• 沒有找到結果。

Performance improvement and mechanism of chlorine-treated InGaN-GaN light-emitting diodes

N/A
N/A
Protected

Academic year: 2022

Share "Performance improvement and mechanism of chlorine-treated InGaN-GaN light-emitting diodes"

Copied!
3
0
0

加載中.... (立即查看全文)

全文

(1)

Performance Improvement and Mechanism of Chlorine-Treated InGaN–GaN Light-Emitting Diodes

Po-Sung Chen, Chi-Sen Lee, Jheng-Tai Yan, and Ching-Ting Lee z

Institute of Microelectronics, Department of Electrical Engineering, Advanced Optoelectronic Technology Center, National Cheng Kung University, Taiwan

The electrical and optical performances of multiple-quantum-well 共MQW兲 InGaN/GaN light-emitting diodes 共LEDs兲 were im- proved by using chlorine to treat the surface of the p-type GaN layer. The chlorine was produced from electrolyzing diluted HCl

共aq兲

. The chlorine reacted with the p-type GaN surface and induced Ga vacancies in the surface region. The specific contact resistance of 4.8 ⫻ 10

−6

⍀ cm

2

was obtained for Ni/Au metals contact with the chlorine-treated p-type GaN due to the creation of more hole carriers via the inducement of Ga vacancies. Compared with the untreated LEDs, the current-voltage 共I-V兲 charac- teristics showed that the forward voltage of the chlorine-treated MQW InGaN/GaN LEDs decreased from 3.3 to 3.0 V at a driving current of 20 mA, and the light output power increases 1.25 times at 300 mA. The reverse leakage current of the chlorine-treated MQW InGaN/GaN LEDs was also significantly decreased due to the passivation of surface states by chlorination treatment of p-type GaN layer.

© 2007 The Electrochemical Society. 关DOI: 10.1149/1.2716314兴 All rights reserved.

Manuscript submitted December 6, 2006; revised manuscript received January 18, 2007. Available electronically March 22, 2007.

Recently, gallium nitride 共GaN兲-based compound semiconduc- tors have attracted the most attention for application on visible-to- ultraviolet light emitters and detectors,

1-4

and high-power electronic devices.

5,6

For those devices, high-quality and reliable metal- semiconductor contacts significantly affect their performance. For n-type GaN-based compound semiconductors, excellent ohmic and Schottky contacts have already been obtained by using suitable metal and surface treatment.

7-9

However, the ohmic contact for p-type GaN is still a challenge due to the difficulty in growing heavily doped p-GaN. In previous reports, additional Ga vacancies generated with the surface treatment have proved helpful to improve the performance of p-type GaN ohmic contact.

10-13

To obtain good ohmic contact of p-type GaN and improve the devices perfor- mances, we present a chlorination treatment of the p-type GaN layer in this work. The chlorine was produced from electrolyzing diluted HCl

共aq兲

. The chlorine reacted with the p-type GaN surface and in- duced Ga vacancies in the surface region. X-ray photoelectron spec- troscopy 共XPS兲 was used to analyze the surface of the chlorine- treated p-type GaN. The chlorination treatment improves the electrical and optical performances of the multiple-quantum-well 共MQW兲 InGaN/GaN light-emitting diodes 共LEDs兲.

Experimental

The epitaxial layers utilized were grown on c-plane sapphire sub- strates using a metallorganic chemical vapor deposition 共MOCVD兲 system. The epitaxial layer structure of the LEDs consists of a 50 nm thick undoped GaN nucleation layer, a 2 ␮m thick undoped GaN buffer layer, a 4 ␮m thick Si-doped GaN layer 共n = 3

⫻ 10

18

cm

−3

兲, an undoped InGaN/GaN MQW active layer, a 50 nm thick Mg-doped Al

0.1

Ga

0.9

N layer 共p = 1 ⫻ 10

17

cm

−3

兲, and a 300 nm thick Mg-doped GaN layer 共p = 5 ⫻ 10

17

cm

−3

兲. The InGaN/GaN MQW active layer was constructed by 10 periods of 3 nm thick In

0.23

Ga

0.77

N well and 7 nm thick GaN barrier. The as- grown samples were annealed for the activation of generating holes at 750°C for 30 min in a N

2

ambient. Using Ni/Au 共50:600 nm兲 as the metal mask, the reactive ion etching 共RIE兲 system was employed to etch through the p-type GaN layer down to the n-type GaN using BCl

3

gas. Following the removal of the metal mask, the Ti/Al/Pt/Au 共25:10:50:150 nm兲 n-type ohmic metals were deposited on the Si- doped GaN layer using an electron-beam evaporator. After thermal annealing at 850°C for 2 min in a N

2

ambient, the samples were then divided into sample A and sample B. Prior to the deposition of the Ni/Au 共20:1000 nm兲 ohmic metals on the p-type GaN layer,

sample B was processed with chlorination treatment. Figure 1 shows the chlorination treatment system. The ohmic contact area of sample B was placed underneath the Pt anodic electrode and a voltage of 20 V was applied on the Pt electrode for 60 min. Dilute HCl 共1 HCl + 10 deionized water兲 was used as the electrolytic solution.

The chlorine used for chlorination treatment was produced by elec- trolyzing dilute HCl

共aq兲

at the Pt anodic electrode. The produced chlorine was adhered and reacted with the p-type GaN surface. The Ga dangling bonds of the Ga-face p-type GaN surface grown by MOCVD system reacted with chlorine to form GaCl

x

. The GaCl

x

can easily be dissolved in the chemical solvent.

14

Therefore, Ga vacancies can be induced on the surface of the p-type GaN layer.

The other part from ohmic contact area of MQW InGaN/GaN LEDs was protected by photoresistant AZ-4620. Both sample A and B treated with and without chlorination treatment were dipped into the chemical solution of aqua regia to remove the GaO

x

layer. Using an electron-beam evaporator, a thin transparent conductive layer of Ni/Au 共2.5:2.5 nm兲 metals was first deposited, and then the pad metal of Ni/Au 共20:100 nm兲 metals was deposited on the ohmic regions of the p-type GaN layer using the lift-off technique. The samples treated with and without chlorination treatment were ther- mally annealed to form ohmic contact at 500°C for 10 min in air ambient.

Results and Discussion

To investigate the function of the chlorination treatment, XPS was used to analyze the chlorine-treated p-type GaN surface. Ac- cording to the XPS measurement, the ratio of Ga/N and Ga/O as a function of depth for the chlorine-treated p-type GaN relative to that without chlorination treatment is indicated in Table I. From the XPS measurement results, the depth of the chlorination treatment can

z

E-mail: [email protected] Figure 1. Chlorination treatment system.

Electrochemical and Solid-State Letters, 10 共6兲 H165-H167 共2007兲

1099-0062/2007/10共6兲/H165/3/$20.00 © The Electrochemical Society H165

Downloaded 23 May 2010 to 140.116.208.53. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp

(2)

thus be deduced to be about 1 nm. The relative ratio of Ga/N for chlorine-treated samples was half that of the samples without chlo- rination treatment. The produced chlorine was adhered to the p-type GaN sample and reacted with the Ga dangling bonds to form GaCl

x

, and then dissolved in the dilute HCl

共aq兲

chemical solution, which in turn induced Ga vacancies on the p-type GaN surface. Therefore, the hole concentration on the p-type GaN surface increased due to the inducement of Ga vacancies on the surface. The relative ratio of Ga/O for the chlorine-treated sample was 0.2 times that without chlorination treatment. The chlorine produced by electrolyzing HCl solution would dissolve in water to form HClO. The generated HClO in turn can oxidize p-type GaN to form GaO

x

. A detailed study of the chlorine-treated p-type GaN was reported previously.

15

Because an interfacial insulating oxide with a thickness of 1–2 nm has a serious influence on the electrical performance of metal con- tact on GaN,

16

it is necessary to remove GaO

x

from the chlorine- treated p-type GaN surface using aqua regia. The removal of GaO

x

can also induce additional Ga vacancy.

11

In comparison with a pre- vious report, using HCl

共aq兲

chemical solution to treat p-type GaN causes the Ga/N ratio to increase to 2.7,

12

which does not corre- spond to our experimental results. Therefore, we can deduce that the effect of our surface treatment can be attributed to the function of chlorine.

Figure 2 shows the current-voltage 共I-V兲 characteristics of the MQW InGaN/GaN LEDs. The chip size of the LEDs is 300

⫻ 300 ␮m. The dc forward voltage of the LEDs with and without chlorination treatment is 3.0 and 3.3 V at the driving current of 20 mA, respectively. The associated total series resistance of the LEDs with and without chlorination treatment is 20.6 and 23.4 ⍀, respectively. The dynamic resistance as a function of dc forward current is shown in Fig. 3. In general, the total series resistance includes the contact resistance between the metal and semiconductor and the series resistance of epitaxial layers of LEDs. Because the effective thickness of the chlorine-treated surface is only 1 nm un- derneath the surface and the series resistance of epitaxial layers is almost the same for all the LEDs, the reduction of the total series resistance can be attributed to the lower contact resistance between Ni/Au and chlorine-treated p-type GaN. Using the transmission line method, the specific contact resistance of the Ni/Au 共20:100 nm兲 ohmic contact of the p-type GaN with and without chlorination treatment is 4.8 ⫻ 10

−6

and 7.2 ⫻ 10

−4

⍀ cm

2

, respectively. There-

fore, the lower contact resistance of the chlorine-treated MQW InGaN/GaN LEDs can be attributed to the formation of better ohmic performances due to the creation of more hole carriers via induce- ment of Ga vacancies.

Figure 4 shows the leakage-current characteristics of the MQW InGaN/GaN LEDs. The reverse current of the LEDs with and with- out chlorination treatment is 2.5 and 40 ␮A, respectively, at a re- verse bias of −10 V. It is well known that the leakage current results recombination centers of surface states. In a previous report,

17

N vacancies at the p-type GaN surface create an n

+

region, which in turn enhances the electron tunneling under reverse bias. The electron tunneling via N vacancies at the p-type GaN surface would induce the increase of leakage current. By using chlorination treatment, the passivation function of the N vacancies can reduce the N vacancies and the surface-state density. Therefore, the leakage current of the chlorine-treated LEDs under reverse bias can be reduced.

For pulse current of pulse width of 100 ␮s and repetition fre- quency of 1 KHz, the light output power-current 共L-I兲 characteristic curves of the MQW InGaN/GaN LEDs with and without chlorina- tion treatment is shown in Fig. 5. The relative light-output power of the chlorine-treated MQW InGaN/GaN LEDs is 1.25 times higher than that without chlorination treatment for the driving current of 300 mA. For the same driving current, the voltage drop at the ohmic contact layer of the chlorine-treated MQW InGaN/GaN LEDs is lower than that without chlorination treatment, which results in a lower operating temperature in LEDs. The lower contact resistance can reduce the Joule heat created at the interface of metal and semi- conductor contact. Besides, the heating effect gives rise to carrier leakage from the InGaN/GaN MQW active region, especially under high current operation.

18

The nonradiative recombination rate would be increased in LEDs operated at a higher temperature. By using the wavelength-shift method,

19

the junction temperature of the LEDs with and without chlorination treatment under a pulse current of Table I. Ga/N and Ga/O ratios of chlorine-treated p-GaN, com-

pared with the sample without chlorination treatment. (The ratio of Ga/N and Ga/O of as-grown GaN film was set to be 1.)

Depth 0 nm 共surface兲 1 nm 3 nm 5 nm

Ga/N ratio 0.5 0.97 1 1

Ga/O ratio 0.2 0.66 1 1

Figure 2. Current-voltage characteristics of the MQW InGaN/GaN LEDs with and without chlorination treatment.

Figure 3. Dynamic resistance as a function of dc forward current for LEDs with and without chlorination treatment.

Figure 4. Leakage current characteristics of the MQW InGaN/GaN LEDs with and without chlorination treatment.

H166 Electrochemical and Solid-State Letters, 10 共6兲 H165-H167 共2007兲 H166

Downloaded 23 May 2010 to 140.116.208.53. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp

(3)

300 mA can be estimated as 130 and 170°C, respectively. Therefore, more light output power can be obtained in the chlorine-treated LEDs. As shown in Fig. 5, the maximum driving current of the MQW InGaN/GaN LEDs with and without chlorination surface treatment is 450 and 340 mA, respectively. In addition to the de- crease of contact resistance, the increase of maximum driving cur- rent can also be attributed to the decrease of the barrier between Ni/Au metals and chlorine-treated p-GaN, which would allow more holes to inject into the MQW active layer.

Conclusions

In summary, the function and mechanism of the optical and elec- trical performances on the chlorine-treated MQW InGaN/GaN LEDs have been investigated. The chlorinated surface treatment can enhance the formation of Ga vacancies and reduce N vacancies and related surface states, thus decreasing the resistance of ohmic con- tact and reducing probability of electron tunneling under reverse

bias. Therefore, a higher light-output power of the chlorine-treated MQW InGaN/GaN LEDs was obtained, and the reverse leakage current was also decreased due to the passivation of the N vacancies and related surface states by chlorinated surface treatment.

Acknowledgment

This work was supported by the National Science Council of the Republic of China under contract no. NSC 95-2221-E006-315.

National Cheng Kung University assisted in meeting the publication costs of this article.

References

1. J. Piprek, R. Farrell, S. DenBaars, and S. Nakamura, IEEE Photon. Technol. Lett., 18, 7 共2006兲.

2. C. T. Lee, U. Z. Yang, C. S. Lee, and P. S. Chen, IEEE Photon. Technol. Lett., 18, 2029 共2006兲.

3. J. K. Kim, T. Gessmann, E. F. Schubert, J. Q. Xi, H. Luo, J. Cho, C. Sone, and Y.

Park, Appl. Phys. Lett., 88, 013501 共2006兲.

4. G. Ariyawansa, Appl. Phys. Lett., 89, 091113 共2006兲.

5. S. J. Hong and K. Kim, Appl. Phys. Lett., 89, 042101 共2006兲.

6. S. Haffouz, H. Tang, J. A. Bardwell, S. Rolfe, E. M. Hsu, I. Sproule, S. Moisa, M.

Beaulieu, and J. B. Webb, J. Vac. Sci. Technol. B, 23, 1199 共2005兲.

7. C. T. Lee and H. W. Kao, Appl. Phys. Lett., 76, 2364 共2000兲.

8. D. F. Wang, F. Shiwei, C. Lu, A. Motayed, M. Jah, S. N. Mohammad, K. A. Jones, and L. Salamanca-Riba, J. Appl. Phys., 89, 6214 共2001兲.

9. C. T. Lee, Y. J. Lin, and D. S. Liu, Appl. Phys. Lett., 79, 2573 共2001兲.

10. J. K. Kim, J. L. Lee, J. W. Lee, Y. J. Park, and T. Kim, J. Vac. Sci. Technol. B, 17, 497 共1999兲.

11. C. S. Lee, Y. J. Lin, and C. T. Lee, Appl. Phys. Lett., 79, 3815 共2001兲.

12. J. Sun, K. A. Rickert, J. M. Redwing, A. B. Ellis, F. J. Himpsel, and T. F. Kuech, Appl. Phys. Lett., 76, 415 共2000兲.

13. C. Huh, S. W. Kim, H. S. Kim, H. M. Kim, H. Hwang, and S. J. Park, Appl. Phys.

Lett., 78, 1766 共2001兲.

14. CRC Handbook of Chemistry and Physics, D. R. Lide and H. P. R. Frederikse, Editors, 75th ed., CRC Press, Boca Raton, FL 共1995兲.

15. P. S. Chen and C. T. Lee, J. Appl. Phys., 100, 044510 共2006兲.

16. X. A. Cao, S. J. Pearton, G. Dang, A. P. Zhang, F. Ren, and J. M. V. Hove, Appl.

Phys. Lett., 75, 4130 共1999兲.

17. S. J. Pearton, J. C. Zolper, R. J. Shul, and F. Ren, J. Appl. Phys., 86, 1 共1999兲.

18. J. K. Sheu, J. M. Tsai, S. C. Shei, W. C. Lai, T. C. Wen, C. H. Kou, Y. J. Su, S. J.

Chang, and G. C. Chi, IEEE Electron Device Lett., 22, 460 共2001兲.

19. J. Cho, C. Sone, Y. Park, and E. Yoon, Phys. Status Solidi A, 202, 1869 共2005兲.

Figure 5. Light output power-current characteristics of the MQW InGaN/GaN LEDs with and without chlorination treatment.

H167 Electrochemical and Solid-State Letters, 10 共6兲 H165-H167 共2007兲 H167

Downloaded 23 May 2010 to 140.116.208.53. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp

數據

Figure 2. Current-voltage characteristics of the MQW InGaN/GaN LEDs with and without chlorination treatment.
Figure 5. Light output power-current characteristics of the MQW InGaN/GaN LEDs with and without chlorination treatment.

參考文獻

相關文件

You are given the wavelength and total energy of a light pulse and asked to find the number of photons it

Wang, Solving pseudomonotone variational inequalities and pseudocon- vex optimization problems using the projection neural network, IEEE Transactions on Neural Networks 17

volume suppressed mass: (TeV) 2 /M P ∼ 10 −4 eV → mm range can be experimentally tested for any number of extra dimensions - Light U(1) gauge bosons: no derivative couplings. =>

Define instead the imaginary.. potential, magnetic field, lattice…) Dirac-BdG Hamiltonian:. with small, and matrix

incapable to extract any quantities from QCD, nor to tackle the most interesting physics, namely, the spontaneously chiral symmetry breaking and the color confinement.. 

• Formation of massive primordial stars as origin of objects in the early universe. • Supernova explosions might be visible to the most

Therefore, every Buddha’s Light member who vows to practice the bodhisattva path needs to cultivate bodhi wisdom and the power of vows in order to change the world and benefit

(Another example of close harmony is the four-bar unaccompanied vocal introduction to “Paperback Writer”, a somewhat later Beatles song.) Overall, Lennon’s and McCartney’s