摘要:
A semiconductor process for fabricating NAND type flash memory devices in a first embodiment includes step which can be performed on a production line which manufactures NOR type flash memory products. A NAND flash memory fabrication process according to a second embodiment simplifies the process and uses fewer masks, thus reducing costs and errors to produce higher yields.
摘要:
In one embodiment, the present invention relates to a method of forming a flash memory cell, involving the steps of forming a tunnel oxide on a substrate; forming a first polysilicon layer over the tunnel oxide by chemical vapor deposition using a silicon containing gas and a mixture of a phosphorus containing gas and a carrier gas, the first polysilicon layer having a thickness from about 800 Å to about 1,000 Å; forming an insulating layer over the first polysilicon layer, the insulating layer comprising a first oxide layer over the first polysilicon layer, a nitride layer over the first oxide layer, and a second oxide layer over the nitride layer; forming a second polysilicon layer over the insulating layer; forming a tungsten silicide layer over the second polysilicon layer by chemical vapor deposition using WF6 and SiH2Cl2; etching at least the first polysilicon layer, the second polysilicon layer, the insulating layer, and the tungsten silicide layer thereby defining at least one stacked gate structure; and forming a source region and a drain region in the substrate, thereby forming at least one memory cell.
摘要:
Polystringers that cause NAND-type memory core cells to malfunction are covered by ONO fence material. ONO fence is removed so that polystringers may then be removed more readily. A SiON layer, tungsten silicide layer, second polysilicon layer, ONO dielectric, and first polysilicon layer are successively removed from between NAND-type flash memory core cells leaving ONO fence that shields some first polysilicon layer material from removal. The device is next exposed to an hydrogen-fluoride solution to remove oxide-based materials, particularly ONO fence. Thereafter, the polystringers are exposed and may thus be removed more readily.
摘要:
In one embodiment, the present invention relates to a method of forming a flash memory device involving the steps of forming a gate oxide layer on a substrate; forming a first poly layer over the gate oxide layer; forming an insulating layer over the first poly layer, the insulating layer comprising a first oxide layer over the first poly layer, a nitride layer over the first oxide layer, and a second oxide layer over the nitride layer; forming a second poly layer over the insulating layer; forming a tungsten silicide layer over the second poly layer; etching a portion of the tungsten silicide layer and the second poly layer, wherein in the etched portion at least about 20% of the second poly is not etched, thereby partially defining at least one stacked gate structure; etching at least a portion of the insulating layer and the unetched portion of the second poly layer thereby defining at least one select gate transistor structure; forming an interlayer dielectric layer over the select gate transistor structure and substrate; etching a contact hole through the interlayer dielectric layer to the first poly layer; and filling the contact hole with a conductive material to thereby form a flash memory device.
摘要:
A polysilicon-based floating gate is formed so as to be resistant to oxidation that occurs during multiple thermo-cycles in fabrication. Accordingly, edge erase times in NOR-type memory devices may be minimized. Additionally, manufacture of oxidation resistant floating gates reduces variations in edge erase times among multiple NOR-type memory devices. A layer of amorphous silicon is deposited on a silicon substrate by directing silane, a phosphene and helium gas mixture, and ammonia at the surface of the silicon substrate thereby doping the amorphous silicon in situ. The amorphous silicon layer is then etched so as to overlap slightly with regions that will later correspond to the source and drain regions. Next, a lower oxide layer of an ONO dielectric is deposited and the device is heated. A thermo-cycle is eliminated by heating the amorphous silicon during formation of the oxide layer rather than immediately following its deposition. Later, the nitride and oxide layers of the ONO dielectric, a second polysilicon layer, a tungsten silicide layer, and SiON layers are successively formed.
摘要:
A layer of polysilicon is deposited over an oxide layer on top of a silicon substrate, with core field oxide and active regions, and patterned. An oxide mask is then added. Next, the oxide mask and the layer of polysilicon are removed from above the core field oxide regions. Next, a second layer of polysilicon is deposited and etched to form polysilicon spacers. Later, an ONO dielectric, a third polysilicon layer, a tungsten silicide layer, and SiON layers are successively formed and patterned. The polysilicon spacers effectively seal any recesses that may occur in the edges of the first polysilicon layer to prevent harboring of subsequently added polysilicon material. Accordingly, NAND-type flash memory core cells cannot be electrically shorted by polysilicon material, so called "polystringers", present in such recesses.
摘要:
In one embodiment, the present invention relates to a method of forming a flash memory cell, involving the steps of forming a tunnel oxide on a substrate; forming a first polysilicon layer over the tunnel oxide; forming an insulating layer over the first polysilicon layer, the insulating layer comprising a first oxide layer over the first polysilicon layer, a nitride layer over the first oxide layer, and a second oxide layer over the nitride layer; forming a second polysilicon layer over the insulating layer; forming a tungsten silicide layer over the second polysilicon layer by chemical vapor deposition using WF6 and SiH2Cl2; etching at least the first polysilicon layer, the second polysilicon layer, the insulating layer, and the tungsten silicide layer thereby defining at least one stacked gate structure; and forming a source region and a drain region in the substrate, thereby forming at least one memory cell.
摘要:
This invention relates to a method for removing contaminate nitrogen from the peripheral gate region of a non-volatile memory device during production of said device, wherein at least some of the contaminate nitrogen has formed a bond with the surface of the silicon substrate in contact with the gate oxide layer in said gate region, said method comprising: contacting said gate oxide layer and contaminate nitrogen with a gas comprising ozone at a temperature of about 850° C. to about 950° C. for an effective period of time to break said bond; and removing said gate oxide layer and contaminate nitrogen from said surface of said silicon substrate.
摘要:
A method (200) of forming a NAND type flash memory device includes the steps of forming an oxide layer (202) over a substrate (102) and forming a first conductive layer (106) over the oxide layer. The first conductive layer (106) is etched to form a gate structure (107) in a select gate transistor region (105) and a floating gate structure (106a, 106b) in a memory cell region (111). A first insulating layer (110) is then formed over the memory cell region (111) and a second conductive layer (112, 118) is formed over the first insulating layer (110). A word line (122) is patterned in the memory cell region (111) to form a control gate region and source and drain regions (130, 132) are formed in the in the substrate (102) in a region adjacent the word line (122) and in a region adjacent the gate structure (107). A second insulating layer (140) is formed over both the select gate transistor region (105) and the memory cell region (111) and first and second contact openings are formed in the second insulating layer (140) down to the gate structure (107) and the control gate region, wherein a depth (X) through the second insulating layer (140) down to the gate structure (107) and down to the control gate region are approximately the same, thereby eliminating a substantial overetch of the gate structure contact opening.
摘要:
A semiconductor device having multiple layers uses different size contacts at different layer in order in order to simply the manufacturing process and the depth of etching required. Contact sizes are selected based on the responsiveness of the material to the etching process. Where a deep etch is required, a larger contact is used. A shallower etch through similar material uses a smaller contact to slow the etching process. As a result, the etches can complete at about the same time. The technique can be employed to etch any number of contacts. An intermediate size contact can be used where the material to be etched results in a slower etching process. A plurality of contact sizes can be used depending on the depths of etching required and the characteristics material to be etched, so that the etching for all the contacts completes at substantially the same time.