1. A method for determining three-dimensional (3D) information of a structural element of a substrate, the method comprising:

generating or receiving one or more properties of the structural element of the substrate, wherein the one or more properties are determined based on a Scanning Electron Microscope (SEM) image of the structural element;

searching for a correlation model from a plurality of models, the correlation model being estimated to predict the 3D information of the structural element of the substrate with at least a predetermined accuracy;

predicting the 3D information of the structural element of the substrate using the correlation model when the correlation model is found; and

in response to failing to find the correlation model.

2. The method of claim 1, wherein responding to the failure comprises computing a new model based on the substrate.

3. The method of claim 1, wherein different models represent different classes of training substrates.

4. The method of claim 3, comprising classifying the different training substrates by applying a classification process.

5. The method of claim 3, wherein the classifying is based at least in part on manufacturing information related to the different training substrates.

6. The method of claim 1, wherein a model represents a class of training substrates that are mutually predictable.

7. The method of claim 6, comprising verifying that the training substrate is included in the certain class of training substrates.

8. The method of claim 1, wherein the searching comprises determining an accuracy of the prediction of at least some of the plurality of models based on a relationship between (a) the one or more properties of the structural element of the substrate and (b) one or more properties associated with each of the plurality of models.

9. The method of claim 1, wherein the searching comprises determining a confidence level of a prediction related to each of the models.

10. The method of any one of claims 1 to 9, wherein the one or more properties of the structural elements of the substrate are generated by compensating for differences in the acquisition of SEM images of different sites of the structural elements of the substrate.

11. The method according to any one of claims 1 to 9, wherein the property of the structural element of the substrate represents information about a set of structural elements of the substrate.

12. The method of claim 11, wherein the set of structural elements of the substrate belong to a single die of the substrate.

13. The method of any of claims 1-9, wherein the one or more attributes are determined based at least in part on information about a manufacturing process of the substrate.

14. A non-transitory computer-readable medium for determining three-dimensional (3D) information of a structural element of a substrate, the non-transitory computer-readable medium storing instructions for:

generating or receiving one or more properties of the structural element of the substrate, wherein the one or more properties are determined based on a Scanning Electron Microscope (SEM) image of the structural element;

searching for a correlation model from a plurality of models, the correlation model being estimated to predict the 3D information of the structural element of the substrate with at least a predetermined accuracy;

predicting the 3D information of the structural element of the substrate using the correlation model when the correlation model is found; and

in response to failing to find the correlation model.

15. A system for determining three-dimensional (3D) information of a structural element of a substrate, the system comprising a processor;

wherein the processor is configured to:

generating or receiving one or more properties of the structural element of the substrate, wherein the one or more properties are determined based on a Scanning Electron Microscope (SEM) image of the structural element;

searching for a correlation model from a plurality of models, the correlation model being estimated to predict the 3D information of the structural element of the substrate with at least a predetermined accuracy;

predicting the 3D information of the structural element of the substrate using the correlation model when the correlation model is found; and

in response to failing to find the correlation model.

16. The system of claim 15, wherein the system comprises an imager configured to acquire the SEM image.

Background

Three-dimensional (3D) metrology is a new area in the semiconductor industry. The shrinkage of planar devices has reached its physical limit and advanced nodes have turned to 3D designs to increase feature density in the devices. Reliable measurement of these 3D structures is crucial to their development process.

Currently, Optical Critical Dimension (OCD) occupies the largest share in 3D non-destructive measurements. However, OCD is limited to measurements on specially designed peripheral targets and cannot be measured on-die. Furthermore, OCDs have low spatial resolution (on the order of 50 μm), can only measure periodic structures, are sensitive to underlying layers, and involve complex and time-consuming recipe setups.

There is an increasing need to provide accurate methods for determining 3D information of structural features of a substrate.

Disclosure of Invention

Methods, non-transitory computer-readable media, and systems for determining 3D information may be provided.

Drawings

The subject matter regarded as embodiments of the disclosure is particularly pointed out and distinctly claimed in the concluding portion of the specification. The embodiments of the disclosure, however, both as to organization and method of operation, together with objects, features, and advantages thereof, may best be understood by reference to the following detailed description when read with the accompanying drawings in which:

FIG. 1 shows an example of a method;

FIG. 2 shows an example of a method;

FIG. 3 shows an example of a method;

FIG. 4 shows an example of a method and various Scanning Electron Microscope (SEM) images and data structures;

FIG. 5 shows an example of a wafer, die, and SEM image; and is

Fig. 6 shows an example of a wafer and system.

Detailed Description

In the following detailed description, numerous specific details are set forth in order to provide a thorough understanding of embodiments of the present disclosure.

However, it will be understood by those skilled in the art that embodiments of the present disclosure may be practiced without these specific details. In other instances, well-known methods, procedures, and components have not been described in detail so as not to obscure the embodiments of the disclosure.

The subject matter regarded as embodiments of the disclosure is particularly pointed out and distinctly claimed in the concluding portion of the specification. The embodiments of the disclosure, however, both as to organization and method of operation, together with objects, features, and advantages thereof, may best be understood by reference to the following detailed description when read with the accompanying drawings.

It will be appreciated that for simplicity and clarity of illustration, elements shown in the figures have not necessarily been drawn to scale. For example, the dimensions of some of the elements may be exaggerated relative to other elements for clarity. Further, where considered appropriate, reference numerals may be repeated among the figures to indicate corresponding or analogous elements.

Because the illustrated embodiments of the present disclosure may be implemented to a great extent using electronic components and circuits known to those skilled in the art, details will not be explained in any greater extent than that considered necessary as illustrated above, for the understanding and appreciation of the underlying concepts of the present embodiments of the disclosure and in order not to obfuscate or distract from the teachings of the present embodiments of the disclosure.

Any reference in this specification to a method shall apply mutatis mutandis to a system capable of performing the method, and shall apply mutatis mutandis to a computer-readable medium that is non-transitory and stores instructions for performing the method.

Any reference in this specification to a system should be applied mutatis mutandis to methods executable by the system and to computer-readable media that are non-transitory and store instructions executable by the system.

Any reference in this specification to a non-transitory computer readable medium should be taken with the necessary modifications to a method applicable when executing instructions stored in the computer readable medium, and to a system configured to execute instructions stored in the computer readable medium.

The term "and/or" means additionally or alternatively.

By structural element is meant a nanoscale structural element such as, but not limited to, a transistor, a portion of a transistor, a memory cell, a portion of a memory cell, an arrangement of conductors, an arrangement of insulators, and the like.

Systems, methods, and non-transitory computer-readable media may be provided for determining three-dimensional (3D) information of a structural element of a substrate.

It should be noted that the systems, methods, and non-transitory computer-readable media can be applied, mutatis mutandis, to determine information other than 3D information related to a structural element of a substrate. The information other than the 3D information may be information that is directly and accurately determined from the SEM image.

The substrate may be a wafer, a MEMS substrate, a solar panel, or the like.

In various examples, for simplicity of explanation, it will be assumed that the substrate is a wafer.

Fig. 1 shows a method 100 for determining 3D information of a structural element of a substrate.

The method 100 may begin at step 105 where a plurality of models are generated or received. The plurality of models are generated during a training process. The model may represent a relationship between one or more SEM images of the structural element and 3D information about the structural element.

Step 105 may include at least one of:

o classifying the different training substrates by applying a classification process and generating a model for each class.

And o classifying the different training substrates based on the predicted estimation accuracy of the different models.

The different training substrates are classified based on information about process parameters of the different wafers.

Step 105 may be followed by step 110 of generating or receiving one or more properties of the structural elements of the substrate.

The one or more properties are determined based on SEM images of the structural elements.

Accordingly, step 110 may include generating one or more attributes of the structural element. Additionally or alternatively, step 110 may include receiving one or more attributes of the structural element.

Step 110 may include compensating for SEM image acquisition process limitations. For example, a relatively low signal-to-noise ratio of a single SEM image is compensated.

The property of the structural element of the substrate may represent information about a set of structural elements of the substrate. For example, the attributes may be calculated per die.

The set of structural elements may include all or only some of the structural elements of a single die of the substrate. The one or more attributes may be generated based on a model and/or based on machine learning techniques.

The set of structural elements may comprise all or only some of the structural elements of the substrate.

The set of structural elements may comprise all structural elements represented by the same 3D information unit.

The set of structural elements may be multiple instances of the same structural element.

Step 110 may be followed by step 120 of searching a correlation model from a plurality of models, the correlation model being estimated as 3D information of the structural element of the substrate with at least a predetermined accuracy.

The predefined accuracy may be determined by a substrate manufacturer, a metrology system operator, and/or in any other manner.

The model may be generated by applying machine learning.

At least some of the models may be generated during a training process. The training process may include receiving or generating one or more properties of the structural elements of the training substrate, and also receiving or generating 3D information about at least some of the structural elements of the training substrate.

The 3D information may be provided in various ways, such as by milling the die and obtaining a Transmission Electron Microscope (TEM) image of the milled die.

During the training process, a model may be generated based on: (a) one or more properties of the structural elements of the training substrate, and (b) 3D information about at least some of the structural elements of the training substrate.

Different models may represent different classes of training substrates. The categories may be determined based on the prediction accuracy obtained by using different models.

The model may be associated with one or more classes of training substrates.

The model should predict 3D information about substrates belonging to a training substrate class associated with the model with at least the predetermined accuracy.

Thus, a model associated with a particular category may not be able to predict, with at least the predetermined accuracy, 3D information relating to substrates that are not associated with the particular category.

The prediction accuracy of a model with respect to a particular substrate may be determined based on a relationship between (a) one or more properties of structural elements of the particular substrate and (b) one or more properties of structural elements of a training substrate associated with the model.

The search for the relevant model may be a matching process that examines the relationship between (a) one or more properties of the structural elements of the particular substrate and (b) one or more properties of the structural elements of the training substrate associated with the model.

The relationship may be, for example, a similarity between one or more attributes, or the like.

If the similarity is below a predetermined threshold, the prediction accuracy of the model may be considered insufficient.

Various matching processes may be applied, such as a sum-of-squared differences Search (SD), a Mahalanobis distance calculation, a KL divergence, or any other regression process.

Using different models for different categories improves the accuracy of the method and has been found to overcome problems such as model overlap and multimodality. Model overlap occurs when different substrates with similar one or more properties are mapped to different 3D information values. Multimodal occurs when different substrates with different one or more properties are mapped to the same 3D information value.

The search (step 120) may have several results.

The first result is to find the correlation model. If multiple correlation models are found, one of the found correlation models may be selected.

The first result results in a step 130 of predicting 3D information of the structural element of the substrate with the associated model.

The second result is the failure to find a correlation model. In this case, it is assumed that none of the plurality of correlation models accurately predicts (with at least the predetermined accuracy) the 3D information of the structural element of the substrate.

The second result results in step 140 in response to failing to find a correlation model.

Step 140 may include at least one of:

o prevents predicting 3D information of structural elements of the substrate.

o predict the 3D information of the structural elements of the substrate, but assign a low level of certainty to the prediction.

o calculating a new model based on the substrate. This would require obtaining 3D information of the substrate.

o requesting calculation of a new model based on the substrate.

o generating a failure indication.

Fig. 2 shows an example of a method 200.

Suppose that:

the o-training substrate is a training wafer.

Each training wafer includes a set of dies.

o for each wafer, the first set of dies has 3D information, while the second set of dies does not have 3D information.

Each die in the first set of dies is represented by a 3D information unit.

The properties of the structural elements of the o-substrate represent the structural elements of the entire die.

Under these assumptions, a model associated with the training wafer is generated based on the one or more attributes and the 3D information about the first set of dies. During the training step, attributes may be determined. The attribute(s) (at least some) and known 3D information (ground state) values and optionally any process information are used to generate a plurality of models. The training step involves determining the attributes themselves, and using all attributes from all training substrates (wafers).

During the testing or inferring step, the attributes of the wafer being tested are used to find the correlation model. If such a correlation model is found, the attributes (all or a subset) of a particular structure may be fed to the correlation model.

The search for the correlation model may be based on one or more attributes associated with the entire wafer.

The method 200 may begin at step 205 where a plurality of models are generated or received. The plurality of models are generated during a training process.

Step 205 may be followed by step 210 of generating or receiving one or more properties of a structural element of a wafer.

Step 210 may be followed by step 220 of searching a correlation model from a plurality of models, the correlation model being estimated to predict 3D information of the structural elements of the certain wafer with at least a predetermined accuracy.

Step 220 may include determining a relationship between (a) one or more properties of the structural elements of a certain wafer and (b) one or more properties of the structural elements of test wafers associated with different classes of training wafers.

The relationship may represent a similarity between (a) one or more properties of structural elements of a wafer and (b) one or more properties of structural elements of test wafers associated with different classes of training wafers.

If for a certain category, the similarity is above a predetermined similarity threshold, the model associated with the certain category may be considered the relevant model.

When the correlation model is found, step 220 is followed by step 230 of predicting 3D information of the structural elements of the certain wafer using the correlation model.

Otherwise, step 220 is followed by step 240 of responding to the failure.

Fig. 3 illustrates a method 300 for generating one or more properties of a structural element of a die.

Method 300 illustrates various steps that increase the signal-to-noise ratio of SEM images and compensate for possible SEM-image-induced variations between SEM images of different sites (such as electron beam intensity illuminating the site, or other variations during illumination and/or collection of electrons). Other steps may be provided.

The method 300 may begin at step 310 with receiving or generating SEM images of a plurality of sites of a wafer. The multiple sites may cover the entire wafer or may cover only one or more portions of the wafer.

Step 310 may be followed by a step 320 of locating a patch (patch) comprising a structural element of interest, in particular a patch comprising multiple instances of certain structural elements in the SEM image. The patch may be a two-dimensional patch.

Step 320 may be followed by step 330 of averaging the patches for each site to provide a patch for each site that is averaged.

Step 330 may be followed by step 340 of converting the patches averaged per site to a site vector. This may be accomplished, for example, by averaging each pixel column of the patch for bit point averaging to provide a bit vector element for each column.

Step 340 may be followed by step 350 of normalizing each location vector.

Step 350 may be followed by step 360 of averaging the site vectors for each die to provide a die vector for each die.

Step 360 may be followed by step 370 of generating wafer properties representing a die vector for each die. The wafer property may be a substrate.

Fig. 4 shows an example of a method 400, along with various SEM images and data structures.

Assume (for ease of explanation only) that there are two classes of training wafers and two training modules: a first model 405 and a second model 406. The first class of training wafers is represented by first 3D information 401 and a first attribute 403. The second category of training wafers is represented by second 3D information 402 and second attributes 404. There may be more than two categories.

One or more new SEM images 410 of new wafers are received.

One or more SEM images are processed by generating one or more attributes 412. The generated one or more attributes are sent to a search for relevant models step 414. If this is followed-step 414 is followed by step 416 of applying a correlation model (which may be the first model or the second model) -for generating 3D information for a new wafer. Otherwise, step 414 is followed by step 418 of requesting generation of a new model based on the one or more SEM images and the 3D information.

Fig. 5 shows an example of an SEM image 510 of wafer 514, die 512, and one of the sites of the die.

Fig. 6 shows a wafer 514 and a system 500 including an imager 710 and a processor 720.

The processor may include one or more processing circuits, such as a microprocessor, a graphics processing unit, a hardware accelerator, a central processing unit, a neural network processor, an image processor, or the like. The processor may be programmed (or otherwise constructed and arranged or configured) to perform any of the steps of any of the methods described in the specification.

The system may further comprise a storage unit, such as a volatile or non-volatile storage unit, for storing information and/or instructions and/or models and/or one or more attributes. The storage unit is an example of a non-transitory computer-readable medium.

The imager 710 may be an electron beam imager, electron beam microscope, ion imager, or the like. The electron beam microscope may be a scanning electron microscope, a transmission electron microscope, or the like.

The system 700 may be configured to perform at least one of the methods 100, 200, 300, and 400.

The imager is configured to generate SEM images and the processor 200 may be configured to perform other steps of at least one of the methods 100, 200, 300, and 400.

In the foregoing specification, embodiments of the present disclosure have been described with reference to specific examples thereof. It will, however, be evident that various modifications and changes may be made therein without departing from the broader spirit and scope of the embodiments of the disclosure as set forth in the appended claims.

Any arrangement of components to achieve the same functionality is effectively "associated" such that the desired functionality is achieved. Hence, any two components herein combined to achieve a particular functionality can be seen as "associated with" each other such that the desired functionality is achieved, irrespective of architectures or intermedial components. Likewise, any two components so associated can also be viewed as being "operably connected," or "operably coupled," to each other to achieve the desired functionality.

Further, those skilled in the art will recognize that the boundaries between the above described operations merely illustrative. Multiple operations may be combined into a single operation, single operations may be distributed in additional operations, and operations may be performed at least partially overlapping in time. Moreover, alternative embodiments may include multiple instances of a particular operation, and the order of operations may be altered in various other embodiments.

Also for example, in one embodiment, the illustrated examples may be implemented as circuitry located on a single integrated circuit or within a same device. Alternatively, examples may be implemented as any number of separate integrated circuits or separate devices interconnected with one another in a suitable manner.

However, other modifications, variations, and alternatives are also possible. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.

In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of other elements or steps than those listed in a claim. Furthermore, the terms "a" or "an," as used herein, are defined as one or more than one. Furthermore, the use of introductory phrases such as "at least one" and "one or more" in the claims should not be construed to imply that the introduction of another claim element by the indefinite articles "a" or "an" limits any claim containing such introduced particular claim element to embodiments of the disclosure containing only one such element, even when the same claim includes the introductory phrases "one or more" or "at least one" and indefinite articles such as "a" or "an". The same holds true for the use of definite articles. Unless otherwise specified, terms such as "first" and "second" are used to arbitrarily distinguish between the elements such terms describe. Thus, these terms are not necessarily intended to indicate temporal or other prioritization of such elements. The mere fact that certain measures are recited in mutually different claims does not indicate that a combination of these measures cannot be used to advantage.

While certain features of the embodiments of the disclosure have been illustrated and described herein, many modifications, substitutions, changes, and equivalents will now occur to those of ordinary skill in the art. It is, therefore, to be understood that the appended claims are intended to cover all such modifications and changes as fall within the true spirit of the embodiments of the disclosure.

Any combination of any module or element listed in any figure, any part of the specification, and or any claim may be provided. In particular, any combination of any claimed features may be provided.

Any reference to the term "comprising" or "having" should also be construed to mean "consisting of … …" or "consisting essentially of … …". For example, a method that includes certain steps is, respectively, one that may include additional steps, may be limited to only those steps, or may include additional steps that do not materially affect the basic and novel characteristics of the method.

The embodiments may also be implemented in a computer program for running on a computer system, the computer program comprising at least code portions for performing steps of a method according to the embodiments when run on a programmable apparatus, such as a computer system, or enabling a programmable apparatus to perform functions of a device or system according to the embodiments. The computer program may cause the storage system to assign disk drives to groups of disk drives.

A computer program is a sequence of instructions, such as a specific application program and/or an operating system. The computer program may for example comprise one or more of the following: a subroutine, a function, a program, an object method, an object implementation, an executable application, an applet, a servlet, a source code, an object code, a shared library/dynamic load library and/or other sequence of instructions designed for execution on a computer system.

The computer program may be stored internally on a computer program product, such as a non-transitory computer readable medium. All or some of the computer programs may be provided on a non-transitory computer readable medium permanently, removably or remotely coupled to an information processing system. The non-transitory computer readable medium may include, for example, but not limited to, any number of the following: magnetic storage media, including magnetic disk and tape storage media; optical storage media such as optical disc media (e.g., CDROMs, CDRs, etc.) and digital video disc storage media; non-volatile memory storage media including semiconductor-based memory units such as flash memory, EEPROM, EPROM, ROM; a ferromagnetic digital memory; an MRAM; volatile storage media include registers, buffers or caches, main memory, RAM, etc. A computer process typically includes an executing (running) program or portion of a program, current program values and state information, and the resources used by the operating system to manage the execution of the process. An Operating System (OS) is software that manages computer resource sharing and provides programmers with an interface for accessing these resources. The operating system processes system data and user input and responds to the system's users and programs by allocating and managing tasks and internal system resources as services. The computer system may, for example, include at least one processing unit, associated memory, and a plurality of input/output (I/O) devices. When executing a computer program, the computer system processes information according to the computer program and generates resultant output information via the I/O devices.

The foregoing description includes specific examples of one or more embodiments. It will, however, be evident that various modifications and changes may be made therein without departing from the broader spirit and scope of one or more embodiments as set forth in the appended claims.