Method for manufacturing a phase-change storage device with a column-shaped bottom electrode

Abstract

Methods for manufacturing a large number of memory cells, including: Providing a substrate (99) that includes an access circuit for the plurality of memory cells and that has a contact surface (100) with an array of conductive contact plugs (120, 141) connected to the access circuit; Forming a layer of bottom electrode material (200) on the contact surface (100) of the substrate (99); Formation of mask structures (201, 202) on the layer of bottom electrode material (200); Preparing the mask structures (201, 202) on the layer of bottom electrode material (200) by isotropic etching of the mask structures (201, 202) to create a pattern of prepared mask structures (201A, 202A); Removal of material from the layer of bottom electrode material (200) by anisotropic etching using the prepared mask structures (201A, 202A) as an etch mask to form a pattern of electrode columns (210, 211) on corresponding conductive contact plugs (120, 141) in the field of conductive contact plugs (120, 141); Forming a layer of dielectric material (212) covering the pattern of electrode columns (210, 211) and exposed sections of the contact surface (100); Planarizing the layer of dielectric material (212) and the electrode columns (210, 211) to create an electrode surface in which one top surface (222, 223) of each electrode column (210, 211) in the pattern of electrode columns is exposed; Forming a layer of programmable resistive material (230) on the electrode surface; Forming a layer of top electrode material (231) over the layer of programmable resistive material (230); and Patterns of the layer of programmable resistive material (230) and the layer of top electrode material (231).

Description

BACKGROUND OF THE INVENTION Area of ​​the invention

[0001] The present invention relates to methods for manufacturing high-density storage devices based on phase-change-based storage materials, including chalcogenide-based materials and other materials, and more specifically to a method for manufacturing a phase-change storage element with a column-shaped bottom electrode. Description of the relevant related technique

[0002] From publication US 6 800 563 B2, a phase-conversion storage cell with a conically shaped bottom electrode tapering upwards and an associated manufacturing process are known.

[0003] From publication DE 102 55 117 A1, a semiconductor storage device with phase conversion storage effect and a method for its manufacture are known, in which a cavity arrangement with at least one cavity in spatial proximity to the respective storage element is provided for each storage element in a semiconductor substrate in such a way that the thermal coupling of the respective storage element to the environment of the storage element is reduced by reducing the thermal conductivity between the storage element and the environment.

[0004] From the publication DE 10 2004 054 558 A1, a method for producing a resistively switching memory cell, in particular a phase-conversion memory cell, is known, wherein the method includes, among other things, the following steps: (a) structuring a hard mask applied above a layer and (b) etching back at least a part of the structured hard mask, in particular by isotropic etching.

[0005] From publication US 2003 / 0189200A1, a method for manufacturing a phase-conversion memory cell is known, in which the bottom electrode is formed as a layer on a pyramidal mandrel of dielectric material and the bottom electrode extends into the layer of phase-conversion material at the top of this pyramid.

[0006] From publication US 2006 / 0 154 185 A1, a method for patterning a resist layer is known with which smaller critical dimensions can be achieved.

[0007] From the publication EP 1 326 254 A1, a phase-conversion memory cell is known which has a cylindrical resistive element between a programmable element and a column-shaped plug.

[0008] Phase-change-based memory elements are widely used in random-access non-volatile memory cells. Materials such as chalcogenides and similar materials can be induced to change their phase between an amorphous and a crystalline state by applying an electric current at levels suitable for implementation in integrated circuits. The generally amorphous state is characterized by a higher resistivity than the generally crystalline state, which can be easily measured or read to display data.

[0009] The transition from the amorphous to the crystalline state generally occurs at low current. The transition from crystalline to amorphous, referred to here as reset, generally occurs at high current and involves a short pulse of high current density to melt or break up the crystal structure, after which the phase-change material cools rapidly, quenching the phase-change process and allowing at least part of the phase-change structure to stabilize in the amorphous state. It is desirable to minimize the strength of the reset current used to induce a transition of the phase-change material from a crystalline to an amorphous state.The strength of the required reset current can be reduced by decreasing the size of the phase change element in the cell and the contact area between the electrodes and the phase change material, so that higher current densities can be achieved with small absolute current values ​​through the phase change element.

[0010] One development is moving towards the formation of small pores in an integrated circuit structure and the use of small amounts of programmable resistive material to fill these small pores. Patents describing this development include: US Patent 5,687,112 A (Ovshinsky, "Multibit Single Cell Memory Element Having Tapered Contact"); US Patent 5,789,277 A (Zahorik et al., "Method of Making Chalogenide Memory Device"); and US Patent 6,150,253 A (Doan et al., "Controllable Ovonic Phase-Change Semiconductor Memory Device and Methods of Fabricating the Same").

[0011] Another memory cell structure under development, sometimes called a mushroom cell because of the shape of the active region at the bottom electrode in a typical setup, is based on forming a small electrode in contact with a larger section of phase-change material, and then a typically larger electrode in contact with an opposite surface of the phase-change material. Current flow from the small contact to the large contact is used to read, set, and reset the memory cell. The small electrode concentrates the current density at the contact point, so that an active region in the phase-change material is confined to a small volume near the contact point.

[0012] Siehe z. B. die Druckschrift von Ahn S. J. [u. a.]: Highly Reliable 50 nm Contact Cell Technology for 256Mb PRAM. In: 2005 Symposium on VLSI Technology Digest of Technical Papers, 2005, S. 98 - 99; die Druckschrift WO 2004 / 055 916 A2 (Denison, „Phase Change Memory and Method Therefor“; und die Druckschrift US 2005 / 0263829 A1 (Song et al., „Semiconductor Devices Having Phase Change Memory Cells, Electronic Systems Employing the Same and Methods of Farbricating the Same“.

[0013] A well-known technique for fabricating very small bottom electrodes, as described in the publication by Ahn et al., is referred to here as the plug-in-via method. It involves forming a dielectric filler layer over a circuit to access the memory cell, etching vias or contact holes into the dielectric filler layer to create an opening for contact with the circuit, and depositing electrode material in the contact hole. The resulting structure is then planarized to expose the electrode material in the contact hole. The phase-change material is deposited in contact with the electrode and patterned. Although this method is suitable for fabricating very small bottom electrode structures using plug-in vias, it has proven to be unreliable and inefficient enough.For example, it has proven difficult to establish reliable contact with the underlying access circuitry at the bottom of very small contact holes. This results in the field being constantly disconnected from the access circuitry in some cells.

[0014] Furthermore, Ahn et al. reported that in the plug-in via method, it is difficult to ensure that the areas of the exposed top surfaces of the plug electrodes are uniform over a large field of such cells after the planarization step. Since the area of ​​the top surface of the bottom electrode influences the current density in the phase-change material and is a critical dimension for phase-change cells of this type, the result is significant variation in the behavior of the cells within a single field. This problem is exacerbated by the methods used in attempts to successfully fill the contact holes, including the deposition of thin films and anisotropic etching of the thin films to form sidewall spacers with the contact holes.The process for forming sidewall spacers is such that it easily leads to the rounding of the upper edges of the contact holes, resulting in the electrode material plug in the contact hole having an upper end with an enlarged cross-section. Since it is difficult to planarize the resulting structure uniformly across the entire field with tolerances sufficient to avoid this enlarged upper end, the back-etching process does not completely remove the enlarged upper end in all cells, leading to significant size variation in the exposed top surface of the bottom electrode plugs.

[0015] Another problem arises in the formation of plug-in via electrodes, namely the difficulty of uniformly filling the contact holes. More precisely, due to the dynamics of thin-film deposition in small holes, the resulting plug can contain a void or vacancy where the top of the contact hole closes before it is completely filled. Planarizing the structure can open this gap, creating a hole in the top of the electrode plug. Such holes cause problems with the successful formation of a phase-change material layer over the electrode.

[0016] The task is to create a reliable method for manufacturing a memory cell structure with good control of the critical dimensions of the bottom electrode and the electrical integrity of connections to the bottom electrode, thereby enabling the manufacture of high-density integrated memory circuit devices. SUMMARY

[0017] The problem is solved by a method for producing a large number of memory cells according to claim 1. Further developments of the invention are specified in the dependent claims. BRIEF DESCRIPTION OF THE DRAWING Fig. Figure 1 is a perspective view of a phase-change storage cell that has a column-shaped bottom electrode. Fig. 2 is a cross-sectional view of a memory cell located in Fig. 1 is shown. Fig. Figure 3 is a cross-sectional representation of an alternative memory cell that includes a column-shaped bottom electrode based on a multilayer stack. Fig. Figure 4 shows a cross-sectional view of an alternative storage cell that includes a phase-change material deposited in a recess on the top of the column-shaped bottom electrode. Fig. Figure 5 is a cross-sectional representation of an alternative storage cell that includes a column-shaped bottom electrode with a larger surface area on the bottom than on the top. Fig. Figure 6 is a cross-sectional representation of a substrate with a contact surface with a field of contacts connected to an access circuit, in a first stage of a manufacturing process for producing phase-change memory cells with column-shaped bottom electrodes. Fig. Figures 7-15 show successive steps in a manufacturing process for producing phase-change storage cells with column-shaped bottom electrodes. Fig. Figure 16 shows a simplified circuit diagram for an array of phase-change storage cells with column-shaped bottom electrodes. Fig. Figure 17 is a block diagram of an integrated circuit device that includes an array of phase-change memory cells with column-shaped bottom electrodes. DETAILED DESCRIPTION

[0018] The following detailed description is given with reference to the figures.

[0019] Regarding the directional descriptions herein, the orientation of the figures determines their respective frame of reference, with "top," "bottom," "left," and "right" referring to the directions depicted in the respective figures. Similarly, "thickness" denotes a vertical dimension and "width" a horizontal one. These directions have no bearing on the practical orientation of the circuits.

[0020] Fig. Figure 1 is a perspective view of the basic structure of a phase-change memory cell, constructed as described herein, with the dielectric filler material omitted from the drawing. The memory cell is formed on a contact (10) with a contact surface 11. The contact (10) comprises a "plug" extending through an interlayer dielectric to an underlying (not shown) access circuit and, in the illustrated embodiment, is formed from tungsten or another conductive material. Other contact structures could also be used. A bottom electrode column 12 is formed on the contact surface 11. The bottom electrode column 12 has a very small contact area on one top surface. The very small contact area leads to a concentration of the current density during operation of the device, enabling low-power operation.A layer 13 of phase-change material is formed on the (not shown) contact surface of the bottom electrode column. A layer 14 of conductive material is formed on the layer 13 of the phase-change material to create a top electrode. The bottom electrode column 12 is fabricated using conductive materials such as titanium nitride or other conductive materials such as TaN, TiAlNi, W, WN, silicides, SiGe, silicon carbide, ruthenium oxide, nickel oxide, iridium oxide, LaNiO3, and other metal oxides and metal nitrides, which are selected for their compatibility with the contact surface 11 and with the phase-change material 13 and which preferably have a resistivity such that the column acts as a heater at the top surface that is in contact with the layer 13 of phase-change material.

[0021] According to the method described herein, the bottom electrode column 12 is fabricated by first depositing a layer of electrode material over the bottom contact, the positions of the columns being defined by a mask, and then removing the electrode material from the layered electrode material according to the mask, leaving the columns behind. This method enables reliable contact with the contact surface, uniform column structures over a large field, and consistent sizes of the column tops where contact with the phase-change material is established.

[0022] Fig. Figure 2 shows a cross-section of the structure of Fig. 1. The basic setup is shown, including the contact 10, the bottom electrode column 12, the phase-change material layer 13, and the conductive material layer 14, which provides a top electrode. The contact area 18 on the top of the bottom electrode column 12 is a small area where the current is concentrated. This results in a small active region 15 in the phase-change material layer 13, which, for heuristic reasons, can be described as having the shape of a mushroom. Thus, the phase-change storage cells of the type shown are referred to as "mushroom cells." Fig. Figure 2 also shows the dielectric material of the intermediate layer. The contact 10 in the illustrated embodiment comprises a plug formed in a contact hole in a dielectric layer 16. The dielectric layer comprises, for example, silicon dioxide, silicon oxynitride, silicon nitride, or other dielectric materials that separate the phase-change storage elements from the underlying access circuitry, as explained in more detail below. The column 12 is surrounded by a dielectric layer 17, which in one embodiment comprises a dielectric material such as silicon nitride or silicon oxynitride that also acts as a diffusion barrier, protecting the phase-change material and the underlying access circuitry from contamination by elements that could diffuse from other layers in the structure.

[0023] Layer 13 of phase-change material consists of a material capable of changing its local configuration within the active region of the cell between a first solid state, in which the material exists in a generally amorphous solid phase, and a second solid state, in which the material exists in a generally crystalline solid phase. The term "amorphous" is used to describe a relatively undetermined structure, less ordered than a single crystal, which exhibits observable properties such as higher electrical resistivity than the crystalline phase. The term "crystalline" is used to describe a relatively more ordered structure, more ordered than the amorphous structure, which exhibits observable properties such as lower electrical resistivity than the amorphous phase.

[0024] Some phase-change materials can electrically switch between different detectable states of their local configuration across the spectrum, ranging from fully amorphous to fully crystalline. Other material properties affected by the transition between amorphous and crystalline phases include atomic order, free electron density, and activation energy. The material can be switched either into different solid phases or into a mixture of two or more solid phases, creating a grayscale between fully amorphous and fully crystalline states. This grayscale enables the creation of multilevel memory cells capable of storing more than one bit per cell.

[0025] Phase-change storage cells can be switched from one phase state to another by applying electrical pulses. It has been observed that a shorter pulse with a higher amplitude causes the phase-change material to transition to a generally amorphous state. A longer pulse with a lower amplitude causes the phase-change material to transition to a generally crystalline state. The energy in a shorter pulse with a higher amplitude is strong enough to break bonds in the crystalline structure and weak enough to prevent the atoms from re-aligning into a crystalline state. Suitable pulse profiles, specifically tailored to a particular phase-change alloy, can be determined without unreasonable experimental effort.

[0026] Chalcogenides are suitable storage materials for use in embodiments of the invention. Chalcogenides include any of the four elements oxygen (O), sulfur (S), selenium (Se), and tellurium (Te) belonging to Group VI of the periodic table. Chalcogenides comprise compounds of chalcogen with a more electropositive element or residue. Chalcogenide alloys include combinations of chalcogenides with other materials, such as transition metals. A chalcogenide alloy typically contains one or more elements from column VI of the periodic table, such as germanium (Ge) and tin (Sn). Chalcogenide alloys often include combinations including one or more of antimony (Sb), gallium (Ga), indium (In), and silver (Ag).Many storage cells based on phase-change storage materials are described in the literature, including alloys such as: Ga / Sb, In / Sb, In / Se, Sb / Te, Ge / Te, Ge / Sb / Te, In / Sb / Te, Ga / Se / Te, Sn / Sb / Te, In / Sb / Ge, Ag / In / Sb / Te, Ge / Sn / Sb / Te, Ge / Sb / Se / Te, and Te / Ge / Sb / S. A wide range of alloy compositions from the Ge / Sb / Te family can be suitable. The compositions can be described as Te. a Ge b Sb 100....(a+b)The alloys can be characterized. One researcher described that the most suitable alloys had an average Te concentration in the deposited materials of well below 70%, generally below about 60%, and generally in the range of only about 23% to about 58% Te, and most preferably about 48% to 58% Te. Ge concentrations were above 5% and ranged from an average of only about 8% to about 30% in the material, generally remaining below 50%. Ge concentrations of about 8% to about 40% were most preferred. The remainder of the main constituent elements in this composition was Sb. These percentages are atomic percentages, representing 100% of the atoms of the constituent elements. (Ovshinsky patent '112, columns 10-11). Special alloys include Ge2Sb2Te5, GeSb2Te4 and GeSb4Te7.(Noboru Yamada, “Potential of Ge-Sb-Te Phase-Change Optical Disks for High-Data-Rate Recording”, SPIE v.3109, pp. 28–37 (1997)). More generally, a transition metal such as chromium (Cr), iron (Fe), nickel (Ni), niobium (Nb), palladium (Pd), platinum (Pt), and mixtures or alloys thereof can be combined with Ge / Sb / Te to form a phase-change alloy exhibiting programmable resistive properties. Specific examples of storage materials that may be suitable are given in Ovshinsky '112 in columns 11–13, and these examples are incorporated herein by reference. One material suitable for implementing a PCRAM cell as described herein is Ge₂Sb₂Tes, commonly referred to as “GST”.

[0027] Several structural variants of the bottom electrode column can be implemented. Representative examples of these variants are given in the Fig. 3, Fig. 4 and Fig. 5 shown, in which the reference figures that are in the Fig. 1 and Fig. 2 can be used for corresponding structures.

[0028] Fig. Figure 3 shows an embodiment in which the electrode column comprises a first segment 20 and a second segment 21, which are formed from a multilayer stack of electrode material. The material used for the first segment is selected for its compatibility with the material used on the contact surface 11 and for a relatively lower resistivity than that of the second segment. The material used for the second segment 21 is selected for its compatibility with the phase-change material 13 and for a relatively higher resistivity than that of the first segment 20. In an embodiment where the contact surface 11 comprises tungsten, the first segment 20 comprises titanium nitride and the second segment 21 comprises tantalum nitride.The second segment 21, due to its higher resistivity, which induces a greater temperature change for a given current flow than would otherwise occur, serves as a “heater”, thereby improving the efficiency of the phase change cycle in the phase change material 13.

[0029] Fig. Figure 4 shows another embodiment in which the electrode column has a first segment 25 comprising a conductor such as titanium nitride, and the second segment 26 comprising a phase-change material, for example, the same phase-change material used for layer 13. In some embodiments, the phase-change material in segment 26 is not the same as that used in layer 13. This setup can be produced by etching a recess on the top surface of the bottom electrode column after planarization. After layer 13 of phase-change material has been deposited, the recess is filled. In this embodiment, the current flow is further limited by segment 26 of the phase-change material, resulting in a reduction of the volume of the active region in the memory cell.

[0030] Fig. Figure 5 shows a further embodiment in which the bottom surface 32 of the bottom electrode column has a larger area in contact with the underlying contact 10 for the access circuit than the top surface 31 of the electrode column, which is in contact with the piece or patch of programmable resistive material. Fig. Figure 5 shows a smooth, conical taper of the column. Of course, columns with larger bottom surfaces than top surfaces can also have other shapes. In this way, the larger bottom surfaces result in a structure that is more robust during handling before and during the deposition of the dielectric filler layer 17 and provides a more reliable and less resistive electrical connection with the contact surface of the underlying contact 10. At the same time, the top surface 31 of the electrode column 30 can be kept quite small, which further concentrates the current flow for actuating the phase-change material. One method for forming a conical column involves using a mask structure that is also conical and etching it in such a way that the conical shape is transferred to the bottom electrode material.One embodiment of an electrode column with this structure can, for example, have a base area at least 30% larger than the top area. Of course, embodiments can be implemented that have any desired difference between the areas, including, for example, a base area only a few percent larger and a top area twice as large, in order to achieve one or more of the objectives described above.

[0031] The Fig. Figures 6-15 show a sequence of steps in an example process for manufacturing a phase-change storage cell with a bottom electrode column as above, with reference to the Fig. 1 - 5 described. Fig. Figure 6 shows a first stage of the fabrication process after typical CMOS processing techniques have been applied, whereby a substrate 99, which includes an access circuit, has been fabricated. The access circuit is formed on a semiconductor substrate 110. Insulating structures such as dielectric trenches 111 and 112 with flat trench insulation (STI) isolate pairs of rows of memory cell access transistors in this example. The access transistors consist of a common source region 116 in the substrate 110 and drain regions 115 and 117 in the substrate 110. Polysilicon word lines 113 and 114 form the gates of the access transistors. The dielectric filler layer 118 is formed over the polysilicon word lines 113 and 114. Contact plug structures 141 and 120 contact individual access transistor drains 115 and 117. A common source line 119 contacts source regions along a row of the field.The common source line 119 contacts the common source region 116. The substrate 99, which includes an access circuit for a plurality of memory cells, has a contact surface 100 with an array of conductive contacts 125, 126 on the contact tops of the contact plugs 121, 141, which in turn are connected to the access circuit. Other access circuit configurations can also be used, including configurations that use diodes instead of transistors.

[0032] Fig. Figure 7 shows the next stage in the manufacturing process after a layer 200 of electrode material has been deposited on the contact surface of the substrate 99. Depending on the embodiment of the desired bottom electrode column, the electrode material layer 200 can comprise a multilayer stack of conductive metals, alloys, semiconductors, phase-change materials, and the like. In one representative embodiment, the bottom electrode column comprises a single-layer column of titanium nitride deposited on the contact surface 100 of the substrate by a process such as physical or chemical vapor deposition. The thickness of the layer 200 is chosen to be slightly greater than the height of the bottom electrode columns to be produced in the described embodiment.

[0033] Fig. Figure 8 shows a next stage of the fabrication process after the deposition and patterning of a photoresist layer to create mask structures 201, 202 that define the location of the pillars above the conductive plugs 120 and 141. The deposition and patterning of a photoresist can be implemented using standard photolithography techniques, including phase-shift masking and other advanced patterning techniques, to form a field of small dots on the electrode material layer 200.

[0034] Fig. Figure 9 shows a further step after etching the mask structures 201 and 202 to create prepared mask structures 201A and 202A with sub-lithographic dimensions. For example, the photoresist mask structures 201 and 202 can be isotropically etched to narrow and thin the structures to form spots with a diameter of less than 50 nm, on the order of 30 nm or less in some embodiments. Photoresist preparation is applied, for example, using an oxygen plasma to isotropically etch the photoresist, thereby preparing its width and thickness. Alternatively, a hard mask layer, such as a low-temperature deposited layer of SiN₂, can be used. xor SiO2, defined by photolithography to define a pattern, followed by finishing by isotropic wet etching, such as dilute HF for silicon dioxide or hot phosphoric acid for silicon nitride, or reactive ion etching based on isotropic fluorine or HBr.

[0035] Fig. Step 10 represents a further step after etching the layer 200 of electrode material to form preliminary electrode columns 210 and 211 on the contact surface 100 over the plugs 120 and 141. A variety of etching recipes can be used to form uniform plugs of a constant diameter over a large field of structures. For example, a TiN layer can be etched using REI with a chlorine-based plasma, and a TaN layer can also be etched using REI with a chlorine-based plasma. An alternative method for forming a heating material layer over a layer of TiN or another metal involves first forming a column as described above, then filling and backpolishing, for example using a chemical-mechanical polish (CMP), to expose a top surface of the column.The top surface of the column can be regenerated using REI with reduced ion energy to regenerate the top surface without significant damage to the dielectric. The heating material is then deposited in a blanket over the resulting structure and again back-etched by CMP to create a heating layer over the electrode column. A similar process can be used for the embodiment of . Fig. 4 are applied, using a GST ceiling separation system. For the embodiment of Fig. 5. For example, a conical column can be formed using a hard silicon nitride mask on TiN by means of a chlorine-based REI etching, with oxygen being added to the etching chemistry, which attacks the hard mask laterally and slowly during the vertical etching of the electrode material, resulting in a slanted column profile.

[0036] Fig. 11 represents the next stage after the formation of the dielectric filling layer 212 over the preliminary column structures 210, 211. The dielectric filling comprises silicon nitride or silicon oxynitride or related dielectric materials deposited by processes such as high-density plasma chemical vapor deposition, based on silane and O2 at 400 to 450 °C, for SiN by adding ammonia to the silane, for oxynitride by using ammonia, silane and oxygen. The material of layer 122 also serves as a diffusion barrier.

[0037] Fig. Figure 12 represents the result of the following step, in which the dielectric filler layer 212 and the preliminary columns are back-etched using a planarization process such as chemical-mechanical polishing to form column structures 20 to 221 with top surfaces 222 and 223, respectively, exposed on a planar top surface 224 of the dielectric layer 212. A dielectric CMP process is chosen. During the CMP process, the thickness of the layer 122 is monitored, for example, using an ellipsometer, an interferometer, or another non-destructive tool. As a result of this process, the surfaces of the top surfaces 222 and 223 of the bottom electrode columns 220 and 221 are highly consistent across the large array of memory cells. In one example of the process, the resulting columns are about 40 to 120 nm high, preferably about 60 nm, and vary in height by less than 10% across the column pattern.

[0038] Fig. Figure 13 shows a subsequent step after the deposition of a layer 230 of phase-change material with a thickness of 40 to 140 nm, preferably about 10 nm, followed by the deposition of layer 231 over the top electrode material with a thickness of, for example, 40 to 80 nm. The top electrode material is selected with regard to conductivity and compatibility with the phase-change material. In one embodiment, the phase-change material comprises GST as described above, and the top electrode material comprises titanium nitride or other metals, metal oxides, and metal nitrides. In this step, the deposition of the dimensionally stable layer 230 of phase-change material on the surface 224 of the dielectric filling 212 forms a reliable electrical connection with the top surfaces of the columns 220 and 221.

[0039] One example method for the formation of chalcogenide material uses a PVD sputtering or magnetron sputtering process with source gas(es) Ar, N₂, and / or He, etc., at a pressure of 0.001333 mbar to 0.1333 microbar (1 mTorr to 100 mTorr). Deposition typically takes place at room temperature. A collimator with an aspect ratio of 1 to 5 can be used to improve the filling efficiency. A DC bias voltage of several tens to several hundred volts is also used to improve the filling efficiency. Likewise, a combination of DC bias and collimator can be used simultaneously.

[0040] A post-deposition annealing treatment in a vacuum or nitrogen environment is optionally performed to improve the crystallization state of the chalcogenide material. The annealing temperature is typically in the range of 100 °C to 400 °C, with an annealing time of less than 30 minutes.

[0041] Fig. Figure 14 shows the result of photolithographic patterning and etching of layer 230 of the phase-change material and layer 231 of top electrode material to form contact patches for individual cells 250, 251, each consisting of a patch 255 of phase-change material covered by a patch 256 of top electrode material with a dimension on the order of the minimum feature size of the lithographic process during fabrication. Thus, as a result of the patterning step, an array of memory cells is formed with a bottom electrode column, such as column 220, a layer of phase-change material, such as patch 255, and a top electrode layer, such as patch 256.

[0042] Fig. Figure 15 shows a subsequent stage of the fabrication process following a series of steps to complete the access circuit. These steps include the deposition of a dielectric filler layer 260 over the individual cells 250, 251. The dielectric filler layer 260 is deposited, for example, by high-density plasma (HDP) chemical vapor deposition (CVD), followed by chemical-mechanical polishing and cleaning. The dielectric filler layer 260 can comprise silicon oxides, silicon nitrides, and other insulating materials, preferably those with good thermal properties as well as good electrical insulating properties. In a subsequent step, contact holes are formed and filled with plugs 261, 262, which are in contact with the top electrode patches 256, 258 and each individual cell in the array.In one embodiment, the plugs comprise tungsten plugs with a diameter on the order of the minimum feature size of the lithographic process used. The formation of the plugs 261, 262 is followed by the deposition of a patterned conductive layer 270. In embodiments of the process, a copper alloy damascene metallization process is employed, in which the patterned conductive layer 270 is formed by depositing fluorosilicate glass (FSG) on the exposed surface and subsequently forming a photoresist pattern in the desired configuration. An etchant is applied, removing the exposed FSG, and then liner and seed layers are deposited in the pattern. A copper plating is then applied to fill the pattern. After plating, an annealing step is applied, followed by a polishing process.Other embodiments may utilize standard Al-Cu processes or other processes known in the art.

[0043] Fig. Figure 16 shows a circuit diagram for the basic array structure for memory cells with column-shaped bottom electrodes as described herein. Each memory cell in an array of the type shown includes an access transistor (or other access device such as a diode), four of which are shown as 1650, 1651, 1652, and 1653, and a phase-changing element, shown as 1635 and 1636, in cells coupled to access transistors 1650 and 1651. The sources of each access transistor 1650, 1651, 1652, and 1653 are connected to a common source line 1628. In one embodiment, the source lines of the selected devices are not electrically connected but can be individually controlled. Word lines 1623 and 1624 run parallel in a first direction.Word lines 1623 and 1624 are electrically connected to a word line decoder 1645, which is referred to as a "Y-decoder" in the figure due to its orientation. The gates of access transistors 1650 and 1652 are connected to a common word line 1623, and the gates of access transistors 1651 and 1653 are connected to a common word line 1624. A bit line 1641 is connected to the top electrodes of phase-changing elements 1635 and 1636. The bit line 1642 is connected to the phase-changing elements coupled to access transistors 1652 and 1653. Bit lines 1641 and 1642 are coupled to the measuring and reading amplifiers and the decoder 1606, which is referred to as the "X-decoder" in the figure due to its orientation. The column-shaped bottom electrodes of the phase-changing elements 1635 and 1636 are coupled to the drains 1632 and 1633 of transistors 1650 and 1651, respectively.It should be noted that four memory cells are shown to simplify the explanation, and that in practice the fields can be much larger.

[0044] In Fig.Figure 17 shows a simplified block diagram of an integrated circuit 1750, which includes an array 1700 consisting of phase-change memory cells with column-shaped bottom electrodes as described herein. The circuit 1750 has a memory array 1700 implemented using phase-change memory cells with column-shaped bottom electrodes. The array 1700 can contain one million cells. A word line or row decoder 1701 is electrically connected to a plurality of word lines 1702. A bit line or column decoder 1703 is electrically connected to a plurality of bit lines 1704 to read data from and write data to the phase-change memory cells in the array. Addresses are supplied to the word line decoder 1701 and the bit line decoder 1703 via the bus 1705.Measuring amplifiers and data input structures in block 1706 are coupled to a bitline decoder 1703 via a data bus 1707. Data is sent via the data input line 1711 from input / output ports on an integrated circuit 1750 or from other data sources inside or outside the integrated circuit 1750 to data input structures in block 1706. Other circuits (not shown) may be enclosed in an integrated circuit 10, such as a general-purpose processor or special-purpose circuit, or a combination of modules providing a system-on-chip function supported by the field 1700. Data is sent via a data output line 1715 from measuring amplifiers in block 1706 to input / output ports on the integrated circuit 1750 or to other data destinations inside or outside the integrated circuit 1750.

[0045] A controller 1709, implemented in this example and utilizing a bias arrangement state machine, controls the application of bias arrangement supply voltages 1708 to read, program, and verify the erasure of voltages. The controller 1709 can be implemented using a special logic circuit known in the art. In alternative embodiments, the controller 1709 includes a general-purpose processor, which can be implemented on the same integrated circuit, to execute a computer program to control the functions of the device. In other embodiments, a combination of a special logic circuit and a general-purpose processor can be used to implement the controller 1709.

[0046] The invention has been described with reference to phase-change materials. However, other storage materials, sometimes referred to as programmable materials, can also be used. As used in this application, storage materials are materials with electrical properties, such as resistance, that can be changed by the application of energy; the change can be a stepwise change, a continuous change, or a combination thereof. Other programmable resistive storage materials can be used in other embodiments of the invention, including N₂-doped GST, Ge x Sb yor another material that uses different crystal phase changes to determine the resistance; PrxCayMnO3, PrSrMnO, ZrOx, or another material that uses an electrical pulse to change the resistive state; TCNQ, PCBM, TCNQ-PCBM, Cu-TCNQ, Ag-TCNQ, C60-TCNQ, TCNQ doped with another metal, or any other polymeric material that exhibits a bistable or multistable resistive state controlled by an electrical pulse. Other examples of programmable resistive memory materials include GeSbTe, GeSb, NiO, Nb-SrTiO3, Ag-GeTe, PrCaMnO, ZnO, Nb2O3, and Cr-SrTiO3.

Claims

[1] Method for manufacturing a large number of memory cells, comprising: Providing a substrate (99) that includes an access circuit for the plurality of memory cells and that has a contact surface (100) with an array of conductive contact plugs (120, 141) connected to the access circuit; Forming a layer of bottom electrode material (200) on the contact surface (100) of the substrate (99); Formation of mask structures (201, 202) on the layer of bottom electrode material (200); Preparing the mask structures (201, 202) on the layer of bottom electrode material (200) by isotropic etching of the mask structures (201, 202) to create a pattern of prepared mask structures (201A, 202A); Removal of material from the layer of bottom electrode material (200) by anisotropic etching using the prepared mask structures (201A, 202A) as an etch mask to form a pattern of electrode columns (210, 211) on corresponding conductive contact plugs (120, 141) in the field of conductive contact plugs (120, 141); Forming a layer of dielectric material (212) covering the pattern of electrode columns (210, 211) and exposed sections of the contact surface (100); Planarizing the layer of dielectric material (212) and the electrode columns (210, 211) to create an electrode surface in which one top surface (222, 223) of each electrode column (210, 211) in the pattern of electrode columns is exposed; Forming a layer of programmable resistive material (230) on the electrode surface; Forming a layer of top electrode material (231) over the layer of programmable resistive material (230); and Patterns of the layer of programmable resistive material (230) and the layer of top electrode material (231). [2] Method according to claim 1, wherein the layer of bottom electrode material (200) includes a stack of materials, including a first material and a second material, wherein the second material forms electrical contacts with the programmable resistive material, and the first material forms electrical contacts with the conductive contact plugs (120, 141), and wherein the second material has a higher resistivity than the first material. [3] Method according to claim 1, wherein the layer of bottom electrode material (200) includes a stack of materials, including a bottom layer comprising titanium nitride and an upper layer comprising tantalum nitride. [4] Method according to claim 1, wherein the programmable resistive material (230) can switch between at least two solid phases by the application of heat or electric current. [5] Method according to claim 1, wherein the bottom electrode material comprises TiN. [6] Method according to claim 1, wherein for each electrode column (210, 211) the electrode column (210, 211) has a bottom surface which is in contact with the corresponding conductive contact plug (120, 141) from the field of conductive contact plugs (120, 140), and the bottom surface has a larger area than the top surface. [7] Method according to claim 1, which includes monitoring the thickness of the dielectric layer (212) during planarizing and stopping the planarizing as soon as a predetermined thickness is detected. [8] Method according to claim 1, which includes selectively removing a part of the top of the electrode column (210, 211) after planarizing, and wherein the formation of the layer of phase-change material includes filling a recess above the electrode column (210, 211) with phase-change material. [9] Method according to claim 1, wherein the patterning of the programmable resistive material layer (230) and the top electrode material layer (231) comprises forming contact spots, which include parts of the phase-change material layer and the top electrode material layer for individual memory cells, wherein the application of a dielectric filler layer (260) over the contact spots comprises forming contacts (261, 262) through the dielectric filler layer (260) and forming a patterned conductor layer (270) including bit lines that are in electrical contact with the contacts (261, 262).

Images