Indium gallium nitride light-emitting diode with reduced distortion

Abstract

A method of forming an LED emitter includes providing a III-nitride layer on a substrate (310), the III-nitride layer having a planar upper surface, the method further includes providing discrete lateral growth regions on the upper surface, and selectively epitaxially growing a base region (1210) comprising In(x)Ga(l-x)N material on each discrete lateral growth region, each base region extending perpendicular to the upper surface, the method further including growing an In(x)Ga(l-x)N material on a portion of the base region (1210). providing a surface of In(x)Ga(1-x)N material, the surface having relaxed strain and characterized by a base lattice constant within 0.1% of its bulk relaxed value, the method further comprising epitaxially growing an LED region on the surface, the LED region comprising a light emitting layer of In(y)Ga(1-y)N material pseudomorphically matched to the surface of the In(x)Ga(1-x)N material, characterized by a lattice constant of the active region (1240) within 0.1% of the base lattice constant, and <x<0.2およびy> It is 0.3.

Description

Technical Field

[0001] Cross - Reference to Related Applications This application claims priority based on U.S. Provisional Patent Application Serial No. 63 / 001,221, filed on March 27, 2020, and the entire disclosure of this application is incorporated herein by reference.

[0002] Field of Disclosure The present disclosure relates to indium gallium nitride (InGaN) light - emitting diodes (LEDs) with reduced strain, and devices including such LEDs.

Background Art

[0003] Background LEDs such as micro - LEDs are candidates for display applications. In some cases, a combination of III - nitride NE LEDs emitting blue, green, and red light is desired.

Summary of the Invention

[0004] Summary Growing a long - wavelength III - nitride light - emitting layer can be difficult for several reasons, such as a large number of defects (point defects, extended defects) that reduce the internal quantum efficiency (IQE), and the difficulty of incorporating a large amount of indium (which is generally required for long - wavelength emission). These effects are generally exacerbated by strain. For example, the strain caused by the lattice mismatch between the GaN matrix and the InGaN light - emitting layer can lead to the incorporation / generation of defects and reduce In incorporation due to lattice tension.

[0005] What is needed is a micro - LED with reduced strain effects for long - wavelength emission. Various aspects of the present invention are summarized as follows.

[0006] Generally, in a first aspect, the present invention features a display light emitter, and the display light emitter Comprising a plurality of three sub-regions each corresponding to a sub-pixel of the display, the plurality of three sub-regions emit blue, green, and red light respectively during the operation of the display emitter, each sub-region includes a light-emitting region (e.g., a plurality of nanowires (NW)), the light-emitting region (e.g., each of the NWs) includes a base region, the base region has an InGaN composition containing at least 5% In, has a relaxed regrowth surface, the regrowth surface has a base lattice constant within 0.1% of its bulk value, the light-emitting region further includes an LED region regrown on the regrowth surface, and the LED region includes at least one light-emitting layer having an InGaN composition containing at least 10% In. The light-emitting layer is pseudo-lattice matched with the regrowth surface and has an active region lattice constant within 0.1% of the base lattice constant.

[0007] Embodiments may include one or more of the following features and / or features of other aspects.

[0008] The base region may be formed by hydride vapor phase epitaxy (HVPE). The LED region may be regrown by metalorganic chemical vapor deposition (MOCVD).

[0009] The base and active region lattice constants may be in-plane lattice constants. The base region may have a uniform composition equal to the InGaN composition.

[0010] The NWs may be grown on a flat III-nitride layer. The base region includes a lateral structure with a lateral dimension less than 300 nm, and relaxation of the base region material may occur inside the lateral structure.

[0011] Generally, in another aspect, the present invention features a method of forming a nanowire LED emitter, the method including providing a substrate, growing a flat GaN layer on the substrate, forming a mask having an opening on the substrate, and growing an In(x)Ga(1-x)N-based region in the opening (e.g., by HVPE) using InCl3 as a precursor, the base region extending vertically above the opening, the method further including providing a regrowth surface on a portion of the base region, the regrowth surface being relaxed and characterized by a base lattice constant within 0.1% of its bulk relaxation value, the method further including growing an LED region on the regrowth surface, the LED region having a light-emitting layer including a composition In(y)Ga(1-y)N that is pseudomorphic to the regrowth surface and characterized by an active region lattice constant within 0.1% of the base lattice constant, where 0.05 < x < 0.15 and y > 0.2.

[0012] Embodiments may include one or more of the following features and / or features of other aspects.

[0013] Generally, in another aspect, the present invention features a method of forming a nanowire LED emitter, the method including providing a substrate and forming a mask having an opening on the substrate, the mask having an upper surface, the method further including growing an In(x)Ga(1-x)N base material in the opening by a first growth method having a first growth parameter that substantially promotes lateral growth, the base material extending above the upper surface and laterally extending outside the opening above the upper surface, the method further including providing at least one regrowth surface on a portion of the base material, the at least one regrowth surface being at least partially relaxed, the method further including growing an LED region on the regrowth surface by a second growth method having a second growth parameter that does not substantially promote lateral growth, the LED region having a light-emitting layer including a composition In(y)Ga(1-y)N that is pseudomorphic to the at least one regrowth surface, where 0.05 < x < 0.2 and y > 0.3, and the misfit strain between the light-emitting layer and the regrowth surface is less than half of the misfit strain between the light-emitting layer and relaxed GaN.

[0014] Embodiments may include one or more of the following features and / or features of other aspects.

[0015] The at least one regrowth surface may be characterized by a base in-plane lattice constant within 0.5% of its bulk relaxation value. The first growth parameter may facilitate growth with a growth rate ratio of the lateral direction to the vertical direction of 1 or more (e.g., 2 or more, 5 or more, 10 or more, 100 or more). The second growth parameter facilitates growth with a growth rate ratio of the vertical direction to the lateral direction of 2 or more (e.g., 10 or more, 100 or more). The first method may be HVPE and the second method may be MOCVD.

[0016] Generally, in a further aspect, the present invention features a method of forming a light emitting diode (LED), the method including growing a buffer layer including gallium nitride (GaN) on a surface of a substrate, forming a mask having an opening on a surface of the buffer layer, and forming a base layer in the opening on the surface of the buffer layer, forming the base layer including epitaxially growing In(x)Ga(1-x)N (e.g., using hydride vapor phase epitaxy (HVPE) with a monohalide or trihalide precursor), the method further including providing a regrowth surface on the base layer, the regrowth surface having a relaxed crystal structure having a base lattice constant within 0.1% of a bulk value of the base lattice constant of In(x)Ga(1-x)N, the method further including epitaxially growing one or more additional layers on the regrowth surface to provide an active region of the LED, the one or more additional layers including at least one layer of In(y)Ga(1-y)N having an active region lattice constant within 0.1% of the base lattice constant of the regrowth surface and being pseudomorphic to the regrowth surface, the method further including forming one or more additional layers on the active region and processing these layers to provide the LED, where 0.05 < x < 0.15 and y > 0.2.

[0017] The implementation example may include one or more of the following features and / or features of other aspects. For example, the mask may include a plurality of spaced-apart openings, and the method may include simultaneously forming LEDs in each opening to provide a plurality of LEDs. The LEDs may be nanowire LEDs (NW LEDs). Each of the plurality of NW LEDs may be configured to emit light at a first peak wavelength λ1 within the visible spectrum. The plurality of NW LEDs may include a first plurality of NW LEDs grouped within a first region of the substrate, and the method may further include forming a second plurality of NW LEDs grouped within a second region of the substrate, each of the second plurality of NW LEDs being configured to emit light at a second peak wavelength λ2 within the visible spectrum, and the method may further include forming a third plurality of NW LEDs grouped within a third region of the substrate, each of the third plurality of NW LEDs being configured to emit light at a third peak wavelength λ3 within the visible spectrum, and λ1, λ2, and λ3 may be different from each other.

[0018] The opening of the mask may have a first dimension d1 in a first direction within the plane of the surface of the buffer layer, and the regrowth surface may have a second dimension d2 in the first direction, where d2 > d1.

[0019] The precursor may be a chloride precursor. The chloride precursor may be selected from the group consisting of GaCl, GaCl3, InCl, and InCl3.

[0020] The GaN layer may be a flat layer. The LED may extend perpendicular to the surface of the substrate above the opening.

[0021] The substrate may include a substrate material selected from the group consisting of sapphire, silicon, and GaN.

[0022] The one or more additional layers may include one or more quantum well (QW) layers configured to emit light having a peak wavelength λ in the visible portion of the electromagnetic spectrum.

[0023] One or more additional layers may be formed using metalorganic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE).

[0024] Generally, in a further aspect, the present invention features a light-emitting device that includes a GaN buffer layer on a surface of a substrate and a light-emitting diode (LED) extending from the GaN buffer layer. The LED includes an LED region having at least one layer of In(y)Ga(1-y)N and a base region between the LED region and the GaN buffer layer. The base region includes a layer of In(x)Ga(1-x)N having a regrown surface, and the regrown surface has a relaxed crystal structure having a base lattice constant within 0.1% of the bulk value of the base lattice constant of In(x)Ga(1-x)N. The layer of In(y)Ga(1-y)N within the LED region is pseudo-lattice matched to the regrown surface and has an active region lattice constant within 0.1% of the base lattice constant of the regrown surface, where 0.05 < x < 0.15 and y > 0.2.

[0025] Embodiments may include one or more of the following features and / or features of other aspects. For example, the LED region may include an active region having one or more quantum well (QW) layers supported by a layer of In(y)Ga(1-y)N. The LED region may further include a second InGaN layer supported by the active region.

[0026] At the surface of the buffer layer, the base region may have a first dimension d1 in a first direction in the plane of the surface of the buffer layer, and the regrown surface may have a second dimension d2 in the first direction, where d2 > d1.

[0027] The light-emitting device may include a mask on the surface of the buffer layer, and the NW LED is formed within an opening of the mask.

[0028] The layer of In(x)Ga(1-x)N within the base region may be formed using hydride vapor phase epitaxy (HVPE) with a monohalide or trihalide precursor.

[0029] The In(y)Ga(1-y)N layer in the LED field may be formed using metalorganic chemical vapor deposition (MOCVD) or molecular beam epitaxy (MBE).

[0030] In another aspect, the present invention features a display device including a plurality of NW LEDs, each of the NW LEDs extending from a GaN buffer layer, and the NW LEDs being spaced apart from each other. Each of the plurality of NW LEDs may be configured to emit light at a first peak wavelength λ1 within the visible spectrum. The plurality of NW LEDs includes a first plurality of NW LEDs grouped within a first region of the substrate, and the display device further includes a second plurality of NW LEDs grouped within a second region of the substrate, each of the second plurality of NW LEDs being configured to emit light at a second peak wavelength λ2 within the visible spectrum, and the display device further includes a third plurality of NW LEDs grouped within a third region of the substrate, each of the third plurality of NW LEDs being configured to emit light at a third peak wavelength λ3 within the visible spectrum, λ1, λ2, and λ3 being different from each other, and the first, second, and third regions constituting one pixel of the display device.

[0031] Generally, in a further aspect, the present invention features a method of forming a light emitting diode (LED), the method comprising providing a growth substrate having a textured upper surface, and selectively growing a first substrate over a region of the upper surface using a first growth technique to form features having at least one lateral dimension less than 500 nm, the first substrate including a relaxed In(x)Ga(1 - x)N layer, the method further comprising growing a second substrate over the first substrate using a second growth technique, such that at least one lateral dimension of the features extends greater than 1 micron upon growth of the second substrate, the second substrate including a layer of relaxed In(x)Ga(1 - x)N, the second growth technique being more favorable for lateral growth than vertical growth, the method further comprising growing an LED structure over the features using a third growth technique, the LED structure including at least one light emitting layer of In(y)Ga(1 - y)N that is pseudomorphic to the second substrate, the third growth technique being more favorable for vertical growth than lateral growth, 0.05 < x < 0.2 and y > 0.3, and the misfit strain between the light emitting layer and the second substrate being less than half of the misfit strain between the light emitting layer and relaxed GaN.

[0032] Embodiments may include one or more of the following features and / or features of other aspects. For example, the textured upper surface of the growth substrate may include a planar surface covered with a mask having openings with lateral dimensions less than 500 nm, and the selective growth occurs within the openings.

[0033] The features may be mesas. The first and second substrates may have in - plane lattice constants within 0.5% of their respective strain - free equilibrium values.

[0034] The light emitting layer may be flat, may have a wurtzite crystal structure, and the plane of the light emitting layer may be along the c - plane of the wurtzite structure.

[0035] The features may have an average dislocation density of less than 1E7 cm -1 ^ - 2. The first and second growth techniques may be hydride vapor phase epitaxy (HVPE), and the third growth technique may be metalorganic chemical vapor deposition (MOCVD).

[0036] The first growth technique may be HVPE, and the second and third growth techniques may be MOCVD.

[0037] The method may include preparing a characteristic surface to facilitate growth of the LED structure after growth of the second substrate and before growth of the LED structure.

[0038] One or more advantages will become apparent from the present disclosure, the drawings, and the claims.

Brief Description of the Drawings

[0039]

Figure 1A

Figure 1B

Figure 2A

Figure 2B

Figure 3A

Figure 3B

Figure 3C

Figure 3D

Figure 4A

Figure 4B

Figure 4C

Figure 5A

Figure 5B

Figure 5C

Figure 5D

Figure 6A

Figure 6B

Figure 7A

Figure 7B

Figure 7C

Figure 7D

Figure 8A

Figure 8B

Figure 8C

Figure 8D

Figure 8E

Figure 8F

Figure 8G

Figure 8H

Figure 8I

Figure 9

Figure 10A

Figure 10B

Figure 10C

Figure 10D

Figure 10E

Figure 11A

Figure 11B

Figure 11C

Figure 11D

Figure 11E

Figure 11F

Figure 12A

Figure 12B

Figure 13A

Figure 13B

Figure 13C

Figure 14

Figure 15A

Figure 15B

Figure 15C

Figure 16A

Figure 16B

Figure 16C

Figure 17A

Figure 17B

Figure 17C

Figure 17D

Figure 18A

Figure 18B

Figure 18C

Figure 18D

Figure 18E

Figure 18F

Figure 18G

Figure 18H

Figure 19A

Figure 19B

Figure 19C

Figure 19D

Figure 19E

Figure 19F

Figure 19G

[0040] Detailed Description LED Structure Some embodiments disclosed herein are nanowire (NW) LEDs, and each LED includes one or several NWs. The NW has a base region including an InGaN layer, an active region (light-emitting region) having an n-doped layer and at least one light-emitting InGaN layer, and an LED region including a p-doped layer. The base region may be grown by a first technique such as HVPE. The LED region may be regrown on the base region by a second technique such as MOCVD or MBE.

[0041] Figs. 1A - 1B respectively show examples of such LEDs 100 and 100'. In Fig. 1A, the active region 110 is flat and grows perpendicular to each NW axis 101. In Fig. 1B, the active region 111 and the p-GaN layer 121 are conformal with the base NW (this is called a core-shell structure). In both examples, the NWs 100 and 101 include an n-InGaN layer 130 grown on a GaN layer 140 on a substrate 150. A mask layer 160 is supported on the GaN layer 140 and includes an opening, and NWs are formed in this opening. The shape of the NW does not have to be vertical and may have inclined sidewalls, and the core-shell structure may have an active region along various facets (including horizontal, vertical, and inclined).

[0042] The base region can have an appropriate strain state, which may be fully relaxed, or may be partially relaxed. The base region functions as a base for growing the epitaxial layer of the LED region. Some embodiments are contrasted with conventional LEDs in that they include a high concentration of In in various layers surrounding the light emitting layer to reduce the strain of the light emitting layer.

[0043] Figures 2A - 2B contrast a conventional LED epitaxial stack 200 with an embodiment 201. In Figure 2A, most of the layers are GaN. Specifically, the stack of layers forming the LED 200 from bottom to top is the GaN buffer 210, the n-GaN layer 220, the GaN / InGaN underlying layer 230, and the GaN spacer 240 supporting the active region 250. The active region 250 is composed of In0.2GaN quantum well layers separated by GaN barrier layers. On top of the active region 250, the stack includes the GaN spacer 260, the AlGaN EBL 270, the p GaN layer 280, and finally the p++GaN layer 290. These layers grow on the relaxed GaN buffer 210 and are pseudomorphic to GaN. The In0.2GaN QW within the active region 250 may experience high strain due to a large lattice mismatch with GaN.

[0044] In FIG. 2B, the stack forming the LED 201 grows on an In0.05GaN base layer 211 with a relaxed surface and has a lattice constant larger than that of GaN (in the case of complete relaxation, the lattice constant is the lattice constant of bulk In0.05GaN). The LED layer may contain InGaN instead of GaN. The LED layer may have the same content as the base region, thereby obtaining low strain (however, other compositions including GaN and AlGaN layers are also possible). In the example shown in FIG. 2B, the stack grown on the base 211 includes, from bottom to top, an In0.05GaN layer 221, an InGaN / InGaN underlayer 231, and an In0.05GaN spacer 241 below the active region 251. Above the active region 251, the stack includes an In0.05GaN spacer 261, an AlGaN EBL 271, a p In0.05GaN layer 281, and a p++In0.05GaN layer 291. The same In0.2GaN QW in the active region 251 maintains lower strain than in FIG. 2A. This can facilitate improvement in material quality and performance.

[0045] In some embodiments, the LED emitter includes pixels, and each pixel has three sub-pixels (for example, emitting blue, green, and red light). Each sub-pixel includes a nanowire (NW) emitter. The nanowire is characterized by an InGaN base region grown by HVPE and an InGaN LED region having quantum wells grown by MOCVD. The red nanowire is characterized by an InGaN base region (5% < [In] < 15%) having a relaxed surface region and InGaN quantum wells ([In] > 20%) grown to be pseudomorphic on the surface region.

[0046] At least one layer in the InGaN base region may be substantially relaxed and have an [In] composition of 50% or less (for example, 40% or less, 30% or less, 25% or less, 20% or less, 15% or less). In some embodiments, at least one layer in the InGaN base region is substantially relaxed and has an [In] composition of 5% or more (for example, 7.5% or more, 10% or more, 12.5% or more, 15% or more).

[0047] Growth Growth includes growth of the base region and growth of the LED region on the base region. These will be described next.

[0048] Base region growth In some embodiments, the base region of the LED includes a III-nitride region having an In composition.

[0049] Generally, growth of a thick InGaN layer is difficult. It has been found that high-quality InGaN layers can be grown by hydride vapor phase epitaxy (HVPE). For example, trihalide HVPE (THVPE) using InCl3 and / or GaCl3 precursors can efficiently incorporate a large amount of In into the crystal, resulting in an InGaN composition in the range of 0 to 100%. THVPE InGaN growth has been demonstrated in both planar and NW geometries.

[0050] Accordingly, embodiments can use a growth technique suitable for growing an InGaN base region. The base region may have an In content in the range of at least 3% (e.g., 5% or more, 8% or more, 10% or more, 12% or more, e.g., up to 15%), e.g., in the range of 5 to 15% (e.g., 5 to 10%, 10 to 15%, 5 to 12%).

[0051] The InGaN material can be characterized by its strain state. In some embodiments, the region of the InGaN material is fully relaxed and the lattice constant (e.g., in-plane and / or vertical) is equal to the lattice constant of the InGaN material in its bulk state. Although the present disclosure focuses on InGaN, other materials (e.g., AlInN, AlInGaN) that provide a suitable lattice constant can be used according to the techniques disclosed herein.

[0052] The base region may be formed on a substrate such as sapphire or Si or GaN (including a GaN template or pseudo-bulk or bulk GaN). The substrate may be configured to reduce the strain of the base region (for example, it may include several III-nitride layers of various compositions to provide strain engineering). The substrate surface may be functionalized with some other material to promote nucleation and growth. The substrate may have an offcut including an offcut in the range of 0 to 3° (for example, 0 to 1°, 0.1 to 1°) in the +m direction (or -m, +a, -a, +c, -c).

[0053] The base region may be grown as a flat layer. This flat layer may then be used as is, or etched to form a lateral structure (such as a mesa or NW). In some embodiments, these lateral structures have typical lateral dimensions of 5 microns or less (for example, 3 microns or less, 1 micron or less, 500 nm or less, 300 nm or less, 150 nm or less, for example, just 100 nm).

[0054] Figures 3A to 3D show such a process flow for manufacturing a base region having a lateral structure. In Figure 3A, a substrate 310 (for example having a buffer layer 320) is provided. In Figure 3B, a flat base layer 330 is grown (for example by HVPE). This base layer may be InGaN. In Figure 3C, the base layer 330 is patterned and etched (for example by dry etching) to form a lateral structure 340 (for example NW). The etching may stop at the interface of the base layer (as shown in Figure 3C), or a part of the base layer may be left unetched, or may extend under the base layer (for example into the buffer and / or substrate). The structure of Figure 3C may be used as a base region for regrowing an LED region. Alternatively, this structure may be further regrown (for example by another technique such as HVPE or MOCVD) to grow extra material in the base region. This regrowth may occur on all surfaces of the lateral structure, or may be selective. Figure 3D shows the resulting structure including the regrown lateral structure 341. This structure may be used for regrowth of an LED region.

[0055] In the approach shown in Figures 3A to 3C, strain relaxation can occur by etching of the lateral structure and / or by regrowth.

[0056] In some embodiments, the base region is grown as a structured layer. For example, the base region has a non-flat shape facilitated by a lateral structure. These lateral structures may have typical lateral dimensions of 5 microns or less (e.g., 3 microns or less, 1 micron or less, 500 nm or less, 300 nm or less, 150 nm or less, e.g., just 100 nm). This structuring can be obtained by forming a mask (such as a hard mask that is patterned and etched to define an opening for access to the growth substrate) on the substrate and growing the base region within the opening of the mask. The opening may have typical lateral dimensions of 5 microns or less (e.g., 3 microns or less, 1 micron or less, 500 nm or less, 300 nm or less, 150 nm or less, e.g., just 100 nm). In some embodiments, the growth is substantially vertical above the opening, but in some embodiments the growth extends laterally outside the opening, and in some embodiments the growth progresses inwards above the opening. Inward growth can be characterized by a negative lateral growth rate and thus a negative ratio of lateral to vertical growth. This ratio may be 0 or less (e.g., -0.1 or less, -0.3 or less, -1 or less).

[0057] Figures 4A - 4C show such a process flow for manufacturing a base region having a lateral structure. In Figure 4A, a growth structure is provided having a substrate 310, a buffer 320 (the buffer is optional), and a growth mask 410 having an opening 411. In Figure 4B, growth of the base layer proceeds within the opening 411, forming, for example, a lateral structure 420 of InGaN. Depending on the dimensions of the mask opening 411, the lateral structure may be a mesa, a micromesa (e.g., having typical lateral dimensions of 10 microns or less, 5 microns or less, 3 microns or less, 1 micron or less, etc., 1 micron or a few microns), or a NW. In Figure 4C, growth occurs not only vertically but also laterally, providing a lateral structure 430 that extends over the mask 410 beyond the opening 411.

[0058] By lateral growth, the size of the lateral structure can be substantially increased. In some embodiments, the openings of the growth mask are relatively small (as taught herein), and the lateral size of the lateral structure is relatively large. The lateral structure may have a lateral size that is at least twice as large as the mask opening (e.g., 5 times or more, 10 times or more, 20 times or more, 30 times or more). The openings may have a lateral dimension selected to facilitate strain relaxation as taught herein. The final size of the lateral structure may be selected to obtain a device of a desired size, such as a size sufficient to facilitate a planar LED manufacturing process. As an example, the openings have a lateral dimension less than 500 nm, and the lateral structure has a lateral dimension greater than 1 micron.

[0059] The growth parameters may be selected to promote lateral growth. In some embodiments, a portion of the base layer grows with a lateral-to-vertical growth rate ratio greater than 2 (e.g., 5 or more, 10 or more, 50 or more, 100 or more). In some embodiments, first growth parameters are used to grow a substrate into the openings, and when the substrate protrudes above the mask, second growth parameters promote lateral growth.

[0060] In some embodiments, other growth parameters that do not promote (or even suppress) lateral growth are used. For example, once the base structure is obtained, the LED is regrown under such growth conditions. In some embodiments, a layer is grown with a vertical-to-lateral growth rate ratio greater than 2 (e.g., 5 or more, 10 or more, 50 or more, 100 or more). For clarity, the lateral growth rate characterizes the plane with the fastest growth among non-vertical planes. This may be an m-plane, an a-plane, a semi-polar plane.

[0061] In some embodiments, the base region has a plurality of sub-regions. Each sub-region has a lateral structure, and the characteristics of the structure vary by sub-region. For example, the dimensions of the lateral structure may be different (e.g., the NWs may have different diameters or lateral dimensions or heights), and / or the composition of the lateral structure may be different (e.g., some NWs have 5% In and other NWs have 10% In).

[0062] The lateral structure may be a mesa or NW. The lateral structure may have a height in the range of 10 nm or more (e.g., 50 nm or more, 100 nm or more, e.g., 10 microns or less, 3 microns or less, 2 microns or less, 1 micron or less), such as 50 nm to 10 microns, 10 nm to 1 micron, 100 nm to 2 microns, 500 nm to 3 microns, and a typical lateral dimension in the range of 500 nm or more (e.g., 1 micron or more, e.g., 20 microns or less, 10 microns or less, 5 microns or less, 3 microns or less), such as 500 nm to 20 microns, 1 to 3 microns, 1 to 5 microns, 1 to 10 microns.

[0063] A part of the base region can be characterized by its strain state and / or its lattice constant. In particular, the base region has a regrowth surface on which the LED region is formed. These regrowth surfaces can be the upper part and / or sidewalls of a lateral structure (NW or mesa). The regrowth surface can be fully relaxed and have a lattice constant (in-plane and / or perpendicular) equal to that of a bulk material of the same composition. In some embodiments, the relaxation is partial and the lattice constant is within 1% (e.g., within 0.5%, within 0.3%, within 0.1%, within 0.05%, within 0.03%, within 0.01%) of that of a bulk material of the same composition. The lattice constant of the bulk relaxed material is also known as the equilibrium lattice constant. Similarly, the relaxation can be expressed in absolute units rather than relative units. Relaxed layers (including base region layers, active region layers, quantum wells, superlattices) can have a lattice constant within 5E-3 nm (e.g., within 3E-3 nm, within 1E-3 nm, within 0.5E-3 nm, within 0.1E-3 nm) of their equilibrium lattice constant. A pseudomorphic layer grown on top of a first layer can have a lattice constant within 5E-3 nm (e.g., within 3E-3 nm, within 1E-3 nm, within 0.5E-3 nm, within 0.1E-3 nm) of the lattice constant of the first layer.

[0064] The cross-section of the lateral structure may have various shapes, which can affect strain relaxation. In some embodiments, the lateral structure is elongated (e.g., a stripe that is much longer than its width). This can facilitate relaxation along the narrower direction. In some embodiments, the lateral structure has a length L and a width w, and L / W > 3 (e.g., 5 or more, 10 or more, 50 or more, 100 or more). In some embodiments, w < 300 nm (e.g., < 200 nm, < 150 nm, < 100 nm, < 75 nm, e.g., just 50 nm). In some embodiments, the strain is uniaxially relaxed along the width direction. In some embodiments, two equivalent crystal directions (e.g., two so-called a-directions or m-directions) in the plane of the lateral structure are characterized by different relaxations, and the lattice constants differ by more than 0.1% between the directions. In contrast, in other embodiments, the shape is substantially regular (such as circular, square, triangular or hexagonal), which can facilitate biaxial strain relaxation. In some embodiments, two equivalent crystal directions (e.g., two a-directions or m-directions) in the plane of the lateral structure are characterized by equal or similar relaxations, and the lattice constants differ by less than 0.1% between the directions.

[0065] In some embodiments, the strain in the base region varies spatially. For example, the base region is InGaN grown on GaN. Initially, the InGaN is pseudomorphically lattice-matched to the GaN, and as growth progresses, the InGaN relaxes the strain (by forming defects and / or by lateral expansion), and after sufficient growth, the InGaN achieves partial or complete relaxation. Relaxation by lateral expansion can occur as soon as the substrate is free to grow laterally, such as when growth reaches the top of a patterned mask. In the case of relaxation by defect formation, the extended defects can terminate at the sidewalls of the lateral structure (e.g., can bend towards the sidewalls) rather than propagating in the growth direction.

[0066] Some embodiments are characterized by a small thickness necessary to achieve relaxation, and the relaxation can occur within a growth range of less than 1 micron (e.g., 500 nm or less, 200 nm or less, e.g., only 100 nm). This can be facilitated by the presence of a lateral structure that provides a relaxation mechanism beyond the relaxation mechanism available in the pseudo-bulk layer. In some embodiments, the base region includes an InGaN composition characterized by a bulk critical thickness t for relaxation, and the relaxation occurs within a thickness range of less than 80% of t (e.g., 50% or less, 20% or less, 10% or less, e.g., only 1%).

[0067] In some embodiments, the relaxation remains elastic and plastic relaxation is avoided. In some embodiments, plastic relaxation occurs and is associated with the formation of extended defects (such as dislocations). The structure is configured such that these extended defects remain far (e.g., at least 100 nm) from the light-emitting layer (this can be achieved by growing a sufficiently thick material on top of the plastic relaxation zone).

[0068] In some embodiments, the regrown surface has a low defect density. The regrown surface may have a threading dislocation density (TDD) of less than 5E8 cm -2 (e.g., less than 1E8 cm -2 or less, 5E7 cm -2 or less, 1E8 cm -2 or less, 5E6 cm -2 or less, only 1E6 cm -2 ). This low TDD may be achieved by growing the base region on a low TDD substrate. This may be achieved by moving the TDD to the lateral region or by growing sufficiently to annihilate each other. The density of defects (including TDD, stacking defects, V-pits) may be configured to be less than 1 per lateral structure (e.g., 1 per NW, 1 or less per 10 lateral structures, 1 or less per 100 lateral structures).

[0069] In some embodiments, the base region includes NWs having a non-uniform composition. For example, the base region NWs may include InGaN of a first composition (e.g., 5%) and InGaN of a second composition (e.g., 8%).

[0070] Figures 5A - 5D show base regions having different In contents. In Figure 5A, the base region has NW500. NW500 has a plurality of regions 510, 520, and 530 having different compositions (e.g., 1, 2, 3). Regions 510, 520, and 530 may be flat or may have another shape. In the example of Figure 5A, the first composition coincides with the thickness of mask 410, but this is not necessary. Strain and strain relaxation can vary by region. As shown in Figure 5B, the In composition may increase during growth (e.g., from 3% to 5% and then to 10%), which can facilitate gentle strain relaxation. Thus, the change in In% along the main epitaxial direction may be at least 2%. The regions may be discrete, or the composition may vary continuously as in Figure 5C. As shown in Figure 5D, a low In region (e.g., 5% above 8%) may follow a high In region. In this case, the high In region may be partially relaxed and defective, and the low In region may have greater relaxation and fewer defects (e.g., because it grows pseudomorphic on the high In region or because it grows low / strain-free).

[0071] In some embodiments, the regrown surface has a uniform composition, and the composition of each element varies by 2% or less (e.g., 1% or less, 0.5% or less) across the regrown surface.

[0072] In some embodiments, the base region or a portion thereof is doped (e.g., n-doped or p-doped). For example, the base region has an n-doped InGaN region due to the presence of one or more dopants (such as O, Si, etc.), or a p-doped InGaN region due to the presence of dopants (such as Mg, Ge, etc.). The doping level may be sufficient to provide good carrier conductivity and may be at least 1E16 (e.g., 5E16 or more, 1E17 or more, 5E17 or more, 1E18 or more, 5E18 or more, 1E19 or more, 5E19 or more). The doping level may be low enough to avoid free carrier absorption and may be less than 1E20 (e.g., 5E19 or less, 1E19 or less, 5E18 or less, 1E18 or less). The appropriate doping upper and lower limits may depend on the doping species (due to variations in activation levels and optical cross-sections between species). In some embodiments, the doping species is O and the doping level is in the range of 1E17 to 1E19 cm -3 ^-3. The doping species can form various states in the crystal, including complexes and interstitial sites. Embodiments where the base region is p-doped open the possibility of a reverse polarity structure (i.e., a junction where n is on top of p in the LED stack). In such cases, LED growth can start with undoped InGaN, an undoped layer including the active region can be grown, and finally n-GaN can be grown. Optionally, the base region may be activated prior to regrowth. The reverse polarity structure may be inverted during processing to expose the p-layer of the base region. A p regrowth step may be performed on these exposed p-layers to create a contact layer for forming a p-contact.

[0073] The base region may be grown in an HVPE reactor. The reactor may use various precursors including monohalides (e.g., GaCl, InCl) and / or trihalides (e.g., GaCl3, InCl3). MCl and MCl3 (where M is a Group III metal including Ga, In, Al) can be obtained by pre-reacting M with HCl or a gaseous Cl species (including Cl2), or from the solid form (e.g., sublimation of MCl or MCl3 solid / powder in a suitable carrier gas such as N2 or H2). The precursor may be formed in the reactor, for example, in a region separate from the growth region. The reactor may use NH3 as a source of N. In some embodiments, InCl3 facilitates incorporating a high indium composition, for example, higher than 3% (e.g., 5% or more, 8% or more, 10% or more) into the base region. A carrier gas including H2 / N2 / Ar / He and combinations thereof may be mixed with these precursor gases. Growth may use one of the following precursor combinations, namely, GaCl / InCl, GaCl / InCl3, GaCl3 / InCl, GaCl3 / InCl3. Some precursor combinations may be suitable for some growth directions. In some embodiments, the precursor is GaCl3 / InCl3 and growth occurs along the -c plane. In some embodiments, the precursor is GaCl / InCl3 and growth occurs along the +c plane. The InCl3 precursor may facilitate a high incorporation of In. In some embodiments, InCl3 is used and the base region has an In(x)Ga(1-x)N material with x > 0.05. More generally, possible precursors include MX and MX3, where M is a Group III metal (In, Ga, Al) and X is one of Cl, Br, I.

[0074] The composition of the base region may be selected to provide a sufficiently small lattice constant mismatch with the light-emitting layer of the LED region. In some cases, this requires an InGaN material having a composition within the miscibility gap. Embodiments of the present invention facilitate this by using growth techniques (e.g., HVPE) and growth parameters that are advantageous for dynamic growth such that the miscibility gap disappears. The pressure (or partial pressure of the seed) may be maintained at a predetermined value to ensure dynamic growth. A predetermined supersaturation of the seed may be obtained.

[0075] The doping element may further be introduced into the HVPE reactor. The dopant source may be a gas containing an O-containing gas (including O 2 ), or a gas containing a Si-containing gas (including silane, dichlorosilane), or a solid (e.g., an oxide or solid form Si containing high-purity crystalline Si). The doping species and concentration may be selected to limit the strain caused by doping.

[0076] In some embodiments, the base region growth conditions are selected to reduce defect formation. In particular, a low vacancy density (including N or Ga or In) is required, where the vacancy density is less than 1E18 cm -3 (e.g., less than 1E17 cm -3 or less, less than 1E16 cm -3 or less, less than 1E15 cm -3 or less, less than 1E14 cm -3 or less, less than 1E13 cm -3 or less, less than 1E12 cm -3 or less, less than 1E11 cm -3 or less, for example less than 1E10 cm -3 or less). The low density may be achieved by using a relatively low growth temperature such as 900°C or less (e.g., 850°C or less, 800°C or less, 750°C or less, 700°C or less, 650°C or less, 600°C or less, 550°C or less, 500°C or less). The low density may also be achieved by using a high partial pressure of the corresponding species.

[0077] The composition of the base region may be controlled to limit the optical absorption of the light emitted from the LED region. In some embodiments, the LED region has sub-regions that emit light at different wavelengths (e.g., blue / green / red), and thus, the re-absorption of the shortest wavelength is most likely. The composition of the base region is selected to limit the optical absorption of the shortest wavelength. In some embodiments, the sub-regions of the LED region emit short-wavelength light (e.g., blue light) having a peak wavelength, and the bulk absorption coefficient of the base layer at the peak wavelength (i.e., the absorption it would have in bulk form) is 10 cm -1 less than (e.g., 5 cm -1 or less, 2 cm -1 or less, 1 cm -1 or less). In some embodiments, after the complete device is formed, the net power absorption of the short-wavelength light by the base region is less than 10% (e.g., 5% or less, 2% or less, 1% or less). This net power absorption quantifies how much of the total light is absorbed by the base layer and directly competes with the net extraction efficiency of the device. In other words, the extraction efficiency (for a sub-pixel of a given color) can be written as Cex = 1 - Abase - Aother, where Abase is the net base layer absorption and Aother is the absorption from all other sources (metal, active region, free carrier absorption, etc.). In some embodiments, for blue sub-pixels, Abase < 10% (e.g., < 5%, < 2%, < 1%).

[0078] Absorption may be reduced by selecting the composition and thickness of the absorption material, as taught above. Other means may be used alone or in combination with the material composition to limit absorption. This includes forming, for example, an optical separation layer (e.g., a reflector, a mirror) between sub-pixels to form an LED device in which the optical path between sub-pixels (e.g., from a blue LED to a red LED) is reduced or blocked. This includes selecting an appropriate physical layout of the sub-pixels. This includes removing the absorption material (e.g., by etching, grinding, and other techniques disclosed herein). In some embodiments, an absorption material (e.g., a substrate, an epitaxial layer, a portion of a substrate) is present during some of the epitaxial steps and is removed or partially removed (e.g., at least 25%, at least 50%, at least 90% of the material is removed) during device processing.

[0079] Thus, the In composition of the base region may be high enough to reduce strain in the active region but low enough to reduce optical absorption. In some embodiments, the In composition of the base region ranges from 2% to 20% (e.g., 5 - 10%, 2 - 5%, 3 - 10%, 5 - 8%, 5 - 12%, 5 - 15%, 10 - 20%).

[0080] In some embodiments, a regrowth region is prepared for regrowth. Surface treatment may be performed to ensure that the regrowth region is in a state where epitaxial growth is possible. The surface treatment may include one or several wet etchings (including acids, bases, solvents). By several wet etchings, several crystal planes may be selectively etched. The wet etching may include KOH or H3PO4 etching. In some embodiments, a polishing step is performed to obtain a smooth surface with an RMS roughness of less than 5 nm (e.g., 3 nm or less, 1 nm or less, 5 Å or less, 3 Å or less). The polishing may be mechanical, chemical, chemical-mechanical, grinding, and other techniques known in the art. In some embodiments, a dry etching step (such as ICP, RIE, etc.) is used to etch the material. Several techniques may be combined to achieve a desired thickness and a desired surface state. In some embodiments, a first step (such as dry etching) removes the material, and a second step (such as polishing or wet etching) facilitates a low roughness. In some embodiments, the regrowth region has a surface with a desired offcut from the crystal direction. For example, the regrowth surface may have an offcut angle in the range of 0.1 to 5° (e.g., 0.1 to 1° or 1 to 5°) in a specific direction (including the a-plane or m-plane) and be slightly deviated from the c-plane. The offcut may be obtained by a polishing step.

[0081] Figures 6A - 6B show examples of a base region having NWs with offcuts. In Figure 6A, the upper surface of the base region is tilted over a macroscopic distance (e.g., the entire wafer) to provide NWs 610 sharing the offcut. In Figure 6B, the offcut occurs individually for each NW 620.

[0082] The growth reactor may operate at a pressure selected for high material quality and desired material properties. High pressure may be desirable to reduce the presence of some defects including vacancies. In some embodiments, the pressure is atmospheric pressure or higher than 1 atm (such as at least 1.2 atm, at least 1.5 atm, at least 2 atm, at least 5 atm, at least 10 atm, etc.). In some embodiments, the partial pressure of the N-containing species is high to reduce the presence of N vacancies in the crystal. In some embodiments, the pressure is selected to promote strain relaxation as disclosed herein.

[0083] The growth parameters may be selected to ensure a sufficient growth rate. In some embodiments, the growth rate is at least 1 micron / hour (e.g., 5 microns / hour), which may be sufficient to grow a base region having a thickness in the range of 100 nm to 10 microns. In some embodiments, the growth rate is at least 20 microns / hour (e.g., 50 microns / hour, 100 microns / hour), which facilitates the growth of a thick base layer and / or a pseudo-bulk base layer.

[0084] The HVPE reactor may adopt various shapes, including those more frequently seen in other growth technologies that can provide advantages in embodiments. This may have a longitudinal / horizontal shape. This may have a vertical flow. The reactor may be a dual-flow reactor having a carrier gas flow in a given direction and a secondary gas flow in a second direction that facilitates control of the carrier gas flow. This may have a showerhead design. The shape may be selected to enhance growth uniformity. Growth may occur on at least one wafer having a radius of at least 4 inches, and the In composition of the base layer material may vary by less than 3% (e.g., 2% or less, 1% or less, 0.5% or less) over an area of at least 60% (e.g., 80% or more, 90% or more) of the wafer area. The reactor may include a quartz material. This may be a cold-wall reactor. This may be a hot-wall reactor in which the temperature of the inner wall of the reactor is maintained higher than a desired temperature including at least 400°C (e.g., 500°C or more, 600°C or more, 650°C or more, 700°C or more). The reactor may be designed to limit the presence of certain atomic species in the crystal. This includes species such as Fe, Cu, Sn, C, B, Mn, etc. The concentration of the selected species may be less than 1E15 cm -3 (e.g., less than 1E14 cm -3 , less than 1E13 cm -3 , less than 1E12 cm -3 , less than 1E11 cm -3 , less than 1E10 cm -3 ). In some embodiments, the portion of the reactor where growth occurs is set to a higher temperature than other portions of the reactor to reduce the incorporation of defects and / or parasitic nucleation. The temperature difference may be at least 50°C (e.g., 100°C or more, 150°C or more). In some embodiments, the wafer temperature may be kept below the maximum temperature to limit the formation of some defects having high formation energies in the substrate (e.g., N vacancies and / or group III vacancies). Thus, the wafer temperature may be in the range of 400 to 1000°C (e.g., 500 to 600°C, 400 to 800°C, 450 to 750°C, 550 to 650°C).

[0085] In some embodiments, the base region is first grown in a lateral structure and combined to form a continuous layer. FIG. 7A shows such a shape where the lateral structures 710 are combined to provide a continuous surface 711. The growth conditions may be selected to promote lateral growth (including promoting growth along planes such as semi-polar or non-polar planes). In some cases, growth occurs along six equivalent crystal planes of the wurtzite structure. Materials from separate lateral structures may combine at the growth front. The combination front may be a plane, a point (i.e., a vertex), or other regions. In some cases, defects such as dislocations, stacking faults, and other crystal misalignments may be formed at the combination front. Alternatively, the combination front may be defect-free. The shape and growth conditions of the lateral structure may be controlled to reduce combination defects. The base region may be grown on a patterned mask having a shape selected to reduce defects in accordance with the behavior of lateral growth. For example, the mask has a triangular lattice and the growth front propagates along six equivalent directions of the wurtzite structure. As shown in FIGS. 7B and 7C, by aligning the mask openings 720 such that the growth front 730 is parallel or perpendicular to the triangular lattice of the mask, a planar (FIG. 7C) or punctiform (FIG. 7B) combination front can be obtained. This may respectively correspond to the alignment of the mask lattice with the m-plane or a-plane. The alignment may be intentionally tilted from the a-plane or m-plane by a preselected angle such as 1 degree or 2 degrees or 5 degrees or 10 degrees. In some embodiments, the mask is aligned with the crystal lattice within a range of + / −5 degrees (e.g., + / −1 degree, + / −0.1 degree). In some embodiments, on average, one or fewer dislocations (or dislocation bundles) are created at the punctiform combination growth front. This is shown in FIG. 7D which shows the mask opening 720 with respect to the dislocation 750. The surface of the continuous layer of the base region may be flat after growth or may be planarized by a process.

[0086] The integration can be obtained using only one growth technique (e.g., HVPE, MOCVD) or by successive growth steps using different techniques. In some embodiments, a base layer having a lateral structure is grown by HVPE and the lateral structure is integrated using MOCVD regrowth. Next, the LED region is grown either in the same MOCVD reactor as the integration step or in a third growth step (i.e., the integration region and the LED region are grown separately by MOCVD).

[0087] Before growing the LED region, the base region may be processed. Before growing the LED region, the base region may be transferred to a submount. Optionally, the base region has a flat upper surface. The upper surface may have group III polarity (i.e., along the +c direction). With the upper surface attached to the submount, this is transferred once. The growth substrate and buffer (if any) may be removed by techniques disclosed herein including grinding and polishing and / or laser lift-off. A part of the exposed base region may be removed / thinned, which may include non-integrated parts (i.e., parts having a lateral structure). This can be achieved by techniques disclosed herein including grinding and polishing. The base region after this step may be flat. The base region may be transferred a second time to a second submount. After this, the base region may be a flat layer continuously attached to the second submount with the upper surface exposed again. This transferred base region may be used as a growth substrate / template for LED growth. The transferred base region may be further patterned (e.g., in a mesa shape) to form small-sized mesas such as micron-scale mesas suitable as sub-pixels. Such mesas may be formed at various parts of the process, for example, when the base region is thinned on the first submount or after the base region is transferred to the second submount.

[0088] Figures 8A - 8I illustrate such a process flow. (A) For example, a combined base region 710 grown on a substrate 310 having a buffer 320 and a mask 410 is provided. (B) The base region is attached to a submount 810. (C) The substrate / mask / buffer is removed by laser lift - off 820. (D) The base region is thinned to obtain a thinned base region 820 having a plane 821. (E) The base region is attached to a second submount 830. (F) The first submount 810 is removed. Alternatively, after step (D), (G) the base region is patterned to form a mesa 825. (H) The base region having the mesa 825 is brought into contact with a second submount 835. (I) A portion of the mesa 825 is transferred. This selective transfer may be achieved by various techniques (e.g., by having a patterned material on the second submount that facilitates selective bonding of some mesas, or by mechanical forces applied to the selective mesas such as forces applied through the first and / or second substrates to push some mesas).

[0089] Using the base region transfer method disclosed herein, different base regions may be combined on the same submount. For example, several combined base regions are combined and they have different compositions and / or strain states. This can be achieved by repeating the process of Figure 8I. The base regions may be spatially combined on the submount such that each base region corresponds to a sub - pixel type. For example, three base regions having mesas with increasing In composition form three sets of sub - pixels on the submount and are used for regrowing an LED region. A repair process of replacing defective mesas may be used before or after regrowth.

[0090] LED region growth The LED region grows on the regrowth surface of the base region.

[0091] The LED region may have an InGaN layer (doped and / or undoped) for carrier transport, similar to a normal GaN n-layer and p-layer, and lower and upper barriers as seen in conventional III-nitride LEDs. The LED region may have an AlGaN or AlGaInN or AlInN layer that serves as an electron blocking layer. The LED region may have an active region with a light-emitting quantum well (QW) / barrier consisting of InGaN / GaN or InGaN / InGaN. The LED region may have a defect reduction layer (such as a uniform InGaN or AlInN layer, or a superlattice of InGaN / InGaN, InGaN / GaN, InGaN / AlInN and other types of III-nitride layers). These various layers may have compositions selected to reduce the strain of the light-emitting layer.

[0092] In some embodiments, the regrowth surface is the upper surface of the NW as shown in FIG. 1A. In some embodiments, the regrowth surface includes some surfaces of the NW, such as the upper surface and side walls as shown in FIG. 1B for example.

[0093] In some embodiments, the LED region is pseudomorphic or near pseudomorphic to the regrowth surface (all layers within the LED region have in-plane lattice constants that differ from the regrowth surface by 0.1% or less than 0.01%). Thus, the lattice constant of the regrowth surface is important as it determines the strain state of the active region.

[0094] In some embodiments, the active region includes one or several QWs having a composition. The composition may include at least 10% (such as 15% or more, 20% or more, 25% or more, 30% or more, 35% or more, 40% or more) of In.

[0095] The layers of the base region and the LED region may be configured to reduce the strain of the QW. For example, the base region has a base composition (e.g., In0.1GaN), and some of the n-layers and p-layers in the LED region have the same base composition so that there is no strain. The barriers between the QWs may also have the same base composition or a similar composition. The composition of the barriers may be configured to compensate for the stress of the QW. For example, the barrier has less In than the base region and has a tensile strain that compensates for the compressive strain of the QW.

[0096] An example of the composition in the stack is as follows (all layers except the first layer are regrown as part of the LED region).

[0097]

Table 1

[0098] The strain state can be quantified by various amounts. A convenient amount is the misfit strain (or basal strain field) between the in-plane lattice parameters of two layers as follows.

[0099] e=(a_b - a_l) / a_l Wherein, a_b is the equilibrium in-plane lattice constant of the base layer (i.e., the layer where pseudo-lattice matching growth occurs), and a_l is the in-plane equilibrium lattice constant of the growing layer.

[0100] In some embodiments, the misfit strain of the QW is reduced to less than 80% (e.g., 50% or less, 30% or less, 20% or less, 10% or less) of the misfit strain when the QW is grown pseudomorphic on a relaxed GaN surface. For example, the misfit strain value of a c-plane In20GaN QW grown pseudomorphic on c-plane GaN is -2.2%. In some embodiments, the same In20GaN QW is grown pseudomorphic on a relaxed In10GaN layer, and its misfit strain is about -1.1%, which is about half of the strain when grown on GaN.

[0101] The following table shows possible embodiments. The embodiments may be configured according to the minimum and maximum boundaries taught in this table. For example, an embodiment may have a base layer with an in-plane lattice constant having a value greater than a selected value (e.g., 3.22 A), and an active region having at least an InGaN composition greater than a selected value (e.g., 30%), and may be configured to have a misfit strain ratio less than a selected value (e.g., 67%).

[0102]

Table 2

[0103] This table assumes that the base layer is relaxed InGaN. However, other materials (including In-containing III nitride compounds and other materials) having similar in-plane lattice constants are also suitable. The misfit strain ratio is the ratio of the actual misfit strain (between the base layer and the active layer) to the value of the misfit strain when the active layer is grown pseudomorphic on GaN.

[0104] In some embodiments, the strain component epsilon_3 is approximately proportional to the misfit strain, and thus the polarization field due to the strain is approximately proportional to the misfit strain. Thus, the values of the misfit strain ratio possible herein may also correspond to the values of the polarization field ratio, which is defined as the ratio of the actual polarization field in the active layer to the polarization field when the structure is pseudomorphic to GaN.

[0105] In some embodiments, at least 50% (e.g., 80% or more, 90% or more) of the light emitted from the LED is emitted from one or several active layers, and the active layers are further characterized by the properties (composition, misfit strain, misfit strain ratio, polarization field ratio, etc.) taught herein.

[0106] In some embodiments, the active region has a composition and a thickness that is at least 1.5 times (e.g., 2 times or more, 3 times or more) the critical thickness for pseudomorphic lattice-matched growth of the composition on relaxed GaN. This is facilitated by strain reduction in the active region. In some embodiments, the QW has a thickness in the range of 2 - 4 nm and a composition in the range of 30 - 60%.

[0107] In some embodiments, the active region has a light-emitting layer having an In content of at least 30% (e.g., 35% or more, 40% or more, 50% or more, e.g., in the range of 30 - 60%) and a thickness of at least 2 nm (e.g., 2.5 nm or more, e.g., in the range of 2 - 5 nm).

[0108] In some embodiments, the active region undergoes lateral relaxation during its growth. For example, the active region is flat (perpendicular to the axis of the base NW / platelet / mesa) as shown in FIG. 1A. As the active region grows, it has a high In composition and thus has a larger in-plane lattice constant than the underlying layer (including the base region), and thus undergoes compressive strain. The typical lateral dimension of the NW is small enough such that the strain is relaxed by the lateral expansion of the active region. The relaxation may be partial. The active region may include quantum wells having an in-plane lattice constant that is at least 0.01% (e.g., 0.03% or more, 0.05% or more, 0.1% or more) larger than the regrowth plane of the base region.

[0109] The prior art has demonstrated structures that utilize InGaN relaxation (e.g., plastic relaxation by forming defects in a very thick InGaN layer). However, such structures are thought to have very poor IQE. In contrast, some embodiments of the present invention maintain a low defect rate and / or high IQE while reducing the strain in the active region. This is facilitated by growth on a relaxed base region.

[0110] Thus, some embodiments are characterized by low defect levels in the active region. The active region has less than 5E8 cm -2 (e.g., less than 1E8 cm-2 Less than 5E7 cm -2 Less than 1E8 cm -2 Less than 5E6 cm -2 Less than 1E6 cm -2 It may have a TDD of (less than). The active region may have a stacking defect density or a misfit dislocation density of less than 1E5 cm -1 Less than (e.g., less than 1E4 cm -1 Less than 1E3 cm -1 Less than 1E2 cm -1 Less than 1E1 cm -1 Less than). In embodiments having a lateral structure (e.g., NW or micromesa), the density of defects (including TDD, stacking defects, V pits) may be configured to be less than 1 per lateral structure (e.g., less than 1 per NW), or 1 per 10 lateral structures, or less than 1 per 100 lateral structures.

[0111] Some embodiments are characterized by a high internal quantum efficiency (IQE). This high IQE may be significantly higher than that obtainable by conventional strain growth on GaN. This can be facilitated by reducing the strain in the active region.

[0112] Figure 9 shows the relationship between wavelength and IQE, contrasting the prior art with the embodiments. The prior art curve is based on published data for planar LEDs grown by MOCVD having various emission wavelengths, and as the well-known green gap appears, the IQE decreases significantly at longer wavelengths. Strain is thought to be the cause of (at least part of) this decrease. Thus, the effect of strain on IQE can be modeled and the improvement in IQE due to strain reduction can be predicted. The curve for the embodiments shows the results of this procedure and indicates the expected performance of some embodiments. Here, a fully relaxed base layer with In0.05GaN is assumed. The EQE of the best prior art devices reported within the red range (620 - 630 nm) is about 2 - 2.5%, which corresponds to an IQE of about 3%. In contrast, the IQE of some embodiments is at least 5% (e.g., 10% or more, 15% or more, 20% or more, 30% or more) at a peak emission wavelength of at least 610 nm (e.g., 620 nm or more, 625 nm or more, 630 nm or more). Figure 9 is illustrative and other values of In composition and strain relaxation may be required to achieve the desired IQE. Embodiments include selecting a desired peak emission wavelength and at least one criterion for a figure of merit of an optoelectronic device (including a minimum desired value of IQE or external quantum efficiency (EQE) or wall plug efficiency (WPE)), and configuring a light emitter as taught herein to achieve this at least one criterion (including selecting the composition and strain state of the base region). In one embodiment, the emission wavelength is at least 615 nm, the IQE is at least 15%, and the base region has an In composition of at least 5% and is substantially fully relaxed.

[0113] In some embodiments, the base region has a plurality of sub-regions. The sub-regions have different base layers of In composition. This can be realized as disclosed herein (e.g., according to the processes of FIGS. 8A - 8I). For example, there are sub-regions having GaN NWs, In0.05GaN NWs, and In0.1GaN NWs. Regions with more In are suitable for growing long-wavelength LEDs. Due to the lattice tensile effect, under the same growth conditions, regions with more In can naturally incorporate more In during LED growth. In some embodiments, the growth of the LED regions occurs simultaneously on different sub-regions, and due to lattice tension, different sub-regions have different active region compositions and different emission wavelengths. In some embodiments, there are three sets of sub-regions, and LEDs emitting blue, green, and red light respectively are grown simultaneously thereon. The base region may have three sets of sub-regions with different compositions. LED regions emitting blue, green, and red light are regrown respectively on these three sets of sub-regions.

[0114] In some embodiments, the sub-regions have NWs of different diameters. For example, there are sub-regions with diameters of 80 nm, 120 nm, and 150 nm. This facilitates the change in In incorporation during the growth of the LED region. For example, regions with a smaller diameter have a smaller growth volume, so In atoms are consumed, facilitating more In incorporation. In some embodiments, there are three sets of sub-regions, and LEDs emitting blue, green, and red light respectively are grown simultaneously thereon.

[0115] The two concepts presented above may be combined. The sub-regions may have different compositions and different NW sizes, and the size and composition tensile effects may be combined to realize different emission wavelengths (including the simultaneous growth of LEDs having desired wavelengths such as blue / green / red).

[0116] In some embodiments, the regrowth of the LED region over some sub-regions is performed in the same regrowth step. In other embodiments, several regrowth steps are performed. For example, the base region has three sets of sub-regions. The first set is exposed and the remaining two sets are covered with a growth mask. The mask includes an oxide material (including SiOx, AlOx), a nitride material (SiNx, AlNx), a dielectric layer, and a metal (including Mo). The growth step is performed (e.g., by MOCVD) to form an LED region having a first wavelength over the first set of sub-regions, which may form, for example, blue sub-pixels. This process is repeated for the remaining two sets of sub-regions to form other LED regions emitting at other wavelengths (e.g., green and red).

[0117] Figures 10A - 10E show a process involving LED region regrowth. (A) A member 1000 having three sets of base sub-regions 1020, 1030, and 1040 is provided on a substrate 1010. These sets have InGaN compositions 1, 2, and 3. (b) A first growth mask 1050 is formed over sub-regions 1030 and 1040. (C) The growth of LED sub-region 1060 is performed over base sub-region 1020. (D) The first growth mask 1050 is removed and a second growth mask 1070 is formed over sub-regions 1020 and 1030. (E) After repeating the masking and epi-step several times, LED sub-regions 1060, 1080, and 1090 are grown over all base sub-regions 1020, 1030, and 1040.

[0118] In some embodiments, successive regrowth steps are performed for different wavelengths, with the longer wavelength step being performed last. For example, a red LED is grown last. This can facilitate good material quality because the longer wavelength active region requires a high In content that can have a low thermal budget. In some embodiments, a red LED is obtained by one regrowth step, and this regrowth step is performed with a low thermal budget. The low thermal budget can be defined by a maximum temperature Tm, and each sub-step in this step is performed below Tm. Tm may be 900 °C or less (e.g., 850 °C or less, 800 °C or less, 750 °C or less, 700 °C or less, 650 °C or less, 600 °C or less, 550 °C or less, 500 °C or less). The low thermal budget can be defined by a maximum temperature Tm and a maximum period tm, and each sub-step in this step is performed below Tm, and the steps performed at or near Tm continue for less than tm.

[0119] In some embodiments, the LED region (or its layer) is grown using a pulse growth technique, for example, by flowing different group III precursors (such as TMG and TMI) at different times. This can facilitate the growth of high In content layers.

[0120] In some embodiments, the LED regrowth occurs over the entire free surface of the base layer. For example, as already shown in FIG. 1B, in a core-shell shape, this occurs on the top and sidewalls of the base NW.

[0121] In contrast, in some embodiments, the LED growth occurs only on a part of the base layer, for example, only on the top facet of the base NW, as already shown in FIG. 1A. This can result in a flat active region along the axis of the NW. The active region may be disk-shaped or, more generally, may have the same cross-section as the NW.

[0122] Various techniques can be used to achieve growth only on this upper portion. The growth parameters may be selected to promote nucleation on the top surface. For example, the top surface is a c-plane, and the growth conditions promote nucleation on the c-plane more than on other planes (e.g., m-plane, c-plane, semi-polar planes). Suitable growth conditions may include temperature, pressure, partial pressures of various precursors, III / V ratio, growth rate, use of pulsed growth. In some embodiments, a low temperature is used to promote top surface growth. The temperature for growing the light-emitting layer may be less than 700 °C (e.g., 675 °C or less, 650 °C or less, 625 °C or less, 600 °C or less, 550 °C or less, 500 °C or less, 450 °C or less, 400 °C or less). Growth techniques suitable for low temperature growth (including MBE, sputtering, plasma-assisted CVD, and other CVD techniques suitable for low temperature) may be used.

[0123] The sidewalls may be covered to prevent epitaxial growth on the sidewalls. In some embodiments, a dielectric material (e.g., SiOx, AlOx, SiNx, AlNx, TiOx, TaOx, ZrOx) is deposited on the sidewalls. The dielectric material may cover all of the sidewalls, or cover substantially all of the sidewalls, leaving the upper region of the NWs for growth.

[0124] Figures 11A to 11F show a process flow for covering the NW sidewall. Figure 11A shows the base region NW1110 after the growth of the base region 1100. In Figure 11B, NW1110 is coated with a dielectric material 1120, where the deposition is conformal (this may be obtained by atomic layer deposition and other known processes). In Figure 11C, the upper part of the dielectric material 1120 is removed to expose the upper surface 1130 of NW1110. This may be achieved by techniques including mechanical processes (e.g., grinding, polishing) and dry etching (e.g., RIE, ICP). The mechanical process includes a selective mechanical process where the etching rate of the dielectric is faster than that of the semiconductor. Dry etching includes a directional dry etching process (which can be achieved by etching the upper surface faster than the sidewalls of the dielectric material and appropriately selecting the etching parameters [pressure, composition, power]) and a selective dry etching process where the etching rate of the dielectric is faster than that of the semiconductor. The final shape before regrowth can be various. Figure 11D shows the case where the dielectric 1121 is etched below the upper surface 1130. Figure 11E shows the case where the dielectric 1122 extends above the upper surface 1130. Figure 11F shows the case where the dielectric 1123 extends around NW1110 non-conformally.

[0125] In the case of Figure 11D, a part of the lateral growth can occur where the sidewall (or the inclined wall if present) is exposed. In some embodiments, it is desirable to limit the amount of lateral growth. Thus, the recess of the filling material may be small enough such that the protruding portion of the NW is small, for example, less than 100 nm (e.g., 50 nm or less, 25 nm or less, 10 nm or less, 5 nm or less).

[0126] In some embodiments, lateral growth occurs, and the growth conditions are selected to reduce In incorporation and / or growth rate for lateral growth such that lateral growth occurs but the In-containing layer (i.e., QW) has less In or is thinner on the sidewalls of the NW than on top of the NW. For example, the QW on the top surface has a thickness t and an In composition x, and the corresponding layer on the sidewalls has a thickness less than t*0.8 (or 0.5) and / or a composition less than x*0.8 (or 0.5). This can prevent light emission and / or absorption by the lateral material.

[0127] As already disclosed, a regrowth surface may be prepared for regrowth (i.e., to be in an epitaxially growable state). Such a preparation step may be performed before or after the dielectric coating step described herein.

[0128] The NWs may be covered with a material other than a dielectric (e.g., metal). Various portions of the NWs, such as sidewalls, tops, inclined walls, semi-polar facets, etc., may be covered to prevent nucleation of regrowth. In some embodiments, regrowth is facilitated by covering some crystal facets and not others. In some embodiments, the uncovered facet is the c-plane (or m-plane, a-plane, semi-polar). Depending on the crystal orientation of the structure, various planes may correspond to the top facet.

[0129] In some embodiments, the layer of the LED region is configured to realize a specific polarization field and control the overlap of the electron and hole wave functions (WFs) in the light-emitting layer. For various crystal directions, the III-nitride heterostructure exhibits both a naturally occurring polarization field and a polarization field due to strain. These fields have various effects, such as the separation of WF overlap (which may be unfavorable for radiative efficiency), an increase in the emission wavelength (which may be beneficial especially for reaching longer wavelengths with a given material composition). Thus, embodiments of the present invention may seek a field of a given intensity or may aim to reduce the trade-off between these effects within a given range. In some embodiments, the magnitude of the polarization field in the active region becomes smaller thanks to the reduction of strain (for example, an In0.3GaN QW has a lower polarization field when pseudomorphically lattice-matched to an In0.1GaN-based region than when pseudomorphically lattice-matched to a conventional GaN layer). In some embodiments, the layers around the active region are selected to manipulate the strain difference and thus the field. For example, the active region may include an In(x)Ga(1-x)N light-emitting QW and at least one layer (such as a barrier) adjacent to the QW, including In(y)Ga(1-y)N (where y < x), or GaN, or AlGaN, or AlInGaN. In some embodiments, the barrier between QWs is composed of multiple layers. For example, the stack between two QWs may be as follows (either the p-side or the n-side is on the left).

[0130] InGaN QW / InGaN / InGaN / InGaN QW InGaN QW / InGaN / GaN / InGaN QW InGaN QW / InGaN / AlGaN / InGaN QW InGaN QW / GaN / AlGaN / InGaN QW In some embodiments, the QW is 1 - 4 MV / cm -1 (for example, 1 - 2, 2 - 2.5, 2.5 - 3, 3 - 4 MV / cm -1) has a polarization field within the range. The polarization field may be selected together with the thickness (since the product of the thickness and the polarization field is equal to the voltage drop across the QW). In some embodiments, the product of the QW thickness and the polarization field across the QW ranges from 0.1 to 1 V (for example, 0.1 to 0.3 V, 0.25 to 0.5 V, 0.5 to 0.75 V, 0.75 V to 1 V, less than 1 V, less than 0.5 V, less than 0.3 V). In some embodiments, the above values are obtained despite the QW having a composition In(x)Ga(1 - x)N (where x > 0.2 (for example, > 0.25, > 0.3, > 0.4)). In one embodiment, the QW has a composition In(x)Ga(1 - x)N (where x > 0.3) and a thickness t > 1 nm, and is grown pseudomorphically on a base layer having a composition In(y)Ga(1 - y)N (where y > 0.05), and the voltage drop across the QW is less than 0.5 V thanks to an appropriate configuration of parameters including y, t, and the composition of the layers surrounding the QW.

[0131] Some embodiments include a bottom layer, for example, a layer configured to improve the IQE of the active region by incorporating defects. The bottom layer may contain In and may be a continuous InGaN or AlInGaN or AlInN layer, or a superlattice of In-containing compounds. Alternatively, if the In concentration of other layers (i.e., the base layer, InGaN n-layer, and barriers) already effectively traps point defects, another bottom layer may not be required.

[0132] Figures 12A to 12B show the shape and strain state of an embodiment. Figure 12A shows the structure after the growth of the base region 1210. A GaN buffer 320 with a mask 410 is provided on a substrate 310, and an InGaN base region 1210 is grown in the mask opening. As the growth of the base region 1210 progresses, its strain is relaxed. Here, this is shown as a lateral expansion, but other relaxation mechanisms are possible as disclosed herein. The base region 1210 ends at a relaxed regrowth surface 1220. At various positions along the growth direction, the lattice constant increases. For example, at the mask opening (position 1211), the growth is pseudomorphic lattice-matched, and the in-plane lattice constant is the in-plane lattice constant of the GaN buffer. At the intermediate position 1212, the relaxation is partial, and the in-plane lattice constant is intermediate between the in-plane lattice constant of bulk GaN and the in-plane lattice constant of bulk InGaN. At position 1213, the material is fully relaxed, and the in-plane lattice constant is the in-plane lattice constant of bulk InGaN. Figure 12B shows the regrowth of the LED region on the regrowth surface 1220, including n-InGaN 1230, an active region 1240, and p-InGaN 1250. The LED region may be pseudomorphic lattice-matched to the regrowth surface 1220, i.e., may have the same in-plane lattice constant as position 3 (within some tolerance, such as a lattice constant value of + / - 0.1%).

[0133] Figures 13A to 13C show various amounts of development when NWs are grown according to Figures 12A to 12B. Figure 13A shows the layers, i.e., the base layer 1210 and the LED layers 1230, 1240, and 1250 (layers with the same composition as the base layer and including QWs with a higher In composition). Figure 13B shows the development of the in-plane lattice constant along the axis of the NW starting from the base region 1210. Initially, the base region is pseudomorphic lattice-matched to the GaN buffer (lattice constant a1), then relaxation begins and the lattice constant increases towards its bulk value (lattice constant a2), reaching full relaxation. After the growth of the base region is completed, the pseudomorphic lattice-matched growth of the LED region continues. Figure 13C shows a simplified schematic of the corresponding strain. The strain decreases during the relaxation of the base layer. The strain increases in the QW due to the high In content.

[0134] The dimensions of the NWs (or other lateral structures) may be selected to facilitate relaxation. In some embodiments, the diameter of the NWs (or typical lateral dimension) is less than 200 nm (e.g., 150 nm or less, 120 nm or less, 100 nm or less, 80 nm or less, 50 nm or less). A small diameter can facilitate lateral relaxation. In some embodiments, several groups of lateral structures are present on the base surface and are characterized by different relaxation levels. This can be facilitated by their different dimensions. In some embodiments, there are three groups of NWs having three different typical lateral dimensions, and the NWs with larger dimensions have less relaxation. In such embodiments, the three groups may correspond to blue, green, and red pixels, and it may be advantageous to increase the relaxation of the red pixels and decrease the relaxation of the blue pixels.

[0135] In some cases, the relaxation in the base region is due not to lateral expansion but to other effects such as the formation of defects (including stacking defects, dislocations, vacancies, interstitials, and other defects) and / or voids. Relaxation may start immediately when the pseudomorphic lattice-matched InGaN material hardly or not at all grows on the GaN buffer. Sometimes, the relaxation is incomplete and the lattice constant of the regrown surface may be smaller than that of the bulk InGaN of the same composition. The difference in lattice constant between the regrown surface and the bulk material of the same composition can be less than half of the difference in lattice constant between the bulk GaN and the bulk material of the same composition.

[0136] FIG. 14 shows a process flow 1400 as an example for obtaining an LED according to an embodiment. In step 1410, a substrate is provided. In step 1420, a GaN buffer is grown on the substrate. In step 1430, the surface of the GaN buffer layer is patterned. In step 1440, an InGaN base region is grown into the opening of the pattern by HVPE. In step 1450, an InGaN LED is grown on the base region by MOCVD. In step 1460, the sample is further processed to form an LED device. Some steps are optional, and some steps can be omitted or the order can be changed.

[0137] In some embodiments, an etching step is performed to remove epitaxial material from the LED region after growth. The etching step may be a selective chemical etching including KOH, H3PO4 and other etching, and may etch some crystal planes faster than other crystal planes. The etching step may be dry etching (including ICP, RIE), or photochemical or electrochemical or photoelectrochemical etching. The etching may etch non-polar facets (including m and / or a) fast and etch c-plane facets (including +c) slowly. Using this, the sidewall material may be removed from the NW or mesa without removing the upper material.

[0138] In some embodiments, the LED region growth occurs conformally such that the material grows on top of and along the sides of the NW (or mesa). An etching step is then performed to remove the sidewall material while leaving the top material. This etching step may be performed until the p-type material and the active region material on the sidewalls are removed. This may expose the n-type material or the material of the base region of the LED region. The top surface may be coated or otherwise protected prior to the etching step such that only the sidewall material is removed. The structure after etching may have substantially vertical sidewalls. The etching may be selected to reduce sidewall recombination by resulting in a high crystalline quality with few defects (e.g., dangling bonds). The sidewalls may further be passivated (e.g., by a dielectric layer) after growth to reduce recombination. Etching may be used to control the lateral dimensions of the lateral structure. In some embodiments, the etching removes defective material from the sidewalls. The sidewall material may have defects (including threading dislocations, misfit dislocations, dangling bonds), and the material may be removed by etching until there are no defects in the NW.

[0139] In some embodiments, the etching removes material that emits light at an undesirable wavelength. In one example, the NW has a light-emitting region (e.g., a disk-shaped active region in the case of a circular NW) perpendicular to its axis, and the emission wavelength of the active region is non-uniform in the radial direction (i.e., varies from the center to the edge), so that a first FWHM of the emission is obtained. By removing the material near the periphery by etching, a second FWHM that is narrower than the first FWHM is obtained. In another example, the LED region growth is conformal, the top and sidewalls of the active region emit at different wavelengths, and the sidewall material is removed by etching such that the emission comes only from the top of the active region. The etching may form a nanoporous material. In some embodiments, the etching facilitates strain relaxation. For example, by removing the lateral material by etching, the NW diameter is reduced and the lateral lattice expansion of the active region is facilitated, reducing the strain.

[0140] The epitaxial layer may have tensile strain or compressive strain. In some embodiments, an In(x)Ga(1-x)N layer (where x > y) grown on GaN or on an In(y)Ga(1-y)N layer has compressive strain. A layer having tensile strain may be grown in the vicinity of a layer having compressive strain to balance the strain. Tensile strain may be realized, for example, by adding Al to a III-nitride compound (e.g., using an AlGaN layer, an AlInN layer, or an AlInGaN layer with an appropriate composition). For example, an AlGaN barrier may be grown in the vicinity of an InGaN quantum well. In some embodiments, a layer having compressive misfit strain e1 is grown in the vicinity of a layer having tensile strain e2, where 0.25 < |e1 / e2| < 4 (e.g., 0.5 < |e1 / e2| < 2). The vicinity may be 10 nm or less (e.g., 5 nm or less, 2 nm or less, 1 nm or less).

[0141] Shape Some embodiments include NWs or mesas or plateletlets. Although any one of these configurations may be selected as an example below, the teachings are generally applicable.

[0142] The upper surface of the wire may be substantially flat. For example, the transition from the sidewall to the upper surface occurs without an inclined sidewall or with an inclined sidewall in a limited range (less than 20 nm, less than 10 nm, less than 5 nm, less than 3 nm, etc.).

[0143] Figures 15A - 15D show cross - sections of NWs (or plateletlets) along their axis 1501 according to some embodiments. In Figure 15A, NW 1510 has a vertical cross - section and its sidewall 1511 is parallel to the axis 1501. In Figure 15B, the NW has a cross - section with an inclined sidewall 1521 and an active region 1522 that grows only in a plane perpendicular to the NW axis 1501. In Figure 15C, NW 1530 has an inclined sidewall 1531 and the growth of the LED active region 1532 occurs along all facets (core - shell structure). In Figure 15D, NW 1540 has a core - shell structure, but the masking layer 1545 prevents growth along a part of the sidewall. Depending on the dimensions of the masking layer 1545, most of the LED active region 1542 may grow on a planar facet and hardly grow along other planes.

[0144] The NW may be grown substantially along the 0001(+c) direction or along the 000 - 1( - c) direction. The NW may have substantially no domain inversion (i.e., a domain where the polarity switches between +c and - c). In some embodiments, at least 90% (e.g., 95% or more, 99% or more) of the upper surface of the base region has a constant polarity.

[0145] The height of the NW may be in the range of 10 nm to 10 microns (e.g., 10 nm to 1 micron, 100 nm to 10 microns, 100 nm to 3 microns). The diameter (or typical lateral dimension) of the NW may be in the range of 10 nm to 1 micron (e.g., 10 - 100 nm, or 10 nm to 500 nm, or 30 nm to 1000 nm, or 10 nm to 300 nm).

[0146] Some embodiments include micro - plateletlets (i.e., having a lateral structure with lateral dimensions on the order of 1 micron or several microns).

[0147] Figures 16A to 16C show an example of the manufacturing process of a platelet LED according to an embodiment. In Figure 16A, a substrate 1610 having a selective growth opening in a mask 1620 is provided, and a relaxation base layer 1630 is grown. The substrate may be, for example, Si or sapphire having a GaN or Si or AlN nucleation layer. The growth of the base layer 1630 starts at the opening of the mask 1620. The base layer 1630 is relaxed as taught herein. For example, by selecting appropriate growth conditions, the growth along the lateral direction is promoted. At the end of the formation of the structure shown in Figure 16A, the base lateral structure may have a lateral dimension in the range of about 1 micron or several microns, or 500 nm to 10 microns (for example, 1 micron to 5 microns, 500 nm to 10 microns, 1 micron to 3 microns). Due to the growth mode, it may not be flat. Therefore, by using a planarization step as shown in Figure 16B, a flat layer 1632 in a desired form as taught herein may be obtained. In Figure 16C, a micro-LED structure 1640 is grown on the base region 1632 as taught herein. In this figure, the LED has a flat active region 1642 that extends to the end of the micro-LED.

[0148] The micro-LED may have vertical or inclined sidewalls (for example, corresponding to a semi-polar plane). In some embodiments, the active region substantially extends to the end of the micro-LED, but the configuration of the active region changes in the lateral direction. The thickness of the active layer may decrease near the end of the platelet. The thickness of the active layer at the end of the platelet may be less than 90% (for example, 80% or less, 50% or less) of the thickness of the same active layer at the center of the platelet. The composition of the active layer may decrease near the end of the platelet. The In composition of the active layer at the end of the platelet may be at least 1% (for example, 2% or more, 5% or more) less than the In composition of the same active layer at the center of the platelet.

[0149] Such a change may facilitate a reduction in the injection of carriers near the end of the platelet. In some embodiments, an exclusion region exists around the end of the platelet LED. The area of the exclusion region may be from 5% to 50% of the total area of the active region, and may be at least 5% (e.g., 10% or more, 20% or more, 30% or more) and less than 50% (e.g., 30% or less, 20% or less). Less than 20% (e.g., 10% or less, 5% or less, 1% or less) of the total emitted light may originate from the exclusion region.

[0150] Figures 17A - 17D show the lateral variation of the active region characteristics from the central structure to the end structure. Figure 17A shows the central structure and the end structure. Figures 17B - 17D show how the characteristics may vary from the center to the end as a function of the relative lateral distance (0 being the center and 1 being the end). The characteristics may be substantially constant up to a distance from the center (0.8 in this example, but other values such as about 0.7, about 0.9, etc. are also possible) and may vary from that distance to the end. The thickness of the active layer 1642 may decrease by at least 5% (e.g., 10% or more, 20% or more, 30% or more, 40% or more, 50% or more). The In composition of the active layer 1642 may decrease by at least 1% (e.g., 2% or more, 5% or more, 10% or more). The emission intensity may decrease by at least 50% (e.g., 80% or more, 90% or more, 95% or more).

[0151] In addition to thinning the base lateral structure prior to LED growth, processing steps that affect the lateral dimensions of the base lateral structure may be used. The lateral dimensions may be reduced, for example, by lithography and etching, by dry etching or chemical etching, or by other methods known in the art.

[0152] Platelets of various dimensions may be formed on the same substrate. In some embodiments, there are three dimensions, which correspond to red / green / blue emission. In some embodiments, the various dimensions are obtained by varying the size of the mask openings, such that platelets of various sizes are obtained after base region growth. In some embodiments, the various dimensions are obtained by selectively reducing the dimensions of some of the platelets (e.g., by masking some of the platelets and etching others).

[0153] Platelets having substrates in various strain states may be formed on the same substrate. In some embodiments, prior to growth of the substrate, openings of different sizes are formed in the growth mask. Small openings facilitate high relaxation. Large openings facilitate low relaxation.

[0154] Due to the various lateral dimensions or strain states, differences in In incorporation during growth of the LED, as taught herein, are facilitated, and co-growth of LEDs having various emission wavelengths (e.g., red / green / blue), as taught herein, may result. Due to differences in strain states, different lattice tensile effects can be obtained, such that less strained substrates have more In incorporated above them. In some embodiments, the active layer is grown and the difference in In% across different platelets is at least 5% (e.g., 10% or more, 15% or more). This can facilitate a difference in emission wavelength of at least 50 nm (e.g., 100 nm or more).

[0155] Processing After growth of the base region and the LED region, the semiconductor material is processed into an LED device. Various known device architectures can be used, including horizontal, vertical, flip-chip.

[0156] Figures 18A to 18H show an example of a process flow. (A) Provide NW 1830 grown on substrate 1810 / buffer layer 1820. NW 1830 has an n-doped region and a p-doped region (with an active region between the n-region and the p-region). (B) Etch buffer 1820 (e.g., by dry etching) to unify sub-pixels. (C) Form n-contact 1840 in the n-doped material (this material may be made in a part of the buffer as shown, or may be made in other places such as the base region, LED region, etc.). (D) Form planarization material 1850 to obtain a plane along the upper surface of NW 1830. The planarization material 1850 may be a dielectric deposited by various known techniques (electron beam, CVD, PEVCD, ALD), or may be a curable fluid or spin-on material (e.g., spin-on glass). This may be thinned to expose the p-side of the NW. (E) Form p-contact 1860 on NW 1830. The n and p contacts may be transparent (e.g., TCO contacts) or may be reflective (e.g., including a metal layer). (F) Remove p-contact 1860 and planarization material 1850 to open access to n-contact 1840. The device of (F) may be used as it is with contacts formed to n-contact 1860 and p-contact 1840. This device may be a top-emission type device or a bottom-emission type device depending on whether the substrate and contacts are opaque or transparent. (G) Alternatively, flip-chip the device onto submount 1880 having metal wirings 1870 and 1875. The wirings may be in various shapes (here shown as surface layers / rows, but may also be wirings embedded in the submount such as vias, redistribution wirings, and other shapes known in the art). Next, connect wirings 1870 and 1875 to an electrical driver to drive various sub-pixels. (H) Remove substrate 1810 if necessary. For example, this is useful when the substrate is opaque or its presence disturbs the optical system of the system.Substrate removal can be achieved by various techniques such as wet etching, dry etching, mechanical etching (e.g., grinding, polishing), laser lift-off, photochemical or electrochemical or photoelectrochemical etching, and combinations thereof. This is only an example of the process and modifications and substitutions are possible. For example, a vertical LED shape may be obtained by forming an n contact on a subpixel after step H. In this case, next, a member may be attached on the assembly after forming the n contact to provide a current path. Various materials in the assembly, including the substrate, submount, contact, and the members described above, may be transparent, and thus light emission is possible from both sides of the assembly.

[0157] Figures 19A - 19G show another example of a process flow. (A) Provide NWs 1930 grown on a substrate layer 1910 / buffer layer 1920. (B) Form a planarization material 1940 between the NWs 1930. (C) Form a p contact 1950 on the NWs 1930. (D) Flip-chip the wafer onto a submount 1970 having a metal wiring 1960. (E) Remove the substrate 1910. (F) Remove the buffer 1920 (e.g., by laser lift-off). (G) Form an n contact 1980 and a pad 1990 on the n side of the NWs 1930. This is only an example of the process and modifications and substitutions are possible. For example, the substrate 1910 may be transparent and may be removed in the same step as the buffer (e.g., the laser lift-off step). The n pad 1990 may be formed to connect the n contact 1980 to a metal trace on the submount (e.g., by using a directional deposition to form an n contact on the sidewall of the passivated NW).

[0158] To remove the buffer and expose the n-side of the NW, various techniques can be used. These techniques include selective chemical etching, grinding and polishing processes, dry etching processes, laser lift-off (LLO) processes, mechanical breaking / splitting, ion implantation and breaking / splitting (similar to the Smart Cut process), laser ablation or micro-ablation processes (such as stealth processes) and in some cases subsequent mechanical breaking, etc.

[0159] In some embodiments, the buffer has a larger bandgap than the NW. For example, the buffer is GaN and the NW has an InGaN region. This facilitates selective LLO using a radiation source (such as a pulsed laser) that is not absorbed by the buffer but is absorbed by the NW. For example, the wavelength is 390 nm which is not strongly absorbed by GaN but is strongly absorbed by In0.1GaN. In some embodiments, a particular layer of the NW has a high In composition and absorbs, while other InGaN layers of the NW do not absorb. For example, the NW has a core region with In0.1GaN and a sacrificial layer with In0.2GaN. The LLO is performed using a laser that is absorbed by In0.2GaN but not by In0.1GaN. The sacrificial layer may be grown during the growth of the LED region.

[0160] Also, photochemical etching (or photoelectrochemical etching) may be used using a layer of a particular composition that has high absorption for photons and is easily etched. Transparency / absorption refers to the wavelength used in the etching step. For example, one embodiment has a transparent substrate, a transparent layer (such as GaN), a base InGaN layer with high absorption, and an LED layer. The structure is illuminated through the substrate (which may be polished and / or optically finished). The base layer is etched by the illumination being absorbed by the base layer through the transparent layer.

[0161] Wet etching may be used to remove the buffer or the substrate (including the case where the substrate is Si).

[0162] In some embodiments, several techniques are used in succession. For example, the LLO process is used to remove the buffer to expose a portion of the NW. Next, a material removal step (e.g., mechanical polishing, dry etching, etc.) is used to thin the exposed NW to a desired thickness and then form contacts to the polished NW. The NW may be thinned to obtain a planar surface. The NW may be thinned to reach a doped layer. In some embodiments, a portion of the NW is not doped and another portion is doped, and a material removal process is used to remove the undoped material to reach the doped material.

[0163] In some embodiments, the base region may include voids (e.g., the base of the NW may have voids). This weakens the connection between the NW and the underlying layer, making the NW more likely to break near the voids.

[0164] In some embodiments, a surface preparation step is used on the doped surface before forming contact to the doped surface of the NW. This process may include cleaning (including by solvents, acids, bases), wet etching, and dry etching. The surface may be n-doped, and this process may be a dry etching including O or Si that reduces the contact resistance when forming contact to the surface by facilitating higher doping of the surface. In some embodiments, the surface preparation region has a higher doping than the semiconductor before surface preparation. In some embodiments, the surface of the InGaN-based region is exposed, and the base region has a doping level D (e.g., about 1E16, 5E16, 1E17, 5E17, 1E18, 5E19, 1E19) after epitaxy, and the doping increases by at least 10 times D by surface treatment. Thereby, good contact resistance can be provided despite moderate doping during growth. Moderate doping may be desirable, for example, to limit strain caused by doping. The doping level of the base region may be selected to ensure a sufficiently low resistivity of the NW subpixel at a desired current density. In some embodiments, the maximum operating current density is moderate (e.g., 50 A / cm -2 or 10 A / cm -2 or 1 A / cm -2 or 0.1 A / cm -2 less than), and thus a moderate doping level may be tolerated.

[0165] Other Although part of this disclosure describes NW LEDs, part of the teachings apply to LEDs that do not feature NWs. For example, mesa LEDs (also called platelet as disclosed herein) may be used instead of a set of NWs. The mesa may have small dimensions (e.g., 10 microns or less, 5 microns or less, 3 microns or less) and may be a subpixel. The mesa may be formed by growth of a base region having a patterned mesa and regrowth of the LED region. Strain relaxation may be achieved with small mesas as disclosed herein.

[0166] Although an InGaN layer is mentioned, it should be understood that other compounds (e.g., AlInGaN, AlInN) may also be suitable if they provide appropriate strain. For example, the base region of the NW may be AlInN having the same in-plane lattice constant as the desired content of InGaN. Such a base region reduces the lattice mismatch with the InGaN light-emitting layer that occurs in an InGaN base region. The present teachings are applicable to material systems other than III-nitride systems, including III-V and II-VI compound semiconductors.

[0167] For example, other crystals having appropriate lattice constants (including semiconductor and insulating crystals) may be used as a substrate to achieve strain reduction in the active region as disclosed herein. In some embodiments, the substrate material has a crystal symmetry and lattice constant that facilitate the growth of strain-reduced InGaN. The symmetry may be hexagonal (including wurtzite-type symmetry). The lattice constant may facilitate a misfit strain of less than half of the misfit strain in the case of pseudomorphic growth on GaN. In some embodiments, an InGaN base layer is grown on the substrate material, and an InGaN active region is grown on the InGaN base layer. The InGaN base layer may be substantially relaxed or may be pseudomorphically lattice-matched to the substrate material. The base layer may have a base layer In composition, and the active region may have an active region In composition, and the active region In composition is at least 3% (e.g., 5% or more, 8% or more, 10% or more, 12% or more, 15% or more, 20% or more, 25% or more, 30% or more) higher than the base layer In composition.

[0168] The in-plane lattice constant generally refers to the lattice constant in a direction perpendicular to the growth. For example, in a common example of growing a wurtzite-type material along the c-axis (or a direction close to the c-axis), the in-plane lattice constant refers to the lattice constant perpendicular to the c-axis.

[0169] When the elemental compositions are disclosed in this specification, they should be understood as the fractional compositions of the elements of a given group (e.g., Group III or Group V) according to common practice. For example, In0.2GaN represents In0.2Ga0.8N, and the sum of the atomic numbers of In and Ga is equal to the atomic number of N.

[0170] Some embodiments have a lateral structure (e.g., mesa or NW) whose cross-section is not circular (e.g., square, rectangular, hexagonal, elliptical, etc.). Such a structure can nevertheless be characterized by a typical lateral dimension. When the cross-section has an area A, the typical lateral dimension is defined herein as 2*sqrt(A / pi). This definition coincides with the diameter of a circular cross-section.

[0171] The LED light emitters described herein may be used in displays including microdisplays. A microdisplay typically has a plurality of pixels, each of which has red, green, and blue subpixels. The distance between two pixels may be less than 20 μm (e.g., 15 μm, 10 μm, 7 μm, 5 μm, 3 μm). The distance between two subpixels may be less than 10 μm (e.g., 7.5 μm, 5 μm, 3.5 μm, 2.5 μm, 1.5 μm). The microdisplay may be integrated into a display system such as an augmented reality or virtual reality headset. The individual subpixels of the display can operate electrically to emit light and form an image.

[0172] Strain, strain relaxation, and lattice constants can be measured by techniques known in the art. This includes X-ray diffraction, X-ray reciprocal space mapping (RSM), grazing incidence X-ray, lateral electron microscopy, Raman spectroscopy, and other techniques known in the art. For example, RSM measurements along an appropriate direction (such as the (10-15) direction in III nitrides) can indicate whether the layer is pseudomorphic, partially relaxed, or fully relaxed, and enable the measurement of the in-plane lattice constant.

[0173] Such measurements also define the degree of relaxation of the second material grown on the first material, as is known in the art. A pseudomorphic lattice-matched layer is 0% relaxed, a layer with an in-plane lattice constant equal to its bulk equilibrium value is 100% relaxed, and a layer grown on the first material with a lattice constant intermediate between that of the first material (e.g., GaN) and its equilibrium value is 50% relaxed. In other words, the degree of relaxation = (a2 - a1) / (a2_relaxed - a1).

[0174] Accordingly, embodiments provide an InGaN material (e.g., a base layer taught herein) having a sufficient In composition (e.g., at least 5%, 10% or more, 15% or more, 20% or more, 25% or more, 30% or more) and a sufficient degree of relaxation with respect to GaN (e.g., at least 30%, 50% or more, 60% or more, 70% or more, 80% or more). The surface of the InGaN material / base layer can provide such relaxation. Additional layers, such as a light-emitting layer / quantum well, may be grown on the relaxed surface. Such an active layer may have a sufficient In composition (e.g., at least 20%, 30% or more, 35% or more, 40% or more, 50% or more, 60% or more) and a limited degree of relaxation with respect to the InGaN material (e.g., 50% or less, 40% or less, 30% or less, 20% or less, 10% or less). In some embodiments, the device has a buffer layer (e.g., GaN) and a base layer grown on the buffer layer, the base layer having an InGaN surface with In > 10% and a degree of relaxation of at least 50% (with respect to the buffer), and the device further has an LED structure with an active region on the base layer, the active region having In > 35% and a degree of relaxation of less than 20% (with respect to the lattice constant of the first material).

[0175] Numerous embodiments have been described. Other embodiments are within the scope of the following claims.

Claims

1. A method of forming an LED light emitter, the method comprising: providing a III-nitride layer on a substrate, the III-nitride layer having a flat upper surface, the method further comprising: providing discrete lateral growth regions on the flat upper surface; and selectively epitaxially growing a base region containing an In(x)Ga(1-x)N material on each discrete lateral growth region, each base region extending in a direction perpendicular to the flat upper surface, the method further comprising: providing a surface of the In(x)Ga(1-x)N material on a portion of the base region, the surface of the In(x)Ga(1-x)N material having a relaxed strain, the surface of the In(x)Ga(1-x)N material being characterized by a base lattice constant within 0.1% of its bulk relaxation value, the method further comprising: epitaxially growing an LED region on the surface, the LED region including a light-emitting layer of an In(y)Ga(1-y)N material that is pseudomorphic to the surface of the In(x)Ga(1-x)N material, the active region being characterized by a lattice constant within 0.1% of the base lattice constant; where 0.05 < x < 0.15 and y > 0.2; wherein the base region and the LED region form a mesa, and each surface of the mesa opposite the flat upper surface has a lateral dimension smaller than the lateral dimension of the lateral growth region.

2. The method of claim 1, wherein the epitaxial growth of the mesa promotes a reduction in the lateral dimension of the mesa above the lateral growth region.

3. The method of claim 1, wherein the mesa includes inclined sidewalls.

4. The method of claim 3, wherein the sidewalls correspond to semi-polar facets.

5. The method of claim 1, wherein the lateral dimension of each surface of the mesa is shaped by at least one of dry etching, wet etching, cutting, and polishing.

6. The method of claim 1, further comprising passivating the mesa by depositing a material on the sidewalls of the mesa after the epitaxial step.

7. The light-emitting region has a total area including a central area corresponding to a central region and a peripheral area corresponding to a peripheral region, the peripheral area being 10% or more of the total area, and during operation of the LED light emitter, 5% or less of the light emitted from the active region is generated from the peripheral region.

8. The method according to claim 7, wherein the active region has an In composition that is at least 1% In higher than that of the peripheral region in the central region.

9. A method of forming an LED emitter (200), the method comprising: providing a III-nitride layer on a substrate (310), the III-nitride layer having a flat upper surface, and the method further comprising: Forming first, second, and third sets of base regions, each including an InGaN material, on discrete lateral regions provided on the upper surface, the first, second, and third sets of base regions having In compositions x １ , x ２ , x ３ characterized by respectively, x １ > 5% and x １ > x ２ > x ３ and the first base region is relaxed and characterized by an in-plane lattice constant within 1% of its bulk relaxation value, the method further comprising comprising simultaneously growing first, second, and third sets of LED regions on each of said base regions, wherein the active regions of said first, second, and third sets of LED regions have In compositions y １ , y ２ , y ３ characterized by y １ > y ２ > y ３ respectively, and said first, second, and third sets of LED regions emit blue, green, and red light respectively when the LED emitter operates, a method, wherein at least one of the first, second, and third base regions comprises a nanoporous material.

10. The method according to claim 9, wherein the lateral dimension of each lateral region is 5 μm or less.

11. The first, second, and third base regions are characterized by first, second, and third in-plane lattice constants a １ , a ２ and a ３ respectively, where a １ > a ２ > a ３ The method according to claim 9, wherein this is the case.

12. Ordering a of lattice constants １ > a ２ > a ３ is facilitated by lattice stretching, the ordering y of the composition １ > y ２ > y ３ The method according to claim 11, which facilitates this.

13. The method according to claim 9, wherein the second and third base regions are covered with a mask during the processing of the first base region or the first LED region prior to the simultaneous growth of the first, second, and third sets of LED regions.

14. An LED emitter (200), comprising: a III-nitride layer on a substrate (310), the III-nitride layer having a flat upper surface, and the LED emitter further comprising: a plurality of base regions each epitaxially grown on a corresponding discrete lateral growth region on the flat upper surface, each base region comprising an In(x)Ga(1 - x)N material, each base region extending in a direction perpendicular to the flat upper surface, each of the base regions having a surface of the In(x)Ga(1 - x)N material, the surface having a relaxed strain with respect to the In(x)Ga(1 - x)N material of the base region on the flat upper surface, the surface of the In(x)Ga(1 - x)N material being characterized by a base lattice constant within 0.1% of its bulk relaxation value, and the LED emitter further comprising: a plurality of LED regions each supported by a corresponding one of the base regions, each of the LED regions comprising a light-emitting layer of an In(y)Ga(1 - y)N material that is pseudomorphic lattice-matched to the surface of the In(x)Ga(1 - x)N material, and being characterized by a lattice constant of the active region within 0.1% of the base lattice constant; 0.05 < x < 0.15 and y > 0.2; wherein the base regions and the LED regions form mesas, and the surface of each mesa on the side opposite to the flat upper surface has a lateral dimension smaller than the lateral dimension of the lateral growth region.

15. An LED emitter (200), comprising: An LED emitter comprising a III-nitride layer on a substrate (310), the III-nitride layer having a flat upper surface, and the LED emitter further comprising comprising first, second, and third sets of base regions each formed in a corresponding discrete lateral region above the upper surface, each of the first, second, and third sets of base regions including an InGaN material, the first, second, and third sets of base regions being characterized by In compositions x １ , x ２ and x ３ respectively, where x １ > 5% and x １ > x ２ > x ３ and the first base region is relaxed and characterized by an in-plane lattice constant within 1% of its bulk relaxation value, the LED emitter further comprises Comprising first, second, and third sets of LED regions on each of the base regions, the active regions of the first, second, and third sets of LED regions have an In composition of y １ , y ２ and y ３ respectively characterized by, y １ > y ２ > y ３ and the first, second, and third sets of LED regions emit blue, green, and red light respectively during operation of the LED emitter (200). an LED emitter, wherein at least one of the first, second, and third base regions includes a nanoporous material.

Images