T 0876/00 29-07-2003

Reasons for the Decision

1. The appeal is admissible.

2. Procedural Matters

With the withdrawal of the opposition during the appeal proceedings, both Respondents ceased to be parties to these proceedings in respect of the substantive issues (T 789/89, OJ EPO 1994, 482, cf. the headnote; T 616/98 of 1 February 2001, section 2.1 of the reasons; T 642/97 of 15 February 2001, Section 2 of the reasons).

3. Wording of the Claims

3.1. Article 123(2) EPC

3.1.1. The consolidation of "Set A" and "Set B" does not extend the content of the patent in suit beyond that of the application as originally filed, since it means only the limitation of "y" to the preferred value of "1" (page 12, lines 17/18 of the published application) and the consequential limitation of "x" to an integer value of from 0, 1, 2, 3 or 4, already individually disclosed in Claim 1 as filed.

3.1.2. The replacement of "may be" by "is" and "are", respectively, in Claims 1, 10 and 18 of "Set B" and in Claim 1 of "Set C" limit the definition of Q to the meanings actually disclosed in the respective paragraphs of these claims. Other meanings had never been contemplated, let alone defined in the application from which the patent in suit is derived.

3.1.3. The deletion of the passages "such as a halogen atom, or (an) alkoxide radical" is supported by page 4, lines 5 to 7, page 10, lines 5 to 17, page 11, line 19 to page 12, line 16 and page 13, Table 1 in the published application.

3.1.4. The corrections of clerical errors made in Claims 1, 2. and 18 of "Set B" as indicated in section VI, above, comply with Rule 88 EPC for the following reasons:

The wording in these claims as granted were clearly erroneous. Thus, the comparison of the definition of component B.(2) of the said Claim 1 with the definition of the anion [A]- derived therefrom in Claims 10 and 18, or of Claim 2 with Claims 11 and 23 further demonstrates that nothing else would have been intended than what has been offered as the correction. The same argument is valid for the formula in Claim 18. It is evident that the anion was missing, which had already been defined in the last paragraph of the claim.

3.2. Hence the amended and corrected claims comply with Article 123(2) EPC. The objection that they did not was already held unsuccessful in the decision under appeal (section III, above), and was not further pursued in appeal.

3.3. Article 123(3) EPC

From the above considerations, it also is evident that the amendments in the claims always result in a limitation of the scope of the claims, by deletion of alternative embodiments and of a reference to a non- limiting example of such an embodiment, respectively, so that the requirements of Article 123(3) EPC are also met by the present wording of these claims.

3.4. Article 84 EPC

In view of the above amendments, the Board is also satisfied that the above amendments comply with the clarity and conciseness requirements of Article 84 EPC.

4. Article 100(b) EPC

In the decision under appeal, the Opposition Division had come to the conclusion that the requirements of Article 83 EPC had been fulfilled by the patent in suit. In the absence of any arguments from the former Opponents, which have not contested this finding, the Board has no reason to take a different view.

If follows that the requirements of Article 83 EPC are met.

5. Novelty

Since the revocation of the patent in suit was based on the finding that Claim 10 of "Set B" lacked novelty, it is this claim to which reference is made in the following considerations.

5.1. The patent in suit is based on a previous application filed in the USA on 22 June 1990. Each of D1, D2 and D3 was published after that date (D1 on 20 March 1991, D2 on 13 March 1991 and D3 on 10 April 1991).

Moreover, in each of these documents, the priorities of more than one previous patent application were claimed, at least one of which, in each case, had a filing date later than 22 June 1990. Consequently, those parts of the disclosures of D1, D2 and D3, respectively, derived from previous applications filed on or after 22 June 1990 or not derived at all from any previous patent application are not state of the art with respect to the patent in suit. In other words, only those parts of each of D1, D2 and D3 can be taken into account as state of the art in this decision, for which proof has been provided that they benefit of a filing date prior to 22 June 1990 (Articles 54(3), 87, 88 and 89 EPC).

Together with the Notices of Opposition, copies of previous US patent applications "US1", "US2" and "US3" having filing dates of before 22 June 1990 were filed (section 11(a)).

Consequently, novelty can only be assessed with regard to those parts of the above documents derived from these previous US patent applications for the assessment of novelty under Article 54(3) and (4) EPC.

5.2. However, the disclosure of a given cited document is not limited a priori to specific parts of its specification or particular embodiments, hence the skilled person does not disregard those parts, which in the course of further detailed enquiries may turn out as not enjoying a certain priority date. He will take the whole contents of the document into account (cf. T 12/81, OJ EPO 1982, 296 , point 5 of the reasons). The only guidance to deviate from this approach to a document is prima facie given by clear preferences for specific embodiments in the document. Hence, the document cited as prior art has to be considered at first as disclosed. Those parts which then appear to anticipate the claimed subject-matter have, in a second step, to be examined whether they are state of the art in the sense of Article 54(3) and (4) EPC. Any other way of examination eg by early removing - in the light of the previous foreign application (priority document) - any "younger" parts of the disclosure from consideration might change the overall information in the cited document and, thus, would include the risk of ex post facto analysis (in the sense of T 305/87, OJ EPO 1991, 429; point 5.3 of the reasons).

These considerations are consistent with the Opinion G 2/98 (above), in particular, with regard to the arguments relating to the so-called " 'OR'-claim" (points 6.5, 6.7 and 6.8 of the reasons).

5.3. Document D1 as such relates to monocyclopentadienyl or substituted monocyclopentadienyl metal complex containing compounds of the formula

CpMXn+A-,

wherein Cp denotes an optionally substituted Eta5- cyclopentadienyl ligand, optionally covalently bound to the metal by a covalent group -Z-Y-, M is a metal of Groups 3 to 10 (or IIIB to VIIB and VIII in terms of the previous denominations of the transition metal groups, as used in "US1") or the Lanthanide Series of the Periodic Table of Elements, each X, independently, relates to hydride or a moiety selected from halo, alkyl, aryl, silyl, germyl, aryloxy, alkoxy, amide, siloxy, combinations thereof and neutral Lewis base ligands, having up to 20 non-hydrogen atoms; or optionally, one X together with Cp forms metallocycle with M of up to 20 non-hydrogen atoms; n is 1 or 2 depending on the valence of M; and A is a non- coordinating, compatible anion of a Bronstedt acid salt. Z is a divalent moiety comprising oxygen, boron or a member of Group 14 (the former IVA main group) and Y is a linking group covalently bonded to the metal comprising nitrogen, phosphorus, oxygen or sulfur, or optionally Z and Y together from a fused ring system.

5.3.1. As argued in the Statement of Grounds of Appeal, D1 encompasses a broad variety of possible combinations of metals and ligands built up from the different subgroups and radicals, each of which requires selections to be made for each group of the formula from the different lists of constituents and radicals (Cp, M, X and A). This is demonstrated by a long list of more than 25 exemplary species in four different sub-lists (column 4, line 25 to column 5, line 7). Only eight compounds in this list relate to compounds having a "bridged" Cp ligand (ie a Eta5-ligand additionally bonded via a -Z-Y-bond to the metal), while fifteen compounds (optionally additionally) contain halogen and/or hydrocarbyloxy ligands (sections IV(a) to (c) and (e), above).

5.3.2. However, the meaning of group -Z-Y- is further specified in D1 as denoting a group of the structure - (ER"'2)p-Y'(R"')- wherein E is carbon, silicon or germanium; Y' is nitrogen or phosphorus; R"' is alkyl, aryl, silyl or a combination thereof having up to 10 carbon or silicon atoms, p is an integer of from 1 to 4 (column 3, lines 27 to 43).

In highly preferred compounds of the formula

FORMULA

M is Zr or Hf, Cp* is a cyclopentadienyl or substituted cyclopentadienyl bound in a Eta5 bonding mode to M; Z is SiR*2, CR*2, SiR*2SiR*2, CR*2CR*2, CR*=CR*, CR*2SiR*2, or GeR*2; Y is a nitrogen or phosphorus containing group of the formulae -N(R"")- or -P(R"")- (with R* being hydrogen or alkyl, aryl, silyl, halogenated alkyl, halogenated aryl groups and combinations thereof having up to 20 non-H atoms; R"" being C1-10 alkyl or C6-10 aryl, or two or more R* and R"" groups together form a fused ring system of up to 30. non-H atoms); X is halo, alkyl, aryl, alkoxy, or aryloxy of up to 20 carbons; n is 1 or 2; and A- is a non-coordinating, compatible anion of a Bronsted acid salt.

Amongst those eight compounds in the list containing a bridged ligand, referred to in section 5.3.1, above, only two are free of a halogen (chloride) ligand, ie (tert-butylamido)dibenzyl(tetramethyl-Eta5- cyclopentadienyl)silane zirconium dibenzyl and (phenylphosphido)dimethyl(tetramethyl-Eta5- cyclopentadienyl)silane zirconium dibenzyl (column 5, lines 1, 2, 5 and 6).

5.3.3. Finally, whilst Examples 1 and 2 relate to catalysts on the basis of Eta5-[C5(CH3)5] titanium compounds which are free of any -Z-Y- group, Example 3 of D1 discloses a catalyst of the above general structure, wherein the relevant groups and ligands would be in terms of the symbols as defined in Claim 10 of the patent in suit: R = methyl; x = 4; B = Si(CH3)2; (JR'z-1-y) = tert-butylamido; Q = methyl; A = pentafluorophenyl borate. A comparison with Claim 10 of the patent in suit indicates that the catalyst of Example 3 would fall within the definitions in that claim.

5.3.4. These considerations show that D1 relates to a broad range of conceivable complex compounds with or without a bridged Cp ligand, without putting a clear emphasis on one of these two possibilities. However, only those containing such a ligand can play a further role in this examination.

5.3.5. Apart from the examples, the compounds according to Claim 10 of the patent in suit differ, according to the Appellant, from the above preferred compounds in D1 identified in the previous paragraphs by a purposive selection of particular meanings of ligand Q out of the various denotations of ligand X in the document.

Thus, ligand Q in Claim 10 is a hydride, a C1 to C50 hydrocarbyl radical, a C1 to C50 substituted hydrocarbyl radical wherein one or more hydrogen atoms is replaced by an electron-withdrawing group, or a C1 to C50 hydrocarbyl substituted metalloid radical wherein the metalloid is selected from the Group IV-A (= Group 14), provided that Q is different from the Cp- ligand, whilst halogen, alkoxy and aryloxy are excluded.

5.3.6. The arguments of the Appellant (see section IV(c), above; Statement of Grounds of Appeal, section III, page 4 et seq.) that ligands X having these latter meanings represent poor leaving groups and are excluded from Q, and that compounds containing such ligands X are not active catalysts, on the one hand, for this reason or, on the other, because they would split off catalyst poisons (hydrogen halide or alcohol) when reacted with the second component of the catalyst (D1: column 8, lines 36 to 49) have not been disputed, let alone refuted by the former Respondents.

This information could neither be taken nor derived from D1. On the contrary, the list of preferred compounds in column 4, line 25 et seq. contains numerous species containing halogen and/or hydrocarbyloxy ligands (sections 5.3.1 and 5.3.2, above).

Hence, the patent in suit contains further essential information not derivable from D1, and the claimed subject-matter is limited accordingly. Therefore, the Board accepts that a purposive selection was made.

However, Example 3 of D1 and the two species mentioned in the last paragraph of section 5.3.2, above, meet these requirements in the claims of the patent in suit.

5.3.7. As already stated above, D1 is a document of the state of the art in accordance with Article 54(3) EPC. Therefore, it is necessary to identify those parts of the document which can be taken into account.

5.3.8. In "US1", the definitions of the symbols in the above formula CpMXn+A- differ in certain respects from those given in D1: Cp may optionally be covalently bonded to M through an unspecified substituent; M is a metal of Groups IIIB to VIB and VIII or a Lanthanide element; X denotes halo, alkyl, aryl, N-R2, aroxy or alkoxy of up to 20 carbon atoms, or two or more X groups together form a metallocycle with M, R being alkyl or aryl of up to 10 carbons. One of the optional substituents of Cp may be a group R" according to one of the following structures:

FORMULA

wherein p is an integer from 1 to 4, Y is nitrogen or phosphor[o]us and each R"', independently, denotes alkyl, aryl, silyl or combinations thereof having up to 10. carbon or silicon atoms. M is preferable Ti or Zr (Claim 1 and page 5).

A single compound having a metallocyclic structure is disclosed in "US1": (N-t-butylamino)(dimethyl)-Eta5- 2,3,4,5-tetramethylcyclopentadienyl)silane zirconium dichloride. "Such components" are said to be readily prepared by combining the corresponding metal chloride with a dilithium salt of the substituted cyclopentadienyl group such as a cyclopentadienyl- alkane, -silyl amine, or -phosphine compound. Then a further reference is made to an unpublished US patent application said to disclose "certain of these compounds". However, no mention is made of that application in D1 at all, nor are any particulars available about that application as to its availability to the public, or which would indicate that or which part of the further disclosure in D1 was based on that application ("US1": page 6, line 23 to page 7, line 9).

The US application contains two examples corresponding to Examples 1 and 2 of D1.

5.3.9. The comparison of these particulars shows that only those compounds of the formula CpMXn+A- containing one Eta5- (C5R'4R") group can be taken into account for novelty which are within the scope of the following limits:

R' is hydrogen, halogen, alkyl, aryl or haloalkyl of up to 10 carbon atoms (the latter limit is not disclosed in D1);

R" is an -SiR"'2-Y(R"')- group, Y is -N(R"")- or - P(R"")-, wherein R"' (= R*) is alkyl, aryl or silyl within a range not exceeding 10 carbon or silicon atoms (whilst this limit is not contained in D1) and R"' (= R"") is an C1-10 alkyl or C6-10 aryl group (the C6 limit not being disclosed);

X is halogen, alkyl, aryl, alkoxy or aryloxy of up to 20. carbon atoms;

M is an element of Groups 3 to 6 or 8 to 10 or a Lanthanide element, preferably Ti or Zr;

n is 1 or 2 (dependent on the valence of M); and A- is a non-coordinating, compatible anion of a Bronsted acid salt (D1: column 1, lines 31/32; column 3, line 1 to column 4, line 18; "US1": page 2, line 30/31; page 3, lines 1 to 10; page 5, lines 1 to 35).

This list clearly demonstrates that the disclosure of D1, as far as belonging to the state of the art in the sense of Article 54(3) EPC (ie in the light of "US1"), still encompasses a large range of conceivable combinations of the different constituents in the above formula, requiring purposive selections for each of these constituents to be made in order to arrive at an active catalyst within the terms of Claim 10.

Furthermore, the only species in "US1" containing a bridged Cp ligand (but which also had halogen ligands rendering the species unsuitable for the preparation of an active catalyst) was deleted from the subject-matter used to prepare the text of D1 and, hence, not published therein.

In fact, none of the relevant individual species referred to in the general disclosure of D1 (in particular, those referred to in the last paragraph of section 5.3.6, above) nor in Example 3 thereof have an antecedent in "US1". Consequently, neither those species nor Example 3 of D1 are state of the art with respect to the patent in suit.

5.3.10. Moreover, both D1 and "US1" define their ligand X in the same way, inclusive of five preferred radicals, three of which prevent, undisputed by the former Respondents, that active catalysts can be prepared from those complex components containing these radicals, and are, therefore, excluded from Claim 10 of the patent in suit.

Hence, as shown above, the patent in suit is based on further essential information, that specific constituents, in particular specific X ligands, have to be selected in order to obtain active catalysts for the polymerisation of olefins, and the subject-matter claimed is limited accordingly. However, no such information is derivable from D1, nor from "US1".

5.3.11. Consequently, the subject-matter of Claim 10 is novel, in the sense of Article 54(3) and (4) EPC over the disclosure of D1.

5.4. Only a small part of Document D2, Claim 1 of which does not even require its Pi-bonded moiety to be cyclopentadienyl, is supported by "US2". More particularly, the metal coordination complex used as a starting material for the preparation of the catalyst contains a ligand X which at each occurrence is hydride or a moiety selected from halo, alkyl, aryl, silyl, germyl, aryloxy, alkoxy, amide, siloxy and combinations thereof having up to 20 non-hydrogen atoms and neutral Lewis base ligands having up to 20 non-hydrogen atoms. In a preferred embodiment of D2, X is hydride, halo, alkyl, aryl, aryloxy or alkoxy of up to 10 carbons (page 7, lines 14 to 18 and 52). In "US2" support is found for halo, alkyl, aryl, -NR2 (R being hydrogen, silyl, alkyl, aryl and combinations thereof having up to 10 carbons), phenoxy and alkoxy of up to 10 carbons (second half of page 3). It follows that the above considerations concerning a purposive selection are also valid for D2.

Moreover, the above metal coordination complex is activated by means of (b) an organoaluminium compound containing at least one oxygen atom bound to the aluminium atom or between two such atoms, ie an aluminoxane (D2: page 5, lines 27 to 31; "US2": Claim 8; page 4, lines 1 to 6). This means that D2 does not disclose an ionic catalyst containing an anion A- as defined in Claim 10, but corresponds to a catalyst involving a "conventional" method of activation as referred to in the patent in suit as being less advantageous (page 10, lines 46 to 48).

Consequently, the subject-matter of Claim 10 is novel, in the sense of Article 54(3) and (4) EPC over the disclosure of D2.

5.5. D3 relates to a process for polymerising vinyl aromatic monomers, such as styrene, by means of a monocyclopentadienyl metal catalyst having the formula [CpmMXnX'p]+A-.

5.5.1. In the catalyst, Cp denotes an optionally substituted Eta5-cyclopentadienyl which may be bonded to M through a substituent; M is a metal of Groups 3 to 6 or 8 to 10 or a lanthanide; the ligands X are inert anionic ligands; at least one of these ligands is however R, the latter being hydride, hydrocarbyl or silyl, optionally substituted with one or more halogen atoms or alkoxy groups and having up to 20 C- and/or Si-atoms; X' is an inert, neutral donor ligand; m and p are independently 0 or 1; n >= 1; the sum of m + n is two less than the valence of M; A- is a non-coordinating compatible anion of a Bronsted acid salt (D3: Claim 1; page 2, lines 19 to 30). Illustrative examples for groups X other than R, are halo, NR2, PR2, OR, SR, BR2 (page 2, lines 45).

Ti, Zr, Hf, Cr, La are named to illustrate M; Ti and Zr being preferred. According to particular embodiments, the substituent further bonding Cp to the metal may be alkyl haloalkyl of up to 6 C-atoms or C2-5 oxyalkylene or alkylene. However, in the numerous individual examples of compounds useful for the preparation of catalysts within the above definitions, not a single species contains a bridged cyclopentadienyl, as pointed out by the Appellant in the letter of 19 December 1997 (page 14, point 2.33). Moreover, the list contains a significant number of individual compounds without any cyclopentadienyl ligands and of those containing halogen or hydrocarbyloxy groups (D3: page3, line 30 to page 4, line 18) None of the examples of D3 discloses a bridged complex compound either.

Hence, the considerations as to the disclosure of a citation and those concerning the necessity for a purposive selection in order to arrive at an active catalyst for the polymerisation of olefins (sections 5.2, 5.3.1 and 5.3.10, above) are also valid for D3. This document does not provide a clear and unambiguous disclosure of a monocyclopentadienyl titanium or zirconium compound, wherein this ligand is further bonded to the metal by a bridging substituent. In the letter of 22 September 1998, this fact was conceded by former Opponent 01 (page 10, point 6) who also put emphasis on the fact that D3 had been of particular for former "Set A" (wherein y could be 0).

5.5.2. These findings are not rendered invalid by "US3", wherein support is only found for a compound of the formula CpMXn+A-. The symbols in the formula have the same meaning as in "US1", apart from the fact that the definition of X extends even further and includes silyl, PR2, OR, SR, and BR2 having up to 20 C-, Si- or C- and Si-atoms, with the proviso that at least one X is alkyl, aryl, silyl or a combination thereof (Claim 1 and page 2, line 15 to page 3, line 30). Nor is this assessment rendered invalid by the fact that the optional bridging group is also further defined in a way similar to that in "US1" (section 5.3.2, above) and refers within these definitions preferably to oxyalkylene (page 6, line 32).

5.5.3. Consequently, the subject-matter of Claim 10 is novel, in the sense of Article 54(3) and (4) EPC over the disclosure of D3.

6. The above considerations, in particular those relating to the necessity for a purposive selection of the groups and radicals of the catalyst in order to arrive at an active catalyst for the polymerisation of olefins, are also valid for the catalyst system as used in the process of Claim 18 of "Set B" and Claim 1 of "Set C". Therefore, their subject-matter is novel over any one of D1, D2 and D3 for the same reasons.

7. Claim 1 of "Set B" differs from the catalyst system of Claim 10 in that it is defined in terms of a product-by- process claim. The definitions of the relevant groups and radicals in the starting compounds A. and B., which form the activated catalyst, are, however, identical to those in Claim 10. Therefore, Claim 1 also meets the novelty requirement with respect to the cited prior art.

8. In summary and for the reasons given above, the Board has come to the conclusion that the main request meets the requirements of the EPC.