A rapid evaluation method for the particulate matter removal efficiency

Abstract

The present invention discloses a rapid evaluation method for the particle removal efficiency, which is used to evaluate the particle removal efficiency of the surface of a wafer after chemical mechanical polishing in semiconductor manufacturing, and includes: cutting the wafer to obtain a wafer sample, conducting a contamination experiment to obtain a contaminated wafer sample to be cleaned; conducting a topography analysis to obtain a three-dimensional topography image before cleaning; performing a cleaning experiment with a cleaning solution to obtain the cleaned wafer sample; obtaining a three-dimensional topography image after cleaning; respectively intercepting cross-sections at a height H in the vertical direction of the obtained three-dimensional topography images before and after cleaning, correspondingly obtaining images of the particle pollutant conditions on the wafer surface before and after cleaning, analyzing and processing the images to obtain the number or area of particle pollutants before and after cleaning, and calculating the cleaning efficiency, wherein the height H is greater than the average roughness of the surface of the wafer sample and less than half of the average particle size of the particle pollutants. The evaluation method of the present invention can achieve a rapid and low-cost evaluation of the removal efficiency.

Description

Technical Field

[0001] The present invention relates to the post - treatment field of semiconductor production and manufacturing. More specifically, the present invention relates to a method for rapidly evaluating the particulate removal efficiency. Background Art

[0002] In the process of semiconductor production and manufacturing, as the feature size continues to decrease and the wafer size increases, Chemical Mechanical Polishing (CMP) has become the main means to achieve high - level planarization of the wafer surface. After the wafer is processed by CMP, a large amount of pollutants such as metal ions, particles, and organic substances will remain on the wafer surface. The residual organic substances will damage the adhesion of the wafer surface, making the film layers in the device structure easy to peel off and affecting the reliability of the device. Particles adsorbed on the wafer surface may cause short - circuit or open - circuit of the copper interconnection circuit, thus damaging the device performance or even causing failure. Metal ions and their oxides mainly affect the resistance of the copper interconnection, thereby affecting the RC delay of the copper interconnection. Therefore, after CMP, a cleaning solution is needed to remove pollutants as much as possible. The particle removal ability (PRE) of the cleaning solution for the particulate pollutants after CMP is an important indicator for evaluating its cleaning effect. For the particulate pollutants on the wafer surface after CMP, they are mainly the abrasive particles in the abrasive slurry used in the CMP process, such as Al 3 O 2 、CeO 2 ,SiO 2 etc., and SiO 2 is the most commonly used; in addition, there are also a small amount of corrosion products such as Cu 2 O and CuO formed by the reaction of the oxidant component in the polishing solution with copper.

[0003] To study the particle removal ability of the cleaning solution for the particulates after CMP, equipment and process conditions equivalent to the chip production process are usually adopted, as shown in Figure 1 . After the wafer is polished and dried on a CMP machine (such as Reflexion LK of Applied Materials), the overall defects of the wafer are detected by a wafer defect detection system (such as SURFSCAN SP2 of KLA Tencor), and then a certain number of defects are randomly selected and photographed by an SEM device at a magnification of more than 10,000 times (such as SEM VisionG4 MAX of Applied Materials), and the corresponding defects are manually classified (such as SiO 2, organic substances, scratches, etc.), calculate the number of particulate matters on the wafer surface according to the proportion, then clean it with a cleaning solution on a CMP machine tool, dry it, perform defect detection and classification again, and finally calculate the cleaning ability of the cleaning solution for the particulate matters.

[0004] The above method is close to the actual working conditions and conditions of the cleaning solution, but this method has the following obvious disadvantages: it requires a large amount of expensive production equipment, which may affect normal production activities; the operation of the detection equipment is relatively cumbersome, and experienced engineers are required to manually classify and identify the wafer defects and conduct statistics, which takes a long time; the detection cost is high, and a large amount of cleaning solution and wafers (usually 12 inches or 8 inches) are required for each test; because manual random statistics are carried out by SEM at a high magnification, the sampling range and the number of statistics are very small, and it is greatly affected by the operator, and the result repeatability is very poor. Using this method is restricted by many conditions and is not conducive to the rapid evaluation, screening and analysis of the cleaning solution. Therefore, there is an urgent need to find a method that can replace the existing evaluation method and can accurately, efficiently and low-cost evaluate the particulate matter removal efficiency. Summary of the Invention

[0005] An object of the present invention is to solve the above problems and provide the advantages described later.

[0006] Another object of the present invention is to provide a rapid evaluation method for particulate matter removal efficiency, which realizes the rapid and low-cost evaluation of particulate matter removal efficiency and the screening of cleaning solutions through five steps: pretreatment, contamination, cleaning, data collection and data processing.

[0007] To achieve these and other advantages in accordance with the present invention, there is provided a rapid evaluation method for particulate matter removal efficiency, comprising the following steps:

[0008] 1) Cut the wafer to obtain a wafer sample, and perform a contamination experiment on the surface of the wafer sample with particulate contaminants of a known particle size to obtain a wafer sample to be cleaned;

[0009] 2) Use a morphology characterization instrument to perform morphology analysis on the surface of the wafer sample to be cleaned to obtain a three-dimensional morphology image before cleaning;

[0010] 3) Perform a cleaning experiment on the surface of the wafer sample to be cleaned with a cleaning solution to obtain a cleaned wafer sample;

[0011] 4) Use a morphology characterization instrument to perform morphology analysis on the cleaned wafer sample again to obtain a three-dimensional morphology image after cleaning;

[0012] 5) Process the three-dimensional topography images before and after cleaning. The processing method is to respectively intercept the cross-section at height H in the vertical direction of the obtained three-dimensional topography images before and after cleaning, and correspondingly obtain the images of the particle contaminant conditions on the wafer surface before and after cleaning. Then, perform data processing on the images to obtain the number or area of particle contaminants before and after cleaning within the image area;

[0013] Among them, the height H is greater than the average roughness of the wafer sample surface and less than half of the average particle size of the particle contaminants;

[0014] 6) Calculate the removal efficiency PRE through the following formula:

[0015]

[0016] Preferably, in step 5), the height H is not less than the maximum roughness of the wafer sample surface and not greater than half of the minimum particle contaminant size. Among them, the maximum roughness is the maximum surface height in the 95% confidence interval when the rough state follows a normal distribution, and the minimum particle contaminant size is the minimum particle size in the 95% confidence interval when the particle contaminant size follows a normal distribution.

[0017] Preferably, in step 1) the contamination experiment / in step 3) the cleaning experiment, the method is: Place the wafer sample on a spin coater, evenly drop the particulate suspension / cleaning solution for contamination / cleaning, the dropping speed is 0.1 - 20 mL / min, the dropping time is 0.1 - 10 min, rotate the spin coater while dropping, the rotation rate is 50 - 5000 rpm, after dropping, adjust the rotation rate of the spin coater to 100 - 10000 rpm, and spin dry the moisture on the wafer sample surface, the time is 0.1 - 10 min. More preferably, place the wafer sample on a spin coater, evenly drop the particulate suspension / cleaning solution for contamination / cleaning, the dropping speed is 1 - 5 mL / min, the dropping time is 1 - 3 min, rotate the spin coater while dropping, the rotation rate is 500 - 800 rpm, after dropping, adjust the rotation rate of the spin coater to 1000 - 1500 rpm, and spin dry the moisture on the wafer sample surface, the time is 1 - 2 min.

[0018] Preferably, in step 1) the particulate suspension in the contamination experiment is a 0.01 - 10 wt% silica sol, cerium dioxide sol or aluminum oxide sol. More preferably, it is a 1 wt% silica sol.

[0019] Preferably, the topography characterization instrument is a 3D optical surface profiler.

[0020] Preferably, the method for cutting and processing the wafer in step 1) is as follows: First, cut the wafer into small wafers, then soak the small wafers in a pretreatment solution for 1 - 8 minutes, rinse with high-purity water for 1 - 60 seconds, dry with nitrogen, and then conduct a contamination experiment on the obtained wafer sample. More preferably, first cut the wafer into 2 cm × 2 cm small wafers, then soak the small wafers in a pretreatment solution for 3 - 8 minutes, rinse with high-purity water for 8 - 15 seconds, dry with nitrogen, and then conduct a contamination experiment on the obtained wafer sample.

[0021] Preferably, the pretreatment solution is a 1 wt% citric acid solution.

[0022] Preferably, position marks are provided on the wafer sample in step 1), and the three-dimensional topography images before cleaning obtained in step 2) and the three-dimensional topography images after cleaning obtained in step 4) are taken of the same area marked by the position marks.

[0023] Preferably, the size of the wafer sample is 2 cm × 2 cm, and the three-dimensional topography images before / after cleaning are images of a 3 mm × 3 mm marked area obtained by splicing multiple pictures continuously taken corresponding to the situation before / after cleaning.

[0024] Preferably, the evaluation method is used to evaluate or screen cleaning fluids, and evaluates the removal efficiency of particulate matter on the surface of wafers after chemical mechanical polishing in semiconductor manufacturing by the cleaning fluids.

[0025] Among them, the cutting and processing includes steps such as dividing and cleaning traditional wafers. The wafers include but are not limited to silicon wafers with copper grown on the surface. Divide the wafers into wafers of the same size for subsequent operations. Cleaning the wafers means using various chemical reagents to clean the divided wafers to remove contaminants or oxide residues on the wafer surface and obtain a relatively clean wafer surface.

[0026] The contamination experiment is to contaminate the pretreated wafers with particulate contaminants. The particulate contaminants include but are not limited to the types, particle sizes, and concentrations of abrasive particles in common polishing fluids in semiconductor processes. The contamination methods include but are not limited to various methods such as soaking, spraying, and spin coating, and the preferred contamination method is spin coating using a spin coater.

[0027] The cleaning experiment is to clean the contaminated wafers with a cleaning fluid. The cleaning methods in the cleaning experiment include but are not limited to various methods such as soaking, spraying, and spin dropping, and the preferred cleaning method is spin dropping using a spin coater.

[0028] Data processing includes data acquisition. Preferably, a 3D optical surface profiler (white light interferometer) is used to capture the three-dimensional surface topography of the wafer surface before cleaning (after the contamination experiment) and after cleaning. There are several advantages to using a 3D optical surface profiler to capture the wafer surface topography: small magnification, only a 10-fold magnification is required for the surface during shooting, so the single-shot photo has a large shooting range; high resolution, the resolution of the 3D optical surface profiler in the Z direction (height direction) can reach 0.01 nm; seamless stitching can be achieved to capture the three-dimensional surface topography in a larger range.

[0029] Data processing also includes processing the wafer surface topography captured by the above 3D optical surface profiler to obtain quantitative data that can characterize the degree of wafer surface contamination.

[0030] The present invention has at least the following beneficial effects:

[0031] The method for rapidly evaluating the particulate removal efficiency of the present invention is simple and fast, and does not require expensive production equipment; has low technical requirements for operators, good repeatability and high reliability of results; and low test costs, without the need for a large number of wafers and cleaning fluids.

[0032] White light interferometer is used instead of SEM for data acquisition to capture the three-dimensional surface topography, which can achieve rapid and high-resolution acquisition of the three-dimensional surface topography. It solves the disadvantages of large magnification, small shooting range, the need for manual identification and classification statistics of defects, and long detection cycle when using SEM. The method of obtaining the cross-sectional image of the three-dimensional surface topography is used to convert the three-dimensional data into two-dimensional data, and an automatic image processing method is used to complete the quantitative analysis of the particulate contaminants on the wafer surface, realizing simple and rapid data analysis.

[0033] In the present invention, a cross-sectional image at a height H in the vertical direction is obtained to get the three-dimensional topography image, and according to the cleaning principle of particulate contaminants, the height H is defined to be greater than the average roughness of the wafer sample surface and less than half of the average particle size of the particulate contaminants, so as to obtain an accurate cleaning image at the height H, and the calculation of the cleaning efficiency is more accurate and efficient.

[0034] Since it usually takes a long time to evaluate a formulation in the traditional way, including the preparation time of the cleaning solution, the pre-value test, the analysis time, the post-value test and the analysis time, which totals more than 10 hours, while the method of the present invention only takes about 1 hour, so the method of the present invention can improve the evaluation efficiency by more than 90%. In terms of cost, excluding labor costs, the traditional test method uses one 12-inch wafer each time, and the consumption of the cleaning solution is about 20L. When using the method of this embodiment to cut the wafer, only a 2cm×2cm wafer is needed for each test, and only about 20mL of the cleaning solution is needed for each test, and the investment in fixed test equipment is also reduced a lot, and the overall test cost is saved by more than 95%. The method of the present invention is more obvious in improving the evaluation efficiency and reducing the cost when testing in large batches, and has wide popularization and application value.

[0035] Other advantages, objectives and features of the present invention will be partially reflected by the following description, and partially will also be understood by those skilled in the art through the research and practice of the present invention. Brief Description of the Drawings

[0036] Figure 1 The technical roadmap for evaluating the cleaning solution removal efficiency for the prior art;

[0037] Figure 2 The technical roadmap for the rapid evaluation method of the cleaning solution removal efficiency adopted by the present invention;

[0038] Figure 3 The schematic diagram of the principle of data processing of the present invention;

[0039] Figure 4 The surface topography diagram of the wafer taken during the implementation of Example 1 of the present invention;

[0040] Figure 5 The image of the particulate contamination condition on the wafer surface obtained by data processing of Example 1 of the present invention. Detailed Description of the Invention

[0041] The following further describes the present invention in detail with reference to the drawings and embodiments, so that those skilled in the art can implement it according to the description in the specification.

[0042] It should be noted that in the description of the present invention, the orientation or positional relationship indicated by the terms "horizontal", "longitudinal", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present invention and simplifying the description, and does not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation to the present invention.

[0043] The rapid evaluation method for the particulate removal efficiency of the present invention can be extended and applied to any occasion for evaluating the removal efficiency of surface particulate contaminants, and thus should be included within the scope of protection of the present invention.

[0044] In the present invention, unless otherwise specified, the concentrations of all solutions refer to mass percentages.

[0045] The wafers used in the present invention are 12-inch copper-plated silicon wafers adopted in semiconductor manufacturing, and the copper plating thickness is about 700 nm.

[0046] The spin-coating equipment used in the present invention is a spin coater that can achieve quantitative and uniform contamination and cleaning of particulate contaminants and cleaning solutions.

[0047] The technical roadmap for evaluating the cleaning solution removal efficiency in the prior art is as Figure 1 shown. Polishing / cleaning is performed on a CMP machine tool, and images are taken before and after cleaning by combining a wafer defect detection system and an SEM. The number of defects before and after cleaning is classified and counted according to the defect type, so as to calculate the cleaning efficiency.

[0048] The technical roadmap for the rapid evaluation method for the cleaning solution removal efficiency adopted in the present invention is as Figure 2 shown, which includes five steps: pretreatment, contamination, cleaning, data collection (before and after cleaning), and data processing. Specific embodiments are as follows.

[0049] Example 1

[0050] 1) Cut a 12-inch copper wafer into square wafers with a size of 2 cm × 2 cm, and make certain indentations on the wafers. For example, when the shooting position is the 3 mm × 3 mm area in the center part, mark the linear indentation at the lower left corner of the center area. Each time when shooting, select the upper right corner of the marked position as the starting point for continuous shooting, and then splice them into a 3 mm × 3 mm picture. The purpose of marking is to ensure that the pre-value after contamination and the post-value after cleaning are taken from the same area, improving the repeatability of the test results. Immerse the prepared wafers in a 1 wt% citric acid solution for 5 min, then rinse with high-purity water for 10 s, and dry with N2 to obtain wafer samples.

[0051] Place the above-mentioned clean wafer samples on the spin coater, and uniformly drop 1 wt% silica sol with a particle size of 50 ± 5 nm for spin-coating contamination. The dropping speed is 2 mL / min, the dropping time is 1 min, and the spin coater rotates while dropping, with a rotation rate of 500 rpm. After dropping, increase the rotation speed of the spin coater to 1000 rpm to spin-dry the water on the wafer surface for 1 min.

[0052] 2) Place the contaminated wafer in a Bruker 3D optical surface profiler to capture the wafer surface 3D morphology. The Bruker 3D optical surface profiler can achieve seamless splicing of multiple 3D surface morphologies, so it can quickly capture the 3D surface morphology of a large area. The shooting area in this embodiment is the upper right corner area of ​​the pre-marked position, with a size of 3mm×3mm. 42 images are collected, which takes about 4 minutes. The obtained 3D surface morphology of the wafer is shown in Figure 4 .

[0053] 3) The wafer after data collection is placed in a spin coater again, and the cleaning solution to be evaluated is evenly dripped, with a dripping speed of 5mL / min and a dripping time of 2min. The spin coater rotates at a uniform speed while dripping, and the rotation rate is 500rpm. After the dripping is completed, the spin coater speed is increased to 1000rpm to spin-dry the surface moisture of the wafer, and the time is 1min. This method avoids the problem that the physical effects of the actual cleaning process of the wafer are easily ignored in the research of the cleaning solution capacity by the static soaking method. These physical effects may include the impact of the liquid in the acid tank machine, the centrifugal force provided by the spin in the single-chip cleaning machine, and the process of drying by the spin mode, etc. The present invention better simulates the cleaning process of the wafer in the actual production by this method of dripping and rotating, and the evaluation method is more scientific and accurate.

[0054] 4) The cleaned wafer is photographed again at the same position of the wafer surface in three-dimensional shape according to step 3).

[0055] 5) Processing the three-dimensional morphology of the wafer surface collected in step 3) and step 4)

[0056] The basis and principle diagram of the data processing method of the present invention is as follows: Figure 3 As shown. After being contaminated by particles, the surface of the chip will become uneven. The surface morphology of the chip after contamination can be obtained by using surface analysis equipment. The surface roughness caused by the particle contaminants on the chip surface is much greater than the roughness of the chip itself. The obtained three-dimensional surface morphology is cut into a cross section of height H in the vertical direction. The cut height H is 10nm, and an image of the chip surface contaminated by particles can be obtained.

[0057] The height H is greater than the average roughness of the surface of the wafer sample itself, but less than half of the average particle size of the contaminants. If the height H is less than the average roughness of the surface of the wafer sample itself, the captured image may not reflect the actual cleaning situation, and the calculated cleaning efficiency may be too low. If the height H is greater than half of the average particle size of the contaminants, the defects in the image may not be fully displayed, and the calculated cleaning efficiency may be too high.

[0058] Preferably, it is greater than the maximum roughness of the wafer surface itself and less than half of the minimum particle contaminant size, where the maximum roughness and the minimum particle contaminant size are the corresponding values of the 95% confidence interval obtained by assuming that the rough state of the wafer surface and the particle contaminant size follow a normal distribution.

[0059] For example, when assuming that the rough state of the wafer surface and the particle contaminant size follow a normal distribution, the average surface roughness of the wafer sample surface before contamination is 2 nm, and the maximum surface height is 3.6 nm under the 95% confidence interval (standard deviation is 1); if the average particle contaminant size selected is 50 nm, the minimum particle size of the minimum surface height is 40.2 nm under the 95% confidence interval (standard deviation is 5). Therefore, the preferably certain intercept height range is [3.6 nm, 20.1 nm], and the probability that the protrusions greater than the intercept height obtained within this preferred range are particle contaminants is greater than 95%. Therefore, a suitable height range can be selected by calculation to obtain a relatively accurate image of the contaminated situation of the wafer surface. The surface topography image actually taken by the surface analysis equipment is shown in Figure 4 , and the image of the particulate contamination situation obtained after the intercept height H is 10 nm is shown in Figure 5 , and the image has been contrast-adjusted. For the image similar to Figure 5 obtained after processing, particulate matter analysis can be easily carried out through some scientific image analysis software, such as ImageJ, etc., so as to obtain the number or area of particle contaminants on the wafer surface before and after cleaning. The area here is the equivalent area corresponding to the number of particulate matters and is dimensionless.

[0060] By comparing the number or area of particle contaminants before and after cleaning, the removal efficiency of the cleaning liquid for particulate matter can be easily obtained, and its calculation formula is as follows:

[0061]

[0062] Specific implementation effects:

[0063] To verify the reliability of this technical solution, the commonly used X (citric acid - water system) cleaning liquid, Y (organic base - water system) cleaning liquid, and Z (water) were specifically selected. Under the experimental conditions, the removal efficiencies of the three are X > Y > Z. Further verification was carried out through the on - machine - platform tests of the traditional evaluation methods and the method in Example 1, and the control groups x group, y group, z group and the corresponding experimental groups X group, Y group, and Z group were obtained. The test results are shown in Table 1 below.

[0064] Table 1

[0065] Group Before cleaning After cleaning Removal efficiency X 9146 332 96.37％ x 2239 144 93.55％ Y 7543 1073 85.77％ y 2795 299 89.31％ Z 8219 5719 30.42％ z 2601 1673 35.66％

[0066] From the data comparison of different implementation schemes in Table 1, the evaluation method of calculating the cleaning efficiency by the number of particulate pollutants before and after cleaning in the present invention is similar to the results of the existing evaluation method, which proves that the method of the present invention is scientific and accurate.

[0067] To further verify the reliability and reproducibility of the present technical solution, two cleaning liquids, A and B, are selected. The method in Example 1 of the present technical solution is used to calculate the removal efficiency according to the number of particulate contaminants, and three parallel tests are carried out to obtain the results of Group A and Group B. The traditional evaluation method takes a long time. The evaluation test of comparing the particulate removal efficiency is carried out by the traditional evaluation method, and two parallel tests are carried out to obtain the evaluation results of Group a and Group b respectively. The corresponding difference between Group A' and Group B' and Group A and Group B is that the data before and after cleaning are used to calculate the removal efficiency based on the area of particulate pollutants, and the other conditions are the same. Among them, the area of particulate pollutants is the equivalent area corresponding to the number of particulate matters, dimensionless. The test results are shown in Table 2 below.

[0068] Table 2

[0069]

[0070] From the data comparison of different implementation schemes in Table 2, the results of calculating the cleaning efficiency by the number or area of particulate pollutants before and after cleaning in the present technical solution are accurate and the method is reliable. At the same time, the reproducibility of the evaluation results of the cleaning efficiency of particulate pollutants by the method of the present invention is significantly better than the existing evaluation methods, and the results are reliable and efficient.

[0071] Further comparative experiments are carried out. In the comparison groups B1 and B2, the cleaning liquid B is also selected. Except that cross-sections with heights of 2 nm and 30 nm are respectively intercepted in the vertical direction to obtain images, other evaluation conditions are the same as those in Group B. The test results are shown in Table 3 below:

[0072] Table 3

[0073]

[0074] As can be seen from Table 3, in the comparison groups B1 and B2, the results measured from the images not intercepted at the heights defined in the present invention are significantly less accurate than those in Group B, resulting in the calculated removal efficiency being significantly too large or too small, while the evaluation results of the groups using the method of the present invention are more reliable and accurate.

[0075] For the groups C, D, E, F, G, and H adopting the present technical solution, different cleaning liquids C, D, E, F, G, and H are correspondingly used. The method in Example 1 is used to verify the removal efficiency of each cleaning liquid for particulate pollutants. Three parallel tests are carried out for each group of cleaning liquids, and the test results are shown in Table 4 below:

[0076] Table 4

[0077]

[0078] As can be seen from Table 4, for multiple groups of cleaning solutions with different formulations, the evaluation method of the present invention can achieve a rapid evaluation of the cleaning efficiency of the cleaning solutions and the screening of the cleaning solutions. The selected groups D, E, and H have also been proven to have good cleaning effects in subsequent practical applications, thus proving again that the evaluation method of the present invention is rapid, accurate, and can be used to screen cleaning solutions with high removal efficiency.

[0079] Although the embodiments of the present invention have been disclosed as above, they are not limited to the applications listed in the specification and embodiments. It can be fully applied to various fields suitable for the present invention. For those familiar with the field, additional modifications can be easily achieved. Therefore, without departing from the general concept defined by the claims and the equivalent scope, the present invention is not limited to specific details and the examples shown and described herein.

Claims

1. A rapid evaluation method for the particulate removal efficiency, characterized in that, it comprises the following steps: 1) Cut the wafer to obtain a wafer sample, conduct a contamination experiment on the surface of the wafer sample with particulate contaminants of a known particle size to obtain the crystal sample to be cleaned; 2) Use a morphology characterization instrument to conduct a morphology analysis on the surface of the crystal sample to be cleaned to obtain a three-dimensional morphology image before cleaning; 3) Conduct a cleaning experiment on the surface of the crystal sample to be cleaned with a cleaning solution to obtain the crystal sample after cleaning; 4) Use a morphology characterization instrument to conduct a morphology analysis on the crystal sample after cleaning again to obtain a three-dimensional morphology image after cleaning; 5) Process the three-dimensional morphology image before cleaning and the three-dimensional morphology image after cleaning. The processing method is to respectively intercept the cross-section at a height H in the vertical direction from the obtained three-dimensional morphology image before / after cleaning, and correspondingly obtain the images of the particulate contaminants on the wafer surface before / after cleaning, and then conduct data processing on the images to obtain the number or area of the particulate contaminants before and after cleaning in the image area; wherein, the height H is greater than the average roughness of the wafer sample surface and less than half of the average particle size of the particulate contaminants; 6) Calculate the removal efficiency PRE through the following formula:

2. A rapid evaluation method for the particulate removal efficiency according to claim 1, characterized in that, in step 5), the height H is not less than the maximum roughness of the wafer sample surface and not greater than half of the minimum particle size of the particulate contaminants, wherein the maximum roughness is the maximum surface height in the 95% confidence interval in the case where the rough state follows a normal distribution, and the minimum particle size of the particulate contaminants is the minimum particle size of the particulate matter in the 95% confidence interval in the case where the particle size of the particulate contaminants follows a normal distribution.

3. A rapid evaluation method for the particulate removal efficiency according to claim 1, characterized in that, the method in step 1) contamination experiment / step 3) cleaning experiment is: place the wafer sample on a spin coater, uniformly drop the particulate suspension / cleaning solution for contamination / cleaning, the dropping speed is 0.1 - 20 mL / min, the dropping time is 0.1 - 10 min, rotate the spin coater while dropping, the rotation rate is 50 - 5000 rpm, after the dropping is completed, adjust the rotation rate of the spin coater to 100 - 10000 rpm, and spin dry the moisture on the surface of the wafer sample, the time is 0.1 - 10 min.

4. A rapid evaluation method for the particulate removal efficiency according to claim 3, characterized in that, the particulate suspension in step 1) contamination experiment is 0.01 - 10 wt% silica sol, cerium dioxide sol or aluminum trioxide sol.

5. A rapid evaluation method for the particulate removal efficiency according to claim 1, characterized in that, the morphology characterization instrument is a 3D optical surface profiler.

6. A rapid evaluation method for the particulate removal efficiency according to claim 1, characterized in that, the method for cutting and processing the wafer in step 1) is: first cut the wafer into small wafers, then soak the small wafers in a pretreatment solution for 1 - 8 min, rinse with high-purity water for 1 - 60 s, and dry with nitrogen, and then conduct the contamination experiment on the obtained wafer sample.

7. A rapid evaluation method for the particulate removal efficiency as described in claim 6, characterized in that, the pretreatment solution is a 1 wt% citric acid solution.

8. A rapid evaluation method for the particulate removal efficiency as described in claim 1, characterized in that, in step 1), a position mark is provided on the wafer sample, and the three-dimensional topography image before cleaning obtained in step 2) and the three-dimensional topography image after cleaning obtained in step 4) are taken of the same area marked by the position mark.

9. A rapid evaluation method for the particulate removal efficiency as described in claim 8, characterized in that, the size of the wafer sample is 2 cm × 2 cm, and the three-dimensional topography image before / after cleaning is an image of a marked area of 3 mm × 3 mm size obtained by splicing multiple pictures continuously taken corresponding to the situation before / after cleaning.

10. A rapid evaluation method for the particulate removal efficiency as described in claim 1, characterized in that, it is used to evaluate or screen cleaning liquids, and the particulate removal efficiency of the cleaning liquid on the surface of the wafer after chemical mechanical polishing in semiconductor manufacturing is evaluated.

Images