Method for preventing wire bending during metal filling process

Abstract

This invention relates to a method for preventing wire bending during metal-filling processes, and specifically provides a method and apparatus for reducing wire bending: during the deposition of metals such as tungsten, molybdenum, ruthenium, or cobalt into features on a substrate, the features are periodically exposed to nitrogen, oxygen, or ammonia during atomic layer deposition, chemical vapor deposition, or sequential chemical vapor deposition to reduce interactions between the metals deposited on the feature sidewalls. The method is suitable for deposition into V-shaped features.

Description

[0001] This application is a divisional application of patent application No. 201710700258.6, filed on August 16, 2017, entitled "Method for Preventing Wire Bending During Metal Filling Process". Technical Field

[0002] The present invention relates generally to semiconductor processing, and more specifically to a method for preventing wire bending during a metal-filling process. Background Technology

[0003] The deposition of tungsten-containing materials is an essential part of many semiconductor manufacturing processes. These materials are used for horizontal interconnects, vias between adjacent metal layers, contacts between metal layers and devices on a silicon substrate, and high aspect ratio features. In conventional tungsten deposition processes on semiconductor substrates, the substrate is heated to processing temperature in a vacuum chamber, and a very thin portion of a tungsten film is deposited, which serves as a seed or nucleation layer. Subsequently, the remaining portion of the tungsten film (the bulk layer) is deposited on the nucleation layer by simultaneously exposing the substrate to two reactants. The bulk layer is typically deposited faster than the nucleation layer. However, with the shrinking of devices and the use of more complex patterning schemes in industry, the deposition of thin tungsten films has become a challenge. Summary of the Invention

[0004] The present invention provides a method and apparatus for depositing metal into features on a substrate. One aspect relates to a method of filling features on a substrate to form a line, the method comprising: (a) providing a substrate having a plurality of spaced features, the spacing between adjacent features being between about 20 nm and about 40 nm, each feature having a feature opening width, wherein the width of the feature narrows from the top to the bottom of the feature; (b) depositing a first amount of tungsten in the plurality of features on the substrate; (c) after depositing the first amount of tungsten, exposing the first amount of tungsten in the plurality of features to nitrogen gas; and (d) depositing a second amount of tungsten on the first amount of tungsten in the plurality of features.

[0005] In various embodiments, the nitrogen gas reduces tungsten-tungsten bonding interactions between tungsten atoms formed on the sidewalls of each feature.

[0006] In various implementations, the width of the bottom of each feature is between 0 nm and 90% of the width at the top of each feature.

[0007] The method may further include filling the feature with tungsten to form the line, wherein σ = (σ1) 2 +σ2 2 ) 1 / 2The calculated total variance of the lines within the substrate is less than about 5 nm, where σ1 is the variance of the variable line-to-line width and σ2 is the variance of the in-line width.

[0008] In various embodiments, the width at the bottom of 50% of the depth of the feature is between 0 nm and 20 nm.

[0009] In various embodiments, the first amount of tungsten is exposed to the nitrogen gas at a substrate temperature of less than about 500°C.

[0010] In some embodiments, the first amount of tungsten is exposed to nitrogen gas during the deposition of the second amount of tungsten on the first amount of tungsten.

[0011] In some embodiments, the second amount of tungsten is deposited via alternating pulses of hydrogen and a tungsten-containing precursor. The first amount of tungsten may be exposed to nitrogen during the pulses of hydrogen. In some embodiments, the first amount of tungsten is exposed to nitrogen during the pulses of the tungsten-containing precursor. In some embodiments, the first amount of tungsten is exposed to argon between the alternating pulses of hydrogen and the tungsten-containing precursor. When the feature is exposed to argon between the alternating pulses of hydrogen and the tungsten-containing precursor, the first amount of tungsten may be exposed to nitrogen.

[0012] On the other hand, a method relates to filling features on a substrate to form a line is provided, the method comprising: (a) providing a substrate having a plurality of spaced-apart features, the spacing between adjacent features being between about 20 nm and about 40 nm, each feature having a feature opening width, wherein the width of the feature narrows from the top to the bottom of the feature; (b) depositing a first amount of metal in the plurality of features on the substrate; (c) after depositing the first amount of metal, exposing the first amount of metal in the plurality of features to a suppressing gas; and (d) depositing a second amount of the metal in the plurality of features on the first amount of metal. The metal may be any one or more of ruthenium, molybdenum, and cobalt. The suppressing gas may be any one of nitrogen, oxygen, ammonia, and combinations thereof.

[0013] In various embodiments, the suppressant gas reduces metal-metal bonding interactions between the metals formed on the sidewalls of each feature. In some embodiments, the width at the bottom of each feature is between 0 nm and 90% of the width at the top of each feature. The method may also include filling the feature with the metal to form the line, wherein σ = (σ1) 2 +σ2 2 ) 1 / 2The calculated total variance of the lines within the substrate is less than approximately 5 nm, where σ1 is the variance of the variable line-to-line width, and σ2 is the variance of the width within the line. The width at the bottom, representing 50% of the feature's depth, can be between 0 nm and 20 nm.

[0014] On the other hand, an apparatus for processing a semiconductor substrate is disclosed, the apparatus having (a) at least one processing chamber including a base configured to hold the substrate; (b) at least one outlet coupled to a vacuum; (c) one or more process gas inlets coupled to one or more process gas sources; and (d) a controller for controlling operations in the apparatus, including machine-readable instructions for: providing a substrate having a plurality of spaced-apart features, the spacing between adjacent features being between about 20 nm and about 40 nm, each feature having a feature opening width, wherein the width of the feature narrows from the top to the bottom of the feature; introducing a tungsten-containing precursor and a reducing agent to deposit a first amount of tungsten in the plurality of features on the substrate; after depositing the first amount of tungsten, introducing nitrogen into the first amount of tungsten in the plurality of features, and introducing a tungsten-containing precursor and a reducing agent to deposit a second amount of tungsten on the first amount of tungsten in the plurality of features.

[0015] Specifically, some aspects of the present invention can be described as follows:

[0016] 1. A method for filling features on a substrate to form a line, the method comprising:

[0017] (a) A substrate having a plurality of spaced features, the spacing between adjacent features being between about 20 nm and about 40 nm, each feature having a feature opening width, wherein the width of the feature narrows from the top of the feature to the bottom of the feature;

[0018] (b) Depositing a first amount of tungsten in the plurality of features on the substrate;

[0019] (c) After depositing the first amount of tungsten, exposing the first amount of tungsten of the plurality of features to nitrogen gas; and

[0020] (d) Deposit a second amount of tungsten on the first amount of tungsten among the plurality of features.

[0021] 2. The method according to Clause 1, wherein the nitrogen gas reduces tungsten-tungsten bonding interactions between tungsten formed on the sidewalls of each feature.

[0022] 3. The method according to Clause 1, wherein the width of the bottom of each feature is between 0 nm and 90% of the width at the top of each feature.

[0023] 4. The method according to Clause 1, further comprising filling the feature with tungsten to form the line, wherein σ = (σ1) 2 +σ2 2 ) 1 / 2 The calculated total variance of the lines within the substrate is less than about 5 nm, where σ1 is the variance of the variable line-to-line width, and σ2 is the variance of the line width.

[0024] 5. The method according to Clause 1, wherein the width at the bottom of 50% of the depth of said feature is between 0 nm and 20 nm.

[0025] 6. The method according to any one of clauses 1 to 5, wherein the first amount of tungsten is exposed to the nitrogen gas at a substrate temperature of less than about 500°C.

[0026] 7. The method according to any one of clauses 1 to 5, wherein during the deposition of the second amount of tungsten on the first amount of tungsten, the first amount of tungsten is exposed to nitrogen gas.

[0027] 8. The method according to any one of clauses 1 to 5, wherein the second amount of tungsten is deposited by alternating pulse deposition of hydrogen and tungsten-containing precursors.

[0028] 9. The method according to Clause 8, wherein the first amount of tungsten is exposed to the nitrogen gas during the pulse of hydrogen.

[0029] 10. The method according to Clause 8, wherein the first amount of tungsten is exposed to the nitrogen gas during the pulse of the tungsten-containing precursor.

[0030] 11. The method according to Clause 8, wherein the first amount of tungsten is exposed to argon between the alternating pulses of the hydrogen and the tungsten-containing precursor.

[0031] 12. The method according to Clause 11, wherein when the feature is exposed to the argon gas between alternating pulses of hydrogen and the tungsten-containing precursor, the first amount of tungsten is exposed to the nitrogen gas.

[0032] 13. A method of filling a feature on a substrate to form a line, the method comprising:

[0033] (a) A substrate having a plurality of spaced features, the spacing between adjacent features being between about 20 nm and about 40 nm, each feature having a feature opening, wherein the width of the feature narrows from the top of the feature to the bottom of the feature;

[0034] (b) Depositing a first amount of metal in the plurality of features on the substrate;

[0035] (c) After depositing the first amount of metal, exposing the first amount of metal in the plurality of features to a suppressing gas; and

[0036] (d) Deposit a second amount of the metal on the first amount of the metal in the plurality of features.

[0037] 14. The method according to Clause 13, wherein the metal is selected from ruthenium, molybdenum and cobalt.

[0038] 15. The method according to Clause 13, wherein the suppressing gas is selected from nitrogen, oxygen, ammonia, and combinations thereof.

[0039] 16. The method according to Clause 13, wherein the suppressing gas reduces metal-metal bonding interactions between metals formed on the sidewalls of each feature.

[0040] 17. The method according to any one of clauses 13-16, wherein the width of the bottom of each feature is between 0 nm and 90% of the width at the top of each feature.

[0041] 18. The method according to any one of clauses 13-16, further comprising filling the feature with the metal to form the line, wherein σ = (σ1) 2 +σ2 2 ) 1 / 2 The calculated total variance of the lines within the substrate is less than about 5 nm, where σ1 is the variance of the variable line-to-line width, and σ2 is the variance of the line width.

[0042] 19. The method according to any one of clauses 13-16, wherein the width at the bottom of 50% of the depth of said feature is between 0 nm and 20 nm.

[0043] These and other aspects are further described below with reference to the accompanying drawings. Attached Figure Description

[0044] Figure 1 This is a schematic diagram of an exemplary film on a substrate.

[0045] Figure 2A A schematic diagram depicting an example of a dynamic random access memory (DRAM) architecture including buried word lines (bWL) in a silicon substrate.

[0046] Figure 2B A schematic diagram depicting the bending of the line is shown.

[0047] Figure 2C A schematic diagram of the zipper phenomenon is shown.

[0048] Figure 2DIt is a graph showing the interatomic forces that are a function of the tungsten-tungsten bond radius.

[0049] Figure 3A-3I These are illustrative examples of various structures in which metals such as tungsten can be deposited according to certain disclosed embodiments.

[0050] Figures 4A-4D It is a process flow diagram depicting the operation of a method performed according to certain disclosed embodiments.

[0051] Figure 5A-5J and Figure 6 This is a schematic diagram of an example of a mechanism for depositing a film according to certain disclosed embodiments.

[0052] Figure 7-11 This is a timing diagram illustrating exemplary loops in various methods according to certain disclosed implementations.

[0053] Figure 12 This is a schematic diagram of an exemplary processing tool for performing the disclosed embodiments.

[0054] Figure 13 This is a schematic diagram of an exemplary station used to perform the disclosed implementation methods.

[0055] Figure 14 Various time series diagrams were depicted.

[0056] Figure 15 , Figure 16 , Figure 17 , Figures 18A-18B and Figures 19A-19B It is a chart of the experimental results. Detailed Implementation

[0057] In the following description, numerous specific details are set forth to provide a full understanding of the presented embodiments. The disclosed embodiments may be practiced without some or all of these specific details. In other instances, well-known processing operations have not been described in detail to avoid unnecessarily obscuring the disclosed embodiments. While the disclosed embodiments are described in conjunction with specific examples, it should be understood that this is not intended to limit the scope of the disclosed embodiments.

[0058] Metal filling (e.g., tungsten (W) filling) of features is commonly used in semiconductor device fabrication to form electrical contacts. As devices shrink to smaller technology nodes and use more complex patterned structures, tungsten filling presents various challenges. One challenge is reducing the fluorine concentration, or content, in the deposited tungsten film. Smaller features with the same fluorine concentration in their tungsten film as larger features significantly impact device performance. For example, the smaller the feature, the thinner the deposited film. As a result, fluorine in the deposited tungsten film is more likely to diffuse through the thinner film, potentially leading to device failure.

[0059] One method to prevent fluorine diffusion involves depositing one or more barrier layers prior to tungsten deposition to prevent fluorine from diffusing from tungsten to other layers of the substrate, such as oxide layers. For example, Figure 1 An exemplary stack of layers deposited on substrate 190 is shown. Substrate 190 includes a silicon layer 192, an oxide layer 194 (e.g., titanium oxide (TiOx), tetraethyl orthosilicate (TEOS) oxide, etc.), a barrier layer 196 (e.g., titanium nitride (TiN)), a tungsten nucleation layer 198, and a host tungsten layer 199. The barrier layer 196 is deposited to prevent fluorine from diffusing from the host tungsten layer 199 and the tungsten nucleation layer 198 into the oxide layer. However, as the device shrinks, the barrier layer thins, and fluorine may still diffuse from the deposited tungsten layer. Although the chemical vapor deposition of the host tungsten at a higher temperature results in a lower fluorine content, this film has poor step coverage.

[0060] Another challenge is reducing the resistivity of the deposited tungsten film. Thinner films tend to have higher resistivity than thicker films. As features become smaller, the resistance of tungsten contacts or lines increases due to scattering effects in thinner tungsten films. Low-resistivity tungsten films minimize power losses and overheating in integrated circuit designs. Tungsten nucleation layers typically have higher resistivity than the overlying host layer. Barrier layers deposited in contacts, vias, and other features can also have high resistivity. Furthermore, thin barrier and tungsten nucleation films occupy a large percentage of smaller features, thus increasing the overall resistance within the feature. The resistivity of a tungsten film depends on the thickness of the deposited film, such that the resistivity increases with decreasing thickness due to boundary effects.

[0061] Another challenge is reducing the stress in the deposited film. Thinner tungsten films tend to have increased tensile stress. Conventional techniques for depositing bulk tungsten films via chemical vapor deposition (CVD) result in tensile stresses greater than 2.5 GPa for films up to 200 angstroms. This high thermal tensile stress causes substrate curling, which makes subsequent processing difficult. For example, subsequent processes might include chemical mechanical planarization, material deposition, and / or clamping the substrate to a substrate holder to perform the process in a chamber. However, these processes typically rely on a flat substrate, and a curled substrate leads to uneven processing or an unprocessable substrate. While existing methods exist for reducing stress in films of other materials, such as annealing, tungsten lacks surface mobility that allows grains to move or change once deposited due to their high melting point.

[0062] Another challenge is reducing line bending, a phenomenon found, for example, in substrates with multiple features with narrow spacing or in substrates with multiple high aspect ratio features adjacent to each other. During tungsten filling, line bending in DRAM buried word line (bWL) structures is thought to be caused by grain boundary merging (which can be called a “zipping mechanism”). When grain boundaries form, tungsten-tungsten bonding between adjacent tungsten surfaces (e.g., tungsten films grown on feature sidewalls) causes strain that leads to bending of the silicon wafer (line) separating the bWL. Conventional ALD and chemical vapor deposition (CVD) tungsten filling techniques result in severe bending of the bWL structure. This line bending leads to tungsten pitting inhomogeneities and contact landing problems in downstream processes, resulting in DRAM yield losses.

[0063] Conventional 2D growth can produce low-stress, low-fluorine, and low-resistivity tungsten films via ALD, but only on surfaces where such growth is permitted. As devices shrink and features narrow, stretching mechanisms may exist, which can cause tensile stress, high fluorine incorporation, and effects on resistivity, resulting in rough morphologies.

[0064] Specific implementations relate to methods and related apparatus for forming tungsten lines in storage devices. Figure 2A A schematic example of a DRAM architecture including buried word lines (bWL) 11 in a silicon substrate 9 is depicted. The bWL 11 is formed in trenches etched into the silicon substrate 9. The bWL 11 is tungsten deposited in the silicon substrate 9 and covered by SiN passivation 5. A conformal barrier layer 12 and an insulating layer 13 disposed between the conformal barrier layer 12 and the silicon substrate 9 are shown as trench liners. Figure 2AIn the example, insulating layer 13 may be a gate oxide layer formed of a material such as silicon oxide. Examples of conformal barrier layers include titanium nitride (TiN) barrier layers and tungsten-containing barrier layers. Tungsten-containing conformal barrier layers are described in U.S. Patent Application Publication No. 2016 / 0233220 (Serial No. 15 / 040,561), filed February 10, 2006, entitled “TUNGSTEN FOR WORDLINE APPLICATIONS,” which is incorporated herein by reference.

[0065] Conventional deposition processes for DRAM bWL trenches tend to deform the trenches, resulting in significant non-uniformity in the final trench width and resistance R. Figure 2B A typical narrow asymmetric trench structure of a DRAM bWL, with unfilled (201) and filled (205) sections, is shown. Multiple features are depicted on the substrate as shown. These features are spaced apart, with adjacent features having a spacing between approximately 20 nm and approximately 40 nm. The spacing is defined as the distance between the central axis of one feature and the central axis of the adjacent feature. Unfilled features are typically V-shaped, as shown in feature 203, which has sloping sidewalls, where the width of the feature narrows from the top to the bottom. The feature widens from the bottom 213b to the top 213a. Severe line bending is observed in the substrate shown in 205 after tungsten filling. Without being bound by any particular theory, it is believed that the cohesive forces between the opposing surfaces of the trenches pull the trench sides together, as indicated by arrow 207. This phenomenon is... Figure 2C This is illustrated and can be characterized as a "zipping up" feature. When feature 203 is filled, more force is applied from the central axis 299 of feature 203, causing line bending. The tungsten 243a and 243b deposited on the sidewalls of feature 203 thus interact closely together, with a small tungsten-tungsten bond radius r, thereby inducing interatomic cohesive forces between the smooth growth surfaces of the tungsten and pulling the sidewalls together, resulting in line bending. Figure 2D The interatomic forces are shown as a function of the tungsten-tungsten bond radius r. It can be seen that cohesive forces exist at certain values ​​of r.

[0066] Until recently, line bending (bWL) was thought to be caused by intrinsic tungsten film stress during filling. However, as mentioned above, low-stress tungsten films deposited via conventional ALD processes can induce severe line bending during filling. An alternative explanation based on a grain boundary zipping mechanism is proposed to account for the line bending.

[0067] This document describes methods, related systems, and apparatus for filling features with metal, using suppressing gases to reduce the formation of metal-metal bonds and thereby reduce line bends. Suppressing gases include nitrogen, oxygen, ammonia, and combinations thereof, depending on the metal to be deposited and the conditions and chemicals used for depositing the metal. Various embodiments involve exposing features with partially filled metal to suppressing gases without plasma to reduce the formation of metal-metal bonds in the features. As described herein, some disclosed embodiments are particularly suitable for filling V-shaped features.

[0068] Some disclosed embodiments utilize added nitrogen (N2) during tungsten filling to disrupt the formation of tungsten-tungsten bonds, thereby reducing strain in the bWL structure. Nitrogen addition can be performed in a pulsed manner (e.g., during H2 co-reactant pulses or purge pulses in cyclic deposition techniques such as atomic layer deposition (ALD) or sequential chemical vapor deposition (CVD), further described below) or continuously during any suitable deposition technique (e.g., during ALD cycling). While various embodiments and implementations are described herein with respect to tungsten, it should be understood that the disclosed embodiments are applicable to the deposition of a variety of metals, including but not limited to ruthenium, molybdenum, cobalt, etc. Application examples include logic and memory contact filling, DRAM buried word line filling, vertically integrated memory gate / word line filling, and 3D integration utilizing through-silicon vias (TSVs). The methods described herein can be used to fill vertical features, such as tungsten vias, and horizontal features, such as 3D-NAND word lines. These methods can be used for conformal and bottom-up or inside-out filling.

[0069] The addition of nitrogen during CVD and pulsed nucleation layer (PNL) processes is described in U.S. Patent No. 8,551,885, entitled "METHOD FOR REDUCING TUNGSTEN ROUGHNESS AND IMPROVING REFLECTIVITY," filed August 29, 2008, and granted October 8, 2013, which is incorporated herein by reference in its entirety. As described above, nitrogen can be added to control film roughness and improve tungsten packing.

[0070] This describes a method for preventing wire bending by adding a suppressing gas such as nitrogen. The addition of nitrogen is particularly effective during ALD tungsten filling and sequential CVD tungsten filling because the grain stretching mechanism is enhanced by film growth via a 2-D mechanism.

[0071] The disclosed implementation can block the surface of the grown tungsten film in a bWL filling process using nitrogen molecules. As adjacent surfaces of the grown film merge, W-N2 bonding weakens the W-W interaction, thereby reducing strain that would otherwise cause silicon line deflection. Process conditions can be tuned to minimize nitrogen (N) incorporated into the film to maintain the low resistivity of the tungsten filler.

[0072] Nitrogen is used in combination with the tungsten-containing precursor WF6 to allow adsorbed N2 molecules to disrupt the WW bond interactions during grain boundary merging, preventing these interactions from increasing stress on the film. H2 is used to convert the tungsten-containing precursor to tungsten in a dosing reaction to produce HF, which is then desorbed and removed from the chamber. In subsequent cycles of sequential CVD, weakly bonded N2 molecules can remain on the tungsten surface but are often used to reduce WW bond interactions at grain boundaries to facilitate interstitial filling without stressing the deposited tungsten film.

[0073] The disclosed embodiments include methods for depositing tungsten films with low fluorine concentrations to reduce line bending using a sequential CVD process combined with exposure to a suppressor gas such as nitrogen. The deposited films may also have low stress. Some methods involve the cyclic introduction of hydrogen and tungsten-containing precursors, such as tungsten hexafluoride. The disclosed embodiments can be combined with other tungsten deposition processes to deposit low-stress tungsten films with significantly lower fluorine content than films deposited by conventional CVD. For example, sequential CVD processes can be combined with nucleation layer deposition at low pressure, fluorine-free tungsten layer deposition, and / or non-sequential CVD processes. The disclosed embodiments have broad applications. The methods can be used to deposit tungsten into features with high step coverage and can also be used to deposit tungsten into 3D NAND structures, including those with deep trenches. Furthermore, the methods can be implemented for architectures that might otherwise be prone to line bending by adding nitrogen to the process.

[0074] Sequential CVD processes are distinguished from non-sequential CVD, pulsed CVD, atomic layer deposition (ALD), and nucleation layer deposition. Non-sequential CVD processes involve the simultaneous exposure of two reactants, allowing both reactants to flow concurrently during deposition. For example, host tungsten can be deposited by simultaneously exposing the substrate to hydrogen (H2) and tungsten hexafluoride (WF6) for a duration sufficient to fill the feature. The hydrogen and WF6 react during exposure to deposit tungsten into the feature. In pulsed CVD processes, one reactant flows continuously while the other is pulsed, but the substrate is exposed to both reactants during deposition to deposit material during each pulse. For example, when WF6 is pulsed, the substrate can be exposed to a continuous flow of H2, and during the pulse, WF6 and H2 react to deposit tungsten.

[0075] In contrast, sequential CVD processes expose each reactant individually, preventing them from flowing into the chamber simultaneously during deposition. Instead, each reactant stream is introduced sequentially into the substrate-containing chamber in separate pulses over time, repeated once or multiple times in a cycle. Generally, a cycle is the minimum set of operations used to perform a single surface deposition reaction. The result of a cycle is the formation of at least a partial film layer on the substrate surface. The cycle of sequential CVD is described in further detail below.

[0076] ALD and nucleation layer deposition also involve cyclically exposing the substrate to two reactants in a time-separated, pulsed manner. For example, in an ALD cycle, a first reactant flows into a chamber, the chamber is purged, a second reactant flows into the chamber, and the chamber is purged again. Such cycles are typically repeated to form a film thickness. In conventional ALD and nucleation layer deposition cycles, the first reactant stream constitutes the first “dosing” in a self-limiting reaction. For example, the substrate includes a limited number of active sites, whereby the first reactant is adsorbed onto the active sites on the substrate and reaches saturation on the surface, and the second reactant reacts with the adsorbed layer to cyclically deposit material layer by layer.

[0077] However, in sequential CVD, reactants are not necessarily adsorbed onto active sites on the substrate, and in some embodiments, the reaction may not be self-limiting. For example, reactants used in sequential CVD may have low adsorption rates. Furthermore, when a second reactant is introduced, reactants on the substrate surface may not necessarily react with the second reactant. Conversely, in some embodiments of sequential CVD, some reactants on the substrate remain unreacted during cycling and do not react until subsequent cycles. Some reactants may not react due to stoichiometry, steric hindrance, or other effects.

[0078] The methods described herein can be performed on a substrate that can be housed in a chamber. The substrate can be a silicon wafer, such as a 200mm, 300mm, or 450mm wafer, including wafers on which one or more layers of material (e.g., dielectric, conductive, or semiconductive) are deposited. The substrate has features such as vias or contact holes, which can be characterized by one or more of V-shaped sidewalls, narrow and / or recessed openings, contractions within the feature, and high aspect ratios. The feature can be formed in one or more of the aforementioned layers. For example, the feature can be at least partially formed in a dielectric layer. In some embodiments, the feature can have an aspect ratio of at least about 2:1, at least about 4:1, at least about 6:1, at least about 10:1, or higher. An example of a feature is a hole or via in a layer on a semiconductor substrate. Features can be spaced apart on the substrate, with a spacing of about 20 nm to about 40 nm between adjacent features.

[0079] Figure 3A-3GThese are illustrative examples of various structures in which tungsten can be deposited, according to the disclosed embodiments. Figure 3A An example cross-sectional view of a vertical feature 301 to be filled with tungsten is shown. Feature 301 may include a feature aperture 305 within a substrate 303. Aperture 305 or other features may have a certain dimension near the opening, for example, between about 10 nm and 500 nm, such as an opening diameter or linewidth between about 25 nm and about 300 nm. Feature aperture 305 may be referred to as an unfilled feature or simply as a feature. Feature 301 and any feature may be characterized in part by an axis 318 extending the length of the feature and passing through the center of aperture 305, wherein vertically oriented features have a vertical axis and horizontally oriented features have a horizontal axis.

[0080] In some implementations, the feature is a trench in the 3D NAND structure. For example, the substrate may include a word line structure having at least 60 rows, 18 to 48 layers, or hundreds of layers, with trenches at least 200 angstroms deep or reaching depths of many micrometers. Another example is a trench in the substrate or layers. The feature can be of any depth. In many implementations, the feature may have an underlying layer, such as a barrier layer or an adhesion layer. Non-limiting examples of the underlying layer include dielectric layers and conductive layers, such as silicon oxide, silicon nitride, silicon carbide, metal oxides, metal nitrides, metal carbides, and metal layers.

[0081] Figure 3B An example of a feature 301 with a concave profile is shown. The concave profile is a profile that narrows from the bottom, closed end, or interior of the feature toward the feature opening. According to various embodiments, this profile may gradually narrow and / or include a protrusion at the feature opening. Figure 3B An example of the latter is shown, in which a lower layer 313 is lining the sidewall or inner surface of the feature aperture 305 of feature 301. The lower layer 313 can be, for example, a diffusion barrier layer, an adhesive layer, a nucleation layer, a combination thereof, or any other suitable material. Non-limiting examples of the lower layer may include dielectric and conductive layers, such as silicon oxide, silicon nitride, silicon carbide, metal oxide, metal nitride, metal carbide, and metal layers. In a particular embodiment, the lower layer may be one or more of Ti, TiN, WN, TiAl, and W. The lower layer 313 forms a protrusion 315 such that the lower layer 313 is thicker near the opening of feature 301 than inside feature 301.

[0082] In some implementations, a feature having one or more contractions may be filled within the feature. Figure 3C Examples of views showing various filled features with contractions are shown. Figure 3CEach of the examples (a), (b), and (c) includes a contraction 309 at the midpoint within this feature. The width of the contraction 309 can be, for example, between about 15 nm and 20 nm. During the deposition of tungsten in the feature using conventional techniques, the contraction can cause pinch-off, where the deposited tungsten hinders further deposition through the contraction before the portion of the feature is filled, thus creating voids in the feature. Example (b) further includes a pad / block protrusion 315 at the feature opening. This protrusion can also be a potential pinch-off point. Example (c) includes a contraction 312, which is further away from the field region than the protrusion 315 in example (b).

[0083] It can also fill horizontal features such as 3D memory structures. Figure 3D An example of a horizontal feature 350 including a contraction 351 is shown. For example, the horizontal feature 350 may be a word line in a 3DNAND structure.

[0084] In some implementations, shrinkage may occur due to the presence of struts in the 3D NAND or other structures. For example, Figure 3E A plan view of the pillar 325 in the 3D NAND or Vertical Integrated Memory (VIM) structure 348 is shown, while Figure 3F This shows a simplified cross-sectional diagram of the support column 325. Figure 3E The arrows in the diagram represent deposited material; when the strut 325 is positioned between region 327 and the gas inlet or other deposit source, adjacent struts may create contractions 351, which can challenge the void-free filling of region 327.

[0085] For example, structure 348 can be formed by depositing a stack of alternating interlayer dielectric layers 329 and sacrificial layers (not shown) on substrate 300 and selectively etching the sacrificial layers. For example, these interlayer dielectric layers may be silicon oxide and / or silicon nitride layers, and these sacrificial layers may be materials selectively etched with an etchant. Thereafter, etching and deposition processes can be performed to form pillars 325, which may include the channel regions of a completed memory device.

[0086] The main surface of substrate 300 can extend in the x and y directions, while pillar 325 faces the z direction. Figure 3E and 3FIn this example, the pillars 325 are arranged in an offset manner such that adjacent pillars 325 in the x-direction are offset from each other in the y-direction, and vice versa. According to various embodiments, these pillars (and the corresponding contractions formed by adjacent pillars) can be arranged in many ways. Furthermore, the pillars 325 can be any shape, including circular, square, etc. The pillars 325 may comprise annular or circular (or square) semiconducting material. A gate dielectric may surround this semiconducting material. The regions between each interlayer dielectric layer 329 may be filled with tungsten; thus, the structure 348 has multiple horizontally oriented features of a stack to be filled, these horizontally oriented features extending in the x and / or y directions.

[0087] Figure 3G Provide another example of a horizontal view feature of a 3D NAND or other structure, such as including a strut contraction 351. Figure 3G The example is an open-type structure, where the material to be deposited can enter horizontally from both sides as shown by the arrows. (It should be noted that...) Figure 3G An instance can be viewed as a 2D drawing of the 3D features of this structure, and Figure 3G This is a cross-sectional view of the area to be filled, and the strut contraction shown in this figure represents the contraction that can be seen in a plan view rather than a cross-sectional view. In some embodiments, the 3-D structure may utilize along two or three-dimensional space (e.g., in...). Figure 3F In examples, the unfilled region extends in the x and y or x, y and z directions, and may present more filling challenges than filling holes or trenches extending along one-dimensional or two-dimensional space. For example, controlling the filling of 3-D structures can be challenging when deposited gas can enter the feature from multiple dimensions.

[0088] Figure 3H An example of a cross-sectional view of the V-shaped feature is provided. Figure 3HThe feature 301, to be filled with tungsten, includes a feature aperture 305 in a substrate 303. The aperture has dimensions close to the opening (e.g., opening diameter or linewidth w, which may be between about 10 nm and about 20 nm, or about 15 nm). The width is measured as the distance between the sidewalls of the feature. The width can vary from the top of the feature opening (opening diameter or linewidth w) to the bottom of the feature. The feature aperture 305 is partially characterized by an axis 318. The V-shaped feature 301 includes a depth 350 that may be between about 80 nm and about 120 nm, or about 100 nm. In various embodiments, the sidewalls intersect at point 395 at the bottom of the feature, or in some embodiments, the bottom of the feature plateaus to the flat bottom surface may have a distance between one sidewall and another, which is between about 0.1 w and about 0.9 w, or a percentage of the linewidth w at the opening between about 10% and about 90% of the width w. Features can have an aspect ratio between 2:1 and 10:1, or between 6:1 and 8:1, or between 6:1 and 8:1. The line spacing can be between 20 nm and 40 nm. The bottom of the feature (characterized by the area within the bottom at 50% to 70% of the feature's depth) can have a width between the sidewalls between 0 nm and 20 nm.

[0089] Figure 3I Another example of a cross-sectional view of a V-shaped feature is provided. As described herein, a V-shaped feature refers to a feature with a width that gradually narrows from the top field level of the substrate to the bottom of the feature. Figure 3I Feature 301, to be filled with a metal such as tungsten, includes a feature aperture 305 in a substrate 303. The aperture has dimensions close to its opening (e.g., opening diameter or linewidth w, which may be between about 10 nm and about 20 nm, or about 15 nm). The bottom of feature 396 is narrower than the width w. For example, the bottom of feature 396 may have a width between 1% and 90% of the width w, or between 1% and 50% of the width w, or between 10% and 20% of the width w.

[0090] In various disclosed embodiments, multiple V-shaped features exist on the substrate, such as Figure 2B As shown. Multiple features on the substrate are defined as adjacent features with a distance between them not exceeding 20 nm and 40 nm. In various embodiments, such multiple features include all V-shaped features, which may have, for example... Figure 3H Or the shape shown in 3I.

[0091] The following examples illustrate the filling of features for horizontal and vertical orientation features. It should be noted that these examples can be applied to both horizontal and vertical orientation features. Furthermore, it should be noted that in the following description, the term "horizontal" can be used to refer to a direction substantially orthogonal to the feature axis, while the term "vertical" refers to a direction substantially along the feature axis.

[0092] Although the following description focuses on tungsten feature filling, many aspects of the invention can also be implemented using other materials to fill features. For example, feature filling using one or more techniques described herein can be used to fill features with other materials, including other tungsten-containing materials (e.g., tungsten nitride (WN) and tungsten carbide (WC)), titanium-containing materials (e.g., titanium (Ti), titanium nitride (TiN), titanium silicide (TiSi), titanium carbide (TiC), and titanium aluminide (TiAl)), tantalum-containing materials (e.g., tantalum (Ta) and tantalum nitride (TaN)), and nickel-containing materials (e.g., nickel (Ni) and nickel silicide (NiSi)). Furthermore, some of the methods and apparatuses disclosed herein are not limited to feature filling but can be used to deposit tungsten on any suitable surface, including for forming a capping film on a flat surface.

[0093] Figure 4A A process flow diagram of a method performed according to certain disclosed embodiments is provided. Execution Figure 4A Operations 402-410 are performed to deposit a tungsten nucleation layer via ALD. Operation 495 involves exposing the substrate to nitrogen gas. According to various embodiments, operations 402, 404, 406, 408, 495, and 410 are performed to deposit the tungsten nucleation layer. In the various embodiments described herein, operations 402-410 are performed at a lower pressure compared to operation 480. For example, operations 402-410 may be performed at a low pressure of less than about 10 Torr. In some examples, operations 402-410 are performed at a pressure of about 10 Torr or about 3 Torr. Without being bound by any particular theory, it is believed that performing operations 402-410 at a low pressure during thin film deposition reduces the fluorine concentration in the deposited tungsten film due to the lower partial pressure of the fluorine-containing precursor in the chamber, resulting in less fluorine being introduced into the film. Examples of a process for depositing a tungsten nucleation layer under low pressure to achieve a low fluorine concentration in the deposited tungsten are further described in U.S. Patent Application Serial No. 14 / 723,275 (Attorney’s File No. LAMRP183 / 3623-1US), filed May 27, 2015.

[0094] In operation 402, the substrate is exposed to a tungsten-containing precursor, such as WF6. For the purposes described herein, although WF6 is used as an example of a tungsten-containing precursor, it should be understood that other tungsten-containing precursors may be suitable for performing the disclosed embodiments. For example, organometallic tungsten precursors may be used. Organometallic precursors and fluorine-free precursors, such as MDNOW (methylcyclopentadienyl-dicarbonylnitrosyl-tungsten) and EDNOW (ethylcyclopentadienyl-dicarbonylnitrosyl-tungsten), may also be used. Chlorinated tungsten precursors (WCl) may also be used. x Examples include tungsten pentachloride (WCl5) and tungsten hexachloride (WCl6).

[0095] In this example, the tungsten-containing precursor may include a combination of these compounds. In some embodiments, a carrier gas such as nitrogen (N2), argon (Ar), helium (He), or other inert gases may flow during operation 402. In some embodiments, the carrier gas carrying the tungsten-containing precursor may be diverted before being delivered to the substrate.

[0096] While tungsten has been described herein, it should be understood that in some embodiments, another metal can be deposited instead of tungsten by using a suitable metal-containing precursor. For example, to deposit molybdenum into a feature, a molybdenum-containing precursor, such as molybdenum tetrachloride (MoCl4), can be used.

[0097] Operation 402 can be performed for any suitable duration and at any suitable temperature. In some examples, operation 402 can be performed for a duration between about 0.25 seconds and about 30 seconds, about 0.25 seconds and about 5 seconds, or about 0.5 seconds and about 3 seconds. In some embodiments, this operation can be performed for a duration sufficient to saturate the active sites on the substrate surface.

[0098] In operation 404, the chamber may optionally be purged to remove excess WF6 that does not adhere to the substrate surface. Purging can be performed by flowing an inert gas at a constant pressure, thereby reducing the pressure of the chamber and repressurizing it before initiating another gas exposure.

[0099] In operation 406, the substrate is exposed to a reducing agent to deposit a tungsten nucleation layer. The reducing agent can be borane, silane, or germanane. Exemplary boranes include borane (BH3), diborane (B2H6), triborane, alkylborane, aminoborane, carborane, and haloborane. Exemplary silanes include silane (SiH4), disilane (Si2H6), propsilane (Si3H8), alkylsilane, aminosilane, carsilane, and haloborane. Germananes include Ge n H n+4 、Ge n H n+6 、Ge n Hn+8 and Ge n H m Where n is an integer from 1 to 10, and n is an integer different from m. Other germananes can also be used, such as alkylgermananes, aminogermananes, carbogermananes, and halogermananes. Generally, halogermananes may not have significant reduction potentials, but there may be suitable process conditions and tungsten-containing precursors for film formation using halogermananes.

[0100] Operation 406 can be performed for any suitable duration. In some examples, exemplary durations include between about 0.25 seconds and about 30 seconds, between about 0.25 seconds and about 5 seconds, or between about 0.5 seconds and about 3 seconds. In some embodiments, the operation may be sufficient to react with the adsorbed layer of WF6 on the substrate surface. Operation 406 can be performed for durations outside these example ranges. In some embodiments, a carrier gas, such as argon (Ar), helium (He), or nitrogen (N2), may be used.

[0101] Following operation 406, an optional purge step may be performed to remove excess reducing agent still in the gas phase that does not react with WF6 on the characteristic surfaces. Purge can be performed by flowing an inert gas at a fixed pressure, thereby reducing the chamber pressure and repressurizing the chamber before triggering exposure to another gas.

[0102] In operation 408, the chamber is purged to remove any reaction byproducts. This purge can be performed by introducing a purge gas, such as an inert gas, or by purging the chamber. Exemplary inert gases include, but are not limited to, hydrogen, argon, and helium.

[0103] In operation 495, the substrate is exposed to nitrogen. Nitrogen passivates the substrate, thereby reducing tungsten-tungsten bonding on the sidewalls of features on the substrate. In some embodiments, an inert gas may be introduced into the substrate along with nitrogen. Exemplary inert gases include argon, helium, and hydrogen. In embodiments that introduce a combination of nitrogen and hydrogen, the mixture may include at least about 10% nitrogen, or nitrogen gas between about 10% and about 100%.

[0104] Operation 495 is performed at a temperature below approximately 500°C or below approximately 450°C. At temperatures above 500°C, unwanted nitrogen atoms can be incorporated into the tungsten film in the feature. Operation 495 can be performed at the same pressure as used in operations 402-408. In some embodiments, operation 495 is performed at a pressure different from that used in operations 402-408, and the pressure is adjusted between the two pressures in each cycle.

[0105] In various embodiments, other suppressing gases may be used instead of nitrogen. In some embodiments, oxygen may be used instead of nitrogen. In some embodiments, depending on the metal to be deposited and the metal-containing precursor used for deposition, the suppressing gas may be nitrogen, oxygen, ammonia, or a combination thereof. For example, in some embodiments, ammonia (NH3) may be used instead of nitrogen or flow with nitrogen to prevent metal-metal bonding of the deposited material from the sidewalls of the substrate. If ammonia is used, the tungsten-containing precursor is not co-flowed with ammonia to prevent reaction between the tungsten-containing precursor and ammonia. For example, as further described below, in some embodiments, nitrogen is introduced continuously or in a pulsed manner. When tungsten hexafluoride is used as the tungsten-containing precursor, ammonia is introduced only in a pulsed manner or only when tungsten hexafluoride is not introduced into the substrate.

[0106] In operation 410, it is determined whether the tungsten nucleation layer has been deposited to a sufficient thickness. If not, operations 402-408 are repeated until a tungsten nucleation layer of the desired thickness is deposited on the feature surface. Additionally, operation 495 may be performed in each repeated cycle, or every two cycles, or every three cycles, or every four cycles, or less frequently. Each repetition of operations 402-408 may be referred to as an ALD “cycle.” In some embodiments, the order of operations 402 and 406 may be reversed, such that the reducing agent is introduced first.

[0107] After the tungsten nucleation layer has been deposited to a sufficient thickness, in operation 499, the substrate may be exposed to nitrogen (or oxygen or a nitrogen-containing gas, such as ammonia). Following exposure to nitrogen in operation 499, in operation 480, the host tungsten is deposited by sequential CVD. Although sequential CVD is described herein, in some embodiments, the host tungsten may be deposited by any suitable method (e.g., CVD or ALD). In many embodiments, operation 480 may be performed at a pressure greater than that during operations 402-410. For example, operation 480 may be performed at a pressure greater than or equal to about 10 Torr, such as about 10 Torr or about 40 Torr.

[0108] Figure 4BA process flow diagram is provided for operations that can be performed during operation 480 after exposure to nitrogen in operation 499. It should be understood that operation 499 can be performed continuously, such that nitrogen flows continuously during the deposition of the host tungsten via sequential CVD, or in some embodiments it can be periodically pulsed, for example, pulsed only during reducing agent exposure, or pulsed only during purge gas operations, or pulsed only during a single purge gas operation, or pulsed only during tungsten precursor dosing, or pulsed during one or more of the aforementioned operations. The pulses can occur during each cycle, or every two cycles, or every three cycles, or every four cycles, or less frequently as needed. A combination of continuous and pulsed exposure can also be used in some embodiments. Furthermore, as mentioned above, when ammonia is used to mitigate the reaction between tungsten deposited on the opposite sidewalls of the feature, ammonia is not introduced in a continuous or pulsed dosing manner during tungsten precursor exposure, such as during exposure to tungsten hexafluoride.

[0109] Note that it is optional to not execute. Figure 4A Execute in the case of operation Figure 4B The operation. In Figure 4B In operation 482, the substrate is exposed to a reducing agent such as H2. This operation may be referred to as “pulsed delivery” or “dosing,” and these terms are used interchangeably herein. In some embodiments described herein, H2 is provided as an exemplary reducing agent; however, it should be understood that other reducing agents may be used, including silanes, boranes, germananes, phosphine, hydrogen-containing gases, and combinations thereof. Unlike non-sequential CVD, H2 is delivered in pulses without the flow of another reactant. In some embodiments, a carrier gas may flow. The carrier gas may be as described above regarding… Figure 4A Any of the carrier gases described in operation 404 (e.g., argon or helium). Operation 482 can be performed for any suitable duration. In some examples, exemplary durations include between about 0.25 seconds and about 30 seconds, between about 0.25 seconds and about 5 seconds, or between about 0.5 seconds and about 3 seconds.

[0110] Return to Figure 4B In operation 484, the chamber is purged. In some embodiments, purging is optional. This purging operation removes excess H2 retained in the gas phase. Purging is performed by flowing an inert gas at a constant pressure, thereby reducing the pressure in the chamber and repressurizing it before triggering exposure to another gas. The chamber can be purged for any suitable duration, such as between about 0.1 seconds and about 3 seconds.

[0111] Return to Figure 4BIn operation 486, the substrate is exposed to a tungsten-containing precursor (e.g., WF6) to form a sub-monolayer of film on the substrate. In some embodiments, other tungsten-containing precursors may be used. While WF6 is used as an example of a tungsten-containing precursor, it should be understood that other tungsten-containing precursors may be suitable for implementing the disclosed embodiments. For example, organometallic tungsten precursors may be used. Organometallic precursors and fluorine-free precursors, such as MDNOW (methylcyclopentadienyl-dicarbonylnitrosyl-tungsten) and EDNOW (ethylcyclopentadienyl-dicarbonylnitrosyl-tungsten), may also be used. Chlorinated tungsten precursors (WCl) may also be used. x Examples include tungsten pentachloride (WCl5) and tungsten hexachloride (WCl6).

[0112] In this example, the tungsten-containing precursor may include a combination of these compounds. In some embodiments, a carrier gas such as nitrogen (N2), argon (Ar), helium (He), or other inert gases may flow. In some embodiments, the carrier gas may be diverted before the tungsten-containing precursor is delivered to the substrate.

[0113] While tungsten has been described herein, it should be understood that in some embodiments, another metal can be deposited instead of tungsten by using a suitable metal-containing precursor. For example, to deposit molybdenum into a feature, a molybdenum-containing precursor, such as molybdenum tetrachloride (MoCl4), can be used.

[0114] For the purposes of this example, WF6 is used. In various embodiments, WF6 flows into the chamber during the operation for a duration between about 0.1 seconds and about 3 seconds, or about 0.5 seconds. In some embodiments, WF6 may be transferred to fill the gas line and the line may be changed prior to dosing. In some embodiments, WF6 flows into the chamber but does not completely react with all H2 molecules on the substrate surface.

[0115] exist Figure 4B During operation 486, some WF6 may react with H2 retained on the surface from previously dispensed feed. Figure 4B During operation 486, some WF6 may not react completely with H2 retained on the surface from previously dispensed feed. Figure 4B During operation 486, some WF6 may not react with H2 at all, but can be physically adsorbed onto the substrate surface, where no H2 is physically adsorbed or retained on the substrate surface. In some embodiments, WF6 may be retained on the substrate surface, but cannot be physically or chemically adsorbed onto that surface. The following refers to the exemplary mechanism diagram. Figure 5A-5J Describe these examples.

[0116] In many implementations, Figure 4BOperation 486 can thus form a sub-monolayer of tungsten. For example, a sub-monolayer with a thickness of about 0.3 angstroms can be deposited after performing operations 482-486.

[0117] In some implementations, operations 486 and 482 can be reversed, such that operation 486 is performed before operation 482. In some implementations, operation 482 can be performed before operation 486.

[0118] exist Figure 4B In operation 488, the chamber is purged to remove reaction byproducts and WF6 from the gas phase. In some embodiments, an excessively short purge duration in operation 488 can increase discontinuous CVD reaction characteristics, resulting in the deposition of a high-stress film. In some embodiments, the purge duration is between about 0.1 seconds and about 2 seconds, and may prevent the removal of all WF6 from the substrate surface due to the low adsorption rate of WF6 on the tungsten surface. In some embodiments, the purge duration is between about 0.1 seconds and about 15 seconds, for example, about 7 seconds. For example, to fabricate a 3D NAND structure, the chamber may be purged for about 7 seconds during operation 488. The purge duration depends on the substrate and stress.

[0119] exist Figure 4B In operation 490, it is determined whether the host tungsten has been deposited to a sufficient thickness. If not, operations 482-488 are repeated until the desired thickness is deposited. In some embodiments, operations 482-488 are repeated until the feature is filled. In some embodiments, operation 499 is also repeated in combination with operations 482-488 when the host tungsten is deposited by repeating deposition cycles via sequential CVD.

[0120] Figure 4C A process flow diagram of the method performed according to the disclosed embodiments is provided. In operation 480, tungsten is deposited on the host via sequential CVD. The process conditions and chemicals can be as described above. Figure 4B and 5A -5J describes any of those. In operation 498, host tungsten is deposited by non-sequential CVD. During non-sequential CVD, the substrate is simultaneously exposed to a tungsten-containing precursor and a reducing agent to deposit the host tungsten. Exemplary tungsten-containing precursors include fluorine-containing precursors (e.g., WF6) and chlorine-containing precursors (e.g., WCl). xTungsten hexacarbonyl (W(CO)6) and hydrogen are used as reducing agents. In some embodiments, non-sequential CVD is performed by exposing the substrate to WF6 and H2. Operations 480 and 498 can be performed sequentially, or any operation in operation 480 can be performed once or more before or after operation 498. In some embodiments, operations 480 and 498 are performed in pulses, such that operation 498 is performed after every two or more cycles of operation 480. Therefore, a combination of sequential and non-sequential CVD can be used to deposit the host tungsten. In operation 499, the substrate is exposed to nitrogen. In some embodiments, operation 499 is performed in combination with operations 480, 498, or both. In many embodiments, the substrate is continuously exposed to nitrogen during operations 480, 498, or both. In many embodiments, the substrate is exposed to nitrogen in pulses during operations 480, 498, or both. In many embodiments, oxygen or other nitrogen-containing gases (such as ammonia) are used instead of nitrogen, or in combination with nitrogen. Operations 480, 498, and 499 can be performed sequentially, or any one of operations 480 or 499 can be performed once or multiple times before or after operation 498. In some embodiments, operations 480, 498, and 499 are performed in pulses, such that operation 498 is performed every two or more cycles of operation 480, and operation 480 is performed every two or more cycles of operation 499. Therefore, tungsten substrates can be deposited using a combination of sequential and non-sequential CVD with continuous exposure to nitrogen, oxygen, and combinations thereof. Similarly, tungsten substrates can be deposited using a combination of sequential and non-sequential CVD with periodic pulses of exposure to nitrogen, oxygen, ammonia, and combinations thereof.

[0121] Figure 4D A process flow diagram of the method performed according to the disclosed embodiments is provided. In operation 420, a substrate having adjacent V-shaped features is provided. The V-shaped features are as described above. Figure 3H and 3IAs defined, the distance between adjacent features on the substrate is no greater than a distance between approximately 20 nm and approximately 40 nm. In operation 430, a first amount of metal is deposited in the V-shaped feature. In many embodiments, the metal is tungsten. In some embodiments, the metal is ruthenium, or cobalt, or molybdenum. The metal is deposited using any suitable technique such as CVD, ALD, sequential CVD, etc. In some embodiments, the metal is tungsten deposited by sequential CVD using a tungsten-containing precursor (e.g., WF6, WCl6, or WCl5). In operation 439, the substrate is exposed to a suppressant gas, which can be nitrogen, oxygen, ammonia, or a combination thereof, depending on the metal to be deposited, the technique used to deposit the metal, and the precursor used to deposit the metal. For example, in some embodiments, the suppressant gas is nitrogen, and the metal to be deposited is tungsten, using WF6 as the tungsten-containing precursor. As stated above, it should be understood that if ammonia is used as the suppressant gas and WF6 is used as the tungsten-containing precursor for tungsten deposition, the exposure to the suppressant gas and the exposure to WF6 for tungsten deposition are time-separated to reduce the reaction between ammonia and WF6. It should be understood that operation 439 can be part of operation 430, such that during the deposition of the first amount of metal, the substrate is periodically or continuously exposed to the suppressant gas. (Refer to below...) Figure 8-11 These embodiments are further described. In operation 440, a second amount of metal is deposited on the first amount of metal. Any suitable deposition technique can be used. In some embodiments, operation 440 is performed after operation 439. For example, in one embodiment, a first amount of tungsten is deposited by CVD, the deposited tungsten is exposed to nitrogen, and a second amount of tungsten is deposited on the nitrogen-exposed deposited tungsten to fill the V-shaped feature. In some embodiments, operation 439 is performed as part of operation 440. For example, when the second amount of metal is deposited by sequential CVD, the suppressor gas can flow with a hydrogen pulse, or with an argon pulse, or with a tungsten-containing precursor pulse, or continuously throughout the sequential CVD cycle, wherein the sequential CVD cycle includes operations of exposure to hydrogen, exposure to argon, exposure to a tungsten-containing precursor, and exposure to argon.

[0122] The disclosed implementation is applicable to reducing line bending. Line bending analysis is performed by measuring the linewidth and roughness of trenches filled with metal (i.e., tungsten). Line bending analysis involves imaging the metal at the top of the device opening using a planar viewing microscope and measuring the metal width at multiple points along multiple lines. For each line, the linewidth is measured at 100 points. Then, for each line, the average linewidth and the variation in linewidth are calculated, which can also be defined as roughness. The “average linewidth” is the average of the average linewidths of all individual lines measured during the analysis.

[0123] For line curvature, two main metrics are defined as follows: (i) Line-to-line (LTL) variation is the standard deviation of the average line width, thus capturing the variable of line width variation across different lines in the image; and (ii) Line width roughness (LWR) is the average of the line roughness (the variable of line width within each line) from all measured lines, thus capturing the average line width variable within a single line. These two metrics, LTL and LWR, are combined into a univariate metric total σ, which is given by σ = (σ1...) 2 +σ2 2 ) 1 / 2 Determined. Furthermore, LTL and total σ are normalized relative to the average linewidth and described as LTL% and total σ% respectively.

[0124] In various embodiments, the disclosed embodiments produce substrates with a total variance of less than about 5 nm, or less than about 1.5 nm, or less than about 7.2% as a percentage, wherein the total variance percentage is calculated by normalizing the total variance using the average linewidth. Experiments conducted to determine the relationship between film thickness and line curvature for top-down SEM / trench images show that with V-shaped features or trenches (as described above)... Figure 3H and 3I As the metal thickness increases (as shown and depicted), the line bending phenomenon becomes more severe. This analysis is based on a top-down SEM / trench top analysis.

[0125] Nitrogen exposure can be used during the deposition of the tungsten nucleation layer and / or the bulk tungsten to reduce line bending. For example, see reference... Figure 4A Nitrogen exposure may be performed during any operation of 402, 404, 406, 408 and combinations thereof, or during all operations 402-408. (Reference) Figure 4B Nitrogen exposure can be performed during any of operations 482, 484, 486, 488 and combinations thereof, or during all operations 482-488. In some embodiments, nitrogen exposure is performed during sequential CVD deposition of the tungsten substrate while the features on the substrate are filled with the tungsten substrate to reduce line bending.

[0126] Figure 5A-5J This is a schematic diagram illustrating an exemplary mechanism of sequential CVD cycles. It should be understood that... Figure 5A-5J Exemplary mechanisms of nitrogen exposure are not included; such examples are found in Figure 6 Provided by China.

[0127] Figure 5AAn exemplary mechanism is described in which H2 is introduced onto a substrate 500 on which a tungsten nucleation layer 501 is deposited. Hydrogen (511a and 511b) is introduced in the gas phase, and some H2 (513a and 513b) is on the surface of the tungsten nucleation layer 501, but may not necessarily be adsorbed on the surface. For example, H2 may not necessarily be chemisorbed onto the nucleation layer 501, but in some embodiments, H2 may be physicosorbed onto the surface of the nucleation layer 501.

[0128] Figure 5B An example diagram is shown, where previously in the gas phase ( Figure 5A H2 in 511a and 511b is swept out of the chamber, and H2 previously on the surfaces (513a and 513b) remains on the surface of the tungsten nucleation layer 501.

[0129] Figure 5C It shows Figure 4B An exemplary diagram illustrating operation 486. Figure 5C In this process, the substrate is exposed to WF6, some of which are in the gas phase (531a and 531b), and some of which are located on or near the surface of the substrate (523a and 523b).

[0130] As mentioned above Figure 4B As described in operation 486, and as Figure 5D As shown in the example, WF6 can react with H2 to temporarily form intermediate 543b, thereby... Figure 5E In the process, intermediate 543b reacts completely to leave tungsten 590 on the nucleation layer 501 on the surface of substrate 500 and HF in the gas phase (e.g., 551a and 551b).

[0131] As mentioned above Figure 4B As described in operation 486, and as Figure 5D As shown, WF6 can partially react with H2 to form intermediate 543a, thereby... Figure 5E In this process, intermediate 543a remains partially reacted on the surface of substrate 500 at nucleation 501. Due to activation barriers and steric effects, the reaction mechanism involving WF6 and H2 may be slower than the reaction between borane, silane, or germanane and WF6 used for depositing tungsten nucleation layers. For example, without being bound by a specific theory, the stoichiometry of WF6 can be achieved by reacting at least three H2 molecules with one WF6 molecule. The partial molecular reaction of WF6 with H2 may not form tungsten, but rather an intermediate. For example, based on stoichiometric principles (e.g., using 3 H2 molecules to react with one WF6 molecule), this may occur even when there is not enough H2 near WF6 to react with it, thus leaving intermediate 543a on the surface of the substrate.

[0132] Figure 5FAn exemplary schematic diagram of the substrate is provided when the chamber is purged. Note that compound 543c may be a formed but not fully reacted intermediate, while some tungsten 590 may form on the substrate. Each cycle thus forms a sub-monolayer of tungsten on the substrate.

[0133] For example, Figure 5G It shows the repeating loop Figure 4B In operation 482, H2 511c from the gas phase is introduced into the substrate, which has deposited tungsten 590 and partially reacted intermediate 543d. Note that the introduced H2 can now react completely with intermediate 543d on the substrate, such that... Figure 5H As shown, the reactant compound 543d leaves the deposited tungsten layers 590b and 590c, and the byproducts HF551c and 551d form a gas phase. Some H2 511c can remain in the gas phase, while some H2 513c can remain on the tungsten layer 590a. Figure 5I In the middle, the cleaning room (therefore corresponding to) Figure 4B Operation 484) left behind deposited tungsten 590a, 590b, and 590c, as well as some H2 513c. Figure 5J In the process, WF6 is introduced again in the formulation, so that molecules 531c and 523c can then be adsorbed and / or react with H2 and the substrate. Figure 5J It can correspond to Figure 4B Operation 486. After WF6 is dispensed, the chamber can be cleaned again and the cycle can be repeated until the desired tungsten thickness is deposited.

[0134] The tungsten film deposited using the disclosed embodiments has a lower fluorine concentration than tungsten deposited by non-sequential CVD, for example, about two orders of magnitude lower. Deposition conditions such as temperature, pulse time, and other parameters can be varied depending on hardware or process modifications. The total tensile stress of the film may be less than about 1 GPa.

[0135] Figure 6 An example of a substrate with a V-shaped feature 603 is shown, wherein nitrogen 670 on the surface of the deposited tungsten 650 along the sidewalls of feature 603 prevents tungsten-tungsten bonding, thereby reducing line bending. Without being bound by any particular theory, it is believed that nitrogen desorbs and therefore little or no nitrogen is incorporated into the deposited tungsten film when additional tungsten is deposited. The disclosed embodiments are suitable for depositing metals such as tungsten into multiple features spaced apart on a substrate, wherein the spacing between adjacent features is between about 20 nm and about 40 nm.

[0136] Figure 7-11 Exemplary timing diagrams are provided according to various embodiments for executing exemplary loops of certain disclosed embodiments in response to continuous and pulsed variations in nitrogen exposure.

[0137] Figure 7 A timing diagram depicting an exemplary cycle of sequential CVD in process 700 is provided, which includes periodic nitrogen exposure during each sequential CVD cycle. Figure 7 The stages described in the examples shown include various process parameters such as carrier or purge gas flow, hydrogen flow, WF6 flow (an example used as a tungsten-containing precursor for tungsten deposition; other suitable metal-containing precursors can be used for deposition of suitable metals such as ruthenium, or cobalt, or molybdenum), and nitrogen flow. Lines indicate when the flow is turned on and off. Note that in many embodiments, the plasma is not ignited and is not depicted as a process parameter. Figure 7 Additional process parameters, not shown but which can be modulated as needed, include substrate temperature and processing chamber pressure.

[0138] Process 700 includes two deposition cycles 711A and 711B; however, it should be understood that more than two deposition cycles may be used in some disclosed embodiments. Deposition cycle 711A includes five stages: nitrogen dosing 799A, hydrogen dosing 720A, purging stage 740A, WF6 dosing 760A, and purging stage 770A. Nitrogen dosing 799A may correspond to... Figure 4B Operation 499. During nitrogen dosing 799A, the carrier gas flow can be initiated. The hydrogen and WF6 gas flows are shut off, and the nitrogen flow is initiated. Hydrogen dosing 720A can correspond to... Figure 4B Operation 482. During hydrogen dosing 720A, the carrier gas flow can be turned on. The hydrogen flow is turned on, while the WF6 and nitrogen flows are turned off. The purge phase 740A can correspond to... Figure 4B Operation 484. During purge phase 740A, the carrier gas may continue to flow to act as the purge gas. In some embodiments, this involves allowing the carrier gas to flow into the chamber rather than diverting it, which may be diverted during nitrogen, hydrogen, or WF6 gas flow. During purge phase 740A, the hydrogen, WF6, and nitrogen flows are shut off. Tungsten hexafluoride dosing 760A may correspond to... Figure 4B Operation 486. Although Figure 7 WF6 is described, but it should be understood that other tungsten-containing precursors can be used. Furthermore, although the combination... Figure 7 While tungsten is mentioned, it should be understood that other metal-containing precursors can be used to deposit other metals. For example, ruthenium-containing precursors can be used to deposit ruthenium, molybdenum-containing precursors can be used to deposit molybdenum, and cobalt-containing precursors can be used to deposit cobalt. During WF6 dosing 760A, the carrier gas can flow to introduce WF6 gas into the chamber, and the WF6 flow is also activated. During this dosing period, the hydrogen and nitrogen flows are shut off. The purge phase 770A can correspond to... Figure 4BOperation 488. During the purging phase 770A, when the carrier gas is used as the purging gas, the carrier gas flow is turned on, while the hydrogen, WF6, and nitrogen flows are turned off.

[0139] Determine as Figure 4B As depicted in operation 490, insufficient tungsten has been deposited, and the deposition cycle is repeated as depicted in deposition cycle 711B. Deposition cycle 711B includes nitrogen dosing 799B, hydrogen dosing 720B, a purging stage 740B, WF6 dosing 760B, and a purging stage 770B. During nitrogen dosing 799B, the nitrogen and carrier gas flows are turned on, while the hydrogen and WF6 flows are turned off. During hydrogen dosing 720B, the carrier and hydrogen flows are turned on, while the WF6 and nitrogen flows are turned off. During purging stage 740B, the carrier gas flows remain on, while the hydrogen, WF6, and nitrogen flows are turned off. During WF6 dosing 760B, the carrier and WF6 flows are turned on, while the hydrogen and nitrogen flows are turned off. During purging stage 770B, the carrier gas flows remain on, while the hydrogen, WF6, and nitrogen flows are turned off.

[0140] Figure 8-11 An exemplary tungsten deposition cycle pulse sequence is shown. Such a cycle can be an ALD deposition cycle or a sequential CVD deposition cycle. Although four examples are described, these examples are not limiting. While N2 is described in these embodiments, it should be understood that in some implementations, oxygen or ammonia may be used instead. Figure 8-11 WF6 is described, but it should be understood that other tungsten-containing precursors can be used. Furthermore, although the combination... Figure 8-11 While tungsten has been mentioned, it should be understood that other metal-containing precursors can be used to deposit other metals. For example, ruthenium-containing precursors can be used to deposit ruthenium, molybdenum-containing precursors can be used to deposit molybdenum, and cobalt-containing precursors can be used to deposit cobalt.

[0141] Figure 8 The sequence shown illustrates process 800, in which N2 flows continuously during sequential CVD, thereby alternatingly pulsed delivery of tungsten-containing precursors and reducing agent H2, with purge gas or carrier gas (such as argon) flowing between pulses. Figure 8A process 800 with two deposition cycles 811A and 811B is described. Deposition cycle 811A includes hydrogen dosing 820A, a purge stage 840A, WF6 dosing 860A, and a purge stage 870A. Nitrogen gas flows continuously throughout deposition cycles 811A and 811B. During hydrogen dosing 820A, the carrier gas flow, hydrogen flow, and nitrogen flow are open, while the WF6 gas flow is closed. During purge stage 840A, the carrier gas flow and nitrogen flow are open, while the H2 and WF6 flow are closed. Assuming continuous flow of the gas used to prevent tungsten-tungsten bonding in this example, and using WF6 as the tungsten-containing precursor in the example, it should be understood that NH3 is not used for continuous flow to reduce tungsten-tungsten bonding, to avoid reactions between NH3 and WF6 that may produce undesirable byproducts. During WF6 dosing 860A, the carrier gas flow, WF6 flow, and nitrogen flow are open, while the H2 gas flow is closed. During purge phase 870A, the carrier gas and nitrogen gas flows are turned on, while the H2 and WF6 gas flows are turned off. This cycle is repeated in deposition cycle 811B, which includes hydrogen dosing 820B, purge phase 840B, WF6 dosing 860B, and purge phase 870B. During hydrogen dosing 820B, similar to hydrogen dosing 820A, the carrier gas, hydrogen gas, and nitrogen gas flows are turned on, while the WF6 gas flows are turned off. During purge phase 840B, the carrier gas and nitrogen gas flows are turned on, while the H2 and WF6 gas flows are turned off. During WF6 dosing 860B, the carrier gas, WF6 gas, and nitrogen gas flows are turned on, while the H2 gas flows are turned off. During purge phase 870B, the carrier gas and nitrogen gas flows remain on, while the hydrogen and WF6 gas flows are turned off.

[0142] Figure 9 The sequence shown illustrates process 900, in which N2 is used during the dosing of reducing agent H2. Figure 9A process 900 with two deposition cycles 911A and 911B is described. Deposition cycle 911A includes hydrogen dosing 920A, a purging stage 940A, WF6 dosing 960A, and a purging stage 970A. During hydrogen dosing 920A, the carrier gas flow, hydrogen flow, and nitrogen flow are turned on, while the WF6 flow is turned off. During purging stage 940A, the carrier gas flow is turned on, while the H2 flow, WF6 flow, and nitrogen flow are turned off. During WF6 dosing 960A, the carrier gas flow and WF6 flow are turned on, while the hydrogen and nitrogen flow remain turned off. During purging stage 970A, the carrier gas flow remains on, while the H2 flow, WF6 flow, and nitrogen flow are turned off. This cycle is repeated in deposition cycle 911B, which includes hydrogen dosing 920B, a purging stage 940B, WF6 dosing 960B, and a purging stage 970B. During hydrogen dosing (920B), the carrier gas flow, hydrogen flow, and nitrogen flow are open, while the WF6 flow is closed. During purging (940B), the carrier gas flow remains open, while the H2, WF6, and N2 flows are closed. During WF6 dosing (960B), the carrier gas flow and WF6 flow are open, while the H2 and nitrogen flows are closed. During purging (970B), the carrier gas flow remains open, while the H2, WF6, and N2 flows are closed.

[0143] Figure 10 The sequence shown illustrates process 1000, in which N2 is used during an argon pulse following the H2 dosing and before the tungsten precursor dosing. Deposition cycle 1011A includes hydrogen dosing 1020A, a purge stage 1040A, WF6 dosing 1060A, and a purge stage 1070A. During hydrogen dosing 1020A, the carrier gas and hydrogen gas flows are on, while the nitrogen and WF6 gas flows are off. During purge stage 1040A, the carrier gas and nitrogen gas flows are on, while the hydrogen and WF6 gas flows are off. During WF6 dosing 1060A, the carrier gas and WF6 gas flows are on, while the hydrogen and nitrogen gas flows are off. During purge stage 1070A, the carrier gas flows remain on, while the H2, WF6, and N2 gas flows are off. This cycle is repeated in deposition cycle 1011B, which includes hydrogen dosing 1020B, a purging phase 1040B, WF6 dosing 1060B, and a purging phase 1070B. During hydrogen dosing 1020B, the carrier gas and hydrogen gas flows, while the WF6 and nitrogen gas flows remain closed. During purging phase 1040B, the carrier gas and nitrogen flow, while the H2 and WF6 gas flows are closed. During WF6 dosing 1060B, the carrier gas and WF6 gas flows, while the hydrogen and nitrogen gas flows are closed. During purging phase 1070B, the carrier gas flows, while the H2, WF6, and N2 gas flows are closed.

[0144] Figure 11The sequence shown illustrates process 1100, in which N2 is used during the argon pulse applied before the H2 reducing agent dosing and after the tungsten-containing precursor dosing. Deposition cycle 1111A includes hydrogen dosing 1120A, a purge stage 1140A, WF6 dosing 1160A, and a purge stage 1170A. During hydrogen dosing 1120A, the carrier gas and hydrogen gas streams are on, while the WF6 and nitrogen gas streams are off. During purge stage 1140A, the carrier gas stream is on, while the H2, WF6, and N2 gas streams are off. During WF6 dosing 1160A, the carrier gas and WF6 gas streams are on, while the hydrogen and nitrogen gas streams are off. During purge stage 1170A, the carrier gas and nitrogen gas streams are on, while the hydrogen and WF6 gas streams are off. This cycle is repeated in deposition cycle 1111B, which includes hydrogen dosing 1120B, a purging phase 1140B, WF6 dosing 1160B, and a purging phase 1170B. During hydrogen dosing 1120B, the carrier gas flow and hydrogen flow are turned on, while the WF6 and nitrogen flow are turned off. During purging phase 1140B, the carrier gas flow remains on, while the H2, WF6, and nitrogen flow are turned off. During WF6 dosing 1160B, the carrier gas flow and WF6 flow are turned on, while the hydrogen and nitrogen flow are turned off. During purging phase 1170B, the carrier gas flow and nitrogen flow are turned on, while the hydrogen and WF6 flow are turned off.

[0145] While WF6 is depicted in these embodiments, it should be understood that other tungsten-containing precursors, such as WCl6 and / or WCl5, may be used. Furthermore, suitable ruthenium-containing, molybdenum-containing, or cobalt-containing precursors may be used accordingly for depositing other metals, such as ruthenium, molybdenum, or cobalt.

[0146] The disclosed embodiments can have a variety of applications in tungsten deposition processes. For example, in some embodiments, features can be filled by depositing a tungsten nucleation layer through an ALD cycle of alternating pulses of a reducing agent (e.g., borane, silane, or germanane) and WF6 (with periodic exposure to nitrogen), followed by bulk tungsten deposition via sequential CVD with periodic exposure to nitrogen, as referenced above. Figure 4B As stated above.

[0147] In another example, in some implementations, a tungsten nucleation layer can be deposited using an ALD cycle with a reducing agent and WF6, followed by CVD of fluorine-free tungsten utilizing a reducing agent and a fluorine-free tungsten precursor (e.g., a metal-organic tungsten precursor) and as described above. Figure 4B The sequential CVD combination described above is used for host tungsten deposition, wherein the substrate is periodically exposed to nitrogen gas to prevent wire bending. Fluorine-free tungsten precursors may also include carbonyl tungsten (W(CO)6) and tungsten chloride (WCl). x Examples include tungsten pentachloride (WCl5) and tungsten hexachloride (WCl6).

[0148] In another example, the tungsten nucleation layer can be deposited on the feature via alternating pulses of reducing agent and WF6 in an ALD cycle, and the host tungsten can be deposited via as described above. Figure 4B The deposition is performed by alternating between sequential CVD and non-sequential CVD, wherein the substrate is periodically exposed to nitrogen gas to prevent wire bending. For example, multiple cycles of sequential CVD can be used between predetermined durations of non-sequential CVD to deposit the host tungsten. In a specific example, the host tungsten can be deposited as follows: approximately 5 cycles of sequential CVD, followed by 5 seconds of non-sequential CVD, then 5 cycles of sequential CVD, and another 5 seconds of non-sequential CVD.

[0149] In another example, the feature can be filled by first depositing a tungsten nucleation layer through an ALD cycle of alternating pulses of reducing agent and WF6, then partially filling the feature using sequential CVD, and filling the remainder of the feature by non-sequential CVD, wherein the substrate is periodically exposed to nitrogen to prevent line bending.

[0150] In another example, the feature can be filled by depositing a tungsten nucleation layer via alternating pulses of a reducing agent and WF6 using an ALD cycle, followed by partial deposition of the host tungsten via sequential CVD, and complete filling of the host via fluorine-free tungsten (e.g., using a metal-organic tungsten precursor) CVD, wherein the substrate is periodically exposed to nitrogen to prevent line bending. For example, multiple cycles of sequential CVD can be performed to partially fill the feature with the host tungsten, and then the remainder of the feature can be filled using CVD simultaneously exposed to MDNOW and H2. Note that in some embodiments, the feature can be filled without depositing a nucleation layer, but a nucleation layer can help reduce the growth delay of the host tungsten.

[0151] It should be understood that various combinations of applications described herein can be used to deposit tungsten, and the methods are not limited to the examples provided herein, in which the substrate is periodically exposed to nitrogen gas to prevent wire bending. For example, tungsten chloride-containing precursors (WCl5) such as tungsten pentachloride (WCl5) and tungsten hexachloride (WCl6) can be used. x It can be used in place of WF6 in the embodiments described herein or in combination with WF6.

[0152] In various embodiments, the immersion or surface treatment operation can be performed prior to the deposition of the nucleation layer. Exemplary immersion or surface treatments include exposing the substrate to silane (SiH4), disilane (Si2H6), propane (Si3H8), germanane (GeH4), argon (Ar), tungsten hexafluoride (WF6), diborane (B2H6), hydrogen (H2), nitrogen (N2), or combinations thereof. In some embodiments, one or more gases may be used to immerse the substrate. For example, in some embodiments, the substrate may be exposed to silane for a first duration, and then exposed to diborane for a second duration. Such operations may also be repeated cyclically. In another example, the substrate may be exposed to diborane for a first duration, and then exposed to silane for a second duration. In another example, the substrate may be exposed to diborane for a first duration, and then exposed to hydrogen for a second duration. In yet another example, the substrate may be exposed to silane for a first duration, and then exposed to hydrogen for a second duration. In some embodiments, any of the above-described immersion processes may be combined to expose the substrate to nitrogen. In any of the disclosed embodiments, the chamber containing the substrate can be purged between one or more immersion operations. Purging can be performed by introducing an inert gas, such as argon, into the chamber. For example, in one instance, the substrate may be exposed to diborane for a first duration, then the chamber may be purged, followed by exposure of the substrate to silane for a second duration.

[0153] According to some disclosed embodiments, the nucleation layer deposited prior to the deposition of the host tungsten layer can be deposited by alternating between a tungsten-containing precursor and a reducing agent (e.g., silane (SiH4), disilane (Si2H6), propane (Si3H8), germanane (GeH4), or diborane (B2H6)). In some embodiments, the nucleation layer is deposited by exposing the substrate to alternating pulses of a tungsten-containing precursor and silane. In some embodiments, the nucleation layer is deposited by exposing the substrate to alternating pulses of a tungsten-containing precursor and silane, and then exposing the substrate to alternating pulses of a tungsten-containing precursor and diborane. In some embodiments, the nucleation layer is deposited by exposing the substrate to alternating pulses of a tungsten-containing precursor and diborane, and then exposing the substrate to alternating pulses of a tungsten-containing precursor and silane. In some embodiments, a nucleation layer is deposited by exposing the substrate to alternating pulses of a tungsten precursor and silane, then to alternating pulses of a tungsten precursor and diborane, followed by alternating pulses of a tungsten precursor and silane. In any disclosed embodiment, the chamber housing the substrate can be purged between one or more feeding operations for depositing the nucleation layer. Purging can be performed by infusing the chamber with an inert gas such as argon. Any suitable inert gas can be used for purging. For example, in some embodiments, the substrate may be exposed to a pulse of a tungsten precursor, then the chamber may be purged, then the substrate may be exposed to a pulse of silane, and the chamber may be purged again, and such operations can be repeated cyclically.

[0154] Nucleation layer deposition that can be used in any of the above embodiments may include co-flowing of any one of hydrogen (H2), argon (Ar), nitrogen (N2), or combinations thereof during the entire nucleation deposition process, or during silane dosing, or during diborane dosing, or during tungsten precursor dosing (e.g., WF6 dosing), or during any purge time. In some embodiments, surface treatment operations may be performed during or after nucleation growth by exposing the substrate to any one of silane, disilane, propane, germanane, diborane, hydrogen, tungsten hexafluoride, nitrogen, argon, or combinations thereof. For example, during nucleation layer deposition, the substrate may be exposed to alternating pulses of silane and WF6, then the substrate may be immersed in silane, and then the substrate may be restored to alternating pulses of silane and WF6 exposure. Such operations may be cyclical. For example, in some embodiments, the following cycle may be repeated once or more to deposit a nucleation layer: alternating pulses of SiH4 and WF6 followed by surface treatment.

[0155] In some embodiments, the nucleation layer can be deposited by exposing the substrate to a tungsten-containing precursor and any combination of one or more of the following gases in one or more cycles, in any sequence and order: diborane, silane, disilane, propane, hydrogen, nitrogen, and germanane (GeH4). For example, in some embodiments, the nucleation layer can be deposited by exposing the substrate to diborane, exposing the substrate to tungsten hexafluoride, exposing the substrate to silane, and exposing the substrate to hydrogen. Such operation can be repeated one or more cycles. In another example, in some embodiments, the nucleation layer can be deposited by exposing the substrate to silane, exposing the substrate to tungsten hexafluoride, and exposing the substrate to hydrogen. Such operation can be repeated one or more cycles. In another example, in some embodiments, the nucleation layer can be deposited by exposing the substrate to diborane, exposing the substrate to hydrogen, and exposing the substrate to tungsten hexafluoride. Such operation can be repeated one or more cycles. In yet another example, in some embodiments, the nucleation layer can be deposited by exposing the substrate to nitrogen, exposing the substrate to diborane, and exposing the substrate to tungsten hexafluoride. Such operation can be repeated one or more cycles. In another example, in some embodiments, the nucleation layer can be deposited by exposing the substrate to silane, exposing the substrate to nitrogen, and exposing the substrate to tungsten hexafluoride. Such an operation can be repeated one or more cycles. In any of the described embodiments, the substrate can be surface-treated and / or immersed before, during, or after the deposition of the nucleation cycle using any available gas. In some embodiments, an additional gas can flow in tandem with any of the aforementioned gases during one or more exposures in the nucleation deposition process. In any disclosed embodiment, the chamber containing the substrate can be purged between one or more feeding operations for depositing the nucleation layer. Purging can be performed by infusing the chamber with an inert gas such as argon. Any suitable inert gas can be used for purging. It should be understood that in some embodiments, the substrate can be periodically exposed to nitrogen during the deposition of the tungsten nucleation layer.

[0156] The host tungsten deposit can be deposited using any of the disclosed embodiments described herein and in U.S. Patent Application Serial No. 14 / 723,275 (Attorney's File No. LAMRP183 / 3623-1US), filed May 27, 2015, which is incorporated herein by reference in its entirety. In any of the above embodiments, the host tungsten can also be deposited periodically, wherein re-nucleation and / or immersion and / or surface treatment and / or conventional CVD deposition operations are performed between host deposits. For example, in some embodiments, the above references can be used... Figure 4BThe disclosed embodiment is used to deposit a host tungsten, then the host tungsten deposition can be paused, and the substrate can then be exposed to alternating pulses of silane and WF6, or alternating pulses of diborane and WF6, to re-nucleate the substrate surface, and then can be used as described above. Figure 4B The disclosed embodiments are used to recover the host tungsten deposition. Such an operation can be repeated any number of cycles. In another instance, in some embodiments, the above references can be used. Figure 4B The disclosed embodiment is used to deposit the host tungsten, and then the host tungsten deposition can be paused. The substrate can then be immersed or surface-treated by flowing any of silane, disilane, propane, germanane, diborane, hydrogen, tungsten hexafluoride, nitrogen, argon, or combinations thereof to treat the substrate surface. The substrate surface can then be treated using the method described above. Figure 4B The disclosed embodiments are used to recover the host tungsten deposition. Host tungsten deposition can be performed by exposing the substrate to a tungsten-containing precursor (such as WF6) and any one or more of the following gases: hydrogen, silane, disilane, propane, diborane, nitrogen, argon, and germanane. Host tungsten can also be deposited using a combination of sequential CVD and conventional CVD as described above. Conventional CVD can be performed before, during (e.g., by cycling between sequential and conventional CVD) host tungsten deposition. It should be understood that in some embodiments, the substrate may be periodically exposed to nitrogen gas during the deposition of the tungsten host layer.

[0157] In some embodiments, the substrate can be annealed at any suitable temperature before and after the deposition of the host tungsten layer. In some embodiments, the substrate can be annealed at any suitable temperature after the deposition of the host tungsten layer. In some embodiments, the substrate can be annealed at any suitable temperature during an intermediate period of host tungsten deposition. Annealing can be performed in any suitable gaseous environment, such as environments containing one or more of the following gases: tungsten-containing gases (e.g., WF6), hydrogen, silane, disilane, propane, diborane, nitrogen, argon, and germane.

[0158] In various embodiments, the chamber housing the substrate can be as described above. Figure 4B The tungsten-containing precursor and reducing agent for depositing the host tungsten in the disclosed embodiments are evacuated or purged before or after dosing. In some embodiments, the delay time may be incorporated into the dosing or purge steps of sequential CVD deposition as described herein. In some embodiments, one or more gases may flow co-currently during dosing or purge operations using one or more of any of the following gases: WF6, hydrogen, silane, disilane, propane, diborane, nitrogen, argon, and germane.

[0159] The substrate temperature during nucleation deposition is the same as mentioned above. Figure 4B The substrate temperature may vary during the sequential CVD process. The substrate temperature is understood to represent the temperature set by the pedestal holding the substrate. The disclosed embodiments can be performed at any suitable pressure, such as greater than or less than about 10 Torr. For multi-station chambers, each pedestal may be set at a different temperature. In some embodiments, each pedestal is set at the same temperature. According to the disclosed embodiments, the substrate can be circulated from station to station during any or all of the above operations. The chamber pressure may also be modulated in one or more operations of some disclosed embodiments. In some embodiments, the chamber pressure during nucleation deposition differs from the chamber pressure during host deposition. In some embodiments, the chamber pressure during nucleation deposition is the same as the chamber pressure during host deposition.

[0160] During any of the above-described exposures, the gas can be pulsed or flowed continuously. For example, in some embodiments, during WF6 dosing in a sequential CVD operation, WF6 can be pulsed once or multiple times during a single dosing. Similarly, in some embodiments, during purging, the inert gas can be pulsed once or multiple times during a single purging operation. Such pulsed operation can be performed in any operation of nucleation deposition or any operation of host deposition, or any combination thereof. In some embodiments, one or more variations of one or more parameters (e.g., pressure, flow rate, and temperature) can be used. In some embodiments, the pedestal can be moved during any operation of nucleation deposition or host deposition, or both, such that the gap between the substrate and the nozzle on the pedestal can be adjusted. Moving the pedestal can be used in combination with changing one or more parameters (e.g., pressure, temperature, or flow rate). Adjusting the gap between the substrate and the nozzle can affect the pressure, temperature, or flow rate that can be used according to some disclosed embodiments. It should be understood that any process described herein is applicable to techniques involving ALD.

[0161] Device

[0162] Any suitable chamber can be used to implement the disclosed embodiments. Exemplary deposition apparatuses include various systems, such as... and The Max, available from Lam Research Corp. in Fremont, California, or any of a variety of other commercially available processing systems, can be used. In some embodiments, sequential chemical vapor deposition (CVD) can be performed at a first station, which is one of two, five, or more deposition stations located within a single deposition chamber. Thus, for example, hydrogen (H2) and tungsten hexafluoride (WF6) can be alternately introduced to the substrate surface at the first station using separate gas supply systems that create a localized atmosphere on the surface of the semiconductor substrate. Another station can be used for fluorine-free tungsten deposition, or non-sequential CVD. Another station can be used to deposit a tungsten nucleation layer at low pressure. Another station can be used for periodic nitrogen exposure. In some embodiments, periodic nitrogen exposure is performed in the same station as deposition. Two or more stations can be used to deposit tungsten in a parallel processing manner. Alternatively, the wafer can be transposed to perform deposition operations sequentially on two or more stations.

[0163] Figure 12 This is a block diagram of a processing system suitable for performing a tungsten thin film deposition process according to an embodiment of the present invention. The system 1200 includes a transfer module 1203. The transfer module 1203 provides a clean, pressurized environment to minimize the risk of substrate contamination as the substrate is moved between different reactor modules. Mounted on the transfer module 1203 is a multi-station reactor 1209 capable of performing atomic layer deposition (ALD) and sequential CVD using nitrogen or suppressing gas exposure according to embodiments. In some embodiments, the multi-station reactor 1209 can also be used to perform fluorine-free tungsten deposition and / or non-sequential CVD. The reactor 1209 may include multiple stations 1211, 1213, 1215, and 1217 that can operate sequentially according to the disclosed embodiments. For example, the reactor 1209 may be configured such that station 1211 performs nucleation layer deposition via ALD, station 1213 performs sequential CVD, station 1215 performs fluorine-free tungsten deposition, and station 1217 performs non-sequential CVD. The station can be configured to expose the wafer to periodic pulses or a continuous flow of nitrogen, oxygen, or ammonia gas to prevent line bending on the substrate. The station may include a heating base or substrate support, one or more gas inlets or nozzles, or a dispersion plate. An example of a deposition station 1300 is... Figure 13 As shown, it includes a substrate support 1302 and a nozzle 1303. A heater may be disposed within the base portion 1301.

[0164] Alternatively, one or more single- or multi-station modules 1207 may be mounted on the transfer module 1203, capable of plasma or chemical (non-plasma) pre-cleaning. This module can also be used for various processes, such as preparing substrates for deposition processes. The system 1200 also includes one or more wafer source modules 1201, where wafers are stored before and after processing. An atmospheric manipulator (not shown) in the atmospheric transfer chamber 1219 can first move the wafer from the source module 1201 to the loading lock 1221. Wafer transfer devices (typically robotic arm units) in the transfer module 1203 move the wafer from the loading lock 1221 to modules mounted on the transfer module 1203 and move the wafer between these modules.

[0165] In various implementations, a system controller 1229 is used to control the process conditions during the deposition process. The controller 1229 will typically include one or more memory devices and one or more processors. The processor may include a CPU or computer, analog and / or digital input / output connections, a stepper motor controller board, etc.

[0166] The controller 1229 controls the activities of all deposition equipment. The system controller 1229 runs system control software, which includes a set of instructions for controlling timing, gas mixing, chamber pressure, chamber temperature, wafer temperature, radio frequency (RF) power levels, wafer chuck or pedestal position, and other process-specific parameters. In some embodiments, additional computer programs stored on memory devices associated with the controller 1229 may be used.

[0167] Typically, a user interface will be associated with the controller 1229. The user interface may include a display screen, a graphical software display of the device and / or process conditions, and user input devices such as pointing devices, keyboards, touch screens, microphones, etc.

[0168] The system control logic can be configured in any suitable manner. Generally, this logic can be designed or configured in hardware and / or software. Instructions for controlling the drive circuitry can be hard-coded or provided as software. These instructions can be provided through "programming." Such programming is understood to include any form of logic, including hard-coded logic in digital signal processors, application-specific integrated circuits (ASICs), and other devices with specific algorithms implemented in hardware. Programming is also understood to include software or firmware instructions executable on a general-purpose processor. The system control software can be encoded in any suitable computer-readable programming language.

[0169] The computer program code used to control the germanium-containing reducing agent pulse, hydrogen flow rate, and tungsten-containing precursor pulse, as well as other processes, in the process sequence can be written in any conventional computer-readable programming language: for example, assembly language, C, C++, Pascal, Fortran, or others. The compiled object code or script is executed by the processor to perform the tasks identified in the program. Also, as indicated, the program code can be hard-coded.

[0170] The controller parameters relate to process conditions such as, for example, process gas composition and flow rate, temperature, pressure, cooling gas pressure, substrate temperature, and chamber wall temperature. These parameters are provided to the user in the form of a recipe and can be entered using a user interface.

[0171] Signals for monitoring the process can be provided via analog and / or digital input connections to system controller 1229. Signals for controlling the process are output via analog and digital output connections to deposition apparatus 1220.

[0172] The system software can be designed or configured in many different ways. For example, multiple chamber assembly subroutines or control targets can be written to control the operation of the chamber assembly required to perform the deposition process according to the disclosed embodiments. Examples of programs or program segments used for this purpose include substrate positioning code, process gas control code, pressure control code, and heater control code.

[0173] In some embodiments, controller 1229 is part of a system that may be part of the embodiments described above. Such systems include semiconductor processing apparatuses that include one or more processing tools, one or more chambers, one or more platforms for processing, and / or specific processing components (wafer pedestals, gas flow systems, etc.). These systems may be integrated with electronics to control their operation before, during, or after the processing of a semiconductor wafer or substrate. The electronics may be referred to as a “controller” that controls various components or sub-sections of one or more systems. Depending on the processing requirements and / or the type of system, controller 1229 may be programmed to control any of the processes disclosed in this invention, including controlling the delivery of processing gases, temperature settings (e.g., heating and / or cooling), pressure settings, vacuum settings, power settings, radio frequency (RF) generator settings in some systems, RF matching circuit settings, frequency settings, flow rate settings, plasma pulse frequency settings, fluid delivery settings, position and operation settings, wafer entry / exit tools and other transfer tools, and / or the transfer of load locks connected to or interfaced with a particular system.

[0174] In a broad sense, a controller can be defined as an electronic device with various integrated circuits, logic, memory, and / or software that receives instructions, issues instructions, controls operations, enables cleaning operations, enables endpoint measurements, etc. The integrated circuit may include a chip storing program instructions in firmware form, a digital signal processor (DSP), a chip defined as an application-specific integrated circuit (ASIC), and / or one or more microprocessors or microcontrollers that execute program instructions (e.g., software). Program instructions may be instructions delivered to the controller or system in various settings (or program files) that define operational parameters for performing specific processes on or for a semiconductor wafer. In some embodiments, these operational parameters may be part of a recipe defined by a process engineer to perform one or more processing steps in the fabrication process of one or more layers, materials, metals, oxides, silicon, silica, surfaces, circuits, and / or bare dies of a wafer.

[0175] In some implementations, controller 1229 may be part of or coupled to a computer integrated with, coupled to, or network-connected to the system or a combination thereof. For example, controller 1229 may be in the “cloud” or be a whole or part of a fab mainframe computer system, allowing remote access to wafer processing. The computer may enable remote access to the system to monitor the current progress of manufacturing operations, examine the history of past manufacturing operations, examine trends or performance criteria of multiple manufacturing operations to change parameters of the current process, set processing steps to follow the current process, or start a new process. In some embodiments, the remote computer (e.g., a server) may provide process recipes to the system via a network, which may include a local network or the Internet. The remote computer may include a user interface that allows input or programming of parameters and / or settings, which are then transmitted from the remote computer to the system. In some examples, the controller receives instructions in the form of data specifying parameters for each processing step to be performed during one or more operations. It should be understood that these parameters may be specific to the type of process to be performed and the type of tool, to which the controller is configured to connect to or control. Therefore, as described above, the controller can be distributed, for example, by comprising one or more discrete controllers connected together via a network and working toward a common goal (e.g., the process and control described herein). An example of a distributed controller for these purposes would be one or more integrated circuits located indoors that communicate with one or more remote integrated circuits (e.g., at the platform level or as part of a remote computer), which together control the process within the indoor environment.

[0176] Exemplary systems may include, but are not limited to, plasma etching chambers or modules, deposition chambers or modules, rotary rinsing chambers or modules, metal plating chambers or modules, cleaning chambers or modules, chamfering edge etching chambers or modules, physical vapor deposition (PVD) chambers or modules, CVD chambers or modules, ALD chambers or modules, atomic layer etching (ALE) chambers or modules, ion implantation chambers or modules, orbital chambers or modules, and any other semiconductor processing systems that may be associated with or used in the fabrication and / or manufacturing of semiconductor wafers.

[0177] As described above, depending on one or more process steps the tool is to perform, the controller may communicate with one or more other tool circuits or modules, other tool components, combined tools, other tool interfaces, adjacent tools, neighboring tools, tools located throughout the plant, a host, another controller, or tools used in material handling that move wafer containers to and from tool locations and / or loading ports within the semiconductor manufacturing plant.

[0178] Controller 1229 may include various programs. A substrate positioning program may include program code for controlling a chamber assembly used to load the substrate onto a pedestal or chuck and controlling the spacing between the substrate and other components of the chamber, such as gas inlets and / or targets. A process gas control program may include code for controlling gas composition, flow rate, pulse duration, and optionally for allowing gas to flow into the chamber prior to deposition to stabilize the pressure within the chamber. A pressure control program may include code for controlling the pressure within the chamber by adjusting, for example, a throttle valve in the chamber's exhaust system. A heater control program may include code for controlling the current of a heating unit used to heat the substrate. Alternatively, the heater control program may control the delivery of a heat transfer gas, such as helium, to the wafer chuck.

[0179] Examples of chamber sensors that can be monitored during the deposition process include mass flow controllers, pressure sensors such as pressure gauges, and thermocouples located in a base or chuck. Appropriately programmed feedback and control algorithms, along with data from these sensors, can be used to maintain the desired process conditions.

[0180] The foregoing describes embodiments of the invention implemented in single-chamber or multi-chamber semiconductor processing tools. The apparatus and processes described herein can be used in conjunction with photolithography patterning tools or processes, for example, for the fabrication or manufacture of semiconductor devices, displays, LEDs, photovoltaic panels, etc. Typically, although not necessarily, these tools / processes will be used together or operated in a common manufacturing facility. Photolithography patterning of films typically includes some or all of the following steps, each step enabling multiple feasible tools: (1) applying a photoresist to a workpiece, i.e., a substrate, using a spin coater or spray coater; (2) curing the photoresist using a hot plate or oven or a UV curing tool; (3) exposing the photoresist to visible light or ultraviolet or X-rays using a tool such as a wafer stepper; (4) developing the photoresist to selectively remove the photoresist and thereby patterning it using a tool such as a wet cleaning station; (5) transferring the photoresist pattern onto the underlying film or workpiece using a dry or plasma-assisted etching tool; and (6) removing the photoresist using a tool such as an RF or microwave plasma stripper.

[0181] experiment

[0182] Experiment 1

[0183] Four processes were tested for depositing host tungsten at a pressure of 40 Torr and a temperature of 395 °C. In each process, atomic layer deposition (ALD) using alternating cycles of diborane (B2H6) and tungsten hexafluoride (WF6) was used to deposit host tungsten on the deposited tungsten nucleation layer. Figure 14 Exemplary pulse schemes are provided for each of these four processes. In process 1, H2 and WF6 flow into the chamber simultaneously and continuously, as during conventional chemical vapor deposition (CVD). In process 2, H2 flows continuously when WF6 is pulsed (e.g., pulsed CVD). In process 3, WF6 flows continuously when H2 is pulsed (e.g., pulsed CVD). In process 4, a method such as the one described above is used. Figure 4B In methods such as those described above (e.g., sequential CVD), H2 and WF6 are pulsed alternately. The thickness, stress, non-uniformity, and resistivity of the tungsten nucleation layer of the film deposited using each of these four methods are measured and compiled in Table 1 below.

[0184] Table 1. Resistivity and Stress

[0185]

[0186] As shown in Table 1, the tungsten film deposited using process 4 has significantly lower stress and resistivity than the films deposited using any of processes 1-3.

[0187] Experiment 2

[0188] Processes for depositing host tungsten on two substrates were experimented with. Both substrates included a titanium nitride (TiN) barrier layer and a tungsten nucleation layer deposited by alternating cycles of B₂H₆ and WF₆ via ALD. One substrate involved the deposition of host tungsten using a non-sequential CVD, which involved simultaneously exposing the substrate to WF₆ and H₂ at 300 °C. The other substrate involved the use of [reference above] Figure 4B The sequential CVD deposition of the host tungsten is described, which involves alternating pulses of WF6 and H2 at a chamber pressure of 10 Torr. The fluorine concentrations of both substrates were measured. The experimental conditions are shown in Table 2. The results are plotted in... Figure 15 middle.

[0189] Table 2. Conditions for Experiment 2

[0190]

[0191] Line 1500 represents the fluorine concentration of a substrate with tungsten deposited by non-sequential CVD. Line 1501 shows the fluorine concentration of a substrate with tungsten deposited by sequential CVD. (Approximately...) The W / TiN interface line at that location represents the interface between the tungsten nucleation layer and the TiN barrier layer. At approximately... The dashed line at the TiN / oxide interface represents the interface between the TiN barrier layer and the oxide. Note that the fluorine concentration on the y-axis of the figure is on the order of magnitude, and the fluorine concentration of 1501 in sequential CVD is significantly lower than that in non-sequential CVD—until the order of magnitude decreases between the two fluorine concentrations at a certain substrate depth.

[0192] Experiment 3

[0193] Experiments were conducted on the deposition of bulk tungsten on substrates under different pressures. Three substrates each contained a TiN barrier layer. One substrate involved the deposition of a tungsten nucleation layer by alternating cycles of B₂H₆ and WF₆ ALD at 10 Torr, followed by bulk tungsten CVD with the substrate exposed to WF₆ and H₂ at 300 °C. Another substrate involved the deposition of a tungsten nucleation layer by alternating cycles of B₂H₆ and WF₆ ALD at 10 Torr, followed by sequential CVD of bulk tungsten with alternating pulses of WF₆ and H₂ at 10 Torr. A third substrate involved the deposition of a tungsten nucleation layer by alternating cycles of B₂H₆ and WF₆ ALD at 3 Torr, followed by sequential CVD of bulk tungsten with alternating pulses of WF₆ and H₂ at 10 Torr. The fluorine concentration was measured for all three substrates. The experimental conditions are shown in Table 3. The results are plotted in... Figure 16 middle.

[0194] Table 3. Conditions for Experiment 3

[0195]

[0196] Line 1600 represents the fluorine concentration of the first substrate on which host tungsten was deposited by non-sequential CVD. The dashed line 1601 represents the fluorine concentration of the second substrate on which a nucleation layer was deposited at 10 Torr, followed by host tungsten deposition by sequential CVD. The dotted line 1603 represents the fluorine concentration of the third substrate on which a nucleation layer was deposited at 3 Torr, followed by host tungsten deposition by sequential CVD. The results show that the low-pressure nucleation layer subsequently subjected to sequential CVD (1603) exhibits a lower fluorine concentration than the second substrate (1601), even at the W / TiN interface and even in the TiN layer (on...). and The same applies between (the two). This indicates that, due to the reduced amount of fluorine concentration in the tungsten film, the amount of fluorine diffusing into the TiN layer and oxides also decreases.

[0197] Experiment 4

[0198] Experiments were conducted on processes for depositing bulk tungsten on substrates using different combinations of tungsten deposition methods. Three substrates were compared. One substrate included... thermal oxides, TiN, ALD deposited using alternating pulses of WF6 and B2H6 at 3 Torr. A tungsten nucleation layer, and a host tungsten substrate deposited by pulsed sequential CVD using WF6 and H2 at 10 Torr. The fluorine concentration of the substrate is determined by... Figure 17 The dashed number 1712 indicates this. Another substrate includes... thermal oxides, TiN, Fluorine-free tungsten, using alternating pulses of WF6 and B2H6, ALD deposited at 3 Torr. A tungsten nucleation layer, and a bulk tungsten substrate deposited at 10 Torr using pulsed sequential CVD with WF6 and H2. The fluorine concentration of this second substrate is determined by... Figure 17 Line 1711 in the diagram indicates that the third type of substrate includes... Oxides deposited on TEOS Fluorine-free tungsten, ALD deposited using alternating pulses of WF6 and B2H6 under 3 Torrard conditions A tungsten nucleation layer, and a host tungsten substrate deposited by pulsed sequential CVD using WF6 and H2 at 10 Torr. The fluorine concentration of the substrate is determined by... Figure 17 The dotted line 1713 indicates the layer. The layers deposited on each substrate in this experiment are summarized in Table 4.

[0199] Table 4. Conditions for Experiment 4

[0200]

[0201] like Figure 17 As shown, films deposited using a combination of fluorine-free tungsten, low-pressure nucleation layers, and sequential CVD exhibit lower fluorine concentrations and less fluorine diffusion (see Depth greater than...). (Lines 1711 and 1713 beyond the W / TiN interface). For depositing more fluorine-free tungsten films on the substrate, near the nucleation layer, between and The fluorine concentration was lowest between these values, while the bulk tungsten of the film deposited using sequential CVD and low-pressure nucleation without the use of a fluorine-free tungsten layer was approximately [missing information]. and The fluorine concentrations were low (see line 1712). These results suggest that the combination of fluorine-free tungsten deposition and sequential CVD of tungsten may result in tungsten films with extremely low fluorine concentrations and reduced fluorine diffusion.

[0202] Experiment 5

[0203] Experiments were conducted to investigate the relationship between process films deposited via sequential CVD combined with low-pressure nucleation layer deposition and process films deposited via sequential CVD combined with high-pressure nucleation layer deposition. One substrate comprises a tungsten nucleation layer deposited using ALD with alternating cycles of WF6 and B2H6 at 10 Torr, and a process film deposited via sequential CVD combined with high-pressure nucleation layer deposition as described above. Figure 4B The sequential CVD method used alternating pulses of WF6 and H2 to deposit the bulk tungsten at 10 Torr. The stress and resistivity of the films were measured at different thicknesses, and... Figure 18A and 18B The diagram shows line 1801, "Low-Pressure Nucleation". Another substrate includes a tungsten nucleation layer deposited using alternating cycles of WF6 and B2H6 in an ALD process at 40 Torr, and is constructed as described above according to... Figure 4B The sequential CVD method used alternating pulses of WF6 and H2 to deposit the bulk tungsten at 10 Torr. The stress and resistivity of the films were measured at different thicknesses, and... Figure 18A and 18B The figure is shown as line 1802, "high-pressure nucleation". The conditions for nucleation and host layer deposition are shown in Table 5.

[0204] Table 5. Conditions for Experiment 5

[0205]

[0206] As the results show, substrates with nucleation layers deposited under low pressure have significantly lower stress than substrates with nucleation layers deposited under high pressure, while the resistivity remains approximately the same.

[0207] Experiment 6

[0208] Experiments were conducted to investigate the relationship between process films deposited via sequential CVD combined with low-temperature nucleation layer deposition and process films deposited via sequential CVD combined with high-temperature nucleation layer deposition. One substrate comprises a tungsten nucleation layer deposited using ALD with alternating cycles of WF6 and B2H6 at 10 Torr and 250 °C, and a process film deposited via sequential CVD combined with high-temperature nucleation layer deposition as described above. Figure 4B The sequential CVD method used alternating pulses of WF6 and H2 to deposit the bulk tungsten at 10 Torr. The stress and resistivity of the films were measured at different thicknesses, and... Figure 19A and 19B The diagram shows line 1902, "Low-T nucleation". Another substrate includes a tungsten nucleation layer deposited using ALD deposition of WF6 and B2H6 alternately cycled at 10 Torr and 300°C, and processed as described above according to... Figure 4B The sequential CVD method used alternating pulses of WF6 and H2 to deposit the bulk tungsten at 10 Torr. The stress and resistivity of the films were measured at different thicknesses, and... Figure 19A and 19B The nucleation is shown as line 1904, "high-T nucleation". The conditions for nucleation and host layer deposition are shown in Table 6.

[0209] Table 6. Conditions for Experiment 6

[0210]

[0211] As the results show, substrates with nucleation layers deposited at low temperatures exhibit significantly lower stress than substrates with nucleation layers deposited at high temperatures, while the resistivity of films deposited at higher temperatures is slightly lower than that of films deposited at lower temperatures. These results indicate that combining lower-temperature nucleation layer deposition with sequential CVD bulk deposition can significantly reduce film stress.

[0212] Experiment 7

[0213] Experiments were conducted on bWL filling with and without nitrogen addition. During hydrogen exposure, nitrogen was added to repeat the following cycle: tungsten precursor exposure, argon purging, reducing agent hydrogen exposure, and argon purging. For substrates without nitrogen, deposition was performed at 430°C, including repeated deposition cycles of nucleation layer deposition and tungsten deposition as described above. Line bending analysis was performed by measuring the linewidth and roughness of the trenches filled with metal (i.e., tungsten). Line bending analysis involved imaging the metal at the top of the device opening using a planar viewing microscope and measuring the metal width at multiple points along multiple lines. For each line, the linewidth was measured between 100 points. The average linewidth and linewidth variation were then calculated for each line, which is sometimes defined as roughness. The “average linewidth” is the average of the average linewidths of all individual lines measured during the analysis. For line curvature, two main metrics are defined as follows: (i) Line-to-line (LTL) variation is the standard deviation of the average linewidth, thus capturing the variable of linewidth variation across different lines in the image; and (ii) Linewidth roughness (LWR) is the average of the line roughness (the variable of linewidth within each line) from all measured lines, thus capturing the average linewidth variable within a single line. These two metrics, LTL and LWR, are combined into a univariate metric, total σ, as described above. Furthermore, LTL and total σ are normalized relative to the average linewidth and described as LTL% and total σ% respectively. Examples of these calculations are described in Table 7 below.

[0214] For the substrate (in which a nitrogen flow is introduced during hydrogen exposure as a reducing agent, with 50% of the flow rate being nitrogen), deposition was performed at 435 °C. A nucleation layer was deposited, followed by cycles of a tungsten precursor, argon purging, hydrogen and nitrogen co-flow, and argon purging. The resulting average linewidth, LTL, LWR, total variance, LTL percentage, and total variance percentage are described in Table 7 below.

[0215] Table 7. Results of Experiment 7

[0216]

[0217] These results are based on cross-sectional and top-down SEM image analysis. The pulse sequence used involved N2 exposure during the H2 conversion process. The results show good filling and minimized line bending at 50% N2 in the bulk process. Line bending in the substrate using N2 was significantly less than that found in the substrate without N2.

[0218] in conclusion

[0219] While the embodiments described above have been described in detail to a certain extent for clarity of understanding, it will be apparent that certain variations and modifications may be implemented within the scope of the appended claims. It should be noted that there are many alternative processes, systems, and apparatuses for implementing embodiments of the invention. Therefore, the embodiments of the invention should be considered illustrative rather than restrictive, and are not limited to the details given herein.

Claims

1. A method for filling features on a substrate to form a line, the method comprising: (a) A substrate having a plurality of spaced features, wherein each of the plurality of features is a trench, a hole or a via, and each feature has one or more of the following: V-shaped sidewalls, narrow and / or recessed openings, a constriction within the feature and a high aspect ratio; (b) Depositing a first amount of metal in the plurality of features on the substrate; (c) After depositing the first amount of metal, exposing the first amount of metal in one of the plurality of features to a suppressing gas to reduce metal-metal bonding interactions between the metals formed on the sidewalls of each feature; and (d) Depositing a second amount of metal on the first amount of metal among the plurality of features, The metal is tungsten, ruthenium, cobalt or molybdenum, and the suppressing gas is nitrogen, oxygen, ammonia or a combination thereof.

2. The method of claim 1, wherein each of the plurality of features is a feature having an opening and a vertically oriented bottom.

3. The method of claim 2, wherein each of the plurality of features narrows from the opening to the bottom of the feature.

4. The method of claim 1, wherein reducing metal-metal bonding interactions comprises exposing the feature to a nitrogen-containing gas.

5. The method according to claim 1, wherein the metal is molybdenum.

6. The method of claim 1, wherein the plurality of features have dielectric sidewalls.

7. The method of claim 1, wherein reducing metal-metal bonding interactions comprises exposing the plurality of features to ammonia.

8. The method of claim 1, wherein during the deposition of the second amount of metal, metal-metal bonding interactions between the metals formed on the sidewalls of each feature are reduced.

9. The method of claim 1, wherein metal-metal bonding interactions between the metals formed on the sidewalls of each feature are reduced prior to depositing the second amount of metal.

Images