Center for Advanced Study & Research on Innovation Policy


CASRIP Newsletter - Fall 1997, Volume 4, Issue 3

TI's Kilby Patent was Found to be Invalid and thus Unenforceable

On September 10, 1997, the Tokyo High Court affirmed the Tokyo District Court's judgment denying relief against infringement. The court surprised both parties in finding patent being obviously invalid and denied relief, stating that a request to enforce such obviously invalid patent is an abuse of right. Following the German tradition, Japanese courts have long established a case law to refrain from examining the patentability in infringement litigation. This rule was developed because (1) courts do not have power to issue a decision extending to third parties,1 and/or (2) courts are refrained from invading power exclusive to the Japanese Patent Office in deciding the fate of a patent, so as to maintain the separation of jurisdiction between the patent office, an administrative branch, and courts, the judicial branch.2 But, in Fujitsu Ltd.v. Texas Instruments Inc., the Tokyo High Court not only examined validity issue but also noted the patent being obviously invalid although parties' disputes concentrate on infringement issues.

The patented invention of this case related to a basic invention for semiconductor technology made by Dr. Kilby of Texas Instruments Inc. The disputed patent was granted for a divisional application based on the Japanese patent application filed February 6, 1959. Fujitsu was one of the Japanese licensees of the patents resulting from the parent and its divisional applications but resisted to renew the license when the parent patent expired. After the licensing negotiation ended, Fujitsu filed for a declaratory judgement of no infringement and TI counterclaimed an infringement and filed an action for preliminary injunction to enjoin Fujitsu from manufacturing and selling alleged infringing products.

One important point for understanding this case is that the old Japanese Patent law controls the patent prosecution of this Kilby invention because the original Japanese patent on which the divisional application was based was filed before the current law has become effective. Under the old law, an applicant could include only a single claim in one application. However, in the United States, more than 30 claims issued from the U.S. application on which the parent application was based on for the priority. To obtain a comparable scope of protection, TI needed to select one claim for the original Japanese application and file divisional applications for remaining claims. TI's patent prosecution was extensively delayed because the Japanese Patent Office refused to examine the parent and divisional applications in parallel because the JPO needed to avoid double patenting issues by determining the scope of claim in a pending application by making reference to issued claims. Of the several divisional applications filed by TI, only four have been granted patents. After more than thirty-years of delay, the disputed patent was issued.

The Tokyo District Court granted a judgment of no infringement, by narrowly interpreting the claim in light of the specification, drawings and the prosecution history.3 Although the sole claim was very broadly drafted and may literally cover all existing IC and microcomputers,4 the specification disclosed embodiments using bulk resistances to separate electronic elements formed on the semiconductor substrate. Fujitsu's products are 1 megabit DRAM using MOSFET technologies and 32-kilobit PROM using PNP and NPN bipolar technologies. Both MOSFET and NPN bipolar technologies were developed after the Kilby invention.

First, the court agreed with that a semiconductor thin plate must include only circuit elements that meet the character further recited in the claim for literally infringe the claim, although TI intentionally adopted the open claim format by using a transition of "including". The court then found that Fujitsu's products do not meet the requirements because it includes electronic circuit elements which are not formed "distance-wise" spaced apart as recited in the TI's claim. Although the court recognized that the plain meaning of " formed distance-wise" may mean to provide any physical separation, it noted that the term should be understood in the context of technology of the invention, i.e., creation of multiple electronic circuits on a single semiconductor substrate, and concluded that the term meant to form an electrical separation between such circuits. After looking at the specification and finding that only a bulk resistance was disclosed to provide an electrical separation, the court adopted Fujitsu's interpretation that electronic elements must be formed with a distance to provide electrically separation through a bulk resistance to meet the claim limitation. Because the technology of using a bulk resistance to provide an electrical separation was long replaced with more advanced technologies to increase the density of elements on a semiconductor substrate, none of Fujitsu's products was found to infringed the TI's claim. Accordingly, the Tokyo District Court issued a declaratory judgement of no infringement and denied the preliminary injunction requested by TI.

TI appealed and the Tokyo High Court again denied infringement but relied on a different reason. The court's analysis focused on a comparison between the patented claim in the suit (the disputed claim) and the claim of the parent application from which the divisional application of the patent in the suit was filed (the parent claim).

The parent claim was rejected by the Japanese Patent office for lack of inventive step. The JPO's decision of rejection was affirmed by the Tokyo High Court and has already became final. Finding that both claims are directed to the same subject matter, the court found that the patent is obviously invalid for violating the doctrine of double patenting. The court also noted that the disputed patent is obviously invalid for lack of inventive because the parent claim directed to the same subject matter was rejected. The court also noted that the patent should have been long expired under the current patent law because an improperly filed divisional application cannot enjoy the filing date of its parent application. Stating that enforcement of obviously invalid patent should not be allowed as an abuse of right, the court granted judgement for Fujitsu and dismissed TI's appeal.

First, the Tokyo High Court examined limitations that are cited only in the disputed claim. For example, the court found limitations of lead lines are not included in the parent claim but are inherently included because the lead lines are disclosed in the specification accompanied with the parent claim, and is obvious to one skilled in the art. Thus, the court concluded that the addition of the limitations does not make the disputed claim distinct from the parent. With respect to limitations, forming circuit elements having a thin domain defined by junction in various regions in the thin plate, the court noted that the limitation is simply an embodiment of the corresponding limitations in the parent claim and cannot make the disputed claim distinct from the parent claim.

After reviewing each limitations, the court found the subject matter of the disputed and parent claims is identical except for the limitations of electronic circuits (1) being laid out "in a plane" and (2) formed "distance-wise" separated apart. With respect to the limitations (2), the court found that the limitations correspond to those in the parent claim that an active circuit element must formed separated apart from a passive circuit element so as to provide necessary insulation between the active and passive circuit elements. In clarifying the meaning of "insulation," the court particularly relied upon the statement made by the patentee during the proceeding of an appeal from a rejection of the parent application by the Japanese Patent Office. It then concluded that the term "insulation" should mean to prevent circuit elements from electrically affecting each other and that the term "spaced apart" should mean to form circuit elements physically separated or distance wise spaced apart. Finding that the specification of the disputed claim also includes the same description to explain the terms "insulation" and " spaced apart", the court concluded that the limitation in the disputed claim requiring electronic circuit elements being formed "distance wise" separated apart is identical to the limitations of the parent claim on "spaced apart".

Finally, the court noted that the technical idea to laid out electronic circuits in a plane was commonly disclosed in the specifications of the original U.S. application, and the applications resulting in disputed and parent claims and thus concluded that the parent claim should read to inherently include the limitations on electronic circuit being laid out in a plane. This interpretation was also supported by the fact that the specification discloses the same multivibrator circuits as embodiments of the subject matter in the parent claim as well as the disputed claim. Because none of limitation in the disputed claims has technical significance to distinguish its subject matter from that of the parent claim, the court found that both claims are directed to the same and violates the doctrine of double patenting. The court rejected TI's argument that the Japanese Patent office reviewed the double patenting issue and affirmed the appropriateness of the divisional application having the disputed claim during the opposition procedure for the prosecution of the disputed claim, stating that the Japanese Patent Office did not review the double patenting issue in interpreting the claim limitations with respect to their technical significance and the decision did not undergo judicial review.

The court also touched upon infringement issues. It denied literal infringement, stating that the broadly drafted claim term "attached onto" should not read to cover the circuit elements of Fujitsu's products which are deposited by a technology (CVD or sputtering) which later developed by other than the inventor after the application date of the patent in the suit. Responding to TI's argument that the accused infringer's use of a technology which was not developed as of the application date of the patent cannot escape infringement, the court agreed with TI, recognizing that a patent protection may extend to an accused device which is not literally covered by the claim language if it directly uses the technical idea of the patented invention, and involves substitutions the interchangeability of which are known to one skilled in the as of the date of infringement. Nevertheless, the court denied to examine infringement by applying the known interchangeability test since the disputed patent has an inherent flaw of being obviously invalid for the reasons discussed above.

Interestingly, the Tokyo High Court's analysis of the double patenting doctrine closely parallels U.S. courts' old double patenting analysis as typically illustrated in Miller v. Eagle Manufacturing Co.5 in applying the double patenting doctrine where (1) two claims are supported by the same embodiments, identical or substantially the same drawings and specification; and (2) the difference in two claims is a mere distinction of their scope. In the United States, this practice has long modified for enabling inventors to effectively cover their incremental developments.

Under the current U.S. practice, two claims do not direct to the same invention if there is a distinctive difference in their scope.6 If one of two claims could be literally infringed without literally infringing the other, the claims do not define identically the same invention. Although the Tokyo High Court ignored the presence of limitations in the disputed claim, noting that they are simply an embodiment of the corresponding limitation of the parent claim, had the same case been brought to a U.S. court, the claims would have obviously been passed the test and found to be distinctive.

Under Japanese Patent Law, the double patenting doctrine extends only to the identical invention, but not to obvious inventions. However, the Tokyo High Court's analysis suggests that it also examined the second part test adopted by U.S. court to see if a terminal disclaimer was necessary: whether the disputed claim define merely obvious variation of the invention disclosed and claimed in the parent application.7 By doing so, the court extended the scope of the double patenting doctrine to obvious modifications. The reason for court's confusion between the identical and obvious invention is that the Japanese identity standard which is also the novelty standard includes the substantial identity test which makes it difficult to clearly distinguish novelty from nonobviousness (inventive step) issues.8 Like the U.S. inherency doctrine, under the substantial identity test, technical features which are obvious to one skilled in the art is considered to present in the claim are inherent in the claim, the court found the two claims identical, ignoring obvious limitations as having no technical significance.

Further, the court used the disclosure of the parent claims to examine the double patenting issue because the court ignored some limitations stating that the technical features of the limitations are disclosed in the specification of the parent claim and was obvious to one skilled in the art. In the United States, case law clearly forbids courts and the USPTO from using the disclosure of the inventor's earlier patent as prior art.9 Even under Japanese patent Law, courts should not use the disclosure of its earlier application as prior art since the disclosure was not published as of the filing date of the parent application and thus the provision for lack of novelty is not applicable. What the Tokyo high court should have done was to simply compare the ideas represented in two claims. As admitted by the Federal Circuit, this is often a very difficult task and confuses judges.10 It appears that this is exactly what happened in Fujitsu v. TI case.

The problem uniquely presented in this case was the inequitable result caused by the enforcement of a patent covering the technology that has long been believed to fall in the public domain. Thus, the court faced a dilemma for wanting to deny the enforcement of such patent but unable to find a reason to do so because of the flaw of the old patent system. Since the recent case development of the doctrine of equivalents in Japanese courts has made it difficult for the court to deny infringement by narrowly interpreting the claim to cover only disclosed embodiments with no equivalent scope, the court may have decided to seek a resort to the double patenting doctrine which is typically effective to prevent an issuance of patent covering an outdated technology. However, the resolution adopted by the court seems to be unfair to the patentee since the patentee has no fault in delaying the issuance of the patent and the inequitable result would not have occurred had the patent being patented long time ago.

The court's less respect to the patent office's finding of the appropriateness of divisional application is also very surprising. The validity of Japanese patents has been protected more than the presumed validity standard by preventing accused infringers from raising the invalidity as a defense against infringement charge in patent infringement litigation. By denying the patent office's authority in determining the patentability of an invention as not considering the technical significance and not being subjected to the judicial review, no Japanese patent is safe from the validity challenge if Japanese courts apply the same rule in the future.

Since the facts involved in this case are very exceptional, the court's analysis on the double patenting doctrine may have a very limited significance. A more significant impact of this case on the future Japanese case law development is that the Tokyo High court's approval on the new case law direction indicated by the Osaka High Court in Genentech.11 Although the court did not use the term "the doctrine of equivalents" or even "equivalency", the court clearly affirmed the protection beyond the literal claim scope, and even extended such protection, stating that the timing for finding the known-interchangeability is infringement. In the past, Japanese courts consistently used the application date to decide infringement both within and beyond the literal claim scope, although many cases took account of the technology developed after the application date of the patent in the suit. Accordingly, at least in theory, the Japanese doctrine of equivalents is now perfectly in line with its U.S. counterpart.12

-- Toshiko Takenaka

1Judgement of Osaka High Court, Sept. 11, 1958, Hanrei Jiho No. 162, 23. Claim exclusion or estoppel theories as broad as those under the U.S. law do not exist under Japanese Law.
2Judgement of Osaka High Court, Oct. 30, 1984, Hanrei Taimuzu No. 543, 263
3For an English summary of the Tokyo District Court decision, Okuyama, Latest Developments in Japanese IP Cases, 20 AIPPI Journal 69 (1995).
4The sole claim in the disputed patent read:

A semiconductor device for an electronics circuit, comprising:
A single semiconductor thin plate including a plurality of circuit elements, and a main surface and a back surface;
A plurality of lead lines electrically connected to those of said circuit elements which need to be connected outside said thin plate;
Said semiconductor device being characterized in that:
  1. said plurality of circuit elements formed distance-wise spaced apart from each other in various regions of said thin plate;
  2. said plurality of circuit elements each of which has at least one thin domain defined by a junction which ends at said main surface of said thin plate;
  3. an inert insulator and a plurality of conductor pieces for circuit connection attached onto said insulator are formed on said main surface, on which said thin domain is formed;
  4. a selected thin domain in said plurality of circuit elements formed distance-wise spaced apart from each other is electrically connected by means of said plurality of conductor pieces for electrical connections, so as to make electronic circuit connections necessary for providing electronic circuit among said plurality of circuit elements; and said electronic circuit is laid out essentially in a plane by means of said plurality of circuit elements and said conductor pieces for circuit connection on said Inert Insulator.

5151 U.S. 186 (1894).
6In re Vogel, 422 F.2d 438, 164 U.S.P.Q. 619 (C.C.P.A. 1970).
8Takenaka, A Unique Feature of the Japanese Novelty Standard, 9 UCLA Pacific Basin Law Journal 220 (1991).
10In re Braat, 937 F.2d 587, 19 U.S.P.Q.2d 1289 (Fed. Cir. 1991). To make this task easier, the Federal Circuit recommeded to compare the one claim with a atangible embodiments which is disclosed and which falls within the scope of the other claim.
11Genentech Inc. v. Sumitomo Seiyaku K.K., Judgment of Osaka High Court, March 29, 1996, Hanrei Jiho No. 1586, 117 (1996). An English translation and comments on the case is reported in Toshiko Takenaka, "New Policy in Interpreting Japanese Patents," CASRIP Newsletter 3 (CASRIP, University of Wash. Seattle, Spring/Summer 1996)
12Hilton Davis Chem. Co. v. Warner-Jenkinson Co., 62 F.3d 1512, 35 U.S.P.Q.2d 1641 (Fed. Cir. 1995).

Connect with us:

© Copyright 2015, All Rights Reserved University of Washington School of Law

4293 Memorial Way Northeast, Seattle, WA 98195