Method for depositing tungsten and other metal in 3D NAND structure
Patent Information
- Application Number
- JP2024114404
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2018-05-03
- Filing Date
- 2024-07-18
- Publication Date
- 2024-11-07
- Estimated Expiration
- 2039-05-03
AI Technical Summary
Conventional tungsten deposition methods face challenges in filling complex high aspect ratio structures such as 3D NAND structures, leading to issues like fluorine diffusion, electromigration, and high tensile stress, which affect device performance and reliability.
A method involving multiple deposition cycles with alternating pulses of reducing agents like hydrogen and inert gas, followed by metal precursors such as tungsten chloride, to deposit tungsten or molybdenum into features, minimizing erosion and stress while ensuring uniform filling.
The method achieves fluorine-free tungsten deposition with reduced tensile stress, improving the filling of complex structures and enhancing device reliability by preventing fluorine diffusion and substrate erosion.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
[Technical field]
[0001] [Cited by reference] The PCT application is filed herewith as a part of this application. Each application to which this application claims benefit or priority, as listed in the PCT application filed herewith, is hereby incorporated by reference in its entirety for all purposes. [Background technology]
[0002] The deposition of tungsten-containing materials is an essential part of many semiconductor manufacturing processes. Such materials may be used for horizontal interconnects, vias between adjacent metal layers, and contacts between metal layers and devices. In a conventional tungsten deposition process, a substrate is heated to a processing temperature in a vacuum chamber to deposit a very thin portion of tungsten film that acts as a seed layer (also called a nucleation layer). The remaining portion of the tungsten film (called the bulk layer) is then deposited on top of the nucleation layer by exposing the substrate to two reactants simultaneously in a chemical vapor deposition (CVD) process. In general, the bulk layer is deposited at a faster rate than the nucleation layer. However, as the industry shrinks devices and employs more complex patterning schemes, the deposition of thin tungsten films becomes a challenge. Deposition into complex high aspect ratio structures such as 3D NAND structures is particularly challenging. Summary of the Invention [Means for solving the problem]
[0003] Provided herein are methods and apparatus for filling features with metal-containing materials. One aspect of the disclosure relates to a method for filling a structure with a metal-containing material, the method comprising preparing a structure to be filled with a metal-containing material and exposing the structure to a plurality of deposition cycles, each deposition cycle comprising exposure to one or more alternating reducing agent (e.g., hydrogen (H2)) dose pulses / inert gas purge pulses, followed by exposure to one or more alternating metal precursor dose pulses and inert gas purge pulses. The metal may be tungsten (W) or molybdenum (Mo) in some embodiments.
[0004] In some embodiments, the structure is a partially fabricated three-dimensional (3D) NAND structure having a sidewall and a plurality of openings in the sidewall leading to a plurality of features having a plurality of interior regions fluidly accessible through the openings. In some embodiments, the metal precursor is a chlorine-containing metal precursor, such as tungsten hexachloride, tungsten pentachloride, tungsten tetrachloride, molybdenum pentachloride, molybdenum dioxide, molybdenum oxide tetrachloride, and mixtures thereof. In some embodiments, the pulse of chlorine-containing metal precursor comprises between about 0.1% and about 5.0% chlorine-containing metal precursor by volume. In some embodiments, the exposure to multiple alternating metal precursor and inert gas purge pulses comprises turning off the inert gas purge flow during the metal precursor pulses. In some embodiments, the duration of the inert gas purge pulse is at least 1.5 times the duration of the metal precursor pulse. In some embodiments, each deposition cycle comprises at least 5 or at least 10 alternating metal precursor and inert gas purge pulses. In some embodiments, each deposition cycle includes only one H2 pulse. In other embodiments, each deposition cycle includes multiple alternating H2 and inert gas pulses.
[0005] Another aspect of the present disclosure relates to a method of filling a structure with a metal-containing material, the method comprising preparing a structure to be filled with a metal-containing material and exposing the structure to multiple deposition cycles, each deposition cycle comprising exposure to a reducing agent (e.g., hydrogen (H2)) dose pulse followed by exposure to an inert gas pulse and further exposure to multiple alternating metal precursor dose pulses and inert gas purge pulses. In some embodiments, the structure is a partially fabricated three-dimensional (3D) NAND structure having a sidewall and multiple openings in the sidewall leading to multiple features having multiple interior regions fluidly accessible through the openings. In some embodiments, the metal precursor is a chlorine-containing metal precursor, such as tungsten hexachloride, tungsten pentachloride, tungsten tetrachloride, molybdenum pentachloride, molybdenum dioxide dichloride, molybdenum oxide tetrachloride, and mixtures thereof. In some embodiments, the pulse of chlorine-containing metal precursor comprises between about 0.1% and about 5.0% chlorine-containing metal precursor by volume. In some embodiments, the exposure to multiple alternating metal precursor and inert gas purge pulses includes turning off the inert gas purge flow during the metal precursor pulse. In some embodiments, the duration of the inert gas purge pulse is at least 1.5 times the duration of the metal precursor pulse. In some embodiments, each deposition cycle includes at least 5 or at least 10 alternating metal precursor and inert gas purge pulses.
[0006] Another aspect of the disclosure relates to an apparatus comprising one or more process chambers each configured to hold a substrate, one or more process gas inlets for coupling to a reducing agent (e.g., hydrogen (H2)) gas source, a metal precursor gas source, and an inert purge gas source, and a controller for controlling operation in the apparatus, the controller including machine-readable instructions for performing a plurality of deposition cycles, each deposition cycle including injecting a hydrogen (H2) pulse into the one or more process chambers via the one or more process gas inlets, injecting an inert purge gas pulse into the one or more process chambers via the one or more process gas inlets after the injection of the H2 pulse, and injecting a plurality of alternating metal precursor and inert gas purge pulses into the one or more process chambers via the one or more purge gas inlets after the injection of the inert purge gas pulse. In some embodiments, the metal precursor is a chlorine-containing metal precursor. In some embodiments, the instructions include instructions for turning off the inert gas purge flow during the metal precursor pulse. In some embodiments, the duration of the inert gas purge pulse is at least 1.5 times the duration of the metal precursor pulse. In some embodiments, each deposition cycle includes at least 5 alternating metal precursor and inert gas purge pulses, hi some embodiments, each deposition cycle includes at least 10 alternating metal precursor and inert gas purge pulses.
[0007] Another aspect of the disclosure relates to a method of filling a structure with a metal-containing material, the method comprising: preparing a structure to be filled with a metal-containing material; and exposing the structure to multiple deposition cycles, each deposition cycle comprising exposure to multiple alternating hydrogen (H2) pulses followed by an inert gas pulse, and further exposure to a metal precursor pulse followed by an inert gas purge pulse. In some embodiments, the structure is a partially fabricated three-dimensional (3D) NAND structure having a sidewall and multiple openings in the sidewall leading to multiple features having multiple interior regions fluidly accessible through the openings. In some embodiments, the metal precursor is a chlorine-containing metal precursor. In some embodiments, the pulses of chlorine-containing metal precursor comprise between about 0.1% and about 5.0% chlorine-containing metal precursor by volume. In some embodiments, the exposure to multiple alternating H2 and inert gas purge pulses comprises turning off the inert gas purge flow during the metal precursor pulses. In some embodiments, each deposition cycle comprises only one metal precursor pulse. In other embodiments, each deposition cycle includes multiple alternating metal precursor and inert gas pulses.
[0008] Another aspect of the present disclosure relates to an apparatus comprising one or more process chambers each configured to hold a substrate, one or more process gas inlets for coupling to a hydrogen (H2) gas source, a metal precursor gas source, and an inert purge gas source, and a controller for controlling operation in the apparatus, the controller including machine-readable instructions for performing a plurality of deposition cycles, each deposition cycle including injecting a plurality of alternating H2 pulses and inert gas purge pulses into one or more process chambers via one or more purge gas inlets, and injecting a metal precursor pulse followed by an inert gas pulse. In some embodiments, the metal precursor is a chlorine-containing metal precursor. In some embodiments, the instructions include instructions for turning off the inert gas purge flow during the H2 pulse.
[0009] These and other aspects are further described below with reference to the drawings. [Brief description of the drawings]
[0010] [Figure 1A] FIG. 1A is a schematic diagram of an exemplary film on a substrate.
[0011] [Figure 1B] FIG. 1B is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1C] FIG. 1C is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1D] FIG. 1D is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1E] FIG. 1E is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1F] FIG. 1F is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1G] FIG. 1G is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1H] FIG. 1H is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1I] FIG. 1I is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments. [Figure 1J] FIG. 1J is a schematic example of various structures onto which tungsten or molybdenum may be deposited according to some disclosed embodiments.
[0012] [Figure 2A]FIG. 2A is a process flow diagram illustrating the operation of methods according to some disclosed embodiments. [Figure 2B] FIG. 2B is a process flow diagram illustrating the operation of methods according to some disclosed embodiments. [Figure 2C] FIG. 2C is a process flow diagram illustrating the operation of methods according to some disclosed embodiments.
[0013] [Diagram 3] FIG. 3 is a timing sequence diagram illustrating an example cycle in a method for depositing a film according to some disclosed embodiments.
[0014] [Figure 4A] FIG. 4A is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4B] FIG. 4B is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4C] FIG. 4C is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4D] FIG. 4D is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4E] FIG. 4E is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4F] FIG. 4F is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4G] FIG. 4G is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4H] FIG. 4H is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4I] FIG. 4I is a schematic diagram of an example of a mechanism for depositing a film according to some disclosed embodiments. [Figure 4J] FIG. 4J illustrates a membrane according to some disclosed embodiments.
[0015] [Diagram 5] FIG. 5 shows a diagram of experimental results comparing tungsten filling of 3D NAND structures using a H2 / Ar / WClx / Ar sequence versus a H2 / Ar / n(WClx / Ar) sequence.
[0016] [Figure 6] FIG. 6 is a schematic diagram of an exemplary process tool for implementing some disclosed embodiments.
[0017] [Figure 7] FIG. 7 is a schematic diagram of an exemplary station for implementing some disclosed embodiments.
[0018] [Figure 8] FIG. 8 is a schematic diagram illustrating basic features of a manifold system that may be used in accordance with some embodiments. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0019] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the presented embodiments. The disclosed embodiments may be practiced without some or all of these specific details. In other instances, well-known process operations have not been described in detail in order to avoid unnecessarily obscuring the disclosed embodiments. It will be understood that, although the disclosed embodiments are described in conjunction with specific embodiments, they are not intended to limit the disclosed embodiments.
[0020] Feature filling with tungsten (W) is often used to form electrical contacts in the fabrication of semiconductor devices. Traditional methods of depositing tungsten films first deposit a tungsten nucleation layer in the via or contact. Generally, the nucleation layer is a thin conformal layer that functions to facilitate the subsequent formation of bulk material over it. The tungsten nucleation layer may be deposited to conformally cover the sidewalls and bottom of the feature. Conformity to the bottom and sidewalls of the underlying feature may be important to support high quality deposition. Nucleation layers are often deposited using atomic layer deposition (ALD) or pulsed nucleation layer (PNL) techniques.
[0021] In the PNL process, pulses of reactants are injected sequentially and then purged from the reaction chamber, typically with a pulse of purge gas between reactants. The first reactant is allowed to adsorb on the substrate and react with the next reactant. This process is repeated cyclically until the desired thickness is achieved. The PNL process is similar to the ALD process. PNL is generally distinguished from ALD by its higher operating pressure range (greater than 1 Torr) and higher growth rate per cycle (greater than one atomic layer of film growth per cycle). Chamber pressures during PNL deposition can range from about 1 Torr to about 400 Torr. In the context of the description presented herein, PNL broadly embodies any cyclic process of sequentially adding reactants for reaction on a semiconductor substrate. Thus, this concept embodies the technique traditionally referred to as ALD. In the context of the disclosed embodiments, chemical vapor deposition (CVD) embodies a process in which reactants are introduced together into a reactor for gas-phase or surface reaction. PNL and ALD processes are distinct from CVD processes, and vice versa.
[0022] After depositing the tungsten nucleation layer, bulk tungsten is deposited, typically by a CVD process, by reducing tungsten hexafluoride (WF6) with a reducing agent such as hydrogen (H2). Bulk tungsten is different from tungsten nucleation layer. Bulk tungsten as used herein refers to tungsten used to fill most or all of a feature, such as at least about 50% of the feature. Unlike a nucleation layer, which is a conformal thin film that functions to facilitate the subsequent formation of bulk material thereon, bulk tungsten is used to carry electrical current. It may be characterized by larger grain size and lower resistivity compared to the nucleation film. In various embodiments, bulk tungsten is tungsten deposited to a thickness of at least 50 Å.
[0023] As devices scale to smaller technology nodes and more complex patterning structures are used, various challenges arise in tungsten filling. Traditional tungsten deposition involves the use of tungsten hexafluoride (WF6), a fluorine-containing precursor. However, the use of WF6 results in some incorporation of fluorine in the deposited tungsten film. The presence of fluorine can cause electromigration and / or fluorine diffusion to adjacent components and can damage contacts, thereby reducing device performance. One challenge is reducing the fluorine content in the deposited tungsten film. The impact of a particular fluorine concentration becomes greater as feature size decreases. This is because smaller features result in thinner films being deposited and fluorine in the deposited tungsten film has a greater chance of diffusing through the thinner film.
[0024] One method to prevent fluorine diffusion is to deposit one or more barrier layers prior to depositing tungsten to prevent fluorine diffusion from tungsten to other layers of the substrate, such as oxide layers. For example, FIG. 1A shows an exemplary stack of layers deposited on a substrate. Substrate 190 includes silicon layer 192, oxide layer 194 (e.g., titanium oxide (TiOx), tetraethyl orthosilicate (TEOS) oxide, etc.), barrier layer 196 (e.g., titanium nitride (TiN)), tungsten nucleation layer 198, and bulk tungsten layer 199. Barrier layer 196 is deposited to prevent fluorine diffusion from bulk tungsten layer 199 and tungsten nucleation layer 198 to the oxide layers. However, as devices shrink, the barrier layer becomes thinner and fluorine can still diffuse from the deposited tungsten layer. Performing the chemical vapor deposition of bulk tungsten at higher temperatures results in lower fluorine content, but such films may have poor step coverage.
[0025] Another challenge is to reduce stress in the film after deposition. Thinner tungsten films tend to have higher tensile stress. Conventional techniques for depositing bulk tungsten films by chemical vapor deposition have tensile stresses of over 2.5 GPa for 200 Å films. High thermal tensile stresses cause the substrate to curl, which makes subsequent processing difficult. For example, subsequent processes may include chemical mechanical planarization, material deposition, and / or clamping the substrate to a substrate holder to perform processes in a chamber. However, these processes often rely on the substrate being flat, and a curled substrate results in non-uniform processing or makes it impossible to process the substrate. Although there are existing methods for reducing stress in films of other materials, such as annealing, tungsten does not have the surface mobility to allow for particle movement or change after deposition due to its high melting point.
[0026] To avoid such reliability and integration problems or device performance problems, fluorine-free tungsten (FFW) precursors are useful. Current FFW precursors include organometallic precursors, but undesirable trace elements from the organometallic precursors, such as carbon, hydrogen, nitrogen, and oxygen, may be incorporated into the tungsten film. In addition, some fluorine-free organometallic precursors are not easy to implement or integrate in tungsten deposition processes.
[0027] Some implementations described herein use tungsten chloride (WCl x The present invention relates to the deposition of tungsten using a tungsten chloride precursor. Tungsten chlorides include tungsten pentachloride (WCl5), tungsten hexachloride (WCl6), tungsten tetrachloride (WCl4), tungsten dichloride (WCl2), and mixtures thereof. Although the examples herein refer to WCl5 and WCl6 as examples, it is understood that other tungsten chlorides may be used in the disclosed embodiments. Films deposited using some of the disclosed embodiments are fluorine-free. Some of the disclosed embodiments relate to the deposition of bulk tungsten using alternating pulses of chlorine-containing tungsten precursors and hydrogen.
[0028] Deposition with WCl5 and WCl6 presents challenges not present with WF6 due to the possibility of erosion by tungsten chloride. As a result of the lower reactivity of tungsten chloride, deposition is performed at higher temperatures than deposition with WF6. The evaporated WCl6 has a high enough vapor pressure to be carried into the tungsten deposition chamber. However, WCl6 may be more likely to erode the substrate than WCl5. Although WCl5 is less likely to erode the substrate, WCl5 also has a higher vapor pressure than WCl6. The lower vapor pressure is useful for depositing low resistivity tungsten films, but step coverage may be poor in some deposition operations.
[0029] The methods described herein may also be used for the deposition of molybdenum (Mo) from molybdenum chloride or molybdenum oxychloride precursors. Molybdenum may be used to form low resistance metallization stack structures and may be a substitute for tungsten.
[0030] The disclosed embodiments have a wide variety of applications: the method may be used to deposit tungsten or molybdenum with high step coverage in features, and may be used to deposit tungsten in 3D NAND structures.
[0031] The methods described herein are performed on a substrate that may be contained within the chamber. The substrate may be a silicon or other semiconductor wafer, e.g., a 200 mm wafer, a 300 mm wafer, or a 450 mm wafer, including a wafer having one or more layers of material deposited thereon, such as a dielectric material, a conductive material, or a semiconductor material. The methods are not limited to semiconductor substrates and may be performed to fill any feature with a metal, such as tungsten.
[0032] The substrate may have features, such as vias or contact holes, that may be characterized by one or more of narrow and / or reentrant openings, constrictions within the features, and high aspect ratios. The features may be formed in one or more of the above layers. For example, the features may be at least partially formed in a dielectric layer. In some embodiments, the features may 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, at least about 25:1, or even higher. One example of a feature is a hole or via in a semiconductor substrate or in a layer on the substrate.
[0033] 1B-1H are schematic examples of various structures in which metal may be deposited according to disclosed embodiments. FIG. 1B shows an example of a cross-sectional view of a vertical feature 101 to be filled with a metal such as tungsten or molybdenum. The feature may include a feature hole 105 in a substrate 103. The hole 105 or other feature may have, for example, an opening diameter or line width, in a dimension near the opening, between about 10 nm and 500 nm, for example, between about 25 nm and about 300 nm. The feature hole 105 may be referred to as an unfilled feature, or simply a feature. Such a feature 101, and any feature, may be characterized in part by an axis 118 that extends the length of the feature, with vertically oriented features having a vertical axis and horizontally oriented features having a horizontal axis.
[0034] In some embodiments, the features are word line features in a 3D NAND structure. For example, the substrate may include a word line structure having any number of word lines (e.g., 50-150) with a vertical channel at least 200 Å deep. Another example is a trench in the substrate or layer. The features may be of any depth. In various embodiments, the features may have an underlayer such as a barrier layer or an adhesion layer. Non-limiting examples of underlayers include dielectric layers and conductive layers, such as layers of silicon oxide, silicon nitride, silicon carbide, metal oxide, metal nitride, metal carbide, and metal.
[0035] FIG. 1C illustrates an example of a feature 101 having a reentrant profile. A reentrant profile is a profile that narrows from the closed bottom end or interior of the feature toward the feature opening. According to some implementations, the profile can be gradually narrowing and / or include an overhang at the feature opening. FIG. 1C illustrates an example of the latter, lining the sidewalls or interior of the feature hole 105 with an underlayer 113. The underlayer 113 can be, for example, a diffusion barrier layer, an adhesion layer, a nucleation layer, a combination thereof, or any other suitable material. Non-limiting examples of underlayers can include dielectric layers and conductive layers, including, for example, layers of silicon oxide, silicon nitride, silicon carbide, metal oxide, metal nitride, metal carbide, and metal. In specific implementations, the underlayer can be one or more of titanium, titanium nitride, tungsten nitride, titanium aluminide, tungsten, and molybdenum. In some embodiments, the underlayer is free of tungsten. In some embodiments, the underlayer is free of molybdenum. Underlayer 113 forms an overhang 115 such that underlayer 113 is thicker near the opening of feature 101 than inside feature 101 .
[0036] In some implementations, features may be filled that have one or more constrictions within the feature. FIG. 1D shows example views of various filled features with constrictions. Each of the examples (a), (b), and (c) of FIG. 1D includes a constriction 109 at a midpoint within the feature. The constriction 109 may be, for example, between about 15 nm and 20 nm wide. During deposition of tungsten or molybdenum into the feature using conventional techniques, the constriction may cause pinch-offs by preventing further deposition beyond the constriction before that portion of the feature is filled, resulting in voids within the feature. Example (b) further includes a liner / barrier overhang 115 at the feature opening. Such an overhang may also be a potential pinch-off point. Example (c) includes a constriction 112 further away from the field region than the overhang 115 of example (b).
[0037] It is also possible to fill horizontal features such as 3D memory structures. FIG. 1E shows an example of a horizontal feature 150 that includes a constriction 151. For example, the horizontal feature 150 can be a word line in a 3D NAND structure. In some implementations, the constriction can be due to the presence of a pillar in a 3D NAND or other structure. For example, FIG. 1F presents a side cross-sectional view of a 3D NAND (also called vertical NAND or VNAND) structure 110 (formed on a semiconductor substrate 103) having a VNAND stack (on the left side 125 and right side 126), a central vertical structure 130, and multiple horizontal features 120 stacked with openings 122 on opposing side walls 140 of the central vertical structure 130. It should be noted that while FIG. 1F shows two “stacks” of the presented 3D NAND structure 110 that together form a “trench-like” central vertical structure 130, in some embodiments, more than two “stacks” may be arranged side-by-side and extending spatially parallel to one another, with the gap between each pair of adjacent “stacks” forming the central vertical structure 130 as shown in FIG. 1F. In this embodiment, the horizontal features 120 are 3D memory word line features that are fluidly accessible from the central vertical structure 130 through openings 122. Although not shown explicitly in the drawings, the horizontal features 120 in both 3D NAND stacks 125 and 126 shown in FIG. 1F (i.e., the left 3D NAND stack 125 and the right 3D NAND stack 126) are also accessible from the other side of the stack (the leftmost side and the rightmost side, respectively) through a similar vertical structure formed by additional 3D NAND stacks (not shown, but to the left and right). That is, each 3D NAND stack 125, 126 includes a stack of word line features that are fluidly accessible from both sides of the 3D NAND stack through a central vertical structure 130. In the specific example shown generally in FIG. 1F, each 3D NAND stack includes six pairs of stacked word lines, although in other embodiments, the 3D NAND memory layout may include any number of vertically stacked word line pairs.
[0038] Word line features in a 3D NAND stack are typically formed by depositing an alternating stack of silicon oxide and silicon nitride layers, and then selectively removing the nitride layers, leaving a stack of oxide layers with gaps between them. These gaps are the word line features. Given the available techniques for forming the word lines, as well as the available techniques for successful (nearly) void-free filling of the vertical features, any number of word lines may be stacked vertically in such a 3D NAND structure. Thus, for example, a VNAND stack may include between 2 and 256 horizontal word line features, or between 8 and 128 horizontal word line features, or between 16 and 64 horizontal word line features, etc. (with the recited ranges being understood to include the recited endpoints).
[0039] FIG. 1G presents a horizontal cross-sectional view of the same 3D NAND structure 110 shown in side view in FIG. 1F, with the cross-section of horizontal section 160 as indicated by the horizontal dashed line in FIG. 1F. The cross-section in FIG. 1G shows several columns of pillars 155, which are shown in FIG. 1F to extend vertically from the base of the semiconductor substrate 103 to the top of the 3D NAND stack 110. In some embodiments, these pillars 155 are formed of polysilicon material and are structurally and functionally important to the 3D NAND structure 110. In some embodiments, such polysilicon pillars may function as gate electrodes for stacked memory cells formed within them. The top view in FIG. 1G shows that the pillars 155 form a constriction at the opening 122 to the word line feature 120, i.e., the pillars 155 inhibit the fluid accessibility through the opening 122 from the central vertical structure 130 to the word line feature 120 (as indicated by the arrow in FIG. 1G). In some embodiments, the size of the horizontal gap between adjacent polysilicon pillars is between about 1-20 nm. This reduced fluid accessibility increases the difficulty of uniformly filling the wordline features 120 with tungsten material. The structure of the wordline features 1 and the difficulty of uniformly filling them with tungsten or molybdenum material due to the presence of pillars 155 are further illustrated in Figures 1H, 1I, and 1J.
[0040] FIG. 1H presents a vertical cut of a 3D NAND structure similar to that shown in FIG. 1F, but in this case focusing only on a pair of wordline features 120, and further shows a schematic metal filling process resulting in the formation of a void 175 in the filled wordline feature 120. The void 175 is also shown in FIG. 1I, but in this drawing shown by a horizontal cut of the pillar 155, similar to the horizontal cut presented in FIG. 1G. FIG. 1J shows the retention of tungsten or molybdenum material around the constriction forming pillar 155, which results in a pinch-off of the opening 122, preventing further tungsten or molybdenum material from being deposited in the region of the void 175. 1H and 1I that void-free tungsten or molybdenum fill relies on a sufficient amount of deposition precursor migrating through vertical structure 130, through opening 122, past constriction-forming pillar 155, and into the innermost part of wordline feature 120 before tungsten deposition builds up around pillar 155, causing pinch-off of opening 122 and impeding further precursor migration into wordline feature 120. Similarly, FIG. 1J presents only one wordline feature 120 in a cross-sectional view from above, showing how the interior pinch-off of wordline feature 120 begins due to the generally conformal deposition of tungsten or molybdenum material causing a significant width of pillar 155 to act to partially close and / or narrow and / or constrict what would otherwise be an open through wordline feature 120. (Note that the example of FIG. 1J can be understood as a 2D rendering of the 3D features of the pillar constriction structure shown in FIG. 1I, and thus shows the constriction as it would be seen in plan view rather than in cross section.)
[0041] Three-dimensional structures may require longer and / or more intensive exposure to the precursors to allow filling of the innermost regions. xThis can be particularly challenging when using precursors whose tendency to corrode means that longer and more intensive exposure can cause more corrosion in some parts of the structure. These challenges can also be present in the case of molybdenum chloride precursors.
[0042] Examples of feature filling for horizontally and vertically oriented features are described below. It should be noted that the examples are often applicable to either horizontally or vertically oriented features. It should also be noted that in the following description, the term "lateral" may be used to refer to a direction generally perpendicular to the feature axis, and the term "vertical" may be used to refer to a direction generally along the feature axis.
[0043] According to various embodiments, the methods described below involve exposing a structure to multiple reducing agent / purge cycles and / or multiple metal precursor / purge cycles within a single ALD cycle. In some implementations, one or more of the following advantages may be realized: In some embodiments, multiple tungsten-containing precursor doses (e.g., W / Ar / W / Ar / W / Ar / W / Ar, where W represents a pulse of tungsten-containing precursor and Ar represents a pulse of argon purge gas) interspersed with purge gas pulses allow the tungsten-containing precursor to be driven deep into 3D NAND wordline structures or other difficult to fill structures. In some implementations, one or more of the following advantages may be realized: In some embodiments, multiple molybdenum-containing precursor doses (e.g., Mo / Ar / Mo / Ar / Mo / Ar / Mo / Ar, where Mo represents a pulse of molybdenum-containing precursor and Ar represents a pulse of argon purge gas) interspersed with purge gas pulses allow for the delivery of the molybdenum-containing precursor deep within 3D NAND wordline structures or other difficult to fill structures.
[0044] An inert gas purge pulse after each metal-containing pulse can ensure that one atomic layer of the metal-containing precursor is adsorbed throughout the structure. In some embodiments, multiple pulses, each followed by a purge gas, can ensure that multiple atomic layers of the metal-containing precursor do not accumulate in more accessible parts of the structure. This can prevent undesired erosion of the structure. In some embodiments, alternating pulses can be used to pressurize the metal precursor bath before each pulse. This can promote flux to the bottom of the feature. In some embodiments, multiple reducing agent dose pulses (e.g., H2 / Ar / H2 / Ar / W / Ar / W / Ar, where H2 is the reducing agent, W represents a pulse of a tungsten-containing precursor, and Ar represents a pulse of argon purge gas, or H2 / Ar / H2 / Ar / Mo / Ar / Mo / Ar, where Mo represents a pulse of a molybdenum-containing precursor) interspersed with purge gas pulses allow sufficient time for the reducing agent to react with the adsorbed chloride and allows the reaction products (e.g., HCl) to leave the surface to make way for the reducing agent in the next pulse.
[0045] Although the following description focuses on feature filling with tungsten and feature filling with molybdenum, aspects of the present disclosure may be implemented in filling features with other materials. For example, feature filling using one or more techniques described herein may be used to fill features with other materials, such as other tungsten-containing materials (e.g., tungsten nitride (WN), tungsten carbide (WC)), titanium-containing materials (e.g., titanium (Ti), titanium nitride (TiN), titanium silicide (TiSi), titanium carbide (TiC), titanium aluminide (TiAl)), tantalum-containing materials (e.g., tantalum (Ta), tantalum nitride (TaN)), nickel-containing materials (e.g., nickel (Ni), nickel silicide (NiSi)), and the like. Additionally, the methods and apparatus disclosed herein are not limited to feature filling, but may be used to deposit tungsten on any suitable surface, such as to form a blanket film on a flat surface.
[0046] 2A-2C provide a process flow diagram for a method performed in accordance with disclosed embodiments. As described below, the method is performed to fill structures on a substrate with tungsten. Example structures are described above with reference to FIGS. 1B-1H.
[0047] The temperature may vary depending on the chemical employed. x and W.O. x For Cl4 precursors, some disclosed embodiments may be carried out at a substrate temperature between about 400° C. and about 600° C., such as about 525° C. By substrate temperature, it is meant the temperature set on the pedestal that holds the substrate.
[0048] MoCl x and MoO x Cl y For precursors, some disclosed embodiments may be carried out at a substrate temperature between about 400° C. and about 600° C., such as about 525° C. By substrate temperature, it is meant the temperature set on the pedestal that holds the substrate.
[0049] Some disclosed embodiments may be performed at a chamber pressure between about 3 Torr and about 60 Torr. x The chamber pressure for tungsten deposition using the precursor is between 5 Torr and 20 Torr, for example 10 Torr.
[0050] Referring to FIG. 2A, in operation 200, a structure to be filled with tungsten or molybdenum is exposed to a reducing agent pulse. In some embodiments, the reducing agent pulse is hydrogen (H2). Other reducing agents may be used, such as silane, borane, germane, phosphine, hydrogen-containing gases, and combinations thereof. In various embodiments, bulk tungsten deposition or bulk molybdenum deposition is performed using hydrogen as the reducing agent. The reducing agent is pulsed without flowing other reactants. In some embodiments, a carrier gas may be flowed. In some embodiments, in operation 200, a carrier gas such as nitrogen (N2), argon (Ar), helium (He), or other inert gas may be flowed.
[0051] Operation 200 may be performed for any suitable duration, including between about 0.25 seconds and about 30 seconds, between about 0.25 seconds and about 20 seconds, between about 0.25 seconds and about 5 seconds, or between about 0.5 seconds and about 3 seconds.
[0052] In operation 202, the chamber is purged to remove excess hydrogen that did not adsorb to the substrate surface. Purging may be performed by flowing an inert gas at a constant pressure to reduce the chamber pressure and repressurizing the chamber before beginning exposure to another reactant gas. Examples of inert gases include nitrogen (N2), argon (Ar), helium (He), and mixtures thereof. Purging may be performed for a duration between about 0.25 seconds and about 30 seconds, between about 0.25 seconds and about 20 seconds, between about 0.25 seconds and about 5 seconds, or between about 0.5 seconds and about 3 seconds.
[0053] Operations 200 and 202 are then repeated until performed k times, where k is an integer that is at least 21. In one example, the duration of operation 200 may be 0.5 seconds and the duration of operation 202 between reducing agent pulses may be 1 second. For k=10, the total tungsten precursor time is 0.5×10=5 seconds and the total purge time is 1×10=10 seconds.
[0054] In operation 204, the substrate is exposed to a tungsten precursor or a molybdenum precursor. Exemplary precursors include chlorine-containing tungsten precursors such as tungsten chloride and tungsten oxychloride. Tungsten chloride is represented by WCl, where x is an integer between 2 and 6, inclusive, such as 2, 3, 4, 5, or 6. x Examples include WCl5 and WCl6. Chlorine-containing tungsten precursors include WCl x Tungsten oxychloride may include a mixture of compounds, where x and y are numbers greater than 0, such as WO x Cl y To deposit molybdenum, molybdenum precursors such as molybdenum chloride and molybdenum oxychloride are used, including molybdenum pentachloride (MoCl5), molybdenum dioxide dichloride (MoO2Cl2), and molybdenum oxide tetrachloride (MoOCl4).
[0055] In some embodiments, operation 206 may flow a carrier gas, such as nitrogen (N2), argon (Ar), helium (He), or other inert gas.
[0056] Operation 204 may be performed at any suitable temperature for any suitable duration. In some examples, operation 206 may be performed for a duration between about 0.25 seconds and about 30 seconds, between about 0.25 seconds and about 20 seconds, between about 0.25 seconds and about 5 seconds, or between about 0.5 seconds and about 3 seconds. This operation may be performed for a duration sufficient to saturate active sites on the substrate surface in some embodiments. In some embodiments, the precursor may be diverted for gas line filling and line changing before dosing. The carrier gas may be any of those described above with respect to operation 202.
[0057] According to various embodiments, in operation 204, some WCl x may react with the H2 remaining on the surface from operation 200, resulting in some WCl x may not completely react with the H2 remaining on the surface. Also, in various embodiments, some H2 may be reacted with the WCl x Sometimes there is no reaction at all with WCl x Instead, H may physisorb onto the substrate surface in the absence of H physisorbed or remaining on the substrate surface. In some embodiments, H may remain on the substrate surface but may not be physisorbed or chemisorbed to the surface. Operation 204 of FIG. 2A may thereby form a subatomic layer of tungsten in some embodiments. Similarly, operation 204 may form a subatomic layer of molybdenum in some embodiments.
[0058] To purge excess precursor that has not reacted with hydrogen or other reducing agent on the surface of the feature and is still in the gas phase, a purge is performed in operation 206. Purging may be performed by reducing the chamber pressure by flowing in an inert gas at a constant pressure and repressurizing the chamber before beginning exposure to another gas.
[0059] The chamber may be purged for any suitable duration. The chamber may be purged for a duration between about 0.25 seconds and about 30 seconds, between about 0.25 seconds and about 20 seconds, between about 0.25 seconds and about 5 seconds, or between about 0.5 seconds and about 3 seconds. In some embodiments, the purge duration is between about 0.1 seconds and about 2 seconds, which reduces the amount of WCl on the tungsten surface. x The low adsorption rate of WCl x or other precursors from being completely removed. In some embodiments, the purge duration is between about 0.1 seconds and about 15 seconds, such as about 7 seconds. For example, for the fabrication of 3D NAND structures, the chamber may be purged for about 2 seconds in operation 206. The purge gas may be any of the gases described above with respect to operation 202.
[0060] Operations 204 and 206 are then repeated until performed n times, where n is an integer that is at least 2. As will be further described below, by performing multiple tungsten pulses separated by purges in a deposition sequence, undesired erosion at the top and edges of the structures can be reduced. In one example, the duration of operation 204 can be 0.7 seconds and the duration of operation 206 between tungsten precursor pulses can be 2 seconds. For n=10, the total tungsten precursor time is 0.7×10=7 seconds and the total purge time is 2×10=20 seconds.
[0061] In operation 208, it is determined whether the tungsten or molybdenum layer has been deposited to an appropriate thickness. If not, operations 200-206 are repeated until a tungsten or molybdenum layer of the desired thickness is deposited on the surface of the feature. Each repetition of operations 200-206 may be referred to as a "cycle." In some embodiments, the order of operations 200 / 202 and 204 / 206 may be reversed such that the tungsten or molybdenum precursor is introduced prior to the introduction of the reducing agent.
[0062] Figure 2B presents a process flow diagram for an embodiment in which each ALD cycle includes only one reducing agent pulse and multiple tungsten or molybdenum precursor pulses, such that operations 200 and 202 are not repeated during the cycle. Figure 2C presents a process flow diagram for an embodiment in which each ALD cycle includes only one tungsten or molybdenum precursor pulse and multiple reducing agent pulses, such that operations 204 and 206 are not repeated during the cycle. The process blocks in Figures 2B and 2C are otherwise as described above with respect to Figure 2A.
[0063] Figure 3 shows the relationship between H2 and WCl x 3 presents a timing sequence diagram illustrating exemplary deposition cycles 311A and 311B in a process for depositing tungsten using H2. FIG. 3 illustrates a H2 pulse in deposition cycle 311A, which is an example of an embodiment of operation 200 of FIGS. 2A-2C. In the example of FIG. 3, during the H2 pulse, the flows of Ar and WCl x In other embodiments, the flow of purge gas may continue during the reducing agent pulse. In some embodiments, turning off the purge gas during the reducing agent pulse is effective to increase exposure to the reducing agent. An Ar pulse is shown immediately after the H2. This is an example of an embodiment of operation 202 of FIGS. 2A-2C. During the purge pulse, the H2 and WCl flows are turned off. x The flow is turned off. The dotted boxes at 320 show any number of H2 pulses / Ar purge pulses, performed without any intervening tungsten precursor pulses. This is an example of an embodiment of operations 200 and 202 of Figures 2A and 2C. Although five H2 / Ar pulse sequences are shown, the number of H2 / Ar pulse sequences can range from 1 to k, as discussed above. The dotted boxes at 340 show repeated WCl x2A and 2B. x During the pulse, the Ar valve was closed and during the Ar pulse, WCl x The valves are closed. Hydrogen flow is turned off throughout this process. x Although a H2 / Ar pulse sequence is shown, the number of H2 / Ar pulse sequences can range from 1 to n as discussed above.
[0064] It should be noted that in some embodiments, the H2 pulse and / or WCl x During the pulse, the Ar flow can remain on, while the H2 and WCl x In addition, an increased Ar flow is used in some embodiments, with an AR pulse characterized by no flow. However, during the tungsten precursor pulse, it may be effective to turn off the purge gas to avoid further dilution of the tungsten precursor gas (which may be delivered diluted). The filling of the 3D NAND structure benefits from a higher W concentration in the wafer, which, if diluted, results in insufficient filling. Furthermore, turning off the purge gas and pressurizing the purge manifold during the tungsten precursor pulse allows for a more efficient purge of the chamber than continuing the purge flow with W pulses in between. Similarly, during the reducing agent pulse, it may be effective to turn off the purge gas to avoid dilution of the reducing agent gas. As shown in FIG. 3, the deposition cycle 311A is terminated by an Ar pulse.
[0065] For illustrative purposes, the example in Figure 3 is x3 is shown, the timing sequence of Figure 3 may be used with other precursors, such as tungsten oxychloride precursors and molybdenum chloride and oxychloride precursors. Similarly, other purge gases and / or other reducing agents may be used in place of Ar and H2.
[0066] The process depicted in Figures 2A-2C and the pulse sequence described with reference to Figure 3 are one example of a pulse sequence. Those skilled in the art will appreciate that there are various variations that can be implemented. For example, in some embodiments, different purge gases may be used, as described above. Similarly, the described process is particularly useful for chlorine-containing tungsten or molybdenum precursors, which tend to erode the substrate at high concentrations, although the process may be used to fill structures with other precursors and materials, such as fluorine-containing W and Mo precursors. In particular, the process may be effective for filling high aspect ratio and / or three-dimensional structures.
[0067] 4A-4J are schematic diagrams of an exemplary mechanism for deposition cycles in tungsten deposition from WCl6. FIG. 4A shows an exemplary mechanism for the incorporation of H2 into a substrate 400 on which an underlayer 401 is deposited. Underlayer 401 can be a barrier layer in some embodiments. For example, in some embodiments, underlayer 401 is a titanium nitride (TiN) layer. It should be noted that in some embodiments, substrate 400 does not include a tungsten nucleation layer. In other embodiments, it can include a tungsten nucleation layer. Hydrogen is introduced in the gas phase (411a and 411b) and some H2 (413a and 413b) is on the surface of underlayer 401, where hydrogen can catalytically dissociate into chemically active adsorbed hydrogen atoms on the metal surface (e.g., H2) or physisorb onto a catalytically inactive surface (e.g., TiN). For example, H2 may not necessarily chemisorb to underlayer 401, but in some embodiments may physisorb to the surface of underlayer 401.
[0068] FIG. 4B provides an example showing that H2 that was previously in the gas phase (411a and 411b in FIG. 4A) is purged from the chamber, while H2 that was previously on the surface (443a and 413b) remains on the surface of underlayer 401.
[0069] FIG. 4C provides a schematic example showing a substrate being exposed to WCl6, with some of the WCl6 being in the gas phase (431a and 431b) and some of the WCl6 being at or near the surface of the substrate (423a and 423b).
[0070] In operation 200, some H2 may react with WCl6 remaining on the surface from a previous dose. In FIG. 4D, WCl6 may react with H2 to temporarily form intermediate 443b, which in FIG. 4E, intermediate 443b is fully reacted to leave tungsten 490 on underlayer 401 on the surface of substrate 400, as well as HCl remaining in the gas phase (e.g., 451a and 451b). It should be noted that in this example, tungsten 490 is grown directly on underlayer 401 without depositing a nucleation layer and without treating underlayer 401 prior to depositing tungsten. It will be appreciated that in some embodiments, underlayer 401 may be soaked, such as by exposure to diborane, prior to exposing underlayer 401 to a hydrogen or chlorine-containing tungsten precursor. In some examples, underlayer 401 is soaked after WCl x The Si-BH layer may be pretreated with a mixture of SiH4, B2H6, and H2 to form a solid Si-BH interface layer capable of sacrificially reacting upon exposure to / H2. The stoichiometry of the Si-BH layer may vary widely, from low (10%) to high (95%) boron concentrations.
[0071] In operation 200 of Figures 2A-2C, some H2 may not fully react with WCl6 (or other tungsten chloride) remaining on the surface from a previous dose. As shown in Figure 4D, WCl6 may partially react with H2 to form intermediate 443a, which in Figure 4E leaves intermediate 443a partially reacted on the surface of substrate 400 above underlayer 401. In various embodiments, films deposited with a chlorine-containing tungsten precursor and hydrogen have lower resistivity than films deposited with borane, silane, or germane for deposition thicknesses up to about 50 Å. For example, without being bound to a particular theory, a WCl6 stoichiometry may require at least three H2 molecules to react with one WCl6 molecule. WCl6 may partially react with H2 molecules to form an intermediate rather than forming tungsten. For example, this may occur when there is not enough H2 in the vicinity to react with WCl6 based on stoichiometric principles (e.g., three H2 molecules are used to react with one WCl6 molecule), leaving an intermediate 343a on the substrate surface. Tungsten chloride may be chemically inert to H2 molecules and may require the presence of adsorbed hydrogen atoms generated by catalytic dissociation of H2 molecules. In this case, a significant excess of simple stoichiometric ratios of H2 molecules may be required to completely reduce tungsten chloride to metallic tungsten (100 to 1 H2 molecules / WCl6). x ratio).
[0072] Figure 4F presents a schematic example of the substrate as the chamber is purged. This may correspond to operation 202 in Figures 2A-2C. Note that while compound 443c in Figure 4F may be an intermediate formed without fully reacting, some tungsten 490 may be formed on the substrate. Each cycle thereby forms a sub-atomic layer of tungsten on the substrate.
[0073] As an example, FIG. 4G shows a diagram of a repeated cycle whereby H2 411c in the gas phase is incorporated into the substrate having deposited tungsten 490 and partially reacted intermediate 443d thereon. This may correspond to operation 200 of FIGS. 2A-2C in a repeated cycle after determining in operation 208 that tungsten has not been deposited to an adequate thickness. Note that the incorporated H2 may now react completely with intermediate 443d on the substrate as shown in FIG. 4G, leaving behind deposited tungsten 490b and 490c by reacted compound 443d as shown in FIG. 4H, and by-product HCl 451c and 451d formed in the gas phase. Some H2 411c may remain in the gas phase, while some H2 413c may remain on top of tungsten layer 490a.
[0074] In Figure 4I, the chamber is purged (which corresponds to operation 202 in Figures 2A-2C), leaving behind deposited tungsten 490a, 490b, 490c and some H2 413c. In Figure 4J, a dose of WCl6 is again introduced, this time with molecules 431c and 423c that may adsorb and / or react with H2 and the substrate. Figure 4J may correspond to operation 204 in Figures 2A-2C. After the dose of WCl6, the chamber may again be purged, and the cycle may again be repeated until the desired thickness of tungsten has been deposited.
[0075] Tungsten films deposited using some of the disclosed embodiments are fluorine-free. Molybdenum films may also be fluorine-free. The overall film tensile stress may be less than about 0.2 GPa. Fluorine-free tungsten precursors include tungsten carbonyl (W(CO)6) and tungsten chlorides (WCl), such as tungsten pentachloride (WCl5) and tungsten hexachloride (WCl6). x Fluorine-free molybdenum precursors include the molybdenum oxychloride and chloride precursors described herein, as well as molybdenum hexacarbonyl (Mo(CO)6).
[0076] The disclosed embodiments may be applied in various ways in tungsten and molybdenum deposition processes. It will be understood that various combinations of the applications described herein may be used to deposit tungsten or molybdenum, and that these methods are not limited to the examples presented herein.
[0077] For example, in some embodiments, a feature may be filled by depositing a tungsten nucleation layer by ALD cycles of alternating pulses of a reducing agent (e.g., borane, silane, or germane with hydrogen) and a tungsten precursor such as WCl6, followed by depositing bulk tungsten by alternating pulses of hydrogen and a chlorine-containing tungsten precursor as described above with respect to Figures 2 and 3. Similarly, in some embodiments, a feature may be filled by depositing a molybdenum layer by ALD cycles of alternating pulses of a reducing agent and a molybdenum precursor. Deposition of bulk tungsten by a process using alternating pulses of a tungsten chloride precursor and the reducing agent hydrogen is sometimes referred to as an ALD process due to the alternating pulses of reactants. The methods described herein typically involve purging between reactant cycles to ensure the absence of growth components from the CVD mechanism.
[0078] However, in some embodiments, these methods may be used in conjunction with "sequential CVD" processes such as those described in U.S. Patent Application Publication No. 2017 / 0117155 and U.S. Patent No. 9,613,818, both of which are incorporated herein by reference.
[0079] By employing multiple cycles of alternating precursor and purge gas pulses in one deposition cycle, precursors can be driven deep into the 3D NAND word lines while minimizing erosion at the top and side edges of the structure. An atomic layer or sub-atomic layer of precursor can be adsorbed throughout the structure. In some embodiments, alternating precursor and purge pulses can be used to pressurize the precursor bath before delivering the precursor dose, resulting in better delivery into the structure.
[0080] Table 1 shows the H2 / Ar / WCl ratio with n=10. x / Ar sequence and H2 / Ar / n (WCl x 1 shows the conditions and resulting erosion of the TiN underlayer for a 300 Å PVD TiN blanket film exposed to a 300 Å / Ar) sequence. [Table 1] The conditions and results in Table 1 show that for the same total exposure to the tungsten precursor, multiple pulses of the tungsten precursor eroded significantly less TiN, especially for the sequence with a single tungsten precursor pulse, all of the TiN was eroded.
[0081] Figure 5 shows the H2 / Ar / WCl x / Ar sequence and H2 / Ar / n(WCl x 5 shows a schematic diagram of experimental results comparing tungsten filling of a 3D NAND structure using a H2 / Ar / WCl (H2 / Ar) sequence. Prior to these sequences, a nucleation layer was deposited in each structure. The diagram in FIG. 5 shows a cross-section of the center section of the 3D NAND structure. The 3D NAND structure 503 is shown in FIG. 5 with a H2 / Ar / WCl (H2 / Ar) sequence. x The image was taken after the H2 / Ar / n(WCl) sequence, showing insufficient tungsten deposition in the center word line, as shown at 511. xA 3D NAND structure 503 exposed to 1000 Å (Ar / TiO2) is observed to have uniform tungsten deposition throughout. The total exposure to the tungsten precursor in Torr-seconds was the same for both 501 and 503.
[0082] Each of the n tungsten pulses may be short and delivered at a constant pressure. The total amount of tungsten precursor is sufficient to provide enough material for the tungsten precursor to reach the bottom of the structure and to laterally coat the entire structure. An intervening inert gas purge may remove the tungsten precursor at the top and edges of the structure, reducing or eliminating erosion of the substrate. This beneficial effect may not be observed if the inert gas purge is insufficient. Additionally, a pinch effect around the opening may be observed if the purge between multiple tungsten pulses is insufficient.
[0083] Thus, in some embodiments, the duration of the inert gas purge is at least as long as the duration of the tungsten or molybdenum precursor. In some embodiments, the inert gas purge is at least 1.5 times, or at least 2 times, or at least 3 times the duration of the tungsten or molybdenum precursor pulse. The precursor to purge gas exposure may also be characterized by partial pressure times duration. The precursor may be diluted to 1%-5% of the flow rate, while the purge is performed with 100% purge gas.
[0084] While the above description focuses on the deposition of tungsten and tungsten-containing layers from tungsten chloride, the process described above with reference to Figures 2 and 3 may be implemented for feature filling with other precursors and / or other materials. As noted above, this method is effective for chemicals that may erode the structure. Additionally, methods involving multiple precursor pulse / purge cycles may be effective to obtain growth without CVD-type characteristics by purging unadsorbed molecules.
[0085] The disclosed embodiments may be performed at any suitable pressure, such as pressures above about 10 Torr or pressures below about 10 Torr. In the case of a multi-station chamber, each pedestal may be set to a different temperature. In some embodiments, each pedestal is set to the same temperature. The substrate may be cycled from station to station during some or all of any of the above-described operations according to the disclosed embodiments. Also, the chamber pressure may be modulated during one or more operations in some disclosed embodiments. In some embodiments, the chamber pressure during nucleation deposition is different from the chamber pressure during bulk deposition. In some embodiments, the chamber pressure during nucleation deposition is the same as the chamber pressure during bulk deposition.
[0086] [Device] Any suitable chamber may be used to carry out the disclosed embodiments. Examples of deposition equipment include various systems, such as ALTUS® and ALTUS® Max available from Lam Research Corp., Fremont, Calif., or any of various other commercially available processing systems. In some embodiments, sequential chemical vapor deposition (CVD) may be performed in a first station that is one of two, five, or even more deposition stations arranged in a single deposition chamber. In that case, in the first station, for example, hydrogen (H2) and tungsten hexachloride (WCl6) may be introduced in alternating pulses to the surface of the semiconductor substrate using individual gas delivery systems that create a local atmosphere at the substrate surface. Other stations may be used for fluorine-free tungsten deposition, or CVD. Two or more stations may be used to deposit tungsten in a parallel process. Alternatively, the wafer may be indexed such that sequential CVD operations are performed sequentially across two or more stations.
[0087] FIG. 6 is a schematic diagram of a processing system suitable for performing a deposition process according to some embodiments. The system 600 includes a transfer module 603. The transfer module 603 provides a clean pressurized environment to minimize the risk of contamination of the substrate being processed as it is transferred between the various reactor modules. A multi-station reactor 609 capable of performing ALD and CVD according to various embodiments is mounted on the transfer module 603. The multi-station reactor 609 may have multiple stations 611, 613, 615, 617 that may sequentially perform operations according to the disclosed embodiments. For example, the multi-station reactor 609 may be configured to perform a tungsten nucleation layer deposition by PNL using a chlorine-containing tungsten precursor or a fluorine-containing precursor at station 611, and a tungsten deposition operation by ALD according to various embodiments at station 613. In some embodiments, station 615 may also perform a tungsten deposition operation by ALD, and station 617 may perform a non-sequential CVD operation. In some embodiments, the number (n) of tungsten precursor pulses can vary between stations 613-615.
[0088] These stations may have a heated pedestal or substrate support and one or more gas inlets or showerheads or distribution plates. An example of a deposition station 700 is shown in Figure 7, which has a substrate support 702 and a showerhead 703. The pedestal portion 701 may be provided with a heater.
[0089] Referring again to FIG. 6, one or more single or multi-station modules 607 capable of performing plasma or chemical (non-plasma) pre-cleaning, other deposition operations, or etching operations may also be mounted on the transfer module 503. This module may also be used for various processes, for example to prepare the substrate for a deposition process. The system 600 further comprises one or more wafer source modules 601 for storing pre- and post-processed wafers. An atmospheric robot (not shown) in an atmospheric transfer chamber 619 may first remove the wafer from the source module 601 to a load lock 621. A wafer transfer device (typically a robot arm unit) in the transfer module 603 transfers the wafer from the load lock 621 to and between modules mounted on the transfer module 603.
[0090] In various embodiments, a system controller 629 is employed to control process conditions during deposition. The controller 629 typically includes 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, stepper motor controller boards, etc.
[0091] The controller 629 may control all of the operations of the deposition apparatus. The system controller 629 executes system control software that includes sets of instructions for controlling the timing, mixture of gases, chamber pressure, chamber temperature, wafer temperature, radio frequency (RF) power levels, wafer chuck or pedestal position, and other parameters of a particular process. Other computer programs stored in memory devices associated with the controller 629 may be employed in some embodiments.
[0092] Typically, there is a user interface associated with the controller 629. The user interface may include a display screen and a graphical software representation of equipment and / or process conditions, as well as user input devices such as a pointing device, keyboard, touch screen, microphone, etc.
[0093] The system control logic may be configured in any suitable manner. In general, logic may be designed or configured in hardware and / or software. Instructions for controlling the drive circuits may be hard-coded or provided as software. Those instructions may be provided by "programming". Such programming is understood to include any form of logic, such as hard-coded logic in digital signal processors, application specific integrated circuits, and other devices with specific algorithms implemented as hardware. Programming is further understood to include software or firmware instructions that may be executed on a general purpose processor. The system control software may be coded in any suitable computer readable programming language.
[0094] The computer program code for controlling the germanium-containing reducing agent pulses, hydrogen flow, and tungsten-containing precursor pulses, as well as other processes in the process sequence, can be written in any conventional computer-readable programming language, such as, for example, assembly language, C, C++, Pascal, Fortran, etc. Execution of the compiled object code or script by a processor performs the tasks indicated in the program. Additionally, as discussed above, the program code may be hard-coded.
[0095] The controller parameters relate to process conditions, such as, for example, process gas composition and flow rates, temperature, pressure, cooling gas pressure, substrate temperature, chamber wall temperature, etc. These parameters are provided to the user in the form of a recipe and may be entered using a user interface.
[0096] Signals for monitoring the process may be provided by analog and / or digital input connections of the system controller 629. Signals for controlling the process are output to analog and digital output connections of the deposition apparatus 600.
[0097] The system software may be designed or configured in many different ways. For example, various chamber component subroutines or control objects may be created to control the operation of chamber components necessary to carry out a deposition process according to the disclosed embodiments. Examples of programs or program portions for this purpose include substrate positioning code, process gas control code, pressure control code, and heater control code.
[0098] In some implementations, the controller 629 is part of a system that may be part of the examples described above. Such systems may include semiconductor processing equipment such as a processing tool or tools, a chamber or chambers, a processing platform or platforms, and / or specific processing components (wafer pedestal, gas flow systems, etc.). These systems may be integrated with electronics to control their operation before, during, and after processing of a semiconductor wafer or substrate. These electronics may be referred to as "controllers" that may control various components or subparts of the system or systems. Depending on the processing requirements and / or type of system, the controller 629 may be programmed to control any of the processes disclosed herein, including supply of process gases, temperature settings (e.g., heating and / or cooling), pressure settings, vacuum settings, power settings, settings of radio frequency (RF) generators in some systems, settings of RF matching circuits, frequency settings, flow settings, settings of fluid supplies, position and motion settings, wafer transfer to and from the tool and other transfer tools and / or load locks connected or interfaced to the particular system.
[0099] A controller may be broadly defined as an electronic device having various integrated circuits, logic, memory, and / or software that receives instructions, issues instructions, controls operations, performs cleaning operations, performs endpoint measurements, etc. Integrated circuits may include chips in the form of firmware that store program instructions, digital signal processors (DSPs), chips defined as application specific integrated circuits (ASICs), and / or one or more microprocessors or microcontrollers that execute program instructions (e.g., software). Program instructions may be instructions communicated to the controller in the form of various individual settings (or program files) that define operational parameters for performing a particular process on or for a semiconductor wafer or for a system. The operational parameters may, in some embodiments, be part of a recipe defined by a process engineer to perform one or more processing steps in the manufacture of one or more layers, materials, metals, oxides, silicon, silicon dioxide, surfaces, circuits, and / or dies of a wafer.
[0100] The controller 629 may, in some implementations, be part of or coupled to a computer integrated or otherwise networked to the system, or a combination thereof. For example, the controller 629 may be in the "cloud" or may be all or part of a fab host computer system that may allow remote access for wafer processing. The computer may provide remote access to the system to monitor the current progress of a manufacturing operation, to study the history of past manufacturing operations, to study trends or performance metrics from multiple manufacturing operations, to change parameters of a current process, to set process steps according to a current process, or to start a new process. In some examples, a remote computer (e.g., a server) may provide a process recipe to the system over a network that may include a local network or the Internet. The remote computer may have a user interface that allows for input or programming of parameters and / or settings that are then communicated to the system from the remote computer. In some examples, the controller receives instructions in the form of data that specify parameters of each of the process steps to be performed in one or more operations. It should be understood that these parameters may be specific to the type of process being performed and the type of tool the controller is configured to interface with or control. In that case, the controller may be distributed, such as by having one or more separate controllers that are networked together as described above and cooperate toward a common purpose, such as the processes and controls described herein. One example of a distributed controller for such purposes is one or more integrated circuits mounted on the chamber that communicate with one or more remotely located integrated circuits (e.g., at the platform level or as part of a remote computer) to jointly control the process in the chamber.
[0101] Exemplary systems may include, but are not limited to, a plasma etch chamber or module, a deposition chamber or module, a spin rinse chamber or module, a metal plating chamber or module, a clean chamber or module, a bevel edge etch chamber or module, a physical vapor deposition (PVD) chamber or module, a CVD chamber or module, an ALD chamber or module, an atomic layer etch (ALE) chamber or module, an ion implantation chamber or module, a track chamber or module, and any other semiconductor processing system that may be associated with or used in the creation and / or manufacturing of semiconductor wafers.
[0102] As described above, the controller may communicate with one or more of other tool circuits or modules, other tool components, cluster tools, other tool interfaces, adjacent tools, nearby tools, tools located throughout the factory, a main computer, other controllers, or tools used in material transport to move containers of wafers to and from tool locations and / or load ports in a semiconductor manufacturing factory, depending on the processing step or steps being performed in the tool.
[0103] The controller 629 may include various programs. A substrate positioning program may include program code for controlling chamber components used to load the substrate onto the pedestal or chuck, as well as to control the spacing between the substrate and other members of the chamber, such as gas inlets, and / or targets. A process gas control program may include code for controlling gas composition, flow rates, pulse times, and optionally flowing gas into the chamber to stabilize the pressure in the chamber prior to deposition. A pressure control program may include code for controlling the pressure in the chamber, for example, by adjusting a throttle valve in the exhaust system of the chamber. A heater control program may include code for controlling the current to a heating unit used to heat the substrate. Alternatively, the heater control program may control the supply of a heat transfer gas, such as helium, to the wafer chuck.
[0104] Examples of chamber sensors that may be monitored during deposition include mass flow controllers, pressure sensors such as manometers, thermocouples located on the pedestal or chuck, etc. Appropriately programmed feedback and control algorithms may be used in conjunction with the data from these sensors to maintain desired process conditions.
[0105] The apparatus may include a gas manifold system that provides line fill to the various gas distribution lines, as shown diagrammatically in FIG. 8. Manifold 804 has an input 802 from a source of tungsten-containing or molybdenum-containing precursor gas (not shown), manifold 811 has an input 809 from a source of hydrogen or other reducing gas (not shown), and manifold 821 has an input 819 from a source of inert purge gas (not shown). Manifolds 804, 811, 821 supply precursor, reducing, and purge gases to the deposition chamber via valved distribution lines 805, 813, 825, respectively. To provide line fill, i.e., to pressurize the distribution lines, various valves are opened or closed. For example, to pressurize distribution line 805, valve 806 to vacuum is closed and valve 808 is closed. After an appropriate time increment, valve 808 is opened to supply precursor gas to the chamber. After a suitable time for gas delivery, valve 808 is closed. The chamber can then be purged to vacuum by opening valve 806 to vacuum.
[0106] A similar process is used to deliver reducing and purge gases. For example, to introduce reducing gas, delivery line 813 is filled by closing valve 815 and closing valve 817 to vacuum. Valve 815 is opened to allow the reducing gas to be delivered to the chamber. Similarly, to introduce purge gas, delivery line 825 is filled by closing valve 827 and closing valve 823 to vacuum. Valve 827 is opened to allow the chamber to be delivered with argon or other inert purge gas. The amount and timing of the initial gas delivery will vary depending on the amount of time allowed for the lines to fill.
[0107] 6 further shows a vacuum pump that can open valves 806, 817, 823, respectively, to purge the system. The supply of gases through the various distribution lines is controlled by a controller, such as a mass flow controller controlled by a microprocessor, digital signal processor, or the like, programmed with flow rates, flow durations, and process sequencing.
[0108] It should be noted that the above process may require precise timing of the valves and mass flow controllers (MFCs) that deliver pulses of reagents to the semiconductor substrate during deposition. One way to achieve this is to issue valve and MFC commands to an embedded digital input / output controller (IOC) in a discrete packet of information that contains instructions for all time-critical commands for all or part of the PNL deposition sequence. Lam Research's ALTUS system provides at least one IOC sequence. The IOCs can be physically located in various locations within the equipment, for example, within the process module or on a standalone power rack some distance away from the process module. Typically, there are multiple IOCs within each module (e.g., three per module). In terms of the actual instructions contained in a sequence, all commands for controlling the valves and setting the MFC flow rates (for all of the carrier and reactant gases) may be included in a single IOC sequence. This ensures that the timing of all devices is tightly controlled in absolute terms and relative to each other. Typically, at any given time, multiple IOC sequences are running. This allows, for example, a PNL sequence to be run at stations 1-2, controlling all the timing of all the hardware components required to deposit a PNL-W nucleation layer at those stations. A second sequence may be running simultaneously at other deposition stations in the same module to deposit bulk tungsten using the above timing sequence. The relative timing of the devices controlling the delivery of reagents to stations 3-4 is important within that group, but the relative timing of the PNL process at stations 1-2 can be offset from the relative timing of stations 3-4. The IOC translates the information in the packetized sequence to issue digital or analog command signals directly to the MFCs or to the pneumatic solenoid banks that control the valves.
[0109] A pulse of tungsten-containing gas may be generated as follows: First, the system is primed with WCl for a period of time while the MFC or other flow control device stabilizes. x to the vacuum pump. This may be done for a period of between about 0.5-5 seconds in one example. The system then pressurizes the tungsten gas supply manifold by closing both the diversion outlet 606 and the outlet 608 to the deposition chamber. This may be done for a period of between about 0.1-5 seconds, for example, to generate an initial burst of reagent when the outlet to the deposition chamber is opened. This is accomplished by opening the outlet valve 808 for a period of between about 0.1-10 seconds in one example. The tungsten-containing gas is then purged from the deposition chamber using an appropriate purge gas. Pulsed flows of other reagents may be done in a similar manner. Pulses of molybdenum-containing gas may be generated in a similar manner.
[0110] Above, implementation of the disclosed embodiments in single or multi-chamber semiconductor processing tools is described. The apparatus and processes described herein may be used in combination with lithographic patterning tools or processes, for example, for the fabrication or manufacture of semiconductor devices, displays, LEDs, solar panels, etc. Typically, such tools / processes are used or performed together in a common manufacturing facility, although this is not necessarily the case. Lithographic patterning of films typically includes some or all of the following steps, each of which is performed using several possible tools: (1) Spreading photoresist onto a workpiece or substrate using a spin or spray tool; (2) curing the photoresist using a hotplate or furnace or UV curing tool; (3) exposing the photoresist to visible or ultraviolet light or x-rays using a tool such as a wafer stepper; (4) developing the resist to selectively remove the resist, thereby forming a pattern, using a tool such as a wet bench; (5) transferring the resist pattern to the underlying film or workpiece using a dry or plasma-assisted etching tool; (6) stripping the resist using a tool such as an RF or microwave plasma resist stripper.
[0111] [Conclusion] Although the above embodiments have been described in some detail for purposes of clarity of understanding, it will be apparent that certain changes and modifications may be made within the scope of the appended claims. It should be noted that there are many alternative ways of implementing the process, system, and apparatus of the embodiments of the present invention. Thus, the embodiments of the present invention should be considered as illustrative, not restrictive, and the embodiments are not limited to the details presented herein.
Claims
1. 1. A method comprising: preparing a structure to be filled with a metal-containing material, the structure being a partially fabricated three-dimensional (3D) NAND structure having a sidewall and a plurality of openings in the sidewall, the openings connecting to a plurality of features having a plurality of interior regions fluid-accessible through the openings; exposing the structure to a plurality of deposition cycles to deposit the metal-containing material throughout the structure without etching a top portion of the structure, the structure including an innermost region of the structure, each deposition cycle comprising: exposure to a metal precursor pulse followed by an inert gas purge pulse; Multiple hydrogen (H 2 ) pulse and an inert gas purge pulse.
2. 2. The method of claim 1 , The method of claim 1, wherein the metal precursor is a chlorine-containing metal precursor.
3. 3. The method of claim 2, The method wherein the metal precursor is delivered from a pressurized vessel.
4. 2. The method of claim 1 , A method wherein each deposition cycle includes exposure to multiple metal precursor pulses followed by an inert gas purge without intervening hydrogen pulses.
5. 2. The method of claim 1 , The method wherein the hydrogen reacts with the adsorbed metal precursor to form hydrogen chloride (HCl).
6. 2. The method of claim 1 , The metal precursor is MoO 2 C 2 and MoOCl 4 A method, which is either
7. 2. The method of claim 1 , Multiple alternating sex 2 The exposure to a pulse and an inert gas purge pulse 2 turning off the flow of the inert gas purge during a pulse.
8. 2. The method of claim 1 , The multiple hydrogen and inert gas pulses, without any intervening metal precursor pulses, allow reducing agents to react with adsorbed chlorides and allow reaction products to leave the surface to make room for the reducing agent in the next pulse.
