US20180269351A1

US20180269351A1 - Indium gallium nitride light emitting devices

Info

Publication number: US20180269351A1
Application number: US15/986,253
Authority: US
Inventors: Michael R. Krames; Mark P. D'Evelyn; Akinori Koukitu; Yoshinao Kumagai; Hisashi Murakami
Original assignee: Soraa Inc
Current assignee: Ecosense Lighting Inc
Priority date: 2012-10-16
Filing date: 2018-05-22
Publication date: 2018-09-20
Also published as: JP2014090169A; US20140103356A1; US9978904B2

Abstract

InGaN-based light-emitting devices fabricated on an InGaN template layer are disclosed.

Description

CROSS REFERENCE TO RELATED APPLICATIONS

The present application claims the benefit under 35 U.S.C. § 119(c) of U.S. Provisional Application No. 61/714,693 filed on Oct. 16, 2012, and is a continuation of U.S. application Ser. No. 14/054,234, both of which are incorporated herein by reference.
Certain embodiments of the present application are related to material disclosed in U.S. Pat. No. 8,482,104, and U.S. Publication No. 2012/0091465, each of which is incorporated by reference in its entirety.

FIELD

The disclosure relates to InGaN-based light-emitting devices fabricated on an InGaN template layer.

BACKGROUND

Today's state-of-the-art visible-spectrum light-emitting diodes (LEDs) and laser diodes (LDs) in the ultraviolet to green (380 nm to 550 nm) regime are based on InGaN active layers grown pseudomorphic to wurtzite GaN. This is true whether the growth substrate is GaN itself, or a foreign substrate such as sapphire or SiC, since in the latter cases GaN-based template layers are employed. Because the lattice constants of GaN and InN are significantly different, InGaN grown pseudomorphically on GaN substrates or layers has significant stress, where the magnitude increases as the In/Ga ratio in the InGaN layer increases.
The built-in stress within the InGaN active layers can make it difficult to achieve high quality material and good device operation. Obtaining high quality material and good device operation becomes progressively more difficult as the InN mole fraction increases, which is a requirement for longer wavelength devices. In addition, for c-plane grown devices, increasing the InN mole fraction also increases the built-in electric fields across the active layers due to spontaneous and piezoelectric polarization fields, reducing the overlap between electrons and holes and decreasing the radiative efficiency. Moreover, there is evidence that material breakdown occurs once the stress level becomes too high, resulting in so-called “phase separation” (see N. A. El-Masry, E. L. Piner, S. X. Liu, and S. M. Bedair, “Phase separation in InGaN grown by metalorganic chemical vapor deposition,” Appl. Phys. Lett., vol. 72, pp. 40-42, 1998). Phase separation is exhibited beyond a critical limit of a certain InN mole fraction combined with a certain layer thickness. Such a limit is commonly observed for InGaN layers of about 10% InN grown more than 0.2 μm thick, for example, resulting in “black” or “grey” wafers.
The use of substrates comprising non-polar (1-100), (11-20), and semi-polar planes of GaN can address some of the problems above. In particular, for certain growth planes, the combined spontaneous and piezoelectric polarization vector can be reduced to zero or near-zero, eliminating the electron-hole overlap problem prevalent in c-plane-based devices. Also, improved material quality with higher InN mole fraction can be observed, such as is demonstrated in semi-polar material, which has resulted in continuous-wave (cw) true-green laser diodes (LDs) (see Enya et al., “531 nm green lasing of InGaN based laser diodes on semi-polar (20-21) free-standing GaN substrates,” Appl. Phys. Express 2, 082101, 2009; J. W. Raring et al., “High-efficiency blue and true-green-emitting laser diodes based on non-c-plane oriented GaN substrates,” Appl. Phys. Express 3, 112101 (2010)). However, the performance of longer-wavelength devices grown on these structures still suffers considerably compared to that of their shorter-wavelength counterparts. Also, it is not clear that growth plane orientation would eliminate the material quality problems associated with strain. Indeed, recent characterization of semi-polar (Al,In,Ga)N heterostructures reveals the formation of a large density of misfit dislocations at heterointerfaces between AlGaN and GaN, for example (see A. Tyagi et al., “Partial strain relaxation via misfit dislocation generation at heterointerfaces in (Al,In)GaN epitaxial layers grown on semipolar (11-22) GaN free standing substrates,” Appl. Phys. Lett. 95, 251905, 2009). These dislocations are likely to act as non-radiative recombination centers, and these dislocations may also provide potential degradation mechanisms which may prevent long-life operation (e.g., as is necessary for applications such as solid-state lighting). Further, reported external quantum efficiencies vs. wavelength for LEDs generally show a strong reduction in external quantum efficiency with increasing InN mole fraction, which is often referred to as the “green gap,” regardless of growth plane orientation.

SUMMARY

Disclosed herein are light emitting devices. In one embodiment, the light emitting device is formed on a gallium- and indium-containing nitride substrate having an n-type layer overlying the substrate, and having an active layer overlying the n-type layer with a p-type layer overlying the active layer. In this specific embodiment, the gallium- and indium-containing nitride substrate comprises a thickness greater than 4 m and an InN composition greater than 0.5%.
In another embodiment, a light emitting device is formed of a substrate, an n-type layer overlying the substrate, an active layer overlying the n-type layer, and a p-type layer overlying the active layer. Each of the afbrementioned layers is characterized by an in-plane lattice constant greater, by at least 1%, than that of similarly oriented GaN.
In certain aspects, light emitting devices are provided comprising a gallium- and indium-containing nitride substrate; an n-type layer overlying the substrate; an active layer overlying the n-type layer, and a p-type layer overlying the active layer; wherein the gallium- and indium-containing nitride substrate comprises a thickness greater than 4 m and an InN composition greater than 0.5%.
In certain aspects, light emitting devices are provided comprising a substrate; an n-type layer overlying the substrate; an active layer overlying the n-type layer, and a p-type layer overlying the active layer, each of the n-type layer, the active layer, and the p-type layer is characterized by an in-plane lattice constant, wherein each of the in-plane lattice constants is greater, by at least 1%, than an in-plane lattice constant of similarly oriented GaN.
In certain aspects, methods of fabricating a light emitting devices are provided comprising providing a substrate; selecting an InN composition; fabricating an InGaN template by growing an InGaN epitaxial layer having the selected InN composition on the substrate by hydride vapor phase epitaxy; and growing an optoelectronic device structure on the InGaN template.
Further details of aspects, objectives, and advantages of the disclosure are described below and in the detailed description, drawings, and claims. Both the foregoing general description of the background and the following detailed description are exemplary and explanatory, and are not intended to be limiting as to the scope of the claims.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a chart 100 showing variation in energy bandgap vs. basal-plane lattice constant.

FIG. 2 is a chart 200 showing variation in external quantum efficiency vs. emission wavelength for several light emitting diodes based on InGaN or AlInGaP active layers (from Denbaars, “Fundamental Limits to Efficiency of LEDs,” DOE Solid State Lighting Workshop, Raleigh, N.C., Feb. 2-4, 2010.)

FIG. 3 shows a schematic view 300 of the InGaN HVPE reactor, which consists of the source zone and the deposition zone, according to some embodiments.

FIG. 4 presents a photograph 400 of a susceptor showing a (000-1) GaN substrate after InGaN growth by HVPE.

FIG. 5 presents a graph 500 showing solid composition of InGaN as a function of growth temperature, according to certain embodiments.

FIG. 6 is a chart 600 showing growth rate and solid composition percentage as a function of an input partial pressure of group III sources GaCl₃and InCl₃, according to certain embodiments.

FIG. 7 shows an SEM image 700 of an In_0.16Ga_0.84N epitaxial layer grown on a (000-1) GaN substrate, according to certain embodiments.

FIG. 8 depicts cross-sectional maps 800 showing energy dispersive X-ray spectroscopy (EDS) data of InGaN on a (000-1) GaN substrate, according to certain embodiments.

FIG. 9 shows a plot 900 of an X-ray reciprocal space map (RSM) of (−1-14) diffraction for InGaN on a (000-1) GaN substrate grown by HVPE, according to certain embodiments.

FIG. 10 shows structure 1000 for thick InGaN growth on GaN, according to certain embodiments.

FIG. 11 depicts a comparison chart 1100 showing relative strain for optoelectronic device layer structures on GaN templates as compared with relative strain of optoelectronic device layer structures on InGaN templates, according to certain embodiments.

DETAILED DESCRIPTION

The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawings will be provided upon request and payment of the necessary fee.
Today's state-of-the-art visible-spectrum light-emitting diodes (LEDs) and laser diodes (LDs) in the ultraviolet to green (380 nm to 550 nm) regime are based on InGaN active layers grown pseudomorphic to wurtzite GaN. FIG. 1 exemplifies this situation. Chart 100 shows the variation in energy bandgap vs. basal-plane lattice constant. Pseudomorphic strained-to-GaN curve 102 is shown in juxtaposition to the relaxed InGaN curve 104. Moreover, external quantum efficiencies for LEDs decreases with increasing InN mole fraction, regardless of growth plane orientation. This is depicted in chart 200 of FIG. 2 that shows variations in external quantum efficiency vs. emission wavelength for several light emitting diodes.
In this disclosure, the above problems are circumvented by fabricating InGaN-based light-emitting devices on an InGaN template layer rather than on a layer whose lattice constant is pseudomorphic to GaN. Devices fabricated using this technique exhibit a lower strain mismatch between the template layer and device layers, which results in improved optical performance (e.g., via reduced polarization fields) as well as improved reliability (e.g., resulting from higher crystalline quality). Moreover, in accordance with this disclosure, the long-wavelength range of high-performing light-emitting devices can be extended.
When carrying out certain embodiments of this disclosure, a seed substrate is placed into a reaction chamber for hydride vapor phase epitaxy deposition of InGaN, to form an InGaN template. FIG. 3 shows a schematic view 300 of an InGaN HVPE reactor that consists of a source zone and a deposition zone (K. Hanaoka, H. Murakami, Y. Kumagai and A. Koukitu, “Thermodynamic analysis on HVPE growth of InGaN ternary alloy,” Journal Crystal Growth, vol. 318 (2011) 441-445). In the source zone, two boats for gallium metal and indium metal are located, and the mixture gas of Cl₂and IG (Inert Gas such as nitrogen, helium and argon), or the mixture gas Cl₂, IG, and hydrogen feeds into the source boats. In FIG. 5, the reactions representing formation of GaCl₃, InCl₃, and NH₃, are identified as reactions 302, 304, and 306, respectively. The precursors are deposited on substrate 308 and unreacted species exhausted 310 from the HVPE growth apparatus. The group III precursors, GaCl₃and InCl₃, are generated by reactions (1) and (2) (below), and are transported into the deposition zone. In substitution for the above reactions, the vapor pressures of GaCl₃and/or InCl₃vaporized from solid GaCl₃and InCl₃sources can be used as the group III precursors. Thus, the relevant reactions are:
Ga(s)+1.5Cl₂→GaCl₃ (1)
In(s)+1.5Cl₂→InCl₃ (2)
In the deposition zone, the group III precursors and NH₃as the group V precursor are mixed to deposit InGaN alloy on a substrate. The reactions in the deposition zone are expressed by reactions (3) and (4), where the InGaN alloy comprises GaN(alloy) and InN(alloy). The reactions are:
GaCl₃+NH₃→GaN(alloy)+3HCl (3)
InCl₃+NH₃→InN(alloy)+3HCl (4)
A substantially indium-free seed substrate may be provided. The substrate may comprise one of sapphire, silicon carbide, gallium arsenide, silicon, germanium, a silicon-germanium alloy, MgAl₂O₄spinel, ZnO, BP, ScAlMgO₄, YFeZnO₄, MgO, Fe₂NiO₄, LiGa₅O₅, Na₂MoO₄, Na₂WO₄, In₂CdO₄, LiAlO₂, LiGaO₂, Ca₅La₂(PO₄)₆O₂, lithium aluminate, gallium nitride, indium nitride, or aluminum nitride. The substrate may have a wurtzite crystal structure and the surface orientation may be within 5 degrees, within 2 degrees, within 1 degree, within 0.5 degree, within 0.2 degree, within 0.1 degree, within 0.05 degree, within 0.02 degree, or within 0.01 degree of the (0 0 0 1)+c plane, the (0 0 0 −1) −c plane, the {1 0 −1 0} m-plane, the {1 1 −2 0} the a-plane, or ta (h k i l) semi-polar plane, where l and at least one of h and k are nonzero and i=−(h+k). In a specific embodiment, the crystallographic orientation of the substrate is within 5 degrees, within 2 degrees, within 1 degree, within 0.5 degree, within 0.2 degree, within 0.1 degree, within 0.05 degree, within 0.02 degree, or within 0.01 degree of {1 0 −1 ±1}, {1 0 −1 ±2}, {1 0−1 ±3}, {1 1 −2 ±2}, {2 0−2 ±1}, {3 0 −3 ±1}, {3 0 −3 ±2}, {2 1 −3 ±1}, or {3 0 −3 ±4}. The substrate may have a cubic crystal structure and the surface orientation may be within 5 degrees, within 2 degrees, within 1 degree, within 0.5 degree, within 0.2 degree, within 0.1 degree, within 0.05 degree, within 0.02 degree, or within 0.01 degree of (1 1 1), (−1 −1 −1), {0 0 1}, or {1 1 0}. The substrate may have a diameter greater than about 5 millimeters, greater than about 10 millimeters, greater than about 15 millimeters, greater than about 25 millimeters, greater than about 40 millimeters, greater than about 70 millimeters, greater than about 90 millimeters, greater than about 140 millimeters, or greater than about 190 millimeters.
In a specific embodiment, the seed substrate is wurtzite GaN and the growth surface is substantially N-face GaN. The seed substrate may have a dislocation density below 10⁷cm⁻², below 10⁶cm⁻², below 10¹cm⁻², below 10⁴cm⁻², below 10³cm⁻², or below 102 cm⁻². The seed substrate may have a stacking-fault concentration below 10³cm⁻¹, below 10²cm⁻¹, below 10 cm⁻¹or below 1 cm⁻¹. The seed substrate may have an optical absorption coefficient below 100 cm⁻¹, below 50 cm⁻¹, below 5 cm⁻¹, below 2 cm⁻¹, below 1 cm⁻¹, or below 0.3 cm⁻¹at wavelengths between about 390 nm and about 700 nm. The seed substrate may have an optical absorption coefficient below 100 cm⁻¹, below 50 cm⁻¹, below 5 cm⁻¹, below 2 cm⁻¹, below 1 cm⁻¹, or below 0.3 cm⁻¹at wavelengths between about 700 nm and about 3077 nm and at wavelengths between bout 3333 nm and about 6667 nm. The top surface of the seed substrate may have an X-ray diffraction ω-scan rocking curve full-width-at-half-maximum (FWHM) less than about 300 arc sec, less than about 200 arc sec, less than about 100 arc sec, less than about 50 arcsec, less than about 40 arcsec, less than about 30 arcsec, less than about 20 arcsec, or less than about 10 arcsec for the lowest-order symmetric and non-symmetric reflections. The top surface of the seed substrate may have been prepared by chemical mechanical polishing and may have a root-mean-square surface roughness less than 1 nanometer, less than 0.5 nanometer, less than 0.2 nanometer, or less than 0.1 nanometer, for example, as measured by atomic force microscopy over an area of at least 10 micrometers by 10 micrometers. In certain embodiments, the crystallographic orientation of the top (growth) surface is within about 0.1 degree of the (000-1) N-face. In certain embodiments, the crystallographic orientation of the top surface is miscut from (000-1) N-face by between about 0.1 and about 10 degrees toward a {10-10} m-plane and is miscut by less than about 0.5 degrees towards an orthogonal {11-20} a-plane. In certain embodiments, the crystallographic orientation of the top surface is miscut from the (000-1) N-face by between about 0.1 and about 10 degrees toward a {11-20} a-plane and is miscut by less than about 0.5 degrees towards an orthogonal {10-10} m-plane. In certain embodiments, the crystallographic orientation of the top surface is miscut from the (000-1) N-face by between about 0.1 and about 10 degrees toward a {10-10} m-plane and is miscut by between about 0.1 and about 10 degrees towards an orthogonal {11-20} a-plane.
In certain embodiments an indium-containing nitride layer is deposited onto the substrate prior to initiating HVPE bulk growth on the substrate. An indium-containing nitride layer may be deposited at relatively low temperature by at least one of molecular beam epitaxy, hydride vapor phase epitaxy, metalorganic chemical vapor deposition, and atomic layer epitaxy. In certain embodiments, alternating layers of a higher-indium composition and a lower-indium composition are deposited. The alternative compositions may be deposited by periodic changes in the gas phase composition above a stationary substrate or by physical transport of the substrate between regions of the reactor providing higher-indium and lower-indium growth environments. Deposition of high-crystallinity layers at low temperature may be facilitated by providing one or more hot wires to assist in the decomposition of gas-phase precursor species, by a plasma, or by similar means. Further details are described in U.S. Publication No. 212/0199952, which is incorporated by reference in its entirety. In certain embodiments, partial or full relaxation of an indium-containing nitride layer on the substrate is performed prior to initiating HVPE bulk growth on the substrate. In certain embodiments, the substrate or a layer deposited thereupon is patterned to facilitate atom transport along glide planes to form misfit dislocations. If desired, a pattern, for example to provide stripes, bottom pillars, holes, or a grid, is formed on the substrate or on an epitaxial layer on the substrate by conventional photolithography or by nanoimprint lithography. Generation of misfit dislocations may be facilitated by roughening the growth surface before deposition, for example, by deposition of nano-dots, islands, ion bombardment, ion implantation, or by light etching. Misfit dislocations may also preferentially be formed by modifying the lattice parameter of the substrate near an epitaxial layer by a process such as atomic diffusion, atomic doping, ion implantation, and/or mechanically straining the substrate. Generation of misfit dislocations may also be facilitated by deposition of a thin layer of Al_xGa_(1-x)N, for example, thinner than about 10 nanometers to 100 nanometers, followed by annealing to a temperature between about 1000 degrees Celsius and about 1400 degrees Celsius in an ammonia-rich atmosphere. Further details are described in U.S. Publication No. 2012/0091465, which is incorporated by reference in its entirety. In certain embodiments, the indium-containing nitride layer is relaxed, having a-axis and c-axis lattice constants within 0.1%, within 0.01%, or within 0.001% of the equilibrium lattice constants for the specific indium-containing nitride composition.
Recent results using an N-face (000-1) GaN substrate are disclosed. A mirror-like surface can be obtained. However, in some cases there may be cracks on the surface of the InGaN layer caused by the expansion stress. A photograph 400 of the susceptor showing the (000-1) GaN substrate after InGaN growth is shown below in FIG. 4.
FIG. 5 depicts a graph 500 showing the solid composition of InN in InGaN as a function of the growth temperature. The solid composition of InN in InGaN decreases with increasing temperature. In FIG. 5, previous data is also displayed for InGaN grown by using solid sources, GaCl₃and InCl₃(N. Takahashi, R. Matsumoto, A. Koukitu and H. Seki, “Vapor phase epitaxy of InGaN using InCl₃, GaCl₃and NH₃sources,” Jpn. Appl. Phys., 36 (1997) pp. L601-L603). The content of the solid compositions of the present samples are similar to that of the previous samples. However, photoluminescence of the previous samples was not observed due to the high impurity concentration in solid sources. In the figure, P_IIIand P_Vindicate the input partial pressures of the group III source (P_In+P_Ga) and the group V source (P_NH3).
FIG. 6 depicts a chart 600 showing the growth rate (602) and the solid composition (604) as a function of the input partial pressure of the group III sources. The growth rate linearly increases with increasing input pressure, while the solid composition, InN in InGaN, is constant at about 0.16.
FIG. 7 shows a scanning electron microscope (SEM) image 700 of an In_0.16Ga_0.84N epitaxial layer grown on a (000-1) GaN substrate of the methods of the present disclosure. The thickness of the epitaxial layer is about 800 nm and the growth rate is about 8.5 μm/h. Because the thickness of InGaN layer is large (about 800 nm), the expansion stress may have caused the cracks on the InGaN surface. Some methods serve to prevent crack formation, resulting in a high quality thick InGaN layer. One such method is the “multi-step grading method.” For example, to grow a crack-free In_xGa_1-xN film, where x is greater than about 0.05, 0.1, 0.15, 0.2, 0.25, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, or 1.0, it may be desirable to grow a series of layers of graded composition. For example, a graded layer may consist of successive 1-micron thicknesses of In_0.2xGa_1-0.2xN, In_0.4xGa_1-0.4xN, In_0.6xGa_1-0.6xN, In_0.8xGa_1-0.8xN, followed by a thicker In_xGa_1-xN film. In certain embodiments, the thickness of the intermediate-composition layers lies between 10 nanometers and 5 microns. In certain embodiments, the composition is graded continuously rather than as discrete layers with each layer having a fixed InGaN composition.
FIG. 8 depicts cross-sectional maps 800 showing energy dispersive X-ray spectroscopy (EDS) data of InGaN on −C GaN, using the same sample shown in FIG. 7. Measurement conditions are as follows: accelerating voltage: 20 kV, emission current: 20 μA, spot size: 7, WD: 15 mm, wavelength resolution: 130 eV to 140 eV. From the mapping data, it can be seen that the InGaN layer exhibits a relatively homogeneous solid composition.
FIG. 9 presents a plot 900 showing the reciprocal space map (RSM) of (−1-14) diffraction for InGaN on −C GaN grown by HVPE (same sample shown in FIG. 7). FIG. 9 shows that the InGaN layer is fully relaxed with respect to the GaN substrate because the grown thickness of InGaN is about 800 nm. The lattice constant of InGaN layer is estimated to be a=0.32657 nm and c=0.52712 nm. Assuming that the lattice constant of InGaN follows Vegard's law of linear function between InN and GaN, solid composition x in In_xGa_1-xN should be 0.216 (estimation from the a-axis lattice constant). Since the value is somewhat different from that estimated from the c-axis lattice constant of InGaN (x=0.152), residual strain and/or some impurities are thought to be included in the layer.
FIG. 10 shows a new structure 1000 for thick InGaN growth without cracks grown using a graded structure, as described above. This structure can be grown using a HVPE reactor described in the present disclosure. The growth rate of InGaN is very fast and it is easy to grow 1 μm-thick multi-step layers (see layers). In certain embodiments, this structure is deposited onto a substrate 1004. A GaN layer 1002 deposited upon the substrate may be patterned to facilitate atom transport along glide planes to form misfit dislocations, as described in U.S. Patent Application Publication. No. 2012/0091465, which is incorporated by reference in its entirety.
There are several ways to provide the final InGaN template suitable for subsequent epitaxial growth and device fabrication. In one embodiment, the InGaN layer is left on its seed substrate. If necessary, a polishing step (e.g., chemical-mechanical polishing) may be provided to prepare the surface for epitaxial growth. In another embodiment, the N-face InGaN material is grown very thick, to several hundred microns or more, to provide a free-standing InGaN crystal or boule. In certain embodiment, the indium-containing nitride crystal or boule is relaxed, having a-axis and c-axis lattice constants within 0.1% within 0.01%, or within 0.001% of the equilibrium lattice constants for the specific indium-containing nitride composition. The InGaN boule may be machined by well-known wafering techniques to provide one or more InGaN wafers. Steps for wafering may include wire-sawing, lapping, polishing, and chemical cleaning steps. A typical wafer thickness may be between 100 and 750 microns, more typically between 250 and 350 microns. In certain embodiments, an InGaN boule or wafer is used as a seed for further bulk growth, for example, by HVPE, by an ammonothermal technique, by a flux technique, or by a solution growth technique. In the case of polar or semi-polar orientations, either the +c or −c orientation surface may be selected preferentially for final preparation for epitaxial growth (that is, the “backside” need not have all the same polishing and or cleaning steps). In certain embodiments it is preferable to remove the original seed substrate at some point in the process.
The InGaN template is provided for device fabrication as follows. Typically, the template is placed in a metal-organic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE) reactor for deposition of device quality layers. This is accomplished by reacting tri-methyl (-ethyl) indium, tri-methyl (-ethyl) gallium, and/or tri-methyl (-ethyl) aluminum in the presence of ammonia (NH₃) in nitrogen and/or hydrogen carrier gas, at elevated temperatures (500° C. to 1000° C.). The ratio of precursors is selected in order to provide a close lattice match to the InGaN template, and can be controlled by monitoring of mass flow controllers (MFCs), or growth temperature, or by in situ wafer bow measurements. Typically, an n-type InGaN layer is deposited first. The n-type characteristic may be achieved unintentionally (i.e., background oxygen) or intentionally by introducing donor species (e.g., Si via silane). Then, an InGaN or AlGaInN active layer structure is deposited. A common active layer structure is a multiple-quantum-well structure. The light-emitting layers are grown with a higher InN mole fraction than the template and the barrier layers (if any) separating the light-emitting layers. The active region may be undoped, n-type doped, or p-type doped. The p-type characteristic is typically achieved by introducing acceptor species (e.g., Mg via Cp₂Mg). After the active layers are deposited, a spacer layer may be provided, before growing the p-type layers. The p-type layers may include an InGaN, GaN, AlGaN, or AlGaInN “electron blocking” layer to assist in carrier confinement in the active layers during device operation. After the electron blocking layer, a p-type InGaN layer is grown, followed by an InGaN or GaN p+ contact layer. Typically, after MOCVD growth, the entire structure is annealed at elevated temperature to activate the p-type layers through redistribution of hydrogen which has been grown into the crystal.
A device structure is illustrated schematically in comparison chart 1100 of FIG. 11. The top illustration (see features 1102, 1104, and 1106) indicates a device structure where the growth template is a GaN template (on, for example, sapphire, silicon carbide, or silicon) or a GaN substrate. The lattice constant along the in-plane direction is that of GaN and the epitaxial layers are grown pseudomorphically to the GaN template or substrate. The device layers grown thereon include InGaN active (e.g., quantum well) layers and AlGaN (e.g., electron blocking) layers. The native in-plane lattice constants of these tertiary alloys are larger (InGaN) or smaller (AlGaN) than that of GaN, resulting in strong compressive or tensile stress, respectively, in these pseudomorphically grown layers. For long wavelength devices, significant InN is required, resulting in severe compressive stress that can result in poor quality crystal growth and consequently, poor device performance and reliability.
Referring to the bottom illustration (see features 1108 1110, and 1112), and in accordance with the present disclosure, the InGaN template is chosen to provide larger in-plane lattice constant than that of GaN, providing for a strain-compensated design, wherein compressive and tensile stress are balanced out so that the entire crystal structure has strong mechanical integrity. This results in more design freedom, higher crystal quality, and better performance and reliability in devices.
In certain embodiments, the active layer is designed so that the InGaN template is substantially transparent with respect to the wavelength(s) of light emitted by the active layer. For example, if the composition of the InGaN substrate is specified by In_yGa_1-yN, where 0.05 5≤y≤1, and the composition of the active layer is designated by In_xGa_1-xN, where 0.05≤x≤1, x may be chosen to be greater than y by at least about 0.01, at least about 0.02, or at least about 0.05. The resulting compressive strain may be compensated by tensile strain associated with GaN, AlN, AlInGaN, or In_zGa_1-zN, where 0.0≤z≤y, barrier layers, electron-blocking layers, p-type layers, and the like. Of course, many quaternary (AlInGaN) compositions may be suitable for selecting a strain state while providing desired bandgap engineering for the device layer stack.
In certain embodiments, the InGaN template is conductive and a vertical light-emitting device may be fabricated, in which case an ohmic contact electrode is made to the back of the InGaN template. If a lateral or flip-chip device is desired, the n-type electrode is provided after etching through the p-type and active layers down to the n-type epitaxial or n-type InGaN template layers. Suitable n-type ohmic contact metallizations include Ti and Al, and combinations thereof. The p-type electrode is provided on the p+ contact layer. Suitable p-type ohmic contact metallizations include Ni, Au, Pt, Pd, and Ag, and combinations thereof. For light-emitting diodes (LEDs), reflective metallizations are sometimes preferred.
In the case of LEDs, extraction features may be incorporated into one of more exposed surfaces of the wafer. Extraction features may include roughened aspects as well as ordered texturing, including photonic crystal structures. After metal patterning and extraction feature implementation, the wafer may be diced into multiple LED chips, by means such as laser scribe and break, diamond tool scribe and break, sawing, etc. After dicing, further extraction features may be incorporated into the chip side surfaces, by means such as wet chemical etching.
In the case of laser diodes, the wafer may be laser scribed and broken into bars to provide mirror facets for multiple laser stripes. The mirror facets may be coated to provide high-reflectivity or anti-reflection properties, to optimize total laser light output. Then, the individual laser diode chips may be obtained in a further singulation step, which may include another scribe and break step.
After dicing, the light-emitting device is packaged into a suitable housing, and electrical connections are made to the n and p ohmic contacts electrodes. Thermal management is provided by providing a thermally conductive path from the active layers to the package housing. Optical encapsulation and/or lensing may be provided by primary optics comprising, for example, transparent materials such as silicones or glass. The final packaged device may be then incorporated into a solid state lighting product such as a lamp, luminaire, or light engine for displays.
The present disclosure is applicable to various crystal orientations. For example, the growth substrate maybe be polar (+c or −c plane), non-polar (a or m plane), or semi-polar (e.g., 11−2±2, 10−1±1, 20−2±1, 30−3±1, 30−3±2, etc.), which will provide for an InGaN template of similar orientation. This allows one to utilize the benefits of the chosen plane (e.g., polarization field reduction) for the InGaN device.
In the foregoing specification, the disclosure has been described with reference to specific embodiments thereof. It will, however, be evident that various modifications and changes may be made thereto without departing from the broader spirit and scope of the disclosure. For example, the above-described process flows are described with reference to a particular ordering of process actions. However, the ordering of many of the described process actions may be changed without affecting the scope or operation of the disclosure. The specification and drawings are, accordingly, to be regarded in an illustrative sense rather than restrictive sense. In the foregoing specification, the disclosure has been described with reference to specific embodiments thereof.

Claims

1. A light emitting device comprising:

a gallium- and indium-containing nitride substrate; an n-type layer overlying the substrate;

an active layer overlying then-type layer; and a p-type layer overlying the active layer;

wherein the gallium- and indium-containing nitride substrate comprises a thickness greater than 4 μm and an InN composition greater than 0.5%.

2. The light emitting device of claim 1, wherein the substrate is characterized by a wurtzite crystal structure and a crystallographic orientation of a large area surface within about 5 degrees of one of the (0001) +c-plane and the (000-1) −c-plane.

3. The light emitting device of claim 1, wherein the substrate is substantially crack-free.

4. The light emitting device of claim 1, wherein the gallium- and indium-containing nitride substrate is relaxed, having a-axis and c-axis lattice constants within 0.1% of the equilibrium lattice constants for a specific indium-containing nitride composition.

5. The light emitting device of claim 1, wherein the gallium- and indium-containing nitride substrate is relaxed, having a-axis and c-axis lattice constants within 0.01% of the equilibrium lattice constants for a specific indium-containing nitride composition.

6. The light emitting device of claim 1, wherein the gallium- and indium-containing nitride substrate is relaxed, having a-axis and c-axis lattice constants within 0.001% of the equilibrium lattice constants for a specific indium-containing nitride composition.

7. The light emitting device of claim 1, further comprising at least one of, a lamp, a luminaire, and a lighting system, wherein the light emitting device is incorporated into the lamp, the luminaire, or the lighting system.

8. A light emitting device comprising: a substrate;

an n-type layer overlying the substrate;

an active layer overlying the n-type layer; and a p-type layer overlying the active layer;

each of the n-type layer, the active layer, and the p-type layer is characterized by an in-plane lattice constant, wherein each of the in-plane lattice constants is greater, by at least 1%, than an in-plane lattice constant of similarly oriented GaN.

9. The light emitting device of claim 8, wherein the substrate is characterized by a wurtzite crystal structure and a crystallographic orientation of a large area surface within about 5 degrees of one of the (0001) +c-plane and the (000-1) −c-plane.

10. The light emitting device of claim 8, wherein the substrate is substantially crack-free.

11. The light emitting device of claim 8, wherein the substrate comprises gallium- and indium-containing nitride and is relaxed, having a-axis and c-axis lattice constants within 0.1% of the respective equilibrium lattice constants for a specific indium-containing nitride composition.

12. The light emitting device of claim 8, wherein the substrate comprises gallium- and indium-containing nitride and is relaxed, having a-axis and c-axis lattice constants within 0.01% of the respective equilibrium lattice constants for a specific indium-containing nitride composition.

13. The light emitting device of claim 8, wherein the substrate comprises gallium- and indium-containing nitride and is relaxed, having a-axis and c-axis lattice constants within 0.001% of the respective equilibrium lattice constants for a specific indium-containing nitride composition.

14. The light emitting device of claim 8, further comprising at least one of, a lamp, a luminaire, and a lighting system, wherein the light emitting device is incorporated into the lamp, the luminaire, or the lighting system.

15. A method of fabricating a light emitting

device comprising: providing a substrate;

selecting an InN composition;

fabricating an InGaN template by growing an InGaN epitaxial layer having the selected InN composition on the substrate by hydride vapor phase epitaxy; and

growing an optoelectronic device structure on the InGaN template.

16. The method of claim 15, wherein the growing of the InGaN epitaxial layer is characterized by a crystallographic orientation within about 5 degrees of [000-1].

17. The method of claim 15, wherein the substrate is substantially indium-free.

18. The method of claim 15, wherein the substrate comprises wurtzite GaN.

19. The method of claim 15, further comprising patterning the substrate to facilitate atom transport along glide planes to form misfit dislocations.

20. The method of claim 15, wherein growing the InGaN epitaxial layer having the selected InN composition on the substrate is performed by growing a series of graded layers having an intermediate InN composition.

21-25. (canceled)