Method and apparatus for surveying semiconductor multilayer structures based on second harmonic waves

Abstract

The present invention provides a surveying method and apparatus based on second harmonic for surveying the entire area of ​​a wafer. The surveying method includes three forms: fixed point surveying, scanning surveying, and a combination of fixed point surveying and scanning surveying. The scanning surveying method can survey the entire area of ​​a wafer, obtain the position, size, and relative density distribution of electrical defects, and realize the location and investigation of wafer abnormal points on the premise of ensuring high efficiency of the surveying, which cannot be achieved by the current second harmonic surveying technology. The present invention further provides a new set of equations describing the second harmonic, which can unify the actual surveying results and the theoretical model in the three cases of fixed point surveying, scanning surveying, and the combination of fixed point surveying and scanning surveying, so that the second harmonic surveying technology can be applied not only as a qualitative analysis method but also as a quantitative analysis method, which can improve the completeness and accuracy of product inspection on the premise of ensuring efficiency, and improve the quality checking ability in advanced semiconductor processes.

Description

[Technical Field]

[0001] The present invention relates to measuring semiconductor device manufacturing and processing processes, and more particularly to measuring or monitoring the interface characteristics or oxide layer quality of semiconductor wafers using second harmonics. [Background technology]

[0002] When processing semiconductor wafers, it is always important to inspect the quality of the interface between the oxide layer and the semiconductor layer, because the interface states and volume charges in the oxide layer at this interface can capture and release carriers, causing unstable threshold voltages and electrical leakage in devices, which can significantly reduce the performance and service life of the devices.

[0003] Currently, the main methods commonly used in this field to quantify electrical properties include electrical conductivity measurement and capacitance measurement. The principle is that electrical properties affect the flatband voltage of a device, but the measurement results may deviate from the theoretical CV or IV curve due to different methods or mechanisms. Therefore, by applying a bias voltage to the sample under test and measuring the resulting current or capacitance signal, the degree of deviation between the actual measurement results and the theoretical results can be analyzed to quantitatively calculate the electrical properties of the sample under test. However, these methods suffer from low efficiency and low resolution, and can even result in wafer destruction. Therefore, they are gradually failing to meet the requirements of advanced processes, such as high efficiency, high resolution, and low material loss.

[0004] Second harmonics are a nonlinear effect that allows a material to generate light with a frequency twice that of incident light under certain conditions. Several current studies have shown that second harmonics are relatively sensitive for detecting the quality of surfaces and interfaces where centrosymmetricity is broken in intrinsic materials, making them particularly valuable as a qualitative analysis method, particularly in the field of quality testing for silicon semiconductors. Furthermore, the measurement process is extremely simple; second harmonics can be received by simply irradiating a photon at the measurement point. However, the value of second harmonics technology in the actual production and processing of wafers is currently very limited, for three main reasons: (1) In conventional surveying techniques, it is difficult to unify the actual survey results with the theoretical model. The main reason why this technique is only a qualitative analysis method and not a quantitative analysis method is that it is unable to decouple the various factors that generate second harmonics. (2) According to the conventional technology or theoretical model, it is not possible to survey the entire area of ​​the wafer on the premise of ensuring high survey efficiency. (3) Conventional second harmonic detection devices lack precision, especially in measuring the initial value, which is one of the most important parameters. Therefore, when analyzing sample parameters using this value, errors will occur in the results. Summary of the Invention [Problem to be solved by the invention]

[0005] The problems that the present invention aims to solve are as follows. (1) Conventional surveying techniques, on the premise of ensuring high surveying efficiency, cannot survey the entire area of ​​a wafer, and cannot locate and investigate abnormal points on the wafer. (2) With conventional surveying techniques, it is difficult to unify actual survey results with theoretical models, and only qualitative analysis of survey results is possible, not quantitative analysis. (3) Conventional surveying techniques lack the precision required to measure initial values, resulting in errors in the analysis results.

[0006] The current state of technology for using second harmonic waves to inspect wafer defects is as follows: Fixed-point measurement has been given some attention over the past 20 years. Scientific research institutions have primarily focused on theoretical modeling and application expansion of second harmonic wave wafer inspection. However, for industrial applications, this technology requires relatively high defect analysis accuracy and defect location capability. This is the main reason why second harmonic wafer inspection technology has not yet been adopted industrially. In light of this, this application provides for the first time a second harmonic scanning technology that can improve the accuracy of second harmonic wave measurement and achieve defect distribution analysis that is not possible with fixed-point mode. It also provides a new theoretical model and corresponding device configuration applicable to both fixed-point mode and scanning mode. [Means for solving the problem]

[0007] The present invention provides a second-harmonic surveying method for measuring the entire area of ​​a wafer, and also provides an apparatus configuration for implementing this surveying method. The surveying modes of this method and apparatus include three types: fixed-point surveying, scanning surveying, and a combination of fixed-point surveying and scanning surveying. Fixed-point surveying has already been applied in current surveying technology. Scanning surveying, which has not yet been applied in the field of second-harmonic, is the original invention of this invention. The scanning surveying method provided by this invention can measure the entire area of ​​a wafer and locate and investigate abnormal points on the wafer while ensuring high surveying efficiency by controlling the shape, size, and luminous intensity of the light spot and adjusting the height of the point to be measured in real time during the surveying process. Furthermore, the present invention provides a correspondence between fixed-point surveying and scanning surveying, allowing for in-depth combination of these two modes. The combined fixed-point surveying and scanning surveying mode of the present invention not only reduces noise in the data results of the fixed-point surveying using the accurate initial value measured in the scanning mode, but also allows the results of the fixed-point surveying mode to be reprocessed, and the data from the scanning mode can be scanned, surveyed, investigated, and located for abnormal points on the wafer before performing fixed-point surveying for the abnormal points. Furthermore, by controlling the light intensity distribution and scanning speed of the light spot, the present invention provides a single "probing" function for scanning and surveying, i.e., the sample is probed only when the distribution of the internal charge in the sample remains almost unchanged and the measured signal is related to the initial state of the measured area. This method can also be used to improve the accuracy of the initial value. Furthermore, the surveying method of the present invention goes beyond analyzing data to establish a theoretical model based on the actual sample configuration and establish a more accurate set of theoretical equations, thereby unifying the actual survey results and the theoretical model in the three cases of fixed-point surveying, scanning surveying, and the combination of fixed-point surveying and scanning surveying.Based on the theoretical model of the present invention, the second harmonic surveying technique is applied not only as a qualitative analysis method but also as a quantitative analysis method.

[0008] Currently, scanning modes for detecting defects in semiconductors mainly include both physical scanning and microscopic imaging scanning. However, such scanning can only detect positions on the surface of the sample or a few nanometers below the surface, so the main targets of detection are physical defects such as scratches and metal particles in these areas. However, the present invention uses optical scanning based on second harmonic waves to detect electrical defects at the subsurface and within the film, such as volume charge and interface states. Currently, there is no scanning device that can locate such electrical defects. The second harmonic scanning technology developed by the present invention fills the gaps in second harmonic surveying technology, not only newly applying scanning surveying to the field of second harmonic waves, but also enabling the identification and location of electrical defects at the subsurface and within the film.

[0009] The present invention provides a correspondence between fixed-point measurement and scanning measurement, deeply integrating these two modes. The distinction between these two modes in terms of the type of electron excitation is explained as shown in Figure 1. Because the light intensity of the light spot is Gaussian, fixed-point measurement results in a Gaussian distribution of electrons excited in the illuminated area. Therefore, differences in the second-harmonic values ​​contributed by each location in the area will appear. However, in scanning mode, although the light intensity of the light spot remains Gaussian, the total length of time that all measurement points are illuminated and the total number of incident photons are the same in the direction of wafer movement, resulting in uniform redistribution of the internal charge by the laser light. If the electrical properties of the wafer to be tested are relatively uniform, the second-harmonic values ​​will inevitably be roughly the same everywhere. Therefore, in scanning measurement mode, a curve graph of "second-harmonic vs. coordinate" can be used to show the uniformity of the entire area of ​​the sample. Furthermore, if the size and light intensity of the light spot are the same, a correspondence exists between the two measurement modes.

[0010] By adjusting the relative moving speed and changing the amount of time the area to be tested is irradiated with the laser light, the matching between the scanning mode and the fixed point surveying mode can be realized; JPEG0007730578000001.jpg38170 is the equivalent size of a light spot.

[0011] The concept of equivalent size can be used in actual measurements when the shape of the light spot projected onto the sample is not circular (for example, elliptical). Generally, if the shape is centrally symmetric, JPEG0007730578000002.jpg6170 and the non-circular light point is JPEG0007730578000003.jpg12170

[0012] The total irradiation time does not exceed one account cycle, and this time is short enough (typically within the range of 0.1 ms to 1 ms) that the internal charge distribution at the point being measured on the wafer has not yet changed, and can be used to represent the initial state of the wafer. In fixed-point measurement, there is a delay (typically within the range of 15 ms to 30 ms) between the time when the laser beam is fully irradiated and the time when the signal is collected, so the internal state of the wafer has already changed at the time corresponding to the initial value in fixed-point measurement. Therefore, the scanning measurement mode significantly improves the accuracy of the initial value and can also be used to reduce noise and calibrate the results of fixed-point measurement.

[0013] It should be noted that the concept of equivalent size is introduced here for convenience of description, but the present technical means is not limited by the equivalent size. Using other parameters to indicate the size of the light spot is equivalent to substituting for the present technical means. What is important here is to establish a correspondence between fixed point measurement and scanning measurement, but does not change the specific method for realizing the correspondence.

[0014] If the scanning speed is slow enough, each measurement point on the wafer is irradiated with the laser light for a long enough time (e.g., 10 seconds or more), so that the charge distribution within it reaches a state of dynamic balance, which corresponds to the final state of fixed-point measurement.

[0015] According to the combined fixed point surveying and scanning surveying mode of the present invention, the abnormal points on the wafer may be scanned, surveyed, inspected and located, and then fixed point surveying may be performed on the abnormal points.

[0016] Furthermore, by controlling the shape, size, intensity and scanning speed of the light spot, the present invention provides a single "probing" function for scanning surveying, i.e., probing the sample only when the internal charge distribution in the sample remains substantially unchanged and the measured signal is related to the initial state of the surveyed area.

[0017] Generally, in second harmonic surveying techniques, the laser light source has two roles: sensing and excitation.

[0018] "Probing" refers to the coupling of light waves generated by the light source with non-centrosymmetric structures to generate second harmonics, and since defects in the crystal structure (e.g., interface states, fixed charges, impurity atoms, etc.) are usually coupling centers, the generated second harmonics can indicate the defect density in the crystal structure of the sample.

[0019] "Excitation" refers to the absorption of photons of incident light incident on the detection region by electrons, causing bound electrons in the valence band of the sample semiconductor layer to acquire sufficient energy to be excited into the conduction band of the semiconductor and become free electrons. These electrons may be captured by interface state defects or absorb more photons, gaining enough energy to overcome the barrier and reach the oxide layer, ultimately forming a large amount of charge accumulation at the interface or surface of the sample. This is called the electron accumulation process. Its essence is that more defects are created and can be "probed" by second-harmonic waves.

[0020] Laser light has two closely related roles that cannot be separated: "excitation" continuously changes the internal charge distribution state of the sample to be tested, while "probing" indicates this continuous change in the sample. Therefore, by collecting signals from the point to be measured over a long period of time, it is possible to measure and analyze a layered sample.

[0021] However, to probe the electrical properties of a sample in its initial state, it is necessary to aim for a single "probing" or "excitation" objective. This is difficult to achieve with current measurements. This is because photons from the probing light source itself can induce energy level transitions in electrons. Even if a relatively low-power probing light source is selected (which may result in a relatively low probe signal and relatively large noise), prolonged irradiation can still change the internal charge distribution in the sample. Therefore, it is unclear whether some of the energy from the "probing" light source is absorbed by the sample and used for electron ionization and excitation, or some of the energy is used for second-harmonic generation. In addition, because there is a time delay between the generation and reception of second-harmonic waves, there is a discrepancy between the timing at which the signal is collected in fixed-point measurements and the length of time the area is actually illuminated. This also leads to discrepancies between the measurement results and the theoretical results. The signal error due to the delay interval (Δt) can be evaluated using the following formula: JPEG0007730578000005.jpg11170

[0022] In the scanning surveying method provided by the present invention, as explained above, the scanned area is scanned by a laser beam. JPEG0007730578000006.jpg10170. The effect of delay errors is significantly reduced. Therefore, the scanning mode has a natural advantage when used to measure a specific second harmonic value, for example, to measure the second harmonic initial value and to reduce noise.

[0023] The second harmonic initial value can show the state when the sample is not yet changed, but the time delay between the signal excitation and the measurement causes some error in the fixed point measurement. However, in the scanning mode, any measurement in the sample JPEG0007730578000007.jpg20170

[0024] By appropriately setting the light spot size and the relative movement speed, the time value can be controlled within a relatively small range, for example, [0.1 ms, 1 ms]. In this time range, the internal charge distribution in the sample remains almost unchanged, and the measured signal is related only to the initial state of the measured area. This is the basic reason why the scanning measurement method according to the present invention can be realized to "measure" the initial value in a single step.

[0025] Furthermore, the surveying method of the present invention also offers advances in data analysis. By constructing a more accurate physical model and a more complete theoretical equation summary, the actual survey results and the theoretical model are unified in three cases: fixed-point surveying, scanning surveying, and a combination of fixed-point surveying and scanning surveying. Therefore, second-harmonic surveying technology can be applied not only as a qualitative analysis method but also as a quantitative analysis method. The following explanation takes the two-layer structure shown in Figure 2 as an example. The sample structure shown in Figure 2 is a commonly seen semiconductor two-layer structure. It has an oxide layer-semiconductor layer structure, but the application of the present invention is not limited to bilayer or semiconductor materials. In actual application, the data analysis method needs to be modified depending on the structure of the sample to be tested (e.g., number of layers, layer thickness, film layer material, reflectivity, etc.). For example, if the test object is a sample made of a different semiconductor material, the material attributes of the material can be used to recalculate the items related to each equation (e.g., Debye length, penetration depth, etc.).

[0026] Figure 2 shows defects such as interface states (i.e., dangling bonds) and volume charges near the interface of a two-layer sample. By selecting an appropriate wavelength of incident light, the probability of ionizing bound electrons in the medium layer is extremely small. Furthermore, when a photon penetrates the upper medium and arrives at the interface, electrons in the valence band in the space charge region (the region of the semiconductor layer near the interface) absorb the energy of a single photon, transition to the conduction band, and become free electrons, gradually accumulating at the interface. Some of these free electrons are captured by interface states and become bound electrons. The remaining unabsorbed electrons absorb more photon energy and transition to the conduction band of the upper medium. Due to the combined effects of the concentration gradient and the built-in electric field, these electrons move through the upper medium and are likely to be captured by volume charges. Finally, some electrons reach the surface and are captured by surface states. During the entire process, from electron ionization to capture at the surface, the internal charge distribution in the sample is changing, and so is the built-in electric field at the interface, until a new balance is finally reached. Therefore, the second harmonic also exhibits a time-varying trend that eventually stabilizes (Figure 3B). This is why second harmonic techniques can be used to reveal electrical defects related to this internal charge distribution, such as interface states and fixed charges.

[0027] JPEG0007730578000008.jpg10170JPEG0007730578000009.jpg15170. However, during the surveying process, the light intensity is not uniformly distributed but is Gaussian, and the electric field strength in the surveyed area also changes differently. Therefore, there is a relatively large difference between the theoretical basis of this formula and actual surveying, making it impossible to unify the actual surveying results with the theoretical model, and therefore making it impossible to effectively calculate quantitative surveying results. To address the above problem, the present invention further deduces this formula to obtain the following Equation 1: JPEG0007730578000010.jpg10170

[0028] JPEG0007730578000011.jpg14170 is the polar coordinate position with the center of the light spot as the origin, z is the distance of the light spot from the interface to be tested in the vertical direction, and t is the length of time the measurement point is irradiated with the laser light.

[0029] JPEG0007730578000012.jpg5170JPEG0007730578000013.jpg11170In the formula 2, P is the peak power of the laser light, and w is the beam waist width. JPEG0007730578000014.jpg11170JPEG0007730578000015.jpg11170JPEG0007730578000016.jpg37170

[0030] Introducing the built-in field created by applying an external bias voltage into Equation 3 has various drawbacks, such that the flatband voltage of the sample under test does not become zero. Therefore, when calculating quantitatively, the second harmonic contribution of the flatband voltage must be taken into account. The present invention achieves the goal of accurate measurement by adding an external voltage device to change the initial conditions before the sample under test is irradiated with the laser light by focusing. For example, if a bias voltage of the same magnitude but opposite direction as the flatband voltage is simultaneously applied to zero the initial built-in field of the sample under test, the resulting second harmonic will be minimized.

[0031] Figures 3A-3D show several second harmonic curves that appear in actual fixed-point measurements. By fitting these data curves according to the above summary, the corresponding electrical information can be obtained. Within the entire second harmonic curve, the most notable features are three signal points, each corresponding to one of the three states of the sample being tested. The initial state point indicates the state before photons are incident on the sample, i.e., the first point in Figures 3A-3D. At this time, free electrons have not yet been excited into the oxide layer, and the built-in electric field at this time is the initial built-in electric field. The time-series state points indicate the dynamic process of internal charge redistribution when photons are injected into the sample, reflecting the process of electrons being trapped by interface states or volume charges. This appears as a line segment moving in the middle in Figures 3A-3D. During this process, the built-in electric field changes, generating a time-series second harmonic. The figures show several different time-series states: a monotonically increasing trend shown in Figures 3A and 3B, a trend that decreases first and then increases later shown in Figure 3C, and a trend that increases first and then decreases later shown in Figure 3D. These different second harmonic change trends represent different sample initial states and defect densities, and are an important basis for quantitatively calculating defect density. Furthermore, the signal change rate is related to the thickness of the oxide layer, the band gap width of the material, and the frequency of the incident photons. For silicon semiconductor materials, JPEG0007730578000017.jpg5170Calculate the curve increase speed. JPEG0007730578000018.jpg11170JPEG0007730578000019.jpg29170When the thickness exceeds the threshold, Equation 4 is close to 0 and the resulting second harmonic does not change with time again. The final state point refers to the state after photons are injected into the sample and it becomes stable, or the state at the end of the measurement time. Stability means that during the measurement time, the internal charge distribution in the sample is dynamically balanced, i.e., the excitation rate of free electrons is equal to the recombination rate of electron-hole pairs, and the built-in electric field does not change again. Even if the measurement time is extended after that, there is no significant change in the signal value. This is shown as the line segment after the signal becomes stable in Figure 3B. However, when using low-power excitation light, the sample may not yet be in a stable state even when the measurement ends. If the measurement time is then extended, the second harmonic will still tend to change over time. In this case, the last point measured, such as the last point marked in Figure 3A, can be defined as the final state.

[0032] The present invention provides calculations using the above-mentioned Equations 1, 2, 3 and 4, and unifies the actual measurement results with the theoretical model in three cases: fixed point measurement, scanning measurement, and a combination of fixed point measurement and scanning measurement. This allows the second harmonic measurement technology to be applied not only as a qualitative analysis method but also as a quantitative analysis method.

[0033] To minimize noise, the device employs three methods: excitation light wave noise reduction, S exit light path noise reduction, and noise reduction performed on fixed point measurement signals by scanning measurement signals.

[0034] Excitation light wave noise reduction involves collecting the split light beams from one of the exit light paths until they reach the wafer in the incident light, monitoring the stability of the incident light, and performing noise reduction processing on the second harmonic when monitoring the luminous intensity density at the light spot in accordance with the monitoring of the imaging system for the shape and size of the light spot.

[0035] The S-exit optical path noise reduction is theoretically based on second harmonics. When the azimuth angle is zero and the polarization direction of the incident optical path is P, theoretically, no second harmonics are generated in the S direction. However, in actual measurements, the second harmonics in the S-exit optical path cannot be zero due to slight deviations in the azimuth angle or polarization angle, or the influence of other interference factors. Therefore, this signal value is used to perform noise reduction processing on the P-exit optical path. It should be noted that this optical path is not only used for noise reduction; in fact, when measuring with an adjusted azimuth angle, the second harmonics in the S direction are also significant when analyzing the symmetry or defects of the sample's crystal structure.

[0036] The use of scanning signals for noise reduction in fixed-point surveying is based on the previous interpretation of scanning modes. In rapid scanning surveying, the time spent irradiating each survey point with laser light is extremely short, and does not affect the internal charge distribution on the wafer. Therefore, by repeating the scanning process along the same scanning path and averaging the results of each survey, shot noise can be significantly reduced. Furthermore, because the signal value obtained in rapid scanning mode is closer to the initial value, the value obtained in this mode can be used to perform noise reduction on the initial value in fixed-point surveying.

[0037] To achieve the above-mentioned surveying objectives, the present invention further provides a surveying apparatus and a series of improvements to the apparatus, as shown in Fig. 4. The apparatus includes a light source, an incident optical path system, a sample, a mounting table, an exit optical path system, a signal receiving system, a monitoring system, an input system, a display system, and a central processing system, of which Fig. 4 shows only a part of the configuration. [Effects of the Invention]

[0038] The beneficial effects of the present invention are as follows: The present invention provides a second-harmonic surveying method for surveying an entire area on a wafer, and also provides an apparatus configuration for implementing this surveying method. The surveying means includes three types: fixed-point surveying, scanning surveying, and a combination of fixed-point surveying and scanning surveying. The scanning surveying means provided by the present invention can survey an entire area on a wafer, obtain the location, size, and relative density distribution of electrical defects, and locate and investigate abnormal points on the wafer, while ensuring high surveying efficiency. This is something that cannot be achieved with previous second-harmonic surveying techniques. Conventional second-harmonic techniques have a delay effect, making it impossible to repeatedly survey the same point within a short period of time. The scanning surveying means provided by the present invention allows the laser light to hit the surveying point for an extremely short time, without any impact on the wafer. Therefore, repeated surveying is possible, greatly improving the accuracy of the surveying. The combined fixed-point surveying and scanning surveying mode provided by the present invention provides a correspondence between fixed-point surveying and scanning surveying, and a method of scanning, surveying, investigating, and locating abnormal points on a wafer, and then performing fixed-point surveying on the abnormal points, thereby deeply combining these two modes. Furthermore, the present invention provides a single "probing" function for scanning surveying, allowing a single probing of a sample when studying the initial state of the sample. Furthermore, the present invention provides a new summary of the second-harmonic equation, unifying the actual survey results and theoretical models for the three cases of fixed-point surveying, scanning surveying, and the combination of fixed-point surveying and scanning surveying, allowing second-harmonic surveying technology to be applied as a quantitative analysis method rather than a qualitative analysis method. Furthermore, based on the theory of fixed-point surveying and scanning surveying, the present invention provides a noise reduction processing means for the device and analysis method, thereby significantly improving the signal-to-noise ratio of the surveying results.In summary, the surveying method and surveying device provided by the present invention can improve the completeness and accuracy of product inspection, enhance the ability to check the quality of advanced semiconductor processes, increase product yield, and ultimately provide more reference and reliable information for further process improvement, on the premise of further increasing the surveying speed, increasing efficiency, and ensuring efficiency.In addition, the technical means provided by the present invention do not require pre-treatment or post-treatment of the samples to be tested, do not require the discharge of any wastewater or exhaust gas, and do not cause any pollution to the environment. [Brief explanation of the drawings]

[0039] In order to more clearly explain the technical means of the embodiments of the present invention, the following briefly introduces drawings necessary for explaining the embodiments or prior art. It is obvious that the drawings described below are only some embodiments of the present invention, and those skilled in the art can obtain other drawings based on these drawings as long as it is an effort that does not involve an inventive step. [Figure 1] 1 is a diagram illustrating the difference in the principles of fixed-point surveying and scanning surveying in a specific embodiment of a surveying device for monitoring the quality of a semiconductor multilayer structure based on second harmonics according to the present invention; FIG. [Figure 2] 1 is a schematic diagram of the internal charge distribution of a sample composed of two layers in a specific embodiment of the measurement method for monitoring the quality of a semiconductor multi-layer structure based on second harmonics according to the present invention; FIG. [Figure 3A] 4 is a graph of a fixed-point measurement signal in one specific embodiment of the measurement method for monitoring the quality of a semiconductor multi-layer structure based on second harmonic waves according to the present invention; [Figure 3B] 4 is a graph of a fixed-point measurement signal in one specific embodiment of the measurement method for monitoring the quality of a semiconductor multi-layer structure based on second harmonic waves according to the present invention; [Figure 3C] 4 is a graph of a fixed-point measurement signal in one specific embodiment of the measurement method for monitoring the quality of a semiconductor multi-layer structure based on second harmonic waves according to the present invention; [Figure 3D]4 is a graph of a fixed-point measurement signal in one specific embodiment of the measurement method for monitoring the quality of a semiconductor multi-layer structure based on second harmonic waves according to the present invention; [Figure 4] 1 is a schematic diagram of a part of a configuration of a specific embodiment of a surveying device for monitoring the quality of a semiconductor multilayer structure based on second harmonics according to the present invention; [Figure 5] 1 is a schematic diagram of the degrees of freedom of a sample stage in a specific embodiment of a surveying device for monitoring the quality of a semiconductor multilayer structure based on second harmonics according to the present invention; FIG. [Figure 6] 1 is a schematic diagram of an optical path angle in a specific embodiment of a surveying instrument for monitoring the quality of a semiconductor multilayer structure based on second harmonics according to the present invention; FIG. [Figure 7] 3 is a graph of a scanning signal in a specific embodiment of the quality monitoring method for semiconductor multi-layer structures based on second harmonic waves according to the present invention; [Figure 8] 1 is a graph of defect density distribution in a specific embodiment of the quality monitoring method for semiconductor multilayer structures based on second harmonic waves according to the present invention; [Figure 9] 1 is a flow chart of a scanning-fixed point test for monitoring and measuring the quality of a semiconductor multi-layer structure based on second harmonic in accordance with the present invention; [Figure 10] 1 is a schematic diagram of a part of a configuration of a specific embodiment of a surveying device for monitoring the quality of a semiconductor multilayer structure based on second harmonics according to the present invention; DETAILED DESCRIPTION OF THE INVENTION

[0040] In this specification, the term "and / or" merely describes a relationship between related objects and indicates the existence of three relationships. For example, A and / or B can indicate three cases: the presence of only A, the presence of both A and B, and the presence of only B. Hereinafter, technical solutions according to the embodiments of the present application will be clearly and completely described with reference to the drawings in the embodiments of the present application. It is understood by those skilled in the art that these embodiments are only a part of the embodiments of the present application, and are not all of the embodiments. Any modifications and variations, as well as other embodiments, can be obtained by those skilled in the art based on the embodiments of the present application.

[0041] According to a first aspect, an embodiment of the present invention provides a method for surveying a semiconductor multi-layer structure based on second harmonic waves. The method includes generating laser light from a light source, passing it through an incident light path system to form a light spot on a sample, and directing photons onto the sample, i.e., irradiating the sample with the laser light.

[0042] The sample is a probeable structure having two or more layers of materials and an interface therebetween, at least one of the materials being a semiconductor layer, and the light source is capable of reaching the interface to be probed. The sample and the light spot can be moved relatively on a plane, and the relative movement JPEG0007730578000020.jpg20170 During the process of scanning and measuring the samples in the same batch, the shape and size of the light spot are maintained constant. The power of the light source is maintained constant, i.e., if the shape and size of the light spot are both maintained constant, the luminous intensity of the light spot is also maintained constant. The relative moving speed between the sample and the light spot is maintained constant.

[0043] When the photon is absorbed by an electron and the internal charge distribution in the sample changes, the change curve of the second harmonic wave received by the signal receiver from the exit optical path system is described as the second harmonic equation, which is the filtered single wavelength light with the same exit angle and incident angle.

[0044] The change curve of the second harmonic wave has the following characteristics: the time-varying characteristics of the second harmonic at which the predetermined survey point occurred, as recorded when entering the fixed point survey mode; and When entering the scanning surveying mode, the information recorded during the relative movement includes the spatial distribution characteristics of the second harmonic generated by the scanning surveying point.

[0045] In this embodiment, the light source achieves a single function: probing. In conventional second-harmonic measurement techniques, both "probing" and "excitation" are performed simultaneously. This is because the photons from the probing light source itself can cause electron energy level transitions. Even if a relatively low-power probing light source (which results in a relatively low measurement signal and relatively large noise) is used, prolonged irradiation can still cause changes in the charge distribution inside the sample. The scanning method provided by this embodiment can effectively achieve "probing" without changing the charge distribution inside the sample. By controlling the shape, size, luminous intensity, and scanning speed of the light spot, the scanning method provided by this embodiment combines scanning and measurement into a single function: probing. That is, the charge distribution inside the sample remains almost constant, and the measurement signal is only detected when it is related to the initial state of the area being measured. Therefore, this light source can be truly used as a single-purpose light source through technical means. Of course, for fixed-point measurement, the light source can also be used for both probing and excitation.

[0046] Laser light passes through the incident light path system and forms a light spot on the sample. The area of ​​the light spot affects the number of incident photons per unit area (light intensity density). For the same power, the smaller the light spot area, the more concentrated the incident photons are, and the greater the impact on the sample during the measurement process. However, the system also has a higher resolution.

[0047] The sample is a probeable structure consisting of two or more layers of material, with an interface present, at least one of which is a semiconductor layer. Because the technical means of this embodiment probes the interface states of semiconductors, the probed material must contain at least one semiconductor layer. Otherwise, the so-called "interface" will not be formed. The semiconductor material configuration includes currently common semiconductor materials (e.g., Si, Ge, SiC, GaN, etc.), new oxide semiconductors (e.g., ZnO, CdO, Fe2O3, etc.), one-dimensional semiconductors (e.g., graphene, black phosphorus, etc.), and other semiconductor materials with wide bandgaps (e.g., diamond, etc.). The light source must be able to reach the interface to be probed. This is because there are limitations on the thickness of each layer of the sample to be measured (especially material layers with relatively high absorption coefficients). If the material layer through which the light source penetrates is too thick or has a large absorption coefficient, insufficient photons can reach the interface to be tested, making the probe function impossible. If the sample is patterned, it is also necessary that the surface area of ​​the area to be tested be larger than the size of the light spot.

[0048] The sample and the light spot may move relative to each other on the plane, or the sample may move, the light spot may move, or both may move. Fixed Point Survey Mode JPEG0007730578000021.jpg26170

[0049] The scanning survey mode requires stable light source power. Because the second-harmonic value is directly proportional to the square of the light intensity, it is necessary to ensure stable light intensity (i.e., stable power and light spot shape and size) for actual scanning surveys. Otherwise, it is impossible to know whether the data fluctuations in the second-harmonic scanning signal are due to changes in the internal electrical properties of the sample being tested or changes in light intensity. Therefore, unless the shape, size, or light intensity of the light spot is controlled, an accurate second-harmonic scanning signal graph cannot be obtained. The shape and size of the light spot are adjusted by the incident light path system. Of course, this adjustment includes, but is not limited to, adjusting the parameters of the incident light by the incident light path system based on monitoring by a monitoring system, maintaining constant parameters of the light source and incident light system while the sample moves with the stage, or moving the light source while the stage remains stationary and parameters remain unchanged. Although light intensity can be controlled by adjusting the power, the power must be constant during the scanning process. Maintaining a constant speed of relative movement ensures that each scanned measurement point is irradiated with the laser beam for the same length of time during the scanning process. The "same lot" refers to the same sample, the same interface position to be tested, and the same test purpose and test method. Maintaining consistency in the measurement and maintenance of products within the same lot ensures comparability of the measurement data. Maintaining a certain degree of parameter does not necessarily mean that the parameters cannot be changed at all; it is preferable to maintain them constant when measuring "the same lot."

[0050] When receiving the second harmonic, it should be noted that the emitted signal contains multiple types of lightwave signals, so it is necessary to focus and select valuable information, i.e., perform filtering. The direction of the emitted light signal must be adjusted so that the angle of reception is the same as the incident angle. The second harmonic is measured and depicted as a curve, and the curve is described by an equation. The equations here include equations related to the prior art and equations that have been improved as technology has developed, and related equations can be used in combination. The equations related to the prior art include the standard equation for second harmonic mentioned in the section on means for solving the problem, but they will not be repeated here to avoid redundancy.

[0051] The change curve of the second harmonic wave contains a wealth of information. When the device is in fixed-point measurement mode, the recorded second harmonic wave curve changes over time, and thus has the characteristic of changing over time. When the device is in scanning measurement mode, the scanning process is limited, i.e., "the shape and size of the light spot are kept constant, the power of the light source is kept constant, i.e., if the shape and size of the light spot are both kept constant, the luminous intensity of the light spot is also kept constant, and the relative moving speed between the sample and the light spot is kept constant." Therefore, the illumination time for each scanned measurement point is equal, i.e., the value of each scanned measurement point is the value illuminated for the same amount of time at each point. Therefore, the value is not a function of time. However, since the positions of each scanned measurement point are different, the value changes with spatial position and has the characteristic of being spatially distributed. It should be noted that in fixed point surveying, v>0 occurs during the transition from predetermined survey point A to predetermined survey point B, but the survey has not yet been carried out at this time, so the second harmonic does not yet have a corresponding change curve.

[0052] Figure 1 shows the change in the internal charge distribution when an ideal, uniform sample is irradiated with laser light in two modes: fixed-point measurement and scanning measurement. Figure 1A shows the fixed-point measurement mode. As can be seen, the light intensity in the measurement area appears as a Gaussian distribution, and the internal charge redistribution caused by laser light excitation also follows a Gaussian distribution. Therefore, the internal charge distribution is not uniform, and there is a deviation between the obtained second-harmonic value and the theoretical value, making it difficult to represent the state of the sample. Figure 1B shows the scanning measurement mode. Although the light intensity of the light spot remains Gaussian, the internal charge redistribution caused by the laser light excitation becomes uniform in the direction of wafer movement. The total length of time that every measurement point is irradiated in that direction and the total number of photons incident are the same. Therefore, the curve graph called "second-harmonic vs. coordinate" created in scanning mode can be used to represent the uniformity of the entire area of ​​the sample. This representation is based on and defined by the above equation. Furthermore, the scanning survey mode has particular application in identifying defect locations, sizes, relative defect density distributions, etc., which will be described in the following examples and which cannot be achieved by the fixed point survey mode.

[0053] In some embodiments, the present invention provides a method for measuring a semiconductor multilayer structure based on second harmonic waves, wherein the sample and the light spot have a degree of freedom of relative movement in the X direction in the plane.

[0054] In this embodiment, as shown in Figure 5, the relative movement between the sample and the light spot on the plane has a degree of freedom in the X direction. The X direction means a certain direction, that is, the relative movement is along a straight line, which is a one-dimensional movement.

[0055] In some embodiments, the present invention provides a method for surveying a semiconductor multi-layer structure based on second harmonic waves, wherein the relative movement further has a degree of freedom in the Y direction.

[0056] In this embodiment, as shown in Figure 5, the relative movement between the sample and the light spot on the plane has an additional degree of freedom in the Y direction. The Y direction refers to a direction other than the X direction, meaning that the relative movement has two degrees of freedom. The degrees of freedom in the X and Y directions allow any type of movement to be completed on the XY plane, and various types of scanning can be realized. For example, linear scanning, arc scanning, etc. are possible, and scanning can be performed along any planar curve.

[0057] In some embodiments, the present invention provides a method for surveying a semiconductor multi-layer structure based on second harmonic waves, wherein the relative movement further comprises a rotation.

[0058] In this embodiment, the relative movement between the sample and the light spot further includes a rotation, that is, the scanning measurement is performed along a direction of a circular movement around a predetermined point, which may be any point on the plane, as shown in Figure 5.

[0059] In some embodiments, the present invention provides a method for surveying a semiconductor multi-layer structure based on second harmonic waves, wherein the relative movement is linear.

[0060] In this embodiment, the relative movement is linear, for example, grating or grid scanning, and the direction of the linear movement may be the same as or different from the X or Y direction.

[0061] In some embodiments, the present invention provides a method for surveying a semiconductor multi-layer structure based on second harmonic waves, wherein the relative movement is arc-shaped.

[0062] In this embodiment, the relative movement is an arc, for example, a measurement is performed during a planar rotation.

[0063] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on second harmonic waves, which adjusts the azimuthal angle of the sample in the range of [0°, 360°] by rotating the sample.

[0064] In this embodiment, the azimuth angle of the sample is adjusted by rotation, i.e., the incident surface and the sample JPEG0007730578000022.jpg7170 is processed for separation and analysis. The adjustment range of the azimuth angle can be selected according to actual needs. In this embodiment, the range of [0°, 360°] is preferred.

[0065] In some embodiments, the present invention provides a method for measuring semiconductor multilayer structures based on second harmonic waves, in which the relative positions of the sample and the light spot can be adjusted in the Z direction.

[0066] In this embodiment, the relative positions of the sample and the light spot can be adjusted in the Z direction (i.e., height direction). Preferably, the accuracy is at the micrometer level, even 0.1 micrometers or higher. During the scanning process, if the height of the sample changes, the light intensity also changes. Therefore, by adjusting the sample height in real time, JPEG0007730578000023.jpg6170 The purpose is to keep the height constant. This is because the surface of the sample itself is not flat, and the levelness of the mounting table is not uniform, which causes differences in the height of the measurement point, and also changes in the optical distance, light spot area, and light intensity density of the laser light. The scanning mode provides a function to adjust the height in real time, achieving better scanning and measurement results.

[0067] In some embodiments, in the method of the present invention for surveying semiconductor multilayer structures based on second harmonic generation, the laser light generated by the light source has adjustable power.

[0068] In this embodiment, the power of the laser light generated by the light source is adjustable. Adjusting the power allows for control of the total number of photons arriving at the measurement area per unit time. The higher the power, the more stable the generated signal, but the greater the number of incident photons per unit time, which increases the impact on the measurement sample. The power value is maintained constant throughout the measurement of the same lot to ensure that the intensity of the light spot remains constant. However, the power may be changed as needed. Depending on the measurement purpose, it may be necessary to adjust the power of the light source when irradiating laser light at different powers to detect changes in second harmonics. Alternatively, depending on the sample being measured, for example, if the material layer penetrated by the light source is too thick or has a large absorption coefficient, it may be necessary to adjust the power to ensure that a sufficient number of photons arrive at the interface to be tested. When the power is low, the signal change rate is relatively slow, and information about the internal charge redistribution when the wafer is irradiated with laser light is clearly visible, but it takes a relatively long time for the signal to reach saturation. When the power is high, the signal change rate is relatively fast, and although there may be some loss of information at the initial timing of laser light irradiation, the time required to reach saturation is relatively short, improving the efficiency of the measurement. At the same time, when the power is high, the signal becomes more stable. In this case, as described in this specification, it is not "the same lot." The laser light source can adjust the power to achieve different measurement purposes.

[0069] In some embodiments, the present invention provides a method for surveying semiconductor multilayer structures based on second harmonic generation, wherein the laser light has a power adjustable range of [0, 1000mW].

[0070] In this embodiment, the laser light preferably has a power adjustment range of [0, 1000 mW].

[0071] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on second harmonic generation, wherein the incident light path system is capable of adjusting the incident angle of the laser light.

[0072] In this embodiment, the incident light path system can adjust the incident angle of the laser light. The incident angle mainly affects the P-polarized component of the second harmonic wave because the P-polarized component is the sum of vectors in the incident direction and the perpendicular direction. Therefore, by adjusting the incident angle, a component in the second polarization tensor of the sample to be tested can be quantitatively analyzed.

[0073] In some embodiments, in the method of the present invention for surveying a semiconductor multilayer structure based on second harmonic waves, the incident angle of the laser light can be adjusted in the range of [10°, 90°].

[0074] In this embodiment, the incident angle of the laser light is adjustable in the range of [10°, 90°].

[0075] In some embodiments, in the method for surveying semiconductor multilayer structures based on second harmonic generation provided by the present invention, the incident light path system can adjust the polarization direction of the laser light to P polarization or S polarization.

[0076] In this embodiment, the incident light path system can adjust the polarization direction of the incident laser light to P polarization or S polarization.

[0077] In some embodiments, the present invention provides a method for surveying semiconductor multilayer structures based on second harmonic waves, including one or more signal receiving systems.

[0078] In this embodiment, there is one or more signal receiving systems, which can receive the second harmonic and / or incident optical split signals processed by the launching system and use them for different functions by multiple sensing devices.

[0079] In some embodiments, the present invention provides a method for surveying semiconductor multilayer structures based on second harmonics, wherein the signal receiving system includes a P signal receiving system and an S signal receiving system, where the P signal receiving system receives only the second harmonics in the P polarization direction, and the S signal receiving system receives only the second harmonics in the S polarization direction.

[0080] In this embodiment, the signal receiving system includes two P signal receiving systems and two S signal receiving systems, which are for receiving the second harmonic waves of the P polarization direction and the S polarization direction, respectively, and selectively receive the second harmonic waves.

[0081] In some embodiments, in the method for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, when the azimuth angle is zero and the polarization direction of the incident light path system is P, a non-zero signal received by the S signal receiving system performs noise reduction on the second harmonic received by the P signal receiving system.

[0082] In this embodiment, when the azimuth angle is zero and the polarization direction of the incident light path is P, theoretically, no second harmonic wave is generated in the S direction. However, in actual measurements, due to slight deviations in the azimuth angle or the angle of the polarization plate, or the influence of other interference factors, the second harmonic wave from the S exit light path is not zero. Therefore, this signal value is used for noise reduction processing in the P exit light path. Note that this light path is not only used for noise reduction; in fact, when adjusting the azimuth angle during measurement, the second harmonic wave in the S direction is also meaningful for analyzing the symmetry or defects in the crystal structure of the sample.

[0083] In some embodiments, the present invention provides a method for surveying semiconductor multi-layer structures based on second harmonic waves, and the signal receiving system includes three signal receiving systems: a P signal receiving system, an S signal receiving system, and an incident light splitting signal system.

[0084] In this embodiment, the signal receiving system includes three parts: a P signal receiving system, an S signal receiving system, and an incident light splitting signal system. The incident light splitting signal system is a light beam directly introduced from the incident light, not the second harmonic wave after measurement.

[0085] In some embodiments, in the method for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, when the azimuth angle is zero, a non-zero signal received by the S signal receiving system is used to perform noise reduction on the second harmonic received by the P signal receiving system, and the stability of the incident optical split signal system is monitored in real time and used for noise reduction.

[0086] In this embodiment, there are two noise reduction modes. When the azimuth angle is zero, a non-zero signal received by the S signal receiving system reduces noise on the second harmonic wave received by the P signal receiving system. This is the same as the noise reduction mode mentioned in the previous specification. At the same time, the stability of the incident light splitting signal system is monitored in real time and used for noise reduction. The incident light splitting signal system collects the split light beam of the incident light before it reaches the sample, primarily to monitor the fluctuation of the laser light power reaching the sample surface. In conjunction with the imaging system's monitoring of the shape and size of the light spot, this is used to monitor the light intensity density at the light spot, thereby enabling noise reduction on the second harmonic wave.

[0087] In some embodiments, the present invention provides a method for measuring a semiconductor multilayer structure based on second harmonics, the method further comprising changing the internal charge distribution state at the measurement point in the sample by a single excitation.

[0088] In this embodiment, the method further includes varying the initial state of the survey point on the sample with a single excitation. In some cases, the initial state is varied as needed until the sample is scanned, i.e., only "excited." The variation of the initial state must be uniform and stable, and may be achieved by optical or electrical devices.

[0089] In some embodiments, the present invention provides a method for second harmonic based interrogation of semiconductor multilayer structures, wherein the single excitation is achieved by a pump light source.

[0090] In this embodiment, the single excitation is realized by another light source, i.e., a pump light source. The pump light source excites the internal electrons in the sample, allowing them to move through the sample with sufficient energy and accumulate at the interface. Therefore, the pump light source must have a relatively wide power range and wavelength tunability. This technical solution is applicable to both scanning and fixed-point measurement, and can be applied before and during measurement. However, since the pump light source has a relatively small light spot, excitation is directed to the point to be measured.

[0091] In some embodiments, the present invention provides a method for surveying semiconductor multilayer structures based on second harmonic waves, wherein the single excitation is achieved by a flashlight.

[0092] In this embodiment, the excitation light source is a flashlight. Compared with the pump light source, the flashlight completely covers the entire wafer, so it affects the entire wafer. This technical solution is applicable to both fixed-point measurement mode and scanning measurement mode, and is only applied before measurement.

[0093] In some embodiments, the present invention provides a method for surveying semiconductor multilayer structures based on second harmonics, wherein the single excitation is achieved by applying a voltage to the sample by a bias voltage device sufficient to alter the built-in electric field of the sample.

[0094] In this embodiment, the single excitation is achieved by applying a voltage to the sample sufficient to alter the internal electric field within the sample to be measured. That is, a bias voltage device is added to alter the initial state of the sample. This method applies an electric field across the sample that sufficiently affects the internal charge arrangement, causing charges to build up at the interface. The density of the final accumulated charge is related not only to the applied electric field but also to other electrical properties (e.g., interface state density, fixed charge density, etc.). Once the charges have stably built up, the internal charge distribution within the sample to be tested can be effectively measured in a scanning mode, evaluating the electrical properties of the entire scanned area. The electric field can be applied directly using a DC voltage device or an AC voltage device, or by a magnetic field bias device or a corona device.

[0095] The DC voltage device utilizes the fact that charges of different polarities react differently to an electric field. When a DC voltage is applied, mobile charges (e.g., carriers, mobile ions, etc.) move in the direction of the electric field lines and eventually pile up at the field / surface. During this process, the presence of internal interface state defects or other electrical defects in the sample will result in slight differences in the final charge pile-up state. This difference is accounted for by performing feedback analysis on the signal in second harmonic scanning mode. This device can be used before or during second harmonic measurement.

[0096] AC voltage devices take advantage of the fact that different types of charges respond differently to AC voltage. For example, mobile charges move periodically with AC voltage (different types of mobile charges move at different speeds or displacements), while fixed charges do not. Furthermore, defects of the same type also respond differently to voltages of different frequencies. For example, interface states exhibit low resistance at high-frequency voltages but high resistance at low-frequency voltages. Therefore, applying AC voltages allows for more detailed analysis of various internal electrical properties of a sample. To ensure measurement accuracy, the frequency of the AC voltage device must be much smaller than the receiving frequency of the signal in the system.

[0097] In the two devices mentioned above, the sample is placed on a conductive tray (which is required to have as low an electrical resistance as possible), and the tray is grounded. A direct or alternating current voltage is applied to the other end of the sample by a non-contacting probe or a covered but non-contacting conductive plate. It is important to note that a small hole must be provided for the light beam to pass through when the voltage is applied by the conductive plate.

[0098] The magnetic field biasing device places the sample in a changing electromagnetic field, which induces an electric field that causes charges to move through the sample and accumulate at the sample's interface, where the applied magnetic field strength and the induced electric field strength are described by Maxwell's equations.

[0099] Regarding the corona device, a corona device with high voltage ionizes the water molecules and carbon dioxide molecules in the surrounding air, generating positive H3O2 + Ions and negative CO3 - Ions are generated, and optionally one of the ions can be focused and scattered onto the surface of the sample, which changes the surface potential of the sample and aims to redistribute the charge by attracting or repelling internal charges in the sample.

[0100] In this embodiment, various types of excitation devices may be used in combination or may be applied in combination with the pump light source or flashlight, and these excitation devices may operate independently and not interfere with each other.

[0101] In the various methods mentioned above, the scanning mode itself is independent of the "excitation" device. In the actual measurement process, a single scanning mode can be used, or the two can be combined to perform multiple types of measurements. In short, this is because the initial interface state of the sample can be indicated by scanning and measurement. At the same time, the most notable advantage of the scanning method over measuring a single point is that it can increase the number of measurement points and improve measurement efficiency and accuracy.

[0102] In some embodiments, in the method of the present invention for surveying a semiconductor multilayer structure based on second harmonic waves, the time-varying feature and the space-varying feature simultaneously or individually satisfy the following: The time-varying features include an initial state that indicates the state of the sample before the photons are poured into it, a final state that indicates the state of the sample when the measurement is completed, and a time-series state that indicates the dynamic process of the internal charge being redistributed as the photons are poured into the sample. The spatial variation characteristics include normal points, which are measurement points where the value of the second harmonic is within a 5% range above and below the average value, and abnormal points, which are measurement points where the value of the second harmonic is outside a 5% range above and below the average value.

[0103] In this embodiment, the time-varying feature and the space-varying feature may be provided simultaneously, but are not necessarily provided simultaneously. The fixed point surveying mode with the time-varying feature and the scanning surveying mode with the space-varying feature may be used one after the other, or may be used simultaneously (similarly performing surveys on the same sample using devices with the same model number), or may be used separately.

[0104] The time-varying features include the initial state, the final state, and the time-series state. The initial state refers to the initial state of the sample before measurement. The initial state may be the original state of the sample or the state of the sample after a single excitation. In other words, it is the state up to measurement. The final state refers to the state of the sample when measurement is completed. After measurement is completed, the sample is no longer irradiated, but its state may be continuously changing. However, the device does not receive the newly generated second harmonic. Therefore, the state at the end of measurement cannot be considered the final state. Of course, if the measurement takes a long enough time, the state of the sample may not change further and may be stable when measurement is completed. The time-series state refers to the process by which the sample dynamically changes from the initial state to the final state.

[0105] The spatial variation feature includes a normal point and an abnormal point. The normal point is where the second harmonic value is within 5% of the average value, i.e., close to the average value. The abnormal point is where the second harmonic value is outside the 5% range of the average value, i.e., significantly deviates from the average value. The 5% value is a preferred value provided in this embodiment, but other values ​​may be used depending on the actual situation.

[0106] Figures 3 and 7 show examples of combining the fixed-point measurement mode and the scanning measurement mode. Figure 3 shows fixed-point measurement of several points. Figure 7 shows the scanning measurement mode. Fixed-point measurement is used to quantitatively analyze the measured defect density, while scanning mode is used to measure different measurement points relatively. Therefore, by combining the two measurement modes, quantitative analysis of the defect density of the entire area of ​​a sample can be achieved. Figure 7 shows an example of data for linear scanning. In actual measurements, different scanning modes may be used depending on the type of sample. Therefore, corresponding data processing methods are required. For example, for patterned wafers, the area to be measured is extremely small (usually several times the size of the light spot). Therefore, when selecting the scanning mode, the scanning area is set to be slightly larger than the test piece (typically 50 micrometers x 50 micrometers). During data processing, signal points outside the test piece area are deleted. On the other hand, for unpatterned wafers, a diverse measurement method can be selected to ensure randomness and diversity in the points to be tested.

[0107] In some embodiments, the method for measuring a semiconductor multilayer structure based on second harmonics provided by the present invention further includes a fixed-scanning correspondence relationship, the fixed-scanning correspondence relationship being a correspondence relationship between the fixed point measurement mode and the scanning measurement mode, i.e., a mean value of the second harmonic wave acquired in the scanning measurement mode and a mean value of the second harmonic wave acquired in the fixed point measurement mode are used. JPEG0007730578000024.jpg25170The length of time that the specified measurement point on the sample is irradiated with the laser light. JPEG0007730578000025.jpg18170JPEG0007730578000026.jpg8170

[0108] In this embodiment, a correspondence relationship between the fixed point measurement mode and the scanning measurement mode has been established. The method of combining the fixed point measurement mode with the scanning measurement mode can analyze the electrical properties of the interface of the sample to be tested. The advantages of this measurement method include the ability to perform noise reduction on the fixed point measurement data results using the accurate initial state measured in the scanning mode, and the ability to reprocess the scanning mode data using the results of the fixed point measurement mode.

[0109] JPEG0007730578000027.jpg14170JPEG0007730578000028.jpg38170, what is acquired by the scanning surveying mode is the final state of the fixed point surveying mode. Generally, the scanning surveying mode is required to acquire the final state. The relative movement speed needs to be very slow, and it is related to the overall length of the fixed point surveying time, that is, related to the final state of the fixed point surveying. During the scanning surveying process, the time that the measurement point is irradiated with the laser light is very short. If ts is small enough, the time that is achieved by the scanning surveying mode becomes the initial value of the fixed point surveying mode.

[0110] In some embodiments, the present invention provides a method for measuring semiconductor multilayer structures based on second harmonic JPEG0007730578000030.jpg7170 The average value of the second harmonic is the initial value in the fixed point survey mode. JPEG0007730578000031.jpg5170JPEG0007730578000032.jpg8170

[0111] The total irradiation time does not exceed the accounting cycle, and this time is sufficiently short (typically within the range of 0.1 ms to 1 ms) that the charge distribution within the sample at the scanning measurement point remains unchanged. In the scanning measurement mode, the average value of the second harmonic is used to indicate the initial state of the sample. This point is the first measurement point (initial value) in the fixed-point measurement mode. In the fixed-point measurement mode, there is a delay (typically within the range of 15 ms to 30 ms) between the time the laser beam is fully irradiated and the time the signal is collected, so the interior of the wafer has already changed at the time corresponding to the initial value of the fixed-point measurement. Therefore, the scanning measurement mode can significantly improve the accuracy of the initial value.

[0112] Furthermore, when performing rapid scanning and measurement, the time it takes for the laser beam to irradiate each measurement point is extremely short, so the internal charge distribution on the wafer is not affected. Therefore, by repeating the scanning process along the same scanning path and averaging the measurement results, shot noise can be significantly reduced. In other words, the values ​​obtained by this scanning measurement mode can be used for noise reduction and calibration of the initial values ​​in the fixed-point measurement mode.

[0113] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on second harmonic, where the equation for the second harmonic is written as Equation 1: JPEG0007730578000034.jpg11170JPEG0007730578000035.jpg14170 is the polar coordinate position with the center of the light point as the origin, z is the distance of the light point from the interface to be tested in the vertical direction, and t is the length of time the measurement point is irradiated with the laser light.

[0114] JPEG0007730578000036.jpg10170JPEG0007730578000037.jpg10170In Equation 2, P is the peak power of the laser light, and w is the beam waist width. JPEG0007730578000038.jpg11170JPEG0007730578000039.jpg11170

[0115] JPEG0007730578000040.jpg37170

[0116] In this example, as shown in Figure 1, the distribution of internal charges changes when an ideal and uniform sample is irradiated with laser light in two modes: fixed-point surveying and scanning surveying. Figure 1A shows the fixed-point surveying mode. As can be seen, the light intensity of the surveyed area appears as a Gaussian distribution, so the redistribution of internal charges due to the excitation of laser light also becomes a Gaussian distribution. Therefore, data processing can no longer be analyzed using the formulas disclosed in previous literature, and analysis must be performed using formulas 1, 2, and 3. Formula 1 is the usual formula for the second harmonic. JPEG0007730578000041.jpg26170 Because the luminous intensity is not uniformly distributed and the changes in the electric field strength in the survey area are different, this embodiment provides a new formula, namely, Formula 1. In Formula 1, the built-in electric field is the most important parameter and the most basic element for extracting the time series second harmonic.

[0117] Equation 2 is a conventional equation for the distribution of Gaussian spot luminous intensity. However, conventional second-harmonic wave technology calculates the luminous intensity distribution as if the spot were a uniform spot, which can lead to errors during actual measurement. Therefore, the present invention introduces a more accurate physical model by incorporating the conventional equation.

[0118] The electric field formula according to Equation 3 is more consistent with the scanning surveying mode. This is because the distribution of incident photons during the scanning process is uniform, and the resulting charge buildup is also uniform, which is consistent with the theoretical model in Equation 3. Here, these two values ​​are specified as This value is sufficient to change the initial state of the sample being tested. N is the cumulative charge density at the interface during the measurement process, and is the change in charge distribution at the interface due to electrons being trapped by interface states or fixed charges during the transfer process. Equation 3 is a deeper development of the second-harmonic equation by the present invention, which more accurately describes the built-in electric field at a semiconductor interface. While conventional second-harmonic techniques attribute the change in the built-in electric field to charge buildup at the interface, in fact, the built-in electric field is the result of the combined action of various electrical defects and charge buildup. However, the effect of electrical defects is described by the flat-band voltage (i.e., the first term on the right in Equation 3) and is adjusted by the external voltage. The effect of charge buildup is described by the second term on the right in Equation 3. Therefore, as can be seen, the model described in Equation 3 is closer to the true case. Therefore, the fixed point survey mode has a better role in description.

[0119] The combination of the scanning survey mode and the bias voltage technique allows for the The advantages of combining measurements are: (1) increasing the signal-to-noise ratio of the signal; (2) allowing the sample to be tested in different operating states (major carrier build-up, major carrier depletion, minor carrier build-up, depletion) for more accurate analysis of the sample's performance; and (3) applying different biases. The answer is JPEG0007730578000044.jpg14170.

[0120] Calculations are performed using Equations 1, 2, and 3 provided by the present invention, and the actual measurement results and theoretical models are unified in three cases: fixed point measurement, scanning measurement, and a combination of fixed point measurement and scanning measurement, and the technology based on second harmonic measurement is applied not only as a qualitative analysis method but also as a quantitative analysis method.

[0121] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on second harmonics, the method further comprising, upon performing the scanning mode surveying, generating a scanning signal graph with the scanning direction as the abscissa and the value of the second harmonic as the ordinate. In the scanning signal graph, the signal peaks and signal valleys are abnormal points of the data. The defect density at the abscissa of the peak value corresponding to the signal peak is greater than that of the surrounding area, the abscissa of the peak value is the defect center, the width of the signal peak is the size width of the defect, and the height of the signal peak indicates the defect density.

[0122] The defect density at the abscissa of the valley value corresponding to the signal valley is smaller than that of the surrounding area, the abscissa of the valley value is the defect center, the width of the signal valley is the size width of the defect, and the height of the signal valley indicates the defect density.

[0123] In this embodiment, as shown in Figure 7, the abscissa is the coordinate of the scanning direction, and the ordinate is the second harmonic wave obtained by scanning. Since the second harmonic wave is related to the defect density, in Figure 7, the defect density at the abscissa of the peak value corresponding to the signal peak is higher than that of the surrounding area, the abscissa of the peak value is the defect center, the width of the signal peak is the size width of the defect, and the height of the signal peak indicates the defect density. The defect density at the abscissa of the valley value corresponding to the signal valley is lower than that of the surrounding area, the abscissa of the valley value is the defect center, the width of the signal valley is the size width of the defect, and the height of the signal valley indicates the defect density. Therefore, the scanning mode can be used to measure the relative defect distribution in the sample.

[0124] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on a second harmonic, the method further comprising creating a graph of defect density distribution based on the value of the second harmonic and the coordinate position of the scanning survey point in the scanning survey mode. In the defect density distribution graph, bright points correspond to the signal peaks in the scanning signal graph and indicate that the defect density at the coordinate position at the scanning measurement point corresponding to the bright points is greater than that of the surrounding area, while dark points correspond to the signal valleys in the scanning signal graph and indicate that the defect density at the coordinate position at the scanning measurement point corresponding to the dark points is less than that of the surrounding area.

[0125] In this embodiment, as shown in Figure 8, a graph of the defect density distribution in the scanning survey mode is shown, where the defect density at the bright spot is higher than that of the surrounding area, corresponding to the peak in Figure 7, while the defect density at the dark spot is lower than that of the surrounding area, corresponding to the valley in Figure 7.

[0126] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on second harmonics, the method further comprising creating a graph of the second harmonic over time in the fixed point surveying mode, with the surveying time as the abscissa and the value of the second harmonic as the ordinate.

[0127] The time series second harmonic graph has the following characteristics: The initial state point is the first point in the time series second harmonic graph and represents the state before the photons are poured into the sample. The final state point is the last point in the time series second harmonic graph and indicates the state of the sample when the survey ends. The time series state points are points between the initial state point and the final state point in the time series second harmonic graph, and represent the dynamic process by which the internal charge is redistributed as the photons are poured into the sample.

[0128] In this embodiment, the three most notable aspects of the entire time series second harmonic curve are the initial state, the final state, and the time series state.

[0129] The initial state point indicates the state before photons are incident on the sample, i.e., the first point in Figures 3A, 3B, 3C, and 3D. At this time, free electrons have not yet been excited into the oxide layer, and the built-in electric field at this time becomes the initial built-in electric field.

[0130] The time-series state points show the dynamic process of the internal charge redistribution when photons are incident on the sample, resulting in electrons being captured by interface states or fixed charges. This is shown as the changing line segments in Figures 3A, 3B, 3C, and 3D. This process changes the built-in electric field, generating a time-series second harmonic. The rate of change of this signal is related to the thickness of the oxide layer, the bandgap width of the material, and the frequency of the incident photons. JPEG0007730578000045.jpg5170Calculate the rate at which the second harmonic curve increases. JPEG0007730578000046.jpg11170JPEG0007730578000047.jpg29170When the thickness exceeds a threshold, Equation 4 approaches zero. The generated second harmonic does not change with time. Therefore, for relatively thick oxide layers, increasing the frequency of the incident photons increases the probability of electron energy level transitions and arrival at the oxide surface.

[0131] The final state point indicates the state where the sample becomes stable as photons are injected into it, or the state at the end of the measurement time. Stability means that during the measurement time, the excitation and binding of new free electrons are balanced, the internal charge distribution in the sample is dynamically balanced, and the built-in electric field remains unchanged. At this time, even if the measurement time is extended, the signal value does not change significantly. This is shown as the line segment where the signal becomes stable in Figures 3B, 3C, and 3D. However, when using low-power excitation light, the sample may not yet reach a stable state when the measurement ends. In this case, if the measurement time is extended, the second harmonic wave will still tend to change over time. In this case, the last point measured, for example, the last point shown in Figure 3A, may be defined as the final state.

[0132] In some embodiments, the present invention provides a method for surveying a semiconductor multilayer structure based on second harmonics, the method further comprising selecting some or all of the abnormal points of the signal as the predetermined survey points and entering the fixed point survey mode.

[0133] In this embodiment, a special point is selected as the location of the fixed point survey, which means that during the surveying process, a fixed point survey is performed at the abnormal point (peak or valley) of the signal, with the purpose of conducting a more detailed survey analysis there.

[0134] In some embodiments, in the method for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, the method further includes randomly selecting the survey point as the predetermined survey point in advance and entering the fixed point survey mode.

[0135] In this embodiment, the location of the fixed point survey is randomly selected, which is selected randomly before the survey and is mainly used to calculate the defect density.

[0136] 9 shows a flowchart of a specific measurement process that comprehensively applies the techniques provided in the above embodiments. The measurement steps include inserting a wafer, adjusting parameters, selecting a scan line, moving the stage, measuring until all scanning and measurement tasks are completed, analyzing abnormal points, and entering fixed point measurement mode. After all the measurements are completed, the data is analyzed.

[0137] In a second aspect, an embodiment of the present invention provides an apparatus for measuring a semiconductor multilayer structure based on second harmonic waves, the apparatus comprising: The light source is for generating laser light, passing through an incident light path system to form a light spot on the sample, and pouring photons onto the sample, that is, irradiating the sample with laser light. The incident light path system is for adjusting optical parameters of the laser light generated by the light source. The sample is a probeable structure having two or more layers of materials and an interface therebetween, at least one of the materials being a semiconductor layer, and the light source is capable of reaching the interface to be probed. The stage is used to place the sample and move it in a plane. JPEG0007730578000048.jpg20170 The output optical path system is for adjusting second harmonic parameters of the output second harmonic. The signal receiving system is for receiving the second harmonic wave adjusted by the output optical path system, and the second harmonic wave has a variation curve described as a second harmonic equation, the variation curve including information on the time variation characteristics of the second harmonic wave generated by the predetermined measurement point, which is recorded when entering the fixed point measurement mode, and the spatial distribution characteristics of the second harmonic wave generated by the scanning measurement point, which is recorded during the relative movement when entering the scanning measurement mode. The monitoring system is for monitoring the operating status of the device in real time and transmitting real-time feedback information. An input system is a system for receiving user input information in human-computer interaction. The display system is for displaying the output information of the device in human-computer interaction. The central processing system is for receiving the input information and the real-time feedback information, controlling the operation of the device, processing the second harmonic according to the second harmonic formula, and outputting the output information.

[0138] During the process of scanning and measuring the samples in the same batch, the shape and size of the light spot are maintained constant, the power of the light source is maintained constant, i.e., the shape and size of the light spot are both maintained constant, and the luminous intensity of the light spot is also maintained constant, and the relative moving speed between the sample and the light spot is maintained constant.

[0139] In this embodiment, a device for measuring semiconductor multilayer structures based on second harmonic waves is provided. Figure 4 shows a partial configuration of the device. The device includes the light source, the incident light path system, the sample, the mounting stage, the exit light path system, the signal receiving system, the monitoring system, the input system, the display system, and the central processing system. In conventional second harmonic wave measurement techniques, both "probing" and "excitation" are performed simultaneously. This is because photons from the probing light source itself can cause electron energy level transitions. Even if a low-power probing light source (which results in a relatively low measurement signal and relatively large noise) is used, prolonged irradiation will still cause changes in the charge distribution inside the sample. The scanning method provided in this embodiment can effectively achieve "probing" without changing the charge distribution inside the sample. By controlling the shape, size, luminous intensity, and scanning speed of the light spot, the scanning device provided in this embodiment combines scanning and measurement into a single function called "probing." That is, when the internal charge distribution in the sample remains almost constant and the measured signal is related only to the initial state of the measured area, it can be realized to probe the sample. Therefore, the light source is a true light source that can be realized with a single purpose by technical means. Of course, in fixed-point measurement, the light source simultaneously performs probing and excitation. The excitation light source in FIG. 4 performs the function of excitation, but the excitation light source is not necessarily an independent component. The excitation device can be realized by the light source itself or by an independent excitation light source.

[0140] Laser light passes through the incident light path system and forms a light spot on the sample. The area of ​​the light spot affects the number of incident photons per unit area (light intensity density). For the same power, the smaller the light spot area, the more concentrated the incident photons are, which increases the impact on the sample during the measurement process. However, the system also has a higher resolution.

[0141] The sample is a probeable structure consisting of two or more layers of material, with an interface present, at least one of which is a semiconductor layer. Because the technical means of this embodiment probes the interface states of semiconductors, the probed material must contain at least one semiconductor layer. Otherwise, the so-called "interface" will not be formed. The semiconductor material configuration includes currently common semiconductor materials (e.g., Si, Ge, SiC, GaN, etc.), new oxide semiconductors (e.g., ZnO, CdO, Fe2O3, etc.), one-dimensional semiconductors (e.g., graphene, black phosphorus, etc.), and other semiconductor materials with wide bandgaps (e.g., diamond, etc.). The light source must be able to reach the interface to be probed. This is because there are limitations on the thickness of each layer of the sample to be measured (especially material layers with relatively high absorption coefficients). If the material layer through which the light source penetrates is too thick or has a large absorption coefficient, insufficient photons can reach the interface to be tested, making the probe function impossible. If the sample is patterned, it is also necessary that the surface area of ​​the area to be tested be larger than the size of the light spot.

[0142] The sample is placed on the stage, and by moving the stage, the sample and the light spot move relative to each other on the plane. JPEG0007730578000049.jpg18170

[0143] The output optical path system is for adjusting second harmonic parameters of the output second harmonic, including but not limited to output light polarization direction and output angle.

[0144] The signal receiving system is for receiving the second harmonic wave adjusted by the output optical path system. The second harmonic wave is expressed as a second harmonic wave formula. When receiving the second harmonic wave, it is important to note that the output signal contains multiple types of light wave signals, so focusing and selecting valuable information, i.e., filtering, is required. The signal receiving direction of the output light must be adjusted to be the same as the incident angle. The second harmonic wave is measured and depicted as a curve, and the curve is described using a formula. The formula here includes formulas related to the prior art and formulas that have been improved as technology has developed, and related formulas can be used in combination. The formulas related to the prior art include the standard formula for second harmonic wave mentioned in the section on means for solving the problem, but to avoid redundancy, they will not be repeated here.

[0145] The monitoring system is for monitoring the operating status of the device in real time and transmitting feedback information in real time, including the power of the light source, the size of the light spot, the height of the mounting table, etc.

[0146] The input system is for receiving user input information in human-computer interaction. The input system may be directly connected to the body of the device, or may send and receive data wirelessly or via a wire. It is only necessary to ensure that the input information can be sent to the device.

[0147] The display system is used to display the output information of the device in human-computer interaction. The display system may be directly connected to the body of the device, or may transmit and receive data wirelessly or via a wire. It is sufficient to ensure that the output information can be obtained from the device.

[0148] The central processing system receives the input information and the real-time feedback information, controls the operation of the device, processes the second harmonic based on the second harmonic formula, and outputs the output information. The central processing system may be integrated into the device or located in the cloud. It may be a single computer, multiple computers, a physical computer, or a virtual machine. Processing the second harmonic based on the second harmonic formula includes the variation curve of the second harmonic. The variation curve of the second harmonic contains rich information. When the device is in fixed-point surveying mode, the recorded second harmonic curve changes over time, and therefore has a time-varying characteristic. When the device is in scanning measurement mode, the scanning process is limited, i.e., "the shape and size of the light spot remain constant, the power of the light source remains constant, i.e., if the shape and size of the light spot remain constant, the luminous intensity of the light spot remains constant, and the relative moving speed between the sample and the light spot remains constant." Therefore, the illumination time for each measurement point measured by scanning is equal. In other words, the value of each measurement point measured by scanning is the value obtained when each point is illuminated for the same amount of time. Therefore, the value is not a function of time. However, the position of each measurement point measured by scanning is different. Therefore, the value changes depending on the spatial position and has the characteristic of being distributed in space. It should be noted that in the case of fixed-point measurement, even though v>0 is achieved in the process from specified measurement point A to specified measurement point B, no measurement is performed. Therefore, the change curve for the second harmonic does not yet exist.

[0149] Specifically, Figure 10 shows an example of the configuration of an optical path system. The incident light is generated from a laser light source (10), passes through a polarizing means (20) and a collimating lens group (30), and reaches the sample (90). This excites a second harmonic wave, generating a reflected light. The reflected light is then filtered through a collimating lens group (40) and a filter (50). The remaining excited second harmonic wave passes through a beam splitter (60) to generate second harmonics in two polarization directions. Of these, the second harmonic wave in the P direction passes through the beam splitter (60) and is directly received by the signal receiver (110), while the second harmonic wave in the S direction passes through the beam splitter (60) and is reflected by a mirror (51) and received by the other signal receiver (110). During the surveying process, information on the area to be tested is monitored by the transmission electron microscope (80) and the lens group (81, 70), the height, coordinates and moving speed of the sample are controlled by the mounting table (100), and all information including the surveying data is finally collected and processed in the central processing system (120).

[0150] Figure 1 shows the change in the internal charge distribution when an ideal, uniform sample is irradiated with laser light in two modes: fixed-point measurement and scanning measurement. Figure 1A shows the fixed-point measurement mode. As can be seen, the light intensity of the measured area appears as a Gaussian distribution, and the internal charge redistribution due to laser light excitation also follows a Gaussian distribution. Therefore, the charge distribution becomes non-uniform, and there are deviations in the acquired second-harmonic values, making it difficult to indicate the state of the sample. Figure 1B shows the scanning measurement mode. Although the light intensity of the light spot still follows a Gaussian distribution, the internal charge redistribution due to laser light excitation is uniform in the direction of wafer movement. This means that the total length of time and the total number of photons irradiated at every measurement point in that direction are the same. Therefore, the "second-harmonic vs. coordinate" curve graph created in scanning mode can indicate the uniformity of the entire area of ​​the sample. This is based on and defined by the above formula. Additionally, the scanning survey mode has particular application in identifying defect locations, sizes, and relative defect density distributions, as will be described in the following embodiments, which cannot be achieved by the fixed point survey mode.

[0151] The scanning survey mode requires stable light source power. Because the second-harmonic value is directly proportional to the square of the light intensity, it is necessary to ensure stable light intensity (i.e., stable power and light spot shape and size) for practical scanning surveys. Otherwise, it is impossible to know whether the data fluctuations in the second-harmonic scanning signal are due to changes in the internal electrical properties of the sample being tested or changes in light intensity. Therefore, unless the shape, size, or light intensity of the light spot is controlled, an accurate second-harmonic scanning signal graph cannot be obtained. The shape and size of the light spot are adjusted by the incident light path system. Of course, this adjustment includes adjusting the parameters of the incident light by the incident light path system based on monitoring by a monitoring system, and also includes maintaining constant the parameters of the light source and the incident light system even as the sample moves along with the stage. Although the light intensity can be controlled by adjusting the power, the power must be constant during the scanning process. Maintaining a constant relative movement speed ensures that each scanned measurement point is irradiated with the laser beam for the same length of time during the scanning process. The "same lot" refers to the same sample, the same interface position to be tested, and the same test purpose and test method. Maintaining consistency in the measurement and maintenance of products within the same lot ensures comparability of the measurement data. Maintaining a certain degree of parameter does not necessarily mean that the parameters cannot be changed at all; it is preferable to maintain them constant when measuring "the same lot."

[0152] In some embodiments, in the apparatus for surveying semiconductor multi-layer structures based on second harmonic waves provided by the present invention, the planar movement of the mounting table has a degree of freedom in the X direction.

[0153] In this embodiment, the planar movement of the mounting table has a degree of freedom in the X direction, as shown in Fig. 5. The X direction means a certain direction, that is, the relative movement is performed along a straight line, and is a one-dimensional movement.

[0154] In some embodiments, in the apparatus for measuring semiconductor multi-layer structures based on second harmonic waves provided by the present invention, the mounting stage further has a degree of freedom in the planar movement in the Y direction.

[0155] In this embodiment, as shown in Figure 5, the planar movement of the mounting table also has a degree of freedom in the Y direction. The Y direction refers to a direction other than the X direction, meaning that the relative movement has two degrees of freedom. The degrees of freedom in the X and Y directions allow any type of movement in the XY plane and realizes different types of scanning. For example, linear scanning, arc scanning, etc. can be performed along any planar curve.

[0156] In some embodiments, in the apparatus for surveying semiconductor multi-layer structures based on second harmonic waves provided by the present invention, the stage is further rotatable.

[0157] In this embodiment, the planar movement of the mounting table further includes rotation, that is, the scanning survey is performed along a circular movement direction around a predetermined point, which may be any point on the mounting table, as shown in Figure 5.

[0158] In some embodiments, the present invention provides an apparatus for surveying semiconductor multi-layer structures based on second harmonic waves, wherein the mounting stage has an adjustable height position.

[0159] In this embodiment, the mounting stage is adjusted in height to control the shape and size of the light spot relative to the light source and the incident light path system.

[0160] In some embodiments, in the apparatus for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, in the scanning surveying mode, the position of the mounting table can be adjusted in real time in the height direction based on real-time feedback of the height of the surveying point on the sample.

[0161] In this embodiment, the mounting stage adjusts the relative position in the height direction. Preferably, the precision is at the micrometer level, even 0.1 micrometer or higher. During the scanning process, as the height of the sample changes, the density of the light intensity also changes. Therefore, the light intensity is controlled by adjusting the height of the sample in real time. This is mainly due to the light spot area. JPEG0007730578000050.jpg5170. This is because the height of the measurement point varies depending on factors such as the surface of the sample itself being uneven and the levelness of the mounting table being uneven, which in turn changes the optical distance, light spot area, and luminous intensity density of the laser light. In scanning mode, the height can be adjusted in real time by monitoring the feedback, resulting in better scanning and measurement results.

[0162] In some embodiments, the present invention provides an apparatus for surveying semiconductor multi-layer structures based on second harmonic waves, wherein the planar movement is linear.

[0163] In this embodiment, the planar movement is linear, for example, grating or grid scanning, and the direction of the linear movement may be the same as or different from the X or Y direction.

[0164] In some embodiments, the present invention provides an apparatus for surveying semiconductor multi-layer structures based on second harmonic waves, wherein the planar movement is an arc.

[0165] In this embodiment, the planar movement is an arc, for example, a survey is performed during a planar rotation.

[0166] In some embodiments, the present invention provides an apparatus for measuring semiconductor multilayer structures based on second harmonic waves, wherein the azimuthal angle of the sample can be adjusted by rotating the mounting table, and preferably, the azimuthal angle can be adjusted within a range of [0°, 360°].

[0167] In this embodiment, the azimuth angle of the sample is adjusted by rotation, i.e., the incident surface and the sample JPEG0007730578000051.jpg7170 components can be divided and analyzed. The adjustment range of the azimuth angle can be selected according to actual needs, and the range of [0°, 360°] is preferred.

[0168] In some embodiments, in the apparatus for measuring a semiconductor multilayer structure based on second harmonic waves provided by the present invention, the apparatus adjusts the scanning size of the light spot and the moving speed of the mounting table so that the time length during which the measurement point on the sample is irradiated with the laser light is in the range of [0.1 ms, 1 ms].

[0169] In this embodiment, by controlling the moving speed, the total length of time each measurement point is illuminated is in the range of [0.1 ms, 1 ms] and generally does not exceed one accounting cycle. This time length is generally short enough. Therefore, the internal charge distribution of the wafer at the measurement point has not yet changed. This is used to indicate the initial state of the wafer. For fixed-point measurement, there is a delay (generally in the range of [15 ms, 30 ms]) between the time when the laser light is fully illuminated and the time when the signal is collected. Therefore, the internal state of the wafer has already changed at the time corresponding to the initial value of the fixed-point measurement. Therefore, the technical means provided in this embodiment can significantly improve the accuracy of the initial value and can also be used to reduce noise and calibrate the results of fixed-point measurement.

[0170] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic generation, wherein the power of the light source is adjustable.

[0171] In this embodiment, the power of the laser light generated by the light source is adjustable. Adjusting the power controls the total number of photons arriving at the measurement area per unit time. The higher the power, the more stable the generated signal, but the greater the number of incident photons per unit time, which increases the impact on the measurement sample. The power value is maintained constant throughout the measurements of the same lot to ensure that the intensity of the light spot remains constant. However, the power may be changed as needed. Depending on the measurement purpose, for example, when irradiating laser light with different powers, it may be necessary to adjust the power of the light source if changes in the second harmonic wave need to be detected. Alternatively, depending on the sample being measured, for example, if the material layer penetrated by the light source is too thick or has a large absorption coefficient, it may be necessary to adjust the power to ensure that a sufficient number of photons arrive at the interface to be tested. When the power is low, the signal change rate is relatively slow, and information about the internal charge redistribution when the wafer is irradiated with the laser light is clearly visible. However, it takes a relatively long time for the signal to reach saturation. When the power is high, the signal change rate is relatively fast, and although there may be some loss of information at the initial timing of laser light irradiation, the time required for saturation is relatively short, improving the efficiency of the measurement. At the same time, when the power is high, the signal becomes more stable. In this case, as described in this specification, it is not "the same lot." The laser light source can adjust the power to achieve different measurement purposes.

[0172] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic waves, wherein the light source has a power range of [0, 1000 mW].

[0173] In this embodiment, preferably, the light source has an adjustable power range of [0, 1000mW].

[0174] In some embodiments, in the apparatus for measuring semiconductor multilayer structures based on second harmonic waves provided by the present invention, the beam parameters that the incident light path system can adjust include the incident angle and / or the incident light polarization direction.

[0175] In this embodiment, the beam parameters that can be adjusted by the incident light path system include the incident angle and / or the polarization direction of the incident light. The incident light path system can adjust the incident angle of the laser light. The incident angle mainly affects the P-polarized component of the second harmonic wave. This is because the P component is the sum of vectors in two directions, the incident direction and the perpendicular direction. Therefore, by adjusting the incident angle, a component in the second polarization tensor of the sample to be tested can be quantitatively analyzed.

[0176] In some embodiments, the present invention provides an apparatus for measuring semiconductor multi-layer structures based on second harmonic waves, wherein the incident angle can be adjusted in a range of [10°, 90°].

[0177] In this embodiment, the incident angle of the laser light is preferably adjusted within a range of [10°, 90°].

[0178] In some embodiments, in the apparatus for measuring semiconductor multi-layer structures based on second harmonic waves provided by the present invention, the polarization direction of the incident light can be P polarization or S polarization.

[0179] In this embodiment, the incident light path system can adjust the polarization direction of the incident laser light to P polarization or S polarization. In some cases, the output light may have certain characteristics, as described in the following embodiments.

[0180] In some embodiments, in the apparatus for surveying semiconductor multilayer structures based on second harmonics provided by the present invention, the second harmonic parameters that the output optical path system can adjust include the output light polarization direction, i.e., the half-wave plate allows only second harmonics in a predetermined polarization direction to pass through.

[0181] In this embodiment, the output optical path system uses a beam splitter to allow only the second harmonic wave in a predetermined polarization direction to pass through and focus for measurement, so that the P-polarized and S-polarized components of the second harmonic wave can be separated for noise reduction and specific data analysis.

[0182] In some embodiments, the present invention provides an apparatus for measuring semiconductor multilayer structures based on second harmonic waves, wherein the polarization direction of the emitted light is P polarization or S polarization.

[0183] In this embodiment, the polarization direction of the emitted light is P-polarized or S-polarized. The second harmonic in P-polarized light is excited by both the charge in the oxide layer and the interface state in the sample to be tested, while the second harmonic in S-polarized light is excited only by the charge in the oxide layer. Therefore, by analyzing the second harmonic in different polarization directions, the two electrical properties can be decoupled.

[0184] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic waves, the apparatus comprising one or more signal receiving systems.

[0185] In this embodiment, there is one or more signal receiving systems, which can receive the second harmonic and / or incident optical split signals processed by the launching system and use them for different functions by multiple sensing devices.

[0186] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonics, the apparatus having two signal receiving systems, a P signal receiving system and an S signal receiving system, where the P signal receiving system receives only second harmonics in the P polarization direction, and the S signal receiving system receives only second harmonics in the S polarization direction.

[0187] In this embodiment, the signal receiving system includes two P signal receiving systems and two S signal receiving systems, which are for receiving the second harmonic waves of the P polarization direction and the S polarization direction, respectively, and selectively receive the second harmonic waves.

[0188] In some embodiments, in the apparatus for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, when the azimuth angle is zero and the polarization direction of the incident light path system is P, a non-zero signal received by the S signal receiving system performs noise reduction on the second harmonic received by the P signal receiving system.

[0189] In this embodiment, when the azimuth angle is zero and the polarization direction of the incident light path is P, theoretically, no second harmonic wave is generated in the S direction. However, in actual measurements, due to slight deviations in the azimuth angle or the angle of the polarization plate, or the influence of other interference factors, the second harmonic wave from the S exit light path is not zero. Therefore, this signal value is used for noise reduction processing in the P exit light path. Note that this light path is not only used for noise reduction; in fact, when adjusting the azimuth angle during measurement, the second harmonic wave in the S direction is also meaningful for analyzing the symmetry or defects in the crystal structure of the sample.

[0190] In some embodiments, the present invention provides an apparatus for measuring semiconductor multilayer structures based on second harmonic waves, which has three signal receiving systems: a P signal receiving system, an S signal receiving system, and an incident light splitting signal system.

[0191] In this embodiment, the signal receiving system includes three parts: a P signal receiving system, an S signal receiving system, and an incident light splitting signal system. The incident light splitting signal system is a light beam directly introduced from the incident light, not the second harmonic wave measured.

[0192] In some embodiments, in the apparatus for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, when the azimuth angle is zero, a non-zero signal received by the S signal receiving system is used to perform noise reduction on the second harmonic received by the P signal receiving system, and the stability of the incident optical split signal system is monitored in real time and used for noise reduction.

[0193] In this embodiment, there are two noise reduction modes. When the azimuth angle is zero, a non-zero signal received by the S signal receiving system reduces noise on the second harmonic wave received by the P signal receiving system. This is the same as the noise reduction mode mentioned in the previous specification. At the same time, the stability of the incident light splitting signal system is monitored in real time and used for noise reduction. The incident light splitting signal system collects the split light beam of the incident light before it reaches the sample, mainly to monitor the fluctuation of the laser light power arriving at the sample surface. In conjunction with the imaging system's monitoring of the shape and size of the light spot, this is used to monitor the light intensity density at the light spot, thereby processing noise reduction on the second harmonic wave.

[0194] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic waves, the apparatus further comprising a sample preparation system for pre-treating the sample to alter the state of internal charge distribution in the sample.

[0195] In this embodiment, the method further includes varying the initial state of the survey point on the sample with a single excitation. In some cases, the initial state is varied, i.e., only "excited," until the sample is scanned, as needed. The variation of the initial state must be uniform and stable, and may be achieved by optical or electrical devices.

[0196] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic generation, in which the pre-processing system is a pump light source.

[0197] In this embodiment, the single excitation is realized by another light source, i.e., a pump light source. The pump light source excites the internal electrons in the sample, allowing them to move through the sample with sufficient energy and accumulate at the interface. Therefore, the pump light source must have a relatively wide power range and wavelength tunability. This technical solution is applicable to scanning and fixed-point measurement, and can be applied before and during measurement. Since the pump light source has a relatively small light spot, it can provide directional excitation to the point to be measured.

[0198] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic waves, wherein the pre-processing system is a flashlight.

[0199] In this embodiment, the excitation light source is a flashlight. Compared with the pump light source, the flashlight completely covers the entire wafer, so it affects the entire wafer. This technical solution is applicable to both fixed-point measurement mode and scanning measurement mode, and is only applied before measurement.

[0200] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonic waves, wherein the pre-treatment system is an electric field bias device for applying an electric field to the sample to change the charge distribution near the interface to be probed.

[0201] In this embodiment, the single excitation is achieved by applying a voltage to the sample sufficient to alter the internal electric field within the sample to be measured. That is, a bias voltage device is added to alter the initial state of the sample. This method applies an electric field across the sample that sufficiently affects the internal charge arrangement, causing charges to build up at the interface. Ultimately, the accumulated charge density is related not only to the applied electric field but also to the electrical properties (e.g., interface state density, fixed charge density, etc.) involved. Once the charges have stably built up, a scanning mode can effectively detect the charge distribution within the sample to be tested and evaluate the electrical properties of the entire scanned area. The electric field can be applied directly using a DC voltage device, an AC voltage device, or a magnetic field bias device or a corona device.

[0202] The DC voltage device takes advantage of the fact that charges of different polarities react differently to an electric field. When a DC voltage is applied, mobile charges (e.g., carriers, mobile ions, etc.) move in the direction of the electric field lines and eventually pile up at the field / surface. During this process, the presence of internal interface state defects or other electrical defects in the sample will result in some difference in the final charge pile-up state. This difference is accounted for by performing feedback analysis on the signal in second harmonic scanning mode. This device is used before or during second harmonic measurement.

[0203] AC voltage devices take advantage of the fact that different types of charges respond differently to AC voltage. For example, mobile charges move periodically with AC voltage (different types of mobile charges move at different speeds or displacements), while fixed charges do not. Furthermore, defects of the same type also respond differently to voltages of different frequencies. For example, interface states exhibit low resistance at high-frequency voltages but high resistance at low-frequency voltages. Therefore, applying AC voltages allows for more detailed analysis of various internal electrical properties of a sample. To ensure measurement accuracy, the frequency of the AC voltage device must be much smaller than the receiving frequency of the signal in the system.

[0204] In the two devices mentioned above, the sample is placed on a conductive tray (which is required to have as low an electrical resistance as possible), the tray is grounded, and a DC or AC voltage is applied to the other end of the sample by a non-contacting probe or a covered but not contacting conductive plate. It is important to note that when a voltage is applied by the conductive plate, a small hole must be provided for the light beam to pass through.

[0205] The magnetic field biasing device places the sample in a changing electromagnetic field, which induces an electric field that causes charges to move through the sample and accumulate at the sample's interface, where the applied magnetic field strength and the induced electric field strength are described by Maxwell's equations.

[0206] Regarding the corona device, a corona device with high voltage ionizes the water molecules and carbon dioxide molecules in the surrounding air, generating positive H3O2 + Ions and negative CO3 - Ions are generated, one of which may be focused and scattered onto the surface of the sample, if desired, to alter the surface potential of the sample, with the goal of redistributing charge by attracting or repelling internal charges in the sample.

[0207] In this embodiment, multiple excitation devices may be applied in combination with the pump light source or flashlight, and these excitation devices may operate independently and not interfere with each other.

[0208] In the various methods mentioned above, the scanning mode itself is independent of the "excitation" device. In the actual measurement process, a single scanning mode can be used, or the two can be combined to perform multiple types of measurements. This is because the characteristics of the sample processing interface state can be displayed by scanning and measurement. At the same time, the most notable advantages of the scanning method over single-point measurement are the increase in measurement points and the improvement of measurement efficiency and accuracy.

[0209] In some embodiments, in the apparatus for surveying semiconductor multilayer structures based on second harmonic waves according to the present invention, the time-varying characteristics and the space-varying characteristics simultaneously or individually satisfy the following: The time-varying features include an initial state that indicates the state of the sample before the photons are poured into the sample, a final state that indicates the state of the sample when the measurement is completed, and a time-series state that indicates the dynamic process of charge redistribution as the photons are poured into the sample. The spatial variation characteristics include normal points, which are measurement points whose second harmonic value is within a 5% range above and below the average value, and abnormal points, which are measurement points whose second harmonic value is outside the 5% range above and below the average value.

[0210] In this embodiment, the time-varying feature and the space-varying feature may be provided simultaneously, but are not necessarily provided simultaneously. The fixed point surveying mode with the time-varying feature and the scanning surveying mode with the space-varying feature may be used one after the other, simultaneously (performing the same survey on the same sample using devices with the same model number), or separately.

[0211] The time-varying features include the initial state, the final state, and the time-series state. The initial state refers to the initial state of the sample until measurement. The initial state may be the original state of the sample or the state of the sample after a single excitation. In other words, it is the state until measurement. The final state refers to the state of the sample when measurement ends. After measurement ends, the sample is no longer irradiated, but its state may be continuously changing. However, the device does not receive the newly generated second harmonic. Therefore, the state at the end of measurement cannot be considered the final state. Of course, if the measurement time is long enough, the state of the sample may not change further and may be stable when measurement ends. The time-series state refers to the process by which the sample dynamically changes from the initial state to the final state.

[0212] The spatial variation feature includes normal points and abnormal points. The normal points have second harmonic values ​​within 5% of the average value, i.e., they are close to the average value. The abnormal points have second harmonic values ​​outside the 5% range of the average value, i.e., they are significantly different from the average value.

[0213] Figures 3 and 7 show examples of combining the fixed-point measurement mode and the scanning measurement mode. Figure 3 shows fixed-point measurement of several points. Figure 3 shows the scanning measurement mode. Fixed-point measurement is used to quantitatively analyze the measured defect density, while scanning mode is used to relatively measure different measurement points. Therefore, by combining the two measurement modes, quantitative analysis of the defect density of the entire area of ​​a sample can be achieved. Figure 3 shows an example of data in the case of linear scanning. In actual measurement, different scanning modes can be adopted for different sample types. Therefore, corresponding data processing methods are required. For example, for patterned wafers, the area to be measured is extremely small (usually several times the size of the light spot). Therefore, when selecting the scanning mode, the scanning area is set to be slightly larger than the test piece (typically 50 micrometers x 50 micrometers). During data processing, signal points outside the test piece area are deleted. On the other hand, for unpatterned wafers, a diverse measurement method can be selected to ensure randomness and diversity in the points to be tested.

[0214] In some embodiments, in the apparatus for measuring semiconductor multi-layer structures based on second harmonic waves provided by the present invention, the central processing system further includes a fixed-point / scanning correspondence module, which is used to establish a correspondence between the fixed-point measurement mode and the scanning measurement mode, i.e., to calculate the second harmonic wave acquired in the scanning measurement mode. JPEG0007730578000052.jpg11170JPEG0007730578000053.jpg18170JPEG0007730578000054.jpg24170 versus movement speed.

[0215] In this embodiment, the central processing system further includes a fixed-scanning correspondence module for establishing a correspondence relationship between the fixed-point measurement mode and the scanning measurement mode. The combined method of the scanning measurement mode and the fixed-point measurement mode can indicate the electrical characteristics at the interface of the sample to be tested. The advantage of this measurement method is that it can not only perform noise reduction on the data results for the fixed-point measurement based on the accurate initial state measured by the scanning mode, but also enable reprocessing of the data in the scanning mode using the results of the fixed-point measurement mode.

[0216] In this case, what is acquired by the scanning surveying mode is the final state of the fixed point surveying mode. Generally, the scanning surveying mode is required to acquire the final state. The relative movement speed needs to be very slow, and it is related to the overall length of the fixed point surveying time, that is, the final state of the fixed point surveying. JPEG0007730578000057.jpg20170This is the initial state in amount mode.

[0217] In some embodiments, the present invention provides a device for measuring semiconductor multilayer structures based on second harmonic waves, wherein the second harmonic wave formula is as follows: JPEG0007730578000058.jpg11170JPEG0007730578000059.jpg14170 is the polar coordinate position with the center of the light spot as the origin, z is the vertical distance of the light spot from the interface to be tested, and t is the length of time the measurement point is irradiated with the laser light.

[0218] JPEG0007730578000060.jpg10170JPEG0007730578000061.jpg11170In the formula 2, P is the peak power of the laser light, and w is the beam waist width. JPEG0007730578000062.jpg6170This is written as Equation 3 below. JPEG0007730578000063.jpg11170JPEG0007730578000064.jpg37170

[0219] In this embodiment, when an ideal and uniform sample is irradiated with laser light, the internal charge distribution changes as shown in Figure 1, in two modes: fixed-point measurement and scanning measurement. Figure 1A shows the fixed-point measurement mode. As can be seen, the light intensity of the area being measured appears as a Gaussian distribution, so the internal charge redistribution due to the excitation of the laser light also becomes a Gaussian distribution. In this embodiment, the data is processed not by the formulas disclosed in the prior art, but by Equation 1, Equation 2, and Equation 3. Equation 1 is the JPEG0007730578000065.jpg26170 is the phase difference. As described above, luminous intensity is not uniformly distributed, and the changes in the electric field strength in the surveyed area are also different. Therefore, this embodiment provides a new formula, Equation 1. In Equation 1, the built-in electric field is the most important parameter and is the most fundamental reason for extracting the time series second harmonic.

[0220] Equation 2 is a conventional formula for Gaussian spot luminous intensity distribution. However, in conventional second-harmonic wave technologies, the luminous intensity distribution is calculated assuming the spot is a uniform spot, which causes errors in actual measurements. Therefore, the present invention uses this conventional formula to establish a more accurate physical model.

[0221] The electric field formula according to Equation 3 is more consistent with the scanning surveying mode because the distribution of incident photons during the scanning process is uniform, and the resulting charge accumulation is also uniform, which is consistent with the theoretical model in Equation 3. Here, two values ​​are particularly noted: JPEG0007730578000066.jpg6170 is sufficient to change the initial state of the sample being tested. Furthermore, N is the charge density accumulated at the interface during the measurement process, which is the change in charge distribution at the interface due to electrons being trapped by interface states or fixed charges during the transfer process. Equation 3 is a deeper development of the second-harmonic equation by the present invention and more accurately describes the built-in electric field at a semiconductor interface. While conventional second-harmonic techniques attribute changes in the built-in electric field to charge accumulation at the interface, in reality, the built-in electric field is the result of the combined action of various electrical defects and charge accumulation. However, the effect of electrical defects is described by the flat-band voltage (i.e., the first term on the right in Equation 3) and is adjusted by the external voltage. The effect of charge accumulation is described by the second term on the right in Equation 3. Therefore, as can be seen, the model described in Equation 3 is closer to the true case. Therefore, the fixed point survey mode has a better description role.

[0222] The combination of the scanning survey mode and the bias voltage technique allows for the The advantages of this combination for measurement are: (1) increasing the signal-to-noise ratio of the signal; (2) allowing the sample to be tested in different operating states (major carrier build-up, major carrier depletion, minor carrier build-up, depletion) to more accurately analyze the performance of the sample; and (3) applying different biases. The point is that JPEG0007730578000068.jpg13170.

[0223] The present invention provides calculations using Equations 1, 2, and 3, and unifies the actual measurement results with the theoretical model in three cases: fixed point measurement, scanning measurement, and a combination of fixed point measurement and scanning measurement. This allows the second harmonic measurement technology to be applied not only as a qualitative analysis method but also as a quantitative analysis method.

[0224] In some embodiments, the present invention provides an apparatus for surveying semiconductor multilayer structures based on second harmonics, wherein the central processing system includes a scanning signal graph creation module for creating a scanning signal graph, the scanning signal graph having the scanning direction as the abscissa and the value of the second harmonic as the ordinate. In the scanning signal graph, signal peaks and signal valleys are abnormal points in the data. The defect density at the abscissa of the peak value corresponding to the signal peak is greater than that of the surrounding area, the abscissa of the peak value is the defect center, the width of the signal peak is the size width of the defect, and the height of the signal peak is intended to indicate the defect density. The defect density at the abscissa of the valley value corresponding to the signal valley is smaller than that of the surrounding area, the abscissa of the valley value is the defect center, the width of the signal valley is the size width of the defect, and the height of the signal valley is intended to indicate the defect density.

[0225] In this embodiment, the scanning signal graph generated by the scanning signal graph generation module is shown in Figure 7, where the abscissa is the scanning direction and the ordinate is the scanned second harmonic. Since the second harmonic is related to defect density, in Figure 7, the defect density at the abscissa of the peak value corresponding to the signal peak is greater than that of the surrounding area, the abscissa of the peak value is the defect center, the width of the signal peak is the size width of the defect, and the height of the signal peak represents the defect density. The defect density at the abscissa of the valley value corresponding to the signal valley is smaller than that of the surrounding area, the abscissa of the valley value is the defect center, the width of the signal valley is the size width of the defect, and the height of the signal valley represents the defect density. Therefore, the scanning mode can be used to measure the relative defect distribution in a sample.

[0226] In some embodiments, the apparatus for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention further includes a defect density distribution graph creation module for creating a defect density distribution graph based on the value of the second harmonic and the coordinate position of the scanning survey point in the scanning survey mode. In the defect density distribution graph, bright points correspond to the signal peaks in the scanning signal graph and indicate that the defect density at the coordinate position of the scanning survey point corresponding to the bright points is higher than that of the surrounding areas. Dark points correspond to the signal valleys in the scanning signal graph and indicate that the defect density at the coordinate position of the scanning survey point corresponding to the dark points is lower than that of the surrounding areas.

[0227] In this embodiment, the defect density distribution graph obtained by the scanning surveying mode is shown in Figure 8. The defect density at the bright spot is higher than that at the surrounding area, which corresponds to the peak in Figure 7, and the defect density at the dark spot is lower than that at the surrounding area, which corresponds to the valley in Figure 7.

[0228] In some embodiments, the apparatus for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention further includes a time series second harmonic graph creation module, which is configured to create a time series second harmonic graph in the fixed point surveying mode, with the surveying time as the abscissa and the value of the second harmonic as the ordinate.

[0229] The time series second harmonic graph has the following characteristics: The initial state point is the first point in the time series second harmonic graph and indicates the state before the photons are poured into the sample. The final state point is the last point in the time series second harmonic graph and indicates the state of the sample when the survey ends. The time series state points are points between the initial state point and the final state point in the time series second harmonic graph, and represent the dynamic process by which the internal charge is redistributed as the photons are poured into the sample.

[0230] In this example, the three most noticeable aspects of the entire second harmonic curve are the initial state, the final state, and the time series state.

[0231] The initial state point indicates the state before photons are incident on the sample, i.e., the first point in Figures 3A, 3B, 3C, and 3D. At this time, there are no free electrons excited in the oxide layer, and the built-in electric field at this time is the initial built-in electric field.

[0232] The time series of state points shows the dynamic process of injecting photons into the sample and redistributing the charge, reflected as the process of electrons being captured by interface states or fixed charges. This appears as a line segment moving in the middle in Figures 3A, 3B, 3C, and 3D. During this process, the built-in electric field changes, generating a time series of second harmonics. The rate of change of the signal is related to the thickness of the oxide layer, the band gap width of the material, and the frequency of the incident photons. JPEG0007730578000069.jpg5170 Calculate the rate of increase of the second harmonic curve when JPEG0007730578000070.jpg11170JPEG0007730578000071.jpg29170, Equation 4 is close to 0, and the generated second harmonic does not change with time. Therefore, for a relatively thick oxide layer, increasing the frequency of the incident photons can increase the probability that the electron energy level transitions and reaches the surface of the oxide layer.

[0233] The final state point indicates the state after photons are injected into the sample and it becomes stable, or the state at the end of the measurement time. Stability means that during the measurement time, the excitation and binding of new free electrons are balanced, the internal charge distribution in the sample is dynamically balanced, and the built-in electric field remains unchanged. Even if the measurement time is extended, there is no significant change in the signal value. This is shown as the line segment where the signal becomes stable in Figures 3B, 3C, and 3D. However, when using low-power excitation light, if the sample does not yet become stable after the measurement is completed, if the measurement time is extended, the second harmonic will still tend to change over time. In this case, the last point measured, such as the last point marked in Figure 3A, can be defined as the final state point.

[0234] In some embodiments, in the apparatus for surveying a semiconductor multilayer structure based on second harmonics provided by the present invention, the central processing system further includes a point selection analysis module, which is configured to select some or all of the signal abnormality points and / or the survey points randomly selected in advance as the predetermined survey points and enter the fixed point survey mode.

[0235] In this embodiment, random points and / or special points may be selected as fixed point survey locations. Random points are randomly selected before surveying and are mainly used to calculate defect density. Special points are fixed point surveys performed on abnormal points (peaks or valleys) of the signal during the surveying process, with the purpose of conducting more detailed surveys and analysis of those locations. Surveys may be performed on random points and special points separately. If both random points and special points are surveyed together, they may be analyzed separately or together in the subsequent analysis process.

[0236] 9 shows a flowchart of a specific measurement process that comprehensively applies the techniques provided by the above embodiments. The measurement steps include inserting a wafer, adjusting parameters, selecting a scan line, moving the stage, measuring until all scanning and measurement tasks are completed, analyzing abnormal points, and entering fixed point measurement mode. After all the measurement is completed, the data is analyzed.

Claims

1. 1. A method for surveying a semiconductor multilayer structure based on second harmonic waves, comprising: generating a laser beam from a light source, passing through an incident light path system to form a light spot on a sample, and directing photons onto the sample, i.e., irradiating the sample with the laser beam; the sample is a probeable structure having two or more layers of materials and an interface therebetween, at least one of the materials being a semiconductor layer, and the light source is capable of reaching the interface to be probed; The sample and the light spot are capable of relative movement in a plane, and the relative movement In the scanning direction, the internal charge redistribution induced by the laser light is uniform, and the total time during which the laser light is irradiated and the total number of injected photons are the same at all measurement points; In the process of scanning and measuring the samples in the same lot, the shape and size of the light spot are maintained constant, the power of the light source is maintained constant, that is, when the shape and size of the light spot are both maintained constant, the luminous intensity of the light spot is also maintained constant, and the relative moving speed between the sample and the light spot is maintained constant; When the photon is absorbed by an electron and the charge distribution inside the sample changes, the change curve of the second harmonic wave received by the signal receiver from the exit optical path system can be described as a second harmonic wave equation, and the second harmonic wave is a filtered single wavelength light having the same exit angle and incident angle; The change curve of the second harmonic includes: a time-varying characteristic of the second harmonic at which the predetermined survey point occurred, recorded when entering the fixed point survey mode; and A method for surveying a semiconductor multilayer structure based on second harmonics, characterized in that when entering the scanning surveying mode, information is recorded during the relative movement process, and includes information on the spatial distribution characteristics of the second harmonics generated at the scanning surveying point.

2. 2. The method of claim 1, wherein the sample and the light spot have a degree of freedom in the relative movement on the plane in the X direction.

3. The method of claim 2 , wherein the relative movement further includes a degree of freedom in the Y direction.

4. The method of claim 1 , wherein the relative movement further comprises a rotation.

5. The method of claim 1 , wherein the relative movement is linear.

6. The method of claim 1 , wherein the relative movement is an arc.

7. 2. The method of claim 1, wherein the azimuthal angle of the sample can be adjusted in the range of [0°, 360°] by rotating the sample.

8. 2. The method of claim 1, wherein the sample and the light spot are adjustable in relative position in the Z direction.

9. 2. The method of claim 1, wherein the power of the laser light generated by the light source is adjustable.

10. 10. The method of claim 9, wherein the laser light has a power adjustment range of [0, 1000 mW].

11. The method of claim 1 , wherein the incident light path system is capable of adjusting the angle of incidence of the laser light.

12. The method according to claim 11, wherein the incident angle of the laser light is adjustable in the range of [10°, 90°].

13. 2. The method according to claim 1, wherein the incident optical path system is capable of adjusting the polarization direction of the laser light to P polarization or S polarization.

14. 2. The method of claim 1, wherein the signal receiving system is one or more.

15. The method of claim 14, wherein the signal receiving system includes two signal receiving systems, a P signal receiving system and an S signal receiving system, the P signal receiving system receiving only the second harmonic wave in the P polarization direction, and the S signal receiving system receiving only the second harmonic wave in the S polarization direction.

16. 16. The method of claim 15, wherein when the azimuth angle is zero and the polarization direction of the incident optical path system is P, a non-zero signal received by the S signal receiving system performs noise reduction on the second harmonic received by the P signal receiving system.

17. 15. The method of claim 14, wherein the signal receiving systems include a P signal receiving system, an S signal receiving system, and an incident light splitting signal system.

18. 18. The method of claim 17, wherein when the azimuth angle is zero, a non-zero signal received by the S signal receiving system performs noise reduction on the second harmonic received by the P signal receiving system, while monitoring the stability of the incident optical split signal system in real time and using it for noise reduction.

19. 19. The method according to any one of claims 1 to 18, further comprising changing the internal charge distribution state of the measurement point in the sample by a single excitation.

20. 20. The method of claim 19, wherein the single excitation is achieved by a pump light source.

21. 20. The method of claim 19, wherein the single excitation is achieved by a flashlight.

22. 20. The method of claim 19, wherein the single excitation is achieved by applying a voltage to the sample using a bias voltage device sufficient to alter a built-in electric field in the sample.

23. The time-varying features include: an initial state representing the state before the photons are injected into the sample; a final state indicating the state of the sample when the measurement ends; and a time series of states representing a dynamic process in which internal charges are redistributed as the photons are injected into the sample; The spatially varying features include: A normal point is a measurement point where the value of the second harmonic is within a range of 5% above and below the average value; and The method of claim 22, wherein the measurement points where the second harmonic value is outside a 5% range above or below the average value are included as abnormal points.

24. The method further includes establishing a fixed-scan correspondence between the fixed-point surveying mode and the scanning surveying mode, and combining the two modes to analyze electrical characteristics of an interface in the sample to be tested; The average value of the second harmonic wave acquired in the scanning survey mode and the fixed point survey mode In the scanning survey mode, the scanning survey points obtained by the scanning on the sample are

25. The average value corresponds to an initial value which is a value in an initial state in the fixed point survey mode.

25. The method of claim 24, wherein v is the relative movement velocity.

26. The second harmonic equation is: It is is the distance of the light point from the interface to be tested, and t is the length of time the measurement point is illuminated by the laser light;

27. the method includes taking a survey in the scanning mode and then creating a scan signal graph with the scanning direction as the abscissa and the value of the second harmonic as the ordinate; In the scanning signal graph, signal peaks and signal valleys are abnormal points of the data; The defect density on the abscissa of the peak value corresponding to the signal peak is higher than that of the surrounding area, the abscissa of the peak value is the defect center, the width of the signal peak is the size width of the defect, and the height of the signal peak indicates the defect density; 27. The method of claim 26, wherein the defect density at the abscissa of the valley value corresponding to the signal valley is smaller than that of the surrounding area, the abscissa of the valley value is the defect center, the width of the signal valley is the size width of the defect, and the height of the signal valley is used to indicate the defect density.

28. the method includes creating a defect density distribution graph based on the value of the second harmonic and the coordinate position of the scanning survey point in the scanning survey mode; The method of claim 27, wherein in the defect density distribution graph, bright points correspond to the signal peaks in the scanning signal graph, indicating that the defect density at the coordinate positions of the scanning survey points corresponding to the bright points is greater than that of the surrounding areas, and dark points correspond to the signal valleys in the scanning signal graph, indicating that the defect density at the coordinate positions of the scanning survey points corresponding to the dark points is less than that of the surrounding areas.

29. The method includes creating a time series second harmonic graph using the fixed point survey mode, with survey time as the abscissa and the value of the second harmonic as the ordinate; The time series second harmonic graph is a first point in the time series second harmonic graph, which is an initial state point indicating the state of the sample before the photons are irradiated; a final state point in the time series second harmonic graph that indicates the sample state when the survey ends; 27. The method of claim 26, characterized by points in the time series second harmonic graph between the initial state point and the final state point, the time series state points representing a dynamic process in which the internal charge is redistributed as the photons are poured into the sample.

30. 24. The method of claim 23, wherein the method includes selecting some or all of the signal anomaly points as the predetermined survey points and entering the fixed point survey mode.

31. 24. The method of claim 23, wherein the method includes randomly selecting the survey point as the predetermined survey point in advance and entering the fixed point survey mode.

32. 1. An apparatus for surveying semiconductor multilayer structures based on second harmonic waves, comprising: The device comprises: a light source for generating laser light, passing through an incident light path system to form a light spot on the sample, and injecting photons onto the sample, i.e., irradiating the sample with laser light; the incident light path system for adjusting the laser generated by the light source; The sample is a probeable structure having an interface formed of two or more layers of material, at least one of the layers being a semiconductor layer, and the light source is capable of reaching the interface to be probed; a mounting table whose movement direction is the scanning direction; an output optical path system for adjusting second harmonic parameters of the output second harmonic; a signal receiving system for receiving the second harmonic wave adjusted by the output optical path system, the change curve of the second harmonic wave being described as a second harmonic equation, the change curve of the second harmonic wave including information on the time change characteristics of the second harmonic wave generated at the predetermined measurement point, which is recorded when entering the fixed point measurement mode, and the spatial distribution characteristics of the second harmonic wave generated at the scanning measurement point, which is recorded during the process of relative movement when entering the scanning measurement mode; a monitoring system for monitoring the operating status of the device in real time and transmitting real-time feedback information; an input system for receiving input information from a user in a human-computer interaction; a display system for displaying output information of said device in a human-computer interaction; a central processing system for receiving the input information and the real-time feedback information, controlling the operation of the device, processing the second harmonic wave according to the second harmonic wave formula, and outputting the output information; In the direction of the scan, the internal charge redistribution induced by the laser light is uniform, and the total irradiation time and the total number of injected photons are the same at all measurement points; In the process of scanning and measuring the sample in the same lot, the shape and size of the light spot are maintained constant, the power of the light source is maintained constant, that is, when the shape and size of the light spot are both maintained constant, the luminous intensity of the light spot is also maintained constant, and the relative moving speed between the sample and the light spot is maintained constant.

33. 33. The apparatus for surveying a semiconductor multi-layer structure based on second harmonic waves according to claim 32, wherein the planar movement of the mounting table has a degree of freedom in an X direction.

34. 34. The apparatus for surveying a semiconductor multi-layer structure based on second harmonic waves according to claim 33, wherein the planar movement of the mounting table further has a degree of freedom in the Y direction.

35. 33. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as recited in claim 32, wherein the stage is further rotatable.

36. 33. The apparatus for surveying a semiconductor multilayer structure based on second harmonic waves according to claim 32, wherein the position of the mounting table is adjustable in a height direction.

37. 37. The apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 36, wherein in the scanning surveying mode, the position of the mounting table is adjusted in real time in the height direction correspondingly based on real-time feedback of the height of the surveying point on the sample.

38. 33. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as recited in claim 32, wherein said planar movement is linear.

39. 33. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as recited in claim 32, wherein the planar movement is an arc.

40. 33. The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as recited in claim 32, wherein the azimuthal angle of the sample can be adjusted by rotating the stage.

41. 41. The apparatus for surveying a semiconductor multilayer structure based on second harmonic waves as claimed in claim 40, wherein the azimuth angle has an adjustment range of [0°, 360°].

42. The apparatus for measuring a semiconductor multilayer structure based on second harmonics, as described in claim 32, characterized in that the apparatus is capable of adjusting the scanning size of the light spot and the moving speed of the mounting table so that the value of the length of time during which the measurement point on the sample is irradiated with the laser light is in the range of [0.1 ms, 1 ms].

43. 33. The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as claimed in claim 32, wherein the power of the light source is adjustable.

44. 44. The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as claimed in claim 43, wherein the power of the light source is in the range of [0, 1000 mW].

45. 33. The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as claimed in claim 32, wherein the light beam parameters that the incident light path system is able to adjust include the incident angle and / or the incident light polarization direction.

46. 46. ​​The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as claimed in claim 45, wherein the adjustment range of the incident angle is [10°, 90°].

47. 46. ​​The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as claimed in claim 45, wherein the incident light polarization direction can be P polarization or S polarization.

48. The apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 32, characterized in that the second harmonic parameters that can be adjusted by the output optical path system include the output light polarization direction, i.e., a half-wave plate is used to allow only second harmonics of a predetermined polarization direction to pass through.

49. 49. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as recited in claim 48, wherein the polarization direction of the emitted light is P polarization or S polarization.

50. 33. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as claimed in claim 32, characterized in that the apparatus has one or more signal receiving systems.

51. The apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 50, characterized in that the apparatus has two signal receiving systems, a P signal receiving system and an S signal receiving system, wherein the P signal receiving system receives only second harmonics in the P polarization direction, and the S signal receiving system receives only second harmonics in the S polarization direction.

52. 52. An apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 51, characterized in that when the azimuth angle is zero and the polarization direction of the incident optical path system is P, a non-zero signal received by the S signal receiving system performs noise reduction on the second harmonic received by the P signal receiving system.

53. 51. The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as claimed in claim 50, characterized in that the apparatus has three signal receiving systems: a P signal receiving system, an S signal receiving system and an incident light splitting signal system.

54. 54. The apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 53, characterized in that when the azimuth angle is zero, a non-zero signal received by the S signal receiving system is used to perform noise reduction on the second harmonic received by the P signal receiving system, while monitoring the stability of the incident light split signal system in real time and using it for noise reduction.

55. 55. An apparatus for measuring semiconductor multilayer structures based on second harmonics, as described in any one of claims 32 to 54, characterized in that the apparatus further comprises a sample preparation system for pre-treating the sample to change the internal charge distribution state of the sample.

56. 56. The apparatus for surveying semiconductor multilayer structures based on second harmonic waves as recited in claim 55, wherein said pre-processing system is a pump light source.

57. 56. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as recited in claim 55, wherein said pre-processing system is a flashlight.

58. the pre-treatment system is an electric field bias device; 56. An apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 55, wherein the electric field bias device is for applying a voltage to the sample sufficient to change the internal charge distribution state of the sample.

59. The time-varying features include: an initial state representing the state before the photons are injected into the sample; a final state indicating the state of the sample when the measurement is completed; a time series of states representing a dynamic process in which internal charges are redistributed as the photons are injected into the sample; The spatial distribution features include: A normal point is a measurement point where the value of the second harmonic is within a range of 5% above and below the average value. The apparatus for surveying a semiconductor multilayer structure based on second harmonics of claim 58, characterized in that abnormal points are included, which are survey points where the value of the second harmonic is outside a 5% range above or below the average value.

60. the central processing system further includes a fixed-point / scanning correspondence module for establishing a correspondence between the fixed-point surveying mode and the scanning surveying mode and combining the two modes to analyze electrical characteristics of an interface in a sample to be tested; The average value of the second harmonic wave acquired in the scanning survey mode and the fixed point survey mode 60. The apparatus for surveying semiconductor multi-layer structures based on second harmonic waves as recited in claim 59, wherein v is the relative movement speed.

61. The second harmonic equation is: is the polar coordinate position with the center of the light spot as the origin, z is the distance of the light spot from the interface to be tested in the vertical direction, and t is the length of time the measurement point is irradiated with the laser light; In the formula 2, P is the peak power of the laser light, and w is the beam waist width. A device for measuring semiconductor multilayer structures.

62. the central processing system includes a scan signal graphing module; the scanning signal graph creation module is for creating a scanning signal graph with the scanning direction as the abscissa and the value of the second harmonic as the ordinate; In the scanning signal graph, signal peaks and signal valleys are abnormal points of the data; 62. The apparatus for surveying a semiconductor multilayer structure based on second harmonics of claim 61, wherein the defect density at the abscissa of the peak value corresponding to the signal peak is greater than that of the surrounding area, the abscissa of the peak value is the defect center, the width of the signal peak is the size width of the defect, and the height of the signal peak is intended to indicate the defect density; and the defect density at the abscissa of the valley value corresponding to the signal valley is smaller than that of the surrounding area, the abscissa of the valley value is the defect center, the width of the signal valley is the size width of the defect, and the height of the signal valley is intended to indicate the defect density.

63. The apparatus further includes a defect density distribution graph creation module; the defect density distribution graph creation module is for creating a defect density distribution graph based on the value of the second harmonic and the coordinate position of the scanning survey point in the scanning survey mode; 63. The apparatus for surveying semiconductor multilayer structures based on second harmonics, as described in claim 62, characterized in that in the defect density distribution graph, bright points correspond to the signal peaks in the scanning signal graph, indicating that the defect density at the coordinate positions of the scanning survey points corresponding to the bright points is greater than that of the surrounding areas, and dark points correspond to the signal valleys in the scanning signal graph, indicating that the defect density at the coordinate positions of the scanning survey points corresponding to the dark points is less than that of the surrounding areas.

64. The apparatus further includes a time series second harmonic graphing module; the time series second harmonic graph creation module is for creating a time series second harmonic graph in the fixed point survey mode, with survey time as the abscissa and the value of the second harmonic as the ordinate; The time series second harmonic graph is a first point in the time series second harmonic graph, which is an initial state point indicating the state before the photons are incident on the sample; a final state point in the time series second harmonic graph that indicates the state of the sample when the survey ends; 62. The apparatus for measuring semiconductor multilayer structures based on second harmonics of claim 61, characterized by points between the initial state point and the final state point in the time series second harmonic graph, which are time series state points that indicate the dynamic process of the internal charge being redistributed as the photons are injected into the sample.

65. The central processing system further includes a point selection analysis module; The apparatus for surveying a semiconductor multilayer structure based on second harmonics, as described in claim 59, characterized in that the point selection analysis module is configured to select some or all of the signal anomaly points and / or the survey points selected randomly in advance as the predetermined survey points and enter the fixed point survey mode.

Images