Jie Sun1, Åsa Haglund1, August Yurgens1, Kun Xu2, Chen Xu2, Weiling Guo2, 1Mikroteknologi och Nanovetenskap, Chalmers Tekniska Högskola AB, Göteborg 41296, Sweden; 2Key Laboratory of Optoelectronics Technology, Beijing University of Technology, Beijing 100124, China.
 Kun Xu, et al, Applied Physics Letters, 103, 222105(1-5), (2013)
 Kun Xu, et al, Applied Physics Letters, 102, 162102(1-5), (2013)
 Jie Sun, et al, IEEE Transactions on Semiconductor Manufacturing, 25, 494-501, (2012)
GaN compounds are widely used to fabricate light emitting diodes (LEDs) covering the spectrum from yellow to ultraviolet. However, due to its high refractive index, photons generated by active regions can be totally reflected at the GaN/air interface easily. To solve this problem, ordered and dense GaN light emitting nanorods are studied with polycrystalline graphene grown by chemical vapor deposition (CVD) as suspended transparent electrodes . As the substitute of indium tin oxide (ITO), the graphene avoids complex processing to fill up the gaps between nanorods and subsequent surface flattening and offers high conductivity to improve the carrier injection. The as fabricated devices have 32% improvement in light output power compared to conventional planar GaN-graphene diodes, mainly due to the much more enlarged light emitting areas .
But, due to the Fermi level mismatch between the graphene and GaN, their electrical contact is not ohmic. To understand this issue further, CVD graphene is used in (planar) GaN LEDs as transparent electrodes, where 7–10 nm ITO contact layer is inserted between the graphene and p-GaN to enhance hole injection . Devices with forward voltage and transparency comparable to those using traditional 240 nm ITO are achieved with better ultraviolet performances . This result indicates that the large turn-on voltage can be indeed attributed to the poor graphene-GaN contact, which can be solved by the thin ITO interlayer.
However, the ITO interlayer is not a sustainable solution due to the scarcity of indium resources. Therefore, we suggest depositing graphene directly on GaN by CVD. The graphene-GaN interface is produced in high temperature and high vacuum CVD chamber, resulting in improved electrical properties. Furthermore, in situ doping of graphene can be carried out which will tune the Fermi level further to match that of p-GaN. Also, it is a reproducible and scalable technique, getting rid of all the uncertainty and irreproducibility associated with the complex transfer process of CVD graphene. Some preliminary results will be presented regarding the direct growth method , indicating its promising future in real industrial applications.
Collaboration for joint projects is warmly welcome (e.g. EU and national funds application). Contacts: Jie.firstname.lastname@example.org