T 0076/03 () of 25.10.2005

Summary of Facts and Submissions

I. The grant of European patent No. 0 659 821 in respect of European patent application No. 94 309 753.5, filed on 23 December 1994 and claiming the priority of 27 December 1993 of an earlier patent application in Japan (332991/93), was announced on 22 March 2000 (Bulletin 2000/12) on the basis of a set of five claims, Claim 1 reading as follows:

"1. A tread rubber composition for tires, which comprises rubber components consisting, when the weight of the rubber components is assumed to be 100 parts by weight, of 20 to 80 parts by weight of a diene rubber and 80 to 20 parts by weight of one or more styrene-butadiene copolymers having a glass transition temperature of from -50ºC to -25ºC and satisfying a relation shown by the equation: 4.8X-Y >= 32, in which a bonded styrene content is represented by X% by weight and a 1,2-bond content in butadiene portion is represented by Y% by weight; characterised in that a temperature dispersion curve of tan delta of the rubber composition having two peaks; the temperature difference between the peak on the highest temperature side thereof and a single peak of a rubber composition using the styrene-butadiene copolymer alone as a rubber component being within 10ºC; and a tan delta measured at 50ºC being 0.08 to 0.20.".

The remaining dependent Claims 2 to 5 related to elaborations of this subject-matter.

Styrene-butadiene copolymer rubber(s) will be referred herein below as "SBR" or "SB-rubbers", "Tg" will be used to indicate the glass transition temperature of the SBR, and the loss factor measured at a given temperature TºC will be symbolised by "tan delta|T". Thus, the last feature in Claim 1, as quoted above, would be referred to as "tan delta|50 being 0.08 to 0.20". The above requirement "4.8X-Y >= 32" will referred to as the "equation".

II. On 20 December 2000, a Notice of Opposition was filed, in which revocation of the patent in its entirety was requested on the grounds of Articles 100(a) and (b) EPC, because of lack of novelty and of inventive step and because the patent would not disclose the invention in a manner sufficiently clear and complete for it to be carried out by a person skilled in the art. The grounds of opposition under Article 100(a) EPC were based on the following documents:

D1: US-A-4 906 697,

D2: US-A-4 485 205,

D3: EP-A-0 447 066,

D4: ASTM E 1356-91, Standard Test Method for Glass Transition Temperatures by Differential Scanning Calorimetry or Differential Thermal Analysis",

D5: G. Kraus et al., "Properties of Random and Block Copolymers of Butadiene and Styrene. I. Dynamic Properties and Glassy Transition Temperatures", J. Appl. Polym. Sci., 11 (1967), 1581 to 1591,

D6: EP-A-0 048 618,

Appendix I:

"Calculation of Glass Transition Temperature" and

Appendix II:

"SBRs of E1" (amended version of Fig.8 of the patent in suit)

The Board uses the denotation of the documents used in the decision under appeal (D1 to D6 instead of E1 to E6 as used in the Notice of Opposition).

(1) Firstly, an objection of lack of clarity of Claim 1 was raised by the Opponent (Article 84 EPC). Whilst it was accepted that this was not a ground for opposition, the practice of the EPO to disregard features of a claim, which were not clear, should, according to the Opponent, also be followed in this case, and, therefore, the Tg range of from -50 to -25ºC in Claim 1, above, should not be regarded as a distinguishing feature. This view was explained by means of the argument that, according to D4, Tg could be determined by different methods, differential scanning calorimetry (DSC) or differential thermal analysis (DTA), and there was, along the respective curve, measured by either method, more than one conceivable point which could be taken as the Tg. Moreover, Tg could be calculated, according to D5, by means of the so-called Gordon-Taylor equation (herein below: "G-T-equation"), as presented by the Opponent in Appendix I.

(2) Secondly, on the basis of its interpretation of page 3, lines 24 to 26 of the patent in suit, the Opponent took the view that the claimed subject-matter was insufficiently disclosed, if tan delta, which was an essential feature of the claimed subject-matter, was affected by some feature or variable other than the Tg.

(3) This argument was, in particular, used in the context of an objection of lack of novelty raised with regard to D1. According to the Opponent, this document related, like the patent in suit, to tread rubber compositions for tyres comprising rubber components, which consisted of diene rubber and SBR, and these known compositions showed also a bimodal tan delta-temperature curve, ie two peaks. However, D1, admittedly, did not explicitly disclose that the temperature difference between the peak on the high temperature side of the above loss factor-temperature curve of the rubber composition and the single peak of a rubber composition using SBR alone (herein below: the "deltaT-value") was within 10ºC as required in Claim 1. However, the Opponent submitted that the compositions of D1 inherently had this feature.

(4) In particular, the 50:50 composition of natural rubber and SBR of Example 4 of D1 allegedly anticipated the claimed subject-matter of the patent in suit. Inferring from the fact that (i) the "SBR-D" fulfilled the "equation" and (ii) its composition was very close to that of "SBR-A" used in some examples of the patent in suit (facts and arguments: page 4, penultimate paragraph), all of which gave acceptable results, and further inferring from the patent specification (page 3, lines 14 to 18) that (iii) the deltaT-value would tend to become larger than 10ºC, when the "equation" was not fulfilled, the Opponent submitted that SBR-D in Example 4 of D1 would automatically have a deltaT-value as required in Claim 1. Any argument of the Patent Proprietor used to dispute this finding would indicate that the subject-matter was insufficiently disclosed (facts and arguments: at the bottom of page 4).

(5) If, despite the arguments presented above, the Tg range required in Claim 1 was nevertheless taken into account, the Opponent argued on the basis of Appendix I (above), that even the Tg requirement was met by SBR-D of D1, although the document had been silent in this respect.

(6) Instead of tan delta|50-values, Table 2 of D1 provided tan delta|60-values for all its compositions, all of which were in the range of 0.08 to 0.2. As shown for its Example 1 in Fig.1 of D1, the loss factors measured at those two temperatures were not, in the Opponent's view, significantly different from each other and both would in any case be < 0.2. This could also be expected for the other examples of D1. Nor did the patent in suit itself appear to distinguish between tan delta-values measured at different temperatures between 50 and 70ºC; rather, it was only required that tan delta|50-70-values were sufficiently low. On page 3, lines 24 to 26, the patent in suit, moreover, indicated that a Tg of <-25ºC was required to give adequate tan delta|50-70-values. Since, the SBR-D in Example 4 of D1 appeared to have such a low Tg, it could be inferred that it would also have a suitable tan delta, otherwise the disclosure in the patent in suit would be insufficient (cf. section II (2), above).

(7) In order to fill the gaps of missing Tg-, tan delta|50- and deltaT-values in D1, the Opponent additionally filed an experimental report with a letter dated 14 August 2002 (Declaration signed on 9 August 2002 by Mr Nakamura).

(8) Document D1 was also identified by the Opponent as the closest state of the art for its arguments concerning inventive step, because its tread rubber composition "aimed at exactly the same purpose as the patent in suit", ie it was designed to improve tyre grip and to reduce frictional resistance. In order to achieve this, a bimodal tan delta-temperature curve of the composition was required. Moreover, as shown in Table 2 of D1, the desired properties had been achieved by those compositions, so that the primary objects of the patent in suit had already been met by D1. If novelty of the claimed subject-matter over D1 was acknowledged, the only distinguishing features could be seen in the tan delta|50- and deltaT-values. However, no beneficial effect had been demonstrated which could rely on these features. Consequently, there was no inventive step.

(9) According to the Notice of Opposition, similar comments applied to documents D2 and D3.

(10) These arguments were disputed by the Patent Proprietor in letters dated 14 August 2001 and 16 August 2002. Together with the latter submission, it filed three auxiliary requests and copies of the following Exhibits A to E:

A: ASTM D 3418-82,

B: "Rubber Chemistry and Technology", Vol. 53, No. 3, July-August 1980, pages 437 to 445,

C: "JSR Handbook", 1988, pages 17 to 19, (in Japanese)

D: Handbook of Rubber Industry (in Japanese), 4**(th )Edition, 1994, pages 217 and 1278 to 1279, together with a translation of paragraph 10.2.1 (page 1278) into English, and

E: an experimental report of the determination of the Tg of SBR 1500 on 5 July 2002.

(11) The Auxiliary Requests differed from the claims as granted by the following modifications (cf. section I, above):

first Auxiliary Request: in Claim 1, the deltaT-value of the composition was to be "within 6ºC";

second Auxiliary Request: Claim 1 contained a narrower Tg-range of from "-40ºC to -30ºC" and Claims 3 and 5 as granted had been deleted;

third Auxiliary Request: the "equation" in Claim 1 read "4.8X-Y >= 70".

III. In the decision orally announced at the end of oral proceedings on 16 October 2002 and issued in writing on 12 November 2002, the patent was revoked.

(1) The Opposition Division rejected the ground for opposition under Article 100(b) EPC (insufficient disclosure), accepted that the amendments in the auxiliary requests did not contravene Article 123(2) EPC and acknowledged novelty of the subject-matter according to the Main Request and each of the Auxiliary Requests with regard to each of D1, D2, D3 and D6.

(2) In particular, the difference with regard to D1 was seen in the lack of a clear disclosure of the Tg, the calculation of which according to D5 would have required the knowledge of "additional features such as the content of the cis/trans-configuration". Moreover, although requiring two tan delta peaks on the loss factor-temperature curve, D1 was found to be silent with respect to the temperature difference between these peaks; nor was the temperature dependency of tan delta explicitly disclosed in D1. Neither D2 nor D3 disclosed the "equation" 4.8X-Y >= 70 (in Claim 1 of the third Auxiliary Request) or >= 32 (in each Claim 1 of the higher ranking requests), nor could the deltaT-feature be determined from the loss factor-temperature curve. As regards D6, it was held that it did not unambiguously disclose all the essential features of Claim 1.

(3) Concerning the question of inventive step, the decision under appeal started from an initially formulated technical problem to be solved vis-à-vis the closest state of the art, D1, relating to the provision of a styrene-butadiene rubber composition having a superior balance between rolling resistance and braking.

However, since no convincing evidence had been provided to show that this problem had been solved by means of the distinguishing feature or features, the Opposition Division reformulated the problem so as to relate to the provision of further rubber compositions, and found that it would have been obvious to the skilled person to test rubbers similar to those as known from D1. Consequently, it was held that the claimed subject-matter according to each of these requests did not involve an inventive step with respect to D1.

IV. On 15 January 2003, a Notice of Appeal was filed by the Patent Proprietor/Appellant against this "decision in its entirety". The prescribed fee was paid on the same date. In the Statement of Grounds of Appeal, received on 19 March 2003, the Appellant requested that the decision under appeal be set aside and that the patent in suit be maintained on the basis of those requests dealt with in the decision under appeal, ie the set of claims as granted, as the Main Request, or, in the alternative, on the basis of one of the sets of claims filed on 16 August 2002, as first to third Auxiliary Requests (section II (11), above).

(1) The Appellant pointed out that Claim 1 of the Main Request required (i) the SBR to have a Tg of from -25 to -50ºC and to satisfy the equation 4.8X-Y >= 32 and (ii) the tread rubber composition as a whole to show two peaks in the tan delta-temperature-dispersion curve and to have a deltaT-value <= 10ºC in order to provide tyres manufactured therefrom having a good balance between rolling resistance and wet grip characteristic. Moreover, several examples given in the patent in suit (Table 2) would, in comparison with the comparative examples, show how to compound SBR as required (cf. item (i), above) in order to obtain a composition fulfilling the above requirements as mentioned in item (ii), above. Hence, Article 83 EPC would be met.

(2) In regard of the reasons in the decision under appeal, the Appellant assumed that the Opposition Division had misinterpreted Claim 1, and pointed out that all features in this claim concerning the tan delta-temperature dispersion curve related to the rubber composition as a whole, but not to the SBR alone.

(3) With respect to the assessment of inventive step in the decision under appeal, the Appellant emphasised that each of the measurements of the rolling resistance and of the wet grip characteristics as given in the examples was an index value relative to a reference value (set to "100" in each case) of a tyre made from the conventional tread rubber composition described in Comparative Example 5 of the patent in suit. As opposed to the comparative examples, wherein at least one of these properties was inferior to the above reference tyre, the examples in accordance with the claimed subject-matter showed, according to the Appellant, improved results in respect of both features. Therefore, the conclusion drawn by the Opposition Division was based on errors and misinterpretations, and it was, consequently, not justified.

V. The Respondent, however in a letter dated 25 July 2003, maintained its initial request. In particular, it reiterated its previous objections under Article 100(b) EPC. With respect to both items of Article 100(a) EPC at issue, novelty and inventive step, the Respondent supported the Opposition Division's assessment and disputed that they had clearly been based on misinterpretations of the claims.

(1) In particular, the Respondent maintained its previous opinion that the composition of Example 4 of D1 destroyed the novelty of the claimed subject-matter. Furthermore, the Respondent expressed its belief that the decision under appeal had been wrong, when it had held that the Tg of SBR-D could not be derived from D1. It rather believed that is was derivable that the Tg fell within the scope of D1 (paragraph bridging pages 6 and 7).

(2) Starting from a technical problem as to "furnish a tread rubber composition for tyres enabling the rolling resistance to be decreased without lowering the grip characteristics of the tyre on a wet road surface", the Respondent argued that the first property was directly measurable by the tan delta|50-70-value, and went on to assert that the problem set out in the patent already appeared to have been solved by document D1, because the tan delta|60-value in Example 4 meant even an improvement over the tan delta|50-values in Examples 1 and 2 of the patent in suit. Accordingly, the Opposition Division had to ask what problem remained to be solved and identified it as "being to have a balance between rolling resistance and braking". It could, however, find no effect upon this balance which was due to the Tg, and came to the conclusion that there had been no inventive step (comments: page 5, last two paragraphs).

(3) The Respondent continued that there had been no experimental results showing that the assertion of the Patent Proprietor "selecting the correct values of Tg can produce a better balance between rolling resistance and braking properties (presumably meaning that both are increased)" was correct (page 6, paragraph 2).

(4) In any case, Example 4 of D1 was, according to the Respondent, closer to the claimed subject-matter than either of Comparative Example 2 and 5, because Example 4 had shown the correct value of 4.8X-Y and a bimodal distribution. Moreover, its deltaT-value was to be <= 10ºC, and, as argued in the opposition proceedings, its tan delta|50-value fell within Claim 1. The poor performance in the two comparative examples of the patent in suit appeared "to be nothing to do with Tg. These examples do not demonstrate that the Tg selected solves any problem." (page 6, paragraph 6).

(5) In summary, the Respondent expressed the view that the conclusions in the decision under appeal had not clearly been based upon errors and misinterpretations. If any, they did not affect the conclusion reached.

VI. The Appellant disputed the above arguments of the Respondent in a further letter dated 18 February 2004.

(1) Thus, it emphasised that the Respondent had never shown that SBR-D of D1 satisfied all the features required in the patent in suit (such as the Tg and the deltaT-value in Example 4 of D1). The Declaration filed by the Respondent did not describe a repetition of that example, because it was based on SBR differing from SBR-D of D1 in its styrene and 1,2-bond contents.

(2) Nor had the Opponent/Respondent ever proved that the rubber composition of Example 4 in D1 showed a balance between rolling resistance and wet grip characteristics as good as the claimed rubber composition (page 6, paragraph 1). The measured grip and roll resistance values in the patent in suit and in D1 were based on different reference compositions, ie in the patent in suit on Comparative Example 5, in D1 on Comparative Example 1 (letter: page 4, paragraph 3) and could not, therefore, be compared with one another.

(3) The Respondent's assertion on the basis of individual parameter values in examples of D1 and of the patent in suit, that the technical problem had already been solved in D1, was disputed by the Appellant, who asserted inconsistencies between measured values provided in Table 2 of D1 and the content of Figure 1 of this document and pointed out that the amounts of carbon black in those compositions had been different and that tan delta|50-values would be higher than tan delta|60-values (page 5).

(4) The decision under appeal was, in the Appellant's opinion, wrong, because D1 did not provide any guidance to test "similar" rubbers (section III, above, last paragraph). D1 referred not only to SBR-D, but also to SBR-C and SBR-E, all of which met the requirements of D1 and were used in its Examples 1 (see Appendix II, above). However, neither SBR-C nor SBR-E fulfilled the "equation", contrary to the comparative rubbers SBR-F and SBR-G mentioned in D1. These latter SB-rubbers even yielded compositions having bimodal tan delta-temperature curves, but lacking sufficient wet skid resistance. Moreover, rather than suggesting to the skilled person that the Tg would have any significance, D1 did not even mention the Tg of the SBR (page 6, paragraph 4).

In the Appellant's view, the patent in suit related to a solution, quite different from that of D1, of the technical problem of achieving sufficient wet skid resistance and adequate fuel saving. This difference was demonstrated by SBR-B as used in Examples 5 and 6 of the patent in suit. This rubber had a 1,2-bond content outside the range of from 60 to 73% required in D1, but solved the problem underlying the patent in suit (page 6, last line and page 7).

VII. In further letters dated 7 May 2004 and 22 September 2005, the Respondent reiterated its objection raised under Article 100(b) EPC.

Moreover, the Respondent relied again on D5, according to which it would be possible to calculate that the Tg of -25ºC or less corresponded to the case where the ratio of 1,4-trans to 1,4-cis was 1.9:1 or less than 1.9:1. On the balance of probabilities, the ratio of 1,4-trans to 1,4-cis-butadiene would fall inside this range and, consequently, the Tg would also fall within the range defined in Claim 1. If, however, the above ratio were above 1.9:1 the calculation would give a higher Tg. "It appears to be beyond doubt that the glass transition temperature of SBRD of D1 will have a glass transition temperature which either falls within the claimed range of -25ºC to -50ºC or is above it." (last letter: page 2, lines 5 to 7).

VIII. Oral proceedings were held on 25 October 2005 in the presence of both parties. The essentials of the hearing and additional arguments can be summarised as follows:

(1) With regard to the meaningfulness of the Tg as used in the patent in suit and the disputed connection between the "equation" and the deltaT-value, the parties maintained their respective views.

(2) Thus, the Appellant argued that the Tg was often used for the characterisation of rubbers and well-known in this art. This was demonstrated eg by Claim 1 of the Respondent's own D3, requiring a polymer rubber to have a Tg of not lower than -50ºC, without any indication of how the Tg was determined. However, the skilled person would know that this parameter was to be understood as being the result of the ASTM method. Whilst D4 referred to the measurement of Tg of a broad range of amorphous and crystalline materials in general, the ASTM-method of Exhibit A was specifically designed for polymers, and from its item 10.2.2 the temperature program to be used was known. From item 10.2.7 and Figure 1, it was also clear that the extrapolated onset temperature Tf was more meaningful for most applications and "may be designated as the Tg temperature in place of the midpoint of the Tg curve". This was also confirmed by Exhibit B [page 440, chapter II.A.: "DTA and DSC provide identical information ..." and "DSC has become the preferred technique for elastomer studies"; page 444, below Table III: "TE0 is the extrapolated onset value obtained by ... (see Figure 2). TE0 is the most often quoted value for Tg and is generally reproducible ..."; NB.: TE0 corresponds to Tf in Exhibit A]. Moreover, Exhibits C and D gave the Tg-values of those commercially available SB-rubbers mentioned in the patent in suit, these values (indicated as being determined in accordance with ASTM) tallied well with the values in Table 1 of the patent in suit, and Exhibit E provided details (including a DSC curve measured on 5 July 2002) of the determination of the Tg of SBR 1500 as used in reference Comparative Example 5 of the patent in suit. Hence, Tg (in terms of a ASTM-measurement) was common general knowledge and the specification gave enough information in this respect.

As regards D5, the Appellant pointed out that the Tg was determined by dilatometry using an ethanol-water mixture (page 1582). It did not know whether the results of different methods of determination were in agreement with one another. Hence, Figure 4 on page 1586 of the document could not serve to support the Respondent's position that the G-T-equation of D5, would allow to reliably calculate the Tg of the SBR-D of Example 4 in D1. Rather, the Appellant argued that the Tg of butadiene-styrene copolymers was primarily a function of butadiene microstructure and of the amount and distribution of styrene in the polymer (page 1585, last paragraph). Neither the microstructure nor the distribution of the styrene were, however, known from D1. In the absence of this information, the G-T-equation could not, however, be used for the determination of the Tg.

(3) The Respondent stated, that it did not raise any further detailed objection with regard to Tg, and informed the Board, that no data were available to the Opponent about the relation between the results of the different experimental methods for the determination of the Tg. Nevertheless, in its view, the G-T-equation was a good model for the determination of the Tg, although the microstructure must be known when using it for the calculation of the Tg. In the further discussion, the Respondent conceded, however, not to have any information that the dilatometric method of D5 would be preferred for determining the Tg, and finally stated that it did not maintain an objection under Article 100(b) EPC with regard to Tg.

(4) With respect to the further feature of the claimed composition missing from D1, the Appellant referred to the "repetition" of Example 4 of D1 as submitted by the Opponent (section II (7), above) and argued that, although the composition in this "repetition" differed from the recipe used in the examples in the patent in suit by the SBR used (SBR-A of the patent in suit instead of SBR-D of D1), by the ratio of the two rubbers used, by the use of a different carbon black in a different amount and by omission of antioxidant, a deltaT-value had been achieved as required by the claim. Nor had the Respondent submitted that it had suffered repeated failures when carrying out a multitude of such experiments. Hence, it was evident that the patent in suit contained enough information for carrying out the claimed subject-matter in the sense of Article 83 EPC.

(5) The Respondent reiterated its initial argument concerning the deltaT-value (see section II (4), above), that the only guidance of how to obtain the required deltaT-value would be found on page 3, lines 14 to 30 of the patent in suit. Moreover, there would be an inconsistency between the information in this passage of the description, that of paragraph [0014] and the results in Table 2 of the patent in suit. The Respondent also doubted that 20 to 30 parts by weight of the SBR in the composition would work.

(6) The Appellant, however, put emphasis on the argument that the effects achieved by the claimed subject-matter could not be attributed to a single feature of a single component, but that it was the whole rubber composition (optionally inclusive of any additives), which gave the desired result. Moreover, both requirements of the SBR must be fulfilled.

(7) Having regard to the issue of novelty vis-à-vis D1, the Appellant disputed that the SBR-D in Example 4 of D1 clearly and unambiguously had a Tg of <= -25ºC. Nor could D5, in its opinion, provide the necessary information, since the microstructure of the rubber, namely the ratio of 1,4-trans- to 1,4- cis-bonds, was unknown. The Respondent, however, had only argued on the basis of assumptions, this way of argument was not, however, appropriate for the assessment of novelty.

(8) The question of whether it knew, that the ratio of trans- to cis-configuration had been 1.9:1 or less in the SBR-D of D1 (cf. section VII, above, last paragraph), was answered by the Respondent in the negative. Nor could D6, which had additionally been mentioned by the Respondent to fill this gap, provide the missing information. It was withdrawn by the Respondent with regard to novelty.

(9) With reference to inconsistencies between the data in Example 1 and Figure 1 of D1, the Appellant denied that the tan delta|50-value could be unambiguously derived from a tan delta|60-value as provided in Table 2 of D1.

Whilst disputing that Figure 1 of D1 would allow to plot back to the facts in Example 4 of the document and maintaining that the tan delta|50-70-values would not be significantly different, the Respondent then argued on the basis of the values given for tan delta|0 and tan delta|60 in Table 2 of D1 that the tan delta|50-value in Example 4 would have, in any case, fulfilled the requirement of Claim 1, ie it would have been between 0.08 and 0.20. To this end, the Respondent presented a calculation of the tan delta|50-value on the basis of a straight line provided by the two data 0.475 (at 0ºC) and 0.145 (at 60ºC). This calculated value of 0.20 would in any case be higher than the real tan delta|50-value. This result was, however, disputed by the Appellant who criticised that too few data had been available.

(10) In D1 and the patent in suit, different measuring conditions had been used in the determination of the loss factors (D1: strain: 10±2%, frequency: 20 Hz, temperatures: 0ºC and 60ºC; patent in suit: strain 10±1.00%, frequency: 10 Hz, temperature: 50ºC). It was not disputed by the parties that this had some influence of the results, but, according to the Appellant, the influence of the exact composition on the curve was higher. However, no reliable data were available in this respect to either party.

(11) Then neither party wished to comment further on novelty.

(12) Like the decision under appeal, both parties considered D1, in particular its Example 4, as representing the closest state of the art. However, the conclusions drawn by the Opposition Division, that no improvements had been achieved with regard to this document, was disputed by the Appellant, whilst the Respondent concurred therewith.

(13) In particular, the Appellant pointed out that a direct comparison between the results in Example 1 of the patent in suit with those in Example 4 of D1 was not possible due to different compositions (60 and 50 parts by weight of carbon black, respectively, and different measuring conditions). These differences were, according to the Appellant, decisive for the tan delta|50-value.

Moreover, the results of the determinations of the rolling resistance and of the wet grip characteristics tested in these examples had been given as index values relative to measurements of these features at tread rubber compositions in Comparative Example 1 in D1 and Comparative Example 5 in the patent in suit. Due to these different reference compositions, no reliable conclusions would be possible. Hence, the decision under appeal was wrong in this respect.

The relevant technical problem was seen by the Appellant in the provision of a tread rubber composition showing a rolling resistance as low as possible together with, at the same time, an excellent wet grip characteristic. This problem would have been solved by the composition fulfilling all the features as defined in Claim 1 in combination.

In D1, ranges of the styrene and 1,2-bond contents in the SBR had been defined without any hint on the importance of the relation between these features. This was confirmed, in the Appellant's view, by the examples of D1. Thus, SBR-C, SBR-D and SBR-E in its Table 1 complied with the teaching of D1. Table 2 of D1 did not, however, indicate that the use of SBR-D, the only SBR meeting the requirements of the patent in suit, would provide better results than SBR-C or SBR-E (cf. section VI (4), above). Hence, D1 did not put special emphasis on SBR-D, nor did it suggest to use SB-rubbers having a particular Tg, let alone one within the range of between -25 and -50ºC. Consequently, D1 itself could not, in the Appellant's view, make the subject-matter of the patent in suit obvious.

Like D1, D2 contained no hint either that the relation between styrene and 1,2-bond contents of the SBR used in those compositions would have any importance. D2 required only that the SBR (ie a "high vinyl amorphous butadiene-styrene copolymeric rubber") contained not less than 60 % by weight of 1,2-bond in the butadiene unit and 3 to 30 % by weight of styrene (D2: Claim 1) and recommended the Tg of this SBR to be -35 to 0ºC (D2: column 3, lines 22 to 24).

Amongst those individual SB-rubbers (denoted A to E) in Table 1 of D2 which fulfilled the above requirements and also the Tg requirement of the claims at issue (-25ºC to -50ºC), only SBR-E also complied with the "equation". Example 5 in Table 3 of the document, being the only example wherein this SBR was used in combination with a diene rubber, however, showed no advantages over any other examples. Moreover, according to D2, the high vinyl amorphous SB-rubbers described therein were compatible with natural or synthetic polyisoprene rubbers (D2: column 3, line 53; column 4, lines 26 to 30). This meant, as explained in D1, that all the examples (D2: Table 2) contained compatible blends, which failed, however, fully to produce the beneficial effect of improving the tyre grip (D1: column 1, lines 45 to 59). The Appellant concluded therefrom that D1 and D2 presented diverging solutions and provided no incentive to combine their teachings.

As in the passage of D1, mentioned in the previous paragraph, the use of high vinyl amorphous SBR, which was also addressed in D3 (page 2, line 24 et seq.), did not yet provide sufficiently "low values recently required for the rolling resistance" (D3: page 2, line 31). The solution for achieving a combination of good wet grip and low rolling resistance found in D3 was the addition of 10 to 150 parts by weight (pbw) of a silica filler, 0 to 150 pbw of carbon black, and 0.2 to 10 pbw of at least one silane coupling agent of a specific chemical formula to 100 pbw of a polymer rubber having a Tg of not lower than -50ºC obtained by polymerisation of 1,3-butadiene or copolymerisation of 1,3-butadiene and styrene with an organic alkali metal initiator, or a rubber blend of >=30 pbw of such a rubber and <=70 pbw of another diene series rubber (Claim 1). In a further alternative, the above butadiene or butadiene-styrene rubber was itself further modified by terminal silane groups. No reference was, however, made in D3 to the relation between the styrene and 1,2-bond contents. This led the Appellant to argue that the document would have guided the skilled reader to orientate itself in a completely different direction. Nor would a combination of D1 and D3, in any case, have led the skilled person to the claimed subject-matter.

(14) The Respondent saw three "candidate distinctions" between D1 and the patent in suit: the Tg and the deltaT- and tan delta|50-values. The slight deviation of the Tg of the SBR used in Comparative Example 2 (-23ºC instead of <=-25ºC) did not, in the Respondent's view, convincingly demonstrate that the Tg would have an effect, because the value of -23ºC lay, according to Exhibit A, within the error margin of -25ºC. Nevertheless, this comparative example was not accepted by the Respondent as a valid comparison to Examples 1 or 5, due to a different amount of oil (24 pbw instead of 18 or 16 pbw used in those examples).

As regards the technical problem solved by the feature deltaT-value, the Respondent argued that Example 4 of D1 was closer to the claimed subject-matter than Comparative Example 1 of the patent in suit, because its deltaT-value would not have been in the range of 20ºC, even if its SBR would have had a Tg (as calculated by means of D5) slightly higher than -25ºC. According to page 3, line 14 et seq. of the patent in suit, it could rather be expected that the deltaT-value in Example 4 of D1 was below 10ºC.

Concerning the third candidate distinction, the Respondent set out that it had been common general knowledge to reduce the carbon black content and, by doing so, the tan delta|50 in order to avoid a high rolling resistance. Therefore, the comparison of Comparative Example 6 and Example 1 of the patent in suit confirmed only common general knowledge.

In summary, the Respondent concluded that Claim 1 did not solve any technical problem with regard to D1.

Moreover, the Respondent took the view that the teaching of Claim 1 did not exclude deltaT-values close or equal to 0ºC, thus, giving rise to the question of whether there was only one peak or a couple of peaks and, thus, making the discussion about compatibility and incompatibility moot.

(15) The Appellant pointed out that the features of Claim 1 must not be considered separately, because it was their combination which led to the solution of the technical problem. The examples in the patent in suit demonstrated that their results had been clearly better than those achieved in the comparative examples.

With regard to the number of peaks in the loss factor-temperature dispersion curve of the claimed composition, the Appellant referred to the fact that Claim 1 required it to be bimodal ("having two peaks").

Finally, it took the view that the Respondent as the opponent had not discharged its burden of proof to demonstrate that its allegation had been true that no technical problem had been solved.

Then the discussion was closed, since both parties had indicated that did not want to further comment on inventive step.

IX. The Appellant requested that the decision under appeal be set aside and that the patent be maintained as granted (Main Request) or in the alternative on the basis of one of the Auxiliary Requests 1 to 3 as filed with letter of 16 August 2002.

The Respondent requested that the appeal be dismissed.

Reasons for the Decision

1. The appeal is admissible.

2. Interpretation of the claims

Before turning to the substantive matters, it appears to the Board to be necessary to interpret Claim 1 as such, in particular with regard to the meaning of the individual features defined in this claim, namely in view of objections, interpretations and statements of the parties and in the decision under appeal (cf. the decision under appeal: No. 3.2ai.3 of the reasons; sections II (1), VIII (5), (14) and (15), above).

2.1 The claim relates to a tread rubber composition for tyres comprising rubber components and optionally, according to the description, further additives (paragraph [0016]; see also the table before paragraph [0020]). The rubber components, assumed to be 100 parts by weight (pbw), consist of 20 to 80 pbw of a diene rubber and 80 to 20 pbw of one or more SB-rubbers.

Subsequently, the claimed subject-matter is further explained in Claim 1 in terms of parameters and properties, some of which are those of the SBR, whilst the others clearly concern the composition as a whole.

2.1.1 Thus, the SBR component is characterised by the Tg and the relation between the styrene (x) and 1,2-bond (y) contents required to satisfy the equation 4.8X - Y >= 32 and >= 70, respectively (sections I and II (11), above).

2.1.2 One requirement for the composition as a whole is the tan delta-temperature dispersion curve having two peaks, another one is the tan delta|50 being from 0.08 to 0.20. Furthermore, the temperature difference (deltaT-value) between (i) the peak at the highest temperature on the above dispersion curve (ie the curve of the rubber composition as claimed) and (ii) the single peak on the corresponding curve of a rubber composition differing from the above composition (as in (i)) only in that it contains the same SBR as the sole rubber component must not exceed 10ºC (section II (3), above; patent in suit: paragraph [0010]).

2.2 Hence, an interpretation of the passage relating to the deltaT-value in the decision under appeal (No. 3.2ai.3 of the reasons) reading "It was explained that D1 discloses two peaks of tan delta over temperature. However, there is no explicit disclosure in D1 of the temperature difference between the peaks." as meaning different peaks on the curve of one composition has no basis in the claim.

2.3 In order to support its objection of lack of novelty with regard to Example 4 of D1 (which discloses neither of the parameters, below), the Respondent assumed that, according to the patent in suit, (i) a Tg below -25ºC was "required to give the tan delta in the range 50ºC to 70ºC" and that (ii) meeting the requirement concerning the "equation" would automatically mean that the deltaT-value was within 10ºC (Notice of Opposition, reasons for lack of novelty vis-à-vis D1: page 3, paragraph 8 and page 4, penultimate paragraph; sections II (2) to (4), above).

This interpretation of the two passages in the description by the Respondent requires, however, that there is, in each of the two cases, a 1:1 correlation between one feature of the SBR and one particular property of the composition prepared therewith, irrespective of other features and properties.

The above interpretation of the two passages of the description, mentioned above, by the Respondent is, however, not convincing for the following reasons:

2.3.1 On the one hand, Example 1 and Comparative Example 6 of the patent in suit show the influence of a change of the amounts of carbon black and oil (ie additives in the sense of paragraph [0016], cf. section 2.1, above) on the tan delta|50 in otherwise identical compositions. In a different context, the Respondent itself argued that the skilled person had been aware of such an influence of further components on the properties of the compositions, when it criticised Comparative Example 2 as being without merit as a comparison to Examples 1 and 5, because of the different amounts of oil (section VIII (14), above).

2.3.2 Moreover, according to D1 (column 2, lines 42 to 60), the tan delta-peak on the high-temperature side of SBR having "too much 1,2-bond" (ie more than 73%, cf. lines 44/45) "tends to shift toward a low-temperature side" and to give a monomodal curve, as opposed to a blend "of an incompatible nature taking advantage of a high-temperature peak of SBR" on the grip properties. These statements in the prior art are clearly contrary to the above assumption of the Respondent concerning the deltaT-value.

2.3.3 On the other hand, however helpful it might have been in view of the above statements in the prior art, no evidence, which unambiguously supported its above assumptions, has been submitted by the Respondent, on whom the onus of proof for its allegations had lain. This burden has, however, not been discharged by the Respondent.

2.3.4 Consequently, the above interpretation, by the Respondent, of the two passages of the description cannot be taken by the Board for granted.

2.4 Nor can the Board concur with the initial suggestion of the Opponent (section II (1), above) to disregard the Tg of the SBR, because it is, as demonstrated by the Appellant (section VIII (2), above) and by D2 (Claims 1 and 2; column 3, line 22 et seq.), a parameter which is well-known to the person skilled in this art and often used to characterise rubbers, even without identifying the method of its determination, as confirmed by the Respondent's own document D3 (D3: Claim 1 and page 5, lines 17 to 19).

This point of view is also supported by the fact that the Tg-data of the commercial SB-rubbers in Exhibits C, D and E tally well with the Tg-data of the same rubbers given in the patent in suit, thus confirming the arguments of the Appellant in this respect. According to these arguments, additionally supported by Exhibits A and B (section II (10), above), the ASTM method has been the usual method of determination of the Tg in this art. The arguments of the Appellant in this respect (section VIII (2), above) are accepted by the Board.

Apart from this finding, the Board concurs with the acknowledgement in the Notice of Opposition, that this objection referred to clarity, which is not a ground for opposition (Article 100 EPC).

3. Problem and solution

The patent in suit concerns tread rubber compositions comprising blends of diene rubber and SBR and having two peaks on their tan delta-temperature dispersion curve.

3.1 Document D1 discloses such a composition of "an incompatible nature" comprising (a) natural rubber or synthetic polyisoprene rubber or both having a cis content of not less than 80% and (b) SBR having a styrene content of from 5% to 50% and a 1,2-bond content in the range of 60% to 73%, the SBR resulting from solution polymerisation in the presence of an organic lithium compound. The known composition has a bimodal loss factor-temperature curve as determined by dynamic viscoelasticity (Claim 1; column 2, lines 3 to 14 and 58 to 60). The importance of the 1,2-bond content on this curve and on the tyre properties have already been addressed in section 2.3.2, above.

3.1.1 The document aims at the provision of a rubber composition suitable for tread formation which excels in skid resistance on dry and wet roads and also in fuel economy by providing a good balance of loss factors, higher tan delta at about 0ºC and lower tan delta at from 50 to 70ºC, at from 10 to 30 Hz (D1: column 1, line 66 to column 2, line 2 in conjunction with column 1, lines 32 to 37 and column 2, lines 56 to 68).

3.1.2 In its examples and comparative examples, seven different SB-rubbers were used, denoted SBR-A to SBR-G. All these SB-rubbers had a styrene content within the above range of 5 to 50%; however, only SBR-C, SBR-D, and SBR-E fulfilled the other requirement of D1 of 60 to 73% of 1,2-bond content:

FORMULA/TABLE/GRAPHIC

Consequently, Examples 1 to 7 of D1 were based on the use of these SB-rubbers (see Table 2, below). However, from Table 1, it can furthermore be derived, but only in the knowledge of the patent in suit, that the condition of the "equation" was only fulfilled by SBR-D (46.4), and comparative rubbers SBR-F (66.0) and SBR-G (87.0).

FORMULA/TABLE/GRAPHIC

The wet skid and abrasion resistances in the table were evaluated relative to the composition of Comparative Example 1, wherein the reference values for these two properties had been set to "100" and whereby, in the examples and all the other comparative examples, a higher value represented a better result, and vice versa.

Further details of the examples are not disclosed in D1. Nor has it been disputed between the parties, that in D1 no mention is made at all of the Tg and of the requirement, that the deltaT-value must not exceed 10ºC. The issue of whether the tan delta|60-values in the table can validly substitute the tan delta|50-value required in Claim 1 will be dealt with below.

3.2 According to paragraph [0008] of the patent in suit, the technical problem to be solved by the patent in suit may be seen in the provision of a tread rubber composition for tyres enabling the rolling resistance to be decreased without lowering the grip characteristics of the tyre on the wet road surface, whilst, up to then, it had not yet been possible to obtain a "tread rubber both lowering the rolling resistance and enhancing the grip characteristics on the wet road surface" at the same time (patent in suit: paragraphs [0003] to [0007]). In the oral proceedings before the Board, the Appellant saw the technical problem in the provision of rubber composition for tyre treads showing, at the same time, (i) a rolling resistance being as low as possible and (ii) an excellent wet grip characteristic.

3.2.1 As mentioned in section III (3), above, the technical problem underlying the claimed subject-matter had been reformulated in the decision under appeal as relating to the provision of further rubber compositions, because, according to the decision, no convincing evidence had been provided to show that a superior balance between rolling resistance and braking had, in fact, been achieved. Moreover, the Respondent even disputed that that there had been a technical problem at all, which had been solved by any one of the features which it addressed in the oral proceedings as "three candidates for a distinction", ie the Tg, the deltaT- and the tan delta|50-values. Rather, in its opinion, even better tan delta|50 results had been obtained in D1 than in the examples of the patent in suit.

3.2.2 In the patent in suit, the experiments (the results of which are shown in Table 2, below) had been based on SB-rubbers which had the following particulars:

FORMULA/TABLE/GRAPHIC

(This version of Table 1, which differs from the version in the printed patent specification only by the headings of the last two columns, was taken from the application as published, in order to avoid the printer's error in the printed patent specification.)

The further ingredients of the compositions exemplified in the examples and comparative examples of the patent in suit were natural rubber, carbon black and aromatic oil in amounts indicated in Table 2, below, and the following amounts of further additives in terms of parts by weight: stearic acid 2, ZnO 3, antioxidant 1, accelerator (CBS as in D1, above) 1 and sulphur 1.75 (patent in suit: the table at the top of page 4).

The indices of the "Characteristics of tire" in Table 2 are based on the reference value, set to 100 for each of the grip property and the rolling resistance in Comparative Example 5. A higher value represents a better result, and vice versa.

FORMULA/TABLE/GRAPHIC

3.2.3 This means, however, that different compositions were used as reference for the evaluation in the experimental results in the respective Tables 1 and 2 of D1 and the patent in suit (sections VI (2), 3.1.2 and 3.2.2, above). Additionally, the Appellant underlined that the experimental results in D1 and the patent in suit could not be compared, because of the different amounts of carbon black having a significant effect on the properties of the respective compositions, namely higher tan delta|50-values (and higher rolling resistance) resulting from higher contents of carbon black. Neither of these facts (different reference, different contents of carbon black) has been disputed by the Respondent.

Consequently, comparing an index value of a given relevant characteristic in Table 2 of D1 with an index value in Table 2 of the patent in suit cannot give a meaningful result. This becomes even clearer when comparing the SB-rubbers in both reference examples, ie SBR type "A" of D1, which was a copolymer "derived from solution polymerization with use of alkyl lithium catalyst", for which X- and Y-values and the Mooney viscosity are given, but no Tg, and "SBR 1500" (SBR-G) in the patent in suit. As very clearly shown in the Opponent's Appendix II (section II, above), these SB-rubbers are significantly different from each other.

3.2.4 Furthermore, the compositions used in the examples of D1, namely in its Example 4 to which particular reference had been made by the Respondent, differ from those in the examples in the patent in suit not only by the amount of carbon black (section 3.2.3, above), but also by the absence of an antioxidant in D1. Nor does D1 provide any information as to the Tg of the SB-rubbers used therein (sections 3.1.2 and 3.2.2, above).

3.2.5 Therefore, the discussion between the parties of whether Figure 1 of D1, as argued by the Appellant, indicated that the determination of the tan delta of such compositions at 60ºC tended to result in lower tan delta-values than the corresponding tan delta|50-values or, in the end, whether tan delta|50- and tan delta|60-values can directly be compared with each other is moot.

Even though the amount of the carbon black in the composition, according to the Appellant, has a greater effect on the resulting tan delta|50-values, this assessment is further complicated by the fact that the measurements of the loss factor in D1 and the patent in suit were furthermore carried out not only at different temperatures, but also at different frequencies and with different strain (section VIII (10), above; patent in suit: paragraph [0025]; and D1: column 4, lines 19 to 23).

3.2.6 In view of these facts, arguments and findings, the Board takes the view, that there is no common ground between the experiments in D1 and in the patent in suit which would have allowed to compare the data of the tyre characteristics as given in both Tables 2, above.

3.2.7 In view of this lack of comparability of the above experimental results, the argument of the Respondent that no technical problem was solved by the patent in suit, because the desired results had already been obtained in D1, is not convincing and, consequently, cannot be accepted by the Board.

Nor, in view of all the differences mentioned above, can the Board accept that Example 4 of D1 would allow to draw any conclusions to the effect that the relevant technical problem as considered in the patent in suit (section 3.2, above) was not solved by the subject-matter of Claim 1 of the patent in suit.

3.2.8 On the contrary, the patent in suit itself contains comparative examples which can be compared with its examples and which demonstrate the influence of the features defined in Claim 1. Thus, the SB-rubbers used in Comparative examples 1, 3 and 4 did not fulfil the "equation", those used in Comparative examples 2, 4 and 5 had a Tg outside the required range. Comparative example 6, although based on the same constituents as Example 1 gave too high a tan delta|50-value due to the higher contents of carbon black and aromatic oil. In each of these comparative examples the results were inferior to those in the examples of the patent in suit.

Moreover, at least Comparative Example 5 can be seen as a comparison to the usual prior art tyres on the basis of commercial SBR (see paragraph [0031] of the patent in suit).

Therefore, the Board cannot concur with the Opposition Division and the Respondent that no evidence for the solution of the technical problem as suggested by the Appellant and by the patent in suit had been provided.

3.2.9 Rather the Board takes the view that it would have been the Respondent who, as the Opponent, should have demonstrated, eg by providing data comparable with each other, that no improvement had been achieved by the claimed subject-matter over the properties of D1. However, this burden of proof has not been discharged by this party.

3.2.10 Thus, the experimental report (Declaration of Mr Nakamura; section II (7), above) submitted by this party "In order to test the formulation of Example 4 of E1" (= D1) was not based on the SBR-D as defined in D1 (section 3.1.2, above, Table 1), but on the "commercially available material Nipol NS116 available from NIPPON ZEON Co LTD", see page 2, items A.6 and B.1 of the "WRITTEN SUBMISSIONS …" accompanying the report and the copy of "Document: 847544" also filed therewith and mentioned in item 6 of the Declaration:

FORMULA/TABLE/GRAPHIC

In view of the inconsistencies concerning (i) the nomenclature of the SBR used in the Declaration (ie items 4 and 5: NS116; item 6: NS-116R, respectively) and (ii) the Tg-values reported, on the one hand, for NS-116R in this document (-25ºC) and, on the other hand, for NS116 in the patent in suit (-33ºC) (section 3.2.2, above), the Board can, at most, assume that the SBR used in the experiment of the Declaration was identical to SBR-A as identified in the patent in suit. This assumed identity has not been disputed by either party. From the additions made by the Opponent (on the basis of Table 1 of D1) in Appendix II (section II, above) to the Figure 8 of the patent in suit (which was based on Table 1 of the patent in suit), it is, however, evident that SBR-A of the patent in suit was clearly different from SBR-D used in Example 4 of D1 (cf. section 3.2.3 and both Tables 1 in sections 3.1.2 and 3.2.2, all as above).

3.2.11 Consequently, the experiment in the Declaration cannot provide any features to remedy the deficiency of Example 4 of D1 in respect of missing data relative to the compositions as defined in the patent in suit (viz. the Tg of the SBR-D used therein), let alone support the Respondent's view that the claimed subject-matter would not solve any technical problem which had not yet been overcome by D1 (sections 3.2.1 and 3.2.7, above).

3.3 In view of these facts and findings, the Board has, therefore, come to the conclusion that the technical problem as suggested by the Appellant has credibly been solved by a composition as defined in Claim 1 of the patent in suit. Consequently, the relevant technical problem to be considered in the assessment of inventive step is the one, mentioned in section 3.2, above.

Main Request

4. Novelty

4.1 Although having raised novelty objections in the Notice of Opposition with regard to D1, D2 and D3, the Respondent neither commented on the findings in the decision under appeal that neither D2 nor D3 anticipated the claimed subject-matter, nor did it maintain the novelty objections with respect to D2 and D3 in reply of a question to this end in the oral proceedings before the Board. The Board does not see any reason to take a different view with regard to D2 and D3 either.

4.2 The objection of lack of novelty with regard to D1 was, however, further pursued by the Respondent, who additionally referred to the G-T-equation of D5 (section VII, above, paragraph 2), in order to remedy one of the deficiencies of D1, as mentioned in the decision under appeal (section III (2), above,) ie the missing disclosure of the Tg-value. In view of the arguments of the parties in sections VIII (2) and (3), above, and of the Respondent's own statements in section VII, above, paragraph 2, the G-T-equation does not, however, provide, on the basis of the available data in D1, the information, that the Tg-requirement of Claim 1 would clearly and unambiguously be met by the compositions of D1.

Rather, the Respondent has never shown that any one of the SB-rubbers in D1, in particular SBR-D in its Example 4, fulfilled all the requirements of Claim 1 under consideration (cf. section VI (1), above).

Nor has evidence been presented by the Respondent, who as the Opponent had the burden of proof for its assertions, that the requirement of deltaT being < 10ºC in Claim 1 was inevitably fulfilled by a tread rubber composition comprising SBR in compliance with the "equation". Any such statement would, anyway, be contrary to statements in D1 (see section 2.3.1, above).

Nor has the Respondent disputed that SBR-A, as used in some examples of the patent in suit and as a substitute for "SBR-D" of Example 4 of D1 in the Declaration to provide the above missing features, is different from SBR-D (sections II (7) and 3.2.10, above). This finding alone renders the declaration useless for the purpose of proving lack of novelty by Example 4 of D1.

Consequently, neither D1 as a whole nor its Example 4, in particular, anticipates the subject-matter of Claim 1 of the patent in suit.

4.3 In view of these findings, in combination with those in section 4.1, above, it follows, therefore, that the subject-matter of Claim 1 meets the requirements of Article 54 EPC. By the same token, this is also valid for the elaborations as defined in the dependent claims.

5. Inventive step

It remains to be decided whether the solution of this problem, as claimed, derives in an obvious way from cited documents.

5.1 Like the parties and the Opposition Division, the Board also considers D1 as representing the closest state of the art. In particular, both parties relied on Example 4 of the document, because it was the only passage describing a tread rubber composition having a bimodal loss factor-temperature curve and including SBR fulfilling the condition of the "equation".

5.1.1 Besides SBR, a second rubber component, which is natural and/or synthetic polyisoprene rubber, is required to be present in the composition as defined in Claim 1 of the document. The specific requirements to be met by this known composition are the second rubber component to have a cis content of not less than 80%, and the composition as a whole to have a bimodal loss factor-temperature curve as determined by dynamic viscoelasticity (as opposed to a monomodal curve meaning "a curve designating a compatible rubber blend"; D1: column 2, lines 47/48). As already pointed out in the decision under appeal and not disputed by the parties, D1 is, however, absolutely silent about the Tg, the deltaT- and the tan delta|50-values.

Moreover, as shown in sections 3.2.3 to 3.2.6, above, the compositions of Example 4 in D1 and of Example 1 in the patent in suit were also different and, in each of the patent in suit and D1, the values used to demonstrate the wet grip characteristic and rolling resistance of the subject-matter in accordance with the respective claims were given as index values relative to different reference examples. Hence, there has been no common ground for a direct and meaningful comparison of the experimental results in the patent in suit and in D1 (cf. also the Appellant's comments in section VIII (13), above).

5.1.2 On the basis of these findings, the Board has come to the conclusion that there was neither a basis for replacing the technical problem as suggested in paragraph [0008] of the patent in suit (section 3.2, above) by the less ambitious version as used in the decision under appeal (section III (3), above), nor a basis for the assumption that the relevant technical problem had not been solved (sections 3.2 to 3.3, above).

5.1.3 The Respondent approached the question of inventive step on the basis of an analysis of features, identifying "three candidate distinctions" between D1 and the patent in suit (section VIII (14), above). Such an approach is, however, clearly based on inappropriate hindsight.

5.1.4 Moreover, D1 provides no suggestion that, by modification of certain features of its composition, a solution to the relevant problem might be found, let alone is any hint derivable therefrom as to which of the features of the known composition could or should be modified for this purpose.

In particular, nowhere in D1, can the slightest hint be found that a Tg of the SBR within a particular range and in combination with the further properties required by Claim 1, viz. a particular ratio between the styrene and 1,2-bond contents in accordance with the "equation", thereby avoiding a deltaT-value of >10ºC, would be essential for the solution of the relevant technical problem. Nor does Example 4 of D1 provide additional information to this end (cf. sections VIII (4), (5) and (8), above). For the reasons given in sections 3.2.7 and 3.2.8, above, it provides, in the Board's view, even less information than the comparative examples of the patent in suit.

Hence the Respondent's argument that this example would be closer to the claimed subject-matter than any one of the comparative examples in the patent in suit is neither convincing, nor does it change the situation that certain features essential to the claimed subject-matter are clearly missing from D1. Nor are these features foreshadowed in or by this document.

5.1.5 Therefore, D1 itself does not render the solution of Claim 1 obvious.

5.2 As argued by the Appellant (section VIII (13), above), D2 does not disclose, either, that the relation between the styrene and 1,2-bond contents of the SBR might be important for the solution of the relevant technical problem.

5.2.1 Apart from the additional presence of dienic rubber having a Tg <-60ºC, D2 requires the SBR to contain styrene in amounts of from 3 to 30, preferably not more than 10 % by weight (x) and not less than 60, preferably more than 80 to 95 % by weight of 1,2-bonds (y). When considering these preferred ranges, it becomes clear that this document points in a specific direction, namely to the use of "high vinyl" SBR. More particularly, D2 teaches clearly away from the solution of the patent in suit, because the requirement of the "equation" requiring a certain minimum X-value for each Y-value or, vice versa, a maximum Y-value for each X-value, cannot be derived from the document.

5.2.2 Such high vinyl or vinyl-rich SBR is referred to in D1 as suffering from deficiencies in respect of the rolling resistance and wet grip, viz. "because too much 1,2-bond renders the rubber blend compatible" (D1: column 1, lines 45 to 59; section 2.3.1, paragraph 2). Such blends give monomodal loss factor-temperature curves and reduced grip expressed in terms of a reduced tan delta|0 (D1: column 2, lines 46 to 55). Reference can be made in this context to the patent in suit (page 3, lines 14 to 20).

5.2.3 In these circumstances, the Board cannot see any convincing argument for why it should follow plainly and logically from D2 that and how the teaching of D1 should/could be modified in order to solve the relevant technical problem, let alone for achieving this goal by means of something falling within the terms of Claim 1.

5.2.4 The above finding in respect of D2 is further supported by D3 which, like D1, discourages to use vinyl-rich SB-rubbers for this purpose (page 2, lines 24 to 31).

5.3 Instead, D3 suggests a solution completely different from the one in Claim 1 of the patent in suit for obtaining tread rubber compositions having good rolling and wet grip properties.

5.3.1 As referred to by the Appellant and undisputed by the Respondent, its solution was based on the mandatory use of particular amounts of specific additives in combination with one or two rubbers having a Tg not lower than -50ºC (see last paragraph of section VIII (13), above). The rubber(s) may comprise an (furthermore optionally silane-modified) SBR (page 3, lines 12 to 19), which was not required to have a Tg within the range of the patent in suit, nor was the composition required to fulfil the requirements of the "equation", of the deltaT- and of the tan delta|50- values as defined in the patent in suit (cf. the tables of its examples).

5.3.2 In view of this teaching, the Board cannot see which features of D3 should be used in isolation from the other mandatory features of its claimed subject-matter and contrary to its teaching to modify the teaching of D1. It follows therefrom, that this document does not provide any incentive either to modify the composition of D1 in a particular way in order to solve the relevant problem and so to arrive plainly and logically at something within the terms of Claim 1 of the patent in suit.

5.4 No arguments on the basis of the other documents D4 to D6 have been provided by the parties with regard to inventive step. Nor does the Board see any reason why any one of these documents should be relevant in this respect.

5.5 In view of the above facts, arguments and findings, the Board has, therefore, come to the conclusion that the subject-matter of Claim 1 is based on an inventive step.

By the same token, this is also valid for the elaborations in the dependent claims.

5.6 Consequently, the requirements of Articles 52, 54 and 56 EPC are met by the Main Request.

6. Article 100(b) EPC

As mentioned in sections II (4) to (6), above, the basis for this ground for opposition are conditional objections of the Respondent in support of its novelty objection to the effect that, if it was found that tan delta depended not only on the Tg and if it was not accepted the deltaT-value is the inevitable result of the compliance with the "equation", then there was insufficient disclosure.

These arguments relate to the meaning of explanations given in the description of the patent in suit (which has already been dealt with in sections 2 to 2.3.4, above), rather than to the practicability of the claimed subject-matter.

Moreover, the Declaration (sections II (7) and VIII (4), above) confirms, in the Board's view, that the patent in suit provided ample information to carry out the claimed subject-matter, because it shows that the Respondent did not have any difficulties in carrying out an experiment in accordance with the teaching of the patent in suit.

This ground for opposition is, therefore, rejected.

7. Auxiliary Requests

Since the Main Request of the Appellant is successful, there is no need to consider the Auxiliary Requests.

ORDER

For these reasons it is decided that:

1. The decision under appeal is set aside.

2. The patent is maintained as granted.