T 1353/14 () of 6.3.2019

Summary of Facts and Submissions

I. The appellant (patent proprietor) lodged an appeal against the decision of the opposition division revoking European patent No. 2165178.

The opposition filed by the respondent (opponent) against the patent as a whole was based on the grounds for opposition of

- added subject-matter (Article 100(c) EPC),

- exclusion from patentability (Article 100(a) EPC together with Article 52(2)(a) and (c) EPC),

- lack of sufficiency of disclosure (Article 100(b) EPC), and

- lack of novelty and lack of inventive step (Article 100(a) EPC, together with Articles 54(1) and 56 EPC).

During the first-instance proceedings reference was made, among other documents, to the following document:

E1: US 2006 0167 651 A1.

II. In its decision the opposition division held inter alia that

- the ground for opposition under Article 100(c) EPC was not substantiated and, in addition, was not prima facie prejudicial to the maintenance of the patent;

- regarding the ground for opposition under Article 100(a) EPC together with Article 52(2)(a) and (c) EPC), the subject-matter of dependent claim 10 as granted was not excluded from patentability under Article 52(2)(a) and (c) EPC;

- the ground for opposition raised under Article 100(b) EPC with respect to the second variant defined in dependent claim 2 as granted was not convincing;

- the subject-matter of claim 1 as granted was new over the disclosure of document E1 but did not involve an inventive step in view of the disclosure of document E1 (Article 56 EPC); and

- none of the auxiliary requests 1 to 6 then on file was admitted into the proceedings.

III. With the statement setting out the grounds of appeal the appellant submitted sets of claims according to a first to a fifth auxiliary request.

By its letter of reply the respondent filed, among other documents, the following document:

A2: "Handbook of Mathematics", I. N. Bronshtein et al.; Springer Verlag, 5th Ed., 2007; three bibliographic pages, and pages 778 to 783, 794 to 797, and 892 to 895.

IV. Oral proceedings were held before the board on 6 March 2019.

The appellant requested that the decision under appeal be set aside and that the opposition be rejected (main request) or, in the alternative, that the patent be maintained as amended on the basis of the claims of one of the first to fifth auxiliary requests filed with the statement of grounds of appeal dated 25 August 2014.

The respondent requested that the appeal be dismissed.

At the end of the oral proceedings the chairman announced the decision of the board.

V. Claim 1, dependent claim 2, and independent claim 15 of the patent as granted (main request) read as follows:

"1. A method for characterizing properties of an article having a multi-layer structure comprising a multiplicity of sites comprising different periodic patterns, the method comprising:

providing a theoretical model of prediction indicative of optical properties of different stacks defined by geometrical and material parameters of corresponding sites, said sites being common in at least one of geometrical parameter and material parameter;

performing optical measurements on at least two different stacks of the article and generating optical measured data indicative of the geometrical parameters and material composition parameters for each of the measured stacks;

processing the optical measured data, said processing comprising simultaneously fitting said optical measured data for the multiple measured stacks with said theoretical model and extracting said at least one common parameter, thereby enabling to characterize the properties of the multi-layer structure within the single article."

"2. The method of Claim 1, wherein said at least two different stacks have at least one of the following configurations: (a) include stacks associated with different locations, respectively, on the article; (b) include stacks associated with the same location on the article and corresponding to different process steps applied to the article; (c) include at least one patterned site; (d) include periodic patterns with one or more different pattern parameters."

"15. A system for use in characterizing properties of an article having a structure comprising a plurality of different periodic patterns, the system comprising a control unit (12) adapted for receiving optical measured data indicative of geometrical parameters and material composition parameters of a measured area on the article and comprising: a memory utility (12A) storing reference data comprising a theoretical model of prediction, said model being indicative of optical properties of different stacks a [sic] multi-layer structure defined by geometrical and material parameters of corresponding sites, where said sites are common in at least one of geometrical parameters and/or at least one of material parameters; and a processor utility (12B) configured and operable for processing and analyzing the optical measured data, said processing and analyzing comprising simultaneously fitting said optical measured data for the multiple measured patterns with said theoretical model and extracting said at least one common parameter, thereby enabling to characterize the properties of the multi-layer structure within the single article."

The remaining claims of the patent as granted, i.e. claims 3 to 14, are all dependent claims referring back to the method of claim 1.

Reasons for the Decision

1. The appeal is admissible.

2. Main request - grounds for opposition under Article 100(c) EPC, and Article 100(a) together with Article 52(2)(a) and (c) EPC

In the decision under appeal the opposition division found that the grounds for opposition raised by the respondent under Article 100(c) EPC, and under Article 100(a) together with Article 52(2)(a) and (c) EPC, did not prejudice the maintenance of the patent as granted (see point II above). These findings were not disputed by the respondent during the appeal proceedings, and the board sees no reason to deviate from the opposition division's conclusion in this respect.

3. Main request - ground for opposition under Article 100(b) EPC

In its decision the opposition division also found that the ground for opposition raised by the respondent under Article 100(b) EPC with respect to the second variant defined in dependent claim 2 as granted did not prejudice the maintenance of the patent as granted. The opposition division held inter alia that the second variant of dependent claim 2 as granted (variant labelled "(b)" and reading "said at least two different stacks [...] include stacks associated with the same location on the article and corresponding to different process steps applied to the article") meant that the "location" was on the article, whereby merely the X and Y coordinates - and not the X, Y and Z coordinates as submitted by the respondent - would be fixed, and that therefore a modification of the stack in the Z direction - for instance, by addition of a layer - would, contrary to the respondent's submissions, not lead to a change in the "location" of the stack on the article. The opposition division concluded that, contrary to the respondent's submissions, there was no contradiction or inconsistency in the second variant of dependent claim 2 that would prejudice the technical feasibility of this variant.

During the appeal proceedings the respondent contested the opposition division's view in this respect and submitted that the location of the stack was defined by three coordinates, and not only by two coordinates as assumed by the opposition division. In addition, when a process step of the stack added or removed a layer, the Z coordinate of the stack, and therefore its location, changed, with the consequence that it was not possible to add or to remove a layer as specified in the second variant of dependent claim 2 without changing the location of the stack.

The board, however, is of the opinion that the skilled person would not understand the expression "location on the article" [emphasis added] in the context of the claimed subject-matter as referring to the three spatial coordinates, but as referring to the position of the stack in relation to the surface of the article and determined, as held by the opposition division, in terms of two coordinates (X and Y). In particular, the skilled person would understand that the "stacks associated with the same location on the article and corresponding to different process steps applied to the article" [emphasis added] referred to in variant "(b)" of dependent claim 2 do not refer to different stacks located at a same spatial point, but to the temporal sequence of stacks resulting from the different stages of the processing of a same stack at a predetermined location or site of the article itself. This construction of the mentioned claimed feature is, in addition, supported by several passages of the description of the patent (see paragraphs [0013] (in particular, the first sentence), [0014] (in particular, the last sentence), and [0057] (in particular, the second sentence)).

Therefore, the counter-arguments submitted by the respondent in this respect are not convincing, and the board sees no reason to question the opposition division's conclusion that the ground for opposition raised under Article 100(b) EPC does not prejudice the maintenance of the patent as granted.

4. Main request - ground for opposition under Article 100(a) together with Article 54 EPC - Novelty of claim 1

4.1 Document E1 is directed to a method of characterizing the properties of stacks formed on a substrate, and in particular of two stacks formed at different locations, or of two stacks within the temporal sequence of stacks resulting from the different processing stages of a same stack at a predetermined location (see document E1, paragraph [0017], last sentence, and paragraph [0026]; compare with variant "(b)" of dependent claim 2 of the contested patent, together with point 3 above, third paragraph). The two stacks are defined by geometrical and material parameters, wherein some of the parameters (the "common parameters" referred to in document E1, and also referred to in the document as "shared" or "global" parameters, see paragraphs [0017], [0001] and [0018]) are common to both stacks, and the remaining parameters of each of the stacks (the "local parameters") are different from the parameters of the other one of the stacks.

4.1.1 The methods considered in document E1 (see independent claims 1 and 9, together with Fig. 1, 2(a) and 2(b) and the corresponding description) essentially involve

- providing a theoretical model of prediction indicative of optical properties of the two stacks,

- performing optical measurements on the two stacks, the respective optical measured data being indicative of the geometrical and the material parameters of the corresponding stack, and

- processing the optical measured data by fitting the optical measured data with the theoretical model, thus obtaining information on the values of the parameters of the stacks.

4.1.2 As regards the handling of the common and the local parameters of the two stacks in the processing step of the optical measured data, document E1 refers to a first prior art processing method (Fig. 2(a) and the corresponding description, in particular paragraph [0023]) in which

- a first fitting process is applied to the optical measured data of the first stack and to the local (P1 and P2) and common (P3 and P4) parameters of the first stack, resulting in the determination of the local and the common parameters of the first stack (see in particular paragraph [0023], lines 15 to 20), and

- a second fitting process is then applied to the optical measured data of the second stack and to the local parameters (P5 and P6) of the second stack, with the values of the common parameters (P3 and P4) of the second stack being set equal to the values of the common parameters (P3 and P4) of the first stack previously obtained in the first fitting process, resulting in the determination of the local parameters of the second stack.

Document E1 then discloses a modification of the first processing method consisting in a second processing method (see Fig. 2(b) and the corresponding description, in particular paragraphs [0024] and [0025]) in which the optical measured data of the two stacks is simultaneously fitted with the theoretical model so as to determine the local parameters (P1, P2, P5, P6) of both stacks. The fitting process is based on an optimization routine called MTSA (Multiple Tool and Structure Analysis) consisting in a parallel or concurrent regression methodology applied to the theoretical model and to the measured data of the two stacks (paragraphs [0016] and [0018]).

4.2 In its decision the opposition division held inter alia that

- the common parameters of the two stacks (P3 and P4) were also obtained in the second processing method disclosed in document E1 and extracted as required in claim 1 as granted, and that

- the method of claim 1 as granted differed from the second processing method of document E1 only in that the article on which the stacks were formed had a multilayer structure comprising a multiplicity of sites comprising different periodic patterns, but that this distinguishing feature did not involve an inventive step.

4.3 The appellant disputed the opposition division's findings in this respect, and in particular the finding relating to the common parameters being extracted as required by the method defined in claim 1. This point was one of the main issues of dispute between the parties in the appeal proceedings.

4.3.1 As regards the interpretation of the claimed feature relating to the extraction of the at least one common parameter, the respondent submitted that claim 1 required that the step of "processing the optical measured data" comprised the first sub-step of "simultaneously fitting said optical measured data [...] with said theoretical model" and the second sub-step of "extracting said at least one common parameter", without requiring a logical link between the two sub-steps, i.e. without requiring that the common parameter was extracted from or determined on the basis of the simultaneous fitting process. The respondent concluded that claim 1 did therefore not exclude that the extracted common parameters were calculated or determined at some other stage and/or that the values of the common parameters were previously fed to the claimed simultaneous fitting process of the optical measured data and then "extracted" in the sense that they would already be present, and then read, in the output of the simultaneous fitting process.

In the opinion of the board, however, the skilled person in the technical field under consideration (see paragraph [0001] of the patent specification) would understand the claimed method as requiring extracting the at least one common parameter from the output of the simultaneous fitting of the optical measured data of the stacks with the theoretical model. In particular, the skilled person would rule out construing the two mentioned sub-steps as two steps technically unconnected to each other, among other reasons because such an interpretation of the claimed method would then require the step of providing the theoretical model and also the step of carrying out the fitting process involving the theoretical method, without a recognisable connection of these steps with the claimed purpose, i.e. with the characterization of the properties of the article being measured.

4.3.2 In the second processing method disclosed in document E1 with reference to Fig. 2(b) the common parameters are set to a common value for the two stacks (see paragraph [0018], lines 4 to 9; paragraph [0019], lines 8 to 11; paragraph [0020], lines 11 to 15; page 4, right column, lines 13 to 28; and paragraph [0026], first sentence) and, as emphasized by the appellant, the claims of document E1 also require that the common parameters are "held fixed", and therefore set to a common, fixed value, in the process of simultaneously fitting the optical measured data of the stacks with the theoretical model for the purpose of determining the local parameters (see in particular independent claims 1 and 9). Consequently, the common parameters are not extracted from the process of simultaneous fitting the optical measured data of the stacks with the theoretical model.

In addition, as also submitted by the appellant, there is no clear explicit disclosure in document E1 on how the common parameters are determined in the second processing method disclosed in document E1.

4.3.3 In its decision the opposition division - while acknowledging that document E1 did not disclose specifically in detail how the values of the global parameters (P3 and P4) were obtained and also acknowledging that the common parameters were not fit simultaneously with the local parameters, but fixed in the parallel processing by regression resulting in the determination of the local parameters - held that it would appear that the common parameters were established in another part of an overall regression model before the local parameters were fitted and determined. The opposition division referred in this respect to several passages of document E1 (sentence bridging the two columns on page 4; page 4, second column, lines 13 to 16; paragraph [0028]; paragraph [0021]; and paragraph [0033], lines 9 to 18) which, in its view, would support, or would be consistent with, or would at least not contradict its interpretation of the disclosure of document E1.

However, in the board's view none of the passages of document E1 referred to by the opposition division constitute a direct and unambiguous disclosure of the determination of the common parameters in a preliminary stage of an overall regression model involving optical measured data and the theoretical model. In document E1 the process of simultaneously fitting the optical measured data of the two stacks with the theoretical model results in the determination of the local parameters. Even if the skilled person would interpret the passages mentioned by the opposition division in the sense that the common parameters would be determined in a previous regression model before being fed to the fitting process involving the determination of the local parameters, there is then no direct and unambiguous disclosure in document E1 that the mentioned previous regression model carried out for determining or extracting the common parameters would also specifically involve a simultaneous fitting of the optical measured data of the two stacks with the theoretical model.

4.3.4 The respondent submitted that the teaching in paragraphs [0016] to [0022] of document E1 anticipated claim 1. In particular, he submitted that paragraph [0019] proposed a parallel processing of multiple data sets using parameter globalization, that in paragraphs [0020] and [0021] and in Fig. 2(b) together with the corresponding description the common or global parameters were shared between the data sets, that according to the last sentence of paragraph [0018] the simultaneous fitting of the two data sets sharing a common parameter reduced the number of parameters to be fit, that the common parameters were referred to in the passages at lines 4 to 9 of paragraph [0018] and at lines 11 to 15 of paragraph [0020] as "fitting parameters", and that the fitting process involved the reduction or minimization (see document A2, chapter 19.2.1.3) of the total residue Rtotal defined in paragraph [0021], and therefore also the common parameters. In addition, the sentence bridging the two columns on page 4 stated that the common parameters "can be fit at the same time". According to the respondent it was therefore clear that the common parameters were also fitted in the parallel processing of the optical measured data to thereby result in an estimate for the corresponding values and in a characterization of the stacks.

However, the mere fact of using parameter globalization in a parallel processing of measured data sets and sharing the common parameters between the data sets does not necessarily imply that the common parameters are used as variables in the parallel fitting processing involving the optical measured data of the stacks and the mathematical model. In addition, the passage in paragraph [0018] of document E1 only states that the number of parameters for each data set "can be minimized by sharing the global [i.e. the common] parameters between data sets, allowing for fewer parameters to be fit during the regressions", and the passage in paragraph [0020] states that there is at least one "global fitting parameter" in the data sets, but these passages are silent as to how the processing and regressions are performed, and in particular silent as to how the parameters, and more particularly the common parameters, are handled in the regressions and/or fitted. Analogous considerations apply to the minimization of the total residual for the data sets referred to in paragraph [0021] of document E1 since, as submitted by the appellant, this passage is silent as to the common and the local parameters being simultaneously fitted.

In addition, the reference in the sentence bridging the two columns on page 14 to the common parameters being "fit at the same time" does not amount to a direct and unambiguous disclosure that the common parameters are fitted in a previous processing specifically involving the optical measured data of the two stacks and the mathematical model. Furthermore, an interpretation of the mentioned sentence in the sense that the common parameters are fitted together with the local parameters would be at variance with the subsequent passage according to which the regression model can be "adjusted" so that the values of the common parameters would be accurate within an acceptable amount (document E1, page 4, right column, lines 13 to 17), also at variance with the subsequent passage according to which the regressions done for the local parameters of each of the stacks are reduced to two fitting parameters instead of four (page 4, right column, lines 17 to 23; see also paragraph [0026]) and, in addition, in contradiction with the claims of document E1 (see in particular independent claims 1 and 9) which require that the common parameters are held fixed when performing the parallel regression analysis involving the theoretical model and the optical measured data of the stacks.

Therefore, none of the passages mentioned by the respondent disclose in a direct and unambiguous manner that the common parameters are also fitted when the optical measured data of the two stacks is being simultaneously fitted with the theoretical model.

4.3.5 The board concludes that the claimed feature relating to extracting the at least one common parameter as claimed, i.e. as a result of the simultaneous fitting of the optical measured data of the stacks with the theoretical model (see point 4.3.1 above, second paragraph), is not directly and unambiguously disclosed in document E1 - irrespectively of whether or not other features of the claimed method, and in particular the feature referred to in point 4.2 above, second sub-paragraph, are anticipated by document E1. Therefore, claim 1 is novel over the disclosure of document E1.

4.4 During the oral proceedings held before the board the respondent did not raise lack of novelty of the method of claim 1 in respect of any other document on file.

4.5 In view of the above considerations, the board concludes that the subject-matter of claim 1 is new over the documents considered during the proceedings (Articles 52(1) and 54(1) EPC).

5. Main request - ground for opposition under Article 100(a) together with Article 56 EPC - Inventive step of claim 1

5.1 It was undisputed by the parties that document E1 represented the closest state of the art.

5.2 As regards the formulation of the objective problem solved by the claimed method over the disclosure of document E1 in view of the distinguishing feature mentioned in point 4.3.5 above, the respondent submitted that claim 1 represented a mere juxtaposition of features and that consequently the claimed method did not solve any technical problem.

This line of argument is based on the respondent's interpretation of claim 1 referred to in point 4.3.1 above, first paragraph, and the board cannot follow this construction of the claimed method for the reasons already given in the second paragraph of point 4.3.1 above.

The respondent also submitted that the objective problem was to be formulated in terms of improving the speed in the characterization of the properties of the article, or in obtaining a faster, simpler and cheaper determination of the common parameters.

The appellant for its part submitted that the objective technical problem was to provide a more accurate determination of the common or global parameters.

The board adheres to the formulation proposed by the appellant because, while the extraction, and therefore the determination, of the common parameters as claimed results in a better fit of the common parameters to the optical measured data, this feature does not necessarily leads to a speedier characterization of the article, or to a faster, simpler and cheaper determination of the common parameters.

5.3 According to the respondent, the skilled person would consider a better characterization of the properties of the stacks and would recognise that there was a need for verifying whether the parameters of the manufactured stacks conformed with the expected values. In addition, document E1 already proposed to perform the best fit comparison to the empirical results (paragraph [0009]) and to apply its teaching to other iterative algorithms where multiple data sets were processed in parallel by combining common parameters between data sets (paragraph [0036]). The skilled person would therefore receive a clear suggestion to also consider the determination of the values of the common parameters. In addition, consideration of the common parameters in the fitting process was also straightforward in view of the approach disclosed in paragraph [0021] of document E1 involving the minimization of a total residual, and in view of the indication in paragraphs [0018] (lines 4 to 9) and [0025] (sentence bridging the two columns) that the common parameters were parameters that could be fitted.

However, as submitted by the appellant, document E1 focuses on the determination of the values of the local parameters of two stacks by simultaneously fitting the measured data with the theoretical model, and in this context neither the general teaching relating to performing a best fit comparison to the empirical results, nor that relating to processing in parallel multiple data sets by combining common parameters between the data sets suggest in the board's opinion also determining the values of the common parameters by setting these parameters as variables when simultaneously fitting the optical measured data of the stacks with the theoretical model. In addition, as also submitted by the appellant, document E1 aims at the reduction of parameters to be considered for the purpose of reducing the complexity of the regression or fitting process (paragraph [0020]), and the document only teaches "adjusting" the regression model so that the values of the global parameters are accurate within an acceptable amount (page 4, right column, lines 13 to 17). In this context, the inclusion of the common parameters as additional variables in the regression analysis involving the local parameters would be at variance with the teaching of document E1.

For analogous reasons, the further argument of the respondent that, in view of the need disclosed in document E1 for a rapid evaluation of the parameters (paragraph [0007], last sentence), it would be obvious to fit all the parameters together, also fails to convince the board.

5.4 According to an alternative line of argument of the respondent, when using fitting algorithms to fit variables to data sets, it was common general knowledge to expand or enrich the data sets to obtain improved values for the variables (see document A2), and it was therefore obvious for the skilled person to improve the accuracy in the determination of the parameters by the method disclosed in document E1 by adding more optical measured data.

However, the fact of feeding additional optical measured data in a process of fitting the optical measured data relating to objects characterized by parameters with a theoretical model involving said parameters for the purpose of determining the values of the parameters might possibly improve the accuracy in the determination of the parameters, but it has no impact on the specific way the parameters themselves - and in particular the local and the common parameters considered in document E1 - are being handled in the fitting process. Therefore, also this line of argument fails to convince the board.

5.5 During the oral proceedings held before the board the respondent did not raise any additional line of argument in respect of the issue of inventive step of the claimed method.

5.6 In view of the above considerations, none of the lines of argument submitted by the respondent allows the conclusion that the method defined in claim 1 as granted is obvious with regard to the prior art, and the board concludes that the method of claim 1 involves an inventive step over the documents considered during the proceedings (Article 52(1) and 56 EPC).

6. Main request - grounds for opposition under Article 100(a) together with Article 54 or 56 EPC - Claims 2 to 15

6.1 Independent claim 15 is directed to a system for use in characterizing properties of an article, the system comprising structural means the functional features of which correspond in substance to the steps of the method of claim 1. In particular, the claimed system comprises a processor utility configured and operable for simultaneously fitting the optical measured data and extracting the at least one common parameter.

The respondent did not raise any further line of argument of lack of novelty or lack of inventive step other than those already raised in respect of claim 1, and the board is of the opinion that the system of independent claim 15 is new and involves an inventive step for analogous reasons to those given in points 4 and 5 above in respect of the method of claim 1.

6.2 Claims 2 to 14 are dependent claims referring back to claim 1 and, by virtue of the reference to claim 1, the corresponding subject-matter is also new and involves an inventive step.

7. Alleged procedural violations

During the appeal proceedings the appellant submitted in support of the admissibility of some of the auxiliary requests that the first-instance proceedings were tainted by procedural violations. As confirmed by the appellant during the oral proceedings, these submissions concerned only some of the auxiliary requests, and not the main request. In addition, as already noted by the board in the communication annexed to the summons to oral proceedings, the alleged procedural violations would not justify the remittal of the case to the opposition division under Article 11 RPBA and/or the reimbursement of the appeal fee in the event that the appeal would be allowed. Therefore, there is no need to consider the allegations of procedural violation submitted by the appellant.

8. In view of the above considerations, the board concludes that none of the grounds for opposition raised by the respondent in respect of the patent as granted prejudices the maintenance of the patent as granted and that therefore the opposition must be rejected (Article 101(2) EPC).

Order

For these reasons it is decided that:

1. The decision under appeal is set aside.

2. The opposition is rejected.