This invention relates to the processing of high-purity tantalum to produce a sputtering target with a microstructure that is desirable for uniform sputtering. In particular, the invention relates to the manufacture of high-purity tantalum with a mean grain size of less than 100 μm and a uniform, predominately (111)<uvw> crystallographic texture throughout the target thickness.
Tantalum is currently used extensively in the electronics industry, which employs tantalum in the manufacture of highly effective electronic capacitors. Its use is mainly attributed to the strong and stable dielectric properties of the oxide film on the anodized metal. Both wrought thin foils and powders are used to manufacture bulk capacitors. In addition, thin film capacitors for microcircuit applications are formed by anodization of tantalum films, which are normally produced by sputtering. Tantalum is also sputtered in an Ar—N2 ambient to form an ultra thin TaN layer which is used as a diffusion barrier between a Cu layer and a silicon substrate in new generation chips to ensure that the cross section of the interconnects can make use of the high conductivity properties of Cu. It is reported that the microstructure and stoichiometry of the TaN film are, unlike TiN, relatively insensitive to the deposition conditions. Therefore, TaN is considered a much better diffusion barrier than TiN for chip manufacture using copper as metallization material. For these thin film applications in the microelectronics industry, high-purity tantalum sputtering targets are needed.
The typical tantalum target manufacturing process includes electron-beam (EB) melting ingot, forging/rolling ingot into billet, surface machining billet, cutting billet into pieces, forging and rolling the pieces into blanks, annealing blanks, final finishing and bonding to backing plates. The texture in tantalum plate is very dependent on processing mechanisms and temperatures. According to Clark et al. in the publication entitled “Effect of Processing Variables on Texture and Texture Gradients in Tantalum” (Metallurgical Transactions A, September 1991), the texture expected to develop in cold-rolled and annealed bcc metals and alloys consists of orientations centered about the ideal orientations, {001}<110>, {112}<110>, {111}<110>, and {111}<112>. Generally conventionally processed tantalum is forged or rolled from ingot to final thickness, with only one (1) or no intermediate annealing stages. A final anneal is usually applied to the plate simply to recrystallize the material. The direction of the deformation influences the strengths of resulting annealed textures but generally little attention is given to the resulting distribution of textures. In conventionally processed tantalum, significant texture variation exists in the cross-section of the plate, as described by Clark et al. (August 1992), Raabe et al. (1994), Michaluk (1996). Typically the above mentioned textures exist in stratified bands through the thickness of the rolled plate, or form a gradient of one texture on the surface usually {100}<uvw>, with a gradual transition to a different texture at the centerline of the plate, usually {111}<uvw>, Wright et al. (1994). Another cause of texture variation through the target thickness is the non-uniformity of the deformation processes used to form the plate. Texture non-uniformity results in variable sputter deposition rates and sputter surface irregularities, which in turn is believed to be a source of micro-arcing. Micro-arcing is believed to believed to be the principle cause of particle generation and is thus undesirable in the semiconductor industry.
In
In accordance with the present invention there is provided a processing route for producing high purity tantalum sputtering targets with a mean fine grain size of less than 100 microns and uniform crystallographic texture throughout the target thickness.
The method comprises forging, rolling and annealing high-purity, vacuum-melted tantalum ingots in such a way as to eliminate remnant as-cast grain structure, and produce a homogeneous fine-grain size (mean <100 μm) microstructure with a uniform, predominately {111}<uvw> texture throughout the thickness of the target. Significant sputtering problems have been reported when the texture of the target is not uniform throughout the target thickness. Sputtering rates and film deposition rates change as a function of target crystallographic texture. This variable sputter rate across a target surface causes film thickness uniformity problems and also produces unwanted surface topography in the form of ‘ridging’, which in turn is believed to cause micro-arcing.
This invention uses a series of deformation techniques, with a minimum of three (3) intermediate, high-temperature inert-atmosphere anneals, preferably vacuum, to produce a combination of fine-grain size (mean <100 μm) tantalum targets with a uniform, predominately {111}<uvw> texture throughout the target thickness, until now unseen in the industry today. ‘Uniform texture throughout the target thickness’ refers to a homogeneous distribution of textural components with no visible banding at a resolution of 20× from the target surface to at least mid-thickness. ‘Inert’ refers to an atmosphere that is non-reactive with the tantalum-comprising mass.
Experiments associated with this invention also revealed that by controlling the annealing temperature the most desirable texture for collimated sputtering, the (111) texture, can be generated. The (111) texture is the only texture that has one of the close-packed directions aligned normal to the target surface. This direction is a dominant emission direction and is, therefore, the texture required for collimated sputtering.
The high-purity tantalum material of the present invention is preferably 3N5 (99.95%) pure and comprises less than 500 ppm total metallic impurities, excluding gases. The methods of chemical analysis used to derive the chemical descriptions set forth herein are the methods known as glow discharge mass spectroscopy (GDMS) for metallic elements and LECO gas analyzer for non metallic elements.
For the purposes of this invention, the term “sputtering target” covers not only sputtering targets in the traditional sense, but also any other component within the sputtering chamber that is likely to sputter.
FIG. 9(a) is a photograph of the experimental sputtering targets manufactured by the conventional method (Process 4).
FIG. 9(b) is a photograph of an experimental sputtering target manufactured by the process of the present invention (Process 12).
Electron beam (EB), Vacuum Arc Melted (VAR) or other vacuum melted tantalum ingots are deformed perpendicular to the ingot centerline to break up the as-cast grain microstructure. This deformation can be forging, rolling or extrusion whereby significant cross-sectional area or thickness reduction takes place. The reduction in cross-sectional area should be greater than a reduction ratio of 3:1 (cross-sectional area of ingot to cross-sectional area of the forged billet), or equivalent to no less than about 40% strain reduction from starting thickness to final thickness. The forged billet should then be annealed in an inert atmosphere, preferably vacuum, at a high temperature (1500 F.-2800 F.), preferably between 2200 F. and 2400 F., in order to achieve a recrystallized microstructure. The resulting billet/plate is then deformed no less than an additional 35%, preferably 45-65%, of its thickness and subjected to a second high-temperature inert atmosphere anneal between 1500 F. and 2800 F., preferably between 2200 F. and 2400 F. The process of the present invention comprises an additional deformation step with a strain >60% followed by a final inert-atmosphere anneal (1500 F.-2800 F.) to recrystallize the microstructure to the desired fine grain size.
Twelve high-purity tantalum ingots were processed according to conventional methods or the process of this invention. The parameters for each experiment and the corresponding grain size and texture results are summarized in Table 1. Texture uniformity was measured by cutting samples from the target and analyzing them using an EBSD system on a scanning electron microscope (SEM). The mapped area was 7 mm×7 mm and was measured from the target surface to at least the plate mid-thickness. The lighter areas depict {111}<uvw> textures and the darker areas depict {100}<uvw> textures.
The ingots processed by conventional methods (Processes 1 through 7) exhibited a banded microstructure in both grain size and texture.
Although the experimental data shows the grain size results to be less than about 50 μm it is expected that a grain size of less than 100 μm will produce similar sputtering results, so long as the texture is uniform throughout the target thickness.
Sputter trials were conducted on a conventional high-purity tantalum target and a target processed according to this invention in order to compare the sputtering characteristics. FIG. 9(a) and FIG. 9(b) are photographs of the used conventional and invented targets, respectively. The conventional target exhibits extensive surface roughness which is associated with non-uniform sputtering. This surface ‘ridging’ in turn increases the likelihood of micro-arcing and sputter film non-uniformity. In contrast, the target processed according to this invention exhibits a smooth evenly-sputtered surface.
This application resulted from a divisional application of U.S. patent application Ser. No. 09/497,079, filed on Feb. 2, 2000 now U.S. Pat. No. 6,331,233.
Number | Name | Date | Kind |
---|---|---|---|
3268328 | Torti, Jr. | Aug 1966 | A |
3497402 | Douglass | Feb 1970 | A |
3616282 | Bodway | Oct 1971 | A |
3653981 | Watanabe et al. | Apr 1972 | A |
3849212 | Thornburg | Nov 1974 | A |
4000055 | Kumagai | Dec 1976 | A |
4020222 | Kausche et al. | Apr 1977 | A |
4374717 | Drauglis et al. | Feb 1983 | A |
4466940 | Siewert et al. | Aug 1984 | A |
4517032 | Goto et al. | May 1985 | A |
4525417 | Dimigen et al. | Jun 1985 | A |
4619695 | Oikawa et al. | Oct 1986 | A |
4663120 | Parent et al. | May 1987 | A |
4762558 | German et al. | Aug 1988 | A |
4842706 | Fukasawa et al. | Jun 1989 | A |
4844746 | Hormann et al. | Jul 1989 | A |
4883721 | Nalepka et al. | Nov 1989 | A |
4889745 | Sata | Dec 1989 | A |
4960163 | Fang et al. | Oct 1990 | A |
5074907 | Amato et al. | Dec 1991 | A |
5087297 | Pouliquen | Feb 1992 | A |
5171379 | Kumar et al. | Dec 1992 | A |
5194101 | Worcester et al. | Mar 1993 | A |
5231306 | Meikle | Jul 1993 | A |
5282946 | Kinoshita et al. | Feb 1994 | A |
5330701 | Shaw et al. | Jul 1994 | A |
5400633 | Segal et al. | Mar 1995 | A |
5413650 | Jarrett et al. | May 1995 | A |
5415829 | Ohhashi et al. | May 1995 | A |
5418071 | Satou et al. | May 1995 | A |
5456815 | Fukuyo et al. | Oct 1995 | A |
5468401 | Lum et al. | Nov 1995 | A |
5508000 | Satou et al. | Apr 1996 | A |
5513512 | Segal | May 1996 | A |
5590389 | Dunlop et al. | Dec 1996 | A |
5600989 | Segal et al. | Feb 1997 | A |
5608911 | Shaw et al. | Mar 1997 | A |
5623726 | Kiiski et al. | Apr 1997 | A |
5673581 | Segal | Oct 1997 | A |
5693203 | Ohhashi | Dec 1997 | A |
5722165 | Kobayashi et al. | Mar 1998 | A |
5766380 | Lo et al. | Jun 1998 | A |
5772795 | Lally et al. | Jun 1998 | A |
5772860 | Sawada et al. | Jun 1998 | A |
5780755 | Dunlop et al. | Jul 1998 | A |
5798005 | Murata | Aug 1998 | A |
5809393 | Dunlop et al. | Sep 1998 | A |
5826456 | Kawazoe et al. | Oct 1998 | A |
5850755 | Segal | Dec 1998 | A |
5993575 | Lo et al. | Nov 1999 | A |
5993621 | Liu | Nov 1999 | A |
5994181 | Hsieh et al. | Nov 1999 | A |
6024852 | Tamura | Feb 2000 | A |
6085966 | Shimomuki | Jul 2000 | A |
6113761 | Kardokus et al. | Sep 2000 | A |
6123896 | Meeks, III et al. | Sep 2000 | A |
6130451 | Hasegawa | Oct 2000 | A |
6139701 | Pavate et al. | Oct 2000 | A |
6192969 | Bunn et al. | Feb 2001 | B1 |
6193821 | Zhang | Feb 2001 | B1 |
6221178 | Torizuka et al. | Apr 2001 | B1 |
6348113 | Michaluk | Feb 2002 | B1 |
6348139 | Shah et al. | Feb 2002 | B1 |
6391163 | Pavate et al. | May 2002 | B1 |
6454994 | Wang | Sep 2002 | B1 |
6521173 | Kumar et al. | Feb 2003 | B1 |
6887356 | Ford et al. | May 2005 | B1 |
6896748 | Perry et al. | May 2005 | B1 |
20010023726 | Koenigamann et al. | Sep 2001 | A1 |
20020041819 | Kumar et al. | Apr 2002 | A1 |
20050268999 | Oda | Dec 2005 | A1 |
Number | Date | Country |
---|---|---|
252442 | Sep 1960 | AU |
284905 | Nov 1990 | DE |
0281141 | Mar 1988 | EP |
0 590 904 | Apr 1994 | EP |
0902102 | Aug 1998 | EP |
822 813 | Dec 1998 | EP |
0902102 | Mar 1999 | EP |
55-179784 | Dec 1980 | JP |
59227992 | Dec 1984 | JP |
62-089543 | Apr 1987 | JP |
362089543 | Apr 1987 | JP |
62-297463 | Dec 1987 | JP |
03-082773 | Apr 1991 | JP |
H03-197640 | Aug 1991 | JP |
HO3-197640 | Aug 1991 | JP |
6-10107 | Jun 1992 | JP |
6-93400 | Sep 1992 | JP |
6-256919 | Mar 1993 | JP |
6-264232 | Sep 1994 | JP |
08-136406 | May 1996 | JP |
08146201 | Jun 1996 | JP |
8-232061 | Sep 1996 | JP |
8-269701 | Oct 1996 | JP |
10008244 | Jan 1998 | JP |
WO 8707650 | Dec 1987 | WO |
WO 9201080 | Jan 1992 | WO |
WO 9902743 | Jan 1999 | WO |
WO 9927150 | Jun 1999 | WO |
WO 9966100 | Dec 1999 | WO |
WO 0031310 | Jun 2000 | WO |
WO 0129279 | Apr 2001 | WO |
Number | Date | Country | |
---|---|---|---|
20020125128 A1 | Sep 2002 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 09497079 | Feb 2000 | US |
Child | 09999095 | US |