T 0002/81 (Methylenebis(phenylisocyanate)) 01-07-1982

Reasons for the decision

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

2.1. All the claims have been restricted to a quenching of the reaction mixture once the desired isocyanate content has been reached. This measure was already emphasised as preferred type of cooling in the first documents. All examples illustrating the overall process were performed under such conditions. The applicant has a right to restrict the claims and the description to those subject-matters which are conceptually clear and supported by the desription in combination with other features. By means of further illustrations the appellant has clarified the scope of the term "quenching" in accordance with the generally available specialist knowledge. From the formal point of view the amendment is therefore admissible.

2.2. Supported by documents, literature or generally avaible specialist knowledge, the Examining Division can of course doubt whether the amendment or introduction of a feature is actually technically relevant, particularly when such a correction of the claim should be critical for novelty or inventive step. The burden of proof then lies with the applicant for clearing the doubts and substantiating the converse. In the present case this requirement has been met in the appeal proceedings by submission of comparative experiments with quenching as against cooling. The facts of the matter support the argument that this measure is essential.

3. The present main claim has also been restricted in regard of the PO concentration and reaction temperature. The range claimed (from 0.05 to 10 ppm) is supported by the description (page 7, lines 28 and 29). Although a general range "from 1 ppb to 10 ppm, preferably from 0.05 to 5 ppm" is stated there, the part-range now claimed may be regarded as disclosed. The end-points are specifically named, and the two part-ranges of the general lying outside the preferred range would be unequivocally and immediately apparent to the person skilled in the art. The simple sub-combination of these part-ranges of the concentration values as claimed would not merit novelty as "selection", so that the restriction does not represent any new subject-matter within the meaning of Article 123(2). The same applies to the temperature range from 180 to 300°C now claimed, which is similarly directly derivable by combination of the general range from 150 to 300°C with the preferred temperature range from 180 to 240°C (original text page 6, line 26). The amendments in the claim are therefore admissible. The remaining patent claims agree with the original version except for the admissible change from cooling to quenching, and are likewise not open to objection.

3.1. The first instance recognised the novelty of the subject-matter of the application in relation to the five prior publications cited. The latter also refer to other documents and the applicant himself included further publications characterising the state of the art. None of these documents gives cause to call into question the novelty of the subject-matter of the application.

3.2. For the assessment of the inventive step of the subject-matter of the application the development of the state of the art is of importance. It was indeed known that MBP can be converted by partial carbodiimidisation into a liquid mixture having a lower melting point than pure MBP. But the mixture can be employed as substitute for the pure material in all conventional polycondensation reactions without the liquid first having to be filtered or pre-heated. The art has made great efforts to produce such MBP derivatives in good quality. The US-A-3 152 162(1) uses high temperatures (150° to 300°C, preferably 180° to 220°C) for a longer period (10 to 15 hours) in order to achieve the desired carbodiimidisation without catalyst. The reaction was terminated by cooling. The product was liquid in the case of isomer mixtures, but crystallised when pure MBP had been used (Example 3). This example was aimed at a higher degree of conversion, but apparently no acceptable product was obtained, as is also shown by the appellant's experiments (see Appendix to the Statement of Grounds of the appeal).

3.3. The further development of carbodiimidisation then turns to acceleration of the reaction by means of a catalyst. For this purpose significant amounts of catalyst (3-5%), e.g., triethyl phosphate or biuret, are employed at temperatures above 150° (see DE-A- 1 593 619(2) or DE-A-1 668 083(3)). The last-named patent specification emphasises the necessity of employing a catalyst which is completely inactive at room temperature, and utilises the method of quenching to prevent the formation of dimers.

3.4. It is significant that after the first phase of processing at high temperature with or without a comparatively inactive catalyst recourse was had to the highly active phospholine oxides (PO). The intention was to reduce the reaction temperature to below 150°C, and particularly below 100°C. The processes according to FR-A-2 306 197(4), FR-A-2 322 129(5) and FR-A-2 341 605(6) all use PO in combination with various catalyst poisons in order to prevent further activity on the part of the catalyst during storage at room temperature. The only alternative solution to this was conversion of the permanently active PO catalyst to an insoluble and separable state, i.e., to a heterogeneous catalyst complex (FR-A-229 353(7).

3.5. From the remarks in the above-mentioned documents of the prior art and from corresponding cross-references in patent specifications the Board concludes that certain facts in this special field of technology were generally perceivable. Firstly, the catalyst corresponds either to the type which because of its comparatively much lower activity is only active at higher temperatures, or it represents a highly active PO which is to be used (a) as a homogeneous catalyst at low temperatures and (b) in conjunction with a poison when it is not employed in an insoluble and easily separable form. Secondly, the first-mentioned type has to be inactivated by cooling or quenching (3), whereas the last-named does not require any artificial cooling (4, 5 and 6).

3.6. The third conclusion concerns the unfavourable quality of the conversion products, since in the case of no or little catalyst a higher temperature or a longer reaction time is required (1,4 etc.). Doubt was also expressed as to the the efficacy of these poisons, and also even about the complete inactivity of the catalyst of the first-named type at room temperatures. All these facts were common technical knowledge in this field.

4.1. The appellant essentially regards three features for the solution of the above-mentioned problem as being essential to the invention: (1) small amounts of PO, (2) elevated temperatures and (3) quenching instead of cooling. It is also recognisable from comparison with the background art that the absence of the catalyst poison could be an important additional characteristic. The use of reduced amounts of PO is actually not new per se, although the preferred range of concentration according to the application is somewhat lower than it is in the prior art. Previously, low temperatures were preferred and catalyst poison was indispensable. The earlier PO processes represented a departure from the high-temperature technique with catalysts which could be largely inactivated by cooling. It was accepted that darker products were to be expected under these conditions. The development trend was towards increasingly better poisons (4, 5 and 6) or complex formation (7). But it was specifically not aimed at dispensing with the catalyst poison.

4.2. The solution found according to the application is based on the discovery that PO is completely eliminated stepwise at elevated temperatures; for without this reproducible and controllable decomposition of the catalyst the solution of the task, i.e., an improved storage stability, would not be possible. There were indeed some doubts about the reliability of terminating the reaction with poison, but under no circumstances would one have omitted the poison from the reaction mixture. The applicant's experimental results (see with letter dated 21 February 1980) show that, on reworking the processes according to (4) and (5) with an excess of poisons, only a dubious long-term stability develops.

4.3. Without the discovery that PO may be completely eliminated at higher temperatures, the employment of the poison would have continued to remain the best solution. Temperature elevation could not be expected to bring anything but disadvantages, so that there was no incentive to take this route - not even from the viewpoint of optimisation. The argument of the Examining Division, that a need for reduction of the PO concentration existed and that the temperature elevation automatically followed from this, is one with which the Board cannot concur. This represents an inadmissible ex post facto approach. The foregoing exposition in fact leads to the conclusion that, if anything, there was a prejudice against reduction in the amount of catalyst with simultaneous elevation of the temperature.

5.1. Whereas the main task of improvement in the long-term stability is solved by the complete decomposition of the catalyst, the quenching as against the much slower cooling in itself represents the solution of another partial task, with immediate advantages. The comparative experiments were performed under otherwise identical conditions, so that the absence of turbidity and browning must be ascribed to the speed of cooling.

5.2. None of the three PO methods in the prior art point to the necessity for artificial cooling, let alone quenching. The patent specification (6) mentions (3) as prior art, and the latter document recommends quenching to prevent side-reactions. However, this process does not use any PO catalyst, but a catalyst of the above-mentioned type (see 3.3.) having entirely different properties and effects. This could also be of importance for the direction and nature of the side-reactions. It was known that the PO processes on the one hand and the earlier catalytic methods on the other employ mutually exclusive conditions. It is therefore not justified to combine these documents, i.e. (6) with (3). The advantageous effect therefore provides a contribution to the inventive step of the complete process.

6. In summary it is clear that, given the problem to be solved, neither the methods of the prior art individually, nor their respective combination with the generally available specialist knowledge, would make the solution according to the invention with the advantageous effects achieved foreseeable. While it is inadmissible to combine unrelated or conflicting documents in order to deny inventive step, it is indeed permissible to consider various documents together mosaically in order to prove a prejudice or a general trend pointing away from the invention. The idea of departing from the catalyst poison regarded as indispensable, in conjunction with the teaching that the PO catalysts decompose at higher temperature, represents a valuable simplification of the state of the art which could not have been found without an inventive step.

7. The appellant has appropriately restricted the wording of the claims by letter of 29 March 1982. The variants claimed are credibly supported by the examples and the experimental results submitted, and appear correspondingly suitable for solution of the task set. While the absence of the catalyst poison is not mentioned in the claim as filed, the Board did not regard this as necessary, being a negative feature.

8. For the reasons given the Boards holds the appeal to be well founded.

9. No application has been made for reimbursement of the appeal fee in accordance with Rule 67 EPC. As crucial evidence was not submitted until the appeal proceedings the facts of the case would not justify such a measure.