JPH02115364A - Tellurium target and production thereof - Google Patents

Abstract

PURPOSE:To obtain a Te target for sputtering having improved cracking resistance by forming a target with Te having a specified purity or above and a specified relative density or above and making the density of the Te target as uniform as possible over the entire surface. CONSTITUTION:A target is formed with Te having >=99.99% purity and >=95% relative density and the density of the Te target is made as uniform as possible over he entire surface so that the occurrence of stress due to the target material itself can be prevented. In order to evaluate the uniformity of the density of the target material, measured values of hardness are used as substitutive characteristics of density from the correlation between hardness and density. The Vickers hardness Hv of the Te target material is measured at many points over the entire surface and the standard deviation sigma of the measured values is regulated to <=10% of the mean value. A Te target having improved cracking resistance is obtd.

Description

[Detailed Description of the Invention] [Industrial Application Field] The present invention relates to a target material for sputtering, and in particular, a sintered target material made of metallic tellurium suitable for use as a vapor deposition source for depositing a metallic tellurium thin film on the surface of a substrate. and its manufacturing method.

[Prior Art] Conventionally, various metals or alloys have been selected and used as sputtering target materials depending on the purpose. The use of metallic tellurium and tellurium alloys as target materials for sputtering is disclosed, for example, in Japanese Patent Application Laid-open No. 1982-
No. 13589 "Sintered target for sputtering made of rTe or Te alloy".

Te and Te alloy targets are very brittle and therefore very difficult to form by casting methods, and are usually produced by powder metallurgy. The target thus produced is incorporated into a sputtering device and put into practical use; however, during sputtering, the target generates heat when power is applied, and its surface is constantly exposed to argon plasma and in a high-energy atmosphere. Under such severe usage conditions, the target material must be used while preventing cracks, abnormal discharge, etc.

[Problems to be Solved by the Invention] However, existing targets are not yet satisfactory enough to withstand the above-mentioned usage conditions, and cracks and melts occur particularly when the input power is increased.

In the past, various studies have been carried out to overcome the above-mentioned drawbacks, and in the aforementioned Japanese Patent Application Laid-Open No. 2002-120003, the thermal conductivity and strength of the target material can be improved by reducing the oxygen concentration contained in the target material to 1100 pp or less. It is stated that the problem could be improved and the problem could be solved to some extent.

However, the maximum output during sputtering of the target disclosed in JP-A-Sho is 3.5 W/cdS in pure Tc.
Even at Tc -50%sb, it is only 4.4 W/cd, which is by no means satisfactory.

Therefore, it is an object of the present invention to develop a tellurium target with improved performance.

[Means for Solving the Problems] All research by the present inventors was conducted on a sintered body of metallic tellurium with a purity of 99.99% or more. therefore,
The condition that the oxygen concentration is below looppm is automatically met. Nevertheless, some Te metal sintered products have a maximum output of only about 3.2 W/c- during sputtering, and will be destroyed if more power is applied. The product had individual target materials that were not destroyed at all even when the sputtering output was 5.0 W/cj. This suggests that there are other important factors besides oxygen concentration that affect the strength and stability of the target material. In order to find out what those factors are,
Our research progressed.

In proceeding with the research, the following considerations were first considered. That is, the target material is usually subjected to surface grinding after pressure forming and sintering, and the surface portion contaminated by the forming and sintering jig is removed, and finally it is joined to the backing plate. In these steps, the molded product is constantly affected by grinding acid, adhesive, heated brazing material, flux, etc., and if there are pores in the molded product, those portions are impregnated. It is also susceptible to oxidation, and oxidation causes an increase in electrical resistance, which is undesirable.

Therefore, it is thought that the performance of target materials can be improved by creating products that are less susceptible to these impregnations and oxidations. To achieve this, it is necessary to increase the density of the sintered body to achieve the true density of metallic tellurium (6
, 24). This is because the fact that the relative density to the true density of sintered Te (expressed as a percentage of the true density) is small is considered to mean that there are voids inside the sintered body. be.

Next, the inventors predicted that the density unevenness of the target material would have a large influence on the destruction of the target material, and so the density of the target material was adjusted over the entire flat target material.
We came up with the idea that if we make the density as uniform as possible, the stress caused by the target material itself will be eliminated and the target material will be less likely to be destroyed.Based on this idea, we took measures to make the density uniform. did.

By the way, in order to know whether the density is uniform over a wide area of the target material, it is necessary to know a method of easily measuring the density in each local area. However, since there is no suitable method for easily measuring local density, we considered using surface hardness measurements as a proxy for density. In order to examine the i'iJ or not, we created Tc molded bodies with various densities, measured the hardness of each, and investigated the relationship between density and hardness, and found the results shown in Figure 2. Obtained.

In other words, it was confirmed that there is a strong positive correlation between the relative density of the target material and Vickers hardness, and it was found that the measured value of hardness can be used as a proxy characteristic for density as a means of uniformizing the density. .

Although the present inventors used Vickers hardness H as the hardness, it is clear in principle that hardness measured by other methods may be used.

First, in order to investigate the relationship between hardness, impregnation, and oxidation using hardness (H) as a proxy property for density, we attempted to impregnate each test piece with wax, and after surface grinding, we tested it in vacuum. The pieces were heated to evaporate the Te and examined for wax impregnation. As a result, no impregnation was observed in the test pieces with hardness H=32 kg/mJ or more, and a lot of impregnation and oxides were observed in the test pieces with hardness H-29 kg/- or less. Oxide was observed in wax impregnation in 15
Oxidation seems to have progressed because it was carried out in air at temperatures between 0'' and 20G°C.For H-30kg/- and above, internal oxidation did not progress despite similar treatment, making the molded products extremely stable. It turned out to be.

The above experimental results demonstrate that Te targets of H-30 kg/- or more, and even H-32 kg/- or more, have a dense structure with sufficient resistance to impregnation and oxidation. It shows.

The relative density corresponding to each hardness is 94 from Figure 2.
, 7% and 9B, 8%. However, as will be explained later, targets with a density of 95% or more or 97% or more are not always the most difficult to destroy, and even if you limit your investigation to the range with a density of 95% or more, It was found that the magnitude of the damage does not necessarily correspond to the difficulty of destruction.

However, if the density is 95% or more or even 97% or more (or H-30kg/- or H-v
v32kg
It can be concluded from the experimental results of impregnation and oxidation resistance that a hardness of /- is one of the major prerequisites for TQ Tutat to be stable against fracture stress. Good morning.

Once such conditions are met, what further provisions should be made?
The inventors further researched the question of whether or not one should be satisfied and discovered the following.

That is, as a result of detailed research in a density range greater than 95.0%, it was found that the fracture resistance of a target during sputtering is most influenced by whether or not the target density is uniform over the entire target surface; It was found that there is a strong correlation with the value of σ, which represents the magnitude of variation in the measured values of Vickers hardness H, which the inventors adopted as a proxy characteristic. That is, it has been found that by controlling the value of σ to be below a certain value, it is possible to manufacture a Tc metal target with extremely excellent fracture resistance. Furthermore, one of the major factors that influences the value of flatness is the filling method when molding the powder. It was found that the filling method did not provide sufficient density uniformity.

That is, in general, when filling the powder 3 into a pressure molding mold consisting of the die 1, the lower bunch 2, and the upper bunch (not shown) (Fig. 3a), if the powder 3 is filled without particular care, for example, the third
As shown in figure b, the part A in the mold where the powder was initially introduced and the part B filled by operations such as scraping after being put into the mold appear uniform visually, but in reality the filling is They have different densities, and if they are compressed using a spindle or two-wheel press, density unevenness will occur due to the filling method.

Sintering in this state results in a sintered body with defects such as deformation, cracks, and uneven density. Even when hot pressing is used, a sintered body with similar defects can be obtained. Products that are deformed or cracked can be easily removed during the manufacturing process, but
Because non-destructive testing is difficult for density unevenness, no special measures were taken, and if there were no problems with the appearance, no particular attention was paid. Furthermore, similar defects may occur due to other factors such as unevenness in sintering temperature with respect to density unevenness. When a target material with such defects is used under severe usage conditions, it generates heat, causing uneven heat generation in areas with different densities, and various stresses resulting from this. The Te target, which originally has low mechanical strength and a low melting point, is destroyed or partially melted due to such stress and uneven heat generation. Therefore, if we pay attention to the above points and devise and improve the process, and in particular, fill the metal powder into the pressure molding mold in a manner that ensures extremely high homogeneous filling, It was confirmed that the variation in density could be made extremely small, and as a result, it was possible to manufacture a Te target that would not cause cracks even when sputtered at a very high output.

The present invention will be explained in more detail with reference to Examples below.

Example 1 A Te target of 5'φ was prepared, sputtering was performed using an Rr magnetron sputtering device, and the breakdown power of the target was investigated by varying the amount of input power, and the results shown in Table 1 below were obtained.

The hardness was measured by pressing a diamond indenter against each sample of the target with a constant load, measuring the size of the indentation, and determining the Vickers hardness (H) using a conversion table.

Average value of H measurement value or direct V
H was measured at one location per cd per target with a V diameter of 127 mm (5 inches), and the number of measurements was calculated as n.

Next, based on the results in Table 1, in order to test the correlation between density and breakdown power, we rearranged the data in descending order of density. The results are shown in Table 2.

The results in Table 2 show that there is no particular correlation between density and breakdown power over the density range tested.

Next, based on the results in Table 1, in order to examine the correlation between the variation in the measured values of H (σ) and the breakdown power (Wb),
When we rearranged the data in descending order of variation (σ), we obtained Table 3.

The results in Table 3 give the impression that there is some kind of correlation, although not clear, between the variation in H measurement values (standard V deviation) σ and the breakdown power (W).

Therefore, for each sample, consider a value representing the percentage of the average value of H that the variation σ corresponds to, that is, σ/nuX100, and in order to test the correlation between this and the breakdown power, σ/ When I rearranged the data in descending order of the value of , it became as shown in Table 4.

The results in Table 4 clearly show a tendency for the value of breakdown power to increase as the value of σ/or decreases.

If this result is plotted on orthogonal coordinates, it will look like Figure 1.

From Fig. 1, when σ/X is 100-10.0, the breakdown power is 3.7 W/c-, and a/RX 100<
10.0, the breakdown power is definitely 3.7 W/
It can be seen that it can be made larger than c-. As a trial, when σ/X is 100-84, the breakdown power W is 4. OW/c-
So, when σ/also×100-7.6, Wb-4.5W
/c-, and when σ/RX100-6.4, W, -5
, QW/c- or more can be predicted with great certainty. Therefore, if the manufacturing process is controlled and the density of the product target plate surface is averaged so that the value of σ/or It is believed that it is possible to produce a Te target that can be used in output.

In addition, although the above explained the metallic tellurium target,
It is clear that essentially the same principle of the invention can be applied to the tellurium alloy target as well, with the only difference being the specific value of H.

[Effects of the Invention] Based on the idea that if the density of the target material is at least 95% and the density is made as uniform as possible over the entire plate surface of the target, stress caused by the target material itself will be eliminated. By aiming to equalize the density of the target material and focusing on the strong correlation between hardness and density as an evaluation method, we developed Te with improved crack resistance by using the hardness DI fixed value as a substitute characteristic for density.
I succeeded in creating a target. The target material obtained in this way has a hardness distribution of, for example, H-33,
[i±3.7 (kg/mj) target material is 3.5W
/cd cracks occurred, whereas H-32,6±2
．． 1 (kg/a+J), no cracks occur even at an output of 5 W/cd, even though the density of the target material is much lower than that of the former.

In other words, we recognized the importance of uniform density over the entire surface of the product plate when manufacturing Tc targets, confirmed that hardness can be used as a surrogate property for density to achieve this, and developed Te targets with improved crack resistance. We were able to make a significant contribution to manufacturing purposes.

[Brief explanation of the drawing]

Figure 1 is a graph showing the relationship between the percentage of the standard deviation σ of the measured values (σ/or It is. FIG. 2 is a graph (1) showing the relationship between relative density (%) and Vickers hardness (H) of a sintered Te target. FIG. 3a is a schematic cross-sectional view showing the state immediately after metal powder is charged into a mold for the purpose of powder compaction by a conventional method. Fig. 3b is a schematic cross-sectional view for explaining the difference in the packing density of the powder depending on the position in the mold in the state shown in Fig. 3a in which the powder is smoothed so that the top surface is horizontal by cutting. . The symbols in the figure represent the following, respectively. 1... Dice 2... Lower punch 3...
metal powder

Claims (3)

[Claims] (1) The purity of the metallic tellurium is 99.99% or more, the relative density is 95% or more, and as a proxy characteristic of the relative density, a large number of measurements of Vickers hardness H_v are measured uniformly over the entire surface of the target material. The standard deviation σ of the values is
A sintered metal target for sputtering made of tellurium metal, characterized in that H_v is less than a selected value of 10.0% or less of the average measured value @x@. (2) The average value of a large number of measured values of Vickers hardness H_v uniformly measured over the entire surface of the target material is
99.99% metal content, characterized in that @x@ = 30.0 (kg/mm^2) or more, and the standard deviation thereof is σ = 2.5 (kg/mm^2) or less A sintered metal target for sputtering made of the metal tellurium mentioned above. (3) In the manufacturing process of the method of manufacturing a sintered metal target by compacting and sintering metallic tellurium powder, the operation of filling the metal powder into a pressure molding mold is performed with extremely high level homogeneous filling. A sample of the sintered metal obtained by sintering the powder compact obtained by powder compacting is taken, and the Vickers is uniformly distributed over at least one surface of the sample. Measure hardness H_v, H_
The average value of the measured values of v is @x@=30.0 (kg/mm^
2) or more, and its standard deviation is σ = 2.5 (kg
/mm^2) A method for manufacturing a sintered metal target for sputtering made of tellurium metal, characterized in that the manufacturing process is controlled so that the following conditions are satisfied: