Nanoporous III-N semiconductors made by selective sublimation for optoelectronic – NAPOLI
Nanoporous III-N semiconductors made by selective area sublimation for optoelectronics
In this project, we propose to explore a new route to make high quality GaN-based devices on Si substrates by fabricating nanoporous GaN with a sublimation approach.
Selective area sublimation and nanomasking: a new route to make high quality GaN-based devices on Si substrate.
In this project, we propose to explore a new route to make high quality GaN-based devices on Si substrates by fabricating nanoporous GaN. By controlling the pore dimensions and density we aim to improve the material quality while keeping the planar geometry. A two-step fabrication is proposed: polar GaN (0001 plane) will be grown in a 2D mode and then transformed into a nanoporous material using selective area sublimation (SAS) process.<br />The main goal of NAPOLI is to demonstrate efficient nanoporous (Ga,In)N based blue LEDs made from thin III-nitride LED structures grown on Si substrate. The target is to obtain a LED performance comparable to the one achieved on thick LED epilayers (4-5 µm) on Si substrate. Our final goal is to demonstrate thin porous LEDs (<1 µm) with an external quantum efficiency of 2% for LEDs on Si substrates and 20% for LEDs flip-chipped and with the absorbing Si substrate removed. <br />The objective is also to go deeper in the understanding of the SAS formation process, the optical and structural properties of this new class of nitride materials.
WP1- The two main tasks are the epitaxial growth of the structures (using a 2D growth mode) and the process to make these structures nanoporous. The samples are grown by metal-organic chemical vapor deposition (MOCVD) on Si(111) substrates. Different kind of structures will be grown with two main categories: simple GaN layers or InGaN/GaN and light emitting diode structures including n-type and p-type GaN layers and an InGaN/GaN multiple quantum well. The grown samples are then porosified using a self-organized or a controlled nanomask and a sublimation step at high temperature in vaccuum.
WP2- The samples are analyzed through detailed structural and optical characterizations. One of the main objectives is to determine the impact of the nanoporous fabrication process on the dislocations. 2D thin GaN layers grown on silicon substrate are affected by a large density of dislocations and making nanoporous these layers improves the room temperature photoluminescence efficiency that we suppose related to an enhanced structural quality. The main goal of the optical analysis is to discriminate between radiative and non-radiative recombination rates (internal quantum efficiency) and to determine the photoluminescence properties as a function of the nanopores size and distribution.
WP3- To achieve efficient nanoporous LEDs it is necessary to address the key fabrication challenges related to the passivation of the pore high specific surface and to the transparent top contacting. In particular, we explore new contact strategies using layered materials such as graphene. Different processing approaches will be evaluated by measuring the electrical characteristics, electroluminescence (EL) spectra and the external quantum efficiency. A flip-chip approach with wafer bonding to a metal holder and Si substrate removal is also planned. This is the usual approach to get efficient blue LEDs on Si substrates because silicon is absorbing the visible light.
WP1- GaN simple layers, InGaN/GaN quantum wells, and several runs of blue light emitting diode structures with an InGaN/GaN multiple quantum well active region were grown by MOCVD. These structures were then porosified in a molecular beam epitaxy reactor. The porosity of the samples is varied depending on the Si exposure time of the sample's surface. Nanoporous LEDs with a porosity varying from 0.15 to 0.5 were achieved and delivered to the partners for the optical and structural analysis and for device processing.
WP2- For TEM experiments it was necessary to set up a specific protocol for the sample preparation enabling close to 100% filling of the pores. The results obtained following this optimization are conclusive: from the high resolution TEM observations, it comes out that the remaining GaN after porosification is free of extended defects. Nanoporous LEDs were characterized by temperature and time-resolved photoluminescence experiments. The measured decay-times are typical for InGaN/GaN multiple quantum wells and the porosity does not induce a strong modification of the Purcell coefficient as observed for porous GaN. The current conclusion is that the recombination mechanisms in the InGaN quantum wells are dominated by the effects of localization due to In composition and quantum well thickness fluctuations.
WP3- The LED fabrication required a specific technological process development to account of the nanoporosity. Several approaches were attempted. The best results were achieved using the process described in the following. First, the pores were filled with parylen. A dry-etching step was then necessary to grant access to the surface for the deposition of ITO contacts defined by optical lithography. These first porous LEDs made by sublimation were characterized by EBIC microscopy and electroluminescence. Room temperature CW operation is demonstrated at a wavelength of 470 nm. The optimization of flip-chip fabrication process is on-going.
The plan view and cross section samples have been investigated using atomic resolution high annular dark field imaging. Although the number of samples is still limited in sublimation conditions, these observations demonstrate clearly that the remaining GaN after porosification can be totally free of extended defects. We showed before the beginning of the project that the photoluminescence efficiency at room temperature of GaN thin layers on Si (affected by a large density of dislocations) can be improved by 3 orders of magnitude after porosification using selective area sublimation. The present result is a clear explanation of the reason of this improvement.
Two issues were identified regarding the electrical injection in porous light emitting diodes:
- the crack density affecting the initial LED structures becomes even larger after porosification,
- the electrical injection in porous LEDs is difficult to control (difficulty to find the optimal insulation : either a short-circuit is obtained for insufficient insulation or the device surface is insulating, which prevent the current injection).
The cracking issue can be solved by using thin AlGaN interlayers to compensate the tensile strain in the LED structure coming from the thermal coefficient lattice mismatch with the Si substrate.
Several routes will be explored to limit the problems of electrical injection in the porous LEDs:
- the porosity degree can be limited (target ~0.25),
- the p-GaN layer of the initial LED structure can be grown thicker (to avoid eventual shortcuts induced by metal deposition during the process),
- the porous LED structure can be overgrown by an additional coalesced p-GaN layer, still to improve electrical injection.
- Ngo T H, Gil B, Shubina T V, Damilano B, Vezian S, Valvin P and Massies J 2018 Enhanced excitonic emission efficiency in porous GaN Scientific Reports 8
- Damilano B, Coulon P-M, Vézian S, Brändli V, Duboz J-Y, Massies J and Shields P A 2019 Top-down fabrication of GaN nano-laser arrays by displacement Talbot lithography and selective area sublimation Appl. Phys. Express 12 045007
- Sergent S, Damilano B, Vézian S, Chenot S, Takiguchi M, Tsuchizawa T, Taniyama H and Notomi M 2019 Subliming GaN into Ordered Nanowire Arrays for Ultraviolet and Visible Nanophotonics ACS Photonics 6 3321–30
- Coulon P-M, Damilano B, Alloing B, Chausse P, Walde S, Enslin J, Armstrong R, Vézian S, Hagedorn S, Wernicke T, Massies J, Zúñiga-Pérez J, Weyers M, Kneissl M and Shields P A 2019 Displacement Talbot lithography for nano-engineering of III-nitride materials Microsyst Nanoeng 5 52
GaN compounds grown on sapphire are currently at the heart of solid-state lighting. However, sapphire substrates present limitations in terms of size, price, thermal conductivity, etc. Therefore, in the last decade a large research effort has been devoted to the growth of GaN on Si. The use of Si rise stress and quality issues necessitate today the growth of complex buffer layer structures hindering the industrial development. We propose to explore a new and simple route based on a selective area sublimation process to make high quality (Ga,In)N devices on Si substrates. Indeed, we have recently demonstrated that this process gives rise to thin nanoporous (GaIn)N layers for which the optical properties are considerably improved. The main objective of the project is to demonstrate that nanoporous nitride layers can be used to fabricate efficient optoelectronic devices on Si substrates.
Project coordination
Benjamin Damilano (Centre de recherche sur l'hétéroepitaxie et ses applications)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
Partnership
L2C Laboratoire Charles Coulomb
UPSud / C2N Université Paris Sud / Centre de Nanosciences et de Nanotechnologies
CIMAP CENTRE DE RECHERCHE SUR LES IONS, LES MATÉRIAUX ET LA PHOTONIQUE
CRHEA Centre de recherche sur l'hétéroepitaxie et ses applications
Help of the ANR 474,315 euros
Beginning and duration of the scientific project:
September 2018
- 42 Months