Adhesive bonding and miscellaneous chemical manufacture – Differential fluid etching apparatus – With plasma generation means remote from processing chamber
Reexamination Certificate
2000-08-11
2002-12-17
Mills, Gregory (Department: 1763)
Adhesive bonding and miscellaneous chemical manufacture
Differential fluid etching apparatus
With plasma generation means remote from processing chamber
C156S345480, C118S7230IR, C118S7230IR, C118S7230AN
Reexamination Certificate
active
06494986
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field
The invention concerns plasma reactors used in processing workpieces in the manufacturing of items such as microelectronic circuits, flat panel displays and the like, and in particular plasma sources therefor.
2. Background Art
The trend in microelectronic circuits toward ever increasing densities and smaller feature sizes continues to make plasma processing of such devices more difficult. For example, the diameter of contact holes has been reduced while the hole depth has increased. During plasma-enhanced etching of a dielectric film on a silicon wafer, for example, the etch selectivity of the dielectric material (e.g. silicon dioxide) to photoresist must be sufficient to allow the etch process to etch a contact hole whose diameter is ten to fifteen times its depth, without appreciably disturbing the photoresist mask defining the hole. This task is made even more difficult because the recent trend toward shorter wavelength light for finer photolithography requires a thinner photoresist layer, so that the dielectric-to-photoresist etch selectivity must be greater than ever. This requirement is more readily met using processes having relatively low etch rates, such as dielectric etch processes employing a capacitively coupled plasma. The plasma density of a capacitively coupled plasma is relatively less than that of an inductively coupled plasma, and the capacitively coupled plasma etch process exhibits good dielectric-to-photoresist etch selectivity. The problem with the capacitively coupled process is that it is slow and therefore relatively less productive. Another problem that arises in such etch processes is non-uniform plasma distribution.
In order to increase productivity or etch rate, higher density plasmas have been employed. Typically, the high density plasma is an inductively coupled plasma. However, the process precursor gases tend to dissociate more rapidly in such a high density plasma, creating a higher plasma content of free fluorine, a species which reduces the etch selectivity to photoresist. To reduce this tendency, fluoro-carbon process gases such as CF
2
are employed which dissociate in a plasma into fluorine-containing etchant species and one or more polymer species which tend to accumulate on non-oxide containing surfaces such as photoresist. This tends to increase etch selectivity. The oxygen in the oxygen-containing dielectric material promotes the pyrolization of the polymer over the dielectric, so that the polymer is removed, allowing the dielectric material to be etched while the non-oxygen containing material (e.g., the photoresist) continues to be covered by the polymer and therefore protected from the etchant. The problem is that the increase in contact opening depth and decrease in photoresist thickness to accommodate more advanced device designs has rendered the high density plasma process more likely to harm the photoresist layer during dielectric etching. As the plasma density is increased to improve etch rate, a more polymer-rich plasma must be used to protect the non-oxygen containing material such as photoresist, so that the rate of polymer removal from the oxygen-containing dielectric surfaces slows appreciably, particularly in small confined regions such as the bottom of a narrow contact opening. The result is that, while the photoresist may be adequately protected, the possibility is increased for the etch process to be blocked by polymer accumulation once a contact opening reaches a certain depth. Typically, the etch stop depth is less than the required depth of the contact opening so that the device fails. The contact opening may provide connection between an upper polysilicon conductor layer and a lower polysilicon conductor layer through an intermediate insulating silicon dioxide layer. Device failure occurs, for example, where the etch stop depth is less than the distance between the upper and lower polysilicon layers. Alternatively, the possibility arises of the process window for achieving a high density plasma without etch stop becoming too narrow for practical or reliable application to the more advanced device designs such as those having contact openings with aspect ratios of 10:1 or 15:1.
What would be desirable at present is a reactor that has the etch rate of an inductively coupled plasma reactor (having a high density plasma) with the selectivity of a capacitively coupled reactor. It has been difficult to realize the advantages of both types of reactors in a single machine led reactor.
One problem with high density inductively coupled plasma reactors, particularly of the type having an overhead coil antenna facing the wafer or workpiece, is that as the power applied to the coil antenna is increased to enhance the etch rate, the wafer-to-ceiling gap must be sufficiently large so that the power is absorbed in the plasma region well above the wafer. This avoids a risk of device damage on the wafer due to strong RF fields. Moreover, for high levels of RF power applied to the overhead coil antenna, the wafer-to-ceiling gap must be relatively large, and therefore the benefits of a small gap cannot be realized.
If the ceiling is a semiconductive window for the RF field of an inductively coupled reactor, or a conductive electrode of a capacitively coupled reactor, then one benefit of a small wafer-to-ceiling gap is an enhanced electric potential or ground reference that the ceiling could provide across the plane of the wafer at a relatively small gap distance (e.g., on the order of 1 or 2 inches).
Therefore, it would be desireable to have a reactor not only having the selectivity of a capacitively coupled reactor with the ion density and etch rate of an inductively coupled reactor, but further having none of the conventional limitations on the wafer-to-ceiling gap length other than a fundamental limit, such as the plasma sheath thickness, for example. It would further be desireable to have a reactor having the selectivity of a capacitively coupled reactor and the etch rate of an inductively coupled reactor in which the ion density and etch rate can be enhanced without necessarily increasing the applied RF plasma source power.
SUMMARY OF THE DISCLOSURE
A plasma reactor for processing a workpiece includes an enclosure, a workpiece support within the enclosure facing an overlying portion of the enclosure, the workpiece support and the overlying portion of said enclosure defining a process region therebetween extending generally across the diameter of said wafer support. The enclosure has at least first and second openings therethrough near generally opposite sides of said workpiece support. At least one hollow conduit outside of said process region is connected to the first and second openings, providing a first torroidal path extending through the conduit and across the process region. A first coil antenna adapted to accept RF power is inductively coupled to the interior of the hollow conduit and is capable of maintaining a plasma in the torroidal path. Furthermore, the enclosure further includes at least third and fourth openings therethrough near generally opposite sides of the workpiece support and disposed along an axis transverse to an axis of the first and second openings. A second hollow conduit generally transverse to the one hollow conduit is similarly disposed outside of the process region and connected to the third and fourth openings, whereby to provide a second closed torroidal path, the second torroidal path extending through the second conduit and across said process region between said third and fourth openings in a direction transverse to the first torroidal path. A second coil antenna capable of being coupled to an RF power source is inductively coupled to the interior of the second hollow conduit and capable of maintaining a plasma current in the second torroidal path.
REFERENCES:
patent: 2344138 (1944-03-01), Drummond
patent: 3109100 (1963-10-01), Papp
patent: 3291715 (1966-12-01), Anderson
patent: 4431898 (1984-02-01), Reinberg et al.
patent: 4778561 (1988-10-01
Collins Kenneth S
Hanawa Hiroji
Nguyen Andrew
Ramaswamy Kartik
Tanaka Tsutomu
Applied Materials Inc.
Bach Joseph
Hassanzadeh P.
Wallace Robert M.
LandOfFree
Externally excited multiple torroidal plasma source does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Externally excited multiple torroidal plasma source, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Externally excited multiple torroidal plasma source will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2948495