| Nor-type channel-program channel-erase contactless flash memory on soi -> Monitor Keywords |
|
Nor-type channel-program channel-erase contactless flash memory on soiNor-type channel-program channel-erase contactless flash memory on soi description/claimsThe Patent Description & Claims data below is from USPTO Patent Application 20090029511, Nor-type channel-program channel-erase contactless flash memory on soi. Brief Patent Description - Full Patent Description - Patent Application Claims This application is a divisional patent application of the U.S. patent application Ser. No. 10/781,112, filed on Feb. 18, 2004, titled “NOR-Type Channel-Program Channel-Erase Contactless Flash Memory on SOI”, the content of which is hereby incorporated by reference. TECHNICAL FIELDThe present invention relates in general to a nonvolatile semiconductor memory having an electric programmable/erasable function. In particular, the present invention relates to contactless memory arrays on Silicon-On-Insulator (SOI) for flash electrically erasable programmable read-only memory (EEPROM) devices. BACKGROUND OF THE INVENTIONThere are a widespread variety of flash memories with different cell structures, program/erase methods, and array organizations. Flash memories can be classified into two groups based on their program/erase methods: (1) employing channel hot electron (CHE) injection for programming and employing Fowler-Nordheim (FN) tunneling for erasing; and (2) employing Fowler-Nordheim (FN) tunneling for both programming and erasing. Method (1) is the most common method for flash memories, and particularly for ETOX (EPROM Tunnel Oxide) type flash memories. The CHE program consumes more than 300 μA per cell, hence only a few bits can be programmed at a time by an on-chip charge pump. To improve the hot electron generation efficiency, the drain junction needs to be an abrupt junction, and sometimes covered by a p+ region to enhance the impact ionization. The FN tunneling can be divided into source/drain edge FN tunneling and channel FN tunneling. The edge FN tunneling is to extract electrons from the floating gate (FG) by applying a negative voltage (e.g. −10 V) to the control gate (CG) and a positive voltage (e.g. 5 V) to the source or drain junction. The source or drain junction needs to be a lighter and deeper junction to (a) sustain a high voltage without breakdown, (b) reduce the band-to-band tunneling (BBT) current, (c) reduce the hot hole injection, and (d) increase the overlap area with the floating gate. The edge FN tunneling consumes about 10 nA per cell, which is mostly constituted of the substrate leakage current due to the band-to-band tunneling. All the memory cells relying on the edge effects (such as edge program and edge erase) require process optimization on the source/drain junctions to enhance the program/erase efficiencies. Such process optimization normally results in asymmetric source/drain junctions, which adds more complexity to the manufacturing process. Moreover, the endurance characteristics of the memory cells employing the edge program and/or edge erase are deteriorated with repeated program/erase cycles due to the trapped electron and/or holes in the tunnel oxide. The FN tunneling via the channel region consumes the least current, in the order of 10 pA per cell, among all program and erase mechanisms. Therefore, a large number of flash cells can be programmed and erased simultaneously by the on-chip charge pumps, which can be also smaller than the charge pumps used for edge program and edge erase. The memory cell employing uniform channel program and channel erase also shows the least deterioration in the endurance characteristics because the trapped carriers are neutralized by the alternative electric fields. Since the memory cell does not rely on the source/drain edge in program or erase, the source/drain junctions can be symmetric, which help to simplify the fabrication process. The physical dimensions of CMOS devices will be continuously scaled down in the future semiconductor technologies. The device dimensions of memory cells also need to be scaled down in the future flash memory technologies. Many of the challenges for bulk CMOS devices can be relaxed if the devices are fabricated on silicon-on insulator (SOI) wafers. A SOI flash memory technology has been proposed in U.S. Pat. No. 5,796,142 and U.S. Pat. No. 5,885,868 to achieve the goals of high density and low power consumption. The memory cell employs channel program and channel erase, which consume very low currents. The memory cell has a symmetric device structure. The memory cells are arranged in a NOR-type contactless flash memory array. Every two adjacent columns share the source/drain line in between. There is no field oxide within a memory array. The memory cell size (about 4 F2) is indeed very small, only about one third of a typical ETOX cell size (about 12 F2). Detailed disclosures in U.S. Pat. No. 5,796,142 and U.S. Pat. No. 5,885,868 are now discussed. FIG. 1 schematically shows the flash memory cell structure of U.S. Pat. No. 5,796,142 and U.S. Pat. No. 5,885,868. The memory cells are fabricated on a SOI wafer, which consists of a silicon substrate 10, an oxide layer 11, and a p-type doped silicon thin film. Each of the memory cell transistor is constituted of a tunnel oxide film 12, a first polysilicon (poly-1) floating gate 13, an oxide-nitride-oxide (ONO) insulating film 14, and a second polysilicon (poly-2) control gate (CG) 15. The n+ source/drain are formed by arsenic implantation into the p-type silicon thin film after the poly-1 floating gate 13 is patterned. The n+ source/drain is shared between two adjacent cells. FIG. 2 is a circuit diagram showing the memory array portion of the flash memory device disclosed in U.S. Pat. No. 5,796,142 and U.S. Pat. No. 5,885,868. The source lines and the drain lines are shared between two adjacent columns. The body line (e.g. BLm) of each column is isolated from the body lines (e.g. BLm−1 and BLm+1) of adjacent columns by the n+ source/drain lines and the oxide layer 11 beneath the p-type body. Memory cell program, erase, and read bias configurations are summarized in TABLE 1. Both program and erase cell operations are accomplished by the Fowler-Nordheim (FN) tunneling effect between the floating gate and the body. It is known that the FN tunneling current is much smaller than the hot-electron injection (HEI) current by orders of magnitude. FIG. 3a shows the cross section view of the program operation of a memory cell. To program a memory cell, a positive high voltage (e.g. 13 V) is applied to the word line and a negative high voltage (e.g. −7 V) is applied to the body line. According to the descriptions of the prior art, the memory cell is programmed by charging up the floating gate. The floating gate potential is coupled to the control gate voltage and the body voltage through the CG-to-FG and body-to-FG coupling coefficients. A voltage difference is therefore created between the floating gate and the body. Electrons are injected from the transistor body to the floating gate through the tunnel oxide by the Fowler-Nordheim tunneling effect. According to the suggested programming condition, the breakdown voltage of the source/drain-to-body junctions needs to be larger than 7 V. Such a large breakdown voltage imposes a serious limitation to scale down the physical dimensions of the memory cell for the future technologies.
Thank you for viewing the Nor-type channel-program channel-erase contactless flash memory on soi patent info. IP-related news and info Results in 3.55441 seconds Other interesting Feshpatents.com categories: Medical: Surgery , Surgery(2) , Surgery(3) , Drug , Drug(2) , Prosthesis , Dentistry orig |
* Protect your Inventions * US Patent Office filing
PATENT INFO |
|