Dual-wavelength light-emitting LED epitaxial structure and preparation method therefor
By vertically separating two active regions in the LED epitaxial structure and employing a photoexcitation mechanism, the problem of current sensitivity in the emission ratio of dual-wavelength LEDs in existing technologies is solved, achieving stable dual-wavelength emission output and efficient spectral combination.
Patent Information
- Application Number
- PCT/CN2025/096552
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2025-05-07
- Filing Date
- 2025-05-22
- Publication Date
- 2026-11-12
AI Technical Summary
The emission ratio of existing dual-wavelength LEDs is sensitive to current, resulting in unstable color temperature changes and limiting their application range.
In the epitaxial structure of an LED, two active regions are vertically separated. Only one active region emits light directly through current injection, while the other active region emits light through photoexcitation. By adjusting the epitaxial structure layers and electrode settings, a photoexcitation mechanism can be achieved to output two wavelengths of light.
It achieves stability in the dual-wavelength emission ratio, improves the stability of emission color changes with current, enhances luminous efficiency, and possesses flexible spectral combination capabilities.
Smart Images

Figure CN2025096552_12112026_PF_FP_ABST
Abstract
Description
A dual-wavelength LED epitaxial structure and its fabrication method Technical Field
[0001] This invention relates to the fields of light-emitting diode (LED) epitaxial growth and semiconductor lighting device technology, and particularly to an LED epitaxial structure capable of emitting dual-wavelength light and its fabrication method. Background Technology
[0002] LEDs can efficiently convert electrical energy into light energy, releasing light through the recombination of electrons and holes, and are widely used in lighting, displays, and other fields. With the development of LED technology, the demand for full-spectrum output in white light lighting is constantly increasing. Integrating multiple emission wavelengths on the same chip can effectively reduce the manufacturing cost of achieving full-spectrum devices. Currently, common dual-wavelength LED devices typically grow two sets of quantum wells emitting different wavelengths of light sequentially within the same active region, simultaneously producing two wavelengths of light output after current injection. However, in this structure, the relative intensity ratio of the two wavelengths changes with the driving current. This is because as the current density increases from low to high, the spatial distribution of charge carriers in the active region shifts—due to the presence of V-pits in gallium nitride (GaN) materials, at lower current densities, charge carriers mainly recombine and emit light in the quantum wells closer to the N-type GaN side, while at higher current densities, the recombination position of charge carriers shifts towards the quantum wells closer to the P-type GaN side. Therefore, the color temperature of the dual-wavelength LED output light will change inconsistently under different currents.
[0003] The existing dual-wavelength LED luminous intensity ratio is not only sensitive to current and difficult to control precisely, but also limits its application range. To address the shortcomings of the existing technology, it is necessary to provide a new dual-wavelength LED epitaxial structure and growth method to achieve LED devices with stable dual-wavelength luminous intensity ratio and controllable spectrum under varying driving current. Summary of the Invention
[0004] The purpose of this invention is to overcome the shortcomings of existing dual-wavelength LEDs where the emission ratio varies with current, and to provide a dual-wavelength LED epitaxial structure and its fabrication method, so as to achieve stable emission output of two wavelengths on a single chip.
[0005] To achieve the above objectives, this invention vertically separates two active light-emitting regions emitting different wavelengths within an LED epitaxial wafer. By adjusting the epitaxial structure layers and electrode settings, only one active region emits light directly through current injection, while the other active region is excited to emit light by absorbing photons emitted by the former. Specifically, the LED epitaxial structure of this invention includes several semiconductor layers formed sequentially from bottom to top: an N-type semiconductor layer, a first active region, an intermediate semiconductor layer, a second active region, an electron blocking layer, and a P-type semiconductor layer. The first and second active regions are each composed of a multi-quantum-well structure, emitting light peaks at different wavelengths. The two active regions are vertically separated by the intermediate semiconductor layer, and an effective PN junction is formed only in the region containing the second active region for current injection. When energized, the second active region located at the PN junction generates spontaneous emission light. This light passes through the intermediate layer and is absorbed by the first active region, thereby exciting the first active region to also generate spontaneous emission light output. By utilizing the aforementioned photoexcitation mechanism, the structure of this invention can simultaneously output two different wavelengths of light, and avoids chromaticity drift caused by uneven carrier distribution in direct dual-injection structures.
[0006] Furthermore, this invention also provides a method for fabricating the aforementioned LED epitaxial structure, comprising the following steps: First, growing a first N-type semiconductor layer and a first active region (selectively growing a first electron blocking layer and a first P-type semiconductor layer) on a substrate; then, after completing the first active region, continuing to grow an intermediate semiconductor layer and a second active region thereon, and growing an electron blocking layer and a P-type semiconductor layer on the second active region. By adjusting the doping type and hierarchical structure during the first and second stage growth processes, only one PN junction for electroluminescence exists in the final epitaxial wafer, while the other active region is not included in the current path of the PN junction. Subsequently, electrodes are fabricated using conventional chip manufacturing processes, with the P electrode in contact with the P-type semiconductor layer and the N electrode in contact with the N-type semiconductor layer, to obtain a dual-wavelength LED chip.
[0007] Compared to existing technologies, the dual-wavelength LED structure of this invention, due to its photoexcitation of the secondary light-emitting region, has only one active region subjected to electrical drive. Therefore, the output ratio of the two colors of light remains essentially constant under different currents, greatly improving the stability of the emitted color as the current changes. By growing the two light-emitting quantum wells separately, this invention also allows for independent optimization of the materials and structure of each active region, improving overall luminous efficiency and providing the flexibility to combine different emission wavelengths as needed. Furthermore, single-chip integration of dual-wavelength emission facilitates achieving broader spectral coverage (e.g., achieving high color rendering white light illumination) and reduces the cost of generating a full spectrum through multi-chip combinations. Attached Figure Description
[0008] Figure 1 is a cross-sectional schematic diagram of the dual-wavelength light-emitting LED epitaxial structure of Embodiment 1 of the present invention.
[0009] Figure 2 is a cross-sectional schematic diagram of the dual-wavelength light-emitting LED epitaxial structure of Embodiment 2 of the present invention.
[0010] Figure 3 is a simplified schematic diagram of the light-emitting principle of the dual-wavelength LED structure in Embodiment 1 of the present invention.
[0011] Figure 4 is a simplified schematic diagram of the light-emitting principle of the dual-wavelength LED structure in Embodiment 2 of the present invention.
[0012] Figure 5 is a comparison of the spectral output of the structure in Embodiment 1 of the present invention and the comparative example "conventional structure" under different driving currents. Detailed Implementation
[0013] [Correction 13.10.2025 according to Rule 91] The present invention will be further described below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the technical features in the following embodiments can be combined with each other to obtain new implementation schemes. [0013.1] [Correction 13.10.2025 according to Rule 91] Example 1 Example
[0014] As shown in Figure 1, the dual-wavelength LED epitaxial structure provided in this embodiment has a P-type semiconductor layer as the top layer of the epitaxial wafer. Its fabrication method is as follows:
[0015] First, a patterned sapphire substrate (PSS) 100 is provided, and a low-temperature GaN buffer layer 101 and an undoped GaN layer 101 are epitaxially grown sequentially on the substrate. Next, a first N-type GaN layer 103 with a thickness of about 2~3 μm is grown on the undoped GaN layer 120. This layer is doped with silicon to improve n-type conductivity.
[0016] Then, a first stress-relief layer 104 is grown on the first N-type GaN layer 103, and a first active region 105 is grown on the first stress-relief layer 104. Specifically, the first active region 105 adopts an InGaN / GaN multiple quantum well structure, for example, including several periods of InGaN well layers and GaN barrier layers, and can be designed as a light-emitting region emitting green light (first wavelength approximately 520 nm). After the first active region 105 is grown, in this embodiment, an electron blocking layer and a P-type layer are not grown on it (i.e., a complete PN junction is not formed for the time being), but the next step of growth continues.
[0017] A second N-type GaN layer 106 (i.e., an intermediate semiconductor layer) is epitaxially grown above the first active region 105. This layer also uses silicon-doped GaN material, with a thickness of, for example, 1~2 μm, and serves as both an n-type carrier transport layer and an electrode lead-out layer for the upper light-emitting region. To reduce stress accumulation during the continuous growth of the heterostructure, after growing the second N-type GaN layer 106, this embodiment grows a second stress-relieving layer 107 on it. This stress-relieving layer can be, for example, an InGaN / GaN superlattice structure or a thin GaN layer grown at low temperature, to relax some of the lattice stress.
[0018] Subsequently, a second active region 108 is grown on the second stress-relieving layer 107. The second active region 108 also employs an InGaN / GaN multi-quantum-well structure, but its emission peak differs from that of the first active region 105. In this embodiment, the second active region 108 can be designed as a quantum well emitting region that emits blue light (a second wavelength of approximately 450 nm). After the second active region 108 is grown, an electron blocking layer (EBL) 109 is grown on it. This EBL can be a magnesium-doped AlGaN layer, used to block electrons from escaping from the second active region 108. Finally, a p-type GaN layer 110 (i.e., the top p-type semiconductor layer) is grown on the EBL. Magnesium doping gives it p-type conductivity, and a high doping concentration can be formed on the top layer to facilitate ohmic contacts.
[0019] Next, chip fabrication is performed on the epitaxial wafer. First, windows are created on the epitaxial wafer using photolithography and etching processes to remove the top P-type GaN layer 110, EBL 109, and the second active region 108 in a localized area until the underlying second N-type GaN layer 106 is exposed. Then, an N-electrode 111 (a metal electrode, such as a Ti / Al layer) is deposited and alloyed on the exposed second N-type GaN layer 106. Simultaneously, a P-electrode 112 (such as a Ni / Au transparent electrode) is deposited and alloyed on the top P-type GaN layer 110 of the chip. In the LED chip thus fabricated, the P-electrode is located on the top P-type GaN layer 110, and the N-electrode is in ohmic contact with the intermediate second N-type GaN layer 106, with the two electrodes isolated from each other. A top-mounted structure can be used for final packaging, allowing both the P-electrode and N-electrode to extend from the top surface of the chip.
[0020] As shown in Figure 3, when the chip is powered on, current is injected from the P electrode 112, passes through the second active region 108, and is led out from the N electrode 111. Electroluminescence, emitting blue light, is mainly achieved in the second active region 108 through electron-hole recombination. Since the first active region 105 is not in the direct current path, it hardly generates any current injection for light emission. However, some of the blue light emitted from the second active region 108 is transmitted downwards and absorbed by the first active region 105 below when passing through the intermediate layer. After the InGaN quantum well in the first active region 105 absorbs high-energy blue light photons, it is excited to generate electron-hole pairs and undergo radiative recombination, thereby emitting green light.
[0021] [Correction 13.10.2025 based on Rule 91] In this way, the chip simultaneously generates blue and green light output. The luminous intensity of the first active region 105 depends on the photon flux provided by the excitation of the second active region 108, while the luminous intensity of the second active region 108 is determined by the driving current. Therefore, the relative intensity of the two lights is mainly determined by factors such as the absorption efficiency of the first active region 105 for the light emitted by the second active region 108 during structural design, and will not change significantly with current variations as in existing dual-wavelength co-injection structures, thus achieving stable dual-wavelength light output. [0021.1] [Correction 13.10.2025 according to Rule 91] Example 2 Example
[0022] As shown in Figure 2, the other dual-wavelength LED epitaxial structure provided in this embodiment includes a P-type semiconductor layer located at the bottom of the epitaxial wafer. Its fabrication method is as follows:
[0023] First, the basic LED structure layer, i.e., a complete single-wavelength LED epitaxial structure, is epitaxially grown on a patterned sapphire substrate (PSS) 200. Specifically, a GaN buffer layer 201, an undoped GaN layer 202, and a silicon-doped first N-type GaN layer 203 with a thickness of approximately 2 μm are sequentially grown on the PSS 100. Then, a first stress-relieving layer 204 is grown, and a first active region 205 is grown on the first stress-relieving layer 204. An InGaN / GaN multi-quantum-well structure can be used, for example, designed to emit blue light (wavelength approximately 450 nm). After the first active region 205 is grown, an AlGaN electron blocking layer, i.e., the first electron blocking layer 206, is grown to block electrons from entering the subsequent P-type layer. Next, a first P-type GaN layer 207 is grown on the first electron blocking layer 206, and p-type conductivity is achieved through magnesium doping. Preferably, a thin layer of highly doped p++ GaN contact layer (not shown in the figure) can also be grown on the first p-type GaN layer 207 to reduce the ohmic contact resistance. At this point, the first LED structure at the bottom is complete.
[0024] Next, a second light-emitting structure is grown on the structure already containing the p++ GaN contact layer. First, a silicon-doped N-type GaN layer, i.e., the second N-type GaN layer 208, is epitaxially grown on the surface of the first P-type GaN layer 207 or the p++ GaN contact layer using a re-filming process. Since directly growing N-type GaN on the first P-type GaN layer 207 may present problems such as lattice polarity and doping diffusion, interface quality can be ensured by optimizing growth conditions or inserting a transition layer at the interface. In this embodiment, a second stress-relieving layer 209 is grown after the second N-type GaN layer 208 to alleviate the stress of subsequent layers. Then, a second active region 210 is grown on the second stress-relieving layer 209, for example, using an InGaN / GaN quantum well structure designed to emit green light (wavelength approximately 520 nm). A second electron-blocking layer 211 (such as a Mg-doped AlGaN layer) is grown above the second active region 210, and finally, a top layer of the second P-type GaN layer 212 (Mg-doped) is grown.
[0025] Subsequently, the epitaxial wafer is fabricated using chip technology. This embodiment can employ a flip-chip structure to facilitate the removal of the electrodes of the lower P1 layer. Specifically, the second P-type GaN layer 212, second electron blocking layer 211, second active region 210, second stress relief layer 209, second N-type GaN layer 208, first P-type GaN layer 207, first electron blocking layer 206, first active region 205, and first stress relief layer 204 in a localized area can be removed on the top surface of the wafer using photolithography and etching, exposing the underlying first N-type GaN layer 203. Then, metal is deposited to form the N electrode 213. Next, metal is deposited on the first P-type GaN layer 207 or the p++ GaN contact layer to form the P electrode 214. After bonding and packaging, the dual-wavelength LED chip has the P electrode connected to the first P-type GaN layer 207 or the p++ GaN contact layer, and the N electrode connected to the first N-type GaN layer 203; the second active region 210 is located on the upper part of the chip and is not directly connected to any electrode.
[0026] As shown in Figure 4, when a voltage is applied to the chip, the PN junction formed by the first P-type GaN layer and the first N-type GaN layer becomes conductive. Current is injected from the P electrode 214, passes through the first active region 205, and is led out from the N electrode 213, causing the first active region 205 (blue light region) to generate electroluminescence and output blue light. At this time, there is no direct current injection into the upper second active region 210, but a portion of the blue light emitted from the first active region 205 is transmitted upwards and illuminates the second active region 210. The quantum well of the second active region 210 absorbs the blue light photons from below and is excited, generating green light emission.
[0027] Since the light emitted by the upper second active region 210 is entirely excited by the light from the lower first active region 205, when the current is changed, the change in the blue light intensity of the first active region 205 will cause a corresponding proportional change in the green light of the second active region 210. The relative intensity ratio between the two remains essentially constant, unlike traditional dual-wavelength LEDs where the ratio becomes unbalanced with changes in current. This structure is particularly suitable for lighting devices that require stable dual-color output or wide-spectrum output.
[0028] As shown in Figure 5, the left figure shows the emission spectrum of the structure in Embodiment 1 of the present invention. Within a current range of 10 mA to 150 mA, the peak positions of blue and green light remain essentially unchanged, and the green / blue intensity ratio (G / B) remains between approximately 30% and 44%, showing minimal variation. This indicates that the dual-wave output ratio of the structure of the present invention is stable under different currents. The right figure shows the emission spectrum of a conventional dual-wave structure LED sample. As the driving current increases from 20 mA to 300 mA, its G / B intensity ratio drops sharply from approximately 69% to below 10%, indicating that its emission ratio is severely affected by the current. This figure clearly demonstrates the significant improvement of the present invention's structure in terms of dual-wave spectral stability compared to the conventional structure.
[0029] It should be noted that those skilled in the art can select appropriate electrode formation methods according to packaging and application requirements, and these variations are all within the protection scope of this invention.
Claims
1. A dual-wavelength LED epitaxial structure, comprising a first N-type semiconductor layer, a first active region, an intermediate semiconductor layer, a second active region, an electron blocking layer, and a P-type semiconductor layer sequentially formed on a substrate, characterized in that: The first active region and the second active region emit light of different wavelengths, and an intermediate semiconductor layer is disposed between the first active region and the second active region to separate the two active regions from each other; only one PN junction for current injection is formed in the LED epitaxial structure, and the light generated by the current through the PN junction is emitted by one of the active regions and absorbed by the other active region to excite the latter to emit light.
2. The LED epitaxial structure according to claim 1, characterized in that, No P-type semiconductor layer is disposed on the first active region, and the P-type semiconductor layer is located above the second active region; the PN junction for current injection is located between the intermediate semiconductor layer and the P-type semiconductor layer.
3. The LED epitaxial structure according to claim 1, characterized in that, A first P-type semiconductor layer is disposed above the first active region, and a second P-type semiconductor layer is disposed above the second active region, wherein the first P-type semiconductor layer is located below the intermediate semiconductor layer; the PN junction for current injection is formed between the first N-type semiconductor layer and the first P-type semiconductor layer.
4. The LED epitaxial structure according to claim 1, characterized in that, Both the first and second active regions are InGaN / GaN multiple quantum well (MQW) structures; the light emitted by the first active region is blue or green light, and the light emitted by the second active region is visible light with a wavelength different from that emitted by the first active region.
5. The LED epitaxial structure according to claim 1, characterized in that, The first N-type semiconductor layer and the intermediate semiconductor layer are both silicon-doped N-type GaN layers, the P-type semiconductor layer is a magnesium-doped P-type GaN layer, and the electron blocking layer is a magnesium-doped AlGaN layer.
6. [Correction 13.10.2025 according to Rule 91] The method for preparing an LED epitaxial structure according to any one of claims 1-5 is characterized in that, Includes the following steps: a) Epitaxially growing a first N-type semiconductor layer on a substrate, and epitaxially growing a first active region on the first N-type semiconductor layer; b) After the first active region is grown, an intermediate semiconductor layer and a second active region are epitaxially grown on it. c) An electron blocking layer and a P-type semiconductor layer are epitaxially grown on the second active region to form a PN junction structure containing the second active region; d) Fabricate electrodes by contacting and connecting the P-type semiconductor layer and the N-type semiconductor layer to the P-type semiconductor layer and the N-type semiconductor layer respectively, so that current is injected only into the second active region and the first active region emits light by absorbing light from the second active region.
7. The preparation method according to claim 6, characterized in that, In step a), after the first active region is grown, no P-type semiconductor layer is formed, and the intermediate semiconductor layer and the second active region are grown directly in step b). In step d), the N electrode contacts the intermediate semiconductor layer through an etching window, and the P electrode is disposed on the surface of the P-type semiconductor layer.
8. The preparation method according to claim 6, characterized in that, In step a), after the first active region is grown, the first P-type semiconductor layer continues to grow; in step b), the second N-type semiconductor layer is grown on the first P-type semiconductor layer, and then the second active region is grown; in step c), the second P-type semiconductor layer is grown on the second active region; in step d), the P electrode is electrically connected to the first P-type semiconductor layer and the N electrode is electrically connected to the first N-type semiconductor layer by flip-chip bonding.
9. The preparation method according to claim 6, characterized in that, Before growing the second active region in step b), a stress relief layer is first grown on the intermediate semiconductor layer.
10. The preparation method according to claim 6, characterized in that, The quantum well structures of the first active region and the second active region are InGaN / GaN multiple quantum wells, respectively. The emission wavelength of the first active region is in the blue light band, and the emission wavelength of the second active region is in the green light band.
