The appeal ruling reported on today kept people waiting for a very long time, but the wait was worthwhile for case law observers, as it contains a number of interesting aspects.
The patent at stake is EP 0570484. It was filed on February 4, 1992 and claimed a priority back to February 4, 1991. The patent proprietor, Trikon Technologies Inc., had an infringement seizure conducted in September 2006 and then sued Alcatel Vacuum Technology France (AVTF) for infringement of the patent.
In case you are wondering what the usual pun in the title of this post is about, the patent relates to high density plasma machines for depositing or etching a substrate.
Over the course of the litigation, the original companies disappeared, the patent was sold, so, in order not to get lost, I will simply refer to the parties as the plaintiff and defendant, or similar designations.
In the first instance judgment dated March 3, 2009, the patent claims asserted by the proprietor were revoked by the Tribunal de grande instance for lack of novelty.
The plaintiff appealed, and this is when the case got dormant for a very long time. Presumably both sides dragged their feet. Well, when I say the case got dormant, this is not entirely accurate, as a lot actually happened.
First, the proprietor initiated central limitation proceedings at the EPO – and later abandoned it.
Then, a national request for limitation was filed at the French patent office (INPI), and was finally granted in September 2012, after a number of objections by the patent office and the defendant acting as a third party.
In May 2014, the defendant filed a motion for canceling the order for infringement seizure. Never too late! The appeal judgment does not mention what the objection was, but it was certainly a serious one, as the plaintiff gave up and withdrew its seizure exhibits from the record. Finally, the appeal went to trial in December 2016.
As I said, there are several interesting aspects in the ruling, but I will focus on the appraisal of the limited claims by the appeal judges.
First, an unconventional objection was raised by the defendant. According to article 564 of the Code de procédure civile, “new claims” may not be filed on appeal. By “new claims” is meant legal claims, not patent claims of course. The question is which claims filed on appeal are really new and which ones are not so new. In particular, according to article 566:
Parties may also make requests explicit if they were virtually comprised in the claims and defenses submitted to the first judge, and add thereto any claims which are an accessory thereto, a consequence thereof, or a complement thereto.
In this case, claim 1 of the patent as limited comprised new features which had thus not been debated in first instance. The defendant then argued that the infringement case based on claim 1 as limited was a “new claim” according to article 564 and thus inadmissible. A very interesting argument indeed, but a long shot, as the court noted that the patent proprietor was entitled to limit its claims “at any time“, and that the limitation was retroactive, pursuant to article L. 613-24 of the Code de la propriété intellectuelle.
Therefore, this inadmissibility defense was rejected. It is true that if it had been successful, the benefits of limitation during litigation would have been considerably affected.
The second way the defendant challenged the limitation was by asserting that the subject-matter of the limited claims extended beyond the contents of the application as filed.
Here is claim 1 as limited, wherein the highlighted portion represents the main changes relative to claim 1 in the application as filed:
A system for generating a high density plasma comprising:
a plasma confinement chamber of cylindrical form;
means for injecting a gas to be ionized into the chamber;
means disposed adjacent the chamber for generating a longitudinal magnetic field in the chamber; and
means for generating high frequency energy at a frequency of 13.56 MHz and at power in the range of 100 W to 5 kW, comprising an impedance adaptation circuit;
the system being characterized in that an antenna forming means comprises:
a single loop element encompassing the cylindrical chamber, the loop element being disposed in a plane at an angle of in excess of 45° to the central axis of the chamber, and positioned in an intermediate region along the length of the chamber, said single loop element being coupled to the means for generating high frequency energy.
The court started by restating the standard to be applied, which was accurately imported from the EPO doctrine:
The content of the application as filed must be interpreted as encompassing the description, the claims and drawings, as well as any teaching, even implicit, directly and unambiguously deriving therefrom, as understood by the skilled person.
This was followed by a long and complex discussion of the teaching of the application, which I will not attempt to fully reproduce here.
In a nutshell, the value of a frequency of 13.56 MHz was mentioned in claim 4 of the application, but (1) this value was not disclosed in association with a power range of 100 W to 5 kW, and (2) claim 4 depended on claims 2 and 3 which recited further features not included in claim 1 as limited.
So one issue was whether the value of 13.56 MHz was generally applicable to the power range of 100 W to 5 kW. And another issue was whether this value could be isolated from the other features disclosed in original claims 2 and 3.
Regarding the first issue, the court considered that the application generally taught a power range of 100 W to 5 kW and a frequency range of 2 to 50 MHz. One sentence in the description read: “to effect wave coupling and establish a high plasma current density, measured in mA/cm2, the antenna loop 12 is driven at 13.56 MHz and with RF energy of the order of 2.0 kW (in the range of 100 W to 5 KW) by the RF energy source 19“. This was interpreted as meaning that the whole power range could be achieved with the preferred frequency of 13.56 MHz, and that the value of 2.0 kW was only an example. The interpretation was confirmed by looking at the examples and figures.
The defendant argued that generating helicon waves required matching a certain frequency with a certain power, but the court did not agree that the invention was only about generating such helicon waves.
Regarding the second issue, original claim 2 mentioned a magnetic field of les than 1000 gauss, a plasma density of more than 1013/cm3 and a loop element at an angle of about 90° relative to the magnetic field; and original claim 3 mentioned a first plasma current and density peak in the grange of about 50 gauss and a second plasma and density peak in the range of about 400 gauss.
The court analyzed the application in detail again and came to the conclusion that these various features were not mandatory or essential, and not inextricably linked with the frequency value of 13.56 MHz.
Thus, there was no extension of subject-matter beyond the content of the application as filed.
Why on earth original claim 4 was then exclusively dependent on claim 3, which itself depended on claim 2, and why this claim did not directly depend on the broadest claim 1, is unclear to me. Generally, this is a hint that the various technical features at stake are probably interconnected. Not in this case, apparently – but as I said, the technical discussion was quite complex.
Anyway, the patent as limited survived the added matter challenge. It also successfully passed novelty and inventive step challenges.
On the other hand, the court found that there was no evidence of infringement, and this is another interesting aspect of this case. As explained above, the evidence seized in 2006 was no longer part of the record, further to the late challenge brought by the defendant against the seizure order.
The plaintiff relied on technical manuals and technical specifications of the defendant’s various machines. The court did not even examine whether the technical information contained in these documents reproduced the subject-matter of the patent claims.
In fact, it was sufficient for the court to note that there was no clear evidence of any potentially infringing act (such as manufacturing, marketing or the like). The new assignee of the patent filed an affidavit by its vice president and CFO, but this affidavit did not convince the court:
The affidavit […] contains tables originating from Alcatel’s documentation and states that “Alcatel made and sold 129 litigious machines during the litigious time period”. This does not make it possible to identify the machines at stake. And, due to the quality of its author, it cannot represent objective evidence of the manufacturing or marketing of machines reproducing the features of the EP’484 patent. The court notes that, in view of the documents relied on by SPTS as to the number of machines allegedly manufactured and marketed, they did not undertake any effort making it possible to physically identify any one of the allegedly infringing machines.
Technical documents do not demonstrate the actual manufacturing of the machines that they describe. And even if they were indeed manufactured, it is no possible to determine the date of manufacturing nor the manufacturer.
Remarkably, the court also rejected a request for the appointment of an expert to assess the extent of infringement and the corresponding damages.
The take-away message here is that infringement seizures are a powerful tool for IP right holders in this country. But if for some reason the infringement seizure falls apart, it becomes much more difficult for them to prove their case, and courts are generally unwilling to help them complete it.
CASE REFERENCE: Cour d’appel de Paris, Pôle 5 – chambre 2, March 10, 2017, TTI Liquidating Inc. et al. v. Pfeiffer Vacuum et al., RG No. 15/01226.