Device and method for rapid non-destructive non-contact measurement of surface strain of silicon material

Abstract

This invention provides a rapid, non-destructive, and non-contact device and method for measuring strain on silicon surfaces, comprising: a femtosecond laser for generating femtosecond laser pulses; a time-domain stretching component disposed in the output optical path of the femtosecond laser; an optical signal amplification component disposed in the output optical path of the time-domain stretching component; a beam splitter located in the output optical path of the optical signal amplification component, for splitting the femtosecond laser pulses into a test beam and a reference beam with a specific power ratio; a spatial dispersion component disposed in the output optical path of the test beam of the beam splitter; a microscope objective disposed in the output optical path of the spatial dispersion component; a high-speed photodetector located in the output optical path of the reference beam of the beam splitter; a high-speed oscilloscope electrically connected to the high-speed photodetector; and a data processing device electrically connected to the high-speed oscilloscope. This invention addresses the problems of strong dependence on speckle and low detection frame rate in existing technologies.

Description

Technical Field

[0001] This invention belongs to the technical field of optical mechanics and deformation measurement, specifically relating to a measurement method for a rapid, non-destructive, non-contact measurement device for strain on silicon material surfaces. Background Technology

[0002] Currently, chip dimensions have entered the nanometer era. Due to the reduction in chip package size and the increase in overall chip power consumption, heat generated by the chip easily accumulates, causing the chip temperature to rise. Furthermore, chip substrates are often made of composite materials, and different components have different coefficients of thermal expansion. Under temperature changes, internal thermal mismatches can occur, resulting in stress and strain, causing warping deformation, and even leading to solder joint failure, chip cracking, and delamination. Therefore, the detection of chip strain is receiving increasing attention.

[0003] Due to the small size of the chip, which falls under microscopic measurement, traditional contact strain measurement methods based on strain gauges or extensometers are insufficient to meet the needs of this field. Currently, Digital Image Correlation (DIC) is a widely used non-contact measurement method in this field. This method first creates speckle patterns on the sample surface using methods such as spraying, immersion, or etching. Then, it uses an optical microscope (OM) or scanning electron microscope (SEM) to capture speckle images before and after deformation. By comparing the changes in the position of the speckle patterns before and after deformation, the strain field distribution of the sample is analyzed.

[0004] However, traditional strain measurement methods have two major limitations: First, traditional DIC technology is based on speckle patterns created on the surface of the sample for analysis. By analyzing the speckle patterns before and after deformation, algorithms are used to calculate the strain distribution and changes on the sample surface. The measurement quality largely depends on the quality of the speckle pattern creation. Second, to achieve high-speed measurements with high frame rates, traditional DIC technology requires expensive high-speed cameras. The frame rate is limited by the high-speed camera, with the highest frame rate being in the tens of thousands of frames per second, and it is also constrained by the field of view size and resolution. Summary of the Invention

[0005] One objective of this invention is to address the shortcomings of existing technologies by providing a rapid, non-destructive, and non-contact measurement device for strain on silicon surfaces, thereby solving the problems of strong dependence on speckle and low detection frame rate in existing technologies.

[0006] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0007] A rapid, non-destructive, non-contact device for measuring strain on silicon surfaces includes:

[0008] Femtosecond lasers are used to generate femtosecond laser pulses that are sufficient to reflect and image the surface of a sample without damaging the sample.

[0009] A time-domain stretching component, which is connected to the near-infrared / near-ultraviolet femtosecond laser, is used to stretch the femtosecond laser pulse in the time domain.

[0010] An optical signal amplification component, which is connected to a time-domain stretching component, is used to compensate the power of the femtosecond laser pulse stretched by the time-domain stretching component.

[0011] The beam splitter, located on the output optical path of the optical signal amplification component, is used to split the femtosecond laser pulse into a test beam and a reference beam with a specific power ratio. The test beam is reflected by the sample under test and returns to the beam splitter along the optical path, interfering with the reference beam to form an interference pulse with stable power.

[0012] A spatial dispersion component is disposed in the output optical path of the beam splitter test beam. It is used to spatially disperse the output pulse into a spatial pulse, and after the spatial pulse is reflected by the sample under test, it is irradiated onto the spatial dispersion component in the opposite direction of the original optical path, and then re-converged by the spatial dispersion component to obtain a point light.

[0013] A microscope objective, which is placed in the output light path of the spatial dispersion component, is used to focus the spatial pulse onto the sample to be tested;

[0014] A high-speed photodetector, located in the output optical path of the beam splitter reference beam, is used to convert interference pulse optical signals into analog electrical signals;

[0015] A high-speed oscilloscope, electrically connected to the high-speed photodetector, is used to collect sampled analog electrical signals;

[0016] A data processing device, electrically connected to the high-speed oscilloscope, is used to process and calculate the strain on the surface of the sample to be tested by collecting and sampling analog electrical signals.

[0017] Furthermore, the femtosecond laser is either a near-infrared femtosecond laser or a near-ultraviolet femtosecond laser.

[0018] Furthermore, it also includes an optical power meter, which is placed between the microscope objective and the sample to be tested before testing, and removed from the optical path during testing.

[0019] Furthermore, it also includes a lens assembly disposed between the spatial dispersion component and the microscope objective, wherein the first lens assembly is used to adjust the size of the pulse spot and the angle at which the pulse is incident on the microscope objective.

[0020] Furthermore, it also includes a first collimator and a second collimator, wherein the first collimator is located between the optical signal amplification component and the beam splitter, and the first collimator is used to incident the pulse amplified by the optical signal amplification component onto the beam splitter at a specific angle and in the form of spatial light; the second collimator is located between the reference beam exit side of the beam splitter and the high-speed photodetector, and the second collimator is used to couple the interference pulse into the optical path.

[0021] Furthermore, the spatial dispersion component includes an acousto-optic deflector and a diffraction grating sequentially disposed on the outgoing optical path of the optical signal amplification component. The acousto-optic deflector is used to change the incident position and direction of the pulse incident on the diffraction grating, and the diffraction grating is used to disperse the femtosecond laser pulse into a one-dimensional spatial pulse.

[0022] Another object of the present invention is to provide a measurement method based on the above-described rapid, non-destructive, non-contact measurement device for silicon surface strain, comprising the following steps:

[0023] Step 1: Clean the sample to be tested and fix it in the outgoing light path of the microscope objective;

[0024] Step 2: Activate the femtosecond laser to generate femtosecond laser pulses sufficient to image the sample under test without damaging it.

[0025] Step 3: The time-domain stretching component stretches the femtosecond pulse in the time domain, then amplifies it through the optical signal amplification component, and then generates a test beam and a reference beam after passing through the beam splitter. The test beam is incident on the spatial dispersion component and then focused onto the sample under test by the microscope objective.

[0026] Step 4: The spatial pulse reflected by the sample under test interferes with the reference beam generated by the spatial dispersion component and the beam splitter to form an interference pulse. After the high-speed photodetector detects the interference pulse, it is converted into an analog signal. The high-speed oscilloscope collects and samples the analog signal and transmits it to the data processing device. The data processing device recovers the test pulse signal distribution image sequence reflected by the sample under test based on the analog electrical signal sampled by the high-speed oscilloscope. Based on the test pulse signal distribution image sequence, it obtains the change of the light intensity distribution image contour boundary before and after the strain of the sample under test, calculates the boundary strain, and calculates the strain of the sample surface through interpolation.

[0027] Furthermore, methods for obtaining the test pulse signal distribution image sequence include:

[0028] The interference pulse signal is analyzed and reconstructed using a data processing device to obtain the test pulse distribution reflected from the sample surface. A two-dimensional light intensity distribution image sequence reflecting from the sample surface is then constructed, with a frame rate equal to the repetition frequency of the near-infrared / near-ultraviolet femtosecond laser. The sequence ranges from the sampled light pulse signal I(x) to the light intensity signal I... s The formula for transforming (x) is as follows:

[0029]

[0030] Among them, I R (x) represents the reference pulse intensity, T represents the high-frequency component of the sampled optical pulse signal I(x) after filtering and Hilbert transform, and re(T) and im(T) represent the real and imaginary parts of T, respectively.

[0031] Furthermore, the method by which the data processing device calculates the surface strain of the sample under test based on the sequence of test pulse signal distribution images includes:

[0032] The nonlocal mean algorithm is used to filter the distribution image sequence of the test pulse signal. Then, the binarized image is extracted from the filtered light intensity distribution image sequence. The eigenvalue and eigenvector method is used to track the distribution of each contour region of the sample before and after strain. The position coordinate matrix A of each region of the sample boundary is represented as: A = QΛQ, where Q is the matrix of eigenvectors and Λ is the matrix of eigenvalues.

[0033] The deformation of each region before and after strain is expressed as follows:

[0034]

[0035] In the formula: e i It is a unit vector; and E represents the transformation matrix of each contour region before and after strain; j These are the characteristic values ​​of the region;

[0036] Then the deformation tensor Represented as:

[0037]

[0038] Strain in each region It is calculated by the following formula:

[0039]

[0040] Furthermore, the method for filtering the test pulse signal distribution image sequence using a nonlocal means algorithm includes:

[0041]

[0042] Where NL[v](i) is the image of pixel i after filtering, v(j) is the original image without filtering; I is the neighborhood of pixel i; w(i,j) represents the similarity between pixels i and j, and its value is determined by the squared Euclidean distance between the search windows:

[0043]

[0044] in,

[0045] Where h is the smoothing parameter; Z is the decay function for the weighted Euclidean distance between pixels i and j, where a is the standard deviation of the Gaussian kernel; i This is the normalization coefficient.

[0046] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention uses near-infrared / near-ultraviolet light to perform ultrafast laser imaging on the sample under test. The imaging frame rate is comparable to the pulse repetition frequency of the laser, which can reach about MHz, thereby enabling rapid measurement of the strain on the silicon surface. In addition, due to the high reflectivity of near-infrared / near-ultraviolet light to silicon, the image sequence generated by the method of the present invention contains the structural information of the surface of the chip or device under test. After the image sequence is processed by the data processing device, it can accurately reflect the strain change on the surface of the chip or device under test, thereby obtaining more accurate strain information on the silicon surface.

[0047] Compared to traditional non-contact DIC technology, this invention overcomes the problem of needing to prepare speckle patterns in DIC technology. It uses images derived from the structural characteristics of the sample chip itself for calculation and analysis, eliminating the need to create additional speckle patterns on the sample and avoiding damage to the sample surface. By controlling the laser power, the generated laser of a specific wavelength can be amplified to have a specific power, so that the laser does not damage the sample while passing through it and forming the image required for the test, thereby achieving non-contact non-destructive testing.

[0048] Furthermore, this invention does not require expensive high-speed cameras and is not constrained by frame rate, field of view size, or resolution. It only requires commonly used near-infrared / near-ultraviolet femtosecond lasers to quickly measure strain changes inside the chip or device under test, which greatly reduces costs and improves measurement efficiency. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of a rapid, non-destructive, non-contact measurement device for silicon surface strain provided in Embodiment 1 of the present invention. Detailed Implementation

[0050] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0051] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other.

[0052] The present invention will be further described below with reference to specific embodiments, but these are not intended to limit the scope of the invention.

[0053] Example 1:

[0054] Example 1 provides an ultrafast, non-destructive, non-contact measurement device for strain on silicon surfaces, such as... Figure 1 The system includes: a femtosecond laser 101, a time-domain stretching component, an optical signal amplification component, a beam splitter, a spatial dispersion component, a microscope object, a high-speed photodetector, a high-speed oscilloscope, and a data processing device. In this embodiment, the femtosecond laser 101 is either a near-infrared femtosecond laser or a near-ultraviolet femtosecond laser, both of which can be reflected on the silicon surface. The time-domain stretching component is a single-mode fiber 102, which is disposed in the output optical path of the near-infrared / near-ultraviolet femtosecond laser 101; the optical signal amplification component is an erbium-doped fiber amplifier 103, which is disposed in the output optical path of the single-mode fiber 102. The beam splitter 105 is located in the output optical path of the optical signal amplification component and is used to split the femtosecond laser pulse into a test beam (first direction) and a reference beam (second direction) with a specific power ratio; the test beam is reflected by the sample under test and returns to the beam splitter 105 along the optical path, interfering with the reference beam to form an interference pulse with stable power. In order to direct the pulse amplified by the erbium-doped fiber amplifier 103 onto the beam splitter 105 at a specific angle in the form of spatial light, a first collimator 104 is provided in the optical path between the erbium-doped fiber amplifier 103 and the beam splitter.

[0055] The spatial dispersion component includes an acousto-optic deflector 106 and a diffraction grating 107 sequentially disposed on the output optical path of the beam splitter 105. The acousto-optic deflector 106 is used to change the incident position and direction of the near-infrared / near-ultraviolet femtosecond laser pulse incident on the diffraction grating 107, and the diffraction grating 107 is used to disperse the near-infrared / near-ultraviolet femtosecond laser pulse into a one-dimensional spatial pulse. In this embodiment, a microscope objective 110 is disposed on the output optical path of the acousto-optic deflector 106, and is used to focus the spatial pulse onto the chip or device under test. To ensure that the pulses generated by the near-infrared / near-ultraviolet femtosecond laser 101 have a first power and a first wavelength after being amplified by the erbium-doped fiber amplifier 103, the first power is sufficient to reflect and image the sample surface without damaging the sample, and the first wavelength is sufficient to form an image of sufficient quality after penetrating the silicon and other polymer components of the sample, an optical power meter 111 is placed between the microscope objective 110 and the sample before the test. After the test begins, the optical power meter 111 is moved out of the optical path.

[0056] In addition, a lens assembly is provided, in which a first plano-convex lens 108 and a second plano-convex lens 109 are sequentially arranged in the optical path between the first diffraction grating 107 and the microscope objective 110. The first lens assembly is used to adjust the size of the pulse spot and the angle at which the pulse is incident on the microscope objective.

[0057] A high-speed photodetector 113 is disposed on the outgoing optical path of the reference beam (second direction side) split by the beam splitter, and is used to convert the interference pulse signal into an analog signal. In this embodiment, the high-speed photodetector 113 is selected as a MHz-level or higher high-speed free-space photodetector 113. To facilitate the coupling of the interference pulse into the optical path for detection by the high-speed photodetector 113, a second collimator 112 is disposed between the beam splitter 105 and the high-speed photodetector 113. A high-speed oscilloscope 114 for collecting and sampling analog electrical signals is electrically connected downstream of the high-speed photodetector 113. In this embodiment, the high-speed oscilloscope 114 is selected as a MHz-level or higher high-speed oscilloscope. A data processing device 115 is electrically connected to the high-speed oscilloscope 114 and is used to recover the test pulse signal distribution image sequence reflected by the sample under test based on the sampled analog electrical signal; and to obtain the change of the light intensity distribution image contour boundary before and after strain of the sample under test based on the test pulse signal distribution image sequence, calculate the boundary strain, and calculate the strain of the sample surface by interpolation. In this embodiment, the data processing device 115 is a computer or other device capable of data processing and calculation.

[0058] Example 2:

[0059] Example 2 provides an ultrafast, non-destructive, and non-contact method for measuring strain on silicon surfaces, which is implemented using the device provided in Example 1, and mainly includes the following steps:

[0060] Step 1: Clean and dry the sample to be tested using chemical detergent and deionized water, and fix it in the optical path of microscope objective 110;

[0061] Step 2: Before testing, set the optical power meter 111 in the optical path between the microscope objective 110 and the sample to be tested. Adjust the near-infrared / near-ultraviolet femtosecond laser 101 and the erbium-doped fiber amplifier 103 so that the femtosecond laser pulse generated by the near-infrared / near-ultraviolet femtosecond laser 101, after being amplified by the erbium-doped fiber amplifier 103, has a first power and a first wavelength. The first power is sufficient to penetrate the sample to be tested without damaging it, and the first wavelength is sufficient to form an image of sufficient quality after penetrating the silicon and other polymer components of the sample to be tested. After the test starts, remove the optical power meter 111 from the optical path, keeping the pulse power and wavelength emitted and amplified by the near-infrared / near-ultraviolet femtosecond laser 101 and the erbium-doped fiber amplifier 103 unchanged.

[0062] Step 3: Temporally stretch the near-infrared / near-ultraviolet femtosecond laser pulse through single-mode fiber 102. The temporally stretched pulse is incident on beam splitter 105 at a specific angle and in the form of spatial light through first collimator 104. Beam splitter 105 splits the femtosecond laser pulse into a test beam (first direction) and a reference beam (second direction) with a specific power ratio.

[0063] Step 4: The test beam is incident on the first acousto-optic deflector 106. By controlling the acoustic wave driving frequency, the crystal refraction characteristics inside the acousto-optic deflector are changed, thereby changing the incident position and angle of the femtosecond laser on the first diffraction grating 107. Spatial light incident at different positions and angles on the first diffraction grating 107 is dispersed into one-dimensional spatial pulses. Then, it is focused on different positions on the sample under test through the first plano-convex lens 108, the second plano-convex lens 109, and the microscope objective 110, thereby realizing a large-scale two-dimensional scan of the one-dimensional femtosecond laser on the sample under test.

[0064] Step 5: After the pulse reflected from the surface of the object under test is reflected, it is irradiated onto the microscope objective 110, the second plano-convex lens 109, and the first plano-convex lens 108 in the opposite direction of the original optical path. Then, it is re-formed into a one-dimensional pulse through the first diffraction grating 107 and the first acousto-optic deflector 106 at the opposite angle to the incident pulse. The pulse is then interfered with by the reference pulse split by the beam splitter and coupled to the high-speed photodetector 113 of MHz level and above through the second collimator 112.

[0065] Step 6: The interference pulse is converted into an analog electrical signal by the high-speed photodetector 113, and then sampled by the high-speed oscilloscope 114 and sent to the computer 115. The computer 115 analyzes and reconstructs the analog signal sampled by the high-speed oscilloscope 114 to obtain a sequence of test pulse signal distribution images reflected from the sample under test. The frame rate is equal to the repetition frequency of the near-infrared / near-ultraviolet femtosecond laser, from the sampled light pulse signal I(x) to the light intensity signal I... s The formula for transforming (x) is as follows:

[0066]

[0067] Among them, I R (x) represents the reference pulse intensity, T represents the high-frequency component of the sampled optical pulse signal I(x) after filtering and Hilbert transform, and re(T) and im(T) represent the real and imaginary parts of T, respectively.

[0068] For the test pulse signal distribution image sequence obtained from the above formula, in order to reduce the influence of test noise such as photodetector noise and oscilloscope measurement noise on the measurement results, a nonlocal mean algorithm is used for filtering to clarify the boundaries of each contour region in the image. The calculation formula of this algorithm is as follows:

[0069] NL[v](i)=∑ j∈I v(j)w(i,j) (2)

[0070] Where NL[v](i) is the image of pixel i after filtering, v(j) is the original image without filtering; I is the neighborhood of pixel i; w(i,j) represents the similarity between pixels i and j, and its value is determined by the squared Euclidean distance between the search windows:

[0071]

[0072] in,

[0073]

[0074] In the formula, h is a smoothing parameter that controls the degree of attenuation of the Gaussian function. The specific value is determined based on the image noise level. Z is the decay function for the weighted Euclidean distance between pixels i and j, where a is the standard deviation of the Gaussian kernel; i This is the normalization coefficient.

[0075] After the above filtering, a binarized image is extracted. The eigenvalue and eigenvector methods are used to trace the contour region distribution of the test object before and after strain. The position coordinate matrix A of each region relative to the sample boundary can be expressed as A = QΛQ, where Q is the matrix of eigenvectors and Λ is the matrix of eigenvalues. The deformation of each region before and after strain can be expressed as:

[0076]

[0077] In the formula: e i It is a unit vector; and E represents the transformation matrix of the region before and after strain, respectively; j Let be the eigenvalues ​​of the region. Then the deformation tensor... It can be represented as:

[0078]

[0079] Strain in each region It can be calculated using the following formula:

[0080]

[0081] Obtain the strain in each region Then, the strain of each region is used to form a node matrix, which is then imported into MATLAB and the two-dimensional spline interpolation toolbox is called. The strain distribution at each point on the surface of the measured sample can be obtained through interpolation calculation.

[0082] Example 2 utilizes the femtosecond-level characteristics of near-infrared / near-ultraviolet femtosecond laser pulses and their high reflectivity on silicon surfaces. A diffraction grating and acousto-optic deflector are used to disperse the spatial pulses, forming a two-dimensional spatial pulse on the surface of the object under test. This allows for ultrafast acquisition of the light intensity distribution after reflection from the object's surface. Combined with the aforementioned algorithm, the strain distribution on the object's surface can be calculated. The entire measurement process is non-contact, enabling ultrafast, non-destructive, non-contact strain measurement of the observed object, with a sampling frequency as high as 10-1. 6 Hz and above can effectively observe the transient strain distribution on the surface of electronic chips or devices.

[0083] The above are merely preferred embodiments of the present invention and are not intended to limit the implementation methods and protection scope of the present invention. Those skilled in the art should recognize that any equivalent substitutions and obvious changes made based on the content of this specification should be included within the protection scope of the present invention.

Claims

1. A rapid, non-destructive, non-contact device for measuring strain on silicon material surfaces, characterized in that, include: Femtosecond lasers are used to generate femtosecond laser pulses that are sufficient to reflect and image the surface of a sample without damaging the sample. A time-domain stretching component is disposed in the output optical path of a near-infrared / near-ultraviolet femtosecond laser and is used to stretch the femtosecond laser pulse in the time domain. An optical signal amplification component is connected to the output optical path of the time-domain stretching component and is used to perform power compensation on the femtosecond laser pulse stretched by the time-domain stretching component. The beam splitter, located on the output optical path of the optical signal amplification component, is used to split the femtosecond laser pulse into a test beam and a reference beam with a specific power ratio. The test beam is reflected by the sample under test and returns to the beam splitter along the optical path, interfering with the reference beam to form an interference pulse with stable power. A spatial dispersion component is disposed in the output optical path of the beam splitter test beam. It is used to spatially disperse the output pulse into a spatial pulse, and after the spatial pulse is reflected by the sample under test, it is irradiated onto the spatial dispersion component in the opposite direction of the original optical path, and then re-converged by the spatial dispersion component to obtain a point light. A microscope objective, which is placed in the output light path of the spatial dispersion component, is used to focus the spatial pulse onto the sample to be tested; A high-speed photodetector, located in the output optical path of the beam splitter reference beam, is used to convert interference pulse optical signals into analog electrical signals; A high-speed oscilloscope, electrically connected to the high-speed photodetector, is used to collect sampled analog electrical signals; A data processing device, electrically connected to the high-speed oscilloscope, is used to process and calculate the strain on the surface of the sample to be tested by collecting and sampling analog electrical signals.

2. The rapid, non-destructive, non-contact measurement device for silicon surface strain according to claim 1, characterized in that, The femtosecond laser is either a near-infrared femtosecond laser or a near-ultraviolet femtosecond laser.

3. The rapid, non-destructive, non-contact measurement device for silicon surface strain according to claim 1, characterized in that, It also includes an optical power meter, which is placed between the microscope objective and the sample to be tested before testing, and removed from the optical path during testing.

4. The rapid, non-destructive, non-contact measurement device for silicon surface strain according to claim 1, characterized in that, It also includes a lens assembly disposed between the spatial dispersion component and the microscope objective, wherein the first lens assembly is used to adjust the size of the pulse spot and the angle at which the pulse is incident on the microscope objective.

5. The rapid, non-destructive, non-contact measurement device for silicon surface strain according to claim 1, characterized in that, It also includes a first collimator and a second collimator, wherein the first collimator is located between the optical signal amplification component and the beam splitter, and the first collimator is used to incident the pulse amplified by the optical signal amplification component onto the beam splitter at a specific angle and in the form of spatial light; the second collimator is located between the reference beam exit side of the beam splitter and the high-speed photodetector, and the second collimator is used to couple the interference pulse into the optical path.

6. The rapid, non-destructive, non-contact measurement device for silicon surface strain according to claim 1, characterized in that, The spatial dispersion component includes an acousto-optic deflector and a diffraction grating sequentially arranged in the output optical path of the optical signal amplification component. The acousto-optic deflector is used to change the incident position and direction of the pulse incident on the diffraction grating, and the diffraction grating is used to disperse the femtosecond laser pulse into a one-dimensional spatial pulse.

7. A method for measuring the surface strain of silicon material using a rapid, non-destructive, non-contact measuring device according to any one of claims 1-6, characterized in that, Includes the following steps: Step 1: Clean the sample to be tested and fix it in the outgoing light path of the microscope objective; Step 2: Start the femtosecond laser to generate a femtosecond laser pulse sufficient to reflect and image the surface of the sample under test without damaging the sample. Step 3: The time-domain stretching component stretches the femtosecond pulse in the time domain, then amplifies it through the optical signal amplification component, and then generates a test beam and a reference beam after passing through the beam splitter. The test beam is incident on the spatial dispersion component and then focused onto the sample under test by the microscope objective. Step 4: The spatial pulse reflected by the sample under test interferes with the reference beam generated by the spatial dispersion component and the beam splitter to form an interference pulse. After the high-speed photodetector detects the interference pulse, it is converted into an analog signal. The high-speed oscilloscope collects and samples the analog signal and transmits it to the data processing device. The data processing device recovers the distribution image sequence of the test pulse signal reflected by the sample under test based on the analog electrical signal sampled by the high-speed oscilloscope. The changes in the light intensity distribution image contour boundary of the sample before and after strain are obtained based on the test pulse signal distribution image sequence, the boundary strain is calculated, and the strain on the sample surface is obtained by interpolation.

8. The measurement method of the rapid, non-destructive, non-contact strain measurement device for silicon material surface according to claim 7, characterized in that, Methods for obtaining test pulse signal distribution image sequences include: The interference pulse signal is analyzed and reconstructed using a data processing device to obtain the test pulse distribution reflected from the sample surface. A two-dimensional light intensity distribution image sequence reflecting from the sample surface is then constructed, with a frame rate equal to the repetition frequency of the near-infrared / near-ultraviolet femtosecond laser. The sequence ranges from the sampled light pulse signal I(x) to the light intensity signal I... s The formula for transforming (x) is as follows: Among them, I R (x) represents the reference pulse intensity, T represents the high-frequency component of the sampled optical pulse signal I(x) after filtering and Hilbert transform, and re(T) and im(T) represent the real and imaginary parts of T, respectively.

9. The measurement method of the rapid, non-destructive, non-contact strain measurement device for silicon material surface according to claim 7, characterized in that, The data processing device calculates the surface strain of the sample under test based on the sequence of test pulse signal distribution images, including: The nonlocal mean algorithm is used to filter the distribution image sequence of the test pulse signal. Then, the binarized image is extracted from the filtered light intensity distribution image sequence. The eigenvalue and eigenvector method is used to track the distribution of each contour region of the sample before and after strain. The position coordinate matrix A of each region of the sample boundary is represented as: A = QΛQ, where Q is the matrix of eigenvectors and Λ is the matrix of eigenvalues. The deformation of each region before and after strain is expressed as follows: In the formula: e i It is a unit vector; and E represents the transformation matrix of each contour region before and after strain; j These are the characteristic values ​​of the region; Then the deformation tensor Represented as: Strain in each region It is calculated by the following formula:

10. The measurement method of the rapid, non-destructive, non-contact strain measurement device for silicon material surface according to claim 9, characterized in that, Methods for filtering test pulse signal distribution image sequences using nonlocal means algorithms include: Where NL[v](i) is the image of pixel i after filtering, v(j) is the original image without filtering; I is the neighborhood of pixel i; w(i,j) represents the similarity between pixels i and j, and its value is determined by the squared Euclidean distance between the search windows: in, Where h is the smoothing parameter; Z is the decay function for the weighted Euclidean distance between pixels i and j, where a is the standard deviation of the Gaussian kernel; i This is the normalization coefficient.

Images