Electricity: measuring and testing – Magnetic – Combined
Reexamination Certificate
2001-02-21
2002-10-29
Lefkowitz, Edward (Department: 1753)
Electricity: measuring and testing
Magnetic
Combined
C204S192120
Reexamination Certificate
active
06472867
ABSTRACT:
BACKGROUND
1. Field of the Invention
The invention relates generally to physical vapor deposition (PVD) of metal films.
The invention relates more specifically to DC magnetron sputtering of ferromagnetic metals such as nickel (Ni) onto semiconductor substrates and the like for forming metallization such as found in the electrically-conductive interconnect layers of modern integrated circuits.
2. Description of the Related Art
The electrically-conductive interconnect layers of modern integrated circuits (IC) are generally of very fine pitch (e.g., 10 microns or less) and high density (e.g., hundreds of lines per square millimeter).
If there is nonuniformity of thickness or nonhomogeneity in other attributes of the precursor metal films that ultimately form the metallic interconnect layers of an IC, such lack of uniformity can lead to out-of-tolerance topographies and improper semiconductor fabrication. The latter can be detrimental to the operational integrity of the ultimately-formed IC. As such it is desirable to form metal films with good uniformity across each of mass-produced wafers.
The metal films of integrated circuits may be formed by physical vapor deposition (PVD). One low cost approach uses a DC magnetron sputtering apparatus such as the Endura™ system available from Applied Materials Inc. of California for sputtering metals onto semiconductor wafers or other like workpieces.
Aluminum (Al) is the most common metal that is deposited by DC magnetron PVD sputtering. Aluminum can be characterized as a polycrystalline, electrically conductive material whose crystals have a face-centered cubic (FCC) structure. One of the characteristics of Al is that it is an essentially nonmagnetic material. (Al may be considered paramagnetic though.)
Recently it has been proposed that magnetic metals such as nickel (Ni) may also be deposited using the Endura™ or like DC magnetron PVD systems.
Because nickel (Ni) is a ferromagnetic material, it presents new problems that had not been earlier posed by nonmagnetic materials such as aluminum. In particular, magnetic flux fields generated within the DC magnetron PVD system may be significantly altered due to shunting or short circuiting of the magnetic flux through the magnetically conductive material of ferromagnetic (e.g., Ni) targets. Such shunting can make it difficult to strike a plasma or sustain a generally-uniform plasma over time and can lead to associated problems such as nonuniform deposition. There is a question as to whether ferromagnetic targets of practical thicknesses (e.g., 3 millimeters or greater) can be used for sputtering with a DC magnetron PVD system.
The present inventors have through experimentation, isolated a number of physical attributes of ferromagnetic targets (e.g., nickel targets) that collectively correlate with how uniform the deposited film is across the substrate and how efficiently the material of the target is used. These collective correlations are disclosed herein together with designs for improved ferromagnetic targets.
SUMMARY OF THE INVENTION
It has been determined that fairly stable plasmas can be struck and sustained in DC_magnetron PVD systems even if ferromagnetic targets are used, and even if the targets have a thickness of as much as 3 mm or more.
Three attributes of nickel-based targets have been found to collectively correlate with uniform deposition thickness. They are in order of importance (with no one factor being dominant by itself): (1) the mix of crystallographic textures in the target, (2) the target's initial pass-through flux factor (% PTF), and (3) the maximum metal grain size in the target.
More particularly it has been found that; where the commercially useful life of nickel targets is limited by cross-workpiece deposition uniformity, an improvement can be obtained in the form of: (1) better deposition uniformity through the commercially useful life of nickel targets (e.g., a useful life of at least 60 KiloWatt Hours {kWHrs}), and/or (2) a longer commercially useful life for each nickel target in view of given limit on acceptable nonuniformity (e.g., cross-wafer resistivity variation of about 5% or less (at 3&sgr;) over target life).
Such improvement in target longevity and/or deposition uniformity may be obtained first by providing, in ferromagnetic targets that have a thickness of as much as 3 mm or more: an average (with per-sample-point restrictions), and more preferably, a homogeneous crystalline texture mix that is at least 20% of the <200> oriented texture. More preferably, the texture mix should at the same time be less than about 50% of the <111> oriented texture. Even more preferably, an average, and more preferably, a homogeneous texture mix should be provided that is at least 32% <200> texture, while further keeping at less than about 10% the <111> oriented texture. Yet more preferably, an average, and more preferably, a homogeneous texture mix should be provided that is at least 35% <200> texture, while further (optionally) keeping at less than 9% the <111> oriented texture. Yet more preferably, the latter homogeneous texture mix should further keep at less than 30% the <113> oriented texture. The remainder of the homogeneous texture mix can be of the <220> texture.
The above-mentioned average with per-sample-point restrictions may be determined for each value of texture in the texture mix by averaging over a multi-point symmetric pattern such as for example a star having four outer points and one central point. Star patterns with greater numbers of points can, of course, be alternatively used. The phrase, “with per-sample-point restrictions”, indicates that each of the sample points participating in the average must further comply with a limited deviation such as being plus or minus 10% of the calculated average. By way of a more specific example, calling for a 20% average content of the <200> oriented texture with a per-sample-point restriction of +/−10% means that anyone of the sample points can be as low as 18% in content of the <200> oriented texture or as high as 22% in content of the <200> oriented texture, so long as the unweighted average is still 20%. In one embodiment, each of the above average specifications for each given type of oriented texture carries with it a per-sample-point restriction (PSPR) of +/−10%. In more tightly specified, second embodiment, each of the above average specifications for each given type of oriented texture carries with it a per-sample-point restriction of +/−5%. Other restriction values may be used provided they are no tighter than the margin of error for per-sample-point measurements and not so loose as to make the average value meaningless with respect to physical consequences (e.g., a per-sample-point restriction of greater than about +/−50%).
Improvement in target longevity and/or through-life deposition uniformity may be further obtained for such thick ferromagnetic targets (e.g. 3 mm or greater thickness) by simultaneously providing (for any one of the texture mixtures specified immediately above) an initial through-target pass-through flux factor (% PTF) that is at least high enough to initially strike a plasma and preferably a higher TPTF. An initial TPTF of about 30% or greater on average with a per-sample-point restriction of between +/−10% and +/−5% across the active (sputtering) portion of the target has been found workable for a permanent driving magnet of about 400 to 500 Gauss. More preferably, the initial % PTF of about 30% or greater should be found homogeneously across the active (sputtering) portion of the target rather than merely on a 5-point or other average.
Improvement in target longevity and/or through-life deposition uniformity may be further obtained by simultaneously providing with said texture mixtures and/or said initial % PTF, an average, and more preferably, a homogeneous grain size in the target of about 200 &mgr;m (200 microns) or less, where the
Abburi Murali
Ramaswami Seshadri
Applied Materials Inc.
Fliesler Dubb Meyer & Lovejoy
Lefkowitz Edward
Zaveri Subhash
LandOfFree
Target for use in magnetron sputtering of nickel for forming... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Target for use in magnetron sputtering of nickel for forming..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Target for use in magnetron sputtering of nickel for forming... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2921333