An OPC model construction method suitable for FPD multi-layer process
Patent Information
- Application Number
- CN202611224609.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-13
- Publication Date
- 2026-10-23
AI Technical Summary
两类效应叠加会造成当前层关键尺寸出现收缩、异常增大、边缘畸变等尺寸偏移缺陷，直接降低图形良率
1、显著降低传统单一维度建模带来的模型拟合误差，有效缩小仿真轮廓与实际光刻图形的尺寸偏差，大幅提升OPC模型整体精度与通用性。
Smart Images
Figure CN122962879A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of FPD lithography modeling and optical proximity correction technology, specifically to an OPC model construction method suitable for FPD multi-layer processes. Background Technology
[0002] As high-end flat panel display manufacturing processes continue to iterate towards higher resolution and higher density patterns, optical proximity correction (OPC) is a core process for controlling the critical dimensions (CD) of lithographic patterns and ensuring the yield and accuracy of mass production of panels.
[0003] Integrated circuit (IC) manufacturing processes commonly employ chemical mechanical polishing (CMP) to planarize the wafer surface globally, significantly reducing optical interference caused by differences in the height of the underlying thin films and the stacking of patterns. However, flat panel display (FPD) panel manufacturing processes do not involve CMP planarization. The thin films and underlying patterns continue to maintain their original morphological stacking state, resulting in a complex structure with numerous differences in height, multiple overlapping layers, and uneven film thickness.
[0004] In multi-layer stacking exposure conditions, residual patterns on the bottom layer, differences in film thickness, and surface undulations can simultaneously generate two types of optical interference on the photolithography imaging of the current exposure layer: near-field short-range light and shadow interference and far-field long-range structural diffraction effects. The superposition of these two effects can cause dimensional displacement defects such as shrinkage, abnormal enlargement, and edge distortion in the critical dimensions of the current layer, directly reducing the pattern yield.
[0005] Current common OPC modeling solutions in the FPD industry have significant drawbacks: they only perform simple linear fitting for the vertical unidirectional CD attenuation and shrinkage effect caused by the bottom layer, resulting in a single modeling dimension and an inability to distinguish and decouple short-range and long-range independent optical interferences; the models cannot reproduce the long-range diffraction effects caused by the height difference of the bottom layer, and there are prediction blind spots for abnormal CD amplification in cross-layer regions. Traditional models have poor adaptability to multi-layer stacking processes with two or more bottom layers, and the CD simulation prediction deviation is large, making it difficult to meet the high-precision mass production modeling requirements of high-end, high-density display panels. Summary of the Invention
[0006] To address the shortcomings of existing technologies, this application provides an OPC model construction method applicable to multi-layer FPD processes. By using dual feature inputs of the underlying structure, long- and short-range optical interference decoupling and fitting are achieved, thereby improving the simulation accuracy and process versatility of the OPC model under multi-layer stacking processes and adapting it to FPD stacking processes without CMP planarization.
[0007] To achieve the above objectives, this application provides a method for constructing an OPC model applicable to multiple underlying processes in FPD, comprising the following steps: S1. Collect measured data of the current exposure layer's key dimension CD, and clean and standardize the measured CD data to generate a Gauge modeling dataset; S2. Extract short-range optical features and long-range optical diffraction features from each bottom layer pattern; S3. A single-layer bottom-level double-bottom-level input mechanism is adopted. Each bottom-level layer corresponds to a set of short-range optical features and a set of long-range optical diffraction features. If the bottom layers are stacked in a total of N layers, 2N sets of bottom-level features are input into the model to achieve decoupling and fitting of near-field short-range light and shadow interference and far-field long-range structural diffraction effect. The model is iteratively trained until the error converges to obtain a high-precision OPC model. S4. Perform photolithography contour simulation on the entire panel layout. Based on the deviation between the simulated contour and the design target size, iteratively optimize the OPC correction formula to obtain the final OPC correction scheme for mask preparation, and complete the optical proximity correction of the mass production pattern.
[0008] Furthermore, in step S1, for the FPD multilayer stacking exposure process without CMP planarization, the measured data of the current exposure layer key dimension CD is collected. The sample covers the cross-layer area of multiple bottom layers, the cross-layer edge transition area, and the blank area without bottom layers. The measured CD data is cleaned and standardized to generate a Gauge modeling dataset.
[0009] Furthermore, in step S1, the same pattern is matched with three types of Gauge measurement samples: cross-layer sample directly above the bottom layer, cross-layer edge transition sample, and flat sample without bottom layer blank.
[0010] Furthermore, in step S2, the optical effect range is limited by the effLength parameter, and the interference intensity of the underlying layer on the current layer imaging is characterized by the coefficient parameter. effLength is a fixed threshold set manually before modeling. During the model iterative training process, only the coefficient coefficient is adjusted, and the effLength parameter is not modified.
[0011] Furthermore, in step S2, the effLength of the short-range optical features is fixed at 500 nm, and the effLength of the long-range optical diffraction features is fixed at 10000 nm; the coefficients corresponding to each set of features are automatically optimized by the model iterative training.
[0012] Furthermore, in step S3, the model convergence criterion is: the model error of 95% of the simulation samples is controlled within 100nm.
[0013] Furthermore, when the number of underlying stacking layers is two, three, or four, the corresponding inputs are 4, 6, or 8 sets of long-range and short-range feature terms. The change in the number of underlying layers only adjusts the number of feature input groups, while the long-range and short-range dual-input modeling logic remains unchanged.
[0014] To achieve the above objectives, this application also provides an OPC model suitable for FPD multi-layer processes, which is constructed using the OPC model construction method for FPD multi-layer processes described above, and is used for CMP-free multilayer thin film stacking FPD lithography pattern simulation and optical proximity correction.
[0015] To achieve the above objectives, this application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the OPC model construction method applicable to FPD multi-low-level processes as described above.
[0016] To achieve the above objectives, this application also provides a computer-readable storage medium having a computer program stored thereon, the computer program being loaded and executed by a processor to implement the OPC model construction method applicable to multiple underlying processes of FPD as described above.
[0017] The OPC model construction method applicable to multi-layer FPD processes proposed in this application has the following advantages compared with the prior art: 1. Significantly reduces model fitting error caused by traditional single-dimensional modeling, effectively reduces the size deviation between simulation contour and actual lithography pattern, and greatly improves the overall accuracy and versatility of OPC model.
[0018] 2. By decoupling modeling of long and short-range optical effects, the unique dimensional deviation problem of increased CD in cross-layer regions during FPD multilayer stacking process can be accurately predicted, making up for the prediction blind spot of traditional models.
[0019] 3. The model can simultaneously identify and fit the imaging deviations caused by the unique adhesive buildup and morphological undulations of the FPD stacking process, perfectly adapting to the special process characteristics of FPD without CMP planarization.
[0020] 4. The modeling closely matches the real physical imaging laws of FPD, effectively expanding the application scenarios of OPC technology in multi-layer, high-stack, and complex topography panel processes, and improving mass production pattern yield and process stability.
[0021] Other features and advantages of this application will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing this application. Attached Figure Description
[0022] The accompanying drawings are provided to further illustrate the present application and form part of the specification. Together with the embodiments of the present application, they serve to explain the present application but do not constitute a limitation thereof. In the drawings: Figure 1 This is a flowchart of an OPC model construction method applicable to multiple underlying processes of FPD according to an embodiment of this application; Figure 2 This is a schematic diagram illustrating the impact of the underlying structure on the current layer's CD size; Figure 3 Diagram illustrating the requirements for Gauge information input; Figure 4 This is a flowchart illustrating OPC modeling applicable to multiple underlying processes in FPD according to an embodiment of the present invention. Figure 5 This is a schematic diagram of an electronic device structure according to an embodiment of this application.
[0023] The above figures are only used to illustrate the technical solution of this application and are not intended to limit the scope of protection of this application. Detailed Implementation
[0024] The preferred embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit this application.
[0025] Embodiments of this application will now be described in more detail with reference to the accompanying drawings. While some embodiments of this application are shown in the drawings, it should be understood that this application can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this application. It should be understood that the drawings and embodiments of this application are for illustrative purposes only and are not intended to limit the scope of protection of this application.
[0026] The term "comprising" and its variations as used in this application are open-ended, meaning "including but not limited to". The term "based on" means "at least partially based on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". Definitions of other terms will be given in the description below.
[0027] It should be noted that the terms "first" and "second" may be used in this application only to distinguish different devices, components or parts, and are not used to define the order of functions performed by these devices, components or parts or their interdependence.
[0028] It should be noted that the terms "one" and "more" used in this application are illustrative rather than restrictive, and those skilled in the art should understand that, unless explicitly stated otherwise in the context, they should be understood as "one or more". "More" should be understood as two or more.
[0029] Definitions: FPD: Flat Panel Display; CD: Critical Size; CMP: Chemical Mechanical Polishing; effLength parameter: Effective range; coefficient parameter: Interference intensity coefficient; Recipe: OPC modified recipe.
[0030] Example 1 The embodiments of this application use a CMP-free planarization-free, multi-layer thin-film stacked FPD metal layer as a typical modeling object.
[0031] Figure 1 This is a flowchart of an OPC model construction method applicable to multiple underlying processes of FPD according to an embodiment of this application. The following will refer to... Figure 1 This application provides a detailed description of the OPC model construction method applicable to FPD multi-layer process in its embodiments.
[0032] First, in step 101, a full-scene, multi-dimensional Gauge dataset is constructed.
[0033] Taking the CMP-free planarization and multi-layer thin film stacked FPD metal layer as the typical modeling object, batch-collected measured data of the key dimension CD of the current exposure layer; The sample layout fully covers three typical structures: overlapping cross-layer areas with multiple underlying layers, cross-layer edge transition areas, and flat blank areas without underlying layers. The collected CD raw data is cleaned, normalized and standardized to generate a unified modeling metric Gauge dataset; the same map pattern must be accompanied by three types of measurement samples: cross-layer Gauge1, transition area Gauge2, and blank area Gauge3, as the training base for the OPC model.
[0034] Figure 2 This is a schematic diagram illustrating the impact of the underlying structure on the current layer CD size, such as... Figure 2 As shown, the Gauge collection structure is as follows: the left side is the overlapping area of the bottom layer, including Gauge1 (directly above the bottom layer) and Gauge2 (cross-layer transition area); the right side is the flat area without a bottom layer, corresponding to the blank sample Gauge3; the three types of samples together constitute the complete modeling dataset. Figure 2In the design, Pattern1, Pattern2, and Pattern3 are designed with the same CD. Pattern1 is far from the bottom layer, Pattern2 is placed above the bottom layer, and Pattern3 is within the long-range effect range. In actual production, the CD of Pattern3 is greater than that of Pattern1, while the CD of Pattern2 is smaller than that of Pattern1.
[0035] Figure 3 A diagram illustrating the requirements for Gauge information input, such as... Figure 3 As shown, the left diagram shows the overlapping area of the bottom layer, including Gauge1 (measurement position directly above the bottom layer) and Gauge2 (measurement position for transition between layers); the right diagram shows the flat area without a bottom layer, corresponding to the blank measurement position of Gauge3; the three types of Gauges are necessary supporting samples for modeling.
[0036] In step 102, the short-range optical features and long-range optical diffraction features of the underlying layer are extracted.
[0037] Based on ray tracing algorithms, two sets of independent optical features are extracted from each bottom layer of the image: short-range optical features and long-range optical diffraction features; the range and intensity of feature influence are controlled by two core parameters: Effective influence length effLength: limits the range of optical effect. The short-range characteristic effLength is fixed at 500nm, and the long-range characteristic effLength is fixed at 10000nm. Influence coefficient: Characterizes the interference intensity of the underlying structure on the imaging of the current layer, and is automatically optimized by the model during iterative training to match process parameters.
[0038] Understandably, effLeng can be manually set and adjusted in advance according to different factories and different film layer processes, and is not a variable that the model automatically iterates and optimizes. Understandably, the coefficient is the core variable for model iterative optimization: during the modeling and training cycle, the optimal coefficient is automatically found by relying on the residual between the measured CD and the simulated CD of the Gauge. Each set of long-range and short-range features has an independent iterable coefficient.
[0039] In step 103, the model is iteratively trained to obtain a high-precision OPC model.
[0040] A single-layer, double-underlayer input strategy is adopted. The single-layer underlayer inputs two sets of feature terms: short-range features and long-range features. If N layers are stacked at the bottom of the process, then 2×N sets of underlayer feature terms are input into the model to achieve decoupling and fitting of near-field short-range light and shadow interference and far-field long-range structural diffraction effect.
[0041] Taking the four-layer bottom stacking process as an example, a total of 8 independent bottom features are input into the model; the model is continuously trained iteratively, and when the model error of 95% of the simulation samples is controlled within 100nm, the model is judged to have converged, and the high-precision OPC model is built.
[0042] In step 104, OPC simulation is performed on the entire panel layout, and the OPC modified recipe is iteratively optimized based on the simulation results.
[0043] The converged OPC model is used to perform lithography contour simulation prediction on the entire panel layout. Based on the deviation between the simulated contour and the lithography target size, the OPC correction recipe is iteratively optimized in reverse so that the simulated contour fits the target pattern size infinitely. The final OPC correction rule is output to adapt to multiple bottom layer stacking processes for mass production lithography pattern correction.
[0044] Meanwhile, the constructed OPC model suitable for multi-layer FPD processes can accurately match the actual process in terms of resist buildup, edge residue, and diffraction deviations due to height differences. The simulated contour closely matches the actual exposed wafer morphology. The overall model error is significantly reduced, and the Recipe correction accuracy is greatly improved, effectively improving the pattern distortion problem under multi-layer stacking processes and enhancing product mass production yield and process stability.
[0045] This embodiment fully verifies that, for the special process of FPD without planarization and multiple layer stacking, the modeling method of long and short-range layered input is more in line with the real physical mechanism of photolithography, and has strong engineering applicability and mass production promotion value.
[0046] Example 2 This embodiment takes the mass production modeling scenario of a four-layer bottom-layer stacked metal layer in a high-end display panel as an example. This process does not involve CMP planarization, resulting in significant undulations in the film layer stacking and a high proportion of complex cross-layer structures. It solves the problems of cross-layer CD prediction distortion and inability to fit the adhesive deposition morphology in traditional models. The specific implementation process of this invention is as follows: Step 1: Layout sampling. Design test layouts with three types of structures: multi-layer overlap, edge transition, and blank flatness. Collect three types of CD test data: Gauge1 (graphics directly above the bottom layer), Gauge2 (graphics with cross-layer edge transition), and Gauge3 (graphics without bottom layer blankness). After removing abnormal noise data, standardize and generate Gauge modeling dataset.
[0047] Step 2: Optical feature extraction. For the four bottom layers Layer1 to Layer4, short-range features (effLength=500nm) and long-range features (effLength=10000nm) are extracted for each layer, generating a total of 8 sets of Underlayer feature terms. Each set of features is matched with an iteratively optimized coefficient.
[0048] Step 3: Model training and fitting. The Gauge dataset, exposure machine parameters, light source information, photoresist parameters, and 8 sets of low-level long and short-range features are input into the OPC modeling framework. The coefficients of each set of features are continuously adjusted iteratively. The model converges when the simulation error of 95% of the samples is ≤100nm.
[0049] Step 4: Mass production simulation verification. The convergence model is used to simulate the entire product layout, which can simultaneously restore the imaging deviation caused by the bottom layer height difference, glue buildup, and edge residue, and accurately predict the abnormal increase of CD in the cross-layer area. Based on the simulation contour deviation, the OPC correction recipe is repeatedly optimized to eliminate pattern shrinkage, enlargement, and distortion defects, which are used for lithography correction in panel mass production.
[0050] Underlying layer number adaptation instructions: If the number of underlying layers in the process is 2 or 3, only the number of underlayer input groups needs to be adjusted to 4 or 6 respectively. The long and short range dual input logic remains unchanged, and it is universally compatible with all multi-underlying FPD stacking processes.
[0051] Figure 4 This is a flowchart illustrating OPC modeling applicable to multiple underlying processes in FPD according to an embodiment of the present invention, such as... Figure 4 As shown, the complete data flow from the original input data to optical computation, low-level double feature input, and finally to model iteration convergence only represents the model building and training process, excluding subsequent layout simulation and OPC-Recipe iterative correction stages. (See figure.) 1. Data (Data Source): Gauge: Gauge dataset (actual measured ground values of CD samples); GDS: Layout geometry data (current layer + all underlying graphics).
[0052] 2. Optical, optical computing module: Inputs: Gauge, GDS, exposure unit parameters, light source information; Function: Performs Hopkins imaging and ray tracing optical calculations, outputs basic optical information, and is divided into two main branches: Resist: Photoresist response model branch; Underlayer input: Low-level feature input branches.
[0053] 3. Underlayer Input (Core: Double feature input from a single bottom layer) Note: For each underlayer problem, enter the term twice; for example, for 4 layers, enter the term 8 times. (1) Underlayer1 short range: effLength=500nm, model iterative optimization coefficient (weight coefficient); (2) Underlayer1 long range: effLength=10000nm, model iterative optimization coefficient (weight coefficient); (3) ... up to Underlayer4 short range and Underlayer4 long range.
[0054] 4. Iterative convergence: The Resist photoresist term and all 8 groups of Underlayer long-range and short-range feature terms are fed into the model together; the residuals are obtained by using the ground truth values of the Gauge dataset CD, and the weights of each coefficient are iteratively updated. The model training converges after the error reaches the target.
[0055] Figure 3 In the process, the Data module takes Gauge dataset and GDS layout data as input; the Optical computing module receives data from the Data module, exposure machine parameters, and light source information, and outputs Resist photoresist model terms and Underlayer input terms, respectively; for each bottom layer, short-range optical feature terms and long-range optical diffraction feature terms are generated, with the effLength of the short-range term preset to 500nm and the effLength of the long-range term preset to 10000nm; a total of 8 sets of bottom layer feature terms are output from the 4 bottom layers, and the coefficients corresponding to each set of terms are obtained by iterative optimization of the model; the photoresist model terms and all bottom layer feature terms are fed into the model, iterative convergence is performed, and the OPC model training is completed.
[0056] The core innovation of this application lies in the following: During the modeling stage, a dual underlayer input strategy is adopted for each bottom layer structure, corresponding to independent fitting modeling of short-range and long-range optical effects respectively. Specifically, a single bottom layer is input with one short-range effect feature and one long-range effect feature; for a four-layer stacked structure, a total of eight sets of bottom-layer feature inputs are completed, achieving decoupled fitting of short-range and long-range optical interference. This solves the technical defects of FPD OPC modeling technology, such as the inability to decouple the short-range and long-range optical effects of the bottom layer, insufficient simulation accuracy of multi-layer stacking, and the existence of blind spots in cross-layer CD prediction.
[0057] An OPC model was built based on an optimized multi-dimensional feature dataset, accurately simulating and reproducing the real lithography imaging performance of the current layer. This enables high-precision prediction of the pattern size of products using similar stacked processes. The simulation results then guide the iterative optimization of the OPC recipe, ensuring that the simulation contour infinitely matches the lithography target size (PhotoTarget).
[0058] This solution effectively solves the problem that traditional models can only fit the shrinkage deviation of the bottom layer CD. It can accurately predict the special process phenomenon of abnormal CD increase around the cross-layer region, and greatly improve the modeling adaptability and prediction accuracy under multi-layer stacked structures.
[0059] Example 3 In embodiments of this application, an electronic device is also provided. Figure 5 This is a schematic diagram of the electronic device structure according to an embodiment of this application, such as... Figure 5 As shown, the electronic device of this application includes a processor 501 and a memory 502, wherein, The memory 502 stores a computer program, which, when read and executed by the processor 501, performs the steps described above in the embodiment of the OPC model construction method applicable to FPD multi-low-level processes.
[0060] Example 4 In the embodiments of this application, a computer-readable storage medium is also provided, which stores a computer program, wherein the computer program is configured to execute the steps described above in the embodiments of the OPC model construction method applicable to FPD multi-low-level processes.
[0061] In this embodiment, the aforementioned computer-readable storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.
[0062] It will be understood by those skilled in the art that the above descriptions are merely preferred embodiments of this application and are not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.
Claims
1. A method for constructing an OPC model applicable to multiple underlying processes in FPD, comprising the following steps: S1. Collect measured data of the current exposure layer's key dimension CD, and clean and standardize the measured CD data to generate a Gauge modeling dataset; S2. Extract short-range optical features and long-range optical diffraction features from each bottom layer pattern; S3. A single-layer bottom-level double-bottom-level input mechanism is adopted, with each bottom-level layer corresponding to a set of short-range optical features and a set of long-range optical diffraction features. If there are N layers in total, then 2N sets of bottom-layer features are input into the model to achieve decoupling and fitting of near-field short-range light and shadow interference and far-field long-range structural diffraction effect; the model is iteratively trained until the error converges to obtain a high-precision OPC model; S4. Perform photolithography contour simulation on the entire panel layout. Based on the deviation between the simulated contour and the design target size, iteratively optimize the OPC correction formula to obtain the final OPC correction scheme for mask preparation, and complete the optical proximity correction of the mass production pattern.
2. The OPC model construction method applicable to FPD multi-layer process as described in claim 1, characterized in that, In step S1, for the FPD multilayer stacking exposure process without CMP planarization, the measured CD data of the current exposure layer key dimension is collected. The sample covers the cross-layer area of multiple bottom layers, the cross-layer edge transition area, and the flat blank area without bottom layers. The measured CD data is cleaned and standardized to generate a Gauge modeling dataset.
3. The OPC model construction method applicable to multi-layer FPD processes according to claim 2, characterized in that, In step S1, the same pattern is matched with three types of Gauge measurement samples: cross-layer sample directly above the bottom layer, cross-layer edge transition sample, and flat sample without bottom layer blank.
4. The OPC model construction method applicable to FPD multi-layer process as described in claim 1, characterized in that, In step S2, the optical effect range is limited by the effLength parameter, and the interference intensity of the underlying layer on the current layer imaging is characterized by the coefficient parameter. effLength is a fixed threshold set manually before modeling. During model iteration training, only the coefficient coefficient is adjusted, and the effLength parameter is not modified.
5. The OPC model construction method applicable to multi-layer FPD processes according to claim 4, characterized in that, In step S2, the effLength of the short-range optical features is fixed at 500 nm, and the effLength of the long-range optical diffraction features is fixed at 10000 nm; the coefficients corresponding to each set of features are automatically optimized by the model iterative training.
6. The OPC model construction method applicable to FPD multi-layer process as described in claim 1, characterized in that, In step S3, the model convergence criterion is: the model error of 95% of the simulation samples is controlled within 100nm.
7. The OPC model construction method applicable to multi-layer FPD processes according to claim 1, characterized in that, When the number of bottom stack layers is two, three, or four, the corresponding inputs are 4, 6, or 8 sets of bottom long-range and short-range feature terms. The change in the number of bottom layers only adjusts the number of feature input groups, while the long-range and short-range dual-input modeling logic remains unchanged.
8. An OPC model suitable for multi-layer FPD processes, characterized in that, The OPC model is constructed using the method described in any one of claims 1 to 7 for multi-layer FPD process, and is used for CMP-free multilayer thin film stacking FPD lithography pattern simulation and optical proximity correction.
9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor is configured to execute the computer program stored in the memory to implement the OPC model construction method applicable to FPD multi-layer process as described in any one of claims 1 to 7.
10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the OPC model construction method for FPD multi-layer process as described in any one of claims 1 to 7.