摘要:
A method of manufacturing non-volatile memory devices, comprises the following steps:
depositing a first layer (3) onto a semiconductor substrate; defining and selectively removing said first layer (4) to form a portion (5) of said first layer (4); depositing a second layer (7) to a first thickness over the entire memory device; forming a screening layer (8) on the second layer (7) to leave uncovered at least a portion (10) of the second layer (7) aligned to the portion (5) of the first layer (3); and partly removing the portion (10) of the second layer (7) such that the thickness of the portion (10) of the second layer (7) is made smaller than the first thickness.
摘要:
A nonvolatile memory array has a plurality of PMOS two transistor (2T) memory cells. Each 2T cell (40) includes a PMOS floating gate transistor (40a) and a PMOS select transistor (40b) and is connected between a bit line and a common source line. The select gate and the control gate of each 2T cell in a common row are connected to a word line and to a control gate line, respectively. The 2T cells of the array are programmed using a combination of FN tunneling and BTBT induced hot electron injection, and are erased using FN tunneling. In some embodiments, the array is divided into sectors, where each sector is defined by an n-well region and includes a predetermined number of rows of the 2T cells. Here, the source of each 2T cell in a sector is coupled to a common source line of the sector. In other embodiments, the bit lines of the array are segmented along sector boundaries.
摘要:
To obtain a small contact-less memory device, a memory device includes a semiconductor chip having a first surface and a second surface located at a level lower than that of the first surface, a memory cell array formed on the second surface, a peripheral circuit, for operating the memory cell array, formed on the first surface, and a connecting portion, for electrically connecting the memory cell array to the peripheral circuit, formed on the first surface.
摘要:
A P-channel MOS memory cell has P+ source (14) and drain (16) regions formed in an N-well (18). A thin tunnel oxide (24) is provided between the well surface and an overlying floating gate (22). In one embodiment, the thin tunnel oxide (24) extends over a substantial portion of the active region (12) of the device. An overlying control gate (26) is insulated from the floating gate (22) by an insulating layer (28). The device is programmed via hot electron injection from the drain end of the channel region (12) to the floating gate (22), without avalanche breakdown, which allows the cell to be bit-selectable during programming. Erasing is accomplished by electron tunneling from the floating gate (22) to the N-well (18) with the source (14), drain (16), and N-well (18) regions equally biased. Since there is no high drain/well junction bias voltage, the channel length of the cell may be reduced without incurring and destructive junction stress.
摘要:
In accordance with the present invention, a silicon nitride layer (20) in a semiconductor device (10) is anisotropically etched selectively to both silicon dioxide, for example gate oxide layer (16), and to silicon, for example, silicon substrate (12) and polysilicon gate electrode (18). The silicon nitride layer is etched in a plasma etch system using CF₄, O₂, and argon gases. In other applications of the present invention, the etch method is used to remove an ONO dielectric stack and to remove a silicon nitride etch stop layer selectively to both active regions and isolation regions to form contacts or local interconnects across these regions.
摘要:
A process flow for fabricating a self-aligned stacked gate EPROM cell that uses a CVD tantalum oxide film to replace ONO as a control gate dielectric. Tungsten replaces polysilicon as the control gate. Both the dielectric deposition and cell definition steps of the process flow are performed in a back-end module to improve dielectric integrity in the memory cells by minimizing high temperature exposure of the tantalum oxide film.
摘要:
A contactless EPROM cell array in accordance with the present invention comprises an N+ source line formed in the silicon substrate. First and second N+ drain lines are formed in parallel with and spaced-apart from the source line on opposite sides of the source line. First and second field oxide strips are formed in parallel with but spaced-apart from the first and second drain lines, respectively, such that the source line/drain line structure is bounded on both sides by the first and second field oxide strips to separate this structure from adjacent similar source/drain line structures. First and second polysilicon 1 lines overlie the channel regions between the first drain line and the source line and the second drain line and the source line, respectively, and are separated therefrom by a first layer of dielectric material. A plurality of spaced-apart, parallel polysilicon 2 word lines overlie and run perpendicular to the first and second polysilicon 1 lines and are spaced-apart therefrom by a second dielectric material. Thus, the EPROM cells of the array are defined at each crossing of the polysilicon 1 lines and the polysilicon 2 word lines.
摘要:
A method for integrating Non-Volatile Memory (NVM) circuitry (18) with logic circuitry (20) is provided. The method includes depositing a first layer of gate material (16) over the NVM area and the logic area of a substrate (12). The method further includes depositing multiple adjoining sacrificial layers (22, 24, 26) comprising nitride, oxide and nitride (ARC layer) overlying each other. The multiple adjoining sacrificial layers (22, 24, 26) are used to pattern a select gate (16) and a control gate (32) of a memory transistor in the NVM area, and an ARC layer (22) of the multiple adjoining sacrificial layers (22, 24, 26) is used to pattern a gate (16) of a logic transistor in the logic area (20).