Chemistry: electrical and wave energy – Processes and products – Coating – forming or etching by sputtering
Reexamination Certificate
1997-02-07
2001-07-24
Nguyen, Nam (Department: 1753)
Chemistry: electrical and wave energy
Processes and products
Coating, forming or etching by sputtering
C204S298210, C204S298220
Reexamination Certificate
active
06264803
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates generally to a rotating cylindrical magnetron apparatus, and more specifically to the magnetic structure or array used to provide a magnetic field or “race track” to sputter the material contained on the surface of the cylindrical magnetron target, and the configuration of the target material relative to the magnetic structure.
2. Description of the Prior Art
This invention relates directly to an apparatus used for magnetron sputtering utilizing rotating, cylindrical shaped targets (cathodes). The cylindrical targets are used to deposit various materials onto a variety of substrates. The materials are deposited in a vacuum chamber using a gas reactive magnetron sputtering technique. The principles and operation of such apparatus are described in detail in Boozenny et al. U.S. Pat. No. 5,096,062, describing a large-scale cylindrical sputtering apparatus (sold by Airco Coating Technology under the trademark “C-MAG”) used for sputtering coatings onto architectural glass panels, automobile windshields, and the like. Further teachings pertaining to such apparatus are described in detail in Wolfe et al. U.S. Pat. No. 5,047,131.
In these prior patents, the magnetic structure or array (typically a closed loop) used to produce the magnetic field to confine the plasma is discussed, and the resulting sputtering area is referred to as the “race track”. Two distinct magnetic arrays which are used to produce the “race track” are discussed, one being the “W” configuration and the other being the “U” configuration, both of which produce a singular “race track” for confining the plasma on a target's surface. The “race track” consists of two parallel lines running longitudinally across the target surface, and an area at each end of the magnetic structure commonly called the “turn-around” area. Together, the parallel lines and turn-arounds form a magnetic loop that comprises the “race track”.
One of the claimed advantages for the rotating cylindrical target is the greater amount of material available for sputtering as compared to a planar magnetron target. In the case of the planar magnetron, the race track and target are in a static condition, therefore, the erosion groove along the race track is narrow, causing rapid depletion of the target material being sputtered within this localized area, and resulting in low material utilization (approximately 25%) before necessary replacement of the target. In the case of the rotating cylindrical target, because of its cylindrical shape and the fact that it rotates around a fixed magnet array, more target material is available to sputter and the target utilization is thus increased.
While the foregoing is true, the material utilization of the cylindrical targets, using the “W” or “U” magnet configurations, falls far short from the theoretical expectation of 85% to 95% of target material usage. In actuality the utilization for the cylindrical target is approximately 40%, making the cylindrical cathode much less advantageous than was originally anticipated. The reason the utilization ended up being less than expected was due to the localized areas at each end of the race track, called the turn-around areas. The turn-around area at each end of the race track has approximately 2.5 times the power density as compared to the medial area along the parallel lines running longitudinally across the target surface. As a result of the higher power density at the turn-around areas, the ends of the cylindrical target wear through (erode) more rapidly than the area between the turn-around areas.
Perhaps an obvious solution for the poor utilization might be to simply increase the target material thickness at the localized turn-around areas of the target. However, experimentation has shown that simply increasing the material thickness by 2.5 times while using the single magnetic race track configuration is not a feasible approach, because this methodology produces two undesirable effects. The first problem is the reduction of the magnetic field strength at the turn-around areas, which causes a loss of plasma confinement, and which can create instability in the plasma.
The second problem with this approach is that some of the sputter materials have too much internal stress, limiting the available thickness of the target material. This situation becomes evident when trying to apply increased thickness for Silicon target materials. As the Silicon thickness increases, so does the stress, and cracking or delamination of the target material results. This is particularly true when the Silicon target material is subjected to high power densities, as would be the situation with the single race track configuration at the turn-around areas. Therefore, in the case of Silicon targets, applying the “thicker end” methodology to exclusively overcome the erosion problem cannot be accomplished by itself, due to the high stress in thicker material located at the target ends. However, it is true that thicker material at the target ends can be used, but only in moderation, and not exclusively.
A further problem using the “W” or “U” magnet configurations of a single race track is that high power densities are concentrated onto the target surface due to the tight plasma confinement created by the magnetic field. This situation becomes apparent while considering low melting point materials (e.g., Zinc and Tin), or materials having a propensity for arcing (e.g., Silicon) at higher power densities. In the cases of the aforementioned materials it becomes desirable to distribute the power being applied to the target over a larger surface area, thereby reducing the localized areas of high power density.
OBJECTS AND ADVANTAGES
It is therefore an object of the present invention to improve the overall material utilization for a rotating cylindrical magnetron target, thereby reducing the effective target cost.
It is a further object of the present invention to improve the sputtering process conditions for cylindrical magnetron target materials such as Zinc, Tin, and Silicon to reduce the risk of melting and/or arcing.
SUMMARY OF THE INVENTION
The method and apparatus of this invention provides improved structural features and techniques used to accomplish the foregoing objectives for improving the material utilization and process stability for the rotating cylindrical magnetron targets used in a sputtering system.
The first methodology used in this invention is to incorporate an additional (i.e., second) magnet structure or “race track” immediately adjacent the first magnet structure or “race track”. Having two distinct magnetic race tracks allows the power density on the target surface to be reduced by approximately 50% for any given point where magnetic confinement of the plasma exists, as compared to the single race track configuration. This double race track configuration, by itself, does not overcome the rapid erosion problem at the magnetic turn-around areas, as described above, but it does provide other advantages (discussed infra).
The second methodology recognizes that since the localized power density is reduced by the double “race track” configuration, the power density at the turn-around areas can be controlled by shifting (offsetting) one race track from the other, along the longitudinal axis of the race tracks and target, so that the ends of “turn-around” areas are offset from one another. As previously discussed, the single race track turn-around areas have approximately 2.5 times more power density than along the parallel and longitudinal race track lines in the medial portion of the target. Therefore, a properly offset dual race track configuration only has 1.25 times the power density at each turn-around area (relative to the longitudinal area), and because they are offset to each other, the erosion at the turn-around areas can thus be minimized to 1.25 times, as compared to the longitudinal areas between the magnetic turn-arounds. This methodology will thus yield approximately 80% target utilization, or
Gobin Guy
Morgan Steven V.
Vanderstraeten Erwin
Vanderstraeten Johan
Johnson Larry D.
McDonald Rodney G.
Nguyen Nam
LandOfFree
Apparatus and method for sputtering does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Apparatus and method for sputtering, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Apparatus and method for sputtering will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2559660