UNITED STATES PATENT AND TRADEMARK OFFICE
`___________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`___________________
`
`APPLIED MATERIALS, INC.
`Petitioner,
`
`v.
`
`DEMARAY LLC
`Patent Owner.
`___________________
`
`Case IPR2021-00106
`Patent No. 7,381,657
`___________________
`
`DECLARATION OF DR. ALEXANDER GLEW IN SUPPORT OF PATENT
`OWNER’S PRELIMINARY RESPONSE
`
`Mail Stop “PATENT BOARD”
`Patent Trial and Appeal Board
`U.S. Patent and Trademark Office
`P.O. Box 1450
`Alexandria, VA 22313-1450
`
`Demaray Ex. 2002-p. 1
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`TABLE OF CONTENTS
`
`I.
`
`II.
`
`Qualification ................................................................................................. 1
`
`Knowledge of a POSITA .............................................................................. 3
`
`III. Technology Background ............................................................................... 5
`
`A. Magnetron Sputtering System ............................................................ 5
`
`1.
`
`2.
`
`3.
`
`4.
`
`Targeted Reaction On Substrates Instead Of Targets .............. 7
`
`Controlled Flow Rate Of Reactive Gases .............................. 10
`
`Arc Suppression ..................................................................... 11
`
`Synergy of different techniques ............................................. 13
`
`Reactor Systems ............................................................................... 13
`
`The Importance Of Filter Type For The Claimed Reactor
`System .............................................................................................. 19
`
`B.
`
`C.
`
`IV. Cited References ......................................................................................... 23
`
`A.
`
`B.
`
`C.
`
`Barber ............................................................................................... 23
`
`Licata ................................................................................................ 24
`
`Kelly ................................................................................................. 28
`
`D. Hirose ............................................................................................... 30
`
`E.
`
`F.
`
`Collins............................................................................................... 35
`
`Other Filter References .................................................................... 38
`
`1.
`
`2.
`
`3.
`
`Ex. 1023 (“Miller”) ................................................................ 38
`
`Exhibits 1009, 1011 & 1012 .................................................. 41
`
`Ex. 1013 ................................................................................. 42
`
`
`
`
`
`- i -
`
`
`
`Demaray Ex. 2002-p. 2
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`4.
`
`5.
`
`6.
`
`7.
`
`8.
`
`9.
`
`Ex. 1057 (“Sill”) .................................................................... 42
`
`Ex. 1058 (“Zennamo”) ........................................................... 44
`
`Ex. 1016 (“Celestino”) ........................................................... 45
`
`Ex. 1014 ................................................................................. 47
`
`Ex. 1017 ................................................................................. 48
`
`Ex. 1018 ................................................................................. 49
`
`10. Ex. 1019 ................................................................................. 50
`
`11. Ex. 1020 ................................................................................. 52
`
`12. Ex. 1021 ................................................................................. 53
`
`13. Exs. 1024, 1025, 1026, 1062 and 1067 .................................. 55
`
`V.
`
`The ’657 Patent ........................................................................................... 56
`
`
`
`
`
`
`
`- ii -
`
`
`
`Demaray Ex. 2002-p. 3
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`I.
`
`Qualification
`
`1. My name is Alexander D. Glew, Ph.D., P.E. My qualifications are
`
`summarized below and are addressed more fully in my CV attached as EXHIBIT
`
`A.
`
`2.
`
`For 33 years I have been involved with engineering practice. A large
`
`portion of my work has involved semiconductor fabrication, including product
`
`design, semiconductor device analysis, semiconductor equipment design and
`
`analysis, thin film processing, equipment, characterization, and project
`
`development. I was intimately involved in this field during the time of the patents
`
`at issue in this case.
`
`3.
`
`I received my Bachelor of Science in Mechanical Engineering from
`
`the University of California, Berkeley, in 1985; I received my Master of Science in
`
`Mechanical Engineering from the University of California, Berkeley, in 1987; I
`
`received my Master of Science in Materials Science and Engineering from
`
`Stanford University in 1995.
`
`4.
`
`I received my Doctor of Philosophy in Materials Science and
`
`Engineering from Stanford University in 2003. My dissertation involved plasma
`
`CVD of diamond-like carbon, fluorinated diamond-like carbon, and low k
`
`dielectrics.
`
`
`
`
`
`- 1 -
`
`
`
`Demaray Ex. 2002-p. 4
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`5.
`
`I began my career with Applied Materials, Inc., one of the leading
`
`companies that supplies equipment for semiconductor manufacturer. My services
`
`to Applied Materials included various engineering roles: product development,
`
`project management, core technology, and supplier quality management. I
`
`remained at Applied Materials for ten years.
`
`6.
`
`I hold six patents on technologies such as tungsten chemical vapor
`
`deposition, and ultra-high purity and high-temperature valves, and thin film heater
`
`and chuck design for processing chambers. I have authored or co-authored over
`
`nine articles, presentations, and seminars on topics including semiconductor thin
`
`film processing and diamond like carbon.
`
`7.
`
`I have been asked by Demaray LLC (“Patent Owner”) to explain the
`
`technologies involved in U.S. Pat. Nos. 7,381,657 and 7,544,276 (collectively “the
`
`Demaray Patents”), the technologies described in the cited references, the
`
`knowledge of a person of ordinary skill in the art at the time of the invention, and
`
`other pertinent facts and opinions regarding IPR2021-00103 through 106. For the
`
`purpose of this declaration, I apply the same skill level as proposed in the Petition,
`
`although I reserve the right to explain why this level is too high. I am being
`
`compensated for my work on this case at a fixed, hourly rate, plus reimbursement
`
`for expenses. My compensation does not depend on the outcome of this case or
`
`any issue in it, and I have no interest in this proceeding.
`
`
`
`
`
`- 2 -
`
`
`
`Demaray Ex. 2002-p. 5
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`II. Knowledge of a POSITA
`
`8. My testimony below is from the perspective of a person of ordinary
`
`skill in the art at the time of the invention. I apply Petitioner’s definition of such a
`
`skilled artisan for the purpose of this declaration, but I reserve the right to offer an
`
`alternative level later in the proceeding or in the Texas Litigations.
`
`9.
`
`I am familiar with the skill level and knowledge of the skilled artisan
`
`as defined by Petitioner because of my 33 years of experience in the industry and
`
`my interactions with such people.
`
`10.
`
`In general, these skilled artisans would know that, at the time of the
`
`invention, plasma processing was a complex and unpredictable process (and it still
`
`is today). They would know that filters were specifically designed for a given
`
`reactor system and thus could not just be applied to a totally different reactor
`
`system. For example, they would know that a filter coupled to an RF power supply
`
`was (and still is generally) designed to work with an RF matching circuit sized for
`
`the RF power supply and the connected reactor system to make sure the designed
`
`impedance was maintained for the plasma system. They would recognize that an
`
`RF matching circuit was designed to enable as much of the RF power supply
`
`output as possible to flow from the RF power source to the reactor and to minimize
`
`the amount of energy that would travel in the reverse direction from the reactor to
`
`the RF power source. They would know that an RF filter was designed to
`
`
`
`
`
`- 3 -
`
`
`
`Demaray Ex. 2002-p. 6
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`supplement the tuning offered by the RF matching circuit and they generally would
`
`not work without such an RF matching circuit.
`
`11. A skilled artisan would also know that an RF matching circuits would
`
`not work with a DC power source including pulsed DC power source because an
`
`RF matching circuit, functioning like a big capacitor coupled to an associated
`
`power supply, required a change in voltage output from the power supply while a
`
`DC power source generally outputted a constant voltage. They would know that,
`
`even a pulsed DC power supply would output a voltage that was constant most of
`
`the time and that the change in voltage potential was only designed to occur as the
`
`polarity of the DC power supply was switched or when the DC power supply was
`
`de-energized (or turned off). They would recognize this difference between DC
`
`and AC/RF power supplies, that is, while an AC or RF power supply’s voltage
`
`output would change constantly, a DC power supply’s voltage output was designed
`
`to be constant in general. As a result, a skilled artisan would know that a filter
`
`designed with an RF matching circuit would not have worked with a DC power
`
`supply (including a pulsed DC power supply) that had no associated RF matching
`
`circuit.
`
`12. A skilled artisan would also not have viewed bipolar pulsed DC
`
`power supplies as necessary or helpful if the surface of the sputtering target
`
`remained conductive. This is because, as Petitioner’s own reference suggested, at
`
`
`
`
`
`- 4 -
`
`
`
`Demaray Ex. 2002-p. 7
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`the time of the invention, it was thought that arcing was caused by charge
`
`accumulation on an insulating target surface that could not be effectively
`
`conducted away and the sudden discharge of those accumulated charges. Ex. 1036
`
`at 3 (“When an insulating material forms on the surface of the sputtering target
`
`during deposition, those insulating surfaces build up a charge and then discharge
`
`during dc reactive sputtering, which results in arcing.”). As a result, a skilled
`
`artisan would not view it necessary or beneficial to alternate the target surface from
`
`positive to negative using a pulsed DC power supply if the target surface remained
`
`conductive during deposition and arcing was not an issue.
`
`13. Below I first provide some background in the relevant sputtering
`
`technology and then discuss Petitioner’s references in detail.
`
`III. Technology Background
`
`A. Magnetron Sputtering System
`
`14. Fabrication of integrated circuits involves formation of defined
`
`features on a substrate, such as a silicon wafer. The formation of these defined
`
`features often begins with deposition of thin films of the appropriate materials,
`
`such as titanium, tungsten and titanium nitride as barrier or nucleation layers. Ex.
`
`1012, 1:10-24. In modern manufacturing, thin film are often deposited by physical
`
`vapor deposition (PVD), also referred to as sputtering.
`
`15. Sputtering involves the production of a vapors, which then deposit by
`
`
`
`
`
`- 5 -
`
`
`
`Demaray Ex. 2002-p. 8
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`physical means on the substrate. As the feature sizes on integrated circuit become
`
`smaller and smaller, the industry has gravitated towards plasma-assisted sputtering.
`
`16.
`
`In a plasma-assisted sputtering, “ions in a plasma are attracted to the
`
`target at sufficient energy….” Ex. 1012 at 1:10-16. The “target is bombarded with
`
`ions, and the atoms of the target material are mechanically ejected from the target
`
`and deposited onto a nearby substrate.” Ex. 1005, 2:1-5. An ion is a charged
`
`particle. One can use electrical attraction of the ion toward an electrically charged
`
`substrate to direct it in a more vertical trajectory toward the surface, which has
`
`certain advantages.
`
`17. Sputtering can also involve a reactive gas, which “is introduced into
`
`the deposition chamber and reacts with the target material to produce a film [on the
`
`target] that is sputtered onto the substrate….” Ex. 1005, 2:5-9. The freed material
`
`is deposited onto the substrate either directly or upon further reaction with the
`
`reactive gas. Id. This ejected processing involving reactive gases (such as
`
`nitrogen or oxygen) is known as reactive sputtering. Reactive sputtering
`
`traditionally suffered from low deposition rate, arcing due to forming of an
`
`insulating film on the target and poorer film quality due to varying stoichiometry
`
`over the course of the deposition. Ex. 1048 at 2.
`
`18. Several techniques were developed to improve the deposition rate as
`
`well as density and adhesion of film formed by reactive sputtering. Ex. 1036 at 1.
`
`
`
`
`
`- 6 -
`
`
`
`Demaray Ex. 2002-p. 9
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`Some techniques help direct the reaction away from the target and toward the
`
`substrate; some provide more precise control of the amount of reactive gases in the
`
`reactor (e.g., control of partial pressure of reactive gases); and still others help
`
`neutralize the charge build-up during deposition so as to reduce the incidence of
`
`arcing (e.g., bipolar pulsed DC power to the target).
`
`1.
`
`Targeted Reaction On Substrates Instead Of Targets
`
`19.
`
`“Magnetron sputtering is a principal method of depositing. Magnets
`
`have been coupled to targets to increase the directionality and deposition rates. In a
`
`magnetron sputtering systems, a magnetron having opposed magnetic poles is
`
`disposed at the back of the target “to generate a magnetic field close to and parallel
`
`to the front face of the target.” Ex. 1012, 1:35-37. The “magnetically-enhanced
`
`targets are used to confine the plasma discharge along a particular path and
`
`enhance the flow of target material.” Ex. 1005, 2:27-29. The magnetic field
`
`generated near the target surface “traps electrons, and, for charge neutrality in the
`
`plasma, additional argon ions are attracted into the region adjacent to the
`
`magnetron to form there a high-density plasma.” Ex. 1012 at 1:37-41. The higher
`
`concentration of plasma increases the sputtering rate. Id.
`
`20. Traditionally, “the strength of the inner and outer magnets is roughly
`
`equal,” and “the magnetron is said to be balanced.” Ex. 1036 at 1-2. In a balanced
`
`magnet, “most of the magnetic field lines will loop between the inner and out
`
`
`
`
`
`- 7 -
`
`
`
`Demaray Ex. 2002-p. 10
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`magnets as is shown” below. Ex. 1036 at 1-2. In such a system, any electrons
`
`escaping from the primary magnetic trap between the inner and outer magnets will
`
`go to anode and be lost. Id. The primary magnetic trap “is responsible for the
`
`formation of the dense plasma directly in front of the sputtering target and for the
`
`higher deposition rate of the magnetron cathode compared to a diode cathode.”
`
`Ex. 1036 at 1. But the location of the reaction also increases the likelihood of
`
`forming an insulating film on the target, therefore leading to arcing and variance in
`
`stoichiometry over the course of deposition.
`
`
`
`21.
`
`In contrast, in unbalanced magnetron, one set of the magnets—
`
`typically the outer magnets—is made stronger than the other, and “some field lines
`
`from the stronger magnets will radiate away from the magnetic surfaces as is
`
`shown” below. Id.
`
`
`
`
`
`- 8 -
`
`
`
`Demaray Ex. 2002-p. 11
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`
`
`22.
`
`In an unbalanced magnetron system, “the escaping energetic electrons
`
`are trapped by excess magnetic field lines, and the electron spiral along the field
`
`lines” to “undergo ionizing collisions with gas atoms.” Ex. 1036 at 1. “A
`
`secondary plasma is formed away from the target surface from these ionizing
`
`collisions and this secondary plasma can be used for ion-assisted deposition of the
`
`growing film.” Id. at 2. Current density on the substrate in an unbalanced
`
`magnetron system is greater than in a comparable balanced magnetron system. Id.
`
`23. One effect of the unbalanced magnetron system is to form secondary
`
`plasma away from the target and to energize substrate surface to stimulate reaction
`
`of the reactive gas directly on the substrate (as on the target). Strongly unbalanced
`
`magnetron, however, may create non-uniformity near the substrate.
`
`24. Several other techniques, such as thermally energizing the substrate,
`
`biasing the substrate to form a plasma near the substrate, providing a secondary
`
`
`
`
`
`- 9 -
`
`
`
`Demaray Ex. 2002-p. 12
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`plasma generation source for creating a plasma in the volume between the target
`
`and substrate and for creating a magnetic field line that is directed towards the
`
`substrate, also produce similar technical effects. These other techniques were
`
`described, for example, in Licata. Ex. 1010, 10:23-31, 7:11-25.
`
`2.
`
`Controlled Flow Rate Of Reactive Gases
`
`25. Better control of the deposited film’s stoichiometry as well as reaction
`
`rates may be achieved by a more precise control of the flow rate of the reactive
`
`gas. The goal is to control the amount of reactive gas in the reactor for a given
`
`reactor system.
`
`26. One technique, as described by Sproul, is via control of the partial
`
`pressure of the reactive gas at a desired set point. Ex. 1036 at 2-3. According to
`
`Sproul, when done correctly, TiN can be deposited as the same rate as Ti and it
`
`becomes possible to “produce the same compound material” with the repeatable
`
`stoichiometry “in every run.” Ex. 1036 at 3.
`
`27. Barber also provides a method for controlling the amount of reactive
`
`gas in a reactor. Barber’s method involves first determining a relationship between
`
`the chamber total pressure and the flow rate of the reactive gas such as nitrogen
`
`and oxygen with a preset flow rate of the inert gases such as argon. See Ex. 1005,
`
`8:49-56; 9:29-36. A cross-over point is then determined, where the cross-over
`
`point corresponds to the point at which the increase in total chamber pressure
`
`
`
`
`
`- 10 -
`
`
`
`Demaray Ex. 2002-p. 13
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`becomes non-linear with the flow of the reactive gases. Id.; see also 6:34-50.
`
`28. The reactive gases are then introduced into the reaction chamber at a
`
`flow rate that is about 3 sccm above the cross-over point. Id., 8:59-64, 9:35-36.
`
`Assuming ideal gas law applies, the partial pressure of the reactive gas is chamber
`
`pressure × molar ratio of reactive gas, with the molar ratio of reactive gas being
`
`(cid:3045)(cid:3032)(cid:3028)(cid:3030)(cid:3047)(cid:3036)(cid:3049)(cid:3032) (cid:3034)(cid:3028)(cid:3046) (cid:3033)(cid:3039)(cid:3042)(cid:3050) (cid:3045)(cid:3028)(cid:3047)(cid:3032)
`
`(cid:3045)(cid:3032)(cid:3028)(cid:3030)(cid:3047)(cid:3036)(cid:3049)(cid:3032) (cid:3034)(cid:3028)(cid:3046) (cid:3033)(cid:3039)(cid:3042)(cid:3050) (cid:3045)(cid:3028)(cid:3047)(cid:3032)(cid:2878)(cid:3002)(cid:3045) (cid:3033)(cid:3039)(cid:3042)(cid:3050) (cid:3045)(cid:3028)(cid:3047)(cid:3032)
`
`. Because for a given reactor system (e.g., at a
`
`preset argon flow rate, like below), a given reactive gas flow rate produces a given
`
`chamber pressure, setting the reactive gas flow rate to a fixed value effectively
`
`fixes the partial pressure of the reactive gas as well. In other words, Barber
`
`improves the deposition rate and film quality of the deposited insulating films
`
`effective by controlling the partial pressure of the reactive gases.
`
`3.
`
`Arc Suppression
`
`29. As mentioned before, one drawback associated with reactive
`
`
`
`
`
`
`
`- 11 -
`
`
`
`Demaray Ex. 2002-p. 14
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`sputtering is the build-up of an insulating film such as nitrides, oxides, oxynitrides
`
`on the target surface. Ex. 1048 at 1; Ex. 1036 at 2. The build-up of such an
`
`insulating film makes the target’s surface also more insulating; and “those
`
`insulating surfaces build up a charge and then discharge during dc reactive
`
`sputtering, which results in arcing.” Ex. 1036 at 3. The burst of high energy
`
`released in arcing can damage power supplies to target and eject particles, even
`
`liquid drops, from the target, (id.), resulting in contamination and poor film quality.
`
`30. Arcing occurs when the accumulated charge causes a breakthrough in
`
`the insulating film on the target surface. Each dielectric material has a dielectric
`
`breakdown strength characterized by voltage divided by distance. When a high
`
`level of electrons are emitted from the breakthrough area, (i.e., when discharge
`
`occurs), an arc appears. Arcing is an unpredictable process causing changing
`
`impedance in the reactor’s electrical circuits.
`
`31. RF power with alternating positive and negative voltages theoretically
`
`can alleviate the problem associated with charge accumulation on the insulating
`
`surface of the target. But because sputtering occurs only in the negative voltage
`
`regime, essentially half of the power is wasted. Ex. 1036 at 3.
`
`32. Bipolar pulsed DC power that alternates between negative and
`
`positive potentials has thus been suggested to suppress arcing. Ex. 1036 at 4; Ex.
`
`1048 at 2. In bipolar pulsed DC power systems, the electric potential formed
`
`
`
`
`
`- 12 -
`
`
`
`Demaray Ex. 2002-p. 15
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`between the cathode and the anode in the chamber is reversed periodically to
`
`prevent charge accumulating on the target insulating film. More specifically, the
`
`positive portion of the applied voltage neutralizes accumulation of positive charge
`
`on the surface of the target insulating film. The negative portion of the applied
`
`voltage, if sufficient, causes ions from the accumulated layer to sputter off the
`
`target material, removing ions and allowing them to accumulate on the substrate.
`
`4.
`
`Synergy of different techniques
`
`33. Magnetron sputtering is a complicated process. The effects of
`
`modifying reactor designs and process parameters are often unpredictable. For
`
`example, one paper stated that “[w]ithout [a] combination of pulsed dc power and
`
`partial pressure control, it really was not possible to reactively sputter Al2O3 at
`
`high deposition rates in a practical way.” Ex. 1036 at 5.
`
`34. As another example, the inventors noticed that using a pulsed DC
`
`power supply at the target may not by itself provide for the desired quality of
`
`deposited film. Ex. 1001, 5:66-6:6, 10:3-4. By applying an RF bias to the
`
`substrate during deposition, as discussed above, the directionality of the deposition
`
`is enhanced and a simultaneous deposition and etching of the film on the substrate
`
`help densify the films by eliminating columnar structures in the deposited film. Id.
`
`B. Reactor Systems
`
`35. There are different plasma-assisted deposition systems, some with a
`
`
`
`
`
`- 13 -
`
`
`
`Demaray Ex. 2002-p. 16
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`magnetron and others without. Some apply an RF power to the target, others apply
`
`continuous DC and yet others apply pulsed DC (unipolar or bipolar). Some apply
`
`an RF bias to the substrate, some apply a DC bias and still others may rely on self-
`
`bias. Some may include an RF coil to surround the reactor walls to generate a
`
`plasma field between the target and the substrate, while others do not.
`
`36. For example, Hirose involves a plasma processing apparatus for
`
`etching that apply one higher frequency (e.g., 60 MHz) RF power 14 to the target
`
`and apply another lower frequency (e.g., 2 MHz) RF bias 15 to the substrate. E.g.,
`
`Ex. 1006 at 5:26-30, 5:40-43. Hirose’s reactor does not include a magnetron. As
`
`with most RF power sources, matching circuits are coupled to the RF power
`
`sources for impedance matching to maximize power transfer and minimize signal
`
`reflection from the load. In Hirose, the filter 20 is tuned so that is slightly off
`
`resonance with the RF power source 15. See Ex.1005, abstract, 5:58-62.
`
`
`
`
`
`- 14 -
`
`
`
`Demaray Ex. 2002-p. 17
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`
`
`37. Collins is another example where RF power is applied to both the
`
`target and the substrate. Ex. 1071 at Fig. 23. In Collins, an induction field is
`
`generated from the coil antenna 145. Id., 16:1-5. An RF power 300 of frequency
`
`f1 is coupled to the ceiling 110 and an RF power 305 of frequency f2 is coupled to
`
`wafer pedestal 120. See Fig. 23. RF matching circuits are coupled to each of the
`
`RF sources; and “isolation filters 310 and 315 prevent the RF energy from either
`
`one of the RF power generators 300, 305 from reaching the other.” 24:7-9.
`
`Ground filters 320 and 325 are grounded to “permit each one of the ceiling and
`
`pedestal 110, 120 to return to ground the RF power radiated across the chamber
`
`
`
`
`
`- 15 -
`
`
`
`Demaray Ex. 2002-p. 18
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`100 by the other” but “prevent the RF power applied to either one of the ceiling
`
`and pedestals 110, 120 from being shorted directly to the ground.” Id., 24:9-14.
`
`Isolation filters 330, 335 coupled to matching circuits 280 and 290 prevent RF
`
`power from 305 and 300 respectively “from interfering with the operation of the
`
`impedance match circuit of the other.” Id., 24:32-38.
`
`38. Licata is an example of a magnetron IPVD system that applies a DC
`
`power to the target and an RF bias power to the substrate. An RF generator
`
`
`
`
`
`
`
`- 16 -
`
`
`
`Demaray Ex. 2002-p. 19
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`coupled to a coil generates a plasma in volume 26 between the target and substrate
`
`that assists the targeted bombardment of the substrate.
`
`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`
`
`39. Licata’s system is similar to Smolanoff discussed in detail during the
`
`prosecution of the parent ’356 patent (indeed Licata is an inventor on both patents).
`
`Both systems provide a DC power to the target via an “RF FILTER” 22. Both
`
`apply an RF bias power 27 via a matching network to the substrate holder. Both
`
`include an RF generator 32 with a matching network that is inductively coupled
`
`into the chamber to generate a secondary plasma that assists in in-flight ionization
`
`of ejected particles and better directs the targeted flow of the ejected particles.
`
`Neither reference, however, explains the type of filter 22 used.
`
`
`
`
`
`- 17 -
`
`
`
`Demaray Ex. 2002-p. 20
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`
`
`40. Still other systems apply bipolar DC pulses to the target. This
`
`includes, for example, Barber (Ex. 1005), Kelly (Exs. 1048, 1059), Belkind (Ex.
`
`1008), and Sproul (Ex. 1011). Some systems applied an RF bias to the substrate,
`
`some applied a DC bias and still others allowed the substrate to develop a self-bias.
`
`None of these references, however, mention the use of a filter coupled with the
`
`bipolar pulsed DC power sources. Id. This is not surprising as it is possible to run
`
`a plasma processing chamber without any filters with the right control of circuit
`
`elements and frequency. Parameters such as pressure, temperature, gas density,
`
`gas and gas ratio all determine the gas’s electrical impedance. For example, pulse
`
`frequency affects plasma conditions because positively charged gas ions are much
`
`heavier than electrons. Thus, electrons move much faster than the positively
`
`charged gas ions. By choosing the right combination of pulse frequency (how
`
`
`
`
`
`- 18 -
`
`
`
`Demaray Ex. 2002-p. 21
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`frequently positive voltage is applied) and pulse width (which determines how long
`
`positive voltage is applied), one can suppress arcing. Thus, in many
`
`circumstances, one can prevent RF coupling into the DC power source and thus
`
`obviates the need for a filter to the bipolar pulsed DC power source.
`
`41.
`
`It should be noted, however, given the complexity of the reactor
`
`system and the unpredictable nature of the deposition process, a skilled artisan
`
`would not treat components of the reactor systems as modular pieces. For
`
`example, they would not expect that a filter designed for a reactor system applying
`
`only RF powers to target and substrate would be suitable for a reactor system using
`
`a DC power to the target and RF power to the substrate. The electrical design of
`
`the filter depends very strongly on the frequencies involved. Whereas pulsed DC
`
`voltage does change from positive to negative as does AC, RF is AC, the DC
`
`change is sharp compared to AC, thus it actually turns out that the DC pulse
`
`generates many harmonics and odd frequencies. It is important not to filter out
`
`portions of the DC signal while trying to filter out the undesired AC. The pulsed
`
`DC power is not a perfect square wave, and the filter must accommodate it.
`
`C. The Importance Of Filter Type For The Claimed Reactor System
`
`42. While none of references on prior bipolar pulsed DC systems
`
`mentioned the use of a filter with the bipolar pulsed DC power source, the
`
`inventors discovered that not only a filter is needed for a claimed reactor system
`
`
`
`
`
`- 19 -
`
`
`
`Demaray Ex. 2002-p. 22
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`with a bipolar pulsed DC power to a target and an RF bias power to a substrate, but
`
`it needs to be a specific type of filter, i.e., a claimed narrow band rejection filter
`
`that operates (or rejects) at a frequency of the RF bias power to the substrate.
`
`During the prosecution of the parent ’356 patent, the inventors repeatedly
`
`explained the unexpected discovery of the importance of the type of filter for the
`
`claimed reactor system:
`
`Ex. 1052 at 1134.
`
`43. The inventors also disputed the examiner’s view that selection of a
`
`filter type is merely a design choice and the right filter would be expected to work
`
`
`
`
`
`
`
`- 20 -
`
`
`
`Demaray Ex. 2002-p. 23
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`under the right frequency:
`
`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`
`
`
`
`Ex. 1052 at 1456-57.
`
`44. The examiner accepted the explanation and thereafter allowed claims
`
`that recited a combination of a claimed filter and a claimed reactor system with
`
`bipolar pulsed-DC power to the target and an RF bias on the substrate. Ex. 1052 at
`
`1448-53 (pending claims), 1471 (notice of allowance).
`
`45.
`
`In the ’657 patent’s prosecution, the examiner similarly allowed the
`
`claims after the inventors successfully traversed that examiner’s incorrect
`
`impression that filter choice was merely “a design choice.” Ex. 1004 at 957, 978-
`
`79, 992.
`
`
`
`
`
`- 21 -
`
`
`
`Demaray Ex. 2002-p. 24
`Applied Materials v Demaray
`IPR2021-00106
`
`

`

`Case IPR2021-00106
`Patent No. 7,381,657
`
`
`
`
`46. Applicants disagreed with the examiner’s assessment and explained
`
`the unexpected results of using the claimed type of filter:
`
`The narrow band rejection filter allows the combination of pulsed-DC
`power to the target (where the target voltage is oscillated between
`positive and negative voltages) and an RF bias on the substrate. A
`filter that blocks too many of the constituent frequencies of the pulsed
`DC waveform results in the target voltage not attaining a positive
`voltage. A filter that does not block the RF bias voltage can resu