Semiconductor device manufacturing: process – Having magnetic or ferroelectric component
Reexamination Certificate
2002-09-30
2003-08-12
Nelms, David (Department: 2818)
Semiconductor device manufacturing: process
Having magnetic or ferroelectric component
C257S295000
Reexamination Certificate
active
06605477
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention pertains to the field of thin films for use in integrated circuits, and particularly thin film layered superlattice materials. More specifically, a specialized interface buffer layer enhances the performance of thin film layered superlattice materials.
2. Statement of the Problem
Ferroelectric materials are characterized by their ability to retain an induced polarization state even in the absence of an applied electric field. If the polarization state in one direction is identified as a logic “0” polarization state and the polarization state in the opposite direction is identified as a logic “1” polarization state, and appropriate circuitry is provided to sense the polarization state, a ferroelectric material can be used as the information storage medium of a high speed nonvolatile computer memory. It is known that such ferroelectric memory device can be made by substituting a ferroelectric material for the dielectric capacitor material of a conventional DRAM capacitor circuit and making appropriate changes in the read and write circuits and manufacturing processes to utilize the ferroelectric film as the information storage medium. See, for example, U.S. Pat. No. 5,784,310 issued Jul. 21, 1998 to Cuchiaro et al. This substitution converts the DRAM cell to a nonvolatile memory cell due to long-term retention of an induced polarization state in the ferroelectric material even in the absence of an applied field. It is also possible to make a ferroelectric memory cell consisting of a single field effect transistor due to the nonvolatile polarization state of ferroelectric thin films, as described in U.S. Pat. No. 5,780,886 issued to Yamanobe et al.
A problem arising in the use of ferroelectric memory devices is that point charge defects at the thin film layered superlattice material layer surfaces have the effect of screening the applied field due to the presence of induced charge at the thin film surface creating a field opposite to the applied field. Thus, some of the interior ferroelectric domains of the crystal are never exposed to a field having sufficient magnitude to completely polarize the domains. The polarization performance of the ferroelectric memory devices suffer as a result of this field screening.
The most serious problems associated with screening, i.e., ferroelectric fatigue, leakage, and imprint problems, can be largely overcome through the use of layered thin film superlattice materials, as reported in U.S. Pat. No. 5,784,310 issued Jul. 21, 1998 to Cuchiaro et al. Ferroelectric perovskite-like layered thin film superlattice materials are a known class of self-ordering crystals, and have been used in thin films suitable for use in integrated circuits, e.g., as reported in U.S. Pat. No. 5,519,234 issued May 21, 1996 to Araujo et al. The term “perovskite-like” usually refers to a number of interconnected oxygen octahedra. A primary cell is typically formed of an oxygen octahedral positioned within a cube that is defined by large A-site metals where the oxygen atoms occupy the planar face centers of the cube and a small B-site element occupies the center of the cube. In some instances, the oxygen octahedra may be preserved in the absence of A-site elements.
The thin film layered superlattice materials layers are characterized by an ability to find thermodynamic stability in layered structures. Disordered solutions of superlattice-forming metals, when exposed to thermal treatments, spontaneously form a single layered superlattice material compound having intercollated layers of perovskite-like octahedrons and a superlattice generator such as bismuth oxide. The resultant self-ordered structure forms a superlattice by virtue of a dual periodicity corresponding to the repeated layers. The layered thin film superlattice materials have this self-ordering ability and, consequently, are distinct from semiconductor heterolattices which require the deposition of each layer in a separate deposition step.
The thin film layered superlattice materials, while much superior than prior art materials, still present problems relating to integration into conventional integrated circuit processes. For example, crystallization of the best of the layered superlattice materials require high temperatures that can damage conventional integrated circuit components such as wiring layers and semiconductors. In addition, diffusion of chemical elements between these materials and conventional integrated circuit materials degrades both the layered materials and the conventional materials. Further, the interfaces between the layered superlattice materials and conventional integrated circuit materials are often problematic: evaporation of highly volatile elements of the layered superlattice material, such as bismuth, can create deficiency of the volatile element in the layered superlattice material, leading to defects. Screening, surface roughness, and adhesion problems can also occur at these interfaces. Because of such problems, current ferroelectric memories are usually made with the ferroelectric material isolated from the conventional materials by a thick layer of an insulator. This, of course, increases the bulk of the memory and decreases the density of the memory.
Bismuth oxide and Sr
3
Bi
2
O
6
buffer layers between the layered superlattice materials and the conventional integrated circuit components have been proposed as a solution to the bismuth deficiency problem. See H. Yamawaki, S. Miyagaki, T. Eshita and Y. Arimoto, “Ultra Thin SrBi
2
Ta
2
O
9
Ferroelectric Films Grown By Liquid Source CVD Using BiOx Buffer Layers, in
Extended Abstracts of the
1998
Inter. Conf. on Solid State Devices and Materials,
1998, pp.102-103 and U.S. Pat. No. 6,194,227 B1 issued Feb. 27, 2001 to Takashi Hase. While analysis of the material has shown that such buffer layers solve the bismuth deficiency problem, the other problems mentioned above remained, and the electrical properties of the layered superlattice materials made with such buffer layers were not significantly improved.
It is also known that the polarizability of layered thin film superlattice materials is reduced if stoichiometric precursors are used, since some elements, such as bismuth, are more volatile and are disproportionately removed from the materials during drying and annealing. Therefore, precursors using excess amounts of these volatile elements are often used so that, after drying and annealing, the resulting material is approximately stoichiometric. Bismuth gradients have also been used to obtain essentially stoichiometric final layered superlattice materials. See, for example, U.S. Pat. No. 5,439,845 issued Aug. 8, 1995 to Watanabe et al. While the devices using a gradient show enhanced polarizability, they also must be relatively thick because of the multiple layers, resulting in lower density of the thin film superlattice material memory.
There remains a need to obtain greater residual polarization values and improve the interface between the thin film superlattice material layers and the conventional materials. The solution of these problems will result in increasing the density of thin film superlattice material memories and other integrated circuits that contain thin film superlattice materials as well as more reliable memories.
Solution
The present invention advances the art and overcomes the aforementioned problems by providing improved thin film ferroelectric devices having an enhanced magnitude of residual polarization. These improvements derive from the use of an interface buffer layer between the electrode and the thin film superlattice material layer. An integrated circuit memory device according to the invention includes a substrate supporting a thin film superlattice material layer. The thin film superlattice material layer is “interfaced” on one or both the top and bottom side by an interface buffer layer. Preferably, the interface buffer layer is a non-ferroelectric material. An interface buffer layer is preferably placed directly on
Ho Tu-Tu
Matsushita Electric Industrial Co Ltd.
Nelms David
LandOfFree
Integrated circuit device including a layered superlattice... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Integrated circuit device including a layered superlattice..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Integrated circuit device including a layered superlattice... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3098262