Etching a substrate: processes – Gas phase etching of substrate – Application of energy to the gaseous etchant or to the...
Reexamination Certificate
1996-10-21
2001-05-29
McDonald, Rodney G. (Department: 1753)
Etching a substrate: processes
Gas phase etching of substrate
Application of energy to the gaseous etchant or to the...
C216S067000, C216S079000, C438S707000, C438S710000, C438S719000, C438S723000, C438S725000, C204S192320, C204S192370
Reexamination Certificate
active
06238588
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Technical Field
The invention is related to a high pressure high non-reactive diluent gas content high plasma ion density plasma oxide etch process.
2. Background Art
In a plasma processing chamber, and especially in a high density plasma processing chamber, RF (radio frequency) power is used to generate and maintain a plasma within the processing chamber. As disclosed in detail in the abovereferenced applications, it is often necessary to control temperatures of surfaces within the process chamber, independent of time varying heat loads imposed by processing conditions, or of other time varying boundary conditions. This is particularly true in the case of a reactor chamber having a window electrode which acts as both a capacitively coupled electrode and a window for admitting therethrough RF power inductively coupled from an inductive antenna. In some cases where the window/electrode is a semiconducting material, it may be necessary to control the temperature of the window/electrode within a particular temperature range to obtain the proper electrical properties of the window. The application of RF power to generate and maintain the plasma leads to heating of surfaces within the chamber, including windows (such as used for inductive or electromagnetic coupling of RF or microwave power) or electrodes (such as used for capacitive or electrostatic coupling of RF power, or for terminating or providing a ground or return path for such capacitive or electrostatic coupling of RF power) or for combination window/electrodes.
In the above-referenced parent application it is disclosed how to overcome the foregoing problems by, among other things, employing multiple solenoid windings at respective radial locations over the reactor chamber ceiling, while enduring the conventional limitations with regard to chamber pressure. As discussed above, the chamber pressure in a high ion density (e.g., 10
11
ions/cc) plasma reactor (e.g., an inductively coupled RF plasma reactor) typically is limited by plasma electron recombination losses that increase with chamber pressure. Such losses prevent electron diffusion that would otherwise enhance plasma ion distribution uniformity. The same is generally true of microwave electron cyclotron resonance plasma reactors. In the case of inductively coupled RF plasma reactors, a typical chamber pressure range is between about 1 mT and 10 mT, while 20 mT is considered above the typical range. Given the definition stated above for “high pressure”, in which the inductive field skin depth is greater than {fraction (1/10)} of the gap between the top electrode and the wafer, 100 mT is definitely “high pressure”. The uniformity of etch rate and etch selectivity is reduced as skin depth (or chamber pressure) increases because non-uniformities in the overhead antenna pattern are more strongly mapped to the wafer surface as skin depth increases. For example, it has been demonstrated that reducing chamber pressure from 75 mT to 20 mT greatly enhances etch selectivity uniformity across the wafer. Thus, conventional wisdom has been to limit chamber pressure in a high density plasma reactor.
Such problems are particularly acute in plasma etching of silicon dioxide layers over underlying non-oxygen-containing layers (such as polysilicon, silicon, silicon nitride, and so forth). This is because the silicon-oxygen bond is much stronger than the bonds in the underlying layer, necessitating the passivation of the underlying layer by a polymer deposited from polymer precursor species in the plasma. Without such passivation, the etch selectivity of the silicon dioxide to the underlying non-oxygen containing layer is inadequate. As is well known, the preferred process gases include fluorocarbon or fluoro-hydrocarbon gases because such gases are precursors for both the etchant species (fluorine) and the polymerizing species. Selectivity is enhanced by increasing the chamber pressure (by decreasing the chamber vacuum pump rate or “throttling back” the pump), because such a pressure increase increases the net residence time of the polymer precursor species so that more polymer precursor species are formed in the plasma. (As understood in this specification, the term “residence time” refers to a particular gas species and is the pressure of that gas multiplied by the volume encompassed between the wafer or workpiece and the plasma source power applicator (typically an overhead inductive antenna) divided by the flow rate at which the gas is supplied into the reactor chamber.) Under such conditions, a stronger polymer tends to form on the underlying passivated layers, thereby enhancing etch selectivity.
In order to counteract the stronger polymer formation on the silicon dioxide surfaces to be etched, the ion energy in the plasma must be increased well above the usual level (the silicon-oxygen bond energy) to overcome polymer deposition on the silicon dioxide surfaces. As a result, the process window is decreased to the extent a higher ion energy is required to prevent etch stopping. To avoid such difficulties, it has been desirable to limit the chamber pressure (by increasing the chamber vacuum pump rate), which limits the selectivity which is enhanced by increasing the pressure. Thus, a certain tradeoff exists between etch selectivity and avoidance of etch stopping.
The problem with having to so limit the chamber pressure is that the polymer formation is weakened accordingly. As stated above, the higher chamber pressure produces stronger polymer and, conversely, lower chamber pressure produces weaker polymer passivation layers. The resulting limitation on polymer strength is manifested in photolithographic layers on the wafer, for example, in a phenomenon sometimes referred to as photoresist mask faceting, in which the polymer passivation layer exhibits a certain weakness around the edges of a contact opening in the photoresist mask layer, permitting the plasma to attack the photoresist at those edges or “facets”. Typically, the silicon dioxide-to-photoresist selectivity at the facets is about 3:1. The result is that the top of the contact opening widens during the etch process, so that the diameter of the opening cannot be controlled. Such a problem is reduced by increasing the chamber pressure to strengthen the polymer passivation layer over the photoresist, but such an increase in pressure requires a corresponding increase in plasma ion energy to avoid etch stopping near the wafer center, thereby narrowing the process window. Thus, there has seemed to be no real solution to such problems.
It is therefore an object of the present invention to strengthen the polymer passivation layer particularly near photoresist facets without risking etch stopping or requiring an increase in plasma ion energy to prevent such etch stopping or incurring other disadvantages typically associated with an increase in chamber pressure.
SUMMARY OF THE INVENTION
It is a discovery of the present invention that increasing the chamber pressure of a high ion density RF plasma reactor by introducing a non-reactive gas (such as an inert gas), rather than by throttling back the chamber vacuum pump, increases the polymer passivation layer strength, particularly near photoresist facets, without a concomitant increase in risk of etch stopping. Thus, the present invention is embodied in a process in which the etchant and polymer precursor gas, such as a fluorocarbon or fluoro-hydrocarbon gas, is diluted with an inert gas such as argon to increase chamber pressure without a corresponding significant change in the chamber vacuum pump rate. Preferably, the etchant/polymer precursor gas is fed into the chamber at a gas flow rate which, by itself, would maintain the chamber pressure below the high pressure regime, and the non-reactive gas is added at a flow rate which, in combination with the flow rate of the precursor gas, is sufficient to raise the chamber pressure into the high pressure regime. By thus refraining from significantly throttling back the chamber vacuum pump, the polymer pre
Collins Kenneth
Cui Chunshi
Ding Jian
Groechel David
Hung Raymond
Applied Materials Inc.
McDonald Rodney G.
Michaelson and Wallace
LandOfFree
High pressure high non-reactive diluent gas content high... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with High pressure high non-reactive diluent gas content high..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and High pressure high non-reactive diluent gas content high... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2563861