Electrically programmable and eraseable memory device

This application is a continuation of U.S. patent application Ser. No. 12/284,002, filed Sep. 17, 2008, which is incorporated by reference herein in its entirety and for all purposes.

FIGS. 1-4 illustrate a prior art electrically programmable and erasable memory device and a method of fabrication thereof. Initially, with reference to FIG. 1 , a substrate, for example a P type silicon substrate 30 , is provided. A silicon dioxide (SiO 2 ) layer 32 is grown thereon, and successive layers of silicon nitride (SiN) 34 , silicon dioxide (SiO 2 ) 36 , silicon nitride (SiN) 38 , and P+ polysilicon 40 are deposited. A layer of photoresist is then provided on the polysilicon layer 40 , and portions of the photoresist layer are removed using a we known procedure, so that the remaining photoresist 42 is provided on the polysilicon layer 40 as shown in FIG. 1 . Using this remaining photoresist 42 as a mask, an etching step is undertaken to remove portions of the layers 40 , 38 , 36 , 34 , 32 , forming a stack 44 including remaining layers of silicon dioxide 32 A, silicon nitride 34 A, silicon dioxide 36 A, silicon nitride 38 A and polysilicon 40 A on the substrate 30 ( FIG. 2 ).

An on implantation step 45 is undertaken, in this example using n-type dopant, to form N+ source 46 and drain 48 in the substrate 40 , the photoresist 42 and the layers therebeneath acting as a mask. ( FIG. 3 ) The photoresist 42 is then removed, resulting in the structure 49 shown in FIG. 4 .

This structure 49 is an electrically programmable and erasable memory device wherein the silicon dioxide layer 32 A is a tunneling gate layer, the silicon nitride layer 34 A is a storage layer (floating gate), the silicon dioxide layer 36 A and silicon nitride layer 38 A together form a dielectric 39 between the storage layer 34 A and the polysilicon layer 40 A, and the polysilicon layer 40 A is the control gate of the device 49 .

In programming the device 49 , as is well known, a positive voltage is applied to the drain 48 , and the source 46 is grounded. Applying a positive voltage pulse of sufficient magnitude to the control gate 40 A causes electrons to travel from the substrate 30 through the tunneling gate layer 32 A and into the storage layer 34 A, such electrons stored in the storage layer 34 A causing the device 49 to be in its programmed state. In the erasing of the device 49 , a positive voltage is applied to the control gate 40 A, while the substrate 30 , source 46 and drain 48 are grounded. Applying a positive voltage pulse of sufficient magnitude to the control gate 40 A causes holes in the gate 40 A to travel through the dielectric 39 (which includes layers 36 A, 38 A) and into the storage layer 34 A, wherein they are combined with electrons stored in the storage layer 34 A to cause the device 49 to be in its erased state.

While the device 49 thus far shown and described is highly efficient in operation, it will be understood that it is always desirable to improve operational speed thereof. In furtherance thereof, the present approach improves erase speed as compared to the prior device as will be shown and described.

Broadly stated, in the present approach a method of forming a memory device is provided, comprising providing a substrate, providing a tunneling layer over the substrate, providing a floating gate over the tunneling layer, providing a dielectric over the floating gate comprising providing a first insulating layer over the floating gate, providing a second insulating layer over the first insulating layer, and providing a third insulating layer over the second insulating layer, and providing a control gate over the dielectric.

Broadly stated, in the present approach, is memory device comprises a substrate, a tunneling layer over the substrate a floating gate over the tunneling layer, a dielectric over the floating gate and comprising silicon oxynitride, and a control gate over the dielectric.

The present invention is better understood upon consideration of the detailed description below, in conjunction with the accompanying drawings. As will become readily apparent to those skilled in the art from the following description, there are shown and described embodiments of this invention simply by way of the illustration of the best mode to carry out the invention. As will be realized, the invention is capable of other embodiments and its several details are capable of modifications and various obvious aspects, all without departing from the scope of the invention. Accordingly, the drawings and detailed description will be regarded as illustrative in nature and not as restrictive.

The novel features believed characteristic of the invention are set forth in the appended claims. The invention itself, however, as well as said preferred mode of use, and further objects and advantages thereof, will best be understood by reference to the following detailed description of illustrative embodiments when read in conjunction with the accompanying drawings, wherein:

FIGS. 1-4 illustrate process steps in formation of a prior art memory device;

FIGS. 5-8 illustrate process steps in formation of the present memory device;

FIGS. 9-16 illustrate various alternative process steps in formation of particular portions of the present memory device;

FIGS. 17 and 18 illustrate different states of energy band diagrams for the present memory device;

FIGS. 19-21 illustrate different states of energy band diagrams for the prior art memory device; and

FIG. 22 is a timing diagram comparing operation of the present memory device with the prior art memory device.

Reference is now made in detail to specific embodiments of the present invention which illustrate the best mode presently contemplated by the inventors for practicing the invention.

FIGS. 5-8 illustrate the present electrically programmable and erasable memory device and method of fabrication thereof. Similar to the above, initially, with reference to FIG. 5 , a substrate 50 , for example a P type silicon substrate, is provided, and a silicon dioxide (SiO 2 ) layer 52 is grown thereon. A layer of silicon nitride (SiN) 54 is deposited on, over and in contact with the layer 52 , and a layer of silicon dioxide (SiO 2 ) 56 is deposited on, over and in contact with the layer 54 . Next, and different from the previous from the previous process, a layer of silicon oxynitride (SiON) 58 is provided on, over and in contact with the layer 56 . This silicon oxynitride layer 58 may be provided in a variety of different manners, which will be described in detail with reference to FIGS. 9-16 . Then, silicon nitride (SiN) layer 60 is deposited on, over and in contact with the layer 58 , and P+ polysilicon layer 62 is deposited on, over and in contact with the layer.

A layer of photoresist is then provided on the polysilicon layer 62 , and portions of the photoresist layer are removed using a well known procedure, so that the remaining photoresist 64 is provided on the polysilicon layer 62 as shown in FIG. 5 . Using this remaining photoresist as a mask, an etching step is undertaken to remove portions of the layers 62 , 60 , 58 , 56 , 54 , 52 , forming a stack 61 including remaining layers of silicon dioxide 52 A, silicon nitride 54 A, silicon dioxide 56 A, silicon oxynitride 58 A, silicon nitride 60 A, and polysilicon layer 62 A on the substrate 50 ( FIG. 5 ).

An ion implantation step 66 is undertaken, in this example using n-type dopant, to form N+ source 68 and drain 70 in the substrate 50 , the photoresist 64 and the layers therebeneath acting as a mask ( FIG. 7 ). The photoresist 64 is then removed, resulting in the structure 72 shown in FIG. 8 .

This structure 72 is an electrically programmable arid erasable memory device wherein the silicon dioxide layer 52 A is a tunneling gate layer, the silicon nitride layer 54 A is a storage layer (floating gate), the silicon dioxide layer 56 A, silicon oxynitride layer 58 A, and silicon nitride layer 60 A together form a dielectric 63 between the storage layer 54 A and the polysilicon layer 62 A, and the polysilicon layer 62 A is the control gate of the device 72 .

Again, in programming the device 72 , a positive voltage is applied to the drain 70 , and the source 68 is grounded. Applying a positive voltage pulse of sufficient magnitude to the control gate 62 A causes electrons to travel from the substrate 50 through the tunneling gate layer 52 A and into the storage layer 54 A, such electrons stored in the storage layer 54 A causing the device 72 to be in its programmed state. In the erasing of the device 72 , a positive voltage is applied to the control gate 62 A. a positive voltage is applied to the drain 70 , the source 68 is grounded, and the substrate 50 is allowed to float. Applying a positive voltage pulse of sufficient magnitude to the control gate 62 A causes holes in the control gate 62 A to travel through the dielectric 63 (which includes layers 56 A, 58 A, 60 A) and into the storage layer 54 A, wherein they are combined with electrons stored in the storage layer 54 A to cause the device 72 to be in its erased state. However, due to the inclusion of the silicon oxynitride layer 58 in the memory device 72 , the erase mechanism of the device 72 is fundamentally different from that as previously shown and described above. This mechanism will be shown and described further on. The various approaches for providing the silicon oxynitride layer 58 will now be described with reference to FIGS. 9-16 .

In a first approach, and referring to FIG. 9 , silicon oxynitride layer 58 is deposited on, over and in contact with silicon dioxide layer 56 by for example Plasma-Enhanced Chemical Vapor Deposition (PECVD), Or Low Pressure Chemical Vapor Deposition (LPCVD).

In another approach, referring to FIG. 10 , the silicon oxynitride layer 58 is formed by nitridation of a top portion of the silicon dioxide layer 56 , by nitrogen-containing (e.g., NH 3 , N 2 O, or N 2 ) plasma treatment, forming silicon oxynitride layer on, over and in contact with silicon dioxide layer 56 .

In yet another approach, referring to FIG. 11 , a silicon dioxide layer 80 is provided on the silicon dioxide layer 56 by deposition (by for example PECVD or LPCVD) or growth by thermal oxidation, and nitridation of that silicon dioxide layer Nis undertaken, by nitrogen-containing (e.g., NH 3 , N 2 O, or N 2 ) plasma treatment, forming silicon oxynitride layer 58 on, over and in contact with silicon dioxide layer 56 .

In yet another approach, referring to FIG. 12 , a silicon nitride layer 82 is deposited on the silicon dioxide layer 56 (by for example PECVD or LPCVD), and oxidation of that silicon nitride layer 82 is undertaken by for example direct contact oxygen or ozone plasma (e.g., PECVD or High Density Plasma (HDP)), or by remote oxygen or ozone plasma, or by a steam oxidation, forming silicon oxynitride layer 58 on, over and in contact with silicon dioxide layer 56 .

In yet another approach, referring to FIG. 13 , nitrogen is implanted 84 into the top portion of the silicon dioxide layer 56 , and a high temperature anneal step is undertaken (by for example furnace, or by rapid thermal anneal (RTA), or by laser anneal), causing bonding of silicon, oxygen and nitrogen to form silicon oxynitride layer 58 on, over and in contact with silicon dioxide layer 56 .

In yet another approach, referring to FIG. 14 , a silicon dioxide layer 86 is provided on the silicon dioxide 56 layer by deposition (by for example PECVD or LPCVD) or growth by thermal oxidation, nitrogen is implanted 88 into the silicon dioxide layer 86 , and a high temperature anneal step is undertaken (by for example furnace, by RTA, or by laser anneal), causing bonding of silicon, oxygen and nitrogen to form silicon oxynitride layer 58 on, over and in contact with silicon dioxide layer 56 .

In yet another approach, referring to FIG. 15 , a silicon nitride layer 90 is deposited on the silicon dioxide layer 56 (by example PECVD or LPCVD), oxygen is implanted 92 into the silicon nitride layer 90 , and a high temperature anneal step is undertaken (by for example furnace, or RTA, or by laser anneal), causing bonding of silicon, oxygen and nitrogen to form silicon oxynitride 58 layer on, over and in contact with silicon dioxide layer 56 .

In yet another approach, referring to FIG. 16 , alternate silicon nitride and silicon dioxide monolayers 94 , 96 , 98 , 100 . . . are deposited on the silicon dioxide layer 56 by a atomic layer deposition (ALD) and a high temperature anneal step is undertaken (by for example furnace, by RTA, or by laser anneal), causing bonding of silicon, oxygen and nitrogen of the monolayers 94 , 96 , 98 , 100 . . . to form silicon oxynitride layer 58 on, over and in contact with silicon dioxide layer 56 .

FIG. 17 illustrates the energy band diagram 110 for the device 72 of FIG. 8 , without electrical potential applied thereto. This diagram illustrates conduction energy E c and valence energy E v . As shown, the layer 56 A, layer 58 A, and layer 60 A have successively smaller energy band gaps. As noted above, in the erasing of the memory device 72 , holes must travel from the polysilicon gate 62 A through the layers 60 A, 58 A, 56 A and into the storage layer 54 A. Movement of holes from the gate 62 A to the storage layer 54 A with the energy band diagram 110 as shown in FIG. 17 requires sufficient energy to overcome energy E v of the oxide layer 56 A (hole energy increasing in a downward direction in FIG. 17 ). This level of energy is substantial, creating a significant barrier to such hole movement. Applying a selected positive voltage of a magnitude V g1 to the polysilicon gate 62 A while holding the substrate 50 , source 68 and drain 70 at ground causes the energy bands to shift to levels shown in FIG. 18 , greatly reducing the barriers presented by the layers 60 A, 58 A, and 56 A. In the present example, voltage V g1 is sufficient to substantially reduce the barrier presented by layer 60 A to B 1 , eliminate the barrier presented by layer 58 A, and substantially reduce the barrier presented by layer 56 A to B 2 . With the barriers B 1 , 82 facing hole transit being minimal, holes injected from the polysilicon gate 62 A under the application of voltage V g1 to the gate 62 A (chosen at a magnitude to provide a hole energy of E H as shown) have a high probability of containing enough energy to tunnel through the barrier B 1 presented by the layer 60 A, pass over the barrier presented by the layer 58 A, and tunnel through the barrier 82 presented by the layer 56 A. Thus, the device 72 erases rapidly with application of voltage of magnitude V g1 to the gate 62 A.

FIG. 19 illustrates the energy band diagram 120 for the prior art device 49 of FIG. 4 , without electrical potential applied thereto. In the erasing of this memory device 49 , holes must travel from the polysilicon gate 40 A through the layers 38 A, 36 A and into the storage layer 34 A. Again, movement of holes from the gate 40 A to the storage layer 34 A with the energy band diagram as shown in FIG. 19 requires sufficient energy to overcome E v of the oxide layer 36 A (hole energy increasing in a downward direction in FIG. 19 ). Applying the same positive voltage of magnitude V g1 to the polysilicon gate 40 A while holding the substrate 30 , source 46 and drain 48 at ground causes the energy bands to shift to levels shown in FIG. 20 . As will be seen in FIG. 20 , while the energy barrier presented by the layer 38 A is lowered (B 3 ), and the energy barrier presented by the layer 36 A is also lowered (B 4 ), the layer 36 A still presents a substantial barrier to hole movement from the polysilicon gate 40 A to the storage layer 34 A. With the barrier B 4 presented by the oxide layer 36 A still being substantial, holes injected from the polysilicon gate 40 A under the application of voltage Vg 1 to the gate 40 A have a low probability of having enough energy to tunnel through the barrier B 4 presented by the layer 36 A. Thus, the device 49 erases relatively slowly with application of voltage of magnitude V g1 to the gate 40 A.

In the prior art device 49 , applying a positive voltage of magnitude V g2 greater than V g1 to the polysilicon gate 40 A while holding the substrate 30 , source 46 and drain 48 at ground causes the energy bands to shift to levels shown in FIG. 21 . While this magnitude of voltage V g2 applied to the gate 40 A may be sufficient to reduce the barrier presented by the layer 36 A to that achieved for layer 56 A in FIG. 18 , the time delay in building up the voltage applied to the gate 40 A from voltage V g1 voltage V g1 to achieve this sufficiently lowered barrier (i.e., time delay Td in going from voltage V g1 to voltage V g2 as illustrated in FIG. 22 ) results in a longer erase time than that achieved in the present device 72 .

It will therefore be seen that inclusion of the silicon oxynitride layer as shown and described significantly decreases erase speed of the device as compared to the prior art.

The foregoing description of the embodiments of the invention has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the invention to the precise form disclosed. Other modifications or variations are possible in light of the above teachings.

The embodiments were chosen and described to provide the best illustration of the principles of the invention and its practical application to thereby enable one of ordinary skill of the art to utilize the invention in various embodiments and with various modifications as are suited to the particular use contemplated. All such modifications and variations are within the scope of the invention as determined by the appended claims when interpreted in accordance with the breadth to which they are fairly, legally and equitably entitled.