Ultra-wide bandgap semiconductor devices containing magnesium germanium oxide

Abstract

Various forms of Mg x Ge 1-x O 2-x has been disclosed, and Mg x Ge 1-x O 2-x is an epitaxial layer formed on a substrate comprising a substantially single crystalline substrate material. x Ge 1-x O 2-x The epitaxial layers of Mg have a crystal symmetry that matches the substrate material. x Ge 1-x O 2-x Semiconductor structures and devices including the epitaxial layers are disclosed, along with methods for making the epitaxial layers and the semiconductor structures and devices.

Description

[Technical Field]

[0001] Related Applications This application is related to U.S. Patent Application No. 16 / 990,349, filed August 11, 2020, entitled "Metal Oxide Semiconductor-Based Light Emitting Device," International Application No. PCT / IB2021 / 060413, filed November 10, 2021, entitled "Epitaxial Oxide Materials, Structures and Devices," and International Application No. PCT / IB2021 / 060427, filed November 10, 2021, entitled "Epitaxial Oxide Materials, Structures and Devices," all of which are incorporated herein by reference for all purposes. [Background technology]

[0002] Electronic and optoelectronic devices rely on semiconductor materials to provide properties that allow the devices to function in a variety of ways for various applications. Optoelectronic devices include light-emitting diodes (LEDs), which emit light at infrared, visible, or ultraviolet wavelengths depending on the bandgap of the material used in the LED. For example, ultraviolet (UV)-emitting LEDs utilize wide-bandgap (WBG) semiconductors such as gallium nitride (GaN) and aluminum nitride (AlN). Photodetectors are another type of optoelectronic device that senses the presence of light. For example, UV detection of light is used in diverse applications such as communication systems, data storage, biosensing, and fluorescence measurements. Lasers and solar-blind detectors are other types of optoelectronic devices.

[0003] Digital power management systems use electronic devices for direct current (DC) and alternating current (AC) power conversion, such as DC-DC and AC-DC conversion. These power inverters and power conversion devices are primarily constructed using power transistors with high breakdown voltages and fast switching times. The properties of the semiconductor materials used in electronic devices affect performance characteristics such as breakdown voltage resistance, on-state electrical losses, transistor switching speed, and overall switching efficiency. High radio frequency switching transistors are used in telecommunications and radar systems, which further require the use of low insertion loss materials and high signal gain. Summary of the Invention [Problem to be solved by the invention]

[0004] The continued development of semiconductor materials is critical to the continued improvement and advancement of electronic and optoelectronic devices. [Means for solving the problem]

[0005] In some embodiments, the semiconductor structure comprises a substrate comprising a substantially single crystalline substrate material and a Mg x Ge 1-x O 2-x and an epitaxial layer of Mg, where x has a value of 0≦x<1. x Ge 1-x O 2-x The epitaxial layer has a crystal symmetry that matches the substrate material.

[0006] In some embodiments, a semiconductor device includes a substrate comprising a substantially single-crystalline substrate material and an active region on the substrate. The active region comprises Mg x Ge 1-x O 2-x The epitaxial layer of Mg is composed of 0≦x<1. x Ge 1-x O 2-x The epitaxial layer has a crystal symmetry that matches the substrate material.

[0007] In some embodiments, the method of forming a semiconductor device includes: x Ge 1-x O 2-x providing a substrate comprising a substantially single crystalline substrate material having a crystal symmetry compatible with an epitaxial layer of Mg; x Ge 1-x O 2-x (where the value of x is 0≦x<1), the material includes at least two elements selected from Mg, Ge, and oxygen according to the value of x, and Mg, Ge, and oxygen are supplied by a Mg source, a Ge source, and an activated oxygen source, respectively.

[0008] In other embodiments, Mg x Ge 1-x O 2-x Additional elements are incorporated into the crystal. [Brief explanation of the drawings]

[0009] [Figure 1A] FIG. 1 is a perspective view of a cubic symmetric MgO structure showing the atomic positions of magnesium and oxygen atoms in the unit cell, according to some embodiments. [Figure 1B] FIG. 1B is a plan view along the axis of cubic MgO shown in FIG. 1A. [Figure 2A] FIG. 1 is a perspective view of a single crystal unit cell comprising trigonal germanium oxide, GeO 2 , according to some embodiments. [Figure 2B] 2B is a plan view of the single crystal unit cell of trigonal germanium oxide GeO2 shown in FIG. 2A. [Figure 3A] FIG. 1 is a perspective view of a single crystal unit cell comprising tetragonal germanium oxide, GeO 2 , according to some embodiments. [Figure 3B] 3B is a plan view of the single crystal unit cell of tetragonal germanium oxide GeO2 shown in FIG. 3A. [Figure 4] 1 is a table of single crystal compositions that can be formed from various proportions of Mg, Ge, and O, according to some embodiments. [Figure 5]1 is a table showing examples of possible single crystal compositions that can be formed from Mg, Ge, and O that exhibit low formation energies and therefore stable structures, according to some embodiments. [Figure 6] 1 is a plot showing calculated formation energies of different single crystal compositions formed from Mg, Ge, and O (i.e., MgGeOz), according to some embodiments. [Figure 7A] FIG. 1 is a representation of a fundamental unit cell defining a Mg2GeO4(Pnma)-type single crystal composition, according to some embodiments. [Figure 7B] FIG. 1 is a representation of a fundamental unit cell defining a MgGeO(C2 / c) type single crystal composition, according to some embodiments. [Figure 7C] FIG. 1 is a representation of a fundamental unit cell defining a Mg2GeO4(Fd3m)-type single crystal composition, according to some embodiments. [Figure 8A] FIG. 7D is an electron energy-momentum (Ek) diagram for Mg2GeO4(Fd3m) showing the calculated valence and conduction band structures at critical points within the Brillouin zone of the crystal disclosed in FIG. 7C, according to some embodiments. [Figure 8B] FIG. 8B is an Ek diagram detailing the valence band structure of the Ek diagram for Mg2GeO4(Fd3m) shown in FIG. 8A, in accordance with some embodiments. [Figure 9A] FIG. 7B is an Ek diagram for Mg2GeO4(Pnma) showing the calculated valence and conduction band structures at critical points within the Brillouin zone of the crystal disclosed in FIG. 7A, according to some embodiments. [Figure 9B] FIG. 9B is an Ek diagram detailing the valence band structure of the Ek diagram for Mg2GeO4(Pnma) shown in FIG. 9A, in accordance with some embodiments. [Figure 10A] FIG. 7C is an E diagram for MgGeO(C / c) showing the calculated valence and conduction band structures at critical points within the Brillouin zone of the crystal disclosed in FIG. 7B, according to some embodiments. [Figure 10B]FIG. 10B is an Ek diagram detailing the valence band structure of the Ek diagram for MgGeO3(C2 / c) shown in FIG. 10A, in accordance with some embodiments. [Figure 11] Disclosed are local bonding configurations in AB2O4-type Fd3m cubic crystals in which A and B cations are bonded to oxygen (O) atoms, according to some embodiments. [Figure 12] 1 is a table showing examples of compatible substrate compositions and orientations on which single crystalline compositions Mg2GeO4(Pnma), MgGeO3(C2 / c), and Mg2GeO4(Fd3m) can be epitaxially formed, according to some embodiments. [Figure 13A] Two different free-standing crystalline structures corresponding to the film and the substrate, according to some embodiments, both having cubic or tetrahedral symmetry, with the film having a larger parallel (i.e., in-plane) lattice constant than the substrate.

number

number

number

number

number

number

[0010] Semiconductor structures having magnesium germanium oxide epitaxial layers are disclosed, in which the magnesium germanium oxide layer and substrate materials are carefully designed and utilized in a specific combination to enable the formation of high-quality crystalline structures. Embodiments of the semiconductor structures include a substantially single-crystalline substrate with a predetermined crystal symmetry that matches the specific morphology of magnesium germanium oxide in the epitaxial layer. The semiconductor structures include a new form of magnesium germanium oxide, forming a substantially single-crystalline structure. The semiconductor structures include at least one region of magnesium germanium oxide with a predetermined single-crystalline symmetry type. Conductively doped embodiments of magnesium germanium oxide materials and methods for forming semiconductor structures are disclosed. Embodiments of the magnesium germanium oxide materials exhibit ultra-wide bandgaps, such as from 5.2 eV for the monoclinic crystalline form to 6.4 eV for the orthorhombic crystalline form. Some embodiments have direct bandgaps, such as a cubic crystalline form of magnesium germanium oxide with a direct bandgap of 5.8 eV. The disclosed materials, structures, and methods enable high quality, efficient semiconductor devices, including optoelectronic devices operating in the deep ultraviolet region, as well as electronic devices for radio frequency (RF) communications, digital computing, and power switching.

[0011] In this disclosure, magnesium germanium oxide (Mg x Ge y O z ) comprises the elements germanium (Ge), magnesium (Mg), and oxygen (O), where x, y, and z are mole fractions of x having any value in the range of 0≦x≦1.5, y having any value in the range of 1.0≦y≦2.5, and z having any value in the range of 2≦z≦5. Embodiments are also contemplated in this disclosure as including Mg a Ge b O c (a, b, c are Mg x Gey O z (corresponding to x, y, z in the formula), or x has any value in the range of 0≦x<1 x Ge 1-x O 2-x as, or 0 <y≦1のMg 1-y Ge y O 1+y One expression for magnesium germanium oxide (e.g., Mg x Ge y O z ), as described in this paragraph, may be used in other expressions (e.g., Mg x Ge 1-x O 2-x ) can be similarly substituted.

[0012] In the electronics and optoelectronics industries, semiconductors with novel material and functional properties are increasingly required for modern applications. While silicon is the workhorse of the digital computing and power switching industries, the performance of power switches in digital power management systems must be improved for new application areas to become commercially viable. In particular, the development of wide-bandgap (WBG) semiconductor materials is required. Furthermore, optoelectronic devices such as light-emitting diodes (LEDs) require unique, material-specific criteria that enable efficient emission at infrared, visible, and ultraviolet wavelengths. UV LEDs use WBG semiconductors such as GaN and AlN. Those skilled in the art are well aware that efficient optical emission from semiconductor materials requires the fundamental property of possessing a direct bandgap energy-momentum configuration.

[0013] However, advantageous direct bandgaps are only possible for a small subset of crystalline structure types and atoms that make up single-crystalline compositions of semiconductors. Conventional UV LEDs (e.g., AlN-based UV LEDs) have technological and fundamental physical challenges that prevent them from achieving deep-ultraviolet wavelengths shorter than about 260 nm, such as less than about 220 nanometers (nm). It would be desirable to provide UV LEDs that can produce wavelengths even shorter than 250 nm and increase the light output of UV LEDs for applications such as inactivation of viruses and biocidal contamination of surfaces, liquids, and air.

[0014] Short UV wavelength sensors are generally limited by available semiconductor bandgap energies and available technology, so short UV wavelength sensors are another type of device that can benefit from further development of WBG semiconductor materials.

[0015] Another desirable application of WBG semiconductor materials is directed towards digital power management systems such as DC-DC and AC-DC conversion. Power switches in such systems are traditionally made of Si(E g =1.1 eV), GaN (E g =3.4 eV), and silicon carbide (SiC, E g =3.1 eV), where E g is the energy gap between the conduction band minimum and the valence band maximum. Although GaN and SiC are classified as WBGs, there is a pressing need to further increase the available bandgap energy to further improve breakdown voltage resistance, reduce on-state electrical losses, and increase transistor switching speed and therefore overall switching efficiency.

[0016] Clearly, conventional WBG materials have technical limitations for many applications, including ultrashort UV LEDs and lasers, high-efficiency power switching systems, and solar-blind detectors. There is an urgent need to develop new materials to overcome these drawbacks and limitations. The wide band gap of epitaxial oxide materials results in high breakdown voltages, allowing them to be used in electronic devices that require large voltage biases and can withstand high electric fields (e.g., high-voltage switches and impact ionization devices). The band gap of epitaxial oxide materials also makes them suitable for use in optoelectronic devices that emit or detect light in the UV range. Materials with band gaps of approximately 4.5 eV to approximately 8 eV can be used to emit or detect UV light with wavelengths between approximately 150 nm and 280 nm. Other disclosed ultrawide band gap oxide materials can be used for extreme ultraviolet applications below 150 nm. Semiconductor heterostructures can also be formed using wide band gap materials as emitter or absorber layers, and materials with wider band gaps than the emitter or absorber layers can be used in other layers of the structure to make them transparent to the wavelengths being emitted or absorbed.

[0017] The magnesium germanium oxide materials disclosed herein can be utilized in a variety of electronic and optoelectronic devices, including, but not limited to, light-emitting devices (e.g., UV LEDs), optical sensors (e.g., short UV wavelength sensors), and power switches, as described above. Other applications include functional oxide surfaces forming bioelectronic interfaces for biosensing. In embodiments, the magnesium germanium oxide materials can be used in one or more layers of the active region of a semiconductor device, as described herein (e.g., n-type active region or layer(s), i-type active region or layer(s), and / or p-type active region or layer(s)). The i-type region shall also be referred to as the intrinsic region or region that is not intentionally doped.

[0018] As described herein, an "epitaxial oxide" material is a material containing oxygen and other elements (e.g., a metal such as gallium or a non-metal such as germanium and silicon) with an ordered crystalline structure configured to form on a single crystal substrate or one or more layers formed on the single crystal substrate. Epitaxial oxide materials have a defined crystalline symmetry and crystalline orientation relative to the substrate. Epitaxial oxide materials can form intimately contacted layers with the single crystal substrate and / or one or more layers formed on the single crystal substrate. Epitaxial oxide materials can exist within layers of strained semiconductor structures, where the crystals of the epitaxial oxide material are distorted compared to a relaxed state. Epitaxial oxide materials can also exist within layers of unstrained or relaxed semiconductor structures.

[0019] In the present disclosure, the crystal symmetry of a substrate and an epitaxial oxide material may be compatible if they have compatible crystal symmetry and the in-plane (i.e., parallel to the surface of the substrate) lattice constants and atomic positions of the substrate surface provide a suitable template for the subsequent growth of the epitaxial oxide material. For example, a substrate and an epitaxial oxide material may be compatible if the in-plane lattice constant mismatch between the substrate and the epitaxial oxide material is less than 0.5%, 1%, 1.5%, 2%, 5%, or 10%. For example, in some embodiments, the crystal structure of the substrate material has a lattice mismatch with the epitaxial layer of 10% or less. In some cases, the crystal symmetry of a substrate and an epitaxial oxide material may be compatible even if they have different types of crystal symmetry if the in-plane (i.e., parallel to the surface of the substrate) lattice constants and atomic positions of the substrate surface provide a suitable template for the subsequent growth of the epitaxial oxide material. In some cases, a plurality (e.g., 2, 4, or other integer) of unit cells of the atomic arrangement of the substrate surface can provide a suitable surface for the growth of an epitaxial oxide material having a unit cell larger than that of the substrate. In other cases, the epitaxial oxide layer may have a smaller lattice constant (e.g., about half) than that of the substrate. In some cases, the unit cell of the epitaxial oxide layer may be rotated (e.g., 45 degrees) relative to that of the substrate.

[0020] The epitaxial oxide materials described herein can be formed using epitaxial growth techniques such as molecular beam epitaxy (MBE), metalorganic chemical vapor deposition (MOCVD), atomic layer deposition (ALD), and other physical vapor deposition (PVD) and chemical vapor deposition (CVD) techniques.

[0021] Crystal structure The present invention relates to the formation of epitaxial Mg x Ge y O z Specific forms of Mg were synthesized using substrates carefully selected based on their crystal symmetry compatibility with the membrane. x Ge y O zThe discovery of the ability to fabricate epitaxial single crystal semiconductor structures of Mg is then disclosed. x Ge y O z To provide background on film formation and exemplary property tuning, insight into crystal structure is provided.

[0022] Density functional theory (DFT) allows for the prediction and calculation of band structures of crystalline oxides based on quantum mechanics without the need for phenomenological parameters. DFT calculations applied to understand the electronic properties of solid oxide crystals are essentially based on treating the nuclei of the atoms that make up the crystal as fixed by the Born-Oppenheimer approximation, thereby generating a static external potential incorporating many-body electron fields. The crystal structure symmetry of the atomic positions and atomic species imposes an effective potential of the basic structure on the interacting electrons. The effective potential for many-body electron interactions in three-dimensional spatial coordinates can be realized by utilizing functionals for the electron density. This effective potential includes exchange and correlation interactions representing interacting and non-interacting electrons. For applications to solid semiconductors and oxides, various improved exchange functionals (XCFs) exist that improve the accuracy of DFT results. Within the DFT framework, the many-electron Schrödinger equation is divided into two groups: (i) valence electrons and (ii) core electrons. The inner-shell electrons are strongly bonded, partially shielding the nucleus and forming an inert nucleus with it. Atomic bonding in crystals is primarily due to valence electrons. Therefore, in many cases, inner-shell electrons can be ignored, and the atoms that make up a crystal are reduced to ionic nuclei that interact with the valence electrons. This effective interaction, called a pseudopotential, approximates the potential felt by the valence electrons. One notable exception to the influence of inner-shell electrons is the case of lanthanide oxides, where the partially filled 4f orbitals of lanthanide atoms are surrounded by closed electron orbitals. The DFT band structures disclosed herein account for this effect. Many refinements to XCF exist to achieve greater accuracy in band structures applied to oxides.For example, known local density approximation (LDA), generalized gradient approximation (GGA) hybrid exchange (e.g., HSE (Heyd-Scuseria-Ernzerhof), PBE (Perdew-Burke-Ernzerhof), and BLYP (Becke, Lee, Yang, Parr)) include the use of the Tran-Blaha modified Becke-Johnson (TBmBJ) exchange function and further modifications such as the KTBmBJ, JTBSm, and GLLBsc formalisms. In accordance with the present disclosure, and particularly for the presently disclosed materials, the TBmBJ exchange potential has been found to be capable of predicting the electron energy-momentum (Ek) band structure, band gap, lattice constant, and several mechanical properties of epitaxial oxide materials. An additional advantage of TBmBJ is its reduced computational cost compared to HSE when applied to large numbers of atoms in large supercells used to simulate smaller perturbations to ideal crystal structures, such as the incorporation of impurities. It is also expected that further improvements beyond TBmBJ, particularly as applied to the present oxide system, can be achieved. DTF calculations are used extensively in this disclosure to provide ab initio insight into the electronic and physical properties of the epitaxial oxide materials described herein, including the band gap and the nature of the band gap, whether direct or indirect. The electronic and physical properties of the epitaxial oxide materials can be used in the design of semiconductor structures and devices utilizing the epitaxial oxide materials. In some cases, experimental data has also been used to verify the properties of the epitaxial oxide materials and structures described herein.

[0023] Described herein are calculated Ek band diagrams for epitaxial oxide materials derived using DFT calculations. The Ek diagrams have several features that can be used to understand the electronic and physical properties of epitaxial oxide materials. For example, the energies and k vectors of the valence and conduction band extrema indicate the approximate energy width of the band gap and whether the band gap is direct or indirect in nature. The curvature of the valence and conduction band branches near the extrema is related to the effective masses of holes and electrons, which in turn are related to the carrier mobility within the material. As verified by experimental data, DFT calculations using the TBmBJ exchange functional more accurately indicate the size of the band gap of a material compared to previous exchange functionals. The calculated band diagrams for epitaxial materials in this disclosure may differ in some respects from the actual band diagrams of the epitaxial materials. However, certain features, such as the valence and conduction band extrema and the curvature of the valence and conduction band branches near the extrema, may closely correspond to the actual band diagrams of the epitaxial materials. Thus, even if some details of the band diagram are inaccurate, the calculated band diagrams of the epitaxial materials of the present disclosure provide useful insight into the electronic and physical properties of epitaxial oxide materials and can be used in the design of semiconductor structures and devices that utilize epitaxial oxide materials.

[0024] 1A and 1B show perspective and plan views, respectively, of the cubic symmetry MgO structure. The figures show the atomic positions of the magnesium ("Mg") and oxygen ("O") atoms in a single unit cell characterized by the lattice constant a0. Cubic MgO is the most stable form of this composition and rarely forms any other polytypes whose unit cells fall into the cubic Fm3m space group. In this crystal structure, Mg atoms are octahedrally bonded to oxygen atoms. Cubic MgO is known and can be readily formed as bulk substrates or films. However, embodiments of the present disclosure focus on the specific form of epitaxial MgO. x Ge y O zThis paper provides unique insights into the use of materials such as MgO to serve as low lattice mismatch substrates for growing ZnO as single crystal structures.

[0025] 2A and 2B show perspective and plan views, respectively, of an example of a low formation energy crystalline structure of GeO2, according to some embodiments. The low formation energy corresponds to a compositionally stable structure. The structure in FIGS. 2A and 2B is trigonal GeO2, characterized by lattice constants a, b, and c, where a ≠ b ≠ c. As shown, Ge atoms ("Ge") are tetrahedrally bonded to oxygen atoms ("O").

[0026] 3A and 3B show perspective and plan views, respectively, of a second example of a low-formation energy crystalline structure of GeO2, according to some embodiments. This structure is tetragonal GeO2, characterized by lattice constants a0 and c0 (a0 ≠ c0). Again, Ge atoms are tetrahedrally bonded to O atoms. The low-energy forms of FIGS. 2A-2B and 3A-3B provide a stable structure. Other types of GeO2 crystalline structures exist, but they are less stable than those shown in FIGS. 2A-2B and 3A-3B.

[0027] In practice, large single-crystal structures of GeO2 are difficult to produce, and at best, polycrystalline or amorphous films can be formed. According to this disclosure, after thorough investigation of growth parameters, it has been found that crystalline GeO2 can be stabilized by adding other elements such as Mg, Zn, and even Al and Ga.

[0028] This disclosure focuses on the formation of single crystal structures using carefully selected substrates. x Ge y O z The present invention uniquely utilizes the relatively stable morphology of octahedral Mg-O and tetrahedral Ge-O bonds to enable the growth of layers. Embodiments also utilize the properties of the octahedral and tetrahedral bonding sites to bond Mg in various ways. x Ge y O zIt allows the material to be doped.

[0029] Mg x Ge y O z Many possible structures of Mg, Ge, and O have been investigated in connection with this disclosure. Figure 4 shows an exemplary table 400 of potential single crystal compositions that can be formed from Mg, Ge, and O, listed in order going from GeO at the top of the table to MgO at the bottom. In this table, the unit cell is x Ge 1-x O 2-x They are classified according to the element (x ranges from 0 to 1), where x = 0 represents binary GeO2 and x = 1 represents binary MgO. As can be seen, the most stable composition corresponds to x = 2 / 3, and the resulting composition is Mg2GeO4. A particular unit cell of a crystal structure contains a unique range of spatial positions for the Mg and Ge cations, as well as the oxygen anions. The space group and point group of a particular crystal are determined by the atomic arrangement within the unit cell. The general formula in Figure 4 is used to describe the various possible configurations.

[0030] From the possible compositions of Mg, Ge, and O shown in FIG. 4, the selection of possible single crystal compositions formed from Mg, Ge, and O was modeled, taking into account the stability of the compositions. Examples of single crystal compositions are shown in table 500 of FIG. 5, with structures in group 510 being the most stable of the structures investigated, and structures in group 520 being the next most stable. Table 500 indicates the corresponding space group and symmetry classification for each composition (or structure). Table 500 also indicates the bonding type for both Ge and Mg, where T d indicates a tetrahedral bond to an O atom, and O h indicates an octahedral bond to an O atom. Mg 14 GeO 24 In the case of , the Ge bond to the O atom is T d Bonds and O h The most stable bond is Mg x Ge 1-x O 2-xThe structures exhibit crystal symmetry groups of the x=2 / 3 cubic (Fd3m), x=2 / 3 orthorhombic (Pnma), and x=1 / 2 monoclinic (C2 / c) types, which were of particular interest.

[0031] Density functional theory (DFT) calculations using the Tran-Blaha modified Becke-Johnson (TBmBJ) exchange potential can predict the electron energy-momentum (Ek) diagram, band gap, lattice constant, and some mechanical properties of epitaxial oxide materials.

[0032] FIG. 6 illustrates in graphical form the structural characteristics of table 500. Plot 600 shows the structure of different single crystal compositions and space groups (i.e., Mg, Ge, and O) formed from Mg, Ge, and O. x Ge y O z The calculated relative energy difference (energy above hull) of the compound z is shown in electron volts (eV). Thermodynamically stable compounds have an energy difference E above the convex hull. hull = 0.0 eV / atom, with higher values ​​indicating increasing metastability. A value of 0 for the energy difference from the convex hull represents the most stable crystalline form, while increasing the energy difference from the convex hull (i.e., increasing the formation energy) progresses to less stable crystalline forms. The most stable compositions (i.e., low formation energy) begin on the left side of plot 600, and progressing to the right along plot 600, compositions with decreasing stability (i.e., higher formation energies) are shown. As shown, the single crystalline compositions Mg2GeO4 with Pnma space group (orthorhombic symmetry) at point 610 on plot 600, MgGeO3 with C2 / c space group (monoclinic symmetry) at point 620, and Mg2GeO4 with Fd3m space group (cubic symmetry) at point 630 are comparable in terms of stability, possessing thermodynamically stable structures with low formation energies. Plot 600 also demonstrates that changing the crystal structure can alter the crystal structure, even if the Mg x Ge y O zThis shows that even with the same stoichiometry, different stabilities can be obtained. For example, MgGeO3 with a crystal structure having an R3 space group (point 650) is less stable than MgGeO3 with a Pbca space group (point 640), and MgGeO3 with a Cmcm space group (point 660) is even less stable.

[0033] 7A, 7B, and 7C illustrate Mg single crystal semiconductor structures that may be formed in a single crystal semiconductor structure, according to some embodiments. x Ge y O z 7A and 7B are crystal lattice diagrams of examples of single crystalline compositions formed from Mg, Ge, and oxygen (O). Each diagram shows the base unit cell of a single crystalline composition formed from Mg, Ge, and oxygen (O). g and a unit cell 710 of Mg2GeO4 (Pnma space group) with a formation energy of -163.1986 eV per atom. g 7C is a unit cell 720 of MgGeO (C2 / c space group) with a band gap E of 5.8056 eV and a formation energy of -117.6576 eV per atom. g and a unit cell 730 of Mg2GeO4 (Fd3m space group) with a formation energy of -81.9710 eV per atom. Of the three forms shown, Mg2GeO4(Fd3m) is It has a bandgap that shows a potential direct bandgap in the UVC band (i.e., about 200-280 nm), an optimal formation energy, and a cubic symmetry that is expected to be easily grown by epitaxial methods. The combination of

[0034] 7A, 7B, and 7C. x Ge y O zThe bandgap characteristics of the Mg2GeO4(Fd3m) structure were calculated and are shown in Figures 9A-9B, 10A-10B, and 8A-8B, respectively. Figures 8A and 8B show bandgap modeling for the Mg2GeO4(Fd3m) structure of Figure 7C. Figure 8A is an Ek diagram 800 for Mg2GeO4(Fd3m), showing the calculated valence band 820 and conduction band 810 structures. Diagram 800 plots energy (eV) on the y-axis relative to the Brillouin zone (BZ) path along the x-axis. The direct bandgap between the conduction band 810 and valence band 820 is shown at the center of the Brillouin zone, with the bandgap energy Eg being 5.8056 eV. Figure 8A shows that Mg2GeO4(Fd3m) is a direct bandgap semiconductor material, with the maximum in the valence band 820 and the minimum in the conduction band 810 occurring at the center of the zone at k=0. Figure 8B is a close-up view 825 of the valence band structure. As shown in Figure 8B, the valence band structure in the zone center region 827 has a relatively large curvature and is parabolic, which indicates good hole mobility (e.g., 1 cm 2 / Vs or more than 10cm 2 / Vs or more than 100cm 2 As can be seen from Figure 8A, the Fermi energy (E f = 0) is located approximately in the middle of the gap between the valence band and the conduction band, indicating that Mg2GeO4(Fd3m) is a semiconductor with high resistivity (e.g., an intrinsic or lightly doped semiconductor).

[0035] Figure 9A shows the Ek diagram 900 for Mg2GeO4(Pnma) (see Figure 7A) with the calculated valence band 920 and conduction band 910 structures. Figure 9B shows a close-up 925 of the valence band structure. The direct band gap is shown at the Brillouin zone center, and the band gap energy E gis 6.4066 eV. As shown, Mg2GeO4(Pnma) is a direct bandgap semiconductor material, and the valence band maximum and conduction band minimum occur at the zone center where k=0. Figure 9B shows that the valence band structure in the zone center region 927 has a high curvature, thus indicating good hole mobility. As can be seen from Figure 9A, the Fermi energy (E f ) is located approximately halfway between the gap between the valence band and the conduction band, indicating that Mg2GeO4(Pnma) is a semiconductor with high resistivity (e.g., an intrinsic or lightly doped semiconductor).

[0036] FIG. 10A shows the Ek diagram 1000 for MgGeO3(C2 / c) (see FIG. 7B), with the calculated valence band 1020 and conduction band 1010 structures. FIG. 10B shows a close-up 1025 of the valence band structure. As shown, MgGeO3(C2 / c) is an indirect bandgap semiconductor material because the valence band maximum and conduction band minimum occur at different k values, and therefore is expected to be a poor optical emitter. That is, in some embodiments, Mg x Ge 1-x O 2-x is MgGeO3 with a C2 / c space group, which is an indirect band gap material. The indirect band gap is observed at a different k vector between the conduction band minimum and the valence band maximum, an example of which is shown by the arrow 1005. As shown in Figure 10B, the valence band structure in the zone center region 1027 has a curvature, indicating good electron-hole mobility. As can be seen from Figure 10A, the Fermi energy (E f ) is located approximately halfway between the gap between the valence band and the conduction band, indicating that MgGeO3(C2 / c) is a semiconductor with high resistivity (e.g., an intrinsic or lightly doped semiconductor).

[0037] Referring now to Figure 11, a local tetrahedral bonding configuration 1100 in an Fd3m cubic crystal of the AB2O4 type is shown, where the A and B cations are tetrahedrally and octahedrally bonded to oxygen (O) atoms, respectively. The critical bond length d between the A and O atoms isA-O is annotated in Figure 11. Equation 1 is x Ge y O z In order to study the stability and cation exchangeability of possible configurations of the structure, they are presented according to the present disclosure. In the case of an ideal Fd3m unit cell, the tetrahedrally coordinated cations are located at the special position 8a (Wyckoff position).

[0038]

number

[0039] The octahedrally coordinated cation occupies a special position 16d (

[0040]

number

[0041] The anions are located at the isoponent positions 32e, which requires a single position parameter, u. When u = 0.25, ideal cubic closed packing of the anions is achieved, so variation of u reflects adjustments of the structure to accommodate cations of different sizes at the octahedral and tetrahedral positions. For example, decreasing u below 0.25 moves the anions along the

[0111] direction toward the nearest tetrahedral cation, thereby reducing the size of the tetrahedron within the range of the octahedron size.

[0042] Bond length between tetrahedral Ge cations and O anions

[0043]

number

[0044] is given by where a0 is the lattice constant of the cubic unit cell. For the bulk free-standing Fd3mMg2GeO4 crystal structure,

[0045]

number

[0046] therefore

[0047]

number

[0048] It turns out to be. For example, if the B atom has a valence coordination number of +2, such as magnesium (i.e., cation site B = Mg), and the A atom has a valence coordination number of +4, such as germanium (i.e., cation site A = Ge), the cubic crystalline composition AB2O4 = GeMg2O4 (i.e., Mg2GeO4) is formed. x Ge y O z A consideration when studying the growth of the structure is whether Mg2GeO4 adopts a conventional spinel structure or a less stable spinel structure in which the cations on the tetrahedral and octahedral sites are interchanged to some extent: the conventional spinel structure is expected to self-assemble with only Mg cations located on the octahedral sites and Ge cations located on the tetrahedral sites within the crystal, allowing for codeposition, which is generally desirable for epitaxial growth.

[0049] Distortion of the Fd3m unit cell can also alter the oxygen u parameter, for example, via biaxial strain during pseudomorphic epitaxial deposition on a lattice-mismatched surface. Tetragonal distortion clearly influences the structural and electronic properties of Mg2GeO4.

[0050] Using these crystal structure considerations, various forms of Mg x Ge y O zIn Figure 12, table 1200 shows examples of compatible substrate compositions and substrate orientations on which single-crystal compositions 1210 of (001)- or (100)-oriented Mg2GeO4 (Pnma), MgGeO3 (C2 / c), and Mg2GeO4 (Fd3m) can be epitaxially formed. Currently, there are no examples of bulk, large-area, single-crystal native Mg2GeO4 substrates in the industry, and therefore heteroepitaxial methods are needed. The space group "SG" and crystal symmetry type are listed for each composition. Possible substrates 1230 for composition 1210 are disclosed, and it should be noted that the seeded substrate may not be ideally lattice-matched (i.e., non-natively lattice-matched) to the epitaxially formed MgGeO composition. As indicated by unit cell construction parameters 1220, including lattice constants a, b, c, and angles α, β, and γ, the Mg x Ge y O z The crystal structure of each unit cell of Mg is also disclosed. x Ge y O z The magnitude of the lattice mismatch between the film and the substrate is preferably selected to allow growth of a single crystal epitaxial film with a low defect density.

[0051] In some embodiments, the crystal symmetry of the substrate material may be cubic or tetrahedral. In some embodiments, the substrate material comprises MgO(001), MgGaO(001), MgAlO(001), or LiF(100). In some embodiments, the substrate material comprises β-GaO(100), LiAlO(100), ZrO(100), LiNbO(001), LiTaO(001), FeO(100), BN(001), LiGaO(001), TiO(001), AlN(100), SiC(100), BaF(100), BN(100), or CdWO(001).

[0052] Mg on the substrate x Ge y O zSchematic diagrams of the single crystal structure of the film (i.e., epitaxial layer) are shown in Figures 13A-13B and 14A-14B, and the substrate material is Mg x Ge y O z The selected (i.e., predetermined) type is selected based on its compatibility with the type of

[0053] 13A is a schematic diagram showing two different free-standing crystalline structures corresponding to a film 1310 and a substrate 1300, both with cubic or tetrahedral symmetry. The growth direction Z is indicated. The film 1310 has a larger in-plane lattice parameter (also called the parallel lattice parameter) than the substrate. That is,

[0054]

number

[0055] A free-standing cubic crystal has equal in-plane and perpendicular lattice constants.

[0056]

number

[0057] where i = {substrate, film}. That is, before the formation of the epitaxial layer (i.e., when the film material is free-standing and not grown on a substrate), due to the cubic or tetrahedral symmetry of the film material,

[0058]

number

[0059] , the parallel lattice constant and perpendicular lattice constant of the film are equal to each other.

[0060]

number

[0061] Similarly, the parallel and perpendicular lattice constants of the substrate are equal due to the cubic or tetrahedral symmetry of the substrate.

[0062]

number

[0063] Figure 13B is a schematic diagram showing the film 1310 of Figure 13A epitaxially grown on a substrate 1300. Figure 13B illustrates the effect of epitaxially forming a layer with cubic symmetry on a substrate that also has cubic symmetry, where the free-standing in-plane lattice constant of the film is larger than the in-plane lattice constant of the substrate. Upon forming the epitaxial layer 1310', the film material elastically deforms and becomes strained compared to its free-standing (or relaxed) state (film 1310), thereby increasing the parallel or in-plane lattice constant of the epitaxial layer.

[0064]

number

[0065] will match the parallel or in-plane lattice constant of the substrate 1300 (i.e.,

[0066]

number

[0067] The resulting epitaxial layer 1310' is formed in a compressed state. Due to the Poisson effect, this deformation causes the vertical lattice constant of the resulting epitaxial layer 1310' to

[0068]

number

[0069] increases and becomes larger than the normal lattice constant of the original membrane material according to the Poisson's ratio of the material (i.e.,

[0070]

number

[0071] ). Therefore, high-quality epitaxial layers can be formed by elastically deforming the unit cell of the film during growth to accommodate the in-plane lattice mismatch, which is defined herein as pseudomorphic epitaxy. Figures 13A-13B illustrate that semiconductor structure designs in which both the epitaxial layer and the substrate have cubic and tetrahedral symmetries can be advantageously used in accordance with embodiments of the present disclosure. In some embodiments, the substrate and film have equal in-plane lattice constants.

[0072]

number

[0073] , having cubic or tetrahedral crystal symmetry such that the in-plane lattice constant of the film is larger than the in-plane lattice constant of the substrate.

[0074]

number

[0075] In such cases, growing an epitaxial film on a substrate increases its vertical lattice constant, resulting in the film generally accumulating in-plane compressive stress. Moreover, in some cases, compression has been observed to facilitate the growth of epitaxial layers.

[0076] 14A-14B are similar to FIGS. 13A-13B, except that the epitaxial layers are under tension. FIG. 14A is a schematic diagram showing two different free-standing crystalline structures corresponding to film 1410 and substrate 1400, both of which have cubic or tetrahedral symmetry. Film 1410 has a smaller in-plane (i.e., parallel) lattice constant than substrate 1400. That is,

[0077]

number

[0078] Before the epitaxial layer is formed, the parallel lattice constant and perpendicular lattice constant of the film are equal.

[0079]

number

[0080] Similarly, due to the cubic or tetrahedral symmetry of the material, the parallel and perpendicular lattice constants of the substrate are equal.

[0081]

number

[0082] . FIG. 14B shows that when a layer with cubic (or tetrahedral) symmetry is epitaxially formed on a substrate also with cubic symmetry, the free-standing in-plane lattice constant of the original film 1410 is smaller than the in-plane lattice constant of the substrate 1400 (i.e., as shown in FIG. 14A).

[0083]

number

[0084] 14 is a schematic diagram illustrating the effect of forming epitaxial layer 1410', where the film material elastically deforms to change the in-plane lattice constant

[0085]

number

[0086] becomes consistent with the in-plane lattice constant of the substrate (i.e.,

[0087]

number

[0088] Due to the Poisson effect, the vertical lattice constant of the resulting epitaxial layer is smaller than that of the original film material (i.e.,

[0089]

number

[0090] By matching the in-plane lattice constant of the film with that of the substrate, the tetragonal unit cell of the film is elastically distorted, resulting in a decrease in the perpendicular lattice constant of the film (i.e.,

[0091]

number

[0092] The resulting epitaxial layer is formed under tension, with the layer generally storing in-plane tensile stress.

[0093] 15A-15B show crystal structure diagrams of a single cubic Fd3mMg2GeO4 crystal 1510 formed on a substrate 1500. In FIG. 15A, the unit cell of the cubic Fd3mMg2GeO4 crystal 1510 elastically deforms equally along the in-plane a- and b-axes, conforming to twice the lattice constant of a freestanding cubic MgO crystal. That is, the Mg2GeO4 crystal 1510 has tensile biaxial strain conforming to MgO(001). This tensile elastic deformation transforms the crystal symmetry of the cubic Fd3m space group into a new tetragonal I4amd space group. The lattice constants a, b, and c in angstroms (Å) are shown in FIG. 15A, along with the angles α, β, and γ.

[0094] FIG. 15B is a representation of the lattice-mismatched formation of an epitaxial layer 1511 of Mg2GeO4(Fd3m) crystals grown on a cubic MgO(001) substrate 1500, showing that the in-plane lattice constant of the film closely matches an integer multiple of the smaller MgO unit cell. In this example, the epitaxial layer 1511 will have tetragonal symmetry with space group I4amd. In this example, the formation of the epitaxial layer is

[0095]

number

[0096] , resulting in a two-fold unit cell match, resulting in a Mg2GeO4(Fd3m) epitaxial layer 1511 formed under tension on the MgO substrate 1500.

[0097] Figures 16 and 17 show the effect of strain on the bandgap for epitaxial layers of cubic Mg2GeO4 (Fd3m) on different cubic substrate materials. Figure 16 is an electronic band structure Ek diagram 1600 for an epitaxial layer of cubic space group Fd3mMg2GeO4 formed on a substrate or rigid layer of cubic Fd3m space group magnesium aluminate MgAl2O4. In this example, the lattice constant mismatch between Mg2GeO4 and MgAl2O4 is approximately +2.2%, resulting in compressive in-plane stress and causing Mg2GeO4 to form in compression. Epitaxial layers of the cubic space group Fd3mMg2GeO4 still have a direct bandgap, with the minimum in the conduction band 1610 and the maximum in the valence band 1620 occurring at the center of the Brillouin zone at k = 0, just as in the case of unstrained Mg2GeO4Fd3m in Figure 8A. However, the bandgap E of compressively deformed or strained Mg2GeO4 layers at the zone center in Figure 1600 g is E in Figure 8A. g It increases to 6.004 eV in contrast to 5.8056 eV.

[0098] Figure 17 shows the electronic band structure Ek diagram 1700 for an epitaxial layer of cubic space group Fd3mMg2GeO4 formed on a substrate or rigid layer of cubic Fd3m space group magnesium oxide MgO. In this example, the lattice constant mismatch between Mg2GeO4 and MgO is approximately -1.9%, resulting in in-plane tensile stress, i.e., Mg2GeO4 is formed in tension. The strained Mg2GeO4 band gap E between the conduction band 1710 and the valence band 1720 at the Brillouin zone center g It is observed that the .DELTA..times ...

[0099] As can be seen from this disclosure, Mg x Ge y O z The Mg epitaxial layer can be formed as an epitaxial single crystal structure with the substrate. x Ge y O z Compatible (i.e., low lattice mismatch) combinations of crystal structures and substrate materials are disclosed. In embodiments, the epitaxial oxide material and substrate material are selected so that the epitaxial layer of the semiconductor structure has an in-plane lattice constant (or crystal plane spacing) within 0.5%, 1%, 1.5%, 2%, 5%, or 10% of the in-plane (i.e., parallel to the surface of the substrate) lattice constant (or crystal plane spacing) of the substrate.

[0100] doping In one embodiment, Mg x Ge y O z doping, which allows the material to be used as an epitaxial layer in various types of electronic and optoelectronic devices, such as an n-type or p-type layer. x Ge y O z The material may be a direct bandgap material or an indirect bandgap material. In some embodiments, the direct bandgap material Mg2GeO4(Fd3m) is doped. In the crystal structure of Mg2GeO4(Fd3m), the Ge sites are tetrahedronized with oxygen (T d) bond, and the Mg site has an octahedral (O h ) bonds. The following are some of the embodiments of the present disclosure. Ga doping is Ge Td sites (e.g., every single Ge Td Site) is Ga Td Calculations showing that substitution at the Ge site results in direct bandgap p-type material Td The site can potentially be replaced. Ga doping, Mg Oh site (e.g., every single Mg Oh Site) is Ga Oh Calculations show that substitution at the Mg site results in a direct bandgap n-type material. Oh The site can potentially be replaced. Calculations show that Al doping leads to a direct bandgap n-type material. Oh The site can potentially be replaced. Calculations show that Al doping leads to a direct bandgap p-type material. Td The site can potentially be replaced. Li doping is a direct bandgap p-type material, as calculated Td or Mg Oh The site can potentially be replaced. Ni doping is a key factor in the formation of defect bands and the reduction of the band gap. Oh The site can potentially be replaced. · N doping can potentially substitute for the O site, with calculations indicating a direct bandgap p-type material. Antisite substitutions can be made within the unit cell, and calculations show p-type materials. Antisite substitutions can involve the exchange of Mg and Ge atoms, resulting in a p-type material. Td Atoms are arranged in octahedral bonding sites (Ge Td →Ge Oh ), Mg Oh Atoms are arranged in tetrahedral bonding sites (Mg Oh→Mg Td ). Doping can be done using either excess Mg or excess Ge.

[0101] Although embodiments are described using Mg2GeO4 as an example, embodiments may also be used with other forms of Mg x Ge 1-x O 2-x It can also be applied to Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg is to codeposit the material on the substrate. x Ge 1-x O 2-x where the value of x is 0≦x<1. The material includes at least two elements selected from Mg, Ge, and oxygen according to the value of x, and Mg, Ge, and oxygen are supplied by a Mg source, a Ge source, and an activated oxygen source, respectively. The method includes forming an epitaxial layer of Mg on a substrate. x Ge 1-x O 2-x The method may include depositing a buffer layer between the epitaxial layer of the first epitaxial layer and the second epitaxial layer of the second epitaxial layer. The codeposition may be performed using a molecular beam epitaxy process.

[0102] 18A and 18B, there are shown tables of the normalized x, y, and z positions in Cartesian coordinates of the cation (i.e., Mg and Ge) sites (FIG. 18A) and anion (i.e., O) sites (FIG. 18B) of the Mg2GeO4(Fd3m) unit cell referred to throughout this disclosure. In this example, the cation unit cell has 16 different Mg sites (Mg i , i = 0 to 15) and eight different Ge sites (Ge j , j = 16 to 23). The anion unit cell contains 32 different sites (O k , k=24-55). Accordingly, Figures 18A-18B disclose the crystal structure of Mg2GeO4(Fd3m), according to an embodiment of the present disclosure.

[0103] In some embodiments, as illustrated in Figures 19-21B, Ge Td sites with Ga atoms (as a result, Ga Td 18A), direct bandgap p-type materials can be created. Figure 19 is a table 1900 summarizing the semiconductor properties of single-site trivalent gallium impurity atom substitutions in Mg2GeO4 (Fd3m) resulting from various single-site substitutions of Ga at the Ge site. Table 1900 includes the Ge i The specific Ge site 1910 that is substituted, corresponding to the cation site, is shown. For example, in the first row of table 1900, a Ge site "Ge16" is substituted with a Ga atom, resulting in a Ga site "Ga16." Table 1900 also includes normalized x, y, and z coordinates, band gap classification 1920 and band gap energy value 1925, and semiconductor type 1930 (E fermi ) and formation energy 1940 (E Formation Calculated semiconductor properties such as ρ, ρ, and ρ are also shown. As shown, for the various Ge sites shown, the resulting doped Mg2GeO4(Fd3m) is a direct bandgap p-type material. x Ge 1-x O 2-x The epitaxial layers of the semiconductor structures and devices are direct bandgap p-type materials and include Ga dopants (e.g., Ga atoms). The Ga dopants are not present in the corresponding undoped (i.e., the corresponding undoped structures are nominally undoped Mg x Ge 1-x O 2-x Mg x Ge 1-x O 2-x Located at the Ge site in the crystal structure. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O 2-x This involves substituting Ga for Ge sites in the crystal structure to provide p-type conductivity.

[0104] Figure 20 shows the crystal structure of doped Mg2GeO4(Fd3m), in this example showing the Ga-substituted Ge 23 21A and 21B show the Ek diagrams for Mg2GeO4(Fd3m) resulting from the substitution of Ga for the Ge sites shown in FIG. 20. Diagram 2100 of FIG. 21A shows the structure of the valence band 2120 and conduction band 2110, and diagram 2125 of FIG. 21B shows a zoomed-in view of the valence band structure. As can be seen, doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor material, with the maximum of the valence band 2120 and the minimum of the conduction band 2110 occurring at the center of the k=0 zone. As best shown in FIG. 21B, the valence band maximum 2127 is located at the Fermi energy (E f ), indicating that the doped Mg2GeO4(Fd3m) in this example is a p-type material. In summary, the degenerate Fermi energy E f is located within the valence band, and the direct band gap energy is reduced by about 140 meV, compared to E of 5.8056 eV for the undoped case in Figure 8A. g In comparison, this Ge-site substitutional form has an Eg of 5.6650 eV. The doped Mg2GeO4(Fd3m) still has an ultra-wide bandgap of over 4 eV.

[0105] In some embodiments, as illustrated in FIGS. 22-24, Mg Oh sites with Ga (resulting in Ga Oh2200 discloses the calculated semiconductor properties of Ga-doped Mg2GeO4(Fd3m) resulting from various single-site substitutions of Ga at distinct Mg sites. Table 2200 shows the substituted Mg sites 2210 and their normalized x, y, and z coordinates. Examples of semiconductor properties, such as bandgap classification 2220 and bandgap energy 2225, semiconductor type 2230, and formation energy 2240, are shown for selected substitution cases. The semiconductor properties shown have very similar values ​​to each other. Other substitution embodiments shown in table 2200 are expected to show similar results. As shown, for the various Mg sites shown, the resulting doped Mg2GeO4(Fd3m) is a direct bandgap n-type material. That is, single-site substitution of a trivalent cation at a divalent Mg site results in n-type conductivity. The embodiment shows the Mg x Ge 1-x O 2-x The epitaxial layer of the present invention is a direct bandgap n-type material and includes a Ga dopant (e.g., Ga atoms). The Ga dopant is doped with a corresponding undoped (i.e., undoped compared to nominal) Mg x Ge 1-x O 2-x Located at the Mg site in the crystal structure. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O 2-x It involves substituting Ga for Mg sites in the crystal structure to provide n-type conductivity.

[0106] FIG. 23 shows a crystal structure diagram 2350 of doped Mg2GeO4(Fd3m), in this example where the Mg1 site is substituted with Ga. Also shown in FIG. 23 is the associated Ek diagram 2300 for an n-type doped semiconductor material. As shown, the resulting doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor material. The minimum 2315 of the conduction band 2310 is located at the Fermi energy (E f ), indicating that the doped Mg2GeO4(Fd3m) in this example is an n-type material.

[0107] Figure 24 is similar to Figure 23, except that the Mg2 site has been substituted with Ga. Figure 24 includes a crystal structure diagram 2450 of doped Mg2GeO4(Fd3m) and an associated Ek diagram 2400 for the n-type doped semiconductor material. As shown, the resulting doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor material, with the minimum 2415 of the conduction band 2410 at the Fermi energy (E f ), indicating that the doped Mg2GeO4(Fd3m) in this example is an n-type material.

[0108] In some embodiments, Mg is used to produce a direct bandgap n-type material, for example by co-deposition with an additional elemental aluminum source. Oh sites can be substituted with Al (see, for example, Figure 42B). Figure 25 is an Ek diagram 2500 illustrating the structure of the valence band 2520 and conduction band 2510 of doped Mg2GeO4(Fd3m) resulting from the substitution of Al at the octahedral Mg1 sites. As shown, the resulting doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor material. The minimum 2515 of the conduction band 2510 is located at the Fermi energy (E f ), indicating that the doped Mg2GeO4(Fd3m) in this example is an n-type material. That is, the Fermi energy is located within the conduction band 2150, and is expected to exhibit n-type conductivity. x Ge 1-x O2-x The epitaxial layer of the Mg epitaxial layer is a direct bandgap n-type material and includes an Al dopant (e.g., Al atoms). The Al dopant is a dopant that is present in a corresponding undoped Mg x Ge 1-x O 2-x Located at the Mg site in the crystal structure. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O 2-x It involves substituting Al for Mg sites in the crystal structure to provide n-type conductivity.

[0109] In some embodiments, Ge Td Al can be substituted at the SiO2 site to create a direct bandgap p-type material. Figure 26 shows the Al-substituted tetrahedral Ge 17 FIG. 26 is an Ek diagram 2600 illustrating the structure of the valence band 2620 and conduction band 2610 of doped Mg2GeO4(Fd3m) resulting from the ZnSe site. As shown, the resulting doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor material. The maximum 2625 in the valence band 2620 is located at the Fermi energy (E f ), indicating that the doped Mg2GeO4(Fd3m) in this example is a p-type material. That is, the Fermi energy is located within the valence band 2620, and is expected to exhibit p-type conductivity. x Ge 1-x O 2-x The epitaxial layer of the present invention is a direct bandgap p-type material and includes an Al dopant (e.g., Al atoms). The Al dopant is a dopant that is present in a corresponding undoped Mg x Ge 1-x O 2-x Located at the Ge site in the crystal structure. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O2-x This involves substituting Al for Ge sites in the crystal structure to provide p-type conductivity.

[0110] In some embodiments, as illustrated in FIGS. 27-30, Ge Td or Mg Oh The sites can be substituted with Li doping to create direct bandgap p-type materials. Figure 27 shows a monovalent Li tetrahedrally bonded to oxygen. + The unit cell crystal structure diagram of cubic lithium oxide, Li2O(Fm3m), showing the atoms. Li2O is a cubic lithium oxide in which the Li atoms are bonded tetrahedrally (T d ) is a stable oxide.

[0111] Figure 28 shows the unit cell crystal structure diagram of doped Mg2GeO4(Fd3m), in this example with Li as the impurity atom. + Tetrahedral Ge substituted with 17 Indicates the site. + and Ge 4+ Both oxides of Li and Li2 prefer tetrahedral bonding with oxygen. + is expected to preferentially substitute for the Ge site compared to the octahedral site.

[0112] FIG. 29A is an Ek diagram 2900 for Li-doped Mg2GeO4(Fd3m) resulting from Li substitution at the Ge sites shown in FIG. 28. The valence band 2920 and conduction band 2910 structures are shown. FIG. 29B is a close-up 2925 of the valence band structure 2920 of FIG. 29A. As shown, doped Mg2GeO4(Fd3m) is an indirect bandgap semiconductor with a bandgap energy of 5.5563 eV between the conduction band minimum 2915 and the valence band maximum 2927. As best shown in FIG. 29B, the doped Mg2GeO4(Fd3m) in this example is a degenerate p-type material. That is, the Fermi energy

[0113]

number

[0114] is located within the valence band 2920 (diagram 2925), resulting in p-type conductivity, and the valence band maximum 2927 is no longer at the Brillouin zone center.

[0115] FIG. 30 is an Ek diagram 3000 corresponding to doped Mg2GeO4(Fd3m), where the Mg1 site is substituted with Li. The conduction band 3010 and valence band 3020 structures are shown. FIG. 30 shows that Li doping at the Mg1 site results in an indirect bandgap semiconductor. Looking at FIGS. 29A and 30, both show that Ge 17 The effect of substituting Li for one or more Mg atoms produces a p-type material, while Ge substitution results in a degenerate p-type material. x Ge 1-x O 2-x The epitaxial layer is a direct bandgap p-type material, and Li + Dopants (e.g., Li + This includes semiconductor structures and devices containing Li + The dopants are the corresponding undoped Mg x Ge 1-x O 2-x Located at the Ge or Mg site in the crystal structure. x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O 2-x Li occupies the Ge or Mg site in the crystal structure. + to provide p-type conductivity.

[0116] In some embodiments, as shown in Figures 31-33B, Mg Oh Ni atoms can be substituted at the ZnO site to create defect bands and reduce the band gap compared to the undoped form. Figure 31 shows the crystal structure diagram of a cubic NiO (space group Fd3m) unit cell. NiO is a structure in which Ni atoms are octahedral bonded with oxygen atoms (Ni Oh ) is a stable oxide.Oh Atoms are chosen instead of Mg.

[0117] Figure 32 shows the unit cell crystal structure diagram of Ni-doped Mg2GeO4(Fd3m), in this example, Ni as the impurity atom. + The octahedral Mg1 site is shown substituted with

[0118] Figure 33A is an Ek diagram 3300 for Ni-doped Mg2GeO4(Fd3m) shown in Figure 32. In Figure 32, the structure of the valence band 3320 and conduction band 3310 is shown. As shown, the resulting doped material has a reduced bandgap energy E of approximately 2 eV. g It has an indirect band gap with . Furthermore, in addition to the normal valence band 3320, several defect bands 3328 and 3329 exist, and parasitic absorption also exists in the gap between the defect bands. Substitution of Ni atoms at Mg sites creates deep levels in the band gap due to Ni p and d orbitals. These calculation results show that Ni does not significantly change the conductivity type of Mg2GeO4, but still introduces parasitic optical absorption in the band gap.

[0119] FIG. 33B is a density of states (DOS) diagram for Ni-doped Mg2GeO4(Fd3m) shown in FIG. 32. The DOS shows unfavorable sub-bandgap energy states due to Ni p- and d-orbitals. x Ge 1-x O 2-x The epitaxial layer of Ni + Dopants (e.g., Ni + This includes semiconductor structures and devices containing Ni. + The dopants are the corresponding undoped Mg x Ge 1-x O 2-x Located at the Mg site in the crystal structure. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O 2-xThe Mg site in the crystal structure is occupied by Ni + This includes replacing

[0120] In some embodiments, as illustrated in Figures 34-35B, O sites can be doped with N atoms (nitrogen doping, N 3+ ) to create a direct bandgap p-type material. Figure 34 shows a unit cell crystal structure diagram of N atoms substituted into oxygen sites in Mg2GeO4(Fd3m). In this example, the anion O 45 The site is substituted with trivalent nitrogen as an impurity atom,

[0121]

number

[0122] It is displayed as follows. Figure 35A is an Ek diagram 3500 for N-doped Mg2GeO4(Fd3m) resulting from the N-site substitution shown in Figure 34. The structure of the valence band 3520 and conduction band 3510 is shown. Figure 35B shows the Ek diagram 3500 for N-doped Mg2GeO4(Fd3m) resulting from the N-site substitution shown in Figure 34. A As can be seen from FIG. 35A, N-doped Mg2GeO4(Fd3m) has a valence band structure similar to that of the undoped E g As best shown in Figure 35B, in N-doped Mg2GeO4 (Fd3m), the Fermi energy due to nitrogen impurities is

[0123]

number

[0124] A relatively shallow energy band 3528 near the Mg x Ge 1-x O 2-x The epitaxial layer is a direct bandgap p-type material, and N 3+ Dopants (e.g., N 3+N 3+ The dopants are the corresponding undoped Mg x Ge 1-x O 2-x Located at the oxygen site in the crystal structure. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of Mg x Ge 1-x O 2-x Oxygen sites in the crystal structure are filled with N 3+ This includes replacing

[0125] In some embodiments, atoms within a unit cell can be substituted for one another to produce p-type materials. This substitution technique is referred to in this disclosure as antisite substitution or antisite doping. For example, Mg and Ge are substituted with atoms of other types (Ge and Mg, respectively) at their respective tetrahedral and octahedral bonding sites. Antisite substitution is illustrated in Figures 36A-37F.

[0126] Figure 36A shows the stoichiometric unit cell crystal structure of cubic Mg2GeO4(Fd3m), identifying the normal structural locations of selected Mg and Ge atomic sites used in the cation antisite doping process. In this example, tetrahedrally bonded Ge 20 Location

[0127]

number

[0128] and octahedrally bonded Mg 12 Location

[0129]

number

[0130] The atom is replaced by another atom. FIG. 36B illustrates an embodiment of antisite mutual replacement, in which the exchange process involves substitution by Ge atoms during the formation of the crystal structure.

[0131]

number

[0132] Nominal Mg 12 sites and are replaced by Mg atoms during the formation of the crystal structure.

[0133]

number

[0134] Nominal Ge 20 Note that the total number of Mg and Ge cations in the unit cell does not change, so the total number of Mg and Ge atoms in the unit cell is maintained. As a result of the antisite substitution, the original Fd3m (cubic) space group crystal becomes a Cm (monoclinic) space group.

[0135] Figure 37A shows the resulting Ek diagram 3700 for antisite exchange in Mg2GeO4 (Fd3m) shown in Figure 36B. Figure 37A shows the structure of the valence band 3720 and conduction band 3710. In this example, the selective Mg⇔Ge antisite inter-substitution changes the symmetry of the new unit cell space group from the normal cubic symmetry (i.e., without antisite exchange) to a monoclinic structure, resulting in the non-ideal nature of the Ek band structure. Antisite exchange is energetically unfavorable compared to simply exchanging atomic positions, moving one Mg atom to a local tetrahedral site and one Ge atom to a local octahedral site. The band structure is reduced in magnitude compared to the most stable normal Fd3m configuration. Intermediate bands (e.g., defect bands 3728, 3729) created deep in the gap are located at the Fermi level (E f) is located between the lowest energy s orbital and the highest energy p orbital across the k vector in the Brillouin zone, indicating that the electronic structure is self-compensated. Figure 37B shows the electronic density of states for the Mg⇔Ge antisite exchange example in Figure 36B.

[0136] Figure 37C shows the distorted crystal unit cell 3740 resulting from the geometry optimization by an energy minimization process of a single Mg⇔Ge antisite atom exchange in the Mg2GeO4 crystal structure shown in Figure 36B. An example of the local environment 3750 of a Ge atom is highlighted.

[0137] Figure 37D is a schematic diagram of the cubic bonding local environment 3750 of a Ge atom (tetrahedrally bonded) in a nominal Fd3m crystal and the distorted crystal field 3751 that results when a Ge atom is substituted into an octahedral bonding site. The lattice distortion minimizes the resulting unit cell formation energy.

[0138] Shown in Figure 37E are tables 3760 and 3770 showing the positions of the cations and anions, respectively, in the distorted crystal structure for a single antisite substitution. In this example of Figure 37E, Mg 10 The site is Ge (Ge 10 ) and Ge 23 The site is Mg (Mg 23 ) is replaced.

[0139] Figure 37F shows Mg 11 The site is Ge (Ge 11 ) and Ge 24 The site is Mg (Mg 24 ) is the E band structure for a geometrically optimized single antisite substitution. The resulting band gap E g is indirect and is smaller than that for the nominal stoichiometric Fd3m. x Ge 1-x O 2-xThe epitaxial layer of SiO2 has a doped unit cell structure and a corresponding undoped unit cell structure (i.e., the corresponding undoped structure is the same as that of undoped nominal Mg x Ge 1-x O 2-x The doped unit cell structure includes semiconductor structures and devices. x Ge 1-x O 2-x is an indirect bandgap p-type material. Here, the doped unit cell structure contains a Ge atom in the first position occupied by Mg in the corresponding undoped unit cell structure. The doped unit cell structure contains a Mg atom in the second position occupied by Ge in the corresponding undoped unit cell structure. In the doped unit cell structure, the Ge atoms are octahedrally bonded and the Mg atoms in the second position are tetrahedrally bonded. Mg x Ge 1-x O 2-x The method of doping the epitaxial layer of x Ge 1-x O 2-x and disposing a Mg atom in a second position occupied by Ge in the corresponding undoped unit cell structure.

[0140] In some embodiments, either excess Mg atoms or excess Ge atoms can be utilized to create doped materials, as illustrated by Figures 38-41. x Ge 1-x O 2-x The present invention also includes semiconductor structures and devices in which the epitaxial layer of Mg is doped with excess Ge atoms or excess Mg atoms. x Ge 1-x O 2-xMethods for doping epitaxial layers include doping with excess Ge atoms or excess Mg atoms. Figure 38 shows a unit cell crystal structure diagram of the cation lattice sites of space group Fd3m converted from the normal configuration to the germanium-rich configuration. In this example, the excess Ge atoms are created by, for example, converting the Mg4 site to an additional Ge atom. The resulting composition exhibits a new space group symmetry of R3m, and the resulting composition corresponds to:

[0141]

number

[0142] FIG. 39A shows Mg crystals with respective crystal space groups 3920 according to some embodiments. x Ge y O z 39 is a table 3900 of possible compositions (formula 3910) of MgO. The number of Mg atoms (N_Mg), Ge atoms (N_Ge), and oxygen atoms (N_O) in each unit cell is listed, along with the relative number of cations and anions 3940 in the crystal unit cell. The ratio 3930 of Mg:Ge atoms in each unit cell is calculated for a given space group. In the example of the nominal Mg2GeO4Fd3m crystal space group, the cation to anion ratio Mg:Ge is 2.0 (indicated by arrow 3932). With an excess of Ge atoms (9 atoms instead of the nominal 8 atoms), the Mg:Ge ratio is 1.7 (arrow 3934), and with an excess of Mg atoms (17 atoms instead of the nominal 16 atoms), the ratio Mg:Ge is 2.4 (arrow 3936).

[0143] FIG. 39B shows the exemplary Mg x Ge y O zTable 3901 shows the formation energies 3950 and lowest energy band gaps 3960 of the compounds. In the case of Fd3mMg2GeO4 crystals, when modified with an Mg excess, the composition exhibits P-43m symmetry (indicated by arrow 3926), while with a Ge excess, the composition exhibits R-3m symmetry (indicated by arrow 3924). Either the Mg-rich or Ge-rich structure has an inherently higher energy of formation and is less energetically favorable, meaning the doped forms are less stable.

[0144] In Figure 39C, table 3902a shows the atomic positions in a regular Fd3m crystal, and table 3902b shows an Fd3m crystal formed with an excess of Ge. In this embodiment, the excess Ge atoms are inserted by substituting Ge for Mg sites (Mg4 in this example). The total number of cations (Mg and Ge) is maintained, but the number of Ge atoms per unit cell is increased.

[0145] Figure 40 is an Ek diagram 4000 for Ge-doped Mg2GeO4(Fd3m), where excess Ge is present. As shown, the doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor with a bandgap energy of 4.3942 eV at the zone center, which is smaller than the undoped Mg2GeO4(Fd3m) in Figure 8A. The Fermi energy (E f ) is located deep within the conduction band 4010 and therefore exhibits n-type conductivity.

[0146] Figure 41 is an Ek diagram 4100 for Mg-doped Mg2GeO4(Fd3m), where excess Mg is present. As shown, the doped Mg2GeO4(Fd3m) is a direct bandgap semiconductor with a bandgap energy of 5.6405 eV, which is smaller than the undoped Mg2GeO4(Fd3m) of Figure 8A. The Fermi energy (E f ) is located near the edge of the valence band 4120 and therefore exhibits p-type conductivity.

[0147] FIG. 42A is a diagram summarizing some of the embodiments of the doping strategies described with respect to FIGS. 19-41. The diagram illustrates the resulting type (i.e., n-type or p-type) of doped Mg2GeO4(Fd3m). Embodiments also include other forms of Mg x Ge 1-x O 2-x The Fermi energy is plotted on the y-axis relative to the undoped intrinsic bands, labeled as the conduction band 4202 and the valence band 4204. As shown, the previously very difficult possibility of p-type conductivity is made possible by substitutional doping of the Ge site with Ga and Li, and substitutional doping of the Mg site with Ge (i.e., Mg excess). Lithium is distinguished by its ability to create p-type characteristic materials for substitution at either or both the Ge and Mg sites. Substitution of Ga results in p-type or n-type doping, depending on whether Ga is substituted at the Ge or Mg site, respectively. Control of the location of Ga insertion can be achieved, for example, by creating defects at the desired lattice site during formation. For example, Mg x Ge 1-x O 2-x The structure (e.g., epitaxial layer) can be grown to be Ge deficient, providing open Ge sites to be substituted with Ga. Similarly, Ga substitution at Mg sites can be grown to be Mg deficient. x Ge 1-x O 2-x This can be achieved by growing the structure. Aluminum doping (not shown in this chart) can be achieved using a similar approach. In this case, Mg-deficient growth promotes substitution of Al at the Mg site, resulting in n-type material. Also, Ge-deficient growth promotes substitution of Al at the Ge site, resulting in p-type material. These substitution strategies are further explained by the formation energies in Figure 42B.

[0148] FIG. 42B illustrates the relative formation energies for selected substitutional impurity doping embodiments of Mg or Ge sites in Mg2GeO4(Fd3m). In this diagram, the impurity species are selected from Li, Al, Ga, and Ge. However, others, such as Si, Bi, and / or Zn, may also be possible. Using the dopant Li as an example, bar 4210 represents the formation energy in eV of a Ge site substituted with Li, and bar 4220 represents the formation energy in eV of a Mg site substituted with Li. The diagram in FIG. 42B shows that the Li doping case exhibits the lowest formation energy among the impurity species presented, and both cases result in p-type conductivity. In another example considering Ga as the dopant, FIG. 42B illustrates that insertion of Ga at a Ge site requires more energy than insertion of Ge at a Ge site. However, insertion of Mg with a Ge deficiency requires more energy. x Ge 1-x O 2-x As the structure grows, Ga atoms are allowed to substitute into Ge sites.

[0149] Mg x Ge y O z Film growth and semiconductor structure Mg x Ge y O z Methods for forming epitaxial layers of SiO2 require the thermokinetic insights described herein to properly grow the structure, such as with molecular beam epitaxy (MBE).

[0150]

number

[0151] and ternary oxides

[0152]

number

[0153] Figure 1 shows a plot of the thermodynamic oxidation Ellingham diagram for the formation of . For a given growth temperature and metal species, there exists an oxygen partial pressure at which the metal begins to oxidize. For each oxide reaction, the boundary line is shown below which the metal remains oxide-free (below the boundary line) or the respective oxide is formed (above the boundary line).

[0154] In general, growth phase diagrams for depositing oxide compounds selected from constituent species indicate that sufficient oxygen species are required at a given growth temperature to optimize the reaction pathway. The oxygen source may be selected from molecular oxygen, excited metastable molecular oxygen, and atomic oxygen. In one embodiment, activated oxygen, such as atomic oxygen, is utilized for the growth of oxides described in this disclosure. The activated oxygen may be from an excited plasma source or an ozone source. Mg from an elemental source (e.g., a solid-state flow source such as a Knudsen cell) may be used. x Ge y O z For thin film growth, the Ellingham diagram in Figure 43A (oxygen partial pressure (Torr) vs. 1000 / temperature T (Kelvin)) shows the x Ge y O z This confirms that the incident Ge flux reaching the surface must be supplied in excess of the Mg flux over the entire temperature range to form the compound. If insufficient Ge is supplied during codeposition, the germanium oxide formed on the surface is reduced to elemental Ge in the presence of Mg.

[0155] Figures 43B, 43C, and 43D show schematic diagrams illustrating the surface kinetics of film formation on the epi surface 4310 using constituent elemental fluxes. Figure 43B shows the surface kinetics of incident Ge or Mg atoms in the absence of oxygen. The arrival rate of the incident atoms minus the desorption rate results in the net deposition surface coverage. The desorption rate depends on the surface temperature. For growth conditions of interest in this disclosure, the desorption rate of Mg dominates due to its lower activation energy for desorption compared to Ge, and follows Arrhenius behavior. Furthermore, both Ge and Mg exhibit sticking probabilities of less than 1 for chemisorption and physisorption to the epi surface 4310, which also depend on the bond energy of the crystal-surface complex (which is closely related to the surface reconstruction of the bound surface atoms and is a function of crystal symmetry, surface temperature, and oxygen environment). In general, higher growth temperatures promote surface migration of cations and a layer-by-layer growth mode. The high growth temperature of the surface 4310 allows the oxygen molecules to easily dissociate. However, it has been found that activated forms of oxygen are preferred, ideally neutral or ionized oxygen atoms, with neutral species being more preferred. Figure 43B shows the co-deposition of cations and anions towards the surface. When co-depositing Ge and O, the ratio of Ge:O determines where the oxide can form. Furthermore, GeO bound to the surface in crystalline form readily decomposes into GeO(g) vapor at sufficiently high temperatures, such as above 500°C. In the example of growth on an MgO substrate, the sticking coefficients δ of Mg and Ge are i Considering this, according to the present disclosure, it is experimentally possible to obtain δ Mg is approximately 10×δ Ge , δ Mg It was experimentally found that the ρ is about 0.5, which means that Mg tends to deposit more easily on MgO substrates than Ge.

[0156] Figure 43C shows the arrival of cations (filled circles) and anions (open circles) at the epi surface 4310. When cation = Mg and anion = O, MgO formation occurs. Similarly, when cation = Ge, surface formation of GeO is possible. If the surface temperature is high enough, the solid GeO formed on the epi surface will decompose into volatile GeO vapor, with an activation temperature of approximately 500°C for GeO decomposition.

[0157] Figure 43D shows the case where a high Ge flux is incident relative to the co-deposited oxygen flux. In this scenario, Ge may etch either GeO or MgO formed on the epi surface 4310. The growth limit between deposition and etching can be controlled by the Ge:[active-O] ratio. Here, deposition is performed under conditions where Ge:O<1 and the growth temperature is less than about 600°C. In the surface etching process, the etch rate increases at surface temperatures between 400 and 800°C, with Ge:O>1.

[0158] Figures 43C and 43D show the desired Mg x Ge y O z This indicates that for successful layer formation, a careful balance between the growth temperature and the flux ratio between Mg, Ge, and O must be maintained to deposit Ge while preventing removal of Ge and / or Mg from the surface. For example, as illustrated in Figure 43A, x Ge y O z A large excess of Ge flux relative to Mg flux is required to form and stabilize the compound. However, too much Ge flux can lead to problems with Mg, such as Ge becoming a volatile vapor as shown in Figure 43C, or Ge etching GeO2 or MgO as shown in Figure 43D. x Ge y O z This can have adverse temperature-dependent effects on the formation of GeO. Furthermore, the formation of pure GeO single crystal films has proven challenging due to the possibility of polymorphism, which can result in polycrystalline and amorphous structures, as well as surface dynamics. In embodiments of the present disclosure, Mg and Zn are ideal crystal growth modifiers that stabilize the formation of GeO-based compounds, allowing these compounds to achieve extremely high crystallinity in a given crystal structure.

[0159] Figure 43E shows a schematic ternary phase diagram of the Mg-Ge-O system calculated using DFT, showing the stable compositions of MgO, GeO2, Mg2GeO4 and MgGeO3, all on the same line from point 4322 (MgO) to point 4324 (GeO2). The relative molar fractions are represented by the composition points on the state diagram. Note that since the compositions shown represent the most stable forms, there are no continuous alloy compositions.

[0160] Figure 44A shows representative growth state diagrams 4400 for the epitaxial co-deposition of the compound Mg 1-y Ge y O 1+y (0 ≦ y < 1) (corresponding to 0 < x ≦ 1). Note that x = 0 and y < 1 are excluded in the above formula only to emphasize that it is difficult to realize the single crystal form of GeO2 as a single crystal film. However, there is a possibility that a small amount of stabilizing species acting as a surfactant can actually stabilize the crystalline GeO2 film. For example, Bi can be used as a surfactant. For a given growth temperature and incident Ge:Mg flux ratio, the resulting surface adsorbed atom ratio y x Ge 1-x O 2-x is plotted in Figure 44A. Ge

[0161] Isothermal lines in the range of 1 to 10 of the incident flux ratio

[0162]

Number

[0163] are shown. The constructed state diagram (y Ge , T g , k) shows the underlying kinetic process where the Mg deposited on the surface shows an Arrhenius desorption rate as a function of temperature. The Ge deposited on the epi-surface reacts with oxygen to form an oxide compound or decomposes into GeO monoxide vapor at an activation temperature of about 500 - 550 °C. Therefore, the elemental incident flux ratio

[0164]

number

[0165] is modified by the deposition surface temperature to maintain the adsorbed surface species ratio given by Figure 44B shows the same growth stage diagram 4400 as Figure 44A. Like Figure 44A, it also shows an experimentally determined growth window 4410 for high-quality single-crystalline Mg2GeO4 with the Fd3m crystal structure deposited on an MgO substrate. The surface growth temperature in this window 4410 is between about 400°C and about 500°C.

[0166]

number

[0167] The value of k is approximately k = 3 to 9, and the surface adatom ratio y Ge The line indicated by arrow 4420 corresponds to the stoichiometric value y Ge = 1 / 3 (i.e., the stoichiometry value to achieve the desired stoichiometry of Mg2GeO4). Growth window 4410 shows the conditions under which the actual growth of Mg2GeO4 was experimentally verified, and growth window 4410 is overlaid with the theoretical predictions graphed on the chart.

[0168] Figure 45 shows

[0169]

number

[0170] In the case of Mg 1-y Ge y O 1+yshows the growth state diagram 4500. The resulting change in the growth rate of the film 4510 is shown, and the units shown on the Y-axis on the right are related to the relative growth rate. Zero corresponds to the adsorption species on the growth surface being negligible, and values greater than 1 indicate a net accumulation of surface species. The growth rate is limited by two kinetic effects, namely the desorption of the Mg surface and the decomposition of GeO. When the incident Ge:Mg flux ratio is fixed, the stoichiometric control window of y Ge =1 / 3 has an optimal growth temperature of 450 °C < Tg < 500 °C, as discussed above. Due to the inherent self-organization of stable Mg2GeO4 during co-deposition, a slight deviation from the optimal flux and temperature is allowed. In Figure 45, when the growth temperature rises above the growth window, the growth rate drops dramatically, as indicated by arrow 4515. Arrow 4525 indicates that as the temperature decreases, the adsorption atom y Ge ratio decreases.

[0171] A method of forming a semiconductor device includes preparing a substrate comprising a substantially single-crystalline substrate material having a crystal symmetry compatible with the epitaxial layer of Mg x Ge 1-x O 2-x and co-depositing a material on the substrate to form Mg x Ge 1-x O 2-xand forming an epitaxial layer of Mg, Ge, and oxygen, where the value of x is 0≦x<1. The material includes at least two elements selected from Mg, Ge, and oxygen according to the value of x, with Mg, Ge, and oxygen provided by a Mg source, a Ge source, and an activated oxygen source, respectively. In some embodiments, the epitaxial layer is substantially monocrystalline and strained to the lattice-mismatched layer or substrate. In yet other embodiments, the epitaxial film and the substrate can have different space group symmetries by configuring them to have a surface configuration favorable for a heterointerface with low lattice mismatch. In some embodiments, the co-deposition is carried out using a molecular beam epitaxy process or an elemental flux co-deposition process. Metalorganic or other precursor species can also be used to transport Ge, Mg, and O to the growth surface. In some embodiments, the method comprises: x Ge 1-x O 2-x In some embodiments, the co-deposition includes depositing a buffer layer between the epitaxial layer of MgO and the epitaxial layer of MgO, for example, the buffer layer includes MgO or other forms of magnesium germanium oxide. x Ge 1-x The epitaxial layer of O2 self-assembles. In one embodiment, the adsorbed surface species ratio of Ge to Mg is

[0172]

number

[0173] In order to maintain the deposition surface temperature, the elemental incident flux ratio between the Ge source and the Mg source is adjusted according to the deposition surface temperature.

[0174]

number

[0175] In some embodiments, the co-deposition may include determining a growth temperature of 400-500° C. and a flux ratio of Ge source to Mg source (k) of k=3-9.

[0176]

number

[0177] For example, the flux ratio k has a value of 3 to 7.5, and Mg x Ge 1-x O 2-x The method can include forming a superlattice on a substrate, the superlattice having a unit cell including a first layer and a second layer, the first layer in the superlattice being Mg x Ge 1-x O 2-x In some embodiments, the second layer of the superlattice is an epitaxial layer of Mg y Ge 1-y O 2-y where y ranges from 0 to 1 and x≠y.

[0178] Various semiconductor structures for electronic and optoelectronic devices are manufactured in accordance with the present disclosure using Mg grown as epitaxial single crystal structures on purposefully selected substrates. x Ge 1-x O 2-x For example, any of the substrate materials listed in FIG. 12 can be used to grow epitaxial layers of MgGeO or MgGeO. In some embodiments, the semiconductor structure comprises a substrate comprising a substantially single crystalline substrate material and an epitaxial layer of MgGeO on the substrate. x Ge 1-x O 2-x epitaxial layer (where x is 0≦x<1), and Mg x Ge 1-x O 2-x The epitaxial layer has a crystal symmetry that matches the substrate material.

[0179] 46A shows a cross-sectional view of an epitaxially formed layered semiconductor structure 4600 including a substrate 4601, a substantially single crystalline epilayer (i.e., epitaxial layer) 4602 (which may include one or more epilayers), and an optional cap 4603. In some embodiments, substrate 4601 may be MgO, epilayer 4602 may include Mg2GeO4(Fd3m), and cap 4603 may be formed of MgO. In this example, epilayer(s) 4602 are formed via a deposition process on a prepared substrate 4601 having a surface crystal structure that is favorable for epilayer growth as described herein.

[0180] Figure 46B is a table 4605 of crystalline structure properties of exemplary epitaxial film materials 4610 and substrates compatible with Mg2GeO4, according to some embodiments. Experiments have shown that the mismatch in lattice match between Mg2GeO4 and the substrate or other listed cubic oxides can be managed to form highly matched, extremely low defect density structures. The smallest lattice mismatch between Mg2GeO4 and the substrate was found to be the substrate material MgO (column 4620), followed by Al2MgO4 (column 4622) and LiF (column 4624). These substrates are important because of their high optical transparency in the extreme ultraviolet region. All listed compounds are cubic, with MgO and LiF having approximately half the lattice constant of the AB2O4 compound, where {A, B} are selected from {Al, Ga, Ge, Zn}.

[0181] Figure 46C shows a cubic crystal structure diagram of a substrate selected from MgO or LiF, or similar crystal structures, with symmetric lattice constant a for each unit cell 4650. Other substrates, such as monoclinic beta-phase Ga2O3, can also be used, using a unique (100) oriented surface.

[0182] FIG. 46D shows a cubic crystal structure diagram of a substrate or epilayer selected from Mg2GeO4, MgGa2O4, ZnGa2O4 or MgAl2O4 (or the like), where the symmetry lattice constant of each unit cell 4652 is about twice that shown in FIG. 46C, i.e., about 2a.

[0183] Figure 46E shows a schematic diagram illustrating the formation of coherent epitaxial layers along the growth direction of two distinct cubic crystal structures with large differences in lattice constants. Layer 4660 has a lattice constant a (see Figure 46C), and layer 4670 has a lattice constant of approximately 2 × a (see Figure 46D). For example, epilayer 4670 of Mg2GeO4 (or γ-Ga2O3, MgGa2O4, ZnGa2O4, or MgAl2O4) can be formed with a low dislocation density at the growth interface between layers 4660 and 4670 by matching the lattice constant of the epilayer to a multiple (e.g., integer) of the lattice constant of the underlying film or substrate (layer 4660). In the example of Figure 46E, the lattice constants are matched by approximately an integer multiple of 2 in the plane of the heterointerface. That is, the lattice constant a of layer 4660 matches the lattice constant 2a of layer 4670. In other words, in the xy growth plane perpendicular to the growth direction, the lattice constants are integer multiples of 2, corresponding to a 2 × 2 arrangement of unit cells. For example, this can be achieved with cubic Mg2GeO4 deposited on an MgO substrate.

[0184] 47-50 show graphs of different flux values ​​according to some embodiments.

[0185]

number

[0186] 1 shows X-ray diffraction (XRD) diagrams (intensity vs. angle in Ω-2θ scans) of exemplary layered semiconductor structures for 1000 sq. m and 1000 sq. m, illustrating the effect of the flux ratio of impinging species on the resulting grown layer.

[0187] Figure 47 shows an experimentally determined triaxial XRD diagram of the layered semiconductor structure shown in Figure 46A. A 240 nm Mg2GeO4(Fd3m) epitaxial layer was grown using

[0188]

number

[0189] The MgO substrate was grown with an incident Ge and Mg flux of 0.01 μm and a growth temperature of 425°C. A thin MgO cap (approximately 11 nm) was deposited on the surface to form an X-ray Fabry-Perot structure. In this example, the extremely narrow full width at half maximum (FWHM) of the Mg2GeO4 peak 4710 compared to the MgO substrate peak 4720 clearly indicates the high quality of the structure. Furthermore, the MgO cap layer further induces high-frequency Penderosung oscillations 4730 in the XRD curve, indicating a completely coherent structure with a low defect density and very high quality. The MgO substrate, cubic Mg2GeO4 film, and MgO cap are highly crystalline, plane-parallel at the atomic scale, and coherent throughout the structure, thereby inducing the X-ray Fabry-Perot effect, as shown. The separation of the bulk Mg2GeO4(400) 4710 peak from the substrate MgO(200) peak 4720 is 1395.9 arcsec.

[0190] FIG. 48 is an experimentally determined triaxial XRD diagram of the layered semiconductor structure shown in FIG. 46A, showing the Mg2GeO4(Fd3m) layer, approximately 254 nm thick,

[0191]

number

[0192] The structure was grown using an incident Ge and Mg flux of 4800 Å and a growth temperature of 425°C. The MgO substrate (peak 4820) and Mg2GeO4 peak 4810 indicate the formation of a crystalline film of Mg2GeO4. A MgO cap similar to that in Figure 47 is also included in the structure, but there are fewer high frequency fringes 4830, indicating a less coherent crystalline structure than in Figure 47.

[0193] FIG. 49 is an experimentally determined triaxial XRD diagram of the layered semiconductor structure shown in FIG. 46A, showing the Mg2GeO4(Fd3m) layer as

[0194]

number

[0195] The growth temperature was 425° C. The formation of the Mg2GeO4 (Fd3m) layer was minimal, as can be seen from the substrate peak 4920 and the Mg2GeO4 peak 4910.

[0196] Figure 50 is an experimentally determined triaxial XRD diagram of the layered semiconductor structure shown in Figure 46A. In Figure 50, the incident Ge and Mg fluxes are

[0197]

number

[0198] The growth temperature was 425°C. As shown, no Mg2GeO4 (Fd3m) layer was formed, and the layered semiconductor structure contained only MgO (peak 5020). Figures 47-50 show that the higher relative amount of Ge flux in the molecular beam epitaxy process

[0199]

number

[0200] This indicates that the addition of SiO2 improved the quality of the epitaxial Mg2GeO4(Fd3m) layers under the growth conditions used. As shown in Figure 47, very high quality Mg2GeO4(Fd3m) epitaxial layers on MgO were achieved.

[0201] multilayer structure The Mg disclosed herein x Ge y O zThe epitaxial layers of the present invention can be used in a variety of semiconductor device structures, including multilayer structures and chirped layers. The multilayer structure can be, for example, a superlattice ("SL," which may include a short period superlattice), multiple superlattices, a compositionally varied (or graded) layer, or a compositionally varied (or graded) multilayer structure (or region). Embodiments of multilayer structures are illustrated in Figures 51-53. While specific combinations of substrate materials, buffer layers, and multilayer layers are described, as described throughout this disclosure, Mg x Ge y O z Other combinations using materials and compatible substrates are possible.

[0202] Embodiments include semiconductor structures and devices having a superlattice, the superlattice including a unit cell having a first layer, Mg x Ge 1-x O 2-x is the first layer in the superlattice. In some embodiments, the unit cell of the superlattice is y Ge 1-y O 2-y where y ranges from 0 to 1 and x≠y. For example, the first layer may be MgO having an Fm3m space group, and in some embodiments, the unit cell of the superlattice further comprises a second layer made of Mg2GeO4 having an Fd3m space group. In some embodiments, the unit cell of the superlattice further comprises a second layer of MgGa2O4 or gamma-Ga2O3.

[0203] 51 shows a cross-sectional view of an epitaxially formed layered semiconductor structure 5100 including a substrate 5110, an optional buffer layer 5120, a superlattice or multilayer periodic structure 5130, and an optional cap 5140. The periodic structure 5130 (e.g., superlattice) comprises a repeating unit cell 5135, which in this example consists of two layers 5132 and 5134. The illustration shows that layer 5132 is thicker than layer 5134. However, in other embodiments, layer 5132 may be thinner than layer 5134, or layers 5132 and 5134 may be of equal thickness. In an exemplary embodiment, the layered semiconductor structure 5100 includes an MgO(100) oriented substrate 5110, a superlattice (periodic structure 5130) comprising a unit cell 5135 of MgGeO and MgO with a constant period, and a final cap 5140 of MgO. The substrate 5110 may be prepared for deposition using an oxygen plasma source to terminate the MgO surface with oxygen species.

[0204] 52A-52F show experimentally determined XRD diagrams for the embodiment of the semiconductor structure shown in FIG. 51. In FIGS. 52A-52F, the structure included a superlattice. In FIGS. 52A-52F, the peak representing the substrate is marked as "SUB." Additional arrows point to "satellite peaks," which are features of the XRD pattern caused by the periodic SL structure. FIGS. 52H-52I show experimentally determined XRD diagrams for the embodiment of the semiconductor structure shown in FIG. 52G.

[0205] Figure 52A is an experimentally determined XRD diagram 5200 of the layered semiconductor structure shown in Figure 51. Figure 52A is a diagram of a superlattice 5130 in which layer 5132 is Fd3mMgGa2O4 and layer 5134 is Mg2GeO4 (Fd3m). The Mg2GeO44 (Fd3m) layer was deposited on an MgO buffer layer 5120 which was deposited on an MgO(100) substrate 5110.

[0206]

number

[0207] As shown, the high-frequency fringes (one example is labeled fringe 5210) show very high crystalline quality, suggesting excellent matching between the in-plane lattice constants of the MgGa2O4 and Mg2GeO4(Fd3m) layers that form the superlattice. The sharp satellite peaks (arrows SL n The SL period is 26 nm, and the SL n=0 The separation of the MgO(200) substrate from the peak is 1024.8 s.

[0208] Figure 52B is an experimentally determined XRD diagram 5201 of the layered semiconductor structure shown in Figure 51. Figure 52B is a diagram of a superlattice 5130 in which layer 5132 is Mg2GeO4(Fd3m) and layer 5134 is MgGa2O4(Fd3m). The superlattice was deposited on an MgO buffer layer 5120 and an MgO(100) substrate 5110. Sharp satellite peaks (SL n ) shows a consistently grown structure with high quality. Figure 52B shows a superlattice with a unit cell in which the first layer of the unit cell is Mg2GeO4 with the Fd3m space group and the second layer of the unit cell is MgGa2O4 with the Fd3m space group. The SL period is 27.2 nm with 10 periods, and the SL n=0 The MgO(200) substrate separation from the peak is 1206.7 s, which indicates a thicker MgGe2O4 layer compared to Figure 52A.

[0209] Figure 52C is an experimentally determined XRD diagram 5202 of the layered semiconductor structure shown in Figure 51. Figure 52C is a diagram of a superlattice 5130 in which layer 5132 is Mg2GeO4(Fd3m) and layer 5134 is MgO(Fm3m). The superlattice was deposited on an MgO buffer layer 5120 and an MgO(100) substrate 5110. Sharp satellite peaks (SL n) again shows a consistently grown structure with high quality. Figure 52C shows a superlattice with a unit cell in which the first layer of the unit cell is MgO with the Fm3m space group and the second layer of the unit cell is Mg2GeO4 with the Fd3m space group. This result demonstrates that Mg2GeO4 can form on MgO and that MgO can form on Mg2GeO4. The SL period is 26.6 nm with N=10 periods and N-2 thickness oscillations between satellite peaks. SL for MgO(200) substrate n=0 The separation is 614.2 seconds,

[0210]

number

[0211] This indicates that Figure 52D shows an experimentally determined XRD diagram 5203 of the layered semiconductor structure shown in Figure 51. Figure 52D is for a superlattice 5130 in which layer 5132 is Mg2GeO4 (Fd3m) and layer 5134 is MgO (Fm3m). The layers of the structure depicted in Figure 52D have more satellite peaks (SL n 52C, as evidenced by the presence of a MgO buffer layer 5120 and a MgO(100) substrate 5110. The sharp satellite peaks indicate a consistently grown structure with high quality. The SL period is 35.6 nm and is grown with 10 periods.

[0212] Figure 52E is an experimentally determined XRD diagram 5204 of the layered semiconductor structure shown in Figure 51. Figure 52E is a diagram of a superlattice 5130 in which layer 5132 is Mg2GeO4(Fd3m) and layer 5134 is stabilized cubic gamma-Ga2O3. The superlattice was deposited on an MgO buffer layer 5120 and an MgO(100) substrate 5110. Sharp satellite peaks (SL n ) shows a consistently grown structure with high quality. The superlattice period is 18.4 nm with 10 periods. SL n=0The separation of the MgO(200) substrate from the peak is 1337.6 s;

[0213]

number

[0214] The cubic γGa2O3 layer is stabilized by intentionally growing the layer below the critical layer thickness (CLT), making it energetically favorable to form βGa2O3 unless stabilized by another species.

[0215] Figure 52F is an experimentally determined XRD diagram 5205 of the layered semiconductor structure shown in Figure 51. Figure 52F is a diagram of a superlattice 5130 in which layer 5132 is Mg2GeO4(Fd3m) and layer 5134 is MgO. After deposition of the 10x period SL [MgO / Mg2GeO4], a thick beta-Ga2O3 cap layer was formed with a thickness above the CLT (note: stabilized gamma-Ga2O3 can form below the CLT). The superlattice was deposited on an MgO buffer layer 5120 and an MgO(100) substrate 5110. In this example, a thick beta-phase Ga2O3 film 5220 was grown as the cap layer 5140, which exhibits a highly coherent structure. Sharp satellite peaks (SL n ) shows a consistently grown structure with high quality. The SL serves as a pseudosubstrate for the formation of high-quality beta-Ga2O3 (100) oriented films with biaxial strain suitable for tuning the valence band edge. Figure 52F shows a superlattice with a unit cell in which the first layer is Mg2GeO4 with Fd3m space group and the second layer is MgO with cubic symmetry, and a thick epitaxial beta-phase Ga2O3 is grown on the SL acting as a pseudosubstrate.

[0216] Figure 52G shows, in a further example, a complex epilayer structure 5206 of heterogeneous cubic oxide layers integrated into a superlattice or multi-heterojunction structure. Large lattice constant cubic GeMg2O4 (i.e., Mg2GeO4) and small lattice constant MgO layers are shown grown along the growth direction to form a superlattice with a repeat period Λ of N repeats. The MgO(100) oriented substrate allows for a lattice "2x" cube-on-cube matching to GeMg2O4. The direct band gap Ek of both materials allows for tuning of unique electronic band structures with preselected quantized energy levels from the specific layer thicknesses that make up the SL period. The thickness "L" of the layers that make up the SL is small, while the L GeMgO and L MgO When the thickness of each of the layers is less than about 10-20 times the thickness of the respective unit cell (i.e., the layer thickness is less than about 150 nm),

[0217]

number

[0218] A digital pseudoalloy having the following structure can be formed: An optional MgO cap layer is shown that can be used to protect the final surface of the structure.

[0219] FIG. 52H shows experimental XRD data 5207 for Fd3m crystal structure Mg2GeO4 deposited as a high quality bulk layer on a Fm3mMgO(100) substrate and further including an MgO cap.

[0220] FIG. 52I shows experimental XRD data 5208 for the Fd3m crystal structure Mg2GeO4 when incorporated as an SL structure containing a 20x periodic SL [Mg2GeO4 / MgO] on a Fm3mMgO(100) substrate.

[0221] As shown in Figure 52H, the extremely high quality of Mg2GeO4 is evidenced by the small FWHM diffraction peaks of the epilayer (labeled "GeMgO4400") and the high frequency thickness oscillations generated by the X-ray Fabry-Perot effect of the parallel atomic planes of the film and MgO cap layer, which are strain-coherent with the underlying substrate crystal. This high degree of lattice matching between Mg2GeO4 and MgO can be further exploited to form complex SL structures, as shown in Figure 52I. Figure 52I shows the 20× periodic SL [Mg2GeO4 / MgO] / MgO. sub (100) is shown. In this case, there are also many sharp SL satellite peaks. n demonstrates a coherent strain structure. Both Mg2GeO4 and MgO constituent materials

[0222]

number

[0223] is the direct band gap. A thin layer of a low-bandgap material, roughly 1-5 crystal unit cell thick, sandwiched between a wide-bandgap material such as MgO, can quantum confine the conduction band minimum and valence band maximum. The transition energy between the lowest quantized energy level of the conduction band and the highest quantized energy level of the valence band in Mg2GeO4 can be tuned by varying the thickness through quantum confinement effects. This tuning method allows the transition energy to be varied from 5.81 eV to 7.69 eV. This energy range is optimal for optoelectronic light-emitting devices operating in the deep ultraviolet (161-213 nm) portion of the electromagnetic spectrum.

[0224] FIG. 53 illustrates a physically realized Mg x Ge y O z1 is a summary table of properties of exemplary structures including epilayers. The table lists the constructed epilayer structure, the space group of the structure, whether the structure layer is bulk or superlattice, the quality, and the substrate material. As demonstrated in these examples, the suitability of MgGa2O4, Mg2GeO4, MgO, beta-Ga2O3, and gamma-Ga2O3 is advantageous for bandgap energy engineering of electronic and optoelectronic devices.

[0225] Conduction Device The Mg disclosed herein x Ge y O z Semiconductor structures, such as conductive devices, are described that include epitaxial oxide materials and compatible substrates. Semiconductor structures that include the epitaxial oxide materials described herein can be single layers on a substrate or multiple layers on a substrate. Such multiple-layer semiconductor structures can include single quantum wells, multiple quantum wells, superlattices, multiple superlattices, compositionally varied (or graded) layers, compositionally varied (or graded) multilayer structures (or regions), doped layers (or regions), and / or multiple doped layers (or regions). Such semiconductor structures with one or more doped layers (or regions) can include pn, pn, nin, pip, npn, pnp, p-metal (to form a Schottky junction), and / or n-metal (to form a Schottky junction) doped layers (or regions).

[0226] In some embodiments, a semiconductor device includes a substrate comprising a substantially single-crystalline substrate material and an active region on the substrate. The active region comprises Mg x Ge 1-x O 2-x where x has a value of 0≦x<1. x Ge 1-x O 2-xThe epitaxial layer has a crystal symmetry that matches the substrate material. In some embodiments, the semiconductor device is an optoelectronic device, including a light emitting diode or a photodetector. For example, the optoelectronic device emits or absorbs light having a wavelength in the range of 150 nm to 280 nm, e.g., 150 nm to 260 nm, or less than about 260 nm, or less than about 250 nm, or less than about 220 nm. In some embodiments, the semiconductor device is an electronic device, including a diode or a transistor. In some embodiments, Mg x Ge 1-x O 2-x The epitaxial layer of Mg is the intrinsic layer of the semiconductor device. x Ge 1-x O 2-x The epitaxial layer is a doped layer containing n-type or p-type conductivity, doped by the methods disclosed herein.

[0227] Devices including semiconductor structures comprising the magnesium germanium oxide materials described herein can include electronic and optoelectronic devices. For example, the devices described herein can be resistors, capacitors, inductors, diodes, transistors (switches, e.g., field-effect transistors, RF power switches), amplifiers, sensors, photodetectors, LEDs, and lasers. In some embodiments, devices including semiconductor structures comprising the magnesium germanium oxide materials described herein are optoelectronic devices, such as photodetectors, LEDs, and lasers, that detect or emit ultraviolet light (e.g., having a wavelength of 150 nm to 280 nm). In some cases, the device includes an active region in which the detection or emission of light occurs, and the active region includes a magnesium germanium oxide material with a bandgap selected to detect or emit ultraviolet light (e.g., having a wavelength of 150 nm to 280 nm).

[0228] Doped layers in conductive devices may contain donor (n-type) or acceptor (p-type) impurity (or dopant) materials. The impurities can function as external dopants that impart n-type or p-type conductivity to the doped layer. For example, a doped layer may contain a high concentration (e.g., a doping concentration of about 1e17 cm) of impurities (or dopants) that can act as donor (n-type) or acceptor (p-type) materials. -3 ~About 1e20cm -3 , or about 1e20cm -3 The doping can be achieved by various doping strategies disclosed herein, such as those described in connection with Figures 18A-42B.

[0229] Reference is now made to Figure 54A, which shows an epitaxial multi-layer structure 5400 for forming an electronic or optoelectronic device, including a substrate 5410, an optional buffer layer 5420, a first conductivity type layer 5430, a second conductivity type region 5440, a third conductivity type region 5450, and an electrical contact layer 5460. The structure 5400 is formed sequentially layer by layer along a growth direction Z. The first, second, and third conductivity types can be selected from p-type, i-type, and n-type, and can be used in various combinations, such as pn, np, pin, nin, pip, npn, pnp, etc. The Mg x Ge y O z Bulk layers or superlattices of epitaxial oxide material may be used for layers 5420, 5430, 5440, 5450 and / or 5460.

[0230] FIG. 54B illustrates a Mg a Ge b O c The possible conductivity types of Mg are shown. The process of creating conductivity is depicted schematically. a Ge b O c5490 is shown to have doping 5492 resulting in conductivity type 5494. In one example, Mg a Ge b O c may be Mg2GeO4 (a=2, b=0, c=4). The conductivity type of FIG. 54B can use co-doping or crystal growth modification during epitaxial layer formation. Doped layers can be formed in any of the regions described in FIG. 54A. In some embodiments, electrically activated doped or conductivity type regions can have atomically abrupt interfaces with other different regions, or the regions can be graded over a predetermined distance.

[0231] Figure 55A shows Mg a Ge b O c 5 shows a multi-layer structure 5500 used to form an electronic device having distinct regions including at least one layer of a substrate 5510 having an epitaxial layer 5520 deposited along a growth direction Z. n (e.g., a film or region) 5520 that make up the device. n is selected from at least one Mg2GeO4 form, e.g., Zn x Ge y O z , Zn x Ga y O z , Al x Ge y O z , Al x Zn y O z , Al x Mg y O z , Mg x Ga y O z , Mg x Zn y O z , and Ga x O z The composition may be combined with a type selected from the following: where x, y, and z represent relative molar fractions.

[0232] Figure 55B shows the Mg a Ge b O c This is a representation of an example of a composition that can be combined with Mg a Ge b O c 5590 plus heterostructure material 5592, in this example the heterostructure material composition 5594 is Mg x Ge y O z , Zn x Ge y O z , Zn x Ga y O z , Al x Ge y O z , Al x Zn y O z , Al x Mg y O z , Mg x Ga y O z , Mg x Zn y O z and Ga x O z FIG. 55C is a plot of the minimum energy gap (eV) versus lattice constant (c, Angstroms) of Mg2GeO4 and other materials that may be used in heterostructures of semiconductor structures of the present disclosure. The plot can be used to determine the lattice matching of suitable crystal structures for a combination of materials. x Ge 1-x O 2-x on a substrate, where x has a value of 0≦x<1, and the second epitaxial layer is Mg x Ge 1-x O 2-x The second epitaxial layer is a Zn epitaxial layer. x Ge y O z , Zn x Ga y Oz , Al x Ge y O z , Al x Zn y O z , Al x Mg y O z , Mg x Ga y O z , Mg x Zn y O z , or Ga x O z where x, y, and z are mole fractions.

[0233] Figure 56A is a band structure diagram of bandgap energy (eV) as a function of growth direction Z, representing a homojunction device with a p-i structure. The structure is formed along the growth direction Z using spatial control of the doping regions. Moving from left to right along the growth direction, an n-type region is formed first, followed by an unintentionally doped region (intrinsic "i" region), and then a p-type region. In various embodiments, the doping transition between the n, i, and p regions can be abrupt or step-wise over distance. The bandgap height of each region is the same, and the bandgap energies E g The p and n regions form a diode. An electric field between the p and n regions is applied across the central intrinsic region along the Z axis, causing electrons and holes to be injected into the i region. In an embodiment, a device having the structure of Figure 56A may have one or more pin layers made of Mg, such as Mg2GeO4. x Ge y O z may include:

[0234] FIG. 56B is a band structure diagram representing a homojunction device, such as a diode, having a nI n structure. The nI n structure is formed along the growth direction Z using spatial control of doping regions. In various examples, the nI n local junctions can have abrupt or graded doping concentrations over a predetermined distance. In embodiments, a device having the structure of FIG. 56B can have one or more nI n layers doped with Mg, such as Mg2GeO4. x Ge y O z may include:

[0235] Figure 56C is a band structure diagram representing a heterojunction device, such as a diode, with a p-i-n structure. The structure is formed continuously along the growth direction Z using spatial control of composition and doping of discrete regions. In various embodiments, the composition and doping can be abrupt or stepped over a predetermined distance. The band gap energy E between the p and n regions is g The conduction band offset ΔE of the heterojunction is not necessarily the same, and in this example the bandgap of the n region is larger than the bandgap of the p region. c and the valence band offset ΔE v provides an energy barrier to control carrier flow / confinement. As shown, the p-i structure forms a diode, with a built-in electric field applying an electric field along the Z direction across the i region, as shown. Heterojunction structures are useful for light-emitting devices because light generated in the central region escapes without being absorbed by the p and n regions. That is, the heterojunction structure of FIG. 56C can be advantageously used as a light-emitting device (e.g., an LED) because the wide bandgap n and p regions have a low absorption coefficient for light emitted from the narrow bandgap i layer. In embodiments, a device having the structure of FIG. 56C may include one or more layers made of Mg, such as Mg2GeO4. x Ge y O z may include:

[0236] Figure 56D is a band structure diagram representing a double heterojunction device such as a quantum well. The structure is formed continuously along the growth direction Z using spatial control of composition. This structure has a wide bandgap E g1 Layer composition and narrow band gap region / layer E g2 It consists of E g2 <E g1 The narrow bandgap region is between two wide bandgap regions. A sufficiently thin narrow bandgap region allows quantization to occur at the allowed energy levels within the quantum well. In various examples, this can be used in optoelectronic and electronic devices. In embodiments, a device having the structure of Figure 56D may have one or more layers made of Mg, such as Mg2GeO4. x Ge y O z Since Mg2GeO4 is a direct bandgap material, optical emission is possible due to spatial recombination of electrons and holes.

[0237] Figure 56E shows the band structure diagram for a metal-insulator-semiconductor (MIS) structure. The semiconductor region has a band gap E g1 and the insulating region has a band gap E g2 In an embodiment, a device having the structure of FIG. 56E may have one or more layers containing Mg, such as Mg2GeO4. x Ge y O z For example, a Mg2GeO4 layer can be used as an insulator or a semiconductor. When a Mg2GeO4 semiconductor is used to form a metal-insulator-semiconductor structure, the insulator can be selected from a suitable wider bandgap material, as shown in Figure 55C.

[0238] Figure 56F is a band structure diagram representing a multiple heterojunction device, such as a diode with a pin structure and a single quantum well (QW). In this example, the band gaps of the n and p regions (E gn ,E gp ) is the barrier (band gap) of the QW region.

[0239]

number

[0240] ) and quantum wells

[0241]

number

[0242] where,

[0243]

number

[0244] Electrons and holes are injected from their respective reservoir regions into the intrinsic region. The conduction band offset of the heterojunction, ΔE c and the valence band offset ΔE v provides an energy barrier to control carrier flow / confinement. Heterojunction structures are useful for light-emitting devices because light generated in the central region escapes without being absorbed by the p and n regions. In other words, the n and p regions with wide band gaps have a low absorption coefficient for light emitted from the quantum well of the i layer with narrow band gap. Band gap

[0245]

number

[0246] A quantum well with a band gap

[0247]

number

[0248] The quantized energy levels of the conduction band and valence band confined between a barrier with a thickness of L QWIn other embodiments, the structure can have two or more quantum wells in the intrinsic region. The energy levels of a multiple quantum well structure affect various properties of the structure, such as the minimum effective bandgap. In some cases, such as light-emitting devices, having multiple quantum wells improves light emission by increasing the quantum well capture rate of carriers injected from the p and n regions into the i region. In embodiments, a device having the structure of Figure 56F can have one or more layers containing MgGeO, such as Mg2GeO4. x Ge y O z may include:

[0249] Figure 56G is a band structure diagram of a pin structure with a superlattice (SL) in the i region. The pin structure has multiple quantum wells, and the band gap of the barrier layer of the multiple quantum well structure in the i region is larger than the band gap of the n layer and p layer. In other cases, the band gap of the barrier layer of the multiple quantum well may be narrower than the band gap of the n layer and p layer. The lower half of Figure 56G shows a single quantum well in the multiple quantum well structure. The thickness L of the barrier layer is QB can be made thin enough to allow electrons and holes to tunnel (e.g., within the i-region and / or between the n-layer and / or p-layer to and / or from the i-region). Such a multiple quantum well structure can operate as a digital alloy, with properties depending on the materials comprising the barriers and wells and their thicknesses. In embodiments, a device having the structure of Figure 56G may contain MgGeO4, such as Mg2GeO4, in one or more layers. x Ge y O z may include:

[0250] Figure 56H shows the band structure of a pin structure with a superlattice in the p, i, and n regions. In this complete superlattice structure of p(SL)-i(SL)-n(SL), the p-, i-, and n-regions can be of the same or different compositions. The n-region is N n SL The pair of wells (thickness L1 and band gap EgW1 ) and barrier (thickness L2 and band gap E gB1 ) and the i region contains N i SL The pair of wells (thickness L3 and band gap E gW2 ) and barrier (thickness L4 and band gap E gB2 ) and the p region contains N p SL The pair of wells (thickness L5 and band gap E gW3 ) and barrier (thickness L6 and band gap E gB3 ). In this example, the band gap of the barriers and wells in the i-region is narrower than the band gaps of both the n-layer and the p-layer. In other cases of structures with multiple quantum wells, the band gap of the barrier layers may be wider than the band gaps of the n-layer and the p-layer. Furthermore, in some cases, the thickness and / or band gap of the barriers and / or wells in the n-, i-, and / or p-regions may vary across the individual regions (e.g., to form graded or chirped layers). The barrier layer thicknesses L2, L4, and / or L6 may be thin enough to allow electrons and holes to tunnel (e.g., within the i-region and / or between the n-layer and / or p-layer to and / or from the i-region). Each region of the structure may behave as a digital alloy, whose properties depend on the materials comprising the barriers and wells and their thicknesses. For example, the materials and layer thicknesses may be selected so that the n- and p-regions have wide band gaps and are therefore transparent (or have low absorption coefficients) to the wavelengths of light emitted from the i-region superlattice. In an embodiment, a device having the structure of Figure 56H may have one or more layers containing Mg, such as Mg2GeO4. x Ge y O z may include:

[0251] Figure 56I is a band structure diagram for a pin structure similar to Figure 56H. The band gaps and barrier and well thicknesses in the n, i, and p regions are defined as in Figure 56H. The superlattices in the n, i, and p regions in this example have alternating pairs of the same material, and the different well (or well and barrier) thicknesses in the i region adjust the optical properties. The structure has material A and material B, with the barriers in the n region superlattice composed of material A and the wells in the n region superlattice composed of material B. In this example, the barriers in the i and p regions are also composed of material A, and the wells in the i and p regions are also composed of material B. The wells in the i region are thickened so that the quantized energy levels of the potential wells are lower in energy relative to the band edges of the host wells, thereby resulting in an effective band gap in the i region superlattice that is narrower than the band gaps in the n and p regions superlattices (i.e., closer to the band gap of material A in the bulk state). Therefore, such structures can be used in light-emitting devices (e.g., LEDs) as described herein. In an embodiment, a device having the structure of FIG. 56I may have one or more layers made of Mg, such as Mg2GeO4. x Ge y O z may include:

[0252] In the following conductive devices shown in Figures 57-71, one or more layers of the active region are Mg x Ge y O zThe semiconductor layer region may be formed as a single layer of a single composition, multiple layers of different compositions, or a multilayer structure including repeating layers, each formed from individual sublayers. In conductive devices, the electrical materials forming contacts to the electron and hole injection regions are selected from low and high work function metals, respectively. In one example, metal ohmic contacts are formed in situ directly on the final metal oxide surface, thereby reducing any intermediate level traps / defects that occur at the semiconductor oxide-metal interface. Some examples of high work function metals that can be used for ohmic (or low resistance) contacts to p-type epitaxial oxide layers include Ni, Os, Se, Pt, Pd, Ir, Au, and their alloys. Examples of low work function materials that can be used for ohmic (or low resistance) contacts to n-type epitaxial oxide layers include Cs, Na, and lanthanides, but also Al, Ti, Ti-Al alloys, and the common metal titanium nitride (TiN). In some cases, the metal contact layer can include two or more layers of metals having different compositions (eg, a Ti layer and an Al layer).

[0253] Figure 57 shows an in-plane conduction device comprising, in this example, an insulating substrate and a semiconductor layer region formed on the substrate, with electrical contacts disposed on the upper semiconductor layer of the device. In this example, a first electrical contact or electrode (Contact 1) is disposed on the top surface of the semiconductor layer and a second electrical contact (Contact 2) is embedded in the semiconductor layer and laterally spaced from the first electrical contact, allowing in-plane current flow, as indicated by the large arrows.

[0254] Figure 58 shows a vertical conduction device, in this example comprising a conductive substrate and a semiconductor layer region formed on the substrate, with electrical contacts located on the top and bottom of the device. In this example, a first of the electrical contacts (Contact 1) is located on the top (either buried or on the top surface) of the semiconductor layer region. A second electrical contact (Contact 2) is located on the underside of the substrate and is vertically spaced from the first electrical contact, providing vertical current flow as indicated by the large arrow.

[0255] Figure 59 shows a representational cross-sectional view of a vertical conduction device for light emission (e.g., a light-emitting diode) configured as a parallel-plane waveguide for light emission, having the electrical contact configuration shown in Figure 58. The device comprises a substrate, a first semiconductor layer (Semi 1) having a first conductivity type, a second semiconductor layer (Semi 2) having a second conductivity type, and a third semiconductor layer (Semi 3) having a second conductivity type. For example, the first, second, and third conductivity types can be n-, i-, and p-, as described throughout this disclosure. A first electrical contact (Contact 1) is on the top surface of the device, and a second electrical contact (Contact 2) is on the bottom surface. Electrons and holes are injected into the middle semiconductor layer, and light is emitted in a plane parallel to the plane of the layers (i.e., perpendicular to the growth direction).

[0256] FIG. 60 shows a representational cross-sectional view of a vertical conduction device (e.g., a light-emitting diode) for emitting light and configured as a vertical light-emitting device, having the electrical contact configuration shown in FIG. 58. The device comprises a substrate, a first semiconductor layer (Semi 1) having a first conductivity type, a second semiconductor layer (Semi 2) having a second conductivity type, and a third semiconductor layer (Semi 3) having a second conductivity type. For example, the first, second, and third conductivity types can be n-, i-, and p-, as described throughout this disclosure. A first electrical contact (Contact 1) is on the top surface of the device, and a second electrical contact (Contact 2) is on the bottom surface. Electrons and holes are injected into the middle semiconductor layer. The substrate and other layers of the device can be designed to be transparent to the wavelength of light to be emitted, such that light is emitted through one or both of the top and / or bottom surfaces of the device. As shown, first and second electrical contacts are disposed on each surface to allow light to pass through.

[0257] Figure 61 is a representational cross-sectional view of an in-plane conduction device for light detection (e.g., a photodetector) having the electrical contact configuration shown in Figure 57 and configured to receive light passing through a semiconductor layer region and / or substrate. The device includes a substrate and a semiconductor layer region formed on the substrate, with electrical contacts disposed on an upper semiconductor layer of the device. In this example, a first electrical contact or electrode (Contact 1) is disposed on the top surface of the semiconductor layer, and a second electrical contact (Contact 2) is embedded in the semiconductor layer and laterally spaced from the first electrical contact. The substrate material is transparent to the wavelength of interest. Light received by the device generates an electric current, which is measured at the first and second electrical contacts.

[0258] Figure 62 shows a representational cross-sectional view of a light-emitting, in-plane conduction device (e.g., a light-emitting diode) having the electrical contact configuration shown in Figure 57 and configured to emit light vertically or in-plane. The device includes a substrate and a semiconductor layer region formed on the substrate, with electrical contacts disposed on the upper semiconductor layer of the device. In this example, a first electrical contact or electrode (Contact 1) is disposed on the top surface of the semiconductor layer, and a second electrical contact (Contact 2) is embedded in the semiconductor layer and laterally spaced from the first electrical contact. In embodiments where light is emitted vertically, the substrate material is transparent to the wavelengths to be generated.

[0259] Figure 63 shows a representational cross-section of an in-plane surface MSM conductive device including a substrate and a semiconductor layer region including multiple semiconductor layers (Semi 1, Semi 2, Semi 3). The top layer of metal includes a pair of planar interdigitated electrical contacts (Contact 1, Contact 2) separated by a distance "a". The width of the repeating portion of the device is Λ cell In this example, the in-plane MSM conduction device includes an optional third electrical contact (Contact 3) located on the bottom surface of the substrate. In the case of a conductive substrate, Contact 3 functions as a vertical conductive collector or drain. In the case of an insulating substrate, Contact 3 can function as a back gate for the field-effect device.

[0260] Figure 64A shows a top view of an in-plane dual-metal MSM conduction device with a first electrical contact (Contact 1) formed of a first metal material interdigitated with a second electrical contact (Contact 2) formed of a second metal material. As can be seen in the magnified view of a portion of the interdigitated contacts, the first electrical contact has a finger width of w1, and the second electrical contact has a finger width of w2, with a spacing of g between the contacts. The lateral gap g between the respective electrodes determines the in-plane electric field strength. Contact 1 and Contact 2 can be formed of different metals, for example, high and low work function metals. In other embodiments, the metal-Semi 1 heterointerface can form a Schottky barrier.

[0261] FIG. 64B shows a representational cross-sectional view of the in-plane dual metal MSM conduction device shown in FIG. 64A formed with a substrate and a semiconductor layer region epitaxially formed on the substrate, showing the electrical contact unit cell arrangement.

[0262] Figure 65 shows a representational cross-sectional view of a multilayer semiconductor device having a first electrical contact (Contact 1) formed on a mesa surface and a second electrical contact (Contact 2) spaced both horizontally and vertically from the first electrical contact. The device includes a substrate and semiconductor layers (Semi 1, Semi 2, Semi 3, Semi 4). In this exemplary embodiment, the first electrical contact is formed on the top surface of a first semiconductor layer region that is etched to expose a sublayer for disposing the second electrical contact. In this example, the multilayer semiconductor device further includes a third electrical contact (Contact 3) located on the underside of the substrate. A three-terminal device composed of Contact 1, Contact 2, and Contact 3 can function as a vertical heterojunction bipolar transistor or a vertical conduction FET switch.

[0263] Figure 66 shows a representational cross-sectional view of an in-plane MSM conduction device comprising multiple unit cells of the mesa structure device shown in Figure 65. Unit cell Λcell are laterally adjacently disposed. The cells may form elongated fingers in the plane of the drawing.

[0264] FIG. 67 shows a representational cross-sectional view of a multi-electrical terminal device having multiple semiconductor layers (Semi1, Semi2, Semi3, Semi4). The device has a first electrical contact (Contact1) formed on a first mesa structure (Mesa1). A second electrical contact (Contact2) is spaced both horizontally and vertically from the first electrical contact and is formed on a second mesa structure (Mesa2). A third electrical contact (Contact3) is spaced both horizontally and vertically from the second electrical contact. In this exemplary embodiment, the first electrical contact is formed on an initial top surface of a semiconductor layer region (Semi4) that is etched to expose a first sublayer (Semi3) for disposing the second electrical contact. The first sublayer is further etched to expose another second sublayer (Semi2) for disposing the third electrical contact. In this example, the multi-electrical terminal device further comprises a fourth electrical contact (Contact4) located on the underside of the substrate. For electrically insulating substrates, the fourth electrical contact is optional.

[0265] FIG. 68A shows a representational cross-sectional view of a planar field effect transistor (FET) with source (S), gate (G), and drain (D) electrical contacts. The source and drain electrical contacts are formed on a semiconductor layer region (Semi1) formed on an insulating substrate. The gate electrical contact is formed on a gate layer formed on the semiconductor layer region. Mg x Ge y O y The material layer can be used in two different ways: x Ge y O yOne function of the layer is to provide the active conductive channel region Semi1 with a wider bandgap material than that used to form the gate layer. For example, the gate layer may itself be epitaxially formed on Semi1 (e.g., cubic gamma-Al2O3, MgO, or MgAl2O4) or may be substantially amorphous (e.g., amorphous Al2O3). Alternatively, Mg x Ge y O y The composition can also be used as the gate layer, for example, where the active channel Semi is a smaller bandgap material. The metals forming the S and D contacts are ideally ohmic, and the gate metal can be selected to control the threshold voltage of the FET.

[0266] Figure 68B shows a top view of the planar FET shown in Figure 68A, illustrating the distance D1 between the source and gate electrical contacts and the distance D2 between the drain and gate electrical contacts. Cross section BB shows a cross section according to Figure 68A. The distance D2>D1 can be used to control the breakdown voltage along channel Semi 1 between the G and D regions.

[0267] Figure 69A shows a representative cross-sectional view of a planar FET with a configuration similar to that shown in Figures 68A and 68B. In Figure 69A, a source electrical contact (S) is implanted (Implant 1) into the substrate through a semiconductor layer region (Semi 1), and a drain electrical contact (Implant 2) is implanted only in the semiconductor layer region. Using selective area ion implantation to spatially modify the conductivity of specific regions, such as the S and D regions, is advantageous for improving lateral contact to the channel layer Semi 1. It is expected that p-type and n-type conductivity regions can be achieved using a selection of ion implant species, such as Ga, Al, Li, and Ge. O implantation can also be used to create locally insulating compositions. An alternative to ion implantation is to use a diffusion process. In this case, material can be spatially formed on the surface of Semi 1 and then driven into the interior of Semi 1 by a thermally activated diffusion process. For example, a Li-based glass can be deposited and the Li can be driven into Semi 1 by an annealing process in an inert environment. Such a rapid thermal annealing process is possible.

[0268] Figure 69B shows a top view of the planar FET shown in Figure 69A. Section BB shows the cross section according to Figure 69A.

[0269] Figure 70 shows a top view of a planar FET comprising multiple interconnected unit cells of the planar FET shown in Figure 68A or Figure 69A. cell is shown, and in this embodiment a three terminal device is shown.

[0270] Figure 71 shows a process flow diagram for forming a conductive device including a conformal semiconductor layer region regrown on the exposed, etched mesa sidewalls. First, a semiconductor device is formed having a substrate (SUB) and an epitaxially formed semiconductor layer region (EPI). This semiconductor layer region is then etched to leave a mesa-structured semiconductor layer region. An additional conformal semiconductor layer region (Semi2) is then grown on the mesa structure, which can then be planarized in a subsequent planarization step, if desired. For example, the conformal coating Semi2 can be another oxide deposited by atomic layer deposition. Semi2 can be used as a passivation region or as an active region to form a FET.

[0271] It will be understood that the terms "comprise" and "include," as well as their derivatives (e.g., comprises, comprising, includes, including), as used herein, are to be construed as inclusive of the features referred to by the term, and do not imply the exclusion of the presence of additional features, unless otherwise stated or implied.

[0272] The reference to any prior art in this specification is not, and should not be construed as, an acknowledgment of any form of suggestion that such prior art forms part of the common general knowledge.

[0273] Reference has been made in detail to embodiments of the disclosed invention, one or more examples of which are illustrated in the accompanying drawings. Each example is provided by way of explanation of the present technology, not by way of limitation. Indeed, while the specification has been described in detail with reference to specific embodiments of the present invention, it will be appreciated that those skilled in the art, upon gaining an understanding of the foregoing, will readily conceive of alternatives, variations, and equivalents to these embodiments. For example, features illustrated or described as part of one embodiment may be used with another embodiment to yield a still further embodiment. It is therefore intended that the present subject matter cover all such modifications and variations that come within the scope of the appended claims and their equivalents. These and other modifications and variations to the present invention can be practiced by those skilled in the art without departing from the scope of the present invention, which is particularly set forth in the appended claims. Moreover, those skilled in the art will appreciate that the foregoing description is illustrative only and is not intended to limit the invention.

Claims

1. 1. A semiconductor structure comprising: a substrate comprising a substantially single crystalline substrate material; Mg on the substrate ｘ Ge １－ｘ O ２－ｘ (wherein the value of x is 0≦x<1), The Mg ｘ Ge １－ｘ O ２－ｘ the epitaxial layer of the semiconductor structure having a crystal symmetry that matches the substrate material.

2. 10. The semiconductor structure of claim 1, wherein the crystalline structure of said substrate material has a lattice mismatch of 10% or less with said epitaxial layer.

3. The Mg ｘ Ge １－ｘ O ２－ｘ 2. The semiconductor structure of claim 1, wherein said epitaxial layer is an active region of said semiconductor structure.

4. 10. The semiconductor structure of claim 1, wherein the crystal symmetry of the substrate material is cubic or tetrahedral.

5. The substrate material is MgO (001), MgGa ２ O ４ (001), MgAl ２ O ４ 5. The semiconductor structure of claim 4, comprising LiF(001), or LiF(100).

6. The semiconductor structure includes the substrate and the Mg ｘ Ge １－ｘ O ２－ｘ 6. The semiconductor structure of claim 5, further comprising a buffer layer between said epitaxial layer and said first epitaxial layer, said buffer layer comprising MgO.

7. The substrate material is β-Ga ２ O ３ (100), LiAlO ２ (100), ZrO ２ (100), LiNbO ３ (001), LiTaO ３ (001), Fe ２ O ３ (100), BN(001), LiGaO ２ (001), TiO ２ (001), AlN (100), SiC (100), BaF ２ (100), BN(100), or CdWO ４ The semiconductor structure of any one of claims 1 to 4, comprising (001).

8. x=2 / 3, and the Mg ｘ Ge １－ｘ O ２－ｘ The epitaxial layer of Mg has cubic symmetry and Fd3m space group. ２ GeO ４ Contains, or x=2 / 3, and the Mg ｘ Ge １－ｘ O ２－ｘ The epitaxial layer of Mg has orthorhombic symmetry and Pnma space group. ２ GeO ４ Contains, or x=1 / 2, and the Mg ｘ Ge １－ｘ O ２－ｘ The epitaxial layer of MgGeO has monoclinic symmetry and C2 / c space group. ３ 10. The semiconductor structure of claim 1, comprising:

9. The Mg ｘ Ge １－ｘ O ２－ｘ 10. The semiconductor structure of claim 1, wherein is a direct bandgap material.

10. The Mg ｘ Ge １－ｘ O ２－ｘ 10. The semiconductor structure of claim 1, wherein is an indirect bandgap material.

11. The Mg ｘ Ge １－ｘ O ２－ｘ MgGeO has the C2 / c space group ３ 10. The semiconductor structure of claim 1, wherein:

12. The Mg ｘ Ge １－ｘ O ２－ｘ is a direct bandgap p-type material, containing Ga dopants, Al dopants, Li ＋ Dopant, or N ３＋ The semiconductor structure of claim 1 , comprising a dopant.

13. The Ga dopant is the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ The semiconductor structure of claim 12 located at a Ge site of the crystalline structure.

14. The Mg ｘ Ge １－ｘ O ２－ｘ 10. The semiconductor structure of claim 1, wherein is a direct bandgap n-type material and comprises a Ga dopant or an Al dopant.

15. The Ga dopant is the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ 15. The semiconductor structure of claim 14, located at Mg sites of the crystal structure.

16. The Al dopant is the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ The semiconductor structure of claim 12 located at a Ge site of the crystalline structure.

17. The Al dopant is the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ 15. The semiconductor structure of claim 14, located at Mg sites of the crystal structure.

18. The Li ＋ The dopants are the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ The semiconductor structure of claim 12 located at a Ge site or an Mg site of the crystal structure.

19. The Mg ｘ Ge １－ｘ O ２－ｘ But Ni ＋ The semiconductor structure of claim 1 , comprising a dopant.

20. The Ni ＋ The dopants are the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ 20. The semiconductor structure of claim 19, located at Mg sites of the crystal structure.

21. The N ３＋ The dopants are the corresponding undoped Mg ｘ Ge １－ｘ O ２－ｘ The semiconductor structure of claim 12 located at oxygen sites of the crystal structure.

22. The Mg ｘ Ge １－ｘ O ２－ｘ the epitaxial layer has a doped unit cell structure and a corresponding undoped unit cell structure; In the doped unit cell structure, the Mg ｘ Ge １－ｘ O ２－ｘ is an indirect bandgap p-type material, the doped unit cell structure includes a Ge atom at a first position occupied by Mg in the corresponding undoped unit cell structure; 2. The semiconductor structure of claim 1, wherein the doped unit cell structure includes a Mg atom in a second position occupied by Ge in the corresponding undoped unit cell structure.

23. 23. The semiconductor structure of claim 22, wherein in the doped unit cell structure, the Ge atoms are octahedrally bonded and the Mg atoms in the second positions are tetrahedrally bonded.

24. The Mg ｘ Ge １－ｘ O ２－ｘ 2. The semiconductor structure of claim 1, wherein the unit cell of is doped with an excess of Ge atoms or an excess of Mg atoms.

25. further comprising a superlattice, the superlattice comprising a unit cell having a first layer; The Mg ｘ Ge １－ｘ O ２－ｘ 2. The semiconductor structure of claim 1, wherein the epitaxial layer of is the first layer in the superlattice.

26. The unit cell of the superlattice is Mg ｙ Ge １－ｙ O ２－ｙ 26. The semiconductor structure of claim 25, further comprising a second layer of: wherein y ranges from 0 to 1 and x≠y.

27. 26. The semiconductor structure of claim 25, wherein the first layer is MgO having a Fm3m space group.

28. The unit cell of the superlattice further comprises a second layer, the second layer having a Mg ２ GeO ４ 28. The semiconductor structure of claim 27, wherein the semiconductor structure is made of

29. The unit cell of the superlattice is MgGa ２ O ４ or gamma-Ga ２ O ３ 26. The semiconductor structure of claim 25, further comprising a second layer of

30. The Mg ｘ Ge １－ｘ O ２－ｘ 4. The semiconductor structure of claim 3, wherein the epitaxial layer of is an active region of an optoelectronic device that is a light emitting diode or a photodetector, said optoelectronic device emitting or absorbing light having a wavelength in the range of 150 nm to 280 nm.

31. The Mg ｘ Ge １－ｘ O ２－ｘ 10. The semiconductor structure of claim 1, wherein the epitaxial layer is an active region of an electronic device.

32. The Mg ｘ Ge １－ｘ O ２－ｘ 10. The semiconductor structure of claim 1, further comprising a second epitaxial layer forming a heterostructure with the first epitaxial layer.

33. The second epitaxial layer is Zn ｘ Ge ｙ O ｚ , Zn ｘ Ga ｙ O ｚ , Al ｘ Ge ｙ O ｚ , Al ｘ Zn ｙ O ｚ , Al ｘ Mg ｙ O ｚ , Mg ｘ Ga ｙ O ｚ , Mg ｘ Zn ｙ O ｚ , or Ga ｘ O ｚ 33. The semiconductor structure of claim 32, comprising:

34. 10. A method of forming the semiconductor structure of claim 1, comprising: Mg ｘ Ge １－ｘ O ２－ｘ providing a substrate comprising a substantially single crystalline substrate material having a crystal symmetry compatible with the epitaxial layer of Co-depositing materials onto the substrate to form the Mg ｘ Ge １－ｘ O ２－ｘ (wherein the value of x is 0≦x<1), The method, wherein the material comprises at least two elements selected from Mg, Ge, and oxygen according to the value of x, and the Mg, Ge, and oxygen are supplied by a Mg source, a Ge source, and an activated oxygen source, respectively.

Images