JP2000252332A - Inspection system for semiconductor and similar device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate alteration to optimal functional ratio by executing mor than one of dimensional measurement, pattern defect inspection and review/ defect sort functions in a part of a plurality of semiconductor inspection SEMs(scanning electron microscopes) connected through a network. SOLUTION: Five semiconductor inspection SEMs(scanning electron microscopes) are connected through a network and a wafer 5 set on a stage is aligned using a part of optical microscope 13 and an alignment pattern formed on the wafer 5. The optical microscope 13 of the alignment pattern is compared with an alignment pattern reference image previously registered in a memory section 15 and the coordinates of stage position is corrected such that the visual field thereof is just superposed on that of a reference image. After the wafer 5 is positioned, pattern dimensions at a specified inspecting part are measured and then the wafer 5 is mounted on a defect detecting function section 17 and a defect sorting function section 18 where defects are detected and sorted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a scanning electron microscope (SEM) and similar devices used in inspection processes such as semiconductor device production, imaging device production, and display device production.

[0002]

2. Description of the Related Art The basic principle and configuration of a scanning electron microscope for in-line inspection (semiconductor inspection SEM) in the manufacture of semiconductor devices will be described with reference to FIG.

An electron beam 2 emitted from an electron gun 1 is
After being accelerated, it is narrowed down by the converging lens 3 and the objective lens 4 and focused on the surface of the wafer 5 as a sample. At the same time, the trajectory of the electron beam 2 is bent by the deflector 6, and the electron beam 2 scans the wafer surface two-dimensionally or one-dimensionally. On the other hand, secondary electrons 7 are emitted from the wafer portion irradiated with the electron beam as a result of the interaction between the electron beam and the wafer material. The secondary electron 7 is a secondary electron detector 8
After the detection and conversion into an electric signal, the signal processing unit 1
In 4, signal processing such as A / D conversion is performed. The processed image signal is stored in the memory unit 15. The stored image signal is used for luminance modulation or Y modulation of the display 9. The display 9 is scanned in synchronization with the scanning of the electron beam 2 on the wafer surface, and a sample image is formed on the display 9. If two-dimensional scanning is performed and luminance modulation is performed, a sample image is displayed. If one-dimensional scanning is performed and Y modulation is performed, a line profile is drawn. The formed sample image and line profile are mainly used for inspecting the pattern processing quality. For example, there are processes such as pattern dimension measurement and pattern defect inspection. Note that the pattern defect inspection step is further divided into a defect detection step and a defect classification step.

On the other hand, such a semiconductor inspection SEM is manufactured and used as a dedicated SEM for realizing each inspection function, and a length measurement SEM and a pattern defect inspection SE, respectively.
M and review SEM.

A procedure for performing an in-line inspection using a semiconductor inspection SEM is, for example, as follows.

After the wafer 5 to be measured is taken out of the wafer cassette 10, it is pre-aligned. The pre-alignment is an operation for adjusting the direction of the wafer based on an orientation flat, a notch, or the like processed on the wafer. After the pre-alignment, the wafer 5 is transferred and mounted on the XY-stage 12 in the sample chamber 11 held in a vacuum. XY-stage 12
The wafer 5 loaded above is aligned using the optical microscope 13 mounted on the upper surface of the sample chamber 11. The alignment is for correcting the position coordinate system of the XY-stage and the pattern position coordinate system in the wafer, and is executed using the alignment pattern formed on the wafer: about several hundred times the alignment pattern. The enlarged optical microscope image is compared with a reference image of the alignment pattern registered in the memory unit 15 in advance, and the stage position coordinates are corrected so that the field of view just overlaps the field of view of the reference image. After the alignment, the stage is moved to a predetermined observation point, and precise positioning and focusing of the observation point are performed. For positioning, a pattern matching method based on the above alignment is used. After positioning, measurement SE
M, pattern defect inspection SEM and review SEM, etc.
Work such as defect inspection and defect review is performed.

In dimension measurement using a length measuring SEM, an electron beam is one-dimensionally scanned across a measurement pattern in a direction in which a dimension is to be determined, and a pattern edge is determined from an obtained line profile according to a predetermined pattern edge determination algorithm. Then, the dimension of the measurement pattern is calculated from the obtained interval between the pattern edges, and the obtained calculated value is output as the dimension measurement result. The output result is stored in the inspection data management system via the network, and is used for yield improvement and production management.

In the defect detection using the pattern defect inspection SEM, the entire inspection area on the wafer is raster-scanned while combining electron beam scanning and stage movement, and the formed sample image is stored in the memory unit 15. The stored sample image is compared with a reference sample image similarly stored in the memory unit, and a difference between the two images is detected as a defect.
As the reference sample image used at this time, a sample image of the same part of the chip or cell inspected immediately before is generally used. The position coordinate data of the detected defect is output as defect data together with a sample image of the defective portion. The output result is stored in the inspection data management system via the network, and is used for yield improvement and production management.

In a review SEM, a sample image of a detected defective portion is observed, and the defects are classified according to the cause of the defect or the degree of affecting the yield based on the features appearing on the defect image. The classification result is stored in the inspection data management system via the network, and is used for yield improvement and production management.

[0010]

Inspection items, inspection steps, and inspection frequencies to be inspected in-line generally vary for each product phase such as development / prototype, small-quantity production, and mass production. In addition, in order to cope with a sudden occurrence of a defect or defect, the inspection contents may be temporarily changed.

Looking at the semiconductor inspection SEM, the development /
In the prototype stage, it is urgently necessary to establish a pattern processing technique, and a pattern dimension measuring operation using a length measuring SEM is frequently performed. Then, when the pattern processing conditions are determined and the stage of small-quantity production is started, the importance of pattern dimension monitoring becomes low, and improving the yield becomes an important issue. In order to increase the yield, the demand for a pattern defect inspection SEM and a review SEM increases for the purpose of early detection of a foreign matter source or a defect source. Furthermore, as the yield increases and the stage of mass production begins,
The biggest challenge is to maintain a high yield stably. Make sure that sudden defects do not reduce yield.
In order not to overlook the occurrence of the sudden defect, the demand for the pattern defect inspection SEM increases. That is, depending on the product phase, the demand for the length measurement SEM, the pattern defect inspection SEM, and the review SEM fluctuates, and a divergence occurs between the demand and its processing capability. If there is too much processing capacity, the device is idle or an ineffective inspection is being performed. If the processing capacity is insufficient, the probability of overlooking the occurrence of defects increases, leading to a delay in product development and a decrease in yield. At present, there is no good way to eliminate such inconvenience. In other words, it is almost impossible to respond flexibly when a sudden defect / sudden defect occurs.

[0012]

The functions of the semiconductor inspection SEM for dimensional measurement, pattern defect inspection, and review / defect classification are different depending on the product phase as described above. The total number of semiconductor inspection SEMs including the length measurement SEM, the pattern defect inspection SEM, and the review SEM is, based on experience, the development / prototype /
It does not change significantly throughout the product phases of low-volume production and high-volume production.

Focusing on this point, as a system configuration, (1) at least a part of a plurality of semiconductor inspection SEMs connected by a network is provided with functions of dimension measurement, pattern defect inspection, and review / defect classification. Multifunctional SE capable of performing at least two or more functions of
The configuration with M is one direction of the solution.

In order to realize this, from the viewpoint of device management, (2) a system in which one semiconductor inspection SEM has a plurality of device identification numbers corresponding to each processing function, or from the viewpoint of data processing. (3) The format of the inspection data used for intra-network communication is configured to include dimension measurement data, defect inspection data, and review / defect classification data.

In addition, the semiconductor inspection SEM itself has the following basic functions: (4) In order to realize a rational and economical multi-functional SEM, the commonly used functions are used as basic parts. (5) One semiconductor inspection SEM can function with a plurality of recipe systems corresponding to its processing functions so that it can be equipped with three applied functions of dimension measurement, defect detection, and defect classification. To

This makes it possible to use the three functions of dimension measurement, pattern defect inspection, and review / defect classification properly according to the time of heavy use.

The three applied functions of dimension measurement, pattern defect inspection, and review / defect classification can be properly used according to the time of heavy use. For this reason, it becomes possible to change and use the limited number of owned devices to the most effective function ratio in response to the fluctuation of the demand amount for each application function. For example, as shown in FIG. 4, a total of five semiconductor inspection SEMs are started with three dimension measurement functions and two defect inspection functions in the initial development / prototyping stage, and two dimension measurement functions in the small-quantity production stage. Defect inspection function + review function 3
Was changed to one, and in the stage of mass production, one dimension measurement function
It is used like three defect inspection functions and one defect inspection function + review function.

On the other hand, when making such a function change, the apparatus is packaged with a basic function section and has a hierarchical structure of an easily detachable applied function section. No need to do.
Therefore, the required labor, time, and cost can be minimized, and the interruption of the continuous production operation can be avoided.

Further, from the viewpoint of device manufacturing, since the number of devices manufactured for the basic function part increases, it is possible to reduce the device manufacturing cost due to the effect of mass production.

[0020]

DESCRIPTION OF THE PREFERRED EMBODIMENTS Referring to FIG. The inspection apparatus for dimension measurement, pattern defect inspection and review is composed of five semiconductor inspection SEMs connected by a network. At least four SEMs from No. 2 to No. It can have three functions of pattern defect inspection and review / defect classification. And
Each semiconductor inspection SEM is a system having a plurality of device identification numbers corresponding to the respective processing functions of dimension measurement, pattern defect inspection, and review / defect classification. In addition, these four SEMs are packaged on top of the following: (1) Commonly used wafer attachment / detachment function, wafer alignment function, observation location positioning function, focus alignment function, and sample image formation function. Easy and detachable dimension measurement function,
Defect detection function and defect classification function can be installed,
It is configured as shown in FIG.

Here, the dimension measuring function can measure not only the dimension / interval between patterns formed in the same process but also the interval between patterns processed in different processes, that is, the overlay accuracy. . Defect detection functions include sample image / reference image acquisition, sample image / reference image comparison,
It has a function of forming a difference image, detecting a difference portion as a defect, and outputting coordinate data of a position where the defect exists, a defect size, a sample image of the defect portion, and the like. The defect classification function uses the sample image /
Reference image acquisition, sample image / reference image comparison / difference image formation, feature extraction of difference part, that is, defect, comparison of extracted feature with previously learned feature data and code classification of defect,
It is configured with a function of outputting a defect classification code together with a sample image of a defective portion.

(2) It is configured to be operable with a plurality of recipe systems corresponding to the respective processing functions of dimension measurement, pattern defect inspection, and review / defect classification.

At the development / prototype stage, the dimension measurement function is No.
From No.1 to No.3, the defect inspection function started with No.4 and No.5. In the small-lot production stage, replace the No. 3 dimension measurement function with a defect inspection function + review function.
A review function is added to No. 4 and No. 5, and inspection work is performed by two dimension measuring functions and three defect inspection functions + three review functions. At the stage of mass production, the dimension measurement function of No. 2 was changed to a defect inspection function, the review function was removed from No. 3 and 4, and one dimension measurement function, three defect inspection functions, and a defect inspection function + Perform inspection work with one review function. An example of the inspection work according to the present invention will be described with reference to FIG. At the development / prototype stage, the dimension measurement function unit 16 is used. After the wafer 5 to be measured is taken out of the wafer cassette 10, it is pre-aligned. After the pre-alignment, the wafer number formed on the wafer 5 is read by a wafer number reader (not shown). The wafer number is unique to each wafer. Using the read wafer number as a key,
A recipe corresponding to the dimension measurement function of the wafer registered in the memory unit 15 in advance is read. The dimension inspection recipe defines the dimension inspection procedure and the dimension inspection conditions of the wafer. Subsequent operations are performed according to this dimension inspection recipe.
Automatic or semi-automatic. After reading the dimension inspection recipe, the wafer 5 is transported and mounted on the XY-stage 12 in the sample chamber 11 held in a vacuum. The wafer 5 loaded on the stage is aligned using the optical microscope 13 mounted on the upper surface of the sample chamber 11 and the alignment pattern formed on the wafer 5. The optical microscope image of the alignment pattern is compared with an alignment pattern reference image registered in the memory unit 15 in advance, and the stage position coordinates are corrected so that the field of view just overlaps the field of view of the reference image. After the alignment, the wafer 5 is positioned using the positioning pattern formed on the wafer 5.
Precise positioning and focusing of the pattern to be inspected formed thereon are performed. The positioning pattern is obtained by moving the stage immediately below the electron beam, and then performing electron beam irradiation, focusing, and image formation. The positioning pattern image is compared with a positioning pattern reference image registered in the memory unit 15 in advance, similarly to the alignment operation.
The scanning area of the electron beam is finely adjusted so that the two images just overlap. With respect to the positioned wafer 5, the pattern size of a predetermined inspection portion is measured. The dimensions of all the inspection locations specified in advance on the wafer are measured by repeatedly performing the operations after the positioning and focusing of the pattern to be inspected. The inspection of one wafer is thus completed. When a plurality of wafers to be measured remain in the wafer cassette, the next wafer is taken out of the wafer cassette, and the measurement is repeatedly performed according to the above-described operation procedure.
The dimension measurement result is output together with the position coordinate data of the measurement point, the image of the measurement point, and the like, and is stored in the inspection data management system, accompanying the identification number of the apparatus for dimension measurement expressed as, for example, CD-05. Used for analysis for yield improvement and production control. The device identification number for dimension measurement is
In this case, the apparatus is used for dimension measurement, and (1)
The key code is for reading the dimension inspection recipe, and (2) is the key code for storing and managing the dimension measurement data. In addition, 05 of CD-05 indicates a device number as a semiconductor inspection SEM.

In the small-quantity production stage and the mass production stage, the dimension measurement function unit 16 is removed, and the defect detection function unit 17 and the defect classification function unit 18 are mounted. The conversion from the length measurement SEM to the pattern defect inspection SEM + review SEM is easily performed because each function is packaged. There is no complicated and expensive work such as replacement of equipment, and the influence on production work is minimized. The operation procedure after the diversion conforms to the dimension measurement described above, but a recipe of pattern defect inspection + review is used, and defect detection and defect classification are performed instead of the dimension measurement. In the defect detection, the sample image and the reference image are compared over a predetermined inspection area, and a difference between the two images is detected as a defect. In the detected defect SEM image, a feature amount such as a shape or surface roughness specified in advance is automatically extracted, and is automatically classified based on learning data of the feature amount registered in advance. The classified result is output together with the position coordinate data of the defect, the defect SEM image, etc., along with the device identification number for pattern defect inspection + review expressed as, for example, DR-05, and is output to the inspection data management system. After being stored, it is used for analysis for yield improvement and production control. In this case, the device identification number for pattern defect inspection + review is such that the present device is used for pattern defect inspection + review.
(2) It is a key code for storing and managing pattern defect inspection data and review data.

Here, sample image / reference image acquisition, sample image /
The reference image comparison / difference image formation may be configured to be commonly used for defect detection and defect classification. Further, as shown in FIG. 3, a wafer attaching / detaching function, a wafer alignment function,
Observation point positioning function, focus position adjustment function, and sample image forming function as basic functions, dimension measurement function that can be arbitrarily attached and detached packaged, sample image / reference image and / or difference image and sample image acquisition It is configured so as to be equipped with an output function of position coordinate data, and the defect detection and / or defect classification is installed independently of the SEM, and is performed by a defect detection function terminal and / or a defect classification function terminal connected via a network. Is also good. In this case, the defect inspection sample image / reference image output function of the semiconductor inspection SEM includes the defect inspection sample image / reference image acquisition, the defect inspection sample image / reference image output, and / or the difference image formation / output and the defect position coordinate data output function. It consists of.

The reference image used for defect detection and / or defect classification can be registered in advance before the inspection work, newly registered or re-registered during the inspection work. For example, it is necessary to repeatedly register the image of the same pattern portion of the chip or cell inspected immediately before as the reference image during the inspection operation. Further, pattern design information can be used as a reference image instead of the sample image.

In the case where the defect detection is performed continuously over a certain area, it is preferable that the focus positioning operation is performed in parallel with the defect detection operation, and the observation point positioning operation can be skipped. It is possible.

In this embodiment, the dimension measuring function is removed, but it may be mounted as it is. that way,
If sudden dimensional defects occur at the production stage and you want to increase the frequency of dimensional inspection, you can respond immediately.

In this embodiment, in order to discriminate whether the inspection data is the dimension measurement data, the pattern defect inspection data, or the review data, the dimension measurement,
The system has a plurality of device identification numbers corresponding to the processing functions of pattern defect inspection and review / defect classification. Instead, the format of the inspection data used for communication in the network is dimension measurement data, defect inspection. It may be configured to include data and review / defect classification data.

For an insulating sample which takes a long time to saturate the charge-up, it is preferable to irradiate an electron beam for a predetermined time and then to load the sample image into a memory. It is also possible to determine the irradiation time from the saturation characteristic data obtained in advance and make a recipe.

Although the XY-stage is used here,
If an XYT-stage is used instead of the Y-stage,
Defect detection can be performed while the sample is tilted.

Although only the detection of a defect has been described here,
The characteristic X
It is also possible to attach an analysis function such as a line analyzer or an Auger electron analyzer.

Here, the case of one probe and one pixel has been described, but a method of forming an image with multiple probes or multiple pixels may be used.

Here, the case of observing a semiconductor wafer has been described, but a wafer for an image pickup device or a display device may be used instead, or a sample shape other than a wafer may be used.

In this embodiment, an electron beam has been exemplified as a probe that causes interaction with a sample, and a scanning electron microscope has been exemplified as an apparatus. However, a focused ion beam apparatus using an ion beam, a laser scanning microscope using a laser beam,
An atomic force microscope using a mechanical probe or the like may be used.

[0036]

According to the present invention, it is possible to easily change to the optimum function ratio in response to a change in the demand for each inspection function. Therefore, the labor, time, and cost associated with the change can be minimized. Further, it is possible to reduce the device manufacturing cost due to the mass production effect.

[Brief description of the drawings]

FIG. 1 is a diagram for explaining the principle of a semiconductor inspection SEM and the device configuration of the present invention.

FIG. 2 is a diagram illustrating a configuration of each function of a semiconductor inspection SEM according to the present invention.

FIG. 3 is a diagram for explaining another configuration of each function of the semiconductor inspection SEM according to the present invention.

FIG. 4 is a diagram for explaining an example of a function change of the semiconductor inspection SEM according to the present invention.

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Electron gun, 2 ... Electron beam, 3 ... Convergent lens, 4 ... Objective lens, 5 ... Wafer, 6 ... Deflection coil, 7 ... Secondary electron, 8 ... Secondary electron detector, 9 ... Display, 10 ... Wafer Cassette, 11: sample chamber, 12: XY-stage,
13 optical microscope, 14 signal processing unit, 15 memory unit, 16 dimension measurement function unit, 17 defect inspection function unit, 1
8. Defect classification function unit.

Continued on the front page F-term (reference) 2G001 AA03 AA05 AA10 AA13 BA05 BA07 BA09 CA03 CA10 DA06 FA01 FA06 GA03 GA04 GA06 GA13 HA09 JA02 JA08 JA11 JA15 KA01 KA03 KA20 LA11 PA01 PA07 2G051 AA51 AB02 BA10 BA20 BC05 CA03 CA04 EA07 EA12 EA14 EA21 EB01 EB09 EC01 4M106 AA01 BA02 CA39 CA42 CA43 CA50 DB05 DB21 DB30 DH33 DJ04 DJ07 DJ15 DJ17 DJ18 DJ20 DJ21 DJ26 DJ32

Claims (6)

[Claims] An inspection system and method comprising a plurality of scanning electron microscopes and similar devices connected by a network and used for device inspection in the manufacture of semiconductors and similar devices. An inspection system and method, wherein at least a part of an electron microscope and similar devices have at least two or more functions of dimension measurement, pattern defect inspection and review. 2. The apparatus according to claim 1, wherein one scanning electron microscope and similar devices have a plurality of device identification numbers corresponding to respective functions of dimension measurement, pattern defect inspection and review. Inspection system and method. 3. The inspection system and method according to claim 1, wherein the format of the inspection data communicated through the network has a configuration including dimension measurement data, defect inspection data, and review data. 4. A scanning electron microscope and a similar device constituting an inspection system, at least a part of which has a sample attaching / detaching function, a sample alignment function, an observation point positioning function, a focus positioning function, and a sample image forming function. As a basic function, at least one (or one or more) of the three functions of a dimension measurement function, a defect detection function, and a defect classification function packaged thereon can be optionally mounted. The inspection system and method according to claim 1. 5. A scanning electron microscope and at least a part of a similar device constituting an inspection system have a basic function of a sample attaching / detaching function, a sample alignment function, an observation point positioning function, a focus positioning function, and a sample image forming function. 2. The apparatus according to claim 1, wherein a dimension measurement function packaged thereon and an output function of sample image / reference image and / or difference image and sample image acquisition position coordinate data can be mounted. Inspection system and method as described. 6. A manufacturing system and a manufacturing method for semiconductors and similar devices, characterized by using the inspection system and method according to claim 1.