Chapter 4 GaN based MCLEDs using ITO as transparent contact layer
4.6 Summary
In summary, 3λ GaN-based MCLED using ITO as transparent contact layer was
fabricated and measured. The device structure composed of high-reflectivity AlN/GaN bottom DBRs (95%) and SiO2/Ta2O5 top DBRs (99%). The turn on voltage of the fabricated device is a comparable value with Ni/Au device to be about 3.4V and 530Ω, respectively. On the other hand, the emission peak wavelength of the ITO MCLED was located at 458nm with a narrow line-width of 2nm. Compared to the emission spectrum of the conventional Ni/Au device, the device using ITO as transparent contact layer shows a relatively excellent line-width and possesses high Q factor of 229, which is wholly consistent with our prediction by theoretical calculation. Moreover, the “bright spots” within the emission aperture were also discussed.
We found the Q factors at the bright spots are relatively higher than those within the dark regions, even as high as 894. The Q factor shows the best value compared with that of MCLED published in the recent literatures. In conclusion, the improvement of Q factor proves that non-absorbed transparent contact indeed plays an important role to fabricate this kind of high Q device like VCSEL.
Chapter5
Mg
+ion implantation for current confinement in GaN-based Micro-cavity Light Emitting Diodes (MCLEDs)
5.1 Recent Status
To fabricate the first GaN based VCSEL, the subject of current confinement is important and required to be overcome. In our conventional device, we use SiNx insulating layer to define the aperture size and to confine the current spreading. The schematic diagram of the conventional device is shown in Figure 5.1(a), and the emission image from top view is also shown in Figure 5.1(b). The leakage-current-induced light could be obviously observed near the mesa side walls in our conventional device. The leakage current might be resulted from the poor quality of SiNx film or carriers flow along P-GaN to sidewall before crossing p-n junction. To improve current confinement, replacing SiNx insulating layer with another way is required. Up to now, several outstanding reviews have appeared in the literature addressing various aspects of ion implantation into GaN. Recently, ion implantation becomes a popular and flexible way for current confinement, such as GaAs VCSEL and MOS transistors. In this chapter, we discuss Mg+ ion implantation for GaN based MCLED, and the results were also compared with the conventional device to understand the current confinement effect.
Ti/Al/Ni/Au n-contact
ITO Ni/Au p-contact
50um
n-GaN
25 pairs of AlN/GaN DBR stack
SiNx insulating layer
Figure 5.1 (a) The schematic diagram of conventional GaN based MCLED. (b) The emission image from top view
5.2 Simulation Design and SIMS results of Mg+ ion implantation in GaN 5.2.1 Introduction to TRIM simulation software
TRIM, a simulation program based on Monte Carlo method, is a common program to simulate ion range and atomic displacement of ion bombardment. For example, Figure 5.2 shows calculated profiles of vacancies produced in GaN by implantation with 40 keV C, 100 keV Au, and 300 keV Au ions. It is seen from this figure that ion implantation causes
Gaussian-like profiles of atomic displacements. The atomic displacements are determined by ion mass and energy, as illustrated in Figure 5.2. Such theoretical calculation of atomic displacements only take into account only ballistic processes and completely neglect dynamic annealing (i.e. defect interaction processes). In the following, we simply introduce the
interface of the TRIM simulation software in Figure 5.3. The information which we must key in the interface is the ion data and target data. The ion data is set for giving the information ion and includes the element and bombardment energy. The target data is set for providing our sample structure which would like to be implanted. However, simple point defects, which survive after quenching of collision cascades, may migrate through the lattice and experience annihilation and cluster formation. As a result, experimental damage-depth profiles caused by ion bombardment may be not consistent with predictions considering collisional processes. It is well-known that dynamic annealing processes in solids under ion bombardment highly depends on implant conditions such as ion mass, energy, dose, substrate temperature, and beam flux. In contrast to our ability to calculate atomic displacements in solids, lattice disorder produced by ion bombardment is usually difficult to predict. Anyway, the TRIM simulation software provides us an effective method to predict the ion range before ion implantation.
Figure 5.2 The TRIM profiles of the total number of vacancies produced in GaN by 40 keV C, 100 keV Au, and 300 keV Au ions
Figure 5.3 The interface of the TRIM simulation software 5.2.2 Introduction to Secondary Ion Mass Spectrometer (SIMS)
SIMS is a measurement technique that is being used for the compositional analysis of small samples. In a SIMS instrument a high energy primary ion beam is directed at an area of the sample whose composition is to be determined The interaction of the primary ions with the sample surface has three major effects: (1) It leads to a mixing of the upper layers of the sample, resulting in an amorphization of the surface; (2) atoms from the primary ion beam are implanted in the sample and (3) some secondary particles (atoms and small molecules) are ejected from sample. Among the ejected particles are electrically neutral, as well as positively and negatively charged species. Charged particles of one polarity can then be extracted from the sputtering area with the help of an electrical field between the sample and an extraction lens. These accelerated secondary ions constitute a secondary ion beam which is then led into a mass spectrometer. There, the secondary ions are sorted by mass (and energy) and finally counted in an ion detector. The count rates of different secondary ion species give information about the composition of the sample in the sputtered area. Since the size of the sputtered area depends only on the primary ion beam diameter, which typically is in the order of
micro-meter, a SIMS analysis has a relatively high lateral resolution. SIMS can be used for
practically all elements of the periodic table, including hydrogen. SIMS allows the routine measurement of many trace elements at very low concentration. And since ions of different mass are measured separately, SIMS is ideally suited for the study of isotopic compositions of small samples.
5.2.3 TRIM Simulation and SIMS results of ion implantation
The TRIM simulation and its corresponding SIMS result of the device with 80keV Mg with does of 2E15 are both shown in the Figure 5.4(a), (b). The SIMS result shows the ion range of the device is about 80nm from the top surface of p-GaN, which corresponds with the TRIM simulation result. However, the result shows the implanted ions pass through the 100nm thick p-GaN and stop in the MQWs. Since the damage induced defect inside the MQWs might affect the illuminant efficiency of our device, using SiNx film as buffer layer during the implantation process is required and a feasible way to avoid this problem.
(a) (b)
Figure 5.5 shows the SIMS data of Mg+ ion implanted sample with 50nm/100nm thick SiNx buffer layer. The ion ranges are 46nm and 15nm from the top surface of p-GaN layer, which are almost consistent with the TRIM simulation profiles, respectively. However, the range of 80keV Mg+ ion implanted sample with SiNx buffer layer of 50nm shows the ions pass through the 100nm thick p-GaN and stop in the MQW, whereas the implanted induced damage is almost rest in the p-GaN in the implanted sample with SiNx buffer layer of 100nm.
In order to avoid damage in MQWs, using 100nm thick SiNx as buffer layer before 80keV Mg+ ion implantation process might be a feasible way to fabricate implanted device.
Figure 5.4 (a) Depth profiles of the distribution of the damage created in p-GaN during the implant process with 80 keV Mg ions. (b) TRIM simulation under the same condition
0 1000 2000 3000 4000 5000 6000 7000 MQW
Mg
Depth (A)
TRIM Simulation
800A
In SIMS
TRIM Simulation SIMS
Figure 5.5 The SIMS profiles and the corresponding TRIM simulation of the damage created in MCLED structures using 50nm/100nm thick SiNx buffer layer during the implantation process of 80keV Mg ions with a does of 2E15
0 1000 2000 3000 4000 5000 6000 7000
Mg
Depth (A) p-GaN MQW
460A
In
Si3N4film 500A
0 1000 2000 3000 4000 5000 6000 7000 p-GaN
Mg
MQW 150A
Depth (A)
Si
3N
4film 1000A
In
5.3 Fabrication of GaN based MCLED using Mg+ ion implantation for current confinement
5.3.1 Process Procedure
Initial clean (I.C.) and photolithography technique
During process of GaN-based MCLED, two basic skills will be frequently used. One is the initial clean (I.C.), the other is photolithography technique. The purpose of the I.C. is to remove the small particle, and organism on the sample surface. The steps of I.C. are described as below.
Degreasing by ultrasonic baths in acetone (ACE) 5min.
Dipping by ultrasonic baths in isopropyl alcohol (IPA) 5min for organism removed.
Rising in de-ionized water (D.I. water) 5min for surface clean.
Blowing with N2 gas for surface drying.
Baking by hot plate 120oC, 5min, for wafer drying.
The purpose of the photolithography is to transfer the pattern drawn on the mask to the photoresist (PR) on the wafer. In the process of photolithography, a special positive
photoresist AZ 5214E was used. Although it is positive photoresist (and may even be used in that way), it is capable of image reversal (IR) resulting in a negative pattern of the mask. In fact AZ 5214E is almost exclusively used in the IR-mode which is proper to be used in the lift-off process. Both positive exposure and IR exposure photolithography technique were employed in the fabrication of the MCLED. These photolithography techniques are described as below.
Positive exposure technique
8. Spin coating by photoresist: AZ 5214E.
a. First step: 1000 rpm for 10sec.
b. Second step: 3500 rpm for 30sec.
9. Soft bake: hot plate 90oC, 90sec.
10. Alignment and exposure: 23sec.
11. Development: dipping in AZ-300 for 30sec.
12. Fixing: rising in D.I. water 30sec.
13. Blowing with N2 gas for drying.
14. Hard bake: hot plate 120oC, 4min.
IR exposure technique
10. Spin coating by photoresist: AZ 5214E.
a. first step : 1000 rpm for 10sec.
b. second step : 3500 rpm for30sec.
11. Soft bake: hot plate 90oC, 90sec.
12. Alignment and exposure: 6sec.
13. Hard bake: hot plate 120oC, 1min50sec.
14. Flood exposure: 57sec.
15. Development: dipping in AZ-300 for 30sec.
16. Fixing: rising in D.I. water 30sec.
17. Blowing with N2 gas for drying.
18. Hard bake: hot plate 120oC, 4min.
Process flowchart
Figure 5.6 show the schematic diagrams of nitride structure of MCLED grown by MOCVD. The Mg+ implanted MCLED was fabricated by seven process steps. In the beginning, SiO2 mesa etching mask and ICP dry etching machine were used to define the mesa region, as shown in Figure 5.7. In Figure 5.8, the 100nm thick SiNx film as buffer layer was deposited using PECVD system. In Figure 5.9, the Au (200nm) was deposited to serve as blocking metal using E-gun evaporator and also defined as aperture size. The aperture size is defined as 15μm、20μm 、25μm、30μm、35μm、40μm. Then, the p-GaN film was implanted by 80keV Mg ions with a dose of 2×1015 cm-2. Then, we use KI and BOE to
remove Au and SiNx film, respectively, as shown in Figure 5.10. In Figure 511, the indium tin oxide (ITO) of 240nm deposited on the sample using E-gun evaporation system and annealed at 525oC under nitrogen ambient was used for current spreading layer, as shown in Figure 5.12. Finally, the Ti/Al/Ni/Au (20/150/20/150 nm) and Ni/Au (20/150 nm) were deposited to serve as n-type and p-type electrode, respectively, as shown in Figure 5.13 and Figure 5.14.
n-GaN ~ 380nm
25 pairs of AlN/GaN DBR stack
10 pairs InGaN(2.5nm)/GaN(7.5nm) MQW p-GaN ~ 100nm
Sapphire
Figure 5.6 The 2D schematic diagram of nitride structure of MCLED grown by MOCVD
Buffer layer : SiNx film 100 nm
Figure 5.8 2nd step of process: buffer layer Figure 5.7 1st step of process: mesa
Blocking metal : Au 2000A Blocking metal : Au 2000A Blocking metal : Au 2000A
Figure 5.9 3rd step of process: blocking metal
Mg+ : Energy : 80Kev , Dose : 2E15 Mg+ : Energy : 80Kev , Dose : 2E15
Figure 5.10 4th step of process: Implant
15,20,25,30,35,40μm
Figure 5.11 Remove Au metal and SiNx film
ITO ITO
Figure 5.12 5th step of process: TCL
Ti/Al/Ni/Au n-contact
Figure 5.13 6th step of process: N-contact
Ti/Al/Ni/Au n-contact
Figure 5.14 7th step of process: P-contact Table 5.1 Process flowchart
Step Process Conditions
1 Mesa
(1) Deposition of 300nm SiO2 by PECVD.
(2) Definition of mesa pattern by photography (positive exposure).
(3) Dry etching by ICP to define the SiO2 mesa etching mask.
(4) Dry etching by ICP to form the mesa.
(5) Remove SiO2 mesa etching mask by BOE.
2 Buffer layer (1) Deposition of 100nm SiNx by PECVD.
3 Blocking metal
(2) Definition of Implantation pattern by photography (IR exposure).
(3) Deposition of Au (200nm) using E-gun evaporator and lift-off.
4
(4) After implantation, remove Au and SiNx by KI/I2 etchant and BOE, respectively.
5 TCL
(1) Deposition of ITO (240nm) using E-gun evaporator.
(2) Definition of TCL pattern by photography (positive exposure).
(3) Wet etching ITO by ITO etchant.
(4) ITO annealing at 525oC , 15min, under N2 ambient.
6 N-contact
(1) Definition of N-contact pattern by photography (IR exposure).
(2) Deposition of Ti/Al/Ni/Au (20nm/150nm/20nm/150nm) using E-gun evaporator and lift-off.
7 P-contact
(1) Definition of P-contact pattern by photography (IR exposure).
(2) Deposition of Ni/Au (20/150nm) using E-gun evaporator and Lift-off.
5.4 Characteristics of GaN based MCLED using Mg ion implantation for current confinement
5.4.1 The electrical characteristics measurement setup
The electrical characteristics of ion implanted and conventional MCLEDs were both measured by probe station system and evaluated by injecting different current density. The device is driven by Keithley 238 CW current source, and its light output from top view could be observed by CCD. Current-voltage (I-V) measurements were performed using the probe station and the data could be fee-backed to the computer from these facilities by a GPIB card.
Figure 5.15 shows the electrical measurement system.
5.4.2 The emission images of implanted and conventional device
Figure 5.15 Probe station measurement instrument setup
Figure 5.16 shows the emission images of Mg implanted MCLED with six aperture size.
The aperture size is defined as 5μm, 10 μm, 15 μm, 20 μm, 25 μm and 30μm. We could clearly find light emission didn’t exist in the region near side walls of these six devices.
Figure 5.17 shows the emission images of the conventional and implanted device under high magnification at the same current injection. Compared with these two images, the main difference was light emission exists in the region near side walls of the conventional MCLED.
To confirm the effect of current confinement in implanted device, Figure 5.18 shows the emission image and structure of implanted MCLED after the deposition of the ITO transparent contact layer. We can distinctly find the different light emission sizes are fully
controlled and Figure 5.18(b) shows it is totally consistent with our design. It powerfully proves that using Mg+ implantation is effective than using SiNx insulating layer for current confinement. The absence of side wall emission of implanted MCLED not only confirms the existence of leakage current in our conventional device but also verifies the current
confinement was successful.
5
5µµmm 110µ0µmm 15µ15µmm 20µ20µmm 2525µµmm 30µ30µmm
Figure 5.16 Emission images of Mg implanted MCLED with six aperture size.
(a) (b)
Conventional
Conventional Mg Mg
++ion implanted ion implanted
Ti/Al/Ni/Au n-contact
ITO Ni/Au p-contact
40um n-GaN
pairs of AlN/GaN DBR stack
SiNx insulating layer
Ti/Al/Ni/Au n-contact ITO
40um n-GaN
25 pairs of AlN/GaN DBR stack
Ni/Au p-contact
25
Figure 5.17 The emission images and structures of Mg implanted and conventional MCLEDs. (a) Conventional MCLED. (b) Implanted MCLED
(a)
Mg+ implanted sample
15
15µµmm 20µ20µmm 2525µµmm 3030µµmm 35µ35µmm 40µ40µmm
(b) (c)
55 55µµmm
40 40µµmm
10 10µµmm
ITO
15、20、25、30、35、40um
n-GaN
25 pairs of AlN/GaN DBR stack
Mg implanted region
Figure 5.18 (a) The emission images of Mg+ implanted MCLED after deposition of ITO transparent contact layer. (b) The emission image of aperture size of 40μm under high magnification. (c) The structure of implanted MCLED after deposition of ITO transparent contact layer
5.4.3 The I-V curves of implanted and conventional device
The current-voltage (I-V) characteristics of implanted and conventional device are both shown in Figure 5.19. The turn on voltage and resistance of the conventional MCLED was about 3.9V and 120Ω, respectively. However, the implanted device shows the same turn on voltage but slightly higher resistance compared with the conventional device. Because ion bombardment could be served as a kind of crystal destruction, such results are realizable and acceptable. In order to make the electrical characteristic to be optimization, post annealing of the implanted device is required to remove the implantation-produced lattice disorder in our future work.
5.5 Summary
GaN based micro-cavity light emitting diode using Mg ion implantation for current confinement is fabricated and measured. The implantation condition is 80keV Mg with does of 2E15 is determined. Moreover, 100nm thick SiNx as buffer layer is required for avoiding damage induced defect in MQWs. The absence of side wall emission of implanted MCLED not only confirms the existence of leakage current in our conventional device but also verifies the current confinement was successful. These characteristic suggests that Mg ion implantation during the process of GaN-based MCLED is an effective and feasible way for current confinement.
0 5 10 15 20
0 5 10 15 20
Conventional MCLED Implanted MCLED
Voltage (V)
Current (mA)
Figure 5.19 The I-V curve of the conventional and Mg implanted MCLED.
Chapter6 Conclusions
6.1 Conclusions
GaN-based MCLED using ITO as transparent contact layer
3λ GaN-based MCLED using ITO as transparent contact layer was fabricated and measured. The device structure composed of high-reflectivity AlN/GaN bottom DBRs (95%) and SiO2/Ta2O5 top DBRs (99%). The turn on voltage of the fabricated device is a comparable value with Ni/Au device to be about 3.4V and 530Ω, respectively. The emission peak wavelength of the ITO MCLED was located at 458nm with a narrow line-width of 2nm.
Compared to the emission spectrum of the conventional Ni/Au device, the device using ITO as transparent contact layer shows a relatively narrow line-width and possesses excellent Q factor of 229, which is wholly consistent with our prediction by theoretical calculation.
Moreover, the “bright spots” within the emission aperture were also discussed. We found the Q factors at the bright spots are relatively higher than those within the dark regions, even as high as 894. The Q factor is the best value compared with that of MCLED published in the recent literatures. In conclusion, the improvement of Q factor proves that non-absorbed transparent contact indeed plays an important role to fabricate this kind of high Q device like VCSEL.
Using Mg+ ion implantation for current confinement in GaN-based MCLED
GaN based micro-cavity light emitting diode using Mg ion implantation for current confinement is fabricated and measured. The implantation condition is 80keV Mg with does of 2E15 is determined. Moreover, 100nm thick SiNx as buffer layer is required for avoiding damage induced defect in MQWs. The absence of side wall emission of implanted MCLED not only confirms the existence of leakage current in our conventional device but also verifies the current confinement was successful. These characteristic suggests that Mg ion implantation during the process of GaN-based MCLED is an effective and feasible way for current confinement.
Reference
[1] S. Nakamura, M. Senoh, N. Iwasa and S. Nagahama, Jpn. J. Appl. Phys., 34, L797 (1995) [2] S. Nakamura, T. Mukai and M. Senoh, Appl. Phys. Lett., 64, 1687 (1994)
[3] S. Nakamura, M. Senoh, S. Nagahama, N. Iwasa, T. Yamada, T. Matsushita, Y. Sugimoto and H. Kiyoku, Appl. Phys. Lett., 70, 868 (1997)
[4] S. Nakamura, Science, 281, 956 (1998)
[5] S. Nakamura, S. Pearton and G. Fasol, The Blue Laser Diode: The Complete Story, Springer-Verlag, Berlin, (2000)
[6] S. Nakamura and S. F. Chichibu, Introduction to Nitride Semiconductor Blue Lasers and Light Emitting Diodes, Taylor & Francis, (2000)
[7] Hadis Morkoc, Nitride Semiconductors and Devices, Springer-Verlag, Berlin, (1999) [8] T. Someya, R.Werner, A. Forchel, M. Catalano, R. Cingolani and Y.Arakawa, Science, 285,
1905 (1999)
[9]T. Tawara, H. Gotoh, T. Akasaka, N. Kobayashi and T. Saitoh, Appl. Phys. Lett., 83, 830 (2003)
[10]Y.-K. Song, H. Zhou, M. Diagne, A. V. Nurmikko, R. P. Schneider, Jr., C. P. Kuo, M. R.
Krames, R. S. Kern, C. Carter-Coman and F. A. Kish, Appl. Phys. Lett., 76, 1662 (2000) [11] H. Zhou, M. Diagne, E. Makarona, A. V. Nurmikko, J. Han, K. E. Waldrip and J. J.
Figiel, Electron. Lett., 36,1777 (2000)
[12] T. Someya, K. Tachibana, J. Lee, T, Kamiya and Y. Arakawa, Jpn. J. Appl. Phys., 37, L1424 (1998)
[13] J. M. Redwing, D. A. S. Loeber, N. G. Anderson, M. A. Tischler, and J. S. Flynn, Appl.
Phys. Lett., 69, 1 (1996)
[14] S. H. Park, J. Kim, H. Jeon, T. Sakong, S. N. Lee, S. Chae, Y. Park, C. H. Jeong, G. Y.
Yeom, and Y. H. Cho, Appl. Phys. Lett., 83, 2121 (2003)
[15] M. Diagne, Y. He, H. Zhou, E. Makarona, A. V. Nurmikko, J. Han, K. E. Waldrip, J. J.
Figiel, T. Takeuchi, and M. Krames, Appl. Phys. Lett., 79, 3720 (2001)
Figiel, T. Takeuchi, and M. Krames, Appl. Phys. Lett., 79, 3720 (2001)