A nanoscale displacement sensing method based on composite grating structure

Abstract

The present invention discloses a nanometer-scale displacement sensing method based on a composite grating structure. The composite grating structure is constructed. The composite grating structure includes two layers of grating groups. Each layer of the grating group includes a plurality of gratings and an intermediate waveguide layer. The gratings are arranged at equal intervals on one side surface of the intermediate waveguide layer. The intermediate waveguide layers are in contact with each other. When the intermediate waveguide layers move relative to each other, the gratings on the two intermediate waveguide layers produce translational dislocation. One side of the composite grating structure is used as an input end and a normally incident plane wave is used as a source excitation. The other side is used as an output end. Reflection spectra and / or transmission spectra corresponding to different displacements d are obtained at the output end. Fano spectra are obtained based on the reflection spectra and / or transmission spectra. The obtained Fano spectra are used as inputs for deep learning. The displacement corresponding to the current Fano spectra is output by using deep learning, thereby realizing high-precision displacement sensing.

Description

Technical Field

[0001] The present invention belongs to the technical field of micro-nano photonics, and in particular relates to a nano-scale displacement sensing method based on a composite grating structure. Background Art

[0002] With the development of nanotechnology and micro-electromechanical systems (MEMS), high-sensitivity and high-resolution displacement sensors are increasingly used in various fields, and higher requirements are also placed on the sensitivity and resolution of displacement sensors. For example, in the field of environmental monitoring, displacement sensors can provide timely and accurate displacement data, which is helpful for earthquake monitoring and structural health monitoring, and preventing and mitigating the impact of natural disasters. In high-precision processing fields such as semiconductor manufacturing and nanomanufacturing, the application of nanoscale displacement measurement can significantly improve the control accuracy of manufacturing processes and product quality.

[0003] However, the existing displacement sensors still have the following shortcomings:

[0004] (1) Existing structures, such as displacement sensors based on laser interferometers, are relatively complex, bulky, and expensive.

[0005] (2) Existing structures, such as inductive displacement sensors, generally have problems such as narrow measurement range, complex structure, and limited application scenarios.

[0006] (3) Existing displacement sensors realize displacement sensing by detecting physical quantities, which may cause large errors, affect detection accuracy, and have shortcomings such as low frequency band.

[0007] (4) Existing displacement sensors generally have problems such as high material cost and small dynamic detection range.

[0008] (5) Existing structures such as optical fiber displacement sensors still have some defects, such as: small dynamic measurement range, low linearity, poor anti-pollution ability, etc. Summary of the invention

[0009] In order to address the deficiencies in the prior art, the present invention proposes a nanoscale displacement sensing method based on a composite grating structure. By constructing a simple composite grating structure and combining it with deep learning, high-precision detection of displacement sensing at the nanometer level is achieved.

[0010] The technical solution adopted by the present invention is as follows:

[0011] A nanometer-scale displacement sensing method based on a composite grating structure comprises the following steps:

[0012] Step 1, constructing a composite grating structure, wherein the composite grating structure includes two layers of grating groups, each layer of the grating group includes a plurality of gratings and an intermediate waveguide layer, and the gratings 1 are arranged at equal intervals on one side surface of the intermediate waveguide layer; and the other side surfaces of the intermediate waveguide layers are in contact with each other;

[0013] Step 2: When the intermediate waveguide layers move relative to each other, the gratings on the two intermediate waveguide layers are displaced by translation, and the displacement between the gratings is recorded as d;

[0014] Step 3, based on the composite grating structure, one side of the composite grating structure is used as the input end and a normally incident plane wave is used as the source excitation, and the other side is used as the output end, and the reflection spectrum and / or transmission spectrum corresponding to different displacements d are obtained at the output end, and the Fano spectrum line is obtained based on the reflection spectrum and / or transmission spectrum;

[0015] Step 4: Use the Fano spectrum line obtained in step 3 as the input of deep learning, and use deep learning to output the displacement corresponding to the current Fano spectrum line, thereby achieving high-precision displacement sensing.

[0016] Furthermore, the deep learning adopts an artificial neural network (ANN) model, and the artificial neural network (ANN) model includes an input layer, an output layer and three hidden layers.

[0017] Furthermore, the nodes of the hidden layer and the output layer use the ReLU function as the activation function.

[0018] Furthermore, each node data of the output layer is converted into a probability distribution output through the Softmax function.

[0019] Further, the processing method of the artificial neural network (ANN) model input data:

[0020] The displacement d from 0 to 2500 nm was evenly divided into 2501 values ​​with an interval of 1 nm.

[0021] For each value, the reflection spectrum was calculated at 501 frequency points in the frequency range of 15.4 THz to 15.45 THz to obtain a spectral line;

[0022] Each spectral line represents a data sample as a 1×501 data array;

[0023] The data was normalized and the input sequence of 2501 samples was randomly set.

[0024] Furthermore, the structures of the two layers of grating groups in the composite grating structure are identical.

[0025] Furthermore, the grating is made of glass and the intermediate waveguide layer is made of silicon.

[0026] Furthermore, the gratings in the grating group are arranged in a periodic manner, and the period a is 8 μm.

[0027] Furthermore, in the grating group, the width w between the grating and the adjacent grating G The grating thickness is 4 μm. G Both are 10um, the thickness of the middle waveguide layer is h W is 2μm.

[0028] An application of a nanoscale displacement sensing method based on a composite grating structure for displacement sensing in semiconductor manufacturing, ultra-precision engineering, micro-electromechanical systems or aerospace.

[0029] The beneficial effects of the present invention are:

[0030] (1) In the present invention, by constructing a composite structure composed of two periodic gratings and combining the translational misalignment of the two gratings, the inversion and continuous deformation of the Fano spectrum are achieved in the composite structure. Compared with the displacement sensor based on laser interferometer, the structure of the present invention is simpler, easier to implement, and has lower manufacturing cost.

[0031] (2) Based on the above structural design, the present invention applies deep learning methods to displacement sensing. Compared with the existing technology, it can quickly and efficiently realize nano-displacement detection, making displacement sensing detection at the nano-level more accurate and reliable.

[0032] (3) The present invention can be applied to high-precision displacement sensing in the fields of semiconductor manufacturing, ultra-precision engineering, micro-electromechanical systems, and aerospace. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 is a schematic diagram of a composite grating structure with marked parameters;

[0034] Figure 2 is the Q value in the frequency band d = 0 and the energy band;

[0035] Figure 3 are the reflection spectra with different displacements d, where (a) is the transmission spectrum and (b) is the reflection spectrum;

[0036] Figure 4 It is a schematic diagram model of artificial neural network ANN;

[0037] Figure 5 The results of deep learning, where (a) is the average loss with the number of iterations, and (b) is the accuracy with the number of iterations;

[0038] In the figure, 1, grating, 2, intermediate waveguide layer. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0040] A nanometer-scale displacement sensing method based on a composite grating structure comprises the following steps:

[0041] Step 1: construct a composite grating structure; the composite grating structure is as follows: Figure 1 As shown, it includes two layers of grating groups, each layer of grating group includes a plurality of gratings 1 and an intermediate waveguide layer 2, and the plurality of gratings 1 are fixed on one side surface of the intermediate waveguide layer 2; and the gratings 1 are arranged at equal intervals. The intermediate waveguide layers 2 of the two grating groups are in contact with each other, so the two grating 1 groups are symmetrically distributed about the intermediate waveguide layer 2.

[0042] In this embodiment, the specific parameters of each component in the composite grating structure are as follows: grating 1 is glass with a refractive index n G is 1.5; the space between the gratings 1 is air, and the refractive index of air is 1; the intermediate waveguide layer 2 is made of silicon Si, and the refractive index of the intermediate waveguide layer 2 is n W is 3.4.

[0043] In the same grating group, the grating period a is 8μum, where the glass and air widths w G Both are 4μm, the upper and lower grating thickness h G Both are 10um, the thickness of the middle waveguide layer is h W is 2μm, which is reflected in Figure 1 The corresponding parameters in .

[0044] Step 2, moving the intermediate waveguide layer 2, causing relative sliding between the two, so that the gratings 2 in the upper and lower grating groups are displaced, and a displacement d is generated between the gratings in the upper and lower groups;

[0045] Step 3: Based on the composite grating structure with translational dislocation in step 2, one side of the composite grating structure is used as the input end and the plane wave with normal incidence is used as the source excitation, and the other side is used as the output end, and the reflection spectrum and / or the transmission spectrum are obtained at the output end, and the Fano spectrum line is obtained based on the reflection spectrum and / or the transmission spectrum. As the displacement d changes, the Fano spectrum line will reverse and change continuously.

[0046] Step 4: Use the Fano spectrum line obtained in step 3 as the input of deep learning, and use deep learning to output the displacement corresponding to the current Fano spectrum line, thereby achieving high-precision displacement sensing.

[0047] In the present invention, an artificial neural network (ANN) model is adopted, such as Figure 4The model diagram shows that high-precision nano displacement detection is achieved through training of Fano line spectrum data. The specific implementation is as follows:

[0048] The displacement d from 0 to 0.25a (2500 nm) was divided into 2501 values ​​with an interval of 1 nm.

[0049] Furthermore, for each value, the reflection spectrum is calculated at 501 frequency points in the frequency range of 15.4 THz to 15.45 THz to obtain a spectral line.

[0050] Furthermore, each spectral line represents a data sample as a 1×501 data array.

[0051] Furthermore, the data were normalized and the input sequence of 2501 samples was randomly set.

[0052] The structure of the artificial neural network (ANN) model is as follows Figure 4 As shown, it includes an input layer, an output layer, and three hidden layers. The input layer has 501 nodes and the output layer has 2501 nodes.

[0053] The nodes of the hidden layer and the output layer use the ReLU function as the activation function:

[0054]

[0055] Among them, x is each node data.

[0056] Each node data of the output layer is applied to the Softmax function to convert the output into a probability distribution:

[0057]

[0058] Among them, x i is the data of the i-th node in the output layer, and n is the total number of nodes in the output layer.

[0059] The output layer corresponds to 2501 displacement samples, and the data of each node represents the probability of the corresponding displacement sample.

[0060] Furthermore, the cross entropy loss function is used to optimize the weights and biases:

[0061]

[0062] Where C is the total number of output nodes, y i The value is 1 when the input sample is the i-th one, otherwise it is 0.

[0063] The initial learning rate was set to 0.001, and 5000 iterations were performed using the gradient descent optimization technique.

[0064] Furthermore, after each iteration, all samples are input sequentially for testing. If the i-th input sample maximizes the probability of the i-th output node, the test is correct.

[0065] Furthermore, accuracy is the ratio of the number of correct answers to the total number of samples. After the number of iterations exceeds 2400, the accuracy increases dramatically and eventually remains above 99%, as shown in Figure 5 The accuracy increases with the number of iterations.

[0066] The current resolution of the sensing system reaches 1nm, which can be further improved to the sub-nanometer level by increasing the number of samples.

[0067] The mechanism of the nanoscale displacement sensing method based on a composite grating structure proposed in the present invention is as follows: based on the composite grating structure of the present invention, since the composite grating structure produces translational dislocation, it can combine the topological transformation principle of the grating and waveguide in the bound state and the resonant state to realize the continuous change and reversal of the Fano spectrum, and combine this with the deep learning method to realize the design of a high-precision displacement sensor.

[0068] The present invention analyzes the influence of grating translation misalignment on the spectral Fano line shape through rigorous coupled wave analysis (RCWA) and S-parameter simulation of the Comsol port model, and reveals its high sensitivity and high-precision sensing capability. The specific process is as follows:

[0069] The S-parameter simulation method of RCWA and Comsol port model is used to calculate the reflection spectrum under normal incidence conditions (d = 0). The spectrum results calculated by the two methods are consistent, which verifies the effectiveness of the algorithm.

[0070] The reflectance spectrum shows an asymmetric Fano line shape, indicating the existence of interference effects between narrowband channels (discrete spectrum) and broadband channels (continuous spectrum).

[0071] The effect of grating translation misalignment on the position of the Fano line is analyzed, and it is found that the misalignment changes the phase difference between the GMR mode and the cavity resonance without changing the center position of the Fano line.

[0072] Furthermore, the transmittance and reflectance spectra were calculated at different d values, and it was found that the shape of the Fano line changes continuously with the increase of d. When d is greater than 0.1a, the direction of the Fano line reverses, which is reflected in Figure 3 This directional feature plays an important role in nano-displacement sensing in the Fano spectral line changes.

[0073] The intersections of the spectral lines at different d values ​​show different topological properties. The bound states and resonant states in the quasi-continuum spectrum exhibit different optical properties at these intersections.

[0074] The bound states in the quasi-continuum spectrum have phase singularities and topological charges, which give them high sensitivity in nano-displacement sensing.

[0075] The frequency f = 15.1427 THz is fixed, and the spectra of transmittance T and reflectance R are calculated by changing d. In the range of d from 0.045a to 0.35a, the values ​​of T and R remain almost unchanged, which is consistent with the intersection point. Figure 3 The Fano spectral lines in the spectrum change.

[0076] Furthermore, when d approaches 0 or 0.5a, the spectral lines of T or R change sharply, indicating that the resonant state cannot be converted into the bound state in the quasi-continuous spectrum through continuous spectral line changes.

[0077] Furthermore, by utilizing the changes in the shape and direction of the Fano lines, highly sensitive detection of nanometer-scale displacements can be achieved. The translational misalignment of the grating directly affects the shape and direction of the Fano lines, and by monitoring these changes, nanometer-scale displacements can be accurately measured.

[0078] Since the BIC state has high quality factor and topological protection characteristics, it is reflected in Figure 2 The Q value in the Q value curve is very high, and the sensor has high accuracy and stability during the detection process. Even in the presence of environmental noise and interference, the sensor can still maintain efficient detection performance.

[0079] The effectiveness of the topological transformation of bound states and resonant states in the quasi-continuous spectrum based on the composite grating structure in nano-displacement sensing was verified. This design not only has high sensitivity and high precision detection capabilities, but also has good tunability and potential for multifunctional integration, providing a new solution for the field of nano-optical sensing.

[0080] Application of the present invention:

[0081] In the field of semiconductor manufacturing, precise displacement control is crucial to the photolithography process and directly affects the performance and yield of the chip. The invented sensor can monitor and adjust the displacement in real time, ensuring that each process step is carried out in the best condition, significantly improving production efficiency and product quality. In ultra-precision engineering, such as the manufacture and alignment of optical lenses, errors at the nanometer level may cause the performance of the finished product to fail to meet the standards. Using the invented high-precision sensor, displacement errors can be detected and corrected in real time during the assembly process to ensure that the final product meets the design requirements.

[0082] In micro-electromechanical systems (MEMS), displacement control of sensors and actuators is crucial. The invention's high-resolution sensors can detect tiny displacement changes, thereby achieving precise control of MEMS devices and improving their performance and stability. For example, in micro-gyroscopes and accelerometers, the accuracy of displacement detection is directly related to the accuracy and reliability of the measurement results. The invention's sensing technology can significantly improve the performance of these devices and broaden their application range.

[0083] In the field of aerospace, high-precision positioning and control of equipment and instruments are crucial to mission success. Especially in the space environment, tiny displacements may cause equipment failure or mission failure. The invented sensor can provide reliable displacement detection in extreme environments, ensuring the normal operation of spacecraft and the smooth completion of missions. For example, in the alignment and calibration of space telescopes, nanometer-level displacement sensing can ensure precise adjustment of lenses to obtain clear images of the universe.

[0084] The invention also has important applications in biomedical engineering. For example, in microsurgery and nano-drug delivery, it can provide high-precision displacement detection to ensure the safety and effectiveness of medical operations. In materials science research, the invented sensor can achieve high-precision displacement detection in material testing and characterization, obtain accurate material property data, and contribute to the development of new materials and the optimization of existing material performance. In the manufacture of optical devices, the invented sensing technology can help achieve precise positioning and adjustment of optical components, and improve the application effect of optical devices in communication, sensing and imaging systems.

[0085] In robotics, the high-precision displacement sensor of this invention is crucial for the positioning and motion control of robot arms, especially in high-precision tasks such as micro-manipulation and nano-manufacturing, providing reliable displacement detection support. In the field of quantum computing and quantum communication, nanoscale displacement sensors can be used for precise positioning and manipulation of quantum bits to ensure the stability and reliability of quantum systems. In terms of earthquake monitoring and disaster prevention and mitigation, the invented sensor can detect tiny displacements of the earth's crust, predict the possibility of earthquakes in advance, and protect human life and property.

[0086] In summary, the topological transformation of Bloch interface states and resonant states based on composite grating structures and their application in nano-displacement sensing not only plays an important role in traditional fields such as semiconductor manufacturing, ultra-precision engineering, micro-electromechanical systems, and aerospace, but also shows broad application prospects in emerging fields such as biomedical engineering, materials science, optical device manufacturing, robotics, quantum computing and quantum communication, earthquake monitoring, and disaster prevention and mitigation. The promotion and application of this invention will bring about technological progress and economic benefits, and promote the development of various industries.

[0087] The above embodiments are only used to illustrate the design ideas and features of the present invention, and their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The protection scope of the present invention is not limited to the above embodiments. Therefore, any equivalent changes or modifications made based on the principles and design ideas disclosed by the present invention are within the protection scope of the present invention.

Claims

1. A nanoscale displacement sensing method based on a composite grating structure, characterized in that: The steps include: Step 1, constructing a composite grating structure, the composite grating structure comprising two layers of grating groups, each layer of the grating group comprising a plurality of gratings (1) and an intermediate waveguide layer (2), the gratings (1) being arranged at equal intervals on one side surface of the intermediate waveguide layer (2); and the other side surfaces of the intermediate waveguide layer (2) being in contact with each other; Step 2: When the intermediate waveguide layers (2) move relative to each other, the gratings (1) on the two intermediate waveguide layers (2) are displaced by translation, and the displacement between the gratings (1) is recorded as d; Step 3: Based on the composite grating structure, one side of the composite grating structure is used as an input end and a normally incident plane wave is used as a source excitation, and the other side is used as an output end, and a reflection spectrum or / and a transmission spectrum corresponding to different displacements d are obtained at the output end, and a Fano spectrum line is obtained based on the reflection spectrum or / and the transmission spectrum; combined with the translational misalignment of the two gratings (1), the inversion and continuous deformation of the Fano spectrum line are realized in the composite grating structure; Step 4: Use the Fano spectrum obtained in step 3 as the input of deep learning, and use deep learning to output the displacement corresponding to the current Fano spectrum, thereby realizing high-precision displacement sensing; the deep learning adopts an artificial neural network model, and the processing method of the input data of the artificial neural network model is as follows: Displace d From 0 to 2500 nm, it is divided into 2501 values ​​with an interval of 1 nm; For each value, the reflection spectrum was calculated at 501 frequency points in the frequency range of 15.4 THz to 15.45 THz to obtain a spectral line; Each spectral line represents a data sample as a 1×501 data array; The data was normalized and the input sequence of 2501 samples was randomly set.

2. The nanoscale displacement sensing method based on a composite grating structure according to claim 1, characterized in that: The artificial neural network model includes an input layer, an output layer and three hidden layers.

3. The nanoscale displacement sensing method based on a composite grating structure according to claim 2, characterized in that: The nodes of the hidden layer and the output layer use the ReLU function as the activation function.

4. The nanoscale displacement sensing method based on a composite grating structure according to claim 2, characterized in that: Each node data of the output layer is converted into a probability distribution output through the Softmax function.

5. The nanoscale displacement sensing method based on a composite grating structure according to claim 2, characterized in that: The two layers of grating groups in the composite grating structure have the same structure.

6. The nanoscale displacement sensing method based on a composite grating structure according to claim 5, characterized in that: The grating (1) is made of glass, and the intermediate waveguide layer (2) is made of silicon.

7. The nanoscale displacement sensing method based on a composite grating structure according to claim 5, characterized in that: The gratings (1) in the grating group are arranged in a periodic manner, and the period a is 8 μm.

8. The nanoscale displacement sensing method based on a composite grating structure according to claim 5, characterized in that: In a grating group, the width between grating (1) and adjacent grating (1) w G The thickness of the grating (1) is 4 μm. h G The thickness of the intermediate waveguide layer (2) is 10 μm. h W is 2μm.

9. An application of a nanoscale displacement sensing method based on a composite grating structure, characterized in that: The nanoscale displacement sensing method based on a composite grating structure as described in claim 1 is used for displacement sensing in semiconductor manufacturing, ultra-precision engineering, micro-electromechanical systems or aerospace.

Images