T 2197/09 20-11-2014

1. The appeal is admissible.

2. As announced in advance, the duly summoned appellant did not attend the oral proceedings. According to Rule 71(2) EPC 1973, the proceedings could however continue without the appellant. In accordance with Article 15(3) RPBA, the board relied for its decision only on the appellant's written submissions. The board was in a position to decide at the conclusion of the oral proceedings, since the case was ready for decision (Article 15(5) and (6) RPBA), and the voluntary absence of the appellant was not a reason for delaying a decision (Article 15(3) RPBA).

3. The document D5

3.1 In support of its position that there is no insufficiency of disclosure, the appellant cited an excerpt from the PhD thesis of one of the inventors, for which no publication date was given. The complete thesis may, however, easily be found online (document D5), and on a page bearing the heading "Proefschrift" it is stated (in Dutch) that the author defended the thesis in public on 10 February 2003. As there is no evidence that the thesis was available to the public before that date, and as the priority date of the present application is 7 November 2000, document D5 is regarded as a post-published document.

Although the disclosure requirement of Article 83 EPC 1973 has to be met at the filing date, or, if applicable, at the priority date (see e.g. T 891/02, Reasons, point 3, first paragraph), post-published documents can be used as evidence in deciding whether the invention was indeed reproducible without undue burden at the relevant filing date (see T 994/95, Reasons, point 8; T 157/03, Reasons, point 9).

It is clear that from the appellant's submissions that document D5 has been cited for this purpose, i.e. to support the contention that "there is no undue burden for the skilled person in carrying out the invention". The evidential significance of document D5 in this respect will be considered in the following.

4. Main Request: Articles 83 and 84 EPC 1973

4.1 Claim 1 is directed to a lithographic projection apparatus comprising inter alia a projection system having precisely six imaging mirrors; an "incidence angle classification, C" is introduced, and the projection system is defined to correspond to one of twelve alternative classifications listed in the claim.

To produce such a projection system it is not disputed that a skilled person, following normal optical design procedures, would firstly have to arrive at a suitable first-order layout (mirror spacings, curvatures etc.) subject to the system constraints. The further stages would involve optimization of the initial design to conform to the design objectives, followed by manufacturing, testing etc. It has never been alleged that a skilled person would be unable to carry out these further stages using standard procedures, and the Board also sees no reason to doubt this.

Hence, in relation to Article 83 EPC 1973, the point at issue is whether a skilled person would be able to arrive at a workable initial layout for each of the systems corresponding to the claimed incidence angle classifications (apart from 9(+), which was not objected to) on the basis of the disclosure of the application, and having regard to the common general knowledge in the art.

4.2 It is undisputed that, with the exception of incidence angle classification 9(+) (see Table 7), the application does not explicitly disclose starting layouts for the claimed incidence angle classifications. The only other layouts explicitly disclosed for six mirror systems (Tables 8 and 9) concern the non-claimed classification 37(+).

4.3 In the absence of explicit prescriptions, it is reasonable to enquire whether it would be a trivial matter for the skilled person to arrive at the required starting layouts. This is, in the opinion of the Board, essentially a question of fact, and hence it is permissible to consider evidence on this matter even if it appears in a post-published document such as document D5.

In the first paragraph on page 51 of document D5 it is stated that:

- "due to the large number of variables and constraints involved [in EUV reflective ring-field projection systems], choosing unobstructed starting configurations for subsequent optimization is a nontrivial issue."

In the third paragraph on the same page (page 51) the difficulties are explained as follows:

- "The use of mirrors may lead to obstruction. The absence of obstructions is an essential and, as we will show, a very restrictive demand. Geometrical obstruction confines the possible system configurations to small domains of the parameter space."

Post-published document D5, cited by the appellant, is therefore considered by the Board to provide clear evidence that, as a matter of fact, the determination of a suitable starting layout is not a trivial matter.

4.4 Despite the above conclusions, the requirements of Article 83 EPC could still be met if the application disclosed a method by which the skilled person could reliably arrive at the required layouts without undue burden.

According to the method set out in the application (see page 19, lines 5-7), a six mirror system is represented as having 13 variables, i.e. 6 curvatures (c1-c6) and 7 distances (d0-d6). The stop position is defined as being on one of the mirror surfaces.

Imposing a number of constraints on the system (e.g. constraints G1-G4, see page 14, lines 8-20) reduces the number of independent variables. For example, four equations may be derived for d0, d1, d5 and d6 in terms of the other variables (equations 45-48, 51-54, 57-60 and 63-66), thus reducing the number of independent variables from 13 to 8.

The pertinent question, therefore, is whether a skilled person, using the information disclosed in the application and common general knowledge in the art, would be able to proceed from this point to a fully worked out starting layout (similar to that of Table 7) for the classification classes other than 9(+).

4.5 In the statement of grounds, the appellant suggests that the skilled person could find starting layouts without undue burden by searching through the entire solution space, i.e. the realistic ranges for the mirror curvatures and the distances between the mirrors, using sampling.

The only mention of sampling in the application, however, appears on page 12 (lines 12-13), where it is stated that "the paraxial approach and the constraints yield a limited number of variables that are sampled to identify solutions." Whether the skilled person would understand this brief comment to imply the exhaustive systematic search described by the appellant and disclosed in the post-published document D5 is open to question.

4.6 However, even if a skilled person did arrive at an understanding that the starting layouts are to be found by a systematic search of solution space, the task facing the skilled person would, in the opinion of the Board, be formidable.

4.7 Even for a single sampled point of solution space, it is doubtful that the suggested procedure would be an entirely straightforward matter.

A skilled person could reasonably be expected to be able to operate commercially available optical design software (e.g. as suggested on page 10, lines 30-33), which would enable the basic ray tracing to be carried out. However, a "particular problem in designing mirror systems" is obscuration (page 29, lines 7-16), and each configuration would have to be evaluated for this, for example by the procedure suggested on page 29, line 17 - page 30, line 4). In addition, according to document D5 (page 64, first paragraph) "workspace" and "feasibility" (i.e. no unrealistic values) need to be checked also. No evidence has been provided that commercial software would, without modification, be capable of providing such evaluations.

4.8 Furthermore, even with appropriate software the appellant concedes that, according to document D5, "it takes several personal computers about a week to evaluate 20 billion systems" (page 64, section 4.2.2), and that for a six-mirror system, typically 2.6 X 10**(12) configurations need to be evaluated, a factor of 130 more. Although the appellant suggests ways in which the time required could be reduced (e.g. using a lower sample rate or a supercomputer), it can hardly be denied that the skilled person would be faced with a significant, lengthy and burdensome research project.

4.9 In the present case, the burden placed on the skilled person is, in the opinion of the Board, particularly unwarranted. The claimed invention concerns the 12 classification classes out of all possible 6-mirror systems which the inventors determined to be "feasible" (page 11, lines 5-10). The applicant (now the appellant) must be presumed to have derived workable starting solutions with no obscuration, no unrealistic parameters etc. for the 12 classes of "feasible" 6-mirror systems before the priority date of the present application. If this were not the case, how would the applicant have known that precisely these classes are "feasible"?

In the application as filed, however, the applicant chose to disclose a working layout for only one (9(+)) of the claimed 12 incidence angle classifications, thereby obliging the skilled person to perform a laborious and time-consuming research project to arrive at workable layouts for the other 11 claimed classes. In the opinion of the Board, this amounts to a failure to discharge the duty of disclosure set out in Article 83 EPC 1973.

4.10 In the light of the above, the Board judges that the subject-matter of claim 1 of the main request does not meet the requirements of Article 83 EPC 1973.

Consequently, the Board does not consider it necessary to decide on the separate objection of the Examining Division that the subject-matter of claim 1 of the main request fails to meet the requirements of Article 84 EPC 1973.

5. Auxiliary Request: Inventive Step

5.1 Claim 1 of the auxiliary request is restricted to the incidence angle classification 9(+). The Board concurs with the finding of the Examining Division that this subject-matter is adequately disclosed.

5.2 The closest prior art is considered to be one of the six-mirror EUV lithography systems referred to in the application (page 3, final paragraph), for example those of documents D1 (class 41(+)), D2 (class 41(+)), D3 (class 41(+)) or D4 (class 45(-)). Claim 1 differs in having a layout corresponding to another incidence angle classification, i.e. class 9(+).

5.3 According to the application, the technical problem is "to provide alternative and improved projection systems for EUV radiation and a methodology for designing such systems." (Page 4, lines 1-2.)

Hence, the Board's understanding is that what is aimed for is a projection system capable of delivering either an improved performance in relation to the closest prior art, or at least a level of performance broadly comparable to that of the prior art systems, achieved by means of an alternative optical design. Understood in this sense, the Board can accept that this is a reasonable technical problem to pose.

5.4 The Board stresses, however, that a convincing argument in support of inventive step could not be based on a "problem" of merely providing any alternative projection system for EUV radiation, even one delivering a performance inferior to systems of the prior art. An inventive step cannot be acknowledged on the basis of a purely disadvantageous modification of the closest prior art (see Case Law of the Boards of Appeal, 7th edition 2013, I.D.9.18.1).

5.5 It is therefore necessary to enquire whether modifying the closest prior art by incorporating a projection system having an incidence angle classification of 9(+) could plausibly be regarded as solving the technical problem set out under point 5.3, above.

5.6 The Board accepts that there is no general requirement under the EPC for an applicant to provide experimental proof for patentability and the establishment of plausibility in this respect is "only relevant when examining inventive step if the case at hand allows the substantiation of doubts about the suitability of the claimed invention to solve the technical problem addressed and when it is thus far from straightforward that the claimed invention solves the formulated problem." (see T 578/06, Reasons, points 13 and 15).

5.7 The Board takes the view, however, that the present case gives rise to legitimate doubts in this respect.

As mentioned above, the design of optical systems typically involves determining a suitable first-order layout, and then applying conventional optimization techniques to arrive at the finished design. The design is then assessed using performance evaluation methods familiar to the skilled person.

According to the present application, obtaining a "feasible" first-order layout for a 6-mirror system (free of obscuration, no unrealistic parameters etc.) is a particular problem, and as a result of its researches, the applicant concludes that among all possible 6-mirror systems, the only feasible layouts are those listed on page 11, lines 8-10.

These layouts therefore represent the starting points for the further stages of the procedure necessary to solve the above-mentioned problem, i.e. optimization and evaluation to determine which - if any - of the optimized systems can deliver performance superior (or at least comparable) to that of the prior art systems.

However, no performance data is presented in the application which would give credibility to the idea that at least some of these systems - and in particular 9(+) - solve the above problem. In fact, the application does not even appear to contain any assertion or claim - even an unsubstantiated one - that this is the case. Indeed, there is nothing in the application to indicate that any attempt at optimization and comparison with the prior art had even been made at the priority date.

The Board therefore concludes that the application does not make it plausible that a projection system having an incidence angle classification 9(+) could provide superior - or even equivalent - performance compared to the systems of the prior art based on incidence angle classifications 41(+) or 45(-).

5.8 The only other source of evidence available for the technical effects provided by a projection system having an incidence angle classification 9(+) is the the post-published document D5 (although cited in relation to the issue of sufficiency of disclosure in relation to the main request).

The jurisprudence of the boards of appeal on the extent to which such evidence appearing in a post-published document may be taken into account is well-established (see "Case Law", op. cit., I.D.4.6).

In the present case, however, it is not necessary examine this jurisprudence in detail, since the only evidence actually found in post-published document D5 in this respect points in exactly the opposite direction.

5.9 In particular, in document D5 two of the requirements of an EUV projection system are listed as "extremely small distortion" (page 24, line 1) and an rms "wavefront error smaller than lambda/50 in image space" (page 24, "Number of mirrors", first paragraph).

In Fig 6.8 (page 89) a six mirror system in class 9(+) is shown, and the caption reads as follows:

- "A positive six-mirror system in class 9+. Although the angles of incidence are large, the root-mean-squared wavefront error can decrease down to lambda/2. The object heights are between 114 and 118 mm. the numerical aperture is 0.3 but the distortion is large."

This system, referred to as an "exotic design" (page 88, final paragraph) therefore has an rms wavefront error of lambda/2 or greater and large distortion, and hence does not meet the requirements for a EUV projection system mentioned above.

This is further confirmed on page 127 (fourth paragraph) where it is conceded that systems in the useful six-mirror classes "are already known from the patent literature". The inference is clear: the 9(+) system is interesting as an "exotic design" but is not "useful" as an actual EUV projection system.

5.10 In summary, there is no evidence in the application or elsewhere that incorporating the the distinguishing feature of claim 1 of the auxiliary request into the closest prior art would solve the problem of providing an improved (or even equivalently performing) EUV projection system. In the absence of any other realistic technical problem credibly solved by the claimed subject-matter, it is not possible to acknowledge an inventive step.

Consequently, the Board judges that the subject-matter of claim 1 of the auxiliary request does not involve an inventive step within the meaning of Article 56 EPC 1973.