Parameter extraction and alignment processing method and system for DLCP test data

Abstract

The application provides a parameter extraction and alignment processing method and system for DLCP test data, which comprises the following steps: DLCP tests are carried out at multiple different test frequencies to obtain the capacitance bias curve data corresponding to each test frequency; the validity of the capacitance bias curve data is determined, and each curve data determined as valid is subjected to function fitting, and the capacitance parameters corresponding to each test frequency are extracted from the fitting results; based on the capacitance parameters, the spatial position parameters corresponding to each test frequency are calculated through a first model; the spatial position parameters corresponding to a preset high frequency are taken as a reference for alignment processing, and a multi-frequency data set in a unified depletion region width coordinate system is output; the trap density distribution is obtained by performing multi-frequency capacitance data difference calculation, and trap density spatial distribution data is output. Through the spatial alignment processing of the capacitance bias data at different test frequencies based on a unified physical model, the accuracy and stability of the test results are improved.

Description

Technical Field

[0001] This invention relates to the field of semiconductor electrical characterization and defect analysis technology, and in particular to a method and system for parameter extraction and alignment processing of DLCP test data. Background Technology

[0002] DLCP (Driving-Level Capacitance Profiling) is a semiconductor defect analysis method based on capacitance-voltage (CV) measurements. It analyzes the spatial distribution of charge carriers and traps within the depletion region of a semiconductor device by measuring the relationship between capacitance and bias voltage under different AC drive levels or equivalent test conditions, and combining this with differential calculations of multiple sets of test data. Existing DLCP techniques typically assume that the same bias voltage conditions correspond to the same depletion region width at different test frequencies, and calculate the trap density based on this multi-frequency capacitance differential calculation.

[0003] However, in actual testing, due to differences in frequency response, noise interference, and testing non-idealities, the capacitance-voltage curves obtained at different frequencies exhibit significant inconsistencies in parameter fitting and depletion region width calculation. This leads to: difficulties in directly aligning the capacitance curves corresponding to each driving amplitude on the bias axis or physical distance axis; systematic shifts in the equivalent parameters obtained from curve fitting under different driving conditions; and the easy introduction of spatial coordinate errors and numerical instability when directly calculating carrier or trap densities based on raw data. Furthermore, there is a lack of a unified and reproducible data processing workflow between test results of different frequencies and driving amplitudes. The fitting process is highly sensitive to outlier data points, requiring extensive manual intervention and exhibiting poor repeatability.

[0004] Therefore, there is an urgent need for a processing method that can robustly extract parameters from DLCP capacitance data and achieve unified alignment of multi-frequency data in physical space, so as to improve the accuracy and consistency of trap density calculation. Summary of the Invention

[0005] The purpose of this invention is to solve the problems of difficulty in aligning multi-frequency DLCP data and unstable differential results in the prior art.

[0006] A first aspect of the present invention provides a method for parameter extraction and alignment processing of DLCP test data, comprising:

[0007] DLCP tests were performed at multiple different test frequencies to obtain capacitor bias curve data corresponding to each test frequency.

[0008] The validity of the capacitor bias curve data is determined, and each curve data determined to be valid is fitted with a function. The capacitor parameters corresponding to each test frequency are extracted from the fitting results. The capacitor parameters include at least one of the zero-order equivalent capacitor parameters and derivative parameters or equivalent higher-order parameters that characterize the trend of capacitor change with bias voltage.

[0009] Based on the capacitor parameters, the spatial position parameters corresponding to each test frequency are calculated using the first model.

[0010] Alignment is performed using the spatial location parameters corresponding to the preset high frequencies as a reference, and a multi-frequency DLCP dataset under a unified depletion region width coordinate system is output.

[0011] The trap density distribution is obtained by performing differential calculation on multi-frequency capacitor data, and the spatial distribution data of trap density is output.

[0012] Furthermore, the validity determination includes at least:

[0013] Remove capacitance data points that are non-positive or have abnormal amplitudes;

[0014] Determine whether the number of valid capacitance data points meets the preset minimum number of fitting points requirement;

[0015] Missing or invalid data points are marked and will not be included in subsequent parameter fitting steps.

[0016] Furthermore, the fitting is performed using one of the following methods: polynomial fitting, spline fitting, or other continuous function fitting.

[0017] Furthermore, a goodness-of-fit threshold is introduced during the fitting process to discard data results that do not meet the fitting conditions.

[0018] Furthermore, boundary constraints are introduced at the beginning or end of the data points during the fitting process to improve the stability of parameter extraction in the low or high bias regions.

[0019] Furthermore, the first model includes a capacitance-depletion width physical model based on semiconductor depletion region theory, which is used to calculate the corresponding spatial location parameters based on the capacitance parameters.

[0020] Furthermore, the spatial location parameters include the depletion region width or profile depth calculated based on the capacitance-depletion width physical model, wherein the depletion region width or the profile depth is used to characterize the physical spatial distribution location of traps or charge carriers inside the device.

[0021] Furthermore, in the alignment process, the spatial position parameters corresponding to each test frequency are remapped to a unified depletion region width coordinate system using the first model.

[0022] Furthermore, the preset high frequency is the highest test frequency among the multiple test frequencies, or a pre-set high frequency test frequency that meets the physical authenticity threshold.

[0023] Furthermore, the spatial distribution data of the trap density is subjected to moving average or gradient smoothing to suppress the influence of noise on the calculation results.

[0024] A second aspect of the present invention provides a parameter extraction and alignment processing system for DLCP test data, employing the parameter extraction and alignment processing method for DLCP test data as described in any of the preceding claims, comprising:

[0025] The capacitor test unit is used to perform DLCP tests at multiple different test frequencies and obtain capacitor bias curve data corresponding to each test frequency.

[0026] A data validity determination unit is used to determine the validity of the capacitor bias curve data.

[0027] The capacitance parameter extraction unit is used to perform function fitting on each curve data that is determined to be valid, and extract the capacitance parameters corresponding to each test frequency from the fitting results.

[0028] The spatial parameter calculation unit calculates the spatial position parameters corresponding to each test frequency based on the capacitor parameters and through the first model.

[0029] The spatial alignment processing unit is used to perform alignment processing with reference to the spatial position parameters corresponding to the preset high frequency, and outputs a multi-frequency DLCP dataset under a unified depletion region width coordinate system.

[0030] The trap density calculation output unit is used to perform differential calculation of multi-frequency capacitor data to obtain the trap density distribution and output the trap density spatial distribution data.

[0031] Compared with existing technologies, this invention has at least the following beneficial effects: By introducing an automated data validity judgment and function fitting mechanism, outliers and noise interference in the original test data are effectively eliminated, reducing the need for manual intervention in the parameter extraction process and improving the efficiency and repeatability of data processing; At the same time, by adopting a unified first model throughout the spatial parameter calculation and multi-frequency data pairing process, and using the preset high-frequency data with the strongest physical realism as a benchmark for spatial coordinate mapping, the problem of difficulty in aligning multi-frequency capacitance curves and systematic spatial coordinate offset caused by frequency response differences in existing technologies is solved, achieving accurate unification of multi-frequency DLCP data under a unified depletion region width coordinate system; Finally, based on the aligned high-precision dataset, trap density difference calculation is performed, effectively suppressing numerical instability and improving the accuracy and reliability of semiconductor device trap density distribution test results. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention 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 embodiments of the present invention. For those skilled in the art, other drawings can be obtained as provided without creative effort.

[0033] Figure 1 This is a schematic diagram of the steps in a parameter extraction and alignment processing method for DLCP test data according to an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of a module for a parameter extraction and alignment processing system for DLCP test data in one embodiment of the present invention;

[0035] Figure 3 This is a schematic diagram comparing the depletion region width of the DLCP capacitor parameters before and after alignment processing at different test frequencies in one embodiment of the present invention. Detailed Implementation

[0036] The present invention will now be described in more detail with reference to the accompanying drawings, which illustrate preferred embodiments of the invention. It should be understood that those skilled in the art can modify the invention described herein while still achieving its advantageous effects. Therefore, the following description should be understood as being broadly known to those skilled in the art and is not intended to limit the invention.

[0037] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0038] The invention is described more specifically by way of example in the following paragraphs with reference to the accompanying drawings. The advantages and features of the invention will become clearer as explained below. It should be noted that the drawings are in a very simplified form and use non-precise proportions, and are only used to facilitate and clarify the illustration of the embodiments of the invention.

[0039] Example 1

[0040] This embodiment provides a method for parameter extraction and alignment processing of DLCP test data. Please refer to [link / reference]. Figure 1 ,include:

[0041] DLCP tests were performed at multiple different test frequencies to obtain capacitor bias curve data corresponding to each test frequency.

[0042] The validity of the capacitor bias curve data is determined, and each curve data determined to be valid is fitted with a function. The capacitor parameters corresponding to each test frequency are extracted from the fitting results. The capacitor parameters include at least one of the following: zero-order equivalent capacitor parameters, derivative parameters or equivalent higher-order parameters that characterize the trend of capacitor change with bias voltage.

[0043] Based on the capacitor parameters, the spatial position parameters corresponding to each test frequency are calculated using the first model.

[0044] Alignment is performed using the spatial location parameters corresponding to preset high frequencies as a reference, and a multi-frequency DLCP dataset under a unified depletion region width coordinate system is output.

[0045] The trap density distribution is obtained by performing differential calculation on multi-frequency capacitor data, and the spatial distribution data of trap density is output.

[0046] Specifically, DLCP is a method that obtains capacitor-bias curves under different test frequencies or equivalent driving conditions, maps the data under different test conditions to a unified depletion region spatial coordinate through a unified physical model, and obtains the spatial distribution of trap density through multi-frequency capacitor differential calculation.

[0047] Regarding the parameter extraction and alignment method for DLCP test data in this embodiment, DLCP testing is first performed on the semiconductor device under test at multiple different test frequencies. An automated test platform is programmed to control the tester to collect and automatically store the capacitance bias curve data corresponding to each frequency. Subsequently, the collected data undergoes automated validity assessment. An outlier is automatically removed and missing data is marked using a preset algorithm. After selecting valid curve data, continuous function fitting (polynomial, spline, etc.) is performed on each curve. The core capacitance parameters corresponding to each frequency are extracted from the fitting results, thus avoiding subjective errors caused by manual intervention throughout the process.

[0048] In this embodiment, the capacitance parameters include at least a zero-order equivalent capacitance parameter and one of a derivative-type parameter or an equivalent higher-order parameter characterizing the capacitance change trend with bias voltage. The zero-order equivalent capacitance parameter is used to characterize the quasi-static capacitance characteristics under the corresponding bias voltage condition; the derivative-type parameter or equivalent higher-order parameter characterizing the capacitance change trend with bias voltage is used to characterize the capacitance change trend. Preferably, multi-frequency capacitance measurement can be performed at each bias voltage point according to a preset bias voltage scanning sequence.

[0049] Building upon this foundation, based on the capacitance parameters, the spatial location parameters corresponding to each frequency are calculated using the first model. A unified model ensures that the calculation paths for different frequencies align with the physical assumptions, achieving a precise mapping from electrical parameters to the physical spatial dimension. Furthermore, using the preset spatial location parameters corresponding to high frequencies as a benchmark, alignment is achieved through reverse mapping using the first model. This precisely maps all frequency data to a unified depletion region width coordinate system, outputting a multi-frequency DLCP dataset in a unified coordinate system, thus resolving the multi-frequency data misalignment problem. Finally, multi-frequency capacitance differential calculations are performed on the aligned dataset. Trap density is derived using semiconductor defect theory, and noise is suppressed through smoothing, ultimately outputting spatial data that accurately characterizes the trap distribution within the device.

[0050] By implementing automated verification, unified model calculation, and high-frequency benchmark alignment, the problems of difficult alignment of multi-frequency curves, parameter offset, and coordinate error are effectively improved, enhancing the consistency and accuracy of DLCP test data and providing reliable support for semiconductor defect analysis.

[0051] In one possible embodiment of the invention, the plurality of different test frequencies are at least two different AC excitation frequencies.

[0052] Specifically, the AC excitation frequency refers to the frequency of the small AC signal applied to the semiconductor device during DLCP testing. DLCP testing characterizes deep-level defects by measuring the capacitance response of the semiconductor device at different frequencies and bias voltages. The AC excitation frequency is one of the key parameters in DLCP testing, determining whether traps with different response times within the device can be effectively detected. Using at least two different AC excitation frequencies ensures that the capacitance characteristics of the semiconductor device at different frequency responses can be captured. Different traps have different capture and emission time constants, which are closely related to the test frequency. By using at least two different frequencies, traps that may not respond at one frequency can respond at another, providing more comprehensive information. This frequency combination yields broader trap response information, providing richer and more accurate raw data for subsequent parameter extraction and trap density calculation.

[0053] Furthermore, the validity determination includes at least:

[0054] Remove capacitance data points that are not positive or have abnormal amplitude.

[0055] Determine whether the number of valid capacitance data points meets the preset minimum number of fitted points requirement.

[0056] Missing or invalid data points are marked and will not be included in subsequent parameter fitting steps.

[0057] Specifically, capacitance data should physically be positive; non-positive data usually indicates measurement error or equipment malfunction. Abnormal capacitance data points refer to values ​​that significantly deviate from the normal range, such as far exceeding the device's expected maximum or minimum capacitance value, or exhibiting drastic fluctuations within a localized area. These anomalies may be caused by transient noise, poor contact, or unstable test conditions. In practice, non-positive data can be eliminated by setting physical thresholds (e.g., capacitance must be greater than zero). Alternatively, reasonable upper and lower limits for capacitance can be set based on the device's physical characteristics; data points exceeding these limits are considered abnormal.

[0058] The accuracy of function fitting largely depends on the number of data points involved in the fitting process. If there are too few valid data points, the fitted curve may not accurately reflect the true physical characteristics of the capacitor bias curve, and may even cause the fitting algorithm to fail to converge or produce unreasonable fitting results. The preset minimum number of fitting points is an empirical value or a threshold set according to the characteristics of the fitting algorithm, designed to ensure that the fitting process has sufficient data support. Before fitting, the system counts the number of valid data points after preliminary screening (such as removing non-positive values ​​and abnormal amplitude points) and compares it with the preset minimum number of fitting points. If the requirement is not met, the curve data is considered unsuitable for fitting, or further data processing measures are required.

[0059] Labeling refers to identifying data points that are judged to be non-positive, have abnormal amplitudes, or are missing due to other reasons (such as data acquisition interruption), so that they can be identified and distinguished in subsequent data processing. By excluding these data points from the fitting process, their negative impact on the fitting results can be avoided, thereby ensuring the accuracy of the fitted curve and the reliability of parameter extraction.

[0060] Furthermore, the fitting is performed using one of the following methods: polynomial fitting, spline fitting, or other continuous function fitting.

[0061] By introducing polynomial fitting, spline fitting, or other continuous function fitting methods, the most suitable fitting algorithm can be flexibly selected based on the specific characteristics and complexity of the DLCP test data. This allows the fitting process to more accurately capture the physical behavior and local details of the capacitor bias curve, effectively reducing fitting errors and improving the accuracy and stability of extracting capacitance parameters from the fitting results. When performing function fitting on the capacitor bias curve data obtained from DLCP tests to extract capacitance parameters, if the fitting results cannot reflect the true physical characteristics or lack accuracy, it will lead to large errors in subsequent calculations of spatial location parameters and trap density, affecting the reliability of data processing.

[0062] To address this, in one possible embodiment of the invention, two optimizations are implemented to improve the fitting quality: First, a goodness-of-fit threshold is introduced. After fitting with continuous functions such as polynomials and splines, indicators such as the coefficient of determination and root mean square error are calculated and compared with a preset threshold. Results that do not meet the goodness-of-fit standard and their corresponding capacitance parameters are discarded to avoid error propagation. Second, boundary constraints are introduced for the first and last data points during the fitting process to solve the problem of fitting deviating from actual physical behavior due to noise and data sparsity in such regions. The boundary constraints are set based on the physical characteristics of the device, with the first point corresponding to the lowest bias voltage region and the last point corresponding to the highest bias voltage region.

[0063] Furthermore, the first model includes a capacitance-depletion width physical model based on semiconductor depletion region theory, which is used to calculate the corresponding spatial location parameters based on the capacitance parameters.

[0064] Specifically, the capacitance-depletion width physical model is a theoretical model describing the electrical characteristics of the depletion layer in a semiconductor device. Based on semiconductor physics principles, this model correlates the device's capacitance with parameters such as the width of the depletion layer, doping concentration, and applied bias voltage. The capacitance-depletion width physical model is used to calculate the corresponding spatial location parameters based on these capacitance parameters. Using a unified model means that the mathematical algorithms and physical principles used to convert capacitance parameters into spatial location parameters are consistent when processing capacitance data at different test frequencies. For example, regardless of whether the test frequency is 1kHz, 10kHz, or 100kHz, the same capacitance-depletion width physical model is used to calculate the corresponding depletion layer width or space charge distribution.

[0065] Furthermore, the spatial location parameters include the depletion region width or profiling distance calculated based on the capacitance-depletion width physical model. The depletion region width or the profiling distance is used to characterize the physical spatial distribution of traps or charge carriers inside the device.

[0066] The depletion region width is the physical thickness of the region without free carriers formed by the space charge region in a reverse-biased PN junction or Schottky junction. The profiling distance is the coordinate representation of the depletion region width in the profiling analysis, used as the depth coordinate axis for the trap density distribution.

[0067] Furthermore, in the alignment process, the spatial position parameters corresponding to each test frequency are remapped to a unified depletion region width coordinate system using the first model.

[0068] Furthermore, the preset high frequency is the highest test frequency among the multiple test frequencies, or a pre-set high frequency test frequency that meets the physical authenticity threshold.

[0069] Specifically, alignment aims to unify the spatial location parameters acquired at different test frequencies into a common depletion region width coordinate system. Since the response depth and detection range of semiconductor materials differ at different frequencies during DLCP testing, directly comparing or superimposing these data may lead to inconsistencies in spatial location. Therefore, alignment is a crucial step in ensuring the accuracy of subsequent data analysis (such as trap density calculation). It adjusts or transforms the spatial location parameters at each frequency to ensure they physically correspond to the same spatial depth. "Re-applying the first model" refers to re-applying the first model for refinement during the initial calculation and alignment of the spatial location parameters for each test frequency. Based on the inherent logic of the physical model, the spatial location parameters are recalculated or calibrated to ensure they conform to the physical laws of semiconductors.

[0070] Furthermore, the spatial distribution data of the trap density is subjected to moving average or gradient smoothing to suppress the influence of noise on the calculation results.

[0071] Specifically, moving average is a commonly used data smoothing technique. Its principle is to replace the value of a given point with the average of the values ​​of its neighbors. When processing spatial distribution data of trap density, a suitable window size can be selected to perform a weighted average of the trap density value at each spatial location and the trap density values ​​of its neighbors, thereby effectively filtering out local random fluctuations. Gradient smoothing, on the other hand, is a method based on data gradient information. While smoothing noise, it can better preserve edge or abrupt changes in the data.

[0072] Please refer to Figure 3 Before alignment, the capacitance (C) curves at different test frequencies were intertwined due to differences in frequency response, resulting in a chaotic state where they could not directly correspond to the same physical location. After alignment, the capacitance curves of all frequencies (10kHz, 100kHz, 500kHz) can be clearly displayed, and all are accurately mapped to a unified depletion region width coordinate system with the 500kHz high-frequency data, which has the strongest physical authenticity, as the reference benchmark. This achieves an orderly arrangement of multi-frequency data under a unified physical dimension.

[0073] Example 2

[0074] This embodiment provides a parameter extraction and alignment system for DLCP test data, employing the parameter extraction and alignment method for DLCP test data as described in Embodiment 1. Please refer to [link / reference]. Figure 2 ,include:

[0075] The capacitor test unit is used to perform DLCP tests at multiple different test frequencies and obtain the capacitor bias curve data corresponding to each test frequency.

[0076] The data validity determination unit is used to determine the validity of the capacitor bias curve data.

[0077] The capacitance parameter extraction unit is used to perform function fitting on each curve data that is determined to be valid, and extract the capacitance parameters corresponding to each test frequency from the fitting results.

[0078] The spatial parameter calculation unit calculates the spatial position parameters corresponding to each test frequency based on the capacitor parameters and using the first model.

[0079] The spatial alignment processing unit is used to perform alignment processing with reference to the spatial position parameters corresponding to the preset high frequency, and outputs a multi-frequency DLCP dataset under a unified depletion region width coordinate system.

[0080] The trap density calculation output unit is used to perform differential calculation of multi-frequency capacitor data to obtain the trap density distribution and output the trap density spatial distribution data.

[0081] Specifically, the capacitance testing unit includes an AC excitation signal source, a DC bias source, and a capacitance measurement module, used to measure the capacitance of the semiconductor device under test under different AC excitation frequencies. Preferably, the capacitance testing unit can perform multi-frequency capacitance measurement at each bias point according to a preset bias scanning sequence. The data validity determination unit is used to preprocess the acquired capacitance data. The capacitance parameter extraction unit is used to perform function fitting operations on the capacitance-bias data at each test frequency to extract the corresponding capacitance parameters. The spatial parameter calculation unit calculates the depletion region width or profile depth corresponding to the capacitance parameters based on the capacitance parameters and in conjunction with a first model (i.e., a unified capacitance-depletion width physical model).

[0082] At different test frequencies, the spatial parameter calculation unit uses the same physical model and calculation path to ensure the physical consistency of the spatial parameters. The spatial alignment processing unit uses a preset high-frequency test result as a reference and, through a unified physical model, maps the depletion region width or profile depth calculated at different test frequencies to a unified depletion region width coordinate system, achieving spatial alignment of multi-frequency data. The trap density calculation and output unit performs differential operations on the capacitance parameters at different test frequencies based on the spatially aligned data, outputting the trap density distribution results at the corresponding physical locations.

[0083] Through modular functional unit division of labor and cooperation, a complete DLCP test data processing closed loop is formed, from automated test acquisition, data verification and fitting, spatial coordinate mapping, to the final output of accurate trap density distribution. This can efficiently solve problems such as multi-frequency data misalignment and error accumulation, and improve the accuracy and repeatability of semiconductor defect analysis.

[0084] The above examples illustrate the present invention only to aid in understanding it and are not intended to limit the scope of the invention. Those skilled in the art can make various simple deductions, modifications, or substitutions based on the principles of this invention.

Claims

1. A method for parameter extraction and alignment processing of DLCP test data, characterized in that, include: DLCP tests were performed at multiple different test frequencies to obtain capacitor bias curve data corresponding to each test frequency. The validity of the capacitor bias curve data is determined, and each curve data determined to be valid is fitted with a function. The capacitor parameters corresponding to each test frequency are extracted from the fitting results. The capacitor parameters include the zero-order equivalent capacitor parameter and one of the derivative parameters or equivalent higher-order parameters that characterize the trend of capacitor change with bias voltage. Based on the capacitor parameters, the spatial position parameters corresponding to each test frequency are calculated using the first model. Alignment is performed using the spatial location parameters corresponding to the preset high frequencies as a reference, and a multi-frequency DLCP dataset under a unified depletion region width coordinate system is output. The trap density distribution is obtained by performing differential calculation on multi-frequency capacitor data, and the spatial distribution data of trap density is output.

2. The parameter extraction and alignment method for DLCP test data as described in claim 1, characterized in that, The validity determination includes at least: Remove capacitance data points that are non-positive or have abnormal amplitudes; Determine whether the number of valid capacitance data points meets the preset minimum number of fitting points requirement; Missing or invalid data points are marked and will not be included in subsequent parameter fitting steps.

3. The parameter extraction and alignment method for DLCP test data as described in claim 1, characterized in that, The fitting process employs one of the following methods: polynomial fitting, spline fitting, or other continuous function fitting.

4. The parameter extraction and alignment method for DLCP test data as described in claim 3, characterized in that, A goodness-of-fit threshold is introduced during the fitting process to discard data results that do not meet the fitting conditions.

5. The parameter extraction and alignment method for DLCP test data as described in claim 4, characterized in that, Boundary constraints are introduced at the beginning or end of the data points during the fitting process to improve the stability of parameter extraction in the low or high bias regions.

6. The parameter extraction and alignment method for DLCP test data as described in claim 1, characterized in that, The first model includes a capacitance-depletion width physical model based on semiconductor depletion region theory, which is used to calculate the corresponding spatial location parameters based on the capacitance parameters.

7. The parameter extraction and alignment method for DLCP test data as described in claim 6, characterized in that, The spatial location parameters include the depletion region width or profile depth calculated based on the capacitance-depletion width physical model. The depletion region width or the profile depth is used to characterize the physical spatial distribution of traps or charge carriers inside the device.

8. The parameter extraction and alignment method for DLCP test data as described in claim 1, characterized in that, In the alignment process, the spatial position parameters corresponding to each test frequency are remapped to a unified depletion region width coordinate system using the first model.

9. The parameter extraction and alignment method for DLCP test data as described in claim 1, characterized in that, The preset high frequency is the highest test frequency among the multiple test frequencies, or a pre-set high frequency test frequency that meets the physical authenticity threshold.

10. A system for parameter extraction and alignment processing of DLCP test data, employing the parameter extraction and alignment processing method for DLCP test data as described in any one of claims 1-9, characterized in that, include: The capacitor test unit is used to perform DLCP tests at multiple different test frequencies and obtain capacitor bias curve data corresponding to each test frequency. A data validity determination unit is used to determine the validity of the capacitor bias curve data. The capacitance parameter extraction unit is used to perform function fitting on each curve data that is determined to be valid, and extract the capacitance parameters corresponding to each test frequency from the fitting results. The spatial parameter calculation unit calculates the spatial position parameters corresponding to each test frequency based on the capacitor parameters and through the first model. The spatial alignment processing unit is used to perform alignment processing with reference to the spatial position parameters corresponding to the preset high frequency, and outputs a multi-frequency DLCP dataset under a unified depletion region width coordinate system. The trap density calculation output unit is used to perform differential calculation of multi-frequency capacitor data to obtain the trap density distribution and output the trap density spatial distribution data.

Images