Semiconductor device manufacturing: process – Coating with electrically or thermally conductive material – To form ohmic contact to semiconductive material
Reexamination Certificate
2002-07-16
2004-06-01
Nelms, David (Department: 2818)
Semiconductor device manufacturing: process
Coating with electrically or thermally conductive material
To form ohmic contact to semiconductive material
Reexamination Certificate
active
06743716
ABSTRACT:
TECHNICAL FIELD
The technical field relates generally to semiconductor structures. More particularly, it pertains to metallization layers in semiconductor structures.
BACKGROUND
One of the main issues confronting the semiconductor processing industry is that of the capacitive-resistance problem in metallization layers. An industry-wide effort has undertaken to address the problem. Since the beginning, the semiconductor processing industry has relied on aluminum and aluminum alloys to serve as metallization layers. Silicon dioxide was selected as the insulator of choice although polyimide, a polymer, was used in a number of products by IBM for a number of years. With each succeeding generation of technology, the capacitive-resistance problem grows. Because each generation requires that the dimensions of the semiconductor structure be reduced, the minimum line-space combination must also decrease. As the line-space combination decreases, the capacitance and resistance of the semiconductor structure increases. Thus, these increases contribute to the problem.
Copper metallurgy has been proposed as a substitute for aluminum metallurgy as a material for the metallization layers since copper exhibits greater conductivity than aluminum. Yet several problems have been encountered in the development of copper metallurgy. The main issue is the fast diffusion of copper through an insulator, such as silicon dioxide, to form an undesired copper oxide compound. Another issue is the known junction-poisoning effect of copper. These issues have led to the development of a liner to separate the copper metallization layer from the insulator. The use of titanium nitride as a liner was proposed by C. Marcadal et al., “OMCVD Copper Process for Dual Damascene Metallization,” VMIC Conference Proceedings, p. 93-7 (1997). The use of tantalum nitride as a liner was proposed by Peijun Ding et al., “Copper Barrier, Seed Layer and Planarization Technologies,” VMIC Conference Proceedings, p. 87-92 (1997). The use of titanium as a liner was proposed by F. Braud et al., “Ultra Thin Diffusion Barriers for Cu Interconnections at the Gigabit Generation and Beyond,” VMIC Conference Proceedings, p. 174-9 (1996). The use of tungsten silicon nitride as a liner was proposed by T. Iijima et al., “Microstructure and Electrical Properties of Amorphous W—Si—N Barrier Layer for Cu Interconnections,” VMIC Conference Proceedings, p. 168-73 (1996). The use of zirconium, hafnium, or titanium as a liner was proposed by Anonymous, “Improved Metallurgy for Wiring Very Large Scale Integrated Circuits,” International Technology Disclosures, v. 4 no. 9, (Sep. 25, 1996). The use of titanium as a liner was proposed by T. Laursen, “Encapsulation of Copper by Nitridation of Cu—Ti Alloy/Bilayer Structures,” International Conference on Metallurgical Coatings and Thin Films in San Diego, Calif., paper H1.03 p. 309 (1997). The use of tantalum, tungsten, tantalum nitride, or trisilicon tetranitride as a liner is currently favored by the industry. See Changsup Ryu et al., “Barriers for Copper Interconnections,” Solid State Technology, p. 53-5 (1999).
Yet another solution to the problem of fast diffusion of copper through an insulator was proposed by researchers at Rensselaer Polytechnic Institute (hereinafter, RPI). See S. P. Muraka et al., “Copper Interconnection Schemes: Elimination of the Need of Diffusion Barrier/Adhesion Promoter by the Use of Corrosion Resistant, Low Resistivity Doped Copper,” SPIE, v. 2335, p. 80-90 (1994) (hereinafter, Muraka); see also Tarek Suwwan de Felipe et al., “Electrical Stability and Microstructural Evolution in Thin Films of High Conductivity Copper Alloys,” Proceedings of the 1999 International Interconnect Technology Conference, p. 293-5 (1999). These researchers proposed to alloy copper with a secondary element, which is either aluminum or magnesium. In their experiments, they used copper alloys with at least 0.5 atomic percent aluminum or 2 atomic percent magnesium. When the copper alloy is brought near the insulator, silicon dioxide, the secondary element and silicon dioxide form dialuminum trioxide or magnesium oxide. The formed dialuminum trioxide or magnesium oxide acts as a barrier to the fast diffusion of copper into the silicon dioxide.
Along the same technique as proposed by RPI, Harper et al. discuss in U.S. Pat. No. 5,130,274 (hereinafter, IBM) the use of a copper alloy containing either aluminum or chromium as the secondary element. As above, the secondary element with the insulator, such as silicon dioxide or polyimide, forms a barrier to the fast diffusion of copper.
Semiconductor products with some of the discussed solutions to the fast diffusion of copper have begun to ship, on a limited basis, and yet the problem of reducing the resistivity in ever smaller line dimensions is still present. It has been shown by Panos C. Andricacos, “Copper On-Chip Interconnections,” The Electrochemical Society Interface, pg. 32-7 (Spring 1999) (hereinafter Andricacos), that the effective resistivity obtainable by the use of barrier layers was approximately 2 microhm-centimeters with a line width greater than 0.3 micrometer. The effective resistivity undesirably increases for lines narrower than that. The alloy approach investigated by RPI had similar resistivity values as found by Andricacos. RPI also found that the use of 0.5 atomic percent aluminum, in the copper, was apparently insufficient to give complete protection from copper diffusion into the silicon dioxide although a significant reduction in the rate of copper penetration through the silicon dioxide was achieved. It should be noted that the maximum solubility of aluminum in copper is 9.2 weight percent or approximately 18 atomic percent whereas the maximum solubility of magnesium in copper is 0.61 weight percent or approximately 0.3 atomic percent. Thus, the alloys used by RPI were saturated with magnesium but far below the saturation limit when aluminum was used as the secondary element in the alloy.
Other researchers have focused on the capacitive effect. The capacitive effect has been studied with respect to polymers, such as polyimide, which are used to substitute for silicon dioxide as insulation in semiconductor structures. Some of these polymers have dielectric constants that are considerably lower than silicon dioxide, and a presumption can be made that the use of these polymers should lessen the undesired capacitive effect. Yet, when one of these polymers is cured to form an insulator near the vicinity of the copper metallization layer, the polymer reacts with the copper metallization layer to form copper dioxide, a conductive material. See D. J. Godbey et al., “Copper Diffusion in Organic Polymer Resists and Inter-Level Dielectrics,” Thin Solid Films, v. 308-9, p. 470-4 (1970) (hereinafter, Godbey). This conductive material is dispersed within the polymer thereby effectively raising the dielectric constant of the polymer and in many cases even increasing its conductivity. Hence, the undesired capacitive effect continues even with the use of lower dielectric polymer materials.
Andricacos points out that the use of copper along with cladding offers a significant improvement in conductivity over the titanium/aluminum-copper alloy/titanium sandwich structure now in widespread use throughout the industry. Andricacos also noted that as the line width decreases even a thin liner would undesirably effect the line resistance. The proposals by RPI and IBM attempt to address this problem by forming the liner using a copper alloy. The liner formed using a copper alloy displaces a portion of an area that was occupied by the insulator.
However, in solving one problem, RPI and IBM introduce another problem. The copper alloys used by RPI and IBM essentially lack the desirable properties of copper that initially drove the industry to use it. As was pointed out by RPI, the use of an alloy containing aluminum, even at a concentration so low as to not be completely effective in preventing the diffusion of copper, shows a measurable increase in resistance
No affiliations
Le Thao P
Nelms David
Schwegman Lundberg Woessner & Kluth P.A.
LandOfFree
Structures and methods to enhance copper metallization does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Structures and methods to enhance copper metallization, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Structures and methods to enhance copper metallization will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3354660