Abstract
This study reports the experimental demonstration of a GaN-based resonant cavity microLED towards future optical interconnect applications. GaN-based microLEDs can enable high bandwidth and low power consumption optical interconnect applications due to GaN’s excellent material properties. Meta-microLED is fabricated by positioning a GaN LED inside a cavity sandwiched between a nanoporous GaN distributed Bragg reflector mirror and a nanopatterned TiO2 metasurface mirror. With the presence of the resonant cavity, the divergence angle of the meta-microLED has been successfully reduced by 25° for the 100 µm devices and 29° for the 5 µm devices, and the emission linewidth is also reduced by 5.5 nm. Applying the mirrors as Fabry-Pérot resonators, the meta-microLED also demonstrates a more stable light emission with increasing device temperature up to 250°C, where the reference LED shows a wavelength shift of 13 nm and the meta-microLED only shows a shift of 4 nm. These results demonstrate meta-microLED’s great potential towards future optical interconnect applications.
1 Introduction
For decades, chip-to-chip communication has relied on traditional copper interconnects. However, with the increasing demand for higher performance integrated circuits, copper interconnects are reaching the limits of performance due to the continuing shrinking device size and the increasing chip size. With the increasing wire length from the increasing chip size and the smaller more closely packed interconnects due to shrinking device size, copper interconnects have a continuous increase in resistance and capacitance which leads to longer resistance-capacitance interconnect delays (). Thus, the speed is becoming a bottleneck for copper interconnects due to the degrading trends with scaling. For longer distance communications, the copper interconnect would consume higher power as a result of the resistance heating which can also lead to copper interconnect breakdowns. Optical interconnect can be an ideal replacement and improvement to the existing copper interconnects. Compared to the traditional copper interconnects, optical interconnects can effectively transmit signals through longer distances without worrying about the increased latency and power consumption. Moreover, the application of optical interconnects could support the ongoing scaling of current architectures while also making it possible to develop new architectures with higher interconnectivity or bandwidth (). They could help address a wide variety of design challenges, including crosstalk, voltage isolation, wave reflections, impedance mismatches, and pin inductance. This makes optical interconnects an ideal replacement for copper interconnects in the future development of chip-to-chip communication technologies.
Gallium nitride is a proven semiconductor material that has a wide area of applications due to its fantastic properties such as wide bandgap, direct bandgap, high breakdown voltage and high electron mobility (). Moreover, GaN growth is very versatile since it can be grown epitaxially via MOCVD on various types of substrates, and the bandgap and wavelength of GaN films can be tuned by incorporating In and Al during the material growth. Thus, GaN’s excellent properties make GaN-based LEDs an ideal candidate for applications in optical interconnect systems. Since the invention of the blue LED, GaN-based c-plane LEDs have been widely adopted in many technologies such as display and lighting (; ; ). As a result, GaN c-plane LEDs are proven in high volume manufacturing with mature fabrication and growth processes. However, GaN c-plane LEDs fail to meet the electrical and optical performance requirements for the optical interconnect application since the c-plane LED is considered to have Lambertian emission with a large divergence angle and emission linewidth with modulation frequency of less than 1 GHz (; ). Resonant cavity (RC) LED was a novel LED structure invented in 1992 and first demonstrated in a GaAs LED system (). A resonant cavity improves the performance of LED by using reflective surfaces to confine light and create optical resonance, which leads to a narrower controlled emission spectrum and higher spectral purity. In this structure, the active region is sandwiched within a resonant optical cavity formed by a highly reflective and a moderately reflective mirror. With the application of these mirrors as Fabry-Pérot resonators, the light bounces back and forth within the cavity, forming standing waves at specific resonant wavelengths. Inside this cavity, only wavelengths that satisfy the resonance condition are reinforced through constructive interference, and other wavelengths are suppressed. To set the resonance condition, the top and bottom mirrors of the cavity are designed to reflect a specific wavelength at high reflectance. Thus, spontaneous emission from the active region is restricted to the optical modes of the cavity and the optical mode density of the on-resonance wavelengths is strongly enhanced (; ). Due to the Lambertian emission of the conventional LED, a lot of light in the LED would be trapped due to internal reflections. Because of the mirrors, the resonant cavity would be able to direct the light emission into preferred directions and couples light more effectively out of the device, resulting in higher light output power and efficiency. As a result, the RC LED has enhanced spectral purity over the traditional LEDs, where the spectral linewidth is narrower and the light-output intensity of the RC LED is increased (). While RC LEDs would be a great improvement over the GaN c-plane LED in the electrical and optical performances, they have not been proven in high volume manufacturing for nitrides material systems due to the complex requirements on the distributed Bragg reflector (DBR) mirrors to form a resonant cavity. Previous GaN-based RC-LED utilizes DBRs as both top and bottom mirrors. Such mirrors include dielectric oxide DBRs such as SiO2/Ta2O5 (), and nitride DBRs such as AlN/GaN (). However, these conventional DBRs require deposition and stacking multiple layers of oxide or nitride materials, which leads to longer and more difficult material deposition and fabrication procedures. Also, due to the requirement of many layers in a DBR stack to reach high reflectivity, these mirrors would cause an increased device thickness as a result of the thick DBR mirrors.
Drawing inspiration from these existing LED structures, we propose a microLED structure that combines the benefits of both GaN c-plane LED devices manufacturability and the superior device characteristics of a RC LED. Our design, the meta-microLED, will address the need for a high-volume-manufacturing-friendly device that may be used for high speed and low power data communication systems. In this structure, an active cavity is sandwiched between the top and bottom mirrors, where the bottom mirror has a high reflectivity and a lower reflectivity for the top mirror to allow topside light emission. Using a metasurface material as mirror would significantly reduce the mirror thickness and the fabrication procedure, since it does not rely on refractive index contrast of different material layers for reflection. With a single layer of metasurface material deposition and nanopattern fabrication, a highly reflective mirror can be achieved with a reduced thickness and simpler process flow. With the addition of mirrors, it is now also possible to avoid a coupling lens to efficiently couple light from the LED to the optical fiber due to the narrowed divergence angle, which could reduce manufacturing cost and improve production yield. Since the linewidth could also be less than 5 nm at peak wavelength around 450 nm, a wavelength division multiplexing data communication system could be fabricated ().
In this paper, we have developed a fabrication process for meta-microLED to demonstrate the potential of this novel device structure. The bottom mirror is designed with a distributed Bragg reflector (DBR) formed by stacking layers of nanoporous (NP) GaN and UID GaN, while the top mirror is fabricated by forming nanopatterns on a layer of TiO2 metasurface material which gives the name of meta-microLED. With the application of mirrors, the meta-microLED can show a reduction of divergence angle and emission linewidth.
2 Materials and methods
The schematics of the epitaxial structure is illustrated in Figure 1a. The proposed structure is grown epitaxially on bulk GaN substrates using metal-organic chemical vapor deposition (MOCVD) reactor. The DBR stack layers are grown by having a layer of n+-GaN with over 3 × 1019 cm-3 Si doping and a UID GaN layer. The thickness of the n+- GaN layer is 70 nm while the UID GaN layer is 45 nm. Then, the LED structure is grown on top of the DBR stack, where n-GaN layer and the p-GaN layer are each 380 nm thick, and the 4-pair MQW has a InGaN quantum well thickness of 3.2 nm with barrier thickness of 15 nm for each pair. The p-GaN layer reaches a doping level of 1 × 1020 cm-3 near the surface so that the contact resistance could be reduced when applying metal contacts. After the MOCVD growth, the Mg in p-GaN layer is activated in a 900 °C rapid thermal annealing process under N2 gas.
FIGURE 1
When designing the cavity length, the thickness of the n-GaN and p-GaN surrounding the active layer (MQW of InGaN/GaN) should satisfy the resonance condition for the desired wavelength. The cavity total length must satisfy the equation (), where is the optical mode number, is the refractive index of GaN. With a longer cavity length, a narrower angular light emission could be achieved due to the increased phase sensitivity to angle. A high optical mode was selected for the design of our meta-microLED structure in order to have a longer cavity length that can reduce the emission divergence angle. As a result, 380 nm of p-GaN and 380 nm of n-GaN are chosen in this work.
The fabricated meta-microLED structure is shown in Figure 1b. First, the bottom DBR mirror needs to be fabricated with the etching of n+-GaN layers. Electrochemical (EC) etching is performed in 0.2 M oxalic acid solution to transform the n+-GaN layers into the nanoporous GaN layers, using platinum wire as the counter electrode. Then, the sample is dry etched using the inductively coupled plasma (ICP) etching process to create the mesa structure and to isolate each device. Followed by metal evaporation to deposit the n-type and p-type GaN metal contact. 30 nm of Ni and 100 nm of Au are deposited on the p-GaN layer and then annealed at 450 °C in air to form the anode contact, and 30 nm of Ti and 120 nm of Au are also evaporated onto the n-GaN layer to form the cathode contact. A device with no mirrors is fabricated using the same procedure as a conventional LED reference device. Following the LED device fabrication, TiO2 is deposited onto the sample using DC magnetron sputtering of a TiO2 target. A layer of Al is evaporated on the TiO2 layer to be used as a hard mask for TiO2 nanopatterning. Electron-beam lithography and ICP etching is used to create the nanopatterns on Al and TiO2, and then Al is removed via Al-etchant afterwards. Finally, a thin layer of SiO2 is deposited using PECVD to passivate the sidewalls and pads are opened with the ICP etching process, and Ti/Au was evaporated again on top of the SiO2 layer to form larger probing pads.
An optical microscope and a scanning electron microscope are employed to capture the images of microLEDs and light emission, as well as the cross-sectional image of the DBR mirror and the nanopatterns of the TiO2 mirror. The current-voltage characteristics of the devices are measured using a Keysight B1505A power device analyzer. Ocean Optics Flame spectrometer is used to collect the light emission spectrum of the devices.
For the bottom mirror, a nanoporous GaN/UID GaN DBR stack is chosen which has been previously applied in VCSELs, where the NP GaN DBR mirror works by utilizing the refractive index difference between the UID GaN layer and the nanoporous GaN layer (). By stacking multiple layers of NP GaN and UID GaN layers, high reflectivity at desired wavelength can be achieved. The thickness of the NP GaN layer is determined by the requirements for the DBR, where the thickness of each DBR layer should be ¼ of the desired emission wavelength inside of the layer. In the case of NP GaN DBR, the refractive index of the NP GaN layer is determined by the porosity of the NP GaN layer. Based on the reported literature (), when the NP GaN layer has a porosity of 50%–60%, the refractive index of the NP GaN layer and refractive index of UID GaN will provide sufficient contrast for high reflectivity at the designed emission wavelength. From Figure 2a), the NP GaN layer is estimated to have a porosity of 56%, and based on effective material approximation, the effective refractive index at the target wavelength is calculated to be 1.6. According to previous literature (), a 12-pair stack would yield a DBR with 97.1% reflectivity at a central wavelength of 442.3 nm. Other publications have shown that increasing the number of pairs of DBR would increase the reflectance of the DBR mirror (). As a result, our NP GaN DBR mirror design has 15 pairs of NP GaN and UID GaN stack, and the NP GaN layer is designed to be 70 nm and the UID GaN layer to be 45 nm. Reflectivity of the 15-pair NP GaN DBR mirror is measured, where the NP GaN DBR has a reflectivity of 91% at 430 nm as shown in Figure 2c.
FIGURE 2
With different Si doping concentrations, n-GaN layers can be etched into NP GaN at different voltages. It has been previously reported that with higher doping concentration, n-GaN layers can be etched into the porous states at lower etching voltage (). Selective etching is then achieved by doping high Si concentration in the desired GaN layers and etching at a low voltage. Since there is also Si doping in the active LED structure, the Si doping concentration in the n+-GaN layer needs to be significantly higher than the doping concentration in the LED layers. Thus, a Si doping concentration of 3 × 1019 cm-3 is selected for the n+-GaN layer so etching can be performed at only 8 V without damaging the GaN layers in the active region. Several 3 µm deep trenches are also created via plasma thermal dry etching across the sample in order to expose more side wall areas of the DBR stack before EC etching. This will help increase the lateral EC, etch rate since more oxalic acid solution can come into contact with the n+-GaN due to the larger sidewall surface area, and results in a more even, etch and smoother surface.
Dielectric metasurface materials such as SiN and TiO2 have been applied towards optical amplitude and phase control due to the benefit of high-index all-dielectric nanostructures (; ). Thus, TiO2 metasurface material is chosen as the top mirror of the meta-microLED. In order to have the right reflectivity for the blue LED emission, the thickness of the TiO2 layer should be half of the cavity wavelength, which means approximately 215 nm thick TiO2 layer needs to be deposited on top of the LED structure. DC magnetron sputtering of a TiO2 target is selected as the TiO2 deposition method because of the faster deposition rate and good film quality. After TiO2 deposition, ellipsometry is used to determine the TiO2 thickness and the refractive index of the layer. AFM is used to measure the surface roughness of the TiO2 layer, and also to verify the layer thickness measured by ellipsometry. Using AFM and ellipsometry, the thickness of the deposited TiO2 layer is approximately 240 nm with 2.94 nm roughness measured over a 5✕5 μm2 area, and the refractive index of the TiO2 layer is measured to be 2.23 at 430 nm wavelength. Furthermore, dispersive phase gradient method is utilized for divergence angle and emission linewidth reduction with TiO2 metasurface mirror (; ). The dispersive phase gradient method is a metasurface design approach where the phase of light emission and the phase change with wavelength is engineered. A metasurface mirror utilizes Mie resonance in subwavelength dielectric structures to control the phase and scattering of light (). The mirror is composed of meta-atoms arranged across a surface at intervals smaller than the emission wavelength, and each meta-atom imparts a local phase shift to the reflected wavelength. In this work, the TiO2 metasurface is formed by an array of gratings as meta-atoms. Each meta-atom is defined by the width, height, material index and spacing, which together determine its optical response. By arranging these meta-atoms, a phase gradient is created across the surface that can be used to control beam steering, focusing, collimation, and tailored far-field shaping (). Also, with different dimensions of the grating, the dispersion can be engineered to control how the phase varies with wavelength, meaning the metasurface mirror can be tuned for a specific emission wavelength (). As a result, by designing and spatially arranging the TiO2 gratings, light emission direction and emitted light wavelength can be simultaneously controlled via the phase gradient and the dispersion. Thus, utilizing the dispersive phase gradient method towards the nanopattern design, the TiO2 metasurface mirror can be designed to specifically reflect the desired emission wavelength, leading to a narrower emission divergence angle and emission linewidth. In this work, the nanopatterned line grating of TiO2 is designed to have a grating width of 86 nm, the height of 215nm, and the period of 180 nm. CF4 is chosen as the, etch gas due to its good, etch selectivity between GaN and TiO2, which makes it significantly easier to, etch through the entire TiO2 layer without damaging the GaN LED structure underneath. The results of the etched TiO2 layer can be seen in the SEM images in Figure 2b, which indicate that Al hard mask and CF4 gas provided a good etching method for the nanopattern designs. From the SEM image, the line patterns also seem to be approximately 86 nm wide, which means the resulting cavity wavelength would be around 430 nm.
3 Experimental results and discussion
3.1 Light emission measurements of Meta-MicroLED
Currently, we are able to successfully fabricate meta-microLED devices that have diameters of 100 μm and 5 µm. The current-voltage graphs of the devices are measured by applying a forward bias from 0 to 5 V and recording the current response using a Keysight B1505A analyzer. The I-V characteristics are plotted and the current density of each device is calculated at the 5 V applied bias. For the reference microLEDs at 5 V, a peak current density of 2020 A/cm2 for the 100 µm device and 2400 A/cm2 for the 5 µm device is obtained. For the meta-microLEDs at 5V, the peak current is 2300 A/cm2 for the 100 µm device and 2930 A/cm2 for the 5 µm device. As illustrated in Figure 3, these devices exhibit similar electrical performances with comparable current density and turn-on voltage.
FIGURE 3
The light emission spectra of the devices are measured with a spectrometer that can be positioned at different angles while the 100 µm sized devices are injected with the same current density of 1000 A/cm2. From Figure 4a, the reference LED has a peak emission wavelength of 421.1 nm, and 427.3 nm for the meta-microLED. Emission linewidth is defined as the full width at half maximum (FWHM) of the emission peaks. The emission spectra of both devices show that the meta-microLED has an emission linewidth of 18.8 nm and reference LED has a linewidth of 24.3 nm, indicating the presence of resonance cavity could reduce linewidth by 5.5 nm. The peak emission wavelength of the meta-microLED is consistent with the cavity wavelength design, which is much closer to the intended 430 nm cavity wavelength than the reference LED. As shown in Figures 4a,b, by positioning the spectrometer at different angles, there is a change in the relative emission intensity between the reference LED and meta-microLED. With the spectrometer at 90-degrees position, the meta-microLED exhibits significantly higher light emission intensity than the reference LED, while at 45-degree the reference LED possesses higher light emission intensity instead. This shows that the meta-microLED has significantly more focused light emission in the vertical direction and the reference LED’s light emission is spread more laterally, which means the meta-microLED has a reduced emission divergence angle. Captured images of light emission with the same applied current of 1000 A/cm2 from both reference and meta-microLED are shown in Figure 4. It can be seen that the radius of light illumination on the backside of the sample is significantly larger in the reference device than the meta-microLED. Based on the radius of the backside light pattern and the thickness of the sample, we calculated the divergence angle of the reference device to be 121° and the divergence angle of meta-microLED to be 96° for the 100 µm size devices. For the 5 µm size devices, the divergence angle of the reference device is calculated to be 94° and the divergence angle of meta-microLED to be 65°. This means that the meta-microLED has effectively reduced divergence angle by 25° for the 100 µm devices and 29° for the 5 µm devices. With the reduced divergence angle and emission linewidth, these results clearly demonstrate that the application of the NP GaN DBR mirror and the TiO2 metasurface mirror help the meta-microLED emit light with a higher spectral purity.
FIGURE 4
3.2 Thermal stability of Meta-MicroLED light emission
To maintain the high intensity light emission, Joule heating could become a major concern for the design of microLEDs towards optical interconnect applications. Previous studies have shown that the LED performance can be greatly affected by the increasing junction temperature as a result of non-radiative recombination and ohmic heating by the series resistance (; ; ). Not only will the internal quantum efficiency decrease with increasing temperature resulting in a lowered light emission intensity, but also an emission wavelength shift towards red would occur at high temperature. This is because the bandgap energy of GaN and InGaN would decrease with increasing temperature as a result of the lattice expansion and the electron-phonon interactions (). When the bandgap of GaN and InGaN decreases, the photon energy would decrease which results in the emission wavelength increase. However, shifting wavelengths could be detrimental to the performance of optical interconnects and the WDM application, since the high spectral purity of the emitted light is crucial. With the presence of the resonant cavity, the meta-microLED could still help maintain higher spectral purity at high temperatures compared to the reference LED.
In order to demonstrate the thermal benefits of meta-microLED structure, the sample is heated from room temperature to 250 °C on a thermal chuck and the emission wavelength peak shift of the 100 µm sized devices are observed with the spectrophotometer. It can be seen from Figure 5 that at room temperature, the reference LED peak wavelength is approximately at 421 nm and the meta-miroLED with NP GaN DBR has peak wavelength at around 427 nm. Moreover, there is a clear reduction in emission linewidth at room temperature with meta-microLED at 14.7 nm and reference LED at 22.5 nm. As previously stated, the difference in peak emission wavelength and the narrowed linewidth at room temperature are direct results of the meta-microLED structure with the resonant cavity. After the devices are heated to 250 °C, the reference LED exhibits a peak wavelength at around 434 nm, while the meta-microLED only has a peak wavelength of 431 nm. This indicates that after being heated from room temp to 250 °C, the reference sample had a wavelength shift of 13 nm, while the meta-microLED only shifted by 4 nm. Also, the emission linewidth of meta-microLED at 250 °C is 25.4 nm, lower than the reference LED’s 32.7 nm. The significantly reduced wavelength shift is consistent with previous literature (; ). Because of the temperature dependency of the refractive index and thermal expansion of the metasurface material, the cavity resonance wavelength would also shift towards longer wavelengths at increasing temperature just as the intrinsic emission wavelength would. However, the shift of the cavity resonance would be significantly smaller than the shift of the peak wavelength emitted from the active semiconductor material. As a result, the emission wavelength from the meta-microLED at higher temperature is dominated by the cavity wavelength, which shifts significantly less with increasing temperature.
FIGURE 5
4 Conclusion
In summary, we have designed, fabricated and tested a meta-microLED with the application of NP GaN DBR mirror and TiO2 metasurface mirror for the first time. With the extra NP GaN structures and fabrication process, the meta-microLED has similar I-V characteristics as the conventional reference LED. With the addition of NP GaN DBR mirror and TiO2 metasurface mirror, the divergence angle of meta-microLED is reduced by 25° and the emission linewidth by 5 nm. Moreover, the formation of resonant cavity via the mirrors also provides additional thermal benefit where meta-microLED would have significantly more stable light emission at increased temperature as a result of the reduced wavelength shift of the cavity wavelength. This work marks a significant achievement in the GaN-based RC LED technology, and demonstrates meta-microLED’s great potential towards optical interconnect applications.
Statements
Data availability statement
The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.
Author contributions
CM: Writing – original draft, Writing – review and editing, Conceptualization, Data curation, Formal Analysis, Investigation, Methodology. RS: Writing – original draft, Writing – review and editing, Data curation, Investigation, Methodology. ZB: Writing – original draft, Writing – review and editing, Data curation, Methodology. JP: Writing – review and editing, Data curation, Methodology. SC: Writing – original draft, Writing – review and editing, Conceptualization, Funding acquisition, Investigation, Project administration, Supervision, Validation, Visualization.
Funding
The author(s) declared that financial support was received for this work and/or its publication. The author(s) declared that this work is a sponsored research program supported by Intel Corporation. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article, or the decision to submit it for publication.
Acknowledgments
We would like to thank Dr. Khaled Ahmed for the support and technical guidance, and Dr. Mozhgan Mansuri and Dr. Junyi Qiu for the insightful discussions. We would also like to thank Professor Dan Congreve and Divine Mbachu for their assistance in optical measurements.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
BabichevaV. E.EvlyukhinA. B. (2024). Mie-resonant metaphotonics. Adv. Opt. Photonics16 (3), 539–658. 10.1364/aop.510826
2
BaetsR. G.DelbekeD. G.BockstaeleR.BienstmanP. (2003). Resonant-cavity light-emitting diodes: a review. Proc. SPIE, Light-Emitting Diodes Res. Manuf. Appl. VII4996, 74–86. 10.1117/12.476588
3
ChenZ.YanS.DaneshC. (2021). MicroLED technologies and applications: characteristics, fabrication, progress, and challenges. J. Phys. D Appl. Phys.54 (12), 123001. 10.1088/1361-6463/abcfe4
4
DenBaarsS. P.FeezellD.KelchnerK.PimputkarS.PanC.-C.YenC.-C.et al (2013). Development of gallium-nitride-based light-emitting diodes — material, technology, and applications. Semicond. Sci. Technol.28 (7), 074008. 10.1016/j.actamat.2012.10.042
5
ElafandyR. T.KangJ.-H.MiC.KimT. K.KwakJ. S.HanJ. (2021). Study and application of birefringent nanoporous GaN in the polarization control of blue vertical-cavity surface-emitting lasers, american chemical society ACS photonics. ACS Photonics8 (4), 1041–1047. 10.1021/acsphotonics.1c00211
6
FattalD.LiJ.PengZ.FiorentinoM.BeausoleilR. G. (2010). Flat dielectric grating reflectors with focusing abilities. Nat. Photonics4, 466–470. 10.1038/nphoton.2010.116
7
HuangW. T.LeeT. Y.BaiY. H.WangH. C.HungY. Y.HongK. B.et al (2024). InGaN-based blue resonant cavity micro-LEDs with staggered multiple quantum Wells enabling full-color and low-crosstalk micro-LED displays. Next Nanotechnol.5, 100048. 10.1016/j.nxnano.2024.100048
8
KapurK. P.McVittieJ. P.SaraswatK. C. (2002). Technology and reliability constrained future copper interconnects—Part I: resistance modeling. IEEE Trans. Electron Devices49 (4), 590–597. 10.1109/16.992867
9
KhaidarovE.LiuZ.Paniagua-DomínguezR.HaS. T.ValuckasV.LiangX.et al (2020). Control of LED emission with functional dielectric metasurfaces. Laser and Photonics Rev.14, 1900235. 10.1002/lpor.201900235
10
LiuW.LiZ.ChengH.ChenS. (2020). Dielectric resonance-based optical metasurfaces: from fundamentals to applications. iScience23 (12), 101868. 10.1016/j.isci.2020.101868
11
LuT.-C.ChenS.-W.WuT.-T.TuP.-M.ChenC.-K.ChenC.-H.et al (2010). Continuous-wave operation of current-injected GaN vertical-cavity surface-emitting lasers at room temperature. Appl. Phys. Lett.97 (7), 071114. 10.1063/1.3483133
12
MekaweyH.ElsayedM.IsmailY.SwillamM. A. (2022). Optical interconnects finally seeing the light in silicon photonics: past the hype. Nanomaterials12 (3), 485. 10.3390/nano12030485
13
MeneghiniM.TrevisanelloL.-R.MeneghessoG.ZanoniE. (2008). A review on the reliability of GaN-based LEDs. IEEE Trans. Device Mater. Reliab.8 (2), 323–331. 10.1109/tdmr.2008.921527
14
MillerD. A. B. (2000). Rationale and challenges for optical interconnects to electronic chips. Proc. IEEE88 (6), 728–749. 10.1109/5.867687
15
MiyataM. (2024). Dispersion-engineered metasurfaces for high-sensitivity color image sensors. Opt. Rev.31 (3), 290–298. 10.1007/s10043-024-00882-8
16
NakamuraS.MukaiT.SenohM. (1994). Candela-class high-brightness InGaN/AlGaN double-heterostructure blue-light-emitting diodes. Appl. Phys. Lett.64 (13), 1687–1689. 10.1063/1.111832
17
ParkS. H.ZhangC.YuanG.ChenD.HanJ. (2015). Applications of electrochemistry for novel wide bandgap GaN devices. ECS Trans.66 (1), 0143. 10.1149/06601.0143ecst
18
SchubertE. F.WangY.-H.ChoA. Y.TuL.-W.ZydzikG. J. (1992). Resonant cavity light-emitting diode. Appl. Phys. Lett.60 (8), 921–923. 10.1063/1.106489
19
SchubertE. F.HuntN. E. J.MalikR. J.MicovicM.MillerD. L. (1996). Temperature and modulation characteristics of resonant-cavity light-emitting diodes. Journ. Light. Techn.14 (7), 1721–1729. 10.1109/50.507950
20
ShiuG.-Y.ChenK.-T.FanF.-H.HuangK.-P.HsuW.-J.DaiJ.-J.et al (2016). InGaN light-emitting diodes with an embedded nanoporous GaN distributed bragg reflectors. Sci. Rep.6, 29138. 10.1038/srep29138
21
VarshniY. P. (1967). Temperature dependence of the energy gap in semiconductors. Physica34 (1), 149–154. 10.1016/0031-8914(67)90062-6
22
WangJ. C.FangC. H.WuY. F.ChenW. J.KuoD. C.FanP. L.et al (2012). The effect of junction temperature on the optoelectrical properties of InGaN/GaN multiple quantum well light-emitting diodes. J. Luminescence132 (2), 429–433. 10.1016/j.jlumin.2011.09.001
23
WooK.BianZ.NoshinM.Pérez MartinezR.MalakoutianM.ShankarB.et al (2024). From wide to ultrawide-bandgap semiconductors for high power and high frequency electronic devices. J. Phys. Mater.7 (2), 022003. 10.1088/2515-7639/ad218b
24
WuT.LiuZ.WangY.ZhangH.YangZ.CaoW.et al (2021). All-dielectric phase-gradient metasurface performing high-efficiency anomalous transmission in the near-infrared region. Nanoscale Res. Lett.16 (1), 158. 10.1186/s11671-021-03616-w
25
YuN.CapassoF. (2014). Flat optics with designer metasurfaces. Nat. Mater.13, 139–150. 10.1038/nmat3839
26
ZhangC.ElAfandyR.HanJ. (2019). Distributed bragg reflectors for GaN-Based vertical-cavity surface-emitting lasers. Appl. Sci.9 (8), 1593. 10.3390/app9081593
Summary
Keywords
distributed bragg reflector (DBR), MicroLED, nanoporous GaN, optical interconnect, resonant cavity LED (RC LED), TiO2 metasurface material
Citation
Meng C, Soman R, Bian Z, Plutowski J and Chowdhury S (2026) Meta-MicroLEDs enabled by metasurface materials and nanoporous GaN. Front. Mater. 13:1836468. doi: 10.3389/fmats.2026.1836468
Received
22 March 2026
Revised
12 May 2026
Accepted
18 May 2026
Published
08 July 2026
Volume
13 - 2026
Edited by
Tongbo Wei, Chinese Academy of Sciences (CAS), China
Reviewed by
Meng Liu, Qilu Normal University, China
Yonghui Zhang, Hebei University of Technology, China
Updates
Copyright
© 2026 Meng, Soman, Bian, Plutowski and Chowdhury.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Chuanzhe Meng, jmeng29@stanford.edu
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.