US20180265977A1 - Deposition system with vacuum pre-loaded deposition head - Google Patents

Abstract

A thin film deposition system includes a vacuum-preloaded gas bearing deposition head positioned in an external environment having an ambient pressure, the deposition head having an output face including a plurality of source openings through which gaseous materials are supplied and one or more exhaust openings. An exhaust pressure at the exhaust openings is less than ambient pressure, and a source pressure at the source openings is greater than that at the exhaust openings, with the pressure at the outermost source openings being greater than ambient pressure. A motion control system moves a substrate unit over the output face in the in-track direction without constraining its motion in a direction normal to the output face to a point where a center of gravity of the substrate unit is beyond the first edge of the output face.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS
• Reference is made to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002126), entitled “Modular thin film deposition system,” by Spath et al.; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002128), entitled “Dual gas bearing substrate positioning system,” by Spath; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002129), entitled “Deposition system with moveable-position web guides,” by Spath et al.; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002130), entitled “Deposition system with repeating motion profile,” by Spath et al.; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002131), entitled “Deposition system with modular deposition heads,” by Spath et al.; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002132), entitled “Porous gas-bearing backer,” by Spath; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002133), entitled “Deposition system with interlocking deposition heads,” by Tutt et al.; to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002154), entitled “Vertical system with vacuum pre-loaded deposition head,” by Spath et al.; and to commonly assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002157), entitled “Heated gas-bearing backer,” by Spath, each of which is incorporated herein by reference.
FIELD OF THE INVENTION
• This invention generally relates to the deposition of thin-film materials, and more particularly to a deposition system including a vacuum-preloaded deposition head.
BACKGROUND OF THE INVENTION
• There is a growing interest in depositing thin-film materials from gaseous precursors on a wide range of substrates for a wide variety of applications. Substrates of interest include both rigid substrates, such as flat-panel glass, and flexible substrates, such as plastic webs or metal foils. Flexible supports are of particular interest since they can be more mechanically robust, lighter weight, and allow for more economic manufacturing (e.g., by enabling roll-to-roll processing) than rigid substrates. Thin-film deposition systems, similar to their liquid coating counterparts, are advantaged if the deposition head, or gas delivery device, is smaller in area than the area of the substrate to be coated. For substrates that are continuous, such as webs and foils, the use of a deposition head that is smaller than the area of the substrate is a requirement not just an advantage.
• Among the techniques widely used for thin-film deposition is chemical vapor deposition (CVD), which uses chemically reactive molecules that react to deposit a desired film on a substrate. Molecular precursors useful for CVD applications comprise elemental (atomic) constituents of the film to be deposited and typically also include additional elements. CVD precursors are volatile molecules that are delivered, in a gaseous phase, to a chamber in order to react at the substrate, forming the thin film thereon. The chemical reaction deposits a thin film with a desired film thickness.
• Atomic layer deposition (ALD) is a thin-film deposition technology that provides excellent thickness control of conformal thin-films. The ALD process segments the thin-film deposition process of conventional CVD into single atomic-layer deposition steps. Advantageously, ALD steps are self-terminating and can deposit one atomic layer when conducted up to or beyond self-termination exposure times. An atomic layer typically ranges from about 0.1 to about 0.5 molecular monolayers, with typical dimensions on the order of no more than a few angstroms. In ALD, deposition of an atomic layer is the outcome of a chemical reaction between a reactive molecular precursor and the substrate. In each separate ALD reaction-deposition step, the net reaction deposits the desired atomic layer and substantially eliminates “extra” atoms originally included in the molecular precursor. In its most pure form, ALD involves the adsorption and reaction of each of the precursors in the absence of the other precursor or precursors of the reaction. In temporal vacuum ALD, thin-film growth is accomplished by alternating the delivery of two or more reactive materials, or precursors, into a vacuum chamber in time. Sequentially, a first precursor is applied to react with the substrate, the excess of the first precursor is removed, and a second precursor is then applied to react with the substrate surface. The excess of the second precursor is then removed and the process is repeated. In all ALD processes, the substrate is exposed sequentially to a series of reactants that react with the substrate and are kept isolated from each other to avoid CVD or gas phase reactions. An ALD cycle is defined by the steps required to form a single layer of the overall thin-film material; for a process using two precursors a cycle is defined as the first precursor exposure, a purge step, the second precursor exposure, and a second precursor purge step.
• A version of ALD processes known as spatial atomic layer deposition (SALD) employs a continuous (as opposed to pulsed) gaseous material distribution from a deposition head. As distributed from the deposition head, the gaseous precursors are separated in space by the flow of an inert gas, rather than being separated in time. While vacuum chambers can be used with SALD, they are no longer necessary due to the physical separation of the gas flows rather than a temporal separation of the precursors within a single chamber. In SALD systems, the required sequential exposures are accomplished by relative movement between the substrate and the delivery head such that any given point on the substrate sees the necessary sequence of gaseous materials. This relative movement can be accomplished by moving a substrate relative to a fixed delivery head, moving a delivery head with respect to a fixed substrate, or moving both the delivery head and the substrate in order to achieve the desired gas exposure at the substrate. Exemplary SALD processes, are described in commonly-assigned U.S. Pat. No. 7,413,982, U.S. Pat. No. 7,456,429, U.S. Pat. No. 7,789,961, and U.S. Patent Application Publication 2009/0130858, the disclosures of which are incorporated herein by reference. SALD enables operation at atmospheric or near-atmospheric pressures and is capable of operating in an unsealed or open-air environment, making it compatible with web coating.
• SALD offers considerable promise as a technique for thin film deposition on a range of substrates. However, in spite of its inherent technical capabilities and advantages, a number of technical hurdles still remain. As in all ALD processes, the thickness of the SALD deposited thin-film is controlled by the number of ALD cycles to which the substrate is exposed, where a cycle is defined by the exposure of the substrate to the minimum required reactant and purge gas flows to form the desired thin-film composition. Due to the process being limited to an atomic layer of growth per cycle, repeated cycles are required to deposit a thin-film having an appreciable thickness. In order to effectively achieve repeated cycles, SALD requires either motion of the substrate past the deposition head or the development of complex equipment such that the delivery head moves with its gas connections, relative to the substrate. Thin-films of appreciable thickness can be accomplished by either 1) using a deposition head containing a sufficient number of gas distribution cycles and moving a substrate (or head) in a unidirectional motion relative to the head (or substrate) or 2) using a head with a limited number of cycles and using relative reciprocating motion. In instances where the substrate or the deposition head are moved by a reciprocating movement, there remains a technical challenge to manage the sequence of gas exposures since the substrate can be exposed to the gases in a different sequence during a forward stroke and a backward stroke. Furthermore, in order to deposit a thin-film over an entire substrate, the substrate or the head may have to travel a long distance in order to expose substrate to the process gases. There remains a need to provide alternative arrangements to both the very large deposition heads and long distance motion profiles such that large substrates may be easily coated.
• One alternative to a single large deposition head is to use multiple deposition heads, or modules, within a larger deposition section. Commonly-assigned U.S. Pat. No. 8,182,608 (Kerr et al.), which is incorporated herein by reference, relates to an apparatus for maintaining the alignment or positional relationship between at least two modules in an SALD system. U.S. Pat. No. 8,182,608 describes aligning multiple delivery heads in a 1-D array, addressing the ability to coating longer substrates or provide thicker thin-film coatings. While simplifying the manufacturing of the deposition head, it does not address the challenge of making coatings of different thicknesses using the same tool, or the footprint required for providing a large deposition section in a manufacturing environment. Additionally, there remains a need for a way to arrange modular heads to be able to coat wider substrates without coating defects or non-uniformity. Additionally, there remains a need for a motion profile that enables the use of small deposition heads in order to build up a sufficient layer thickness from an SALD. Furthermore, there remains a need for a substrate handling means for coating on roll-to-roll webs that enables exposure of the substrate to multiple SALD cycles during deposition, while simultaneously moving the substrate smoothly from the feed roll to the take-up roll.
• In order to function properly, an SALD system must maintain the separation of the reactant gases. Although separated in space and by a purge gas as delivered by the deposition head, the system must be further designed to insure that the gases do not mix in the region between the deposition head and the substrate. Commonly-assigned U.S. Patent Application Publication 2009/0130858 (Levy), relates to an SALD deposition system and method using a delivery head where the distance between the substrate and the deposition head is maintained by gas pressure. In this device, the pressure of flowing reactive and purge gases is used as a means to control the separation between the deposition head and the substrate. Due to the relatively large pressures that can be generated in such a system, gases are forced to travel in well-defined paths and thus eliminate undesired gas intermixing.
• The system of U.S. Patent Application Publication 2009/0130858 operates as a gas-bearing SALD system. The gas bearing operation maintains a close proximity of the substrate to the deposition head, and either the substrate or head must be free to move in the direction normal the deposition head. The use of a gas bearing SALD head is advantaged due to the resultant pressure profiles that separate the precursor gasses by the purge gas and prevent undesired gas intermixing. There remains a need for SALD systems that utilize a gas-bearing deposition head to coat large substrates, particularly for depositions systems with small manufacturing footprints. There remains a need to coat long substrates with deposition heads that are considerably smaller than the coating length, both for piece-parts and particularly for roll-to-roll webs; this need further necessitates novel motion control profiles and substrate handling. There remains a further need for roll-to-roll SALD systems that utilize a gas-bearing deposition head having a simple construction, as well as roll-to-roll systems that can manage potential substrate distortions and can isolate the motion needed for deposition from the global motion of the web through the system. Additionally, there remains a need, for a modular system that can accommodate different substrate form factors, including roll-to-roll webs of substrate, and provide a system that is relatively low in cost and easy to use.
SUMMARY OF THE INVENTION
• The present invention represents a thin film deposition system, including:
• a vacuum-preloaded gas bearing deposition head positioned in an external environment having an ambient pressure, the deposition head having an output face facing a first substrate surface of a substrate and being configured to expose the first substrate surface to a plurality of gaseous materials including one or more reactive gaseous materials, wherein the output face includes a plurality of source openings through which the gaseous materials are supplied and one or more exhaust openings;
• wherein an exhaust pressure at the exhaust openings is less than ambient pressure and a source pressure at the source openings is greater than the exhaust pressure that at the exhaust openings, and wherein the pressure at the outermost source openings is greater than ambient pressure;
• a substrate unit including the substrate, the substrate having a length in the in-track direction;
• a motion control system that moves the substrate unit over the output face of the deposition head in the in-track direction in accordance with a specified motion profile without constraining motion of the substrate unit in a direction normal to the output face of the deposition head, wherein the motion profile includes moving a center of gravity of the substrate unit beyond the first edge of the output face; and
• wherein an integrated pressure across the output face provides a net force on the first substrate surface in a direction normal to the output face, wherein a sum of the forces on the substrate unit in a direction normal to the output face and a sum of moments on the substrate unit are zero at all positions in the motion profile and wherein a gap exists between the first substrate surface and the output face at all positions in the motion profile.
• This invention has the advantage that it enables utilization of greater area of substrate than possible without vacuum-preloading. For example, a vacuum-preloaded deposition head enables the application of SALD coatings to overhung substrates such that the coated area exceeds the in-track size of the deposition head.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1 is schematic block diagram showing the functional elements of an SALD deposition system;
• FIGS. 2A-2C are cross-sectional side views of SALD deposition heads useful in the present invention having a single ALD cycle;
• FIG. 3A is a cross-sectional side view of an alternative embodiment of an SALD deposition head having 1.5 ALD cycles;
• FIG. 3B is a plan view of the SALD head ofFIG. 3A;
• FIGS. 4A-4C are cross-sectional views illustrating the range of motion of a substrate relative to a non-vacuum-preloaded gas bearing deposition head;
• FIGS. 5A-5D are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate relative to a vacuum-preloaded gas bearing deposition head;
• FIGS. 6A-6C are cross-sectional views illustrating the range of motion of a substrate unit including a backer device relative to a non-vacuum-preloaded gas bearing deposition head;
• FIGS. 7A-7C are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head;
• FIG. 8 illustrates a pressure profile between a substrate unit and a vacuum-preloaded gas bearing deposition head when the center of gravity of the substrate unit is moved beyond the edge of the deposition head;
• FIG. 9 is a simplified pressure profile corresponding to the example ofFIG. 8;
• FIG. 10 illustrates a pressure profile between a substrate unit and a vacuum-preloaded gas bearing deposition head having alternating output and exhaust zones when the center of gravity of the substrate unit is moved beyond the edge of the deposition head;
• FIG. 11A-11C are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head, wherein outboard support elements including rollers are used to extend the range of motion;
• FIG. 12A-12C are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head, wherein outboard support elements including air bearings are used to extend the range of motion;
• FIG. 13A-13C are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head, wherein a cantilever spring is used to extend the range of motion;
• FIG. 14A-14C are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head, wherein a vertical element and springs are used to extend the range of motion;
• FIG. 15A-15B are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head, wherein a counter weight mechanism is used to extend the range of motion;
• FIG. 16A-16B are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vacuum-preloaded gas bearing deposition head, wherein a counter weight mechanism including linkages is used to extend the range of motion;
• FIG. 17A-17C are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit relative to a vertically-oriented vacuum-preloaded gas bearing deposition head;
• FIG. 18A-18B are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vertically-oriented vacuum-preloaded gas bearing deposition head, wherein the substrate unit include roller bearings that fit within yoke features on the substrate positioner;
• FIG. 19A-19B are cross-sectional views of an exemplary embodiment illustrating the range of motion of a substrate unit including a backer device relative to a vertically-oriented vacuum-preloaded gas bearing deposition head, wherein the substrate unit is connected to the substrate positioner using flexures;
• FIG. 20A-20B are cross-sectional views of exemplary embodiments similar to that ofFIGS. 19A-19B wherein a force mechanism is used to apply a constant force onto the back surface of the substrate unit;
• FIG. 21 illustrates various quantities affecting the maximum allowable substrate motion during ALD deposition with a one-cycle ALD deposition head;
• FIG. 22 illustrates various quantities affecting the minimum amount of substrate motion required for ALD deposition with a one-cycle ALD deposition head;
• FIG. 23 illustrates an exemplary 2.5-cycle ALD deposition head;
• FIG. 24 illustrates various quantities affecting the maximum allowable substrate motion during ALD deposition for the deposition head ofFIG. 22;
• FIG. 25 illustrates the use of a repeating motion profile for performing ALD deposition using the deposition head ofFIG. 22;
• FIG. 26 illustrates exemplary repeating motion profiles useful for performing ALD deposition using the deposition head ofFIG. 22;
• FIG. 27 illustrates exemplary repeating motion profiles useful for performing ALD deposition using the deposition head ofFIG. 22;
• FIG. 28 illustrates an exemplary repeating motion profile useful for performing ALD deposition using a deposition head where the inert zones and the deposition zone have equal lengths;
• FIG. 29 illustrates various quantities affecting the maximum allowable substrate motion during ALD deposition with a 2.5-cycle ALD deposition head having reactive gas zones with unequal length;
• FIG. 30 illustrates the use of a repeating motion profile for performing SALD deposition using the deposition head ofFIG. 22 where the forward motion distance is equal to a repeat cycle;
• FIG. 31 illustrates various quantities relating to the computation of the number of ALD growth events;
• FIG. 32 illustrates various quantities relating to the computation of the number of ALD growth events;
• FIG. 33 shows a table of flow rates used to provide exemplary SALD coatings; and
• FIG. 34 shows a table of experimental SALD coatings that were made according to exemplary embodiments.
• It is to be understood that the attached drawings are for purposes of illustrating the concepts of the invention and may not be to scale. Identical reference numerals have been used, where possible, to designate identical features that are common to the figures.
DETAILED DESCRIPTION OF THE INVENTION
• Throughout the specification and claims, the following terms take the meanings explicitly associated herein, unless the context clearly dictates otherwise. The meaning of “a,” “an,” and “the” includes plural reference, the meaning of “in” includes “in” and “on.” Additionally, directional terms such as “on,” “over,” “top,” “bottom,” “left,” and “right” are used with reference to the orientation of the figure(s) being described. Because components of embodiments of the present invention can be positioned in a number of different orientations, the directional terminology is used for purposes of illustration only and is in no way limiting.
• The invention is inclusive of combinations of the embodiments described herein. References to “a particular embodiment” and the like refer to features that are present in at least one embodiment of the invention. Separate references to “an embodiment” or “particular embodiments” or the like do not necessarily refer to the same embodiment or embodiments; however, such embodiments are generally not mutually exclusive, unless so indicated or as are readily apparent to one of skill in the art. The use of singular or plural in referring to the “method” or “methods” and the like is not limiting. It should be noted that, unless otherwise explicitly noted or required by context, the word “or” is used in this disclosure in a non-exclusive sense. Even though specific embodiments of the invention have been described herein, it should be noted that the present invention is not limited to these embodiments. In particular, any features described with respect to one embodiment may also be used in other embodiments, where compatible. The features of the different embodiments can be exchanged, where compatible.
• It is to be understood that elements not specifically shown, labeled, or described can take various forms well known to those skilled in the art. In the following description and drawings, identical reference numerals have been used, where possible, to designate identical elements. It is to be understood that elements and components can be referred to in singular or plural form, as appropriate, without limiting the scope of the invention.
• The example embodiments of the present invention are illustrated schematically and are not to scale for the sake of clarity. One of ordinary skill in the art will be able to readily determine the specific size and interconnections of the elements of the example embodiments of the present invention. Therefore, the provided figures are not drawn to scale but are intended to show overall function and the structural arrangement of some embodiments of the present invention.
• The embodiments of the present invention relate components for systems useful for thin-film deposition. In preferred embodiments, the thin-film deposition is done using a spatial atomic layer deposition (SALD) process. For the description that follows, the term “gas” or “gaseous material” is used in a broad sense to encompass any of a range of vaporized or gaseous elements, compounds, or materials. Other terms used herein, such as: reactant, precursor, vacuum, and inert gas, for example, all have their conventional meanings as would be well understood by those skilled in the materials deposition art. Reactant gas flows can include multiple reactive species together with inert gaseous species. In some embodiments, the reactive gases can include a reactive plasma, such as supplied by a remote plasma source. One type of remote plasma source that can be used includes a surface dielectric barrier discharge source. As such, plasma-enhanced spatial ALD (PE-SALD) arrangements are considered to be useful in some embodiments. While the exemplary embodiments are described in the context of SALD systems, those skilled in the art will recognize that aspects of the present invention can also be used for any application which involves exposing a substrate to one or more gaseous substances, such as chemical vapor deposition processes.
• Unless otherwise explicitly noted or required by context (for example, by the specified relationship between the orientation of certain components and gravity), the term “over” generally refers to the relative position of an element to another and is insensitive to orientation, such that if one element is over another it is still functionally over if the entire stack is flipped upside down. As such, the terms “over”, “under”, and “on” are functionally equivalent and do not require the elements to be in contact, and additionally do not prohibit the existence of intervening layers within a structure. The term “adjacent” is used herein in a broad sense to mean an element next to or adjoining another element. The figures provided are not drawn to scale but are intended to show overall function and the structural arrangement of some embodiments of the present invention.
• Embodiments of the present invention are illustrated and described with a particular orientation for convenience; and unless indicated specifically, such as by discussion of gravity or weight vectors, no general orientation with respect to gravity should be assumed. For convenience, the following coordinate system is used: the z-axis is perpendicular to the output face of the deposition head, the x-axis is parallel to the primary motion direction (in the plane of the output face), and the y-axis is perpendicular to the primary motion axis (in the plane of the output face). Roll, pitch, and yaw are as used herein have their commonly understood definitions. To facilitate interpretation of relative motion and degrees of freedom, the following clarifications are provided. Roll is the rotation about an axis parallel to the primary motion axis (x-axis). Pitch is the rotation about the y-axis in the plane of the output face of the delivery device and perpendicular to the primary motion axis. Yaw is the rotation about the z-axis which is normal to the output face of the delivery device.
• An ALD process accomplishes thin-film growth on a substrate by the alternating exposure of two or more reactive materials, commonly referred to as precursors, either in time or space. A first precursor is applied to react with the substrate. The excess of the first precursor is removed and a second precursor is then applied to react with the substrate surface. The excess of the second precursor is then removed and the process is repeated. In all ALD processes, the substrate is exposed sequentially to a series of reactants that react with the substrate. The thickness of the ALD (and SALD) deposited thin-films is controlled by the number of ALD cycles to which the substrate is exposed, where a cycle is defined by the exposure to the minimum required reactant and purge gas flows to form the desired thin-film composition. For example, in a simple design, a single cycle can provide one application of a first reactant gaseous material G1 and one application of second reactant gaseous material G2. In order to effectively achieve repeated cycles, SALD requires either motion of the substrate past the deposition head or the development of complex equipment such that the delivery head with its gas connections, can be moved relative to the substrate. Thin-films of appreciable thickness can be accomplished by either 1) using a deposition head containing a sufficient number of gas distribution cycles and moving the substrate (or the deposition head) in a unidirectional motion relative to the deposition head (or substrate) or 2) using a deposition head with a limited number of cycles and using relative reciprocating motion.
• In order to effectively use an SALD deposition head for thin-film deposition, it is commonly employed within a larger SALD system, or apparatus. Typically, such systems are specifically designed to deposit thin films on a particular type of substrate (for example, either rigid or flexible). Furthermore, SALD systems typically utilize a singular motion profile type that is chosen as a result of the design of the deposition head and the type of substrate being coated. In many cases, SALD systems are further designed for a specific application, and as such are configured to coat a single material at a given thickness on a substrate having a particular form factor.
• As known by one skilled in the art, each SALD system requires at least three functional elements in order to effectively deposit a thin-film, namely a deposition unit, a substrate positioner and a means of relative motion. To date, the specific design of each functional element has generally differed from system to system. As will be described, preferred embodiments of the SALD systems of the present invention are modular in nature, and as such includes a range of components of differing design that can be exchanged to perform the function of a particular functional element within the novel SALD platform. The design and advantages of specific components useful in a range of SALD systems, and design and advantages of inventive elements and configurations of the novel modular SALD platform of the present invention will be better understood with respect to the Figures.
• As shown in schematic block diagram ofFIG. 1,SALD system200 of the present invention is preferably one in which asubstrate97 is moved relative to a fixeddeposition unit210. As such,substrate97 is positioned over theoutput face134 of adeposition unit210 bysubstrate positioner module280, and relative motion between thesubstrate97 and thedeposition unit210 is accomplished by motion of thesubstrate positioner module280 using relative motion means270, which can also be referred to as a motion controller or a motion control means. Thedeposition unit210,substrate positioner module280 and relative motion means270 are functional elements ofdeposition subsystem205 ofSALD system200. In various embodiments of the present invention, thedeposition unit210 can be asingle deposition head30 or can be a deposition unit that include an array of deposition heads30. The relative motion means270 interacts with thesubstrate positioner module280 to move thesubstrate97 relative to thedeposition unit210.
• Thesubstrate positioner module280 is preferably an interchangeable substrate positioning module, with the modular system having multiple substrate positioning modules that can be easily exchanged into theSALD system200, where the different substrate positioning modules are configured to handle different types ofsubstrates97 and different substrate form factors.
• Many types of substrates can be coated with theSALD system200. Thesubstrates97 used in the present invention can be any material that acts as a mechanical support for the subsequently coated layers. Thesubstrate97 can include a rigid material such as glass, silicon, or metals. The substrate can also include a flexible material such as a polymer film or paper. Useful substrate materials include organic or inorganic materials. For example, the substrate can include inorganic glasses, ceramic foils, and polymeric materials. The thickness ofsubstrate97 can vary, typically from about 25 μm to about 1 cm. Using aflexible substrate97 allows for roll processing, which can be continuous, providing economy of scale and economy of manufacturing relative to flat or rigid supports.
• In some example embodiments, thesubstrate97 can include a temporary support or support material layer, for example, when additional structural support is desired for a temporary purpose, e.g., manufacturing, transport, testing, or storage. In these example embodiments, thesubstrate97 can be detachably adhered or mechanically affixed to the temporary support. For example, a flexible polymeric support can be temporarily adhered to a rigid glass support to provide added structural rigidity during the deposition process. The glass support can be removed from the flexible polymeric support after completion of the manufacturing process. Thesubstrate97 can be bare indicating that it contains no substantial materials on its surface other the material from which it is composed. Thesubstrate97 can include various layers and patterned materials on the surface.
• The relative motion means270 is adapted to connect to the interchangeable substrate positioner modules, and as such, the relative motion means270 and the interchangeable substrate positioner modules preferably contain appropriate mating features. Thesubstrate positioner module280 is designed to position thesubstrate97 in the x- and y-directions relative to theoutput face134 of thedeposition unit210. TheSALD system200 may also include a secondary substrate positioner (not shown) which is designed to control the position of thesubstrate97 in the z-direction.
• In various configurations, thesubstrate97 can be attached to a backer device during deposition. The backer device can be used as heat source for the substrate, or to stiffen otherwise flexible substrates. A backer that is temporarily attached to the substrate, by vacuum for example, is intended to move with the substrate during relative motion between the substrate and a fixed deposition head. The backer attachment can provide greatly increased rigidity and flatness to flexible substrates. A backer device useful in the present invention can be larger than the substrate, as might be used to stabilize piece-parts of flexible substrate or approximately the same size as the substrate, or significantly smaller than the substrate when the substrate is rigid and self-supporting. As used herein, the “substrate unit” refers to either thesubstrate97 alone or asubstrate97 with an attached backer device; the substrate unit has relative motion relative to thedeposition unit210.
• Thedeposition unit210 can use any type of SALD deposition head that is known in the art.FIGS. 2A-2C illustrate deposition heads30 that are configured to simultaneously supply a plurality of gaseous materials from the output face in different gas zones within adeposition zone305. In all three figures, thedeposition zone305 contains the necessary gas zones for a single two-step ALD deposition cycle. Moving from left to right within thedeposition zone305, there is a first reactive gas zone313 (G1) followed by an inert gas purge zone314 (P), and a second reactive gas zone315 (G2). As the relative motion means270 (FIG. 1) moves thesubstrate97 relative to the deposition head30 (the x-direction being the primary motion direction as indicated by motion arrow98), a particular location on thesubstrate97 sees the above sequence of gases which results in ALD deposition. Deposition heads30 of the present can include adeposition zone305 with gas zones for any number of ALD deposition cycles, the single ALD cycle illustrated is for clarity of understanding.
• The SALD systems of the present invention can use any deposition head geometry so long it has the required gas delivery to form gas zones between thedeposition head30 and thesubstrate97 in the required order to accomplish an ALD cycle, as illustrated by thesimplified deposition head30 ofFIG. 2A. In preferred embodiments, the reactive gases (G1 and G2, for example) have little or no intermixing to avoid a CVD component during film deposition or gas phase reactions. The purge zone314 (P) serves to separate the reactive gases G1, G2 and allows for the removal of any reaction byproducts from the substrate surface as it moves through thepurge zone314.
• A single deposition cycle (moving from left to right) is defined by an inert gas flow I, followed by a first reactive gas flow G1, followed by an inert purge gas flow P, and lastly by a second reactive gas flow G2. Thedeposition zone305 has a deposition zone length that spans the distance from the start of the first reactive gas zone to the end of the last reactive gas zone (e.g., from the firstreactive gas zone313 to the secondreactive gas zone315 inFIG. 2A).
• The deposition heads30 illustrated inFIGS. 2A-2C, have extendedinert zones308,309 on either side of thedeposition zone305. The firstinert zone308 has a first inert zone length that spans the distance from theleft edge321 of thedeposition head30 to the boundary of the firstreactive gas zone313. The secondinert zone309 has a second inert zone length that spans the distance from the boundary of the secondreactive gas zone315 to theright edge322 of thedeposition head30. The extendedinert zones308,309 isolate thedeposition zone305 from theexternal environment15 and enable thedeposition head30 tocoat substrates97 that are substantially longer than the length of thedeposition head30 without exposing the growth region to theexternal environment15. Deposition heads of the prior art are typically operated within a larger system where the external environment is controlled to be inert, under vacuum, or both. In preferred embodiments of the present invention, thedeposition head30 can be used at atmospheric pressure without any additional environmental controls for theexternal environment15. One of the advantages of the present invention is that thedeposition head30 andSALD system200 containing it can be used to coat onsubstrates97 whose length is much larger than the length of thedeposition zone305. A further advantage of some embodiments of the present invention is the ability to control the environment of the region of the substrate being actively coated during deposition. Additionally, the relatively small deposition head size allows for lower cost manufacturing of the deposition head.
• It is known that ALD is self-limiting, meaning that when all available sites on a substrate surface have reacted with a precursor there is no further reaction during that half-step. When both half-reactions in a deposition cycle have sufficient time and available precursor to reach this state, it is said that the ALD cycle has reached “saturation”. ALD depositions done in these conditions are by definition, saturated ALD, and continued exposure to the precursors does not change significantly the deposition amount. In SALD, the substrate velocity and length of reaction zones determine the exposure time to a give precursor. For a given velocity, there is a minimum zone length required to reach saturation (i.e., a “saturation length”) and zone lengths longer than the saturation length do not add film thickness during material deposition. SALD systems of the present invention can be used in both saturated and sub-saturated conditions. One advantage of the present invention is that sub-saturated growth can still be deterministic, since each point on thesubstrate97 will see the same concentration of precursors for a time which is set by the substrate velocity and motion profile.
• Themotion arrow98 indicates one known motion of thesubstrate97 useful in SALD which is to move thesubstrate97 in a smooth oscillating, or reciprocating, motion through theentire deposition zone305 such that the substrate “sees” the required number of cycles to produce the desired coating thickness (as discussed above). In preferred embodiments of the present invention the substrate motion is controlled such that the region being actively coated is prevented from experiencing the external environment during coating. This has the advantage of avoiding contamination of the thin-films during growth by preventing exposure to any reactive species or dust particulates or other contaminates that may be present in the external environment outside of the controlled environment defined by the region between thedeposition head30 and thesubstrate97.
• Thedeposition head30 ofFIG. 2B illustrates an embodiment where one or more of the gas zones use a transverse arrangement, such as that disclosed in the aforementioned commonly-assigned U.S. Pat. No. 7,456,429 (Levy et al.), entitled “Apparatus for atomic layer deposition.” In a transverse flow arrangement, the flow of gases during deposition is orthogonal, or transverse, to the direction of substrate motion and is exhausted either out the edges of thedeposition head30, or into exhaust slots along the perimeter of thedeposition head30. As illustrated, thedeposition head30 has gas slots110 (i.e., output slots112) that are configured to supply the gases into their corresponding gas zones. In other embodiments, thedeposition head30 provides gas to the elongated parallel gas zones through an array of orifices, rather than through the illustrated output slots112 (elongated channels).
• Thedeposition head30 ofFIG. 2C illustrates a preferred gasbearing deposition head30 of the present invention. The principles and design of gas bearing deposition heads30 has been described in detail in the aforementioned U.S. Patent Application Publication 2009/0130858, as well as in commonly-assigned U.S. Pat. No. 7,572,686 (Levy et al.), entitled “System for thin film deposition utilizing compensating forces.” As shown inFIG. 2C, anexemplary deposition unit210 includes adeposition head30 that operates on a vacuum-preloaded gas bearing principle having an output face134 (facing upward) havinggas slots110 which provide gases into the gas zones and exhaust gases from the gas zones. Gases are provided into the gas zones by spatially separated elongated output slots112 (extending in the y-direction). Each gas zone includes acorresponding output slot112.Adjacent exhaust slots114 remove (or exhaust) gas from the gas zones. Theexhaust slots114 are positioned to define the boundaries of the various gas zones. As illustrated, the gas zones are equivalent to those ofFIGS. 2A and 2B.
• In these preferred embodiments wherein thedeposition head30 operates using a gas bearing principle thesubstrate97 is positioned above theoutput face134 of thedeposition head30 and is maintained in close proximity to theoutput face134 by an equilibrium between the pull of gravity, the flow of the gases supplied to theoutput face134 through theoutput slots112, and a slight amount of vacuum at theexhaust slots114. While the gas openings in this example are gas slots110 (also referred to as gas channels) that extend in the y-direction, one skilled in the art will recognize that the gas openings could also have other geometries, such as a row of nozzles or circular orifices, so long as the proper gases are delivered into and exhausted from the gas zones between the deposition head and the substrate.
• As shown inFIG. 2C, the gases are introduced and exhausted in alternatingoutput slots112 andexhaust slots114 in theoutput face134 of thedeposition head30. The flow of gases between theoutput slots112 during deposition is primarily in the direction of substrate travel (forward and backward) toward theadjacent exhaust slots114. As discussed earlier, the region that spans the reactive gas zones can be referred to as thedeposition zone305, which is preferably surrounded by twoinert zones308,309. The individual gas zones within thedeposition zone305, where thesubstrate97 is exposed to each gas, generally extend outward from thecorresponding output slot112 to the twoadjacent exhaust slots114 as illustrated for the firstreactive gas zone313, thepurge zone314, and the secondreactive gas zone315. In the illustrated configuration, the extendedinert zones308,309 extend from the inertgas output slots112 to the edges of thedeposition head30. In alternative embodiments, the extendedinert zones308,309 can includeadditional output slots112 or other gas supply features. Additionally, the extendedinert zones308,309 can includeexhaust slots114, or other exhaust features, to provide additional protection/separation from theexternal environment15.
• Using any of the embodiments ofdeposition head30 ofFIGS. 2A-2C, an SALD deposition process can be accomplished by oscillating the position of thesubstrate97 across the deposition head30 (in the in-track direction indicated by the motion arrow98) for the number of cycles necessary to obtain a uniform deposited film of the desired thickness for the given application.
• FIG. 3A is a cross-sectional view of adeposition head30 illustrating a preferred embodiment of the present invention where thedeposition zone305 is arranged to be symmetric, so that as thesubstrate97 is moved relative to the deposition head30 a position can “see” a full cycle exposure in either a forward or reverse direction.FIG. 3B illustrates a plan view corresponding to the cross-sectional view ofFIG. 3A, where the cross-sectional view is taken along the line A-A′ of the plan view. In common parlance, thedeposition head30 illustrated inFIG. 3A-3B can be referred to a “one-and-a-half cycle head” or a “1.5 cycle head.” Moving from left-to-right through thedeposition zone305, thesubstrate97 is exposed to (in order) a firstreactive gas zone313 where the substrate is exposed to a first reactive gas G1, aninert purge zone314 where the substrate is exposed to an inert purge gas P, a secondreactive gas zone315 where the substrate is exposed to a second reactive G2, anotherinert purge zone314 where the substrate is exposed to the purge gas P, and another firstreactive gas zone313 where the substrate is exposed to the first reactive gas G1. Moving in the reverse direction from right-to-left through thedeposition zone305, thesubstrate97 is exposed to the same sequence of gases as in the forward (left-to-right) direction, namely the first reactive gas G1, the inert purge gas P, the second reactive gas G2, the inert purge gas P, and the first reactive gas G1. The advantage of this symmetry is that feeding thesubstrate97 from left-to-right or right-to-left results in equivalent exposure, and entrance and exit sides of thedeposition head30 depend of the direction of relative motion of thesubstrate97 not the design of thedeposition head30.
• As with the previous embodiments, the gas zones (or regions) are between thesubstrate97 and thedeposition head30. The labels inFIG. 3A are placed above the substrate for clarity and to further emphasize thesmall working distance94 between the process-side ofsubstrate97 and theoutput face134 of thedeposition head30 enabled by the use of a vacuum-preloaded gasbearing deposition head30. As illustrated in the plan-view ofFIG. 3B, in addition to the output slots112 (shown as black lines) and the exhaust slots114 (shown as gray lines) in the deposition zone305 (shown as a shaded area), there areadditional output slots401 orthogonal to thegas slots110 in thedeposition zone305. The additionalgas output slots401 provide inert gas to the cross-track edge region of thedeposition head30, providing further isolation of thedeposition zone305 from theexternal environment15.
• The exemplary gasbearing deposition head30 ofFIG. 3A hasgas slots110 corresponding to 1.5 ALD cycles to provide the proper sequence of gas exposure in the forward and reverse directions. As thesubstrate97 is oscillated back and forth over thedeposition head30, it will provide only a single ALD cycle (one G1 and one G2 exposure) per single direction pass over thedeposition head30, therefore a round trip oscillation provides two ALD cycles. Furthermore, when the second precursor G2 is reactive with the external environment, while the first precursor G1 is not, this arrangement provides additional protection against unwanted reactions involving G2. An example of a precursor pair that would benefit from this arrangement is water and trimethylaluminum (TMA), where water is the non-reactive precursor G1 and TMA is the highly reactive precursor G2.
• Thedeposition head30 is preferably constructed of a material which does not react with the precursor gases and can withstand the required temperatures without significant deformation. One preferable material is stainless steel. It is recognized that other materials can also be used, but differential thermal expansions must be kept low to prevent distortions. As described, thedeposition head30 delivers multiple reactive and inert process gasses throughoutput face134. Connection of the various gas sources to thedeposition head30 can be accomplished using individual pipe or tubing connections distributed about the periphery of thedeposition head30. In an exemplary configuration, commercially available fittings, such as Swagelok VCR series components, are used for gas source connections. In preferred embodiments, the gases are supplied to thedeposition head30 via a manifold.
• A relatively clean external environment is useful to minimize the likelihood of contamination, but is not necessary. Full “clean room” conditions or an inert gas-filled enclosure can be used in systems of the present invention, however preferred embodiments do not be require control of the external environment and are advantaged for that reason. The apparatus of the present invention is advantaged in its capability to perform deposition onto asubstrate97 over a broad range of temperatures, including room temperature, or near-room temperature, in some embodiments. The apparatus of the present invention can operate in a vacuum environment, but is particularly well suited for operation at or near atmospheric pressure. In preferred embodiments, the SALD process can be performed at or near atmospheric pressure and over a broad range of ambient and substrate temperatures, preferably at a temperature of under 300° C.
• An SALD deposition head operating as a vacuum-preloaded gas bearing has been described in the aforementioned U.S. Pat. No. 7,572,686 (Levy et al.), which is incorporated herein by reference. As noted, the use of a vacuum-preloaded gas bearing can provide efficiency of materials utilization, freedom from gas intermixing, and fast reaction kinetics due to the very small gap between the substrate (deposition side) and the output face of the deposition head. Desired gaps for prior art optimum operation are less than 50 μm, in some cases less than 30 μm. The present invention has a preferred operation with a substrate-head fly height (i.e., process gap, d_P) of less than 30 μm, or even less than 10 μm which is practical and preferred for stable vacuum-preloaded bearing operation with minimal gas flows. Experimental measurements using the process parameters of the present invention have more preferred gaps of approximately 5 μm, enabled by careful co-optimization of the mass flows of reactant and inert gasses, as well as exhaust vacuum levels. This preferred operation condition is within a range more restrictive than previously contemplated and enables the coating of much larger substrates than could be accomplished using a similar geometry with a gap of 30 μm. Furthermore, in preferred embodiments of the present invention, the coating of large substrates can be accomplished with the use of a vacuum attached substrate backer, with an associated weight.
• As noted, the use of vacuum preloading is a means to improve gap control between the SALD deposition head and the substrate to improve gas confinement, avoid intermixing, and increase reaction kinetics. U.S. Pat. No. 7,572,686 describes an operation condition wherein the vacuum “pressure” at the exhaust slots is at least 2× the weight per unit area of the substrate (expressed in equivalent pressure units evaluated over the area of the output face). In this previous vacuum-preloaded gas bearing design it is taught that the deposition head may be traversed across the fixed substrate wherein the center of gravity of the deposition head would always be over the substrate. Embodiments of the present invention enable the coating of substrates that are significantly longer than a fixed-position deposition head, and provide for a greater extent of substrate motion in gas bearing SALD systems, and are desirable for use in a modular SALD system. The present invention addresses issues of operation during loading, preheating, and depositing when using a gas bearing deposition head that is smaller than the substrate being coated, and in particular for systems where the region of the substrate not over the deposition head is subject to an external environment.
• Referring back toFIGS. 3A and 3B, the vacuum-preloaded gasbearing deposition head30 provides gas pressure againstsubstrate97, such that the workingdistance94 is maintained, at least in part, by the force of pressure that is exerted. By maintaining some amount of gas pressure betweenoutput face134 and the process surface of thesubstrate97, the apparatus of the present invention provides at least some portion of an air bearing, or more properly a gas fluid bearing, between thedeposition head30 and thesubstrate97 as illustrated inFIG. 3A. Importantly, the effect of controlling the workingdistance94, or gap, between thedeposition head30 and thesubstrate97 using a gas fluid bearing generates a pressure profile in thedeposition zone305 which advantageously maintains the isolation of the reactive gas zones. By allowing thesubstrate97 to float on the gas flows from thedeposition head30, pressure fields are set up in thereactive gas zones313,315 and purgezones314 cause the gases to be directed from theoutput slots112 to theexhaust slots114 with little or no intermixing with other gas streams, particularly when operated as a vacuum-preloaded gas bearing. One advantage of maintaining the workingdistance94 using this gas bearing design is the allowance of essentially frictionless motion of thesubstrate97 over thedeposition head30 using any suitable type of transport mechanism. Alternatively, thedeposition head30 can be caused to “hover” above the surface ofsubstrate97 as it is channeled back and forth, sweeping across the surface ofsubstrate97 during materials deposition.
• A deposition process can be accomplished using the SALD system ofFIG. 3A by oscillating thesubstrate97 across thedeposition head30 for the number of cycles necessary to obtain a uniform deposited film of the desired thickness for the given example. In alternative embodiments, other motion profiles are used which have an oscillatory component with net forward movement of the substrates. A substrate positioner module280 (FIG. 1) is used to move thesubstrate97 in preferred embodiments of the present invention with a specified motion profile. As previously described, thesubstrate positioner module280, or substrate positioner, is configured to position and move the substrate in the x-y plane. Thesubstrate positioner module280 can be considered as the payload for a motion actuator (not shown) such as a linear motion controller. As shown inFIG. 3A, thesubstrate97 has a length in the in-track direction (x-direction) that is longer than the length of theoutput face134 and longer than the length of thedeposition zone305. In preferred embodiments of the present invention thesubstrate positioner module280 is configured, via the motion actuator, to control the motion of thesubstrate97 in a specified motion profile that prevents a region on thesubstrate97 where material has been deposited from being exposed to theexternal environment15 prior to having achieved a final material deposition amount.
• Advantageously, thedeposition head30 can be fabricated at a smaller size than the substrates that it will be used to coat. Furthermore, the geometry of the deposition head shown inFIG. 3A allows for theoutput slots112 and theexhaust slots114 to be positioned in close proximity to each other, often at distances which are difficult to achieve for other types of deposition heads. For example, in one embodiment, theoutput slots112 have a width of 0.005 inches (0.127 mm) and is extended in length to about 3 inches (75 mm). Advantageously, the SALD system of the present invention may be used tocoat substrates97 that are significantly longer than the length of thedeposition head30 using a simple oscillating motion while simultaneously maintaining the distance between thedeposition head30 and thesubstrate97 using a vacuum-preloaded gas bearing design. Without the use of a vacuum-preloaded gas bearing design,long substrates97 would require mechanical positioning and constraint in the z-direction due to tipping forces that will be better understood with respect to the discussion ofFIGS. 4A-4C.
• The resultant force vector of a gas film emanating from theoutput face134 of thedeposition head30 can be determined by the area integral of the gas pressure distribution over theoutput face134. For purposes of simplified discussion, adeposition head30 having elongated output gas slots in the cross-track direction, as inFIGS. 3A-3B, will be considered illustrative of deposition heads30 of various comparable geometries. In the following examples, the pressure distribution along the direction of theelongated output slots112 will be considered uniform and edge effects along the lateral edges ignored. This enables a 2-D representation which is illustrated in the following figures. For non-vacuum-preloaded gas bearing designs having an output face with a uniform pattern of pressurized slots, orifices, or porous media, the resultant force vector acts at the area centroid of the pattern and is a directed outward along the normal of the output face. An example non-vacuum-preloaded gas bearing design is illustrated inFIGS. 4A-4C, which includes three equidistantly spaced parallel elongated output slots.Force vectors113 are shown corresponding to the net force from each output slot.
• FIGS. 4A-4C illustrate the issues associated with the use of a gasbearing deposition head30 without vacuum preloading. These figures are cross-sectional views through thesubstrate positioner288,substrate97 anddeposition head30. The interaction with the center ofgravity71 of therigid substrate97 is shown in various substrate positions. As was described above, a motion actuator (not shown) will move a substrate positioner module280 (FIG. 1); thesubstrate positioner288 illustrated is a component of asubstrate positioner module280. In this example, thesubstrate positioner288 consists of aframe285 having aframe opening286. Thesubstrate97 fits within theframe opening286, positioning it in x-y, but allowing for free movement in the z-direction. Moving thesubstrate positioner288 causes thesubstrate97 to move laterally over theoutput face134 of thedeposition head30 in accordance with a specified motion profile without constraining motion of thesubstrate97 in a direction normal to theoutput face134 of thedeposition head30.
• As shown inFIGS. 4A-4B, such an arrangement is stable as long as the center ofgravity71 of therigid substrate97 is located above theoutput face134 of thedeposition head30. InFIG. 4B thesubstrate97 has been moved to bring the center ofgravity71 to the edge of the deposition zone (i.e., over the outermost slot). Thesubstrate97 has a weight vector acting in a downward direction, and the weight vector can pass through the area bounded by the outermost output slots while maintaining levitation and staying “attached” to the head provided that the pressure from thedeposition head30 was adequate. Under these conditions, the substrate can be traversed or reciprocated while maintaining the weight vector within the bounds of the deposition zone. In the case ofFIG. 4B, the gas flow from thedeposition head30 is such that there is only a region of positive pressure as indicated byforce vectors113, which provides no restoring force to stabilize thesubstrate97. As shown inFIG. 4C when the motion profile includes moving a center of gravity of therigid substrate97 beyond the edge of the last output slot, thesubstrate97 will tip and the coating integrity is lost, and therefore the non-vacuum-preloadeddeposition head30 ofFIGS. 4A-4C is unable to coatlonger substrates97 without additional mechanical constraints and complexities. This represents a substantial limitation on the range of motion for thesubstrate97 relative to thedeposition head30 and results in a very small area in the center of thesubstrate97 being available for deposition.
• FIGS. 5A-5D illustrate the use of a vacuum-preloaded gasbearing deposition head30 in accordance with an exemplary embodiment of the present invention in a similar arrangement to that shown inFIGS. 4A-4C. For anoutput face134 comprised of uniformly distributed alternating positive and negative pressure regions, such as illustrated inFIG. 3A, the resultant force vector is also normal to theoutput face134 but may be directed toward or away from the output face134 (depending on the relative intensity of the positive and negative force vectors). In a vacuum-preloaded gasbearing deposition head30, the direction and magnitude of the resultant force depends on the gap distance between theoutput face134 and thesubstrate97 and changes from a positive (repulsive) condition when thesubstrate97 is close, to negative (attractive) condition when thesubstrate97 is further away. One minimalist example of this preloaded bearing is a single elongated vacuum exhaust slot positioned centrally between two parallel output slots (or output channels) as illustrated inFIGS. 5A-5D. Theforce arrows113 correspond to regions of positive pressure formed above the output slots112 (FIG. 3A) and theforce arrow123 corresponds to regions of negative pressure formed above the exhaust slots114 (FIG. 3A). When the center ofgravity71 of thesubstrate97 has been moved beyond the edge of thedeposition head30 as illustrated inFIG. 5D, the exhaust flow into thedeposition head30 contributes to an integrated pressure across theoutput face134 that provides a net force on thesubstrate97 in a direction towards theoutput face134. This net force enables the system to counterbalance the weight of thesubstrate97 so that the sum of moments on thesubstrate97 about the first edge of theoutput face134 is zero even when the center ofgravity71 is beyond theedge311 of thedeposition head30. This compensating force advantageously allows for a much larger coating area using a reciprocating motion than could be accomplished using a non-vacuum-preloaded gas bearing design, or stated another way enables a much larger range of motion in the motion profile.
• In general, vacuum-preloaded gas bearing deposition heads30 have anoutput face134 with a plurality of output openings (e.g.,output slots112 in andFIG. 3A) and one or more exhaust openings (e.g.,exhaust slots114 inFIG. 3A). The output openings can also be called source openings. The exhaust pressure at the exhaust openings is less than ambient pressure, and the output pressure at the output openings is greater than that at the exhaust openings. Generally, the pressure at the outermost output openings will be greater than ambient pressure.
• In some embodiments, thesubstrate97 is attached to a backer device during deposition. The backer device can be used as heat source for the substrate, or to stiffen otherwise flexible substrates. A substrate backer that is temporarily attached to thesubstrate97, by vacuum for example, is intended to move with thesubstrate97 during relative motion between thesubstrate97 and a fixeddeposition head30. The backer attachment can provide greatly increased rigidity and flatness toflexible substrates97. A backer device useful in the present invention can be larger than thesubstrate97, as might be used to stabilize piece-parts offlexible substrate97 or approximately the same size as thesubstrate97, or significantly smaller than thesubstrate97 when thesubstrate97 is rigid and self-supporting. The backer device may have a similar aspect ratio and/or surface area as thesubstrate97, or may have a different design. For instance, a backer device with a square contact area could be attached to roundplanar substrate97, like a wafer for example. Many other configurations are possible and fall within the scope of the present invention.
• FIGS. 6A-6C illustrate the issues associated with the use a gasbearing deposition head30 without vacuum preloading to support asubstrate unit74 that includes thesubstrate97 and abacker device70, and are analogous toFIGS. 4A-4C. As shown, thesubstrate97 is attached tobacker device70. The attachment force used to hold thebacker device70 andsubstrate97 in contact can be provided using any means known in the art including mechanical means, and is preferably achieved by a vacuum force. In the illustrated configuration, thebacker device70 fits within frame opening286 inframe285 of thesubstrate positioner288, and thesubstrate97 is positioned below theframe285. As was described above, a motion actuator (not shown) will move thesubstrate positioner288 during thin film deposition. Thesubstrate positioner288 is configured to move thebacker device70 and the attachedsubstrate97 laterally over theoutput face134 of thedeposition head30 in accordance with a specified motion profile without constraining motion of thebacker device70 and thesubstrate97 in a direction normal to theoutput face134 of thedeposition head30. Thebacker device70 has a weight vector that passes through the mass centroid of thesubstrate97, and the combination of thebacker device70 andsubstrate97 is herein referred to as thesubstrate unit74. The combinedsubstrate unit74 has a center ofmass71. Thesubstrate unit74 is levitated in the condition where the weight vector is over thedeposition head30 and the pressure in the bearing region is adequate to support the weight of thesubstrate unit74. In a non-vacuum-preloaded configuration, thesubstrate unit74 can be traversed or reciprocated over a comparable distance similar to the case illustrated inFIGS. 4A-4C, where thesubstrate unit74 contains only the substrate97 (with no backer).
• As shown inFIGS. 6A-6B, at best, such an arrangement is stable for as long as the center ofgravity71 of thesubstrate unit74 is located above the region bounded by the positive pressure of the output slots, typically the deposition zone.FIG. 6B is analogous toFIG. 4B, illustrating the condition where thesubstrate unit74 has been moved to a position bringing the center ofgravity71 to the edge of the last output slot. Thesubstrate unit74 is now at (or beyond) the tipping point in the case ofFIG. 6B, where the gas flow from the non-vacuum-preloaded gasbearing deposition head30 is such that there is only a region of positive pressure as shown byforce vectors113, which provides no restoring force to stabilize thesubstrate unit74. As shown inFIG. 6C when thesubstrate unit74 is moved to a position where the center ofgravity71 of thesubstrate unit74 is beyond the last output channel, the substrate will tip and the coating integrity is lost. As noted above, this represents a substantial limitation on the range of motion for thesubstrate unit74 relative to thedeposition head30 and results in a very small area in the center of thesubstrate97 being available for deposition.
• FIGS. 7A-7C illustrate an exemplary embodiment using a vacuum-preloaded gasbearing deposition head30 together with asubstrate unit74 including asubstrate97 and abacker device70. Theforce arrows113 correspond to regions of positive pressure formed above the output slots112 (FIG. 3A) and theforce vector123 corresponds to a region of negative pressure formed above an exhaust slot114 (FIG. 3A). As shown inFIG. 7C, in this embodiment, the exhaust flow into thedeposition head30 contributes to an integrated pressure across theoutput face134 that provides a net force on thesubstrate97 in a direction towards theoutput face134 to counterbalance the weight of thesubstrate unit74 so that the sum of moments on thesubstrate unit74 about theedge311 of theoutput face134 is zero at a much larger range of positions in the motion profile. This compensating force advantageously allows for a much larger coating area using a reciprocating motion than could be accomplished using a non-vacuum-preloaded gas bearing design, or stated another way, a much larger range of motion in the motion profile. This basic arrangement is compatible with a wide range ofsubstrate units74, including those described above.
• As illustrated inFIGS. 2A, 2B, 3A, and 3B, thedeposition head30 of the present invention preferably has extendedinert zones308,309 on either side of thedeposition zone305. Theseinert zones308,309, as previously described, have a length and contain an output slot which provides an inert gas. The inert zones have an associated pressure which is positive, with respect to the air pressure in theexternal environment15, and so provides a repulsive normal force on thesubstrate unit74. When thesubstrate unit74 is moved to a position where the center ofgravity71 is displaced from the area centroid of the active area of the vacuum-preloaded gasbearing deposition head30, and in particular outboard of theoutput slot112 of theinert zone308,309, there will be a position where the resultant force from the preloaded gas bearing changes from repulsive to attractive. The actual displacement where this occurs depends on the inert zone length and pressure. When the center ofgravity71 of thesubstrate unit74 is outboard of the center of the inert zone length, and still over thedeposition head30, the force in the active zone is attractive. In practice, the location that the net force that thedeposition head30 imparts on thesubstrate unit74 changes from repulsive (floating) to attractive (preventing thesubstrate unit74 from detaching/tipping) is determined by the pressure gradient in theinert zone308. Referring toFIGS. 5D and 7C, when thedeposition head30 is configured consistent with a vacuum preload gasbearing deposition head30 as illustrated inFIGS. 2C, 3A, and 3B, the vertical forces include the downward force of thesubstrate unit74 weight, the upward lift from theinert zones308,309, and the net force from the preloaded active zone (from the distributedoutput slots112 and exhaust slots114). The forces in theinert zones308,309 anddeposition zone305 are variable depending on the position of thesubstrate unit74. At positions wherein the center ofgravity71 of thesubstrate unit74 is outboard of an inert zone center, the force component in the active zone is downward, as described above.
• In the present invention, static equilibrium of thesubstrate unit74 is accomplished when the sum of forces acting in the vertical direction is zero and the sum of the moments acting about a point is also zero. In equation form, these equilibrium conditions are given by:
• 
  Σ_z=0  (1a)
• 
  ΣM₀=0  (1b)
• where F_zis the z-component of the forces on the substrate unit74 (where the z-direction is normal to the output face134), and M₀are the moments acting around a specified point. It is convenient to compute the moments about thesubstrate unit74 centroid so that thesubstrate unit74 weight is not a factor. For the 2-D simplified example with an arbitrary pressure distribution for rectangular substrate having a weight per unit width, it can be shown that these equilibrium conditions yield the following equations:
• $\begin{array}{cc}\frac{W}{w}={\int }^{\mathrm{Substrate}}P\left(x\right)\mathrm{dx}& \left(2a\right)\\ 0={\int }^{\mathrm{Substrate}}P\left(x\right)·\left(x-x_0\right)\mathrm{dx}& \left(2b\right)\end{array}$
• where P(x) is the gauge pressure distribution above thedeposition head30 as a function of x (i.e., the position in the in-track direction), W is the weight of thesubstrate unit74, w is the width of the substrate in the cross-track direction (i.e., the y-direction), and x₀is the x-position of the center ofgravity71 of thesubstrate unit74.
• When theoutput face134 is discretized into regions, such asinert zones308,309 and adeposition zone305, the pressure distribution over those regions can be represented by equivalent force vectors according to the following integrals:
• $\begin{array}{cc}{F}_e=w{\int }_A^BP\left(x\right)\mathrm{dx}& \left(3a\right)\\ \mathrm{Xe}=\frac{{\int }_A^Bx·P\left(x\right)\mathrm{dx}}{{\int }_A^BP\left(x\right)\mathrm{dx}}& \left(3b\right)\end{array}$
• where F_eis the magnitude of an equivalent normal force vector, X_eis the x position at which the equivalent force vector acts, and the limits of integration A and B correspond to the x positions of the boundaries of the respective zones.
• For the example of a symmetric head having twooutput slots112 and oneexhaust slot114, the characteristic continuous pressure distribution and equivalent force vectors when the mass centroid offset is outside of the deposition head boundary are shown in thepressure distribution plot500 ofFIG. 8. (The position of thesubstrate unit74 relative to thedeposition head30 is the same as that illustrated inFIG. 7C.) For ease of understanding, there are dashed-vertical lines extending from the edges of thedeposition head30 to the corresponding pressure-distribution plot. The weight of thesubstrate unit74 is shown by the vector acting through the center ofgravity71 of thesubstrate unit74. Alternatively,FIG. 9 illustrates the piecewise uniform distributed loads for the same arrangement ofsubstrate unit74 anddeposition head30.
• Referring toFIG. 9, the vertical forces include the downward force of the weight of thesubstrate unit74, the upward lift from the inert zone (outboard of the last output channel), and the net force from the preloaded central-active zone. The forces in the inert zone and active zone are variable depending on the position of thesubstrate unit74. At positions where the center ofgravity71 of thesubstrate unit74 is outboard of the inert zone center, the force component in the active zone is downward. The preloaded active zone is capable of a finite maximum downward force based on the exhaust port area, the vacuum level, and to a degree the friction factor of the source slots. In the limit, the downward force can be only as high as 1 atmosphere times the area of the deposition zone.
• In some embodiments, the vacuum level is set by controlling a restriction (or orifice) to a vacuum source, and as such is not dynamically controlled. In these embodiments, the attractive force is set when the center ofgravity71 of thesubstrate unit74 is centered over thedeposition head30 and is adjusted to an optimal value for a static equilibrium preload. In other embodiments, the vacuum level may be dynamically controlled such that the attractive force is increased as the center ofgravity71 of thesubstrate unit74 is extended beyond theedge311 of thedeposition head30.
• The actual pressure distribution is determined by the design of thedeposition head30, and as such may be arbitrarily complex; an example deposition head with4 output channels and 3 exhaust channels is illustrated inFIG. 10 along with the corresponding pressure distribution plot. It is advantageous that the pressure in theoutermost output slots112, which is the source for the inert zone, be independently determined relative to the source pressures in the deposition zone.
• The maximum allowable displacement of thesubstrate97 relative to a vacuum-preloaded gas-bearingdeposition head30 is not obvious; prior art disclosures suggest that the limit is reached when the trail edge of the substrate uncovers an exhaust (vacuum) slot and thus compromises the vacuum at other commonly ported slots. The present invention operates at conditions where the center of gravity of thesubstrate unit74 is beyond the edge of thedeposition head30, allowing for greater coating area than prior art systems. When the center of gravity of thesubstrate unit74 is beyond the edge of thedeposition head30, the sum of forces and moments about the edge of thedeposition head30 for equilibrium is given by:
• $\begin{array}{cc}0={\stackrel{\_}{P}}_{\mathrm{inert}}·L_{\mathrm{inert}}+{\stackrel{\_}{P}}_{\mathrm{deposition}}·L_{\mathrm{deposition}}-\frac{W}{w}& \left(4a\right)\\ 0={\stackrel{\_}{P}}_{\mathrm{inert}}·L_{\mathrm{inert}}·\frac{L_{\mathrm{inert}}}{2}+{\stackrel{\_}{P}}_{\mathrm{deposition}}·L_{\mathrm{deposition}}·\frac{L_{\mathrm{Head}}}{2}-\frac{W}{w}·X_0& \left(4b\right)\end{array}$
• where L_head, L_inertand L_depositionare the lengths of the deposition head, the inert zone and the deposition zone, respectively, andP_inertandP_depositionare the average pressures in the inert and deposition zones, respectively.
• In order to maintain an equilibrium condition such that thesubstrate unit74 remains in a vacuum-preloaded condition with thedeposition head30, the sum of the moments must equal zero. If the center ofgravity71 of thesubstrate unit74 is extended beyond the distance from theedge311 of thedeposition head30 where the sum of the moments is zero, then an equilibrium can't be achieved and thesubstrate unit74 will detach and tip away from the deposition head. It should be understood from examination of Eqs. (4a)-(4b) that minimizing the weight of thesubstrate unit74 and maximizing the vacuum preload (vacuum at the exhaust channels), advantageously provides an increased range of motion of thesubstrate unit74, and therefor coating area. Embodiments of the present invention enable the coating of the maximum possible area on a piece-part substrate. In a preferred operating mode, the motion profile is configured such that the extent of substrate motion moves the center ofgravity71 of thesubstrate unit74 beyond the edge of thedeposition head30 during a portion of the substrate motion cycle. In embodiments of the present invention, the extent of motion is less than the tipping point, where thesubstrate97 would become detached from thedeposition head30. Preferred motion profiles limit the range of substrate motion so that the substrate edge is prevented from entering the deposition zone, ensuring that the deposition zone is bounded by the substrate at all positions in the motion profile. In preferred embodiments of the present invention, the motion profile is optimized for thesubstrate unit74, such that these limiting distances are designed to be the same so that neither is limiting before the other.
• FIGS. 11A-11C andFIGS. 12A-12C illustrate embodiments which include additionaloutboard support elements95 together with the vacuum-preloadeddeposition head30 in order to increase the in-track length of thesubstrate97 that can be coated without thesubstrate97 tipping. Theoutboard support elements95 are arranged such that they engage with thesubstrate97 and provide a support force before it reaches a position that would otherwise result in tipping, or detachment from the vacuum-preloadeddeposition head30. These additionaloutboard support elements95 are preferably placed on either side of thedeposition head30, in order to maximize possible substrate travel, so that its area may be best utilized without tipping. Theoutboard support elements95 can be mechanical elements such as shelves or rollers, as illustrated inFIGS. 11A-11C, placed to engage thesubstrate97 before tipping. In alternative embodiments, theoutboard support elements95 can be independent gas bearings as illustrated inFIGS. 12A-12C. Theoutboard support elements95 keep thesubstrate97 from tipping without interfering with the operation of thedeposition head30.
• It is also possible to counteract the tipping tendency by applying counter moments to the substrate as a function of the substrate position via elements connected to thebacker device70. Various mechanisms known in the art can be used to apply the counter moments including torsion springs, cantilever springs, coil springs, leaf springs, magnetic actuators, electromagnetic actuators, gas bellows and pneumatic cylinders. In one embodiment illustrated inFIGS. 13A-13C, avertical cantilever spring65 is attached to the top of thebacker device70. Thevertical cantilever65 engages stop pins66 which are located at an offset distance, providing a counter moment to the backer as thecantilever spring65 is deflected by further lateral motion of thesubstrate97. Alternatively, a rigidvertical element75 can be connected to thebacker device70 as shown inFIGS. 14A-14C. Thevertical element75 engages withsprings76 which are located at an offset distance on either side of thedeposition head30, providing a counter moment to thebacker device70 as thespring76 is compressed by thevertical element75.
• FIGS. 15A-15B, andFIGS. 16A-16B illustrate alternative embodiments in which the backer system includes one or more counter weights77 (having a center of gravity67) that move in a lateral direction opposite to thesubstrate97 to produce counter moments and postpone tipping. Thecounter weights77 can be moved by means such as connecting linkages and levers, or cables, or independent motion actuators. The embodiment ofFIGS. 15A-15B uses connected cables having a length between theframe285 of the substrate positioner and thecounter weight77.FIG. 15B illustrates the position of thesubstrate97 andcounter weight77 after movement from a starting position ofFIG. 15A. The embodiment ofFIGS. 16A-16B useslinkages87 between thebacker70 and thecounter weight77.FIG. 16B illustrates the position of thesubstrate97 andcounter weight77 after movement from a starting position ofFIG. 16A.
• In alternative embodiments, thedeposition head30 is oriented such that theoutput face134 of thedeposition head30 is in a vertical plane (with respect to gravity) and with the primary motion axis also in the vertical direction. For convenience, this orientation will be referred to as the vertical orientation. With thedeposition head30 in a vertical orientation, there are no longer concerns with tipping due to the force of gravity acting on anoverhung substrate97. As in the previously described horizontal orientation, the vacuum preload from the vacuum-preloaded gasbearing deposition head30 provides pitch and roll stability to thesubstrate unit74, and defines and maintains the z position of thesubstrate unit74 at the equilibrium fly height (head-to-substrate gap).
• In alternative embodiments, the SALD system of the present invention can utilize an additional compensating weight on the back of thesubstrate97. In these alternative embodiments, the compensating weight will impart a downward force vector toward theoutput face134. The force will preferably be within the area of thedeposition head30 defined by theoutermost output slots112, more preferably will be centered on thedeposition head30. The force imparted by the compensating weight enables thesubstrate97 to be levitated given a sufficient positive pressure vector and to stay loaded to maintain a constant and small gap between the substrate and thedeposition head30. Furthermore, in some embodiments, the compensating weight can be constrained to remain over thedeposition head30, enabling thesubstrate97 to be moved over thedeposition head30, and for the center ofmass71 to be moved beyond the edge of the deposition zone. (For example, see the dual gas-bearing system described in commonly-assigned, co-filed U.S. patent application Ser. No. ______ (Docket K002128) to Spath, entitled “Dual gas bearing substrate positioning system,” which is incorporated herein by reference.)
• A vertically oriented SALD system for coating long substrates can be better understood with respect toFIGS. 17A-17C. (Note that the axes have been rotated so that the orientation relative to theoutput face134 of thedeposition head30 is the same as it had been in the horizontal orientation.) As illustrated inFIG. 17A, thesubstrate unit74 includes asubstrate97 and a rigidly attachedbacker device70. Thesubstrate positioner288 engages withbacker device70 at contact location(s) along the line of action of the gravity vector from the center ofgravity71 of the substrate unit74 (the combinedsubstrate97 and backer device70). This is consistent with horizontally oriented systems, where it is preferred that the contact of thesubstrate positioner288 and the center of gravity of thesubstrate unit74 be in the plane of the center ofgravity71 to minimize acceleration effects on pitching moment. In the vertically oriented system, this contact location of thesubstrate positioner288 prevents pitching moments due to gravitational acceleration as well as acceleration of thesubstrate positioner288. In practice,substrates97 to be coated using vertically-oriented SALD system can be composed of a wide range of materials with a range of density properties, and a range of sizes (width, length and thickness). To enable the vertically-oriented SALD system to compensate for variations in the substrate properties, anadjustable counter mass78 is provided on thebacker unit70 which allows the position of the net center ofgravity71 to be maintained in line with the pivot axis, as illustrated inFIGS. 17A-17C. As illustrated, thecounter mass78 can be a simple threaded cylinder arranged on a stud; any other common means which provide anadjustable counter mass78 are also within the scope of the present invention.
• In this vertical configuration, as thesubstrate unit74 is moved by thesubstrate positioner288, the center ofgravity71 of thesubstrate unit74 can be moved outside of the range of the deposition head without impacting the gap between theoutput face134 and the process-side of thesubstrate97 as illustrated inFIG. 17B. As seen, the center ofgravity71 of thesubstrate unit74 is directly above the lower contact with thesubstrate positioner288, and has been moved to be above thedeposition head30, outside of the area of thesubstrate97 which is interacting with thedeposition head30. Thesubstrate unit74 is stabilized by the vacuum-preloadeddeposition head30. Furthermore, the vacuum preload from thedeposition head30 provides pitch and roll stability to thesubstrate unit74, and also defines and maintains the z-position of thesubstrate unit74.
• To allow freedom for thesubstrate unit74 to equilibrate in pitch, it is preferred that the contact betweenbacker device70 andsubstrate positioner288 be confined to a line, which allows for unencumbered pitch rotation. This “line” may be defined by two discrete points, according to basic geometry. The two point (minimum) contact between thesubstrate positioner288 and thesubstrate unit74 desirably constrains the yaw orientation of thesubstrate unit74.
• In an alternative embodiment illustrated inFIG. 17C, the line contact is provided by aroller281 incorporated into thesubstrate positioner288. Alternatively, theroller281 could be incorporated into thebacker unit70 of thesubstrate unit74. In preferred embodiments where the line contact is provided by aroller281, there is greater freedom for thesubstrate unit74 to equilibrate in pitch and z. In more preferred embodiments, the line contact is defined by two independent roller segments that are collinear but free to rotate individually, advantageously allowing thesubstrate unit74 to have the desirable freedom to seek equilibrium in roll, in addition to pitch and yaw.
• In an alternative embodiment illustrated inFIGS. 18A-18B, a single pair ofroller bearings282 are provided on thebacker device70 which engage with asubstrate positioner288 having mating yoke features283. The roller bearings rotate around an axis extending in the cross-track (y) direction. The yoke features283 are oriented to enable theroller bearings282 to freely move in the direction normal to theoutput face134 of the deposition head30 (i.e., the y-direction). This configuration provides the desired constraints and degrees of freedom with a preferred engagement of thesubstrate positioner288 with the relative motion means (primary motion actuator). As illustrated, the depth of theyoke feature283 is sufficient to provide clearance to prevent bottoming of the rollers. In the embodiment illustrated inFIG. 18A, the roller axis is arranged to pass through the center ofgravity71 of thesubstrate unit74 and provides neutral pitch stability. Alternatively, the roller axis can be arranged to pass above the center ofgravity71 of thesubstrate unit74 and providing positive pitch stability. This configuration of the roller axis provides the advantage that thesubstrate unit74 will not tip, even if vacuum preload provided by thedeposition head30 is interrupted. The open yoke features283 on thesubstrate positioner288 advantageously provide for easy removal and loading ofsubstrates97 over the output face of thedeposition head30.
• In vertically oriented SALD systems, thesubstrate positioner288 bears a load component due to the weight of thesubstrate unit74. Therefore, it is desirable that the contact (interface) between the substrate unit74 (load) and thesubstrate positioner288 be as low friction as possible in the direction normal to theoutput face134 of the deposition head30 (i.e., the z-axis) so as to avoid interference with the vacuum-preloaded gas bearing functionality of thedeposition head30. Low friction bearing devices known to one skilled in precision mechanical design can be employed in accordance with the present invention to minimize this friction component. Exemplary embodiments include one or more ceramic hybrid ball bearings, such as of silicon nitride or similar balls, and high speed steel races. These are particularly low friction and tolerant of elevated temperatures, and will not artificially restrict the process temperatures of the SALD system (e.g., up to 350° C.). In addition to the rolling element bearings discussed above, conventional independent gas bearings acting on a face perpendicular to the output face of the deposition head can be employed in alternative embodiments. The stability of thesubstrate unit74 is provided by the vacuum-preloadeddeposition head30, therefore, these additional bearings do not need to provide orientation stability other than yaw, which is readily accomplished by two “point source” bearings at a suitable distance apart.
• The functional requirement of the support element on thesubstrate positioner288 is that it can constrain three degrees of freedom of motion of thesubstrate unit74, including one rotation and two translations. In some embodiments of the present invention, such as the exemplary configuration shown inFIGS. 19A-19B andFIGS. 20A-20B, the support elements of the substrate positioner areflexures284, which provide a zero friction means of support for thesubstrate unit74 to thesubstrate positioner288. Theflexures284 can comprise sheet or wire elements. The flexures can be provided by a single planar sheet or two parallel sheets joined by a common rigid member, as shown inFIGS. 19A-19B. In the illustrated configuration, the attachment of the “free end” of theflexures284 to thesubstrate unit74 is by means of bosses having surfaces aligned with the axis of the center ofgravity71 of thesubstrate unit74. The plane of theflexures284 is parallel to theoutput face134 of thedeposition head30. The “fixed” end of theflexures284 is attached to a bracket on thesubstrate positioner288. Thesubstrate positioner288 is attached to the relative motion means270 (i.e., the primary motion actuator).
• Alternative embodiments providing equivalent constraints utilize threeflexures284, where two of theflexures284 are located in the plane of the center ofgravity71 of thesubstrate unit74 and oriented parallel to theoutput face134 of thedeposition head30. Theseflexures284 constrain motion in x and yaw. A third flexure284 (not shown) is arranged in the direction parallel to theoutput face134 of thedeposition head30 and perpendicular to the primary motion axis (i.e., the x-axis), intersecting the center ofgravity71 of thesubstrate unit74 in a plane parallel to thefirst flexures284, and is preferably in the same plane as the first two wire flexures. Thisflexure284 constrains the payload from translation in y. The necessary degrees of freedom to allow equilibration of thesubstrate97 to theoutput face134 of thedeposition head30, including pitch, roll, and translation in z, are advantageously preserved in this alternative embodiment.
• In some embodiments of the SALD system of the present invention, thebacker device70 provides heat to thesubstrate97. In these embodiments, in addition to supporting thesubstrate97 by either vacuum attachment or gas bearing floatation, theflexures284 between thesubstrate unit74 and thesubstrate positioner288 can be advantageously utilized to deliver electrical energy or fluid flows. In one embodiment, two parallelsheet metal flexures284 are utilized to provide electrical current to heater elements or thermo-electric Peltier modules incorporated in thebacker device70. Suitable conductive and insulating materials are used as necessary to define the current flow. In some embodiments electrically resistive properties of a porous membrane utilized in thebacker device70 can be used as a heating element.
• In another embodiment, two of the wire flexures are used to source and sink electrical energy while the third element is tubular and is used to convey vacuum or air to the face of thebacker device70 that faces thesubstrate97. The principle mechanical load is a result of acceleration in the primary motion direction, so the solid metallic wires are preferably oriented in that direction. The tubular element may be soft tubing, such as silicone, because the forces in the y direction are minor and y position is not critical. The straight path of the gas supply tube avoids disturbing forces caused by Bourdon tube effects (i.e., forces due to pressure applied to unequal areas on inside on outside of bends).
• The electrical energy conveyed to thebacker device70 can be modulated as a means to communicate process conditions to or from the backer to the system controller by means known to one skilled in electronics. Properties and process conditions that can be communicated include temperature, pressure, acceleration, sample presence, and amount of thin-film growth. Thebacker device70 can include sensors and signal conditioning electronics in such embodiments. In some embodiments, communication can be accomplished by means of RF or optical links. While, as described, theflexures284 are clear implementations of the necessary functionality, they are not exhaustive of the possible arrangements that could provide similar functionality which are also included within the scope of the present invention.
• In configurations such as that shown inFIGS. 7A-7C, the weight of the backer unit including thebacker device70 can provide a useful contribution to preload thesubstrate97 anddeposition head30 when theoutput face134 of thedeposition head30 is in a horizontal plane. This component of preload is particularly useful for enhancing safety in the event of loss of exhaust vacuum, and thus loss of vacuum preload in thedeposition head30. This weight component may also be a critical factor in a proven deposition head design and operating condition. When theoutput face134 of thedeposition head30 is oriented in a vertical plane as in the configuration ofFIGS. 19A-19B, the weight of thesubstrate unit74 no longer contributes to preload. When desired, an equivalent preloading force can be introduced in the horizontal direction by means of additional hardware.
• The apparatus illustrated inFIGS. 20A-20B illustrate exemplary means for providing a substantially constant horizontal force which is independent of substrate thickness variation, and without introducing any disturbing moments to thesubstrate unit74. The weight of amass60 is redirected by alever61 which pivots aroundpivot point62 and applied to a bearing which in turn conveys the force to thesubstrate unit74. Thepivot point62 oflever61 is fixed with respect to machine ground. The line of action of the force imparted to thesubstrate unit74 passes through the center of theoutput face134 of thedeposition head30. InFIG. 20A, the bearing is aconventional gas bearing63, such a porous graphite air bearing, having an axi-symmetric design and a spherical mounting seat so that its alignment conforms to any imperfection in planarity or parallelism of the outside face of thesubstrate unit74. As shown inFIG. 20B, the bearing that interfaces to thesubstrate unit74 can also be a rolling element bearing64, such as a ball bearing, whose axis is in a plane parallel to theoutput face134 of thedeposition head30 and aligned in the transverse direction. The outer race of the rolling element bearing64 may have a spherical contour such that point contact between the outer race and the outside face of the substrate unit does not constrain roll of thesubstrate unit74 about the primary motion axis. The rolling element bearing may advantageously be constructed of high temperature compatible materials, such as ceramic balls and high speed steel races so that thesubstrate unit74 may be operated at temperatures up to 350° C.
• The use of a fixed weight and a lever provides an easy to understand illustration of a means to provide a constant horizontal force to thesubstrate unit74 without introducing problematic forces or moments that would disturb the ability of thesubstrate unit74 to be freely guided by the preloaded gas bearing interface of thedeposition head30. It is understood that other means of imparting a horizontal force component thesubstrate unit74 can be employed, such as described in the context of other embodiments and configurations of the present invention and are within the scope of the invention (e.g., mechanisms employing magnetic attraction or repulsion, or pneumatic bellows). Linear bearings and/or flexures can be used to (re)direct the applied force in the desired direction normal to theoutput face134 of thedeposition head30. Preferably, the force producing apparatus is advantageously designed to be easily disengaged to facilitate loading and unloading of thesubstrate97.
• As mentioned earlier, one of the advantages of the present invention is the ability tocoat substrates97 that are much longer than the length of the deposition head, including continuous substrates in a roll-to-roll mode. In preferred embodiments of the present invention theexternal environment15 is not required to be strictly controlled, and as such the motion of thesubstrate97 during deposition is controlled so as to protect the “active growth region” from entering theexternal environment15.FIG. 21 illustrates the limits of substrate motion when subjected to the constraint of controlled environment, using the simple one-cycle ALD deposition head discussed with respect toFIGS. 2A-2C. For clarity, inFIG. 21 thedeposition head30 is not shown, and rather focuses on the gas zones formed between thesubstrate97 and the deposition head30 (which are common to all head embodiments and geometries).FIG. 21 illustrates thesubstrate97 at three different positions in the motion profile. AtPosition #1, thesubstrate97 is centered with respect to thedeposition zone305. UsingPosition #1 as a starting place, four positions have been marked on the substrate—a diamond corresponding to the outer edge of the firstinert zone308, a hexagon for the outer edge of the firstreactive gas zone313, a triangle for the outer edge of the last reactive gas zone (i.e., second reactive gas zone315), and a circle for the outer edge of the secondinert zone309. The region ofsubstrate97 in thedeposition zone305 can be considered as an “active growth region”. In order to keep the hexagonal point in a controlled environment once it has experienced an exposure to the first precursor (i.e., the first reactive gas) it must be controlled to remain within the firstinert zone308, as illustrated byPosition #2 inFIG. 21. Similarly, to keep the triangle point in a controlled environment after it has experienced exposure to the last precursor (i.e., the second reactive gas) it must be controlled to remain within the secondinert zone309, as illustrated byPosition #3 inFIG. 21. When the substrate is moved in a simple oscillating, or reciprocating, motion profile with equal forward and backward motion, the area of the substrate that can be coated while controlling the environment of the active coating area is limited by the size of theinert zones308,309. The motion profiles of the present invention, which provide the ability for longer coatings while preventing the active growth area ofsubstrates97 from experiencing the external environment during deposition, are constrained as a function of the lengths of theinert zones308,309. InFIG. 21, L_I1is the length of the firstinert zone308, L_G1is the length of the firstreactive gas zone313, L_Pis the length of thepurge zone314, L_G2is the length of the secondreactive gas zone315, and L_I2is the length of the secondinert zone309. Regardless of the number of gas zones within the deposition zone, there will be afirst reactant zone316 adjacent to the firstinert zone308 and alast reactant zone317 adjacent to the secondinert zone309.
• In order to “see” a full ALD cycle a point on thesubstrate97 must be exposed to each of the reactive gases. By examination ofFIG. 22, one can see that the distance required for a full cycle is slightly longer than the distance of thepurge zone314 such that the point will move from a portion of thefirst reactant zone313 containing the first reactive precursor G1, through thepurge zone314 and into a portion of thesecond reactant zone315, containing the second reactive precursor G2 thereby experiencing the sequential gas exposure of G1-P-G2. Similarly, in the reverse, or backward, direction a point on thesubstrate97 will experience exposure to the gases in the deposition zone in the order of G2-P-G1. “Position #2” ofFIG. 22 illustrates this minimum required movement where thesubstrate97 has been moved a distance of L_Pto the right from the starting position (i.e., “Position #1”). The point on thesubstrate97 indicated by a black triangle has moved from the edge of the firstreactive gas zone313 to the edge of the secondreactive gas zone315, and would therefore see the two necessary reactants. However, it can also be seen that the point on thesubstrate97 indicated by the lighter triangle while having a full exposure to G1, will not experience the G2 precursor; and in a simple oscillatory motion, no growth would occur at this point. In motion profiles with continuous forward motion, there would similarly be no further growth at the solid dark triangle point.
• “Position #3” ofFIG. 22 illustrates the position of thesubstrate97 after it has been moved a distance of L_P+L_G1to the right from the starting position (i.e., “Position #1”). The points on thesubstrate97 indicated by both the dark and light triangle markers have moved from exposure in the G1 region to a position in G2 region, and would see the two necessary reactants. This distance of L_P+L_G1represents the minimum movement (in either direction) of sufficient interest for simple oscillatory motion, and for movement in a motion profile with continuous forward motion. As shown, the region of thesubstrate97 between the two triangle markers will be exposed to bothreactive gas zone313,315. However, it can also be seen that the lighter triangle while having a full exposure to G1, will experience a minor amount of the G2 precursor. Therefore, in some embodiments it is desirable that the minimum motion be greater than L_P+L_G1. “Position #4” illustrates a preferred minimum motion requirement of the present invention, showing substrate after it has been moved a distance of L_G1+L_P+L_G2to the right from the starting position (i.e., “Position #1”), which is equal to the length of thedeposition zone305. In all cases, in order to maintain control of the growth region environment, the length of theinert zones308,309 (L_I1and L_I2) must be greater than or equal to the longest length of movement in any motion profile.
• In preferred embodiments, the motion profile of the present invention is used with deposition units having multiple ALD cycles; while increasing the number of cycles is useful, there is a relationship between the number of cycles in adeposition head30 and the physical size of thedeposition head30, which will limit the minimum substrate size that can be coated and impact the cost of the system. The size and cycle count of thedeposition head30 can be optimized for a given application. The deposition head illustrated inFIG. 23 has adeposition zone305 with symmetry containing 2.5 ALD cycles, and operates in a similar manner to the deposition heads ofFIG. 2C andFIG. 3A and should be understood from the previous description. As illustrated, thefirst reactant zone316 and thelast reactant zone317 of the deposition zone both contain the first reactive gas G1. As shown inFIG. 23, the gases are introduced and exhausted in alternating output andexhaust channels112,114 in theoutput face134 of a vacuum-preloaded gasbearing deposition head30. For the purposes of useful motion profiles for SALD deposition, the arrangement of the gas zones between thedeposition head30 and the process side of thesubstrate97 is the relevant feature. However, vacuum-preloaded gas bearing designs are preferreddeposition head30 embodiments for reasons discussed earlier. Thedeposition head30 illustrated inFIG. 23 illustrates a number of desirable aspects of exemplary embodiments of the present invention, firstly the two extendedinert zones308,309 have the same length, the gas regions have symmetry, each of thepurge zones314 within thedeposition zone305 have equivalent lengths, and finally the firstreactive gas zones313 and the secondreactive gas zones315 each have the same length. In preferred embodiments, all of the various process gas zones (i.e., the firstreactive gas zones313, thepurge zones314, and the second reactive gas zones315) within thedeposition zone305 each have the same length.
• InFIG. 23, theinert zones308 and309 are shown with asingle output slot112 which supplies the inert gaseous materials. In other embodiments (not shown), theinert zones308 and309 can include a plurality ofoutput slots112 through which the inert gaseous materials are supplied, and optionally can also includeadditional exhaust slots114. In some configurations, this approach can be used to extend the vacuum-preloading region.
• FIG. 24 illustrates an alternative embodiment withinert zones308,309 having different lengths (i.e., L_I1≠L_I2). All other aspects are the same as the gas zones ofFIG. 23. “Case A” illustrates the movement limitation on the motion profile when portions of thesubstrate97 that have been exposed to the reactive gas in thefirst reactant zone316 and the last reactant zone317 (in this example, the first reactive gas G1) cannot be allowed to travel outside of theinert zones308,309 during the deposition process. In this case, as withFIG. 21, the maximum movement of the motion profile during deposition is the minimum length of the either inert zone: Min(L_I1, L_I2).
• “Case B” illustrates the movement limitation on the motion profile when the reactive gas in thefirst reactant zone316 and the last reactant zone317 (in this example, the first reactive gas G1) does not react with theexternal environment15 such that portions of thesubstrate97 that have been exposed to G1 can be allowed to travel outside of theinert zones308,309 during the deposition process. For example, in instances where G1 is a water vapor, passage of a point on thesubstrate97 that has been exposed to G1 into a relatively clean external environment will not cause any adverse surface reactions. That point will traverse the inert zone and be “re-exposed” to G1 prior to having a next exposure to G2. In this case, the first exposure to G2 sets the initiation point of the active deposition region such that the maximum movement of the motion profile during deposition is the minimum length from the edge of the deposition head to the outer edge of the second reactive gas zone315: Min(L_I1+L_G1+L_P, L_P+L_G1+L_I2). In general, the maximum movement of the motion profile will be limited by the minimum distance between a reactant zone having a reactive gas that will react with theexternal environment15 and the outer edge of the inert zone.
• From the descriptive discussion of SALD above, it should be clear that the reactive gasses G1, G2 need to be isolated from each other (i.e., by the inert purge gas flows), and contained between theoutput face134 of thedeposition head30 and thesubstrate97 to prevent them from reacting with each other away from the substrate surface. The reactive gasses G1, G2 also generally need to isolated from the external environment to maintain purity and avoid unwanted reactions. Lastly, the surface of thesubstrate97 that is reacting with the reactive gasses to build up a deposited layer should be isolated from theexternal environment15 during deposition to avoid contamination, as discussed above. One of the advantages of the current invention is that it provides a motion profile that enables the coating oflong substrates97 using asmall deposition head30 and simultaneously prevents the growth region of thesubstrate97 from being exposed to theexternal environment15 until the completion of the desired coating thickness, or equivalently until the desired number of ALD cycles has been achieved.
• The previous discussion of the motion profile has focused on a simple reciprocating, or oscillatory, motion profile, where the center ofgravity71 of thesubstrate unit74 is moved back and forth an equal distance relative to the center of adeposition head30. This motion profile limits the region of uniform coating that can be achieved with asmall deposition head30. There is a need for a motion profile that can provide a net forward motion in the in-track direction (i.e., the x-direction). Preferred embodiments of the present invention use a motion profile that enables a large coating area while preventing atmospheric (external environment15) exposure during growth. An exemplary motion profile for use in the present invention uses a repeating “ooching” motion, wherein the forward in-track motion is incrementally larger than the reverse, or backward, in-track motion resulting in a net-forward motion of the substrate during each cycle of the motion profile. In order for thesubstrate97 to have a net-forward motion a repeating motion profile of the present invention uses a forward motion portion “f” which is longer than a backward motion portion “b”. The difference between the forward motion portion and the backward motion portion will herein be referred to as the “ooch distance” given by Δx=f−b.
• FIG. 25 illustrates aspects of an exemplary motion profile of the present invention that enable the dual benefits of long coating and protected coating region. The gas zones illustrated inFIG. 25 are those that would be generated fromdeposition head30 ofFIG. 23 having symmetry, includinginert zones308,309 of equal length, and should be understood from the previous description. The repeating motion profile is provided by repeating amotion profile oscillation150 that includes a forward motion portion f and a backward motion portion b, providing a net forward motion of Δx. In this example, the substrate surface after exposure to the first reactive gas G1 is not reactive with theexternal environment15, and therefore the movement limitation is set to the broader limitation discussed earlier such that f≤L_I+L_G1+L_P. To ensure that each point on thesubstrate97 receives a full exposure of both reactive gases in both the forward and backward directions, the minimum backward motion should be greater than the sum of the lengths of a firstreactive gas zone313 and apurge zone314 such that b≥L_P+L_G1. In themotion profile oscillation150 ofFIG. 25, the backward motion is longer than the sum of the lengths of the firstreactive gas zone313, apurge zone314 and a secondreactive gas zone315 such that b≥L_G2+L_P+L_G1. Given the repeating nature of the motion profile, subsequentmotion profile oscillations150 are initiated at the end of the backward motion portion of the previousmotion profile oscillation150.
• A repeatingmotion profile165 including a sequence of threemotion profile oscillations150 is shown for illustration. A point on thesubstrate97 that starts at the position of the tail of the forward motion vector in the firstmotion profile oscillation150 will first pass through a firstreactive gas zone313 where it will be exposed to the first reactive gas G1 providing a 1^st“half reaction” H1. After passing through apurge zone314, it will then pass through a secondreactive gas zone315 where it will be exposed to the second reactive gas G2 providing a 2^ndhalf reaction H2. Then, after passing through anotherpurge zone314, it will then enter another firstreactive gas zone315 where it will be exposed again to the first reactive gas G1 providing a 3^rdhalf reaction H3. Thesubstrate97 will then change directions and move backwards for the backward motion portion. The initial portion of the backward motion portion will still be a part of the 3^rdhalf reaction H3 since the point never leaves that gas zone. During the backward motion portion that point on thesubstrate97 will then experience a 4^thhalf reaction H4 and a 5^thhalf reaction H5. At the end of the illustrated three repeat profile segments, it can be seen that the point on thesubstrate97 will have undergone 12 half reactions (H1-H12). For a two reactive precursor ALD system two half reactions equals one ALD cycle, and therefore thesubstrate97 will have experienced six ALD cycles.
• FIG. 26 illustrates the movement of a single point on thesubstrate97 moving using a repeating motion profile with themotion profile oscillation150 ofFIG. 25. As illustrated inFIG. 26, the “gray” arrows indicate movement prior to and after the time when active growth (i.e., deposition) is occurring, and the “black” arrows indicate movement during the time when active growth is occurring. As can be seen, during deposition the environment experienced by the active growth region is limited to that which is provided by thedeposition head30 using the inventive motion profile such that the active growth region is not exposed to theexternal environment15. The total number of ALD cycles experienced by a point on thesubstrate97 can be controlled by using an appropriate combination of the forward motion distance f and the ooch distance Δx for a given head geometry. For an appreciable coating thickness, it is necessary that the substrate experience multiple ALD cycles. Typical ALD growth rates range from sub-angstrom to 3 angstroms per cycle, therefor multiple cycles are needed for even nanometer thicknesses of material. In order to manage the overall size of thedeposition head30, the ooch distance Δx is typically quite small, and therefore the forward motion distance f and the backward motion distance b are nearly equivalent. In preferred embodiments Δx is less than the length of any single process gas zone within the deposition zone305 (i.e., the length L_G1of the firstreactive gas zones313, the length L_Pof thepurge zones314, and the length L_G2of the second reactive gas zones315) as shown inFIGS. 25 and 26. Any repeating motion profile where f is greater than b, and b is greater than L_P, will have growth as the substrate travels in a net in-track direction and is within the scope of the present invention.
• The motion profile of the current invention advantageously prevents an active growth region of thesubstrate97 from being exposed to theexternal environment15 prior to the completion of the desired deposition. In general, to prevent exposure to the external environment during the forward motion portion, the forward distance f of the repeating motion profile should be less than the sum of the lengths of the secondinert zone309, thelast reactant zone317 and thepurge zone314. Likewise, to prevent exposure to the external environment during the backward motion portion, the backward distance b of the repeating motion profile should be less than the sum of the lengths of the firstinert zone308, thefirst reactant zone316 and thepurge zone314. In the more restrictive case where the gas in thelast reactant zone317 will react with the external environment, the forward distance f of the repeating motion profile should be less than the length of the secondinert zone309. Likewise, if the gas in thefirst reactant zone316 will react with theexternal environment15, the backward distance b of the repeating motion profile should be less than the length of the firstinert zone308.
• FIG. 27 illustrates the movement of a single point on the substrate moving using a repeating motion profile withmotion profile oscillations150 which satisfies the more restrictive conditions appropriate for cases where the gases in thefirst reactant zone316 and the last reactant zone will react with theexternal environment15. In this figure, the light gray arrows indicatemotion profile oscillations150 where the point on thesubstrate97 is not exposed to either of the reactive gases G1, G2; the dark gray arrows indicatemotion profile oscillations150 where the point on the substrate is exposed to only the first reactive gas G1, and the black arrows correspond tomotion profile oscillations150 where the point on the substrate is exposed to both of the reactive gases G1, G2 such that ALD growth is actively occurring. As can be seen, during deposition the environment experienced by the active growth region is limited to that which is provided by thedeposition head30 using the inventive motion profile.
• As thesubstrate97 moves into thedeposition zone305 it may experience multiple exposures to thefirst reactant zone316, however without exposure to a secondreactive gas zone315 no ALD cycle growth will occur. Once thesubstrate97 has progressed into thedeposition zone305 to a point where the growth region of interest has entered the left-most secondreactive gas zone315, signally the start of ALD cycle growth, the substrate motion is controlled such that the growth region is prevented from moving beyond theinert zones308,309 and into theexternal environment15. After the point of the substrate has exited the right-most secondreactive gas zone315 for the final time, it is no longer necessary to make sure that it does not move beyond the secondinert zone309 and into theexternal environment15 since ALD growth will no longer occur.
• In accordance with preferred embodiments, the motion profile of the present invention provides a net-forward motion, with the forward motion portion being longer than the backward motion portion. Since the backward distance is the smaller, or limiting travel distance, the substrate must move backward by an amount sufficient to experience an ALD cycle. As described above, the backward motion distance of the motion profile of the present invention is therefore greater than the purge zone length (L_P), and more preferably is greater than the sum of the lengths of thepurge zone314, a first reactive gas zone313 (L_P+L_G1) or the sum of the lengths of thepurge zone314, a second reactive gas zone315 (L_P+L_G2), whichever is larger.
• The repeatingmotion profile165 of the present invention is used tocoat substrates97 whose length is greater than the length of adeposition unit30, wherein thedeposition unit30 has extended endinert zones308,309 and acentral deposition zone305. As described, preferred embodiments provide the gas zones using asingle deposition head30. In alternative embodiments, multiple heads maybe be arranged in a deposition unit to provide the necessary gas zone, as described in commonly-assigned, co-pending U.S. patent application ______ (Docket K002131), which is incorporated herein by reference, so long as they provide a controlled environment with the previously described gas zones between the surface of the deposition unit and the process side of thesubstrate97 during active growth. The maximum size ofsubstrate97 that can be coated using the repeatingmotion profile165 of the present invention is only limited by the substrate handling capability of the overall deposition system, and not by either the deposition unit or the profile. As such, the repeating motion profile of the present invention is advantaged for depositing on continuous web substrates using systems such as those described in commonly-assigned, co-pending U.S. patent application Ser. No. ______ (Docket K002129), which is incorporated herein by reference.
• Due to the cyclic-nature of ALD, there will be “lead-in” and “trailing” sections of growth on thesubstrate97 on either side of a region of uniform thickness. The lead-in and trailing sections are a result of the requirement that thedeposition zone305 be covered by a substrate during the flow of reactive gases, and as such the points on the substrate in the lead-in (or trailing) region will begin (or end) the motion profile at different points in the ALD cycle and therefore receive varying amounts of deposition. In preferred embodiments of the present invention, the length of the substrate is sufficient to achieve the desired uniform coating region.
• Since increasing the number of ALD cycles (and thereby increasing the layer thickness) can be easily obtained by decreasing the ooch distance Δx of the repeatingmotion profile165, SALD deposition heads with a fewer number of cycles andshort deposition zones305 are viable alternatives to those withlong deposition zones305 having many cycles, advantageously serving to minimize both the footprint of the deposition system (and the deposition head30) and the substrate waste during operation. Deposition heads30 that provide gas zones having symmetry are most preferred. Deposition heads30 havinginert gas zones308,309 that are substantially equal in length to thedeposition zone305 as illustrated inFIG. 28 have some advantages in the present invention. By increasing the length of theinert zones308,309, the maximum allowable forward distance f of the repeatingmotion profile165 is increased allowing for exposure to more ALD cycles during a single forward motion. Since reversing direction inherently requires the motion controller to slow down, stop, reverse direction and accelerate, minimizing the number of turns will increase the overall deposition rate out of the deposition system. However,inert zones308,309 that are longer than thedeposition zone305 do not result in the active growth region being exposed to additional reactive gases with increasing forward distance f, and come at a penalty of increased deposition system size. In some exemplary embodiments, the lengths of theinert gas zones308,309 are greater than or equal to half the length of thedeposition zone305 and less than or equal to the length to the deposition zone305 (i.e., L_D/2≤L_I2)≤L_D). In the case where high temperature deposition is required, longerinert gas zones308,309 can be used to insure that thesubstrate97 comes to thermal equilibrium before it reaches the active deposition region.
• The previous illustrations described embodiments wherein the variousreactive gas zones313,315 and purgezones314 had the same length. This is not a requirement of the current invention, and in some embodiments the lengths of thepurge zones314 andreactive gas zones313,315 can be different. For example, in some cases it may be desirable to have the length of thepurge zones314 be longer than the length of thereactive gas zones313,315. An exemplary configuration where thepurge zones314, the firstreactive gas zones313, and the secondreactive gas zone315 have different lengths is illustrated inFIG. 29. In this case, the motion limitations will be equivalent to those discussed earlier with respect toFIG. 24. Other configurations are possible, and are only limited by fabrication and practical constraints and are considered part of the present invention as long as they provide continuous forward motion using a repeating motion profile, while keeping the active growth region protected during the growth event. For example,deposition zones305 including two ALD cycles such that thelast reactant zone317 is a secondreactive gas zone315. In this case, the motion limitations will be somewhat different for the case where only the second reactive gas G2 reacts with the external environment. In this case, the maximum motion will be limited by: Min(L_I1+L_G1+L_P, L_I2).
• It is useful to consider how one would use the repeating motion profile of the present invention to achieve a desired number of ALD cycles, or a given coating thickness. As should be apparent from the earlier figures, during the motion profile any point on thesubstrate97 can spend a variable amount of time in any given reactant zone. The necessary values of the forward motion distance f, the backward motion distance b and the ooch distance Δx can be determined experimentally or calculated, such as by using the relative kinetics of the half reactions and the “dwell time” in any gas zone. As the growth during different portions of the motion profile may deviate from conditions that yield saturated growth, the number of half reaction gas exposures necessary will increase and therefore the associated ooch distance will decrease in order to achieve the desired layer thickness. Therefore, using a fully saturated ALD approximation to calculate the necessary number of repeats (oscillations) within the deposition zone to achieve the necessary number of half-cycles, as well as the associated ooch distance will provide an approximation of the useful operating conditions.
• The following example illustrates how the growth process can be modeled as a function of the system parameters. Consider a point on the surface of thesubstrate97 which remains within thedeposition zone305 during amotion profile oscillation150. At any time, the point can be said to have a surface state according to the last reactive gas to which the point was exposed. For anSALD deposition head30 having output slots for two reactive gasses G1 and G2, the point on the surface can be in two states. The “original” state, for this discussion can be considered to correspond to the first reactive gas (e.g., G1) that the point was exposed to on its initial entry into thedeposition zone305. Subsequent exposure of the point on the surface to the second reactive gas (e.g., G2) changes the surface state from its original state. A subsequent exposure to first reactive gas causes the surface state to return to its “original” surface state. A “growth” event is counted each time the surface state is returned to its “original” state. Exposure to a purge gas (e.g., P), an inert gas (e.g., I1 or I2), or atmosphere in the case of non-atmospherically sensitive reactants, does not cause a change to the surface state in either direction. Repeat exposure to the last seen reactive gas, even with an exposure to a non-reactive gas in between, does not cause a change in surface state.
• Consider the example illustrated inFIG. 30 of a symmetric 2.5cycle deposition head30 having adeposition zone305 includingrepeat units318 of an ALD cycle (where eachrepeat unit318 includes a firstreactive gas zone313, apurge gas zone314, a secondreactive gas zone315 and another purge gas zone314) The repeat units are followed by a terminating ½ cycle having a final firstreactive gas zone313 and a terminating purge gas zone (which is coincident aninert zone309 in the illustrated embodiment). In the illustrated example, the forward motion distance f is set to be equal to the length L_Rof a single repeat unit318 (i.e., f=L_R=L_G1+L_P+L_G2+L_P). In preferred embodiments, the ooch distance Δx=f−b is small relative to the forward and backward motion distance (i.e., Δx<<f, b) so that b≈f. The number of growth events completed during a single oscillatory motion of the point on the surface depends on the zones in which the turnaround points of themotion profile oscillation150 occur. When f=L_R, the turnaround points in a particularmotion profile oscillation150 will always occur in the same gas zone type (i.e., each turn around occurs in the same gas composition). In a preferred embodiment, the length L_Iof theinert zones308,309 is greater than or equal to the length L_Rof therepeat unit318, and more preferably is at least 125% of the length L_Rof therepeat unit318. In some exemplary configurations, the length L_Iof theinert zones308,309 is substantially equal to an integer multiple of the length L_Rof therepeat unit318. Within the context of the present invention, “substantially equal” means equal to within 5%.
• By examination ofFIG. 30, it can be seen that when the turnaround points are within areactive gas zone313,315 (i.e., within a G1 or G2 zone), the point on the surface will experience two growth events per oscillation (e.g., if the point starts in a firstreactive gas zone313, it will start in a G1 state, then will go to a G2 state, a G1 state, a G2 state, then back to a G1 state). Similarly, when the turnaround points are located within apurge zone314, the point on the surface will experience a single growth event per oscillation.
• For this exemplary analysis, it will be assumed that thereactive gas zones313,315 have a same length so that L_G=L_G1=L_G2. The purge gas zones will also have a same length so that L_P=L_P1=L_P2The length of the repeat cycle is therefore L_R=2·L_G+2·L_P., where L_Gis not necessarily equal to L_P. The average number of growth events per oscillation N_Gcan be calculated by weighting the number of growth events for amotion profile oscillation150 starting in the different zone types by the probability of starting in those zone types:
• $\begin{array}{cc}{N}_G=2\left(\frac{2L_G}{L_R}\right)+1\left(\frac{2L_P}{L_R}\right)& \left(5\right)\end{array}$
• The above equation indicates that growth is bounded by 1<N_G<2, where faster growth is favored by having thereactive gas zones313,315 occupy a greater percentage of the repeat length L_R. This is strictly true only when the purging operation is “perfect” and all reactions are saturated (i.e., fully completed in the available exposure time). In practice, optimization for a particular application may be useful. For the simplest case where L_G=L_P, Eq. (5) simplifies to N_G=3/2.
• With thedeposition zone305 having an integer N_Rfull ALD cycles plus ½ terminating cycle, the number of oscillations that it takes for a point to completely transit the working portion of the deposition zone (i.e., inclusive of all repeats318, where the growth rate is given by Eq. (5)), is N_O=N_R·L_R/Δx. The total number of growth events N_Tthat the point on the substrate experiences as it transits the working portion of the head (inclusive of all repeats318) is given by:
• 
  N_T≈N_G·N_O  (6)
• Note that the number of oscillations for the point on the substrate to transit across the full expanse of thedeposition zone305 will be somewhat more than N_Oas it is necessary to clear the terminating half cycle, and is: N_OD=(N_R+1/2)·L_R/Δx.).
• This simplified example neglects the transition regions that occur where the point is entering and leaving thedeposition zone305 where it will experience a reduced number of growth events per oscillation (0<N_G<1 when f=L_R). A more generalized analysis will be presented later which accounts for the transition regions at the start and end of the active growth region where the substrate will experience growth, but where the growth has not yet reached the steady state.
• An interesting class of solutions exist where the number of cycles is N_R+1/2 and N_R>1, which affords an opportunity for the forward motion distance f to be chosen as an integer multiple of L_Rsuch that f=K·L_R, where K is an integer ≤N_R. If the ooch distance Δx is scaled in proportion to K but is still small relative to L_Pand L_G, the total number of growth events for a point will approximately be independent of K but will still be directly dependent on N. The actual growth amount may be slightly dependent on K due to interactions between the reaction kinetics and velocity profile (i.e., the acceleration and terminal velocity), which may be a minor function off.
• The above examples illustrate that the total number of growth events N_Tis deterministic based on knowledge of the deposition head geometry and the motion profile imparted to thesubstrate97. For a given apparatus, the total number of growth events, and therefore the amount of material that is deposited on the substrate, can be modified by choosing appropriate values for the motion profile (for example, Δx given a fixed f=L_R). Other combinations of head design and motion can also be readily analyzed using a similar approach.
• These examples should not be taken to be limiting. One skilled in the art will recognize that other choices can be made in deposition head design and motion parameter selection in order to co-optimize other system level considerations such as overall physical size and mechanical dynamics. For example, to minimize the size of thedeposition head30, the pattern of gas zones in thedeposition zone305 can include as little as one cycle (i.e., N_R=1) with no terminating ½ cycle, such as the example discussed earlier with respect toFIG. 21. To minimize the required size for theinert zones308,309, the forward motion distance f can be reduced to slightly larger than L_Pwhile still enabling growth to occur (albeit at a reduced average rate per oscillation). In this case, there will be only one growth event per cycle.
• In practice, the forward motion distance f is not restricted to integer multiples of L_R, and may be greater or less than that physical distance as illustrated by the motion profiles inFIGS. 25-28. The following analysis is generalized to consider other deposition head geometries and values for f. For this analysis, a fully saturated ALD approximation will be used. For the purposes of this discussion, the growth from the motion profile and the associateddeposition zone305 will be treated separately from the constraints of theinert zones308,309 (i.e., the constraints for maintaining the active deposition region on thesubstrate97 in an inert environment separate from the external environment15).
• From inspection ofFIGS. 25-28, it can be seen that themotion profile oscillation150 results in exposure of a point on the substrate to a number of half reactions, which should be understood to be equivalent to the state-change treatment in the previous discussion. It is useful to understand the number of half reactions during any portion of the repeating motion profile.FIG. 31 illustrates a general method for calculating the number of half reactions, using the deposition head geometry and motion profile ofFIG. 26, with the simplification that L_G=L_G1=L_G2. It should also be understood that thedeposition zone305 contains a number ofreactive gas zones313,315, and the length of the deposition zone is equal to L_D=N_rgz·L_G+(N_rgz−1)·L_P, where N_rgzis the number ofreactive gas zones313,315. As discussed above, the minimum forward motion distance to achieve a reaction is the length of the purge zone (i.e., f>L_P). As illustrated inFIG. 31, if the forward motion is initiated at the boundary of thefirst reactant zone316 and the followingpurge zone314, but slightly inside of thepurge zone314, then the number of half reactions N_Havailable is equal to an integer n, where it can be shown that n=int[(f+L_G)/(L_G+L_P)], where int[·] is the integer operator. If instead, the forward motion is initiated at the boundary of thefirst reactant zone316 and the followingpurge zone314, but slightly inside of inside of thefirst reactant zone316, then it can be shown that the number of half reactions in equal to N_H=n+1. In the example ofFIG. 31, f is slightly longer than 3·L_P+2·L_Gso that n=3. A “maximum average growth region” where a point on thesubstrate97 experiences the maximum growth allowed by the combination of themotion profile oscillation150 and the geometry of thedeposition head30. The maximum average growth region can also be considered the “motion-profile-defined growth region”, or the “steady growth region” having a length L_steady=L_D+2L_P−f, as can be seen inFIG. 32 this region begins at a distance L_Pbefore the start of thedeposition zone305 and ends at a point at the start of a forward motion portion which terminates at a distance L_Pafter then end of thedeposition zone305, provided that L_I1≥L_Pand L_I2≥L_P. The steady growth region can also be considered a “plateau” region or “uniform growth” region, where a point on the substrate in this steady growth region with have, on average, the same number of growth events in each repeat of the repeating motion profile. Within this region, each forward and backward motion portion will experience either n or n+1 half reactions, which can be used to determine the average growth per oscillation.
• FIG. 32 illustrates the definition of some terms that can be useful for understanding the growth characteristics. As the substrate moves into the deposition zone from the left-to-right, ALD growth is only initiated after exposure to both reactive gases G1 and G2. Therefore, the first motion profile cycle that will cause growth on thesubstrate97 corresponds to a forward motion portion that just enters the left-most secondreactive gas zone315. Similarly, the last motion profile cycle that will cause growth on thesubstrate97 corresponds to a forward motion portion that starts just within the right-most secondreactive gas zone315. The “active growth region” corresponds to motions profile cycles having starting positions between these two extremes, and is shown with a length L_growth. It can be seen that the length of the active growth region is given by:
• $\begin{array}{cc}\begin{array}{c}{L}_{\mathrm{growth}}=\left(f-L_G-L_P\right)+(L_D-L_G-L_P)\\ =f+L_D-2·\left(L_G+L_P\right)\end{array}& \left(7\right)\end{array}$
• For exemplary motion profile cycles of the present invention, the net forward movement of each repeat unit is the ooch distance Δx, therefore the number of motion profile cycles where growth occurs is equal to L_growth/Δx. Similarly, the net forward velocity V_forwardthat thesubstrate97 moves through the deposition region is equal to the Δx divided by the time to complete one motion profile cycle, or approximately:
• $\begin{array}{cc}{V}_{\mathrm{forward}}\approx \frac{\Delta x}{t_{\mathrm{osc}}}=\frac{\Delta x}{\left(f+b\right)/\stackrel{\_}{V_{\mathrm{motion}}}}=\stackrel{\_}{V_{\mathrm{motion}}}\frac{\Delta x}{\left(f+b\right)}& \left(8\right)\end{array}$
• where|V_motion| is the average speed that the substrate is moved during the forward and backward motion portions. Note that the time for acceleration at the turn arounds of the motion profile will reduce the average V_forwardto a value less than the peak speed and the actual average speed will be dependent on the acceleration, terminal speed, and f+b used in the motion control.
• As thesubstrate97 moves into and out of the active grown region, there will be transition regions at the start and end of the active growth region where the substrate will experience growth, but where the motion profile has not yet advanced far enough to be fully with in the steady state growth region. In the transition regions, the number of half reactions that the substrate is exposed to is less than the number of half reactions that the substrate experiences when the motion profile is fully within the steady growth region. For a symmetric head, the transition regions will also be symmetric, and the difference between the length of the active growth region L_growthand the length of the steady growth region L_steadywill be the twice the length of the transition region L_transition:
• 
  L_growth=L_steady+2·L_transition  (9)
• In the illustrated example, the number of half-reactions will be 2 throughout the entire transition region. For cases where n>3, the number of half-reactions will transition from 2 at the outer boundary to n at the boundary of the transition regions with the steady growth region.
• Symmetric deposition heads, and therefore symmetric transition regions, are not a constraint of the present invention. It should therefore be understood that the size of the transition regions, and the number of half reactions as a function of position within the transition regions, can be independently calculated/determined for the leading and trailing transition zones.
• As illustrated inFIG. 25, at every “turnaround-event” the subsequent exposure to a reactive gas is the same as the previous exposure in the opposite direction. This is true within the repeat unit, and between repeat units. Using the simplification that Δx is very small relative to f and b in preferred embodiments, and therefore that f≈b (i.e., f/b≈1), it can be shown that the average number of half reactions H_rin the steady growth region for a full forward and backward motion profile cycle is given by:
• $\begin{array}{cc}\begin{array}{c}{H}_r=2·F_r-2\\ =2·\left(\left(f+L_G\right)/\left(L_G+L_P\right)\right)-2\end{array}& \left(10\right)\end{array}$
• where F_r=(f+L_G)/(L_G+L_P) is the average number of half reactions in the forward motion portion within the steady growth region. H_ris bounded by the number of half reactions available in the deposition zone such that H_ris the minimum of 2·F_r−2 and 2·N_rgz−2, where N_rgzis the number ofreactive gas zones313,315.
• As discussed earlier, in the transition regions, the number of half reactions transitions from n at the boundary of the steady growth region to 2 at the edge of the active growth region. It can be shown that the average number of half reactions in the transition region can be calculated using the following equation:
• $\begin{array}{cc}{T}_r=\frac{n·\left(F_r-0.5\right)-0.5·n^2-1}{F_r-2}& \left(11\right)\end{array}$
• Using these equations, the bounds on f and Δx can be determined for any deposition zone having symmetry. For any defined ooch distance Δx and forward motion f, the number of oscillations (i.e., the number of repeats of the motion profile oscillation150) that have growth will be equal to N_O=L_growth/Δx, of which the number of oscillations in the steady growth region will be given by N_OS=L_steady/Δx and the number of oscillations in the transition region will be given by N_OT=2L_transition/Δx cases where the lengths of the transition zones (L_entrance, L_exit) are equivalent, more generally N_OT=(L_entrance+L_exit)/Δx. The total number of growth events N_T(which will be equal to the total number of half reactions divided by 2) will therefore be given by:
• $\begin{array}{cc}\begin{array}{c}{N}_T=\frac{1}{2}\left[2·T_r·N_{\mathrm{OT}}+H_r·N_{\mathrm{OS}}\right]\\ =\frac{1}{2}\left[2·T_r\left(\frac{L_{\mathrm{transition}}}{\Delta x}\right)+H_r\left(\frac{L_{\mathrm{steady}}}{\Delta x}\right)\right]\end{array}& \left(12\right)\end{array}$
• These equations can be rearranged to solve for the required ooch distance Δx to give a desired total number of growth events N_T, given a head geometry and a specified value of the forward motion distance f. Similarly, they can be used to solve for the forward motion distance f given a specified ooch distance Δx. These equations make the assumption that any exposure to a reactive gas results in a full half reaction of growth, the exact growth for any motion profile will be dependent on the reaction kinetics and these equations should therefore be seen to provide bounding values; for example, these equations can be used to determine the maximum ooch value for a desired value of N_T. Since the number of growth events N_Twill experience a range of gas exposure times, the coating thickness may be less than what would be expected based on a steady-state ALD growth per cycle (GPC) calculation. Similar equations can be derived to apply fordeposition zones305 without symmetry. As noted, the length of each transition zone will need to be handled separately.
• As discussed earlier, there is no particular benefit to a forward motion portion f that is substantially longer than the length ofdeposition zone305, but it causes no harm if theinert zones308,309 are sufficiently long. One limiting example corresponds to the case where the length of the steady growth region is zero (i.e., L_steady=0). This will correspond to the case where f=L_D+2−L_P. The length of the growth region will be L_growth=2L_D−2L_G, and the length of the transition region will be L_transition=L_growth/2. The value of n will be given by one more than the total number ofreactive gas zones313,315 (n=N_rgz+1), and it can be seen that F_r=n so that from Eq. (11), T_r=(0.5·n²−0.5 n−1)/(n−2). Substituting into Eq. (12) gives:
• $\begin{array}{cc}\begin{array}{c}{N}_T=T_r\left(\frac{L_{\mathrm{transition}}}{\Delta x}\right)\\ =T_r\left(\frac{{\mathrm{<figure-callout\; label="L\; growth"\; filenames="US20180265977A1-20180920-D00023.png,US20180265977A1-20180920-D00024.png"\; state="\{\{state\}\}">L</figure-callout>}}_{\mathrm{growth}}}{2\Delta x}\right)\\ =T_r\left(\frac{L_D-L_G}{\Delta x}\right)\\ =\frac{\left(0.5·n^2-0.5·n-1\right)\left(L_G+L_P\right)}{\Delta x}\end{array}& \left(13\right)\end{array}$
• Solving for the ooch distance Δx needed to provide a desired number of growth events N_Tgives:
• $\begin{array}{cc}\Delta x=\frac{\left(0.5·n^2-0.5·n-1\right)\left(L_G+L_P\right)}{N_T}& \left(14\right)\end{array}$
• The ratio of the forward motion portion to the backward motion portion (f/b) is a useful attribute for characterizing a repeat motion profile. At the limit where L_steady=0, it can be shown that:
• $\begin{array}{cc}\frac{f}{b}=\frac{\left(n-1\right)+L_P/\left(L_G+L_P\right)}{\left(n-1\right)+L_P/\left(L_G+L_P\right)-\left(0.5·n^2-0.5·n-1\right)/N_T}& \left(15\right)\end{array}$
• It can be seen that the absolute size of the deposition zone L_Dis not a factor in the ratio of f/b, rather it is the relative size of the purge zone length L_Pto the reactive gas zone length L_G. Furthermore, it can be seen that, consistent with the figures and earlier discussion, as the number of desired growth events N_Tincreases, the ratio of f/b decreases (approaching 1.0 as growth events N_Tgrows and the corresponding Δx approaches 0.0). Additionally, as the number of reactive gas zones (N_rgz) decreases, the ratio of f/b will also decrease; therefore, deposition heads30 having a smaller footprint due to fewerreactive gas zones313,315, will operate using a motion profile where f/b is close to 1.0. At the limit where the number ofreactive gas zones313,315 in the deposition head is equal to the desired number of growth events N_Tthen, the ratio of f/b is roughly 2. Therefore, motion profiles of the present invention have a ratio of f/b less than 2, preferably less than 1.5, more preferably less than 1.2. In most preferred embodiments, the ratio of f/b will be less than 1.1. As should be understood, the actual value of f/b will be determined based on the geometry of the actual deposition head and desired number of growth events N_T.
• The present invention can be used with existing deposition head geometries, where the motion profile will be designed to get the desired coating given a specific deposition zone arrangement. In alternative embodiments, the present invention can be used to design optimized deposition zones, as well as overall output face features for deposition heads for specific applications.
EXAMPLES
• Embodiments of the present invention enable coating lengths that are substantially longer than the in-track length of thedeposition head30. Coatings were generated to illustrate different aspects of the present invention. The thin film coatings of the material layers on glass substrates were accomplished in each of the examples using a spatialALD deposition head30 having a similar configuration to that illustrated inFIG. 23. The gas flows through thedeposition head30 were controlled such that during operation it was operated as a vacuum-preloaded gas-bearingdeposition head30. As thesubstrate97 was moved relative to thedeposition head30, as described in each of the examples, the surface was exposed to the sequence of gaseous materials to perform ALD deposition.
• Glass substrates97 that were 2.5×2.5 inch squares or 2.5×5.0 inch rectangles were used. Eachsubstrate97 was attached to aheated backer device70 to form asubstrate unit74 that was positioned over theoutput face134 of the vacuum-preloaded gas-bearingdelivery head30. As described above, the separation between thesubstrate97 and theoutput face134 of thedelivery head30 was maintained by equilibrium between the flow of the gaseous materials supplied to theoutput slots112 on theoutput face134 and a slight amount of vacuum supplied to theexhaust slots114. For all of the examples, the exhaust supply pressure was approximately 40 inches of water below atmospheric pressure. The purge gas P supplied in thepurge zones314 and the inert gas I supplied in theinert zones308,309 were pure nitrogen. The first reactive gas G1 supplied in the firstreactive gas zones313 was a mixture of nitrogen, water vapor, and optionally ammonia vapor (an oxygen reactive precursor O). The second reactive gas G2 supplied in the secondreactive gas zones315 was a metal reactive precursor M, and was one or a mixture of active metal alkyl vapors in nitrogen.
• The metal alkyl precursors used in these examples were trimethylaluminum (TMA) and diethyl zinc (DEZ). The flow rate of the active metal alkyl vapor was controlled by bubbling nitrogen through the pure liquid precursor contained in an airtight bubbler by means of individual mass flow controllers. This saturated stream of metal alkyl was mixed with a dilution flow of nitrogen before being supplied to thedeposition head30. The flow of water vapor was controlled by adjusting the flow rate of nitrogen passed through pure water in a bubbler. This saturated stream of water vapor was mixed with a dilution flow before being supplied to thedeposition head30. The flow of ammonia vapor was controlled by passing pure ammonia vapor from the head space of a compressed fluid tank through a mass flow controller and mixing with the water vapor stream. All bubblers were held at room temperature. The temperature of the coating operation was established by controlled heating of both thedeposition head30 and thebacker device70 to a desired temperature. Experimentally, the flow rates of the individual gasses (in units of standard cubic centimeter per minute) were adjusted to the settings shown in table580 ofFIG. 33 for each of the material layers (Al₂O₃, ZnO and ZnO:N) coated in the examples contained herein. The flows shown are the total flows supplied to the coating device, and thus are partitioned equally among the individual gas zones of a common type. In all of the experimental examples, the growth temperature used was 200° C.
• FIG. 34 shows a table590 containing details for a series of comparative and exemplary SALD coatings made in accordance with embodiments of the invention. In all examples, deposition was initiated after placing thesubstrate97 to be coated in a position where it was centered over thedeposition head30. After establishing the z-position via the gas bearing effect, thesubstrate97 was moved a distance of half of the total forward motion in an in-track direction. For samples that were coated using a simple oscillating motion, the ooch distance is equal to zero, and therefore backward distance is equal to the forward motion distance. After completing the number of repeat motion units, the sample was returned to the center position for unloading from the deposition system. Thedeposition head30 used for the examples was constructed such that the total in-track length of the head was 2 inches, and the length of thedeposition zone305 was L_D=0.9 inches. There were fivereactive gas zones313,315 separated by fourpurge gas zones314 having the pattern illustrated inFIG. 23. The zone lengths in the deposition zone were L_G1=L_G2=L_P=0.1 inches. The inert zones were roughly equivalent, and nominally were L_I1=L_I2=0.55 inches (14 mm).
• As seen in Table 2 ofFIG. 34, three materials were grown to evaluate the ooching motion profile. Starting with comparative example C1, a 2.5 inch square substrate was coated using a simple oscillating motion profile (with equal forward and backwards motion distances) for 30 oscillations to grow nominally 20 nm of ZnO:N. During the deposition, the center of gravity of thesubstrate unit74 never traveled beyond the edge of thedeposition head30. The growth per cycle (GPC) of ZnO:N for the deposition temperature of 200° C. and a velocity of 50.8 mm/sec is approximately 1.9 Angstroms/cycle, therefore the total thickness of the deposited material was nominally about 23 nm.
• For inventive example I1, a 2.5×5.0-inch substrate was coated using a simple oscillating motion profile (with equal forward and backwards motion distances) that moved the center of gravity of thesubstrate unit74 beyond the edge of thedeposition head30 by about 15.25 mm. The number of oscillations for I1 was also30, with the difference between I1 and C1 being the length of travel during the oscillations to enable coating a larger substrate area. As discussed earlier, the vacuum-preloadeddeposition head30 was able to maintain the gap between thesubstrate97 and theoutput face134 without tipping during deposition, and a uniform coating of ZnO:N was obtained over a coating length of approximately 2.5 inches. In example I1, the active deposition region of the substrate moved beyond the edge of thedeposition head30 so that it was exposed to theexternal environment15 during growth.
• Inventive example I2 demonstrates the use of a repeating motion profile where the forward motion distance f was larger than the backward motion distance b by an ooch distance Δx. In this example, the center of gravity of thesubstrate unit74 was not moved beyond the edge of thedeposition head30. In order to prevent the exposure of the active deposition region to the external environment, the forward motion distance f was chosen to be 7.5 mm. The required ooching distance was then calculated and used to achieve the same number of ALD cycles as in the simple oscillating motion profile of comparative example C1. Inventive sample I2 produced an equivalent ZnO:N coating thickness to that in examples C1 and I1, but obtained it using the ooching motion profile. The uniform coating length was very short for I2, approximately 0.18 inches due to the prior art imposed limitation of maintaining the center of gravity over the deposition head.
• Inventive example I3, combined the inventive aspects of examples I1 and I2 to coat a 2.5×5.0-inch substrate using the same ooching motion profile of example I2 for a larger number of oscillations to cover a larger substrate area. In this example, the center of gravity of thesubstrate unit74 was moved beyond the edge of thedeposition head30 during both the beginning and end of the coating process by 15.25 mm. As discussed earlier, the vacuum-preloadeddeposition head30 was able to maintain the gap between thesubstrate97 and theoutput face134 without tipping during deposition, and a uniform coating of ZnO:N was obtained over a coating length of approximately 2.1 inches. In example I3, the active deposition region of the substrate was protected from the external environment during growth.
• Comparative example C2, is similar to comparative example C1 with the following exceptions. A 2.5-inch square substrate was coated using a simple oscillating motion profile for 154 oscillations to grow nominally 50 nm of Al₂O₃. The growth per cycle (GPC) of Al₂O₃for the deposition temperature of 200° C. and a velocity of 50.8 mm/s is approximately 0.812 Angstroms/cycle, therefore the total thickness of the deposited material was nominally about 50.0 nm.
• Inventive examples I4 and I5 are similar to I1 and I3, respectively. Inventive example I4 illustrates the deposition of a uniform coating of Al₂O₃with a larger number of oscillations using the same motion profile as Inventive example I1, where the center of gravity of thesubstrate unit74 traveled beyond the edge of thedeposition head30 by about 15.25 mm. The number of oscillations for I4 was 154 (the same as comparative example C2). As with I1, the vacuum-preloadeddeposition head30 was able to maintain the gap between thesubstrate97 and theoutput face134 without tipping during the 154 oscillations, and a uniform coating of Al₂O₃was obtained (over a length of approximately 2.5 inches). In example I1, the active deposition region of the substrate moved beyond the edge of thedeposition head30 so that it was exposed to the environment.
• Inventive example I5 used a similar repeating motion profile to that of example I3. The same forward motion distance f of 7.5 mm was used, and the ooch distance was adjusted to achieve the desired number of ALD cycles for Al₂O₃. Inventive sample I4 produced an equivalent Al₂O₃coating thickness to that in examples C2 and I5, but obtained it using the ooching motion profile for a uniform coating length of 2.1 inches which was protected from the external environment during growth. Furthermore, the vacuum-preloaded condition was maintained during the deposition event.
• In Comparative example C3, a 2.5-inchsquare substrate97 was coated using a simple oscillating motion profile (with equal forward and backwards motion distances) for 75 oscillations to grow nominally 57 nm of ZnO. During the deposition, the center of gravity of thesubstrate unit74 never traveled beyond the edge of thedeposition head30. The growth per cycle (GPC) of ZnO for the deposition temperature of 200° C. and a velocity of 50.8 mm/s is approximately 1.9 Angstroms/cycle, therefore the total thickness of the deposited material was nominally about 57 nm.
• In inventive example I6, a repeating motion profile was used where the forward motion distance f was larger than the backward motion distance b by an ooch distance Δx. In this example, the center of gravity of thesubstrate unit74 was moved beyond the edge of thedeposition head30 by 3.8 mm. The forward motion distance f was chosen to be 10.16 mm (0.4 inches) which is equal to the length of one repeat unit (G1-P-G2-P). An ooch distance of Δx=0.1016 mm was used to achieve the same number of ALD cycles as in the simple oscillating motion profile of comparative example C3. Inventive sample I6 produced an equivalent ZnO central uniform coating thickness to that in examples C3, having a length of approximately 1.0 inch. Inventive example I7 is similar to and has the same uniform coating thickness as I6, however the total number of oscillations was increased to coat a longer region of the substrate, for a uniform coating region of approximate 1.9 inches.
• As can be seen from the description of the inventive examples, and the information in Table 2 (FIG. 34) the present invention can be used tocoat substrates97 that are significantly longer than the in-track length of thedeposition head30. Furthermore, a repeat motion profile can be used to control the exposure of the active growth region to theexternal environment15. The parameters of the repeat motion profile f and Δx can be selected to achieve the desired coating thickness given the deposition conditions used (temperature, velocity, and precursor concentration). The number of oscillations used with a given repeat motion profile can independently control the length of uniform coating area, which for the coating of continuous substrates can be set to a value to coat the length of web in a roll-to-roll operation.
• The invention has been described in detail with particular reference to certain preferred embodiments thereof, but it will be understood that variations and modifications can be effected within the spirit and scope of the invention.
PARTS LIST
• 15 external environment
• 30 deposition head
• 60 mass
• 61 lever
• 62 pivot point
• 63 air bearing
• 64 rolling element bearing
• 65 cantilever spring
• 66 pin
• 67 center of gravity
• 66 stop pin
• 70 backer device
• 71 center of gravity
• 74 substrate unit
• 75 vertical element
• 76 springs
• 77 counter weight
• 78 counter mass
• 87 linkage
• 94 working distance
• 95 outboard support elements
• 97 substrate
• 98 motion arrow
• 110 gas slot
• 112 output slot
• 113 force vector
• 114 exhaust slot
• 123 force vector
• 134 output face
• 150 motion profile oscillation
• 165 motion profile
• 200 SALD system
• 205 deposition subsystem
• 210 deposition unit
• 270 relative motion means
• 280 substrate positioner module
• 281 roller
• 282 roller bearing
• 283 yoke feature
• 284 flexure
• 285 frame
• 286 frame opening
• 288 substrate positioner
• 305 deposition zone
• 308 inert zone
• 309 inert zone
• 311 edge
• 313 first reactive gas zone
• 314 purge zone
• 315 second reactive gas zone
• 316 first reactant zone
• 317 last reactant zone
• 318 repeat unit
• 321 left edge
• 322 right edge
• 401 output slots
• 500 pressure distribution plot
• 580 table
• 590 table

Claims (15)

1. A thin film deposition system, comprising:

a vacuum-preloaded gas bearing deposition head positioned in an external environment having an ambient pressure, the deposition head having an output face facing a first substrate surface of a substrate and being configured to expose the first substrate surface to a plurality of gaseous materials including one or more reactive gaseous materials, wherein the output face includes a plurality of source openings through which the gaseous materials are supplied and one or more exhaust openings;

wherein an exhaust pressure at the exhaust openings is less than ambient pressure and a source pressure at the source openings is greater than the exhaust pressure at the exhaust openings, and wherein the pressure at the outermost source openings is greater than ambient pressure;

a substrate unit including the substrate, the substrate having a length in the in-track direction;

a motion control system that moves the substrate unit over the output face of the deposition head in the in-track direction in accordance with a specified motion profile without constraining motion of the substrate unit in a direction normal to the output face of the deposition head, wherein the motion profile includes moving a center of gravity of the substrate unit beyond the first edge of the output face; and

wherein an integrated pressure across the output face provides a net force on the first substrate surface in a direction normal to the output face, wherein a sum of the forces on the substrate unit in a direction normal to the output face and a sum of moments on the substrate unit are zero at all positions in the motion profile and wherein a gap exists between the first substrate surface and the output face at all positions in the motion profile.

2. The thin film deposition system ofclaim 1, wherein the length of the substrate is longer than the length of the output face in the in-track direction.

3. The thin film deposition system ofclaim 1, wherein the output face includes:

a first inert zone having wherein an inert gaseous material is supplied;

a second inert zone having wherein an inert gaseous material is supplied; and

a deposition zone located between the first inert zone and the second inert zone along the in-track direction where the first substrate surface is exposed to the one or more reactive gaseous materials to deposit a substance onto the first substrate surface.

4. The thin film deposition system ofclaim 3, wherein the deposition zone is fully covered by the first substrate surface at all positions in the motion profile.

5. The thin film deposition system ofclaim 1, wherein the source openings include two outermost openings and one or more inner source openings between the outermost source openings, and wherein the source pressure at outermost source openings is controlled independently of the source pressure at the one or more inner source openings.

6. The thin film deposition system ofclaim 1, further including one or more outboard support elements located outboard of the deposition head in the in-track direction which are positioned to engage with the substrate to provide a support force that prevents the substrate from tipping.

7. The thin film deposition system ofclaim 6, wherein the outboard support elements include shelves, rollers or gas bearings.

8. The thin film deposition system ofclaim 1, wherein the substrate unit includes a backer device rigidly attached to a second substrate surface of the substrate, the second substrate surface being opposite to the first substrate surface.

9. The thin film deposition system ofclaim 8, wherein the backer device is attached to the second substrate surface using a vacuum force.

10. The thin film deposition system ofclaim 8, further including a counter moment inducing system that applies a counter moment to the substrate unit which is a function of the in-track position of the substrate unit.

11. The thin film deposition system ofclaim 10, wherein the counter moment inducing system includes a first component attached to the backer device which engages with a second component mounted in a fixed position.

12. The thin film deposition system ofclaim 10, wherein the counter moment inducing system includes a moveable counter mass which moves in an opposite direction to the substrate unit, and wherein the counter mass connects to the backer unit such that the counter mass transfers a torque to the backer unit.

13. The thin film deposition system ofclaim 10, wherein the counter moment inducing system includes a torsion spring, a cantilever spring, a coil spring, a leaf spring, a magnetic actuator, an electromagnetic actuator, gas bellows, pneumatic cylinder.

14. The thin film deposition system ofclaim 1, wherein the motion profile includes a repeating motion pattern that includes a forward motion portion that moves the substrate a forward distance in the in-track direction and a backwards motion portion that moves the substrate a backward distance in the in-track direction.

15. The thin film deposition system ofclaim 14, wherein the forward distance is larger than the backward distance such that the repeating motion pattern provides a net forward motion.

Images