A control gate electrode and a memory gate electrode of a memory cell of a non-volatile memory are formed in a memory cell region of a semiconductor substrate, and a dummy gate electrode is formed in a peripheral circuit region. Then, n.sup.+-type semiconductor regions for a source or a drain of the memory cell are formed in the memory cell region and n.sup.+-type semiconductor regions for a source or a drain of MISFET are formed in the peripheral circuit region. Then, a metal silicide layer is formed over the n.sup.+-type semiconductor regions but the metal silicide layer is not formed over the control gate electrode, the memory gate electrode, and the gate electrode. Subsequently, the gate electrode is removed and replaced with the gate electrode for MISFET, Then, after removing the gate electrode and replacing it with a gate electrode for MISFET, a metal silicide layer is formed over the memory gate electrode and the control gate electrode.

A semiconductor device includes: a stack structure including conductive patterns and insulating layers, which are alternately stacked; a channel structure penetrating the stack structure; and a memory layer penetrating the stack structure, the memory layer being disposed between the channel structure and the stack structure. The memory layer includes memory parts and dummy parts, which are alternately arranged. Each of the memory parts includes a first part between the insulating layers and a second part between the dummy parts. The first part of the memory parts have ferroelectricity.

A method of manufacturing a memory structure including the following steps is provided. A first pad layer is formed on a substrate. Isolation structures are formed in the first pad layer and the substrate. At least one shape modification treatment is performed on the isolation structures. Each shape modification treatment includes the following steps. A first etching process is performed on the first pad layer to reduce a height of the first pad layer and to form first openings exposing sidewalls of the isolation structures. After the first etching process is performed, a second etching process is performed on the isolation structures to modify shapes of the sidewalls of the isolation structures exposed by the first openings. The first pad layer is removed to form a second opening between two adjacent isolation structures.

An operation method of a multi-bits read only memory includes a step of applying a gate voltage to a conductive gate, a first voltage to a first electrode, and a second voltage to a second electrode. The multi-bits read only memory of the present invention includes a substrate and a transistor structure with the conductive gate mounted between the first electrode and the second electrode. A multiplicity of M nanowire channels is mounted between the first electrode and the second electrode, and M is a positive integer greater than one. The present invention breaks multiple states of the multi-bits read only memory. The multiple states are programmable and include an i.sup.th state, and 1 <i <M . The aforementioned states allow storage of multiple bits on the read only memory, instead of just storing a single bit on the read only memory.

An operation method of a multi-bits read only memory includes a step of applying a gate voltage to a conductive gate, a first voltage to a first electrode, and a second voltage to a second electrode. The multi-bits read only memory of the present invention includes a substrate and a transistor structure with the conductive gate mounted between the first electrode and the second electrode, a first oxide located between the first electrode and the conductive gate, and a second oxide located between the second electrode and the conductive gate. The present invention creates an initial state wherein the transistor structure is not conducting, an intermediate state wherein the first oxide is punched through by the first voltage, and a fully opened state wherein both the first oxide and the second oxide are punched through. The aforementioned states allow storage of multiple bits on the read only memory.

The present invention provides a semiconductor device and a method of forming the same, and the semiconductor device includes a first insulating layer, a source and a drain, a stacked structure, a second insulating layer, and a gate. The first insulating layer is disposed on a substrate. The source and the drain are disposed on the first insulating layer, and the stacked structure is also disposed on the first insulating layer, between the source and the drain. The stacked structure includes a charge storage layer and an oxide semiconductor (OS) layer disposed on the charge storage layer. The second insulating layer covers the source, the drain and the OS layer. The gate is disposed on the second insulating layer.

A control gate electrode and a memory gate electrode of a memory cell of a non-volatile memory are formed in a memory cell region of a semiconductor substrate, and a dummy gate electrode is formed in a peripheral circuit region. Then, n.sup.+-type semiconductor regions for a source or a drain of the memory cell are formed in the memory cell region and n.sup.+-type semiconductor regions for a source or a drain of MISFET are formed in the peripheral circuit region. Then, a metal silicide layer is formed over the n.sup.+-type semiconductor regions but the metal silicide layer is not formed over the control gate electrode, the memory gate electrode, and the gate electrode. Subsequently, the gate electrode is removed and replaced with the gate electrode for MISFET. Then, after removing the gate electrode and replacing it with a gate electrode for MISFET, a metal silicide layer is formed over the memory gate electrode and the control gate electrode.

Memory cells having isolated charge sites and methods of fabricating memory cells having isolated charge sites are described. In an example, a nonvolatile charge trap memory device includes a substrate having a channel region, a source region and a drain region. A gate stack is disposed above the substrate, over the channel region. The gate stack includes a tunnel dielectric layer disposed above the channel region, a first charge-trapping region and a second charge-trapping region. The regions are disposed above the tunnel dielectric layer and separated by a distance. The gate stack also includes an isolating dielectric layer disposed above the tunnel dielectric layer and between the first charge-trapping region and the second charge-trapping region. A gate dielectric layer is disposed above the first charge-trapping region, the second charge-trapping region and the isolating dielectric layer. A gate electrode is disposed above the gate dielectric layer.

A semiconductor device and method of manufacturing the same are provided. In one embodiment, method includes forming a first oxide layer over a substrate, forming a silicon-rich, oxygen-rich, oxynitride layer on the first oxide layer, forming a silicon-rich, nitrogen-rich, and oxygen-lean nitride layer over the oxynitride layer, and forming a second oxide layer on the nitride layer. Generally, the nitride layer includes a majority of charge traps distributed in the oxynitride layer and the nitride layer. Optionally, the method further includes forming a middle oxide layer between the oxynitride layer and the nitride layer. Other embodiments are also described.

A semiconductor device includes a polysilicon substrate, a first oxide layer formed on the polysilicon substrate, an oxygen-rich nitride layer formed on the first oxide layer, a second oxide layer formed on the oxygen-rich nitride layer, and an oxygen-poor nitride layer formed on the second oxide layer.