WO2008039166A2

WO2008039166A2 - Non-volatile switching and memory devices using vertical nanotubes

Info

Publication number: WO2008039166A2
Application number: PCT/US2006/028464
Authority: WO
Inventors: Toshiharu Furukawa; Mark C. Hakey; Steven J. Holmes; David V. Horak; Charles W. Koburger, Iii.
Original assignee: International Business Machines Corporation
Priority date: 2005-07-26
Filing date: 2006-07-21
Publication date: 2008-04-03
Also published as: JP5004960B2; TW200719471A; JP2009516362A; WO2008039166A3; CN101542630A; US20080137397A1; KR20080055784A; EP1938330A2; US7352607B2; US7668004B2; KR101004650B1; US20070025138A1

Abstract

Non-volatile and radiation-hard switching and memory devices (225) using vertical nano-tubes (155) and reversibly held in state by van der Waals' forces and methods of fabricating the devices. Means for sensing the state of the devices include measuring capacitance, and tunneling and field emission currents.

Description

INVENTION TITLE

NON-VOLATILE SWITCHING AND MEMORY DEVICES USING VERTICAL NANOTUBES

TECHNICAL FIELD

The present invention relates to the field of non-volatile memory devices; more specifically, it relates to non-volatile switching and memory devices using vertical nanotubes and the method of fabricating non-volatile switching and memory devices using vertical nanotubes.

BACKGROUND ART

There is a continuing need to improve the performance, decrease the power consumption and decrease the dimensions of solid-state electronic devices, particularly those used as cells for memory devices and as switching devices. Further, as semiconductor device sizes decrease, various sources of radiation have been shown to cause changes in state of semiconductor-based memory and switching devices.

Therefore, there is a need for memory and switching devices that are both non-volatile and radiation hard.

DISCLOSURE OF INVENTION

A first aspect of the present invention is a structure, comprising: an insulating layer on a top surface of a substrate; an electrically conductive bitline formed in the insulating layer or on a top surface of the insulating layer, a top surface of the bitline parallel to the top surface of the substrate; a first electrically conductive wordline having a bottom surface, a top surface and a first sidewall; a second electrically conductive wordline having a bottom surface, a top surface and a second sidewall, the top and bottom surfaces of the first and second wordlines parallel to the top surface of the bitline, the first and second sidewalls about perpendicular to the top surface of the bitline, the first and second wordlines spaced apart, the first and second sidewalls facing each other; a dielectric layer between the bottom surfaces of the first and second wordlines and the top surface of the bitline; a dielectric first spacer on the first sidewall and a dielectric second spacer on the second sidewall; the first and second spacers spaced apart, the first and second spacers facing each other, and the top surface of the bitline exposed between the first and second spacers; and at least one electrically conductive nanotube having a first end and an opposite second end, the first end permanently attached to the bitline, the at least one nanotube extending away from the top surface of bitline.

A second aspect of the present invention is the first aspect of the present invention wherein the at least one nanotube is flexible and has a length between the first and second ends such that a portion of the one or more nanotubes proximate to the second end may reversibly contact either the first or the second spacers.

A third aspect of the present invention is the first aspect of the present invention wherein the at least one nanotube is reversibly held in contact with the first or second spacers by van der Waals' forces.

A fourth aspect of the present invention is the first aspect of the present invention wherein the at least one nanotube is a carbon nanotube.

A fifth aspect of the present invention is the first aspect of the present invention wherein the at least one nanotube is a single- wall carbon nanotube.

A sixth aspect of the present invention is the first aspect of the present invention further including: means for voltage biasing the first wordline and the bitline opposite to the second wordline and for voltage biasing the second wordline and the bitline opposite to the first wordline.

A seventh aspect of the present invention is the first aspect of the present invention further including: means for detecting a spike of current on the first or second wordline or on the bitline or means for sensing a change in capacitance between the first wordline and the bitline or between the second wordline and the bitline.

An eighth aspect of the present invention is the first aspect of the present invention further including a third spacer on top of the first spacer, a total thickness of the first and second spacers measured perpendicular to the first sidewall greater than a thickness of the second spacer measured perpendicular to the second sidewall.

A ninth aspect of the present invention is the first aspect of the present invention further including: means for sensing a tunneling current through the second spacer, the current flow between the second wordline and the bitline, the current flowing through the one or more nanotubes.

A tenth aspect of the present invention is the first aspect of the present invention further including: a first dielectric cap having a bottom surface, a top surface and a third sidewall, the bottom surface of the first dielectric cap in direct physical contact and coextensive with the top surface of the first wordline; a second dielectric cap having a bottom surface, a top surface and a fourth sidewall, the bottom surface of the second dielectric cap in direct physical contact and coextensive with the top surface of the second wordline, the third and fourth sidewalls facing each other, the first spacer extending over and in direct physical contact with the third sidewall and the second spacer extending over and in direct physical contact with the fourth sidewall; and an electrically conductive third spacer on the first spacer and an electrically conductive fourth spacer on the second spacer, the third and fourth spacers spaced apart, the third and fourth spacers facing each other, a bottom surface of the third spacer facing and overhanging the top surface of the bitline exposed between the first and second spacers, and a bottom surface of the fourth spacer facing and overhanging the top surface of the bitline exposed between the first and second spacers.

An eleventh aspect of the present invention is the tenth aspect of the present invention wherein: when an upper portion of the at least one nanotube proximate to the second end of the at least one nanotube is in contact with the first spacer, the second end of the at least one nanotube is positioned under but not touching a bottom surface of third spacer; and when the upper portion of the at least one nanotube proximate to the second end of the at least one nanotube is in contact with the second spacer, the second end of the at least one nanotube is positioned under but not touching the bottom surface of fourth spacer.

A twelfth aspect of the present invention is the eleventh aspect of the present invention further including: means for voltage biasing the first wordline and the third spacer opposite to the second wordline and the bitline and for voltage biasing the second wordline and the fourth spacer opposite to the first wordline.

A thirteenth aspect of the present invention is the eleventh aspect of the present further including: means for sensing a field emission current across a first gap between the second end of the at least one nanotube and the bottom surface of the third spacer when the upper portion of the at least one nanotube proximate to the second end of the at least one nanotube is in contact with the first spacer; and means for sensing a field emission current across a second gap between the second end of the at least one nanotube and the bottom surface of the fourth spacer when the upper portion of the at least one nanotube proximate to the second end of the at least one nanotube is in contact with the second spacer. A fourteenth aspect of the present invention is the first aspect of the present invention wherein the bitline comprises a catalytic material for the formation of carbon nanotubes.

BRIEF DESCRIPTION OF THE DRAWINGS

The features of the invention are set forth in the appended claims. The invention itself, however, will be best understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying drawings, wherein:

FIGs. I A through I G are cross-sectional views illustrating fabrication of a device according to a first embodiment of the present invention;

FIG. 2 is an isometric cross-section of a device according to the first embodiment of the present invention;

FIGs. 3A through 3G are cross-sectional views illustrating fabrication of a device according to a second embodiment of the present invention;

FIG. 4 is an isometric cross-section of a device according to the second embodiment of the present invention;

FIGs. 5A through 5K are cross-sectional views illustrating fabrication of a device according to a third embodiment of the present invention;

FIG. 6 is an isometric cross-section of a device according to the third embodiment of the present invention; and FIGs. 7, 8 and 9, are plan views illustrating memory arrays using devices according the embodiments of the present invention.

BEST MODE FOR CARRYING OUT THE INVENTION

Nanotubes are more correctly called fullerenes, which are closed-cage molecules comprised of atoms arranged in hexagons and pentagons. There are two types of fullerenes, namely closed spheroid cage fullerenes also called "bucky balls", and fullerene tubes. Fullerene tubes come in two types, single-wall fullerenes tubes, which are hollow tube-like structures or and multi-wall fullerene tubes. Multi-wall fullerenes resemble sets of concentric cylinders. Single- wall fullerenes are hereinafter called single-wall nanotubes (SWNT) and multi-wall fullerenes are hereafter called multi-wall nanotubes (MWNT).

While the present invention is described using electrically conductive single-wall and multiple-wall carbon nanotubes comprised of sp²-hybridized carbon, electrically conductive or semi-conductive single-wall and multiple wall nanotubes comprised of other electrically conductive or semi-conductive materials may be substituted for electrically conductive or semi-conductive single or multi-wall carbon nanotubes. For the purposes of the present invention, the term carbon nanotube (CNT) denotes either a carbon SWNT or a carbon MWNT unless otherwise specified.

CNTs used in the embodiments of the present invention are grown on electrically conductive bitlines formed on or embedded in an insulating layer by exposing bitlines to a vapor mixture of a CNT precursor and optionally a CNT catalyst at an elevated temperature. In one example, the CNT precursor is hydrocarbon or hydrocarbon isomer mixture and the bitline comprises iron (Fe), cobalt (Co), nickel (Ni) or other materials known in the art. In one example, formation of CNTs is performed at elevated temperatures between about 400 ⁰C to about 900 ⁰C.

When non-carbon SWNTs and MWNTs are substituted for CNTs, besides changes to reactants used to form the non-carbon SWNTs and MWNTs, appropriate changes to the composition of the bitline may be required, however, the material of the bitline remains an electrically conductive material.

FIGs. I A through I G are cross-sectional views illustrating fabrication of a device according to a first embodiment of the present invention. In FIG. I A, formed on a top surface of a substrate 100 is a first insulating layer 105. The top surface of substrate 100 defines a horizontal plane and a line perpendicular to the top surface of substrate 100 defines a vertical direction. Formed on a top surface of first insulating layer 105 is a bitline 1 10. Alternatively, bitline 1 10 may be damascened into first insulating layer 105, top surfaces of the first insulating layer 105 and bitline 1 10 being coplanar (see FIG. 2). Formed on a top surface of bitline 1 10 (and exposed top surface of first insulating layer 105) is a first dielectric layer 1 1 5.

In one example, first insulating layer 105 comprises Siθ2.

In one example, bitline 1 10 comprises Fe, Co, Ni, other conductive CNT-catalytic material, or combinations thereof. In one example, bitline 1 10 comprises a layer of Fe, Co, Ni₁ or other CNT catalytic material and combinations thereof. In one example, bitline 1 10 comprises a layer Fe, Co, Ni, or other CNT catalytic material over a layer or layers of tungsten (W), titanium (Ti), tantalum (Ta), titanium nitride (TiN), tantalum nitride (TaN), copper (Cu), aluminum (Al) or combinations thereof. In one example, first dielectric layer 1 1 5 comprises silicon nitride (SJsN₄).

In FIG. 1 B, a first electrically conductive wordline 120A and a second electrically conductive wordline 120B are formed on a top surface of first dielectric layer 1 1 5. First and second wordlines 120A and 1 20B are each covered by a respective dielectric cap 1 25 formed on respective top surfaces of the wordlines. First and second wordlines 120A and 120B and dielectric caps 125 may be formed, for example, by deposition of a conductive layer on the top surface of first dielectric layer 1 1 5, deposition of a capping layer on a top surface of the conductive layer followed by a photolithographic masking process and an anisotropic etch to define the first and second wordlines and dielectric caps.

A photolithographic masking process includes, applying a layer of photoresist, exposing the photoresist to actinic radiation through a patterned mask that will block the radiation from reaching regions of the photoresist layer, and developing the photoresist layer to generate a pattern of photoresist. After etching of underlying structure, the islands of photoresist are removed. In one example, first and second wordlines 1 2OA and 1 2OB comprise doped polysilicon, W, Ti, Ta, Cu, TiN, TaN, Al and combinations thereof. In one example, dielectric caps 125 comprise

Si₃N₄.

In FIG. 1 C, a conformal second dielectric layer 1 30 is formed over all exposed surfaces of first dielectric layer 1 1 5, first and second wordlines 1 2OA and 12OB, and dielectric caps 1 25. In one example, second dielectric layer 1 30 comprises SJsN₄. In one example, second dielectric layer 1 30 is a high K (dielectric constant) material, examples of which include but are not limited metal oxides such as Ta2θ5, BaTiO3, Hfθ2, Zrθ2, AI2O3, or metal silicates such as HfSi_xOy or HfSi_xOyNz or combinations of layers thereof. A high K dielectric material has a relative permittivity above about 10. In one example, second dielectric layer 1 30 is about 7 nm to about 20 nm thick..

In FIG. 1 D, a reactive ion etch (RIE) process is performed to form, from second dielectric layer 1 30 (see FIG. 1 C), dielectric sidewall spacers 1 35 on sidewalls of first and second wordlines 120A and 120B.

In FIG. 1 E, a second insulating layer 140 is formed. Top surfaces of second insulating layer 140 are coplanar with top surfaces of dielectric cap 125. In one example, second insulating layer 140 may be formed by deposition of a dielectric material to a depth greater than the distance between the top surface of first dielectric layer 1 1 5 and top surfaces of dielectric caps 125, followed by a chemical- mechanical-polish (CMP). In one example second insulating layer 140 comprises S1O2.

In FIG. 1 F, a trench 145 is formed between first and second wordlines 120A and 120B exposing a top surface 1 50 of bitline 1 10 by removing second insulating layer 140 and first dielectric layer 1 1 5 from between the wordlines. In one example, trench 145 is formed using a photolithographic masking process followed by a RIE.

In FIC. I G, one or more electrically semi-conductive or conductive CNTs 1 55 are grown on exposed top surface 1 50 of bitline 1 10 between first and second wordlines 1 20A and 1 20B. CNTs 1 55 have two opposite ends. The first ends are permanently attached to bitline 1 10 but the second ends are not permanently attached to any other structure. CNTs 1 55 extend upward from bitline 1 10 in the generally vertical direction. In one example, growth conditions for CNTs 1 55 are selected so as to grow at least one and up to a number of CNTs sufficient to cover exposed dielectric spacers 135 on one or the other of wordlines 120A or 120B with about a single layer of upper ends of CNTs 155. In one example, the growth of CNTs 1 55 is limited such that the CNTs do not extend above the top surfaces of dielectric caps 125. CNTs 1 55 are flexible so as to be able to bend and temporally touch sidewall spacer 1 35 on wordline 120A (shown) or touch sidewall spacer 135 on wordline 12OB. It is expected that single-wall CNTs will be more flexible than multi-wall CNTs.

Bitline 1 10, wordlines 120A and 120B, dielectric caps 125, dielectric spacers 1 35 and CNTs 1 55 comprise a switching or memory device (or memory cell) 225 according to the first embodiment of the present invention. Substrate 100 may be a semiconductor substrate, for example a bulk silicon or silicon-on-insulator (SOI) substrate, and include devices such as transistors, capacitors, resistors, diodes and inductors which are wired together to form support circuits for device 225.

Operation of device 225 is described infra, in relationship to FIG. 2, but a discussion of van der Waals' forces is required first. Though not entirely understood, in general, van der Waals' forces are attractive forces between molecules. Bonding in a molecule is caused by orbiting electrons. Any given electrons may be thought of being on one side or the other of a molecule in any one instance of time creating a surplus of negative charge on one side of the molecule and a lack of charge (positive charge) on the opposite side of the molecule, i. e. a dipole. When the dipoles on adjacent molecules are aligned positive pole to negative pole, negative pole to positive pole, there is a weak and transient electrostatic attraction. Since an object is made up of many molecules, there are always a finite number of pairs molecules having attracting dipoles. Van der Waals' forces are very small forces and can be easily broken, but absent an external force to force two objects apart, objects attached to one another by van der Waals' forces will remain attached. At the nano-scale, van der Walls; forces are significant forces.

Because van der Waals' forces do not require externally supplied power, the devices of the embodiments of the present invention when de-powered will retain the state in which they remained when last powered, and are thus non-volatile memory devices. Because van der Waals' forces are not effected by ionizing radiation, the devices of the embodiments of the present invention will retain their state even when struck by ionizing radiation and are thus radiation-hard devices.

FIG. 2 is an isometric cross-section of a device according to the first embodiment of the present invention. In FIG. 2, with a positive (negative) charge on first wordline 1 2OA and a negative (positive) charge on second wordline 12OB and bitline 1 10, CNTS 1 55 will become negatively (positively) charged and be attracted toward first wordline 1 20A. With sufficient voltage applied between first wordline 1 2OA and bitline 1 10, the upper ends of CNTs 1 55 will press against dielectric sidewall spacer 1 35 on first wordline 120A. With the voltage differential removed (first and second wordlines 1 20A and 1 20B and bitline all at the same potential), CNTS 1 55 will continue to stick to sidewall spacer 1 35 on first wordline 120A because of van der Wall attraction between molecules in CNTs 1 55 and molecules in dielectric sidewall spacer 1 35.

The location of CNTs 1 55 can be "flipped" by placing a positive (negative) charge on second wordline 120B and a negative (positive) charge on first wordline 120A and bitline 1 10, CNTS 1 55 will become negatively (positively) charged and be attracted toward second wordline 120B. With sufficient voltage, applied between second wordline 1 2OB and bitline 1 10, the upper ends of CNTs 1 55 will release from sidewall spacer 1 35 of first wordline 1 2OA and move to and press against sidewall spacer 1 35 on second wordline 1 2OB. With the voltage differential removed (first and second wordlines 1 2OA and 12OB and bitline all at the same potential), CNTS 1 55 will continue to stick to dielectric sidewall spacer 1 35 on second wordline 120B because of van der Wall attraction between molecules in CNTs 1 55 and molecules in dielectric sidewall spacer 1 35.

The state (whether CNTs are attached to dielectric sidewall spacer 1 35 of first wordline 1 20A or attached to dielectric sidewall spacer 1 35 of second wordline 120B) can be sensed as a change in capacitance in the wordlines or a spike in current flow through the bitline.

FIGs. 3A through 3G are cross-sectional views illustrating fabrication of a device according to a second embodiment of the present invention. The initial fabrication steps for the second embodiment of the present invention are that same as illustrated in FIGs. I A, I B, 1 C and 1 D and described supra. FIG. 3A is the same as FIG. I D.

In FIG. 3B, a photolithographic masking process followed by an isotropic etch is performed to remove dielectric sidewall spacer 1 35 from a sidewall 1 6OB of second wordline 1 20B that is adjacent to first wordline 120A, exposing sidewall 16OB of the first wordline 1 20B. Sidewall 16OA of first wordline 120A is still covered by sidewall spacer 1 35. In FIG. 3C, a conformal third dielectric layer 165 is formed over all exposed surfaces of first dielectric layer 1 1 5, dielectric caps 125, dielectric sidewall spacers 1 35, and exposed sidewall 16OB of second wordline 120B. In one example, third dielectric layer 165 comprises S13N4. In one example third dielectric layer 165 is a high K (dielectric constant) material, examples of which include but are not limited to metal oxides such as Ta2θs, BaTiθ3, Hfθ2, Zrθ2, AI2O3, or metal silicates such as HfSi_xOy Or HfSi_xOyNz or combinations of layers thereof. A high K dielectric material has a relative permittivity above about 10. In one example, third dielectric layer 165 is about 1 nm to about 5 nm thick.

In FIC. 3D, a RIE process is performed to form, from third dielectric layer 165 (see FIG. 3C), dielectric sidewall spacers 1 70 on dielectric sidewall spacers 1 35 of first wordline 12OA, dielectric sidewall spacer 135 of second wordline 12OB, and sidewall 16OB of second wordline 1 2OB.

In FIG. 3E, second insulating layer 140 is formed as described supra in reference to FIC 1 E. In FIC. 3F, trench 145 is formed between first and second wordlines 1 2OA and 1 2OB exposing a top surface 1 50 of bitline 1 10 as described supra in reference to FIG. 1 F. In FIG. 3G, one or more electrically conductive CNTs 1 55 are grown on exposed top surface 1 50 of bitline 1 10 between first and second wordlines 120A and 120B as described supra in reference to FIG. I G. Bitline 1 10, wordlines 12OA and 12OB, dielectric caps 125, dielectric sidewall spacers 1 35, dielectric spacers 1 70, and CNTs 1 55 comprise a switching or memory device (or memory cell) 230 according to the second embodiment of the present invention. Substrate 1 00 may be a semiconductor substrate, for example a bulk silicon or silicon-on-insulator (SOI) substrate, and include devices such as transistors, capacitors, resistors, diodes, and inductors which are wired together to form support circuits for device 230. Operation of device 230 is described infra, in relationship to FIG. 4.

FIG. 4 is an isometric cross-section of a device according to the second embodiment of the present invention. Change of state of device 230 is similar to that described for device 225 in reference to FIG. 2. However, the method of sensing the state of device 230 is different from that described for device 225 (see FIG. 2). The state (whether CNTs 1 55 are attached to dielectric spacer 1 70 on dielectric sidewall spacer 1 35 of first wordline 1 20A or attached to dielectric spacer 170 of second wordline 1 20B) can be sensed as a flow of tunneling leakage current (a flow of electrons through the dielectric) from second wordline 120B, through dielectric spacer 1 70 , through CNTs 1 55 to bitline 1 10 when CNTs 1 55 are attached to dielectric spacer 1 70 of second wordline 1 20B by van der Waals' forces. There is no current flow (or a much smaller current flow) from first wordliαe 1 20A to bitline 1 10 when CNTs 1 55 are attached to dielectric spacer 1 70 on dielectric sidewall spacer 1 35 of first wordline 120A. Tunneling leakage current is a flow of current through dielectric spacer 1 70 is similar to tunneling current flow in a field effect transistor (FET). Tunneling current flow in a FET is current flow from the gate, through the gate dielectric into the channel and thence to either the source or the drain. Tunneling leakage current is different from normal current flow from the source to the drain (or vice versa) in a FET when the gate of the FET is at the threshold voltage (VT) of the device. Tunneling leakage current is different from subthreshold leakage current flow from the source to the drain (or vice versa) in a FET when the gate of the FET is below threshold voltage (VT) of the device

Therefore, a dielectric spacer 1 70 should to be thin enough to allow tunneling leakage current while the thickness of a dielectric sidewall spacer 1 35 or the combined thickness of a dielectric spacer 1 70 and a dielectric sidewall spacer 1 35 should be thick enough to preclude tunneling leakage current or at least prevent it rising above a predetermined current level.

FICs. 5A through 5K are cross-sectional views illustrating fabrication of a device according to a third embodiment of the present invention. The initial fabrication steps for the third embodiment of the present invention are that same as illustrated in FIGs. I A, I B, 1 C and 1 D and described supra, with the exception that dielectric cap 125 of FIGs 1 B, 1 C and 1 D is replaced with a significantly thicker dielectric cap 1 75 in FIG. 5A. Otherwise FIG. 5A is similar to FIG. 1 D. In one example, dielectric cap 1 75 comprises SJ3N4 and is about 0.85 to about 1 .5 times as thick as wordlines 12OA and 120B as measured in a direction perpendicular to the top surface of substrate 100.

In FIG. 5B, a second insulating layer 1 80 is formed. A top surface of a second insulating layer 1 80 is coplanar with top surfaces of dielectric cap 1 75. In one example, second insulating layer 180 may be formed by deposition of a dielectric material to a depth greater than the distance between the top surface of first dielectric layer 1 1 5 and top surfaces of dielectric caps 1 75, followed by a CMP. In one example second insulating layer 1 80 comprises SiCh.

In FIG. 5C, a trench 1 85 is formed between first and second wordlines 120A and 120B exposing top surface 1 50 of bitline 1 10 by removing second insulating layer 180 and first dielectric layer 1 1 5 from between the wordlines. In one example, trench 1 85 is formed using a photolithographic masking process followed by a RIE.

In FIG. 5D, one or more electrically conductive or semi- conductive CNTs 190 are grown on exposed top surface 1 50 of bitline 1 10 between first and second wordlines 120A and 120B. CNTs 190 have two ends. The first ends are permanently attached to bitline 1 10 but the second ends are not permanently attached to any other structure. In one example growth conditions for CNTs 190 are selected so as to grow at least one and up to a number of CNTs sufficient to cover exposed dielectric sidewall spacers 1 35 on one or the other of wordlines 12OA or 1 20B with about a single layer of upper ends of CNTs 1 90. CNTs 1 90 may extend (as shown) above the top surfaces of dielectric caps 1 75 or may be shorter and not extend above the top surfaces of dielectric caps 1 75.

In FIC. 5E, the space (trench 1 85 of FIG. 5C) between wordlines 120A and 120B is filled with fill material 195 and a CMP performed to so a top surface of fill material 1 95 is coplanar with top surfaces of dielectric caps 1 75. This CMP also polishes away any portions of CNTs 190 that extended above top surface of dielectric caps 1 75, forming CNTs 19OA. In one example fill material 195 comprises poly-crystalline or amorphous germanium (Ge).

In FIG. 5F, upper portions of fill layer 1 95 are removed to form a trench 200 partially filled with fill material 195. Upper portions of second insulating layer 1 80 can also be removed. Next, CNTs 190A of FIG. 5E are also reduced in height, to the same height as the remaining portion of fill layer 195, to form CNTS 190B. These three operations may be performed as one, two or three distinct etch operations. In the case of two operations, the CNTs 190 may be etched along with the fill material 195, the CNTs 1 90 may be etched along with the second insulating layer 180, or the fill material 1 95 may be etched along with the second insulating layer 1 80.

In FIG. 5G, a conformal conductive layer 200 is formed over all exposed surfaces of second insulating layer 180, fill material 195, dielectric sidewall spacers 1 35, and dielectric caps 1 75. In one example, conductive layer 200 comprises WSi_x, TiSi2, TiN, TaN, doped polysilicon, or combinations thereof. In FIG. 5H, an RIE is performed to form conductive spacers 205 on exposed sidewall surfaces of dielectric sidewall spacers 1 35.

In FIG. 51, fill material 1 95 (see FIG. 5F) is removed. In the example of fill material 195 being Ge, an etch in aqueous hydrogen peroxide (H2O2) or other oxidizing solution may be used to remove the fill material 195. CNTs 1 9OB are now free-standing.

In FIG. 5J, an isotropic etch (for example a wet etch or a high pressure plasma etch) is performed to remove a small amount of material from conductive spacers 205 (see FIG. 51) generating conductive spacers 2O5A and a field emission gap 235 having of dimension G between bottom edges of conductive spacers 205A and top ends of CNTS 1 9OB. In one example, G is between about 4 nm and about 10 nm.

In FIG. 5K, CNTs 1 55 are flexible so as to be able to bend and temporally touch dielectric sidewall spacer 1 35 on first wordline 1 20A (shown) or touch dielectric sidewall spacer 1 35 on second wordline 1 20B. Bitline 1 10, wordlines 1 20A and 1 20B, dielectric sidewall spacers 135, dielectric caps 1 75, conductive spacers 205A, and CNTs 1 9OB comprise a switching or memory device (or memory cell) 240 according to the third embodiment of the present invention. Substrate 100 may be a semiconductor substrate, for example a bulk silicon or silicon-on-insulator (SOI) substrate, and include devices such as transistors, capacitors, resistors, diodes and inductors which are wired together to form support circuits for device 240. Operation of device 240 is described infra, in relationship to FIG. 6. FIC. 6 is an isometric cross-section of a device according to the third embodiment of the present invention. Change of state of device 240 is similar to that described supra for device 225 in reference to FIG. 2. However, the method of sensing the state of device 230 is different from that described for device 225 (see FIG. 2). The state (whether CNTs 1 90B are attached to dielectric sidewall spacer 1 35 of first wordline 12OA or attached to dielectric sidewall spacer 1 35 of second wordline 1 20B by van der Waals' forces) can be sensed as a flow of field emission current from bitline 1 10 to conductive spacer 2O5A of first wordline 120A to conductive spacer 2O5A of second wordline 120B. To enhance the amount of field emission current, polarities may be adjusted so that conductive spacers 205A are anodes and CNTs 19OB cathodes. Therefore it is useful to keep bitline 1 10 negative and opposite polarities on wordlines 1 20A and 1 20B, the conductive spacers associated with the positive wordline being the anode through which current will flow.

FIGs. 7, 8 and 9, are plan views illustrating memory arrays using devices according the embodiments of the present invention. FIG. 7 illustrates a first array of memory cells 210 using devices according to the first and second embodiments of the present invention. In FIG. 7, a first cell comprises wordlines WLl and WL2, CNTs CNTl and bitline BLl . A second cell comprises wordlines WL3 and WL4, CNTs CNT2 and bitline BLl . A third cell comprises wordlines WLl and WL2, CNTs CNT3 and bitline BL2. A fourth cell comprises wordlines WL3 and WL4, CNTs CNT4 and bitline BL2. To write the first cell of array 210, wordline WLl is brought up while wordlines WL2, WL3, and WL 4 are brought down. The state of BLl will then determine whether CNTs CNTl are attracted to, and attach via van der Waals' forces to, WLl or WL2. All cells between wordlines WLl and WL2 must be written simultaneously. While writing the first cell of array 210 WL3 and WL4 are shorted together so that the second cell is not disturbed.

FIG. 8 illustrates a second array of memory cells 21 5 using devices according to the first and second embodiments of the present invention. In FIG. 8, a first cell comprises wordlines WLl and WL2, CNTs CNTl and bitline BLl . A second cell comprises wordlines WL2 and WL3, CNTs CNT2 and bitline BLl . A third cell comprises wordlines WL3 and WL4, CNTs CNT3 and bitline BLl . A fourth cell comprises wordlines WL4 and WL5 (not shown), CNTs CNT4 and bitline BLl . A fifth cell comprises wordlines WLl and WL2, CNTs CNT5 and bitline BL2. A sixth cell comprises wordlines WL2 and WL3, CNTs CNT6 and bitline BL2. A seventh cell comprises wordlines WL3 and WL4, CNTs CNT7 and bitline BL2. An eighth cell comprises wordlines WL4 and WL5 (not shown), CNTs CNT8 and bitline BL2.

To write the second cell of array 21 5, wordline WLl and WL2 are brought up while wordlines WL3, WL4 and WL 5 are brought down. In the array of FIG. 8, in order not to disturb inactive wordline bits, all wordlines to the "left" of the active cell (i. e. wordline WLl ) are "shorted" to the active wordline, wordline WL2, and all wordlines to the "right" of the active cell (i. e. wordlines WL3 and WL4) are shorted together and to wordline WL2. (They are held at the opposite polarity to the "left-hand" wordlines).

FIG. 9 is similar to FIG. 7, except a third array of memory cells 220 comprises devices according to the third embodiment of the present invention. In FIG. 9, a first cell comprises wordlines WLI and WL2, CNTs CNTl , bitline BLl , and anodes (conductive spacers) Al and A2. A second cell comprises wordlines WLl and WL2, CNTs CNT2 bitline BL2, and anodes Al and A2. A third cell comprises wordlines WL3 and WL4, CNTs CNT3, bitline BLl , and anodes A3 and A4. A fourth cell comprises wordlines WL3 and WL4, CNTs CNT4, bitline BL2, and anodes A3 and A4. To write the first cell of array 220, wordline WLl is brought up, wordlines WL2, WL3 and WL 4 are brought down, and the bitline potential is set appropriately, depending on which "side" of the first cell CNTs CNTl are to attach. To read the first cell, bitline BLl is biased negatively, and other bitlines and anodes Al and A2 are biased positively. A tunnel current will become established only between CNTs CNTl and only the anode on the side of the first cell to which the CNTs CNTl are attached.

Devices according to the third embodiment of the present invention may be arranged into arrays similar to those depicted in FIGs. 7 and 8. The number of cells illustrated in FIGs. 7, 8 and 9 are to be taken as exemplary and any number of cells arranged in any number of rows and columns may be fabricated. Thus, the embodiments of the present invention provide memory and switching devices that are both non-volatile and radiation hard.

The description of the embodiments of the present invention is given above for the understanding of the present invention. It will be understood that the invention is not limited to the particular embodiments described herein, but is capable of various modifications, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, it is intended that the following claims cover all such modifications and changes as fall within the true spirit and scope of the invention.

Claims

What is clai med is : [Claim 1 ] 1 . A structure, comprising: an insulating layer on a top surface of a substrate; an electrically conductive bitline formed in said insulating layer or on a top surface of said insulating layer, a top surface of said bitline parallel to said top surface of said substrate; a first electrically conductive wordline having a bottom surface, a top surface and a first sidewall; a second electrically conductive wordline having a bottom surface, a top surface and a second sidewall, said top and bottom surfaces of said first and second wordlines parallel to said top surface of said bitline, said first and second sidewalls about perpendicular to said top surface of said bitline said first and second wordlines spaced apart, said first and second sidewalls facing each other; a dielectric layer between said bottom surfaces of said first and second wordlines and said top surface of said bitline; a dielectric first spacer on said first sidewall and a dielectric second spacer on said second sidewall said first and second spacers spaced apart, said first and second spacers facing each other, and said top surface of said bitline exposed between said first and second

spacers; and

at least one electrically conductive nanotube having a first end

and an opposite second end, said first end permanently attached to

said bitline, said at least one nanotube extending away from said top

surface of bitline.

[Claim 2] 2. The structure of claim 1 , wherein said at least one nanotube is flexible and has a length between said first and second ends such that a portion of said one or more nanotubes proximate to said second end may reversibly contact either said first or said second spacers.

[Claim 3] 3. The structure of claim 1 , wherein said at least one nanotube is reversibly held in contact with said first or second spacers by van der Waals' forces.

[Claim 4] 4. The structure of claim 1 , wherein said at least one nanotube is a carbon nanotube.

[Claim 5] 5. The structure of claim 1 , wherein said at least one nanotube is a single-wall carbon nanotube.

[Claim 6] 6. The structure of claim 1 , further including: means for voltage biasing said first wordline and said bitline opposite

to said second wordline and for voltage biasing said second wordline

and said bitline opposite to said first wordline.

[Claim 7] 7. The structure of claim 1 , further including: means for detecting a spike of current on said first or second wordline or on said bitline or means for sensing a change in capacitance between said first wordline and said bitline or between said second wordline and said bitline.

[Claim 8] 8. The structure of claim 1 , further including a third spacer on top of said first spacer, a total thickness of said first and second spacers measured perpendicular to said first sidewall greater than a thickness of said second spacer measured perpendicular to said second sidewall.

[Claim 9] 9. The structure of claim 8, further including: means for sensing a tunneling current through said second

spacer, said current flow between said second wordline and said

bitline, said current flowing through said one or more nanotubes.

[Claim 1 0] 10. The structure of claim 1 , further including: a first dielectric cap having a bottom surface, a top surface and a third sidewall, said bottom surface of said first dielectric cap in direct physical contact and coextensive with said top surface of said first wordline; a second dielectric cap having a bottom surface, a top surface and a fourth sidewall, said bottom surface of said second dielectric cap in direct physical contact and coextensive with said top surface of said second wordline, said third and fourth sidewalls facing each other, said first spacer extending over and in direct physical contact with said third sidewall and said second spacer extending over and in direct physical contact with said fourth sidewall; and an electrically conductive third spacer on said first spacer and an electrically conductive fourth spacer on said second spacer, said third and fourth spacers spaced apart, said third and fourth spacers facing each other, a bottom surface of said third spacer facing and overhanging said top surface of said bitline exposed between said first and second spacers, and a bottom surface of said fourth spacer facing

and overhanging said top surface of said bitline exposed between said

first and second spacers.

[Claim 1 1 ] 1 1 . The structure of claim 10, wherein: when an upper portion of said at least one nanotube proximate

to said second end of said at least one nanotube is in contact with said

first spacer, said second end of said at least one nanotube is

positioned under but not touching a bottom surface of third spacer;

and

when said upper portion of said at least one nanotube

proximate to said second end of said at least one nanotube is in

contact with said second spacer, said second end of said at least one

nanotube is positioned under but not touching said bottom surface of

fourth spacer.

[Claim 1 2] 12. The structure of claim 1 1 , further including:

means for voltage biasing said first wordline and said third spacer

opposite to said second wordline and said bitline and for voltage biasing said second wordline and said fourth spacer opposite to said first wordline.

[Claim 1 3] 1 3. The structure of claim 1 1 , further including: means for sensing a field emission current across a first gap between said second end of said at least one nanotube and said bottom surface of said third spacer when said upper portion of said at least one nanotube proximate to said second end of said at least one nanotube is in contact with said first spacer; and means for sensing a field emission current across a second gap between said second end of said at least one nanotube and said bottom surface of said fourth spacer when said when said upper portion of said at least one nanotube proximate to said second end of said at least one nanotube is in contact with said second spacer.

[Claim 1 4] 14. The structure of claim 1 , wherein said bitline comprises a catalytic material for the formation of carbon nanotubes.

[Claim 1 5] 1 5. A method, comprising: forming a first insulating layer on a substrate; forming an electrically conductive bitline on a top surface of said first insulating layer or forming an electrically conductive bitline in said first insulating layer such that top surface of said first insulating layer is coplanar with a top surface of said bitline; forming a dielectric layer on exposed surfaces of said first insulating layer and on exposed surfaces said bitline; forming a first electrically conductive wordline having a top surface, a first sidewall and forming a second electrically conductive wordline having a top surface and a second sidewall, said first and second wordlines spaced apart, said first and second sidewalls facing each other; forming a first dielectric cap in direct physical contact and coextensive with said top surface of said first wordline and forming a second dielectric cap in direct physical contact and coextensive with said top surface of said second wordline; forming a first dielectric spacer on said first sidewall and forming a second dielectric spacer on said second sidewall, said first and second spacers spaced apart, said first and second spacers facing

each other;

forming a second insulating layer on said first insulating layer, a

top surface of said second insulating layer coplanar with a top surfaces

of said first and second dielectric caps;

removing said second insulating layer and said dielectric layer

between said first and second spacers; and

growing at least one nanotube on a top surface of said bitline

exposed between said first and second spacers.

[Claim 1 6] 16. The method of claim 1 5, wherein said at least one nanotube is flexible and has a length between said first and second ends such that a portion of said one or more nanotubes proximate to said second end may reversibly contact either said first or said second spacers.

[Claim 1 7] 1 7. The method of claim 1 5, wherein said at least one nanotube is a carbon nanotube.

[Claim 1 8] 1 8. The method of claim 1 5, wherein said at least one nanotube is a single-wall carbon nanotube.

[Claim 1 9] 1 9. The method of claim 1 5, wherein said bitline comprises a catalytic material for the formation of carbon nanotubes.

[Claim 20] 20. The method of claim 1 5, further including, removing said second spacer and forming a dielectric third spacer on said first spacer and a fourth spacer on said second sidewall of said second wordline.

[Claim 21 ] 21. The method of claim 20, wherein: said first spacer has a first thickness in a direction perpendicular

to said top surface of said first wordline;

said third spacer has a second thickness in a direction

perpendicular to said top surface of said first wordline;

said fourth spacer has said second thickness in a direction

perpendicular to said top surface of said second wordline; and

said second thickness is less than said first thickness.

[Claim 22] 22. The method of claim 1 5, wherein:

said first dielectric cap has a third sidewall coextensive with first

sidewall of said first wordline, said second dielectric cap has a fourth

sidewall coextensive with second sidewall of said second wordline,

said third and fourth sidewalls facing each other, said first spacer

extending over and in direct physical contact with said third sidewall and said second spacer extending over and in direct physical contact with said fourth sidewall; and further including forming an electrically conductive third spacer on said first spacer and an electrically conductive fourth spacer on said second spacer, said third and fourth spacers spaced apart, said third and fourth spacers facing each other, a bottom surface of said third spacer facing and overhanging said top surface of said bitline exposed between said first and second spacers, and a bottom surface of said fourth spacer facing and overhanging said top surface of said bitline exposed between said first and second spacers.

[Claim 23] 23. The method of claim 1 5, further including after forming said at least one nanotube: filling the space between said first and second spacers with a fill material; removing a portion of said fill material to form a recessed top surface of said fill material below coplanar top surfaces of said first and second dielectric caps and a portion of said at least one nanotube proximate to said second end of said at least one nanotube to form a new second end of said at least one nanotube; forming an electrically conductive third spacer on said first spacer and an electrically conductive fourth spacer on said second spacer said third and fourth spacers spaced apart, side surfaces of said third and fourth spacers facing each other, a bottom surface of said third spacer in direct physical contact with said recessed top surface of said fill material, and a bottom surface of said fourth spacer direct physical contact with said recessed top surface of said fill material; removing all remaining fill material; and isotropically removing a layer of said third spacer from all exposed surfaces of said third spacer and isotropically removing a layer of said fourth spacer from all exposed surfaces of said fourth spacer.

[Claim 24] 24. A method, comprising providing a device comprising: a bitline; a first electrically conductive wordline having a bottom surface, a top surface and a first sidewall; a second electrically conductive wordline having a bottom surface, a top surface and a second sidewall, said top and bottom surfaces of said first and second wordlines parallel to said top surface of said bitline, said first and second sidewalls about perpendicular to said top surface of said bitline, said first and second wordlines spaced apart, said first and second sidewalls facing each other; a dielectric layer between said bottom surfaces of said first and second wordlines and said top surface of said bitline; a dielectric first spacer on said first sidewall and a dielectric second spacer on said second sidewall, said first and second spacers spaced apart, said first and second spacers facing each other, said top surface of said bitline exposed between said first and second spacers; and at least one electrically conductive nanotube having a first end and an opposite second end, said first end permanently attached to said bitline, said at least one nanotube extending away from said top surface of bitline; and electro-statically attracting said at least one nanotube to either said

first or to said second spacer, an upper portion of said at least one

nanotube proximate to said second end of said at least one nanotube

temporarily attached to either said first or second spacer by van der

Waals' forces; and

sensing to which of said first or said second spacers said at least

one nanotube is temporarily attached.

[Claim 25] 25. The method of claim 24, wherein said sensing includes measuring a capacitance change between said first wordline and said bitline or measuring a capacitance change between said second wordline and said bitline or measuring a spike in current flow on said first wordline, said second wordline or said bitline.

[Claim 26] 26. The method of claim 24, wherein said first spacer has a first thickness said second spacer has a second thickness, said second thickness less than said first thickness and wherein said sensing includes measuring a tunneling current flow between said second wordline and said bitline, said tunneling current flow passing through said second spacer, and said at least one nanotube.

[Claim 27] 27. The method of claim 24, wherein said device further comprises an electrically conductive third spacer on said first spacer and an electrically conductive fourth spacer on said second spacer, said third and fourth spacers spaced apart, said third and fourth spacers facing each other, a bottom surface of said third spacer facing and overhanging said top surface of said bitline exposed between said first and second spacers, and a bottom surface of said fourth spacer facing and overhanging said top surface of said bitline exposed between said first and second spacers.

[Claim 28] 28. The method of claim 27, wherein said sensing includes measuring a field emission current between in either said third or said fourth spacers and said bitline, said field emission current jumping a gap between said second end of said at least one nanotube and said third spacer when said at least one nanotube is temporarily attached to said first spacer and said a field emission current jumping a gap between said second end of said at least one nanotube and said fourth spacer when said at least one nanotube is temporarily attached to said second spacer. [Claim 29] 29. A device comprising: an array of cells, each cell comprising:

a first electrically conductive wordline having a bottom surface, a top

surface and a first sidewall;

a second electrically conductive wordline having a bottom surface, a

top surface and a second sidewall, said top and bottom surfaces of said first and second wordlines parallel to said top surface of said bitline, said first and second sidewalls about perpendicular to said top surface of said bitline, said first and second wordlines spaced apart, said first and second sidewalls facing each other; a dielectric layer between said bottom surfaces of said first and second wordlines and said top surface of said bitline; a dielectric first spacer on said first sidewall and a dielectric second spacer on said second sidewall, said first and second spacers spaced apart, said first and second spacers facing each other, said top surface of said bitline exposed between said first and second spacers; and at least one electrically conductive nanotube having a first end and an opposite second end, said first end permanently attached to said bitline, said at least one nanotube extending away from said top surface of bitline; means for selectively attracting for each cell each said at least one nanotube of to either said first or said second wordline; and means for selectively sensing for each cell which of said first or said second wordlines said at least one nanotube has been attracted to. [Claim 30] 30. The method of claim 29, wherein said means for sensing, senses a capacitance between particular wordlines and bitlines of said device, a spike in current flow on particular wordlines or bitlines of said device, a tunneling current flow between particular wordlines and bitlines of said device, or a current drain on particular bitlines of said device.