`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`
`SPTS TECHNOLOGIES LTD.,
`Petitioner,
`
`v.
`
`PLASMA-THERM LLC,
`Patent Owner.
`____________
`
`Case IPR2017-01674
`Patent 8,802,545 B2
`____________
`
`Record of Oral Hearing
`Held: October 10, 2018
`____________
`
`
`
`
`
`
`
`
`
`
`Before WILLIAM V. SAINDON, ELIZABETH M. ROESEL, and AMANDA F.
`WIEKER, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`
`APPEARANCES:
`
`
`ON BEHALF OF THE PETITIONER, SPTS TECHNOLOGIES LTD.:
`ADAM C. VOLENTINE, ESQUIRE
`VOLENTINE, WHITT & FRANCOS, PLLC
`One Freedom Square
`11951 Freedom Drive Suite 1300
`Reston, Virginia 20190
`avolentine@volentine.com
`
`
`ON BEHALF OF THE PATENT OWNER, PLASMA-THERM LLC:
`RYAN M. CORBETT, ESQUIRE
`HARVEY S. KAUGET, ESQUIRE
`BURR & FORMAN LLP
`One Tampa City Center, Suite 3200
`201 North Franklin Street
`Tampa, Florida 33602
`rcorbett@burr.com
`hkauget@burr.com
`
`
`
`The above-entitled matter came on for hearing on Wednesday, October 10, 2018,
`
`commencing at 10:00 AM ET, at the U.S. Patent and Trademark Office, 600 Dulany Street,
`Alexandria, Virginia.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` 2
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE WIEKER: Good morning, everyone. Please be seated.
`This is an oral hearing for case number IPR2017-01674 between Petitioner,
`SPTS Technologies, and Patent Owner, Plasma-Therm, LLC. The
`proceeding concerns U.S. Patent Number 8,802,545. I'm Judge Wieker. I'm
`accompanied today by Judge Saindon and Judge Roesel. I would like to ask
`Petitioner's counsel to please introduce yourself for the record.
`MR. VOLENTINE: Adam Volentine of Volentine, Whitt and
`Francos representing the Petitioner, SPTS Technologies, Limited.
`JUDGE WIEKER: Thank you, Mr. Volentine. And for Patent
`Owner?
`MR. CORBETT: Ryan Corbett from the Burr & Forman law firm
`representing Patent Owner, Plasma-Therm, LLC. And with me -- he can
`introduce himself.
`MR. KAUGET: Harvey Kauget of Burr & Forman on behalf of
`Patent Owner.
`JUDGE WIEKER: Good morning. Thank you for being here
`today. Each party will have a total of 45 minutes to present their argument.
`Petitioner will proceed first. Patent Owner will then respond to Petitioner's
`case. Using any reserved rebuttal time, Petitioner may then respond to
`Patent Owner's case. Finally, using any reserved sur-rebuttal time, Patent
`Owner may respond to Petitioner's rebuttal argument. I would like to
`
`
`
`
`
`
` 3
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`remind the parties to please identify any demonstrative exhibit into the
`record by either slide number or screen number.
`With that, Mr. Volentine, would you like to reserve any rebuttal
`time from your 45 minutes?
`MR. VOLENTINE: I expect 15 minutes of rebuttal.
`JUDGE WIEKER: Okay. And you may begin when you are
`
`ready.
`
`MR. VOLENTINE: I'm at slide 2. I would first like to point out
`that we think our Petition and Dr. Spencer’s, our expert's, declaration and
`our Reply are sufficient to establish the invalidity of the claims. What I
`want to do here is address the major issues that have come up over the past
`eight months or so and in depositions. So I won't be rehashing claim charts
`or anything of that nature.
`So let's go to slide 2, then. As you know, Petitioner relies on two
`references. One is Fischer, which unfortunately is misspelled throughout
`these demonstratives. And it's issued to Robert Bosch GmbH, which is a
`German company credited with inventing the Bosch process. The second
`reference is Donohue. That was issued to Plasma-Therm, Inc., which we
`believe to be a predecessor of Patent Owner in this case, Plasma-Therm,
`Limited.
`Fischer -- just briefly, Fischer is directed to plasma dicing of a
`wafer, meaning separating the wafer into chips using plasma. There are
`many, many references out there that teach plasma dicing. This is one of
`them.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 4
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`Donohue is directed to plasma etching the silicon layer of a silicon
`on insulating layer substrate, SOI substrate. The intention of Donohue is as
`you etch through the silicon, you want to avoid notching at the interface
`between the underlying oxide layer and the silicon. That's the -- what
`Donohue is about, for the most part. It is admittedly not a dicing case.
`In paragraph 75 of his original declaration, our expert, Dr. Spencer,
`opined that the '545 patent in its description is describing the Patent Owner's
`previous patent to Donohue. And after going through all this discovery and
`so forth, we stand by that. We think that's pretty clear. That is described in
`detail at pages -- beginning at page 43 of our Petition. So those are the
`references at play here.
`Moving on to slide 4, one of the primary issues that came up was
`whether the POSITA would even want to avoid undercut when dicing a
`wafer. This goes, I guess, to motivation to combine. Not surprisingly, the
`Patent Owner says, no, that wouldn't be the case. We disagree. We say the
`record shows that the POSITA would indeed find undercut during dicing to
`be undesirable at the least.
`So what evidence do we have, the evidence of the desire to avoid
`undercut? We have our expert's testimony. I'm going to, if you'll indulge
`me, I'll read most of that. He says, “Initially, I agree with both the '545
`patent,” which is the patent under review, “and Donohue that undercutting of
`silicon at the etch stop interface is ‘undesirable.’” So we have Dr. Spencer,
`an expert in the art, agreeing with the '545 observation on that point and
`agreeing with Donohue on that point. Spencer goes forward to say this is
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 5
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`true whether one is plasma etching silicon to form trenches in an SOI
`structure as in Donohue or one is plasma etching silicon to dice a wafer into
`individual chips. So that expert testimony is on the record.
`I'm moving now to slide 6. So what do I want to point out here is
`that Dr. Spencer, his testimony, is not shown to be incredible anywhere
`during these proceedings. It hasn't even been alleged as far as I know. And
`he's corroborated the statements about the undesirability of undercut. As I
`mentioned before, one of the corroborations is the '545 patent itself. It
`reads -- it talks about undercutting, and this is in a dicing environment. It is
`the '545 patent. I'll skip ahead a little bit, but such problems, which means
`undercut, include local severe undercutting at the substrate/insulator
`interface. Now, the insulator in the dicing world is the dicing tape which
`holds the wafer in place during this etch process. There's a ring around that
`as well, but that's not really relevant to this particular proceeding.
`What is relevant is you have a wafer on dicing tape, and the dicing
`tape is insulative. So the '545 patent is correctly making the observation that
`you can suffer undercutting by the accumulation of charges at the bottom of
`that recess. This is all explained in the patent itself and in Donohue.
`The '545 patent goes on to further corroborate Dr. Spencer by
`saying this undercutting is “undesirable during die separation, since this
`affects the performance of the singulated die.” Again, this is from the patent
`itself.
`
`We have been charged with improperly using admitted prior art in
`this argument. And my response to that is we are not using these admissions
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 6
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`as rejections. We are just using them -- Dr. Spencer is using them to
`corroborate his own testimony. We have an expert in the field who agrees
`with what is being said in the '545 patent.
`There's further corroboration, and that is by one of the co-inventors
`of the '545 patent. In an earlier patent that's patent number 7,781,310, which
`is Exhibit 1021, that is directed to plasma dicing, that patent, and it has
`technical distinctions for sure relative to what Fischer is doing. But that's
`not the point. The point is Grivna, that patent, presents a technique to avoid
`undercut. It's expressly trying to avoid undercut. To us the implication is
`undercut is not desirable. And that is just to further corroborate
`Dr. Spencer's testimony of the lack of desirability of undercutting in a
`plasma dicing environment.
`JUDGE WIEKER: But isn't Grivna's undercutting of a different
`type than that at issue here?
`MR. VOLENTINE: A different type?
`JUDGE WIEKER: For example, Grivna is directed to isotropic
`etching, correct?
`MR. VOLENTINE: Yeah, the processes are different. It's the fact
`that the final product, once the chips are separated and moved to the next
`packaging process, will those chips have undercutting or will they not have
`undercutting. Grivna says we would rather they did not have undercutting.
`It goes to the desirability to avoid undercutting. I agree the processes of
`Grivna are distinct for sure. But the lesson of Grivna, if you will, is that we
`don't want undercutting.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 7
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`JUDGE WIEKER: But, for example, my understanding is that the
`undercutting experienced by Donohue, for example, is due to the build up of
`charge at the bottom of the trench. That's not what's being discussed in
`Grivna, correct?
`MR. VOLENTINE: I don't know for sure. Grivna is very brief to
`go from -- I forget. I don't remember the details, but it was very brief, one
`paragraph. But again, we are not concerned with -- for example, Grivna
`could be sawing. If Grivna says when we are sawing we want to avoid
`undercut, that means there is a teaching of the desire to avoid undercut.
`That's the only reason we rely on Grivna. Not for its processes. We have a
`co-inventor of the '545 patent. He has a publication out there that is dicing --
`plasma dicing, and he presents a technique for avoiding undercut. We are
`not suggesting that we take that technique and use it. We are suggesting we
`take the desire to avoid undercutting and apply it in Fischer. If Grivna wants
`to avoid undercutting, then Fischer wants to avoid undercutting.
`JUDGE SAINDON: Counsel, I have a question for you. So let's
`take as a given that a person of ordinary skill in the art doesn't like
`undercutting. I'm wondering if there's a separate issue given that, for this
`question, would a person of ordinary skill in the art really think undercutting
`was a problem in Fischer to begin with, such as to look to avoid
`undercutting? Maybe you don't want undercutting but you have to know it's
`a problem before you would go and look to solve that.
`MR. VOLENTINE: Well, Donohue, among other references,
`teaches that when you etch through silicon and encounter an insulating
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 8
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`underlayer, you will suffer undercutting unless you take remedial action. It's
`just going to happen. And the remedial action of one reference might be
`different from another. Grivna's is different from Donohue's. Fischer makes
`no mention of undercutting, but it's not about undercutting. What Donohue
`teaches you is that if you run those etch processes of Fischer, you will
`encounter undercutting. You meet all of the conditions of undercutting.
`JUDGE SAINDON: I guess my question is in Donohue's slightly
`different context, we are making the functional components of the chip
`versus just slicing through it. So maybe there's a reason why Donohue is
`concerned with undercutting that maybe isn't applicable to Fischer where
`you just cut through the chip.
`MR. VOLENTINE: So those reasons are not apparent to us. The
`tape is an insulator, just like the insulator of an SOI structure. The charges
`will be trapped in the same manner in either case. We also have the patent
`itself saying that undercutting is a problem if you don't do something about
`it. And then that's in a plasma dicing environment.
`I'm going to get to the Patent Owner makes a big deal over
`thicknesses of the silicon layer trying to distinguish, saying Donohue is this
`thick and Fischer is this thick and that silicon layer and there are a lot of
`calculations and analysis done based on that to prove that they are not the
`same, that you wouldn't -- the worlds are too far apart.
`JUDGE ROESEL: So counsel, aside from the challenged patent,
`what evidence is there in this record that undercutting is a problem when it
`comes to plasma dicing?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 9
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`MR. VOLENTINE: We have the testimony of our expert that is
`corroborated by not only the patent but by Grivna. Grivna wants to avoid
`plasma -- I mean, undercut.
`JUDGE ROESEL: Grivna is not a plasma dicing process, right?
`MR. VOLENTINE: It is a plasma dicing process.
`JUDGE ROESEL: Yeah, but isotropic method, right?
`MR. VOLENTINE: Yeah. And that's why it's so important. It's
`plasma dicing, and there's a suggestion that undercutting is not something
`we want. And then there's also, I don't want to get -- this is going to come
`across as nit-picky maybe, but the Board noticed this in their Decision to
`Institute. I'm going to go ahead, if it's okay, to -- I'm reluctant to bring it up
`because it's lawyerly, but nevertheless, in their original preliminary reply the
`Patent Ownersaid the undercutting problem is much more problematic in
`Donohue than it is in Fischer. And then the Board reasonably interpreted
`that to mean, well, it's at least somewhat problematic in Fischer. It may not
`be as problematic, but it's at least somewhat problematic. And that came
`across as an admission that at least there are some problems associated with
`undercutting in plasma dicing.
`JUDGE WIEKER: The Board's finding at the institution stage are
`merely preliminary. So turning back to the evidence supporting Petitioner's
`argument, I believe it's paragraph 90 of Dr. Spencer's declaration, he states
`that -- let me read it, that undercutting would occur in Fischer at the carrier
`film for the same reasons that Donohue describes undercut to occur at the
`SOI layer. What evidence -- is there any evidence to support Dr. Spencer's
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 10
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`opinion, knowing that Fischer says that the etching process stops when you
`reach the carrier film?
`MR. VOLENTINE: We are going to get to that issue. I can move
`straight to it now if you like.
`JUDGE WIEKER: That would be great.
`MR. VOLENTINE: I'm going to skip ahead. So this argument,
`I'm sorry, is entirely misplaced. Let me find it in my notes. Okay. Fischer,
`as I understand Patent Owner, Patent Owner says that while you are etching,
`you got all these streets and so forth, that the etch rates are not all the same
`and you will at one point reach the tape before you have the other points.
`It's just the nature of the beast. The process is not perfect. So the suggestion
`by Patent Owner is that the instant that that tape is exposed, you stop etching
`in Fischer. Okay. That is nowhere in Fischer.
`JUDGE WIEKER: Before we go too much further, just so we are
`all on the same page, I'm looking at Fischer column 6, line 58. It says “the
`etching operation is first stopped on the boundary surface to carrier film 28.
`Thus, substrate wafer 6 is completely sectioned into substrate chips 20.” So
`is that -- maybe I'm not understanding your argument, then.
`MR. VOLENTINE: So the argument is Dr. Shanfield, in his
`original testimony, his original declaration, he quotes almost verbatim that
`portion of Fischer. And that portion reads -- it's slightly off because he
`changed a couple of insignificant words. But he stated, According to
`Fischer, after the etching process is completed, several lines 24 in the
`regions not covered by the etching mass -- separating lines from in the
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 11
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`regions not covered by the etching mass, as shown in Figure 6 below, in
`these lines, the silicon of the substrate wafer is completely removed and the
`etching operation is stopped on the boundary surface up to the carrier film.
`Complete removal, then stopping. There's no suggestion there of stopping
`the instant that tape is exposed. Then the result is the substrate wafer is
`completely sectioned into substrate chips.
`That was the original testimony of Dr. Shanfield. And that seems
`clear to me all the chips have been separated as a result of the plasma
`process.
`JUDGE WIEKER: So I think I know where you are going, and
`before you move on to there, let me just return to my first question, which
`was whether there is any discussion by Dr. Spencer, for example, of why
`undercut would occur in that circumstance you just described. So where you
`have completely sectioned the wafer into individual chips and you stop at
`boundary layer 28, is there any discussion of whether there's sufficient
`charge built up to result in undercutting?
`MR. VOLENTINE: I don't recall specifically other than that he
`said there would be undercutting for the same reasons that Donohue says
`there would be undercutting.
`JUDGE WIEKER: But in Donohue, the etching doesn't stop as
`soon as the interface is exposed, does it?
`MR. VOLENTINE: No, but that's my point, is it doesn't in Fischer
`either. So you have these streets, I don't know how many of them there are,
`and you are etching. Somewhere in this grid the tape will be exposed, and it
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 12
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`will be exposed before everywhere else in the grid because of the uneven
`etch rates that are unavoidable. So you don't stop. You just keep etching
`until -- and Fischer teaches the entire grid gets etched out.
`JUDGE WIEKER: And where in Fischer does it specify that,
`other than that portion of column 6 that says the etching is first stopped and
`then they are completely sectioned? To me that's not clear that the entire
`grid, as you describe it, is necessarily being --
`MR. VOLENTINE: I don't know. To me it's clear. The drawings,
`for one, don't show any residual -- there's no mention in Fischer of residual
`silicon in these trenches anywhere. The drawings show every trench being
`cleared. And then Fischer itself says in these lines, these are the dicing lines,
`the silicon of the substrate wafer is completely removed.
`JUDGE WIEKER: Where are you looking?
`MR. VOLENTINE: It's --
`JUDGE SAINDON: Looks like column 6, line 57 to 58.
`MR. VOLENTINE: It's in that area. So yeah, line 57.
`JUDGE SAINDON: So I guess what is the significance of that
`word "first" in the following sentence? What are we to make of that word,
`"first stopped"?
`MR. VOLENTINE: I cannot explain that word. There's no
`“second” that I can find in there. So in order -- to me, in my opinion, the
`evidence is overwhelming that Fischer is not describing an embodiment
`where there is residual silicon throughout the grid of dicing lines. It's not
`mentioned. It's not shown in the drawings. None of that.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 13
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`We have Dr. Shanfield, who testified to the same thing in his first
`declaration. He makes no mention that there might be residual silicon in
`these dicing lines. Later in the supplemental declaration, this comes up for
`the first time. And the argument for it relies partially on a reference to
`Sekiya, and it is entirely misplaced. I'm sorry, Sekiya, let me see if I can get
`the right -- so Sekiya is Exhibit 2014, I believe, and it's Patent Number
`6,465,158. Bear with me for a second. I got out of order.
`Okay. So in his supplemental declaration, Shanfield, in our view,
`departing from his original declaration, looks to Sekiya. He states that --
`again, this is Exhibit 2010, paragraph 21, of Shanfield's supplemental
`declaration. For example, in Sekiya, Sekiya states that a slight thickness of
`unetched region may remain. Sekiya, by the way, is directed to plasma
`dicing, in which case, a slight bending force is later exerted on the
`semiconductor wafer along the streets to break the remaining unetched
`regions. Therefore, it is my opinion that a POSITA would understand
`Fischer as disclosing that the etching process is stopped upon reaching the
`carrier film.
`Now, it's apples and oranges. In this field, you can stop, whether
`it's plasma dicing or sawing or whatever, you can stop before you
`completely travel through these dicing lines. But the idea is you stop
`everywhere, then you break it. It's not like you expose the tape and then
`stop.
`
`What Sekiya is teaching is you stop at a height above the tape and
`then you break that up. And even if Sekiya didn't teach that, Sekiya was
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 14
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`saying, well, you go the instant -- it doesn't say this, but even if it said the
`instant the tape is exposed, we stop and then we break it apart, that's a
`teaching of doing that. Nothing in Fischer suggests that kind of operation.
`The other thing Sekiya describes is that this idea of stopping short
`and breaking is the alternative to the main embodiment. The main
`embodiment of Sekiya is etching all the streets all the way through.
`JUDGE ROESEL: So it's petitioner's position that a person of
`ordinary skill in the art reading Fischer would recognize that undercutting is
`occurring; is that right? And it's a problem that needs to be solved in
`Fischer?
`MR. VOLENTINE: That's correct.
`JUDGE ROESEL: Can you direct us exactly what step, etching
`step in Fischer would be causing this undercutting and what is the cause of
`the undercutting?
`MR. VOLENTINE: So in Fischer there's a single etch carried out
`through the entire plasma dicing process. It's a -- they have the example of
`using a microwave, microwave plasma. That's a high density plasma in tech
`speak, in that world. High density plasma is what causes undercut,
`according to Donohue. It's these charges that accumulate. And they will
`accumulate at the interface because the interface is an insulator. It's tape.
`JUDGE ROESEL: So you are saying the interface between the
`carrier film and the silicon wafer?
`MR. VOLENTINE: And the wafer, that's correct. Donohue
`explains very nicely how undercut occurs and why.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 15
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`JUDGE ROESEL: Is there anything else in Fischer that you can
`point to to say that the person of ordinary skill in the art would have
`recognized that undercutting is a problem?
`MR. VOLENTINE: In Fischer itself, no. No, Fischer was chosen
`among many, many plasma dicing patents for the reason that it has a feature
`relevant to one of the other claims. So we didn't have to have multiple
`patents. But it's just basic plasma dicing. The fact that there would be
`undercutting that would be desired to be avoided is from the teachings of
`Donohue and the fact that Grivna says, hey, we don't want undercutting
`anyway.
`JUDGE ROESEL: So petitioner looked at many plasma dicing
`patents before selecting Fischer?
`MR. VOLENTINE: Yeah, we selected Fischer because one of the
`claims of this '545 patent states that an element -- I don't remember the exact
`wording, but essentially that an element of the circuit functions as a mask.
`I'm sure they'll correct me if I'm wrong on that, but that's the basic idea, that
`an element -- an actual circuit element operates as a mask during the plasma
`dicing process.
`JUDGE ROESEL: So in those many other plasma dicing patents
`that petitioner looked at, was there any recognition of undercutting as a
`problem?
`MR. VOLENTINE: Just Grivna as far as I know. I mean, we
`found what we thought we needed and we stopped.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
`
`
` 16
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`JUDGE WIEKER: Judge Roesel, do you have any other
`questions? We are at the end of your 30 minutes, if you would like to --
`MR. VOLENTINE: Yeah, I think that's a good time.
`JUDGE WIEKER: We'll save your other 15 minutes for rebuttal.
`MR. VOLENTINE: Thank you.
`JUDGE WIEKER: Mr. Corbett, whenever you are ready, would
`you like to reserve time for sur-rebuttal?
`MR. CORBETT: Yes, Your Honor, I would like to reserve 10
`minutes for sur-rebuttal.
`May it please the Board, again, my name is Ryan Corbett from the
`Burr Forman law firm. I represent Plasma-Therm, LLC, who is the Patent
`Owner in this inter partes review. Now, my colleague, as he stated,
`Petitioner takes a position that the challenged claims are obvious in view of
`Fischer and Donohue. And it's Patent Owner's position that Petitioner has
`failed to meet its burden for two primary reasons. First, there's no
`motivation to combine Fischer and Donohue. And there's two sub-reasons
`for that. First, as the Board had already asked about with respect to
`undercutting, there's actually no evidence that Fischer suffers from undercut.
`Petitioner makes a big deal about that undercut is known to be undesirable in
`dicing, and patent owner agrees with that. The '545 patent actually explains
`that that's undesirable. But that's not sufficient to meet petitioner's burden.
`Petitioner needs to show that Fischer actually suffers from undercut such
`that there would be a motivation to combine Donohue with Fischer to avoid
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
`
`
` 17
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`the undercut. And without that problem, there's no motivation to combine
`Donohue to solve a problem that doesn't exist.
`The second reason there's no motivation to combine is that Fischer
`states that its primary objective is to create a time-saving dicing process, and
`they mention the comparison to sawing, which is a much slower process.
`Now, Donohue relates to a much different process in which very thin film
`layers are being etched when you are actually forming the device, which is
`in contrast to Fischer, where you are dicing the wafer after the devices are
`already formed.
`And the reason this is important is that Donohue uses an RIE
`process which stands for reactive ion etching. And that is a much slower
`process than the high etch rate process that Fischer uses. So using a slow
`process of Donohue would frustrate the time-saving purpose of Fischer. So
`that's a separate reason why there's no motivation to combine Fischer and
`Donohue.
`The second primary reason is that the cited references do not teach
`or suggest all of the recited claim elements. In particular, the second time
`division multiplex process or TDM process. As that claim term is properly
`construed, the second TDM process requires a high density plasma source
`which stands in contrast to the low density RIE process of Donohue. So
`because the two references don't teach two high density TDM processes,
`petitioner has failed to meet its burden for that separate reason.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`
`
`
`
`
`
`
` 18
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`
`JUDGE ROESEL: Is it Patent Owner's position that undercutting
`was not recognized for plasma dicing generally or just not recognized for
`Fischer?
`MR. CORBETT: It's Patent Owner's position that undercut was a
`known problem in dicing but that it just simply does not occur in Fischer.
`JUDGE ROESEL: Why is that?
`MR. CORBETT: Why is it that it doesn't occur in Fischer?
`JUDGE ROESEL: Yes.
`MR. CORBETT: Well, in Fischer -- and I can skip to a slide that
`shows this. Well, in short -- and I'll supplement with the slide here in a
`moment. In short, Fischer teaches anisotropic etching which means that the
`positive ions from the plasma impinge in a vertical direction down onto the
`substrate. So there's no -- there's very minimal undercut that occurs before
`you get to the tape because the ions are going in a vertical direction.
`And Fischer even more so goes on to say that there's a process
`where they alternate between a deposition step and an etching step. In this
`deposition step, they deposit a polymer on the side walls of the trench, and
`that further prevents undercutting. So that's a different kind of undercut.
`That's actually the kind of undercut that Grivna discusses.
`Now, the kind of undercut that petitioner refers to is the undercut
`that occurs when the charge accumulates on the tape layer. Now, this also
`doesn't occur in Fischer, because as was discussed previously, Fischer
`explains that the etching process first stops when the tape is reached. And so
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`
`
`
`
`
`
` 19
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`if the etch process stops when you reach the tape, there's no accumulation of
`the charge, and therefore, there's no undercut.
`JUDGE WIEKER: How do you respond to Petitioner's argument
`that even though Fischer says it stops, it doesn't actually stop until all the
`streets are etched away?
`MR. CORBETT: Right. Well, Fischer does say that it first stops
`and it's completely etched. Now, Petitioner raises for the very first time this
`argument about uneven etch rates and interpreting how -- what that means in
`Fischer. And Patent Owner would submit that's an improper new argument
`that has not been raised before in either the petition or the reply.
`But if the Board is inclined to consider it, the way to reconcile that
`is that the two teachings of Fischer is that the etch rate is uniform. So when
`you hit the tape, all of the silicon is gone from the street. So still there's no
`undercut. You can still stop when you hit the tape and have no silicon left in
`the street and also no undercut because there's no charge accumulation.
`JUDGE WIEKER: Isn't there some evidence that the wafers
`necessarily have varying thicknesses to a degree such that, for example, you
`could stop when tape is exposed in area A of the wafer but it not be exposed
`yet in area B?
`MR. CORBETT: Yes, the '545 patent does explain that there are
`circumstances where that is the case. Dr. Shanfield, in his initial declaration,
`also explained that in some instances that is the case, that there is a variance
`in the thickness or the etch rate, and so it's possible that you can have some
`tape exposed at one part of the street where there's still some silicon left in
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
`
`
` 20
`
`
`
`IPR2017-01674
`Patent 8,802,545 B2
`
`other parts of the street. And that's actually the problem that the '545 patent
`is trying to solve. That's why it switches to a lower etch rate second TDM
`process.
`But Fischer is silent with respect to that. So there's no evidence
`that Fischer -- that's actually the case in Fischer. So it's Patent Owner's
`position that what Fischer is describing is a more uniform process where you
`hit the tape and stop, and yet it's still completely removed from the streets.
`The key point being that there's no undercut because it stops when you hit
`the tape.
`JUDGE WIEKER: How does Dr. Shanfield's testimony about the
`teaching in Sekiya, for ex