Structure and Fabrication Method for Non-Planar Memory Elements

1. Field of the Invention

This invention pertains to two fabrication methods for forming the dielectric structures in non-planar capacitors and ferroelectric memory cells, and to memory device structures utilizing these fabrication methods.

2. Background

The incorporation of high dielectric constant materials into small geometry capacitors suitable for Gigabit scale DRAM introduces fabrication challenges relating to topography, electrode material patterning, reaction of high-epsilon materials with Si contact and ultimate density/scalability. Similar challenges pertain to the fabrication of ferroelectric memory cells for ferroelectric RAM (FRAM) and other non-volatile RAM (NVRAM). To date, most fabrication strategies for making non-planar memory cells (a category of devices including both capacitors or "dielectric memory cells" and ferroelectric memory cells) is that the dielectric or ferroelectric deposition process be a conformal one which leaves a uniformly thin coating of film over all features of the sample topography. This requirement for a conformal process (such as chemical vapor deposition) tends to limit the use of promising spin-on deposition techniques (such as sol-gel) which have a tendency to gap fill and planarize.

An example of a fabrication method incorporating a sol-gel deposited cell dielectric into a three-dimensional memory device is found in U.S. Pat. No. 5,081,559, issued on Jan. 14, 1992 to Fazan et al. In this method, two electrodes are formed sequentially, prior to sol-gel deposition of the cell dielectric material, and the gap defined by a disposable sidewall spacer.

This invention describes two fabrication methods for making non-planar capacitors and ferroelectric memory cells, and memory device structures utilizing these fabrication methods. A key feature of these fabrication methods is that the ferroelectric or high-epsilon dielectric material is deposited to completely fill a cavity whose geometrical width is the sole determinant of the thickness of the electrically active portion of the ferroelectric or high-epsilon dielectric layer in the final device. In a first embodiment, the cavity into which the dielectric is deposited is defined by the gap between the plate and stack electrodes which are deposited and patterned in a through-mask plating step prior to the deposition of the memory cell dielectric or ferroelectric. This embodiment differs from typical memory cell designs in which at least one of the electrodes is put on after the dielectric or ferroelectric layer. Use of the invention makes it possible, for example, to build non-planar capacitors using relatively inexpensive and more easily developable dielectric deposition processes such as sol-gel. This is particularly advantageous when conformal deposition processes (e.g., CVD) do not exist for the desired dielectric material.

Another advantage of this embodiment of this invention is that pinholes in the dielectric will not lead to shorts between the electrodes, as would be the case if the second electrode were applied to a dielectric layer already in the structure as a coating on the first electrode.

In a second embodiment, the electrode adjacent to the device contact material is deposited after the deposition of the cell dielectric. In this embodiment, (i) the gap filled by the cell dielectric material is first defined by a disposable sidewall spacer, and (ii) the cell dielectric is deposited between the first electrode and a temporary "dummy electrode" which is replaced by the final electrode material after deposition of the cell dielectric. This embodiment has the advantage of minimizing the oxidation and contact material reactions that might occur were the final electrode material in place during cell dielectric deposition.

The ferroelectric or high-epsilon dielectric layer structures of the present invention have a thickness that is defined by the width of the cavity or gap into which the dielectric layer is to be deposited. When the dielectric is deposited between the plate and stack electrodes, the electrode height defines the width of the capacitor area, and the perimeter or boundary region between the plate and stack electrodes, which may take any of a number of shapes (circle, oval, hollow cross, etc.), defines the length of the capacitor area.

For an annular capacitor, the effective capacitor area is given by the product of the electrode height and the circumference of the annulus. The dielectric thickness is the same as the gap thickness.

This invention enables the fabrication of non-planar capacitors incorporating dielectrics deposited with non-conformal deposition processes such as sol-gel. This is important because sol-gel processes have a number of advantages over CVD, such as cost, ease of development, and availability for a wider range of dielectric materials. Combined with the electrode fabrication process of the present invention as described below, memory cell fabrication for DRAM, FRAM, and/or NVRAM can be relatively simple and inexpensive.

FIG. 1(a) is a cross-sectional view of one embodiment of the capacitor structure of the present invention showing the geometry of first and second electrodes.

FIGS. 1(b) and 1(c) are top views of two variations of the shape of the gap between the first and second electrodes of the structure of FIG. 1(a).

FIG. 2 is a cross-sectional view of another embodiment of the structure of FIG. 1 having an additional barrier layer beneath one of the electrodes.

FIG. 3 is a cross-sectional view of another embodiment of the structure of FIG. 1 having an additional dielectric layer above the first and second electrodes.

FIG. 4 is another embodiment of the structure of FIG. 3 in which the height of one of the electrodes is higher than the other electrode.

FIG. 5 is a cross-sectional view of another embodiment of FIG. 4 having an alternate shape for the second electrode.

FIGS. 6(a)-6(g) are cross-sectional views depicting various stages in one method of fabricating the structures of FIGS. 1-3.

FIGS. 7(a)-7(f) are cross-sectional views depicting various stages of a prior art method of fabricating a structure similar to that of FIG. 6(e).

FIGS. 8(a)-8(k) are cross-sectional views depicting various stages in a method of fabricating the structures of FIGS. 4 and 5.

The present invention is directed to a semiconductor electrical device comprising an insulating or semiconducting substrate, a first conductive region formed in the substrate from a first conducting material, and a first dielectric layer formed above the substrate. The first dielectric layer has a first opening or contact via above some portion of said first conductive region. The first opening is substantially filled with a second conductive material to make a "plug." A substantially planar layer is formed above the first dielectric layer, and the substantially planar layer contains separated first and second electrodes, made of a third conductive material.

The first or "plate" electrode surrounds and is isolated from the second or "stack" electrode. The second electrode is positioned over the contact via.

The first and second electrodes are separated from one another by a gap whose width is equal to the spacing of the electrode sidewalls, which are substantially vertical and transverse to the layer of third conductive material. A layer of second dielectric material fills the gaps between the electrodes.

In the fabrication of a memory device, a large number of these devices or memory cells are formed on a single substrate. In the formation of such a memory device, the plate electrodes will be electrically connected to each other, while the stack electrodes will be insulated from each other.

The steps to make the electrical device are such that the second dielectric material is deposited to completely fill a gap whose sidewalls are present when the second dielectric material is being deposited, and whose sidewalls define the width of the second dielectric material in the final structure, and further whose sidewalls may or may not be constructed from the final electrode materials.

In one embodiment of the present invention, the second electrode is made from the third conductive material, and the first electrode is made from a fourth conductive material, where the third and fourth conductive materials may be the same or different. In another embodiment, the first and second electrodes are made from a bilayer of two electrically conductive materials or a multilayer of two or more electrically conductive materials, the layers not necessarily being the same for both electrodes.

The first conductive material consists of the conductive elements of the semiconductor devices in the substrate. The second conductive material may consist of doped polysilicon, tungsten, or any suitable conductive material. The third and fourth electrode materials may be selected from the groups consisting of noble metals (such as Au, Pt, Pd, Ir, Rh), alloys of noble metals with noble or non-noble metals, metals whose oxides are conducting (such as Ru and Mo), electrically conducting oxides (such as RuO.sub.2, IrO.sub.2, Re.sub.2 O.sub.3, etc.), nitrides such as TiN or WN, electrically conductive, oxidation-resistant nitrides (such as TaN, TaSiN) and silicides (such as TaSi.sub.2) Electrodes being deposited after deposition of the second dielectric material may also be selected from the additional group of electrically conducting materials whose oxides are insulating.

The first dielectric material may substantially consist of SiO.sub.2, PSG (phosphosilicate glass), BPSG (borophosphosilicate glass), flowable oxide, spin-on-glass, or other conventional dielectric or combination of these. The second dielectric material is selected from the group consisting of ferroelectric, paraelectric, perovskite, pyrochlores, relaxors, layered perovskites, or any material with a dielectric constant greater than or equal to 20. Examples of such materials are Ta.sub.2 O.sub.5, (Ba,Sr)TiO.sub.3 (BST or BSTO), BaTiO.sub.3, SrTiO.sub.3, PbZr.sub.1-x Ti.sub.x O.sub.3 (PZT), PbZrO.sub.3, Pb.sub.1-x La.sub.x TiO.sub.3 (PLT), Pb.sub.1-x La.sub.x (Zr.sub.y Ti.sub.1-y).sub.1-x/4 O.sub.3 (PLZT) , and SrBi.sub.2 Ta.sub.2 O.sub.9 (SBT).

In another embodiment, a layer of an electrically conducting material, which is both a barrier to oxygen diffusion and a barrier to silicon or tungsten diffusion, is located at the top surface of the conductive plug. The diffusion barrier material may or may not be etch-resistant. Examples of possible barrier materials include TiN, Ta.sub.1-x Si.sub.x N.sub.y (with O<x<1 and y>1), and similar materials. Alternatively, this electrically conducting barrier material may be located at the bottom surface of the conductive plug, or replace the conductive plug entirely.

In yet another embodiment of the electrical device of the present invention, the first or second electrodes further include additional dielectric layers on or above their top surfaces, an example being excess second dielectric material left over from filling the gaps between the first and second electrodes.

In another variation, one or more additional dielectric layers may be placed between the first insulating material and the first (plate) electrode to function as an etch-stop, buffer layer, or adhesion promoter. One or more of these additional dielectric layers could be formed from a material such as SiO.sub.2 or SiN.sub.x, from an etch-resistant material such as Al.sub.2 O.sub.3 or HfO.sub.2, or from a material which might or might not be identical to the second dielectric material.

One electrical device made in accordance with the present invention is a capacitor having any of the cross-sectional structures described above where the sidewall gap filled with the second dielectric material may have the form of a circle, oval, hollow cross, etc. In such a capacitor, the sidewall gap can have a width in the range 300-1200 .ANG. and a height between 500 .ANG. and 3 .mu.m, preferably between 1000 .ANG. and 1 .mu.m. Additional devices formed by the structures described above include a capacitive memory element for DRAM, and a ferroelectric memory element for NVRAM or FRAM.

A method for fabricating the structures above in which the first and second electrodes are deposited at the same time may consist of the following process steps:

a) blanket deposition of an electrically conducting seed or "plating base" layer by a process selected from the group consisting of sputtering, evaporation, electroless deposition, or any other physical or chemical deposition process;

b) application and patterning of a first disposable material to define and temporarily fill the region that will later be the gap between the electrodes, this first disposable material being a photoresist, or x-ray, ion-beam or e-beam resist;

c) electroplating regions not covered by the first disposable material with an electrically conductive material ("third conductive layer"), which may or may not be the same as the plating base material, to form first (or plate) and second (or stack) electrodes;

d) removing the first disposable material from the gap between the newly formed electrodes;

e) electrically isolating the plated plate and stack electrode structures which are still electrically connected by the plating base remaining at the base of the gap between the electrodes. Electrical isolation can be achieved either by removing the plating base between the formed plate and stack electrodes, or by rendering it insulating. This portion of the plating base may be removed by a process selected from the group consisting of ion beam sputtering, RIE, and any other physical etching or chemical etching/reaction technique. Alternatively, said portion of the plating base may be rendered insulating by a process such as oxidation, anodization, etc.

Thereafter a non-conformal process such as sol-gel or a conformal process such as chemical vapor deposition is used to deposit a second dielectric material to fill the gaps between the plate and stack electrodes.

An alternative method for fabricating the structures described above allows one of the electrodes to be deposited after the deposition of the second dielectric. In this process, (i) the gap filled by the second dielectric material is first defined by a temporary sidewall spacer, and (ii) the second dielectric material is deposited between a first electrode and a temporary "dummy electrode." This method (when used in the plate electrode first, stack electrode last variation) has the advantages that electrode/contact material reactions are minimized (since the electrode adjacent to the contact plug material is formed after cell dielectric deposition) and that plug oxidation and cell dielectric/plug reactions are minimized (since the plug is protected by the material of the temporary dummy electrode during cell dielectric deposition).

The method includes blanket deposition of the third conducting material, an optional third dielectric material which may also function as a polish or etch-stop, and an optional first disposable material, by processes selected from the group consisting of sputtering, evaporation, or any other physical or chemical deposition process; anisotropic etching of the layered stack to leave a structure having the lateral dimensions of the plate (or stack) electrode by a process selected from the group consisting of RIE, ion beam etching, etc.; forming sidewall spacers from a readily etched second disposable material which may or may not be the same as the first disposable material, on the exposed vertical surfaces of the layered stack containing the formed plate (or stack) electrode; conformally depositing a third disposable material having the property of being resistant to the etching conditions used to remove the first and second disposable materials, to a thickness approximately equal to the layered stack thickness; planarizing the third disposable material to expose the second disposable material by a process such as CMP; removing the second and first disposable materials by a process such as wet etching, leaving the third disposable material; filling the gap between the plate (or stack) electrode and the third disposable material with the second dielectric material; planarizing the second dielectric material to expose the third disposable material; and replacing the third disposable material with a fourth conducting material, which may or may not be the same as the third conducting material, to make the stack (or plate) electrode.

In one embodiment, the first and second disposable materials are SiO.sub.2, and the third disposable material is silicon nitride. Alternatively, the first and second disposable materials are silicon nitride, and the third disposable material is SiO.sub.2. Also, the third dielectric material, which can also function as a polish or etch-stop, can be Al.sub.2 O.sub.3. In addition, the second electrode to be deposited can consist of one or more materials, one of which may or may not be TiN. Also, the third dielectric material can be a multilayer of one or more dielectric materials.

Turning now to the drawings, several embodiments of the structure and method of fabrication of the present invention are shown, in which like numerals will be used to represent like elements. FIG. 1(a) shows a cross-sectional view of one embodiment of the present invention. The capacitor structure 100 of this embodiment consists of a first or plate electrode 1 which surrounds a second or stack electrode 2. The first electrode 1 is separated from the second electrode 2 by a narrow gap filled with the capacitor dielectric 3. A conducting plug 4 embedded in a dielectric 5 provides contact to electrically conducting regions 6 on a substrate 20. The embodiment of FIG. 2 is similar to that of FIG. 1(a), differing only in that structure 102 includes the addition of a conductive barrier layer 7 at the top of the plug 4. The purpose of this barrier layer is to prevent out-diffusion of plug material into the electrode material, and to protect the plug material from reactions with oxygen. The remaining embodiments are illustrated only for the case of no barrier layer, but a barrier layer may be included in each if desired.

FIGS. 1(b) and 1(c) are top views of two embodiments illustrating variations in the manner in which first electrode 1 surrounds second electrode 2. The shape of the gap filled by dielectric 3 may be any shape as may be necessary according to the application and particular device being formed.

The structure 104 of FIG. 3 is an embodiment that is similar to that of FIG. 1(a), but differs in the presence of additional capacitor dielectric material 3 over the plate and stack electrodes. FIG. 4 shows a structure 106 similar to that of FIG. 3, differing only in that the top surface of the stack electrode 2 extends above the height of the plate electrode 1 and is approximately even with the capacitor dielectric material 3 overlying the plate electrode. In this embodiment, the lateral extent of the stack electrode is defined in a self-aligned process (e.g., chemical mechanical polishing). FIG. 5 shows a structure 108 similar to that of FIG. 4, differing only in that the lateral dimensions of the stack electrode's top surface are defined by a non-self aligned etching process.

The key process step in the fabrication of these capacitors is the filling of a pre-formed gap with the capacitor dielectric material, where the width of the gap defines the thickness of the capacitor dielectric. The gap may be formed between the first and second electrodes that will be in the final structure, or between one or more sacrificial materials and a single first or second electrode. Such a process allows one to sidestep the requirement that the capacitor dielectric be deposited by a conformal process such as chemical vapor deposition. A suitable deposition process for the embodiments of FIGS. 1-5 would be sol-gel deposition.

FIGS. 6-8 show the key steps for three methods of forming the gap structures that are filled with the capacitor dielectric in the fabrication of the capacitors of FIGS. 1-5. FIGS. 6(a)-6(g) show a plating method for forming a gap surrounded by both plate and stack electrodes. FIG. 6a shows a conductive plating base layer 8 deposited on a substrate 20 consisting of a conductive region 6, dielectric layer 5, plug material 4, and barrier material 7. FIG. 6b shows the structure after the deposition and patterning of a disposable material 9 to define and temporarily fill the region that will later be the gap between the plate and stack electrodes. This disposable material may be a polymer resist patterned by e-beam, ion-beam, or x-ray lithography, and should have a thickness that exceeds the thickness desired for the electrodes. FIG. 6c shows the structure after electroplating plate 1 and stack 2 electrodes. FIG. 6d shows the structure after removal of the disposable material, and FIG. 6e shows the structure after the removal of the exposed plating base. As shown in FIGS. 6(f) and 6(g), the capacitor dielectric 3 is deposited and planarized as desired to make the structures of FIGS. 2 or 3.

FIGS. 7(a)-7(f) (adapted from prior art) describe a temporary sidewall spacer method for forming a gap surrounded by plate and stack electrodes. The method is illustrated for the case of structures having the electrode and substrate geometries of FIGS. 1-6. FIG. 7a shows the deposition of a first electrode material 1 on a substrate 20 consisting of a conductive region 6, dielectric layer 5, plug material 4, and barrier material 7. FIG. 7b shows the structure after an etching step to form a cavity 10. FIG. 7c shows the structure after conformal deposition of a temporary spacer layer 11; FIG. 7d shows the structure after the temporary spacer layer has been anisotropically etched to form temporary sidewall spacers 11. FIG. 7e shows the structure after deposition and planarization of the second electrode material 2. FIG. 7f shows the structure after removal of the temporary spacer 11. The capacitor dielectric is then deposited and planarized as desired to make the structures of FIGS. 2 or 3.

FIG. 8 describes an alternative temporary sidewall spacer method which consists of forming both a gap surrounded by a plate or stack electrode and a temporary "dummy" electrode 13. The dummy electrode is replaced with the final electrode after deposition of the capacitor dielectric. The method is illustrated for the case in which the plate electrode is deposited first, but the order could be reversed. FIG. 8a shows the deposition of a first electrode material 1 and an overlayer of an additional disposable material 12 on a substrate 20 consisting of a conductive region 6, dielectric layer 5, and plug material 4. FIG. 8b shows the structure after an etching step to form a cavity 10. FIG. 8c shows the structure after conformal deposition of a temporary spacer layer 11; FIG. 8d shows the structure after the temporary spacer layer 11 has been anisotropically etched to form sidewall spacers 11. FIG. 8e shows the structure after deposition and planarization of the dummy electrode material 13. FIG. 8f shows the structure after removal of temporary material 12 and temporary spacer 11. After deposition of capacitor dielectric 3 (FIG. 8g), the structure is planarized to expose the dummy electrode 13 (FIG. 8h). The dummy electrode 13 is then removed (FIG. 8i) and replaced with the final electrode material 2 to make the structures of FIG. 8j or 8k, which correspond to FIGS. 4 and 5, respectively.

While the invention has been particularly shown and described with respect to preferred embodiments thereof, it will be understood by those skilled in the art that the foregoing and other changes in form and details may be made therein without departing from the spirit and scope of the invention.