Impacts of ammonia background flows on structural and photoluminescence properties of InN dots grown on GaN by flow-rate modulation epitaxy

(1)

Impacts of ammonia background flows on structural and photoluminescence

properties of InN dots grown on GaN by flow-rate modulation epitaxy

W. C. Ke, L. Lee, C. Y. Chen, W. C. Tsai, W.-H. Chang, W. C. Chou, M. C. Lee, W. K. Chen, W. J. Lin, and Y. C. Cheng

Citation: Applied Physics Letters 89, 263117 (2006); doi: 10.1063/1.2425038

View online: http://dx.doi.org/10.1063/1.2425038

View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/89/26?ver=pdfcov Published by the AIP Publishing

Articles you may be interested in

Non-polar (11-20) InGaN quantum dots with short exciton lifetimes grown by metal-organic vapor phase epitaxy Appl. Phys. Lett. 102, 251905 (2013); 10.1063/1.4812345

Compositional instability in InAlN/GaN lattice-matched epitaxy Appl. Phys. Lett. 100, 092101 (2012); 10.1063/1.3690890

Hole carrier concentration and photoluminescence in magnesium doped InGaN and GaN grown on sapphire and GaN misoriented substrates

J. Appl. Phys. 108, 023516 (2010); 10.1063/1.3466768

Nucleation of InN quantum dots on GaN by metalorganic vapor phase epitaxy Appl. Phys. Lett. 87, 263104 (2005); 10.1063/1.2152110

Structural properties of InN on GaN grown by metalorganic vapor-phase epitaxy J. Appl. Phys. 85, 7682 (1999); 10.1063/1.370571

This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP: 140.113.38.11 On: Thu, 01 May 2014 01:40:03

(2)

Impacts of ammonia background flows on structural

and photoluminescence properties of InN dots grown on GaN

by flow-rate modulation epitaxy

W. C. Ke, L. Lee, C. Y. Chen, W. C. Tsai, W.-H. Chang,a兲,b兲 W. C. Chou, M. C. Lee, and W. K. Chena兲,c兲

Department of Electrophysics, National Chiao Tung University, Hsinchu 300, Taiwan

W. J. Lin and Y. C. Cheng

Chung-Shan Institute of Science and Technology, Tao-Yuan 325, Taiwan

共Received 29 August 2006; accepted 26 November 2006; published online 29 December 2006兲 Structural and photoluminescence 共PL兲 properties of InN dots grown on GaN by metal organic vapor phase epitaxy using the flow-rate modulation technique, and their dependence on growth conditions, were investigated. An ammonia 共NH3兲 background flow was intentionally supplied

during indium deposition periods to control the kinetics of adatoms and hence the morphology of InN dots. Samples prepared under lower NH3 background flows generally exhibit narrower and

more intense PL signals peaked at lower emission energies. The authors point out that the NH3

Indium nitride共InN兲 has attracted much attention in re-cent years due not only to its superior electronic transport properties,1 such as high theoretical mobility and large drift velocity, but also to the recent revision of its band gap energy to ⬃0.7 eV.2–4 This discovery further spans the spectral range of III-nitride materials into near infrared, covering the wavelength range for telecommunications applications. High crystalline quality InN films have been obtained by a number of growth techniques in the past few years, including mo-lecular beam epitaxy共MBE兲共Ref.5兲 and metal organic

va-por phase epitaxy 共MOVPE兲.6 Recent progress has also shown that nanometer-scale InN dots with controllable size and density can be formed on GaN surface during the initial stage of heteroepitaxial growth using MBE 共Refs. 7–9兲 or

MOVPE.10,11 After proper capping with GaN, this kind of InN dots shows superior photoluminescence共PL兲 properties and pronounced quantum confined effects.11 However, the growth of high-quality InN dots with tailored optical proper-ties to act as “quantum dots” for practical applications is still of great challenge. The difficulty is believed to arise from the low growth temperature for InN due to the low InN disso-ciation temperature and the high equilibrium N2vapor

pres-sure over the InN film.1This peculiar feature not only inher-ently restricts the temperature for growing capping layers 共e.g., GaN or AlN兲 but also hinders the surface migrations of adatoms. A low growth temperature implies a short migration length, which means that In adatoms are less mobile to find an energetically favored site. A number of studies with re-spect to nucleation processes and structural properties of ei-ther MBE—or MOVPE-grown InN dots have been reported.12,13 However, it remains less clear how the optical properties of InN dots relate to their structural properties and nucleation processes.

In MOVPE growth, a well-known technique that can achieve a lower growth temperature while keeping high crys-talline quality is the flow-rate modulation epitaxy共FME兲.14

Due to the alternative supply of group-III and group-V sources in FME growth, surface migrations can be consider-ably enhanced, which is believed to be very useful for the fabrication of nanostructures. In this letter, structural and PL properties of FME-grown InN dots on GaN, and their depen-dence on growth conditions, were investigated. During the FME growth of InN dots, an NH₃background flow was in-tentionally supplied in addition to the alternating precursors. By tuning the NH3background flow, it is possible to control

the kinetics of In adatoms and hence the morphology of InN dots. PL investigations further reveal the information about intrinsic properties of InN dots prepared under different growth conditions. We point out that the NH₃ background flow is an important parameter that controls not only the nucleation process but also the emission property of InN dots.

Samples in this study were grown on sapphire 共0001兲 substrates in a MOVPE system using trimethylgallium 共TMGa兲, trimethylindium 共TMIn兲, and ammonia 共NH3兲 as

source precursors. After nitridation of the substrate at 1120 ° C, a thin GaN nucleation layer was first grown at 520 ° C, followed by the growth of a 1-␮m-thick undoped GaN buffer layer at 1120 ° C. After the growth of GaN buffer layers, the substrate temperature was then reduced to 600 ° C to grow InN dots. The gas-flow sequence of FME consists of four steps in one cycle: a 20 s TMIn flow period, a 20 s NH3

flow period, and two 10 s purge periods intervened in be-tween. The TMIn and NH3 flow rates are 150 and

18 000 SCCM共SCCM denotes cubic centimeter per minute at STP兲, respectively. During TMIn flow periods, an NH₃ background flow was intentionally supplied throughout the growth. A series of samples using different NH3background

flow rates r₀, ranging from 0 to 10 000 SCCM, has been grown. The growth of InN dots was completed by a total of six cycles for all samples. For comparison purpose, we have also prepared a sample of InN dots by the conventional MOVPE growth mode with continuous flows of 10 000 SCCM NH3 and 150 SCCM TMIn for 120 s at the

same growth temperature of 600 ° C. Surface morphologies,

a兲_{Authors to whom correspondence should be addressed.} b兲_{Electronic mail: whchang@mail.nctu.edu.tw}

c兲_{Electronic mail: wkchen@cc.nctu.edu.tw}

APPLIED PHYSICS LETTERS 89, 263117共2006兲

(3)

sizes, and densities of InN dots were analyzed using atomic force microscopy共AFM兲. PL measurements were performed at 10 K using the 488 nm line of an Ar+laser as an excita-tion source. The luminescence signals were dispersed by a 0.5 m monochomator and detected by a liquid-nitrogen-cooled extended InGaAs detector共with a cutoff wavelength at 2.05␮m兲 using the standard lock-in technique.

Figure1shows AFM micrographs of InN dots grown by FME with different NH3background flow rates共r0兲 ranging

from 0 to 10 000 SCCM 关Figs. 1共a兲–1共e兲兴 and by the con-ventional growth mode关Fig.1共f兲兴. The dot density and aver-age diameter as function of r0are plotted in Fig.1共g兲. As r0

was increased from 0 to 10 000 SCCM, the dot density in-creased from 6.2⫻108_{to 1.0⫻10}10 _cm−2_{, whereas the}

aver-age base diameter decreased from 350 to 95 nm. Comparing the dot density and size to those obtained by conventional growth mode关indicated by arrows in Fig.1共g兲兴, FME growth under low NH3background flows共r0⬍1000 SCCM兲

gener-ally yields much larger and less dense dots. This can be attributed to the enhanced surface migration of In adatoms in low NH3 ambience during the deposition of In species.15,16

As more NH₃are injected during TMIn periods, surface mi-gration of In adatoms is considerably hindered due to the exposure to a highly N-rich ambience, leading to higher dot densities and smaller dot sizes.

Apart from the dot size and density, their shape also changes with the NH3background flow. Figure2 gives line

profiles across typical InN dots grown by FME with different

r0 关Figs. 2共a兲–2共e兲兴 and by the conventional method 关Fig.

2共f兲兴. In order to represent the typical dot size in each case, we have selected the dot having based diameter and height very close to their average values. In general, FME-grown InN dots under low NH₃ background flows reveal a mesa shape, or more precisely, a truncated hexagonal pyramid with a flattop and faceted sidewalls. The aspect ratio共i.e., height-to-diameter ratio兲 for the case of r0= 0 is found to be ⬃1/8.

Such a low aspect ratio indicates that the lateral growth rate is at least a few times共⬃8兲 faster than the vertical one. With the increasing NH₃ background flow, the mesa-type dots gradually evolved to lens-shaped dots without clear faceting.

Furthermore, the aspect ratio also increases remarkably to ⬃1/3 as the r0 reaches 10 000 SCCM. As for InN dots

grown by conventional MOVPE, the dot shape is also lens-shape-like, with an aspect ratio of about⬃1/5.

The above observations clearly demonstrate the impact of the NH3background flow on the kinetics of adatoms

dur-ing the nucleation of InN dots. Indeed, reducdur-ing the NH3

background flow during In deposition periods not only leads to a longer migration length of In adatoms but also prevents the In adatoms from moving uphill to island tops, due to the preferential formation of In-terminated surface on the InN island tops under such In-rich conditions. As a consequence, In adatoms tend to nucleate at the edge of InN dots, yielding a faster lateral growth rate than the vertical one and hence mesa-type islands with lower aspect ratios. On the contrary, when a high NH3background flow was used, the exposure to

a highly N-rich ambience not only hinders the surface migra-tion of In adatoms, but also tends to form N-stabilized sur-faces on islands tops. This facilitates uphill transfer of In adatoms to island tops, due possibly to the lower chemical potential thereon, leading lens-shaped dots with a signifi-cantly higher aspect ratio.

The effect of the NH3background flow on PL properties

of InN dots is shown in Fig.3. The PL spectra shown in Fig.

3共a兲were measured at 10 K under an excitation power den-sity of ⬃0.15 kW/cm2_{. All samples exhibit near-infrared}

emission bands in the range of 0.7– 0.9 eV. Since the dot sizes are still too large to produce pronounced quantum size effects, differences in peak energy between samples are mainly due to variations of the electron concentration in InN dots. With the decreasing NH3background flow, a redshift of

peak energy can be clearly seen. This implies that the NH₃ background flow is an important parameter for controlling the electron concentration of InN dots. To give a quantitative estimation, we have further employed a line shape model considering “free-to-bond” radiative recombination to ana-lyze the PL spectra.17 Since our main interest in modeling aims at finding the Fermi energy共Ef兲 and the relative

elec-FIG. 1.共Color online兲 AFM micrographs of InN dots grown by FME with different r0:共a兲 0 SCCM, 共b兲 500 SCCM, 共c兲 1000 SCCM, 共d兲 5000 SCCM,

and共e兲 10000 SCCM, and 共f兲 by the conventional growth mode. 共g兲 The density and diameter of the InN dots as function of r0.

FIG. 2.共Color online兲 Line profiles across typical InN dots grown by FME with different r₀: 共a兲 0 SCCM, 共b兲 500 SCCM, 共c兲 1000 SCCM, 共d兲 5000 SCCM, and共e兲 10000 SCCM, and 共f兲 by the conventional method. The AFM micrographs shown on the right are the corresponding morpholo-gies in an area of 500⫻500 nm2_.

263117-2 Ke et al. Appl. Phys. Lett. 89, 263117共2006兲

(4)

tron concentration 共ne兲 in different samples, the model was

further simplified by assuming a deltalike function for the distribution of photogenerated holes near the valence band edge, so that the shape of emission band can be approxi-mated by the electron energy distribution in the conduction band. As illustrated in Fig.3共b兲, the simplified model repro-duces our PL spectra very well. The fitted band gap values are in the range of 0.69– 0.72 eV, in good agreement with the value reported in literature.17In Fig.3共c兲, the deduced Ef

was lowered from 179 to 74 meV as r₀was decreased from 10 000 to 500 SCCM. If the electron effective mass was chosen to be 0.07m0,17 the change in Ef corresponds to a

reduction of ne from 8.6⫻1018to 2.0⫻1018cm−3. This

seems to imply that more donor-type impurities, most likely the hydrogen, may be incorporated under a high NH₃ back-ground flow. On the other hand, we noted that the emission property of InN dots is getting worse when further changing the NH3 background flow from 500 to 0 SCCM. This fact

suggests that another mechanism governs the source of elec-tron concentration in InN dots grown under such an NH3-free condition for the deposition of In species. For the

growth of InN at 600 ° C, the decomposition of InN is sig-nificant, while the desorption of In is still negligible.18If the amount of active nitrogen is insufficient, N atoms are ex-pected to desorb rapidly from InN during the TMIn flow periods, probably in the form of N2 molecules, resulting in

the accumulation of metallic In on the surface. This may facilitate the formation of nitrogen vacancies and/or metallic In droplets, leading to degraded optical quality.

Besides, we also found a general correlation between the PL intensity共IPL兲 and the deduced carrier concentration ne.

In Fig.3共d兲, we can see that the IPL decreases with the

de-duced ne as a power function ⬃共ne兲−1.9. This is a stronger

dependence than for a dominant Auger recombination ⬃共ne兲−1.

19

Therefore, we conclude that the PL efficiency

in-creases not only due to the reduced electron concentration but also to the reduced nonradiative centers with the decreas-ing n-type background. The maximum PL intensity occurred at r0= 500 SCCM, which is about ⬃40 times higher that at

10 000 SCCM and is still ⬃ eight times higher than those prepared by conventional MOVPE growth mode. Such an improvement in PL efficiency provides a new approach to the fabrication of high optical quality InN dots for practical applications.

In summary, structural and PL properties of FME-grown InN dots grown on GaN and their dependence on growth conditions were investigated. By tuning the NH₃background flow during deposition periods of In species, it is possible to control the kinetics of adatoms and hence the morphology of InN dots. Samples prepared under lower NH3 background

flows generally exhibit narrower and more intense PL signals at lower emission energies. We point out that the amount of NH₃background is very important not only to the morphol-ogy but also to the emission property of InN dots.

This work is supported in part by the project of MOE-ATU and the National Science Council of Taiwan under Grant Nos. NSC 047, NSC 009-012, NSC 009-044-MY3, and NSC 95-2112-M-009-020.

1_{For a review, see for example, A. G. Bhuiyan, A. Hashimoto, and A.}

Yamamoto, J. Appl. Phys. 94, 2779共2003兲.

2_{V. Yu. Davydov, A. A. Klochikhin, R. P. Seisyan, V. V. Emtsev, S. V.}

Ivanov, F. Bechstedt, J. Furthmüller, H. Harima, A. V. Mudryi, J. Aderhold, O. Semchinova, and J. Graul, Phys. Status Solidi B 229, R1 共2002兲.

3_{J. Wu, W. Walukiewicz, K. M. Yu, J. W. Ager III, E. E. Haller, H. Lu, W.}

J. Schaff, Y. Saito, and Y. Nanishi, Appl. Phys. Lett. 80, 3967共2002兲.

4_{T. Matsuoka, H. Okamoto, M. Nakao, H. Harima, and E. Kurimoto, Appl.}

Phys. Lett. 81, 1246共2002兲.

5_{H. Lu, W. J. Schaff, J. Hwang, H. Wu, G. Koley, and L. F. Eastman, Appl.}

Phys. Lett. 79, 1489共2001兲.

6_{A. Yamamoto, T. Tanaka, K. Koide, and A. Hashimoto, Phys. Status Solidi}

A 194, 510共2002兲.

7_{C. Nörenberg, R. A. Oliver, M. G. Martin, L. Allers, M. R. Castell, and G.}

A. D. Briggs, Phys. Status Solidi A 194, 536共2002兲.

8_{Y. F. Ng, Y. G. Cao, M. H. Xie, X. L. Wang, and S. Y. Tong, Appl. Phys.}

Lett. 81, 3960共2002兲.

9_{A. Yoshikawa, N. Hashimoto, N. Kikukawa, S. B. Che, and Y. Ishitani,}

Appl. Phys. Lett. 86, 153115共2005兲.

10_{O. Briot, B. Maleyre, and S. Ruffenach, Appl. Phys. Lett. 83, 2919}

共2003兲.

11_{W. C. Ke, C. P. Fu, C. Y. Chen, L. Lee, C. S. Ku, W. C. Chou, W.-H.}

Chang, M. C. Lee, W. K. Chen, W. J. Lin, and Y. C. Cheng, Appl. Phys. Lett. 88, 191913共2006兲.

12_{L. Zhou, T. Xu, D. J. Smith, and T. D. Moustakas, Appl. Phys. Lett. 88,}

231906共2006兲.

13_{J. G. Lozano, A. M. Sánchez, R. García, D. González, D. Araújo, S.}

Ruffenach, and O. Briot, Appl. Phys. Lett. 87, 263104共2005兲.

14_{N. Kobayashi, T. Makimoto, and Y. Horikoshi, Jpn. J. Appl. Phys., Part 2}

24, L962共1985兲; Y. Horikoshi, J. Cryst. Growth 201/202, 150 共1999兲.

15_{H. Lu, W. J. Schaff, J. Hwang, H. Wu, W. Yeo, A. Pharkya, and L. F.}

Eastman, Appl. Phys. Lett. 77, 2548共2000兲.

16_{R. S. Qhalid Fareed, R. Jain, R. Gaska, M. S. Shur, J. Wu, W.}

Waluk-iewicz, and M. Asif Khan, Appl. Phys. Lett. 84, 1892共2004兲.

17_{B. Arnaudov, T. Paskova, P. P. Paskov, B. Magnusson, E. Valcheva, B.}

Monemar, H. Lu, W. J. Schaff, H. Amano, and I. Akasaki, Phys. Rev. B

69, 115216共2004兲.

18_{E. Dimakis, E. Iliopoulos, K. Tsagaraki, Th. Kehagias, Ph. Komninou, and}

A. Georgakilas, J. Appl. Phys. 97, 113520共2005兲.

19_{T. Stoica, R. J. Meijers, R. Calarco, T. Richter, E. Sutter, and H. Lüth,}

Nano Lett. 6, 1541共2006兲. FIG. 3. 共Color online兲共a兲 PL spectra measured at 10 K under the same

excitation condition.共b兲 The measured PL spectra and the calculated line shapes for samples with r0= 500 and 10000 SCCM.共c兲 The deduced Fermi

energy relative to the conduction band edge as a function of r0.共d兲

Corre-lation of the estimated electron concentration newith the integrated PL

intensity IPL.

263117-3 Ke et al. Appl. Phys. Lett. 89, 263117共2006兲