IEEE TRANSACTIONS ON ELECTRON DEVICES, VOL. 60, NO. 9, SEPTEMBER 2013 2821 GaN-Based Dual-Color LEDs With p -Type Insertion Layer for Controlling the Ratio of Two-Color Intensities Kai-Lun Chi, Shu-Ting Yeh, Yu-Hsiang Yeh, Kun-Yan Lin, Jin-Wei Shi, Yuh-Renn Wu, Ming Lun Lee, and Jinn-Kong Sheu Abstract—In this paper, a novel GaN-based dual-color LED for phosphor-free white-light generation has been demonstrated. By inserting p-type layers with different p-type doping density and thickness into active regions of dual-color GaN LEDs, we can control the ratio of output light intensities from quantum- wells near n- and p-sides. With an optimum sheet charge density of such insertion layer, the intensities of these two colors can be balanced under a much lower driving-current density (45 versus 450 A/cm 2 ) compared with that of reference device without such insertion layer. A 2-D finite-element Poisson and drift–diffusion self-consistent solver including the indium fluctuation is used to design and simulate the device performances. The experimental and simulation results match very well. Index Terms— Carrier dynamic, efficiency droop, gallium nitride, LEDs. I. I NTRODUCTION G aN-BASED LEDs have had a huge impact on the solid- state lighting market. Commercial white-light LEDs are usually composed of blue GaN LEDs and a mixture of red and yellow phosphors. However, the internal quantum efficiency may be degraded because of the loss during the processes of optical pumping and re-emission [1]. The utilization of GaN-based multiple quantum-wells (MQWs) with dual center wavelengths [2]–[4], Si and Zn codoped active layers [5], and carbon implantation [6] have made it possible to generate white-light without needing to use phosphor wavelength con- verters. However, for the case of dual wavelengths, MQWs, the spectrum of output light would only concentrate on the central wavelength of MQW near the p-side cladding layers Manuscript received May 14, 2013; revised June 20, 2013; accepted July 3, 2013. Date of publication July 22, 2013; date of current version August 19, 2013. This work was supported by the National Science Council of Taiwan under Grant NSC-96-2221-E-008-106-MY3, Grant 97-2221-E-006-242-MY3, Grant NSC 102-2221-E-002-194-MY3, and Grant NSC 99-2221-E-002-058- MY3. The review of this paper was arranged by Editor E. G. Johnson. K.-L. Chi, K.-Y. Lin, and J.-W. Shi are with the Department of Electrical Engineering, National Central University, Taoyuan 320, Taiwan (e-mail: porpoise5233@msn.com; jumping2395@gmail.com; jwshi@ee.ncu.edu.tw). S.-T. Yeh and Y.-R. Wu are with Institute of Photonics and Opto- electronics, National Taiwan University, Taipei 10617, Taiwan (e-mail: pipi0923.eo96@g2.nctu.edu.tw; yrwu@cc.ee.ntu.edu.tw). Y.-H. Yeh and J.-K. Sheu are with the Department of Photon- ics, National Cheng-Kung University, Tainan 70101, Taiwan (e-mail: l78001119@mail.ncku.edu.tw; jksheu@mail.ncku.edu.tw). M. L. Lee is with the Department of Electro-Optical Engineer- ing, Southern Taiwan University, Tainan 71001, Taiwan (e-mail: minglun@mail.stust.edu.tw). Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TED.2013.2272803 [2]–[4] and the dual wavelength emission with (near) balanced intensities only appears under a very high bias current density. This is because the small GaN hole mobility would accompany a small diffusion length and lead to the accumulation of injected hole near p-side in the GaN-based active layers [7]. The cascade LED structure would be a good solution to uniform the internal carrier distribution among different MQWs [8], [9]. However, the values of turn-on voltage and differential resistance of LED would be linear proportional to the number of cascade units [8], [9]. In our previous work, we have demonstrated a transverse junction LED structure [10]–[12] which allowed for uniform injection of carriers among MQWs with different center wavelengths. However, the additional regrowth process would degrade the Ohmic contacts and I –V curve performance of LEDs [11], [12]. In this paper, we demonstrated a novel dual-color LED structure. Compared with our previous work [13], a more detailed study in various device structures and device modeling techniques have been reported in this paper. By inserting an additional p-type GaN layer in the active region with different doping density and thickness, we can manipulate the ratio of output light intensi- ties from quantum-wells near n- and p-sides. With an optimum p-type sheet charge density, the intensities of these two colors can be balanced under a much lower driving-current density (45 versus 450 A/cm 2 ) compared with that of reference device without such an insertion layer. This result is also confirmed with the simulation work, which is based on the 2-D finite- element Poisson and drift–diffusion self-consistent solver including the indium fluctuation [14], presented in this paper. II. DEVICE STRUCTURE AND SIMULATION The epi-layer structures of the demonstrated dual-color LEDs is shown in Fig. 1(a) and the inset shows the picture of fabricated devices, which have a 250 μm active diameter. The z -axis, which specifies on the epi-layer structure, is used for our device simulation as will be discussed latter in Figs. 6–8. Three different epi-layer structures (A–C) have been fabricated and tested. The active region of such three structures is composed of MQWs with two different central wavelengths at 470 (blue) and 430 nm (violet), which is near topmost p- and bottom n-type cladding layers, respectively. The thickness of each In x Ga 1-x N well and each GaN barrier in our active region is the same as 2 and 12 nm, respectively. The number of blue and violet MQW pairs is the same as 0018-9383 © 2013 IEEE