The present invention relates to light emitting diodes (LEDs) and, more particularly, to light emitting unit cells and light emitting chips which recycle total internal reflection (TIR) light as a photocurrent source, and methods of forming the same.
Light emitting diodes (LEDs) generally convert electrical energy to light, and are known to be used as light sources. For example, LEDs may be used in full-color displays, image scanners, optical communication systems and various signal systems. LEDs are generally formed from semiconductor materials and typically include an active layer of semiconductor material located between two oppositely doped layers. When a bias is applied across the doped layers, holes and electrons are injected into the active layer, where they recombine to generate light. The light generated by the active region may be emitted in all directions and may escape from the LED through any exposed surfaces. The material of the active layer may be selected for emission of a particular wavelength of light. For example, gallium nitride (GaN) and zinc selenide (ZnSe) semiconductor materials may be used to emit green or blue light. Other examples of semiconductor materials include gallium phosphide (GaP) for green light, gallium arsenide phosphide (GaAsP) for yellow, orange and red light, and gallium aluminum arsenide (GaAlAs) for red light.
The efficiency of conventional LEDs may be limited by their inability to emit all of the light that is generated by the active layer. When an LED is energized, the light that is emitted from the active layer may reach the emitting surfaces/adjacent surfaces at many different angles. LEDs are typically formed from semiconductor materials having relatively high refractive indices (for example, a refractive index of about 2.2-3.8) compared to a refractive index of air (of about 1.0). According to Snell's law, light traveling from a region with a high index of refraction (the semiconductor material) to a region with a low index of refraction (air) that is less than a critical angle (relative to the surface normal direction) may propagate out of the LED. Light that reaches the surface at an angle greater than the critical angle does not pass, but instead experiences total internal reflection (TIR). Because of total internal reflection, much of the light generated by conventional LEDs is not emitted, thereby reducing the external quantum efficiency of the LED.
The present invention relates to light emitting chips and methods of forming light emitting chips. The light emitting chip includes a light emission structure comprising a p-type semiconductor layer, an n-type semiconductor layer and an active layer between the p-type semiconductor layer and the n-type semiconductor layer. The light emitting chip includes at least one light emitting unit comprising a light emitting diode (LED) portion formed from the light emission structure and a plurality of light receiving diode (LRD) portions formed from the light emission structure. The plurality of LRD portions are serially connected and configured to surround the LED portion. The plurality of LRD portions are optically coupled to the LED portion to receive total internal reflection (TIR) light from the LED portion and are configured to convert the TIR light to a photocurrent.
The present invention also relates to a light emitting unit cell comprising a first light emitting diode (LED) electrically connected to a power source, a plurality of light receiving diodes (LRDs) connected in series and a second LED. The plurality of LRDs are optically coupled to the first LED to receive total internal reflection (TIR) light from the first LED and are configured to convert the TIR light to a photocurrent. The second LED is electrically connected in parallel with the plurality of LRDs.
The present invention further relates to a light emitting unit cell comprising a light emitting diode (LED) electrically connected to a power source and a plurality of light receiving diodes (LRDs) connected in series. The plurality of LRDs are optically coupled to the LED to receive total internal reflection (TIR) light from the LED and are configured to convert the TIR light to a photocurrent. The plurality of LRDs feed back the photocurrent to the LED.
The invention may be understood from the following detailed description when read in connection with the accompanying drawing. It is emphasized that, according to common practice, the various features of the drawing are not to scale. On the contrary, the dimensions of the various features are arbitrarily expanded or reduced for clarity. Included in the drawing are the following figures:
Referring to
Conventional LED 100 includes substrate 110, buffer layer 112, n-type GaN layer 104, active layer 114 containing a multi-quantum well (MQW) structure, p-type GaN layer 116 and transparent electrode 108, which are sequentially laminated on substrate 110. Transparent electrode 108 may be used, for example, to enhance a current spreading effect.
Portions of transparent electrode 108, p-type layer 116 and active layer 114 may be removed by mesa-etching such that a portion of the upper surface of n-type layer 104 is exposed. A negative electrode (n-electrode) 102 is formed on the exposed upper surface of n-type layer 104. A positive electrode (p-electrode) 106 is formed on an upper surface of transparent electrode 108.
In active layer 114, electrons and holes are recombined so as to generate and emit light. The MQW structure of active layer 114 is formed by alternately laminating well layers and barrier layers (not shown). The well layer includes a semiconductor layer with a smaller band gap than n-type layer 104, p-type layer 116, and the barrier layer, thereby providing quantum wells in which electrons and holes may be recombined.
Referring to
One conventional technique to improve the extraction efficiency is related to ray redirection using, for example, surface roughening, gratings and volume holograms to circumvent TIR. However, these techniques tend to improve the extraction by no more than about 60% from 40% (which is only a 50% increase in efficiency).
In general, LED efficiency may include an internal quantum efficiency, an extraction efficiency and an external quantum efficiency (also referred to herein as power efficiency). The internal quantum efficiency, extraction efficiency and external quantum efficiency may be defined by respective equations (1)-(3), below as:
Because the external quantum efficiency (power efficiency) is the product of internal quantum efficiency (eq. 1) and extraction efficiency (eq. 2), the external quantum efficiency may be improved by improving either the internal quantum efficiency or the extraction efficiency.
As shown in
Referring next to
LRDs 306-1, 306-2, 306-3, 306-4 are connected to each other in series. Anode p4 of LRD 306-4 is electrically connected to anode p5 of second LED 308. Cathode n1 of LRD 306-1 is electrically connected to cathode n5 of second LED 308 and cathode n0 of first LED 304.
In operation, first LED 304 is powered by power source 302 and second LED 308 is powered by LRDs 306 (i.e., LRDs 306 supply current and voltage to second LED 308 for light emission). LRDs 306 may absorb light trapped inside the layers of first LED 304 (i.e., due to TIR) and convert the absorbed light to a photocurrent. LRDs 306 may be configured to absorb the TIR light without emitting light. Thus, each of LRDs 306 may act as a photodiode. Accordingly, light emitting unit cell 300 may recycle photocurrent that would be lost due to TIR and apply the photocurrent to power second LED 308.
Although four LRDs 306 are shown in
Although one light emitting unit cell 300 is shown in
Referring to
A first LRD, 306-1, is connected to first LED 304 at cathodes n0, n1. Cathode n1 is the n-side of LRD 306-1 which is the same as cathode n0. A p-side of LRD 306-1, at anode p1, is connected to the n-side of LRD 306-2, at cathode n2, to form a series connection. LRDs 306-2, 306-3, 306-4 are similarly connected to each other. Because a p-type layer is formed as an upper layer and an n-type laser is formed as a lower layer, non-planar contacts 402 are provided for serial connection of LRDS 306-1, 306-2, 306-3, 306-4.
The p-side of LRD 306-4, at anode p4, is in contact with the p-side of second LED 308, at anode p5. The n-side of LED2, at cathode n5, is connected to cathodes n0, n1. Because the photocurrent received from LRDs 306 may be a fraction of the power supplied to first LED 304, it may be desirable for second LED 308 to be formed with an area that is smaller than first LED 304.
As shown in
Referring to
Substrate 502 may be formed of a transparent material such as sapphire (for example with a (0001) plane orientation). Substrate 502 may be formed from other materials including, but not limited to, silicon carbide (SiC), GaN or magnesium aluminum oxide (MgAlO2).
Buffer layer 504 may be used to enhance a lattice matching between substrate 502 and n-type semiconductor layer 506. Buffer layer 504 may be omitted depending on a process condition and diode characteristic. According to an exemplary embodiment, buffer layer 504 may be formed from about a 10 nm thickness undoped GaN semiconductor layer on a (0001) surface of sapphire substrate 502 for the lattice matching. In addition to GaN, buffer layer 504 may be formed from, but not limited to, (undoped) GaN, aluminum nitride (AlN), aluminum gallium nitride (AlGaN) or aluminum indium nitride (AlInN).
N-type semiconductor layer 506 and p-type semiconductor layer 510 may be formed of semiconductor materials including, but not limited to, AlmGa1-mN (for 0≦m≦1), to emit blue light. According to an example embodiment, n-type semiconductor layer 506 of about 2 μm thickness may be grown from GaN semiconductor material doped with n-type conductive impurities, such as silicon (Si) or germanium (Ge). According to an example embodiment, p-type semiconductor layer 510 of about 200 nm thickness may be grown from GaN semiconductor material doped with p-type conductive impurities, such as magnesium (Mg), zinc (Zn) or beryllium (Be).
To produce light emitting chips for emitting other colors, n-type and p-type semiconductor layers 506, 510 may be formed from different materials. For example, to emit blue light: ZnSe or indium gallium nitride (InGaN) may be used. To emit green light: InGaN, GaP, aluminum gallium indium phosphide (AlGaInP) or aluminum gallium phosphide (AlGaP) may be used. To emit yellow light: gallium arsenide phosphide (GaAsP), AlGaInP or GaP may be used. To emit orange light: GaAsP, AlGaInP or GaP may be used. To emit red light: aluminum gallium arsenide (AlGaAs), GaAsP, AlGaInP or GaP may be used.
Active layer 508 may include a single quantum well (SQW) structure or a MQW structure. According to an exemplary embodiment, active layer 508 of about 50 nm total thickness may be formed from alternating layers of InGaN/GaN. Active layer 508 may be formed of semiconductor material including, but not limited to, InmAlnGa1-m-nN (for 0<m≦1, 0≦n≦1, 0<m+n≦1) or InmGa1-mN (for 0<m<1). Active layer 508 may be omitted depending on a desired process condition and a desired diode characteristic. According to another embodiment, active layer 508 may be omitted for portions of the light emitting structure corresponding to the LEDs or to the LRDs. According to a further embodiment, light emitting structure 501 may be formed with different active layer materials for the portions corresponding to the respective LEDs and LRDs.
Buffer layer 504, n-type semiconductor layer 506, active layer 508 and p-type semiconductor layer 510 may be grown by using any suitable deposition process, including, but not limited to, metal organic chemical deposition (MOCVD) or molecular beam epitaxy (MBE).
As shown in
According to an exemplary embodiment, a silicon dioxide (SiO2) film (not shown) may be deposited on the p-type semiconductor layer 510 layer, for example, by chemical vapor deposition (CVD). A photoresist (not shown) may be subsequently spun on the SiO2 film. A binary chromium (Cr) mask with a desired insulating pattern may be applied for patterning the photoresist. The photoresist may be exposed to ultraviolet (UV) light by a mask aligner or stepper and may be subsequently developed by a developer. The SiO2 film may be etched through the photoresist by, for example, reactive ion etching (RIE) with insulating pattern 512.
The patterned SiO2 film may be used as a mask for a GaN full etching process. All layers of light emitting structure 501 may be etched via the SiO2 mask through to substrate 502. The SiO2 film may be subsequently removed after the etching process is completed.
As shown in
The patterns for n-GaN mesa-etching may include: an entire n-electrode portion 514, a portion 516 of a first LED portion 520, and a portion 518 of each LRD portion 522. On each portion (514, 516, 518), a top 0.55 μm thickness may be etched away such that the n-type layer 506 is partially exposed.
As shown in
Transparent electrode layer 528 may be deposited to cover p-electrode portion 526, first LED portion 520 (except for the mesa-etched portion 516), LRD portion 522 (except for mesa-etched portion 518), and second LED portion 524. Transparent electrode layer 528 is desirably formed to be substantially transparent to light having a predetermined wavelength. Transparent electrode layer 528 may be formed so that light escaping from the underlying layer (p-type layer 510) may be effectively extracted and so that electrons are spread over transparent electrode layer 528 from p-type layer 510 to p-electrode portion 526. P-electrode portion 526 may also be formed from any suitable metallic materials, such as Au, copper (Cu), a combination of platinum (Pt) and Au, a combination of nickel (Ni) and Au or a combination of chromium (Cr) and Au, for example by sputtering, CVD and evaporation.
As shown in
As shown in
As shown in
Referring to
Next, a theoretical extraction efficiency (ηextr) for light emitting unit cell 300 (
where the conversion efficiency (α) is the respective efficiencies for converting light to a photocurrent within LRDs 306.
A summary of theoretical extraction efficiencies under different conversion efficiencies (α=0.8 and α=1.0) and a different number of LEDs 304, 308 (m=2, m=3) are shown in Table 1 below. Referring to
Referring next to
LRDs 806-1, 806-2, 806-3, 806-4 are connected to each other in series. Anode p4 of LRD 806-4 is electrically connected to anode p0 of LED 804. Cathode n1 of LRD 806-1 is electrically connected to cathode n0 of LED 804. Because LRDs 806 are serially connected, each LRD 806 is powered by a fraction of the received voltage (based on the number of LRDs 806). Consequently, none of the LRD's can emit light. Instead, each of the LRD's 806 operates as a photodiode.
In operation, LED 804 is powered by power source 802. LRDs 806 may absorb light trapped inside the layers of LED 804 (i.e., due to TIR) and convert the absorbed light to a photocurrent. The photocurrent generated in LRDs 806 is fed back to LED 804, to supply a photocurrent to light emitting unit cell 800. Because LRDs 806 are serially connected (and receive a fraction of the voltage), LRDs 806 may be configured to absorb the TIR light without emitting light. Accordingly, light emitting unit cell 800 may recycle photocurrent that would be lost due to TIR and apply the photocurrent to further power light emitting unit cell 800.
Although four LRDs 806 are shown in
Although one light emitting unit cell 800 is shown in
A first LRD, 806-1, is connected to LED 804 at respective cathodes n0, n1. A p-side of LRD 806-1, at anode p1, is connected to the n-side of LRD 806-2, at cathode n2, to form a series connection. LRDs 806-2, 806-3, 806-4 are similarly connected to each other. Because a p-type layer is formed as an upper layer and an n-type laser is formed as a lower layer, non-planar contact 902 is provided for serial connection of LRDs 806-1, 806-2, 806-3, 806-4.
The p-side of LRD 806-4, at anode p4, is in contact with the p-side of LED 804, at anode p0, and a cathode of the system power source. The n-side of LRD 806-1, at cathode n1, is also connected with an anode of the system power source.
As shown in
Referring to FIGS. 5A and 10A-10F, an exemplary method of forming light emitting chip 1000 (
Light emitting chip 1000 may be formed from the same light emission structure 501 described above with light emitting chip 500. As described with respect to
As shown in
As shown in
As shown in
As shown in
As shown in
As shown in
Referring to
Along line 11A-11A′, bridge contact layer 1022 connects the mesa-etched n-type layer 506 on one side of one LRD 1012 to a p-type layer 510 of an adjacent LRD 1012. Bridge contact layer 1022 is formed over passivation layer 1020. Accordingly, photo-electrons created in one LRD 1012 travel from n-type layer 506 to a p-type layer 510 of an adjacent LRD 1012 through bridge contact 1022. Along line 11B-11B′, bridge contact layer 1022 bridges over passivation layer 1020 to connect a p-type layer 510 of LRD 1012 to a p-type layer 510 of LED portion 1010.
Next, a theoretical extraction efficiency (ηextr) for light emitting unit cell 800 (
where the conventional extraction efficiency (ηo) and the conversion efficiency (α) are the respective efficiencies for converting light to a photocurrent within LRDs 806. In equation (5), for the sake of simplicity, any series resistances of LED 804 and LRDs 806 have been neglected.
A summary of theoretical extraction efficiencies under different conversion efficiencies (α=0.6, α=0.8 and α=1.0) are shown in Table 2 below. Referring to
Referring to
Micro-pixelated LED 1300 includes an n-electrode 1302 formed on an exposed upper surface of n-type layer 1304. A p-electrode 1306 is formed on an upper surface of transparent electrode 1308.
Although p-side up configurations (where light is emitted from the p-type layer) of light emitting unit cells 300, 800 are described above, n-side up configurations (where light is emitted from the n-type layer) of unit cells 300, 800 may also be used. Light emitting unit cells 300, 800 of the present invention may be applied to any configuration which provides photocurrent recycling.
An n-side up configuration may be formed, for example, by direct growth of a p-type layer and an n-type layer in reverse order. An n-side up configuration may also be formed by growing a p-side up configuration first and further depositing a suitable metal layer (such as Au and/or Sn)) and an appropriate submount (such as Au-coated Si) on top of the p-type layer. A laser lift-off (LLO) process may be used to detach the n-type layer from the substrate. For example, a laser beam (such as a krypton fluoride (KrF) laser beam) may be applied through the transparent substrate side to detach the n-type layer from the substrate.
An n-side up configuration may also be formed using a flip-chip technique. In this configuration, a p-side up configuration is initially formed, except that the transparent electrode may be replaced by a thick reflective electrode layer. The device may be flipped and bonded onto a submount, typically a Si wafer, by using a bonding material such as solder, a Au bump or a Au ball. In this device, light is emitted from the substrate bottom surface.
Several embodiments of the invention have been described herein. It is understood that the present invention is not limited to these embodiments and that different embodiments may be used together. In addition, it is understood that any conventional technique such as surface roughening, grating, volume hologram and photonic crystal may be incorporated with embodiments of the present invention, for example, for further enhancement of the extraction efficiency.
Although the invention is illustrated and described herein with reference to specific embodiments, the invention is not intended to be limited to the details shown. Rather, various modifications may be made in the details within the scope and range of equivalents of the claims and without departing from the invention.