Confined growth of 2D materials and their heterostructures

Abstract

Two-dimensional (2D) materials and their heterostructures offer a promising path toward next-generation electronics. However, challenges exist regarding (i) controlling the growth of 2D materials per monolayer (ML), (ii) maintaining single-domain growth, and (iii) controlling the number of layers and crystallinity at the wafer scale. The deterministic confined growth technique disclosed herein simultaneously addresses these challenges and fabricates wafer-scale single-domain 2D-MLs and their heterostructures on arbitrary substrates. A SiO2 mask (110) is patterned on a 2-inch substrate (100) to define selective or confined growth areas, thereby confining the growth of first nuclei (122). Each growth area or trench (102) is only a few microns wide and is filled with single-domain MLs (124) before a second group of nuclei (132) is introduced. Growth of the second group of nuclei (132) within these trenches results in an array of single-domain bilayers (124, 134) at the 2-inch wafer scale. Devices fabricated with the single domain bilayer exhibit superior performance throughout the device.

Description

[Technical Field]

[0001] [CROSS-REFERENCE TO RELATED APPLICATIONS] This application claims the benefit of priority under 35 U.S.C. §119(e) to U.S. Provisional Patent Application No. 63 / 374,090, filed August 31, 2022, the entire contents of which are incorporated herein by reference for any purpose. [Background technology]

[0002] Two-dimensional (2D) transition metal dichalcogenides (TMDs) and their heterostructures are promising platforms for next-generation electronics, spintronics, valleytronics, and optoelectronics. However, the integration of semiconducting 2D heterostructures onto industrial platforms has so far been challenging due to a lack of scalability. Common 2D heterostructure construction methods, mechanical exfoliation and stacking of 2D flakes, are trial-and-error-based operations, which are plagued by strict size limitations on producible structures. Fabricating 2D heterostructures using mechanical exfoliation and stacking also takes a long time. Summary of the Invention [Problem to be solved by the invention]

[0003] In recent years, significant progress has been made in improving the scalability of obtaining single-crystalline single-layer (ML) TMDs on single-crystalline hexagonal substrates, such as sapphire, using epitaxial growth methods. However, significant challenges remain in growing large-scale 2D heterostructures due to the lack of a layer-by-layer growth strategy for single-domain TMDs. Furthermore, some current growth methods require the undesirable step of applying 2D materials within silicon devices to grow them on hexagonal non-silicon substrates. Single-domain TMD arrays can also be grown via laser irradiation of a nucleation spot. However, growth via laser irradiation presents challenges because the second heterolayer tends to nucleate at the edge of the first single-domain patch. Currently, no viable solution exists for achieving wafer-scale single-domain 2D heterostructures. [Means for solving the problem]

[0004] The technologies involve the layer-by-layer growth of 2D materials on arbitrary substrates. They can be used to grow single-domain homojunction and heterojunction TMDs at the wafer scale. They also incorporate a non-epitaxy strategy to grow single-domain TMDs on amorphous materials, allowing the growth of single-crystalline 2D materials on Si wafers coated with arbitrary layers.

[0005] This technology addresses a fundamental kinetic problem in TMD growth. The SiO2 mask, located on an amorphous Al2O3 or HfO2 layer, rests on a Si substrate, confining the growth of a first set of TMD nuclei to an array of selective growth areas, called pockets or trenches (102), whose lateral dimensions (width and length) are each less than a few microns (1 micron = 1 μm). Density functional theory (DFT) calculations confirm that the binding energy of the TMDs to the substrate (100) is greater than that to the SiO2 mask (110). As a result, TMD nucleation is concentrated on the surface of the substrate (100) rather than on the walls of the SiO2 mask (112). The reduction in size of the pockets or trenches in the SiO2 mask (110) significantly shortens the duration of the growth, resulting in a full first set of nuclei (122) (i.e., a single domain (124) completely filling one pocket) during the incubation period of the second set of nuclei (132) for the second TMD layer (134). This results in a single crystalline domain of ML-TMD layer in each trench on the wafer. The confined geometry allows for precise control of multiple layers, thus allowing TMD-MLs to be grown on top of each other to fill the trenches.

[0006] Using this technology, single-domain bilayer (BL)-WSe2 can be fabricated on a 2-inch wafer scale (1 inch = approximately 2.54 cm) by subsequent confined growth of WSe2. FETs fabricated on single-domain WSe2 arrays spanning an entire 2-inch wafer exhibit performance approaching that of mechanically exfoliated WSe2 flakes, e.g., effective mobility of 72.8 cm for ML-WSe2. 2 V -1 s -1 , 103.5 cm for BL-WSe2 2 V -1 s -1 This will extend to.

[0007] Furthermore, these techniques can be used for the layer-by-layer confinement growth of MoS2 / WSe2 heterostructures on a wafer scale. Valley lifetime measurements for single-domain MoS2 / WSe2 heterostructure arrays are comparable to those obtained from single-domain flakes of TMDs. Our confinement growth techniques enable the fabrication of single-domain ML-by-ML homo- or heterojunctions on a wafer scale.

[0008] One method of the present invention for confined growth of a 2D material (e.g., a TMD, such as MoS2, MoSe2, WS2, or WSe2) on a substrate (100) (e.g., HfO2 or amorphous Al2O3) that exhibits a first binding energy with the 2D material can be carried out as follows: A mask layer (110) (e.g., SiO2) that exhibits a second binding energy with the 2D material that is less than the first binding energy is formed on the substrate (100). A trench (102) is formed in the mask layer (110) with sidewalls (112) surrounding an exposed portion of the substrate (100). A nucleus (122) of the 2D material is disposed on the exposed portion of the substrate (100) and grown into a single-domain monolayer (124) of the 2D material. The single-domain monolayer (124) can fill the trench (102). The single domain monolayer (124) can be used to form semiconductor devices, such as valleytronic devices, forked sheet field effect transistors (FETs), or complementary FETs.

[0009] The maximum lateral dimension of the trench (102) can be approximately equal to the incubation time of a separate nucleus of the 2D material on the single-domain monolayer multiplied by the growth rate of the 2D material. For example, the maximum lateral dimension of the trench (102) can be approximately 2 microns. The trench (102) can be an individual trench among many trenches formed in the mask layer (110) in the form of a trench array, with a single nucleus (122) of the 2D material deposited within each trench in the trench array. Formation of such a single nucleus (122) of the 2D material is achieved if the critical trench size for each pocket is below a threshold value, e.g., 2 microns or less. These nuclei can grow into individual single-domain monolayers (124) of the 2D material, thereby filling the trenches in the trench array.

[0010] In some cases, the nuclei 122 are referred to as first nuclei 122, and the single-domain monolayer 124 is referred to as first single-domain monolayer 124. In these cases, a second nuclei 132 of the 2D material can be deposited on the first single-domain monolayer 124 and grown into a second single-domain monolayer 134 of the 2D material, thereby forming a bilayer of the 2D material in the trench. Alternatively, the 2D material can be referred to as a first 2D material, and a nuclei 132 of a second 2D material, distinct from the first 2D material, can be deposited on the single-domain monolayer 124 and grown to form a heterostructure of the first 2D material 124 and the second 2D material 134 in the trench 102. For example, the first and second 2D materials may be selected from the group consisting of MoS2, MoSe2, WS2, and WSe2, respectively.

[0011] In one embodiment of the present invention, a method is provided comprising the steps of: (a) providing a substrate (100) of a first material; (b) depositing a mask material (110) on the substrate (100); and (c) forming a trench array on the mask material (110), the trench array comprising a plurality of trenches (102), each having a trench geometry with lateral dimensions of l x w, where l and w are each selected to have a maximum dimension of 2 μm, and each trench (102) having an exposed portion of the substrate (100), i.e., a portion surrounded by sidewalls (112) formed by the mask material (110). (d) depositing adatoms (120) of a second material on the exposed portions of the substrate (100), wherein a first binding energy between the first material and the second material is greater than a second binding energy between the mask material (110) and the second material; (e) selectively nucleating the adatoms (120) into nuclei (122) within each trench (102) in the trench array; and (f) growing the nuclei (122) within each trench (102) in the trench array, wherein growth of the nuclei (122) into a single-domain monolayer (124) of the second material is limited by the lateral dimensions of the trench geometry.

[0012] In some embodiments, the first material comprises one of hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HZO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO). In some embodiments, the second material comprises one of graphene, carbon nanotubes (CNTs), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide (e.g., vanadium disulfide (VS2), vanadium diselenide (VSe2), cobalt sulfide (CoS2), cobalt selenide (CoSe2), titanium disulfide (TiS2), or titanium diselenide (TiSe2)), a semiconducting transition metal dichalcogenide (e.g., molybdenum disulfide (MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), or tungsten diselenide (WSe2)), and a high-k material (e.g., Bi2SeO5 or Sb2O3). In some embodiments, the mask material comprises amorphous silicon dioxide (a-SiO2), amorphous silicon (a-Si), amorphous silicon nitride (a-SiN x ), amorphous carbon (a-carbon), hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HfZrO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO).

[0013] In an expanded embodiment, the nuclei (122) are first nuclei (122) and the single-domain monolayer (124) is a first single-domain monolayer (124), and the method further comprises the steps of (g) waiting an incubation period; (h) depositing another adatom (130) of a third material on top of the first single-domain monolayer (124) in at least one trench in the trench array, wherein a third bonding energy between the second material and the third material is greater than a second bonding energy between the mask material (110) and the third material; (j) selectively nucleating another adatom (130) of a third material into a second nucleus (132) on top of the first single-domain monolayer (124) in at least one trench; and (j) growing the nucleus (132) in at least one trench in the trench array, wherein the growth of the second nucleus (132) into the second single-domain monolayer (134) of the third material is limited by the lateral dimensions of the trench geometry, and the first single-domain monolayer (124) and the second single-domain monolayer (134) form a bilayer.

[0014] In some embodiments, the third material comprises one of graphene, carbon nanotubes (CNTs), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide (e.g., vanadium disulfide (VS2), vanadium diselenide (VSe2), cobalt sulfide (CoS2), cobalt selenide (CoSe2), titanium disulfide (TiS2), or titanium diselenide (TiSe2)), a semiconducting transition metal dichalcogenide (e.g., molybdenum disulfide (MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), or tungsten diselenide (WSe2)), and a high-k material.

[0015] In some embodiments, the bilayer is a heterojunction bilayer, in which the second and third materials are different from one another, and in other embodiments, the bilayer is a homojunction bilayer, in which the second and third materials are similar to one another.

[0016] In one embodiment, l and w are each selected to have a value equal to the incubation time of additional nuclei of the second material on the single-domain monolayer multiplied by the growth rate of the second material, and in another embodiment, l and w are each selected to be 2 microns.

[0017] In another embodiment, the method further comprises forming a semiconductor device (a valleytronic device, a forked-sheet field effect transistor (FET) or a complementary FET) comprising the single-domain monolayer.

[0018] In one embodiment, the method includes depositing at least one of hafnium oxide (HfO), aluminum oxide (AlO), hafnium zirconium oxide (HZO), titanium dioxide (TiO), zinc oxide (ZnO), iron oxide (FeO), tin oxide (SnO), nickel oxide (NiO), and copper oxide (CuO) onto silicon to form the substrate before depositing the mask material on the substrate.

[0019] Another embodiment of the present invention provides a method comprising the steps of: (a) providing a substrate (100) of a first material, the first material having a first Gibbs free energy; (b) depositing a mask material (110) on the substrate (100), the mask material (110) having a second Gibbs free energy, the second Gibbs free energy being greater than the first Gibbs free energy; and (c) forming a trench array on the mask material (110), the trench array comprising a plurality of trenches. (d) depositing adatoms (120) of a second material on the exposed portions of the substrate (100); and (e) depositing the adatoms (120) of a second material on each of the trenches (102) in the trench array. The adatoms (120) are then immersed in the first nuclei (122) within the trenches (102). ... (f) growing first nuclei (122) in each trench (102) in the trench array, wherein growth of the first nuclei (122) into a first single-domain monolayer (124) of a second material is limited by the lateral dimensions of the trench geometry; (g) waiting for an incubation period; (h) depositing another adatom (130) of a third material on top of the first single-domain monolayer (124) of at least one trench in the trench array; (i) growing the trench array. (j) selectively nucleating another adatom (130) of a third material into a second nucleus (132) on top of the first single domain monolayer (124) within at least one trench in the array of trenches; and (j) growing the nuclei (132) within at least one trench in the array of trenches, wherein the growth of the second nuclei (132) into the second single domain monolayer (134) of the third material is limited by the lateral dimensions of the trench geometry, and the first single domain monolayer (124) and the second single domain monolayer (134) form a bilayer.

[0020] In some embodiments, the first material comprises one of hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HZO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO).

[0021] In some embodiments, the second material or the third material comprises one of graphene, carbon nanotubes (CNTs), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide (e.g., vanadium disulfide (VS2), vanadium diselenide (VSe2), cobalt sulfide (CoS2), cobalt selenide (CoSe2), titanium disulfide (TiS2), or titanium diselenide (TiSe2)), a semiconducting transition metal dichalcogenide (e.g., molybdenum disulfide (MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), or tungsten diselenide (WSe2)), and a high-k material (e.g., Bi2SeO5 or Sb2O3).

[0022] In some embodiments, the mask material is amorphous silicon dioxide (a-SiO2), amorphous silicon (a-Si), amorphous silicon nitride (a-SiN x ), amorphous carbon (a-carbon), hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HfZrO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO).

[0023] Any combination of the concepts listed above with additional concepts discussed below (to the extent such concepts are not mutually inconsistent) is part of the inventive subject matter of this disclosure. In particular, any combination of subject matter set forth in the claims presented at the end of this disclosure is part of the inventive subject matter of this disclosure. Terminology used in this application, and which may appear in any incorporated-by-reference disclosure, should be associated with the meaning that most closely matches the specific concepts of this disclosure.

[0024] As will be appreciated by those skilled in the art, the drawings are primarily for illustrative purposes and are not intended to limit the scope of the inventive subject matter described herein. The drawings are not necessarily to scale, and in some cases, various aspects of the inventive subject matter disclosed herein may be exaggerated or enlarged in the drawings to facilitate an understanding of the various features. In the drawings, like reference numerals generally refer to similar features (e.g., functionally and / or structurally similar elements). [Brief explanation of the drawings]

[0025] [Figure 1A] FIG. 1 illustrates a conventional transition metal dichalcogenide (TMD) growth process. [Figure 1B] FIG. 1 depicts the confined growth process of the present invention for selective single-domain synthesis of single-domain ML-TMDs, which addresses limitations of conventional TMD growth processes. [Figure 1C] FIG. 10 depicts the creation of a single-domain MoS2-WSe2 heterostructure by confined growth of a second MoS2 layer on a WSe2ML in each trench in a trench array. [Figure 1D] Figure 1 shows the calculation of the binding energies of W3O9, Se2 and W3Se6 clusters to c-Al2O3, a-HfO2 and a-SiO2 substrates. [Figure 2A] FIG. 10 shows selective single-domain synthesis and ML-by-ML confined growth of WSe in a 10 μm pocket on a sapphire substrate with SiO sidewalls. [Figure 2B] FIG. 10 shows selective single-domain synthesis and ML-by-ML confined growth of WSe in a 10 μm pocket on a sapphire substrate with SiO sidewalls. [Figure 2C] FIG. 10 shows selective single-domain synthesis and ML-by-ML confined growth of WSe in a 2 μm pocket on a sapphire substrate with SiO sidewalls. [Figure 2D]FIG. 10 shows selective single-domain synthesis and ML-by-ML confined growth of WSe in a 2 μm pocket on a sapphire substrate with SiO sidewalls. [Figure 2E] FIG. 16 shows a Raman mapping image of ML-WSe2 confined in a 2 μm pocket on a sapphire substrate. [Figure 2F] Photoluminescence (PL) mapping image of confined ML-WSe in a 2 μm pocket on a sapphire substrate. [Figure 2G] FIG. 1 shows a Raman mapping image of confined bilayer (BL)-WSe2 in a 2 μm pocket on a sapphire substrate. [Figure 2H] Figure 1 shows PL mapping images of BL-WSe2 confined in a 2 μm pocket on a sapphire substrate. [Figure 2I] FIG. 1 shows a cross-sectional high-resolution (HR) transmission electron microscope (TEM) image of ML-WSe confined in a pocket on a sapphire substrate. [Figure 2J] FIG. 1 shows a cross-sectional HR-TEM image of BL-WSe2 confined in a pocket on a sapphire substrate. [Figure 3A] Photograph of an integrated, confined BL-WSe field-effect transistor (FET) array on a 5.1 cm × 5.1 cm SiO / Si wafer. The inset shows a micrograph of an individual FET array with 20 FETs integrated (scale bar: 10 μm). [Figure 3B] Figure 1 shows the transfer characteristics of the confined BL-WSe2FET with a channel length LCH of 0.7 μm at a drain-source voltage VDS = -1 V. As shown, the maximum on-state current density reaches 155.8 μA μm-1 and the field-effect mobility reaches 103.5 cm2V-1s-1. [Figure 3C] FIG. 10 shows the output characteristics of a confined BL-WSe2FET. [Figure 3D]Figure 1 shows the on-state current density Ion versus effective mobility μeff for various WSe2 FETs, including a FET of the present invention (star; upper right) and FETs with chemical vapor deposition (CVD)-grown single-crystalline 1-3ML-WSe2 (dark squares), CVD-grown polycrystalline 1-3ML-WSe2 films (open squares), and exfoliated WSe2 flakes (pentagons). The drain-source voltage and channel length are −1 V and approximately 1 μm, respectively. [Figure 3E] Histograms of on-state current density Ion (left) and effective mobility μeff (right) for FETs fabricated with exfoliated ML / BL-WSe2 flakes and those fabricated with confined ML / BL-WSe2 films. [Figure 3F] FIG. 10 is a plot of the statistical distribution of on-state current density and effective mobility in a confined BL-WSe2 FET array (the various triangles represent the position of the BL-WSe2 film on the wafer). [Figure 4A] FIG. 1 shows photographs and schematic images of HfO 2 deposited on a Si wafer in pockets with lateral dimensions of 1 μm (scale bar: 5 μm). [Figure 4B] FIG. 10 shows the nucleation of single-domain MoS in a 1 μm HfO pocket with SiO sidewalls. [Figure 4C] FIG. 10 shows an array of confined single-domain MoS2 within a 1 μm HfO2 pocket with SiO2 sidewalls. [Figure 4D] Figure 10 is a plot of the transfer characteristics of 16 FETs fabricated with ML-MoS2 (8 FETs, lower curves) or BL-MoS2 (8 FETs, upper curves) confined on a HfO2 substrate. [Figure 4E] Histograms of the average and maximum values ​​of the on-state current density Ion (left) and effective mobility μeff (right) in ML-MoS2 and BL-MoS2 FETs. [Figure 4F]Figure 1 shows a cross-sectional high-angle annular dark-field scanning transmission electron microscope (HAADF-STEM) image of a heterointerface [MoS2 (top layer) and WSe2 (bottom layer)] with overlaid energy dispersive X-ray (EDX) spectra for MoKα (top left peak), SKA (top right peak), WLA (bottom left peak), and SeKα (bottom right peak). [Figure 4G] Figure 1 shows plots of the time-resolved circular dichroism (CD) response of ML-WSe2 at 300 K (lower curve) and the time-resolved CD response of hetero-BL (MoS2 / WSe2) at 300 K (middle curve) and 77 K (upper curve). [Figure 5] FIG. 1 is a plot showing the relationship between lateral growth rate and secondary nucleation time in confined growth. [Figure 6A] A comparison of the formation of the second WSe2 layer with increasing H2 content is depicted. [Figure 6B] A comparison of the formation of the second WSe2 layer with increasing H2 content is depicted. [Figure 6C] A comparison of the formation of the second WSe2 layer with increasing H2 content is depicted. [Figure 6D] A comparison of the formation of the second WSe2 layer with increasing H2 content is depicted. [Figure 6E] A comparison of the formation of the second WSe2 layer with increasing H2 content is depicted. DETAILED DESCRIPTION OF THE INVENTION

[0026] Figure 1A shows a conventional transition metal dichalcogenide (TMD) growth process. First, a first group of TMD adatoms 12 are introduced onto the surface of a substrate 10 (top). These TMD adatoms 12 nucleate to form nuclei 14 on the substrate 10. Because the orientation of the nuclei 14 is random, they are generally not aligned with the substrate 10. As the nuclei 14 grow laterally and encounter each other (middle), they form grains 16 and coalesce, resulting in a continuous polycrystalline layer due to the random orientation of the nuclei 14. This polycrystalline growth ultimately degrades the intrinsic properties of the TMDs. Additionally, additional nucleations 18 may occur on some of the grains 16 (bottom). If this process were repeated without controlling the additional nucleation, the result would be a TMD layer growing with an irregular thickness due to the overlap of grains 16 and the nucleation of a second group of nuclei 18 on the initial layer of grains 16.

[0027] Figure 1B illustrates a confined growth process that addresses these issues by precisely controlling the thickness and crystallinity of TMD growth. First, c-plane Al2O3, HfO2, or other materials with relatively low Gibbs free energies, such as amorphous or crystalline metal oxides, are deposited on a Si wafer to form a substrate 100 on which the TMDs are grown. Other suitable substrate materials include graphene, hexagonal boron nitride (hBN), hafnium zirconium oxide (HZO), TiO2, ZnO, Fe2O3, SnO2, NiO, CuO, or TMD-coated materials.

[0028] A thin layer of mask material 110 (e.g., 5 nm, 10 nm, 50 nm, 100 nm, or even hundreds of nanometers thick, potentially up to 250 nm, 500 nm, 750 nm, or 1 μm thick) is then deposited onto the c-plane Al2O3 or HfO2 surface of substrate 100. Suitable mask materials include materials with relatively large Gibbs free energies, such as amorphous SiO2 (a-SiO2), a-Si, a-SiN, and the like. x and a-carbon.

[0029] Next, confined growth areas 102, also known as recesses, pockets, trenches, wells, or cavities, each with lateral dimensions ranging from tens of nanometers to approximately 2 microns, are patterned into a thin layer 110 of a-SiO. The patterned a-SiO layer 110 is also referred to as a mask or mask layer. These pockets 102 can be of any suitable shape (e.g., square, rectangular, triangular, or circular) and can be formed in 1D or 2D arrays (e.g., square, rectangular, or hexagonal arrays). The pockets 102 extend through the entire thickness of the a-SiO layer 110, exposing a portion of the AlO or HfO surface of the substrate 100, surrounded by a-SiO sidewalls 112. The pockets 102 may also extend partway (e.g., over several nanometers) into the c-plane AlO or HfO (Figure 1B, top).

[0030] After the pockets 102 are formed in the a-SiO2 layer 110, 2D material adatoms 120 are introduced into each pocket (top). As an example, it is noted that within a 1 nm x 1 nm area, there would be 42 (W and Se) adatoms. Thus, within a 2 μm x 2 μm trench, there would be 84,000 adatoms per pocket if a monolayer is required, or 168,000 adatoms if a bilayer is required. Suitable 2D materials include semiconducting TMDs, such as WSe2 (here), MoS2, MoSe2, and WS2. Other suitable 2D materials include, but are not limited to, graphene, carbon nanotubes (CNTs), hBN, metallic TMDs (e.g., VS2, VSe2, CoS2, CoSe2, TiS2, TiSe2), and high-k 2D materials (e.g., Bi2SeO5, Sb2O3).

[0031] The size of each pocket 102 is small enough that only one nucleation 122 (represented by a single triangle in FIG. 1B) occurs within each pocket (middle). Each nucleation 122 grows on the exposed c-plane Al2O3 or HfO2 at the bottom of its pocket 102, eventually reaching the a-SiO2 sidewall 112 and filling the entire trench 102, thereby forming a TMD film 124 (bottom). Each TMD film 124 is both a single domain and a multi-layer structure (bottom, inset).

[0032] Figure 1C shows how the adatom introduction, nucleation, and single-crystal growth steps of the process in Figure 1B can be repeated to obtain single-domain MoS2 / WSe2 heterostructures and single-domain homobilayers (BLs) of WSe2. (A BL can be thought of as two MLs connected by van der Waals interactions.) After WSe2 MLs 124 are formed in the pockets 102, MoS2 adatoms 130 are introduced into the cavities 102 above the WSe2 MLs 124 (Figure 1C, top). These MoS2 adatoms 130 undergo nucleation to form nuclei 132 (middle), and then grow (bottom) to form single-crystalline MoS2 MLs 134 on the WSe2 MLs 124 (bottom inset), resulting in the development of MoS2 / WSe2 heterostructures 140 within each pocket 102. DFT calculations confirm this growth selectivity. Alternatively, WSe2 adatoms can be deposited on a WSe2 ML for nucleation and growth to form a single-domain BL of WSe2.

[0033] Figure 1D shows the DFT calculations of the binding energies of WO and Se precursor and product WSe clusters with c-AlO, a-HfO, and a-SiO. These DFT calculations reveal that the WO cluster (WO, left), Se cluster (Se, center), and WSe cluster (WSe, right) exhibit stronger bonding interactions with c-AlO and a-HfO ​​than with the SiO surface. This indicates that the clusters preferentially bond to the substrate surface at the bottom of the pockets rather than to the a-SiO sidewalls, leading to selective WSe growth within the pockets.

[0034] This selectivity has been confirmed by the simultaneous growth of WSe2 on Al2O3, HfO2, and SiO2 substrates under the same CVD growth conditions. Atomic force microscopy (AFM) images show that WSe2 nucleates exclusively on Al2O3 and HfO2, but not on SiO2, during 20 min of growth. This led to the successful selective confinement of WSe2 growth on the exposed substrate surface of fine-patterned SiO2 trench arrays.

[0035] The trench size is selected to ensure exclusively single-domain ML-WSe2 formation and depends on both the lateral growth rate of WSe2 and the incubation period for secondary nucleation. The measured lateral growth rate of WSe2 and the incubation period for secondary nucleation are approximately 0.4 μm / min and 5 min, respectively. Consequently, each trench should be approximately 2 μm wide or less to avoid nucleation of a second WSe2 layer. The lateral growth rate and incubation period (and thus the maximum trench width) can be controlled by varying the TMD powder composition, gas ratio (Ar / H2), and growth temperature. For example, the composition of the S / Se powder can be varied from 100 mg to 1500 mg, and the composition of the MoO3 / WO3 powder can be varied from 10 mg to 100 mg to affect such control. The gas ratio of argon (Ar) to hydrogen (H2) can be adjusted between Ar-100% / H2-0% and Ar-0% / H2-100%, and the growth temperature can be adjusted between 200°C and 1000°C.

[0036] A plot showing the relationship between lateral growth rate and secondary nucleation time for confined growth is shown in Figure 5. In this non-limiting example, after an initial incubation time of 5 minutes, WSe2 was grown laterally up to 10 minutes, with secondary nucleation occurring after maintaining a confined monolayer (ML) for an additional 2 minutes (i.e., the second incubation period from the 10 minute point).

[0037] Figures 6A–6E show the comparison of the formation of the second WSe2 layer with increasing H content. Scanning electron microscope (SEM) images of the second WSe2 layer before (see Figure 6A) and after (see Figure 6B) a 15% increase in H content are shown. Figure 6C shows an optical microscope image of the AA′ stacking structure in Figure 6B. With increasing H content, the nucleation density on the first ML-WSe2 layer saturates, and the second WSe2 layer begins to nucleate at the center of the first ML. Therefore, we compared the SEM morphology of the confined ML with bilayers (BLs) obtained through ML-by-ML confinement growth of WSe2 at different H content levels (see Figures 6A and 6B). Compared to polycrystalline WSe2, the surface of the confined WSe2 was significantly smoother, and the contrast between the confined ML (see Figure 6(d)) and the confined BL (see Figure 6(e)) was similar.

[0038] One embodiment of the present invention provides a method comprising the steps of: (a) providing a substrate 100 of a first material; (b) depositing a mask material 110 on the substrate 100; and (c) forming a trench array on the mask material 110, the trench array comprising a plurality of trenches 102, each having a trench geometry with lateral dimensions of l×w, where l and w are each selected to have a maximum dimension of 2 μm, and each trench 102 having an exposed portion of the substrate 100 surrounded by sidewalls 112 formed by the mask material 110. (d) depositing adatoms 120 of a second material on the exposed portions of the substrate 100, wherein a first bonding energy between the first material and the second material is greater than a second bonding energy between the mask material 110 and the second material; (e) selectively nucleating the adatoms 120 into nuclei 122 within each trench 102 in the trench array; and (f) growing the nuclei 122 within each trench 102 in the trench array, wherein the growth of the nuclei 122 into a single-domain monolayer 124 of the second material is limited by the lateral dimensions of the trench geometry.

[0039] In some embodiments, the first material comprises one of hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HZO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO). In some embodiments, the second material comprises one of graphene, carbon nanotubes (CNTs), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide (e.g., vanadium disulfide (VS2), vanadium diselenide (VSe2), cobalt sulfide (CoS2), cobalt selenide (CoSe2), titanium disulfide (TiS2), or titanium diselenide (TiSe2)), a semiconducting transition metal dichalcogenide (e.g., molybdenum disulfide (MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), or tungsten diselenide (WSe2)), and a high-k material (e.g., Bi2SeO5 or Sb2O3). In some embodiments, the mask material comprises amorphous silicon dioxide (a-SiO2), amorphous silicon (a-Si), amorphous silicon nitride (a-SiN x ), amorphous carbon (a-carbon), hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HfZrO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO).

[0040] In an expanded embodiment, the nuclei 122 are first nuclei 122 and the single-domain monolayer 124 is a first single-domain monolayer 124, and further comprising the steps of (g) waiting an incubation period; (h) depositing another adatom 130 of a third material on top of the first single-domain monolayer 124 in at least one trench in the trench array, wherein a third bonding energy between the second material and the third material is greater than a second bonding energy between the mask material 110 and the third material; and (i) depositing another adatom 130 of a third material on top of the first single-domain monolayer 124 in at least one trench in the trench array. (j) selectively nucleating another adatom 130 of a third material into second nuclei 132 on top of the first single domain monolayer 124 in at least one trench; and (j) growing the nuclei 132 in at least one trench in the trench array, wherein the growth of the second nuclei 132 into the second single domain monolayer 134 of the third material is limited by the lateral dimensions of the trench geometry, and the first single domain monolayer 124 and the second single domain monolayer 134 form a bilayer.

[0041] In some embodiments, the third material comprises one of graphene, carbon nanotubes (CNTs), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide (e.g., vanadium disulfide (VS2), vanadium diselenide (VSe2), cobalt sulfide (CoS2), cobalt selenide (CoSe2), titanium disulfide (TiS2), or titanium diselenide (TiSe2)), a semiconducting transition metal dichalcogenide (e.g., molybdenum disulfide (MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), or tungsten diselenide (WSe2)), and a high-k material.

[0042] In some embodiments, the bilayer is a heterojunction bilayer, in which the second and third materials are different from one another, and in other embodiments, the bilayer is a homojunction bilayer, in which the second and third materials are similar to one another.

[0043] In some embodiments, l and w are each selected to have a value equal to the product of the incubation time of additional nuclei of the second material on the single-domain monolayer and the growth rate of the second material. For example, if the measured lateral growth rate of WSe2 and the incubation time of the second nuclei are approximately 0.4 μm / min and 5 minutes, respectively, l and w are selected to have a value equal to 5 minutes × 0.4 μm / min = 2 μm. In such an embodiment, l and w are each selected to be 2 microns.

[0044] In another embodiment, the method further comprises forming a semiconductor device (a valleytronic device, a forked-sheet field effect transistor (FET) or a complementary FET) comprising the single-domain monolayer.

[0045] In one embodiment, the method includes depositing at least one of hafnium oxide (HfO), aluminum oxide (AlO), hafnium zirconium oxide (HZO), titanium dioxide (TiO), zinc oxide (ZnO), iron oxide (FeO), tin oxide (SnO), nickel oxide (NiO), and copper oxide (CuO) onto silicon to form the substrate before depositing the mask material on the substrate.

[0046] Another embodiment of the present invention provides a method including the steps of: (a) providing a substrate 100 of a first material, the first material having a first Gibbs free energy; (b) depositing a mask material 110 on the substrate 100, the mask material 110 having a second Gibbs free energy, the second Gibbs free energy being greater than the first Gibbs free energy; and (c) forming a trench array on the mask material 110, the trench array (d) depositing adatoms 120 of a second material on the exposed portions of the substrate 100; and (e) selecting the adatoms 120 into first nuclei 122 within each trench 102 in the trench array. (f) growing first nuclei 122 in each trench 102 in the trench array, wherein growth of the first nuclei 122 into a first single-domain monolayer 124 of a second material is limited by the lateral dimensions of the trench geometry; (g) waiting for an incubation period; (h) depositing another adatom 130 of a third material on top of the first single-domain monolayer 124 of at least one trench in the trench array; (i) growing the first nuclei 122 in each trench 102 in the trench array; (i) selectively nucleating another adatom 130 of a third material into a second nucleus 132 on top of the first single domain monolayer 124 in at least one trench in the trench array; and (j) growing the nuclei 132 in at least one trench in the trench array, wherein the growth of the second nuclei 132 into the second single domain monolayer 134 of the third material is limited by the lateral dimensions of the trench geometry, and the first single domain monolayer 124 and the second single domain monolayer 134 form a bilayer.

[0047] In some embodiments, the first material comprises one of hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HZO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO).

[0048] In some embodiments, the second material or the third material comprises one of graphene, carbon nanotubes (CNTs), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide (e.g., vanadium disulfide (VS2), vanadium diselenide (VSe2), cobalt sulfide (CoS2), cobalt selenide (CoSe2), titanium disulfide (TiS2), or titanium diselenide (TiSe2)), a semiconducting transition metal dichalcogenide (e.g., molybdenum disulfide (MoS2), molybdenum diselenide (MoSe2), tungsten disulfide (WS2), or tungsten diselenide (WSe2)), and a high-k material (e.g., Bi2SeO5 or Sb2O3).

[0049] In one embodiment, the mask material (110) is amorphous silicon dioxide (a-SiO2), amorphous silicon (a-Si), amorphous silicon nitride (a-SiN x ), amorphous carbon (a-carbon), hafnium oxide (HfO2), aluminum oxide (Al2O3), hafnium zirconium oxide (HfZrO), titanium dioxide (TiO2), zinc oxide (ZnO), iron oxide (Fe2O3), tin oxide (SnO2), nickel oxide (NiO), and copper oxide (CuO).

[0050] Any combination of the concepts listed above with additional concepts discussed below (to the extent such concepts are not mutually inconsistent) is part of the inventive subject matter of this disclosure. In particular, any combination of subject matter set forth in the claims presented at the end of this disclosure is part of the inventive subject matter of this disclosure. Terminology used in this application, and which may appear in any incorporated-by-reference disclosure, should be associated with the meaning that most closely matches the specific concepts of this disclosure.

[0051] Figures 2A-2D are images of WSe2 grown in an array of confinement areas (trench) of different widths for different durations using the process in Figure 1B. Figures 2A and 2B show WSe2 grown in 10 μm-wide trenches for 20 and 30 minutes, respectively. Figures 2C and 2D show WSe2 grown in 2 μm-wide trenches for 7 and 10 minutes, respectively.

[0052] As shown in Figure 2A, growth of WSe2 in 10-μm-wide trenches for 20 min (significantly longer than the 5-min second-nucleation incubation period) resulted in multiple nuclei in some trenches. Approximately 70% of the trenches in Figure 2A contained only a single WSe2 triangle, while other trenches contained multiple nuclei. Extending the growth time to 30 min resulted in the formation of multidomain polycrystalline WSe2, as shown in Figure 2B. In contrast to the single-domain areas, a significant decrease in photoluminescence (PL) intensity was observed for the multidomain areas. Of the trenches with multiple nuclei, 25% contained more than two nuclei at the edge, while the remaining nuclei were located in the center. However, as shown in Figures 2C and 2D, reducing the trench width to 2 μm resulted in only one nucleus per trench, regardless of the nucleation location, whether homogeneous in the center or heterogeneous at the edge.

[0053] Without wishing to be bound by any theory, confined growth effectively suppresses additional heterogeneous nucleation in the trenches by reducing the size of the potential nucleation area. Further growth of the nuclei results in single-crystalline MLs filling the entire corresponding trench, as shown in Figures 2C and 2D. Because each confined WSe2 ML originates from a single nucleus, the individual WSe2 MLs in the trench array become single crystals on the wafer. Approximately 97% of the 2 μm-wide trenches were completely filled with WSe2 during the incubation period of the second nuclei on top of the first WSe2 layer.

[0054] Figures 2E and 2F show the E and E spectra of WSe grown in a 2-µm-wide trench, respectively. 1 2g Raman mapping of the peak position and photoluminescence (PL) mapping at 1.65 eV are shown. The height of each trench can be varied from 5 to 100 nm. These measurements confirm that the WSe2 grown in the trenches is entirely ML-WSe2. The average full width at half maximum (FWHM) of the PL spectrum of the single-domain WSe2 in the trenches was measured to be approximately 55 meV at room temperature, which is similar to that of high-quality single-domain WSe2 flakes mechanically isolated from bulk WSe2.

[0055] Measurements showed that growing another WSe2 layer on top of these single ML-WSe2 arrays could result in uniform homogeneous BLs that are electrically superior to the MLs. As shown in Figure 1B, each trench contained only a single domain, filling each pocket. Further extending the growth time resulted in the formation of a single-domain BL-WSe2 that completely filled the trench. Within each trench, there was only one additional WSe2 nucleus, regardless of its nucleation mode; i.e., approximately 60% of the secondary nuclei were heterogeneously formed, seeded from the sidewall of the SiO2 mask. This suggests that the A associated with the confined MLs is a key factor in determining the BL-WSe2 nuclei. 1g Peak (approx. 259.6 cm -1 ) from B related to confinement BL 1 2g Peak (approx. 308.5cm -1 ) was also confirmed by the Raman spectral shift from 1.65 eV to 1.6 eV. In addition, the transition from the direct gap to the indirect gap was confirmed by the PL spectral peak shift from 1.65 eV to 1.6 eV.

[0056] B 1 2gRaman mapping at the peak and PL mapping at 1.6 eV, shown in Figures 2G and 2H, respectively, confirm the presence of uniform wafer-scale BL-WSe2 material within these trenches across the wafer. High-resolution transmission electron microscopy (HRTEM) images of 0.8 nm-thick ML-WSe2 and 1.6 nm-thick BL-WSe2 are shown in Figures 2I and 2J, respectively. Scanning tunneling electron microscopy (STEM) images confirm that BL-WSe2 has grown without strain at the center or edge of the SiO2 trenches. In addition, planar HAADF-STEM analysis reveals that the BL-WSe2 is epitaxially aligned with AA' stacking.

[0057] Confined growth of single-domain TMD homo- or hetero-BLs at precisely defined locations on a wafer enables the use of TMDs as CMOS channel materials. Figure 3A shows a field-effect transistor (FET) array on BL-WSe2 grown on a 2-inch wafer using the confined growth method depicted in Figures 1B and 1C. The wafer in Figure 3A is a sapphire wafer, with back-gated devices fabricated using metal-induced migration. Other substrates, such as Si wafers with appropriate masking materials, and other top-gating strategies are also possible.

[0058] Figure 3B shows a typical drain-source current vs. gate-source voltage (I ds -V gs ) characteristics are shown. These FETs are 8 On / off current ratio of over 240.5mV / dec, subthreshold swing (SS) of 155.8μAμm -1 The maximum on-current (I on ) density, V ds = 103.5cm at -1V 2 V -1 s -1 The field-effect mobility (μ eff =g m L?(WC g Vds ), C g =11.6nFcm -2 As shown in Figure 3C, the saturation current is 465 μA μm -1 This has led to the following.

[0059] Figure 3D shows benchmark results of the turn-on current / effective mobility characteristics of the WSe2 FET array in Figure 3A against WSe2 FETs fabricated using other methods. As shown in Figure 3E, the electrical characteristics of FETs fabricated with confined-grown ML / BL-WSe2 are comparable to the best reported performance for single-crystal WSe2-based FETs and are similar to (or better than) those of exfoliated flake-based ML / BL-WSe2 FETs.

[0060] Figure 3F shows the turn-on current per unit width, I on and effective mobility μ eff A statistical analysis of the I per unit width of FET arrays is presented. on and μ eff Both show Gaussian distributions, and their mean and variance are I on The density is 89.9 μA μm -1 ,17.3%,μ eff Regarding the dimensions, each is 79.1cm. 2 V -1 s -1 , 24.1%. 213 FETs were fabricated with confined BL-WSe2, achieving an estimated yield of 93.9%. In addition, the electrical performance of FETs fabricated with confined ML-WSe2 was comparable to that of FETs fabricated with confined BL-WSe2. FETs fabricated with confined BL-WSe2 exhibited performance comparable to that of flake-based FETs, uniformly across the entire array.

[0061] Our confined growth technique can also be used to fabricate logic devices and memory. Figure 4A shows a single-crystalline TMD (here, MoS2) grown on a 10 nm layer of amorphous HfO2 deposited on a Si wafer with a SiO2 mask defining a trench approximately 1 μm wide. DFT calculations confirm that the binding energy of the TMD to HfO2 is larger than that to SiO2. However, the binding energy difference between sapphire and SiO2 is larger than that between HfO2 and SiO2. This can lead to a shortened incubation time between the first and second nucleation events when growing TMDs within a SiO2 mask on a HfO2 surface. This shortened incubation time can be mitigated by reducing the trench size to 1 μm and adjusting the carrier gas and growth conditions, thereby further enhancing the HfO2-SiO2 selectivity.

[0062] As shown in Figures 4B and 4C, only one nucleation event occurs within each 1 μm trench. Successful filling of the trenches by further growth results in single-crystalline ML-MoS2 within the confined region at the amorphous HfO2 surface. Layer-by-layer growth is also possible. Using the confined ML / BL-MoS2 on the HfO2 substrate, we were able to fabricate FETs to verify the electrical properties of the confined ML / BL-MoS2. Figure 4D shows the measured transfer characteristics (I ds -V ds ) are shown. These FETs have a current rating of 86.7 μA μm -1 (ML-MoS2) and 129.3 μA μm -1 (BL-MoS2) on The density is C g HfO2 =600nFcm -2 and V ds = 62.2cm when 1V 2 V -1 s -1 and 88.61 cm 2 V -1 s -1 μ eff(Fig. 4E). The electrical properties of these MoS2 FETs are similar to those of single-crystal MoS2-based FETs fabricated using other techniques.

[0063] Subsequent confined growth of single-domain MLs can be used to fabricate hetero-BL-TMD semiconductors. For example, ML-MoS2 can be grown on an array of single-domain ML-WSe2. A single-domain MoS2 / WSe2 hetero-BL can be formed by confining and growing a single MoS2 nucleus within a WSe2-filled trench over the entire area. Raman mapping and PL spectroscopy confirmed the uniform formation of the MoS2 / WSe2 hetero-BL. In addition, as shown in the cross-sectional HAADF-STEM image in Figure 4F, the van der Waals (vdW) hetero-interface between the confined ML-MoS2 and ML-WSe2 is sharp and completely free of alloying. The uniform hetero-interface without secondary nucleation can be observed in low-magnification HAADF-STEM images.

[0064] The valleytronic performance of the MoS2 / WSe2 hetero-BL array is depicted in Figure 4G. Specifically, Figure 4G shows the valley-polarized carrier dynamics in the confined ML-WSe2 and hetero-BL (MoS2 / WSe2) arrays, characterized by ultrafast circular dichroism (CD) based on time-resolved pump-probe spectroscopy. While a valley lifetime (e.g., about tens of picoseconds) was observed in the confined ML-WSe2 arrays due to fast valley depolarization caused by exchange interactions, a longer valley lifetime (e.g., about hundreds of picoseconds) was observed in the confined hetero-BL arrays at 300 K due to ultrafast charge separation. Additionally, an extended valley lifetime (about several nanoseconds) was also observed at 77 K due to reduced phonon scattering.

[0065] Using confined growth techniques, arrays of single-domain 2D TMDs can be synthesized at the wafer scale. These confined growth techniques enable layer-by-layer synthesis via critical Gibbs free energy differences, enabling the fabrication of homo-layer (WSe2 / WSe2) and hetero-layer (MoS2 / WSe2) array structures. In addition, arrays of confined-layer WSe2 transistor devices have been shown to exhibit performance comparable to that of devices fabricated with exfoliated WSe2 at the wafer scale. Thus, confined growth techniques address the difficulty of controlling 2D material kinetics at the wafer scale, a major obstacle associated with 2D TMDs. Their application to large-scale integration of single-crystal vdW provides a new route for the construction of 2D material-based electronics platforms.

[0066] [Simulation, Creation and Measurement] DFT calculations for selective confinement growth of TMDs. DFT calculations were performed using the Vienna Ab Initio Simulation Package (VASP) with projector-augmented wave (PAW) pseudopotentials and a plane-wave basis set. The generalized gradient approximation (GGA) of the Purdue-Burke-Ernserhof (PBE) functional was used to describe the electronic exchange-correlation interactions. The valence electron configurations of W, Se, O, Al, and Si are 6s, respectively. 2 5d 4 ,4s 2 4p 4 ,2s 2 2p 4 ,3s 2 3p 1 ,3s 2 3p 2 The energy cutoff for the plane wave expansion was set to 420 eV. Since the DFT calculations used large cells (lattice constants > 10 Å), the Brillouin zone was sampled using a Γ-point only k-point grid (1 Å = 10 -10m). Surface bonding interactions were investigated by disposing WO3, Se, and WSe2 clusters on a-HfO2, Al2O3 (0001), and a-SiO2 slabs, respectively. The HfO2, SiO2, and Al2O3 surfaces were passivated with H atoms to mimic the Ar / H2 ambient growth environment. Amorphous HfO2 and SiO2 atomic structures were obtained by subjecting the crystal structures to a melt-quenching process simulated by ab initio molecular dynamics. The structures were optimized by relaxing the top adsorbate atoms anchored to the substrate atoms. The structural relaxation criterion was that the force acting on the yeast atoms was less than 0.01 eV / Å. The system energy difference between two successive iterations was 10 -5 The electronic minimization occurred when the surface binding energy of adsorbent A to substrate B was less than ξ eV. b =E A / B -E A -E B , but E A / B , E A and E B were calculated using the adsorption system A / B, the energy of the adsorbent A when isolated, and the energy of the substrate B, respectively.

[0067] Confinement pattern fabrication: For the confined growth of TMDs, a sapphire substrate was coated with LOR 3A and photoresist (PR) S1805 and patterned using an AS200 inline stepper (AutoStep200). An approximately 25-nm-thick a-SiO2 layer was deposited on the PR-patterned sapphire substrate using an electron beam evaporator. To create sapphire pockets (trenches), the SiO2 pattern was lifted off using Remover PG (Kayaku Advanced Materials) and rinsed in acetone and isopropanol for 15 min each.

[0068] Synthesis of WSe2 and MoS2. Confined TMDs were synthesized in a 4-inch diameter quartz tube. With the zone distance fixed at 33 cm, 300 mg of Se or S powder was placed in (zone I) and 30 mg of WO3 or MoO3 powder was placed in (zone II). A SiO2-patterned sapphire substrate was vertically loaded 6 cm behind the WO3 or MoO3 powder, and the substrate was covered with quartz plates on both sides to reduce or minimize direct reaction.

[0069] Before synthesizing the confined ML-WSe2, the air in the quartz tube was removed by a vacuum pump. After closing the vacuum valve, the tube was filled with a carrier gas of Ar (50 sccm) / H2 (50 sccm) and then the air valve was opened. The Ar / H2 ratio was continuously maintained. The growth temperatures in (Zone I) and (Zone II) were set at 15 °C min ?1 ,30℃min ?1 The temperature was increased at a ramp rate of 450°C (zone I) and 890°C (zone II) for 10 minutes, and then the temperature was naturally reduced to room temperature.

[0070] For the confined BL-WSe2, the synthesis of the second WSe2 layer was carried out with a carrier gas ratio of Ar (35 sccm) / H2 (65 sccm). For the confined heterostructure (MoS2 / WSe2), MoS2 synthesis was carried out at 8 °C min ?1 ,30℃min ?1 The reactions were carried out at 200 °C (Zone I) and 750 °C (Zone II) with different ramp rates. Specifically, to improve the growth selectivity on the HfO substrate, the size of the SiO trenches was reduced to 1 μm, and the total Ar / H flow rate was increased from 100 sccm [Ar (50 sccm) / H (50 sccm)] to 200 sccm [Ar (100 sccm) / H (100 sccm)]. All reactions were carried out at atmospheric pressure, resulting in TMD powders with a purity of 99.99%.

[0071] Characterization of the Confined TMDs. Raman and PL spectra were measured using a Renishaw® InViaReflex® microspectrometer with a 532 nm pump laser. The light was dispersed by a holographic grating with 2400 grooves per mm. To obtain Raman and PL mapping images, the specimen was scanned on an xy piezoelectric stage with laser illumination. SEM images were acquired using a high-resolution scanning electron microscope (ZEISS® Merlin®) with an in-lens detector. The working distance was 6 mm at an accelerating voltage of 2 kV and a probe current of 70 pA. TEM characterization was performed using a JEOL® JEM-2100F® microscope at an accelerating voltage of 200 kV and a STEM (Titan ThemisZ G3 Cs-Corrected® microscope) at an accelerating voltage of 60 kV. EDX line profiles were collected using characteristic X-ray signals of MoKα, SKα, WLα, and SeKα with Velox® software in STEM mode. XPS spectra were measured with a magnesium Kα source (MultiLab2000®, Thermo VG), and the peak energy was calibrated by the C1s peak at 284.8 eV. AFM morphology analysis was performed using an XE100® (Park Systems Corp.).

[0072] Device fabrication and electrical measurements. For device fabrication using confined ML / BL-WSe2, a 600-nm-thick Au film was deposited on the confined WSe2 / sapphire by electron beam evaporation. The Au / WSe2 stack was peeled off and transferred onto a 300-nm-thick SiO2 / heavily p-doped silicon wafer using thermal release tape as a handling layer. After removing the thermal release tape on a hot plate at 120 °C, tape residue was removed from the Au film by oxygen plasma treatment. The Au film was then etched with an Au etchant and rinsed with deionized water. (Few-layer WSe2 flakes were also transferred in the same manner to compare their electrical properties.)

[0073] After transferring the confined ML / BL-WSe2 to the SiO2 substrate, alignment marks for electron beam lithography (EBL) were patterned on the SiO2 substrate using an optical lithography process, followed by deposition of 2.5 nm thick Ti and 7.5 nm thick Au using an electron beam evaporator. Then, drain and source contact regions with a width of 2 μm were patterned using EBL. Polymethyl methacrylate (PMMA) A4 and PMMA A6 were spin-coated at 3000 rpm and baked at 180 °C for 150 s to form EBL photoresists. After the PMMA was developed, a 10 nm thick Pt layer and an 80 nm thick Au layer were deposited using an electron beam evaporator. Finally, the areas other than the source / drain contact metal regions were removed by a lift-off process. The same process was followed for fabricating devices on HfO2 using the confined ML / BL-MoS2, from patterning the alignment marks for EBL to depositing PMMA. Then, a 10 nm thick Ni layer and an 80 nm thick Au layer were deposited using an electron beam evaporator, followed by lift-off.

[0074] Its current-voltage characteristics were measured with an Agilent® B2900A® source / measurement unit. All measurements were performed in air at room temperature. In addition, the 2-inch confinement BL-WSe2 was measured in a 5.1 × 5.1 cm 2 The BL-WSe2 was then transferred onto a 300-nm-thick SiO2 / Si substrate of varying size. A highly uniform work function distribution (5.08 eV) was confirmed on the confined BL-WSe2 by Kelvin probe force microscopy (KPFM). CH ) 0.7 μm source and drain electrodes were integrated on it using platinum, and the hole barrier height was estimated to be 0.31 eV through modified Richardson plotting.

[0075] Time-resolved pump-probe spectroscopy. We investigated valley-polarized carrier dynamics by measuring ultrafast CD using time-resolved pump-probe spectroscopy. Femtosecond laser pulses were provided by a 100 kHz Yb-based regenerative amplifier system (Light Conversion PHAROS™), and wavelength-tunable pump-and-probe pulses resonant with the A exciton resonance of WSe2 were generated by a sequential optical parametric amplifier (ORPHEUS™) with a pulse duration of 50 fs and a spectral bandwidth of 50 meV. The specimen on a cryostat was illuminated by the pump excitation pulses through a 40x objective. Pump-induced changes in the probe reflectivity were recorded as a function of the time delay introduced by a mechanical translation stage and a lock-in amplifier. The polarization profiles of the pump-and-probe pulses were independently controlled by a half-wave plate and a quarter-wave plate. The signals were measured for pump-probe pulses with circular polarizations of the same helicity (co-polarization) and those with circular polarizations of the opposite helicity (cross-polarization). The valley-dependent ultrafast CD response shown in Figures 4F and 4G was captured by the difference between the co-polarized and cross-polarized pump-probe responses.

[0076] [Conclusion] While various embodiments of the present invention have been described and illustrated herein, various other means and / or structures which can perform the functions and / or obtain the results and / or one or more advantages described herein, and each variation and / or modification thereof, are deemed to be within the scope of the embodiments of the invention described herein, as will be readily apparent to those skilled in the art. More generally, those skilled in the art will readily appreciate that any parameters, dimensions, materials, and configurations described herein are intended to be exemplary, and that the actual parameters, dimensions, materials, and / or configurations will depend on the particular application or applications for which the teachings of the present invention are used. Those skilled in the art will recognize or be able to ascertain, using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Accordingly, it is to be understood that the foregoing embodiments are presented by way of example only, and that, within the scope of the appended claims and their equivalents, embodiments of the invention may be practiced otherwise than as specifically described and claimed. Inventive embodiments of the present disclosure are directed to each individual feature, system, article, material, kit, and / or method described herein. In addition, any combination of two or more such features, systems, articles, materials, kits, and / or methods, provided that such combinations are not mutually inconsistent, is encompassed within the inventive scope of the present disclosure.

[0077] Also, various concepts of the present invention can be implemented as one or two methods, examples of which have been presented. The operations performed as part of the method can be ordered in any suitable manner. Thus, embodiments can be constructed in which operations are performed in an order different from that depicted, for example, some operations can be performed simultaneously, even if shown as sequential operations in the illustrative embodiments.

[0078] All definitions provided and used herein should be understood to control dictionary definitions, definitions in incorporated documents, and / or ordinary meanings of the defined words.

[0079] The indefinite articles "a" and "an" as used in the specification and claims of this application should be understood to mean "at least one" unless a clear indication to the contrary is provided.

[0080] The phrase "and / or," as used in the specification and claims of this application, should be understood to mean "one or both" of the elements it conjoins, i.e., elements that are present jointly in some cases and not jointly in other cases. Multiple elements listed with "and / or" should also be interpreted in the same manner, i.e., "one or more" of the elements it conjoins. Other elements, related or unrelated to the elements specifically identified by the "and / or" clause, may optionally be present. Thus, as a non-limiting example, a reference to "A and / or B" used in conjunction with an open term, such as "comprising," may, in one embodiment, refer to only A (optionally including elements other than B); in another embodiment, it may refer to only B (optionally including elements other than A); in yet another embodiment, it may refer to both A and B (optionally including other elements); and so forth.

[0081] As used in the specification and claims of this application, "or" should be understood to have the same meaning as "and / or" as defined above. For example, "or" or "and / or" when setting forth list terms will be understood to be inclusive, i.e., including at least one of the elements of a plurality or list, and optionally including more than one, and optionally additional terms not in the list. Exclusionary terms, such as "only one of" or "exactly one of," or "consisting of" as used in the claims, will refer to the inclusion of exactly one element of a plurality or list. In general, the word "or" as used herein will be interpreted exclusively as indicating exclusive alternatives (i.e., "one or the other, but not both") when preceded by exclusive terms, such as "any of," "one of," "only one of," or "exactly one of,". "Consisting essentially of," as used in the claims, shall be understood to have its ordinary meaning as used in the field of patent law.

[0082] In the specification and claims of this application, the phrase "at least one," when used in reference to a list of one or more elements, should be understood to mean at least one element selected from any one or more of the elements in the list of elements, but not necessarily including at least one of each individual element explicitly listed within the list of elements, and not excluding any combination of elements in the list of elements. By this definition, elements other than those specifically identified in the list of elements referred to by the phrase "at least one," whether related or unrelated to the elements specifically identified, may optionally be present. That is, as a non-limiting example, "at least one of A and B" (or its equivalent "at least one of A or B" or its equivalent "at least one of A and / or B") can refer in one embodiment to at least one, and optionally more than one, A but no B (optionally including elements other than B); in another embodiment to at least one, and optionally more than one, B but no A (optionally including elements other than A); in yet another embodiment to at least one, and optionally more than one, A and at least one, and optionally more than one, B (optionally including other elements); and so on.

[0083] In the claims, as well as in the foregoing specification, all transitional phrases, such as "comprise," "include," "carry," "have," "include," "accompany," "hold," "composed of," etc., are to be understood in an open sense, i.e., inclusive but not limited to. Only the transitional phrases "consisting of" and "consisting essentially of" shall be closed or semi-closed transitional phrases, respectively, as explained in MPEP Section 2111.03.

Claims

1. 1. A method comprising: (a) providing a substrate (100) of a first material; (b) depositing a mask material (110) on the substrate (100); (c) forming a trench array on the mask material (110), the trench array comprising a plurality of trenches (102), each having a trench geometry with lateral dimensions of l x w, where l and w are each selected to have a maximum dimension of 2 μm, each trench (102) having an exposed portion of the substrate (100) surrounded by sidewalls (112) formed by the mask material (110); (d) depositing adatoms (120) of a second material on the exposed portions of the substrate (100), wherein a first bond energy between the first material and the second material is greater than a second bond energy between the mask material (110) and the second material; (e) selectively nucleating the adatoms (120) into nuclei (122) within each trench (102) in the trench array; (f) growing the nuclei (122) in each trench (102) in the trench array; wherein growth of the nuclei (122) into a single-domain monolayer (124) of the second material is limited by the lateral dimensions of the trench geometry.

2. 2. The method of claim 1, wherein the first material is hafnium oxide (HfO ２ ), aluminum oxide (Al ２ O ３ ), hafnium zirconium oxide (HZO), titanium dioxide (TiO ２ ), zinc oxide (ZnO), iron oxide (Fe ２ O ３ ), tin oxide (SnO ２ ), nickel oxide (NiO), and copper oxide (CuO).

3. 10. The method of claim 1, wherein the second material comprises one of graphene, carbon nanotubes (CNT), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide, a semiconducting transition metal dichalcogenide, and a high-k material.

4. 4. The method of claim 3, wherein the metallic transition metal dichalcogenide is vanadium disulfide (VS ２ ), vanadium diselenide (VSe ２ ), cobalt sulfide (CoS ２ ), cobalt selenide (CoSe ２ ), titanium disulfide (TiS ２ ) and titanium diselenide (TiSe ２ ) is one of the methods.

5. 4. The method of claim 3, wherein the semiconducting transition metal dichalcogenide is molybdenum disulfide (MoS ２ ), molybdenum diselenide (MoSe ２ ), tungsten disulfide (WS ２ ) and tungsten diselenide (WSe ２ ) is one of the methods.

6. 4. The method of claim 3, wherein the high-k material is Bi. ２ SeO ５ and Sb ２ O ３ A method that is one of the above.

7. 2. The method of claim 1, wherein the mask material (110) is amorphous silicon dioxide (a-SiO ２ ), amorphous silicon (a-Si), amorphous silicon nitride (a-SiN ｘ ), amorphous carbon (a-carbon), hafnium oxide (HfO ２ ), aluminum oxide (Al ２ O ３ ), hafnium zirconium oxide (HfZrO), titanium dioxide (TiO ２ ), zinc oxide (ZnO), iron oxide (Fe ２ O ３ ), tin oxide (SnO ２ ), nickel oxide (NiO), and copper oxide (CuO).

8. 2. The method of claim 1, wherein the core (122) is a first core (122) and the single-domain monolayer (124) is a first single-domain monolayer (124), further comprising: (g) waiting for an incubation period; (h) depositing another adatom (130) of a third material on top of the first single-domain monolayer (124) in at least one trench in the trench array, wherein a third bond energy between the second material and the third material is greater than the second bond energy between the mask material (110) and the third material; (i) selectively nucleating said separate adatoms (130) of said third material into second nuclei (132) on top of said first single-domain monolayer (124) in said at least one trench in said trench array; (j) growing the nuclei (132) in the at least one trench in the trench array; wherein growth of the second nuclei (132) into the second single-domain monolayer (134) of the third material is limited by the lateral dimensions of the trench geometry; and The method wherein said first single-domain monolayer (124) and said second single-domain monolayer (134) form a bilayer.

9. 9. The method of claim 8, wherein the third material comprises one of graphene, carbon nanotubes (CNT), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide, a semiconducting transition metal dichalcogenide, and a high-k material.

10. 10. The method of claim 9, wherein the metallic transition metal dichalcogenide is vanadium disulfide (VS ２ ), vanadium diselenide (VSe ２ ), cobalt sulfide (CoS ２ ), cobalt selenide (CoSe ２ ), titanium disulfide (TiS ２ ) and titanium diselenide (TiSe ２ ) is one of the methods.

11. 10. The method of claim 9, wherein the semiconducting transition metal dichalcogenide is molybdenum disulfide (MoS ２ ), molybdenum diselenide (MoSe ２ ), tungsten disulfide (WS ２ ) and tungsten diselenide (WSe ２ ) is one of the methods.

12. 9. The method of claim 8, wherein the bilayer is a heterojunction bilayer in which the second and third materials are different from one another.

13. 9. The method of claim 8, wherein the bilayer is a homojunction bilayer in which the second and third materials are similar to each other.

14. 2. The method of claim 1, wherein forming the trench (102) includes etching partway through the mask material (110) into the substrate (100).

15. 2. The method of claim 1, wherein l and w are each selected to have a value equal to the product of the incubation time of another nucleus of said second material on said single-domain monolayer and the growth rate of said second material.

16. 2. The method of claim 1, wherein l and w are each selected to be 2 microns.

17. 10. The method of claim 1, further comprising: A method of forming a semiconductor device comprising said single domain monolayer (124).

18. 20. The method of claim 18, wherein the semiconductor device comprises one of a valleytronic device, a fork-sheet field effect transistor (FET), and a complementary FET.

19. 10. The method of claim 1, further comprising, before depositing a mask material (110) on the substrate (100): Hafnium oxide (HfO ２ ), aluminum oxide (Al ２ O ３ ), hafnium zirconium oxide (HZO), titanium dioxide (TiO ２ ), zinc oxide (ZnO), iron oxide (Fe ２ O ３ ), tin oxide (SnO ２ ), nickel oxide (NiO), and copper oxide (CuO) are deposited on silicon to form the substrate (100).

20. 1. A method comprising: (a) providing a substrate (100) of a first material, the first material having a first Gibbs free energy; (b) depositing a mask material (110) on the substrate (100), the mask material (110) having a second Gibbs free energy, the second Gibbs free energy being greater than the first Gibbs free energy; (c) forming a trench array on the mask material (110), the trench array comprising a plurality of trenches (102), each having a trench geometry with lateral dimensions of l x w, where l and w are each selected to have a maximum dimension of 2 μm, each trench (102) having an exposed portion of the substrate (100) surrounded by sidewalls (112) formed from the mask material (110); (e) depositing adatoms (120) of a second material on the exposed portions of the substrate (100); (f) selectively nucleating the adatoms (120) into first nuclei (122) within each trench (102) in the trench array; (g) growing the first nuclei (122) in each trench (102) in the trench array, wherein growth of the first nuclei (122) into a first single-domain monolayer (124) of the second material is limited by the lateral dimensions of the trench geometry; (h) waiting for an incubation period; (i) depositing another adatom (130) of a third material on top of the first single-domain monolayer (124) in at least one trench in the trench array; (j) selectively nucleating the separate adatoms (130) of the third material into second nuclei (132) on top of the first single-domain monolayer (124) within the at least one trench in the trench array; (k) growing the nuclei (132) in the at least one trench in the trench array; wherein growth of the second nuclei (132) into the second single-domain monolayer (134) of the third material is limited by the lateral dimensions of the trench geometry; and The method wherein said first single-domain monolayer (124) and said second single-domain monolayer (134) form a bilayer.

21. 21. The method of claim 20, wherein the first material is hafnium oxide (HfO ２ ), aluminum oxide (Al ２ O ３ ), hafnium zirconium oxide (HZO), titanium dioxide (TiO ２ ), zinc oxide (ZnO), iron oxide (Fe ２ O ３ ), tin oxide (SnO ２ ), nickel oxide (NiO), and copper oxide (CuO).

22. 21. The method of claim 20, wherein the second material or the third material comprises one of graphene, carbon nanotubes (CNT), hexagonal boron nitride (h-BN), a metallic transition metal dichalcogenide, a semiconducting transition metal dichalcogenide, and a high-k material.

23. 23. The method of claim 22, wherein the metallic transition metal dichalcogenide is vanadium disulfide (VS ２ ), vanadium diselenide (VSe ２ ), cobalt sulfide (CoS ２ ), cobalt selenide (CoSe ２ ), titanium disulfide (TiS ２ ) and titanium diselenide (TiSe ２ ) is one of the methods.

24. 23. The method of claim 22, wherein the semiconducting transition metal dichalcogenide is molybdenum disulfide (MoS ２ ), molybdenum diselenide (MoSe ２ ), tungsten disulfide (WS ２ ) and tungsten diselenide (WSe ２ ) is one of the methods.

25. 23. The method of claim 22, wherein the high-k material is Bi ２ SeO ５ and Sb ２ O ３ A method that is one of the above.

26. 21. The method of claim 20, wherein the mask material (110) is amorphous silicon dioxide (a-SiO ２ ), amorphous silicon (a-Si), amorphous silicon nitride (a-SiN ｘ ), amorphous carbon (a-carbon), hafnium oxide (HfO ２ ), aluminum oxide (Al ２ O ３ ), hafnium zirconium oxide (HfZrO), titanium dioxide (TiO ２ ), zinc oxide (ZnO), iron oxide (Fe ２ O ３ ), tin oxide (SnO ２ ), nickel oxide (NiO), and copper oxide (CuO).

27. 21. The method of claim 20, wherein the bilayer is a heterojunction bilayer in which the second and third materials are different from one another.

28. 21. The method of claim 20, wherein the bilayer is a homojunction bilayer in which the second and third materials are similar to each other.

Images