Oral argument — WELLMAN INC v. EASTMAN CHEMICAL

Perhaps I should note, but I think you have observed, that the Court's honored today to have Judge Ronald White from the Northern District of California joining us. And we feel that's at least doubled the IQ of the bench. Hand of counsel before you. Please proceed. Thank you. May it please the Court, my name is Barbara Moore. I represent the plaintiff appellant Wellman Incorporated in the appeal before you. The district court below granted summary judgment against Wellman on two grounds. The first was indefiniteness, and the second was failure to disclose the best mode of practicing the invention. With respect to both of those grounds, the district court basically accepted the characterization of the defendant appellee Eastman Inc., of what was in the record, and of what the evidence was, without really searching the record to see that that's what it said. So my first, yes. The district court's opinion says that Wellman doesn't contest that the inventors viewed titanium 818 as the preferred embodiment. That's incorrect. That's incorrect? That is incorrect. How is it incorrect? The very first sentence of our opposition, which is JA 5436, we say that there was no best mode. We argued no best mode existed. It's a JA 5439. And that has been Wellman's position since the beginning, that there wasn't a consistent best mode. And frankly, it's perplexing to me how the judge, the district But didn't the court rely on the testimony from the inventors themselves, that TI 818 was their preferred way of doing it? The court believed that the court was relying on the inventors' testimony. But this gets to what I was saying before. It seems as though the court accepted the characterizations of what the inventors said that was argued to it by Eastman and really didn't fully appreciate what the inventors were saying. I have examples right here, if your honors would like to. The evidence is ambiguous at best. The court relied on testimony from Stephen Nichols, who was one of the inventors and is a doctor. The first run of the product that we called TI 818 was done at our Pearl River. That was one of the better ones? Yes. The question was, and I'm looking at 297 of the joint appendix. He was asked whether there was a preferred composition. And he says, in what respect do you mean a recipe? Recipe is the word that they use for production, the recipe of what they're doing at the plant. And the date they're asking about is May 21, 2004, because that's the filing date of the application. Here's what Dr. Nichols actually says. The question is, which ones were the better ones? He says, well, the better ones would have been the first run of the product that we called TI 818, which was done at our Pearl River plant in the fall of 2003. Now, there's no evidence that that recipe, using the word of the production run, is the same as what they were making in May 21, 2004, which is actually the recipe that Eastman contends was the preferred embodiment. Where in the specification is the best mode disclosed? Well, as Eastman has made a lot of, there's not a particular recipe disclosed in the patent. However, the- Doesn't that QED, that the statute requires a disclosure of the best mode? There was no best mode. There was no best mode. There was no best mode. There was no best mode. The first prong of the best mode analysis, whether the inventors subjectively had a best mode at the time of the filing of the patent application, there is no evidence, much less clear and convincing evidence, that that existed as of May 21, 2004. How about the carbon black 990? Again, in the patent, there are disclosures regarding different heat upgrade additives. There's a heat upgrade additive in the resin. And the patent, I think you see it, talks about different carbon black additives. It talks about spinels. And it says that copper chromium spinels are particularly outstanding. And that's at JA0052. So it talks about a number of different heat upgrade additives. As of the time of the patent application, the inventors considered the particularly outstanding were the chrome iron nickel black spinel and the copper chromite black spinel, as set forth in the patent. There's much ado about this carbon black, the N990. That was the heat upgrade additive that was used in the commercial embodiment. The court seems to have assumed that because it was the commercial product, it must have been the best mode. And there's no evidence of that. Most commercial products are the best that companies have, aren't they? I don't know that that's true, Your Honor. And the testimony in this case is not consistent with that. The testimony in this case is that although spinels were considered the best, as the patent says, and as it's presumed to be valid, and as it says, that there was concern. And this is in the deposition testimony of the lawyer who prosecuted the patent, Mr. Addeton, that there was concern about freedom to practice that mode, that is the spinels, because there was a patent called Woo that covered the use of spinels. So even though the research that had been done just previously to filing the patent application that led to these patents, and that was in May of 2004, although that research clearly shows, and the spreadsheets are in the record before you, that the spinels were the better heat upgrade additive, there was concern because there was a patent that covered them. What about TCH, indefiniteness? Indefiniteness. I would like actually to focus on indefiniteness, because from a purely practical point of view, if we win on indefiniteness but lose on best mode, we still have claims to try. The best mode decision of the court. But not the reverse. But not the reverse. That's correct. So being purely pragmatic, I would like to focus on the indefiniteness argument. And again, the court seems to have sort of almost shifted the burden to Wellman to prove that the claims were definite. The term that the court focused on is TCH, as Your Honor has pointed out. And the patent claims that TCH allows, it delays the onset of crystallization, thereby facilitating the formation of high-clarity bottles. That's what the patent says. Now, the patent is very clear with respect to what testing needs to be done with respect to determining the TCH. It's set forth in it's JA 0050, column 7, line 64 through 86. It talks about, it sets forth four steps that should be followed on the sample that's being tested. It refers to figures 1 through 8 of the patents, all of which are graphs of TCH for certain samples that were tested. Each of those recites at the bottom the same four steps to test for the TCH. And DCS testing, or the type of testing here, differential scanning calorimetry, sorry, was well known at the time of the patent application. So I don't think there's really any evidence that the testing was not clearly disclosed in the patent. I think what the argument is more about is whether the sample upon which the testing is to be done was clearly described. And the real key there is that the patent says, in at least four places, that the resin has to be amorphous. And that's the key here. And that's what differentiates this case, frankly, from the Honeywell case, or one thing that differentiates it. The sample is described at four places. JA 0050 says figures 1, 3, 5, and 7 correspond to the heating of amorphous polymer. Again, on the same page, column 8, lines 55 through 58, says those having skill in the art will recognize that TCH is determined on a non-crystalline PET resin. How do you know if a sample is amorphous or not? Those with skill in the art would know. There are tests that can be done to determine whether it's amorphous. In this bottle. What's the definition of amorphous? Non-crystalline and non-oriented. And we get that definition from Dr. Schiraldi, who was Wellman's expert. And his expert report covers it in quite a bit of detail. And it's in the joint appendix at 1047 through 1136. And most importantly, we get it from Eastman's own in-house expert, Dr. Quillen. She testified, and it's a JA 7552, that in an amorphous polymer, the chains are not oriented. And she says there's no crystallinity in an amorphous polymer. How much crystallinity is allowed? The patent defines it as less than 4%. And then when they performed the test on the figures, they were all above 4%. I think what you're thinking of there, Your Honor, is there was a formula that was given by Stephen Nichols, Dr. Nichols again, in one of his declarations for a kind of a guesstimate formula for determining crystallinity on a second scan after a quench, which is not what the figures in this patent are. So if you take that sort of guesstimate formula and apply it, which is what I think you're thinking of in Eastman's brief. That's what they did. They just took that math and did it with respect to these and came up with different crystallinities. I think one of them they came up with was a negative crystallinity, which makes no sense. So again, that's taking a formula that Dr. Nichols talked about with respect to something different, a second scan resin after a quench, and applying it to these figures and saying, therefore, you can't tell what the crystallinity is. You can do, and in Dr. Schiraldi's expert report, he talks about tests that are easy, that are well-known, and the industry that you can do to get exact crystallinity. And again, most importantly, the patent says in two places less than 4%. And it's at 0.052 and 0.058. And again, there's no evidence, and the burden is on Eastman here, there's no evidence that that's wrong, improper, not a good estimate, or anything else. There's no evidence here that if you say 4% and you go to 6%, there's no evidence that there's anything wrong with that. Again, if Eastman wanted to prove that that 4% number is not good, it doesn't work, where's the evidence? There really isn't any. So I would submit that the combination of the four steps that I think are clearly set forth in the patent with respect to the testing, combine that with the requirement that the sample be amorphous, which is further defined as less than 4% crystallinity, I believe in and of itself that renders the patent definite. And particularly under the reasoning of the Exxon case and others that say the court should bend over backwards to find claims definite, inferences should be resolved in favor of the patentee in a definiteness analysis. I think there's plenty in the patent itself to render it that's definite. And more to the point, there's no evidence, clear and convincingly, to persuade anybody that it's not definite. You have to read amorphous into the claim, though, right? Well, you have to take the specification where it says amorphous. The word amorphous is not in the claim, that's correct. But there's no confusion, I don't think, in the specification, because it appears four times that it's on an amorphous material. But is the references in the specification dealing with testing? Yes. I believe they are dealing with testing. In fact, let me see, the injection molding conditions give a clear test parison that is predominantly amorphous. And then the 0058, it talks about. That's called main, yeah. That was column eight you just read from. Yeah. It says an amorphous bottle, preform, less than about 4% crystallinity. And this is talking about making the preform for testing. And in fact, the 863 patent only talks about preforms. So the argument that because it appears in a part of the patent that talks about preforms is wrong doesn't frankly make any sense to me, because the whole 863 patent talks about preforms. So I believe that that is fairly throughout the specification and has to be read. The claims have to be read in light of the specification. So I just don't think you can close your eyes to the requirement that the resin be amorphous. And again, Dr. Schiraldi testified, put in his declaration. I see that I'm eating into my rebuttal time here. It looks like I am. Would you like to reserve the rest of your rebuttal time? Why don't I reserve just until I hear what I see? But I just see I'm eating into my three minutes. Thank you, Ms. Moore. Thank you. The expression is ape. The word is ape. Ape, OK. Past tense. I see, ape, that's right. Actually, we're going to give you your whole time back. Could you give Mr. Adamo an additional two minutes, should he need to use that? And then the time will be even on both sides. Thank you. Mr. Adamo, you may proceed. Judge White hit on the first of the problems of arguing that these claims are definite. You've got to read the word amorphous into the claim. But the spec is littered with references to amorphous. It's a little hard to say this is insolubly ambiguous, isn't it? Not at all. The reference that we use, the standard's actually one or the other. If you look at Exxon, it doesn't say insolubly ambiguous. There's an or in there. So that the standard is not rigorously, insolubly ambiguous. You can't give a construction that a person of ordinary skill can tell whether they're inside or outside the claim. Now, let me go back to amorphous. The word that got defined, the phrase that got defined in this section the Court was just asking my opponent about, is predominantly amorphous. The claim doesn't say predominantly amorphous. And this is exactly the point that Judge Robinson pointed to in her opinion, amorphous enough. And if you look at the various uses of amorphous in Dr. Sheraldi's report and other places in the record, it's predominantly amorphous, it's generally amorphous, et cetera, et cetera. Non-crystalline means. If you look at figure one there, right at the bottom it's got the four-step protocol. If you've got a crystalline sample, that's a six-step protocol. Wouldn't want a skill in the art just instantly recognize, we're not talking crystalline samples here, because you're using a four-step protocol. No, a six-step protocol does not, per se, define that the sample is crystalline. And the easy answer to that is, they're saying amorphous can go up to 4%. Their whole argument is, is those figures and the TCH can only be measured on an amorphous sample. And this is where the problem starts. You have to read into amorphous, amorphous, how much crystallinity? How much is amorphous? The opposing definition, or the other definition, amorphous means non-crystalline. More plainly, that's not the position that they took. That's not the position that the specification showed to Judge Robertson. When you apply the calculations to the curves in figures 1, 3, 5, and 7, Your Honor, you're exactly right. The crystallinity numbers come up from between 4 at the low end, all the way up to just around 9 at the top end. And the response to that was that there's no evidence that one of skill in the art wouldn't consider 6% still amorphous, non-crystalline, and clearly within their definition. And that's the first evidence of the fact that there is ambiguity here. Their position is, no, you're not even allowed to take the formula. But one of skill in the art would find this insolubly ambiguous. You've got to show that by clear and convincing evidence. And the fact that we're debating this suggests that it's a factual question, and this is summary judgment. So don't you have about three strikes there that you have to deal with? Not at all. Eastman Chemical, I'm sorry, Eastman Chemical. Exxon, bad histories with Exxon, Your Honor, my apologies. Too much of my life fighting with those people. Exxon Research, that case makes explicit, and Judge Robinson clearly cited it for this purpose, that simply because there are fact issues in indefiniteness. Indefiniteness is an issue of law, just like there are fact issues when you're construing claims. That doesn't bar summary judgment. So this court has already resolved that issue. On the question of, so that doesn't stop the appropriateness of granting summary judgment on indefiniteness. The court's done this a number of times has affirmed summary judgment. Datomize, Halliburton, a variety of cases, even in Honeywell, in fact. Now, going back to the spread of the crystallinity, my opponent, Wellman, is saying, no, the crystallinity is 4% or less. One ordinary skill in the art comes to the figures and does the calculation, which, by the way, Dr. Nichols did say you can do the calculation on those figures. It's not just limited to one or two scans. You end up with numbers that are higher than that. Now, here's the first conundrum. That means, under their definition, the scans, the only data of TCH and the spec aren't within the scope of the claim. That's a red flag right there. So you've got that problem. So you've got to add amorphous. Amorphous could mean who knows what. So there's ambiguity there. Now, orientation. Orientation has an effect on TCH. Orientation is different than amorphous. It's amorphous crystalline. That's the contrast, oriented, non-oriented. Amorphous PET absolutely can be oriented. Honeywell, this court noted that in Honeywell. There are a variety of emissions to that effect. So orientation is not specified. That has an effect on TCH. Orientation, according to the record, JA27 note 19, citing JA361, which is a Wellman R&D document, and also Dr. Schiraldi's deposition testimony, JA8719. Orientation can change the TCH by 17 degrees centigrade. One way or the other, the spec is dead silent about orientation. Mr. Adamo, what about best mode? Yes, there is one. Dr. Nichols said, this is a JA300, page 23, lines 21 to 25. Question. When I asked you before about what was preferred and what was better, you identified TI818. Do you recall that? Yes. And the question of whether there was any other recipe other than TI818 before May 21, 2000 that you felt was better to that question, Dr. Nichols responded no. And the other inventors also supported the fact that TI818 was the preferred recipe at this time. This argument about the 2003 samples was waived. This is a new argument. This is a brand new argument. But Nichols said at a later point, I don't know if we had a preferred body at that time. Well, Your Honor, the point here is you've got two clean admissions from Nichols. You've got admissions from other of the inventors. You've got, as Judge Laurie pointed out, TI818 was commercialized and had been for several years by the time the first non-provisional application was filed. So the boots on the ground, their footprints all over the place in the snow, the TI818 was the preferred embodiment. You've got Nichols admitting it. You've got one of the other inventors admitting it. Excuse me. What if the inventors disagree as to the best mode? Each one has a different best mode. Interesting that Your Honor brought that up. Doesn't make a difference. As long as any one of them has a best mode, that mode's got to be disclosed. They try to argue that Schiavone here at one point, there's something in the record that Dr. Schiavone thought Tungsten was better. Well, all that would mean is if he thought Tungsten was the best mode and Dr. Nichols thought that the best mode was TI818, which doesn't use a metal HUR and uses the trade secret N990, then both of them should have been disclosed. Is there specific intent to conceal here? Doesn't need to be, first of all. Doesn't? No. Best mode can be concealed or not disclosed inadvertently. But is there specific intent? Boy, I don't think I've ever seen better instance of specific intent. The changes, Your Honor, from the provisionals where the spinels were pressed as the best HUR over to the, I'm sorry, I said it backwards. Carbon blacks were pressed as the best HURs in the provisional. But they didn't use the spinels because of the WU patent. No, that's just wrong. Well, and here's why, OK? They never used spinel, period. So to say that we were going to have to come away from the HUR being carbon black, Your Honor. But see, what you're doing is you're urging us to examine products, not claims. It's the best mode of practicing the claims. And your evidence that you bring forward is that, well, they did this in their product. Are we missing something here? Well, I would disagree, Your Honor, that that's what we're doing and what our briefing indicates we're doing. The TI-818, by concession from Wellman, is within all of the claims in issue, the asserted claims, except, as Your Honor found, five of the dependent claims. And there are only two of those dependent claims left. Was one of the inventors told by his boss not to disclose N-990, that they wanted to keep it as a trade secret? Yes. And that, Your Honor, is just Penn Gilly recites N-990. No, it does not, Your Honor. Far from it. Penn Gilly, if anything, hides the needle even further in the haystack. Penn Gilly talks about a preferred range of 10 to 30 nanometers. N-990 is 290 nanometers. And the record is clear that that 290 nanometers, particularly according to Thompson, not only was a big deal, it was a big enough deal that he thought it was patentable. So Penn Gilly is not disclosing in the preferred ranges N-990. And, Judge, well, you're absolutely right. This reminds me, and I hate to date myself by this, this reminds me of the Union Carbide Sixth Circuit best mode case, where there was a disclosure of an extruder, but the key to it was it was a double-barreled extruder with the lands on the twists of very important. The patent said, yeah, there's an extruder. Problem was, it didn't disclose the rest of the detail. It maintained it as a trade secret. Interestingly, because there, Union Carbide was running a program where they were selling the know-how besides the patent. There's no indication that that selling things was going on here, but there is no doubt that they were hiding N-990 because they thought it gave them a competitive advantage. And that hiding, right up to the day we argued this, we were in Judge Robinson's courtroom arguing these motions. And they asked for the courtroom to be sealed because they didn't want anybody knowing what N-990 was and the fact that it had the 290 nanometers right up to that point. They only waived trade secret protection when they had lost below and the opinion was going to be published. So we argued to Her Honor, and her opinion, I believe, supports this, particularly in view of Exxon, that there are really two best modes, one within the other. TI-818, the recipe itself, the exact formulation, best mode, not disclosed. But N-990 is a sub-issue. The specific HUR with the carbon black of a certain size, that wasn't disclosed either. And it not only wasn't disclosed, this was affirmatively concealed. That's beyond peradventure on this record. Our view is, as far as changing the preference between the provisionals and the non-provisional, and thank you, Your Honor, for correcting me. I'm going too quickly here. So sometimes the synopsis don't fire fast enough. That change, it can't be explained on the basis of woe. By the way, Addison, who was the lawyer whose testimony was I had to go read that a little more closely than the way counsel presented it. He didn't mention woe. Did the court err by not doing a claim construction before it reached a summary judgment on best mode? It has to be a best mode of the claimed invention. Don't you need a claim construction first? Under the circumstances here, no. Where do the claims recite that you must use N-990? They don't. As an HUR, the claims recite a carbon HUR. But different, Your Honor, than Teleflex, even if you say the specific. Tell me what the claims recite again. I'm sorry? They give us a. I'm sorry, Your Honor. I got ahead of myself. They recite absorbance properties. Yeah, they recite characteristics. Dr. Schiraldi, I left a step out. Thank you for my colleagues. Absorbance. Dr. Schiraldi testified you need an HUR to get the absorbance property that is recited. And then he went on to say that part of it. Isn't it your view that 818 is the best mode of the claimed invention? Yes. And that 990 is the best mode of that particular HUR component of 818? Yes. Brought in, though, and more properly, I'm sorry I was not trying to mislead the court, brought in by the absorbance value that's reciting, that Dr. Schiraldi testified was then required a carbon black HUR. Part of what they alleged the invention here was, and Schiraldi said, well, the big deal was not just that you needed an HUR, but that you had to learn that any HUR of any size, of any concentration, wouldn't work. That's why the 7 and 1 half parts per million and the 290 nanometers is so important here. But going back to your honor's point, Judge Robinson, citing, controlling, Third Circuit precedent, as well as precedent of this court, said you don't have to construe when no one is making an issue out of it. And in parallel, just so I don't know if the court's aware of how Judge Robinson does that, but the same day, at the same time, we argued claim construction, Markman, we had our Markman hearing. We then argued the indefiniteness motion. We then argued the best mode motion. And there were one or two others that was a very busy day. So front and center, besides what you can see on the face of her opinion, where she's reciting what you need to do to construe a claim, and then she's doing it where there's an issue, she's doing the necessary claim construction. This was a situation, however, that because all of the terms, parameters, the indicia of the construction of the product, there wasn't any challenge to it. And as the case law says, where there's no challenge, you don't have to construe where no one's giving you an argument. Perfect web technologies, this court's opinion, 587 Fed 3rd at 1332, makes exactly this point. And on top of which, when they responded, they only challenged certain limitations in those five dependent claims as to whether or not TI-18 read. And if you look at perfect web technologies, when you make that type of a concession, that takes the issue off the board because it's viewed as being satisfying. So everything Judge Robinson did, and just at this last point as my time is about to expire, Your Honor, is to say that Judge Robinson just accepted the characterizations of Eastman Chemical regarding this record in view of the detail in her opinion on both of these motions, plus her normal practices, this record just doesn't support this. The trial court was not gone for the day in phoning this in. It simply wasn't the way it went. Thank you, Mr. Adamo. Thank you, Your Honor. Ms. Moore, you have three minutes. Thank you. I'd like to start with something that Mr. Adamo just said with respect to this trade secret issue about someone told one of the inventors to keep N-990 as a trade secret. First off, as we pointed out in our brief, there's a real timeline problem here. The inventor testified that that conversation occurred in 2002. The person who had the conversation with him said it happened in the 1990s. Again, this is exhibit A for why best modes shouldn't be decided on summary judgment. I mean, there's a lot of questions of fact here. But Mr. Adamo just stood up and told this court that Wellman kept this trade secret through the date of the argument in November 19th. And I actually looked at the citation that they have for that just the other day. And if you look at Joint Appendix 5231, which is the site, it's Mr. Adamo that requested that the court be sealed at that hearing. Because Eastman, I'm saying right here, during the infringement arguments, Eastman is requesting that the courtroom be sealed for those arguments, both because of the technical information about Eastman's product and also because of information about Eastman's customer agreements. So again, this highlights perfectly my point that you really have to look at the record and see what the record says as opposed to the characterization of it. Wellman didn't ask for the court to be sealed. Eastman did because of its concerns about confidentiality. And again, I think that the whole trade secret thing is a red herring because of the timeline issue that I just related. Let me mention, again, you asked about one of you, and I'm sorry, I don't remember which one now, asked about what if there's disagreement among inventors as to the best mode. Again, I think this is Exhibit B for why best mode should be determined on a full evidentiary record as opposed to on summary judgment. Here we have Nichols maybe saying that some run of the product in 2003 from a particular plant was the one he thought was the best. We've got Schiavone in a declaration saying he thought Tungsten was the best. I don't know if that's an obvious answer. If there's disagreement on that fact within the inventorship community, you disclose them both. Well, I don't. Isn't that the essence of the patent system? If you've got one, if you've got two that are very close, you disclose them? If that were the case. But again, and to comment on the Panu case, I guess, which is the one footnote here from this court, what it was talking about was, what I think that footnote said was, if you're going from having one inventor to two inventors, and you're adding another inventor to the mix, then you're going to have to revisit best mode. Because if you find that the first inventor didn't have a best mode, but you're adding an inventor to the mix and he did, maybe you've got a best mode issue. Now, I don't know whether you really extrapolate that to four inventors. But in this case, there's no clear. There's a difference between very best mode and next best mode. Right. Do you recite inventor A thinks Tungsten, inventor B? But again, you don't even get to that here, because there's a very substantial question of fact as to whether anybody thought there was a best mode as of the time of filing. Thank you, Ms. Moore. OK. Thank you, Your Honor. Appreciate it.