JP2001011610A - Sputtering target and its manufacture - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a sputtering target having specific resistance comparable to that of pure Cu, increased in electromigration resistance, minimal in the defects of a sputtered film used for the formation of a wiring part in a wafer of a semiconductor device, etc., and capable of improving productive efficiency, and its manufacturing method. SOLUTION: The sputtering target for formation of Cu alloy wiring, excellent in electromigration resistance, has a composition in which 0.02-10 atomic %, in total, of one or more elements selected from the group consisting of Ce, Dy, Er, La, Pr, Sc, Sr, Tb, and Y are added to Cu. This sputtering target for formation of Cu alloy wiring can be manufactured by carrying out the addition of one or more elements selected from the group consisting of Ce, Dy, Er, La, Pr, Sc, Sr, Tb, and Y in the form of hydrides.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a sputter target used for forming a wiring portion in a wafer of a semiconductor device such as an LSI and a method of manufacturing the same, and more particularly, to reducing the wiring portion to a low resistance and without lowering the yield. And a sputtering target capable of significantly improving electromigration resistance and a method of manufacturing the same.

[0002]

2. Description of the Related Art As the degree of integration of LSIs increases, Al
Alternatively, pure Cu wiring having low resistance and excellent electromigration (EM) resistance has been used instead of Al alloy wiring. However, in recent years, as the degree of integration has further increased, the required characteristics have become more severe, and the electromigration resistance of pure Cu has become insufficient.
To improve electromigration resistance, C
It is effective to add an additional element to u.
u has been proposed to add Al or Si (Japanese Unexamined Patent Publication No.
-177117). The present invention has electromigration resistance and is still useful at present. However, the specific resistance of the wiring is pure Cu (1.73 μΩ · c).
m) is 1.5 times or more, and there is a drawback that the specific resistance is too high to substitute for pure Cu especially in applications requiring conductivity.

[0003]

From the above, it can be seen that the present invention has a specific resistance comparable to that of pure Cu, a high electromigration resistance, and a sputtered film used for forming a wiring portion in a wafer of a semiconductor device or the like. Provided are a sputter target having few defects and capable of improving production efficiency, and a method for manufacturing the same.

[0004]

[Means for Solving the Problems] 1 Cu, Ce, D
at least one element selected from the group consisting of y, Er, La, Pr, Sc, Sr, Tb, and Y in a total amount of 0.02 to 10 a
A sputtering target for forming a Cu alloy wiring having excellent electromigration resistance characterized by adding t%, and a reflection intensity of 20 using an ultrasonic flaw detector whose sensitivity is adjusted so that the reflection intensity from a 0.3 mm defect becomes 100%. %, The area ratio of oxide inclusions in the target detected at 0.5% or more is 0.5% or less. The sputtering target 3 Cu according to 1 above, wherein Ce, Dy, Er, La, Pr, Sc , S
A hydride of at least one element selected from the group consisting of r, Tb, and Y is added, and the hydride is used as a target by a melting method or a powder metallurgy method to form a Cu alloy wiring excellent in electromigration resistance. Method of manufacturing sputtering target for 4 、 The composition of the target is Ce, Dy, Er, La, P
At least one element selected from the group consisting of r, Sc, Sr, Tb, and Y contains a total of 0.02 to 10 at%, and is a copper alloy or a copper alloy sintered body composed of unavoidable impurities and Cu. 5. The method for producing a sputtering target according to the above 3, wherein the target detected at a reflection intensity of 20% or more by an ultrasonic flaw detector whose sensitivity is adjusted so that the reflection intensity from a 0.3 mm defect becomes 100%. 5. The method for producing a sputtering target according to the above item 3 or 4, wherein the area ratio of the oxide inclusions is 0.5% or less.

[0005]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, the present invention will be described in detail. In order to improve the electromigration resistance, it is effective to reduce the diffusion speed at the crystal grain boundary, which is a portion where the diffusion speed of atoms is high, and at the wiring surface. In addition, the wiring surface of the Cu wiring is covered with a barrier layer such as Ta or TaN provided for preventing diffusion of Cu to the insulating portion, so that surface diffusion is effectively suppressed. On the other hand, in order to suppress the diffusion of crystal grain boundaries as described above, it is effective to add an additive element to Cu to form precipitates at the grain boundaries. However, for Cu, Al, S
When an element such as i is added, the electromigration resistance is improved, but the specific resistance is also increased. Therefore, it is not suitable for practical use when conductivity is required. Therefore, even in the case where a precipitate is formed at the grain boundary by the added element to improve the electromigration resistance, a large increase in specific resistance must not be caused by this, and it must be effectively suppressed. It came to the knowledge that it had to be done.

As a result of the study, the present inventors have found that the elements (Ce, Dy, E) forming an intermetallic compound having a low solid solubility in Cu.
r, La, Pr, Sc, Sr, Tb, Y) was found to be extremely effective. As a result of further studies, it has been found that the inclusion of at least one element selected from these elements in a total of 0.02 to 10 at% copper alloy improves the electromigration resistance and increases the specific resistance. It turned out to be suppressed. When the total amount of the added elements is less than 0.02 at%, sufficient electromigration resistance cannot be obtained, and when it exceeds 10 at%, the specific resistance becomes high, which is suitable for practical use.

Generally, when an alloy film is formed, CVD is performed.
Method, MBE method or a sputtering method using an alloy target is used. Particularly, in a semiconductor process, the sputtering method is mainly used because the handling, maintenance, and throughput are good. I have. However, in this sputtering method, Ce, Dy, Er, La, Pr, Sc, S
at least one element selected from the group consisting of r, Tb and Y
In addition, when a Cu alloy target was prepared and sputtered, there was a new problem that many sputter-shaped defects called water drops were formed in the sputtered film.

This significantly impairs the properties of the film. For this reason, the cause of Splats was investigated. As a result, since these additional elements are very easily oxidized, when these additional elements are simply added to Cu, regardless of a production method such as a powder metallurgy method or a melting method, the obtained Cu alloy target is It was found that the particles contained many oxide particles of the additional element. Since these oxides are insulators, they are charged during sputtering, and when the charged amount exceeds a certain value, a discharge occurs, and the surroundings of the oxides are locally heated, melted, and scattered, and a sputtered film is formed. Was found to form splats. Therefore, it is absolutely necessary to reduce oxides in the target and reduce defects in the sputtered film.

Therefore, the production conditions of the Cu alloy target were examined, and these elements, namely, Ce, Dy, E
When adding at least one element selected from the group consisting of r, La, Pr, Sc, Sr, Tb, and Y to Cu, a hydride of these elements is added, and a target is formed by a melting method or a powder metallurgy method. As a result, almost no oxide particles of the added element were detected in the Cu alloy target. Further, when sputtering was performed using this target, the splats of the sputtered film were significantly reduced. Then, the amount of oxide in the ingot and in the target can be evaluated by the ultrasonic flaw detector, and the reflection intensity from the ultrasonic flaw detector adjusted to have a reflection intensity from a 0.3 mm defect of 100% is 20% or more. By reducing the area ratio of oxide inclusions in the ingot or target to be detected to 0.5% or less, it is possible to reduce oxides in the sputtering target and reduce splats and other defects in the sputtered film. . If the area ratio of the oxide inclusions in the target exceeds 0.5%, the amount of oxides in the sputtering target increases, and splats and other defects in the sputtered film increase.

[0010]

Next, the present invention will be described based on examples and comparative examples. Note that these embodiments show preferred examples to facilitate understanding, and the present invention is not limited by these examples. That is, other embodiments and other examples and modifications within the scope of the technical idea of the present invention are naturally included in the present invention.

(Example 1) C was placed in a graphite crucible.
A Cu ingot weighed so as to have u-0.03 at% Dy and a hydrogenated Dy powder were put therein, melted and cast in vacuum. Analysis of the composition of the Cu-Dy alloy ingot thus formed revealed that the target was Cu-0.03 at% Dy.
Had become. Further, the area ratio of oxide inclusions in the target, which was detected at a reflection intensity of 20% or more by an ultrasonic flaw detector whose sensitivity was adjusted so that the reflection intensity from a 0.3 mm defect became 100%, was measured. ,
It was less than 0.1%. Further, this ingot was processed as a target, and a sputter film was formed on an 8-inch wafer to obtain a Cu-Dy alloy film having a thickness of 0.2 μm. The sputtering conditions are as follows. (Sputtering conditions) Target φ12.98 ″ (for φ8 ″ wafer) Input power 2 kW Ar pressure 1 Pa Substrate temperature Room temperature Film pressure 0.2 μm Substrate φ8 ″ thermally oxidized silicon substrate The average number of splats on the wafer after sputtering is 0.
The number was less than 05. The specific resistance of this film is 1.77.
μΩ · cm. In addition, wiring of this film
When an electromigration test was performed, it was confirmed that the average rupture time was about 50% longer than that of the pure copper wiring having the same shape, and the electromigration resistance was improved. Table 1 shows the results.

[0012]

[Table 1]

(Example 2) C was placed in a graphite crucible.
A Cu ingot weighed so as to have u-15 at% Dy and a hydrogenated Dy powder were charged and melted and cast in a vacuum.
The composition of the Cu-Dy alloy ingot is Cu-15.2a
t% Dy. This ingot was subjected to the same ultrasonic flaw detection test as in Example 1, and the area ratio of the oxide in the target was measured. As a result, the area ratio was 0.2%.
This ingot was processed into a target under the same conditions as in Example 1, and a sputter film was formed on an 8-inch wafer to obtain a Cu-Dy alloy film having a thickness of 0.2 μm. The average number of splats on this alloy film formation was 0.05. The specific resistance of this film was 2.5 μΩ · cm. In addition, when an electromigration test was performed on a wiring using this film, it was confirmed that the average rupture time was about five times longer than that of the pure copper wiring having the same shape, and the electromigration resistance was improved. These results are also shown in Table 1.

(Example 3) C was placed in a graphite crucible.
A Cu ingot weighed so as to obtain u-0.5 at% Er and a hydrogenated Er powder were put therein and melted and cast in a vacuum. The composition of the Cu-Er alloy ingot is Cu-0.5
It was 2 at% Er. This ingot was subjected to the same ultrasonic flaw detection test as in Example 1, and the area ratio of the oxide in the target was measured. As a result, the area ratio was 0.2%. Further, under the same conditions as in Example 1, this ingot was processed as a target, and a sputter film was formed on an 8-inch wafer to obtain a Cu-Er alloy film having a thickness of 0.2 μm.
The average number of splats on the wafer after sputtering is 0.
The number was less than 05. The specific resistance of this film is 1.8.
It was 2 μΩ · cm. Furthermore, when an electromigration test was performed on a wiring using this film, it was confirmed that the average rupture time was about 60% longer than that of a pure copper wiring having the same shape, and the electromigration resistance was greatly improved. Table 1 shows the results. In this example, Dy and E were added as additional elements.
Although only r was shown, similar properties and effects were obtained with other alloy elements included in the present invention.

Comparative Example 1 In a graphite crucible, C was added.
A Cu ingot weighed to become u-0.01 at% Dy and a hydrogenated Dy powder were put therein and melted and cast in a vacuum. The composition of the Cu-Dy ingot is Cu-0.01a.
t% Dy. This ingot was subjected to the same ultrasonic flaw detection test as in Example 1, and the area ratio of the oxide in the target was measured. As a result, the area ratio was less than 0.1%. Furthermore, processing this ingot as a target,
Sputter deposition on an 8-inch wafer, thickness 0.2μ
m-Cu-Dy alloy film was obtained. The average splats number on the wafer after sputtering was 0.05. And
The specific resistance of this film was 1.75 μΩ · cm. Also,
When an electromigration test of a wiring using this film was performed, it was not confirmed that the average rupture time was superior to a pure copper wiring having the same shape. Thus, it was found that the Cu-0.01 at% Dy alloy target less than the composition of the present invention had good average Splats number on the wafer and specific resistance of the film, but had poor electromigration resistance. These results are shown in Table 1 in comparison with the examples.

Comparative Example 2 C was placed in a graphite crucible.
A Cu ingot and a hydrogenated Dy powder were weighed so as to obtain u-23 at% Dy, and were melted and cast in a vacuum.
The composition of the Cu-Dy alloy ingot is Cu-24.5a
t% Dy. The ingot was subjected to the same ultrasonic inspection test as in Example 1 and the area ratio of the oxide in the target was measured. As a result, the area ratio was less than 0.5%. Furthermore, processing this ingot as a target,
Sputter deposition on an 8-inch wafer, thickness 0.2μ
m-Cu-Dy alloy film was obtained. The average number of splats on the wafer after sputtering was 0.05. The average rupture time is about 10 times longer than that of pure copper wiring of the same shape, and it has been confirmed that the electromigration resistance is improved. However, the specific resistance of this film is as high as 3.5 μΩ · cm, which is suitable for practical use. Was not. These results
Table 1 shows a comparison with the examples.

Comparative Example 3 C was placed in a graphite crucible.
A Cu ingot weighed so as to be u-5 at% Dy and a small lump of metal Dy were put and melted and cast in vacuum. C
The composition of the u-Dy alloy ingot is Cu-4.5 at%.
Dy. This ingot was subjected to the same ultrasonic flaw detection test as in Example 1, and the area ratio of the oxide in the target was measured. As a result, the area ratio was 0.6%. Further, this ingot is processed into a target, and a sputter film is formed on an 8-inch wafer, and a Cu film having a thickness of 0.2 μm is formed.
-A Dy alloy film was obtained. As a result, it was confirmed that the average rupture time was about six times longer than that of the pure copper wiring having the same shape, and the electromigration resistance was improved. Further, the specific resistance of this film was 1.92 μΩ · cm, but the wafer was The average number of splats above was very large at 12, and could not be put to practical use. These results are shown in Table 1 in comparison with the examples.

[0018]

According to the present invention, Cu, Ce, Dy, Er, La, P
By adding at least one element selected from the group consisting of r, Sc, Sr, Tb, and Y in a total amount of 0.02 to 10 at%, Cu having excellent electromigration resistance can be obtained.
An object of the present invention is to provide a sputtering target for forming an alloy wiring, and by using the wiring according to the present invention in a wiring portion in a wafer of a semiconductor device such as an LSI,
The resistance to electromigration can be greatly improved without minimizing the increase in the resistance of the wiring portion due to alloying and without reducing the yield.
When one or more elements selected from the group consisting of y, Er, La, Pr, Sc, Sr, Tb, and Y are added, an oxide generated in the copper alloy target by adding a hydride of the element. Prevents the generation of grains, thereby causing the charge during sputtering caused by the oxide being an insulator, the discharge caused by the charge amount exceeding a certain value, and the local heating, melting and melting of the oxide. It has an excellent effect of suppressing formation of splats on a sputtered film due to scattering.

Claims (5)

[Claims] 1. Cu, Ce, Dy, Er, La, P
A sputtering target for forming a Cu alloy wiring excellent in electromigration resistance, wherein at least one element selected from the group consisting of r, Sc, Sr, Tb, and Y is added in a total amount of 0.02 to 10 at%. . 2. The reflection intensity from a 0.3 mm defect is 100.
2. The area ratio of oxide inclusions in the target detected at a reflection intensity of 20% or more by an ultrasonic flaw detector whose sensitivity has been adjusted to be 0.5% or less is 0.5% or less.
The sputtering target according to the above. 3. Cu, Ce, Dy, Er, La, P
Excellent in electromigration resistance, characterized by adding a hydride of at least one element selected from the group consisting of r, Sc, Sr, Tb, and Y and using the hydride as a target by a melting method or a powder metallurgy method. A method for manufacturing a sputtering target for forming a Cu alloy wiring. 4. The composition of the target is Ce, Dy, E
at least one element selected from the group consisting of r, La, Pr, Sc, Sr, Tb, and Y contains 0.02 to 10 at% in total, and a copper alloy or copper alloy sintered consisting of unavoidable impurities and Cu as a balance; The method for producing a sputtering target according to claim 3, wherein the sputtering target is a union. 5. The reflection intensity from a 0.3 mm defect is 100.
%, Wherein the area ratio of oxide inclusions in the target detected at a reflection intensity of 20% or more by an ultrasonic flaw detector adjusted to have a sensitivity of 0.5% or less is 0.5% or less.
Or the manufacturing method of the sputtering target of 4.