Non-reproducible commercial products and inventive step – T 1044/23
In decision G 1/23, the Enlarged Board of Appeal found that non-reproducible commercial products are part of the state of the art. More details can be found in an earlier blog post.
We now have a decision (T 1044/23) of the Technical Board of Appeal of the EPO applying the principles of G 1/23; the decision includes some interesting comments from the Board on the reliance on a non-reproducible commercial product as a starting point for assessing inventive step, and on assessing obviousness starting from such products.
Background
The patent in suit was directed to a polyethylene composition characterized by a number of parameters, such as density and melt flow indices.
The opponent cited two commercial products (DMDA-8904 NT7 and DMDA-8907 NT7) which, they alleged, anticipated the claimed composition.
Claim 1 of the Main Request was found to be novel by the Opposition Division - the OD held that DMDA-8904 NT7 and DMDA-8907 NT7 were not prior art at all, as the manufacture of these commercial products was not enabled (the OD decided this case 2 years before G 1/23).
On appeal, the Board applied the teaching of G 1/23 (‘reproducibility is not relevant to the question of whether a commercial product can be prior art’) and found that claim 1 of the Main Request lacked novelty over DMDA-8904 NT7 and DMDA-8907 NT7. The Board was persuaded by the opponent-appellant’s evidence that the two resins were made available to the public (by sale) before the priority date of the patent and that they satisfied the relevant features of claim 1.
Claim 1 of AR2 before the Board was held to be novel over the commercial resins (due to the difference in density), so the remaining issue was that of inventive step.
The closest prior art
The opponent argued that either of the commercial resins could be selected as the closest prior art; the patentee disagreed, essentially relying on alleged irreproducibility of the resins.
The Board said that G 1/23 did not exclude the possibility of selecting a non-reproducible product as the closest prior art, but that the specific circumstances of the case at hand should be analysed.
Referring to the evidence on file, the Board said that it was known before the priority date that both DMDA-8904 NT7 and DMDA-8907 NT7 could be used for injection moulding and had good impact stress crack resistance and processability (and so were in the technical field of the claimed invention and were said to possess the advantages of the invention). This information made the resins “immediately relevant for a skilled person working in the present technical field and wishing to provide alternative or even improved compositions”.
According to the Board, that those resins were possibly non-reproducible was irrelevant in this context. They were commercially available, and so the skilled person “could use them as such or modify these resins, within the limits of its competence, to obtain different properties”. Alternatively, the skilled person could have “analysed [their] internal composition […] and, within certain limits, attempted to reproduce some of their properties”.
The Board also said that there was “no apparent reason why the skilled person would have considered a priori (before identifying the differences and establishing the objective technical problem solved) that any further improvement necessarily required a modification of the specific synthesis conditions employed for the preparation of the commercial product, which were not known, and excluded the product as a possible starting point on that basis”
The Board also noted that a non-reproducible commercial product is not comparable to a defective or speculative disclosure - “the former is a physical, tangible product that the skilled person can use, analyse and modify, whereas the latter are non-concrete disclosures lacking the necessary technical reality”.
Importantly, while the Board acknowledged that the knowledge of the method used to obtain the commercial resins might be essential in order to achieve a resin defined in claim 1 of AR2, they considered that this was not a valid criterion for selecting the closest prior art. In the Board’s words:
“the extent to which a non-reproducible product must be modified to obtain the claimed subject-matter, and the level of knowledge about that product and its manufacture required to achieve it, are not considered relevant for determining whether the product can be regarded as the closest prior art. These aspects instead relate to later stages of the problem-solution approach, namely identifying the distinguishing features between the product and the claimed subject-matter and/or assessing whether the skilled person, starting from that product, would have been able to obtain the claimed one”.
DMDA-8904 NT7 and DMDA-8907 NT7 were thus held to be valid starting points for assessing inventive step in the present case.
The distinguishing feature was the density of the polymer composition, and the objective technical problem was formulated as the provision of an alternative.
Obviousness
The opponent-appellant necessarily argued that:
(i) it was common general knowledge that density of a polyethylene composition could be increased (e.g. by reducing the amount of comonomer units), and DMDA resin compositions were, in fact, reproducible - the prior art gave information on the type of the process and the catalyst used (‘Ziegler-Natta’, UNIPOL process), and the nature of comonomers could have been easily identified; density could then be adjusted as noted above; and
(ii) the remaining parameters characterizing the compositions defined in claim 1 of AR2 were all dependent on the molecular weight of polyethylene; the skilled person could have easily adjusted those independently of density, if necessary.
The patentee essentially argued that:
(a) DMDA-8904 NT7 and DMDA-8907 NT7 were irreproducible due to lack of crucial information about the process conditions and, in particular, the exact Ziegler-Natta catalyst used and its characteristics; reverse-engineering the commercial products would not have enabled reconstructing the specific manufacturing process; there were many process-relevant features at play that governed the mechanical and structural properties of polymers, and those features were unknown to the skilled person; and
(b) even if the skilled person had reproduced the commercial resins, further modifications would still have been required to increase the density without altering the remaining parameters of claim 1 of AR2; there was no evidence on file that such modifications were feasible.
The Board sided with the patentee. The question of reproducibility (even though irrelevant in the context of selecting the closest prior art) proved to be central to the assessment of obviousness.
Reproducibility
Even if the selection of parameters in claim 1 was arbitrary (as the OTP was the provision of an alternative - thus, no particular pointer in the prior art was required), it was still necessary to establish that the preparation of a polyethylene composition meeting those parameters would have been obvious. Would the skilled person, relying on prior art and the common general knowledge, have been able to identify the measures leading to a composition that fulfils those parameters, without undue burden?
The Board said that the exact process leading to the commercial resins was not disclosed in the prior art. While the appellant generally explained which practical steps could be followed to prepare the resins (e.g. a Ziegler-Natta catalyst in a particular set up), “it has not been shown that, by following these indications and with only routine experimentation, the skilled person would have been able to reproduce the DMDA resins or obtain materials having comparable properties”.
The Board pointed to the examples of the patent, according to which the comparative compositions were all obtained using a Ziegler-Natta catalyst and yet the parameters of the obtained products were outside the scope of claim 1 of AR2. Thus, the choice of any Ziegler-Natta catalyst would not have been sufficient to obtain the claimed composition. Similar considerations applied to other process parameters which would need to be adjusted.
In the Board’s view, it was not made credible that the skilled person could have reproduced the commercial resins (or obtained analogous material with the parameters relevant to claim 1 at issue) without undue burden.
Modifications/parametric values
Finally, the Board was not persuaded that, having adjusted the density of the 'reproduced' compositions (as put forward by the opponent) the skilled person would have no difficulty in maintaining other claimed characteristics of the obtained polymer.
In particular, the value of one of the parameters of claim 1 for the commercial resins was “relatively close to the upper limit of the range defined in claim 1”, and the opponent did not provide any concrete steps as to how to adjust this parameter.
The Board was also not persuaded by the opponent’s argument that the parameters defined in the claim (minus density) were all dependent on the molecular weight of the polymer. Rather, the results in the patent suggested that the remaining parameters were uncorrelated.
According to the Board, it was not made credible that the skilled person would have known how to modify the commercial resins to obtain a composition having the parameters as defined in the claim at issue. The skilled person could not have done so without resorting to undue burden.
Claim 1 of AR2 was thus found to be inventive over DMDA-8904 NT7 and DMDA-8907 NT7.
Conclusion
While the aspect of reproducibility may not be relevant to the prior art status of a commercial product, it does remain relevant to the later stages of the problem-solution approach.
If you have any questions about EPO practice, please do not hesitate to get in touch.
