T 1088/16 20-02-2019

Reasons for the Decision

1. Status of documents cited in the procedure

1.1 D26, D27 (Appellant), D28, D29, D31, D32 (Respondent) - Admittance

D31 is cited in paragraph [0042] of the patent in suit in respect of the single site catalyst to be used. Insofar as this document is presented in the patent as an integral part of the disclosure, and its admittance is not contested, its content is de facto in the procedure.

D26 and D27 had been cited by the appellant with the statement of grounds of appeal in order to address the findings of the decision in respect of the question of whether the definition of the catalyst could serve as a characterising feature of the product (minutes of the oral proceedings before the opposition division, section 3, second and fourth paragraphs; decision, section 2.3 of the reasons; statement of grounds of appeal, section 5.2).

The respondent did not object to the citation of these documents but instead provided further documents in response, i.e. D28, D29, and D31 submitted with the rejoinder to the statement of grounds of appeal and D32 submitted with a subsequent letter.

The Board is satisfied that the submission of these documents by the respective parties was a direct response to a matter central to the decision under appeal and submissions in the appeal proceedings.

It has not been rendered credible by either party that there would have been a reason to submit such documents at any earlier stage of the procedure (cf. Article 12(4) RPBA).

The arguments of the appellant regarding the documents invoked by the respondent, i.e. D28, D29, D31 and D32 are directed not to demonstrating that these are prima facie not relevant, but on the contrary enter into a detailed discussion and analysis of the respective teachings to establish whether they in fact demonstrate what the respondent states (appellant's letter of 22 August 2018, pages 10-13 in respect of D28 and D29 and the entirety of the letter of 17 September 2018, regarding all of D28, D29 and D32). To the extent that a detailed analysis of these documents was undertaken it is not possible to conclude that it has been shown that these are prima facie not relevant. On the contrary appellant has by means of the detailed discussions advanced in effect admitted the documents to the procedure.

Accordingly the Board can identify no reason to exclude any of D26, D27, D28, D29, D31 or D32 from the procedure.

1.2 D30 - experimental report submitted by respondent - admittance

This report was cited to address the findings of the decision in respect to the question of whether the flexural modulus was disclosed by the cited documents (decision under appeal, section 2.4), and hence can be seen as directed to addressing a finding of the opposition division. No objection against its admittance was raised by the appellant.

Accordingly the Board can identify no reason to exclude this document.

1.3 In summary, documents D26 to D32 are admitted into the proceedings.

2. Main request

2.1 Novelty

2.1.1 D5

D5 relates to polyethylene moulding compositions which are blends of high density polyethylene (HDPE) and linear low density polyethylene (LLDPE), which according to claim 1 have:

- a density of 0.930 to 0.940 g/cm**(3)

- a MFI(MFR) 190/2.16 of 0.05 to 1.0 g/10 min

- having 50-80 wt.% of HDPE having a density of 0.940-960 g/cm**(3), a MFI 190/2.16 of 0.01 to 0.5 g/10 min, a broad bimodal molecular weight distribution, wherein the HDPE is produced by a two-stage polymerisation process, the product of the first stage having inter alia a density of 0.910 to 0.925 g/cm**(3 ) and MFI 190/2.16 of 0.5 to 2.0 g/10 min.

According to claim 4 the first - HDPE - component can be a copolymer.

In Table 3 (columns 5 and 6) Examples 1-3 are reported, indicating among the properties given Young's modulus of the compositions.

In the examples of D5 a ZN catalyst system is used for preparation of the HDPE component (column 1, lines 66-67).

Regarding examples 1-3, which were invoked by the respondent, reference can be made to Table 3 of D5:

FORMULA/TABLE/GRAPHIC

These examples are based on a bimodal HDPE, which was prepared by a ZN process (column 1, line 62 - column 3, line 3, column 4 lines 9-69) and two different LLDPEs.

In example 1 a unimodal LLDPE but-1-ene/4-methylpent-1-ene/ethylene terpolymer was employed whilst in examples 2 and 3 the LLDPE was an oct-1-ene/ethylene copolymer having a unimodal molecular weight distribution (MWD) as well as further additives.

Thus the LLDPE of the indicated examples of D5 reflects fraction A of operative claim 1 and the HDPE reflects fraction B.

On the basis of the reported properties - see Table above - it is established that, as required by the operative claim, the copolymer fraction is of lower molecular weight than the homopolymer fraction (higher MFI at two different loadings), and that the density of the total composition and the MFR thereof are within the ranges required.

Two features of the operative claims are not explicitly disclosed in the examples of D5:

- insofar as the preparation of the polymers was disclosed, this was on the basis of ZN catalysis, not by means of a SSC.

- the flexural modulus is not reported, instead Young's modulus is given.

(a) Regarding the product-by-process feature "in which a single-site catalyst (SSC) is used in the polymerisation of at least one of the fractions (A) and (B)" it was the position of the appellant that this provided a distinguishing feature.

For the following reasons the Board does not share this view.

According to D26, introduction, second paragraph, the first paragraph in the section "Polymer Characterization" and the second paragraph on the third page of the document, metallocene (SSC) prepared LLDPEs exhibit narrower MWD and more uniform comonomer distribution than "conventional" LLDPEs (i.e. those prepared with ZN catalysts). In the final paragraph of the third page it is stated that metallocene prepared LLDPE has improved toughness.

A similar teaching can be derived from D27, page 173 second paragraph in section 1.1 which states that LLDPE prepared with ZN catalysts exhibit considerable heterogeneity in their microstructure and melting behaviour, and that the resulting polymers are considered to be a mixture of fractions of polyethylene copolymers with a range of molecular weights and short chain branch content. This is contrasted with SSC prepared polymers which have a narrow molecular weight distribution and more homogeneous distribution of short chain branches.

Thus according to what may be derived from D26 and D27 the two different catalysts result in - detectably - different polymers.

However according to D28, page 3 lines 1-5 the outcome of polymerisation with SSC depends on whether they are used in supported or unsupported form. Thus supported SSC lead to LLDPE having broader Composition Distribution Breadth Index (CDBI) and MWD than unsupported SSC. This is confirmed in the final partial paragraph of page 4, where it is also stated that the CDBI of SSC prepared resins is generally (emphasis of the Board) narrower than that of copolymers prepared using ZN catalysts. According to D28, page 8, 3rd complete paragraph, the LLDPEs have CDBI generally in the range of 50-90%, usually 55-85%. According to page 12, second complete paragraph and Table V SSC prepared polymers have narrower MWD and at equivalent melt index, lower Mw, Mz and Mz+1 than ZN prepared catalysts.

However according to D29, although these trends are generally confirmed - column 20, lines 20-24 - the purpose of the document is to prepare polymers with broadened MWD and CDBI despite using SSC (column 1, lines 14-23, statements of problem in columns 2 and 3 and the summary of the invention, first three paragraphs). This document demonstrates that it is possible to adjust SSC systems to mimic the properties arising from ZN catalyst.

Although operative claim 1 does not provide any definition of the catalyst, the patent contains in paragraphs [0042] and [0079] a reference to D31 for the purposes of defining the SSCs to be used. Claim 1 of this document provides only a very general, broad definition of the catalyst. Indeed this broad definition encompasses the catalysts of D29, which as noted above, can mimic the results obtained with ZN catalysts.

Hence even assuming - in favour of the appellant - that the term "SSC" would be understood as denoting one of those disclosed in D31 this still would not impose any restriction in terms of the properties of the resulting polymers which would necessarily and reliably establish a distinction over polymers prepared under action of another type of catalyst, in particular ZN.

D32 is directed to an investigation of different in situ supported SSCs. Figure 3 demonstrates that, depending on the precise nature of the catalyst, widely differing distributions of comonomer throughout the molecular weight spectrum of the polymer result, as shown by the Crystaf profiles. This diagram shows that either a rather narrow distribution can be obtained, or at the other extreme a distribution with three peaks. Similarly Figure 2 and Table 1 of D32 provide evidence for wide variations in the MWD and a significant difference in the length of the low molecular weight tails depending on the catalyst employed.

The conclusions that are to be drawn from the documents provided by the parties (D26, D27, D28, D29, D31 and D32) is that whilst in a direct side by side comparison it may well be possible to ascertain which of two otherwise identical polymers had been prepared by a particular type of catalyst system, the variations that are possible with SSC are so broad, extending even to mimicking the outcome obtained with a ZN catalyst that it has not been credibly shown that on the basis of an isolated polymer with no reference material it would be possible unambiguously to determine by which means the polymer had been prepared. Hence it cannot be concluded, that the feature "obtainable by" a SSC is suitable to characterises the polymer.

Accordingly this "product-by-process" feature cannot serve as a distinguishing feature compared to the prior art.

(b) Flexural modulus

In the context of D5 the question to be addressed is whether, based on the disclosure of Young's modulus, the flexural modulus can be derived.

As an aside, the Board notes that, in contrast to the "product-by process" feature it has not been disputed that the flexural modulus is suitable and appropriate to characterise the subject-matter claimed.

The respondent has provided an analysis, based on the data plotted in D30, with which the correlations between the flexural and Young's moduli and density are shown. These data are derived from two of the documents cited in the procedure - D3 and experimental report D11, as well as the patent and a study conducted by the respondent. It is argued that, given certain information about the polymer, e.g. one modulus, it is possible to determine whether the flexural modulus will fall within the range claimed.

The provided data show plots of the relationships between flexural modulus and density; Young's (E) Modulus and density; and flexural and Young's moduli. Although in all cases there is a broadly linear relationship, this is subject to variability, as indicated by the degree of scattering in the plots.

These data demonstrate at most that given one of the moduli it is possible - within a broad approximation - to estimate the other.

It is recalled that the Case Law of the Boards of Appeal applies a very strict standard of proof in respect of novelty namely "beyond all reasonable doubt" (T 793/93 of 27 September 1995, see the catchword specific to the case of what is or is not the inevitable outcome of an express literal disclosure which clearly applies here). To the extent that the correlations adduced by the respondent are demonstrably subject to variability and uncertainty (scattering), it cannot be concluded that the arguments based thereon can serve to establish - to the required standard - the corresponding properties of the compositions of the cited documents.

Accordingly it has not been shown that the examples of D5 disclose compositions having the required flexural modulus.

Consequently the subject-matter of claim 1 is not anticipated by the disclosure of D5.

2.1.2 D7

The sole example of D7 does not formally disclose the required density (a value of 940 kg/m**(3) is disclosed while a value below 940 kg/m**(3) is required in claim 1), and the flexural modulus value has a reported value that is above the range required (955 MPa vs a value from 400 to 820 MPa in claim 1) which moreover has been determined by a different method to that set out in the patent. With D11 and D11a the respondent has stated that it has provided replications of the teaching of the example of D7. However the Board observes that many details of the examples, e.g. catalysts used, reaction conditions etc. are not disclosed in D7, meaning that various assumptions have to be made. The respondent has submitted that it has provided a cluster of examples within the ambit of what is explicitly disclosed in D7 and that the conclusion to be reached from these is that the subject-matter claimed is anticipated.

However, since there are a number of details lacking from D7 it is inherently impossible to replicate the teaching thereof and accordingly any attempt to do so is condemned to failure in the light of the high standards set by the Case Law of "beyond all reasonable doubt" (i.e. the inevitable outcome of the disclosure) as noted above with respect to decision T 793/93.

2.1.3 The subject-matter of claim 1 of the main request is therefore novel.

2.2 Inventive step

As explained in its first paragraph, the patent is directed to the provision of a polyethylene composition and the use thereof in the production of pipes and to pipes, particularly pressure pipes, made from the composition. The pipes are used for transport of liquid and gaseous materials (paragraph [0002]). Specifically the problem addressed is to provide a polyethylene composition for the manufacture of pressure pipes which composition has both good processability, is sufficiently flexible for easy handling and complies with pressure class PE63 (paragraph [0009]).

2.2.1 Closest prior art

D5 is directed to the provision of a polyethylene composition suitable for the production of gas pipes, landfill belts and sheets (column 1 lines 6-9). The resulting pipes are stated to have excellent characteristics (paragraph bridging columns 3 and 4) in terms of processing and mechanical properties (Young's modulus, impact strength, long term stress rupture and long service life).

According to claim 1 this is achieved by a composition having:

- a density of 0.930-0.940g/cm**(3)

- a MFI (MFR)190/2.16 of 0.05 to 1.0 g/10 minutes

- 50-80 wt% HDPE of MFI 190/2.16 of 0.01 to 0.5 g/10 min

- 20-50 wt% of linear polyethylene of low density and MFI 190/2.16 of 0.5 to 2.0g/10 min.

Regarding the argument of the appellant as to the non-relevance of D5 since the application envisioned therein does not require flexibility (see section XI.(c), above) the Board observes that D5 aims at broadly the same field of application as the patent meaning that fundamentally different pipe properties are not inherent to the two sets of end uses. In particular it is noted that both the patent (paragraph [0002]) and D5 (column 1, line 8) envisage gas pipes. Accordingly there are no grounds for disregarding D5 in the assessment of inventive step.

D7 is directed to the provision of a multimodal polyethylene to prepare shaped articles, in particular pipes for water or gas distribution (page 1, top to page 2, line 6). The problem addressed is to provide polyethylene materials which provide a combination of advantageous thermal, mechanical and processing properties, with the emphasis on high temperature resistance, high stress resistance, good mechanical properties (tensile, impact) and excellent processability (page 2, lines 18-24).

According to claim 1 of D7 this problem is solved by a multimodal polyethylene having:

- a density of 0.925 to 0.950 g/cm**(3)

- a Melt Index (MFR) 190°C/2.16 Kg of 0.05 to 5 g/10 mins

- HMW ethylene interpolymer

- LMW ethylene polymer.

Both of D5 and D7 address therefore the same problem as the patent in suit and hence are equally suitable to serve as closest prior art.

2.2.2 Analysis starting from D5

(a) Distinguishing feature

As noted above with respect to novelty (section 2.1.1) the subject-matter of operative claim 1 differs from the examples in D5 in that it has not been established that the flexural modulus necessarily falls within the range specified in the claim.

(b) Technical effect

The patent contains an example and a comparative example, which are presented in Table 1 of the patent:

FORMULA/TABLE/GRAPHIC

FORMULA/TABLE/GRAPHIC

The following is noted. In the comparative example the LMW fraction is a homopolymer, whereas the claim requires a copolymer. Moreover, the example and comparative example further differ from each other in a plethora of other features. For example in the copolymer component (prepared in the gas phase reactor) different monomers are used, the splits are different as are the conditions in the second reactor.

Accordingly the example and comparative example are not suitable to show any technical effect arising from a specific feature, nor in particular an effect with respect to the compositions of D5.

(c) Objective technical problem - its solution

On this basis the only technical problem which can be formulated is the provision of further polyethylene compositions with satisfactory mechanical properties.

This problem was solved by undertaking - in the light of the available evidence - arbitrary modifications to the compositions explicitly known from D5.

(d) Obviousness

Such arbitrary modifications are within the ordinary and routine activities of the skilled person when seeking to solve the problem of providing a composition which is merely in some unspecified manner different from those which are known. Further it has not been made credible that there were any particular technical hurdles or obstacles to undertaking the so defined modifications. In this respect while the information available (in particular the data in D30) were not sufficient to conclude that the examples in D5 necessarily had a flexural modulus within the range of claim 1, they equally show that values within the range are normal when working according to the teaching of D5.

In conclusion the subject-matter of claim 1 is obvious in the light of the teachings of D5.

2.2.3 Analysis starting from D7 as closest prior art

The analysis and conclusion are similar.

As noted, the subject-matter of operative claim 1 differs from that of the examples of D7 - marginally - by the density (940 kg/m**(3 )vs below 940 kg/m**(3)) and the absence of a disclosure of the flexural modulus.

However due to the absence of any evidence the formulation of the problem remains the same as for the situation with respect to D5, and the solution - arbitrary modification - is similarly obvious.

In this respect D7 itself discloses a density range between 925 and 950 kg/m**(3) and the same considerations as above are valid for the flexural modulus.

2.3 In the light of these negative findings with respect to inventive step it is not necessary for the Board to address any other objection.

3. Auxiliary requests 1 and 1a

As noted above, claim 1 of these requests differs from claim 1 of the main request by

- the density of fraction A was specified - 925 to 945 kg/m**(3)

- MFR2 of the base resin was 0.10 to 0.50g/10 min

- shear thinning index SHI2.7/210 was specified as below 20

- flexural modulus was 450 to 650 MPa.

However as for the main request, there is no evidence for any technical effect arising from any of these (further) features which also appear to be within the general disclosure of D5 and D7.

Regarding the specification of the SHI, it has not been shown that this imposes any technical difference on the compositions compared to those of the prior art since this feature is not reported in the documents and apparently has been used only by the patent proprietor.

Although in D11 and D11a - the (attempted) replication of D7 - values for SHI are reported, due to the uncertainty concerning the disclosure of D7 (see discussion of novelty in section 2.1.2, above), no weight can be attached to these data.

In any case the absence of any specification of SHI in the prior art does not serve as evidence that the definition of an additional (unusual) parameter imbues the claimed subject-matter with a different technical character, and the contrary has not been demonstrated.

Consequently the amendments made to auxiliary requests 1 and 1a do not give cause to reach a different conclusion with respect to inventive step to that for the main request.

4. Auxiliary requests 2 and 2a

As observed above, claim 1 of these differs from claim 1 of auxiliary request 1 in that:

- the density of fraction A was defined as 925 to 940 kg/m3

- it was required that fractions (A) and (B) were polymerised using the same single-site catalyst

- the density of the base resin was from 920 to below 939 kg/m**(3)

- the SHI was below 10.

As for the previous two requests considered, there is no evidence for any technical effect associated with these (further) restrictions which remain within the general disclosure of D5 and D7. As to the specification of the catalyst, no further difference can be acknowledged in line with the above analysis developed for novelty (see section 2.1.1(a)). Accordingly there are no grounds for reaching a different conclusion with respect to inventive step.

5. Auxiliary requests 3 and 3a

The following further restriction compared to auxiliary request 2 was made:

- The MFR2 (190°C/2.16 kg) of fraction A was specified to be between 50 to 140 g/10 min.

Similarly to the situation with the previous requests there are no data to demonstrate the existence of a technical effect associated with this further difference, which remains within the general disclosure of D5 and D7, leading to the same conclusion of lack of inventive step.