T 0613/13 () of 18.12.2014

Summary of Facts and Submissions

I. The appeal by the patent proprietor lies against the decision of the opposition division posted on 6 November 2012 revoking European patent

No. EP 1 883 677, based on application

No. 06 755 047.5.

II. The granted patent comprised the following claim:

"1. A polypropylene composition comprising (per cent by weight):

a) 65-77% of a crystalline propylene polymer having an amount of isotactic pentads (mmmm), measured by **(13)C-MNR on the fraction insoluble in xylene at 25° C, higher than 97.5 molar % and a polydispersity index ranging from 5 to 10;

b) 8 to less than 13% of an elastomeric copolymer of ethylene and propylene, the copolymer having an amount of recurring units deriving from ethylene ranging from 30 to 70%, and being partially soluble in xylene at ambient temperature; the polymer fraction soluble in xylene at ambient temperature having an intrinsic viscosity value ranging from 2 to 4 dl/g; and

c) 10-23% of polyethylene having an intrinsic viscosity value ranging from 1.5 to 4 dl/g and optionally containing recurring units deriving from propylene in amounts lower than 10%,

the said composition exhibiting the following properties:

- a flexural modulus value higher than 1300 MPa according to ISO method 178,

- stress-whitening resistance values corresponding to a diameter of the whitened area of at most 1.7 cm caused by a ram falling from a 76 cm height and a diameter of the whitened area of at most 1.2 cm caused by a ram falling from a 20 cm height, and

- a value of Izod impact resistance at 23° C more than 14 kJ/m**(2) and the one at -20° C at least 5 kJ/m**(2) according to ISO method 180/1A. " |

III. Notice of opposition to the patent was filed requesting revoca­tion of the patent in its entirety on the grounds of Art. 100 (a) EPC (lack of an inventive step) and Art. 100 (b) EPC.

IV. During the opposition procedure the following documents were inter alia cited:

D9: Viscosimetry of polymers and polyelectrolytes, W.M. Kulicke and C. Clasen, 2004, page 98

D10: Journal of Molecular Liquids, 112, 2004, pages 161-169

D12: Grounds of opposition of EP 1 456 294 dated 17 February 2012 (pages 1-12)

D20: Comparison of calculated and measured intrinsic viscosity of polyolefin compositions A to G fractionated by the TREF technique

V. The decision under appeal was based on the patent as granted as the main request and on one auxiliary request. Claim 1 of said auxiliary request differed from granted claim 1 in that it was specified in each of features b) and c) that the intrinsic viscosity was "measured in tetrahydronaphthalene at 135 °C".

The opposition division held, inter alia, that the patent in suit did not satisfy the requirements of Art. 100(b) EPC because the intrinsic viscosity of component c) specified in granted claim 1 could not be established. Therefore, neither the patent in suit nor the auxiliary request satisfied the requirements of Art. 83 EPC.

Besides, whereas document D20 was admitted to the proceedings, D12 was not (see last paragraph on page 3 of the reasons).

VI. The patent proprietor (appellant) lodged an appeal against the above decision. In its statement of grounds of appeal the appellant requested that the decision of the opposition division be set aside and the patent be maintained as granted (main request) or, in the alternative, that it be maintained in amended form according to the auxiliary request filed therewith. Claim 1 of said auxiliary request corresponded to claim 1 of the auxiliary request on which the contested decision is based.

The following document was further simultaneously filed:

D12a: Grounds of opposition of EP 1 456 294 (pages 1 and 14)

VII. With letter of 14 October 2013 the opponent (respondent) requested the dismissal of the appeal and filed a further document, which is not relevant for the present decision.

Additional arguments were submitted with letter of 18 November 2014. The following documents were, together with additional documents, further simultaneously filed:

D24: granted patent EP 1 828 302 of appeal case T 868/13

D24a: appellant's statement of grounds of appeal dated 11 April 2013 in respect of T 868/13 (three pages)

VIII. In a communication issued by the Board on 22 October 2014 accompanying the summons to oral proceedings, issues to be discussed at the oral proceedings to be held on 18 December 2014 were specified.

IX. The appellant's arguments relevant for the present decision may be summarised as follows:

Main request

a) Granted claim 1 was directed to compositions obtained by sequential polymerisation. Such compositions had, as a consequence of said process, a specific morphology, which was essential in order to obtain the specific combination of properties specified in granted claim 1. During the oral proceedings it was further indicated that the combination of properties specified in granted claim 1 could only be met if features a), b) and c) were satisfied.

b) It was well known in the art that the intrinsic viscosity of a polymer blend was equal to the weight average of the intrinsic viscosity of each component following the Philippoff rule, as indicated e.g. in D9 and D10. The respondent himself commonly used it, as shown in section 5.12 of D12a. It was further demonstrated in D20 that the Philippoff rule validly applied for compositions according to the granted claims. That method of determination being so usual in the art, there was no need that the patent contained any indication that such a calculation was made.

In the examples of the patent in suit, the intrinsic viscosity of the fraction soluble in xylene of component b) could be calculated by the Philippoff rule by measuring the intrinsic viscosity of the fraction soluble in xylene of component a) and the intrinsic viscosity of the fraction soluble in xylene of components a)+b). In the same manner it was possible to calculate the intrinsic viscosity of component c) e.g. from the measurement of the intrinsic viscosity of components a)+b) and the intrinsic viscosity of the whole composition.

c) Although the compositions exemplified in D20 were not exactly according to granted claim 1, they were very similar. Therefore, D20 showed that the Philippoff rule could validly be used for such ternary systems. The fact that no precise information was indicated in D20 regarding the experimental conditions used for fractionating the polymer blend by TREF (temperature rising elution fractionation) was not an issue because that method was well known in the art. Questioned by the Board, the appellant indicated that TREF was applicable to determine the intrinsic viscosity specified in the granted claims, in particular if one knew which fractions were obtained.

d) Another possibility to determine the intrinsic viscosity specified in features b) and c) of granted claim 1 would be to perform the sequential polymerisation in different orders (e.g. step c) first, then step b) then step a)).

e) During the oral proceedings before the Board the appellant indicated that to his knowledge each method (Philippoff rule, TREF, various step sequences) could equally be used to determine intrinsic viscosity, the Philippoff rule being however probably the more convenient. All methods would further lead to the same results, taking into account the inherent measurement errors.

f) Regarding the data given in Tables 2 and 3 of the patent in suit, the following held true:

- no xylene-solubles were produced in the third step of the polymerisation process used in the examples of the patent in suit. Therefore, the amounts of xylene-soluble fractions obtained at the end of the second step were somewhat diluted in the third step and the skilled person would know that the xylene-soluble fraction values indicated in Table 3 for the final composition were wrong (they should be lower than those listed in Table 2 for the propylene-ethylene copolymer and amounting to components a)+b));

- as indicated in Table 1, only minor amounts of C3 were copolymerised in the second gas phase reactor. Therefore, the ethylene content in PE was very close to 100 wt%. The value of - exact - 100 wt% indicated in Table 2 was not an error but corresponded to a rounded-up value;

- the values of ethylene content reported in Table 3 were correct when calculated from the polymer content and ethylene content values indicated in Table 2 for the propylene-ethylene copolymer and the polyethylene.

g) It was explained in D10 that the Philippoff rule could not be valid when strong interactions existed between the polymer chains of different constituents of a blend. However, such interactions did not occur for the dilute systems according to the patent in suit.

h) The respondent had not provided any evidence that the invention could not be carried out over the whole breadth of the claims.

i) Information was provided in paragraph [0053] of the patent in suit on the catalyst system used in the examples. Should further information be required, the skilled person could consult either the description and/or the documents indicated in paragraph [0039] of the patent in suit.

j) The respondent's arguments were related to clarity issues not sufficiency.

k) Therefore the requirements of Art. 83 EPC were met.

Auxiliary request

l) The same argumentation was valid as for the main request.

X. The respondents' arguments relevant for the present decision were essentially as follows:

Main request

a) According to paragraphs [0008] and [0009] of the patent in suit the combination of properties specified in granted claim 1 was only achieved by appropriately selecting components a), b) and c) also defined in claim 1. Therefore, in order to carry out the invention, it was essential to be in a position to determine the intrinsic viscosity mentioned in features b) and c).

According to the patent in suit the compositions being claimed were obtained by sequential polymerisation in a 3-reactor system. Such a process conferred a specific morphology to those compositions. In particular, intimate mixtures were obtained at the end of the second and third reactors, so that the individual components could not be isolated and characterised. Therefore, it was not possible to determine the intrinsic viscosity specified in features b) and c) of granted claim 1.

b) The patent in suit was silent in respect of how to determine those intrinsic viscosities.

c) It could not be derived from the information provided in the patent in suit that the intrinsic viscosity was to be determined using the Philippoff rule. In particular, the intrinsic viscosity values necessary for those calculations were not indicated.

As indicated in D10, the Philippoff rule was developed for cellulosic systems. There was no indication in the cited documents that the Philippoff rule applied to the compositions now being claimed, which consisted in a three phase system obtained by a specific sequential polymerisation process.

D10 further failed to disclose that the Philippoff rule applied to heterophasic systems and even taught that it did not work for immiscible systems (pages 161, left column; page 164, right column; page 165, left column and Fig. 4).

d) As shown in D24 and D24a, the appellant argued in a different case that intrinsic viscosities for the same kind of compositions as those being claimed had to be determined using TREF, not the Philippoff rule. The respondent's position was that none of the methods applied.

e) D20 was filed late. None of the examples reported in D20 corresponded to a composition within the scope of granted claim 1. D20 further contained too little information regarding e.g. how the TREF fractionation was exactly done (temperature conditions, nature of various fractions, solvent used) so that it could not be verified. Finally, since the intrinsic viscosity of each of the fractions was not indicated, it could not be concluded whether or not the Philippoff rule worked. Therefore, D20 was not highly relevant and should not be admitted to the proceedings.

f) D12/D12a were concerned with binary compositions comprising two miscible fractions, not a ternary multiphasic system as that being claimed. Besides, on page 15 of said document, which had not been filed by the appellant, it was indicated that the value determined using the Philippoff rule differed from that actually measured. Therefore, D12/D12a in their complete version showed that even for a binary system, there were difficulties in applying the Philippoff rule.

g) The patent in suit contained little information in respect of the catalyst systems used in the examples. Besides, some of the data given in Tables 2 and 3 appeared to be incorrect e.g.

- ethylene content of 100 wt% reported in Table 2, although some C3 is used for the polymerisation;

- same xylene-soluble fractions reported for the propylene-ethylene copolymer in Table 2 and the final composition in Table 3;

- ethylene content reported in Table 3.

Under these circumstances, the examples of the patent in suit could not be reworked and no information could be derived therefrom in order to find out how to determine the intrinsic viscosity specified in the granted claims.

h) Therefore, the patent in suit did not provide sufficient information to determine the intrinsic viscosity specified in features b) and c) of granted claim 1 and, thus, to carry out the claimed invention without undue burden.

Auxiliary request

i) The same argumentation applied as for the main request.

XI. The appellant (patent proprietor) requested that the decision under appeal be set aside and the patent be maintained as granted (main request), or the patent be maintained in amended form on the basis of the auxiliary request filed with the statement of grounds of appeal.

The respondent (opponent) requested that the appeal be dismissed.

XII. The Board announced its decision at the end of the oral proceedings.

Reasons for the Decision

1. The appeal is admissible.

Main request (patent as granted)

2. Sufficiency of disclosure

2.1 In order to meet the requirements of Art. 83 EPC, an invention has to be disclosed in a manner sufficiently clear and complete for it to be carried out by the skilled person over the whole area claimed without undue burden, on the basis of the information provided in the patent specification and, if necessary, using common general knowledge. This means in the present case that the skilled person should in particular be capable to prepare with reasonable effort a polypropylene composition according to granted claim 1.

2.2 Granted claim 1 is directed to a polypropylene composition

- comprising three components, namely a crystalline propylene polymer, an elastomeric copolymer of ethylene and propylene, and a polyethylene (co)polymer, each of those components being present in specific amounts and characterised by specific parameters as defined in features a) to c), and

- having a combination of properties in terms of flexular modulus, stress-whitening resistance and Izod impact resistance at 23 °C and -20 °C.

2.2.1 According to paragraphs [0008] and [0009] of the patent in suit and as agreed by both parties in particular during the oral proceedings before the Board, the combination of properties specified in granted claim 1 is achieved by producing a polypropylene composition comprising the three specific components characterised by features a) to c) according to granted claim 1. It further follows from the patent specification as a whole that such compositions are prepared by means of a sequential copolymerisation process comprising at least three sequential polymerisation stages with each subsequent polymerisation being conducted in the presence of the polymeric material formed in the immediately preceding polymerisation reaction. More particularly, a polymerisation of propylene producing the crystalline polymer a) is first carried out in at least one stage, followed by a copolymerisation stage of mixtures of ethylene with propylene (and optionally a diene) to elastomeric polymer b), and finally a polymerisation stage of ethylene to polyethylene c) is carried out (paragraphs [0022] and [0023] and examples 1-2 of the patent in suit). It was further agreed by the parties that such a sequential polymerisation process leads to compositions having a specific morphology, namely a three phase system wherein the polypropylene a) forms the matrix, the elastomeric polymer b) being dispersed in said matrix and the ethylene (co)polymer c) being dispersed in b).

2.2.2 In the present case, the compositions defined in granted claim 1 are characterised both in terms of their constituents (features a) to c)) and their properties. In that respect, according to the appellant, the combination of properties in terms of flexural modulus, stress-whitening resistance and Izod impact resistance at 23°C and -20°C is only to be achieved if features a), b) and c) specified in claim 1 are met. Moreover, it cannot be concluded from the evidence on file, nor was it argued by the appellant, that features a) to c) are implicitly fulfilled by any compositions satisfying the combination of properties indicated in granted claim 1.

Under these circumstances, in order to prepare compositions having the combination of properties specified in granted claim 1 the skilled person has to be provided with sufficient guidance to prepare, with a good chance of success, a composition comprising an intimate mixture of the three polymers specified therein and each of them meeting the parametric definition indicated in features a) to c) according to granted claim 1.

2.3 According to the patent specification, those compositions may be prepared by sequential polymerisation in at least three reactors as explained above. Although the catalyst to be used is defined in a broad manner in paragraph [0024], it is derivable from paragraphs [0025]-[0040] and from examples 1-2 of the patent in suit that catalyst systems comprising Mg, Ti, halogen, an internal donor, an alkylaluminum compound and an external donor are preferred. Broad indications regarding the amounts of said components, the kind of reactors, residence time, the reaction temperature and pressure are further provided in paragraphs [0029], [0032] and [0042]-[0045] of the patent in suit. However, considering that all the information is very general, it is neither credible nor was it argued by the appellant, that using e.g. any of the preferred catalytic systems and any reaction conditions indicated in the description would automatically lead to compositions comprising the three polymers satisfying the parameters and having the specific combination of properties specified in granted claim 1. Nor is the skilled person provided with any information in the description of the patent in suit regarding which process conditions should be used in order to reliably prepare such compositions, in particular compositions satisfying features b) and c) characterised in terms of intrinsic viscosity.

2.4 Regarding the determination methods of the parameters and properties specified in granted claim 1, indications are provided in paragraphs [0049] and [0050] of the patent in suit. However, regarding intrinsic viscosity, the sole information provided is that it is measured in tetrahydronaphthalene at 135 °C (page 6, line 14 of the specification). The patent in suit contains no other indication with respect to the methodology to be used, including the extrapolation method that would be necessary in order to precisely determine the intrinsic viscosity of the polymer fraction soluble in xylene of the elastomeric copolymer of ethylene and propylene and/or of the ethylene (co)polymer specified in features b) and c) of granted claim 1.

2.5 Considering the specific morphology of the compositions being claimed, namely in the form of an intimate mixture of the three polymeric components defined in granted claim 1, it is obvious that the mere indication of the solvent and the temperature indicated in the patent specification is not sufficient for precisely determining the intrinsic viscosity of the polymer fraction soluble in xylene of the elastomeric copolymer of ethylene and propylene and/or of the ethylene (co)polymer specified in features b) and c).

The appellant did not show the existence of a usual method in the art for determining the intrinsic viscosity of some of the components of the present type of intimate mixtures so that the skilled person would have to rely on common general knowledge to fill the lack of information in the patent in suit in that respect.

2.6 The question arose if that information could be derived from the examples of the patent in suit.

Although the patent in suit provides some information regarding the preparation process used (paragraphs [0053]-[0055], Tables 1, 2), many essential features are missing such as the alkylaluminium/Ti ratio of the catalyst system, residence time, catalyst feed, which are all known in the art to have major impact on the polymerisation process.

Besides, the data listed in Tables 2 and 3 show some inconsistency. The appellant acknowledged during the oral proceedings before the Board that the xylene-soluble fraction values indicated in Table 3 were obviously erroneous since they could not be identical to those indicated in Table 2. At least some doubts further exist regarding the value of ethylene content given in Table 3: even if, to the appellant's benefit, its calculation method is followed, one ends up with an ethylene content of (10.5x0.55)+(17x1)= 22.8 and of (8x0.46)+(16x1)= 19.7 for examples 2 and 1c (instead of 22.4 and 23, respectively, as indicated in the patent in suit). Those errors raise doubts as to the information provided in the experimental part of the patent in suit.

Under these circumstances the lack of experimental details and the inconsistencies in the experimental part of the patent in suit do not allow the skilled person to complete, by reworking the examples, the lack of information of the patent in suit in respect of the method for determining the intrinsic viscosity indicated in features b) and c) of granted claim 1.

2.7 The appellant argued that determination methods of the intrinsic viscosity of the various components of an intimate mixture belonged to common general knowledge and that the skilled person could determine those values e.g. either using the Philippoff rule, TREF or by varying the sequence of the polymerisation stages.

2.7.1 The appellant in particular relied on D9 and D10, which were considered to show that the Philippoff rule was commonly used in the art to determine the intrinsic viscosity of mixtures of polymers.

However, D9 does not specifically deal with polymer mixtures, let alone with multiphasic compositions as those according to granted claim 1. According to equation 8.6 of D9, the Philippoff rule is merely used in order to determine the viscosimetric average Meta, which is usually used to characterise a given polymer, not polymeric blends, in particular multiphasic compositions as claimed.

D10 is directed to the determination of the viscosity of poly(ethylene-copropylene) and poly(alkyl methacrylates) in dilute xylene solutions. Since those systems neither comprise the three components specified in granted claim 1 nor exhibit the specific morphology of the compositions being claimed, it is highly questionable if any information derived from D10 would mandatorily apply to the patent in suit. Besides, D10 teaches (see in particular the paragraph preceding section "3. Experimental" on page 163; page 164, first full paragraph) that polymer blends may not follow the Philippoff rule.

Therefore, neither D9 nor D10 support the appellant's argumentation.

2.7.2 That the Philippoff rule would be of any relevance for the patent in suit can also not be derived from the data provided in Tables 2 and 3 of the patent in suit, in particular because all the intrinsic viscosity values necessary to calculate the intrinsic viscosity values specified features b) and c) of granted claim 1 using the Philippoff rule are not indicated in the patent in suit, in particular not in Tables 2-3 (the intrinsic viscosity of the xylene soluble fractions of a) and a)+b) as well as the intrinsic viscosity of the total composition are in particular missing). Therefore, it could also not be implicitly deduced from the information provided in the patent in suit that the Philippoff rule was or could be used.

2.7.3 The appellant further argued that D20 showed that the Philippoff rule applied to the compositions being claimed, while the respondent requested that D20 be not admitted to the proceedings.

D20 was, although late-filed, admitted to the proceedings by the first instance (see Reasons for the decision, middle of page 2 and bottom of page 3). The respondent's reason for requesting not to admit D20 into the appeal procedure is merely that that piece of evidence would lack the necessary relevance. However, in the absence of any legal basis for not considering D20 which had been already admitted into the proceedings before the first instance and was referred to by the appellant in its statement of grounds, the Board must take into account that piece of evidence and the arguments based on it (Art. 12(4) and 12(1) RPBA).

In the experimental report D20, seven compositions comprising a polypropylene homopolymer, a propylene ethylene copolymer (rubber) and a polyethylene homopolymer were prepared by a sequential process according to the patent in suit, as indicated during the oral proceedings before the Board. Those compositions were fractionated using the TREF technique with various solvents. The data given in Table 1 of D20 show that the Philippoff rule validly applied since the intrinsic viscosity of the total composition is equal to the weight average sum of the intrinsic viscosities of each fractions. However, it was clarified during the oral proceedings before the Board that the various fractions obtained in D20 could not be related to either the polymer fraction soluble in xylene of feature b) or to the polyethylene (co)polymer according to feature c) of granted claim 1. Therefore, D20 does not show that the Philippoff rule could be applied to determine the intrinsic viscosities specified in features b) and c) of granted claim 1. More importantly, that experimental report was not known to the public before the filing date of the patent in suit. Hence, it cannot constitute evidence that the skilled person would have known, at the date of filing, that the Philippoff rule would have to be used for determining the intrinsic viscosity in features b) and c).

2.7.4 Finally, D12/D12a and D24a are submissions by the parties before the EPO in a different case that had been made years after the filing date of the patent in suit. Hence, those submissions are of no relevance for establishing whether or not the skilled person would have applied at the date of filing of the patent in suit the Philippoff rule for carrying out the claimed invention.

2.7.5 Regarding the TREF technique, it was not contested during the proceedings that that method was known in the art. During the oral proceedings before the Board, the appellant, questioned by the Board, stated that the TREF technique could be used in order to separate either the polymer fraction soluble in xylene of feature b) or to the polyethylene (co)polymer according to feature c) of granted claim 1, and hence, to determine the intrinsic viscosity of those fractions. Although it is credible that the skilled person might contemplate to use the TREF technique to separate the various components of the mixture defined in claim 1, no evidence is available that the details of the methodology to be used to obtain reliable measurements would be known to him. D20, as shown above in section 2.7.3, is of no help in that respect as it relates to the separation of different fractions.

2.7.6 During the oral proceedings before the Board, the appellant argued that the intrinsic viscosity specified in features b) and c) of granted claim 1 could also be determined by performing the sequential polymerisation process using different sequences of polymerisation i.e. by varying the order in which the polymers were prepared.

However, the appellant has in particular neither shown that the same products would be obtained by varying the sequence of the polymerisation stages, nor that the intrinsic viscosities specified in granted claim 1 would, for each sequence, remain the same.

2.7.7 In the present case, the appellant has therefore not demonstrated by any evidence that the determination of the intrinsic viscosity specified in features b) and c) of granted claim 1 could be reliably established by a well known method and/or using common general knowledge. The appellant further argued that at least three methods existed for that determination. However, there is no evidence on file nor was it convincingly shown by the appellant that all these methods, in the absence of necessary detailed explanations to put them in practice, would lead to the same results in terms of intrinsic viscosity. In that respect, the methodology used for precisely measuring the intrinsic viscosity is also not indicated (see above point 2.4).

In these circumstances, the skilled person wanting to prepare a polypropylene composition inevitably satisfying features a) to c), which is necessary in order to obtain the combination of properties in terms of flexural modulus, stress-whitening resistance and Izod impact resistance at 23°C and -20°C specified in granted claim 1, is left with the task of performing an elaborate program in order to find out essential aspects of the preparation conditions and determination method of the intrinsic viscosity to be used. In other words, the skilled person can only establish by trial and error whether or not his particular choice of working conditions and intrinsic viscosity determination method will provide compositions according to granted claim 1, which amounts to an undue burden and is contrary to the requirements of Art. 83 EPC.

As explained above, the question in the present case is not whether or not the skilled person knows if he is working within or outside the scope of granted claim 1 because of an ambiguity in the determination method of a parameter specified in the granted claims but rather if the skilled person would have had enough guidance in respect of the working conditions and determination method of intrinsic viscosity in order to prepare a composition satisfying all the requirements of granted claim 1. Therefore, the lack of information in that respect in the patent in suit effectively amounts to a lack of sufficient disclosure according to Art. 83 EPC and not merely to a lack of clarity under Art. 84 EPC. Therefore, the appellant's argument that the respondent's insufficiency objections were only related to clarity is dismissed.

2.8 For these reasons, the main request does not meet the requirements of sufficiency of disclosure (Art. 83 EPC) and has to be refused.

Auxiliary request

2.9 The amendment made in the auxiliary request consists in inserting in the claim information regarding the solvent and the temperature to be used for the determination of intrinsic viscosity. Considering that the assessment of sufficient disclosure has to be made considering the patent as a whole and since that information was already taken into account for assessing sufficiency of disclosure with respect to the main request, the amendment made cannot change the reasoning and the conclusion given for the main request. Therefore, the auxiliary request does not meet the requirements of Art. 83 EPC either.

3. Since none of the appellant/patent proprietor's requests is allowable, the appeal has to be dismissed.

Order

For these reasons it is decided that:

The appeal is dismissed.