Molybdenum deposition

Abstract

Methods of depositing molybdenum in 3D NAND structures include: providing a 3D NAND structure having wordline features; forming molybdenum directly on dielectric surfaces of the wordline features; filling the features with molybdenum by an ALD process using a molybdenum reactant; and adjusting deposition conditions to maintain complete saturation as the deposition proceeds.

Description

Attorney Docket No. LAM1P127WO-12314-1WOMOLYBDENUM DEPOSITIONRELATED APPLICATION(S)

[0000] A PCT Request Form is filed concurrently with this specification as part of the present application. Each application that the present application claims benefit of or priority to as identified in the concurrently filed PCT Request Form is incorporated by reference herein in its entirety and for all purposes.BACKGROUND

[0001] Deposition of conductive materials is an integral part of many semiconductor fabrication processes. These materials may be used for horizontal interconnects, vias between adjacent metal layers, contacts between metal layers and devices, and as lines in memory devices. In an example of deposition, a tungsten (W) layer may be deposited on a titanium nitride (TiN) barrier layer to form a TiN / W bilayer by chemical vapor deposition (CVD) process using tungsten hexafluoride (WFe). However, as devices shrink and more complex patterning schemes are utilized in the industry, the deposition of thin tungsten becomes a challenge. The continued decrease in feature size and film thickness brings various challenges to TiN / W film stacks. These include high resistivity forthinnerfilms and deterioration of TiN barrier properties. Deposition in complex high aspect ratio structures such as 3D NAND structures and DRAM buried wordline (bWL) is particularly challenging.

[0002] The background description provided herein is for the purpose of generally presenting the context of disclosure. Work of the presently named inventors, to the extent it is described in this background section, as well as aspects of the description that may not otherwise qualify as prior art at the time of filing, are neither expressly nor impliedly admitted as prior art against the present disclosure.SUMMARY

[0003] In some aspects, the techniques described herein relate to a method including: providing a 3D NAND structure having wordlines feature to be filled with molybdenum (Mo) to a semiconductor processing tool, wherein each wordline feature is having a first opening and a second opening, the first opening and the second opening being at opposite ends of the featureAttorney Docket No. LAM1P127WO-12314-1WO and wherein the first opening opens to a first vertical slit of the 3D NAND structure and the second opening opens to a second vertical slit and wherein the feature is flu id ica lly accessible via the first and second vertical slit; forming molybdenum directly on dielectric surfaces of the wordline features; filling the features with molybdenum by an ALD process using a molybdenum reactant; and adjusting deposition conditions throughout the ALD process to maintain complete saturation ofthe molybdenum reactant.

[0004] In some aspects, the techniques described herein relate to a method, wherein adjusting the deposition conditions during the ALD process includes increasing molybdenum reactant exposure, wherein molybdenum reactant exposure is a product of the partial pressure of the molybdenum reactant and a dose time ofthe molybdenum reactant.

[0005] In some aspects, the techniques described herein relate to a method, wherein the molybdenum reactant exposure is increased every N cycles, wherein N is a number ranging from 3 to 50.

[0006] In some aspects, the techniques described herein relate to a method, wherein adjusting the deposition conditions during the ALD process further includes increasing post-molybdenum reactant purge gas exposure, wherein reactant purge gas exposure reactant exposure is scaled relative to the product ofthe total purge gas flow rate and the purge time.

[0007] In some aspects, the techniques described herein relate to a method, wherein adjusting the deposition conditions during the molybdenum reactant exposure, wherein molybdenum reactant exposure is a product of the partial pressure of the molybdenum reactant and a dose time ofthe molybdenum reactant.

[0008] In some aspects, the techniques described herein relate to a method, further including estimating transient molybdenum reactant mass fraction inside a wordline feature (OJMO) during an ALD process.

[0009] In some aspects, the techniques described herein relate to a method, further including determining a molybdenum exposure during an ALD process to maintain saturation of the molybdenum reactant based on precursor mass fraction OJMO.

[0010] In some aspects, the techniques described herein relate to a method, wherein the ALD process includes reacting a co-reactant with the molybdenum reactant.Attorney Docket No. LAM1P127WO-12314-1WO

[0011] In some aspects, the techniques described herein relate to a method, adjusting the deposition conditions during the ALD process includes increasing co-reactant exposure, wherein co-reactant exposure is a product of the partial pressure of the co-reactant and a dose time of the co-reactant.

[0012] These and other aspects are described more fully below with reference to the Figures.BRIEF DESCRIPTION OF DRAWINGS

[0013] Figures 1A-1D show examples of material stacks featuring a substrate and a molybdenum layer deposited thereon.

[0014] Figures 1E-1G show examples of stages of deposition of molybdenum in a feature according to various embodiments.

[0015] Figures 2A-2L depicts examples of features in which molybdenum may be deposited.

[0016] Figure 3 shows top-down and cross-sectional views of a flash memory array geometry.

[0017] Figures 4A and 4B are graphs illustrating the fraction of NAND wordline volume occupied by metal (excluding the liner volume) for various liner thicknesses and the NAND wordline resistance for various liner thicknesses, respectively.

[0018] Figure 4C is a process flow diagram illustrating operations in a method of depositing a conductive material.

[0019] Figure 5 is an SEM image of ALD-Mo deposited on an alumina surface.

[0020] Figure 6 shows molybdenum growth rates on AI2O3 of various ALD processes.

[0021] Figure 7 shows resistivity as a function of film thickness for ALD-Mo films and TiN / W stacks.

[0022] Figure 8 illustrates transient diffusion into 3D NAND Flash wordlines in a cell array.

[0023] Figure 9A illustrates wordline saturation time as a function of molybdenum thickness in the wordline features for various molybdenum reactant concentrations.Attorney Docket No. LAM1P127WO-12314-1WO

[0024] Figures 9B and 9C are flow diagrams showing operations in examples of methods of filling wordline features.

[0025] Figure 9D illustrates wordline saturation time as a function of number of wordlines for various molybdenum reactant concentrations.

[0026] Figures 10A and 10B depicts schematic illustrations of embodiments of ALD process stations.

[0027] Figures 11A and 11B depicts schematic illustrations of embodiments of semiconductor processing systems including single-station and / or multi-station chambers.DETAILED DESCRIPTION

[0028] In the following descriptions, numerous specific details are set forth to provide a thorough understanding of the presented embodiments. The disclosed embodiments may be practiced without some or all these specific details. In other instances, well-known process operations have not been described in detail to not unnecessarily obscure the disclosed embodiments. While the disclosed embodiments will be described in conjunction with the specific embodiments, it will be understood that it is not intended to limit the disclosed embodiments.

[0029] Provided herein are methods for deposition of molybdenum (Mo). In some embodiments, molybdenum is deposited directly on a dielectric surface. Deposition of Mo on dielectrics may be implemented without delay by an initial atomic layer deposition (ALD) process that includes a long reactant (e.g.,H2) at high temperature and high pressure.

[0030] Figures 1A-2L show examples of material stacks and structures into which the methods of the disclosure may be used to deposit a Mo liner layer and / or deposit Mo for feature fill.

[0031] Figures 1A and IB are schematic examples of material stacks that include Mo layers according to various embodiments. Figures 1A and IB illustrates the order of materials in examples of particular stacks and may be used with any appropriate architecture application, as described further below with respect to Figures 2A-2L. Figure 1A shows a first material stack 111 featuring a substrate 102 and a molybdenum layer 108 deposited thereon. The substrate 102 may be a silicon or other semiconductor wafer, e.g., a 200-nm wafer, a 300-nm wafer, or a 450-nm wafer, including wafers having one or more layers of material, such as dielectric, conducting, orAttorney Docket No. LAM1P127WO-12314-1WO semiconductive materials deposited thereon. In some embodiments, the substrate 102 may be or include silicon (Si) or silicon germanium (SiGe). The methods may also be applied to form metallization stack structures on other substrates, such as glass, plastic, and the like.

[0032] The stack 111 has a dielectric layer 104 on the substrate 102. The dielectric layer 104 may be deposited directly on a semiconductor surface (e.g., a Si or SiGe surface) of the substrate 102, or there may be any number of intervening layers. For example, the substrate 102 may include any number of layers deposited in various arrangements on a semiconductor substrate.

[0033] Examples of dielectric layers include doped and undoped silicon oxide, silicon nitride, and aluminum oxide layers, with specific examples including doped or undoped layers of silicon nitride (SiN), silicon dioxide (SiCh), and aluminum oxide (AI2O3). The stack 111 has a layer 106 disposed between the molybdenum layer 108 and the dielectric layer 104. The layer 106 may be a diffusion barrier and / or adhesion layer, for example. A diffusion barrier is a layer that prevents the diffusion of species between layers. An adhesion layer is a layer that promotes adhesion of one layer to an underlying layer. Examples of diffusion barrier and adhesion layers include titanium nitride (TiN), titanium / titanium nitride (Ti / TiN), tungsten (W), tungsten nitride (WN), and tungsten carbon nitride (WCN). In the example of Figure 1A, the molybdenum layer 108 may be the main conductor of the structure. In some embodiments, the molybdenum layer 108 may or may not include a molybdenum nucleation layer. In some embodiments, the molybdenum layer is an amorphous molybdenum-containing layer. In the depicted example of Figure 1A, the molybdenum layer 108 is deposited directly on the layer 106. In other embodiments (not depicted), the molybdenum layer 108 may be deposited on a separate layer, such as a growth initiation layer that includes another material, such as a tungsten (W) or W-containing growth initiation layer. The growth initiation layer may be used to facilitate nucleation and growth of molybdenum layer 108. Still further, in some embodiments, the molybdenum layer 108 may be deposited on a conductive layer that is the main conductor of an underlying structure.

[0034] Figure IB shows another example of a stack 121. In this example, the stack 121 includes the substrate 102, dielectric layer 104, with molybdenum layer 108 deposited directly on the dielectric layer 104, without an intervening diffusion barrier or adhesion layer. In some embodiments, the molybdenum layer 108 is a main conductor as described with respect to Figure 1A. By using molybdenum as the main conductor, low-resistivity thin films can be obtained.Attorney Docket No. LAM1P127WO-12314-1WO Examples of low-resistivity thin films include films with resistivity less than 40 ufim-cm at 60 angstroms thickness and less than 15 ufim-cm at 200 angstroms thickness.

[0035] In some embodiments, a stack may include a substrate, a conductive layer, and a molybdenum layer deposited onto the conductive layer. As used herein, a conductive layer or conductive material is a layer or material having a conductivity of at least 104Q '-cm1at room temperature. Examples include molybdenum on a metal layer (e.g., a W layer, or another Mo layer). In these embodiments, there is no dielectric layer between the molybdenum layer and the conductive layer. Similarly, the stack may include molybdenum deposited directly on a metal compound layer. Examples include molybdenum on a metal nitride layer (e.g., TiN, WN, or MoN). In still some other embodiments of stack (not shown), the stack may include a substrate and a molybdenum layer deposited directly on the substrate, including directly on a semiconductor surface, on a dielectric surface, or a conductive surface.

[0036] Figure 1C shows another example of a stack 131. In this example, the stack 131 includes the substrate 102, a conductive layer 103, a molybdenum layer 108, and an overlying conductive layer 103. In some embodiments, molybdenum may be a diffusion barrier layer with another layer deposited thereon. The conductive layers 103 may be the same or different materials. Figure ID shows another layer example of a stack 141. In this example, the stack 141 includes the substrate 102, a dielectric layer 104, molybdenum layer 108, and a conductive layer 103.

[0037] Figures 1A-1D illustrates examples of the order of materials in a particular stack and may be used with any appropriate architecture and applications. One example of a feature into which molybdenum is deposited according to certain embodiments is shown in Figure IE. In the example of Figure IE, a feature is formed in a dielectric layer 104. The dielectric layer overlies a conductive material 103, such that the feature is defined by dielectric sidewall surfaces 104a and a conductive bottom surface 103a. While the conductive bottom surface 103a and the sidewall surfaces 104a are depicted as being uniform and even horizontal or vertical surfaces in the example of Figure IE, they may be angled, curved, rough, or uneven according to various embodiments.

[0038] In Figure IF, a conformal Mo layer 108 is shown deposited on both dielectric sidewall surfaces 104a and conductive surface 103a. Thus, the architecture in Figure IF includes Mo on both dielectric and conductive surfaces. In some embodiments, a conductive Mo layer 103 fills the remainder of the feature, as shown in Figure 1G. In the examples of Figures 1E-1G, the MoAttorney Docket No. LAM1P127WO-12314-1WO layer 108 can be a diffusion barrier layer. The conductive material in each of the conductive layers 103 may be the same or different. In some embodiments, the diffusion barrier is a Mo layer with the conductive layers being tungsten (W) layers. Other examples of materials in conductive layers include metals such as cobalt (Co), ruthenium (Ru), copper (Cu), nickel (Ni), iridium (Ir), rhodium (Rh), tantalum (Ta), and titanium (Ti). Titanium nitride (TiN) may also be used. Other conductive nitridesthat may be used in zirconium nitride (ZrN), hafnium nitride (HfN), vanadium nitride (VN), niobium nitride (NbN), tantalum nitride (TaN), as well as MoN and WN. As discussed further below, in some embodiments, the Mo layer at the bottom surface of the feature is etched to reduce resistance between the conductive layers 103.

[0039] Further examples of appropriate architecture and applications for the material stacks shown in Figures 1A-1G are described below with respect to Figures 2A-2L. The methods described herein are performed on a substrate that may be housed in a chamber. The substrate may be a silicon or other semiconductor wafer, including wafers having one or more layers of materials, such as dielectric, conducting, or semiconducting material deposited thereon. The methods are not limited to semiconductor substrates and may be performed to deposit a molybdenum layer in any feature. Moreover, in some embodiments, the molybdenum compound layer (e.g., molybdenum nitride) may be deposited rather than a molybdenum metal layer (also referred to as an elemental molybdenum layer).

[0040] Examples of features include vias, trenches, and contact holes. Features may be characterized by one or more narrow and / or re-entrant openings, constrictions within the feature, and high aspect ratios. A feature may be formed in one or more of the above-described stacks or layers within a stack. For example, the features may be formed at least partially in a dielectric layer. In some embodiments, a feature 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 higher. One example of a feature is a hole or via in a semiconductor substrate or a layer on the substrate.

[0041] Figure 2A depicts a schematic example of a DRAM architecture, including a buried wordline (bWL) 208 in a silicon substrate 202. The bWL 208 is formed in a trench etched in the silicon substrate 202. Lining the trench is a conformal barrier layer 206 and an insulating layer 204. The conformal barrier layer 206 is disposed between the insulating layer 204 and the silicon substrate 202. In this example, the insulating layer 204 may be a gate oxide layer formed from a high-k dielectric material such as silicon oxide or silicon nitride material. In some embodimentsAttorney Docket No. LAM1P127WO-12314-1WO disclosed herein, the conformal barrier layer 206 is TIN or tungsten-containing layer, such as WN or WCN layer. In some embodiments, a conformal tungsten-containing growth initiation layer (not shown) may be present between the conformal barrier layer 206 and the molybdenum bWL 208. Alternatively, the molybdenum bWL 208 may be deposited directly on a TiN or other diffusion barrier. In some embodiments, one or both of layers 204 and 206 is not present. Still further, in some embodiments, the conformal barrier layer 206 is molybdenum or a molybdenum containing layer. The bWL is a conductive material and may be any of the metals described above in some embodiments. The bWL structure shown in Figure 2A is one example of an architecture that includes a conductive fill layer. During fabrication of the bWL, molybdenum is deposited into a feature that may be defined by an etched recess in the silicon substrate 202.

[0042] Figures 2B-2H are additional schematic examples of various structures into which a conductive material may be deposited in accordance with disclosed embodiments. In any of these examples, a Mo liner layer (or a Mo-containing compound liner layer) may be deposited prior to deposition of the main conductive material. In some embodiments, molybdenum is the main conductive material.

[0043] Figure 2B shows an example of a cross-sectional depiction of a vertical feature 201 to be filled with a conductive material. The feature can include a feature hole 205 in a silicon substrate 202. The feature hole 205 may have an underlayer 203 lining the sidewall or interior of the feature hole 205 and may form the interior surfaces. The feature hole 205 or other features may have a dimension near the opening, e.g., an opening diameter or line width of between about 10 nm to 500 nm, for example, between about 25 nm to about 300 nm. The feature hole 205 can be referred to as an unfilled feature or simply a feature. The vertical feature 201, and any feature, may be characterized in part by an axis 218 that extends through the length of the feature, with vertically oriented feature having vertical axes and horizontally oriented feature having horizontal axes. The underlayer 203 can be, for example, a diffusion barrier layer, an adhesion layer, a nucleation layer, a combination thereof, or any other applicable material. Non-limiting examples of underlayerscan include dielectric layersand conducting layers. Examplesof dielectric materials include oxides, such as SiO2, AI2O3; nitrides, such as SiN; carbides, such as nitrogen-doped silicon carbide (NDC) and oxygen-doped silicon carbide (ODC); and low k dielectrics, such as carbon doped SiO2. In particular implementations, an underlayer can be one or more of titanium, titanium nitride, tungsten nitride, titanium aluminide, tungsten, and molybdenum. In someAttorney Docket No. LAM1P127WO-12314-1WO embodiments, the underlayer is tungsten-free. In some embodiments, the underlayer is mo ly bde n u m-f ree .

[0044] In some embodiments, features are wordline features in a 3D NAND structure. For example, a substrate may include a wordline structure having an arbitrary number of wordlines (e.g., 50 to 450) with vertical channels at least 200 A deep. Examples of wordline features are described further below. Another example of a feature is a trench in a substrate or layer. Features may be of any depth. In various embodiments, the features may have an underlayer, such as a barrier layer or adhesion layer. Non-limiting examples of underlayers include dielectric layers and conducting layers, e.g., silicon oxides, silicon nitrides, silicon carbides, metal oxides, metal nitrides, metal carbides, and metal layers.

[0045] Figure 2C shows an example of a vertical feature 201 that has a re-entrant profile. A reentrant profile is a profile that narrows from the bottom, closed end, or interior of the feature to the feature opening. According to various implementations, the profile may narrow gradually and / or include an overhang at the feature opening. Figure 2C shows an example of the latter, with an underlayer 213 lining the sidewall or interior surfaces of the feature hole 205. Similar to Figure 2B, underlayer 213 can be a diffusion barrier layer, an adhesion layer, a nucleation layer, a combination thereof, or any other applicable material. Non-limiting examples of underlayers can include dielectric layers and conducting layers. The underlayer 213 forms an overhang 215 such that the underlayer 13 is thicker near the opening of the vertical feature 201 than inside the vertical feature 201.

[0046] In some implementations, features have one or more constrictions within the feature may be filled. Figure 2D shows examples of views of various filled features having constrictions. Each of the examples (a), (b), and (c) in Figure 2D includes a constriction 209 at a midpoint within the feature. The constriction 209 can be, for example, between about 15 nm - 20 nm wide. Constrictions can cause pinch off during the deposition of molybdenum in the feature using conventional techniques, with deposited metal blocking further deposition past the constriction before that portion of the feature is filled, resulting in voids in the feature. Example (b) further includes an overhang 215 (such as, a liner / barrier overhand) at the feature opening. Such an overhang could also be a potential pinch-off point. Example (c) includes a constriction 212 further away from the field region than the overhang 215 in example (b).Attorney Docket No. LAM1P127WO-12314-1WO

[0047] Horizontal features, such as in 3D memory structures, can also be filled. Figure 2E shows an example of a horizontal feature 250 that includes a constriction 251. For example, horizontal feature 250 may be a wordline in a 3D NAND (also referred to as vertical NAND or VNAND) structure. In some implementations, the constrictions can be due to the presence of pillars in a 3D NAND or other structure. Figure 2F presents a cross-sectional side view of a 3D NAND structure 210 (formed on a silicon substrate 202) having 3D NAND stack (left 225 and right 226), central vertical structure 230, and the plurality of stacked horizontal wordline features 220 with opening 222 on opposite sidewalls 240 of central vertical structure 230. Note that Figure 2F displays two "stacks" of the exhibited 3D NAND structure 210, which together form the "trench-like" central vertical structure 230. However, in certain embodiments, there may be more than two such stacks arranged in sequence and running spatially parallel to one another, the gap between each adjacent pair of s stacks forming a central vertical structure 230, like that explicitly illustrated in Figure 2F. These may be referred to as slits. In this embodiment, the horizontal wordline features 220 are 2D memory wordline features that are fluidically accessible from the central vertical structure 230 through the openings 222. Although not explicitly indicated in the figure, the horizontal wordline feature 220 present in both the 3D NAND stacks 225 and 226 shown in Figure 2F (i.e., the left 3D NAND stack 225 and the right 3D NAND stack 226) are also accessible from the other sides of the stacks (far left and far right, respectively) through similar vertical structures formed by additional 3D NAND stacks (to the far left and far right, but now shown). Each 3D NAND stack 225, 226 contains a stack of wordline features that are fluidically accessible from both sides of the 3D NAND stack through a central vertical structure 230. In the particular example schematically illustrated in Figure 2F, each 3D NAND stack contains 6 pairs of stacked wordlines. However, 3D NAND memory layout may contain any number of vertically stacked pairs of wordlines.

[0048] The wordline features in a 3D NAND stack can be formed by depositing an alternating stack of silicon oxide and silicon nitride layers, and then selectively removing the nitride layers leaving a stack of oxides layers having gaps between them. These gaps are the wordline features. Any number of wordlines may be vertically stacked in such a 3D NAND structure so long as there is a technique for forming them available, as well as a technique available to successfully accomplish (substantially) void-free fills ofthe vertical features. Thus, for example, a VNAND stack may include between 2 and 512 horizontal wordline features, between 2 and 256 horizontalAttorney Docket No. LAM1P127WO-12314-1WO wordline features, between 8 and 128 horizontal wordline features, or between 16 and 64 wordline features, and so forth (the listed ranges understood to include either recited endpoint).

[0049] Figure 2G presents a cross-sectional top-down view of the same 3D NAND structure 210 shown in the side view in Figure 2F with the cross-section taken through the horizontal section 260 as indicated by the dashed horizontal line in Figure 2F. The cross-section of Figure 2G illustrates several rows of pillars 255, which are shown in Figure 2F to run vertically from the base of the substrate 202 to the top of the 3D NAND structure 210. In some embodiments, the pillars 255 are formed from a polysilicon material and are structurally and functionally significant to the 3D NAND structure 210. In some embodiments, such polysilicon pillars may serve as gate electrodes for stacked memory cells formed within the pillars. The top-view of Figure 2G illustrates that the pillars 255 form constrictions in the opening 222 to wordline feature 220. Fluidic accessibility of wordline features 220 from the central vertical structure 230 via opening 222 (as indicated by the arrows in Figure 2G) is inhibited by pillars 255. In some embodiments, the size of the horizontal gap between adjacent polysilicon pillars is between about 1 and 20 nm. This reduction in fluidic accessibility increases the difficulty of uniformly filling wordline features 220 with the material. The structure of the wordline features 220 and the challenges of uniformly filling them with molybdenum material due to the presence of pillars 255 is further illustrated in Figures 2H, 21, and 21.

[0050] Figure 2H exhibits a vertical cut through a 3D NAND structure similar to that shown in figure 2F, but here focused on a single pair of wordline features 220 and additionally schematically illustrating a fill process which resulted in the formation of a void 275 in the filled wordline feature 220. Figure 21. also schematically illustrates void 275, but in this figure illustrated via a horizontal cut through pillars 255, similar to the horizontal cut exhibited in Figure 2G. Figure 21 illustrates the accumulation of molybdenum material around the constriction-forming pillars 255, the accumulation resulting in the pinch-off of opening 222, so that no additional molybdenum material can be deposited in the region of voids 275. Apparent from Figures 2H and 21 is that void-free molybdenum fill relies on migration of sufficient quantities of deposition precursor down through central vertical structure 230, through openings 222, past the constricting pillars 255, and into the furthest reaches of the wordline feature 220, priorto the accumulated deposition of molybdenum around pillars 255 causing a pinch-off of the openings 222 and preventing further precursor migration into wordline features 220. Similarly, Figure 2J exhibits a single wordline feature 220 viewed cross-sectionally from above and illustrates how a generally conformalAttorney Docket No. LAM1P127WO-12314-1WO deposition of molybdenum material begins to pinch-off the interior of wordline feature 220 due to the fact that the significant width of pillars 255 acts to particularly block, and / or narrow, and / or constrict what would otherwise be an open path through wordline feature 220. (It should be noted that the example in Figure 2J can be understood as a 2D rendering of the 3D features of the structure of the pillar constrictions shown in Figure 21, thus illustrating constrictions that would be seen in plan view rather than in a cross-sectional view.)

[0051] Three-dimensional structures may need longer and / or more concentrated exposure to precursors to allow the innermost and bottommost areas to be filled. Three-dimensional structures can be particularly challenging when employing molybdenum halide and / or molybdenum oxyhalide precursors because of their proclivity to etch, with longer and more concentrated exposure allowing for more etch as the precursors diffuse to other parts of the structure.

[0052] Figures 2K and 2L show examples of asymmetric trench structure DRAM bWL. Some fill processes for DRAM bWL trenches can distort the trenches such that the final trench width and resistance Rs are significantly non-uniform. Figure 2K shows an unfilled feature 261 and filled feature 265 that exhibits the line bending after fill. In this example, the features are a narrow asymmetric trench structure DRAM bWL. As shown, multiple features 283 are depicted on a substrate. These features 283 are spaced apart, and in some embodiments, adjacent features have a pitch between about 20 nm and about 60 nm or between about 20 nm and 40 nm. The pitch is defined as the distance between the middle axis of one feature to the middle axis of an adjacent feature. The unfilled feature 261 may be generally V-shaped, as shown in feature 283, having sloped sidewalls where the width of the feature narrows from the top of the feature to the bottom of the feature. The features widen from the feature bottom 273b to the feature top 273a. After some fill operations, line bending may be observed within the filled feature 265. In some situations, a cohesive force between opposing surfaces of a trench pulls the trench sides together, as depicted by arrows 267. The phenomenon is illustrated in Figure 2L and may be characterized as "zipping up" the feature. As feature 283 is filled, more force is exerted from the center axis 299 of feature 283, causing line bending. For example, molybdenum may be deposited on the sidewalls of the feature 283. Deposited molybdenum 284a and 284b on sidewalls of feature 283 thereby interact in close proximity, where molybdenum-molybdenum bond radius r is small, thereby causing cohesive interatomic forces between the smooth growing surfaces of molybdenum and pulling the sidewalls together, thereby causing line bending.Attorney Docket No. LAM1P127WO-12314-1WO

[0053] The methods described herein may be used for Mo feature fill in any of the structures and on any of the materials stacks described above. In some embodiments, they are advantageous for deposition in 3D-NAND structures as described with respect to Figures 2F-2K. Deposition in these structures is challenging for several reasons. First, these structures have dielectric surfaces on which the molybdenum is deposited for wordline fill. Metal nucleation on dielectric surfaces is difficult. Second, the structures have high aspect ratios and complex geometry. Uniform reactant distribution from the bottom to the top of the structure and throughout the wordline features is challenging.Molybdenum Deposition

[0054] Deposition of molybdenum as described herein involves reacting a Mo-containing precursor, also referred to as a molybdenum precursor. In some embodiments, a molybdenum halide compound as described above is used. In methods including surface treatment using a molybdenum halide compound, the same or different compound may be used for deposition.

[0055] In some embodiments, a Mo precursor is a molybdenum chloride (MoClx) compound also referred to as a molybdenum chloride precursor or MoClxprecursor. Molybdenum chloride precursors are given by the formula MoClx, where x is 2, 3, 4, 5, or 6, and include molybdenum dichloride (M0CI2), molybdenum trichloride (M0CI3), molybdenum tetrachloride (M0CI4), molybdenum pentachloride (M0CI5), and molybdenum hexachloride (MoCk). In some embodiments, M0CI5 or MoCkare used. While the description chiefly refers to MoClxprecursors, in other embodiments, other molybdenum halide precursors may be used. Molybdenum halide precursors are given by the formula MoXz, where X is a halogen (fluorine (F), chlorine (Cl), bromine (Br), or iodine (I)) and z is 2, 3, 4, 5, or 6. Examples of MoXzprecursors include molybdenum fluoride (MoFe). In some embodiments, a non-fluorine-containing MoXzprecursor is used to prevent fluorine etch or incorporation. In some embodiments, a non-bromine-containing and / or a non-iodine-containing MoXzprecursor is used to prevent etch or bromine or iodine incorporation.

[0056] In some embodiments, the feature may be filled using a molybdenum oxyhalide precursor. Molybdenum oxyhalide precursors are given by the formula MoOyXz, where X is a halogen (fluorine (F), chlorine (Cl), bromine (Br), or iodine (I)), and y and z are numbers greater than 0 such that MoOyXzforms a stable compound. Examples of molybdenum oxyhalides include molybdenum dichloride dioxide (MOO2CI2), molybdenum tetrachloride oxide (MoOCk),Attorney Docket No. LAM1P127WO-12314-1WO molybdenum tetrafluoride oxide (M00F4), molybdenum dibromide dioxide (MoChBr?), and the molybdenum iodides MoC l, and MO4O11I. It should be understood that as used herein the term molybdenum oxyhalide precursor may refer to a molybdenum oxyhalide precursor as described above or a molybdenum-containing oxyhalide precursor that includes molybdenum, oxygen, a halide and one or more other elements. In some embodiments, molybdenum oxyhalide or molybdenum-containing oxyhalides may include multiple different halogens (e.g., F and Cl and / or I and / or Br, etc.). A feature may be filled with molybdenum using a MoXxprecursor, MoOyXzprecursor, or a combination thereof.

[0057] A molybdenum precursor may be provided in a carrier gas, e.g., Ar in some embodiments. As an example, MoOzCh may be provide as 10%(vol) in a process gas with the balance being argon. Reference to partial pressure of the precursor in this document refer to the precursor only, and not the carrier gas.

[0058] For deposition of molybdenum into the feature, the molybdenum precursor may be reacted with a co-reactant. Examples of co-reactants include hydrogen (H2), silane (Sil-14), diborane (B2H6), germane (GeF ), ammonia (NH3), and hydrazine (N2H4). Ammonia and hydrazine may be used to deposit molybdenum nitrides or molybdenum oxynitrides.

[0059] In some embodiments, deposition of molybdenum may use a plasma-based process. Gas may be fed into a remote or in-situ plasma generator to generate plasma species. Examples of gas that may be used to generate plasma may be a hydrogen-containing gas, such as H2, nitrogencontaining gas, such as nitrogen (N2) and other gases, such as Ar and NH3. The plasma species may be inert or react with the molybdenum precursor to form a film.

[0060] A feature may be filled with molybdenum by atomic layer deposition (ALD) or chemical vapor deposition (CVD). Thermal ALD or plasma enhanced ALD (PEALD) may be used. Similarly, thermal CVD or plasma enhanced CVD (PECVD) may be used.

[0061] ALD is a surface-mediated deposition technique in which doses of a precursor and a reactant are sequentially introduced into a deposition chamber. One or more cycles of sequential doses of a molybdenum precursor and reactant may be used to deposit Mo. For example, in the deposition of an initial molybdenum layer, M0CI5 may be used as a precursor and F as a reducing agent. Doses of MoCk and H2 are sequentially introduced into the deposition chamber with a purge gas, such as argon, flowed between. For ALD, the temperature of the substrate and theAttorney Docket No. LAM1P127WO-12314-1WO pressure of the chamber may be controlled. For example, the substrate may be heated between 200°C and 800°C, e.g., between 250°C and 550°C or between 300°C and 500°C between 350°C and 450°C. In some embodiments, the chamber may be pressurized between 10 Torr and 200 Torr, e.g., between 50 Torr and 90 Torr. In some embodiments, the temperature and / or pressure may be used to control the rate of reactions. In some embodiments, the temperature and / or pressure may be used to control selectivity.

[0062] In some embodiments, the Mo precursor is a molybdenum fluoride (MoFx) compound, also referred to as a molybdenum fluoride precursor or MoFxprecursor. Molybdenum fluoride precursors are given by the formula MoFx, where x is 4, 5, or 6, and include molybdenum tetrafluoride (M0F4), molybdenum pentafluoride (M0F5), and molybdenum hexafluoride (MoFg).

[0063] MoFe can be advantageous as it has a boiling point of 34°C. Being a gas at standard pressure and 35°C allows MoFe to be delivered through a mass flow controller (MFC) at room temperature, without heating and without condensing and forming particles. However, MoFe is an aggressive etchant and exposure to MoFe during a process can result in etching instead of or in addition to Mo deposition. In some embodiments, deposition using MoFe involves providing a flow of MoFe in a process gas with the MoFe at a molar concentration of 0.01% or less. Concentration may be significantly lower in some embodiments, for example, 0.008% or less, 0.005% or less, or 0.004% or less. These values can also be expressed as parts per million (ppm) of a gas: 100 ppm (100 MoFe molecules per 1 million gas particles (atoms, molecules)) or less, 80 ppm or less, or 40 ppm or less. At temperatures between 200°C and 650°C, for example, a molar concentration at or below 0.004% results in CVD deposition when flowed with H2 and argon. Higher temperatures may be used to favor the deposition reaction and allow higher concentrations of MoFe, e.g., up to 0.01%. In some embodiments, concentrations may be 0.0039% or .0035% or less. In some embodiments, the MoFg concentration is at least 0.00004% or at least 0.0001%. Concentration may be very low with an exposed metal surface to grow on, for example.

[0064] Deposition using MoF6with H2 as reducing agent occurs only at unusually low concentration. As an example, for 0.5 seem of MoFg, a total flow rate of 13,500 seem may be used, for a MoFg concentration of 0.0037%. Deposition using metal halides and hydrogen generally involves much higher concentrations. For example, deposition of molybdenum usingAttorney Docket No. LAM1P127WO-12314-1WO molybdenum hexachloride and hydrogen can be performed using concentrations 5 to 10 times higher than those used for MoFe.

[0065] In some embodiments, MoFe may be used at higher concentrations and lower temperatures with a reducing agent that is stronger than that of hydrogen. Lower temperatures can reduce or prevent etching with MoFe; however, at low temperatures H2 may not result in deposition. Stronger reducing agents such silane, disilane, polysilanes and diborane may be used for deposition at lower temperatures (e.g., below 200°C). The resulting films may not be pure molybdenum and in some cases are more resistive than those deposited using H2 as the reducing agent. For these reasons they may not be appropriate for some applications.

[0066] In some embodiments, molybdenum fill may involve CVD. In a CVD process, the molybdenum precursor and reactant are in vapor phase together in the deposition chamber. Generally speaking, a CVD process fills a feature faster than an ALD process. In one example, the precursor may be a molybdenum oxychloride, such as MOO2CI2, and is flowed into the chamber with a reactant, such as H2. In this example, the wafer is simultaneously exposed to the precursor and reactant, which react and fill features with Mo. In one example, MoFe is flowed into the chamber with a reactant, such as H2. In this example, the wafer is simultaneously exposed to the precursor and reactant, which react and fill features with Mo.

[0067] In still some other embodiments, a feature may be filled using a pulsed CVD process. The pulsed CVD process continuously flows a reactant intoa chamberwhile pulses ofa precursorflow into the chamber. For example, H2gas may be flowed into the chamber and is continuously flowing into the chamber while the molybdenum-containing precursor is intermittently flowing into the chamber. The temperature of the substrate and pressure in the chamber may be controlled during a CVD operation.

[0068] Plasma-enhanced CVD may be used in which a plasma is ignited during the deposition. In a pulsed CVD process, a plasma may be ignited during deposition cycle or during, e.g., pulses of the hydrogen reactant. In some embodiments, a remote plasma may be used. The plasma may be remotely-generated or direct. Further it may be generated by any appropriate plasma generator including a capacitively-coupled plasma generator or an inductively-coupled plasma generator. A microwave plasma generator may be used.Attorney Docket No. LAM1P127WO-12314-1WO

[0069] In addition to the molybdenum halides and molybdenum oxyhalides described herein, the molybdenum films may be deposited using organometallic and / or sulfur-containing precursors. Organometallic molybdenum-containing compounds and / or sulfur-containing molybdenum-containing compounds may be used as molybdenum precursors in some embodiments. Examples of these are given in PCT publication W02023250500, incorporated by reference herein.Selective deposition

[0070] Molybdenum may be selectively deposited into a feature using the methods described herein. Selective deposition refers to preferential deposition on a first material with respect to a second material. Molybdenum deposition and growth may be easier on a metal material relative to molybdenum deposition and growth on a dielectric material. For example, a feature may have a sidewall surface of SiC and a TiN plug in a bottom portion of the feature. In selective deposition, molybdenum is deposited intothe feature and may grow on the TiN plug but not grow (or grow to a lesser extent) on the SiCh sidewall surfaces.

[0071] Process conditions such as the precursor gas, the reducing agent, process temperature, process pressure, and exposure time may affect the selectivity of the molybdenum film being deposited. Process temperatures for selective deposition of the molybdenum film may be between 200°C to 800°C, e.g., 250°C to 550°C, or 300°C to 500°C. At these temperatures, the molybdenum film is selectively deposited on conductive metal or metal compound surfaces, such as a TiN surface, in a feature relative to dielectric surfaces.

[0072] Different precursor gases may have different process windows in which molybdenum film may be selectively deposited. Generally speaking, MoCLgas has a large process window, i.e., large temperature and pressure range, where the precursor gas retains its selectivity. For example, M0CI5 may be selectively deposited on a metal material with respect to a dielectric material where the process temperature is 200°C to 800°C, e.g., 250°C to 550°C, or 300°C to 500°C. Generally speaking, higher process temperatures and higher process pressures reduce the selectivity of the deposited gas. M0CI5 deposits selectively on metals, titanium nitride (TiN) and other conductive materials relative to dielectric materials at a wide range of temperatures.

[0073] M0CI5 may react with a different reactant to deposit a molybdenum film. Described below are examples of deposition of molybdenum film within a feature using a M0CI5 precursorAttorney Docket No. LAM1P127WO-12314-1WO and different process controls. In a first example, the M0CI5 precursor is reacted with a hydrogen (H2) reactant using the deposition methods described above. In the description herein, the metal precursors are reacted with H2 as a co-reactant (also referred to as a hydrogen reactant or H2 reactant). Other reactants may be used instead of hydrogen including other hydrogen-containing reactants such Sil- , B2H6, NH3, as appropriate. Reactants such as B2H6 and / or Sil- are stronger reducing agents and generally show reduced selectivity. They can also result in higher resistivity. Thus, in some embodiments, using H2 as described herein is advantageous. As noted above, process temperatures for selective deposition of the molybdenum film from M0CI5 may be between 200°C to 800°C, e.g., 250°C to 550°C, or 300°C to 500°C. At these temperatures, the molybdenum film is selectively deposited on conductive metal or metal compound surfaces, such as a TiN surface, in a feature relative to dielectric surfaces. The molybdenum film grows from the locations where the conductive surfaces are located in a feature. If the conductive surface is a TiN plug at the bottom of the feature, the molybdenum film may be deposited and grown from the bottom of the feature. In a second example, the molybdenum film may be deposited using the M0CI5 precursor and the H2 reactant, but at higher temperatures, i.e., above 800°C. This process window may have the molybdenum film deposited on both the dielectric and conductive surfaces within the feature. The deposition of the molybdenum film on the dielectric surface may be used to create a barrierless molybdenum layer in the feature.

[0074] In some embodiments, selective deposition is performed using a MoFxprecursor. Molybdenum fluoride precursors are given by the formula MoFxas described above. As indicated above, MoFe can be advantageous for ease of delivery. Deposition of molybdenum from MoFg at the low concentrations disclosed above results in high (at least 100:1) selectivity on elemental metal surfaces (e.g., W, Mo, Cu) relative to oxides and nitrides such as silicon oxide and titanium nitride. MoFealso deposits selectively on metals with respect to dielectric materials, though it is less selective than M0CI5. As can be seen, after a delay, MoFe deposits on thermal oxide. Selectivity of molybdenum halides can also be affected by operating at conditions (e.g., concentration, temperature, etc.) at which the molybdenum halide also etches.

[0075] Selective deposition using a molybdenum oxyhalide precursor is much more d ifficu It than using a molybdenum halide precursor. However, the surface treatments described above significantly improve selectivity of Mo deposition from MOO2CI2. As indicated above, examples of MoOyXz precursors include M0O2CI2, M0OCI4, M00F , MoChB^, MOO2I, and MO4O11I. The feature may be filled using ALD, plasma enhanced ALD, chemical vapor deposition (CVD), or plasmaAttorney Docket No. LAM1P127WO-12314-1WO enhanced CVD. For ALD or CVD, H2 may be the reducing agent. Molybdenum deposits more quickly using a molybdenum oxyhalide precursor than the MoClxprecursor used in the surface treatment. For example, a MoOyXzprecursor may deposit molybdenum at a deposition rate at least twice as fast as a MoClxprecursor for a non-plasma process.Non-selective Deposition

[0076] The selectivity described above may be reduced or eliminated using plasma deposition in some embodiments, such that the molybdenum is deposited on different materials. This may be referred to as non-selective deposition. When ALD processes are used, the non-selective deposition may be conformal to the contours of surface. The plasma is generally an in-situ or direct plasma for non-selective deposition.

[0077] Examples of plasma processes include plasma-enhanced ALD (PEALD) or plasma enhanced CVD (PECVD) processes using a molybdenum halide precursor. In some embodiments, the molybdenum halide precursor is M0CI5 or MoFg. A molybdenum oxyhalide may also be used, with examples including MOO2CI2 or M0OCI4. Hydrogen (H2) or other reducing agent may be used for the PEALD or PECVD deposition.

[0078] For PECVD deposition, the molybdenum precursor can be co-flowed with the reducing agent. For MoFe, the concentration of the MoFe is as described above, with the mixture flowed into a plasma generator. Remote or direct plasmas may be used. In some embodiments, a capacitively-coupled direct plasma that is generated in the chamber is employed.

[0079] Non-selective deposition may also be a thermal process using molybdenum oxyhalides. For example, thermal MOO2CI2 and H2 may be used to deposit a molybdenum layer non-selectively. Temperatures at or above 450°C may be used in for thermal deposition from MOO2CI2 and H2.

[0080] To reduce selectivity, an ALD process may be performed to deposit a Mo-containing nucleation layer. For nucleation layer deposition, a stronger reducing agent than hydrogen is employed. This can allow the film to grow on surfaces that face nucleation delay with hydrogen as reducing agent. As described further below, such a reducing agent can be a silicon-containing or boron-containing reducing agent such as silane (SJH ) or diborane (B2H6). Germanium-containing reducing agents (e.g., GeHz may be used. These may be used to deposit an elemental molybdenum film. In other embodiments, a reducing agent such as ammonia (NH3) may be used.Attorney Docket No. LAM1P127WO-12314-1WO In such cases, the molybdenum layer may be a molybdenum nitride or molybdenum oxynitride layer, depending on the presence of oxygen in the molybdenum precursor. This oxynitride layer or nitride layer may be converted into an elemental molybdenum layer in the subsequent process.

[0081] When using MoFe, the concentration of MoFe in the MoFe dose may as described above, i.e., 0.01% or less, 0.008% or less, 0.005% or less, or 0.004% or less of the total gas flowed into the chamber. Alternatively, because a stronger reducing agent than hydrogen is used in the subsequent operation, a higher concentration (e.g., up to 0.1% molar) may be used during the MoFe. Some amount of a reducing agent may be present to suppress etching. As described above, this can be between 0.5% and 10% or between 1% and 9% H2. Another reducing agent may be included instead of or in addition to hydrogen. The balance is wholly or predominately argon or other inert gas. During the reducing agent dose, the dose is wholly or predominately the reducing agent, with some amount (e.g., up to 10%, or between 1% and 9%) being argon in some embodiments, and the remainder the reducing agent. After deposition of the nucleation layer, a bulk molybdenum layer can be deposited using H2 as a reducing agent by any of the methods described above, including thermal or plasma-enhanced ALD or CVD.Integration processes including etch and / or inhibition

[0082] Etch operations may be used in the methods for filling features with Mo films. Etch operations remove materials such as metals and nitrides from the feature. For example, an etch process may partially or completely remove a liner (e.g., a TiN) layer from a feature. In another example, the etch process may be used to reduce the thickness of a liner layer. Etch processes may be performed as part of a pre-treatment process as described elsewhere in the disclosure and / or as part of a deposition-etch-deposition process in which molybdenum is etched.

[0083] An etchant is any compound used to remove a material such as a layer, byproduct or contaminant from a surface. In some embodiments, the etchant is a halogen-containing etchant such as chlorine (CI2), fluorine (F2), bromine (Brz), iodine (I2), hydrogen chloride (HCI), hydrogen fluoride (HF), hydrogen iodide (HI), chlorine trifluoride (CIF3), ferric chloride (FeCU), trifluoromethane (CHF3), fluoromethane (CH3F), octafluorocyclobutane (C4F8), hexafluorobutadiene (C4F6), hexafluorocyclopentad iene (CsFe), carbon tetrafluoride (CF4), carbon tetrafluoride (CCI4), nitrogen trifluoride (NF3), boron trichloride (BCI3), boron trifluoride (BF3), hydrogen bromide (HBr), sulfur tetrafluoride (SF4), sulfur hexafluoride (SFe), thionyl chloride (SOCh), phosphorus pentafluoride (PFs), phosphorus trifluoride (PF3), silicon tetrabromide (SiBu),Attorney Docket No. LAM1P127WO-12314-1WO or a combination thereof. In some embodiments, a single etchant may be sufficiently effective. In some embodiments a combination including more than one etchant may be utilized. Examples of combinations include oxygen (O2) with one of the above halogen-containing etchants such as chlorine and oxygen; or fluorine and oxygen. Alternatively, carbon dioxide (CO2) may be combined with one of the above halogen-containing etchants. If a combination of etchants is utilized, they may be flowed through delivery lines together (concomitantly), or sequentially (one following the other). The etchant may be co-flowed with an inert gas, such as argon. In some embodiments, etchants are combined. For example, the halogen-containing etchant may be coflowed with a non-halogen containing etchant.

[0084] In some embodiments, the etchant is M0CI5, MoFe, WFe, WCI5, or any of the other metal halides described above. For example, an etch operation, in some embodiments, may involve soaking the feature soaked in a Mo halide. In some embodiments, an etch operation involves soaking the feature with a MoClxsuch as M0CI5. In some embodiments, the soak may be done continuously with the Mo halide gas. In some embodiments, the soak may be pulsed, cycling the Mo halide with a purge gas, such as argon (Ar).

[0085] A molybdenum halide precursor may be used for both deposition and etch operations. For example, in certain process windows, a M0CI5 precursor may concurrently grow a Mo film and etch away a metal or metal compound film in the feature. The process is considered a net etch operation if the rate of material removed is greater than the material deposited by the precursor. The speed at which the precursor deposits material and etches material may be controlled by a variety of process conditions, including the type of reactant used and the process temperature. Generally speaking, the lowerthe temperature, the higher the ratio of etching away material is relative to deposition of material. At higher temperatures, the same precursor and reactant may be used as a net deposition operation, i.e., the amount of material deposited is greater than the material removed. For example, M0CI5 precursor and H2 reactant may be used in an etch operation when the process temperature is below 400°C. M0CI5 and H2 may be used in a deposition operation when the process temperature is above 550°C.

[0086] In some embodiments, the MoClxprecursor at high temperatures, e.g., above 550°C, may continue to etch material at a faster rate than depositing material. For example, M0CI5 may be used to etch a feature by a soak without a reactant. In this example, the temperature may be as high as 700°C and will continue to etch away material from the feature. In operations where theAttorney Docket No. LAM1P127WO-12314-1WO feature is soaked in a M0CI5 without a reactant, the increased temperature may increase the rate at which material is etched from the feature.

[0087] A feature may have surface oxide or contaminants on it. For example, the surface of an underlying TiN, WN, or W layer may be oxidized. If left, the oxidized surface can result in higher resistivity. Clean operations are used to remove such oxides and contaminants. In some embodiments, the clean operation may have the feature soaked in a Mo precursor gas, typically a Mo halide. Similar to the etch operations described above, the precursor gas may be a MoClxprecursor. In some embodiments, the soak may be done continuously. In some embodiments, the soak may be pulsed, cycling MoClxand a purge gas, such as argon (Ar). The precursor may be a non-oxygen Cl-containing Mo compound able to remove oxidation from the feature's surfaces. Examples of MoClxcompounds are given above. A Cl-containing precursor may be used where traditional cleaning with thermal or plasma H2 does not work, such as where the oxidized surface is stable on the surface material. A Cl-containing precursor is less likely to over-etch a feature's liner layer or attack a feature's surfaces than an F-containing compound.

[0088] An etch may be thermal or plasma-enhanced. In some embodiments in which material in lateral features is etched, a thermal etch to allow the etchant chemistry to diffuse into the feature.

[0089] Inhibition operations may be used in the methods for filling features with Mo films. Inhibition operations inhibit molybdenum nucleation on a surface. As an example, an inhibition operation may be used to inhibit nucleation on only part of a feature, extending from the feature opening to some depth within the feature. In some embodiments, an incoming structure may be treated to inhibit molybdenum nucleation. For example, a feature having dielectric sidewalls and a conductive bottom surface may be treated such that nucleation is inhibited on the upper portion of the sidewalls, facilitating selective deposition. The inhibition treatment may be repeated during the subsequent deposition to maintain its effectiveness.

[0090] A dielectric material may be treated with a halogen-containing chemistry to inhibit molybdenum nucleation. Examples include F2, NF3, BCh, M0CI5, and CI2. Each of these chlorinates or fluorinates oxides inhibiting further nucleation.

[0091] Inhibition operations may also be performed as part of deposition-inhibition-deposition (DID) techniques. In some embodiments, a portion of a molybdenum film is treated to inhibitAttorney Docket No. LAM1P127WO-12314-1WO subsequent deposition. Examples of inhibition chemistries include nitrogen-containing chemistries including N2, and NH3, and well as halide-containing chemistries such as alkyl halides. An inhibitor such as N2 may be co-flowed with a molybdenum precursor and / or H2, for example. The inhibition may be a plasma or thermal operation. If plasma, a remote or direct plasmas may be used. Other examples of inhibition operations can include exposure to oxygen-containing, carbon-containing, and phosphorous-containing thermal or plasma chemistries. In some embodiments in which material in lateral features is etched, a thermal inhibition allows the inhibition chemistry to diffuse into the feature.

[0092] Alkyl halides may be used to inhibit nucleation on molybdenum-containing surfaces for DID operations as well as to modify other surfaces including metal nitrides such as TiN. In some embodiments, the halogen-containing compound is an alkyl halide (e.g., a tertiary alkyl halide, such as t-butyl chloride or t-butyl iodide). In some embodiments, the halogen-containing compound is an iodine-containing compound. Further examples of inhibitors include trimethylsilylchloride [(CH3)3SiCI] and trimethylsilyl-dimethylamide [(CHshSiNfCHah- Chlorine (CI2) is an etchant and can also inhibit growth on molybdenum. Inhibition is observed at substrate temperatures of about 450°C to 600°C for non-plasma exposure to CI2-

[0093] De-inhibition operations may be used to reduce the effect of inhibition, either before or after the subsequent deposition. This can be used to further tailor the fill profile. Examples of de-inhibition operations include H2 soak, NH3 soak, and H2 plasma exposure. Soak operations may be continuous flow or pulsed.

[0094] Also provided herein are deposition-etch-deposition (DED) techniques and depositioninhibition-deposition (DID). These may be used to tailor deposition into features during interconnect metallization and for memory applications. The DED operations described herein may be used for logic applications such as interconnects as well as memory applications. Filling a 3D NAND structure using a DED technique may also be performed. In some embodiments, multiple DED operations are used to fill a feature. The same or different chemistries may be used for each deposition. The molybdenum precursor may be a molybdenum halide or molybdenum oxyhalide as described above or a molybdenum organometallic precursor. The same or different chemistries may be used for each etch operation.

[0095] During the etch, a high flow short dose time may be employed to achieve an anisotropic etch. As indicated above, a pre-treatment may be used to increase etch rate as well as tailor etchAttorney Docket No. LAM1P127WO-12314-1WO profile. For example, etch may be preceded by an anisotropic oxidation or nitridation. This can help etch only in the top of the feature (for vertical features) or outer part of the feature (e.g., outer wordlines in a 3D NAND structure). Examples of oxidation operations include exposure to O2 or O3 or oxygen-containing plasmas. Examples of nitridation operations include exposure to NH3 or N2 or nitrogen-containing plasmas. Post-treatments can be used to remove impurities after etch. For example, exposure to a halosilane may be used to remove fluorine or chlorine. Exposure to H2 can be used to remove impurities. According to various embodiments, a posttreatment may be performed after every dose of the etchant or less frequently, for example, at the end of multiple cycles that include etching.

[0096] In some embodiments, the DED sequences may include one or more inhibition operations. An inhibition operation is an operation to inhibit nucleation or formation of molybdenum film in a subsequent deposition. It may be used to tune a deposition profile. Examples of inhibition chemistries include nitrogen-containing chemistries including NF3, N2, and NH3, and well as halide-containing chemistries such as alkyl halides, EhHe, and CI2. An inhibitor such as N2 may be co-flowed with a molybdenum precursor and / or H2, for example. The inhibition may be a plasma or thermal operation.

[0097] De-inhibition operations may be used to reduce the effect of inhibition, either before or after the subsequent deposition. This can be used to further tailor the fill profile. Examples of de-inhibition operations include H2 soak, NH3 soak, and H2 plasma exposure. Soak operations may be continuous flow or pulsed. Prolonged precursor and / or reactant dose time after an inhibition treatment may also be used to reduce or eliminate inhibition effects.

[0098] A process may use various permutations of Depl, Dep2, Etch, Inhibition and de-lnhibition operations to tailor fill. Examples of process sequences are:Dep - Etch - DepDep - Inhibition - DepDep - Etch(x) - Inhibition(y) - DepDep - Etch(x) - Inhibition(y) - Dep - de-lnhibition - DepDep - Inhibition - Etch - DepDep - Etch - Dep - Inhibition - DepAttorney Docket No. LAM1P127WO-12314-1WO Dep - Etch - Dep - Inhibition - Dep - de-lnhibition - DepDep - Inhibition - Dep - Etch - DepDep - Oxidation - Etch - DepDep - Nitridation - Etch - Dep

[0099] In some embodiments, the dep-etch-dep operations disclosed herein may be integrated into single chamber metallization processes as described above.Barrierless ALD Molybdenum for NAND

[0100] Embodiments of the disclosure include metallization schemes that include fluorine-free barrierless ALD molybdenum for both memory and logic applications. For the NAND Flash wordlines, the ALD-Mo process described herein nucleates well on dielectrics (e.g., AI2O3), enabling low wordline resistance in geometrically constrained NAND Flash memory arrays. The methods may be used to for wordline deposition as NAND Flash devices transition from 300 to 1000 wordlines.

[0101] When integrated with a conformal TiN barrier, ALD tungsten (ALD-W) has outstanding fill capability, near-zero electromigration (high melting point), well developed wet and dry etch processes, and relatively low cost. But the resistivity of the TiN-W metallization stack is rapidly becoming a bottleneck. The product of (electron mean free path A) and (metal resistivity p) scales well with line resistance. By this metric, any of Mo, Ru, Ni, Co, Rh, Cu, Al, Os, and Ir can be used to produce lines with lower resistance than TiN-W. Of these, Mo, Ru, Ni, Co, Rh, Cu, Al may be used with suitable wet or dry etch processes. Ni, Co, Ru, and Mo have melting points above 1400°C without being prohibitively costly. The methods described herein allow Mo to be deposited directly on dielectrics using ALD. This eliminates the need for a high-resistivity Liner or Barrier layer, such as TiN, freeing up valuable volume for more low resistance metallic conductor.

[0102] Molybdenum chlorides and molybdenum oxychlorides with vapor pressures advantageous for ALD include MOO2CI2, MoOCU, and M0CI5. Self-etching behavior of these precursors is roughly correlated with their Cl / Mo ratios. Self-etching in NAND Flash word lines can lead to upper wordlines being etched away completely (high concentration), normal deposition in middle wordlines (ideal concentration), and precursor depletion resulting in no deposition or etch at the bottom (low concentration). Because of its high vapor pressure and relatively etching, M0O2CI2 is advantageous for ALD in NAND wordlines.Attorney Docket No. LAM1P127WO-12314-1WO > > >

[0103] As described above, ALD of Mo is performed directly on dielectric surfaces, allowing barrier-free metallization. This is challenging - it is difficult to grow metal films directly on dielectrics because Metal-Metal attraction is typically much stronger than Metal-Dielectric attraction. This can result in long nucleation delays, poor adhesion, and metal agglomeration. These problems can be overcome by introducing a metal nitride between the dielectric and the metal (e.g., TiN-W). Dielectric-nitride attraction is comparable to nitride-nitride attraction and nitride-metal attraction is comparable to metal-metal attraction). Metal nitrides create a bridging layer that nucleate well on dielectrics and create strong nitride-metal interfaces.

[0104] Thin metal-nitride layers have high resistivities (102to 103micro-Ohm-cm) and consume volume that could otherwise be filled with low resistivity (on the order of 101micro-Phm-cm). Figure 3 shows top-down and cross-sectional views of a flash memory array geometry.

[0105] 300-pair Flash wordline pillar-pillar opening widths are on the order of 10 to 20 nm and will are shrinking fast. In a typical 300-pair Flash memory array, a 3 nm interfacial liner would leave just 30% of the available wordline volume for low resistivity metal. This results in wordline resistance 5 times that of wordline with no interfacial liner. This is illustrated in Figures 4A and 4B (geometric parameters a, d are defined in Figure 3). Figure 4A shows the fraction of NAND flash wordline volume occupied by a metal (excluding interfacial liner volume) and Figure 4B shows the normalized wordline resistance for various liner thicknesses. Even a relatively modest lnm interfacial liner has only 75% of its available wordline volume filled with metal, with wordline resistance increased by 50% relative to all-metal fill.

[0106] Figure 4C is a process flow diagram illustrating operations in a method of depositing a conductive material. In operation 402, a conformal nucleation layer is formed on a structure by an atomic layer deposition (ALD). In an ALD method, the substrate may be exposed in cycles suchAttorney Docket No. LAM1P127WO-12314-1WO that the substrate is first exposed to a pulse of a suitable metal-containing precursor, then the precursor is optionally purged, then the substrate is exposed to a pulse of a reducing agent, and then the reducing agent is optionally purged, and such cycles may be repeated until a desired thickness of the nucleation layer is formed on the substrate. It will be appreciated that the order of precursor and reducing agent may be reversed such that the sequence may be initiated by a reducing agent dose followed by a metal containing precursor dose.

[0107] In some embodiments, the reducing agent is ammonia (NH3) or other nitrogencontaining reducing agent such hydrazine (N2H4). NH3 chemisorption on dielectrics is more favorable than that of hydrogen (H2). In some embodiments, the reducing agent and precursor are selected such that they react without reducing agent dissociation. NH3 reacts with metal oxychlorides and metal chlorides without dissociation. This is in contrast to, for example, ALD from metal oxychlorides that uses H2 as a reducing agent; H2 dissociates on the surface to form adsorbed atomic hydrogen, which results in very low concentrations of reactive species and low surface coverage during initial nucleation of metal on the dielectric surface. By using NH and metal oxychloride or metal chloride precursors, nucleation delay is reduced or eliminated at deposition temperatures upto hundreds of degrees lower than used by H2 reduction of the same metal precursors.

[0108] In some embodiments, the reducing agent may be a boron-containing or silicon-containing reducing agent such as diborane (B2H6) or (Sil- ). These reducing agents may be used with metal chloride precursors; with metal oxychlorides, however, the B2H6 and Sil-U will react with water formed as a byproduct during the ALD process and form solid B2C>3 and SiC>2, which are insulating and will remain in the film, increasing resistivity. Use of NH3 also has improved adhesion over B2H6 and SiFU ALD processes on certain surfaces including AI2O3.

[0109] Examples of metal oxychloride and metal chloride precursors include molybdenum pentachloride (M0CI5), molybdenum oxychlorides such as molybdenum dichloride dioxide (MOO2CI2) and molybdenum oxytetrachloride (MoOCk), tungsten pentachloride ( WCI5), tungsten hexachloride (WCI6), tungsten tetrachloride (WCI4), tungsten dichloride (WCI2), and tungsten oxychlorides (WOxCly) such as tungsten oxytetrachloride (WOCI4).

[0110] The metal chloride and metal oxychloride may be useful in embodiments in which fluorine incorporation is a concern. However, in some embodiments, fluorine-containingAttorney Docket No. LAM1P127WO-12314-1WO precursors may be used. These include metal fluorides such as tungsten hexafluoride (WFe), molybdenum hexafluoride (MoFe), and molybdenum pentafluoride (M0F5).

[0111] The resulting nucleation layer is generally not a pure elemental film but a metal nitride or metal oxynitride film. In some embodiments, there may be residual chlorine or fluorine from the deposition, particularly if the deposition is performed at low temperatures. In some embodiments, there are no more than trace amounts of residual chlorine or fluorine. In some embodiments, the nucleation layer is an amorphous layer. Impurities in the film (e.g., oxygen, NH3, chlorine or other halogen) facilitate growth of an amorphous microstructure. In some embodiments, the nucleation layer as deposited is an amorphous metal oxynitride layer or an amorphous metal nitride layer. The amorphous character templates large grain growth in the subsequently deposited conductor. The surface energy of nitride or oxynitride relative to an oxide surface is much more favorable than that of a metal on an oxide surface, facilitating formation of a continuous and smooth film on the dielectric. This allows formation of thin, continuous layers. Example thicknesses of the nucleation layer range from 5-30A as deposited. Depending on the temperature, this may be about 5-50 ALD cycles for example.

[0112] As described below, during subsequent processingthe nucleation layer may be converted to a pure (or less impure) elemental metal film with the thickness decreasing.

[0113] Substrate temperature for nucleation layer deposition may range, for example, from 300°C-600°C. In some embodiments, low temperatures may be used. Such temperatures may be less than 500°C, less than 550°C, less than 450°C, less than 400°C, or less than 350°C. Low temperatures may be used for improved step coverage. In addition, low temperatures may increase the amount of impurities in the nucleation layer, increasing the amorphous character, which in turn may increase grain size of the subsequently deposited conductor.

[0114] The surface on which the nucleation layer is deposited depends on the particular application. In some embodiments, the nucleation layer is deposited directly on a dielectric (e.g., silicon oxide, aluminum oxide, silicon nitride, etc.) surface. In some embodiments, the nucleation layer is deposited directly on a titanium nitride or other surface. As discussed further below, by performing operation 402, the subsequent elemental metal deposition may be performed on any surface.Attorney Docket No. LAM1P127WO-12314-1WO

[0115] After deposition of the nucleation layer, an optional operation 404 may be performed. In operation 404, lower temperature ALD cycles a metal conductor and a reducing agent are performed. The "lower" temperature refers to the temperature in operation 404, if performed, being lower than the subsequent operation 406. Example temperatures may be less than 500°C, less than 550°C, less than 450°C, less than 400°C, or less than 350°C. In this operation, the reducing agent is different than in operation 402, and in particular examples may be hydrogen (H2). In particular, H2 may result in deposition of an elemental film with significantly fewer impurities than in the nucleation layer. The temperature may be the same temperature as used in operation 402 in some embodiments. The metal precursor may also be the same ora different precursor than in employed in operation 402. In some embodiments, the same precursor is used, with only the reducing agent changed. In some embodiments, operation 404 may facilitate conversion of the metal nitride or metal oxynitride nucleation layer to an elemental metal film. According to various embodiments, operation 404 may or may not deposit an appreciable amount of film of the main conductor.

[0116] In a further optional operation 406, the substrate temperature is raised. In embodiments, in which operation 404 is performed, operation 406 is also performed. In other embodiments, operation 406 may be performed. For example, if the nucleation layer deposition occurs at relatively low temperatures (e.g., below 400°C), the temperature may be raised in operation 406 to a higher temperature at which deposition of the main conductor will be performed. In some embodiments, the temperature may be greater than 500°C, and in some embodiments, greater than 600°C. In some embodiments, a lower temperature (e.g, between 400°C and 500°C, endpoints inclusive) may be used for bulk deposition. The temperature may or may not be raised, depending on the temperature of previous operations.

[0117] The method may then proceed to an operation 408 (from any of operation 402, 404, or 406) in which the main conductor is deposited by ALD. As in operation 404 (if performed), H2 may be used as a reducing agent.

[0118] Examples of metal oxychloride and metal chloride precursors that may be employed in operations 404 and 408 include molybdenum pentachloride (M0CI5) and molybdenum hexachloride (MoCk), molybdenum oxychlorides such as molybdenum dichloride dioxide (MOO2CI2) and molybdenum oxytetrachloride (MoOCU), tungsten pentachloride (WCI5), tungstenAttorney Docket No. LAM1P127WO-12314-1WO hexachloride (WCk), tungsten tetrachloride (WCI4), tungsten dichloride (WCI2), and tungsten oxychlorides (WOXCIY) such as tungsten oxytetrachloride (WOCI4).

[0119] During one or more of operations 404-408, the nucleation layer is converted to an elemental metal layer. This may also be characterized as removing impurities, i.e., any non-metal constituent. The nucleation layer may have greater impurities than the subsequently deposited elemental layer, but they are sufficiently removed such that the stack resistivity is the same or similar to a stack that does not include a nucleation layer. The thickness will also decrease; for example, a 30A as-deposited film may contribute about 10A metal to the stack.

[0120] According to various embodiments, one or more of the following may be employed to facilitate conversion of the nucleation layer to an elemental metal film: 1) depositing the bulk conductor at a higher temperature (e.g., 550°C) than the nucleation layer is deposited, 2) performing lower temperature ALD Hz / metal precursor cycles as described with reference to operation 404 above, and 3) in-situ deposition of the bulk layer, such that the nucleation layer is not exposed to air or otherwise oxidized before bulk deposition. Mo oxychlorides and W oxychlorides in particular are relatively easy to convert to elemental metal. The resulting converted nucleation layer and pure metal layer may each be characterized as having fewer than 1% atomic impurities.

[0121] Deposition of pure ALD-Mo on dielectrics creates low resistivity ALD-Mo films without a high resistance layer at the metal-dielectric interface. See, e.g., Figure 5, which shows an SEM of ALD-Mo on AI2O3. No interfacial layer is present.

[0122] Moreover, there is no nucleation delay. See Figure 6, which shows subsequent Mo growth from H2 and molybdenum oxychloride ALD cycles on a nucleation layer deposited on an AI2O3 surface as described above as compared to Mo growth from H2 and molybdenum oxychloride ALD cycles directly on an AI2O3 surface at 550°C. Curve A shows Mo growth on approximately 10A nucleation layer deposited from NH3 / M0 oxychloride and curve B shows Mo growth on AI2O3. As can be seen, ALD deposition from H2 reduction of the Mo precursor shows a steady growth on the Mo nucleation layer deposited from NH3 reduction of the Mo precursor, with no nucleation delay. By contrast, the H2 / M0 precursor ALD cycles result in no deposition on AI2O3. As shown in Figure 6, stand ALD-Mo (e.g., H2-M0) is unable to nucleate directly on dielectrics. Resistivity is low, as shown in Figure 7. In particular, it is significantly lower than the resistivity of TiN-W stacks.Attorney Docket No. LAM1P127WO-12314-1WO

[0123] Additional processes to deposit ALD-Mo directly on dielectrics are described in US Provisional Patent Applications 63 / 752,562, filed January 31, 2025; 63 / 752,526 filed January 31, 2025; and 63 / 741,767, filed January 3, 2025. These applications are incorporated by reference herein for all purposes.Modeling 3D Wordline Fill and Deposition Processes Derived Therefrom

[0124] Provided herein methods of filling wordlines and methods of optimizing wordline fill processes. In some embodiments, the methods involve adjusting deposition conditions from the beginning of the process, when the wordlines are wide open, to seam closure at the end, we can maintain complete saturation with reasonable dose times.Chemical Kinetics Modeling

[0125] The overall ALD-Mo reaction for wordline fill according embodiments of the disclosure:Mo + 2H2O + 2HCI (1) can be broken down as follows:1. Dissociative adsorption of MOO2CI2 on available Mo sites creates adsorbed Cl and surface MoOx- Not Rate Limiting.2. Dissociative adsorption of H2 on available metallic Mo sites - Not Rate limiting.3. Reaction of adsorbed H* with surface MoOx- Unlikely Rate Limiting.4. Desorption of reaction byproducts (H2O, HCI) creates new metal sites for H2, M0O2CI2 adsorption - Likely rate limiting.

[0126] From kinetic theory, the overall precursor mass flux rate to the substrate ( / can be expressed as:"where £ the precursor sticking coefficient, is also closely related to the step coverage of thin film deposition (< / >,= mass fraction, p = density, c, = mean thermal speed). The precursor mass flux (jt) can also be expressed in terms of a chemical rate expression"Attorney Docket No. LAM1P127WO-12314-1WO

[0127] is a rate constant, At;are dose times, and Ea\s an overall reaction activation energy [adapted from 3]. By combining equations (2) and (3), the sticking coefficient governing precursor deposition, £f, can be solved for and used to model ALD-Mo step coverage.Transport modeling

[0128] The geometry of NAND Flash wordlines is extremely complicated (see, Figure 3 and Figure 8), but with appropriate simplifications process trends are extracted from transient wordline vapor transport models. Figure 8 shows NAND Flash wordlines in a cell array. Precursor molecules diffuse laterally into the wordline array from vertical slits on either side of the array at the start of a precursor dose. Slit-to-slit spacing is 2L with x being the distance from the center of the word line.

[0129] The diffusion equation (4) is used to model transient precursor mass fractions inside a wordline (tMo) during ALD:

[0130] The effective diffusivity coefficient De^ in these equations is the free-molecular diffusivity of a precursor molecule in a porous media comprised of a regular array of cylinders in crossflow. With appropriate scaling and boundary conditions, the solution of equation (4) yields useful estimates of transient precursor concentration profiles inside NAND Flash memory arrays, as illustrated in Figure 8.

[0131] Wordline saturation time increases as ALD-Mo deposition progresses towards wordline seam closure as shown in Fig. 9A. However, by adjusting deposition conditions from the beginning of the process, when the wordlines are wide open, to seam closure at the end, complete saturation with reasonable dose times is maintained. Figure 9B is flow diagram showing operations in a method of filling wordline features. At operation 902, a 3D NAND structure having features to be filled is provided. It may be provided to a semiconductor processing apparatus capable of ALD. In an operation 904, molybdenum is formed directly on dielectric surfaces of the wordlines features. Then, at an operation 906, the wordline features are filled by ALD. ALD conditions are adjusted to maintain saturation. In some embodiments, the ALD conditions to maintain saturation can be determined based on the kinetic and transport modeling described above, with the space remaining to fill becoming narrower as molybdenum is deposited in theAttorney Docket No. LAM1P127WO-12314-1WO wordline features. In some embodiments, the ALD conditions to maintain saturation can be determined experimentally. In some embodiments, experiments informed by modeling may be used to determine the ALD conditions.

[0132] For example, Mo dose times or dose amounts may be increased as the wordlines fill to maintain saturation. In some such embodiments, the Mo precursor exposure is increased, with exposure during a cycle measured in (Mo precursor partial pressure) x (dose time). In another example, exposure of H2 or other co-reactant is increased during the ALD process. Exposure is also measured in H2 (partial pressure) x (dose time).

[0133] According to various embodiments, the exposure may be increased each cycle or every N cycles, with N remaining the same or varying throughout the ALD process. In an example, the Mo exposure may be increased every 5 cycles or every 10 cycles. In another example, the Mo exposure may be increased every 5 cycles for multiple increases, then by every 10 cycles. Similarly, H2 exposure may be increased every N cycles. Figure 9C shows an example of a method of filling features. First, a first set of ALD cycles is performed at conditions to achieve complete saturation in an operation 922. Then, conditions (e.g., increase Mo precursor and H2 exposure) are changed in an operation in operation 924. In some embodiments, for example, this change reflects that diffusion into the wordline feature is increasingly difficult as the wordline fills. A subsequent set of ALD cycles is performed at the new conditions in an operation 926. Operations 924 and 926 are then repeated one or more times to fill the features in an operation 928. According to various examples, 300 to 1000 cycles may be performed to fill wordline features, with conditions changed every 5 to 50 cycles. The methods described herein can be used to adjust deposition conditions as the wordlines fill for structures with more and more wordline pairs, as is shown in Figure 9d.

[0134] In some embodiments, adjusting ALD conditions involves adjusting purge conditions. Byproduct desorption is the rate limiting process during purge, so chamber pressure has limited impact on purge time.

[0135] As one can see from in-feature modeling, vapor transport changes dramatically during the ALD-Mo fill process. ALD systems disclosed herein and described below enable recipe optimization at the beginning, middle, and end of the process. Certain embodiments disclosed below use a 4-station chamber architecture, which allows adjusting wafer temperature, gasAttorney Docket No. LAM1P127WO-12314-1WO timing, and gas flows from station to station. This process adaptability is especially important as Flash memory scales beyond 300 wordlines.Apparatus

[0136] Figure 10A depicts a schematic illustration of an embodiment of an ALD process station 1000 having a process chamber 1002 for maintaining a low-pressure environment. In some embodiments, a plurality of ALD process stations may be included in a common low-pressure process tool environment. For example, Figures 11A and 11B (described below) depict embodiments of a system 1100. In some embodiments, one or more hardware parameters of ALD process station 1000, including those discussed in detail below, may be adjusted programmatically by one or more computer controllers 1050. In some other embodiments, a process chamber may be a single station chamber.

[0137] ALD process station 1000 fluidly communicates with reactant delivery system 1001a for delivering process gases to a distribution showerhead 1006. Reactant delivery system 1001a includes a mixing vessel 1004 for blending and / or conditioning process gases, such as a Mo precursor-containing gas, a hydrogen-containing gas, an argon or other carrier gas, or other reactant-containing gas, for delivery to showerhead 1006. One or more mixing vessel inlet valves 1020 may control introduction of process gases to mixing vessel 1004. In various embodiments, deposition of an initial Mo layer is performed in process station 1000 and in some embodiments, other operations such as a pre-treatment may be performed in the same or another station of the system 1100 as further described below with respect to Figure 11A.

[0138] As an example, the embodiment of Figure 10A includes a vaporization point 1003 for vaporizing liquid reactant to be supplied to the mixing vessel 1004. In some embodiments, vaporization point 1003 may be a heated vaporizer. In some embodiments, a liquid precursor or liquid reactant may be vaporized at a liquid injector (not shown). For example, a liquid injector may inject pulsesofa liquid reactant into a carrier gas stream upstream of the mixing vessel 1004. In one embodiment, a liquid injector may vaporize the reactant by flashing the liquid from a higher pressure to a lower pressure. In another example, a liquid injector may atomize the liquid into dispersed microdroplets that are subsequently vaporized in a heated delivery pipe. Smaller droplets may vaporize faster than larger droplets, reducing a delay between liquid injection and complete vaporization. Faster vaporization may reduce the length of piping downstream fromAttorney Docket No. LAM1P127WO-12314-1WO vaporization point 1003. In one scenario, a liquid injector may be mounted directly to mixing vessel 1004. In another scenario, a liquid injector may be mounted directly to showerhead 1006.

[0139] Reactant delivery system 1001a may also include one or more solid precursor delivery components including one or more on-board ampoules 1013 and / or bulk delivery components 1015.

[0140] In some embodiments, a liquid flow controller (LFC) upstream of vaporization point 1003 may be provided for controlling a mass flow of liquid for vaporization and delivery to process chamber 1002. For example, the LFC may include a thermal mass flow meter (MFM) located downstream of the LFC. A plunger valve of the LFC may then be adjusted responsive to feedback control signals provided by a proportional-integral-derivative (PID) controller in electrical communication with the MFM. However, it may take one second or more to stabilize liquid flow using feedback control. This may extend a time for dosing a liquid reactant. Thus, in some embodiments, the LFC may be dynamically switched between a feedback control mode and a direct control mode. In some embodiments, this may be performed by disabling a sense tube of the LFC and the PID controller.

[0141] Showerhead 1006 distributes process gases toward substrate 1012. In the embodiment shown in Figure 10A, the substrate 1012 is located beneath showerhead 1006 and is shown resting on a pedestal 1008. Showerhead 1006 may have any suitable shape and may have any suitable number and arrangement of ports for distributing process gases to substrate 1012.

[0142] In some embodiments, pedestal 1008 may be raised or lowered to expose substrate 1012 to a volume between the substrate 1012 and the showerhead 1006. In some embodiments, pedestal 1008 may be temperature controlled via heater 1010. Pedestal 1008 may be set to any suitable temperature, such as between about 200°C and about 800°C during operations for performing various disclosed embodiments. It will be appreciated that, in some embodiments, pedestal height may be adjusted programmatically by a suitable computer controller 1050. At the conclusion of a process phase, pedestal 1008 may be lowered during another substrate transfer phase to allow removal of substrate 1012 from pedestal 1008.

[0143] In some embodiments, a position of showerhead 1006 may be adjusted relative to pedestal 1008 to vary a volume between the substrate 1012 and the showerhead 1006. Further, it will be appreciated that a vertical position of pedestal 1008 and / or showerhead 1006 may beAttorney Docket No. LAM1P127WO-12314-1WO varied by any suitable mechanism within the scope of the present disclosure. In some embodiments, pedestal 1008 may include a rotational axis for rotating an orientation of substrate 1012. It will be appreciated that, in some embodiments, one or more of these example adjustments may be performed programmatically by one or more suitable computer controllers 1050.

[0144] In some embodiments where plasma may be used as discussed above, showerhead 1006 and pedestal 1008 electrically communicate with a radio frequency (RF) power supply 1014 and matching network 1016 for powering a plasma. In some embodiments, the plasma energy may be controlled by controlling one or more of a process station pressure, a gas concentration, an RF source power, an RF source frequency, and a plasma power pulse timing. For example, RF power supply 1014 and matching network 1016 may be operated at any suitable powertoform a plasma having a desired composition of ionic and / or radical species. Likewise, RF power supply 1014 may provide RF power of any suitable frequency. In some embodiments, RF power supply 1014 may be configured to control high- and low-frequency RF power sources independently of one another. Example low-frequency RF frequencies may include, but are not limited to, frequencies between 0 kHz and 900 kHz. Example high-frequency RF frequencies may include, but are not limited to, frequencies between 1.8 MHz and 2.45 GHz, or greater than about 13.56 MHz, or greater than 27 MHz, or greater than 80 MHz, or greater than 60 MHz. It will be appreciated that any suitable parameters may be modulated discretely or continuously to provide plasma energy for the surface reactions. In some embodiments, the showerhead is powered and the pedestal is grounded. In some embodiments, the pedestal is powered.

[0145] In some embodiments, the plasma may be monitored in-situ by one or more plasma monitors. In one scenario, plasma power may be monitored by one or more voltage, current sensors (e.g., VI probes). In another scenario, plasma density and / or process gas concentration may be measured by one or more optical emission spectroscopy sensors (OES). In some embodiments, one or more plasma parameters may be programmatically adjusted based on measurements from such in-situ plasma monitors. For example, an OES sensor may be used in a feedback loop for providing programmatic control of plasma power. It will be appreciated that, in some embodiments, other monitors may be used to monitor the plasma and other process characteristics. Such monitors may include, but are not limited to, infrared (IR) monitors, acoustic monitors, and pressure transducers.Attorney Docket No. LAM1P127WO-12314-1WO

[0146] In some embodiments, instructions for a controller 1050 may be provided via input / output control (IOC) sequencing instructions. In one example, the instructions for setting conditions for a process phase may be included in a corresponding recipe phase of a process recipe. In some cases, process recipe phases may be sequentially arranged, so that all instructions for a process phase are executed concurrently with that process phase. In some embodiments, instructions for setting one or more reactor parameters may be included in a recipe phase. For example, a first recipe phase may include instructions for setting a flow rate of e.g., H2 and time delay instructions for the first recipe phase. A second, subsequent recipe phase may include instructions for modulating or stopping a flow rate of the gas, and instructions for modulating a flow rate of a carrier or purge gas and time delay instructions forthe second recipe phase. A third recipe phase may include instructions for modulating a flow rate of a reactant gas (e.g., MOO2CI2) and time delay instructions for the third recipe phase. A fourth, subsequent recipe phase may include instructions for modulating or stopping a flow rate of an inert and / or a reactant gas, and instructions for modulating a flow rate of a carrier or purge gas and time delay instructions for the fourth recipe phase. It will be appreciated that these recipe phases may be further subdivided and / or iterated in any suitable way within the scope of the present disclosure.

[0147] Further, in some embodiments, pressure control for process station 1000 may be provided by butterfly valve 1018. As shown in the embodiment of Figure 10A, butterfly valve 1018 throttles a vacuum provided by a downstream vacuum pump (not shown). However, in some embodiments, pressure control of process station 1000 may also be adjusted by varying a flow rate of one or more gases introduced to the process station 1000.

[0148] Figures 11A and 11B show examples of processing systems. Figure 11A shows an example of a processing system including multiple chambers. The system 1100 includes a transfer module 1103. The transfer module 1103 provides a clean, vacuum environment to minimize risk of contamination of substrates being processed as they are moved between various modules. Mounted on the transfer module 1103 is a multi-station chamber 1109 capable of performing ALD processes described above.

[0149] Also mounted on the transfer module 1103 are one or more single station modules 1107a and 1107b. In some embodiments, an optional preclean is performed in a module 1107a with formation of a boron-based layer formed in a module 1107b. Bulk deposition of molybdenum may be performed in chamber 1109. In some embodiments, multiple stations 1111, 1113, 1115,Attorney Docket No. LAM1P127WO-12314-1WO and 1117 that may sequentially perform operations in accordance with disclosed embodiments. For example, chamber 1109 may be configured such that station 1111 perform ALD of a first bulk layer using a molybdenum oxyhalide precursor and H2. Stations 1113, 1115, and 1117 may be configured to perform 1 / 3 deposition of the remaining Mo. In another example, deposition of the first Mo bulk layer is performed in module 1107b with deposition of the second layer performed in the multi-station chamber 1109. The multi-station chamber 1109 may also be operated in static mode such that the same deposition occurs in parallel on different substrates. In another example, operations 303-307 of Figure 3 are all performed in multi-station chamber 1109. In some embodiments, a single station may perform both high H2:Mo ratio and low H2:Mo ALD. For example, station 1113 may be configured to perform an initial set of ALD cycles using a higher flow of H2, followed by a second set of ALD cycles at a lower H2 flow rate. ALD of the Mo bulk layer at the lower rate may be continued in station 1115 and / or 1117.

[0150] Two or more stations may be included in a multi-station chamber, e.g., 2-6, with the operations appropriately distributed. For example, a two-station chamber may be configured to perform ALD of an initial Mo bulk layer in a first station followed by ALD of a second layer of bulk metal in a second station. Stations may include a heated pedestal or substrate support, one or more gas inlets or showerhead or dispersion plate.

[0151] Stations may include a heated pedestal or substrate support, one or more gas inlets or showerhead or dispersion plate.

[0152] The system 1100 also includes one or more wafer source modules 1101, where wafers are stored before and after processing. An atmospheric robot (not shown) in the atmospheric transfer chamber 1119 may first remove wafers from the source modules 1101 to loadlocks 1121. A wafer transfer device (generally a robot arm unit) in the transfer module 1103 moves the wafers from loadlocks 1121 to and among the modules mounted on the transfer module 1103.

[0153] Chamber 1109 may have one or more of the following features:Individually addressable plasma power generators associated with each station;Individually addressable reactant inputs associated with each station;Multi-plenum showerheads on each station;Dual solid precursor delivery systems.Attorney Docket No. LAM1P127WO-12314-1WO

[0154] Solid precursor delivery systems may include bulk delivery systems and / or on-board ampoules.

[0155] Figure 11B is an embodiment of a system 1100. The system 1100 in Figure 11B has wafer source modules 1101, a transfer module 1103, atmospheric transfer chamber 1119, and loadlocks 1121, as described above with reference to Figure 11A. The system in Figure 11B has three single station modules 1157a-1175c. The system 1100 may be configured to sequentially perform operations in accordance with the disclosed embodiments. For example, the single station modules 1157a-1157c may be configured so that a first module 1157a does a pretreatment and deposits a first bulk layer, a second module 1357b deposits a second bulk layer, and module 1357c performs an etch. In accordance with the embodiments disclosed above, any of the modules in Figures 11A and 11B may be configured to perform etch.

[0156] Stations may include a heated pedestal or substrate support, one or more gas inlets or showerhead or dispersion plate as described above with reference to Figure 10A. Reactants and / or purge gases may be delivered to a station via a showerhead from charge vessels as shown in Figure 10B.

[0157] Returning to Figure 11A and 11B, in various embodiments, a system controller 1129 is employed to control process conditions during deposition. The controller 1129 will typically include one or more memory devices and one or more processors. A processor may include a CPU or computer, analog and / or digital input / output connections, stepper motor controller boards, etc. Such a system controller may be employed in control of any of the processes and apparatus described herein.

[0158] The controller 1129 may control all the activities of the apparatus. The system controller 1129 executes system control software, including 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 on memory devices associated with the controller 1129 may be employed in some embodiments.

[0159] Typically, there will be a user interface associated with the controller 1129. The user interface may include a display screen, graphical software displays of the apparatus and / orAttorney Docket No. LAM1P127WO-12314-1WO process conditions, and user input devices such as pointing devices, keyboards, touch screens, microphones, etc.

[0160] System control logic may be configured in any suitable way. In general, the logic can be designed or configured in hardware and / or software. The instructions for controlling the drive circuitry may be hard coded or provided as software. The instructions may be provided by "programming." Such programming is understood to include logic of any form, including hard coded logic in digital signal processors, application-specific integrated circuits, and other devices which have specific algorithms implemented as hardware. Programming is also understood to include software or firmware instructions that may be executed on a general-purpose processor. System control software may be coded in any suitable computer readable programming language.

[0161] The computer program code for controlling the Mo precursor pulses, hydrogen pulses, and argon flow, and other processes in a process sequence can be written in any conventional computer readable programming language: for example, assembly language, C, C++, Pascal, Fortran, or others. Compiled object code or script is executed by the processor to perform the tasks identified in the program. Also as indicated, the program code may be hard coded.

[0162] The controller parameters relate to process conditions, such as, for example, process gas composition and flow rates, 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 may be entered utilizing the user interface.

[0163] Signals for monitoring the process may be provided by analog and / or digital input connections of the system controller 1129. The signals for controlling the process are output on the analog and digital output connections of the deposition apparatus.

[0164] The system software may be designed or configured in many ways. For example, various chamber component subroutines or control objects may be written to control operation of the chamber components necessary to carry out the deposition processes in accordance with the disclosed embodiments. Examples of programs or sections of programs for this purpose include substrate positioning code, process gas control code, pressure control code, and heater control code.

[0165] In some implementations, a controller 1129 is part of a system, which may be part of the above-described examples. Such systems can include semiconductor processing equipment,Attorney Docket No. LAM1P127WO-12314-1WO including a processing tool or tools, chamber or chambers, a platform or platforms for processing, and / or specific processing components (a wafer pedestal, a gas flow system, etc.). These systems may be integrated with electronics for controlling their operation before, during, and after processing of a semiconductor wafer or substrate. The electronics may be referred to as the "controller," which may control various components or subparts of the system or systems. The controller 1129, depending on the processing requirements and / or the type of system, may be programmed to control any of the processes disclosed herein, includingthe 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, fluid delivery settings, positional and operation settings, wafer transfers into and out of a tool and other transfer tools and / or load locks connected to or interfaced with a specific system.

[0166] Broadly speaking, the controller may be defined as electronics having various integrated circuits, logic, memory, and / or software that receive instructions, issue instructions, control operation, enable cleaning operations, enable endpoint measurements, and the like. The 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), defining operational parameters for carrying out a particular process on or for a semiconductor wafer or to a system. The operational parameters may, in some embodiments, be part of a recipe defined by process engineers to accomplish one or more processing steps during the fabrication of one or more layers, materials, metals, oxides, silicon, silicon dioxide, surfaces, circuits, and / or dies of a wafer.

[0167] The controller 1129, in some implementations, may be a part of or coupled to a computer that is integrated with, coupled to the system, otherwise networked to the system, or a combination thereof. For example, the controller 1129 may be in the "cloud" or all or a part of a fab host computer system, which can allow for remote access of the wafer processing. The computer may enable remote access to the system to monitor current progress of fabrication operations, examine a history of past fabrication operations, examine trends or performance metrics from a plurality of fabrication operations, to change parameters of current processing, to set processing steps to follow a current processing, or to start a new process. In some examples,Attorney Docket No. LAM1P127WO-12314-1WO a remote computer (e.g. a server) can provide process recipes to a system over a network, which may include a local network or the Internet. The remote computer may include a user interface that enables entry or programming of parameters and / or setings, which are then communicated to the system from the remote computer. In some examples, the controller receives instructions in the form of data, which specify parameters for each of the processing steps to be performed during one or more operations. The parameters may be specific to the type of process to be performed and the type of tool that the controller is configured to interface with or control. Thus, as described above, the controller may be distributed, such as by including one or more discrete controllers that are networked together and working towards a common purpose, such as the processes and controls described herein. An example of a distributed controllerfor such purposes would be one or more integrated circuits on a chamber in communication with one or more integrated circuits located remotely (such as at the platform level or as part of a remote computer) that combine to control a process on the chamber.

[0168] Without limitation, example systems may include 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 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 systems that may be associated or used in the fabrication and / or manufacturing of semiconductor wafers.

[0169] As noted above, depending on the process step or steps to be performed by the tool, the controller might communicate with one or more of other tool circuits or modules, other tool components, cluster tools, other tool interfaces, adjacent tools, neighboring tools, tools located throughout a factory, a main computer, another controller, or tools used in material transport that bring containers of wafers to and from tool locations and / or load ports in a semiconductor manufacturing factory.

[0170] The controller 1129 may include various programs. A substrate positioning program may include program code for controlling chamber components that are used to load the substrate onto a pedestal or chuck and to control the spacing between the substrate and other parts of the chamber such as a gas inlet. A substrate tilt and rotation program may include code for tilt and rotation. A process gas control program may include code for controlling gas composition, flowAttorney Docket No. LAM1P127WO-12314-1WO rates, pulse times, and optionally for flowing gas into the chamber prior to deposition in orderto stabilize the pressure in the chamber. A pressure control program may include code for controlling the pressure in the chamber by regulating, e.g., 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 that is used to heat the substrate. Alternatively, the heater control program may control delivery of a heat transfer gas such as helium to the wafer chuck.

[0171] Examples of chamber sensors that may be monitored during deposition include mass flow controllers, pressure sensors such as manometers, and thermocouples located in the pedestal or chuck. Appropriately programmed feedback and control algorithms may be used with data from these sensors to maintain desired process conditions.

[0172] The foregoing describes implementation of disclosed embodiments in a single or multichamber semiconductor processing tool. The apparatus and process described herein may be used in conjunction with lithographic patterning tools or processes, for example, for the fabrication or manufacture of semiconductor devices, displays, LEDs, photovoltaic panels, and the like. Typically, though not necessarily, such tools / processes will be used or conducted together in a common fabrication facility. Lithographic patterning of a film typically includes some or all of the following steps, each step provided with a number of possible tools: (1) application of photoresist on a workpiece, i.e., substrate, using a spin-on or spray-on tool; (2) curing of photoresist using a hot plate or furnace or UV curing tool; (3) exposing the photoresist to visible or UV or x-ray light with a tool such as a wafer stepper; (4) developing the resist so as to selectively remove resist and thereby pattern it using a tool such as a wet bench; (5) transferring the resist pattern into an underlying film or workpiece by using a dry or plasma-assisted etching tool; and (6) removing the resist using a tool such as an RF or microwave plasma resist stripper.

Claims

Attorney Docket No. LAM1P127WO-12314-1WO CLAIMSWhat is claimed is:

1. A method comprising:providing a 3D NAND structure having wordlines feature to be filled with molybdenum (Mo) to a semiconductor processing tool, wherein each wordline feature is having a first opening and a second opening, the first opening and the second opening being at opposite ends of the feature and wherein the first opening opens to a first vertical slit of the 3D NAND structure and the second opening opensto a second vertical slit and wherein the feature is flu id ica lly accessible via the first and second vertical slip¬ forming molybdenum directly on dielectric surfaces of the wordline features; filling the features with molybdenum by an ALD process using a molybdenum reactant; andadjusting deposition conditions throughout the ALD process to maintain complete saturation of the molybdenum reactant.

2. The method of claim 1, wherein adjusting the deposition conditions during the ALD process comprises increasing molybdenum reactant exposure, wherein molybdenum reactant exposure is a product of the partial pressure of the molybdenum reactant and a dose time of the molybdenum reactant.

3. The method of claim 2, wherein the molybdenum reactant exposure is increased every N cycles, wherein N is a number ranging from 3 to 50.

4. The method of claim 2, wherein adjusting the deposition conditions during the ALD process further comprises increasing post-molybdenum reactant purge gas exposure, wherein reactant purge gas exposure reactant exposure is scaled relative to the product of the total purge gas flow rate and the purge time.

5. The method of claim 2, wherein adjusting the deposition conditions during the molybdenum reactant exposure, wherein molybdenum reactant exposure is a product of the partial pressure of the molybdenum reactant and a dose time of the molybdenum reactant.Attorney Docket No. LAM1P127WO-12314-1WO6. The method of claim 1, further comprising estimating transient molybdenum reactant mass fraction inside a wordline feature (LOMO) during an ALD process.

7. The method of claim 5, further comprising determining a molybdenum exposure during an ALD process to maintain saturation of the molybdenum reactant based on precursor mass fraction WMO-8. The method of claim 1, wherein the ALD process comprises reacting a co-reactant with the molybdenum reactant.

9. The method of claim 8, adjusting the deposition conditions during the ALD process comprises increasing co-reactant exposure, wherein co-reactant exposure is a product of the partial pressure of the co-reactant and a dose time of the co-reactant.

Images