This Letter describes the impact of shape on micro light-emitting diodes (µLEDs), analyzing 400 µm2 area µLEDs with various mesa shapes (circular, square, and stripes). Appropriate external quantum efficiency (EQE) can yield internal quantum efficiency (IQE) which decreases with increasing peripheral length of the mesas. However, light extraction efficiency (ηe) increased with increasing mesa periphery. We introduce analysis of Jpeak (the current at peak EQE) since it is proportional to the non-radiative recombination. Etching the sidewalls using tetramethylammonium hydroxide (TMAH) increased the peak EQE and decreased the sidewall dependency of Jpeak. Quantitatively, the TMAH etching reduced non-radiative surface recombination by a factor of four. Hence, shrinking µLEDs needs an understanding of the relationship between non-radiative recombination and ηe, where analyzing Jpeak can offer new insights.

We have demonstrated a fabrication process for the Ohmic contact on low-doping-density p-type GaN with nitrogen-annealed Mg. An Ohmic contact with a contact resistance of 0.158 Ω cm2 is realized on p−-GaN ([Mg] = 1.3 × 1017 cm−3). The contact resistance of p-type GaN with higher Mg concentration ([Mg]=1.0 × 1019 cm−3) can also be reduced to 2.8 × 10−5 Ω cm2. A localized contact layer is realized without any etching or regrowth damage. The mechanism underlying this reduced contact resistance is studied by scanning transmission electron microscopy with energy dispersive x-ray spectroscopy and secondary ion mass spectrometry, representing a mutual diffusion of Ga and Mg atoms on the interface. Reductions in the barrier height and surface depletion width with the nitrogen-annealed Mg layer are confirmed by XPS and Hall effect measurements qualitatively.

We demonstrated the formation of excellent 1 Ohmic contact to p-type GaN (including the plasma 2 etching-damaged p-type GaN which otherwise exhibited 3 undetectable current within ±5V) by the post-growth 4 diffusion of magnesium. The specific contact resistivity 5 on the order of 10 -4 Ω.cm 2 (extracted at V=0V) was 6 achieved on the plasma-damaged p-GaN with linear 7 current-voltage characteristics by the transfer length 8 method (TLM) measurement. The improvement in current 9 by a factor of over 10 9 was also obtained on the plasma-10 damaged p-n junction diode after the same Mg-treatment.

Herein, the operation of dopant‐free GaN‐based p‐n junctions formed by distributed polarization doping (DPD) is experimentally demonstrated and their space charge profiles and carrier transport properties are investigated. The device exhibits ideal space charge profiles explained by polarization effects and demonstrates the excellent controllability of DPD. In addition, it shows rectification and electroluminescence under forward‐biased conditions. The carrier transport properties could be explained by the conventional recombination/diffusion model used for impurity‐doped p‐n junctions. Repeatable breakdowns are also observed in all devices and the temperature‐dependent breakdown voltages reveal that the breakdowns are caused by avalanche multiplication, which is also the same as those reported in impurity‐doped GaN p‐n diodes. These results indicate that DPD is a promising doping technology for GaN‐based power devices overcoming any issues associated with conventional impurity doping.

Inductively coupled plasma–reactive ion etching (ICP–RIE)-induced damage in heavily Mg-doped p-type GaN ([Mg] = 2 × 1019 cm−3) was investigated by low-temperature photoluminescence (PL) and depth-resolved cathodoluminescence (CL) spectroscopy. From PL measurements, we found broad yellow luminescence (YL) with a maximum at around 2.2–2.3 eV, whose origin was considered to be isolated nitrogen vacancies (V N), only in etched samples. The depth-resolved CL spectroscopy revealed that the etching-induced YL was distributed up to the electron-beam penetration depth of around 200 nm at a high ICP–RIE bias power (P bias). Low-bias-power (low-P bias) ICP–RIE suppressed the YL and its depth distribution to levels similar to those of an unetched sample, and a current–voltage characteristic comparable to that of an unetched sample was obtained for a sample etched with P bias of 2.5 W.