Classifications

• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32009—Arrangements for generation of plasma specially adapted for examination or treatment of objects, e.g. plasma sources
  • H01J37/32082—Radio frequency generated discharge
  • H01J37/321—Radio frequency generated discharge the radio frequency energy being inductively coupled to the plasma
• • C—CHEMISTRY; METALLURGY
  • C23—COATING METALLIC MATERIAL; COATING MATERIAL WITH METALLIC MATERIAL; CHEMICAL SURFACE TREATMENT; DIFFUSION TREATMENT OF METALLIC MATERIAL; COATING BY VACUUM EVAPORATION, BY SPUTTERING, BY ION IMPLANTATION OR BY CHEMICAL VAPOUR DEPOSITION, IN GENERAL; INHIBITING CORROSION OF METALLIC MATERIAL OR INCRUSTATION IN GENERAL
  • C23C—COATING METALLIC MATERIAL; COATING MATERIAL WITH METALLIC MATERIAL; SURFACE TREATMENT OF METALLIC MATERIAL BY DIFFUSION INTO THE SURFACE, BY CHEMICAL CONVERSION OR SUBSTITUTION; COATING BY VACUUM EVAPORATION, BY SPUTTERING, BY ION IMPLANTATION OR BY CHEMICAL VAPOUR DEPOSITION, IN GENERAL
  • C23C14/00—Coating by vacuum evaporation, by sputtering or by ion implantation of the coating forming material
  • C23C14/22—Coating by vacuum evaporation, by sputtering or by ion implantation of the coating forming material characterised by the process of coating
  • C23C14/34—Sputtering
  • C23C14/35—Sputtering by application of a magnetic field, e.g. magnetron sputtering
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/32431—Constructional details of the reactor
  • H01J37/32623—Mechanical discharge control means
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J37/00—Discharge tubes with provision for introducing objects or material to be exposed to the discharge, e.g. for the purpose of examination or processing thereof
  • H01J37/32—Gas-filled discharge tubes
  • H01J37/34—Gas-filled discharge tubes operating with cathodic sputtering
  • H01J37/3402—Gas-filled discharge tubes operating with cathodic sputtering using supplementary magnetic fields
  • H01J37/3405—Magnetron sputtering
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01J—ELECTRIC DISCHARGE TUBES OR DISCHARGE LAMPS
  • H01J2237/00—Discharge tubes exposing object to beam, e.g. for analysis treatment, etching, imaging
  • H01J2237/02—Details
  • H01J2237/022—Avoiding or removing foreign or contaminating particles, debris or deposits on sample or tube

Definitions

• This invention relates to sputter coating, and more particularly, to a method and an apparatus for conducting Ionized Physical Vapor Deposition (IPVD) of coating material onto substrates.
• IPVD Ionized Physical Vapor Deposition
• VLSI and ULSI Very and Ultra Large Scale Integration
• contacts on the bottoms of such features often must be coated with liners, and the features often must be filled with conductive metals.
• PVD physical vapor deposition
• the deposition of films on the bottoms of narrow high aspect ratio features (apertures) by physical methods requires the achievement of a high degree of directionality in movement of the material to be deposited toward the substrate. Higher aspect ratio features require greater directionality.
• To effectively coat contacts, for example, on the bottoms of narrow high aspect ratio holes on the surface of a substrate it is necessary for the particles of coating material to move at angles to the normal that are not substantially larger than the angular openings of the features.
• CVD processes require equipment that is more complex and expensive than PVD process equipment.
• the CVD processes due to their chemical nature, often involve environmental factors and employ chemical precursors that can provide a source of device contamination, and typically these systems require higher frequencies of preventive maintenance resulting in non-productive down time.
• PVD processes are faster than CVD processes, yielding better throughput and therefore lower cost.
• CVD processes might not exist or may not be practical for many deposition materials, such as by requiring complex precursors and delivery systems which may preclude the CVD deposition.
• Acceptable manufacturing CVD processes exist for titanium, titanium nitride, and tungsten.
• CVD processes for aluminum, copper, tantalum, and tantalum nitride either do not exist or if they exist, they are immature or not commercially viable.
• CVD can subject devices that are partially formed on the substrates to heat for prolonged periods, which can cause migration and diffusion of material at material boundaries or can subject the devices to other heat induced damage or exceed the thermal budget for the process in question.
• a sputter coating process is typically carried out by placing a substrate and a target of high purity coating material into a vacuum chamber filled with an inert gas such as argon and creating a plasma in the gas.
• the plasma is typically generated by maintaining the target, either constantly or intermittently, at a negative potential, so that the target functions as a cathode to supply electrons that excite the gas in the chamber and form a plasma adjacent the target surface.
• the plasma generation is usually enhanced with a magnetron cathode assembly in which magnets behind the target trap these electrons over the surface of the target where they collide with atoms of the process gas, stripping electrons from atoms of gas and converting them into positive ions.
• the gas ions are accelerated toward the negatively charged target where they collide with the surface and eject from the target surface atoms and atomic clusters or particles of target material and secondary electrons.
• the secondary electrons play a major role in sustaining the plasma.
• the ejected particles of target material are neutral in charge and propagate through the vacuum space in various directions with some striking the substrate, to which they adhere to form the film.
• Increasingly narrower features and higher aspect ratio features on the substrates act to decrease the acceptance angle of the aperture, thereby shadowing the sides of the features, resulting in increasingly more of the incident particles being intercepted by the sides of, and areas around, the features with increasingly fewer of the particles available to deposit on the feature bottoms.
• Various schemes have been used for causing the propagating particles to move in straight lines toward and normal to the substrate surface.
• One approach involves the use of a physical collimator plate between the target and the substrate to achieve normal distribution of angles of incidence and improve incident particle directionality by intercepting particles directed at low angles to the collimator so that the particles that pass through the collimator are only those that are normal or nearly normal to the substrate.
• Another approach known as long-throw sputtering, calls for increasing the target to substrate spacing so that only particles moving normal or close to normal angles to the substrate travel the length of the chamber to strike the substrate.
• Collimators provide a source of particulate contamination because the intercepted particles are deposited on the collimator, where film builds up and tends to eventually flake off.
• Both the collimated deposition and long- throw schemes achieve directionality by eliminating material that is moving at low angles to the substrate, which drastically reduces the percentage of the sputtered material incident on the substrate and thereby substantially decreases the deposition rates. It also increases preventive maintenance, decreases target material utilization, and reduces throughput.
• IPVD Ionized Physical Vapor Deposition
• additional plasma is created in the gas in the space between the target and the substrate through which the sputtered particles pass on their way to the substrate.
• the additional plasmas have been formed in the space by various methods, such as by capacitively coupling RF energy into the chamber downstream of the target, or have been formed remote from the space by electron cyclotron resonance (ECR) or other microwave plasma generating techniques and then caused to flow into the space.
• ECR electron cyclotron resonance
• the particles of sputtered material passing through this space collide with the electrons or metastable neutrals of the ionized process gas. The collisions tend to strip electrons from the atoms of the sputtered particles leaving the particles positively charged.
• These positive ions of sputtered material are then electrically accelerated toward the substrate, for example, by application of a negative bias to the substrate.
• IPVD processes in the prior art indicate a number of drawbacks and problems which preclude their practical use in a manufacturing environment. Such processes have, for example, produced low overall efficiency. In particular, IPVD processes typically yield low deposition rates. Furthermore, the prior art processes have produced high levels of film contamination. In particular, with IPVD proposals of the prior art, the filling of high aspect ratio features has been found to deteriorate as sputtering power at the target increases. Such deterioration has limited the sputtering of aluminum alloys to less than 3 kW of DC power with a 12 inch magnetron target as compared to 12 to 30 kW that is typically achievable for such a target/magnetron assembly.
• the low sputtering power results in a low deposition rate resulting in low productivity and throughput as well as increased contamination, e.g., 10 to 40 minutes of sputtering time per wafer compared with the typical wafer processing times of about 45 seconds to one minute.
• the fractional ionization of sputtered material has been found to be low unless the apparatus is operated at a relatively high pressure in the sputtering chamber, such as 20 to 40 mTorr. With argon process gas, this pressure is higher than the desired sputtering pressures that are typically less than 15 mTorr or in the low millitorr range. The higher pressures have had a tendency to reduce the quality of deposited film properties and to increase film contamination.
• IPVD apparatus and method overcomes the drawbacks and problems of the prior art.
• IPVD apparatus that produces acceptably high overall efficiency, particularly high deposition rates, high sputtered material ionization efficiency and low contamination of the deposited film.
• an apparatus that produces a film of high uniformity and quality, while providing productivity sufficiently high for the process to be commercially useful.
• a primary objective of the present invention is to provide a method and apparatus for depositing thin films on the bottoms, and to some extent on the sides, of narrow high aspect ratio holes and trenches in VLSI and ULSI semiconductor wafers. It is also a primary objective of the present invention to provide a method and apparatus for conducting ionized physical vapor deposition that has a high overall efficiency, and particularly, that yields a high deposition rate at a wide range of pressure with a high ionization efficiency of the coating material. It is a further objective of the present invention to provide an effective IPVD method and apparatus that maintains low deposited film contamination. Yet, it is another objective of the present invention to provide an IPVD process and hardware which has low preventive maintenance requirements.
• a particular objective of the present invention is to provide an apparatus and method of IPVD that will allow the sputtering power at the target to be sustained at least at a moderate level and that will provide a high RF energy coupling efficiency into sputtered material without the need to maintain the chamber at a relatively high sputtering pressure.
• a further objective of the present invention is to provide a method and apparatus in which adverse interaction between the plasmas in the chamber and the electrode or element used to couple the
• RF energy into the plasma to ionize the sputtered material is kept low, particularly the sputtering of, the flaking of sputtered material from, or the potential shorting of the electrode.
• an IPVD apparatus and method in which a main plasma is formed adjacent to the target to sputter material from the target while an RF element couples energy into the PVD processing chamber to produce a secondary plasma in a volume of the chamber between the main plasma and a substrate.
• the secondary plasma is supplemental to the main plasma that is typically confined close to the sputtering target.
• the secondary plasma generally fills the chamber, but primarily occupies at least a portion of the space between the target and the substrate, thereby ionizing the sputtered material particles in flight while they are moving from the target so that the particles can be electrostatically accelerated toward the substrate in the ion-assisted deposition of the sputtered material onto the substrate.
• the ionized sputtered material is preferably accelerated toward the substrate by a negative bias applied to the substrate, which may be controlled to provide for optimal steering of the moving ions without damaging the surface of the wafer.
• the chamber may be further surrounded with permanent magnets or electromagnets that produce an axial magnetic field in the chamber between the substrate and the target to aid in confining the ionized sputtered particles in paths parallel to the axis of the chamber and normal to the surface of the substrate.
• the RF ionization energy coupling element may be an RF electrode, preferably an inductive element, such as, for example, one to a plurality of coils surrounding the chamber.
• the RF element may either be positioned within the chamber, preferably insulated from the chamber process gas, or positioned outside of the chamber.
• the preferred apparatus is also provided with a protective structure comprising an electrically non- conductive and non-magnetic dielectric material that protects or isolates the RF element from adverse interaction with plasmas in the chamber, such as interaction with the main plasma as well as with the secondary plasma produced by the RF element.
• the protective structure is such that sputtered material that impinges upon it, if any, adheres to the structure in such a way that it does not tend to flake off of the structure to result in a source of contamination.
• the parts of the protective structure are further preferably configured to prevent eddy currents therein or in layers of sputtered material deposited thereon and to prevent electrostatic shielding of the RF element.
• an RF coil surrounds the chamber behind the protective structure, which forms part of the vacuum tight inner wall of the chamber surrounding the processing space, with the RF coil covered with an external conductive enclosure.
• the RF coil is within the vacuum of the processing chamber outside of and downstream of the target periphery, with the protective structure separating the RF coil from interaction with the plasmas.
• an RF coil is provided that is covered with protective insulating material, either with a solid insulator that completely covers the coil conductor or with a slotted or segmented insulator having slots sufficiently narrow to discourage the formation of plasma adjacent the conductor.
• the RF coil and protective structure are preferably cylindrical in shape and surround the processing space.
• the preferred apparatus further includes a shield array provided to shield the protective structure so that the function of the protective structure is not compromised by the effects of sputtered material deposition thereon.
• a shield array provided to shield the protective structure so that the function of the protective structure is not compromised by the effects of sputtered material deposition thereon.
• the RF element comprises a helical coil surrounding the chamber from behind a mostly cylindrical quartz window utilized as the protective structure.
• the mostly cylindrical quartz window may form part of the vacuum tight inner wall of the chamber, or it may be in the form of an insulator encasing the coil inside of the chamber, or in some other form that isolates the coil conductor from the process gas.
• a substantially cylindrical shield is provided that surrounds the chamber in close proximity to the window that separates the coil from a PVD processing chamber.
• the shield is slit preferably in a direction parallel to its axis of the chamber.
• close proximity is meant spaced from the window a distance that is sufficiently short as to prevent the formation of plasma between the shield and the window.
• the slit shield follows the shape of the dielectric window that separates the coil from the vacuum chamber and process gas.
• the shield prevents the deposit of coating material onto the window. When the material is electrically conductive, electrical short circuiting of the coil could occur preventing RF energy from being transmitted into the chamber.
• the shield is preferably slit in such a way as to prevent the shield itself from providing a circumferential path around the chamber in which circumferential currents could be induced, which would consume energy from the RF coil and detract from the efficiency of energy coupled into the supplemental plasma.
• the shield is further extended sufficiently far in an axial direction to short axial electrical fields across the RF coil and thereby optimizing the efficiency of the inductive coupling of energy into the plasma and reducing capacitive components of the coupled energy.
• the shield is maintained in close spaced proximity to the window so as to prevent generation of plasma behind the shield so that the plasma is more efficiently generated in the space through which the sputtered particles travel.
• this spacing of the shield from the window is not greater than the mean free path of atoms of the process gas or the minimum diffusion length of the plasma in the space.
• the slit in the shield is sufficiently wide to allow the formation of plasma therein, so that the plasma will continually remove by resputtering any deposition of coating material on the window as a result of material from the source passing through the slit.
• the location and configuration of the shield relative to the protective covering of the coil contributes to a high efficiency of plasma generation in the space of the chamber, avoiding losses due to the generation of plasma between the shield and the coil. As a result, a high ionization efficiency of the sputtered material is provided.
• plasma generation in ineffective areas of the chamber, such as between shield structure and the coil-protecting insulator or window, is prevented, and the loss of ionization efficiency thereby is avoided.
• an enclosure, a dielectric window and solid or segmented insulation is used, alone or in combination, to mutually protect the RF element from the plasmas and the sputtered material.
• the shield array is preferably in the form of a plurality of shield sections, which may be biased to control the sputtering thereof by the plasmas.
• the shield array has a plurality of gaps to at least partially electrically interrupt the shield sections to prevent induced eddy currents from consuming the energy and countering the coupling of energy to the plasma.
• the individual shield sections are preferably electrically separate so that they may be individually biased to optimize the uniformity of coating on, and the directionality of ionized material onto, the substrate. Spaces between the shield segments facilitate the propagation of plasma from behind the shield into the processing space.
• a helical coil surrounds the chamber behind a protective structure. So arranged, the coil is protected from contact with the plasmas formed within the chamber.
• An array of shield sections is provided within the chamber, also surrounding the space between the target and substrate, and preferably biased to control the sputtering thereof by the plasmas.
• the shield array has a plurality of gaps to at least partially electrically separate the shield sections, in part to prevent induced eddy currents from consuming the energy and countering the coupling of energy to the plasma.
• the shield sections are configured and oriented, and the gaps are defined, so that the shield sections shadow the protective structure from the target while minimally affecting the coupling of energy from the coil and the formation and location of the secondary plasma.
• the shield array is situated relative to the protective structure so that coating does not build up on it that would support eddy currents therein or that would produce electrostatic shielding of the RF coil.
• no part of the target is visible from any part of the protective structure, and if any part of the protective structure can see the target so as to accumulate a coating of conductive sputtering material, the coated area would not be so shaped as to allow it to support eddy currents or cause a significant shielding of the coil.
• the protective structure comprises a dielectric window.
• the shield array is formed of angled sections that collectively entirely block all paths between the target and the window.
• the sections are also angled so that much of the volume of the space between the main plasma and the substrate is visible from the coil. As such, the window is protected from sputter deposition from the target while providing the greatest effective coupling of energy to form the secondary plasma for ionizing the sputtered material.
• the shield sections are preferably spaced from the window and have sufficient space between adjacent sections to allow some part of the coil to be in sight of the volume of the chamber where the secondary plasma is desired to be formed, so that plasma can be formed adjacent the window and extend into the volume where sputtered material can be ionized.
• the shield array is formed of a plurality of axially spaced frusto-conical sections positioned inside of the window and inclined at an angle generally perpendicular to a path from the target to the window.
• the shield sections may be inclined at the same angle relative to the axis of the chamber or with the sections inclined at different angles, for example, with the sections at greater distances from the target at smaller angles to the axis.
• the sections do not shadow adjacent sections from every point on the target, although a minimal overlap may tend to reduce scattered sputtered particles from striking the window.
• the shield sections are preferably further segmented circumferentially by gaps in the sections, thereby interrupting potential induced current paths.
• the shield array is formed of a plurality of flat or slightly curved axially extending rectangular blades circumferentially spaced around the chamber inside of the window. These sections are spaced from the window, and are each inclined at an angle relative to the radius of the chamber to collectively shadow the window, or at least the portion of the window substantially within the field of the coil, from the entire area of the target, but to allow some part of the coil to be in sight of the volume of the chamber where the secondary plasma is desired to be formed. In this way a secondary plasma can form adjacent the window and readily extend into the volume through which sputtered particles will pass.
• the shield sections of this embodiment are preferably inclined at the same angle relative to the radii of the chamber.
• the sections do not shadow adjacent sections from every point on the target, although a minimal overlap may be effective to further reduce coating of the window.
• the shield sections are preferably circumferentially spaced from each other and spaced from the target and the substrate by a distance equal to at least the mean free path of molecules in the gas in the vacuum chamber.
• sputtering power can be maintained at a high level, thereby maintaining a high deposition rate and a high sputtered material ionization rate.
• the present invention also provides advantages in PVD processes that employ evaporation sources or other sources of vaporized material that are deposited by substantially physical techniques. Reactive processes and physical processes that are enhanced by or otherwise include chemical deposition of the material may also benefit from the present invention.
• the invention has particular utility in connection with the deposition of metal films, but can also provide advantages in the deposition of other materials, particularly oxides and nitrides.
• Fig. 1 is an elevational diagrammatic representation of a IPVD sputtering apparatus according to one embodiment of the present invention.
• Fig. 1A is an enlarged elevational diagram of a portion of Fig. 1 illustrating an alternative form of protection for the coil.
• Fig. 2 is a perspective diagram of a shield of the apparatus of Fig. 1.
• Fig. 3 is a diagrammatic representation of a IPVD sputtering apparatus according to one embodiment of the present invention.
• Figs. 4A-4D are diagrams illustrating alternative coil configurations of the apparatus of Fig. 3.
• Fig. 5 is a diagrammatic representation of an IPVD sputtering apparatus having an alternative configuration of the secondary plasma RF coupling element and protective structure in contrast to that shown in Fig. 3.
• Fig. 6 is a diagrammatic representation of an IPVD sputtering apparatus having another alternative configuration of the secondary plasma RF coupling element and protective structure in contrast to those shown in Figs. 3 and 5.
• Figs. 7A-7D are diagrammatic representations illustrating forms of coil insulating protective structure in the alternative to that shown in the embodiment of Fig. 6.
• Fig. 8 is a diagrammatic representation of a IPVD sputtering apparatus according to one embodiment of the present invention.
• Fig. 9 is a diagram of a portion of Fig. 8 illustrating an alternative configuration of a shield array.
• Fig. 10 is a cross-sectional view of the shield array embodiment of Fig. 9 taken along the line 3-3. of
• Fig. 1 diagrammatically illustrates a sputter coating apparatus 10 according to principles of the present invention.
• the apparatus 10 includes a vacuum tight processing space 11 enclosed in a chamber 12. Mounted in the chamber 11 at one end thereof is a substrate support or susceptor 14 for supporting a semiconductor wafer 15 mounted thereon.
• the wafer 15, when mounted on the substrate support 14, is parallel to and faces a target 16.
• the target 16 is formed of a sputter coating material that is to be deposited as a thin film on the wafer 15.
• the processing space 11 is a generally cylindrical space that is maintained at an ultra high vacuum pressure level and is filled with a processing gas, such as argon, during processing.
• the space 11 lies in the chamber 12 between the substrate support 14 and the target 16.
• the target 16 is part of a cathode assembly 17 mounted in the chamber 12 at an end thereof opposite the substrate support 14.
• the cathode assembly 17 includes a target holder 18 to which the target 16 is secured.
• a magnet structure 19 is typically provided behind the target holder 18 on the opposite side thereof from the substrate support 14.
• a dark space shield 13 may also be provided around the periphery of the target 16.
• the magnet structure 19 preferably includes magnets that produce a closed magnetic tunnel over surface 21 of the target 16 that traps electrons given off into the chamber 12 by the cathode assembly 17 when it is electrically energized to a negative potential as is familiar to one skilled in the art.
• the magnet structure 19 may include fixed or rotating or otherwise moving magnets, which may be permanent or electromagnets, of any one of a number of magnetron sputtering assemblies known in the art.
• a power supply or source 20 of electrical energy is connected between the cathode assembly 17 and the wall of the chamber 12, which is usually grounded and serves as the system anode.
• the cathode assembly 17 is insulated from the wall of the chamber 12.
• the power supply 20 is preferably connected to the cathode assembly 17 through an RF filter 22.
• An auxiliary source of energy such as an RF generator 24 may also be optionally connected to the cathode assembly 17 through a matching network 25.
• a bias circuit 27 is also provided and connected to the substrate support 14 through a matching network 28. The bias circuit 27 applies a bias to a wafer 15 mounted on the substrate support 14.
• a bipolar DC supply or an RF supply can be used for this purpose.
• Power from the steady or pulsed DC power supply 20 and/or RF generator 24 produces a negative potential on the surface 21 which causes electrons to be emitted from surface 21 of the target 16.
• the emitted electrons remain trapped over the surface 21 by the magnetic field generated by the magnet structure 19 until they strike and ionize atoms of process gas in close proximity to the surface 21 of the target 16, forming a main plasma 23 adjacent to the target surface 21.
• This main plasma 23 becomes a source of positive ions of gas that are accelerated toward and against the negatively charged surface 21, where they eject particles of coating material from the target 16.
• the space 11 between the target surface 21 and the substrate support 14 can be considered as formed of two parts. One part is that primarily occupied by the plasma 23, which is shaped to produce a desired erosion pattern on the sputtering surface 21 of the target 16, while the second part of the space 11 is a remaining volume 26 that lies between the plasma 23 and the substrate 15 on the substrate support 14.
• the particles of sputtered material from the target 16 generally originate as electrically neutral particles that can propagate only by momentum through the space 11, where some, but not all, pass through the plasma 23 and the volume 26 to impinge upon the substrate 15.
• neutral sputtered particles passing through the plasma 23 are not ionized significantly since the plasma 23 occupies a small volume near target surface 21, and at operating pressures of interest, few collisions occur between the neutral sputtered particles and particles of the plasma 23.
• the neutral sputtered particles exit the plasma 23 mostly neutral and stay neutral until deposited as a thin film on substrate 15.
• VLSI and ULSI semiconductor device manufacturing For coating contacts at the bottom of high aspect ratio holes and other features and for metallizing holes by filling them with sputtered conductive material, it is highly preferred in VLSI and ULSI semiconductor device manufacturing that the particles impinge onto the substrate surface in a narrow angular distribution around the normal to the substrate, so that they can proceed directly into the features and onto the feature bottoms, and not strike or be shadowed by the feature sides. This perpendicular impingement of particles onto the substrate is facilitated in the apparatus 10 by ionizing the sputtered particles as they pass through the volume 26, so that the particles develop an electrical charge.
• the particles can be electrostatically accelerated or otherwise electrically or magnetically directed into paths that are parallel to the axis of the chamber and perpendicular to the surface of the substrate 15.
• Such a process is known in the art as ionized physical vapor deposition (IPVD) or ion-assisted sputter coating.
• in-flight ionization of sputtered particles in the space 26 is carried out by reactively and preferably inductively coupling RF energy into the volume 26 by provision of an RF element that surrounds the volume 26 but does not occupy the space 11.
• the RF element is preferably in the form of a helical coil assembly 30, though coil configurations other than helical may be used. Examples of possible configurations 30a-30d of coil assemblies are illustrated in Figs. 4A-4D.
• the configurations of the coil assembly 30 should include coils, windings and/or coil and winding arrangements.
• the coil assembly 30 inductively couples energy into the process gas in the volume 26 thereby forming a secondary plasma 29 that generally fills the volume 26 and is distinct from the main plasma 23.
• An RF generator 32 preferably operative in, but not limited to, the range of from 0.1 or 0.2 MHz to 60 or 80 MHz, is connected to the coil assembly 30 through a matching network 33 to provide the energy to the coil assembly 30 to form the secondary plasma 29 in the volume 26.
• a source of processing gas 40 is connected to the chamber 11, through a flow control device 41.
• the supply gas 40 is typically an inert gas such as argon.
• additional gases such as nitrogen and oxygen can be introduced through auxiliary flow controllers.
• a high vacuum pump 39 is also connected to the chamber 12 to pump the chamber 12 to a vacuum level in the milli Torr or sub-milli Torr range. Pressures in the 5 to 50 milli Torr range are preferred.
• the pump 39 maintains the ultra high vacuum with a flow rate of process gas in a range of 5 to 300 standard cubic centimeters per second (seem).
• the apparatus 10 also includes a main controller 50 that is preferably a microprocessor-based programmable controller operative to sequence and control the operation of the components discussed above.
• the controller 50 has outputs for controlling the energization of the cathode power supplies 20 and 24, the substrate bias power supply 27, the RF generator 32 for energizing the secondary plasma generating element that is the coil assembly 30, the gas flow control 41, the pump 39 and other controllable components of the apparatus 10.
• an electrical potential gradient may be maintained in the plasma sheath in front of substrate support 14 by negatively biasing the substrate 15 relative to the secondary plasma 29 with a bias power supply 27 connected to the substrate support 14 through a matching network 28, to provide force to accelerate the positively ionized sputtered particles toward and onto the substrate surface.
• a bipolar DC supply or an RF supply can be used for this purpose.
• a magnet 80 may be provided around the chamber 12 to produce a magnetic field in the axial direction in the chamber 12.
• the magnet 80 may be an electromagnet or formed of one or more permanent magnets. The fields from the magnet 80 cause the charged particles to gyrate about the lines, thereby increasing their confinement in the radial direction. In the presence of an axial electric field, the charged particles can be directed in the axial direction, moving toward the substrate and minimizing radial losses.
• a protective structure that prevents the plasmas 23 and 29 from contacting and electrically interacting with the coil assembly 30.
• This structure is an electrically non-conductive material that does not impede the magnetic field surrounding the coil assembly 30 from reaching into the volume 26.
• One preferred form of protective structure is that of a window 60, made of a vacuum-compatible dielectric material such as quartz, in the wall of the chamber 12, that is mounted to form a vacuum-tight seal with the chamber wall.
• the window 60 may be a single piece of electrically-insulating and magnetically-transparent material or it may be formed in joined segments thereof, to form a generally cylindrical protective structure.
• the coil assembly 30 depicted in the foregoing embodiments is wound around the chamber 12, preferably outside of the window 60.
• a conductive metal enclosure 61 which forms a sealed cavity 62, which isolates the coil assembly 30 and also prevents electromagnetic energy from radiating from the coil assembly 30 and from within the chamber 12 to the outside of the chamber 12.
• the cavity 62 may be sealed from the chamber 11 but may be in communication with the outside atmosphere or it may be filled with an inert gas, at atmospheric or low pressure, such that formation of a plasma is not supported by the gas in the cavity 62 when the coil assembly 30 is energized.
• the window 60 itself is not electrically conductive, it is susceptible to the accumulation of a coating of conductive material sputtered from the target 16. Electrical conductivity in or on the window 60 supports the induction of currents around the chamber which reduce, cancel or otherwise undermine the effectiveness of the RF coupling of energy from the coil assembly 30 to the secondary plasma 29 in the volume 26. Such conductivity of coating on the window 60, particularly in the azimuthal (circumferential) direction, that is, a direction that extends around the chamber 12, produces an inductively coupled short circuit, can negate all or much of the energy inductively coupled into the volume 26.
• the preferred apparatus further includes a shield array, various embodiments of which are described below.
• Fig. 1 illustrates a slit cylindrical shield 100 provided between the space 11 and the window 60, in close proximity to the inside surface of the window 60.
• the shield 100 shadows the window 60 from material sputtered from the target 16, and preferably blocks all direct line-of-sight paths between any point on the surface 21 of the target 16 and the window 60.
• the shield 100 has a longitudinal slit 103 therein that is parallel to the axis of the chamber 12. Shields with a single or plurality of slits fashioned to interrupt circumferential currents can also be used.
• the slit 103 of the preferred embodiment substantially interrupts circumferential paths in the shield 100 around the chamber 12. This prevents the induction of circumferential or azimuthal currents in the shield 100.
• the shield 100 has an axial extent beyond the axial extent of the coil assembly 30 that reaches substantially the full effective axial extent of the electric field from the coil assembly 30.
• the electrically conductive shield 100 effectively suppresses electric fields in the secondary plasma 29 that are parallel to the axis of the chamber 12, preventing such axial electric fields that would capacitively shield the coil assembly 30 from the volume 26 and thereby undermine the coupling efficiency of energy to the volume 26 from the coil assembly 30.
• the shield 100 extend axially from behind the plane of the surface 21 of the target 16 to beyond the window 60 and coil assembly 30. With this configuration, the shield 100 more effectively shorts out axial electric fields in the secondary plasma 29, thereby enhancing the inductive coupling of energy from the coil assembly 30 into the secondary plasma 29.
• the preferred embodiment of the invention also produces a high coupling efficiency of energy from the coil assembly 30 into the volume 26 due to a close spacing of the shield 100 from the window 60.
• This spacing is maintained at a distance that is preferably not more than the mean free path of atoms or molecules in the gas or the minimum diffusion length of the secondary plasma 29 within the chamber 12.
• This close shield-to-window spacing is in contrast to other embodiments described below, which permit formation of plasma adjacent a window or coil protecting non-conductive structure and behind any shield structure that is provided. Avoiding plasma formation behind the window has a tendency of increasing the percentage of energy from the coil or other plasma-generating electrode into the volume through which the sputtered particles pass, thereby increasing the effective plasma and thus the ionization efficiency of the sputtered material.
• processing gas pressures in the range of about 5 to 50 milli-Torr will be used.
• the mean free path of argon gas at such pressures is from 11 mm to 1.0 mm, respectively.
• the preferred spacing of the shield 100 from the window 60 is approximately 1.0 to 15 mm.
• the slit 103 is preferably made greater than approximately 15 mm in width.
• the width of the slit 103 is sufficiently wide to allow the secondary plasma 29 to form in the slit 103 in order to clean sputtered material that might deposit on the edges of the shield 100 adjacent the slit 103, or on the window 60 as a result of sputtered material that passes through the slit 103.
• Such plasma 29 that forms in the slit 103 will extend against the window 60 in the vicinity of the slit 103 and continuously remove, by resputtering the material that deposits on the window 60 at the slit 103.
• the coil assembly 30 may be alternatively embedded in an insulation block 66 within the chamber 12, as illustrated in Fig. 1A, where the insulation block 66 functions in a manner similar to that of the window 60 to isolate the coil assembly 30 from the plasma in the chamber 11 and from sputtered material.
• the shield 100 is configured relative to the insulation 66 in the same way it is configured relative to the window 60, in Fig. 1.
• Fig. 3 illustrates a shield array 200, which is an alternative to shield 100 of Fig. 1, but is provided between the space 11 and the window 60 in less close proximity to the inside surface of the window 60.
• the shield array 200 at least partially shadows the window 60 from material sputtered from the target 16, but has sufficient spaces or gaps 204 therein to facilitate the coupling of energy from the coil assembly 30 into the volume 26.
• the shield array 200 is preferably a plurality of individual shields or shield segments 202 that shadow at least axial strips of the window 60 so that circumferential conductive paths are not formed by a coating of sputtered material.
• the gaps 204 are configured to substantially interrupt circumferential current paths in the shield array 200 and are configured to extend entirely or partially across the axial dimension of the array 200.
• the shield segments 202 should be made of a metal or other material selected to retain sputtered material coating thereon when such coating forms on the shield segments 202. Otherwise, such deposits will flake off and cause contamination of the chamber 12 and of wafers 15 being processed.
• the shield segments 202 may be electrically biased.
• the segments 202 are also preferably individually biased with their biases separately controllable for use in optimizing the distribution of film being deposited on the substrate, such as by optimizing the uniformity of coating on, and the directionality of ionized material onto, the substrate 15.
• the gaps 204 will completely separate and electrically isolate each of the separately biased shield segments 202 from each other.
• the bias is provided by a generator 206 that is connected through a filter or matching network 207, with each shield separately connected through a current limiting resistor 208.
• the resistors 208 may be variable or other means may be provided for individually controlling the bias of the shield segments 202 in response to the controller 50.
• Fig. 5 illustrates an alternative embodiment 10a to the apparatus 10, in which a coil assembly 30 is situated in the vacuum chamber 12, inside the wall of the chamber 12 but still outside of the space 11.
• the protective structure is in the form of a window 60a that is situated between the coil assembly 30 and the space 11, on an enclosure 61a that extends inwardly from the inside of the wall of the chamber 12 and encloses the coil assembly 30.
• the enclosure 61a contains ports 62 that vent the inside of the enclosure in which the coil assembly 30 lies to the vacuum of the chamber 12.
• the shield array 200 is situated as in the embodiment above to shadow the window 60a from the target 16.
• the 6 utilizes a protective structure in the form of an insulative coating 86 to cover the conductor of the coil assembly 30.
• the coil assembly 30 is situated in the chamber 12 outside of the space 11, surrounding the volume 26.
• the shield array 200 is situated as in the embodiments above to shadow the insulating layer 86 from the target 16.
• the insulator 86 may be in any one of a number of forms, such as that of a solid insulator 86a that totally covers the surface of the conductor of the coil assembly 30, as illustrated in Fig. 7A, or may be in the form of a plurality of discrete insulator segments 86b, as illustrated in Fig.7B.
• gaps 87 between the segments facilitate the effectiveness of the coil assembly 30 while the narrowness of the gaps 87 is preferably maintained to less than the mean free path of the gas molecules in the chamber 12 so that it will not communicate plasma to the conductor of the coil assembly 30.
• insulating material may encase the coil assembly 30, such as insulators 86c and 86d of Figs.7C and 7D, respectively.
• the shield array 300 illustrates a shield array 300 provided between the space 11 and the window 60, in close proximity to the inside surface of the window 60.
• the shield array 300 shadows the window 60 from material sputtered from the target 16, and preferably blocks all direct line-of-sight paths between any point on the surface 21 of the target 16 and the window 60. Nonetheless, according to this embodiment of the present invention, the shield array 300 provides spaces or gaps 305 therein that provide a substantially uninterrupted area between the coil assembly 30 behind the window 60 and the volume 26 into which the plasma 29 is to be coupled, thereby facilitating the coupling of energy from the coil assembly 30 and the propagating plasma, into the volume 26.
• the shield array 300 is preferably in the form of a plurality of shields or shield segments 302 that collectively shadow the window 60 from every point on the target 16. This shadowing eliminates most of the tendency for a build-up of sputtered film to form on the window 60. Accordingly, neither conductive paths nor electrostatic shielding occurs.
• the shield segments 302 are frusto conical in shape, with the insides thereof forming an angle #with a plane parallel to the surface 21 of the target 16 and to the substrate 15 on the support 14.
• the angles ⁇ of each shield segment 302 may be the same, but the effectiveness of the shield segments 302 can be enhanced or optimized by decreasing the angles ⁇ as the distance between the segments 302 and the target 16 increases, so that the top surface 303 of the segment 302 directly faces the target 16, providing maximum shadowing of the target 16 from the window 60 for a given segment area.
• the shield segments 302 lie outside of the space 11 and circumferentially surround the space 11, and are axially- spaced from each other by the circumferential gaps or spaces 305.
• the maximum widths of the gaps 305 are the widest gaps S that still completely shadow the surface of the target 16 from the window 60, as illustrated by lines 79, so that circumferential bands of window 60 are not exposed to the target 16 so as to deposit an annular conductive strip around the window 60. Therefore, the maximum width S of the gaps 305 can be greater at greater distances from the target 16.
• the gaps 305 may be narrower, but should not be less than the mean free path of atoms of the process gas at the temperature and pressure of the chamber 12, and should be optimally spaced, given the process conditions, to facilitate the most efficient diffusion of the RF plasma into the volume 26.
• the segments 302 each have heights H that may be the same for each segment 302 or varied to optimize the shadowing and spaces between the segments 302.
• the ideal number of the shield segments 302 is dependent on the geometry of the chamber 12. While a single shield segment 302 can be used, typically two to six segments 302 will be employed. The number of segments 302 should be limited and the cumulative shield segment area should be minimized to minimize RF plasma losses. Furthermore, to prevent the formation of closed circumferential paths for eddy or other currents induced by the RF coil assembly 30, the segments 302 should have at least one gap 304 interrupting each of them.
• the gaps 304 of the adjacent segments 302 may be in line as shown, or preferably may be staggered to prevent the deposition of a continuous line of film axially across the window 60.
• the gaps 304 should be - In sufficiently wide to prevent arcing, which, depending on the process parameters, will require widths of about l/4th to 1 inch.
• the gaps 304 are configured to substantially interrupt current paths in the shield array 300 and are configured to extend entirely or partially across the axial dimension of the array 300.
• the shields segments 302 may be made of a metal, a vacuum-compatible dielectric material such as a ceramic or quartz, or some other compatible material selected to retain sputtered material coating thereon when such coating forms on the shields segments 302. Otherwise, such deposits will flake off and cause contamination of the chamber 12 and of wafers 15 being processed.
• the shield segments 300 may be electrically-biased, and made of metal for that purpose.
• the shield segments 302 are also preferably individually-biased with their biases separately controllable for use in optimizing the distribution of film being deposited on the substrate, such as by optimizing the uniformity of coating on, and the directionality of ionized material onto, the substrate 15.
• the gaps 304 will completely separate and electrically isolate each of the separately-biased shield segments 302 from each other.
• the bias is provided by a generator 306 that is connected through a filter or matching network 307, with each shield segment 302 separately connected through a current limiting resistor 308.
• the resistors 308 may be variable or other means may be provided for individually controlling the bias of the shield segments 302 in response to the controller 50.
• the shield array 300 may be realized by an alternative embodiment having a shield array 300a, as illustrated in Figs. 9 and 10.
• the array 300a is formed of a plurality of flat or slightly curved rectangular segments 302a that are arranged as an array of blades or vanes around the perimeter of the space 11 inside of the window 60.
• the segments 302a are circumferentially spaced from each other by axial spaces or slots 304a, which provide space between the segments 302a for plasma to couple into the volume 26 as well as interrupt potential circumferential current paths around the array 300a.
• the orientation of the segments 302a is such that they each define an angle ⁇ W ⁇ h a radial plane 311 through axis 312 of chamber 12.
• the spacing W between adjacent ones of the shield segments 302a and between the shield segments 302a and the window 60 should not be less than the mean free path of the gas in the chamber 12 so that plasma can effectively form adjacent the window 60 and propagate into the volume 26 in the gaps 304a between the segments 302a.
• the segments 302a are sufficiently long in the axial direction as to prevent any circumferential band of coating to form on the window 60 at the ends of the segments 302a, and are preferably set at angle and spaces W relative to each other to shadow the entire target 16 at the window 60.
• the shield arrays described above may be used with a dielectric window provided inside of the chamber or with a coil in the chamber that is protected from the plasmas by insulation.

Applications Claiming Priority (7)

Publications (1)

Family

ID=27420268

Family Applications (1)

Country Status (7)

Families Citing this family (46)

Family Cites Families (8)

• 1998
  • 1998-04-21 WO PCT/US1998/008033 patent/WO1998048444A1/en active IP Right Grant
  • 1998-04-21 AU AU69779/98A patent/AU6977998A/en not_active Abandoned
  • 1998-04-21 KR KR1019997009690A patent/KR100322330B1/ko not_active IP Right Cessation
  • 1998-04-21 JP JP54628098A patent/JP3775689B2/ja not_active Expired - Fee Related
  • 1998-04-21 CN CNB988043351A patent/CN1228810C/zh not_active Expired - Fee Related
  • 1998-04-21 EP EP98915648A patent/EP0978138A1/de not_active Withdrawn
  • 1998-06-10 TW TW087109240A patent/TW460602B/zh not_active IP Right Cessation

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events