Etching effect prediction method and related equipment

Abstract

The invention discloses an etching effect prediction method and related equipment. The etching effect prediction method comprises the following steps: acquiring an input etching parameter; and performing etching effect prediction processing on the input etching parameters by adopting a preset etching effect prediction model to obtain a corresponding etching effect prediction result. According to the technical scheme provided by the embodiment of the invention, the robustness of etching effect prediction can be improved.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of semiconductor manufacturing, and in particular to an etching effect prediction method and related equipment. Background Art

[0002] Plasma etching processes, such as reactive ion etching (RIE), generate plasma at low pressure and introduce an etchant into the plasma. The etchant is ionized or radicalized, and the ions or radicals can react with the material to be etched.

[0003] In a plasma etching process, once the etching process forms an opening or pattern in the etched film and is about to reach the underlying film layer, the etching process must be accurately stopped to avoid damaging the underlying film layer. To control the etching process, the chemical composition of the gas in the plasma processing chamber is analyzed to infer whether the etching process has reached the underlying film layer with a different chemical composition from the etched layer.

[0004] However, the robustness of etching effect prediction of existing plasma etching processes still needs to be improved. Summary of the Invention

[0005] The problem solved by the embodiments of the present invention is to provide an etching effect prediction method and related equipment, which can improve the robustness of etching effect prediction in a plasma etching process.

[0006] To solve the above problems, an embodiment of the present invention provides an etching effect prediction method, comprising:

[0007] Get input etching parameters;

[0008] A preset etching effect prediction model is used to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

[0009] Optionally, the etching effect prediction model is obtained in the following manner:

[0010] Acquire a training sample set including a plurality of training samples, wherein each training sample in the training sample set includes information on etching sample parameters and corresponding etching effects;

[0011] The deep neural network training is performed using the training samples in the training sample set to obtain the etching effect prediction model.

[0012] Optionally, obtaining a training sample set including a plurality of training samples includes:

[0013] Acquire an initial training sample set, where the initial training sample set includes a plurality of initial training samples, and the plurality of initial training samples respectively include information on etching sample parameters and corresponding etching effects;

[0014] Obtaining a correlation score between an etching sample parameter and a corresponding etching effect in each initial training sample of the initial training sample set;

[0015] Initial training samples whose correlation scores between etching sample parameters and corresponding etching effects are within a preset value range are selected from the initial training sample set to form the training sample set.

[0016] Optionally, obtaining a correlation score between an etching sample parameter and a corresponding etching effect in each initial training sample in the initial training sample set includes:

[0017] A preset correlation analysis algorithm is used to obtain a correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample in the initial training sample set.

[0018] Optionally, the etching sample parameters include OES parameters.

[0019] Optionally, the etching effect includes at least one of a critical dimension, a depth, a top angle and a bottom angle of the etching pattern.

[0020] Accordingly, an embodiment of the present invention further provides an etching effect prediction device, comprising:

[0021] an acquisition unit, adapted to acquire input etching parameters;

[0022] The prediction unit is adapted to perform etching effect prediction processing on the input etching parameters using a preset etching effect prediction model to obtain corresponding etching effect prediction results.

[0023] Optionally, the etching effect prediction device also includes: a model training unit, suitable for obtaining a training sample set including multiple training samples, each of the training samples in the training text set includes etching sample parameters and corresponding etching effect information; using the training samples in the training sample set to perform deep neural network training to obtain the etching effect prediction model.

[0024] Optionally, the model training unit is suitable for obtaining an initial training sample set, the initial training sample set including multiple initial training samples, and the multiple initial training samples respectively including information on etching sample parameters and corresponding etching effects; obtaining the correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample in the initial training sample set; selecting the initial training samples whose correlation score between the etching sample parameters and the corresponding etching effect is within a preset numerical range from the initial training sample set to constitute the training sample set.

[0025] Optionally, the model training unit is adapted to adopt a preset correlation analysis algorithm to obtain a correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample of the initial training sample set.

[0026] Optionally, the etching sample parameters include OES parameters.

[0027] Optionally, the etching effect includes at least one of CD, depth, top angle and bottom angle of the etching pattern.

[0028] Accordingly, an embodiment of the present invention also provides a computer device comprising at least one memory and at least one processor, wherein the memory stores one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the etching effect prediction method as described in any one of the above items.

[0029] Accordingly, an embodiment of the present invention further provides a computer program product, including a computer program / instruction, which is used to implement any of the above etching effect prediction methods when executed by a processor.

[0030] Correspondingly, an embodiment of the present invention further provides a storage medium, wherein the storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement any one of the etching effect prediction methods described above.

[0031] Compared with the prior art, the technical solution of the embodiment of the present invention has the following advantages:

[0032] The etching effect prediction method provided by the embodiment of the present invention includes: obtaining input etching parameters; using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

[0033] The etching effect prediction method provided by an embodiment of the present invention uses a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters. The etching effect prediction model can effectively learn the complex nonlinear relationship between the optical emission spectrum (OES) data generated in the plasma etching process and the etching effect. Therefore, regardless of whether the aperture ratio of the etched structure is high or low, the etching effect can be accurately predicted, which is beneficial to improving the robustness of the etching effect prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 It is a flow chart of an embodiment of an etching effect prediction method provided by the technical solution of the present invention;

[0035] Figure 2 It is a structural diagram of an embodiment of a fully connected neural network system in the process parameter optimization method provided by the technical solution of the present invention;

[0036] Figure 3 It is a structural diagram of an embodiment of an intermediate layer of a fully connected neural network system in the process parameter optimization method provided by the technical solution of the present invention;

[0037] Figure 4 It is a structural schematic diagram of an embodiment of an etching effect prediction device provided by the technical solution of the present invention;

[0038] Figure 5 This is a schematic diagram of an optional hardware structure of a computer device provided by the technical solution of the present invention. DETAILED DESCRIPTION

[0039] As can be seen from the background art, the robustness of etching effect prediction of existing plasma etching processes still needs to be improved.

[0040] In order to solve the above technical problems, an embodiment of the present invention provides an etching effect prediction method, comprising: obtaining input etching parameters; using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

[0041] The etching effect prediction method provided by an embodiment of the present invention uses a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters. The etching effect prediction model can effectively learn the complex nonlinear relationship between the optical emission spectrum (OES) data generated in the plasma etching process and the etching effect. Therefore, regardless of whether the aperture ratio of the etched structure is high or low, the etching effect can be accurately predicted, which is beneficial to improving the robustness of the etching effect prediction.

[0042] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0043] Figure 1 The flowchart of an embodiment of the etching effect prediction method provided by the technical solution of the present invention is shown. Figure 1 A method for predicting etching effects may include the following steps:

[0044] Step S110: obtaining input etching parameters;

[0045] Step S120: using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

[0046] Please continue to refer to Figure 1 , execute step S110 to obtain input etching parameters.

[0047] The input etching parameters are obtained to provide a basis for subsequently using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters and obtain corresponding etching effect prediction results.

[0048] In an exemplary embodiment, the etching effect prediction method is used to predict the etching effect of a plasma etching process. During the plasma etching process, the upper film is gradually removed, and when the underlying material is exposed, the gas environment in the plasma undergoes significant changes. This change is caused by the volatile etching by-products released by the underlying material, which directly affects the element composition and concentration in the plasma, thereby causing changes in the spectral characteristics. Therefore, during the plasma etching process, optical emission spectroscopy (OES) determines the composition and characteristics of a substance by measuring the spectrum emitted by the substance under specific conditions. OES can detect elements including metal elements (such as aluminum, copper, iron, etc.), non-metal elements (such as silicon, oxygen, nitrogen, etc.) and elements in volatile compounds that may be produced.

[0049] Accordingly, the OES parameters include information on the composition of elements generated in the plasma etching process, such as CN chemical bonds, CO chemical bonds, NO chemical bonds, Si-O chemical bonds, Si-Cl chemical bonds, Si-OF chemical bonds, CF chemical bonds or CH chemical bonds.

[0050] Please continue to refer to Figure 1 , executing step S120, using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

[0051] A preset etching effect prediction model is used to perform etching effect prediction processing on the input etching parameters. The etching effect prediction model can effectively learn the complex nonlinear relationship between the OES data generated in the plasma etching process and the etching effect. Therefore, regardless of whether the aperture ratio of the etched structure is high or low, the etching effect can be accurately predicted, which is beneficial to improving the robustness of the etching effect prediction.

[0052] In an exemplary embodiment, the etching effect prediction model is a deep neural network model (DNN). A deep neural network is composed of a large number of neurons connected together to form a complex network structure. Its powerful learning and representation capabilities enable it to perform well in various complex tasks.

[0053] Reference Figure 2 The described deep neural network model 20 includes an input layer 210, an intermediate layer 220, and an output layer 230. Accordingly, the step of using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters and obtaining corresponding etching effect prediction results includes: using the input layer 210 to obtain the input etching parameters; using the intermediate layer 220 to extract features of the input etching parameters; and using the output layer 230 to output the etching effect prediction results for the input etching parameters.

[0054] As an example, the deep neural network model is a fully connected application network model. A fully connected neural network is a classic feedforward neural network structure, typically composed of multiple stacked layers of neurons, with each layer of neurons connected to all neurons in the previous layer. This allows the fully connected neural network to perform complex feature extraction and nonlinear transformations, giving the model strong expressive power.

[0055] Accordingly, the intermediate layer 220 includes at least one fully connected neural network layer, wherein each of the at least one fully connected neural network layers includes at least one network block.

[0056] In an exemplary embodiment, referring to Figure 3 The network block 30 described herein includes a cascaded fully connected layer 310 and an activation function layer 320. The fully connected layer 310 receives multiple input features and performs a fully connected operation on the received multiple input features to obtain corresponding multiple fully connected features; the activation function layer 320 performs a nonlinear transformation on the multiple fully connected features to obtain corresponding transformed features.

[0057] The fully connected layer 310 performs a weighted sum operation on the input from the previous layer of neurons and adds a bias term, and then the activation function layer 320 performs a nonlinear transformation operation on the fully connected features from the fully connected layer 310, so that at least one fully connected neural network layer can capture the nonlinear relationship in the data and improve the expressive power of the neural network.

[0058] Common activation functions include Sigmoid activation function, ReLU (Rectified Linear Unit), Tanh activation function, etc. In an exemplary embodiment, the activation function layer 320 is a ReLU activation function.

[0059] In other embodiments, the etching effect prediction model can also be other types of deep neural network models other than the fully connected neural network model, such as a convolutional neural network model, etc. Those skilled in the art can select it according to actual needs and there is no limitation here.

[0060] Correspondingly, the steps of constructing the etching effect prediction model include: obtaining a training sample set including multiple training samples, each training sample in the training text set includes information on etching sample parameters and corresponding etching effects; using the training samples in the training sample set to perform deep neural network training to obtain the etching effect prediction model.

[0061] In an exemplary embodiment, the step of obtaining a training sample set including multiple training samples includes: obtaining an initial training sample set, the initial training sample set including multiple initial training samples, and the multiple initial training samples respectively including information on etching sample parameters and corresponding etching effects; obtaining a correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample in the initial training sample set; and selecting initial training samples from the initial training sample set whose correlation scores between the etching sample parameters and the corresponding etching effects are within a preset numerical range to constitute the training sample set.

[0062] In an exemplary embodiment, the etching sample parameters include OES parameters, and the etching effect includes at least one of a critical dimension (CD), a depth, a top angle, and a bottom angle of an etched pattern.

[0063] The correlation score between the etching sample parameters in the initial training sample and the corresponding etching effect reflects the correlation between the etching sample parameters in the initial training sample and the corresponding etching effect. Specifically, when the correlation score between the etching sample parameters in the initial training sample and the corresponding etching effect is within a preset numerical range, it indicates that the etching sample parameters in the initial training sample have a significant impact on the corresponding etching effect; conversely, it indicates that the etching sample parameters in the initial training sample have a smaller impact on the corresponding etching effect.

[0064] By using the correlation scores between the etching sample parameters in the initial training samples and the corresponding etching effects, the initial training samples in the initial training sample set whose correlation scores are within the preset numerical range are retained, and the initial training samples in the initial training sample set whose correlation scores are not within the preset numerical range are removed. This can effectively improve the quality of the training samples and thereby improve the prediction accuracy of the etching effect prediction model obtained through training.

[0065] The preset value range can be flexibly set according to the need to improve the quality of training samples. As an example, the preset value range is [-0.2, +0.2].

[0066] In an exemplary embodiment, the step of obtaining the correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample of the initial training sample set includes: using a preset correlation analysis algorithm to obtain the correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample of the initial training sample set.

[0067] The correlation analysis algorithm, such as the Pearson correlation coefficient algorithm, the Spearman rank correlation coefficient algorithm, the point-biserial correlation coefficient algorithm, etc., can be selected by those skilled in the art according to actual needs and is not limited here.

[0068] In an exemplary embodiment, the training process of the etching effect prediction model includes: performing an iterative training on the etching effect prediction model with initial weights using a preset number of training data sets each time to complete an adjustment of the weights of the etching effect prediction model; performing multiple iterative trainings to continuously iteratively update the weights of the etching effect prediction model until the loss value of the etching effect prediction model on the preset verification set converges.

[0069] Among them, the process of each iterative training includes: using a preset number of training data to train the etching effect prediction model to be trained, and obtaining a preset number of training results; according to the difference between the training results and the corresponding correct results, respectively using a preset loss function to calculate the loss value; performing network back propagation derivation based on the calculated loss value to obtain the gradient value; according to the gradient value obtained by back propagation derivation, adjusting the weight of the etching effect prediction model once.

[0070] The loss value of the etching effect prediction model on the preset verification set converges, which means that the loss value of the etching effect prediction model on the preset verification set reaches a minimum value.

[0071] For more detailed information about the multiple iterative training processes of the etching effect prediction model, please refer to the iterative training process of the convolutional neural network model in the prior art, which will not be repeated here.

[0072] Correspondingly, an embodiment of the present invention further provides an etching effect prediction device.

[0073] Figure 4 The schematic diagram of the structure of an embodiment of the etching effect prediction device provided by the technical solution of the present invention is shown. Figure 4 An etching effect prediction device 40 may include: an acquisition unit 401, suitable for acquiring input etching parameters; a prediction unit 402, suitable for using a preset etching effect prediction model to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

[0074] In an exemplary embodiment, the etching effect prediction device 40 may further include: a model training unit 403, adapted to obtain a training sample set comprising a plurality of training samples, each of the training samples in the training text set comprising information on etching sample parameters and corresponding etching effects; and performing deep neural network training using the training samples in the training sample set to obtain the etching effect prediction model.

[0075] In an exemplary embodiment, the model training unit 403 is suitable for obtaining an initial training sample set, wherein the initial training sample set includes multiple initial training samples, and the multiple initial training samples respectively include information on etching sample parameters and corresponding etching effects; obtaining the correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample in the initial training sample set; and selecting initial training samples from the initial training sample set whose correlation score between the etching sample parameters and the corresponding etching effect is within a preset numerical range to constitute the training sample set.

[0076] In an exemplary embodiment, the model training unit 403 is adapted to adopt a preset correlation analysis algorithm to obtain a correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample in the initial training sample set.

[0077] In an exemplary embodiment, the etching sample parameters include OES parameters.

[0078] In an exemplary embodiment, the etching effect includes at least one of a CD, a depth, a top angle, and a bottom angle of the etching pattern.

[0079] The etching effect prediction device in the embodiment of the present invention can be used to execute the aforementioned etching effect prediction method, or other functional modules can be used to execute the aforementioned etching effect prediction method. For the etching effect prediction method provided by the embodiment of the present invention, please refer to the detailed description in the previous part, which will not be repeated here.

[0080] Correspondingly, an embodiment of the present invention further provides a computer device, which can implement the etching effect prediction method provided by the embodiment of the present invention by loading the above-mentioned etching effect prediction method in the form of a program.

[0081] refer to Figure 5 , which shows an optional hardware structure diagram of a computer device provided in an embodiment of the present invention. The computer device in the embodiment of the present invention includes: at least one processor 01, at least one communication interface 02, at least one memory 03 and at least one communication bus 04.

[0082] In this embodiment, the number of each of the processor 01 , the communication interface 02 , the memory 03 and the communication bus 04 is at least one, and the processor 01 , the communication interface 02 and the memory 03 communicate with each other via the communication bus 04 .

[0083] The communication interface 02 may be an interface of a communication module for network communication, such as an interface of a GSM module.

[0084] The processor 01 may be a central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the etching effect prediction method of this embodiment.

[0085] The memory 03 may include a high-speed RAM memory, or may also include a non-volatile memory, such as at least one disk memory. The memory 03 stores one or more computer instructions, which are executed by the processor 01 to implement the etching effect prediction method provided in the aforementioned embodiment.

[0086] It should be noted that the above-mentioned computer device may also include other devices (not shown) that may not be necessary for understanding the contents disclosed in the embodiments of the present invention; since these other devices may not be necessary for understanding the contents disclosed in the embodiments of the present invention, the embodiments of the present invention will not introduce them one by one.

[0087] Accordingly, an embodiment of the present invention further provides a computer program product, including a computer program / instruction, which is used to implement the etching effect prediction method described in the embodiment of the present invention when executed by a processor.

[0088] An embodiment of the present invention further provides a storage medium, wherein the storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement the etching effect prediction method provided in the above embodiment.

[0089] The embodiments of the present invention described above are combinations of elements and features of the present invention. Unless otherwise mentioned, elements or features may be considered as optional. Each element or feature may be put into practice without being combined with other elements or features. In addition, embodiments of the present invention may be constructed by combining some elements and / or features. The order of operations described in the embodiments of the present invention may be rearranged. Some configurations of any one embodiment may be included in another embodiment and may be replaced by the corresponding configuration of another embodiment. It is obvious to those skilled in the art that claims that do not have a clear reference relationship to each other in the appended claims may be combined into embodiments of the present invention, or may be included as new claims in amendments after submitting this application.

[0090] The embodiments of the present invention may be implemented by various means such as hardware, firmware, software, or a combination thereof. In a hardware configuration, the method according to the exemplary embodiment of the present invention may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, etc.

[0091] In firmware or software configurations, embodiments of the present invention may be implemented in the form of modules, procedures, functions, and the like. Software codes may be stored in a memory unit and executed by a processor. The memory unit may be located inside or outside the processor and may send and receive data to and from the processor via various known means.

[0092] The above description of the disclosed embodiments will enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein, but is to be construed in the widest possible manner consistent with the principles and novel features disclosed herein.

[0093] Although the present invention is disclosed as above, the present invention is not limited thereto. Any person skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention should be based on the scope defined by the claims.

Claims

1. A method for predicting etching effects, characterized in that: include: Get input etching parameters; A preset etching effect prediction model is used to perform etching effect prediction processing on the input etching parameters to obtain corresponding etching effect prediction results.

2. The etching effect prediction method according to claim 1, wherein: The etching effect prediction model is obtained in the following way: Acquire a training sample set including a plurality of training samples, wherein each training sample in the training sample set includes information on etching sample parameters and corresponding etching effects; The deep neural network training is performed using the training samples in the training sample set to obtain the etching effect prediction model.

3. The etching effect prediction method according to claim 2, wherein: The obtaining of a training sample set including a plurality of training samples comprises: Acquire an initial training sample set, where the initial training sample set includes a plurality of initial training samples, and the plurality of initial training samples respectively include information on etching sample parameters and corresponding etching effects; Obtaining a correlation score between an etching sample parameter and a corresponding etching effect in each initial training sample of the initial training sample set; Initial training samples whose correlation scores between etching sample parameters and corresponding etching effects are within a preset value range are selected from the initial training sample set to form the training sample set.

4. The etching effect prediction method according to claim 3, wherein: The obtaining of the correlation score between the etching sample parameter and the corresponding etching effect in each initial training sample of the initial training sample set includes: A preset correlation analysis algorithm is used to obtain a correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample in the initial training sample set.

5. The etching effect prediction method according to claim 2, wherein: The etching sample parameters include OES parameters.

6. The etching effect prediction method according to claim 1, wherein: The etching effect includes at least one of a critical dimension, a depth, a top angle, and a bottom angle of the etching pattern.

7. An etching effect prediction device, characterized in that: include: an acquisition unit, adapted to acquire input etching parameters; The prediction unit is adapted to perform etching effect prediction processing on the input etching parameters using a preset etching effect prediction model to obtain corresponding etching effect prediction results.

8. The etching effect prediction device according to claim 7, wherein: Also includes: A model training unit, adapted to obtain a training sample set comprising a plurality of training samples, wherein each training sample in the training sample set comprises information on etching sample parameters and corresponding etching effects; The deep neural network training is performed using the training samples in the training sample set to obtain the etching effect prediction model.

9. The etching effect prediction device according to claim 8, wherein: The model training unit is adapted to obtain an initial training sample set, wherein the initial training sample set includes a plurality of initial training samples, and the plurality of initial training samples respectively include information on etching sample parameters and corresponding etching effects; Obtaining a correlation score between an etching sample parameter and a corresponding etching effect in each initial training sample of the initial training sample set; Initial training samples whose correlation scores between etching sample parameters and corresponding etching effects are within a preset value range are selected from the initial training sample set to form the training sample set.

10. The etching effect prediction device according to claim 9, wherein: The model training unit is adapted to adopt a preset correlation analysis algorithm to obtain a correlation score between the etching sample parameters and the corresponding etching effect in each initial training sample of the initial training sample set.

11. The etching effect prediction device according to claim 7, wherein: The etching sample parameters include OES parameters.

12. The etching effect prediction device according to claim 7, wherein: The etching effect includes at least one of a critical dimension, a depth, a top angle, and a bottom angle of the etching pattern.

13. A computer device, characterized in that: The method comprises at least one memory and at least one processor, wherein the memory stores one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the etching effect prediction method according to any one of claims 1 to 6.

14. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instruction is executed by a processor, it is used to implement the etching effect prediction method according to any one of claims 1 to 6.

15. A storage medium, characterized in that: The storage medium stores one or more computer instructions, and the one or more computer instructions are used to implement the etching effect prediction method according to any one of claims 1 to 6.

Images