`Trials@uspto.gov
`571-272-7822
`
`
`
`
`
`
`
`Paper No. 66
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`NICHIA CORPORATION
`Petitioner,
`
`v.
`
`EMCORE CORPORATION
`Patent Owner.
`____________
`
`Case No. IPR2012-00005 (JYC)
`Patent 6,653,215
`____________
`
`Held: November 6, 2013
`____________
`
`Before KEVIN F. TURNER, STEPHEN C. SIU, and JONI Y. CHANG,
`Administrative Patent Judges.
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`MATTHEW SMITH, ESQUIRE
`
`
`Turner Boyd, LLP
`
`
`2570 West El Camino Real, Suite 380
`Mountain View, California 94040
`
`
`
`
`
`
`
`
`
`
`and
`
`
`
`
`
`
`
`
`
`
`STEPHEN B. MAEBIUS, ESQUIRE
`CHASE J. BRILL, ESQUIRE
`Foley & Lardner LLP
`Washington Harbour
`3000 K Street, NW, Suite 600
`Washington, DC 20007-5109
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`
`
`
`
`
`
`
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`MICHAEL TOMASULO, ESQUIRE
`
`
`DAVID K. LIN, ESQUIRE
`
`
`Winston & Strawn LLP
`
`
`333 South Grand Avenue
`
`
`Los Angeles, California 90071
`
`
`
`The above-entitled matter came on for hearing on Wednesday,
`November 6, 2013, commencing at 1:00 p.m., at the U.S. Patent and
`Trademark Office, 600 Dulany Street, Alexandria, Virginia.
`
`
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`27
`28
`29
`30
`31
`
`
`
`
`
`
`
`P R O C E E D I N G S
`- - - - -
`JUDGE CHANG: Good afternoon. Welcome to the PTO. This is an
`oral hearing for case IBR2012-00005. The Board instituted this Inter Partes
`review for patent 6,653,215 on February 12th, 2013.
`At this time, we would like counsel to introduce yourself,
`beginning with petitioner.
`MR. SMITH: Thank you, Your Honor, this is Matthew Smith
`from Turner Boyd, LLP for the petitioner, Nichia Corporation. I have with
`
`
`
`
` 2
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`me Chase Brill at counsel table, and Mr. Maebius, from Foley & Lardner
`and also a representative from Nichia, Mr. Miki.
`JUDGE CHANG: Thank you.
`MR. TOMASULO: Good afternoon, Your Honor, my name is
`Michael Tomasulo for patent owner. With me is David Lin who is an
`associate of Winston & Strawn. I have Mr. Richard Merisier with Trial
`Graphics who is going to help us with our onscreen visual presentation. And
`I have a representative from Emcore, Mr. Daniel McGlynn, and a
`representative from Emcore's licensee, Everlight, Ms. Cindy Chou. Both
`Mr. McGlynn and Ms. Chow are attorneys.
`JUDGE: Welcome. Thank you.
`Each party has one hour to present their arguments. This is
`consistent with our order granting the parties' request for an oral hearing.
`Petitioner will proceed first, presenting its case as to the challenged-to
`claims. Since in this case the patent owner filed a motion to amend, so the
`petitioner may reserve rebuttal time for its case and time for responding to
`the patent owner's case with regard to the motion to amend claims.
`Thereafter, patent owner will respond to the petitioner's case,
`and also present its case with regard to the motion to amend. Patent owner
`may reserve rebuttal time for its case as to the motion to amend claims. And
`after that, the petitioner will use up the rest of the time to respond to the
`patent owner's presentation on all matters.
`And I just want to make sure that Judge Turner, are you okay
`over there? Judge Turner?
`JUDGE TURNER: I can hear you, yes, thank you, I'm here.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 3
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`JUDGE CHANG: Great. So, we can start.
`MR. TOMASULO: May I ask a procedural clarification before
`
`we start?
`
`JUDGE CHANG: Sure.
`MR. TOMASULO: With respect to the order that one of the
`recent orders instituted regarding the scope of oral argument, it was made
`pretty clear that the parties are restricted to arguments that they made. And,
`so, if Mr. Smith, for instance, has an objection that I've gone beyond proper
`argument, would it be customary for him to wait until his rebuttal time to
`reserve that? I would appreciate that I wouldn't interject my objections
`during his presentation and reserve my objections until once he's concluded,
`and similarly he would do the same for me, but I want to make sure that to
`the extent either one of us do have objections, we understand the proper
`procedure for preserving them for the record.
`JUDGE CHANG: Okay, that's a good question. Usually, so
`far, we have several AIA hearings, it's very similar to our other oral
`hearings, most of the time we do have a lot of questions, so in answering the
`questions, you might go outside the scope of the papers or argument that you
`filed, but we do ask that the counsel stay with the argument that was
`presented in your papers. But I realize sometimes it might go outside of it,
`and we try to give some leeway, but if your whole presentation is outside the
`scope of your paper, I mean, I think that to reserve the time and also
`efficiency of this proceeding, we would like to probably, you know, stop
`counsel from that line of argument.
`
`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
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`So, I'm just going to try to be reasonable. I can't say at this
`point that, you know, I hope that not everybody's going to jump up and say
`objection to every single sentence, but we're going to give you leeway to go
`outside the scope of your paper, but like I said, if your main argument for ten
`to 15 minutes to be all outside the scope of your brief, then we will probably
`stop you. Is that fair?
`MR. TOMASULO: It's certainly fair, but I'm still not entirely
`clear. I would prefer to not interrupt his presentation with an objection.
`JUDGE CHANG: Okay.
`MR. TOMASULO: May I just reserve my objections and make
`them at the time of my presentation?
`JUDGE CHANG: You can.
`MR. TOMASULO: I would just as soon not have Mr. Smith to
`interrupt me as well.
`MR. SMITH: I'm happy to agree to that, Your Honor.
`JUDGE CHANG: How's this? We're going to conduct it in a
`nice manner. So, I will ask Mr. Smith not to interrupt you, but I do want to
`make sure you understand that we may be able to interrupt your presentation
`with questions as well as may ask you where is that argument in your brief,
`so that way we can understand whether the argument was originally
`presented in your paper.
`MR. TOMASULO: Understood, I'm certain that we would
`both welcome your questions.
`JUDGE CHANG: Okay, that sounds great.
`
`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
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`MR. SMITH: Thank you very much for the clarification, Your
`Honor. Judge Turner, are you getting video?
`JUDGE TURNER: I can hear you, yes.
`MR. SMITH: Wonderful. Are you able to see the video
`
`screen?
`
`JUDGE TURNER: Please make sure you speak into the mic'.
`MR. SMITH: May it please the Board, Mr. Brill, if you could
`bring up the decision instituting trial, page 3, I would like to start with just a
`brief explanation of claim 1 of the '215 patent. I know we're all relatively
`familiar with it, but I will get through it relatively quickly.
`Claim 1 of the '215 patent, of course, is shown in the diagram in
`the decision instituting trial as a method claim directing the person carrying
`out the method to deposit four contact layers onto a III-V semiconductor
`base. The four contact layers are, in order,
`aluminum/titanium/platinum/gold, of course. The titanium and platinum are
`said to be barrier layers and, of course, the final step is the annealing step
`that we have discussed in the papers.
`I wonder, Mr. Brill, if you could bring up the petition for the
`Inter Partes review, page 35.
`Let's talk about the ground 4 references, and in particular, the
`Kidoguchi reference, and I will discuss the ground 4 references, most of the
`prima facie case, if I could reserve maybe a half an hour, some for rebuttal
`and some for responding to the patent owner's motion to amend, that would
`be good.
`
`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
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`The Kidoguchi reference, of course, is very, very close to claim
`1 of the '215 patent. Kidoguchi shows an n-type gallium nitride layer that's a
`III-V semiconductor. Figure 14 has the inside contact, it has a molybdenum
`base layer, a titanium barrier layer, a platinum barrier layer and a gold top
`layer. So, we are already very close to claim 1 of the '215 patent.
`The molybdenum layer is not the same as the aluminum layer,
`but, of course, the Kidoguchi reference expressly says that one can use an
`aluminum layer for the molybdenum base layer in the portion of the
`reference that has been highlighted by Mr. Brill on the screen.
`I want to stop there for the moment, because Emcore has made
`a number of arguments about whether or not it would have actually been
`obvious to use that aluminum-based layer in molybdenum. And honestly, I
`think we're already past that, because the Kidoguchi reference expressly says
`that you can use aluminum as a base layer in that context stack. So, that is a
`disclosed species. It's not a question under section 103 whether you can
`substitute aluminum in for molybdenum, and this is also not the sort of case
`where there's a very large genus disclosed and the species in the genus are
`not discernible to a person of ordinary skill in the art. There is a very small
`number of species disclosed here, only four additional metals next to the
`molybdenum layer.
`So, Kidoguchi discloses everything in claim 1, except that the
`decision instituting trial found that Kidoguchi did not disclose annealing.
`One of the things I do want to mention with respect to Kidoguchi and the
`base layer argument is even if Emcore were allowed to make an obviousness
`argument for the substitution of aluminum into molybdenum, that
`
`
`
`
` 7
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`substitution would not have shocked anyone of ordinary skill in the art. And
`that's because there are any number of ground 4 references that also use
`aluminum as a base layer to n-type gallium nitride.
`I'm wondering, Mr. Brill, if you could go to page 49 of the
`
`petition.
`
`Page 49 of the petition for clarification is not ground 4, but it
`discusses Shibata, which is a ground 4 reference. The Shibata reference, of
`course, has n-type gallium nitride, additional LED or other gallium nitride
`device. It has an aluminum base layer, a titanium barrier layer and a gold
`top layer, and it yields the structure.
`So, here is another ground 4 reference that uses aluminum as a
`base layer to n-type gallium nitride just as Kidoguchi does. A person of
`ordinary skill in the art would not have been shocked by putting the
`molybdenum in for the aluminum.
`JUDGE CHANG: Counsel, can I ask you a question?
`MR. SMITH: Yes, Your Honor?
`JUDGE CHANG: Can you give us a reason why you would
`want to use aluminum?
`MR. SMITH: Yes, I can give you a reason, Your Honor. In the
`prior art, and in particular, I will say the admitted prior art, Shibata,
`Fujimoto, as well as various other background references that we'll probably
`discuss today, it's very, very clear that aluminum was an excellent choice for
`a base layer to n-type gallium nitride, and, in fact, that one could achieve a
`very low contact resistance and a good element contact under annealing with
`
`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
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`aluminum on n-type gallium nitride. And I will get into a couple of those
`references in a second.
`Mr. Brill, could you go to page 16 of the petition.
`This is another ground 4 reference, this is Fujimoto, again,
`n-type gallium nitride, aluminum base layer, a barrier layer, gold top layer.
`Fujimoto also yields the entire contact structure. So, we're seeing very
`clearly the pattern of people using n-type gallium nitride with aluminum
`base layers.
`I would also like to add that the resulting contact structure that
`is disposed in Kidoguchi, aluminum/titanium/platinum/gold, which is the
`same as the resulting structure in claim 1 of the '215 patent, would not have
`surprised anyone, because that structure was known in the prior art.
`I'm wondering, Mr. Brill, if you could bring up Exhibit 1033 at
`page 23. And this is the deposition testimony of Professor Goorsky, who
`was the testifying expert for Emcore.
`In Professor Goorsky's deposition testimony, we asked him
`whether he would agree that the aluminum/titanium/platinum/gold contact
`was known as a semiconductor contact in 1999-2000, and that's the time
`frame immediately preceding the filing of the provision application that
`leads to the patent in this case.
`And he said, "So, I mean at the time of this, the 1999-2000, was
`it known as a semiconductor contact? That's -- the layer stack existed in
`other publications." So, he's admitting that that contact stack was known in
`the prior art and certainly we see it in a number of references that we also
`cited in the petition outside of ground 4.
`
`
`
`
` 9
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`So, in sum, for this portion of it, Kidoguchi teaches at least
`everything in the claims, aside from the annealing step that was described in
`the petition instituting trial, Kidoguchi expressly teaches using
`aluminum/titanium/platinum/gold contact on n-type gallium nitride as one of
`its species, and even if we were to make a sort of obviousness type case in
`terms of substituting aluminum in, absolutely nobody would be shocked by
`the choice of aluminum out of those four metals there.
`JUDGE CHANG: Can I ask you a question?
`MR. SMITH: Absolutely.
`JUDGE CHANG: I am looking at the patent owner response
`on page 32 through 33 where the patent owner made an argument that there
`is a mistranslation, there's that translation error.
`MR. SMITH: Yes, Your Honor.
`JUDGE CHANG: Can you comment on that?
`MR. SMITH: Absolutely. So, the patent owner's position on
`the mistranslation is that paragraph 31 of Kidoguchi said that titanium and
`aluminum were conventionally used -- conventionally used as base layers,
`okay? They retranslated that to say, in effect, that titanium and aluminum
`were used together to form contacts to n-type gallium nitride. I think that
`issue is largely irrelevant. That is, I don't think that you as the Board have to
`decide between Emcore's translation and our original translation. And that's
`because it will be very clear from the evidence that aluminum was
`conventionally used as a base layer to n-type gallium nitride, as is shown in
`the prior art that we just went through. The fact that Kidoguchi mentions it
`expressly, Shibata mentions it expressly, Fujimoto mentions it expressly,
`
`
`
`
` 10
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`and the admitted prior art, I will put up on the screen very shortly, will
`mention it expressly as well.
`JUDGE CHANG: I have another question.
`MR. SMITH: Yes, Your Honor?
`JUDGE CHANG: Regarding the base layer. The patent owner
`offers a claim construction of that term "base layer." Can you comment on
`their claim construction?
`MR. SMITH: Yes, the construction of base layer, as well as the
`construction of annealing, both seek to introduce the terms ohmic and low
`resistance into claim 1, and I think, first of all, there's no basis to introduce
`ohmic into claim 1, I'm just not seeing that in the '215 patent specification.
`The low resistance introduction into any portion of claim 1
`presents some challenges for Emcore, I think, and it presents challenges
`because Emcore's expert has testified that what he considers to be low
`resistance is about ten to the minus fifth ohm centimeters squared, but that's
`exactly the limitation of claim 15. So, as soon as you introduce that concept
`into claim 1, you've got a claim differentiation problem.
`Now, as the Board found in the decision instituting trial, the
`annealing step, which I think is also an interesting construction along the
`same lines, the annealing step could be carried out in the '215 patent for a
`variety of purposes, including making the contact transparent. There has to
`be some physical property change, presumable that remains there after you
`cool down the annealing -- from the annealing temperature.
`JUDGE CHANG: So, do you agree with the patent owner's
`claim construction on the term "annealing," then?
`
`
`
`
` 11
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`MR. SMITH: Nichia favors the Board's construction for the
`purposes of this proceeding. I think the patent owner's attempt to
`incorporate low resistance into claim 1 would violate the claim
`differentiation doctrine with respect to claim 15, because the patent owner
`essentially agrees with Nichia that the dividing line between poor resistance
`and good resistance was around ten to the minus fifth ohm centimeters
`squared at the time the application was filed, and Professor Goorsky has
`testified that that's what he considered to be about a good resistance.
`JUDGE CHANG: So, let me understand you. So, when one of
`ordinary skill in the art will understand what the term means sufficiently to
`form a contact with low resistance?
`MR. SMITH: I think a person of ordinary skill in the art in that
`time frame would probably know what that means. In reality, I think
`probably reasonable minds could disagree about that point, but the experts in
`this case have agreed that that's going to be something around ten to the
`minus fifth ohm centimeters squared, or a little bit less. Understanding that
`contact resistance is one of those quantities that varies a lot, all right? So,
`it's not something you get really precise measurements on, it can vary by a
`factor of two or ten and it's not going to be something that we're going to be
`too concerned about. But around ten to the minus fifth ohm centimeters
`squared is what the experts have pulled out as one skilled in the art would
`recognize at that time as the dividing line between good and poor resistance.
`JUDGE CHANG: Thank you.
`JUDGE TURNER: May I raise a question?
`MR. SMITH: Absolutely.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 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
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`JUDGE TURNER: One of the arguments patent owner makes
`has to do with the unpredictable nature of the art. So, you've sort of gone
`through this whole context of how aluminum can be used as a base layer, but
`if we accept their argument about the unpredictability, you know, you can
`use any conventional layer as a base layer, but why would it then, if it's
`unpredictable, be obvious to therefore include other layers?
`MR. SMITH: I think the first thing we need to specify to that,
`Judge Turner, is that the Kidoguchi reference actually discloses all four
`layers. So, it's really not a question of being obvious to use one layer or the
`three other layers, but even if we go beyond Kidoguchi, if you look at the
`level of generality of the '215 patent, the disclosure in the '215 patent is at a
`very high level. Emcore's evidence of unpredictability tends to be from the
`early 1990s. And all of the experts agree in this case that gallium nitride
`technology was born in the early 1990s. And there was an intense
`development period between 1991 and the year 2000.
`So, in the year 2000, we are looking at a very different world
`with respect to gallium nitride. In the year 2000, when the provisional
`application was filed, at the level of disclosure of the '215 patent, which is
`very general, it simply put the four layers on top of gallium nitride, anneal it
`and it's going to work. At that level, it's completely predictable, because the
`work in deciding that aluminum is a good base layer has already been done
`by the other workers in the art.
`JUDGE TURNER: Okay.
`MR. SMITH: I think where Emcore argues with the most
`volume is in the patent owner response, if you could pull that up, page 9.
`
`
`
`
` 13
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`This is a summary of the patent owner's argument, the patent
`owner says, "Nichia-cited references, as well as other references at the time,
`powerfully teach away from annealing an n-type contact with an aluminum
`base layer for a III-V semiconductor." Okay, so that's where I think Emcore
`is putting most of its effort into the argument.
`And I really think this argument is a more or less
`litigation-driven fiction. And I want to spend the next couple of minutes
`telling you about the evidence that demonstrates that. I think the first is the
`'215 patent itself. As I just explained in response to Judge Turner's question,
`if there were really some problem with putting an aluminum base layer on a
`III-V semiconductor and annealing it, one might have expected the inventors
`to actually say something about that in the '215 patent. Some disclosure
`like, people in the prior art have had a problem with this in the past, and
`here's how we solved that problem. Here's the technique we're going to use
`that's novel. But none of that is in the '215 patent.
`The '215 patent is absolutely reliant on prior art for its
`enablement on this point, and that's why we cite in re: Epstein in our reply
`brief. A patent owner who is relying on the prior art for enablement is also
`bowing to that admission of enablement in the prior art for obviousness
`purposes.
`
`The next piece of evidence, I think, is certainly the Shibata
`reference, which we just looked at. Shibata teaches an aluminum base layer,
`titanium barrier layer, gold top layer, under annealing at 600 degrees C for
`one minute.
`
`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
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`Chase, can you actually pull up Shibata, which is the translation
`1019 and go to paragraph 26.
`So, the Shibata reference, which already has a structure very
`similar to the one that results from claim 1 of the '215 patent, it's just missing
`the platinum layer, specifically recommends an alloying treatment 600 C.
`What's the result of that? They get a contact with good contact resistance, as
`you can see, it's ten to the minus fifth, Professor Schubert has testified that
`that is, in fact, ohm centimeters squared, and Shibata does actually say
`contact resistance there. They get a good ohmic contact and it's suitable for
`wire bonding in the gold layer.
`Now, Emcore doesn't have a strong teaching away argument
`based on Shibata. Emcore does say one thing. If you could go up to the top
`of column 28. It's up here.
`Emcore looks at this language "dysfunctional as a bonding
`pad." Okay? And what Emcore says about this is Shibata teaches that if you
`anneal this contact, there's a risk that the top layer, that gold layer, will
`become dysfunctional when you try to connect a wire to it and solder it.
`But if you actually look at Shibata's disclosure leading into this
`dysfunctional as a bonding pad sentence, what they say is the optimal
`thickness -- this is the bottom of the left-hand column -- the optimal
`thickness of the titanium layer 82, that's the barrier layer, is 1,000 angstroms
`to one micron. If it is 1,000 angstroms or less, aluminum and gold would
`react with one another in the alloying treatment, rendering the layer 83
`dysfunctional as a bonding pad.
`
`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
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`So, what Shibata is actually saying here is, you need a barrier
`layer and that barrier layer has to be thick enough to prevent the known
`interaction between gold and aluminum. That's not a teaching away as
`expressed in the patent owner response. There's a critical difference.
`In the patent owner response, Emcore is saying, any time you
`have an aluminum base layer on a III-V semiconductor and you anneal it,
`there's a problem. There's some problem at the interface between aluminum
`and the III-V n-type semiconductor, okay?
`The '215 patent doesn't solve that problem, but what Shibata
`says is, if you anneal and your barrier layer is too small, then you might have
`a problem in the gold layer, top. It's very different from what Emcore is
`saying in the patent owner response.
`The next, of course, is Fujimoto, which we have already looked
`at. I wonder if you could pull up, Chase, the Fujimoto patent, which is
`Exhibit 1007. Go to column 16.
`So, the Fujimoto patent, of course, teaches an aluminum base
`layer, a platinum layer, a gold layer, and that the electrode is preferably
`annealed at a temperature around 400 degrees C. Very hard to construe this
`as a recommendation to not anneal a contact with an aluminum base layer on
`n-type gallium nitride.
`The next ground 4 reference I want to talk about is the admitted
`prior art. Can you please bring up Exhibit 1004.
`Your Honor, when did we start?
`JUDGE CHANG: You have ten minutes.
`MR. SMITH: I have ten minutes?
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`
`
` 16
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`
`JUDGE CHANG: Because you reserved 30 minutes.
`MR. SMITH: Thank you very much.
`So, the admitted prior art starts out with this section, "Typical
`low works function metal/metal stack which yield low contact resistance to
`n-type gallium nitride on annealing is aluminum, titanium/aluminum." And
`let me say that there's no dispute about what that sentence means.
`Chase, could you bring up Professor Goorsky's deposition
`testimony again, page 156.
`So, we asked Professor Goorsky, Emcore's expert, what that
`sentence in the background means. If you could go down just a little bit,
`Chase.
`
`We asked him, "It states that aluminum is typically used" --
`typically -- "to achieve low contact resistance to n-type gallium nitride upon
`annealing, correct?" And he says, "Well, it is not described quantitatively
`there, but, yes, that's what it states." Okay?
`So, both experts in this case interpreted the admitted prior art as
`saying, the use of aluminum as a base layer on n-type gallium nitride is
`typical in the prior art. Annealing that is used to form low contact
`resistance.
`
`Could you go back to the admitted prior art again.
`On the second page of this, the inventors say it again, starting
`here, I'm pointing at the sentence that says, "For this reason, most
`metallization schemes to n-type gallium nitride use titanium,
`titanium/aluminum, or aluminum followed by nickel/gold." So, again
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`
`
`
`
`
` 17
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`they're saying that out there in the prior art, most metallization schemes use
`an aluminum base layer to n-type gallium nitride.
`JUDGE CHANG: Counsel? I'm sorry to interrupt.
`MR. SMITH: Yes?
`JUDGE CHANG: There is an argument in the patent owner
`response that your expert did not provide reasons to combine. So, can you
`comment on that?
`MR. SMITH: Right. So, the motivation to combine in ground
`4 is really asking whether it would be obvious to anneal the
`aluminum/titanium/platinum/gold stack of Kidoguchi, right? Professor
`Schubert testified in paragraphs 21 -- 29 to 31 and 63 to 65 of his original
`declaration that annealing was an extremely commonly used technique to
`improve contacts in the relevant time frame. So, almost all contacts were
`annealed. Non-annealed contacts were a kind of exception.
`And that testimony, as far as I can see, is not disputed by
`Emcore. Instead, what Emcore says is, for the specific situation where you
`have an aluminum base layer on a III-V semiconductor n-type, then the prior
`art teaches you not to do that. Okay?
`But as we're seeing, all of the ground 4 prior art actually teaches
`you to do that. It's a specific motivation to combine, and if you want some
`sort of express teaching, suggestion, motivation, the prior art is certainly
`teaching you that annealing will lower the contact resistance. Okay?
`Shibata says that he got a good contact resistance, ten to the
`minus fifth ohm centimeters squared, upon annealing at 600 degrees
`Centigrade, which was within the temperature range of claim 2 of the '215
`
`
`
`
` 18
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`patent. The admitted prior art says that in the first sentence of the
`background, aluminum is typically used as a base layer, which on annealing,
`forms a contact n-type gallium nitride with low resistance. That's
`motivation.
`Here they say, annealing of the metallization is carried out at
`temperatures between 400 and 900 C for minimum contact resistance, and
`everybody agrees that low contact resistance is a good thing in these devices.
`That's express motivation to combine right there.
`Given the time, I would like to quickly touch on the advantages
`of the invention and the secondary considerations that Emcore is recently
`arguing.
`
`Emcore claims that there are four advantages of the '215 patent,
`they actually call it remarkable break-through. The first of these is the very
`fact that you can use aluminum as a base layer. They say that's a surprising
`advantage because aluminum is reflective, and if you put it on a
`light-emitting device or a laser or something like that, the reflectivity will
`help contain the light within the device and project it in the right direction.
`Essentially.
`But Professor Goorsky admitted in his deposition, and that's
`cited in our reply brief, that the reflective properties of aluminum were
`known, prior to the year 2000, as you would expect, and that it was known
`to be desirable to have reflective properties on these types of light-emitting
`devices.
`
`So, Emcore's purported advantage for using aluminum, not only
`is it factually untrue, because we see plenty of references that are using
`
`
`
`
` 19
`
`
`
`Case No. IPR2012-00005
`Patent 6,653,215
`
`aluminum before the year 2000, it is actually a motivation to use aluminum
`in the prior art. Everybody knew that aluminum had reflective properties
`and that reflective properties are good, that's a reason to use the aluminum
`layer.
`
`The second advantage that Emcore brings up is the purported
`low contact resistance, which they claim as about -- and the word "about" is
`critical -- ten to the minus fifth ohm centimeters squared, or less. But
`Dr. Schubert testified, initially, and Emcore embraced the fact that the
`person of ordinary skill in the art would recognize that ten to the minus fifth
`ohm centimeters square is about the dividing line between good and bad. It
`is a middle of the road contact resistance at that time.
`Now, understanding that contact resistance has changed
`throughout the 1990s, because we had this period of intense development
`and research going on from 1991 to 2000, and that's disclosed in Emcore's
`expert declaration in paragraphs 18 and 35, but by 1999-2000, ten to the
`minus fifth ohm centimeters squared was a middle of the road contact
`resistance, it's exactly what you would expect, and it is, in fact, what Shibata
`reports from his contact.
`The last two advantages were the thermal stability and the
`bonding, those two are related in my mind because they relate to the gold
`layer, you don't want the gold layer to break down when it reacts wi