System and method for detecting wafers using nonlinear optical signals

Abstract

To provide a system and method for detecting wafers using nonlinear optical signals. [Solution] A system and method for detecting a wafer using a nonlinear optical signal acquires the emitted light from the sample under test using a noise reduction module. The core assembly of the noise reduction module includes a series of beam splitter assemblies and a photodetector assembly that receives the ratio beams of the beam splitter assemblies. By reducing the signal from the homogeneous region of the sample under test and the background noise that randomly occurs during the photoelectric conversion process, the overall background noise is reduced, thereby clarifying the frequency multiplier signal to be retained and improving the signal-to-noise ratio.

Description

Technical Field

[0001] This invention claims priority based on Taiwan Patent Application No. 113143874.

Background Art

[0002] For example, wide bandgap (WBG) materials such as silicon carbide (SiC) have excellent properties such as high breakdown voltage, high frequency, high temperature resistance, and low loss, so they are widely used in fields such as high-power and high-frequency electronic components. However, fine defects inside SiC wafers reduce the performance of components and may even lead to loss of function, so detection technology is extremely important.

[0003] There are mainly two types of conventional methods for detecting defects in SiC wafers.

[0004] Destructive inspection: For example, KOH etching can detect crystal defects, but it damages the wafer, so the material is wasted.

[0005] Non-destructive inspection: For example, X-ray diffraction, optical microscopy, etc. Usually, these methods can only detect defects on the wafer surface and cannot effectively detect defects in crystals deep inside the wafer.

[0006] Nonlinear optical detection technology is a new non-destructive and non-contact detection method. Nonlinear optics is highly sensitive to the crystal structure and defects of materials, so it can detect fine defects that were difficult to discover by conventional methods. Moreover, since there is no need to directly contact the material, damage to the material is avoided.

[0007] The prior art document Patent Document 1, "Wafer Detection System and Detection Method," relates to a technique for detecting crystal defects using nonlinear optics. The detection system includes a mounting stage, a light source module, an optical detection module, and a beam splitter. The emitted light reflected from the sample under test is separated into a first harmonic signal and a second harmonic signal by the beam splitter. Then, by analyzing the first harmonic signal and the second harmonic signal respectively using the optical detection module, surface characteristic parameters of the sample under test (including the thickness of the thin film on the surface of the sample under test, refractive index, and absorption coefficient) are obtained based on the first harmonic signal, and "electrical characteristic data" of the sample under test is obtained based on the second harmonic signal. By combining the surface characteristic parameters and electrical characteristic data of the sample under test in this way, more accurate measurements of the wafer can be performed. Regarding the "electrical characteristic data," it is determined whether or not there is an abnormality in the electrical performance of the sample under test based on whether or not the ratio of the actual p-polarization component and the ratio of the actual s-polarization component are within a preset range of reference values.

[0008] Nonlinear optical detection can also acquire two-dimensional or three-dimensional images of defects within a wafer, intuitively presenting (visualizing) the distribution and morphology of the defects. However, the low signal-to-noise ratio (SNR) of the images is one of the bottlenecks of this technology. The main factors contributing to the low SNR of images are the inclusion of noise into the homogeneous regions of the sample under test, and background noise randomly generated by the photodetector during the photoelectric conversion process. These noises obscure the frequency multiplier signals representing the defective regions of the wafer, thereby reducing the SNR and affecting the final measurement or detection accuracy.

[0009] In the prior art described in Patent Document 2, "Reducing Shot Noise by Frame Averaging," an image processing algorithm is used to suppress noise within the image and further identify defects. In this method, a sample (an unpatterned wafer) is irradiated with an illumination light source (including any laser system) to acquire multiple images. These multiple images are divided into a detection image and a reference image. The multiple images are repeated images from a single region of a single sample. This region may be a detection region and a reference region. By averaging the multiple images, an averaged image is generated for the region. Averaging is performed by adding the intensity values ​​of each common pixel position and then dividing by the number of images. It is also possible to generate a difference image by subtracting the reference image from the average detection image and then analyze the difference image to detect defects in the detection region of the sample. [Prior art documents] [Patent Documents]

[0010] [Patent Document 1] Chinese Patent No. 117491384 Specification [Patent Document 2] Taiwan Patent Application Publication No. 202405409 [Overview of the Initiative] [Problems that the invention aims to solve]

[0011] The signal-to-noise ratio (SNR) of the image is improved, but without using image processing algorithms. Instead, a noise reduction module reduces noise in the homogeneous region of the sample and background noise during the photoelectric conversion process. [Means for solving the problem]

[0012] A system and method for detecting defects in a wafer using optical frequency multiplication technology, wherein a noise reduction module acquires emitted light at the same position and time in a designated area of ​​the sample under test. The core assembly of the noise reduction module includes a series of beam splitter assemblies and a photodetector assembly that receives the ratio beams of the beam splitter assemblies. The noise reduction module reduces the signal from the homogeneous region of the sample under test and the background noise randomly generated by the photodetector during the photoelectric conversion process, thereby reducing the overall background noise, clarifying the frequency multiplied signal to be retained, and improving the signal-to-noise ratio. [Brief explanation of the drawing]

[0013] [Figure 1] This is Example 1 of the noise reduction module according to the present invention. [Figure 2] This is an embodiment 2 of the noise reduction module according to the present invention. [Figure 3] This is an embodiment 3 of the noise reduction module according to the present invention. [Figure 4] This is an embodiment 4 of the noise reduction module according to the present invention. [Figure 5] This is the fifth embodiment of the noise reduction module according to the present invention. [Figure 6] This is an embodiment 6 of the noise reduction module according to the present invention. [Figure 7] This is an embodiment 7 of the noise reduction module according to the present invention. [Figure 8] This is an example 8 of the noise reduction module according to the present invention. [Figure 9] This is the ninth embodiment of the noise reduction module according to the present invention. [Figure 10] This is an image of the test sample used in the present invention. [Modes for carrying out the invention]

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings and examples. The drawings provided in the following examples only schematically illustrate the basic idea of the present invention. In the drawings, only the assemblies related to the present invention are shown, and they are not described based on the number of parts, shape, and size in actual implementation. The above is explained in advance.

[0015] As shown in FIGS. 1 to 9, in a first aspect, the wafer detection system of the present invention includes the following.

[0016] The mounting table 10 is used to mount the test sample 11. The test sample 11 is a test wafer, particularly a SiC wafer. A specular reflection structure can be installed on the mounting table 10. The specular reflection structure faces the test sample 11.

[0017] The light source module 20 generates fundamental frequency light 21 and guides it to the test sample 11. Also, emitted light 13 is generated from the test sample 11. The light source module 20 carries fundamental frequency light of a specific wavelength. The light source includes, but is not limited to, a pulsed laser with an irradiation intensity of 100 to 200 mW and a repetition rate of 30 to 80 MHz. The fundamental frequency light 21 emitted from the light source module 20 is converged and / or amplified via a series of optical elements including, but not limited to, lenses, mirrors, and objective lenses for magnifying magnification. Thereby, the fundamental frequency light 21 can be accurately irradiated onto a specified region on the test sample 11, and it is surely guaranteed that the emitted light 13 that holds the frequency doubling signal is generated from the specified region.

[0018] The noise reduction module 30 is used to receive the emitted light 13 at the same position and the same time in the specified area of the test sample 11. The core assembly of the noise reduction module 30 includes a series of beam splitter assemblies and a photodetector assembly that receives the ratio beams of the beam splitter assemblies. The photodetector assembly includes a plurality of photomultiplier tubes (PMTs). A photomultiplier tube can convert a weak optical signal into a measurable current and amplify it significantly. Based on this, in the present invention, instead of the photomultiplier tube, other devices including a photodiode (PD), an avalanche photodiode (APD), and a charge-coupled device (CCD) that can convert an optical signal into an electrical signal and amplify it (but not limited to these) may be used. The noise reduction module 30 is used to reduce the signal of the homogeneous region of the test sample 11 and enhance the signal of the defective region. Since there are multiple embodiments of the noise reduction module 30, they will be described in detail later. In a preferred embodiment, the photodetector assembly includes a plurality of photodetectors. Each of the photodetectors can enhance the amplification ability of the photodetector for the incident optical signal by adjusting the output gain. Different gain adjustment methods are available depending on the type of photodetector (e.g., photomultiplier tube PMT or photodiode). Taking the photomultiplier tube (PMT) as an example, the gain is mainly determined by the voltage applied between the photocathode and the dynodes. When the high voltage is increased, the electric field strength between the dynodes is improved, so that the energy when electrons are multiplied at each electrode is further increased, and the final output gain is increased. Usually, the relationship between the gain and voltage of the PMT is an exponential relationship: G∝V n is shown. G is the gain, V is the applied voltage, and n is the exponent related to the number of dynodes (usually between 5 and 8). Usually, a high voltage power supply module is installed in the PMT, and the voltage can be adjusted by manually turning a knob, entering a numerical value for setting, or by a computer program.

[0019] The signal processing and imaging module 40 is connected to the noise reduction module 30 and acquires and processes electrical signals from the noise reduction module 30 to generate an image of the defective region of the sample under test 11. The signal processing and imaging module 40 can enhance the frequency-multiplied image by utilizing (but not limited to) known image processing techniques, including (but not limited to) color addition, grayscale multiplication, and grayscale sharpening multiplication.

[0020] An embodiment of the noise reduction module 30 is as follows.

[0021] Figure 1, Example 1 The beam splitter assembly includes one beam splitter, and the photodetector assembly includes two photodetectors.

[0022] The beam splitter is defined as a first beam splitter 313. The first beam splitter 313 is provided on the optical path of the emitted light 13 from the sample under test 11. The spectral ratio (branching ratio) of the beam splitter is 50:50, and the tolerance is ±5%. The emitted light 13 from the sample under test 11 is spectrally split into a first beam 51 and a second beam 52. The first beam 51 is 50% ratio light, and the second beam 52 is 50% ratio light. The first beam 51 is received by a first photodetector 321, and the second beam 52 is received by a second photodetector 322.

[0023] Figure 2, Example 2 The beam splitter assembly includes one beam splitter, and the photodetector assembly includes two photodetectors. Furthermore, it includes a reflector 60.

[0024] The beam splitter is provided in the optical path of the emitted light 13. The spectral ratio of the beam splitter is 50:50, and it spectrally separates the emitted light 13 from the sample under test 11 into a first beam 51 and a second beam 52. The first beam 51 is received by the first photodetector 321. The reflector 60 is positioned in the optical path of the second beam 52 and focuses and reflects the second beam 52 for reception by the second photodetector 322.

[0025] Figure 3, Example 3 The beam splitter assembly includes two beam splitters, and the photodetector assembly includes three photodetectors.

[0026] The two beam splitters are defined as the first beam splitter 313 and the second beam splitter 314, respectively. The beam splitters may be general beam splitters or polarized beam splitters. The three photodetectors are defined as the first photodetector 321, the second photodetector 322, and the third photodetector 323, respectively.

[0027] The first beam splitter 313 is provided in the optical path of the emitted light 13 from the sample under test 11. The spectral ratio of the first beam splitter 313 is 30:70, and the tolerance is ±5%. The emitted light 13 is spectrally split into a first beam 51 and a second beam 52. The first beam 51 is 30% ratio light, and the second beam 52 is 70% ratio light. The first beam 51 is received by the first photodetector 321.

[0028] The second beam splitter 314 is provided in the optical path of the second beam 52. The spectral ratio of the second beam splitter 314 is 50:50, and the tolerance is ±5%. As a result, the second beam 52 (70% ratio light) is further spectrally split into a third beam 53 and a fourth beam 54. Thus, the third beam 53 is 35% ratio light, and the fourth beam 54 is 35% ratio light. The third beam 53 is received by the second photodetector 322, and the fourth beam 54 is received by the third photodetector 323.

[0029] Figure 4, Example 4 The beam splitter assembly includes two beam splitters, and the photodetector assembly includes three photodetectors. Furthermore, it includes a reflector 60.

[0030] The two beam splitters are defined as the first beam splitter 313 and the second beam splitter 314, respectively. The beam splitters may be general beam splitters or polarized beam splitters. The three photodetectors are defined as the first photodetector 321, the second photodetector 322, and the third photodetector 323, respectively.

[0031] The first beam splitter 313 is provided in the optical path of the emitted light 13 from the sample under test 11. The spectral ratio of the first beam splitter 313 is 30:70, and the tolerance is ±5%. The emitted light 13 is spectrally split into a first beam 51 and a second beam 52. The first beam 51 is 30% ratio light, and the second beam 52 is 70% ratio light. The first beam 51 is received by the first photodetector 321.

[0032] The second beam splitter 314 is provided in the optical path of the second beam 52. The spectral ratio of the second beam splitter 314 is 50:50, and the tolerance is ±5%. The second beam 52 (70% ratio light) is further spectrally split into a third beam 53 and a fourth beam 54. Thus, the third beam 53 is 35% ratio light, and the fourth beam 54 is 35% ratio light. The third beam 53 is received by the second photodetector 322. The reflector 60 is provided in the optical path of the fourth beam 54, and focuses and reflects the fourth beam 54 so that it is received by the third photodetector 323.

[0033] Figure 5, Example 5 The beam splitter assembly includes three beam splitters, and the photodetector assembly includes four photodetectors.

[0034] The three beam splitters are defined as the first beam splitter 313, the second beam splitter 314, and the third beam splitter 315, respectively. The beam splitters may be general beam splitters or polarized beam splitters. The four photodetectors are defined as the first photodetector 321, the second photodetector 322, the third photodetector 323, and the fourth photodetector 324, respectively.

[0035] The first beam splitter 313 is provided in the optical path of the emitted light 13 from the sample under test 11. The spectral ratio of the first beam splitter 313 is 20:80, and the tolerance is ±5%. The emitted light 13 is spectrally split into a first beam 51 and a second beam 52. The first beam 51 is 20% ratio light, and the second beam 52 is 80% ratio light. The first beam 51 is received by the first photodetector 321.

[0036] The second beam splitter 314 is positioned in the optical path of the second beam 52. The spectral ratio of the second beam splitter 314 is 30:70, and the tolerance is ±5%. As a result, the second beam 52 (80% ratio light) is further spectrally split into a third beam 53 and a fourth beam 54. Thus, the third beam 53 is 24% ratio light, and the fourth beam 54 is 56% ratio light. The third beam 53 is received by the second photodetector 322.

[0037] The third beam splitter 315 is positioned in the optical path of the fourth beam 54. The spectral ratio of the third beam splitter 315 is 50:50, and the tolerance is ±5%. As a result, the fourth beam 54 (56% ratio light) is further spectrally split into the fifth beam 55 and the sixth beam 56. Thus, the fifth beam 55 is 28% ratio light, and the sixth beam 56 is 28% ratio light. The fifth beam 55 is received by the third photodetector 323, and the sixth beam 56 is received by the fourth photodetector 324.

[0038] Figure 6, Example 6 The beam splitter assembly includes three beam splitters, and the photodetector assembly includes four photodetectors. Furthermore, it includes a reflector 60.

[0039] The three beam splitters are defined as the first beam splitter 313, the second beam splitter 314, and the third beam splitter 315, respectively. The beam splitters may be general beam splitters or polarized beam splitters. The four photodetectors are defined as the first photodetector 321, the second photodetector 322, the third photodetector 323, and the fourth photodetector 324, respectively.

[0040] The first beam splitter 313 is provided in the optical path of the emitted light 13 from the sample under test 11. The spectral ratio of the first beam splitter 313 is 20:80, and the tolerance is ±5%. The emitted light 13 is spectrally split into a first beam 51 and a second beam 52. The first beam 51 is 20% ratio light, and the second beam 52 is 80% ratio light. The first beam 51 is received by the first photodetector 321.

[0041] The second beam splitter 314 is positioned in the optical path of the second beam 52. The spectral ratio of the second beam splitter 314 is 30:70, and the tolerance is ±5%. As a result, the second beam 52 (80% ratio light) is further spectrally split into a third beam 53 and a fourth beam 54. Thus, the third beam 53 is 24% ratio light, and the fourth beam 54 is 56% ratio light. The third beam 53 is received by the second photodetector 322.

[0042] The third beam splitter 315 is positioned in the optical path of the fourth beam 54. The spectral ratio of the third beam splitter 315 is 50:50, and the tolerance is ±5%. As a result, the fourth beam 54 (56% ratio light) is further spectrally split into the fifth beam 55 and the sixth beam 56. Thus, the fifth beam 55 is 28% ratio light, and the sixth beam 56 is 28% ratio light. The fifth beam 55 is received by the third photodetector 323. The reflector 60 is positioned in the optical path of the sixth beam 56 and focuses and reflects the sixth beam 56 for reception by the fourth photodetector 324.

[0043] Figures 7, 8, and 9, Example 7, Example 8, and Example 9 A lens 61, a light-reducing filter 62, and a polarizing element 63 can be arranged selectively or in combination in the optical path before the photodetector.

[0044] As is clear from the above-mentioned multiple embodiments, the embodiments of the noise reduction module 30 have the following principles.

[0045] 1. Two or more photodetectors are connected in parallel.

[0046] 2. A beam splitter is used to allocate the beam to the photodetector in the appropriate ratio.

[0047] 3. The beam splitters are arranged so that the difference in light intensity received by each photodetector is within 10%.

[0048] 4. By adjusting the gain value of the photodetector in each path, the difference in signal intensity output by each path is kept within 10%.

[0049] 5. Each pathway simultaneously acquires light intensity signals from the same location and time of the test sample 11.

[0050] This invention uses a noise reduction module consisting of multiple beam splitters and multiple photodetectors to reduce the signal in the homogeneous region of the sample under test and the background noise randomly generated by the photodetectors during the photoelectric conversion process. By reducing the background noise overall, the frequency-multiplied signal to be preserved is clarified, and the signal-to-noise ratio (SNR) is improved. For related images, please refer to Figure 10. The bright spots indicated by frames in the image are wafer defects represented by frequency-multiplied light, and there is almost no noise in the background of the image. [Explanation of symbols]

[0051] 10 Mounting platform 11 Test samples 13 Emitted light 20 Light Source Modules 21 fundamental frequency light 30 Noise Reduction Modules 40 Signal Processing and Imaging Modules 313 First Beam Splitter 314 Second Beam Splitter 315 Third Beam Splitter 321 First photodetector 322 Second photodetector 323 Third photodetector 324. Fourth photodetector 51. First Beam 52. Second Beam 53 Third Beam 54. Fourth beam 55. Fifth Beam 56. Beam 6 60 Reflector 61 lenses 62 Neutral Density Filters 63 Polarizing elements

Claims

1. A system for detecting wafer defects using nonlinear optical signals, A light source module equipped with fundamental frequency light of a specific wavelength, the fundamental frequency light is irradiated onto a designated area of ​​the sample under test, and light is emitted from the designated area, A noise reduction module comprising at least one beam splitter assembly and a photodetector assembly that receives the ratio beam of the beam splitter, wherein the photodetector assembly converts the optical signal into an electrical signal and amplifies it, and the noise reduction module receives the emitted light at the same position and time in the designated region, and reduces the signal from the homogeneous region of the designated region and the background noise randomly generated by the photodetector assembly during the photoelectric conversion process, thereby reducing the background noise overall, clarifying the frequency multiplier signal to be retained and improving the signal-to-noise ratio. A signal processing and imaging module connected to the noise reduction module, which acquires and processes electrical signals from the noise reduction module to generate a frequency-multiplied image of the sample under test. A system that includes this.

2. The system for detecting wafer defects using a nonlinear optical signal according to claim 1, wherein the photodetector assembly includes a plurality of photodetectors, and the output gain of the photodetectors is adjustable.

3. A system for detecting defects in a wafer using a nonlinear optical signal according to claim 1, wherein the frequency multiplication signal is a second harmonic generation signal (SHG).

4. A system for detecting defects in a wafer using a nonlinear optical signal according to claim 1, wherein the photodetector includes one or a combination of a photodiode (PD), an avalanche photodiode (APD), a charge-coupled device (CCD), and a photomultiplier tube (PMT).

5. A system for detecting defects in a wafer using a nonlinear optical signal as described in claim 1, wherein the beam splitter assembly includes a first beam splitter, the photodetector assembly includes a first photodetector and a second photodetector, the first beam splitter spectrally separates the emitted light into a first beam and a second beam, the first beam is received by the first photodetector, and the second beam is received by the second photodetector.

6. A system for detecting defects in a wafer using a nonlinear optical signal according to claim 5, wherein the spectral ratio of the first beam splitter is 50:50 and the tolerance is ±5%.

7. A system for detecting defects in a wafer using a nonlinear optical signal as described in claim 1, wherein the beam splitter assembly includes a first beam splitter and a second beam splitter, the photodetector assembly includes a first photodetector, a second photodetector and a third photodetector, the first beam splitter spectrally separates the emitted light into a first beam and a second beam, the spectral ratio of the second beam is greater than that of the first beam, the first beam is received by the first photodetector, and the second beam splitter is provided in the optical path of the second beam and further spectrally separates the second beam into a third beam and a fourth beam, the third beam is received by the second photodetector and the fourth beam is received by the third photodetector.

8. A system for detecting defects in a wafer using a nonlinear optical signal according to claim 7, wherein the spectral ratio of the first beam splitter is 30:70 and the tolerance is ±5%, and the spectral ratio of the second beam splitter is 50:50 and the tolerance is ±5%.

9. A system for detecting defects in a wafer using a nonlinear optical signal as described in claim 1, wherein the beam splitter assembly includes a first beam splitter, a second beam splitter, and a third beam splitter, the photodetector assembly includes a first photodetector, a second photodetector, a third photodetector, and a fourth photodetector, the first beam splitter spectrally separates the emitted light into a first beam and a second beam, the first beam is received by the first photodetector, the spectral ratio of the second beam is greater than that of the first beam, the second beam splitter is provided in the optical path of the second beam and further spectrally separates the second beam into a third beam and a fourth beam, the spectral ratio of the fourth beam is greater than that of the third beam, the third beam is received by the second photodetector, and the third beam splitter is provided in the optical path of the fourth beam and further spectrally separates the fourth beam into a fifth beam and a sixth beam, the fifth beam is received by the third photodetector, and the sixth beam is received by the fourth photodetector.

10. A system for detecting defects in a wafer using a nonlinear optical signal as described in claim 9, wherein the spectral ratio of the first beam splitter is 20:80 and the tolerance is ±5%, the spectral ratio of the second beam splitter is 30:70 and the tolerance is ±5%, and the spectral ratio of the third beam splitter is 50:50 and the tolerance is ±5%.

11. A system for detecting wafer defects using a nonlinear optical signal according to claim 1, wherein a lens, a light-reducing filter, and a polarizing element are selectively or in combination arranged in the optical path before the photodetector.

12. A method for detecting defects in a wafer using a nonlinear optical signal, A noise reduction module consisting of a beam splitter assembly and a photodetector assembly is used to acquire emitted light from the same position and time in a specified area of ​​the sample under test, The noise reduction module reduces the homogeneous signal in the specified region and the background noise randomly generated during the photoelectric conversion process, thereby reducing the overall background noise, clarifying the frequency-multiplied signal, and improving the signal-to-noise ratio. A method that includes this.

Images