DE19860704B4 - Method for checking at least a part of a semiconductor wafer with a scanning electron microscope - Google Patents

Abstract

Method for checking at least part of a semiconductor wafer with the following method steps:
- reading image data of a scanning electron microscope (100) for the part of the semiconductor wafer;
- Identification of image data for a feature of the semiconductor wafer from the data for the part of the semiconductor wafer;
- calculating a parameter associated with the feature from the image data of the feature;
Comparison of the parameter with a range of parameter values to be accepted;
- Classification of the characteristic according to the comparison with the parameter values to be accepted;
characterized by the following features:
Superimposed on the image of the scanning electron microscope (100) for the inspecting part of the semiconductor wafer is a grid with grid axes defining a plurality of meshes, each mesh having a predetermined number of pixels, such that each mesh or selected meshes has a predetermined number of pixels includes features to be analyzed simultaneously; in which
- by adjusting the distances of the grid axes the number of pixels ...

Description

Field of the invention

The The present invention relates to the verification of semiconductor devices. It particularly concerns the detection of faulty contacts, such as for example non-open contact openings, using a Scanning electron microscope.

Background of the invention

integrated Circuits are made by first discrete semiconductor devices be formed in silicon wafers. In the components will now formed in several levels a metallic interconnection network, that makes contact with the active parts of the components and they to produce the desired Connecting circuits together. The connecting layers become formed as follows: depositing an insulating layer over the discrete components, patterning and etching of contact openings in this layer and depositing a conductive material in these openings. Above the Insulating layer will now typically be a conductive layer applied. The conductive layer is now patterned and etched to to form a first circuit level connections between the Make contacts of the components. By generating a circuit with several levels, the following process steps are repeated carried out: Deposition of an insulating layer, formation of contact openings or Vias, formation of conductive material layers, structuring, etc.

In dependence from the complexity The integrated overall circuit is typically a plurality of metallic ones Layers, for example, two to four levels, required to complete the force To create connections and these connections with contact points to connect, which has an external connection to the completed circuit enable. A high density of integrated circuits with dimensions in the Submicron range makes an extremely accurate control the dimensions required. It will be very sensitive surveillance procedures for examining the connection structure and / or the contact openings needed about the accuracy of the dimensions and the flawlessness of the structure of the draft. These requirements are increasing with density and miniaturization circuits are more stringent, such as in mass production of semiconductor memory devices such as 64M DRAMs or 256M DRAMs, for the currently Typically, circuit dimensions of 0.25 to 0.30 μm are required are.

• JP 07220077 A
• JP 10074813 A
• JP 06347246 A
• US 5 544 256 A
• US 5 493 116 A
• JP 06295330 A
• EP 0 441 375 A2
• US 5 479 535 A
• DE 197 02 862 A1
• JP 61124810 A

Checking contact openings for their condition, such as a non-open or non-free condition, is becoming increasingly important as the aspect ratio (AIR) of contact openings, ie the ratio of their depth to their diameter, increases with the need for higher densities has increased the components. However, a conventional light microscopy method with a wavelength of 488 nm in the visible region is technically limited in view of checking the internal structure of the contact holes because it has too low a resolution for checking the internal structure of the contact holes having a size of 200 nm or less , It is also not possible to produce light spot sizes of 1 μm or less using light microscopy. The prior art refers to the following publications:

• JP 07220077 A
• JP 10074813 A
• JP 06347246 A
• US 5 544 256 A
• US 5,493,116
• JP 06295330 A
• EP 0 441 375 A2
• US 5,479,535 A
• DE 197 02 862 A1
• JP 61124810 A

The various documents describe procedures for reviewing a Semiconductor wafer, wherein also scanning electron microscopes are used reach.

The from the document JP 07220077 A known features are summarized in the preamble of claim 1.

Summary of the invention

The The object of the present invention is to provide a verification method for faulty Contacts or contacting errors, by means of which a precise check by means of digitized values on contact images and not by the naked eye or through the microscope is, and by which the disadvantages of the prior art avoid it. The invention should also be advantageous for detection of contact errors on contacts with a large aspect ratio, the is called a big one relationship applicable between the depth and the diameter of a contact be and especially by a high speed and the Fitness for the application in mass production and in the production of semiconductor devices allow greater yield. Advantageously, a faulty photoresist structure after a Development processing step in a photolithography process allows become.

The inventive task is achieved by a method having the features of claim 1. In the dependent claims particularly advantageous embodiments of the invention are given.

According to the invention the image data of a Scanning Electron Microscope (SEM = Scanning Electron Microscope) for the Part of the semiconductor laser read. From this SEM image data will be Image data for a particular feature or property of the wafer identified. A parameter related to this feature is calculated and with a range of acceptable values for this parameter compared. Based on this comparison between the parameter and the range of values still to be accepted can classify the feature.

at an embodiment of the invention is used as a computer parameter the dimension or the size of the feature used.

If the feature, for example, a contact opening in an integrated Circuit is, for example, the parameter is the diameter the opening measured in image data pixels. A certain contact opening can For example, be determined by having a width of 20 pixels has. In another embodiment can use as parameter an average pixel intensity for pixels used, which are within the feature. If the feature a contact opening is, for example, as the parameter, the average of the pixel intensities for the pixels used that the contact opening assigned. If the measured parameter is within a range of acceptable values for is this parameter, the feature is classified as acceptable. If the parameter is outside this range of acceptable values for this parameter is classifies the characteristic as defective. If the feature is a contact opening is, the opening becomes detected as faulty, for example, if it is not open.

at an embodiment of the invention be for the feature calculates two parameters. For example, the two parameters a dimension of a feature, such as a contact opening, be measured in pixels associated with the feature. The second Parameter can be the average or middle pixel intensity of the Feature associated pixels. Both parameters are predetermined Ranges compared to acceptable values for these parameters. If both parameters are simultaneously within their respective range to be accepted values becomes the feature, such as a contact opening, in one embodiment classified as acceptable. A contact opening is under these circumstances as open, properly measured and properly shaped classifiable. The relationship between the parameters and their corresponding ranges be used to place the feature in one of several categories classify. The parameters can depending on whether they are below, within, or above one still to be accepted Range of values are used to classify a feature become.

at an embodiment become the SEM image data of both secondary and higher energy backscattered Generates electrons in the scanning electron microscope. The data values are digitized and can in the form of digitized grayscale pixel levels or pixel levels or color-coded pixel values.

In an exemplary embodiment according to the invention, a grid or mesh structure is used to characterize the feature to be checked, wherein, for example, the position and / or the dimensions of the feature to be checked are determined for characterization. The grid or mesh typically comprises two orthogonal axes to analyze the image of the superimposed on the wafer part. Alternatively, the grid axes may have any geometric dependence on each other, that is, they may be triangular, trapezoidal, etc., for example. In one embodiment, the grid location method determines the position, shape, and / or periodic structure of the feature by analyzing the pixel values along a line that is parallel to one of the orthogonal axes that successively positions at pixel positions along the other orthogonal axis becomes. For example, the grid method may include positioning a vertical line at multiple horizontal pixel locations and adding the vertical pixel intensity values at the respective horizontal location. The summed intensities may be compared at the respective horizontal location to detect an increase in intensity used as an indication of the presence of a feature, such as a contact opening. This method can be repeated for multiple pixel positions in one dimension or direction. It can then be repeated in the direction perpendicular thereto, so that the structure, the shape and the dimension of the features can be determined.

This Method can also be used to the optimal dimension a sub-grid or sub-mesh unit to determine the includes features to be analyzed. The grid method can be used, for example, the optimal dimension of a grid unit in pixels to determine the 100th includes simultaneously to be analyzed contact openings. These Procedure makes the inventive method for checking Characteristics are much more efficient because of unnecessary processing steps Optimization of the area of the area to be checked eliminate it.

at an embodiment become the SEM image data pixels for calculating an intensity profile the characteristic to be checked, d. H. a contact opening, used. In one embodiment becomes the intensity profile first by summing the pixel intensity values of a feature along a orthogonal axis generated at each of several pixel positions, which are arranged along the axis perpendicular thereto. At every For example, the horizontal pixel position becomes the pixel intensity values summed in the vertical direction, averaged and plotted as a function of pixel position on the horizontal axis. The pixel intensity profile is for classification according to the invention of the feature.

at an embodiment is used to normalize the intensity profiles of all features of all intensity values in the respective grid unit, the background intensity value is subtracted. This causes that the background value for the respective intensity profile reduced to zero. The normalized profile can now enter next Threshold can be set to have pixel intensities above of the threshold value to the characteristic to be checked be assigned. From the profile can now the above first and calculating the above second parameter. The dimensions of the feature for example, by counting the number of pixel positions in a first direction be, the summed intensity values in the perpendicular to this standing in the direction that exceed the threshold. Since it is assumed that a Sum of pixel intensity, that exceeds the threshold, associated with the feature is the number of pixel positions with a sum exceeding the threshold, a measure of the characteristic in one direction, measured in pixels. The second parameter can by calculating a mean intensity value exceeding the threshold be calculated. These two parameters can be set with their respective predetermined Range to be accepted values to the predetermined Classify feature in one of the predetermined classes to feature types.

The inventive verification method has many advantages over the prior art. For example, in certain prior art methods optical methods, such as light microscopy or verification with naked Eye, used to detect contact or contact errors. These systems are unsuitable for resolving minor feature errors, which lead to the failure of the circuits. The use according to the invention a scanning electron microscope ensures a much better resolution, with which can also detect minor errors. The present invention is therefore also applicable to the characteristics of current circuits, whose dimensions are in the submicrometer range. Due to the lattice process according to the invention In addition, the processing of the pixel data according to the invention is extraordinary efficient. The processing and error detection can be very be done efficiently and quickly, so that the inventive verification method and the verification system of the invention very good for the mass production of wafers and integrated circuits suitable is.

Another aspect of the present invention relates to a verification method for bonding defects in semiconductor devices, comprising the steps of: mounting a processing cartridge with wafers, on the surface of each of which a plurality of contact openings are formed; Removing a particular wafer from the cassette and loading a table in a reference chamber of a SEM with the wafer; Aligning the wafer for scanning by an electron beam; Moving the table with the wafer mounted thereon to a specific location associated with an incident direction of the electron beam of the SEM; Opening a shutter for the scanning electron beam to a specific location of the wafer; Self-addressing or self-drive for detecting the inspection position by detecting a prestructured reference image formed on the wafer; Deflecting the electron beam of the SEM into the inspection position; Self-focusing to produce another clear image by repeating the electron beam excursion; Closing the shutter to separate the self-focused wafer from the electron beam; Checking for an erroneous contact or a contact failure by comparing the electron signal values detected from a unit area with at least one contact opening after deflecting the electron beam with an electron signal value corresponding to a normal contact; further checking a contacting error in another position of the wafer by moving the table to the other position and repeating the same method steps; and further checking all of the other wafers in the cassette for contacting errors by taking the wafer being inspected and inserting the other wafers into the reference chamber and repeating the same process steps.

According to one Another aspect of the present invention includes a method of manufacture of semiconductor devices, the following process steps: training from contact openings for certain insulating material layers formed on a semiconductor substrate; Review of the Contacts of the contact openings by comparing each one of a surface with at least one contact opening detected electron signal values with the electron signal values corresponding to a normal contact and implementation the subsequent processing steps of the manufacturing process for semiconductor devices after the review took place Introducing conductive material layers into the contact openings.

The verification step for contacting errors can for a certain scanning position are performed on the semiconductor substrate, for example, the verification step for contact errors to use a production line for mass production. To Completion of development processing to form the photoresist pattern can the verification step for contacting errors also at the bottom of the photoresist pattern to form contact openings carried out become.

One Another aspect of the present invention relates to a method for the production of semiconductor devices with the following process steps: Formation of a photoresist contact opening structure for the formation of contact openings for on insulating material layers formed on a semiconductor substrate; and review of the respective contact of the contact openings by comparing the of a unit area with at least one contact opening structure detected Electron signal values corresponding to those of a normal contact structure Electron signal values.

Brief description of the drawings

In the drawings, which are not necessarily drawn to scale, are the same components designated by the same reference numerals. The Drawings serve to illustrate the underlying of the invention Principles.

It demonstrate:

1 a schematic block diagram illustrating the operation of a Scanning Electron Microscope (SEM);

2 a schematic representation of the energy spectrum of the electrons emitted in a SEM, including the secondary electrons and the backscattered electrons, when irradiating a comparison sample with an electron beam;

3 a schematic block diagram illustrating the construction of an in-line SEMs;

4 a schematic block diagram of an embodiment of a verification system for contact failure in semiconductor devices;

5 a schematic block diagram of an alternative embodiment of a verification system for contact failure in semiconductor devices;

6 a schematic block diagram of another alternative embodiment of a contact failure checking system in semiconductor devices;

7 a schematic block diagram of another alternative embodiment of a contact failure checking system in semiconductor devices;

8th a schematic block diagram of a functional block diagram of an embodiment of a verification method according to the invention for contacting errors and a corresponding system;

9 a schematic block diagram of a functional block diagram of another embodiment of a method according to the invention for contacting errors and a corresponding system;

10 a schematic block diagram of a functional block diagram of another embodiment of a method according to the invention for contacting errors and a corresponding system;

11 a schematic diagram of a flowchart illustrating the logical sequence in an embodiment of a method according to the invention for contacting errors in semiconductor devices;

12 a schematic flow diagram illustrating the logical sequence in an alternative embodiment of a method according to the invention for contacting errors in semiconductor devices;

13 a schematic flow diagram illustrating the logical sequence in another alternative embodiment of a method according to the invention for contacting errors in semiconductor devices;

14 a schematic flow diagram illustrating the logical sequence in another alternative embodiment of a method according to the invention for contacting errors in semiconductor devices;

15 a schematic diagram of a pattern of chip scanning on a wafer for the verification of Kontaktierungsfehlern invention;

16 a schematic diagram illustrating details of the scanning areas at the chip scanning according to 15 ;

17 a schematic cross-sectional view of a semiconductor device with contact openings formed therein, in which the method according to the invention for contacting errors is applicable;

18 SEM image data for contact openings after a grid has been determined by a contact point detection method according to the invention;

19 a schematic representation of a grid for performing an embodiment of the contact detection method according to the invention;

20 a schematic representation of a grid for performing an alternative embodiment of the contact detection method according to the invention;

21 a schematic representation of the relationship between a contact unit and horizontal and vertical pixel units for an inventive embodiment;

22 an intensity profile of a contact unit before performing a background value correction in an embodiment according to the invention;

23 an intensity profile of the contact unit according to 22 after performing a background value correction;

24 an intensity profile of a SEM image of contact units after performing a background value correction;

25 a table of codes for identifying results in an embodiment of the inventive method for contacting errors according to the invention;

26 a table with a part of the results in the examination of contacting errors according to an embodiment of the invention;

27 a flow chart illustrating the logical sequence of a semiconductor device processing sequence according to the invention;

28 a schematic diagram of a flowchart for illustrating the logical sequence in an embodiment of the contact verification method according to the invention;

29 a schematic diagram of a flowchart for illustrating the logical sequence when reading the image data of the scanning electron microscope according to the illustrated method in 28 ;

30A to 30D in a schematic representation of a flowchart for illustrating the logical sequence in the detection of contact opening positions according to the illustrated method in 28 ;

31A to 31D a schematic diagram of a flowchart for illustrating the logical sequence in the calculation of contact opening profiles according to the illustrated method in 28 ;

32A to 32B in a schematic representation of a flowchart for illustrating the logical sequence in the examination of contact openings according to the illustrated method in 28 ,

Detailed description more preferred embodiments

1 shows a schematic block diagram of a scanning electron microscope system 100 , which can be used for the inventive examination of contact openings in semiconductor devices.

1 shows an electron gun 102 passing an electron beam through a condenser lens 104 shoots. The beam passes through a deflection coil 122 , an iris diaphragm 106 , an object lens 108 and a panel 124 , The focused electron beam strikes a reference or sample surface 110 which he sweeps or scans. At the surface 110 it can be the surface of a semiconductor wafer to be checked. Secondary electrons and backscattered electrons emitted from the reference surface are detected by a signal detector 112 which generates signals indicative of the received electrons. The detected electron beam is passed through a signal amplifier 114 strengthened. The amplified signal is transmitted across the fluorescent surface in a cathode ray tube (CRT). 118 deflected so that a visible image of the reference surface is formed.

The sweeping of the surface of the CRT 118 through the beam is through a deflection coil 116 which is correlated with the scanning of the reference surface passing through the deflection coil 122 is controlled. In the SEM, the scanned surface of the reference surface is divided into fine pixels or picture elements. The respective electron signals detected by the pixel elements are successively transmitted one after the other to form the SEM image. That the signal amplifier 114 passing electron signal becomes a sampling circuit 120 through which the deflection angle of the electron beam in the second deflection coil 122 is controlled.

In addition, the amplified electron signal data for the individual pixels may become a processing unit 115 which fulfills numerous signal conditioning and signal processing functions. The processing unit 115 may convert the electron signal for each pixel to a discrete gray scale value or a color coded value to produce an image. The gray scale value can assume one of 256 possible levels or states digitally coded by a binary value between 0 and 255. For storing the gray scale values for the individual pixels, a memory device may be used. A computer, which is also part of the processing unit, can process the image values as desired. In one embodiment, the computer may be programmed to analyze the gray scale data to perform the inventive contact verification described in detail below.

3 shows a schematic diagram of a block diagram with the characteristic components of an in-line SEM system, in which the contact verification can be done in-line. In prior art methods, the SEM image data is collected and analyzed off-live, ie apart from the manufacturing process. Since the in-line method of the present invention is much more efficient, the SEM image data can be collected and analyzed during the manufacturing process, eliminating the need for additional verification steps common to conventional methods. The in-line SEM system includes an electron optical section, a reference section, a vacuum section, and an electrical section. The electron optical section comprises an electron beam generator 14 , a deflection device 15 for the electron beam and a signal detector 16 , The reference section comprises a reference transport part 12 for transporting the reference sample, ie the wafer, from a cassette to a reference chamber and a reference alignment part 13 , The vacuum section includes a vacuum generating part 11 to maintain a negative pressure in the reference chamber. The electrical section includes a main computer 10 with a main controller 21 for controlling the electron optical section, the reference chamber, the vacuum section and the other system components. The electrical section comprises a main memory unit or a main memory 22 for storing the detected signal data values of the signal detector 16 , The electrical section also includes a main display or main screen 19 to display the images generated from the detected electronic signals. An automatic focusing control device 18 provides automatic focusing for clear or sharp images.

In a typical scanning electron microscope, the reference surface is irradiated with the electron beam, whereby secondary electrons are generated and emitted from the reference surface. The electrons of the beam can also be scattered through the surface. 2 shows an energy spectrum of the emitted from the reference surface and scattered at her electrons when irradiating the reference surface. As in 2 As can be seen, the largest number of secondary electrons (SE) is found in an electron energy band of less than 50 eV, while most backscattered electrons (BSE) can be found in a much higher energy band. Commonly used in-line SEMs use the secondary electrons (SE) generated in the low energy band of about 20 eV to give clear or sharp images of surfaces and edges. When checking certain features, such as contact openings with a large aspect ratio, ie the ratio of the opening depth to the opening diameter, the secondary electrons generated in the contact openings can be dissipated as they pass through the contact openings, so that the contact openings are not clearly or sharply imaged. Since these features are typically visually inspected with the naked eye, clear images are essential for detecting bad contacts.

The 4 to 7 show in schematic representation block diagrams of different embodiments of a verification system for contacting errors in semiconductor devices, which makes use of the inventive method. This in 4 The illustrated system includes some of the components of the in-line SEMs 3 , It is covered in a computer 20 but also a verification module 60 for contacting errors. As described above, the in-line SEM includes an electron optical section with an electron beam generator 14 a deflector 15 for the electron beam and a signal detector part 16 , As a signal detector, a detector is preferably used which can detect both secondary electrons (SE) and backscattered electrons (BSE), which are emitted after the irradiation of the reference surface by the electron beam. The system also includes a reference chamber with a reference alignment part 13 for rotating or tilting the table on which the reference wafer to be inspected is located while being moved along the X, Y or Z axis. A vacuum generating part 11 serves to maintain a desired negative pressure in the reference chamber. A reference transport part 12 serves to transport the reference sample into the reference chamber. The electrical section includes a main computer 20 with a main control part 21 for controlling the electron optical section, the reference chamber, the vacuum generating part 11 and the other subsystems. A main storage unit or main storage 22 stores those from the signal detector 16 detected signals. A main display or main display 19 shows that from the detected electroni signal derived from the signal. An automatic focusing control device 18 automatically focuses to produce a clear or sharp image on the display. This embodiment also includes a verification module 60 for contacting errors for analyzing the information contained in the electron beam received from the signal detector 16 transferred and in the main computer 20 are stored in order to check the contacts according to the invention.

5 FIG. 12 illustrates another embodiment of a contactor failure checking system in semiconductor devices. FIG. This in 5 The system illustrated also includes some of the components of the in-line SEMs 3 , One difference is that he also has a verification module 60 for contacting errors and a measuring module 70 for Critical Dimensions (CD) of contacts residing in the main computer 20 can be located. A critical dimension is the dimension of a particular feature to be checked. For example, in the case of a circular opening, the critical dimension may be the diameter of the opening. In one embodiment, the measurement is performed by the measuring module 70 for critical contact dimensions (CD) by comparing the contact diameter determined from the contact images generated by the SEM with pre-stored standard values.

6 FIG. 12 illustrates another embodiment of an in-line device for contact failure in semiconductor devices. FIG. The system according to 6 also includes some of the components of the in-line SEMs 3 , One difference is that he also has a sub-computer 80 may include, via an interface with the main computer 10 connected is. In the sub-computer 80 there is a verification module 60 for contacting errors. As a subcomputer 80 For example, a conventional commercially available personal computer having a sub-screen and a sub memory unit or a sub memory may be used. The occurrence of contacting errors is determined by analysis in the main memory unit 22 stored contact electron signal data through the contacting error checking module 60 in the subcomputer 80 detected.

7 FIG. 12 illustrates another embodiment of a contactor failure checking system in semiconductor devices. FIG. The system according to 7 that also includes some components of the in-line SEMs 6 includes one via an interface with the main computer 40 connected subcomputers 80 , The system according to 7 includes in the sub-computer 80 also a contacting error checking module 60 as well as in the main computer 40 a measuring module 70 for critical contact dimensions.

The 8th to 10 show in schematic representation functional block diagrams for different embodiments of the contacting error checking module 60 , As in 8th can be seen, a contacting error checking module 60 a reading module 60a for SEM signals receiving the SEM signals indicative of the electrons received by a wafer irradiated by an electron beam. The SEM signal is passed through a contact position detection module 60d analyzed to determine the positions of the contact openings and / or other features to be checked. A module 60e For calculating contact profiles and eliminating the background, intensity profiles for the contact openings are generated using the SEM signal data. The intensity profiles are typically normalized by eliminating the data generated by background intensity effects, so that the shape of the intensity profiles can be checked independently of background effects. A contact failure check module 60f analyzes the intensity profiles of the contact openings to identify contact errors. As will be described in detail below, in one embodiment of identifying a faulty contact, the average intensity value of the contact is compared to a predetermined threshold. A result display module 60g is used to display the results of the error analysis.

In one embodiment, the read module reads 60a for the SEM signals, the digitized electron signal information of the contacts stored in the main memory unit or the main memory 22 of the main computer 20 get saved. The in-line SEM digitizes the intensity of the electron signals detected due to the scanning by the electron beam, and stores the intensities as a gray scale level or a color coded level. In one system, the grayscale value associated with the pixels corresponds to one of 256 possible values between 0 and 255. The highest intensity is defined as 255, while the lowest intensity is defined as 0. The digitized intensity values are color-coded by the individual pixel elements, ie, as grayscale value. The contact images are generated by reading the grayscale values for the individual pixels sequentially in time and displaying the pixel images on a cathode ray tube, a screen and / or a printer. The gray scale values can also be converted to a color for a color screen.

In the 9 illustrated embodiment of the contacting error checking module is a modification of the embodiment according to 8th , In 9 lie the contacting profile calculation module 60e (1) and the background value elimination or correction module 60e (2) as separate modules before while in 8th as a common module 60e are shown.

10 shows a schematic representation of a functional block diagram of another embodiment of the contacting error checking module 60 , The illustrated contacting error checking module 60 includes a graphics file transmission network module 60b , a conversion module 60c for converting the graphics file into an SEM signal, a contact position detection module 60d , a combined contact profile calculation and background correction module 60e , a contact failure verification module 60f and a display module 60g for the results.

The graphics file transmission network module 60b is a device for signal transmission between the main computer 10 . 40 and the sub-computer 80 (please refer 6 ). The module 60b converts the information in the main storage unit 22 of the main computer 10 . 40 stored digitized electron signals of a contact in a graphics file and transmits them to the sub-computer 80 ,

The conversion module 60c from the graphics file to an SEM signal reads the color-coded values, ie gray scale values, to the sub-computer 80 transferred graphics file and converts them into a digitized SEM signal. The contact position detection module 60d , the contact profile calculation and background correction module 60e , the contact error verification module 60e and the module 60g for the presentation of the results were already above on the basis of 8th and 9 described.

The 11 to 14 contain flowcharts for different embodiments of the in-contact contact failure verification method for semiconductor devices by contact fault checking systems. The contact error checking part 60 according to 4 In a wafer having a plurality of contact holes, the contact error check is performed using the one in the main computer 40 installed in-line SEMs (see 11 ). First, at a predetermined position of the in-line SEM, a cassette having a plurality of wafers mounted thereon is set up with a plurality of contact holes each (S10). After removing a wafer to be checked from the cassette, a table is now loaded in the reference chamber of the SEM with the wafer (S12). Now, the bevel or segment portion of the wafer is aligned. Subsequently, the wafer for scanning by the electron beam is aligned (S14), and the table with the wafer deposited thereon is moved to a predetermined position in the incident direction of the electron beam of the SEM (S16).

The under the object lens arranged aperture is opened to a certain point to irradiate the wafer with the electron beam, wherein a self-drive or self-addressing (S20). By self-addressing a particular location is recognized by a pre-structured Standard image is applied to the specific location so that a check for the Standard picture can be done.

Now becomes the body to be checked irradiated with the electron beam of the SEM (S22) and the scanning through the electron beam is used to produce a clear or sharp contact image by the autofocusing control part repeatedly performed (S24). Subsequently the shutter is closed to scan the wafer through the End electron beam (S26).

Now is for all contacts detected by the scanning electron beam according to the invention have the intensity profile the electron signals are checked (S28). Then it is determined whether there is also a check on other parts of the wafer Contacting error should occur (S30). If necessary should be done in the flow chart a return to the step S16. The table is moved to other parts of the wafer and the above steps are repeated. After completion the contact error check becomes the wafer removed (S32). It is determined, if even more Wafer to be checked (S34). In this case, the reference chamber is with another Wafer loaded from the cassette and become the above steps repeated, so that for all wafers a contact error check is made in the cassette. After checking all Wafer is taken out of the cassette (S36) and the process thereby completed.

12 FIG. 12 shows the contacting-failure checking method for multi-contact hole wafers using the in-line SEM according to FIG 5 , in its main computer 30 the contacting error checking part 60 and the measuring part 70 are designed together for the critical contact dimensions. The check for contacting errors takes place according to 11 , but unlike the 11 the shutter is closed (S26), the contacting error check or detection is determined (S27), and the measurement of the critical contact dimensions is made (S29) if the contacting error check (S28) is not performed.

13 FIG. 12 shows an in-vivo contacting failure checking method for a multi-contact hole wafer using the in-line SEM according to FIG 6 in which the contacting error checking part 60 not in the main computer 10 but in the sub-computer 80 is housed. As in 13 2, after closing the shutter (S26), the SEM signal stored in the main memory of the main computer is transferred to the sub-computer, and it is determined whether or not to make contact-error check for other locations of the wafer (S31). The sub-computer receives the transmitted signal to perform the contacting error check (S37).

At the in 14 illustrated embodiment, the arrangement or position of the sub-computer and the main computer is reversed, if the contacting error checking part 60 is installed in the sub-computer and not in the main computer, as in 6 is shown. By the same process steps as in 11 when the iris is closed (S26), the transmission of the SEM signals stored in the main memory part of the main computer to the subcomputer is determined (S31-1), and the subcomputer checks for contacting errors (S31-2). It is also determined whether a contacting check is required elsewhere (S31-3, S31-4).

15 shows numbered shaded areas (# 2 to # 37) on a wafer 110 to be checked by the contacting error checking method of the present invention. A range labeled "AP" indicates an alignment point, while "# 1" indicates the chip-free focus location.

In the shaded areas in 15 each time several sampling points can be specified. For the chip or area # 2 in 15 shows 16 For example, five sampling points, the top left (2.1), top right (2.2), bottom right (2.3), bottom left (2.4) and in the middle (2.5). The sampling points or the sampling numbers can be selected in different ways within a chip unit to be scanned. In the present embodiment, 175 locations of 35 chip or sample areas are scanned, with five samples checked for each chip. In one embodiment, where an in-line SEM with a gain of 12.5k is used, there may be 98 contacts for each sample in a 480x480 pixel image. With five sampling points for each of the 35 areas, 17150 contacts are checked.

17 shows a schematic representation of a cross section through a semiconductor device with inventively to be checked contact openings. 17 shows the buried contact fabrication process of a 64M DRAM. Through a field oxide layer 131 an active region is determined that is above a semiconductor substrate 130 is trained. Above the active area is a gate electrode 132 formed by a spacer layer 133 is covered. After forming a first insulating layer 134 a high-temperature oxide film on the surface becomes a first contact hole 137 trained as direct contact to a bit line 135 serves. After forming the bit line 135 becomes a second insulating layer on the surface 136 trained as a BPSG. In addition, a second contact opening 138 formed as a word line.

As an example of the verification method according to the invention, during the manufacturing process for the 64 M DRAM, the buried contact for the formation of the word line is checked. As in 17 can also be the direct contact 137 be checked. The check can also be made after the development process of the photoresist pattern for the formation of these contacts.

at the verification method according to the invention for contacting errors is for the bodies to be checked each first an optimal image size selected on the dimension the characteristic to be checked, such as the diameter of a round contact opening based. In one embodiment comprises a typical SEM picture 480 × 480 Pixel.

Such an image can be used for each of the numbered digits in 16 to be created. Depending on the size and the distance between the contact openings, an optimal image size for the individual contacts is determined.

18 shows an example of a contact image of an in-line SEM with a gain of 12.5 K for a sample point of the semiconductor device. It consists of 480 × 480 pixels. The number in the picture contacts is 98, ie 14 in the horizontal direction and 7 in the vertical direction.

It will be the optimal resolution determined and based on the dimension of the feature to be checked, d. H. of the Contact opening and the distance between the openings checked. at a system can For example, the pixels each resolve 12 nm in the SEM. contact openings currently known to have diameter in the range of 200 nm. The number of pixels covering a feature is chosen to ensure, that irregularities or errors of the feature in the image are reliably recognizable. If an area to be examined, for example, 100 openings in the form of a grid with uniform mesh size, can a set of 100 sub-gratings with 48 × 48 pixels can be used to all openings including the area between the openings cover each opening a subgrid is assigned. The 480 × 480 pixel field can be a right-angled Grid or a rectangular mesh of horizontal and superimposed on vertical lines to check the openings each used to form 100 sub-grids with 48 × 48 pixels.

According to the invention will now determines if a grid with 48 × 48 Pixel is sufficient to resolve irregularities or errors of the opening to be checked. The Dimension of the opening is compared with the distance in the respective subgrid to to determine if the number of pixels covering the aperture is sufficient is to analyze the opening. The resolution is determined by the fact that the critical dimension, d. H. the diameter of the opening, by the number of the opening covering Dividing pixels. The resolution is compared to a threshold, such as 12 nm / pixel threshold to determine if the resolution is adequate.

To the determination of pixel resolution the grid structure is used to the contact openings to locate and determine their dimensions. In one embodiment become vertical and horizontal lines of a grid or mesh structure for localizing the openings used.

18 illustrates the contact position detection process of the module 60d in which a grid or a mesh is placed over the matrix-aligned contact images, the horizontal and vertical axis distances within a particular search area being adjusted so that the contacts each lie in a mesh of the mesh. At this time, the distances can be controlled by increasing or decreasing the number of pixels forming the image. The search area of the mesh line is preferably set to include the area where the same pattern or structure repeats at contact openings.

At the in 18 The contact position recognition method utilizing the mesh search operation is set so that the mesh units or sub-meshes each consist of at least 32 pixels in the horizontal direction and at least 62 pixels in the vertical direction. The search area is determined by the imaginary horizontal axis grid line 150 and the imaginary vertical axis grid line 152 within the areas comprising at least 32 pixels on the horizontal axis and at least 62 pixels on the vertical axis. In this case, the position of the lowest or weakest digitized electron signal value is detected, so that the contacts in the contact images do not overlap the grid line in each case.

at an embodiment the stitch search method is performed so that either the vertical line or the horizontal line is positioned at a first location. The intensity values along the line are summed up to the overall intensity for the line to investigate. The line will gradually move to the next position emotional. For example, the vertical line may be stepped along the horizontal axis to the next Position are moved in which the intensity values along the vertical Line in turn be summed up. In each position, the total intensity value is included a predetermined threshold and the previous total value compared. An increase in intensity can serve as an indication that the Edge of the opening was reached, assuming that the openings have a higher intensity than the background intensity. In other embodiments can the openings a lower intensity own as the background. This process is about the entire grid structure continued to locate the openings respectively and / or to determine their dimensions and shape.

After calculating all the totals in one direction, the process is repeated for the other direction, resulting in a complete characterization of the size, shape and position of the openings. This information can be used for different purposes during subsequent processing. By knowing the positions and shapes of the openings, unnecessary processing of pixels not associated with any opening can be avoided. If at the following If an irregularity or an error is detected, the exact position of the errors or of the defective opening can also be easily determined.

For example, after performing the contact position detection method, the origin of the first grid unit is at the pixel number XO = 13 on the horizontal axis and the pixel number YO = 23 on the vertical axis (see FIG 18 ). Units of the same size can be compared. This is the reason why the contact position detection method is performed in the above-described manner.

The mesh or grid unit can be set differently. As in 19 is shown, a contact opening 153 be arranged in a mesh unit, wherein between two such units, a mesh unit is skipped. As in 20 can also be at least two contact openings 153 be arranged in a mesh unit. The position detecting method of the above-described meshing method may also be used for different image forms when repeating a pattern of images in the unit areas, such as square or oval pad or pad images, in contrast to the above-mentioned round contact images.

18 Fig. 12 shows an SEM image of a contact unit (480 x 480 pixels) for setting the grid or mesh for performing a contact position detection method according to the present invention. As described above, in this example, the grating unit of the present test is set at 32 × 62 pixels, while the pixel number (X0, XY) at the origin is (13, 23). The above-described determination of the lattice unit is by the movement of the horizontal grid line 150 and the vertical grid line 152 determined within the search range, which are determined in the manner described above by the respective grid line and are defined according to the respective spacing of the grid unit with about 60 pixels or about 30 pixels. In another embodiment, the lines are each analyzed to identify the lowest intensity value corresponding to the respective grid line and to determine the position of the openings.

The 19 and 20 show different types of grating fixings of the type described above. 21 shows how the grid unit is occupied with pixel units to explain the generation of contact intensity profiles. 22 Figure 11 shows a first intensity profile in a grid unit, wherein the intensity values with respect to the vertical axis prior to the normalization of the profile are represented by elimination of the background intensity values. 23 shows the intensity profile according to 22 after elimination of the background values with a specified contact threshold. In the present test, the electronic threshold signal is set at 5, while the pixel count at the threshold is determined at 23 20 is. The checking contact opening spans the pixels 20 to 40 along the vertical axis.

24 shows the intensity profiles of the SEM images of the contacts according to 18 after normalization by elimination of the background value. 25 shows a table with coded results of the contact fault checking method according to the invention for the contacts according to 18 , The circle contacts in 24 agree with the place in 25 is identified by the code 4 as a contact or contacting error.

26 shows a table with a part of the contact error check results for the in 15 in the present test with respect to the sampling points of each chip, the number of all contacts being indicated which correspond to the respective classification criterion for the contacts according to the invention. At each of the five locations of the chips or test areas, the number of classifications for each type for the opening found at the location is listed. For example, at the position (1,3), 87 contact holes were classified as belonging to the type D, while three contact holes were classified as the type E. Five contact holes were classified as type G, while three of the contact holes were classified as belonging to type H. It should be noted that 98 openings were checked and classified at each location, corresponding to a total of 17510 openings on the 35 areas checked. In one embodiment, this test can be completed within one hour due to the inventive saving in processing time. The method is therefore suitable for mass production.

The contact profile calculation module 60e (1) is used to generate a first intensity profile of the above-sensed electronic signal values for the individual grating units of the gratings specified above. The background value elimination or correction module 60e (2) is used to generate a second inte sitätsprofils from the first intensity profile by a performed for each grid unit subtraction of the background value of the first intensity profile used.

The first intensity profile and the second intensity profile are calculated using electronic signal values, corresponding to the pixels contained in the grid units be digitized. The electron signal values obtained from the lattice units however, both include the electronic signal value of the corresponding one Kontaktes as well as an electronic signal value of the the Contact surrounding outside area is produced. To get accurate electronic signal values, the only intensity values from the inside of the contact in a grid unit, becomes the electronic signal background value from outside the contact area, d. H. the area surrounding the contact, according to the invention of the intensity profile subtracted to produce the second normalized intensity profile. This is called elimination of the "iscolor effect".

In an embodiment according to the invention, the calculation of the contact intensity profiles and the elimination of the background values are performed by the module 60e according to the following equation 1:

X

die Summe der Elektronensignalwerte oberhalb eines vorbestimmten Schwellenwertes innerhalb einer Maschen- oder Gittereinheit;

B

die Summe der Elektronensignale unterhalb eines vorbestimmten Schwellenwertes innerhalb einer Maschen- oder Gittereinheit;

B_c

die Anzahl der Elektronensignale, deren Wert innerhalb einer Maschen- oder Gittereinheit unterhalb eines vorbestimmten Schwellenwertes liegt;

X

die Anzahl der Elektronensignale, deren Wert innerhalb einer Maschen- oder Gittereinheit oberhalb eines vorbestimmten Wertes liegt;

Y

der Elektronensignalwert mit der Hintergrundeliminierung oder Hintergrundkorrektur in einer Maschen- oder Gittereinheit.

Where X is:

X

the sum of the electron signal values above a predetermined threshold within a mesh or lattice unit;

B

the sum of the electron signals below a predetermined threshold within a mesh or lattice unit;

B _c

the number of electron signals whose value within a mesh or lattice unit is below a predetermined threshold value;

X

the number of electron signals whose value within a mesh or grating unit is above a predetermined value;

Y

the electron signal value with background elimination or background correction in a mesh or grid unit.

In Equation 1 can be the predetermined value given above by elimination of the background value for obtaining accurate measurement results. In one embodiment is For example, the value is 100, but you are not at this value limited.

Of the Y value, d. H. the result of equation 1, is the sum of the relative to the Background value corrected electron signals within the grating or mesh units. In one embodiment, for the value Y in equation 1 defines an upper and a lower limit. If the value of Y for one certain contact is below the lower limit, the Contact considered defective. In one embodiment, a value is used below the predetermined lower limit as an indication of an unopened faulty one Contact opening.

The Equation 1 is typically used to check contact openings with irregular shapes used. For example, equation 1 may precede the formation of openings to check the Photoresist layer used to form the openings is used.

PNm = Grundlinie [(PNk )'] (4) In another embodiment, the calculation of the contact profiles and the elimination of the background value are carried out according to the following equations 2 to 4:

P N m = Baseline [(P N k ) '] (4)

n

die Anzahl der Pixel auf der horizontalen Achse;,

k

die Anzahl der Pixel auf der vertikalen Achse;

P_nk

der digitalisierte Signalwert eines Pixels an der Stelle n auf der horizontalen Achse und an der Stelle k auf der vertikalen Achse;

N

die Nummer der überprüften Gitter- oder Mascheneinheit;

h_i ^N

die anfängliche Pixelanzahl auf der horizontalen Achse innerhalb einer Gitter- oder Mascheneinheit;

h_f ^N

die Pixelendanzahl auf der horizontalen Achse innerhalb einer Maschen- oder Gittereinheit sind.

Here is:

n

the number of pixels on the horizontal axis ;,

k

the number of pixels on the vertical axis;

P _nk

the digitized signal value of a pixel at the location n on the horizontal axis and at the location k on the vertical axis;

N

the number of the checked grid or mesh unit;

h _i ^N

the initial number of pixels on the horizontal axis within a grid or mesh unit;

h _f ^N

the pixel end number on the horizontal axis are within a mesh or grid unit.

The equations 2 to 4 are based on the 21 and 22 described. The 21 shows a graphical representation pixel units in a mesh or grid unit of a contact, which are used for calculating the contact intensity profile according to the invention. 22 shows a schematic representation of the first intensity profile within the mesh or lattice unit, as it was before the subtraction of the background value for a contact image according to 18 was calculated. In one embodiment, the intensity profile is generated by stepwise variation along an axis and by summing the intensity values along the axis perpendicular thereto at the individual positions. The accumulated pixel intensities are displayed as a function of the pixel number along the other axis. The profile according to 22 For example, it is determined so as to move stepwise over the pixel positions along the vertical axis and to summate the pixel intensities in the horizontal direction. In the profile according to 22 The result is an intensity profile with a peak near the center indicating the presence of a contact opening in the particular grid or mesh unit. The aperture extends approximately from the pixel position 16 to the pixel position 44 so that the aperture extends in the vertical direction over approximately 28 pixels. The profile has a slight indentation in the center of the opening at the top of the spike, indicating a decrease in the detected intensity at the bottom of the opening. This form of intensity profile is indicative of a normal trained contact opening.

The value (P _k ^N ) 'from Equation 3 is the average electron signal value per pixel of the pixel k on the vertical axis. It is obtained by taking the total sum (the total level of the digitized electron signal values) which corresponds to the respective pixel with the pixel line k on the vertical axis ( 21 ; k = 20), ie the height of the curve in 22 , divided by the number of horizontal pixels at the vertical axis position k given by h _f ^N - h _i ^N. 22 shows the also equation 3 resulting profile. P _m ^N is a minimum value of (P _k ^N ) ', ie it corresponds to the intensity background or the value of the baseline. P _k ^N therefore corresponds to a mean electron ^signal value per pixel with subtraction of the background value.

23 shows a second intensity profile after the subtraction of the background value or baseline P _m ^N according to Equation 2. 24 shows the normalized second intensity profile for contacts of the contact image according to 18 ,

In one embodiment, the results of equations 2 to 4 are also analyzed by the contacting error checking method according to the invention in order to qualify the contacts as defective and, if appropriate, to be able to identify the types of errors. The second intensity profile ( 24 ) of the openings is analyzed in each case with regard to the identification and classification of irregularities or errors.

As in 23 For example, in one embodiment, a threshold, for example, 5, is used for the second intensity profile (the average electron signal value per pixel after the subtraction of the background value). The critical dimension CD ^{N of} the contact is defined as the length (or width) of the peak in the profile at the threshold. As in 23 with a threshold value of 5 as the critical dimension for the contact CD ^N = 40-20 = 20 pixels. The critical dimension CD ^N , which may, for example, be the diameter of the contact opening, may be calculated according to Equation 5 below:

v_i ^N

die anfängliche Pixelanzahl auf der vertikalen Achse innerhalb der Gitter- oder Mascheneinheit;

V_f ^N

die Pixelendanzahl auf der vertikalen Achse innerhalb der Gitter- oder Mascheneinheit;

Pk N = (Pk N)' – Pm N; und

W_k ^N

gibt an, ob die Pixelintensität oberhalb des Schwellenwertes liegt, d. h. es gilt:

W_k ^N

= 1, falls P_k ^N größer oder gleich dem Schwellenwert ist 0, falls P_k ^N kleiner ist als der Schwellenwert

Here is:

v _i ^N

the initial number of pixels on the vertical axis within the grid or mesh unit;

V _f ^N

the pixel end number on the vertical axis within the grid or mesh unit;

P k N = (P k N ) '- P m N ; and

W _k ^N

Indicates whether the pixel intensity is above the threshold, ie:

W _k ^N

= 1 if P _k ^{N is} greater than or equal to the threshold 0 if P _k ^{N is} less than the threshold

Now, for the pixels above the threshold according to the following equation 6, a mean pixel intensity BSE ^{N is} calculated.

It should be noted that Equations 5 and 6 represent an alternative verification method to Equation 1. The average pixel intensity BSE ^{N in} accordance with Equation 6 corresponds to the value Y calculated in Equation 1. The value CD ^N in Equations 5 and 6 is replaced in Equation 1 by the value X _c .

After calculating the pixel count for CD ^N and the average pixel intensity BSE ^N for the contact under test, these values are used to classify the contact state. In one embodiment, the pixel count CD ^{N is set to} an upper limit NOC2 and a lower limit NOC1. This limit is used to set a range of acceptable pixel counts for normal contacts. It is also possible to set a limit for the average pixel intensity BSE ^N. An upper limit NOT2 and a lower limit NOT1 may be used to determine a range of average pixel values to accept for normal contact.

The values CD ^N and BSE ^{N of} the contacts checked are compared to classify the contacts with their corresponding ranges. In one embodiment, the contacts are classified so as to be divided into one of nine possible classes corresponding to each type according to the comparison of the values for CD ^N and BSE ^N with their respective ranges. An example of the nine possible conditions and their corresponding type classification as well as the corresponding numerical codes are summarized in the following Table 1: classification (BSE ^N ) ≤ (NOT1) (NOT1) ≤ (BSE ^N ) ≤ (NOT2) (BSE ^N ) ≤ (NOT2) (CD ^N) ≤ (NOC1) A type (Code1) B type (Code2) C type (Code3) (NOC1) ≤ (CD ^N) ≤ (NOC2) D type (Code4) E type (Code5) F type (Code6) (CD ^N) ≤ (NOC2) G type (Code7) H type (Code8) 1 type (Code9) Table 1

The Three columns in Table 1 determine the three depth conditions the contact openings. They are arranged in order of decreasing depth. The first Column determines three states, namely the types A, D and G, of relatively deep contact openings. The second column comprises three states for normal depth contact openings, namely the types B, E and H. The third column determines three states of insufficient contact hole depth, namely types C, F, and E. These types of contact openings are typically partially open contact openings or non-uniform contact openings certainly. The rows of Table 1 are after the increasing contact hole diameter arranged. The first line includes the contact hole types A, B and C with insufficient small diameter.

By the second category with the types D, E and F are contact openings determined with a normal diameter. By the third category with the types G, H and I are contact holes with an excessively large diameter certainly.

As shown in Table 1, a classification of Type E is made when both CD ^N and BSE ^{N are} within their respective predetermined ranges, indicating normal contact. The other types, where either or both of these values are out of range, are placed in the remaining type classes, which serve as an indication of differential error propagation and different types of contact errors.

The result screen or result display 60g is used to display the results of the classification of normal contacts and / or contact errors made by the contacting error checking module 60f were classified. The results can be displayed as digitized values for the position of the individual contacts.

25 FIG. 12 shows a table with examples of the classification and the pixel position of the contact openings according to FIG 24 , Whether a contact is normal or faulty is shown in accordance with the second intensity profile as a numerical code for the position of the individual contacts. The code "5" represents the type E and corresponds to a normal contact, while the code "4" represents a type D contacting error. In one embodiment, type "D" represents a non-open contact. The X values in 25 represent the initial pixel count for each mesh or grid unit on the horizontal axis, while the Y values represent the initial pixel count on the vertical axis. 26 contains a table of the results of a verification according to the invention at five locations on each of the seven areas of the semiconductor wafer. The table shows the number of contacts for each classification type at each location.

The values for CD ^N and BSE ^N can be used to classify the contacts in different ways. The type classification for a particular contact can thus serve to specify a particular type of contact failure. If the BSE ^N value for a contact is below a minimum value NOT1, which typically indicates an unopened contact opening, the opening is classified into one of the A, D or G type classes. If the BSE ^N value is greater than a maximum value NOT2, the opening is open, but for some reason it is not yet acceptable. For example, the opening may be unevenly shaped to have, for example, a broadening or narrowing towards its bottom or bottom. In this case, the opening is classified into one of the type classes C, F or I.

If the CD ^N value is below the minimum value NOC1, the detected error may indicate an opening that is too narrow or has an irregular shape, such as an oval. When the CD value ^N is above the maximum value NOC2, this is an indication of an irregularly shaped opening.

The 27 contains a flow chart for illustrating the logical sequence of the processing steps in a semiconductor device manufacturing method according to the invention. At a certain processing step in the fabrication of semiconductor devices, first, a photoresist pattern corresponding to the contact openings is formed after the photoresist has been applied to certain insulating layers such as a nitride film or an oxide film. Now, a photolithographic process is used (S40). The photoresist pattern is generated by exposure and development steps.

Now the photoresist pattern becomes an etching mask used and lying under the photoresist structure insulating Layer is etched to form the contact holes (S42). Subsequently, will the inside of the contact openings cleaned and the wafers are placed in the in-line SEM device, where, in the manner described above, the verification method according to the invention is performed on faulty contact openings. Now, the inside of the contact holes with a conductive Material provided and it will be the subsequent steps for manufacturing semiconductor devices (S46).

• N = Anzahl der Pixel des SEM-Bildes in Richtung der Y-Achse;
• M = Anzahl der Pixel des SEM-Bildes in Richtung der X-Achse;
• VP (Vertical Pitch = senkrechter Abstand) = Y-Achsen-Kontaktabstand in Richtung der Y-Achse des Gitters oder des Maschennetzes;
• HP (Horizontal Pitch = horizontaler Abstand) = X-Achsen-Kontaktabstand in Richtung der X-Achse des Gitters oder des Maschennetzes;
• MX = X-Achsen-Pixelbereich für das Gitterverfahren;
• MY = Y-Achsen-Pixelbereich für das Gitterverfahren;
• bse = Grundschwellenwert für ein charakteristisches Kontaktprofil in einer Gitter- oder Mascheneinheit;
• NO1 = unterer Grenzwert für die charakteristische Profilintensität normaler Kontakte;
• NO2 = oberer Grenzwert für die charakterisitische Profilintensität normaler Kontakte;
• CD1 = unterer Grenzwert für die Pixelanzahl eines charakteristischen Profils normaler Kontakte;
• CD2 = oberer Grenzwert für die Pixelanzahl eines charakteristischen Profils normaler Kontakte;
• XN = überprüfte SEM-Gesamtbildanzahl (Endzahl oder Ergebnis) pro Chip oder SHOT-Einheit eines Wafers;
• YN = überprüfte SEM-Gesamtbildanzahl (Endzahl oder Ergebnis) in einem Chip oder einer SHOT-Einheit;
• X = überprüfte SEM-Bildordnung oder Reihenfolge pro Chip oder SHOT-Einheit eines Wafers;
• Y = überprüfte SEM-Bildordnung oder Reihenfolge in einem Chip oder einer SHOT-Einheit;
• cdata [j][i] = SEM-Bildsignalpegel bei (pro) Pixeleinheit.

28 shows a schematic representation of a flowchart with the logical sequence in an embodiment of the contact verification method according to the invention. In the process step 500 the used parameters are read in. In one embodiment, the method uses the following parameters:

• N = number of pixels of the SEM image in the Y-axis direction;
• M = number of pixels of the SEM image in the X-axis direction;
• VP (Vertical Pitch) = Y-axis contact distance in the direction of the Y-axis of the grid or mesh;
• HP (Horizontal Pitch) = X-axis contact distance in the direction of the X-axis of the grid or mesh;
• MX = X-axis pixel area for the grid method;
• MY = Y-axis pixel area for the grid process;
• bse = basic threshold for a characteristic contact profile in a grid or mesh unit;
• NO1 = lower limit for the characteristic profile intensity of normal contacts;
• NO2 = upper limit for the characteristic profile intensity of normal contacts;
• CD1 = lower limit for the number of pixels of a characteristic profile of normal contacts;
• CD2 = upper limit for the number of pixels of a characteristic profile of normal contacts;
• XN = checked SEM total number of pictures (final number or result) per chip or SHOT unit of a wafer;
• YN = checked SEM total number of pictures (final number or result) in a chip or a SHOT unit;
• X = checked SEM image order or order per chip or SHOT unit of a wafer;
• Y = checked SEM order or order in a chip or a SHOT unit;
• cdata [j] [i] = SEM image signal level at (per) pixel unit.

Now in the step 502 the x-axis value is initialized to 0 while in the step 504 the Y-axis value is initialized with 0. The verification process will now take place along the Y axis in one of the steps 506 to 520 formed inner loop continues until the maximum Y-axis value is reached. Now the X-axis value is incremented and the inner loop is repeated again for all Y-axis values. Finally, the outer loop ends when the final values for the X and Y axis values are reached. In the inner loop according to 28 The SEM image data at the location (X, Y) and cdata [j] [i] in the step 506 read that in 29 is shown in detail. It should be noted that the mesh or gantry method described herein uses a rectangular grid structure with orthogonal X and Y axes. It should be noted that a rectangular grid is not required. Other meshes or meshes may be used. For example, a triangular or a trapezoidal grid may be used. The grid or mesh structure is selected to detect any periodically repeating structure of contacts.

In the step 508 in 28 Now the position of the contact opening is determined. The step 508 is in the 30A - 30D shown in detail. The detection of the contact opening position involves selecting the type and structure of the grid or mesh used to check the contact openings. In the in the 30A - 30D In detail, the method is moved pixel-by-pixel along a first selected direction (in the horizontal direction) and sums up all pixel intensity values in a second orthogonal direction (the vertical direction). By detecting a noticeable change (jump) in intensity, the edge of an aperture is fixed. The process is continued until detection of a significant drop in the intensity by which the opposite ing edge of the opening is fixed. This procedure is used until all openings are located. It should be noted that in the steps 550 and 582 in the 30A - 30D the absolute value of the intensity differences is used. This is because the amount of intensity differences or contrast is important for determining the pads. In this method, different conventions are used to define the apertures by high or low intensity.

At the step 514 in 28 the contact opening profiles are calculated. This process is in the flowcharts in the 31A - 31D shown in detail. The profiles are calculated by analysis of the individual contact openings, which, in conjunction with the 30A - 30D described methods have been identified according to the invention. For each opening a profile is calculated. In one embodiment, the profile is generated by summing the intensity values in one direction at each location along the other orthogonal thereto. The intensity values at the individual points are averaged and displayed to generate the profile. It should be noted that in the flowcharts of 31A - 31D general variables F and F2 are used. These variables, in a preferred embodiment of the present invention, are interchangeable with the variables BSE ^N and CD ^N , respectively, defined in equations 5 and 6 above.

In the step 516 in 28 the contact openings are checked according to the invention. This process is in the 32A - 32B shown in detail. As already described above, according to the 31A - 31D analyzed certain values to classify the individual contact openings in one of nine possible contact classes. As related to the 32A - 32B already mentioned, the variables F and F2 are interchangeable with the variables BSE ^N and CD ^N.

In the step 518 in 28 the Y-axis value is incremented and in the step 520 it is determined if the maximum Y-axis value has already been reached. If this is not the case, the flowchart returns to the beginning of the inner loop. Otherwise, the x-axis value in the step becomes 522 is incremented and it is done in the flow chart over the block 524 a return to the beginning of the outer loop in the step 504 in which the Y-axis value is initialized. After completing the outer loop, the results of the review in step 526 being represented.

As already mentioned above, shows 29 a schematic diagram of a flow chart for the detailed presentation of the step 506 in 28 for reading the SEM image data. In step 528 index j is initialized to 0 while in step 530 the index i is initialized with 0. In the step 532 the data cdata [j] [i] is read while in the step 534 the index i is incremented. In the step 536 it is determined whether the index i has already reached its maximum value M. If this is not the case, the flowchart returns to the step 532 in which the data is read in again. Otherwise, in the step 538 the index j increments and in the step 540 determines whether j has already reached its maximum value M. In this case, the process ends. Otherwise, the flowchart returns to the step 530 in which the index i is again initialized with 0. The process is then repeated.

at an embodiment the check according to the invention is carried out faulty contacts after the formation of the contact openings and the cleaning of the inside of the contact openings (ACI = After Cleaning Inspection = check after execution the cleaning step). However, the error checking can also be done on the exposed insulation layer over the wafers in the development process, wherein the photoresist structure to form the contact openings is formed (ADI = After Development Inspection) execution development).

The The present invention is not limited to the type described and Way for contact openings applicable, but also for Vias at each step for direct connection of all contact openings and for contacting with the semiconductor substrate and the conductive Layers. The present invention is also for verification of Patterning errors after the development process during the photo processing step for the formation of contact openings suitable.

In addition to this the present invention is not only suitable for checking round contact openings, but also to review Patterns or structures by capturing the different species of structural images that repeat regularly.

Contact or contacting errors are inventively accurately detected by digitized values, without the contact images are checked with the naked eye or a microscope. For contacts with a larger aspect ratio, contact errors can be easily and very accurately confirmed. In addition, the Checking for defective contacts for the entire wafer surface within a short period of time, in which the results required for the detection of contact errors are generated. The process has high efficiency and enables high productivity in mass production.

Claims (30)

Method for checking at least a part of a semiconductor wafer with the following method steps: - reading image data of a scanning electron microscope ( 100 ) for the part of the semiconductor wafer; - Identification of image data for a feature of the semiconductor wafer from the data for the part of the semiconductor wafer; - calculating a parameter associated with the feature from the image data of the feature; Comparison of the parameter with a range of parameter values to be accepted; - Classification of the characteristic according to the comparison with the parameter values to be accepted; characterized by the following features: - the image of the scanning electron microscope ( 100 ) for the inspecting part of the semiconductor wafer, superimposing a grid of grid axes defining a plurality of meshes, each mesh having a predetermined number of pixels, such that each mesh or selected mesh comprises a predetermined number of features to be analyzed simultaneously; wherein, by adjusting the pitches of the grating axes, the number of pixels per stitch is set or changed depending on the size of the feature to be inspected such that the feature is covered by a predetermined number of pixels and irregularities or defects of the feature in the image can be reliably recognized are. The method of claim 1, wherein the feature is a contact opening in an integrated circuit. The method of claim 2, wherein the contact opening as not classified open if the parameter is outside range of acceptable parameter values. The method of claim 1, wherein the feature is as is classified incorrectly if the parameter is outside range of acceptable parameter values. The method of claim 1, wherein the feature is as is classified acceptable if the parameter within the Range of acceptable parameter values. The method of claim 1, wherein the SEM image data by secondary electrons and backscattered electrons be generated. The method of claim 1, wherein the parameter is a Dimension of the feature includes. The method of claim 1, wherein the parameter is a Number of SEM image data pixels comprises which are associated with the feature. The method of claim 1, wherein the parameter is a medium intensity the pixel associated with the feature. The method of claim 1, wherein in an additional Process step calculates an image pixel intensity profile of the feature becomes. The method of claim 10, wherein the calculation of the image pixel intensity profile the subtraction of a background intensity value from the intensity values which comprises pixels in a region encompassing the feature. The method of claim 1 with the following additional Steps: - Calculation a second parameter associated with the feature from the image data the feature; - Comparison of the second parameter with a range of acceptable values for the second parameter; - Classification of the characteristic according to the comparison with the range of to be accepted Values for the second parameter. The method of claim 12, wherein the second parameter a dimension of the feature. The method of claim 12, wherein the second parameter comprises a number of SEM image data pixels associated with the feature are. The method of claim 12, wherein the second parameter a medium intensity the pixel associated with the feature. The method of claim 12, wherein the feature is only then classified as acceptable if the first parameter within the range of acceptable values for the first Parameter is and if the second parameter within the range on acceptable values for the second parameter lies. The method of claim 1 with an additional Characterization of the feature using a coordinate system, wherein the characterization comprises the following steps: - Overlay an image of the part of the semiconductor wafer with the coordinate system; and - Analysis the intensity values arranged along a second axis of the coordinate system Pixels at several points along a first axis of the coordinate system. The method of claim 17, wherein the analysis is the Sum of the intensity values the pixel arranged along the second axis. The method of claim 18, wherein the analysis for Acquisition of the feature also the detection of a change the accumulated intensity values in several places along the first axis. The method of claim 17, wherein the analysis is a Averaging the ordered intensity values the pixel arranged along the second axis. The method of claim 20, wherein the analysis for Acquisition of the feature also the detection of a change the averaged intensity values in several places along the first axis. The method of claim 17, wherein the characterization includes detecting a dimension of the feature. The method of claim 17, wherein the characterization determining the position of the feature. The method of claim 17, wherein the characterization the identification of a pattern or a structure of several Features includes. The method of claim 24, wherein the structure is a is periodic structure. The method of claim 17, wherein the coordinate system is a rectangular coordinate system. Method according to claim 17, wherein the coordinate system is a triangular coordinate system. The method of claim 17, wherein the coordinate system a trapezoidal Coordinate system is. The method of claim 1, wherein the SEM image data in the form of digitized gray scale pixel values. The method of claim 1, wherein the SEM image data in the form of digitized color-coded pixel values.