TW402769B - Apparatus and method for contact failure inspection in semiconductor devices - Google Patents

Abstract

There is provided a contact failure inspection system and method for semiconductor devices and a method of manufacturing semiconductor devices. Using digitized values for electron signals detected using a scanning electron microscope, contacts can be inspected to identify failures such as non-open contact holes. The contact failure inspection is performed by comparing the electron signal value detected from a unit area including at least one contact hole with values representative of the electron signal corresponding to a normal contact.

Description

40276S a? ------______________ B7 V. Description of Invention (1) Tongguan Application I This case is based on US Provisional Patent Application No. 60 / 090,137, and the application date is June 22, 1998. This invention relates to the field of inspection of semiconductor devices. In particular, the present invention relates to the use of a scanning electron microscope to inspect contact failures such as "air vent test contact holes." _ BACKGROUND OF THE INVENTION The manufacturing method of integrated circuits is to first form discrete semiconductor elements on a silicon wafer. Then a multi-level metal interconnection network is formed on the element, contacting its active element and connecting the active elements together to produce a predetermined circuit. The formation of the interconnect layer is by depositing an insulating layer on the discrete element, making patterns and etching contact openings. Inside the interconnect layer, a conductive material is then deposited inside the opening. A conductive layer is then typically applied over the insulating layer. The conductive layer is then patterned and etched to form interconnections between device contacts to produce a first-order circuit. Depositing an insulating layer, forming a contact hole or a through hole, forming a conductive material layer, and making a pattern are repeated until a multi-level circuit is generated. Depending on the complexity of the overall integrated circuit, multi-levels such as second to fourth-order metals are typically required to form the required interconnections and connect the interconnections to the contacts,

It allows external connection to the finished circuit. Designing submicron-sized high-density integrated circuits requires extremely precise dimensional control and highly sensitive inspection methods to inspect interconnect patterns and / or contact holes to ensure the size and structural integrity of the design pattern. As the circuit becomes denser and smaller, this demand becomes more stringent, such as mass production of semiconductor memory components such as 64M (('NS) Λ4 specification (210X297 male t)% ^ 1-installed-(阅读 Read the notes on the back before filling this page), Thread 4 A7 B7 2

40Z76B invention description (DRAM or 256M DRAM, currently requires circuit size of 0.25 to 30 micron)

I It is more and more important to check the condition of the contact hole, such as the case without opening, because the aspect ratio (A / R) of the contact hole, that is, the depth-to-diameter ratio increases as the semiconductor method and the degree of demand increase. However, generally, the optical microscope with visible light with a wavelength of 488 millimeters is used to check the internal characteristics of the contact hole. There is a technical limitation because it cannot obtain a high resolution to test the internal characteristics of the contact hole. The inside of the contact hole may be about 200 nm or less Of the size. Optical microscopy also cannot provide a spotlight of 1 micron or less. SUMMARY OF THE INVENTION It is an object of the present invention to provide a device for a method for inspecting a contact failure of a semiconductor device, which utilizes a digitized value instead of providing a precise contact failure inspection of the contact_image through the naked eye or a microscope, thus substantially eliminating one or Various problems caused by the limitations and disadvantages of related technologies. Another object of the present invention is to provide a method for inspecting a contact failure of a semiconductor device, and inspecting whether a contact having a contact having a high aspect ratio and a hole depth to diameter ratio exists. Still another object of the present invention is to provide a method for inspecting a contact failure of a semiconductor device, and a contact failure inspection system for inspecting a wafer in a short period of time and going to a girl, so it can be applied to mass production and equipment. Another object of the present invention is to provide a method for manufacturing a semiconductor device using a contact failure inspection method and a contact failure inspection system. Λ Another object of the present invention is to provide a method that can measure the paper size of the contact iS ", related materials (TKS) Λ4 ^ Τ (2! 0X297 ^ f ----- ^ ---- -装 ----_-- Γ— Order, ------ M / ί--(Neighbor read the notes on the back before filling in this page) Printed by the Staff Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs 402768 Λ7 — -_ B? 1 " 1 ----------------- I 1 «i— · 、 —β-· .- — ·-__ V. Description of the invention (3) Malfunction A method and system for detecting contact failure at a location where the yield of a semiconductor device is improved. Another object of the present invention is to provide a testing method and a testing system for testing whether a semiconductor device is detected after development during a lithography process. There are pattern failures and photoresist pattern failures. In order to achieve these and other objectives, the present invention is directed to a method and apparatus for inspecting at least semiconductor wafers. In the present invention, scanning electrons that read a portion of a semiconductor wafer Microscope (SEM) image data. Among the SEM image data, the image data identifying the features on the wafer. The parameters about the features are calculated and can be related to a certain range. Comparison of the parameter values. Features can be classified based on the comparison between the parameter and the acceptable value range. In a specific example, the calculated parameter is the size or size of the feature. For example, if the feature is a contact hole of an integrated circuit, the parameter It can be the diameter of the hole measured in the pixels of the image data. For example, the specific contact hole can be determined to be 20 pixels wide. Another-specific example, the parameter can be the average pixel intensity of the pixels in the feature. Again, if the feature is a contact hole, the parameter can be The average pixel intensity of the pixel associated with the contact hole. When the measured parameter is within the acceptable parameter value range, the feature is classified as acceptable. If the parameter is outside the acceptable parameter value range, the feature is classified It is a fault. For example, if the feature is a contact hole, the hole is classified as a fault if it is not open. In the specific example of the present invention, two parameters are calculated for the feature. The two parameters may be the feature size measured at the feature-associated pixel, for example. Such as contact holes. The second parameter may be the average pixel intensity of the pixels associated with the feature. The two parameters are compared with a predetermined acceptable range of values for these parameters. In a specific example, this paper scales quietly --------- Approval ----- 1T -------...--I-(Please read the note on the back first ^^ Fill in this page) 6 V. 402769 Λ7 Η 7 Description of the invention (4 Both parameters are not acceptable parameters. For example, contact holes are classified as accessible. For example, contact holes in this case are classified as The openings are moderate in size and shape. The relationship between the parameters and their individual ranges can be used to classify whether the feature belongs to one of several types or categories. For example, each of these parameters can be used to determine whether the parameter is below or within its range or above its acceptable level Value range while classifying features. In a specific example, the SEM image data is generated from both the secondary electrons of the scanning electron microscope and higher energy backscattered electrons. Data values are digitized and can be digitized in grayscale or color-coded pixel values. In a specific example of the present invention, the grid or grid structure is used to characterize the features to be tested, such as determining the location and / or size of the features to be tested. The grid or grid structure typically includes a pair of interactive orthogonal axes that are superimposed on the part of the wafer image under analysis. The H grid axis can form any other suitable geometric relationship such as a triangle, trapezoid, etc. In a specific example, the grid positioning program may determine the position, shape, and / or periodic pattern of the feature by analyzing the pixel values along a line. The line is one of the parallel orthogonal axes, which is successfully positioned along another-orthogonal The position of the axis's pixels. For example, the grid method includes positioning at most horizontal pixel positions—vertical lines, and adding vertical pixel intensity values to each horizontal position. The summed intensity can be compared at each horizontal position. 俾 Identification can be used to indicate whether there are features such as increased intensity of contact holes. This process is repeated for a plurality of pixel positions along a single dimension. It can then be repeated in the orthogonal dimension, so the pattern, shape and size of all features can be determined. This method can also be used to determine the size of the paper containing the feature to be analyzed. The Chinese paper standard (CNS) Α4 specification (metrics) Ding (Please read the note on the back ^^ Fill this page first)-Install ----- Order III line. Printed by the staff of the Central Bureau of Standards of the Ministry of Economic Affairs 7 ^ 02768 Five Λ7 Description of the invention (5 grids or grids) The optimal size of the unit. For example, the grid program can be used to select the optimal grid unit size for pixels containing j-hundreds of contact holes to be analyzed at one time. This method can make the method for checking features of the present invention more effective because Optimizing the area of each area to be inspected by $ | can eliminate unnecessary process procedures. Preparation · In a specific example, the 'SEM image pixel data is used to calculate the strength of the various features that are inspected | I. Contact holes Profile (pr〇file). In a specific example | _ j, the first way to generate an intensity profile is by pairing one of the orthogonal axes | 彳 丨 丨 characterized by each complex number set along an orthogonal rectangular axis. Pixel location image The prime intensity values are read ^ and added up. For example, at each horizontal pixel position, the pixel weights in the vertical direction are totaled | the intensity value is averaged and compared with the horizontal axis pixel position. Then the pixel [intensity profile is used for classification according to the invention In a specific example, in order to normalize the intensity profile for all features-\, the background intensity value is subtracted from all the intensity values of each grid unit. So $ * 7 reduces the background value of each intensity profile to The effect of zero. Secondly, the threshold can be recorded by regularization, so the pixel intensity higher than the intermediate value can be attributed to the feature of _ 彡 [test. Secondly, the first and second parameters of the aforementioned identification can be recorded by side recording. Calculated by j. For example, the feature size can be calculated along the first-dimension with Yu Zheng ㈣ f printed by the Shellfish Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs.

Calculate the number of exercises when you add the total strength of the money. The pixel intensity of the verification material I threshold is related to this feature, and has

The number of prime positions can be measured by the feature. The H parameter can be determined by averaging the intensity values exceeding _. These two parameters 彳 | compare their individual predetermined acceptable value ranges, and categorize a specific feature as one of the predefined, predetermined feature type classifications. The financial inspection method of the present invention provides countless excellent materials. This paper ruler is printed on the homepage of the paper. It is printed by the Zhengong Consumer Cooperative of the Central Bureau of the Ministry of Economic Affairs. ^ 02768 λ7--; ____ Η? 5. Description of the invention (6 )… The advantages of the method. For example, some previous methods used optical methods such as optical microscopy or visual inspection to detect contact failures. These systems are unable to resolve small irregularities in the characteristics that cause the circuit to fail. The scanning electron microscope used in the present invention provides far better resolution, so it can detect smaller irregularities. Therefore, the present invention is suitable for presenting circuit characteristics whose size belongs to the sub-micron range. In addition, because of the grid method of the present invention, the pixel data processing of the present invention is extremely effective. Processing and fault identification can be performed very efficiently and quickly, so the inspection method and system of the present invention are highly suitable for wafer and circuit mass production equipment. In another aspect, the present invention is directed to a method for inspecting a contact failure of a semiconductor device, which includes the following steps: setting a processing cassette with a wafer mounted thereon, the wafer having a plurality of contact holes formed on its surface; and picking up by the cassette Take out a specific wafer and load it on the table of the SEM reference chamber; perform electron beam scanning on the wafer being loaded; move the stage on which the wafer is mounted to a specific position related to the incident direction of the electron beam of the SEM; open The shutter is used to scan the electron beam to a specific position on the wafer; it automatically detects the reference position by identifying a reference image formed in advance on the wafer and checks the inspection position: the electron beam of the scan line is at the inspection position; Obtain a clearer image; close the shutter to isolate the auto-focused wafer from the electron beam. By comparing the electron signal detected from the surface of the unit containing at least one contact hole after the electron beam is traced, the normal contacts are defined. The electronic signal value to check the contact failure; move the station to another position and repeat the same steps to further inspect the other positions of the wafer The contact failures of all the wafers inside the cassette are further checked by unloading the wafer product and loading other wafers into the reference chamber and repeating the same steps. This paper size applies the Chinese National Standard (CNS) A4 specification (210 × 297 cm);. Clothing-(Please read the note on the back first: fill out this page) 'βτ LineΗ 印 Printed by the Consumers' Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs Λ7 Η 7 V. Description of the Invention (7) According to another aspect of the present invention, a method for manufacturing a semiconductor element includes the following steps: forming a contact hole for a specific insulating material layer formed on a semiconductor substrate; The electronic signal value detected on the surface of a contact hole and the corresponding normal connector signal value are used to test the contact of each contact hole; and after the inspection, a conductive material layer is filled into the contact hole to perform the subsequent process of the semiconductor device manufacturing method. The contact failure inspection step can be performed on a specific sampling position on a semiconductor substrate, for example, the contact failure inspection step is applied to a mass production line. After the development process for forming the photoresist pattern is completed, the failure inspection step may be further performed on the bottom of the photoresist pattern forming the contact hole. According to yet another aspect, the present invention includes a method of manufacturing a semiconductor device, including the steps of: forming a photoresist contact hole pattern; forming a contact hole for an insulating material layer formed on a semiconductor substrate; The electronic signal value detected per unit area of a contact hole pattern and the electronic signal value corresponding to the normal contact pattern are used to check the contact of each contact hole. Brief description of the Zhou style. The foregoing and other objects, features, and advantages of the present invention will be explained more specifically by the preferred specific examples of the present invention. It will be clearly clarified as illustrated in the drawings. Similar reference signs in the drawings indicate several figures. Of the same parts. The drawings are not necessarily drawn to scale, but emphasis is placed on illustrating the principles of the present invention. Figure 1 is a schematic block diagram illustrating the operation of a scanning electron microscope (SEM). This paper is scaled by it's wealth); \ Private clothing-(Please read the note on the back first ^ Fill in this page)

， 1T 10 Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs -------- 7. The electron spectrum includes the energy spectrum of secondary electrons and backscattered electrons. Figure 3 is a schematic block diagram illustrating the configuration of the online sem. Fig. 4 is a schematic block diagram showing a specific example of a contact failure inspection system for a semiconductor device according to the present invention. Fig. 5 is a schematic block diagram showing another specific example of a contact failure inspection system for a semiconductor device according to the present invention. Fig. 6 (i) is a schematic block (ii) showing another specific example of a contact failure inspection system for a semiconductor device according to the present invention. Fig. 7 is a schematic block diagram showing another specific example of a contact failure inspection system for a semiconductor device according to the present invention. Fig. 8 is a schematic functional block diagram showing a specific example of a contact failure inspection method and system according to the present invention. Fig. 9 is a schematic functional block diagram showing another specific example of the contact failure inspection method and system according to the present invention. Fig. 10 is a schematic functional block diagram showing another specific example of a method and a system for checking a contact failure according to the present invention. Fig. 11 contains a schematic flowchart illustrating a logic flow chart of a specific example of a method for detecting a contact failure of a semiconductor component according to the present invention. Fig. 12 contains a schematic flowchart illustrating a logic flow diagram of another specific example of a method for inspecting a contact failure of a semiconductor element according to the present invention. Figure 13 contains a schematic flow chart, which illustrates that the paper size according to the present invention is applicable to the Chinese National Standard (CNS) A4 specification (210X297): i-- (Please read the note on the back first also # fill this tile )

.1T line 11 4Q276B Α7 ---------- B? V. Description of the invention (9) ~ '= Another way to check the contact failure of physical parts-specific logic Figure, which illustrates another example of the logic of the full range of the examples of the contact transfer test method of the two-body components of the hair core. Fig. 15 is a schematic diagram illustrating a wafer sampling position pattern on a wafer for contact failure inspection according to a specific example of the present invention. Fig. 16 is a schematic diagram illustrating details of the sampling area at one of the wafer sampling positions in Fig. 15. FIG. 17 is a schematic cross-sectional view of a semiconductor element having a contact hole formed therein, which can be used for a contact according to a specific example of the present invention. FIG. 18 is an example illustrating confirmation of a contact position by a specific example according to the present invention. Methods After setting the grid, the SEM image data of the contact holes. Fig. 19 illustrates a grid setting for performing a contact confirmation method according to a specific example of the present invention. Fig. 20 illustrates a grid setting for performing a contact confirmation method according to an alternative example of the present invention. Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs— · —— ^ ----- Approval— (Please read the precautions on the back first and fill out this page) Figure 21 shows a contact unit and a specific example of the present invention. A schematic representation of the relationship between horizontal and vertical pixel units. Fig. 22 contains a profile of the intensity of the contact unit before the background value is removed according to a specific example of the present invention. Figure 23 contains the intensity of the contact unit in Figure 22 after the background value is removed. This paper size is applicable to China National Standard for Ladder (CNS) A4 (210X297). -12-Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs; τ- ------------ ---------------- ^ 、 Explanation of the invention (10) Figure 24 contains the side record of the intensity of the SEM image of the contact unit after the background value is removed. Fig. 25 contains a code chart for identifying the result of a contact failure inspection according to a specific example of the present invention. Fig. 26 contains a graph showing the results of a partial contact failure inspection according to a specific example of the present invention. Fig. 27 contains a flowchart illustrating a logic flow of a semiconductor element processing sequence according to a specific example of the present invention. Fig. 28 is a schematic block diagram illustrating a logic flow of a specific example of a contact inspection method according to the present invention. Fig. 29 is a schematic flowchart illustrating the logic flow for reading the scanning electron microscope image data according to the method of Fig. 28. Figures 30A-30D contain schematic flowcharts illustrating the logic flow for confirming the position of the contact hole according to the method of Figure 28. Figure 3 1A-3 1D contains a schematic flowchart illustrating the logic flow for calculating the contact hole profile according to the method in Figure 28. Figures 32A-32B contain schematic flow diagrams illustrating the logic flow for checking contact holes according to the method of circle 28. Detailed description of a preferred specific example FIG. 1 contains a schematic block diagram of a scanning electron microscope system 100, which can be used to inspect contact holes of a semiconductor device according to the present invention. Referring to FIG. 1, the electron grabber 102 projects an electron beam through a condenser 104. The electron beam passes through the deflection yoke 122, the aperture 106 of the objective lens 108 and the shutter 124. The focused electron beam is projected and scanned on the reference or inspection surface. This paper size is applicable to China National Standard (CNS) A4 specification (210 father 297 public #) -----. ^ I; JTS (Please read the notes on the back / 41 to complete this page first) 13 Printed by the Consumer Cooperatives of the Central Government Bureau of the Ministry of Economic Affairs ___402768___ V. Description of the invention (11) It may be the side of the semiconductor wafer under inspection. By reference The secondary electrons and backscattered electrons emitted by the product are detected by the signal detector 112, which generates a signal indicating the received electrons. The detected electronic signals are amplified by the signal amplifier 114. The amplified signals can be amplified by the cathode ray tube ( CRT) 118 is scanned on the fluorescent surface, so a visible image of the reference surface is formed. The scanning of CRT 118 is controlled by the deflection coil 116, which is related to the scanning of the reference surface controlled by the deflection coil 122. In SEM, the reference The scanning surface of the product is divided into small pixels, and the electronic signals detected by each pixel element are time-series shifted, so a SEM image is formed. In order to achieve this project, the power of the signal amplifier 114 is used. The signal is transferred to the scanning circuit 120, so the deflection angle of the electron beam is controlled by the second deflection coil 122. In addition, the amplified electronic signal data of each pixel can be transmitted to the processing unit 115 ', which can perform a variety of signal conditioning and processing functions The processing unit can convert the electronic signal of each pixel into a gray scale or color code value to form an image. The gray scale value can be assumed to be one of the 256 possible gray scales that are digitally encoded from a binary value of 0 to 255 The memory is used to store the gray level values of each pixel. The computer also constitutes a processing unit and can process image values as needed. In one configuration, the computer can analyze the gray level data through program planning to perform the contact inspection of the present invention. Details will be described later. Section 3 contains schematic block diagrams showing the characteristics of the online SEM system, in which the contact inspection can be performed online. In the previous method, the SEM image data was taken offline, that is, collected and analyzed from the manufacturing process. The method has greatly improved efficiency, so SEM image data can be collected and analyzed during the manufacturing process. This can eliminate the additional inspection used by the previous method The size of this paper on line is applicable to China National Standard (CNS) A4 specification (210X297 public trip) t; I r order --------- y line (please read the precautions on the back ^ 1 to fill in this page) '14 V. Description of the Invention (ι2) —' The sem system includes an electro-optical section, a reference section, a vacuum section and an electric section.

The I sub-optical section includes an electron beam generator 14, an electron beam deflector 15, and a signal detector 16. The reference section includes the reference 00 transfer section 12 and the reference product alignment section 13 for transferring the reference product, that is, the wafer from the cassette to the reference cavity. The vacuum section includes a vacuum forming portion 11 for maintaining a vacuum inside the reference chamber. The power section includes a main computer 10 with a main controller 21 for controlling the electro-optical section, reference chamber, vacuum section, and other system components. The power section also includes a main storage unit or memory 22, which can store the detected signal data values from the signal detector 16, and a main display 19 for displaying images generated by the detected electronic signals. The autofocus controller 18 can perform autofocus to provide a clear image. In a typical scanning electron microscope, a reference surface illuminated by an electron beam generates primary electrons and is emitted from the reference surface. Electrons from the electron beam are also scattered by the surface. Figure 2 shows the energy spectrum of electrons emitted and scattered from the reference surface when the reference surface is illuminated. As shown in Fig. 2, the maximum secondary electrons (SE) were found in the energy band of electrons below 50 eV, and most of the backscattered electrons (BSE) appeared in far more pseudo-energy bands. The commonly used online SEM is used for secondary electrons (SE) generated in the low energy band of about 20 eV2, so the surface and edge images are clear. However, for testing features that have a high aspect ratio, that is, hole depth to diameter ratio, such as a contact hole, the secondary electrons generated inside the contact hole may dissipate when passing through the contact hole, so the image of the contact hole is not clear. Therefore, the characteristics are typically optically inspected with the naked eye, so a clear image is necessary for contact failure detection. Figures 4-7 contain schematic block diagrams of various specific examples of the contact failure inspection system for semiconductor components according to the present invention. . The system in Figure 4 contains several paper standards applicable to the China National Standard (CNS) A4 ^ (2; ^^ 7 公 妙 ~} -------— printed by the Bayer Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs

402769 Λ? ---_..___ 5. Description of the invention (13) " Same component as SEM on line shown in Fig.3. But it also includes the gallop 6Q of the contacts of the main computer paste. As mentioned above, the on-line structure includes the electronic optical section, including the electron beam generator 14, the electron beam deflection center, and the signal detector section 16. The signal detector is preferably a detector capable of detecting secondary electrons (SE) and backscattered electrons (hall) emitted from the reference surface after the reference surface is irradiated with an electron beam. The system also includes a reference chamber including a reference alignment section 13 for rotating or tilting the table on which the reference wafer to be inspected is set to move simultaneously along the X, Y, and Z axes 4. The vacuum forming unit u can maintain the vacuum state of the reference chamber to a predetermined degree. The reference product transfer unit 12 can transfer the reference product to the reference chamber. The power section includes a host computer 20 with a main controller component for controlling the electro-optical section, reference chamber, and other subsystems of the vacuum forming component. The main storage list 7G or the memory 22 stores a signal detected by the signal detector 16. The main display 19 displays an image deduced from the detected electronic signals. The AF controller 18 can automatically perform the focusing operation to obtain a clear image display. This specific example also includes a contact failure inspection module 60 for analyzing the information contained in the electronic signals sent by the signal detector 16 and stored in the host computer 20, and inspecting the contacts according to the present invention. Fig. 5 illustrates another specific example of a contact failure inspection system for a semiconductor device according to the present invention. The system of Figure 5 also contains several of the same components as the online SEM of Figure 3. Part of the difference is that it also includes a contact failure inspection module 60 and a contact critical dimension (CD) measurement module 70, which can be located inside the main computer 20. The critical dimension is the dimension of a particular feature that is tested. For example, taking a circular hole as an example, the critical dimension is the diameter of the hole. In a specific example, the contact critical dimension (CD) measurement module 70 is a contact image generated through a comparative SEM. The paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm). Please read the precautions below. 3 (Fill out this page-order five. Description of the invention (14 Λ7 B7 Measurement of contact diameter and pre-stored standard value printed by the Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs.) Figure 6 illustrates an example of a semiconductor device according to the present invention. Another specific example of the contact fault online inspection system. The system in Fig. 6 also includes some of the same components as the online SEM in Fig. 3. Part of the difference is that it also includes a sub-computer 80 and a contact interface with the main computer 10. The fault checking module 60 is inside the auxiliary computer 80. The auxiliary computer 80 can use a standard commercially available personal computer, including the auxiliary display and the auxiliary storage unit or the memory. The point failure inspection module 60 analyzes and obtains the detection from the contact electronic signal data (which can be stored in the main storage unit 22). FIG. 7 illustrates an example of a contact failure of a semiconductor element according to the present invention. Another specific example of the inspection system. The system in Fig. 7 also contains some of the same components as the online SEM in Fig. 6, with the interface of the auxiliary computer 80 and the main computer 40. It also has a contact fault inspection module 60 on the auxiliary The computer 80 and the contact CD measurement module 70 are on the host computer 40. Figures 8-10 contain schematic functional block diagrams showing various specific examples of the contact failure inspection module 60 according to the present invention. Referring to Figure 8, the contacts The fault inspection module 60 includes a SEM signal reading module 60a, which receives the SEM signal indicating the electrons received by the wafer when the wafer is irradiated with the electron beam. The contact position confirmation module 60d analyzes the SEM signal to determine the contact hole and / Or the location of other features to be tested. The contact profile calculation and background removal module 60e can use the SEM signal data to generate an intensity profile of the contact hole. The intensity profile is typically removed by removing the data due to the background intensity effect Regularization, so the strength profile can be independently checked for background effects. The contact fault inspection module 60f analyzes the strength profile of the contact hole to identify contact failures. In a specific example (please read the details later) Of 41 Pei-book paper size applies Chinese National Standard (CNS) A4 specification (210X297 public #) 17 Λ7 40276 & 5. Description of Invention (15)), the average strength value of the contact is compared with a predetermined interval to identify the fault. The result display module 60g can display the failure analysis result. In a specific example, the SEM signal reading module 60a reads the digitalized electronic signal information of the contact, which is stored in the main storage unit or memory of the host computer 20 22. On-line SEM digitizes the electronic signal intensity detected by the electron beam scan and stores the intensity as a grayscale or color code value. In a system, the grayscale value assigned to each pixel ranges from 256 possible values in the range of 0 to 255 One. The highest intensity is defined as 255 and the lowest intensity is defined as 0.1. The digitized intensity value is color coded by each pixel element, that is, gray scale. The contact image reads the grayscale value of each pixel in time series and displays the pixel image on the cathode ray tube, monitor and / or printer. The gray level can be changed to color for color display. Refer to Figure 9, the specific example of the contact failure inspection module explained here includes the modification of the specific example in Figure 8. In FIG. 9, in contrast to the merge module 60 e in FIG. 8, a contact profile recording module 60 e (1) and a background value removal module 60 e (2) are provided as separate modules. Fig. 10 is a schematic functional block diagram showing another specific example of the contact failure inspection module 60 according to the present invention. Referring to FIG. 10, the contact failure inspection module 60 includes a graphic file transmission network module 60b, a graphic broadcast_8 £ 1 ^ signal conversion module 60c, a contact position confirmation module 60d, and a combination Contact profile calculation and background value removal module 60e 'Contact failure inspection module 60f and result display module 60g. Referring to FIG. 6, the graphic file transmission network module 60b is a signal transmission device between the main computer 10 '40 and the sub-computer 80. The module will be stored in the main computer 10 '40 of the main storage unit 22 contacts digitalized electronic signal information. The paper size applies the Chinese National Standard (CNS) A4 specification (210X297 mm). -18-40 ^ 769 ***. 'Invention note (16) is converted into a graphic file and then transferred to the auxiliary computer 80. The graphic broadcast-SEM signal conversion module 60c reads the color of the graphic file transmitted to the auxiliary computer 80, that is, the gray scale, and converts it into a digitalized SEM signal. The contact location green recognition module 60d, the contact profile calculation and background value removal module 60e, the contact failure inspection module 60f and the result display module 60g are as described above with reference to Figures 8 and 9. 11 to 14 are flowcharts showing a plurality of specific examples of a contact failure online inspection process of a semiconductor element by the contact failure inspection system according to the present invention. Referring to FIG. 11, the contact failure inspection section 60 of FIG. 4 performs contact failure inspection on a wafer having a plurality of contact holes on the wafer by using an SEM mounted on the host computer 40. First, a cassette is set at a predetermined position of the online SEM, and a wafer (810) with a plurality of contact holes is installed on it. Secondly, after a wafer to be inspected is taken out from the cassette, it is loaded on a table inside the SEM reference chamber. (S12). Align the wafer plane next. The loaded wafer is then scanned by the positron beam (S14) and the stage on which the wafer is mounted is moved to be set at a position in the electron beam incident direction of the SEM (S16). The shutter set under the objective lens is opened to irradiate the electron beam at a certain position on the wafer 'and performs automatic addressing (S20). The automatic addressing is performed by applying a standard image of a pre-made pattern to a certain position, so that it can be checked against the standard image to identify a certain position. Next, the position to be inspected is irradiated with an electron beam from the SEM (S22), and the electron beam scanning is repeated by the autofocus control unit to obtain a clear contact image (S24). Then, the shutter is closed and the electron beam scanning of the wafer is completed (S26). Secondly, check the contacts detected by the electron beam scanning according to the present invention 403768 at -------- H7 V. Record the intensity of the electronic signal of the invention description (17) (S28). It is then decided whether to perform a contact failure inspection at another position of the wafer (S30). If yes, the flow returns to step s ^ 6, the stage moves to other positions on the wafer and repeats the foregoing steps. When the contact failure inspection is completed, the wafer is unloaded (S32p determines whether there are other wafers to be inspected (S34). If so, the other wafer in the cassette is loaded in the reference chamber and the previous steps are repeated to thereby check the card. All wafers inside the cassette are inspected for contact failure. If all wafers are inspected, the cassette is removed (S36) and the process is completed. Refer to the specific example in Figure 12 for wafers with multiple contact holes. The contact fault inspection method uses the online SEM in Figure 5 to show that it has a host computer 30, which together forms a contact fault inspection section 60 and a contact CD measurement ap 70. The contact fault inspection is performed as described in Figure 11 However, unlike the specific example in figure η, the shutter is closed (S26), the contact failure inspection is determined (S27), and the contact CD measurement is performed when the contact failure inspection is not performed (S28) (S29). Printed in Figure 13 by the Consumer Bureau of Standard Bureau. Figure 13 shows the online fault detection method for the contact failure of wafers with multiple contact holes on the line SEM in Figure 6. Host computer 10. As shown in Figure 13 After the shutter is closed (S26), the SEM signal stored in the main storage unit of the main computer is transmitted to the sub-computer to determine whether it is necessary to perform contact failure inspection on other positions of the wafer (S31). The sub-computer receives the transmitted signal The contact failure inspection is performed (S37). In the specific example of FIG. 14, the contact failure inspection section 60 is installed on the sub-computer instead of the main computer shown in FIG. 6, the sub-computer and the main computer. Reciprocating exchange instruction. By the same steps as shown in item 11), the shutter is closed. The paper size applies the Chinese National Standard (CNS) A4 (210X297 mm) 20 i) i) A7 H? V. Description of the invention (18 (S26 ), The SEM signal of the main storage part stored in the main computer is transmitted to the sub-computer (S31-1), and the sub-computer is used to perform the contact failure inspection (S312). Decide whether to perform the contact inspection of other locations (S31_3, S31_4 Figure 15 shows the numbered shaded area (No. 2 to No. 37) to be tested on wafer 11 under the contact failure inspection of the present invention. The area marked "AP" shows the alignment point and No. 1 shows Contains the focus position of the chip. In each numbered shaded area in Figure 15, several sampling positions can be defined. For example, for wafer 2 or area in Figure 15, Figure 16 shows five sampling positions', that is, top left (2, 1), top right (2 , 2), bottom right (2,3), bottom left (2,4), and center (2,5). The sampling position or number of samples can be selected in a single way in a single sampled wafer unit. In the exemplary embodiment of the present invention, 175 positions were sampled from 35 wafers or sampling area. Each wafer was tested at 5 sampling positions. In a specific example, the online SEM was used to zoom in 12_5 k, and there were 98 contacts in the 480 X 480 pixel image of each sampling position. For the five sampling locations in each of the 35 zones, check the 17,150 contacts. Section Π is a schematic cross-sectional view of a semiconductor device having a contact hole to be inspected according to the present invention. Figure 17 shows the process of forming a 64M DRAM embedded contact. The field oxide layer 131 defines an active region and is formed on the semiconductor substrate 130. The gate electrode 132 is formed on the active region and is covered with the spacer 133. After the first insulating layer 134 of the high-temperature oxide film is formed on the surface, a first contact hole 137 is formed as a direct contact of the bit line 135. After the bit line 135 is formed, a second insulating layer 136 is formed on the surface as BPSG, and a second contact hole 138 is formed for the word line. As for the inspection example of the present invention, the inspection is made on 64M DRAM. The paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 male dust)-^^ 衣 ^ Order -------- / line-- (Please read the note on the back first # Fill in this page) Yin Fan 21, Λ7 4〇3ϊβ3 A, Shellfish Consumer Cooperative, Central Bureau of Standards, Ministry of Economic Affairs, and the description of the invention (19). As shown in Fig. 17, the inspection is also performed on the direct contact 137 'or after the photoresist-like development process of forming the contact. In the contact failure inspection method of the present invention, first, an optimal image size is selected for each inspection location based on the characteristic size of the inspection, such as the diameter of a circular contact hole. In a specific example, a 'typical SEM image includes 480 x 480 pixels. Images can be taken for each numbered position in Figure 16. According to the contact hole size and spacing, determine the optimal image size of each contact. Figure 8 shows an example of a 12.5 K magnified SEM contact image on the line of a sampling position of a semiconductor device. It consists of 480 X 480 pixels, and the number of contacts illustrated in the image is 98, that is, 14 in the horizontal direction and 7 in the vertical direction. Based on the feature size under test, that is, the contact hole and the distance between holes, determine and check the best resolution. For example, a system can analyze each pixel at about 12 nm in SEM. The diameter of the contact hole is currently known to be about 200 nm. Choose the number of pixels covering a feature to ensure that feature irregularities can be detected in the image. For example, each of the squares in the test has 100 holes in a spaced average sentence. The set of 100 48 × 48 pixel sub-squares can be used to cover all the holes, including the space between the holes. Each sub-square is related to one. hole. Rectangular grids or grids of horizontal and vertical lines can be stacked on a 48 () < 48 pixel array to form 100 48 X 48 pixel sub-squares for testing each hole. It was subsequently decided according to the present invention that a 48 by 48 pixel grid is sufficient to resolve any irregularities in the inspection hole. The size of the hole is compared with the amount of space in each sub-square. Determine whether the number of pixels covering the hole itself is sufficient to analyze the hole. Resolution is determined by dividing the critical dimension, ie the diameter of a hole, by the number of pixels covering the hole. The dimensions of this paper are applicable to the Chinese National Standard (CNS) M regulations; --- Λ '-22-I ~ „: Attaching ^ nails ----- j line./1.--. (Please read the back first Please pay attention to this page, please fill in this page) Printed by the Consumer Cooperatives of the Central Bureau of Standards, Ministry of Economic Affairs Λ7 402Ϊ68 V. Description of the invention (20) The resolution is determined by comparing the threshold with a threshold value such as 12 nm / pixel threshold. After determining the pixel resolution The grid structure can be used to locate and determine the size of the contact hole. In a specific example, the vertical and horizontal lines of the grid or grid structure can be used to locate the hole. Refer to Figure 18 for an example of the contact position identification method of the module 60d. The grid or grid is placed on the contact image of the alignment matrix, and the horizontal and vertical axis pitches are adjusted in a certain study area, so each contact is placed in each grid. At this time, the pitch can be increased or decreased. Control the number of pixels that can generate a contact image. The search area of the grid lines is preferably set to include areas that repeat the same contact hole pattern. Referring to Figure 18, the grid position search process is used to determine each grid unit or The sub-square has at least 32 images in the horizontal direction And the vertical direction is composed of at least 62 pixels, and the search area is determined by moving the imaginary horizontal axis grid line 50 and the imaginary vertical axis grid line 152 on the horizontal axis including at least 32 pixels and the vertical axis including at least 62 pixels. Detect the position of the lowest digitized electronic k value, so each contact in the contact image will not be disturbed by grid lines. In a specific example, the grid search is performed by positioning the vertical or horizontal line at the first position. Along the line The density values are summed to determine the total intensity of the line. Then the line is stepped to the next position. For example, the vertical line can be stepped to the next position along the horizontal axis, and the intensity values along the vertical line are again added here. Position, the total intensity value is compared with a predetermined threshold and the previous total value. The increased intensity is used to indicate that the edge of the hole is reached, assuming that the hole strength is higher than the background. In other specific examples, the hole strength is lower than the background. This process can continue across the entire grid structure Positioning V. 402768 Λ7 B? Print the invention description (21) 'and / or define the size and shape of each hole by the Consumer Consumption Cooperative of the Central and Southern Bureau of the Ministry of Economic Affairs. Calculate all totals in one direction. After the value is repeated, the process is repeated for other directions, the result completely characterizes the size, shape and position of the hole. This information can be used for multiple purposes in subsequent processes. With known hole position and shape, unnecessary pixels that are not related to the hole can be deleted In addition, if a fault is identified in the subsequent process, it is easy to determine the correct position of the fault hole. 0 After performing contact position recognition, the initial grid unit origin shown in Figure 18 is, for example, the horizontal axis pixel number XQ = 13 And vertical axis pixel number γ〇 == 2 3. Units with the same size are acceptable for comparison, which is why the contact position identification is performed as above. The grid unit can adopt multiple settings, that is, as shown in Figure 9 As shown in FIG. 20, 'where at least two contact holes i 53 can be positioned in a -grid unit.' In addition, if the position recognition process of the aforementioned grid method is repeated for each unit area image pattern, such as a square or oval pad image, it can be used for a variety of image shapes, as opposed to the aforementioned circular contact image. Figure 18 shows an SEM image of the contact unit (480 X 480 pixels) of the grid setting value used for the contact position identification process according to the present invention. As mentioned above, in this example, the grid unit of this test is set to 32 X 62 pixels, and the origin pixel number (XO, YO) is (13, 23). The setting value of the grid unit is determined by moving the horizontal grid line 150 and the vertical grid line 152 within the search range defined by the aforementioned grid lines, and it can be set to correspond to each grid unit with a pitch of about 60 pixels. And about 30 pixels. In another specific example, each line is analyzed to identify the lowest intensity value corresponding to each grid line and determine the location of the hole. Please k, read the note on the back 1¾, and then fill in this page. The size of this paper applies the Chinese National Standard (CNS) A4 specification (210 × 297). 24 5. Description of the invention (22) 402762 Λ7 Η 7 19th and 20th The figure shows the aforementioned different types of grid settings, and Figure 21 does not specify the pixel unit for the contact unit. It explains the generation of contact strength profile. Figure 22 shows the first intensity profile recorded in the -grid unit, showing the intensity values before the background value is removed and the regular profile is relative to the vertical axis. Figure 23 shows the intensity profile of Figure 22 with the contact threshold after removing the background intensity value. In this test, the electronic signal cabinet value is set to 5, and the pixel threshold is 20, as shown in Figure 23. The inspection contact hole spans 20-40 pixels along the vertical axis. -Order Figure 24 shows the intensity profile of the SEM image of the contacts in Figure 18 after normalization by removing the background value. Fig. 25 is a table and contains the coding results of the node 8 of the present invention to perform the contact fault inspection of the present invention. The round contact match in Figure 24 defines code 4 in Figure 25 as the contact fault location. Fig. 26 is a graph showing a portion of the contact failure inspection result of each sampling position relative to each wafer sampling position shown in Fig. 15 in this test, indicating all the contact numbers corresponding to each contact classification standard according to the present invention. In other words, for each of the five positions of the wafer or test area, list the number of holes in each type of classification seen at that position. For example, at position (1, 3), 87 contact holes are classified as d-type, 3 holes are classified as E-type, 5 holes are classified as G-type, and 3 holes are classified as ^ -type. Attention should be paid to the inspection and classification of 98 holes in each position, and a total of 17, 50 holes in 35 inspection areas. In a specific example, since the method of the present invention can save processing time, the test can be completed within one hour. Therefore, this method is applicable to mass production. The contact side record calculation module 60 0 (1) is used to generate the first intensity side record of the aforementioned detection electronic signal value for each grid unit of the specific network card. The background value removal module 6 0 e (2) is used to deduct the smart scale scale of each grid unit from the first intensity profile. Gj Ma County (CNS) Tiege ⑽χ297 ^ 25 Consumption cooperative printing 402769 Λ7 _______B? V. Description of the invention (23) 'Background value and the second intensity profile are generated from the first intensity profile. The first intensity profile and the second intensity profile are calculated using electronic signal values, which correspond to the digitization of each pixel contained in each grid unit. However, the electronic signal value obtained from each grid unit includes the electronic signal value obtained from the corresponding contact and the electronic signal value generated by the outer area around the contact. In the present invention, in order to achieve an accurate electronic signal value, it includes the intensity of the electronic signal only from the inside of a grid unit contact, and the background electronic signal value from the contact area, that is, the area outside the contact area. The second intensity profile is recorded. This is called "decoloring effect" removal. In the specific example of the present invention, the calculation of the contact strength profile and the removal of the background value are performed by the borrowing module 60e according to the following formula (1). 4 {BIBc) X ^ ⑴; where X is the sum of electronic signal values above a predetermined threshold within a grid unit f B is the sum of electronic signal values below a predetermined threshold within a grid unit;

Bc & —the number of electronic signals S in the unit whose internal value is lower than a predetermined threshold;

Xc is the number of electronic signals whose internal value of a grid is higher than a predetermined threshold; and Y is the value of electronic signals having a background compensation value in a grid unit; This paper size applies to China National Standard (CNS) A4 specifications ( 2ΐ〇χ297 公 梦).. Binding (please read the notes on the back and fill out this page first) 26 402769 Λ7 V. Description of the invention (24) 'In the formula (1), the aforementioned predetermined value can remove the background value and reach The method of quasi-breaking the measurement result is decided. For example, in a specific example, the value is i 00, but is not limited to this value. The γ value of the result of the formula (1) is the electronic signal and ° after the compensation of each grid unit. In a specific example, the lower limit and the upper limit of the γ value of the formula (1) are set. If the γ value calculated for a particular contact is below the lower limit, the contact is classified as a fault. In a specific example, a reading of this type below a predetermined lower limit indicates a failure of the open contact hole. Equation U) is typically used to inspect irregularly shaped contact holes. For example, formula (1) can be used to test a photoresist layer that forms a contact hole before the hole is formed. In another specific example, the calculation of the contact profile and the removal of the background value are performed according to the following equations (2)-(4).

V = (pkN) '-pmN · ^ Meal ----- ^-Order, ί-(Please read the notes on the back first # Fill this page) (2) (w) (3) i% Ministry of Economy Printed by the Consumer Standards Cooperative of the Central Bureau of Standards

PmN = baseline value [(PkN) '] (4); where n is the number of pixels on the horizontal axis; k is the number of pixels on the vertical axis;

Pnk is the digitized signal value of the pixels at the horizontal axis position η and the vertical axis position k; N is the number of grids to be analyzed; This paper size is applicable to China National Interpolation (CNS) A4 specification (210 × 297 mm) ^ 27 ^ _4〇276S Λ1 r --------- Men, invention description (25) ------ is the initial number of pixels inside a grid unit on the horizontal axis, and W is a grid unit inside the horizontal The number of final pixels of the axis. Expressions (2)-(4) are described with reference to FIGS. 21 and 22. Figure 2ι is a graphic representation of the graph, which shows the calculation of the strength of the contact according to the current issue—the pixel unit of the contact in a grid unit. Figure 22 is a schematic diagram illustrating the calculation of the -intensity profile in the -grid unit before the background value of the contact image in Figure 18 is subtracted. In a specific example, the intensity profile is generated by stepping to a discrete position along one of the axes and summing the intensity values along the orthogonal direction of each position and plotting the pixel intensity and the number of pixels along the pixel axis. For example, the side recording of a picture can be formed by stepping along the vertical axis through the pixel position and adding up the pixel intensity in the horizontal direction. In the profile in Figure 22, the result is an intensity profile with spikes entering the center, which indicates the presence of a contact hole in a particular grid unit. The hole extends from about pixel position 16 to pixel position 44 and therefore spans about 28 pixels in the vertical direction. The side recording includes a slight decrease in intensity at the peak center at the center of the hole, indicating a decrease in intensity detected at the bottom of the hole. This intensity profiled shape indicates a normal contact hole. Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs; ^ --- (Please read the notes on the back and fill in this tile) The line is obtained from the formula (3) where (PkN) 'is the pixel number on the vertical axis 1 ^ The average electronic signal value per pixel. Because the vertical axis pixel number 1 ^ line (Figure M; k = 20) corresponds to the sum (degree) of the digitized electronic signal value corresponding to each pixel divided by the number of horizontal pixels at the vertical axis position k (calculated as hfN-hiN ) Calculated. Figure 22 shows the side record generated by equation (3). PmN is (pkN), and the minimum value is the background value or baseline value. Therefore PkN shows the average electronic signal value per pixel minus the background value. Figure 23 is the second strength after the background value pmN has been deducted according to formula (2). The paper size applies the Chinese National Standard (CNS) A4 specification (210X297 cm) 28 40276S V. Description of the invention (26 Λ7 Η 7 side record) Fig. 24 shows the regularization of the second intensity profile of the butt joints in the contact image of Figs. I 8. In a specific example, the contact fault inspection method of the present invention further analyzes the results of equations (2)-(4) and The contact is classified according to whether it is faulty, and if so, what type of fault it belongs to. The second intensity profile of each hole (Figure 24) is analyzed to identify and classify the fault. The specific example shown in Figure 23 For example, the threshold value of 5 is applied to the second intensity profile (average electronic signal value per pixel after subtracting the background value). The critical dimension CDN of the contact is defined as the peak length (or width) of the profile recorded at the threshold. As shown in the figure, the critical dimension CDN where the threshold is set to 5 ′ contacts is CDN = 40-20 pixels. The critical dimension CDn may be the contact hole diameter, for example, which can be obtained by the following formula (5).

CD Σ

N (5), printed here by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs

ViN is the initial number of pixels on the vertical axis inside the grid unit. ViN is the number of pixels on the vertical axis inside the grid unit. PkN = (PkN), -PmN: and WkN indicate whether the pixel intensity value is higher than the threshold, especially W, 1, Right corpse 2 threshold 値 .0, if PkN < threshold paper size is applicable to Chinese National Standard (CNS) A4 specifications (210X297 consolation --- ---- I | Siyi II 1 order, II evening line., ^- (Please read the notes on the back before filling out this page) 29

40276S Λ7 Η 7 V. Description of the invention (27)

Intensity BSEN Second ’Calculates the average pixel for pixels above the cabinet value according to the following formula

PkNwkN

BSE

CD

N (6) Printed by the Shellfish Consumer Cooperative of the Central Bureau of Standards, Ministry of Economic Affairs. Note the alternative inspection methods of expressions (5) and (6). The average pixel intensity bsen referred to in Equation 为 is a similar value to the γ value calculated in Equation (1). The CDN values of equations (5) and (6) can be substituted for the value of 2Xc of equation (l). After calculating the number of pixels cdn and the average pixel intensity value BSEN for the contact under test, it can be used to classify the contact situation. In a specific example, the upper limit value N0C2 and the lower limit value N0C1 can be set for the pixel number cdn. These limits are used to define the number of pixels that a normal contact can accept. These limits can also be set for the average pixel intensity BSEN. The upper limit NOT2 and the lower limit 1 ^ 〇11 can be used to define an acceptable range of average pixel values for normal contacts. The cdn and BSEN values of each contact under analysis are compared with their individual ranges to classify contacts. "In a specific example, each point can be classified into one of 9 possible types based on the comparison of the CDN value and the BSEN value with its individual range. ^ 9 types A list of possible situations and their corresponding type classifications and numerical code examples is shown in Table 1. This paper size is applicable to Zhongguanjiazhuan (CNS) A4 specification (21QX297) (Please read the precautions on the back before filling out this page} --- Installation ----- T7T ------: M-- ------------- 30

40276S Λ7 Η 7 V. Description of the invention (28) Classification (BSEN) ^ (N0T1) (N0T1) ^ (BSEN) ^ (N0T2) (BSEN) ^ (N0T2) (CD ^ CNOCl) A-type (Code 1) B -Type (Code 2) C-Type (Code 3) (NOC ^^ CCC ^) ^ (N0C2) Type (Code 4) B · Type (Shima 5) F-Type (Code 6) (CD ^) ^ XN0C2) G-type (code 7) Η-type (code 8) I-type (code 9) Table 1. Contact classification Three types of contact hole depths are printed in the third column of Table 1 printed by the staff consumer cooperative of the Central Standards Bureau of the Ministry of Economic Affairs condition. They are arranged in decreasing depth. In other words, the first column defines the three conditions for the relatively deep contact holes, namely the A, D and G types. The second column contains the three conditions B, E and Η for normal contact depth. The third column defines the three conditions C, 卩 and I for which the contact depth is insufficient. These contact hole types typically define partially open contact holes or non-uniform hook contact holes. The columns in Table 1 are organized in order of increasing contact hole diameter. The first column includes too small or insufficient diameters of A, B, and C contact holes. The second category includes D, E and F types defining contact holes with normal diameter. The third category includes G, Η, and I-type contact holes with excessively large diameters. As shown in Table 1, the E-type classification is performed on CDN & BSEN, and all belong to their individual predetermined ranges, indicating normal contacts. One or both of the other types are outside the scope and classified as one of the remaining types, which can be used to indicate unequal degree or type of contact failure. The result shows that 60g shows the normal contact and / or contact failure result classified by the contact failure inspection module 60f. The result is displayed as a digitized value with respect to each contact position. Fig. 25 contains a table showing an example of the classification of the contact holes and pixel positions in Fig. 24. Whether each contact is normal or faulty corresponds to the second (please read the note on the back 44 to fill out this page) This paper size is applicable to China National Standard (CNS) A4 specification (210X297). 31 Central Bureau of Standards, Ministry of Economic Affairs Printed by the consumer cooperative 40276S ;; 5. Description of the invention (29) The intensity side record is displayed with a numerical code relative to the position of each contact. Code "5" indicates E-type, indicating normal contact. Code "4" indicates D-type contact failure. In a specific example, the D type indicates an unopened contact. The X value in Fig. 25 represents the initial number of pixels on the horizontal axis of each grid unit. The Y value represents the number of pixels on the initial vertical axis. FIG. 26 contains a table illustrating the results of the present invention's inspection of 5 locations in 7 regions of a semiconductor wafer. The table shows the number of contacts for each classification type at each location. The CDN & BSEN value can be used to classify contacts in different ways. In other words, the type classification to which a particular contact belongs can indicate a particular type of contact failure. For example, when the BSEN of the contact is lower than the minimum value of NOT1, it typically indicates that the contact hole is not opened and the hole classification is one of A, D and G types. When BSEN is greater than the maximum value N0T2, the hole is open, but it is still unacceptable for some reason. For example, the holes may be irregularly shaped, such as widening or narrowing toward the bottom of the hole. In this case the holes are classified as one of the C, F and I types. Similarly, when the CDN is lower than the minimum value N0C1, the fault indicates that the hole is too small or has an irregular shape such as an oval. If the CDN is higher than the maximum value! ^ 〇 (: 2, it indicates an irregularly shaped hole. Fig. 27 contains a process chart, which illustrates the logic of a process sequence showing a method of manufacturing a semiconductor device according to a specific example of the present invention. First, in a specific process step of the semiconductor device manufacturing method, a photoresist pattern (S40) e photoresist pattern system for photoresist on a specific insulating layer such as a nitride or oxide film and corresponding contact holes is formed after lithography. Formed through the exposure process and the development process. Second, use the photoresist pattern as the mask, and the pattern (please read the precautions on the back before filling in this page)-binding and binding 32 402769 Λ7 Η 7 V. Description of the invention (3〇 ) The insulating layer printed by the Consumer Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs is etched to form a contact hole (S42). Next, the inside of the contact hole is cleaned, and the wafer is moved to the online SEM according to the present invention. The contact hole failure inspection system according to the present invention Proceed as described above. Then, the inside of the contact hole is filled with a conductive material, and subsequent processing is performed for manufacturing a semiconductor element (S46). Fig. 28 contains an unintended flowchart, which is an example. The logic flow of a specific example of the contact inspection method of the present invention. The parameters used by the method are read in step 500. The parameters used by the method in a specific example are as follows: N = number of pixels in the Y-axis direction of the SEM image M = Number of pixels in the X-axis direction of the SEM image VP (vertical pitch) = γ-axis contact pitch HP (horizontal pitch) in the γ-axis direction of the grid = X-axis contact pitch MX in the X-axis direction of the grid MX = X Axial grid search pixel range MY = Y-axis grid search pixel range bse = Basic threshold of contact feature profile recording of unit grid NO 1 = Normal contact feature profile recording lower limit value N02-Normal contact feature side Upper limit of recording intensity CD 1-Lower limit of the number of pixels recorded on the side of the normal contact feature CD2 = Upper limit of the number of pixels recorded on the side of the normal contact feature XN =-Total number of wafers inspected by wafer or firing unit Number of SEM images (count) YN = —Total number of SEM images (count) of wafers tested by wafers or firings X = S EM image sequence tested on a wafer by wafers or firing units Please read ii. Note 1 (Fill in the binding line on this page 33 _ 术 76S V. Description of the invention (3i) Printed by the Consumer Cooperatives of the Central Bureau of Standards, Ministry of Economic Affairs Λ7 Η 7 γ = sequence of SEM images on a chip or firing for inspection cdata [j] [i] = degree of SEM image signal in (each) unit pixel Next, at step 502, the X-axis value is initialized to 0; and at step 504, the γ-axis value is initialized to 0. The inner loop formed by the step 5606 52 along the γ-axis inspection system continues to reach the maximum y-axis Then, the χ-axis value is incremented, and the inner loop repeats all the Υ-axis values again. Finally, the outer loop is terminated when the final χ & γ-axis value is reached. In the inner circle of FIG. 28, read the SEM image data at (X, γ) and cdata [i] [j] in step 506. A detailed example is shown in FIG. 29. Note that the grid method described here uses a rectangular grid structure with orthogonal X and γ axes. Know that no rectangular grid is required. Other grid shapes can be used. For example, a triangular or trapezoidal mesh can be used. The grid structure is selected to ensure that periodically repeated contact patterns will be detected. Next, in step 508 of FIG. 28, the position of the contact hole is determined. Step 508 is detailed and illustrated in Figures 30A-30D. The identification of the contact hole position includes selecting the grid type and pattern that can be used to verify the contact hole. The process details of Figures 30A-30D are moved pixel by pixel along the first selected dimension (horizontal), and all pixel intensity values are summed in the second orthogonal dimension (vertical). When a significant change (jump) in intensity is measured, ′ defines the margin. The process continues until the measured strength drops significantly and the relative edges of the holes are defined. This method is used until all holes are positioned. It should be noted that steps 550 and 582 in Figures 30A-30D are absolute values of the difference in intensity of use. The reason for this is the difference in intensity or the magnitude of the contrast, which is very important in defining the location of the joint. This approach can be adapted to the different habits that define holes as high or low density. This paper size is applicable to China National Standard (CNS) M specification (210X297mmt) (please read the notes on the back first and fill in this page)

34 Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs 40276S Λ_ _____ Η 7 V. Description of the invention (32) In step 514 of Fig. 28, the contact hole profile is calculated. The details of this process are explained in the flow chart of No. 31A-3D. The side recording is calculated by analyzing each contact hole ′ according to the process described in the previous figure 30A-30D. Calculate a profile for each well. In a specific example, the 'side recording' is a sum of the intensity values in one dimension via positions along another orthogonal dimension. The intensity values at each position can be averaged and lumped to generate side recordings. It should be noted that the usual variables F and F2 are used in the flowcharts of Figures 31A-31D. These variables can be exchanged with the variables BSEN and CDN as defined in the formulas (5) and (6) according to the preferred embodiment of the present invention. In step 516 of Fig. 28 ', the contact hole is inspected according to the present invention. This side

Examples of method details are illustrated in Figures 32A-32B. As mentioned above, the values determined according to Figures 31A-31D are analyzed to classify each hole based on 9 contact types. As mentioned above, the variables F and F2 in Figures 32A-32B can be interchanged with the variables BSEN and CDN. ○ Referring to Figure 28 again, the Y-axis value is incremented at step 518, and it is determined at step 520 'whether to reach the maximum Y-axis value. If not, the process returns to the top of the inner loop. If so, the X-axis value is incremented at step 522, and the flow returns to the outer loop top through block 524 at step 504 'where the γ-axis value is initialized. When the outer loop ends, the result of the 'check process is displayed in step 526. As described above, FIG. 29 is a schematic flowchart illustrating the details of step 506 of reading SEM image data shown in FIG. 28. In step 528, the index j is initialized to 0 'and in step 530, the index i is initialized to 0. cdata [j] [i] is read at step 532 and index i is incremented at step 534. In step 536, it is determined whether the bow | i reaches the maximum value M. If not, the flow returns to step 532, where the data is read again. If it is, the index j is incremented at step 538 and the paper size is adapted to the Chinese National Standard (CNS) A <) specification (21〇 × 297) while-~~~- --Λ line (please read the notes on the back before filling this page) 35 Λ7 402769 ___— V. Description of the invention (33) Step 540 determines whether j reaches the maximum value n. If yes, the process ends. If not, the process Returning to step 530, the index i is initialized to 0 again and the process is repeated. In a specific example, the contact failure inspection of the present invention is performed after the contact hole is formed and the inside of the contact hole is cleaned (post-cleaning inspection: ACI). The failure inspection can also be performed on the exposed insulating layer on the wafer during the development process of forming the photoresist pattern on the contact hole to form the contact hole (post-development inspection: ADI). The invention can be applied not only to the contact hole but also to the contact hole. Each step of the through-hole is used to connect all the contact holes to directly contact the semiconductor substrate and the conductive layer. Also, the present invention can be used to check the pattern formation failure after the development process during the photo-processing contact hole formation. In addition, except for the circular shape contact In addition, the present invention is applied to inspect patterns by detecting various types of pattern images that are repeated by the rules. According to the present invention, it is possible to accurately and accurately detect the presence of a contact failure by digitizing a numerical value without inspecting the contact image with the naked eye or a microscope. For contacts with high aspect ratio, it is convenient and extremely accurate to confirm contact failure. In addition, the contact failure detection of the whole wafer surface can be performed in a short time to provide the result of contact failure detection 'and display the height of the mass production system line. Efficiency and productivity. It is obvious to those skilled in the art that various modifications and changes of the present invention can be made without departing from the essence or scope of the present invention. In this way, the present invention covers the present invention which belongs to the scope of the attached patent application and its equivalent scope. Modifications and changes. The paper size, financial standard, household standard rate (CNS), A4 size (2K), X297, 幻, ζ ^ 线 ~~-/ / __ (Please read the notes on the back first and fill in this page) Ministry of Economy Printed by the Consumer Standards Cooperative of the Central Standards Bureau 36 ΓΓ Printed by the Consumer Cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs ~ ----- Description of the invention (34 Λ7 Η 7 component label pair 100 Scanning electron microscope system 102 Electron 104 Condenser 106 Aperture 110 Reference or inspection surface Π4 Signal amplifier 118 Cathode ray tube processing unit Vacuum forming part Reference alignment part Electron beam deflector Autofocus controller Main controller contact fault detection mode Group 115 11 13 15 18 21 60 80 108 objective lens 112 signal detector 116, 122 deflection coil 120 scanning circuit 10 '20, 30' 40 host computer 12 reference transfer unit 14 electron beam generator 16 signal detector 19 main display 22 main Storage unit, memory 70 contact critical dimension measurement module 60a SEM signal reading unit, please read ii back ¢). Order a secondary computer 60b Graphic file transmission network module 60c Graphic file / SEM signal conversion module 60 d Contact position confirmation module 60e contact profile calculation and background removal module 60e (l) contact profile calculation module 60e (2) background value removal module 60f contact failure detection The result of module 60g shows that the module 110 wafer I30 semi-conductor Zhao base material paper size 剌 t ϋ ® housekeeping standard (CNS) A4 specification (210 × 29ϋ1 line 37

40276S Description of the invention (35) 131 Field oxide layer 132 Gate 133 Spacer 134 First insulating layer 135 Bit line 136 Second insulating layer 137 First contact hole 138 Second contact hole 150 Imaginary horizontal axis grid line 152 Imaginary Vertical axis grid line 153 Contact hole --------- f ---- ^ — 4T ------ ά (Please read the notes on the back first ^^ Fill in this page) Central standard of the Ministry of Economy The paper size printed by the Bureau Shell Consumer Cooperative is applicable to the Chinese National Standard (CNS) A4 specification (210 ×: 297 g) 38

Claims (1)

六 〇a76S Α8 Β8 C8 D8, the scope of patent application 2. 3.-A method for inspecting at least a part of the semiconductor wafer, including: "Haikou (5 knife semiconductor wafer read scanning electron microscopy image data; ^ 4 ^ image data identifying a feature on the semiconductor wafer from the data of the semiconductor wafer; calculating the parameter associated with the feature from the image data of the feature; comparing the parameter with the acceptable range of the parameter; and according to the parameter Acceptable value comparison results categorize this feature. For example, in the scope of item i of the patent application, one of the + _ circuit contact holes. 2 Please be full-time «2 slave method, if the parameter is in the range of acceptable values Otherwise, the contact hole can be classified as non-open. 5. 5. 6. Printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs 7. 8. 9. If the application of the first scope of the patent, the ancient, earth ^ Δ ^ Μ The money parameter is outside the range of the acceptable value of the parameter. This feature can be classified as the method of Taibei, such as the 1st patent application range. If the parameter is within the acceptable range of the number, Heading This feature can be categorized as a method that can be used as a method of drinking, such as Tai Yegan A ^ ^, where the SEM image data is generated by secondary electrons and backscattered electrons. Dimensions such as the method of applying for the scope of the patent application item 1. The number of revolutions includes features such as the method of applying for the scope of the patent application item No. 1 and the method of too A whistle, where the parameter includes the number of pixels of the SEM image data associated with it. ^ 特 如Apply for full-time ㈣1 item M, where this parameter contains the special 丨 ~ W --.----- install -----.- I order ----- J line-.,, (Please read first Note on the back page to fill in this page) This paper size applies the Chinese National Standard (CNS) Α4 ^ Γ (2Ϊ〇 ^ 297 ^ 'The average intensity of the pixels associated with the scope of patent application for the printed consumer patent cooperatives of the Central Standards Bureau of the Ministry of Economic Affairs. Ίο .. For example, the method of applying for the scope of the patent scope of item 1 further includes a profile of the pixel intensity of the image. 11. The method of applying for the scope of the patent scope of the item 10, wherein the calculating the pixel intensity of the profile includes The pixel intensity value of the area of the feature is deducted from the background intensity value. $ 1 2. If the method of item 1 of the scope of patent application, the further steps include: calculating the second parameter t related to the feature from the image data of the feature; comparing the acceptable value range of the second parameter with the second parameter; A comparison of the acceptable range of the second parameter categorizes the sign. 13. The method of item 12 of the patent application range, wherein the second parameter includes one of the dimensions of the feature. 14. The method of applying the patent _ item 12 , Where the second parameter includes the number of pixels of the SEM image data associated with the feature. 15. For the method of claim 12 in the scope of patent application, wherein the second parameter includes the average intensity of the pixels associated with the feature. 16. In the method of claim 12, the feature is classified as acceptable only if the first parameter belongs to the acceptable range of the first parameter and the second parameter belongs to the acceptable range of the second parameter. 17. The method according to the scope of patent application, further comprising characterizing the feature using a standard system, the characterizing includes: superimposing a coordinate system on the image of the portion of the semiconductor wafer; and See (Please read the note on the back ^ Fill this page first) The size of the bound and bound paper is applicable to China National Standard (CNS) A4 (21 〇χ297 公 酱) 40 40276 & Patent application scope ABCD Central Standards Printed by the Bureau Cooperative Consumer Cooperative, a number of locations on the first axis of the ~ k cliff drilling system are analyzed for pixel intensity values set on the second axis along the coordinate system. ^ The method of claim 17 in the patent application range, wherein the analysis includes summing the pixel intensity values set along the second axis. 1 The method of claim 18 of the patent scope, wherein the analysis further includes detecting the sum of the intensity values of a plurality of positions along the-axis to detect the feature. 20. The method of claim 17 of the scope of patent, wherein the The analysis includes averaging the intensity values of the pixels set along the second axis. 21 · ^ The method of applying for the 2G item, wherein the analysis further includes detecting changes in the average intensity value for multiple positions on the-axis. Detect the feature. 22. If the method of claim 17 of the scope of patent application, the feature includes determining the feature size. 23. If the method of claim 17 of the scope of patent application, the feature includes determining the location of the feature. The method of item 17 of the patent, wherein the characterization includes identifying most feature patterns. 25. The method of item 24 of the patent application, wherein the pattern is a periodic pattern. 26. The item 7 of the patent application, Gule / Negative method, where the coordinate system is a moment coordinate system. 27. For the method in the scope of patent application No. 17 呗, where the coordinate system is a triangle coordinate system. Please first note the note i (10 × 297 mm) 41 A8 68 C8 D8 where the coordinates are trapezoidal, where the SEM image data, which is the SEM image data, 40276S, the scope of the patent application 2 8 · If the method of the scope of the patent application, the method coordinate system 17. 29. For example, the method of applying for the first item of patent scope is in the form of digital pixel gray scale value. 30. The method of applying for the first item of patent scope is in the form of digital pixel code color value. 31_-Inspection of at least a part of a semiconductor wafer The device includes: a reading device for reading scanning electron microscope (SEM) image data of the semiconductor wafer; and an identification device for identifying the semiconductor wafer in the data of the semiconductor wafer. Image data of a feature; a computing device that calculates a parameter associated with the feature from the image data of the feature; a comparison device that compares the parameter with an acceptable range of the parameter; and a classification device that It is used to classify the feature according to the comparable result with the parameter. 32. For the device in the scope of patent application No. 31, the feature is in the product The contact hole of the road "33. If the device in the scope of patent application No. 32, if the parameter is outside the acceptable range of the parameter, the device with the classification feature can classify the contact hole as unopened. I 34. If The device in the scope of patent application No. 31, in which if the parameter is outside the acceptable range of the parameter, the device with the classification characteristic can classify the M-scale scale applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) 4 3 Install IJ 7 order line (please read and read the notes on the back page to complete this page) Printed by Zhengong Consumer Cooperative, Central Standards Bureau, Ministry of Economic Affairs, 42 A8 B8 C8 D8 40276 6. The contact hole for patent application is faulty. 35. If the device in the scope of patent application No. 31, if the parameter is within the acceptable range of the number of parameters ^ — c Jingxianwen on the back of the mackerel, fill in this page), the device of the classification characteristics can be The classification section is characterized as acceptable. Λ 36. The device according to item 3 丨 of the scope of patent application, wherein the SEM image data is generated by secondary electrons and backscattered electrons. 37. The device as claimed in claim 31, wherein the parameter includes the dimension of the feature. 38. For the device in the scope of application for item 31, the parameter includes the number of pixels of the SEM image data associated with the feature. 39. The device as claimed in claim 31, wherein the parameter includes the average intensity of the pixels associated with the feature. 40. The device according to item 31 of the scope of patent application, which further includes a device for calculating an image pixel intensity profile for the feature. 41. For example, the device in the scope of patent application No. 40, wherein the device for calculating the pixel intensity profile of the image includes a device that subtracts the background intensity value from the pixel intensity value of the area including the feature. Printed by the Central Ministry of Economic Affairs # Associate Bureau Consumer Cooperatives 42. For the device in the 31st scope of the patent application, it further includes a computing device, which is used to calculate the second parameter of the relevant feature from the image data of the feature A comparison device for comparing the second parameter with an acceptable value range of the second parameter; and a classification device for classifying the feature based on the comparison with the acceptable value range of the second parameter. This paper card applies Chinese code standard (CNS) Ba Na 43 A8 B8 C8 D8 40276 ^ Patent application range 43. For the device with patent application item 42, where the second parameter includes one dimension of the feature "44. Such as patent application The device of the range 42 'wherein the second parameter includes the number of pixels of the SEM image data associated with the feature. 45. The device according to item 42 of the patent application, wherein the second parameter includes the average pixel intensity associated with the feature. 46. If the device of the scope of patent application 42 is applied, the feature is classified as acceptable only if the first parameter belongs to the acceptable value range of the -parameter and the second parameter belongs to the acceptable value range of the second parameter. 47. If the device of the 31st scope of the patent application, further includes a device for characterizing the feature using a coordinate system, the characterization includes: 4-position device 'It is used to superimpose the coordinate system on the part of the semiconductor wafer On the image; and an analysis device for analyzing the pixel intensity values set on the second axis along the coordinate system at a plurality of positions along the first axis of the coordinate system. 48. The device of claim 47, wherein the analysis device includes a device for summing the pixel intensity values set along the second draw. 49. For example, the apparatus of item 48 of the patent scope, wherein the analysis apparatus further includes detecting a change in the total intensity value for a plurality of positions along the first axis, and a device for detecting the feature. 50. The device according to item 47 of the patent, wherein the analysis device includes a device for averaging pixel intensity values along the second axis. 51. If the device of the scope of application for patent No. 50, wherein the analysis device further includes a change in the average intensity value of a plurality of positions along the first axis (please read the precautions on the back first \? ^ Fill this page)-binding line Printed by the Consumer Cooperatives of the Central Government Bureau of the Ministry of Economic Affairs-H · ^^ 1 —.1 40276 Scope of patent application: The device is printed by the Consumer Cooperatives of the Central Bureau of Standards of the Ministry of Economic Affairs to detect this feature. 52. The device of claim 47, wherein the characterization device includes a device that determines the size of the feature. It is assumed that the device under the scope of patent application No. 47, wherein the characterization device includes a device that determines the feature position. 54. The device of claim 47, wherein the characterizing device includes a device for identifying a plurality of characteristic patterns. & The device according to item 54 of the patent, wherein the pattern is a periodic pattern. 56. The device according to item 47 of the patent application, wherein the coordinate system is a rectangular coordinate system. 57. The device according to item 47 of the patent application, wherein the coordinate system is a triangular coordinate system. 58. The device according to item 47 of the patent application scope, wherein the coordinate system is a trapezoidal coordinate system. 59. As for the 31st item in the scope of patent application, the SEM image is a gray scale value of digital pixel 60. As the HI item 3 of the scope of patent application, the SEM image data is a digital color coded pixel Value form. This paper size applies to Chinese National Standard (CNS) A4 (210X297 mm) I I— I I ^ Packing ^ —1 '" I Line 4 * 1 (Please read the notes on the back first ^ rFill this page) 45