`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`
`SAMSUNG ELECTRONICS CO. LTD.,
`Petitioner,
`
`v.
`
`PROMOS TECHNOLOGIES, INC.,
`Patent Owner.
`____________
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`____________
`
`Record of Oral Hearing
`Held: August 16, 2018
`____________
`
`
`
`
`
`
`
`
`
`
`
`Before JAMESON LEE, KEVIN F. TURNER, and JOHN A. HUDALLA,
`Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`PAUL M. ANDERSON, ESQUIRE
`NAVEEN MODI, ESQUIRE
`CHETAN BANSAL, ESQUIRE
`DANIEL ZEILBERGER, ESQUIRE
`Paul Hastings
`875 15th Street, N.W.
`Washington, D.C. 20005
`
`
`ON BEHALF OF PATENT OWNER:
`KEVIN JONES, ESQUIRE
`JERRY CHEN, ESQUIRE
`CRAIG R. KAUFMAN, ESQUIRE
`MICHAEL TING, ESQUIRE
`Tech Knowledge Law Group, LLP
`100 Marine Parkway
`Suite 200
`Redwood Shores, California 94065
`
`
`
`
`
`
`
`
`The above-entitled matter came on for hearing on Thursday, August
`16, 2018, commencing at 2:03 p.m., at the U.S. Patent and Trademark
`Office, 600 Dulany Street, Alexandria, Virginia.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` 2
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`P R O C E E D I N G S
`- - - - -
`JUDGE LEE: Welcome to the Board. This is the oral hearing for
`IPR2017-01418, petitioner, Samsung Electronics Company, LTD, and patent
`owner is ProMOS Technologies, Inc. For the record, let's begin with
`counsel introduction, first with petitioner, followed with patent owner.
`MR. BANSAL: Your Honor, this is Chetan Bansal for petitioner,
`Samsung. Along with me I have lead counsel, Naveen Modi, and Paul
`Anderson.
`JUDGE LEE: Thank you.
`MR. KAUFMAN: Good afternoon, Your Honor. Craig Kaufman
`from the Tech Knowledge Law Group. I'll be arguing this one. With me, of
`course, are Mr. Ting and Mr. Chen.
`JUDGE LEE: Thank you. Each side has 30 minutes. Let me
`know when you are ready and I'll start the timer, Mr. Bansal.
`MR. BANSAL: Your Honor, before I get started, can I approach
`the bench with physical copies of the demonstratives?
`JUDGE LEE: Please. Would you like to reserve some time for
`rebuttal?
`MR. BANSAL: Yes, Your Honor, 10 minutes, please.
`Good afternoon, Your Honors. May it please the Board, based on
`the petition and the supporting evidence, the Board instituted trial in this
`matter. The record includes even more evidence now that supports the
`
`
`
`
`
`
`
` 3
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Board's institution decision. The Board should therefore, confirm its initial
`findings and find all challenged claims unpatentable. Let me explain why.
`Can you please turn to slide number 2. On this slide, we have
`summarized the grounds of which trial was instituted. For purposes of
`today, I plan on addressing only a subset of the grounds, but of course, I'm
`happy to take any questions the Board may have.
`I'll begin by addressing the first ground and specifically claim 1.
`Can we please turn to slide 3. On this slide, I have reproduced claim 1 of the
`'044 patent. The claim includes quite a few limitations, but there's really
`only one aspect of claim 1 that is disputed between the two parties. And this
`aspect concerns the limitation of providing a mask over a dielectric layer and
`then providing a photoresist over that mask. There's no dispute that the prior
`art discloses this feature. The dispute is only whether a person of ordinary
`skill in the art would have found it obvious to combine the teachings of the
`prior art in the manner proposed by petitioner.
`Let me turn to slide 4. Slide 4 reproduces a figure from one of our
`primary references, Fujimoto. If you see this figure, etching mask 24 is
`provided over a dielectric 15. But Fujimoto is silent on what the etching
`mask 24 is. And in fact, it is undisputed that etching mask 24 can only be
`one of two things. It can either be a photoresist or it can be a hard mask
`structure.
`Petitioner explained that a person of ordinary skill in the art
`starting with Fujimoto would know that -- or would be motivated to use a
`hard mask for etching mask 24. The hard mask --
`
`
`
`
`
`
`
` 4
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`JUDGE LEE: But why do you need a hard mask? Because the
`claim doesn't recite a hard mask.
`MR. BANSAL: The claim doesn't recite a hard mask, you are
`right. But 24 -- sorry, the claim also requires providing a photoresist hard
`mask. And if you look at just Fujimoto, you do not have an explicit
`disclosure of photoresist hard mask. So what petitioner did was, we
`combined Fujimoto with Ho and said it would have been obvious to use a
`hard mask structure which includes hard mask --
`JUDGE LEE: I understand that. So Fujimoto has a mask but it
`doesn't have the photoresist. So what the petitioner did is to substitute a
`multi-tiered structure from Ho into Fujimoto to account for both the mask
`and the photoresist.
`MR. BANSAL: That's correct, Your Honor.
`JUDGE LEE: Okay. That's what I thought happened, but if you
`look at claim 16, it requires just providing a mask. It doesn't have the
`additional requirement of putting a second photoresist over that mask. So
`for claim 16 at least, I can't figure out why the petitioner is going to Ho.
`MR. BANSAL: Your Honor, for claim 16, you don't necessarily
`have to go to Ho. But what we did was in order to be concise in our
`explanation, we were relying on the same explanation that we gave for claim
`1 for claim 16. So, yes, if you look at claim 16, there is a mask as recited in
`claim 16.
`JUDGE LEE: And you accounted for 16 essentially by piggy-
`backing that on top of the arguments made for 1 because 1 actually requires
`
`
`
`
`
`
`
` 5
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`one additional element, and you didn't have to make any more presentation
`for 16?
`
`MR. BANSAL: Exactly, Your Honor.
`JUDGE LEE: Technically speaking, 16 doesn't require a
`photoresist, and if a mask doesn't have to be a hard mask, then Fujimoto has
`a mask, the etch mask 24.
`MR. BANSAL: I would agree, Your Honor.
`JUDGE LEE: But that's not the way it's presented, right? So
`essentially the arguments for 16 would stand or fall with your arguments for
`claim 1.
`MR. BANSAL: That's correct.
`JUDGE LEE: So you can't come back and say, well, I lost on
`claim 1, but 16, it doesn't require a photoresist, therefore, Fujimoto by itself
`would satisfy claim 16. You can't do that later.
`MR. BANSAL: Yes, Your Honor. We are not relying on simply
`Fujimoto's mask 24 for the mask of claim 16. We are actually relying on the
`combination which we presented for claim 1.
`JUDGE LEE: I was trying to figure out the reasoning. There's no
`technical reasoning but simply as a legal strategy, you have chosen to do
`that?
`
`MR. BANSAL: Yes, sir.
`JUDGE LEE: Thank you.
`MR. BANSAL: Can we please turn to slide number 6. So before
`Judge Lee asked me this question, what I was talking about was Fujimoto
`
`
`
`
`
`
`
` 6
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`has an etching mask 24. That etching mask 24 can only be one of two
`things, either a hard mask or a photoresist. There's ample evidence in the
`record suggesting that one of ordinary skill would have been motivated to
`use a hard mask structure and in particular a hard mask as etching mask 24.
`Here is testimony from Dr. Rubloff. Dr. Rubloff explained that a hard mask
`is a much more robust material compared to a photoresist. And that is
`undisputed.
`Let me show you some testimony from their declarant. If you can
`turn to slide number 7, please, during cross-examination, we asked their
`declarant at the time of the invention of the '044 patent, what would be one
`advantage of using a hard mask over a photoresist as an etching mask?
`Well, the first thing that comes to mind is a hard mask is a material
`that is more robust, and therefore, it would provide you with better CD
`control.
`
`Now, if you turn to slide number 8, there is additional testimony.
`Mr. Brahmbhatt said, And therefore, if you have a mask like a hard mask
`that is not getting damaged or altered during the etching process, then that
`would be able to better provide the dimensions with a better control.
`So again, there cannot really be a dispute that a hard mask is a
`much more superior etching mask when compared to photoresist. And given
`that you only have two choices really for an etching mask, one of ordinary
`skill in the art would have been motivated to use a hard mask structure for
`etching mask 24 in Fujimoto.
`
`
`
`
`
`
`
` 7
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`So what does patent owner argue? Can we turn to slide 9, please.
`Patent owner's argument is Ho's hard mask cannot be used in Fujimoto's
`process because Fujimoto's contact etch is performed using precisely the
`type of fluorine-based etchant that Ho says should not be used with its mask.
`So let me unravel what they are trying to say. What they are
`saying is they are really making two assumptions. Assumption number 1,
`Fujimoto is limited to a fluorine-based etchant. Assumption number 2, Ho
`tells a person of ordinary skill in the art do not use a hard mask if you have a
`fluorine-based etchant. If either one of those assumptions fail, their
`argument fails. And as the record shows, both of those assumptions are
`actually incorrect.
`Let me talk about the first assumption. The first assumption in the
`argument is Fujimoto is limited to a fluorine-based etchant. There is no such
`statement in Fujimoto. Fujimoto only gives a CF-based etchant which is a
`fluorine-based etchant as an example. And even the Board recognized that
`in reaching the decision. The Board said patent owner only speculates that
`Fujimoto must use fluorine-based etchants that Ho indicates should not be
`used Ho's hard mask.
`Can we please turn to slide number 13. So let's talk about the
`second assumption. The second assumption which patent owner makes is
`there is some strong statement in Ho, an explicit statement, hey, do not ever
`use fluorine with a hard mask. And that's not what Ho says. If you look at
`the language in Ho, the first box, all that Ho says is a problem may exist
`with the use of fluorine. First, the use of fluorine may have a negative
`
`
`
`
`
`
`
` 8
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`impact on critical dimension controlled during etching. May have. Not that
`it necessarily will. In fact, there is a passage of Ho that talks about using
`fluorine during etching if you want a faster rate of etching.
`So again, as I said, there were two assumptions that they made.
`Both of those assumptions are incorrect.
`Can you please turn to slide number 10. Using the two
`assumptions that they made, they made the argument that using the Ho hard
`mask in the Fujimoto process would result in lateral etching and make it
`difficult to maintain the particular dimensions. So what they are saying is if
`you use a hard mask with a fluorine-based etchant, that hard mask is not
`going to work. But that is contradicted by testimony from their own expert
`or their declarant, and it's also contradicted by the '044 patent.
`Let me explain. If you look at the second box, Mr. Brahmbhatt,
`their declarant, testified on cross-examination there might be situations
`where you will not see any notice of etching at all in the hard mask even
`with fluorine.
`If we turn to slide number 11, this is an excerpt from the '044
`patent. The '044 patent talks about using a hard mask as the etching mask.
`And guess what etchant it uses. A fluorine-based etchant. So how come is it
`that the '044 patent is able to use a fluorine-based etchant with a hard mask
`but that in the prior art that would not be a possibility?
`Finally, can I turn to slide number 14, please. The other argument
`they make is that Ho's hard mask would be severely eroded if not entirely
`consumed because Fujimoto's etch etches through oxide and nitride, and
`
`
`
`
`
`
`
` 9
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`those are the hard mask materials that Ho identifies. But again, there is a
`problem with this argument. This argument assumes that the combination is
`somehow limited to those two mask materials. But that's not true because
`Ho explicitly states that hard mask materials other than those two materials,
`silicon dioxide and silicon nitride, were well known. And the Board even
`recognized that during institution. If we turn to slide number 15, here is an
`excerpt from the institution decision. The Board said, indeed, the portion of
`Ho cited by petitioner states hard mask may be formed of SiO2, which is
`silicon dioxide, silicon nitride or other hard mask materials. We are satisfied
`that one of ordinary skill in the art would have known to select other hard
`mask materials that were compatible with Fujimoto's etchants.
`Again, we are talking about obviousness here. The claim simply
`requires a mask and a photoresist over the mask. You have Fujimoto which
`gives you the mask. Ho gives you a hard mask. A hard mask is something
`that is very well known. It's not something novel that the '044 patent came
`up with. If a person of ordinary skill in the art wanted to combine the
`teachings of Fujimoto and Ho and put a hard mask into Fujimoto, they
`would also know what gas to select. They would also know which material
`to select. And again, we have all the information in the references
`themselves for the person of ordinary skill in the art to actually make the
`combination.
`With that, Your Honor, if I can turn to slide number 17 -- any
`questions on claim 1?
`
`
`
`
`
`
`
` 10
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`JUDGE LEE: Yeah, you have explained why you would want to
`use Ho's hard mask in Fujimoto. But Fujimoto isn't missing a mask. It
`already shows an etch mask 24. So you need to bring in other stuff from Ho.
`So just bringing in the hard mask isn't enough, right, because you need to
`account for the second photoresist. That's the whole reason to go to Ho.
`MR. BANSAL: Yes, Your Honor. So when I say bring in the
`hard mask, if you see our petition, we are bringing the hard mask structure
`which includes the photoresist. Maybe I can turn back to --
`JUDGE LEE: So all the time you were referring to the hard mask,
`you were thinking of the hard mask plus the photoresist that's on the hard
`mask?
`
`MR. BANSAL: Yes. So the reason I kept referring to hard mask
`is because their argument focuses on the hard mask. Remember, the
`photoresist is only there to make the holes in the desired locations in the
`hard mask. After that the photoresist is gone and all that remains is the hard
`mask.
`
`JUDGE LEE: So as a shorthand, you were referring as a hard
`mask, you were using that word to refer to the combination of the
`photoresist and the hard mask?
`MR. BANSAL: Yes, Your Honor. I apologize for any confusion
`that may have caused.
`JUDGE LEE: In that regard, earlier you said there are only two
`types of masks. One is a photoresist and the other one is a hard mask. I
`guess in that context, the more accurate statement would be the two choices
`
`
`
`
`
`
`
` 11
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`are photoresist as one choice and the other choice is a combination hard
`mask and a photoresist?
`MR. BANSAL: I think you can look at it that way. But really the
`choice is still between a photoresist and a hard mask as the final etching
`mask. You are going to have a structure -- when you start the process, you
`are going to have a hard mask, right. You would deposit the hard mask, but
`there are no holes in the hard mask. So you have to deposit a photoresist in
`order to make the holes in the hard mask at the locations where you
`ultimately want the holes in your dielectric layers.
`JUDGE LEE: But anyway, there's no argument from the other
`side that when you use Ho's hard mask, you wouldn't bring over the
`photoresist?
`MR. BANSAL: Yes, Your Honor, there is no argument in the
`record from the other side.
`JUDGE LEE: So the parties agree that if you want to use the hard
`mask of Ho, you essentially bring over the two multi-tiered structure hard
`mask?
`
`MR. BANSAL: Yes, Your Honor. So if I may turn to slide
`number 17, the second claim limitation that I wanted to talk about is the one
`that deals with etching selectivity. Can we turn to slide number 18, please.
`The claim requires that when you are making the first opening in Fujimoto,
`that that etch have an etching selectivity between silicon nitride and silicon
`oxide of approximately 10 to 40. So what does that mean?
`
`
`
`
`
`
`
` 12
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`On slide 19, please, what this means is if we look at Fujimoto, 15
`in Fujimoto is an oxide. And prior to the etch, there was a nitride layer
`covering that blue silicide layer 13. And these two materials were etched
`away in a single etch. And what the claim is saying is that this etch will etch
`a nitride 10 to 40 times faster than it would etch oxide.
`Fujimoto says that the selectivity is between 1.2 to 2. And that's an
`example that Fujimoto gives. So Fujimoto is silent on the 10 to 40. But
`Your Honor, it is undisputed that this etching selectivity is a variable that
`one of ordinary skill in the art would optimize because it affects the profile
`of the etch. It is also undisputed that this 10 to 40 number that is in claim 7
`and 19 has no criticality. There is no unexpectedly good result that comes
`out of this range. And in view of that, we would submit that the claim is
`rendered obvious.
`JUDGE LEE: Well, I want to ask, is it enough simply to say that
`that's a result-effective variable or you have to couple it with optimization
`for you to prevail?
`MR. BANSAL: Your Honor, the result-effective variable is -- for
`a variable to be recognized as a result-effective variable, it is understood that
`a variable would only be result-effective if one would optimize that variable.
`And here in the record we have testimony from Dr. Rubloff that a person of
`ordinary skill in the art starting with Fujimoto would optimize the etching
`selectivity to get the desired profile, the desired etching profile.
`JUDGE LEE: Yeah, but he doesn't say the optimum result is 10 to
`
`40.
`
`
`
`
`
`
`
` 13
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`MR. BANSAL: Your Honor, the optimum result with being 10 to
`40, I don't think is necessarily required because then that would be saying
`that the 10 to 40 is critical. I mean, there is no evidence that 10 to 40 is
`critical. When there is a difference between the claimed range and what's
`shown in the prior art, so if the prior art has range A and the claim has range
`B, and if there's a distinction between those two ranges, those two numbers,
`that claim cannot be obvious -- sorry, cannot be patentable if the claimed
`range is not critical or doesn't give you --
`JUDGE LEE: But then you are flipping the burden. Otherwise it
`would be easy to show obviousness every time you have a difference to say
`you can do it that way. The art shows a selectivity of 1.3 to 2. Not 10 to 40.
`You can't flip it around on patent owner to say, well, you haven't shown
`there's any criticality, so I can change it to whatever I want. You got to have
`a motivation to change it to what the claim requires. So the art only has 1.3
`to 2. Not 10 to 40 as required. So what is your motivation to modify the
`selectivity from 1.3 to 2 to 10 to 40?
`MR. BANSAL: Well, for one, you would want to increase the
`selectivity because you could get an increase in the etching rate. So if you
`wanted to etch the nitride faster, you could do that. But again, Your Honor,
`I might quibble with you a little bit. I am not sure that you necessarily need
`a motivation to get to 10 to 40 if the variable has been identified as a result-
`effective variable.
`JUDGE LEE: That's what I'm saying, but usually I have read that
`in the case law as coupled with optimization. It's got to be, one, result-
`
`
`
`
`
`
`
` 14
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`effective variable, and two, you can use it to optimize. And therefore, you
`will optimize. But there's no indication that the selectivity of 10 to 40 is the
`optimal selectivity to have.
`MR. BANSAL: So Your Honor, may I pull up one of the cases
`that we cited in our papers. I know I'm in my rebuttal time, but --
`JUDGE LEE: Well, this is an important issue.
`MR. BANSAL: Your Honor, if you look at the In re Aller case,
`what it says is it is to be expected that a particular change in temperature and
`concentration or in both would be an unpatentable modification. Under
`some circumstances, however, changes such as these may impart
`patentability to a process if the particular ranges claimed produce a new and
`unexpected result which is different in kind and not merely in degree from
`the results of the prior art.
`Now, what is the point of this case law? The point of this case law
`is you do not want to incentivize an applicant for a patent to basically throw
`in some numbers in their claims and argue for patentability because I could
`draft 100 claims and throw it to an examiner and say claim 1 says 1.1 to 1.5;
`claim 2 says 1.3 to 1.4; claim 5 says 2 to 3. That's the whole point of this
`case law.
`JUDGE LEE: I understand. Is that In re Aller?
`MR. BANSAL: That is In re Aller.
`JUDGE LEE: I understand exactly what you are saying, but I'm
`just saying you put it into the pigeon hole of result-effective variable which
`normally requires optimization. I think I would be fine with it if you simply
`
`
`
`
`
`
`
` 15
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`said one of ordinary skill in the art would know to have the range at
`whatever you want. It could be 1.3 to 2 or it could be 10 to 40 because it
`simply makes no difference. And I wouldn't be citing cases that say it's
`optimization because I don't know that it's an optimal selectivity to be 10 to
`40.
`
`MR. BANSAL: Sure, Your Honor. If you look at the evidence,
`Your Honor, Dr. Rubloff testified that discovering an optimum value of a
`variable to achieve a recognized result in a known process would have been
`available within the skill of the art. I believe this is also the passage that the
`Board cited in the institution decision. And he also said, Moreover,
`consistent with Fujimoto's disclosure of etching the nitride at a faster rate
`than the oxide, the prior art in the speed envisioned setting the etching
`selectivity of such an etch in the claimed range.
`So again, there is evidence in the record that one of ordinary skill
`in the art would have optimized this selectivity to achieve --
`JUDGE LEE: For 10 to 40? Why is 10 to 40 more optimal than
`1.3 to 2?
`MR. BANSAL: Your Honor, there is no evidence that 10 to 40 is
`more optimal than 1.3 to 2.
`JUDGE LEE: That's what bothers me when you rely on case law
`that talks about optimization and there's no evidence that 10 to 40 is more
`optimized than what the art has.
`MR. BANSAL: Your Honor, I think our position is, and I can
`certainly see maybe the difference in our views, but our position is that once
`
`
`
`
`
`
`
` 16
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`a variable has been recognized as result-effective, and here it is undisputed,
`and if there is no criticality for the claimed range, then that range is obvious.
`And in fact, patent owner is not disputing any of this, right. All that they are
`saying is that if you use the 10 to 40 range in Fujimoto, certain undesirable
`things may happen which is that the nitride spacers might get etched away.
`But that is something that we have explained.
`JUDGE LEE: Alright. You'll have five minutes left.
`MR. BANSAL: Thank you, Your Honor.
`JUDGE LEE: Actually, you are over by four, so you have six left.
`MR. BANSAL: Thank you, Your Honor.
`MR. KAUFMAN: Good afternoon, Your Honors. May it please
`the Board, may I approach with hard copies of the slides?
`JUDGE LEE: Yes, please.
`MR. KAUFMAN: So why don't we take up the argument first
`where we left off with Mr. Bansal, which is this issue of optimization and
`whether the 10 to 40 etching rate is optimization of a, quote/unquote, result-
`effective variable that once they deem it result-effective, then becomes
`obvious no matter what the changes are and why they were made.
`And we would submit, Your Honor, that this is not one of the cases
`where the etch rate is such a variable. We explained in detail some of the
`problems that would occur with this particular selectivity in our response.
`But let's start for a moment with Fujimoto because Fujimoto, we
`had this talk about optimization, and Fujimoto has optimized this particular
`etch rate, the point at 1.3 to 2. The point of Fujimoto, one of its key
`
`
`
`
`
`
`
` 17
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`disclosures is to use a single etchant gas, and that avoids changing etchant
`gases which avoids the formation of carbides. And to do so they need an
`etchant gas that etches oxide and etches nitride, and to achieve that, they
`came up with this 1.3 to 2 rate.
`Samsung then says, well, it's etching and selectivity is well known,
`but that's really inconsistent with what Fujimoto is doing. If we focus on
`what the art is teaching us here, the art is teaching us that there is a value to
`keeping the etch in the range that is taught by the art and to disregard that is
`something that should not be done. So we look, for example, at Exhibit
`1005, which explains the ratio of 1.3 to 2 and is used in Fujimoto. And
`Fujimoto goes on to explain the reason for this and why it is important.
`Jeng, on the other hand, which is their secondary reference, which
`discloses an etch ratio of 5 to 20, is a different type of etch completely. It is
`trying to etch silicon nitride selectively because it is the second step of a
`contact etch where it is trying to remove silicon nitride. And so it's a
`different context than the careful balance that Fujimoto is trying to strike.
`So this combination is not suggested by the prior art. There's no suggestion
`in Fujimoto or Jeng that you should upset Fujimoto's carefully crafted
`balance of its etchants and its etching rate in order to achieve the claimed 10
`to 40 rate.
`And so we would submit, Your Honor, that consistent with the
`case law that Your Honor referred to, this is not a result-effective variable,
`this is not a minor change and that we are not obligated to show unexpected
`
`
`
`
`
`
`
` 18
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`
`results or superior results. We are just obligated to point out why the art
`doesn't teach somebody to do this which we have in our opposition.
`And even their expert on cross-examination acknowledged that the
`point of the Jeng etch was to remove the nitride. Whereas, in contrast, if you
`look at Fujimoto, the point of that etch is primarily to remove silicon oxide
`but also to remove silicon nitride as well. As our expert explained, if you
`have too selective of an etch in favor of silicon nitride, then as you etch
`through your oxide, you are going to take away silicon nitride that you don't
`want to take away.
`So with that, we will return back to providing a mask and the
`combination of Fujimoto and Ho. The question here is can you combine
`Fujimoto -- the tiered mask of Ho with Fujimoto. Our answer is respectfully
`no. The teachings of the references teach away from that combination. Ho
`teaches away from the use of fluorine. And Fujimoto only teaches the use of
`fluorine as he said. That's the only exemplified etchant and flow in
`Fujimoto. And indeed, our expert provided evidence that fluorine etchants
`were the most common etchants for etching silicon dioxide which is the
`material in Fujimoto that is being etched.
`So considering the standard from Applied Materials which is one
`of the cases cited by petitioner for teaching away, a reference may teach
`away when a person of ordinary skill in the art, upon reading the reference,
`would be discouraged from following the path set out in the reference or be
`led in a direction divergent from the path taken by the applicant. We would
`
`
`
`
`
`
`
` 19
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`submit, Your Honors, that's what's happened here and that particularly is the
`teaching of Ho.
`Samsung, in its argument, effectively raised the standard of
`teaching away to a "don't you ever, ever, ever do it, we really mean it,
`seriously" level. And that's not required by the case law and it's certainly not
`appropriate here.
`So we have here on slide 8 one example of the teaching of
`Fujimoto which refers to the use of a fluorine-based etching gas. And every
`time in Fujimoto that a specific etchant is mentioned, it is a fluorine-based
`gas. It appears approximately eight times throughout the disclosure of
`Fujimoto.
`Now, it's undisputed as well that Fujimoto does not teach the
`combined hard mask second photoresist unit that is necessary here, and that
`is why they resort to Ho. The problem is that the teachings of Ho argue
`against making that combination. Ho discloses that the hard mask can be
`formed from silicon oxide, from silicon nitride or from other materials, other
`inorganic hard mask materials. They then argue it would have been obvious
`for a person to consider using a hard mask to do so, particularly using silicon
`nitride.
`
`But what does Ho say? Ho says the problem exists with the use of
`fluorine. And Ho says that to solve this problem, to avoid the problem that
`occurs with the use of fluorine, the etching of the hard mask and the
`resulting reduction in critical dimension accuracy, the use of fluorine should
`be avoided. So they used ethane and other organic compounds that don't
`
`
`
`
`
`
`
` 20
`
`
`
`Case IPR2017-01418
`Patent 6,559,044 B1
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`contain fluorine to replace fluorine in the etching process. And that's the
`disclosure of Ho. This is one example of it at column 3, lines 52 to 56. And
`there's another example of it at the end. Since fluorine is not used in the
`invention, critical dimensions control may be improved and bowing may be
`reduced.
`And Dr. Rubloff acknowledges that this is what the reference says
`and that that's -- it says not to use fluorine in its invention because critical
`dimension control can be improved without it.
`Obviously, of course, a mask that gets etched during the etching
`process can't maintain the prescribed pattern. It becomes sloppy and less
`accurate. That can lead to electrical defects, and so that is obviously bad.
`So we would submit, Your Honors, that Fujimoto and Ho are not
`appropriately combined. And because they are not appropriately combined,
`all ground