (SiON), and / or silicon carbonitride (SiOCN) can be deposited using CVD, ALD, PVD, and / or another suitable deposition technique.
[0058] As mentioned above, Figure 2A and Figure 2B Provided as an example. Other examples may be related to... Figure 2A and Figure 2B The descriptions differ from those in the text.
[0059] Figure 3A and Figure 3B This is an illustration of an example implementation 300 of the source / drain recess formation process described herein. Example implementation 300 includes an example of forming source / drain recesses, where the source / drain regions of the nanostructure transistor of semiconductor device 105 are formed in the source / drain recesses. In some implementations, the operations described in conjunction with example implementation 300 are combined with... Figures 1A-1C , Figure 2A and / or Figure 2B The operation described is executed afterward.
[0060] Figure 3AA perspective view of a semiconductor device 105 having an active / drain recess 305 is shown. Figure 3B It shows Figure 3A A cross-sectional view of the semiconductor device 105 along line AA, which runs along fin structure 135 and crosses multiple occupied gate structures 205 in the x direction.
[0061] like Figure 3A and Figure 3B As shown, the nanostructure layer stack 115 and the fin substrate 140 of the underlying fin structure 135 can be etched to form a source / drain recess 305, which extends through the nanostructure layer stack 115 and into the underlying fin substrate 140. The source / drain recess 305 can be formed to be adjacent to the opposite side of the occupied gate structure 205 in the x-direction.
[0062] The etching operation can be performed using etching tools and may be referred to as a strained source / drain (SSD) etching operation. In some implementations, the etching operation includes the use of plasma etching, wet chemical etching, and / or another type of etching technique. Portions through the spacer layers 230a, 230b can be etched to expose the top portion of the fin structure 135, and portions of the nanostructure layer stack 115 between the occupier gate structures 205 can be etched to form source / drain recesses 305.
[0063] like Figure 3A and Figure 3B As further shown, etching the fin substrate 140 into the fin structure 135 to form source / drain recesses 305 results in the formation of a channel layer stack 310 from the nanostructure layer stack 115 and a mesa structure 315 in the fin substrate 140. The mesa structure 315 (also referred to as a base) corresponds to the portion of the fin substrate 140 located below the channel layer stack 310 on the fin structure 135, which is defined by etching the nanostructure layer stack 115 to define the source / drain recesses 305. The channel layer stack 310 may include interposers 320-1 to 320-3 alternating with nanostructure channels 325-1 to 325-3 above the mesa structure 315. Interposer 320-1 may correspond to a portion of interposer 125-1, interposer 320-2 may correspond to a portion of interposer 125-2, and interposer 320-3 may correspond to a portion of interposer 125-3. Similarly, nanostructured channel 325-1 may correspond to a portion of semiconductor layer 130-1, nanostructured channel 325-2 may correspond to a portion of semiconductor layer 130-2, and nanostructured channel 325-3 may correspond to a portion of semiconductor layer 130-3.
[0064] The vacant gate structure 205 can be used as a self-aligned mask to form the channel layer stack 310. As a result, the ends of the nanostructured channels 325-1 to 325-3 in the x-direction (and the ends of the intermediate layers 320-1 to 320-3) can be aligned with the outer surfaces of the spacer layers 230a and 230b on the sidewalls of the vacant gate structure 205.
[0065] As mentioned above, Figure 3A and Figure 3B Provided as an example. Other examples may be related to... Figure 3A and Figure 3B The descriptions differ from those in the text.
[0066] Figures 4A-4G This is an illustration of an example implementation 400 of the interposer replacement process described herein. Example implementation 400 includes removing interposers 320-1 to 320-3 (e.g., silicon-germanium interposers) of the replacement channel layer stack 310 and replacing them with a dielectric interposer. In some implementations, the operations described in conjunction with example implementation 400 are combined with... Figures 1A-1C , Figure 2A , Figure 2B , Figure 3A and / or Figure 3B The operation described is executed afterward.
[0067] Intermediate layers 320-1 to 320-3 can be replaced with dielectric intermediates because dielectric intermediates are less likely to degrade surrounding layers and / or structures during subsequent high-temperature processes. For example, the semiconductor material of intermediates 320-1 to 320-3 (e.g., silicon-germanium (SiGe)) may readily diffuse into the nanostructured channels 325-1 to 325-3 at high temperatures. This phase mixing may degrade the electrical properties of the nanostructured channels 325-1 to 325-3. Therefore, intermediates 320-1 to 320-3 can be replaced with dielectric intermediates prior to high-temperature processes such as epitaxial deposition of source / drain regions in the source / drain recesses 305. The dielectric material of the dielectric intermediates is better able to withstand material degradation during such high-temperature processes.
[0068] Figure 4A A perspective view of a semiconductor device 105 with the interposers 320-1 to 320-3 removed is shown. Figure 4B It shows Figure 4A A cross-sectional view of the semiconductor device 105 along line AA, which runs along fin structure 135 and crosses multiple occupied gate structures 205 in the x direction.
[0069] like Figure 4A and Figure 4BAs shown, intermediate layers 320-1 to 320-3 can be removed from the channel layer stack 310. Removal of intermediate layers 320-1 to 320-3 results in the formation of spaces 405 in the vertical direction between the nanostructured channels 325-1 to 325-3 of the channel layer stack 310. An etching tool can be used to perform an etching operation (e.g., wet etching, dry etching) using an etchant that selectively etches the material of the intermediate layers 320-1 to 320-3 while minimizing etching of the material of the nanostructured channels 325-1 to 325-3.
[0070] Figure 4C It shows Figure 4A Another cross-sectional view of semiconductor device 105 along line AA. (See diagram below.) Figure 4C As shown, a dielectric layer 410 can be deposited in the space 405 left by removing interposers 320-1 to 320-3. The dielectric layer 410 can also be blanket-deposited on the spacer layers 230a, 230b on the sidewalls of the occupying gate structure 205 and / or on the sidewalls and bottom surface of the source / drain recess 305. The dielectric layer 410 may include a dielectric material, such as silicon oxide (SiO2). x ) and / or silicon nitride (Si x N y The deposition tool can be used to deposit the material of the dielectric layer 410 using PVD, ALD, CVD, and / or another suitable deposition technique.
[0071] In some implementations, a wet cleaning operation may then be performed to clean the dielectric layer 410. In some implementations, a high-temperature sulfuric acid peroxide mixture (HTSPM) is used in the wet cleaning operation. The wet cleaning operation may also promote the reflow of the dielectric material in the dielectric layer 410 to improve the convergence and filling of gaps and / or other voids in the dielectric layer 410 and / or space 405.
[0072] like Figure 4D As shown, a semiconductor device 105 can be processed by performing one or more first etch operations using a semiconductor processing tool 415 to etch the dielectric layer 410 back and forth to remove portions of the dielectric layer 410 from the spacer layers 230a, 230b on the sidewalls of the occupying gate structure 205 and the sidewalls and bottom surface of the source / drain recess 305. Figure 4E and 4F As shown, some portions of the dielectric layer 410 can be retained in the space 405 between the nanostructured channels 325-1 to 325-3 of the channel layer stack 310 as a dielectric interlayer.
[0073] like Figure 4DAs shown, the semiconductor processing tool 415 may include a processing chamber 420 and a chuck 425 supported by a base 430 within the processing chamber 420, on which the semiconductor device 105 is placed. The semiconductor processing tool 415 may be configured to perform chemical oxide removal (COR), which is performed to etch the oxide material of the dielectric layer 410.
[0074] To perform a COR operation (e.g., one or more first etch operations) to etch the dielectric layer 410, process gases 1 and 2 can be supplied to the processing chamber 420 through gas inlet 435. Process gases 1 and 2 can be supplied from one or more gas sources 440 through gas inlet 435 and are supplied to the processing chamber 420 through nozzles 445, which facilitates the distribution of process gases 1 and 2. In some implementations, process gases 1 and 2 are supplied to the processing chamber 420 separately and mixed within the processing chamber 420 to induce a reaction between them. In some implementations, process gases 1 and 2 are premixed, and the mixture is supplied to the processing chamber 420.
[0075] Processing chamber 420 can be evacuated to a low vacuum of approximately 50 mTorr to approximately 500 mTorr to facilitate the removal of contaminants and to promote the flow of process gases 1 and 2 within processing chamber 420. Process gases 1 and 2 can be supplied to processing chamber 420 at approximately constant flow rates similar to each other. In some implementations, gas 1 includes a hydrogen-containing gas, such as ammonia (NH3), and gas 2 includes a fluorine-containing gas, such as hydrofluoric acid (HF). The hydrogen in gas 1 and the fluorine in gas 2 are adsorbed onto the surface of semiconductor device 105 by physical adsorption (through weak van der Waals forces). Desorption and surface diffusion occur, resulting in the adsorption of hydrogen from silicon oxide (SiO2) onto the dielectric layer 410. x The material undergoes a chemical adsorption surface reaction, forming byproducts. These byproducts are discharged through vacuum line 450, thereby removing the material from dielectric layer 420. For example, hydrogen in gas 1 and fluorine in gas 2 react with silicon oxide (SiO2) in dielectric layer 410. x The material reaction is as follows: Ammonia and hydrofluoric acid react with silicon dioxide to form silicon tetrafluoride (SiF4) and water vapor (H2O) as byproduct gases. Ammonia and hydrofluoric acid continue to react with silicon tetrafluoride byproducts to form fluorosilicates, such as ammonium fluorosilicate ((NH4)2SiF6 or AFS).
[0076] like Figure 4E and 4FAs shown, the remaining material of the dielectric layer 410 may correspond to the interposers (e.g., dielectric interposers) 455-1 to 455-3. Interposer 455-1 may be located vertically between the mesa structure 315 above the fin substrate 140 and the bottom nanostructure channel 325-1 in the channel layer stack 310 above the mesa structure 315. Interposer 455-2 may be located vertically between the bottom nanostructure channel 325-1 and the middle nanostructure channel 325-2 in the channel layer stack 310. Interposer 455-3 may be located vertically between the middle nanostructure channel 325-2 and the top nanostructure channel 325-3 in the channel layer stack 310.
[0077] like Figure 4E and Figure 4F As further shown, one or more first etch operations may cause the ends of the interposers 455-1 to 455-3 to be etched laterally, thereby forming cavities 460 at the ends of the interposers 455-1 to 455-3. The cavities 460 may correspond to the space defined by the recessed interposers 455-1 to 455-3 and the adjacent nanostructure channels in the vertical direction.
[0078] like Figure 4G As shown in detailed cross-sectional view 465, after one or more first etching operations, cavity 460 may have a curved or arcuate inner surface 470 in the z-direction along the thickness of the end of interposer layer 455. Specifically, inner surface 470 may be concave, such that the center point of inner surface 470 extends inwardly into cavity 460 from a base 475 corresponding to the top and bottom ends of inner surface 470. Base 475 corresponds to the arcuate bottom of arc of inner surface 470. After one or more first etching operations, cavity has a dimension D3, which corresponds to a first distance between approximately the center point of arc of inner surface 470 and approximately the center point of base 475 of arc. After one or more first etching operations, inner surface 470 has a first radius of curvature. The first radius of curvature R corresponds to the curvature of inner surface 470 (…). κ The reciprocal of (e.g.) ).
[0079] The top and bottom corners of cavity 460 correspond to the transition between the inner surface 470 and the sidewalls of cavity 460, where the sidewalls of cavity 460 are the top and bottom surfaces of cavity 460. After one or more first etching operations, the cavity has a dimension D4 that corresponds to a first angle of the corner of cavity 460 (e.g., a first angle of the transition between the inner surface 470 and the z-direction plane in semiconductor device 105). In some implementations, dimension D4 is included in the range of about 40 degrees to about 45 degrees. However, other values in this range are also within the scope of this disclosure.
[0080] Figure 4G Detailed top view 480 shows a top view of a cross-section of the interposer layer 455 in the xy plane. As shown in detailed top view 480, after one or more first etch operations, cavities 460 are formed in opposite ends of the interposer layer 455 in the x-direction, resulting in a distance between approximately the center points 485 of the cavities 460 in the y-direction (in Figure 4G The dimension D5 is smaller than the distance between the lateral endpoints 490 on opposite sides of the cavity 460 in the y-direction. Figure 4G The dimensions are represented as D6 and D7. The rounding of the inner surface 470 of the cavity 460 can correspond to half the average of the difference between the dimensions D6 and D5 and the difference between the dimensions D7 and D5 (e.g., rounding = [AVG(D6-D5) & (D7-D5)] / 2).
[0081] As mentioned above, Figures 4A-4G Provided as an example. Other examples may be related to... Figures 4A-4G The descriptions differ from those in the text.
[0082] Figures 5A-5D This is an illustration of an example implementation 500 of the intermediate trimming process described herein. Example implementation 500 includes trimming the ends of intermediate layers 455-1 to 455-3 (e.g., dielectric intermediate layers) of the channel layer stack 310 to reduce the amount of corner rounding at the ends of intermediate layers 455-1 to 455-3, preparing for the formation of internal spacers. In some implementations, the operations described in conjunction with example implementation 500 are combined with... Figures 1A-1C , Figure 2A , Figure 2B , Figure 3A , Figure 3B and / or Figures 4A-4G The operation described is executed afterward.
[0083] like Figure 5A and 5B As shown, one or more second etching operations can be performed using a semiconductor processing tool 505 to trim the ends of the interposers 455-1 to 455-3 to flatten or reduce the amount of corner rounding (vertical corner rounding and lateral corner rounding) of the inner surface 470 of the cavity 460 formed at the ends of the interposers 455-1 to 455-3 to process the semiconductor device 105. A portion of the dielectric layer 410 can be retained as a dielectric interposer in the space 405 between the nanostructure channels 325-1 to 325-3 of the channel layer stack 310.
[0084] like Figure 5AAs shown, the semiconductor processing tool 505 may include multiple processing chambers 510a and 510b, each comprising a chuck 515a supported by a base 520a and a chuck 515b supported by a base 520b, on which the semiconductor device 105 can be processed. The processing chambers 510a and 510b may also include gas inlets 525a and 525b, nozzles 530a and 530b, and vacuum lines 535a and 535b, respectively.
[0085] Processing chamber 510a can be a chemical processing system for semiconductor processing tool 505, and processing chamber 510b can be a thermal processing system for semiconductor processing tool 505. Environmental conditions such as pressure, temperature, humidity, and / or gas composition can be independently controlled for the chemical processing of semiconductor device 105 in processing chamber 510a and the thermal processing of semiconductor device 105 in processing chamber 510b. Semiconductor device 105 can be transferred between processing chambers 510b and 510a through an inner door 540 between processing chambers 510a and 510b.
[0086] Semiconductor processing tool 505 can be configured to perform a chemical atomic layer removal (CAR) process (e.g., one or more second etch operations) to etch the oxide material at the ends of the interposers 455-1 to 455-3. The CAR process can be similar to... Figures 4A-4G The described COR process involves etching the oxide material of the dielectric layer 410. However, in the CAR process, a dual-chamber semiconductor processing tool 505 can be used to enhance process control, combining gas pulse technology to finely control the etching profile in the CAR process and minimize over-etching during the CAR process. Essentially, the CAR process is capable of removing one atomic layer of material from the interposers 455-1 to 455-3 at a time, similar to (but in reverse) the deposition of atomic layers using ALD.
[0087] To perform a CAR operation (e.g., one or more second etch operations) to etch interposers 455-1 to 455-3, the semiconductor device 105 may be placed on a chuck 515a in a processing chamber 510a (e.g., a chemical processing system), and process gases 1 and 2 may be supplied to the processing chamber 510a through a gas inlet 525a. In some implementations, gas 1 comprises ammonia (NH3), and gas 2 comprises a fluorine-containing gas, such as hydrofluoric acid (HF). In some implementations, process gas 3 may be used as a substitute for gas 1 or in addition to gas 1. Gas 3 may include a methyl-based gas. For example, gas 3 may include a methyl acid. In some implementations, the methyl-based gas is methyl methacrylate (MMA) gas and / or N-methylpyrrolidone (NMP) gas, and gas 3 is included to improve the cleaning of residues and / or byproducts in the CAR process.
[0088] like Figure 5A As further shown, a gas pulse unit 545a can be used to pulse the gas 1 / gas 3 flow into the processing chamber 510a, and an air pulse unit 545b can be used to pulse the gas 2 flow into the processing chamber 510b. The gas pulse units 545a and 545b can independently control the flow rate, pulse rate, duty cycle, and / or other parameters of gas 1 / gas 3 and gas 2.
[0089] Gas 1 / Gas 3 and Gas 2 can be pulsed in overlapping pulse 550, wherein the pulse of Gas 2 overlaps with a portion of the pulse of Gas 1 / Gas 3. In other words, in pulse 550, the duration of the pulse of Gas 1 / Gas 3 can be longer than the duration of the pulse of Gas 2. In pulse 550, Gas 1 / Gas 3 and Gas 2 are pulsed as follows: Figure 5A The flow rates shown are supplied to processing chamber 510a. Between pulses 550, the flows of gas 1 / gas 3 and gas 2 are cut off and not supplied to processing chamber 510a. During pulse 550, the flow rates of gas 1 / gas 3 can be approximately uniform and constant (e.g., in a monopole square wave mode), while the flow rate of gas 2 increases first during pulse 550 and then decreases after reaching a peak flow rate. In other words, the flow rate of gas 2 during pulse 550 can have a non-uniform flow rate similar to a monopole triangular wave.
[0090] In some implementations, the pulse of gas 2 in pulse 550 may occur at the end of the pulse of gas 1 / gas 3, so that the processing chamber 510a can be filled with gas 1 / gas 3 before the flow of gas 2 is supplied, thereby promoting the reaction between gas 1 / gas 3 and gas 2. In some implementations, the duration of the pulse of gas 2 is included in the range of about 0.5 seconds to about 2 seconds. In some implementations, the duration of the pulse of gas 1 / gas 3 may be about 2 to about 4 times the duration of the pulse of gas 2. However, other values and ranges are also within the scope of this disclosure.
[0091] The flow rate of gas 2 in pulse 550 can be higher than the flow rate of gas 1 / gas 3 in pulse 550 to achieve the mixture of gas 1 / gas 3 and gas 2 required for the reaction with the surface of semiconductor device 105. For example, the flow rate of gas 2 in pulse 550 can be about 2 to about 4 times the flow rate of gas 1 / gas 3 in the pulse to provide a sufficient amount of fluorine to effectively etch the interposers 455-1 to 455-3. However, other values and ranges are also within the scope of this disclosure. Furthermore, the flow rate of gas 2 in pulse 550 in the CAR process can be... Figures 4A-4GThe flow rate of gas 2 in the COR process is approximately 2 to 4 times higher to provide sufficient fluorine for efficient etching of the interposers 455-1 to 455-3 in the CAR process. However, other values and ranges are also within the scope of this disclosure.
[0092] Process gases 1 / 3 and 2 can be supplied from gas pulse units 545a and 545b respectively through gas inlet 525a and enter processing chamber 510b through nozzle 530a in processing chamber 510a, which facilitates the distribution of process gases 1 / 3 and 2. Processing chamber 510a can be evacuated to a vacuum level higher than that of gas 3 / 2. Figures 4A-4G The vacuum generated in the COR process is weak. In other words, the pressure in the processing chamber 510a for the CAR process can be higher than the pressure in the processing chamber 420 for the COR process. For example, the ratio of the pressure in the processing chamber 510a for the CAR process to the pressure in the processing chamber 420 for the COR process can be included in the range of about 4:1 to about 10:1. In some implementations, the pressure in the processing chamber 510a for the CAR process is included in the range of about 500 mTorr to about 2 Torr. The higher pressure in the processing chamber 510a for the CAR process, for example in the range of about 500 mTorr to about 2 Torr, promotes an increase in the adsorption of hydrofluoric acid gas molecules on the surface of the semiconductor substrate 110. This can enhance the etch selectivity between different oxide layers of the semiconductor device 105, for example selectively promoting adsorption to the interposers 455-1 to 455-3 and suppressing adsorption to the STI region 150 and other oxide-based dielectric layers and / or structures (such as stop layer 210). Furthermore, the increased adsorption can promote the smoothing of the contours of the intermediate layers 455-1 to 455-3, which can flatten the ends of the intermediate layers 455-1 to 45.5-3.
[0093] Molecules of gases 1 / 3 and 2 are adsorbed onto the surface of semiconductor device 105 via physical adsorption (physical adsorption via weak van der Waals forces). Semiconductor device 105 can then be transferred from processing chamber 510a to processing chamber 510b via inner door 540 for heat treatment, which may include increasing the temperature in processing chamber 510b to induce desorption and surface diffusion. This results in the adsorption of silicon oxide (SiO2) onto the interposers 455-1 to 455-3. x The material undergoes a chemical adsorption surface reaction to form byproducts, which are then discharged through vacuum line 535b, thereby removing the material from the ends of the intermediate layers 455-1 to 455-3. For example, hydrogen in gas 1 and fluorine in gas 2 react with silicon oxide (SiO2) in the intermediate layers 455-1 to 455-3. x The material reaction is as follows: where ammonia and hydrofluoric acid react with silicon dioxide to form silicon tetrafluoride (SiF4) byproduct gas and water vapor (H2O) byproduct gas. Ammonia and hydrofluoric acid continue to react with the silicon tetrafluoride byproduct to form fluorosilicates, such as ammonium fluorosilicate ((NH4)2SiF6 or AFS).
[0094] In some implementations, the chemical operation in processing chamber 510a (e.g., using the gas pulse technique described above) and the thermal operation in processing chamber 510b may be performed for multiple cycles (e.g., 5 cycles, 10 cycles) to remove material from the ends of the intermediate layers 455-1 to 455-3 atom by atom, thereby achieving a high degree of control over the profile shape of the ends of the intermediate layers 455-1 to 455-3. To initiate another cycle, the semiconductor device 105 may be transferred back from processing chamber 510b to processing chamber 510a for chemical processing.
[0095] as Figure 5B shown, the CAR process (e.g., one or more second etching operations) results in the ends of the intermediate layers 455-1 to 455-3 having a flatter vertical profile along the thickness of the intermediate layer 455 (as Figure 5B shown in the detailed cross-sectional view 555), and having a flatter lateral profile between the opposite sides of the intermediate layers 455-1 to 455-3 along the y direction (as Figure 5B shown in the detailed top view 560).
[0096] As shown in the detailed cross-sectional view 555, after the CAR process (e.g., after one or more second etching operations), the cavity 460 in the end of the intermediate layer 455 between adjacent nanostructure channels 325 in the z direction in the vertical direction has a dimension D8, which corresponds to the second distance between the approximate center point of the arc of the inner surface 470 and the approximate center point of the base 475 of the arc. Since the CAR process achieves a reduction in the cross-sectional curvature of the inner surface 470, the second pitch after one or more second etching processes is smaller than the first pitch (e.g., D8<D3). After one or more second etching operations, the inner surface 470 has a second radius of curvature. Due to the reduction of the cross-sectional curvature of the inner surface 470 achieved by one or more second etching operations ( κ ), after the CAR process, the second radius of curvature R of the inner surface 470 is greater than the first radius of curvature of the inner surface 470. In other words, the inner surface 470 is flatter along the z direction after the CAR process.
[0097] As further shown in the detailed cross-sectional view 555, after the CAR process, the cavity 460 has a dimension D9 that corresponds to a second angle of a corner of the cavity 460 (for example, a second angle of the transition between the inner surface 470 and the z-direction plane in the semiconductor device 105). Since the CAR process achieves a reduction in the cross-sectional curvature of the inner surface 470, the second angle of the corner of the cavity 460 after the CAR process is smaller than the first angle of the corner of the cavity (for example, D9 < D4). In some implementations, the dimension D9 is less than about 10 degrees. However, other values of the dimension D9 are also within the scope of the present disclosure.
[0098] As shown in the detailed top view 560, after one or more second etching operations, the distance between the approximate center points 485 of the cavities 460 in the opposite ends of the interposer 455 in the x-direction (in Figure 5B denoted as dimension D10) may be less than the distance between the end points 490 on the opposite sides of the interposer 455 in the y-direction after the CAR process. However, after the COR process and before the CAR process, the difference between the dimension D10 and the dimensions D11 and D12 is smaller than the difference between the dimension D5 and the dimensions D6 and D7. Thus, after the CAR process, the rounding of the inner surface 470 of the cavity 460 is less than the rounding of the inner surface 470 of the cavity 460 before the CAR process. In some implementations, due to the CAR process, the difference between the shortest distance between the cavities 460 in the opposite ends of the interposer 455 in the x-direction (which may be the distance across the interposer 454 between the approximate center points 485 of the cavities 460) and the longest distance between the cavities 460 in the x-direction (which may be the distance between the end points 490 on one side of the interposer 445) may be less than about 5 nanometers, down to about 0.1 nanometers. After the CAR process, the rounding of the inner surface 470 of the cavity 460 may correspond to half of the average of the difference between dimension D11 and dimension D10 and the difference between D12 and dimension D10 (for example, rounding=[AVG(D11-D10) & (D12-D10)] / 2).
[0099] Figure 5C A detailed cross-sectional view 565 and a detailed top view 570 of another implementation for performing the CAR process to form a substantially planar profile for the inner surface 470 of the cavity 460 and form substantially rectangular vertical (z-direction) and lateral (y-direction) profiles for the ends of the interposers 455-1 to 455-3 are shown, as Figure 5CDetailed views are shown in the diagram. In this implementation, the angle of the corner of cavity 460 can be approximately 0 degrees, and the distance between the approximate center point of the arc of inner surface 470 and the approximate center point of the base 475 of the arc of cavity 460 can also be approximately 0 nm. After the CAR process, the distance between the approximate center points 485 of cavity 460 in opposite ends of interposer 455 in the x-direction (dimension D10) and the spacing between the endpoints 490 of opposite sides of interposer 455 in the y-direction (dimensions D11 and D12) can be approximately equal.
[0100] Figure 5D A detailed top view 575 is shown of an alternative implementation in which the CAR process results in a notch being formed on the side of the cavity 460. In this implementation, after the CAR process, the distance (dimension D10) between the approximate center points 485 of the cavities 460 at opposite ends of the interposer layer 455 in the x-direction can be greater than the distance between the endpoints 490 of opposite sides of the interposer layer 455 in the y-direction.
[0101] As mentioned above, Figures 5A-5D Provided as an example. Other examples may be related to... Figures 5A-5D The descriptions differ from those in the text.
[0102] Figures 6A-6D This is an illustration of an example implementation 600 of the internal spacer formation process described herein. Example implementation 600 includes an example of forming an internal spacer in a cavity 460 at the ends of interposers 455-1 to 455-3 (e.g., dielectric interposers) of a channel layer stack 310. In some implementations, the operations described in conjunction with example implementation 500 are performed in conjunction with... Figures 1A-1C , Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figures 4A-4G and / or Figures 5A-5D The operation described is executed afterward.
[0103] like Figure 6A As shown, to form the internal spacers, a deposition tool can be used to deposit material for the dielectric layer 605 in the cavity 460 and along the sidewalls and bottom surface of the source / drain recess 305. The material of the dielectric layer 605 may include silicon nitride (Si). x N y ), silicon dioxide (SiO) x Silicon oxynitride (SiON), silicon oxycarbonate (SiOC), silicon carbonitride (SiCN), silicon carbonitride (SiOCN), and / or another dielectric material. CVD, PVD, ALD, and / or another deposition technique can be used to deposit the dielectric layer 605.
[0104] Figure 6B A perspective view of a semiconductor device 105 is shown after the dielectric layer 605 has been trimmed to form internal spacers 610-1 to 610-3 in the cavity 460. Figure 6C It shows Figure 6B A cross-sectional view of semiconductor device 105 along line AA, which is in the x-direction of semiconductor device 105.
[0105] like Figure 6B and 6C As shown, an etching tool is then used to remove excess material from the source / drain recess 305 of the dielectric layer 605, so that the remaining portion corresponds to the internal spacers 610-1 to 610-3 formed in the cavity 460 at the ends of the intermediate layers 455-1 to 455-3.
[0106] like Figure 6D A detailed cross-sectional view in the x-direction is shown in Figure 615, due to the combination Figures 5A-5D The described CAR process, which adjusts the contour of the cavity 460 forming the internal spacers 610-1 to 610-3, may have one or more cross-sectional dimensions in the x-direction, with intermediate layers 455-1 to 455-3 and / or internal spacers 610-1 to 610-3. Intermediate layer 455 between vertically adjacent nanostructure channels in the channel layer stack 310 may have a cross-sectional width in the x-direction corresponding to dimension D13 (e.g., at approximately the center of the portion of intermediate layer 455 between adjacent internal spacers 610 in the x-direction). In some implementations, dimension D13 is included in the range of about 12 nanometers to about 18 nanometers. However, other values in this range are also within the scope of this disclosure. The internal spacers 610 adjacent to intermediate layer 455 may have cross-sectional widths in the x-direction corresponding to dimension D14 at approximately the center of internal spacer 510, dimension D15 at the bottom of internal spacer 410, and dimension D16 at the top of internal spacer 710. In some implementations, the difference between dimensions D14 and D15, and the difference between dimensions D14 and D16, can each be less than about 5 nanometers, and in some implementations, are included in the range of about 0.5 nanometers to about 1.5 nanometers, thereby achieving high uniformity in the lateral thickness of the internal spacer 610. This makes it less likely that etching will pass through the internal spacer 610 (e.g., at the top or bottom of the internal spacer 610). However, other values and ranges are also within the scope of this disclosure.
[0107] like Figure 6D A detailed top view of the xy plane is shown in 620, due to the combination Figures 5A-5DThe described CAR process adjusts the contour of the cavity 460 forming the internal spacers 610-1 to 610-3, which may have one or more top-view dimensions in the x-direction. The internal spacer 610 adjacent to the interposer 455 may have a cross-sectional width in the x-direction between an inner surface 625 facing the interposer 455 and an outer surface 630 facing the source / drain recess 305. The cross-sectional width in the x-direction may include a cross-sectional width corresponding to dimension D17 at approximately the center 635 of the internal spacer 610, a cross-sectional width corresponding to dimension D18 at a first side of the internal spacer 610 in the y-direction, and a cross-sectional width corresponding to dimension D19 at a second side of the internal spacer 610 opposite to the first side in the y-direction. In some implementations, the difference between dimensions D17 and D18, and the difference between dimensions D17 and D19, can each be less than about 5 nanometers, and in some embodiments, are included in the range of about 0.5 nanometers to about 1.5 nanometers, thereby enabling high uniformity of the lateral thickness of the internal spacer 610. This makes it possible to achieve low etching probability through the internal spacer 610 (e.g., at the top or bottom of the internal spacer 610), which can improve gate control of the nanostructured channels 325-1 to 325-3, and / or enable low gate-to-source / drain contact distance defect rates. However, other values and ranges are also within the scope of this disclosure.
[0108] As mentioned above, Figures 6A-6D Provided as an example. Other examples may be related to... Figures 6A-6D The descriptions differ from those in the text.
[0109] Figure 7 This is an illustration of an example implementation 700 of the source / drain region formation process described herein. Example implementation 700 includes an example of forming the source / drain region 705 of a nanostructured transistor of semiconductor device 105 in a source / drain recess 305. In some implementations, the operations described in conjunction with example implementation 700 are combined with... Figures 1A-1C , Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figures 4A-4G , Figures 5A-5D and / or Figures 6A-6D The process described is performed afterward.
[0110] like Figure 7As shown, the material of buffer 710 can be deposited at the bottom of source / drain recess 305. In some implementations, buffer 710 (sometimes referred to as "L0") comprises undoped semiconductor material, such as undoped silicon (Si) and / or undoped silicon germanium (SiGe), and deposition tools can be used to epitaxially grow the semiconductor material of buffer 710.
[0111] like Figure 7 As further shown, the material of the isolation layer 715 may be deposited on and / or over the buffer 710 in the source / drain recess 305. The isolation layer 715 (sometimes referred to as a flexible bottom isolation layer) may comprise a silicon-containing dielectric material, such as silicon oxide (SiO2). x Such as SiO2), silicon nitride (Si x N y Silicon materials such as Si3N4, silicon carbonitride (SiCN), silicon oxynitride (SiO2), and / or silicon carbonitride (SiOCN) can be deposited using CVD, PVD, ALD, and / or another suitable deposition technique.
[0112] like Figure 7 As further shown, source / drain regions 705 may be formed on an isolation layer 715 in the source / drain recess 305. Depending on the context, "source / drain region" may be referred to individually or collectively as source or drain. In some implementations, source / drain regions 705 may be included as being adjacent to the opposite side of the occupied gate structure 205, such that the nanostructured channels 325-1 to 325-3 of the channel layer stack 310 beneath the occupied gate structure 205 extend between and are electrically coupled to the source / drain regions 705.
[0113] like Figure 7 As further shown, the source / drain region 705 may include a multilayer structure. For example, a deposition tool may be used to deposit a first epitaxial layer 720 (sometimes referred to as "L1") of the source / drain region 705 on the ends of the nanostructure channels 325-1 to 325-3. As another example, the deposition tool may deposit a second epitaxial layer 725 (sometimes referred to as "L2") of the source / drain region 705 in the source / drain recess 305. The second epitaxial layer 725 may fill the remaining region in the source / drain recess 305 and may contact the first epitaxial layer 720 and the internal spacers 610-1 to 610-3 still exposed in the source / drain recess 305. In some implementations, the second epitaxial layer 725 includes multiple layers, such as layer 725a (e.g., layer L2-1) and another layer 725b (e.g., layer L2-2), which have the same semiconductor material but different dopant concentrations. In some implementations, the top surface of the second epitaxial layer 725 can be higher than the top of the top nanostructure channel 325-3.
[0114] The first epitaxial layer 720 and the second epitaxial layer 725 of the source / drain region 705 may each comprise a semiconductor material, such as silicon (Si), silicon germanium (SiGe), silicon arsenide (SiAs), silicon phosphide (SiP), and / or another semiconductor material. The first epitaxial layer 720 and the second epitaxial layer 725 may each be doped with one or more dopants, such as arsenic (As), phosphorus (P), and / or boron (B), and other examples. The first epitaxial layer 720 and the second epitaxial layer 725 of the source / drain region 705 may each be epitaxially grown, deposited (e.g., using CVD, PVD, ALD), and / or formed using one or more other deposition techniques. In some implementations, the dopant concentration in the first epitaxial layer 720 and the dopant concentration in the second epitaxial layer 725 may be different.
[0115] For a p-type metal-oxide-semiconductor (PMOS) nanostructure transistor, the first epitaxial layer 720 and the second epitaxial layer 725 of the source / drain region 705 may each comprise boron (B)-doped silicon germanium (SiGe). The germanium (Ge) concentration of the second epitaxial layer 725 may be greater than that of the first epitaxial layer 720. For example, the germanium (Ge) concentration of the second epitaxial layer 725 may be in the range of about 40% to about 60%, while the germanium (Ge) concentration of the first epitaxial layer 720 may be in the range of about 10% to about 20%. However, other values and ranges are also within the scope of this disclosure. The boron (B) dopant concentrations of the second epitaxial layer 725 and the first epitaxial layer 720 may each be in the range of about 5 × 10⁻⁶. 20 cm -3 Up to approximately 5×10 21 cm -3 However, other values and ranges are also within the scope of this disclosure.
[0116] For an n-type metal-oxide-semiconductor (NMOS) nanostructure transistor, the first epitaxial layer 720 and the second epitaxial layer 725 of the source / drain region 705 may each comprise silicon (Si) doped with arsenic (As) and / or phosphorus (P), among other examples. The dopant concentration of the second epitaxial layer 725 may be greater than that of the first epitaxial layer 720. For example, the dopant concentration of the second epitaxial layer 725 may be included in approximately 2 × 10⁻⁶. 21 cm -3 Up to approximately 9×10 21 cm -3 Within the range, the dopant concentration of the first epitaxial layer 720 can be included in approximately 1 × 10⁻⁶. 20 cm -3 Up to approximately 1×10 21 cm -3 However, other values and ranges are also within the scope of this disclosure.
[0117] A capping layer 730 may be formed on the source / drain region 705 and may comprise silicon, silicon-germanium, doped silicon, doped silicon-germanium, and / or another material. The capping layer 730 may be included to reduce dopant diffusion and protect the underlying source / drain region 705 during semiconductor processing operations of the semiconductor device 105 prior to contact formation. Furthermore, the capping layer 730 may facilitate the formation of a metal-semiconductor alloy (e.g., silicide). In some implementations, the capping layer 730 is omitted.
[0118] As mentioned above, Figure 7 Provided as an example. Other examples may be related to... Figure 7 The descriptions differ from those in the text.
[0119] Figure 8 This is an illustration of an example implementation 800 of the dielectric layer formation process described herein. Example implementation 800 includes an example of forming one or more dielectric layers over the source / drain regions 705 of a nanostructured transistor in a semiconductor device 105. In some implementations, the operations described in connection with example implementation 800 are combined with... Figures 1A-1C , Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figures 4A-4G , Figures 5A-5D , Figures 6A-6D and / or Figure 7 The process described is performed afterward.
[0120] like Figure 8 As shown, a contact etch stop layer (CESL) 805 may be conformally deposited (e.g., using a deposition tool) on the source / drain regions 705 and the sidewalls of the occupant gate structure 205. An interlayer dielectric (ILD) layer 810 may be formed on the CESL 805 such that the ILD layer 810 fills the space above the source / drain regions 705 between the occupant gate structures 205. The CESL 805 may provide a mechanism to stop the etching process when forming contacts or vias for the source / drain regions 705, and the ILD layer 810 may provide electrical isolation between the contacts and the metal gate structure to replace the occupant gate structure 205. The source / drain regions 705 may be located below the ILD layer 810 such that the source / drain regions 705 are laterally adjacent to the ends of the nanostructured channels 325-1 to 325-3 in the x-direction.
[0121] CESL 805 may be formed of a dielectric material different from that of ILD layer 810 and may have different etch selectivity than ILD layer 810. The thickness of CESL 805 along the sidewall of the occupancy gate structure 205 (e.g., in the x-direction) is less than the thickness of ILD layer 810 in the x-direction.
[0122] CESL 805 may include or may be nitrogen-containing materials, silicon-containing materials, and / or carbon-containing materials. Furthermore, CESL 805 may include or may be silicon nitride (Si). x N y Silicon carbonitride (SiCN), carbon nitride (CN), silicon oxynitride (SiON), silicon oxycarbonate (SiCO), or combinations thereof, and other examples. CESL 805 can be deposited using deposition processes such as ALD, CVD, or another deposition technique.
[0123] The ILD layer 810 may include oxides (e.g., silicon oxide (SiO2)). x The dielectric material can be an undoped silicate glass (USG), a borosilicate glass (BSG), a fluorosilicate glass (FSG), tetraethyl orthosilicate (TEOS), a porous hydrogen silsesquioxane (HSQ), and / or another suitable dielectric material. In some implementations, the ILD layer 810 comprises an extremely low dielectric (ELK) dielectric material having a dielectric constant less than about 2.5. Examples of ELK dielectric materials include carbon-doped silicon oxide (C-SiO₂). x amorphous fluorinated carbon (aC) x F y Poly(p-xylene), dibenzocyclobutene (BCB), polytetrafluoroethylene (PTFE), silicon carbide (SiOC) polymers, porous hydrogen silsesquioxane (HSQ), porous methyl silsesquioxane (MSQ), porous polyarylene ether (PAE), and / or porous silica (SiOx), among other examples. The ILD layer 510 can be deposited using a deposition process such as ALD, CVD, or another deposition technique.
[0124] CESL 805 and ILD layer 810 may be planarized and / or etched to remove CESL 805 and ILD layer 810 from the top of the placeholder gate structure 205. ILD layer 810 may also be etched to form a cavity on top of ILD layer 810. Cap layer 815 is deposited in the cavity on top of ILD layer 810. Cap layer 815 may be planarized and / or etched to remove excess material from above the placeholder gate structure 205. Cap layer 815 may include a dielectric material, such as silicon nitride (Si). x N ySilicon carbonitride (SiCN), carbon nitride (CN), silicon oxynitride (SiON), silicon carbon oxynitride (SiCO), or combinations thereof, and other examples. The capping layer 815 may be included as a sacrificial layer for subsequently planarizing the metal gate structure formed as a replacement gate for the occupier gate structure 205.
[0125] As mentioned above, Figure 8 Provided as an example. Other examples may be related to... Figure 8 The descriptions differ from those in the text.
[0126] Figures 9A-9C This is an illustration of an example implementation 900 of the gate replacement process described herein. Example implementation 900 includes an example of replacing the occupied gate structure 205 with a metal gate structure of a nanostructured transistor of semiconductor device 105. In some implementations, the operations described in connection with example implementation 900 are in conjunction with... Figures 1A-1C , Figure 2A , Figure 2B , Figure 3A , Figure 3B , Figures 4A-4G , Figures 5A-5D , Figures 6A-6D , Figure 7 and / or Figure 8 The process described is performed afterward.
[0127] like Figure 9A As shown, the gate replacement process includes a gate removal operation. The gate removal operation includes removing the polysilicon layer 215 and stop layer 210 of the occupier gate structure 205 from the semiconductor device 105, leaving spacer layers 230a and 230b. The removal of the occupier gate structure 205 leaves an opening (or recess) in the ILD layer 810 and provides access to the interposers 455-1 to 455-3 of the underlying channel layer stack 310. The occupier gate structure 205 may be removed in one or more etching operations. Such etching operations may include plasma etching, wet chemical etching, and / or another type of etching technique.
[0128] like Figure 9B As shown, the gate replacement process includes an interposer release operation. The interposer release operation is performed to remove the interposers 455-1 to 445-3 of the channel layer stack 310. This results in the formation of an opening between the nanostructured channels 325-1 to 325-3 of the channel layer stack 310 (e.g., the region surrounding the nanostructured channels 325-1 to 325-3).
[0129] Intermediate layers 455-1 to 455-3 can be removed through the space previously occupied by the occupier gate structure 205. Specifically, removing the occupier gate structure 205 exposes the lateral sides of the intermediate layers 455-1 to 455-3 in the y-direction, so that these lateral sides of the intermediate layers 455-1 to 455-3 can be etched laterally.
[0130] The interposer release operation may include using an etching tool to perform an etching operation to remove the interposers 455-1 to 455-3 based on the difference in etch selectivity between the materials of the interposers 455-1 to 455-3 and the materials of the nanostructured channels 325-1 to 325-3, and between the materials of the interposers 455-1 to 455-3 and the materials of the internal spacers 610-1 to 610-3. The internal spacers 610-1 to 610-3 may be used as etch stop layers in this etching operation to protect the source / drain regions 705 from etching.
[0131] like Figure 9C As shown, the gate replacement operation includes forming a gate structure (e.g., a replacement gate structure or a metal gate structure) 905 in the openings between the source / drain regions 705 and between the internal spacers 610-1 to 610-3, and in the spaces previously occupied by the respective occupier gate structures 205 above the channel layer stack 310. Specifically, the gate structure 905 fills the regions between and around the nanostructure channels 325-1 to 325-3 previously occupied by the interposers 455-1 to 455-3, such that the gate structure 905 completely surrounds and encloses each set of nanostructure channels 325-1 to 325-3. This increases the control of the gate structure 905 over the conductivity of the nanostructure channels 325-1 to 325-3, increases the drive current of the nanostructure transistors of the semiconductor device 105, and / or reduces the short-channel effect (SCE) of the nanostructure transistor(s) of the semiconductor device 104.
[0132] like Figure 9C As further shown, the gate structure 905 may include gate portions 910-1 to 910-3 alternating with the nanostructured channels 325-1 to 325-3 above the mesa structure 315. The bottom gate portion 910-1 may be positioned vertically between the mesa structure 315 and the bottom nanostructured channel 325-1 of the channel layer stack 310. The middle gate portion 910-2 may be positioned vertically between the bottom nanostructured channel 325-1 and the middle nanostructured channel 325-2 of the channel layer stack 310. The top gate portion 910-3 may be positioned vertically between the middle nanostructured channel 325-2 and the top nanostructured channel 325-3 of the channel layer stack 310.
[0133] The gate structure 905 may further include a top portion 915 formed in the space previously occupied by the occupier gate structure 205 between portions of the CESL 805 and the ILD layer 810 above the channel layer stack 310. Spacer layers 230a, 230b may be positioned laterally between the top portion 915 and portions of the CESL 805 and ILD layer 810. The top portion 915 of the gate structure 905 may be planarized, which may result in the removal of the capping layer 815.
[0134] like Figure 9C As further shown, the gate structure 905 may include multiple layers. For example, the gate structure 905 may include a gate dielectric layer 920 and one or more metal layers 925a, 925b.
[0135] The gate dielectric layer 920 may include one or more high-k materials (e.g., dielectric materials with a dielectric constant greater than that of silicon dioxide (SiO2 – dielectric constant approximately 3.9)). Examples include lanthanum oxide (La). x O y Such as La2O3), hafnium oxide (H f O x , such as H f O2), zirconium oxide (ZrO) x (e.g., ZrO2) and / or aluminum oxide (Al) x O y Materials such as Al2O3, etc. The dielectric constant of the gate dielectric layer 920 can be greater than the dielectric constant of the dielectric materials of adjacent layers (such as spacer layers 230a, 230b and / or ILD layer 810). In other words, the dielectric constant of the materials of spacer layers 230a, 230b can be less than the dielectric constant of the material of gate dielectric layer 920. In some implementations, the thickness of gate dielectric layer 920 is less than the thickness of spacer layer 230a and / or the thickness of spacer layer 230b.
[0136] One or more metal layers 925a, 925b may include one or more work function metal layers 925a and / or a gate electrode layer 925b, wherein the work function metal layers 925a are located between the gate dielectric layer 920 and the gate electrode layer 925b. In some implementations, the gate dielectric layer 920 and the work function metal layers 925a fill the space between the nanostructure channels 325-1 to 325-3, such that the gate portions 910-1 to 910-3 include the gate dielectric layer 920 and the work function metal layers 925a, and such that the gate electrode layer 925b is omitted from the gate portions 910-1 to 910-3. In some implementations, the gate portions 910-1 to 910-3 further include the gate electrode layer 925b.
[0137] One or more work function metal layers 925a may include one or more n-type metal materials and / or one or more p-type metal materials. A PMOS nanostructure transistor may include a gate structure comprising one or more work function metal layers 925a containing a p-type metal material. The p-type metal material may include one or more p-type metals, such as tungsten (W), cobalt (Co), titanium nitride (TiN), tungsten nitride (WN), and / or another metal with a work function greater than about 4.7 eV. An NMOS nanostructure transistor may include a gate structure comprising one or more work function metal layers 925a containing an n-type metal material. The n-type metal material may include one or more n-type metals, such as titanium aluminum (TiAl), titanium aluminum carbon (TiAlC), and / or another type of n-type metal. A gate electrode layer 925b may include one or more metal materials, such as titanium (Ti), titanium-containing materials (e.g., titanium nitride (TiN)), ruthenium (Ru), tungsten (W), cobalt (Co), copper (Cu), and / or molybdenum (Mo).
[0138] like Figure 9C As shown in the detailed cross-sectional diagram 930, the gate portion 910 of the gate structure 905 located between vertically adjacent nanostructure channels 325 in the channel layer stack 310 can have cross-sectional widths in the x-direction corresponding to dimensions D20 (approximately the center of the gate portion 910 between opposing internal spacers 610 of the gate structure 905 in the x-direction), D21 (the bottom of the gate portion 910 of the gate structure 905), and D22 (the top of the gate portion 910 of the gate structure 905). In some implementations, the difference between dimensions D20 and D21, and the difference between dimensions D20 and D22, can be less than about 5 nanometers, as low as about 0.1 nanometers, which is combined with Figures 5A-5DThe described CAR process results in planarizing the inner surface 470 of the cavity 460 of the internal spacer 610. For example, the difference between dimensions D20 and D21 and / or the difference between dimensions D20 and D22 can be included in the range of about 0.5 nm to about 1.5 nm. However, other values and ranges are also within the scope of this disclosure. The difference between dimensions D20 and D21 and / or the difference between dimensions D20 and D22 can be less than the z-direction thickness of the nanostructure channel 325, which can be included in the range of about 3 nm to about 10 nm. However, other values in this range are also within the scope of this disclosure. In some implementations, dimensions D20, D21, and D22 range from about 12 nm to 18 nm. However, other values in this range are also within the scope of this disclosure. In some implementations, the ratio of dimensions D20 to D21 and / or the ratio of dimensions D20 to D22 is included in the range of about 1.03:1 to about 1.12:1. However, other values in this range are also within the scope of this disclosure.
[0139] like Figure 9C As shown in the detailed top view 935, the gate portion 910 of the gate structure 905, located between adjacent nanostructure channels 325 in the vertical direction within the channel layer stack 310, can have cross-sectional widths in the x-direction corresponding to dimensions D23 (approximately the center point 940 of the gate portion 910 in the y-direction between opposite sides of the gate portion 910), D24 (at the first side 945a of the gate portion 910 in the y-direction), and D25 (at the second side 945b of the gate portion 910 opposite to the first side in the y-direction). In some implementations, the difference between dimensions D23 and D24, and the difference between dimensions D23 and D24, can be less than about 5 nanometers, as low as about 0.1 nanometers, which is combined with Figures 5A-5DThe CAR process described above is a result of flattening the inner surface 470 of the cavity 460 of the internal spacer 610. For example, the difference between dimensions D23 and D24 and / or the difference between dimensions D23 and D25 may be included in the range of about 0.5 nm to about 1.5 nm. However, other values and ranges are also within the scope of this disclosure. The difference between dimensions D23 and D24 and / or the difference between dimensions D23 and D25 may be less than the z-direction thickness of the nanostructure channel 325, which may be included in the range of about 3 nm to about 10 nm. However, other values in this range are also within the scope of this disclosure. In some implementations, dimensions D23, D24, and D25 range from about 12 nm to 18 nm. However, other values in this range are also within the scope of this disclosure. In some implementations, the ratio of dimensions D23 to D24 and / or the ratio of dimensions D23 to D25 is included in the range of about 1.03:1 to about 1.12:1. However, other values in this range are also within the scope of this disclosure. The rounding of the end of the gate portion 910 of the gate structure 905 may correspond to half the average of the difference between size D24 and size D23 and the difference between D25 and size D23 (e.g., rounding = [AVG(D24-D23) & (D25-D23)] / 2).
[0140] As mentioned above, Figures 9A-9C As an example. Other examples may be related to... Figures 9A-9C The descriptions differ from those in the text.
[0141] Figure 10 This is a flowchart of an example process 1000 associated with forming the semiconductor device described herein. In some implementations, this process is performed using one or more semiconductor processing tools, such as deposition tools, exposure tools, development tools, etching tools, planarization tools, ion implantation tools, annealing tools, wafer / die transfer tools, and / or another type of semiconductor processing tool. Figure 10 One or more process blocks.
[0142] like Figure 10 As shown, process 1000 may include forming a first plurality of semiconductor layers and a second plurality of semiconductor layers, the second plurality of semiconductor layers alternating with the first plurality of semiconductor layers in a direction substantially perpendicular to the semiconductor substrate of the semiconductor device (block 1010). For example, as described herein, one or more semiconductor processing tools may be used to form the first plurality of semiconductor layers (e.g., interposers 125-1 to 125-3) and the second plurality of semiconductor layers (semiconductor layers 130-1 to 130-3), the second plurality of semiconductor layers alternating with the first plurality of semiconductor layers in a direction substantially perpendicular to the semiconductor substrate (e.g., semiconductor substrate 100) of the semiconductor device (e.g., semiconductor device 105).
[0143] like Figure 10 As further shown, process 1000 may include etching a first plurality of semiconductor layers, a second plurality of semiconductor layers, and a semiconductor substrate to define a fin substrate and a layer stack above the fin substrate (block 1020). For example, as described herein, one or more semiconductor processing tools may be used to etch the first plurality of semiconductor layers, the second plurality of semiconductor layers, and the semiconductor substrate to define a fin substrate (e.g., fin substrate 140) and a layer stack above the fin substrate (e.g., nanostructure layer stack 115).
[0144] like Figure 10 As further shown, process 1000 may include etching a layer stack and a fin substrate to form source / drain recesses (block 1030). For example, as described herein, one or more semiconductor processing tools may be used to etch the layer stack and fin substrate to form source / drain recesses (e.g., source / drain recess 305). In some implementations, etching the layer stack and fin substrate to form source / drain recesses results in the formation of a mesa structure extending over the fin substrate (e.g., mesa structure 315) and a channel layer stack (e.g., channel layer stack 310) over the mesa structure. In some implementations, the channel layer stack includes semiconductor interposers formed from a first plurality of semiconductor layers (e.g., interposers 320-1 to 320-3) and nanostructured channels formed from a second plurality of semiconductor layers (e.g., nanostructured channels 325-1 to 325-3).
[0145] like Figure 10 As further shown, process 1000 may include etching a semiconductor interposer to remove the semiconductor interposer from between the nanostructure channels of the channel layer stack (block 1040). For example, as described herein, one or more semiconductor processing tools may be used to etch the semiconductor interposer to remove the semiconductor interposer from between the nanostructure channels of the channel layer stack.
[0146] like Figure 10 As further shown, process 1000 may include material (block 1050) for depositing a dielectric layer along the sidewalls of the source / drain recesses and in the spaces formed between nanostructure channels by removing a semiconductor interposer. For example, as described herein, one or more semiconductor processing tools may be used to deposit a dielectric layer (e.g., dielectric layer 410) along the sidewalls of the source / drain recesses and in the spaces (e.g., space 405) formed between nanostructure channels by removing a semiconductor interposer.
[0147] like Figure 10As further shown, process 1000 may include performing a first etch operation to etch a dielectric layer, thereby forming a dielectric interposer layer (block 1060) in the space between the nanostructure channels. For example, as described herein, one or more semiconductor processing tools may be used to perform the first etch operation to etch the dielectric layer, thereby forming a dielectric interposer layer (e.g., interposers 455-1 to 455-3) in the space between the nanostructure channels. In some implementations, the first etch operation is performed using an approximately constant flow rate of a first process gas and a second process gas, resulting in the formation of a cavity (e.g., cavity 460) at the end of the dielectric interposer layer.
[0148] like Figure 10 As further shown, process 1000 may include performing a second etching operation to etch the ends of the dielectric interposer, thereby modifying the shape of the inner surface of the cavity (block 1070). For example, as described herein, one or more semiconductor processing tools may be used to perform the second etching operation to etch the ends of the dielectric interposer, thereby modifying the shape of the inner surface of the cavity (e.g., inner surface 470). In some implementations, the second etching operation is performed using a first variable flow rate of a first process gas and a second variable flow rate of a second process gas.
[0149] like Figure 10 As further shown, process 1000 may include forming internal spacers in the cavity (block 1080) after performing the second etching operation. For example, as described herein, after performing the second etching operation, one or more semiconductor processing tools may be used to form internal spacers in the cavity (e.g., internal spacers 610-1 to 610-3).
[0150] like Figure 10 As further shown, process 1000 may include forming a source / drain region in the source / drain recess such that the source / drain region contacts an internal spacer, and such that the source / drain region includes a plurality of epitaxial layers comprising the same semiconductor material but with different dopant concentrations (block 1090). For example, as described herein, one or more semiconductor processing tools may be used to form a source / drain region (e.g., source / drain region 705) in the source / drain recess such that the source / drain region contacts an internal spacer, and such that the source / drain region includes a plurality of epitaxial layers (e.g., first epitaxial layer 720, second epitaxial layer 725) comprising the same semiconductor material but with different dopant concentrations.
[0151] Process 1000 may include additional implementations, such as any single implementation or any combination of implementations described below and / or combined with one or more other processes described elsewhere in this document.
[0152] In the first implementation, the first variable flow rate of the first process gas is achieved by sending a first flow pulse of the first process gas into the processing chamber of the semiconductor processing tool for the second etching operation, and the second variable flow rate of the second process gas is achieved by sending a second flow pulse of the second process gas into the processing chamber of the semiconductor processing tool for the second etching operation.
[0153] In the second implementation, either alone or in combination with the first implementation, a first stream of the first process gas and a second stream of the second process gas are supplied to the processing chamber during a gas pulse, followed by a shutdown duration during which the first stream of the first process gas and the second stream of the second process gas are stopped.
[0154] In the third implementation, either alone or in combination with one or more of the first and second implementations, the first flow velocity of the first process gas in the gas pulse is approximately constant, while the second flow velocity of the second process gas in the gas pulse is variable.
[0155] In the fourth implementation, either alone or in combination with one or more of the first to third implementations, the first process gas includes ammonia (NH3), and the second process gas includes hydrofluoric acid (HF).
[0156] In the fifth implementation, either alone or in combination with one or more of the first to fourth implementations, during the first portion of the gas pulse, the first flow rate of the first stream of the first process gas is higher than the second flow rate of the second stream of the second process gas.
[0157] In the sixth implementation, either alone or in combination with one or more of the first to fifth implementations, during the second portion of the gas pulse, the first flow rate of the first stream of the first process gas is less than the second flow rate of the second stream of the second process gas.
[0158] although Figure 10 An example block of process 1000 is shown, but in some implementations, process 1000 includes... Figure 10 The blocks shown can be different additional blocks, fewer blocks, different blocks, or blocks with different arrangements. Alternatively, two or more blocks of process 1000 can be executed in parallel.
[0159] Figure 11 This is a flowchart of an example process 1100 associated with forming the semiconductor device described herein. In some implementations, this process is performed using one or more semiconductor processing tools, such as deposition tools, exposure tools, development tools, etching tools, planarization tools, ion implantation tools, annealing tools, wafer / die transfer tools, and / or another type of semiconductor processing tool. Figure 11 One or more process blocks.
[0160] like Figure 11 As shown, process 1100 may include forming a layer stack over a fin substrate of a semiconductor device, the layer stack including a plurality of semiconductor interposers and a plurality of semiconductor channel layers, the plurality of semiconductor channel layers alternating with the plurality of semiconductor interposers in a first direction (block 1110). For example, as described herein, one or more semiconductor processing tools may be used to form a layer stack (e.g., nanostructure layer stack 115) over a fin substrate (e.g., fin substrate 140) of a semiconductor device (e.g., semiconductor device 105), the layer stack including a plurality of semiconductor interposers (e.g., interposers 125-1 to 125-3) and a plurality of semiconductor channel layers (e.g., semiconductor layers 130-1 to 130-3), the plurality of semiconductor channel layers alternating with the plurality of semiconductor interposers in a first direction (e.g., the z-direction). In some implementations, the fin substrate and the layer stack extend in a second direction (e.g., the x-direction).
[0161] like Figure 11 As further shown, process 1100 may include an STI region (block 1120) formed adjacent to the fin substrate and extending in a second direction. For example, as described herein, one or more semiconductor processing tools may be used to form the STI region (e.g., STI region 150) adjacent to the fin substrate and extending in a second direction. In some implementations, the STI region has a non-planar top surface along a third direction (y-direction) that is substantially perpendicular to the second direction.
[0162] like Figure 11 As further shown, process 1100 may include etching a layer stack and a fin substrate to form source / drain recesses (block 1130). For example, as described herein, one or more semiconductor processing tools may be used to etch the layer stack and fin substrate to form source / drain recesses (e.g., source / drain recess 305). In some implementations, etching the layer stack and fin substrate to form source / drain recesses results in the formation of a mesa structure extending over the fin substrate (e.g., mesa structure 315) and a channel layer stack (e.g., channel layer stack 310) over the mesa structure. In some implementations, the channel layer stack includes portions of a plurality of semiconductor interposers (e.g., interposers 320-1 to 320-3) and nanostructured channels formed from the plurality of semiconductor channel layers (e.g., nanostructured channels 325-1 to 325-3).
[0163] like Figure 11As further shown, process 1100 may include etching portions of the semiconductor interposer to remove portions of the semiconductor interposer from between the nanostructure channels of the channel layer stack (block 1140). For example, as described herein, one or more semiconductor processing tools may be used to etch portions of the semiconductor interposer to remove portions of the semiconductor interposer from between the nanostructure channels of the channel layer stack.
[0164] like Figure 11 As further shown, process 1100 may include material (block 1150) for depositing a dielectric layer along the sidewalls of the source / drain recesses and in the spaces formed between nanostructure channels by removing a semiconductor interposer. For example, as described herein, one or more semiconductor processing tools may be used to deposit material for a dielectric layer (e.g., dielectric layer 410) along the sidewalls of the source / drain recesses and in the spaces (e.g., space 405) formed between nanostructure channels by removing a semiconductor interposer.
[0165] like Figure 11 As further shown, process 1100 may include performing a first etch operation to etch a dielectric layer, thereby forming a dielectric interposer (block 1160) in the space between the nanostructure channels. For example, as described herein, one or more semiconductor processing tools may be used to perform the first etch operation to etch the dielectric layer, thereby forming a dielectric interposer (e.g., interposers 455-1 to 455-3) in the space between the nanostructure channels. In some implementations, the first etch operation results in the formation of a cavity (e.g., cavity 460) at the end of the dielectric interposer. In some implementations, after the first etch operation, the inner surface of the cavity (e.g., inner surface 470) has a first corner rounding amount in a third direction.
[0166] like Figure 11 As further shown, process 1100 may include performing a second etch operation to etch the ends of the dielectric interposer (block 1170). For example, as described herein, one or more semiconductor processing tools may be used to perform the second etch operation to etch the ends of the dielectric interposer. In some implementations, after the second etch operation, the inner surface of the cavity has a second corner rounding amount in a third direction. In some implementations, the second corner rounding amount is less than the first corner rounding amount.
[0167] like Figure 11 As further shown, process 1100 may include forming internal spacers in the cavity (block 1180) after performing the second etching operation. For example, as described herein, after performing the second etching operation, one or more semiconductor processing tools may be used to form internal spacers in the cavity (e.g., internal spacers 610-1 to 610-3).
[0168] Process 1100 may include other implementations, such as any single implementation or any combination of implementations described below and / or combined with one or more other processes described elsewhere herein.
[0169] In the first implementation, a first process gas and a second process gas are used in the first etching operation, and in the second etching operation, a first process gas, a second process gas, and a third process gas are used.
[0170] In the second implementation, either alone or in combination with the first implementation, the first process gas includes ammonia (NH3), and the second process gas includes hydrofluoric acid (HF).
[0171] In the third implementation, the third process gas includes methyl acid gas, either alone or in combination with one or more of the first and second implementations.
[0172] In the fourth implementation, either alone or in combination with one or more of the first to third implementations, the first process gas and the third process gas are supplied to the processing chamber of the semiconductor processing tool using a first gas pulse unit of the semiconductor processing tool, and the second process gas is supplied to the processing chamber of the semiconductor processing tool using a second process gas pulse unit of the semiconductor processing tool.
[0173] In the fifth implementation, either alone or in combination with one or more of the first to fourth implementations, the duration of the start pulse that introduces the first process gas and the first flow pulse of the third process into the processing chamber is longer than the duration of the start pulse that introduces the second flow pulse of the second process gas into the processing chamber.
[0174] In the sixth embodiment, either alone or in combination with one or more of the first to fifth embodiments, performing the first etching operation includes performing the first etching operation using a first semiconductor processing tool (e.g., semiconductor processing tool 415) having a single processing chamber (e.g., processing chamber 420), and performing the second etching operation includes performing the second etching operation using a second semiconductor processing tool (e.g., semiconductor processing tool 505) having multiple processing chambers (e.g., processing chambers 510a, processing chambers 510b), and in the second etching operation, the semiconductor device is transferred between the multiple processing chambers.
[0175] In the seventh embodiment, performing the first etching operation, either alone or in combination with one or more of the first to sixth embodiments, includes performing the first etching operation at a first pressure in a first processing chamber (e.g., processing chamber 420) of a first semiconductor processing tool (e.g., semiconductor processing tool 415), and performing the second etching operation includes performing the second etching operation at a second pressure in a second processing chamber (e.g., processing chamber 510a, processing chamber 510b) of a second semiconductor processing tool (e.g., semiconductor processing tool 505), wherein the second pressure is higher than the first pressure.
[0176] In the eighth implementation, either alone or in combination with one or more of the first to seventh implementations, the ratio of the second pressure to the first pressure is approximately 4:1 to 10:1.
[0177] although Figure 11 An example block of process 1100 is shown, but in some implementations, process 1100 includes... Figure 11 The blocks shown can be different additional blocks, fewer blocks, different blocks, or blocks with different arrangements. Alternatively, two or more blocks of process 1100 can be executed in parallel.
[0178] In this way, internal spacers are formed in nanostructured transistors of semiconductor devices to reduce the likelihood of etching through the internal spacers and into adjacent source / drain regions. Specifically, a multi-etch process as described herein is performed to form cavities in which the internal spacers are to be formed. The multi-etch process includes one or more first etch operations to form the cavity, and one or more second etch operations to refine the corners of the cavity to reduce corner rounding. Due to the performance of one or more second etch operations, the corners of the cavity have a sharper transition between the sidewalls and the inner surface of the cavity (e.g., increased orthogonality). This results in smaller lateral thickness variations and increased uniformity of the internal spacers subsequently formed in the cavity. The increased uniformity of the lateral thickness of the internal spacers reduces the likelihood of etching through any particular portion of the internal spacers during a replacement gate operation that replaces the sacrificial nanostructure layer with the gate structure of the nanostructured transistor. Thus, the techniques described herein can reduce the likelihood of nanostructured transistor failure and / or improve the yield of nanostructured transistors formed on semiconductor devices.
[0179] The following provides an overview of some aspects of this disclosure: Aspect 1: A method comprising: forming a first plurality of semiconductor layers and a second plurality of semiconductor layers, the second plurality of semiconductor layers alternating with the first plurality of semiconductor layers in a direction substantially perpendicular to a semiconductor substrate of a semiconductor device; etching the first plurality of semiconductor layers, the second plurality of semiconductor layers, and the semiconductor substrate to define a fin substrate and a layer stack above the fin substrate; etching the layer stack and the fin substrate to form a source / drain recess, wherein etching the layer stack and the fin substrate to form the source / drain recess results in the formation of a mesa structure extending above the fin substrate and a channel layer stack above the mesa structure, wherein the channel layer stack includes a semiconductor interposer formed from the first plurality of semiconductor layers and a nanostructure channel formed from the second plurality of semiconductor layers; etching the semiconductor interposer to remove the semiconductor interposer between the nanostructure channels of the channel layer stack; and etching along the sidewalls of the source / drain recess and the nanostructure... Material of a dielectric layer is deposited in the space formed between the channels by removing a semiconductor interposer; a first etch operation is performed to etch the dielectric layer to form a dielectric interposer and a cavity at the end of the dielectric interposer, wherein the first etch operation is performed using an approximately constant flow rate of a first process gas and a second process gas, resulting in the formation of a cavity at the end of the dielectric interposer; a second etch operation is performed to etch the end of the dielectric interposer to modify the shape of the inner surface of the cavity, wherein the second etch operation is performed using a first variable flow rate of the first process gas and a second variable flow rate of the second process gas; after performing the second etch operation, an internal spacer is formed in the cavity; and a source / drain region is formed in the source / drain recess such that the source / drain region contacts the internal spacer and such that the source / drain region includes a plurality of epitaxial layers comprising the same semiconductor material but with different dopant concentrations.
[0180] Aspect 2: According to the method of aspect 1, wherein the first variable flow rate of the first process gas is achieved by sending a first flow pulse of the first process gas into the processing chamber of the semiconductor processing tool for the second etching operation, and wherein the second variable flow rate of the second process gas is achieved by sending a second flow pulse of the second process gas into the processing chamber of the semiconductor processing tool for the second etching operation.
[0181] Aspect 3: According to the method of aspect 2, wherein a first stream of the first process gas and a second stream of the second process gas are provided to the processing chamber during a gas pulse, followed by a shutdown duration during which the first stream of the first process gas and the second stream of the second process gas are stopped.
[0182] Aspect 4: According to the method of aspect 3, the first flow velocity of the first process gas in the gas pulse is approximately constant, and the second flow velocity of the second process gas in the gas pulse is variable.
[0183] Aspect 5: According to the method of aspect 4, wherein during the first portion of the gas pulse, the first flow rate of the first stream of the first process gas is higher than the second flow rate of the second stream of the second process gas.
[0184] Aspect 6: According to the method of aspect 5, wherein, during the second portion of the gas pulse, the first flow rate of the first stream of the first process gas is less than the second flow rate of the second stream of the second process gas.
[0185] Aspect 7: The method according to any one of aspects 1-6, wherein the first process gas comprises ammonia (NH3), and wherein the second process gas comprises hydrofluoric acid (HF).
[0186] Aspect 8: A method comprising: forming a layer stack over a fin substrate of a semiconductor device, the layer stack including a plurality of semiconductor interposers and a plurality of semiconductor channel layers, the plurality of semiconductor channel layers alternating with the plurality of semiconductor interposers in a first direction, wherein the fin substrate and the layer stack extend in a second direction; forming a shallow trench isolation (STI) region extending in the second direction adjacent to the fin substrate, wherein the STI region has a non-planar top surface along a third direction, the third direction being substantially perpendicular to the second direction; etching the layer stack and the fin substrate to form a source / drain recess, wherein etching the layer stack and the fin substrate to form the source / drain recess results in the formation of a mesa structure extending over the fin substrate and a channel layer stack over the mesa structure, wherein the channel layer stack includes portions of the plurality of semiconductor interposers and nanostructured channels formed from the plurality of semiconductor channel layers; The process involves etching a portion of the semiconductor interposer to remove a portion of the semiconductor interposer from between the nanostructure channels of the stacked channel layers; depositing a dielectric layer material along the sidewalls of the source / drain recesses and in the spaces formed between the nanostructure channels by removing the semiconductor interposer; performing a first etching operation to etch the dielectric layer to form a dielectric interposer and a cavity at the end of the dielectric interposer, wherein after the first etching operation, the inner surface of the cavity has a first corner rounding amount in the third direction; performing a second etching operation to etch the end of the dielectric interposer, wherein after the second etching operation, the inner surface of the cavity has a second corner rounding amount in the third direction, wherein the second corner rounding amount is less than the first corner rounding amount; and using a first process gas, a second process gas, and a third process gas in the second etching operation.
[0187] Aspect 9: According to the method of aspect 8, a first process gas and a second process gas are used in the first etching operation, and the first process gas, the second process gas and a third process gas are used in the second etching operation.
[0188] Aspect 10: The method according to aspect 9, wherein the first process gas comprises ammonia (NH3), and wherein the second process gas comprises hydrofluoric acid (HF).
[0189] Aspect 11: According to the method of aspect 10, wherein the third process gas includes methyl acid gas.
[0190] Aspect 12: According to the method of aspect 9, wherein the first process gas and the third process gas are provided to the processing chamber of the semiconductor processing tool by using a first gas pulse unit of the semiconductor processing tool; and wherein the second process gas is provided to the processing chamber of the semiconductor processing tool by using a second gas pulse unit of the semiconductor processing tool.
[0191] Aspect 13: According to the method of aspect 9, wherein the duration of the start pulse for pulsedly introducing the first stream of the first process gas and the third process gas into the processing chamber is longer than the duration of the start pulse for pulsedly introducing the second stream of the second process gas into the processing chamber.
[0192] Aspect 14: The method according to any one of aspects 8-13, wherein performing the first etching operation comprises: performing the first etching operation using a first semiconductor processing tool having a single processing chamber, and wherein performing the second etching operation comprises: performing the second etching operation using a second semiconductor processing tool having a plurality of processing chambers, wherein in the second etching operation, the semiconductor device is transferred between the plurality of processing chambers.
[0193] Aspect 15: According to the method of aspect 14, performing the first etching operation includes performing the first etching operation in a first processing chamber of a first semiconductor processing tool at a first pressure, and wherein performing the second etching operation includes performing the second etching operation in a second processing chamber of a second semiconductor processing tool at a second pressure, wherein the second pressure is higher than the first pressure.
[0194] Aspect 16: The method according to aspect 14, wherein the ratio of the second pressure to the first pressure is about 4:1 to about 10:1.
[0195] Aspect 17: A semiconductor device comprising: a fin substrate; a shallow trench isolation (STI) region extending in a first lateral direction adjacent to the fin substrate; a channel layer stack above a mesa region of the fin substrate, including a plurality of nanostructured channels; a gate structure surrounding the plurality of nanostructured channels and extending in a second lateral direction above the STI region, the second lateral direction being substantially perpendicular to the first lateral direction, wherein the gate structure includes a gate dielectric layer and one or more metal layers, and wherein at least one of the one or more metal layers includes a titanium-containing material; and an internal spacer adjacent to an end of a gate portion of the gate structure located in a vertical direction between adjacent nanostructured channels of the plurality of nanostructured channels, wherein the internal spacer has a first width in the first lateral direction at approximately its center in the second lateral direction, wherein the internal spacer has a second width in the first lateral direction at its endpoint in the second lateral direction, and wherein the difference between the first width and the second width is in the range of about 0.1 nanometers to about 5 nanometers.
[0196] Aspect 18: The semiconductor device according to aspect 17, wherein, at approximately the center point of the gate portion in the second lateral direction, the internal spacer has a third width in the first lateral direction, wherein, at one side of the gate portion in the second lateral direction, the gate portion has a fourth width in the first lateral direction, and wherein the difference between the third width and the fourth width is in the range of about 0.1 nanometers to about 5 nanometers.
[0197] Aspect 19: The semiconductor device according to aspect 18, wherein the interface between the gate portion and the internal spacer is curved.
[0198] Aspect 20: The semiconductor device according to aspect 18, wherein the interface between the gate portion and the internal spacer is approximately rectangular.
[0199] The terms “about” and “substantially” can indicate that the value of a given quantity varies within 5% of that value (e.g., ±1%, ±2%, ±3%, ±4%, ±5% of that value). These values are merely examples and are not intended to be limiting. It should be understood that, according to this disclosure, the terms “about” and “substantially” can refer to a percentage of the value of a given quantity.
[0200] The foregoing disclosure outlines features of several embodiments, enabling those skilled in the art to better understand various aspects of this disclosure. Those skilled in the art should understand that they can readily use this disclosure as a basis for designing or modifying other processes and structures to perform the same purposes and / or achieve the same advantages as the embodiments described herein. Those skilled in the art should also recognize that these equivalent structures do not depart from the spirit and scope of this disclosure, and that various changes, substitutions, and modifications can be made to them without departing from the spirit and scope of this disclosure.
Claims
1. A method for forming a semiconductor device, comprising: A first plurality of semiconductor layers and a second plurality of semiconductor layers are formed, the second plurality of semiconductor layers alternating with the first plurality of semiconductor layers in a direction substantially perpendicular to the semiconductor substrate of the semiconductor device; Etch the first plurality of semiconductor layers, the second plurality of semiconductor layers and the semiconductor substrate to define the fin substrate and the layer stack above the fin substrate; The layer stack and the fin substrate are etched to form source / drain recesses. Etching the layer stack and the fin substrate to form the source / drain recess results in the formation of a mesa structure extending above the fin substrate and a channel layer stack above the mesa structure. The channel layer stack includes a semiconductor interposer formed from the first plurality of semiconductor layers and a nanostructured channel formed from the second plurality of semiconductor layers; The semiconductor interposer is etched to remove the semiconductor interposer from between the nanostructure channels of the channel layer stack; Material for depositing a dielectric layer along the sidewalls of the source / drain recesses and in the spaces formed between the nanostructure channels by removing the semiconductor interposer; A first etching operation is performed to etch the dielectric layer to form a dielectric interlayer and to form a cavity at the end of the dielectric interlayer. The first etching operation is performed using a near-constant flow rate of a first process gas and a second process gas, resulting in the formation of a cavity at the end of the dielectric interlayer. A second etching operation is performed to etch the ends of the dielectric interposer to modify the shape of the inner surface of the cavity. The second etching operation is performed using a first variable flow rate of the first process gas and a second variable flow rate of the second process gas; After performing the second etching operation, an internal spacer is formed in the cavity; as well as A source / drain region is formed in the source / drain recess such that the source / drain region contacts the internal spacer, and the source / drain region includes a plurality of epitaxial layers, the epitaxial layers comprising the same semiconductor material but with different dopant concentrations.
2. The method of claim 1, wherein, The first variable flow rate of the first process gas is achieved by sending a first pulse of the first process gas into the processing chamber of the semiconductor processing tool used for the second etching operation, and The second variable flow rate of the second process gas is achieved by sending a second flow pulse of the second process gas into the processing chamber of the semiconductor processing tool used for the second etching operation.
3. The method of claim 2, wherein, The first stream of the first process gas and the second stream of the second process gas are supplied to the processing chamber during a gas pulse, followed by a shutdown duration during which the first stream of the first process gas and the second stream of the second process gas are stopped.
4. The method of claim 3, wherein, The first flow velocity of the first process gas in the gas pulse is approximately constant, and The second flow rate of the second process gas in the gas pulse is variable.
5. The method of claim 4, wherein, During the first portion of the gas pulse, the first flow rate of the first stream of the first process gas is higher than the second flow rate of the second stream of the second process gas.
6. The method of claim 5, wherein, During the second portion of the gas pulse, the first flow rate of the first stream of the first process gas is less than the second flow rate of the second stream of the second process gas.
7. The method of claim 1, wherein, The first process gas includes ammonia (NH3), and The second process gas includes hydrofluoric acid (HF) gas.
8. A method of forming a semiconductor device, comprising: A layer stack is formed over a fin substrate of a semiconductor device. The layer stack includes a plurality of semiconductor interposers and a plurality of semiconductor channel layers, wherein the plurality of semiconductor channel layers alternate with the plurality of semiconductor interposers in a first direction. The fin base and the layer stack extend in the second direction; A shallow groove isolation (STI) zone is formed next to the fin base extending in the second direction. The STI region has a non-planar top surface along a third direction, which is substantially perpendicular to the second direction; The layer stack and the fin substrate are etched to form source / drain recesses. Etching the layer stack and the fin substrate to form the source / drain recess results in the formation of a mesa structure extending above the fin substrate and a channel layer stack above the mesa structure. The channel layer stack includes portions of the plurality of semiconductor interposers and nanostructured channels formed from the plurality of semiconductor channel layers; Etch a portion of the semiconductor interposer to remove a portion of the semiconductor interposer from between the nanostructure channels of the channel layer stack; Material for depositing a dielectric layer along the sidewalls of the source / drain recesses and in the spaces formed between the nanostructure channels by removing the semiconductor interposer; A first etching operation is performed to etch the dielectric layer to form a dielectric interlayer and to form a cavity at the end of the dielectric interlayer. After the first etching operation, the inner surface of the cavity has a first corner rounding amount along the third direction; Perform a second etching operation to etch the ends of the dielectric interposer. After the second etching operation, the inner surface of the cavity has a second corner rounding amount in the third direction, and Wherein, the second corner rounding amount is less than the first corner rounding amount; and An internal spacer is formed in the cavity after the second etching operation is performed.
9. The method of claim 8, wherein, In the first etching operation, a first process gas and a second process gas were used, and The second etching operation uses the first process gas, the second process gas, and the third process gas.
10. A semiconductor device, comprising: Fin base; The shallow groove isolation (STI) area extends in the first lateral direction next to the fin base; The channel layer stack is located above the mesa region of the fin substrate and includes multiple nanostructured channels; A gate structure surrounds the plurality of nanostructure channels and extends over the STI region along a second lateral direction, which is substantially perpendicular to the first lateral direction. The gate structure includes a gate dielectric layer and one or more metal layers, and At least one of the one or more metal layers comprises a titanium-containing material; and The internal spacer is adjacent to the end of the gate portion of the gate structure located in the vertical direction between adjacent nanostructure channels in the plurality of nanostructure channels. Wherein, at approximately the center of the inner spacer in the second lateral direction, the inner spacer has a first width in the first lateral direction. Wherein, at the endpoint of the inner spacer in the second lateral direction, the inner spacer has a second width in the first lateral direction, and The difference between the first width and the second width is in the range of about 0.1 nanometers to about 5 nanometers.
