Resistive random access memory with preformed filaments
12127486 ยท 2024-10-22
Assignee
Inventors
Cpc classification
H10N70/068
ELECTRICITY
H10N70/884
ELECTRICITY
H10N70/826
ELECTRICITY
H10B63/80
ELECTRICITY
H10N70/00
ELECTRICITY
H10B63/00
ELECTRICITY
H10N70/063
ELECTRICITY
International classification
H10N79/00
ELECTRICITY
H10B63/00
ELECTRICITY
Abstract
A method for fabricating a plurality of resistive random access memory (RRAM) cells includes providing a substrate including a memory medium arranged on an underlying layer; creating channel holes in the memory medium having a first critical dimension in a range from 1 nm to 20 nm; depositing switching material defining a filament of the RRAM cells in the channel holes; depositing a top electrode of the RRAM cells on the memory medium and the switching material; and separating adjacent ones of the RRAM cells by etching the top electrode and the memory medium between adjacent ones of the channel holes.
Claims
1. A method for fabricating a plurality of resistive random access memory (RRAM) cells, comprising: providing a substrate including a memory medium arranged on an underlying layer; creating channel holes in the memory medium; and depositing switching material by depositing a conformal layer doped with a metal in the channel holes, the metal segregating at a surface of the conformal layer defining a filament of the RRAM cells in the channel holes.
2. The method of claim 1, wherein creating the channel holes in the memory medium includes: depositing a first mask layer on the memory medium; patterning the channel holes having a first critical dimension in the first mask layer; etching the memory medium through the channel holes to the underlying layer; and removing the first mask layer.
3. The method of claim 1, wherein creating the channel holes in the memory medium includes: depositing a first mask layer on the memory medium; patterning the channel holes having a second critical dimension that is greater than a first critical dimension in the first mask layer; etching the memory medium through the channel holes to the underlying layer; removing the first mask layer; and filling the channel holes with the doped conformal layer to reduce the second critical dimension of the channel holes to less than or equal to the first critical dimension.
4. The method of claim 1, wherein the doped conformal layer is deposited using atomic layer deposition.
5. The method of claim 1, wherein the memory medium is selected from a group consisting of silicon (Si), silicon germanium (Si.sub.xGe.sub.y), silicon dioxide (SiO.sub.2), silicon nitride (Si.sub.aN.sub.b), metal oxides, and metal nitrides.
6. The method of claim 1, wherein the switching material is selected from a group consisting of doped silicon (Si), silicon germanium (Si.sub.xGe.sub.y), silicon dioxide (SiO.sub.2), silicon nitride (Si.sub.aN.sub.b), metal oxides, and metal nitrides.
7. The method of claim 1, wherein filling the channel holes with the switching material includes depositing a doped semi-conformal layer in the channel holes and wherein the doped semi-conformal layer is thicker at a bottom portion of the channel holes than at a upper portion of the channel holes.
8. The method of claim 1, wherein filling the channel holes with the switching material includes depositing a plurality of alternating layers of metal and dielectric.
9. The method of claim 8, wherein the plurality of alternating layers of metal and dielectric define a plurality of filaments.
10. The method of claim 8, wherein the plurality of alternating layers of metal and dielectric are deposited using atomic layer deposition.
11. The method of claim 1, further comprising, after depositing the switching material, performing chemical mechanical polishing to remove the switching material from a field region of the memory medium.
12. The method of claim 1, wherein the memory medium is selected from a group consisting of amorphous, polycrystalline and single crystal.
13. The method of claim 1, wherein a first critical dimension of the channel holes is greater than or equal to 1 nm and less than or equal to 5 nm.
14. The method of claim 1, wherein a first critical dimension of the channel holes is in a range from 1 nm to 20 nm.
15. The method of claims 1, further comprising depositing a top electrode of the RRAM cells on the memory medium and the switching material.
16. The method of claim 15, further comprising, after depositing the top electrode, separating adjacent ones of the RRAM cells by etching the top electrode and the memory medium between adjacent ones of the channel holes.
Description
BRIEF DESCRIPTION OF THE DRAWINGS
(1) The present disclosure will become more fully understood from the detailed description and the accompanying drawings, wherein:
(2)
(3)
(4)
(5)
(6)
(7)
(8) In the drawings, reference numbers may be reused to identify similar and/or identical elements.
DETAILED DESCRIPTION
(9) One source of unreliability in RRAM cells involves resistance variations (cycle-to-cycle and/or device-to-device) in the low resistance state (LRS) and the high resistance state (HRS). The primary source of the resistance variations is caused by inconsistent formation and disruption of the filaments in the RRAM cells.
(10) The present disclosure relates to RRAM cells and methods for creating RRAM cells. The RRAM cells include switching material (defining one or more filaments) that are patterned and etched. In some examples, the switching material is etched with a critical dimension that is in a range from 1 nm to 20 nm. In other examples, the switching material is etched with a critical dimension that is greater than or equal to 1 nm and less than or equal to 5 nm.
(11) Some RRAM cells include a memory medium with epitaxially-grown, single crystal materials. Filaments are formed in the line defects. Dislocation control is used to generate reliable formation of the line defects and filaments in SiGe/Si epitaxial memory. The memory has high endurance and long retention, high on-off ratio, cycle-to-cycle uniformity, device-to-device uniformity, suppression of sneak paths and linear conductance update. However, memory using this approach is difficult to manufacture at higher volumes.
(12) The method for fabricating RRAM cells according to the present disclosure includes switching material with filaments that are formed in the memory medium by patterning holes, etching and/or deposition. In some examples, the channel holes in the memory cells (for the switching material) have a critical dimension that is in a range from 1 nm to 20 nm. In other examples, the channel holes in the memory cells (for the switching material) have a critical dimension that is greater than or equal to 1 nm and less than or equal to 5 nm. This approach ensures that a filament channel will be located on every memory cell. Furthermore, the memory cell will have reduced variability since the critical dimension (CD) and location of the variable resistance channel is controlled.
(13) The method for fabricating RRAM cells allows a wider variety of materials to be used. For example, the switching material doesn't need to be single crystalline. It can be amorphous or poly crystalline (even though polycrystalline materials may increase variability).
(14) Referring now to
(15) In
(16) In
(17) In
(18) In
(19) Referring now to
(20) This arrangement produces a higher electric field due to the decreasing thickness from the lower portion 264 to the upper portion 260. In some examples, an inhibitor plasma may be used prior to atomic layer deposition (or between ALD cycles) to inhibit deposition in the upper portion 260 of the channel holes 122 relative to the lower portion 264 of the channel holes 122 and to create the semi-conformal layer. An example of inhibitor plasma is disclosed in commonly assigned U.S. Pat. No. 9,425,078, which is entitled Inhibitor Plasma Mediated Atomic Layer Deposition for Seamless Feature Fill, which is incorporated herein by reference in its entirety.
(21) Referring now to
(22) At 332, the memory medium is etched through the channel holes in the mask layer. At 336, a switching material is deposited in the channel holes of the memory medium. In some examples, atomic layer deposition is used. In some examples, chemical mechanical polishing, etching or another process is used to remove the switching material deposited on a field region of the substrate. At 340, a top electrode is deposited.
(23) At 344, a mask layer is deposited and patterned to define inter-cell holes between the memory cells. At 348, the areas below the inter-cell holes and between the memory cells are etched using conductor etching and dielectric etching. At 350, the mask layer is removed. Further processing of the substrate may be performed.
(24) Referring now to
(25) Referring now to
(26) At 536, film is deposited in the channel holes to reduce the critical dimension from the second critical dimension to the first critical dimension, respectively. In some examples, ALD is performed. At 538, a switching material is deposited in the channel holes of the memory medium. In some examples, atomic layer deposition is used to deposit a metal doped material. In some examples, chemical mechanical polishing (CMP), etching or another process is used to remove the switching material on a field region of the substrate. At 540, a top electrode is deposited.
(27) At 544, a mask layer is deposited and patterned to define inter-cell holes arranged between the memory cells. At 548, the areas below the inter-cell holes (and between the RRAM cells) are etched using conductor etching and dielectric etching. At 550, the mask layer is removed. Further processing of the substrate may be performed.
(28) Referring now to
(29) In some examples, L is an even number and there are L/2 layers of metal and L/2 layers of dielectric. In other examples, L is an odd number and there are an uneven number of layers of metal and dielectric. In some examples, each of the layers 614 are deposited using multiple ALD cycles. For example, R ALD cycles are used to deposit the dielectric layers and S ALD cycles are used to deposit the metal layers, where R and S are integers greater than one. The multi-layer arrangement allows multiple filaments to be formed in a single memory channel.
(30) The foregoing description is merely illustrative in nature and is in no way intended to limit the disclosure, its application, or uses. The broad teachings of the disclosure can be implemented in a variety of forms. Therefore, while this disclosure includes particular examples, the true scope of the disclosure should not be so limited since other modifications will become apparent upon a study of the drawings, the specification, and the following claims. It should be understood that one or more steps within a method may be executed in different order (or concurrently) without altering the principles of the present disclosure. Further, although each of the embodiments is described above as having certain features, any one or more of those features described with respect to any embodiment of the disclosure can be implemented in and/or combined with features of any of the other embodiments, even if that combination is not explicitly described. In other words, the described embodiments are not mutually exclusive, and permutations of one or more embodiments with one another remain within the scope of this disclosure.
(31) Spatial and functional relationships between elements (for example, between modules, circuit elements, semiconductor layers, etc.) are described using various terms, including connected, engaged, coupled, adjacent, next to, on top of, above, below, and disposed. Unless explicitly described as being direct, when a relationship between first and second elements is described in the above disclosure, that relationship can be a direct relationship where no other intervening elements are present between the first and second elements, but can also be an indirect relationship where one or more intervening elements are present (either spatially or functionally) between the first and second elements. As used herein, the phrase at least one of A, B, and C should be construed to mean a logical (A OR B OR C), using a non-exclusive logical OR, and should not be construed to mean at least one of A, at least one of B, and at least one of C.