JP5879083B2 - Semiconductor wafer evaluation method - Google Patents

Description

  The present invention relates to a method and apparatus for evaluating the breaking strength of an edge portion of a semiconductor wafer used in a semiconductor device manufacturing process or the like.

When a crack occurs in a semiconductor wafer such as a silicon wafer, which is a material in the semiconductor device manufacturing process, a large loss occurs. For this reason, there is a high demand for wafers that are difficult to break during device manufacturing.
In the manufacturing process of semiconductors and liquid crystals, especially in the processes such as dry etching, ion implantation, and vapor deposition, high temperature / rapid heating / rapid cooling is progressing, and the manufacturing processes performed by vacuum and drying are also increasing. In addition, silicon wafers and glass substrates as substrates have been increased in diameter, and resistance to impacts and the like has become more important.

As a cause of the destruction of the semiconductor wafer, there are many cases mainly caused by hitting the wafer edge portion, and it is important to evaluate the impact strength of the edge portion.
Since silicon wafers and the like are crystalline brittle materials, there is a large variation in measured values in a general material evaluation technique. Standard equipment for evaluating and inspecting the easiness of cracking of a wafer is not commercially available. For this reason, for example, an apparatus as shown in Patent Documents 1-4 has been devised.

JP 2006-287139 A JP 2004-101258 A JP-A-5-288663 Japanese Patent Laid-Open No. 6-201533

However, even if the apparatus of Patent Documents 1-4 is used, the evaluation of the fracture strength of the semiconductor wafer varies, and it is difficult to evaluate with high accuracy.
For example, silicon is a cubic system with a diamond structure, and there is anisotropy in physical properties such as Young's modulus, and anisotropy is also clearly seen in fracture strength depending on the crystal orientation. However, conventionally, this anisotropy has not been considered. Even in Patent Documents 1-4, since the influence on the fracture strength of anisotropy is not taken into account, the ability of sensitivity and accuracy is insufficient to evaluate a slight difference in the strength of the semiconductor wafer. There is no JIS standard for evaluating the anisotropy of the strength of a silicon wafer.

  The present invention has been made in view of the above problems, and provides a method and an apparatus capable of reducing the variation in measurement results and performing a highly accurate evaluation in the evaluation of the breaking strength of the edge portion of a semiconductor wafer. The purpose is to do.

  In order to achieve the above object, the present invention is a method for evaluating the fracture strength of an edge portion of a semiconductor wafer, applying a load to the edge portion corresponding to a predetermined crystal orientation of the semiconductor wafer to be evaluated, There is provided a method for evaluating a semiconductor wafer, wherein the breaking strength of an edge portion of the semiconductor wafer is evaluated.

  In this way, by applying a load to the edge portion corresponding to a predetermined crystal orientation and evaluating it, it is possible to reduce the influence of the crystal anisotropy on the evaluation, thereby suppressing variation in evaluation and improving sensitivity. be able to. Further, if a load can be applied to the edge portion corresponding to a predetermined crystal orientation, a conventional apparatus or the like can be used without changing the design and can be easily implemented. Therefore, according to the present invention, the breaking strength of the edge portion of the semiconductor wafer can be evaluated with high accuracy and efficiency.

At this time, it is preferable that the predetermined crystal orientation is a crystal orientation in which the breaking strength of the edge portion of the semiconductor wafer is highest.
Thus, if a load is applied to the edge portion corresponding to the crystal orientation having the highest fracture strength, the influence on the measurement result due to the crystal anisotropy can be further reduced, and the evaluation accuracy can be further improved.

At this time, it is preferable to apply the load by a falling weight impact method or a horizontal static pressure load method.
With such a method, the breaking strength can be evaluated efficiently and accurately.

At this time, a test piece including an edge portion corresponding to the predetermined crystal orientation is cut out from the semiconductor wafer to be evaluated, and the test piece is held between two holding jigs. It is preferable to apply a load to the edge portion and evaluate the breaking strength of the edge portion of the semiconductor wafer.
Thus, by applying a load and evaluating, the influence on the measurement result by the bending of the test piece at the time of load application, etc. can be suppressed, and a fracture strength can be evaluated accurately.

At this time, it is preferable that the two holding jigs are provided with notches that expose the periphery of the edge portion corresponding to the predetermined crystal orientation.
By using such a holding jig, since the portion other than the portion to which the load is applied can be held and held by the holding jig, the influence on the measurement result due to the deflection of the test piece or the like can be reliably prevented.

  According to another aspect of the present invention, there is provided an apparatus for evaluating the breaking strength of an edge portion of a semiconductor wafer, wherein a load is applied to the edge portion corresponding to a predetermined crystal orientation of the semiconductor wafer to be evaluated. A wafer evaluation apparatus is provided.

  In this way, if the apparatus evaluates by applying a load to the edge portion corresponding to a predetermined crystal orientation, the influence on the evaluation due to the crystal anisotropy can be reduced. Can be improved. Further, since only a load is applied to the edge portion corresponding to a predetermined crystal orientation, a simple apparatus can be obtained. Therefore, the apparatus can accurately and efficiently evaluate the breaking strength of the edge portion of the semiconductor wafer.

At this time, it is preferable that the predetermined crystal orientation is a crystal orientation in which the breaking strength of the edge portion of the semiconductor wafer is highest.
In this way, if the load is applied to the edge corresponding to the crystal orientation with the highest fracture strength, the effect of the crystal anisotropy on the fracture strength can be further reduced, and the evaluation accuracy is further improved. It becomes a device that can.

At this time, it is preferable that the load is applied by a falling weight impact method or a horizontal static pressure load method.
With such a system, the apparatus can be efficiently and accurately evaluated.

At this time, it is preferable that the evaluation apparatus has two holding jigs for holding a test piece including an edge portion corresponding to the predetermined crystal orientation cut out from the semiconductor wafer to be evaluated.
If it has such a holding jig, it becomes an apparatus which can suppress the influence by the bending of the test piece at the time of load provision, etc., and can evaluate fracture strength correctly.

It is preferable that the two holding jigs are provided with a notch that exposes the periphery of the edge portion corresponding to the predetermined crystal orientation.
If it is a holding jig provided with such a notch, it can be held by holding the portion other than the portion to which the load is applied, so that it is possible to reliably prevent the influence on the evaluation due to the deflection of the test piece. .

  As described above, according to the present invention, the breaking strength of the edge portion of the semiconductor wafer can be evaluated accurately and efficiently.

It is a figure which shows the level of the fracture strength by the crystal orientation of a semiconductor wafer. It is the schematic which shows an example of the destructive testing machine of the falling weight type impact system which can be used for this invention. It is the schematic which shows an example of the measuring apparatus of the horizontal direction static pressure load system which can be used for this invention ((a) is a side view, (b) is a top view). In this invention, it is explanatory drawing which shows an example of the process after cutting out a test piece from a semiconductor wafer and performing a test. In this invention, it is explanatory drawing which shows the other example of the process after cutting out a test piece from a semiconductor wafer and performing a test. It is the schematic which shows an example of the destructive testing machine of the falling weight type impact system which can be used for this invention. In this invention, it is explanatory drawing which shows an example of a process until holding a semiconductor wafer and performing a test. It is the schematic which shows the other example of the measuring apparatus of the horizontal direction static pressure load system which can be used for this invention ((a) is a side view, (b) is a top view).

In silicon wafers several generations ago, there were cases where precipitation, strain, etc. remained in the outermost periphery, and there were cases where the fracture strength in one lot exceeded “maximum value ÷ minimum value = 10 times”. However, recent silicon wafers have greatly improved crystal manufacturing methods and wafer processing methods. In the conventional evaluation method, it is difficult to evaluate the maximum to the minimum value of the fracture strength in one silicon wafer, and improvement of the evaluation method has been demanded.
On the other hand, the present inventor found that a highly accurate evaluation can be performed by conducting a test in consideration of the anisotropy of the crystal of the semiconductor wafer to be evaluated. Completed.

  Hereinafter, the present invention will be described in detail as an example of an embodiment with reference to the drawings, but the present invention is not limited thereto.

  The present invention is a method for evaluating the breaking strength of an edge portion of a semiconductor wafer by applying a load to the edge portion corresponding to a predetermined crystal orientation of the semiconductor wafer to be evaluated.

An anisotropic crystal such as silicon has a fragile surface called cleaving in a specific direction because of its crystal structure. Cleavage is a surface where the bonding force between atoms is weak in the crystal structure, and is completely different from the hardness of the wafer. Therefore, the present inventors have found that when a load is applied regardless of the crystal orientation as in the prior art, the direction in which the load is applied and the extension direction of cracks due to the effect of cleaving are different, and the variation in measured strength is large. It was.
For this reason, if a load is applied to the edge portion corresponding to a predetermined crystal orientation as in the present invention for evaluation, the variation in measurement intensity due to the crystal anisotropy as described above can be reduced, and the accuracy of the evaluation is improved. it can.

At this time, it is preferable that the predetermined crystal orientation is a crystal orientation in which the fracture strength of the edge portion of the semiconductor wafer is highest.
In the crystal orientation in which the breaking strength is highest as described above, a load can be applied to the position or direction where the influence of cleaving is the least, so that the breaking strength of the semiconductor wafer can be evaluated with high accuracy.

For example, when evaluating a (100) silicon wafer with a notch in (011), a high to low fracture strength at the edge in the silicon wafer is determined by a specific angle with the {110} cleaved surface. Can be evaluated. FIG. 1 is a diagram showing the difference in fracture strength depending on crystal orientation.
As shown in FIG. 1, the edge portion having an angle θ = 20 ° to 28 ° with the {110} cleaved surface has the highest fracture strength of the edge portion in one silicon wafer and is difficult to break. The inventor found.

In a silicon single crystal, the {111} plane (atomic distance 3.13Å, Young's modulus 190 Gpa) has the highest atomic density and Young's modulus, so the bonding force between {111} planes is smaller than that of other planes, { The cleavage is most likely to occur on the 111} plane. The cleaved surface that is most likely to cleave is the {110} plane.
Since the (100) wafer is easily cracked at 45 ° vertically and horizontally with respect to the {110} cleavage plane, the angle θ = 20 ° to 28 ° with the {110} cleavage plane is the {110} plane, {100 } Surface and {111} surface are considered to be the least susceptible to influence.

Therefore, it can be seen that the edge portion corresponding to the angle θ = 26.565 ° from the {110} cleaved plane (the {210} crystal plane group can be expressed as a representative plane) has the highest fracture strength. It is preferable to apply a load perpendicular to the part.
In addition, it is preferable to examine in advance each crystal orientation with such a high fracture strength.

Further, a test piece including an edge portion corresponding to a predetermined crystal orientation is cut out from the semiconductor wafer to be evaluated, the test piece is held between two holding jigs, and a load is applied to the edge portion of the held test piece. It is preferable to evaluate the breaking strength of the edge portion of the semiconductor wafer.
If the load is applied while holding the test piece cut out from the semiconductor wafer in this way, the deflection due to buckling during the load application hardly occurs, and the evaluation accuracy can be improved. Moreover, since a test piece is cut out and evaluated, a plurality of loads can be applied and strength measurements can be performed using a single wafer, which is efficient and low cost.

At this time, it is preferable that the two holding jigs are provided with notches that expose the periphery of the edge portion corresponding to a predetermined crystal orientation.
With such a holding jig, since the periphery of the edge portion to which the load is applied is exposed, the load applying device and the holding jig do not interfere with each other, and the other portions can be held and sandwiched. Deflection can be reliably prevented.

  Further, the method for applying the load is not particularly limited, and accurate fracture strength measurement can be performed if the falling weight impact method or the horizontal static pressure load method is used.

The method of the present invention as described above can be implemented by a semiconductor wafer evaluation apparatus characterized in that a load is applied to an edge portion corresponding to a predetermined crystal orientation of a semiconductor wafer to be evaluated.
Hereinafter, the present invention will be described more specifically.

FIG. 2 is a schematic explanatory view of a falling weight type impact tester that can be used in the present invention.
A single-axis slider robot 9 is erected on the device base 17 of the destructive testing machine 10, an electromagnet 12 is set on the slider 11, and a cylindrical striking pin 13 (chrome steel) is adsorbed by a magnetic force. A polycarbonate cover 14 is installed on the holding jig 15. The test piece 16 is exposed from the hole at the bottom of the cover 14 and is designed so that the falling cylindrical hitting pin 13 collides.

In this destructive testing machine 10, it is possible to apply an impact in the vertical direction to the test piece 16 held by the holding jig 15 by moving the slider 11 up and down and dropping the cylindrical hitting pin 13 from an arbitrary height. is there.
By using this tester 10, it is possible to perform an analysis by the staircase method by mainly performing a “falling weight type impact fracture test using the principle of the staircase method at a constant drop weight”.

The steer case method is a method of statistically analyzing the impact fracture strength based on the presence or absence of fracture by moving the stress level up and down to each level (for example, Dixon, WJ and Mood, AM, J. Amer. Stat. Assn., Vol. 43, pp. 109-126, 1948).
The fracture strength of the semiconductor wafer is calculated by calculating the “50% impact fracture energy (E ₅₀ ), 50% impact fracture energy standard deviation (SE)” from the presence / absence of fracture of the sample and the impact force distribution. To evaluate.

Next, a method for carrying out the present invention will be briefly described with the measuring device of the horizontal static pressure load method shown in FIG.
In the measuring device 20 of the horizontal static pressure load method of FIG. 3, after placing the holding jig 15 holding the test piece 16 on the mounting table 21, the tip of the load shaft 22 is moved to the edge portion of the test piece 16. Then, a static pressure load is applied to the edge portion in the horizontal direction while adjusting the gas from the cylinder 23 with the pressure control valve 24. And it is the method of measuring the breaking strength when this test piece 16 is destroyed.

In the test using the apparatus of FIGS. 2 and 3 described above, in the present invention, the test piece 16 can be produced and held on the holding jig 15 as shown in FIG.
As shown in FIG. 4A, from a semiconductor wafer W to be evaluated, a predetermined crystal orientation to which a load is applied (for example, in the case of a (100) wafer, an angle θ from a {110} cleavage plane θ = 26.565. The test piece 16 including the edge portion corresponding to (°) is cut out with a dicer or the like.

Next, as shown in FIG. 4B, the two holding jigs 15 provided with the notches 18 exposing the edge portions to which the load is applied are sandwiched and held.
And as shown to FIG.4 (c1), (c2), it can install in an apparatus and a load can be provided. As shown in FIGS. 4C1 and 4C2, the edge of the test piece 16 to which the load is applied is held and installed so that a load and an impact are applied in the vertical direction.

Alternatively, as shown in FIG. 5, a fan-shaped test piece 16 ′ can be produced and held by a holding jig 15 ′.
As shown in FIG. 5A, the semiconductor wafer W to be evaluated is divided into four, and a fan-shaped test piece 16 ′ is produced so that the edge corresponding to a predetermined crystal orientation to which a load is applied is in the center.

  Next, as shown in FIG. 5 (b), the test piece 16 'is sandwiched and held by a holding jig 15' provided with a notch 18 ', and FIGS. 5 (c1), 5 (c2), and 6 are used. As shown, a load can be applied to this.

Alternatively, as shown in FIG. 7, the semiconductor wafer W to be evaluated can be held by a holding jig 15 ″ provided with a notch 18 ″ (FIG. 7A), and a load can be applied thereto. (FIG. 7 (b1), (b2)).
Also in this case, as shown in FIG. 7A, the semiconductor wafer W is held so that a load is applied to the edge portion corresponding to the predetermined crystal orientation.

When the load is applied to the semiconductor wafer W without cutting out the test piece as shown in FIG. 7, the mounting table without sandwiching the semiconductor wafer W as shown in FIGS. 8 (a) and 8 (b). It is also possible to apply the load by placing it on 21 and supporting it at two or more points by the support means 25. However, in consideration of deflection due to buckling, it is preferable to apply a load in a state of being held by a holding jig as shown in FIG.
In addition, if the test is performed by cutting the test piece as described above, the number of measurements per wafer can be increased, and the cost can be reduced.

  When the evaluation is performed by selecting the load point regardless of the crystal orientation as in the prior art using the apparatus of FIGS. 2, 3 and 8 as described above, the strength of the semiconductor wafer is caused by the crystal anisotropy. It was difficult to evaluate slight differences in However, in the present invention, by using the holding jig as described above, impact and load are applied to a predetermined crystal orientation of the test piece or wafer, and the fracture strength is measured, thereby greatly improving the sensitivity and accuracy. I can confirm.

  For example, the apparatus and method of Patent Documents 2 to 4 are evaluation methods in which strength = the number of repeated hits. However, considering that a silicon wafer or the like is a brittle material, the evaluation method overestimates the effect of repeated hits, and the evaluation accuracy is considered to be low.

  The above-described evaluation according to the present invention is suitable for evaluation of, for example, a silicon wafer having a thickness of 0.1 to 5 mm. Also, sapphire and SiC crystal wafers used as semiconductor wafers have physical properties such as Young's modulus. The present invention is preferred because of the presence of anisotropy.

  In the present invention as described above, for example, by selecting and evaluating the edge portion and direction that are not affected by the cleavage, the variation in the breaking strength of the edge portion of the semiconductor wafer is reduced. Can be evaluated with high accuracy.

EXAMPLES Hereinafter, although an Example and a comparative example are shown and this invention is demonstrated more concretely, this invention is not limited to these.
Example 1
As a wafer to be evaluated, a low oxygen concentration product (product type A) and a medium oxygen concentration product (product type B) were prepared, and the breaking strength of the edge portion was evaluated by the present invention using the apparatus shown in FIG.
As a low oxygen concentration product (type A), five [100] silicon wafers having a thickness of 0.78 mm, a diameter of 300 mm, a P-type, an oxygen concentration of 12 ppma, and a resistivity of 20 Ωcm were prepared.

Next, as shown in FIG. 4, eight substantially trapezoidal silicon piece samples were produced so as to include an edge portion corresponding to an angle θ = 26.5 ° from the {110} cleaved surface.
40 pieces of silicon piece samples were prepared from 8 wafers each for approximately trapezoidal samples from one wafer.

As a medium oxygen concentration product (product type B), five [100] silicon wafers having a thickness of 0.78 mm, a diameter of 300 mm, a P-type, an oxygen concentration of 14 ppma, and a resistivity of 22 Ωcm were prepared.
Next, a total of 40 pieces of silicon pieces were produced in the same manner as the product A.

"Measurement procedure for impact evaluation"
(1) Determine the basic conditions from the material and thickness of the sample. In Example 1, the size of the falling weight (cylindrical striking pin (bearing steel)) is 8 mm in diameter, 20 mm in length, and 8.6 g in weight, and the falling ball jig, the reference height (100 cm) at the start, Select the change level (10 cm).
(2) The sample is sandwiched and held by a silicon piece holding jig and set on the apparatus base. Also, the cylindrical striking pin is set on the electromagnet.
(3) Set the robot controller to a height of 100 cm and raise the slider. Then, the electromagnet is turned off to drop the cylindrical hitting pin, and an impact force is applied to the edge portion corresponding to the angle θ = 26.5 ° from the {110} cleaved surface of the sample. The destruction is judged by the naked eye for the presence or absence of cracks, fractures or fractures on the surface of the impact site of the sample.
(4) If the sample is not destroyed, replace it with a new sample and set it to raise the drop height by one level (+10 cm).
(5) If the sample is destroyed by impact, replace it with a new sample and set it to lower the drop height by one level (-10 cm).
(6) Then, the drop height condition is selected from the previous test result, and (1)-(5) is repeated 40 times.
(7) Calculate the staircase method from the presence or absence of sample breakage and the impact force distribution.
(8) Calculate “50% impact fracture energy (E ₅₀ ), 50% impact fracture energy standard deviation (SE)”.

The 50% impact fracture energy (E ₅₀ ) is, for example,
E ₅₀ = weight of impact material × gravity acceleration × 50% impact fracture height (H ₅₀ )
It can be calculated from The 50% impact fracture height (H ₅₀ ) is the height at which 50% of the number of tests is estimated to be destroyed.

The impact strength of the low oxygen concentration product (variety A) was as follows.
50% impact fracture energy (E ₅₀ ) = 0.044J
Standard deviation of 50% impact fracture energy (SE) = 0.013J

Moreover, the impact strength of the medium oxygen concentration product (variety B) was as follows.
50% impact fracture energy (E ₅₀ ) = 0.063J
Standard deviation of 50% impact fracture energy (SE) = 0.013J

A significant difference was found in the population mean when the test (significance level 0.05) of the population mean was performed from the distribution of the fracture strength of the low oxygen concentration product (variety A) and the medium oxygen content product (variety B). .
Further, the distribution of fracture strength than a test of normality (x ² goodness of fit test) is considered to be a normal distribution.

(Comparative Example 1)
As wafers to be evaluated, low oxygen concentration products (variety A) and medium oxygen concentration products (variety B) similar to Example 1 were prepared and evaluated using the apparatus shown in FIG.
In this comparative example 1, the angle θ between each wafer and the {110} cleave = any angle (0 °, 10 °, 20 °, 30 °, 40 °, 50 °, 60 °, 70 °, 80 °, (90 °) under 10 types of conditions, a dicer was used to cut each of the four letters into four parts, and 40 fan-shaped samples were produced with 10 wafers × 4 wafers. By measuring these samples at random, a conventional test in which the crystal orientation of the angle at which the load is applied was not specified was reproduced.
As in Example 1, the holding jig used the holding jig 15 ′ shown in FIGS. 5 and 6 to give an impact to the center of the edge portion of the sample to obtain the impact strength.

The impact strength of the low oxygen concentration product (variety A) was as follows.
50% impact fracture energy (E ₅₀ ) = 0.035J
Standard deviation of 50% impact fracture energy (SE) = 0.028J

The impact strength of the medium oxygen concentration product (variety B) was as follows.
50% impact fracture energy (E ₅₀ ) = 0.037J
Standard deviation of 50% impact fracture energy (SE) = 0.023J

From the distribution of the fracture strength of the low oxygen concentration product (variety A) and the medium oxygen concentration product (variety B), when the population average test (significance level 0.05) is performed, there is a significant difference in the population average. There wasn't.
Also, from a test of normality (x ² goodness of fit test), the distribution of breaking strength is skewed, the normal distribution was observed. It is probable that the fracture strength of the sample has varied due to cleaving.

As described above, in Example 1 in which the load application angle of the present invention is limited to a predetermined crystal orientation, the impact strength of the wafer edge portion can be more statistically evaluated. As a result, the effect of improving the evaluation ability was obtained as compared with the conventional inspection machine.
In addition, since the test can be performed with a test piece cut out in a substantially trapezoidal shape, the number of measurement n per wafer can be doubled, and the measurement accuracy can be improved and the cost can be reduced.

(Example 2)
As a wafer to be evaluated, a low oxygen concentration product (type C) was prepared and evaluated according to the present invention using the apparatus shown in FIG.
As a low oxygen concentration product (product type C), five [100] silicon wafers having a thickness of 0.78 mm, a diameter of 300 mm, a P-type, an oxygen concentration of 10 ppma, and a resistivity of 21 Ωcm were prepared.

Next, as shown in FIG. 4, 8 pieces of substantially trapezoidal silicon piece samples are included from one wafer so as to include an edge portion corresponding to an angle θ = 26.5 ° from a {110} cleaved surface. Individually produced.
A total of 40 pieces of silicon pieces were prepared from 8 wafers each x 8 wafers of approximately trapezoidal samples from one wafer.

These samples are sandwiched and held by a silicon piece holding jig, and a load is applied to an edge corresponding to an angle θ = 26.5 ° from the {110} cleaved surface, and a horizontal static pressure load method is used. The test was performed with the apparatus of FIG. The measurement results are shown below.
Mechanical strength: MIN680N to MAX888N
Average mechanical strength = 776 N
Standard deviation of mechanical strength = 67N
Distribution of fracture strength than a test of normality (x ² goodness of fit test) is considered to be a normal distribution.

(Example 3)
As in Example 2, however, as shown in FIG. 8, a load was applied using the wafer as it was while being supported by the support means. Also in this case, a test was performed with the apparatus shown in FIG. 8 by applying a load to the edge portion corresponding to the angle θ = 26.5 ° from the {110} cleaved surface of the wafer and using a horizontal static pressure load method. . The measurement results are shown below.

Mechanical strength: MIN703N to MAX1008N
Average mechanical strength = 806N
Standard deviation of mechanical strength = 105N

From the fracture strength distributions of Examples 2 and 3, when the population average test (significance level 0.05) was performed, no significant difference was found in the population average.
On the other hand, from a test of normality (x ² goodness of fit test), the distribution of fracture strength in Example 3, the normal distribution is skewed was observed.

In Example 3, it is considered that the entire wafer was buckled due to the load, causing a large deflection.
From these points, it is considered that Example 2 in which a load is applied while being held by a holding jig is an excellent evaluation method. In addition, compared to the measurement in Example 3, the number of measurements n can be increased eight times in Example 2, so that the measurement accuracy can be improved and the cost can be reduced.

(Comparative Example 2)
As a wafer to be evaluated, a low oxygen concentration product (type C) was prepared and evaluated using the apparatus shown in FIG.
As a low oxygen concentration product (type C), ten [100] silicon wafers having a thickness of 0.78 mm, a diameter of 300 mm, a P-type, an oxygen concentration of 10 ppma, and a resistivity of 21 Ωcm were prepared.
In this comparative example 2, for each wafer, an angle θ with {110} cleaved = an arbitrary angle (0 ° · 10 ° · 20 ° · 30 ° · 40 ° · 50 ° · 60 ° · 70 ° · 80 ° A conventional test in which the crystal orientation of the angle at which the load is applied is not specified and reproduced by applying a load to 10 positions at 90 ° and measuring the strength was reproduced. The measurement results are shown below.

Mechanical strength: MIN445N to MAX954N
Average mechanical strength = 645N
Standard deviation of mechanical strength = 183N

  Conventionally, the apparatus of Patent Document 1 as shown in FIGS. 3 and 8 does not select load points in consideration of wafer anisotropy. However, when the population average test (significance level 0.05) was performed from the fracture strength distributions of Example 3 and Comparative Example 2, a clear significant difference was observed in the population average. This is considered to be due to wafer anisotropy.

  In this way, in the evaluation of the breaking strength of the edge portion of the semiconductor wafer, the strength of the wafer edge portion is statistically evaluated more finely by selecting the load point in consideration of the anisotropy of the wafer as in the present invention. It became possible to do. Thereby, the effect of the improvement of evaluation capability was acquired compared with the conventional evaluation method.

  The present invention is not limited to the above embodiment. The above-described embodiment is an exemplification, and the present invention has substantially the same configuration as the technical idea described in the claims of the present invention, and any device that exhibits the same function and effect is the present invention. It is included in the technical scope of the invention.

9 ... single axis slider robot, 10 ... falling weight type impact test machine,
11 ... Slider, 12 ... Electromagnet, 13 ... Cylindrical striking pin, 14 ... Cover,
15, 15 ', 15''... holding jig, 16, 16' ... test piece,
17 ... Device base, 18, 18 ', 18''... Notch,
20 ... Measuring device of horizontal hydrostatic load system, 21 ... Mounting table,
22 ... Load shaft, 23 ... Cylinder, 24 ... Pressure control valve,
W: Semiconductor wafer.

Claims (3)

A method for evaluating the breaking strength of an edge portion of a semiconductor wafer,
The crystal orientation in which the breaking strength of the edge portion of the semiconductor wafer to be evaluated is highest is examined in advance, and the falling weight impact method or the horizontal direction is applied only to the edge portion corresponding to the crystal orientation in which the examined breaking strength is highest. A method for evaluating a semiconductor wafer, comprising applying a load by a hydrostatic load method and evaluating a breaking strength of an edge portion of the semiconductor wafer. From the semiconductor wafer to be evaluated, a test piece including an edge portion corresponding to the crystal orientation with the highest fracture strength is cut out and held by sandwiching the test piece with two holding jigs. 2. The semiconductor wafer evaluation method according to claim 1, wherein a load is applied to the edge portion to evaluate a breaking strength of the edge portion of the semiconductor wafer. 3. The semiconductor wafer according to claim 2, wherein the two holding jigs are provided with a notch exposing a periphery of an edge portion corresponding to a crystal orientation in which the fracture strength is highest . 4. Evaluation method.

Images