Etching a substrate: processes – Gas phase etching of substrate – Application of energy to the gaseous etchant or to the...
Reexamination Certificate
1997-09-17
2001-07-10
Gulakowski, Randy (Department: 1746)
Etching a substrate: processes
Gas phase etching of substrate
Application of energy to the gaseous etchant or to the...
C438S731000, C156S345420
Reexamination Certificate
active
06258287
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to the preparation of etched substrates. More particularly, the present invention relates to an improved process for low-damage, anisotropic etching of substrates such as semiconductors and insulators, and improved anisotropically etched substrates.
BACKGROUND OF THE INVENTION
Dry etching is an absolutely critical process in the fabrication of all micrometer and nanometer scale features on high speed electronic and opto-electronic devices. In brief, the fabrication of such chips and devices involves the following process. A substrate of some semiconductor or metal is selected and a pattern is laid down over it, the pattern having open areas in it. The overlying structure containing the pattern is sometimes called a mask. Etching chemistry is then performed through the open areas, which means that in effect some of the underlying material exposed through the open areas is dissolved away so that the pattern is transferred into the underlying layer(s). Then, the mask is stripped away and what is left behind is the original substrate, but now the pattern has been transferred into it. The process is similar to silk screening or stamping a pattern into material. The resulting pattern has a three dimensional structure.
In the early days of integrated circuit fabrication, most etching was done using a wet chemical process that is quite similar to conventional photography. For example, to etch an array of grooves in a silicon wafer, the wafer is first placed in a high temperature, oxidizing environment and a layer of silicon dioxide is grown on the top surface of the wafer. Then, the oxidized wafer is covered with a thin photosensitive layer of gelatinous organic material called a “photoresist.” Next, a piece of material analogous to a photographic negative, called a “photomask,”is placed over the photoresist. Ultraviolet light is shined through openings in this photomask. The ultraviolet light changes the solubility of the photoresist. Thus, areas of photoresist that have been illuminated with the ultraviolet light display a different solubility than areas which have not been exposed to the light. Finally, a solvent is used which dissolves away only the areas of the photoresist which have had their solubility increased by the ultraviolet light. At this point, the original pattern on the photomask has been transferred to the photoresist layer. Some people refer to this patterned photoresist layer as a “soft mask.”
Subsequently, a wet chemical, hydrofluoric acid (HF) dip is used to dissolve away the silicon dioxide which has been exposed through each of the openings in the overlaying photoresist. Then, the photoresist is stripped off. At this point, it is apparent that the pattern originally appearing on the photomask has been transferred to the silicon dioxide layer overlying the silicon wafer. This patterned layer of silicon dioxide is sometimes referred to as a “hard mask.”
Finally, the wafer is dipped into a caustic etch, such as potassium hydroxide (KOH) which etches away the silicon exposed under the openings in the hard mask. After stripping away the hard mask, the desired silicon wafer with the etched grooves remains.
As an alternative to following the hard masking step described above with a KOH silicon etch, ion implantation or high temperature diffusion could optionally be used to place dopant atoms through the openings in the hard mask.
Many other structural and chemical variations using the sorts of wet processing steps described above are possible and are well known to those of skill in the art. In each case however, the corresponding process suffers from a problem that is inherent with the associated etching using wet chemicals. In particular, at the same time that the wet chemistry is etching down into the wafer, it is also etching laterally under the mask. Indeed, this undesired lateral etching tends to extend approximately as far as the desired vertical etching. This tendency for wet chemical etching to proceed equally in all directions without prejudice is called “isotropic etching.”
Isotropic etching is adequate for making a line that is 20 microns wide through a film that is 1 micron deep. The resulting inaccuracy in the edges of such a feature is a small percentage of the overall device structure; and therefore, it does not compromise performance. However, as smaller and smaller structures are fabricated, isotropic etching is inadequate. The industry is moving toward fabrication of structures with so called sub-micron features, which are essential for high speed computer chips, optical structures, and electronic and optoelectronic devices. In other words, the accurate transfer of a pattern which is half a micron wide into a material which is half a micron thick requires absolutely straight vertical side walls, or anisotropic etching. Isotropic etching is inadequate because the associated rounded undercut would be a very high percentage of the active device material and would destroy its performance.
Presently, it is thought that the only way to get straight side walls is by a technique called reactive-ion etching (RIE). Rather than dipping the device in wet chemicals, it is exposed to reactive gases in plasmas. Energetic ions formed in the plasma are accelerated in the normal direction to the substrate where they enhance the etching chemistry at the bottom of the open area defined by the mask and not on the sidewalls. Thus, straight side walls can be achieved with reactive-ion etching.
Reactive-ion etching provides anisotropic etching. However, the ions are heavy ions like argon or CF
3
+
and are traveling at a few hundred electron volts of kinetic energy. Thus, they carry enough momentum to displace lattice atoms from their normal position. This damages the surfaces and often, the optical and electrical properties of the substrate have been detrimentally changed.
Fabrication of ultra small electronic and optoelectronic devices requires dry etching processes that give high anisotropy, high selectivity between different materials, and minimal surface damage. Currently, ion enhanced plasma etching processes (e.g. reactive ion etching (RIE) and electron cyclotron enhanced RIE (ECR)) create high aspect ratio nanometer scale features; however, etch induced damage has become increasingly troublesome as critical dimensions shrink. To minimize etch damage, reactive species generated in the plasma should have energies larger than the activation energy of the etch reaction (a fraction of an eV), but less than the energy required for atomic displacement (3 to 10 eV for III-V semiconductors). Given these limitations, the ion energies available in reactive-ion (about 300 eV) and electron cyclotron resonance plasma etching (about 50 eV) are not ideally suited for fabricating nanometer scale devices.
Placing the sample to be etched on the anode within a DC plasma environment is one way to ensure precise control over the anisotropic etching process while minimizing damage to the substrate and is described in commonly assigned copending U.S. patent application Ser. No. 08/705,902, filed Aug. 28, 1996 and titled “METHOD AND APPARATUS FOR LOW ENERGY ELECTRON ENHANCED ETCHING OF SUBSTRATES”. This technique is called Low Energy Electron Enhanced Etching (LE4, for convenience), and operates by placing the etching substrate on the anode of a DC glow discharge. This method works well for conducting and semi-conducting substrates, but is inherently problematic for etching non-conducting substrates such as insulators because, in the aforementioned method, the substrate sample is physically and electrically connected to the anode in the plasma, thus becoming a conducting element of the electrical circuit within the plasma. An insulator, by definition does not efficiently conduct electrical current. Therefore, placing an insulating substrate on the anode will impede the electrical flow and will be an inefficient and nearly impossible way to etch a non conducting substrate.
What is needed and was apparently not available unt
Choutov Dmitri A.
Gillis Harry P.
Martin Kevin P.
Georgia Tech Research Corporation
Gulakowski Randy
Olsen Allan
Thomas Kayden Horstemeyer & Risley, L.L.P.
LandOfFree
Method and apparatus for low energy electron enhanced... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus for low energy electron enhanced..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus for low energy electron enhanced... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2480553