CN113847883B - Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure - Google Patents

Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure Download PDF

Info

Publication number
CN113847883B
CN113847883B CN202111056510.7A CN202111056510A CN113847883B CN 113847883 B CN113847883 B CN 113847883B CN 202111056510 A CN202111056510 A CN 202111056510A CN 113847883 B CN113847883 B CN 113847883B
Authority
CN
China
Prior art keywords
image
scanning
sample
top end
pixel point
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202111056510.7A
Other languages
Chinese (zh)
Other versions
CN113847883A (en
Inventor
高志山
赵雨晴
孙一峰
马剑秋
袁群
周俊涛
郭珍艳
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing University of Science and Technology
Original Assignee
Nanjing University of Science and Technology
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing University of Science and Technology filed Critical Nanjing University of Science and Technology
Priority to CN202111056510.7A priority Critical patent/CN113847883B/en
Publication of CN113847883A publication Critical patent/CN113847883A/en
Application granted granted Critical
Publication of CN113847883B publication Critical patent/CN113847883B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B11/00—Measuring arrangements characterised by the use of optical techniques
    • G01B11/24—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures
    • G01B11/2441—Measuring arrangements characterised by the use of optical techniques for measuring contours or curvatures using interferometry

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Length Measuring Devices By Optical Means (AREA)

Abstract

The invention discloses an interferometry method suitable for detecting three-dimensional morphology of a high aspect ratio structure, which is used for solving the problem that the three-dimensional morphology of a sample with the aspect ratio larger than 10 is difficult to recover from a complex coherent signal in the conventional vertical scanning interferometry. According to the method, light intensity information and envelope information are extracted from coherent signals corresponding to all pixel points, the light intensity information and the envelope information are combined to obtain a synthetic image, binarization segmentation is carried out on the image according to the difference of top-end and bottom-end pixel points of a high-aspect-ratio structure in the synthetic image, the top end and the bottom end are distinguished, and transverse resolution is achieved; aiming at the problem that a bottom-end pixel point has a plurality of coherent signal envelopes, the scanning position corresponding to the maximum position of the coherent signal envelope in a selected window is positioned for the pixel point at the bottom end, so that the accurate height value of the structure bottom end is obtained, the automatic processing of a complex coherent signal in vertical scanning interferometry is realized, and the three-dimensional appearance of a sample with the depth-to-width ratio larger than 10.

Description

Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure
Technical Field
The invention relates to the technical field of precision optical measurement engineering, in particular to an interference measurement method suitable for detecting a three-dimensional shape of a high-depth-to-width ratio structure.
Background
With the development of micro-electromechanical system MEMS, the requirement for measuring the microstructure is higher and higher, and in the MEMS processing technology based on silicon, the aspect ratio is one of the main indexes and directly influences the performance of MEMS devices; the trench width of the MEMS high-aspect-ratio microstructure is 3-10 μm, the depth is 10-300 μm, and the high-aspect-ratio microstructure is generally 10-100. The existing geometric measurement method for the microstructure device with the high aspect ratio at home and abroad is roughly divided into contact measurement and non-contact measurement; for contact measurements, the most common instruments include scanning electron microscopes SEM and atomic force microscopes, which require the device to be cut open from the side to measure the trench bottom; the non-contact measurement mainly refers to an optical measurement technology, and compared with other measurement technologies, the non-destructive measurement can be realized.
The vertical scanning interferometry is a non-contact surface micro-topography measuring technology, and can obtain visual three-dimensional topography of a sample on the basis of not damaging the sample to be measured. And the broadband light is adopted as a light source, so that the zero optical path difference position of each point in the field of view to be measured can be accurately judged, the problem of phase ambiguity is avoided, and high-precision measurement is realized. The traditional vertical scanning interferometry technology relies on scanning devices such as piezoelectric ceramics and the like to scan a sample to be measured according to preset scanning steps, images in a field of view to be measured under each step are recorded in the scanning process, coherent signals of all pixel points are extracted from the images to carry out signal envelope extremum positioning, and finally the three-dimensional shape of the sample is obtained.
However, for high aspect ratio structures, the coherent signal corresponding to the pixel point at the top of the structure mainly contains a signal envelope, which is a conventional signal form. However, the abnormal coherent signal corresponding to the pixel point located at the bottom end of the structure contains a plurality of signal envelopes, and the signal envelope corresponding to the real bottom end of the structure cannot be determined, so that the three-dimensional morphology of the sample cannot be calculated by locating the position of the envelope maximum. Therefore, in the measurement process, the top end and the bottom end of the structure must be distinguished firstly, the position of the maximum value of the envelope of the coherent signal is directly positioned for the pixel point positioned at the top end, an effective window is selected for the pixel point positioned at the bottom end, the interior of the window only contains the envelope of the coherent signal corresponding to the bottom end of the structure, the position of the maximum value of the envelope is positioned, and finally, the automatic detection of the three-dimensional shape of the high-aspect ratio structure is realized.
Chinese patent (CN 200710053292.5) discloses a method and a device for measuring a micro-nano deep groove structure, wherein the method comprises the steps of projecting an infrared beam to the surface of a silicon wafer containing the deep groove structure, and analyzing interference light formed by reflection of each interface of the deep groove structure to obtain a measured reflection spectrum; and constructing a theoretical reflection spectrum of the equivalent multilayer thin film stack optical model of the deep groove structure by adopting an equivalent medium theory, fitting the measured reflection spectrum through the theoretical reflection spectrum, and further extracting the set characteristic parameters such as the depth, the width and the like of the groove. The method disclosed by the patent needs to model the groove structure of the sample to be measured in advance and calculate to obtain a theoretical reflection spectrum, the accuracy of the measurement result is influenced by a theoretical model established in advance, the modeling difficulty of the sample to be measured with a complex structure or an unknown structure is large, and the accuracy of the measurement result is difficult to ensure.
Chinese patent "method for detecting defects in depth features" (cn201980002469. X) which uses a broadband deep ultraviolet beam to illuminate the bottom of a structure such as a hole, slit or trench, obtains a bright field illumination image containing structural features by receiving reflected or scattered light, and detects defects by contrast between an abnormal image caused by defects and a normal image reflected and received by a standard sample. The method described in this patent is, however, only suitable for under-etching defects and does not specifically reveal the abnormal image features caused by the defects. In addition, the detection result can only qualitatively identify the defect, and accurate shape information cannot be obtained to evaluate the size of the defect.
Disclosure of Invention
The invention aims to provide an interferometry method suitable for detecting three-dimensional topography of a high aspect ratio structure, which is used for solving the problem that the three-dimensional topography of a sample with the aspect ratio larger than 10 is difficult to recover from a complex coherent signal in the conventional vertical scanning interferometry.
The technical solution for realizing the purpose of the invention is as follows: an interferometry method suitable for detecting three-dimensional shape of a high aspect ratio structure comprises the following steps:
step 1, focusing an interference objective of an interferometer with a Linnik structure to the bottom end of a sample, and recording the position b of a PZT (piezoelectric transducer) of a test arm at the moment 1 Performing time sequence vertical coarse scanning on the sample according to the scanning direction from bottom to top, simultaneously controlling the CCD to collect images, calculating the images by using a gray variance evaluation function, and further automatically positioning the top end of the sample and recording the PZT position a 1 In the position a 1 Time sequence vertical middle scanning is carried out for centering, the top end of a sample is precisely positioned, and a corresponding PZT position a is recorded 2 In the position a 2 Defining a scan range of + -wide for the center 1 And performing time sequence vertical fine scanning with step length, controlling the CCD to synchronously acquire a sample top interference image C, and turning to the step 2.
Step 2, controlling PZT to return to position b 1 At position b 1 Time-series vertical middle scanning is carried out for centering, the bottom end of a sample is precisely positioned, and the corresponding PZT position b is recorded 2 In the position b 2 Defining a scan range of + -wide for the center 1 And (4) performing time sequence vertical fine scanning with step length, controlling the CCD to synchronously acquire the interference image D at the bottom end of the sample, and turning to the step 3 and the step 4.
Step 3, extracting coherent signals corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, calculating the maximum value of each pixel point coherent signal, and splicing all the obtained maximum values into a light intensity information image I max And (5) turning to the step.
Step 4, extracting a coherent signal corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, obtaining coherent signal envelopes in a contrast calculation mode, positioning the coherent signal envelope gravity center position by adopting a gravity center method and obtaining corresponding scanning positions, and splicing all scanning positions pixel by pixel to obtain an envelope information image E cen And (5) turning to the step.
Step 5, the light intensity information image I max With the envelope information image E cen And combining to obtain a composite image, increasing the difference of pixel point signals at the top end and the bottom end of the high-aspect-ratio structure, converting the composite image into a gray histogram, calculating an image segmentation threshold value t through an OTSU algorithm, performing binarization segmentation on the composite image to distinguish the top end and the bottom end of the high-aspect-ratio structure, realizing transverse resolution, and turning to the step 6.
Step 6, positioning a scanning position corresponding to the maximum value of the envelope of the coherent signal for each pixel point positioned at the top end of the high-aspect-ratio structure; for the pixel point at the bottom of the high aspect ratio structure, the position b 2 Defining windows ± wide for the center 2 Positioning a scanning position corresponding to the maximum value of the envelope of the coherent signal in the window; position a recorded from PZT 2 、b 2 Calculating the distance L of the automatic focus-fixing jump between the top end and the bottom end in the time sequence vertical fine scanning process, and adding the distance L into the scanning position result of each pixel point at the top end of the high-aspect-ratio structure; according to the above methodAnd calculating and splicing all scanning positions pixel by pixel to obtain the three-dimensional shape of the high-aspect-ratio structure.
Compared with the prior art, the invention has the remarkable advantages that:
(1) The effective scanning interval is positioned by adopting a mode of automatically fixing focus from bottom to top, so that a scanning mode of full-height coverage in the existing method is avoided, the data acquisition speed is greatly improved, the data redundancy is reduced, and the high-efficiency and automatic three-dimensional shape detection of the high-aspect-ratio structure is realized.
(2) Light intensity information and envelope information are extracted from the coherent signals, the light intensity information and the envelope information are fused to obtain a synthetic image, the difference of pixels at the top end and the bottom end of the high-aspect-ratio structure is further reflected, the synthetic image is processed by adopting an OTSU threshold segmentation algorithm, the top end and the bottom end of the structure are effectively distinguished, and transverse resolution is realized.
(3) The influence of abnormal coherent signals of the structure bottom pixel points on the detection result is eliminated by adopting different signal processing modes for the pixels at the top end and the bottom end of the structure with the high depth-to-width ratio, so that the automatic processing of the complex coherent signals in the field of view in the vertical scanning interferometry is realized, and the accuracy of morphology detection is ensured.
Drawings
FIG. 1 is a flow chart of an interferometric method suitable for detecting three-dimensional features of high aspect ratio structures according to the present invention.
FIG. 2 is a schematic diagram of a corresponding optical path structure of an interferometric method suitable for detecting a three-dimensional shape of a high aspect ratio structure.
Fig. 3 is a corresponding interference image when the sample tip is finely positioned.
Fig. 4 is a light intensity information image obtained by splicing the maximum coherent signals of each pixel point.
Fig. 5 is an envelope information image obtained by the center-of-gravity method.
Fig. 6 is a binarized image obtained after the lateral resolution.
FIG. 7 is a diagram of anomalous coherence signals at the bottom pixels of a high aspect ratio structure.
FIG. 8 is a diagram of conventional coherent signals at the top pixel of a high aspect ratio structure.
FIG. 9 shows the two-dimensional topography results of the recovered sample.
FIG. 10 is a two-dimensional topography curve of a high aspect ratio silicon-based MEMS sample obtained by the interferometry method applicable to three-dimensional topography detection of high aspect ratio structures.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below.
With reference to fig. 1, an interferometric method suitable for detecting a three-dimensional shape of a high aspect ratio structure includes the following steps:
step 1, focusing an interference objective of an interferometer with a Linnik structure to the bottom end of a sample, and recording the position b of a PZT (piezoelectric transducer) of a test arm at the moment 1 . The time sequence vertical coarse scanning is carried out according to the scanning direction from bottom to top, the CCD is controlled to collect images, the gray variance evaluation function calculates the images, the top end of a sample is automatically positioned, the time sequence vertical coarse scanning step length must ensure that a series of interference images collected by the CCD must have a frame falling in the range where interference occurs at the top end, and the gray variance evaluation function calculation result corresponding to the frame becomes an extreme value. The calculation formula of the gray variance evaluation function of the image is as follows:
Figure BDA0003254793380000041
Figure BDA0003254793380000042
in the formula, var i Is the gray variance value of a certain frame image, W is the row number of the image gray matrix, H is the column number of the image gray matrix, and I (x, y) is the gray value at the position of the image gray matrix (x, y); u represents an intermediate function; image frame number i =1,2,3, … ….
The range of the interference generated at the top end is twice of the light source coherence length Δ L, which is expressed as follows:
Figure BDA0003254793380000043
wherein, delta lambda is the spectral width of the spectral distribution of the light source, and lambda 0 The center wavelength of the light source.
Positioning frame gray variance value var in coarse scanning process p Greater than the gray variance value var of the front and back two frames of images p-1 、var p+1 Recording the scanning position a of the corresponding PZT 1 。
Then at position a 1 Performing time sequence vertical middle scanning for precisely positioning the top end of a sample, simultaneously controlling a CCD to collect an image, wherein the middle scanning range is equal to 2 times of coarse scanning step length, the middle scanning step length needs to be scanned for 14 times in the middle scanning range, calculating the image by adopting a gray variance evaluation function, and recording the PZT scanning position corresponding to a frame of image with the maximum gray variance value var as a 2 . In the position a 2 Defining a scan range of + -wide for the center 1 And performing time sequence vertical fine scanning with step length to control the CCD to synchronously acquire a sample top end interference image C, wherein wide 1 ＝4×ΔL，step＝λ 0 /8。
Step 2, controlling PZT to return to position b 1 At position b 1 Time-series vertical middle scanning is carried out for centering, the bottom end of a sample is precisely positioned, and the corresponding PZT position b is recorded 2 In the position b 2 Defining a scan range of + -wide for the center 1 And performing time sequence vertical fine scanning with step length, and controlling the CCD to synchronously acquire the interference image D at the bottom end of the sample.
Step 3, extracting coherent signals corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, calculating the maximum value of each pixel point coherent signal, and splicing all the obtained maximum values into a light intensity information image I max Wherein:
I max ＝max (I(x,y,z)) (4)
in the formula, I (x, y, z) represents a coherent signal of any pixel point extracted from the sample top end interference image C and the sample bottom end interference image D.
Step 4, extracting a coherent signal corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, and obtaining a coherent signal envelope by adopting a contrast calculation mode, wherein the calculation mode of the contrast M of a certain pixel point in the sample top end interference image C and the sample bottom end interference image D is as follows:
Figure BDA0003254793380000051
in the formula, M (I) and I i Respectively calculating the contrast value and the light intensity value of the pixel point in the ith frame of the interference image, and obtaining the contrast of the pixel point, namely the signal envelope, after calculating the contrast values of all the frames of the interference image.
And then, positioning the envelope gravity center position h of the coherent signal by adopting a gravity center method and obtaining a corresponding scanning position, wherein the calculation expression of the gravity center method is as follows:
Figure BDA0003254793380000052
in the formula, N is the total frame number of the acquired interference image.
Obtaining scanning positions according to the gravity center position h, calculating pixel by pixel and splicing all the scanning positions into an envelope information image E cen 。
Step 5, multiplying the light intensity information image I max With the envelope information image E cen And combining to obtain a composite image, and increasing the difference of pixel point signals at the top end and the bottom end of the high-aspect-ratio structure. And converting the synthesized image into a gray histogram, and calculating an image segmentation threshold value t through an OTSU algorithm. The OTSU algorithm divides the image gray histogram into A, B two parts, and determines the optimal gray threshold for image segmentation by searching the maximum variance between the two parts, wherein the inter-class variance
Figure BDA0003254793380000061
The calculation expression of (a) is as follows:
Figure BDA0003254793380000062
in the formula w A 、w B Denotes the probability of each part in the population, u A 、u B Representing the mathematical expectation of each part and E representing the mathematical expectation of the ensemble.
Threshold t is the between-class variance
Figure BDA0003254793380000063
And at the maximum gray level position, regarding the pixel points higher than the threshold value t as the points at the top end of the structure, regarding the pixel points lower than the threshold value t as the points at the bottom end of the structure, binarizing the synthesized image, distinguishing the top end and the bottom end of the structure with the high depth-to-width ratio, finishing the transverse resolution of the sample structure, and solving the problem that the step edge is difficult to accurately position when the appearance of the structure with the high depth-to-width ratio with the transverse characteristic close to the diffraction limit is restored at present.
Step 6, positioning a scanning position corresponding to the maximum value of the envelope of the coherent signal for each pixel point positioned at the top end of the high-aspect-ratio structure; for the pixel point at the bottom of the high aspect ratio structure, the position b 2 Defining windows ± wide for the center 2 Locating the scanning position corresponding to the maximum value of the envelope of the coherent signal in the window, wherein wide 2 =40 × step; position a recorded from PZT 2 、b 2 And calculating the distance L between the top end and the bottom end in the scanning process due to automatic fixed focus jump, wherein the calculation mode is as follows:
L＝|a 2 -b 2 | (8)
and adding the distance L into the scanning position result of each pixel point at the top end of the high-aspect-ratio structure, calculating pixel by pixel according to the method, and splicing all the scanning positions to obtain the three-dimensional shape of the high-aspect-ratio structure.
Effective coherent signals are extracted by carrying out differential processing on the coherent signals at the top end and the bottom end, the influence that abnormal coherent signals of pixel points at the bottom end of the structure contain a plurality of signal envelopes is eliminated, and the automatic processing of complex coherent signals in a Cheng Zhongshi field measured by a sample with a high depth-to-width ratio is realized.
Example 1
With reference to fig. 1 to 10, in an interferometric method suitable for detecting a three-dimensional shape of a high aspect ratio structure, an optical system is mounted with an interferometric objective system with a magnification of 20 × during an experiment process to detect a sample having a comb-shaped structure with a high aspect ratio, wherein the depth of the sample is 80 μm, the width of a trench is 3 μm, and the aspect ratio is about 27. The central wavelength lambda of the light source at the position of the CCD through the micro interference system 0 576nm, a spectral width Δ λ of 260nm, a coherence length Δ L of 1.276 μm, a time-series vertical fine scanning step size of 72nm, a time-series vertical middle scanning step size of 576nm, a time-series vertical coarse scanning step size of 4.032 μm, a time-series vertical middle scanning range of + -4.032 μm, and a time-series vertical fine scanning range of + -wide 1 Is + -5.104 μm, window range of coherent signal envelope + -wide 2 Is + -2.88 μm. The interferometric method suitable for detecting the three-dimensional shape of the high aspect ratio structure, disclosed by the patent, comprises the following implementation steps:
referring to fig. 2, a light source 1 is connected to a kohler illumination system 2, the positions of the kohler illumination system 2 and a spectroscope 3 are adjusted to enable light emitted by the kohler illumination system 2 to be received and reflected by the spectroscope 3, the position of a test light path system 6 is adjusted to enable light reflected by the spectroscope 3 to pass through an objective lens 4 and then reflected by an object 5 to be tested to form test light, the positions of a reference light path system 9 and the spectroscope 3 are adjusted to enable light transmitted by the spectroscope 3 to pass through an objective lens 7 and then reflected by a reference mirror 8 to form reference light, and the reference light and the test light form an interference light signal. The interference light signal is incident on the target surface of the CCD11 through the spectroscope 3 and the tube mirror 10, and an interference image is received.
After the system is built, a white light source 1 is turned on, white light is irradiated on a spectroscope 3 through a Kohler lighting system 2, reflected light generated by the spectroscope 3 is transmitted downwards and is irradiated on an object 5 to be measured through an objective lens 4, and the reflected light is irradiated on the spectroscope 3 to form test light. The transmitted light generated by the spectroscope 3 is irradiated to the reference mirror 8 through the objective lens 7, and the reference mirror 8 is reflected and irradiated to the spectroscope 3 again to form the reference light. The reference light and the test light interfere to generate an interference light signal, and the interference light signal is transmitted upwards through the spectroscope 3 and the tube mirror 10 and then received by the CCD 11. The control and data processing system 12 controls the scanning system 13 to perform time-sequential vertical scanning according to the flow.
Step 1, focusing an interference objective of an interferometer with a Linnik structure to the bottom end of a sample, and recording the position b of a PZT (piezoelectric transducer) of a test arm at the moment 1 The method comprises the steps of carrying out time sequence vertical coarse scanning on a sample according to a scanning direction from bottom to top, simultaneously controlling a CCD to collect images, calculating the images by using a gray variance evaluation function, and further automatically positioning the top end of the sample, wherein the coarse scanning step length must ensure that a series of interference images collected by the CCD must have a frame falling in a range where the top end interferes, and enabling a gray variance evaluation function calculation result corresponding to the frame to become an extreme value. The calculation formula of the gray variance evaluation function of the image is as follows:
Figure BDA0003254793380000071
Figure BDA0003254793380000072
in the formula, var i Is the gray variance value of a certain frame image, W is the row number of the image gray matrix, H is the column number of the image gray matrix, and I (x, y) is the gray value at the position of the image gray matrix (x, y); u represents an intermediate function; image frame number i =1,2,3, … ….
The range of the interference generated at the top is twice the coherent length Δ L of the light source, and the coherent length Δ L of the light source is expressed as follows:
Figure BDA0003254793380000081
wherein, delta lambda is the spectral width of the spectral distribution of the light source, lambda 0 The center wavelength of the light source.
Positioning frame gray scale variance value var in coarse scanning process p Greater than the gray variance value var of the front and back frames p-1 、var p+1 Recording the scanning position a of the corresponding PZT 1 。
Then at position a 1 Performing time sequence vertical middle scanning for precisely positioning the top end of a sample, simultaneously controlling a CCD to collect an image, wherein the middle scanning range is equal to 2 times of coarse scanning step length, the middle scanning step length needs to be scanned for 14 times in the middle scanning range, calculating the image by adopting a gray variance evaluation function, and recording the PZT scanning position corresponding to a frame of image with the maximum gray variance value var as a 2 . In the position a 2 Defining a scan range of + -wide for the center 1 And performing time sequence vertical fine scanning with step length to control the CCD to synchronously acquire a sample top interference image C, wherein wide 1 ＝4×ΔL，step＝λ 0 /8。
Step 2, controlling PZT to return to position b 1 At position b 1 Time-series vertical middle scanning is carried out for centering, the bottom end of a sample is precisely positioned, and the corresponding PZT position b is recorded 2 In the position b 2 Defining a scan range of + -wide for the center 1 And performing time sequence vertical fine scanning with step length, and controlling the CCD to synchronously acquire the interference image D at the bottom end of the sample.
Step 3, extracting coherent signals corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, calculating the maximum value of each pixel point coherent signal, and splicing all the obtained maximum values into a light intensity information image I max As shown in fig. 4, wherein:
I max ＝max(I(x,y,z)) (4)
in the formula, I (x, y, z) represents a coherent signal of any pixel point extracted from the sample top interference image C and the sample bottom interference image D.
Step 4, extracting a coherent signal corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, and obtaining a coherent signal envelope by adopting a contrast calculation mode, wherein the calculation mode of the contrast M of a certain pixel point in the sample top end interference image C and the sample bottom end interference image D is as follows:
Figure BDA0003254793380000082
in the formula, M (I) and I i Respectively calculating the contrast value and the light intensity value of the pixel point in the ith frame of the interference image, and obtaining the contrast of the pixel point, namely the signal envelope, after calculating the contrast values of all the frames of the interference image.
Then, a gravity center method is adopted to obtain a coherent signal envelope gravity center position h and obtain a corresponding scanning position, wherein a calculation expression of the gravity center method is as follows:
Figure BDA0003254793380000091
in the formula, N is the total frame number of the acquired interference image.
Obtaining scanning positions according to the gravity center position h, calculating pixel by pixel and splicing all the scanning positions into an envelope information image E cen As shown in fig. 5.
Step 5, multiplying the light intensity information image I max With the envelope information image E cen And combining to obtain a composite image, and increasing the difference of pixel point signals at the top end and the bottom end of the high-aspect-ratio structure. And converting the synthesized image into a gray histogram, and calculating an image segmentation threshold t by an OTSU algorithm. The OTSU algorithm divides the image gray histogram into A, B two parts, and determines the optimal gray threshold for image segmentation by searching the maximum variance between the two parts, wherein the inter-class variance
Figure BDA0003254793380000092
The calculation expression of (c) is as follows:
Figure BDA0003254793380000093
in the formula w A 、w B Representing the probability of each part in the population, u A 、u B Representing the mathematical expectation of each part and E representing the mathematical expectation of the ensemble.
Threshold t is between classesVariance (variance)
Figure BDA0003254793380000094
At the maximum gray level position, the pixel points higher than the threshold value t are regarded as the points at the top end of the structure, the pixel points lower than the threshold value t are regarded as the points at the bottom end of the structure, the composite image is binarized, as shown in fig. 6, the transverse resolution of the sample structure is completed, and the problem that the step edge is difficult to accurately position when the structural form of the high aspect ratio structure with the transverse feature close to the diffraction limit is restored at present is solved.
Step 6, positioning a scanning position corresponding to the maximum value of the envelope of the coherent signal for each pixel point positioned at the top end of the high-aspect-ratio structure; for the pixel point at the bottom of the high aspect ratio structure, the position b 2 Defining windows ± wide for the center 2 Locating the scanning position corresponding to the maximum value of the envelope of the coherent signal in the window, wherein wide 2 =40 × step; position a recorded from PZT 2 、b 2 And calculating the distance L between the top end and the bottom end in the scanning process due to automatic fixed focus jump, wherein the calculation mode is as follows:
L＝|a 2 -b 2 | (8)
the method provided by the invention collects 213 frames of images from the top and the bottom of the sample respectively, and positions a obtained by fine positioning 2 270.14 μm, position b obtained by fine positioning 2 And for 346.04 μm, calculating to obtain a distance L which skips 75.9 μm, adding L into the scanning position result of each pixel point at the top end of the high aspect ratio structure, calculating and splicing all scanning positions pixel by pixel according to the method, and completing automatic detection of the three-dimensional morphology of the high aspect ratio structure, wherein the final three-dimensional morphology result is shown in fig. 9 and 10.
In the embodiment, the measurement of the high-aspect-ratio comb-shaped structure sample with the height reaching the micrometer level is realized through a series of measures, and the three-dimensional appearance of the sample is finally restored. The experimental result shows that the coincidence degree of the measured value and the actual value is high, the scanning speed is high, the calculation time of the algorithm is obviously shorter than that of the traditional vertical scanning measurement algorithm, and the method has the advantages of good real-time performance, high speed and high measurement precision.

Claims (9)

1. An interference measurement method suitable for detecting three-dimensional shape of a high aspect ratio structure comprises the following steps:
step 1, focusing an interference objective lens of an interferometer with a Linnik structure to the bottom end of a sample, and recording the position b of a PZT of a test arm at the moment 1 Performing time sequence vertical coarse scanning on the sample according to the scanning direction from bottom to top, simultaneously controlling a CCD (charge coupled device) to collect an image, calculating the image by using a gray variance evaluation function, and further automatically positioning the top end of the sample and recording the PZT (lead zirconate titanate) position a 1 In the position a 1 Scanning in time sequence vertical for precisely positioning sample top end and recording corresponding PZT position a 2 In the position a 2 Defining a scan range of + -wide for the center 1 Performing time sequence vertical fine scanning with step length, controlling the CCD to synchronously acquire a sample top end interference image C, and turning to the step 2;
step 2, controlling PZT to return to position b 1 At position b 1 Scanning in time sequence vertical for precisely positioning the bottom end of the sample and recording the corresponding PZT position b 2 In the position b 2 Defining a scan range of + -wide for the center 1 Performing time sequence vertical fine scanning with step, controlling the CCD to synchronously acquire a sample bottom end interference image D, and turning to the step 3 and the step 4;
step 3, extracting coherent signals corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, calculating the maximum value of each pixel point coherent signal, and splicing all the obtained maximum values into a light intensity information image I max Turning to step 5;
step 4, extracting a coherent signal corresponding to each pixel point in the sample top end interference image C and the sample bottom end interference image D, obtaining coherent signal envelopes in a contrast calculation mode, positioning the coherent signal envelope gravity center position by adopting a gravity center method and obtaining corresponding scanning positions, and splicing all scanning positions pixel by pixel to obtain an envelope information image E cen And then, turning to the step 5;
step 5, the light intensity information image I max With the envelope information image E cen Combining to obtain a composite image, increasing the difference of pixel point signals at the top end and the bottom end of the high-aspect-ratio structure, converting the composite image into a gray histogram, calculating an image segmentation threshold value t through an OTSU algorithm, performing binarization segmentation on the composite image to distinguish the top end and the bottom end of the high-aspect-ratio structure, realizing transverse resolution, and turning to step 6;
step 6, positioning a scanning position corresponding to the envelope maximum value of the coherent signal for each pixel point positioned at the top end of the high-aspect-ratio structure; for the pixel point at the bottom of the high aspect ratio structure, the position b 2 Defining a window for the center + -wide 2 Positioning a scanning position corresponding to the maximum value of the envelope of the coherent signal in the window; position alpha recorded from PZT 2 、b 2 Calculating the distance L of the automatic focus-fixing jump between the top end and the bottom end in the time sequence vertical fine scanning process, and adding the distance L into the scanning position result of each pixel point at the top end of the high-aspect-ratio structure; and calculating and splicing all the scanning positions pixel by pixel according to the method to obtain the three-dimensional shape of the high-aspect-ratio structure.
2. The interferometry method suitable for detecting the three-dimensional shape of a high-aspect-ratio structure according to claim 1, wherein: in the step 1, the time sequence vertical coarse scanning step length must ensure that a series of interference images acquired by the CCD must fall within the range where interference occurs at the top end by one frame, and the gray variance evaluation function calculation result corresponding to the frame becomes an extreme value; the calculation formula of the gray variance evaluation function of the image is as follows:
Figure FDA0003731333740000021
Figure FDA0003731333740000022
in the formula, var i Is the gray variance value of a certain frame image, W is the number of rows of the image gray matrix, H is the number of columns of the image gray matrix, I (x, y)) Is the gray value at the image gray matrix (x, y) location; u represents an intermediate function; image frame number i =1,2,3, … …;
positioning frame gray scale variance value var in coarse scanning process p Greater than the gray variance value var of the front and back frames p-1 、var p+1 Recording the scanning position a of the corresponding PZT 1 。
3. The interferometry method suitable for detecting the three-dimensional profile of a high-aspect-ratio structure according to claim 2, wherein: the range of the interference at the top is twice the coherence length Δ L of the light source, which is expressed as follows:
Figure FDA0003731333740000023
wherein, delta lambda is the spectral width of the spectral distribution of the light source, lambda 0 The center wavelength of the light source.
4. The interferometry method suitable for detecting the three-dimensional profile of a high-aspect-ratio structure according to claim 3, wherein: in the step 1, the position a is used 1 Defining a scanning range for the center to perform time sequence vertical middle scanning, simultaneously controlling a CCD to acquire an image, wherein the scanning range is equal to 2 times of coarse scanning step length, the middle scanning step length needs to be scanned for 14 times in the middle scanning range, calculating the image by adopting a gray variance evaluation function, and marking a PZT scanning position corresponding to a frame of image with the maximum gray variance value var as a 2 。
5. The interferometry method suitable for detecting the three-dimensional profile of a high-aspect-ratio structure according to claim 4, wherein: in the step 1, the position a 2 Defining a scan range of + -wide for the center 1 Performing a time-sequential vertical fine scan with step size, wherein wide 1 ＝4×ΔL，step＝λ 0 /8。
6. According to the rightThe interferometry method suitable for detecting the three-dimensional shape of the high aspect ratio structure according to claim 5, wherein the interferometry method comprises the following steps: in the step 3, the maximum value of the coherent signal of each pixel point in the sample top end interference image C and the sample bottom end interference image D is calculated and spliced into the light intensity information image I max Wherein:
I max ＝max(I(x,y,z)) (4)
in the formula, I (x, y, z) represents a coherent signal of any pixel point extracted from the sample top end interference image C and the sample bottom end interference image D.
7. The interferometry method suitable for detecting the three-dimensional profile of a high-aspect-ratio structure according to claim 6, wherein: in the step 4, when signal envelope calculation is performed on the coherent signal, a signal envelope is obtained by adopting a contrast calculation mode, and the calculation modes of the contrast M of a certain pixel point in the sample top interference image C and the sample bottom interference image D are as follows:
Figure FDA0003731333740000031
in the formula, M (I) and I i Respectively calculating the contrast value and the light intensity value of the pixel point in the ith frame of interference image, and then obtaining the contrast of the pixel point, namely the signal envelope, after calculating the contrast values of all the frames of interference images;
and then processing the coherent signal envelopes of each pixel point in the sample top end interference image C and the sample bottom end interference image D by adopting a gravity center method to obtain the gravity center position h of the signal envelopes, wherein the calculation expression of the gravity center method is as follows:
Figure FDA0003731333740000032
in the formula, N is the total frame number of the acquired interference image;
obtaining scanning positions according to the gravity center position h, calculating pixel by pixel and splicing all the scanning positionsEnvelope information image E cen 。
8. The interferometry method suitable for detecting the three-dimensional profile of a high-aspect-ratio structure according to claim 7, wherein: in the step 5, the light intensity information image I is multiplied by the light intensity information image I max With the envelope information image E cen Combining to obtain a synthetic image, increasing the difference of pixel point signals at the top end and the bottom end of the high-aspect-ratio structure, converting the synthetic image into a gray histogram, calculating an image segmentation threshold value t through an OTSU algorithm, and performing binarization segmentation on the synthetic image to distinguish the top end and the bottom end of the high-aspect-ratio structure so as to realize transverse resolution, wherein the method specifically comprises the following steps:
the OTSU algorithm divides the image gray histogram into A, B two parts, and determines the optimal gray threshold for image segmentation by searching the maximum variance between the two parts, wherein the between-class variance is delta 1 The calculation expression of (a) is as follows:
Figure FDA0003731333740000033
in the formula w A 、w B Denotes the probability of each part in the population, u A 、u B Representing the mathematical expectation of each part, E representing the mathematical expectation of the population;
threshold t is between-class variance
Figure FDA0003731333740000034
And at the maximum gray level position, regarding the pixel point higher than the threshold value t as the point at the top end of the structure, regarding the pixel point lower than the threshold value t as the point at the bottom end of the structure, binarizing the synthesized image, and finishing the transverse resolution of the sample structure.
9. The interferometry method suitable for detecting the three-dimensional profile of a high-aspect-ratio structure according to claim 8, wherein: in the step 6, for each pixel point positioned at the top end of the high-aspect-ratio structure, positioning a scanning position corresponding to the envelope maximum value of the coherent signal; for the structure with high depth-width ratioEach pixel point at the bottom end by position b 2 Defining a window for the center + -wide 2 Locating the scanning position corresponding to the maximum value of the envelope of the coherent signal in the window, wherein wide 2 =40 × step; for the distance L between the top end and the bottom end time sequence vertical fine scanning due to the automatic fixed focus jump, the calculation mode is as follows:
L＝|a 2 -b 2 | (8)
and adding the L into the scanning position result of each pixel point at the top end of the high-aspect-ratio structure, calculating pixel by pixel, and splicing all the scanning positions to obtain the three-dimensional shape of the high-aspect-ratio structure.
CN202111056510.7A 2021-09-09 2021-09-09 Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure Active CN113847883B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111056510.7A CN113847883B (en) 2021-09-09 2021-09-09 Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111056510.7A CN113847883B (en) 2021-09-09 2021-09-09 Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure

Publications (2)

Publication Number Publication Date
CN113847883A CN113847883A (en) 2021-12-28
CN113847883B true CN113847883B (en) 2022-10-14

Family

ID=78973791

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111056510.7A Active CN113847883B (en) 2021-09-09 2021-09-09 Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure

Country Status (1)

Country Link
CN (1) CN113847883B (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110779464A (en) * 2019-11-18 2020-02-11 重庆邮电大学 Time domain and frequency domain joint analysis broad spectrum coherence measurement method and system

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6426501B1 (en) * 1998-05-27 2002-07-30 Jeol Ltd. Defect-review SEM, reference sample for adjustment thereof, method for adjustment thereof, and method of inspecting contact holes
US8970850B2 (en) * 2012-12-14 2015-03-03 University Of New Brunswick Method and apparatus for surface profilometry
CN103791853A (en) * 2014-01-20 2014-05-14 天津大学 Microstructure measuring device and measuring method based on color strip information processing
CN106767496A (en) * 2016-11-18 2017-05-31 中国科学院光电技术研究所 3D morphology recovery method combining phase shift interference and vertical scanning interference
CN109751971B (en) * 2019-01-23 2020-10-02 南京理工大学 High-precision short coherent light three-dimensional morphology rapid measurement algorithm
CN112097645B (en) * 2020-08-31 2021-12-10 南京理工大学 High depth-width ratio micro-structure reflection type interference microscopic nondestructive measuring device
CN113091634B (en) * 2021-03-01 2022-09-13 南京理工大学 Rapid microscopic morphology measuring method suitable for white light scanning interference

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110779464A (en) * 2019-11-18 2020-02-11 重庆邮电大学 Time domain and frequency domain joint analysis broad spectrum coherence measurement method and system

Also Published As

Publication number Publication date
CN113847883A (en) 2021-12-28

Similar Documents

Publication Publication Date Title
CN109269438B (en) Structured light illumination microscopic measurement method for detecting multilayer complex micro-nano structure
CN104964982B (en) Glass surface true and false defect identification method and system based on OCT complex signals
CN110715616B (en) Structured light micro-nano three-dimensional morphology measurement method based on focusing evaluation algorithm
CN113091634B (en) Rapid microscopic morphology measuring method suitable for white light scanning interference
CN112097645B (en) High depth-width ratio micro-structure reflection type interference microscopic nondestructive measuring device
CN101131317A (en) Method and device for measuring micro-nano deep groove structure
CN115325963B (en) Wafer surface three-dimensional shape measuring device and measuring method thereof
EP0647310B1 (en) Method of an apparatus for interferometrically inspecting a surface of an object
CN113588682A (en) Large-range high-precision rapid defect detection system for 3D parts
CN113551614B (en) Three-dimensional topography interferometry rapid vertical scanning method based on spectral distribution characteristics
CN103226001A (en) Measuring device and measuring method for tiny surface defect through post-magnification digital holographic microscopy
WO2024051857A1 (en) Translation-difference-based microscopic non-destructive measurement method for micro-structure linewidth
CN113916154A (en) Self-calibration structured light measuring method based on constant half width of modulation degree
CN113847883B (en) Interferometric method suitable for detecting three-dimensional shape of high aspect ratio structure
US11733034B2 (en) Fast measurement method for micro-nano deep groove structure based on white light interference
CN115235345B (en) High aspect ratio microstructure transmission type interference microscopic nondestructive measurement device and measurement method
CN215574729U (en) Large-range high-precision rapid defect detection system for 3D parts
CN115165889A (en) Sample characteristic analysis method and device based on through-focus scanning imaging
CN111829457B (en) Method for detecting three-dimensional morphology of ultrathin film device based on structured light illumination microscope system
CN114252007A (en) Optical detection device and method for three-dimensional structure of micro-nano device
CN117685899B (en) Method for measuring pattern structure morphology parameters
Conroy et al. Recent advances in non-contact metrology, high speed measurement, steep slope measurement and correlation with stylus data
Alvarez et al. Toward online non-contact roughness profile measurements with a sensor based on conoscopic holography: current developments
TWI667448B (en) Inspecting device of gratings and inspecting method of same
CN117198913A (en) Device and method for detecting defects of pattern wafer and measuring three-dimensional morphology

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant