CN210120150U

CN210120150U - Defect-removing monocrystalline substrate adopting 2D material epitaxy

Info

Publication number: CN210120150U
Application number: CN201920473733.5U
Authority: CN
Inventors: 王晓靁; 刘家桓; 宋高梅
Original assignee: Individual
Current assignee: Individual
Priority date: 2019-04-10
Filing date: 2019-04-10
Publication date: 2020-02-28
Anticipated expiration: 2029-04-10

Abstract

The utility model discloses an adopt 2D material epitaxy to go bad single crystal substrate, adopt the 2D material van der Waals epitaxial growth 2D material ultrathin layer of separation effect as the mesosphere on low-cost GaN single crystal substrate or other low-cost GaN quasi-homogeneous material single crystal substrates, van der Waals epitaxial growth high quality GaN or GaN base epitaxial layer on 2D material ultrathin layer, 2D material ultrathin layer comprises single material or more than one materialStacking layers are formed, and the range of the epitaxial conditions of the quasi-homogeneous material single crystal substrate is as follows: lattice constant mismatch of not more than 5% and difference in coefficient of thermal expansion of not more than 1.5X 10^‑6℃^‑1. The utility model discloses can obtain high-quality gaN single crystal substrate, simplify epitaxy and subassembly process for the base plate material of adoption selects the possibility more widely, and manufacturing cost greatly reduced is favorable to marketing and application.

Description

Defect-removing monocrystalline substrate adopting 2D material epitaxy

Technical Field

The utility model relates to a LED's technical field, in particular to adopt the ultra-thin intermediate level of 2D material homogeneous or accurate homogeneous epitaxy to go bad single crystal substrate.

Background

In the fabrication of light emitting diode (led) or Laser Diode (LD) devices, epitaxy has a significant impact on the quality of the product. Wherein the influence on the quality includes even the luminous efficiency, the durability, etc. The reason is that the light emitting diode particularly requires that electrons and holes cooperate with each other when the crystal is excited to generate photons smoothly. In contrast, if a defect is generated on a material structure or a texture, the possibility that the mutual combination of electrons and holes is hindered by the defect increases, resulting in deterioration of the light emitting effect. The main luminescent material of the led is gallium nitride (GaN), which is usually epitaxially grown on a substrate, and the crystal structure and structure of the produced GaN are largely affected by the substrate. In order to improve the light emitting efficiency, durability and other characteristics related to the quality of the led, several conditions are generally considered in the art when selecting a suitable substrate material. In general, the substrate material is desirably a single crystal material that minimizes the defect density, and crystal quality of the led is not affected during the epitaxy process as much as possible when the crystal structure, lattice constant (lattice constant), and Coefficient of Thermal Expansion (CTE) are matched to the epitaxial material.

According to the prior art, the most commonly used substrate material is single crystal Sapphire (Sapphire), and the advantages of good chemical stability, mature manufacturing technology and the like are mainly considered; and due to recent increases in productivity, sapphire substrates are becoming a relatively popular alternative to other alternatives, such as: aluminum nitride (AlN), and even gallium nitride (GaN) substrates, etc., are more economically desirable. However, because sapphire is not ideal in terms of matching crystal structure, lattice constant (lattice constant), Coefficient of Thermal Expansion (CTE) and epitaxial material, the defect density of GaN or AlGaN epitaxial layer is high, which affects the application of Laser Diode (LD) and the performance improvement of ultraviolet light emitting diode (UV LED); the UVC LED light-emitting wavelength which belongs to the deep ultraviolet light range has the most disinfection and sterilization efficiency, the current mercury lamp with low efficiency energy consumption and harmful environment is effectively replaced, the UVC LED light-emitting wavelength has great development potential in civil and daily disinfection and sterilization application, but the existing aluminum nitride substrate production technology which is most suitable for the UV LED has a bottleneck, the development of the UVC LED mainly focuses on a sapphire substrate with poor matching degree, and the performance improvement is greatly hindered.

In other words, if a single crystal substrate of the above two materials is directly produced by a melt-growth method, not only the production cost is increased, but also relatively more waste heat is generated, which causes unavoidable environmental pollution. In the Vapor Phase growth process, the Hydride Vapor Phase Epitaxy (HVPE) method is currently used for growing gallium nitride crystals to produce single-crystal gallium nitride substrates, and due to the limitations of production cost and yield conditions, the current mass production technology reaches 4 inches of substrates and the cost is extremely high. In fact, the defect density of the vapor phase method is still higher than that of other liquid phase crystal growth processes, but the crystal growth rate of the rest processes is too slow, the volume production cost is higher, and the commercial main flow is still limited to the HVPE method under the consideration of market demand, device performance and substrate cost and supply trade-off. The literature indicates that the vapor phase method GaN growth rate still has the possibility of increasing several times and maintaining good crystallinity, but is limited by the deterioration of defect density and is not currently oriented to reduce the cost of GaN substrates. As for the aluminum nitride crystal growth technology, a Physical Vapor Transport (PVT) method, which is one of Vapor phase methods, is used to produce the single crystal aluminum nitride substrate, because of the limitations of production technology and yield, only two manufacturers have mass production capability globally, the cost is very high when the current mass production technology only reaches 2 inches of substrates, and the capacity cannot be widely supplied to the market because of the occupation of a few manufacturers. Due to the chemical characteristics of aluminum nitride and the limitation of hardware components by a physical vapor transport method, carbon (C) and oxygen (O) impurities exist in a single crystal finished product to a certain degree inevitably, and the component characteristics are also influenced to a certain degree.

TABLE 1

Zinc oxide (ZnO) single crystal materials are a suitable choice of substrate materials in the previous section in terms of crystal structure, thermal properties, and lattice constant, and therefore have attracted research efforts by technical developers. However, zinc oxide is not widely used in the art today, and the main reasons include that zinc oxide has high chemical activity and is easily corroded by hydrogen-containing substances during the subsequent epitaxy process, which results in poor quality of the epitaxy layer, as shown in fig. 1, zinc rapidly diffuses into the epitaxy layer while hydrogen etching occurs on the zinc oxide substrate during the epitaxy process, which results in poor quality of the epitaxy layer, and zinc and oxygen diffuse and dope into the crystal grains of the light emitting diode while the process is adjusted to improve the epitaxy quality, which causes the light emitting characteristics not to meet expectations, so that the structure cannot meet the actual market requirements.

The same situation may also exist in other opto-electronic component substrate-epitaxial combinations currently in use, such as silicon carbide (SiC) or gallium arsenide (GaAs) etc.; the single crystal silicon carbide substrate is a substrate material of a high-performance power semiconductor and a high-end light emitting diode at present, a single crystal growing process is a Physical Vapor Transport (PVT) method in a vapor phase method, the growing technology of the high-quality large-size silicon carbide single crystal is difficult, the high-end mass production technology is mastered by a few manufacturers, and the influence on the application cost is still a great progress space.

Two-dimensional (2D) materials are an emerging field of rapid development, the most well-known material in the 2D material family, which most early attracts a great deal of research and development investment, is graphene (graphene), the two-dimensional layered structure of which has special or excellent physical/chemical/mechanical/photoelectric properties, no strong bonding exists between layers, and only van der waals bonding exists, which also means that the surface of the layered structure does not have any dangling bond (dangling bond), and graphene is currently identified to have a wide range and excellent propertiesExcellent application potential; graphene development work is widely carried out all over the world, and meanwhile, the development of more 2D materials is also driven, including hexagonal Boron nitride hbn (hexagonal Boron nitride), transition metal dichalcogenides tmds (transition metal dichalcogenides), black phosphorus and the like are also accumulated in the 2D material family, as shown in fig. 2 and 3, the materials respectively have specific material characteristics and application potential, and the development of manufacturing technologies of related materials is continuously and actively promoted. MoS of one of graphene, hBN and TMDs material in addition to excellent photoelectric characteristics₂Are considered to have excellent diffusion barrier properties and varying degrees of high temperature stability, and in particular hBN is considered to have excellent chemical inertness (inertness) and high temperature oxidation resistance.

Due to the nature of the above-mentioned layered structure and the characteristics of van der waals bonding between layers, the technical feasibility of fabricating two or more materials in the 2D family of materials into a layered-stacked heterostructure (hetero-structure) is greatly opened, and it is possible for the heterostructure to create new application characteristics or fabricate new devices in addition to combining different characteristics, and the development in the field of optoelectronics and semiconductors is currently very active, as shown in fig. 4a and 4b, which are schematic diagrams of mechanically composed layers, and fig. 5a and 5b, which are schematic diagrams of physical or chemical vapor deposition.

The van der Waals force binding properties of 2D materials have also gained attention for the use of epitaxial substrates for conventional 3D materials, focusing on the fact that epitaxial materials in epitaxial technology must match very well with the substrate material in terms of crystal structure, lattice constant (lattice constant), Coefficient of Thermal Expansion (CTE), but in reality they are subject to situations such as lack of suitability for the substrate material, or high or not easily available substrate material, when 2D materials offer another solution for heteroepitaxial substrates, namely the so-called van der Waals epitoxy. The mechanism by which van der waals epitaxy may be favored over heteroepitaxy is that the direct chemical bonding at the conventional epitaxial interface is replaced by van der waals bonding, which allows some relaxation of the stress or strain energy from lattice and thermal expansion mismatch during the epitaxy process, thereby improving the quality of the epitaxial layer, or alternatively, some of the previously impractical heteroepitaxy techniques are possible by the 2D material and van der waals epitaxy introduction. Related studies have also shown that when the above 2D materials are stacked on top of each other in a heterostructure, the interaction forces are dominated by van der waals forces; when the Epitaxy of the 3D material is performed on the 2D material, the Epitaxy is not substantially pure van der Waals epitaxiy (van der Waals epitaxiy) or more precisely can be regarded as Quasi van der Waals epitaxiy (Quasi van der Waals epitaxiy) because the existence of dangling bonds (dangling bonds) of the 3D material on the interface simultaneously contributes to the bonding force of the interface; in any case, the degree of lattice and thermal expansion matching still certainly contributes to the final epitaxial quality, and the overall matching degree is contributed by the 2D material interposer and the substrate material. The above 2D layer material has a hexagonal or honeycomb structure, and is considered to be structurally compatible with Wurtzite and Zinc Blende (Zinc-blend) structure materials at an external time delay, and the related art main epitaxial materials of the utility model all belong to such a structure.

Taking mass production of blue laser diodes as an example, since laser diode components are very sensitive to the defect density of the epitaxial active layer, according to the prior art, the adopted substrate material is a single crystal gallium nitride (GaN) substrate produced by Hydride Vapor Phase Epitaxy (HVPE) which is one of vapor Phase methods, and the cost is extremely high when the current mass production technology reaches 4 inches of substrates due to the limitations of production cost, yield conditions and the like. In fact, the defect density of the vapor phase method is still higher than that of other liquid phase crystal growth processes, but the crystal growth rate of the rest processes is too slow, the volume production cost is higher, and the commercial main flow is still limited to the HVPE method under the consideration of market demand, device performance and substrate cost and supply trade-off.

Table 2 shows the comparison of various GaN crystal growth methods under different conditions (1atm ═ 1.01325X 10)⁵Pa，1inch＝2.54cm)。

As shown in FIG. 6, the prior art directly epitaxially grows a high-quality GaN or GaN-based epitaxial layer on a high-quality GaN single crystal substrate, but on the premise of extremely high cost of the GaN single crystal substrate, the GaN single crystal substrate by HVPE method is used as trade-off of the business flow, and the defect density is about 10% higher than that of other GaN single crystal or sapphire single crystal substrates in the liquid phase crystal growth process²To 10⁴And (4) doubling.

As described above, the growth rate of GaN by vapor phase method such as HVPE is still possible to be increased several times and maintain good crystallinity, but it is limited by deterioration of defect density and has not been considered as an orientation for reducing the cost of GaN substrate. If the crystal growth condition can be relaxed to improve the production efficiency, the cost of the produced GaN single crystal substrate can be effectively reduced, and the GaN single crystal substrate with high defect density but relatively low cost can be obtained. In fact, GaN epitaxial growth with high crystallinity (XRD FWHM can reach 200arcsec) has been successfully achieved on the surface of 2D materials; the barrier property of the 2D material layer and the van der Waals force between the 2D material layer and the 3D material are mainly combined, so that the damage of defects (dislocation, impurities and the like) in the substrate material to the quality of the epitaxial layer can be effectively prevented; when the main body of the epitaxial substrate and the epitaxial layer are still homogeneous GaN, the matching of the thermal properties of crystal lattices is also ensured, and the favorable foundation of homogeneous epitaxy is maintained.

SUMMERY OF THE UTILITY MODEL

In order to solve the above problems, an object of the present invention is to provide a defect-removing single crystal substrate using a 2D material epitaxy.

In order to achieve the above purpose, the solution of the present invention is:

using 2D material to epitaxially debulk a single crystal substrate, growing a 2D material ultra-thin layer on a low-cost GaN single crystal substrate (e.g. from the constraint of relaxing the growth conditions to improve the yield, such as HVPE method to increase the growth rate) or other low-cost GaN quasi-homogeneous material single crystal substrate (e.g. ZnO material) with good barrier effect by using 2D material Van Der Waals Epitaxy (VDWE) as an intermediate layer, growing a high-quality (i.e. high crystallinity, XRD FWHM can be lower than 400arcsec) GaN or GaN-based epitaxial layer on the 2D material ultra-thin layer, wherein the 2D material ultra-thin layer is made of a single material or a material stack of more than one materialLayer formation, Quasi-homogeneous material single crystal substrate epitaxy (Quasi-homogeneous epitaxial) having lattice constant mismatch (CTE) of not more than 5% and Coefficient of Thermal Expansion (CTE) difference of not more than 1.5 × 10^-6℃^-1。

The thickness of the 2D material ultrathin layer ranges from 0.5nm to 1000 nm.

The 2D material ultrathin layer is a single material with a good blocking effect, such as hexagonal boron nitride (hBN) and graphene (graphene).

The 2D material ultrathin layer is of a composite layer structure, and the top layer is made of a 2D material which is well matched with GaN in lattice or has high surface energy and is beneficial to epitaxy, such as WS₂Or MoS₂And the bottom layer is made of 2D material with good barrier effect, such as hexagonal boron nitride (hBN) and graphene (graphene).

A metal catalyst layer is added between the substrate and the intermediate layer, the total thickness of the metal catalyst layer ranges from 0.5nm to 3000nm, and the metal catalyst layer comprises Fe, Co, Ni, Au, Ag, Cu, W, Mo, Ru or Pt.

After the scheme of the adoption, the utility model discloses a good 2D material of separation effect covers the substrate material surface and regards as the epitaxial intermediate layer of high-quality GaN, carry out van der Waals epitaxy or accurate van der Waals epitaxial technique and use, the ultrathin layer of 2D material is as the barrier layer, come the harm that defect in the separation substrate material caused epitaxial layer quality and subassembly performance, defect in the base plate includes point defect (like oxygen ion or other impurity) and line defect (like the dislocation), can obtain high-quality GaN single crystal substrate, simplify epitaxy and subassembly process, make the substrate material optional possibility of adoption wider, and manufacturing cost greatly reduced is favorable to marketing and application.

Drawings

FIG. 1 is a schematic diagram of a zinc oxide substrate being attacked during epitaxy;

FIG. 2 is a schematic diagram of the structure of a two-dimensional transition metal dichalcogenide TMDs;

FIG. 3 is a schematic structural diagram of hexagonal boron nitride hBN, a two-dimensional material;

FIGS. 4a and 4b are schematic views of a mechanically composed laminate;

FIGS. 5a, 5b are schematic illustrations of physical and chemical vapor deposition;

FIG. 6 is a schematic view showing the growth of a conventional GaN single crystal substrate;

fig. 7 is a schematic structural diagram of an embodiment of the present invention;

FIG. 8 is a schematic structural diagram of a second embodiment of the present invention;

fig. 9 is a schematic view of the manufacturing assembly of the present invention.

Detailed Description

The present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.

As shown in fig. 7 and 8, the 2D epitaxial defect removal single crystal substrate disclosed in the present invention is formed by growing a 2D ultrathin layer 3 of material by Van Der Waals Epitaxy (VDWE) with good barrier effect on a low-cost GaN single crystal substrate 1 or other low-cost GaN quasi-homogeneous single crystal substrates 1, and growing a high-quality GaN or GaN-based epitaxial layer 2 on the 2D ultrathin layer 3 by van der waals epitaxy.

Among them, the low-cost GaN single crystal substrate 1 is a GaN single crystal substrate obtained by relaxing the limitation of the growth conditions to improve the productivity, such as a GaN single crystal substrate obtained by increasing the growth rate by the HVPE method. The other low-cost GaN quasi-homogeneous material single crystal substrate 1 is a material such as ZnO which is produced at a lower cost than the conventional HVPE GaN. The range of the epitaxial conditions of the quasi-homogeneous material single crystal substrate 1 is as follows: lattice constant mismatch of not more than 5% and difference in coefficient of thermal expansion of not more than 1.5X 10^-6℃^-1. The 2D ultra-thin layer is made of a single material or a laminate of more than one material. By high quality GaN or GaN-based epitaxial layer 2 is meant high crystallinity, XRD FWHM (full width at half maximum XRD pattern) can be lower than 400 arcsec. The thickness of the 2D ultrathin layer 3 ranges from 0.5nm to 1000 nm. The ultra-thin 2D material layer 3 shown in fig. 7 is a single material with good barrier effect, such as hexagonal boron nitride hBN and graphene (graphene). The ultra-thin layer 3 of 2D material shown in FIG. 8 is a composite intermediate layer, and the top layer 31 is made of 2D material with good lattice matching with GaN or high surface energy for epitaxy, such as WS₂Or MoS₂While the bottom layer 32 is made of a resist2D materials with good isolation effect, such as hexagonal boron nitride (hBN) and graphene (graphene). The lattice constants of the various materials are shown in Table 3.

TABLE 3

The utility model discloses a good 2D material ultra-thin layer of separation effect 3 or bottom 32 cause the harm to epitaxial layer quality and subassembly performance as barrier layer (barrier) come the defect in the separation base plate material, defect in the base plate includes point defect (like oxygen ion or other impurity) and line defect (like the dislocation), can obtain high quality GaN single crystal substrate, simplify epitaxy and subassembly process, make the base plate material of adoption select the possibility more broadly, manufacturing cost greatly reduced is favorable to marketing and application.

In order to obtain a better structure, the utility model discloses can increase metal catalysis layer 4 at the surface of 2D material cover base plate 1 material, metal catalysis layer 4 can include Fe, Co, Ni, Au, Ag, Cu, W, Mo, Ru or Pt etc. and metal catalysis layer 4 grows in advance or deposits on base plate 1 surface, also can need the heat treatment process, and metal catalysis layer 4 gross thickness scope is at 0.5nm to 3000 nm.

The utility model also discloses a preparation method of adopting 2D material epitaxy to remove defect monocrystalline substrate, the step is as follows:

in a first step, the substrate 1 (chip) material is subjected to epitaxial growth-level polishing as a starting material and is prepared for a subsequent fabrication process via appropriate pre-treatments (including chip cleaning).

After the first step and before the second step, the metal catalyst layer 4 can be added at a proper time according to the growth requirement of the 2D material. The growth or deposition process of the 2D material covering the surface of the substrate 1 may require the metal catalyst layer 4 including Fe, Co, Ni, Au, Ag, Cu, W, Mo, Ru, Pt, or the like to be grown or deposited on the surface of the substrate 1 in advance, or may require a heat treatment process. The total thickness of the metal catalyst layer 4 is in the range of 0.5nm to 3000 nm.

Secondly, covering the 2D material with good barrier effect on the surface of the substrate 1 material as an intermediate layer by utilizing Van der Waals epitaxy or quasi Van der Waals epitaxy technology; there may be a single or multiple layer 2D material ultra-thin layer 2 covering. The 2D material may be applied to the surface of the substrate 1 material by existing processes including growth, deposition, transfer, coating, etc., and associated necessary pre-and post-treatment processes. The total thickness of the single layer or the multiple layers ranges from 0.5nm to 1000 nm.

And thirdly, growing a high-quality GaN or GaN-based epitaxial layer 2 on the intermediate layer by utilizing Van der Waals epitaxy or quasi Van der Waals epitaxy technology.

As shown in FIG. 9, various photoelectric semiconductor device products manufactured by the method of the present invention can be directly manufactured into GaN template (GaN template) products with high-quality GaN epitaxial layers on the surfaces by the aforementioned method; the GaN template finished product can also be prepared by peeling the high-quality GaN epitaxial layer on the surface prepared by the method from the original substrate, bonding the substrate with other substrate materials, and then performing necessary processes to prepare a component or bonding the component with other substrate materials.

The foregoing is only a preferred embodiment of the present invention, and is not intended to limit the scope of the invention. It should be noted that after reading this description, those skilled in the art can make equivalent changes according to the design concept of the present application, which fall within the protection scope of the present application.

Claims

1. Adopt 2D material epitaxy to remove defective monocrystal base plate, its characterized in that: 2D material van der Waals epitaxial growth 2D material ultrathin layer with blocking effect is adopted on a GaN single crystal substrate or a GaN quasi-homogeneous material single crystal substrate to serve as an intermediate layer, GaN or a GaN-based epitaxial layer is subjected to van der Waals epitaxial growth on the 2D material ultrathin layer, the 2D material ultrathin layer is formed by single materials or stacking more than one material, the mismatching degree of the epitaxial lattice constant of the quasi-homogeneous material single crystal substrate is not more than 5%, and the difference of the thermal expansion coefficient is not more than 1.5 multiplied by 10^-6℃^-1。

2. An epitaxial defect-removing single crystal substrate using a 2D material according to claim 1, wherein: the thickness of the 2D material ultrathin layer ranges from 0.5nm to 1000 nm.

3. An epitaxial defect-removing single crystal substrate using a 2D material according to claim 1, wherein: the 2D material ultrathin layer is a single material with a barrier effect.

4. An epitaxial defect-removing single crystal substrate using a 2D material according to claim 1, wherein: the 2D material ultrathin layer is of a composite layer structure, the top layer is made of a 2D material matched with GaN in lattice or high in surface energy, and the bottom layer is made of a 2D material with a blocking effect.

5. An epitaxial defect-removing single crystal substrate using a 2D material according to claim 1, wherein: the GaN or GaN-based epitaxial layer is high in crystallinity, and XRD FWHM is lower than 400 arcsec.

6. An epitaxial defect-removing single crystal substrate using a 2D material according to claim 1, wherein: a metal catalyst layer is added between the substrate and the intermediate layer, the total thickness of the metal catalyst layer ranges from 0.5nm to 3000nm, and the metal catalyst layer comprises Fe, Co, Ni, Au, Ag, Cu, W, Mo, Ru or Pt.