T 0009/86 (Polyamide-6) of 6.11.1986

Summary of Facts and Submissions

I. European patent No. 5 466 was granted on 27 April 1983 (Patent Bulletin 83/17) on the basis of European patent application No. 79 101 260.2 filed on 26 April 1979 claiming the priority of a previous application (DE 2 821 686) of 18 May 1978. It had nine claims of which Claim 1 read as follows: "Process for producing polyamide-6 or other polyamides with a low residual monomer and dimer concentration and consisting of at least 60% by weight Zeta-caprolactam units by the hydrolytic polymerisation of Zeta-caprolactam and if appropriate further monomers in a reactor with subsequent separation of the low-molecular-weight by- products existing in equilibrium with the polyamide and of the unreacted Zeta-caprolactam from the polyamide melt in a partial vacuum and at elevated temperatures in an evaporator and their recycling to the reactor, characterised in that the low-molecular-weight by-products and the-caprolactam separated off from the polyamide melt in gaseous form are condensed directly on the Zeta-caprolactam melt still to be polymerised and the resulting mixture is conveyed to the reactor, if appropriate together with further monomers."

II. On 11 November 1983 the opponents filed notice of opposition to the European patent, requesting that it be revoked in its entirety and citing new documents in support.

III. By decision dated 7 November 1985 the Opposition Division rejected the opposition, essentially on the following grounds: (i) Where the citations related at all to processes for producing polyamide from caprolactam and subsequently separating off and recycling unreacted monomers or low-molecular-weight by-products, they failed to give the teaching of the contested patent whereby the products separated off in gaseous form are condensed directly on the monomer melt still to be polymerised; the subject-matter of the patent was thus novel. (ii) Although there was a similarity between the process of the contested patent and that of the nearest prior art (DE-A-2 503 308 (1)) in that a sufficiently high proportion of monomers was present in the condenser to prevent the dimers precipitating out, in (1) this result was achieved by halting the polymerisation process before equilibrium was reached. This teaching does not point to the solution proposed to the same problem by the patent in suit, which lies in the apparatus used rather than in influencing the kinetics of reaction.

IV. The opponents filed notice of appeal against this decision, paying the prescribed fee on 3 January 1986 and submitting a statement of grounds on 6 March 1986. They based their arguments on (1) and on a new citation, DE-A-1 949 911(5). They argued that it was known from (5) that components of a polyamide melt that could be separated off, i.e. monomers and oligomers, can be drawn off with a spray device and the condensable parts condensed therein, caprolactam being particularly recommended as the propellant liquid. This left as the only patentable feature at most the fact that the resulting mixture was recycled for polymerisation, something suggested in (1) and in any case obvious. A comparison of the process described in Example 1 of the patent in suit and that claimed in Claim 1 of (5) reveals no difference: in both cases the vapour is condensed in a large amount of caprolactam, which is then recirculated by pumping. V.The respondents, on the other hand, claim that according to (5) the parts drawn off from the polyamide melt and condensed with the aid of a spray device must be filtered; filtration is essential as the propellant liquid for the spray device is recirculated, with the result that the limit of solubility of the dimer in caprolactam is rapidly exceeded, causing it to precipitate out. The process claimed cannot be inferred from such a teaching. With regard to (1) the respondents cite the arguments put forward by the Opposition Division.

VI. In oral proceedings on 6 November 1986 the parties substantiated their arguments: (i) The appellants referred to the breadth of the teaching in (5). In a preferred embodiment (Claim 6) filtration was essential only if a condensate was used as the propellant liquid. If, on the other hand, pure caprolactam was used - as also suggested -the only thing not mentioned was the recycling to the reactor of the mixture plus the extractable components. This, however, was obvious for reasons of economy and was, moreover, already suggested in (1). (ii) The respondents for their part stressed that according to (1) the polymer melt used for the evaporation process contained more than 9% of monomer; this called for a pressure of between 6 and 60 torr (p. 3, first paragraph), considerably higher than the 0.4 to 6.5 mbar referred to in the contested patent; this poor vacuum, quite consciously opted for, was in fact necessary if the dimer was to remain soluble. This meant that in (1) the dimer component was not wholely eliminated, a particular disadvantage when the polyamide was to be used for packaging foodstuffs. A further serious disadvantage of the process of (1) was that polymerisation could not be continued to equilibrium. The major drawbacks of the process described in (5) were the use of a spray device and the presence of a filter. Although recirculation of the caprolactam ensured that the vapour was washed, the technique was not suitable in practice because the capacity of the vacuum device, operated in the manner of a water jet vacuum pump, was too small to cope with the large volume of caprolactam to be conveyed through the continuous process tube reactor, making the process uneconomical. For this reason there is nothing in (5) to suggest the present invention.

VII. The appellants request that the contested decision be set aside and the patent revoked. The respondents in turn request that the appeal be dismissed and the patent maintained with the following amendments to Claim 1: -Column 6, line 33: "Verfahren" (process) to be replaced by "kontinuierliches Verfahren" (continous process); -Column 6, lines 51 - 52: "kondensiert werden" (are condensed) to be replaced by "derart condensiert werdenen, daß sich kein Dimer abscheidet" (are condensed ... in such a manner that no dimer is precipitated out), the description being brought into line accordingly.

Reasons for the Decision

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

2. ... (re amendments to Claim 1 and description)

3.The first comment to be made is that when polyamide-6 is produced from Zeta-caprolactam the hydrolytic polymerisation of the monomer proceeds until an equilibrium is reached dependent on the temperature and represented, since the working temperature normally used is between 240 and 280°C, by a content of Zeta-caprolactam and low-molecular- weight by-products (see preamble to the description of the contested patent) of 7 to 14%. The low-molecular-weight by-products contain unreacted monomer and oligomers, predominantly the dimer. Such products, more especially the dimer, must be eliminated before the polymer is processed further since they impair the mechanical properties of polyamide mouldings.

In (1), which is to be viewed as the nearest prior art, the kinetics of the reaction are used to obviate these difficulties and instead of being allowed to continue until an equilibrium is established between low-molecular- weight by-products, unreacted monomer and polymer, the polymerisation process is interrupted once the monomer content of the polyamide melt exceeds 9% and reaches, for example, 14 or 16% (cf. claim and examples). In the course of the subsequent vacuum treatment the dimer - with the monomer - passes into the condensate, where it is kept in solution by the high monomer content, after which the condensate can be fed back into the continuous process tube reactor. The disadvantage of this procedure is the low yield that results from failure to achieve equilibrium and the low utilisation of the tube reactors. Furthermore, the high monomer concentrations in the condensate lead to increased consumption of energy. Compared with this state of the art the technical problem the patent in suit sets out to solve is to be seen as reducing consumption of energy and making the continuous process more economical without running the risk of the dimer precipitating in the condenser. The solution offered by the contested patent is to continue the polymerisation process to equilibrium and to condense the components separated off from the polyamide melt in gaseous form directly on the monomer melt still to be polymerised in such a manner that no dimer is deposited, subsequently conveying this mixture to the reactor. The advantages thus achieved speak for themselves; in the Board's view, therefore, the problem at issue is actually solved.

4. The proposed solution cannot be inferred from any of the citations as none of them discloses the combination of the three features essential to the claimed process. Since this fact has not been disputed, any more detailed comments on the novelty of the process of the contested patent would be superfluous.

5. It must therefore be considered whether the process involves inventive step. Viewed in isolation from the evolving state of the art, the solution proposed by the contested patent appears at first sight to be obvious. On the one hand, a reduction in energy consumption was to be achieved by once again allowing polymerisation to proceed to a point of equilibrium and, on the other, the threatened deposit of dimers in the condenser was prevented by the use of suitably large amounts of caprolactam, as indicated in the claim in (1). If, however, the process is viewed against the background of the actual technical development in this field one comes to the opposite conclusion.

5.1 Very different methods were adopted to remove the volatile components, namely:

-Production of granulates from the melt in equilibrium and removal of the unwanted products with water;

-Reducing the extract content before spinning the melt, and

-Vacuum treatment in the screw reactors or in the thin-layer evaporators.

Apart from being expensive these methods do not satisfactorily solve the specific problem of eliminating the cyclic dimer which because of its physical properties is a particularly inconvenient by-product for the continuous evaporation process. It sublimates easily, has a very high melting point and a low solubility in Zeta-caprolactam. It consequently has a marked tendency to precipitate in the evaporator condenser systems, which leads to clogging and interruptions in the evaporation process.

Special devices suggested to obviate these difficulties include

-condensers provided with scrapers,

-double condensers used alternately,

-rinsing systems.

All these necessitate additional investment in plant without guaranteeing continuous processing.

5.2 Citation (5) also proposes a continuous polyamide melt evaporation process in which the volatile products are drawn off under a vacuum of, for example, 0.1 to 3 torr using a spray device and the condensable components are condensed there in the propellant liquid, which may be either pure caprolactam or the collected condensate (p. 1, fourth paragraph, to p. 2, first paragraph). The products drawn off by means of a spray device and condensed in the propellant liquid are filtered (p. 2, line 9; p. 3, line 13, and Figure) and recirculated. Because filtration is not specified in Claim 1 of (5) it is unclear how this idea can be carried out. The appellants' assertion that this could have been done exactly as described in the contested patent reflects an ex post facto approach since given the economic significance of such a possibility for the continuous process with which (5) is solely concerned (cf. p. 1, third paragraph) this idea would have been mentioned, possibly by an indication that after use as a propellant liquid the pure caprolactam is fed back into the continuous process tube reactor.

5.3 That (5) did not provide a satisfactory solution is shown also by the fact that a completely different approach was then adopted in (1), which in order very simply to avoid the difficulties associated with dimer deposits advises against allowing the polymerisation reaction to proceed to the stage of equilibrium; allowance is here consciously made, however, for the additional energy required to evaporate the increased amount of monomer, as already explained in point 3 above. Given, therefore, the evolving state of the art there was clearly, despite a multitude of experiments and proposals, still need for a cost-efficient solution to the problem of dimer deposits.

6.Contrary to the teaching of the nearest prior art the contested patent abandons the idea put forward in (1) of not continuing the polymerisation reaction to equilibrium and falls back on earlier suggestions that equilibrium should be achieved. The additional measures indicated in the characterising portion of Claim 1 ensure in a suprisingly simple manner that the dimer content in the condenser remains below the solubility limit and that the dimer-containing caprolactam is used economically in the continuous process as a whole. Given the way in which the state of the art developed in this area, to which considerable attention was devoted because of its commercial importance, the very simplicity of the solution proposed in the contested patent is an argument for the existence of inventive step. Those involved in the field who for a long time had sought a solution to the same problem in vain and were recommending possible, but more costly, ways of achieving the desired end, missed the solution disclosed here that in retrospect appears so simple.

7. Claims 3 to 9 - now 2 to 8 - are directed to preferred embodiments of the process as claimed in Claim 1; since it is allowable, so too are they.

ORDER

For these reasons, it is decided that:

1. The contested decision is set aside.

2.The case is remitted to the department of first instance with the order that the patent be maintained in the following form: ... .