JPH05113406A - Inspecting apparatus for defect of wafer - Google Patents

Abstract

PURPOSE:To reliably inspect a defect on the surface of a wafer at high speeds by photographing the surface of the wafer by means of a plurality of line image sensor elements shifted one another while moving the wafer relatively to a plurality of line image sensors. CONSTITUTION:A plurality of CCD line image sensors 1-A, 1-B, 1-C constituting an image detecting mechanism to photograph the surface of a wafer have a plurality of sensor elements A-1-A-3, B-1-B-3, C-1-C-3 corresponding to pixels. The line image sensors 1-A-1-C are shifted one another by a distance obtained by dividing the size of one pixel by the number of the line image sensors. While the wafer is moved relatively to the line image sensors 1-A-1-C by a moving means, the wafer is photographed and, the obtained image data is analyzed by an image analyzing means thereby to detect a defect.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a wafer defect inspection apparatus for inspecting defects on the surface of a wafer, and more particularly to an OSF (Oxidation-induced Stacking) for a silicon wafer.
Wafer defect inspection apparatus for inspecting faults.

[0002]

2. Description of the Related Art Conventionally, a defect inspection of a surface of a wafer, particularly an OSF inspection of a silicon wafer, is performed by a person observing the surface of the wafer with a microscope. Such an inspection
Although it is desirable to perform the entire surface of the wafer, in practice, as shown in FIG.
A method of inspecting a restricted area called a cross method in which only the direction observation area 3 and the Y direction observation area 4 are inspected is adopted. In FIG. 1, 2 is an orientation flat.

As described above, the inspection method in which a human observes using a microscope has problems such as variations depending on the measurer, fluctuations due to the measurement time, physical condition of the measurer on the day, and inconsistent judgment criteria. There has been a demand for a method of automatically inspecting the above, and recently, some OSF inspection apparatuses, that is, apparatuses capable of automatically inspecting the OSF in a designated area have been developed and put into practical use.

In such a conventional OSF inspection apparatus, the surface of the wafer, which is fixed on the stage and whose position is grasped, is observed with an optical microscope, and the image information obtained by this microscope is used as a lens barrel of the microscope or an eyepiece lens. The image is processed by an image pickup tube or a CCD area sensor attached to the unit and supplied to an image processing apparatus, and the image processing apparatus analyzes the image to distinguish the OSF from other flaws, foreign matters, and the like.
By repeating such a process for all the target surfaces of the wafer, the O
The number and density of SFs are calculated.

[0005]

It has been found that the state of the raw silicon wafer has a great influence on the characteristics and yield of the device due to the recent high integration and high performance of the device. There is a growing need to inspect large areas or nearly the entire surface of a silicon wafer, rather than inspecting only partial areas.

However, in order to inspect the entire surface of the silicon wafer in this way, it takes a very long time by the conventional manual method, and it is practically impossible from the viewpoint of uniformity within a long time. There is a problem that is.

Further, in the conventional automated system in which an area sensor is attached to a microscope, even if a 10 × objective lens is used, one target area is as small as 0.5 mmφ, and for measuring the entire surface, for example, It is necessary to measure several tens of thousands of points on a 6 '' wafer, and it takes a very long time to stop the movement of the stage. For example, even an apparatus capable of performing a relatively fast process requires 10 hours or more. is there.

The present invention has been made in view of the above,
It is an object of the invention to provide a wafer defect inspection apparatus capable of accurately inspecting defects on the surface of a wafer at a practically high speed.

[0009]

In order to achieve the above object, a wafer defect inspection apparatus of the present invention comprises a plurality of line image sensors for imaging the surface of a wafer, each line image sensor being a line image sensor of the line image sensor. Number 1
A plurality of line image sensors arranged so as to be displaced from each other by a distance obtained by dividing the size of the pixel, a moving means for relatively moving the wafer with respect to the plurality of line image sensors, and a relative means by the moving means. The image analysis means for analyzing the image information of the wafer obtained by capturing the image of the wafer moving to the above with the plurality of line image sensors to detect the defect.

[0010]

In the wafer defect inspection apparatus of the present invention, while moving the wafer relative to the plurality of line image sensors, the surface of the wafer is imaged by the plurality of line image sensors arranged so as to be offset from each other. Defects are detected by analyzing the image information of the imaged wafer.

[0011]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 shows a plurality of (three in the figure) CCD line image sensors 1-A, 1-B constituting an image detecting mechanism used in a wafer defect inspection apparatus according to an embodiment of the present invention. , 1-C. These line image sensors 1-A, 1-B, 1-C include a plurality of sensor elements A-1, A-2, respectively corresponding to respective pixels.
A-3, ..., B-1, B-2, B-3, ..., C
-1, C-2, C-3, ... More specifically, each line image sensor has 1728 sensor elements corresponding to 1728 pixels. The size of each pixel is a rectangle with one side of 15 μm. Therefore, each line image sensor has an imaging width of 15 μm × 1728≈25 mm.

The line image sensors 1-A, 1-
B and 1-C are not arranged so that the positions of the pixels are arranged in a line, but are slightly displaced as shown in the drawing, specifically, are displaced by 5 μm in the direction in which the pixels are arranged. hand,
It is arranged. By arranging in such a shifted manner, it is possible to detect defects up to such a shifted length. More specifically, a line image sensor having a width x (μm) of one pixel is shifted by a length y (μm) to n.
When arranged individually, the length up to the minimum y (μm) can be determined. Then, in order to have a resolution of y (μm) at any position of the line image sensor, n
= X / y line image sensors, as shown in FIG.
It is necessary to shift them by (μm). The line image sensor of FIG. 1 has x = 15 μm, y = 5 μm, n =
It is 3.

FIG. 2 is a diagram showing a method of measuring an object 11 having a length shorter than 15 μm by the line image sensors 1-A, 1-B and 1-C shown in FIG.
The line image sensors 1-A, 1-B, and 1-C are configured to move with respect to the object 11 by a moving unit (not shown) in a method indicated by an arrow 31. Then, initially, the object 11 existing in front of the moving direction of the line image sensors 1-A, 1-B and 1-C is indicated by 11 'in FIG. 2 due to the movement of the line image sensor after a certain time has elapsed. When it comes to the position, the object 11 'is detected by the sensor element A-1 of the line image sensor 1-A. As a result, it can be seen that the length of the object 11 is 15 μm or less.

Next, after a certain time has passed, the object 1
If 1 comes to the position of 11 '', this object 11 '' is detected only by the sensor element B-2 of the line image sensor 1-B. Here, the fact that the objects 11 'and 11''are the same object 11 means that the moving speed of the line image sensor in the direction of the arrow 31 and the sensor elements A-1 and B-2.
It can be seen from the distance between and the difference between the times when the objects 11 'and 11''are detected. Further, after a certain time has passed, the object comes to the position of 11 '''and is detected by the sensor element C-2. As a result, the object 11 is the sensor element A-
Since it is detected only by 1, B-2, C-2 and not by other sensor elements, it is identified that the length of the object 11 is 5 μm or less. The lengths of objects having other lengths can be detected by the same method as described above.

FIG. 3 is an explanatory diagram in the case of detecting an object 12 whose length is longer than one pixel. By the same method as described with reference to FIG.
1, A-2, B-1, B-2, C-2, C-3. In addition, it is determined whether the object 12 detected by each of the sensor elements A-1, A-2 or B-1, B-2 or C-2, C-3 is the same as the object 12. Sensor element A-1, A-2 or B-1, B
-2 or both of C-2 and C-3 have the same time, and it can be determined. From such a result, it is understood that the length of the object 12 is longer than 10 μm and 20 μm or less.

Next, with reference to FIG. 4, a method of determining whether or not the object detected by the line image sensor is a defect (OSF) will be described.

In order to discriminate whether or not the object detected by the line image sensor is OSF, it is necessary to detect the object by the line image sensor and then detect the width of the detected object. .. Therefore, in FIG. 4, the line image sensors 1-A, 1-B, and 1-C are moved at a certain speed S (μm / sec) in the same manner as in FIG.
The time when a certain sensor element detects an object in this movement is defined as t1, the time when the sensor element deviates from this detection is defined as t2, and the width of the pixel corresponding to each sensor element is x (μm) ( In this embodiment, 15 μm),
The width w of the object can be obtained from the following equation.

W = S / (t2-t1) -x (1) In this embodiment, the moving speed S is 1000 μm / sec.

Generally, the OSF is determined within a certain range of values of the length and width, and other than the OSF, the ratio of the length to the width is usually smaller than that of the OSF. Also, the OS
Since the width of F is usually 0.5 to 1 μm, whether or not it is OSF can be clearly discriminated from these facts.

For example, in FIG. 4, the OS of the object 13
For F, the width Wc can be obtained by the method described in FIGS. 2 and 3, and when this width Wc is found to be 1 μm,
It turns out that it is OSF. Moreover, since the length of the dust of the object 14 is measured to be 10 to 15 μm and the width Wd is about 5 μm from the method shown in FIGS. 2 and 3 and the method described in the above equation (1), Since the width ratio is smaller than the OSF, the object 14 can be determined to be dust.

However, even in the OSF, in the case of a silicon wafer having a crystal orientation of (100), an OSF 15 perpendicular to the OSF 13 appears. In this case, the length and the width may be reversed and the ratio may be determined, or the wafer may be rotated by 90 ° and scanning may be performed again.

Further, in the case of a silicon wafer having a crystal orientation of (111), OSFs appear in two directions (a total of three directions) that are deviated by 60 ° from a certain OSF direction. In this case, , Can be determined by rotating the wafer twice by 60 ° and re-scanning.

FIG. 5 and FIG. 6 are explanatory views showing a method of utilizing the change of the light amount detected by the sensor element to judge the defect on the surface of the wafer. That is, in this method, the object is an OSF using the change in the amount of light from the object when the sensor element detects the object and the length of the object obtained by the method shown in FIGS. 2 and 3. It is to determine whether or not there is.

In FIG. 5, the object 18 is a sensor element A.
If -8 detects that the object 18 is at a certain time t1
At 8 ', the position reaches 18' in FIG. 5 and a part of it is detected by the sensor element A-8. The light amount at this time is the light amount at time t18 'shown in (c) of FIG. .. After that, the object 18 is 18 '' in FIG. 5 at time t18 ''.
When it reaches the position indicated by and is detected by the sensor element A-8 as a whole, the light amount at this time becomes the light amount shown at time t18 '' in (c) of FIG. Further, at the subsequent time t18 ''', the object 18 comes to the position indicated by 18''' in FIG.
A part thereof is detected by the sensor element A-8, and the light amount at this time becomes the light amount shown at time t18 '''in FIG. 6C.

By investigating the change of the light quantity of the object in this way, the change of the light quantity becomes as shown in FIG. 6 (c). Then, the ratio of the length to the width of the object 18 is calculated from the shape of the change in the amount of light shown in this figure, particularly the maximum amount of light obtained as a result, and the length of the object obtained by the method shown in FIGS. 2 and 3. Therefore, the object 18 can be determined to be the OSF.

The change in the amount of light in the case of the object 17 shown in FIG. 5 is as shown in FIG. 6 (b).
It can be determined to be something other than. Further, the object 16 shown in FIG. 5 is an L-shaped object in which two OSFs are in contact with each other, and the change in the amount of light in this case is shown in FIG.
As shown in. Whether or not it is an OSF can be determined from such a change in the light amount and the length obtained by the above-described method.

In the above description, the defect has been described as being contained within one pixel, but the same applies to a defect extending over two or more pixels. In this case, the light amount is detected as the sum of the light amounts of the pixels corresponding to the length of the defect.

Next, the operation of the above-described wafer defect inspection apparatus of this embodiment will be described with reference to the flow chart shown in FIG.

In FIG. 7, when the inspection is started,
Inspection conditions are set (steps 110 and 120). When the inspection conditions are set, the wafer is transported, and the wafer is set on the stage of the present wafer defect inspection apparatus with the orientation flat aligned (step 130). When the orientation flat of the wafer is set, the movement of the wafer is started by the driving mechanism or the like (step 140). The sensor element corresponding to each pixel of the line image sensor shown in FIG. 1 starts capturing a surface image of the wafer, and pixel passing time and position coordinate capturing processing is performed (step 150),
The OSF is judged (step 160). This OS
The determination of F is performed based on the number of pixels of the detected object, the pixel passage time, the change in the light amount of the detected object, and other parameters used in the above-described methods. When the OSF is determined, the number of OSFs is counted, the wafer is transferred, and the inspection is completed (steps 170, 180, 190).

FIG. 8 shows the results of comparing the inspection time per wafer and the measurement accuracy using the above-described wafer defect inspection apparatus of the present invention and the conventional defect inspection apparatus. The wafer used for this inspection is N type,
10 silicon wafers with a crystal orientation (100) and a diameter of 150 mm, which are considered to have an OSF density of several pieces / cm ² to 100 pieces / cm ² in order to compare the OSF measurement accuracy. did.

Then, the prepared silicon wafer 1
After cleaning 0 sheets, heat treatment is performed in an oxygen atmosphere at 1000 ° C. for 16 hours, and then oxide film peeling treatment with hydrofluoric acid is performed.
Further, Et'g for 2 minutes with Wright Et'g solution, and
The SF was made observable. 10 processed in this way
The wafers were measured by the wafer defect inspection device of the present invention and the conventional defect inspection device, respectively, and the measurement time and the measurement accuracy were compared. The measurement conditions are the same as those shown in FIG. 1 in the embodiment of the present invention, one pixel size is 15 μm square, and three line image sensors with 1728 pixels are 5 μm.
The actual inspection width is set to 2 by using ones staggered by one.
5 mm, the scanning speed S is 1000 μm / sec,
Further, the area excluding the 5 mm width around the wafer is the wafer 1
6 scans were performed per sheet. In addition, the measurement condition of the conventional inspection device is a normal measurement magnification of 1 FOV area of 0.25 mm ^D (total magnification of 80 times), and the peripheral area of the wafer is 5 m.
The entire area except the m width was measured.

The above measurement results are shown as a table in FIG. 8, and the unit of each measured value excluding the percentage difference in the table of FIG. 8 is K / cm ² . FIG. 9 is a graph showing the correlation between the OSF measurement values (OSF densities) of the wafer defect inspection apparatus of the embodiment of the present invention and the conventional inspection apparatus based on the measurement results shown in FIG. From FIGS. 8 and 9, it can be seen that the measurement error between the wafer defect inspection apparatus of this embodiment and the conventional apparatus is within 7%, and there is sufficient correlation.
Therefore, it is understood that the defect measurement accuracy of the wafer defect inspection apparatus of this embodiment is the same as that of the conventional one.

Regarding the measuring time, the time per wafer by the conventional inspection apparatus is ten and several hours, while the wafer defect inspection apparatus of this embodiment is significantly 14 minutes per wafer. I found it fast.

As described above, the present wafer defect inspection apparatus has the same OSF inspection accuracy as the conventional one, and the inspection speed is as fast as 70 to 80 times in the whole surface scan of a 6 ″ φ wafer, and can be sufficiently put into practical use. Turned out to be speed. As a result, it is possible to surely evaluate the entire surface of the wafer, significantly improve the yield of semiconductor products, and save resources and energy.

In the above embodiment, the case where the line image sensor is used has been described, but the present invention is not limited to the line image sensor and may be an area image sensor or the like.

[0037]

As described above, according to the present invention,
While moving the wafer relative to the plurality of line image sensors, the surface of the wafer is imaged by the plurality of line image sensors arranged so as to be displaced from each other, and the image information of the imaged wafer is analyzed to detect defects. Since it is detected, defects on the wafer surface can be accurately inspected at a very high speed.

[Brief description of drawings]

FIG. 1 is a diagram showing a line image sensor forming an image detection mechanism used in a wafer defect inspection apparatus according to an embodiment of the present invention.

FIG. 2 is a diagram showing a method of measuring an object having a short length by the line image sensor shown in FIG.

FIG. 3 is an explanatory diagram for detecting an object whose length is longer than one pixel.

FIG. 4 is an explanatory diagram showing a method of determining whether or not an object detected by a line image sensor is a defect (OSF).

FIG. 5 is an explanatory diagram showing a method that utilizes a change in the amount of light detected by a sensor element to determine a defect on the surface of a wafer.

FIG. 6 is a diagram showing changes in the amount of light detected by a sensor element for determining a defect on the surface of a wafer.

FIG. 7 is a flowchart showing the operation of the wafer defect inspection apparatus according to the embodiment of the present invention.

FIG. 8 is a table showing the results of comparison of the inspection time per wafer and the measurement accuracy using the wafer defect inspection apparatus of the embodiment of the present invention and the conventional defect inspection apparatus.

9 is a graph showing the correlation between the OSF measurement values of the wafer defect inspection apparatus of the embodiment of the present invention and the conventional inspection apparatus based on the measurement results shown in FIG.

FIG. 10 is a diagram showing a normal observation area in a defect inspection by a conventional optical microscope.

[Explanation of symbols]

 1-A, 1-B, 1-C Line image sensor A-1 to C-2 ... Sensor element 11 Object

Claims (1)

[Claims] 1. A plurality of line image sensors for imaging the surface of a wafer, wherein each line image sensor is arranged offset from each other by a distance obtained by dividing the size of one pixel by the number of the line image sensors. A plurality of line image sensors, a moving unit that relatively moves the wafer with respect to the plurality of line image sensors, and an image of the wafer in which the plurality of line image sensors image the wafer that is relatively moved by the moving unit. A wafer defect inspection apparatus comprising: an image analysis unit that analyzes information to detect a defect.