T 0192/82 (Moulding composition) of 22.3.1984

Summary of Facts and Submissions

1. The oppositions filed against European Patent No. 1625 (Application No. 78 101 148.1) were rejected by the decision of the Opposition Division of the European Patent Office dated 29 June 1982. The decision was based on the sole claim worded as follows: "Thermoplastic moulding compositions of (a) 25 to 95% by weight of a graft polymer of 70 to 30% by weight of a mixture of styrene (95 to 50% by weight) and acrylonitrile (5 to 50% by weight) on 30 to 70% be weight of polybutadiene or a butadiene-styrene copolymer, and (b) 5 to 75% by weight of a terpolymer of acrylonitrile, maleic acid anhydride and styrene, characterised in that the terpolymer contains 10 to 30 parts by weight of acrylonitrile, 7,5 to 15 parts by weight of maleic acid and anhydride, and 82.5 to 55 parts by weight of styrene, in copolymerised form and has been produced by continuous bulk polymerisation in an ideally mixed tank reactor under stationary conditions and with incomplete conversions of from 25 to 60 mole%, with volume time yields of from 200 to 2000 g/l and at temperatures of from 60 to 150°C in the presence of from 0.01 to 0.5% by weight, based on monomers, of an initiator decomposing into radicals with a decomposition rate constant at 100°C of greater than 5 x 10-3 sec.-1 and, in another continuous process step, has been freed from the residual monomers to a content of less than 0.1% by weight, based on the terpolymer."

II. (a) The reason given for the refusal of the opposition was that the subject-matter of the claim was new and inventive. The claimed mixtures of graft polymers and terpolymers represented a selection from the moulding compositions covered in DE-B-1 949 487 (1). Examples 25 and 27 in the cited document described mixtures in which either the graft polymer or the second polymer compound was of the kind used in the patent under appeal, but each of these components was mixed with further components different from those specified in the patent. (b) The Opposition Division accepted that the additional citations revealed some of the other conditions relevant to the invention claimed in the patent, but nevertheless decided that none of the disclosures had taught the general concept involved therein. The problem with which the invention was concerned was to provide thermoplastic moulding compositions with improved heat deflection resistance, better flowability and processability, whereby in particular the former should be increased by a least 12°C, without loss of impact resistance, when compared to the corresponding mouldings composed of graft polymers and styrene-acrylonitrile (S-AN) copolymers (column 4, lines 6 to 11). The selection according to the patent had involved a considerable restriction of the available choice in document (1), and none of the disclosures available in the state of the art had ever suggested the criteria of selection which were now recommended in the patent. (c) In addition, the claim also required that the terpolymer be prepared by a process in accordance with some specially selected conditions which afforded particularly homogeneous terpolymers. The presence of an unexpected effect, which had been shown by the applicants in the substantive examination, had never been refuted by the opponents. There had been no hints in the prior art that the suggested moulding compositions could be further improved. The instructions for preparing the second polymer component in the citation (1) were rather concerned with certain bispolymers without specifically mentioning the methods for the preparation of terpolymers. There was even a hint that adding the optional third component copolymer to the mixture of graft polymer and copolymer could be of particular advantage. (d) Although several documents in the state of the art described a number of the required processing conditions for the second polymer, and it was known that the increase in the maleic anhydride component therein would improve its heat deflection resistance, the specific choice of the selective requirements was in no way obvious to the skilled person. The argument that the comparative test was not informative since it had been carried out under extreme conditions and under circumstances where a negative result would have been expected in any case, i.e. under the conditions of DE-A-2 343 871 Example 2a, was rejected by the Opposition Division on the grounds that the opponent should have provided comparisons which fell nearer to within the claimed range to show that different results would have been obtained under those conditions.

III. One of the opponents filed an appeal against the decision on 30 August 1982 with payment of the fee, and submitted a Statement of Grounds within the prescribed time limit. The respondents, i.e. the patent proprietor, filed his reply within the term provided.

IV. The Board expressed its own concern about the patentability of the claim in a communication to the parties and both filed their observations. Oral proceedings took place on 22 March 1984.

V. The appellants argued in their submissions and during the oral proceedings substantially as follows: (a) The main citation (1) discloses the kind of compositions which are also the subject of the claim under appeal and the only remaining relevant difference is a set of process conditions for making the terpolymer components. Some of these conditions were already envisaged in document (1) and in the article published by Hanson and Zimmerman (Ind. Eng. Chem. 1957, 49, 1803-1807)(4) referred to therein, and other conditions are described in further citations (e.g. DE-A-2 513 253(5)). It is within the knowledge of a skilled person to use and select values within such conditions to obtain improved, i.e. optimised, results. To use such improved terpolymer material must be obvious for the suggested purpose. This argument is also supported by the reasoning of the decision of a Board of Appeal ("Electromagnetically-operated switch / ALLEN-BRADLEY", T 21/81, OJ 1/1983, 15-21). (b) The comparative example used by the patent proprietor to show an inventive step was carried out on the basis of DE-A-2 343 872 Example 2a, which employed a very prolonged dwell time (40 hours), although it was clear from the same document that such a condition would lead to a mixture of incompatible products. As bad results had been expected with such extremely low volume-time throughput rates (Raumzeitausbeute), the comparison was meaningless.

VI. The respondent submitted during the proceedings before the Board substantially the following arguments: (a) Contrary to the wide range of possibilities represented by the compositions in document (1), the patent under appeal was restricted to two-component mixtures of a specific graft polymer with a terpolymer. Thus the optional addition of a third acrylonitrile-containing component was excluded, whilst the second component could not merely be a S-MA bispolymer of styrene and maleic anhydride but must also contain acrylonitrile to form a terpolymer (S-AN-MA). In addition, the ranges for the proportions of the ingredients for the terpolymer were restricted and the material must be prepared in accordance with strict conditions. Such a terpolymer component improved the properties of the moulding composition considerably. (b) The instructions in document (1) were insufficient to indicate how the terpolymers were to be prepared. Copolymers which contained less maleic anhydride than styrene were not easy to prepare and the reference to the Hanson/Zimmerman article (4) was only relevant to bispolymers containing such compositions in various proportions and not to terpolymers. It was therefore difficult to know exactly how relevant terpolymers in the cited art were to be reproduced in the first place. It was a fair assumption that the suggested continuous polymerisation would follow the model of Example 2 of the specification DE-A-2 343 871 as far as the dwell time was concerned. This was in order to demonstrate the essentiality of the suggested specific conditions for preparing terpolymers according to the invention, in comparison with the prior art. No other relevant close prior art could be identified for the purpose. (c) Although it was never the intention of the patent proprietor to protect the process of preparing the homogeneous terpolymer component and the resulting terpolymer itself, there was no reference in the state of the art which fully anticipated them. Even if the terpolymers were not novel, the provision of improved moulding compositions with them was undoubtedly inventive. (d) Document (5) gave no hint whether or not the terpolymer prepared according to a different, three-stage, process was suitable for admixture with graft polymers. Although the disclosure suggested the improvement in the relevant properties of the product it was silent as to the basis of the comparison. No conclusions could be drawn from the preparation of terpolymers and their properties as to the behaviour of the products used in the present case.

VII. The appellant requested that the decision of the Opposition Division be set aside and the patent be revoked. The respondent requested that the appeal be rejected and the patent maintained. Alternatively maintenance of the patent with a specifically amended claim was requested. Finally, as a further alternative, the respondent requested that the board specify what experimental results should be submitted on his part in support of the existence of an inventive step.

Reasons for the Decision

1. The appeal complies with Articles 106 to 108 and Rule 64 EPC and is, therefore, admissible.

2. The suggested problem with which the invention was concerned was to provide polymer compositions with improved heat deflection resistance and better flowability, i.e. processability, without the loss of impact resistance, when compared to moulding compositions of the state of the art which contain the same graft polymer and a S-AN bispolymer (i.e. styrene with acrylonitrile)(cf. lines 6 of 11 page 4 of the patent). The proposed solution according to the claim under consideration comprised the addition of a maleic anhydride constituent to certain bispolymers to provide a terpolymer having its ingredients in specified proportions and to prepare this material according to a process which is characterised in essence by the following requirements: (i) the use of continuous bulk polymerisation , (ii) in an ideally mixed reactor, (iii) under stationary conditions, (iv) to complete conversion (25 to 60 mole%), (v) at a specified volume-time through-put rate (200 to 2000 g/l.h.), (vi) at a temperature between 60 to 150°C, (vii) in the presence of a specified kind of catalyst (0.01 to 0.5 w/w%), (viii) and the use of a further step to remove residual monomers to a content less than 0.1 w/w%.

3. It is admitted by the respondent that the claimed compositions fall within the general scope of moulding compositions described and claimed in document (1). The cited art also relates to the problem of improving the heat deflection resistance and the mechanical properties, i.e. processability, of such compositions (cf. page 2, 2nd paragraph). Since the patent under appeal aims at the further improvement of properties of the same kind, the selection must be associated with some unexpected advantage in this respect. In addition to this, the cited art (1) also discloses a specific combination of a terpolymer in Example 25, and the subject-matter of the present appeal must therefore be construed as a modification of such material. The modification should result in some unforeseen improvement in the properties of the composition so constituted. Nevertheless, the combination of features which restrict the choice of components and requires a particular process for the preparation of the terpolymer imparts novelty to the claimed compositions.

4. The allegation on the part of the respondent that the cited document (1) gives insufficient instructions as to how to provide the suggested terpolymer products cannot be accepted by the Board. To show that a disclosure in the state of the art is not an enabling one involves a heavy burden of proof. There was no suggestion in (1) that the preparation of such polymers was problematic except when reduced amounts of maleic anhydride had to be combined with large amounts of styrene. The disclosure recommended the use of a continuous mixing polymerisation for such instances, which involves vigorous mixing at high temperature with a steady removal of the amount of the product which corresponds to the input. Furthermore the reference in this respect to the Hanson- Zimmerman article (4) revealed that a well-controlled copolymerisation of styrene with maleic anhydride could be achieved in various proportions up to 45% of the latter. The copolymerisation of styrene with acrylonitrile is also discussed in the document. It would have been within the ordinary skill of a polymer technologist to envisage the analogous conditions for a process with all three necessary ingredients together. The respondent submitted no convincing arguments which would suggest that the skilled person could not have applied and adapted the Hanson-Zimmerman teachings for the purposes of making a terpolymer to be used as suggested in document (1).

5. In view of the considerable literature which was concerned with the technique of colopymerisation in general at the relevant time, it cannot be assumed that there was insufficient common knowledge available to render document (1) fully operative in 1971, and even less that the necessary information for making the required terpolymer was still unavailable in 1977, the respondent's priority date. This means that the particular products exemplified in the cited document (1), together with their reported properties, must be construed as being part of the state of the art at the priority date of the patent under opposition, in the absence of any relevant evidence from skilled practitioners to the contrary. The general teaching of the document, including the presentation of graft and terpolymer mixtures as compositions with a range of heat resilience and impact resistance properties, must therefore be assumed to be disclosed and known.

6. The respondent has neither claimed the class of terpolymers used in the formation of their moulding compositions nor the process for their making. Consequently no search was carried out to establish the novelty or the inventiveness of these features on their own. The patentability of the composition cannot, therefore, be based on the assumption or allegation that these features, as components, already possess the necessary requisites in this respect.

7. Since the terpolymer component of Example 25 of document (1) comprises S-AN-MA in a proportion of 62.8:27.1:10.1, it falls well within the area of variation in this respect which the claim under appeal represents. The difference lies partly in the particular graft polymer which is required by the said claim and partly in the use of particular processing conditions specified therein. The particular graft polymers preferred for the new compositions nevertheless fall within those also recommended in the general description of (1).

8. Notwithstanding this, it is impossible to envisage a patentable selection solely on the basis of differences in the graft polymer. Document (1) emphasises that "Component A is responsible for the good heat deflection resistance of the mixtures". Component A was the bis- or terpolymer and not the graft polymer of the mixture. Selection on the basis of the graft polymer was never even mentioned, let alone argued, although formally there was a clear restriction of the scope within the variants available according to document (1). The applicants themselves stated (cf. letter of 18.2.82) that "This question finally comes down to examining whether a skilled person could expect the particular terpolymers B of our application to improve significantly the properties of the moulding composition".

9. The unexpected properties of the claimed compositions must therefore be in consequence of the use of the specific processing conditions for the preparation of the terpolymer component alone. The nearest state of the art, Example 25 of document (1), discloses a composition of a graft polymer (42%) and a terpolymer (58% S-AN-MA 62.8:27.1:10.1) with a Vicat number (for heat deflection resistance) of 106°C and impact resistance of 87 (at 26°C). The best example of the patent, on the other hand, Experiment 3 of Table 1, represents a graft polymer (35%) and terpolymer (65% S-AN-MA 64:24:12) with a Vicat number of 110°C and an impact resistance of ca. 92. The rest of the examples in the patent are apparently worse than the nearest state of the art in both respects. Whether or not the slight increase in the figures for the best example was, in any case, in consequence of the somewhat higher terpolymer and maleic anhydride contents is irrelevant since the differences are hardly significant, let alone substantial.

10. Even if the shift from the specific example of the state of the art to the now claimed compositions were to represent some real improvement in quality, the question arises whether this is unexpected in view of knowledge already available in the state of the art. As regards the measures taken within the selection range, the results are well within the fluctuation range of properties disclosed in document (1). Vicat values up to 110°C and impact resistances up to 93 were already achieved and it was already well established in the cited art that the presence of the maleic anhydride as an ingredient in the bis- or terpolymer component increased the Vicat value, whilst the acrylonitrile contributed to the impact resistance. Such capabilities are already manifest in the bispolymers with styrene (cf. top paragraph on page 4, and examples 1 to 5 in comparison with 7 and 8). A balanced compromise with regard to the incorporation of both contributors was called for in the light of the general explanations in the document and the results demonstrated in the examples. The example with the terpolymer in the cited document was already clearly one of the best compositions disclosed in this respect.

11. In view of the statement in document (1) which suggested that the particular properties of the component to be added to the graft polymer might indeed be "maintained", it was already reasonable to look for material for admixture which showed impressive characteristics in this respect. It is therefore highly relevant that the specification (5) discloses a terpolymer which not only falls within the "selected" range of proportions in the present case as far as the three ingredients are concerned but also possesses an outstanding quality. According to Example 1, the product (first stage S-AN-MA 55.3:29.7:15, final stage 57:28.5:14.5) shows a Vicat number of 124°C, an impact resistance of 20 and a notch impact resistance of 3.0, which is a better overall result than anything disclosed in the present patent (cf. the properties of the terpolymers in Table 1(a) to (c)). The respondent also argued that the quality of the homogeneous terpolymer component is the primary cause of the relevant improvement, if any, in the composition. It appears that appropriately selected improved homogeneous terpolymers were generally and particularly available from the above document in the respondent's own name, ready for incorporation in any moulding composition according to document (1).

12. It is no less significant that the disclosure for obtaining the above homogeneous terpolymer includes without exception all those critical features which are now in the claim as essential for the preparation of the terpolymer. In particular, the main claim for a process in document (5) expressly refers to conditions (i), (ii), (iv), (vi) and (viii), listed above in paragraph 2, as further characterising the process according to the present patent. In addition, condition (iii) is explained on page 13, last paragraph, in the same document. Again, the volume-time throughput rate range (v) overlaps substantially that directly implied by the dwell time range specified in the claim of document (5), and the character and quality of the initiator (viii), e.g. tert-butylperoctoate or benzylperoxide in the present patent, correspond to those recommended in the cited art (cf. last paragraph page 23 to line 3, page 24). It is very apparent that the critical conditions of the state of the art are re-employed in the present case.

13. It is irrelevant that the cited process includes two further steps and thereby additional conditions which might have also contributed to the improvements in the quality of the material. The claim in the present appeal might be construed as an open definition which does not necessarily exclude any added features or conditions from the process, as long as these are not inconsistent with the requirements of the definition. Even if the processing conditions in the respondent's claim in their patent were not directly transferrable to the product of Example 1 of document (5), the arbitrary shortening or extending of the list of requirements for obtaining various grades of improved terpolymers cannot alter the fact that such kinds of terpolymer was already disclosed in the state of the art. Unless an in itself patentable terpolymer is incorporated in the composition - notwithstanding the possible merits of improving or simplifying any technology leading to already known products - the use of known or obvious terpolymers to provide a known type of combination product must itself demonstrate an unexpected effect to render the composition patentable on a selection basis.

14. The basic prior art established the possibility of moulding compositions, and this general teaching covers existing and future combinations of graft polymers with bis- or terpolymers of the kind specified. Whilst improved new terpolymers may, as suggested above, still be patentable as such and in combination with graft polymers as a selection, the same would not be possible after such terpolymers became known or obvious. If an article is known as a combination product or mixture of components fulfilling known functions, the generation and application of an improved novel component for the same purpose may be patentable as such and also as an improved article incorporating the same. If the component in question forms, on the other hand, part of the state of the art together with its relevant properties, the incorporation thereof in the same article will be obvious in view of its predictable beneficial effect. Not unlike cases of "analogous use", where existing means are employed in an obvious manner to obtain an improved performance in known kinds of processes or devices solely on the basis of the known properties of the means, here is an "analogous substitution" of a component in a known combination or mixture with an already existing entity in order to improve the performance of the composition in consequence of the known properties of the said entity.

15. The terpolymers specified in the claim of the patent under consideration embrace known materials described in document (1) when the manner of preparation is disregarded. It has already been demonstrated in document (5) that such terpolymers can be prepared with improved properties under specific processing conditions. The submission of the respondent that it was not stated in the cited document by what standard the improvement had to be assessed, is not persuasive in view of the quantitative disclosure of the properties of the terpolymer in Example 1 of (5). When these are compared with Example 6 in (1), the substantial improvements, both in respect of heat deflection resistance and impact resistance, become apparent. The respondent submitted no evidence as to the unsuitability of this (his own) material for the purpose of forming compositions with graft polymers, contrary to its advantageous properties.

16. In view of the respondent's own argument that there was no or hardly any terpolymer available in the art for such purposes, it is all the more likely that the skilled man would have inevitably turned to the above materials with excellent published properties. The skilled man must be free to employ the best means already available for his purposes, although the use of means leading to some expected improvements may well be patentable if relying on an additional effect, provided this involves a choice from a multiplicity of possibilities. The lack of alternatives in this respect may, therefore, create a "one-way-street" situation leading to predictable advantages which remain obvious in spite of the existence of some unexpected "bonus" effect. The compositions claimed in the patent under consideration fall into category and are therefore obvious having regard to the state of the art. This applies equally to the amended claim presented with the statement of the grounds for appeal incorporating a minor correction.

17. It was evident that the tests submitted on behalf of the applicants during the substantive examination had no relevance to the problem with which the invention was concerned. The main feature of the problem was to improve heat deflection resistance and flowability (i.e. processability), and this was to be achieved without losing impact resistance. Whilst even the evidence available from the patent proprietor demonstrated that no improvement in flowability was obtained, the only remaining basis for patentability may have been an unexpected increase in the Vicat test figures. This requires, of course, positive results which cannot be replaced by negative results generated by reducing impact resistance under extreme conditions. The requirement that impact resistance be maintained clearly implied that prior art had already achieved a satisfactory degree thereof and that the improvement in heat deflection resistance should therefore be assessed in comparison with such art. In particular, the use of extreme conditions to demonstrate "unexpected" improvement in this respect must be strongly objected to since this could easily lead to a situation where everything becomes patentable in view of its advantageous performance in comparison with its own modified version which carries one of its features in a distorted from or to an exaggerated degree.

18. For the above reasons, the Opposition Division was wrong to conclude that the results of a comparison between the applicants' process and another one lying outside their claim was relevant to the inventive step. The choice of conditions was in any case unfair since the extremely long dwell time in the process described in the specification of DE-A-2 343 871 Example 2 was designed to demonstrate the beneficial effect of a different technique in a situation in which the formation of incompatible products was already observed. Notwithstanding this, the heat deflection resistance of the material remained virtually unchanged, which rendered the results meaningless with respect to the stated problems. It was, therefore, equally erroneous to suggest that the opponents should have provided results under less extreme conditions. Whenever an invention resides in the modification of a known article in order to improve its known capability, the modifying feature must not only characterise the invention in the claim, i.e. distinguish it from the prior art, but must contribute causally to the improvement of the capability thereby achieved. Thus, if no property implying a new use is involved, the onus is on the applicant to make the improvement credible, if necessary with evidence, as long as said improvement is still unexpected in the light of the state of the art.

19. However, since the incorporation of a high quality terpolymer in the known type of composition was obvious irrespective of the actual degree of improvement, there is no good reason to grant the respondent's request for the submission of an comparative test results. This is in consequence of the character of the claimed composition which includes a component already known to confer advantages on the whole combination. Whilst it is the firm opinion of the Board that the door for real selection inventions in the field of compositions and admixtures should be left open, these must represent testable and significant improvements which are neither expected nor necessarily obtained through a direct and simple optimisation of the available ranges of parameters in the routine development of the products.

ORDER

For these reasons, it is decided that:

1. The decision of the Opposition Division of the European Patent Office dated 29 June 1982 is set aside.

2. European Patent No. 1625 is revoked.