JP2009014346A - Wafer surface inspection method and wafer surface inspection device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enhance a yield by measuring a surface roughness of a wafer with a pattern having fine irregular patterns in a midway of wafer working, and by preventing the yield from getting worse caused by dispersions of file thicknesses and film qualities among the wafers, in a wafer surface inspection method and a wafer surface inspection device. <P>SOLUTION: The wafer surface inspection device has a method and a means for searching an area capable of measuring the surface roughness existing on the wafer with the pattern, and for measuring the measurable area. The wafer surface inspection method/device measures the surface roughness, using a TEG (Test Element Group) existing in a scribe area on the wafer, in the measurable area. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a surface inspection method and a surface inspection apparatus for a semiconductor wafer used for manufacturing a semiconductor device.

The latest semiconductor devices in recent years have been pursuing miniaturization to the limit, and not only the line width has been miniaturized but also the thinning of the generated film has been accelerated. A thin film of 1 to 2 nanometers is required for the gate oxide film. For this reason, the unevenness of the film surface and the crystal state of the generated film, which have not been a problem with conventional devices, have greatly affected the characteristics of the latest devices. Furthermore, the form of the wafer processing apparatus has also changed. For example, a vertical oxide diffusion apparatus is conventionally used in the gate oxide film forming process, and a form in which batches of dozens to hundreds of wafers are batch processed is used. Although it is general, single wafer processing is spreading rapidly recently, and variations in film thickness and film quality between wafers are a major factor in yield reduction. Therefore, it is necessary to set appropriate conditions for the manufacturing process in consideration of minute variations in the generated film and line width, which can be ignored in the state-of-the-art device process, and strictly control the set manufacturing process conditions.
Therefore, wafer material evaluation and surface roughness management have become one of the important factors for improving the yield, and the need for measuring the surface roughness of the wafer has increased.

  Conventional techniques for measuring the surface roughness of a wafer include (1) a method of observing the cut surface with an electron microscope by cutting or FIB (Focused Ion Beam) processing, and (2) AFM (Atomic Force). A method for measuring the surface roughness of a wafer with a microscope (atomic force microscope) is known. However, these methods (1) have a problem that it takes a long time to prepare for observing the cross section, and (2) has a problem that the throughput is very low.

  As a technique for measuring the surface roughness of a wafer that solves such a problem, a laser beam as shown in Patent Document 1 is focused, the laser beam is incident on the surface of the substrate, and the scattered light from the substrate surface is two-dimensionally reflected. It is known to receive light on an array light sensor, calculate a scattered light intensity distribution from the received light signal, and convert the scattered light intensity distribution into a numerical value of surface roughness.

Further, as prior arts related to the present invention using such a laser beam or the like, there are those shown in Patent Documents 2 and 3.
Patent Document 2 discloses that a single laser beam emitted from a laser is subjected to deflection surface adjustment and is incident on an angle of incidence that generates only scattered light while minimizing reflected light in order to collect light with a condenser lens. Then, the sample surface is irradiated. The generated scattered light from the sample surface is sequentially converted into an electrical signal from an imaging lens and a CCD (Charge Coupled Device) image sensor. And the method of memorize | storing and analyzing an image and obtaining the digitized data which shows the surface roughness of a sample is disclosed.

  In Patent Document 3, beam X-rays parallel to the incident direction are incident on a wafer at a low angle less than a critical angle, and the intensity of fluorescent X-rays emitted from the wafer top surface or scattered X-rays scattered at the wafer top surface is disclosed. Is repeated for a plurality of wafers with known surface roughness. From this operation, the correlation between the X-ray intensity and the surface roughness of the wafer is obtained, and the X-ray intensity is measured for the wafer to be measured whose surface roughness is unknown. Then, based on the measured value, the surface roughness of the wafer to be measured is obtained from the correlation, and the surface roughness of the wafer including a semiconductor wafer having a multilayer structure such as an SOI (Silicon On Insulator) wafer is nondestructively and noncontacted. The usage of measuring with is disclosed.

JP 2006-64496 A JP 2002-340537 A Japanese Patent Laid-Open No. 7-19844

  In the above prior art, it is impossible to measure the surface roughness of a patterned wafer having a fine uneven pattern during wafer processing. That is, when the illumination light (laser light) is incident on the patterned wafer, the scattered light generated from the fine uneven pattern becomes noise, so that only the surface roughness of the unpatterned wafer can be measured.

  Further, in the cross-sectional observation method in which the wafer is cut or processed by FIB (focused ion beam) and the cut surface is observed with an electron microscope, it takes a long time to prepare for observing the cross section. Furthermore, there is a problem that a wafer that has been subjected to cutting or FIB processing does not become a product.

In observation with an AFM (atomic force microscope), the surface roughness of the patterned wafer can be measured with high accuracy, but there is a problem that the throughput is very low and the measurement takes too much time.
Therefore, an object of the present invention is to automatically measure the surface roughness of a patterned wafer at high speed and with high accuracy.
It is also an object to enable early detection of process abnormality at the process stage of semiconductor devices and improve yield by using measurement data such as wafer surface roughness and grain size (granularity) measured using the present invention. And

In order to solve the above problems and measure the surface roughness of a patterned wafer, the wafer surface roughness measuring method of the present invention searches an area where the surface roughness existing on the patterned wafer can be measured. And means for measuring the surface roughness of the patterned wafer.
The problem can also be solved by using a TEG (Test Element Group) present in the scribe area on the wafer as the area where the surface roughness of the wafer can be measured.

According to the present invention, there is an effect that the surface roughness of a patterned wafer can be automatically measured at high speed and with high accuracy.
Further, by using measurement data such as wafer surface roughness and grain size (granularity) measured according to the present invention, SPC management (Statistical Process Control) can be performed at the process stage of a semiconductor device. Early detection of process anomalies is possible, yields can be greatly improved by quickly analyzing the anomalies and feeding them back to the process.

  An example in which the present invention is applied to an optical wafer surface inspection apparatus for inspecting foreign matters and defects on the wafer surface will be described below. Here, in order to measure the surface roughness of the patterned wafer, a flat region capable of measuring the surface roughness existing in the patterned wafer is used.

[First Embodiment]
An embodiment of the present invention will be described below with reference to FIGS.
An example of a system according to the present invention is shown in FIG. The system includes an optical wafer surface inspection apparatus 110, a data server 120, and a design information server 130, and each apparatus is connected via a network 140.

  Here, the optical wafer surface inspection apparatus 110 has means for irradiating the wafer surface with light and acquiring the intensity of light scattered on the wafer (scattered light intensity) and its position (coordinates). Details will be described later.

The data server 120 includes measurement data obtained from the optical wafer surface inspection apparatus 110, measurement data of a review SEM (Scanning Electron Microscope) and CD-SEM (Critical Dimension-Scanning Electron Microscope), test data of an electrical test apparatus, This is a computer capable of storing measurement data obtained by measuring a wafer using a conventional wafer inspection apparatus, such as analysis data of a foreign component analysis apparatus.
In addition, the data server 120 has a calibration curve database 121, and the calibration curve database 121 stores a calibration curve created from measurement data stored in the data server 120.

  The design information database 130 is a database for storing wafer design information. For example, layout data (floor plan data) of functional blocks constituting a semiconductor device and layers created as a result of the layout (element activation) Information such as the mask data for forming the layer, wiring layer, etc., the shape of the pattern of the semiconductor chip, the position of the semiconductor chip on the wafer, the position of the scribe area, the position of the TEG, etc. This is a database responsible for linkage with measurement data measured using a conventional wafer inspection apparatus stored in the data server 120.

  The network 140 is a communication network based on Ethernet (registered trademark), for example.

  FIG. 2 shows the concept of the optical wafer surface inspection apparatus. Optical wafer surface inspection apparatus 110 includes sample inspection table 210, illumination light source 220, scattered light detection unit 230, signal synthesis unit 240, overall control unit 250, stage control unit 260, information display unit 270, input operation unit 280, and storage. Part 290 and communication part 295.

The sample inspection table 210 includes a sample stage 211 on which a sample such as the wafer 200 is placed, a rotation drive unit 213 that rotates the sample stage 211 around the rotation shaft 212, and a slide drive unit 214 that moves the sample stage 211 in the radial direction. I have.
Here, the rotation drive unit 213 and the slide drive unit 214 are controlled by a later-described stage control unit 260 that receives a command signal from the later-described overall control unit 250.

  The illumination light source 220 is installed so that the irradiated light (illumination light 221) irradiates a certain point (spot) on the sample stage 211. Therefore, under the control of the stage control unit 260, the slide drive unit 214 moves in the radial direction while the rotation drive unit 213 of the sample inspection table 210 rotates the rotation shaft 212, so that every place on the sample stage 211 is spotted. The illumination light 221 can be irradiated to a specific position of the wafer 200 on the sample stage 211.

  Then, the stage controller 260 converts the specific position of the wafer 200 irradiated with the illumination light 221 into XY coordinates by the rotation angle of the rotation drive unit 213 and the radial movement distance of the slide drive unit 214. it can. The acquired XY coordinate data is stored in the storage unit 290 via the overall control unit 250.

  Here, in order to make the area where the illumination light 221 hits as small as possible, light having a high degree of convergence of light such as laser light is preferable.

The scattered light detection unit 230 includes detectors 231a to 231d that detect light. FIG. 2 shows a total of four detectors, detectors 231a and 231d arranged at low angle positions and detectors 231b and 231c arranged at high angle positions, but the number of detectors is not limited. Two or more detectors should just be arrange | positioned so that the detectors 231a-231d may differ in at least one of the azimuth angle and elevation angle of each spot and a detector. Each detector 231a-231d is irradiated with illumination light (laser light) 221 from the illumination light source 220 onto the surface of the wafer 200, and detects scattered light generated at the spot. The detection signals output from the detectors 231a to 231d include a foreign matter or defect signal (defect signal) and a surface roughness signal (Haze signal; haze signal).
In the scattered light detection unit 230, the detectors 231a to 231d are connected to the amplifiers 232a to 232d, respectively, and then connected to the A / D converters 233a to 233d. Thereby, the detection signals of the detectors 231a to 231d are amplified by the amplifiers 232a to 232d and converted into digital signals by the A / D converters 233a to 233d.

  The signal synthesis unit 240 creates a synthesized signal by synthesizing the detection signals of the detectors 231a to 231d converted into digital signals in accordance with designated operation conditions. The data of the synthesized signal synthesized by the signal synthesis unit 240 and the data of the detection signals of the detectors 231a to 231d that are converted into digital signals based on the synthesized signal are stored in the storage unit 290 via the overall control unit 250. The

The overall controller 250 controls the entire optical wafer surface inspection apparatus. For example, an operation signal from the input operation unit 280 is received, a program stored in the storage unit 290 is executed, processing corresponding to the operation signal is performed, and the stage control unit 260 has a rotation driving unit included in the sample inspection table 210. Command signals for controlling 213 and the slide drive unit 214 are output, and calculation conditions for synthesizing the detection signals of the detectors 231a to 231d converted into digital signals by the signal synthesis unit 240 are changed.
In addition, the overall control unit 250 causes the storage unit 290 to store the combined signal data combined by the signal combining unit 240 and the detection signal data of the detectors 231a to 231d converted into digital signals based on the combined signal. Alternatively, the processing program stored in the storage unit 290 is executed, the data is processed, and displayed on the information display unit 270.

  The input operation unit 280 is for the user to input detection signal synthesis conditions by the signal synthesis unit 240 as described above, and to instruct the operation of each device.

  The storage unit 290 stores programs and constants necessary for various types of control / arithmetic processing, measurement results (synthetic signals and detection signals), synthesis conditions set by the input operation unit 280, and the like. The combined signal data and detection signal data of the detectors 231a to 231d are stored together with the measurement position (coordinates) of the scattered light on the wafer obtained from the stage controller 260.

  The communication unit 295 is connected to the network 140, and the overall control unit 250 transmits and receives data to and from the data server 120 and the design information database 130 via the communication unit 295.

  FIG. 3 is a scattered light intensity distribution diagram when a region having no pattern on the wafer surface is measured using the optical wafer surface inspection apparatus according to the present invention. The wafer is irradiated with illumination light (laser light), and scattered light generated from the wafer surface is acquired by a detector. By performing signal processing on the detected scattered light, a scattered light intensity distribution diagram 141 is created, and the foreign component and the surface roughness component are separated.

For example, as shown in FIG. 3, the surface of the wafer to be measured includes foreign objects 301 a and 301 b, a defect 302, a certain point 310 on the flat surface, and a point other than the point 310. By irradiating the wafer with illumination light (laser light), the scattered light intensity W310 is obtained from the scattered light generated at the point 310 on the flat surface. Similarly, the scattered light intensity W301a can be acquired from the foreign object 301a, the scattered light intensity W311 can be acquired from the point 311 on the flat surface, the scattered light intensity W302 can be acquired from the defect 302, and the scattered light intensity W301b can be acquired from the foreign object 301b. Then, a scattered light intensity distribution diagram 141 can be created from the measured scattered light intensity at points on the surface of the wafer. Here, the foreign substance component and the surface roughness component are separated using a threshold value designated by the input operation unit 280 shown in FIG. For example, by setting the threshold value to L, the scattered light intensity W301a and W301b of the foreign matter and the scattered light intensity W302 of the defect can be used as foreign matter components. Haze which is the scattered light intensity of the surface roughness of the wafer to be measured is obtained by taking an average from all scattered light intensities (scattered light intensity of the surface roughness component) such as W310 and W311 remaining after removing the foreign component. A value can be obtained.
Here, the foreign substance component means the scattered light intensity at which the scattered light intensity of the scattered light generated by the foreign substance, the defect or the like is equal to or greater than the threshold value L.

  Next, an example of a flow of processing for measuring the surface roughness of a wafer with a pattern using an optical wafer surface inspection apparatus and an example of a configuration of information used in the processing will be described with reference to FIGS.

FIG. 4 is a diagram showing a flow of processing for measuring the surface roughness of the patterned wafer.
First, design information of a wafer with a pattern to be measured (hereinafter referred to as an inspection wafer with a pattern) is acquired from the design information database 130 (step S401).
The measurable area is searched and searched from the acquired design information (step S402).

  As a search condition for the measurable area, the optical wafer surface inspection apparatus 110 is used for measurement, so that the scattered light intensity is determined from the scattered light generated by irradiating the laser beam having a beam spot diameter of about 10 micrometers as illumination light. A place that has a sufficiently large area that can be measured (about 50 μm × 50 μm or more) and is flat enough not to generate noise that affects the measurement of scattered light intensity It is a condition that the wafer has.

  FIG. 5 shows an example of a measurable area of a patterned wafer. On the wafer 200, there are 1000 or more chips on the wafer. A certain chip 510 has patterns 520, 521, and 522. It has an area (50 micrometers x 50 micrometers or more) large enough to measure the surface roughness at locations other than the patterns 520, 521, and 522, and affects the measurement of scattered light intensity. A flat region 530, which is a region that satisfies a condition that is flat enough to prevent generation of noise, is a measurable region.

  Here, when the wafer has a plurality of layers and the transparency between the layers is high and thin, noise that affects the measurement of scattered light intensity is likely to occur due to the pattern shape of the lower layer, etc. . In that case, it is necessary to search for a measurable area in consideration of design information of the lower layer.

Next, measurement procedure information is created for a plurality of measurable areas found by the search (step S403). It is possible to automatically create measurement procedure information such as a measurement position, a measurement range, and a focus from position information and layer information of a measurable area from the design information database 130.
Based on the created measurement procedure information, the scattered light intensity of the measurable area of the inspection wafer with a pattern is measured (step S404).

  Next, a scattered light intensity distribution map is created from the scattered light intensity measured in the measurable region of the inspection wafer with a pattern (step S405). Then, the scattered light intensity analysis is performed, the foreign matter component and the surface roughness component are separated using the threshold value, and only the surface roughness component is extracted (step S406). By taking the average from the scattered light intensity of the surface roughness component of all the measurement points measured in the flat area of the inspection wafer with pattern, the scattered light intensity of the surface roughness component of the inspection wafer with pattern to be measured A certain Haze value is acquired (step S407).

  Next, a calibration curve stored in advance in the calibration curve database 121 is acquired (step S408). As will be described later, the calibration curve is created from data obtained by measuring a plurality of wafers. In other words, the calibration curve is a standard value of a wafer. This calibration curve includes surface roughness (arithmetic mean roughness Ra, root mean square roughness Rms, etc.) and grain size based on the Haze value.

  Next, the Haze value of the inspection wafer with a pattern acquired in step S407 is applied to the calibration curve acquired in step S408 (step S409), and the surface roughness of the inspection wafer with pattern (arithmetic average roughness Ra, mean square) (Roughness Rms, etc.) and grain size are acquired (step S410), and the acquired data is converted into management target data and stored in the storage unit 290. (Step S411).

Here, an example of a calibration curve creation procedure and an acquired calibration curve will be described with reference to FIGS.
FIG. 6 is a flowchart showing the procedure for creating a calibration curve.
First, an appropriate wafer (hereinafter referred to as a specimen wafer) is obtained. (Step S601) Since this sample wafer is a data acquisition wafer for creating a calibration curve, it may be made of various film types (eg, Si, Poly-Si, Cu, etc.) and layers. However, since there is a condition to be selected as a sample wafer and measurement is performed using the optical wafer surface inspection apparatus 110, scattering is performed from scattered light generated by irradiating laser light having a beam spot diameter of about 10 micrometers as illumination light. It has an area large enough to measure the light intensity (about 50 micrometers x 50 micrometers or more) and is flat enough not to generate noise that affects the measurement of scattered light intensity. The condition is that the wafer has a certain location.

Next, using the optical wafer surface inspection apparatus 110, the Haze value of the sample wafer is measured from the scattered light intensity at the flat location of the sample wafer (step S602).
Next, the surface roughness, grain size, etc. of the sample wafer are measured using an AFM (atomic force microscope) or the like (step S603).
The measurement in step S603 is performed using a method in which the specimen wafer is processed by FIB (focused ion beam) without using an AFM (atomic force microscope), and the cut surface is observed and measured with an electron microscope or the like. You may measure surface roughness, grain size (particle size), etc.

Next, a calibration curve is created from the Haze value of the sample wafer acquired using the optical wafer surface inspection apparatus 110, the surface roughness of the sample wafer measured with an AFM (Atomic Force Microscope), the grain size, and the like. (Step S604). The created calibration curve is stored in the calibration curve database 121 (step S605). The form of the calibration curve information stored here may be a map, a graph, a table, a function, or the like.
Although a calibration curve is used in the present embodiment, it is only necessary to be able to show the correspondence between information A and information B having some relationship, and it may be a map, a graph, a table, a function, or the like.

FIG. 7 is an example in which a calibration curve is shown in a table. The haze value of the sample wafer measured by the optical wafer surface inspection apparatus 110 and the surface roughness of the sample wafer measured by an AFM (atomic force microscope) ( (Arithmetic mean roughness Ra, root mean square roughness Rms, etc.) and grain size (granularity) of the sample wafer measured by SIM or the like. The particle diameter is clarified by the EBSP method (Electron Back-Scattering Diffraction), and the relationship with the Haze value is used as a calibration curve.
In the example of FIG. 7, the table has calibration curves of different film types (for example, Si, Poly-Si), but the relationship with various other films (for example, Cu, Al, etc.) is calibrated in advance. By creating as a line, the surface roughness of a desired film type can be measured.

FIG. 8 shows the sample wafer Haze value measured by the optical wafer surface inspection apparatus 110 in the table of FIG. 7 and the sample wafer surface roughness (root mean square roughness Rms) measured by an AFM (Atomic Force Microscope) or the like. ) Is indicated by a calibration curve.
By using the graph shown in FIG. 8 and applying the Haze value of the inspection wafer with pattern performed in step S409 of FIG. 4 to the calibration curve, the root mean square roughness Rms indicating the surface roughness of the inspection wafer with pattern can be acquired. . For example, when the Haze value of the inspection wafer with pattern is 40, the calibration curve shows that the root mean square roughness Rms indicating the surface roughness of the inspection wafer with pattern is 0.16.

  Here, the overall control unit 250 can output the management target data stored in the storage unit 290 in step S410 of FIG. 4 to the information display unit 270. When output to the information display unit 270, the management target data can be converted into GUI (Graphical User Interface) display, text display, Haze map, surface roughness map, and the like and displayed. Further, the display can be stored in the storage unit 290.

  Here, the Haze map is a contour display or a three-dimensional conversion display showing the positional relationship between the magnitude of the Haze value and the flat area measured on the inspection wafer with a pattern. Similarly, the surface roughness map shows contour line display and three-dimensional conversion display indicating the positional relationship between the size of the surface roughness acquired from the haze value using the calibration curve and the flat region measured on the inspection wafer with a pattern.

  Also, a plurality of inspection wafers with patterns are prepared, each inspection wafer with a pattern is measured, a haze value and management target data (data converted in step S411) are obtained, and time series are obtained based on the obtained data. By creating a plot or control chart and performing SPC management (Statistical Process Control), it is possible to detect an abnormality in the film forming process. The surface roughness is affected by conditions in which various conditions such as temperature, humidity, and water at the time of manufacturing a semiconductor device are complicatedly integrated. Therefore, by using the present invention, the yield can be greatly improved by detecting a process abnormality at an early stage in the process stage of the semiconductor device, immediately analyzing the abnormality, and feeding back to the process.

[Second Embodiment]
The method for searching the measurable area described in the processing flow for measuring the surface roughness of the patterned wafer shown in FIG. 4 is an example of searching the flat area 203 on the chip from the design information database 130. In recent years, semiconductor devices have become very complicated, and a flat region 203 (measurable region) in which surface roughness can be measured in an optical wafer surface inspection apparatus does not always exist in a chip.
A second embodiment capable of measuring the surface roughness even when no measurable area exists in the chip will be described with reference to FIG. FIG. 9 is a second embodiment in which a scribe area is used instead of the flat area that is the measurement target in the first embodiment.

A large number of semiconductor chips 910 are formed on the wafer 200. A scribe region 920 is provided between the semiconductor chips, and the semiconductor chip 910 is completed by cutting in this region.
Normally, a TEG (Test Element Group; test element group) 930 is designed in the scribe area 920. The TEG 930 is subjected to a test process including a specific element, which is created in order to evaluate a specific element in the development stage of a semiconductor product, which cannot be evaluated after being cut into a semiconductor chip. It is a test area (test pattern) that has been set.

  Due to the miniaturization and high integration of semiconductor products, it is no longer possible to individually evaluate the characteristics and reliability of internal elements in actual products. Therefore, TEG930s that are suitable for evaluation purposes at each design stage are individually designed. By doing so, each element can be evaluated efficiently and with high accuracy, and it can be fed back to establishment of optimum conditions for the semiconductor manufacturing process, process control, and the like. One of the TEGs 930 includes a film pressure measurement test area, and one of the embodiments of the present invention can measure the surface roughness of the wafer in the film pressure measurement test area.

  The film pressure measurement test region TEG930 is the same as the material of the semiconductor device, and the film pressure is substantially equal to that of the semiconductor device. And since it measures using the optical wafer surface test | inspection apparatus 110, it can fully measure scattered light intensity from the scattered light which arises by irradiating the laser beam whose beam spot diameter which is illumination light is about 10 micrometers. The surface of the wafer can be obtained by using a TEG930 that has a large area (about 50 micrometers × 50 micrometers or more) and is flat enough not to generate noise that affects the measurement of scattered light intensity. Roughness can be measured.

  In the present embodiment, a film pressure measurement test area which is one of the TEGs is shown as a preferred example. However, as long as the area satisfies the above conditions, measurement can be performed in other TEG and scribe areas.

  By using this embodiment, even if there is no flat region having an area large enough to measure the surface roughness in the chip on the wafer, the scribe region on the wafer is measured to measure the wafer surface. It is possible to measure the surface roughness.

  The position of the test area used for the surface roughness measurement is also stored in the design information database 130, and measurement procedure information can be easily created automatically from the position information of the TEG 930 and the surface roughness can be measured.

  For a wafer without a pattern, an accurate Haze map can be created by measuring the entire surface of the wafer, but a Haze map can usually be created if there are about 20 measurement points per wafer. Since there are 1000 or more chips on an actual wafer, 1000 or more scribing areas exist. For this reason, if this invention is used, it is possible to produce a highly accurate Haze map even on a patterned wafer at high speed.

[Third Embodiment]
Next, a third embodiment will be described. In the system configuration of the first embodiment shown in FIG. 1, the optical wafer surface inspection apparatus 110, the data server 120, the design information database 130, and as shown in FIG. 10, the network 140, the review SEM 150, the CD-SEM 160, This is a system configuration for connecting a foreign component analysis apparatus 170, an AFM (atomic force microscope) 180, and an electrical test apparatus 190.

The review SEM 150 is an apparatus that observes and classifies defects on a wafer using a scanning electron microscope. In general, the optical defect inspection device is used to check the presence and position of foreign matter, and based on the information, the review SEM is used to observe and classify the foreign matter.
The CD-SEM 160 is a device that automatically measures the pattern width in a chip.
The foreign substance component analyzer 170 is an apparatus that analyzes foreign substance components, and is, for example, an EDX (energy dispersive X-ray analyzer).
The AFM 180 is an apparatus that measures the atomic force received from the wafer surface by the tip of the probe and observes it as the wafer surface state (unevenness).
The electrical test device 190 is a device that applies a signal waveform to the chip and compares the output waveform output from the chip with a normal waveform stored in advance in the electrical test device 190.

  Here, data of the synthesized signal synthesized by the signal synthesizing unit 240, which is data output from the signal synthesizing unit 240 included in the optical wafer surface inspection apparatus 110, and a digital signal detector based on the synthesized signal. The detection signal data 231a to 231d may be configured to be output to the review SEM 150, the CD-SEM 160, and other devices via the network 140.

Here, the operation of the system of the third embodiment will be described.
The surface roughness of the wafer is measured by the system shown in the first embodiment, and a haze value is acquired. From the haze value, the wafer surface roughness (arithmetic average roughness Ra, root mean square roughness Rms, etc.) ) And grain size, etc., and the acquired data is converted into management target data and stored in the storage unit 290.

  By performing the scattered light intensity analysis from the created scattered light intensity distribution diagram, foreign matter components can also be acquired. For example, information on the selected foreign matter is added to the foreign matter position (coordinates) information and the foreign matter size information, and the information is transmitted to the data server 120 via the network 140. Here, as a method of adding information on the selected foreign object, for example, a flag indicating whether countermeasures are necessary may be added to the inspection result. Then, in order to examine the foreign matter detected by the optical wafer surface inspection apparatus 110 in more detail, the inspection object is moved to the review SEM 150 or the CD-SEM 160. This movement may be manual conveyance or mechanical conveyance.

  After the inspection object is moved to the review SEM 150 or CD-SEM 160, the data server 120 is accessed from the review SEM 150 or CD-SEM 160, and the inspection result is received from the data server 120 via the network 140. Then, review or length measurement is started using this inspection result. During the review or length measurement, the design information database 130 is accessed, the inspection position is searched from the design information, the position information is matched, etc., the position information is accurately and quickly determined, and the review or length measurement is performed efficiently. It is possible. At this time, by using the information added by the optical wafer surface inspection apparatus 110 to preferentially review the foreign matter that needs countermeasures, it becomes possible to quickly analyze the foreign matter that causes the failure. Similarly, the foreign substance component analyzer 170 can preferentially analyze a foreign substance requiring countermeasures based on information added by the optical wafer surface inspection apparatus 110, and promptly analyze the cause of the defect. Can do.

  These review data and analysis results are stored in the data server 120 and checked with the test results in the electrical test apparatus 190 to confirm whether or not they will eventually become defective. If it does not eventually become defective, the data server 120 transmits data to the optical wafer surface inspection apparatus 110 to change the criteria for selecting foreign substances that need countermeasures, and the optical wafer surface inspection apparatus 110. By changing the standard of necessity of countermeasures, it becomes possible to select foreign substances that need countermeasures with higher accuracy, and it is possible to take countermeasures against defects in semiconductor manufacturing more quickly.

  Although the above explanation has been given taking the case of transmitting and receiving data via a network as an example, it is not necessarily performed via a network. For example, data is transferred via a removable storage medium or printed paper. May be performed.

  The present invention is suitable for use in semiconductor element inspection / measurement and semiconductor manufacturing process management in the semiconductor element manufacturing field.

The system configuration figure of a 1st embodiment. The conceptual diagram of an optical wafer surface inspection apparatus. FIG. 6 is a scattered light intensity distribution diagram when a region having no pattern on the wafer surface is measured using an optical wafer surface inspection apparatus. The flowchart of the process which measures the surface roughness of a wafer with a pattern. The figure which shows the measurable area | region of a wafer with a pattern. The flow figure showing the creation procedure of a calibration curve. The figure which showed the analytical curve in the table | surface. FIG. 8 is a graph showing a calibration curve showing the relationship between the Haze value of a sample wafer whose film type is Si in the table of FIG. 7 and the surface roughness (root mean square roughness Rms) of the sample wafer. The figure which made the scribe area | region of the wafer the measurable area | region. The system block diagram of 3rd Embodiment.

Explanation of symbols

110 Optical Wafer Surface Inspection Device 200 Wafer (Sample)
210 Sample Inspection Table 220 Illumination Light Source 221 Illumination Light (Laser Light)
230 scattered light detection units 231a to 231d detector 240 signal synthesis unit 250 overall control unit 260 stage control unit 270 information display unit 280 input operation unit 290 storage unit 295 communication unit

Claims (6)

In a method of measuring the surface roughness of a patterned wafer using scattered light generated by irradiated light,
The control means
Obtaining pre-stored design information of the patterned wafer;
And a step of searching for an area where the surface roughness of the patterned wafer can be measured using design information of the patterned wafer. 2. The wafer surface inspection method according to claim 1, wherein the control means executes a step of searching for a TEG in a scribe region of the patterned wafer using design information of the patterned wafer. The control means is
Creating a calibration curve of the sample wafer indicating the relationship between the surface roughness of the sample wafer measured and stored in advance and the scattered light intensity generated on the surface of the sample wafer stored in advance; ,
Calculating the scattered light intensity of the scattered light generated by irradiating illumination light to the area where the surface roughness of the searched wafer with a pattern can be measured;
Applying the scattered light intensity of the scattered light generated by irradiating the surface of the patterned wafer with illumination light to the sample wafer calibration curve, and calculating the surface roughness of the patterned wafer. The wafer surface inspection method according to claim 1. In an apparatus for measuring the surface roughness of a patterned wafer using scattered light generated by irradiated light,
Means for acquiring design information of the patterned wafer stored in advance;
A wafer surface inspection apparatus comprising: means for searching for an area where the surface roughness of the patterned wafer can be measured using design information of the patterned wafer. further,
5. The wafer surface inspection apparatus according to claim 4, further comprising means for retrieving a TEG of a scribe area of the patterned wafer. Storage means for storing a calibration curve of the sample wafer indicating the relationship between the surface roughness of the sample wafer and the scattered light intensity generated on the surface of the sample wafer;
Means for calculating the scattered light intensity of the scattered light generated by irradiating illumination light to the area where the surface roughness of the searched wafer with a pattern can be measured;
Means for calculating the surface roughness of the patterned wafer by applying the scattered light intensity of the scattered light generated by irradiating the surface of the patterned wafer with illumination light to the stored specimen wafer calibration curve. The wafer surface inspection apparatus according to claim 4, wherein the apparatus is a wafer surface inspection apparatus.

Images