BACKGROUND OF THE INVENTION 1. Field of the Invention
Embodiments of the present invention generally relate to a sputtering apparatus for forming films on large area substrate such as flat panel, large screen televisions, and solar panels.
2. Description of the Related Art
As demand for larger flat panel display screens increases, so must the sputtering target area. As the sputtering target become larger, it becomes increasingly more difficult to adequately provide a uniform distribution of reactive gas to the sputtering target. In the past, reactive gas has been introduced to the sputtering chamber through a gas inlet. The gas inlet is typically located on the side of the chamber. With an increase in target size, the reactive gas tends to not adequately reach the center of the target. When the reactive gas does not uniformly reach the entire target, the film deposited on the substrate will not have a uniform composition across the substrate. For a large area sputtering targets with the gas introduced from the periphery of the target, the gas concentration is highest at the chamber wall where the gas inlet is located. The gas concentration decreases moving across the chamber to a low point in the center of the chamber.
U.S. Pat. No. 5,346,601 to Barada et al. shows a sputtering apparatus in which two gas introduction tubes are provided within a collimator. The gas introduction tubes are perpendicular to each other and within the same plane. The gas tubes extend across the processing area. By providing the gas introduction tubes within the collimator, the reactive gas can adequately be provided to the substrate while not shadowing the wafer from sputtering material. A plurality of gas outlets are present across the tube. In order to remove the gas tubes, the entire collimator structure must be removed. The collimator cannot be removed without also removing the gas tubes.
As shown by Barada et al., sometimes gas introduction tubes can extend across the processing space between the target and the substrate. The gas introduction tubes, such as that used by Barada et al., usually only introduce gas through a series of gas outlet holes formed in a gas introduction tube. A problem with prior art gas introduction tubes is that they must provide the gas at a high pressure through the tube in order to have a uniform pressure passing through the tiny holes in the tube. When the process is stopped and the gas is stopped, gas will continue to flow out of the holes because of the pressure buildup within the tube. The gas will continue to disperse into the processing chamber even after the process has stopped. The excess gas introduced into the chamber may contaminate the wafer or cause further, undesirable reactions with the substrate.
There is a need in the art to provide reactive sputtering gas to a chamber uniformly across a large area sputtering target. There is also a need in the art to provide easily removable reactive gas introduction tubes without disassembling the sputtering chamber.
SUMMARY OF THE INVENTION The present invention generally involves a sputtering apparatus for forming films on large area substrates such as flat panel, large screen televisions.
In a first embodiment, a sputtering apparatus has a vacuum chamber, a sputtering target, a substrate support, and a plurality of parallel gas introduction tubes. The gas introduction tubes extend across the vacuum chamber in an area between the target and the substrate support.
In a second embodiment, a sputtering apparatus has a vacuum chamber, a sputtering target, a substrate support, and one or more gas introduction tubes extending across the vacuum chamber in an area between the target and the substrate support. Each tube has at least one inner tube having a plurality of openings and an outer tube having a plurality of openings. The outer tube surrounds the at least one inner tube.
In a third embodiment, a method of sputtering a sputtering target in a sputtering apparatus comprises sputtering the target to deposit a layer on a substrate. The apparatus comprises a vacuum chamber, a sputtering target, and a plurality of gas introduction tubes extending across the vacuum chamber in an area between the target and the substrate wherein no collimator is present between the target and the substrate.
In a fourth embodiment, a method of sputtering a sputtering target in a sputtering apparatus is provided that comprises sputtering the target to deposit a layer on the substrate. The sputtering apparatus comprises a vacuum chamber, a sputtering target, and one or more gas introduction tubes extending across the vacuum chamber in an area between the target and the substrate. Each tube comprises at least one inner tube comprising a plurality of openings and an outer tube comprising a plurality of openings. The outer tube surrounds the at least one inner tube.
In a fifth embodiment, a gas introduction tube is disclosed that comprises at least one inner tube comprising a plurality of openings and an outer tube comprising a plurality of openings. The outer tube surrounds the at least one inner tube.
BRIEF DESCRIPTION OF THE DRAWINGS So that the manner in which the above recited features of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.
FIG. 1 is a side view of a gas introduction tube used in the instant invention.
FIG. 2 is a sputtering apparatus according to the instant invention showing the spacing between the target, substrate, and gas introduction tubes.
FIGS. 3A-3D are cross sectional representations of gas introductions tubes.
FIG. 4 is an isometric view of a lower chamber assembly in an exemplary physical vapor deposition chamber.
FIG. 5 is an isometric cross-sectional view of gas introduction tubes formed in an exemplary physical vapor deposition chamber.
FIG. 6 is an isometric cross-sectional view of a lower chamber assembly in an exemplary physical vapor deposition chamber according to this invention.
FIG. 7 is an isometric cross-sectional view of gas introduction tubes formed in an exemplary physical vapor deposition chamber according to this invention.
DETAILED DESCRIPTION The present invention generally provides an apparatus to introduce reactive gas to a sputtering target apparatus. The sputtering apparatus can be small enough to process semiconductor wafers or sufficiently large to process large area substrates used in making flat panel television screens.
FIG. 2 is an exemplary schematic of a sputtering apparatus incorporating the instant invention. An exemplary sputtering apparatus which can be modified to incorporate the instant invention is shown in U.S. patent application Ser. No. 11/247,705 filed Oct. 11, 2005 and hereby incorporated by reference in its entirety. The target9 rests on abacking plate8 within avacuum chamber7. Thesubstrate12 is positioned in opposition to the target9. Thesubstrate12 rests on apedestal13. Between the target9 and thepedestal13,gas introductions tubes14 are found. Thesetubes14 extend across the chamber. As few as one and as many as necessary can be provided. If more than onetube14 is provided, thetubes14 are substantially parallel to each other and in the same plane as each other. In one embodiment, thegas introduction tubes14 are spaced about 100 mm to about 300 mm apart. In another embodiment, thegas introduction tubes14 are spaced about 150 mm to about 180 mm apart as shown by arrow E. Thetubes14 are located about halfway between the target9 and thesubstrate12, but thetubes14 should be greater than about 30 mm away from thesubstrate12 and greater than about 30 mm away from the target9. If thegas tubes14 are closer than 30 mm to the target9, thetubes14 will likely disturb the plasma by sinking an excessive fraction of electrons from it and form a layer on the target9. If thegas tubes14 are closer than 30 mm to thesubstrate12, then thegas tubes14 will block material from evenly reaching thesubstrate12 so that a non-uniform film will be formed. The target9 andsubstrate12 can be separated by about 300 mm to about 360 mm. The target9 would then be about 150 mm to about 180 mm from thetubes14 as shown by arrow C. Thesubstrate12 would then be about 150 mm to about 180 mm from thetubes14 as shown by arrow D.
Thegas tubes14, while located between the target9 and thesubstrate12, are not intended to provide any collimating effect. In fact, it is preferable that thegas tubes14 do not provide any collimating effect. No collimator should be present. A collimator will interfere with the uniform distribution of material on thesubstrate12. Therefore, only asmany gas tubes14 as are necessary to ensure a uniform gas distribution should be present within the chamber. One ormore gas tubes14 could be present. Thegas tubes14 can run in a substantially 2-dimensional plane and could be parallel or substantially parallel. Alternatively, thegas tubes14 can intersect or even overlap.
By extending thegas introduction tubes14 across the processing area, the reactive gas can be evenly provided to the target9 for reaction. When the gas is provided at the periphery, the gas is not evenly distributed to the target9. When the gas is not evenly distributed, the resulting film will not be uniform in composition across the surface. By providing the reactive gas along the length of the target9, the reactive gas will be uniformly provided to the target9 and the deposited film will have a uniform composition across its surface. It is especially difficult to provide reactive gas uniformly to the target9 when the target9 is a large area target9 used for forming flat panel television screens.
As noted above, the prior art gas introduction tubes will still emit gas into the processing region even after the gas has been turned off and the process stopped. The gas continues to flow into the processing region because of the high pressure buildup within the gas introduction tube and the tiny holes through which the gas will pass. Simply increasing the hole size would certainly decrease the pressure and allow the reactive gas to stop flowing upon shutdown, but more gas will leave the tube at the hole closest to the edge than will leave at each additional hole along the gas tube. The larger holes will decrease the pressure and, thus, allow less gas to be introduced at the center of the target9.
Because increasing the hole size will not solve the gas introduction problem, another solution was found to maintain sufficient pressure within the tube to uniformly provide gas to the whole target surface and to also quickly reduce pressure within the tube at shutdown.FIG. 1 shows an exemplary gas introduction tube according to the instant invention. Thegas introduction tube3 contains numerous other tubes within the tube. For clarity, only three tubes are shown in the figure, but it is understood that as many tubes as are practically necessary could be provided. The outer tube has awall4 with numerous holes along its length. Within the outer tube is a middle tube that has awall5. The middle tube has numerous holes along its length as well. Within the middle tube is an inner tube that has awall6 with numerous holes along its length.
Thegas introduction tube3 provides a uniform gas pressure along the length of the tube and quick dispersion of gas at shutdown. Gas flows into thetube3 through the inner tube at a high pressure. The gas passes through thetube3 and passes through thewall6 to the middle tube. The gas in the middle tube is then dispersed through thewall5 to the outer tube. The gas in the outer tube is then dispersed throughwall4 to the processing chamber. The holes in the walls of the tubes are not lined up. If the holes are lined up, then the gas would disperse directly from the inner tube to the chamber. Such a situation would render thetube3 exactly the same as the tube of the prior art in effectiveness. By misaligning the holes, the gas must snake through the processing tubes and decrease in pressure as it passes through the holes. By providing the gas in the inner tube at a high pressure, the gas will snake through the middle tube and outer tube until it reaches the processing chamber. Each time the gas passes through a hole to another tube, the pressure will drop. When the gas is turned off, the pressure will rapidly drop within the tube and prevent unwanted gas introduction.
FIGS. 3A-3D show cross sections of gas tubes3A-3D. The gas tubes3A-3D can be round (FIG. 3A), oval (FIG. 3B), square (FIG. 3C) or any conventional shape. So long as the shape of the tubes does not interfere with sputtering target material, the shape of the gas introduction tubes is not restricted.FIG. 3D shows a circulargas instruction tube3D that has a plurality of holes F. The holes F are the outlets for the gas to pass into the chamber. The holes F can face downward towards the substrate, upwards toward the target, or sideways away from both the target and the substrate. In one embodiment, the holes F face away from the target and the substrate. The holes F should be present on less than about 10% of thegas introduction tube3D. In one embodiment, about 10 to about 50 holes F can span the length of the tube. In another embodiment, about 25 to about 35 holes F can span the length of the tube. The size of the holes F should be much smaller than the diameter of the tubes. In one embodiment, the holes F are about 5 times smaller than the diameter of the tube. In another embodiment, the holes F are about 10 times smaller than the diameter of the tube. The diameter of the tubes can be about ⅛″ to about ⅞″. In another embodiment, the diameter of the tubes can be about ¼″ to about ¾″.
Thegas introduction tubes14 can have a bias applied to them. The bias can be applied to anindividual tube14 or collectively to alltubes14. Thetubes14 can have an RF bias applied so that thegas introduction tubes14 will function not only as a gas source, but also as an ionization source. Thegas tubes14 could also have an AC, DC, or pulsed bias applied from apower source2 or thetubes14 can be grounded (seeFIG. 2). The gas introduction tubes could be used as an additional sputtering target if desired. The gas tubes should be made of the same material as the sputtering target to prevent contamination.
Thegas introduction tubes14 can also be tailored to suit the needs of the user. For example,multiple tubes14 can be used with eachtube14 providing a different processing gas. Additionally, the instant invention provides the added benefit of functionality. Thetubes14 can be easily removed through the access port. By removing thetubes14 through the access ports, the entire chamber does not need to be disassembled to simply change afew tubes14. Benefits of such an easy removal are clear. Downtime is significantly reduced.
In one embodiment of theprocess chamber10, illustrated inFIG. 4, thelower chamber assembly35 may contain one or moregas introduction tubes14. In one embodiment, eachtube14 extends through theprocessing region15. In this configuration thetubes14 are in electrical contact with the groundedshield50, so that current flowing through thetubes14 passes through theshield50 to ground. In another configuration, thetubes14 are biased and not in contact with theshield50. In one embodiment, thetubes14 are positioned over the stationaryconductive member support97 and is used to hide or isolate theconductive member support97 from the plasma generated in the processing region15 (FIG. 6). The ability to hide or isolate theconductive member97 from the plasma will reduce the amount of deposition that will land on the stationaryconductive member support97 and thus minimize particle generation as thetubes14 are removed from processingregion15 of theprocess chamber10. In one embodiment, thetubes14 are longer than the target surface in the dimension in the direction in which thetubes14 extend and thus the conductive member support(s)97 are not positioned below the target surface so as to limit the interaction between the plasma generated in theprocessing region15 and the conductive member support(s)97.
InFIG. 4 the lid assembly has been removed, and is not shown, to more clearly illustrate some of the components in the lowerprocessing chamber assembly35. In the embodiment shown inFIG. 4, thelower chamber assembly35 generally contains asubstrate support assembly60,chamber body assembly40, agas delivery system14 and ashadow frame52. In one aspect, as shown inFIG. 4 thechamber body assembly40 generally contains aprocess kit holder140, one ormore chamber walls41 and achamber base42. Theprocess kit holder140 is positioned on thechamber walls41 and is adapted to support theshield50, anupper shield50E and one or more tubes14 (e.g., three shown inFIG. 4). In one aspect, theprocess kit holder140 electrically connects theshield50 and theupper shield50E to thechamber walls41 which are grounded. Theshield50 andupper shield50E are generally sized and adapted to prevent the plasma and sputtered target material from escaping from theprocess region15 and depositing on the components in thelower chamber assembly35. In the configuration illustrated inFIG. 4 thelower chamber assembly35 contains threetubes14 that are positioned above thesubstrate support61. In one aspect, as shown inFIG. 4, theconductive member support97 is mounted on and electrically connected to the groundedshield50.
It should be noted that the cross-sectional area and the material used to form the components in thetube14, theconductive member93, and theconductive member support97 is important since it will affect the ability to withstand the high temperatures that it will be seen during processing (e.g., resistive heating and interaction with the plasma). The number oftubes14 and the surface area of theconductive member93 exposed in theprocessing region15 is important since it will have an effect amount of current carried by eachconductive member93 and thus the maximum temperature achieved by eachconductive member93 andconductive member support97 during processing. The total surface area of theconductive member93 can be defined by the length of theconductive member93 in the processing region times the length of the exposed perimeter of theconductive member93 times the number of conductive members positioned in the processing region. In one aspect, the number ofgas tubes14 positioned in theprocessing region15 may be between about one and about twenty depending on the desired process uniformity, cost and complexity allowed for a desired application. Preferably, the number ofgas tubes14 that pass through theprocessing region15 is as few as possible with a preferred range of between about two and about ten. The exposed perimeter of the embodiment of theconductive member93 illustrated inFIG. 7 can generally be defined as twice the vertical length plus the horizontal length of surface of theconductive member93. In one example, for a substrate that is 1800 mm×1500 mm in size the exposed surface area of all of theconductive members93 was about 5.0 m2, which is spread across sevenconductive members93 that were 1.9 meters long. In one aspect, the cross-sectional area of theconductive member93 is sized to carry the current delivered to theconductive members93 from the plasma generated by the target bias. In one example, the total current that could be carried by all of the conductive members is about 1000 amps.
WhileFIGS. 4-7 illustrate embodiments of thetubes14 that are generally straight and are generally rod or bar shaped, this configuration is not intended to limit the scope of the invention described herein. In general, the term bar, or rod, shaped as used herein is intended to described a component that is longer (e.g., X-direction) than its cross-section is wide or high. In one aspect, the bar or rod shapedtubes14 are not straight and thus have one or more regions along their length that are curved or coiled. In one embodiment, thetubes14 are positioned throughout the processing region to improve the sputter deposited film uniformity on the substrate surface by increasing the tube surface area and not appreciably obstructing or altering the amount and/or direction of the flux of sputtered material passing from the target to the substrate surface. Referring toFIGS. 3A-3D, in one embodiment, the cross-section of thetubes14 are oval, round, rectangular, or other cross-sectional shape that will not appreciably obstruct or alter the amount and/or direction of the flux of sputtered material passing from the target to the substrate surface.
FIG. 5 illustrates an exploded isometric view of atube14 that has a conductive memberelectrical connection point105 that is adapted to electrically contact a supportelectrical connection point104 of thesupport102. In one aspect, the conductive memberelectrical connection point105 and the supportelectrical connection point104 act as apivot point106 that allows thetube14 to be positioned in and/or removed from the processing region15 (discussed below). To hide the pivot point106 asupport cover103 is positioned over this region to prevent the sputtered material deposition from inhibiting the removal of these components from theprocess region15. Theconductive member support97 may have apivot point106 at one end and an end that is detachable from the other vertical support.
In one embodiment, not shown, thetubes14 are cantilevered over the substrate surface and thus do not extend all the way across the substrate. In one aspect, the cantilevered end of thetubes14 may only extend to a point that is above the center of the substrate positioned on the substrate support. In one aspect, the cantileveredtubes14 are evenly distributed throughout theprocessing region15.
While the embodiments of theprocess chamber10 illustrated herein all show thetubes14 in contact with theshield50, this configuration is not intended to be limiting to the scope of the invention described herein. Therefore, in some embodiments the vertical support may be mounted on a bracket or supporting surface positioned in thechamber body assembly40.
Gas Tube Removal
FIG. 6 is an isometric cross-sectional exploded view as viewed from outside theprocess chamber10 that illustrates thetubes14 andplates99 in a position that is partially removed from theprocessing region15 of theprocess chamber10. In one embodiment of the invention, thetubes14 are adapted to be removed from theprocess chamber10 through anaccess port98 formed in theprocess kit holder140. In one aspect, theaccess port98 may be formed in thechamber wall41. InFIG. 6, the lid assembly has been removed to more clearly illustrate some of the components in the lowerprocessing chamber assembly35. Thetube14 has ahandle93A that is attached or welded to the surface of thegas introduction tubes14 to facilitate the insertion and/or removal of thegas introduction tubes14 through theaccess port98 formed in theprocess kit holder140.
When thetube14 has reached its useable lifetime, thetube14 can be removed from theprocessing region15 by venting theprocess chamber10 and removing aplate99 that is sealably attached to theprocess kit holder140 so that a user can access thetube14 through theaccess port98. The process of removing thetube14 may include shutting “off” the vacuum pumps (not shown) and then delivering a flow of an inert gas, such as argon, into the vacuum processing area from thetubes14 to create a pressure greater than atmospheric pressure in the vacuum processing area. Creating a positive pressure in the processing area during the removal of thetube14 may be advantageous since it can prevent the contamination of the chamber components positioned in theprocessing region15 due to the exposure of the process kit components to atmospheric contamination (e.g., atmospheric gases, vapors or particles). In one aspect, theaccess ports98 are purposely kept as small as possible to minimize the area through which atmospheric contamination can enter theprocessing region15. The down time of theprocessing chamber10 can thus be minimized since there is no need to remove and reposition the chamber lid assembly20 and/or other major chamber components, there is no need to bake out of the chamber to remove adsorbed gases and water from processing chamber components, and there is no need to replace contaminated components due to their exposure to atmospheric contamination.
Gas Introduction Tube Bias
In one embodiment of theprocess chamber10, abiasable shield50F may be positioned in the processing region to change the electric field and the plasma density generated near the edge of the target and substrate.FIG. 7 illustrates one embodiment of thebiasable shield50F that is positioned around the periphery of thesubstrate12 and is electrically connected to theshield50, which is grounded, by use of anelectrical component50G. In one aspect, theelectrical component50G may be used as a “stand-off” to physically space thebiasable shield50F from theshield50. It should be noted that the term “grounded” is generally intended to describe a direct or in-direct electrical connection between a component and the anode. Thebiasable shield50F may be purposely biased at a different potential versus the tube surfaces due to the introduction of theelectrical component50G that may add resistive, capacitive and/or inductive type elements to the electrical path between thebiasable shield50F and the tube surfaces. In one aspect, during processing a bias voltage, which will generally be less anodic, may be “passively” induced in thebiasable shield50F due to a bias applied between the target and anodic surface (e.g., shield50) and the interaction of thebiasable shield50F with the plasma generated in the processing region. In another aspect, not shown, thebiasable shield50F may be separately biased by use of a power supply (not shown) which is in electrical communication with thebiasable shield50F and the anode surfaces. In this configuration theelectrical component50G may act as an insulator.
In another embodiment of theprocessing chamber10, thetubes14 may be purposely biased at a different potential versus the anode surfaces by the introduction of a resistive, capacitive and/or inductive components to the electrical path between thetubes14 and the anode surfaces. In one embodiment, as shown inFIG. 7, a secondelectrical component50H may be positioned in the electrical path between thetube14 and theshield50 to allow thetube14 to be biased at a different potential than theshield50. In one aspect, during processing a bias voltage, which will generally be less anodic, may be “passively” induced in thetube14 due to a bias applied between the target and anodic surface (e.g., shield50) and the interaction of thetube14 with the plasma generated in the processing region. In another aspect, thetube14 may be separately biased by use of a power supply (seeFIG. 2) which is in electrical communication with thetubes14. In this configuration the secondelectrical component50H may act as an insulator.
While the foregoing is directed to embodiments of the present invention, other and further embodiments of the invention may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.