Friday, 6 April 2012

T 2451/09 – The Dangers Of Welding

This decision – dealing with an appeal against the rejection of the opposition against the patent under consideration – is not brand new but it contains some interesting statements on sufficiency of disclosure and in particular on the burden of proof.

[2.1] The invention concerns a sealing slide fastener with opposing elastic tapes that are pressed together edgewise by teeth on the tapes that are made to mesh by a slider - basically a zipper. The teeth are formed of two halves on opposite sides of the tapes. The invention’s main thrust is towards preventing infiltration of water through holes in the tape that allow the halves to be joined […]. To this end the patent essentially proposes using materials for teeth and tape that can be “welded together chemically”, in such a way that when the teeth are injection moulded in position on the tapes, they are welded by chemical bonding” […]. Therefore, […] the teeth halves are injection moulded onto outer tape layers made of thermoplastic elastomer (TPE), with the teeth halves “welded together chemically, so that [they] adhere by chemical bonding ... onto the opposite outer layers of [the] tapes”.

[2.2] As acknowledged by the [patent proprietor] the expression “welded together chemically” is not standard usage. It does not refer to any particular recognized type of welding. The skilled person – a plastics engineer involved in the manufacture of a plastics fastener with extensive textbook knowledge of plastics – must therefore try to understand the expression from context, that is considering it in conjunction with the further terms of the claims, description and figures, given their usual meaning, and drawing on his common general knowledge (CGK) if necessary.

[2.2.1] “Welding (together)”, for example, on its own is a common term in the plastics industry where it denotes the process of “joining pieces of suitable plastics by raising the temperature at the joint so that the pieces are united by fusing or by forging or under pressure”, see e.g. Chambers Science and Technology Dictionary, 1988. However, the juxtaposition with “chemically” is unusual; it does not appear elsewhere in the cited or consulted literature. In its usual sense that term will suggest to the skilled person an interaction at chemical level during welding of the two components, and indeed claim 1 (in all versions) goes on to state that teeth and tapes “adher[e] by chemical bonding”. […]

[2.2.2] The particular type of chemical bond – whether covalent, ionic or hydrogen, primary or secondary, strong or weak – is not specified in the patent. Nor will it be apparent to the skilled person from his common knowledge whether a particular bond type is meant and which one, as all are known to play roles in polymer chemistry. All he has is the information that the bond must result in teeth and tapes welding.

[2.2.3] Giving the terms their normal meaning the skilled person, who is trying to make technical sense of the patent’s teaching, will therefore infer the meaning of this defining feature of the invention as welding together of the materials during injection moulding in a way that they interact to form chemical bonds.

[2.2.4] With the above understanding complemented with CGK in the field he must then be able to choose appropriate tooth and tape materials as well as the necessary conditions to produce the requisite chemical bond during injection-moulding, if he is to be able to successfully put the invention into practice. Though the patent does not provide any specific examples of combinations of materials, it does instruct him, see claim 1, to use TPE’s for the tape outer layer, and […] one of a list of synthetic resin groups for the teeth. This information offers an obvious starting point in determining which materials to choose.

[2.3] In the light of the CGK in the field it becomes clear that the skilled person faces a daunting task. For one, it shows that TPE’s form a large class of elastomeric materials, which vary widely in their ability to weld with selected resins […]. More importantly, it emerges from the CGK, that it is not yet fully understood how elastomers adhere, whether the mechanism involved is purely physical or chemical in nature, let alone which materials produce a chemical bond and by what means.

[2.3.1] D13/D14 is a textbook treatise on adherence of rubbers. It focuses on the role of primary (covalent or ionic) chemical bonds in adherence, discussing a variety of different models with varying degrees of primary and secondary (weak) bonds interacting in a diffusive boundary layer […]. A purely mechanical model of molecular entanglement is also briefly mentioned […]. Though the role of primary bonds is underlined, […] D14 makes clear that the various models are tentative and that research is ongoing. In fact, D14 […] states that it is not yet possible to quantify the primary bond. The Board notes that the models in D13/D14 are not specific to rubber but apply to the broader class of elastomers, to which TPE’s belong. In figure 12-6, for example, the terms “rubber” and “elastomer” appear side by side, while throughout the text the terms are interchanged frequently. The skilled person, who is trying to understand the patent’s teaching, will thus also draw on this textbook knowledge.

[2.3.2] The view in D13/D14, which stresses the role of primary bonding, conflicts with that of Osen, a paper published 5 years later and which presents an overview of TPE’s and their properties, and with which the average plastic practitioner will also be familiar. Osen considers physical interdiffusion and molecular entanglement to be the predominant mechanism in the adhesion in TPEs […] (“bonding ... is based almost entirely on diffusion ... The resulting molecular entanglement can provide high bond strength”). Where it mentions “grafting reactions”, […] it does not refer to a chemical adhesion mechanism. Instead, it explains how normally non-miscible polymers can be made compatible and thus miscible by changing the chemistry of the individual molecules. Grafting normally refers to side chains being added to a main chain of a given polymer molecule, not cross-linking between molecules as suggested by the [patent proprietor]. The basic mechanism by which such “compatibilized” polymers weld according to Osen is still that of interdiffusion or physical entanglement.

[2.3.3] D13/D14 and Osen thus offer opposing views of how elastomers, in particular TPEs might bond. Nor does the expert opinion D15 offer a more conclusive view. At best it suggests that some materials may adhere physically (by entanglement), others chemically, either by strong, primary covalent bonds or by or weak or secondary hydrogen bonds […]. None of the models is substantiated by empirical proof. The tentative formulation in point 3 (“In principle ... could adhere ... could create ... could abstract ... could be accelerated ...”; emphasis added) indeed suggests that Prof. Camino’s interpretation of chemical welding is based on conjecture rather than fact, and that he differentiates between hypothetical models, not empirically proven mechanisms.

[2.4] In view of these conflicting adhesion models, none of which has been verified, the skilled person would at the priority date not have known with any certainty from his CGK how elastomers, and in particular TPEs, adhere. He would have been unable to say whether they adhere by physical entanglement or by any of a number of chemical bonding mechanisms, or, indeed whether some materials adhere physically, while others adhere chemically, much less which ones and under which conditions. This lack of knowledge might lead him to try and determine whether and which materials bond chemically, and possibly under which conditions. Given that hitherto none of the proposed models has been proven – and indeed concrete proof regarding the role of chemical bonding has remained forthcoming during opposition and appeal procedures – , the skilled person would then need to invest a great deal of effort in merely proving that chemical bonds are responsible. Starting from the information given in the patent, he would have to do so for a large number of possible combinations of different TPE’s and listed resins. Finally, he might not even be rewarded for his efforts, as he might find that adhesion is purely physical as posited by Osen.

[2.5] In the Board’s view the effort involved goes far beyond that, say, of routine trial and error. It amounts to an extensive research programme that puts undue burden on the skilled person in his attempts to put the invention into practice, and which may not even lead to success. Following established jurisprudence, that for an invention to be sufficiently disclosed if the disclosure is reproducible without undue burden, see the Case Law of the Boards of Appeal, 6th Edition, 2010, II.A.4.2 and the case law cited therein, e.g. T 265/85, the Board can but conclude that the invention is not sufficiently clearly and completely disclosed in the sense of A 100(b) and A 83. This fundamental flaw of the patent is a result of the speculative nature of a central, defining feature of the invention. It is not remedied by any of the amendments to the claims formulated in the main, first or second auxiliary requests.

[2.6] This speculative nature of a central defining feature of the invention naturally also means that the skilled person is not able to ascertain that he has indeed successfully carried out the invention. The requirement of sufficiency necessarily implies that he must be able to do so. The [patent proprietor’s] assertions to the contrary, for example his mention of IR spectrometry as a means of determining that the weld is formed by chemical bonds, are unsubstantiated and thus without merit. This confirms the Board’s finding of insufficiency.

[2.7] The assertion that where the skilled person reads welding by chemical bonds as synonymous with strong adhesion assumes that when there is strong adhesion it is due to (primary, covalent) chemical bonds. This assumption is again based on speculation as to the nature of the bond rather than substantiated fact. Nor has the [patent proprietor] provided any evidence that the sole example he cites, Hytrel with PBT, which is known to adhere well, see D6, page 18, adheres by chemical bonding. In this conjunction the Board also notes that the [patent proprietor] has previously, in examination (see the reply of March 1st 2005, the paragraph bridging pages 4 and 5), argued that the main mechanism of chemical bonding was in fact “molecular interdiffusion”, as posited by Osen.

2.8 Equally speculative is the assertion that all TPE’s weld chemically under the right conditions so that specifying TPE would by itself be sufficient. Not only is this assertion unsubstantiated, it also conflicts with table 2 of Osen showing specific combinations of TPE´s and thermoplastics that weld well and others that do not. Even if it were so, it would again be left entirely to the skilled person to work out what those conditions, for which the patent provides no information, might be.

[2.9] Finally, though it is true that the [opponent] carries the burden of proof that an invention is insufficiently disclosed, there is no requirement that that proof take any particular form or that it conclusively demonstrate that adequate information is lacking in the patent. The latter requirement, proving that something is missing, would set a high standard indeed. The departure point for assessing sufficiency of disclosure is the legal presumption that a granted patent is sufficiently clear and complete, see in particular T 63/06 [3]. It is incumbent on an opponent to prove that that presumption is invalid. This depends on the strength of the presumption, normally determined by the amount of detail given in the patent on how to put the invention into practice. In the present case the information in the patent is scant, providing little or no practical detail regarding a central defining feature, which could have been illustrated by a single specific material combination. It therefore suffices when the [opponent] on the basis of arguments and/or documentary evidence is able to raise serious doubts with the Board regarding the feasibility of this feature for it to find the disclosure insufficient.

In the present case the [opponent’s] arguments based on D13/D14 have indeed raised such doubts, which the [patent proprietor’s] counterarguments and mention of Osen have been unable to dispel.

[3] The Board concludes that the invention is not disclosed sufficiently clearly and completely for it to be successfully carried out by the skilled person, A 100(b) and A 83. As the patent, taking into considerations the amendments proposed in all of the requests, fails to meet this fundamental requirement of the EPC, the Board must revoke the patent pursuant to A 101(3)(b).

Should you wish to download the whole decision, just click here.

The file wrapper can be found here.

NB: This decision has been commented on Le blog du droit européen des brevets some time ago (here).