A semiconductor defect detection device and a detection method

Abstract

This application discloses a semiconductor defect detection device and method, relating to the field of semiconductor defect detection technology. When detecting defects in a semiconductor sample, a first light source component emits a pulsed laser with photon energy greater than the bandgap of the semiconductor sample, exciting a first photoluminescence signal. Then, while maintaining the first light source component's pulsed laser emission, a second light source component emits a continuous laser with photon energy equal to the energy interval between the energy level and valence band of the target deep-level defect within the semiconductor sample, exciting a second photoluminescence signal. Based on these two photoluminescence signals, defect information of the semiconductor sample can be obtained. A synchronous optical path calibration module is used to detect the focused spot positions of the pulsed laser and the continuous laser on the surface of the semiconductor sample in real time, ensuring that the offset between the two focused spots is less than a preset threshold. This effectively reduces detection errors caused by spot offset and improves detection accuracy.

Description

Technical Field

[0001] This application relates to the field of semiconductor defect detection technology, and in particular to a semiconductor defect detection device and detection method. Background Technology

[0002] Time-resolved photoluminescence spectroscopy is a core technique for detecting carrier lifetime in semiconductor materials. Carrier lifetime is closely related to semiconductor defects (especially deep-level defects). Deep-level defects act as carrier recombination centers, accelerating carrier annihilation and leading to a shortened carrier lifetime.

[0003] To detect deep-level defects in semiconductor samples, a dual-laser excitation method can be used to generate photoluminescence. One laser has a photon energy greater than the bandgap of the semiconductor sample and is used to excite the semiconductor sample to generate photoluminescence. The photon energy of the other laser corresponds to the energy level position of a certain deep-level defect in the semiconductor sample and is used to saturate the electronic state of the deep-level defect. By comparing the photoluminescence spectral intensity when the electronic state of the deep-level defect is saturated with that when it is unsaturated, it is possible to detect or determine whether the deep-level defect is an effective carrier recombination center.

[0004] However, existing detection devices, which use dual lasers to detect deep-level defects in semiconductor samples, are prone to large detection errors. Summary of the Invention

[0005] To address the aforementioned technical problems, this application provides a semiconductor defect detection device and method to reduce the error in detecting deep-level defects in semiconductor samples using dual lasers and improve detection accuracy.

[0006] To achieve the above objectives, this application provides the following technical solution:

[0007] A semiconductor defect detection device for detecting defects in semiconductor samples, the detection device comprising:

[0008] A light-emitting module includes a first light source assembly, a second light source assembly, and a beam combiner assembly. The first light source assembly emits pulsed laser light, the photon energy of which is greater than the bandgap of the semiconductor sample. The second light source assembly emits continuous laser light, the photon energy of which is equal to the energy interval between the energy level and the valence band of a target deep-level defect within the semiconductor sample. The beam combiner assembly focuses the pulsed laser light onto the surface of the semiconductor sample to generate a first photoluminescence signal, and combines the pulsed laser light and the continuous laser light beam and focuses them onto the surface of the semiconductor sample to generate a second photoluminescence signal.

[0009] The optical path synchronization calibration module is used to detect the focused spot positions of the pulsed laser and the continuous laser on the surface of the semiconductor sample in real time. When the offset between the two is not less than a preset threshold, the beam combining component is adjusted so that the offset is less than the preset threshold.

[0010] A light separation and collection module is used to transmit the first photoluminescence signal and the second photoluminescence signal, and to filter out the pulsed laser and the continuous laser.

[0011] The detection module is used to obtain defect information of the semiconductor sample based on the first photoluminescence signal and the second photoluminescence signal.

[0012] Optionally, the beam combining assembly includes a laser beam combiner, a beam splitter, and a first lens;

[0013] The laser beam combiner is used to combine the pulsed laser and the continuous laser into the same optical path;

[0014] The beam splitter is used to project the pulsed laser and the continuous laser, which share the same optical path, onto the first lens;

[0015] The first lens is used to focus the pulsed laser onto the surface of the semiconductor sample to generate the first photoluminescence signal, and to focus the combined pulsed laser and the continuous laser onto the surface of the semiconductor sample to generate the second photoluminescence signal.

[0016] The first lens and the beam splitter are also used to project the first photoluminescence signal and the second photoluminescence signal onto the light separation and collection module.

[0017] Optionally, the optical path synchronization calibration module includes a position sensor, a piezoelectric ceramic adjustment stage, and a controller, wherein the piezoelectric ceramic adjustment stage is fixedly connected to the first lens;

[0018] The position sensor is used to detect in real time the position of the focused spot of the pulsed laser and the continuous laser on the surface of the semiconductor sample;

[0019] The controller is used to determine whether the offset between the pulsed laser and the continuous laser on the surface of the semiconductor sample is not less than a preset threshold, based on the focused spot positions of the two lasers. When the offset is not less than the preset threshold, the controller controls the piezoelectric ceramic adjustment stage to adjust the spatial position of the first lens so that the offset is less than the preset threshold.

[0020] Optionally, the light separation and collection module includes a tunable filter assembly, a second lens, and a monochromator arranged along the optical path transmission direction;

[0021] The tunable filter component is used to match the wavelengths of the pulsed laser and the continuous laser in real time, filter out the pulsed laser and the continuous laser, and allow the first photoluminescence signal and the second photoluminescence signal to pass through.

[0022] The second lens is used to focus the first photoluminescence signal and the second photoluminescence signal into the monochromator;

[0023] The monochromator is used to filter out the first photoluminescence signal and the second photoluminescence signal, and transmit them to the detection module.

[0024] Optionally, the second light source assembly includes a continuous laser source and a continuous laser photon energy dynamic scanning unit;

[0025] The continuous laser photon energy dynamic scanning unit is electrically connected to the continuous laser source and is used to control the continuous laser source to emit the continuous laser. The photon energy of the continuous laser is matched step by step to match the energy interval between the energy level and the valence band of different deep energy level defects in the semiconductor sample, so that the electronic states of different deep energy level defects in the semiconductor sample are saturated respectively. At the same time, the pulsed laser excites the second photoluminescence signal corresponding to the continuous laser with different photon energies.

[0026] Optionally, the detection module includes a photodetector, a time-correlated single-photon detector, a fluorescence lifetime multi-channel parallel processing unit, and a defect impact quantification analysis unit;

[0027] The photodetector is used to convert the first photoluminescence signal and each of the second photoluminescence signals into electrical pulse signals;

[0028] The time-correlated single-photon detector is used to perform time stamping and statistical analysis on the electrical pulse signal converted by the photodetector based on the electrical pulse signal synchronized with the pulsed laser, so as to obtain the intensity decay curves of the first photoluminescence signal and each of the second photoluminescence signals over time.

[0029] The fluorescence lifetime multi-channel parallel processing unit is used to simultaneously acquire the curves of different wavelength bands and extract the fluorescence lifetime corresponding to the first photoluminescence signal and each of the second photoluminescence signals.

[0030] The defect impact quantification analysis unit is used to calculate the impact coefficient of each deep-level defect in the semiconductor sample on the fluorescence lifetime based on the difference between the fluorescence lifetime corresponding to the extracted first photoluminescence signal and the fluorescence lifetime corresponding to each of the second photoluminescence signals.

[0031] Optionally, the first light source component includes a pulsed laser source and a pulsed laser source control power supply;

[0032] The pulsed laser source control power supply is electrically connected to the pulsed laser source and is used to control the frequency and pulse width of the pulsed laser emitted by the pulsed laser source. It is also used to transmit an electrical pulse signal synchronized with the pulsed laser to the time-correlated single-photon detector.

[0033] A method for detecting semiconductor defects, applied to the detection apparatus described in any one of the preceding claims, the detection method comprising:

[0034] The first light source component in the light emission module emits a pulsed laser, the photon energy of which is greater than the bandgap of the semiconductor sample. The pulsed laser is focused on the surface of the semiconductor sample by the beam combining component in the light emission module, thereby exciting a first photoluminescence signal. The first photoluminescence signal is then transmitted to the detection module through the light separation and collection module.

[0035] While the first light source component emits the pulsed laser, the second light source component in the light emission module emits a continuous laser. The photon energy of the continuous laser is equal to the energy interval between the energy level and the valence band of the target deep energy level defect in the semiconductor sample. The pulsed laser and the continuous laser are combined and focused on the surface of the semiconductor sample by the beam combining component to excite a second photoluminescence signal. The second photoluminescence signal is then transmitted to the detection module through the light separation and collection module.

[0036] The optical path synchronization calibration module is used to detect the focused spot positions of the pulsed laser and the continuous laser on the surface of the semiconductor sample in real time. When the offset between the two is not less than a preset threshold, the beam combining component is adjusted so that the offset is less than the preset threshold.

[0037] The detection module uses the first photoluminescence signal and the second photoluminescence signal to obtain the defect information of the semiconductor sample.

[0038] Optionally, the photon energy of the continuous laser is matched stepwise to the energy interval between the energy level and the valence band of different deep energy level defects in the semiconductor sample, so that the electronic states of different deep energy level defects in the semiconductor sample are saturated respectively. At the same time, the pulsed laser excites the second photoluminescence signal corresponding to the continuous laser with different photon energies.

[0039] The detection module includes a photodetector, a time-correlated single-photon detector, a fluorescence lifetime multi-channel parallel processing unit, and a defect impact quantification analysis unit. Based on the first photoluminescence signal and the second photoluminescence signal, the detection module obtains defect information of the semiconductor sample, including:

[0040] The photodetector is used to convert the first photoluminescence signal and each of the second photoluminescence signals into electrical pulse signals;

[0041] Using the time-correlated single-photon detector based on the electrical pulse signal synchronized with the pulsed laser, the electrical pulse signal converted by the photodetector is time-stamped and statistically analyzed to obtain the intensity decay curves of the first photoluminescence signal and each of the second photoluminescence signals over time;

[0042] The fluorescence lifetime multi-channel parallel processing unit is used to simultaneously acquire the curves of different wavelength bands, and the fluorescence lifetimes corresponding to the first photoluminescence signal and each second photoluminescence signal are extracted.

[0043] The defect impact quantification analysis unit calculates the impact coefficient of each deep-level defect in the semiconductor sample on the fluorescence lifetime based on the difference between the fluorescence lifetime corresponding to the extracted first photoluminescence signal and the fluorescence lifetime corresponding to each of the second photoluminescence signals.

[0044] Optionally, the light separation and collection module includes a tunable filter component; the detection method further includes:

[0045] The tunable filter component is used to match the wavelengths of the pulsed laser and the continuous laser in real time, filtering out the pulsed laser and the continuous laser, and allowing the first photoluminescence signal and the second photoluminescence signal to pass through.

[0046] Compared with existing technologies, the above technical solution has the following advantages:

[0047] The semiconductor defect detection apparatus and method provided in this application are used to detect defects in semiconductor samples. The detection apparatus includes a light emission module, an optical path synchronization calibration module, a light separation and collection module, and a detection module. The light emission module includes a first light source assembly, a second light source assembly, and a beam combiner assembly. When detecting defects in a semiconductor sample, the first light source assembly first emits a pulsed laser with photon energy greater than the bandgap of the semiconductor sample. The pulsed laser is focused onto the surface of the semiconductor sample by the beam combiner assembly, exciting a first photoluminescence signal. The first photoluminescence signal is then transmitted to the detection module through the light separation and collection module. Then, while the first light source assembly continues to emit the pulsed laser, the second light source assembly emits a continuous laser. The photon energy of the continuous laser is equal to the energy interval between the energy level and valence band of the target deep-level defect within the semiconductor sample. The pulsed laser and the continuous laser are combined and focused onto the surface of the semiconductor sample by the beam combiner assembly. The continuous laser causes the semiconductor sample to... The electronic states of the target deep-level defects in the semiconductor sample are saturated, and a second photoluminescence signal is excited by a pulsed laser. This second photoluminescence signal is then transmitted to the detection module through the light separation and collection module. The detection module can then compare the intensities of the first and second photoluminescence signals to obtain defect information from the semiconductor sample. During the focusing of the pulsed and continuous laser beams onto the surface of the semiconductor sample, the optical path synchronization calibration module monitors the focused spot positions of the pulsed and continuous laser beams on the semiconductor sample surface in real time. When the offset between the two focused spots is not less than a preset threshold, the beam combining component is adjusted so that the offset between the two focused spots is less than the preset threshold. This ensures precise control of the coaxiality of the focusing of the pulsed and continuous laser beams on the semiconductor sample surface, effectively reducing the fluorescence lifetime detection error caused by insufficient saturation of defect electronic states due to the offset of the focused spots on the semiconductor sample surface, thus improving detection accuracy. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A schematic diagram of the structure of a semiconductor defect detection device provided in an embodiment of this application;

[0050] Figure 2 A schematic diagram illustrating the principle of generating the first photoluminescence signal by exciting a semiconductor sample with a single pulsed laser.

[0051] Figure 3A schematic diagram illustrating the principle of generating a second photoluminescence signal by simultaneously irradiating a semiconductor sample with pulsed laser and continuous laser.

[0052] Figure 4 A schematic diagram comparing the first fluorescence lifetime corresponding to the first photoluminescence signal generated by a single pulsed laser excitation of a semiconductor sample with the second fluorescence lifetime corresponding to the second photoluminescence signal generated by simultaneous irradiation of the semiconductor sample by pulsed laser and continuous laser.

[0053] Figure 5 The fluorescence lifetime comparison diagram obtained by the semiconductor defect detection device provided in the embodiments of this application after detecting three deep-level defects of 0.6eV, 0.8eV and 1.0eV in a copper indium gallium selenide semiconductor thin film.

[0054] Explanation of reference numerals in the attached figures:

[0055] The system includes: a light emission module 100; a first light source assembly 110; a second light source assembly 120; a beam combiner assembly 130; an optical path synchronization calibration module 200; a light separation and collection module 300; a detection module 400; a pulsed laser source 1; a pulsed laser source control power supply 2; a continuous laser source 3; a continuous laser photon energy dynamic scanning unit 4; a first reflecting mirror 10; a laser beam combiner 11; a beam splitter 9; a first lens 6; a semiconductor sample 5; a position sensor 8; a piezoelectric ceramic adjustment stage 7; a controller 20; a tunable filter assembly 12; a second reflecting mirror 13; a second lens 14; a monochromator 15; a photodetector 16; a time-correlated single-photon detector 17; a fluorescence lifetime multi-channel parallel processing unit 18; a defect influence quantification analysis unit 19; a first fluorescence lifetime t1; and a second fluorescence lifetime t2. Detailed Implementation

[0056] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0057] The terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms are interchangeable where appropriate; this is merely a way of distinguishing objects with the same attributes in the embodiments of this application. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, so that a process, method, system, product, or apparatus that comprises a series of elements is not necessarily limited to those elements, but may include other elements not explicitly listed or inherent to those processes, methods, products, or apparatuses.

[0058] Given the problem of significant detection errors in existing detection devices described in the background art when using dual lasers to detect deep-level defects in semiconductor samples, the inventors discovered that, in principle, the laser used to saturate the electronic states of deep-level defects and the laser used to excite photoluminescence should be focused at the same position on the surface of the semiconductor sample. However, in reality, the focused spots of the two lasers on the surface of the semiconductor sample tend to shift, which can easily lead to insufficient saturation of the electronic states of the corresponding deep-level defects on the surface of the semiconductor sample, thus introducing fluorescence lifetime detection errors. Therefore, reducing the shift in the focused spots of the two lasers on the surface of the semiconductor sample is an important means to reduce detection errors and improve detection accuracy.

[0059] Based on this, embodiments of this application propose a semiconductor defect detection device for detecting defects in semiconductor samples. Figure 1 This diagram illustrates the structure of a semiconductor defect detection device according to an embodiment of this application, where solid lines represent the transmission relationship of optical or electrical signals in each component, such as... Figure 1 As shown, the detection device includes a light emission module 100, an optical path synchronization calibration module 200, a light separation and collection module 300, and a detection module 400.

[0060] The light emitting module 100 includes a first light source assembly 110, a second light source assembly 120, and a beam combiner assembly 130. The first light source assembly 110 is used to emit pulsed laser light, the photon energy of which is greater than the bandgap of the semiconductor sample 5. The second light source assembly 120 is used to emit continuous laser light, the photon energy of which is equal to the energy interval between the energy level and the valence band of the deep energy level defect of the target in the semiconductor sample 5. The beam combiner assembly 130 is used to focus the pulsed laser light onto the surface of the semiconductor sample 5 to generate a first photoluminescence signal, and to combine the pulsed laser light and the continuous laser light onto the surface of the semiconductor sample 5 to generate a second photoluminescence signal.

[0061] When performing defect detection on a semiconductor sample, a pulsed laser with photon energy greater than the bandgap of the semiconductor sample 5 is first emitted by the first light source component 110. The pulsed laser is focused on the surface of the semiconductor sample 5 by the beam combining component 130, thereby exciting a first photoluminescence signal. Figure 2 A schematic diagram illustrating the principle of generating the first photoluminescence signal by exciting a semiconductor sample with a single pulsed laser is shown, as follows: Figure 2 As shown, E V For the price band, E C For the conductor, E T As a defect energy level, it can be understood that when a pulsed laser with photon energy greater than the material's bandgap irradiates a semiconductor sample, the valence band (E)... V Electrons in the electron ion absorb photon energy and jump to the conduction band (E0). C In the conduction band, free electrons are generated, leaving holes in the valence band, forming electron-hole pairs. Some of these electron-hole pairs recombine directly to emit light, producing the first photoluminescence signal. However, some charge carriers are also absorbed by defect levels (E0). T The capture process consumes charge carriers, causing the intensity of the photoluminescence signal to decay faster.

[0062] Then, while the first light source component 110 emits pulsed laser, the second light source component 120 emits continuous laser with photon energy equal to the energy interval between the energy level and the valence band of the deep energy level defect of the target in the semiconductor sample 5. The pulsed laser and the continuous laser are combined and focused on the surface of the semiconductor sample 5 by the beam combining component 130 to excite the second photoluminescence signal. Figure 3 A schematic diagram illustrating the principle of generating a second photoluminescence signal by simultaneously irradiating a semiconductor sample with pulsed and continuous laser light is shown, as follows: Figure 3 As shown, it can be understood that since the photon energy of a continuous laser is exactly equal to the valence band energy (E), V The energy level of the target deep-level defect (i.e., the defect energy level E) T The energy interval of ) so it was originally in the valence band (E) V Electrons from the target will be preferentially pumped by continuous laser light to the energy level of the deep-level defect (i.e., the defect energy level E). T This causes the target deep-level defect states to be filled with electrons (i.e., electronic state saturation), preventing carrier capture. Simultaneously, the pulsed laser empties the valence band (E... V Electrons in the electrons are excited to the conduction band (E) C The electron-hole pairs are generated, and the direct recombination of the electron-hole pairs produces light, generating a second photoluminescence signal. Because there is no target deep-level defect to capture and consume the charge carriers, the intensity of the photoluminescence signal decays more slowly.

[0063] The light separation and collection module 300 is used to transmit the first photoluminescence signal and the second photoluminescence signal to the detection module 400 through the first photoluminescence signal and the second photoluminescence signal, and to filter out pulsed laser and continuous laser to avoid interference of pulsed laser and continuous laser on the photoluminescence signal.

[0064] The detection module 400 can obtain defect information of the semiconductor sample based on the first photoluminescence signal and the second photoluminescence signal. For details, refer to... Figure 4 As shown, based on the decay of the intensity of the first photoluminescence signal over time, the corresponding first fluorescence lifetime t1 can be calculated. This reflects the recombination of electrons from the conduction band to the valence band with holes and the overall decay of carriers due to defect trapping. Based on the decay of the intensity of the second photoluminescence signal over time, the corresponding second fluorescence lifetime t2 can be calculated. If the target deep-level defect is an effective carrier recombination center, then t2 > t1. Therefore, Δt = t2 - t1 can be calculated. This difference represents the emission time lost due to carrier trapping by the target deep-level defect. If Δt > 0, it proves that the continuous laser does indeed saturate the electronic states of the target deep-level defect in the semiconductor sample, resulting in a longer carrier lifetime excited by the pulsed laser. This verifies that the target deep-level defect is an effective carrier recombination center. Furthermore, the larger Δt is, the stronger the ability of the target deep-level defect to trap carriers and the stronger its effect on the decay of fluorescence lifetime. Thus, it is possible to determine whether the target deep-level defect is an effective carrier recombination center and the degree of its influence.

[0065] The detection device provided in this application embodiment uses an optical path synchronization calibration module 200 to detect the position of the focused spot of the pulsed laser and the continuous laser on the surface of the semiconductor sample 5 in real time during the process of focusing the pulsed laser and the continuous laser on the surface of the semiconductor sample 5. When the offset of the two focused spots is not less than a preset threshold, the beam combining component 130 is adjusted so that the offset of the two focused spots is less than the preset threshold. Optionally, the preset threshold can be 0.1 μm. In this way, the coaxiality of the focusing of the pulsed laser and the continuous laser on the surface of the semiconductor sample can be accurately controlled, effectively reducing the fluorescence lifetime detection error caused by insufficient saturation of defect electronic states due to the offset of the focused spot of the pulsed laser and the continuous laser on the surface of the semiconductor sample, and improving the detection accuracy.

[0066] Experiments show that the optical path synchronization calibration module 200 can reduce the focusing coaxiality error of pulsed laser and continuous laser on the surface of semiconductor sample 5 to less than or equal to 0.1 μm, and reduce the fluorescence lifetime detection error to less than or equal to 5%, while the detection error of existing detection devices is usually greater than 15%.

[0067] Optional, such as Figure 1As shown, the first light source assembly 110 includes a pulsed laser source 1, which emits pulsed laser light with photon energy greater than the bandgap of the semiconductor sample. Additionally, the first light source assembly 110 may also include a pulsed laser source control power supply 2, which is electrically connected to the pulsed laser source 1. The pulsed laser source control power supply 2 is used to control the frequency and pulse width of the pulsed laser light emitted by the pulsed laser source 1. For example, the frequency of the emitted pulsed laser light can be controlled within the range of 100Hz-1MHz, and the pulse width within the range of 1ps-10s. Compared to a standalone pulsed laser source 1, adding a pulsed laser source control power supply 2 allows for precise control of the frequency and pulse width of the pulsed laser light, which is beneficial for emitting more stable pulsed laser light and improving the stability and detection accuracy of the detection device.

[0068] Optional, such as Figure 1 As shown, the second light source assembly 120 includes a continuous laser source 3, which emits a continuous laser with photon energy equal to the energy interval between the energy level and the valence band of a target deep energy level defect within the semiconductor sample. Specifically, the wavelength range of both pulsed laser and continuous laser can be 300nm-2000nm, and the full width at half maximum (FWHM) of the spectrum can be 0.01nm-10nm.

[0069] Understandably, if only a single deep-level defect in a semiconductor sample needs to be detected, the photon energy of the continuous laser can be directly set to the energy interval between the energy level and the valence band of the deep-level defect to be detected. However, if multiple deep-level defects in a semiconductor sample need to be detected, the photon energy of the continuous laser needs to be adjusted to correspond to the energy interval between the energy level and the valence band of the deep-level defects to be detected.

[0070] Alternatively, the photon energy of the continuous laser can be manually adjusted and the detection process repeated, but this is cumbersome and time-consuming.

[0071] Another option, such as Figure 1As shown, the second light source assembly 120 can be configured to include not only the continuous laser source 3, but also a continuous laser photon energy dynamic scanning unit 4. The continuous laser photon energy dynamic scanning unit 4 is electrically connected to the continuous laser source 3 and is used to control the continuous laser source 3 to emit photon energy in a step-by-step manner to match the energy interval between the energy level and valence band of different deep energy level defects in the semiconductor sample. For example, the continuous laser photon energy dynamic scanning unit 4 can step-by-step adjust the photon energy of the continuous laser in the wavelength range of 300nm-2000nm, and the step interval can be less than or equal to 0.01eV, thereby sequentially saturating the electronic states of different deep energy level defects in the semiconductor sample. At the same time, the pulsed laser excites each second photoluminescence signal corresponding to the continuous laser with different photon energies, thereby realizing the continuous detection of multi-deep energy level defects in the semiconductor sample and greatly improving the detection efficiency.

[0072] It is also understandable that, in order to combine pulsed laser and continuous laser beams into a single optical path, options such as... Figure 1 As shown, the second light source assembly 120 may further include a first reflector 10, which is used to project the continuous laser emitted by the continuous laser source 3 onto the laser beam combiner 11 to save optical path space and reduce the size of the detection device. Of course, alternatively, a reflector may also be provided in the first light source assembly 110 to project the pulsed laser emitted by the pulsed laser source 1 onto the laser beam combiner 11, depending on the specific circumstances.

[0073] Optional, such as Figure 1 As shown, the beam combining assembly 130 includes a laser beam combiner 11, a beam splitter 9, and a first lens 6; wherein, the laser beam combiner 11 is used to combine pulsed laser and continuous laser into the same optical path; the beam splitter 9 is used to project the pulsed laser and continuous laser of the same optical path onto the first lens 6; the first lens 6 is used to focus the pulsed laser on the surface of the semiconductor sample 5 to excite a first photoluminescence signal, and to focus the combined pulsed laser and continuous laser on the surface of the semiconductor sample 5 to excite a second photoluminescence signal.

[0074] It is understandable that when the first light source assembly 110 is turned on alone, the pulsed laser emitted by the first light source assembly 110 passes through the laser beam combiner 11, the beam splitter 9, and the first lens 6 in sequence and is focused on the surface of the semiconductor sample 5. When the first light source assembly 110 and the second light source assembly 120 are turned on simultaneously, the pulsed laser emitted by the first light source assembly 110 and the continuous laser emitted by the second light source assembly 120 are combined by the laser beam combiner 11, and then pass through the beam splitter 9 and the first lens 6 in sequence and are focused on the surface of the semiconductor sample 5. The first lens 6 is a key component that focuses the combined pulsed laser and the continuous laser at the same position on the surface of the semiconductor sample 5.

[0075] And, as Figure 1 As shown, the first lens 6 and the beam splitter 9 can also be used to project the first photoluminescence signal and the second photoluminescence signal excited from the semiconductor sample 5 to the light separation and collection module 300.

[0076] Based on this, optional, such as Figure 1 As shown, the optical path synchronization calibration module 200 includes a position sensor 8, a piezoelectric ceramic adjustment stage 7, and a controller 20. The piezoelectric ceramic adjustment stage 7 is fixedly connected to the first lens 6. The position sensor 8 is used to detect the position of the focused spot of the pulsed laser and the continuous laser on the surface of the semiconductor sample 5 in real time. For example, the position sensor 8 can be a laser displacement sensor, which can detect the position of the focused spot directly facing the surface of the semiconductor sample 5, and its detection accuracy can be less than or equal to 0.01 μm. The controller 20 is used to determine whether the offset of the two focused spots is not less than a preset threshold (e.g., 0.1 μm) based on the position of the focused spot of the pulsed laser and the continuous laser on the surface of the semiconductor sample 5. When the offset of the two focused spots is not less than the preset threshold, the controller controls the piezoelectric ceramic adjustment stage 7 to adjust the spatial position of the first lens 6 so that the offset of the two focused spots is less than the preset threshold, for example, to ensure that the coaxiality error of the two focused spots is less than or equal to 0.1 μm.

[0077] In practical applications, before detecting defects in semiconductor samples, the optical path synchronization calibration module can be activated first. The optical path synchronization calibration module 200 is used to adjust the beam combining component 130 (specifically, the spatial position of the first lens 6), so that the focusing coaxiality requirement can be met directly after the subsequent pulsed laser and continuous laser are turned on.

[0078] Optional, such as Figure 1 As shown, the light separation and collection module 300 includes a tunable filter component 12, a second lens 14, and a monochromator 15 arranged along the optical path transmission direction; wherein, the tunable filter component 12 is used to match the wavelengths of pulsed laser and continuous laser in real time, filter out pulsed laser and continuous laser, and transmit the first photoluminescence signal and the second photoluminescence signal; the second lens 14 is used to focus the first photoluminescence signal and the second photoluminescence signal into the monochromator 15; the monochromator 15 is used to filter out the first photoluminescence signal and the second photoluminescence signal and transmit them to the detection module 400.

[0079] Compared to existing detection devices that use fixed filters that can only filter out pulsed and continuous lasers of fixed wavelengths and cannot match laser wavelength changes in real time, potentially leading to incomplete excitation light filtering and affecting the purity of the photoluminescence signal, and if multiple deep-level defects in a semiconductor sample need to be detected, it is not only necessary to adjust the photon energy of the continuous laser and repeat the detection process, but also to replace the filters one by one and adjust the laser parameters; the detection device provided in this application uses a tunable filter assembly 12 composed of two independent electrically controlled tunable filters that match the pulsed and continuous laser wavelengths in real time to filter out both pulsed and continuous lasers and allow the photoluminescence signal to pass through. The tunable filter assembly 12 has an automatic calibration function and can match the excitation light wavelength in real time, thus eliminating the need for manual filter replacement. At the same time, it can ensure the purity of the transmitted photoluminescence signal (transmittance greater than or equal to 95%) and can also be used to detect multiple deep-level defects.

[0080] In practical applications, before detecting defects in semiconductor samples, the automatic calibration function of the tunable filter component 12 can be activated to match the initial excitation wavelength.

[0081] Furthermore, the tunable filter assembly 12 is electrically connected to a computer. When the wavelengths of the pulsed laser and the continuous laser change, the computer controls the tunable filter assembly 12 to perform an automatic calibration function, matching the changed pulsed laser wavelength and the continuous laser wavelength. The computer can provide... Figure 1 The controller 20 in the middle.

[0082] Specifically, the wavelength adjustment range of the tunable filter component 12 is not less than the wavelength range of pulsed laser and continuous laser. For example, the wavelength adjustment range of the tunable filter component 12 can be 300nm-2000nm, its automatic calibration cycle can be less than or equal to 10ms, and its adjustment accuracy can be less than or equal to 0.001nm. Thus, the filtering range is adjusted in real time according to the wavelength changes of pulsed laser and continuous laser to ensure that pulsed laser and continuous laser are filtered out in real time and to ensure the purity of the transmitted photoluminescence signal.

[0083] The monochromator 15 has a grating coverage wavelength range of 200nm-3000nm, and can scan the photoluminescence signal of semiconductor samples across the entire wavelength range, with a scanning speed of 100nm / s.

[0084] Optional, such as Figure 1 As shown, the light separation and collection module 300 may also include a second reflector 13, which is used to reflect the first photoluminescence signal and the second photoluminescence signal after passing through the tunable filter component 12 to the second lens 14, so that the reflected first photoluminescence signal and the second photoluminescence signal are then focused by the second lens 14 into the monochromator 15, thereby saving optical path space and reducing the size of the detection device.

[0085] Based on the second light source assembly 120, which includes a continuous laser source 3 and a continuous laser photon energy dynamic scanning unit 4, and the step-by-step detection of multi-deep energy level defects in semiconductor samples, optional features include... Figure 1 As shown, the detection module 400 includes a photodetector 16, a time-correlated single-photon detector 17, a fluorescence lifetime multi-channel parallel processing unit 18, and a defect impact quantification analysis unit 19.

[0086] The photodetector 16 is used to convert the first photoluminescence signal and the second photoluminescence signal into an electrical pulse signal. The spectral response range of the photodetector 16 is perfectly matched with the photoluminescence spectrum range of the semiconductor sample, ensuring that the photoluminescence signal is converted into an electrical pulse signal without attenuation.

[0087] The time-correlated single-photon detector 17 is used to perform time stamping and statistical analysis on the electrical pulse signal converted by the photodetector 16 based on the electrical pulse signal synchronized with the pulsed laser, to obtain the intensity decay curves of the first photoluminescence signal and each second photoluminescence signal over time (e.g., Figure 5 (As shown); the detection range of the time-correlated single-photon detector 17 is 25 ps to 10 s, and the resolution is 1 ps. It should be noted that the pulsed laser source control power supply 2 in the first light source assembly 110 can also synchronously emit electrical pulse signals synchronized with the pulsed laser to the time-correlated single-photon detector 17.

[0088] The fluorescence lifetime multi-channel parallel processing unit 18 is used to synchronously acquire the curves of different wavelength bands and extract the fluorescence lifetimes corresponding to the first photoluminescence signal and the second photoluminescence signal. The fluorescence lifetime multi-channel parallel processing unit 18 can have 8 channels and a sampling frequency greater than or equal to 1 MHz.

[0089] The defect impact quantification analysis unit 19 is used to calculate the influence coefficient of each deep energy level defect in semiconductor sample 5 on fluorescence lifetime based on the difference in fluorescence lifetime corresponding to the extracted first photoluminescence signal and the second photoluminescence signal.

[0090] Based on the foregoing analysis, when detecting multi-level defects in a semiconductor sample, a pulsed laser is first focused onto the surface of the semiconductor sample 5 to excite a first photoluminescence signal. The first fluorescence lifetime t1 can be calculated based on the intensity decay curve of the first photoluminescence signal over time. Then, the pulsed laser and a continuous laser beam are combined and focused onto the surface of the semiconductor sample 5. The photon energy of the continuous laser is progressively matched to the energy intervals between the energy levels and valence bands of different deep-level defects within the semiconductor sample, thus exciting second photoluminescence signals corresponding to different photon energies. The corresponding fluorescence lifetime t1 can be calculated based on the intensity decay curve of each second photoluminescence signal over time. The second fluorescence lifetime t2 is calculated; then, the difference Δt between the second fluorescence lifetime t2 and the first fluorescence lifetime t1 corresponding to continuous lasers with different photon energies is calculated. If Δt > 0, it indicates that the corresponding deep-level defect is an effective carrier recombination center. Furthermore, the larger Δt is, the stronger the carrier capture ability of the corresponding deep-level defect and the larger the influence coefficient. Thus, quantitative analysis of multi-deep-level defects in semiconductor samples can be realized, and the influence coefficient of different deep-level defects can be output. This provides a quantitative basis for defect control of semiconductor materials and makes up for the shortcomings of existing detection devices, which can only realize qualitative analysis of defects and cannot quantify the degree of influence of defects on fluorescence lifetime, making it difficult to meet the needs of precise optimization of semiconductor materials.

[0091] Furthermore, the use of the continuous laser photon energy dynamic scanning unit 4 enables dynamic scanning of the continuous laser, and combined with the multi-channel parallel detection of the fluorescence lifetime multi-channel parallel processing unit 18, the efficiency of continuous detection of multi-deep level defects can be further improved (≥80%). Experiments show that if five deep level defects need to be detected, the detection device provided in this application embodiment only requires 250ms, while existing detection devices require more than 1250ms.

[0092] In practical applications, the defect impact quantification analysis unit 19 can ultimately transmit the defect analysis results to a computer, which can then perform... Figure 1 The controller 20 in the middle.

[0093] For example, the implementation process of the detection device of this application is described in detail, taking the detection of three deep-level defects of 0.6eV, 0.8eV and 1.0eV in copper indium gallium selenide (CIGS) semiconductor thin films as an example.

[0094] First, set the parameters of the detection device.

[0095] In the optical emission module 100, since the bandgap of copper indium gallium selenide (CIGS) is 1.01eV-1.68eV, the wavelength of the pulsed laser can be set to 635nm (corresponding to a photon energy of 1.95eV), the frequency can be 1kHz, and the pulse width can be 10ns. At the same time, the initial wavelength of the continuous laser can be set to 1550nm (corresponding to a photon energy of 0.8eV), the dynamic scanning range of the continuous laser can be 1240nm-2067nm (corresponding to a photon energy range of 0.6eV-1.0eV), and the step interval is 0.01eV. The step speed of the continuous laser photon energy dynamic scanning unit 4 can be 5000eV / s to ensure that the detection time for a single defect is less than or equal to 50ms.

[0096] In the optical path synchronization calibration module 200, the position sensor 8 has a measurement range of 0-100μm and a measurement accuracy of 0.01μm; the piezoelectric ceramic adjustment stage 7 has an adjustment range of ±50μm and a response time of 1ms.

[0097] In the light separation and collection module 300, the wavelength adjustment range of the tunable filter component 12 can be 300nm-2000nm, the adjustment accuracy can be 0.001nm, and the automatic calibration cycle can be 10ms; the grating wavelength range of the monochromator 15 can be 200nm-3000nm, and the scanning speed can be 100nm / s.

[0098] In the detection module 400, the detection range of the time-correlated single-photon detector can be 25ps-10s, and the resolution can be 1ps; the number of channels of the fluorescence lifetime multi-channel parallel processing unit can be 8, and the sampling frequency can be 1MHz.

[0099] Secondly, the detection steps for deep-level defects in copper indium gallium selenide (CIGS) semiconductor thin films are as follows:

[0100] Step 1: Optical path calibration;

[0101] The position sensor 8 and piezoelectric ceramic adjustment stage 7 in the optical path synchronization calibration module 200 are activated to detect the focused spots of the pulsed laser and the continuous laser, ensuring that the coaxiality error of the two focused spots is less than a preset threshold. Furthermore, the tunable filter assembly 12 in the light separation and collection module 300 is automatically calibrated, with its two internal filters matching the wavelength of the pulsed laser (635nm) and the initial wavelength of the continuous laser (1550nm), respectively.

[0102] Step 2: Single-pulse laser excitation;

[0103] The first light source component 110 is activated to emit a pulsed laser. The pulsed laser passes sequentially through the laser beam combiner 11, the beam splitter 9, and the first lens 6, and is then focused onto the surface of the indium gallium selenide (CIGS) semiconductor thin film, exciting the generation of a first photoluminescence signal. The first photoluminescence signal passes through the first lens 6, the beam splitter 9, the tunable filter component 12 (to filter out the pulsed laser), the second reflector 13, the second lens 14, and the monochromator (to filter the 900nm-1200nm CIGS photoluminescence band), and is then transmitted to the photodetector 16. Finally, the time-correlated single-photon detector 17, combined with the fluorescence lifetime multi-channel parallel processing unit 18, acquires the first fluorescence lifetime t1 = 12.5ns corresponding to the first photoluminescence signal.

[0104] Step 3: Dual laser excitation combined with dynamic scanning detection;

[0105] While the first light source component 110 is turned on to emit pulsed laser, the second light source component 120 is turned on to emit continuous laser, and the wavelength is adjusted in 0.01eV steps. The pulsed laser and the continuous laser pass through the laser beam combiner 11, the beam splitter 9, and the first lens 6 in sequence, and are then focused onto the surface of the indium gallium selenide (CIGS) semiconductor thin film to generate a second photoluminescence signal. The second photoluminescence signal passes through the first lens 6, the beam splitter 9, the tunable filter component 12 (to filter out the pulsed laser and the continuous laser), the second reflector 13, the second lens 14, and the monochromator (to filter the 900nm-1200nm CIGS photoluminescence band), and is then transmitted to the photodetector 16. Finally, the time-correlated single-photon detector 17, combined with the fluorescence lifetime multi-channel parallel processing unit 18, collects the second fluorescence lifetime t2 corresponding to the second photoluminescence signal.

[0106] Specifically, for a 0.6 eV deep-level defect in a copper indium gallium selenide (CIGS) semiconductor thin film, the corresponding continuous laser wavelength is 1240 nm, and the second fluorescence lifetime corresponding to the excited second photoluminescence signal is t2 = 15.0 ns; for a 0.8 eV deep-level defect in a CIGS semiconductor thin film, the corresponding continuous laser wavelength is 1550 nm, and the second fluorescence lifetime corresponding to the excited second photoluminescence signal is t2 = 15.6 ns; for a 1.0 eV deep-level defect in a CIGS semiconductor thin film, the corresponding continuous laser wavelength is 2067 nm, and the second fluorescence lifetime corresponding to the excited second photoluminescence signal is t2 = 15.3 ns. The comparison of the fluorescence lifetimes of the obtained CIGS semiconductor thin films is as follows: Figure 5 As shown.

[0107] Furthermore, for a 0.6 eV deep-level defect in a copper indium gallium selenide (CIGS) semiconductor thin film, the difference between the fluorescence lifetime corresponding to the first photoluminescence signal and the fluorescence lifetime corresponding to the second photoluminescence signal is Δt = 15.0 − 12.5 = 2.5 ns, so the influence coefficient of the 0.6 eV deep-level defect can be 2.5 ns / ppm; for a 0.8 eV deep-level defect in a CIGS semiconductor thin film, the difference between the fluorescence lifetime corresponding to the first photoluminescence signal and the fluorescence lifetime corresponding to the second photoluminescence signal is Δt = 15.6 − 12.5 = 3.1 ns, so the influence coefficient of the 0.8 eV deep-level defect can be 2.5 ns / ppm. The influence coefficient can be 3.1 ns / ppm; for a 1.0 eV deep-level defect in a copper indium gallium selenide (CIGS) semiconductor thin film, the difference between the fluorescence lifetime corresponding to the first photoluminescence signal and the fluorescence lifetime corresponding to the second photoluminescence signal is Δt = 15.3 − 12.5 = 2.8 ns, so the influence coefficient of the 0.6 eV deep-level defect can be 2.8 ns / ppm; thus, it can be seen that the three deep-level defects of 0.6 eV, 0.8 eV and 1.0 eV are all effective carrier recombination centers, among which the influence coefficient of the 0.8 eV deep-level defect is the largest, indicating that the 0.8 eV deep-level defect has the strongest effect on the decay of fluorescence lifetime.

[0108] Compared with existing detection devices, the detection device provided in this application has a detection error of only 3.2% for deep-level defects in copper indium gallium selenide (CIGS) semiconductor thin films (the detection error of existing detection devices is about 18%), improving the multi-defect detection efficiency by 88.8%, and for the first time realizing quantitative impact analysis of CIGS multi-deep-level defects, providing accurate data support for defect control of CIGS solar cells.

[0109] Accordingly, this application also provides a method for detecting semiconductor defects, applied to the detection apparatus provided in any of the above embodiments, with reference to... Figure 1 As shown, the detection method includes:

[0110] S100: The first light source component 110 in the light emission module 100 emits a pulsed laser. The photon energy of the pulsed laser is greater than the bandgap of the semiconductor sample 5. The pulsed laser is focused on the surface of the semiconductor sample 5 by the beam combining component 130 in the light emission module 100, which excites a first photoluminescence signal. The first photoluminescence signal is then transmitted to the detection module 400 through the light separation and collection module 300.

[0111] S200: While emitting pulsed laser using the first light source component 110, a continuous laser is emitted using the second light source component 120 in the light emission module 100. The photon energy of the continuous laser is equal to the energy interval between the energy level and the valence band of the target deep energy level defect in the semiconductor sample 5. The pulsed laser and the continuous laser are combined and focused on the surface of the semiconductor sample 5 by the beam combining component 130, which excites a second photoluminescence signal. The second photoluminescence signal is then transmitted to the detection module 400 through the light separation and collection module 300.

[0112] Specifically, the optical path synchronization calibration module 200 detects the focused spot positions of pulsed laser and continuous laser on the surface of semiconductor sample 5 in real time. When the offset between the two focused spots is not less than a preset threshold, the beam combining component 130 is adjusted so that the offset is less than the preset threshold. Optionally, the preset threshold can be 0.1 μm.

[0113] S300: The detection module 400 obtains the defect information of the semiconductor sample 5 based on the first photoluminescence signal and the second photoluminescence signal.

[0114] Understandably, during the process of focusing the pulsed laser and continuous laser beams onto the surface of the semiconductor sample 5, the optical path synchronization calibration module 200 detects the position of the focused spot of the pulsed laser and continuous laser on the surface of the semiconductor sample 5 in real time. When the offset of the two focused spots is not less than a preset threshold, the beam combining component 130 is adjusted so that the offset of the two focused spots is less than the preset threshold. In this way, the precise control of the focusing coaxiality of the pulsed laser and continuous laser on the surface of the semiconductor sample can be ensured, effectively reducing the fluorescence lifetime detection error caused by insufficient saturation of defect electronic states due to the offset of the focused spot of the pulsed laser and continuous laser on the surface of the semiconductor sample, and improving the detection accuracy.

[0115] In practical applications, before detecting defects in semiconductor samples, the optical path synchronization calibration module can be activated first. The optical path synchronization calibration module 200 is used to adjust the beam combining component 130 (specifically, the spatial position of the first lens 6), so that the focusing coaxiality requirement can be met directly after the subsequent pulsed laser and continuous laser are turned on.

[0116] Understandably, if only a single deep-level defect in a semiconductor sample needs to be detected, the photon energy of the continuous laser can be directly set to the energy interval between the energy level and the valence band of the deep-level defect to be detected. However, if multiple deep-level defects in a semiconductor sample need to be detected, the photon energy of the continuous laser needs to be adjusted to correspond to the energy interval between the energy level and the valence band of the deep-level defects to be detected.

[0117] Alternatively, the photon energy of the continuous laser can be manually adjusted and the detection process repeated, but this is cumbersome and time-consuming.

[0118] Another option, such as Figure 1 As shown, the second light source assembly 120 can be configured to include not only the continuous laser source 3, but also a continuous laser photon energy dynamic scanning unit 4. The continuous laser photon energy dynamic scanning unit 4 is electrically connected to the continuous laser source 3 and is used to control the continuous laser source 3 to emit photon energy in a step-by-step manner to match the energy interval between the energy level and valence band of different deep energy level defects in the semiconductor sample. This sequentially saturates the electronic states of different deep energy level defects in the semiconductor sample. At the same time, the pulsed laser excites each second photoluminescence signal corresponding to the continuous laser with different photon energies, thereby realizing the continuous detection of multi-deep energy level defects in the semiconductor sample and greatly improving the detection efficiency.

[0119] Based on this, optional, such as Figure 1 As shown, the detection module 400 includes a photodetector 16, a time-correlated single-photon detector 17, a fluorescence lifetime multi-channel parallel processing unit 18, and a defect impact quantification analysis unit 19; then, step S300 uses the detection module 400 to obtain the defect information of the semiconductor sample 5 based on the first photoluminescence signal and the second photoluminescence signal, including:

[0120] S310: The first photoluminescence signal and each of the second photoluminescence signals are converted into electrical pulse signals using the photodetector 16;

[0121] S320: Using a time-correlated single-photon detector 17 based on an electrical pulse signal synchronized with a pulsed laser, the electrical pulse signal converted by the photodetector 16 is time-stamped and statistically analyzed to obtain the intensity decay curves of the first photoluminescence signal and each second photoluminescence signal over time.

[0122] S330: The fluorescence lifetime multi-channel parallel processing unit 18 is used to synchronously acquire the curves of different wavelength bands and extract the first photoluminescence signal and the fluorescence lifetime corresponding to each second photoluminescence signal.

[0123] S340: Using the defect influence quantification analysis unit 19, the influence coefficient of each deep level defect in the semiconductor sample on the fluorescence lifetime is calculated based on the difference in fluorescence lifetime corresponding to the extracted first photoluminescence signal and each second photoluminescence signal.

[0124] Based on the foregoing analysis, when detecting multi-level defects in a semiconductor sample, a pulsed laser is first focused onto the surface of the semiconductor sample 5 to excite a first photoluminescence signal. The first fluorescence lifetime t1 can be calculated based on the intensity decay curve of the first photoluminescence signal over time. Then, the pulsed laser and a continuous laser beam are combined and focused onto the surface of the semiconductor sample 5. The photon energy of the continuous laser is progressively matched to the energy intervals between the energy levels and valence bands of different deep-level defects within the semiconductor sample, thus exciting second photoluminescence signals corresponding to different photon energies. The corresponding fluorescence lifetime t1 can be calculated based on the intensity decay curve of each second photoluminescence signal over time. The second fluorescence lifetime t2 is calculated; then, the difference Δt between the second fluorescence lifetime t2 and the first fluorescence lifetime t1 corresponding to continuous lasers with different photon energies is calculated. If Δt > 0, it indicates that the corresponding deep-level defect is an effective carrier recombination center. Furthermore, the larger Δt is, the stronger the carrier capture ability of the corresponding deep-level defect and the larger the influence coefficient. Thus, quantitative analysis of multi-deep-level defects in semiconductor samples can be realized, and the influence coefficient of different deep-level defects can be output. This provides a quantitative basis for defect control of semiconductor materials and makes up for the shortcomings of existing detection devices, which can only realize qualitative analysis of defects and cannot quantify the degree of influence of defects on fluorescence lifetime, making it difficult to meet the needs of precise optimization of semiconductor materials.

[0125] Furthermore, the use of the continuous laser photon energy dynamic scanning unit 4 enables the continuous laser to perform dynamic scanning, and combined with the multi-channel parallel detection of the fluorescence lifetime multi-channel parallel processing unit 18, the efficiency of continuous detection of multi-deep level defects can be further improved (≥80%).

[0126] Compared to existing detection methods that use fixed filters to filter out pulsed and continuous lasers of fixed wavelengths, which cannot match laser wavelength changes in real time and may lead to incomplete excitation light filtering, affecting the purity of the photoluminescence signal, and if multiple deep-level defects in a semiconductor sample need to be detected, it is necessary not only to adjust the photon energy of the continuous laser and repeat the detection process, but also to replace the filters one by one and adjust the laser parameters; the detection device provided in this application, such as Figure 1 As shown, the light separation and collection module 300 may include a tunable filter component 12, and the detection method provided in this application may further include:

[0127] S400: The tunable filter component 12 is used to match the wavelengths of pulsed laser and continuous laser in real time, filter out pulsed laser and continuous laser, and transmit the first photoluminescence signal and the second photoluminescence signal.

[0128] In other words, the tunable filter assembly 12 has an automatic calibration function, which can match the excitation light wavelength in real time, so that there is no need to manually replace the filter. At the same time, it can ensure the purity of the transmitted photoluminescence signal (transmittance greater than or equal to 95%), and can also be matched for multi-level defect detection.

[0129] In practical applications, before detecting defects in semiconductor samples, the automatic calibration function of the tunable filter component 12 can be activated to match the initial excitation wavelength.

[0130] Since the structure and implementation steps of the detection device provided in this application have been described in detail in the foregoing embodiments, and the implementation steps of the detection method and detection device provided in this application correspond to each other, the specific details of the detection method can be referred to the foregoing embodiments, and will not be repeated here.

[0131] In summary, the embodiments of this application propose a detection device and method that integrates "precise optical path calibration, dynamic wavelength adaptation, parallel detection of multiple defects, and quantitative analysis of impact", which significantly improves the accuracy, efficiency and data value of semiconductor defect detection and can overcome the shortcomings of low detection accuracy, low efficiency and lack of quantitative analysis in the prior art.

[0132] The various parts of this manual are described in a combination of parallel and progressive methods. Each part focuses on the differences between the other parts, and the same or similar parts can be referred to each other.

[0133] The features described above regarding the disclosed embodiments can be substituted or combined with each other to enable those skilled in the art to implement or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A semiconductor defect detection device, characterized in that, The detection device is used for defect detection of semiconductor samples, and includes: A light-emitting module includes a first light source assembly, a second light source assembly, and a beam combiner assembly. The first light source assembly emits pulsed laser light, the photon energy of which is greater than the bandgap of the semiconductor sample. The second light source assembly emits continuous laser light, the photon energy of which is equal to the energy interval between the energy level and the valence band of a target deep-level defect within the semiconductor sample. The beam combiner assembly focuses the pulsed laser light onto the surface of the semiconductor sample to generate a first photoluminescence signal, and combines the pulsed laser light and the continuous laser light beam and focuses them onto the surface of the semiconductor sample to generate a second photoluminescence signal. The optical path synchronization calibration module is used to detect the focused spot positions of the pulsed laser and the continuous laser on the surface of the semiconductor sample in real time. When the offset between the two is not less than a preset threshold, the beam combining component is adjusted so that the offset is less than the preset threshold. A light separation and collection module is used to transmit the first photoluminescence signal and the second photoluminescence signal, and to filter out the pulsed laser and the continuous laser. The detection module is used to obtain defect information of the semiconductor sample based on the first photoluminescence signal and the second photoluminescence signal.

2. The semiconductor defect detection device according to claim 1, characterized in that, The beam combining assembly includes a laser beam combiner, a beam splitter, and a first lens; The laser beam combiner is used to combine the pulsed laser and the continuous laser into the same optical path; The beam splitter is used to project the pulsed laser and the continuous laser, which share the same optical path, onto the first lens; The first lens is used to focus the pulsed laser onto the surface of the semiconductor sample to generate the first photoluminescence signal, and to focus the combined pulsed laser and the continuous laser onto the surface of the semiconductor sample to generate the second photoluminescence signal. The first lens and the beam splitter are also used to project the first photoluminescence signal and the second photoluminescence signal onto the light separation and collection module.

3. The semiconductor defect detection device according to claim 2, characterized in that, The optical path synchronization calibration module includes a position sensor, a piezoelectric ceramic adjustment stage, and a controller, wherein the piezoelectric ceramic adjustment stage is fixedly connected to the first lens; The position sensor is used to detect in real time the position of the focused spot of the pulsed laser and the continuous laser on the surface of the semiconductor sample; The controller is used to determine whether the offset between the pulsed laser and the continuous laser on the surface of the semiconductor sample is not less than a preset threshold, based on the focused spot positions of the two lasers. When the offset is not less than the preset threshold, the controller controls the piezoelectric ceramic adjustment stage to adjust the spatial position of the first lens so that the offset is less than the preset threshold.

4. The semiconductor defect detection device according to claim 1 or 2, characterized in that, The light separation and collection module includes a tunable filter assembly, a second lens, and a monochromator arranged along the optical path transmission direction; The tunable filter component is used to match the wavelengths of the pulsed laser and the continuous laser in real time, filter out the pulsed laser and the continuous laser, and allow the first photoluminescence signal and the second photoluminescence signal to pass through. The second lens is used to focus the first photoluminescence signal and the second photoluminescence signal into the monochromator; The monochromator is used to filter out the first photoluminescence signal and the second photoluminescence signal, and transmit them to the detection module.

5. The semiconductor defect detection device according to claim 1, characterized in that, The second light source assembly includes a continuous laser source and a continuous laser photon energy dynamic scanning unit; The continuous laser photon energy dynamic scanning unit is electrically connected to the continuous laser source and is used to control the continuous laser source to emit the continuous laser. The photon energy of the continuous laser is matched step by step to match the energy interval between the energy level and the valence band of different deep energy level defects in the semiconductor sample, so that the electronic states of different deep energy level defects in the semiconductor sample are saturated respectively. At the same time, the pulsed laser excites the second photoluminescence signal corresponding to the continuous laser with different photon energies.

6. The semiconductor defect detection device according to claim 5, characterized in that, The detection module includes a photodetector, a time-correlated single-photon detector, a fluorescence lifetime multi-channel parallel processing unit, and a defect impact quantification analysis unit. The photodetector is used to convert the first photoluminescence signal and each of the second photoluminescence signals into electrical pulse signals; The time-correlated single-photon detector is used to perform time stamping and statistical analysis on the electrical pulse signal converted by the photodetector based on the electrical pulse signal synchronized with the pulsed laser, so as to obtain the intensity decay curves of the first photoluminescence signal and each of the second photoluminescence signals over time. The fluorescence lifetime multi-channel parallel processing unit is used to simultaneously acquire the curves of different wavelength bands and extract the fluorescence lifetime corresponding to the first photoluminescence signal and each of the second photoluminescence signals. The defect impact quantification analysis unit is used to calculate the impact coefficient of each deep-level defect in the semiconductor sample on the fluorescence lifetime based on the difference between the fluorescence lifetime corresponding to the extracted first photoluminescence signal and the fluorescence lifetime corresponding to each of the second photoluminescence signals.

7. The semiconductor defect detection apparatus according to claim 6, characterized in that, The first light source assembly includes a pulsed laser source and a pulsed laser source control power supply; The pulsed laser source control power supply is electrically connected to the pulsed laser source and is used to control the frequency and pulse width of the pulsed laser emitted by the pulsed laser source. It is also used to transmit an electrical pulse signal synchronized with the pulsed laser to the time-correlated single-photon detector.

8. A method for detecting semiconductor defects, characterized in that, The detection method, applied to the detection apparatus according to any one of claims 1-7, comprises: The first light source component in the light emission module emits a pulsed laser, the photon energy of which is greater than the bandgap of the semiconductor sample. The pulsed laser is focused on the surface of the semiconductor sample by the beam combining component in the light emission module, thereby exciting a first photoluminescence signal. The first photoluminescence signal is then transmitted to the detection module through the light separation and collection module. While the first light source component emits the pulsed laser, the second light source component in the light emission module emits a continuous laser. The photon energy of the continuous laser is equal to the energy interval between the energy level and the valence band of the target deep energy level defect in the semiconductor sample. The pulsed laser and the continuous laser are combined and focused on the surface of the semiconductor sample by the beam combining component to excite a second photoluminescence signal. The second photoluminescence signal is then transmitted to the detection module through the light separation and collection module. The optical path synchronization calibration module is used to detect the focused spot positions of the pulsed laser and the continuous laser on the surface of the semiconductor sample in real time. When the offset between the two is not less than a preset threshold, the beam combining component is adjusted so that the offset is less than the preset threshold. The detection module uses the first photoluminescence signal and the second photoluminescence signal to obtain the defect information of the semiconductor sample.

9. The semiconductor defect detection method according to claim 8, characterized in that, The photon energy of the continuous laser is matched stepwise to the energy interval between the energy level and the valence band of different deep energy level defects in the semiconductor sample, thereby saturating the electronic states of the different deep energy level defects in the semiconductor sample. At the same time, the pulsed laser excites the second photoluminescence signal corresponding to the continuous laser with different photon energies. The detection module includes a photodetector, a time-correlated single-photon detector, a fluorescence lifetime multi-channel parallel processing unit, and a defect impact quantification analysis unit. The defect information of the semiconductor sample obtained by the detection module based on the first photoluminescence signal and the second photoluminescence signal includes: The photodetector is used to convert the first photoluminescence signal and each of the second photoluminescence signals into electrical pulse signals; Using the time-correlated single-photon detector based on the electrical pulse signal synchronized with the pulsed laser, the electrical pulse signal converted by the photodetector is time-stamped and statistically analyzed to obtain the intensity decay curves of the first photoluminescence signal and each of the second photoluminescence signals over time; The fluorescence lifetime multi-channel parallel processing unit is used to simultaneously acquire the curves of different wavelength bands, and the fluorescence lifetimes corresponding to the first photoluminescence signal and each second photoluminescence signal are extracted. The defect impact quantification analysis unit calculates the impact coefficient of each deep-level defect in the semiconductor sample on the fluorescence lifetime based on the difference between the fluorescence lifetime corresponding to the extracted first photoluminescence signal and the fluorescence lifetime corresponding to each of the second photoluminescence signals.

10. The method for detecting semiconductor defects according to claim 8 or 9, characterized in that, The light separation and collection module includes a tunable filter component; the detection method further includes: The tunable filter component is used to match the wavelengths of the pulsed laser and the continuous laser in real time, filtering out the pulsed laser and the continuous laser, and allowing the first photoluminescence signal and the second photoluminescence signal to pass through.

Images