Sionyx, LLC v. Hamamatsu Photonics K.K. (330 F. Supp. 3d 574, 2018)

I. Background

A. Factual Background

The following facts appear to be undisputed.

1. The Parties

SiOnyx, LLC was founded in 2006 by Eric Mazur, a physics professor at Harvard University, and James Carey, his former doctoral student. (ECF 337-40 at 7:5-6, 8:3-9:7). Their goal was to commercialize laser-textured black silicon photodetectors, which had been the topic of Carey's Ph.D. dissertation and postdoctoral work in Mazur's laboratory. (ECF 337-40 at 9:15-11:11). Stephen Saylor joined SiOnyx in the fall of 2006 as President and CEO. (ECF 337-40 at 9:8-14). SiOnyx owns U.S. Patent No. 8,680,591, which it asserts in this lawsuit. (ECF 163-2 ¶¶ 12-17, 19).

The President and Fellows of Harvard College ("Harvard") are the assignees of the other patent asserted in this lawsuit, U.S. Patent No. 8,080,467, which covers Mazur and Carey's work. SiOnyx is the exclusive licensee of that patent. (ECF 342 Ex. G).

Hamamatsu Photonics K.K. is a Japanese integrated photonics company that researches, develops, and manufactures optical devices and image sensors. (ECF 163-2 ¶¶ 54-55; ECF 178 ¶¶ 54-55; ECF 337-41 at 114:5-18). It is the assignee of U.S. Patent Nos. 8,564,087 ; 8,629,485 ; 8,742,528 ; 8,884,226 ; 8,916,945 ; 8,994,135 ; 9,190,551 ; 9,293,499 ; and 9,614,109, in addition to several Japanese patents covering similar inventions. (ECFs 337-22 through 337-31).

Hamamatsu Corporation ("HC") is the marketing and sales company responsible for distributing HPK's products in North America. (ECF 97 at 3). It is a New Jersey corporation with its principal place of business in New Jersey. (ECF 97 at 3). HC is a wholly owned subsidiary of Photonics Management Corp., which is a holding company owned by HPK. (ECF 97 at 3, 13). HC purchases products from HPK at a price set by HPK. (ECF 97 at 3; ECF 382-1 at 62:21-65:7). HC has the authority to set its own resale prices, and it separately profits from its sales to end users. (ECF 97 at 3; ECF 382-1 at 62:21-65:7).

Ocean Optics is a Florida corporation with its principal place of business in Florida. (ECF 163-2 ¶ 6). It primarily sells spectrometers, some of which incorporate photodiodes purchased from HC. (See ECF 529-2 at 6, 9; ECF 529-3 at 107:13-108:20, 146:2-11).

2. The Technology at Issue

The technology at issue involves silicon photodetectors where one surface has been irradiated by a pulsed laser beam.

The photodetectors use p-n photodiodes, which work by transforming light into electrical current. The photodiode is formed from a silicon semiconductor substrate that has two types of charge-neutral impurities: (1) those that donate electrons (n-type impurities) and (2) those that accept electrons (p-type impurities), which can be said to have electron "holes." (ECF 377-1 at 87). When n-doped silicon is placed next to p-doped silicon, it creates a p-n junction, around which the electrons and holes rearrange themselves until they reach an equilibrium. (ECF 377-1 at 88-89). At equilibrium, there is a thin insulating layer at the juncture where the electrons and holes (charge carriers) have recombined (depletion region), and an electric field-created by the ions left behind when the electrons and holes diffused away-preventing further diffusion. (ECF 377-1 at 89; see ECF 201 at 13:10-14:20).

The outermost electrons associated with the silicon substrate are said to be in the "valence band," and have a certain energy. The next-highest energy state available is in the "conduction band." The difference in energy between the valence band and the conduction band is a physical property of the semiconductor material; for silicon, the band-gap energy is about 1.07 eV, which corresponds to light with a wavelength around 1100 nm. (ECF 377-1 at 1, 63-64).

If a photon of sufficient energy (that is, for silicon, one with a wavelength of less than 1100 nm) interacts with the silicon substrate, it may transfer its energy to an electron in the valence band and promote it to the conduction band; in other words, the photon is absorbed. (ECF 377-1 at 63; see ECF 201 at 10:21-11:21). Higher-energy photons will be absorbed closer to the light-incident surface, while lower-energy photons are absorbed deeper in the substrate. (ECF 386 Ex E at HPK0022535; see ECF 201 at 12:8-13:9, 15:11-16:16).

When a photon is absorbed, it creates an electron-hole pair (by promoting an electron to the conduction band). (ECF 377-1 at 93). If the photon is absorbed in the depletion region of the photodiode, the electron and the hole are immediately separated because of the electric field, which creates a current. (ECF 377-1 at 93-94). Photons absorbed too far away from the depletion region are much less likely to produce a current. (ECF 201 at 15:11-16:16).

Thus, in an ordinary p-n photodiode, light enters through one surface of the photodiode and, to some extent, is absorbed in the depletion region, resulting in electric current. (ECF 377-1 at 63). Light that is not absorbed will either go right through the photodiode (in which case it does not contribute to the sensitivity of the photodiode) or reflect off the back surface of the photodiode back into the photodiode, in which case it has another opportunity to be absorbed and turned into current. ( '109 patent, col. 7 ll. 24-34). Whatever portion of that light is still unabsorbed after a second trip through the photodiode will either pass through the light-incident surface (again without contributing to the sensitivity of the photodiode) or be reflected by that surface, and so on. ( '109 patent, col. 7 ll. 34-37). Infrared light is more likely to go through the photodiode without being absorbed than visible light, because its longer wavelength (and correspondingly lower energy) is absorbed deeper in the substrate and its energy may be insufficient to bridge the band gap of the silicon semiconductor. (ECF 377-1 at 6-7, 63-64; see ECF 201 at 9:3-13; 12:1-7).

The technology at issue seeks to improve the sensitivity of the photodiode to near-infrared light by irradiating a surface of the silicon substrate with a laser. That irradiation creates an irregular texture on the surface, so that, instead of being smooth, it has micro- or nanometer-scale features that cause the surface to look black to the human eye. (ECF 377-1 at 6-7). Changing the parameters of the irradiation protocol can change the size and shape of those features. (ECF 377-1 at 55 tbl.3.2).

When applied to the back surface of a photodiode, the irregular asperity has the effect of improving the sensitivity of the photodiode to infrared light. In that case, the light enters the photodiode from one surface, and, as before, some is absorbed by the substrate. But instead of meeting a smooth surface on the backside of the photodetector, the unabsorbed light meets the irregular asperity. Light components that hit the asperity an angles greater than or equal to 16.6° will be totally reflected, and because the asperity is irregular, they will be reflected back toward the first surface and the side surfaces in many different directions. ( '109 patent, col. 7 ll. 39-50). Because they are arriving from all different directions, they "are extremely highly likely to be totally reflected" on the first and side surfaces, and therefore to be "repeatedly totally reflected on different faces to further increase their travel distance" inside the photodiode. ( '109 patent col. 7 ll. 51-59). By increasing the travel distance of light inside the photodiode, the asperity makes a thinner piece of silicon act "thicker," and infrared light that otherwise would pass through can be absorbed "deeper" than the photodiode actually is. (See ECF 201 at 16:12-18:3). The longer the light is trapped within the photodiode, the more likely it is to be absorbed and generate current, and the more sensitive the photodiode will be. ( '109 patent, col. 7 l. 59-col.8 l. 2; see ECF 201 at 12:8-13:9; 17:20-18:3; 20:6-21:3).

3. HPK's Knowledge of Harvard's Early Black Silicon Work

On January 12, 2006, James Roberts (of HC's University Sales and Marketing department) sent a report about Harvard's black-silicon technology to Koei Yamamoto (Senior Executive Managing Director and General Manager of HPK's Solid State Division). (ECF 482-3). It summarizes ten academic publications and two published patent applications that list Carey and Mazur as inventors: U.S. Patent Application Publication Nos. 2003/0029495 (Application No. 10/155,429) and 2005/0127401 (Application No. 10/950,230). Those applications are ancestors of the '467 patent, which is asserted in this action, and the '446 patent, which has been dropped from this action. Roberts updated the report in October 2006, to explain that Carey had left Mazur's lab to found SiOnyx and that SiOnyx had received seed funding, and emailed it to Yamamoto again. (ECF 474-9; ECF 526-2). The updated report cites to the same ten publications and two published patent applications that the original report cites to.

HPK produced a slide presentation created by Terumasa Nagano (of HPK's Microelectrical Mechanical Systems Manufacturing Development Group) in November 2006, around the time that it began discussions with SiOnyx. That presentation describes Mazur and Carey's laser-texturing work at length. (ECF 377-14 at HPK0068525-39; see ECF 536-1 (Harvard's U.S. Pat. App. Pub. No. 2005/0127401 ) ). It also includes two device architectures with surfaces textured by a silicon-dry-etching technique. The first shows a textured surface on the same side as the p-n junction (like Carey's preferred architecture, described below). The second has a textured surface opposite the p-n junction (like Carey's Alternative # 1, described below). (ECF 377-14 at HPK0068533). Neither diagram indicates the surface through which the light enters the device, and the second needs to be flipped upside-down to match up with Alternative # 1. An earlier slide in the same presentation states: "During dry-etching of Si, debris, naturally oxidized film, and so forth that stuck to the Si surface become a micro-mass, and the pillar-shaped structure is formed. ? A failure for Si etching." (ECF 377-14 at HPK0068525).

4. The 2007 Nondisclosure Agreement

In late 2006, SiOnyx, through Mazur, reached out to HPK about a possible business relationship. (ECF 337-40 at 14:3-15:15; ECF 353 Ex. B at 119:4-24). Mazur, Carey, and Saylor went to Japan in November 2006 to meet with representatives of HPK, where Mazur gave a presentation introducing the technology. (ECF 386 Ex. F at 22:4-25:8; see id. Ex. E).

SiOnyx and HPK entered into a mutual nondisclosure agreement on January 11, 2007, to facilitate a possible business relationship. (ECF 337-2). The nondisclosure agreement provides:

Each of the parties has developed certain products, technology and methodologies, including information that each party regards as confidential, proprietary, trade secret information. Each party proposes to disclose certain of such information to the other party, to be used by the other party solely for the limited purpose of EVALUATING APPLICATIONS AND JOINTS [sic] DEVELOPMENT OPPORTUNITES OF PULSED LASER PROCESS DOPED PHOTONIC DEVICES and for no other purposes whatsoever (the "Permitted Purpose.").

(ECF 337-2 at 1). SiOnyx and HPK agreed that any breach of the nondisclosure agreement would constitute irreparable harm, so that the "Disclosing Party" would be entitled to equitable relief to enforce the agreement. (ECF 337-2 at 2). And they agreed that all ownership rights in any intellectual property arising from "Confidential Information" would remain with the "Disclosing Party" in the absence of a separate written instrument expressly granting those rights. (ECF 337-2 at 2).

5. The Confidential Architectures

On January 16, 2007, Saylor and Carey met with Keith Kobayashi (of HPK's International Division) and Yamamoto by telephone to plan possible experimental prototypes. (ECF 337-3; ECF 337-40 at 17:13-19, 19:19-20:9).

The following day, on January 17, 2007, Carey emailed Kobayashi "the first draft of a device architecture we would like to pursue." (ECF 337-3 at 2). He explained that "[t]he suggested device architecture was selected based on our past experiments and what we believe is compatible with current Hamamatsu photodetectors," and he "also included two possible alternatives if the preferred architecture is difficult." (ECF 337-3 at 2-3). The "preferred" device architecture showed a laser-processed layer on the top of the device. (ECF 337-3 at 4). The first alternative-"Alternative # 1"-showed a p-n photodiode where the laser-textured layer was positioned on the back of the device, opposite the side where light would enter the device. (ECF 337-3 at 5; ECF 337-40 at 19:13-20:21).

Carey testified that prior to his January 17, 2007 email, neither he nor anyone else at SiOnyx had ever discussed the idea of locating the laser-textured layer as depicted in Alternative # 1 with anyone outside SiOnyx. (ECF 337-4 at 345:20-346:2; ECF 337-40 at 20:10-21:1).

6. The SiOnyx-HPK Collaboration

On April 4, 2007, Saylor traveled to Japan to meet with representatives of HPK. At the meeting, HPK representatives showed a presentation created by Akira Sakamoto (of HPK's Solid-State Production Development Group) at the direction of Yamamoto. The presentation outlined a plan for HPK to make four types of silicon test wafers, each created by up to three different processes. (ECF 337-8 at 55:5-57:17; ECF 337-7). The wafers would be laser textured by SiOnyx, sent back to HPK for final processing, and tested for their optical-response characteristics. (ECF 337-8 at 14:6-15:3). That presentation showed architectures very similar to those proposed by Carey, and included devices where the texture was placed on the light-incident surface. (Compare ECF 337-3 with ECF 337-7).

Between April and November 2007, SiOnyx and HPK jointly tested 38 wafers. HPK fabricated the test devices up until the laser-texturing step, and mailed them to SiOnyx for texturing. SiOnyx then returned them to HPK for final processing and testing. (ECF 337-8 at 14:3-15:3; ECF 337-40 at 44:21-45:20). The results of the testing showed that at least some of the devices that were laser-textured by SiOnyx had improved infrared photosensitivity as compared to one of HPK's standard devices. (ECF 337-9 at HPK0010411). The testing also showed that the photosensitivity of devices having the laser-textured surface on the side of the device opposite from the direction of incident light (front illuminated, in this case where the textured surface is on the back) had stronger performance than devices having the laser-textured surface on the same side as the incident light (back illuminated), which performed significantly worse than the reference device. (ECF 337-11 at 14).

Some knowledge about laser-texturing devices was in the public domain-due, in part, to Mazur's many academic publications on the topic. Carey testified, however, that SiOnyx had discovered that there was a preferred target size for the structures making up the texture, which balanced optical response against certain disadvantageous properties. That texture, and the process for making it, was confidential information of commercial value to SiOnyx. (ECF 337-40 at 21:2-22:1). SiOnyx, when it textured the test devices from HPK, used its confidential process to produce its preferred texture. (ECF 337-40 at 49:21-50:14). But SiOnyx shared very limited information as to the process parameters for achieving that texture with HPK-nothing except the identity of the ambient gas in the laser-processing chamber. (ECF 337-10; ECF 45-2 at 69:4-15, 94:3-7). As part of the testing procedure, HPK took scanning-electron-microscope ("SEM") images of the textures, which showed detailed images of textures achieved and allowed structural features of the textures to be measured. (ECF 337-9 at HPK0010405-08, HPK0010415). Carey testified that the size of the features shown in these SEM images were within the target range identified by SiOnyx, which he considered to be confidential. (ECF 337-40 at 49:5-50:14).

An employee of HC was at the initial November 2006 meeting, and another employee of HC actually signed the nondisclosure agreement on behalf of HPK, but following that there were no communications between SiOnyx and HC. (ECF 97 at 13; ECF 345 Ex. Q at 172:10-19, Ex. S at 255:3-256:21, Ex. U at 121:10-15).

7. The End of the Collaboration and HPK's Further Activities

After the testing was finished, Saylor and Kobayashi exchanged a few emails concerning the possibility of further tests. (ECF 337-12). But on January 15, 2008, Kobayashi responded as follows:

After discussing with related people mainly from technical aspects, we reached to the following conclusions.

As a commercial entity, [HPK], of course, has to study the possibility to enhance our product capability. However, we are not confident that black silicon technology will greatly contribute. We would rather like to stick with our own technique/technology for that purpose because we would like to keep our own pace of development and accumulate our own know-hows. In [HPK] culture, business decisions almost always come after full technical evaluation, which seems to be quite different from your company. Therefore, we would like to do study by ourselves without further reference to proprietary information of SiOnyx.

(ECF 337-12 at 1). The collaboration effectively ended at that time.

That same day, someone from HPK's Central Research Laboratory emailed Kobayashi to say that technology regarding the processing of black silicon was a top priority. (ECF 337 ¶ 45 (Pl. SMF); see ECF 377 ¶ 45 (Def. Response) ).

Nagano testified that HPK kept the test wafers than had been textured by SiOnyx until at least 2010. (ECF 337-13 at 84:5-16). Prototype reports from April 29 and May 22, 2008, show that HPK's Central Research Laboratory was attempting to replicate the quantum efficiency of the SiOnyx-textured device and comparing the textures they were able to achieve to those SiOnyx achieved. (ECF 337-36 at SIONYXHARVARD_00107908, SIONYXHARVARD_00107916; ECF 337-43 at HPK0012401, HPK0012412, HPK0012425; see ECF 337 ¶ 53). A presentation dated July 17, 2008-authored by Sakamoto and Nagano and titled "Black Silicon Technology In-House Production"-has, as its first page, a slide titled "PD sensitivity characteristics increase by SiOnyx Company laser processing." (ECF 337-13 at 131:22-132:25; see also ECF 337-14 (mostly in Japanese, but containing the words "Black Si" and "SiOnyx" in English) ). That slide shows a graph of photosensitivity at different wavelengths of light (the same characteristics measured during the SiOnyx-HPK collaboration) that compares the performance of SiOnyx's device (labeled "SPL Si PD") with HPK's standard silicon photodiode (labeled "STD Si PD"). (ECF 337-13 at 134:11-136:10; ECF 337-14 at HPK0038509). Another slide in that presentation states that HPK had attempted its own laser texturing, but was not able to achieve the same infrared-sensitivity improvement as SiOnyx. (EDF 337-13 at 137:16-139:15; ECF 337-14 at HPK0038511). And another slide contains a schedule, as to which Nagano testified:

Q. Just looking at row 6, the first element in the rightmost column says prototype 1 comparison with SiOnyx; is that correct?

A. Yes.

Q. And in the next row down, row 7, does that refer to progress?

A. Yes.

Q. And the first item under progress is completed data acquisition of existing patterns SiOnyx laser processing; is that correct?

A. Yes.

(ECF 337-13 at 140:4-20; see ECF 337-14 at HPK0038512).

Another HPK presentation, dated July 31, 2008, contains a slide titled "laser process conditions," and shows three scanning electron microscope ("SEM") images. (ECF 337-13 at 142:15-23; ECF 337-17 at HPK0024515). The leftmost image is labeled "SiOnyx." (ECF 337-13 at 142:24-25; ECF 337-17 at HPK0024515). Under the third image, the text reads "visually equivalent to SiOnyx's wafer" and that, like SiOnyx's wafer, "no silicon scum is attached." (ECF 337-13 at 145:4-13; ECF 337-17 at HPK0024515).

Sakamoto testified that HPK's Central Research Laboratory had examined SiOnyx's publications and attempted to estimate the process conditions for the laser texturing. However, it was unable to recreate the texture shown in those publications, and saw almost no surface roughness. (ECF 337-8 at 119:5-122:16). He also testified that "[t]he samples provided to us by SiOnyx, we looked at them in the prototyping of the silicon photodiode black silicon processing." (ECF 337-8 at 122:1-16). Yamamoto testified that for the photodiode product that HPK produced, the "wafers were prepared by Hamamatsu without using the wafers from SiOnyx." (ECF 337-19 at 140:12-15). He further testified:

Q. And that's because the Solid State Division had wafers that SiOnyx had created the texture and worked to replicate that texture and the performance of that texture before you released a commercial product; isn't that true?

Mr. Simmons: Objection to form.

The Witness: No.

By Mr. Belanger:

Q: And you agree that if [HPK] had done that, that would have been a violation of your agreement with SiOnyx, correct?

Mr. Simmons: Objection to form.

The Witness: Yes.

(ECF 337-19 at 140:24-141:13).

8. The 2009 Photonics Fair Emails

A few years later, on February 9, 2009, Kobayashi sent an email to Saylor and Mazur at SiOnyx stating that "[HPK] will introduce various products under development at our general exhibition, PHOTON FAIR 2009 in February in Hamamatsu, Japan" and that "[o]ne of these products is Silicon Photodiode with higher sensitivity covering through 1200nm range." (ECF 316 Ex. E at 1). It reminded Saylor and Mazur that when the HPK-SiOnyx talks terminated, HPK had "informed you that we, by ourselves, would like to focus our development efforts on photovoltaic type," and explained that "while we are greatly appreciating having given us an opportunity to think about further enhancement/improvement of our accumulated know-how as a photonics company, we do not think we are infringing any of your IP or originality, or breaching any obligation of confidentiality." (Id. ).

The email attached a file showing the architecture of HPK's device and a chart of its spectral photosensitivity. (Id. at 2). The email also described the device, stating that it "has PN junction on the one side and the back side consists of an accumulation layer by ion implantation over the black silicon surface fabricated by laser in the inert gas atmosphere." (Id. at 1). Saylor responded by email ten days later, as follows:

Regarding our prior collaboration and information exchanged under Mutual Non-Disclosure Agreement, SiOnyx is confident that [HPK] will ensure the integrity of SiOnyx confidential information. While the diagram provided in your email is insufficient for our understanding, it looks very similar to the work product of our collaboration. Should [HPK] wish to provide SiOnyx detailed specifications and English translation versions of the presentation materials planned for the Photon Fair, we may be able to comment on your conclusion that your laser processed NIR enhanced Silicon Photodiode does not utilize SiOnyx IP or violate any provisions of our prior agreement.

(ECF 316 Ex. H at 1).

On February 24, 2009, Kobayashi replied with the following assurances:

First of all, we would like to emphasize that we only applied our own know-how and technology to the developed products, which we will introduce at our Photon Fair.... Although we briefly discussed the results, our structure and processes in the attached file are our own idea. Therefore, we strongly believe that our developed product does not infringe your patents or use any of your proprietary information disclosed to us because the development is based on only our own wafer process technologies. We believe that following points are explicit differences.

We do not use Femto second laser for fabrication of black silicon surface[.]

Laser treatment is in the inert gas atmosphere[.] We form backside accumulation layer by ion implantation and high temperature annealing[.]

All of above technologies are our own technologies.... [HPK] believes that we only applied our own know-how and technologies to the development of photovoltaic type Silicon Photodiode to be introduced at the Photon Fair.

(ECF 316 Ex. I at 1). That email attached a step-by-step process flow diagram, in English, describing how HPK's photodiode was made. (ECF 316 Ex. I at 4). HPK admits that the process-flow diagram was very similar to one discussed by the parties during the collaboration. (ECF 315 at 2-3; see ECF 316 Ex. B at 2).

Saylor replied on March 10, 2009, that SiOnyx would "review the information provided," and reiterated its interest in a possible business relationship with HPK. (ECF 316 Ex. K at 1). That same day, Kobayashi replied that HPK was not interested in a business relationship with SiOnyx at that time. (ECF 316 Ex. L at 1). On April 15, 2009, Kobayashi contacted Saylor again to say that he hadn't heard from him about the proposed disclosure at the Photon Fair and to thank him for his "understanding that our technique explained in our e-mails is not infringing your intellectual property." (ECF 316 Ex. M at 1). There was no other communication between the parties about the photodiode HPK planned to show at the Photon Fair.

The parties were asked about those communications in their depositions. Carey testified that HPK's first email was concerning because it seemed to him that the diagram HPK sent depicted a device that would be covered by SiOnyx's intellectual property. (ECF 316 Ex. F at 265:16-266:11, 273:17). Mazur also testified that he thought the device shown in the diagram would be covered by SiOnyx's intellectual property. (ECF 316 Ex. G at 157:14-24). He further testified that the differences HPK pointed out in their second email failed to convince him that the HPK product did not infringe SiOnyx's rights. (ECF 316 Ex. G at 161:5-24, 163:1-165:24, 167:1-171:24). Saylor, testifying as SiOnyx's Rule 30(b)(6) deposition witness, stated that the architecture of HPK's photodiode was "the same architecture from an optical perspective as what Jim Carey disclosed." (ECF 316 Ex. J at 191:16-23). But he also testified that while Carey and another employee had suspicions because it was close to what SiOnyx and HPK had worked on together, Saylor "emphasized with them that [HPK] is clearly articulating to us that they had independently developed all of this." (ECF 316 Ex. J at 192:4-19). He read the emails "to mean that none of the people who worked on the project with us would have worked on this, because in big companies, you typically-what you would do is, you would firewall your team, so I'm assuming that's what they've done; otherwise, I don't know how he can claim that it's all their own know-how. They must have firewalled the team." (ECF 316 Ex. J at 189:3-10). He found Kobayashi's final email to be "amusing" because it thanked him for an understanding that he never gave, but he never corrected him or sent a response. (ECF 316 Ex. J at 201:14-202:15).

A certified translation of an email sent on February 6, 2009, from Sakamoto to Kobayashi contains a draft of the email to SiOnyx. The text above that draft states:

Thank you for all you'd done during the infrared high sensitivity device studies performed with SiOnyx the year before last. One sample prepared at that time was commercialized this time and the technology will be displayed at the Photon Fair. Although considered not likely to be in conflict with the patent filed by Harvard College, because there is a description, "photodetector that detects even at the wavelength of energy at or less than the Si bandgap due to a femtosecond laser processing" in one claim, and the N2 gas atmosphere is also described, we have been reviewing how to make the presentation. As a result, if it is commercialized and becomes successful business in the future, rather than hiding a risky notation, we have decided to appeal that our processing is completely different than the one performed by SiOnyx by clarifying the specifics. It is a fact that it was jointly studied in 2007 and since the structure of the sample fabricated at that time and the technique used were already defined, please convey this and notify of the commercialization.

(ECF 326-4 at 6-7). In response to the draft, Kobayashi asked, "Was the sample prepared at that time the one we prepared? And what kinds of information did we receive from SiOnyx for our preparing the sample?" (Id. at 5). Sakamoto responded that the sample was the one HPK had prepared, but that SiOnyx had done the black-silicon laser processing; that he thought they had received some information concerning the laser conditions from SiOnyx; and that while they had also received information concerning annealing conditions from SiOnyx, they were not currently using those annealing conditions. (Id. ).

When asked about that email exchange, Kobayashi testified as follows:

Q. [Sakamoto] suggests that you tell SiOnyx that you intend to commercialize the structures and methods that you jointly studied in 2007, correct?

....

THE WITNESS: Well, if you put the two parts together, it may come out to that. What we wanted to impress upon SiOnyx was that what we are doing is something completely different. The background to this is that there is the fact that the way of making the structures was something that we had worked on jointly and wanted to make that clear to SiOnyx that what we were doing was something different.

(ECF 326-6 at 109:5-23).

Q. So you-your understanding of this email is Sakamoto was trying to give the impression that the [HPK] products were based only on information in the public domain?

A. I believe what it was, was to give the impression that it was information in the public domain and also information that belonged to HPK itself.

(Id. at 123:18-25).

9. The Patents of HPK

As relevant to this lawsuit, HPK owns nine U.S. patents relating to silicon photodetectors with a textured surface that improves absorption of near-infrared light. U.S. Patent Nos. 8,564,087 ("the '087 patent") ; 8,629,485 ("the '485 patent") ; 8,742,528 ("the '528 patent") ; 8,884,226 ("the '226 patent") ; 8,916,945 ("the '945 patent") ; 8,994,135 ("the '135 patent") ; 9,190,551 ("the '551 patent") ; 9,293,499 ("the '499 patent") ; and 9,614,109 ("the '109 patent"). (ECFs 337-22 through 337-31).

HPK started applying for its Japanese patents in February 2009, and U.S. patents in February 2010. One or more of Sakamoto, Yamamoto, Nagano, and Kazuhisa Yamamura (another employee of HPK) are listed as inventors on every patent except the '226 patent. Seven of the nine patents (the '087, '528, '945, '485, '135, '551, and '109 patents ) cite to patents or published patent applications in the same family as Harvard's patent, discussed below. Six of the nine patents (the '485, '528, '945, '135, '551, and '109 patents ) claim priority to Japanese Patent Application 2009-041078. The "invention submission form" of that Japanese application mentions the collaboration between SiOnyx and HPK as part of the "background and motive of the invention." (ECF 337-34; ECF 337-35 at SIONYXHARVARD_00107991; see ECF 337 ¶ 70 (Pl. SMF); ECF 377 ¶ 70 (Def. Response) ).

Those same six patents describe an embodiment wherein the photodiode is configured to have a laser-processed layer on the surface opposite to the light-incident surface, and contain, as Figure 11, a diagram of an architecture very close to the architecture Carey provided to HPK as "Alternative # 1." ( '485 patent, fig.11, col. 7 ll. 8-12; '528 patent, fig.11, col. 8 ll. 18-23; '945 patent, fig.11, col. 6 ll. 31-36; '135 patent, fig.11, col. 8 ll. 25-30; '551 patent, fig.11, col. 6 l. 66-col. 7 l. 4; '109 patent, fig.11, col. 7 ll. 18-23). The '087 patent shows a more complicated architecture at Figure 6 that nonetheless incorporates the basic idea of an irregular asperity opposite the p-n junction and incident light. ( '087 patent, fig.6, col. 7 ll.7-26).

The six patents containing an architecture close to Carey's "Alternative # 1" also contain a figure showing the invention's increased photosensitivity in the near-infrared range, which is similar to the results obtained from the HPK-SiOnyx joint testing. ( '485 patent, fig.12, col. 8 ll. 31-35; '528 patent, fig.12, col. 9 ll. 32-46; '945 patent, fig.12, col. 7 ll. 44-58; '135 patent, fig.12, col. 9 ll. 38-52; '551 patent, fig.12, col. 8 ll. 13-27; '109 patent, fig.12, col. 8 ll. 32-46). The '087 patent contains a figure showing somewhat improved sensitivity from illuminating the device from opposite the textured surface as compared to illuminating from the side with the textured surface, and much improved sensitivity from the use of texturing as compared to no texturing. ( '087 patent, fig.13, col. 10 l. 64-col. 11 l. 13).

All nine patents teach an "irregular asperity," and contain a figure showing an SEM image of the asperity. ( '087 patent, fig.5, col. 6 ll. 63-66; '485 patent, fig.8, col. 6 ll. 17-20; '528 patent, fig.8, col. 7 ll. 27-31; '226 patent, fig.6, col. 7 ll. 25-29; '945 patent, fig.8, col. 5 ll. 40-44; '135 patent, fig.8, col. 7 ll. 33-37; '551 patent, fig.8, col. 6 ll. 8-12; '499 patent, fig.8, col. 7 l. 61-col. 8 l. 1).

10. The Patents of SiOnyx and Harvard

One patent asserted in this lawsuit is owned by SiOnyx outright: U.S. Patent No. 8,680,591, which names six SiOnyx employees as inventors (but not Carey or Mazur). The other patent asserted in this lawsuit is owned by Harvard: U.S. Patent No. 8,080,467 (naming both Mazur and Carey as inventors). Harvard, however, licensed that patent to SiOnyx on an exclusive basis. (ECF 342 Ex. G). The Third Amended and Restated Exclusive Patent License Agreement dated July 27, 2015, gives SiOnyx the exclusive right to develop, make, use, sell, offer to sell, lease, or import the technologies covered by the claims. (ECF 342 Ex. G at 7-8). In return, Harvard received a partial ownership interest in SiOnyx and the rights to certain royalties. (ECF 342 Ex. G at 11-17).

11. HPK's S9840 CCD Sensor

HPK, HC, and Ocean Optics rely on HPK's S9840 CCD sensor in the invalidity contentions for the '591 patent. That sensor has been sold by HPK since 2005. (ECF 485-10 at 154; ECF 523 at C-4 through C-6). Defendants' invalidity contentions cite to an August 2007 datasheet and an October 2007 manufacturing specification, (ECF 523 at C-4 through C-6), as does the opinion of Dr. Shukri J. Souri, defendant's expert, (e.g. , ECF 485-10 at 155-56, 229, 236, 239-41). The datasheet was publicly distributed by HPK, but did not show the internal layers of the device or the manufacturing steps. (ECF 522-1 ¶ 20 Response). The manufacturing specification does show those layers and manufacturing steps, but is an internal HPK document that was not published or distributed to customers. (ECF 523-1 ¶¶ 3-5).

Defendants' expert, Dr. Souri, testified that in forming his expert opinion, he relied on documents, rather than the physical product itself:

Q. You're relying on a Hamamatsu product as prior art for the '591 patent ; correct?

A. That is correct.

Q. Did you test a sample of that product to support your opinion?

A. No, I relied on publicly available information regarding that product.

Q. Datasheets; correct?

A. Yes.

(ECF 485-12 at 18:19-19:4).

12. Damages Evidence

Ocean Optics manufactures spectrometers, including the Maya2000 Pro and the Maya2000 Pro-NIR. The Maya2000 Pro incorporates Hamamatsu's (not-accused) S10420 CCD image sensor and uses aluminum mirrors. (ECF 529-3 at 50:4-11, 68:6-18). The Maya2000 Pro-NIR model incorporates Hamamatsu's (accused) S11510 CCD image sensor and uses gold mirrors, making it uniquely suited to taking measurements in the near-infrared spectrum. (ECF 529-2 at 9; ECF 529-3 at 68:6-18).

Starting in October 2009 and extending to late 2010, Ocean Optics approached SiOnyx about supplying an image sensor for one of its spectrometers that would be able to extend its spectral range. (ECF 529-11 at 81:12-87:15; ECF 529-16). It appears that SiOnyx supplied some sensors to Ocean Optics at that time, but it is not clear whether Ocean Optics ever sold a spectrometer that incorporated a SiOnyx sensor. (ECF 529-11 at 85:17-87:15). In early 2011, Ocean Optics reached out to SiOnyx again regarding linear arrays. (ECF 529-11 at 89:1-91:4). SiOnyx did not have a sensor that met the specifications of Ocean Optics at that time, but indicated that it could develop one, although there would be nonrecurring engineering costs associated with that effort, which presumably SiOnyx would expect Ocean Optics to share. (ECF 529-11 at 91:10-99:15). Ocean Optics ultimately chose to use the S11510 manufactured by HPK, which it purchased through HC. (ECF 529-6 at 38:4-16; ECF 529-11 at 79:3-12, 99:16-24).

Plaintiffs' damages expert, Pauline Booth, provided a report in which she calculated reasonable-royalty damages based on a "hypothetical negotiation between SiOnyx and HPK and HC together, though the analysis would be the same for one party or the other because all of the relevant unit sales are overlapping and HC's 30(b)(6) witness indicated that 'for anything having to do with patents, we don't deal with it' generally deferring to HPK." (ECF 491-2 ¶ 122). She explained that she had "not analyzed a separate hypothetical negotiation with Ocean." (Id. ). When asked about damages for Ocean Optics specifically at her deposition, she testified:

Q: Do you give an opinion of damages for Ocean Optics in this case?

A. I do not.

Q. Why not?

A. I wasn't requested to split out any damages for Ocean Optics.

(ECF 491-3 at 88:4-9). However, the underlying data on which she based her opinion breaks out sales of the accused sensor from HC to Ocean Optics, and includes Ocean Optics' own data on sales of its accused spectrometer. (ECF 491-2 at 28 n.113, 68 n.284 (explaining that the underlying sales data from HPK was broken down into "sales to the U.S. for Ocean Optics" and "for U.S. other than Ocean Optics" and that sales data produced directly from Ocean Optics, although not utilized by her damages opinion, are nevertheless in the record); ECF 529-18 at 5:9-13:1 (testimony from Ocean Optics' 30(b)(6) deposition witness about spreadsheets detailing sales of the Maya2000 Pro-NIR) )

B. Procedural Background

SiOnyx and Harvard filed this action on October 1, 2015. The second amended complaint, deemed filed on May 8, 2017, asserts 24 counts. Counts 1-14 and 21-24 seek relief under 35 U.S.C. § 256 and assert that Mazur and Carey should be named sole joint inventors on each of HPK's nine patents, or, alternatively, that they be named joint co-inventors with the inventors currently listed on those patents. (ECF 163-2 ¶¶ 127-168, 262-273). Count 15 seeks declaratory judgment of the same. (Id. ¶¶ 169-173). Count 16 alleges that HPK breached the nondisclosure agreement. (Id. ¶¶ 174-181). Count 17 alleges that HPK and HC were unjustly enriched. (Id. ¶¶ 182-188). Count 18 has been voluntarily dismissed. (ECF 422). And Counts 19-20 allege that all defendants have infringed each of SiOnyx and Harvard's patents. (ECF 163-2 ¶¶ 215-261).

On September 30, 2016, HC petitioned the Patent Trial and Appeal Board ("PTAB") for inter partes review ("IPR") of the '591 patent. (ECF 485-6) (PTAB IPR2016-01910). That petition identified HPK as a real party-in-interest, and requested that an IPR be instituted as to all 26 claims of the patent, as to one or more of five grounds. (ECF 485-6 at 1, 18-52). Specifically, the petition requested review of:

• Claims 1, 4, 5, 7-9, 13, 24, and 25 for anticipation by Nakashiba;

• Claims 1-5, 7-9, 11, 13-22, and 24-26 for obviousness over Akahori in view of Mazur;

• Claims 1, 2, 4, 5, 7-11, 13-21, and 23-26 for obviousness over Mabuchi in view of Mazur;

• Claim 6 for obviousness over Mabuchi in view of Mazur and Uematsu; and

• Claim 12 for obviousness over Mabuchi in view of Mazur and Furukawa.

(ECF 485-6 at 18-52).

The PTAB instituted review on March 30, 2017. (ECF 485-8). But, as was its practice at the time, it instituted review only as to certain claims and certain grounds on which it deemed HC to have demonstrated a reasonable likelihood of prevailing:

• Claims 1, 4, 5, 7-9, 13, 24, and 25 for anticipation by Nakashiba;

• Claims 1, 2, 4, 5, 7-11, 13-18, 21, and 23-26 for obviousness over Mabuchi and Mazur;

• Claim 6 for obviousness over Mabuchi, Mazur, and Uematsu; and

• Claim 12 for obviousness over Mabuchi, Mazur, and Furukawa.

(ECF 485-8 at 42-43). The PTAB determined that HC had not demonstrated a reasonable likelihood of prevailing in showing that claims 1-5, 7-9, 11, 13-22, and 24-26 were obvious over Akahori and Mazur because Akahori explicitly discourages texturing on the key surface. (ECF 485-8 at 27-28). It determined that HC had not demonstrated a reasonable likelihood of prevailing in showing that claims 19 and 20 were obvious over Mabuchi and Mazur because those references did not teach "forming surface features to have dimensions that selectively diffuse or selectively absorb a desired wavelength of electro-magnetic radiation" as required by those claims. (ECF 485-8 at 37-38).

On May 26, 2017, in response to SiOnyx's motion for rehearing, the PTAB reversed its decision to institute review of claims 9, 24, and 25 for anticipation by Nakashiba, and the IPR did not go forward as to those claims on that ground. (ECF 522-14 at 2).

The PTAB issued its final written decision on March 28, 2018. It ruled that claims 1, 2, 4-11, 13-18, 21, and 23-25 of the '591 patent were unpatentable, but upheld claims 12 and 26. (ECF 485-18 at 66). SiOnyx filed a notice of appeal on May 24, 2018. That appeal is still pending, and SiOnyx's opening brief is due September 7, 2018. (Fed. Cir. No. 18-2019). HC did not cross-appeal.

While the IPR was proceeding, discovery in this case was ongoing. Fact and expert discovery have now mostly closed, with the exception of some lingering disputes that have required the intervention of the magistrate judge or the undersigned. Defendants previously filed six motions for partial summary judgment, and plaintiffs filed one motion for partial summary judgment and a motion to amend the second amended complaint, which the Court addressed in a previous order.

Defendants have filed five more motions for partial summary judgment asserting the following contentions: (1) that SiOnyx and Harvard have failed to prove willful infringement; (2) that SiOnyx and Harvard have failed to prove indirect infringement; (3) that SiOnyx has failed to prove infringement of the '591 patent ; (4) that SiOnyx and Harvard have failed to prove that Carey and Mazur are the sole co-inventors of the HPK patents; and (5) that SiOnyx and Harvard have failed to prove any claims against Ocean Optics. Defendants have also filed a motion to strike the declarations of Ezekiel Kruglick. SiOnyx in turn has moved for summary judgment regarding inter partes review estoppel.

II. Standard of Review

The role of summary judgment is to "pierce the pleadings and to assess the proof in order to see whether there is a genuine need for trial." Mesnick v. Gen. Elec. Co. , 950 F.2d 816, 822 (1st Cir. 1991) (internal quotation mark omitted). Summary judgment is appropriate when the moving party shows that "there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law." Fed. R. Civ. P. 56(a). "Essentially, Rule 56 [ ] mandates the entry of summary judgment 'against a party who fails to make a showing sufficient to establish the existence of an element essential to that party's case, and on which that party will bear the burden of proof at trial.' " Coll v. PB Diagnostic Sys., Inc. , 50 F.3d 1115, 1121 (1st Cir. 1995) (quoting Celotex Corp. v. Catrett , 477 U.S. 317, 322, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986) ). In making that determination, the court must "view the record in the light most favorable to the nonmovant, drawing reasonable inferences in his favor." Noonan v. Staples, Inc. , 556 F.3d 20, 25 (1st Cir. 2009). When "a properly supported motion for summary judgment is made, the adverse party must set forth specific facts showing that there is a genuine issue for trial." Anderson v. Liberty Lobby, Inc. , 477 U.S. 242, 250, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986) (internal quotation marks and footnotes omitted). The non-moving party may not simply "rest upon mere allegation or denials of his pleading," but instead must "present affirmative evidence." Id. at 256-57, 106 S.Ct. 2505.

III. Analysis

A. Noninfringement of the '591 Patent

SiOnyx asserts that defendants infringe claims 3, 4, 9, 18-20, and 25 of the '591 patent. Claims 3, 4, 9, and 25 depend from claim 1, and claims 18-20 depend from claim 13. Claim 1 recites:

1. A photosensitive imager device, comprising:

a semiconductor substrate having a substantially planar surface and multiple doped regions forming at least one junction;

a textured region coupled to the semiconductor substrate on a surface opposite the substantially planar surface and positioned to interact with electromagnetic radiation;

integrated circuitry formed at the substantially planar surface; and

an electrical transfer element coupled to the semiconductor substrate and operable to transfer an electrical signal from the at least one junction.

( '591 patent, col. 18 ll. 33-45). Claim 13 recites:

13. A method of making a photosensitive imager device, comprising:

forming a textured region on a semiconductor substrate, wherein the semiconductor substrate has a substantially planar surface opposite the textured region and multiple doped regions forming a[t] least one junction, and wherein the textured region is formed in a position to interact with electromagnetic radiation;

forming integrated circuitry on the substantially planar surface; and

coupling an electrical transfer element to the semiconductor substrate such that the electrical transfer element is operable to transfer an electrical signal from the at least one junction.

( '591 patent, col. 19 ll. 16-29). Therefore, all of the asserted claims contain the limitation that the "electrical transfer element" be "operable to transfer an electrical signal from the at least one junction."

In its claim-construction opinion, the Court construed the term "electrical transfer element" to mean "an element used to transfer charge or signal from a photosensitive pixel." (ECF 251 at 36). The Court explained that the specification explicitly provides that the invention "can be incorporated into complementary metal-oxide-semiconductor (CMOS) imager architectures or charge-coupled device (CCD) imager architectures." ( '591 patent col. 6 ll. 45-49). The Court further explained that the language "operable to transfer an electrical signal" "requires that the transfer element must be able to transfer a signal, it does not exclude the ability to transfer a charge." (ECF 251 at 35).

All of the devices SiOnyx accuses of infringement are CCDs or devices that incorporate CCDs. SiOnyx does not dispute that the pixels of a CCD transfer a charge, which is converted to a voltage by an amplifier outside the pixel array (in contrast to a CMOS detector, in which the charge is converted to a voltage at each pixel). (See ECF 175-1 ¶¶ 17-21; ECF 481-5 at 56:8-13).

Defendants contend that they are entitled to summary judgment of noninfringement because the CCDs only transfer a charge, not a signal, and the claims specifically require that the "electrical transfer element" be "operable to transfer an electrical signal. " They contend that by taking the position that both "charge" and "signal" must be included in the construction of "transfer element," SiOnyx has conceded that the two are different. In support, defendants point to the testimony of SiOnyx's expert, Michael Guidash, who stated, "To the extent Defendants' proposed construction [of electrical transfer element] only applies to 'transferring an electrical signal' it would rule out CCDs, which are explicitly defined as within the scope of the invention by the specification." (ECF 175-1 ¶ 55).

Defendants' argument, however, mischaracterizes Guidash's testimony. Guidash has consistently opined that the claims of the '591 patent cover both CMOS and CCD detectors, and that it would be erroneous to read the claims to exclude CCD detectors when the specification clearly contemplates that its teachings could be applied to CCD detectors, which were previously well-known in the art. (See '591 patent, col. 1 ll. 18-22; id. , col. 6 ll. 45-49; id. , col. 11 ll. 11-13). Guidash opined that "[t]o those of ordinary skill of the art, particularly in view of the specification, the terms 'electrical transfer element' or 'transfer element' ... refers [sic] to the circuitry or electrical components used in the pixels," which, as to a CCD pixel, would be "a MOS capacitor, which transfers the charge from the pixel to a shift register and ultimately to read-out circuitry where it is converted into a voltage." (ECF 175-1 ¶ 53). The Court acknowledged that in its claim-construction opinion, in which it explicitly relied on Guidash's unrebutted opinion as to how a skilled artisan would interpret the claims. (See ECF 251 at 35).

Claim construction is an exercise in clarification. As part of that process, SiOnyx (through Guidash) requested that the Court clarify that the "electronic transfer element" includes the transfer of charges. The Court can understand how defendants now view the inclusion of two words in the claim-construction opinion-signal and charge-to suggest a distinction between the two. Nonetheless, it is clear that "charges" are a kind of "signal" encompassed by the claim. Indeed, the patent does not distinguish between "signal" and "charge"-in fact, it does not use the word "charge" except to name "charge-coupled devices (CCDs)." But it does say that its teachings are applicable to CCDs, and Guidash explained that a person of ordinary skill would view the claims as covering them, despite the patent's use of the word "signal." (See ECF 175-1 ¶ 20 (explaining that CCDs transfer "signal charge packets") ).

To be clear, this is not a back-door, after-the-fact construction of the word "signal." Neither party requested a construction of that term, and its application to the accused devices will therefore be a question of fact for the jury. See Hewlett-Packard Co. v. Mustek Sys., Inc. , 340 F.3d 1314, 1320-21 (Fed. Cir. 2003) ; Moba, B.V. v. Diamond Automation, Inc. , 325 F.3d 1306, 1312-13 (Fed. Cir. 2003). The Court only holds that SiOnyx has not conceded, and the Court's claim-construction opinion does not require, that "signal" and "charge" are mutually exclusive categories.

Defendant's motion for summary judgment of noninfringement of the '591 patent will therefore be denied.

B. Motion of Ocean Optics for Summary Judgment

Plaintiffs allege that Ocean Optics is liable for infringing the '467 and '591 patents through its Maya2000 Pro-NIR spectrometer. That theory of infringement depends entirely on the spectrometer's incorporation of an HPK sensor as a component part. Ocean Optics contends that summary judgment should be granted in its favor because (1) plaintiffs have not requested damages from Ocean Optics; (2) even if they had requested damages, the record is bereft of evidence to actually prove those damages; and (3) an injunction would never be available against Ocean Optics. It contends that because plaintiffs thus would have no remedy against it separate from their remedies against HPK and HC, it should be dismissed from the case. The motion will be denied for multiple reasons.

First, the fact that plaintiffs did not specifically identify damages as a form of relief against Ocean Optics in their complaint is inconsequential. If Ocean Optics is judged to have infringed the patents, plaintiffs are entitled to damages. Section 281 provides: "A patentee shall have remedy by civil action for infringement of his patent." 35 U.S.C. § 281. Section 283 further provides: "The several courts having jurisdiction of cases under this title may grant injunctions in accordance with the principles of equity to prevent the violation of any right secured by patent, on such terms as the court deems reasonable." Id. § 283 (emphasis added). And Section 284 provides: "Upon finding for the claimant the court shall award the claimant damages adequate to compensate for the infringement, but in no event less than a reasonable royalty for the use made of the invention by the infringer, together with interest and costs as fixed by the court." Id. § 284 (emphasis added). Taking the three statutes together, it is clear that damages must be awarded against an infringer. Dow Chem. Co. v. Mee Indus., Inc. , 341 F.3d 1370, 1381-82 (Fed. Cir. 2003) ("The statute is unequivocal that the district court must award damages in an amount no less than a reasonable royalty.").

Second, plaintiffs have put forth adequate evidence to prove any damages Ocean Optics may owe. Plaintiffs concede that they can only recover one reasonable royalty for each infringing product. See Glenayre Elecs., Inc. v. Jackson , 443 F.3d 851, 864 (Fed. Cir. 2006) ("[A] party is precluded from suing to collect damages for direct infringement by a buyer and user of a product when actual damages covering that very use have already been collected from the maker and seller of that product."); Shockley v. Arcan, Inc. , 248 F.3d 1349, 1364 (Fed. Cir. 2001) ("Each joint tort-feasor is liable for the full amount of damages (up to a full single recovery) suffered by the patentee."). But plaintiffs have the option of recovering the amount attributable to the devices manufactured by HPK and sold to Ocean Optics through HC from any or every defendant. Shockley , 248 F.3d at 1364 ("[P]arties that make and sell an infringing device are joint tort-feasors with parties that purchase an infringing device for use or resale.") (citing Birdsell v. Shaliol , 112 U.S. 485, 488-89, 5 S.Ct. 244, 28 L.Ed. 768 (1884) ). Although plaintiffs' damages expert calculated those amounts as part of the damages for HPK, the portion of those damages attributable to sales to Ocean Optics appears to be discernable from the underlying evidence, which details sales to Ocean Optics separately from other U.S. sales. (See ECF 491-2 at 28 n.113, 68 n.284; ECF 529-18 at 10:4-13:1). The fact that plaintiffs' expert did not specifically break out the figure attributable to Ocean Optics, or opine as to a separate hypothetical negotiation with it, does not preclude plaintiffs from ever collecting damages from Ocean Optics.

Furthermore, even if there were no expert testimony on damages, the court would still be obligated to evaluate what evidence there was to come up with a reasonable royalty. In Dow Chemical Co. v. Mee Industries, Inc. , the district court judge excluded all of the plaintiff's expert's testimony on reasonable-royalty damages, and then held that the plaintiff had failed to prove any damages. 341 F.3d 1370, 1381 (Fed. Cir. 2003). The Federal Circuit reversed, explaining that 35 U.S.C. § 284 required that damages be awarded, and directed the district court on remand to consider the Georgia-Pacific factors and "award such reasonable royalties as the record evidence may support." Id. at 1382 ; see Apple Inc. v. Motorola, Inc. , 757 F.3d 1286, 1327-28 (Fed. Cir. 2014) ("If a patentee's evidence fails to support its specific royalty estimate, the fact finder is still required to determine what royalty is supported by the record."), overruled on other grounds by Williamson v. Citrix Online, LLC , 792 F.3d 1339 (Fed. Cir. 2015). Therefore, the alleged inadequacy of plaintiffs' expert testimony on reasonable royalty damages, without more, is not a reason to grant summary judgment to Ocean Optics.

Finally, it is inappropriate to determine at this time whether an injunction would be available in the event of infringement. Ocean Optics has presented no example of a case where a court has found, prior to trial, that a permanent injunction would never be available as a matter of law, regardless of liability. There are genuine issues of material fact as to whether the harm is irreparable and how the balance of the hardships, at the time the injunction would enter, would be determined.

Accordingly, the motion of Ocean Optics for summary judgment will be denied.

C. Inter Partes Review Estoppel

SiOnyx currently asserts that defendants infringe claims 3, 4, 9, 18-20, and 25 of the '591 patent, and defendants appear to have limited their counterclaims of invalidity to those same claims. (See ECF 522 at 5 ("The below table summarizes Defendants' current prior art invalidity grounds presented in this case....") ). The PTAB declined to institute review of claims 3, 19, and 20, and has ruled that claims 4, 9, 18, and 25 are unpatentable. SiOnyx has appealed the unpatentability ruling to the Federal Circuit; defendants have not appealed the declination of review.

SiOnyx has now moved for summary judgment on the ground that HC, HPK, and Ocean Optics are estopped from arguing in this action that any of the asserted claims of the '591 patent are invalid based on:

• Prior art actually discussed in the PTAB's written decision, specifically

• anticipation by or obviousness over Nakashiba, and

• obviousness over Mabuchi and Mazur;

• Prior art raised in the petition as to which the PTAB did not institute review, specifically

• obviousness by Akahori and Mazur; and

• Prior art that was not raised in the petition, but could have been, specifically

• anticipation by Adkisson, and

• obviousness over the S9840 datasheets in combination with various other references.

The America Invents Act, Pub. L. No. 112-29, 125 Stat. 284 (2011), created the IPR process. As part of the statute, Congress included an estoppel provision in an attempt to avoid duplicative validity challenges before the PTAB and the district courts. Under the statute, "[t]he petitioner in an inter partes review of a claim in a patent under this chapter that results in a final written decision under section 318(a), or the real party in interest or privy of the petitioner, may not assert in a civil action arising in whole or in part under section 1338 of title 28 ... that the claim is invalid on any ground that the petitioner raised or reasonably could have raised during that inter partes review." 35 U.S.C. § 315(e)(2).

Notwithstanding the straightforward language of the statute, its application is considerably complicated by the PTAB's former practice of instituting IPRs as to only some of the challenged claims of a patent. See SAS Inst., Inc. v. Complementsoft, LLC , 842 F.3d 1223, 1224-25 (Fed. Cir. 2016) (Newman, J., dissenting from denial of the petition for rehearing en banc); SAS Inst., Inc. v. ComplementSoft, LLC , 825 F.3d 1341, 1356-59 (Fed. Cir. 2016) (Newman, J., concurring in part and dissenting in part) ("Among the consequences of the PTO's curious and unforeseen practice, partial review does not estop unreviewed claims as to either validity or invalidity, thereby adding to the litigants' burden rather than lightening it."), rev'd sub. nom. SAS Inst. Inc. v. Iancu , --- U.S. ----, 138 S.Ct. 1348, 200 L.Ed.2d 695 (2018).

In 2016, the Federal Circuit ruled that estoppel did not apply to claims or grounds on which the PTAB declined to institute review. Synopsys, Inc. v. Mentor Graphics Corp. , 814 F.3d 1309, 1316 (Fed. Cir. 2016) ("The validity of claims for which the Board did not institute inter partes review can still be litigated in district court."); Shaw Indus. Grp. v. Automated Creel Sys., Inc. , 817 F.3d 1293, 1300 (Fed. Cir. 2016) (same for grounds); see also Credit Acceptance Corp. v. Westlake Servs. , 859 F.3d 1044, 1