Plasma assisted atomic layer deposition titanium oxide for patterning applications

Abstract

The embodiments herein relate to methods and apparatus for depositing an encapsulation layer over memory stacks in MRAM and PCRAM applications. The encapsulation layer is a titanium dioxide (TiO2) layer deposited through an atomic layer deposition reaction. In some embodiments, the encapsulation layer may be deposited as a bilayer, with an electrically favorable layer formed atop a protective layer. In certain implementations, gaps between neighboring memory stacks may be filled with titanium oxide, for example through an atomic layer deposition reaction or a chemical vapor deposition reaction.

Description

BACKGROUND

Memory applications have become increasingly important with the rise and spread of computing devices. A memory device is a storage device that retains information or data that can be accessed and output as needed. There are several different kinds of memory devices. Dynamic random-access memory (DRAM) is a type of random-access memory that stores each bit of data in a separate capacitor within an integrated circuit. The capacitor can be either charged or discharged, representing the two possible values of a bit (conventionally 0 and 1). DRAM has been used for a relatively long time. More recent developments in memory applications include magnetoresistive random-access memory (MRAM) and phase change random-access memory (PCRAM), for example.

Typically, a memory device will include a number of individual memory cells that encode information. The individual memory cells are separated from one another so that they can be functionally isolated. This isolation ensures that the memory cells do not interfere with one another such that the cells can function as desired. In certain cases, the memory cells are lined with an encapsulation layer. The gaps between adjacent lined memory cells may be filled with a gap filling dielectric material. The encapsulation layer and filled gaps serve to isolate adjacent memory cells.

SUMMARY

Certain embodiments herein relate to methods and apparatus for depositing an encapsulation layer over memory stacks in memory applications such as MRAM and PCRAM. The encapsulation layer is a titanium dioxide (TiO₂) layer deposited through an atomic layer deposition reaction. In certain implementations, the encapsulation layer may be deposited as a bilayer, with an electrically favorable layer formed atop a protective layer. In some embodiments, gaps between neighboring memory stacks may be filled with titanium oxide, for example through an atomic layer deposition reaction or a chemical vapor deposition reaction.

In one aspect of the embodiments herein, a method is provided of depositing a titanium dioxide encapsulation layer over memory stacks on a semiconductor substrate may be characterized by the following operations: (a) receiving a substrate having a plurality of magnetoresistive random access memory (MRAM) stacks or phase change random access memory (PCRAM) stacks thereon in a reaction chamber; (b) flowing a titanium-containing reactant into the reaction chamber and allowing the titanium-containing reactant to adsorb onto the surface of the substrate; (c) flowing an oxygen-containing reactant into the reaction chamber and allowing the oxygen-containing reactant to adsorb onto the surface of the substrate; and (d) exposing the reaction chamber to plasma to drive a surface reaction between the titanium-containing reactant and the oxygen-containing reactant to thereby conformally deposit a titanium dioxide encapsulation layer over the plurality of MRAM stacks or PCRAM stacks. This sequence of operations may be repeated for multiple iterations or cycles, as in an atomic layer deposition process.

In certain embodiments, the oxygen-containing reactant is or includes a weak oxidant, such as an oxidant that is not elemental oxygen (O₃ or O₂). In some cases, the oxygen-containing reactant is substantially free of strong oxidants.

In some implementations, the method is repeated to form a second encapsulation layer on the encapsulation layer. The two encapsulation layers may be deposited under different deposition conditions and/or have different properties. During deposition of the second encapsulation layer, the oxygen-containing reactant may be selected from the group consisting of oxygen, nitrous oxide, carbon monoxide, nitric oxide, nitrogen, nitrogen dioxide, oxygen, ozone, sulfur oxide, sulfur dioxide, oxygen-containing hydrocarbons, water, hydrogen peroxide, alkyl alcohols, and combinations of the foregoing.

The process conditions used to form the encapsulation layer may be varied as appropriate for the applications. In some cases, an RF power used to generate the plasma is maintained below about 125 W/deposition station, and the plasma is exposed to the substrate for a duration of about 0.5 seconds or less during each iteration (e.g., a cycle of an atomic layer deposition process). As a further example, during deposition of the second encapsulation layer, an RF power used to generate the plasma is maintained above about 500 W/station, and the plasma is exposed to the substrate for a duration of about 0.5 seconds or greater during each iteration. In implementations employing a second encapsulation layer, during deposition of the second encapsulation layer, an RF power used to generate the plasma may be maintained above about 500 W/station, and the plasma may be exposed to the substrate for a duration of about 0.5 seconds or greater during each iteration. In certain embodiments, the method is performed at a temperature of about 250° C. or less.

In some implementations, the method includes an additional operation after depositing the titanium dioxide encapsulation layer, which operation involves depositing dielectric material in gaps between adjacent MRAM or PCRAM stacks. In some cases, depositing the encapsulation layer and depositing material in gaps are performed in the same reaction chamber without removing the substrate. The gaps may be filled with titanium dioxide through the same reaction and reaction conditions used to deposit the encapsulation layer, or the gaps may be filled with titanium dioxide through different reaction conditions than used to deposit the encapsulation layer. In the second alternative, the gaps may be filled at a higher deposition rate than used to deposit the encapsulation layer.

In certain embodiments, the encapsulation layer has a thickness between about 20-500 Å. In certain embodiments, the encapsulation layer is a hermetic seal. In certain embodiments, the encapsulation layer does not substantially or selectively oxidize any of the materials in the MRAM or PCRAM stacks. In certain embodiments, the encapsulation layer does not substantially desorb hydrogen up to a temperature of at least about 500° C.

Another aspect of the disclosure concerns apparatus for depositing material on partially fabricated semiconductor substrates. Such apparatus may be characterized by the following features: (a) a reaction chamber; (b) an inlet to the reaction chamber for introducing gas phase reactants; (c) an outlet to the reaction chamber for removing material from the reaction chamber; (d) a plasma generator configured to expose the reaction chamber to plasma; and (e) a controller configured to perform operations including the following: (i) receiving a partially fabricated semiconductor substrate having a plurality of magnetoresistive random access memory (MRAM) stacks or phase change random access memory (PCRAM) stacks thereon in the reaction chamber; (ii) flowing a first titanium-containing reactant into the reaction chamber and allow the first titanium-containing reactant to adsorb onto the surface of the substrate; (iii) flowing a first oxygen-containing reactant into the reaction chamber and allow the first oxygen-containing reactant to adsorb onto the surface of the substrate; and (iv) generating a first plasma with the plasma generator, and expose the reaction chamber to the first plasma to drive a surface reaction between the titanium-containing reactant and the oxygen-containing reactant to thereby conformally deposit a first titanium dioxide encapsulation layer over the plurality of MRAM stacks or PCRAM stacks.

In certain embodiments, the controller is further configured to (v) flow a second titanium-containing reactant into the reaction chamber and allow the second titanium-containing reactant to adsorb onto the surface of the substrate; (vi) flow an second oxygen-containing reactant into the reaction chamber and allow the second oxygen-containing reactant to adsorb onto the surface of the substrate; and (vii) generate a second plasma with the plasma generator, and expose the reaction chamber to the second plasma to drive a surface reaction between the second titanium-containing reactant and the second oxygen-containing reactant to thereby conformally deposit a second titanium dioxide encapsulation layer over the first titanium dioxide encapsulation layer. In some implementations, the first titanium dioxide encapsulation layer is deposited under different conditions than the second encapsulation layer. In certain embodiments, the first oxygen-containing reactant comprises a weak oxidant, and the second oxygen-containing reactant comprises O₂ and/or N₂O.

The process conditions configured in a controller recipe may be varied as appropriate for the applications. For example, the controller may be configured to generate the first plasma at an RF power of about 125 W/station or less, and expose the first plasma to the reaction chamber for a duration of about 0.5 seconds or less per plasma exposure, and the controller may be further configured to generate the second plasma at an RF power of about 500 W/station or more, and expose the reaction chamber to the second plasma for a duration of about 0.5 seconds or more per plasma exposure.

These and other features will be described below with reference to the associated drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIGS. 1, 2A and 2B illustrate a basic magnetic tunnel junction structure.

FIG. 3 presents a more complicated version of a magnetic tunnel junction structure.

FIG. 4 shows an example of a phase change memory stack.

FIG. 5 is a flowchart describing a method of forming an encapsulation layer on a memory stack.

FIG. 6 illustrates a bilayer encapsulation layer.

FIG. 7 illustrates a bilayer encapsulation layer over several memory stacks.

FIG. 8 shows a simplified cross-sectional view of a substrate having memory stacks lined with an encapsulation layer, where the gaps between the memory stacks are filled with a gap filling material.

FIG. 9 presents an example of a reaction chamber that may be used to practice the disclosed embodiments.

FIG. 10 presents a multi-tool apparatus that may be used to practice the disclosed embodiments.

FIG. 11 shows a substrate that has a layer of titanium oxide deposited conformally thereon.

FIG. 12 is a table describing film thickness at different locations on the substrate shown in FIG. 11.

FIG. 13 is a graph depicting the stress drift of a tensile TEOS film, with and without titanium oxide encapsulation layers thereon.

FIG. 14 presents two graphs illustrating film stability and gas desorption for a titanium oxide film and for a silicon nitride film.

FIG. 15 shows a substrate that has a layer of titanium oxide lining and filling gaps between adjacent features.

DETAILED DESCRIPTION

In this application, the terms “semiconductor wafer,” “wafer,” “substrate,” “wafer substrate,” and “partially fabricated integrated circuit” are used interchangeably. One of ordinary skill in the art would understand that the term “partially fabricated integrated circuit” can refer to a silicon wafer during any of many stages of integrated circuit fabrication thereon. A wafer or substrate used in the semiconductor device industry typically has a diameter of 200 mm, or 300 mm, or 450 mm. However, the invention is not so limited. The work piece may be of various shapes, sizes, and materials. In addition to semiconductor wafers, other work pieces that may take advantage of this invention include various articles such as printed circuit boards and the like.

The flow rates and power levels provided herein are appropriate for processing on 300 mm substrates. One of ordinary skill in the art would appreciate that these flows and powers may be adjusted as necessary for substrates of other sizes. These parameters scale linearly based on substrate area. For example, an RF power of 500 W/station for a 300 mm wafer would correspond to an RF power of about 1125 W/station for a 450 mm wafer. Further, a flow rate of 1 mL/min for a 300 mm wafer would correspond to a flow rate of about 2.25 mL/min for a 450 mm wafer.

In the following description, numerous specific details are set forth in order to provide a thorough understanding of the presented embodiments. The disclosed embodiments may be practiced without some or all of these specific details. In other instances, well-known process operations have not been described in detail 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.

A memory chip includes many individual memory cells. The memory cells store bits of information (e.g., in the form of 1s and 0s) in such a way that it can be accessed as needed. In order to store large amounts of information, the memory cells are packed closely together on an integrated circuit. In order for the memory cells to function adequately at high densities, they must be isolated from one another. Otherwise, the memory cells could interfere with one another and prevent the device from working as desired.

An encapsulation layer may be used to help isolate the memory cells from one another. The encapsulation layer may be a relatively thin liner layer that conformally coats the memory cells. The encapsulation layer may be made from a dielectric material. In many cases, a silicon-based material (e.g., silicon oxide, silicon nitride or silicon carbide) is used for an encapsulation layer. After the memory cells are coated with the encapsulation layer, gaps may remain between adjacent encapsulated memory cells. These gaps may be filled with gap filling material. In various cases, the gap-filling material is a dielectric material that further serves to isolate the memory cells, and also provides mechanical support to the device. Common gap filling materials include silicon-based materials such as silicon oxide, silicon nitride and silicon carbide, for example.

Encapsulation layers must meet certain requirements to be commercially useful. In many cases, it is necessary that the encapsulation layer be deposited at a relatively low temperature, typically less than about 250° C. However, the encapsulation layer should be stable even at higher temperatures, e.g., up to about 500° C., in order to withstand thermal anneals during manufacture. Further, the encapsulation layer should exhibit high hermeticity. In other words, the encapsulation layer should be a good moisture barrier. Another desirable characteristic is that the encapsulation layer should not desorb hydrogen. Silicon-based materials have proven adequate in meeting these requirements. However, titanium oxide deposited by atomic layer deposition (ALD) may provide these same characteristics, and provides a new technique for implementing encapsulation layers and even gap filling material. The titanium oxide exhibits excellent conformality (e.g., 100% conformal), is hermetic at room temperature down to about 30 Å thickness, is stable up to at least about 500° C., and does not desorb hydrogen.

One application for the presently disclosed embodiments is in the context of magnetoresistive random-access memory (MRAM). MRAM devices store data by means of magnetic storage elements. MRAM is beneficial because it allows for dense data storage, requires relatively low power, and is non-volatile (i.e., the data is retained even if the device loses power).

MRAM devices utilize a structure known as a magnetic tunnel junction (MTJ). An MTJ includes two ferromagnetic layers separated by a thin insulating layer, sometimes referred to as a tunnel barrier layer. Current can flow from one ferromagnetic layer to another through the insulating layer by tunneling. One of the ferromagnetic layers has a magnetic field that is fixed. This layer is referred to as the fixed layer or reference layer. The other ferromagnetic layer has a magnetic field that can be changed by driving a current. This ferromagnetic layer is referred to as the free layer. The magnetic fields of the fixed and free layers can be in the same direction or in opposite directions, depending on the direction of the free layer field. When the two fields are aligned/parallel, the memory cell is in a low resistance state, representing a “0.” When the two fields are in opposite directions/antiparallel, the memory cell is in a high resistance state, representing a “1.”

FIG. 1 shows an example of a magnetic tunnel junction 100. A first ferromagnetic layer 101 has a variable magnetic field, and is therefore considered to be the free layer 101. A second ferromagnetic layer 105 has a fixed magnetic field, and is therefore considered to be the fixed layer. The free layer 101 and fixed layer 105 are separated by an insulating barrier layer 103, which may be made from AlO or MgO, for example. Top and bottom electrodes (not shown) may be attached to the free and fixed layers 101 and 105, respectively.

FIG. 2A illustrates the MTJ of FIG. 1 in a high resistance anti-parallel state. The arrows in FIG. 2A denote the direction of the magnetic field. Because the two ferromagnetic layers 101 and 105 have fields that are in opposite directions, the anti-parallel high resistance state is present. The memory cell indicates a “1” in this state. FIG. 2B illustrates the MTJ of FIG. 1 in a low resistance parallel state. Like in FIG. 2A, the arrows represent the direction of the magnetic field of the magnetic layers 101 and 105. Here, because both fields are in the same direction, the memory cell is in the low resistance parallel state, which indicates a “0.”

The ferromagnetic layers may be provided as single layers or as stacks of individual sub-layers. FIG. 3 presents one example of a more complex form of a magnetic tunnel junction 300. The MTJ 300 includes a free ferromagnetic layer 304 formed of nickel-iron (NiFe). This free ferromagnetic layer 304 acts as a sensing layer, and is formed over a seed layer 306. In this example, the seed layer 306 is formed of tantalum. The seed layer is formed directly on the substrate 305. The free ferromagnetic layer 304 is separated from a pinned ferromagnetic layer 308 by a tunnel barrier layer 310. In this example, the tunnel barrier layer 310 is formed of aluminum oxide (AlOx), and the pinned ferromagnetic layer 308 is formed of cobalt iron (CoFe). The pinned ferromagnetic layer 308 forms part of a pinned structure 312, which may be provided in the form of a synthetic antiferromagnet (SAF). SAFs and other MRAM structures are further discussed in U.S. Pat. No. 7,072,209 and U.S. Pat. No. 8,535,954, each of which is herein incorporated by reference in its entirety.

The orientation of the magnetic field of the pinned ferromagnetic layer 308 is fixed. The orientation of this layer is maintained in a fixed state by a second ferromagnetic layer 314, which in this example may be made of cobalt-iron or nickel-iron. The second ferromagnetic layer 314 is separated from the pinned ferromagnetic layer 308 by a non-magnetic spacer layer 316. In this example, the non-magnetic spacer layer 316 is formed from a layer of ruthenium. The non-magnetic spacer layer 316 produces anti-ferromagnetic coupling between the two ferromagnetic layers 308 and 314 in the pinned structure 312. Further, an antiferromagnetic pinning layer 318, in this example formed of iridium-manganese (IrMn), may be provided above the second ferromagnetic layer 314. The antiferromagnetic pinning layer 318 operates to orient the magnetic moment of the second ferromagnetic layer 314. A pinning field is generated to fix the magnetic moment of the pinned ferromagnetic layer 308. The MTJ may also include a capping layer 320.

Although MRAM devices can have different structures, a basic MRAM device generally includes at least one magnetic layer that acts as a fixed/reference layer, at least one magnetic layer that acts as a free/sensing layer, and at least one layer in between the magnetic layers to act as a tunnel barrier.

Another application for the disclosed embodiments is in the context of phase change random-access memory (PCRAM). PCRAM exploits the unique behavior of chalcogenide glass, a material that can transition between an amorphous and a crystalline state. FIG. 4 shows a cross-sectional view of a phase change memory cell. The cell includes a conductive layer 410, a lower electrode 430, a phase change material 460, a second electrode 440, and a heater element 450. Lower electrode 430 is in a first dielectric layer 420, and heater element 450 is in a second dielectric layer 470. The cell may be formed on a semiconductor substrate.

Phase change material 460 is typically a chalcogenide glass material, though other materials that are capable of changing from a crystalline to an amorphous state or vice versa upon application or withdrawal of heat may also be used. Because the electrical resistivity of the amorphous state of the chalcogenide material is significantly higher than the crystalline state, the phase change material can be used to store data. The amorphous, highly resistive state is used to represent one bit value (0 or 1) with the crystalline, low resistive state used to represent the other bit value.

Conductive layer 410 may be an address line, e.g., a column line, row line, bit line or word line, used to program or read information stored using the phase change material. The conductive layer may be connected to external addressing circuitry.

The electrodes are made of electrically conductive materials. In the example shown in FIG. 4, the lower electrode 430 is a plug, e.g., a tungsten plug. The first dielectric layer 420 may be silicon oxide, silicon nitride or any other type of electrically insulating material. Second dielectric material 470 is electrically and thermally insulating may be made of the same or different material as dielectric layer 420. In certain embodiments, the dielectric materials 420 and/or 470 may be titanium oxide.

Heater element 450 is a material that generates heat in response to the electrical current passing through the adjacent electrode, lower electrode 430. The resistivity of heater element 450 is high relative to that of electrode 430 to generate heat from the current. In operation a current or voltage pulse is applied to the heater element to generate heat. The heat generated by the heater element heats the phase-change material to a temperature above its crystallization temperature and below its melting temperature to achieve a crystalline state, which typically represents a bit value of 1. To change the phase change material back to its amorphous state, the heater element heats the material above its melting point. The phase change material is then quenched to achieve an amorphous state, which typically represents a bit value of 0. Different possible configurations are possible for PCRAM devices, however, a PCRAM device will typically include a phase change material and a heating material between two conductive electrodes. PCRAM structures are further discussed in U.S. Pat. No. 8,450,772, and U.S. Pat. No. 8,243,506, each of which is incorporated herein by reference in its entirety.

In MRAM and PCRAM applications, each memory cell may be used to encode a bit of information (e.g., a “0” or “1”). The memory cells are arranged in rows and columns, and must be isolated from one another to prevent interference between the cells. One method for isolating adjacent memory cells is to line them in an encapsulation layer. The encapsulation layer helps prevent interference between adjacent cells, and also protects the cells from damage from the environment, for example by preventing moisture from penetrating the memory cells.

There are various standards that encapsulation layers should meet in order to be feasible for certain commercial applications. In many memory applications, the memory stacks (i.e., the layers of material that form a memory cell) are very sensitive to oxidation. Where oxidation of the memory stack occurs, the stack materials may degrade and lose electrical and/or magnetic integrity. As such, the encapsulation layer should be deposited in a way that prevents oxidation of the memory stack materials during deposition. Also, to the relatively small extent that oxidation occurs during deposition of the encapsulation layer, oxidation should not occur in a non-uniform manner throughout the different layers of the memory stack. In other words, a deposition process that results in oxidation of certain memory stack materials but not others is particularly undesirable. Further, the encapsulation layer should provide adequate protection against future oxidation of the underlying memory stack materials after deposition.

One criteria that affects the extent of oxidation of the memory stack materials is the temperature at which the encapsulation layer is deposited. Often, the encapsulation layer is deposited through a vapor-based technique such as atomic layer deposition, which is further explained below. For certain memory applications, the encapsulation layer should be deposited at a temperature of less than about 250° C. At higher temperatures, the materials in the memory stack may start to oxidize, thereby degrading the quality and performance of the memory stack.

Another criteria that affects the extent of oxidation of the memory stack materials is the hermeticity of the encapsulation layer. Hermeticity refers to the ability of a material to block moisture transport. An encapsulation layer should be sufficiently hermetic to prevent moisture from penetrating into the memory stack. If moisture were to penetrate the encapsulation layer, it would likely result in oxidation of the underlying memory stack materials, thereby degrading the memory cell. Hermeticity of an overlying layer may be measured by the decrease in stress of a an underlying tensile TEOS film caused by exposure to the atmosphere. Without a hermetic barrier layer, the stress of a tensile TEOS film will decrease dramatically during exposure to the atmosphere, as shown below in FIG. 13, which is further described in the Experimental section. For example, the stress of a tensile 5 kÅ TEOS film decreases from 10⁹ dynes/cm² to −10⁹ dynes/cm² after approximately one week of exposure to the atmosphere—a decrease of over 100%. Capping the tensile TEOS films with a hermetic film results in a stress change in an underlying tensile TEOS film of less than 10% after one week of exposure time.

The encapsulation layer should also be stable up to temperatures of about 500° C. Otherwise, the encapsulation layer may degrade during subsequent processing, for example during thermal anneals. On a related note, the encapsulation layer should not desorb hydrogen. Otherwise, the hydrogen can penetrate the memory stack materials and cause them to degrade. In the context of MRAM, for example, desorption of hydrogen can affect the resistive switching behavior of the free ferromagnetic layer. In many cases, it is desired that an encapsulation layer is deposited conformally to achieve good step coverage. For the purposes of this description, “step coverage” is defined as a ratio of two thicknesses, i.e., the thickness of the material inside the feature divided by the thickness of the material near the opening. For purposes of this document, the term “inside the feature” represents a middle portion of the feature located about the middle point of the feature along the feature's axis, e.g., an area between about 25% and 75% of the distance or, in certain embodiments, between about 40% and 60% of the distance along the feature's depth measured from the feature's opening, or an end portion of the feature located between about 75% and 95% of the distance along the feature's axis as measured from the opening. The term “near the opening of the feature” or “near the feature's opening” represents a top portion of the feature located within 25% or, more specifically, within 10% of the opening's edge or other element representative of the opening's edge. Step coverage of over 100% can be achieved, for example, by filling a feature wider in the middle or near the bottom of the feature than at the feature opening.

Encapsulation layers are frequently silicon-based materials deposited through chemical vapor deposition (CVD) or atomic layer deposition (ALD) techniques. ALD processes provide superior step coverage but are relatively slow. CVD techniques are faster, but provide poor step coverage. Pulsed plasma CVD techniques have been used to improve step coverage compared to more conventional CVD methods. However, pulsed plasma CVD methods result in low hermeticity films that diffuse hydrogen into the memory stack.

CVD and plasma enhanced chemical vapor deposition (PECVD) are sometimes the method of choice for depositing encapsulation layers. In a CVD reaction, one or more vapor phase reactants are provided to a reaction chamber, where they react in the gas phase before being deposited on the substrate surface. Where PECVD is used, the reaction is driven by exposure to plasma. CVD reactions are characterized by gas phase reactions.

One alternative to CVD is ALD processing and plasma enhanced atomic layer deposition (PEALD) processing. Unless otherwise noted, the term ALD is intended to include PEALD, and the term CVD is intended to include PECVD in the following description. ALD methods involve cyclic, self-limiting adsorption of reactant gases and reactions that occur directly on substrate surfaces. ALD processes can provide thin, conformal dielectric films within high aspect ratio features. ALD methods have been developed for the deposition of silicon oxide and other types of film. Films produced by ALD are very thin (e.g., about one monolayer); therefore, numerous ALD cycles may be repeated to adequately fill a gap feature.

In contrast with a CVD process, where activated gas phase reactions are used to deposit films, ALD processes use surface-mediated deposition reactions to deposit films on a layer-by-layer basis. In one example ALD process, a substrate surface, including a population of surface active sites, is exposed to a gas phase distribution of a first film precursor (P1). Some molecules of P1 may form a condensed phase atop the substrate surface, including chemisorbed species and physisorbed molecules of P1. The reactor is then evacuated to remove gas phase and physisorbed P1 so that only chemisorbed species remain. A second film precursor (P2) is then introduced to the reactor so that some molecules of P2 adsorb to the substrate surface. The reactor may again be evacuated, this time to remove unbound P2. Subsequently, energy provided to the substrate (e.g., thermal or plasma energy) activates surface reactions between adsorbed molecules of P1 and P2, forming a film layer. Finally, the reactor is evacuated to remove reaction by-products and possibly unreacted P1 and P2, ending the ALD cycle. Additional ALD cycles may be included to build film thickness.

Depending on the exposure time of the precursor dosing steps and the sticking coefficients of the precursors, each ALD cycle may deposit a film layer of, in one example, between about 0.5 Å-3 Å thick.

A related deposition technique that presents a feasible alternative to CVD for small feature sizes is conformal film deposition (CFD). Generally, CFD does not rely on complete purges of one or more reactants prior to reaction to form the film. For example, there may be one or more reactants present in the vapor phase when a plasma (or other activation energy) is struck. Accordingly, one or more of the process steps described in an ALD process (e.g., a purge step) may be shortened or eliminated in an example CFD process. In some cases, a reactant may be provided continuously. The continuously provided reactant is often a co-reactant/auxiliary reactant (e.g., an oxygen- or nitrogen-containing reactant, as opposed to a primary reactant such as one that contributes silicon to a final film). It is generally necessary, even in CFD processes, to minimize the amount of non-adsorbed principal reactant in the reaction chamber when the chamber is exposed to plasma. Otherwise, a gas phase CVD reaction may result. Typically, a chamber capable of performing an ALD process is also capable of performing a CFD process, and as used herein, a CFD process is considered to be a type of ALD process.

Encapsulation layers have typically been formed from silicon-based materials. However, it has been found that titanium oxide deposited through PEALD techniques can meet all of the standards set forth above for commercially feasible encapsulation layers. In particular, the titanium oxide films show excellent step coverage (e.g., about 100% step coverage), are stable up to at least about 500° C., and do not desorb hydrogen. Further, oxidation damage to the underlying magnetic stack materials during deposition can be minimized in a number of ways.

FIG. 5 presents a flowchart for a method of depositing a titanium oxide encapsulation layer through a PEALD process. The method 500 begins at operation 501, where a first reactant flows into the reaction chamber and is allowed to adsorb onto the surface of the substrate. In one example, the first reactant includes tetrakis dimethylamido titanium (TDMATi, [(CH₃)₂N]₄Ti). Other potential first reactants include Bis(tert-butylcyclopentadienyl)titanium(IV) dichloride (C₁₈H₂₆Cl₂Ti), Tetrakis(diethylamido)titanium ([(C₂H₅)₂N]₄Ti), Tetrakis(ethylmethylamido)titanium ([(CH₃C₂H₅)N]₄Ti), Titanium diisopropoxidebis(2,2,6,6-tetramethyl-3,5-heptanedionate) (Ti[OCC(CH₃)₃CHCOC(CH₃)₃]₂(OC₃H₇)2), Titanium(IV) isopropoxide (Ti[OCH(CH₃)₂]₄), and Titanium tetrachloride (TiCl₄), for example, though the embodiments are not so limited.

The first reactant may flow into the reaction chamber for a duration between about 0.1-5 seconds, for example about 1 second. The first reactant may flow at a rate between about 0.5-3 mL/min, for example between about 0.5-1 mL/min, as measured before the reactant is vaporized. The temperature in the reaction chamber may be between about 50-300° C. In various cases, the reaction temperature is below about 250° C. The pressure in the reaction chamber may be between about 1-10 Torr, for example between about 2-4 Torr.

The method continues at operation 503, where the reaction chamber is purged. The purge may include evacuating the reaction chamber and/or sweeping the reaction chamber with a gas that does not include the species to be purged. In this example, a suitable sweep gas is one that does not include the first reactant. The purge may have a duration between about 0.1-25 seconds. Where a sweep gas is used, a flow rate of the sweep gas during the purge may be between about 5-60 slm. The purpose of the purge is to remove unadsorbed first reactant from the reaction chamber.

Next, at operation 505 a second reactant is flowed into the reaction chamber and allowed to adsorb onto the substrate surface. In one example, the second reactant includes carbon dioxide (CO₂). Alternatively or in addition, the second reactant may include carbon monoxide (CO), nitric oxide (NO), nitrous oxide (N₂O), nitrogen (N₂), nitrogen dioxide (NO₂), oxygen (O₂), ozone (O₃), sulfur oxide (SO), sulfur dioxide (SO₂), oxygen-containing hydrocarbons (C_xH_yO_z), water (H₂O), hydrogen peroxide (H₂O₂), alkyl alcohols (e.g, methanol, ethanol, etc.), mixtures thereof, etc. The second reactant may be provided for a duration between about 0.1-5 seconds, for example about 0.5-1 seconds.

At operation 507, the reaction chamber is exposed to plasma to drive a reaction between the first and second reactants on the surface of the substrate to form a thin layer of titanium oxide. Each reaction cycle may deposit about a monolayer of material. Various types of plasma may be used to drive this reaction including capacitively coupled plasmas and inductively coupled plasmas. Different types of plasma generators may be used including RF, DC, and microwave plasma generators. Moreover, according to various embodiments, the plasma may be direct or remote. Where an RF generator is employed, an RF power used to drive plasma formation may be between about 50-2500 W/station, for example about 625 W/station. There may be a number of stations in a single chamber. In one example, there are four individual stations in a single reaction chamber, though additional or fewer stations may be used. The frequency used to generate the plasma may include a high frequency (HF) and/or a low frequency (LF) component. In certain embodiments, the frequency used to generate plasma may be between about 13.56-40 MHz, for example about 13.56 MHz. The duration of plasma exposure may be between about 0.1-5 seconds, for example about 0.5-1 seconds. In some embodiments, operations 505 and 507 may occur at least partially at the same time such that plasma is generated while the second reactant is being supplied to the reaction chamber. Next, a purge may occur at operation 509. This purge may be implemented as an evacuation and/or sweep of the reaction chamber, and is done to help remove any unadsorbed species and film byproducts. The method may be repeated to form a film having a desired thickness. As such, in block 511, the thickness of the deposited film is compared to a desired film thickness. If the film has reached a sufficient thickness, the deposition ends at block 513. On the other hand, if the deposition has not reached a sufficient thickness, the method may be repeated starting at block 501. In some cases, an encapsulation layer is deposited to a thickness of at least about 20 Å, or at least about 30 Å. The encapsulation layer may have a thickness between about 20-500 Å, for example between about 30-100 Å in some embodiments. Typically, the encapsulation layer will grow at a rate between about 0.6-1.1 Å/cycle.

In certain embodiments, extra care is taken to minimize any oxidation of underlying memory stack materials during deposition of the titanium oxide encapsulation layer. One method for minimizing oxidation of underlying materials is to use a gentle oxidation reactant. In the context of FIG. 5, the oxidation reactant is the second reactant. Examples of gentle/weak oxidation reactants include CO₂, and oxygen-containing hydrocarbons. One example of a reactant that does not qualify as gentle oxidation reactant is elemental oxygen (O₂ or O₃). Another method for minimizing oxidation of underlying memory stack materials during deposition of an encapsulation layer is to use gentle plasma conditions. These gentle plasma conditions are sometimes referred to as “soft landing” conditions. The relative harshness of plasma conditions depends on the power used to generate the plasma (higher powers lead to harsher plasmas) as well as the duration of plasma exposure (longer plasma exposures are harsher). Where a soft plasma landing is desired, the RF power may be maintained below about 125 W/station, for example between about 50-125 W/station, and the plasma exposure may have a duration of about 0.5 seconds or less, for example between about 0.1-0.5 seconds.

Another method for minimizing oxidation damage to the substrate is to deposit the encapsulation layer as a bilayer. A first layer may be a protective layer (sometimes also referred to as a first encapsulation layer) and a second layer may be an electrically favorable layer (sometimes also referred to as a second encapsulation layer). The protective layer may be deposited directly on the memory stack, and may be deposited under gentle oxidation conditions, for example with a gentle oxidant (e.g., CO₂) and/or with a soft plasma landing (e.g., the RF power used to generate the plasma may be about 125 W/station or less, and the plasma may be exposed for a duration of about 0.5 seconds or less during formation of the protective layer). In some embodiments, the electrically favorable layer is formed at a higher temperature than the protective layer. The first layer prevents or minimizes oxidation of the stack materials. However, this protective layer may have relatively poor electrical properties. For example, the protective layer may have low breakdown voltage (BDV) and high leakage current due to insufficient oxidative conversion of the precursors. The protective layer may have a higher carbon content and/or a higher hydrogen content than an electrically favorable layer. These relatively high carbon and/or hydrogen levels may lead to the poor electrical qualities of the protective layer.

To remedy the poor electrical qualities of the protective layer, this layer may be integrated into a bilayer with a more electrically favorable layer. The electrically favorable layer may be deposited directly on the protective layer. Because the protective layer is present to shield the underlying memory stack materials from oxidation, harsher oxidation conditions may be used to deposit the electrically favorable layer without risking damage to the underlying materials. The electrically favorable layer may be deposited using a less gentle oxidation reactant. For instance, the electrically favorable layer may be deposited using O₂, N₂O, or a mixture thereof. The electrically favorable layer may be deposited using relatively harsh plasma conditions. Where this is the case, the RF power used to deposit the electrically favorable layer may be at least about 500 W/station, or at least about 625 W/station. The duration of plasma exposure for a single ALD cycle during formation of the electrically favorable layer may be at least about 0.5 seconds, for example at least about 1 second.

Where an encapsulation layer is deposited as a bilayer, the protective layer should be sufficiently thick to protect the underlying materials from oxidation (i.e., the layer must be thick enough to provide sufficient hermeticity). Further, the protective layer should be thick enough to form a coalescent layer that does not have pinholes or other form of incomplete coverage. In some embodiments, a protective layer has a thickness of at least about 20 Å. The protective layer may have a thickness between about 20-500 Å, for example between about 30-100 Å in some embodiments.

FIG. 6 shows a bilayer encapsulation layer 602 deposited over a memory stack 601 on substrate 600. The memory stack is depicted as a single block, though it is understood that the stack includes several individual layers that operate together. The encapsulation layer 602 includes protective layer 604 and electrically favorable layer 605.

Previously, titanium oxide has not been used as an encapsulation layer for protecting MRAM and PCRAM stacks in part due to the relatively harsh plasma conditions used during conventional titanium oxide deposition, and the sensitivity of the stack materials to oxidation. The harsh plasma conditions would result in unacceptably high levels of oxidation of the underyling stack materials. More gentle plasma conditions (RF power, plasma exposure time) could be used, but would result in films that are likely to flake off, or that have unacceptably high levels of carbon due to the presence of precursors. However, various methods disclosed herein overcome these problems. In some embodiments, a gentle oxidation reactant is used to prevent oxidation of stack materials. In other embodiments, a bilayer approach is used to prevent oxidation of the stack while ensuring that the encapsulation layer has good electrical properties.

FIG. 7 illustrates a substrate 700 having a plurality of memory stacks 701 deposited thereon. The memory stacks 701 may be MRAM or PCRAM stacks, for example. The stacks 701 are coated with a bilayer of material 702. The bilayer includes a lower protective layer 704 and an upper electrically favorable layer 705. In some embodiments, the spaces between adjacent lined stacks are filled with gap filling material.

After a collection of memory stacks are lined with an encapsulation layer, the encapsulated stacks are separated by gaps, as shown in FIG. 7. These gaps may be filled with a dielectric material through a CVD or ALD process, for example. In certain cases, the gaps are filled with a silicon-based material such as silicon oxide or silicon nitride. In other cases, the gaps may be filled with a non-silicon based material. One example of a non-silicon based material that may be used to fill the gaps is titanium oxide. In some embodiments, both the encapsulation layer and the gap filling material are titanium oxide. In a particular embodiment, the encapsulation layer deposition process is extended in duration such that instead of merely lining the memory stacks, the ALD-deposited titanium oxide fills the entire region between adjacent memory stacks. In general, the encapsulation layer needs to be of high quality to adequately protect underlying materials. In contrast, there is much more flexibility on the quality and properties of a gap filling material.

Where titanium oxide is used to both line the memory stacks and fill the gaps between the stacks, a number of processing methods may be used. In one case, both parts of the process (lining the stacks and filling the gaps between stacks) are performed in a single step, under the same deposition conditions. In this case, there may be substantially no difference between the material lining the stacks and filling the gaps. In another embodiment, both parts of the process are performed under the same conditions, but additional processing steps may be taken between deposition of the encapsulation layer and the gap filling material. One example of a process that may occur in between these two steps is a thermal anneal. During a thermal anneal, the substrate may be raised to an elevated temperature, for example at least about 300° C., or at least about 400° C., or at least about 500° C.

In certain other embodiments, the encapsulation layer and gap filling material may each be deposited through an ALD process, but the deposition conditions may be different between the two processes. For instance, the gap filling material may be deposited at a higher deposition rate, using shorter purges, etc.

In some implementations, the encapsulation layer and gap filling material may be deposited through different kinds of processes. For instance, the encapsulation layer may be deposited through an ALD process, and the gap filling material may be deposited through a CVD process, for example a pulsed plasma CVD process. The encapsulation layer and gap filling material may each be titanium oxide. Other materials may also be used for the gap filling material, as mentioned. For instance, the gap filling material may be SiN or SiC deposited through a pulsed plasma CVD process.

FIG. 8 illustrates a simplified cross-sectional view of a memory device 800 having a number of individual memory stacks 801 covered by an encapsulation layer 802 and positioned on a substrate 805. The gaps between the encapsulated memory stacks 801 are filled with gap filling material 803. Encapsulation layer 802 may be formed from titanium oxide through an ALD process as described herein. The gap filling material 803 may be titanium oxide deposited through an ALD process, titanium oxide deposited through a CVD process, or some other dielectric material. Though gap filling material 803 is shown as having a flat top portion, the shape of film deposited in this region will depend on the deposition conditions and process. The material may need to be planarized before reaching the state shown in FIG. 8.

In certain embodiments, the ALD-deposited TiO₂ may be used as a cap layer in front end of line (FEOL) logic devices. Memory encapsulation layers and FEOL application layers such as high-K (high dielectric constant) layers for MIM capacitors in DRAM, interpoly dielectric in 3D NAND, etc. require similar film properties (protection against oxidation, hermeticity, film stability, etc.). Thus, ALD-deposited TiO₂ is a promising material for this application, as well. Where TiO₂ is used as a cap layer in FEOL logic devices, it will typically involve deposition of the TiO₂ material over an electrically active device/gate. For instance, each gate that covers a metal oxide semiconductor or capacitor may be lined with an encapsulation layer to isolate it from neighboring transistors. In some embodiments, these encapsulation layers may be titanium oxide deposited through an ALD process as described herein.

Apparatus

FIG. 9 schematically shows an embodiment of a reaction chamber process station 900. For simplicity, process station 900 is depicted as a standalone process station having a process chamber body 902 for maintaining a low-pressure environment. However, it will be appreciated that a plurality of process stations 900 may be included in a common process tool environment. For example, FIG. 10 depicts an embodiment of a multi-station processing tool. Further, it will be appreciated that, in some embodiments, one or more hardware parameters of process station 900, including those discussed in detail above, may be adjusted programmatically by one or more computer controllers.

Process station 900 fluidly communicates with reactant delivery system 901 for delivering process gases to a distribution showerhead 906. Reactant delivery system 901 includes a mixing vessel 904 for blending and/or conditioning process gases for delivery to showerhead 906. One or more mixing vessel inlet valves 920 may control introduction of process gases to mixing vessel 904.

Some reactants may be stored in liquid form prior to vaporization at and subsequent delivery to the process station. The embodiment of FIG. 9 includes a vaporization point 903 for vaporizing liquid reactant to be supplied to mixing vessel 904. In some embodiments, vaporization point 903 may be a heated vaporizer. The saturated reactant vapor produced from such vaporizers may condense in downstream delivery piping when adequate controls are not in place. Exposure of incompatible gases to the condensed reactant may create small particles. These small particles may clog piping, impede valve operation, contaminate substrates, etc. Some approaches to addressing these issues involve sweeping and/or evacuating the delivery piping to remove residual reactant. However, sweeping the delivery piping may increase process station cycle time, degrading process station throughput. Thus, in some embodiments, delivery piping downstream of vaporization point 903 may be heat traced. In some examples, mixing vessel 904 may also be heat traced. In one non-limiting example, piping downstream of vaporization point 903 has an increasing temperature profile extending from approximately 100° C. to approximately 150° C. at mixing vessel 904.

As mentioned, in some embodiments the vaporization point 903 is a heated liquid injector. A liquid injector may inject pulses of a liquid reactant into a carrier gas stream upstream of the mixing vessel. In one scenario, a liquid injector may vaporize reactant by flashing the liquid from a higher pressure to a lower pressure. In another scenario, a liquid injector may atomize the liquid into dispersed microdroplets that are subsequently vaporized in a heated delivery pipe. It will be appreciated that smaller droplets may vaporize faster than larger droplets, reducing a delay between liquid injection and complete vaporization. Faster vaporization may reduce a length of piping downstream from vaporization point 903. In one scenario, a liquid injector may be mounted directly to mixing vessel 904. In another scenario, a liquid injector may be mounted directly to showerhead 906.

In some embodiments, a liquid flow controller upstream of vaporization point 903 may be provided for controlling a mass flow of liquid for vaporization and delivery to process station 900. For example, the liquid flow controller (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, the LFC may be dynamically switched from a feedback control mode to a direct control mode by disabling a sense tube of the LFC and the PID controller.

Showerhead 906 distributes process gases toward substrate 912. In the embodiment shown in FIG. 9, substrate 912 is located beneath showerhead 906, and is shown resting on a pedestal 908. It will be appreciated that showerhead 906 may have any suitable shape, and may have any suitable number and arrangement of ports for distributing processes gases to substrate 912.

In some embodiments, a microvolume 907 is located beneath showerhead 906. Performing a CFD process in a microvolume rather than in the entire volume of a process station may reduce reactant exposure and sweep times, may reduce times for altering process conditions (e.g., pressure, temperature, etc.), may limit an exposure of process station robotics to process gases, etc. Example microvolume sizes include, but are not limited to, volumes between 0.1 liter and 2 liters.

In some embodiments, pedestal 908 may be raised or lowered to expose substrate 912 to microvolume 907 and/or to vary a volume of microvolume 907. For example, in a substrate transfer phase, pedestal 908 may be lowered to allow substrate 912 to be loaded onto pedestal 908. During a deposition on substrate process phase, pedestal 908 may be raised to position substrate 912 within microvolume 907. In some embodiments, microvolume 907 may completely enclose substrate 912 as well as a portion of pedestal 908 to create a region of high flow impedance during a deposition process.

Optionally, pedestal 908 may be lowered and/or raised during portions the deposition process to modulate process pressure, reactant concentration, etc., within microvolume 907. In one scenario where process chamber body 902 remains at a base pressure during the process, lowering pedestal 908 may allow microvolume 907 to be evacuated. Example ratios of microvolume to process chamber volume include, but are not limited to, volume ratios between 1:500 and 1:10. It will be appreciated that, in some embodiments, pedestal height may be adjusted programmatically by a suitable computer controller.

In another scenario, adjusting a height of pedestal 908 may allow a plasma density to be varied during plasma activation and/or treatment cycles included in the deposition process. At the conclusion of the deposition process phase, pedestal 908 may be lowered during another substrate transfer phase to allow removal of substrate 912 from pedestal 908.

While the example microvolume variations described herein refer to a height-adjustable pedestal, it will be appreciated that, in some embodiments, a position of showerhead 906 may be adjusted relative to pedestal 908 to vary a volume of microvolume 907. Further, it will be appreciated that a vertical position of pedestal 908 and/or showerhead 906 may be varied by any suitable mechanism within the scope of the present disclosure. In some embodiments, pedestal 908 may include a rotational axis for rotating an orientation of substrate 912. 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.

Returning to the embodiment shown in FIG. 9, showerhead 906 and pedestal 908 electrically communicate with RF power supply 914 and matching network 916 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 914 and matching network 916 may be operated at any suitable power to form a plasma having a desired composition of radical species. Examples of suitable powers are included above. Likewise, RF power supply 914 may provide RF power of any suitable frequency. In some embodiments, RF power supply 914 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 50 kHz and 500 kHz. Example high-frequency RF frequencies may include, but are not limited to, frequencies between 1.8 MHz and 2.45 GHz. It will be appreciated that any suitable parameters may be modulated discretely or continuously to provide plasma energy for the surface reactions. In one non-limiting example, the plasma power may be intermittently pulsed to reduce ion bombardment with the substrate surface relative to continuously powered plasmas.

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.

In some embodiments, the plasma may be controlled via input/output control (IOC) sequencing instructions. In one example, the instructions for setting plasma conditions for a plasma activation phase may be included in a corresponding plasma activation 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 plasma parameters may be included in a recipe phase preceding a plasma process phase. For example, a first recipe phase may include instructions for setting a flow rate of an inert and/or a reactant gas, instructions for setting a plasma generator to a power set point, and time delay instructions for the first recipe phase. A second, subsequent recipe phase may include instructions for enabling the plasma generator and time delay instructions for the second recipe phase. A third recipe phase may include instructions for disabling the plasma generator and time delay instructions for the third 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.

In some deposition processes, plasma strikes last on the order of a few seconds or more in duration. In certain implementations described herein, much shorter plasma strikes may be applied during a processing cycle. These may be on the order of 50 ms to 1 second, with 0.25 seconds being a specific example. Such short RF plasma strikes require quick stabilization of the plasma. To accomplish this, the plasma generator may be configured such that the impedance match is preset to a particular voltage, while the frequency is allowed to float. Conventionally, high-frequency plasmas are generated at an RF frequency at about 13.56 MHz. In various embodiments disclosed herein, the frequency is allowed to float to a value that is different from this standard value. By permitting the frequency to float while fixing the impedance match to a predetermined voltage, the plasma can stabilize much more quickly, a result which may be important when using the very short plasma strikes associated with ALD/CFD cycles.

In some embodiments, pedestal 908 may be temperature controlled via heater 910. Further, in some embodiments, pressure control for process station 900 may be provided by butterfly valve 918. As shown in the embodiment of FIG. 9, butterfly valve 918 throttles a vacuum provided by a downstream vacuum pump (not shown). However, in some embodiments, pressure control of process station 900 may also be adjusted by varying a flow rate of one or more gases introduced to process station 900.

The internal surfaces of the process station 900 may be coated with conditioning layer 950, which may include an undercoat and/or a pre-deposition coat. Examples of surfaces that become coated with the conditioning layer include the chamber walls 902, the chamber ceiling and floor, the pedestal 908 and the showerhead 906. Although FIG. 9 is shown with a substrate 912 in the process station 900, this substrate 912 is not present during deposition of the one or more conditioning layers. Instead, the substrate 912 is introduced to the process station 900 after the conditioning layers are deposited, when the process station 900 is ready to be used for depositing film on the substrate 912.

As described above, one or more process stations may be included in a multi-station processing tool. FIG. 10 shows a schematic view of an embodiment of a multi-station processing tool 1000 with an inbound load lock 1002 and an outbound load lock 1004, either or both of which may comprise a remote plasma source. A robot 1006, at atmospheric pressure, is configured to move wafers from a cassette loaded through a pod 1008 into inbound load lock 1002 via an atmospheric port 1010. A wafer is placed by the robot 1006 on a pedestal 1012 in the inbound load lock 1002, the atmospheric port 1010 is closed, and the load lock is pumped down. Where the inbound load lock 1002 comprises a remote plasma source, the wafer may be exposed to a remote plasma treatment in the load lock prior to being introduced into a processing chamber 1014. Further, the wafer also may be heated in the inbound load lock 1002 as well, for example, to remove moisture and adsorbed gases. Next, a chamber transport port 1016 to processing chamber 1014 is opened, and another robot (not shown) places the wafer into the reactor on a pedestal of a first station shown in the reactor for processing. While the embodiment depicted in FIG. 10 includes load locks, it will be appreciated that, in some embodiments, direct entry of a wafer into a process station may be provided.

The depicted processing chamber 1014 comprises four process stations, numbered from 1 to 4 in the embodiment shown in FIG. 10. Each station has a heated pedestal (shown at 1018 for station 1), and gas line inlets. It will be appreciated that in some embodiments, each process station may have different or multiple purposes. For example, in some embodiments, a process station may be switchable between an ALD mode, a CFD mode, and a CVD process mode. Additionally or alternatively, in some embodiments, processing chamber 1014 may include one or more matched pairs of ALD/CFD/CVD process stations. While the depicted processing chamber 1014 comprises four stations, it will be understood that a processing chamber according to the present disclosure may have any suitable number of stations. For example, in some embodiments, a processing chamber may have five or more stations, while in other embodiments a processing chamber may have three or fewer stations.

FIG. 10 also depicts an embodiment of a wafer handling system 1090 for transferring wafers within processing chamber 1014. In some embodiments, wafer handling system 1090 may transfer wafers between various process stations and/or between a process station and a load lock. It will be appreciated that any suitable wafer handling system may be employed. Non-limiting examples include wafer carousels and wafer handling robots. FIG. 10 also depicts an embodiment of a system controller 1050 employed to control process conditions and hardware states of process tool 1000. System controller 1050 may include one or more memory devices 1056, one or more mass storage devices 1054, and one or more processors 1052. Processor 1052 may include a CPU or computer, analog and/or digital input/output connections, stepper motor controller boards, etc.

In some embodiments, system controller 1050 controls all of the activities of process tool 1000. System controller 1050 executes system control software 1058 stored in mass storage device 1054, loaded into memory device 1056, and executed on processor 1052. System control software 1058 may include instructions for controlling the timing, mixture of gaseous and liquid reactants, chamber and/or station pressure, chamber and/or station temperature, wafer temperature, target power levels, RF power levels, RF exposure time, substrate pedestal, chuck and/or susceptor position, and other parameters of a particular process performed by process tool 1000. These programmed processes may include various types of processes including, but not limited to, processes related to deposition of an undercoat, processes related to deposition of film on substrates, and processes related to cleaning the chamber. System control software 1058 may be configured in any suitable way. For example, various process tool component subroutines or control objects may be written to control operation of the process tool components necessary to carry out various process tool processes. System control software 1058 may be coded in any suitable computer readable programming language.

In some embodiments, system control software 1058 may include input/output control (IOC) sequencing instructions for controlling the various parameters described above. For example, each phase of an encapsulation layer deposition process may include one or more instructions for execution by system controller 1050. The instructions for setting process conditions for an encapsulation layer deposition process phase may be included in a corresponding encapsulation layer deposition recipe phase. In some embodiments, the recipe phases may be sequentially arranged, so that all instructions for a process phase are executed concurrently with that process phase.

Other computer software and/or programs stored on mass storage device 1054 and/or memory device 1056 associated with system controller 1050 may be employed in some embodiments. Examples of programs or sections of programs for this purpose include a substrate positioning program, a process gas control program, a pressure control program, a heater control program, and a plasma control program.

A substrate positioning program may include program code for process tool components that are used to load the substrate onto pedestal 1018 and to control the spacing between the substrate and other parts of process tool 1000. The positioning program may include instructions for appropriately moving substrates in and out of the reaction chamber as necessary to deposit the conditioning layers, deposit films on substrates, and clean the chamber. These may include instructions for ensuring that no substrates are present in the reaction chamber during deposition of the conditioning layers and during the cleaning process.

A process gas control program may include code for controlling gas composition and flow rates and optionally for flowing gas into one or more process stations prior to deposition in order to stabilize the pressure in the process station. A pressure control program may include code for controlling the pressure in the process station by regulating, for example, a throttle valve in the exhaust system of the process station, a gas flow into the process station, etc.

A heater control program may include code for controlling the current to a heating unit that is used to heat the substrate. Alternatively or in addition, the heater control program may control delivery of a heat transfer gas (such as helium) to the substrate. A plasma control program may include code for setting RF power levels, frequencies, and exposure times in one or more process stations in accordance with the embodiments herein.

In some embodiments, there may be a user interface associated with system controller 1050. The user interface may include a display screen, graphical software displays of the apparatus and/or process conditions, and user input devices such as pointing devices, keyboards, touch screens, microphones, etc.

In some embodiments, parameters adjusted by system controller 1050 may relate to process conditions. Non-limiting examples include process gas composition and flow rates, temperature, pressure, plasma conditions (such as RF bias power levels and exposure times), etc. These parameters may be provided to the user in the form of a recipe, which may be entered utilizing the user interface.

Signals for monitoring the process may be provided by analog and/or digital input connections of system controller 1050 from various process tool sensors. The signals for controlling the process may be output on the analog and digital output connections of process tool 1000. Non-limiting examples of process tool sensors that may be monitored include mass flow controllers, pressure sensors (such as manometers), thermocouples, etc. Appropriately programmed feedback and control algorithms may be used with data from these sensors to maintain process conditions.

System controller 1050 may provide program instructions for implementing the above-described deposition processes. The program instructions may control a variety of process parameters, such as DC power level, RF bias power level, pressure, temperature, etc. The instructions may control the parameters to operate in-situ deposition of film stacks according to various embodiments described herein.

The system controller will typically include one or more memory devices and one or more processors configured to execute the instructions so that the apparatus will perform a method in accordance with the present invention. Machine-readable, non-transitory media containing instructions for controlling process operations in accordance with the present invention may be coupled to the system controller.

The various hardware and method embodiments described above 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 comprises some or all of the following steps, each step enabled with a number of possible tools: (1) application of photoresist on a workpiece, e.g., a substrate having a silicon nitride film formed thereon, using a spin-on or spray-on tool; (2) curing of photoresist using a hot plate or furnace or other suitable 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 or a spray developer; (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. In some embodiments, an ashable hard mask layer (such as an amorphous carbon layer) and another suitable hard mask (such as an antireflective layer) may be deposited prior to applying the photoresist.

It is to be understood that the configurations and/or approaches described herein are exemplary in nature, and that these specific embodiments or examples are not to be considered in a limiting sense, because numerous variations are possible. The specific routines or methods described herein may represent one or more of any number of processing strategies. As such, various acts illustrated may be performed in the sequence illustrated, in other sequences, in parallel, or in some cases omitted. Likewise, the order of the above described processes may be changed.

The subject matter of the present disclosure includes all novel and nonobvious combinations and sub-combinations of the various processes, systems and configurations, and other features, functions, acts, and/or properties disclosed herein, as well as any and all equivalents thereof.

EXPERIMENTAL

Experimental results have shown that ALD-deposited TiO₂ is a promising material for use as an encapsulation layer in emerging memory fields such as MRAM and PCRAM. The titanium oxide shows excellent step coverage with no pattern loading (no etching profile differences between film deposited in isolated features and film deposited in dense features), good hermeticity, and good film stability at the relevant temperatures.

FIG. 11 shows SEM images of a substrate having a layer of TiO₂ deposited thereon. The substrate shown here is a test wafer having a base layer of silicon 1101. A layer of silicon oxide 1102 was deposited above the silicon 1101. The silicon oxide layer 1102 was etched to form recessed features. The features were first lined with nitride material 1103 deposited through a low pressure chemical vapor deposition (LPCVD) process. The TiO₂ was deposited on the nitride layer at a temperature of about 150° C. Though three separate panels are shown in FIG. 11, each of the panels shows a different part of the same substrate. In other words, the titanium oxide films in each panel were deposited at the same time through the same process on the same substrate. The top panel shows relatively high aspect ratio (AR) features that are packed relatively densely. The middle panel shows lower aspect ratio features that are packed less densely. The bottom panel shows a field region. In a typical deposition process during manufacture, a substrate may have a variety of features with different aspect ratios and different feature densities. One important goal during deposition is to deposit material uniformly on the substrate, both within a feature (for good step coverage), and between different features. Where a substrate includes both high aspect ratio dense features and low aspect ratio isolated features, it can be challenging to deposit in both types of features uniformly. As shown in FIG. 11, however, the ALD-deposited titanium oxide was deposited very uniformly in both types of features, as well as in the field region.

FIG. 12 presents a table describing the film properties of the titanium oxide film of FIG. 11. The film thickness was measured at the top, sidewall and bottom portions of each feature. The step coverage is calculated by dividing the sidewall thickness by the top thickness (S/T), and is expressed in a percent. The High AR row of FIG. 12 corresponds to the high aspect ratio feature shown in the top panel of FIG. 11. The Low AR row of FIG. 12 corresponds to the lower aspect ratio feature shown in the middle panel of FIG. 11. The Field row of FIG. 12 corresponds to the field region shown in the bottom panel of FIG. 11. The film thickness was substantially uniform throughout the entire film. The high aspect ratio feature showed a step coverage of about 99.9%, and the low aspect ratio feature showed a step coverage of about 99.1%. Further, there was substantially no pattern loading. In other words, the film thickness was very uniform for all feature types. In certain embodiments, the step coverage of a titanium oxide film is between about 98-102%, for example between about 99-101%.

FIG. 13 is a graph showing the stress drift of different films deposited over a tensile TEOS-based silicon oxide film. The films were deposited with N₂O-only plasma at a temperature of about 150° C. RF power was about 2.5 kW, and pressure was about 3.5 Torr. As described above, when a bare tensile TEOS film is exposed to atmosphere, it begins to absorb moisture. The stress of the tensile TEOS film changes in a predictable manner as the moisture enters the film. This stress drift is shown by the Control (t-TEOS) line of FIG. 13. However, this stress drift does not occur (or occurs to a much lesser extent) where the tensile TEOS film has a sufficiently effective hermetic cap to prevent the absorption of moisture. Four different thicknesses of TiO₂ film were tested, including thicknesses between about 30-175 Å. All of the TiO₂ films proved to be sufficiently hermetic to prevent a stress drift in the tensile-TEOS film. This means that even at thicknesses down to about 30 Å, the TiO₂ provides a good hermetic seal to protect underlying layers. In the context of an encapsulation layer for capping memory stacks, this hermeticity is beneficial because moisture will be prevented from entering the memory stacks and oxidizing or otherwise degrading the stack materials.

FIG. 14 presents two graphs illustrating the stability of certain films at different temperatures. The films were formed using N2O-only plasma, an RF power of about 2.5 kW, at a pressure of about 3.5 Torr. The data were generated by thermal dispersive spectrometry (TDS). The left panel of FIG. 14 relates to a TiO₂ film deposited through an ALD process as described herein. The right panel of FIG. 14 relates to a SiN film deposited through a pulsed plasma CVD process. The right panel is provided for the sake of comparison. The ALD-deposited TiO₂ shows very good film stability up through at least about 500° C. Substantially no H₂, N₂, O₂ or N₂O are being desorbed from the film up through this temperature. In contrast, the SiN film shows substantial desorption of H₂ starting around 360° C. While the TiO2 film shows some level of hydrogen even at low temperatures, this hydrogen level corresponds to the baseline level from the chamber ambient. This relatively low level of hydrogen was seen across many samples, and does not signify any change in the film being tested. This same baseline hydrogen level is seen at low temperatures in the SiN film, for example. These results suggest that titanium oxide deposited through ALD as described herein is stable at the relevant operating temperatures. In some embodiments, the encapsulation layer does not substantially desorb hydrogen at temperatures up to about 500° C. As used herein, “does not substantially desorb hydrogen” means that at the relevant temperatures, the level of hydrogen observed through TDS does not rise more than about 25% above the ambient level.

FIG. 15 shows an example where TiO₂ was deposited through an ALD process in order to fill gaps between features. In this example, the TiO₂ both lines the features (like an encapsulation layer) and also fills the spaces between the features (like a gap filling material). The TiO2 deposited in a very conformal manner. The film is slightly thicker in the field region, but this extra thickness is present because the film continued to deposit conformally after the gaps were filled. The extra thickness in the field region does not represent an overburden as would be seen with CVD deposited films, for example. These results suggest that a titanium oxide encapsulation layer may be extended in thickness in order to fill the gaps between adjacent features/memory stacks. Although seams can be observed where two sidewalls meet, the seams are fairly small, and no voids are present in the filled gaps. This technique may be especially useful where the gaps to be filled have smaller aspect ratios, for example an aspect ratio of about 1:2.5 or less (depth:width). The presence of seams is a less serious problem in these small aspect ratio gaps as compared to larger aspect ratio gaps.

Claims

1. A method of depositing a titanium dioxide bilayer over a plurality of structures on a substrate, the method comprising: (a) receiving the substrate in a reaction chamber, the substrate being a semiconductor substrate and comprising the plurality of structures on a surface of the substrate;(b) flowing a first titanium-containing reactant into the reaction chamber and allowing the first titanium-containing reactant to adsorb onto the surface of the substrate;(c) flowing a first oxygen-containing reactant into the reaction chamber and allowing the first oxygen-containing reactant to adsorb onto the surface of the substrate;(d) exposing the reaction chamber to plasma for a duration of about 0.5 seconds or less to drive a surface reaction between the first titanium-containing reactant and the first oxygen-containing reactant to thereby conformally deposit a first titanium dioxide layer over the plurality of structures, wherein the plasma is generated using an RF power density of about 1750 W/m2 or less;(e) repeating (b)-(d) until the first titanium dioxide layer reaches a final thickness;(f) after (e), flowing a second titanium-containing reactant into the reaction chamber and allowing the second titanium-containing reactant to adsorb onto the surface of the substrate;(g) flowing a second oxygen-containing reactant into the reaction chamber and allowing the second oxygen-containing reactant to adsorb onto the surface of the substrate;(h) exposing the reaction chamber to a second plasma to drive a second surface reaction between the second titanium-containing reactant and the second oxygen-containing reactant to thereby conformally deposit a second titanium dioxide layer over the first titanium dioxide layer, wherein the second titanium dioxide layer is deposited under different plasma conditions than the first titanium dioxide layer; and(i) repeating (f)-(h) until the second titanium dioxide layer reaches a second final thickness, wherein the first titanium dioxide layer and the second titanium dioxide layer together form the titanium dioxide bilayer.

2. The method of claim 1, wherein the first oxygen-containing reactant comprises a weak oxidant.

3. The method of claim 2, wherein the first oxygen-containing reactant does not include oxygen (O2).

4. The method of claim 2, wherein during deposition of the first titanium dioxide layer, the first oxygen-containing reactant is substantially free of strong oxidants.

5. The method of claim 2, wherein during deposition of the second titanium dioxide layer, the second oxygen-containing reactant is selected from the group consisting of oxygen, nitrous oxide, carbon monoxide, carbon dioxide, nitric oxide, nitrogen dioxide, ozone, sulfur oxide, sulfur dioxide, oxygen-containing hydrocarbons, water, hydrogen peroxide, alkyl alcohols, and combinations of the foregoing.

6. The method of claim 1, wherein during (h), an RF power density used to generate the second plasma is maintained above about 7075 W/m2, and wherein during each iteration of (h), the second plasma is exposed to the substrate for a duration of about 0.5 seconds or greater.

7. The method of claim 2, wherein during (h), an RF power density used to generate the second plasma is maintained above about 7075 W/m2, and wherein during each iteration of (h), the second plasma is exposed to the substrate for a duration of about 0.5 seconds or greater.

8. The method of claim 1, wherein the final thickness of the first titanium dioxide layer is between about 20-500 Å.

9. The method of claim 1, wherein the first titanium dioxide layer hermetically seals the substrate after the first titanium dioxide layer is deposited in (d)-(e).

10. The method of claim 1, wherein the method is performed at a temperature between about 50-300° C.

11. The method of claim 1, wherein depositing the first titanium dioxide layer does not substantially or selectively oxidize any of the materials in the structures on the substrate.

12. The method of claim 1, wherein the first titanium dioxide layer does not substantially desorb hydrogen up to a temperature of at least about 500° C.

13. An apparatus for depositing a titanium dioxide bilayer on partially fabricated semiconductor substrates, comprising: a reaction chamber;an inlet to the reaction chamber for introducing gas phase reactants;an outlet to the reaction chamber for removing material from the reaction chamber;a plasma generator configured to expose the reaction chamber to plasma; anda controller configured to: receive a partially fabricated semiconductor substrate having a plurality of structures thereon in the reaction chamber;flow a first titanium-containing reactant into the reaction chamber and allow the first titanium-containing reactant to adsorb onto the surface of the partially fabricated semiconductor substrate;flow a first oxygen-containing reactant into the reaction chamber and allow the first oxygen-containing reactant to adsorb onto the surface of the partially fabricated semiconductor substrate;generate a first plasma with the plasma generator, and expose the reaction chamber to the first plasma for a duration of about 0.5 seconds or less to drive a surface reaction between the first titanium-containing reactant and the first oxygen-containing reactant to thereby conformally deposit a first titanium dioxide layer over the plurality of structures on the partially fabricated semiconductor substrate, wherein the first plasma is generated using an RF power density of about 1750 W/m2 or less;flow a second titanium-containing reactant into the reaction chamber and allow the second titanium-containing reactant to adsorb onto the surface of the partially fabricated semiconductor substrate;flow an second oxygen-containing reactant into the reaction chamber and allow the second oxygen-containing reactant to adsorb onto the surface of the partially fabricated semiconductor substrate; andgenerate a second plasma with the plasma generator, and expose the reaction chamber to the second plasma to drive a surface reaction between the second titanium-containing reactant and the second oxygen-containing reactant to thereby conformally deposit a second titanium dioxide layer over the first titanium dioxide layer,wherein the first titanium dioxide layer is deposited under different conditions than the second titanium dioxide layer, and wherein the first titanium dioxide layer and the second titanium dioxide layer together form the titanium oxide bilayer.

14. The apparatus of claim 13, wherein the first oxygen-containing reactant comprises a weak oxidant, and wherein the second oxygen-containing reactant comprises O2 and/or N2O.

15. The apparatus of claim 13, wherein the controller is configured to generate the second plasma at an RF power density of about 7075 W/m2 or more, and to expose the reaction chamber to the second plasma for a duration of about 0.5 seconds or more per plasma exposure.

16. A method of forming a titanium dioxide bilayer over a plurality of structures on a substrate, comprising: (a) receiving the substrate in a reaction chamber, the substrate comprising the plurality of structures on a surface of the substrate;(b) depositing a first titanium dioxide layer on the plurality of structures using a first plasma-assisted atomic layer deposition reaction;(c) after (b), depositing a second titanium dioxide layer on the first titanium dioxide layer using a second plasma-assisted atomic layer deposition reaction, wherein the second plasma-assisted atomic layer deposition reaction is performed under different plasma conditions than the first plasma-assisted atomic layer deposition reaction, and wherein the first titanium dioxide layer and the second titanium dioxide layer together form the titanium dioxide bilayer.

17. The method of claim 16, wherein the first plasma-assisted atomic layer deposition reaction comprises exposing the substrate to a first plasma, wherein the first plasma is generated at an RF power density of about 1750 W/m2 or less.

18. The method of claim 17, wherein the second plasma-assisted atomic layer deposition reaction comprises exposing the substrate to a second plasma, wherein the second plasma is generated at an RF power density of about 7075 W/m2 or greater.

19. The method of claim 18, wherein the substrate is exposed to the first plasma for durations of about 0.5 seconds or less, and wherein the substrate is exposed to the second plasma for durations of about 0.5 seconds or more.

20. The method of claim 16, wherein the first titanium dioxide layer is deposited using a first set of reactants, wherein the second titanium dioxide layer is deposited using a second set of reactants, and wherein the first set of reactants differs from the second set of reactants.