Stacked thin film transistor, non-volatile memory devices and methods for fabricating the same

[0001] This application claims the benefit of U.S. Provisional Application Ser. No. 60/748,911, entitled “The Process of TFT NAND and Nitride Read Only Memory,” filed Dec. 9, 2005, which is incorporated herein by reference as if set forth in full. This application is also related to U.S. patent application Ser. No. 11/425,959, entitled “A Stacked Non-Volatile Memory Device and Methods for Fabricating the Same,” Filed Jun. 22, 2006, and which is also incorporated herein by reference as if set forth in full.

[0002] 1. Field of the Invention

[0003] The embodiments described herein are directed to methods for fabricating a thin film transistor, non-volatile memory devices, and more particularly to methods for fabricating a thin film transistor, non-volatile memory device comprising multiple layers of memory cells.

[0004] 2. Background of the Invention

[0005] Non-volatile memory devices are finding uses in more and more products. For example, flash-based memory devices are being used in MP3 players, digital cameras, as storage devices for computer files, etc. As these uses grow, there is a need for larger memories housed in smaller packages. This requires the fabrication of highly dense memories. Accordingly, research and development has been directed at increasing the density of conventional non-volatile memory devices.

[0006] One approach for increasing the density of non-volatile memory devices is to create a stacked memory device, i.e., a device in which layers of memory cells are stacked on top of each other. Unfortunately, to date little effort has been put into creating certain types of stacked memory devices. For example, there are few stacked nitride read-only memory designs. This is in part because stacked memory devices are not necessarily compatible with the latest fabrication processes, which can make fabricating a stacked memory device inefficient and costly.

[0007] There are other approaches to increasing the density of conventional non-volatile memory devices; however, these approaches do not necessarily address the needs of all applications. Accordingly, there is still a need for further, or other approaches for increasing the density of conventional non-volatile memory devices.

[0008] One particular type of non-volatile memory device is the nitride read-only memory device. FIG. 1 is a diagram illustrating a conventional nitride read-only memory structure 150. As can be seen, nitride read-only memory 150 is constructed on a silicon substrate 152. The silicon substrate can be a P-type silicon substrate or an N-type silicon substrate; however, for various design reasons P-type silicon substrates are often preferred. Source/drain regions 154 and 156 can then be implanted in substrate 152. A trapping structure 158 is then formed on substrate 152 between source/drain regions 154 and 156. Control gate 160 is then formed on top of trapping layer 158.

[0009] Source/drain regions 154 and 156 are silicon regions that are doped to be the opposite type as that of substrate 152. For example, where a P-type silicon substrate 152 is used, N-type source/drain regions 154 and 156 can be implanted therein.

[0010] Charge trapping structure 158 comprises a nitride trapping layer as well as an isolating oxide layer between the trapping layer and channel 166 in substrate 152. In other embodiments, trapping structure 158 can comprise a nitride trapping layer sandwiched between two isolating, or dielectric layers, such as oxide, or more specifically silicon dioxide layers. Such a configuration is often referred to as an Oxide-Nitride-Oxide (ONO) trapping layer.

[0011] Charge can be accumulated and confined within trapping structure 158 next to source/drain regions 154 and 156, effectively storing two separate and independent charges 162 and 164. Each charge 162 and 164 can be maintained in one of two states, either programmed or erased, represented by the presence or absence of a pocket of trapped electrons. This enables the storage of two bits of information without the complexities associated with multilevel cell technology.

[0012] Each storage area in nitride read-only memory cell 150 can be programmed independently of the other storage area. A nitride read-only memory cell is programmed by applying a voltage that causes negatively charged electrons to be injected into the nitride layer of trapping structure 158 near one end of the cell. Erasing is accomplished by applying voltages that cause holes to be injected into the nitride layer where they can compensate for electrons previously stored in the nitride layer during programming.

[0013] A nitride read only memory device is constructed by manufacturing arrays of memory cells such as the cell illustrated in FIG. 1. Arrays are constructed by tying the cells together via word and bit lines. The bit lines are often polysilicon lines, while the word lines can be polysilicon or metal.

[0014] While nitride read-only memory devices, such as the device illustrated in FIG. 1, can be configured to store multiple bits per cell, the density of nitride read-only memory devices can be increased by using a stacked construction. Unfortunately, the stacking of nitride read-only memory devices is rarely done and when it is, the process can be inefficient and therefore more costly.

[0015] Methods for fabricating a stacked non-volatile memory device is disclosed. The methods disclosed use efficient processing techniques in order to fabricate a stacked device. Accordingly, the embodiments described herein can be scaled to achieve various levels of stacking.

[0016] In one aspect, a stacked nitride read-only memory can be fabricated using the methods described herein.

[0017] In another aspect, the stacked nitride read-only memory device can comprise thin film transistors (TFTs) fabricated using Silicon On Insulator (SOI) processing techniques.

[0018] In another aspect, a trapping layer included in the stacked non-volatile memory device can include one of a plurality of structures such as a SONOS, BE-SONOS, SONS, an Hi-K material, to name just a few.

[0019] In another aspect, a stacked memory device fabricated using the methods described herein can be configured for NAND operation.

[0020] These and other features, aspects, and embodiments of the invention are described below in the section entitled “Detailed Description.”

[0021] Features, aspects, and embodiments of the inventions are described in conjunction with the attached drawings, in which:

[0022] FIG. 1 is a diagram illustrating a conventional nitride read-only memory structure;

[0023] FIG. 2 is a diagram illustrating a stacked nitride read-only memory structure in accordance with one embodiment;

[0024] FIGS. 3-17 are diagrams illustrating an example progression of steps for fabricating the stacked nitride read-only memory of FIG. 2 in accordance with one embodiment;

[0025] FIGS. 18A-18H are diagrams illustrating different embodiments for a trapping structure that can be included in the device of FIG. 2;

[0026] FIGS. 19A and 19B are band diagrams for the trapping structure of FIG. 18C;

[0027] FIG. 20 is a diagram illustrating another example of a stacked non-volatile memory structure configured in accordance with one embodiment;

[0028] FIGS. 21-31 are diagrams illustrating processing steps comprising an example process for fabricating the device of FIG. 24 in accordance with one embodiment;

[0029] It will be understood that any dimensions, measurements, ranges, test results, numerical data, etc., presented below are approximate in nature and unless otherwise stated not intended as precise data. The nature of the approximation involved will depend on the nature of the data, the context, and the specific embodiments or implementations being discussed.

[0030] FIG. 2 is a diagram illustrating an example TFT, stacked nitride read-only memory 100 in accordance with one embodiment. In the example of FIG. 2, stacked nitride read-only memory 100 is fabricated on top of an insulating layer 102. Accordingly, device 100 is fabricated using SOI processing techniques. For example, device 100 can be fabricated using TFT processing techniques. A TFT is a special kind of field effect transistor made by depositing thin films for the metallic contacts, semiconductor active layer, and dielectric layer on an insulating layer. The channel region of a TFT is a thin film that is deposited onto a substrate that is often glass.

[0031] Successive bitline layers and wordline layers can then be fabricated on insulating layer 102. For example, in FIG. 2 a first bitline layer 110 is fabricated on insulating layer 102. A first wordline layer 120 is then fabricated on top of first bit layer 110. A second bitline layer 130 is then fabricated on top of first wordline layer 120. Finally, a second wordline layer 140 is fabricated on top of second bitline layer 130.

[0032] Further bitline and wordline layers can be successively fabricated on top of the layers illustrated in FIG. 1. Thus, two bitline layers and two wordline layers are shown for convenience only and the methods described herein should not be seen as limited to a certain number of bitline layers and/or wordline layers. Each bitline layer 110 and 130 comprises a plurality of bitlines 104 separated by insulating regions 106. Each wordline layer 120 and 140 comprises a wordline conductor 105 sandwiched between trapping layers 103 and 107.

[0033] By using the stacked configuration illustrated in FIG. 2, greater memory densities can be achieved. Further, as explained below, efficient processing techniques can be used to fabricate structure 100.

[0034] FIGS. 3-17 are diagrams illustrating an example sequence of steps for fabricating structure 100 in accordance with one embodiment. As illustrated in FIG. 3, a semiconductor layer 204 can be formed on an insulating layer 202. In certain embodiments, for example, insulating layer 202 can comprise an oxide material. Semiconductor layer 204 can comprise a P-type semiconductor material, such as silicon (Si), germanium (Ge), or silicon germanium (SiGe). It can be preferable, for example, for layer 204 to comprise a thin film polysilicon deposited on insulating layer 202. It will be understood that in other embodiments, semiconductor layer 204 can comprise N-type semiconductor material. A cap layer 206 can then be formed over semiconductor layer 204. In certain embodiments, for example, cap layer 206 can comprise a silicon nitride (SiN) material.

[0035] As illustrated in FIG. 4, conventional photolithography techniques can be used to pattern and etch layers 204 and 206. FIG. 5 is a diagram illustrating a top view of the layers comprising the device as fabricated to this point. FIG. 4 is a cross sectional view of FIG. 5 along the line AA′. Thus, as can be seen in FIG. 5, layer 206 and 204 have been patterned and etched into regions 205 that traverse insulating layer 202 from top to bottom. As will be explained below, regions 205 will form the bit lines of first bit line layer 110 shown in FIG. 2.

[0036] Referring to FIG. 6, a dielectric layer 209 can then be formed over insulating layer 202 as illustrated. Dielectric layer 209 can be, for example, an oxide, such as a silicon dioxide (SiO₂) layer and can be formed using High Density Plasma (HDP)—Chemical Vapor Deposition (CVD). Referring to FIG. 7, a portion of dielectric layer 209 is removed to expose the remaining portions of cap layer 206 and part of the remaining portions of semiconductor layer 204. For example, a conventional wet etching, i.e., isotropic, process can be used to remove a portion of dielectric layer 209. Removing the right amount of dielectric layer 209 can be achieved by having a high etching selectivity ratio between dielectric layer 209 and cap layer 206. The etching process produces dielectric regions 210 on top of cap layer 206 as well as dielectric regions 212 in between the remaining portions of semiconductor layer 204.

[0037] FIG. 8 is a diagram illustrating a top view of the layers as fabricated to this point. FIG. 7 is a cross-sectional view of the layers along the line AA′. Thus, as can be seen in FIG. 8, dielectric regions 212 now reside between regions 205. Dielectric regions 210 are illustrated as covering a portion of cap layer 206.

[0038] Referring to FIG. 9, the remaining portions of cap layer 206 can be removed, removing regions 210 of dielectric layer 209 in the process. For example, a hot phosphoric acid can be used to remove the remaining portions of cap layer 206. Regions 210 of dielectric layer 209 will automatically be removed during the removal of the remaining portions of cap layer 206, because portions 210 are disconnected from dielectric regions 212.

[0039] The process illustrated in FIG. 6-9 are described in U.S. Pat. No. 6,380,068, entitled “Method for Planarizing a Flash Memory Device,” assigned to the assignee of the present application, issued Apr. 30, 2002, and incorporated herein by references as if set forth in full. The process described in FIGS. 6-9 result in an efficient planarization of the remaining surfaces illustrated in FIG. 9. Accordingly, the fabrication processes described herein are compatible with newer, efficient processing techniques. This makes the fabrication of stacked non-volatile memory device efficient and cost-effective.

[0040] FIG. 10 is a top view of the layers formed thus far. FIG. 9 is a cross sectional view along the lines AA′ of the layers illustrated in FIG. 10. Accordingly, insulating layer 202 is now covered by alternating oxide regions 212 and bitlines 205 formed from the remaining portions of semiconductor material 204.

[0041] As illustrated in FIG. 11, a trapping structure 216 can then be formed over the remaining portions of semiconductor layer 204 and insulating regions 212. A word line conductor layer 218 can then be formed over trapping structure 216. A SiN layer (not shown) can then be formed over layer 218. The SiN layer (not shown) and layers 218 and 216 can then be patterned and etched using conventional photolithography techniques. The etching can be performed such that HDP oxide regions 212 act as a stop for the etching process. Another HDP oxide layer can then be formed over the etched layers, including the SiN layer (not shown). The HDP layer can then be partially etched and then part of the HDP oxide layer can be removed along with the remaining SiN layer (not shown) in a manner similar as that illustrated in FIG. 6-9, to form word lines 219 illustrated in FIG. 12.

[0042] In the example of FIG. 11, trapping structure 216 can comprise a multi-layered structure. Examples of multi-layered structures are described more fully below with respect to FIGS. 18A-18H. Thus, trapping structure 216 can be formed by sequentially forming the layers comprising trapping structure 216.

[0043] Word line conductor layer 218 can be formed from an N+ or P+ conductor material, such as a polysilicon material, polysilicon/silicide/polysilicon material, or a metal such as aluminum (Al), copper (Cu), or Tungsten (W).

[0044] FIG. 12 is a diagram illustrating a top view of the layers as formed thus far. Accordingly, wordlines 219 are illustrated as overlapping bit lines 205. FIG. 13 is a diagram illustrating a cross-sectional view of the layers illustrated in FIG. 12 along the lines AA′. FIG. 14 is a diagram illustrating a cross-sectional view of the layers illustrated in FIG. 12 along the line BB′.

[0045] As illustrated in FIG. 15, once wordlines 219 are formed over bit lines 205, source and drain regions 220 can be formed in the areas of semiconductor layer 204 comprising bit lines 205 that are not covered by word line conductors 218. Accordingly, source and drain regions 220 can be implanted and heat-driven in the regions 220 of semiconductor layer 204. As will be understood, the process of implanting source and drain regions 220 is a self-aligned process. In the example of FIG. 15, the source and drain regions should be N+regions formed using, e.g., arsenic (As), or phosphorous (P), since semiconductor layer 204 comprises P-type semiconductor material. It will be understood that P+regions should be formed in embodiments that use N-type semiconductor material.

[0046] After formation of source and drain regions 220, semiconductor layer 204 will comprise source/drain regions 220 which are N+ region, and P-type regions under word line conductor 218. As explained below, these P-type regions will form the channel regions for particular memory cells.

[0047] FIG. 16 is a cross-sectional view of the layers formed thus far along the line BB′. As can be seen, N+ source/drain regions 220 are formed between wordlines 219 and are separated by dielectric regions 212. FIG. 13 still illustrates a cross-sectional view of the layers formed thus far along the line AA′. As can be seen, layer 204 still comprises P-type regions 221 under word line conductor 218.

[0048] The layers formed thus far will then form non-volatile memory cells 240-256. The source and drain regions for non-volatile memory cells 240-256 are formed from the N+ source/drain regions 220 on either side of the associated wordlines 219. Referring to FIG. 13, the areas 221 of polysilicon layer 204 forming bit lines 205 under wordlines 219 form the channel regions for non-volatile memory cells 240-256. Trapping structure 216 above these channel regions are used to store charge in each of cells 240-256. The charge trapping structures are described more fully below with respect to FIGS. 18A-18H.

[0049] Accordingly, by applying the correct voltages to word line conductors 218 and source/drain regions 220, charge can be stored in trapping structure 216 of the appropriate cells 240-256. Similarly, cells 240-256 can be erased by applying the appropriate voltages to word line conductors 218 and the associated source/drain regions 220. The programming status of cells 240-256 can also be read by applying the appropriate voltages to the appropriate word line conductors 218 and source/drain regions 220.

[0050] As illustrated in FIG. 17, additional memory cells 260-270 can be formed by forming additional bit line and word line layers over bit line and word line layers 210 and 220 respectively. Thus, an additional trapping structure 222 can be formed over word line conductor 218, and then an additional bit line layer 230 can be formed on trapping structure 222. Bit line layer 230 can be formed using the same steps illustrated above for forming bit line layer 210. Accordingly, bit line layer 230 will comprise the remaining portions of an etched polysilicon layer 224, wherein the remaining portions are separated by dielectric regions 226. The remaining portions of polysilicon layer 224 illustrated in FIG. 17, which reside under the wordlines of word line layer 240, can form the channel regions for additional cells 260-270.

[0051] Source and drain regions can be implanted into the remaining portions of polysilicon layer 224, that reside on either side of the word lines of word line layer 240.

[0052] Word line layer 240 can be formed over bit line layer 230 by forming a trapping structure 228 over the remaining portions of polysilicon layer 224 and dielectric regions 226, forming a word line conductor layer 231 over trapping structure 228, and then forming a trapping structure 232 over word line conductor 231. Again, word line layer 240 can be formed using the same processing techniques described above with regard to word line layer 220. Memory cells 240-244 from FIG. 15 are illustrated in FIG. 17. Thus, additional memory cells 260-270 are formed in the layers above cells 240-244. It will be understood, however, that a device fabricated in accordance with the systems and methods described herein can comprise any number of layers and any number of cells.

[0053] Because of the efficient processing techniques used to form the word line and bit line layers, the process can be scaled to accommodate any number of layers. Thus, it will be understood that two bit line, and two word line layers are illustrated in FIG. 17 by way of convenience only.

[0054] FIGS. 18A-18H are diagrams illustrating example embodiments of various trapping structures that could be used in device 100. For example, referring back to FIG. 11, the structures illustrated in FIGS. 18A-18H can be used for trapping structure 216. The first example embodiment illustrated in FIG. 18A comprises a Silicon-Oxide-Nitride-Oxide-Silicon (SONOS) structure. This structure comprises an oxide layer 272, nitride layer 274, and oxide layer 276 sequentially formed over polysilicon layer 204. Oxide region 272 act as a tunnel dielectric layer and nitride layer 274 access a trapping layer for trapping charge. When the SONOS structure of FIG. 18A is used, charge is stored in trapping structure 274 of a particular cell by injection of holes into trapping layer 274. A cell can be erased through the direct tunneling of holes into trapping structure 274, where they compensate for any electrons previously stored in trapping structure 274. The tunneling of holes in the trapping structure 274 is achieved via Fowler-Nordheim tunneling. Oxide layer 272 can be a thin oxide layer, e.g., less than 3 nanometers thick. Cells formed using the SONOS trapping structure illustrated in FIG. 18A can, e.g., be used for NAND memory applications.

[0055] NAND devices constructed using the SONOS trapping structure illustrated in FIG. 18A, can exhibit somewhat poorer charge retention due to leakage current that results from the direct tunneling of holes into trapping layer 274 during charge retention.

[0056] FIG. 18B illustrates a nitride read only memory trapping structure. Again, the nitride read only memory trapping structure comprises an ONO structure formed by sequentially forming an oxide layer 278, nitride layer 280, and a second oxide layer 282, over polysilicon region 204. Here, however, oxide layer 278 comprises a thickness in the range of approximately 5-7 nanometers. A cell formed using the nitride read only memory structure of FIG. 18B is programmed via injection of electrons into layer 280. A cell formed using the nitride read only memory structure of FIG. 18B can then be erased via hot hole erase techniques. The nitride read only memory structure of FIG. 18B can be used for NOR applications; however, devices constructed using the nitride read only memory structure of FIG. 18B exhibits some degradation due to the hot hole erase procedure.

[0057] FIG. 18C is a diagram illustrating a band-gap engineered (BE)—SONOS structure. The BE-SONOS structure of FIG. 18C is fabricated by sequentially forming an ONO structure 294 followed by a nitride layer 290 and a dielectric layer 292. ONO structure 294 is thus formed by sequentially forming an oxide layer 284, nitride layer 286, and an oxide layer 288 over polysilicon layer 204. As with the SONOS structure of FIG. 18A, the BE-SONOS structure of FIG. 18C uses Fowler-Nordheim hole tunneling to erase the memory cells; however, the BE-SONOS structure of FIG. 18C does not exhibit the poor retention that results from direct tunneling leakage, or device degradation that results from hot hole erase damage. Further, the BE-SONOS structure of 18C can be used for both NOR and NAND applications.

[0058] FIGS. 19A and 19B are band diagrams illustrating the bands for ONO structure 294, of the BE-SONOS structure illustrated in FIG. 18C. FIG. 19A is a band diagram during data retention, and FIG. 19B is a band diagram during erase. As can be seen in FIG. 19A, during retention holes do not have sufficient energy to overcome the potential barriers of the layers comprising ONO structure 294. Data retention occurs when a low electric field exists across trapping structure 294. Because tunneling of holes is blocked by structure 294, there is little tunneling leakage during application of a low field. As illustrated in FIG. 19B, however, when a high field exists across trapping structure 294, the bands shift allowing holes to tunnel across structure 294. This is because the barriers presented by layers 286 and 288 are almost eliminated from the perspective of the holes, due to the band shift when a high field is present.

[0059] FIGS. 18D-18H illustrate other example structures that can be used for the trapping layers included in device 100. For example, FIG. 18D is a diagram illustrating a SONS structure that can be used for the trapping layers included in device 100. The structure illustrated in FIG. 18D comprises a thin oxide layer 302 formed over polysilicon layer 204. A nitride layer 304 is then formed over the thin oxide layer 302. Gate conducting layer 218 can then be formed over nitride layer 304. Thin oxide layer 302 acts as the tunnel dielectric and charge can be stored in nitride layer 304.

[0060] FIG. 18E is an example of a top BE-SONOS structure that can be used for trapping layers included in device 100. Accordingly, the structure illustrated in FIG. 18E comprises an oxide layer 306 formed over polysilicon layer 204. A nitride layer 308 is then formed over oxide layer 306, and ONO Structure 315 comprising an oxide layer 310, nitride layer 312 and oxide layer 314 is then formed over nitride layer 308. In the example of FIG. 18E, oxide layer 306 acts as the tunnel dielectric layer and charge can be trapped in nitride layer 308.

[0061] FIG. 18F is a diagram illustrating a bottom SONOSOS structure that can be used for the trapping layers included in device 100. The structure illustrated in FIG. 18F comprises an oxide layer 316 formed over polysilicon layer 204, and a nitride layer 318 formed over oxide layer 316. A thin oxide layer 320 is then formed over nitride layer 318 followed by a thin polysilicon layer 322. Another thin oxide layer 324 is then formed then over polysilicon layer 222. Accordingly, layers 320, 322 and 324 formed OSO structure 321 near gate conductor 218. In the example of FIG. 18F, oxide layer 316 can act as the tunnel dielectric and charge can be stored in nitride layer 318.

[0062] FIG. 18G is a diagram illustrating a bottom SOSONOS structure. Here, a thin OSO structure 325 is formed over polysilicon layer 204. OSO structure 325 comprises thin oxide layer 326, a thin polysilicon layer 228, and a thin oxide layer 330. A nitride layer 332 can then be formed over OSO structure 325, and an oxide layer 334 can be formed over nitride layer 332. In the example of FIG. 18G, OSO structure 325 can act as the tunnel dielectric and charge can be stored in nitride layer 332.

[0063] FIG. 18H is a diagram illustrating an example SONONS structure that can be used for the trapping structures included in device 100. Here, an oxide layer 336 is formed over polysilicon layer 204 and a nitride layer 338 is formed over oxide layer 336. An ON structure 341 is then formed over nitride layer 338. ON structure 341 comprises a thin oxide layer 340 formed over nitride layer 338, and a thin nitride layer 342 formed over thin oxide layer 240. In the example of FIG. 18H, oxide layer 336 can access the tunnel dielectric and charge can be trapped in nitride layer 338.

[0064] In other embodiments, the trapping structure can comprise a SiN or a SiON, or a Hi-K material such as HfO₂, Al₂O₃, AIN, etc. In general, any trapping structure or material can be used as long as it meets the requirements of a particular application.

[0065] FIG. 20 is a diagram illustrating an examples stacked Non-volatile memory device configured in accordance with another embodiment. FIGS. 21-31 are diagrams illustrating a progression of steps for fabricating the device of FIG. 20. The embodiment described with respect to FIGS. 20-31 present a simpler design in which word lines are not shared between memory cells. As can be seen in FIG. 20, the process illustrated in FIGS. 21-31 produces a stack memory structure that comprises an insulator or dielectric layer 2402 with word line and bit line layers stack on top of insulator 2402 and separated by inter layer, or inter-module dielectric layers 2404. The word line and bit line layers comprise bit lines 2410 separated from word lines 2406 by trapping structures 2408. As described below, a bit line layer can be deposited and then patterned and etched to form bit lines 2410. A trapping structure layer can then be deposited and a word line layer can be deposited over the trapping structure layer. The word line and trapping structure layers can then be patterned and etched to form word lines over bit lines 2410. The trapping structure 2408 above bit lines 2410 and under word lines 2406 can then act as the trapping layer for storing charge in a memory cell.

[0066] FIGS. 21-31 are diagrams illustrating an example process for fabricating the device illustrated in FIG. 20. As illustrated in FIG. 21, a polysilicon layer 2504 can be deposited over an insulating layer 2502. Insulating layer 2502 can comprise an oxide material, e.g., a silicon dioxide material (SiO). Polysilicon layer 2504 can have a thickness in the range of approximately 200-1,000 Å. For example, the thickness of polysilicon layer 2504 can, in certain embodiments, preferably be approximately 400 Å.

[0067] Referring to FIG. 22, polysilicon layer 2504 can then be patterned and etched using conventional photolithography processes in order to produce bit line regions 2506. For example, insulating layer 2502 can be used as an etch stop for the etching process in order to produce regions 2506. The overall thickness of the layers illustrated in FIG. 22 can be between approximately 200-1000 Å, and can prefer with the approximately 400 Å.

[0068] FIGS. 23A-23C, illustrate an alternative process for etching polysilicon layer 2504 in order to produce bit line regions 2506. Referring to FIG. 23A, a cap layer 2508 can be formed over polysilicon layer 2504. For example, cap layer 2508 can comprise a silicon nitride (SiN) layer. Polysilicon layer 2504 and cap layer 2508 can then be patterned and etched using conventional photolithography techniques as illustrated in FIG. 23B. Again, insulating layer 2502 can act as a etch stop for the etching process.

[0069] Referring to FIG. 23C, after layers 2504 and 2508 are etched to produce regions 2506 and 2510, and cap layer 2508, regions 2510 can be removed using conventional processes.

[0070] Referring to FIG. 24, a trapping structure layer 2508 can be formed over insulating layer 2502 and bit line regions 2506. As described above, trapping structure layer 2508 can comprise any of a plurality of trapping structures, such as SONOS, BE-SONOS, top BE-SONOS, SONONS, SONOSLS, SLSLNLS, etc. In other embodiments, trapping structure layer 2508 can comprise a SiN material, SiON material, or a high-K material such as HfO₂, Al₂O₃, AIN, etc.

[0071] Referring to FIG. 25, a word line layer 2510 can then be formed over trapping structure layer 2508. For example, word line layer 2510 can comprise a polysilicon material deposited over trapping structure layer 2508. Layers 2510 and 2508 can then be patterned and etched using conventional photolithography techniques. As illustrated in FIG. 27, this will produce word lines 2510 over bit lines 2506.

[0072] As can be seen in FIG. 26, the etching process can be configured such that it etches through trapping structure layer 2508 in the regions in between word lines 2510. This can produce regions 2506 with regions 2512 of trapping structure layer 2508 remaining on the sides of regions 2506.

[0073] FIG. 27 is a diagram illustrating a top view of the layers as formed thus far. FIG. 25 is a diagram illustrating a cross-sectional view of the layers illustrated in FIG. 27 along the lines AA′. FIG. 26 is a diagram illustrating a cross-sectional view of the layers illustrated in FIG. 27 along the line BB′.

[0074] Referring to FIG. 30, source and drain regions 2514 can be deposited in the areas of bit lines 2506 that are not under word lines 2510. For example, if word lines 2506 are formed from a P-type polysilicon material, then N-type source/drain regions 2514 can be implanted and heat driven into the regions of bit lines 2506 that are not under word lines 2510. Alternatively, if word lines 2506 are formed from an N-type polysilicon material, then P-type source/drain regions can be implanted and heat driven into bit lines 2506.

[0075] FIG. 28 is a diagram illustrating a cross-sectional view of the layers illustrated in FIG. 30 along the line AA′. FIG. 29 is a diagram illustrating a cross-sectional view of the layers illustrated in FIG. 30 along the line BB′. Accordingly, it can be seen that bit lines 2506 now comprised channel regions 2516 under word line layer 2510. The source and drain regions 2514 are then formed on either side of word lines 2510. It will be understood, that formation of source/drain regions 2514 is a self-aligned process.

[0076] Referring to FIG. 31, an interlayer, or inter-module dielectric layer 2518 can then be formed over word line layer 2510. Another bit line and word line layer can then be formed on top of interlayer, or inter-module dielectric 2518 using the same processing steps as described above. In this manner, any number of word line and bit line layers, separated by an interlayer, or inter-module dielectric 2518 can be formed over insulating layer 2502.

[0077] Referring to FIG. 30, memory cells 2520-2526 can then be formed in the structure illustrated. Memory cells 2520 and 2522 are illustrated in FIG. 31 as well. The source and drain regions for the memory cells are formed from source/drain regions 2514 on either side of the associated word lines 2510. The channel region is formed from the regions 2516 of bit lines 2506 under word lines 2510.

[0078] It will be understood that the cells illustrated in FIGS. 30 and 31 are tri-gate devices. Tri-gate devices can exhibit excessive corner effect, but can also have increase cell current due to increased device width.

[0079] While certain embodiments of the inventions have been described above, it will be understood that the embodiments described are by way of example only. Accordingly, the inventions should not be limited based on the described embodiments. Rather, the scope of the inventions described herein should only be limited in light of the claims that follow when taken in conjunction with the above description and accompanying drawings.