A method for spatially resolved non-destructive detection of electrical parameters of a semiconductor wafer based on optically pumped-terahertz detection spectroscopy

Abstract

This invention relates to the field of wafer parameter testing technology, and discloses a spatially resolved non-destructive testing method for the electrical parameters of semiconductor wafers based on optical pump-terahertz probe spectroscopy. The method includes: generating pulsed light using an ultrafast laser; splitting the pulsed light into pump light and terahertz probe light using a beam splitter; focusing the pump light and terahertz probe light onto the same measurement position on the surface of the semiconductor wafer under test; reconstructing the carrier relaxation dynamics curve at this measurement position; obtaining the dynamic parameters and transient conductance changes related to the wafer carrier recombination process; acquiring the electrical parameters of the semiconductor wafer under test at different measurement positions and generating a two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test; and identifying abnormal regions of electrical parameters on the surface of the semiconductor wafer under test. This invention enables high spatial resolution, rapid, and comprehensive detection of the electrical parameters of semiconductor wafers without damaging the wafer surface structure and electrical properties through non-contact optical means.

Description

Technical Field

[0001] This invention relates to the field of wafer parameter detection technology, and more specifically, to a spatially resolved non-destructive testing method for the electrical parameters of semiconductor wafers based on optical pump-terahertz detection spectroscopy. Background Technology

[0002] Semiconductor wafers are the core foundational materials for the manufacture of integrated circuits, power devices, and optoelectronic devices. Their electrical and structural parameters, such as conductivity, mobility, carrier lifetime, and defect density, directly affect the performance and production yield of subsequent devices. Currently, the measurement of wafer electrical properties mainly relies on contact testing methods, including Hall effect measurement, four-probe method, electrode contact sheet resistance measurement, and electrical characterization after fabrication of specific metal structures. These methods require the fabrication of electrodes on the wafer surface or the use of mechanical probes to contact the sample, which can easily introduce problems such as contact resistance, ohmic contact uncertainty, and surface damage. For high-value semiconductor wafers such as GaN, SiC, and GaAs, such contact processes lead to additional losses and higher testing costs. Furthermore, contact methods typically only measure local areas, making it difficult to achieve large-area scanning and two-dimensional uniformity assessment of the entire wafer. As wafer diameters continue to increase, the operational complexity and time consumption of contact methods also increase significantly, gradually limiting their applicability in rapid detection and online monitoring scenarios.

[0003] Therefore, it is necessary to design a spatially resolved nondestructive testing method for the electrical parameters of semiconductor wafers based on optical pump-terahertz detection spectroscopy to solve the problems existing in the current technology. Summary of the Invention

[0004] In view of this, the present invention proposes a spatially resolved non-destructive testing method for the electrical parameters of semiconductor wafers based on optical pump-terahertz detection spectroscopy, aiming to solve the structural problems of current contact methods, such as easy damage, inability to scan large areas, and uncertain contact resistance.

[0005] This invention proposes a spatially resolved non-destructive testing method for the electrical parameters of semiconductor wafers based on optical pump-terahertz detection spectroscopy, comprising the following steps: S100: Uses an ultrafast laser to generate pulsed light, and splits the pulsed light into pump light and terahertz probe light through a beam splitter module; S200: Performs optical path adjustment on the pump light and terahertz probe light, and after optical path adjustment, focuses the pump light and terahertz probe light onto the same measurement position on the surface of the semiconductor wafer under test; S300: By adjusting the time delay in the terahertz probe light path, the terahertz probe light is irradiated onto the surface of the semiconductor wafer under test at different delay times, and the corresponding terahertz probe light signal intensity is collected, thereby obtaining the terahertz probe light signal sequence of the measurement position as a function of time delay, so as to reconstruct the carrier relaxation dynamics curve of the measurement position. S400: Based on a preset semiconductor material dynamics analysis model, the carrier relaxation dynamics curve is fitted to obtain the dynamic parameters and transient conductance changes related to the wafer carrier recombination process. S500: Using dynamic parameters and transient conductance changes as inputs, it inversely solves the dynamic analysis model of semiconductor materials to infer the electrical parameters of the semiconductor wafer under test. S600: Fix the semiconductor wafer under test on a programmable two-dimensional displacement platform, move the position of the semiconductor wafer under test by controlling the displacement platform, and repeat steps S200 to S500 to obtain the electrical parameters of the semiconductor wafer under test at different measurement positions, and generate a two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test. S700: Evaluate the spatial uniformity of the electrical properties of the semiconductor wafer under test based on the two-dimensional spatial distribution map of electrical parameters, and identify abnormal regions of electrical parameters on the surface of the semiconductor wafer under test.

[0006] Furthermore, when adjusting the optical path of the pump light and the terahertz probe light, the following steps are included: An adjustable time delay device is set in the path of the terahertz probe light to control the relative arrival time of the pump light and the terahertz probe light.

[0007] Further, when acquiring the terahertz probe light signal sequence at the measurement location, which varies with time delay, to reconstruct the carrier relaxation dynamics curve at the measurement location, the process includes: By adjusting the optical delay device set in the terahertz probe light path, the terahertz probe light is made to illuminate the surface of the semiconductor wafer under test at different delay times. At each delay time, the corresponding terahertz probe light signal intensity is collected to form a terahertz probe light signal sequence that varies with time delay; The acquired terahertz probe optical signal sequence was denoised and normalized. Based on the physical model of pump light-induced photogenerated carriers, the terahertz probe light signal sequence is mapped as a function of carrier concentration or transient conductance as a function of time. The function of carrier concentration or transient conductance as a function of time obtained by mapping is formed into a continuous curve, which is the reconstructed carrier relaxation dynamics curve.

[0008] Furthermore, the kinetic parameters include carrier lifetime and relaxation time constant.

[0009] Furthermore, the electrical parameters include mobility, conductivity, and defect-related recombination rate.

[0010] Furthermore, the two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test is a two-dimensional image with the coordinates of different measurement positions as the horizontal and vertical axes, and the electrical parameter values ​​at the corresponding positions as color or grayscale values.

[0011] Furthermore, when evaluating the spatial uniformity of the electrical properties of the semiconductor wafer under test based on the two-dimensional spatial distribution map of electrical parameters, and identifying abnormal regions of electrical parameters on the surface of the semiconductor wafer under test, the following steps are taken: Based on the two-dimensional spatial distribution diagram of electrical parameters, the overall analysis area and the sub-region of interest of the semiconductor wafer under test are determined; Statistical analysis is performed on the electrical parameter values ​​within the overall analysis area to obtain overall statistical indicators; among which, the overall statistical indicators include the overall electrical parameter mean, the overall electrical parameter standard deviation, and the overall electrical parameter variance; Calculate the regional statistical index for each sub-region of interest, and identify abnormal electrical parameter regions on the surface of the semiconductor wafer under test based on the overall statistical index, regional statistical index and preset judgment threshold. The spatial uniformity of the output is evaluated based on the abnormal electrical parameters in the region.

[0012] Furthermore, when determining the overall analysis region and the sub-region of interest of the semiconductor wafer under test, the following steps are included: Using the boundary of the semiconductor wafer under test as the boundary, all measurement positions on the surface of the semiconductor wafer under test are set as the overall analysis area; The overall analysis area is divided into multiple non-overlapping sub-regions of interest using a rectangular grid. Each region of interest contains several measurement locations, and the coordinates of the measurement locations are determined by controlling a two-dimensional displacement platform. For each sub-region of interest, record the spatial coordinate range of its upper left and lower right corners.

[0013] Furthermore, when identifying abnormal electrical parameter regions on the surface of the semiconductor wafer under test based on overall statistical indicators, regional statistical indicators, and preset judgment thresholds, the process includes: Calculate the deviation between the regional statistical indicators and the corresponding overall statistical indicators; whereby the regional statistical indicators include the mean, standard deviation, and variance of the regional electrical parameters. The deviation value is compared with a preset judgment threshold. When the deviation value exceeds the preset judgment threshold range, the sub-region of interest is determined to be a candidate region for abnormal electrical parameters. The electrical parameter values ​​at each measurement location within the candidate area of ​​abnormal electrical parameters are verified one by one. The absolute deviation between the electrical parameter value at a single measurement location and the overall average electrical parameter value is calculated. If the absolute deviation is greater than the preset single-point abnormal threshold, the measurement location is marked as an abnormal point. The percentage of abnormal points within the candidate region of abnormal electrical parameters is counted. When the percentage exceeds a preset threshold, the sub-region of interest is finally determined as an abnormal electrical parameter region.

[0014] Furthermore, when evaluating the spatial uniformity of output based on the abnormal electrical parameter regions, the following are included: The percentage of all abnormal electrical parameter areas relative to the total surface area of ​​the semiconductor wafer under test is statistically analyzed and used as a quantitative evaluation index of spatial uniformity. The quantitative evaluation index is output as the spatial uniformity evaluation result.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Truly non-destructive testing: It adopts a pure optical non-contact measurement method, which eliminates the need to make electrodes or use mechanical probes, completely eliminating the risks of contact resistance, interface barriers, metal contamination and sample damage, and perfectly maintaining the structural integrity of high-value wafers.

[0016] 2. Abundant Inversion Information: Leveraging the time-resolution capability of optically pumped-terahertz probe spectroscopy, the entire process of carrier relaxation dynamics can be comprehensively acquired. Through model fitting, multiple key parameters such as carrier lifetime, mobility, and conductivity can be directly and accurately inverted from a single dynamic curve, with an information content and accuracy far exceeding that of steady-state optical measurements.

[0017] 3. High efficiency, speed, and high spatial resolution: Single-point testing speed can reach the second level. Combined with an automated two-dimensional displacement platform, it can complete large-area scanning of the entire wafer in a short time, meeting the needs of online monitoring. At the same time, the micron-level focused spot gives the method high spatial resolution, which can accurately draw two-dimensional distribution maps of electrical parameters, intuitively evaluate the uniformity of material electrical properties, and accurately locate defects.

[0018] 4. High precision and strong applicability: It avoids many uncertainties of contact measurement, ensuring high accuracy and repeatability of measurement results. The system has a stable structure and a high degree of engineering and automation. It is not only suitable for traditional materials such as Si, GaAs, and InP, but also particularly suitable for the inspection of wide bandgap semiconductor wafers such as GaN and SiC, combining the dual value of cutting-edge scientific characterization and industrial mass production monitoring. Attached Figure Description

[0019] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 The flowchart illustrates a spatially resolved nondestructive testing method for the electrical parameters of semiconductor wafers based on optically pumped terahertz detection spectroscopy, provided in an embodiment of the present invention. Detailed Implementation

[0020] Exemplary embodiments of the present disclosure will now be described in more detail with reference to the accompanying drawings. While exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be implemented in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the disclosure to those skilled in the art. It should be noted that, unless otherwise specified, embodiments and features in the embodiments of the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0021] See Figure 1 As shown in some embodiments of this application, this embodiment provides a spatially resolved non-destructive testing method for the electrical parameters of semiconductor wafers based on optically pumped-terahertz detection spectroscopy, including the following steps: S100: Uses an ultrafast laser to generate pulsed light, and splits the pulsed light into pump light and terahertz probe light through a beam splitter module; S200: Performs optical path adjustment on the pump light and terahertz probe light, and after optical path adjustment, focuses the pump light and terahertz probe light onto the same measurement position on the surface of the semiconductor wafer under test; S300: By adjusting the time delay in the terahertz probe light path, the terahertz probe light is irradiated onto the surface of the semiconductor wafer under test at different delay times, and the corresponding terahertz probe light signal intensity is collected, thereby obtaining the terahertz probe light signal sequence of the measurement position as a function of time delay, so as to reconstruct the carrier relaxation dynamics curve of the measurement position. S400: Based on a preset semiconductor material dynamics analysis model, the carrier relaxation dynamics curve is fitted to obtain the dynamic parameters and transient conductance changes related to the wafer carrier recombination process. S500: Using dynamic parameters and transient conductance changes as inputs, it inversely solves the dynamic analysis model of semiconductor materials to infer the electrical parameters of the semiconductor wafer under test. S600: Fix the semiconductor wafer under test on a programmable two-dimensional displacement platform, move the position of the semiconductor wafer under test by controlling the displacement platform, and repeat steps S200 to S500 to obtain the electrical parameters of the semiconductor wafer under test at different measurement positions, and generate a two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test. S700: Evaluate the spatial uniformity of the electrical properties of the semiconductor wafer under test based on the two-dimensional spatial distribution map of electrical parameters, and identify abnormal regions of electrical parameters on the surface of the semiconductor wafer under test.

[0022] In this embodiment, the pump light is used to irradiate the surface of the semiconductor wafer under test to excite the generation of photogenerated carriers, causing a transient change in the reflectivity or transmittance of the semiconductor wafer under test; the terahertz probe light is a terahertz pulse used to obtain the relaxation dynamics information of the carriers of the semiconductor wafer under test.

[0023] In this embodiment, the semiconductor material dynamics analysis model encompasses the physical mapping relationship that maps the terahertz probe optical signal sequence to the carrier dynamics curve, as well as the mathematical fitting function used for quantitative analysis of the dynamics curve.

[0024] The physical mapping relationship is constructed based on the electromagnetic response theory (such as the Drude model) of the interaction between terahertz waves and photogenerated carriers in semiconductors. It aims to convert the detected changes in the time-domain electric field intensity of terahertz waves into transient changes in photoconductivity or carrier concentration of semiconductor materials.

[0025] Mathematical fitting functions are used to describe the relaxation behavior of transient photoconductivity or carrier concentration over time. Depending on the characteristics of the semiconductor material under test, appropriate function forms can be selected or constructed. For example, for materials with relatively simple recombination mechanisms, a single exponential decay function can be used for fitting to extract the average lifetime of carriers; while for materials with two significantly different recombination processes (fast and slow), a double exponential decay function can be used for fitting to obtain the characteristic time constants corresponding to surface recombination and bulk recombination, respectively.

[0026] By fitting the data, intermediate kinetic parameters, including carrier lifetime, transient conductivity variation, and relaxation time constant, can be obtained. Furthermore, based on the inherent relationships between mobility, transient conductivity, and carrier concentration in semiconductor physics (e.g., Δσ=e*μ*Δn), and the correlation model between defect recombination rate and carrier lifetime, electrical parameters such as mobility, conductivity, and defect-related recombination rate of the semiconductor wafer under test can be calculated from the intermediate kinetic parameters.

[0027] In one embodiment of this application, an ultrafast laser is used to generate pulsed laser light. The pulsed laser light first passes through a beam splitter, splitting into pump light and terahertz probe light. The pump light, after optical path adjustment, enters a time delay device. By adjusting the time delay device to change the optical path length of the pump light, the delayed pump light is then focused onto a measurement position on the surface of a semiconductor wafer under test, fixed on a programmable two-dimensional displacement platform. The photon energy of the pump light is greater than the bandgap energy of the semiconductor wafer material, exciting non-equilibrium carriers at the measurement position. The terahertz probe light is further split into terahertz generation light and probe light by the beam splitter. The terahertz generation light is guided to a terahertz emitter to generate a terahertz probe light signal. The terahertz probe light signal, after collimation and focusing by a terahertz optical system, illuminates the measurement position on the surface of the semiconductor wafer under test, coinciding spatially with the pump light. By controlling the time delay device, the pump light and terahertz probe light signals are sequentially applied to the measurement position with different pump-probe delay times. When the pump light arrives at the measurement location before the terahertz probe light signal and excites carriers, the electric field intensity of the subsequently arriving terahertz probe light signal is modulated by the dynamic behavior of the photogenerated carriers as it penetrates or reflects through the semiconductor wafer. The terahertz probe light signal carrying sample response information is ultimately received by a terahertz detector. The terahertz detector employs electro-optic sampling or photoconductive sampling, using the time-synchronized probe light as a gating signal to convert the instantaneous electric field intensity of the terahertz probe light signal into measurable signal intensity data. By scanning a time delay device, the corresponding terahertz probe light signal intensity is acquired at different pump-probe delay times, thereby obtaining the terahertz time-domain signal sequence of the terahertz probe light signal at the measurement location as a function of time delay. Based on this, the carrier relaxation dynamics curve at that measurement location can be reconstructed. Furthermore, by comparing and analyzing the changes in the terahertz time-domain signal sequence before and after pump excitation, and by fitting the data based on a pre-defined semiconductor material dynamics analysis model, the transient conductance change, carrier mobility, carrier recombination lifetime, and other electrical parameters at the measurement location can be quantitatively obtained. By controlling a programmable two-dimensional displacement platform to move the semiconductor wafer under test and repeating the above measurement process at different measurement locations, non-destructive spatial resolution detection of the electrical parameters of the entire semiconductor wafer can be achieved, and a two-dimensional spatial distribution map of the electrical parameters can be generated.

[0028] In this embodiment, an adjustable time delay device can also be provided in the pump light path. The specific steps are as follows: S100: Uses an ultrafast laser to generate pulsed light, and splits the pulsed light into pump light and probe / generation light beam through a beam splitter; the probe / generation light beam is further split into terahertz generation light and terahertz probe light through a second beam splitter; wherein, the terahertz pulsed light includes terahertz generation light and terahertz probe light; S200: The pump light and terahertz pulse light are optically calibrated, and after optical path calibration, the pump light and terahertz pulse are focused onto the same measurement position on the surface of the semiconductor wafer under test. S300: By adjusting the optical delay in the pump light path, the time difference between the pump light pulse and the terahertz detection pulse arriving at the sample measurement point is changed; at each time difference, the terahertz time-domain electric field signal after being acted upon by the sample is acquired; by analyzing the change sequence of this signal with the pump light time delay, the carrier relaxation dynamics curve at the measurement location is reconstructed. S400: Based on a preset semiconductor material dynamics analysis model, the carrier relaxation dynamics curve is fitted to obtain the dynamic parameters and transient conductance changes related to the wafer carrier recombination process. S500: Using dynamic parameters and transient conductance changes as inputs, it inversely solves the dynamic analysis model of semiconductor materials to infer the electrical parameters of the semiconductor wafer under test. S600: Fix the semiconductor wafer under test on a programmable two-dimensional displacement platform, move the position of the semiconductor wafer under test by controlling the displacement platform, and repeat steps S200 to S500 to obtain the electrical parameters of the semiconductor wafer under test at different measurement positions, and generate a two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test. S700: Evaluate the spatial uniformity of the electrical properties of the semiconductor wafer under test based on the two-dimensional spatial distribution map of electrical parameters, and identify abnormal regions of electrical parameters on the surface of the semiconductor wafer under test.

[0029] Understandably, whether the time delay device is placed in the terahertz probe optical path or the pump optical path, the core purpose is to achieve precise control of the relative time delay between the pump light and the terahertz probe light, thereby acquiring terahertz probe light signals under different time delays to fully capture the entire dynamic process of charge carriers from generation to relaxation. The two optical path configurations are fundamentally similar and can both be effectively used to reconstruct the carrier relaxation dynamics curve, ultimately achieving spatially resolved non-destructive testing of the electrical parameters of semiconductor wafers. The specific choice of optical path configuration can be flexibly adjusted based on actual experimental conditions, system complexity requirements, and the research focus on a particular physical process.

[0030] It is understood that the spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped terahertz detection spectroscopy provided in this embodiment can achieve high spatial resolution, rapid, and comprehensive detection of semiconductor wafer electrical parameters without damaging the wafer surface structure and electrical properties through non-contact optical means. This method can not only accurately acquire key electrical parameters such as carrier lifetime, mobility, and conductivity at any measurement location on the wafer, but also generate an intuitive two-dimensional spatial distribution map of electrical parameters through two-dimensional scanning and data processing, clearly presenting the electrical performance distribution of the entire wafer. Based on this, through statistical analysis and deviation determination of the overall and local regions, the spatial uniformity of wafer electrical performance can be effectively evaluated, and abnormal regions of electrical parameters on the surface can be accurately identified. This provides strong technical support for semiconductor wafer quality control, defect analysis, process optimization, and reliability assessment, and is particularly suitable for semiconductor materials with high added value and high precision requirements.

[0031] Specifically, optical path adjustment for the pump light and terahertz probe light includes: An adjustable time delay device is set in the path of the terahertz probe light to control the relative arrival time of the pump light and the terahertz probe light.

[0032] Understandably, introducing an adjustable time delay device into the terahertz probe light path allows for precise control of the time difference between the arrival of the pump light and the terahertz probe light at the same measurement position on the surface of the semiconductor wafer under test. This is crucial for obtaining the carrier relaxation dynamics curve. After the pump light arrives first to excite the generated carriers, the terahertz probe light detects at different delay time points, capturing the transient response of the semiconductor wafer's optical properties from carrier generation to relaxation. For example, the delay time can start from the negative delay before pump light excitation, covering different time windows to completely record the carrier process. The adjustable time delay device can use a mechanical displacement platform to drive the reflector to change the optical path difference, or it can use faster methods such as electro-optic or acousto-optic modulation. The specific choice depends on the delay range, adjustment accuracy, and time resolution requirements. In this way, the terahertz probe light can record the dynamic behavior of carriers like a "high-speed camera," laying the foundation for subsequent curve reconstruction and extraction of key electrical parameters.

[0033] Specifically, when acquiring the terahertz probe light signal sequence at the measurement location, which varies with time delay, to reconstruct the carrier relaxation dynamics curve at that measurement location, the process includes: By adjusting the optical delay device set in the terahertz probe light path, the terahertz probe light is made to illuminate the surface of the semiconductor wafer under test at different delay times. At each delay time, the corresponding terahertz probe light signal intensity is collected to form a terahertz probe light signal sequence that varies with time delay; The acquired terahertz probe optical signal sequence was denoised and normalized. Based on the physical model of pump light-induced photogenerated carriers, the terahertz probe light signal sequence is mapped as a function of carrier concentration or transient conductance as a function of time. The function of carrier concentration or transient conductance as a function of time obtained by mapping is formed into a continuous curve, which is the reconstructed carrier relaxation dynamics curve.

[0034] In this embodiment, the physical model for pump-excited photogenerated carriers is preferably a model based on the theory of photogenerated carrier generation and recombination in semiconductor materials. This model considers various carrier generation and recombination mechanisms, such as interband transitions, impurity-level assisted recombination, and Auger recombination, and incorporates dynamic processes such as carrier diffusion, drift, and interaction with the crystal lattice. This model can quantitatively describe the evolution of photogenerated carrier concentration over time. Its core lies in correlating changes in the intensity of the terahertz probe light signal with changes in carrier concentration or transient conductivity.

[0035] Understandably, denoising and normalizing the terahertz probe optical signal sequence can eliminate interference from stray light, detector noise, and light intensity fluctuations in the experimental environment, ensuring accurate subsequent data mapping. Normalization unifies the dimensions or benchmark of the terahertz probe optical signal intensity at different delay times, using the pre-excitation signal intensity as a reference value, making the signal sequence more intuitively reflect the optical response changes of carrier relaxation. Mapping based on the physical model of pump light-excited photogenerated carriers is crucial for connecting the optical measurement signal with the dynamic behavior of semiconductor carriers. Terahertz pulses are sensitive to changes in carrier concentration or conductivity, and changes in terahertz probe optical signal intensity are physically related to changes in carrier concentration or transient conductivity. Using this model, the preprocessed terahertz probe optical signal sequence can be converted into a curve showing the change in carrier concentration or transient conductivity over time, reflecting the carrier relaxation dynamics characteristics, and ultimately reconstructing a relaxation dynamics curve that illustrates the process of carrier generation to equilibrium recovery. This curve contains information about the semiconductor's electrical properties and is the basis for extracting dynamic and electrical parameters.

[0036] Specifically, the dynamic parameters include carrier lifetime and relaxation time constant.

[0037] As is understandable, carrier lifetime refers to the average time from the generation of photogenerated carriers to their disappearance through various recombination mechanisms, reflecting the carrier recombination efficiency of semiconductor materials. A longer carrier lifetime means fewer recombination centers and defects in the material, which is crucial for the performance of optoelectronic devices, allowing carriers more time to participate in transport and separation, thus improving device conversion efficiency or responsivity. The relaxation time constant, on the other hand, more broadly describes the rate of recovery of a carrier system from a non-equilibrium excited state to an equilibrium state, encompassing multiple timescale processes such as carrier thermalization, cooling, transitions between different energy states, and the final recombination lifetime. By fitting carrier relaxation dynamics curves, different time constants can be analyzed, leading to a deeper understanding of the microscopic dynamic processes of carriers, such as the process of carriers reaching thermal equilibrium after photoexcitation and recombination after thermal equilibrium. These dynamic parameters are important indicators for evaluating the quality of semiconductor materials and are also key bases for constructing and verifying semiconductor material dynamic analysis models; their accuracy affects the reliability of subsequent inverse solutions to electrical parameters.

[0038] Specifically, electrical parameters include mobility, conductivity, and defect-related recombination rate.

[0039] As is understandable, mobility is a physical quantity characterizing the ease with which charge carriers move in semiconductor materials, directly affecting device switching speed and current driving capability. Higher mobility means faster carrier drift, resulting in better device response and conductivity. The mobility obtained by this method reflects the actual transport characteristics of charge carriers in the wafer under test and is one of the core parameters for evaluating the electrical transport performance of semiconductor materials. Conductivity comprehensively reflects the conductivity of semiconductor materials and is closely related to carrier concentration and mobility; it is a macroscopic physical quantity measuring the overall electrical conduction level of the material. In device design and manufacturing, the uniformity and absolute value of conductivity are important indicators to ensure consistent and stable device performance. Defect-related recombination rate reflects the activity of various defects in the semiconductor wafer as carrier recombination centers. A high recombination rate means more harmful defects in the material, which reduces carrier lifetime and degrades device optoelectronic performance and reliability. Accurate measurement of recombination rate can assess wafer material quality and provide crucial information for tracing defect sources and optimizing processes. Accurately obtaining these electrical parameters is the foundation for a comprehensive and in-depth evaluation of the electrical performance of semiconductor wafers.

[0040] Specifically, the two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test is a two-dimensional image with the coordinates of different measurement positions as the horizontal and vertical axes, and the electrical parameter values ​​at the corresponding positions as color or grayscale values.

[0041] Understandably, by dividing the semiconductor wafer surface into multiple discrete measurement locations, reconstructing the carrier relaxation kinetics curve and extracting electrical parameters for each location, the corresponding electrical parameter values ​​can be obtained. Subsequently, these values ​​are correlated with the two-dimensional coordinates (e.g., x-axis and y-axis coordinates) of their respective measurement locations, and interpolation algorithms or direct mapping are used to transform the discrete electrical parameter values ​​into a continuously distributed two-dimensional data field. When forming the two-dimensional image, color mapping (pseudo-color image) or grayscale levels are typically used to visually represent the differences in the magnitude of the electrical parameters. For example, one color (e.g., red) can be set to represent a higher parameter value, another color (e.g., blue) to represent a lower parameter value, and intermediate transition colors to correspond to medium values; or the parameter variation can be reflected by different shades of black to white in the grayscale image, with black representing the minimum value, white representing the maximum value, and gray gradients representing intermediate values. In this way, by observing the two-dimensional spatial distribution map, the distribution characteristics of the electrical parameters on the wafer surface can be clearly identified, such as high-value areas, low-value areas, uniformity, and the presence of local anomalies. This intuitive spatial distribution representation method is of great significance for evaluating the uniformity of wafer manufacturing processes, detecting potential defect distributions, and analyzing problems in the material growth process, providing key spatial dimension information for the quality control and performance optimization of semiconductor devices.

[0042] Specifically, when evaluating the spatial uniformity of the electrical properties of the semiconductor wafer under test based on the two-dimensional spatial distribution map of electrical parameters, and identifying abnormal regions of electrical parameters on the surface of the semiconductor wafer under test, the following steps are included: Based on the two-dimensional spatial distribution diagram of electrical parameters, the overall analysis area and the sub-region of interest of the semiconductor wafer under test are determined; Statistical analysis is performed on the electrical parameter values ​​within the overall analysis area to obtain overall statistical indicators; among which, the overall statistical indicators include the overall electrical parameter mean, the overall electrical parameter standard deviation, and the overall electrical parameter variance; Calculate the regional statistical index for each sub-region of interest, and identify abnormal electrical parameter regions on the surface of the semiconductor wafer under test based on the overall statistical index, regional statistical index and preset judgment threshold. The spatial uniformity of the output is evaluated based on the abnormal electrical parameters in the region.

[0043] Specifically, determining the overall analysis region and the sub-region of interest of the semiconductor wafer under test includes: Using the boundary of the semiconductor wafer under test as the boundary, all measurement positions on the surface of the semiconductor wafer under test are set as the overall analysis area; The overall analysis area is divided into multiple non-overlapping sub-regions of interest using a rectangular grid. Each region of interest contains several measurement locations, and the coordinates of the measurement locations are determined by controlling a two-dimensional displacement platform. For each sub-region of interest, record the spatial coordinate range of its upper left and lower right corners.

[0044] Understandably, clearly defining the overall analysis region and the sub-regions of interest is fundamental to evaluating the spatial uniformity of electrical performance. Designating all measurement locations on the wafer surface as the overall analysis region allows for a macroscopic understanding of the wafer's electrical parameter distribution, ensuring a comprehensive evaluation and avoiding the neglect of overall characteristics. Dividing the overall analysis region into multiple non-overlapping sub-regions of interest using a rectangular grid facilitates refined local analysis. The regular and uniform rectangular grid division facilitates regional statistics and comparisons; each sub-region contains several measurement locations, and its statistical indicators reflect local average characteristics and fluctuations. Through two-dimensional displacement platform control, the coordinates of the measurement locations are accurately recorded, defining the spatial coordinate range of the sub-regions, making the sub-region positions and boundaries clear, and providing coordinate basis for the precise location of anomalies. This division considers both the overall and local aspects, enabling both the evaluation of overall wafer uniformity and in-depth sub-region analysis, laying a spatial framework foundation for subsequent statistics and anomaly identification.

[0045] Specifically, when identifying abnormal electrical parameter regions on the surface of a semiconductor wafer under test based on overall statistical indicators, regional statistical indicators, and preset judgment thresholds, the process includes: Calculate the deviation between the regional statistical indicators and the corresponding overall statistical indicators; whereby the regional statistical indicators include the mean, standard deviation, and variance of the regional electrical parameters. The deviation value is compared with a preset judgment threshold. When the deviation value exceeds the preset judgment threshold range, the sub-region of interest is determined to be a candidate region for abnormal electrical parameters. The electrical parameter values ​​at each measurement location within the candidate area of ​​abnormal electrical parameters are verified one by one. The absolute deviation between the electrical parameter value at a single measurement location and the overall average electrical parameter value is calculated. If the absolute deviation is greater than the preset single-point abnormal threshold, the measurement location is marked as an abnormal point. The percentage of abnormal points within the candidate region of abnormal electrical parameters is counted. When the percentage exceeds a preset threshold, the sub-region of interest is finally determined as an abnormal electrical parameter region.

[0046] Understandably, this method of identifying anomalous regions, combining regional statistical deviation with single-point anomaly verification, can consider both the overall characteristics of the region and local details. First, the deviation between regional statistical indicators and overall statistical indicators is calculated and compared with a preset judgment threshold. This quickly filters out sub-regions that significantly differ from the overall electrical performance level of the wafer as candidate anomalous regions, initially locating potential anomalous regions and improving identification efficiency. Next, each measurement position within the candidate region is verified, calculating the absolute deviation of the electrical parameter value at a single position from the overall average and comparing it with the single-point anomaly threshold. Anomalies are marked, eliminating "false anomaly regions" and improving identification accuracy. Finally, the proportion of anomalies within the candidate region is statistically analyzed. Only when the proportion exceeds a preset threshold is the sub-region determined to be an electrical parameter anomalous region. This considers that true anomalous regions usually contain multiple consecutive or densely packed anomalies, ensuring the reliability and robustness of the judgment and avoiding misjudgments. This multi-level, progressive identification logic can accurately locate electrical parameter anomalous regions on the semiconductor wafer surface, providing precise target areas for subsequent analysis of anomaly causes and optimization of manufacturing processes.

[0047] In this embodiment, carrier mobility is the preferred electrical parameter. It is a core electrical parameter of semiconductor materials, directly determining the transport capability of carriers under an electric field, and playing a decisive role in key performance indicators such as switching speed and conduction current of semiconductor devices. In contrast, conductivity is affected by both carrier concentration and mobility, and defect-related recombination rate reflects the defect state of the material more directly, while mobility more directly characterizes the transport properties and lattice quality of the semiconductor material itself. Focusing on carrier mobility allows for a more accurate assessment of the differences in carrier transport performance at different spatial locations on a semiconductor wafer. Its two-dimensional spatial distribution map can intuitively reflect the uniformity of mobility on the wafer surface, which is valuable for identifying local performance degradation regions caused by uneven doping, lattice distortion, stress distribution, or process defects. This provides a more direct and crucial basis for wafer-level quality control and process optimization.

[0048] Specifically, when evaluating the spatial uniformity of output based on the abnormal electrical parameter regions, the following is included: The percentage of all abnormal electrical parameter areas relative to the total surface area of ​​the semiconductor wafer under test is statistically analyzed and used as a quantitative evaluation index of spatial uniformity. The quantitative evaluation index is output as the spatial uniformity evaluation result.

[0049] Understandably, using the percentage of abnormal electrical parameter regions relative to the total wafer surface area as a quantitative evaluation metric for spatial uniformity transforms abstract distribution characteristics into intuitive numerical values, facilitating performance comparisons between different wafers or different batches of the same wafer. For example, if the percentage of abnormal regions on one wafer is 0.5%, while on another wafer it is 5%, it can be directly concluded that the former has better spatial uniformity than the latter. This quantitative metric provides a quantifiable basis for wafer quality grading and process stability monitoring, making the evaluation results more objective and comparable. Furthermore, outputting this quantitative evaluation metric as the spatial uniformity assessment result allows users to quickly and clearly grasp the overall uniformity of wafer electrical performance, providing concise and crucial reference information for subsequent decisions such as process improvement, device design adaptation, or wafer selection.

[0050] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program goods. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program goods on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0051] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program goods according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0052] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0053] These computer program instructions can also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0054] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A spatially resolved non-destructive testing method for electrical parameters of semiconductor wafers based on optically pumped-terahertz detection spectroscopy, characterized in that, include: S100: Uses an ultrafast laser to generate pulsed light, and splits the pulsed light into pump light and terahertz probe light through a beam splitter module; S200: Performs optical path adjustment on the pump light and terahertz probe light, and after optical path adjustment, focuses the pump light and terahertz probe light onto the same measurement position on the surface of the semiconductor wafer under test; S300: By adjusting the time delay in the terahertz probe light path, the terahertz probe light is irradiated onto the surface of the semiconductor wafer under test at different delay times, and the corresponding terahertz probe light signal intensity is collected, thereby obtaining the terahertz probe light signal sequence of the measurement position as a function of time delay, so as to reconstruct the carrier relaxation dynamics curve of the measurement position. S400: Based on a preset semiconductor material dynamics analysis model, the carrier relaxation dynamics curve is fitted to obtain the dynamic parameters and transient conductance changes related to the wafer carrier recombination process. S500: Using dynamic parameters and transient conductance changes as inputs, it inversely solves the dynamic analysis model of semiconductor materials to infer the electrical parameters of the semiconductor wafer under test. S600: Fix the semiconductor wafer under test on a programmable two-dimensional displacement platform, move the position of the semiconductor wafer under test by controlling the displacement platform, and repeat steps S200 to S500 to obtain the electrical parameters of the semiconductor wafer under test at different measurement positions, and generate a two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test. S700: Evaluate the spatial uniformity of the electrical properties of the semiconductor wafer under test based on the two-dimensional spatial distribution map of electrical parameters, and identify abnormal regions of electrical parameters on the surface of the semiconductor wafer under test.

2. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 1, characterized in that, When adjusting the optical path of the pump light and the terahertz probe light, the following steps are included: An adjustable time delay device is set in the path of the terahertz probe light to control the relative arrival time of the pump light and the terahertz probe light.

3. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 2, characterized in that, When acquiring the terahertz probe light signal sequence at the measurement location, which varies with time delay, to reconstruct the carrier relaxation dynamics curve at that measurement location, the process includes: By adjusting the optical delay device set in the terahertz probe light path, the terahertz probe light is made to illuminate the surface of the semiconductor wafer under test at different delay times. At each delay time, the corresponding terahertz probe light signal intensity is collected to form a terahertz probe light signal sequence that varies with time delay; The acquired terahertz probe optical signal sequence was denoised and normalized. Based on the physical model of pump light-induced photogenerated carriers, the terahertz probe light signal sequence is mapped as a function of carrier concentration or transient conductance as a function of time. The function of carrier concentration or transient conductance as a function of time obtained by mapping is formed into a continuous curve, which is the reconstructed carrier relaxation dynamics curve.

4. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped terahertz detection spectroscopy according to claim 3, characterized in that, The kinetic parameters include carrier lifetime and relaxation time constant.

5. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 4, characterized in that, Electrical parameters include mobility, conductivity, and defect-related recombination rate.

6. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 5, characterized in that, The two-dimensional spatial distribution map of the electrical parameters of the semiconductor wafer under test is a two-dimensional image with the coordinates of different measurement positions as the horizontal and vertical axes, and the electrical parameter values ​​at the corresponding positions as color or grayscale values.

7. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 6, characterized in that, When evaluating the spatial uniformity of the electrical properties of a semiconductor wafer under test based on a two-dimensional spatial distribution map of electrical parameters, and identifying anomalous regions of electrical parameters on the surface of the semiconductor wafer under test, the following steps are taken: Based on the two-dimensional spatial distribution diagram of electrical parameters, the overall analysis area and the sub-region of interest of the semiconductor wafer under test are determined; Statistical analysis is performed on the electrical parameter values ​​within the overall analysis area to obtain overall statistical indicators; among which, the overall statistical indicators include the overall electrical parameter mean, the overall electrical parameter standard deviation, and the overall electrical parameter variance; Calculate the regional statistical index for each sub-region of interest, and identify abnormal electrical parameter regions on the surface of the semiconductor wafer under test based on the overall statistical index, regional statistical index and preset judgment threshold. The spatial uniformity of the output is evaluated based on the abnormal electrical parameters in the region.

8. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 7, characterized in that, When determining the overall analysis region and the sub-region of interest of the semiconductor wafer under test, the following are included: Using the boundary of the semiconductor wafer under test as the boundary, all measurement positions on the surface of the semiconductor wafer under test are set as the overall analysis area; The overall analysis area is divided into multiple non-overlapping sub-regions of interest using a rectangular grid. Each region of interest contains several measurement locations, and the coordinates of the measurement locations are determined by controlling a two-dimensional displacement platform. For each sub-region of interest, record the spatial coordinate range of its upper left and lower right corners.

9. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 8, characterized in that, When identifying abnormal electrical parameter regions on the surface of a semiconductor wafer under test based on overall statistical indicators, regional statistical indicators, and preset judgment thresholds, the following methods are used: Calculate the deviation between the regional statistical indicators and the corresponding overall statistical indicators; whereby the regional statistical indicators include the mean, standard deviation, and variance of the regional electrical parameters. The deviation value is compared with a preset judgment threshold. When the deviation value exceeds the preset judgment threshold range, the sub-region of interest is determined to be a candidate region for abnormal electrical parameters. The electrical parameter values ​​at each measurement location within the candidate area of ​​abnormal electrical parameters are verified one by one. The absolute deviation between the electrical parameter value at a single measurement location and the overall average electrical parameter value is calculated. If the absolute deviation is greater than the preset single-point abnormal threshold, the measurement location is marked as an abnormal point. The percentage of abnormal points within the candidate region of abnormal electrical parameters is counted. When the percentage exceeds a preset threshold, the sub-region of interest is finally determined as an abnormal electrical parameter region.

10. The spatially resolved non-destructive testing method for semiconductor wafer electrical parameters based on optically pumped-terahertz detection spectroscopy according to claim 9, characterized in that, When evaluating the spatial uniformity of output based on the abnormal electrical parameter regions, the following should be included: The percentage of all abnormal electrical parameter areas relative to the total surface area of ​​the semiconductor wafer under test is statistically analyzed and used as a quantitative evaluation index of spatial uniformity. The quantitative evaluation index is output as the spatial uniformity evaluation result.

Images