Flash cell device

1. Field of the Invention

The present invention relates generally to semiconductor memory devices, and more specifically to a memory device and the process of manufacturing such a device.

2. Description of the Prior Art

An electrically erasable programmable read only memory (EEPROM) cell is a nonvolatile writable and erasable memory cell which requires very low operating currents. The unit cell of an EEPROM may be formed by connecting a memory transistor in series with a select transistor. Some EEPROM designs are integrated so that the features of the two transistors are merged. Flash EEPROMs describe a family of single-transistor cell EEPROMs. Cell sizes of Flash EEPROMs are about half that of two transistor EEPROMs.

Flash memory designs differ in their cell structure based on whether they require one or several transistors per cell. Single-transistor self-aligned stacked-gate cells are well known. However, the single-transistor cell suffers from the possibility of over-erasure and consequent current leakage. A split-gate cell provides the equivalent of a two-transistor architecture, but requires only a little more chip area than a single transistor cell. Through a diffusion process, the split-gate creates a phantom transistor that looks like a series transistor. This allows the cell to be isolated from others in a column.

FIG. 1 shows a cross sectional elevation view of a conventional flash memory device at 10 , the memory device including: a first memory cell 12 and a second memory cell 14 formed over a substrate 16 , each of the cells sharing a common source region 18 ; and first and second drain regions 19 and 20 formed in the substrate 16 on opposite sides of the common source region 18 . The first and second memory cells 12 and 14 include a first select gate stack 22 and a second select gate stack 24 respectively, each of the select gate stacks 22 and 24 having: a tunnel oxide layer 26 formed over the substrate; a first conductive layer 28 formed over the tunnel oxide layer 26 and providing a select gate of the respective memory cell; and a first dielectric layer 30 formed over the first conductive layer 26 ; a first spacer 32 formed adjacent an outer sidewall of the corresponding stack 22 ; and a second spacer 33 formed adjacent an inner sidewall of the corresponding stack 22 .

Each memory cells 12 and 14 also include first and second floating gates 40 and 42 respectively. Each of the floating gates 40 and 42 including: a first portion formed over an outer portion of the common source region 18 ; a second portion formed superjacent an area of the substrate disposed between the common source region 18 and the corresponding one of the select gate stacks 22 and 24 ; and a third portion disposed over an inner portion of the corresponding one of the select gate stacks 22 and 24 . A dielectric layer 44 is formed over the first and second drain regions 19 and 20 , the first and second select gate stacks 22 and 24 , the first and second floating gates 40 and 42 , and the common source region 18 . The first and second memory cells 12 and 14 further include first and second control gates 48 and 50 respectively, each of the control gates being formed over portions of the dielectric layer 44 which overly a portion of the corresponding one of the floating gates 40 and 42 , and a portion of the corresponding one of the select gate stacks 22 and 24 .

There are several problems associated with the manufacturing and performance of the prior art flash memory device 10 . One disadvantage associated with the device 10 is that a relatively large area is required for fabricating each of the memory cells 12 and 14 on a semiconductor substrate, and therefore it is difficult to achieve very large scale integration of integrated circuits having such devices. Another disadvantage is that even though the memory cells 12 and 14 share a common source region 18 , the separate control gates 48 and 50 of the two memory cells 12 and 14 must be individually selected by a decoding means (not shown). Therefore, device performance suffers. A further problem associated with the prior art flash memory device 10 is that it is difficult to control the fabrication process or the device because of the overlapping of the outer portion of the floating gates 40 and 42 over the inner portions of the first and second select gate stacks 20 and 24 . Certain lithography steps required to form the device 10 are difficult to control to a degree of accuracy required to avoid shifting on the selective positions of the overlapping pairs of floating gates 40 and 42 as well as select gate stacks 22 and 24 . Such shifting of the selective positions of the overlapping floating gates and select gates can effect performance of the device 10 .

Fukumoto (U.S. Pat. No. 5,753,953, issued May 19, 1998) discloses a semiconductor storage device having a drain region and a source region formed in a silicon substrate, a select gate formed on the substrate between the source and drain regions, and a gate insulating film sandwiched between the select gate and substrate. On one side of the select gate, a floating-gate is formed out of a sidewall formed with an insulating film sandwiched. On the floating-gate and the select gate, a control gate is formed with an insulating film sandwiched. The insulating film directly below the floating-gate is formed as a tunnel oxide film which allows FN tunneling of electrons. In an erase operation, electrons are injected into the floating-gate from the silicon substrate, and in a write operation, electrons are extracted from the floating-gate to the drain region. A current required for writing and erasing each cell can be decreased, a low power supply can be used, and the lifetime of the tunnel insulating film can be increased. One disadvantage associated with the device described by Fukumoto is that adjacent memory cells have separate control gates which must be individually selected by a decoding means (not shown).

It is an object of the present invention to provide a flash semiconductor memory device including a pair of memory cells sharing a common source region, wherein both of the cells may be controlled by a common control gate.

It is another object of the present invention to provide a flash semiconductor memory device including a pair of memory cells sharing a common source region, wherein the amount of space required to fabricate the device on a semiconductor substrate is minimized.

It is a further object of the present invention to provide a process for manufacturing a flash semiconductor memory device including a pair of memory cells sharing a common source region, wherein the manufacturing process is easily controlled.

Briefly, a presently preferred embodiment of the present invention provides a memory device including a first memory cell and a second memory cell both controlled by a common control gate. The device includes: a substrate; a first stack and a second stack, each of the stacks including an insulating layer formed over the substrate, a first conductive layer formed over the insulating layer, and a first dielectric layer formed over the first conductive layer, each of the first and second stacks also including an inner sidewall and an outer sidewall, the inner sidewalls of the first and second stacks opposing each other and being separated by a common area of the substrate, the first conductive layers of the first and second stacks providing select gates of the first and second memory cells respectively, the inner and outer sidewalls of the first and second select gate stacks being coated with a second dielectric layer; first and second spacers are formed adjacent the portions of the second dielectric layer that are coated on the inner sidewalls of the first and second stacks respectively, the first and second spacers being separated by a medial portion of the common source area of the substrate, each of the spacers including a tunnel oxide layer disposed over the substrate, and a second conductive layer disposed over the tunnel oxide layer, the second conductive layers of the first and second spacers providing first and second floating gates of the first and second memory cells respectively, first and second drain regions are formed in the substrate proximate and generally outward from the outer sidewalls of the first and second select gate stacks respectively; a common source region is formed in the substrate generally beneath the common source area. A third dielectric layer is disposed over the medial portion of the common source area of the substrate, over the first and second spacers, and over the first and second stacks; a third conductive layer is disposed over a portion of the third dielectric layer which overlies the inner portions of the first and second select gate stacks, the spacers, and the medial portion of the common source area, the third conductive layer forming the common control gate.

One advantage of the present invention is that it provides a process for manufacturing a semiconductor memory device including a pair of memory cells sharing a common source region, wherein the manufacturing process is easily controlled.

The foregoing and other objects, features, and advantages of the present invention will be apparent from the following detailed description of the preferred embodiment which makes reference to the several figures of the drawing.

FIG. 1 is a cross sectional elevation view of a conventional semiconductor memory device having a pair of memory cells formed over a substrate and sharing a common source region;

FIG. 2A is a cross sectional elevation view of a first embodiment of a semiconductor memory device in accordance with the present invention, the device including a pair of memory cells sharing a common source region, both cells being controlled by a common control gate;

FIG. 2B is a cross sectional elevation view of a second embodiment of a semiconductor memory device in accordance with the present invention;

FIGS. 3A through 3F are cross-sectional views generally illustrating a progression of manufacturing steps in accordance with a process of manufacturing the memory device of FIG. 2A in accordance with the present invention.

FIG. 2A shows a cross sectional elevation view of a first embodiment of a pair of semiconductor flash memory devices at 100 in accordance with the present invention. The device 100 including a first flash memory cell 102 a and a second flash memory cell 102 b both controlled by a common control gate 106 . The pair of flash memory devices 100 includes: a semiconductor substrate 108 ; a first and second stacks 110 a and 110 b each having an insulating layer 114 a and 114 b formed over the substrate 108 . The first and second stacks 110 a and 110 b also include first and second select gates 116 a and 116 b of the first and second flash memory cells 102 a and 102 b respectively, each of the select gates 116 a and 116 b being formed from a conductive layer disposed over an insulating layers 114 a and 114 b respectively as further explained below. First dielectric layers 118 a and 118 b are formed over each of the select gates 116 a and 116 b of the first and second stacks 110 a and 110 b. Each of the first and second stacks 110 a and 110 b also has an inner sidewall 120 a and 120 b and an outer sidewall 122 a and 122 b respectively. The inner sidewalls 120 a and 120 b of the first and second select gate stacks 110 a and 110 b respectively oppose each other and are separated by a common source area 124 of the substrate 108 . The inner sidewalls 120 a and 120 b are each coated with a second dielectric layer 126 a and 126 b respectively. The outer sidewalls 122 a and 122 b are each coated with a third dielectric layer 128 a and 128 b respectively.

The cells 102 a and 102 b of the pair of flash memory devices 100 include first and second spacers 130 a and 130 b respectively being formed adjacent the second dielectric layer 126 a and 126 b respectively, coating on the inner sidewalls 120 a and 120 b of the first and second stacks 110 a and 110 b respectively. The first and second spacers 130 a and 130 b being separated by a medial portion 132 of the common source area 124 of the substrate 108 . Each of the spacers 130 a and 130 b including a tunnel oxide layer 134 a and 134 b respectively, disposed over the substrate 108 . The first and second spacers 130 a and 130 b function as floating gates 136 a and 136 b of each of the flash memory cell 102 a and 102 b respectively. The floating gates 136 a and 136 b being formed over the tunnel oxide layer 134 a and 134 b of the first and second flash memory cell 102 a and 102 b respectively, as further explained below.

The pair of flash memory devices 100 also includes: first and second drain regions 140 a and 140 b formed in the substrate 108 proximate and generally outward from the outer sidewalls 122 a and 122 b of the first and second stacks 110 a and 110 b respectively; and a common source region 144 formed in the substrate 108 generally beneath the common source area 124 and beneath inner portions of the first and second spacers 130 a and 130 b.

The pair of flash memory devices in 100 further includes: a fourth dielectric layer 104 formed supedjacent the medial portion 132 of the common source area 144 of the substrate 108 between the first and second spacers 130 a and 130 b; first and second outer spacers 148 a and 148 b disposed adjacent the third dielectric layer 128 a and 128 b of the first and second stacks 110 a and 110 b and generally above the first and second drain regions 140 a and 140 b of the substrate 108 .

A fifth dielectric layer 154 is disposed superjacent the fourth dielectric layer 104 of the medial portion 132 of the common source area 124 of the substrate 108 , over the first and second spacers 130 a and 130 b, over the first and second stacks 110 a and 110 b and over the first and second outer spacers 148 a and 148 b. The common control gate 106 is formed from a third conductive layer disposed over inner portions of the first and second stacks 110 a and 110 b, the spacers 130 a and 130 b, and a medial portion 132 of the fifth dielectric layer 154 .

FIG. 2B shows a cross sectional elevation view of a second embodiment of a pair of flash memory devices at 160 in accordance with the present invention. The pair of flash memory devices in 160 being substantially similar to the pair of flash memory devices in 100 (FIG. 2A) except that first and second spacer 162 a and 162 b of the devices in 160 extend a distance vertically above the height of the first and second stacks 110 a and 110 b whereas the first and second spacers 130 a and 130 b (FIG. 2A) of the devices in 100 are formed to have substantially the same height as the first and second stacks 170 a and 170 b. The advantage of the structure shown in devices 160 is that it increases the capacitance surface area of the device.

FIGS. 3A through 3F show cross-sectional views generally illustrating a progression of manufacturing steps of a process of manufacturing a pair of flash the memory devices 100 of FIG. 2A in accordance with the present invention. A process of fabricating the pair of flash semiconductor memory devices in 160 (FIG. 2B) in accordance with the present invention is substantially similar to the process of manufacturing the pair of flash memory devices in 100 (FIG. 2A) with the exception of a few differences as further explained below.

Referring to FIG. 3A, a first and second stacks 110 a and 110 b are formed by first forming an insulating layer 114 over the semiconductor substrate 108 . The insulating layer 114 is typically formed using an oxide, commonly referred to as a gate oxide. However, other suitable insulating materials are also known to be used. A first conductive layer 116 is formed over the insulating layer 114 , and a first dielectric layer 118 is formed over the first conductive layer 116 . Subsequently, the insulating layer 114 , first conductive layer 116 , and first dielectric layer 118 are patterned and etched to expose areas of the substrate 108 including the common source area 124 of the substrate 108 between inner sidewalls 120 a and 120 b of the first and second stacks 110 a and 110 b. Each of the stacks including remaining portions of the insulating layer 114 a and 114 b respectively, first conductive layer 116 a and 116 b respectively, and the first dielectric layer 118 a and 118 b resepectively. Each of the stacks further including the outer sidewalls 122 a and 122 b respectively disposed adjacent corresponding outer exposed areas of the substrate 108 . The remaining portions of the first conductive layer 116 form select gates 116 a and 116 b of the first and second flash memory cells 102 a and 102 b (FIG. 2 A).

Referring to FIG. 3B, the inner and outer sidewalls 120 a, 120 b and 122 a, 122 b of the stacks 110 a and 110 b are coated with a second dielectric layer 126 a, 126 b and 128 a, 128 b respectively. In a preferred embodiment, this step of the manufacturing process includes: depositing a second dielectric layer 126 over the first and second stacks 110 a and 110 b in accordance with a chemical vapor deposition (CVD) process; and etching back the second dielectric layer 126 , leaving the coating on the inner and outer sidewalls 126 a, 126 b and 128 a and 128 b respectively as shown. Subsequently, a tunnel oxide layer 134 is grown on the exposed common area 124 of the substrate 108 by exposing the substrate 108 to oxygen at high temperatures. Alternatively, the tunnel oxide layer 134 may be formed via a chemical vapor deposition (CVD) process. The resulting interface between the tunnel oxide, or thermal oxide layer 134 and the substrate 108 has low levels of ionic impurities and defects.

Referring to FIG. 3C, a second conductive layer 130 is initially deposited over the stacks 110 a and 110 b and over the tunnel oxide layer 134 (FIG. 3 B). Subsequently, the second conductive layer 130 and the tunnel oxide layer 134 (FIG. 3B) are patterned in accordance with a lithography process, and etched to expose the medial portion 132 of the common source area 124 of the substrate 108 leaving the spacers 130 a and 130 b adjacent the second dielectric layer 126 a and 126 b respectively on the inner sidewalls 120 a and 120 b of the first and second stacks 110 a and 110 b respectively. The remaining portions of the second conductive layer 130 form the floating gates 136 a and 136 b of the first and second flash memory cells 102 a and 102 b (FIG. 2A) respectively.

Referring to FIG. 3D, the substrate 108 is lightly doped, or diffused, to partially form the first and second drain regions 140 a and 140 b in the substrate 108 adjacent and outward from the outer sidewalls 122 a and 122 b of the first and second select gate stacks 110 a and 110 b respectively using the select gate stacks 110 a and 110 b as masking. The substrate 108 is also doped to form a common source region 144 in the substrate 108 intermediate the first and second select gate stacks 110 a and 110 b using the floating gate spacers 130 a and 130 b as masking.

Referring to FIG. 3E, the outer spacers 148 a and 148 b are formed adjacent the sidewalls 128 a and 128 b adjacent the outer sidewalls 122 a and 122 b (FIG. 3A) of each of the first and second select gate stacks 110 a and 110 b respectively, and above the first and second drain regions 140 a and 140 b respectively. Subsequently, the substrate 108 is doped once again using the outer spacers 148 a and 148 b as ion implantation masks to complete the formation of the first and second drain regions 140 a and 140 b in the substrate 108 . Note that a more lightly doped area of the first and second drain regions 140 a and 140 b is formed under the outer spacers 148 a and 148 b. Each of these more lightly doped areas is referred to as a lightly doped drain extension (LDD). Next, a dielectric layer 104 is formed superjacent the medial portion 132 of the common source area 144 of the substrate 108 intermediate the first and second select gate stacks 110 a and 110 b.

Referring to FIG. 3F, a fourth dielectric layer 154 a is formed over the outer spacers 148 a and 148 b, second dielectric layer 126 a, 126 b, 128 a, and 128 b, the first and second select gate stacks 110 a and 110 b, the spacers 130 a and 130 b and the dielectric layer 104 . In the preferred embodiment, the fourth dielectric layer 154 a is formed from oxide-on-nitride-on-oxide (ONO). The third dielectric layers 104 and fourth dielectric layer 154 a form the third dielectric layer 154 (FIG. 2A) of the flash semiconductor memory device in 100 .

The common control gate 106 is formed by a third conductive layer formed over a medial portion of the dielectric layer 154 a that is disposed over inner portions of the first and second select gate stacks 110 a and 110 b, the first and second spacers 130 a and 130 b, and the third dielectric layer 104 .

The process of fabricating the pair of flash semiconductor memory devices 160 (FIG. 2B) diverges from the process of manufacturing the pair of flash memory devices 100 (FIG. 2A) in the step of forming the dielectric layer 154 (FIG. 2B) of the devices in 160 . Before forming the dielectric layer 154 , an etch back process is performed on the first dielectric layer 118 a and 118 b, as well as the second dielectric layers 126 a, 126 b and third side walls 128 a, 128 b. As a result, spacers 162 a and 162 b extend a distance vertical above the height of the first and second select gate stacks 110 a and 110 b as shown in FIG. 2 B. Otherwise, the process of fabricating the pair of flash memory devices 160 (FIG. 2B) is substantially similar to the process of manufacturing the pair of flash memory devices in 100 (FIG. 2 A).

In the embodiment described above, wherein the first, conductive layer 116 that functions as the select gate, the second conductive layer 130 that functions as the floating gate, and the third conductive layer that functions as the control gate 106 each can be formed with the material of polysilicon. Furthermore, each of the first, second, and third conductive layer may have different material, such as amorphous silicon, recrystalized amorphous silicon, or silicon alloy material, or other conductive material. The first dielectric layer 118 , second dielectric layer 126 , third dielectric layer 104 , and fourth dielectric layer 1545 can be selected from a group consisting the following material; silicon dioxide, silicon nitride, and silicon nitride dioxide, such as oxide-on-nitride-on-oxide (ONO). These dielectric layers can be formed via the process of chemical vapor deposition (CVD).

Although the present invention has been particularly shown and described above with reference to a specific embodiment, it is anticipated that alterations and modifications thereof will no doubt become apparent to those skilled in the art. It is therefore intended that the following claims be interpreted as covering all such alterations and modifications as fall within the true spirit and scope of the invention.