Opinion
No. C 02-03942 MHP.
June 15, 2004
Robert Paul Feldman, Christopher D. Catalano, Michael Barclay, Wilson Sonsini Goodrick Rosati, Palo Alto, CA, for Plaintiff.
Vickie L. Feeman, William Sloan Coats, III, Bas de Blank, Tarek, J. Helou, Orrick Herrington Sutcliffe LLP, Menlo Park, CA, Alexander Charles Johnson, James Edward Harris, Marger Johnson McCollom, P.C., Portland, OR, for Defendant.
MEMORANDUM AND ORDER RE: DEFENDANT'S MOTIONS FOR SUMMARY JUDGMENT OF NON-INFRINGEMENT AND SUMMARY JUDGMENT OF INVALIDITY FOR OBVIOUSNESS
Plaintiff IXYS Corporation ("IXYS") filed this action against defendant Advanced Power Technology, Inc. ("APT"), alleging infringement of two U.S. patents, numbered 5,486,715 (the "'715 patent") and 5,801,419 (the "'419 patent"), that it holds on an improved design for power MOSFET devices. APT has counterclaimed for infringement of its patents, numbered 5,283,202 (the "`202 patent") and 5,262,336 (the "`336 patent"). The parties are now before the court upon APT's motions seeking summary adjudication that its accused products do not infringe IXYS's '715 and '419 patents and summary judgment that IXYS's patents are invalid because obvious. After having considered the parties' arguments and submissions, and for the reasons set forth below, the court rules as follows. BACKGROUND
The background facts are drawn from the parties' moving papers, unless otherwise noted. The technology at issue in this litigation has already been described at length in prior orders issued by this court. For a detailed description, please see this court's original claim construction order, IXYS Corp. v. Advanced Power Technology, Inc., 301 F. Supp.2d 1065 (N.D.C.A. 2004).
Plaintiff IXYS Corporation and defendant Advanced Power Technology, Inc. are both semiconductor manufacturing firms that do business in Santa Clara, California. IXYS filed suit against APT on August 15, 2002, alleging that APT was infringing two related patents detailing an improved design for "high-frequency power MOSFETs" held by IXYS. On October 1, 2002, APT counterclaimed against Ixys for infringement of a patent it held that described an improved design for producing "lifetime control" in semiconductor devices. On January 22, 2004, this court entered an order construing disputed terms in those three patents. That same day, the court also authorized APT to amend its counterclaims to add a claim that IXYS had infringed APT's '336 patent. Eleven days later, APT filed for summary judgment of invalidity with respect to Ixys's two patents; the court denied APT's motion on March 18, 2004. Now before the court are APT's motions requesting summary judgment that its accused products do not infringe IXYS's '715 and '491 patents, either literally or under the doctrine of equivalents, and that IXYS's patents are invalid for reason of obviousness.
LEGAL STANDARD
I. Summary Judgment
Summary judgment is proper when the pleadings, discovery and affidavits show that there is "no genuine issue as to any material fact and that the moving party is entitled to judgment as a matter of law." Fed.R.Civ.P. 56(c). Material facts are those which may affect the outcome of the case. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986). A dispute as to a material fact is genuine if there is sufficient evidence for a reasonable jury to return a verdict for the nonmoving party.Id. The moving party for summary judgment bears the burden of identifying those portions of the pleadings, discovery and affidavits that demonstrate the absence of a genuine issue of material fact. Celotex Corp. v. Cattrett, 477 U.S. 317, 323 (1986). On an issue for which the opposing party will have the burden of proof at trial, the moving party need only point out "that there is an absence of evidence to support the nonmoving party's case." Id.
Once the moving party meets its initial burden, the nonmoving party must go beyond the pleadings and, by its own affidavits or discovery, "set forth specific facts showing that there is a genuine issue for trial." Fed.R.Civ.P. 56(e). Mere allegations or denials do not defeat a moving party's allegations. Id.;see also Gasaway v. Northwestern Mut. Life Ins. Co., 26 F.3d 957, 960 (9th Cir. 1994). The court may not make credibility determinations, Anderson, 477 U.S. at 249, and inferences to be drawn from the facts must be viewed in the light most favorable to the party opposing the motion. Masson v. New Yorker Magazine, 501 U.S. 496, 520 (1991).
The Federal Circuit applies the same standard for summary judgment. See, e.g., Southwall Techs., Inc. v. Cardinal IG Co., 54 F.3d 1570, 1575 (Fed. Cir. 1995); Barmag Barmer Maschinenfabrik AG v. Murata Machinery, Ltd., 731 F.2d 831, 835 (Fed. Cir. 1984). The Federal Circuit has repeatedly held that summary judgment is as appropriate in patent cases as in any other type of case. See, e.g., Paragon Podiatry Lab., Inc. v. KLM Lab., Inc., 984 F.2d 1182, 1190 (Fed. Cir. 1993); Hodosh v. Block Drug Co., 786 F.2d 1136, 1141 (Fed. Cir. 1986).
II. Infringement
A. Literal Infringement
To determine if an accused product infringes a patent, the court must compare the accused product with the asserted claims of the patent. See Southwall, 54 F.3d at 1575. A product literally infringes a patent if "every limitation of the patent claim [can] be found in the accused device." Gen. Mills, Inc. v. Hunt-Wesson, Inc., 103 F.3d 978, 981 (Fed. Cir. 1997).
B. Doctrine of Equivalents
A product may be found to infringe under the doctrine of equivalents if the accused device "performs substantially the same function, in substantially the same way to obtain the same result" as the claimed invention. Graver Tank Mfg. Co. v. Linde Air Prods. Co., 339 U.S. 605, 608, 70 S.Ct. 854, 856 (1950); Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17, 35 (1997). "A device that does not literally infringe a claim may nonetheless infringe under the doctrine of equivalents if every element in the claim is literally or equivalently present in the accused device." Sage Products, Inc. v. Devon Industries, Inc., 126 F.3d 1420, 1423 (Fed. Cir. 1997). "A claim element is equivalently present in an accused device if only `insubstantial differences' distinguish the missing claim element from the corresponding aspects of the accused device." Id. The Supreme Court has emphasized that "[e]ach element contained in a patent claim is deemed material to defining the scope of the patented invention, and thus the doctrine of equivalents must be applied to individual elements of the claim, not to the invention as a whole." Warner-Jenkinson, 520 U.S. at 29.
The essential inquiry in determining equivalents is whether "the accused product or process contain[s] elements identical or equivalent to each claimed element of the patented invention."Id. at 40. One suitable method for making this determination is the "triple identity" test, which requires the court to focus on "the function served by a particular claim element, the way that element serves that function, and the result thus obtained by that element." Id. (emphasis in original). "Although the presence of equivalents is a factual matter normally reserved for the fact finder, the trial court should grant summary judgment in any case where no reasonable fact finder could find equivalence."Sage Products, 126 F.3d at 1423 (citation omitted).
Furthermore, in applying the doctrine of equivalents, the court must also determine the "range of equivalents to which the claimed invention is entitled, in light of the prosecution history, the pioneer-nonpioneer status, and the prior art."Intel Corp. v. U.S. Int'l Trade Com'n, 946 F.2d 821, 842 (Fed. Cir. 1991) (quoting D.M.I., Inc. v. Deere Co., 755 F.2d 1570, 1575 (Fed. Cir. 1985)). The purpose of examining the prior art in the context of the doctrine of equivalents is to ensure that the patent holder does not obtain a broader right to exclude the patent under the doctrine than would have been given by the patent office. Id. Thus, the prior art restricts the extent to which patent protection can be extended beyond the limitations in the claims. Id.
Prosecution history estoppel applies to amendments made to a claim to avoid the prior art, or to otherwise address special concerns such as obviousness that "arguably would have rendered the claimed subject unappealable." Warner-Jenkinson, 520 U.S. at 30-31 (citations omitted). The Federal Circuit has set forth a three-step process for determining whether a patent holder has meaningfully narrowed the scope of the patent through amendment, thus invoking prosecution history estoppel. First, a court must decide "whether an amendment filed in the Patent and Trademark Office (`PTO') has narrowed the literal scope of a claim." Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Ltd., 344 F.3d 1359, 1366 (Fed. Cir. 2003) (en banc) (citing Pioneer Magnetics, Inc. v. Micro Linear Corp., 330 F.3d 1352, 1356 (Fed. Cir. 2003)). "If the amendment was not narrowing, then prosecution history estoppel does not apply." Id. If, however, the court finds that the amendment did narrow the literal scope of the claims, it must proceed to the second step of the process and evaluate "whether the reason for that amendment was a substantial one relating to patentability." Id. "If the patentee successfully establishes that the amendment was not for a reason of patentability, then prosecution history estoppel does not apply." Id. at 1367. On the other hand, if the court concludes that the patent was amended for a substantial reason related to patentability, it must then undertake the third and final step in the process and determine "the scope of the subject matter surrendered by the narrowing amendment." Id. The patentee is presumed to have surrendered "all territory between the original claim limitation and the amended claim limitation," though she "may rebut that presumption of total surrender" as to particular equivalents. Id. (citing Festo Corp. v. Shoketsu Kinzoku Kogyo Kabushiki Co., Ltd., 535 U.S. 722, 740 (2002).
III. Obviousness 35 U.S.C. § 103(a) provides that:
A patent may not be obtained though the invention is not identically disclosed or described as set forth in section 102 of this title, if the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains.
In a motion for invalidity on grounds of obviousness, the moving party "must show prior art references which alone or combined with other references would have rendered the invention obvious to one of ordinary skill in the art at the time of the invention." Al-Site Corp. v. VSI, Int'l, Inc., 174 F.3d 1308, 1323 (Fed. Cir. 1999) (citations omitted). If more than one prior art reference is employed, "there must be some suggestion for [combining prior art references], found either in the references themselves or in the knowledge generally available to one of ordinary skill in the art." In re Jones, 958 F.2d 347, 351 (Fed. Cir. 1992). The "presumption of validity under 35 U.S.C. § 282 carries with it a presumption that the Examiner did his duty and knew what claims he was allowing." Intervet Am., Inc. v. Kee — Vet Labs., Inc., 887 F.2d 1050, 1054 (Fed. Cir. 1989). By consequence, the movant's "burden is especially difficult when the prior art was before the PTO examiner during prosecution of the application."Hewlett-Packard Co. v. Bausch Lomb Inc., 909 F.2d 1464, 1467 (Fed. Cir. 1990).
Whether a patent is invalid for obviousness is a question of fact to be determined by reference to: (1) the scope and content of the prior art; (2) differences between the prior art and the claims at issue; (3) the level of ordinary skill in the art; and, if necessary, (4) secondary evidence of non-obviousness. Graham v. John Deere Co., 383 U.S. 1, 17-18 (1966). Secondary evidence of non-obviousness can include the commercial success of the invention, long-felt but unsolved need, failure of others to solve the problem, licensing of the patented invention, professional recognition and approval, or copying of the invention. Id.; Minnesota Mining Mfg. Co. v. Johnson Johnson Orthopaedics, Inc., 976 F.2d 1559, 1573 (Fed. Cir. 1992).
An accused infringer bears the burden of proving invalidity for reason of obviousness by clear and convincing evidence. See Monarch Knitting Mach. v. Sulzer Morat GMBH, 139 F.3d 877, 881 (Fed. Cir. 1998). When "the defendant in a patent infringement case moves for summary judgment on an affirmative defense, the elements of which the defendant must prove by clear and convincing evidence, the non-moving party must simply produce enough evidence to allow a rational trier of fact to find that there is not clear and convincing evidence." Chiron Corp. v. Abbott Laboratories, 902 F. Supp. 1103, 1110 (N.D. Cal. 1995) (Patel, J.). DISCUSSION
I. Literal Infringement
IXYS has not seriously contested APT's argument that none of the accused devices possesses a second metallization layer overlying the entirety of the insulating layer, and thus that summary judgment of literal non-infringement is warranted. IXYS has presented this court with no indication that APT's accused devices meet the "overlying" limitation, regardless of which of the parties' quantifications of "overlying the entirety" this court selects. In light of this fact, the court hereby grants summary judgment that APT's products do not literally infringe IXYS's '419 and '715 patents.
II. Infringement by the Equivalents
A. Prosecution History Estoppel
Before it was issued as the '715 patent, IXYS's original patent application was first rejected by the patent examiner as indefinite under 35 U.S.C. § 112 ¶ 2. Feeman Dec., Exh. 5, at 2. In specifying his reasons for rejecting the claims, the patent examiner asked: "[A]re the first and second metallization layers, which constitute the gate bus and the source bus, respectively, electrically shorted to each other or merely formed at a second level of metallization?" Id. In response, IXYS amended what would become claim 1 of the '715 patent to read "a second metallization layer comprising a gate bus and a source bus overlying at least said insulating layer." Id., Exh. 6, at 1. The inventor explained further that "[t]he gate bus portion of the second metallization layer contacts the first metallization layer through a gate contact opening in an insulation layer."Id. at 4.
This amendment is representative of the other relevant claim amendments.
In determining whether IXYS surrendered claim scope relevant to this motion during prosecution of its patent, this court must consider each distinct limitation of the claim separately. See Ericsson, Inc. v. Harris Corp., 352 F.3d 1369, 1375 (Fed. Cir. 2003) (". . . the equivalence question relates to whether `the speech signal amplifiers . . . only supply power to the telephone set' when the receiver is off-hook. That limitation was never amended and therefore cannot be subject to the Festo presumption.") (second ellipsis in original);Pioneer Magnetics, 330 F.3d at 1356 ("Prosecution history estoppel serves to limit the doctrine of equivalents by denying equivalents to a claim limitation whose scope was narrowed during prosecution for reasons related to patentability."); ACLARA Biosciences, Inc. v. Caliper Tech. Corp., 125 F. Supp.2d 931, 402-403 (N.D.C.A. 2000) (Breyer, J.) (". . . prosecution history estoppel under Festo must be applied individually to each of the limitations, not to the `plurality of electrodes' clause as a whole.").
Before amendment, claim 1 of the '715 patent contained two separate limitations: 1) there must exist a second metallization layer, and 2) that second metallization layer must overly at least the insulating layer. Feeman Dec., Exh. 6, at 1. As amended, the claim involves three limitations:
1. a second metallization layer
2. that second metallization layer comprising a gate bus and a source bus
3. that second metallization layer overlying at least the insulating layer.
'715 patent, claim 1. The amendment thus served only to add an additional limitation — the gate bus and source bus requirement — not to modify the requirement that the second metallization layer overly at least the insulating layer. This modification maps directly onto the patent examiner's § 112 rationale for rejecting the claims, explaining that the gate bus and source bus are both formed in the second metallization and are not shorted to one another. Feeman Dec., Exh. 5, at 1. In the course of this amendment, IXYS therefore surrendered claim scope relating to the "gate bus and source bus" limitation, not the "overlying said insulating layer" limitation, which remained unchanged. Since neither party debates that the APT devices literally meet the source and gate bus limitation, prosecution history estoppel is not relevant here.
Even if the court were to find that IXYS had modified the "overlying at least said insulating layer" limitation in a fashion relevant to this motion, it still does not appear that IXYS surrendered any claim scope that would alter the outcome of this motion. See Festo, 344 F.3d 1367. The extended description of the claim amendment contained in Mr. Ogawa's letter to the examiner explains that the source and gate buses are not shorted to each other and then states that "[t]he gate bus portion of the second metallization layer contacts the first metallization layer through a gate contact opening in an insulation layer." Feeman Dec., Exh. 6, at 4. The letter is silent regarding whether or not the source bus contacts the first metallization layer. While IXYS might therefore be incapable of claiming infringement by any equivalents in which the gate bus does not contact the second metallization layer or in which the source and gate bus are shorted to one another, this amendment would not appear to preclude IXYS from claiming infringement by an equivalent — such as in APT's accused devices — in which the source bus contacts the first metallization layer.
B. "Insubstantial Differences" Analysis
For the predominant thrust of its analysis of equivalents, this court will employ the "insubstantial differences" test, which the Federal Circuit appears to favor, but will pay particular heed to the "function," "means," and "result" of each claim element, as suggested in Warner Jenkinson. See Toro Co. v. White Consolidated Industries, Inc., 266 F.3d 1367, 1370 (Fed. Cir. 2001). For the purposes of determining whether IXYS's patents are infringed under the doctrine of equivalents, the only claim limitation in question for purposes of this analysis is the "second metallization layer . . . overlying at least said insulating layer." '715 patent, claim 1. The parties appear to agree that the remainder of the claim limitations are literally fulfilled.
The court understands the "function-way-result" test as, in some sense, simply a more robust method for evaluating the same questions of "insubstantial differences" and "hypothetical claims" placed at issue in the other tests.
In conducting this analysis, the court must first ascertain the function or functions of IXYS's second metallization layer. Such an inquiry "entails an examination of the claim and the explanation of it found in the written description of the patent." Vehicular Tech. Corp. v. Titan Wheel Int'l, 141 F.3d 1084, 1090 (Fed. Cir. 1998) (citing Applied Materials, Inc. v. Advanced Semiconductor Materials Am., Inc., 98 F.3d 1563, 1574 (Fed. Cir. 1996)). The court is nevertheless mindful of the Federal Circuit's admonition that "[a]n invention claimed in purely structural terms generally resists functional limitation."Toro, 266 F.3d at 1371 (limiting the "function" of a claim limitation, for purposes of determining infringement according to the doctrine of equivalents, to that function specifically disclosed in the language of the claim itself).
It appears to this court that the conflict betweenVehicular and Toro is nearly irreconcilable. The Toro court seems to have recognized this very fact, as it describedVehicular as having merely held that an accused device must perform every "key objective" of the asserted claims of the patented invention, not that a court should search the patent specifications when determining the functions of certain claim limitations. Toro, 266 F.3d at 1371. This represents a substantial rolling back of the reach of the Vehicular opinion, which derived the functions of the claim directly and explicitly from the specifications. See Vehicular, 141 F.3d at 1090-91. This court will generally hew to the mandates expressed inVehicular, both because the very inquiry into a claim limitation's "functions" — as described by Warner-Jenkinson — seems to demand examination of the specifications which describe the operation of structural elements, and because Vehicular stands as the earlier precedent.
1. First and Second Metal in APT's Devices
The '715 patent specification discloses a number of different functions performed by the second metallization layer. In the "Summary of Invention" section, under the sub-heading that describes improved embodiments of the device, the patent explains that the second metallization layer "provides source contacts 102 and a common source bus 105." '715 patent, 5:32-33. In addition, the "second metallization layer has a gate bus connecting the first metallization layer to a common gate pad." '715 patent, 57-59. APT thus argues that a principal function of the second metal in the '715 and '419 patents is to contact the source regions of the substrate and "collect" current from those regions. In addition, APT claims that this metal "must be coextensive with the active area," implying that it must 1) reach all active source regions, and 2) overly the entirety of the substrate's active area. A second "function" of IXYS's second metallization layer, explains APT, is to conduct gate current to the gate electrodes through a gate bus that is separated from the source bus; this activity, APT"s expert believes, must take place outside of the active area. Shenai Dec. ¶¶ 11 12. Most importantly, APT goes to great length to illustrate the fact that both contacts to source regions and source and gate buses in its devices are formed through the first metallization layer, Tsang Dec. ¶ 12, and thus that APT's devices are operable with only the first metal layer intact. APT reasons that the second metal in its accused devices does not fulfill the same function as IXYS's second metal structure, and argues that this functional difference precludes IXYS from proving infringement by the equivalents.
Though apparently it might be more difficult to bond a wire to these devices.
IXYS's principal argument by way of reply is that APT has improperly imported imagined "functions" from the specifications into the claims for the purposes of the infringement analysis, in violation of the venerable patent maxim prohibiting such relations. See, e.g., Toro, 266 F.3d at 1371. However, the court need not address the question of whether these functional distinctions proposed by APT are properly part of the claims, because a reasonable fact-finder could conclude that the asserted differences are wholly insubstantial.
As an initial matter, APT's argument that the functions allegedly performed by IXYS's second metallization layer are performed in part by APT's first metallization layer and in part by APT's second metallization layer does not present a bar to IXYS's ability to prove infringement by equivalents. A jury is capable of finding infringement where one claim element of the patented invention fulfills a role satisfied by two or more structures in an accused device. See Dolly, Inc. v. Spalding Evenflo Cos., 16 F.3d 394, 398 (Fed. Cir. 1994) ("An accused device may infringe under the doctrine of equivalents even though a combination of its components performs a function performed by a single element in the patented invention."). Furthermore, even if the IXYS patent required that the source metal portions of the second metallization layer overly the active area completely, an APT device that failed to do the same would not necessarily be substantially different; a reasonable jury could conclude that such topographical differences introduce no functional distinction between APT's devices and the patents. Likewise, APT has provided no basis for believing that a device with a gate bus over the active region (such as its own) could not be found to be equivalent to a patent that required that the gate bus (or the spacing between the gate and source buses) exist over non-active areas of the device. These are paradigmatic "insubstantial differences," Sage Products, 126 F.3d at 1423, and a reasonable jury could conclude that APT's accused devices infringe under the doctrine of equivalents despite their existence.
The court is dubious as to the veracity of this asseveration. APT cites two figures found within the '715 patent, Figures 1C and 2B, for support of this position. Figure 1C is a drawing of a prior art device, not of the '715 patent invention itself. See Feeman Dec., Exh. 1, at 3. In addition, APT's expert claims that Figure 2B shows the second metal source bus and the active area as co-extensive, and attempts to illustrate this point by simply applying a bluish shading to that entire area. See Shenai Dec. ¶ 4. However, the "source bus" label in that figure is affixed not to the entire "active area," but to the cross-hatched second metallization depiction that, at least in the figure itself, covers only a portion of the active area.See Feeman Dec., Exh. 1, at 5.
However, if these were the only differences between the accused devices and IXYS's patented inventions, IXYS would be prevented from proving infringement by the equivalents as a matter of law, as APT's prior art device (the "208") also possesses first and second metallization layers that perform the functions described above. See Tsang Dec. ¶¶ 6 12-13; see generally IXYS Corp. v. Advanced Power Technology, Inc., 2004 WL 540513 (N.D.C.A. 2004) (Patel, C.J.) (Memorandum and Order Re: APT's Motion for Summary Judgment of Invalidity). Indeed, with respect to these functions, APT's prior art and APT's accused devices are entirely equivalent. See id. IXYS cannot claim infringement under the doctrine of equivalents against a device if to do so would necessarily "ensnare prior art" that employs similarly "equivalent" structures. Marquip, Inc. v. Fosber America, Inc., 198 F.3d 1363, 1367 (Fed. Cir. 2000); see also Wilson Sporting Goods Co. v. David Geoffrey Assoc., 904 F.2d 677, 684 (Fed. Cir. 1990) ("a patentee should not be able to obtain, under the doctrine of equivalents, coverage which [s]he could not lawfully have obtained from the PTO by literal claims.").
2. Improved Current Flow
By consequence, if IXYS's claims of infringement are to survive, IXYS must uncover a characteristic according to which its patented invention and the allegedly infringing devices differ only "insubstantially," but one that is also sufficiently distinct from the prior art that it would have been non-obvious, and thus patentable, in light of that prior art. See id. According to one formulation suggested by the Federal Circuit, there must exist a "hypothetical claim," similar to the actual patent claim, that would both "cover literally the accused device" and have been patentable (and of particular relevance here, non-obvious) over prior art. Id.; see also Key Mfg. Group, Inc. v. Microdot, Inc., 925 F.2d 1444, 14449 (Fed. Cir. 1991). It is thus IXYS's description of the function of its second metallization layer, and IXYS's attempts to distinguish its patented invention from APT's prior art, that raise the more significant questions here.
For the purposes of this motion, IXYS's approach to the question of "insubstantial differences" has been to measure the proportion of the insulating layer covered by the second metallization layer in APT's various devices and compare that percentage with the claim construction order's requirement that the second metallization layer completely overly the insulating layer (i.e. cover 100% of the insulating layer). See Blanchard Rep. ¶¶ 59-65. According to Dr. Blanchard, IXYS's expert, the second metals of APT's 208 and 208x devices, which are indisputably prior art, cover 1.821% and 2.472% of the insulating layers of those devices, respectively. Id. ¶ 65. Dr. Blanchard calculated the second metal coverage of APT's 526 die — which may or may not be prior art — at 46.14%. Id. By contrast, Dr. Blanchard's states that the second metal of APT's accused infringing device, the 546, covers 80.48% of that device's insulating layer. Id. Unsurprisingly, IXYS argues that the discrepancy between 80.48% and 100% or 93% is insubstantial, while the distance between 1.821% and 93% is sufficiently substantial to separate the '715 and '419 patents from the prior art.
The parties agree that the second metallization need not cover the entirety of the device, since there are portions of the device (e.g. the "vias" connecting the gate bus and first gate metal and the spacing between the source and gate bus in the second metal layer) where no insulating layer that need be overlaid exists. See Def. Mot., at 7; Pl. Opp., at 6. The parties disagree as to the exact percentage of the substrate that must be covered with second metal, but for the purposes of this motion it suffices to indicate that this number is between 93% and 100%. See Shenai Dec. ¶ 24.
This question was left open by the court in its prior adjudication of APT's motion for summary judgment of invalidity.See IXYS, 2004 WL 540513, at 9.
Dr. Blanchard apparently views the 546 die as representative of APT's other allegedly infringing dies, the 846 and 1046.
Numerical discrepancies are not, in and of themselves, a basis for determining that a device does not infringe under the doctrine of equivalents, particularly where the numerical discrepancy plays no meaningful functional role. See, e.g., Hilton Davis Chem. Co. v. Warner-Jenkinson Co., 114 F.3d 1161, 1163-64 (Fed. Cir. 1997) (finding that patent specifying pH "from approximately 6.0 to 9.0" may be infringed by a process that takes place at a pH of 5.0); Abbott Labs. v. Dey, L.P., 287 F.3d 1097, 1105-1106 (Fed. Cir. 2002). Concomitantly, however, it is not immediately clear why the coverage gap between the APT 208 and the IXYS patents is sufficiently large or significant to separate IXYS's patented invention from those earlier devices.See Market St. Cable Ry. Co. v. Rowley, 155 U.S. 621, 629 (1895) (". . . a mere carrying forward of the original thought-a change only in form, proportions, or degree, doing the same thing in the same way, by substantially the same means, with better results-is not such an invention as will sustain a patent.") (citations omitted). In accord with the "function-way-result" test, this court must look not merely to the amount of second metal coverage but to the function of that claim limitation within IXYS's patented invention.
Addressing this question, IXYS avers that the "claimed function of the second metallization layer is improved current flow;" in other words, sufficiently large second metal coverage facilitates the rapid movement of current from an outside power source, through the second metallization layer, and to the source regions and the gate first metal, enabling the device to be "switched" (turned on and off) with greater alacrity. See Pl. Opp., at 14. The patent provides some support for IXYS's contention in its description of the interaction between the gate bus and the first gate metal:
"Thus, the improved device provides a medium comprising substantially metallization for signal propagation. In particular, the signal activating the device propagates from the gate pad through the gate bus (or second metallization), and then through the first metallization layer to turn-on and turn-off the device. This embodiment reduces resistance R and capacitance C by relying upon metallization interconnects instead of polysilicon, and therefore reduces the RC factor of the prior art device. As such, the improved device provides improved switching capabilities.
'715 patent, 2:36-46.
The improved device relies upon this combination of the first and second metallization layer for facilitating device turn-on and turn-off. The signal propagates from the metal gate pad through the metal gate bus (or portion of second metallization layer), and then through the first metallization layer to turn-on or turn-off the device. Since the signal propagates substantially through metallization rather than the resistive polysilicon layer of the prior art, switching performance is improved.Id. at 6:37-46.
While these portions of the specifications do identify improved switching speed as an improvement of the invention over the prior art, they do not explicitly assign responsibility for this improvement to the augmented second metallization coverage. On the contrary, the improvement in switching speed appears to stem from the fact that the switching signal is now propagating through two layers of metallization, and not through polysilicon, rather than anything having to do with the degree of second metal coverage. Id. at 42-44 ("This embodiment reduces resistance R and capacitance C by relying upon metallization interconnects instead of polysilicon. . . ."). Indeed, the size of the second metallization layer is nowhere mentioned within the specifications. Nevertheless, the "second metallization layer . . . overlying at least said insulating layer," the claim limitation at issue here, is described as helping — in concert with other aspects of the claimed invention — to produce this improved switching speed. The specifications do not describe the precise manner in which IXYS alleges that it does so, but the Federal Circuit has not gone so far as to require such specifics. See Vehicular, 141 F.3d at 1089-91. IXYS has therefore put forth an adequate showing that "improved current flow" is a function of the claim limitation at issue here.
IXYS argues that a device with sufficient second metal to realize improved current flow would infringe its patents, whereas a device (such as APT's prior art) lacking a higher proportion of second metal coverage, and thus failing to offer improved current flow, would not be swept up within IXYS's patents' ambit. Not surprisingly, IXYS's and APT's experts disagree regarding the efficacy of higher proportions of second metal in improving current flow and switching speed, and consequently dispute whether improved current flow could even be the purpose for increasing the proportion of second metal coverage. Dr. Blanchard, IXYS's expert, opines that the "function of the second metallization layer that overlaps the insulating layer is improved current flow in the second metal bus regions," and notes that the "second metallization layer in [APT's] MOS V, VI, and VII devices . . . (1) overlies a high percentage of the insulating layer . . . and (2) acts to improve the flow of current, making high frequency operation possible." Blanchard Rep., ¶ 70 73; see also id. ¶ 58 (arguing that the large second metal coverage of APT's accused devices must have some purpose beyond functioning merely as bonding pads). Dr. Blanchard implies also that APT's prior generations of device — due to their comparative lack of second metal coverage — would fail to enable similar high frequency operation. See id. at ¶ 68 (APT's "prior art did not disclose structures where the second metallization layer overlies a high percentage of the insulating layer.").
IXYS's argument maps well onto the "function-way-result" test described in Warner Jenkinson. Essentially, IXYS has identified improved current flow, and thus improved switching speed, as the "function" of the second metallization layer in the '715 and '419 patents. IXYS claims that APT's accused devices perform this same function through the same means (a higher proportion of second metal coverage) and to the same result as IXYS's patented devices, while the prior art devices — lacking a high percentage of second metal or other equivalent means — cannot switch current as rapidly and thus do not perform this function.
In stark contrast, APT's expert, Dr. Krishna Shenai, strenuously disputes Dr. Blanchard's contention that the proportion of second metal coverage has any bearing upon current flow. Explains Dr. Shenai, "[t]he APT second metal pads contribute a vanishingly small amount of the APT devices' resistance to current flow, and thus they are vastly larger than they need to be for this purpose. Furthermore, increasing the second metal overlying the active area actually retards high frequency operation, because it increases input capacitance." Shenai Dec. ¶ 34. Dr. Shenai's opinion squares with calculations performed and reported by Dr. Dah Wen Tsang, APT's Vice President for Engineering and Research and Development. According to his declaration, Dr. Tsang performed a number of experiments in which he reduced the amount of second metal on the APT 546 (an accused device) and determined that doing so lowered the "input capacitance" of the second metallization layer, "unquestionably result[ing] in faster switching speeds for the design with less overlap of second metal on active area, not the design with more overlap as Dr. Blanchard states." Tsang Dec. ¶ 40 (first emphasis added, second emphasis in original). At one point during his deposition, however, Dr. Tsang appeared to contradict this position and indicate that a desire to increase switching speed may have catalyzed APT's interest in increasing the overlay of second metal in its later devices. See Tsang Dep., at 196 (discussing the earlier 526 device and explaining that "there is a delay from one gate pad to the rest of the device. And we have a practice of instructing our customers if they want fast switching, they can bond to both gates.").
Although they do not say so directly, the opinions and calculations of Drs. Tsang and Shenai presumably apply as much to IXYS's devices as they do to APT's.
Although the court is troubled by the fact that Dr. Blanchard's opinions regarding improved current flow through enhanced second metallization coverage do not appear to be undergirded by calculations (or, at the very least, he does not report these calculations), his expert opinion is enough to raise a triable issue of fact. The parties' experts have presented ineluctably conflicting opinions, and this court may not — upon a motion for summary judgment — make the type of credibility determinations necessary to resolve such conflicts. Anderson, 477 U.S. at 249. The court thus finds that a reasonable juror could believe that the "second metallization" claim limitation of the '715 and '419 patents performs the function of improving current flow, and that this function is similarly performed by APT's accused devices but not by the prior art. As such, a hypothetical claim such as "overlying at least 50% of the insulating layer and providing improved current flow due to the larger proportion of second metal" would conceivably be literally infringed by APT's accused devices, but would not impermissibly capture the prior art.
For the purposes of this aspect of APT's motions, neither party has posited any distinction between the claims of the '715 and '419 patents.
The question nevertheless remains whether increasing the amount of second metallization coverage in order to improve current flow — and thus such an hypothetical claim — would have been obvious in light of prior art. In addition to rendering this hypothetical claim unpatentable and preventing IXYS from proving infringement by the equivalents, see Wilson Sporting Goods, 904 F.2d at 684, a finding of obviousness would invalidate IXYS's patents. It is at this point that APT's motions regarding non-infringement and invalidity coalesce, and the court will consider them here as an ensemble.
II. Obviousness
IXYS's expert, Dr. Richard Blanchard, has opined that as of 1993, the year the IXYS patents were granted, one of ordinary skill in the field encompassed by those patents would be "an electrical engineer with a B.S.E.E. degree and two to five years' experience in the design of power semiconductor devices, or could be an electrical engineer with an M.S.E.E. degree with one to three years' experience. Blanchard Rebuttal Rep. ¶ 7.e.iii. APT has not disputed this assessment. The court will thus adopt Dr. Blanchard's estimation as the applicable level of skill here, and all subsequent references to what a prior art reference might teach or suggest — or to what has been rendered obvious by the prior art — are made with regard to what an individual with this level of skill would have understood.
A. Scope and Content of Prior Art With Respect to the Claims at Issue
APT has assembled a broad array of prior art references, including patents, marketed devices, and published articles that — when assembled in various combinations — it believes render IXYS's '419 and '715 patents obvious. In addition to debating APT's contentions on their merits, IXYS argues that APT has failed to produce these prior art references in a timely fashion, or has failed to disclose the combinations it claims invalidate its patents, all in violation of Local Patent Rule 3-3. In its analysis of this issue, the court will confine its consideration to only several of the earliest-disclosed and best-known pieces of prior art, as these prior art references are both representative of the body of art as a whole and, as the subsequent discussion will show, dispositive of APT's motion.
APT's 208 and 208x devices are undisputed prior art to the IXYS patents. Moreover, these devices literally contain every claim limitation of those patents save one: The second metallization layers of those devices do not overly the entirety of the insulating layers, as required by independent claims 1 and 23 of the '715 patent, and 1 and 11 of the '419 patent. See generally IXYS Corp. v. Advanced Power Technology, Inc., 2004 WL 540513 (N.D.C.A. 2004) (Patel, C.J.) (Memorandum and Order Re: APT's Motion for Summary Judgment of Invalidity). APT's motion for summary judgment of invalidity for obviousness thus turns on whether there exist other prior art references (and the suggestion to combine them with the 208 and 208x) that would have made it obvious for one skilled in the art to extend the coverage of this second metallization layer over the entirety of the insulating layer. See Al-Site, 174 F.3d at 1323. Importantly, APT's 208 device was not before the patent examiner who granted IXYS's patents, and thus a presumption of validity of those patents as non-obvious over the 208 does not attach. See Hewlett-Packard, 909 F.2d at 1467.
To the extent that such a finding is necessary, the court finds that the APT 208 renders obvious all claim limitations of the IXYS patents other than the "overlying the insulating layer" limitation. Graham, 383 U.S. at 17-18.
This is particularly true in light of the fact that the 208 device contains the claim limitations that were used by the inventor to distinguish the '715 and '419 patents from the prior art before the patent examiner. Nathan Zommer's patent claims were originally rejected because the examiner believed that the combination of the Korman and Jones patents "taught a polysilicon gate with an overlying metallization layer." Helou Dec., Exh. 24, at 4. Zommer argued back to the examiner that Korman and Jones could not be teaching the use of a first metallization layer "comprising aluminum," because the processes described in the Korman patent involved fabrication temperatures that would melt aluminum. Id. ("However, Korman et al. still fail to teach the use of the claimed first metallization layer comprisingaluminum. . . .") (emphasis in original). The parties do not dispute that the 208 device employs a first metallization layer comprising aluminum, and thus the presumption of accuracy on the part of the examiner is inapplicable here.
Discussion of a second and third reference that might — when combined with the 208 device — render obvious complete second metal coverage begins and ends with the familiar Korman and Jones patents, as well as the admitted prior art cited in Figures 1A through 1C of the '715 and '419 patents themselves ("the IXYS admitted prior art"). Korman discloses a power MOSFET device that employs a substantial layer of metal deposited on top of an insulating layer that itself covers the gate and channel regions of the transistor. See Helou Dec., Exh. 1, at 2 (Korman patent, U.S. Patent No. 5,119,153) (Figure 1). Though not precisely a "top" view, the Korman figure is a three-dimensional cross-section of the device that offers a depth-wise view of the device structure and displays the upper metal layer covering the entirety of the insulating layer (indeed, covering the entire device). Id. The written description of this patent even ties this metallization layer to the need for improved conductivity in a fashion no less direct than that of the '715 and '419 patents. The "Object of the Invention" section of the patent states: "Another object is to provide a minimum cell size, multi-cellular field effect power device structure with a high conductivity portion of the body region in contact with the source electrode." Korman patent, 2:38-41; id. at Figure 1 (showing this upper metallization layer providing source contacts).
IXYS raises a number of objections to the analogy between this broad metallization layer and the second metallization layer described in the claims of its patents. First, IXYS argues that this layer of metal is not truly a "second metallization layer," because the Korman invention does not utilize a "first metallization layer" deposited over the gate polysilicon — Korman teaches depositing a layer of metal silicide, rather than metal, over the gate polysilicon. As an initial matter, IXYS is not entirely correct; the Korman patent does not exclude the use of metal overlying the gate polysilicon, and thus in some embodiments the upper metallization layer may indeed be a "second metal." See id. at 2:65-67 ("Such devices preferrably include a similar metal or metal-silicide layer disposed on the surface of the gate polysilicon."). Likewise, although Korman's broad metallization layer does not specifically include a source or gate bus, the upper metallization layer plays essentially the role of a source bus (the gate bus, if it exists, is apparently located elsewhere) and bears an extremely strong structural resemblance to the IXYS patents' second metallization layer.Compare id. at Figure 1, with Helou Dec., Exh. 2, at Figure 2A (IXYS patent figure). Indeed, Korman's broad metallization even performs a function, viz., conduction of current directly to the source regions, that IXYS's second metal performs but that the APT 208 does not; in this sense, Korman's second metal is even a closer analogue to IXYS's than is APT's. See Helou Dec., Exh. 4, at Figure 1.
IXYS's most meritorious argument is its claim that the Korman patent does not actually supply the necessary feature: The upper metallization layer of the Korman patent does not necessarily overly the entirety of Korman's insulating layer, and no "top view" drawing of the patent or written description exists to prove conclusively that the metallization layer so covers the device. See generally id. Yet a party attempting to prove invalidity for obviousness need not demonstrate that one or more prior art references literally contain all of the claim elements and limitations in the patents alleged to be invalid. The moving party need only establish that the prior art "would have rendered the invention obvious to one of ordinary skill in the art at the time of invention." Al-Site, 174 F.3d at 1323. Whereas a claim of invalidity for anticipation must necessarily focus upon whether claim limitations are met literally by the prior art, an obviousness analysis requires a more holistic approach encompassing a general examination of the prior art and an analysis of the degree to which they prefigure or suggest the patent claims at issue. Compare Scripps Clinic Research Foundation v. Genetech, Inc., 927 F.2d 1565, 1576 (Fed. Cir. 1991) (holding that invalidity for anticipation requires that all of the elements and limitations of the challenged claim are found within a single prior art reference), with Al-Site, 174 F.3d at 1323. With this standard in mind, the court finds that APT has shown, by clear and convincing evidence, that Korman teaches — in the interests of improved conductivity — an extensive upper metallization layer that provides source contacts and covers the underlying insulating layer without limitation, up to and including overlying that insulating layer in its entirety, rendering obvious the "overlying the insulating layer" claim limitation. See Helou Dec., Exh. 4; Graham, 383 U.S. at 17-18.
IXYS argues at various points that a prior art reference must "necessarily" or "inherently" require the claim element that is believed to have been rendered obvious and makes much of the fact that (in its opinion) APT's prior art references do not meet this rigorous standard. Yet in attempting to describe this standard for finding a patent invalid because obvious, IXYS has erroneously cited this court to several cases discussing the standard for invalidity by anticipation, not by obviousness.Compare Trintec Indus., Inc. v. Top-U.S.A. Corp., 295 F.3d 1292, 1295 (Fed. Cir. 2002) ("Inherent anticipation requires that the missing descriptive material is `necessarily present,' not merely probably or possibly present, in the prior art.") (citations omitted); Rosco, Inc. v. Mirror Lite Co., 304 F.3d 1373, 1380 (Fed. Cir. 2002) (same), with Al-Site, 174 F.3d at 1323 (holding that the moving party "must show prior art references which alone or combined with other references would have rendered the invention obvious to one of ordinary skill in the art at the time of the invention."). A plaintiff who seeks to prove invalidity for obviousness need not meet so stringent a standard. See id. Consequently, IXYS's claim that (for instance) "[t]he figures of Korman do not clearly show that the single aluminum layer overlies the entire insulating layer," even if true, would be far from dispositive of this issue.
These teachings are further reinforced and recapitulated by the Jones patent, U.S. Patent No. 5,164,802. Jones teaches a "monolithic" MOSFET device in which a multitude of active regions are combined through the use of a uniform source metal layer that appears to overly nearly all of the upper surface of the substrate. See Helou Dec., Exh. 3, at Figure 3. In similar fashion to the Korman patent, this upper metallization layer does not necessarily include a source bus and a gate bus, although the relevant figure does identify openings for gate connections alongside the source metal and also includes parts labeled "G1 Metal" and "G2 Metal," which may constitute upper gate metallizations. See id. at Figure 3; id. at 5:31-35 (" 26 is an opening for the gate terminal metal connection to the gate of the VDMOS to contact the drain of the lateral MOSFET, and 28 is an opening for gate terminal metal to contact the gate electrode of the VDMOS.").
Much like Korman, the Jones patent does not decisively demonstrate that the upper metallization layer necessarily extends to all corners of the device or covers every square micron of the insulating layer beneath it. See id. at Figure 3. Despite this fact, the extensive coverage of the upper metallization layer in Jones indicates that no restriction on the amount of metal coverage was believed inherent to the functioning of a multi-cellular MOSFET device, nor required for such a device, nor necessarily even desirable; a device may function properly, or even preferably, with the addition of a limitless amount of second metal coverage. The court finds that APT has shown, by clear and convincing evidence, that Jones, alongside Korman, suggests an upper metallization layer that conducts current to the source and overlies an indefinitely large proportion of the device, up to and including the entirety of the insulating layer beneath it, rendering obvious the "overlying the insulating layer" claim limitation of the IXYS patents. Graham, 383 U.S. at 17-18.
Figure 3 identifies an "interlayer," which presumably functions as an insulating layer in the Jones device.
In similar fashion, the IXYS admitted prior art teaches a power MOSFET device with an extensive upper metallization layer that provides source contacts and includes a source bus and gate bus.See Helou Dec., Exhs. 1 2, Figures 1A — 1C 3:5-65. In fact, this prior art, which bears a strong resemblance to the Korman invention, lacks only the first metallization structure of IXYS's patented invention. See generally id. IXYS is again correct that this prior art device is not precisely a "double metal" device (as there is no first metallization), and Figure 1C does not demonstrate conclusively that the second metallization layer overlies the entirety of the insulating layer. Regardless, in precisely the same fashion as Korman this prior art teaches a MOSFET device with broad upper metallization coverage, including source metal that is coextensive with the active area. Id. at Figure 1C. Congruently, the court finds that APT has shown, by clear and convincing evidence, that the admitted prior art the '715 and '419 patents suggest a MOSFET with a polysilicon gate layer and an upper metallization layer that conducts current to the source and overlies an indefinitely large proportion of the device, up to and including the entirety of the insulating layer beneath it, rendering obvious the "overlying the insulating layer" and source and gate bus claim limitations of the IXYS patents. Graham, 383 U.S. at 17-18.
This prior art also may not meet the "high-frequency power MOSFET" limitation or other such restrictions, but this matter is irrelevant for present purposes as those limitations are met by the APT 208 device.
B. Combination of Korman, Jones, and Admitted Prior Art References with the APT 208
When combined, the APT 208 (which contains all of the claim limitations of the '715 and '419 patents, save a second metallization layer overlying the entirety of the insulating layer) and the Korman and Jones patents, or the APT 208 and IXYS's admitted prior art, render the '715 and '419 patents obvious to one skilled in the art. The question thus before this court is whether there exists "some teaching, suggestion, or motivation to combine the references." In re Rouffet, 149 F.3d 1350, 1355 (Fed. Cir. 1998). A court must assiduously scrutinize the knowledge and teaching available to individuals working in the field at the time of the invention in order to avoid finding obvious in hindsight what was contemporaneously novel. In re Kotzab, 217 F.3d 1365, 1369 (Fed. Cir. 2000). Here, three possibilities for the source of this combination present themselves: The suggestion to combine these prior art references may be inherent to the problem to be solved and the nature of the nature of the prior art references, may stem from the ordinary knowledge of one skilled in the art, or may arise from the additional prior art documents presented by APT.
IXYS contends that APT never suggested combining these references in this fashion in its invalidity contentions, as required under Patent Local Rule 3-3, and thus claims that APT should be precluded from relying upon this combination at this stage of the proceedings. (In fact, in its invalidity contentions APT suggested precious few of the myriad combinations of prior art references it now alleges invalidate IXYS's patents.) The court is not unsympathetic to IXYS's concerns. However, the Korman and Jones patents have been at issue from the moment IXYS's patent applications were filed (as has IXYS's admitted prior art, for obvious reasons), and the APT 208 device has been a principle subject of this litigation (for purposes of invalidity) for quite some time now; in fact, it was one of the few devices whose mask layer data APT provided to IXYS in a timely fashion. While APT's motion makes clear that it has not complied fully with the letter of the local rules, the court finds that IXYS has long been on notice of these potential combination. In the interests of privileging substance over form, the court will proceed to address these most pertinent — and most well-known — of combinations.
Hereinafter, the court will use descriptors such as "combining Korman and Jones with the APT 208 device" to indicate generally a combination of a power MOSFET that lacks a second metallization layer overlying at least the entirety of the insulating layer with a similar device that includes that crucial claim limitation. The court is certainly not insisting upon a specific suggestion to combine Korman with the APT 208.
1. Inherent Nature of the Prior Art
The suggestion to combine prior art references may derive in some cases from "the nature of the problem to be solved," or may be "implicit from the prior art as a whole." In re Kotzab, 217 F.3d at 1370. The Korman and Jones patents, the admitted prior art, and the APT 208 are directed towards the same type of invention, viz., a power MOSFET device, and thus the suggestion to combine elements of those various prior art references in new combinations is, in some sense, inherent in the inventions themselves and the nature of the problems they are intended to solve. The broad second metallization coverage of the Korman patent is linked in that patent's written description to the issue of conductivity, providing further indication that the prior art references themselves suggest the relevant combination.See Helou Dec., Exh. 4, at 2:38-41 ("Another object is to provide a minimum cell size, multi-cellular field effect power device structure with a high conductivity portion of the body region in contact with the source electrode.").
IXYS argues at length that it would be technically impossible to combine the Korman patent with another prior art device, such as the APT 208, that employs a first metal layer comprising aluminum. Korman teaches a process that involves reacting the wafer "at a temperature in the range of from 800°-850°C for a period of 40 minutes," a fabrication step that would melt a first metal aluminum layer and render the device unworkable. Id. at 8:9-10; see also Blanchard Rebuttal Rep. ¶ 30.b. Yet this hardly prevents one skilled in the art from combining the existing underlying structures of the APT 208 with the comprehensive upper layer of metallization taught in Korman (rather than the other way around); Korman's upper metallization layer, which comprises aluminum, is entirely compatible with the processes used in manufacturing the 208 device. See Helou Dec., Exh. 4, at 9:33-34 ("An aluminum metallization layer is then sputtered onto the wafer surface. . . ."); Blanchard Rebuttal Rep. ¶ 30.c. ("The only aluminum layer in Korman is deposited after the 800-900°C densification has been performed."). In addition, and crucially so, the admitted prior art of the '419 and '715 patents do not describe any such high-temperature fabrication processes. See generally Helou Dec. Exhs. 1 2, at 3:5-65. There is thus no reason to believe that a first metallization aluminum layer — such as that employed by the APT 208 — would be incompatible with these prior art references.Cf. Blanchard Rebuttal Rep. ¶ 30.b.
Nevertheless, the inherent compatibility of the Korman patent, the IXYS prior art, and the APT 208 does not, by itself, constitute clear and convincing evidence of a suggestion within the prior art to combine those references; the prior art references must do more than obliquely suggest the combination to pass this high evidentiary threshold. See In re Rouffet, 149 F.3d at 1357 ("Furthermore, rejecting patents solely by finding prior art corollaries for the claimed elements would permit an examiner to use the claimed invention itself as a blueprint for piecing together elements in the prior art to defeat the patentability of the claimed invention.").
2. General Skill in the Art
It is conceivable that the general knowledge of one skilled in the art at the time of the IXYS inventions would itself render the combination of Korman and Jones with the APT 208 obvious. "Rarely, however, will the skill in the art component operate to supply missing knowledge or prior art to reach an obviousness judgment." Al-Site, 174 F.3d at 1324 (citing W.L. Gore Assoc., Inc. v. Garlock, Inc., 721 F.2d 1540, 1553 (Fed Cir. 1983) ("To imbue one of ordinary skill in the art with knowledge of the invention in suit, when no prior art reference or references of record convey or suggest that knowledge, is to fall victim to the insidious effect of a hindsight syndrome wherein that which only the inventor taught is used against its teacher.")). APT has not offered any specific evidence, such as through expert testimony, to indicate that the level of general skill in the art was of a level that such a combination would have been obvious, and the court will not presume such facts in the absence of particularized proof.
3. Additional Documents
a. Combining the APT 208 with Korman and Jones
APT has proffered a number of prior art documents — most of them technical articles — that it believes suggest the combination of APT's prior art dual-metal devices with the Korman and Jones patents. These articles describe the design and production of many facets of dual-metal devices and highlight the fact that such technology was apparently well-known among those skilled in the art before the time that the IXYS patents were filed and issued. See, e.g., Helou Dec., Exh. 14, at 8.27 ("Semiconductor Technology Handbook") ("The metal may be used to form a low resistance interconnection layer to replace either polysilicon or aluminum; it may be deposited selectively over exposed silicon to lower the sheet resistance of source and drain regions and polysilicon gates. . . ."); id. Exh. 15, at 902 ("Evolution and Current Status of Aluminum Metallization") ("In order to exploit fully the sophisticated device structures and processing which have resulted in substantial reductions in device size, it is frequently necessary to employ at least two levels of metallization (225). . . . The presence of a second level of metallization requires careful attention to the vias and to crossover areas for the two levels."); id. Exh 16, at 1-2 (Neilson Expert Report) (discussing invention and design of dual metal devices according to varying specifications).
Despite the extensive discussion of second metal devices in these articles, none of them mentions — much less discusses in anything approaching significant detail — the amount of second metal coverage these devices could, or should, employ. To the contrary, a 1988 article describing an APT product seems to trace APT's innovation to the use of a simplified second metal structure that overlies a lower percentage of the layers below it. See Barclay Dec., Exh. 6 ("MOS FET Expands Envelope for Both Power and Speed") ("APT's proprietary self-aligned process — which allows implantation, diffusion, and etching to be done with a single mask — eliminates this source-metal overlap and reduces input capacitance, because the source metal does not overlie the gate structures."). While this language is hardly decisive at first glance, IXYS's expert has read it to teach away from the "overlying the insulating layer" limitation at issue here. See Blanchard Rebuttal Rep. ¶ 35.a.vi. ("Further, the APT design in Cole actively teaches one skilled in the art not to meet the limitation of a second metallization layer that overlies the entire insulating layer. To the contrary, the Cole article teaches that the second metallization layer should overly none of the insulating layer.") (emphasis in original). In light of these contradictory references, the court finds that APT has failed to prove the obviousness of this combination by clear and convincing evidence. See Monarch Knitting, 139 F.3d at 881.
The APT device mentioned in this article is a single metal, rather than dual metal, device. The principle of employing less metal coverage nevertheless translates by analogy from this single metal situation to the dual metal context relevant here.
b. Combining the IXYS Admitted Prior Art with the APT 208
As described in Section II.B.1., supra, a combination of the 208's (or a similar patent's) aluminum first metal layer with the remainder of the Korman invention — essentially the converse of the combination discussed at length in the text above — is infeasible; an aluminum first metallization layer is incompatible with the remainder of Korman's high-temperature fabrication steps. See Helou Dec., Exh. 4, at 8:47-50; Blanchard Rebuttal Rep. ¶ 30(b). However, the IXYS admitted prior art discloses no such high-temperature fabrication techniques and presents no indication that it would be impossible to introduce an aluminum first metallization layer into those devices. See generally '715 patent, Figures 1A — 1C 3:5-65; '419 patent, 1 — 1C 3:5-65. Indeed, this appears to be precisely what Dr. Zommer did in creating the patented invention. See, e.g., '715 patent, 5:23-27 ("The gate ( G), however, is defined by a polysilicon gate layer 93, including portions overlying the channel and field region, on and in contact with a layer of oxide 95 and a first metallization layer 97 on and in contact with the polysilicon layer.").
The prior art provides several separate suggestions for one skilled within the art to import the APT 208's aluminum first metallization layer (and related structures) into the IXYS admitted prior art and place it beneath that device's extensive second metallization. The "Semiconductor Technology Handbook," authored by Richard Blanchard, provides an explicit suggestion to use aluminum as an "interconnection layer" in integrated circuits. See Helou Dec., Exh. 14, at 8.27 ("Other metals that can be deposited by CVD for IC applications are molybdenum and aluminum, although these systems are not as well developed as is the CVD of tungsten."). Moreover, that same reference, published in 1985, describes a method of finding "the contact resistance between aluminum and either single crystal silicon or polysilicon," mirroring the IXYS patents' characterization of the invention's improvement. Compare Helou Dec., Exh. 14, at 9.10,with '715 patent, 2:14-17 ("The improved device provides an interconnection for propagating device turn-on and/or turn-off signals, comprising substantially metallization over a polysilicon gated region.").
In addition — as noted in Section II.B.3. above — a plethora of other prior art references discuss and suggest the use of dual overlapping metallization layers, a feature that was evidently well-known and well-understood by individuals skilled in the art before the filing of the IXYS patents. See Helou Dec., Exh. 14, at 9.10 ("Via Resistance When dual layer metallization is used, the resistance from one layer of metal to another through the "via" is of concern."); Exh. 15, at 902 ("Evolution and Current Status of Aluminum Metallization") ("In order to exploit fully the sophisticated device structures and processing which have resulted in substantial reductions in device size, it is frequently necessary to employ at least two levels of metallization (225). . . . The presence of a second level of metallization requires careful attention to the vias and to crossover areas for the two levels."); Exh 16, at 1-2 (Neilson Expert Report) (discussing invention and design of dual metal devices according to varying specifications); Exh. 8, Figures 4E 4G (Sasaki Patent, U.S. Patent No. 4,521,448) (teaching an invention that involves depositing a layer of aluminum gate metal, followed by an upper layer of metal patterned into "wiring strips"); Exh. 12 ("Fourth Generation Mosfets — Larger Die Size, Lower on Resistance") ("At these higher frequencies, the closed cell polysilicon gate power MOSFETs are starting to have trouble switching fast enough . . . The resistance is in the polysilicon which has a resistance of 30 to 60 ohms per square. . . . A dual layer metal structure would be required for a closed cell. . . ."). Beyond merely describing the general subject matter of the APT 208 and the '715 and '419 patents, these prior art references address directly the particular subject matter at hand — the use of an aluminum-over-polysilicon gate in a dual-metal device — with sufficient specificity to render such a combination obvious to one skilled in the art; these documents describe precisely what one of ordinary skill in the art must do to combine these inventions. In re Jones, 958 F.2d at 351. Because it would have been obvious to one skilled in the art to combine the APT 208 — which captures all elements of the '715 and '419 patents save the extensive second metallization layer — with the similar structures of the IXYS admitted prior art (which suggests complete second metal coverage but lacks an aluminum first metal layer), the court finds that APT has proven by clear and convincing evidence that the inventions described in the '715 and '419 patents would have been contemporaneously obvious to one skilled in the art. Monarch Knitting, 139 F.3d at 881. The court hereby finds '715 patent claims 1, 4, 5, 23, 26, 27, and 30, and '419 patent claims 1, 5, 6, 8, 9, 10, 11, 15, 16, 18, 19, and 20 invalid. CONCLUSION
For the reasons set forth above, the court GRANTS defendant's motion for summary judgment of obviousness and GRANTS defendant's motion for summary judgment of non-infringement.
IT IS SO ORDERED.